Inorganic Reactions and Methods- Reactions Catalyzed by Inorganic Compounds

Inorganic Reactions and Methods Volume 16 Inorganic Reactions and Methods Editor Arlan D. Norman Department of Chemist...

Author: J. J. Zuckerman | A. P. Hagen

83 downloads 1321 Views 37MB 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

Inorganic Reactions and Methods Volume 16

Inorganic Reactions and Methods Editor Arlan D. Norman Department of Chemistry University of Colorado Boulder, CO 80309-0215 Edltorlal Advisory Board Professor N. Bartlett Department of Chemistry University of California at Berkeley Berkeley, California 94720 Professor F.A. Cotton Department of Chemistry Texas A&M University College Station, Texas 77840 Professor E.O. Fischer Anorganisch-chemisches Laboratorium der Technischen Universitat D-8046 Garching Lichtenbergestrasse4 Federal Republic of Germany Professor P. Hagenmuller Laboratoire de Chemie du Solide du C.N.R.S. 351 cours de la Liberation F-33405 Talence France Professor M.F. Lappert The Chemical Laboratory University of Sussex Falmer, Brighton, BN1 9A3 England

Professor A.G. MacDiarmid Department of Chemistry University of Pennsylvania Philadelphia, Pennsylvania 19174 Professor M. Schmidt lnstitut fur Anorganische Chemie der Universitat D-8700 Wurzburg Am Hubland Federal Republic of Germany Professor H. Taube Department of Chemistry Stanford University Stanford, California 94305 Professor L.M. Venanzi Laboratoriumfur Anorganische Chemie der ETH CH-80006 Zurich Universitatsstrasse5 Switzerland

01991 VCH Publishers, Inc., New York Distribution: VCH Verlagsgeseilschafl mbH, P.O. Box 1260/1280, 0-6940 Weinheim, Federal Republic of Germany USA and Canada: VCH Publishers, Inc., 303 N.W. 12th Avenue, Deerfield Beach, FL 33442-1705, USA

Inorganic Reactions and Methods Volume 16 Reactions Catalyzed by Inorganic Compounds Founding Editor

J.J. Zuckerman Editor

Arlan D. Norman

@ WILEY-VCH

Library of Congress Cataloging in Publication Data Inorganic reactions and methods Includes bibliographiesand indexes. Contents: v. 1. The formation of bonds to hydrogenpt. 2, v. 2. The formation of the bond to hydrogen-v. 16. Reactionscatalyzed by inorganic compounds. 1. Chemical reaction, Conditions and laws ofCollected works. 2. Chemistry, Inorganic-SynthesisCollected works. I. Zuckerman, Jerry J. QD501.1623 1987 541.39 85-15627 ISBN 0-895730-250-5(set)

61993 VCH Publishers, Inc, This work is subject to copyright, All rights are reserved, whether the wholeor part of the materialis concerned, specifically those of translation, reprinting,re-useof illustrations,broadcasting, reproductionby photocopying machine or similar means, and storage in data banks. Authorizationto photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by VCH Publishers, Inc. for libraries and other users registeredwith the Copyright Clearance Center (CCC) TransactionalReportingService, provided that the base fee of $?.a0 per copy, plus $0.25 per page is paid directly to CCC, 27 Congress Street, Salem, MA 01970. Registerednames, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotectedby law.

ISBN 0-471-18666-X VCH publishers

ISBN 3-527-26274-1VCH Verlagsgesellschaft

Contents of Volume 16 How to Use this book Preface to the Series Editorial Consultants to the Series Contributors to Volume 16

xiii xix xxiii xxv

14.

Reactions Catalyzed by Inorganic Compounds

1

14.1.

Introduction: Principles of Catalysis

2

14.1.1. 14.1.2. 14.1.2.1. 14.1.2.2. 14.1.2.2.1. 14.1.2.2.2. 14.1.2.3. 14.1.2.4. 14.1.2.5. 14.1.2.5.1. 14.1.2.5.2. 14.1.2.5.3. 14.1.2.6.

14.2. 14.2.1. 14.2.2. 14.2.2.1. 14.2.2.1.1. 14.2.2.1.2.

Catalysis as a Kinetic Phenomenon Basic Processes in Molecular Catalysis Electron Transfer Ligand Dissociation and Association Processes Heterolytic Ligand Dissociation. Homolytic Ligand Dissociation. Promotion of Nucleophilic Reactions by Electron Withdrawal from Reactants Catalysis of Electrophilic Reactions by Proton Loss from a Coordinated Ligand Oxidative Addition/Reductive Elimination Reactions One-Electron Oxidative Addition. Two-Electron Oxidative Addition. Free Radical Chain Mechanism of Oxidative Addition. Insertion Reactions

Types of Catalysts Introduction Solid Catalysts Metallic Catalysts Metal Crystals and Films. Supported Metal Catalysts.

2 5 5

7 7 9 10 11 13 14 15 18 19

21 21 22 22 22 23 V

vi

14.2.2.2. 14.2.3. 14.2.3.1. 14.2.3.2. 14.2.4. 14.2.4.1. 14.2.4.2. 14.2.5. 14.2.6. 14.2.7. 14.2.7.1. 14.2.7.2. 14.2.7.2.1. 14.2.7.2.2. 14.2.7.2.3. 14.2.7.3. 14.2.7.4.

14.3. 14.3.1. 14.3.2. 14.3.2.1. 14.3.2.2. 14.3.2.3. 14.3.3. 14.3.3.1. 14.3.3.2. 14.3.3.3. 14.3.3.4. 14.3.3.5. 14.3.3.6. 14.3.4. 14.3.4.1. 14.3.4.1.l. 14.3.4.1.2. 14.3.4.1.3. 14.3.4.2. 14.3.4.3.

Contents of Volume 16

Metal Oxide and Metal Sulfide Catalysts Soluble Catalysts Selectivity Advantages Process Engineering and Product Recovery Problems Supported Metal Complexes Polymeric Supports Metal Oxide Supports Phase Transfer Catalysis Catalysis in Microscopic Phases Production of Catalysts and Supports General Principles Methods of Production of Nonmetal Catalysts and Supports Precipitation and Gel Formation. Impregnation. Natural Materials, Leaching, Carbon supports. Methods of Production of Supported Metal Catalysts Relationships between Catalyst Production and Performance

Hydrogenation Reactions Introduction Dihydrogen Activation Homolytic Cleavage to Give Metal-Hydrides Heterolytic Cleavage to Give Metal-Hydrides Molecular Hydrogen Complexes Classes of Soluble Catalysts Rhodium(I) Catalysts Cobalt Cyanide Systems Cobalt Carbonyl Catalysts Chromium(0) Carbonyl Catalysts Ziegler Catalysts Ruthenium(l1) Catalysts Hydrogenation of Aliphatic C-C Functions In Simple Olefins Isolated Double Bonds. Olefins Conjugated to Carbonyl, Nitrile, Nitro. Vinyl Functions. In Conjugated Dienes In Unconjugated Dienes

--

26 32 32

34 36 36 40 41 43 45 45 51 51 55 56

56 60

65 65 65 66 71 77 79 80 89 92 94 97 99 102 102 102 114 121 123 132


14.3.4.4. 14.3.4.4.1. 14.3.4.4.2. 14.3.4.5. 14.3.5. 14.3.5.1. 14.3.5.2. 14.3.5.3. 14.3.5.4. 14.3.5.5. 14.3.6. 14.3.6.1. 14.3.6.1.1. 14.3.6.1.2. 14.3.6.1.3. 14.3.6.2. 14.3.6.2.1. 14.3.6.2.2. 14.3.6.2.3. 14.3.6.2.4. 14.3.6.3. 14.3.6.4. 14.3.7. 14.3.7.1. 14.3.7.1.l. 14.3.7.1 -2. 14.3.7.1.3. 14.3.7.1.4. 14.3.7.2. 14.3.7.2.1. 14.3.7.2.2. 14.3.7.2.3. 14.3.7.3.

14.4. 14.4.1. 14.4.2. 14.4.2.1. 14.4.2.2. 14.4.2.3. 14.4.3.

In Acetylenes and Cumulenes In Triple Bonds. In Allenes and Cumulenes. By Asymmetric Hydrogenation Hydrogenation of Arenes By Cobalt Catalysts By Ruthenium Catalysts by Rhodium Catalysts By Palladium and Platinum Catalysts By Miscellaneous Catalysts Hydrogenation of C = O Functions In Aldehydes Saturated Aliphatic Aldehydes. Aromatic Aldehydes. Selectivity. In Ketones Hydrogenation to the Carbinol. Hydrogenolysis and Miscellaneous Reactions. Selectivity. Stereochemistry and Asymmetric Hydrogenation. In Carboxyl Derivatives By Transfer Hydrogenation Hydrogenation of Other Functional Groups Nitrites Hydrogenation to Primary Amines. Coupling Reactions. Reductive Hydrolysis. Hydrogenolysis and Cyclizations. Nitro Compounds Hydrogenation to the Amine. Selective and Partial Reductions. Side Reactions in Polyfunctional Molecules. Miscellaneous

Addition Reactions Introduction Hydrosilylationof Olefins and Acetylenes By Platinum Catalysts By Rhodium and Nickel Catalysts By Other Transition Metal Catalysts Hydrosilylation of Conjugated Dienes

vii 137 137 144 145 157 157 159 162 165 170 172 172 172 174 175 178 178 180 183 185 193 198 202 202 203 205 206 207 209 209 21 1 214 216

218 218 218 218 225 229 233

viii 14.4.3.1 . 14.4.3.2. 14.4.3.3. 14.4.4. 14.4.4.1. 14.4.4.2. 14.4.4.3. 14.4.5. 14.4.5.1. 14.4.5.2. 14.4.6. 14.4.6.1. 14.4.6.2. 14.4.6.3. 14.4.6.4.

14.5. 4.5.1. 4.5.1 .I. 4.5.1 .I.l. 4.5.1.1.2. 4.5.1 -2. 4.5.1.2.1. 4.5.1 -2.2. 4.5.1.2.3. 4.5.1.3. 4.5.2. 4.5.2.1. 4.5.2.2. 4.5.2.2.1. 4.5.2.2.2. 4.5.2.2.3. 4.5.2.2.4. 4.5.2.3. 4.5.2.4. 4.5.2.4.1. 4.5.2.4.2. 4.5.2.4.3.


By Platinum Catalysts By Palladium and Nickel Catalysts By Rhodium, Cobalt, and Chromium Catalysts Hydrosilylation of Carbonyl Compounds By Transition Metal Catalysts Stereoselective and Chemoselective Reactions Asymmetric Synthesis Hydrosilylation of Carbon-Nitrogen Double Bonds Of lmines Of Isocyanates, Carbodiimides, and Nitriles Hydrocyanation of Olefins and Dienes By Nickel Catalysts By Palladium Catalysts By Copper Catalysts By Cobalt Catalysts

Olefin Transformations lsomerization Allylic Hydrogen Transfer By Palladium Catalysts. By Iron Catalysts. Metal Hydride Addition-Elimination By Cobalt Carbonyl Catalysts. By Ruthenium Catalysts. By Platinum Catalysts. Skeletal Rearrangement Olefin Dimerization and Oligomerization Introduction Linear Dirnerization and Oligomerization of Monoolefins Mechanistic Aspects. By Group 4 and 5 Metal Catalysts. By Nickel Catalysts. By Other Group 8-1 0 Metal Catalysts. Cyclodimerization and Cyclooligomerization of Monoolefins Linear Dimerization and Oligomerization of 1,3-DioIefins By Palladium Catalysts. By Nickel Catalysts. By Other Transition Metal Catalysts.

234 235 239 240 24 1 243 247 255 255 257 259 260 265 266 267

269 269 269 269 270 271 27 1 274 274 275 276 276 278 278 280 282 287 289 290 290 292 293

14.5.2.5. 14.5.2.5.1. 14.5.2.5.2. 4.5.3. 4.5.3.1. 4.5.3.2. 4.5.3.2.1. 4.5.3.2.2. 4.5.3.2.3. 4.5.3.2.4. 14.5.3.2.5. 14.5.3.3. 14.5.3.3.1. 14.5.3.3.2. 14.5.3.4. 14.5.3.4.1. 14.5.3.4.2. 14.5.3.4.3. 14.5.4.

14.6. 14.6.1. 14.6.1.1. 14.6.1.2. 14.6.1.3. 14.6.1.4. 14.6.1.5. 14.6.1.6. 14.6.1.7. 14.6.1-8. 14.6.1-9. 14.6.2. 14.6.2.1. 14.6.2.1.l. 14.6.2.1.2.

Contents of Volume 16 Cyclodimerization and Cyclooligomerization of 1,3-DioIefins By Nickel Catalysts. By Other Transition Metal Catalysts. Olefin Polymerization Introduction Ethylene Polymerization By Titanium Catalysts. By Vanadium Catalysts. By Homogeneous Zirconium Catalysts. By Chromium Catalysts. By Nickel-Ylid Catalyst. Propylene Polymerization Heterogeneous Catalysts. Homogeneous Catalysts. Butadiene Polymerization By Titanium Catalysts. By Cobalt and Nickel Catalysts. By Lithium. Olefin Metathesis

Carbon Monoxide Reactions Introduction Oxidation and Disproportionation of Carbon Monoxide Reductions of Carbon Monoxide Base-Catalyzed Reactions of Carbon Monoxide Acid-Catalyzed Reactions of Carbon Monoxide Reactions of Carbon Monoxide with Transition Metals Coordinative Addition of Carbon Monoxide Insertions of Carbon Monoxide Chemisorption of Carbon Monoxide Metal-CatalyzedReactions of Carbon Monoxide Metal Carbonyls Important in Catalysis Chromium, Molybdenum, and Tungsten Carbonyls Preparation of the Hexacarbonyls M(C0)6 (M = Cr,Mo,W). Reactions of Hexacarbonyls of Cr, Mo, and W.

ix

294 294 297 298 298 299 299 303 304 305 306 307 308 309 312 312 312 313 314

315 315 316 317 318 319 320 32 1 323 324 326 329 334 334 336

X

14.6.2.2. 14.6.2.2.1. 14.6.2.2.2. 14.6.2.3. 14.6.2.3.1. 14.6.2.3.2. 14.6.2.4. 14.6.2.4.1. 14.6.2.4.2. 14.6.2.5. 14.6.2.5.1. 14.6.2.5.2. 14.6.3. 14.6.3.1. 14.6.3.2. 14.6.3.3. 14.6.3.4. 14.6.4. 14.6.4.1. 14.6.4.2. 14.6.4.3. 14.6.5. 14.6.5.1. 14.6.5.1.l. 14.6.5.1.2. 14.6.5.1.3. 14.6.5.2. 14.6.5.3. 14.6.5.4. 14.6.5.4.1. 14.6.5.4.2. 14.6.5.5. 14.6.6. 14.6.6.1 . 14.6.6.2. 14.6.6.2.1. 14.6.6.2.2. 14.6.6.3. 14.6.6.3.1. 14.6.6.3.2. 14.6.6.3.3. 14.6.6.4.


Manganese and Rhenium Carbonyls Preparation of the Metal Carbonyls. Reactions of the Carbonyls of Mn and Re. Iron and Ruthenium Carbonyls Preparation of the Metal Carbonyls. Reactions of the Carbonyls of Iron and Ruthenium. Cobalt Carbonyls Preparation of Cobalt Carbonyls. Reactions of Cobalt Carbonyls. Nickel Carbonyls Preparation of Tetracarbonylnickel(0). Reactions of Tetracarbonylnickel. Hydroformylation of Olefins by Cobalt Catalysts by Rhodium Catalysts by Ruthenium Catalysts by Platinum Catalysts Hydrocarboxylation of Olefins by Cobalt Catalysts by Rhodium and Iridium Catalysts by Palladium and Platinum Catalysts Carbonylation and Reductive Carbonylation of C-OH and C-OR Bonds Carbonylation of Alcohols by Cobalt Catalysts. by Rhodium Catalysts. by Other Metal Catalysts. lsomerization of Formates Carbonylation of Esters Reductive Carbonylation of Alcohols by Cobalt Catalysts. by Other Metals. Reductive Carbonylation of Esters Oxidation and Reduction of CO Oxidation In the Water Gas Shift Reaction General Aspects. Applications of the Reaction. In Reduction To Formyl. To Alkyl, Hydroxyalkyl, and Alkoxyl Ligands and Reductive Coupling. To Alcohols and Alkanes. Reduction by H,

338 338 340 343 343 344 348 348 349 35 1 35 1 352 353 354 358 362 363 364 365 367 369 372 373 373 373 375 376 377 377 378 379 380 38 1 382 384 384 387 388 388 389 392 394


14.7. 14.7.1. 14.7.2. 14.7.2.1. 14.7.2.2. 14.7.2.3. 14.7.2.4. 14.7.2.5.

14.8. 14.8.1. 14.8.2. 14.8.2.1. 14.8.2.1 .l. 14.8.2.1.2. 14.8.2.2. 14.8.2.3. 14.8.2.3.1. 14.8.2.3.2. 14.8.2.3.3. 14.8.2.3.4. 14.8.2.3.5. 14.8.2.3.6. 14.8.2.4. 14.8.3. 14.8.3.1. 14.8.3.2. 14.8.3.3. 14.8.3.4. 14.8.3.5. 14.8.4. 14.8.4.1. 14.8.4.1.1. 14.8.4.1.2. 14.8.4.2. 14.8.4.2.1. 14.8.4.2.2. 14.8.4.3. 14.8.4.3.1. 14.8.4.3.2. 14.8.4.3.3.

Oxidation Introduction Oxidation of Saturated Unactivated and Activated C-H Bonds In Methane Oxidation In Butane Oxidation In Cyclohexane Oxidation In Toluene Oxidation In Xylene Oxidation

Bioinorganic Catalysis Introduction Cobalamin Reactions Cobalamin Models Formation of the Cobalt-Carbon Bond. Cleavage of the Cobalt-CarbonBond. Cobalamin-Catalyzed Enzymatic Reactions Bioalkylation Mercury. Arsenic. Lead. Selenium and Tellurium. Tin. Other Metals, Metalloids, and Nonmetals. Biomethylation Mechanisms In Oxygen Transport Nature of the Bound Dioxygen Natural Oxygen Carriers Cobalt(l1) Complexes Iron(l1) Complexes Manganese(l1) and Copper(I) Complexes In Oxidases In Cytochrome Oxidases Reactions. Models. In Copper-Containing Oxidases Reactions. Models. In Peroxidases and Catalases Reactions. Structures. Models.

xi

398 398 398 398 399 402 404 406

408 408 408 408 408 414 419 426 427 428 429 429 430 430 43 1 433 434 435 436 438 44 1 442 442 445 446 448 450 452 454 454 456 457

xii

Contents of Volume 16 ~~

14.8.5. 14.8.6. 14.8.6.1. 14.8.6.2. 14.8.6.2.1. 14.8.6.2.2. 14.8.6.2.3. 14.8.6.3. 14.8.6.4. 14.8.7. 14.8.7.1. 14.8.7.2. 14.8.7.3. 14.8.7.4. 14.8.8. 14.8.8.1. 14.8.8.2. 14.8.8.2.1. 14.8.8.2.2. 14.8.8.2.3.

The Catechol Dioxygenases In Magnesium and Manganese Enzymes Introduction Reactions Current Principles. Models. Physical Methods. Sources Specific Examples In Calcium Binding Proteins Introduction Characteristics of the Ca2+ Ion lntracellular Catalysis Extracellular Enzymes In Selenium Enzymes Introduction Forms of Selenium Present in Biological Molecules Glutathione Peroxidase. Formate Dehydrogenase. Glycine Reductase.

458 463 463 464 464 465 466 467 468 473 473 474 475 484 488 488 488 490 492 493

List of Abbreviations

495

Author Index

501

Compound Index

565

Subject Index

649

How to Use this Book 1. Organization of Subject Matter 1.l.Logic of Subdivision and Add-on Chapters This volume is part of a series that describes all of inorganic reaction chemistry. The contents are subdivided systematically and so are the contents of the entire series. Using the periodic system as a correlative device, it is shown how bonds between pairs of elements can be made. Treatment begins with hydrogen making a bond to itself in H, and proceeds according to the periodic table with the bonds formed by hydrogen to the halogens, the groups headed by oxygen, nitrogen, carbon, boron, beryllium and lithium, to the transition and inner-transition metals and to the members of group zero. Next it is considered how the halogens form bonds among themselves and then to the elements of the main groups VI to I, the transition and inner-transition metals and the zero-group gases. The process repeats itself with descriptions of the members of each successive periodic group making bonds to all the remaining elements not yet treated until group zero is reached. At this point all actual as well as possible combinations have been covered. The focus is on the primary formation of bonds, not on subsequent reactions of the products to form other bonds. These latter reactions are covered at the places where the formation of those bonds is described. Reactions in which atoms merely change their oxidation states are not included, nor are reactions in which the same pairs of elements come together again in the product (for example, in metatheses or redistributions). Physical and spectroscopic properties or structural details of the products are not covered by the reaction volumes; the latter are concerned with synthetic utility based on yield, economy of ingredients, purity of product, specificity, etc. The preparation of short-lived transient species is not described. While in principle the systematization described above could suffice to deal with all the relevant material, there are other topics that inorganic chemists customarily identify as being useful in organizing reaction information and that do not fit into the scheme. These topics are the subject of eight additional chapters constituting the last four volumes of the series. These chapters are systematic only within their own confines. Their inclusion is based on the best judgment of the Editorial Advisory Board as to what would be most useful currently as well as effective in guiding the future of inorganic reaction chemistry.

1.2. Use of Decimal Section Numbers The organization of the material is readily apparent through the use of numbers and headings. Chapters are broken down into divisions, sections, and subsections, xiii

xiv

How to Use this Book

which have short descriptive headings and are numbered according to the following scheme: 1. Major Heading 1.1. Chapter Heading 1.1.1, Division Heading 1.1.1.1. Section Heading 1.1.1.1.1. Subsection Heading Further subdivision of a five-digit “slice” utilizes lower-case Roman numerals in parentheses: (i), (ii), (iii), etc. It is often found that as a consequence of the organization, cognate material is located in different chapters but in similarly numbered pieces, i.e., in parallel sections. Section numbers, rather than page numbers, are the key by which the material is accessed through the various indexes.

1.3. Building of Headings 1.3.1. Headings Forming Part of a Sentence Most headings are sentence-fragmentphrases which constitute sentences when combined. Usually a period signifies the end of a combined sentence. In order to reconstitute the context in which a heading is to be read, superior-rank titles are printed as running heads on each age. When the sentences are put together from their constituent parts, they describe the contents of the piece at hand. For an example, see 2.3 below.

1.3.2. Headings Forming Part of an Enumeration For some material it is not useful to construct title sentences as described above. In these cases hierarchical lists, in which the topics are enumerated, are more appropriate. To inform the reader fully about the nature of the material being described, the headings of connected sections that are superior in hierarchy always occur as running heads at the top of each page.

2. Access and Reference Tools 2.1. Plan of the Entire Series (Front Endpaper) Printed on the inside of the front cover is a list, compiled from all 18 reaction volumes, of the major and chapter headings, that is, all headings that are preceded by a one- or two-digit decimal section number. This list shows in which volumes the headings occur and highlights the contents of the volume that is at hand by means of a gray tint.


xv

2.2. Contents of the Volume at Hand All the headings, down to the title of the smallest decimal-numberedsubsection, are listed in the detailed table of contents of each volume. For each heading the table of contents shows the decimal section number by which it is preceded and the number of the page on which it is found. Beside the decimal section numbers, successive indentations reveal the hierarchy of the sections and thereby facilitate the comprehension of the phrase (or of the enumerative sequence) to which the headings of hierarchically successive sections combine. To reconstitute the context in which the heading of a section must be read to become meaningful, relevant headings of sections superior in hierarchy are repeated at the top of every page of the table of contents. The repetitive occurrences of these headings is indicated by the fact that position and page numbers are omitted.

2.3. Running Heads In order to indicate the hierarchical position of a section, the top of every page of text shows the headings of up to three connected sections that are superior in hierarchy. These running heads provide the context within which the title of the section under discussion becomes meaningful. As an example, the page of Volume 1 on which section 1.4.9.1.3 “in the Production of Methanol” starts, carries the running heads: 1.4. The Formation of Bonds between Hydrogen and 0, S , Se, Te, Po 1.4.9. by Industrial Processes 1.4.9.1. Involving Oxygen Compounds

whereby the phrase “in the Production of Methanol” is put into its proper perspective.

2.4. List of Abbreviations Preceding the indexes there is a list of those abbreviations that are frequently used in the text of the volume at hand or in companion volumes. This list varies somewhat in length from volume to volume; that is, it becomes more comprehensive as new volumes are published. Abbreviations that are used incidentally or have no general applicability are not included in the list but are explained at the place of occurrence in the text.

2.5. Author Index The author index is compiled by computer from the lists of references. Thus it tells whose publications are cited and in that respect is comprehensive. It is not a list of authors, beyond those cited in the references, whose results are reported in the text. However, as the references cited are leading ones, consulting them, along

xvi


with the use of appropriate works of the secondary literature, will rapidly lead to the complete literature related to any particular subject covered. Each envy in the author index refers the user to the appropriate section number.

2.6. Compound Index The compound index lists individual, fully specified compositionsof matter that are mentioned in the text. It is an index of empirical formulas, ordered according to the following system: the elements within a given formula occur in alphabetical sequence except for C, or C and H if present, which always come first. Thus, the empirical formula for Ti(SO,), is O,S,Ti BH,*NH, BH6N Be2C03 CBe,O, CsHBr, Br,CsH Al(HCO,), C,H,AlO, The formulas themselves are ordered alphanumerically without exception; that is, the formulas listed above follow each other in the sequence BH,N, Br,CsH, CBe,O,, C,H,A109, O,S,Ti. A compound index constructed by these principles tells whether a given compound is present. It cannot provide information about compound classes, for example, all aluminum derivatives or all compounds containing phosphorus. In order to open this route of access as well, the compound index is augmented by successively permuted versions of all empirical formulas. Thus the number of appearances that an empirical formula makes in the compound index is equal to the number of elements it contains. As an example, C,H,AlO,, mentioned above, will appear as such and, at the appropriate positions in the alphanumeric sequence, as H,AlO,*C,, A10,*C3H3 and 09*C,H,A1. The asterisk identifies a permuted formula and allows the original formula to be reconstructed by shifting to the front the elements that follow the asterisk. Each nonpermuted formula is followed by linearized structural formulas that indicate how the elements are combined in groups. They reveal the connectivity of the compounds underlying each empirical formula and serve to distinguish substances which are identical in composition but differ in the arrangement of elements (isomers). As an example, the empirical formula C,H,,O might be followed by the linearized structural formulas (CH,CH,),O, CH,(CH,),OCH,, (CH,),CHOCH,, CH,(CH,),OH, (CH,),CHCH,OH and CH,CH,(CH,)CHOH to identify the various ethers and alcohols that have the elemental composition C4H,,0. Each linearized structural formula is followed in a third column by keywords describing the context in which it is discussed and by the number(s) of the section(s) in which it occurs.

2.7. Subject Index The subject index provides access to the text by way of methods, techniques, reaction types, apparatus, effects and other phenomena. Also, it lists compound


xvii

classes such as organotin compounds or rare-earth hydrides which cannot be expressed by the empirical formulas of the compound index. For multiple entries, additional keywords indicate contexts and thereby avoid the retrieval of information that is irrelevant to the user’s need. Again, section numbers are used to direct the reader to those positions in the book where substantial information is to be found.

2.8. Periodic Table (Back Endpaper) Reference to periodic groups avoids cumbersome enumerations. Section headings in the series employ the nomenclature. Unfortunately, however, there is at the present time no general agreement on group designations. In fact, the scheme that is most widely used (combining a group number with the letters A and B) is accompanied by two mutually contradictory interpretations.Thus, titanium may be a group IVA or group IVB element depending on the school to which one adheres or the part of the world in which one resides. In order to clarify the situation for the purposes of the series, a suitable labeled periodic table is printed on the inside back cover of each volume. All references to periodic group designations in the series refer to this scheme.

Preface to the Series Inorganic Reactions and Methods constitutes a closed-end series of books designed to present the state of the art of synthetic inorganic chemistry in an unprecedented manner. So far, access to knowledge in inorganic chemistry has been provided almost exclusively using the elements or classes of compounds as starting points. In the first 18 volumes of Inorganic Reactions and Methods, it is bond formation and type of reaction that form the basis of classification. This new route of access has required new approaches. Rather than sewing together a collection of review articles, a framework has had to be designed that reflects the creative potential of the science and is hoped to stimulate its further development by identifying areas of research that are most likely to be fruitful. The reaction volumes describe methods by which bonds between the elements can be formed. The work opens with hydrogen making a bond to itself in H, and proceeds through the formation of bonds between hydrogen and the halogens, the groups headed by oxygen, nitrogen, carbon, boron, beryllium and lithium to the formation of bonds between hydrogen and the transition and inner-transition metals and elements of group zero. This pattern is repeated across the periodic system until all possible combinations of the elements have been treated. This plan allows most reaction topics to be included in the sequence where appropriate. Reaction types that do not arise from the systematics of the plan are brought together in the concluding chapters on oxidative addition and reductive elimination, insertions and their reverse, electron transfer and electrochemistry,photochemical and other energized reactions, oligomerization and polymerization, inorganic and bioinorganic catalysis and the formation of intercalation compounds and ceramics. The project has engaged a large number of the most able inorganic chemists as Editorial Advisors creating overall policy, as Editorial Consultantsdesigning detailed plans for the subsectionsof the work, and as authors whose expertise has been crucial for the quality of the treatment. The conception of the series and the details of its technical realization were the subject of careful planning for several years. The distinguished chemists who form the Editorial Advisory Board have devoted themselves to this exercise, reflecting the great importance of the project. It was a consequence of the systematics of the overall plan that publication of a volume had to await delivery of its very last contribution. Thus was the defect side of the genius of the system revealed as the excruciating process of extracting the rate-limiting manuscripts began. Intense editorial effort was required in order to bring forth the work in a timely way. The production process had to be designed so that the insertion of new material was possible up to the very last stage, enabling authors to update their pieces with the latest developments. The publisher supported the cost of a computerized bibliographic search of the literature and a second one for updating. xix

xx

Preface to the Series

Each contribution has been subjected to an intensive process of scientific and linguistic editing in order to homogenize the numerous individual pieces, as well as to provide the highest practicable density of information. This had several important consequences. First, virtually all semblances of the authors’ individual styles have been excised. Second, it was learned during the editorial process that greater economy of language could be achieved by dropping conventionally employed modifiers (such as very) and eliminating italics used for emphasis, quotation marks around nonquoted words, or parentheses around phrases, the result being a gain in clarity and readability. Because the series focuses on the chemistry rather than the chemical literature, the need to tell who has reported what, how and when can be considered of secondary importance. This has made it possible to bring all sentences describing experiments into the present tense. Information on who published what is still to be found in the reference lists. A further consequence is that authors have been burdened neither with identifying leading practitioners, nor with attributing priority for discovery, a job that taxes even the talents of professional historians of science. The authors’ task then devolved to one of describing inorganic chemical reactions, with emphasis on synthetic utility, yield, economy, availability of starting materials, purity of product, specificity, side reactions, etc. The elimination of the names of people from the text is by far the most controversial feature. Chemistry is plagued by the use of nondescriptive names in place of more expository terms. We have everything from Abegg’s rule, Adkin’s catalyst, Admiralty brass, Alfven number, the Amadori rearrangement and Adurssov oxidation to the Zdanovskii law, Zeeman effect, Zincke cleavage and Zinin reduction. Even well-practiced chemists cannot define these terms precisely except for their own areas of specialty, and no single source exists to serve as a guide. Despite these arguments, the attempt to replace names of people by more descriptive phrases was met in many cases by a warmly negative reaction by our colleague authors, notwithstanding the obvious improvements wrought in terms of lucidity, freedom from obscurity and obfuscation and, especially, ease of access to information by the outsider or student. Further steps toward universality are taken by the replacement of element and compound names wherever possible by symbols and formulas, and by adding to data in older units their recalculated SI equivalents. The usefulness of the reference sections has been increased by giving journal-title abbreviations according to the Chemical Abstracts Service Source Index, by listing in each reference all of its authors and by accompanying references to patents and journals that may be difficult to access by their Chemical Abstracts citations. Mathematical signs and common abbreviations are employed to help condense prose and a glossary of the latter is provided in each volume. Dangerous or potentially dangerous procedures are highlighted in safety notes printed in boldface type. The organization of the material should become readily apparent from an examination of the headings listed in the table of contents. Combining the words constituting the headings, starting with the major heading (one digit) and continuing through the major chapter heading (two digits), division heading (three digits), section heading (four digits) to the subsection heading (five digits), reveals at once the subject of a “slice” of the plan. Each slice is a self-contained unit. It includes its


xxi

own list of references and provides definitions of unusual terms that may be used in it. The reader, therefore, through the table of contents alone, can in most instances quickly reach the desired material and derive the information wanted. In addition there is for each volume an author index (derived from the lists of references) and a subject index that lists compound classes, methods, techniques, apparatus, effects and other phenomena. An index of empirical formulas is also provided. Here in each formula the element symbols are arranged in alphabetical order except that C, or C and H if present, always come first. Moreover, each empirical formula is permuted successively. Each permuted formula is placed in its alphabetical position and cross referenced to the original formula. Therefore, the number of appearances that an empirical formula makes in the index equals the number of its elements. By this procedure all compounds containing a given element come together in one place in the index. Each original empirical formula is followed by a linearized structural formula and keywords describing the context in which the compound is discussed. All indexes refer the user to subsection rather than page number. Because the choice of designations of groups in the periodic table is currently in a state of flux, it was decided to conform to the practice of several leading inorganic texts. To avoid confusion an appropriately labeled periodic table is printed on the back endpaper. From the nature of the work it is obvious that probably not more than two persons will ever read it entire: myself and the publisher’s copy editor, Arline Keithe. She, as well as Steven Bedney, Production Manager of VCH Publishers, are to be thanked for their unflagging devotion to the highest editorial standards. The original conception for this series was the brainchild of Dr. Hans F. Ebel, Director of the Editorial Department of VCH Verlagsgesellschaft in Weinheim, Federal Republic of Germany, who also played midwife at the birth of the plan of these reaction volumes with my former mentor, Professor Alan G. MacDiarmid of the University of Pennsylvania, and me in attendance, during the Anaheim, California, American Chemical Society Meeting in the Spring of 1978. Much of what has finally emerged is the product of the inventiveness and imagination of Professor Helmut Griinewald, President of VCH Verlagsgesellschaft. It is a pleasure to acknowledge that I have learned much from him during the course of our association. Ms. Nancy L. Burnett is to be thanked for typing everything that had to do with the series from its inception to this time. Directing an operation of this magnitude without her help would have been unimaginable. My wife Rose stood by with good cheer while two rooms of our home filled up with 10,OOO manuscript pages, their copies and attendant correspondence. Finally, and most important, an enormous debt of gratitude toward all our authors is to be recorded. These experts were asked to prepare brief summaries of their knowledge, ordered in logical sequence by our plan. In addition, they often involved themselves in improving the original conception by recommending further refinements and elaborations. The plan of the work as it is being published can truly be said to be the product of the labors of the advisors and consultants on the editorial side as well as the many, many authors who were able to augment more general knowledge with their own detailed information and ideas. Because of the unusually

xxii


strict requirements of the series, authors had not only to compose their pieces to fit within narrowly constrained limits of space, format and scope, but after delivery to a short deadline were expected to stand by while an intrusive editorial process homogenized their own prose styles out of existence and shrank the length of their expositions. These long-sufferingcolleagues had then to endure the wait for the very last manuscript scheduled for their volume to be delivered so that their work could be published, often after a further diligent search of the literature to insure that the latest discoveries were being cited and that claims for facts now proved false were eliminated. To these co-workers (270 for the reaction volumes alone), from whom so much was demanded but who continued to place their knowledge and talents unstintingly at the disposal of the project, we dedicate this series. J.J. ZUCKERMAN Norman, Oklahoma July 4, 1985 The JJZ vision of the Inorganic Reactions and Methods series, and the unique systemization of inorganic chemistry it embodies, continue as the series moves to completion. During the period of A.P. Hagen’s editorship and since, under my direction, every effort has been made to keep the style consistent with what was originally conceived. The continued development of this exciting series has depended upon the untiring efforts and patience of the many series authors and Dr. Barbara A. Chernow of Chernow Editorial Services. A.D. NORMAN Boulder, Colorado July 10, 1993

Editorial Consultants to the Series Professor H.R. Allcock

Pennsylvania State University

Professor J.S. Anderson University of Aberystwyth

Professor F.C. Anson California Institute of Technology Dr. M.G. Barker

University of Nottingham

Professor D.J. Cardin Trinity College

Professor M.H. Chisholm Indiana University

Professor C. Cros

Professor J. R. Etourneau

Laboratoire de Chemie du Solide du C.N.R.S.

Professor G.L. Geoffroy Pennsylvania State University Professor L.S. Hegedus Colorado State University

Professor W.L. Jolly

University of California at Berkeley

Professor C.B. Meyer

University of Washington

Professor H. Noth Universitat Munchen


Professor H. Nowotny

Dr. B. Darriet

Dr. G.W. Parshall

University of Connecticut


E.I. du Pont de Nemours

Professor E.A.V. Ebsworth University of Edinburgh


Professor J. J. Eisch

Professor J. Rouxel

State University of New York at Binghamton

Professor M. Pouchard

Laboratoire de Chemie MinCrale au C.N.R.S.

xxiv

Editorial Consultants to the Series

Professor R. Schmutzler Technische Universitat Braunschweig

Dr. N. Sutin Brookhaven National Laboratory

Professor A.W. Searcy University of California at Berkeley

Professor R.A. Walton Purdue University

Professor D. Seyferth Massachusetts Institute of Technology

Dr. J.H. Wernick Bell Laboratories

Contributors to Volume 16 Professor Michael T. Ashby Department of Chemistry and Biochemistry 620 Parrington Oval University of Oklahoma Norman, Oklahoma 73019 (Sections 14.3.2- 14.3.3.6) Professor Gerald T. Babcock Department of Chemistry Michigan State University East Lansing, Minnesota 48824 (Sections 14.8.4-14.8.4.2.2) Professor Fred Basolo Department of Chemistry Northwestern University Evanston, Illinois 60208 (Section 14.8.3) Dr. Ernst Billig Technical Development Department P.O. Box 8361 Union Carbide Chemicals and Plastics Co., Inc. South Charleston, West Virginia 25303 (Sections 14.6.3-14.6.4) Dr. Fred E. Brinckman Polymers Division National Institutes of Science and Technology Gaithersburg, Maryland 20899 (Sections 14.8.2.3- 14.8.2.4)

Dr. E. S. Brown Technical Center Specialty Chemicals Division Union Carbide Chemicals and Plastics Co., Inc. P.O. Box 8361 South Charleston, WV 25303 (Section 14.4.6) Professor Fausto Calderazzo Dipartimento di Chimica e Chemica Industriale Universith di Pisa Via Risorgimento 35 56100 Pisa, Italy (Sections 14.6-14.6.2.5.2) Professor David H. Dolphin Department of Chemistry University of British Columbia 2036 Main Mall Vancouver, British Columbia V6T 1Z1 Canada (Sections 14.8.2-14.8.2.2) Dr. B. Duane Dombeck Technical Center Research and Development Department Union Carbide Chemicals and Plastics Co., Inc. P.O. Box 8361 South Charleston, West Virginia 25303 (Section 14.4.6)

xxvi

Contributors to Volume 16

Professor H. Brian Dunford Department of Chemistry Faculty of Science University of Alberta Edmonton, Alberta T6G 2G2 Canada (Section 14.8.4.3)

Professor William Dew. Horrocks, Jr. Department of Chemistry Pennsylvania State University University Park, Pennsylvania 16802 (Section 14.8.7)

Professor Richard S. Eisenberg Department of Chemistry 404 Hutchinson Hall, River Campus University of Rochester Rochester, New York 14627 (Section 14.6.6)

Steacie Institute for Molecular Sciences National Research Council of Canada Ottawa, Ontario KIA OR6 Canada (Section 14:7.5)

Professor Bruce C. Gates Department of Chemical Engineering University of California Davis, California 95616 (Sections 14.2- 14.2.6) Professor Jean-Louis Gras Laboratoire de Synthese Organique D12 Maitre de Recherches CNRS, URA 1411 FacultC des Sciences St. Jerome F- 13397 Marseille Cedex 13, France (Sections 14.3.4- 14.3.7.3)

Dr. J. Anthony Howard

Professor Brian R. James Department of Chemistry 2036 Main Mall University of British Columbia Vancouver, British Columbia Canada V6T 1Y6 (Sections 14.3.2- 14.3.2.2, 14.3.3- 14.3.3.2) Professor Walter Kaminsky Institute Tech. and Makromelek. Chem. Martin Luther King Platz 6 Bundestrasse 45 D-2000 Hamburg 13, Germany (Sections 14.5.2- 14.5.3.3.3)

Professor Jack Halpern Department of Chemistry University of Chicago 5735 South Ellis Avenue Chicago, Illinois 60637 (Section 14.1)

Professor Reinhard Kramolowsky Institut fur Anorganische und Angewandte Chemie der Universitat Hamburg Martin Luther King Platz 6 D-2000 Hamburg 13, Germany (Sections 14.5.2- 14.5.3.3.3)

Professor Gordon A. Hamilton Department of Veterinary Science and Department of Chemistry Pennsylvania State University University Park, Pennsylvania 16802 (Section 14.8.8)

Professor Clifford P. Kubiak Department of Chemistry 1393 Brown Building Purdue University West Lafayette, Indiana 47907 (Section 14.6.6)

Contributors to Volume 16

Dr. G . Alex Mills Center for Catalytic Science and Technology Department of Chemical Engineering University of Delaware Newark, DE 19716-3119 (Section 14.2.7) Professor Iwao Ojima Department of Chemistry SUNY Stony Brook Stony Brook, New York 11794-3400 (Sections 14.4- 14.4.5.2) Dr. Gregory Olson Coal Preparation Program Pittsburg Energy Technology Center Pittsburg, Pennsylvania 15236 (Section 14.8.2.3) Professor Milton Orchin Department of Chemistry Mail Location 172 University of Cincinnati Cincinnati, Ohio 4522 1-0172 (Sections 14.4- 14.5.1.3)

xxvii

Dr. Channo C. Reddy

Department of Veterinary Science and Department of Chemistry Pennsylvania State University University Park, PA 16802 (Section 14.8.8)

Dr. Pagona F. Roussi Department of Chemistry University of British Columbia 2036 Main Mall Vancouver, British Columbia V6T 1Z1 Canada (Sections 14.8.2- 14.8.2.2) Professor Frederick C. Wedler Department of Molecular and Cell Biology Pennsylvania State University 108 Althouse Lab University Park, Pennsylvania 16802 (Section 14.8.6)

8431 Willow Oak Lane Harrisburg, North Carolina 28075 (Sections 14.6.3- 14.6.4)

Dr. Richard W. Wegman Technical Center Research and Development Department P.O. Box 8361 Union Carbide Chemicals and Plastics Co., Inc. South Charleston, West Virginia 25303 (Section 14.6.5)

Professor Lawrence Que, Jr. Department of Chemistry Kolthoff and Smith Halls 207 Pleasant Street S.E. University of Minnesota Minneapolis, Minnesota 55455-0431 (Section 14.8.5)

Dr. Lily Yun Xie Department of Chemistry University of British Columbia 2036 Main Mall Vancouver, British Columbia V6T 1Z1 Canada (Sections 14.8.2-14.8.2.2)

Dr. Roy L. Pruett

lnorganic Reactions and Methods, VolumeI 6 Edited by J.J. Zuckerman, Arlan D. Norman Copyright 0 1993 by VCH Publishers, Inc.

14. Reactions Catalyzed by Inorganic

Compounds

14.1. Introduction: Principles of Catalysis

14.1.l.Catalysis as a Kinetic Phenomenon Catalysts function by opening up new pathways for chemical reactions, the contributions from which are reflected in enhanced reaction rates. Sometimes such catalytic pathways are closely related to those which operate in the absence of catalysts; more generally the catalytic mechanism is a distinctive one into which the catalyst enters as a reactant, undergoes chemical transformation, but is ultimately regenerated so that its concentration remains undiminished. Accordingly, such catalytic mechanisms are invariably stepwise processes. For any given catalytic path there can be constructed a corresponding uncatalyzed reaction path that may make a contribution to the rate in the absence of catalysts. In this context the role of the catalyst may be interpreted as that of stabilizing the intermediates of the corresponding uncatalyzed reaction path, this stabilization necessarily being greater than that of the reactants. An illustration of this theme is provided by the catalysis by Cuz+ and its complexes of the oxidation of H, by various oxidants such as 0,, Cr(VI), Fe(III), and Tl(III)’-’. This catalysis is effected through a mechanism in which the rate-determining step is the heterolytic (electrophilic) splitting of H, by Cu2+:

The corresponding uncatalyzed reaction mechanism

H,

+ 2Fe3+

-

2H+

+ 2Fe2+

(f)

is much less favorable since the endothermicity of reaction (d) and, hence, its activation enthalpy can be estimated to exceed 150 kJ/mol, as compared with AHS = 110 kJ/mol for the CuZ+-catalyzedreaction. This lowering of the activation energy can be attributed to stabilization by the catalyst, Cu2+,of the intermediate H-. Other instances of catalysis may be similarly interpreted. Catalytic pathways may be related to the alternative possibilities of “concerted” and “stepwise” pathways for the uncatalyzed reaction, respectively: 2

e

14. Reactions Catal zed by lnor anic Compounds 14.1. Introduction: Fbnciples of atalysis 14.1.1. Catalysis as a Kinetic Phenomenon \

concerted

/c-c\

\

/

3

c-c

r - I I

H

H

/c-

c\

H

H

(g)

/ \

H-H ___f

H-H

I

\

/ __*

/

c-c,

I 1

H

/

H

(h)

exemplified for the simple case of addition of H2 to an olefin. Both of these pathways are expected to be slow; the concerted path because of the high energy of the fourcentered transition state (according to the rules of orbital-symmetry c~nservation)~, and the stepwise path because of the high energy H' atom and free-radical intermediates. The catalysis of such olefin hydrogenation reactions, for example, by Rh(1) complexes, could in principle be attributed to either concerted or stepwise routes, respectively. M

\c-cf / concerted

/

j

I\

\

-M

\

/

~

c-

/I H

/

C\

I

H

According to the first of these interpretations, the role of the metal center (M) is the stabilization of the otherwise high-energy ,four-centered intermediate [through electronic interactions analogous to those involved in the well-known stabilization of the squareplanar, cyclobutadiene ligand through coordination as, e.g., in Fe(CO),(C,H,)]. On the other hand, according to the second interpretation, the role of the metal center is to participate in the formation of discrete hydrido- and organometallic intermediates, such as MH, and M(R)H, in which otherwise high-energy species such as H atoms and alkyl radicals are stabilized by coordination. A specific example of a mechanism of the latter type is that of the rhodium complexcatalyzed hydrogenation of 01efins~-~: Oxidative HZ

+ Reductive Elimination

L,,Rh"'/H

c;-'

[

>

Addition

- H-,c-C

Accessibility of different coordination numbers facilitates ligand dissociation and substrate incorporation into such complexes and accounts, at least in part, for the widespread roles of low-spin d8- [notably Co(I), Rh(I), and Pd(II)] and of d'O- [notably Ni(0) and Pd(O)] complexes in catalysis5v6. The influence of steric factors on ligand dissociation-association processes also is well recognized and documented. Bulky ligands, such as tertiary phosphines, tend to favor ligand dissociation and stabilize complexes of lower coordination number. The widespread roles of bulky phosphine ligands such as PPh, as components of catalytic complexes are related to this theme7. The catalytic activities of such complexes, e.g., of Rh(PPh,),CI as a hydrogenation catalyst, commonly are reduced or eliminated by replacing the bulky phosphines with smaller ones (e.g., PMe,) which exhibit a correspondingly smaller tendency to dissociate. In some cases, where spontaneous ligand dissociation is slow, catalytic activity may be enhanced by promoting such dissociation. One method of accomplishing this is through photoactivation, i.e., photochemical ligand dissociation'. Thus, the catalytic activities of Cr(CO), and Fe(CO), as hydrogenation catalysts are enhanced by UV-irradiation which presumably induces photodissociation of a CO ligand and generates the catalytically active, coordinately unsaturated Cr(CO), and Fe(CO), cr(co)6

hv

' Cr(CO), +

(c>

CO

Other methods of promoting ligand dissociation include protonation: [(NH,),COF]~+

H+

[(NH3),CoFHI3+

[(NH3),Co(0H2)l3'

+

HF

(e)

14.1.2. Basic Processes in Molecular Catalysis 14.1.2.2. Li and Dissociation and Association Processes 14.1.2.2.2. aornolytic Ligand Dissociation.

and metal-ion induced dissociation: [(NH3),CoC112+

9

-

a [(NH3),Co**C1*.Hgl4+ H2O Hg2

[ ( N H ~ > ~ C O ( ~+ H HgClf Z ) ~ ~ ~ (8) Such promotion is particularly effective in inducing the dissociation of chelated ligands as in the example of Eq. (f)". (J. HALPERN)

1. R. G. Wilkins, Kinetics and Mechanism by Reactions of Transition Metal Complexes, 2nd ed., Ch. 5, VCH, Weinheim, 1991. 2. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, Ch. 4, University Science Books, Mill Valley, CA, 1987. 3. J. Halpem, Inorg. Chim. Acta, 50, 1 (1981). 4. C. A. Tolman, P. Z. Meakin, D. L. Lindner, J. P. Jesson, J . Am. Chem. SOC., 96,2762 (1974). 5. J. Halpem, Adv. Chem. Ser., 70, 1 (1968). 6. J. Halpem, Acct. Chem. Res., 3, 386 (1970). 7. C. A. Tolman, Chem. Rev., 77, 313 (1977). 8. M. S . Wrighton, Ann. N . Y . Acad. Sci., 333, 188 (1980). 9. M. A. Schroeder and M. S . Wrighton, J . Am. Chem. SOC., 98,551 (1976). 10. F. Basolo, R. Pearson, Mechanisms of Inorganic Reactions, 2nd. ed., John Wiley, New York, 1967, p. 216. 14.1.2.2.2. Homoiytic Ligand Dissociation.

Whereas homolytic ligand dissociation is not commonly observed for inorganic complexes, it has been identified as an important process in organometallic chemistry where it is favored by the characteristic weakness of transition metal-alkyl o-bonds'. Recent determinations yield metal alkyl band dissociation energies for CH,-Mn(CO), (ca. 120 kJ/mol)2 and for several alkylcobalt complexes (ca. 80-100 kJ/m0l)'9~.Homolytic dissociation of such complexes results in the formation of free radicals and in the opening up of free radical catalytic pathways, e.g., for hydrogenation4. Important biochemical examples of free radical catalytic mechanisms, initiated by the homolytic dissociation of a transition metal-carbon bond (Lea,the 5 '-deoxyadenosyl-cobalt bond of coenzyme Blz) are provided by the coenzyme B,,-promoted rearrangements (see Section 14.8.3),s6. (J. HALPERN)

J. Halpem, A.C.S. Symp. Ser., 428, 100 (1990). D. L. S. Brown, J. A. Connor, H. A. Skinner, J . Organomet. Chem., 81,403 (1974). J. Halpem, Polyhedren, 7 , 1483 (1988). J. Halpem, Pure Appl. Chem., 51, 2171 (1979). J. Halpem, in Vitamin B,,, D. Dolphin, ed., Vol. 1, John Wiley, New York, 1981, p. 501, and references therein. 6. J. Halpem, Bull. Sec. Chim. France, 187 (1988). 1. 2. 3. 4. 5.


14.1.2. Basic Processes in Molecular Catalysis 14.1.2.2. Li and Dissociation and Association Processes 14.1.2.2.2. aornolytic Ligand Dissociation.

and metal-ion induced dissociation: [(NH3),CoC112+

9

-

a [(NH3),Co**C1*.Hgl4+ H2O Hg2

[ ( N H ~ > ~ C O ( ~+ H HgClf Z ) ~ ~ ~ (8) Such promotion is particularly effective in inducing the dissociation of chelated ligands as in the example of Eq. (f)". (J. HALPERN)

1. R. G. Wilkins, Kinetics and Mechanism by Reactions of Transition Metal Complexes, 2nd ed., Ch. 5, VCH, Weinheim, 1991. 2. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, Ch. 4, University Science Books, Mill Valley, CA, 1987. 3. J. Halpem, Inorg. Chim. Acta, 50, 1 (1981). 4. C. A. Tolman, P. Z. Meakin, D. L. Lindner, J. P. Jesson, J . Am. Chem. SOC., 96,2762 (1974). 5. J. Halpem, Adv. Chem. Ser., 70, 1 (1968). 6. J. Halpem, Acct. Chem. Res., 3, 386 (1970). 7. C. A. Tolman, Chem. Rev., 77, 313 (1977). 8. M. S . Wrighton, Ann. N . Y . Acad. Sci., 333, 188 (1980). 9. M. A. Schroeder and M. S . Wrighton, J . Am. Chem. SOC., 98,551 (1976). 10. F. Basolo, R. Pearson, Mechanisms of Inorganic Reactions, 2nd. ed., John Wiley, New York, 1967, p. 216. 14.1.2.2.2. Homoiytic Ligand Dissociation.

Whereas homolytic ligand dissociation is not commonly observed for inorganic complexes, it has been identified as an important process in organometallic chemistry where it is favored by the characteristic weakness of transition metal-alkyl o-bonds'. Recent determinations yield metal alkyl band dissociation energies for CH,-Mn(CO), (ca. 120 kJ/mol)2 and for several alkylcobalt complexes (ca. 80-100 kJ/m0l)'9~.Homolytic dissociation of such complexes results in the formation of free radicals and in the opening up of free radical catalytic pathways, e.g., for hydrogenation4. Important biochemical examples of free radical catalytic mechanisms, initiated by the homolytic dissociation of a transition metal-carbon bond (Lea,the 5 '-deoxyadenosyl-cobalt bond of coenzyme Blz) are provided by the coenzyme B,,-promoted rearrangements (see Section 14.8.3),s6. (J. HALPERN)

J. Halpem, A.C.S. Symp. Ser., 428, 100 (1990). D. L. S. Brown, J. A. Connor, H. A. Skinner, J . Organomet. Chem., 81,403 (1974). J. Halpem, Polyhedren, 7 , 1483 (1988). J. Halpem, Pure Appl. Chem., 51, 2171 (1979). J. Halpem, in Vitamin B,,, D. Dolphin, ed., Vol. 1, John Wiley, New York, 1981, p. 501, and references therein. 6. J. Halpem, Bull. Sec. Chim. France, 187 (1988). 1. 2. 3. 4. 5.


10

14.1. Introduction: Principles of Catalysis 14.1.2. Basic Processes in Molecular Catalysis 14.1.2.3. Promotion of Nucleophilic Reactions by Electron Withdrawal

14.1.2.3. Promotion of Nucleophlllc Reactions by Electron Withdrawal from Reactants

Coordination to a positively charged metal ion is expected to increase the positive charge density on a substrate, thereby rendering it more susceptible to nucleophilic attack. The role of a metal ion catalyst in this context is essentially that of a Lewis acid or superacid and the origin of such catalytic activity is closely related to that underlying many of the familiar catalytic effects of protonic acids. Despite the higher positive charges that they can bear, metal ions are frequently not as effective, because of their more ionic bonding (particularly when coordinated to small electronegative atoms such as oxygen or nitrogen), in polarizing, or transferring positive charge to coordinated substrates. This is reflected, e.g., in the relative acidities of protonated and coordinated acids, some values of which are listed in Table 1. There are, however, several circumstances in which metal ions may be more effective than protons in catalyzing nucleophilic reactions', Among these are the following:

1. With species (e.g., soft or class B bases) such as halide ions, which exhibit low basicity toward protons, but high basicity toward certain metal ions2. Promotion by metal ions of halide ion displacement from both organic and inorganic halides (a class of reactions not generally susceptible to protonic acid catalysis) serves to illustrate this effect, e.g., where R = CH3, (NH3),Co3+, etc.: RCl

~ g + etc. ,

I

I

H*O

'ROH

t

+ H+ + AgCl

2. With species such as CO, olefins, acetylenes, etc., whose coordination to metal ions depends on rr-bonding through the donation of daelectrons from the metal to the ligand, e.g.3*4

TABLE1. COMPARISON OF ACIDITIES OF FREE(XH,), PROTONATED (XH, AND

+

,),

COORDINATED(MXH,) ACIDS

Free Acid

pK,

Protonated Acid

PK,

Coordinated Acid

PK,

14.1, Introduction: Principles of Catalysis 14.1.2. Basic Processes in Molecular Catalysis 14.1.2.4. Catalysis of Electrophilic Reactions by Proton Loss

M"+

M"+

+ CO I [M(CO)]"+

+ CH,=CH,

HZO [MH

e

-OH]("-')+M(n-Z)+

- ')

+ CO, + H C

(b)

H2O

I[M(CH,=CH,)]"+

[M- CH,-CH,-OH]("

11

H+

-M("

-),

+ CH,CHO + H+

(c) Such steps play important roles in the oxidation of C0395or ole fin^^-^ by Hg(II), Tl(III), Pd(II), Co(III), Pt(IV), etc., or in oxidations catalyzed by salts of these metal ions. Reaction (b) also is a key step in the homogeneously catalyzed water-gas shift reaction: +

+

+

+

CO H,OeCO, H, (4 3. When chelation contributes to the binding of the substrate to (or stabilization of the product by) the metal ion as in the catalysis by various metal ions (M = Cu, Ni, Mg, etc.) of the hydrolysis of amino acid esters according to the mechanism'-": R O R O

I

H,NC -C-

It

I

H I

1 M'+

.L

IH-C-C-OR P a + I

H,N

1:

0

\ / M

OR'

-

I

HZO

MZ+

2+

OH-

- OH-

,OR'

H-C-C

1

H,N

It

H,NC-C-OH

I'OH 0

+ R'OH

+ I

\ / M

(J. HALPERN) 1. J. Halpem, Pure Appl. Chem., 20, 59 (1969), and references therein. 2. F. Basolo, R. G . Pearson, Mechanisms of Inorganic Reactions, 2nd ed., John Wiley, New York, 1967, p. 23. 3. J. Halpem, Comments Inorg. Chem., 1 , 3 (1981), and references therein. 4. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, Ch. 7, University Science Books, Mill Valley, CA, 1987. 5. J. E. Bercaw, L. Y. Goh, J. Halpem, J. Am. Chem. SOC.,94, 6534 (1978). 6. P. M. Henry, Palladium Catalyzed Oxidations of Hydrocarbons, D. Reidel, Dordrecht, Holand, 1980. 7. J. E. Backvall, B. Akermark, S. 0. Ljunggren, J . Am. Chem. SOC.,101, 2411 (1979). 8. J. E. Byrd, J. Halpem, J . Am. Chem. SOC., 95, 2586 (1973). 9. J. Halpem, R. A. Jewsbury, J . Organomer. Chem., 181, 223 (1979). 10. H. Kroll, J. Am. Chem. SOC., 74,2036 (1952).

14.1.2.4. Catalysis of Electrophlllc Reactions by Proton Loss from a Coordinated Ligand

Paradoxically (in view of their positive charges), metal ions also may catalyze electrophilic attack by promoting loss of a proton from a coordinated ligand or by stabilizing a reactive (e.g., enol) form of the latter'.


14.1, Introduction: Principles of Catalysis 14.1.2. Basic Processes in Molecular Catalysis 14.1.2.4. Catalysis of Electrophilic Reactions by Proton Loss

M"+

M"+

+ CO I [M(CO)]"+

+ CH,=CH,

HZO [MH

e

-OH]("-')+M(n-Z)+

- ')

+ CO, + H C

(b)

H2O

I[M(CH,=CH,)]"+

[M- CH,-CH,-OH]("

11

H+

-M("

-),

+ CH,CHO + H+

(c) Such steps play important roles in the oxidation of C0395or ole fin^^-^ by Hg(II), Tl(III), Pd(II), Co(III), Pt(IV), etc., or in oxidations catalyzed by salts of these metal ions. Reaction (b) also is a key step in the homogeneously catalyzed water-gas shift reaction: +

+

+

+

CO H,OeCO, H, (4 3. When chelation contributes to the binding of the substrate to (or stabilization of the product by) the metal ion as in the catalysis by various metal ions (M = Cu, Ni, Mg, etc.) of the hydrolysis of amino acid esters according to the mechanism'-": R O R O

I

H,NC -C-

It

I

H I

1 M'+

.L

IH-C-C-OR P a + I

H,N

1:

0

\ / M

OR'

-

I

HZO

MZ+

2+

OH-

- OH-

,OR'

H-C-C

1

H,N

It

H,NC-C-OH

I'OH 0

+ R'OH

+ I

\ / M

(J. HALPERN) 1. J. Halpem, Pure Appl. Chem., 20, 59 (1969), and references therein. 2. F. Basolo, R. G . Pearson, Mechanisms of Inorganic Reactions, 2nd ed., John Wiley, New York, 1967, p. 23. 3. J. Halpem, Comments Inorg. Chem., 1 , 3 (1981), and references therein. 4. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, Ch. 7, University Science Books, Mill Valley, CA, 1987. 5. J. E. Bercaw, L. Y. Goh, J. Halpem, J. Am. Chem. SOC.,94, 6534 (1978). 6. P. M. Henry, Palladium Catalyzed Oxidations of Hydrocarbons, D. Reidel, Dordrecht, Holand, 1980. 7. J. E. Backvall, B. Akermark, S. 0. Ljunggren, J . Am. Chem. SOC.,101, 2411 (1979). 8. J. E. Byrd, J. Halpem, J . Am. Chem. SOC., 95, 2586 (1973). 9. J. Halpem, R. A. Jewsbury, J . Organomer. Chem., 181, 223 (1979). 10. H. Kroll, J. Am. Chem. SOC., 74,2036 (1952).

14.1.2.4. Catalysis of Electrophlllc Reactions by Proton Loss from a Coordinated Ligand

Paradoxically (in view of their positive charges), metal ions also may catalyze electrophilic attack by promoting loss of a proton from a coordinated ligand or by stabilizing a reactive (e.g., enol) form of the latter'.

12

14.1. Introduction: Principles of Catalysis 14.1.2. Basic Processes in Molecular Catalysis 14.1-2.4. Catalysis of Electrophilic Reactions by Proton Loss

A simple illustration of this is the catalysis by Cu2+ and other metal ions of the oxidation of H,, according to the mechanism depicted by Cu2+

+ H, r [CuH]+ + H +

(a)

In this case the catalytic role of Cu2+ can be interpreted in terms of enhancement of the acidity of H, and increasing the effective concentration of the reactive conjugate base, i.e., the H- ion, by stabilization of the latter through coordination (see Section 14.1.1)2,3. Other metal ions which catalyze the oxidation of H, by analogous mechanisms include Ag(I), Cu(I), Hg(II), Rh(III), and Ru(II1). A complete description of the mechanisms of such reactions encompasses the recognition (1) that coordination of H- to the metal ion involves the displacement of one of the original ligands, as in4: [RuC1,I3-

+ H,

[RuHC1,I3-

+ H+ + C1-

and (2) that in addition to the metal ion, such catalyst systems must involve an appropriate base to stabilize the released proton. The role of this base may be assumed by a solvent molecule or, in certain cases, by the displaced ligand the mechanism in such cases being of the form: M-H

+ H2

M --- X I

I

H---- H+

-

M-H-+

X-H+

(c)

The variation of rate with the ligand X, in such cases, reflects the expected direct dependence on the basicity of X, and inverse dependence on the stability of the M-X bond5. Other saturated bonds that can be cleaved heterolytically by electrophilic attack by metal ions include C-H bonds, as in the mercuration of benzene:

and strained C-C

bonds as in the Ag(1)-catalyzed rearrangement of cubane?

Metal ions, notably those with filled or nearly filled d-subshells, such as Ag(I), Hg(II), Cu(II), Pd(II), etc., often are more effective electrophiles than H + for these reactions. This can be attributed to the enhancement of the electrophilic interaction of the metal ion by back-donation of d,-electrons into the antibonding orbitals of the ubond (i.e,, H-H, C-H or C-C), both interactions contributing to dissociation of the latter. Other examples of electrophilic substitution reactions catalyzed by metal ions through mechanisms involving the promotion of proton loss and stabilization of enol forms of reactants include the bromination of P-diketones:

14.1. Introduction: Principles of Catalysis 14.1.2. Basic Processes in Molecular Catalysis 14.1.2.5. Oxidative Addition/Reductive Elimination Reactions

515; CH3CCH2CCH3 + Br,

Pi I

I

MZ+

0

1I

13

0

II

CH3CCHBrCCH3

+ H+ +

Br-

(f)

2+

+ Br-+

H+ +

H+

and of ketoesters. The order of catalytic activities in the case of reaction (f) (Cuz+ > Ni2+ > Zn2+ > MnZ+ > Ca") parallels the order of the stabilities of the acetylacetonate complexes of the metal ions',7. (J. HALPERN)

1. J. Halpem, Pure Appl. Chem., 20, 59 (1969), and references therein. 2. J. Halpem, E. R. Macgregor, E. Peters, J . Phys. Chem., 60, 1455 (1956). 3. J. Halpem, Ann. Revs. Phys. Chem., 16, 103 (1965) and references therein. 4. J. Halpem, B. R. James, Can. J . Chem., 44,671 (1966). 5. J. Halpem, J. B. Milne, Proc. 2nd Int. Cong. Catal., Vol. 1, Ed. Technip, Paris, 1960, p. 445. 6. J. Halpem, in Organic Syntheses via Metal Carbonyls, Vol. 11, I. Wender, P. Pino, eds., John

Wiley, New York, 1977, p. 705. 7. K. Pederson, Acra Chem. Scand., 2,252 (1948).

14.1.2.5. Oxidative Addition/Reductive Elimination Reactions

I

Oxidative-addition reactions (see Section 10) play a widespread role in catalytic processes and constitute one of the principal routes for the activation and dissociation of saturated bonds, including H-H, C-X (X = halogen), strained C-C bonds, and in certain cases, C-H bonds. Correspondingly, the reverse process, i.e., reductive elimination, constitutes a widespread route for the formation of such bonds. The latter process frequently is the product-forming step in catalytic reactions such as hydrogenation, carbonylation and hydr~formylation'-~. Oxidative-addition reactions have their origin in the inverse dependence of coordination number on d-electron population that characterizes low-spin transition metal complexes, notably those with nearly filled (d6-d") d-subshells, e.g., [CO(III)(CN),]~(d6); [Co(I1)(CN),l3- (d7); [Ni(II)(CN),I2- (d'); [Ag(I)(CN),]- (d"); etc. One of the consequences of this trend is the oxidations of such complexes, which tend to be accompanied by increases in the preferred coordination numbers of the metal atoms and hence by the incorporation of additional ligands into their coordination shells'. The ligands required to complete the coordination shells may in certain cases be derived from the oxidant itself and, indeed, such complexes are especially effective as reductants for molecular oxidants, which typically undergo dissociative reduction to yield anionic ligands, i.e., X-Y + eX' Y-;X-Y 2eXY-. Important classes of oxidative addition-reductive elimination reactions are exemplified by

-

+

+

-

+


14.1. Introduction: Principles of Catalysis 14.1.2. Basic Processes in Molecular Catalysis 14.1.2.5. Oxidative Addition/Reductive Elimination Reactions

515; CH3CCH2CCH3 + Br,

Pi I

I

MZ+

0

1I

13

0

II

CH3CCHBrCCH3

+ H+ +

Br-

(f)

2+

+ Br-+

H+ +

H+

and of ketoesters. The order of catalytic activities in the case of reaction (f) (Cuz+ > Ni2+ > Zn2+ > MnZ+ > Ca") parallels the order of the stabilities of the acetylacetonate complexes of the metal ions',7. (J. HALPERN)

1. J. Halpem, Pure Appl. Chem., 20, 59 (1969), and references therein. 2. J. Halpem, E. R. Macgregor, E. Peters, J . Phys. Chem., 60, 1455 (1956). 3. J. Halpem, Ann. Revs. Phys. Chem., 16, 103 (1965) and references therein. 4. J. Halpem, B. R. James, Can. J . Chem., 44,671 (1966). 5. J. Halpem, J. B. Milne, Proc. 2nd Int. Cong. Catal., Vol. 1, Ed. Technip, Paris, 1960, p. 445. 6. J. Halpem, in Organic Syntheses via Metal Carbonyls, Vol. 11, I. Wender, P. Pino, eds., John

Wiley, New York, 1977, p. 705. 7. K. Pederson, Acra Chem. Scand., 2,252 (1948).

14.1.2.5. Oxidative Addition/Reductive Elimination Reactions

I

Oxidative-addition reactions (see Section 10) play a widespread role in catalytic processes and constitute one of the principal routes for the activation and dissociation of saturated bonds, including H-H, C-X (X = halogen), strained C-C bonds, and in certain cases, C-H bonds. Correspondingly, the reverse process, i.e., reductive elimination, constitutes a widespread route for the formation of such bonds. The latter process frequently is the product-forming step in catalytic reactions such as hydrogenation, carbonylation and hydr~formylation'-~. Oxidative-addition reactions have their origin in the inverse dependence of coordination number on d-electron population that characterizes low-spin transition metal complexes, notably those with nearly filled (d6-d") d-subshells, e.g., [CO(III)(CN),]~(d6); [Co(I1)(CN),l3- (d7); [Ni(II)(CN),I2- (d'); [Ag(I)(CN),]- (d"); etc. One of the consequences of this trend is the oxidations of such complexes, which tend to be accompanied by increases in the preferred coordination numbers of the metal atoms and hence by the incorporation of additional ligands into their coordination shells'. The ligands required to complete the coordination shells may in certain cases be derived from the oxidant itself and, indeed, such complexes are especially effective as reductants for molecular oxidants, which typically undergo dissociative reduction to yield anionic ligands, i.e., X-Y + eX' Y-;X-Y 2eXY-. Important classes of oxidative addition-reductive elimination reactions are exemplified by

-

+

+

-

+

14

14.1.2. Basic Processes in Molecular Catalysis 14.1.2.5. Oxidative Addition/Reductive Elimination Reactions 14.1.2.5.1. One-Electron Oxidative Addition'**.

+ j + H,

~[CO(II)(CN),]~- H, [Rh(I)Cl(PPh,),]

2[Co(III)H(CN),I3-

(a)

[Rh(III)H,Cl(PPh3)3]

(b) (J. HALPERN)

1 . J. Halpem, Acct. Chem. Res., 3, 386 (1970) and references therein. 2. A. Nakamura, M. Tsutsui, Principles and Applications of Homogeneous Catalysis, John Wiley, New York, 1980, p. 64. 3. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G . Finke, Principles and Applications of Organotransition Metal Chemistry, Ch. 8 , University Science Books, Mill Valley, CA, 1987. 4. G . W. Parshall, S . D. Ittel, Homogeneous Catalysis, 2nd. ed., John Wiley, New York, 1992. 14.1.2.5.1. One-Electron Oxidative AdditionlJ.

This class of reaction is manifested especially by (typically five-coordinated) lowspin d7 cobalt(I1) complexes, such as [CO(II)(CN),]~- , [Co(II)(DMGH),B] (DMGH, = dimethylglyoxime, B = axial ligand such as pyridine, triphenylphosphine, etc.) and cob(I1)alamin (vitamin B Important examples include the oxidative addition of organic halides leading to the formation of organometallic compounds by free-radical mechanisms:

--

+ CH31 k\ fast [CO(II)(CN),]~++ CH3' [CO(II)(CN,]~-

2[Co(II)(CN),I3-

+ CH31

[CO(III)(CN),I]~-

+ CH;

[CO(II)(CN,CH,]~+ [CO(III)(CN),I]~-

+

(a) (b)

[CO(III)(CN),CH,]~-

(c)

which exhibits the rate-law, ~ [ C O ( C N )-][CH3II3. ,~ Other molecules, for example HO-OH, NH,-OH and I-CN, oxidatively add to low-spin cobalt(I1) complexes by analogous free-radical mechanisms4. Reactions (a) and (b) are also examples of innersphere electron transfer reactions. Thus in the oxidative addition of p-substituted benzyl bromide (p-X-c,H,CH,-Br) to [Co(II)(DMGH),P(p-Y-C,H,),I, the rate of the initial halogen abstraction step corresponding to

-

[CO(II)(DMGH),P(~-Y-C,H~)~] + p-X-c6H4CH,-Br [CO(III)B~(DMGH),P(~-Y-C~H~)~] + p-X-C6H,CH,'

(d)

increases with the electron-withdrawing ability of the substituent X and with the electrondonating ability of the substituent Y5. Such oxidative-addition reactions find catalytic applications, e.g., in the [CO(CN),]~ --catalyzed reduction of organic halides by [CoH(CN),I3-, which proceeds by the free-radical chain mechanism: [Co(CN),I3-

+ RX -+

[CoX(CN),I3-

-

R'

+ [CoH(CN),I3-

-+ RH

RX

+

[Co(cN),l3-

[CoH(CN),13-

+ R'

+ [Co(CN),I3RH

+ [CoX(CN),I3-

(el (f)

(€9

In contrast to the scheme of Eqs. (a) to (c) the oxidative addition of H, to [CO(CN),]~- :


14

14.1.2. Basic Processes in Molecular Catalysis 14.1.2.5. Oxidative Addition/Reductive Elimination Reactions 14.1.2.5.1. One-Electron Oxidative Addition'**.

+ j + H,

~[CO(II)(CN),]~- H, [Rh(I)Cl(PPh,),]

2[Co(III)H(CN),I3-

(a)

[Rh(III)H,Cl(PPh3)3]

(b) (J. HALPERN)

1 . J. Halpem, Acct. Chem. Res., 3, 386 (1970) and references therein. 2. A. Nakamura, M. Tsutsui, Principles and Applications of Homogeneous Catalysis, John Wiley, New York, 1980, p. 64. 3. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G . Finke, Principles and Applications of Organotransition Metal Chemistry, Ch. 8 , University Science Books, Mill Valley, CA, 1987. 4. G . W. Parshall, S . D. Ittel, Homogeneous Catalysis, 2nd. ed., John Wiley, New York, 1992. 14.1.2.5.1. One-Electron Oxidative AdditionlJ.

This class of reaction is manifested especially by (typically five-coordinated) lowspin d7 cobalt(I1) complexes, such as [CO(II)(CN),]~- , [Co(II)(DMGH),B] (DMGH, = dimethylglyoxime, B = axial ligand such as pyridine, triphenylphosphine, etc.) and cob(I1)alamin (vitamin B Important examples include the oxidative addition of organic halides leading to the formation of organometallic compounds by free-radical mechanisms:

--

+ CH31 k\ fast [CO(II)(CN),]~++ CH3' [CO(II)(CN,]~-

2[Co(II)(CN),I3-

+ CH31

[CO(III)(CN),I]~-

+ CH;

[CO(II)(CN,CH,]~+ [CO(III)(CN),I]~-

+

(a) (b)

[CO(III)(CN),CH,]~-

(c)

which exhibits the rate-law, ~ [ C O ( C N )-][CH3II3. ,~ Other molecules, for example HO-OH, NH,-OH and I-CN, oxidatively add to low-spin cobalt(I1) complexes by analogous free-radical mechanisms4. Reactions (a) and (b) are also examples of innersphere electron transfer reactions. Thus in the oxidative addition of p-substituted benzyl bromide (p-X-c,H,CH,-Br) to [Co(II)(DMGH),P(p-Y-C,H,),I, the rate of the initial halogen abstraction step corresponding to

-

[CO(II)(DMGH),P(~-Y-C,H~)~] + p-X-c6H4CH,-Br [CO(III)B~(DMGH),P(~-Y-C~H~)~] + p-X-C6H,CH,'

(d)

increases with the electron-withdrawing ability of the substituent X and with the electrondonating ability of the substituent Y5. Such oxidative-addition reactions find catalytic applications, e.g., in the [CO(CN),]~ --catalyzed reduction of organic halides by [CoH(CN),I3-, which proceeds by the free-radical chain mechanism: [Co(CN),I3-

+ RX -+

[CoX(CN),I3-

-

R'

+ [CoH(CN),I3-

-+ RH

RX

+

[Co(cN),l3-

[CoH(CN),13-

+ R'

+ [Co(CN),I3RH

+ [CoX(CN),I3-

(el (f)

(€9

In contrast to the scheme of Eqs. (a) to (c) the oxidative addition of H, to [CO(CN),]~- :

14.1 -2. Basic Processes in Molecular Catalysis 14.1-2.5. Oxidative Addition/Reductive Elimination Reactions 14.1 -2.5.2. Two-Electron Oxidative Addition.

-

+

~[CO(CN),]~- H, -+ [(NC),CO.*.H***H*.*CO(CN),]~ 2[CoH(CN),I3-

15

(h)

-I2, implying a concerted mechanism in exhibits a third-order rate law, ~[H,][CO(CN),~ which the driving force for dissociation of the H, bond (ca. 430 kJ/mol) is provided by the simultaneous formation of two 240 kJ/mol) Co-H bonds6. In this case the initial step of a stepwise mechanism analogous to that of equations (a) and (b) would be prohibitively endothermic (ca. 190 kJ/mol). Reaction (h) constitutes an important route for the activation of H,, which finds catalytic application in the [Co(CN),I3 --catalyzed hydrogenation of substrates such as butadiene and styrene7s8. Interest in the study of the oxidative addition reactions of low-spin cobalt complexes is enhanced by parallels with corresponding reactions of vitamin B derivatives and hence by their possible relevance as B ,,-model system^^^'^. Analogous oxidative-addition reactions are exhibited by certain complexes of the early transition metals, notably chromium(I1) (high spin d4), as exemplified by .'lvl

+ C6H,CH,Br

c?'(aq)

-4- C6H5CH,'

c?'(aq) 2C?+ (aq)

+ C6H,CH,Br

-+

[Cr(III)(H,0),Br]2+ -4- C6H5CH2'

(i)

[Cr(II)(H,0),CH,C,H,]2+

(j)

[Cr(III)(H,O),Br]Zf 4- [Cr(III)(H,0),CH,C6H5]2f

(k)

(J. HALPERN) 1 . J. Halpern, Acct. Chem. Res., 3, 386 (1970).

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

J. Halpern, Pine Appl. Chem., 51,2171 (1971). P. B. Chock, J. Halpern, J . Am. Chem. SOC., 91, 582 (1969). P. B. Chock, R. B. K. Dewar, J. Halpern, L. Y. Wong, J . Am. Chem. Soc., 91, 82 (1969). J. Halpern, P. F. Phelan, J . Am. Chem. Soc., 91, 77 (1969). J. Halpern, M. Pribanic, Inorg. Chem., 9, 2616 (1970). J. Kwiatek, Cafal.Rev., I , 37 (1967). J. Kwiatek, J. K. Seyler, Adv. Chem. Series, 70,207 (1968). J. Halpern, Bull. Chem. SOC.Jn., 61, 13 (1988). J. Halpern, in Vitamin B,,, D. Dolphin, ed., Vol. 1, John Wiley, New York, 1981, p. 501. C. E. Castro, W. C. Kray, J . Am. Chem. SOC., 85, 2768 (1963). J. K. Kochi, D. D. Davis, J . Am. Chem. SOC.,86,2649 (1964).

14.1.2.5.2. Two-Electron Oxidative Addition.

A widespread class of such reactions is that exhibited by square-planar, four-coordinate low-spin d8-complexes, which oxidatively add molecules to form six-coordinated d6-adducts '9':

[Pt(II)C14]Z-

+ c1,

-

[Pt(IV)C16]2-

(a) The scope of such reactions is extended to include the oxidative addition of other molecules such as H, and organic halides, e.g.: Rh(I)Cl(PPh,), Ir(I)I(CO)(PPh,),

+ H,

+ CH,I

-+

Rh(III)H,Cl(PPh,),

(b)

Ir(III)I,(CH,)(CO)(PPh,),

(c)


14.1 -2. Basic Processes in Molecular Catalysis 14.1-2.5. Oxidative Addition/Reductive Elimination Reactions 14.1 -2.5.2. Two-Electron Oxidative Addition.

-

+

~[CO(CN),]~- H, -+ [(NC),CO.*.H***H*.*CO(CN),]~ 2[CoH(CN),I3-

15

(h)

-I2, implying a concerted mechanism in exhibits a third-order rate law, ~[H,][CO(CN),~ which the driving force for dissociation of the H, bond (ca. 430 kJ/mol) is provided by the simultaneous formation of two 240 kJ/mol) Co-H bonds6. In this case the initial step of a stepwise mechanism analogous to that of equations (a) and (b) would be prohibitively endothermic (ca. 190 kJ/mol). Reaction (h) constitutes an important route for the activation of H,, which finds catalytic application in the [Co(CN),I3 --catalyzed hydrogenation of substrates such as butadiene and styrene7s8. Interest in the study of the oxidative addition reactions of low-spin cobalt complexes is enhanced by parallels with corresponding reactions of vitamin B derivatives and hence by their possible relevance as B ,,-model system^^^'^. Analogous oxidative-addition reactions are exhibited by certain complexes of the early transition metals, notably chromium(I1) (high spin d4), as exemplified by .'lvl

+ C6H,CH,Br

c?'(aq)

-4- C6H5CH,'

c?'(aq) 2C?+ (aq)

+ C6H,CH,Br

-+

[Cr(III)(H,0),Br]2+ -4- C6H5CH2'

(i)

[Cr(II)(H,0),CH,C,H,]2+

(j)

[Cr(III)(H,O),Br]Zf 4- [Cr(III)(H,0),CH,C6H5]2f

(k)

(J. HALPERN) 1 . J. Halpern, Acct. Chem. Res., 3, 386 (1970).

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

J. Halpern, Pine Appl. Chem., 51,2171 (1971). P. B. Chock, J. Halpern, J . Am. Chem. SOC., 91, 582 (1969). P. B. Chock, R. B. K. Dewar, J. Halpern, L. Y. Wong, J . Am. Chem. Soc., 91, 82 (1969). J. Halpern, P. F. Phelan, J . Am. Chem. Soc., 91, 77 (1969). J. Halpern, M. Pribanic, Inorg. Chem., 9, 2616 (1970). J. Kwiatek, Cafal.Rev., I , 37 (1967). J. Kwiatek, J. K. Seyler, Adv. Chem. Series, 70,207 (1968). J. Halpern, Bull. Chem. SOC.Jn., 61, 13 (1988). J. Halpern, in Vitamin B,,, D. Dolphin, ed., Vol. 1, John Wiley, New York, 1981, p. 501. C. E. Castro, W. C. Kray, J . Am. Chem. SOC., 85, 2768 (1963). J. K. Kochi, D. D. Davis, J . Am. Chem. SOC.,86,2649 (1964).

14.1.2.5.2. Two-Electron Oxidative Addition.

A widespread class of such reactions is that exhibited by square-planar, four-coordinate low-spin d8-complexes, which oxidatively add molecules to form six-coordinated d6-adducts '9':

[Pt(II)C14]Z-

+ c1,

-

[Pt(IV)C16]2-

(a) The scope of such reactions is extended to include the oxidative addition of other molecules such as H, and organic halides, e.g.: Rh(I)Cl(PPh,), Ir(I)I(CO)(PPh,),

+ H,

+ CH,I

-+

Rh(III)H,Cl(PPh,),

(b)

Ir(III)I,(CH,)(CO)(PPh,),

(c)

16

14.1.2. Basic Processes in Molecular Catalysis 14.1-2.5. Oxidative Addition/Reductive Elimination Reactions 14.1.2.5.2. Two-Electron Oxidative Addition.

Such oxidative-addition reactions play a widespread role in homogeneous catalytic processes and constitute an important pathway of activation and dissociation of H-H, C-C and C-H bonds. Correspondingly, the reverse reductive-elimination reactions represent an important process for the formation of such bonds, and thus frequently constitute the product-forming steps in catalytic cycles'.'. Available evidence' supports the following schemes for the oxidative additions of H, (as well as other non-polar molecules) and organic halides (RX), respectively, to square planar ds-complexes such as those of Rh(1) and Ir(I), [e.g., Ir(I)C1(CO)(PPh3),]:

q;

I

L

"M..,H /'H

I -

[L4M...R...x] s+

J%H

(d)

L

L

- 7%'

(el

L

X Accordingly, the rate of oxidative addition of H, is relatively insensitive to electronic factors, notably the electron-donating power of L, whereas the rates of oxidative addition of organic halides increase markedly with the electron donating ability of L and (in view of the polar transition state) with solvent polarity. The different stereochemistries of oxidative addition, i.e., cis for Hz vs. trans for RX, also reflect this mechanistic distinction. The influence of variation of the metal on the rate of oxidative addition is less well documented and understood, but available data suggest the reactivity sequence, Ir(I)(5ds) > Co(I)(3d8) > Rh(I)(4d8)3. Other molecules which undergo oxidative addition to ds-complexes include HX (X = C1, Br, I), R,Si-H, and R,Sn-Cl as well as molecules containing strained C-C bonds. A stoichiometric example of the latter is depicted by

Such oxidative addition reactions are key steps in Rh(1)-catalyzed rearrangements of strained-ring compounds such as4*5

Ligands containing appropriately reactive bonds may undergo intramolecular oxidative-addition reactions, the driving force for which is enhanced by proximity and chelation effects. A particularly striking and important example is provided by the metallation of the aromatic rings in ligands such as PPh,, e.g.6*7

14.1.2. Basic Processes in Molecular Catalysis 14.1.2.5. Oxidative Addition/Reductive Elimination Reactions 14.1.2.5.2. Two-Electron Oxidative Addition.

-

17

Five-coordinated, low-spin, d8-complexes also undergo oxidative-addition reactions such as protonation: [Mn(CO),]nucleophilic displacement: [Mn(CO),]or oxidative elimination:

+ CH,I

+ R,SiH

IrCl(CO)PR,),

+ H+

HMn(CO),

(9

+ I-

CH,Mn(CO),

IrHCl(SiR,)(CO)(PR,),

(j)

+ PR,

(k)

Anionic metal complexes such as those of Co(1) (including vitamin BlZsoften exhibit high nucleophilic reactivities in reactions such as equation (j) and have been termed supernucleophiless. Protonation constitutes an important route for formation of metal hydrides, reflected in the roles of acids as cocatalysts, e.g., in the Rh(1)-catalyzed isomerization and dimerization of olefinsg. The d"-complexes of Ni(O), Pd(O), and Pt(0) exhibit patterns of oxidative-addition reactions analogous to those of low-spin, d8-complexes, e,g.'o-i2 Pt(O)(PEt,), Pt(O)(PPh,),

-

+ H, -+Pt(II)H,(PEt,),

+ CH,I

+ PPh,

Pt(II)I(CH,)(PPh,),

(1) (m)

Analogous patterns of oxidative-addition reactions also are observed for some early transition metal complexes, notably those of d2- and d4-configuration'3*'4 Cp,Ti(II) (d2) Cp,Mo(II) (d4)

+ H,

+ H,

Cp,Ti(IV)H, (n)

(do) Cp,Mo(IV)H, (d2)

(0)

Such reactions account for the [CpNbH,]-catalyzed isotopic exchange of C6H6 with D, according to the mechanistic scheme (Cp = C,H,)8,'5 Cp,Nb(V)H,

- H2

Cp,Nb(III)H

D2 C6H6

Cp,Nb(V)HD,

- HD

Cp,Nb(III)D (P)

Cp,Nb(V)HD(CsH,)

Whereas there are numerous examples of the oxidative addition of aromatic C-H bonds to metal complexes [e.g., equations (h) and (p)] attempts to activate paraffinic hydrocarbons by C-H oxidative addition are generally unsuccessful. Since the reverse

18

14.1.2. Basic Processes in Molecular Catalysis 14.1.2.5. Oxidative Addition/Reductive Elimination Reactions 14.1.2.5.3. Free Radical Chain Mechanism of Oxidative Addition.

reaction, i.e., reductive elimination of alkanes from cis-hydroalkymetal complexes, generally is fast, e.g. c~s-FV(II)H(CH~)(PP~~)~ I CH,

+ Pt(0)(PPh3),

the barrier to such oxidative additions would appear to be thermodynamic rather than kineticI6. Thus, the equilibria for reactions such as (q) generally appear to lie far to the right reflecting the characteristic weakness of transition metal-alkyll o-bonds. (J. HALPERN) 1 . J. Halpem, Acct. Chem. Res., 3, 386 (1970), and reference therein. 2. J. P. Collman, L. S. Hegedus, J. Norton, R. G. Finke, Principles and Applications or Organotransition Metal Chemistry, Ch. 5, University Science Books, Mill Valley, CA, 1987, p. 176. 3. L. Vaska, L. S. Chen, W. V. Miller, J. Am. Chem. SOC., 93, 6671 (1971). 4. L. Cassar, P. E. Eaton, J. Halpem, J . Am. Chem. SOC.,92, 3515 (1970). 5. J. Halpem, in Organic Syntheses via Metal Carbonyls, Vol. 11, I. Wender, P. Pino, eds., John Wiley, New York, 1977, p. 705. 6. M. A. Bennett, D. L. Miller, J. Chem. SOC., Chem. Commun., 581 (1967). J . Am. Chem. Soc., 91,6983 (1969). 7. G. W. Parshall, Acct. Chem. Res., 8, 113 (1975). 8. G. N. Schrauzer, E. Deutsch, J . Am. Chem. SOC., 91,3341 (1969). 9. R. Crarner, Acct. Chem. Res., 1, 186 (1968). 10. J. P. Birk, J. Halpem, A. L. Pickard, J. Am. Chem. SOC.,90, 4491 (1968). 11. D. H. Gerlach, A. R. Kane, G. W. Parshall, J. P. Jesson, E. L. Muetterties, J . Am. Chem. Soc., 93,3543 (1971). 12. T. Yoshida, S. Otsuka, J. Am. Chem. Soc., 99, 2134 (1977). 13. I. E. Bercaw, R. H. Marvich, L. G. Bell, H. H. Brintzinger,J. Am. Chem. Soc., 94,1219 (1972). 14. J. L. Thomas,J. Am. Chem. Soc., 95, 1838 (1973). 15. F. N. Tebbe, G. W. Parshall, J. Am. Chem. SOC.,93, 3793 (1971). 16. L. Abis, A. Sen, J. Halpem, J. Am. Chem. SOC.,100,2915 (1978). 14.1.2.5.3. Free Radical Chain Mechanism of Oxidative Addition.

In certain cases, notably with secondary alkyl halides, for which direct oxidative addition [is., according to equation (e), Section 14.1.2.5.21 is disfavored for steric reasons, the same overall reaction may be achieved by a free radical chain mechanism: Initiation:

Propagation:

+ Ir(1) R-Ir(I1) + RX R'

Overall Reaction: Ir(1)

--

+ RX

R-Ir(1I) R-Ir(II1)X

+ R'

R-Ir(II1)X

(4

Such a mechanism has been identified for the oxidative addition of cyclo-C,H, ,Br to [IrCl(CO)(PPh,),] I. Certain oxidative addition reactions of binuclear complexes also proceed through free radical chain mechanisms, e.g.':


18

14.1.2. Basic Processes in Molecular Catalysis 14.1.2.5. Oxidative Addition/Reductive Elimination Reactions 14.1.2.5.3. Free Radical Chain Mechanism of Oxidative Addition.

reaction, i.e., reductive elimination of alkanes from cis-hydroalkymetal complexes, generally is fast, e.g. c~s-FV(II)H(CH~)(PP~~)~ I CH,

+ Pt(0)(PPh3),

the barrier to such oxidative additions would appear to be thermodynamic rather than kineticI6. Thus, the equilibria for reactions such as (q) generally appear to lie far to the right reflecting the characteristic weakness of transition metal-alkyll o-bonds. (J. HALPERN) 1 . J. Halpem, Acct. Chem. Res., 3, 386 (1970), and reference therein. 2. J. P. Collman, L. S. Hegedus, J. Norton, R. G. Finke, Principles and Applications or Organotransition Metal Chemistry, Ch. 5, University Science Books, Mill Valley, CA, 1987, p. 176. 3. L. Vaska, L. S. Chen, W. V. Miller, J. Am. Chem. SOC., 93, 6671 (1971). 4. L. Cassar, P. E. Eaton, J. Halpem, J . Am. Chem. SOC.,92, 3515 (1970). 5. J. Halpem, in Organic Syntheses via Metal Carbonyls, Vol. 11, I. Wender, P. Pino, eds., John Wiley, New York, 1977, p. 705. 6. M. A. Bennett, D. L. Miller, J. Chem. SOC., Chem. Commun., 581 (1967). J . Am. Chem. Soc., 91,6983 (1969). 7. G. W. Parshall, Acct. Chem. Res., 8, 113 (1975). 8. G. N. Schrauzer, E. Deutsch, J . Am. Chem. SOC., 91,3341 (1969). 9. R. Crarner, Acct. Chem. Res., 1, 186 (1968). 10. J. P. Birk, J. Halpem, A. L. Pickard, J. Am. Chem. SOC.,90, 4491 (1968). 11. D. H. Gerlach, A. R. Kane, G. W. Parshall, J. P. Jesson, E. L. Muetterties, J . Am. Chem. Soc., 93,3543 (1971). 12. T. Yoshida, S. Otsuka, J. Am. Chem. Soc., 99, 2134 (1977). 13. I. E. Bercaw, R. H. Marvich, L. G. Bell, H. H. Brintzinger,J. Am. Chem. Soc., 94,1219 (1972). 14. J. L. Thomas,J. Am. Chem. Soc., 95, 1838 (1973). 15. F. N. Tebbe, G. W. Parshall, J. Am. Chem. SOC.,93, 3793 (1971). 16. L. Abis, A. Sen, J. Halpem, J. Am. Chem. SOC.,100,2915 (1978). 14.1.2.5.3. Free Radical Chain Mechanism of Oxidative Addition.

In certain cases, notably with secondary alkyl halides, for which direct oxidative addition [is., according to equation (e), Section 14.1.2.5.21 is disfavored for steric reasons, the same overall reaction may be achieved by a free radical chain mechanism: Initiation:

Propagation:

+ Ir(1) R-Ir(I1) + RX R'

Overall Reaction: Ir(1)

--

+ RX

R-Ir(1I) R-Ir(II1)X

+ R'

R-Ir(II1)X

(4

Such a mechanism has been identified for the oxidative addition of cyclo-C,H, ,Br to [IrCl(CO)(PPh,),] I. Certain oxidative addition reactions of binuclear complexes also proceed through free radical chain mechanisms, e.g.':

-

14.1. Introduction: Principles of Catalysis 14.1.2. Basic Processes in Molecular Catalysis 14.1-2.6. Insertion Reactions

Rh,(OEP),

+ RX

RRh(0EP)

19

+ XRh(0EP)

(el

(OEP = octaethylporphyrin, RX = C,H,CH,Br). (J. HALPERN) 1. J. S. Bradley, D. E. Connor, D. H. Dolphin, J. A. Labinger, J. A. Osbom, J. Am. Chem. SOC.,

94, 4043 (1972). 2. R. S. Paonessa, N. C.Thomas, J. Halpem, J. Am. Chem. SOC.,107,4333 (1985).

14.1.2.6. Insertion Reactions The term insertion reactions (see Section 11) designates a rather broad collection of processes in which the overall reaction corresponds to insertion of an unsaturated molecule (X) into a metal-ligand (M-Y) bond’. Variants of processes encompassed by this designation are depicted schematically by X L

I

M-Y M-Y M-X

+X +Y

--

M-X-Y

(-

L

I

M-X-Y)

(a)

M-X-Y

(b)

M-X-Y

(c) Familiar combinations include Y = H, X = olefin, acetylene, or diene; Y = alkyl, X = CO; Y = OH or OR, X = CO or olefin. Such reactions constitute important routes for addition to unsaturated molecules, and thus play a widespread role in catalytic reaction processes such as hydrogenation, hydrosilylation, and hydrocyanation of olefins, carbonylation, etc? Many diverse mechanisms are identified for insertion reactions, some of which are depicted below. Concerted Inserted: CH -CH, HPt(SnCI,)(PEt,J,

3

‘T

HPt(SnCI,)(PEt,), 0

-

C,H,Pt(SnCI,)(PEt,)z 0

II

(d)

It

CH,Mn(CO), -+ C H , C M n ( C O ) 4 L CH,CMn(CO),L (e) This route normally requires coordination of the inserting molecule (CO, CzH4, etc.) to the metal atom and, hence, the availability of a vacant or accessible coordination site3*,. Many catalytic hydrogenation and carbonylation reactions proceed through mechanisms encompassing such steps. External Nucleophilic Attack: Hg’+

CHR=CHR

[Hg(CHR-CHR)]’+

rrans-addition H20

0

[PtCI(CO)(PEt,),]+

[Ag OH CHR-CHR

I

II + ROH IPtCl(COR)(PEt3)2 + H+

+ H+

(f)

(g)


-

14.1. Introduction: Principles of Catalysis 14.1.2. Basic Processes in Molecular Catalysis 14.1-2.6. Insertion Reactions

Rh,(OEP),

+ RX

RRh(0EP)

19

+ XRh(0EP)

(el

(OEP = octaethylporphyrin, RX = C,H,CH,Br). (J. HALPERN) 1. J. S. Bradley, D. E. Connor, D. H. Dolphin, J. A. Labinger, J. A. Osbom, J. Am. Chem. SOC.,

94, 4043 (1972). 2. R. S. Paonessa, N. C.Thomas, J. Halpem, J. Am. Chem. SOC.,107,4333 (1985).

14.1.2.6. Insertion Reactions The term insertion reactions (see Section 11) designates a rather broad collection of processes in which the overall reaction corresponds to insertion of an unsaturated molecule (X) into a metal-ligand (M-Y) bond’. Variants of processes encompassed by this designation are depicted schematically by X L

I

M-Y M-Y M-X

+X +Y

--

M-X-Y

(-

L

I

M-X-Y)

(a)

M-X-Y

(b)

M-X-Y

(c) Familiar combinations include Y = H, X = olefin, acetylene, or diene; Y = alkyl, X = CO; Y = OH or OR, X = CO or olefin. Such reactions constitute important routes for addition to unsaturated molecules, and thus play a widespread role in catalytic reaction processes such as hydrogenation, hydrosilylation, and hydrocyanation of olefins, carbonylation, etc? Many diverse mechanisms are identified for insertion reactions, some of which are depicted below. Concerted Inserted: CH -CH, HPt(SnCI,)(PEt,J,

3

‘T

HPt(SnCI,)(PEt,), 0

-

C,H,Pt(SnCI,)(PEt,)z 0

II

(d)

It

CH,Mn(CO), -+ C H , C M n ( C O ) 4 L CH,CMn(CO),L (e) This route normally requires coordination of the inserting molecule (CO, CzH4, etc.) to the metal atom and, hence, the availability of a vacant or accessible coordination site3*,. Many catalytic hydrogenation and carbonylation reactions proceed through mechanisms encompassing such steps. External Nucleophilic Attack: Hg’+

CHR=CHR

[Hg(CHR-CHR)]’+

rrans-addition H20

0

[PtCI(CO)(PEt,),]+

[Ag OH CHR-CHR

I

II + ROH IPtCl(COR)(PEt3)2 + H+

+ H+

(f)

(g)

20

14.1. Introduction: Principles of Catalysis 14.1-2. Basic Processes in Molecular Catalysis 14.1.2.6. Insertion Reactions

Such reactions constitute important steps in the oxidation of CO and olefins by metal ions such as Hg(II), Tl(III), and Pd(II)6-8. Free Radical Mechanisms:

-

This mechanism, exemplified by (where X = CN, Ph, COOR, etc.) [HCo(CN),I3-

CH -CHX

CH3kHX

+ ~ o ( C N ) , ~-+ ["if$-Co(CN),]-

(h)

reflects the characteristic weakness of metal-hydrogen bonds (typically ca. 250 kJ/mol in view of which the endothermicity of H-atom transfer from metal hydrides to activated olefins such as styrene is relatively low (ca. 80 kJ/m~l)~-".Unlike concerted olefin insertion [equation (d)] this mechanism does not require a vacant coordination site and, thus, is favored for dissociatively inert metal hydrides such as [HMn(CO),] and [HCo(CN),I3 -. On the other hand, the energetics of such H-atom transfer are unfavorable unless the resulting radical is stabilized by a conjugating substituent such as CN, C,H, or COOR. Thus, the mechanism tends to be restricted to conjugated olefins and a,&unsaturated compounds and does not extend to isolated double bonds. Such free radical mechanisms of hydrogen transfer have been proposed as steps in the [HCo(CN),I3 --catalyzed hydrogenation of styrene" and in the [HCo(CO),]-catalyzed hydrogenation of arenes13and hydroformylation of styrene'. (J. HALPERN) 1. 2. 3. 4.

5. 6. 7. 8. 9. 10. 11. 12. 13.

R. F. Heck,Adv. Chem. Ser., 49, 181 (1965). A. Wojcicki, Adv. Organomet. Chem., I I , 88 (1972). F. Calderazzo, Angew. Chem., Int. Ed., 16,299 (1977). J. P. Collman, L. S. Hegedus, J. Norton, R. G . Finke, Principles and Applications ofOrganotransition Metal Chemistry, Ch. 6, University Science Books, Mill Valley, CA, 1987, p. 259, and references therein. H. C. Clark, C. Jablonski, J. Halpem, A. Mantovani,T. A. Wei1,Inorg. Chem.,13,1541 (1974). J. Halpem, H. B. Tinker, J. Am. Chem. SOC.,89,6427 (1967). J. E. Byrd, J. Halpem, J. Am. Chem. SOC.,93, 1634 (1971). J. Halpem, Comments Inorg. Chem., I , 3 (1981). J. Halpem, Pure Appl. Chem., 51, 2171 (1979). J. Halpem, L. Y. Wong, J . Am. Chem. SOC., 90,6665 (1968). R. L. Sweany, J. Halpem, J . Am. Chem. SOC., 99, 8335 (1977). L. Simandi, F. Nagy, Acta Chim. Hung., 46, 137 (1965); 64, 15037h (1966). H. M. Feder, J. Halpern, J . Am. Chem. Soc., 97, 7186 (1975).


14.2. Types of Catalysts

14.2.1. Introduction A catalyst increases the rate of a chemical reaction without affecting its equilibrium or being substantially consumed, usually by forming bonds with the reactant(s) and opening up new sequences of steps. Catalysis is classified according to the phase(s) of the reactant-catalyst mixture. When the catalyst and reactant(s) are present in the same fluid phase, homogeneous catalysis occurs; when they are present in separate phases, heterogeneous catalysis occurs. The latter usually involves a solid catalyst, and the bonding of reactants to the surface where catalysis occurs is termed chemisorption. Catalyst performance is evaluated in terms of kinetics. Activity is a measure of how fast the catalytic reaction occurs (activity may be a reaction rate, a rate constant, or a conversion); selectivity is a measure of product distribution, such as a ratio of the rate of a desired reaction to the rate of disappearance of reactant; and stability is a measure of the rate of loss of activity or selectivity during operation as the catalyst undergoes changes in structure and composition. The science of surface catalysis is strongly influenced by technology. Catalysis plays a dominant role in chemical industry; many chemical manufacturing processes are catalytic, and the value of the goods manufactured yearly in the United States by processes that at some stage involve catalysis is almost a trillion dollars. Many of the best understood examples of surface catalysis involve industrial reactions or models thereof. The relation between the science and technology of homogeneous catalysis is not so close, but the technology still provides a good part of the impetus for research in inorganic and organometallic chemistry. In the next sections, solid catalysts are classified on the basis of their type of material, the most important types comprising metals, metal oxides, and metal sulfides. It is emphasized, however, that a more fundamental classification is also possible, although it can be developed to only a small degree because the science of catalysis is so complex. The next few paragraphs summarize this alternative classification based on the reactions catalyzed and the structures of intermediates; this classification also is a start toward unifying homogeneous and heterogeneous catalysis, which have traditionally been considered separately. For example, many catalytic cycles involve the transfer of protons. Common intermediates are carbenium ions and carbanions, and the catalysts include soluble and solid acids and bases and enzymes. The catalytic cycles may be similar, whether the proton donor (or acceptor) is a soluble molecule or ion or a functional group on a surface. Similarly, catalysis proceeding via organometallic intermediates may involve soluble transition metal complexes, metalloenzymes, or metal surfaces. Catalysis by metals is, however, much more complicated than acid-base catalysis, and the analogies between soluble metal complexes and surfaces cannot yet be developed beyond a few selected examples. 21

22

14.2.2. Solid Catal sts 14.2.2.1. Metallic Jatalysts 14.2.2.1.1. Metal Crystals and Films.

Table 1 is a summary of some of the most important catalysts grouped according to the types of reactions for which they are active. Even primitive understanding of catalyst functions is of great value to those searching for a catalyst for a particular reaction. (B. C.GATES)

14.2.2. Solid Catalysts 14.2.2.1. Metalllc Catalysts 14.2.2.1 .l. Metal Crystals and Films.

Transition metals, especially those of the platinum group, are important catalysts, displaying a remarkable range of activities and selectivities and finding application in

TABLE1. EXAMPLES OF CATALYSTS AND CATALYTIC REACTIONS Catalyst

Pt, Pd, Rh, Ni (surfaces) Rh complexes

Reactions

Cr,03 (surface) Ru, Fe (surfaces)

Hydrogenation of alkenes and arenes Hydrogenation of alkenes, hydrofonnylation of alkenes, carbonylation of alcohols, hydrocarboxylation of alkenes Hydrogenation of alkenes and arenes, hydroformylation of alkenes Hydrogenation of alkenes 3 H, + N, 2 NH,

Pt, Cr,O, (surfaces)

n-C,H,,

Pt, Ni (surfaces)

H,C-CH-CH,

Co complexes

Fe, Co (surfaces) Cu supported on ZnO (surface) Mo, W, Re complexes Ti, Zr complexes Sulfonic acid resins, SO,-Al,O,, acidic zeolites (surfaces) Si0,-Al,O,, acidic zeolites, AICl, (surfaces) Sulfated ZrO, Co3 , Mn3 complexes Pt, V,O, (surfaces) +

+

Ag (surface) Mixed oxides of Bi and Mo (surfaces)

-

-O

I

C

H

3 + 4 H,

H,C-C-CH,

II

+

H,

OH 0 CO + H, alkanes + alkenes + H,O CO + 2 H2 CH3OH Alkene metathesis Alkene polymerization Alkene isomerization and polymerization, arene alkylation, alcohol dehydration Alkane cracking, alkane-alkene alkylation, alkane disproportionation Alkane isomerization Alkene partial oxidation Oxidation of CO, SO,, and hydrocarbons H,C=CH, + 4 0,-+ H,CPartial oxidation of alkenes

/O\

CH,

J

14.2.2. Solid Catal sts 14.2.2.1. Metallic atalysts 14.2.2.1 -2. Supported Metal Catalysts.

23

some of the largest scale industrial proce~ses'-~. Metal catalysts are often used with a low dispersion (fraction of the metal atoms exposed on a surface). Metal films, wires, powders, and crystals are all used, although the typical forms of technological catalysts are small crystallites highly dispersed on supports. Much recent research with metal catalysts has been done with very clean single crystals under ultrahigh vacuum. The structures of the crystal surfaces can be determined precisely by low-energy electron diffraction (LEED), and catalytic reactions in the same apparatus have allowed comparisons of catalytic activity for various surface structures. For example, Pt single crystals have been cleaved to expose various surfaces exhibiting steps, kinks, and terraces, as shown in Fig. 1. The surfaces have been rigorously cleaned in a vacuum chamber, their compositions determined by Auger electron spectroscopy, their structures determined by LEED, and their catalytic activities measured for reactions of n-heptane. The results demonstrate the structure sensitivity of the hydrogenolysis reaction, which is associated with kink sites, and the structure insensitivity of the dehydrogenation and isomerization reactions4. The structures illustrate the nonuniformity that is typical of catalyst surfaces. The sites at which catalysis occurs are called active centers, and these may be a small fraction of the surface; they are usually unidentified. These results relate catalytic activity to catalyst surface structures. The results are narrowly restricted to metal surfaces under ultrahigh vacuum, however, since otherwise the LEED technique is inapplicable. When brought in contact with reactants at higher pressures, clean catalyst surfaces typically undergo drastic changes; for example, hydrocarbon reactants such as a-heptane react with Pt to give stable carbonaceous overlayers that cover most of the surface, greatly changing its reactivity and making its structure far more complicated than that of the metal single crystal-so complicated as to be undeterminable currently. Reactants may also bond to surfaces by corrosive chemisorption, whereby the structure of the metal surface is significantly changed. Metal films and powders are scarcely applied in practical catalysis, but gauzes are used, for example, in the oxidation of NO for the manufacture of nitric acid. The Pt catalyst is so active that the rate of reaction is largely determined by how fast the reactants are transported through the gas phase to the gauze surface; therefore, there is little practical incentive to increase the surface area of the metal by dispersing it on a support. Occasionally, metals are used as porous solids with internal surface areas of lo5 m2/kg or more. Porous Raney nickel particles are used for hydrogenation of fats. The pore structure in the nickel is formed by extraction of A1 from a Ni-A1 alloy with NaOH. (B. C.GATES)

1. 2. 3. 4.

G. C. Bond, Catalysis by Metals, Academic Press, London, 1962. A standard source, but dated. J. R. Anderson, Structure of Metallic Catalysts, Academic Press, New York, 1975. B. C. Gates, Catalytic Chemistry, Wiley, New York, 1992. Introductory textbook. G . A. Somorjai, in Catalyst Design, Progress and Perspectives, L. L. Hegedus, ed., Wiley, New York, 1987, p. 11. Catalysis on single-crystal surfaces.

14.2.2.1.2. Supported Metal Catalysts.

Metal catalysts are usually applied as small crystallites dispersed on a high-surfacearea porous support such as y-Al,03, SO,, or carbon'-3. The crystallites in such a catalyst are nonuniform in size and structure; crystallite dimensions may range from less than 1 nm to tens of nanometers. Crystallites smaller than about 1 nm are often referred


J

14.2.2. Solid Catal sts 14.2.2.1. Metallic atalysts 14.2.2.1 -2. Supported Metal Catalysts.

23

some of the largest scale industrial proce~ses'-~. Metal catalysts are often used with a low dispersion (fraction of the metal atoms exposed on a surface). Metal films, wires, powders, and crystals are all used, although the typical forms of technological catalysts are small crystallites highly dispersed on supports. Much recent research with metal catalysts has been done with very clean single crystals under ultrahigh vacuum. The structures of the crystal surfaces can be determined precisely by low-energy electron diffraction (LEED), and catalytic reactions in the same apparatus have allowed comparisons of catalytic activity for various surface structures. For example, Pt single crystals have been cleaved to expose various surfaces exhibiting steps, kinks, and terraces, as shown in Fig. 1. The surfaces have been rigorously cleaned in a vacuum chamber, their compositions determined by Auger electron spectroscopy, their structures determined by LEED, and their catalytic activities measured for reactions of n-heptane. The results demonstrate the structure sensitivity of the hydrogenolysis reaction, which is associated with kink sites, and the structure insensitivity of the dehydrogenation and isomerization reactions4. The structures illustrate the nonuniformity that is typical of catalyst surfaces. The sites at which catalysis occurs are called active centers, and these may be a small fraction of the surface; they are usually unidentified. These results relate catalytic activity to catalyst surface structures. The results are narrowly restricted to metal surfaces under ultrahigh vacuum, however, since otherwise the LEED technique is inapplicable. When brought in contact with reactants at higher pressures, clean catalyst surfaces typically undergo drastic changes; for example, hydrocarbon reactants such as a-heptane react with Pt to give stable carbonaceous overlayers that cover most of the surface, greatly changing its reactivity and making its structure far more complicated than that of the metal single crystal-so complicated as to be undeterminable currently. Reactants may also bond to surfaces by corrosive chemisorption, whereby the structure of the metal surface is significantly changed. Metal films and powders are scarcely applied in practical catalysis, but gauzes are used, for example, in the oxidation of NO for the manufacture of nitric acid. The Pt catalyst is so active that the rate of reaction is largely determined by how fast the reactants are transported through the gas phase to the gauze surface; therefore, there is little practical incentive to increase the surface area of the metal by dispersing it on a support. Occasionally, metals are used as porous solids with internal surface areas of lo5 m2/kg or more. Porous Raney nickel particles are used for hydrogenation of fats. The pore structure in the nickel is formed by extraction of A1 from a Ni-A1 alloy with NaOH. (B. C.GATES)

1. 2. 3. 4.

G. C. Bond, Catalysis by Metals, Academic Press, London, 1962. A standard source, but dated. J. R. Anderson, Structure of Metallic Catalysts, Academic Press, New York, 1975. B. C. Gates, Catalytic Chemistry, Wiley, New York, 1992. Introductory textbook. G . A. Somorjai, in Catalyst Design, Progress and Perspectives, L. L. Hegedus, ed., Wiley, New York, 1987, p. 11. Catalysis on single-crystal surfaces.

14.2.2.1.2. Supported Metal Catalysts.

Metal catalysts are usually applied as small crystallites dispersed on a high-surfacearea porous support such as y-Al,03, SO,, or carbon'-3. The crystallites in such a catalyst are nonuniform in size and structure; crystallite dimensions may range from less than 1 nm to tens of nanometers. Crystallites smaller than about 1 nm are often referred

24

14.2.2. Solid Catal sts 14.2.2.1. Metallic C!atalysts 14.2.2.1-2. Supported Metal Catalysts.

TERRACE

A

\

B

STEP

TERRACE

Figure 1. Single crystal surfaces of Pt, cleaved to expose steps and terraces (A) and steps, kinks, and terraces (B). to as clusters, and they are small enough that dispersions approach unity; when the average Pt crystallite size is 10 nm, the dispersion is about 0.1. There are several reasons why highly dispersed metals are preferred: 1. The greater the dispersion, the greater the fraction of the (usually expensive) metal atoms engaged in catalysis; 2. For structure-sensitive reactions (such as alkane hydrogenolysis), the activity per unit of metal surface area depends on the crystallite size, provided that the crystallites are relatively small',*; 3. The selectivity [e.g., the rate of hydrogenolysis relative to the rate of dehydrogenation (both of which are important reactions in the reforming of petroleum distillates)] may also depend on crystallite size4; 4. The influence of the underlying support (which can be considered a multidentate ligand) increases with decreasing crystallite size, and for certain highly dispersed metals (e.g., Rh) on basic metal oxide supports, the selectivity (e.g., in CO hydrogenation) may be significantly altered by changes in the support5; and 5. When bimetallic catalysts are applied, the composition and structure of the metal surface (and, therefore, the catalytic activity) may be different for different crystallite

14.2.2. Solid Catal sts 14.2.2.1. Metallic Zatalysts 14.2.2.1.2. Supported Metal Catalysts.

25

The metal in a supported metal catalyst may be characterized by electron microscopy (providing a distribution of crystallite sizes), X-ray diffraction line broadening (providing an average size for crystallites larger than about 4 nm), and specific chemisorption (titration) with compounds such as H, or CO [providing an estimate of the exposed metal area (vs. the total surface area of metal plus support)]. Spectroscopy also provides structural information about supported metals. The EXAFS (extended X-ray absorption fine structure) technique helps to define average crystallite structures’, and infrared spectroscopy provides structural characterization of chemisorbed species. The challenges of determining structures of supported metals are great because of the nonuniformities of the metal crystallites in almost all catalysts. Crystallite size can be influenced by the methods of catalyst preparation (see 14.2.7), and the size distribution usually changes during the lifetime of a catalyst. The crystallites sinter (i.e., aggregate to form larger crystallites) in operation by processes that include migration on the support and vapor-phase transport of volatile compounds of the metal. Since sintering implies loss of surface area and, therefore, loss of catalytic activity, supported metal catalysts are typically redispersed (regenerated), e.g., in processes whereby they are heated in oxygen to bum off carbonaceous material accumulated on the surface (called coke) and allowed to react with compounds (e.g., HCl and 0, in the case of Pt) that form volatile species that are redeposited then reduced to give redispersed metal. Both the rate of sintering and the rate of coke accumulation influence catalyst stability; stability is also influenced by poisons (e.g., sulfur-containing compounds) present in reactant mixtures. Supported bimetallic catalysts (and unsupported alloys) have been studied intensively; much of the incentive has been provided by the industrial application of Re-Pt/Al,O, and other bimetallics for reforming of naphtha. Alloying brings about only moderate changes in the electronic structures of the alloy component^'-^, and influences chemisorption bond strengths only modestly. However, it often drastically changes structures of chemisorbed species. Changes in catalytic performance that reflect changes in surface composition caused by alloying are attributed primarily to geometric effects. These are important for reactions involving catalytic sites, that are ensembles of surface atoms rather than single atoms. The porous support for a metal catalyst is often prepared (see 14.2.7.1) with a large internal surface. The support must be resistant to sintering and have sufficient mechanical strength to withstand handling and use. The typical support can be considered inert; its function is to maintain the metal in a dispersed form. There are catalysts, however, for which the support also provides a separate catalytic function, complementing the metal. A bifunctional catalyst of this sort is Pt/A1,0?, which preceded Re-Pt/Al,O, as the standard naphtha reforming catalyst. The Pt surface is catalytically active for alkane dehydrogenation and alkene hydrogenation, and the acidic support is active for reactions such as isomerization of alkenes. The optimum catalyst represents a balancing of the separate functions; the activities are adjusted by variation of the respective surface areas and the proton donor strength of the A1203, which can be increased by incorporation of chloride on the oxide surface. Alkenes, formed by dehydrogenation of alkanes on the metal surface, rapidly migrate to the nearby acidic sites on the support surface, where they react further. The alkenes are the key intermediates in the catalysis. (B. C.GATES)

26

14.2. Ty es of Catalysts 14.2.2. %lid Catalysts 14.2.2.2. Metal Oxide and Metal Sulfide Catalysts

1. M. Boudart, J . Mol. Catal., 30,27 (1985). A review of supported metal catalysts. 2. M. Boudart and G. Djtga-Mariadassou,Kinetics of Heterogeneous Catalytic Reactions, Princeton University Press, Princeton, NJ, 1984. 3. B. C. Gates, Catalytic Chemistry, Wiley, New York, 1992. 4. B. C. Gates, J. R. Katzer, G.C. A. Schuit, Chemistry of Catalytic Processes, Chap. 3, McGrawHill, New York, 1979. Pt/A1,03 and the catalytic reforming process. 5. M. Ichikawa, K. Shikahura,Proc. 7th Int. Cong. Catal.,Kodansha, Tokyo, B, 925 (1981). Chem. Abstr., 95 203250r (1981). 6. J. H. Sinfelt, Bimetallic Catalysts, Discoveries, Concepts, and Applications, Wiley, New York, 1983. 7. D. C. Koningsberger, R. Prins, eds., X-Ray Absorption: Principles, Applications, Techniques of EXAFS, SEXAFS, and XANES, Wiley, New York, 1988.

14.2.2.2. Metal Oxlde and Metal Sulfide Catalysts

The principal constituents of most solid catalysts are metal oxides. These may be nearly inert supports, but typically they are catalytically active themselves. Within the class of metal oxides, a great range of catalytic activities is found; the oxides include acids and bases, hydrogenation/dehydrogenation catalysts, hydrocarbon partial oxidation catalysts, etc. (Table 1, 14.2.1). Alumina, silica, magnesia, and other oxides used as catalysts are usually highsurface-area (typically 3 X lo5 m’/kg) solids’. The surfaces are the boundaries of small aggregates of crystallites. Interspersing the aggregates are void spaces (pores). Pores with diameters 1 2 nm are often called micropores, those with diameters between 2 and 50 nm, mesopores, and those larger, macropores. It follows that more surface area per unit volume of catalyst is contained in micropores than in macropores, and the optimum catalyst pore structure represents a tradeoff between the activity associated with high internal surface area and the greater restriction to transport of reactant and product molecules in the smaller pores. If the transport restriction is significant, the interior surface area is poorly accessible to reactants and inefficiently used’. Metal oxide catalysts (and supported metal catalysts) are usually applied in the form of cylinders (extrudates), with the particle dimensions dictated by the hydrodynamics (especially the pressure drop) in a catalytic flow reactor and by the resistance to transport of reactant and product molecules through the pores, which increases with increasing path length (particle dimension). The X-ray crystal structures of many of the simple metal oxides used in catalysis have been determined3. The structures of the surfaces are more pertinent, however, since that is where catalysis takes place. Inferences about surface structure can occasionally be drawn from bulk structure, but in general structures of metal oxide surfaces are complicated and highly nonuniform. With oxide and other surfaces the catalytically active sites constitute only a small fraction of the surface. It is, therefore, generally difficult to determine the structures of catalytic sites. Considering a model of a portion of a transition alumina surface (Fig. I), one can illustrate some surface groups of broad importance in catalysis. Metal oxides, including transition aluminas, are typically terminated by -OH groups, the protons (or other cations) at the surface being necessary to maintain electrical neutrality of the material. The acidic -OH groups can be detected by chemisorption of a base (e.g., pyridine) and observation of the resulting infrared spectrum. There are -OH groups in various surroundings that have various proton-donor strengths. Besides weak Bronsted acid groups, the Al’O, surface also incorporates Lewis acid groups, which can be characterized by


26


1. M. Boudart, J . Mol. Catal., 30,27 (1985). A review of supported metal catalysts. 2. M. Boudart and G. Djtga-Mariadassou,Kinetics of Heterogeneous Catalytic Reactions, Princeton University Press, Princeton, NJ, 1984. 3. B. C. Gates, Catalytic Chemistry, Wiley, New York, 1992. 4. B. C. Gates, J. R. Katzer, G.C. A. Schuit, Chemistry of Catalytic Processes, Chap. 3, McGrawHill, New York, 1979. Pt/A1,03 and the catalytic reforming process. 5. M. Ichikawa, K. Shikahura,Proc. 7th Int. Cong. Catal.,Kodansha, Tokyo, B, 925 (1981). Chem. Abstr., 95 203250r (1981). 6. J. H. Sinfelt, Bimetallic Catalysts, Discoveries, Concepts, and Applications, Wiley, New York, 1983. 7. D. C. Koningsberger, R. Prins, eds., X-Ray Absorption: Principles, Applications, Techniques of EXAFS, SEXAFS, and XANES, Wiley, New York, 1988.

14.2.2.2. Metal Oxlde and Metal Sulfide Catalysts

The principal constituents of most solid catalysts are metal oxides. These may be nearly inert supports, but typically they are catalytically active themselves. Within the class of metal oxides, a great range of catalytic activities is found; the oxides include acids and bases, hydrogenation/dehydrogenation catalysts, hydrocarbon partial oxidation catalysts, etc. (Table 1, 14.2.1). Alumina, silica, magnesia, and other oxides used as catalysts are usually highsurface-area (typically 3 X lo5 m’/kg) solids’. The surfaces are the boundaries of small aggregates of crystallites. Interspersing the aggregates are void spaces (pores). Pores with diameters 1 2 nm are often called micropores, those with diameters between 2 and 50 nm, mesopores, and those larger, macropores. It follows that more surface area per unit volume of catalyst is contained in micropores than in macropores, and the optimum catalyst pore structure represents a tradeoff between the activity associated with high internal surface area and the greater restriction to transport of reactant and product molecules in the smaller pores. If the transport restriction is significant, the interior surface area is poorly accessible to reactants and inefficiently used’. Metal oxide catalysts (and supported metal catalysts) are usually applied in the form of cylinders (extrudates), with the particle dimensions dictated by the hydrodynamics (especially the pressure drop) in a catalytic flow reactor and by the resistance to transport of reactant and product molecules through the pores, which increases with increasing path length (particle dimension). The X-ray crystal structures of many of the simple metal oxides used in catalysis have been determined3. The structures of the surfaces are more pertinent, however, since that is where catalysis takes place. Inferences about surface structure can occasionally be drawn from bulk structure, but in general structures of metal oxide surfaces are complicated and highly nonuniform. With oxide and other surfaces the catalytically active sites constitute only a small fraction of the surface. It is, therefore, generally difficult to determine the structures of catalytic sites. Considering a model of a portion of a transition alumina surface (Fig. I), one can illustrate some surface groups of broad importance in catalysis. Metal oxides, including transition aluminas, are typically terminated by -OH groups, the protons (or other cations) at the surface being necessary to maintain electrical neutrality of the material. The acidic -OH groups can be detected by chemisorption of a base (e.g., pyridine) and observation of the resulting infrared spectrum. There are -OH groups in various surroundings that have various proton-donor strengths. Besides weak Bronsted acid groups, the Al’O, surface also incorporates Lewis acid groups, which can be characterized by

14.2. Ty es of Catalysts 14.2.2. lolid Catalysts 14.2.2.2. Metal Oxide and Metal Sulfide Catalysts

27

Figure 1. Model of the (1 11) face of A1,0,: (0)-OH group; (

the distinctive infrared spectra of chemisorbed bases. The Lewis acid groups are exposed (coordinatively unsaturated) A13+ ions. The distribution of Bransted and Lewis acid groups on the surface can be adjusted by heating the sample; dehydroxylation results in the simultaneous loss of Brcansted acid groups and formation of Lewis acid groups. The proton donor strength of the former groups is increased by the presence of juxtaposed Lewis acid groups. The acid strength can also be increased by incorporation of Si in place of A1 in the oxide lattice at the surface (the resulting solid is called silica-alumina) or by incorporation of electron-withdrawing halides in the surface. The Al,O, surface also incorporates various basic sites, the -OH groups and the 0’- ions, which are proton acceptors. There is a broad range of catalytic characteristics corresponding to the range of surface acid and base groups,. Progress in learning about the surface chemistry of Al,O, and other oxides will be slow because of the complexity of the surfaces and the lack of definitive characterization techniques. Many common metal oxides (SiO,, MgO, Cr203,etc.) are like AI,O, in being terminated by -OH groups, 0,- ions, and coordinative unsaturated metal ions. These exposed metal ions may be responsible for the distinctive catalytic character of a particular oxide surface. Many metal oxides are called amorphous because the small crystalline aggregates fail to give sharp X-ray diffraction patterns, but there is one important class which is exceptional, the crystalline aluminosilicates (zeolites)536.The zeolites are related to silica-alumina (both contain linked AlO, and SiO, tetrahedra) but they have better defined structures than the amorphous metal oxide catalysts. When 24 SiO, and AlO, tetrahedra are put together as shown in Fig. 2, the result is a truncated octahedral structure called a sodalite cage. When these cages are arranged in a regular array so that each square face (four-membered oxygen ring) is connected by bridging oxygen ions to the square face of another cage, the structure of zeolite A is formed (Fig. 3). These crystals have a regular array of molecular-scale pores. The apertures are eight-membered oxygen rings, each having a diameter of 0.42 nm and opening into a cavity called a supercage, surrounded by eight sodalite cages. This cavity is large enough to contain a sphere with a diameter of 1.1 nm.

28


U

o Oxygen 0 Aluminum or silicon Figure 2. Sodalite cage. The truncated octahedron is clear from the representation on the left, where the lines represent 02-ions and the points of intersection represent Si4+ or AP+ ions.

The zeolites that find widest application in catalysis are the faujasites, including zeolites X and Y. Each sodalite cage is connected to four others; each connecting unit consists of six bridging oxygen ions linking the hexagonal faces of the two sodalite cages (Fig. 4). The supercage surrounded by 10 sodalite cages is large enough to contain a sphere with a diameter of 1.2 nm, and the apertures are 12-membered oxygen rings, 0.74 nm in diameter. Another important class of zeolite catalysts is typified by ZSM-5, which has a threedimensional pore structure with apertures intermediate to those of zeolite A and faujasites'. X-Ray diffraction studies have determined the structure of these and other zeolites, including the framework structures and the locations of cations in the cages that must be present to maintain charge neutrality. Infrared spectroscopy has been used, as for the amorphous oxides, with adsorbed organic probe molecules to determine details of the surface chemistry, including characterization of the same types of functional groups present on A1203and SO2-A120,. The intracrystalline spaces in molecular-scale zeolite pores provide unique environments for catalytic reactions, a topic considered in Section 14.2.6; the catalytic groups in zeolites may include acids, transition metal complexes, and metal clusters. The term mixed oxide catalyst is commonly applied to oxides of Bi and Mo, which are used in industrial processes for partial oxidation of hydrocarbons, a prominent example being the ammoxidation of p r ~ p e n e ~ . ~ : H,C=CH-CH,

+ -32

0 2

+ NH3

-

H,C=CHCN

+ 3H2O

(a)


29

Figure 3. The arrangement of truncated octahedra (sodalite cages) in zeolite A.

Figure 4. Perspective views of the arrangement of truncated octahedra (sodalite cages) in faujasite.

30


Modem industrial catalysts may incorporate a large number of components, e.g.,

Co, Fe, Bi, Mo, Na, Ba, P, and 0. The first industrial ammoxidation catalysts were

simpler, the critical constituents being Bi, Mo, and 0. A later generation of catalysts consisted of U, Sb, and A suggested structure of USb3010,believed to be the catalytically active phase present in the catalyst, is shown in Fig. 5. This structure illustrates the complexity of many commercial catalysts; Sb exists in two different environments and 0 in four. Mechanisms for propene oxidation (to give acrolein) and for propene ammoxidation on the mixed-oxide surface^'^^ involve formation of an allylic species from propene on coordinatively unsaturated cation sites at the surface. Subsequent steps include electron and oxygen atom transfers. Many partial-oxidation catalysts allow oxygen from the gas phase to become incorporated in the lattice; the oxygen atoms then migrate through the lattice to be donated to the chemisorbed hydrocarbon; there is a compensating electron transfer through the lattice. The catalyst surface provides a limited amount of reactive oxygen, so that the partially oxidized organic product can desorb before being further oxidized. Another important partial oxidation catalyst, used for the production of ethene oxide from ethene, is silver metal, which functions as its surface is converted to a submonolayer silver oxide that provides the oxygen to the alkene5. Oxidation catalysis nominally involving metals typically involves surface metal oxides. Likewise, reactions of sulfur compounds nominally involving metal oxide catalysts actually involve metal sulfide catalysts". Hydrogenolysis, referred to as hydrodesulfurization, used in petroleum refining to produce clean-burning fuels (those not forming sulfur oxides on combustion)12,involves, e.g.: 0 8 9 9 .

The industrial catalyst is initially a mixed oxide of Mo and Co supported on y-Al,O,. In the presence of H, and the H,S and organosulfur compounds that form H,S, the oxide is converted into a surface sulfide. The catalyst consists of small crystallites of MoS, dispersed on the A1,0, support. Some of the Co is present at the crystallite edges in the MoS, structure and forms the catalytic sites13. (B. C.GATES)

1. B. C. Linsen, J. M. H. Fortuin, C. Okkerse, J. J. Steggerda, eds., Physical and Chemical Aspects of Adsorbents and Catalysrs, Academic Press, New York, 1970. Reviews of the common catalyst supports. 2. C. N. Satterfield, Mass Transfer in Heterogeneous Catalysis, MIT Press, Cambridge, MA, 1970. Standard reference. 3. F. S. Stone, in Chemistry and Chemical Engineering of Catalytic Processes, R. Prins, G . C. A. Schuit, eds., Sijthoff and Noordhoff, Alphen an den Rijn, The Netherlands, 1980. Structures of metal oxide catalysts. 4. (a) H.-P.Boehm, H. Knozinger in Catalysis-Science and Technology, J. R. Anderson, M. Boudart, eds., Vol. 4, Springer, Berlin, 1983, p. 39; (b) H. Knozinger, P. Ratnasamy, Cutal. R e v . P c i . Eng., 17, 31 (1978). 5. B. C. Gates, Catalytic Chemistry, Wiley, New York, 1992 6. F. R. Ribeiro, A. E. Rodrigues, L. D. Rollmann, C. Naccache, eds., Zeolites: Science and Technology, Martinius Nijhoff, The Hague, 1984.


Figure 5. Unit cell of USb30,0'0.

32

14.2. Ty es of Catalysts 14.2.3. %luble Catalysts 14.2.3.1. Selectivity Advantages

7. P. B. Weisz, Proc. 7th Int. Cong. Catal., 1, 1 (1981); Chem. Abst., 95, 192859t (1981). 8. B. C. Gates, J. R. Katzer, G. C. A. Schuit, Chernisrry ofCatulytic Processes, Chap. 4, McGrawHill, New York, 1979. Ammoxidation catalysis and processes. 9. R. K. Grasselli, J. D. Burrington, Adv. Catal., 30, 133, 1981. 10. R. K. Grasselli and D. K. Suresh, J . Catal., 25, 273 (1972). 11. 0. Weisser and S . Landa, Sulphide Catalysts: Their Properties and Applications, Pergamon, London, 1973. 12. Ref. 8, Chap. 5. Sulfide catalysts and hydrodesulfurization processes. 13. R. Prins, V. H. J. de Beer, G. A. Somorjai, Caral. Rev.-Sci. Eng., 31, 1 (1989).

14.2.3. Soluble Catalysts 14.2.3.1. Selectivity Advantages

Prior to 1970 the term homogeneous catalysis usually connoted acid-base catalysis, and this is the best understood type of catalysis'. Since then, however, homogeneous catalysis has come to mean primarily transition-metal-complex catalysis, which has been the object of intensive research motivated in part by a series of technological accompli~hrnents~*~. The first industrial processes involving transition-metal-complex catalysis were largely focused on ethyne conversion', but these are no longer important. Many of the present processes involve alkene and CO conversion, including the following4:

-

1. The Wacker process, which proceeds in the presence of palladium and copper complexes in aqueous solution:

+ H,O + PdC1, CH3CH0 + Pd + 2 HCl PdC1, + 2 CuCl Pd + 2 CuC1, 1 2 CuCl + - 0, + 2 HCl 2 CUCI, + H20 2

H,C=CH,

j

SUM: H,C=CH,

+ -21 0, +CH,CHO

(4 (b) (c)

(4

--

2. The vinyl acetate process, proceeding similarly in solutions of Pd and Cu complexes: H,C=CH,

+ CH3COOH + PdC1, CH,COOCH=CH, + Pd + 2 HC1 PdC1, + 2 CuCl Pd + 2 CuC1, 1 2 CuCl + - 0, + 2 HC12 CuC1, + H20 2

SUM: H,C=CH,

+ -21

0 2

+ CH3COOH

CH,COOCH=CH,

+ H2O

(e) (f)

(g)

(h)

3. The 0x0 (hydrofomylation) process, proceeding in the presence of a Co complex or a Rh complex (see Section 14.6.3):

RCH=CH,

+ CO + H,

--+

RCH,CH,CHO (with R HCH,) IHO

(i)


32

14.2. Ty es of Catalysts 14.2.3. %luble Catalysts 14.2.3.1. Selectivity Advantages

7. P. B. Weisz, Proc. 7th Int. Cong. Catal., 1, 1 (1981); Chem. Abst., 95, 192859t (1981). 8. B. C. Gates, J. R. Katzer, G. C. A. Schuit, Chernisrry ofCatulytic Processes, Chap. 4, McGrawHill, New York, 1979. Ammoxidation catalysis and processes. 9. R. K. Grasselli, J. D. Burrington, Adv. Catal., 30, 133, 1981. 10. R. K. Grasselli and D. K. Suresh, J . Catal., 25, 273 (1972). 11. 0. Weisser and S . Landa, Sulphide Catalysts: Their Properties and Applications, Pergamon, London, 1973. 12. Ref. 8, Chap. 5. Sulfide catalysts and hydrodesulfurization processes. 13. R. Prins, V. H. J. de Beer, G. A. Somorjai, Caral. Rev.-Sci. Eng., 31, 1 (1989).

14.2.3. Soluble Catalysts 14.2.3.1. Selectivity Advantages

Prior to 1970 the term homogeneous catalysis usually connoted acid-base catalysis, and this is the best understood type of catalysis'. Since then, however, homogeneous catalysis has come to mean primarily transition-metal-complex catalysis, which has been the object of intensive research motivated in part by a series of technological accompli~hrnents~*~. The first industrial processes involving transition-metal-complex catalysis were largely focused on ethyne conversion', but these are no longer important. Many of the present processes involve alkene and CO conversion, including the following4:

-

1. The Wacker process, which proceeds in the presence of palladium and copper complexes in aqueous solution:

+ H,O + PdC1, CH3CH0 + Pd + 2 HCl PdC1, + 2 CuCl Pd + 2 CuC1, 1 2 CuCl + - 0, + 2 HCl 2 CUCI, + H20 2

H,C=CH,

j

SUM: H,C=CH,

+ -21 0, +CH,CHO

(4 (b) (c)

(4

--

2. The vinyl acetate process, proceeding similarly in solutions of Pd and Cu complexes: H,C=CH,

+ CH3COOH + PdC1, CH,COOCH=CH, + Pd + 2 HC1 PdC1, + 2 CuCl Pd + 2 CuC1, 1 2 CuCl + - 0, + 2 HC12 CuC1, + H20 2

SUM: H,C=CH,

+ -21

0 2

+ CH3COOH

CH,COOCH=CH,

+ H2O

(e) (f)

(g)

(h)

3. The 0x0 (hydrofomylation) process, proceeding in the presence of a Co complex or a Rh complex (see Section 14.6.3):

RCH=CH,

+ CO + H,

--+

RCH,CH,CHO (with R HCH,) IHO

(i)

h

14.2. Ty es of Catal sts 14.2.3. %luble Cata sts 14.2.3.1. Selectivity dvantages

-

33

Another large-scale process involving CO chemistry is the carbonylation of methanol to give acetic acid catalyzed by a Rh complex, which requires a cocatalyst (promoter), CH31 (see 14.6.5)4: CH,OH

+ CO

CH3COOH

ti)

The details of the chemistry of these and related processes are considered in subsequent sections (cited above); the object of this and the next section is to consider homogeneous catalytic processes from the viewpoint of the chemical engineer and to compare homogeneous and heterogeneous catalysis. The main advantage of a commercial homogeneous transition-metal-complex catalyst is its high selectivity, which results when there is one predominant catalytic cycle. Only rarely can a combination of metal and ligands be found to give high selectivity along with good activity and stability. It is remarkable how frequently Rh offers the best combination of selectivity and activity for reactions involving alkenes, H,, and/or CO. Surface catalysis is typically less selective than molecular catalysis because the surfaces are nonuniform, presenting a variety of catalytic sites and allowing more than one catalytic cycle. We may define a hierarchy of catalysts ordered by their selectivity characteristics. If surfaces rank lowest in the hierarchy, simple molecular species next, and enzymes highest, then it is expected that increasingly complex molecules with appropriately designed structures will be more selective catalysts, approaching the limits defined by nature’s enzyme catalysts. Development of increasingly selective molecular catalysts can be expected to follow developments in inorganic and organometallic synthesis. Improvements in selectivity of surface catalysts may also follow these developments, as methods are found for synthesizing specific molecular structures on supports. The importance of selectivity and the opportunities for modifying catalysts to give improved selectivity are illustrated by the 0x0 process. Formation of the desired straightchain aldehyde is accompanied by formation of branched-chain aldehydes and other products. Incorporating bulky phosphine ligands in the CO complex made it a more selective catalyst for straight-chain product, which has been explained as a steric effecl!. Rhodium complexes with triphenyl phosphine ligands give >90% selectivity to straightchain aldehyde when a large excess of triphenylphosphine to Rh is used, a s in the industrial propene hydroformylation process6. Precise steric control of the environment about an Rh center may be exerted by bulky chelating diphosphine ligands, allowing asymmetric hydrogenation to occur (see 14.3.4.5). A catalyst of this type was applied in the commercial production of L-dopa, a drug used for the treatment of Parkinson’s disease. The accomplishment is a milestone; it perhaps represents the most sophisticated molecular design of a technological catalyst2s3. Recent research in homogeneous catalysis emphasized the identification and development of transition metal complexes for the conversion of synthesis gas (CO + H,)’, which is produced from natural gas or from coal and other hydrocarbon sources by gasification. Solid catalysts are known for conversion of CO + H, into methanol, into mixtures of hydrocarbons (Fischer-Tropsch products), or into other products (see 14.2.1); but there is a need for improved catalysts. Molecular complexes offer good prospects because the most desirable characteristics would be high selectivities for products such as ethene, propene, ethanol, ethylene glycol, etc. Ruthenium complexes convert CO H, into methanol’. Rhodium complexes at very high pressures catalyze the formation of methanol and ethylene glycol in high

+

34

Y

14.2. Ty es of Catal sts 14.2.3. %luble Cata ysts 14.2.3.2. Process Engineering and Product Recovery Problems

yield^^,^. The catalytically active species may be rhodium carbonyl cluster^^.^. Metal clusters are in prospect a promising class of catalysts for CO hydrogenation, since their neighboring metal centers may confer on them some of the properties of metal surfaces," but most metal clusters are unstable under practical conditions.

(B. C.GATES)

1. M. L. Bender, Mechanisms of Homogeneous Catalysis form Protons to Proteins, Wiley Interscience, New York, 1971. Includes a thorough review of acid-base catalysis. 2. G. W. Parshall, Homogeneous Catalysis, Wiley, New York, 1980. Review of chemistry of transition metal complex catalysis. 3. J. P. Collrnan, L. S. Hegedus, J. R. Norton, R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, 2nd ed., University Science Books, Mill Valley, CA, 1987. 4. B. C. Gates, J. R. Katzer, G. C. A. Schuit, Chemistry of Catalytic Processes, Chap. 2, McGrawHill, New York, 1979. Process chemistry and engineering of transition metal complex catalysis. 5. M. Orchin, W. Rupilius, Catal. Rev., 6, 85 (1972). 6. R. L. Pruett, Ann. N . Y.Acad. Sci., 295,239 (1977). 7. R. L. Pruett, Science, 211, 11 (1981). 8. J. S. Bradley, J . Am. Chem. SOC., 101,7419 (1979). 9. J. L. Vidal, W. E. Walker, R. L. Pruett, R. C. Schoening, R. A. Fiato in M. Tsutsui, eds., Fundamental Research in Homogeneous Catalysis, Vol. 3, Plenum, New York, 1979. 10. B. C. Gates, L. Guczi, H. Knbzinger, eds., Metal Clusters in Catalysis, Elsevier, Amsterdam, 1986.

14.2.3.2. Process Engineeringand Product Recovery Problems The above processes are catalyzed by transition metal complexes, and their catalytic cycles are characterized by a number of common elementary steps such as oxidative addition/reductive elimination and ligand dissociation/association. Besides the details of the chemistry, there are many characteristics of the process engineering that unify these processes'. 1. The reactions proceed under mild conditions, at pressures -2 X lo6 N/m2, and temperatures of -425K. Higher temperatures would necessitate higher pressures for maintenance of a liquid phase, and the higher pressures imply higher processing costs. In contrast, solid catalysts stable enough to withstand high temperatures can be used at low pressures with the reactants flowing in a gas phase. 2. The reactants include small molecules (e.g., CO, H,, H,C=CH,, and O,), most of which are gases under ambient conditions. Since the catalysts are dissolved in liquids, the reactants must be introduced by bubbling into the liquid. The reactor design must allow efficient transport of the reactants from the gas bubbles into the liquid; this requires producing swarms of bubbles in a well-mixed reactor. A continuous flow reactor is often a stirred tank, or a long, narrow tube, which may contain solid packing to disperse the flowing fluids and bring about rapid mixing. Without provision for efficient gas-liquid contacting, and with rapid liquid-phase reactions the observed rate (and possibly the product distribution) may be limited by the transport of reactant between phases. Results from such poorly designed reactors give false impressions of catalytic activity and selectivity. Standard chemical engineering textbooks provide guidelines for data interpretation and appropriate reactor 3. Since most of the reactions are exothermic, liberating 120 to 200 kJ/mol of reactant, the well-mixed liquid medium allows rapid convective heat removal from the reactor. 4. Corrosive reactant solutions require expensive materials such as Ti and stainless steels. Consequently, there is a motive to design the process so that a minimum of corrosion


34

Y


yield^^,^. The catalytically active species may be rhodium carbonyl cluster^^.^. Metal clusters are in prospect a promising class of catalysts for CO hydrogenation, since their neighboring metal centers may confer on them some of the properties of metal surfaces," but most metal clusters are unstable under practical conditions.

(B. C.GATES)

1. M. L. Bender, Mechanisms of Homogeneous Catalysis form Protons to Proteins, Wiley Interscience, New York, 1971. Includes a thorough review of acid-base catalysis. 2. G. W. Parshall, Homogeneous Catalysis, Wiley, New York, 1980. Review of chemistry of transition metal complex catalysis. 3. J. P. Collrnan, L. S. Hegedus, J. R. Norton, R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, 2nd ed., University Science Books, Mill Valley, CA, 1987. 4. B. C. Gates, J. R. Katzer, G. C. A. Schuit, Chemistry of Catalytic Processes, Chap. 2, McGrawHill, New York, 1979. Process chemistry and engineering of transition metal complex catalysis. 5. M. Orchin, W. Rupilius, Catal. Rev., 6, 85 (1972). 6. R. L. Pruett, Ann. N . Y.Acad. Sci., 295,239 (1977). 7. R. L. Pruett, Science, 211, 11 (1981). 8. J. S. Bradley, J . Am. Chem. SOC., 101,7419 (1979). 9. J. L. Vidal, W. E. Walker, R. L. Pruett, R. C. Schoening, R. A. Fiato in M. Tsutsui, eds., Fundamental Research in Homogeneous Catalysis, Vol. 3, Plenum, New York, 1979. 10. B. C. Gates, L. Guczi, H. Knbzinger, eds., Metal Clusters in Catalysis, Elsevier, Amsterdam, 1986.

14.2.3.2. Process Engineeringand Product Recovery Problems The above processes are catalyzed by transition metal complexes, and their catalytic cycles are characterized by a number of common elementary steps such as oxidative addition/reductive elimination and ligand dissociation/association. Besides the details of the chemistry, there are many characteristics of the process engineering that unify these processes'. 1. The reactions proceed under mild conditions, at pressures -2 X lo6 N/m2, and temperatures of -425K. Higher temperatures would necessitate higher pressures for maintenance of a liquid phase, and the higher pressures imply higher processing costs. In contrast, solid catalysts stable enough to withstand high temperatures can be used at low pressures with the reactants flowing in a gas phase. 2. The reactants include small molecules (e.g., CO, H,, H,C=CH,, and O,), most of which are gases under ambient conditions. Since the catalysts are dissolved in liquids, the reactants must be introduced by bubbling into the liquid. The reactor design must allow efficient transport of the reactants from the gas bubbles into the liquid; this requires producing swarms of bubbles in a well-mixed reactor. A continuous flow reactor is often a stirred tank, or a long, narrow tube, which may contain solid packing to disperse the flowing fluids and bring about rapid mixing. Without provision for efficient gas-liquid contacting, and with rapid liquid-phase reactions the observed rate (and possibly the product distribution) may be limited by the transport of reactant between phases. Results from such poorly designed reactors give false impressions of catalytic activity and selectivity. Standard chemical engineering textbooks provide guidelines for data interpretation and appropriate reactor 3. Since most of the reactions are exothermic, liberating 120 to 200 kJ/mol of reactant, the well-mixed liquid medium allows rapid convective heat removal from the reactor. 4. Corrosive reactant solutions require expensive materials such as Ti and stainless steels. Consequently, there is a motive to design the process so that a minimum of corrosion

Y


35

resistant material is required, which implies restricting the corrosive solutions to a small number of devices in the process. 5. Separation of soluble catalysts from the reaction products and recycling them to the reactor are usually expensive. Most of the catalysts (e.g., Rh, Pd) are expensive, and the efficiency of catalyst recovery must be high. Considering methanol carbonylation to give acetic acid (see Fig. 1 for a flow diagram for the process) provides an opportunity for posing specific process engineering problems: 1. The solutions containing catalyst and the halide cocatalyst are corrosive, and appro-

priate construction materials are required. 2. Separation by distillation is difficult because the product acetic acid is intermediate in volatility between the volatile cocatalyst (CH,I) and the high-boiling Rh complex, requiring a series of distillation columns. Design of the distillation train requires a knowledge of the stream compositions and flow rates: The side reactions must be-accounted for, i d the reactant flowiates and the conversion in the reactor must be chosen. The side reactions below occur, and to a first approximation they may be considered to reach equilibrium:

2 CH3OH H,O CH,COOH

HZO

+ CH3-O-CH3

(a)

+ C H 3 1 e H I + CH,OH + CH30H IHzO + CH3COOCH3

(b) (c)

Components of widely varying volatilities will be present. It is advantageous to recycle the volatile ether for ultimate conversion to methanol and then acetic acid. The methyl acetate might also be recycled, or taken off as the predominant product. The Rh carbonyl complexes present in the product stream must be stable at low CO partial pressures and the temperatures in the distillation columns. Complexes of the less expensive Co are catalytically active for methanol carbonylation in the presence of an iodide cocatalyst, but the carbonyl complexes require CO partial pressures of lo7N/m2 for stability, which complicates the product-catalyst separation. In the 0x0 process high pressures are required for the Co catalysts to maintain their stability. Product purification

-

Overhead recvcle

C

co_ m

Reactor

Light ends

+ i

Heavy recycle

#

t

uct

Water-acid recycley Heavy ends to incinerator

Recycle

Figure 1. Schematic flow diagram of methanol carbonylation proces~~'~).

36

&

14.2. Ty es of Catalysts ported Metal Complexes 14.2.4. 14.2.4.1. 8olymeric Supports

dominates the process engineeri~~g'.~. Rhodium complexes are preferred for propene hydroformylation, because they are stable at low pressures and can be separated from the products by distillation. Researchers have worked to alleviate the problems of separation and corrosion in processes such as the 0x0 process by designing catalysts that are confined in a separate phase from the reactants (see Section 14.2.4).A commercially successful approach for propene hydroformylation resulted from preparation of water-soluble rhodium complex catalysts by sulfonation of the phenyl rings of the triphenyl phosphine ligands6. The catalyst is used in a reactor with two liquid phases; the propene is concentrated in the organic phase and the catalyst in the aqueous phase near the interface. The CO H, is bubbled into a mixed reactor, and the two-phase liquid product flows to a settler; the organic product flows to downstream separation devices, and the aqueous phase with the catalyst is recycled to the reactor.

+

(B. C. GATES) 1. B. C. Gates, J. R. Katzer, G. C. A. Schuit, Chernistvy of Catalytic Processes, Chap. 2, McGrawHill, New York, 1979. 2. G. Astarita, Mass Transfer with Chemical Reaction, Elsevier, Amsterdam, 1967. 3. C. N. Satterfield, Mass Transfer in Heterogeneous Catalysis, MIT Press, Cambridge, MA, 1970. 4. (a) J. F. Roth, J. H. Craddock, A. Hershman, F. E. Paulik, CHEMTECH, I , 600 (1971); (b) F. E. Paulik, A. Hershman, W. R. Knox, J. F. Roth, U. S . Patent 3,769,329 (1973). 5 . J. Falbe, Carbon Monoxide in Organic Synthesis, transl. C. R. Adams, Chap. 1, Springer, New York, 1970. 6. H. Bach, W. Gick, W. Konkol, E. Wiebus, Proc. 9th Int. Cong. Catal., I , 254 (1988).

14.2.4. Supported Metal Complexes 14.2.4.1. Polymeric Supports

High selectivity and difficulty of separation from products are two quintessential characteristics of soluble catalysts. Attempts to retain the catalyst selectivity and alleviate the separation (and corrosion) problems have resulted in the preparation of catalysts consisting of molecular species bonded to solid supports. The common supports are organic polymers and porous inorganic solids, which are discussed in turn in the next two sections. Organic polymers offer several advantages as catalyst supports:

1. Polymers, especially those incorporating arene groups, can be easily functionalized for bonding of catalytic groups. 2. Hydrocarbon polymers are nearly inert supports; introduction of specific catalytic groups can lead to the formation of a selective catalyst. 3. Polymers can be prepared with wide ranges of physical properties and designed for optimum resistance to transport of reactant and product molecules and for optimum interactions between catalytic groups. Commonly applied polymeric supports are gels, loosely arranged and weakly structured macromolecular solids. A gel may swell to many times its collapsed volume when brought in contact with solvents for which it has a strong affinity. A swelled gel resembles a liquid in many ways, and catalytic groups bonded within a gel may act like catalytic groups in solution.


36

&

14.2. Ty es of Catalysts ported Metal Complexes 14.2.4. 14.2.4.1. 8olymeric Supports

dominates the process engineeri~~g'.~. Rhodium complexes are preferred for propene hydroformylation, because they are stable at low pressures and can be separated from the products by distillation. Researchers have worked to alleviate the problems of separation and corrosion in processes such as the 0x0 process by designing catalysts that are confined in a separate phase from the reactants (see Section 14.2.4).A commercially successful approach for propene hydroformylation resulted from preparation of water-soluble rhodium complex catalysts by sulfonation of the phenyl rings of the triphenyl phosphine ligands6. The catalyst is used in a reactor with two liquid phases; the propene is concentrated in the organic phase and the catalyst in the aqueous phase near the interface. The CO H, is bubbled into a mixed reactor, and the two-phase liquid product flows to a settler; the organic product flows to downstream separation devices, and the aqueous phase with the catalyst is recycled to the reactor.

+

(B. C. GATES) 1. B. C. Gates, J. R. Katzer, G. C. A. Schuit, Chernistvy of Catalytic Processes, Chap. 2, McGrawHill, New York, 1979. 2. G. Astarita, Mass Transfer with Chemical Reaction, Elsevier, Amsterdam, 1967. 3. C. N. Satterfield, Mass Transfer in Heterogeneous Catalysis, MIT Press, Cambridge, MA, 1970. 4. (a) J. F. Roth, J. H. Craddock, A. Hershman, F. E. Paulik, CHEMTECH, I , 600 (1971); (b) F. E. Paulik, A. Hershman, W. R. Knox, J. F. Roth, U. S . Patent 3,769,329 (1973). 5 . J. Falbe, Carbon Monoxide in Organic Synthesis, transl. C. R. Adams, Chap. 1, Springer, New York, 1970. 6. H. Bach, W. Gick, W. Konkol, E. Wiebus, Proc. 9th Int. Cong. Catal., I , 254 (1988).

14.2.4. Supported Metal Complexes 14.2.4.1. Polymeric Supports

High selectivity and difficulty of separation from products are two quintessential characteristics of soluble catalysts. Attempts to retain the catalyst selectivity and alleviate the separation (and corrosion) problems have resulted in the preparation of catalysts consisting of molecular species bonded to solid supports. The common supports are organic polymers and porous inorganic solids, which are discussed in turn in the next two sections. Organic polymers offer several advantages as catalyst supports:

1. Polymers, especially those incorporating arene groups, can be easily functionalized for bonding of catalytic groups. 2. Hydrocarbon polymers are nearly inert supports; introduction of specific catalytic groups can lead to the formation of a selective catalyst. 3. Polymers can be prepared with wide ranges of physical properties and designed for optimum resistance to transport of reactant and product molecules and for optimum interactions between catalytic groups. Commonly applied polymeric supports are gels, loosely arranged and weakly structured macromolecular solids. A gel may swell to many times its collapsed volume when brought in contact with solvents for which it has a strong affinity. A swelled gel resembles a liquid in many ways, and catalytic groups bonded within a gel may act like catalytic groups in solution.

14.2. Ty es of Catalysts 14.2.4. !& ported Metal Complexes 14.2.4.1. olymeric Supports

B

37

Polymers can also be prepared with rigid and porous structures. Solid polymers with high internal surface areas (approaching lo6 m2/kg) and pores significantly larger than the spaces between strands in a gel are called macroporous. Such polymers consist of aggregates of small gel-form particles. The most common polymeric supports are crosslinked polystyrenes; the usual crosslinking agent is divinylbenzene', Commercially available polymers include gels with crosslinks ranging from 1 to >10% divinylbenzene. Macroporous poly(styrene-diviny1benzene) and poly(styrene-ethylvinylbenzene-divinylbenzene)usually have higher crosslinking percentages. The preparations of the commercial polymers are not well documented. Many samples, especially of macroporous polymers, are contaminated with compounds used as emulsifiers in the process of polymerization, and these can interfere with syntheses designed to incorporate catalytic groups and also with the catalysis itself. Thorough extraction with solvents is necessary to prepare catalysts with well-defined catalytic groups. Alternatively, gel-form polymers can be prepared in the laboratory with or without water-soluble emulsifiers. Functional groups can be incorporated in crosslinked polystyrene by functionalized monomers with styrene and divinylbenzene'. The details of the polymerizaton are important. Reaction of styrene, divinylbenzene, and p-bromostyrene gives a nearly random copolymer, i.e., a nearly random distribution of -Br groups in the solid2. On the other hand, reaction of styrene and divinylbenzene with p-styryldiphenylphosphinegives a block copolymer having islands with high concentration of --PPh2 groups2. The common ligands in transition-metal-complex catalysts are incorporated directly into polymers like poly( styrene-divinylbenzene). For example, bromination of the rings proceeds well in the presence of T13+ ions3, and -CH2C1 groups can be incorporated by reaction with chloromethylmethyl ether, CAUTION: THIS ETHER IS A CARCINOGEN. The --CH2C1 groups can be easily converted with amines, e.g., by reaction with a secondary amine to form a tertiary amine. The -CH,Cl groups can be converted into phosphines by reaction with LiPPh,; bound -Br groups can similarly be converted into phosphines4. Metal complexes are incorporated into the functionalized polymers by simple ligand association or ligand exchange; some examples are shown in Fig. 1. Alternatively, with highly crosslinked macroporous polymers, the groups may be bonded almost exclusively near the internal surface (i.e,, on the surfaces of the aggregated gel particles).

Figure 1. Preparation of polymer-supported transition metal complexes by ligand exchange.

38

14.2. Ty es of Catalysts 14.2.4. !fu potted Metal Complexes 14.2.4.1. olymeric Supports

8

The metal complexes most often studied as polymer-bound catalysts have been Rh(1) complexes, such as analogues of Wilkinson's complex. The catalytic activity of a bound metal complex is nearly the same as that of the soluble analogue. Rhodium complexes are active for alkene hydrogenation, alkene hydroformylation, and, in the presence of CH,I cocatalyst, methanol carbonylation, etc. Polymer supports thus allow the chemistry of homogeneous catalysis to take place with the benefits of an insoluble, easily separated cataly~t'~~*~. The principal generalization is the similarity of soluble and anchored catalytic groups and the straightforwardness of design of supported catalysts, but there are also important differences between soluble and supported catalysts-the supports offer unique opportunities for influencing the catalyst performance5. Anchoring a catalyst to a rigid support can sometimes make possible a reaction that does not take place in solution. For example, catalysts have been prepared by functionalization of a highly crosslinked poly( styrene-divinylbenzene), as follows6:

TiCI,

Reduction with butyllithium forms a hydrogenation catalyst, presumably a sandwich complex of Ti with coordinative unsaturation to allow bonding of reactant hydrogen with alkene (or alkyne). The molecular analogue of this complex, titanocene, is ineffective as a catalyst in solution, since Ti-Ti bonds form and it oligomerizes or polymerizes, giving species lacking the coordinative unsaturation necessary for bonding of the reactants. This is an example of self-inhibition of a catalyst; the role of the rigid support is to stabilize coordinative unsaturation by holding nearby catalytic groups apart from each other. The physical properties of the support are critical; if the support is too flexible and solution-like, then the catalyst cannot function. Polymers are unique catalyst supports because their physical properties can be varied so widely. Polymers with low crosslinking (flexibility) have allowed preparation of highly substituted (chelated) metals. For example, 2% crosslinked poly(styrene-divinylbenzene), a block copolymer prepared with p-styryldiphenylphosphine,allows a tetrasubstituted cluster, H,Ru4(CO),(Ph,P),-@. A monosubstituted cluster was prepared with a polymer of randomly distributed -PPh, groups in low concentration'. When the phosphine substitution of Rh complexes is varied by changing the polymer physical properties, there are changes in the selectivity for alkene hydroformylation'. Because they are easily functionalized, polymers present good opportunities for design of bi- and multifunctional catalysts. For example, polymers incorporating palladium complexes and quinone groups,

n

14.2. Ty es of Catalysts 14.2.4. ported Metal Complexes 14.2.4.1. F!olyrneric Supports

&

39

catalyze the Wacker oxidation of ethene, the Pd2+ functioning as in solution and the quinone group playing the role of Cuz+ in solutiong. Poly( styrene-divinylbenzene) incorporating Rh complexes and pseudohalide groups analogous to CH,I,

c1 c1

SCH,

is a bifunctional catalyst for the carbonylation of methanol". Polymers incorporating Co phthalocyanine groups and amine groups catalyze the oxidation of mercaptans to disulfides' I, and polymers incorporating Rh complexes and amine groups catalyze three aldox reactions: propene hydroformylation, aldol condensation, and hydrogenation of the condensation product to give 2-ethylhexanalI2. Only a few simple polymer-supported catalysts have found industrial use. All are strong acids, sulfonated poly(styrene-divinylbenzene), and are used in processes including phenol alkylation, the phenol-acetone condensation reaction to give bisphenol A, and the conversion of methanol and isobutylene into methyl-t-butyl ether, a high-octane gasoline component. Polymers incorporating metal complexes, including the hydroformylation catalyst incorporating Rh(I), have undergone commercial evaluation but they lack stability, even when the metal is bonded to the support through more than one pendant ligand'. There is a slow leaching of Rh from the polymer into the reactant ~ o l u t i o n 'which ~ makes commercial application uneconomical. Polymers also lack mechanical stability and are not usable in stirred reactors, and they do not withstand the high temperatures that may result from exothermic reactions'. (B. C. GATES)

1. P. E. Garrou, B. C. Gates, in Synthesis and Separations Using Functional Polymers, D. C. Shemngton, P. Hodge, eds., Wiley, Chichester, 1988, p. 123. 2. J. Lieto, D. Milstein, R. L. Albnght, J. V. Minkiewicz, B. C. Gates, CHEMTECH, 13, 46 (1983). 3. M. J. Farall, J. M. J. Frechet, J . Org. Chem., 41, 3877 (1976). 4. C. Tamborski, F. E. Ford, W. L. Lehn, G. J. Moore, E. J. Soloski, J . Org. Chem., 27, 619 (1962). 5 . B. C. Gates, Catalytic Chemistry, Wiley, New York, 1992, Chap. 4. 6. W. D. Bonds, Jr., C. H. Brubaker, Jr., E. S . Chandrasekaran, C. Gibbons, R. H. Grubbs, L. C. Kroll, J . Am. Chem. SOC.,97, 2128 (1975). 7. Z. Otero-Schipper, J. Lieto, B. C. Gates, J. Catal., 63, 175 (1980). 8. C. U. Pittman, Jr., Q.Ng, A. Hirao, W. Honnick, R. Hanes, in Relations Entre Catalyse HomogPne et Catalyse Hitirogdne, Editions du Centre National de la Recherche Scientifique, Paris, 1978. Chem. Abst., 92, 136004m (1980). 9. H. Arai, M. Yashiro, J . Mol. Catal., 3, 427 (1978). 10. K. M. Webber, B. C. Gates, W. Drenth, J . Mol. Catal., 3, 1 (1977/78). 1 1 . J. Zwart, H. C. van der Weide, N. Brbker, C. Rummens, G. C. A. Schuit, A. L. German, J . Mol. Card., 3, 151 (1977/78). 12. R. F. Batchelder, B. C. Gates, F. P. J. Kuijpers, Proc. 6th lnt. Cong. Catal., The Chemical Society, London 1,499 (1977). Chem. Abst., 88,28239a (1978). 13. W. H. Lang, A. T. Jurewitz, W. 0. Haag, D. D. Whitehurst, L. D. Rollmann, J . Organomer. Chem., 134, 85 (1977).


40

14.2. Ty es of Catalysts 14.2.4. upported Metal Complexes 14.2.4.2. Metal Oxide Supports

14.2.4.2. Metal Oxide Supports

The instability of polymeric supports leads to the idea of using inorganic solids for supporting molecular catalytic species. Methods have been developed for functionalizing surfaces of SiO, and other metal oxides with groups such as phosphines’.’. Incorporation of a metal complex is straightforward; the methods are like those cited in the preceding section. However, even the least reactive surfaces of oxides-such as that of silica gelare more reactive then those of hydrocarbon polymers, and complications in catalyst synthesis and in the catalysis itself may be caused by the oxide surface. Amine, phosphine, thiol, and other groups can be incorporated onto metal oxide surfaces by reactions involving the surface -OH groups: E O H

+ Ph2P[CH2CH,CH2Si(OEt)3] (

-

+ ~O$iCH,CH,C€€,PPh, support.) I

EtOH

1represents the SiO,

(a)

A variety of Si0,-supported metal complexes have been prepared including Rh(1) complexes’, which are catalytically similar to their soluble and polymer-supported analogues. Detailed comparisons are lacking, and the stablities of the silica-supported catalysts are unreported. A silica-supported Rh(1) catalyst for hydroformylation has not found use. More useful than metal complexes bonded to oxide supports through organic tether ligands are metal complexes bonded directly to the oxide surface^^-^. Supported mononuclear metal complexes catalyze alkene polymerization, metathesis, and hydrogenation. Polymerization catalysts incorporating Cr or Zr complexes are used commercially for the manufacture of p~lyethylene~.~. Ziegler-type catalysts incorporating surface Ti complexes are used for the manufacture of stereoregular polypropylene. Preparation of a supported Cr complex may involve one or two surface -OH groups3:

Coordinatively unsaturated species (still unidentified) derived from these structures catalyze ethene polymerization. Analogous supported Mo and Re complexes are active for alkene metathesis. Supported polynuclear metal complexes have been formed on SiO,, A1,0,, and Mg04.6. The supported clusters catalyze alkene isomerization and CO hydrogenation. The supported metal clusters are of interest because they have unique structures with neighboring metal centers, intermediate between supported mononuclear metal complexes and metal crystallites. Catalysts in this class may offer interesting selectivities for reactions such as CO hydrogenation, but they are limited by their lack of stability. (B. C.GATES)

14.2. Ty es of Catalysts 14.2.5. fhase Transfer Catalysis

41

1. K. G . Allum, R. D. Hancock, S. McKenzie, R. C. Pitkethly, Proc. 5th Int. Cong. Card., The Chemical Society, London, I, 477 (1973). 2. L. L. Murrell, in J. J. Burton, R. L. Garten, eds., Advanced Materials in Catalysis, Academic

Press, New York, 1977.

3. Yu. I. Yennakov, B. N. Kuznetsov, V. A. Zakharov, Catalysis by Supported Complexes, Else-

vier, Amsterdam, 1981. 4. H. H. Lamb,B. C. Gates, H. Kntkinger, Angew. Chem. Inf.Ed. Engl., 27, 1127 (1988). 5. D. G . H. Ballad, Adv. Catal., 23, 263 (1973). 6. B. C. Gates, L. Guczi, H. Knozinger, eds., Metal Clusters in Catalysis, Elsevier, Amsterdam, 1986.

14.2.5. Phase Transfer Catalysis Phase transfer catalysts transport reactants across a liquid-liquid interface 1-3, and are used when the desired reaction involves reactants with far different solubility characteristics, and which, therefore, cannot be brought simultaneously to a high concentration in any single liquid phase. The use of phase transfer catalysis may eliminate the need for anhydrous solvents. An alkyl halide, RX (soluble in an organic phase, but not in an aqueous phase), and an inorganic ion, such as OH- or X - (soluble in water but not in the organic phase) can react near an interface between the aqueous and organic phases, and a phase transfer catalyst can greatly accelerate the reaction by bringing the reaction partners into efficient contact. The mechanism of action of a phase transfer catalyst for the reaction:

RX

+ Y - - + R Y + X-

(a)

is shown schematically in Fig. 1A. The phase transfer agent is Q', which has a strong affinity for an organic solvent, such as a tetraalkylphosphonium ion [e.g., C16H33P+(C,H,),], a tetraalkylammonium ion, or a complexing agent like a crown ether capable of solubilizing organic and inorganic alkali metal salts even in nonpolar organic solvents; the ethers function by providing the cation with an organic mask by complex formation:

In Fig. lA, Y - is more lipophilic than X - . The cation Q' migrates with the anion Y - from the aqueous phase to the phase boundary (the liquid-liquid interface) and into the organic phase, where it comes in contact with RX. In the organic phase, the ion pairs Q c Y - are only slightly solvated or aggregated, and the Y - ions are, therefore, exposed and highly reactive with RX. After reaction the catalytic cation Q' forms an ion pair with X - (the reaction product) and migrates back across the interface into the polar

phase. Properly chosen Q' ions can have such a strong affinity for the organic phase that they exist only there and near the interface. The simplified representation of the process shown in Fig. 1B thus approximates the role of Q'. This ion associates with Y - ions


14.2. Ty es of Catalysts 14.2.5. fhase Transfer Catalysis

41

1. K. G . Allum, R. D. Hancock, S. McKenzie, R. C. Pitkethly, Proc. 5th Int. Cong. Card., The Chemical Society, London, I, 477 (1973). 2. L. L. Murrell, in J. J. Burton, R. L. Garten, eds., Advanced Materials in Catalysis, Academic

Press, New York, 1977.

3. Yu. I. Yennakov, B. N. Kuznetsov, V. A. Zakharov, Catalysis by Supported Complexes, Else-

vier, Amsterdam, 1981. 4. H. H. Lamb,B. C. Gates, H. Kntkinger, Angew. Chem. Inf.Ed. Engl., 27, 1127 (1988). 5. D. G . H. Ballad, Adv. Catal., 23, 263 (1973). 6. B. C. Gates, L. Guczi, H. Knozinger, eds., Metal Clusters in Catalysis, Elsevier, Amsterdam, 1986.

14.2.5. Phase Transfer Catalysis Phase transfer catalysts transport reactants across a liquid-liquid interface 1-3, and are used when the desired reaction involves reactants with far different solubility characteristics, and which, therefore, cannot be brought simultaneously to a high concentration in any single liquid phase. The use of phase transfer catalysis may eliminate the need for anhydrous solvents. An alkyl halide, RX (soluble in an organic phase, but not in an aqueous phase), and an inorganic ion, such as OH- or X - (soluble in water but not in the organic phase) can react near an interface between the aqueous and organic phases, and a phase transfer catalyst can greatly accelerate the reaction by bringing the reaction partners into efficient contact. The mechanism of action of a phase transfer catalyst for the reaction:

RX

+ Y - - + R Y + X-

(a)

is shown schematically in Fig. 1A. The phase transfer agent is Q', which has a strong affinity for an organic solvent, such as a tetraalkylphosphonium ion [e.g., C16H33P+(C,H,),], a tetraalkylammonium ion, or a complexing agent like a crown ether capable of solubilizing organic and inorganic alkali metal salts even in nonpolar organic solvents; the ethers function by providing the cation with an organic mask by complex formation:

In Fig. lA, Y - is more lipophilic than X - . The cation Q' migrates with the anion Y - from the aqueous phase to the phase boundary (the liquid-liquid interface) and into the organic phase, where it comes in contact with RX. In the organic phase, the ion pairs Q c Y - are only slightly solvated or aggregated, and the Y - ions are, therefore, exposed and highly reactive with RX. After reaction the catalytic cation Q' forms an ion pair with X - (the reaction product) and migrates back across the interface into the polar

phase. Properly chosen Q' ions can have such a strong affinity for the organic phase that they exist only there and near the interface. The simplified representation of the process shown in Fig. 1B thus approximates the role of Q'. This ion associates with Y - ions

14.2. Ty 8s of Catalysts 14.2.5. fhase Transfer Catalysis

42

AQUEOUS PHASE interface I

A

ORGANIC PHASE

RX

interface

RX

ORGANIC PHASE

Figure 1. Schematic representations of phase transfer catalysis. at the interface, transfers them in ion pairs to the organic phase, where they react with RX and generate X-;Q' then associates with X- and transfers it to the interface, where it exchanges X- for another Y-. Concentration gradients provide the driving force for the transport processes and, therefore, for the catalysis. The key to the high reactivity of the ion pairs in weakly polar organic solvents is their weak solvation. Examples of phase transfer catalysis include hydrolysis, condensation, oxidation, and p~lymerization'-~. There are many industrial applications. The following reaction C8H17C1

+ NaCN

aqueous NaCN NR4X or pR4x'

C8H,,CN

+ NaCl

proceeds quantitatively within hours under the influence of phase transfer catalysis, whereas without a catalyst, no product is formed even after weeks. Phase transfer catalysis is also useful for the generation of highly reactive dichlorocarbene: CH3Cl

+ NaOH Q', C1,C:

(c)

to form nitriles, isonitriles, and chlorides in high yields. Phase transfer catalysis is a significant advance in preparative organic chemistry, but there is a practical limitation-phase transfer agents sometimes stabilize emulsions, which make product recovery difficult. A variation of the technique, called triphase catalysis, involves bonding the phase transfer agent to a support, such as a gel-form polymer (see 14.2.4.1)4. If the polymer has an affinity for both liquid phases, the phase

14.2. Ty es of Catalysts 14.2.6. tatalysis in Microscopic Phases

43

transfer agent can function much as it would in the absence of the polymer, and the product workup is simplified. (B. C.GATES) 1. C. M. Starks, C. Liotta, Phase Transfer Catalysis-Principles and Techniques, Academic Press, New York, 1978. 2. W. P. Weber, G. W. Gokel, Phase Transfer Catalysis in Organic Synthesis, Springer-Verlag, New York, 1977. 3. H. Alper, Adv. Organomet. Chem., 19, 183 (1981). 4. S . L. Regen, Angew. Chem., Int. Ed. Engl., 18, 421 (1979).

14.2.6. Catalysis in Microscopic Phases Restricting catalysis to solutions and surfaces omits a large number of intriguing and useful catalysts. Enzymes represent the ultimate in catalyst performance, and they are difficult to categorize in traditional terms (see 14.2.1).The catalytic sites of enzymes are flexible three-dimensional clefts where catalytic functional groups are precisely positioned to interact with multifunctional reactant molecules, to the exclusion of virtually all other molecules. Synthetic bi- and multifunctional polymers (see 14.2.4.1)mimic enzymes. Both natural and synthetic organic and organometallic polymer catalysts have the character of solutions, being flexible and able to conform to reactant molecules and bond to them at more than one position, and they are also often insoluble and therefore restricted to a phase separate from reactants and products. Other catalysts have characteristics intermediate between those of solutions and surfaces. These are described below. 1. Zeolites, crystalline solids with regular, molecular-scale pores that incorporate cata-

lytic groups, exert unique catalytic influence because of the molecular-sieving properties of their pore structures and the solvent-like character of their narrow pores. 2. Clays and graphite, materials with layer structures within which catalytic groups can be held, also have catalytic properties strongly influenced by the solvent-like nature of the intralayer space. 3. Supported liquid-phase catalysts, porous solids with liquids held in the narrow pores, function even at high temperatures. 4. Micelles, discrete aggregates of organic molecules or ions that can incorporate catalytic groups, act as individual “microreactors,” their performance being influenced by the distribution of reactants between the micelle and the surrounding liquid. These catalysts provide three-dimensional microscopic media for reaction. The most Faujasites are used to crack petroleum for the important are the zeolites (see 14.2.2.2). manufacture of gasoline, and HZSM-5 to convert methanol into gasoline and in other petrochemical conversion processes including xylene isomerization and toluene disproportionation’. These are examples of acid-base catalysis. The zeolites are distinct catalytically from their amorphous analogues (silica-aluminas), because they are molecular sieves; in ZSM-5, for example, molecules larger than durene (1,2,4,5-tetramethylbenzene) cannot form and migrate through the narrow pores. They are therefore absent from the product formed from methanol, which is cut off near the desired high-boiling end of the gasoline range’.


14.2. Ty es of Catalysts 14.2.6. tatalysis in Microscopic Phases

43

transfer agent can function much as it would in the absence of the polymer, and the product workup is simplified. (B. C.GATES) 1. C. M. Starks, C. Liotta, Phase Transfer Catalysis-Principles and Techniques, Academic Press, New York, 1978. 2. W. P. Weber, G. W. Gokel, Phase Transfer Catalysis in Organic Synthesis, Springer-Verlag, New York, 1977. 3. H. Alper, Adv. Organomet. Chem., 19, 183 (1981). 4. S . L. Regen, Angew. Chem., Int. Ed. Engl., 18, 421 (1979).

14.2.6. Catalysis in Microscopic Phases Restricting catalysis to solutions and surfaces omits a large number of intriguing and useful catalysts. Enzymes represent the ultimate in catalyst performance, and they are difficult to categorize in traditional terms (see 14.2.1).The catalytic sites of enzymes are flexible three-dimensional clefts where catalytic functional groups are precisely positioned to interact with multifunctional reactant molecules, to the exclusion of virtually all other molecules. Synthetic bi- and multifunctional polymers (see 14.2.4.1)mimic enzymes. Both natural and synthetic organic and organometallic polymer catalysts have the character of solutions, being flexible and able to conform to reactant molecules and bond to them at more than one position, and they are also often insoluble and therefore restricted to a phase separate from reactants and products. Other catalysts have characteristics intermediate between those of solutions and surfaces. These are described below. 1. Zeolites, crystalline solids with regular, molecular-scale pores that incorporate cata-

lytic groups, exert unique catalytic influence because of the molecular-sieving properties of their pore structures and the solvent-like character of their narrow pores. 2. Clays and graphite, materials with layer structures within which catalytic groups can be held, also have catalytic properties strongly influenced by the solvent-like nature of the intralayer space. 3. Supported liquid-phase catalysts, porous solids with liquids held in the narrow pores, function even at high temperatures. 4. Micelles, discrete aggregates of organic molecules or ions that can incorporate catalytic groups, act as individual “microreactors,” their performance being influenced by the distribution of reactants between the micelle and the surrounding liquid. These catalysts provide three-dimensional microscopic media for reaction. The most Faujasites are used to crack petroleum for the important are the zeolites (see 14.2.2.2). manufacture of gasoline, and HZSM-5 to convert methanol into gasoline and in other petrochemical conversion processes including xylene isomerization and toluene disproportionation’. These are examples of acid-base catalysis. The zeolites are distinct catalytically from their amorphous analogues (silica-aluminas), because they are molecular sieves; in ZSM-5, for example, molecules larger than durene (1,2,4,5-tetramethylbenzene) cannot form and migrate through the narrow pores. They are therefore absent from the product formed from methanol, which is cut off near the desired high-boiling end of the gasoline range’.

44

14.2. Ty es of Catalysts 14.2.6. 8atalysis in Microscopic Phases

The narrowness of the zeolite pores also influences their catalytic nature. A faujasite in the K + form, even though it is not acidic, catalyzes the cracking of alkanes'. The product distribution in cracking is consistent with free-radical intermediates and the induced homolytic rupture of C-H or C-C bonds. The activity arises from the strong electric fields of the ions in the confining pores. The high activity of the zeolites in cracking by acid groups via carbenium ion intermediates is associated with the high local concentrations of reactants in the pores near the catalytic sites, and the high selectivities for formation of gasoline-range hydrocarbons result from the high rates of bimolecular H-transfer reactions in the microenvironment of the pores. Acidic zeolites are most commonly used, but metal-containing zeolites also have catalytic properties. For example, ion exchange of Rh into faujasites produces methanol carbonylation catalysts; the Rh complexes work in much the same way as the analogous soluble catalyst (see 14.2.3.l)3,4. Metals introduced into zeolites by ion exchange have been reduced to give metal clusters in the cages. These catalysts are used in hydrocracking operations; they are bifunctional with Pd catalyzing dehydrogenation and the acidic zeolite catalyzing carbenium-ion reactions. Zeolite L containing Pt clusters of only about five atoms, on average, are soon to be applied industrially for catalytic dehydrocyclization of n-alkanes to give aromatics5. Zeolites containing small carbonyl clusters of metals like 0 s and Rh work in CO hydrogenation, where the small metal clusters and confining nature of the pores contribute to high selectivities for low-molecular-weight alkenes6. Aluminosilicate clays, like zeolites, offer large intracrystalline volumes where catalysis can be influenced by effects like molecular sieving and solvation. Clays with acidic groups have been used, for example, for cracking hydrocarbons that are too large to enter zeolite pores7. Methods have been devised for incorporating catalytically active Rh complexes between the layers of clays by use of cationic ligands*. The performance of catalytic groups is changed by incorporating them in the spaces between the charged layers of these swellable materials. Another layered material into which catalytic groups have been incorporated is graphite. The lamellar C,K catalyzes reactions such as hydrogenation of alkenes and of Cog. Lewis acids like SbF, can also be incorporated into the intralayer spaces for catalysis'O. The related supported liquid-phase catalysts consist of traditional support materials such as -y-A1,0, having micropores filled with solvent and a dissolved catalyst. In small pores, because of the Kelvin effect, the vapor pressure of the solvent is small so that it will remain in the pore as a liquid, even when the catalyst is used at a high temperature in flowing vapor-phase reactants' ',l'. These catalysts are active for alkene hydroformylation; the soluble catalyst can thus be used without the complications of corrosion and difficult separation from products-provided that it is stable (cf. 14.2.4). Organic polymers are comparable to the above catalysts, having microenvironments different from those of the surrounding media and being swellable (see 14.2.4.1).Micelles, which are colloidal species produced by aggregation of ca. 20 to thousands of surfactant molecules or ions with both polar and nonpolar portions, also have these characteristics. In a typical aggregate, the hydrophobic ends of the molecules are clustered in the core of the micelle, and the polar ends are located at the interface to the aqueous phase. Micelles may be spherical or (in highly concentrated solutions) cylindrical or lamellar. Inverted micelles may form in a hydrocarbon solvent.

14.2. Ty es of Catal sts 14.2.7. froduct ion oYCatalysts and Supports 14.2.7.1. General Principles

45

Micelles concentrate reactants from the surrounding medium and provide microenvironments favorable to rea~tion'~. Rate enhancement arises from electrostatic and hydrophobic interactions between reactants and micelles. Rates may be strongly dependent on the struture of the reactant. For example, the hydrolysis of methyl orthobenzoate is catalyzed by micellar sodium dodecylsulfate, whereas the hydrolysis of methyl orthoformate (which has less hydrophobic character and less affinity for the micelle core) is not14. The more pronounced the hydrophobic nature of the reactant, the more rapid is the micellar cataly~is'~. Rates may be increased by phase transfer agents (see 14.2.5), and specific catalytic groups may be incorporated in the micelles. For example, surfactant Rh complexes are used in micelles to catalyze hydrogenaton and hydroformylation of alkenes". (6.C.GATES) 1 . P. B. Weisz, Proc. 7th Int. Cong. Catal., I , 1, 1981; Chem. Abst. 95, 192859t (1981). 2. J. A. Rabo in F. R. Ribiero, A. E. Rodrigues, L. D. Rollmann, C. Naccache, eds., Zeolites: Science and Technology, Martinius Nijhoff, The Hague, 1984. p. 237. 3. I. Maxwell, Adv. Catal., 31, 1 (1982). 4. P. Gelin, Y. Ben Taarit, C. Naccache, Proc. 7th Int. Cong. Catal., B, 898 1981. Chem. Abst., 92, 83183k (1980). 5. M. Vaarkamp, J. von Grondelle, J. T. Miller, D. J. Sajkowski, F. S . Modica, G . S . Lane, G . W. Zajac, B. C. Gates, D. C. Koningsbarger, Catal. Lett., 6, 369 (1990). 6. P.-L. Zhou, S. D. Maloney, B. C. Gates,J. Cafal.,129,315 (1991). 7. J. Shabtai, R. Lazar, A. G . Oblad, Proc. 7th Int. Cong. Catal., B, 828, 1981; Chem. Abst., 95, 210388~(1981). 8. W. H. Quayle, T. J. Pinnavaia, Inorg. Chem., 18,2840 (1979). 9. K. Tamaru, Am. Sci., 60,474 (1972). 10. P. G. Rodewald, U. S. Patent 3,984,352 (1976); Chem. Abst., 86 124,091a (1977). 11. P. R. Rony, J . Catal., 14, 142 (1969). 12. C. N. Kenney in ACS Symp. Ser. 72, D. Luss, V. W. Weekman, eds., American Chemical Society, Washington, D.C., 1980. 13. J. H. Fendler, E. Fendler, Catalysis in Micellar and Macromolecular Systems, Academic Press, New York, 1975. 14. M. T. A. Behme, E. H. Cordes, J . Am. Chem. Soc., 87,268 (1965). 15. Y. Dror, J. Manassen, Proc. 7th Int. Cong. Card., B , 887 1981; Chem. Abst., 87, 67885g (1978).

14.2.7. Production of Catalysts and Supports 14.2.7.1. General Principles

Catalyst design and production are the judicious application of available knowledge of the effect of production variables on catalyst structure and of the relationship between catalyst structure and performance. The usefulness of a catalyst is governed by certain performance requirements. The chief requirement is selectivity, the capability to accelerate reaction rates of desired reactions. Selectivity also includes the ability to minimize side reactions that are particularly deleterious, for example, those that lead to deactivating coke formation. Second, catalysts must be sufficiently active at reasonable temperatures for commercial application. There are several important physical requirements for useful catalysts. They must have sufficient strength to resist breakage or abrasion in fixed or fluid bed applications,


14.2. Ty es of Catal sts 14.2.7. froduct ion oYCatalysts and Supports 14.2.7.1. General Principles

45

Micelles concentrate reactants from the surrounding medium and provide microenvironments favorable to rea~tion'~. Rate enhancement arises from electrostatic and hydrophobic interactions between reactants and micelles. Rates may be strongly dependent on the struture of the reactant. For example, the hydrolysis of methyl orthobenzoate is catalyzed by micellar sodium dodecylsulfate, whereas the hydrolysis of methyl orthoformate (which has less hydrophobic character and less affinity for the micelle core) is not14. The more pronounced the hydrophobic nature of the reactant, the more rapid is the micellar cataly~is'~. Rates may be increased by phase transfer agents (see 14.2.5), and specific catalytic groups may be incorporated in the micelles. For example, surfactant Rh complexes are used in micelles to catalyze hydrogenaton and hydroformylation of alkenes". (6.C.GATES) 1 . P. B. Weisz, Proc. 7th Int. Cong. Catal., I , 1, 1981; Chem. Abst. 95, 192859t (1981). 2. J. A. Rabo in F. R. Ribiero, A. E. Rodrigues, L. D. Rollmann, C. Naccache, eds., Zeolites: Science and Technology, Martinius Nijhoff, The Hague, 1984. p. 237. 3. I. Maxwell, Adv. Catal., 31, 1 (1982). 4. P. Gelin, Y. Ben Taarit, C. Naccache, Proc. 7th Int. Cong. Catal., B, 898 1981. Chem. Abst., 92, 83183k (1980). 5. M. Vaarkamp, J. von Grondelle, J. T. Miller, D. J. Sajkowski, F. S . Modica, G . S . Lane, G . W. Zajac, B. C. Gates, D. C. Koningsbarger, Catal. Lett., 6, 369 (1990). 6. P.-L. Zhou, S. D. Maloney, B. C. Gates,J. Cafal.,129,315 (1991). 7. J. Shabtai, R. Lazar, A. G . Oblad, Proc. 7th Int. Cong. Catal., B, 828, 1981; Chem. Abst., 95, 210388~(1981). 8. W. H. Quayle, T. J. Pinnavaia, Inorg. Chem., 18,2840 (1979). 9. K. Tamaru, Am. Sci., 60,474 (1972). 10. P. G. Rodewald, U. S. Patent 3,984,352 (1976); Chem. Abst., 86 124,091a (1977). 11. P. R. Rony, J . Catal., 14, 142 (1969). 12. C. N. Kenney in ACS Symp. Ser. 72, D. Luss, V. W. Weekman, eds., American Chemical Society, Washington, D.C., 1980. 13. J. H. Fendler, E. Fendler, Catalysis in Micellar and Macromolecular Systems, Academic Press, New York, 1975. 14. M. T. A. Behme, E. H. Cordes, J . Am. Chem. Soc., 87,268 (1965). 15. Y. Dror, J. Manassen, Proc. 7th Int. Cong. Card., B , 887 1981; Chem. Abst., 87, 67885g (1978).

14.2.7. Production of Catalysts and Supports 14.2.7.1. General Principles

Catalyst design and production are the judicious application of available knowledge of the effect of production variables on catalyst structure and of the relationship between catalyst structure and performance. The usefulness of a catalyst is governed by certain performance requirements. The chief requirement is selectivity, the capability to accelerate reaction rates of desired reactions. Selectivity also includes the ability to minimize side reactions that are particularly deleterious, for example, those that lead to deactivating coke formation. Second, catalysts must be sufficiently active at reasonable temperatures for commercial application. There are several important physical requirements for useful catalysts. They must have sufficient strength to resist breakage or abrasion in fixed or fluid bed applications,

14.2. Ty es of Catalysts 14.2.7. production of Catalysts and Supports 14.2.7.1. General Principles

46

possess a suitable size and shape to permit flow of reactant fluids without undue pressure drop, and have a porosity and pore-size distribution, sometimes bimodal, which avoids diffusional limitations for desired reactions. For catalysts containing molecular sieves, pore sizes must be suitable to confer shape selectivity for reactant and product molecules. In addition, catalysts must have high thermal stability to resist sintering, particularly where there is a need for periodic regeneration by combustion of coke deposited on the catalyst. A further desirable characteristic is the ability to resist deactivating influence of poisons, notaby sulfur compounds, which are often present in reactant streams. Several techniques have been utilized in catalyst production to meet these requirements. Indeed, catalyst preparation must be a critical consideration in every technical activity involving catalysts'-' The successful discovery and development of industrial catalysts have been achieved sometimes through serendipity or intuition, but mainly based on a knowledge of reaction mechanisms and of catalyst constituents that provide functionality for such mechanisms. The classical approach to catalyst research has been to vary composition and method of catalyst preparation and to correlate these variables with characterization measurements and with catalyst performance. Modifications are then made in catalyst preparations based on this information and a general knowledge of the effect of catalyst constituents on the chemistry of desired and undesired reactions. Several catalyst modifications to improve catalyst selectivity have been identified and these serve to assist in catalyst improvements (Fig. 1)l1. Catalyst characterization tests include measurement of surface areas, chemisorption, pore-size distributions, crystal structure as determined by X-ray crystallography, reaction mechanisms as revealed by kinetics, and isotopic tracers and diagnostic catalytic reactions to test functional capabilities. These have been interpreted in terms of variation of catalyst preparation-structure-performance relationships. Recently, new instrumental techniques have become available for determining surface structures on an atomic scale. These include X-ray fine structure (EXAFS), electron spectroscopies, ultraviolet photoelectron spectroscopy (UPS), X-ray photoelectron spectroscopy (XPS), Auger electron spectroscopy (AES), ion spectroscopies, secondary-ion

'.

to poison undesired activity

to inhibit consecutive reactions

to modify surface access

to limit molecular size

inert selective adsorbed poison

inert adsorption competitor for product

inert optically active adsorbent

catalyst in molecular size cage

HCI in C3H4 oxidation

CO in C2H2 hydrogenation

no commercial application at present

ZSM 5 zeolite for CH,OH+ gasoline

14.2. Ty es of Catal sts 14.2.7. firoduction oYCatalysts and Supports 14.2.7.1. General Principles

47

mass spectroscopy (SIMS), and ion scattering spectroscopy (ISS). Nuclear magnetic resonance (NMR) measurements have provided information the nature of catalytic complexes. Scanning tunneling microscopy (STM) shows atoms as they are on the surface, providing topographical information at the atomic level. Finally, catalyst design is being assisted by computer graphics and artificial intelligence techniques. A survey of elements involved in catalysis amounts to a survey of the periodic table. Moreover, combinations of elements for catalytic purposes are numerous. The wealth of materials is expanded even further owing to the ability of many elements, especially transition elements, to exist in a large number of valence states. The unusually wide array of catalyst compositions which have been developed for the many major industrial catalytic applications is illustrated in Table 1. Some important catalysts are specific compounds, such as the inorganic acids HF, H,S04, H,P04, or salts such as AlCl,. The Wacker catalyst is a combination of two compounds, PdC1, and CuCl,. A more complex example is the Ziegler-Natta catalyst, formed by the interaction of TiCl, and Al(C,H,),, a catalyst famous for its capability to bring about stereospecific olefin polymerization. Many transition metal complexes, often containing CO, have found important catalytic applications. One of the most important is cobalt carbonyl, CO,(CO)~,where the active form is believed to be HCo(CO),. Other catalytically active complexes are RhC1CO[P(C6H5),],, and cluster catalysts such as [Rh,3(CO),H,]2-ion. These catalysts are often referred to as homogeneous catalysts because they and the reactants operate in the same liquid phase. However, to facilitate separation of reaction products from the catalyst, some of these catalytically active complexes have been attached or anchored onto solid surfaces. In contrast to these specific compounds, most catalysts are solid oxides or sulfides possessing high surface areas, generally 50-500 m2/g. Some catalysts contain finely divided constituents in metallic form deposited on the surface of a support. While chemical composition is a major factor in determining catalyst properties, catalytic characteristics may vary widely depending on method of catalyst preparation, due to the nature of interaction of catalyst components, their dispersion, crystallographic form, and physical properties such as surface area and pore structure. The composition of the surface is of special importance. Usually only a small fraction of the surface, sometimes called active sites, has a special composition and structure that are responsible for catalyst activity. Hence, small amounts of strategic components situated on the surface can have a profound influence. Catalysts which rely on their acidity can be enhanced during operation by addition of small amounts of H20 or HCl or deactivated by bases such as NH, or Na,O. There are several techniques for preparation of high-area catalysts. Some involve formation of a support or carrier, especially alumina, silica or carbon, onto whose surface is deposited an active catalyst ingredient. However, seldom is the support inactive in the sense that it functions only to spread out the active component. The support usually influences the added ingredient through epitaxial6 or chemical interaction which alters the behavior of the active component. In SiO2-Al20, catalysts, it is the combination of both oxides that provides for the essential acidity. In dual-function catalysts, the support can serve catalytically as the essential acid function. (G.A. MILLS)

Methanol synthesis

1920

I940

1930

Ammonia synthesis Liquid hydrocarbons via coal hydrogenation

1910

Aldehydes and alcohols via hydroformylation of olefin Alkylation of isoparaffin Synthetic rubber Reforming of naphtha

Polyethylene production

Ethylene oxide synthesis

Catalytic cracking

Reaction Catalyst

Heterogeneous/ homogeneous (Wh)

B

Liquid hydrocarbons from synthesis gas

Butter substitutes via fat hydrogenation Methane from synthesis gas

Process

1900

Decade First Commercialized

TABLE1. INDUSTRIAL APPLICATIONS OF CATALYSIS

48 14.2. Ty 8s of Catal sts 14.2.7. roduction o Catalysts and Supports 14.2.7.1. General Principles

Y

+

Acrylonitrile via ammodoxidation Cracking using zeolites Xylene isomerization Toluene disproportionation Olefins by disproportionation Adiponitrile from butadiene/HCN C4H,

+ 2HCN + NC(CH,),CN

2C3H,-+C*H4

Toluene + benzene

+ xylene + C4HS

CH,=CHCH3 + NH, + 3/202+ CH,=CHCN + 3H,O Petroleum + high yields gasoline o,m-Xylene --* p-xylene

+

+

1960

CsH1.5

Stereospecific polymerization "C3H5 (C,H,)" Acetaldehyde from ethylene C,H4 l/2 0,-+ CH3CH0 Hydrocrackingaf Heavy oil H, -+ gasoline petroleum Hydrodesulfurizationof Oil-S H2 -+ oil + H,S petroleum Phthalic acid via p-xylene oxidation H,CPhCH, + 30, -+ HO,CPhCO,H + 3H,O OCNRCO + HOR'OH -+ -[RNHCOOROOCNH]-Polyurethane

+

2C4H8

1950

Paraffin isomerization

Polymerization of otefins Olefins/diolefins by dehydrogenation

+ -

1950s)

h

H

Mo or W on M2O3 Ni-ligand4

H

ZSM 5 zeolite

H H H

h h

Co,Mo complexes N(C2H4),N Bi,O,-MoO, Y Zeolite Si0,-Al,O, ZSM 5 zeolite

H

H

W

Co-(Ni)-Mo/Al,O,

Co-Mo/Al,0,,-SiO~/Al,03

PdCI,-CuCI,

TiCl,-AI(C,H5),

h

H H H H

F%03-K20-Crz03 cr203/A1203 HCl-AICI, (Pt(Cl)/Al,O,

H.

H3P04-kieselguhr

1980

Gasoline from methanol Acetic anhydride via synthesis gas Middle distillate synthesis

Acetic acid via carbonylation Ldopa Emission control catalysts Ether fuels

Process

Reaction

Catalyst

Heterogeneous/ homogeneous (Wh)

14.2. Ty es of Catal sts 14.2.7. roduction o Catalysts and Supports 14.2.7.1. General Principles

B

1970

Decade First Commercialied

TABLE1. (continued)

50

Y

14.2.7. Production of Catalysts and Supports 14.2.7.2. Methods of Production on Nonmetal Catalysts and Supports 14.2.7.2.1. Precipitation and Gel Formation.

51

1. B. Delmon, P. A. Jacobs, C. Poncelet, eds., Preparation of Catalysts, Vol. I, Elsevier, New York, 1976. 2. B. Delmon, P. Grange, P. Jacobs, P. G. Poncelet, eds., Preparation of Catalysts, Vol. 11, Elsevier, New York, 1979. 3. G. Poncelet, P. Grange, P. A Jacobs, eds., Preparation of Catalysts, Vol. 111, Elsevier, New York, 1983. 4. B. Delmon, P. Grange, P. A. Jacobs, G. Poncelet, eds., Preparation of Catalysts, Vol. IV, Elsevier, New York, 1987. 5 . G. Poncelet, P. A. Jacobs, P. Grange, B. Delmon, eds., Preparation of Catalysts, Vol. V, Elsevier, New York, 1991. 6. A. B. Stiles, ed., Catalyst Supports and Supported Catalysts, Butterworths, London, 1987. 7. D. L. Trimm, Design ofIndustria/ Catalysts, Elsevier, New York, 1980. 8. R. B. Anderson and P. T. Dawson, eds., Vol. 11, Experimental Methods in Catalytic Research; Preparation and Examination of P ractical Catalysts, Academic Press, New York, 1976. 9. C. N. Satterfield, Heterogeneous Cataysis in Industrial Practice, 2nd ed., McGraw Hill, New York, 1991. 10. M. V. Twigg, ed., Catalyst Handbook, 2nd ed., Wolfe Pub. Co. Ltd., England, 1989. 11. S. P. S. Andrew, Chem. Engin. Sci., 36, 1431 (1981).

14.2.7.2. Methods of Production of Nonmetal Catalysts and Supports 14.2.7.2.1. Precipitation and Gel Formation.

The production of catalyst particles of suitable configuration and hardness is an essential part of catalyst manufacture'-*. Most heterogeneous catalysts are produced by processes that involve formation of solids from aqueous solutions3. Precipitation is frequently employed in preparation of hydrous oxide catalysts. To avoid occluded o r adsorbed impurities, ammonia or ammonium salts are often used as well as nitrates of the desired metal constituents. Calcination removes the nitrogen-containing components. Anions such as C1- or SO2- or cations such as Na+ are avoided; these often are poisons if they are present in the final catalyst. Despite improvements made in precipitation methods, precipitation rarely produces homogeneous products. The difference between overall composition of the initial solution and the composition of the precipitate at various stages indicates heterogeneity of the latter. The composition of the precipitate cannot be prevented from varying from the beginning to the end of the precipitation (Fig. 1). Gel formation is especially well suited to the preparation of hydrous oxide catalysts such as those of Si, Al, Fe. Also, an all-embracing gel can be used to produce the catalyst particle in a preferred size and spherical form by allowing droplets of the recently rapidly

solutions hydroxides 2.07

F e f samples ~

1.97

2.02

1.99

2.06 2.15

Figure 1. Ferrites prepared by coprecipitation of hydroxides. Atomic ratio Fe(M in coprecipitate and mixtures of starting solutions?



51

1. B. Delmon, P. A. Jacobs, C. Poncelet, eds., Preparation of Catalysts, Vol. I, Elsevier, New York, 1976. 2. B. Delmon, P. Grange, P. Jacobs, P. G. Poncelet, eds., Preparation of Catalysts, Vol. 11, Elsevier, New York, 1979. 3. G. Poncelet, P. Grange, P. A Jacobs, eds., Preparation of Catalysts, Vol. 111, Elsevier, New York, 1983. 4. B. Delmon, P. Grange, P. A. Jacobs, G. Poncelet, eds., Preparation of Catalysts, Vol. IV, Elsevier, New York, 1987. 5 . G. Poncelet, P. A. Jacobs, P. Grange, B. Delmon, eds., Preparation of Catalysts, Vol. V, Elsevier, New York, 1991. 6. A. B. Stiles, ed., Catalyst Supports and Supported Catalysts, Butterworths, London, 1987. 7. D. L. Trimm, Design ofIndustria/ Catalysts, Elsevier, New York, 1980. 8. R. B. Anderson and P. T. Dawson, eds., Vol. 11, Experimental Methods in Catalytic Research; Preparation and Examination of P ractical Catalysts, Academic Press, New York, 1976. 9. C. N. Satterfield, Heterogeneous Cataysis in Industrial Practice, 2nd ed., McGraw Hill, New York, 1991. 10. M. V. Twigg, ed., Catalyst Handbook, 2nd ed., Wolfe Pub. Co. Ltd., England, 1989. 11. S. P. S. Andrew, Chem. Engin. Sci., 36, 1431 (1981).

14.2.7.2. Methods of Production of Nonmetal Catalysts and Supports 14.2.7.2.1. Precipitation and Gel Formation.

The production of catalyst particles of suitable configuration and hardness is an essential part of catalyst manufacture'-*. Most heterogeneous catalysts are produced by processes that involve formation of solids from aqueous solutions3. Precipitation is frequently employed in preparation of hydrous oxide catalysts. To avoid occluded o r adsorbed impurities, ammonia or ammonium salts are often used as well as nitrates of the desired metal constituents. Calcination removes the nitrogen-containing components. Anions such as C1- or SO2- or cations such as Na+ are avoided; these often are poisons if they are present in the final catalyst. Despite improvements made in precipitation methods, precipitation rarely produces homogeneous products. The difference between overall composition of the initial solution and the composition of the precipitate at various stages indicates heterogeneity of the latter. The composition of the precipitate cannot be prevented from varying from the beginning to the end of the precipitation (Fig. 1). Gel formation is especially well suited to the preparation of hydrous oxide catalysts such as those of Si, Al, Fe. Also, an all-embracing gel can be used to produce the catalyst particle in a preferred size and spherical form by allowing droplets of the recently rapidly

solutions hydroxides 2.07

F e f samples ~

1.97

2.02

1.99

2.06 2.15

Figure 1. Ferrites prepared by coprecipitation of hydroxides. Atomic ratio Fe(M in coprecipitate and mixtures of starting solutions?

52

14.2.7. Production of Catalysts and Supports 14.2.7.2. Methods of Production on Nonmetal Catalysts and Supports 14.2.7.2.1. Precipitationand Gel Formation.

mixed ingredients to fall through an immiscible oil at such a rate that the gel sets during the fall through the oil. Silica-alumina catalysts are manufactured on a very large scale. One formulation5 utilizes a solution of A12(SO,), and N-brand water glass (SiO,/Na,O) = 3.2). The solutions are mixed rapidly; they gel in about 45 s. The gel is aged, then washed using a solution of Al,(SO,), to remove Na+ by base exchange, washed free of SO,'-, and finally dried and calcined at 550°C. The catalyst is amorphous. The preparation of mixed oxide catalysts, which is usually very complex, can be described as a succession of many elementary steps (Fig. 2) which are governed largely by empirical rules and where the evolution of composition, structure, and texture often occurs simultaneously. A systematic characterization of each intermediate is necessary to set up a continuous relationship between the first hydrated precursor and final activated oxide catalyst. Catalyst production requires careful control of gel aging, washing, drying, calcination, and activation, with special concern for control of humidity conditions. During gel aging, significant hydrothermal reactions occur. One special case of hydrothermal reaction leads to the formation of molecular sieve zeolite^'^^. Zeolite synthesis generally requires' (1) reactive starting materials such as freshly coprecipitated gels or amorphous solids, (2) relatively high pH introduced in the form of an alkali metal hydroxide or other strong base, (3) low-temperature hydrothermal conditions with concurrent autogenous pressure at saturated water vapor pressure, and (4) a high degree of supersaturation of the components of the gel leading to the nucleation of a large number of crystals. The gels are crystallized in a closed hydrothermal system at temperatures up to 175°C during periods of a few hours to several days. The zeolites crystallize easily due to the high reactivity of the gel, the concentration of the alkali hydroxide, and the high surface activity of the solid phases concerned. The numerous zeolites that have been synthesized have been reviewed8,' as a function of conditions, phase diagram and zeolite structure. Those of the Y (Faujasite)-type are of enormous values for catalytic cracking and other petroleum refining processes. Another type, designated as ZSM-5, is also of great scientific and industrial interest. It has been synthesized from sodium aluminosilicate gels using tetrapropylammonium (TPA) hydroxide, which acts as a template, from a mixture having the composition in moles/Al,O,: TPA, 8.6; Na,O, 10; SO,, 7.7; H,O, 453; using typical conditions of 150°C for 5 to 8 days. Alumina is in many ways the most versatile catalyst upp port'^. Although alumina can be produced by gelation, it is usually produced by thermal decomposition of one of the commercially available hydrates. These are precursors for high area y- and 11-alumina supports. The sequence of thermal decomposition products and the influence of conditions on product formation are shown in Fig. 311. Alumina trihydrate, A1,O3.3H,O, preferably called aluminum trihydroxide, Al(OH),, since it does not contain H,O as such, exists in three crystallographic forms, gibbsite, bayerite, and nordstrandite. These are composed of identical structural elements that are stacked in slightly different patterns. Gibbsite, also known as hydrargillite or a-alumina trihydrate, is precipitated as an intermediate in the Bayer process for the manufacture of aluminum metal from bauxite. When calcined at 400-500"C, y-alumina forms which has a surface area of about 300 m2/g and contains surface OH groups corresponding to a calculated 0.5% H20. The P-trihydrate, (mis)named bayerite, produces 11-alumina upon calcination and exhibits properties superior to y-alumina for certain catalyst applications. Nordstrandite is the third trihydrate. It is not used for the manufacture of commercial


53

Solution Containing Some

lb

Complexation

Coprecipitation

I (Aging) I

(acids-alcoholsaddn)

I

Drying

Washing

I I Drying

Gel Formation

I

Drying

I

I

(Asins)

I

Extrusion or Balls Agglomeration

Removal of Volatile Compounds

(Asins)

Makxing With Oxides Precursors

I

(Asins)

I

Thermal Activation

Thermal Activation

I \

decomposition

( of the complex)

I

Thermal Activation

Thermal Activation

(soliktate reaction)

Addition of Other Elements Impregnation malaxing balls agglomeration

I

Drying Thermal Activation Forming Process tabletting, extrusion, balls agglomeration

I

(Asins)

I I

Drying Thermal Activation

I

CATALYST "READY FOR U S E

Figure 2. Different methods of synthesis of bulk mixed oxide catalysts.6 catalysts. y-Alumina monohydrate, also called boehmite, which can be prepared as a gelatinous precipitate, also produces y-alumina upon calcination. Because of their high areas, favorable pore size distribution, hardness, thermal stability, and, if desired, their ability to act as acids, y- and galumina are employed in a wide variety of catalysts (Table 1, 14.2.7.1).


54

Conditions Favoring Transformations(i) Path a

Conditions

Path b

Pressure Atmosphere Heating Rate

> 1 atm. moist air

> 1"C/min.

2.3 x 104 -- -

PB is the partial pressure of the intermediate (atm), dN/dt is the overall reaction rate (mol/s cm3), T is the reaction temperature (K), D is the diffusivity of the medium [cm2/s],

and R is the catalyst particle size (cm). The intimacy requirements in terms of particle size and partial pressure of intermediate have been calculated5. This suggests the constraints that must be applied in forming a dual function catalyst. Guidance is provided for the possible mixing of two particles, for size reduction or the requirement that one catalyst component be added to the other in solution to obtain required intimacy. Shape selectivity. Crystalline zeolites form a special class of heterogeneous catalysts. They are oxides, generally combinations of silica and alumina, having open-framework

14.2. Ty es of Catal sts 14.2.7. froduct ion or Catalysts and Supports 14.2.7.4. Relationships between Catalyst Production and Performance

63

structures with apertures of molecular dimensions of similar size, 0.4- 1.4 nm, to that of many organic molecules. Also called molecular sieves, those of greatest interest have apertures made of rings of 8 to 12 oxygen ions (Fig. 4). The apertures lead to channels or pores, some of which are interconnecting as illustrated for the zeolite ZSM 5 (Fig. 5). Crystalline zeolites have provided a new method to control catalytic selectivity. Called shape selectivity, this control utilizes the restrictive consequences of the molecular size of the zeolite pores6. Certain molecules, because of their size and shape, are constrained in their ability to enter or leave zeolite pores. Reactant selectivity occurs when only some of the reactants can diffuse into catalyst pores, for example, linear but not

Y Zeolite

A Zeolite

Figure 4. Framework structures and projections of representative large and small pore zeolites. Elliptical 10-ring straight channel (5.7A X 5.1 A)

Near-circular 10-ring zig-zag channel (dia 5 . 4 4

Figure 5. Framework structure of zeolite ZSM 5.

64

14.2. Ty es of Catal sts 14.2.7. 8roduction or Catalysts and Supports 14.2.7.4. Relationships between Catalyst Production and Performance Reactant Selectivity 'V

Restricted Transition State Selectivity

4

bProduct Selectivity

L

Figure 6. Types of zeolite shape selectivity6. branched hydrocarbons (Fig. 6). Product selectivity occurs because some of the products formed within the pores are too bulky to diffuse out. For example, in alkylation of benzene with methanol, p-xylene is produced rather than more bulky 0-or m-xylene due to pore diameter restrictions (Fig. 6). Restricted transition state selectivity occurs when shape restrictions act on intrinsic kinetics rather than diffusional limitations.

(G.A. MILLS) 1 . P. Courty, C. Marcilly, in Preparation ofcatalysis, Vol. I, B. Delmon, P. A. Jacobs, C. Poncelet, eds., Elsevier, New York, 1976, p. 119. 2. R. K. Grasselli, in Heterogeneous Catalysts, B. H. Davis, W. P. Hettinger, eds., Symposium Series 222, American Chemical Society, Washington, D. C., 1983. 3. J. B. Utterhoeven, L. G. Christner, W. K. Hall, Catalysis Annual Review, C . Kemball and D. A. Dowden, eds., The Chemical Society, London, 1980, p. 208. 4. G. A. Mills, H. Heinemann, T. H. Milliken, A. G. Oblad, Ind. Eng. Chem. 45, 134 (1953). 5. P. B. Weisz, in Academic Press Adv. Catal. D. D. Eley, P. W. Selwood, P. B. Weisz, Eds. 13, 137 (1962). 6. S. M. Csissery, Zeolites, 4,202 (1984).


14.3 Hydrogenation Reactions 14.3.1. Introduction This section describes general dihydrogen activation, reactions which form metal hydrides and molecular dihydrogen complexes, and the metal complexes which constitute the main classes of soluble catalysts.

14.3.2. Dihydrogen Activation Catalytic reactions that employ H, are generally associated with hydrogenation of unsaturated moieties. Hydrogenations involving unsaturated organic substrates that contain C=C, C c C , C=O, C=N, C-N, and N = O will be considered in subsequent subsections within 14.3. H, is also used in catalytic systems that precipitate metals or metal oxides by reduction of dissolved metal complexes', and for catalytic reduction of other inorganic substrates including permanganate, dichromate, iron(II1) complexes, nitrite, sulfur, halogens, oxygen, and hydrogen peroxide,. Hydrogenation of CO to CH,, alcohols, and hydrocarbons via Fischer-Tropsch syntheses represents a particularly important example of reduction of an inorganic substrate3. Catalyzed hydrogenolysis reactions of C-OH and C-X (X = halide) bonds are sometimes competitive reactions in catalytic hydrogenations2v4.All these reactions, when catalyzed by transition metal complexes, require activation of H, at the metal center (H, addition to unsaturated groups has also been accomplished with metal-free, base-catalyzed systems5 and with boranecatalyzed systems6). Transition metal-catalyzed isotope exchange reactions (e.g., H,/D,O or D,/H,O) and ortho-para H, conversion are also examples of reactions that require H, activation2,'. Although H, addition reactions are frequently thermodynamically favorable, the high bond dissociation energy for the homolytic splitting of H, into H atoms (435 W mol-' in the gas-phase and ca. 420 W mol-' in water) makes uncatalyzed H, addition reactions kinetically unfavorable. The bond dissociation energy of H, is partially recouped in the form of M-H bonds when H, is oxidatively added to a metal center. Accordingly, the function of the metal catalyst is to provide energetically accessible pathways for the addition of H, to substrates. The energy required for the heterolytic splitting of H, (ca. 155 W mol-' in water) is substantially less than the energy required to effect the corresponding homolytic H-H bond cleavage. This suggests H, activation via heterolytic mechanisms (particularly in polar solvents) is plausible. Although most addition reactions proceed via metal-hydride intermediates, the discovery of metal-dihydrogen complexes raises the question of their possible involvement in hydrogenation reactions. The formation of metal-hydrides and metal-dihydrogen complexes from H, is considered below. (6. R. JAMES, M. T. ASHBY)

65

66

14.3. Hydrogenation Reactions 14.3.2. Dihydro en Activation 14.3.2.1. Hornofjtic Cleavage to Give Metal-Hydrides

1. J. Halpern, J . Organornet. Chem., 200, 133 (1980). 2. B. R. James, Homogeneous Hydrogenation, Wiley, New York, 1973. 3. J. Falbe, ed., New Syntheses with Carbon Monoxide, Springer-Verlag, Berlin, 1980; R. A. Sheldon, Chemicals from Synthesis Gas, Reidel, Dordnecht, 1983. 4. M. Bennett, Chemtech., 10,444 (1980). 5. C. Walling, L. Bollyky, J . Am. Chem. SOC., 86, 3750 (1964). 6. Ref. 2, Ch. XVI; H. C. Brown, M. Srebnik, R. K. Bakshi, T. E. Cole, J . Am. Chem. Soc., 109,

5420 (1987). 7. T. C. Eisenschmid, R. U. Kirss, P. P. Deutsch, S. I. Hommeltoft, R. Eisenberg, J. Bargon, R. G. Lawler, A. L. Balch, J . Am. Chem. SOC., 109, 8089 (1987). 14.3.2.1. Homolytic Cleavage to Give Metal-Hydrides

The term hornolytic splitting of H, by metal complexes, in contrast to heterolytic cleavage (H, + HH + ) to be discussed in the following section, is generally used synonymously with oxidative addition. Equation (a) illustrates an oxidative addition reaction of H, to a univalent metal center:

+

M'

+ H,

6 M"'(H),

(a)

Such reactions are written as a concerted H, addition via a three-center transition state (with q2-H, (dihydrogen) species as intermediates (Section 14.3.2.3)).Hydrogen oxidative additions are particularly common for d8 metals in square planar coordination environments (particularly Rh' and Ir') because this results in a favorable pseudooctahedral d6 configuration. A classic case is shown in equation (b), the reversible addition of H, to trans-IrCl(CO)(PPh,), (Vaska's complex), to give the Ir'" dihydride with cis hydride and trans phosphine ligands. Oxidative addtion of H, and the reverse reaction, reductive elimination, are specific examples within a broad class of reactions that are critical in hydrogenation and homogeneous catalytic reactions'. trans-IrCl(CO)(PPh,),

+ H,

6 Ir(H),Cl(CO)(PPh,),

(b)

Table 1 gives examples of catalytically active dihydrides that are formed by addition of H, to a complex with or without loss of an ancillary ligand (the dissociated ligand is shown in italics); the transients listed seem likely precursors in dihydride formation, e.g., in the synthesis of Cp,MoH, from Cp,MoCl, equation (c). Cp,MoCl,

zn'Hg>

"Cp,Mo"-%

Cp,Mo(H),

(c)

If the hydride ligand is viewed as monoanionic (but see below), the systems of Table 1 involve predominantly the ds + d6 process, although there are examples of the also favored d" + d8, as well as less favored d9 + d7, d6 + d4, d4 + d2, and dZ +-do processes. The reaction shown in equation (d)', with cleavage of an Mo-Mo bond, constitutes a formal d3 + d2 process: [Cp,Mo],

+ 2 H,

-

2 Cp,MoH,

(4 Metal centers that contain an odd number of d electrons, especially d7 and d9 systems, attain the more favored d6 and d8 configurations, respectively, by a net addition of H, at two such centers generally to give monomeric monohydrides. The general equation, shown in equation (e) (n = oxidation state of precursor), is exemplified by the classic d7 case of pentacyanocobaltate(II), equation (f)3, and the d9 system of octacarbonyldicobalt(O), equation (g)4.


66


1. J. Halpern, J . Organornet. Chem., 200, 133 (1980). 2. B. R. James, Homogeneous Hydrogenation, Wiley, New York, 1973. 3. J. Falbe, ed., New Syntheses with Carbon Monoxide, Springer-Verlag, Berlin, 1980; R. A. Sheldon, Chemicals from Synthesis Gas, Reidel, Dordnecht, 1983. 4. M. Bennett, Chemtech., 10,444 (1980). 5. C. Walling, L. Bollyky, J . Am. Chem. SOC., 86, 3750 (1964). 6. Ref. 2, Ch. XVI; H. C. Brown, M. Srebnik, R. K. Bakshi, T. E. Cole, J . Am. Chem. Soc., 109,

5420 (1987). 7. T. C. Eisenschmid, R. U. Kirss, P. P. Deutsch, S. I. Hommeltoft, R. Eisenberg, J. Bargon, R. G. Lawler, A. L. Balch, J . Am. Chem. SOC., 109, 8089 (1987). 14.3.2.1. Homolytic Cleavage to Give Metal-Hydrides

The term hornolytic splitting of H, by metal complexes, in contrast to heterolytic cleavage (H, + HH + ) to be discussed in the following section, is generally used synonymously with oxidative addition. Equation (a) illustrates an oxidative addition reaction of H, to a univalent metal center:

+

M'

+ H,

6 M"'(H),

(a)

Such reactions are written as a concerted H, addition via a three-center transition state (with q2-H, (dihydrogen) species as intermediates (Section 14.3.2.3)).Hydrogen oxidative additions are particularly common for d8 metals in square planar coordination environments (particularly Rh' and Ir') because this results in a favorable pseudooctahedral d6 configuration. A classic case is shown in equation (b), the reversible addition of H, to trans-IrCl(CO)(PPh,), (Vaska's complex), to give the Ir'" dihydride with cis hydride and trans phosphine ligands. Oxidative addtion of H, and the reverse reaction, reductive elimination, are specific examples within a broad class of reactions that are critical in hydrogenation and homogeneous catalytic reactions'. trans-IrCl(CO)(PPh,),

+ H,

6 Ir(H),Cl(CO)(PPh,),

(b)

Table 1 gives examples of catalytically active dihydrides that are formed by addition of H, to a complex with or without loss of an ancillary ligand (the dissociated ligand is shown in italics); the transients listed seem likely precursors in dihydride formation, e.g., in the synthesis of Cp,MoH, from Cp,MoCl, equation (c). Cp,MoCl,

zn'Hg>

"Cp,Mo"-%

Cp,Mo(H),

(c)

If the hydride ligand is viewed as monoanionic (but see below), the systems of Table 1 involve predominantly the ds + d6 process, although there are examples of the also favored d" + d8, as well as less favored d9 + d7, d6 + d4, d4 + d2, and dZ +-do processes. The reaction shown in equation (d)', with cleavage of an Mo-Mo bond, constitutes a formal d3 + d2 process: [Cp,Mo],

+ 2 H,

-

2 Cp,MoH,

(4 Metal centers that contain an odd number of d electrons, especially d7 and d9 systems, attain the more favored d6 and d8 configurations, respectively, by a net addition of H, at two such centers generally to give monomeric monohydrides. The general equation, shown in equation (e) (n = oxidation state of precursor), is exemplified by the classic d7 case of pentacyanocobaltate(II), equation (f)3, and the d9 system of octacarbonyldicobalt(O), equation (g)4.


TABLE1. OXIDATIVE ADDITION OF H, AT ACTIVEDIHYDRIDES: M H, MH,"

+

A

67

SINGLEMETALCENTER TO GIVECATALYTICALLY

Dihydride precursor (ref.) ~

_

_

_

_

~

_

~

~

~

~

~

~

Simple addition [Ti(q-C,Me,),] (35); [FeH,L,] (36,37); [CoHL,] (38,39); [RhXL,lb, [ ( RhXL, I,] (41,42); [R~CI(PBU",)~] (43); [RhCl(PPh,),(CH,=CHCN)] (44); [Rh(PhNC(S)NMe, j(PPh,),] (45); [Rh(dppp),] +,[Rh(dppb),] +,[Rh(diop)J (46); [RhXL,(diene)Y (47); [ML,S,]+d, M = Rh, Ir (48-50); [M(phen)S,]+", M = Rh, Ir (54-56); [MCl(PCy,),], M = Rh, Ir (57); [(MCl),p-@( (Ph,PCH,CH,),P)C,H,, M = Rh, Ir (58); trans-[IrX(CO)LJf (50,59); [IrH,L,], [Ir(diphos),] (4930); [Ir(PPh,)(py)S,] + , [Ir(cod),] , [Ir(Ph,PMe),(cod)] (61); [IrC1(PPh3),(cod)lg (62); [Ir2(~-S)(Co),(d~pm),j (63); [WPEt,),l (64); [Pt (PBu*,(CH,),PBu',)],h (65)pet +

+

+

+

With loss of one ligand (indicated in bold) [Cr( €('OR), t 61 (66); [ { Mo(vC,Me,), t ,I1 (67,681; [FeHz(N2)(PEtPh,),1 (36,691; [Ru(CO),(PPh,),I (70); [RuH,(Nz)(PPh3)31,[RuH2(PPh3)41(5OJ 1 ); [CoH(Nz)(PPh3)31, [Co(NdPPh,),l (38,501; [IrCI(CO)(PMe2Ph)31(72); [IrH(CO),(PPf,)I (73); [IrWCO)(PPh,),l (49,50,74) Transients [Zr(gCsMes),l (75); [NbHCp,IJ (76); "M-ICp,l (77); [MCp,l, M = Mo, W (67,681; [Fe(CO),l? (78,79); [RuL,] (50,71,80); [Ru(diphos),l (8 1 ); [Os,(CO),ol (82); [CoL,(q'-C,H,)] (83); [Pd(PPhd,l (84); [Pt(diphos)I (65) 'Unless stated otherwise, X = anion (usually halide); L = monodentate tertiary phosphine, arsine, or phosphite; S = solvent. Mixed phosphine ligand systems have also been used.@ 'Product is [Rh(H), X L,], L = P(p-ClC,H,),. dFonned from dime precursors such as [&(cod)]+; with M = Rh, Lz also = (RCN)(PR,)5'-52, ( ~ y ) ( p R , ) ~ ' . S~ ~=; solvent or L. Dihydrides well substantiated only with bis(phosphine) systems. eFonned from diene precursors such as [ML,(cod)]+; with M = Rh, L, also = bipy, (amine), (nitrile),5'*52; S = solvent. Dihydrides not well substantiated. 'X here includes halides, NCO, NCS, N,, NO,, SnCl,, GeR, u-carboranes.m gProduct is [(cod)Ir~-Cl),IrH,(PPh,),l. hProduct is cis-[Pt(H),{ PBu',(CH,),PBu',)]. 'Product is [Mo(H),(T&M~,),]; process visualized as involving dissociation of a [Mo( q-C,Me,),] ligand. jForms not only the trihydride by addition of H, but also loses H, reversibly to give [ ( NbHCp(C,H,)),], which probably contains bridging $:$-fulvalene.

2 M" (or Mn2)

2 [Co"(CN),l3- (or [CO~~,(CN),,]~-) + H, Co0,(CO),

+ H,

2 M(" ')H

(el

2 [Co"'H(CN),I3-

(f)

+

F=2

Co'H(CO),

(€9

Other systems that form monohydrides according to equation (e) include d7 cobalt(I1) and rhodium(I1) amine and dimethylglyoximato complexes, d9 cobalt(O), rhodium(O), and iridium(0) carbonyl phosphine dimers, Mn,(CO),,, and [CpCr(CO),];. A rare example in which oxidative addition of H, generates a dimeric product with one hydrogen bound to each metal is seen in equation (h), in which the formally Ir(I1) d7 product maintains an 18-electron configuration by formation of a metal-metal bond6. Analogous systems involving bridged d8 metal complexes, for example, Rh,(pC1),(PPh3), and A-

68

14.3. Hydrogenation Reactions 14.3.2. Dihydro en Activation 14.3.2.1. Hotnottic Cleavage to Give Metal-Hydrides

-

frame complexes such as Irz(p-S)(p-dppm),(CO),, usually add H2 at just one metal center to yield mixed valence products, (M'~M'''H,)7,8. [Ir(p-SBu')(CO)(PPh,)],

+ H,

[IrH(p-SBu')(CO)(PPh,)],

(h)

The mechanisms of reactions such as (e) are difficult to elucidate. Reaction via a direct termolecular step, H, addition to undetectable amounts of dimer, or the process outlined in equation (i) have all been considered for cobalt(I1) systems. M

+ H,

M(H),

M

2MH

(i)

There are many catalytic systems active for hydrogenation of organic substrates where an initial oxidative addition of H, at one or two metal centers has been invoked, although the intermediate di- and monohydride is not detected. Dihydride formation has been postulated, for example, with the suggested Cr(CO)?, Fe(CO),", and chlororuthenate(1)' intermediates, and the Rh,(OAc), complex". Reactions of H, with carbonyl clusters provide many examples of H, addition across several metal centers',. For example, FeRu,(H),( CO) yields FeRu,(H),( CO) Re,(CO) yields Re,(H),( CO) 12 and Re,(H),(CO),,; and OS,(CO),~eventually gives 0s4H4(CO),,. Hydroformylation reactions catalyzed by the Rb(CO),, cluster may proceed via an undetected RhH(CO), intermediate14.There are also a large number of mono-, di-, and polynuclear complexes that form catalytically inactive hydrides (including di- and p~lyhydrides)'~. Dihydride complexes synthesized directly using H, generally have cis dihydride geometry, although a trans concerted addition of H, is allowed in principle". Some data exist for trans H, addition to trans-IrH(CO)(PPh,),, but the evidence was clouded by the possibility of an intramolecular rearrangement process16. Also cis-trans isomerization within M(H),[P(OR),], complexes (M Fe, Ru) has been do~umented'~. Some Pt(PR,), complexes, where R is a bulky substituent, give trans dihydrides18,and transMo(H),(diphos), is formed from tr~ns-Mo(N~)~(diphos),~~. The forward steps of H2 oxidative addition, equations (a) and (e), are usually considered to be promoted by coordinatively unsaturated metal centers in initially low oxidation states (i.e., high metal basicity). Loss of electrons via the oxidation is compensated for by a gain of electrons through an increase in the coordination numbef'. This also explains why the earlier transition metal systems (d'-d4) tend to form complexes of higher coordination number than those of the later d7-d'' systems (especially Group 8-10 metals), and why oxidative addition reactions generally are not favored for steric reasons with the earlier metals, although this constraint should not be too severe for the small hydrogen ligands. Studies of various Vaska-type complex reactivity, transIrX(CO)L, (Table l), toward H, reveal increased reactivity (kinetic and thermodynamic) with increasing metal basicity and decreasing ligand sizez'.22.The nonreactivity of transRhCl(CO)(PPh,), toward H,, in contrast to the iridium analog and RhC1(PPh3), under corresponding conditions, has been attributed similarly to the lower basicity of Rh compared to Ir and the introduction of the r-acid CO ligand, respectivelyz3. Introduction of the stronger wacid CS ligand renders trans-IrCl(CS)(PPh,), unreactive, whereas [1r(CO),(CS)(PCy3),] forms the dihydride [Ir(H),(CO)(CS)(PCy,),] and the less basic PPh, analogue does notz4. Within square-planar d8 I f complexes, introduction of wacid ligands such as CS, PF,, olefins, and alkynes is thought to decrease reactivity toward H, by increasing the "electronic promotion energy" for the formal Ir' to If" oxidation state change". The strong .rr-acceptor ability of the NO ligand is probably

,

+

+

+


69

detrimental for catalytic hydrogenation activity of nitrosyl complexes, despite the ligand's ability to create a vacant coordination site by conversion from a three- to a oneelectron dono?. Studies on equilibria (e) and (i) as a function of T allow determination of the thermodynamic parameters, particularly dH",and hence estimation of the metal-hydride bond dissociation er~ergy'~*'~. For complexes, MnH(CO),, COH(CN),~- , CoH(CO),, CoH(dmgh),PBu",, and Co(H),(PPh,),, the dissociation energies are in the 220-260 kJ mol- range, similar to that for H2 absorbed on iridium metal". All these systems are active for catalytic hydrogenation of organic substrates. The synonymity of H, homolytic splitting with oxidative addition is clear upon examination of equations (a) and (e), and the examples discussed above. Some hydrides, e.g., MnH(CO),, CoH(CO),, and [CoH(CN),I3 - act as hydrogenation catalysts via free radical pathways involving H atom transfer (see 14.3.3.2, 14.3.3.3). The hydrogen is best considered as a stabilized atom rather than hydride, e.g., Co"(*H) rather than Co"'(:H). The manganese hydride itself is considered to be formed from Mn,(CO),, via the radical process outlined in equations (i) and (k) and involves oxidative addition of H, to Mn(CO),'*, cf. equation (i). The formulations of these hydrides differ only in the position of the electron originally associated with the H atom, i.e., the polarity of the metalhydrogen bond; coordinated hydrogen is always formally assigned the classical - 1 oxidation state. Addition of H, to [Ir(cod),] + (Table 1) has been described as reductive rather than oxidative in character, electron density on the metal increases and the hydrogens becoming more acidicz9.From their chemistry, hydrides of the later transition metals tend to be covalent or acidic while those of the earlier transition metals are more hydridi~~~,~'. huord -co H Mnz(CO)lo 2 Mn(CO), Mn(CO), a Mn(H),(CO), (j)

'

TABLE2. VALUES OF pK, FOR REPRESENTATIVE METAL-HYDRIDE COMPLEXES IN WATER' 5'32

Complex

PKa

-0.0 5 -5 7 -7 -9

-10

10.5 11.7 -12 12.8 -18

"See E. J. Moore, J. M. Sullivan, J. R. Norton, J . Am. Chem. Soc., 108, 2257 (1986) for pK, values in acetonitrile.

70

14.3. Hydrogenation Reactions 14.3.2. Dihydrogen Activation 14.3.2.1. Hornolytic Cleavage to Give Metal-Hydrides

Some metal hydride pKa values have been determined (Table 2p3'; the values are markedly ligand dependent, as seen for example, by replacement of one CO of HCo(CO),, a strong acid, by PPh3, which decreases the acidity of 7 pK units! However, the pKa values, which must be a measure of the polarity of the M-H bond under a certain set of conditions (those of the titration procedures), have proved of little value in predicting the chemistry of the metal hydrides, e.g., whether behavior is more typical of Co"'(:H), Co"(-H), or Co'(H). This is critical in catalysis, particularly in the direction of addition of the M-H bond across olefinic groups (i-e., in olefin insertion), which is important in hydrogenation, hydroformylation, and i ~ o m e r i z a t i o n ~ 'This ~ ~ ' ~is~ a. complex question and, as well as electronic factors, stenc factors, solvent polarity, the presence of radical initiators, and even temperature changes, can be important. (B. R. JAMES, M. T. ASHBY) 1. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, University Science Books, Mill Valley, CA, 1987. 2. J. L. Thomas,J. Am. Chem. SOC.,95, 1838 (1975). 3. B. R. James, Homogeneous Hydrogenation, Ch. X, Section A, Wiley, New York, 1973. 4. F. Ungvary, J. Organomet. Chem., 36, 363 (1972). 5. B. R. James, in Comprehensive Organometallic Chemistry,Vol. 8, Ch. 51, G. Wilkinson, F. G. A. Stone, E. Abel, eds., Pergamon Press, Oxford, 1982. 6. J. J. Bonnet, A. Thorez, A. Maisonnat, J. Galy, R. Poilblanc, J. Am. Chem. SOC., 101, 5940 (1979). 7. B. R. James, Adv. Organomet. Chem., 17, 319 (1979). 8. C. Kubiak, C. Woodcock, R. Eisenberg, Inorg. Chem., 19, 2733 (1980); B. C. Y. Hui, W. K. Teo, G . L. Rempel, Inorg. Chem., 12,757 (1973). 9. Ref. 3, Ch. VII, Section A. 10. Ref. 3, Ch. XIV, Section B. 11. B. C. Hui, B. R. James, Can. J. Chem., 52,3760 (1974). 12. B. C. Hui, G. L. Rempel, Chem. Commun., 1195 (1970). 13. H. D. Kaesz, R. B. Saillant, Chem. Rev., 72, 231 (1972); B. F. G. Johnson, ed., Transition Metal Clusters, Wiley, New York, 1980. 14. Ref. 3, Ch. XI, Section F. 15. R. G. Pearson, Theor. Chim. Acra, 16, 107 (1970). 16. J. F. Harrod, G. Hamer, W. Yorke, J. Am. Chem. SOC., 101, 3987 (1979). 17. J. P. Jesson, in Transition Metal Hydrides, E. L. Muetterties, ed., Dekker, New York, 1971, p. 75. 18. T. Yoshida, T. Yamagata, T. H. Tulip, J. A. Ibers, S. Otsuka, J. Am. Chem. SOC., 100, 2063 (1978). 19. M. Hidai, K. Tomirari, Y. Uchida, J. Am. Chem. SOC., 94, 110 (1972). 20. J. Halpern, Acct. Chem. Res., 3, 386 (1970). 21. Ref. 3, Ch. XII, Section A. 22. N. E. Burke, A. Singhal, M. J. Hintz, J. A. Ley, H. Hui, L. R. Smith, D. M. Blake, J. Am. Chem. SOC., 101,74 (1979). 23. Ref. 3, Ch. XI, Section E. 24. M. J. Mays, F. P. Stefanini, J. Chem. SOC. (A),2747 (1971). 25. J. Halpern, Adv. Chem. Ser., 191, 165 (1980). 26. J. L. Hendrikse, J. H. Kaspersma, J. W. E. Coenen, Int. J. Chem. Kinet., 7 , 557 (1975). 27. G . C. Bond, Catalysis by Metals, Academic Press, London, 1962. 28. D. R. Kidd, T. L. Brown, J. Am. Chem. SOC., 100,4095 (1978). 29. R. H. Crabtree, G. G. Hlatky, Inorg. Chem., 19, 571 (1980). 30. C. Masters, Adv. Organomet. Chem., 17, 61 (1979). 31. J. A. Labinger, Adv. Chem. Ser., 167, 149 (1978). 32. D. F. Shriver, Acct. Chem. Res., 3, 231 (1970). 33. Ref. 3, Ch. X, Section F. 34. R. A. Sheldon, Chemicals from Synthesis Gas, Reidel, Dordrecht, 1983, Ch. 2,4. 35. J. E. Bercaw, R. H. Marvich, L. G. Bell, H. H. Brintzinger, J. Am. Chem. SOC.,94, 1219 (1972). 36. V. D. Bianco, S. Doronzo, M. Aresta, J. Organomet. Chem., 42, C63 (1972).

14.3. Hydrogenation Reactions 14.3.2. Dihydrogen Activation 14.3.2.2. Heterolytic Cleavage to Give Metal-Hydrides

71

37. W. E. Newton, J. L. Corbin, P. W. Schneider, W. A. Bulen, J . Am. Chem. SOC., 93, 268 (1971). 38. Ref. 3, Ch. X, Sect. G. 39. M. C. Rakowski, E. L. Muetterties, J . Am. Chem. SOC., 99,739 (1977). 40. R. L. Augustine, R. J. Pellet, J . Chem. SOC.,Dalton Trans., 832 (1979). 41. Ref. 3, Ch. XI, Sect. B. 42. B. R. James, Adv. Organomet. Chem., 17, 1979, Sect. IIA. 43. T. Yoshida, S. Otsuka, M. Matsumoto, K. Nakatsu, Inorg. Chim. Actu, 29, L257 (1978). 44. Y. Ohtani, A. Yamagishi, M. Fujimoto, Bull. Chem. SOC.Jpn., 51, 2562 (1978). 45. A. W. Gal, F. H. A. Bolder, J. Organomet. Chem., 142, 375 (1977). 46. B. R. James, D. Mahajan, Can. J. Chem., 57, 180 (1979). 47. L. A. Oro, J. V. Heras, Inorg. Chim. Acta, 32, L37 (1979). 48. Ref. 3, Ch. XI, Sect. G. 49. Ref. 3, Ch. XII, Sect. B. 50. Ref. 41, Sects. IIB, IE. 5 1. R. Uson, L. A. Oro, J. Artigas, R. Sariego, J. Organomet. Chem., 179, 65 (1979). 52. R. Uson, L. A. Oro, M. C. Carmen, P. Lahuerta, Transition Met. Chem., 4, 55 (1979). 53. R. H. Crabtree, H. Felkin, J. Mol. Cutal., 5 , 75 (1979). 54. Ref. 41, Sect. XI. 55. G. Mestroni, G. Zassinovich, A. Camus, J . Organomet. Chem., 140, 63 (1977). 56. A. Camus, G. Mestroni, G. Zassinovich, J . Mol. Catal., 6 , 231 (1979). 57. S. Hietkamp, D. J. Stufkens, K. Vrieze, J . Organomet. Chem., 152, 347 (1978). 58. M. M. T. Khan, S. S. Ahmed, Rafeequnnisa, in Proceedings of the 16th International Conference on Coordination Chemistry, Abstract R39, Dublin, 1974; Chem. Abstr., 85, 40, 280 (1976). 59. Ref. 3, Ch. XII, Sect. A. 60. B. Longato, F. Morandini, S. Bresadola, Inorg. Chem., 15, 650 (1976). 61. R. Crabtree, Acct. Chem. Res., 12, 331 (1979). 62. M. Gargano, P. Giannoccara, M. Rossi, J . Organomet. Chem., 129, 239 (1977). 63. C. Kubiak, C. Woodcock, R. Eisenbexg, Inorg. Chem., 19,2733 (1980). 64. D. H. Gerlach, A. R. Kane, G. W. Parshall, J. P. Jesson, E. L. Muetterties, J. Am. Chem. Soc., 93, 3543 (1971). 65. T. Yoshida, T. Yamagata, T. H. Tulip, J. A. Ibers, S . Otsuka, J. Am. Chem. SOC., 100, 2063 (1978). 66. S. D. Ittle, C. A. Tolman, U.S. Pat. 4 155 925 (1979); Chem. Abstr., 91,59, 608 (1979). 67. J. L. Thomas, H. H. Brintzinger, J . Am. Chem. SOC.,94, 1386 (1972). 68. J. L. Thomas, J . Am. Chem. SOC.,95, 1838 (1973). 69. E. Koerner von Gustorf, I. Fischler, J. Leitich, H. Dreeskamp, Angew. Chem., Int. Ed. Engl., 11, 1088 (1972). 70. F. Porta, S. Cenini, S. Giordano, M. Pizzotti, J . Organomet. Chem., 150, 261 (1978). 71. Ref. 3, Ch. IX, Sect. B. 72. J. Y. Chen, J. Halpem, J . Am. Chem. SOC., 93,4939 (1971). 73. R. Whyman, J. Organomet. Chem., 94,303 (1975). 74. M. G. Burnett, C. J. Strugnell, J . Chem. Res. (S), 250 (1977). 75. J. E. Bercaw, Adv. Chem. Ser., 167, 136 (1978). 76. F. N. Tebbe, G. W. Parshall, J . Am. Chem. SOC.,93, 3793 (1971). 77. E. K. Barefield, G. W. Parshall, F. N. Tebbe, J . Am. Chem. SOC.,92, 5234 (1970). 78. Ref. 3, Ch. XIV, Sect. B. 79. M. A. Schroeder, M. S. Wrighton, J . Am. Chem. SOC., 98,551 (1976). 80. S. Komiya, A. Yamamoto, J . Mol. Catal., 5, 279 (1979). 81. P. Pertici, G . Vitulli, W. Porzio, M. Zocchi, Inorg. Chim. Acru, 37, L521 (1979). 82. J. B. Keister, J. R. Shapley, J . Am. Chem. SOC.,98, 1056 (1976). 83. Ref. 41, Sect. VII. 84. A. S. Berenblyum, L. I. Lakhman, I. I. Moiseev, E. D. Radchenko, Koord. Khim., 2,841 (1976).

14.3.2.2. Heterolytlc Cleavage to Give Metal-Hydrides The subject of heterolytic activation of H, by transition metal complexes has been critically reviewed'. Equation (a) illustrates the heterolytic splitting of H, by a divalent metal complex. Ligand X is typically anionic, and the reaction involves a net substitution



71

37. W. E. Newton, J. L. Corbin, P. W. Schneider, W. A. Bulen, J . Am. Chem. SOC., 93, 268 (1971). 38. Ref. 3, Ch. X, Sect. G. 39. M. C. Rakowski, E. L. Muetterties, J . Am. Chem. SOC., 99,739 (1977). 40. R. L. Augustine, R. J. Pellet, J . Chem. SOC.,Dalton Trans., 832 (1979). 41. Ref. 3, Ch. XI, Sect. B. 42. B. R. James, Adv. Organomet. Chem., 17, 1979, Sect. IIA. 43. T. Yoshida, S. Otsuka, M. Matsumoto, K. Nakatsu, Inorg. Chim. Actu, 29, L257 (1978). 44. Y. Ohtani, A. Yamagishi, M. Fujimoto, Bull. Chem. SOC.Jpn., 51, 2562 (1978). 45. A. W. Gal, F. H. A. Bolder, J. Organomet. Chem., 142, 375 (1977). 46. B. R. James, D. Mahajan, Can. J. Chem., 57, 180 (1979). 47. L. A. Oro, J. V. Heras, Inorg. Chim. Acta, 32, L37 (1979). 48. Ref. 3, Ch. XI, Sect. G. 49. Ref. 3, Ch. XII, Sect. B. 50. Ref. 41, Sects. IIB, IE. 5 1. R. Uson, L. A. Oro, J. Artigas, R. Sariego, J. Organomet. Chem., 179, 65 (1979). 52. R. Uson, L. A. Oro, M. C. Carmen, P. Lahuerta, Transition Met. Chem., 4, 55 (1979). 53. R. H. Crabtree, H. Felkin, J. Mol. Cutal., 5 , 75 (1979). 54. Ref. 41, Sect. XI. 55. G. Mestroni, G. Zassinovich, A. Camus, J . Organomet. Chem., 140, 63 (1977). 56. A. Camus, G. Mestroni, G. Zassinovich, J . Mol. Catal., 6 , 231 (1979). 57. S. Hietkamp, D. J. Stufkens, K. Vrieze, J . Organomet. Chem., 152, 347 (1978). 58. M. M. T. Khan, S. S. Ahmed, Rafeequnnisa, in Proceedings of the 16th International Conference on Coordination Chemistry, Abstract R39, Dublin, 1974; Chem. Abstr., 85, 40, 280 (1976). 59. Ref. 3, Ch. XII, Sect. A. 60. B. Longato, F. Morandini, S. Bresadola, Inorg. Chem., 15, 650 (1976). 61. R. Crabtree, Acct. Chem. Res., 12, 331 (1979). 62. M. Gargano, P. Giannoccara, M. Rossi, J . Organomet. Chem., 129, 239 (1977). 63. C. Kubiak, C. Woodcock, R. Eisenbexg, Inorg. Chem., 19,2733 (1980). 64. D. H. Gerlach, A. R. Kane, G. W. Parshall, J. P. Jesson, E. L. Muetterties, J. Am. Chem. Soc., 93, 3543 (1971). 65. T. Yoshida, T. Yamagata, T. H. Tulip, J. A. Ibers, S . Otsuka, J. Am. Chem. SOC., 100, 2063 (1978). 66. S. D. Ittle, C. A. Tolman, U.S. Pat. 4 155 925 (1979); Chem. Abstr., 91,59, 608 (1979). 67. J. L. Thomas, H. H. Brintzinger, J . Am. Chem. SOC.,94, 1386 (1972). 68. J. L. Thomas, J . Am. Chem. SOC.,95, 1838 (1973). 69. E. Koerner von Gustorf, I. Fischler, J. Leitich, H. Dreeskamp, Angew. Chem., Int. Ed. Engl., 11, 1088 (1972). 70. F. Porta, S. Cenini, S. Giordano, M. Pizzotti, J . Organomet. Chem., 150, 261 (1978). 71. Ref. 3, Ch. IX, Sect. B. 72. J. Y. Chen, J. Halpem, J . Am. Chem. SOC., 93,4939 (1971). 73. R. Whyman, J. Organomet. Chem., 94,303 (1975). 74. M. G. Burnett, C. J. Strugnell, J . Chem. Res. (S), 250 (1977). 75. J. E. Bercaw, Adv. Chem. Ser., 167, 136 (1978). 76. F. N. Tebbe, G. W. Parshall, J . Am. Chem. SOC.,93, 3793 (1971). 77. E. K. Barefield, G. W. Parshall, F. N. Tebbe, J . Am. Chem. SOC.,92, 5234 (1970). 78. Ref. 3, Ch. XIV, Sect. B. 79. M. A. Schroeder, M. S. Wrighton, J . Am. Chem. SOC., 98,551 (1976). 80. S. Komiya, A. Yamamoto, J . Mol. Catal., 5, 279 (1979). 81. P. Pertici, G . Vitulli, W. Porzio, M. Zocchi, Inorg. Chim. Acru, 37, L521 (1979). 82. J. B. Keister, J. R. Shapley, J . Am. Chem. SOC.,98, 1056 (1976). 83. Ref. 41, Sect. VII. 84. A. S. Berenblyum, L. I. Lakhman, I. I. Moiseev, E. D. Radchenko, Koord. Khim., 2,841 (1976).

14.3.2.2. Heterolytlc Cleavage to Give Metal-Hydrides The subject of heterolytic activation of H, by transition metal complexes has been critically reviewed'. Equation (a) illustrates the heterolytic splitting of H, by a divalent metal complex. Ligand X is typically anionic, and the reaction involves a net substitution

72


of X by hydride. This does not result in a change of the formal oxidation state of the metal, which is the major contrast to activation of H, via homolytic cleavage. Such reactions are often promoted by the presence of base that stabilizes the released proton. External added bases such as NEt, have been used, but the solvent media and/or the ligand X itself can play the same role. M"X

+ H Z F M I I H+ H+ + X-

(a)

Heterolytic cleavage of H, has been cited in various systems. However, often a heterolytic mechanism has been inferred from insufficient data. Since there is no easily accessible higher oxidation states, transition metals in relatively high oxidation states, and especially those with do electron configurations, are not expected to activate H, via oxidative addition. However, metal-hydride bonds appear to be more covalent than most coordination bonds (as evidenced by polyhydride species such as [ReV11H,]2- that exhibit unusually high formal oxidation states); therefore, arguments against homolytic addition of H, based entirely on formal oxidation state arguments may be fallacious. Distinction between the mechanistic pathways of H, activation generally rests on kinetic measurements and labeling studies. Table 1 lists some examples of isolated, catalytically active hydrides that are formed via a net heterolytic cleavage of H,. The term heterolytic cleavage originates from early kinetic work on hydrogenation of inorganic substrates using copper and silver systems2v3.For example, the mechanism of H, reduction of Cu" carboxylates in heptanoic acid solution has been investigated, one of the first reactions for which heterolytic cleavage of H, was proposed, and the rate law in equation (b) was suggested. -d[H,]/dt

= ~,[H,][CU"]

+ k2[H2][Cu']

(b)

The Cu' product is a more effective hydrogenation catalyst than the Cu" species and the rate data exhibits autocatalytic behavior. The initial reaction rate was attributed to activation of H, by Cu". The first-order dependence on Cu' fell to nearly half-order with increasing concentration of the Cur salt. This was attributed to dimerization of the Cu' TABLE1. MONOHYDRIDE FORMATION VIA A NET HETEROLYTIC CLEAVAGE OF H,: MX d + H+ + X - ' -MH

+ H,

Monohyhydride precursor, MX (ref.) [RuXzL31,[RuX,L2S21, [RuX3L,lb (27-29); [RuHCI(PPh,),] (30); [RuCl,(PPh,)(q-C6Me6)], [{RuCl2(17-C6Me6)),Ic (31); [RUC12(DMSO)(q-C6H6)](27,32,33); cis-[RuCl,(DMSO),] (29); [O~Cl,(PPh3)31(34); (CoX(P(OEt)3 ),,Id (35); tr~ns-[RhCl(CO)(PPh,),] (36); [RhHC1,(PBut2Me),l (37); [RhCl(dmgh),(PPh,)l (38); [Rh(HS04)(C2B,H, ,)(PPh3),1 (39); [{MX,(q-C,Me,) ),I: M = Rh, Ir (40-42); [PdCI(SnCl,)(PPh,),] (43); cis-[PtX,L,],f cis[PtX,L,]/SnCl,,"g [Pt(SnCI3),l3- (44-46) 'X = halide, unless stated otherwise. bX = usually halide but with [RuX,L,] is also carboxylate or a-hydroxycarboxylate;L

usually PPh,, but also chiral phosphines; S = solvent. 'Product is [ (Ru(q-C,Me,) J2(p-H)Z(p-Cl)II-. = 3,4. 'Product is [ ( MCl(q-C,Me,)J,(~-H)(CL-CI)]. 'X = halide, L = tertiary phosphine. SProduct is, for example, [PtH(SnCI,),(PEt,),] -

=

tertiary phosphine,


73

species to give an inactive product. Both the Cu' and Cu" species are believed to split H, heterolytically. Additional support for a heterolytic mechanism was obtained for the Cu' species from isotope exchange experiments4. Related studies of CuSO, reduction by H, in aqueous H,SO, solution also suggested heterolytic H, cleavage by both the Cut and Cu" species5,?

-

+ H, Cd'H + H+ 2 Cu' + H' Cu"H + Cu" Cu' + H , F C u ' H + H + Cu'H + Cu" Cu"H + Cu' Cu"

(c) (4 (e) (f)

The rate is inhibited by acid; further evidence of a heterolytic mechanism. Correlation between hydrogen activating ability and ligand basicity for a series of Ag' amine and Cur' salts led to formulation of transition states such as 1, in which coordinated iigand

X helps stabilize the release of proton3. Solvent or added base could equally well be substituted for X. Reaction (g) illustrates a case in which added base promoter (acetate) finishes up as a coordinated ligand'. Coordinated carboxylates may serve a similar role; the asymmetric [ R U ( P P ~ ~ )+ ~ ]H, ~+

MeC0,-

+

RuH(CO,Me)(PPh,), MeC0,H (g) hydrogenation catalyst Ru(BINAP)(O,CMe), apparently cleaves H, heterolytically? Ru(BINAP)(O,CR),

+ H,

MeOH

+

[RuH(BINAP)(O,CR),] H+ and/or [RuH(BINAP)(O,CR)] RCOzH

+

(h)

R = -C(Me)=CH(Me) Reversibility of reactions such as (a) involving isolable hydrides (Table 1) has been demonstrated in only a few cases, e.g., with RuHCl(PPh,),, RuHBr,(AsPh,),, and transPtHCl(PEt,),, although protonation of a metal hydride with generation of H, (the reverse reaction) is well known', and is exemplified by equation (i)". NiHL4+

+ H+

-

H,

+ Ni2+ + 4L,

L = P(OEt),

0) A wide range of systems are active for catalytic hydrogenation in which a nondetectable monohydride has been postulated and the basic process of equation (a) has been invoked. The evidence is largely kinetic, coupled sometimes with isotope exchange data (H,/D,O or D,/solvent)". Many complexes of Ru", Ru"', Rh"', Ni", Pd", and Pt", containing halide ligands alone, or stabilized against reduction to the metal by further ancillary n--acid ligands (e.g. CO, SnC13-, PR,, bipy, arenes) appear to act in this way, particularly in polar solvent systems (see 14.3.2).

74


The net reaction in equation (a) for some systems must certainly occur via an initial oxidative addition of H,, followed by reductive elimination of HX for the dihydride intermediate; the second step may again be base-assisted. Of the systems listed in Table 1, some Pt" phosphine complexes give Pt"(H), species via the equilibrium outlined in (j), L = PEt,',. Cis-PtClZL,

+ H,

F=Pt(H),Cl,L,

5 Wans-PtHClLZ

+ HC1

(j)

Similarly, there are examples of rhodium(1) and iridium(1) tertiary phosphine complexes that form isolable dihydrides, which with separate treatment with external base yield monohydrides, equation (k)". Hydrogenations catalyzed by tran~-RhCl(CO)(PPh,)~~~~'~ may involve frans-RhH(CO)(PPh,), formed according to equation (1) via an undetected dihydride intermediate. In some aminophosphine analogues, a coordinated N atom may act as proton a~ceptor'~.

base M'H + base.H+ + Hz 6 Rh(H),Cl(CO)(PPh,), F= M'

RhCl(CO)(PPh,),

+ H,

& M"'(H),

RhH(CO)(PPh,),

(k)

+ HCl

(1)

The reaction shown in equation (m) establishes that a coordinated amide within an Ir"' complex definitely acts as proton acceptor16. This may be a genuine heterolytic activation of H, with a transition state like 1 but, because Irv polyhydrides are known (see 14.3.2.1, Table l), an initial oxidative addition of Hz cannot be ruled out.

Ti'", Zr'", and Hev species have do electron configurations. Accordingly, oxidative addition is not viable. Catalytic hydrogenation has been observed for Cp,ZrH,, Cp,ZrClMe, Cp,ZrMe,, Cp,TiClMe, and Cp,TiPh,'7~'8. Labeling studies rule out mechanistic pathways that invoke reductive elimination and M" intermediates". Cp,Zr(H),

-

+ olefin +Cp,ZrHR

Cp,ZrHR Cp,Zr

+ H,

Cp,Zr

+ RH

-+ Cp,Zr(H),

(n) (0)

(PI

The mechanism is believed to involve the four-centered transition state 1. Kinetic studies provide a useful probe into the nature of the transition state during metal hydride formation. Activation parameters are available for some cases of H, oxidative addition that give isolable dihydrides (particularly with Rh' and If centers"; @ values are small (up to 60 kl mol-I), and ASt values always negative ( - 155 to - 60 J mol-' K-'). Parameters measured within these ranges for net heterolytic processes giving monohydrides thus offer indirect support for the two-step process of equation (k).

-


75

Data for Ru" and Pd" systems are consistent with such a route. Where data are available, most Ru"' and Rh"' systems (and earlier Cu", Cu', Ag' systems) reveal larger AH' values and sometimes positive AS' values, and H, activation is probably genuinely heterolytic. This also seems realistic chemically in terms of the relatively inaccessible higher oxidation states that would be required for the two-step process. More extensive kinetic data, including volumes of activation20, would be valuable. Kinetic isotope effects (D, vs. H,, or D 2 0 vs. H,O for water-soluble species) are generally small and offer little insight into the different mechanisms of H, activation' 1,21. Active hydrogenation catalysts have been generated also via initial hydride formation followed by proton elimination. The reactions are generally base-promoted and can be viewed as involving a net heterolytic cleavage of H,, followed by reductive elimination. The process is exemplified best by some Rh"' systems, the overall reaction being a two-equivalent reduction of the metal complex by H, equation (q)',. The catalytically active Rh' product with L = PPh, could then follow an "unsaturated route" when the organic substrate coordinates prior to reaction with H,, or a "hydride route" involving an initial oxidative addition of H, (see 14.3.3.1). RhC13L3

+ H2

- HCI

F RhHC12L3

RhClL3

+ HC1

(4)

In the latter case, the active Rh(H),ClL, species may also be the product of heterolytic cleavage of H, (with HCl elimination) by RhHCl,L,; that is catalyst generation from RhC& requires two successive hydride substitution steps, cf. equation (q)23.The mechanism is probably dominated by the pK, of the Rh"' monohydride. Formally analogous to the net heterolytic addition of H, of equation (a) are net hydrogenolysis reactions, where the X ligand, initially a a-bonded alkyl, aryl, or especially a u-or wbonded allyl, is removed as a hydrocarbon moiety. An important example is shown in equation (r) for a general Ziegler transition metal halide/aluminum alkyl system (14.3.3.6),where the active hydride catalyst is probably generated by hydrogenolysis of the metal-alkyl bond. This step is also a key in catalytic hydrogenation of alkenes (14.3.3.1). RMX,

+ H2-RH

+ MHX,

(r)

Similarly tetravalent Ti and Zr dihydride catalysts are formed from alkyl or aryl precursors." A wide range of Group 8-10 metal hydride catalysts has been isolated or formed in situ from precursor allyl complexes. The systems are generally quite active because the catalysts are necessarily ligand deficient with sites available for substrate coordination". For an q3-allyl precursor, equation (s) initial dihydride addition to an M( ql-allyl) intermediate appears very pla~sible,~, cf. equations (i)-(1). M(q3-allyl)

+ H,

M(H),(ql-allyl)

-

MH

+ alkene

6)

Reactions in which an unsaturated ligand is removed by complete hydrogenation to yield catalytically active species are related to the hydrogenolysis process. Common examples are provided by Rh' and Ir' systems, equation (t)11*L3. Depending on L, the product solvated species can subsequently perform hydrogenation via the hydride or unsaturated routes (see 14.3.3.1, 14.3.4.5). M(diene)L,+

+ 2 H,

+

+ML,S, alkane (L = tertiary phosphine type ligand, phen, bipy; S = solvent)

(t)


76

Finally, it should be noted that an extremely wide range of hydridometal complexes, largely Group 8-10 derivatives and many of them catalysts, have been synthesized without using H,. The methods include use of alcohols, water, BH4-, AlH,-, hydrazine, and formic acid as sources of hydrogen. Organomagnesium (Grignard-type) reagents can provide hydrogen via an intramolecular transfer process from an intermediate alkyl, equation (u); hydrogen transfer from coordinated allyls, Cp, carboxylate, formyl, and phosphine or phosphite-type ligands (the orthometallation reactionz5) is also well documented. Oxidative addition of HX, equation (v), where HX may be H20, HC1, H2S04, HClO,, CH,CO,H, HCN, as well as species such as sulfides, silanes and germanes, can yield hydrides. There are also examples of B-H, N-H, and P-H bonds adding oxidatively to metal complexes; Rh hydrogenation catalysts containing carborane ligands have been made by such a route*'-26. MCH,CH,R MI

1. 2. 3. 4. 5. 6.

I.

8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.

+ HX

--

MH(CH,=CHR) M"'HX

-

MH

+ RCH=CH2

(u) (v)

(B. R. JAMES, M. T. ASHBY)

P. J. Brothers, Prog. Znorg. Chem., 28, 1 (1981). J. Halpern, J . Organomet. Chem., 200, 133 (1980). B. R. James, Homogeneous Hydrogenation, Wiley, New York, 1973, Ch. 11. A. J. Chalk, J. Halpern, J. Am. Chem. Soc., 81,5846,5852 (1959). W. J. Dunning, P. E. Potter, Proc. Chem. SOC., 244 (1960). E. A. von Hahn, E. Peters, J . Pys. Chem., 69, 547 (1965). R. W. Mitchell, A. Spencer, G. Wilkinson, J . Chem. Soc., Dalton Trans., 846 (1973). M. T. Ashby, J. Halpern,J. Am. Chem. SOC., 113,589 (1991). J. P. Collman, Ann. N.Y. Acad. Sci., 30, 479 (1968). C. A. To1man.J. Am. Chem. SOC., 92,4217 (1970). B. R. James, in Comprehensive Organometallic Chemistry, Vol. 8, Ch. 51, G. Wilkinson, F. G. A. Stone, E. Abel, eds., Pergamon Press, Oxford, 1982. Ref. 3, Ch. XIII, Sect. C. B. R. James, Adv. Organomet. Chem., 17, 319 (1979). Ref. 3, Ch. XI, Sect. E. C. Pradat, J. G. Riess, D. Bondoux, B. F. Mentzen, I. Tkatchenko, D. Houalla, J . Am. Chem. Soc., 101,2234 (1979). M. D. Fryzuk, P. A. MacNeil, Organometallics, 2, 682 (1983). P. C. Wailes, H. Weigold, A. P. Bell, J . Organomet. Chem., 43, C32 (1972). P. C. Wailes, H. Weigold, A. P. Bell, J . Organomet. Chem., 34, 155 (1972). K. I. Gell, J. Schwartz, J. Am. Chem. Soc., 100, 3246 (1978). D. A. Palmer, H. Kelm, Coord. Chem. Rev., 36,89 (1981). Ref. 3, Ch. XVII. Ref. 3, Ch. XI, Sect. H. Ref. 3, Ch. XI, Sect. B. J. R. Bleeke, E. L. Muetterties, J . Am. Chem. SOC., 103,556 (1981). G. W. Parshall, Acct. Chem. Res., 3, 139 (1970); 8, 113 (1975). C. W. Jung, M. F. Hawthorne, J. Am. Chem. SOC., 102, 3024 (1980). Ref. 3, Ch. IX, Sect. B2. Ref. 13, Sect. IIA. B. R. James, R. S. McMillan, R. H. Morris, D. K. W. Wang, Adv. Chem. Ser., 167,122 (1978). T. I. Eliades, R. 0. Harris, M. C. Zia, Chem. Commun.,1709 (1970). M. A. Bennett, T. N. Huang, T. W. Turney, J . Chem. Soc., Chem. Commun., 312 (1979). R. Iwata, I. Ogata, Tetrahedron, 2753 (1973). A. G. Hinze, Red. Trav. Chim. Pays-Bas, 92,542 (1973). A. Oudeman, F. Van Rantwijk, H. Van Bekkum, J . Coord. Chem., 4 , 1, (1974). M. E. Volpin, I. S. Kolomnikov, Russ. Chem. Rev. (Engl. Transl.), 38, 273 (1969).

14.3. Hydrogenation Reactions 14.3.2. Dih drogen Activation 14.3.2.3. olecular Hydrogen Complexes

77

rJy

36. 37. 38. 39. 40. 4 1. 42. 43. 44. 45. 46.

Ref. 3, Ch. XI, Sect. E. C. Masters, W. S. McDonald, G. Raper, B. L. Shaw, Chem. Commun., 210 (1971). B. G. Rogachev, M. L. Khidekel, Bull. Acad. Sci. USSR, Div. Chem. Sci., 127 (1969). W. C. Kalb, R. G . Teller, M. F. Hawthorne, J. Am. Chem. SOC., 101, 5417 (1979). M. R. Churchill, A. S.Julis, Inorg. Chem., 18, 1215 (1979). P. M. Maitlis, Adv. Chern. Ser., 173, 3 1 (1979). D. S . Gill, C. White, P. M. Maitlis, J. Chem. SOC., Dalton Trans., 617 (1978). Ref. 3, Ch. XIII, Sect. B. Ref. 3, Ch. XIII, Sect. C. F. Van Rantwijk, H. Van Bekkum, J. Mol. Catal., 1 , 383 (1976). J. C. Bailar, Jr., Adv. Chem. Ser., 173, 1, 1979.

14.3.2.3. Molecular Hydrogen Complexes

Molecular H, complexes have been discussed in several recent reviews’-5. The term molecular hydrogen complex (sometimes dihydrogen complex or “nonclassical” hydride complex) refers to transition metal complexes that bind H, in a side-on fashion without cleavage of the H-H bond. As early as 1976 it was suggested that RuH4(PR3), might actually be the Ru” dihydrogen complex Ru(H),( $-H2)(PR3)36; however the first unequivocal example of a molecular hydrogen complex, W( v2-H,)(CO),(PCy3),, was first reported in 1984’. Convincing evidence for formulation of W( v2-H,)(CO),(PCy3), as a molecular H, copplex came from a neutron diffraction study that revealed a HTH distance of 0.84 A, somewhat longer than that of the H-H distance of free H, (0.74 A). Only a few structures of molecular hydrogen complexes have been established by X-ray or neutron diffraction*-”. NMR is more commonly employed to characterize suspected molecular H, complexes. Characterization of molecular H, complexes has sometimes been possible using NMR spectroscopy at low T , but often the metal-hydride complexes exhibit fluxional behavior down to the lowest attainable T. An indirect method of distinguishing between classical and nonclassical metal-hydride structures has been developed that is based on measurement of spin-lattice relaxation times ( T , ) of the hydride ligand”. The “ T I method” assumes that relaxation of the proton signals of the coordinates H, is due primarily to dipole-dipole interactions with the other hydride nuclei of the molecule and, since such a mechanism depends on r6 (where r = interatomic distance between nuclei), that the short H-H separation of the nonclassical hydride ligands will give distinctively short T , relaxation times. As originally defined”, the “T, criterion” for distinguishing between classical and nonclassical hydrides was based on whether T,(min), the minimum value of T , when the temperature is varied, was shorter than 80 ms (nonclassical) or longer than 150 ms (classical) at 250 MHz. The value of T , is proportional to the magnetic field strength. The T , method has been challenged13-15; for example, although ReH,(dppe) [T,(min) = 67 ms at 250 MHz]” is a nonclassical hydride in solution by the T , criterion, a subsequent neutron diffraction study16 demonstrated it has a classical structure in the solid-state. A recent critical reassessment of the T , ~riterion’~ concluded that (1) dipoledipole relaxation is the main mechanism for relaxation of many polyhydrides, including classical hydrides, (2) the contribution of the nonhydride NMR-active nuclei to dipoledipole relaxation of the hydride ligands must in some cases be factored in, (3) for some polyhydrides of undetermined structures [including RuH4(PR3), and other polyhydrides that had previously been identified as nonclassical by the “ T , criterion”], the observed values of T,(min) are consistent with both classical and nonclassical structures. Neutron


14.3. Hydrogenation Reactions 14.3.2. Dih drogen Activation 14.3.2.3. olecular Hydrogen Complexes

77

rJy

36. 37. 38. 39. 40. 4 1. 42. 43. 44. 45. 46.

Ref. 3, Ch. XI, Sect. E. C. Masters, W. S. McDonald, G. Raper, B. L. Shaw, Chem. Commun., 210 (1971). B. G. Rogachev, M. L. Khidekel, Bull. Acad. Sci. USSR, Div. Chem. Sci., 127 (1969). W. C. Kalb, R. G . Teller, M. F. Hawthorne, J. Am. Chem. SOC., 101, 5417 (1979). M. R. Churchill, A. S.Julis, Inorg. Chem., 18, 1215 (1979). P. M. Maitlis, Adv. Chern. Ser., 173, 3 1 (1979). D. S . Gill, C. White, P. M. Maitlis, J. Chem. SOC., Dalton Trans., 617 (1978). Ref. 3, Ch. XIII, Sect. B. Ref. 3, Ch. XIII, Sect. C. F. Van Rantwijk, H. Van Bekkum, J. Mol. Catal., 1 , 383 (1976). J. C. Bailar, Jr., Adv. Chem. Ser., 173, 1, 1979.

14.3.2.3. Molecular Hydrogen Complexes

Molecular H, complexes have been discussed in several recent reviews’-5. The term molecular hydrogen complex (sometimes dihydrogen complex or “nonclassical” hydride complex) refers to transition metal complexes that bind H, in a side-on fashion without cleavage of the H-H bond. As early as 1976 it was suggested that RuH4(PR3), might actually be the Ru” dihydrogen complex Ru(H),( $-H2)(PR3)36; however the first unequivocal example of a molecular hydrogen complex, W( v2-H,)(CO),(PCy3),, was first reported in 1984’. Convincing evidence for formulation of W( v2-H,)(CO),(PCy3), as a molecular H, copplex came from a neutron diffraction study that revealed a HTH distance of 0.84 A, somewhat longer than that of the H-H distance of free H, (0.74 A). Only a few structures of molecular hydrogen complexes have been established by X-ray or neutron diffraction*-”. NMR is more commonly employed to characterize suspected molecular H, complexes. Characterization of molecular H, complexes has sometimes been possible using NMR spectroscopy at low T , but often the metal-hydride complexes exhibit fluxional behavior down to the lowest attainable T. An indirect method of distinguishing between classical and nonclassical metal-hydride structures has been developed that is based on measurement of spin-lattice relaxation times ( T , ) of the hydride ligand”. The “ T I method” assumes that relaxation of the proton signals of the coordinates H, is due primarily to dipole-dipole interactions with the other hydride nuclei of the molecule and, since such a mechanism depends on r6 (where r = interatomic distance between nuclei), that the short H-H separation of the nonclassical hydride ligands will give distinctively short T , relaxation times. As originally defined”, the “T, criterion” for distinguishing between classical and nonclassical hydrides was based on whether T,(min), the minimum value of T , when the temperature is varied, was shorter than 80 ms (nonclassical) or longer than 150 ms (classical) at 250 MHz. The value of T , is proportional to the magnetic field strength. The T , method has been challenged13-15; for example, although ReH,(dppe) [T,(min) = 67 ms at 250 MHz]” is a nonclassical hydride in solution by the T , criterion, a subsequent neutron diffraction study16 demonstrated it has a classical structure in the solid-state. A recent critical reassessment of the T , ~riterion’~ concluded that (1) dipoledipole relaxation is the main mechanism for relaxation of many polyhydrides, including classical hydrides, (2) the contribution of the nonhydride NMR-active nuclei to dipoledipole relaxation of the hydride ligands must in some cases be factored in, (3) for some polyhydrides of undetermined structures [including RuH4(PR3), and other polyhydrides that had previously been identified as nonclassical by the “ T , criterion”], the observed values of T,(min) are consistent with both classical and nonclassical structures. Neutron

14.3. Hydrogenation Reactions 14.3.2. Dihydrogen Activation 14.3.2.3. Molecular Hydrogen Complexes

78

diffraction continues to provide the most convincing evidence for nonclassical hydride complexes, but in the absence of such evidence, when carefully applied and interpreted, T , measurements can provide useful information about the structures of metal-hydride complexes. The possible role of molecular H2 complexes in transition metal-catalyzed hydrogenations has been contemplated. Known hydrogenation catalysts form molecular H, complexes; for example, [RuH(BINAP),] forms the molecular hydrogen complex [RuH(7’-H,)(BINAP),] 17. Three roles for nonclassical hydrides in hydrogenation reactions can be envisaged: (1) they may be viewed as arrested intermediates to classical metal hydrides that react with substrates, (2) they may react directly with substrates, and (3) they may serve as labile ancillary ligands. These possibilities are considered below. The nonclassical hydride complex, W( q2-H,)(CO),(PCy3),, exists in solution in equilibrium with its classical tautomer W(H),(CO)3(PCy3),1s. Similar equilibria have been observed for di- and polyhydrides’,2. Nonclassical hydride ligands must certainly exist in equilibrium with some catalytically active classical hydride complexes. As intermediates to more reactive classical hydrides, non-classical hydrides would simply tie up available catalyst. Experimental evidence exists for the direct reaction of some nonclassic hydrides with bases; the nonclassical hydride (rather than the classical hydride) of [IrH(H,)(bq)L,]+ is deprotonated by RLi19 and [CpRu(dmpe)(H,)] is deprotonated by Et,N in preference to its equilibrium dihydride form”. However, direct involvement of nonclassical hydrides in stoichiometric or catalytic reduction of substrates is still not demonstrated; a stoichiometric hydrogenation of 1-hexene has been reported using a dinuclear Ru-molecular hydrogen complex2’. Coordinated nonclassical hydride ligands are generally labile, and their loss can generate a vacant coordination site at which further reaction (such as coordination of a substrate) can occur, equation (a). However, the difference in lability between classical and nonclassical hydrides may often be small”. +

+

+

MH,

- - -- Hz

MH,-,

+S

MH,-,(S)

(a) For example, the rate of H, loss from the nonclassical W( q2-H,)(CO),(PCy,), is only ca. 13 times faster than from the classical isomer WH,(CO),(PCy,), (469 vs. 37 s-’ at 20°C)23.Furthermore, [ReH4(q2-H2)(Cyttp)]+ , which has been identified as a nonclassical hydride in the solid-state by X-ray diffraction and by the T , method in solution, is remarkably inert to sub~titution~~. Further study of the role of nonclassical hydride complexes in hydrogenation reactions is needed. (M. T. ASHBY)

G . J. Kubas, Acct. Chem. Res., 21, 120 (1988); Comments Inorg. Chem., 7 , 17 (1988). P. G. Jessop, R. H. Moms, Coord. Chem. Rev., 121, 155 (1992). R. H. Crabtree, D. G . Hamilton, Adv. Organomet. Chem., 28, 299 (1988). R. H. Crabtree, Acct. Chem. Res., 23, 95 (1990). A. G. Ginzburg, A. Bagatur’yants, A . Organomet. Chem. USSR,2, 11 1 (1989). T. V. Ashworth and E. Singleton, J . Chem. Soc., Chem. Commun., 705 (1976). G . J. Kubas, R. R. Ryan, B. I. Swanson, P. J. Vergamini, H. J. Wasserman, J . Am. Chem. SOC., 106, 451 (1984). 8. R. H. Moms, J. F. Sawyer, M. Shiralian, J. D. Zubkowski, J . Am. Chem. SOC., 107, 5581

1. 2. 3. 4. 5. 6. 7.

( 1985).

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts

79

9. J. S. Ricci, T. F. Koetzle, M. T. Bautista, T. M. Hofstede, R. H. Moms, J. F. Sawyer, J. Am. Chem. Soc., 111, 8823 (1989). 10. L. S. Van Der Sluys, J. Eckert, 0. Eisenstein, J. H. Hall, J. C. Huffman, S. A. Jackson, T. F. Koetzle, G. J. Kubas, P. J. Vergamini, K. G. Caulton, J . Am. Chem. SOC., 112, 4831 (1990). 11. F. A. Cotton, R. L. Luck, J. Chem. Soc., Chem. Commun., 1277 (1988). 12. D. G. Hamilton, R. H. Crabtree, J . Am. Chem. Soc., 110, 4126 (1988) and references therein. 13. X.-L. Luo, R. H. Crabtree, Inorg. Chem., 28, 3775. 14. F. A. Cotton, R. L. Luck, J . Am. Chem. Soc., 111,5757 (1989). 15. P. J. Desrosiers, L. Cai, Z. Lin, R. Richards, J. Halpem, J . Am. Chem. Soc., 113,4173 (1991). 16. J. A. K. Howard, S. A. Mason, 0. Johnson, I. C. Diamond, S. Crennell, P. A. Keller, J. L. Spencer, J . Chem. SOC., Chem. Commun., 1502 (1988). 17. T. Tsukahara, H. Kawano, Y. Ishii, T. Takahashi, M. Saburi, Y. Uchida, S. Akutagawa, Chem. Lett., 2055 (1988). 18. K. Zhang, A. A. Gonzales, C. D. Hoff, J . Am. Chem. Soc., 111, 3627 (1989). 19. R. H. Crabtree, M. Lavin, L. Bonnevict, J . Am. Chem. Soc., 108,4032 (1986). 20. M. S. Chinn; D. M. Heinekey, J . Am. Chem. Soc., 109,5865 (1987). 21. C. Hampton, W. R. Cullen, B. R. James, J.-P. Charland, J . Am. Chem. Soc., 110,6918 (1988). 22. J. Halpem, L. Cai, P. J. Desrosiers, Z. Lin, J . Chem. SOC.,Dalton, Trans., 717 (1991). 23. K. Zhang, A. A. Gonzales, C. D. Hoff, J. Am. Chem. Soc., I l l , 3627 (1989). 24. Y. Kim, H. Deng, D. W. Meek, A. Wojcicki, J . Am. Chem. SOC., 112,2798 (1990).

14.3.3. Classes of Soluble Catalysts The following sections summarize the developments of some well recognized classes of soluble transition metal catalysts. Discussion of these soluble catalysts will allow for coverage of the various mechanistic pathways available for hydrogenation of unsaturated C-C bonds. Following this, in 14.3.4and 14.3.5,more practical aspects will be presented. Most practical hydrogenation catalysts be heterogeneous (see 14.2),and these remain versatile tools for the synthetic chemist. Good thermal stability, fewer solvent restrictions, and ease of separation from reaction products are all significant advantages of heterogeneous catalysts particularly for commercial processes. However, heterogeneous systems are difficult to study, a major problem being reproducibility of catalyst formulation. The techniques for studying surfaces are generally applicable only at low pressure under conditions very much different than those of the actual catalytic systems. In contrast, homogeneous systems involve discrete molecular catalysts, usually a monomer, that can be studied in detail by standard spectroscopic techniques under catalytic conditions. Geometric and stereochemical features of organometallic intermediates have sometimes been fully elucidated and the kinetics and thermodynamics of the individual steps of catalytic systems may be studied. Such detailed understanding aids in catalyst design in which the ancillary ligands, solvents, conditions, etc. can be vaned in a controlled manner. In this context, the development and understanding of rhodium catalysts using chiral phosphine ligands for asymmetric hydrogenation of prochiral olefink substrates for the commercial manufacture of stereochemically pure amino acid pharmaceuticals is a remarkable success story (see 14.3.3.1and 14.3.4.5). In general for soluble hydrogenation catalysts, activation of the organic substrate involves its coordination to the metal center. If a mono- or dihydride species is formed prior to substrate activation, the catalytic process is said to operate via a “hydride route” [equation (a)]; if substrate binding precedes activation of H,, the catalysis operated via an “unsaturated route” [equation (b)]’.


14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts

79

9. J. S. Ricci, T. F. Koetzle, M. T. Bautista, T. M. Hofstede, R. H. Moms, J. F. Sawyer, J. Am. Chem. Soc., 111, 8823 (1989). 10. L. S. Van Der Sluys, J. Eckert, 0. Eisenstein, J. H. Hall, J. C. Huffman, S. A. Jackson, T. F. Koetzle, G. J. Kubas, P. J. Vergamini, K. G. Caulton, J . Am. Chem. SOC., 112, 4831 (1990). 11. F. A. Cotton, R. L. Luck, J. Chem. Soc., Chem. Commun., 1277 (1988). 12. D. G. Hamilton, R. H. Crabtree, J . Am. Chem. Soc., 110, 4126 (1988) and references therein. 13. X.-L. Luo, R. H. Crabtree, Inorg. Chem., 28, 3775. 14. F. A. Cotton, R. L. Luck, J . Am. Chem. Soc., 111,5757 (1989). 15. P. J. Desrosiers, L. Cai, Z. Lin, R. Richards, J. Halpem, J . Am. Chem. Soc., 113,4173 (1991). 16. J. A. K. Howard, S. A. Mason, 0. Johnson, I. C. Diamond, S. Crennell, P. A. Keller, J. L. Spencer, J . Chem. SOC., Chem. Commun., 1502 (1988). 17. T. Tsukahara, H. Kawano, Y. Ishii, T. Takahashi, M. Saburi, Y. Uchida, S. Akutagawa, Chem. Lett., 2055 (1988). 18. K. Zhang, A. A. Gonzales, C. D. Hoff, J . Am. Chem. Soc., 111, 3627 (1989). 19. R. H. Crabtree, M. Lavin, L. Bonnevict, J . Am. Chem. Soc., 108,4032 (1986). 20. M. S. Chinn; D. M. Heinekey, J . Am. Chem. Soc., 109,5865 (1987). 21. C. Hampton, W. R. Cullen, B. R. James, J.-P. Charland, J . Am. Chem. Soc., 110,6918 (1988). 22. J. Halpem, L. Cai, P. J. Desrosiers, Z. Lin, J . Chem. SOC.,Dalton, Trans., 717 (1991). 23. K. Zhang, A. A. Gonzales, C. D. Hoff, J. Am. Chem. Soc., I l l , 3627 (1989). 24. Y. Kim, H. Deng, D. W. Meek, A. Wojcicki, J . Am. Chem. SOC., 112,2798 (1990).

14.3.3. Classes of Soluble Catalysts The following sections summarize the developments of some well recognized classes of soluble transition metal catalysts. Discussion of these soluble catalysts will allow for coverage of the various mechanistic pathways available for hydrogenation of unsaturated C-C bonds. Following this, in 14.3.4and 14.3.5,more practical aspects will be presented. Most practical hydrogenation catalysts be heterogeneous (see 14.2),and these remain versatile tools for the synthetic chemist. Good thermal stability, fewer solvent restrictions, and ease of separation from reaction products are all significant advantages of heterogeneous catalysts particularly for commercial processes. However, heterogeneous systems are difficult to study, a major problem being reproducibility of catalyst formulation. The techniques for studying surfaces are generally applicable only at low pressure under conditions very much different than those of the actual catalytic systems. In contrast, homogeneous systems involve discrete molecular catalysts, usually a monomer, that can be studied in detail by standard spectroscopic techniques under catalytic conditions. Geometric and stereochemical features of organometallic intermediates have sometimes been fully elucidated and the kinetics and thermodynamics of the individual steps of catalytic systems may be studied. Such detailed understanding aids in catalyst design in which the ancillary ligands, solvents, conditions, etc. can be vaned in a controlled manner. In this context, the development and understanding of rhodium catalysts using chiral phosphine ligands for asymmetric hydrogenation of prochiral olefink substrates for the commercial manufacture of stereochemically pure amino acid pharmaceuticals is a remarkable success story (see 14.3.3.1and 14.3.4.5). In general for soluble hydrogenation catalysts, activation of the organic substrate involves its coordination to the metal center. If a mono- or dihydride species is formed prior to substrate activation, the catalytic process is said to operate via a “hydride route” [equation (a)]; if substrate binding precedes activation of H,, the catalysis operated via an “unsaturated route” [equation (b)]’.

80

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.1. Rhodium(1) Catalysts

M

+ H,

-3

M+S-MS-

M(H), or [MH][H+l

S ---$

(a)

H2

(b) Addition of H, to C-C unsaturated bonds is cis in nearly all cases studied. For example, addition of D, to maleate gives rneso-2,3-dideuterosuccinate,equation (c), while fumarate gives ( +. )-dideuterosuccinate. Cis addition to alkynes is also readily demonstrated; thus, deuteration of phenyl acetylene gives cis-PhCD=CDH2.

RozcHcozR DIIIIII~,

H

H

~ ~ ~ 1 1D 111

R Ho z c ~ Hc o z R

(c)

Catalytic systems involving rhodium are among the most thoroughly studied homogeneous catalytic processes. These are considered below in 14.3.3.1. (6. R. JAMES, M. T. ASHBY) 1. B. R. James, Homogeneous Hydrogenation, Ch. XI, Sect. B, Wiley-Interscience, New York,

1973. 2. B. R. James, in Comprehensive Organometallic Chemistry, Vol. 8, Ch. 51, G. Wilkinson, F. G. A. Stone, E. Abel, eds., Pergamon Press, Oxford, 1982.

14.3.3.1. Rhodium(1) Catalysts

The first use of a rhodium complex as a hydrogenation catalyst was described in 19391-2.Reports on the RhCl(PPh,), catalysts in 1965, and shortly afterward on cationic such as [Rh(diene)(PR,),] + led to great interest in rhodium (and iridium) hydrogenation catalysts5. Use of precursor complexes such as [RhCl(C,H,,),],, which contains labile cyclooctene ligands, gives a very wide range of monomeric catalysts simply by addition of a donor ligand that cleaves the halide bridge5. Scheme 1, which shows reaction pathways available for hydrogenation of alkenes using dihydride catalysts6v7,has been developed largely from studies on Rh' catalystss. The KHzklsteps define the hydride route, and Ksk, the unsaturated route via oxidative addition of H, to the metal-alkene complex. The common key dihydride-alkene intermediate 1 gives the saturate product with regeneration of catalyst M via two successive hydrogen atom transfer steps (k,, k4). The KH,and K , equilibria are usually established

Scheme 1. Mechanistic pathways for dihydride catalysts.


80


M

+ H,

-3

M+S-MS-

M(H), or [MH][H+l

S ---$

(a)

H2

(b) Addition of H, to C-C unsaturated bonds is cis in nearly all cases studied. For example, addition of D, to maleate gives rneso-2,3-dideuterosuccinate,equation (c), while fumarate gives ( +. )-dideuterosuccinate. Cis addition to alkynes is also readily demonstrated; thus, deuteration of phenyl acetylene gives cis-PhCD=CDH2.

RozcHcozR DIIIIII~,

H

H

~ ~ ~ 1 1D 111

R Ho z c ~ Hc o z R

(c)

Catalytic systems involving rhodium are among the most thoroughly studied homogeneous catalytic processes. These are considered below in 14.3.3.1. (6. R. JAMES, M. T. ASHBY) 1. B. R. James, Homogeneous Hydrogenation, Ch. XI, Sect. B, Wiley-Interscience, New York,

1973. 2. B. R. James, in Comprehensive Organometallic Chemistry, Vol. 8, Ch. 51, G. Wilkinson, F. G. A. Stone, E. Abel, eds., Pergamon Press, Oxford, 1982.

14.3.3.1. Rhodium(1) Catalysts

The first use of a rhodium complex as a hydrogenation catalyst was described in 19391-2.Reports on the RhCl(PPh,), catalysts in 1965, and shortly afterward on cationic such as [Rh(diene)(PR,),] + led to great interest in rhodium (and iridium) hydrogenation catalysts5. Use of precursor complexes such as [RhCl(C,H,,),],, which contains labile cyclooctene ligands, gives a very wide range of monomeric catalysts simply by addition of a donor ligand that cleaves the halide bridge5. Scheme 1, which shows reaction pathways available for hydrogenation of alkenes using dihydride catalysts6v7,has been developed largely from studies on Rh' catalystss. The KHzklsteps define the hydride route, and Ksk, the unsaturated route via oxidative addition of H, to the metal-alkene complex. The common key dihydride-alkene intermediate 1 gives the saturate product with regeneration of catalyst M via two successive hydrogen atom transfer steps (k,, k4). The KH,and K , equilibria are usually established

Scheme 1. Mechanistic pathways for dihydride catalysts.


81

rapidly and, if the hydrogen transfer steps are relatively fast, the resulting rate law is of the form shown in equation (a). Thus, reaction via the hydride,

unsaturated, or both routes, leads to a rate law of the same form; kinetics alone do not distinguish the reaction pathways. Complexes that give isolable dihydrides (14.3.2.1, Table 1, especially Rh’ and I+ systems) are generally considered to operate via the hydride route, although both routes have been invoked in cases where both alkene and H, coordinate separately. The well-known RhCl(PPh,); is included in Table 1 together with other complexes that operate wholly or partly by unsaturated routes; Rh systems again dominate. The unsaturated route alone is usually postulated when the dihydride is not detected (KHzis small) and when there is some evidence of alkene coordination (K, is measurable). Ligand dissociation at some stage in the catalysis [e.g., phosphine dissociation from RhCl(PPh,),] leads to rate laws more complex than in equation (a); for example, a dissociation equilibrium ( K ) will add a further term [PPh,]/K to the denominator, which can be reflected as a fractional kinetic order in meta14,5. For some experimental conditions, relatively simple limiting forms of the rate law exist, and these are often analyzed by standard inverse plots, for example (rate- l ) vs. [olefin]- l at fixed metal, H, and ligand concentrations [see equation (a)]. The “full” elucidation of a mechanism requires evaluation of kinetic and thermodynamic parameter for each individual step within the catalytic cycle, so that these can be combined to calculate the experimental rate of the overall reaction6. Such a stage has almost been reached for the widely used RhCl(PPh,), catalyst (see 14.3.4.1-14.3.4.5) in the case of one substrate, c y ~ l o h e x e n e Scheme ~ ~ ~ ~ ~2. shows the important steps that were delineated; the major catalytic pathway, corresponding to the hydride route of Scheme 1, is shown in the rectangle and involves entirely nondetectable species (only species outside the box were observed!). The important solvated dihydride 6 forms more rapidly from 5 than from 4, and the data imply that the alkene insertion step k6 is rate determining. [In the overall mechanism of Scheme 1, if 1 is nondetectable, a measured k , value does not distinguish between olefin binding or hydrogen transfer (the k, step) as rate determining.] Changes

TABLE1. CATALYSTS THATOPERATE VIA DIHYDRIDE INTERMEDIATES ROUTESFOR HYDROGENATION OF ALKENIC SUBSTRATES

WITHIN

UNSATURATED

Catalyst (ref.) [Fe(CO),(diene)] (33); [Co(PPh,Me),(solvent)] (34); Rh(I)/R,Sa (35,36); Chlororhodate(1) (35); Rh(I)/amines (37); [Rh(diphos)(solvent),]i.b (38,39); [Rh(C2H4),(CH3CN),] (40); [M(diene),]+, [M(diene)(MeCN),]+, M = Rh, Ir (41,42); [M($-C,H,,)]+: M = Rh, Ir (43); [ { IrCl(C~H~4)~]2] (42,44,45); tr~ns-[IrCl(CO)(PPh&](46,47); [IrCl(PPh3)(cod)] (48) +

+

‘R = Et, CHzPh. bChiral chelating diphosphines have also been used. ‘Ligand formed from dehydrogenation of 1,3,5-~yclooctatriene.


H

H

ratedetermining

Ik6

H I

fast

Scheme 2. Mechanistic scheme for hydrogenation of cyclohexene (S) catalyzed by [RhCl(PPh,),] in benzene (P = PPh,). The major pathway is shown in the rectangle.

in substrate, solvent, and ligands within RhXL, complexes generally can modify dissociation equilibria, hydrogenation pathways, and selectivity patterns5. For example, although RhCl(PPh,), generally catalyzes hydrogenation of alkenes more rapidly than alkynes, under certain conditions, alkynes can be highly selectively reduced to the corresponding alkene'. Favored binding of the alkyne over the alkene at some stage of the catalytic cycle (Scheme 1) readily accounts for the observed selectivity pattern. Selectivities are usually discussed in terms of the abilities of substrates to coordinate; often within alkene substrates the reactivity decreases with increasing substitution at the alkene bond'^^. Dihydride olefinic intermediates such as 1 have been detected in some [Ir(H),(diene)(PR,),] + systems at low temperature [see 14.3.2.2,equation (t)], where it was also shown that the hydride migratory insertion reaction 1 2 requires a coplanar arrangement of metal, hydride, and alkene .rr-bondlo.This stereospecific reaction is well documented in the formation of metal alkyls via reaction of a metal monohydride with alkene (and the reverse /3-hydride elimination reaction)", and the observed cis addition of the metal hydride is rationalized in terms of a transition state such as 7. The overall cis addition of H2 to alkenes via dihydride catalysts (Scheme 1) thus requires that reductive elimination of the saturated product from the alkyl hydride, the k4 step, occurs with retention of configuration at the metal-bonded carbon. This has been demonstrated for a stoichiometric reaction between Cp,Mo(D), and fumarate, and is pictured as involving the three-center transition state 812:

-


M

83

H

An alkyl hydride intermediate, e.g., 2, has been characterized at low T in the hydrogenation of methyl-(a-a-acetamidocinnamate catalyzed by [Rh(diphos)(MeOH),] + I 3 . This system is considered in more detail in 14.3.4.5. It is of particular interest because related catalysts containing chiral derivatives of diphos are highly effective for asymmetric hydrogenation of prochiral cinnamate substrates to give optically active amino acids. The systems operate via the unsaturated route of Scheme 1. Unlike the bis-monodentate phosphine systems (see 14.3.2.1, Table l), the solvated monochelated diphosphine systems (phosphine = Ph,P(CHR),PPh,; n = 1-4, R = H or alkyl, including chiral derivatives) do not form dihydrides (see Table l)5,'3. With chiral phosphines, the Rh(a1kene) intermediate (Scheme 1) exists as two possible diastereomers in which either one or the other face of the alkene is coordinated to the rhodium. Preferential cis addition of H, to one alkenic face gives the optically active saturated product via the alkyl hydride intermediate (see also 15.3.4.5)5*'3,'4.A wide range of chiral phosphorusdonors has been used with Rh' for asymmetric hydrogenation3914915 (and 14.3.4.5). The mechanisms by which catalyst precursors of the type RhCl,(BH,)(amide)py, give a range of effective hydrogenation^^^'^, including asymmetric syntheses when the amide is chiral (15.3.4.5), have not been elucidated. Reaction schemes for hydrogenation of dienes using dihydride catalysts usually follow the basic hydride or unsaturated pathways of Scheme 1; species 2 will be, however, a hydrido cr-alkenyl or a r-ally1 intermediate [see equation (f) below]. Transfer of the second hydrogen yields a monoene product; if diene coordination is preferred over monoene coordination, selective reduction of diene to monoene can occur. Some Rh' (and Ir') dihydride catalysts that effect such catalysis are listed in Table 2, although systems based on the Cr and Fe carbonyl have been studied most (see also 14.3.3.5)5.

TABLE2. RHODIUMCATALYSTS FOR SELECTIVE REDUCTIONOF DIENES AND POLYENES TO MONOENES'

Hydride Catalyst, or Precursor (Ref.)

aSee ref. 5, Table 5, for a more extensive list of catalysts. bL = polyethynenimine, tyrosine, phen. dL = monodentate tertiary phosphine; S = solvent or L.

'L = H,O, PR,.

84


The equivalence of Scheme 1 written for hydrogenation of alkynes to alkenes requires that species 2 be a hydridovinyl intermediate. These have not been detected within any mononuclear catalytic systems, but treatment of Os,H,(CO),, with acetylene yields Os,H(C,H,)(CO),, that contains both bridging hydride and vinyl groups". Systems based on cationic and neutral Rh' precursors of the type [Rh(diene)LJ+, [Rh(diene)(py)L] , and Rh(diene)LX (L = tertiary phosphine, X = benzoate) seem particularly effective for terminal and internal alkynes. Although these complexes are typical precursors of dihydrides, since base is sometimes required for the catalysis, the monhydride species (see below) may be involved [ 14.3.2.2, equation (k)]3~'*~19, +

M,(or 2M) Scheme 3. Mechanistic pathways for monohydride catalysts. [CpRh(PMe,)(C,&)Hl

+

@

[CpRh(PMe,)(C,H,)L]+ (L = CH3N02 or C2H4) (b) Proposed mechanistic pathways for hydrogenation of monoalkenes using monohydride catalysts are summarized in Scheme 3; step (a) defines the hydride route and step (b) the unsaturated route. Step (b), a net heterolytic cleavage of H, by a metal alkene complex, was first proposed for chlororuthenate(I1) species (14.3.3.7). Similar cleavage prior to alkene coordination, step (a), seems likely for some Pt" systems (14.3.3.4). Step (a) for characterized monohydrides is well documented, especially within rhodium and iridium (see below), cobalt (14.3.3.2), and ruthenium (14.3.3.7) catalytic processes. A key alkyl intermediate 10 is usually assumed to be formed from 9 via the well known insertion reaction discussed above (see 7), although such a process has not been observed directly within a catalytic hydrogenation system. The hydride migration is usually facile; other steps within the cycle are rate determining. So-far-isolated monohydrido-alkene or -diene complexes5*''720*21are not effective in catalytic hydrogenation of that alkene or diene. An isolated (hydrid0)rhodacarbaborane complex with a chelated $-3,4-butenyl side chain attached to a dicarbollide ligand C atom becomes an extremely efficient hy-


85

drogenation catalyst, but only after the butenyl moiety has been hydrogenated off and a vacant coordination site has been generated,'. The example shown in equation (b) is a rare case where a hydrido-olefin species is shown to be in equilibrium with an alkyl, although the system was not reported to effect catalytic hydrogenation of C2H4,l. Formation of saturated product from the alkyl10 under catalytic conditions (Scheme 3) requires reaction via steps (d,g), (e), or (f).The RhH(CO)(PPh,), complex4s22and the iridium a n a l ~ g u edissociate ~ , ~ ~ a phosphine in solution and then operate via the (a,c,d,g) cycle; the net hydrogenolysis using H,, steps (d,g), generally proceeds via oxidative addition of H, to give the dihydridoalkyl 11, followed by reductive elimination of alkyl hydride (the saturated product) and regeneration of the monohydride catalyst [cf. 14.3.2.2, equation (t) and related discussion]. Species such as 11 have never been detected, and the evidence for step (d) is by analogy with known oxidative addition of H, to d6 (Ru") and d8 systems (Pt", Co', Rh', I?) (see 14.3.2.1). Experimental rate laws for catalytic hydrogenations using monohydrides via the (a,c,d,g) cycle are often of the form shown in equation (c). They are interpreted in terms of a rapid equilibrium formation of alkyl [steps (a) and (c), governed by K ] followed by a rate-determining H, addition [step (d), governed by k]; k and its associated activation parameters can thus be determined. Ligand dissociation from the metal at some stage again leads to more complex rate la~s.~,~,~~. -d[H,] - kK[MHl[alkenel[H,] Rate = dt 1 K[alkene]

+

Since step (a) occurs stereospecifically via the transition state 7 (see above), overall cis addition of H, in hydrogenations catalyzed by monohydrides requires that steps (g), (e), or (f) occur with retention of configuration at the metal-bonded carbon. The case for step (g) follows by analogy with the molybdenum model data noted in connection with the dihydride catalytic systems (see 8 above). However, it should be noted that deuteration of a complex RCo(dmgh),B [where R = (R)-l-(methoxycarbony1)ethyl and B is (R)-a-methylbenzylamine] is said to give (S)-methylpropionate-2-d, equation (d). This means that inversion of configuration has occurred at the carbon25; more mechanistic details would be valuable. Cleavage of the M-C bond via protonolysis, step (e), or via reaction with a further mole of metal-hydride, step (f), is discussed in 14.3.3.4, 14.3.3.7, and 14.3.3.2, 14.3.3.3, respectively. Following step (e) the metal reenters the catalytic cycle either through the unsaturated route via formation of M(alkene), or the hydride route via formation of MH; step (f), a binuclear reductive elimination reaction, requires regeneration of catalyst via oxidative addition of H, to the dimer (or 2M species). Co-CHMe(C0,Me)

5 MeCH(D)CO,Me

The reversibility of steps (a) and (c) provides an explanation for the isomerization of alkenes in the absence of H, by addition and elimination of metal hydride, equation (e), and for the exchange between the metal hydride and hydrogens on the alkene. These reactions are common for monohydrides; RhH(CO)(PPh,), provides a good example4. Such olefin isomerization (discussed in more detail in 15.5) may be an accompanying reaction under hydrogenation conditions. (Dihydride systems generally show little alkene isomerization or exchange, which implies that k4 is relatively large within Scheme 1 and is consistent with the general lack of detectable hydridoalkyls.) Isomerization reaction 13 + 15 involves /3-hydrogen abstraction from alkyl 14 formed by Markownikov ad-

86


dition of the metal hydride. With the RhH(CO)(PPh,), reaction’, which functions via a trans-RhH(CO)(PPh,), intermediate, an efficient hydride transfer to terminal aklenes (and fast H exchange) may occur only via anti-Markownikov addition (13 + 12) since formation of 14 in which the alkyl group is mutually cis to trans phosphines is not favored sterically. Thus, hindered Markownikov addition readily explains the relatively slow isomerization observed during efficient hydrogenation of terminal olefins with this catalyst; slow hydrogenation of, and exchange with, 2-alkenes results from hindered addition of Rh-H in either direction. At higher dilution a monophosphine species becomes the catalyst. Steric factors are less important and selectivity for terminal alkenes is less. Such detailed discussion on the important metal hydride addition is generally difficult because of uncertainty about the polarity of the M-H bond (14.3.2.1) and the direction of addition to alkenes. This will be considered further in 14.3.3.3. RCH,CH, C H, I

F= RCH2CH=CH2

M

12

I MH 13

RCH, C HCH, F RCH=CHCH, II I M hH 14 15

(e)

Certain monohydride catalysts, including a few Rh species, exhibit high selectivities toward hydrogenation of conjugated dienes and polyenes to give monoenes (see Table 2). The basic pathways of Scheme 3 becomes modified in that addition of metal hydride now yields (T- or v-ally1 intermediates instead of 10. Associated ligands L can be important because at lower concentrations a more unsaturated metal environment favors formation of wallyls (17); the syn form is preferred:

-CH=CH-

CHsCH-

MHL,

’ -CHZ-CH-CH=CHI

MLn 16

-CH,-CH=CH18

17

11

CH-

(f)

I

MLn

Understanding such processes stems largely from studies on cobalt system^^^,^' (14.3.3.2, 14.3.3.3), but the principles apply generally. Product selectivity (e.g., butadiene can give 1- or 2-butenes) depends on the relative concentrations of 16, 17, and 18, and their rate of subsequent decomposition by any of the available pathways (d,g), (e), or (f)5. For example, net hydrogenolysis of 16 and 18 (d,g) would give l-butene and 2-butenes, respectively. Oxidative addition of H, to the n--ally1 17 could give two dihydro metal complexes analogous to 16 and 18, and thus less selective pathways. Steps (e) or (f) allow formation of l-butene by protonation of 18 via a carbonium ion that undergoes 1,2-elimination,equation (g), or via metal hydride attack at the y-position of 18, equation (f), at the a-position of 16 or on the wallyl17. Either cis- or truns-2-butene could result from y-protonation of 16 or attack of M-H on 18. Formation of 2-butene via attack on M-H on 17 would strongly favor the trans isomer. Generally, selectivity for hydrogenation of dienes in the presence of monoalkenes is attributed to the favored formation of wallyls. Many of the monohydrides listed in Table 2 that effect selective reduction of nonconjugated molecules probably achieve this via a prior isomerization of the substrate to the conjugated form, equation (e).

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.1. Rhodiurn(1) Catalysts

?

ML,-CH26HCH2CH3

-

or MLn-CH2CH=CHCH3 MHL,

2MLn

ML,

87

+ CH2=CHCH2CH3 (8)

+ CH2=CHCH,CH3

Pathways corresponding to those of Scheme 3 are applicable for hydrogenation of alkynes by monohydride catalysts, the key intermediate now is a vinyl rather than an alkyl (10). Some rhodium complexes effective for hydrogenation of alkynes to alkenes were mentioned earlier and are listed in Table 3 with other rhodium catalysts. Many rhodium catalysts are available; however, RhCl(PPh,), is the most widely used homogeneous hydrogenation system. Biphasic systems are used to hydrogenate aqueous solutions of water-soluble alkenes, by shaking with benzene solutions of RhCl(PPh,), under H,. Removal of the aqueous layer leaves the catalyst solution available for further usez8.The same principle has been used in reverse by using water-soluble rhodium phosphine complexes in conjunction with substrate-containing organic phases2*. Organometallic-based catalysis in aqueous media is attractive industriallyz9. Tables of data are published elsewhere for hydrogenations and deuterations using RhCl(PPh3)3. These include as substrates simple alkenes, dienes, cyclic monoenes and dienes, allenes, terpenes, natural products with exocyclic methylene groups, antibiotics, spirocyclic compounds, prostaglandins, nitriles, ketones, aldehydes, and nitro corn pound^^^^*^^,^^. The continuing book series by Fieser and Fiese?' is useful for tracing current applications of RhCl(PPh,), and other metal complex catalysts. In addition to RhCl(PPh,),, the less well understood RhC1,(BH,)(dmf)py2 complex is a versatile catalyst (Tables 2,3),which also effects hydrogenation of aromatics, e.g., pyridine to piperidine and quinoline to 1,2,3,4-tetrahydroq~inoline~~~,'~. Stereoselective hydrogenation of benzenes and polyaromatics has been accomplished using [(~5-C5Me5)RhC12]2 (14.3.2.2Table 1), Rh2(N-phenylanthranilate),C1-, and Rh[P(OPr'),],( 77,-C3H5) complexes (14.3.5.3). Many applications of rhodium complex catalysts are given in 14.3.4-14.3.7. (B. R. JAMES, M. T. ASHBY)

1 . M. Iguchi, J. Chem. SOC.Jpn., 60, 1287 (1939).

TABLE3. RHODIUM CATALYSTS THATEFFECTHYDROGENATION OF ALKYNES TO MONOENES~ Complex (Ref.) ~

[RhCI(PPh,),] (63); [Rh(diene)L,] +,[Rh(diene)LL'] +,[Rh(diene)LXIb (53); [Rh,(O,CMe),]/H+ (64); [RhCl,(BH,)(DMF)py,] (41); [Rh,L,Cl] -'(41); [Rh(TPP)I - d (65); [Rh(H),Cl(PBu',),I (66); [Rh(SnX,)(PPh3)31e(67); tRh(O,CPh)(PPh,)(cod)I, [Rh(PPhd(~~)(cod)l+ (68) "See ref. 5, Table 13, for a more extensive list of catalysts. bL = tertiary phosphine, L' = py, X = benzoate. 'L = phenylacetate,N-phenylanthranilate, tyrosine. dTPP = dianion of meso-tetraphenylporphyrin. = CI, Br.

88


2. B. R. James, Homogeneous Hydrogenation, Ch. XI, Sect. A, Wiley-Interscience, New York, 1973. 3. B. R. James, Adv. Organomet. Chem., 17,319 (1979). 4. Ref. 2., Ch. XI, Sect. D. 5. B. R. James, in Comprehensive Organometallic Chemistry, Vol. 8, Ch. 51, G. Wilkinson, F. G . A. Stone, E. Abel, eds., Pergamon Press, Oxford, 1982. 6. Ref. 2, Ch. XVII. 7. Ref. 5 , Ch. 6. 8. J. Halpem, in Organotransition Metal Chemistry, Y. Ishii, M. Tsutsui, eds., Plenum, New York, 1975, p. 109. 9. Ref. 2, Ch. XI, Sect, B. 10. R. H. Crabtree, Acct. Chem. Res., 12, 331 (1979). 11. J. P. Collman, L. S . Hegedus, J. R. Norton, R. G . Finke, Principles and Applications of Organotransition Metal Chemistry, University Science Books, Mill Valley, CA, 1987. 12. A. Nakamura, S . Otsuka, J. Am. Chem. Soc., 95, 7262 (1973). 13. A. S. C. Chan, J. J. Pluth, J. Halpem, J. Am. Chem. SOC., 102, 5952 (1980). 14. H. Kagan, in Comprehensive Organometallic Chemistry, Vol. 8, Ch. 53, G. Wilkinson, F. G. A. Stone, E. Abel, ed., Pergamon Press, Oxford, 1982. 15. L. Marko, J. Bakos, in Aspects of Homogeneous Catalysis, Vol. 4, R. Ugo, ed., Reidel, Dordrecht, 1981, p. 145; J. D. Momson, ed. Asymmetric Synthesis, Vol. 5 , Academic Press, New York, 1986. 16. F. J. McQuillin, Homogeneous Hydrogenation in Organic Chemistry, Reidel, Dordrecht, 1976. 17. A. J. Deeming, S . Hasso, M. Underhill, J. Chem. SOC.,Dalton Trans., 1614 (1975). 18. R. R. Schrock, J. A. Osbom,J. Am. Chem. SOC., 98,2143 (1976). 19. R. H. Crabtree, A. Gautier, G . Giordano, T. Khan, J. Organomet. Chem., 141, 113 (1977). 20. M. S . Delaney, C. B. Knobler, M. F. Hawthorne, J. Chem. SOC.,Chem. Commun., 849 (1980). 21. H. Werner, R. Feser,Angew. Chem. Int. Ed. Engl., 18, 157 (1979). 22. Ref. 2, Ch. XI, Sect. F. 23. Ref. 2, Ch. XII, Sect. B. 24. Ref. 2, Ch. IX, Sect. B. 25. Y. Ohashi, Y. Sasada, Bull. Chem. SOC.Jpn., 50, 2863 (1977). 26. Ref. 2, Ch. X, Sect. A. 27. Ref. 2, Ch. X, Sect. F. 28. Y. Dior, J. Manassen, J. Mol. Catal., 2 , 219 (1977). 29. E. G. Kuntz, CHEMTECH, 570 (1987). 30. A. J. Birch, D. H. Williamson, Org. React., 24, 1 (1976). 31. A. P. G . Kieboom, E. van Rantwijk, Hydrogenation and Hydrogenolysis in Synthetic Organic Chemistry, Delft University Press, 1977. 32. M. Fieser, L. F. Fieser, Reagents for Organic Synthesis, Wiley, New York; e.g., Vol. 6, 1977. 33. Ref. 2, Ch. IX, Sect. A; Chap. XIV, Sect. B. 34. K. Kawakami, T. Mizoroki, A. Ozaki, J. Mol. Catal., 5, 175 (1979). 35. Ref. 2, Ch. XI, Sect. H. 36. B. R. James, F. T. T. Ng, Can. J. Chem., 53, 797 (1975). 37. F. Pruchnik, Inorg. Nucl. Chem. Lett., 10, 661 (1974). 38. B. R. James, Adv. Organomet. Chem., 17, Sect. IIB, IIIA. 39. A. S. C. Chan, J. J. Pluth, J. Halpem, J. Am. Chem. SOC.,102, 5952 (1980). 40. F. Maspero, E. Perrotti, F. Sirnonetti, Ger. Pat. 2 263 882 (1973); Chem. Abstr., 79, 92 386 (1973). 41. Ref. 2, Ch. XI, Sect. H. 42. Ref. 2, Ch. XII, Sect. C. 43. P. T. Draggett, M. Green, F. W. S . Lowrie, J. Organomet. Chem., 135, C60 (1977). 44. H. Van Gaal, H. A. M. Cuppers, A. van der Ent, J. Chem. SOC., Chem. Commun., 1694 (1970). 45. C. Y. Chan, B. R. James, Inorg. Nucl. Chem. Lett., 9, 135 (1973). 46. Ref. 2, Ch. XII, Sect. A. 47. M. Bumett, R. J. Momson, C. J. Strugnell, J. Chem. SOC., Dalton Trans., 701 (1973). 48. R. N. Haszeldine, R. J. Lunt, R. V. Parish, J. Chem. SOC.(A), 3711 (1971). 49. V. M. Frolov, 0. P. Parenago, L. P. Shuikina, Kinet. Katal., 19, 1608 (1978). 50. V. N. Perchenko, I. S . Mirskova, N. S . Nametkin, Dokl. Akad. Nauk SSSR, 251, 1437 (1980).

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.2. Cobalt Cyanide Systems

89

51. M. V. Klyuev, B. G. Rogachev, M. L. Khidekel, Izv. Akad. Nauk SSSR, Ser. Khirn., 2620 (1978). 52. Ref. 2, Ch. XI, Sect. G. 53. Ref. 3, Sect. IIB. 54. S. Siegel, G. Perot, J . Chern. SOC., Chern. Cornrnun., 114 (1978). 55. A. Spencer, J . Organornet. Chern., 93,389 (1975). 56. Ref. 2, Ch. XI, Sect. D. 57. Ref. 3, Sect. IIA. 58. Ref. 2, Ch. XI, Sect. B. 59. Ref. 3, Sect IIC. 60. Ref. 2, Ch. XII, Sect. B. 61. Ref. 2, Ch. XI, Sect. H; Ch. XII, Sect. C. 62. L. K. Freidlin, Y. A. Kopyttsev, E. F. Litvin, N. M. Nazarova, Zh. Org. Khirn., 10,430 (1974). 63. Ref. 5, Sect. 51.1.6. 64. Ref. 2, Ch. XI, Sect. I. 65. B. R. James, D. V. Stynes, J . Am. Chern. SOC., 94,6225 (1972). 66. C. Masters, W. S . McDonald, G . Raper, B. L. Shaw, J . Chern. SOC., Chern. Cornrnun., 210 (1971). 67. N. V. Borunova, L. K. Freidlin, P. G. Antonov, Y. N. Kukushkin, Y. N. Trink, V. M. Ignatov, A. R. Ganeeva, Izv. Akad. Nauk SSSR,Ser. Khirn., 2045 (1977). 68. R. H. Crabtree, A. Gautier, G . Giovdano, T. Khan, J . Organornet. Chern., 141, 113 (1977).

14.3.3.2. Cobalt Cyanide Systems (i) Reaction of [Co(CN),I3- with HP. Absorption of H, by aqueous solutions of [Co(CN),I3- was discovered 50 years ago’. Since then much effort has focused on characterizing the species present in solution2. The [Co(CN),I3- is generally prepared in situ. Aqueous solutions prepared from Co(CN), and CN- ion with [CN-1 > ~ [ C O ] are green at room T . A violet, diamagnetic solid, isolated from the latter solutions, has been identified as K , [ C O , ( C N ) , ~ ] . ~ H However, ,~~. the green solutions exhibit paramagnetisim that corresponds to one unpaired electron per ~ o b a l f l .Though ~. dimeric in the solid-state, the species apparently exists as a monomer in solution. Pratt and coworkers have deduced from the visible spectrum that the monomer in solution is [Co(CN),(H,0)l3- and not [Co(CN),I3- ,*’; however, data from kinetic studies* and ESR measurements in solutiong and in the solid-state” show that the principle species in solution is [CO(CN),]~-;coordination of water must be weak. Study of [Co(CN),I3- is complicated by its instability in H,O. Aqueous solutions of [Co(CN),I3- “age” on standing. The nature of the aging process is not well understood. Freshly prepared aqueous solutions of [CO(CN),]~-([CO]< 0.2 M) at 0-25°C absorb approximately 0.5 equivalents of H, (1 atm) per Co’v5. Aged solutions of [Co(CN),I3- absorb less H,. Loss of paramagnetisim of the solutions parallels the decrease in reactivity toward H,. This has been attributed to dimerization (2 Co” + CO”,)~, but more likely is due to disproportionation (2Co” + Co’ CO“’)”.Solutions of [Co(CN),I3- in anhydrous EtOH are not subject to “aging”; however, the rate at which such solutions take up H, is substantially slower than corresponding aqueous solutions12. All evidence so far suggests that the mechanism of reaction of [Co(CN),I3- with H, in H,O involves homolytic splitting13:

+

2 [CO(CN),]~- + H,

-

2 [CoH(CN),I3-

(a)

(ii) HydrogenationReactions Catalyzed by [Co(CN),I3-. About 20 years after the discovery of reaction (a), it was demonstrated that the aqueous [Co(CN),I3- system



89

51. M. V. Klyuev, B. G. Rogachev, M. L. Khidekel, Izv. Akad. Nauk SSSR, Ser. Khirn., 2620 (1978). 52. Ref. 2, Ch. XI, Sect. G. 53. Ref. 3, Sect. IIB. 54. S. Siegel, G. Perot, J . Chern. SOC., Chern. Cornrnun., 114 (1978). 55. A. Spencer, J . Organornet. Chern., 93,389 (1975). 56. Ref. 2, Ch. XI, Sect. D. 57. Ref. 3, Sect. IIA. 58. Ref. 2, Ch. XI, Sect. B. 59. Ref. 3, Sect IIC. 60. Ref. 2, Ch. XII, Sect. B. 61. Ref. 2, Ch. XI, Sect. H; Ch. XII, Sect. C. 62. L. K. Freidlin, Y. A. Kopyttsev, E. F. Litvin, N. M. Nazarova, Zh. Org. Khirn., 10,430 (1974). 63. Ref. 5, Sect. 51.1.6. 64. Ref. 2, Ch. XI, Sect. I. 65. B. R. James, D. V. Stynes, J . Am. Chern. SOC., 94,6225 (1972). 66. C. Masters, W. S . McDonald, G . Raper, B. L. Shaw, J . Chern. SOC., Chern. Cornrnun., 210 (1971). 67. N. V. Borunova, L. K. Freidlin, P. G. Antonov, Y. N. Kukushkin, Y. N. Trink, V. M. Ignatov, A. R. Ganeeva, Izv. Akad. Nauk SSSR,Ser. Khirn., 2045 (1977). 68. R. H. Crabtree, A. Gautier, G . Giovdano, T. Khan, J . Organornet. Chern., 141, 113 (1977).

14.3.3.2. Cobalt Cyanide Systems (i) Reaction of [Co(CN),I3- with HP. Absorption of H, by aqueous solutions of [Co(CN),I3- was discovered 50 years ago’. Since then much effort has focused on characterizing the species present in solution2. The [Co(CN),I3- is generally prepared in situ. Aqueous solutions prepared from Co(CN), and CN- ion with [CN-1 > ~ [ C O ] are green at room T . A violet, diamagnetic solid, isolated from the latter solutions, has been identified as K , [ C O , ( C N ) , ~ ] . ~ H However, ,~~. the green solutions exhibit paramagnetisim that corresponds to one unpaired electron per ~ o b a l f l .Though ~. dimeric in the solid-state, the species apparently exists as a monomer in solution. Pratt and coworkers have deduced from the visible spectrum that the monomer in solution is [Co(CN),(H,0)l3- and not [Co(CN),I3- ,*’; however, data from kinetic studies* and ESR measurements in solutiong and in the solid-state” show that the principle species in solution is [CO(CN),]~-;coordination of water must be weak. Study of [Co(CN),I3- is complicated by its instability in H,O. Aqueous solutions of [Co(CN),I3- “age” on standing. The nature of the aging process is not well understood. Freshly prepared aqueous solutions of [CO(CN),]~-([CO]< 0.2 M) at 0-25°C absorb approximately 0.5 equivalents of H, (1 atm) per Co’v5. Aged solutions of [Co(CN),I3- absorb less H,. Loss of paramagnetisim of the solutions parallels the decrease in reactivity toward H,. This has been attributed to dimerization (2 Co” + CO”,)~, but more likely is due to disproportionation (2Co” + Co’ CO“’)”.Solutions of [Co(CN),I3- in anhydrous EtOH are not subject to “aging”; however, the rate at which such solutions take up H, is substantially slower than corresponding aqueous solutions12. All evidence so far suggests that the mechanism of reaction of [Co(CN),I3- with H, in H,O involves homolytic splitting13:

+

2 [CO(CN),]~- + H,

-

2 [CoH(CN),I3-

(a)

(ii) HydrogenationReactions Catalyzed by [Co(CN),I3-. About 20 years after the discovery of reaction (a), it was demonstrated that the aqueous [Co(CN),I3- system

90


effectively catalyzes reduction of various organic and inorganic substrates at room T and 1 atm H,14. Inorganic substrates that are reduced by the [Co(CN),I3-/H2 system include H,O,, O,,s8, halogens, Fe(CN)63-, MnO,-, S,03'-, Cr,O?-, NO,-, NH,OH, and pentacyanocobaltate(II1) c~mplexes'~. The [Co(CN),I3 - hydrogenation system is poisoned by excess 0,; however, the poisoning is reversible if the 0, is removed16. Comparatively few organic substrates are reduced by [Co(CN),I3 -/H2. Conjugated dienes such as butadiene and cyclohexadiene are reduced to m o n o e n e ~ ~ ~Kinetic ,~~-~~. measurements and deuterium labeling studies suggest that [CoH(CN),I3- reacts reversibly with one equivalent of butadiene," and the resulting intermediate alkyl reacts irreversibly with a second equivalent of [CoH(CN),I3- :

+ C4H, S

[CoH(CN),I3[CO(CN),(C,H,)]~-

+ [CoH(CN),I3-

-

[CO(CN),(C,H,>]~-

- -

2 [Co(CN),I3-

+ C4H8

(b) (c)

Conjugated dienes and polyenes are generally reduced via the (a,c,f) cycle (14.3.3.1, Scheme 3): C"'H

+ substrate (S)

Co"'H

Co"'(SH)

SH,

+ (Co"', -% 2 Co"'H)

(d)

Only activated monoenes are hydrogenatedz8. These include carvene, mesityl oxide, 2cyclohexenone, and ben~alacetone'~.Some styrenes are hydrogenated; a-functionalized styrenes react, but P-functionalized styrenes do not17,'8,30,31. Similarly, only activated ketones such as benzil, diacetyl and p-benzoquinone are hydrogenated to a l ~ o h o l s ~ * ~Often ~ ' ~catalytic ~ ~ ~ , reduction ~ ~ ~ ~ ~of. a ketone is observed only in the presence of added OH-. The base is believed to react with an intermediate to give [CO(CN),(OH)]~-and the reduced ~ubstrate'~. Aryl ketones such as acetophenone and benzophenone are not r e d ~ c e d ' ~Several . examples of nitro and nitroso compound reductions have been r e p ~ r t e d ' ~ . ' ~ ' ~ ~ . In contrast to the mechanism of equation (d), which is based on hydride transfer from [CoH(CN),I3-, detailed kinetic studies with some a$-unsaturated acids, esters, and nitriles indicate successive hydrogen atom transfers to uncoordinated substrates from [CO"(CN),(*H)]~-species equations (e), (f)35. Co"(.H)

+S 2 -Co"

-

*Co"

Co"(*H + (*SH 2 *Co" + SH,)

Co",

3 2 Co"(*H)

(e) (f)

See 14.3.2.1 for a discussion of the ambiguity in formulation of metal hydrides. Other activated alkenes including various styrenes appear to be reduced via the same mechanism. Hydrogenation of trans- 1-phenyl- 1,3-butadiene may proceed via the radical mechanism in glycerol-MeOH solution, but via the organometallic mechanism, equation (d), in H,036. Even with the radical mechanism, organocobalt species may be present but they simply tie up available cobalt and inhibit the catalysis. Kinetic data on addition of [CoH(CN),I3 - to a$-unsaturated compounds are generally consistent with formation of the organocobalt product via intermediate radical species [equation (g)] rather than via the usual four-center transition state or electrophilic attack via CO'--H"~~. Indirect evidence supporting a [Co"(CN),( -H)l3- formation includes an observed radical polymerization initiated by the hydride and production of dimeric substrate hydrogenation

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.2.Cobalt Cyanide Systems

91

products (2 .SH + S2H2),’. That monohydride catalysts can operate via the free radical pathways outlined in equations (e); (f) has been established definitively in more recent studies on hydridocarbonyl complexes (14.3.3.3).

RCH=CHX

2 {RCH, ~ [ C Othe ] species [Co(CN),]’- is preferred, a species which is less active than [CoH(CN),(diamine)]-. When [CN-] < ~ [ C O excess ], diamine yields [Co(CN),(diamine),], which under mild conditions does not react with H,. (M. T. ASHBY, B. R. JAMES)

1. M. Iguchi, J . Chem. SOC.Jpn., 63,634, 1752 (1942). 2. B. R. James, Homogeneous Hydrogenation, Wiley, New York, 1973, pp. 106-107 and references therein. 3. B. M. Chadwick, A. G. Sharpe, Adv. Inorg. Chem. Radiochem., 8, 83 (1966) and references therein. 4. A. W. Adamson, J . Am. Chem. Soc., 73,5710 (1951). 5. G. A. Mills, S. Weller, A. Wheeler, J . Phys. Chem., 63, 403 (1959). 6. J. M. Pratt, R. J. P. Williams, J . Chem. SOC., A, 1291 (1967). 7. R. G. S. Banks, J. M. Pratt, J . Chem. SOC.,A, 854 (1968). 8. J. P. Birk, J. Halpem, J . Am. Chem. SOC., 90, 305 (1968). 9. B. M. Chadwick, L. Shields, J . Chem. Soc., Chem. Commun., 650 (1969). 10. J. P. Maher, J . Chem. SOC., A , 2918 (1968). 83,3366 (1961). 11. N. K. King, M. E. Winfield, J. Am. Chem. SOC., 12. G. Pregaglia, D. Morelli, F. Conti, G. Gregorio, R. Ugo, Discuss. Faraday SOC., 46, 110 (1968). 13. J. Halpem, Inorg. Chim. Acta, 77, L105 (1983) and references therein. 14. J. Kawiatek, I. L. Mador, J. Y. Seyler, J . Am. Chem. SOC., 84, 304 (1962). 15. J. Kwiatek, Catal. Rev., I, 37 (1967) and references therein. 16. J. H. Bayston, M. E. Winfield, J. Catal, 3, 123 (1964). 17. J. Kwiatek, I. L. Mador, J. Y. Seyler, J . Am. Chem. Soc., 84, 304 (1962). 18. J. Kwiatek, I. L. Mador, J. Y. Seyler, Advances in Chemisrry, Vol. 37, American Chemical Society, Washington, D.C., 1963, p. 201. 19. T. Suzuki,T. Kwan, J . Chem. SOC. Jpn., 86,713, 1198, 1341 (1965). 20. J. Kwiatek, J. K. Seyler, J. Organomet. Chem., 3, 421, 433 (1965). 21. M. G. Bumett, P. J. Conolly, C. Kemball, J . Chem. SOC., A, 991 (1968). 22. J. Halpem, L. Y. Wong, J . Am. Chem. SOC., 90, 6665 (1968). 23. W. Strohmeier, N. Iglauer, Z. Physi. Chem. (Franw), 61, 29 (1968). 24. T. Funabiki, K. Tarama, Bull. Chem. SOC.Jpn., 44, 945 (1971). 25. T. Funabiki, M. Matsumoto, K. Tarama, Bull. Chem. SOC.Jpn., 45, 2723 (1972). 26. T. Funabiki, M. Mohri, K. Tarama, J . Chem. Soc., Dalton Trans., 1813 (1973). 27. D. L. Reger, M. M. Habib, D. J. Fauth, J . Org. Chem., 45, 3860 (1980). 28. J. Kwiatek, J. Y. Seyler, Advances in Chemistry, Vol. 70, American Chemical Society, Washington, D. C., 1968, p. 207. 29. G. S. R. Subba Rao, J. Rajaram, S. Rathnamala, R. Sivaramakrishnan, Proc. Indian Acad. Sci., Sect. A, 86,435 (1977). 30. W. Strohmeier, N. Iglauer, Z. Physi. Chem. (Frankf), 51, 50 (1966). 31. M. Murakami, J. Kang, Bull. Chem. SOC.Jpn., 36, 763 (1963). 32. A. A. Vlcek, J. Hanzlik, Inorg. Chem., 6,2053 (1967). 33. J. Hanzlik, A. A. Vlcek, Inorg. Chem., 8,669 (1969). 34. M. Murakami, R. Kawai, K. Suzuki, J. Chem. Sac. Jpn., 84,669 (1963).

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.3. Cobalt Carbonyl Catalysts

92

35. 36. 37. 38.

K. Isogai, Y . Hazeyama, J. Chem. SOC.Jpn., 86, 869 (1965). T. Funabiki, M. Mohri, K. Tamara, J. Chem. SOC.,Dalton Trans., 1813 (1973). Ref. 2, Ch. X, Sect. B. T. Funabiki, S. Kasaoka, M. Matsumoto, K. Tarama, J. Chem. Soc., Dalton Trans., 2043

(1974). 39. D. L. Reger, A. Gabrielli,J. Mol. Cufal., 12, 173 (1981).

14.3.3.3. Cobalt Carbonyl Catalysts

(i) Hydrogenation by CoH(CO),. Co,(CO),, catalyzes hydrofonnylation (0x0) reactions:

RCH=CH

+ CO + H,

c02(co)12

> RCH,CH,CHO

100-160°, 100-300 atm

+ RCH(CHO)CH,

The important mechanistic features of the 0x0 reaction are' Co,(CO), + H, 2 CoH(CO), CoH(CO), CoH(CO),

+ RCH=CH,

RCH,CH,Co(CO),

+ CO

RCH,CH,Co(CO), RCH,CH,COCO(CO)~ + H,

F=

CoH(CO),

--

(c)

[CoH(CO),(RCH=CHz)]

RCH,CH,Co(CO),

(4

RCH,CH,Co(CO),

(el

RCH,CH,COCo(CO),

(f)

F=

+ CO

RCH,CH,CHO

+ CoH(CO),

(g)

Hydrogenation of the olefin reactant and the aldehyde products are important competing side reactions during hydroformylation. Orchin and co-workers first demonstrated the presence of CoH(CO), under catalytic hydroformylation conditions,. CoH(CO), in the absence of H, and CO reacts stoichiometrically with olefins at RT to give products that are also observed under catalytic condition^^^^. Stoichiometric reduction of olefins, aldehydes, and ketones by CoH(CO), is also ~ b s e r v e d ~ Hydrogenation -~. of the olefinic substrate only becomes significant if the olefin is branched, conjugated, or bears electronegative substituents. Thus, simple olefins such as propylene and cyclohexene give C 10% saturated hydrocarbons, but branched olefins such as isobutylene and diisobutylene give as much as 65% saturated hydrocarbon product during hydrofonnylation'03' The use of Co,(CO), as a hydrogenation catalyst in the absence of CO generally results in decomposition to metal. Phosphine derivatives of cobalt carbonyl complexes, which are more robust hydrogenation catalysts, will be discussed next. (ii) Hydrogenation by CoH(CO),( PR,),,and Related Species. Cobalt carbonyl complexes bearing tertiary phosphines, arsines, or phosphite ligands have been investigated as hydrogenation catalysts. Hydrogenation of polyene compounds to rnonoenes using COH(CO),(PR,),_~ catalysts, have been studied e.g."-14

'.

1,5,9-~yclododecatriene

[CO(CO)~PBU~I~

20-30 atm H,, 110-180C

cis- and trans-cyclododecene (ca. 1:2)

(h)



92

35. 36. 37. 38.

K. Isogai, Y . Hazeyama, J. Chem. SOC.Jpn., 86, 869 (1965). T. Funabiki, M. Mohri, K. Tamara, J. Chem. SOC.,Dalton Trans., 1813 (1973). Ref. 2, Ch. X, Sect. B. T. Funabiki, S. Kasaoka, M. Matsumoto, K. Tarama, J. Chem. Soc., Dalton Trans., 2043

(1974). 39. D. L. Reger, A. Gabrielli,J. Mol. Cufal., 12, 173 (1981).

14.3.3.3. Cobalt Carbonyl Catalysts

(i) Hydrogenation by CoH(CO),. Co,(CO),, catalyzes hydrofonnylation (0x0) reactions:

RCH=CH

+ CO + H,

c02(co)12

> RCH,CH,CHO

100-160°, 100-300 atm

+ RCH(CHO)CH,

The important mechanistic features of the 0x0 reaction are' Co,(CO), + H, 2 CoH(CO), CoH(CO), CoH(CO),

+ RCH=CH,

RCH,CH,Co(CO),

+ CO

RCH,CH,Co(CO), RCH,CH,COCO(CO)~ + H,

F=

CoH(CO),

--

(c)

[CoH(CO),(RCH=CHz)]

RCH,CH,Co(CO),

(4

RCH,CH,Co(CO),

(el

RCH,CH,COCo(CO),

(f)

F=

+ CO

RCH,CH,CHO

+ CoH(CO),

(g)

Hydrogenation of the olefin reactant and the aldehyde products are important competing side reactions during hydroformylation. Orchin and co-workers first demonstrated the presence of CoH(CO), under catalytic hydroformylation conditions,. CoH(CO), in the absence of H, and CO reacts stoichiometrically with olefins at RT to give products that are also observed under catalytic condition^^^^. Stoichiometric reduction of olefins, aldehydes, and ketones by CoH(CO), is also ~ b s e r v e d ~ Hydrogenation -~. of the olefinic substrate only becomes significant if the olefin is branched, conjugated, or bears electronegative substituents. Thus, simple olefins such as propylene and cyclohexene give C 10% saturated hydrocarbons, but branched olefins such as isobutylene and diisobutylene give as much as 65% saturated hydrocarbon product during hydrofonnylation'03' The use of Co,(CO), as a hydrogenation catalyst in the absence of CO generally results in decomposition to metal. Phosphine derivatives of cobalt carbonyl complexes, which are more robust hydrogenation catalysts, will be discussed next. (ii) Hydrogenation by CoH(CO),( PR,),,and Related Species. Cobalt carbonyl complexes bearing tertiary phosphines, arsines, or phosphite ligands have been investigated as hydrogenation catalysts. Hydrogenation of polyene compounds to rnonoenes using COH(CO),(PR,),_~ catalysts, have been studied e.g."-14

'.

1,5,9-~yclododecatriene

[CO(CO)~PBU~I~

20-30 atm H,, 110-180C

cis- and trans-cyclododecene (ca. 1:2)

(h)


93

The catalyst can be recovered. The PPh, derivative of [Co(CO),(PR,)], under similar conditions decomposes [like Co,(CO),] to give metal and incomplete, unselective hydrogenation. The stability of Co(CO),(PR,) hydrogenation catalysts has been linked with the relative donor ability of the phosphine ligand15.As for the homoleptic cobalt carbonyl catalyst, the corresponding hydride complex CoH(CO),(PR,) is believed to be the actual catalyst. CoH(CO),(PBu,) reacts with 1,3-~yclooctadieneat 0°C to give Co(R)(CO),(PBu,), which at higher T gives cyclooctene presumably via reaction (i) or (j): Co(R)(CO),(PBu,)

+ CoH(CO),(PBu,)

Co(R)(CO),(PBu,)

+ H,

--

[Co(CO),(PBu,)],

CoH(CO),(PBu,)

+ RH + RH

(i) (j)

Cyclooctene also reacts with CoH(CO),(PBu,) to give a cobalt-alkyl complex; however, the latter complex decomposes upon heating to give back the reactants rather than a hydrogenated product. In many studies, CO,(CO), is reacted in situ with phosphine ligands to generate catalytic species16. In contrast to the rhodium analogue, CoH(CO)(PPh,), is not a very effective hydrogenation catalysts. The rate of catalysis can be improved by addition of AlEt,, which may serve to remove a phosphine ligand to give an active species17. Some cobalt-phosphine complexes that bear no carbonyl ligands are good hydrogenation catalyst. The complex [CoX(PPh,),] (X = C1, Br, I) in the presence of Lewis acids exhibits high selectivity in the hydrogenation of dienes. Internal double bonds are reduced preferentially. The Lewis acid presumably abstracts the halide ligand thereby creating a coordinatively unsaturated metal center and a coordination site for the substrate, e.g.', [CoBr(PPh,),]

+ diene + BF,.OEt,

-+

[Co(diene)(PPh,),]+ [BF,Br]-

+ Et,O

(k)

The molecular N, complex CoH(N,)(PPh,), reacts with sodium naphthalide to give a paramagnetic species which hydrogenates styrene under mild conditions''. (M. T. ASHBY) 1. 2. 3. 4.

5.

6. 7. 8.

R. F. Heck, D. S . Breslow, J . Am. Chem. SOC., 83,4023 (1961). M. Orchin, L. Kirch, I. Goldfarb, J . Am. Chem. SOC., 78, 5450 (1956). I. Wender, H. W. Sternberg, M. Orchin, J . Am. Chem. SOC., 75, 3041 (1953). P. Pino, R. Ercoli, F. Calderazzo, Chim. lnd. (Milan), 37, 782 (1955). R. W. Goetz, M. Orchin, J. Org. Chem., 27, 3698 (1962). R. W. Goetz, M. Orchin, J . Am. Chem. SOC.,85,2782 (1963). R. F. Heck, J. Am. Chem. SOC.,85, 1460 (1963). Y. Takelzami. C. Yokokawa. Y. Watanabe, H. Masada, Y. Okuda, Bull. Chem. SOC.J m , 37, 1190 (1k4).

G. Gut, M. H. El-Markhzangi, A. Guyer, Helv. Chim. Acta, 48, 1151 (1965). L. Marko, Chem. Ind. (London), 260, (1962). M. Freund, L. Marko, J. Laki, Acta Chim. Acad. Sci. Hung., 31, 77 (1962). A. Misono, I. Ogata, Bull. Chem. SOC.Jpn., 40, 2718 (1967). I. Ogata, A. Misono, Discuss. Faraday SOC., 46, 72 (1968). I. Ogata, J . Chem. SOC.Jpn., 72, 1710 (1969). B. R. James, Homogeneous Hydrogenation, John Wiley, New York, 1973, p. 183, and references therein. 16. For example, L. Marko, P. Szabo, J. Laky, Hung. Pat. 157,605; Chem. Abstr., 73, 124,016

9. 10. 11. 12. 13. 14. 15.

(1970).

94

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.4. Chrorniurn(0)Carbonyl Catalysts

17. M. Hidai, T. Kuse, T. Hikita, Y. Uchida, A. Misono, Tetrahedron Lett., 1715 (1970). 18. K. Kawakami, T. Mizoroki, A. Ozaki, J . Mol. Catal., 5, 175 (1979). 19. S.Tyrlik, J . Organomet. Chem., 50, C46 (1973).

14.3.3.4. Chromlum(0) Carbonyl Catalysts

Group 6 metal carbonyl complexes, particularly Cr(0) compounds, are used as homogeneous hydrogenation catalysts'-5. The most frequently used classes of zero-valent catalyst precursors are M(CO), (M = Cr, Mo, W), (diene)Cr(CO),, (arene)Cr(CO),, and Cr(CO),(NCMe),. These are of synthetic utility because they generally stereoselectively 1,Chydrogenate conjugated dienes to give cis monoenes exclusively, equation (a).

M(CO), (M = Cr, Mo, W) are employed as thermal hydrogenation catalysts; however, relatively severe conditions are required [e.g., -200°C and 50 atm H, for the catalytic 1,Chydrogenation of a conjugated diene by Cr(CO)6J6-10.In contrast, M(CO), (M = Cr, Mo, W) catalyze the light-induced 1,4-hydrogenation of various conjugated dienes under ambient conditions. Although the first photoinduced reactions were continuously irradiated", it has been shown that irradiation generates a thermally active catalyst". It was also concluded on the bases of the relative reactivities of 2-methyl-1,3butadiene, trans- 1,3-pentadiene, and cis- 1,3-pentadiene that it is the s-cis conformation of 1,3-dienes that undergo hydrogenation. Of these three dienes, only cis- 1,3-pentadiene cannot readily achieve the s-cis conformation. It undergoes hydrogenation at a rate less than one-tenth of the other two dienes. (trans,trans-2,4-Hexadiene)Cr(CO), is an intermediate in the photocatalytic hydrogenation of trans,trans-2,4-hexadienewith Cr(C0),13. It was found after an extensive mechanistic investigation of the photoinduced hydrogenation of (norbomadiene)Cr(CO), that the principal photochemical reaction involves rupture of a metal-olefin bond and subsequent reaction of the coordinatively unsaturated intermediate with H214. Studies of the photochemical substitution reactions of (norbomadiene)M(CO),(M = Cr, Mo, W) complexes show that an axial CO ligand is preferentially lost upon photolysis of (n~rbomadiene)M(CO),'~. The effect of M on the product distribution obtained when norbomadiene is hydrogenated by (norbomadiene)M(CO), and M(CO), (M = Cr, Mo, W) has also been studied. Norbomadiene differs from most other diene substrates in that its reduction is not selective and three hydrogenation products are formed: norbomene, nortricyclene, and norbomane. The ratios of nortricyc1ene:norbomene:norbomane produced when (norbomadiene)Cr(CO), undergoes photoinduced hydrogenation at ambient T and P are about 2.8:1:0'4. A similar result is obtained when norbomadiene is hydrogenated with Cr(CO), (1,9:1:0)15.However, different results are obtained when M(CO), (M = Mo, W) and (diene)M(CO), (M = Mo, W) are used as catalyst precursors; only norbomene is obtained when M(CO), (M = Mo, W) is used and substantial amounts of nortricyclene are formed together with norbomene when (diene)M(CO), (M = Mo, W) is used. This result has been rationalized in terms of parallel mechanisms that involve either loss of an axial CO ligand or metalolefin bond cleavage, Eq. (b). Beginning with M(CO),, if on photochemical loss of a CO ligand a monoolefinic species (norbomadiene)M(CO), is formed that absorbs H, upon subsequent loss of a second CO ligand, the mechanism would predict formation of norbomene. Whereas beginning with (norbomadiene)M(CO),, the product distribution will depend on the relative importance of the CO-dissociation versus the olefin-dissocia-


94

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.4. Chrorniurn(0)Carbonyl Catalysts

17. M. Hidai, T. Kuse, T. Hikita, Y. Uchida, A. Misono, Tetrahedron Lett., 1715 (1970). 18. K. Kawakami, T. Mizoroki, A. Ozaki, J . Mol. Catal., 5, 175 (1979). 19. S.Tyrlik, J . Organomet. Chem., 50, C46 (1973).

14.3.3.4. Chromlum(0) Carbonyl Catalysts

Group 6 metal carbonyl complexes, particularly Cr(0) compounds, are used as homogeneous hydrogenation catalysts'-5. The most frequently used classes of zero-valent catalyst precursors are M(CO), (M = Cr, Mo, W), (diene)Cr(CO),, (arene)Cr(CO),, and Cr(CO),(NCMe),. These are of synthetic utility because they generally stereoselectively 1,Chydrogenate conjugated dienes to give cis monoenes exclusively, equation (a).

M(CO), (M = Cr, Mo, W) are employed as thermal hydrogenation catalysts; however, relatively severe conditions are required [e.g., -200°C and 50 atm H, for the catalytic 1,Chydrogenation of a conjugated diene by Cr(CO)6J6-10.In contrast, M(CO), (M = Cr, Mo, W) catalyze the light-induced 1,4-hydrogenation of various conjugated dienes under ambient conditions. Although the first photoinduced reactions were continuously irradiated", it has been shown that irradiation generates a thermally active catalyst". It was also concluded on the bases of the relative reactivities of 2-methyl-1,3butadiene, trans- 1,3-pentadiene, and cis- 1,3-pentadiene that it is the s-cis conformation of 1,3-dienes that undergo hydrogenation. Of these three dienes, only cis- 1,3-pentadiene cannot readily achieve the s-cis conformation. It undergoes hydrogenation at a rate less than one-tenth of the other two dienes. (trans,trans-2,4-Hexadiene)Cr(CO), is an intermediate in the photocatalytic hydrogenation of trans,trans-2,4-hexadienewith Cr(C0),13. It was found after an extensive mechanistic investigation of the photoinduced hydrogenation of (norbomadiene)Cr(CO), that the principal photochemical reaction involves rupture of a metal-olefin bond and subsequent reaction of the coordinatively unsaturated intermediate with H214. Studies of the photochemical substitution reactions of (norbomadiene)M(CO),(M = Cr, Mo, W) complexes show that an axial CO ligand is preferentially lost upon photolysis of (n~rbomadiene)M(CO),'~. The effect of M on the product distribution obtained when norbomadiene is hydrogenated by (norbomadiene)M(CO), and M(CO), (M = Cr, Mo, W) has also been studied. Norbomadiene differs from most other diene substrates in that its reduction is not selective and three hydrogenation products are formed: norbomene, nortricyclene, and norbomane. The ratios of nortricyc1ene:norbomene:norbomane produced when (norbomadiene)Cr(CO), undergoes photoinduced hydrogenation at ambient T and P are about 2.8:1:0'4. A similar result is obtained when norbomadiene is hydrogenated with Cr(CO), (1,9:1:0)15.However, different results are obtained when M(CO), (M = Mo, W) and (diene)M(CO), (M = Mo, W) are used as catalyst precursors; only norbomene is obtained when M(CO), (M = Mo, W) is used and substantial amounts of nortricyclene are formed together with norbomene when (diene)M(CO), (M = Mo, W) is used. This result has been rationalized in terms of parallel mechanisms that involve either loss of an axial CO ligand or metalolefin bond cleavage, Eq. (b). Beginning with M(CO),, if on photochemical loss of a CO ligand a monoolefinic species (norbomadiene)M(CO), is formed that absorbs H, upon subsequent loss of a second CO ligand, the mechanism would predict formation of norbomene. Whereas beginning with (norbomadiene)M(CO),, the product distribution will depend on the relative importance of the CO-dissociation versus the olefin-dissocia-

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.4. Chrorniurn(0) Carbonyl Catalysts

95

tion pathways. As further support of the latter mechanism, it was noted that (norbornadiene)W(CO), is observed spectroscopically on irradiating W(CO), in the presence of norbomadiene. The ratio of norbomene to nortricyclene increases with increasing P and above 40 atm norbomene is hydrogenated to norbomaneI6. Interestingly, the production of norbomadiene is first-order with respect to H, at low P , but the production of nortricyclene is independent of H2 P . The latter result is due to an increase in H, concentration and not a large negative activation volume16.It has been suggested that the results obtained at high H2 P are inconsistent with this rnechanisml6; however, the results are consistent if the first step of the CO-dissociation mechanism and the second step of the olefin-dissociation mechanism are rate-determining, respectively. Recent infrared spectroscopic studies on the mechanism of photocatalytic hydrogenation of norbomadiene by Group 6 metal carbonyls has implicated a nonclassical dihydrogen complex (q4-norbomadiene) M(CO), (q2-H,) and the species fuc- and mer-( $-norbomadiene) (?~~-norbomadiene)M(CO), ”.

M(CO), has been used to stereoselectively hydrogenate methyl sorbate and truns,trunsconjugated fatty e!ters at ambient T and P by exposing the reaction mixture to UV irradiation (3500 A)I8. Under the same conditions, (heptatriene)Cr(CO), showed lower activity than Cr(CO), and Mo(CO), and (mesitylene)Mo(CO), showed no significant activity toward conjugated substrates. Thermal hydrogenation reactions, catalyzed by (arene)Cr(CO), complexes, have been studied. Conjugated dienes are regioselectively 1,Chydrogenated using (arene)Cr(CO), with high cis stereo~pecificity~~~. Derivatives of (arene)Cr(CO), complexes with relatively electron-donating substituents on the arene ligand thermally catalyze hydrogenation reactions at high P of H, (ca. 50 atm) and high T (150-175°C)19. In contrast, electron-withdrawing substituents on the arene ligand accelerate catalytic hydrogenation reactions”. This has been attributed to a weakening of the metal-arene bond, which is further evidenced by the deformation of the arene ligand observed in the solid-state structures of derivatives that bear electron-withdrawing heteroatom substituents”. The catalytic reaction is also facilitated by polar, coordinating solvents. The observed substituent effect and the induction period that is usually observed when (arene)Cr(CO),

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.4. Chromium(0) Carbonyl Catalysts

96

complexes are used as catalyst precursors suggest that L,Cr(CO), is the true catalytic species in a coordinating solvent (L),,. A common (diene)Cr(CO),(H), intermediate has been proposed for the photoinduced Cr(CO), and the thermal (arene)Cr(CO), catalyzed hydrogenation of conjugated dienes,,. Because Cr(CO), on photodissociation of CO hydrogenates conjugated dienes and (arene)Cr(CO), catalyzes the same reaction at elevated temperature after the apparent loss of the arene ligand, researchers have investigated the catalytic activity of Cr(CO),(NCMe),, which bears labile nitrile ligands,,. Cr(CO),(NCMe), catalyzes the 1,Chydrogenation of 1,3-dienes such as 2-methyl- 1,3-butadiene, truns-l,3-pentadiene, and truns,rruns-2,4-hexadieneat low T (40°C) and low H, P (ca. 1 atm). Reactions can be carried out under neat conditions with turnover numbers in excess of 3000. Cr(CO),(NCMe), is still the most effective group 6 metal carbonyl catalyst for 1,4hydrogenation of conjugated dienes. Chromium (0) hexacarbonyl catalyzes reduction of cyclopentadienones in aqueous organic solvents in the absence of H,, equation When D,O is substituted for H,O, the corresponding 5,6-dideutero compounds are obtained. With reactions performed in D,O in the presence of H,, only deuterium is incorporated into the products. Cr(CO), catalyzes the water gas shift reaction [equation (d)], thereby providing a source of H,. However, since only D is incorporated into the products when the reactions (c) are performed in D,O in the presence of H,, it is unlikely that H, is involved in the reaction. Furthermore, little CO, is detected in the reaction. 0

0

R,=R2=SiMe,, R,=H, R4=OMe; 14% R,=R,=Ph, R,=H, %=OMe; 61% R,=R,=R,=R,=Ph; 92% CO

+ H,O

= H,

+ CO,

(d) It is noteworthy that M(CO), (M = Cr, Mo, W) catalyze the water gas shift reaction in the presence of a large excess of sulfides,,. Since S-containing impurities are expected to be present in many synthesis gas feedstocks, practical catalytic systems must exhibit tolerance to S compounds. (M. T. ASHBY) 1. A. Rejoan, M. Cais, in Progress in Coordination Chemistry, M. Cais, ed., Elsevier, Amsterdam, 1968, p. 32. 2. B. R. James, Homogeneous Hydrogenation, Ch. VII, Sect. A, John Wiley, New York, 1973. 3. M. Cais, D. Fraenkel, K. Weidenbraum, Coord. Chem. Rev.,16, 27 (1975). 4. M. Cais, D. Fraenkel, Ann. N . Y . Acad. Sci, 333, 23 (1980). 5. B. R. James, in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G . A. Stone, E. Abel, eds., Vol. 8, Ch. 51, Pergamon Press, Oxford, 1982. 6. K. Horino, Japanese Patent 03 34,952; Chem Abstr., 115, 8084u (1991). 7. A. A. Vasil’ev, G . V. Cherkaev, L. I. Soldatenko, G . G . Kolomeer, Zh. Org. Khim., 27, 317 ( 1991 ) .

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.5. Ziegler Catalysts

97

8. A. A. Vasil'ev, G. V. Cherkaev, M. A. Nikitina, Khim.-Farm. Zh., 25,57 (1991). 9. A. A. Vasil'ev, S. S. Poddubnaya, G. V. Cherkaev, N. A. Novikov, V. G. Cherkaev, Zh. Vses. Khim., O-va. im. D. I . Mendeleeva, 32, 117 (1987). 10. A. Furuhata, K. Onishi, A. Fujita, K. Kogami, Agric. Biol. Chem., 46, 1757 (1982). 11. J. Nasielski, P. Kirsch, L. Wilputte-Steinert, J . Organomet. Chem., 27, C13 (1971). 12. M. Wrighton, M. A. Schroeder, J. Am. Chem. SOC., 95,5764 (1973). 13. I. Fischler, M. Budzwait, E. A. Koerner von Gustorf, J . Organornet. Chem., 105, 325 (1976). 14. G. Platbrood, L. Wilputte-Steinert,J . Organomet. Chem., 70, 393 (1974). 15. D. J. Darensbourg, H. H. Nelson, 111, M. A. Murphy, J . Am. Chem. SOC.,99, 896 (1977). 16. M. J. Mirbach, D. Steinmetz, A. Saw, J . Organomet. Chem., 168, C13 (1979). 17. F.-W. Grevels, J. Jacke, W. E. Klotzbucher, K. Schaffner, R. H. Hooker, A. J. Rest, J . Organornet. Chem., 382,201 (1990); A. Jackson, P. M. Hodges, M. Poliakoff, J. J. Tumer, F.-W. Grevels, J. Am. Chem. Soc., 112 (1990); P. M. Hodges, S. A. Jackson, J. Jacke, M. Poliakoff, J. J. Turner, F.-W. Grevels, J . Am. Chem. Soc., 112, 1234 (1990). 62, 1044 (1985). 18. J. A. Heldal, E. N. Frankel, J . Am. Oil Chem. SOC., 19. M. Cais, E. N. Frankel, A. Rejoan, Tetrahedron Lett., 1919 (1968). 20. E. N. Frankel, R. 0. Butterfield, J . Org. Chem., 34, 3930 (1969). 21. P. Le Maux, J. Y. Saillard, D. Grandjean, G. Jaouen, J . Org. Chern., 45,4524 (1980). 22. Y. Eden, 0. Fraenkel, M. Cais, E. A. Halevi, Isr. J . Chem., 15, 223 (1976). 23. M. A. Schroeder, M. S. Wrighton, J. Organomet. Chem., 74, C29 (1974). 24. J. W. Herndon, S. U. Turner, W. F. K. Schnatter, J . Am. Chem. SOC.,110, 3334 (1988). 25. J. W. Hemdon, S. U. Tumer, Tetrahedron Lett, 30,295 (1989). Chem. Commun., 529 (1980). 26. A. D. King, R. B. King, D. B. Yang, J . Chem. SOC.,

s.

14.3.3.5. Ziegler Catalysts

Transition metal halides react with alkyls of aluminum, alkaline earth, and alkali metals to give Ziegler catalysts that dimerize and polymerize olefins. Ziegler catalysts are generally heterogeneous, but they can be made homogeneous by modifying their ligands. Olefin hydrogenation in the presence of homogeneous Ziegler-Natta catalysts has been known since the early sixties.'-5. The catalytic activity of acetylacetonates of Co(III), Cr(II), Fe(III), Mn(I1 and 111), Mo(VI), and Ru(II1); alkoxides of Ti(1V) and V(V); and the dichloride complexes Cp2TiC12,Cp2ZrC12,CoCl2(PPh3),, NiC12(P"Bu3)2, and PdCl,(P"Bu,), has been explored. The latter transition metal-halides and pseudohalides, when combined with an alkyllithium (EtLi or "BuLi) or alkylaluminum ('Bu2A1H, Et3Al, 'Bu,Al) co-catalyst, successfully hydrogenate under mild conditions (30-50°C, 3.7 atm H2) one or more of the olefins: cyclohexene, I-octene, 2-methyl-2butene, 2-pentene, tetramethylethylene, and stilbene. The alkylaluminum co-catalysts are generally more effective than the alkyllithium cocatalysts. Kinetic data have been reported for cyclohexene reduction with a 1:6 Cr(acac),-'Bu,Al catalyst in heptane at 30°C, which showed a first-order dependence on catalyst and H,. Hydrogenation rates generally decrease with increasing substitution of the alkene substrate'. Similar kinetic results were independently obtained for the Cr(acac),-'Bu,Al catalyst6. A proposed mechanism involves alkylation of the metalhalide [equation (a)], hydride formation [equation (b)], followed by reversible insertion of the olefin substrate into the metal-hydride bond [equation (c)], and hydrogenoiysis of the resulting metal-alkyl bond [equation (d)]'.


14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.5. Ziegler Catalysts

97

8. A. A. Vasil'ev, G. V. Cherkaev, M. A. Nikitina, Khim.-Farm. Zh., 25,57 (1991). 9. A. A. Vasil'ev, S. S. Poddubnaya, G. V. Cherkaev, N. A. Novikov, V. G. Cherkaev, Zh. Vses. Khim., O-va. im. D. I . Mendeleeva, 32, 117 (1987). 10. A. Furuhata, K. Onishi, A. Fujita, K. Kogami, Agric. Biol. Chem., 46, 1757 (1982). 11. J. Nasielski, P. Kirsch, L. Wilputte-Steinert, J . Organomet. Chem., 27, C13 (1971). 12. M. Wrighton, M. A. Schroeder, J. Am. Chem. SOC., 95,5764 (1973). 13. I. Fischler, M. Budzwait, E. A. Koerner von Gustorf, J . Organornet. Chem., 105, 325 (1976). 14. G. Platbrood, L. Wilputte-Steinert,J . Organomet. Chem., 70, 393 (1974). 15. D. J. Darensbourg, H. H. Nelson, 111, M. A. Murphy, J . Am. Chem. SOC.,99, 896 (1977). 16. M. J. Mirbach, D. Steinmetz, A. Saw, J . Organomet. Chem., 168, C13 (1979). 17. F.-W. Grevels, J. Jacke, W. E. Klotzbucher, K. Schaffner, R. H. Hooker, A. J. Rest, J . Organornet. Chem., 382,201 (1990); A. Jackson, P. M. Hodges, M. Poliakoff, J. J. Tumer, F.-W. Grevels, J. Am. Chem. Soc., 112 (1990); P. M. Hodges, S. A. Jackson, J. Jacke, M. Poliakoff, J. J. Turner, F.-W. Grevels, J . Am. Chem. Soc., 112, 1234 (1990). 62, 1044 (1985). 18. J. A. Heldal, E. N. Frankel, J . Am. Oil Chem. SOC., 19. M. Cais, E. N. Frankel, A. Rejoan, Tetrahedron Lett., 1919 (1968). 20. E. N. Frankel, R. 0. Butterfield, J . Org. Chem., 34, 3930 (1969). 21. P. Le Maux, J. Y. Saillard, D. Grandjean, G. Jaouen, J . Org. Chern., 45,4524 (1980). 22. Y. Eden, 0. Fraenkel, M. Cais, E. A. Halevi, Isr. J . Chem., 15, 223 (1976). 23. M. A. Schroeder, M. S. Wrighton, J. Organomet. Chem., 74, C29 (1974). 24. J. W. Herndon, S. U. Turner, W. F. K. Schnatter, J . Am. Chem. SOC.,110, 3334 (1988). 25. J. W. Hemdon, S. U. Tumer, Tetrahedron Lett, 30,295 (1989). Chem. Commun., 529 (1980). 26. A. D. King, R. B. King, D. B. Yang, J . Chem. SOC.,

s.

14.3.3.5. Ziegler Catalysts

Transition metal halides react with alkyls of aluminum, alkaline earth, and alkali metals to give Ziegler catalysts that dimerize and polymerize olefins. Ziegler catalysts are generally heterogeneous, but they can be made homogeneous by modifying their ligands. Olefin hydrogenation in the presence of homogeneous Ziegler-Natta catalysts has been known since the early sixties.'-5. The catalytic activity of acetylacetonates of Co(III), Cr(II), Fe(III), Mn(I1 and 111), Mo(VI), and Ru(II1); alkoxides of Ti(1V) and V(V); and the dichloride complexes Cp2TiC12,Cp2ZrC12,CoCl2(PPh3),, NiC12(P"Bu3)2, and PdCl,(P"Bu,), has been explored. The latter transition metal-halides and pseudohalides, when combined with an alkyllithium (EtLi or "BuLi) or alkylaluminum ('Bu2A1H, Et3Al, 'Bu,Al) co-catalyst, successfully hydrogenate under mild conditions (30-50°C, 3.7 atm H2) one or more of the olefins: cyclohexene, I-octene, 2-methyl-2butene, 2-pentene, tetramethylethylene, and stilbene. The alkylaluminum co-catalysts are generally more effective than the alkyllithium cocatalysts. Kinetic data have been reported for cyclohexene reduction with a 1:6 Cr(acac),-'Bu,Al catalyst in heptane at 30°C, which showed a first-order dependence on catalyst and H,. Hydrogenation rates generally decrease with increasing substitution of the alkene substrate'. Similar kinetic results were independently obtained for the Cr(acac),-'Bu,Al catalyst6. A proposed mechanism involves alkylation of the metalhalide [equation (a)], hydride formation [equation (b)], followed by reversible insertion of the olefin substrate into the metal-hydride bond [equation (c)], and hydrogenoiysis of the resulting metal-alkyl bond [equation (d)]'.

98

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.5.Ziegler Catalysts

-

+ MX, +RzAlX + RMX, RH + HMX,-, RMX,- + H, HMX,- + olefin 6 alkyl-MX,alkyl-MX, - + H, 6 HMX, - I + product R,Al

-1

(a) (b) (c) (4

An alternative to step (b) was suggested that involves P-hydride elimination [equation (e>l.

+

--+ HMX,RCH=CH, (el RCH,CH,MX,An alternative to the hydrogenolysis step [equation (d)] was also suggested in which the alkyl intermediate reacts with another metal-hydride to yield the product [equation (f)] followed by reaction of the resulting metal species with H, to regenerate the hydride catalyst [equation (g)].

-

+ HMX,product + (MX,-,), (MX,-1)2 + H, +2 HMX,-,

alkyl-MX,-,

(f)

(8)

A chiral Ziegler catalyst that effects enantioselective hydrogenation of olefins has been reported7. The catalyst is derived from ( - )-[ethylene(4,5,6,7-tetrahydro-l(R)indenyl)]zirconium compounds (1) and aluminoxane.

1 Chiral group 4 metallocenes have been developed'. a-Olefins are readily polymerized in the presence of catalyst derived from zirconocenes and aluminoxanes. However, terminal olefins substituted in the 2- or 3-positions and internal olefins are not polymerized but undergo hydrogenation instead. Thus, styrene is hydrogenated at the rate of 12 turnovers/ min at 20 atm of H, at 25°C with the l/aluminoxane catalytic system. When D, is employed in the latter reduction, ( - )-(R)-l,2-dideuterioethylbenzeneis obtained in 93% yield with an optical purity of 65% ee (ee = enantiomeric excess). Interestingly, active hydrogenation catalysts are obtained when methyl and binaphtholate derivatives of 1are used as catalyst precursors, but not when the chloride derivative is used. Other olefins that are hydrogenated by the system include 2-methyl- 1-pentene, 2-phenyl- 1-butene, and cis- and truns-2-hexene. 2-Phenyl- 1-butene was hydrogenated with H, to give ( -)-(I?)2-phenyl-l-butane in 95% chemical yield and 36% ee. Although the nature of the catalytically active species in the l/aluminoxane system is unknown, model studies9-" suggest that an ion pair species of the type

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.6. Ruthenium(l1) Catalysts

99

[Cp,M-R] [X(Al(CH,)-0),] - is involved. Direct involvement of the aluminoxane is implicated since hydrogenation7 and polymerization" reactions that use aluminoxane/ zirconocene catalysts are influenced by the nature of X. The latter results contrast with solution XPS studies that suggest similar species are formed from the dialkyl and dichloride derivatives of metallocenes in the presence of al~minoxane'~. Interestingly, the enantioface of 1-pentene, styrene, and 2-phenyl- 1-butene that is hydrogenated'~'~ by the l/aluminoxane catalyst is opposite of that which is p~lymerized'~-'~for 1-pentene and propene. +

(M. T. ASHBY) 1. M. F. Sloan, A. S. Matlock, D. S . Breslow, J . Am. Chem. SOC.,85,4014 (1963). 2. S . J. Lapporte, Ann. N. Y . Acad. Sci., 158, 510 (1969). 3. J. Boor, Jr., Ziegler-Natta Catalysts and Polymerization, Academic Press, New York, 1979, p. 607. 4. B. R. James, Homogeneous Hydrogenation, John Wiley, New York, 1973, Ch. XV and ref-

erences therein.

5. B. R. James, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. Abel, eds., Vol. 8, Ch. 51, Pergamon Press, Oxford, 1982, p. 340. 6. W. R. Kroll, J. Catal., 15, 281 (1969). 7. R. Waymouth, P. Pino, J . Am. Chem. SOC., 112, 4911 (1990). 8. F. R. W. P. Wild, L. Zsolnai, G. Huttner, H. H. Brintzinger, J . Organomet. Chem., 288, 63 (1985). 9. R. F. Jordan, J . Chem. Educ., 65, 285 (1988), and references therein. 10. G. C. Hlatky, H. W. Turner, R. R. Eckman, J. Am. Chem. SOC., 111,2728 (1989). 11. J. J. Eisch, A. M. Pitrowski, S . K. Brownstein, E. G. Gabe, F. L. Lee, J. Am. Chem. SOC., 107, 7219 (1985). 12. E. Giannetti, G. M. Nicoletti, R. J. Mazzocchi, Polym. Sci. Polym. Chem.,23, 2117 (1985). 13. P. G. Gassman, M. R. Callstrom, J . Am. Chem. SOC.,109, 7875 (1987). 14. P. Pino, M. Galimberti, J . Organomet. Chem, 370, 1 (1989). 15. P. Pino, M. Galimberti, J. Wei, N. Piccolrovazzi, in Transition Metals and Organometallics

as Catalysts for Olejin Polymerization, W. Kaminsky, H. Sinn, eds., Springer-Verlag, New York, 1980, p. 269. 16. P. Pino, P. Cioni, J. Wei, J . Am. Chem. SOC.,109, 6189 (1987).

14.3.3.6. Ruthenlum(l1) Catalysts

A review of homogeneous hydrogenation by Ru catalysts was published in 1970' and the synthesis and properties of ruthenium-hydride complexes known prior to 1977 have been reviewed'. An up to date review of ruthenium-hydride complexes appeared in 19843. We concern ourselves here with hydrogenation reactions that involve Ru(I1) species as catalyst precursors or catalytic intermediates. The first example of activation of molecular H, by a Ru(I1) species was described in 19614.The study employed a solution of Ru(I1) choride in aqueous HC1 at 65 to 95°C and H, P of about 1 atm to catalytically hydrogenate maleic, fumaric, and acrylic acid. The Ru(I1) species that are prepared in situ by reduction of solutions of Ru(II1) with Ti(II1) is thought to be RuCli-. The latter blue solutions turn yellow in the presence of the olefin substrates, and they absorb H, according to a rate law that is first-order each in Ru(I1) and H,. The yellow intermediate is shown spectrophotometrically to be a Ruolefin complex. The following mechanism was proposed4:



99

[Cp,M-R] [X(Al(CH,)-0),] - is involved. Direct involvement of the aluminoxane is implicated since hydrogenation7 and polymerization" reactions that use aluminoxane/ zirconocene catalysts are influenced by the nature of X. The latter results contrast with solution XPS studies that suggest similar species are formed from the dialkyl and dichloride derivatives of metallocenes in the presence of al~minoxane'~. Interestingly, the enantioface of 1-pentene, styrene, and 2-phenyl- 1-butene that is hydrogenated'~'~ by the l/aluminoxane catalyst is opposite of that which is p~lymerized'~-'~for 1-pentene and propene. +

(M. T. ASHBY) 1. M. F. Sloan, A. S. Matlock, D. S . Breslow, J . Am. Chem. SOC.,85,4014 (1963). 2. S . J. Lapporte, Ann. N. Y . Acad. Sci., 158, 510 (1969). 3. J. Boor, Jr., Ziegler-Natta Catalysts and Polymerization, Academic Press, New York, 1979, p. 607. 4. B. R. James, Homogeneous Hydrogenation, John Wiley, New York, 1973, Ch. XV and ref-

erences therein.

5. B. R. James, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. Abel, eds., Vol. 8, Ch. 51, Pergamon Press, Oxford, 1982, p. 340. 6. W. R. Kroll, J. Catal., 15, 281 (1969). 7. R. Waymouth, P. Pino, J . Am. Chem. SOC., 112, 4911 (1990). 8. F. R. W. P. Wild, L. Zsolnai, G. Huttner, H. H. Brintzinger, J . Organomet. Chem., 288, 63 (1985). 9. R. F. Jordan, J . Chem. Educ., 65, 285 (1988), and references therein. 10. G. C. Hlatky, H. W. Turner, R. R. Eckman, J. Am. Chem. SOC., 111,2728 (1989). 11. J. J. Eisch, A. M. Pitrowski, S . K. Brownstein, E. G. Gabe, F. L. Lee, J. Am. Chem. SOC., 107, 7219 (1985). 12. E. Giannetti, G. M. Nicoletti, R. J. Mazzocchi, Polym. Sci. Polym. Chem.,23, 2117 (1985). 13. P. G. Gassman, M. R. Callstrom, J . Am. Chem. SOC.,109, 7875 (1987). 14. P. Pino, M. Galimberti, J . Organomet. Chem, 370, 1 (1989). 15. P. Pino, M. Galimberti, J. Wei, N. Piccolrovazzi, in Transition Metals and Organometallics

as Catalysts for Olejin Polymerization, W. Kaminsky, H. Sinn, eds., Springer-Verlag, New York, 1980, p. 269. 16. P. Pino, P. Cioni, J. Wei, J . Am. Chem. SOC.,109, 6189 (1987).

14.3.3.6. Ruthenlum(l1) Catalysts

A review of homogeneous hydrogenation by Ru catalysts was published in 1970' and the synthesis and properties of ruthenium-hydride complexes known prior to 1977 have been reviewed'. An up to date review of ruthenium-hydride complexes appeared in 19843. We concern ourselves here with hydrogenation reactions that involve Ru(I1) species as catalyst precursors or catalytic intermediates. The first example of activation of molecular H, by a Ru(I1) species was described in 19614.The study employed a solution of Ru(I1) choride in aqueous HC1 at 65 to 95°C and H, P of about 1 atm to catalytically hydrogenate maleic, fumaric, and acrylic acid. The Ru(I1) species that are prepared in situ by reduction of solutions of Ru(II1) with Ti(II1) is thought to be RuCli-. The latter blue solutions turn yellow in the presence of the olefin substrates, and they absorb H, according to a rate law that is first-order each in Ru(I1) and H,. The yellow intermediate is shown spectrophotometrically to be a Ruolefin complex. The following mechanism was proposed4:


100

--

+ olefin SRu(I1) (olefin) Ru(II)(olefin) + H, [Ru(II)(olefin)(H)]- + H + Ru(I1)

[Ru(II)(olefin)(H)][Ru(II)(alkyl)]-

+ H+

(a) (b)

[Ru(II)(alkyl)] -

(c)

Ru(I1)

(d)

+ saturated product

The RuCli- system is only an effective catalyst for the hydrogenation of activated olefins. Unactivated olefins form ruthenium adducts, but are not hydrogenated. In a related system, hydrogenation of maleic acid with H, is catalyzed by [RuCl4(bipy)l2- in aqueous HCl solution5. Although the spectroscopic behavior of the former and latter systems are similar, kinetic studies suggest the following mechanism: [RuCl,(bipy)12-

+ olefin +

-

[RuCl,(bipy)(olefin)] -

[RuCl,(bipy)]’-

+ H,

+ C1-

(none productive side reaction)

(e)

+ HCl (entry into the catalytic cycle)

(f)

-

[RuHCl,(bipy)]’-

+ olefin +[RuHCl,(bipy)(olefin)]- + C1-

[RuHCl,(bipy)]’-

[RuHCl,(bipy)(olefin)] [RuCl,(bipy)(alkyl)]-

+ H, + C1-

[RuCl,(bipy)(alkyl)] -

[RuHCl,(bipy)I2-

+ H + + saturated product

(g) (h) (i)

Many Ru(I1)-phosphine compounds are also active homogeneous hydrogenation catalysts. RuCl,(PPh,), serves as a catalyst precursor, which gives the catalytically active RuHCl(PPh,), upon reaction with H, in situ. Several reviews describe hydrogenation reactions catalyzed by RuHCl(PPh,), lS6. The mechanism of hydrogenation catalyzed by RuHCl(PPh,), is not simple; it is apparently sensitive to reaction conditions. Hydrogenation of terminal alkenes by RuCl,(PPh,), is slow in benzene, but becomes rapid in 1:1 benzene-ethanol. The alcohol may serve as a base that promotes formation of RuHCl(PPh,), with concomitant change in color from brown to violet7: RuCl,(PPh,),

+ H, + base

-

RuHCl(PPh,),

+ base.HC1

(3

Other bases are equally effective: Et,N, NaOPh, and KOH. RuCl,(PPh,), in benzeneethanol also hydrogenates alkynes*. RuHCl(PPh,), has been synthesized by treating RuCl,(PPh,), with NaBH, in wet benzene. RuHCl(PPh,), ( lo-,M) in benzene selectively reduces terminal olefins’. However, in contrast to RuCl,(PPh,), in benzene-ethanol, RuHCl(PPh,), in benzene does not catalyze acetylene hydrogenation. The catalyst is only slightly soluble in benzene, the system is oxygen sensitive and easily poisoned, and the kinetics are rapid and in many cases diffusion controlled. The solubility problem is overcome by employing dimethylacetamide as the solvent. Kinetic data obtained in DMA support the mechanism’: RuHCl(PPh,), S RuHCl(PPh,),

-

+ PPh,

+ alkene FRuCl(PPh,),(alkyl) RuHCl(PPh,), + alkene RuCl(PPh,),(alkyl) + H, RuHCl(PPh,),

(k) (1)

(m)

14.3. Hydrogenation Reactions 14.3.3. Classes of Soluble Catalysts 14.3.3.6. Rutheniurn(l1) Catalysts

101

Observed hydrogen atom exchange and olefin isomerization reactions that are observed when RuHCl(PPh,), is reacted with alkenes in the absence of H, are thought to occur via equilibrium (i)7*8.’0. In the absence of substrate of D, at 1 atm reacts with RuHCl(PPh,), at 25°C to give RuDC~(P(~,~-D,C,H~)~)~’~~ ’,”. The exchange involves reversible intramolecular oxidative addition of the ortho-phenyl C-H bond to the Ru center of the coordinatively unsaturated RuDCl(PPh,), (i.e., orrho-metallation). The rate of olefin isomerization and the rate of hydrogen exchange with the phosphine ligands are slow with respect to the rate of hydrogenations catalyzed by RuHCl(PPh,),. Chiral phosphine derivatives of Ru hydrogenation catalysts have been developed. These represent a significant departure from the Rh systems that previously dominated the field of asymmetric hydr~genation’~. Most of the useful catalysts are based on Ru(I1) (BINAP), where BINAP = 2,3’-bis(diphenylphosphino)-l,l’-binaphthl. The mechanism of equations (n)-(q) has been proposed for the asymmetric hydrogenation of a,p-unsaturated carboxylic acids in methanol solvent catalyzed by Ru(I1) complexes derived from the precursor Ru(I1) (BINAP)(OAc), (1, R = Me)l49I5.

1

102

14.3.4. Hydrogenation of Aliphatic C-C 14.3.4.1. In Simple Olefins 14.3.4.1.1. Isolated Double Bonds. O

A

1-

Functions

R

The mechanism of equations (n)-(q) for the Ru(I1) (BINAP)(OAc), system is similar to the hydrogenation of unsaturated carboxylic acids in aqueous solutions catalyzed by Ru(II)Cl, complexes [equations (a)-(d)], the first characterized homogeneous catalytic hydrogenation of an olefin, in that both systems apparently proceed via a heterolytic cleavage of Hi.Indeed, heterolytic addition of H, in Ru hydrogenation systems is pervasive. But, a difference between the two systems is found in the way the substrate is bound to the Ru center. Olefinic substrates typically bind to homogeneous hydrogenation catalysts via their C=C bonds. However, whereas olefinic substrates binds to the Ru center of the RuCl;--based system via coordination of their C=C bond prior to the heterolytic addition of H2, the u,P-unsaturated carboxylic acid substrates of the Ru(I1) (BINAP)(OAc), system apparently bind to the Ru center via their carboxylate groups. In contrast to other asymmetric hydrogenation systems16, since step (0)is turnoverlimiting, the stereochemistry of the saturated product is determined before coordination of an enantioface of the prochiral olefin. (M. T. ASHBY)

B. R. James, Inorg. Chim. Acta Rev., 4 , 73 (1970). G . L. Geoffroy, J. R. Lehman, Adv. Inorg. Chem. Radiochem., 20, 189 (1977). E. A. Seddon, K. R. Seddon, The Chemistry of Ruthenium, Elsevier, Amsterdam, 1984. J. Halpem, J. F. Harrod, B. R. James, J . Am. Chem. Soc., 83, 753 (1961); 88, 5150 (1966). B. C. Hui, B. R. James, Inorg. Nucl. Chem. Lett., 6 , 367 (1970). B. R. James, Homogeneous Hydrogenation, John Wiley, New York, 1973. 7. P. S. Hallman, B. R. McGarvey, G. Wilkinson, J . Chem. Soc., A, 3143, (1968). 8. D. Evans, J. A. Osbom, F. H. Jardine, G . Wilkinson, Nature (London),208, 1203 (1965). 9. B. Hui and B. R. James, Proc. 4th Intern. Con6 on Organomet. Chem., Bristol, 1969, L6. 10. D. Rose, J. D. Golbert, R. P. Richardson, G . Wilkinson, J . Chem. SOC,A , , 2610 (1969). 11. G . W. Parshall, W. H. Knoth, R. A. Schunn, J . Am. Chem. Soc., 91,4990 (1969). 12. J. J. Levinson, S. D. Robinson, J . Chem. SOC.A., 639 (1970). 13. R. Noyori, H. Takaya, Acct. Chem. Res., 23, 345 (1990). 14. M. T. Ashby, J. Halpem, J . Am. Chem. Soc., 113, 589 (1991). 15. M. T. Ashby, M. A. Khan, J. Halpem, Organometallics, 10, 2011, (1991). 16. C. R. Landis, J. Halpem, J. Am. Chem. Soc., 109, 1746 (1987), and references therein. 1. 2. 3. 4. 5. 6.

14.3.4. Hydrogenation of Aliphatic C-C

Functions

14.3.4.1. In Simple Olefins 14.3.4.1 .l. Isolated Double Bonds.

The reduction of multiple C-C bonds with excess H, in a suitable solvent in the presence of a metal catalyst can achieve controlled transformations with little experimentation. This addition proceeds easily and is widely used in organic synthesis.


102

14.3.4. Hydrogenation of Aliphatic C-C 14.3.4.1. In Simple Olefins 14.3.4.1.1. Isolated Double Bonds. O

A

1-

Functions

R

The mechanism of equations (n)-(q) for the Ru(I1) (BINAP)(OAc), system is similar to the hydrogenation of unsaturated carboxylic acids in aqueous solutions catalyzed by Ru(II)Cl, complexes [equations (a)-(d)], the first characterized homogeneous catalytic hydrogenation of an olefin, in that both systems apparently proceed via a heterolytic cleavage of Hi.Indeed, heterolytic addition of H, in Ru hydrogenation systems is pervasive. But, a difference between the two systems is found in the way the substrate is bound to the Ru center. Olefinic substrates typically bind to homogeneous hydrogenation catalysts via their C=C bonds. However, whereas olefinic substrates binds to the Ru center of the RuCl;--based system via coordination of their C=C bond prior to the heterolytic addition of H2, the u,P-unsaturated carboxylic acid substrates of the Ru(I1) (BINAP)(OAc), system apparently bind to the Ru center via their carboxylate groups. In contrast to other asymmetric hydrogenation systems16, since step (0)is turnoverlimiting, the stereochemistry of the saturated product is determined before coordination of an enantioface of the prochiral olefin. (M. T. ASHBY)

B. R. James, Inorg. Chim. Acta Rev., 4 , 73 (1970). G . L. Geoffroy, J. R. Lehman, Adv. Inorg. Chem. Radiochem., 20, 189 (1977). E. A. Seddon, K. R. Seddon, The Chemistry of Ruthenium, Elsevier, Amsterdam, 1984. J. Halpem, J. F. Harrod, B. R. James, J . Am. Chem. Soc., 83, 753 (1961); 88, 5150 (1966). B. C. Hui, B. R. James, Inorg. Nucl. Chem. Lett., 6 , 367 (1970). B. R. James, Homogeneous Hydrogenation, John Wiley, New York, 1973. 7. P. S. Hallman, B. R. McGarvey, G. Wilkinson, J . Chem. Soc., A, 3143, (1968). 8. D. Evans, J. A. Osbom, F. H. Jardine, G . Wilkinson, Nature (London),208, 1203 (1965). 9. B. Hui and B. R. James, Proc. 4th Intern. Con6 on Organomet. Chem., Bristol, 1969, L6. 10. D. Rose, J. D. Golbert, R. P. Richardson, G . Wilkinson, J . Chem. SOC,A , , 2610 (1969). 11. G . W. Parshall, W. H. Knoth, R. A. Schunn, J . Am. Chem. Soc., 91,4990 (1969). 12. J. J. Levinson, S. D. Robinson, J . Chem. SOC.A., 639 (1970). 13. R. Noyori, H. Takaya, Acct. Chem. Res., 23, 345 (1990). 14. M. T. Ashby, J. Halpem, J . Am. Chem. Soc., 113, 589 (1991). 15. M. T. Ashby, M. A. Khan, J. Halpem, Organometallics, 10, 2011, (1991). 16. C. R. Landis, J. Halpem, J. Am. Chem. Soc., 109, 1746 (1987), and references therein. 1. 2. 3. 4. 5. 6.

14.3.4. Hydrogenation of Aliphatic C-C

Functions

14.3.4.1. In Simple Olefins 14.3.4.1 .l. Isolated Double Bonds.

The reduction of multiple C-C bonds with excess H, in a suitable solvent in the presence of a metal catalyst can achieve controlled transformations with little experimentation. This addition proceeds easily and is widely used in organic synthesis.

14.3.4. Hydrogenation of Aliphatic C-C 14.3.4.1. In Simple Olefins 14.3.4.1.1. Isolated Double Bonds.

Functions

103

Catalysts and procedures improve the selectivity of the metal-catalyst systems and asymmetric syntheses. Hydrogen-transfer and homogeneous catalysts offer alternatives to heterogeneous hydrogenation'-6. The references in the following sections refer to work of practical significance either for preparation or stereochemistry, or as examples of the common problems encountered. (i) Choice of the Catalyst. Isolated carbon-carbon double bonds are hydrogenated with common heterogeneous catalysts, solvents, and operating conditions. It is unnecessary to use particularly active catalysts, except for hindered olefins, e.g., tetrasubstituted double bonds or double bonds at specific positions in steroid systems. Universal rules are not available for the choice of catalysts promoting hydrogenation of unsaturated compounds, but most catalytic reductions of olefins are conducted over Pd or Pt which make active catalysts for this purpose7. Platinum (PtO,) is useful when reduction fails to start or go to completion and Pd is often superior to other catalysts to effect this reaction at low T and P. The cataylst may need to be activated, e.g., to reduce an isolated steroid 5d-double bond that possesses only low-order activity. Perchloric acid activates Pt, which allows cholesterol to be hydrogenated in 30 min to cholestane-3P-01'.

HO

Pt, HCQ H,, 105Pa, 4O-5O0C

'

A 88%

Saturation of a 3P-acetate-A5 steroid can be accomplished over 10% Pd-C, in ethanol. The choice of a catalyst is determined by the olefin and by other functionality. Platinum, Pd, and Ni hydrogenate gaseous or easily volatilized olefins in the vapor. Noble metal catalysts, finely divided and activated prove efficient for liquid olefins, unsaturated alcohols, and cycloalkenes at atm P, in ethanol, glacial acetic acid9 or ethyl acetate. Other Ni catalysts require H, P. Side chains of aromatic compounds can be hydrogenated in the liquid phase with Ni or Pt and in the gas phase with Cu catalysts9. Other catalysts are preferred when selectivity is a problem. Ruthenium and Rh are useful catalysts for selective hydrogenation of the olefinic function in oxygenated compounds, a reaction plagued by hydrogenolysis. Ruthenium dioxide catalyzes the hydrogenation of the otherwise unreactive tetrasubstituted double bond in guaiol, 1, without hydrogenoloysis": RuO,

H,, 107 Pa

'

Palladium or Pt leads to complex mixtures in the hydrogenation of toxol, 2, but Rhon-alumina smoothly affords dihydrotoxol after absorption of one equiv of H,' l :

104

14.3.4. Hydrogenationof Aliphatic C-C 14.3.4.1. In Simple Olefins 14.3.4.1-1. Isolated Double Bonds.

5%Rh-AI,O3 EtOH

0

'

Functions

O

To hydrogenate substrates that contain sulfur impurities or in synthesis of sulfurcontaining organic compounds, sulfide catalysts that are not poisoned by sulfur are recommended. All six noble metals, Pd, Pt, Ru, Ir, and Os, make active homogeneous hydrogenation catalysts which enlarge the capabilities of reaction control. Some of the most active homogeneous catalysts for rapid hydrogenation of olefins are tris(tripheny1phosphine)hydridochlororthenium(II), [Ir(COD)(pyr)(Pi-Pr,)] [PF,] -, and the rhodocarborane prepared by the reaction of tris(tripheny1phosphine)chlororhodium with Cs + [7butenyl-7,8-C,B9H,] - , The rhodocarborane has an initial hydrogenation rate that is 30 times faster than that of [Rh(PPh,),C1]12. However, since the reductions discussed so far can be done better over heterogeneous catalysts, homogeneous catalysts are used only when problems of selectivity are involved, e.g., homogeneous catalysts [RhH(CO)(PPh,),, RuCl,(PPh,), , RhH(DBP),] are better for selective hydrogenation of terminal olefins and for hydrogenations of olefins without hydrogenolysis of carbon-heteroatom bonds. The stereoselective total synthesis of seychellene, 3, uses the RhCl(PPh,),-catalyzed hydrogenation of an exocyclic methylenel,: +

3 Iridium is little used, but gives satisfactory results when stereoselective reduction of steroids is required, e.g., the methylene compound, 4, is transformed to the 16P-methyl compound, 5, by the use of Ir deposited on BaSO, or CaC0,I4.

& &, f OAc

H

H,, Ir-BaSO, >

+ 16a-isomer

(e)

n-octane

cis- and trans-stilbene Pd black bibenzyl 1,l-diphenylethylene

cyclohexene

100%

(W2,

1,l-diphenylethane 100%

(i),,

Reduction of substituted benzyl alcohols of styrenes is achieved using hydrogen transfer from cyclohexene catalyzed by Pd on carbon and AlCl,:

R Rf

R

R'

Limonene or tetralin may also be used as hydrogen donors. Cyclic amines, e.g., indoline and pyrrolidine, are more reactive in catalyzed transferhydrogenation than oxygenated and hydroaromatic corn pound^^^.

o+a;>

RhCKPPh3)3>

0

H n=0,1

(k)

+

H

(ii) Selectivity. Selectivity influences the choice of a catalyst. Ruthenium on activated carbon catalyzes the selective hydrogenation of monosubstituted olefins in the presence of di- and trisubstituted olefins. Thus in mixtures, 6, is reduced in preference to either 7 or SZ5, following the rule that an ethylene can be more easily hydrogenerated the lower the substitution.

(CH,),CHCH,CH=CH2

6

CH,CH,CH,C(CH,)=CH, 7

CH,CH2CH=C(CH,),

8

106

14.3.4. Hydrogenation of Aliphatic C-C 14.3.4.1. In Simple Olefins 14.3.4.1.1. Isolated Double Bonds.

Functions

Selectivity of Ru is also evident from the reduction of octene-1 in the presence of octene-2. The complex formed by a Co(I1) salt and NaBH, (most likely a Co hydride species) reduces alkenes with high steric selectivity: mono + di tri- and tetrasubstituted olefins. Steric effects can best explain the observed selectivity. The reduction of limonene, 9, demonstrates the synthetic utility of this reducing agent?

-

1-octene n-octane cyclohexene --+ cyclohexane

98% 98%

z-BuCO-CHZCH~P~ 90%

i-PrOH, 12 h

PhCO-CH=CH-Ph

PhCO-CH,CH,-Ph

95%

(e)

Similarly Ni catalysts are useful substitutes to Pd for the selective olefin reduction in an a$-unsaturated ketone. Finely divided Ni (and partially deactivated by washing it with a 0.1% MeOH solution of acetic acid) in a mixture 5% HCCl,/EtOH is efficient at RT and atm P". Neutral catalysts, U-Ni-N, obtained by refluxing precipitated Ni with isopropanol, also shows selectivity, mesityl oxide being reduced to methyl- 1 pentanone-212. 4-Cyclopentene- 1,3-dione, 5 , can be converted into 1,3-cyclopentanedione, 6,by hydrogenation over partially deactivated neutral Ni cataly~t'~:

H2, Partially EtOH deactivated Ni

0

6 6

75-81%

(f)

0

Excellent yields also are obtained for the reduction of acyclic enones using Co(C0) as catalyst under H,/COl4. Mesityl oxide can be selectively hydrogenated to methylisobutylketone with homogeneous catalysts such as RhCl(PPh,),". The low-valent Co complex, H,Co[P(OP,,),],, is a soluble selective catalyst for the efficient hydrogenation of a,&unsaturated ketones and amides to the corresponding saturated ketones and amidesI6. Catalytic activity is greatly increased without loss in selectivity by reacting at 70°C. Unsaturated aldehydes are not reduced under similar conditions. Quinones are selectively reduced over homogeneous catalysts as illustrated by the quinone, 717,even in the presence of an olefin as in the quinone 8, over Zn in acetic acid".

kOCH3 - kOCH

14.3.4. Hydrogenation of Aliphatic C-C Functions 14.3.4.1. In Simple Olefins 14.3.4.1.2. Olefins Conjugated to Carbonyl, Nitrile, Nitro.

116

7

(g)

OCH,

OCH,

0

0 0

0

Tris(tripheny1phosphine)rutheniumdichloride is able to transfer hydrogen from primary alcohols to a,&unsaturated carbonyl compounds, e.g., when benzyl alcohol is heated under N, at 200°C for 2 h with benzalacetone and the catalyst, benzal (90%)and 4-phenyl-butane-2-one (92%) are formed”: PhCHzOH

+ Ph-CH=CH-COCH, R-CH=CH-COR

’

RuCI,(PPh,),

Ph-CH,CHz-COCH, RCH,CH,-COR’

+ PhCHO

(i)

The reaction does not proceed well in the case of aldehydes. A wine-red hydridoiron carbonyl complex generated in situ from Fe(CO), and a small amount of base in moist solvents can be used in a stoichiometric manner for the hydrogenation and deuteration of a$-unsaturated ketones, aldehydes, esters, and lactones. The mixture stands at RT for 12 h, and yields are often greater than 90%”:

-

Benzalacetone A4-Cholestenone Dimethylmaleate Cinnamaldehyde 5-Hydroxy-2-hexenoic acid Glactone

[Fe(CO),-OH] [Fe(CO),-OHl[Fe(CO),-OH] [Fe(CO),-OH] -

Benzylacetone Coprostanone Dimethylsuccinate 3-Phenylproprionaldehyde 5-Hydroxyhexanoic acid GLactone

98% 32% 96% 98%

90%

Systems prone to hydrogenolysis require special techniques, eg, 9 is hydrogenated to 10 in 55% yield, over 10% Pd-on-carbon in an apparatus equipped with a Hg-filled manometer and burette”:

9

J&

0

+s

-27‘C, Pd-C Et,O >

J&

& ,

6 10

55%

(j)

(ii) Stereochernistry. When the P-carbon of the a,P-unsaturated ketone is part of a ring junction, the stereochemistry of the saturated molecule is dependent on the reaction


117

parameters, but also on some structural feature of the substrate. First, there is evidence of the participation of the carbonyl function, e.g., through partial 1,4-addition2*,or making easy the hydrogenation over Pd of the usually resistant tetrasubstituted double bond in As-unsaturated 7-keto steroids23.Under all conditions saturation of the hydrindenones, 11 and 12,affords the cis-ring fused corresponding

11

6

0

&-+

12

cis-ringfusion

(k)

R = H, CH,, COOEt However, hydrogenation of p-octalones, 13, leads to mixtures of cis-trans ring fused products in ratios depending on the pH of the reaction medium, on the solvent, and on the C-10 substituent (see Table 1). With some polysubstituted compounds, preTABLE1. INFLUENCE OF PH OF THE MEDIUM, OF THE SOLVENT, OF THE C-10 SUBSTITUENT ON RINGFUSION,DURING fi-OCTALONES HYDROGENATION

Medium26 R = H

Acidic (EtOH, aq. HCl) Basic (EtOH, aq. NaOH) Protic Aprotic

S~bstituent~~.~'

t-BuOH MeOH hexane DMF

R = COOEt R = CH3

10.9 33.6 c = 1.89 E = 38.0

c = E =

93% 62%

7% 38%

91%

9% 59% 52% 21%

41% 48%

79% 0%

80%

100% 20%

TABLE2. ISOMER DISTRIBUTION RELATED TO SUBSTITUENTS FROM THE HYDROGENATION OF 14

R

= R' = H29 R = OH,R' = H30 R = OH, R' = CH2(isoxazolin)30

R = OtBu,R' = H3' R = OtBu,R' = COOH3'

~~

Pd-C Pd-C Pd-C Pd-C Pd-BaSO,

100% 100% 15% 70%

-

-

85% 30%

100%

14.3.4. Hydrogenation of Aliphatic C-C Functions 14.3.4.1. In Simple Olefins 14.3.4.1.2.Olefins Conjugated to Carbonyl, Nitrile, Nitro.

118

dicting the stereochemical result is difficult (see Table 2), e.g., for the hydrogenation of 14. The results are explained on configurational grounds. Hydrogenation of A4-3-keto steroids mostly over Pd result in productin of a mixture of cis- and trans-A/B ring fused isomers. The addition of base (KOH) to the reaction mixture favors the formation of the cis-isomer ( 5 P - h ~ d r o g e n ) ~ ~ s ~ ~ . Functional groups which lie far from the reaction site, such as at C-11 and C-17, influence the stereochemical outcome of hydrogenation according to their nature and ~ r i e n t a t i o n(see ~ ~ Table 3). These results cannot be interpreted in steric terms alone; electronic factors must be involved. Modification of Pd black with pyridine affords a recognized route to 5P-steroids, but 4-methoxypyridine is an even more effective modifying agent, giving high yields (95%) of 5P-steroids, 16,which may contain a variety of substituent~~~: 15

H,, Pd black 4-MeO-pyr

R' = H o r O RZ = H, P-CgH,,, P-OH, P-OAC, =O, P-AC

16 >95% of 5P-H

(1)

R3 = H,CH3

The hydroxyl coordination with catalysts affords an efficient stereocontrol of the homogeneous hydrogenation of hydrindanone 1736.

aoH' [Ir(COD)(PPh,)(pyr)]

O

-H

OH

+ PF,

0

CH2C12

--

cis

17

trans

(iii) Conjugated Aldehyde, Acid, Ester, Nitrile, Nitro Functions. Double bond hydrogenation of a$-unsaturated potentially reducible groups other than ketones is also possible. Aliphatic saturate aldehydes are obtained in high yields by reduction over Pd,

TABLE3. RATIOOF 5P- TO 5a-H IN THE HYDROGENATION OF d4-3-KET0 STEROIDS, 15

H, Pd I-ROH 9

' 0

0

5Pba = R* = H R1 = H, R2 = a-OH R1 = H,RZ = P-OH R1 = H, R2 = P-OAc

R1

R 3 = CH3

1 .o 4.1 0.73 1.89

R3

= CH3

R' = R1 = Rl = R' =

P2 H

P-OH, R2 = P-OAc 0, R2 = P-OAc 0 R2 = 0 0 , R 2 = P-CH,CO

l6

5Pls.

0.62 0.13 0.09 0.03

14.3.4. Hydrogenation of Aliphatic C-C Functions 14.3.4.1. In Simple Olefins 14.3.4.1 -2. Olefins Conjugated to Carbonyl, Nitrile, Nitro.

119

Pt, Rh, and Ru on various support^'^^^; Pd is often the choice catalyst38.Reduction of aromatic a$-unsaturated aldehydes is more difficult than their aliphatic counterpart but borohydride-reduced Ni or Pd exhibit an almost perfect selectivity in the reduction of both aliphatic and aromatic conjugated aldehydes39940: R-CH=CHCHO R-H,

M ~ph ,

H, > borohydride reduced-Ni or Pd

RCHZCHZCHO

(m)

Double reduction of P-chloro-unsaturated aldehyde is possible under basic conditions4’: c1\ Ar’

c=c

/R

/R CHz- C\

H, , Pd-C

‘CHO

Ar’

CHO

32-86% NaOHorK*CO,

,

(0)

4CH0

CHO

On hydrogenation of unsaturated acids the double bond is usually saturated preferentially. Over PtO,, crotonic acid and undecenoic acid are converted to butyric acid and undecanoic acid4’, while the Na salt of maleic acid is hydrogenated to succinic acid (98% yield) at high P (17 X lo6 Pa), 100°C, with finely divided Ni4. Palladium-on-carbon reduces acid conjugated olefins at RT43. Alkylidene carboxylic esters, acrylates, and methylene lactones are hydrogenated over finely divised Ni or P d - ~ n - c a r b o nTetra~~. substituted double bonds conjugated with an ester require black Pd46, Pt47, or Pt028. Ester conjugated olefins may be reduced selectively in the presence of an isolated double bond like in steroid 1849:

Unsaturated acids and esters including fatty acids derivatives are also selectively hydrogenated over homogeneous catalysts5’, e.g., Ir(CO)Cl(PPh,), catalyzes the hydrogenation of styrene, ethylacrylate at lo6 Pa and 80-120°C51. Selective reduction of unsaturated nitriles presents little difficulty, and palladium is a popular catalyst for this5’. Even tetrasubstituted double bonds can be selectively reduced, in the presence of a nitrile function, over platinum dioxide53: COOEt

NcAp EtOOC

COOEt H2.PtOz

~

NC% EtOOC

86%

(9)

120


Nickel boride may offer some advantage since the catalyst is not poisoned by amine~~~: CH,CH=CHCN

H,, Ni-bride > Ch,CH,CH,CN

100%

Unsaturated nitro compounds are reduced to the corresponding saturated nitro compounds55unless the nitro group is bonded to an aromatic nucleus equation (s)? O,N-(C6H,J-CH=CHCOOH

Hz’Ruo2> H2N-(C,H4)-CH2CH,COOH ElOH

(s)

70%

Unsaturated phosphoranes are quantitatively reduced over P ~ O , ~ ~ :

(J.-L. GRAS)

1. W. Taub, J. Szmuszkovicz, J . Am. Chem. SOC., 74,2117(1952). 2. E.Ucciani, L. Tanguy, C . R. Hebd. Seances Acad. Sci., C284,577 (1977). 3. K. Kakiuchi, T. Nakao, M. Takeda, Y. Tobe, Y. Odaira, Tetrahedron Lett., 25,557 (1984). 4. P. D.Hobbs, P. D. Magnus, J . Am. Chem. SOC.,98,4594(1976). 5. A. G. Schulte, J. P. Dittami, J . Org. Chem, 49,2615 (1984). 6. H. J. Ringold, J . Am. Chem. SOC.,82,961 (1960). 7. P. A. Aristoff, P. D. Johnson, A. W. Harrison, J . Am. Chem. SOC.,107,7967(1985). 8. The course and stereospecificity of the reduction of 2-benzylidene-l-indanonesis known: W. Hiickel, M. Macer, E. Jordan, W. Seeger, Justus Liebigs, Ann. Chem. 616,46(1958). 9. A. P.Phillips, J. Mentha, J . Am. Chem. SOC.,78, 140 (1956). 10. H. B. Henbest, J. Trocha-Grimshaw, J . Chem. SOC.,Perkin 1,601(1974). 11. R. Cornubert, H. G. Eggert, P. Thomas, C . R. Hebd. Seances Acad. Sci., C234,2324(1952). 12. M. Kajitani, J. Okada, T. Ueda, A. Sugimori, Y. Urushibara, Chem. Lett., 777 (1973). 13. J. Sraga, P. Hmciar, Synthesis, 282 (1977). 14. E. Ucciani, R. Lay, L. Tanguy, C . R. Hebd. Seances Acad. Sci., C281, 877 (1974). 15. W. Strohmeier, E. Hitzec, J . Organomer. Chem., 91,373 (1975);102,C37 (1975). 16. M. C. Rakowski, E. L. Muetterties, J . Am. Chem. SOC., 99,739 (1977). 17.A. J. Birch, K. A. M. Walker, Tetrahedron Lett., 3457 (1967). 18. A. Bertsch, W. Grimme, G. Reinhardt, Angew. Chem., lnt. Ed. Engl. 25,377 (1986). 19. Y.Sasson, J. Blum, Tetrahedron Lett., 2167 (1971). 20. R. Noyori, I. Umeda, T. Ishigami, J . Org. Chem., 37,1542 (1972). 21. A. Nickon, 3. F. Bagli, J . Am. Chem. SOC.,83,1498 (1961). 22. R. L.Augustine, Adv. Caral., 25,63 (1976). 23. C. Djerassi, E. Bates, M. Velasco, G. Rosenkranz, J . Am. Chem. SOC., 74,1712 (1952). 24. M. Chaykovsky, R. E. Ireland, J . Org. Chem., 28,748 (1963). 25. J. E.McMurry, T. R. Webb, J . Med. Chem., 27,1367 (1984). 26. R. L. Augustine, J . Org. Chem., 23,1853 (1958). 27. F. J. McQuillin, W. 0. Ord, J. Chem. SOC.,2902 (1959). 28. J. B. Jones, D. R. Doods, Can. J . Chem., 25,2397 (1987). 29. R. L. Augustine, A. D. Broom, J . Org. Chem., 25,802 (1960). 30. T. C. McKenzie, J . Org. Chem.,39,629(1974). 31. Z. G. Hajos, D. R. Panish, J . Org. Chem., 38, 3239 (1973). 32. H. J. E.Loewenthal, Tetrahedron, 6 269 (1959). 33. T. Witiak, R. J. Patch, S . J. Emma, Y.K. Fung, J . Med. Chem., 29,1 (1986).

14.3.4. Hydrogenation of Aliphatic C-C 14.3.4.1. In Simple Olefins 14.3.4.1.3. Vinyl Functions. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51.

52. 53. 54. 55. 56. 57.

Functions

121

K. Mori, K. Abe, M. Washida, S . Nishimura, M. Shiota, J . Org. Chem., 36, 231 (1971). N. Tsuji, J. Suzuki, M. Shiota, I. Takahashi, S . Nishimura, J . Org. Chem., 45, 2729 (1980). G . Stork, D. Ekahne, J . Am. Chem. Soc., 105, 1072 (1982). E. Ucciani, R. Lai, L. Tanguy, C . R. Hebd. Seances Acad. Sci., C283, 17 (1976). N. A. Cortese, R. F. Heck, J . Org. Chem., 43, 3985 (1978). T. W. Russel, D. M. Duncan, S . C. Hansen, J . Org. Chem., 42,551 (1977). J. H. Billman, J. I. Stiles, J. Tonnis, Synth. Commun., I , 127 (1971). J. A. Virgilio, E. Heilweil, Org. Prep. Proced. Int., 14, 9 (1982). E. B. Maxted, V. Stone, J . Chem. Soc., 26 (1934). D. Hermeling, H. J. Schafer, Chem. Ber., 121, 1151 (1988). B. B. Allen, B. W. Wyatt, H. R. Henze, J . Am. Chem. SOC., 61, 843 (1939). J. Rebek Jr., D. F. Tai, Y. K. Shue, J . Am. Chem. Soc., 106, 1813 (1984). R. Yamaguchi, M. Bau, N. Kawanisi, Bull. Chem. SOC. Jpn., 61,2909 (1988). J. R. Donaubauer, A. M. Greaves, T. C. McMorris, J . Org. Chem., 49,2833 (1984). S. Hatakeyama, H. Numata, S . Takano, Tetrahedron Lett., 25, 3617 (1984). M. Kocor, M. M. Kabat, J. Wicha, W. Peczynska-Czock, Steroids, 41 (1983). A. J. Birch, D. H. Williamson, Org. React., 24, 1 (1976). W. Strohmeier, M. Lukacs, J . Organomet. Chem., 129, 331 (1977). C. W. Whitehead, J. J. Traverso, H. R. Sullivan, F. J. Marshall, J. Org. Chem., 26,2814 (1961). A. G . Anderson, W. F. Harrison, R. G . Anderson, J . Am. Chem. Soc., 85, 3448 (1963). T. W. Russel, R. C. Hoy, J. C. Cornelius, J . Org. Chem., 37, 3552 (1972). H. Feuer, R. Hannetz,J. Org. Chem., 26, 1061 (1961). I. S. Monakhova, N. S. Smimova, K. I. Karpavichyus, V. G . Kharchenko, Zh. Prikl. Khim., 55, 704 (1982). G . Falsone, U. Wingen, Tetrahedron Lett., 20, 675 (1989).

14.3.4.1.3. Vinyl Functions.

Enol ethers and esters can be hydrogenated over 5% Pd-on-BaSO or SrCO, in ethanol at RT and atm P, but hydrogenolysis is often a concomitant reaction', sensitive to substrate, solvent and catalyst, e.g., triflate of cyclic or steroidal ketones are readily hydrogenated over Pt0;:

wphmpha

Selective hydrogenation of isoflavon, 1, is achieved in almost quantitative yield to isoflavone, 2, in dioxane with Pd-on-carbon, whereas the same reaction conducted in ethanol or aqueous acetone yields a mixture of a- and P-isoflavan-4-01,3, also in quantitative yield, the a-isomer being formed preferentially3:

Ph

3

OH

0 1

0

2

(b)

100%

Platinum favors hydrogenolysis and Pd, Ru, Rh favor hydr~genation~-~, polar solvents favor hydrogenolysis, whereas nonpolar solvent favor hydrogenation7, e.g.':


14.3.4. Hydrogenation of Aliphatic C-C 14.3.4.1. In Simple Olefins 14.3.4.1.3. Vinyl Functions. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51.

52. 53. 54. 55. 56. 57.

Functions

121

K. Mori, K. Abe, M. Washida, S . Nishimura, M. Shiota, J . Org. Chem., 36, 231 (1971). N. Tsuji, J. Suzuki, M. Shiota, I. Takahashi, S . Nishimura, J . Org. Chem., 45, 2729 (1980). G . Stork, D. Ekahne, J . Am. Chem. Soc., 105, 1072 (1982). E. Ucciani, R. Lai, L. Tanguy, C . R. Hebd. Seances Acad. Sci., C283, 17 (1976). N. A. Cortese, R. F. Heck, J . Org. Chem., 43, 3985 (1978). T. W. Russel, D. M. Duncan, S . C. Hansen, J . Org. Chem., 42,551 (1977). J. H. Billman, J. I. Stiles, J. Tonnis, Synth. Commun., I , 127 (1971). J. A. Virgilio, E. Heilweil, Org. Prep. Proced. Int., 14, 9 (1982). E. B. Maxted, V. Stone, J . Chem. Soc., 26 (1934). D. Hermeling, H. J. Schafer, Chem. Ber., 121, 1151 (1988). B. B. Allen, B. W. Wyatt, H. R. Henze, J . Am. Chem. SOC., 61, 843 (1939). J. Rebek Jr., D. F. Tai, Y. K. Shue, J . Am. Chem. Soc., 106, 1813 (1984). R. Yamaguchi, M. Bau, N. Kawanisi, Bull. Chem. SOC. Jpn., 61,2909 (1988). J. R. Donaubauer, A. M. Greaves, T. C. McMorris, J . Org. Chem., 49,2833 (1984). S. Hatakeyama, H. Numata, S . Takano, Tetrahedron Lett., 25, 3617 (1984). M. Kocor, M. M. Kabat, J. Wicha, W. Peczynska-Czock, Steroids, 41 (1983). A. J. Birch, D. H. Williamson, Org. React., 24, 1 (1976). W. Strohmeier, M. Lukacs, J . Organomet. Chem., 129, 331 (1977). C. W. Whitehead, J. J. Traverso, H. R. Sullivan, F. J. Marshall, J. Org. Chem., 26,2814 (1961). A. G . Anderson, W. F. Harrison, R. G . Anderson, J . Am. Chem. Soc., 85, 3448 (1963). T. W. Russel, R. C. Hoy, J. C. Cornelius, J . Org. Chem., 37, 3552 (1972). H. Feuer, R. Hannetz,J. Org. Chem., 26, 1061 (1961). I. S. Monakhova, N. S. Smimova, K. I. Karpavichyus, V. G . Kharchenko, Zh. Prikl. Khim., 55, 704 (1982). G . Falsone, U. Wingen, Tetrahedron Lett., 20, 675 (1989).

14.3.4.1.3. Vinyl Functions.

Enol ethers and esters can be hydrogenated over 5% Pd-on-BaSO or SrCO, in ethanol at RT and atm P, but hydrogenolysis is often a concomitant reaction', sensitive to substrate, solvent and catalyst, e.g., triflate of cyclic or steroidal ketones are readily hydrogenated over Pt0;:

wphmpha

Selective hydrogenation of isoflavon, 1, is achieved in almost quantitative yield to isoflavone, 2, in dioxane with Pd-on-carbon, whereas the same reaction conducted in ethanol or aqueous acetone yields a mixture of a- and P-isoflavan-4-01,3, also in quantitative yield, the a-isomer being formed preferentially3:

Ph

3

OH

0 1

0

2

(b)

100%

Platinum favors hydrogenolysis and Pd, Ru, Rh favor hydr~genation~-~, polar solvents favor hydrogenolysis, whereas nonpolar solvent favor hydrogenation7, e.g.':

14.3.4. Hydrogenation of Aliphatic C-C 14.3.4.1. In Sirn le Olefins 14.3.4.1.3. VinyPFunctions.

122

Functions

-x:+y Pd CH,OH

AcO COOEt

COOMe

COOMe

En01 lactones are hydrogenated to the saturated lactone if the bond is not activated to hydrogenolysis by an a-keto or phenyl groupg. Conversion of a keto-activator to the ketal prevents hydrogenolysis. When bicyclic compounds are involved, e.g., in 4, the cis-trans nature of the hydrogenated product can be reversed according the choice of solvent lo:

b0 4

Pd-BaSO, AcOH

cis 60%

0

trans 70%

\____, Pd-BaSO, EtOH

9''so:

However, the stereochemistry of the hydrogenation of enol ethers is determined more rigidly as shown by the highly stereospecific reduction of lactone 5":

H,,5% Rh-AlZO, MeOH

(e)

'

5

97%

Vinylogous urethanes, P-amino-acyclic esters are sensitive to hydrogenolysis. Rhodium-on-alumina'2, 5% Pd-on-carbon are used with success, but at below 100°C'3. Enamines are hydrogenated over Pt catalyst^'^^'^ and on Pd-on-carbon16. A completely stereoselective reduction of lactone 6 is achieved over Rh17.

+Rx

HO

N

0-t-Bu

Rh(AI,O, 5%). EtOAc H,, 2.8 x 105 Pa, RT

'

+Po % ,

HO

(f)

NCOOtBu

6

Enaminones are hydrogenated (Pt-on-carbon) to the amino alcohols with a good stereochemical regulation exercised by the amino group18. The pyrrole nucleus is reduced to pyrolidine derivatives over Pt" or Rh catalyst with stereochemical control as in hydrogenation of pyrrole 7".

14.3. H drogenation Reactions 14.3.4. Lydrogenation of Aliphatic C-C 14.3.4.2. In Conjugated Dienes 0

Functions 0

Rh-5% alumina, AcOH Me0 %COOMe

Me0 % Y O o M e 7

123

H,, 4

X

I d Pa, 4 h, R;

(g)

79%

Asymmetric hydrogenation of enamines is described in 14.3.4.1.5. (J.-L. GRAS)

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

11.

12. 13. 14. 15. 16. 17. 18. 19. 20.

J. B. Jones, D. R. Dods, Can. J . Chem., 65,2397 (1987). V. B. Jigajinni, R. H. Wightman, Tetrahedron Lett., 23, 117 (1982). V. Szabo, E. Antal, Tetrahedron Lett., 1659 (1973). R. Metternich, W. Ludi, Tetrahedron Lett., 29, 3923 (1988). J. S. Panek, M. A. Sparks, J . Org. Chem., 54, 2034 (1989). D. A. Evans, M. M. Momssey, J . Am. Chem. Soc., 106, 3866 (1984). L. H. Knox, R. Villotti, F. A. Kinel, H. J. Ringold, J. Org. Chem., 26, 501 (1961). A. Rosenthal, K. Shudo, J. Org. Chem., 37, 4391 (1972). A. G. Barrett, H. G. Sheth,J. Org. Chem., 48,5017 (1983). K. W. Rosenmund, G. Kositzke, H. Bach, Chem. Ber., 92,494 (1959). G. Stork, S. D. Rychnovsky, J. Am. Chem. Soc., 109, 1564 (1987). Y. Hamada, A. Kawai, T. Shioiri, Tetrahedron Lett., 25, 5409 (1984). R. L. Augustine, R. F. Bellina, A. J. Gustavsen, J . Org.Chem., 33, 1287 (1968). A. B. Shenvi, E. Ciganek, J. Org. Chem., 49, 2942 (1984). G. Fraenkel, J. Gallucci, H. S . Rosenzweig, J. Org. Chem., 54, 677 (1989). T. Hudlicky, G. Seoane, T. C. Lovelace, J. Org. Chem., 53, 2094 (1988). Y. Hamada, A. Kawai, T. Shiori, Tetrahedron Lett., 25, 5409 (1984). Y. Matsumura, J. Fujiwara, K. Maruoka, H. Yarnamoto, J . Am. Chem. SOC.,105,6312 (1983). H. P. Kaiser, J. M. Muchowski, J . Org. Chem., 49,4203 (1984). W. W. Turner, J. Heterocyclic Chem., 23, 317 (1986).

14.3.4.2. In Conjugated Dienes

The hydrogenation of conjugated dienes can take place by addition of a first mole of H,, followed or not by addition of a second mole. Hydrogenation of 1,3-butadiene with a Pd catalyst occurs by stepwise saturation of the double bonds; only butenes are formed until all the butadiene is consumed; then the butenes are hydrogenated to butane'. Thus, controlling the quantity of absorbed H, makes possible the half-hydrogenation of dienes to monoolefins. When complete saturation is required, use of metal catalyst based on Pt or Ir is recommended. Hydrogenation of 1 over Pd-on-CaCO, stops after 1 equiv of H, is absorbed, leaving the 1,Caddition product, 2. Over Ni or better Pt oxide, complete saturation occurs, affording the trans-6P,9-dimethyldecahydronaphthalene derivate, 3, after two stereoselective stepwise 1,2-additions':

2 1 3 Iridium black effects the stereoselective (-98%) complete hydrogenation of 4 to 5, whereas mixtures of isomers are obtained when 4 is hydrogenated with Pd-on-carbon or finely divided Ni3:


14.3. H drogenation Reactions 14.3.4. Lydrogenation of Aliphatic C-C 14.3.4.2. In Conjugated Dienes 0

Functions 0

Rh-5% alumina, AcOH Me0 %COOMe

Me0 % Y O o M e 7

123

H,, 4

X

I d Pa, 4 h, R;

(g)

79%

Asymmetric hydrogenation of enamines is described in 14.3.4.1.5. (J.-L. GRAS)

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

11.

12. 13. 14. 15. 16. 17. 18. 19. 20.

J. B. Jones, D. R. Dods, Can. J . Chem., 65,2397 (1987). V. B. Jigajinni, R. H. Wightman, Tetrahedron Lett., 23, 117 (1982). V. Szabo, E. Antal, Tetrahedron Lett., 1659 (1973). R. Metternich, W. Ludi, Tetrahedron Lett., 29, 3923 (1988). J. S. Panek, M. A. Sparks, J . Org. Chem., 54, 2034 (1989). D. A. Evans, M. M. Momssey, J . Am. Chem. Soc., 106, 3866 (1984). L. H. Knox, R. Villotti, F. A. Kinel, H. J. Ringold, J. Org. Chem., 26, 501 (1961). A. Rosenthal, K. Shudo, J. Org. Chem., 37, 4391 (1972). A. G. Barrett, H. G. Sheth,J. Org. Chem., 48,5017 (1983). K. W. Rosenmund, G. Kositzke, H. Bach, Chem. Ber., 92,494 (1959). G. Stork, S. D. Rychnovsky, J. Am. Chem. Soc., 109, 1564 (1987). Y. Hamada, A. Kawai, T. Shioiri, Tetrahedron Lett., 25, 5409 (1984). R. L. Augustine, R. F. Bellina, A. J. Gustavsen, J . Org.Chem., 33, 1287 (1968). A. B. Shenvi, E. Ciganek, J. Org. Chem., 49, 2942 (1984). G. Fraenkel, J. Gallucci, H. S . Rosenzweig, J. Org. Chem., 54, 677 (1989). T. Hudlicky, G. Seoane, T. C. Lovelace, J. Org. Chem., 53, 2094 (1988). Y. Hamada, A. Kawai, T. Shiori, Tetrahedron Lett., 25, 5409 (1984). Y. Matsumura, J. Fujiwara, K. Maruoka, H. Yarnamoto, J . Am. Chem. SOC.,105,6312 (1983). H. P. Kaiser, J. M. Muchowski, J . Org. Chem., 49,4203 (1984). W. W. Turner, J. Heterocyclic Chem., 23, 317 (1986).

14.3.4.2. In Conjugated Dienes

The hydrogenation of conjugated dienes can take place by addition of a first mole of H,, followed or not by addition of a second mole. Hydrogenation of 1,3-butadiene with a Pd catalyst occurs by stepwise saturation of the double bonds; only butenes are formed until all the butadiene is consumed; then the butenes are hydrogenated to butane'. Thus, controlling the quantity of absorbed H, makes possible the half-hydrogenation of dienes to monoolefins. When complete saturation is required, use of metal catalyst based on Pt or Ir is recommended. Hydrogenation of 1 over Pd-on-CaCO, stops after 1 equiv of H, is absorbed, leaving the 1,Caddition product, 2. Over Ni or better Pt oxide, complete saturation occurs, affording the trans-6P,9-dimethyldecahydronaphthalene derivate, 3, after two stereoselective stepwise 1,2-additions':

2 1 3 Iridium black effects the stereoselective (-98%) complete hydrogenation of 4 to 5, whereas mixtures of isomers are obtained when 4 is hydrogenated with Pd-on-carbon or finely divided Ni3:

124

14.3. Hydrogenation Reactions 14.3.4. Hydrogenationof Aliphatic C-C 14.3.4.2. In Conjugated Dienes

Ir, H2, lo5 Pa EtOH, 25OC

Functions

'

(b)

5

4

Activation of the catalyst by solvent or an acidic additive may be also Hydrogenation of dienone, 6, over Pd-on-carbon in ethanol or acetic acid leads to a cisdecalone':

3

0 H2, Pd-C EtOH or AcOH

Bzo

(c) Bzo

6 Hydrogenation of 7 to either cholestan-7-one, 8, or A*-cholesten-7-one, 9, takes advantage of this dependence of selectivity on the polarity of the solvent. Hydrogenation of 7 over Pd-on-carbon in benzene in presence of a small amount of aqueous KOH, yields after work-up 77% of 9. The same reaction carried in ethanol or ethyl acetate affords 8 in 78% yield8:

& 9

0 77%

t C6H6 Pd-C --

\

\

7

EtOHPd-C or EtOAc

____)

0

&

(4

0

8

78%

The symmetrical 1,5-~yclooctadieneis hydrogenated to cyclooctene over 5% Pdon-CaCO, in methanol, cyclooctene being reduced slowlyg. Complete reduction can be obtained over Rh-on-carbon at a fast and constant rate. However, the problem in conjugated diene hydrogenation is to achieve the selective reduction of only one carbon-carbon double bond. Half-hydrogenation of conjugated diene may be obtained as the result of a property of the catalyst, mostly of the metal. Butadiene and 1,3-pentadiene are quantitatively hydrogenated to the corresponding monoalkenes in the vapor in the presence of a Ni-zinc chromite catalyst". The Pt metals give selectivities for reduction of 1,3-pentadiene to monoolefin: Pd Rh Ru Pt Ir". Selectivity in reduction of dienes depends on the age of the catalyst; with reuse, catalyst activity diminishes, and selectivity increases. Similar results are obtained by deliberate deactivation through additives. Conjugated dienes in mixture with alkenes are selectively hydrogenated to alkenes without reduction of alkenes, both initial and produced, to alkanes using heterogeneous catalysts with Ni, Pd or Pt intercalated in graphite'*. Although it promotes double-bond migration, Pd is many times the most selective catalyst in hydrogenation of conjugated dienes to monoolefins vs. complete saturation, e.g., Pd is the most selective catalyst in partial hydrogenation of 1,3-pentadiene1* and of 1,3-butadiene, whereas other metals give some butane, which becomes the major product of Ir-catalyzed red~ctions'~. Palladium-on-

14.3. Hydrogenation Reactions 14.3.4. Hydrogenation of Aliphatic C-C 14.3.4.2. In Conjugated Dienes

125

Functions

alumina or Pd black selectively reduces cyclopentadiene to cyclopentene whereas Pt black and Pt-on-barium sulfate are non~elective’~. However, when dienes are potential aromatic systems (e.g., 1,3-~yclohexadieneleading to cyclohexane and benzene), Pd may lead to disproportionation. Such reactions are carried on the catalyst obtained by Na borohydride reduction of Ni ~ h l o r i d e ’Palladium ~. chloride, in DMF proved also efficient both toward cyclic and acyclic dienes. Reduction of 1,3-cyclohexadiene gives 100%yield of cyclohexene:

0

H,

NiCI,/NaBH, PdCI,, DMF

, 84-95% 100%

and isoprene affords a mixture of methylbutenes16.

/h\ h /hh pdC:DMF’

+

+

29%

25%

(f)

44%

Heterogeneous catalysts often afford the same mixture of products from 1,2-, 3,4-, 1,Caddition and double-bond migration. The ratio of these isomers varies from metal to metal. Half-reduction of butadiene to 1-butene, cis-2-butene, and trans-2-butene over various catalysts is an example (see Table 1). Isolation of trans-2-butene suggests that the 1,Caddition product is formed by 1,2hydrogenation with subsequent double-bond migration. Platinum is useful when double bond migration is to be avoided”, and 72% of 1-butene is obtained over this catalyst”. Gold on silica, y-alumina, or boehmite, prepared by decomposition at 383-403°C of impregnated chloroauric acid, HAuCl,, is a diene hydrogenation catalyst affording 60% of 1-butenel’. Magnesium oxide obtained by evacuating Mg(OH), at 1100°C for 2 h presents unusual properties as an active catalyst for hydrogenation of 1,3-butadiene to cis-2-butene (100%). When D, is used insted of H,, the D atoms are located at C-1 and C-42O: D2 + DCH,CH=CHCH,D

CH,=CH-CH=CH,

TABLE1. ISOMER DISTRIBUTION FROM CATALYSTS Catalyst

Pd-BaSO,

Pt, - 12°C Au-silica MgO

CH,=CHCHZCH3 (%)

6 72 60

-

MgO

THE

(g)

cis

HALF-REDUCTION OF BUTADIENE OVER VARIOUS

CH-,CH=CHCH, (%) cis 75 9 20 100

CH,CH=CHCH3 (%o)

trans

Reference

19 18 20 -

14 14 15

16

126


Functions

The same discrepancy is found in the abilities of metals to afford one of the other isomers in hydrogenation of cis-penta- 1,3-diene. Half-reduction over Co leads to 90% yield of trans-pent-2-ene, whereas over Cu 70% pent-1-ene is obtained even though the disubstituted double bond is more difficult to hydrogenate than the terminal one? CH,CH=CH-CH=CH, cis

co cu

CH,CH=CHCH,CH, trans 90%

CHz=CHCHzCHzCH3

(h)

70%

Introduction of an aromatic substituent as in cis- and trans- 1-phenylbutadiene orients the reaction over PtOz to quantitative yields of cis- and trans- 1-phenylbutene”. Selective reduction of a conjugated diene within a polyolefinic substrate is synthetically useful. The nature of the olefin, steric hindrance, strain, and functional vicinity play preeminent roles, e.g., hydrogenation of ergosterol, 10, in ethyl acetate or dioxane with finely divided Ni gives A7-5a-ergostene-3/3-olin almost quantitative yield. Addition of dimethylaniline retards hydrogenation of the 22-, 23-bond; in this way selective reduction to A7~22-ergostadiene-3/3-ol can be achieved in high yieldz3:

#

finely divided AcOEt,25T Ni, PhN(CH,),

’ Ho&

(i)

-

H

H

10

11

Haptophilicity of the hydroxyl function may account for a selective hydrogenationz4. A difference of strain is responsible for the preferred reduction of one of two similarly substituted double bonds in lZ2?

& COOMe

Me0

fl

COOMe

H 2 P i F >

\

12

(j)

Me0

90%

The enol conjugated double bond of 13 is selectively reduced with Pd-on-StCO,, while the tetrasubstituted olefin is not affected? OCH3

OCH3 Pd-SrCO,, EtOH/C,H, RT,3 x lo5 Pa

EtO

13

EtO 71%


Functions

127

Ruthenium-on-carbon achieves the selective saturation of the highly functionalized enol lactone, 14, in methanolz7:

ocN--go zp3u;c;, on=. =.=
=

Ph

Ph

Ph

Ph

Ph

Ph

(R)-citronellol

Chiral cyclopentanones and y-butyrolactones are obtained by asymmetric hydrogenation of the corresponding alkylidene derivatives, catalyzed by binap-Ru complexes in 94-98% e.e.56. Olefin geometry does not affect stereochemistry and enantioselectivity.

14.3. Hydrogenation Reactions 14.3.4. Hydrogenation of Aliphatic C-C Functions 14.3.4.5. By Asymmetric Hydrogenation

155

Appropriate substrate-catalyst matching can achieve interesting enantiomer recognition. Chemical kinetic resolution is observed during hydrogenation of 3-substituted itaconate esters with Rh-dipamp catalysts5*,and of various prochiral secondary alcohols with b i n a ~ ~ ~ .

LI

~

MeOH [(R)-binap Ru(OAc),]

’

+

46%, 99% e.

6 % ,

46%, 95% e.

Conjunction of the newly created asymmetric center with an existing one offers applications e.g., high stereoselectivity for production of all diastereomers of chiral dipeptides are achieved by means of asymmetric hydrogenation of dehydrodipeptides that have the proper choice of chiral ligands6’. R-=(

NHCOR’

R”

I

CONH-CH-

H* Rh-L*

COOR’”

’

R” ANHCORf (m) R CONH- $H -COOR’”

I

The configuration of the new asymmetric center is controlled by the catalyst and not by the asymmetric center within the substrate61@-. Catalytic asymmetric hydrogenation of prochiral olefins now competes with biochemical processes. Its limit is the lack of generality. There are rules for matching the proper ligand with its substrate; Z- and E-isomers do not react the same way with the same ligand, and optical yields are sensitive to the functionalities and substituents of the substrate. However, catalytic asymmetric hydrogenations are synthetic reactions that can be run on a large scale employing economical substrate-catalyst levels. Several industrially important compounds are prepared this way: (S)-phenylalanine 2, (S)-dopa 2, dextromethorphan 11, and naproxen 12. Me0 Me0

-0Me 11 Dextromethorphan 99.5%e.e.

12 Naproxen 97% e.e. (J.-L. GRAS)

156

14.3. Hydrogenation Reactions 14.3.4. Hydrogenation of Aliphatic C-C Functions 14.3.4.5. By Asymmetric Hydrogenation

1. Y. Izumi, I. Chibata, T. Itoh, Angew. Chem. Int. Ed. Eng., 17, 176 (1978). 2. Y. Izumi, Adv. Catal., 32 (1983). 3. K. Harada, in Asymmetric Synthesis, Vol. 5 , J . D. Morrison, ed., Academic Press, New York, 1980. 4. H. Brunner, E. Bielmeir, J. Organomet. Chem., 38, 223 (1990). 5 . A. Coma, M. Iglesias, C. del Pino, F. Sanchez, J . Chem. SOC.,Chem. Commun., 1253 (1991). 6. W. Oppolzer, R. J. Mills, M. Reglier, Tetrahedron Lett., 183 (1986). 7. A. Bemardi, 0. Carugo, A. Pasquarello, A. Sidjinov, G. Poli, Tetrahedron, 47, 7357 (1991). 8. J. P. Vigneron, H. Kagan, H. Moreau, Tetrahedron Lett., 5681 (1968). 9. M. Tamura, K. Harada, Bull. Chem. SOC.Jpn., 53, 561 (1980). 10. M. Takasaki, R. Harada, Chem. Lett., 1745 (1984). 11. H. Poisel, U. Schmidt, Chem. Ber., 106, 3408 (1973). 12. B. W. Bycroft, G. R. Lee, J. Chem. SOC.,Chem. Commun., 988 (1975). 13. N. Izumiya, S . Lee, T. Kanmera, H. Aoyagi, J . Am. Chem. SOC.,99, 8346 (1977). 14. L. Homer, Pure Appl. Chem., 52, 843 (1980). 15. V. Caplar, G. Commisso, V. Sunjic, Synthesis, 85 (1981). 16. J. M. Brown, D. Parker, Organometallics, I , 950 (1982). 17. W. S . Knowles, Accts Chem. Rex, 16, 106 (1983). 18. H. B. Kagan, in Asymmetric Synthesis, Vol. 5 , J . D. Morrison, ed., Academic Press, New York, 1985, p. 1. 19. J. Halpem, in Asymmetric Synthesis, Vol. 5 , J. D. Morrison, ed., Academic Press, New York, 1985, p. 41. 20. K. E. Koenig, in AsymmetricSynthesis, Vol. 5 , J. D. Morrison, ed., Academic Press, New York, 1985, p. 7. 21. J. W. ApSimon, Tetrahedron, 42, 5157 (1986). 22. B. Bosnich in Asymmetric Catalysis, NATO AS1 Series, E 103, Nijhoff, Dordrecht, 1986. 23. M. Nogradi, in Stereoselective Synthesis, VCH Verlagsgesellschaft, Weinheim, 1987. 24. H. Brunner, Synthesis, 645 (1988). 25. E. Cesarotti, R. Ugo, R. Vitiello, J. Mol. Catal., 12, 63 (1981). 26. L. 0. Nindakova, F. K. Schmidt, E. I. Klabunovskii, V. N. Sheveleva, V. A. Pavlov, Izv. Akad. Nauk SSR, Ser. Khim., 2621 (1981). 27. S . Takeuchi, Y. Ohgo, Bull. Chem. SOC.Jpn., 54, 2136 (1981). 28. A. Fischli, J. J. Daly, Helv. Chim. Acta, 63, 1628 (1980). 29. U. Matteoli, P. Frediani, M. Bianchi, C. Botteghi, S . Gadiali, J. Mol. Card., 12, 265 (1979). 30. R. Noyori, Chem. SOC.Rev., 18, 187 (1989). 31. W. S . Knowles, M. J. Sabacky, J. Chem. SOC.Chem. Commun., 145 (1968); B. D. Vineyard, W. S . Knowles, M. J. Sabacky, J. Mol. Caral., 19, 159 (1983). 32. L. Homer, H. Siegel, H. Biithe, Angew Chem. Int. Ed. Engl., 7,942 (1968). 33. Y. Ohgo, Y. Natori, S . Takeuchi, J. Yoshimura, Chem. Lett., 1327 (1974). 34. 0. Korpium, R. A. Lewis, J. Chicos, K. Mislow, J . Am. Chem. SOC., 90, 4842 (1968). 35. M. Mikolajczik, Pure Appl. Chem., 52, 959 (1980). 36. B. D. Vineyard, W. S . Knowles, M. J. Sabacky, G. L. Bachmann, D. L. Weinkauff, J. Am. Chem. SOC.,99,5946 (1977). 37. H. B. Kagan, T.-P. Dang, J. Am. Chem. SOC., 94,6429 (1972). 38. M. Inoue, K. Ohta, N. Ishizuka, S . Enomoto, Chem. Pharm. Bull., 31, 3371 (1983). 39. R. Noyori, Chem. SOC.Rev., 18, 187 (1989). 40. N. Yamamoto, M. Murata, T. Morimoto, K. Achiwa, Chem. Pharm. Bull., 39, 1085 (1991). 41. T. Hayashi, M. Kumada,Acc. Chem. Res., 15, 395 (1982). 42. B. D. Zwick, A. M. Arif, A. T. Patton, J. A. Gladysz, Angew. Chem. Int. Ed. Engl., 26, 910 ( 1987). 43. H. Brunner, R. Eder, B. Hammer, U. Klement, J. Organomet. Chem., 394,555 (1990). 44. K. Inoguchi, K. Achiwa, Synlett, 49 (1991). 45. R. Selke, H. Pracejus, J. Mol. Catal., 37, 213 (1986). 46. H. Kawano, Y. Ishii, T. Akariya, M. Saburi, S . Yoshikawa, Y. Ushida, H. Kumobayashi, Tetrahedron Lett., 28, 1905 (1987). 47. T. Ohta, H. Takaya, R. Noyori, Inorg. Chem., 27, 566 (1988). 48. T. Ohta, H. Takaya, M. Kitamura, K. Nagai, R. Noyori, J. Org. Chem., 52, 3174 (1987).

14.3. Hydrogenation Reactions 14.3.5. Hydrogenationof Arenes 14.3.5.1. By Cobalt Catalysts

157

49. H. Takaya, T. Ohta, N. Sayo, H. Kumobayashi, S. Akutagawa, S. Inoue, I. Kasahara, R. Noyori, J . Am. Chem. SOC.,109, 1596 (1987). 50. B. Heiser, E. A. Broger, Y. Crameri, Tetrahedron Asym.,2, 51 (1991). 51. M. J. Burk,J. Am. Chem. Soc., 113, 8518 (1991). 52. T. Hayashi, N. Kawamura, Y. Ito, J . Am. Chem. Soc., 109, 7876 (1987). 53. T. Ohta, H. Takaya, R. Noyori, Tetrahedron Lett., 31, 7189 (1990). 54. W. D. Lubell, M. Kitamura, R. Noyori, Tetrahedron: Asym.,2, 543 (1991). 55. J. J. Bozell, C. E. Vogt, J. Gozum, J. Org. Chem.,56, 2584 (1991). 56. T. Ohta, T. Miyake, N. Seido, H. Kumobayashi, S. Akutagawa, H. Takaya, Tetrahedron Lert., 33, 635 (1992). 57. B. Imperiali, J. W. Zimmerman, Tetrahedron Lett., 29, 5343 (1988). 58. J. M. Brown, Angew. Chem. Int. Ed, Engl., 26, 190 (1987). 59. M. Kitamura, I. Kasahara, K. Manabe, R. Noyori, H. Takaya, J . Org. Chem., 53, 708 (1988). 60. I. Ojima, T. Kogure, N. Yoda, T. Suzuki, M. Yatabe, T. Tanaka, J . Org. Chem., 47, 1329 (1982). 61. S. El-Baba, J. M. Nuzillard, J. C. Poulin, H. B. Kagan, Tetrahedron, 42, 3851 (1986). 62. J. M. Nuzillard, J. C. Poulin, H. B. Kagan, Tetrahedron Lett., 27, 2993 (1986).

14.3.5. Hydrogenation of Arenes Although hydrogenation of aromatic systems requires more energetic conditions than do those of olefinic double bonds, carbocyclic aromatics are readily hydrogenated to the fully saturated product. Controlling hydrogenolysis, selectivity and stereochemistry are major problems connected with arene hydrogenations. The substituents play an important role, and their effects are not the same for all catalysts'. (J.-L. GRAS) 1. R. H. Fish, Aspects Homogeneous Cataly., 7, 65 (1990).

14.3.5.1. By Cobalt Catalysts

Catalytic hydrogenations over Co,(CO), (using H, and CO) or with stoichiometric quantities of preformed hydridocarbonyl complex CoH(CO), are useful for the partial selective reductions of polycyclic aromatic compounds'. Isolated benzene rings are not affected. Naphthalene is reduced to tetralin', at 200°C under a pressure of 20 X lo3 Wa and anthracene to 9,lO-dihydroanthracene (99%). The substituted phenanthrene nucleus is stable under these conditions as illustrated by hydrogenation of perylene 1 and pyrene 2,. Alkanes, benzene, and diethylether are usual solvents for hydrogenations over cobalt-carbonyl catalysts. Vigorous conditions are necessary for Co,(CO), ( 100-2OO0C, 30 X lo3 Wa), but very mild conditions (25"C, 100 kPa) are sufficient with CoH(CO),. Ziegler-type catalysts derived from cobalt 2-ethylhexanoates with AlEt, hydrogenate the aromatic nucleus4, and 2-naphthol 3 to tetrahydronaphthols 4 and 5 over finely divided Co or Ni [equation (a)I5. The ratio 4/5 depends on the solvent and on promotion by various amines. The benzene nucleus is hydrogenated to cyclohexane (> 95%) over a salicylaldehyde complex of Co, which absorb H, to form the bis( salicylaldehydato) (Co(I1) dihydrate (cosal) catalyst6. Cosal can act alone, or with LiAlH,, which behaves as both reductant and support and enhances the activity of the catalyst. Activity decreases with substitution in the ring. Cobalt-acetylacetonates react with LiAlH, to give black



157







157





14.3. Hydrogenation Reactions 14.3.5. Hydrogenation of Arenes 14.3.5.1. By Cobalt Catalysts

158

solutions, which catalyze the hydrogenation of benzene at 30°C and 100 kPa of H, in THF'.

1

72%

LJ

kJ

2

69%

Raney Co or U-Ni-B

3

4

5

The chemoselective hydrogenation of aromatic regions of complex organic molecules is a major synthetic challenge. The soluble complex trihapto-allyltris(trimethylphosphite)cobalt(I), ( T~-C,H,)CO[P(OCH,),],, at RT catalyzes hydrogenation of arenes more readily than alkenes. Thus benzene is reduced to cyclohexane three to four times faster than cyclohexene is reduced to cyclohexane, and twice as fast as hexenes are reduced to hexanes'. Once transfer of the first hydrogen occurs, the arene must remain tightly bound to the metal center until all the hydrogens have been added. Consequently hydrogenation of C6D6 affords the pure all-cis cyclohexane D6'. The exceptional cis-character of the above H, addition is illustrated by formation of cis ring junction of decalin obtained from naphthalene, and by the stereoselective hydrogenation of xylenes and mesitylenes. The first step in the reaction sequence is an h3-C3H, into h'-C,H, conversion. The allyl group is not lost during reaction. The corresponding hydrido-cobalt complex is inactive. The scope of arene hydrogenation using ( q3-C3H5)Co [P(OCH,),], has been extended to benzenes with substituent groups that include R, OR, COOR, and NR,; however, electron-withdrawing groups, e.g., halogen, NO,, and CN, inhibit the reduction". The allyl Co catalysts have a relatively short lifetime which may affect their practical use. The multistepped character of catalytic Co complex reactions reveals mechanistic details on the hydrogenation of arenes' '. (J.-L. GRAS) 1. R. H. Fish, Aspects Homogeneous Cafal.,7, 65 (1990). 2. P. D. Taylor, M. Orchin, J . Org. Chem., 37, 3913 (1972). 3. S. Friedman, S. Metlin, A. Svedi, I. Wender, J . Org. Chem., 24, 1278 (1959).

14.3.Hydrogenation Reactions 14.3.5.Hydro enation of Arenes 14.3.5.2.By uthenium Catalysts

159

#

4. S. J. Lapporte, W. R. Schuett, J . Org. Chem., 28, 1947 (1963).

5. M. Kajitani, Y. Watanabe, Y. Iimura, A. Sugimori, Bull. Chem. SOC.Jpn., 48, 2848 (1975). 6. P. Patnaik, S. Sarkar, Tetrahedron Lerr., 2531 (1977). 7. N. Murugesan, S. Sarkar, Indian J . Chem., 14 A , 107 (1976). 8. E. L. Muetterties, F. J. Hirsekom, J. Am. Chem. SOC., 96, 4063 (1 974). 9. E. L. Muetterties, M. C. Rakowski, F. J. Hirsekom, W. D. Larson, V. J. Basus, F. A. L. Anet, J. Am. Chem. Soc., 97, 1266 (1975). 10. L. S. Stuhl, M. Rakowski DuBois, F. J. Hirsekom, J. R. Bleeke, A. E. Stevens, E. L. Muetterties, J . Am. Chem. SOC.,100,2405 (1978). 1 1 . E. L. Muetterties, J. R. Bleeke, Acct. Chem Res., 12, 325 (1979).

14.3.5.2. By Ruthenium Catalysts

Ruthenium catalysts hydrogenate arenes mostly at elevated T and P without hydrogenolysis. Water promotes Ru catalysis reaching a 10-fold improvement at a weight ratio, H,O:substrate = 1, either with or without solvent’. Ruthenium has been used in homogeneous catalytic hydrogenation of benzenes. The complex ~ f - c , ( c H ~ ) ~ R u - ~ ~ C,(CH,), in which an arene is tetrahapto bonded to Ru is a long-lived homogeneous catalyst sensitive to alkyl substitution in the arene. Hexamethylbenzene is not hydrogenated, showing that the catalyst has no reducible ligands, a critical element for catalyst life. The catalyst is less stereoselective than allylcobalt systems, affording a 9:l cis:fruns ratio in the dimethylcyclohexanes produced from O-xylenes’. Hydrogenation of benzene to cyclohexane occurs under mild conditions on Ru(H)Cl( $-C6Me,)PPh3, another longlived catalyst. No cyclohexadienes or cyclohexene are detected, suggesting that the hydrocarbon remains strongly coordinated to the metal center throughout all hydrogenation steps3. Partial hydrogenation of benzene to cyclohexene is possible on supported Ru catalyst, or on an unsupported Ru catalyst poisonned by FeCl,, FeSO,, TiCl,, or 0;. Ruthenium catalysts allow selective reduction of anthracene derivatives bridged by a 9,lO-ethano group. Hydrogenation of 1 at high P and moderate T stops after reduction of one benzene ring [equation The second ring is not hydrogenated probably because the substrate molecule is bent and only one ring lies flush on the catalyst surface. Haptophilicity - probably allows the full reduction of 2 [equation (b)],.

H,,9.5

&

X 103kPa, 147OC RU-A1203 RU-C

~

OH

& 84%

1 (a) R1,R’ = -COOCO(b)R1,R2 = H

2

6

H,, 28 X Ru-C lo3 kPa, 150°C

’

OH


14.3.Hydrogenation Reactions 14.3.5.Hydro enation of Arenes 14.3.5.2.By uthenium Catalysts

159

#

4. S. J. Lapporte, W. R. Schuett, J . Org. Chem., 28, 1947 (1963).

5. M. Kajitani, Y. Watanabe, Y. Iimura, A. Sugimori, Bull. Chem. SOC.Jpn., 48, 2848 (1975). 6. P. Patnaik, S. Sarkar, Tetrahedron Lerr., 2531 (1977). 7. N. Murugesan, S. Sarkar, Indian J . Chem., 14 A , 107 (1976). 8. E. L. Muetterties, F. J. Hirsekom, J. Am. Chem. SOC., 96, 4063 (1 974). 9. E. L. Muetterties, M. C. Rakowski, F. J. Hirsekom, W. D. Larson, V. J. Basus, F. A. L. Anet, J. Am. Chem. Soc., 97, 1266 (1975). 10. L. S. Stuhl, M. Rakowski DuBois, F. J. Hirsekom, J. R. Bleeke, A. E. Stevens, E. L. Muetterties, J . Am. Chem. SOC.,100,2405 (1978). 1 1 . E. L. Muetterties, J. R. Bleeke, Acct. Chem Res., 12, 325 (1979).

14.3.5.2. By Ruthenium Catalysts

Ruthenium catalysts hydrogenate arenes mostly at elevated T and P without hydrogenolysis. Water promotes Ru catalysis reaching a 10-fold improvement at a weight ratio, H,O:substrate = 1, either with or without solvent’. Ruthenium has been used in homogeneous catalytic hydrogenation of benzenes. The complex ~ f - c , ( c H ~ ) ~ R u - ~ ~ C,(CH,), in which an arene is tetrahapto bonded to Ru is a long-lived homogeneous catalyst sensitive to alkyl substitution in the arene. Hexamethylbenzene is not hydrogenated, showing that the catalyst has no reducible ligands, a critical element for catalyst life. The catalyst is less stereoselective than allylcobalt systems, affording a 9:l cis:fruns ratio in the dimethylcyclohexanes produced from O-xylenes’. Hydrogenation of benzene to cyclohexane occurs under mild conditions on Ru(H)Cl( $-C6Me,)PPh3, another longlived catalyst. No cyclohexadienes or cyclohexene are detected, suggesting that the hydrocarbon remains strongly coordinated to the metal center throughout all hydrogenation steps3. Partial hydrogenation of benzene to cyclohexene is possible on supported Ru catalyst, or on an unsupported Ru catalyst poisonned by FeCl,, FeSO,, TiCl,, or 0;. Ruthenium catalysts allow selective reduction of anthracene derivatives bridged by a 9,lO-ethano group. Hydrogenation of 1 at high P and moderate T stops after reduction of one benzene ring [equation The second ring is not hydrogenated probably because the substrate molecule is bent and only one ring lies flush on the catalyst surface. Haptophilicity - probably allows the full reduction of 2 [equation (b)],.

H,,9.5

&

X 103kPa, 147OC RU-A1203 RU-C

~

OH

& 84%

1 (a) R1,R’ = -COOCO(b)R1,R2 = H

2

6

H,, 28 X Ru-C lo3 kPa, 150°C

’

OH

160

14.3. Hydrogenation Reactions 14.3.5. H dro enation of Arenes 14.3.5.2. b y futhenium Catalysts

The anionic hydride catalyst K [(Ph,P)2(Ph2PC,H4)RuH2]~C,oH~~Et,0 requires only relatively mild conditions ( lOO"C, 600 kPa) to reduce naphthalene to tetrahydronaphthalone and anthracene to tetrahydroanthracene with 98% selectivity. Under the same conditions, isolated aromatic rings are unaffected7. Ring saturation of benzyl compounds carrying oxygen or nitrogen functions without hydrogenolysis is achieved over catalysts, usually at elevated pressures. Hydrogenation of acetophenone over RuO, at 100°C and 7 X lo3 kPa affords 1-cyclohexylethanol in 88% yield'. The production of methylcyclohexylcarbinol from methylphenylcarbinol is best accomplished over Ru hydroxide at 100°C at 10 x lo3 kPa9. The functionalized dihydroquinolone derivative 3 is smoothly hydrogenated to decal014 with negligible hydrogenolysis [equation (c)] over ruthenium-on-carbon". 0

0

Ruthenium dioxide also catalyzes hydrogenation of the amine ring. The optically active ephedrine 5 is reduced to amine 6 over RuO, with no change in rotation [equation (d)] (when compared with reduction over palladium black); P-phenylamines are saturated with difficulty". H,, 7

X

lo3Ha, 90°C

RuO,, EtOH, 40 min

'

5

6

90%

Reduction of phenols and phenyl ethers can occur with limited loss of the oxygen function. Hydrogenation of 4-hydroxyisophtalic acid at IOO"C, in THF, over RuO, affords 65% ring reduction along with 35% hydrogenolysis12.Hydrogenolysis decreases with increased pressure, and excellent results are obtained with Ru-Al,O, (180"C, 14 X lo3kPa)I3. Methylcyclohexanols are obtained quantitatively from the reduction of cresols over ruthenium hydroxide, without solvent, at 80°C and 10 X lo3 kPa14. Ruthenium hydrogenation of estrone, estradiol, or related steroids 7 gives 3 phydroxy-5 a,10 a-estranes 8 as the major products [equation (e)]. Reaction proceeds through a 3-keto intermediate and the stereochemistry of the product at all three new asymmetric centers follows the pattern of cis-rear attack15.

7

8

Ruthenium-on-carbon and ruthenium-on-alumina also are effective for reduction of a phenolic ring in polycyclic compounds. Anilines are hydrogenated under the same con-

14.3. Hydrogenation Reactions 14.3.5. Hydrogenation of Arenes 14.3.5.2. By Ruthenium Catalysts

161

ditions used for phenols, but are less susceptible to hydrogenolysis. Side reactions include reductive hydrolysis leading to carbonyl compounds and reductive coupling leading to secondary amines. Ruthenium catalysts are recommended for hydrogention of anilines both in laboratory and industrial situations16. Product composition is influenced by the solvent and additives. Reduction of aniline over RuO, affords 11.9% dicyclohexylamine in methanol but only 0.3% of coupled product in i~opropanol'~; lithium hydroxide avoids dicyclohexylamine formation'*. In general LiOH effectively promotes aniline hydrogenation over RuO, by accelerating the rate of reduction, by eliminating the inhibiting effect of ammonia. Reduction of anilines to cyclohexylamines over RuO, works successfully on a series of nuclear substituted substrates, at 90-125"C, 8 X lo3 kPa, in alcohols or without solvent". Yields of 92% are obtained in the preparation of diamines such as bis(4aminocyclohexyl) methane, the product being mostly cis& and cis,truns i~orners'~. Phenylenediamines are reduced to the 1,3-diamine (91%) or to the 1,4-diamine (88%) over ruthenium-on-alumina in ethanol. The cis-isomer predominates (70-84%) in a number of solvents and over a range of experimental conditions2'. Synthetic advantages can be taken from some side reactions. Hydrogenation of 3,4-diaminobenzoic acid can lead to a mixture of bicyclic lactams that lack an amino substituent2'. Selective hydrogenation of trisubstituted aniline 9 affords lactame 10, an intermediate in the total synthesis of ibogarnine,,.

*OH

9

10 (J.-L. GRAS)

1. P. N. Rylander, N. Rakoncza, D. Steele, M. Bollinger, Englehardlnd. Tech. Bull.,4,95 (1963). 2. J. W. Johnson, E. L. Muetterties, J. Am. Chem. Soc., 7395 (1977). 3. M. A. Bennett, T. Huang, A. K. Smith, T. W. Tumey, J . Chem. SOC.,Chem. Commun., 582 (1978). 4. C. U. F. Odenbrand, S. T. Lundin, J . Chem. Technol. Biotechnol., 31, 660 (1981). 5 . M. Kolobielski, J . Org. Chem., 28, 1883 (1963). 6. V. L. Mylroie, J. F. Stenberg, Ann. N.Y.Acad. Sci.,214,255 (1973). 7. R. A. Grey, G. P. Pez, A. Wallo, J . Am. Chem. Soc., 102,5948 (1980). 8. M. Freifelder, T. Anderson, Y. H. Ng, V. Papendick, J . Pharm. Sci., 53, 967 (1964). 9. K. Taya, M. Hiramoto, K. Hirota, Sci. Pap. lnst. Phys. Chem. Res. (Jpn), 62, 145 (1968). 10. J. A. Hirsch, G. Schwartzkopf, J. Org. Chem., 39,2044 (1974). 11. M. Freifelder, G. R. Stone, J . Org. Chem., 27, 3568 (1962). 12. P. S. Wharton, C. E. Sundin, D. W. Johnson, H. C. Kluender, J . Org. Chem., 37, 34 (1972). 13. P. L. Omstein, M. B. Amold, N. K. Augenstein, J. W. Paschal, J . Org. Chem.,56,4388 (1991). 14. Y. Takagi, S. Ishii, S. Nishimura, Bull. Chem. SOC.Jpn., 43, 917 (1970). 15. R. T. Rapala, E. Farkas, J . Org. Chem., 23, 1404 (1958). 16. P. N. Rylander, L. Hasbrouck, I. Karpenko, Ann. N.Y.Acad. Sci.,214, 100 (1973). 17. S . Nishimura, T. Shu, T. Hara, Y. Takagi, Bull. Chem. SOC.Jpn., 39, 329 (1966). 18. S. Nishimura, Y. Kono, Y. Otsuki, Y. Fukaya, Bull. Chem. SOC.Jpn., 44,240, (1971). 19. A. E. Barkdoll, D. C. England, H. W. Gray, W. Kirk, Jr., G. M. Whitman, J . Am. Chem. SOC., 75, 1156 (1953).

162

14.3. Hydrogenation Reactions 14.3.5. Hydrogenationof Arenes 14.3.5.3. by Rhodium Catalysts

20. E. F. Litvin, L. K. Freidlin, G . K. Oparina, V. I. Kheifets, V. V. Yakubenok, L. P. Pivonenkova, M. K. Bychkova, Izv. Akad. Nauk SSSR, Ser. Khim., 854 (1973); Chem. Abstr., 79, 31603j (1973). 21. R. L. Augustine, L. A. Vag, J. Org. Chem., 40, 1074 (1975). 22. J. Witte, V. Boekelheide, J . Org. Chem., 37, 2849 (1972).

14.3.5.3. by Rhodium Catalysts

Hydrogenation of carbocyclic aromatic compounds requires only mild conditions over Rh catalysts. Rhodium is an outstandingly active catalyst for reduction of benzene. Catalyst efficiency is influenced by trace materials that act as inhibitors or promoters'. Hydrogen halides are strong inhibitor for reductions in MeOH over Rh-on-carbon or on alumina2. Small amounts of acetic acid promote reduction of aromatics over Rh-onalumina3. In a clean medium, 5% Rh-on-carbon or Rh-on-alumina in MeOH reduces alkyl benzenes at room temperature, under 500 @a4. Reaction is facilitated at higher T or by adding glacial acetic acid. Selective hydrogenation of one phenyl ring in polyphenyl compounds is hardly possible using Rh catalysts, however, the reduction of l-methyl-4-(2,2-diphenyl-2hydroxyethy1)piperazine 1 as the dihydrochloride salt leads to selectively reduced 2. Rhodium catalysts are less sensitive to poisoning by nitrogen compounds than other metal catalysts when camed out on rhodium-on-alumina5:

1

Rhodium is the most active, on a metal weight basis, of platinum metals for reduction of naphthalene to decalin (90% cis-isomer)6. The composition of the stereoisomer mixture is closely related to the catalyst used; Pt gives 74% cis-isomer. Formation of trunsisomers during hydrogenation of disubstituted arenes is accounted for by partial desorption and readsorption of some partially hydrogenated intermediate. More truns-dimethylcyclohexane isomers are formed by Rh than Ru in hydrogenation of xylenes. The percentage of trans-isomer varies from o-xylene (10.8%) to m-xylene (26.3%) to p-xylene (36.4%)6p7. This percentage also depends on the metal concentration, the catalyst carrier, and markedly on temperature. Weaker olefin adsorption relative to the aromatic makes possible detection of olefin intermediates in the hydrogenation of aromatics over Rh. For example, cis-l,3,5-tri-~-butylcyclohexene forms 65% of the total mixture of the reduction of 1,3,5-tri-t-b~tylbenzene~. Preferential ring reduction in the presence of reducible functionality is seen in the hydrogenation of 2-methoxyphenylacetone 3 to saturated cis-ketone 4 over rhodium-onalumina in acetic acid': 5% Rh -A1,0, AcOH, 10 h

3

4

78%


162


20. E. F. Litvin, L. K. Freidlin, G . K. Oparina, V. I. Kheifets, V. V. Yakubenok, L. P. Pivonenkova, M. K. Bychkova, Izv. Akad. Nauk SSSR, Ser. Khim., 854 (1973); Chem. Abstr., 79, 31603j (1973). 21. R. L. Augustine, L. A. Vag, J. Org. Chem., 40, 1074 (1975). 22. J. Witte, V. Boekelheide, J . Org. Chem., 37, 2849 (1972).


Hydrogenation of carbocyclic aromatic compounds requires only mild conditions over Rh catalysts. Rhodium is an outstandingly active catalyst for reduction of benzene. Catalyst efficiency is influenced by trace materials that act as inhibitors or promoters'. Hydrogen halides are strong inhibitor for reductions in MeOH over Rh-on-carbon or on alumina2. Small amounts of acetic acid promote reduction of aromatics over Rh-onalumina3. In a clean medium, 5% Rh-on-carbon or Rh-on-alumina in MeOH reduces alkyl benzenes at room temperature, under 500 @a4. Reaction is facilitated at higher T or by adding glacial acetic acid. Selective hydrogenation of one phenyl ring in polyphenyl compounds is hardly possible using Rh catalysts, however, the reduction of l-methyl-4-(2,2-diphenyl-2hydroxyethy1)piperazine 1 as the dihydrochloride salt leads to selectively reduced 2. Rhodium catalysts are less sensitive to poisoning by nitrogen compounds than other metal catalysts when camed out on rhodium-on-alumina5:

1

Rhodium is the most active, on a metal weight basis, of platinum metals for reduction of naphthalene to decalin (90% cis-isomer)6. The composition of the stereoisomer mixture is closely related to the catalyst used; Pt gives 74% cis-isomer. Formation of trunsisomers during hydrogenation of disubstituted arenes is accounted for by partial desorption and readsorption of some partially hydrogenated intermediate. More truns-dimethylcyclohexane isomers are formed by Rh than Ru in hydrogenation of xylenes. The percentage of trans-isomer varies from o-xylene (10.8%) to m-xylene (26.3%) to p-xylene (36.4%)6p7. This percentage also depends on the metal concentration, the catalyst carrier, and markedly on temperature. Weaker olefin adsorption relative to the aromatic makes possible detection of olefin intermediates in the hydrogenation of aromatics over Rh. For example, cis-l,3,5-tri-~-butylcyclohexene forms 65% of the total mixture of the reduction of 1,3,5-tri-t-b~tylbenzene~. Preferential ring reduction in the presence of reducible functionality is seen in the hydrogenation of 2-methoxyphenylacetone 3 to saturated cis-ketone 4 over rhodium-onalumina in acetic acid': 5% Rh -A1,0, AcOH, 10 h

3

4

78%

14.3. Hydrogenation Reactions 14.3.5. H dro enation of Arenes 14.3.5.3. y R odium Catalysts

163

t7l

Rhodium is used for mild reductions of benzyl alcohols, ethers, or amines to the corresponding cyclohexyl derivatives. Hydrogenolysis of the carbon-oxygen or -nitrogen bond does not occur, and there is no loss of optical activity even if acetic acid is added to maintain the rate of hydrogenation of optically active amines, or mandelic acid derivatives3. Excellent yields obtain from reduction of meso- and d,l-2,3-diphenyl-2,3butanediol over rhodium-on-alumina, without hydrogenolysis to the corresponding dicyclohexyldiolg. Aromatics undergo hydrogenation homogeneously with the [Rh(r)5-C6Me,)C12], complex which requires an added base (NEt,) as cocatalyst. The stereochemistry of reduction is well defined and the all-cis-isomers are the main products. However, some hydrogenolysis of functional groups is observed when the reaction is carried out under 50 X lo3 kPa". Antranilic acid is used as a ligand with Rh to form Rh(I)(N-phenylanthranilate),, a catalyst effective for hydrogenation of benzene under mild conditions (RT, 100 kPa). The catalyst is easily prepared but different to recover. The bidentate ligand has been anchored to polystyrene beads to form 5 and its reaction with [RhC1,.3H,O] followed by reduction with NaBH,, affords a high activity polymer-supported Rh catalyst for hydrogenation of aromatic hydrocarbons' '. The catalyst has considerable tolerance to poisons (e.g., air) and long-term stability. Catalytic activity depends on retention of beads, and is considerably diminished if fragmentation occurs. Benzene is reduced to cyclohexane (99% yield) and naphthalene is reduced to 1,2,3,4-tetrahydronaphthalene (70°C, 7 X lo4 kPa, 3 h)I2.

5

COOH

Rhodium catalysts saturate phenols and phenyl ethers with little hydrogenolysis. They are recommended for the hydrogenation of sensitive compounds. Rhodium-onalumina promotes hydrogenation of gallic acid 6 to all-cis-hexahydrogallic acid 7.l 3 OH / HOf&COOH

H,, 14 X lo3 kPa, 9OoC,8 h Rh-AIzO,

HO

6

OH

'HO&COOH 7

This catalyst reduces cu-naphthol to 1-decal01 (94%), which contains 33% &,cis-isomer. Hydroquinone and resorcinol afford 90 and 85% 1,4-cyclohexanediol and 1,3-cyclohexanediol, re~pectively'~.Hydrogenation of p-hydroxybenzoic acid over Rh-onalumina15to hydroxycyclohexane-4-carboxylicacid is quantitative and 2,2-dimethylbenzofuran affords stereospecifically cis-2,2-dimethyloctahydrobenzofuran in 94% yield16. Rhodium hydroxide is also effective in ring saturation of the hydrogenolysissensitive hydroquinone dimethyl or diphenylether". Use of Rh-on-carbon as catalyst permits hydrogenation of naphthols to decalols [equation (d)] and of tetralones to decalols [equation (e)] without hydrogenolysis of alkoxy substituents".

164


MeowoH v

EtOH Rh-C

OH

Me0

65%

25%

The stereochemical results of phenols and phenol ether hydrogenation vary with catalyst, solvent, and temperature. Rhodium-on-carbon is preferred for increasing the percentage of cis-isomer in the hydrogenation of cresols; the cis-isomer content decreases from 91 to 64% in the temperature range 30-100°C'9. Rh catalysts can lead to partial ring reduction of phenols to ketones, a result sensitive to the reduction medium. Decalols form in excellent yields on reduction of naphthols over 5% Rh-on-alumina in MeOH or EtOH, but a 70% yield of decalone is obtained in acetic acid". This result is attributed to deactivation of the catalyst during reaction. Low-temperature (25°C) hydrogenation of resorcinol over Rh-on-alumina in alkaline solution is a convenient procedure for preparation of 1,3-~yclohexanedione(85% yield)2'. The 3,5-dihydroxyphenylacetic acid 8 is reduced under the same conditions to the corresponding dione in 77% yieldz2. O

Rh-AIZO,

\COOH

v

o

H

(f)

\COOH 77%

8

Rhodium-on-alumina appears to be the catalyst of choice for low T hydrogenation of anilines, particularly alkoxy derivatives, to cycl~hexylamines~~. Dicyclohexylamine formation, which occurs in the hydrogenation of aniline over Rh catalysts without solvent, is increased by an appropriate choice of support: carbon (30%), BaCO, (20%), A1,03 (18%), BaSO, (14%), CaCO, (9%)',. Partially hydrogenated aromatic rings 9 account for secondary amine formation. Isomerization of 9 gives an imine that can interact with a cyclohexylamine to give an addition product 10, followed by direct hydrogenolysis or by loss of NH, and r e d ~ c t i o n ~ ~ .

' D N H 2

%

R\0_..2"D *...a

-+

9

NH3+ R

o

N

NH

H

a

R

2R

n

1

N

NH2

H

(g)

a

R

10

The coupling reaction in aniline hydrogenation generally is decreased by addition of ammonia, but Rh is poisoned by ammonia.

14.3. Hydrogenation Reactions 14.3.5. Hydrogenation of Arenes 14.3.5.4. By Palladium and Platinum Catalysts

165

Ammonium phenylsulfamate 11 may be reduced to ammonium cyclohexylsulfamate over rhodium catalyst, at room temperaturez6. C,HsNHS03NH4

10

5% Rh-AlZO,

H,,200 Ha, RT

C~HI~NHSO~NH~ 85%

(h)

(J.-L. GRAS)

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

R. Egli, C. H. Eugster, Helv. Chim. Acta, 58, 2321 (1975). M. Freifelder, J. Org. Chem., 26, 1835 (1961). Y. Kobayashi, Y . Takemoto, Y. Ito, S . Terashima, Tetrahedron Lett., 31, 3031 (1990). P. Raddatz, H.-F. Radunz, G. Schneider, H. Schwartz, Ang. Chem. Int. Ed. Engl., 27, 426 (1988). M. Freifelder, J . Org. Chem., 29, 979 (1964). P. N. Rylander, D. R. Steele, Engelhard Ind. Tech. Bull., 5 , 113 (1965). R. D. Schuetz, L. R. Caswell, J . Org. Chem., 27, 486 (1962). H. Van Bekkum, H. M. A. Buurmans, G. Van Minnen-Pathuis, B. M. Wepster, Rec. Trav. Chim. Pays-Bas, 88, 779 (1969). J. H. Stocker, J . Org. Chem., 29, 3593 (1964). M. J. Russel, C. White, P. M. Maitlis, J . Chem. SOC.,Chem. Commun., 427 (1977). N. L. Holy, Tetrahedron Lett., 3703 (1977). N. L. Holy, J . Org. Chem., 44, 239 (1979). A. W. Burgstahler, Z. J. Bithos, Org. Synth., 42, 62 (1962). A. I. Meyers, W. M. Beverung, R. Gault, Org. Synth., 51, 103 (1971). T. A. Giudici, T. C. Bruice, J . Org. Chem., 35, 2386 (1970). A. I. Meyers, K. Barburao, J. Heterocycl. Chem., I , 203 (1964). Y. Takagi, T. Naito, S . Nishimura, Bull. Chem. SOC.Jpn., 38, 2119 (1965). K. Chebaane, M. Guyot, D. Molko, Bull. SOC. Chim. Fr., 244 (1975). Y. Takagi, S . Nishimura, K. Hirota, Bull. Chem. SOC.Jpn., 43, 1846 (1970). A. I. Meyers, W. Beverung, G. Garcia-Munoz, J . Org. Chern., 29, 3427 (1964). J. C. Sircar, A. I. Meyers, J . Org. Chem., 30, 3206 (1965). M. Mokotoff, R. C. Cavestri, J . Org. Chem., 39,409 (1974). M. Freifelder, Y. H. Ng, P. F. Helgren, J . Org. Chem., 30, 2485 (1965). P. N. Rylander, L. Hasbrouck, I. Karpenko, Ann. N.Y. Acad. Sci., 214, 100 (1973). H. Greenfield, J . Org. Chem., 29, 3082 (1964). M. Freifelder, B. Meltsner, G. M. Illich, R. M. Robinson, British Patent 882,952 (1961). Chern. Abstr., 56, 127708 (1962).

14.3.5.4. By Palladium and Platinum Catalysts

Although less reactive than Rh or Ru catalysts, Pt under mild conditions' and Pd at relatively elevated T and P can be used for arenes hydrogenation'. Platinum oxide is used for ring saturation but is sensitive to impurities. Traces of Na salts make the catalyst ineffective for reduction of benzene3. This inhibition is circumvented using a strong acid, e.g., HC1, which interacts with the Na component, or by conducting the reaction in acetic acid4. Platinum-on-alumina allows an industrial reduction of benzene to high-purity cyclohexane', and disodium platinum tetrachloride (Na,PtCl,), which is soluble in acetic acid is a homogeneous catalyst for deuterium exchange6.

Palladium hydrogenates arenes at elevated T (>9OoC) and P (>20 X lo3 kPa), but still finds industrial applications such as the preparation of cyclohexanecarboxylic acid, a caprolactam intermediate, by hydrogenation of benzoic acid in quantitative yield7. Water increases the reduction rate over Pd catalysts, and strained rings can be hydrogenated over palladium-on-carbon in alcohol at Rt and P [equation (a)]*. Compound 1 is readily hydrogenated; the less strained compound 2 is inert.



165

Ammonium phenylsulfamate 11 may be reduced to ammonium cyclohexylsulfamate over rhodium catalyst, at room temperaturez6. C,HsNHS03NH4

10

5% Rh-AlZO,

H,,200 Ha, RT

C~HI~NHSO~NH~ 85%

(h)

(J.-L. GRAS)

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

R. Egli, C. H. Eugster, Helv. Chim. Acta, 58, 2321 (1975). M. Freifelder, J. Org. Chem., 26, 1835 (1961). Y. Kobayashi, Y . Takemoto, Y. Ito, S . Terashima, Tetrahedron Lett., 31, 3031 (1990). P. Raddatz, H.-F. Radunz, G. Schneider, H. Schwartz, Ang. Chem. Int. Ed. Engl., 27, 426 (1988). M. Freifelder, J . Org. Chem., 29, 979 (1964). P. N. Rylander, D. R. Steele, Engelhard Ind. Tech. Bull., 5 , 113 (1965). R. D. Schuetz, L. R. Caswell, J . Org. Chem., 27, 486 (1962). H. Van Bekkum, H. M. A. Buurmans, G. Van Minnen-Pathuis, B. M. Wepster, Rec. Trav. Chim. Pays-Bas, 88, 779 (1969). J. H. Stocker, J . Org. Chem., 29, 3593 (1964). M. J. Russel, C. White, P. M. Maitlis, J . Chem. SOC.,Chem. Commun., 427 (1977). N. L. Holy, Tetrahedron Lett., 3703 (1977). N. L. Holy, J . Org. Chem., 44, 239 (1979). A. W. Burgstahler, Z. J. Bithos, Org. Synth., 42, 62 (1962). A. I. Meyers, W. M. Beverung, R. Gault, Org. Synth., 51, 103 (1971). T. A. Giudici, T. C. Bruice, J . Org. Chem., 35, 2386 (1970). A. I. Meyers, K. Barburao, J. Heterocycl. Chem., I , 203 (1964). Y. Takagi, T. Naito, S . Nishimura, Bull. Chem. SOC.Jpn., 38, 2119 (1965). K. Chebaane, M. Guyot, D. Molko, Bull. SOC. Chim. Fr., 244 (1975). Y. Takagi, S . Nishimura, K. Hirota, Bull. Chem. SOC.Jpn., 43, 1846 (1970). A. I. Meyers, W. Beverung, G. Garcia-Munoz, J . Org. Chern., 29, 3427 (1964). J. C. Sircar, A. I. Meyers, J . Org. Chem., 30, 3206 (1965). M. Mokotoff, R. C. Cavestri, J . Org. Chem., 39,409 (1974). M. Freifelder, Y. H. Ng, P. F. Helgren, J . Org. Chem., 30, 2485 (1965). P. N. Rylander, L. Hasbrouck, I. Karpenko, Ann. N.Y. Acad. Sci., 214, 100 (1973). H. Greenfield, J . Org. Chem., 29, 3082 (1964). M. Freifelder, B. Meltsner, G. M. Illich, R. M. Robinson, British Patent 882,952 (1961). Chern. Abstr., 56, 127708 (1962).

14.3.5.4. By Palladium and Platinum Catalysts

Although less reactive than Rh or Ru catalysts, Pt under mild conditions' and Pd at relatively elevated T and P can be used for arenes hydrogenation'. Platinum oxide is used for ring saturation but is sensitive to impurities. Traces of Na salts make the catalyst ineffective for reduction of benzene3. This inhibition is circumvented using a strong acid, e.g., HC1, which interacts with the Na component, or by conducting the reaction in acetic acid4. Platinum-on-alumina allows an industrial reduction of benzene to high-purity cyclohexane', and disodium platinum tetrachloride (Na,PtCl,), which is soluble in acetic acid is a homogeneous catalyst for deuterium exchange6.

Palladium hydrogenates arenes at elevated T (>9OoC) and P (>20 X lo3 kPa), but still finds industrial applications such as the preparation of cyclohexanecarboxylic acid, a caprolactam intermediate, by hydrogenation of benzoic acid in quantitative yield7. Water increases the reduction rate over Pd catalysts, and strained rings can be hydrogenated over palladium-on-carbon in alcohol at Rt and P [equation (a)]*. Compound 1 is readily hydrogenated; the less strained compound 2 is inert.

166

14.3. Hydrogenation Reactions 14.3.5. Hydrogenation of Arenes 14.3.5.4. By Palladium and Platinum Catalysts 5% Pd-C

-' 1

(a)

2

Under forced conditions, dodecahydrotriphenylene 3 is hydrogenated to perhydrotriphenylene 4 over palladium-on-carbon, despite the extensive substitution on the aromatic ring'. Best yields are obtained in saturated hydrocarbon solvents rather than in methanol:

(% +% H,, 9 X lo3 kF'a 200°C, Pd-C

3

(b) 4

Selective reduction of the benzene ring in 5, which contains a hindered nitro group, is possible over 30% palladium-on-carbon in acetic acid":

@J+ " 2 ,QT-7 COOEt

H

COOEt

5

(c)

H

Catalytic hydrogenation of quinolines and isoquinolines usually occurs preferentially in the pyridine ring. Selective hydrogenation in the benzene ring is possible in the presence of HCl at Rt and atm P, over PtO, [equation (d)]ll. If hydrogenation is conducted in strong acid (conc. HCl), the benzene ring is reduced more rapidly; the nitrogen containing ring is not affected''.

isoquinoline

MeOH 87 13 MeOH, 4 N HC1 13 87 HCl 95 Palladium-on-carbon can be used, but the combination PtO,/CF,COOH is fastest. Several analogues of quinoline (2-Me, 6-Me, 3-Me, 8-Me) and acridine 6 [equation (e)] are hydrogenated to the saturated pyridine deri~ative'~.

6

100%


167

This is a synthetic route to apomorphine and morphinan analogues. Palladium is an excellent catalyst for selective reduction of polyaromatic compounds; Pt is used for specific compounds. Biphenyl leads to maximum yields (94%) of phenylcyclohexane over PdO, in methanol at 100°C'4, selectivity which is the consequence of catalyst choice. However, hydrogenation of 9-chloromethyltriptycene 7 to ihe octadecahydro-9-chloromethyltryptycene8 is possible with palladium-on-carbon [equation (e)]. The solvent must contain HCl to prevent loss of the halogen15.

&-A

(f)

IN HC1-EtOH

c1

c1

7 8 Side reactions may be encountered, for instance, hydrogenation of (2.2)metacyclophane 9 over platinum oxide in acetic acid affords the perhydropyrene 10 after carbon-carbon bond formation16.

PtO,, AcOH 25"C, 100 k P :

9

10

Palladium is especially selective for hydrogenation of fused polycyclic arenes. Hydrogenation of naphthalene over Pd yields pure tetralin". Here, naphthalene should be considered a diolefin attached to an aromatic ring; Pd is able to hydrogenate olefins in the presence of aromatics. Hydrogenation of complex fused ring systems takes place at various rings and Pd and Pt behave differently. In the reduction of a polyaromatic series such as pyrene, various benzoanthracenes, and phenanthrenes, a Pd catalyst affords the internal dihydroarene regiosectively, while the analogous reaction with Pt catalysts occurs on the terminal ring to give the tetrahydroarene, as illustrated in equation (h)I8. Palladium regiospecifically reduces the electron-rich K-region bond (a bond excision that leaves an intact aromatic system). These reductions compliment the Li in liquid NH, reductions (Birch reductions) that generally provide different hydroaromatic products.

95%

97%

Palladium promotes the facile hydrogenolysis of benzyl compounds carrying oxygen [equation (i)]19 or nitrogen functions [equation ( j ) ] * O with a very small tendency to saturate the aromatic ring.

168


CH3

I Ph-C-COOEt I

CH3 Pd-BaSO, EtOH, NEt,

I

Ph-CH-COOEt

(9

Ph-CH,-NHCOR3

0)

OCOPh

NR~R~

I

Ph-CH

-NHCOR3

Pd dioxane

Platinum accomplishes ring saturation of benzyl oxygen functions without hydrogenolysis, although less effectively than Rh or Ru". Platinum oxide and acetic acid, used either as promoter or as solvent, is commonly used for the reduction of benzyl alcohol or alkylphenylcarbinols to saturated carbinols22.p-Aminomethylbenzoic acid is quantitatively hydrogenated to the corresponding saturated amino acid under these conditionsz3. Catalytic hydrogenation of phenols or phenyl ethers over Pd or Pt may occur differently: hydrogenolysis without ring reduction, hydrogenolysis and ring saturation, ring saturation without hydrogenolysis, and partial ring reduction. Clean loss of oxygen without reduction of the aromatic ring does not occur on phenols but is observed on phenyl ethers2, and is avoided under mild conditions. Platinum is very efficient for ring reduction along with hydrogenolysis of phenolic corn pound^^^. The hydrogenolysis rate is enhanced by acids and an increase in the T, e.g., in the reduction of hydroxybenzenesZ6over PtO,. Palladium generally gives much less hydrogenolysis products, but it is less satisfactory than Rh and Ru because of the elevated T or P required to maintain satisfactory hydrogenation rates. Strontium carbonate, BaCO,, and charcoal are the most common carriers used; structural features may contribute to selectively under less vigorous conditions2':

H,. 100 P a . 10% Pd-C AcOH

OMe

oMe

80%

The ability of Pd to avoid hydrogenolysis is attributed to the absence of ally1 or homoallyl ether intermediate formation. A dihydro intermediate selectively hydrogenated to the enol ether, a function that is not susceptible to hydrogenolysis over Pd2*. An attractive property of Pd is its ability to catalyze the partial reduction of phenols to cyclohexanones. This is important industrially, and many adjustments have been made on the choice of carrier and promoter, to limit overhydrogenation. Palladium-on-BaCO,, palladium chloride-on-carbon, palladi~m-on-CaCO,~~, palladium-on-carbon, and palladiurn-~n-alumina~~ have been used, but the last two generally are satisfactory. Phenol can be reduced to cyclohexanone over palladium hydroxide-on-BaSO,, in dilute HC13', but also in high yields over palladium-on-carbon in presence of sodium salts32.The sodium quantity must be controled to avoid overreduction to cyclohexanol. Hydrogenation of 1&dihydroxynaphthalene catalyzed by palladium-on-carbon leads to

14.3.Hydrogenation Reactions 14.3.5.Hydrogenationof Arenes 14.3.5.4.By Palladium and Platinum Catalysts

& $ $

169

8-hydroxy-1-tetralone [equation (l)]33.Likewise, 2,4-dimethylphenol is reduced to cis-

truns-2,4-dimethylcyclohexanonein a 7:93 ratio [equation (m)]34. Pd-C,

(1)

Pd-C,

Platinum catalysts have been used for hydrogenation of anilines to cyclohexylamines but are now surpassed by Rh and Ru catalysts. The Pt and Pd catalysts are recommended to promote reductive coupling or reductive hydrolysis. Hydrogenation of aniline in acetic acid over palladium-on-carbon affords high yields of N-phenyl-cyclohexyl-amine,and formation of the coupled product is favored as the temperature is increased slightly35. Palladium induces reductive hydrolysis of aniline derivatives to cyclohexanone or cyclohexanol. The course of hydrolytic cleavage follows the scheme of equation (n) and it is facilitated by nitrogen atom sub~titution~~. Hydrolytic cleavage can be avoided by use of nonaqueous solvents.

Hydrogenation of N,N-dimethylaniline in dilute aqueous HCl produces cyclohexanone and cyclohexanol (90%yield). Reductive hydrolysis of anilines that are properly substituted with a chiral ligand on the nitrogen atom affords chiral cyclohexanones with 30% optical yield3':

q p

H

*

p HClO,, PdO-TiO, 22T'

Qo

(0)

OH (J.-L. GRAS)

1. 2. 3. 4. 5. 6. 7. 8. 9.

M. Minabe, K. Watanabe, Y. Ayabe, M. Yoshida, T. Toda, J . Org. Chem., 52, 1745 (1987). H. Greenfield, Ann. N.Y. Acad. Sci., 214, 233 (1973). C. W. Keenan, B. W. Giesemann, H. A. Smith, J . Am. Chem. Soc., 76, 229 (1954). A. P. Phillips, J. Mentha, J . Am. Chem. Soc., 78, 140 (1956). J. W. Teter, U.S.Patent 2,898,387 (1959); Chem. Absrr., 54, 10410~(1959). G. E. Calf, J. L. Gamett, J . Chem. SOC., Chem. Commun., 373 (1969). M. Tavema, M. Chita, Hydrocarbon Process, 137 (1970). H. Rapoport, G . Smolinsky, J . Am. Chem. Soc., 82, 1171 (1960). M. Farina, G. Audisio, Tetrahedron, 26, 1827 (1970).

170 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

14.3. Hydrogenation Reactions 14.3.5. Hydrogenation of Arenes 14.3.5.5. By Miscellaneous Catalysts

D. V. Young, H. R. Snyder, J . Am. Chem. SOC.,83, 3160 (1961). J, Z. Ginos, J . Org. Chem., 40, 1191 (1975). F. W. Vierhapper, E. L. Eliel, J . Am. Chem. SOC., 96, 2256 (1974). F. W. Vierhapper, E. L. Eliel, J . Org. Chem., 40, 2729 (1975). P. N. Rylander, in Catalytic Hydrogenation in Organic Syntheses, Academic Press, New York, 1979, p. 180. V. L. Mylroie, J. F. Stenberg, Ann. N.Y. Acad. Sci., 214, 255 (1973). E. Langer, H. Lehner, Monatsch. Chem., 104, 1484 (1973). F. Zymalskowski, T. Schuster, H. Scherer, Arch. Pharm. Weinheim, 302, 272 (1969). A. W. Weitkamp, J . Catal., 6,431 (1966). P. P. Fu, H. M. Lee, R, G. Harvey, J . Org. Chem., 45, 2797 (1980). N. Sakura, K. Ito, M. Sekiya, Chem. Pharm. Bull., 20, 1156 (1972). Y. Ichinohe, H. Ito, Bull. Chem. SOC.Jpn., 37, 887 (1964). S. Nishimura, Bull. Chem. SOC.Jpn., 32, 1155 (1959). M. Levine, R. Sedlecky,J. Org. Chem., 24, 115 (1959). C. F. Barfknecht, R. V. Smith, V. V. Reif, Can. J. Chem., 48, 2128 (1970). W. S. Johnson, E. R. Rogier, J. Ackerman, J. Am. Chem. SOC., 78, 6322 (1956). H. A. Smith, B. L. Stump, J . Am. Chem. SOC., 83,2739 (1961). A. W. Schrecker, J. L. Hartwell, J . Am. Chem. SOC., 75,5917 (1953). S . Nishimura, M. Uramoto, T. Watanabe, Bull. Chem. SOC.Jpn., 45, 216 (1972). G. K. Oparina, M. P. Chemyshova, V. I. Kheifets, S . S. Gluzman, Zh. Vses. Khim. Ova., 18, 346 (1973); Chem. Abstr., 79,77778~(1973). G. D. Lyubarskii, N. E. Buyanova, I. D. Ratner, M. M. Strelets,Kinet. Katal., 14, 1020 (1973). R. Kuhn, H. J. Haas, Ann. Chem., 611,57 (1958). R. J. Duggan, E. J. Murray, L. 0. Winstrom, US.Patent 3,076,810 (1963); Chem. Abstr., 59,

268111 (1963).

I. A. Kaye, R. S . Matthews, J. Org. Chem., 28, 325 (1963). F. Johnson, N. A. Starkovsky, A. C. Paton, A. A. Carlson, J. Am. Chem. SOC., 86, 118 (1964). K. Ikedate, S. Suzuki, Nippon Kagaku Zasshi, 90,91 (1969); Chem. Abstr., 70,96287t (1969). H. H. Baer, F. Kienzle, J . Org. Chem., 34, 3848 (1969). R. Kuhn, H. E. Driesen, H. J. Haas, Justus Liebigs Ann. Chem., 718, 78 (1968).

14.3.5.5. By Miscellaneous Catalysts Other metals have been tested for the catalytic hydrogenation of aromatic compounds. Finely divided Ni (Raney Ni) is useful for the reduction of the benzene ring of benzyl amine without cleavage of the benzyl-nitrogen bond, although hydrogenation of benzene is more difficult and many functional groups will reduce first. Alloying effects from Cu allows benzene hydrogenation over Ni catalysts at both low and high T'. Resorcinols are hydrogenated in alkaline solution at high P in the presence of finely divided Ni (after this has been freed of aluminum), to afford 1,3-cyclohe~anediones~. Phloroglycinol is totally hydrogenated to 1,3,5-cyclohexane trio1 in ethanol [5OoC,300 kPa)3. Hydrogenation of phenanthrene [equation (a)] and anthracene [equation (b)] over finely divided Ni results in reduction of the central aromatic nucleus4. Finely divided-Ni, EtOH H,, 7 X lo3 kPa, 50°C

Finely divided Ni, EtOH H,,7 X 103kPa,50°C

'


170 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

14.3. Hydrogenation Reactions 14.3.5. Hydrogenation of Arenes 14.3.5.5. By Miscellaneous Catalysts

D. V. Young, H. R. Snyder, J . Am. Chem. SOC.,83, 3160 (1961). J, Z. Ginos, J . Org. Chem., 40, 1191 (1975). F. W. Vierhapper, E. L. Eliel, J . Am. Chem. SOC., 96, 2256 (1974). F. W. Vierhapper, E. L. Eliel, J . Org. Chem., 40, 2729 (1975). P. N. Rylander, in Catalytic Hydrogenation in Organic Syntheses, Academic Press, New York, 1979, p. 180. V. L. Mylroie, J. F. Stenberg, Ann. N.Y. Acad. Sci., 214, 255 (1973). E. Langer, H. Lehner, Monatsch. Chem., 104, 1484 (1973). F. Zymalskowski, T. Schuster, H. Scherer, Arch. Pharm. Weinheim, 302, 272 (1969). A. W. Weitkamp, J . Catal., 6,431 (1966). P. P. Fu, H. M. Lee, R, G. Harvey, J . Org. Chem., 45, 2797 (1980). N. Sakura, K. Ito, M. Sekiya, Chem. Pharm. Bull., 20, 1156 (1972). Y. Ichinohe, H. Ito, Bull. Chem. SOC.Jpn., 37, 887 (1964). S. Nishimura, Bull. Chem. SOC.Jpn., 32, 1155 (1959). M. Levine, R. Sedlecky,J. Org. Chem., 24, 115 (1959). C. F. Barfknecht, R. V. Smith, V. V. Reif, Can. J. Chem., 48, 2128 (1970). W. S. Johnson, E. R. Rogier, J. Ackerman, J. Am. Chem. SOC., 78, 6322 (1956). H. A. Smith, B. L. Stump, J . Am. Chem. SOC., 83,2739 (1961). A. W. Schrecker, J. L. Hartwell, J . Am. Chem. SOC., 75,5917 (1953). S . Nishimura, M. Uramoto, T. Watanabe, Bull. Chem. SOC.Jpn., 45, 216 (1972). G. K. Oparina, M. P. Chemyshova, V. I. Kheifets, S . S. Gluzman, Zh. Vses. Khim. Ova., 18, 346 (1973); Chem. Abstr., 79,77778~(1973). G. D. Lyubarskii, N. E. Buyanova, I. D. Ratner, M. M. Strelets,Kinet. Katal., 14, 1020 (1973). R. Kuhn, H. J. Haas, Ann. Chem., 611,57 (1958). R. J. Duggan, E. J. Murray, L. 0. Winstrom, US.Patent 3,076,810 (1963); Chem. Abstr., 59,

268111 (1963).

I. A. Kaye, R. S . Matthews, J. Org. Chem., 28, 325 (1963). F. Johnson, N. A. Starkovsky, A. C. Paton, A. A. Carlson, J. Am. Chem. SOC., 86, 118 (1964). K. Ikedate, S. Suzuki, Nippon Kagaku Zasshi, 90,91 (1969); Chem. Abstr., 70,96287t (1969). H. H. Baer, F. Kienzle, J . Org. Chem., 34, 3848 (1969). R. Kuhn, H. E. Driesen, H. J. Haas, Justus Liebigs Ann. Chem., 718, 78 (1968).

14.3.5.5. By Miscellaneous Catalysts Other metals have been tested for the catalytic hydrogenation of aromatic compounds. Finely divided Ni (Raney Ni) is useful for the reduction of the benzene ring of benzyl amine without cleavage of the benzyl-nitrogen bond, although hydrogenation of benzene is more difficult and many functional groups will reduce first. Alloying effects from Cu allows benzene hydrogenation over Ni catalysts at both low and high T'. Resorcinols are hydrogenated in alkaline solution at high P in the presence of finely divided Ni (after this has been freed of aluminum), to afford 1,3-cyclohe~anediones~. Phloroglycinol is totally hydrogenated to 1,3,5-cyclohexane trio1 in ethanol [5OoC,300 kPa)3. Hydrogenation of phenanthrene [equation (a)] and anthracene [equation (b)] over finely divided Ni results in reduction of the central aromatic nucleus4. Finely divided-Ni, EtOH H,, 7 X lo3 kPa, 50°C

Finely divided Ni, EtOH H,,7 X 103kPa,50°C

'

14.3. Hydrogenation Reactions 14.3.5. Hydro enation of Arenes 14.3.5.5. By discellaneous Catalysts

171

Partial hydrogenation of substituted naphthalene 1 finely divided Ni yields the phenyl ether 2 under slightly acidic conditions [equation (c)] and yields the alkyl ether 3 in neutral solvent [equation (d)]?

woCH3 Finely divided Ni EtOH/AcOH

2

1 Finely divided Ni EtOH

'

(c)

95%

WoMe 3

70%

Copper chromite is effective for saturation of the reactive 9,lO-double bond of phenanthrene6; finely divided Cu selectively hydrogenates 9-aminoalkylanthracenes to the 9,lO-dihydro derivatives'. Although rarely used, Ir catalysts compare to Pt in promoting hydrogenolysis of the aryl-oxygen bond during hydrogenation of phenols'. Hydrogenation of o-xylene gives only cis products':

98.6% Synergistic effects appear in hydrogenation of aromatic compounds when two Pt group metals are used together. A Pt catalyst to which 1% of Pd is added displays enhanced activity allowing hydrogenation of 1,2,3,4-tetrahydroacridinereadily to octahydroacridine (the reaction is very slow over platinum-on-carbon)". Rhodium-PtO is particularly effective for hydrogenation of aromatic ring with no hydrogenolysis of C-0 bond". This reduction converts phenol 4 to lactone 5 with a cis ring fusion as the major product in a total synthesis of the sesquiterpane 9-iso~yanopupukaenane:'~ H2, Rho-F'tO AcOH,H+

-

A

4

'

-

A

5

37%

Maximal activity and selectivity depend on the nature and on the ratio of the mixed metals, and even on the solvent. Mixed Rh-Pt catalysts are superior to PtO alone for reduction of aniline to cyclohexylamine (92% yield) in ethanol. Addition of acetic acid to the hydrogenation mixture affords 75% cyclohexylamine along with 12% dicyclohexylamine' I . Potassium-on-alumina or a lamellar K-graphite catalyst promotes total hydrogenation of benzene rings13. Carbocyclic and heterocyclic aryl rings are reducible with

172

14.3.6. Hydrogenation of C=O Functions 14.3.6.1. In Aldehydes 14.3.6.1.1. Saturated Aliphatic Aldehydes.

Ni-A1 alloy under vigorous conditions that allow partial reduction of naphthalene and ~henanthrene'~. (J.-L. GRAS) 1. G. A. Martin, J. A. Dalmon, J. Catal., 75, 233 (1982). 2. H. J. Teuber, D. Cornelius, U. Wolche, Justus Liebigs Ann. Chem., 696, 116 (1966). 3. K. Pradad, 0. Repic, Tetrahedron Lett., 25, 2435 (1984). 4. R. P. Linstead, W. E. Doering, S. B. Davis, P. Levine, R. R. Whetstone, J . Am. Chem. SOC., 64,2022 (1942). 5. G. Stork, J . Am. Chem. SOC., 69, 576 (1947). 6. D. D. Phillips, Org. Synth., 4, 313 (1963). 7. D. W. Blackburn, J. J. Mylnarski, Ann. N.Y. Acad. Sci., 214, 158 (1973). 8. Y. Takagi, S.Nishimura, K. Hirota, Bull. Chem. SOC.Jpn., 43, 1846 (1970). 9. S. Nishimura, F. Mochizuki, S. Kobayakawa, Bull. Chem. SOC.Jpn., 43, 1919 (1970). 10. E. Hayashi, T. Nagao, Yakugaku Zasshi, 84, 198 (1964). 11. S. Nishimura, H. Taguchi, Bull. Chem. SOC.Jpn., 36, 873 (1963). 12. E. J. Corey, M. Behforouz, M. Ishiguro, J. Am. Chem. SOC., 101, 1608 (1979). 13. M. Ichikawa, Y. Inoue,K. Tamaru, J . Chem. SOC., Chem. Comm., 928 (1972). 14. L. K. Keefer, G. Lunn, Chem. Rev.,89, 459 (1989).

14.3.6. Hydrogenation of C=O Functions 14.3.6.1. In Aldehydes 14.3.6.1 .l. Saturated Allphatlc Aldehydes.

Aldehydes are hydrogenated to the corresponding carbinols with varying ease depending on their environment. Saturated unactivated aliphatic aldehydes are reduced over various heterogeneous or homogeneous catalysts. Reaction is usually slow, even sluggish, but an increase in the T and P or use of a promoter is often satisfactory. Finely divided Ni (Raney Ni, active forms W6 and W7) catalyzes this reduction quite effectively at room T and atmospheric P. Condensation of the carbonyl compound may be caused by the base present in the catalyst, but is avoided at elevated P. @-Aminoaldehyde hydrochlorides are conveniently hydrogenated (90% yield) over finely divided Ni (W2 Raney Ni) in aqueous solution, without amine hydrogenolysis' :

Et,NCH,C(CH,),CH,OH

. HC1

(a)

Nickel is used for reduction of dextrose to sorbital on industrial scale' and of dialdehyde~~.~. Platinum catalysts are used widely on laboratory and industrial scales. Platinum oxides or the supported platinum-on-carbon is preferred for reasons of economy. They quickly deactivate but are regenerated by shaking the reaction mixture with air (nitrogen has no effect). The improvement brought by periodic oxygen reactivation is attributed to reconstitution of the catalyst to its original form by restoring the oxide film and to removing accumulated poisons6. Certain metallic salts promote complete hydrogenation of aldehydes over platinum catalysts, particularly platinum oxide. As little as 0.005 mol of FeC1, per mole of platinum


172


Ni-A1 alloy under vigorous conditions that allow partial reduction of naphthalene and ~henanthrene'~. (J.-L. GRAS) 1. G. A. Martin, J. A. Dalmon, J. Catal., 75, 233 (1982). 2. H. J. Teuber, D. Cornelius, U. Wolche, Justus Liebigs Ann. Chem., 696, 116 (1966). 3. K. Pradad, 0. Repic, Tetrahedron Lett., 25, 2435 (1984). 4. R. P. Linstead, W. E. Doering, S. B. Davis, P. Levine, R. R. Whetstone, J . Am. Chem. SOC., 64,2022 (1942). 5. G. Stork, J . Am. Chem. SOC., 69, 576 (1947). 6. D. D. Phillips, Org. Synth., 4, 313 (1963). 7. D. W. Blackburn, J. J. Mylnarski, Ann. N.Y. Acad. Sci., 214, 158 (1973). 8. Y. Takagi, S.Nishimura, K. Hirota, Bull. Chem. SOC.Jpn., 43, 1846 (1970). 9. S. Nishimura, F. Mochizuki, S. Kobayakawa, Bull. Chem. SOC.Jpn., 43, 1919 (1970). 10. E. Hayashi, T. Nagao, Yakugaku Zasshi, 84, 198 (1964). 11. S. Nishimura, H. Taguchi, Bull. Chem. SOC.Jpn., 36, 873 (1963). 12. E. J. Corey, M. Behforouz, M. Ishiguro, J. Am. Chem. SOC., 101, 1608 (1979). 13. M. Ichikawa, Y. Inoue,K. Tamaru, J . Chem. SOC., Chem. Comm., 928 (1972). 14. L. K. Keefer, G. Lunn, Chem. Rev.,89, 459 (1989).

14.3.6. Hydrogenation of C=O Functions 14.3.6.1. In Aldehydes 14.3.6.1 .l. Saturated Allphatlc Aldehydes.

Aldehydes are hydrogenated to the corresponding carbinols with varying ease depending on their environment. Saturated unactivated aliphatic aldehydes are reduced over various heterogeneous or homogeneous catalysts. Reaction is usually slow, even sluggish, but an increase in the T and P or use of a promoter is often satisfactory. Finely divided Ni (Raney Ni, active forms W6 and W7) catalyzes this reduction quite effectively at room T and atmospheric P. Condensation of the carbonyl compound may be caused by the base present in the catalyst, but is avoided at elevated P. @-Aminoaldehyde hydrochlorides are conveniently hydrogenated (90% yield) over finely divided Ni (W2 Raney Ni) in aqueous solution, without amine hydrogenolysis' :

Et,NCH,C(CH,),CH,OH

. HC1

(a)

Nickel is used for reduction of dextrose to sorbital on industrial scale' and of dialdehyde~~.~. Platinum catalysts are used widely on laboratory and industrial scales. Platinum oxides or the supported platinum-on-carbon is preferred for reasons of economy. They quickly deactivate but are regenerated by shaking the reaction mixture with air (nitrogen has no effect). The improvement brought by periodic oxygen reactivation is attributed to reconstitution of the catalyst to its original form by restoring the oxide film and to removing accumulated poisons6. Certain metallic salts promote complete hydrogenation of aldehydes over platinum catalysts, particularly platinum oxide. As little as 0.005 mol of FeC1, per mole of platinum


173

oxide can be used7. An equivalent of SnCl, per atom of Pt affords the most effective of several salts tested'. These additives serve several functions such as inhibition of reduction of active PtO, to an inactive lower oxidation state, even to the metal, and to prevention of catalyst coagulation during slow reductions, by accelerating the hydrogenation rate9. Despite some success, saturated aliphatic aldehydes are not easily reduced over Pd catalysts, and indeed these are used in hydrogenations in which the aldehydic group should be preserved. Interestingly, 5% palladium-on-carbon catalyzes the selective reduction of p-cyanopropionaldehyde to the cyanoalcohol, a precursor of butyrolactone": NCCH,CH,CHO

H,

5% Pd-C, HZO

NCCH,CH,CH,OH

Copper chromite catalyzes reduction of aliphatic aldehydes to the alcohol, and Ru-oncarbon appears especially effective in aqueous medium. Carbohydrates are very conveniently reduced and polyhydric alcohols are obtained almost quantitatively from polysaccharides, after acid hydrolysis (H,PO,, H,S04) and immediate Ru catalyzed hydrogenation of the resulting aldehydes' . These reductions use elevated T (100-180°C) and P (3.5 X lo3-14 X lo3 Wa). Aliphatic aldehydes also can be reduced over various homogeneous catalysts involving Co, Ir, Rh, and Ru, without particular advantage, except the reaction conditions are milder. Saturated alcohols are obtained from a,p-unsaturated or saturated aldehydes in up to 90% yields over Co,(CO),, but at high T (180"C)'2. Trihydrid~tris(triphenylphosphine)iridium(III)'~in combination with acetic acid affords a complex that is an active catalyst for reduction of aldehydes, but not k e t ~ n e s ' ~ . Cationic Rh complexes effect homogeneous catalytic hydrogenation of aldehydes with a catalytic activity depending on the structure of the phosphorus l i g a n d ~ ' The ~ . dihydride complex of [Rh(NBD)(PEt,),] 'C10,- is stable and active toward aliphatic and aromatic aldehydes. In combination with fully alkylated mono- and diphosphine ligands, the hydrogenation of aldehydes to the corresponding alcohols occurs quantitatively within 5 rnin'?

'

H,, 5 min

R-CHO

[Rh(L-L)(NBD)] C 10, NBD = norbomadiene L-L = PiPr,-(CH,)-PiPr,

'

R-CHZOH n = 3,4

(c)

Several Ru complexes under mild conditions catalyze reduction of aldehydes to the corresponding alcohols. Linear and branched aldehydes are effectively hydrogenated by molecular H, in the presence of RuCl,(CO)~[P(Ph),], or RuCl,(PPh,). High catalyst activities and turnover numbers up to 95,000 are observed, with yields up to 99%17.The most convenient catalyst precursor is RuHCl(CO)(PPh,) in toluene solutions [equation (d)]. Water and acetic acid accelerate reaction without producing byproducts' *: H,, 3 X 10' @a, toluene, 80°C

CH3CH2CH0

RuHCl(CO)(PPh,),

> CH3CH,CH,0H

Minimal side reactions are observed during hydrogenation of aldehydes. However, aldehyde reductions carried out in methanol or ethanol may afford acetals that are

174

14.3.6. Hydrogenation of C=O Functions 14.3.6.1. In Aldehydes 14.3.6.1.2. Aromatic Aldehydes.

hydrogenolyzed to the corresponding ether. Palladium promotes acetal formation. PtO, in alcohol containing dry HCl leads to ether formation (93% yield) after hydrogenolysis of intermediate hemiketal~'~: (CH,),CHCOCH,

Hz' Ptoz > CH,OH, 2.5 M HCl

(CH,),CHC(CH,)OCH,

a-Hydroxymethylene ketones undergo hydrogenolysis during reduction over Pd-oncarbon, affording the selective introduction of a methyl group in steroids2': OH

(J.-L. GRAS)

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

W. Wenner, J. Org. Chem., 15, 301 (1950). L. W. Wright, Chemtech., 42 (1974). E. H. Pryde, C. M. Thierfelder, J. C. Cowan, J . Am. Chem. SOC.,53, 90 (1976). J. A. Schreifels, P. C. Maybury, W. E. Swartz, Jr., J. Org. Chem., 46, 1263 (1981). W. Sucrow, M. Slopianka, P. P. Caldeira, Chem. Ber. 108, 1101 (1975). N. E. Hoffman, A. T. Kanakkanatt, R. F. Schneider, J . Org. Chem., 27, 2687 (1962). W. H. Carothers, R. Adams, J. Am. Chem. SOC.,45, 1071 (1923). P. N. Rylander, J. Kaplan, Engelhardtlnd. Tech. Bull., 2 , 48 (1961); Chem. Abstr., 56, 2918 d (1961). R. Baltzly, J . Org. Chem., 41, 933 (1976). T. Komatsu, R. Iwanaga, J. Kato, U.S.Patent 3,141,895, (1964). Jpn. Patent I. 3421 (1964); Chem. Abstr., 6 I , 2977 b (1964). V. I. Sharkov, Angew. Chem. Int., Ed. Engl. 2,405 (1963). G. L. Aldridge, H. B. Jonassen, J . Am. Chem. SOC., 85, 886 (1963). R. S . Coffey, J. Chem. Soc., Chem. Commun., 923 (1967). W. Strohmeier, H. Steigenvald, J. Organomet. Chem., 129, C 43 (1977). H. Fujitsu, E. Matsumura, K. Takeshita, I. Mochida, J . Org. Chem., 46, 5353 (1981). K. Tani, K. Suwa, E. Tanigawa, T. Yoshida, T. Okano, S . Otsuka, Chem. Lett., 261 (1982). W. Strohmeier, L. Weigelt, J. Organomet. Chem., 145, 189 (1978). R. A. Sanchez-Delgado, A. Andriollo, 0. L. De Ochoa, T. Suarez, N. Valencia, J . Organomet. Chem., 209, 77 (1981). M. Verzele, M. Acke, M. Anteunis, J. Chem. Soc., 5598 (1963). P. D. Ruggieri, U.S.Patent 3,207,752 (1965); Chem. Abstr., 59, 11610b (1963).

14.3.6.1.2. Aromatlc Aldehydes.

Aromatic aldehydes hydrogenate more readily than unactivated aliphatic aldehydes. If not careful, overreduction affords the hydrocarbon after hydrogenolysis of the benzyl hydroxyl function'. Palladium is a choice catalyst for hydrogenation of aryl aldehydes with or without hydrogenolysis. Acidic, solvents, high T and P should be avoided in order to obtain the alcohol. Basic inhibitors such as tertiary amines or alkali hydroxides stop reduction at the hydroxyl stage, e.g., formation of amine 1 prevents hydrogenolysis of the benzylic hydroxyl group even at high H, P2:


174

14.3.6. Hydrogenation of C=O Functions 14.3.6.1. In Aldehydes 14.3.6.1.2. Aromatic Aldehydes.

hydrogenolyzed to the corresponding ether. Palladium promotes acetal formation. PtO, in alcohol containing dry HCl leads to ether formation (93% yield) after hydrogenolysis of intermediate hemiketal~'~: (CH,),CHCOCH,

Hz' Ptoz > CH,OH, 2.5 M HCl

(CH,),CHC(CH,)OCH,

a-Hydroxymethylene ketones undergo hydrogenolysis during reduction over Pd-oncarbon, affording the selective introduction of a methyl group in steroids2': OH

(J.-L. GRAS)

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

W. Wenner, J. Org. Chem., 15, 301 (1950). L. W. Wright, Chemtech., 42 (1974). E. H. Pryde, C. M. Thierfelder, J. C. Cowan, J . Am. Chem. SOC.,53, 90 (1976). J. A. Schreifels, P. C. Maybury, W. E. Swartz, Jr., J. Org. Chem., 46, 1263 (1981). W. Sucrow, M. Slopianka, P. P. Caldeira, Chem. Ber. 108, 1101 (1975). N. E. Hoffman, A. T. Kanakkanatt, R. F. Schneider, J . Org. Chem., 27, 2687 (1962). W. H. Carothers, R. Adams, J. Am. Chem. SOC.,45, 1071 (1923). P. N. Rylander, J. Kaplan, Engelhardtlnd. Tech. Bull., 2 , 48 (1961); Chem. Abstr., 56, 2918 d (1961). R. Baltzly, J . Org. Chem., 41, 933 (1976). T. Komatsu, R. Iwanaga, J. Kato, U.S.Patent 3,141,895, (1964). Jpn. Patent I. 3421 (1964); Chem. Abstr., 6 I , 2977 b (1964). V. I. Sharkov, Angew. Chem. Int., Ed. Engl. 2,405 (1963). G. L. Aldridge, H. B. Jonassen, J . Am. Chem. SOC., 85, 886 (1963). R. S . Coffey, J. Chem. Soc., Chem. Commun., 923 (1967). W. Strohmeier, H. Steigenvald, J. Organomet. Chem., 129, C 43 (1977). H. Fujitsu, E. Matsumura, K. Takeshita, I. Mochida, J . Org. Chem., 46, 5353 (1981). K. Tani, K. Suwa, E. Tanigawa, T. Yoshida, T. Okano, S . Otsuka, Chem. Lett., 261 (1982). W. Strohmeier, L. Weigelt, J. Organomet. Chem., 145, 189 (1978). R. A. Sanchez-Delgado, A. Andriollo, 0. L. De Ochoa, T. Suarez, N. Valencia, J . Organomet. Chem., 209, 77 (1981). M. Verzele, M. Acke, M. Anteunis, J. Chem. Soc., 5598 (1963). P. D. Ruggieri, U.S.Patent 3,207,752 (1965); Chem. Abstr., 59, 11610b (1963).

14.3.6.1.2. Aromatlc Aldehydes.

Aromatic aldehydes hydrogenate more readily than unactivated aliphatic aldehydes. If not careful, overreduction affords the hydrocarbon after hydrogenolysis of the benzyl hydroxyl function'. Palladium is a choice catalyst for hydrogenation of aryl aldehydes with or without hydrogenolysis. Acidic, solvents, high T and P should be avoided in order to obtain the alcohol. Basic inhibitors such as tertiary amines or alkali hydroxides stop reduction at the hydroxyl stage, e.g., formation of amine 1 prevents hydrogenolysis of the benzylic hydroxyl group even at high H, P2:

14.3.6. Hydrogenation of 14.3.6.1. In Aldehydes 14.3.6.1.3. Selectivity.

C=O Functions

175

1 On the other hand, use of acetic acid as the solvent, or addition of a few drops of concentrated HCl or HClO,, facilitates formation of the hydrocarbon. Platinum oxide is rapidly deactivated by aromatic aldehydes through reduction of the catalyst to a lower oxidation state. This difficulty is circumvented by reactivation after shaking the reaction mixture with air or by various additives such as Fe3: Br

Br

OMe

OMe

Platinum oxide-Fe or Cu-containing catalysts allow hydrogenation of furfural to furfurylal~ohol~. Ruthenium catalysts (Ru-C, RuO,) are successful in this specific case5; they have an activity well preserved through reuses. Otherwise Ru exhibits little activity in the heterogeneous hydrogenation of aromatic aldehydes. Other heterogeneous catalysis include platinized (PtCl,) Raney Ni and copper chromite. Under homogeneous conditions, RuC1, [P(Ph)3]3is an effective catalyst for hydrogenation of aromatic aldehydes to benzyl alcohols at 50-80°C and at an H, P of lo3 P a 6 . Aromatic aldehyde hydrogenations are generally performed in ethanol, acetic acid, acetone, or ethyl acetate, with an order of efficiency depending on the catalyst. Acetal formation, catalyzed by traces of acid, are observed in methanol but not in ethanol'. (J.-L. GRAS)

R. W. Meschke, W. H. Hartung, J. Org. Chem., 25, 137 (1960). K. Sato, Y. Fujima, A. Yamada, Bull. Chem. SOC.Jpn., 41, 442 (1968). 3. H. Gardner, T. F. McDonnell, J . Am. Chem. SOC.,63, 2279 (1941). G. Seo, H. Chon, J . Catal., 67,424 (1981). A. A. Ponomarev, A. S. Chegolya, Dokl. Akad. Nauk. SSSR 145,812 (1962);Chem. Abstr., 57, 14467 h (1962). 6. 3. Tsuji, H. Suzuki, Chem. Lett., 1085 (1977). 7. W. H. Carothers, R. Adams, J . Am. Chem. SOC., 46, 1675 (1924). 1. 2. 3. 4. 5.

14.3.6.13.Selectivity.

Hydrogenation of the aldehyde carbonyl group competes with hydrogenation of other functional groups. The cyan0 group of P-cyanoproprionaldehyde is preserved during catalytic hydrogenation of the C=O function over 5% palladium-on-carbon with rigorous exclusion of alkali'. Aldehydes are hydrogenated more rapidly than ketones, and a-hydroxymethylene ketones are thus converted to the keto alcohols upon hydrogenation over PtO,, in ethanol [equation (a)]. Hydrogenolysis of the hydroxymethylene group to a methyl group occurs following esterification':


14.3.6. Hydrogenation of 14.3.6.1. In Aldehydes 14.3.6.1.3. Selectivity.

C=O Functions

175

1 On the other hand, use of acetic acid as the solvent, or addition of a few drops of concentrated HCl or HClO,, facilitates formation of the hydrocarbon. Platinum oxide is rapidly deactivated by aromatic aldehydes through reduction of the catalyst to a lower oxidation state. This difficulty is circumvented by reactivation after shaking the reaction mixture with air or by various additives such as Fe3: Br

Br

OMe

OMe

Platinum oxide-Fe or Cu-containing catalysts allow hydrogenation of furfural to furfurylal~ohol~. Ruthenium catalysts (Ru-C, RuO,) are successful in this specific case5; they have an activity well preserved through reuses. Otherwise Ru exhibits little activity in the heterogeneous hydrogenation of aromatic aldehydes. Other heterogeneous catalysis include platinized (PtCl,) Raney Ni and copper chromite. Under homogeneous conditions, RuC1, [P(Ph)3]3is an effective catalyst for hydrogenation of aromatic aldehydes to benzyl alcohols at 50-80°C and at an H, P of lo3 P a 6 . Aromatic aldehyde hydrogenations are generally performed in ethanol, acetic acid, acetone, or ethyl acetate, with an order of efficiency depending on the catalyst. Acetal formation, catalyzed by traces of acid, are observed in methanol but not in ethanol'. (J.-L. GRAS)

R. W. Meschke, W. H. Hartung, J. Org. Chem., 25, 137 (1960). K. Sato, Y. Fujima, A. Yamada, Bull. Chem. SOC.Jpn., 41, 442 (1968). 3. H. Gardner, T. F. McDonnell, J . Am. Chem. SOC.,63, 2279 (1941). G. Seo, H. Chon, J . Catal., 67,424 (1981). A. A. Ponomarev, A. S. Chegolya, Dokl. Akad. Nauk. SSSR 145,812 (1962);Chem. Abstr., 57, 14467 h (1962). 6. 3. Tsuji, H. Suzuki, Chem. Lett., 1085 (1977). 7. W. H. Carothers, R. Adams, J . Am. Chem. SOC., 46, 1675 (1924). 1. 2. 3. 4. 5.

14.3.6.13.Selectivity.

Hydrogenation of the aldehyde carbonyl group competes with hydrogenation of other functional groups. The cyan0 group of P-cyanoproprionaldehyde is preserved during catalytic hydrogenation of the C=O function over 5% palladium-on-carbon with rigorous exclusion of alkali'. Aldehydes are hydrogenated more rapidly than ketones, and a-hydroxymethylene ketones are thus converted to the keto alcohols upon hydrogenation over PtO,, in ethanol [equation (a)]. Hydrogenolysis of the hydroxymethylene group to a methyl group occurs following esterification':

14.3.6. Hydrogenation of C=O Functions 14.3.6.1. In Aldehydes 14.3.6.1.3.Selectivity.

176

I I

I

QQ 0

H,,100 kPa, RT PtOe, EtOH

’

CHOCOPh

0

Both aliphatic and aromatic aldehydes are hydrogenated to the alcohols in presence of RuCl,(PPh,),; ketones and nitro compounds are not affected3. Even though it is more difficult to hydrogenate a C=O function than a C=C bond, it is possible to obtain all three hydrogenated products (saturated aldehyde, unsaturated alcohol, or saturated alcohol) by selective hydrogenation of unsaturated aldehydes. Provided the C=C bond is not too hindered, selectivity can be reached in high yields by an appropriate choice of reaction variables such as metal, support, catalyst modifiers, and reaction conditions. Reduction of the C=C bond in unsaturated aldehydes is not difficult per se, and the saturated aldehyde is obtained in good yields using Pd on various supports4. Selective catalytic hydrogenation of unsaturated aldehydes to unsaturated alcohols is much harder, but has been achieved in several instances. Platinum-or-carbon or platinum oxide modified by FeC1, and zinc acetate are among the earliest successful systems tested5.Zinc ions inhibit double bond saturation and Fe(I1) ions promote carbonyl reduction, as6: (CH,),C=CHCHO (CH,),C=CHCH,OH

93%

0.5 mg PtO, 0.37 mg FeSO, 0.08 mg Zn(OAc), H,, 3 X lo3 kPa

+

(CH,),CHCH,CH,OH

3.5%

(CH,),CHCH,CHO

1.5%

(b)

Attempts to replace both Fe and Zn by other modifiers failed and Pt is effective only supported on carbon or CaC0;. Modified finely divided Ni or Co (Raney metals) are used in similar reductions, e.g., for the hydrogenation of citral 1 [equation (c)]. Using a Pt0,-Pt black catalyst, the conjugated C=C bond is reduced first, then the aldehyde group, and finally the second C=C bond, to give tetrahydrogeraniol 2. If FeSO, and zinc acetate are added, the aldehyde group is reduced first and the reduction can be stopped at the geraniol 3; absorption of an additional equivalent of H, gives citronellol 48. The same stepwise reductions can be obtained with high selectivity using a CoC1,-modified cobalt (Raney) catalyst’. The promoter decreases the amount of strongly adsorbed relative to the weakly

14.3.6. Hydrogenation of C=O Functions 14.3.6.1. In Aldehydes 14.3.6.13. Selectivity.

177

__

adsorbed H, and increases the rate of formation of solely the unsaturated alcohol". (Citronellol is produced directly from 1 in 94% yield using a chromium-promoted Ni (Raney) catalyst in the presence of methanol". (a) PtO,, FeSO,, Zn(OAc),

1

(b) CoCI,-Ra. Co, EtOH

Copper chromite is successfully used at high P and T but homogeneous Ru catalysts selectively reduce the carbonyl group of unsaturated aldehyde at 35-50°C'2. Aromatic unsaturated aldehydes behave differently, since they are vinylogues of benzaldehydes. In cinnamaldehydes, reduction of the C=O group and C=C bond are almost equally competitive. However, the selective hydrogenation to hydrocinnamaldehydes or cynnamyl alcohols is controlled depending on the metal, on the support, on additives and on solvent4. Hydrocinnamaldehydes form selectively over Pd catalysts modified with potassium salts of weak acids', or FeS02. Cobalt carbonyl catalysts under 0x0 conditions, or Co,(CO), in the presence of amines, efficiently (96.4% yield) catalyze selective red~ction'~:

p-t-Bu-C6H4-CH=C(CH3)CHO

H,:CO, 6.5 X lo3 P a , i-PrOH 107°C Co,(CO),, HNi-F'r,

~-~-Bu-C~H~-CH~CH(CH,)CHOH (d)

The cinnamaldehyde C=C bond is also preferentially reduced under 0x0 conditions using RhCl(PPh,), or RhCl(CO)(PPh,), to produce hydrocinnamaldehyde q~antitatively'~. Cinnamyl alcohols are obtained in >90% yields over appropriate catalysts. Platinum catalysts modified by FeS04 and zinc acetategs7,5% iridium-on-carbon'6 or 5% osmium-~n-carbon~~ smoothly afford cinnamyl alcohol. Cinnamaldehydes, with Rh halide catalysts in the presence of CO and highly basic tertiary amine such as triethylamine and N-methylpyrrolidine, are reduced with 85% selectivity to the unsaturated alcohol [equation (e)]lg. Prereduction of RhC13.3H,0 with CO suggests that reduction of the trichloride occurs during hydrogenation. Activity after this treatment is such that lower reaction temperature can be employed (90°C). PhCH=C(R)CHO

H,:CO, 8 X lo3 P a , 60°C > Rh,CI2(CO),, MeN--(CHJ,-

PhCH=C(R)CH,OH

(e)

(J.-L. GRAS)

1. T.Komatsu, R.Iwanaga, J. Kato, U.S.Patent 3,141,895(1964);Jpn. Patent 1, 3421 (1964); Chem. Abstr., 61, 2977b (1964).

14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.1. Hydrogenationto the Carbinol.

178

2. B. D. Astill, V. Boekelheide, J . Am. Chem. SOC.,77, 4079 (1955). 3. J. Tsuji, H. Suzuki, Chem. Lerr., 1085 (1977). 4. P. N. Rylander, N. Himelstein, Engelhard Ind. Tech. Bull., 4, 131 (1964), Chem. Absrr., 60, 15765e (1964). 5 . W. F. Tuley, R. Adams, J. Am. Chem. Soc., 47, 3061 (1925). 6. Y. Ichikawa, M. Suzuki, T. Sawaki, Jpn. Patent 77,46,193 (1977); Chem. Abstr., 8 7 , 2 0 0 7 9 4 ~ (1977). 7. P. N. Rylander, N. Himelstein, M. Kilroy, Engelhard Ind. Tech. Bull., 4, 49 (1963); Chem. Abstr., 60, 400d (1964). 8. R. Adams, B. S. Gamey, J . Am. Chem. Soc., 48,477 (1926). 9. K. Hotta, S. Watanabe, T. Kubomatsu, Nippon Kagaku Kaishi, 352 (1982). 10. K. Hotta, T. Kubomatsu, Bull. Chem. SOC.Jpn., 46, 3566 (1973). 11. P. S . Gradeff, G . Formica, Tetrahedron Lett., 4681 (1976). 12. J. M. Gosselin, C. Mercier, G . Allmang, F. Grass, Organomerallics, 10, 2126 (1991). 13. M. Dunkel, D. J. Eckardt, A. Stem, US. Patent 3,520,934 (1970); Chem. Abstr., 73, 66251w (1970). 14. K. Kogami, J. Komanotani, Bull. Chem. SOC.Jpn., 46, 3562 (1973). 15. E. Ucciani, R. Lai, L. Tanguy, C . R. Acad. Sci. SCr. C , 283, 17 (1976). 16. M. L. Khidekel, E. N. Bakhanova, A. S. Astakhova, K. A. Brikenshtein, V. I. Savchenko, I. S. Monakhova, V. G. Dorokhov, Izv. Akad. Nauk SSSR,Ser. Khim., 499 (1970); Chem. Abstr., 73, 3130k (1970). 17. P. N. Rylander, D. R. Steele, Tetrahedron Left., 1579 (1969). 18. T. Mizoroki, K. Seki, S. Meguro, A. Ozaki, Bull. Chem. SOC.Jpn., 50, 2148 (1977).

14.3.6.2. In Ketones 14.3.6.2.1. Hydrogenationto the Carbinol.

Hydrogenation of ketones to the corresponding carbinol occurs over various catalysts. Only aromatic ketones undergo easy hydrogenolysis or overhydrogenation of the ring. Most common metals have been successfully used in heterogeneous or homogeneous catalysis. Their relative effectiveness is a consequence of the adsorption rather than the reaction step’. Hydrogenation of aromatic ketones is most frequently done over Pd, sometimes with 100% yields. High activity for ketone reduction and low activity for ring hydrogenation make palladium-on-carbon the catalyst of choice for reduction of acetophenone to phenylethanol and of 2-acetylpyridine to 2-( 1-hydroxyethyl)pyridine’: 5% Pd-C

Qy-Qy 0

OH

Highest yields are obtained using trace amounts of organic bases or alkali3, and by interrupting the reduction after absorption of the theoretical amount H,. Amino ketones are reduced to the corresponding amino carbinols, either as free base or as amine salts. With some hindered carbonyls, small amounts of strong acid promote hydrogenation without concomitant hydrogenoly~is~:

wNa 0

H,, 5% Pd-BaSO,

AcOH, HClO,, 70°C

’

OH

90% yield


14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.1. Hydrogenationto the Carbinol.

178

2. B. D. Astill, V. Boekelheide, J . Am. Chem. SOC.,77, 4079 (1955). 3. J. Tsuji, H. Suzuki, Chem. Lerr., 1085 (1977). 4. P. N. Rylander, N. Himelstein, Engelhard Ind. Tech. Bull., 4, 131 (1964), Chem. Absrr., 60, 15765e (1964). 5 . W. F. Tuley, R. Adams, J. Am. Chem. Soc., 47, 3061 (1925). 6. Y. Ichikawa, M. Suzuki, T. Sawaki, Jpn. Patent 77,46,193 (1977); Chem. Abstr., 8 7 , 2 0 0 7 9 4 ~ (1977). 7. P. N. Rylander, N. Himelstein, M. Kilroy, Engelhard Ind. Tech. Bull., 4, 49 (1963); Chem. Abstr., 60, 400d (1964). 8. R. Adams, B. S. Gamey, J . Am. Chem. Soc., 48,477 (1926). 9. K. Hotta, S. Watanabe, T. Kubomatsu, Nippon Kagaku Kaishi, 352 (1982). 10. K. Hotta, T. Kubomatsu, Bull. Chem. SOC.Jpn., 46, 3566 (1973). 11. P. S . Gradeff, G . Formica, Tetrahedron Lett., 4681 (1976). 12. J. M. Gosselin, C. Mercier, G . Allmang, F. Grass, Organomerallics, 10, 2126 (1991). 13. M. Dunkel, D. J. Eckardt, A. Stem, US. Patent 3,520,934 (1970); Chem. Abstr., 73, 66251w (1970). 14. K. Kogami, J. Komanotani, Bull. Chem. SOC.Jpn., 46, 3562 (1973). 15. E. Ucciani, R. Lai, L. Tanguy, C . R. Acad. Sci. SCr. C , 283, 17 (1976). 16. M. L. Khidekel, E. N. Bakhanova, A. S. Astakhova, K. A. Brikenshtein, V. I. Savchenko, I. S. Monakhova, V. G. Dorokhov, Izv. Akad. Nauk SSSR,Ser. Khim., 499 (1970); Chem. Abstr., 73, 3130k (1970). 17. P. N. Rylander, D. R. Steele, Tetrahedron Left., 1579 (1969). 18. T. Mizoroki, K. Seki, S. Meguro, A. Ozaki, Bull. Chem. SOC.Jpn., 50, 2148 (1977).

14.3.6.2. In Ketones 14.3.6.2.1. Hydrogenationto the Carbinol.

Hydrogenation of ketones to the corresponding carbinol occurs over various catalysts. Only aromatic ketones undergo easy hydrogenolysis or overhydrogenation of the ring. Most common metals have been successfully used in heterogeneous or homogeneous catalysis. Their relative effectiveness is a consequence of the adsorption rather than the reaction step’. Hydrogenation of aromatic ketones is most frequently done over Pd, sometimes with 100% yields. High activity for ketone reduction and low activity for ring hydrogenation make palladium-on-carbon the catalyst of choice for reduction of acetophenone to phenylethanol and of 2-acetylpyridine to 2-( 1-hydroxyethyl)pyridine’: 5% Pd-C

Qy-Qy 0

OH

Highest yields are obtained using trace amounts of organic bases or alkali3, and by interrupting the reduction after absorption of the theoretical amount H,. Amino ketones are reduced to the corresponding amino carbinols, either as free base or as amine salts. With some hindered carbonyls, small amounts of strong acid promote hydrogenation without concomitant hydrogenoly~is~:

wNa 0

H,, 5% Pd-BaSO,

AcOH, HClO,, 70°C

’

OH

90% yield

14.3.6. Hydrogenationof C=O Functions 14.3.6.2. In Ketones 14.3.6.2.1. Hydrogenationto the Carbinol.

179

Ruthenium-on-carbon in aqueous ethanol or platinum oxide also is used but to a much lesser extent than Pd. Palladium shows a low activity for hydrogenation of nonactivated aliphatic ketones, but all the platinum metals can be used in addition to Cu chromite and Ni catalysts. Platinum catalysts have been widely used, platinum-on-carbon in aqueous acid is satisfactog. Rhodium is active under mild conditions and leads to a-hydroxy steroids in excellent yields6:

Rhodium catalysts such as the complexes RhCl(C,H,,)PPh, and Rh,H, Cl,(C,H12)(PPh,), have been developed for hydrogenation of ketones. With strong alkali, the dirhodium species gives best results7. Pretreatment with NaBH, results in higher hydrogenation rates, suggesting the existence of a hydroxo complex active intermediate. Cationic Rh(1) complexes with triethylphosphine 1' or with tetraalkyled diphosphine ligands z9 also show remarkable activity as catalysts for the hydrogenation of ketones to alcohols. [Rh(NBD)i-Pr,P( CH,),PiPr, ]ClO,

[Rh(NB D)(PEt,),] C10,

1

2

NBD = norbomadiene

n = 3,4

Ruthenium catalysts require an induction period at low Plo and are commonly used at elevated P (20 X lo3 Wa). Anionic Ru complexes, 3 and 4, catalyze homogeneous hydrogenation of carbonyl derivatives in THF at 85°C".

3

K [(PPh,),Ph,P-C,H,-RuH,]

4

K, [(PPh,),Ph,PRu,H,]

naphthalin ether

2 diglyme

Iridium and osmium are little used for ketone to carbinol reduction, Ir mainly as a hydrogen transfer catalyst, and 0 s in order to minimize ring saturation in keto-compounds having an aromatic nucleus". Copper-chromite is useful for hydrogenation of hydroxyphenyl carbonylcompounds without hydrogenolysis [equation (d)]. Vigorous conditions are required including activation through refluxing in cycl~hexanol'~ under nitrogen. Activated Cu-CrO, EtOH H,, 25 X lo3 P a , 70°C

HO

HO

98%

Copper on alumina promotes hydrogenation of keto steroid to the carbinol quantitative1y 14. Nickel catalysts are often used under conditions that depend on the catalyst nature and the amount used. Finely divided Ni and nickel boride, as well as NaH/RONa/NI(OAc),, have been studied as efficient catalysts sensitive to steric

180

14.3.6.Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.2. Hydrogenolysisand Miscellaneous Reactions.

factors15. Hydrogenation rate of the keto group depends on the solvent. Better results should be obtained in a weakly adsorbed solvent such as a hydrocarbon16. Small quantities of acids or bases influence the hydrogenation rate or the product obtained. Traces of acid commonly accelerate the reduction rate, supporting the idea that protonated ketones are involved". However, the role of alkali often appears obscure'*. Either types of promoters also act as inhibitors and generalization remains elusive. Ketones are always easily reduced with Ni-A1 alloy in alkaline medium''. Enolized 1,Zdiketones are hydrogenated over Pd-C after protection of enol as methylethe?'. (J.-L. GRAS) 1. J. Simonikova, A. Ralkova, K. Kochloefl, J . Catal., 29, 412 (1973). 2. M. Freifelder, J . Org. Chem., 29, 2895 (1964). 3. P. N. Rylander, L. Hasbrouck, Engelhard Ind. Tech. Bull., 8, 148 (1968); Chem. Abstr., 69, 76710b (1968). 4. C. L. Stevens, C. H. Chang, J . Org. Chem., 27,4392 (1962). 5. E. Breitner, E. Roginski, P. N. Rylander, J . Org. Chem., 24, 1855 (1959). 6. S.Nishimura, M. Ishige, M. Shiota, Chem. Lett., 963 (1977). 7. M. Gargano, P. Giannoccaro, M. Rossi, J . Organomer. Chem., 129, 239 (1977). 8 . H. Fujistu, E. Matsumura, K. Takeshita, I. Mochida, J . Chem. SOC., Perkin Trans. I , 2650 ( 1981). 9. K. Tani, K. Suwa, E. Tankgawa, T. Yoshida, T. Okano, S. Otsuka, Chem. Lett., 261 (1982). 10. E. E. Smissman, J. F. Muren, N. A. Dahle, J . Org. Chem., 29, 3517 (1964). 11. R. A. Grey, G. P. Pez, A. Wallo,J. Am. Chem. Soc., 103,7536 (1981). 12. P. N. Rylander, L. Hasbrouck, Engelhard Ind. Tech. Bull., 10, 50 (1969), Chem. Abstr., 72, 43052x (1970). 13. W. R. Nes, J. Org. Chem., 23, 899 (1958). 14. N. Ravasio, M. Rossi, J . Org. Chem., 56,4329 (1991). 15. J. J. Brunet, P. Gallois, P. Caubere, J . Org. Chem., 45, 1937 (1980). 16. S. Kishida, S . Teranishi, J. Card., 12, 90 (1968). 17. J. H. Brewster, J . Am. Chem. Soc., 76, 6361 (1954). 18. S. Mitsui, H. Saito, Y. Yamashita, M. Kaminaga, Y. Senga, Tetrahedron, 29, 1531 (1973). 19. L. K. Keefer, G. L. Lunn, Chem. Rev., 89,459 (1989). 20. R. H. Bumell, M. Jean, D. Poirier, Can. J . Chem., 65, 775 (1987).

14.3.6.2.2. Hydrogenolyslsand MiscellaneousReactions.

Aromatic ketones are easily reduced to the corresponding aromatic after hydrogenolysis of the intermediate benzyl alcohol. The procedure is an interesting alternative to hydrazine/NaOEt (Wolff-Kishner) or Zn/Hg/HCl (Clemmensen) reductions, because of the high yields and relatively mild conditions'. Hydrogenolysis of phenol ketones occurs readily over Pt and finely divided Ni (WZRaney Ni) at 80°C and 11 X lo3 kPa [equation (a)]*. Low activity toward ring reduction makes Pd (Pd-C, Pd-BaSO,, Pd black) the most useful catalyst for hydrogenolysis in general334. Finely divided Raney Ni, EtOH H,, 1 1 X 103kPa,80"C

II

0

' 87%

Hydrogenolysis occurs at elevated temperatures [equation (b)I5, in polar solvents such as alcohols, acetic acid, and by the addition of a few drops of concentrated HC1, HClO,, or H,S04 [equation (c)]~,while bases inhibit this full reduction to the methylene.


180

14.3.6.Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.2. Hydrogenolysisand Miscellaneous Reactions.

factors15. Hydrogenation rate of the keto group depends on the solvent. Better results should be obtained in a weakly adsorbed solvent such as a hydrocarbon16. Small quantities of acids or bases influence the hydrogenation rate or the product obtained. Traces of acid commonly accelerate the reduction rate, supporting the idea that protonated ketones are involved". However, the role of alkali often appears obscure'*. Either types of promoters also act as inhibitors and generalization remains elusive. Ketones are always easily reduced with Ni-A1 alloy in alkaline medium''. Enolized 1,Zdiketones are hydrogenated over Pd-C after protection of enol as methylethe?'. (J.-L. GRAS) 1. J. Simonikova, A. Ralkova, K. Kochloefl, J . Catal., 29, 412 (1973). 2. M. Freifelder, J . Org. Chem., 29, 2895 (1964). 3. P. N. Rylander, L. Hasbrouck, Engelhard Ind. Tech. Bull., 8, 148 (1968); Chem. Abstr., 69, 76710b (1968). 4. C. L. Stevens, C. H. Chang, J . Org. Chem., 27,4392 (1962). 5. E. Breitner, E. Roginski, P. N. Rylander, J . Org. Chem., 24, 1855 (1959). 6. S.Nishimura, M. Ishige, M. Shiota, Chem. Lett., 963 (1977). 7. M. Gargano, P. Giannoccaro, M. Rossi, J . Organomer. Chem., 129, 239 (1977). 8 . H. Fujistu, E. Matsumura, K. Takeshita, I. Mochida, J . Chem. SOC., Perkin Trans. I , 2650 ( 1981). 9. K. Tani, K. Suwa, E. Tankgawa, T. Yoshida, T. Okano, S. Otsuka, Chem. Lett., 261 (1982). 10. E. E. Smissman, J. F. Muren, N. A. Dahle, J . Org. Chem., 29, 3517 (1964). 11. R. A. Grey, G. P. Pez, A. Wallo,J. Am. Chem. Soc., 103,7536 (1981). 12. P. N. Rylander, L. Hasbrouck, Engelhard Ind. Tech. Bull., 10, 50 (1969), Chem. Abstr., 72, 43052x (1970). 13. W. R. Nes, J. Org. Chem., 23, 899 (1958). 14. N. Ravasio, M. Rossi, J . Org. Chem., 56,4329 (1991). 15. J. J. Brunet, P. Gallois, P. Caubere, J . Org. Chem., 45, 1937 (1980). 16. S. Kishida, S . Teranishi, J. Card., 12, 90 (1968). 17. J. H. Brewster, J . Am. Chem. Soc., 76, 6361 (1954). 18. S. Mitsui, H. Saito, Y. Yamashita, M. Kaminaga, Y. Senga, Tetrahedron, 29, 1531 (1973). 19. L. K. Keefer, G. L. Lunn, Chem. Rev., 89,459 (1989). 20. R. H. Bumell, M. Jean, D. Poirier, Can. J . Chem., 65, 775 (1987).

14.3.6.2.2. Hydrogenolyslsand MiscellaneousReactions.

Aromatic ketones are easily reduced to the corresponding aromatic after hydrogenolysis of the intermediate benzyl alcohol. The procedure is an interesting alternative to hydrazine/NaOEt (Wolff-Kishner) or Zn/Hg/HCl (Clemmensen) reductions, because of the high yields and relatively mild conditions'. Hydrogenolysis of phenol ketones occurs readily over Pt and finely divided Ni (WZRaney Ni) at 80°C and 11 X lo3 kPa [equation (a)]*. Low activity toward ring reduction makes Pd (Pd-C, Pd-BaSO,, Pd black) the most useful catalyst for hydrogenolysis in general334. Finely divided Raney Ni, EtOH H,, 1 1 X 103kPa,80"C

II

0

' 87%

Hydrogenolysis occurs at elevated temperatures [equation (b)I5, in polar solvents such as alcohols, acetic acid, and by the addition of a few drops of concentrated HC1, HClO,, or H,S04 [equation (c)]~,while bases inhibit this full reduction to the methylene.

14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.2. Hydrogenolysis and Miscellaneous Reactions.

Pd-I-;EtOAc

10%PdX;EtOAc

HO

10%

-q--J-q 0

COOH

COOH

COOH

AcOH -H,SO,

MeOOC

181

0

0

COOMe

MeOOC

(c)

COOMe

Activated easily enolizable aliphatic ketones, such as P-diketones, P-ketoamides, and P-ketoesters, are extensively hydrogenolyzed. Hydrogenolysis depends on the catalyst type, amount of catalyst, solvent, and substrate structure'. The methylene compound is favored in acetic acid, alcohols, and H20, with larger amount of catalyst and Pt is often used. For instance, hydrogenation of cohulupone 1 over platinum oxide in methanol affords mainly 2 [equation (d)I8.

1

2

1,2-Diketones are hydrogenolyzed to monoketones through catalytic hydrogenation of an unsaturated oxypho~phorane~: R-

P(OCH3)3

CO -CO-R'

R-C-C-R'

I

0

I

-% R-CH,-CO-R'

0

(e)

\/

P (OCH,), Hydrogenolysis of nonactivated ketones is a side reaction only when hydrogenation is run in acid. In inososes lacking two axial hydroxyls next to the keto group, hydrogenolysis occurs in dilute H2S0,":

0

PtOZ

H2S04

'

(0

HO

HO

OH

OH

Reduction of cyclohexanones in MeOH or EtOH with prereduced palladium hydroxide as catalyst affords cyclohexyl ethers. For example, 5a-cholestane-3-one gives P-methoxy-Sa-cholestane undergoes reduction in 91% yield in methanol. Hydrogenation of cyclohexanones in ethanol leads to ethoxycyclohexanes" :

14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.2. Hydrogenolysis and Miscellaneous Reactions.

182

Go

Pd catalyst

EtOH

'

0

0

.

(g)

96%,90%cis Reaction occurs intramolecularly when appropriate diketones are submitted to hydrogenationI2:

(h)

In fact, diketones in a medium sized ring give several transannular reactions such as pinacol coupling [equation (i)]I3 or aldolization [equation The later reaction also occurs during hydrogenation of aliphatic monoketones under favorable conditions (Pdon-carbon, zeolite, 20O0C,4 X lo3kPa) to produce methyl isobutyl ketone from acetone (in 96%)15: OH

0

0

Ketones in a potentially aromatic system such as quinones can give phenols on catalytic hydrogenation under heterogeneous (PtO,) or homogeneous conditions16:

'Go

-

(Rh(NBD)[i-Pr,P H,, 6(CH,),,-PiPr,] min

)ClO,,

'

H

O

e

O

H (k)

100% However, reductions of aromatic ketones over Rh or Ru in appropriate solventI7 and under elevated pressure1s afford saturated carbinols with maintenance of benzyl oxygen, as in the synthesis of natural products such as d-phyll~cladene'~:

'

'COOMe (J.-L. GRAS)

14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.3. Selectivity.

183

1. E. J. Eisenbraun, C. W. Hinman, J. M. Springer, J. W. Bumham, T. S.Chou, P. W. Flanagan, M. C. Hamming, J . Org. Chem., 36, 358 (1971). 2. H. E. Ungnade, A. D. McLaren, J . Am. Chem. Soc., 66, 118 (1944). 3. W. H. Hartung, R. Simonoff, Urg. React., 7 , 263 (1953). 4. J. G . Berger, S. R. Teller, I. J. Pachter, J . Org. Chem., 35, 3122 (1970). 5. G. N. Walker, J . Urg. Chem., 23, 133 (1958). 6. A. G . Anastassiou, G . W. Griffin, J . Org. Chem., 33, 3441 (1968). 7. P. N. Rylander, S.Stamck, Engelhard Ind. Tech. Bull., 7 , 106 (1966); Chem. Abstr., 67,90361 d ( 1967). 8. J. S. Burton, J. A. Elvidge, R. Stevens, J . Chem. SOC.,3816 (1964). 9. L. M. Stephenson, L. C. Falk, J . Urg. Chem., 4 1 , 2928 (1976). 10. G. G . Post, L. Anderson, J . Am. Chem. Soc., 84, 471 (1962). 11. S.Nishimura, T. Itaya, M. Shiota, J . Chem. SOC.,Chem. Cornmun., 422 (1967). 12. A. J. Birch, F. A. Hochstein, J. A. K. Quartley, J. P. Tumbull, J . Chem. SOC. 2923 (1964). 13. A. C. Cope, F. Kagan, J . Am. Chem. Soc., 80,5499 (1958). 4606 (1963). 14. G . L. Buchanan, J. G. Hamilton, R. A. Raphael, J . Chem. SOC., 15. T. Imai, Y. Mitsuda, H. Ebisawa, T. Kametaka, T. Minoura, Jpn. Patent 7102,643 (1971); Chem. Abstr. 74, 111569 q (1971). 16. K. Tani, K. Suwa, E. Tanigawa, T. Yoshida, T. Okano, S. Otsuka, Chem. Lett., 261 (1982). Ind., Tech. Bull., 8 , 148 (1968): 17. P.N. Rylander, L. Hasbrouk, Ennelhard . , Chem. Abstr.. 69. 76710 d(1968). 18. N. S. Barinov, D. V. Mushenko, Zh. Prikl. Khim., 46, 940 (1973); Chem. Abstr. 79, 4674b (1973). 19. M. Shimagaki, A. Tamara, Tetrahedron Lett., 1715 (1975).

14.3.6.2.3. Selectivity.

Hydrogenation of diketones or unsaturated ketones yields diols or saturated carbinols under forcing conditions': 0

OH

Under appropriate conditions, depending on the catalyst and on the substrate, partial reduction of diketones affords the intermediate hydroxy ketones. Steric crowding may be used advantageously to achieve high selectivity, for instance in the reduction of polycarbonyl compounds 1 [equation (b)]', 2, 3 [equation (c)I3, and 4 [equation (d)I4. In these hydrogenations catalyzed by various Ni catalysts, the more highly hindered carbony1 remains unchanged.

H

H


14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.3. Selectivity.

183

1. E. J. Eisenbraun, C. W. Hinman, J. M. Springer, J. W. Bumham, T. S.Chou, P. W. Flanagan, M. C. Hamming, J . Org. Chem., 36, 358 (1971). 2. H. E. Ungnade, A. D. McLaren, J . Am. Chem. Soc., 66, 118 (1944). 3. W. H. Hartung, R. Simonoff, Urg. React., 7 , 263 (1953). 4. J. G . Berger, S. R. Teller, I. J. Pachter, J . Org. Chem., 35, 3122 (1970). 5. G. N. Walker, J . Urg. Chem., 23, 133 (1958). 6. A. G . Anastassiou, G . W. Griffin, J . Org. Chem., 33, 3441 (1968). 7. P. N. Rylander, S.Stamck, Engelhard Ind. Tech. Bull., 7 , 106 (1966); Chem. Abstr., 67,90361 d ( 1967). 8. J. S. Burton, J. A. Elvidge, R. Stevens, J . Chem. SOC.,3816 (1964). 9. L. M. Stephenson, L. C. Falk, J . Urg. Chem., 4 1 , 2928 (1976). 10. G. G . Post, L. Anderson, J . Am. Chem. Soc., 84, 471 (1962). 11. S.Nishimura, T. Itaya, M. Shiota, J . Chem. SOC.,Chem. Cornmun., 422 (1967). 12. A. J. Birch, F. A. Hochstein, J. A. K. Quartley, J. P. Tumbull, J . Chem. SOC. 2923 (1964). 13. A. C. Cope, F. Kagan, J . Am. Chem. Soc., 80,5499 (1958). 4606 (1963). 14. G . L. Buchanan, J. G. Hamilton, R. A. Raphael, J . Chem. SOC., 15. T. Imai, Y. Mitsuda, H. Ebisawa, T. Kametaka, T. Minoura, Jpn. Patent 7102,643 (1971); Chem. Abstr. 74, 111569 q (1971). 16. K. Tani, K. Suwa, E. Tanigawa, T. Yoshida, T. Okano, S. Otsuka, Chem. Lett., 261 (1982). Ind., Tech. Bull., 8 , 148 (1968): 17. P.N. Rylander, L. Hasbrouk, Ennelhard . , Chem. Abstr.. 69. 76710 d(1968). 18. N. S. Barinov, D. V. Mushenko, Zh. Prikl. Khim., 46, 940 (1973); Chem. Abstr. 79, 4674b (1973). 19. M. Shimagaki, A. Tamara, Tetrahedron Lett., 1715 (1975).

14.3.6.2.3. Selectivity.

Hydrogenation of diketones or unsaturated ketones yields diols or saturated carbinols under forcing conditions': 0

OH

Under appropriate conditions, depending on the catalyst and on the substrate, partial reduction of diketones affords the intermediate hydroxy ketones. Steric crowding may be used advantageously to achieve high selectivity, for instance in the reduction of polycarbonyl compounds 1 [equation (b)]', 2, 3 [equation (c)I3, and 4 [equation (d)I4. In these hydrogenations catalyzed by various Ni catalysts, the more highly hindered carbony1 remains unchanged.

H

H

14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.3.Selectivity.

184

Urushibara Ni-A

0

Urushibara Ni-A

HO" 2

0

77%

3

70%

Ni-Cr,

'

4

Palladium and rhodium are good catalysts for partial reduction of acyclic 1,3-diketones, although the substrate may be determinent for the choice of the catalys?. Limiting H, absorption to one equivalent allows selective hydrogenation of symmetrical cyclic diketones 5 to the corresponding ketols, over platinum oxide in acetic acid6:

Preferential reduction of an a#-unsaturated ketones C=C bond is more easily accomplished than preferential reduction of the C=O bond. However, a keto group can be hydrogenated in preference to a C=C bond after the appropriate choice of the catalyst, usually Pt or Ru7. Additives such as FeCl, (a carbonyl reduction promoter) and zinc acetate (a double bond saturation inhibitor) may be used in conjunction with platinum oxide*. The less hindered ketone in 6 is selectively hydrogenated over PtO, [equation (f)], whereas the C=C bond is selectively reduced over Pd9.

MeV

n 6

73%

14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.4. Stereochemistry and Asymmetric Hydrogenation.

185

Rhenium compounds such as Re,Se, are more reactive hydrogenation catalysts for the reduction of carbonyls than for olefins, except in conjugated systems”. Unsaturated carbonyls have been selectively hydrogenated to unsaturated alcohols, catalyzed by a chromium-promoted finely divided Ni (Raney Ni) catalyst’ ’: 0

?H Cr-finely divided Ni H,O, MeOH, NEt,, KOH

’

88.5%

Hydrogenation of ketones in alkaline medium at room T and atmospheric P, catalyzed by [Rh(2,2’-bipyridine) (diene)]PF, type complexes occurs with selectivity for reduction of carbon-oxygen double bonds to hydroxyl groups even in the presence of olefinic bonds. The system acts as a hydrogenolysis catalyst for molecular 0,, and hydrogenation activity is retained even if large amounts of 0, are present”. (J.-L. GRAS) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

F. Bohlmann, W. Otto, Liebigs Ann. Chem., 186 (1982). J. J. Brunet, P. Gallois, P. Caubere, J . Org. Chem., 45, 1937 (1980). M. Isaige, M. Shiota, F. Suzuki, Can.J. Chem. 54, 2581 (1976). D. Savoia, E. Tagliavini, C. Trombini, A. Umani-Ronchi,J. Org. Chem. 46, 5344 (1981). P. N. Rylander, D. R. Steele, Engelhardhd. Tech. Bull., 5, 113 (1965); Chem. Abstr., 63,5527 d (1965). A. T. Bloomquist, J. Wolinsky, J. Am. Chem. Soc., 77, 5423 (1955). I. A. Kaye, R. S. Matthews, J. Org. Chem., 29, 1341 (1964). Z. Csuros, K. Zech, T. Geczy, Hung. Acta Chim., I , 1 (1946); Chem. Abstr., 41, 109 (1947). D. J. France, J. J. Hand, M. Los, Tetrahedron, 25,4011 (1969). H. S. Broadbent, C. W. Whittle, J. Am. Chem. Soc., 81, 3587 (1959). P. S. Gradeff, G. Formica, Tetrahedron Lett., 4681 (1976). G. Mestroni, G. Zassinovich, A. Camus, J. Organomet. Chem., 140, 63 (1977).

14.3.6.2.4. Stereochemistry and Asymmetric Hydrogenatlon.

Catalytic hydrogenation of substituted cyclic ketones may afford two cis-trans isomers. Predicting the stereochemical outcome of this reduction is well studied with substituted cyclohexanones, for which rules are a~ailablel-~. However, numerous exceptions indicate that generalization is quite uncertain; too many factors affecting stereochemistry are involved. In neutral media and solvents the addition of H, takes place preferentially from the least hindered side (the side opposite the substituent) to yield cis product as illustrated in the hydrogenation of various 2-substituted 1-tetralones4.Hydrogenations in basic media or over catalysts containing alkaline impurities (finely divided Ni, PtO,) increase the ratio of more stable, equatorial hydroxy isomer, due to alcohol epimerization. Hydrogenations in acidic media favor the axial alcohol. Considerable variation in isomers ratio is observed by the use of different catalysts. Nickel is sensitive to carbonyl hindrance, whereas Pd favors the thermodynamic isome?. Urushibara-nickel-A (U-Ni-A) is an excellent catalyst for reduction of cholestane3-one to the 3a-01 [equation (a)], and to obtain epicholesterol in high yield from A5-cholestene-3-one [equation (b)],:



185

Rhenium compounds such as Re,Se, are more reactive hydrogenation catalysts for the reduction of carbonyls than for olefins, except in conjugated systems”. Unsaturated carbonyls have been selectively hydrogenated to unsaturated alcohols, catalyzed by a chromium-promoted finely divided Ni (Raney Ni) catalyst’ ’: 0

?H Cr-finely divided Ni H,O, MeOH, NEt,, KOH

’

88.5%

Hydrogenation of ketones in alkaline medium at room T and atmospheric P, catalyzed by [Rh(2,2’-bipyridine) (diene)]PF, type complexes occurs with selectivity for reduction of carbon-oxygen double bonds to hydroxyl groups even in the presence of olefinic bonds. The system acts as a hydrogenolysis catalyst for molecular 0,, and hydrogenation activity is retained even if large amounts of 0, are present”. (J.-L. GRAS) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

F. Bohlmann, W. Otto, Liebigs Ann. Chem., 186 (1982). J. J. Brunet, P. Gallois, P. Caubere, J . Org. Chem., 45, 1937 (1980). M. Isaige, M. Shiota, F. Suzuki, Can.J. Chem. 54, 2581 (1976). D. Savoia, E. Tagliavini, C. Trombini, A. Umani-Ronchi,J. Org. Chem. 46, 5344 (1981). P. N. Rylander, D. R. Steele, Engelhardhd. Tech. Bull., 5, 113 (1965); Chem. Abstr., 63,5527 d (1965). A. T. Bloomquist, J. Wolinsky, J. Am. Chem. Soc., 77, 5423 (1955). I. A. Kaye, R. S. Matthews, J. Org. Chem., 29, 1341 (1964). Z. Csuros, K. Zech, T. Geczy, Hung. Acta Chim., I , 1 (1946); Chem. Abstr., 41, 109 (1947). D. J. France, J. J. Hand, M. Los, Tetrahedron, 25,4011 (1969). H. S. Broadbent, C. W. Whittle, J. Am. Chem. Soc., 81, 3587 (1959). P. S. Gradeff, G. Formica, Tetrahedron Lett., 4681 (1976). G. Mestroni, G. Zassinovich, A. Camus, J. Organomet. Chem., 140, 63 (1977).

14.3.6.2.4. Stereochemistry and Asymmetric Hydrogenatlon.

Catalytic hydrogenation of substituted cyclic ketones may afford two cis-trans isomers. Predicting the stereochemical outcome of this reduction is well studied with substituted cyclohexanones, for which rules are a~ailablel-~. However, numerous exceptions indicate that generalization is quite uncertain; too many factors affecting stereochemistry are involved. In neutral media and solvents the addition of H, takes place preferentially from the least hindered side (the side opposite the substituent) to yield cis product as illustrated in the hydrogenation of various 2-substituted 1-tetralones4.Hydrogenations in basic media or over catalysts containing alkaline impurities (finely divided Ni, PtO,) increase the ratio of more stable, equatorial hydroxy isomer, due to alcohol epimerization. Hydrogenations in acidic media favor the axial alcohol. Considerable variation in isomers ratio is observed by the use of different catalysts. Nickel is sensitive to carbonyl hindrance, whereas Pd favors the thermodynamic isome?. Urushibara-nickel-A (U-Ni-A) is an excellent catalyst for reduction of cholestane3-one to the 3a-01 [equation (a)], and to obtain epicholesterol in high yield from A5-cholestene-3-one [equation (b)],:


186

& "2-A & & & &&

O

H

O

,\ H

'

4Ho

HO"

6047%

U-Ni-A Hz

,

HO"

H

(a)

40-13%

+

\

HO

94-77%

\

(b)

23-6%

The ratio of epimeric products depends on the solvent'.' and, to a lesser extent, the reaction temperatureg. Aging the catalyst should be also considered. Thus hydrogenation of 2-methyl-cyclohexanone with freshly prepared finely divided Ni (Raney Ni) gives 69% of the cis-alcohol, and use of aged Ni gives 80% of the same axial hydroxy isomer". Very high stereoselective hydrogenation of 4-tertbutylcyclohexanone occurs in ethanol or tetrahydrofuran containing HCl with Rh metal as catalyst": a

0

Hz, 250c,

e

Pd, EtOH Rh, EtOH Rh, THF, HCl

O

H

+

+oEt

(c)

1% (88% cis) 99% (97% cis) 93% (97% cis) 7% (90% cis) 100% (99.3% cis) -

' a+ m

This method is used for synthesis of axial alcohols, especially steroids':

r

r

7

HO"

97.5%

HO

2.3%

(4

H,, Rh, THF, HC1

96.6%

HO"

3.2%

Generally, stereochemistry of the hydrogenated product is a consequence of the substrate configuration either in the first adsorption step or in the second hydrogen transfer ~ t e p ~ ,Steric ' ~ . considerations may be outweighed by the presence within the molecule of other functionality or a basic atom, which can coordinate to the metal. The side at which the substrate adsorbed is thus determined by an "anchor effect," which controls the stereochemical outcome of the reaction. Keto-acid 1 is hydrogenated to hydroxy-acid 2, where the two oxygenated functions are trans1':

WOOH Q€p 14.3.6. Hydrogenation of C=O Functions 14.3.6.2. In Ketones 14.3.6.2.4. Stereochemistry and Asymmetric Hydrogenation.

8.r' gR

187

EtOH 10% Pd-C

1

HO

2

COOH

This effect is illustrated in the reduction of amino ketones, which involves direction of hydride transfer by participation of the amine f ~ n c t i o n l ~ ~ ' ~ : H

H

O

P

N-

t -

___j

R = Me

, (f)

R=H

Hydrogenation of diketones to diols yields mixtures of stereoisomers. Tetramethyl1,3.-cyclobutanedione hydrogenates rapidly over Ru-on-carbon, nearly quantitative yielding an about equimolar mixture of the cis- and tr~ns-diol'~: HO 125°C H2, Ru-C, 8 X 103kF'a MeOH'

OH ' H o ) H

(g)

98%

1,2-Diketones are reduced over a diphosphine coordinated cationic Rh complex to a mixture of meso- and DL-a-dio1sl6: H,, 10 min

CH3C0-c0CH3

'

[Rh(L-L)(NBD)]CIO, CH,CH(OH)CH(OH)CH,

(h)

meso56 D,L:44

a-Hydroxy ketones generally reduce to a diol mixture, but a-amino ketones are reduced more stereospecifically to amino-alcohols. These have the erythro configuration, through a cyclic intermediate involving the carbony, the amino group, and the metall':

P-Hydroxy ketones and enamino ketones are reduced to the 1,3-diol and to the 1,3-amino alcohol with very high selectivity in favor of the syn isomer'8319. Catalytic hydrogenation of the carbonyl group of benzoylformic acid and pyruvic acid bound to an optically active alcohol [( - )-menthol, ( + )-borne011 or an optically active amine or amino isobutylester yields chiral mandelic acid and lactic acid". The e.e. (enantiomeric excess) is up to 96% at low T [equation u)]". Catalysts are RO,, finely divided Ni (Raney Ni) or various forms of Pd. The reaction shows a marked solvent effect.

188


CH3COCOzH

I + H,N-CH-Ph-

Dcc

CH3

I

CH3COCONH-CH-Ph

H,, Pd-C 30"c

OH

CH3

I I P CH3-CH-CONH-CH-Ph

(S)

(S)

(j)

(S)

e.e. 96% Addition of H, to a prochiral carbonyl double bond using catalysts modified with optically active substances yields a chiral alcohol. Despite intensive study of various chiral catalysts or ligands", synthetic utility seems limited. Optical yields are highly dependent on many variables. Solid catalysts can be used, after being impregnated with chiral modifiers such as amines and a-hydroxy acids23. Ketones having a second functionality either a or p to the carbonyl and able to coordinate to the metal center give best optical yields. This suggests that direct chelation of the substrate to the metal surface occurs. The reaction is used to produce a-and p-hydroxy acids or esters and to reduce P-diketones. Rather drastic conditions of P and T limit its scope to simple compounds. Modification of finely divided Ni (Raney Ni) by working at 100°C in aqueous chiral tartaric acid and sodium bromide produces the best heterogeneous catalytic system for asymmetric hydrogenation of acetylacetone [to (2R,4R)- or (2R,4S)-pentane-2,4-diol], 4-hydroxybutanone [to (R)- or (S)-1,3-butanediol], and methyl acetoacetateZ4:

0 L C O O E t CH,

H,, 11 X lo3 kPa, 100°C finely divided NI,tartrate, NaBr

'

OH &COOEt CH,

(k)

e.e. 92%

Ultrasound treatment of Raney Ni prior to modification results in a much faster hydrogenation rate and better enantiodifferentiation in the reduction of 1,3-dienes and 3-0xoalkanoates~~. Other transition metal catalysts modified with tartaric acid have been used26.Tartaric acid is not a good modifier with Pt. Cinchona alkaloids efficiently modify Pt, Pd, Rh, and reduce the carbonyl group of a-ketoesters":

e.e. 95% Catalytically active soluble chiral metal complexes containing a relatively cheap metal such as Co give interesting enantioselectivity in the hydrogenation of 1,2diketonesZ8.Bio dimethylglyoximato Co(I1) 3 (cobaloxime), associated with a cocatalyst and possessing a chiral amine function such as quinine, quinidine, cinchonidine, ephedrine, brucine, or S-(- )-a-methylbenzylamine, forms a chiral system that induces preferentially one enantiomer, as in enzymatic systems:


189

COR" /H..

0

I

R'-CH-CO-R'

..

I

3

''.H/

I

R'-C-CH(0H)-CO-R"

I

32-99% e.e. 62%

0

- quinine Q

+

(m)

OH product of dimerization

-

I

Q

However, soluble metal chiral complex catalysts are better, even though ketone hydrogenation under these conditions is more difficult to run than olefin hydrogenation. Notable achievements have been made using Rh and Ru metals. Phosphine ligands are often used, which contain either an asymmetric alkyl group have the phosphorous atom as the asymmetric center, or have an axial element of chirality ( a t r o p i s o m e r i ~ m ) ~Other ~,~~. chiral ligands commonly used are derived from amino acids, from L-hydroxyproline and from ferrocene.

DIOP

CAMP

pphz

R = Ph BPPM R = c - C ~ H , , BCPM PPh, PPh, BINAP

BPPFOH

Methyl aryl ketones are hydrogenated to the ( R ) secondary alcohol using a Rh/( Diop/Et,N complex in e.e. 80%:

1

+ )-

J

CH,

Ar

Ar

= Ph

H,,[Rh(norbomadiene)Cl],

1-naphthyl

(+)-DIOP, Et,N

'

CH,

Ar

(n)

e.e. 80% ref. 30 e.e. 84% ref. 31

Best results are attained with ketones having a second group capable of coordination to the catalyst. Methyl acetylacetate is reduced in 77% e.e. with a chiral complex of Rh and CAMP3*.Hydrogenation of open-chain a-keto-esters in the presence of Rh(1) com-


190

plexes formed with biphosphine ligands BPM and BCPM (derived from natural L-hydroxyproline) gives chiral lactates33334. The optical yield depends on solvent; benzene is recommended. Neutral cationic DIOP-complexed Rh-catalysts afford hydroxylactame under mild conditions and with no double asymmetric induction3?

qNS”

Rh Cy,-(R,R)-DIOP, THF H,, 100 Ha, 20°C

Ph

0

Ph

0

COOMe

100%

COOMe

72% e.e.

Asymmetric reduction of ketopantoyl lactone is an effective biomimetic route to R)-pantothenic

( - )-pantolactone, an intermediate in the synthesis of the important D-( acid (a component of vitamins B and of coenzyme A):

0

100%

87% e.e. 92% e.e.

[Rh(COD)Cl],-BPPM [Rh(COD)Cl],-BCPM

+

0 ref. 36 ref. 34,37

Matching of ligand to substrate is vital to the optimization of optical yields. The corresponding reduction using the chiral ligand (-)-DIOP results in only 35% e.e. Peralkyl diphosphine ligands enhance both activity and selectivity of Rh catalysts3*.Additional constraints within the transition state arising from an interaction between the ligand and substrate can have a marked and beneficial effect. Thus, hydrogenation of pyruvic to lactic acid using the hydroxyl containing ferrocene compound BppfOH affords optical yields up to 83%. Use of the corresponding dimethylamino derivative, which contains no polar group, in addition to the phosphines that can interact with the carbonyl group in pyruvic acid, gives only 16% e.e.39. Chiral Rh complexes of hydroxyalkyl ferrocenyl phosphine BppfOH catalyze the asymmetric hydrogenation of aminomethyl aryl ketones to 2-amino- 1-aryl ethanols. Both occur in high chemical and optical yield4’:

Ar

-‘II

-cHzNHR*HC1

0

H,, RT [Rh(NBD)-BPPFOH] C10,

’

6

Ar - -CH,NHR.HCl

I

OH

(9)

up to 95% e.e. Conventional chiral hydride reagents are not very effective at forming these intermediates in the synthesis of adrenergic drugs, because active hydrogens are present on the starting ketones and the latter is unstable under basic conditions. The homogeneous hydrogenation of functionalized carbonyl compounds gives remarkable results with catalysts of the type RuX,L (X = halide, L = atropisomeric ligands). These bis(triaIky1 phosphine) Iigands have a C, axis. BINAP is their main representative4’.


I

(R)-BINAP

191

BITAP

The catalyst is prepared conveniently from equimolar amounts of Ru(OAc), [BINAP] and HX4,. Another method uses the easily accessible complex (COD), Ru, (OCOCF,), as a precursor for in situ preparation of the dichloro complex with BINAP. The COD ligands are displaced at 40°C by atropisomeric diphosphine~~~. Both optical forms of BINAP are available. Ketones are reduced in predictable manner with maximal optical yields. P-Keto derivatives (esters, thioesters, amides) are hydrogenated to the secondary alcohol with 93% to 99% e.e. Ketones having a heteroatom (Y to the carbonyl (-OH, -NR,, -halide) also undergo clean, stereochemically directed reduction. Coordination of the carbonyl oxygen and the heteroatom of the other functional group to the metal forms a chelate ring responsible for the enantiomeric differentiation. This clean process favorably competes with the use of bakers’ yeast, and is applied to the synthesis of natural products like carnitine and compactine4:

camitine Diastereoselective hydrogenation with BINAP-Ru combines chirality transfer from the catalyst and intramolecular asymmetric induction, providing an efficient entry to statine analogues45:

Re H,, RuBr, [(S)-BINAP]

NHBoc

,

R

NHBoc

.

+

99:1

R+COOEt OH

(s)

NHBoc

Dynamic kinetic resolution occurs during BINAP-Ru catalyzed hydrogenation of 2-substituted 3-0x0 carboxylic esters [equation (t)Ia and 2-acylamino-3-oxobutyrates47.

91% e.e. syn 97.7% anti 2.3% Cationic complexes in water-saturated CH,CI, or MeOH also yield the syn product48.


192

H,, 5 X lo3 kPa, 50°C, 40 h [BINAP-RuCIPh]*I-

NHCOPh

CH2CI,/H,0

’

%

COOMe

NHCOPh

99% e.e.

(u)

84% de

Other axially disymmetric biphenylbisphosphine ligands have been designed: 6,6’(diphenylphosphino)-3,3’-dimethoxy-2,2’,4,4’-tetrame~yl-l,l’-biphenyl (BIMOP)49 and trans bis(3-diphenylphosphini-phenyl)cyclopentane (BITAP)”. All show some efficiencies in ketone reduction. Chiral Ru complexes are successful in reducing even y-ketoesters, producing 4-substituted y-lactones in high optical purity5’:

R

COOEt

(1) H,, 104 Wa, EtOH 25°C BINAP-Ru(OAc), (2) AcOH, reflux

\

R=Me,96%, 99% e.e. (J.-L. GRAS)

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

D. H. R. Barton, J . Chem. SOC., 1027 (1953). R. J. Wicker, J. Chem. SOC., 2165 (1956). S . P. Findlay, J . Org. Chem., 24, 1540 (1959). K. Hanaya, Bull. Chem. SOC.Jpn., 43,442 (1970). T. F. Buckly 111, H. Rappoport, J . Org. Chem., 48,4222 (1983). M. Ishige, M. Shiota, Can. J. Chem., 53, 1700 (1975). S. Nishimura, M. Katagiri, Y. Kunikata, Chem. Lett., 1235 F(1975). S.Nishimura, M. Ishige, M. Shiota, Chem. Lett., 963 (1977). J. Solodar, J . Org. Chem., 41, 3461 (1976). S. Mitsui, H. Saito, Y. Yamashita, M. Kaminaga, Y. Senda, Tetrahedron, 29, 1533 (1973). S. Nishimura, M. Ishige, M. Shiota, Chem. Lett., 535 (1977). H. 0. House, V. Paragamian, D. J. Wluka, J . Amer. Chem. SOC., 82,2561 (1960). M. Mokotoff, J. Org. Chem., 33, 3557 (1968). H. 0. House, H. C. Muller, C. G. Pitt, P. P. Wickham, J. Org. Chem., 28, 2407 (1963). R. H. Hasek, E. U. Elam, J. C. Martin, R. G. Nations, J . Org. Chem., 26,700 (1961). K. Tani, K. Suwa, E. Tanigawa, T. Yoshida, T. Okano, S . Otsuka, Chem. Lett., 261 (1982). T. Matsurnoto, T. Nishida, H. Shirahama, J . Org. Chem., 27, 79 (1962). F. G. Kathawala, B. Prager, K. Prasad, 0. Repic, M. J. Shapiro, R. S. Stabler, L. Widler, Helv. Chim. Acta, 69, 803 (1986). Y. Matsumura, J. Fujiwara, K. Maruoka, H. Yamamoto,J. Am. Chern. SOC., 105,6312 (1983). K. Harada, Asymmetric Synthesis, Vol. 5 , J. D. Momson ed., Academic Press, New York, 1985, p. 345. K. Harada, T. Munegumi, S . Nomoto, Tetrahedron Lett., 22, 111 (1981). For a review of various chiral ligands see V. Caplar, G. Comisso, V. Sunsic, Synthesis, 85

(198 1). 23. H. U. Blaser, Tetrahedron Asym., 2, 843 (1991). 24. Y. Izumi, Adv. Catal., 32, 3215 (1983). 25. A. Tai, T. Kikukawa,T. Sugimura, Y. Inoue, T. Osawa, S . Fuji, J. Chem. SOC.,Chem. Commun., 795 (1991). 26. H. Brunnel, Synthesis, 645 (1988). 27. H. U. Blaser, H. P. Jalett, J. Wiehl, J . Mol. Card., 68, 215 (1991). 28. Y. Ohgo, S. Takeuchi, Y. Narori, J. Yoshimura, Bull. Chem. SOC. Jpn., 54, 2124 (1981). 29. H. Kagan, in Asymmetric Synthesis, Vol. 5 , J. D. Momson ed., Academic Press, New York, 1985, p. 1.

14.3. Hydrogenation Reactions 14.3.6. Hydrogenation of C=O Functions 14.3.6.3. In Carboxyl Derivatives

193

30. 31. 32. 33. 34. 35. 36. 37.

Sz. Toros, B. Heil, L. Kollar, L. Marko, J. Organomet. Chem., 197, 85 (1980). I. Osima, T. Kogure, K. Achiwa, J. Chem. SOC.,Chem. Commun., 428 (1977). J. Solddar, Chemtech., 421 (1975). K. Achiwa, Tetrahedron Lett., 3735 (1977). H. Takamashi, T. Morimoto, K. Achiwa, Chem. Lett., 855 (1987). K. Tani, E. Tanigawa, Y. Tatsuno, S . Otsuka, Chem. Lett., 737 (1986). I. Ojima, T. Kogure, J. Organomet. Chem., 195, 239 (1980). H. Takahachi, M. Hattori, M. Chiba, T. Morimoto, K. Achiwa, Tetrahedron Lett., 27, 4477

38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

C. Hatat, A. Krim, A. Mortreux, F. Petit, Tetrahedron Lett., 29, 3675 (1988). T. Hayachi, T. Mise, M. Kumada, Tetrahedron Lett., 4351 (1976). T. Hayachi, A. Katsumura, M. Konishi, M. Kumada, Tetrahedron Lett., 425 (1979). R. Noyori, Chem. SOC. Rev., 18, 187 (1989). Y. Motita, M. Suzuki, R. Noyori, J. Org. Chem., 54, 1785 (1989). B. Heiser, E. A. Broger, Y. Crameri, Tetrahedron Asym., 2.51 (1991). M. Kitamura, T. Ohkuma, N. Takaya, R. Noyori, Tetrahedron Lett., 29, 1555 (1988). T. Nishi, M. Kitamura, T. Ohkuma, R. Noyori, Tetrahedron Lett., 29, 6327 (1988). M. Kitamura, T. Ohkuma, M. Tolunaga, R. Noyori, Tetrahedron Asym., 1, 1 (1990). J.-P. Genet, C. Pinel, S.Mallart, S. Juge, S.Thorimbert, J. A. Laffitte, Tetrahedron Asym., 2, 555 (1991). K. Mashima, Y. Matsumura, K. Kusano, H. Kumobayashi, N. Sayo, Y. Hori, T. Ishizaki, S. Akutagawa, H. Takaya, J. Chem. SOC., Chem. Commun., 609 (1991). N. Yamamoto, M. Murata, T. Morimoto, K. Achiwa, Chem. Parm. Bull., 39, 1085 (1991). N. Fukuda, K. Mashima, Y. I. Matsumura, H. Takaya, Tetrahedron Lett., 31,7185 (1990). T. Ohkuma, M. Kitamura, R. Noyori, Tetrahedron Lett., 31,5509 (1990).

48. 49. 50. 51.

(1986).

14.3.6.3. In Carboxyl Derivatives

Reactivity of various carboxyl derivatives decreases in the order, acid chlorides > (aldehydes, ketones) > anhydrides > esters > carboxylic acids > amides'. Carboxylic acids are reduced to alcohols under vigorous conditions. High P (35 X lo3 to 70 X lo3 Wa) and elevated T (up to 230°C) are required to effect conversion in high yields. Ruthenium catalysts, like Ru-on-carbon or RuO,, are most often used, in water as solvent. They reduce dicarboxylic acids or a-hydroxy acids to the corresponding glycols2: HOCHZCOOH

H,,40 X lo3 kPa, 150°C 10% Ru-C, H,O

HOCH,CH,OH

80%

(a)

Several oxides of Re (Re,O,, ReO,, ReO,, ReO) effectively reduce carboxylic acids to primary alcohols3, without saturation of the ring in aromatic carboxylic acids. Activated copper-barium-chromium oxide in dioxane is used at higher temperatures4. Strong acids (HClO,), tertiary amines, and water are promoters for this hydrogenation. When conducted in acetic acid, the reduction may require more than the theoretical amount of H, due to partial reduction of the solvent, especially over Rh or PtO, catalysts'. The carboxyl group is unaffected in the hydrogenation of unsaturated acids. Esters are reduced reluctantly and make very good solvents for the reduction of various functional groups. In substrates with certain structural feature, esters can be hydrogenolyzed to acids, alcohols, ethers, or hydrocarbons, depending mainly on the substrate.



193

30. 31. 32. 33. 34. 35. 36. 37.

Sz. Toros, B. Heil, L. Kollar, L. Marko, J. Organomet. Chem., 197, 85 (1980). I. Osima, T. Kogure, K. Achiwa, J. Chem. SOC.,Chem. Commun., 428 (1977). J. Solddar, Chemtech., 421 (1975). K. Achiwa, Tetrahedron Lett., 3735 (1977). H. Takamashi, T. Morimoto, K. Achiwa, Chem. Lett., 855 (1987). K. Tani, E. Tanigawa, Y. Tatsuno, S . Otsuka, Chem. Lett., 737 (1986). I. Ojima, T. Kogure, J. Organomet. Chem., 195, 239 (1980). H. Takahachi, M. Hattori, M. Chiba, T. Morimoto, K. Achiwa, Tetrahedron Lett., 27, 4477

38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

C. Hatat, A. Krim, A. Mortreux, F. Petit, Tetrahedron Lett., 29, 3675 (1988). T. Hayachi, T. Mise, M. Kumada, Tetrahedron Lett., 4351 (1976). T. Hayachi, A. Katsumura, M. Konishi, M. Kumada, Tetrahedron Lett., 425 (1979). R. Noyori, Chem. SOC. Rev., 18, 187 (1989). Y. Motita, M. Suzuki, R. Noyori, J. Org. Chem., 54, 1785 (1989). B. Heiser, E. A. Broger, Y. Crameri, Tetrahedron Asym., 2.51 (1991). M. Kitamura, T. Ohkuma, N. Takaya, R. Noyori, Tetrahedron Lett., 29, 1555 (1988). T. Nishi, M. Kitamura, T. Ohkuma, R. Noyori, Tetrahedron Lett., 29, 6327 (1988). M. Kitamura, T. Ohkuma, M. Tolunaga, R. Noyori, Tetrahedron Asym., 1, 1 (1990). J.-P. Genet, C. Pinel, S.Mallart, S. Juge, S.Thorimbert, J. A. Laffitte, Tetrahedron Asym., 2, 555 (1991). K. Mashima, Y. Matsumura, K. Kusano, H. Kumobayashi, N. Sayo, Y. Hori, T. Ishizaki, S. Akutagawa, H. Takaya, J. Chem. SOC., Chem. Commun., 609 (1991). N. Yamamoto, M. Murata, T. Morimoto, K. Achiwa, Chem. Parm. Bull., 39, 1085 (1991). N. Fukuda, K. Mashima, Y. I. Matsumura, H. Takaya, Tetrahedron Lett., 31,7185 (1990). T. Ohkuma, M. Kitamura, R. Noyori, Tetrahedron Lett., 31,5509 (1990).

48. 49. 50. 51.

(1986).

14.3.6.3. In Carboxyl Derivatives

Reactivity of various carboxyl derivatives decreases in the order, acid chlorides > (aldehydes, ketones) > anhydrides > esters > carboxylic acids > amides'. Carboxylic acids are reduced to alcohols under vigorous conditions. High P (35 X lo3 to 70 X lo3 Wa) and elevated T (up to 230°C) are required to effect conversion in high yields. Ruthenium catalysts, like Ru-on-carbon or RuO,, are most often used, in water as solvent. They reduce dicarboxylic acids or a-hydroxy acids to the corresponding glycols2: HOCHZCOOH

H,,40 X lo3 kPa, 150°C 10% Ru-C, H,O

HOCH,CH,OH

80%

(a)

Several oxides of Re (Re,O,, ReO,, ReO,, ReO) effectively reduce carboxylic acids to primary alcohols3, without saturation of the ring in aromatic carboxylic acids. Activated copper-barium-chromium oxide in dioxane is used at higher temperatures4. Strong acids (HClO,), tertiary amines, and water are promoters for this hydrogenation. When conducted in acetic acid, the reduction may require more than the theoretical amount of H, due to partial reduction of the solvent, especially over Rh or PtO, catalysts'. The carboxyl group is unaffected in the hydrogenation of unsaturated acids. Esters are reduced reluctantly and make very good solvents for the reduction of various functional groups. In substrates with certain structural feature, esters can be hydrogenolyzed to acids, alcohols, ethers, or hydrocarbons, depending mainly on the substrate.

194


Hydrogenolysis of an ester to the alcohol is a reversible reaction commonly obtained at high T and P however, with massive quantities of catalyst the reaction proceeds below 100°C. Noble metal catalysts are not effective for this purpose6 and hydrogenation to the alcohol usually is done over activated copper-barium-chromite in dioxane'. At relatively low T a-hydroxy, P-hydroxy, P-keto, and P-carbethoxy esters are reduced to the glycols; amides, acids, and the benzene ring are not affected: 0

oacootBu H,,35 X lo3kPa 150°C Cu-CrO, dioxane

*

&OH

(b)

95%

Zinc-chromium oxide and massive amounts of finely divided Ni (W4 or W6 Raney Ni) in ethanol are effective*, but saturation of a benzene ring occurs preferentially. Unsaturated esters are reduced selectively to unsaturated alcohols using a zinc oxide-chromium oxide catalyst, and more generally with zinc or copper promoted with Ca, Cd, or Co9. Hydrogenation of butyrolactone to l,.l.-butanediol, a component of polyesters, occurs in high yields over copper chromite [equation (c)]l0 or a Ni-Co catalyst".

CoL0

cu-cro

H0(CH2)40H 89-98%

Hydrogenolysis of esters to acids can occur provided the R'-0 bond is weakened, for instance when R' is a vinyl, allyl, or aryl group [equation (d)]12. Thus, hydrogenolysis of an enol ester is a synthetic method for facile removal of carbonyl oxygen^'^ that applies to enol trifluoromethanesulfonates [equation (e)] 14:

Yi

14.3. H drogenation Reactions 14.3.6. ydrogenation of C=O Functions 14.3.6.3. In Carboxyl Derivatives

195

Hydrogenolysis to acids occurs also when R' is tertiary, in the presence of strong acidsI5 or in bromo esters like steroid 1 [equation (f)]16. Both reactions proceed via an alkene intermediate.

Lactones reduce more easily than acyclic esters. Lactone 2 with an activated C-0 bond is hydrogenolyzed to dimethyl 2-methyl-terephthalate over Pd17:

Lactones may show some unusual cleavage of oxygen-carbonyl bonds to aldehydes. Sugar lactones are hydrogenated to aldoses over platinum at atm P and RT", and 3 leads to amide 4 over palladium-on-carbon'9:

Ph 4

Ph 3

Several Slactones hydrogenolize to the cyclic ether under mild conditions (RT, 100 Wa) over large quantities of platinum oxide in acetic acid with a catalytic amount of HC10,6 [equation (i)I2'. The reaction is applicable to 4-oxa-3-oxo-5a-cholestane (giving 92% of 4-oxa-5a-cholestane) but not to y- or elactones. However, butyrolactone yields tetrahydrofuran through hydrogenolysis over a Co-Re catalyst in dioxane". AcOHPtOZHCIO,'

0

a

80-90%

(9

Hydrogenolysis and reduction of esters or lactones to hydrocarbons are difficult. The reaction is easy with esters of aromatic acids and pyridinol or quinolino12*:

Qo-coph

-2

Q

OH

+CH3Ph

Acid chloride hydrogenation to aldehydes is an important synthetic reaction conveniently carried out using some improved procedures.

196

14.3. Hydrogenation Reactions 14.3.6. Hydrogenationof C=O Functions 14.3.6.3. In Carboxyl Derivatives

Palladium-on-barium sulfate in q ~ i n o l i n eor ~ ~palladium-on-carbon alonez4 or associated with ethyldiisopropylamine as the HC1 acceptoP is an efficient catalyst to produce aldehydes under very mild conditions: RCOCl

R

=

H,, 100 kPa, RT

Pd-C, EtN(i-Pr),

9

RCHO

alkyl, benzyl, aryl

Anhydrides hydrogenate with relative ease (high catalyst loading prolonged reduction) to various products depending on the substrate, on the catalyst, on reactions conditions, and on the solvent26.Complete hydrogenation over copper chromite yields glycols, but partial reduction is possible2'. Reduction of succinic anhydrides is of great synthetic utility since they are easily available from Diels-Alder additions. Substituted succinic anhydride 10, hydrogenated over platinum oxide in ethyl acetate, afford hydroxylactone 11. The prolonged reaction affords a mixture of lactone 12 and acid 1328: ?H

PtO,, EtOAc RT, 300 kPa

PtO, AcOH

AcO

+*

lo

p

(1)

COOH

AcO 12> -

AcO

13

Hydroxylactone 11 yields 12 and 13; 12 is not hydrogenolyzed to 13 under these conditions. Hydrogenations of maleic and succinic anhydrides produce butyrolactone, tetrahydrofuran, or 1,4-butanediol, all economically i m p ~ r t a n t ~ ~ , ~ ' :

+ CH3( ),CH20H + HOCH2( ),CH,OH 5

0 n = 2,3

7

(m)

6

Catalytic hydrogenation of the anhydrides 5 in presence of technetium catalysts efficiently yields the partially hydrogenated lactones 7, and in presence of rhenium catalysts the oxides 631.Especially interesting is the highly regioselective partial hydrogenation of substituted cyclic anhydrides. Hydrogenation occurs at the least hindered carbonyl group, in contrast to the reduction with LiAlH,, which affects the more hindered car-


197

bonyl. Hydrogenation of 2,2-dimethylsuccinic anhydride over technetium-black affords y-lactones 8 and 9 in a 6.6:l ratio3' and y-lactone 8 using R u C ~ , [ P ( P ~ ) , ] , ~ ~ :

0 Tc black Ru catalyst

0 9 1

8

6.6 10

-

Asymmetric induction occurs by enantioselective catalytic hydrogenation of a cyclic anhydride using an Ru(I1) chiral phosphine complex33:

R

R HZ RuzC1, (DIOP),

'

0 27-79%

e.e.: 5.4-20%

Amides are hydrogenolyzed to amines during hydrogenation over Cu chromite or

Ru under vigorous conditions. Other functional groups are selectively hydrogenated in

the presence of an amide. Some lactams are hydrogenolyzed over large amounts of platinum oxide, and HC1 as solvent34:

0

100% (J.-L. GRAS)

1 . A. J. McAlees, R. McCrindle, J . Chem. SOC. C , 2425 (1969). 2. J. E. Carnaghan, T. A. Ford, W. F. Gresham, W. E. Grigsby, G. F. Hager, J. Am. Chem. SOC., 77, 3766 (1955). 3. H. S. Broadbent, W. J. Bartley, J . Org. Chem., 2345 (1963). 4. A. Guyer, A. Bieler, K. Jaberg, Helv. Chim. Actu, 38, 976 (1955). 5. J. D. Chanley, T. Mezzetti, J . Org. Chem., 29, 228 (1964). 6. R. T. Rapala, E. R. Lavagnino, E. R. Shepard, E. Farkas, J. Am. Chem. Soc., 79,3770 (1957). 7. H. Adkins, H. R. Billica, J . Am. Chem. Soc., 70, 3121 (1948). 8. H. Adkins, Org. React., 8 , 1 (1954). 9. V. G. Cherkaev, N. V. Bliznyak, A. A. Bag, Tr Vses. Nuuchmo-Issled. Inst. Sint. Nut. Dushistykh Veshchestv, 234 (1968); Chem. Abstr. 71,3820 (1969). 10. A. Murata, H. Kobayashi, Ger. Patent 2,043,349 (1972); Chem. Abstr. 76, 126378s (1972). 1 1 . M. Yamaguchi, Y. Kageyama, Ger. Patent 2,144,316 (1972); Chem. Abstr. 76, 12638Om (1972). 12. T. Sakuragi, J . Org. Chem., 23, 129 (1958). 13. M. Harnick, U.S.Patent 3,107,256 (1963); Chem. Abstr., 60, 1824f (1964). 14. V. B. Jigajinni, R. H. Wightman, Tetrahedron Lett., 23, 117 (1982). 15. P. E. Peterson, C. Casey, J. Org. Chem., 29, 2325 (1964). 16. S . G. Levine, M. E. Wall, J . Am. Chem. SOC., 81,2829 (1959).

198

14.3. Hydrogenation Reactions 14.3.6. Hydro enation of C=O Functions 14.3.6.4. By ransfer Hydrogenation

9

17. A. T. Jurewicz, L. S. Fomey, US.Patent 3,651,126 (1972);Chem. Abstr., 77, 20305W (1972). 18. 0. T. Schmidt, H. Miiller, Chem. Ber., 76, 344 (1943). 19. J. C. Martin, K. C. Brannock, R. D. Burpitt, P. G. Gott, V. A. Hoyle, Jr., J . Org. Chem., 36, 2211 (1971). 20. J. T. Edward, J. M. Ferland, Chem. Ind., 975 (1964). 21. H. Hiari, K. Miyata, Jpn. Patent 72, 42,832 (1970); Chem. Abstr. 78, 1 lllOt (19973). 22. C. J. Cavallito, T. H. Haskel1,J. Am. Chem. SOC., 66, 1166 (1944). 23. A. I. Scott, F. McCapra, R. L. Buchanan, A. C. Day, D. W. Young, Tetrahedron 21, 3605 (1965). 24. A. I. Rachlin, H. Gurien, D. P. Wagner, Org. Syn., 51, 8 (1971). 25. J. A. Peters, H. Van Bekkum, Rec. Trav. Chim. Pays-Bas, 90 1323 (1971). 26. T. A. Eggelte, H. De Konning, H. 0. Huisman, Tetrahedron, 29, 2445 (1973). 27. R. Kuhn, I. Butula, Justus Liebigs Ann. Chem., 718, 50 (1968). 28. R. McCrindle, K. H. Overton, R. A. Raphael, J . Chem. SOC.,4799 (1962). 29. B. Miya, F. Hoshino, T. Ono, Chem. Abstr., 82,431128 (1975). 30. J. Kanetaka, S. Mori, Jpn. Patent 72, 34,180 (1972); Chem. Abstr. 78, 1 1 1104 (1973). 31. B. Bayerl, M. Wahren, Zeir. Chem., 21, 149 (1981). 32. P. Morand, M. Kayser, J . Chem. SOC.,Chem. Commun.,314 (1976). 33. K. Osakada, M. Obana, T. Ikariya, M. Saburi, S. Yoshikawa, Tetrahedron Lett., 22, 4297 (1981). 34. F. Galinovsky, E. Stem, Chem. Ber., 76, 1034 (1943).

14.3.6.4. By Transfer Hydrogenation

Reduction of multiple bonds using an organic molecule as the hydrogen donor and a metal catalyst, known as catalytic transfer hydrogenation, can be applied to reduction of carbonyl groups under specific conditions. Aldehydes are easily bound to the metal and may decrease the catalytic activity of some Rh or Ru catalysts, e.g., RhCl[P(Ph),], is transformed by reaction with aldehydes to RhCl(CO)[P(Ph),],, which has no catalytic activity'. So far, the transfer hydrogenation of aldehydes offers few synthetic advantages. However, an unusual selective reduction of a$-unsaturated aldehydes to unsaturated alcohols is achieved using the hydrioiridium sulfoxide catalyst [Ir(H)Cl,(Me,SO),] in isopropanol as the hydrogen source. Under mild conditions, cinnamaldehyde, a-methylcinnamaldehyde, and crotonaldehyde are reduced to the corresponding unsaturated alcohol with >80% selectivity at 90% conversion'. Preferential reduction of the carbonyl function occurs for electronic rather than steric reasons. Complete hydrogenolysis of cu,S-unsaturated aldehydes to methyl derivatives is achieved with high selectivity in steroids?

5% Pd-C cyclohexene

(a)

Me0 CHO Dihydridotetrakis (triphenylphosphine) ruthenium(II), RuH, [P(Ph)3]4,exhibits excellent catalytic activity under mild conditions, transferring hydrogen from ethers, decaline, tertiary amines, and alcohols to aldehydes in about 50% yields'. Coordination of the


198


9

17. A. T. Jurewicz, L. S. Fomey, US.Patent 3,651,126 (1972);Chem. Abstr., 77, 20305W (1972). 18. 0. T. Schmidt, H. Miiller, Chem. Ber., 76, 344 (1943). 19. J. C. Martin, K. C. Brannock, R. D. Burpitt, P. G. Gott, V. A. Hoyle, Jr., J . Org. Chem., 36, 2211 (1971). 20. J. T. Edward, J. M. Ferland, Chem. Ind., 975 (1964). 21. H. Hiari, K. Miyata, Jpn. Patent 72, 42,832 (1970); Chem. Abstr. 78, 1 lllOt (19973). 22. C. J. Cavallito, T. H. Haskel1,J. Am. Chem. SOC., 66, 1166 (1944). 23. A. I. Scott, F. McCapra, R. L. Buchanan, A. C. Day, D. W. Young, Tetrahedron 21, 3605 (1965). 24. A. I. Rachlin, H. Gurien, D. P. Wagner, Org. Syn., 51, 8 (1971). 25. J. A. Peters, H. Van Bekkum, Rec. Trav. Chim. Pays-Bas, 90 1323 (1971). 26. T. A. Eggelte, H. De Konning, H. 0. Huisman, Tetrahedron, 29, 2445 (1973). 27. R. Kuhn, I. Butula, Justus Liebigs Ann. Chem., 718, 50 (1968). 28. R. McCrindle, K. H. Overton, R. A. Raphael, J . Chem. SOC.,4799 (1962). 29. B. Miya, F. Hoshino, T. Ono, Chem. Abstr., 82,431128 (1975). 30. J. Kanetaka, S. Mori, Jpn. Patent 72, 34,180 (1972); Chem. Abstr. 78, 1 1 1104 (1973). 31. B. Bayerl, M. Wahren, Zeir. Chem., 21, 149 (1981). 32. P. Morand, M. Kayser, J . Chem. SOC.,Chem. Commun.,314 (1976). 33. K. Osakada, M. Obana, T. Ikariya, M. Saburi, S. Yoshikawa, Tetrahedron Lett., 22, 4297 (1981). 34. F. Galinovsky, E. Stem, Chem. Ber., 76, 1034 (1943).

14.3.6.4. By Transfer Hydrogenation

Reduction of multiple bonds using an organic molecule as the hydrogen donor and a metal catalyst, known as catalytic transfer hydrogenation, can be applied to reduction of carbonyl groups under specific conditions. Aldehydes are easily bound to the metal and may decrease the catalytic activity of some Rh or Ru catalysts, e.g., RhCl[P(Ph),], is transformed by reaction with aldehydes to RhCl(CO)[P(Ph),],, which has no catalytic activity'. So far, the transfer hydrogenation of aldehydes offers few synthetic advantages. However, an unusual selective reduction of a$-unsaturated aldehydes to unsaturated alcohols is achieved using the hydrioiridium sulfoxide catalyst [Ir(H)Cl,(Me,SO),] in isopropanol as the hydrogen source. Under mild conditions, cinnamaldehyde, a-methylcinnamaldehyde, and crotonaldehyde are reduced to the corresponding unsaturated alcohol with >80% selectivity at 90% conversion'. Preferential reduction of the carbonyl function occurs for electronic rather than steric reasons. Complete hydrogenolysis of cu,S-unsaturated aldehydes to methyl derivatives is achieved with high selectivity in steroids?

5% Pd-C cyclohexene

(a)

Me0 CHO Dihydridotetrakis (triphenylphosphine) ruthenium(II), RuH, [P(Ph)3]4,exhibits excellent catalytic activity under mild conditions, transferring hydrogen from ethers, decaline, tertiary amines, and alcohols to aldehydes in about 50% yields'. Coordination of the


199

?

donor to the complex is rate determining as suggested by the kinetic isotope effect (RH/RD = 0.9). Transfer of hydrogen from the donor on the metal occurs last. Transfer hydrogenation of ketones is of greater applicability. Heterogeneous catalysts are seldom successfully used: keto-groups are not reduced over the commonly used Pd-C catalyst, with the exception of p-quinone and benzil due to their low reduction potentials4. A Lewis acid promoter (FeCl,) is necessary for the reduction of aromatic aldehydes or ketones by hydrogen transfer from cyclohexane or limonene using Pd on carbon as a catalyse. Finely divided Ni (Raney Na) catalyzes hydrogen transfer from alcohols to phenyl ketones to form the hydrogenolyzed methylene compounds6. The Benzyl alcohol is obtained with unsolvated diorganomagnesium amide (41-80% yield)'. Transfer hydrogenation of ketones to the corresponding alcohols can be completed over homogeneous catalysts made from Co, Ru, Rh, Re, Ir complexes containing phosphines, dimethyl sulfoxide, or nitrogen donor ligands. Mainly, secondary carbinols are used as hydrogen donors. The equilibrium indicated in equation (b) is displaced, providing that ketone 1 and ketone 2 have significantly different oxidation potentials. ketone 1

+ carbinol 2

carbinol 1

+ ketone 2

(b)

Tris(tripheny1phosphine) cobalt trihydride can transfer hydrogen atoms between cyclanones or acetophenone and isopropanol. The cobalt hydride reacting with a ketone gives LxCo-OR, and with alcohol it gives LxCoH . . . ROH, the cobalt alkoxide being the probable catalyst for hydrogen migration'. Transfer of hydrogen from secondary carbinols to saturated ketones is also catalyzed by Ru catalysts': PhCH,COCH,Ph

+ PhCH(OH)CH,

RuCI,(PPh,), h, 180"c

9

(PhCH,),CHOH 93%

+ CH,COPh

(c)

Here, the initial formation of an Ru-ketone complex is followed by fast addition to form an Ru-alkoxide. Migration of an a-hydrogen atom from the alkoxy ligand to the bound ketone, followed by protonation by the solvent, leads to extrusion of a carbinol molecule. The rate depends on the ease of metal hydride formation by abstraction of an a-hydrogen atom from the donor, and not on the nature of the ketone. In the presence of an aldehyde or primary alcohol as precursor, reduction of a keto group does not occur'o. However, a wide variety of ketones and aldehydes can be reduced by formic acid, at 125"C, in the presence of ruthenium complexes: RuCl,(PPh,),, RuCl(CO)(PPh,),, Ru(CO),(PPh,),, Ru(CO)1,.9PPh,, Ru(C0) 12". Some soluble Rh complexes that contain basic phosphines as ligands (e.g. PPh,Me, PPhMe,, PMe,) also catalyze reduction of ketones under mild conditions when promoted by small quantities of water',. Coordination of the ketone with Rh, followed by 1,3-hydride migration from the metal to the carbon atom of the keto-group affords an Rh alkoxide. The second proton transfer is carried out by a hydroxyl ion and a water molecule: R2OH H-[Rh]OCHR, + (d) R2C= 0 + [RhH,(PPhMe,),L,]+- Hz'Hzo > G-OH catalyst HO-

F

b

-

200

14.3. Hydrogenation Reactions 14.3.6. Hydro enation of C=O Functions 14.3.6.4. By jansfer Hydrogenation ~

~~~

Reduction from isopropanol to cyclohexanones by hydrogen transfer can be catalyzed by transition metal compounds-tin chloride systems such as the salts mixtures: SnClz~2HzO/MC13~3Hz0, (M=Ir, Rh) and H31rC16.nH20/SnC1z~2Hz0'3. Ketones are reduced to the secondary alcohols by isopropanol with RuCl (PPh,), in the presence of sodium hydroxide',. Chloro complexes of Rh(II1) catalyze dehydrogenation of isopropanol to acetone via transfer of a hydride group from the a-carbon atom of the alcohol to the Rh complex. The rhodium hydride intermediate formed is stabilized by addition of tin chloride; there is no decomposition to give Rh metal under these conditions. Instead, the intermediate can react with a proton present in the solution to form hydrogen, and reduce the carbonyl group with yields up to 97%. Complexes containing Ir metal are the most promising for transfer hydrogenation of ketones. Iridium complexes IrH5(PPh,) and IrH,(PPh,), in trifluoroacetic acid activate the hydrogen atom to hydride ion, and reduce the carbonyl group of ketones to secondary alcohols, esterified in the reaction medium to trifluoroacetic ester^'^: Ir cat. ___,

R'COR2 CF,COOH R'CHR'

I

54-68%

1

OCOCF, The reaction is sensitive to steric hinderance. Aromatic ketones are reduced to hydrocarbons. Unsaturated ketones are fully reduced and with no selectivity. Complexes of the type Ir(Chel)(CH2=CH2),C1, with Chel = 2,2'-bipyridine or phenantholine derivatives, behave as catalyst precursors for hydrogen transfer from isopropanol to ketones and Schiff based6. Potassium hydroxide is required as cocatalyst to convert the isopropanol coordinated to the Ir(1) ion, in the neutral isopropoxy derivative. Enolates that are present would act as inhibitors when coordinated to the cationic derivative. Ethylene complexes are better precursors than the corresponding cyclooctadiene derivatives, because they are activated more easily and more completely, and they show high catalytic activity. The most active complexes is the 3,4,7,8-Me, phen derivative, which, at 83"C, gives turnovers of up to 2850 cycles/min. Reduction of 4-t-butylcyclohexanone affords 97% of the trans-alcohol. Most reagents that reduce ketones to secondary carbinols give a preponderance of the more stable equatorial alcohol when applied to akyl-substituted cyclohexanones (relatively unhindered). Ruthenium or rhenium catalysts yield high proportions of trunsalcohols when used as hydrogen transfer catalysts and the trihydride triphenylphosphine complexes H,Ir(PPh,), and H,Ir(PPh,), also transfer hydrogenate 4-t-Bu-cyclohexanone to 88-90% of the trans-alcohol. The catalyst prepared from iridium chloride and phosphorus acid or an easily hydrolyzed ester give a highly specific method for the reduction of cyclohexanones, to give the cis-axial alcohols in yields exceeding 90%17. Several experiments indicate that phosphite is the reductant. The reaction is sensitive to steric hinderance, 2-methyl cyclohexanone being reduced more slowly than 4-methyl cyclohexanone. This method of reduction finds various synthetic applications such as for the preparation of 3a-hydroxy-5a- and 3P-hydroxy-5P-steroids. The chloroiridic acid and trimethyl phosphite system in 90% aqueous isopropanol is completely selective for reduction of 3-oxo-functions of steroids, in the presence of unprotected oxo-groups at any of the other skeletal positions where they are commonly found in natural steroids (C-6, C-11, C-12, C-17, and C-2O)l8:

+


20 1

1rCl6, P(OMe),

98% axial This stereoselectivity was also useful for synthesis of material reputed to possess sandalwood-like odors [equation (g)]”, and for the preparation of the axial alcohol

trans-9,1 O,cis-8,9-H-8-hydroxy-2-methyldecahydroisoquinoline20.

&-q* &-CJ 0

*

(g)

OH

Tris(tripheny1phosphine)RhCl associated with trimethyl phosphite in isopropanol gives even greater selectivity (100%)of the axial alcohol. Reductions are conveniently effected (24h at 82°C)in a sealed tube21. The reducing agent approaches from the less hindered side of the carbonyl and the mechanism does not proceed via the hydrogenation of an enol. Iridium(1) complexes with Schiff base ligands effectively transfer hydrogenate ketones. Use of optically active ligands allows asymmetric reduction of prochinal ketones to optically active alcohols22:

*

,PhCH(0H)i-h

Ph-CO--i-Pr

[Ir(COD) 4 or 51 ClO,

Me-CO-(CH,),Me

i-PrOH*

+ +

,

*

96%

MeCH(OH)(CH,),Me

80%

R(+)

(h)

S( +)

(i)

e.e.: 19.8% e.e.: 7.5%

4: ( )- and ( - )-2-pyridinalphenylethylimine 5: ( )- and ( - )-2-pyridinal-3-(iminomethyl)pinane

Alkyl phenyl ketones are reduced to the (R)alcohols with greater selectivity than the alkyl methyl ketones are reduced to the (S) alcohols. Optical yields vary depending on the chiral ligand (4 being superior to 5) and increase on increasing the [i-PrOH]/[substrate] ratio (the reverse reaction becoming less thermodynamically favored). Formation of catalytically active species from the complex requires displacement of the cyclooctadiene from the coordination sphere of the metal. The enantioselectivity observed is explained in terms of the equilibria between the diastereoisomeric forms of this catalytically active species. (J.-L. GRAS) 1 . H. Imai, T. Nishigushi, K. Fukuzumi, J. Org. Chem., 41,665 (1976). 2. B. R. James, R. H. Moms, J. Chem. Soc., Chem. Commun., 929 (1978).

14.3. Hydrogenation Reactions 14.3.7. Hydrogenationof Other Functional Groups 14.3.7.1 Nitriles

202

3. D.Bum, D.N. Kirk, V. Petrov, Tetrahedron, 21, 1619 (1965). 4. E. A. Braude, R. P. Linstead, K. R. H. Wooldridge, P. W. D.Mitchell, J . Chem. SOC., 3595 (1954). 5. G . Brieger, T.-H. Fu, J . Chem. SOC.,Chem. Commun.,757 (1976). 6. E. C. Kleiderer, E. C. Komfeld, J . Org. Chem., 13,455 (1948). 7. R. Sanchez, W. Scott, Tetrahedron Lett., 29, 139 (1988). 8. E. Malunowicz, S.Tyrlik, Z. Lasocki, J. Organomet. Chem., 72, 269 (1974). 9. Y. Sasson, J. Blum, J. Org. Chem., 40, 1887 (1975). 10. Y. Sasson, P. Albin, J. Blum, Tetrahedron Lett., 833 (1974). 11. Y. Watanabe, T. Ota, Y. Tsuji, Chem. Lett., 1585 (1980). 12. P. R. Schrock, J. A. Osbom, J. Chem. Soc., Chem. Commun., 567 (1970). 13. J. Kaspar, R. Spogliarich, M. Graziani, J . Organomet. Chem., 231,71 (1982). 14. R. L. Chowdhury, J. E. Backvall, J . Chem. SOC.,Chem. Commun., 1063 (1991). 15. M. I. Kalinskin, S. M. Markosyan, D.N. Kursanov, Z. N. Pames, 1. Akad. Nauk. SSSR, Ser. Khim., 675 (1981); Chem. Abstr., 95,41917j (1981). 16. F. Martineili, G. Mestroni, A. Camus, G . Zassinovich, J . Organomet. Chem.,270, 383 (1981). 17. H. B. Henbest, T. R. B. Mitchell, J . Chem. SOC., C, 785 (1970). 18. P. A. Browne, D.N. Kirk, J. Chem. SOC.,C , 1653 (1969). 19. W. I. Fanta, W. F. Erman, J . Org. Chem., 36, 358 (1971). 20. I. W. Mathison, P. H. Morgan, J. Org. Chem., 39, 3210 (1974). 21. J. C. Om,M. Mersereau, A. Sanford, J . Chem. SOC., Chem. Commun., 162 (1970). 22. G . Zassinovich, C. Del Bianco, G . Mestroni, J . Organomet. Chem., 222, 329 (1981).

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.1 Nitriles

Catalytic hydrogenation of nitriles is a complex process that can yield several final products deriving from a mechanism that assumes the formation of a key imine intermediate 1 as the first step'.'. Hydrogenation of imine 1 affords the usually desired primary amine 2. Addition of imine 1 and newly formed amine 2 gives a-amino amine 3, which can yield secondary amine 4 directly by hydrogenolysis, or it eliminates a molecule of ammonia to afford the Schiff base 5. This imine can be isolated when sterically hindered3, or hydrogenated to secondary amine 4. Imine 1 and secondary amine 4 can interact to form amino-amine 6,which yields tertiary amine 7 from hydrogenolysis. Finally, in proper aqueous media, aldehyde 8 may form directly by hydrolysis of the imine intermediate4.

R-CN

HZ

RCH'NH,

RCH(NH,)NHCH,R

+2

RCH=NH

RCHO 8

+ NH3

//

'L

3 l-NH3

-

U

(RCH,),NH

T 4

+ NH3

RCH=NCHzR

RCH(NH,)N(CH,R), 6

-

(RCH,),N 7

+ NH,

Fortunately, the reaction can be controlled and oriented through each of these paths, and is a very useful synthetic tool. (J.-L. GRAS)


14.3. Hydrogenation Reactions 14.3.7. Hydrogenationof Other Functional Groups 14.3.7.1 Nitriles

202

3. D.Bum, D.N. Kirk, V. Petrov, Tetrahedron, 21, 1619 (1965). 4. E. A. Braude, R. P. Linstead, K. R. H. Wooldridge, P. W. D.Mitchell, J . Chem. SOC., 3595 (1954). 5. G . Brieger, T.-H. Fu, J . Chem. SOC.,Chem. Commun.,757 (1976). 6. E. C. Kleiderer, E. C. Komfeld, J . Org. Chem., 13,455 (1948). 7. R. Sanchez, W. Scott, Tetrahedron Lett., 29, 139 (1988). 8. E. Malunowicz, S.Tyrlik, Z. Lasocki, J. Organomet. Chem., 72, 269 (1974). 9. Y. Sasson, J. Blum, J. Org. Chem., 40, 1887 (1975). 10. Y. Sasson, P. Albin, J. Blum, Tetrahedron Lett., 833 (1974). 11. Y. Watanabe, T. Ota, Y. Tsuji, Chem. Lett., 1585 (1980). 12. P. R. Schrock, J. A. Osbom, J. Chem. Soc., Chem. Commun., 567 (1970). 13. J. Kaspar, R. Spogliarich, M. Graziani, J . Organomet. Chem., 231,71 (1982). 14. R. L. Chowdhury, J. E. Backvall, J . Chem. SOC.,Chem. Commun., 1063 (1991). 15. M. I. Kalinskin, S. M. Markosyan, D.N. Kursanov, Z. N. Pames, 1. Akad. Nauk. SSSR, Ser. Khim., 675 (1981); Chem. Abstr., 95,41917j (1981). 16. F. Martineili, G. Mestroni, A. Camus, G . Zassinovich, J . Organomet. Chem.,270, 383 (1981). 17. H. B. Henbest, T. R. B. Mitchell, J . Chem. SOC., C, 785 (1970). 18. P. A. Browne, D.N. Kirk, J. Chem. SOC.,C , 1653 (1969). 19. W. I. Fanta, W. F. Erman, J . Org. Chem., 36, 358 (1971). 20. I. W. Mathison, P. H. Morgan, J. Org. Chem., 39, 3210 (1974). 21. J. C. Om,M. Mersereau, A. Sanford, J . Chem. SOC., Chem. Commun., 162 (1970). 22. G . Zassinovich, C. Del Bianco, G . Mestroni, J . Organomet. Chem., 222, 329 (1981).

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.1 Nitriles

Catalytic hydrogenation of nitriles is a complex process that can yield several final products deriving from a mechanism that assumes the formation of a key imine intermediate 1 as the first step'.'. Hydrogenation of imine 1 affords the usually desired primary amine 2. Addition of imine 1 and newly formed amine 2 gives a-amino amine 3, which can yield secondary amine 4 directly by hydrogenolysis, or it eliminates a molecule of ammonia to afford the Schiff base 5. This imine can be isolated when sterically hindered3, or hydrogenated to secondary amine 4. Imine 1 and secondary amine 4 can interact to form amino-amine 6,which yields tertiary amine 7 from hydrogenolysis. Finally, in proper aqueous media, aldehyde 8 may form directly by hydrolysis of the imine intermediate4.

R-CN

HZ

RCH'NH,

RCH(NH,)NHCH,R

+2

RCH=NH

RCHO 8

+ NH3

//

'L

3 l-NH3

-

U

(RCH,),NH

T 4

+ NH3

RCH=NCHzR

RCH(NH,)N(CH,R), 6

-

(RCH,),N 7

+ NH,

Fortunately, the reaction can be controlled and oriented through each of these paths, and is a very useful synthetic tool. (J.-L. GRAS)

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.1 Nitriles 14.3.7.1 -1. Hydrogenationto Primary Amines.

203

1. J. Von Braun, G . Blessing, F. Zobel, Ber. Dtsch. Chem. Ges., 856, 1888 (1923). 2. H. Greenfield, Ind. Eng. Chem., Prod. Res. Rev.,6 , 142 (1967). 3. S. Chiavarelli, G. B. Marine-Bettolo, Gazz. Chim. Ital., 86, 515 (1956). 4. K. Miyatake, M. Tsundo, Yakugaku Zasshi, 72, 630 (1952); Chem. Abstr., 47, 2177h (1953). 14.3.7.1 .l.Hydrogenationto Prlmary Amines.

Hydrogenation of a nitrile to a primary amine depends on the catalyst and whether it is an aliphatic or aromatic nitrile. Low-molecular-weight aliphatic nitriles are best converted to primary amines over finely divided Co (Raney Co) catalyst in dioxane or nickel boride', when the presence of base in the reaction mixture is undesirable. High P can limit coupling reactions, as illustrated by the 100% reduction of valeronitrile to pentylamine over Rh,O, in methanol at RT and under 104kPa H, P2. Long-chain nitriles are less sensitive to the catalyst and P as they give little coupled products. Several additives have been used to limit the extent of coupling. Strong acid solutions allow removing primary amine 2 (see 14.3.7.1) by formation of a salt that can react with imine 1 [equation (a)],. Another alternative involves using finely divided Ni (Raney Ni) and sodium acetate in acetic anhydride, a continuation that gives excellent yields of the primary amine acetate [equation (b)I4.

dCH,

OCH,

Primary amine formation is equally well promoted in alkaline medium, e.g., aqueous ethanolic NaOH solution5, that selectively poisons the catalyst for hydrogenolysis reactions. However, saturated NH,/alcohol solutions best afford almost quantitative yields of primary amines from catalytic reduction of nitriles6. Ammonia adds to imine 1 to give a 1,l-diamine, which is hydrogenolyzed to the primary amine. In the presence of NH,, finely divided Ni can be used', platinized finely divided Ni for the hydrogenation of hindered nitriles, and rhodium-on-alumina for sensitive compounds. Mild reduction of 3-indoleacetonitrile to tryptamine [equation (c)] is effected at RT over 5% rhodiumon-alumina in 10% ethanolic NH, with little side reaction', and branched chain amino sugars are conveniently prepared using this selective hydrogenation [equation (d)I9. CN

H

H,, 250 @a, Rh-A120, EtOH, NH,, 25'C

'

H


14.3.7. Hydrogenation of Other Functional Groups 14.3.7.1 Nitriles 14.3.7.1 -1. Hydrogenationto Primary Amines.

203

1. J. Von Braun, G . Blessing, F. Zobel, Ber. Dtsch. Chem. Ges., 856, 1888 (1923). 2. H. Greenfield, Ind. Eng. Chem., Prod. Res. Rev.,6 , 142 (1967). 3. S. Chiavarelli, G. B. Marine-Bettolo, Gazz. Chim. Ital., 86, 515 (1956). 4. K. Miyatake, M. Tsundo, Yakugaku Zasshi, 72, 630 (1952); Chem. Abstr., 47, 2177h (1953). 14.3.7.1 .l.Hydrogenationto Prlmary Amines.

Hydrogenation of a nitrile to a primary amine depends on the catalyst and whether it is an aliphatic or aromatic nitrile. Low-molecular-weight aliphatic nitriles are best converted to primary amines over finely divided Co (Raney Co) catalyst in dioxane or nickel boride', when the presence of base in the reaction mixture is undesirable. High P can limit coupling reactions, as illustrated by the 100% reduction of valeronitrile to pentylamine over Rh,O, in methanol at RT and under 104kPa H, P2. Long-chain nitriles are less sensitive to the catalyst and P as they give little coupled products. Several additives have been used to limit the extent of coupling. Strong acid solutions allow removing primary amine 2 (see 14.3.7.1) by formation of a salt that can react with imine 1 [equation (a)],. Another alternative involves using finely divided Ni (Raney Ni) and sodium acetate in acetic anhydride, a continuation that gives excellent yields of the primary amine acetate [equation (b)I4.

dCH,

OCH,

Primary amine formation is equally well promoted in alkaline medium, e.g., aqueous ethanolic NaOH solution5, that selectively poisons the catalyst for hydrogenolysis reactions. However, saturated NH,/alcohol solutions best afford almost quantitative yields of primary amines from catalytic reduction of nitriles6. Ammonia adds to imine 1 to give a 1,l-diamine, which is hydrogenolyzed to the primary amine. In the presence of NH,, finely divided Ni can be used', platinized finely divided Ni for the hydrogenation of hindered nitriles, and rhodium-on-alumina for sensitive compounds. Mild reduction of 3-indoleacetonitrile to tryptamine [equation (c)] is effected at RT over 5% rhodiumon-alumina in 10% ethanolic NH, with little side reaction', and branched chain amino sugars are conveniently prepared using this selective hydrogenation [equation (d)I9. CN

H

H,, 250 @a, Rh-A120, EtOH, NH,, 25'C

'

H

204

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.1 Nitrites 14.3.7.1-1. Hydrogenation to Primary Arnines.

(a) Rh-AI,O, EtOH, NH, (b) Ac,O

/

NHCOCH,

'

80%

The rhodium hydride RhH(i-Pr,), is an active homogeneous catalyst for the hydrogenation of nitriles to primary amines [equation (e)]. Activity of this catalyst is lower for aromatic than for aliphatic nitriles, and reduction of the C=C bond of unsaturated nitriles proceeds faster than that of the nitrile function".

96%

PhCH,CN PhCN

PhCHzCHzNHz

45%

PhCHzNHz

(e)

12%

CH,CH=CHCN -+ CH,(CHz),NH2 Aromatic nitriles hydrogenate to primary amine over palladium-on-carbon in ethanol-HC1 or acetic acid" finely divided Ni (Raney Ni) in ethanol/NH, [equation (f)]. Rhodium hydroxide and (7:3) rhodium-platinum oxide inhibited by LiOH convert benzonitrile to benzylamine, and isophthalonitrile or terephthalonitrile to the corresponding diamines [equation (g)] 12. finely divided Ni (W2 Raney Ni) EtOH, NH,

(f 1

Many hydrogenation of dinitrile patents have been issued. Hydrogenation of aliphatic dinitriles yields either polymers or cyclic amines if five- or six-member ring formation is possible, but the corresponding diamines are obtained in acidic media (H,S04) over a platinum oxide-palladium-on-carbon catalyst, over Raney Co in the presence of ammonia, or over Ni in ethanolic NaOH s ~ l u t i o n ' Adiponitrile ~. is hydrogenated under mild conditions over the organometallic mixed catalysts N i ( a ~ a c )or ~ Co(acac),/triethylaluminium, to hexamethylenediamine (92% yield)14. One nitrile group only can be reduced under specific conditions (usually over palladium catalysts) or by limiting the hydrogen absorption.

14.3.7. Hydrogenationof Other Functional Groups 14.3.7.1 Nitriles 14.3.7.1.2. Coupling Reactions.

205

The nitrile group can be selectively hydrogenated in the presence of aldehydes, ketones but not in the presence of olefins, acetylenes, and nitro groups. Cyanohydrins are reduced to amino alcohol mostly over platinum oxide in acetic acid”. Low-pressure hydrogenation of a-aminonitriles occurs without hydrogenolysis over platinum oxide in acetic anhydride (to the a#-diacetamido compound) or in alcoholHCl (to the a,P-diamine)I6. This procedure is not applicable to N-substituted aminonitriles that are efficiently reduced using rhodium-on-alumina in alcoholic ammonia [equation (h)]’; the catalytic system also reduces p-, y , &aminonitriles to diamines. CH3-N

n

N-CH2CN

W

5% Rh-A1203 EtOH,NH3

’

CH3-N

n

N-CH2CH2NH2

W

(h)

(J.-L. GRAS) 1. 2. 3. 4. 5. 6. 7.

8.

9. 10. 11.

T. W. Russel, R. C. Hoy, J. E. Cornelius, J . Org. Chem., 37, 3552 (1972). P. N. Rylander, L. Hasbrouk, I. Karpenko, Ann. N.Y. Acad. Sci., 214, 100 (1973). M. A. Schwartz, M. Zoda, B. Vishnuvajjala, I. Mami, J . Org. Chem., 41,2502 (1976). C. G . Overberger, 3. E. Mulvaney, J. Am. Chem. Soc., 81,4697 (1959). R. J. Bergeron, J. S . McManis, J . Org. Chem., 53, 3108 (1988). H. Greenfield, Ind. Eng. Chem., Prod. Res. Dev., 6, 142 (1967). F. Janssens, J. Torremans, M. Janssens, J . Med. Chem. 28, 1934 (1985). M. Freifelder, J. Am. Chem. SOC.,82, 2386 (1960). A. Rosenthal, D. A. Baker, J. Org. Chem., 38, 193 (1973). T. Yoshida, T. Okano, S . Otsuka, J . Chem. SOC.,Chem. Commun., 870 (1979). J. F. J. Engbersen, A. Koudijs, M. H. A. Jootsen, H. C. Van Der Plas, J . Heterocycle Chem., 23,989 (1986).

12. Y. Takagi, S. Nishimura, K. Taya, K. Hirota, Sci. Pup. Inst. Phys. Chem. Res., 61, 114 (1967). 13. R. J. Bergeron, J. R. Garlich, Synthesis, 782 (1984). 14. B. Fell, G. Gurke, Chem. Ztg., 115, 83 (1991). 15. H. J. Ringold,J. Am. Chem. SOC., 82,961 (1960). 16. M. Freifelder, R. B. Hasbrouck, J. Am. Chem. Soc., 82, 696 (1960). 14.3.7.1.2. Coupling Reactlons.

Catalytic hydrogenation of nitriles in neutral media yields large amounts of secondary or tertiary amines. These arise from addition of the primary amine to the imine intermediate, and of the newly formed secondary amine to another imine intermediate, depending mainly on the catalyst. Hydrogenation of aliphatic nitriles can be oriented to secondary amine by running the reaction over Rh, whereas tertiary amine is predominant over Pd or Pt. For example, rhodium-on-carbon affords 100% dipentylamine and dibenzylamine, respectively, from valeronitrile and benzonitrile’. A similar difference is found between rhodium boride, which favors secondary amine and platinum boride, which favors tertiary amine2. However, commercially benzonitrile is reduced to dibenzylamine over platinum-on-carbon in the presence of H203. Generally, coupled products are favored at low T and P and are harder to obtain with long chain (>c6) aliphatic nitriles. Since secondary and tertiary amines are obtained by reaction of a primary and secondary amine with the imine intermediate, selected unsymmetric secondary and tertiary amines can be prepared by substituting an added chosen amine for the reacting amines. The product composition of this reductive condensation over an appropriate catalyst depends on the nitri1e:amine ratio, and to a lesser extent on solvent. Platinum, Pd, and rhodium-on-carbon, in alcohol or hydrocarbon solvent with about 100% excess of added amine, give good yields of n-butyl-n-pentylamine from hydrogenation of valeronitrile in presence of n-b~tylamine~:


14.3.7. Hydrogenationof Other Functional Groups 14.3.7.1 Nitriles 14.3.7.1.2. Coupling Reactions.

205

The nitrile group can be selectively hydrogenated in the presence of aldehydes, ketones but not in the presence of olefins, acetylenes, and nitro groups. Cyanohydrins are reduced to amino alcohol mostly over platinum oxide in acetic acid”. Low-pressure hydrogenation of a-aminonitriles occurs without hydrogenolysis over platinum oxide in acetic anhydride (to the a#-diacetamido compound) or in alcoholHCl (to the a,P-diamine)I6. This procedure is not applicable to N-substituted aminonitriles that are efficiently reduced using rhodium-on-alumina in alcoholic ammonia [equation (h)]’; the catalytic system also reduces p-, y , &aminonitriles to diamines. CH3-N

n

N-CH2CN

W

5% Rh-A1203 EtOH,NH3

’

CH3-N

n

N-CH2CH2NH2

W

(h)

(J.-L. GRAS) 1. 2. 3. 4. 5. 6. 7.

8.

9. 10. 11.

T. W. Russel, R. C. Hoy, J. E. Cornelius, J . Org. Chem., 37, 3552 (1972). P. N. Rylander, L. Hasbrouk, I. Karpenko, Ann. N.Y. Acad. Sci., 214, 100 (1973). M. A. Schwartz, M. Zoda, B. Vishnuvajjala, I. Mami, J . Org. Chem., 41,2502 (1976). C. G . Overberger, 3. E. Mulvaney, J. Am. Chem. Soc., 81,4697 (1959). R. J. Bergeron, J. S . McManis, J . Org. Chem., 53, 3108 (1988). H. Greenfield, Ind. Eng. Chem., Prod. Res. Dev., 6, 142 (1967). F. Janssens, J. Torremans, M. Janssens, J . Med. Chem. 28, 1934 (1985). M. Freifelder, J. Am. Chem. SOC.,82, 2386 (1960). A. Rosenthal, D. A. Baker, J. Org. Chem., 38, 193 (1973). T. Yoshida, T. Okano, S . Otsuka, J . Chem. SOC.,Chem. Commun., 870 (1979). J. F. J. Engbersen, A. Koudijs, M. H. A. Jootsen, H. C. Van Der Plas, J . Heterocycle Chem., 23,989 (1986).

12. Y. Takagi, S. Nishimura, K. Taya, K. Hirota, Sci. Pup. Inst. Phys. Chem. Res., 61, 114 (1967). 13. R. J. Bergeron, J. R. Garlich, Synthesis, 782 (1984). 14. B. Fell, G. Gurke, Chem. Ztg., 115, 83 (1991). 15. H. J. Ringold,J. Am. Chem. SOC., 82,961 (1960). 16. M. Freifelder, R. B. Hasbrouck, J. Am. Chem. Soc., 82, 696 (1960). 14.3.7.1.2. Coupling Reactlons.

Catalytic hydrogenation of nitriles in neutral media yields large amounts of secondary or tertiary amines. These arise from addition of the primary amine to the imine intermediate, and of the newly formed secondary amine to another imine intermediate, depending mainly on the catalyst. Hydrogenation of aliphatic nitriles can be oriented to secondary amine by running the reaction over Rh, whereas tertiary amine is predominant over Pd or Pt. For example, rhodium-on-carbon affords 100% dipentylamine and dibenzylamine, respectively, from valeronitrile and benzonitrile’. A similar difference is found between rhodium boride, which favors secondary amine and platinum boride, which favors tertiary amine2. However, commercially benzonitrile is reduced to dibenzylamine over platinum-on-carbon in the presence of H203. Generally, coupled products are favored at low T and P and are harder to obtain with long chain (>c6) aliphatic nitriles. Since secondary and tertiary amines are obtained by reaction of a primary and secondary amine with the imine intermediate, selected unsymmetric secondary and tertiary amines can be prepared by substituting an added chosen amine for the reacting amines. The product composition of this reductive condensation over an appropriate catalyst depends on the nitri1e:amine ratio, and to a lesser extent on solvent. Platinum, Pd, and rhodium-on-carbon, in alcohol or hydrocarbon solvent with about 100% excess of added amine, give good yields of n-butyl-n-pentylamine from hydrogenation of valeronitrile in presence of n-b~tylamine~:

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.1 Nitriles 14.3.7.1.3. Reductive Hydrolysis.

206

C,H,CN

+ C4H,NHz -% C4H,NHC,Hl1

(a)

>go%

This reaction applied to a mixture of a benzyl cyanide and various amines affords a convenient route to N-substitute phenylethylamines [equation (b)I5. Pharmacologically active amines are obtained in excellent yield, as in the synthesis of epinin dimethyl ether [equation (c)]~: 02N-(C6H4)-cH2CN

+ (CH3)zNH

Pd-BaSO,

EtOH

’

Me0

OMe (J.-L. GRAS)

P. N. Rylander, L. Hasbrouck, I. Karpenko, Ann. N.Y. Acad. Sci., 214 (1979). C. Bamett, Ind. Eng. Chem.,Prod. Res. Dev., 8, 145 (1969). H. Greenfield, Ind. Eng. Chem., Prod. Res. Dev., 15, 5 (1976). P. N. Rylander, in Catalytic in Organic Syntheses, Academic Press, New York, - Hydrogenation ~1979, p 144. 5. K. Kindler. K. Shrader. B. Middelhoff. Arch. Pharm.. 283. 184 (1950). 6. S. Carra, V.Ragaini, Tetrahedron Lett., 1079 (1967): 1. 2. 3. 4.

14.3.7.1 3.Reductive Hydrolysis.

Formation of an aldimine intermediate is the first step in the catalytic hydrogenation of nitriles. When hydrogenation is carried out in aqueous acidic media, aldehydes may form by hydrolysis of the aldimine. Selective hydrogenation of aromatic nitriles to benzaldehydes over finely divided Ni (Raney Ni) is best obtained with an equimolecular amount of HzS04 in a 1O:l mixture of tetrahydrofuran-water1: p-RO--(C&)-CN

R = H, CH,

finely divided Ni, H,SO,

THF-H20, 25°C

’p-RO-(C,H,)-CHO

(a)

86%

If hydrogenation is conducted with excess of semicarbazide hydrochloride, after absorption of one equivalent of hydrogen the intermediate imine reacts with the nuclophilic species and yields the aldehyde semicarbazone [equation (b)]’. Under selected reaction conditions, the semicarbazone is resistant to hydrogenation. In the presence of amines, reaction would lead to ketimines3. R-CN

+ HzNCONHz.HC1

finely divided Ni CH,OH,H,O, 25”c)R--CH=N-NHCONH,

- 70%

(b)

Nitriles are easily reduced by Ni-A1 alloy, to the corresponding primary amine in aqueous alkali, and to the aldehyde in aqueous formic acid solution4. Reductive hydrolysis may lead to the carbinol after further hydrogenation of the aldehydic group, as


14.3.7. Hydrogenation of Other Functional Groups 14.3.7.1 Nitriles 14.3.7.1.3. Reductive Hydrolysis.

206

C,H,CN

+ C4H,NHz -% C4H,NHC,Hl1

(a)

>go%

This reaction applied to a mixture of a benzyl cyanide and various amines affords a convenient route to N-substitute phenylethylamines [equation (b)I5. Pharmacologically active amines are obtained in excellent yield, as in the synthesis of epinin dimethyl ether [equation (c)]~: 02N-(C6H4)-cH2CN

+ (CH3)zNH

Pd-BaSO,

EtOH

’

Me0

OMe (J.-L. GRAS)

P. N. Rylander, L. Hasbrouck, I. Karpenko, Ann. N.Y. Acad. Sci., 214 (1979). C. Bamett, Ind. Eng. Chem.,Prod. Res. Dev., 8, 145 (1969). H. Greenfield, Ind. Eng. Chem., Prod. Res. Dev., 15, 5 (1976). P. N. Rylander, in Catalytic in Organic Syntheses, Academic Press, New York, - Hydrogenation ~1979, p 144. 5. K. Kindler. K. Shrader. B. Middelhoff. Arch. Pharm.. 283. 184 (1950). 6. S. Carra, V.Ragaini, Tetrahedron Lett., 1079 (1967): 1. 2. 3. 4.

14.3.7.1 3.Reductive Hydrolysis.

Formation of an aldimine intermediate is the first step in the catalytic hydrogenation of nitriles. When hydrogenation is carried out in aqueous acidic media, aldehydes may form by hydrolysis of the aldimine. Selective hydrogenation of aromatic nitriles to benzaldehydes over finely divided Ni (Raney Ni) is best obtained with an equimolecular amount of HzS04 in a 1O:l mixture of tetrahydrofuran-water1: p-RO--(C&)-CN

R = H, CH,

finely divided Ni, H,SO,

THF-H20, 25°C

’p-RO-(C,H,)-CHO

(a)

86%

If hydrogenation is conducted with excess of semicarbazide hydrochloride, after absorption of one equivalent of hydrogen the intermediate imine reacts with the nuclophilic species and yields the aldehyde semicarbazone [equation (b)]’. Under selected reaction conditions, the semicarbazone is resistant to hydrogenation. In the presence of amines, reaction would lead to ketimines3. R-CN

+ HzNCONHz.HC1

finely divided Ni CH,OH,H,O, 25”c)R--CH=N-NHCONH,

- 70%

(b)

Nitriles are easily reduced by Ni-A1 alloy, to the corresponding primary amine in aqueous alkali, and to the aldehyde in aqueous formic acid solution4. Reductive hydrolysis may lead to the carbinol after further hydrogenation of the aldehydic group, as

14.3.7. Hydrogenationof Other Functional Groups 14.3.7.1 Nitrites 14.3.7.1.4. Hydrogenolysisand Cyclizations.

207

illustrated in equation ( c ) ~for an aromatic nitrile. The carbinol is obtained from aliphatic nitriles by reductive hydrolysis over finely divided Co (Raney Co) (MeOH, 120°C)6.

(J.-L. GRAS)

1. P. Tinapp, Chem. Ber., 102, 2770 (1969). 2. H. Plieninger, G . Werst, Chem. Ber., 88, 1956 (1955). 3. H. R. Snyder, E. P. Merica, C. G . Force, E. G. White, J . Am. Chem. SOC.,80,4622 (1958). 4. L. K. Keefer, G. Lunn, Chem. Rev., 89,459 (1989). 5 . R. H. Mizzoni, R. A. Lucas, R. Smith, J. Boxer, J. E. Brown, F. Goble, E. Konopka, J. Gelzer, J. Szanto, D. C. Maplesden, G . De Stevens, J . Med. Chem., 13,878 (1970). 6. W. A. W. Cummings, A. C. Davis, J . Chem. Soc., 4591 (1964).

14.3.7.1.4. Hydrogenolyslsand Cyclizations.

Nitriles are reduced completely to methyl groups by an active donor such as pmenthene under conditions normally used for catalytic hydrogen transfer. Reduction proceeds satisfactorily when the cyano group is bound to an aromatic of heterocyclic ring [equation (a)I1, but is slightly more difficult in the case of aliphatic nitriles [equation @)I2.

PhCH=C(COOEt)CN

Pd-C

PhCH,CH(COOEt)CH3

(b)

81%

This hydrogenation-hydrogenolysis reaction is convenient for methylation of an aromatic nucleus, a method useful in the synthesis of thyroxine analogues3:


14.3.7. Hydrogenationof Other Functional Groups 14.3.7.1 Nitrites 14.3.7.1.4. Hydrogenolysisand Cyclizations.

207

illustrated in equation ( c ) ~for an aromatic nitrile. The carbinol is obtained from aliphatic nitriles by reductive hydrolysis over finely divided Co (Raney Co) (MeOH, 120°C)6.

(J.-L. GRAS)

1. P. Tinapp, Chem. Ber., 102, 2770 (1969). 2. H. Plieninger, G . Werst, Chem. Ber., 88, 1956 (1955). 3. H. R. Snyder, E. P. Merica, C. G . Force, E. G. White, J . Am. Chem. SOC.,80,4622 (1958). 4. L. K. Keefer, G. Lunn, Chem. Rev., 89,459 (1989). 5 . R. H. Mizzoni, R. A. Lucas, R. Smith, J. Boxer, J. E. Brown, F. Goble, E. Konopka, J. Gelzer, J. Szanto, D. C. Maplesden, G . De Stevens, J . Med. Chem., 13,878 (1970). 6. W. A. W. Cummings, A. C. Davis, J . Chem. Soc., 4591 (1964).

14.3.7.1.4. Hydrogenolyslsand Cyclizations.

Nitriles are reduced completely to methyl groups by an active donor such as pmenthene under conditions normally used for catalytic hydrogen transfer. Reduction proceeds satisfactorily when the cyano group is bound to an aromatic of heterocyclic ring [equation (a)I1, but is slightly more difficult in the case of aliphatic nitriles [equation @)I2.

PhCH=C(COOEt)CN

Pd-C

PhCH,CH(COOEt)CH3

(b)

81%

This hydrogenation-hydrogenolysis reaction is convenient for methylation of an aromatic nucleus, a method useful in the synthesis of thyroxine analogues3:

14.3.7. H drogenation of Other Functional Groups 14.3.7.1 {itriles 14.3.7.1-4. Hydrogenolysis and Cyclizations.

208

-

NHAc -o*o+cH2~H

Pd-C

I

COOEt

CN

NHAc

I

CH2CH I

MeO-@-O*

(c)

COOEt

Me

Other synthetic applications derive from various cyclizations that occur during reduction of the substrates that contain a reactive function at a specific position. These cyclizations proceed through the intermediate imine or amine to the cyclic amine. Various functional units such as olefins, heterocyclic rings, ketones, acids, amides, amines, nitrile, and nitro have been involved. Palladium, Rh, and finely divided Ni (Raney Ni) under more vigorous conditions are commonly used. Reductive cyclization of ketonitrile 1 over Raney Ni gives myosmine 2 along with some nornicotine 3 as overreduced product4:

CN

H, finelydividedNi , 300 P a , EtOH

~

fl fl +

2

1

(d)

3

Quinolizidine derivatives formed by hydrogenation of the dicyano ketone 4 to d,l-matrine 5 over 5% palladium-on-carbon [equation (e)I5. Increased catalyst loading affects the stereochemistry of cyclization, probably by isomerization.

4

5

Hydrogenation of butyronitrile derivative 6 over platinum oxide in acidic medium yields quinolizidine 7, by reduction of the pyridine ring first and piperidone derivative 8 by reduction of the cyan0 group6:

FN 6

COOEt

0

7

8

I

14.3.7. Hydrogenationof Other Functional Groups 14.3.7.2. Nitro Compounds 14.3.7.2.1. Hydrogenationto the Amine.

209

Reductive cyclization of nitrocyano derivatives provides an elegant entry to a variety of indole-3-carboxamides by formation of a five member ring amine’:

&oy’ CN

Pd-C

,

mcoNR ( g)

H (J.-L. GWS)

1. 2. 3. 4. 5. 6. 7.

K. Kindler, K. Luhrs, Chem. Ber., 99, 227 (1966). K. Kindler, K. Luhrs, Justus Liebigs Ann. Chem., 707, 26 (1967). P. Block, Jr., D. H. Coy, J . Chem. SOC., Perkins. Trans. I , 633 (1972). E. Lette, M. R. Chedekel, G. B. Bodem,J. Org. Chem., 37,4465 (1972). L. Mandell, K. P. Singh, J. T. Gresham, W. J. Freeman, J . Am. Chem. Soc., 87,5234 (1965). W. Boekelheide, W. J. Linn, P. O’Grady, M. Lamborg, J . Am. Chem. Soc., 75, 3243 (1953). J. Bordais, C. Germain, Tetrahedron Lerr., 195 (1970).

14.3.7.2. Nitro Compounds 14.3.7.2.1. Hydrogenation to the Amlne.

Catalytic hydrogenation of an aromatic nitro group to aniline derivatives occurs easily in preference to all other functional groups except olefins or acetylenes. Aliphatic nitro compounds are reduced much more slowly to the corresponding amine, and offer the possibility of partial reduction. Palladium, Pt, and Ni, in supported or unsupported forms, are widely used for hydrogenation of the nitro group. Rhodium or Ru can be used for special purposes. Commercially, platinum metal sulfides are popular because they are insensitive to sulfur compound poisoning. Hydrogen transfer reduction of aromatic nitro compounds has been extensively investigated, and appears to provide some technical improvement as well as higher selectivity’. Nitro group hydrogenation has been carried out in many solvents, and with no solvent at all. Most common alcohols or polyols have been used, beside dilute aqueous HCI, H2S0,, and many common organic solvents, so long as the solvent does not interact with other substrate functions. Reductions can be run in neutral or acidic media (essential in some cases although acetic acid can act as an inhibitor). Since hydrogenation of a nitro group is exothermic, it can be run under ambient T conditions. Exceptions are industrial processes or hydrogenation of aliphatic nitro compounds. An aromatic nitro function hydrogenates to the corresponding amine in high yield over palladium-on-carbon in a process that is slightly superior to that using platinum. Homogeneous reduction of aromatic compounds (1) by a triphenylphosphine complex of palladium yields anilines 2 along with minor quantities of azobenzene and azoxybenzene derivatives2.

1

R = H,C1

2

75-95%


14.3.7. Hydrogenationof Other Functional Groups 14.3.7.2. Nitro Compounds 14.3.7.2.1. Hydrogenationto the Amine.

209

Reductive cyclization of nitrocyano derivatives provides an elegant entry to a variety of indole-3-carboxamides by formation of a five member ring amine’:

&oy’ CN

Pd-C

,

mcoNR ( g)

H (J.-L. GWS)

1. 2. 3. 4. 5. 6. 7.

K. Kindler, K. Luhrs, Chem. Ber., 99, 227 (1966). K. Kindler, K. Luhrs, Justus Liebigs Ann. Chem., 707, 26 (1967). P. Block, Jr., D. H. Coy, J . Chem. SOC., Perkins. Trans. I , 633 (1972). E. Lette, M. R. Chedekel, G. B. Bodem,J. Org. Chem., 37,4465 (1972). L. Mandell, K. P. Singh, J. T. Gresham, W. J. Freeman, J . Am. Chem. Soc., 87,5234 (1965). W. Boekelheide, W. J. Linn, P. O’Grady, M. Lamborg, J . Am. Chem. Soc., 75, 3243 (1953). J. Bordais, C. Germain, Tetrahedron Lerr., 195 (1970).

14.3.7.2. Nitro Compounds 14.3.7.2.1. Hydrogenation to the Amlne.

Catalytic hydrogenation of an aromatic nitro group to aniline derivatives occurs easily in preference to all other functional groups except olefins or acetylenes. Aliphatic nitro compounds are reduced much more slowly to the corresponding amine, and offer the possibility of partial reduction. Palladium, Pt, and Ni, in supported or unsupported forms, are widely used for hydrogenation of the nitro group. Rhodium or Ru can be used for special purposes. Commercially, platinum metal sulfides are popular because they are insensitive to sulfur compound poisoning. Hydrogen transfer reduction of aromatic nitro compounds has been extensively investigated, and appears to provide some technical improvement as well as higher selectivity’. Nitro group hydrogenation has been carried out in many solvents, and with no solvent at all. Most common alcohols or polyols have been used, beside dilute aqueous HCI, H2S0,, and many common organic solvents, so long as the solvent does not interact with other substrate functions. Reductions can be run in neutral or acidic media (essential in some cases although acetic acid can act as an inhibitor). Since hydrogenation of a nitro group is exothermic, it can be run under ambient T conditions. Exceptions are industrial processes or hydrogenation of aliphatic nitro compounds. An aromatic nitro function hydrogenates to the corresponding amine in high yield over palladium-on-carbon in a process that is slightly superior to that using platinum. Homogeneous reduction of aromatic compounds (1) by a triphenylphosphine complex of palladium yields anilines 2 along with minor quantities of azobenzene and azoxybenzene derivatives2.

1

R = H,C1

2

75-95%

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.2. Nitro Compounds 14.3.7.2.1. Hydrogenation to the Amine.

210

Petroleum ether is the most efficient solvent for hydrogenation of nitrobenzene over the organometallic reagent bis(acety1acetonate) palladium(I1) used in conjunction with pyridine3. Aromatic nitro compounds are not reduced by only NaBH,, however with palladium-on-carbon4 or copper(I1) a~etylacetonate~, reduction to the corresponding amine occurs. In general, Pd(I1) has a limited range of hydrogenation or activity; however, NaBH, reduction leads to a black material that catalyzes rapid hydrogenation of bonds of the type C=C, N=N, and N=06. Palladium chloride alone, supported on derivatized polystyrene, hydrogenates nitrobenzene to aniline (97%) and to ben~onitrile~. Aromatic nitro compounds undergo reduction to anilines when refluxed in excess cyclohexene in the presence of ordinary commercial 10% palladium-on-carbon catalyst. The reaction is slower but usually successful with sulfur-containing substrates'. Hydrogen transfer hydrogenations can involve also Ru or Rh trichlorides and various cyclic amines as hydrogen donors'. Rhodium trichloride (N-formylpiperidine) catalyzes hydrogenation of nitrobenzene to aniline in high yield", a reduction effected quantitatively over RuCl,(PPh,), l l . Aliphatic nitro groups hydrogenate less easily. Complete reductions are achieved by higher catalyst loadings, slightly more vigorous conditions, and longer reaction times, thus providing convenient routes to complex amines. For example, hydrogenation of 3 over prereduced 10% palladium-on-carbon in methanol-acetic acid under ambient conditions produces, after acetylation of the new amino group, the amino sugar 4, in excellent yields":

(a) 10%Pd-C, H,

CH,OH-AcOH (b)

(b) AcNH

3

4

Amino alcohols can be synthesized by hydrogenation of the corresponding nitro alcohols over finely divided Ni (Raney Ni)15 or Pd catalysts without hydrogenolysis of the hydroxyl function, even at a benzilic po~ition'~: CH3 I Ph-CH(0H)-C-NO,

Pd-C, H,, 400 kPa

I

EtOH, AcOH, 50"

CH3

' Ph-CH(0H)-C-NH,I

I

Nitroparaffins with (Ph,P),RuCl, hydrogenate to secondary alkyl primary amines. Aqueous work-up procedures are not necessary; the method is recommended for high catalyst turnovers and relatively mild reaction condition^'^. RCH(R' )-NO,

H,, 9

X

lo3 kPa, KOH, EtOH > RuCl,(PPh,),

RCH(R')-NH, up to 88%

(d) (J.-L. GRAS)

1. For a review see G. Brieger, T. J. Nestrick, Chem. Rev., 74,567 (1974). 2. T. K. Banerjee, D. Sen, J . Chem. Technol. Biotechnol., 31, 676 (1981).

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.2. Nitro Compounds 14.3.7.2.2. Selective and Partial Reductions.

211

M. C. Datta, C. R. Saha, D. Sen, Chem. Ind., 1057 (1975). T. Neilson, H. C. S. Wood, A. G . Wylie, J . Chem. SOC.,371 (1962). K. Hanaya, T. Muramats, H. Kudo, Y. L. Chow, J . Chem. SOC., Perkin I , 2409 (1979). T. W. Russell, D. M. Duncan, J . Org. Chem., 39, 3050 (1974). N. L. Holy, J . Chem. SOC.,Chem. Commun., 1074 (1978). I. D. Entwistle, R. A. W. Johnstone, T. J. Povall, J . Chem. SOC.,Perkin I , 1300 (1975). H. Imai, T. Nishiguchi, K. Fukuzumi, J . Org. Chem., 42,431 (1977). I. Jardine, F. J. McQuillin, Tetrahedron Lett., 173 (1972). J. F. Knifton, Tetrahedron Lett., 2163 (1975). T. Takamoto, Y. Yokota, R. Sudoh, T. Nakagawa, Bull. Chem. SOC. Jpn., 46, 1532 (1973). 13. F. H. Marquardt, S. Edwards, J . Org. Chem., 37, 1861 (1972). 14. J. F. Knifton, J . Org. Chem., 40, 519 (1975). 15. B. Imperiali, R. H. Abeles, Tetrahedron Lett., 27, 135 (1986).

3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

14.3.7.2.2. Selective and Partial Reductions.

Selective hydrogenation of one or two nitro groups in an aromatic dinitro compound is the basis for the synthesis of otherwise unattainable molecules. Carefully chosen catalysts, under totally different reaction conditions, have been met with success. Partial reduction of 2,6-dinitroanilines to nitrophenylenediamines occurs in 60-90% yield over 10% palladium-on-carbon at RT [equation (a)]', although most heterogeneous catalysts do not afford such selectivity.

X @iH2

H,, RT,300 EtOHRICCl, P a , Pd-C

+

X6

i

H

2

(a)

n02

NO2 X = COOEt, COOH, cF3, CH3

60-90%

Hydrogen transfer from cyclohexene catalyzed by ordinary commercial 10% Pd-C catalyst is very useful for selective fast reduction of polynitrobenzene2: C6HI2,Pd-C

M e O e O M e O2N

NO2

ref. EtOH, 10 min

*

Meo*oMe O2N

NH2

(b)

In compounds where the two nitro groups are not equivalent, selectivity depends on electronic effects [equation (c)] and steric hindrance3.

X = CF,, C1

Among various heterogeneous catalysts, finely divided Cu (Raney Cu) catalyzes selective hydrogenation of 2,4-dinoalkylbenzenes to the corresponding 4-amino-2-nitroalkylbenzenes [equation (d)I4. The selectivity attributed to steric hindrance is reversed by using



211

M. C. Datta, C. R. Saha, D. Sen, Chem. Ind., 1057 (1975). T. Neilson, H. C. S. Wood, A. G . Wylie, J . Chem. SOC.,371 (1962). K. Hanaya, T. Muramats, H. Kudo, Y. L. Chow, J . Chem. SOC., Perkin I , 2409 (1979). T. W. Russell, D. M. Duncan, J . Org. Chem., 39, 3050 (1974). N. L. Holy, J . Chem. SOC.,Chem. Commun., 1074 (1978). I. D. Entwistle, R. A. W. Johnstone, T. J. Povall, J . Chem. SOC.,Perkin I , 1300 (1975). H. Imai, T. Nishiguchi, K. Fukuzumi, J . Org. Chem., 42,431 (1977). I. Jardine, F. J. McQuillin, Tetrahedron Lett., 173 (1972). J. F. Knifton, Tetrahedron Lett., 2163 (1975). T. Takamoto, Y. Yokota, R. Sudoh, T. Nakagawa, Bull. Chem. SOC. Jpn., 46, 1532 (1973). 13. F. H. Marquardt, S. Edwards, J . Org. Chem., 37, 1861 (1972). 14. J. F. Knifton, J . Org. Chem., 40, 519 (1975). 15. B. Imperiali, R. H. Abeles, Tetrahedron Lett., 27, 135 (1986).

3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

14.3.7.2.2. Selective and Partial Reductions.

Selective hydrogenation of one or two nitro groups in an aromatic dinitro compound is the basis for the synthesis of otherwise unattainable molecules. Carefully chosen catalysts, under totally different reaction conditions, have been met with success. Partial reduction of 2,6-dinitroanilines to nitrophenylenediamines occurs in 60-90% yield over 10% palladium-on-carbon at RT [equation (a)]', although most heterogeneous catalysts do not afford such selectivity.

X @iH2

H,, RT,300 EtOHRICCl, P a , Pd-C

+

X6

i

H

2

(a)

n02

NO2 X = COOEt, COOH, cF3, CH3

60-90%

Hydrogen transfer from cyclohexene catalyzed by ordinary commercial 10% Pd-C catalyst is very useful for selective fast reduction of polynitrobenzene2: C6HI2,Pd-C

M e O e O M e O2N

NO2

ref. EtOH, 10 min

*

Meo*oMe O2N

NH2

(b)

In compounds where the two nitro groups are not equivalent, selectivity depends on electronic effects [equation (c)] and steric hindrance3.

X = CF,, C1

Among various heterogeneous catalysts, finely divided Cu (Raney Cu) catalyzes selective hydrogenation of 2,4-dinoalkylbenzenes to the corresponding 4-amino-2-nitroalkylbenzenes [equation (d)I4. The selectivity attributed to steric hindrance is reversed by using


21 2

an Ru complex such as (Ph,P),RuCl, [equation (e)I5. Generally, homogeneous hydrogenation offers extremely high selectivity in the reduction of dinitroarenes.

h finely divided Cu, H,, 14OO kPa 150OC

P'

,

NO,

NH2

Me

Me RuCI,(PPh,),, H,, 8 135°C

X

lo3kPa

NO2

NO2

Since the nitro group is one of the easiest functional groups to reduce, especially in aromatics, selective hydrogenation is possible in presence of several other functional groups within the same molecule. Aromatic nitro-nitrile compounds are easily reduced to the corresponding aminonitrile derivative, and nitro oximes to the amino oximes6:

R

R

up to 90% Selective hydrogenation of the nitro function in aromatic nitro ketones, esters, and amides takes place over various catalysts such as Pd black in H,SO,, palladiumon-carbon, platinum oxide, and even ruthenium-on-carbon; the nitro function is preferentially adsorbed. For preparation of an intermediate required in the synthesis of 5,5-dimethylpyrrolidone,1 is reduced selectively over finely divided Ni (W5 Raney Ni) in absolute EtOH': (CH3),C-(CH,),COOCH3

I

rso,

Raney Ni, H,, 7

X

EtOH, 55°C

lo3 kPa

,

1

Reduction of aromatic nitro groups using triethylammonium formate, catalyzed by palladium-on-carbon or a soluble triaryphosphine-palladium acetate catalyst is compatible with ester or amide functions*, leading to a-amino esters'. The nitro group of the pyranone 2 is selectively reduced, without hydrogenolysis of the benzyl ether linkage in the presence of finely divided Ni":

14.3.7. Hydrogenationof Other Functional Groups 14.3.7.2. Nitro Compounds 14.3.7.2.2. Selective and Partial Reductions.

NO2

H,, finely divided Ni

'

""h NH2 (h)

OH

OH 2

213

83%

Hydrogenation of halonitro aromatics to the corresponding haloanilines is achieved on an industrial scale with limited dehalogenation. Sulfides of all noble metals have been used, under vigorous conditions' ',12. Platinum, along with a suitable inhibitor, combines minimized dehalogenation and fast rate of reduction of the nitro group13. Inhibitors to minimize loss of halogens include mostly bases like hydroxides and amines13,or compounds of the group X-PH(0)-OH (with X = H, OH, alkyl, or phenyl)14. Selective hydrogenation of the nitro group preferentially to an aliphatic C-C unsaturation is a greater challenge. In the presence of an acetylenic function, selective reduction may be achieved using Ru supported on carbon or alumina. With monosubstituted acetylenes, catalyst turn over is poor, because of the strong absorption of the terminal acetylene onto the catalyst. However, use of a dimethylcarbinol as a blocking group increases steric crowding around the triple bond, and maintains a high rate of hydrogenation of the nitro group [equation (i)]15. The blocking group is removed by a catalytic amount of NaOH.

Reduction of the double bond in an olefinic aliphatic nitro derivatives is usually easy. The olefin linkage may be preserved only in aromatic nitro compounds [equation (i)]16, or in particular structures such as hydroxycoumarins or pyranones'O.

During catalytic hydrogenation of nitro derivatives, partial reduction of the nitro group yields azo, hydrazo, hydroxylamine, nitroso, or oxime functions, in side products, as a result of catalyst deactivation. Partial reduction may become synthetically useful under carefully chosen conditions, and in substrates where the new formed function cannot interact within the molecule. Nitrobenzene hydrogenation over platinum-on-carbon in a mixture alcohol-dimethyl sulfoxide (as promoter) produces high yields of phenylhydr~xylamines'~:

Aliphatic hydroxylamines form efficiently through many patented procedures involving partial reduction of the corresponding nitro compounds over Pd-black and Pd

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.2. Nitro Compounds 14.3.7.2.3. Side Reactions in Polyfunctional Molecules.

214

supported on carbon, alumina, CaCO,, or a partially deactivated Pt metal catalyst. Palladium or platinum catalysts deactivated by heavy metals are patented procedures for production of oximes through catalytic partial hydrogenation of nitro derivatives. Excellent oxime yields arise by hydrogenation of vinyl or styryl nitro compounds over palladium-on-carbon deactivated in pyridine18, over rhodium-on-alumina in ethanol-acetic acid-ethyl acetate [equation (l)]I9, or by reduction in acidic media over Pd catalystsz0. MeO\

Me0\

a = - N o z \NOz

EtOH/AcOH/EtOA: 5% Rh-A1203

M

e

o

w \

N

o

H (1)

n02 (J.-L. GRAS)

1. R. E. Lyle, J. L. LaMattina, Synthesis, 726 (1974). 2. I. D. Entwistle, R. A. W. Johnstone, T. J. Povall, J . Chem. Soc., Perkin I , 1300 (1975). 3. J. L. Miesel, G. 0. P. O'Doherty, J. M. Owen, in Catalysis in Organic Syntheses, P. N. Rylander, H. Greenfield, eds., Academic Press, New York, 1976, p. 273. 4. W. H. Jones, W. F. Benning, P. Davies, D. M. Mulvey, P. I. Pollack, J. C. Schaeffer, R. Tull, L. M. Weinstock, Ann. N.Y.Acad. Sci., 158,471 (1969). 5. J. F. Knifton, Tetrahedron Lett., 2163 (1975). 6. W. L. F. Armareg0,J. Chem. Soc., 5030 (1962). 7. R. B. Moffett, Org. Synth., Cull. vol. 4 , 357 (1963). 8. N. A. Cortese, R. F. Heck, J . Org. Chem., 42, 3491 (1977). 9. S. Ram, R. E. Ehrenkaufer, Synth. Commun., 16, 133 (1986). 10. H. W. R. Williams, Can. J . Chern., 54, 3377 (1976). 1 1 . H. S. Broadbent, Ann. N.Y.Acad. Sci., 145,58 (1967). 12. H. Greenfield, Ann. N.Y.Acad. Sci., 145,108 (1967). 13. J. R. Kosak, Ann. N.Y.Acad. Sci., 172, 175 (1970). 14. J. R. Kosak, U.S. Patent 4,020,107 (1977); Chem. Abstr., 86, 89364t (1977). 15. A. Onopchenko, E. T. Sabourin, C. M. Selwitz, J . Org. Chem., 44, 1233 (1979). 16. B. R. Baker, J. H. Jordaan, J. Med. Chem., 8, 35 (1965). 17. P. N. Rylander, I. M. Karpenko, G. R. Pond, U.S. Patent 3,694,509 (1977); Chem. Abstr., 76, 14082j (1972). 18. W. K.Seifert, P. C. Condit, J. Org. Chem., 28, 265 (1963). 19. L. K. Friedlin, E. F. Litvin, V. M. Chursina, Katal. Vosstanov.GidrirovZhidk. Faze, 59 (1970); Chem. Abstr., 76,45665f (1972). 20. I. Baxter, G. A. Swan, J . Chem. SOC. C , 2230 (1967). 14.3.7.2.3. Side Reactions In Polyfunctional Molecules.

If other active groups are present y or S to the reducible nitro group, hydrogenation can result in formation of nitrogen heterocyclic products. Several such cyclizations provide an entry to indoles, such as reductive cyclizations of dinitrostyrene 1 [equation (a)]', of o-nitrobenzyl ketone 2 [equation (b)]', and of nitro nitrile 3 [equation (c)],, all of them carried out on palladium-on-carbon. 10% Pd-C,45"C EtOH/EtOAc/AcOH

Me0

I

MeO R

R = H,I


14.3.7. Hydrogenation of Other Functional Groups 14.3.7.2. Nitro Compounds 14.3.7.2.3. Side Reactions in Polyfunctional Molecules.

214

supported on carbon, alumina, CaCO,, or a partially deactivated Pt metal catalyst. Palladium or platinum catalysts deactivated by heavy metals are patented procedures for production of oximes through catalytic partial hydrogenation of nitro derivatives. Excellent oxime yields arise by hydrogenation of vinyl or styryl nitro compounds over palladium-on-carbon deactivated in pyridine18, over rhodium-on-alumina in ethanol-acetic acid-ethyl acetate [equation (l)]I9, or by reduction in acidic media over Pd catalystsz0. MeO\

Me0\

a = - N o z \NOz

EtOH/AcOH/EtOA: 5% Rh-A1203

M

e

o

w \

N

o

H (1)

n02 (J.-L. GRAS)

1. R. E. Lyle, J. L. LaMattina, Synthesis, 726 (1974). 2. I. D. Entwistle, R. A. W. Johnstone, T. J. Povall, J . Chem. Soc., Perkin I , 1300 (1975). 3. J. L. Miesel, G. 0. P. O'Doherty, J. M. Owen, in Catalysis in Organic Syntheses, P. N. Rylander, H. Greenfield, eds., Academic Press, New York, 1976, p. 273. 4. W. H. Jones, W. F. Benning, P. Davies, D. M. Mulvey, P. I. Pollack, J. C. Schaeffer, R. Tull, L. M. Weinstock, Ann. N.Y.Acad. Sci., 158,471 (1969). 5. J. F. Knifton, Tetrahedron Lett., 2163 (1975). 6. W. L. F. Armareg0,J. Chem. Soc., 5030 (1962). 7. R. B. Moffett, Org. Synth., Cull. vol. 4 , 357 (1963). 8. N. A. Cortese, R. F. Heck, J . Org. Chem., 42, 3491 (1977). 9. S. Ram, R. E. Ehrenkaufer, Synth. Commun., 16, 133 (1986). 10. H. W. R. Williams, Can. J . Chern., 54, 3377 (1976). 1 1 . H. S. Broadbent, Ann. N.Y.Acad. Sci., 145,58 (1967). 12. H. Greenfield, Ann. N.Y.Acad. Sci., 145,108 (1967). 13. J. R. Kosak, Ann. N.Y.Acad. Sci., 172, 175 (1970). 14. J. R. Kosak, U.S. Patent 4,020,107 (1977); Chem. Abstr., 86, 89364t (1977). 15. A. Onopchenko, E. T. Sabourin, C. M. Selwitz, J . Org. Chem., 44, 1233 (1979). 16. B. R. Baker, J. H. Jordaan, J. Med. Chem., 8, 35 (1965). 17. P. N. Rylander, I. M. Karpenko, G. R. Pond, U.S. Patent 3,694,509 (1977); Chem. Abstr., 76, 14082j (1972). 18. W. K.Seifert, P. C. Condit, J. Org. Chem., 28, 265 (1963). 19. L. K. Friedlin, E. F. Litvin, V. M. Chursina, Katal. Vosstanov.GidrirovZhidk. Faze, 59 (1970); Chem. Abstr., 76,45665f (1972). 20. I. Baxter, G. A. Swan, J . Chem. SOC. C , 2230 (1967). 14.3.7.2.3. Side Reactions In Polyfunctional Molecules.

If other active groups are present y or S to the reducible nitro group, hydrogenation can result in formation of nitrogen heterocyclic products. Several such cyclizations provide an entry to indoles, such as reductive cyclizations of dinitrostyrene 1 [equation (a)]', of o-nitrobenzyl ketone 2 [equation (b)]', and of nitro nitrile 3 [equation (c)],, all of them carried out on palladium-on-carbon. 10% Pd-C,45"C EtOH/EtOAc/AcOH

Me0

I

MeO R

R = H,I

14.3.7. Hydrogenation of Other Functional Groups 14.3.7.2. Nitro Compounds 14.3.7.2.3. Side Reactions in Polyfunctional Molecules. 10%Pd-C, 45’C EtOH/EtOAc/AcOH

coy

215

(b) I4

FONRR’ DMF, Pd-C 80°C

QcJo;NRR’

(c) H

Reductive cyclization of 2-nitroacetanilides over palladium-on-carbon in ethanol yields imidazoles [equation (d)J4. Ring closure between the reduced nitro group and an acid or ester function leads to a lactam [equation (e)J5:

% NHCOMe

5%Pd-C EtOH, 65OC

Ra Ni, 9OoC

Z

C

O

O

M

e

-+

H,, 14 X lo3 kPa

’

Sometimes cyclization occurs readily and cannot be prevented; otherwise it requires

a mild subsequent treatment. Cyclization may occur before reduction of the nitro group

to the amide is complete and a hydroxyl compound is formed?

OH Thus many heterocycles of various sizes can be formed and the reductive cyclization of polyfunctional nitro compounds offers a useful synthetic tool for the synthesis of complex molecules. When the interacting functions are not located at positions conducive to ring formation, bimolecular condensation or even polymerization may occur.

21 6

14.3. Hydrogenation Reactions 14.3.7. Hydrogenation of Other Functional Groups 14.3.7.3. Miscellaneous

Hydroxylamines which are formed by partial reduction of the nitro group may undergo rearrangements in acidic media, providing patented commercial syntheses of p-haloanilines and p-aminophenols'. (J.-L. GRAS)

1. R. A. Heacock, 0. Hutzinger, B. D. Scott, J. W. Daw, B. Witkop, J . Am. Chem. SOC., 85, 1825 (1963). 2. R. L. Augustine, A. J Gustavsen, S. F. Wanat, I. C. Pattison, K. S.Houghton, G. Koletar, J. Org. Chem., 38, 3004 (1973). 3. J. Bordais, C. Germain, Tetrahedron Lett., 195 (1970). 4. A. J. Neale, K. M. Davies, J. Ellis, Tetrahedron, 25, 1423 (1969). 5. C. D. Gutsche, H. R. Zandstra, J . Org. Chem., 39, 324 (1974). 6. A. L. Davies, 0. H. P. Choun, D. E. Cook, T. J. McCord, J . Med. Chem., 7,632 (1964). 7. N. R. W. Benwell, Br. Patent 1,181,969 (1970); Chem. Abstr., 76, 33943g (1972).

14.3.7.3. Miscellaneous

Besides the most commonly encountered functional groups surveyed in preceding sections, there are other chemical functions that can be reduced by hydrogenation. Azo compounds yield the primary amine over Pd catalyst','. Oximes are reduced to the primary amine in alkaline media3 or in acetic acid over Pt oxide. They afford the acetylated amine over Pd-on-carbon in the presence of acetic anhydride4. Hydrazones usually lead to the primary amine, but overreduction leads to the corresponding methylene derivative. If another functional group, one able to participate in a reductive amination reaction is present, secondary amines form. Halides are replaced by hydrogen under some conditions5, over Ni-A1 alloy3 or by hydrogen transfer6. The hydrodehalogenation is enantioselective over a Pd-BaSO, catalyst modified with the alkaloid cinchonine7. In addition to functional group reduction, side reactions may occur during the hydrogenation. Dehydrogenation of cyclohexadienes to the benzene ring' or migration of double bonds is observed over Pd9. Hydrogenolysis is common, especially of allylic and benzylic compounds. Reductive degradation of the N-0 bond of alkoxy amines and oxazoles occurs over finely divided Ni (Raney Ni), PtO, or Pd in alcohol as The hydrazine N-N bond is hydrogenolyzed to the amine over Raney Ni". N-N and N-0 bonds are also cleaved over Ni-A1 alloy3. Regioselective hydrogenolysis of benzyl g l y c ~ s i d e sand ' ~ acetylated furanose is achieved with Raney Nil4. Homogeneous Rh-catalyzed hydrogenation of sodium epoxysuccinate results in asymmetric hydrogenolysis of the epoxy ring, to produce enantioselectively the malic acid sodium salt15. Finally, the t-butyl dimethyl silyl protecting group of hydroxyl functions is removed by PdO catalyzed transfer hydrogenation from cyclohexene16. (J.-L. GRAS)

1. 2. 3. 4. 5.

6. 7. 8. 9.

T. J. Dunn, W. L. Neumann, M. M. Rogic, S. R. Woulfe, J . Org. Chem., 55,6368 (1990). M. Saia, M. Bessodes, K. Antonakis, Tetrahedron Asym., 2, 111 (1991). L. K. Keefer, G. Lunn, Chem. Rev.,89, 459 (1989). M. P. Georgiadis, S. A. Haroutounian, E. A. Couladouros, C. D. Apostopoulos, K. P. Chondros, J . Heterocyclic Chem., 28, 697 (1991). A. P. Pinder, Synthesis, 425 (1980). G. Brieger, T. J. Nestrick, Chem. Rev.,74,567 (1974). H.-U.Blaser, S. K. Boyer, U. Pittelkow, Tetrahedron Asym., 2, 721 (1991). P. G. Sammes, R. J. Whitby, J . Chem. SOC.Perkins Trans. I, 195 (1987). F.-P. Montforts, G. Zimmermann, Angew. Chem., Int. Ed. Engl., 25, 458 (1986).


21 6

14.3. Hydrogenation Reactions 14.3.7. Hydrogenation of Other Functional Groups 14.3.7.3. Miscellaneous

Hydroxylamines which are formed by partial reduction of the nitro group may undergo rearrangements in acidic media, providing patented commercial syntheses of p-haloanilines and p-aminophenols'. (J.-L. GRAS)

1. R. A. Heacock, 0. Hutzinger, B. D. Scott, J. W. Daw, B. Witkop, J . Am. Chem. SOC., 85, 1825 (1963). 2. R. L. Augustine, A. J Gustavsen, S. F. Wanat, I. C. Pattison, K. S.Houghton, G. Koletar, J. Org. Chem., 38, 3004 (1973). 3. J. Bordais, C. Germain, Tetrahedron Lett., 195 (1970). 4. A. J. Neale, K. M. Davies, J. Ellis, Tetrahedron, 25, 1423 (1969). 5. C. D. Gutsche, H. R. Zandstra, J . Org. Chem., 39, 324 (1974). 6. A. L. Davies, 0. H. P. Choun, D. E. Cook, T. J. McCord, J . Med. Chem., 7,632 (1964). 7. N. R. W. Benwell, Br. Patent 1,181,969 (1970); Chem. Abstr., 76, 33943g (1972).

14.3.7.3. Miscellaneous

Besides the most commonly encountered functional groups surveyed in preceding sections, there are other chemical functions that can be reduced by hydrogenation. Azo compounds yield the primary amine over Pd catalyst','. Oximes are reduced to the primary amine in alkaline media3 or in acetic acid over Pt oxide. They afford the acetylated amine over Pd-on-carbon in the presence of acetic anhydride4. Hydrazones usually lead to the primary amine, but overreduction leads to the corresponding methylene derivative. If another functional group, one able to participate in a reductive amination reaction is present, secondary amines form. Halides are replaced by hydrogen under some conditions5, over Ni-A1 alloy3 or by hydrogen transfer6. The hydrodehalogenation is enantioselective over a Pd-BaSO, catalyst modified with the alkaloid cinchonine7. In addition to functional group reduction, side reactions may occur during the hydrogenation. Dehydrogenation of cyclohexadienes to the benzene ring' or migration of double bonds is observed over Pd9. Hydrogenolysis is common, especially of allylic and benzylic compounds. Reductive degradation of the N-0 bond of alkoxy amines and oxazoles occurs over finely divided Ni (Raney Ni), PtO, or Pd in alcohol as The hydrazine N-N bond is hydrogenolyzed to the amine over Raney Ni". N-N and N-0 bonds are also cleaved over Ni-A1 alloy3. Regioselective hydrogenolysis of benzyl g l y c ~ s i d e sand ' ~ acetylated furanose is achieved with Raney Nil4. Homogeneous Rh-catalyzed hydrogenation of sodium epoxysuccinate results in asymmetric hydrogenolysis of the epoxy ring, to produce enantioselectively the malic acid sodium salt15. Finally, the t-butyl dimethyl silyl protecting group of hydroxyl functions is removed by PdO catalyzed transfer hydrogenation from cyclohexene16. (J.-L. GRAS)

1. 2. 3. 4. 5.

6. 7. 8. 9.

T. J. Dunn, W. L. Neumann, M. M. Rogic, S. R. Woulfe, J . Org. Chem., 55,6368 (1990). M. Saia, M. Bessodes, K. Antonakis, Tetrahedron Asym., 2, 111 (1991). L. K. Keefer, G. Lunn, Chem. Rev.,89, 459 (1989). M. P. Georgiadis, S. A. Haroutounian, E. A. Couladouros, C. D. Apostopoulos, K. P. Chondros, J . Heterocyclic Chem., 28, 697 (1991). A. P. Pinder, Synthesis, 425 (1980). G. Brieger, T. J. Nestrick, Chem. Rev.,74,567 (1974). H.-U.Blaser, S. K. Boyer, U. Pittelkow, Tetrahedron Asym., 2, 721 (1991). P. G. Sammes, R. J. Whitby, J . Chem. SOC.Perkins Trans. I, 195 (1987). F.-P. Montforts, G. Zimmermann, Angew. Chem., Int. Ed. Engl., 25, 458 (1986).

14.3. H drogenation Reactions 14.3.7. Lydrogenationof Other Functional Groups 14.3.7.3. Miscellaneous 10. 11. 12. 13. 14. 15. 16.

21 7

D. K. Pide, W. M. Welch, P. D. Weeks, R. A. Volkmann, Tetrahedron Lett., 27, 1549 (1986). V. Mancuso, C. Hootele, Tetrahedron Lett., 29,5917 (1988). A. Alexakis, N. Lenson, P. Mangeney, Tetrahedron Lett., 32, 1171 (1991). T. Bieg, W. Szejia, Carbohydrate Res., 205, C10 (1990). M. Okabe,R.-C. Sun, G . B. Zenchoff, J. Org. Chem., 56,4392 (1991). A. S. C. Chan, J. P. Coleman, J. Chem. SOC, Chem. Commun., 535 (1991). J. F. Cornier, Tetrahedron Lett., 32, 187 (1991).


14.4. Addition Reactions 14.4.1. Introduction Hydrosilylation describes the addition of hydrosilanes to unsaturated bonds. Hydrosilylation of carbon-carbon multiple bonds has been extensively studied for more than 40 years and is used for the industrial production of organosilicon compounds such as adhesives, binders, and coupling agents. Hydrosilylation is a convenient method for the synthesis of various organosilicon compounds. Although hydrosilylation sometimes takes place at high temperatures (>3OO0C) without any catalyst, the reaction is effectively promoted by ultraviolet light, y-irradiation, electric discharge, and catalysts such as peroxides, metals, metal salts, or metal complexes, or, in special cases, amines. Chloroplatinic acid is the most commonly used catalyst, but group VIII transition metal species such as tertiary phosphine complexes of nickel, palladium, platinum, and rhodium are extremely effective. Inorganic and organic syntheses using organosilicon reagents are one of the most remarkable advances in synthetic chemistry in the last decades. The hydrosilylation of various functional groups catalyzed by transition metal complexes serves as a convenient method for the synthesis of such organosilicon reagents as well as a unique and effective method for the asymmetric, chemoselective, or stereoselective reductions of carbonheteroatom bonds. This section summarizes the scope and limitations of hydrosilylation as a synthetic method. (I. OJIMA)

14.4.2. Hydrosilylation of Olefins and Acetylenes Hydrosilylation of carbon-carbon multiple bonds, discovered in the late 1940s, has been one of the most important laboratory and industrial methods of forming siliconcarbon bonds', The reaction is especially valuable for the introduction of organic radicals, which contain reactive functional groups to silicon, since these cannot often be achieved by simple organometallic procedures. In the early days, the reaction was carried out under the influence of UV light, y-irradiation, or electric discharge as well as peroxides and similar free-radical catalysts. The latter were used intensively for some years after the discovery of the reaction.2 Platinum catalysts, especially chloroplatinic acid, are now the most commonly used, along with other transition metal catalysts. In this section the hydrosilylation of olefins and acetylenes promoted by transition metal catalysts is described and some newer aspects of the reaction are disclosed. (I. OJIMA)

14.4.2.1. By Platinum Catalysts

Chloroplatinic acid, platinum on carbon or alumina, platinum(I1) complexes with Among them, olefins, and platinum(0) complexes with phosphines have been 218

14.4. Addition Reactions 14.4.2. H drosilylation of Olefins and Acetylenes 14.4.2.1. b y Platinum Catalysts

21 9

chloroplatinic acid, H,PtC1,.6H2O, is by far the most common and efficient catalyst particularly for the hydrosilylation of olefinic ~ubstratesl-~. Chloroplatinic acid is frequently used as a solution in i~opropanol~. As chloroplatinic acid is an excellent catalyst, only a small quantity is necessary, usually in the range of mol/mol of hydrosilane. In some cases, as little as lo-' mol/mol of reactant has been successfully used6. Solvents are not usually employed, but sometimes benzene, toluene, xylene, chloroform, dichloromethane, and tetrahydrofuran have been used'. In the presence of H2PtC1,, hydrosilanes add to various 1-alkenes to give the corresponding terminal adducts in high yields'-4. Halogenated olefins such as 3,4-dichlorobut- 1-ene', 3,3,3-trifluoropropenep, perfluoroethylene', and 1,l-difluoroethene' are also readily hydrosilylated. 1,l -Difluoroethene is hydrosilylated regioselectively to give 2,2-difluoroethylsilane'. However, the hydrosilylation of styrene with HSiCl, gives a mixture of a- and P-isomers": CH,=CHPh

+ HSiCl,

H2PtC1,.6H20

Cl,SiCH,CH,Ph

+

(58.7%)

Cl,SiCH( CH,)Ph (38.5%)

-

(a)

The proportion of p-isomer can be increased to 92-95% on adding triphenylphosphine or pyridine to the catalyst". Cycloalkenes can be employed as substrate, but the reactivity is considerably lower than that of l-alkenes6s". The hydrosilylation of n-alkenes having an inner double bond proceeds with a remarkable isomerization to give a terminal silyl isomer as predominant product :

P

+

H,PtCl,.6H20

HSiMe20SiMe2H

SiMe,OSiMe,H (87%)

SiMe20SiMe,H i(7%) SiMe'OSiMe'H

+ HSiC1,

SiCl,

H2RC1,-6H2O

(90%)

Dichlorosilane adds to inner olefins without is~merization'~: SiHC1,

+

H2SiC1,

H2RC16'6H20,

(34%)

+ SiHCl,

220

14.4. Addition Reactions 14.4.2. Hydrosilylation of Olefins and Acetylenes 14.4.2.1. By Platinum Catalysts

Many olefin functional groups survive hydrosilylation. Thus, olefins bearing oxirane, acetal, ester, nitrile, amine, amide, nitro, ketone, carbamate, ether, isocyanate, phosphate, phophonic dichloride, dialkoxyborane, sulfide, sulfone, or carborane groups are readily hydrosilylated without affecting the functional groups, as shown in Table 1. In some cases, the hydrosilylation is hampered by severe side reactions. Ally1 chloride can be hydrosilylated, but reduction to propene accompanies the r e a ~ t i o n ~ ’ , ~ ~ : CH,=CHCH,Cl

+ HSiC1,

+ CH,=CHCH,

HZPtCl6.6H20

+ Cl3Si(CH,),C1

(e)

Hydrochlorosilanes do not hydrosilylate 1,3-alkadiene monoxide, but open the oxirane ring34v35: 2 CH,-CH-CH=CH, ‘0’

+

(CH,= CHCHClCH,O),SiHMe HSiMeC1, -+

+

CH, =CHCHClCH,OSiHMeCl

(f)

The hydrosilylation of acetylene with chlorohydrosilanes is effectively catalyzed by the platinum catalyst without any heating to give vinylchlorosilanes in high yield36. Although exclusive terminal addition of hydrosilanes to 1-alkynes is claimed’, careful investigation discloses that substantial amounts of side product(s) are formed. Typically, the hydrosilylation of hex-1-yne with trichlorosilane (1: 1) in the presence of H,PtCl, gives a mixture of 1- and 2-trichlorosilylhex-1-ene (78:22),‘?

+

-%

H,PtC1,*6HZ0

~

TH + TH (g)

HSiCl,

H

SiC13

C1,Si

H

A mixture of 1:l adducts (62%) and two isomeric 2:l adducts, 1,6bis(trichlorosily1)hexane (20%) and 1,2-bis(trichlorosilyl)hexane (18%), is produced when using the catalyst in higher concentration for prolonged periods of time37. The proportion of the 1-silyl isomer in the mixture of 1:l adducts depends upon the substituents of both alkenes and hydrosilanes, and the solvents employed. The trans- 1-silyl isomer is exclusively produced through cis-addition. This is characteristic of the hydrosilylation of alkynes in the presence of platinum catalysts. In contrast trans-addition giving the cis-adduct is the predominant process on using free radical initiators3*.Dichlorosilane also adds to 1-alkynes in the presence of platinum catalyst giving a mixture of 1:l and 1:2 ad duct^^^:

-+

+

H,PtCI,-6H20

H2SiCl,

H

+

SMC1, (52%)

C1,HSi

H (12%)

(h)

Hw H

(16%) H

0

CH=CH, CH,=CHCH,NO, CH,=CH(CH,),COMe CH,=CHCH,OCONH, CH,=CHOPh CH,=CHCH,NCO CH,=CHCH,OP(O)(OEt)Me CH,=CHCH,P(O)CI, CH,=CHB(OBu’), CH,=CHSPh CH,=CHSO,Ph CH,=CH(CH,)Z-C =CH \ / BIOHIO

0I

CH,=CHCH(OEt), CH,=CHCH,CO,Et CH,=CHCN CH,=CHOCOMe CH,=CHCH,NH, CH,=CHCHiNHSiMe,

‘O/

CH2=CHCHzOCHzCH- CH,

Olefin

HSiCl, HSiMeCl, HSiMeEt, HSiEtC1, HSiMeCl, HSi(OEt), HSiMeC1, HSiCl, HSiMeEt, HSiMeCI, HSiMeClOSiMqCl

HSiEt, HSiEt, HSiCl, HSiMeC1, HSiEt, HSi(OEt), HSiMqSiMqPh

HSi(OEt),

Hydrosilane

n

CH,

CH,CH,SiMqSiMe,Ph CI,Si(CH,),NO, Cl,MeSi(CH,),COMe Et,MeSi(CH,),OCONH, Cl,EtSi(CH,),OPh Cl,MeSi(CH,),NCO (EtO),Si(CH,),OP(O)(OEt)Me Cl,MeSi( CH,),P(O)Cl, CI,Si(CH,),B(OBu’), Et,MeSi(CH,),SPh C1,MeSi(CH2),SO2Ph Me,ClSiOSiMeCl(CHd C -CH \-/ B,OHIO

Et,Si(CH,),CH(OEt), Et,Si(CH,),CO,Et Cl,Si(CH,),CN Cl,MeSi(CH,),OCOMe Et,Si(CH,),NH, (EtO),Si(CH,),NHSiMe,

‘O/

(EtO),Si(CH,),OCH,CH-

Product

TABLE1. HYDROSILYLATION OF FUNCTIONALIZED OLEFINS CATALYZED BY CHLOROPLATINIC ACID

99

85

92 84 84 77 78 47 50 73 30 -

71 62 61 23

54

100

Yield (%)

22 23 24 25 24 26 27 28 29 30 31

16 17 18 6 19 20 21

15

Reference

222


Again, the mode of the hydrosilane addition is exclusively cis. The hydrosilylation of acetylenic bonds is faster than that of olefinic bonds, and vinylacetylene is chemoselectively hydrosilylated to give 1-silylbuta- 1,3-diene4’s41:

+

H,F‘tC1,’6H20

> R,Si-CH=CH-CH=CH, (i) HC=C-CH=CH, HSiR, The hydrosilylation of propargyl alcohol gives a mixture of 7- and P-adducts and their dehydrogenatively silylated although earlier publications claimed only the formation of y-adducp. truns-Di-~-hydridobis(~cyclohexylphosphine)bis(silyl)diplatinumcomplexes,e.g., [Pt(PCy3)(SiMe,,CH,Ph)(pH)12 (Cy = cyclohexyl), are effective catalysts for the hydrosilylation of alkynes. The reaction proceeds under mild conditions (-65°C) in high yield (80-98%) at low catalyst/alkyne ratio 10-5)4’,46.Addition of hydrosilanes, X,SiH (X = alkyl, alkoxy or chloro), to alkynes, RC=CH, RCECR, and RC=CR’, proceeds in a stereospecific cis-fashion giving trans-RC=CHSiX,, cisRC(SiX,)=CHR and mixtures of the regioisomers (E)-RC(SiX,)=CHR’ and (E)RCH=C(SiX,)R’, respectively. The mechanism of catalytic hydrosilylation involves oxidative addition of a siliconhydrogen bond to a metal complex as an essential step since it is here the activation of hydrosilane by the catalyst takes place. Thus, many transition metal ions and complexes, especially group VIII metals in low oxidation state containing .rr-acid ligands such as CO, tertiary phosphines or olefins display catalytic activity. The sequence of unit reactions in a typical d8-metal complex-catalyzed hydrosilylation is summarized as47,48 (i) Olefin complex formation

-

-M-

I I

I

RICH 11-MCH,

+ R’CH=CH2

(j)

I

(ii) Oxidative addition R‘CH ll-M-+ CH2

I

I

HSiR,

d

11-MCHd

SiR, I /H

I

(iii) cis-Ligand insertion R’CH M ,-II CH, (iv) Reductive elimination SiR, R’CHZCH,-M-

I/

/I

SiR3 1 ,H

I

SiR, &

+ R’CH2=CH2

-+

R’CH&!HZ-M-

I/

’I

R’CH,CH2SiR3

I + RICH II-MCH2 I

(1)

(m)


223

The active catalyst species arising from chloroplatinic acid involves platinum in several oxidation states since (1) partial reduction to Pt(I1) is observed for the solution of H,PtCI, in i s o p r o p a n 0 1 ~ ~ * ~ ~ H,PtC1,-6H,O

+ Me,CHOH

+ H2PtC14

+ Me,C=O + 2 HCl + 6 H,O

(n)

and the reduction proceeds further to Pt(0) upon storage, and (2)hydrosilanes used can also reduce chloroplatinic acid to lower valent platinum compounds or to the free metal”. It seems clear that catalysis by chloroplatinic acid involves homogeneous rather than heterogeneous processes, and Pt(I1) species are the most effective catalyst in the mixture of H,PtCl,, olefin, and h y d r o ~ i l a n e ~Various ~. Pt(I1) complexes48 such as

bis(triphenylphosphine)dichloroplatinum(II), dichlorobis-(ethylene)-p,p’-dichlorodiplatinum(II), and dichloro(ethylene)(pyridine)platinum(II~~are effective homogeneous catalysts for hydrosilylation with results similar to those achieved by H,PtCl,. On the other hand, platinum metal is also a catalyst used as platinum black, or supported on carriers such as carbon, y-alumina, and silica. These heterogeneous catalysts show much less activity than the homogeneous platinum catalyst^'-^. The regio-, chemo-, and stereoselectivity of hydrosilylation catalyzed by platinum metal resemble but are not always the same as those of the homogeneous catalysis. The most striking difference can be observed in the hydrosilylation of norbomadiene with HSiMeCl,, which affords 3-dichloromethylsilylnortricyclene (1) in 94% yield and endo-dichloromethylsilylbicyclo[2.2.1]hept-Zene (endo-2) in 6% yield on using Pt/C as a catalyst, whereas endo-2 is the major product (64%) accompanied by the formation of 1 (30%) and exo-2 (6%) on using H,PtC123:

A

+ HSiMeCl, &,SiMeCl, I H

1

+

bSiMeC1, I H exo-2

+

b

H (0) SiMeC1, I

endo-2 (I. OJIMA)

1. E. Lukevics, M. G. Voronkov, Organic Insertion Reactions of Group IV Elements, Consultants Bureau, New York, 1966. 2. C. Eabom, R. W. Bott, in The Bond to Carbon, Vol. 1, Part I, A. G. MacDiarmid, ed., Dekker, New York, 1968, pp. 213-279. 3. E. Lukevics, Z. V. Belyakova, M. G. Pomerantseva, M. G. Voronkov, in Organometallic Chemistry Reviews, Journal of Organometallic Chemistry Library, Vol. 5, D. Seyferth, A. G. Davies, E. 0. Fischer, J. F. Normant, 0. A. Reutov, eds., Elsevier Scientific Publ. Co., Amsterdam, 1977, pp. 1-179. 4. J. L. Speier, Adv. Organomet. Chem. 17,407-447 (1979). 5 . This system is referred to as Speier’s catalyst, in recognition of the work done by the American chemist J. L. Speier in the discovery of its usefulness. 6. J. L. Speier, J. A. Webster, G. H. Barnes, J . Am. Chem. SOC., 79, 974 (1957). 7. M. A. Mamedov, I. M. Akhmedov, M. M. Guseinov, S . I. Sadykh-Zade, Zh. Obshch. Khim., 35,461 (1965); J . Gen. Chem. USSR, 35,458 (1965). 8 . 0. W. Steward, 0. R. Pierce, J . Org. Chem., 26, 2943 (1961). 9. V. A. Ponomarenko, V. G. Cherkaev, A. D. Petrov, N. I. Zadorazhnyi, Izv. Akad. Nauk SSSR, Otd. Khim. Nauk, 247 (1958).

224


M. Ciipka, P. Svoboda, J. HetflejS, Collect. Czech. Chem. Commun., 38, 3830 (1973). H. Takahashi, H. Okita, M. Yamaguchi, I. Shiihara, J . Org. Chem., 28, 3353 (1963). H. M. Bank, J. C. Saan, J. L. Speier, J . Org. Chem., 29, 792 (1964). D. Seyferth,J. Org. Chem., 22, 1252 (1957). R. A. Benkeser, W. C. Muench, J. Am. Chem. SOC., 95,285 (1973). E. P. Plueddemann, G. Fanger, J . Am. Chem. Soc., 81, 2632 (1959). A. D. Petrov, S. I. Sadykh-Zade, Bull. SOC.Chim. Fr., 1932 (1959). F. Rijkens, M. J. Janssen, W. Prenth, G. J. M. van der Kerk, J . Organomet. Chem., 2, 347 (1964). 18. 2.V. Belyakova, S. A. Golubtsov, T. M. Yakusheva, Zh. Obshch. Khim., 35, 1183 (1965);J . Gen. Chem. USSR,35, 1187 (1965). 19. N. S. Nametkin, A. V. Topchiev, T. I. Cherysheva, I. N. Lyashenko, Dokl. Akad. Nauk SSSR, 140, 384 (1961); Chem. Abstr., 56,493 (1962). 20. J. C. Saam, J. L. Speier, J . Org. Chem., 24, 119 (1959). 21. L. A. Haluska, U.S. Patent, 3,249,586 (1966); Chem. Abstr., 65, 2296 (1966). 22. S. S.Novikov, V. V. Sevost’yanova, Izv. Akad. Nauk SSSR,Otd. Khim. Nauk, 1485 (1962). 23. N . V. Komarov, V. K. Roman, Zh. Obshch. Khim., 35,2017 (1965);J . Gen. Chem. USSR,35, 2008 (1965). 24. V. F. Mironov, V. P. Kozyukov, V. D. Sheludyakov,Dokl. Akad. Nauk. SSSR,178,358 (1968). 25. G. C. Balezina, G. M. Alekseeva, Izv. Sibirsk. Otd. Akad. Nauk SSSR Ser. Khim. Nauk, 92 (1963); Chem. Abstr., 60, 14534 (1964). 26. E. F. Bugerenko, A. S. Petukhova, E. A. Chemyshev, Zh. Obshch. Khim., 40, 606 (1970). J. Gen. Chem. USSR,40,576 (1970). 27. E. F. Bugerenko, E. A. Chemyshev, A. D. Petrov, Dokl. Akad. Nauk SSSR, 143, 840 (1962). 28. B. M. Mikhailov, P. M. Aronovich, L. V. Tarasova, Zh. Obshch. Khim., 30, 3624 (1960); J . Gen. Chem. USSR,30,3592 (1960). 29. A. A. Dzhafarov, I. A. Aslanov, D. A. Kochkin, Zh. Obshch. Khim., 45,2023 (1975);J. Gen. Chem. USSR,45, 1986 (1975). 30. A. Berger, U. S.Patent, 3,483,241 (1969); Chem. Abstr., 72, 32010f (1970). 31. N. Mayes, J. Green, M. S. Cohen, J . Polymo Sci., A-I, 365 (1967). 32. V .F. Mironov, V. V. Nepomnina, Izv. Akad.Nauk SSSR,Otd. Khim. Nauk, 2140 (1960);Chem. Absrr., 55, 15331 (1961). 33. A. V. Topchiev, N. S. Nametkin, T. I. Chemysheva, S.G. Durgar’yan, Dokl. Akad. Nauk SSSR, 110, 97 (1956). 34. E. V. Nekhorosheva, V. M. Al’bitskaya, Zh. Obshch. Khim., 38, 1511 (1968); J . Gen. Chem. USSR,38, 1461 (1968). 35. I. E. Sharikova, V. M. Al’bitskaya, A. A. Petrov, Zh. Obshch. Khim., 34,2262 (1964).J. Gen. Chem. USSR,34,2275 (1964). 36. V. A. Ponomarenko, V. G. Cherkaev, A. D. Petrov, N. I. Zadorazhnyi, Izv. Akad. Nauk SSSR, Otd. Khim. Nauk, 247 (1958). 37 * R. A. Benkeser, R. F. Cunico, S. Dunny, P. R. Jones, P. G. Nerlekar, J . Org. Chem., 32,2634 ( 1967). 38. R. A. Benkeser, Pure Appl. Chem., 13, 133 (1966). 39. R. A. Benkeser, D. F. Ehler, J. Organomet. Chem., 69, 193 (1974). 40. A. D. Petrov, S. I. Sadykh-Zade, Izv. Akad. Nauk SSSR,Otd. Khim. Nauk, 513 (1958). 41. M. D. Stadnichuk, A. A. Petrov, Zh. Obshch. Khim., 33,3563 (1963); Chem. Abstr., 60,9306 (1964);J . Gen. Chem. USSR,33,3495 (1963). 42. L. L. Shchukovskaya, R. I. Pal’chik,Zh. Obshch. Khim., 35,1122 (1965);J . Gen. Chem. USSR, 35, 1126 (1965). 43. M. G. Voronkov, S.V. Kirpichenko, V. V. Keiko, L. V. Sherstyannikova, V. A. Pestunovich, E. 0. Tsetlina, Izv. Akad. Nauk SSSR,390 (1975); Chem. Abstr., 83, 10252m (1975). 44. S. I. Sadykh-Zade, I. A. Shikhiev, E. M. Kholilova, Zh. Obshch. Khim., 34, 1393 (1964);J. Gen. Chem. USSR,34, 1395 (1964). 45. C. A. Tsipis, J . Organorner. Chem., 187, 427 (1980). 46. M. Green, J. L. Spencer, F. G. A. Stone, C. A. Tsipis, J . Chem. SOC. Dalton Trans., 1525 (1977). 47. A. J. Chalk, Trans. N. Y. Acad. Sci., 11.32, 481 (1970). 48. A. J. Chalk, J. F. Harrod, J . Am. Chem. Soc., 87, 16 (1965). 10. 11. 12. 13. 14. 15. 16. 17.

14.4. Addition Reactions 14.4.2. H drosilylation of Olefins and Acetylenes 14.4.2.2. b y Rhodium and Nickel Catalysts

225

49. M. G . Voronkov, V. B. Pukharevich, S. P. Sushchinskaya, L. I. Kopylova, B. A. Trofimov, Zh. Obshch. Khim., 41,2102 (1971). J . Gen. Chem. USSR,41, 2120 (1971). 50. J. Lahaye, R. Lagarde, Bull. SOC. Chim. Fr., 2999 (1974). 51. C. Eabom, B. C. Pant, E. R. A. Peeling, S. C. Taylor, J . Chem. SOC.,C , 2823 (1969). 52. G . Koemer Ger. Patent, 1,203,776 (1965); Chem. Abstr., 64, 2126 (1966). 53. H. G . Kuivila, C. R. Warner, J . Org. Chem., 29, 2845 (1964).

14.4.2.2. By Rhodium and Nickel Catalysts

Rhodium(1) complexes such as Rh(PPh,),Cl’, Rh(PPh,),(CO)Cl’, Rh(PEt,),(CO)Cl’, Rh(PPh,),(CO)H, [Rh(CO),Cl],’, [Rh(C,H,),Cl],, and Rh(acac)(CO), (acac = acetylacetonato) have similar catalytic activity to platinum catalysts in the hydrosilylation of olefins’, although the efficiency of H,PtCl, is much higher than that of the rhodium catalysts. Accordingly, it is of no advantage to use rhodium complexes as catalysts for simple hydrosilylation of olefins. Nickel exhibits lower catalytic activity than platinum and rhodium catalyst^^*^, and in many cases, phosphine-nickel complexes cause the disproportionation of chlorohydrosilanes giving complex results’. However, in some cases, the regioselectivity in the hydrosilylation of styrene with trichlorosilane catalyzed by nickel complexes is quite different from that achieved by platinum or rhodium catalysts. For instance, a-adduct is exclusively formed by the catalysis of [N~(CO)(T-C,H,)],~: PhCH=CH,

+ HSiCl,

“i(CO)(.rr-C5H5)I2

20°C 2.5 h

> PhCHCH,

AiCI,

(a)

(82%)

The hydrosilylation of acrylonitrile, crotononitrile, and cinnamonitrile catalyzed by Rh(PPh,),Cl’ or Rh(PPh,),(CO)H6 affords a-adduct exclusively: R’CH=CHCN

+ HSiR’R;

Rh(PPh,),Cl

R’CH,CHCN kiRZR:

(b)

in contrast to that obtained using chloroplatinic acid7, platinized asbestos, benzoyl peroxide, amine bases, or tertiary phosphines in a stainless steel vessel, which give the P-adduct selectively’. Asymmetric hydrosilylation of a-methylstyrene with HSiMeC1, catalyzed by [( )BMPP],NiCl, (BMPP = benzylmethylphenylphosphine) gives the corresponding optically active P-adduct, 1, in 31% yield (20.9%enantiomeric excess (e.e.), R)’. When a chiral platinum complex, [( +)BMPP-PtCl,],, is employed, the reaction gives 1 (5.2% e.e., R ) in 56% yield”. Similarly, the asymmetric addition of HSiMe, to a-methylstyrene catalyzed by [( +)BMPP-Rh(NBD)]+ClO, (NBD = norbomadiene) affords 2 (7.0% e.e., R ) in 63% yield’:

+

Ph Me

>

C =CH,

+ HSiMeX,

Ph

(+)BMPP-M

> Me,

‘

EHCH,SiMeX,

1X=C1; 2 X = M e

(c)

Hydrosilylation of 1-hexyne with Et,SiH catalyzed by Rh,(CO),,, Co,Rh,(CO),,, Co,Rh(CO),,, and RhCl(PPh,), gives a mixture of cis- 1-triethylsilyl-1-hexene(3a)



225

49. M. G . Voronkov, V. B. Pukharevich, S. P. Sushchinskaya, L. I. Kopylova, B. A. Trofimov, Zh. Obshch. Khim., 41,2102 (1971). J . Gen. Chem. USSR,41, 2120 (1971). 50. J. Lahaye, R. Lagarde, Bull. SOC. Chim. Fr., 2999 (1974). 51. C. Eabom, B. C. Pant, E. R. A. Peeling, S. C. Taylor, J . Chem. SOC.,C , 2823 (1969). 52. G . Koemer Ger. Patent, 1,203,776 (1965); Chem. Abstr., 64, 2126 (1966). 53. H. G . Kuivila, C. R. Warner, J . Org. Chem., 29, 2845 (1964).

14.4.2.2. By Rhodium and Nickel Catalysts

Rhodium(1) complexes such as Rh(PPh,),Cl’, Rh(PPh,),(CO)Cl’, Rh(PEt,),(CO)Cl’, Rh(PPh,),(CO)H, [Rh(CO),Cl],’, [Rh(C,H,),Cl],, and Rh(acac)(CO), (acac = acetylacetonato) have similar catalytic activity to platinum catalysts in the hydrosilylation of olefins’, although the efficiency of H,PtCl, is much higher than that of the rhodium catalysts. Accordingly, it is of no advantage to use rhodium complexes as catalysts for simple hydrosilylation of olefins. Nickel exhibits lower catalytic activity than platinum and rhodium catalyst^^*^, and in many cases, phosphine-nickel complexes cause the disproportionation of chlorohydrosilanes giving complex results’. However, in some cases, the regioselectivity in the hydrosilylation of styrene with trichlorosilane catalyzed by nickel complexes is quite different from that achieved by platinum or rhodium catalysts. For instance, a-adduct is exclusively formed by the catalysis of [N~(CO)(T-C,H,)],~: PhCH=CH,

+ HSiCl,

“i(CO)(.rr-C5H5)I2

20°C 2.5 h

> PhCHCH,

AiCI,

(a)

(82%)

The hydrosilylation of acrylonitrile, crotononitrile, and cinnamonitrile catalyzed by Rh(PPh,),Cl’ or Rh(PPh,),(CO)H6 affords a-adduct exclusively: R’CH=CHCN

+ HSiR’R;

Rh(PPh,),Cl

R’CH,CHCN kiRZR:

(b)

in contrast to that obtained using chloroplatinic acid7, platinized asbestos, benzoyl peroxide, amine bases, or tertiary phosphines in a stainless steel vessel, which give the P-adduct selectively’. Asymmetric hydrosilylation of a-methylstyrene with HSiMeC1, catalyzed by [( )BMPP],NiCl, (BMPP = benzylmethylphenylphosphine) gives the corresponding optically active P-adduct, 1, in 31% yield (20.9%enantiomeric excess (e.e.), R)’. When a chiral platinum complex, [( +)BMPP-PtCl,],, is employed, the reaction gives 1 (5.2% e.e., R ) in 56% yield”. Similarly, the asymmetric addition of HSiMe, to a-methylstyrene catalyzed by [( +)BMPP-Rh(NBD)]+ClO, (NBD = norbomadiene) affords 2 (7.0% e.e., R ) in 63% yield’:

+

Ph Me

>

C =CH,

+ HSiMeX,

Ph

(+)BMPP-M

> Me,

‘

EHCH,SiMeX,

1X=C1; 2 X = M e

(c)

Hydrosilylation of 1-hexyne with Et,SiH catalyzed by Rh,(CO),,, Co,Rh,(CO),,, Co,Rh(CO),,, and RhCl(PPh,), gives a mixture of cis- 1-triethylsilyl-1-hexene(3a)

226

14.4. Addition Reactions 14.4.2. Hydrosilylation of Olefins and Acetylenes 14.4.2.2. By Rhodium and Nickel Catalysts

(major), trans-isomer (4a) (minor), and a-isomer (5a) (minor) in excellent yield' A rhodium+arbene complex also promotes the reaction giving the cis-product (3a) predominantly, in a similar ratio to that obtained using RhC1(PPh,),14. Under optimum conditions, the yield of l a increases to 96%13.The cisltrans ratio depends on the concentration of catalyst and the substituents of hydrosilane used',. The lower the catalyst concentration, the higher is the cisltrans ratio. Triethylsilane, HSiMe,Ph, HSiEt,Me, and HSiMe,Et give thermodynamically unfavorable cis-isomers as the major products, whereas HSiMe,Cl, HSiMe,Cl, and HSi(OMe), do not give cis-isomer (3) at all under usual conditionsL3. 13'*313.

SiR,

+

\ -

+

HSiR,

SiR,

3

+

(d) SiR,

4

A mechanism is proposed to accommodate the observed stereo~electivity'~. The mechanism includes a first silicon shift to an acetylenic bond and a carbene-type zwitterionic rhodium complex (7) as the key intermediate, which undergoes isomerization from a higher energy form (Z-complex, 6) to a lower energy form (E-complex, 8) followed by reductive elimination to cis-isomer (3) as the kinetic product.

R'-C

+ R,Si-

=C-H

[Rh](H)

Si-shift

H

SiR,

H

x

R'

trans

H

xH

R'

cis

SiR3

HSi(OMe)3, HSiMe2C1, HSiMeC12: k?,,, >> k I HSiEt3, HSiMe2Et, HSiMezPh: kyra,, c k', In contrast to other catalysts, nickel and Et3A1 give coupling products in the hydrosilylation of terminal acetylenes?

2 RC=CH

+

HSiX,

Ni(acac),

*IEt,

CH,=C(R)-C(R)=CHSiX, 9 RCH=CH-C(R)=CHSiX, 10

+

(f)

Although there are four possible structures for the coupling product, even without considering geometrical isomers, only two are found to be formed, and the head-to-head

14.4. Addition Reactions 14.4.2. Hydrosilylation of Olefins and Acetylenes 14.4.2.2. By Rhodium and Nickel Catalysts

227

isomer, 9, predominates (-80%). Reaction with 2-methylbut- 1-en-3-yne gives isomer 10 as the major product. With diethyl(bipyridyl)nickel(II) and chlorohydrosilanes, the hydrosilylation of disubstituted acetylenes undergoes dehydrogenative, stereoselective cis-double silylation to give 1,2-bissilylalkenes, accompanied by normal hydrosilylated products? RC =CR

+

HSiX3

XR + XR

R Ni(bipy)Et,

R

X3Si

Six3

11

Six3

H

(g)

12

The product ratio depends on the substituents on the acetylene, e.g., alkyl substituents favor the formation of bissilylalkene 11(R = Et, X = C1,11/12 = 90/10), while phenyl groups favor normal hydrosilylation (R = Ph, X, = MeCl,, 11/12 = 5/95). Platinum and rhodium complex catalysts can be immobilized on organic polymers or on silica. Various organic and inorganic supports have been studied”-,,. An aminated silica (13) was treated with H2PtC1, and RhCl(PPh,), to give the aminated silica-supported platinum (13-Pt) and rhodium (13-Rh) catalysts, respectively. The metal content in the immobilized catalysts was 0.05 wt% for Pt and 0.35-0.49 wt% for Rh depending on the structure of amine moiety. The hydrosilylation of 1-alkenes, allyl chloride, and allyl chloroacetate by HSiC1, and HSi(OEt,) was studied; 13-Rh (NR, = morpholino) retains its high catalytic activity even after nine times of use22.

/

/

0

\ silica -0--Si-(CH2)3-

/

0

\

NR2 Silica j-O-Si-(CH2)3-NR2 4,

0

M

/

(h)

0

\

\ 13

13-Pt: 13-Rh:

M = Pt M=Rh

Polyamides were successfully used as supports. Polyamides 14 and 15 were allowed to react with H,PtCl6, PtCl,(CH,CN),, and [RhCl(CO),], to give the corresponding immobilized catalysts. The IR study shows that the carbonyl oxygens of amides are the sites of coordination to platinum and rhodium. It was found that the activity of those immobilized catalysts correlates with the crystallinity of the polyamides, viz., the higher the crystallinity, the lower is the catalytic activity. Partial phosphination of 15 did not show any effects on catalytic activity. The polyamide-bound catalysts are thermally stable and reu~able’~.

14 15 Aminophosphine ligands were anchored on polystyrene or silica (16,17) and complexed with rhodium by using [RhCl(CO)], as a precursor. The immobilized catalyst showed good catalytic activities; the catalytic activity of 16-Rh was higher than 17-Rh18, although the activity decreased rapidly.

228


polystyrene or silica

-CH2

polystyrene or silica

p P P h 2 NMez 16

Me2N

PPh2

17

Diphenylphosphinoalkylorganosiloxaneswere immobilized onto silica (18,19) and

complexed with rhodium by using RhCl(PPh,), as a precursor. The metal content of 18-Rh and 19-Rh was 0.9 wt% with the P/Rh ratio of 2.0-2.3. Both immobilized catalysts retained good catalytic activities after being used three times at 120°C.

-0-&-OH

/ \

0

18:xly

=

3 19: x = 0

Diphenylphosphinoalkylated silicas (20; n = 1-6) were complexed with rhodium by using RhCl (PPh3)3or [RhCl (CHz=CH,),]2 as precursors to give the corresponding immobilized catalysts20~21. The catalysts (20-Rh; n = 2 , 3 ) contain 1.93-2.65 mol% Rh metal with the P/Rh ratio of 7.1-10.320. There were essentially no polymer effects on kinetic parameters, in comparison with those for homogeneous counterparts, although a 9- to 10-min induction period was observed". Catalytic activity depends on the methylene chain length. In the hydrosilylation of 1-hexene'l, the diphenylphosphinomethyl group (20-Rh; n = 1) is about 10 times more active than the corresponding longer chain members (20-Rh; n = 2-6). A mononuclear Rh species (21) is postulated, which would be unstable in the homogeneous systems, but substantially stabilized by the immobilization'-'.

20

21 (I. OJIMA)

1. F. DeCharentenay, J. A. Osborn, G. Wilkinson, J. Chem. Soc., A , 787 (1968). 2. E. Lukevics, Z. V. Belyakova, M. G. Pomerantseva, M. G. Voronkov, in Organometallic Chemistry Review, Journal of Organometallic Chemistry Library, Vol. 5, D. Seyferth, A. G. Davies, E. 0. Fischer, J. F. Normant, 0. A. Reutov, eds., Elsevier Scientific Publ. Co., Amsterdam, 1977, pp. 1 - 179. 3. C. Eabom, R. W. Bott, in The Bond to Carbon, Vol.I , Part I , A. G . MacDimid, ed., Dekker, New York, 1968, pp. 213-279. 4. P. Svoboda, P. Sedlmayer, J. HetflejS, Collect. Czech. Chem. Commun., 38, 1783 (1973).

14.4. Addition Reactions 14.4.2. Hydrosilylation of Olefins and Acetylenes 14.4.2.3. By Other Transition Metal Catalysts

229

~

5. I. Ojima, M. Kumagai, J. Organomet. Chem., 111,43 (1976). 6. A. J. Chalk, J . Organomet. Chem., 21, 207 (1970). 7. Z. V. Belyakova, S. A. Golubtsov, T. M. Yakusheva, Zh. Obshch. Khim., 32, 1997 (1962); J . Gen. Chem. USSR,32, 1978 (1962). a. S. Nozakura, S . Konotsune, Bull. Chem. SOC.Jpn., 29, 322 (1956). 9. K. Yamamoto, T. Hayashi, Y. Uramoto, R. Ito, M. Kumada, J. Organomet. Chem., 118, 331 (1976). 10. K. Yamamoto, T. Hayashi, M. Zembayashi, M. Kumada, J . Orgamomet. Chem., 118, 161 (1976). 11. I. Ojima, M. Kumagai, Y. Nagai, J . Organomet. Chem., 66, C14 (1974). 12. H. M. Dickers, R. N. Haszeldine, A. P. Mather, R. V. Parish, J . Organomet. Chem., 161, 91 (1978). 13. I. Ojima, N. Clos, R. J. Donovan, P. Ingallina, Organometallics,9, 3127 (1990). 14. J. E. Hill, T. A. Nile, J . Organomet. Chem., 137, 293 (1977). 15. M. F. Lappert, T. A. Nile, S . Takahashi, J . Organornet. Chem., 72,425 (1974). 16. K. Tamao, N. Miyake, Y. Kiso, M. Kumada, J . Am. Chem. Soc., 97,5603 (1975). 17. E. M. Michalska, B. Ostaszewski, J . Organometal. Chem., 299, 259 (1986). 18. Z. M. Michalska, J . Mol. Catal., 19, 345 (1983). 19. B. Marciniec, W. Urbaniak, J . Mol. Catal., 18,49 (1983). 20. Z. M. Michalska, M. Capka, J. Stoch, J . Mol. Catal., 11, 323 (1981). 21. B. Marciniec, W. Urbaniak, P. Pawlak, J. Mol. Catal., 14, 323 (1982). 22. B. Marciniec, Z. W. Kometka, and W. Uraniak, J . Mol. Catal., 12, 221 (1981).

14.4.2.3. By Other Transition Metal Catalysts

Although many transition metals and their salts other than platinum, rhodium, and nickel, such as Pd/C, Co/C, Ru/Al,O,, Cu, RuCl,, IrCl,, Fe, FeCl,, PdCl,, PdCldSi02, MoCl,, WCl,, OsO,, CrCl,, ZrCl,, Ti(OBu"),, and TiCl,, have been used as catalysts for the hydrosilylation of olefins, their catalytic activity is generally lower than that of chloroplatinic acid'. However, some transition metal complexes bring about synthetically useful reactions that cannot be promoted by the platinum catalyst. Dichloro ( (R)-NJV-dimethyl-l-[(S)-2-(diphenylphosphino)fe~ocenyl]ethylamine ] palladium(II), [(I?)-(S)-PPFAIPdCl,, is an efficient catalyst for the asymmetric hydrosilylation of styrene and norbornadiene with HSiCl,, which gives rise to (S)-a-phenylethyltrichlorosilane and (1R,2S,4S)-norbornyltrichlorosilane, respectively, in good yields,:

+

PhCH =CH2 HSiCl,

HSiCl,

[(R)-(S)-PPFAIPdCI,

Ph, Me'

*

CH -SiCl,

(a)

(52% e.e., >95% yield)

H

(53% e.e., 53% yield) The asymmetric hydrosilylation of 1-alkenes giving highly optically active 2-trichlorosilylalkanes is catalyzed by palladium complexes with chiral monodentate binaphthyl phosphine (MOP) ligands2v4. Palladium catalysts with chiral monodentate phosphine ligands, (S)-Zdiphenylphosphino)-2'-alkoxy-l, 1'-binaphthyl (MOPS), are extremely efficient for the asymmetric hydrosilylation of 1-alkenes with HSiC1,4. Reaction is carried out at 40°C for



229

~

5. I. Ojima, M. Kumagai, J. Organomet. Chem., 111,43 (1976). 6. A. J. Chalk, J . Organomet. Chem., 21, 207 (1970). 7. Z. V. Belyakova, S. A. Golubtsov, T. M. Yakusheva, Zh. Obshch. Khim., 32, 1997 (1962); J . Gen. Chem. USSR,32, 1978 (1962). a. S. Nozakura, S . Konotsune, Bull. Chem. SOC.Jpn., 29, 322 (1956). 9. K. Yamamoto, T. Hayashi, Y. Uramoto, R. Ito, M. Kumada, J. Organomet. Chem., 118, 331 (1976). 10. K. Yamamoto, T. Hayashi, M. Zembayashi, M. Kumada, J . Orgamomet. Chem., 118, 161 (1976). 11. I. Ojima, M. Kumagai, Y. Nagai, J . Organomet. Chem., 66, C14 (1974). 12. H. M. Dickers, R. N. Haszeldine, A. P. Mather, R. V. Parish, J . Organomet. Chem., 161, 91 (1978). 13. I. Ojima, N. Clos, R. J. Donovan, P. Ingallina, Organometallics,9, 3127 (1990). 14. J. E. Hill, T. A. Nile, J . Organomet. Chem., 137, 293 (1977). 15. M. F. Lappert, T. A. Nile, S . Takahashi, J . Organornet. Chem., 72,425 (1974). 16. K. Tamao, N. Miyake, Y. Kiso, M. Kumada, J . Am. Chem. Soc., 97,5603 (1975). 17. E. M. Michalska, B. Ostaszewski, J . Organometal. Chem., 299, 259 (1986). 18. Z. M. Michalska, J . Mol. Catal., 19, 345 (1983). 19. B. Marciniec, W. Urbaniak, J . Mol. Catal., 18,49 (1983). 20. Z. M. Michalska, M. Capka, J. Stoch, J . Mol. Catal., 11, 323 (1981). 21. B. Marciniec, W. Urbaniak, P. Pawlak, J. Mol. Catal., 14, 323 (1982). 22. B. Marciniec, Z. W. Kometka, and W. Uraniak, J . Mol. Catal., 12, 221 (1981).

14.4.2.3. By Other Transition Metal Catalysts

Although many transition metals and their salts other than platinum, rhodium, and nickel, such as Pd/C, Co/C, Ru/Al,O,, Cu, RuCl,, IrCl,, Fe, FeCl,, PdCl,, PdCldSi02, MoCl,, WCl,, OsO,, CrCl,, ZrCl,, Ti(OBu"),, and TiCl,, have been used as catalysts for the hydrosilylation of olefins, their catalytic activity is generally lower than that of chloroplatinic acid'. However, some transition metal complexes bring about synthetically useful reactions that cannot be promoted by the platinum catalyst. Dichloro ( (R)-NJV-dimethyl-l-[(S)-2-(diphenylphosphino)fe~ocenyl]ethylamine ] palladium(II), [(I?)-(S)-PPFAIPdCl,, is an efficient catalyst for the asymmetric hydrosilylation of styrene and norbornadiene with HSiCl,, which gives rise to (S)-a-phenylethyltrichlorosilane and (1R,2S,4S)-norbornyltrichlorosilane, respectively, in good yields,:

+

PhCH =CH2 HSiCl,

HSiCl,

[(R)-(S)-PPFAIPdCI,

Ph, Me'

*

CH -SiCl,

(a)

(52% e.e., >95% yield)

H

(53% e.e., 53% yield) The asymmetric hydrosilylation of 1-alkenes giving highly optically active 2-trichlorosilylalkanes is catalyzed by palladium complexes with chiral monodentate binaphthyl phosphine (MOP) ligands2v4. Palladium catalysts with chiral monodentate phosphine ligands, (S)-Zdiphenylphosphino)-2'-alkoxy-l, 1'-binaphthyl (MOPS), are extremely efficient for the asymmetric hydrosilylation of 1-alkenes with HSiC1,4. Reaction is carried out at 40°C for

230


24-72 h, giving the corresponding 2-trichlorosilylalkane with 91-97% e.e. as the predominant (8 1-94%) product accompanied by a small amount of achiral l-trichlorosilylalkane. The 2-trichlorosilylalkanes can be converted to the corresponding secondary alcohols with high enantiomeric purity.

+ HSiCl,

(cO.1 mol %)

[PdI/MOP

____)

X

I

c13si-R)

EtOH/Et,N

X = OMe, OPr', OCH,Ph, Et The reaction of olefins with hydrosilane and CO is a useful synthetic alternative to the hydroformylation of olefins5, which instead of the corresponding aldehyde, gives silyl enol ethers bearing one more carbon atom than the starting olefins5-,:

+

HSiEt,Me

+ co

[0 ]

c~~~co)B>

-OSiEt,Me

(49%)

+

HSiEt,Me

+

CO

CO,(CO),

140"c

OSiEt,Me (89%)

The catalysts for the reaction include Ru,(CO),,, Co,(CO),, Co,(CO),-PPh,, Co,(CO),-PBu;, Rh(PPh,),Cl, and Rh(PPh,),Cl-Et,N. In contrast, CuCl,, Cu,Cl,, WCO),, Mn2(CO)10, WCO),, Fe,(CO),, Fe(CO),-PPh,, Ru(PPh,),Cl,, RhzO,, Ir(PPh,),(CO)Cl, Ni(CO),, Ni(PPh,),(CO),, Ni(acac),, Pd(PPh,),Cl,, Pd(PPh,),, Pd-black, H,PtC16.6H,0, and Pt-black show little or no activity5. Among the catalysts, Co,(CO), is best?. The hydrosilane-CO combination also brings about unique reactions with cyclic ethers and aldehydes. The former reaction gives silyl ether of ~hydroxyaldehyde~:


0+ 0

0

0

c o (CO)

HSiEt,Me co ;60

+ HSiEt,Me

> MeEt,SiO

23 1

(53%)

(f)

OSiEt,Me c02(co)8 CO (60 atm)

0

(51%)

and the latter gives an cr-siloxyaldehydeJo:

-Lie

OSiEbMe

Co,(CO),-PPh,

(49%) 0 )

Acetates can be used instead of alkenes". Cyclic ketones undergo ring expansion to give a-hydroxyketone precursors". Oxiranes ring-open with siloxymethylation at ambient temperature and CO pressureJ3.

The silylcarbonylation reaction has been expanded to various systems. It provides a unique and effective method for organic syntheses14. Reactions of alkynes with hydrosilanes under CO (1-30 atm) catalyzed by Rh,(CO),, or Rh,Co,(CO),, give /3-formylvinylsilanes in high to excellent yieldsJ5-". This reaction, silylformylation, is also a silylcarbonylation, but it proceeds via a different mechanism from the Co,(CO),-catalyzed silylcarbonylation. In fact, Co,(CO), is inactive for silylf~rmylationl~. The reaction catalyzed by Co,Rh,(CO),, is proposed to involve the Co-Rh mixed metal dinuclear complexes as active catalyst species.I7

232


+

CO/cat.

R3SiH

Rl-=--.

RZ

ttt co

R'

RZ

>-(

+

Rg

(m)

OHC SiR, R3Si cat: Rh4(CO),,, RhzC0z(C0)2

CHO

co

Silylformylation has been applied to the synthesis of a-silylmethylene-P-lactones and a-silylmethylene-y-lactones'6.

The dehydrogenative silylation of olefins, closely related to hydrosilylation, is promoted by a ruthenium carbonyl complex, Ru,(CO),, ". The product vinylsilane is always the trans-isomer":

2 RCH-

CHz

+ HSiR;

Ru3(C0)lz 50-100°C

'

W S i R i + RCHZCH3 R

(p)

The reaction proceeds smoothly with styrene, p-substituted styrenes, and P-vinylnaphthalene to give the corresponding (E)-vinylsilanes in 83- 100% yield. However, isomerization of the double bond gives an allylsilane when a substrate having allylic proton@) is employed: PhCH,CH = CHz

+ HSiEt,

-

PhCHz

\A\ ~ i ~ t ,

Ru3(C0)lz, 50°C

+

ph

SiEt,

(77%)

(9)

(15%)

Trifluoromethylethene and pentafluorophenylethene are good substrates for the Ru,(CO),,-catalyzed reaction. When using HSiMe,Ph and HSiMe,Cl, the reaction gives

14.4. Addition Reactions 14.4.3. Hydrosilylation of Conjugated Dienes

233

P-trifluoro-methyl- or P-pentafluorophenylvinylsilaneas the main products (82- loo%), accompanied by a small amount of apparent hydrosilylation product”. Reaction is also catalyzed by RhCl(PPh), with trialkylsilanes, but selectivity is considerably lower than that of Ru,(CO) ,,19: R,CH =CH,

+ HSiR,

[Ru]or [Rh] Rf-

SiR3

and/or

The proposed mechanism for the dehydrogenative silylation includes dialkyl-M (M = Ru or Rh) species as the key inte~mediate’~’’~.

RrM$ic H H (I. OJIMA) 1. C. Eaborn, R. W. Bott, in The Bond to Carbon, Vol. 1, Part I, A. G. MacDiamid, ed., Dekker, New York, 1968, pp. 213-279. 2. K. Yamamoto, Y. Kiso, R. Ito, K. Tamao, M. Kumada, J . Organomet. Chem., 210, 9 (1981); Y. Kiso, K. Yamamoto, K. Tamao, M. Kumada, J . Am. Chem. Soc., 94,4373 (1972). 3. T. Hayashi, T. Tamao, Y. Katsuro, I. Nakae, M. Kumada, Tetrahedron Lett., 21, 1871 (1980). 4. Y. Uozumi, T. Hayashi, J. Am. Chem. SOC.,113,9887 (1991). 5. S. Murai, N. Sonoda, Angew. Chem., Int. Ed. Engl., 18, 837 (1979). 6. Y. Seki, A. Hidaka, S. Murai, N. Sonoda, Angew. Chem., Int. Ed. Engl., 16, 174 (1977). 7. Y. Seki, A. Hidaka, S. Makino, S. Murai, N. Sonoda, J. Organomet. Chem., 140,361 (1977). 8. Y. Seki, S. Murai, A. Hidaka, N. Sonoda, Angew. Chem., Int. Ed. Engl., 16, 881 (1977). 9. Y. Seki, S. Murai, I. Yamamoto, N. Sonoda, Angew. Chem., Int. Ed. Engl., 16, 789 (1977). 10. S. Murai, T. Kato, N. Sonoda, Y. Seki, K. Kawamoto, Angew. Chem., Int. Ed. Engl., 18, 393 (1979). 11. N. Chatani, S. Fujii, Y. Yamasaki, S. Murai, N. Sonoda, J. Am. Chem. Soc., 108, 7361 (1986). 12. N. Chatani, H. Furukawa, S. Kato, S. Murai, N. Sonoda, J . Am. Chem. SOC., 106,430 (1984). 13. T. Murai, S. Kato, S. Murai, T. Toki, S. Suzuki, N. Sonoda, J. Am. Chem. Soc., 106, 6093 (1984). 14. For other developments, see, e.g., (a) T. Murai, K. Furuta, S. Kato, S. Murai, N. Sonoda, J . Organometal. Chem., 302, 249 (1986). (b) T. Murai, S. Kato, S. Murai, Y. Hatayama, N. Sonoda, Tetrahedron Lett., 26, 2683 (1985). (c) T. Murai, Y. Hatayama, S. Murai, N. Sonoda, Organometallics, 2, 1883 (1983). (d) N. Chatani, Y. Yamasaki, S. Murai, N. Sonoda, Tetrahedron Lett., 24, 5649 (1983). (e) N. Chatani, s. Murai, N. Sonoda, J . Am. Chem. SOC.,105, 1370 (1983). 15. I. Matsuda, A. Ogiso, S. Sato, Y. Izumi, J. Am. Chem. SOC.,111, 2332 (1989). 16. I. Matsuda, A. Ogiso, S. Sato, J . Am. Chem. Soc., 112, 6120 (1990). 17. I. Ojima, P. Ingallina, R. J. Donovan, N. Clos, Organometallics, 10, 38 (1991). 18. Y. Seki, K. Kawamoto, S. Murai, N. Sonoda, J . Syn. Org. Chem. Jpn., 40, 501 (1982). 19. I. Ojima, T. Fuchikami, M. Yatabe, J . Organometal. Chem., 260, 335 (1984). 20. A. Millan, E. Towns, P. M. Maitlis, J . Chem. SOC.,Chem. Commun., 673 (1981).

14.4.3. Hydrosilylation of Conjugated Dienes Until the late 1960s, the investigations of hydrosilylation of dienes involved peroxides, UV irradiation, and platinum as the catalyst’. In 1969, a Ph,P complex of pal-


14.4. Addition Reactions 14.4.3. Hydrosilylation of Conjugated Dienes

233

P-trifluoro-methyl- or P-pentafluorophenylvinylsilaneas the main products (82- loo%), accompanied by a small amount of apparent hydrosilylation product”. Reaction is also catalyzed by RhCl(PPh), with trialkylsilanes, but selectivity is considerably lower than that of Ru,(CO) ,,19: R,CH =CH,

+ HSiR,

[Ru]or [Rh] Rf-

SiR3

and/or

The proposed mechanism for the dehydrogenative silylation includes dialkyl-M (M = Ru or Rh) species as the key inte~mediate’~’’~.

RrM$ic H H (I. OJIMA) 1. C. Eaborn, R. W. Bott, in The Bond to Carbon, Vol. 1, Part I, A. G. MacDiamid, ed., Dekker, New York, 1968, pp. 213-279. 2. K. Yamamoto, Y. Kiso, R. Ito, K. Tamao, M. Kumada, J . Organomet. Chem., 210, 9 (1981); Y. Kiso, K. Yamamoto, K. Tamao, M. Kumada, J . Am. Chem. Soc., 94,4373 (1972). 3. T. Hayashi, T. Tamao, Y. Katsuro, I. Nakae, M. Kumada, Tetrahedron Lett., 21, 1871 (1980). 4. Y. Uozumi, T. Hayashi, J. Am. Chem. SOC.,113,9887 (1991). 5. S. Murai, N. Sonoda, Angew. Chem., Int. Ed. Engl., 18, 837 (1979). 6. Y. Seki, A. Hidaka, S. Murai, N. Sonoda, Angew. Chem., Int. Ed. Engl., 16, 174 (1977). 7. Y. Seki, A. Hidaka, S. Makino, S. Murai, N. Sonoda, J. Organomet. Chem., 140,361 (1977). 8. Y. Seki, S. Murai, A. Hidaka, N. Sonoda, Angew. Chem., Int. Ed. Engl., 16, 881 (1977). 9. Y. Seki, S. Murai, I. Yamamoto, N. Sonoda, Angew. Chem., Int. Ed. Engl., 16, 789 (1977). 10. S. Murai, T. Kato, N. Sonoda, Y. Seki, K. Kawamoto, Angew. Chem., Int. Ed. Engl., 18, 393 (1979). 11. N. Chatani, S. Fujii, Y. Yamasaki, S. Murai, N. Sonoda, J. Am. Chem. Soc., 108, 7361 (1986). 12. N. Chatani, H. Furukawa, S. Kato, S. Murai, N. Sonoda, J . Am. Chem. SOC., 106,430 (1984). 13. T. Murai, S. Kato, S. Murai, T. Toki, S. Suzuki, N. Sonoda, J. Am. Chem. Soc., 106, 6093 (1984). 14. For other developments, see, e.g., (a) T. Murai, K. Furuta, S. Kato, S. Murai, N. Sonoda, J . Organometal. Chem., 302, 249 (1986). (b) T. Murai, S. Kato, S. Murai, Y. Hatayama, N. Sonoda, Tetrahedron Lett., 26, 2683 (1985). (c) T. Murai, Y. Hatayama, S. Murai, N. Sonoda, Organometallics, 2, 1883 (1983). (d) N. Chatani, Y. Yamasaki, S. Murai, N. Sonoda, Tetrahedron Lett., 24, 5649 (1983). (e) N. Chatani, s. Murai, N. Sonoda, J . Am. Chem. SOC.,105, 1370 (1983). 15. I. Matsuda, A. Ogiso, S. Sato, Y. Izumi, J. Am. Chem. SOC.,111, 2332 (1989). 16. I. Matsuda, A. Ogiso, S. Sato, J . Am. Chem. Soc., 112, 6120 (1990). 17. I. Ojima, P. Ingallina, R. J. Donovan, N. Clos, Organometallics, 10, 38 (1991). 18. Y. Seki, K. Kawamoto, S. Murai, N. Sonoda, J . Syn. Org. Chem. Jpn., 40, 501 (1982). 19. I. Ojima, T. Fuchikami, M. Yatabe, J . Organometal. Chem., 260, 335 (1984). 20. A. Millan, E. Towns, P. M. Maitlis, J . Chem. SOC.,Chem. Commun., 673 (1981).

14.4.3. Hydrosilylation of Conjugated Dienes Until the late 1960s, the investigations of hydrosilylation of dienes involved peroxides, UV irradiation, and platinum as the catalyst’. In 1969, a Ph,P complex of pal-

234

14.4. Addition Reactions 14.4.3. Hydrosilylation of Conjugated Dienes 14.4.3.1. By Platinum Catalysts

ladium was discovered to catalyze the addition of HSiMe, to butadiene giving octa-2,6dienyl-1-trimethylsilanenearly quantitatively', and the addition of HSiCl, to butadiene to give but-2-enyl-1-trichlorosilanein excellent yield3. Since then, the hydrosilylation of conjugated dienes such as butadiene and isoprene catalyzed by palladium, nickel, and rhodium has been extensively studied. The predominant mode is 1,4-addition (conjugate addition), in contrast to the results with platinum catalysts such as HzPtC1,.6H20 in which the 1,4- and the 1,Zselectivity depend markedly on the hydrosilane employed and a mixture of isomers is usually produced'. (I. OJIMA)

1. E. Lukevics, Z. V. Belyakova, M. G . Pomerantseva, M. G . Voronkov, in Organometallic Chemistry Review, Journal of Organometallic Chemistry Library, Vol.5 , D. Seyferth, A. G . Davies, E. 0. Fischer, J. F. Normant, 0. A. Reutov, eds., Elsevier Scientific Publ. Co., Amsterdam, 1977, pp. 1-179. 2. S . Takahashi, T. Shibano, N. Hagihara, J . Chem. SOC., Chem. Commun., 161 (1969). 3. M. Hara, K. Ohno, J. Tsuji, Chem. SOC.,Chem. Commun.,247 (1971).


The hydrosilylation of conjugated dienes such as butadiene and isoprene using H,RCl, as the catalyst proceeds only at elevated temperatures giving a mixture of products. The hydrosilylation of butadiene with HSiMeC1, affords a 1:1 mixture of monoadduct and diadduct (combined yield 50%), the former including but-2-enylsilane as major (85-90%) and but-3-enylsilane as minor products (15-10%)':

w SiMeC1z

-k -SiMeCl,

SiMeC1,

+ MeC1,Si

(a)

The hydrosilylation of myrcene with HSiCl, also affords a mixture of mono- and diadduct. Of the two possible monoadducts, the 1,2-adduct is favored2:

+ HSiCl,

H2PtC16*6H,O

u (40.6%)

+

SiCl,

SiC1,

I

(19.0%)

SiCl,

The hydrosilylation of isoprene with chlorohydrosilanes was claimed to give 2-methylbut-2-enylsilanese x c l u ~ i v e l yexcept ~ ~ ~ when using heptamethylcyclotetrasiloxane, which affords 2-methylbut-3-enylsilane(1,2-adduct) as the sole product5. However, a careful reinvestigation revealed that 3-methylbut-Zenylsilanes are the major product'.


234

14.4. Addition Reactions 14.4.3. Hydrosilylation of Conjugated Dienes 14.4.3.1. By Platinum Catalysts

ladium was discovered to catalyze the addition of HSiMe, to butadiene giving octa-2,6dienyl-1-trimethylsilanenearly quantitatively', and the addition of HSiCl, to butadiene to give but-2-enyl-1-trichlorosilanein excellent yield3. Since then, the hydrosilylation of conjugated dienes such as butadiene and isoprene catalyzed by palladium, nickel, and rhodium has been extensively studied. The predominant mode is 1,4-addition (conjugate addition), in contrast to the results with platinum catalysts such as HzPtC1,.6H20 in which the 1,4- and the 1,Zselectivity depend markedly on the hydrosilane employed and a mixture of isomers is usually produced'. (I. OJIMA)

1. E. Lukevics, Z. V. Belyakova, M. G . Pomerantseva, M. G . Voronkov, in Organometallic Chemistry Review, Journal of Organometallic Chemistry Library, Vol.5 , D. Seyferth, A. G . Davies, E. 0. Fischer, J. F. Normant, 0. A. Reutov, eds., Elsevier Scientific Publ. Co., Amsterdam, 1977, pp. 1-179. 2. S . Takahashi, T. Shibano, N. Hagihara, J . Chem. SOC., Chem. Commun., 161 (1969). 3. M. Hara, K. Ohno, J. Tsuji, Chem. SOC.,Chem. Commun.,247 (1971).


The hydrosilylation of conjugated dienes such as butadiene and isoprene using H,RCl, as the catalyst proceeds only at elevated temperatures giving a mixture of products. The hydrosilylation of butadiene with HSiMeC1, affords a 1:1 mixture of monoadduct and diadduct (combined yield 50%), the former including but-2-enylsilane as major (85-90%) and but-3-enylsilane as minor products (15-10%)':

w SiMeC1z

-k -SiMeCl,

SiMeC1,

+ MeC1,Si

(a)

The hydrosilylation of myrcene with HSiCl, also affords a mixture of mono- and diadduct. Of the two possible monoadducts, the 1,2-adduct is favored2:

+ HSiCl,

H2PtC16*6H,O

u (40.6%)

+

SiCl,

SiC1,

I

(19.0%)

SiCl,

The hydrosilylation of isoprene with chlorohydrosilanes was claimed to give 2-methylbut-2-enylsilanese x c l u ~ i v e l yexcept ~ ~ ~ when using heptamethylcyclotetrasiloxane, which affords 2-methylbut-3-enylsilane(1,2-adduct) as the sole product5. However, a careful reinvestigation revealed that 3-methylbut-Zenylsilanes are the major product'.

14.4. Addition Reactions 14.4.3. H drosilylation of Conjugated Dienes 14.4.3.2. b y Palladium and Nickel Catalysts

+ HSiMeCl, LSiMeC12 (72%)

H,PtCI,* 6H,O 165'C

+ LSiMeC12 ( HSiMeCl, >> HSi(OEt),. No reaction takes place with trialkylsilanes. Polymer-anchored catalysts are used for the hydrosilylation of butadiene, giving cisbut-Zenylsilanes in excellent yield7v8. Palladium complexes on inorganic supports are comparable or even more effective than their soluble analogues such as Pd(PhCN),Cl, and Pd(PPh,),Cl,. However, these catalysts are not stable to reuse. The hydrosilylation of cyclopentadiene with chlorohydrosilanes catalyzed by a phosphine-palladium complex gives 3-silylpent- 1-ene in high yield'.

+ HSiX,Y

Pd(PhCN),Cl,-PPh, 89-90°C

(h)

(80-90%)

Methylation with methyl Grignard reagent affords 3-trimethylsilylcyclopent- 1-ene, which is a versatile reagent for the regiospecific introduction of the cyclopentene skeleton in organic synthesis'. Asymmetric hydrosilylation of cyclopentadiene and cyclohexa-1,3-diene using chiral phosphine-palladium complexes gives optically active 3-trimethylsilylcyclohex1-ene after methylation":

0+

0

HSiCl,

+ HSiC1,

Pd(PhCN)2C12-P*

0 s i c ' 3

MeMgBr

(69-8 1%) Pd(PhCN),Cl,-P*

SiCl,

(56-64%)

SiMe,

P*: menthyldiphenylphosphine (MDPP) neomenthyldiphenylphosphine (NMDPP) When 1-aryl-1,3-butadienes were used for the asymmetric hydrosilylation with HSiCl, and PdCl,[(R)-(S)-PPFA] as the catalyst, higher asymmetric inductions were achieved (30-64% e.e.)".

238

14.4. Addition Reactions 14.4.3. Hydrosil lation of Conjugated Dienes 14.4.3.2. By PaIadiurn and Nickel Catalysts

Ph

HSiCI,, 8OoC PdCI,[(R)-(S)-PPFA]

62%

> -

EtOH NEt3

64% e.e. (S)

(96 :4)

30% e.e. ( R )

Nickel complexes also ex..ibit high catalytic activity toward the hydrosilylation of conjugated dienes under mild conditions. The reaction usually occurs by 1,Caddition, but the regio- and stereoselectivity is low compared with that achieved by palladium catalysts. The hydrosilylation of butadiene with HSiMe, catalyzed by nickel(0) complexes, such as Ni(PR3)2(C0)2,Ni(COD),, Ni(CO),, and Ni(CH,=CHCN)PPh,, gives a mixture of but-2-enyltrimethylsilaneand octa-2,6-dienyltrimethylsilanetogether with a considerable amount of cycl~octadiene'~~'~. Product ratio is affected by the donor ligand employed". Nickel(I1) species, such as Ni(v-C,H,),, NiC1, and Ni(acac), are inactive with Uialkylsilanes, but Ni(PPh,)( q-C,H,)-C,H, exhibits good catalytic activity4. In contrast, nickel(I1) salts with tertiary phosphines are good catalysts giving butenylsilanes and bissilylbut-Zene when trichlorosilane is employed although product selectivity is low',. Similarly, nickel(I1)-phosphine complexes, such as Ni(PPh,),Cl, and Ni(dmpf)Cl, [dmpf = 1,l '-bis(dimethylphosphino)ferrocene], catalyze the addition of HSi(OEt), and HSiMeCl, to penta- 1,3-diene15 and isopreneI6, respectively. The combination of nickel(I1) salts and reducing agents, e.g., Et,Al and aluminum hydrides, is an effective catalyst system. For instance, the hydrosilylation of butadiene with HSiMe, or H,SiMe, catalyzed by Ni(acac),-AlEt, gives cis-but-2-enylsilanes in 90-95% yield, or cis,cis-octa-2,6-dienylsilanes in 30-45% yield on using two equivalents of butadiene toward hydro~ilane'~. Triethoxysilane adds to cyclohexa- 1,3-diene giving 3-triethoxysilylcyclohex-1-ene in 97% yield;

0

+ HSi(OEt),

Ni(acac)z-AIEtz(OEt)

2O0C

,o S i ( O E t ) ,

(1)

(97%)

whereas triethylsilane does not react',. Phenyltrihydrosilane adds to isoprene by 1,4addition to give 2-methylb~t-2-enylsilane'~:

(78%) (I. OJIMA) 1. S. Takahashi, T. Shibano, H. Hagihara, J . Chem. SOC., Chem. Commun., 161 (1969). 2. M. Hara, K. Ohno, J. Tsuji, J . Chem. SOC.,Chem. Commun., 247 (1971). 3. Z. V. Belyakova, M. G. Pomerantseva, K. K. Popkov, L. A. Efremova, S. A. Golubtsov, Zh. Obshch. Khim., 42, 889 (1972); J . Gen. Chem. USSR, 42, 879 (1972). 4. J. Tsuji, M. Hara, K. Ohno, Tetrahedron, 30, 2143 (1974).

14.4. Addition Reactions 14.4.3. Hydrosilytationof Conjugated Dienes 14.4.3.3. By Rhodium, Cobalt, and Chromium Catalysts 5. 6. 7. 8. 9. 10. 11. 12.

239

I. Ojima, M. Kumagai, J. Organomet. Chem., 157, 359 (1978). V. Vaisarovfi, M. C$ka, J. HetflejS, Syn. Inorg. Met.-Org. Chem.,2, 289 (1972). V. Vaisarovh, J. HetflejS, G . Krause, H. Prackjus, Z. Chem., 14, 105 (1974). M. Cipka, J. HetflejS, Collect. Czech, Chem. Commun., 39, 154 (1974). I. Ojima, M. Kumagai, Y. Miyazawa, Tetrahedron Lett., 1385 (1977). Y. Kiso, K. Yamamoto, K. Tamao, M. Kumada, J . Am. Chem. SOC.,94, 4373 (1972). T. Hayashi, K. Kabeta, Tetrahedron Lett., 25, 3023 (1985). S. Takahashi, T. Shibano, H. Kojima, N. Hagihara, Organomet. Chem. Syn., I , 193

(1970/ 1971). M. Cspka, J. HetflejS, Collect. Czech. Chem. Commun., 40, 3020 (1975). M. Cipka, J. HetflejS, Collect. Czech. Chem. Commun., 40, 2073 (1975). A. J. Cornish, M. F. Lappert, T. A. Nile,J. Organomet. Chem., 132, 133 (1977). Y. Kiso, M. Kumada, K. Tamao, M. Umeno, J. Organomet. Chem., 50, 297 (1973). V. P. Yur’ev, I. M. Salimareeva, G . A. Tolstikov, 0. Zh. Zhebarov, Zh. Obshch. Khim., 45, 955 (1975); J . Gen. Chem. USSR,45,945 (1975). 18. V. P. Yur’ev, I. M. Salimareeva, 0. Zh. Zhebarov, G . A. Tolstikov, Izv. Akad. Nauk SSSR, Ser. Khim., 1888 (1975).

13. 14. 15. 16. 17.

14.4.3.3. By Rhodium, Cobalt, and Chromium Catalysts

The hydrosilylation of butadiene with HSiMe, using Rh(PPh,),Cl or Co2(CO), as catalyst proceeds smoothly at RT giving but-2-enylsilane (major) and but-3-enylsilane (minor) products’. The reaction catalyzed by Co(acac),-AlEt, was reported to afford but2-enylsilane (cis and trans-mixture) exclusively in 95% yield2. However, careful reinvestigation3 reveals that the rhodium(1) complex catalyzed reaction produces a mixture of three butenylsilanes and bis-silylbut-Zene, like the case of the nickel catalyzed reaction4. Similarly, HSiMe2Et adds to 2,3-dimethylbutadiene in the presence of Co2(CO), giving the 1,2- and 1,Cadducts and bis-~ilylbutene~. Selectivity of the reaction catalyzed by rhodium(1) complexes depends on the structure of the hydrosilane employed. The hydrosilylation of butadiene with HSi(OEt), catalyzed by a rhodium(1) complex, e.g., Rh(PPh,),Cl, Rh(PPh,),-(CO)Cl, Rh(PPh,),(CO)H, Rh(PPh,),( 77-C,H5), or Rh(PPh,),CH,, proceeds by 1,4-addition to give but-2-enyltrethoxysilane as sole product3:

+ HSi(OEt), Rh(PPh,),Br or Rh(PPh,),I is by far less effective. The hydrosilylation of isoprene using Rh(PPh,),Cl as catalyst gives 3-methyl-but2-enylsilane as the major product in addition to lesser amounts of (Z)-2-methyl-but-2enylsilane6:

+ HSiMe2Ph

Rh(PPh,),Cl 8ooc

SiMe2Ph + PhMe2Si (70.6%)

(27.4%)

The regioselectivity of the reaction is opposite to that observed in isoprene hydrosilylation catalyzed by other transition metal complexes. The product ratio depends on the structure of hydrosilane. The regioselectivity observed in the hydrosilylation of myrcene is the same as with isoprene6, but is much higher with ocimene. The reaction of ocimene with HSi(OEt), affords 2,6-dimethyl-8-triethoxysilylocta-2,6-diene in 98% regioselectivity6:


14.4. Addition Reactions 14.4.3. Hydrosilytationof Conjugated Dienes 14.4.3.3. By Rhodium, Cobalt, and Chromium Catalysts 5. 6. 7. 8. 9. 10. 11. 12.

239

I. Ojima, M. Kumagai, J. Organomet. Chem., 157, 359 (1978). V. Vaisarovfi, M. C$ka, J. HetflejS, Syn. Inorg. Met.-Org. Chem.,2, 289 (1972). V. Vaisarovh, J. HetflejS, G . Krause, H. Prackjus, Z. Chem., 14, 105 (1974). M. Cipka, J. HetflejS, Collect. Czech, Chem. Commun., 39, 154 (1974). I. Ojima, M. Kumagai, Y. Miyazawa, Tetrahedron Lett., 1385 (1977). Y. Kiso, K. Yamamoto, K. Tamao, M. Kumada, J . Am. Chem. SOC.,94, 4373 (1972). T. Hayashi, K. Kabeta, Tetrahedron Lett., 25, 3023 (1985). S. Takahashi, T. Shibano, H. Kojima, N. Hagihara, Organomet. Chem. Syn., I , 193

(1970/ 1971). M. Cspka, J. HetflejS, Collect. Czech. Chem. Commun., 40, 3020 (1975). M. Cipka, J. HetflejS, Collect. Czech. Chem. Commun., 40, 2073 (1975). A. J. Cornish, M. F. Lappert, T. A. Nile,J. Organomet. Chem., 132, 133 (1977). Y. Kiso, M. Kumada, K. Tamao, M. Umeno, J. Organomet. Chem., 50, 297 (1973). V. P. Yur’ev, I. M. Salimareeva, G . A. Tolstikov, 0. Zh. Zhebarov, Zh. Obshch. Khim., 45, 955 (1975); J . Gen. Chem. USSR,45,945 (1975). 18. V. P. Yur’ev, I. M. Salimareeva, 0. Zh. Zhebarov, G . A. Tolstikov, Izv. Akad. Nauk SSSR, Ser. Khim., 1888 (1975).

13. 14. 15. 16. 17.

14.4.3.3. By Rhodium, Cobalt, and Chromium Catalysts

The hydrosilylation of butadiene with HSiMe, using Rh(PPh,),Cl or Co2(CO), as catalyst proceeds smoothly at RT giving but-2-enylsilane (major) and but-3-enylsilane (minor) products’. The reaction catalyzed by Co(acac),-AlEt, was reported to afford but2-enylsilane (cis and trans-mixture) exclusively in 95% yield2. However, careful reinvestigation3 reveals that the rhodium(1) complex catalyzed reaction produces a mixture of three butenylsilanes and bis-silylbut-Zene, like the case of the nickel catalyzed reaction4. Similarly, HSiMe2Et adds to 2,3-dimethylbutadiene in the presence of Co2(CO), giving the 1,2- and 1,Cadducts and bis-~ilylbutene~. Selectivity of the reaction catalyzed by rhodium(1) complexes depends on the structure of the hydrosilane employed. The hydrosilylation of butadiene with HSi(OEt), catalyzed by a rhodium(1) complex, e.g., Rh(PPh,),Cl, Rh(PPh,),-(CO)Cl, Rh(PPh,),(CO)H, Rh(PPh,),( 77-C,H5), or Rh(PPh,),CH,, proceeds by 1,4-addition to give but-2-enyltrethoxysilane as sole product3:

+ HSi(OEt), Rh(PPh,),Br or Rh(PPh,),I is by far less effective. The hydrosilylation of isoprene using Rh(PPh,),Cl as catalyst gives 3-methyl-but2-enylsilane as the major product in addition to lesser amounts of (Z)-2-methyl-but-2enylsilane6:

+ HSiMe2Ph

Rh(PPh,),Cl 8ooc

SiMe2Ph + PhMe2Si (70.6%)

(27.4%)

The regioselectivity of the reaction is opposite to that observed in isoprene hydrosilylation catalyzed by other transition metal complexes. The product ratio depends on the structure of hydrosilane. The regioselectivity observed in the hydrosilylation of myrcene is the same as with isoprene6, but is much higher with ocimene. The reaction of ocimene with HSi(OEt), affords 2,6-dimethyl-8-triethoxysilylocta-2,6-diene in 98% regioselectivity6:

14.4. Addition Reactions 14.4.4 Hydrosilylationof Carbonyl Compounds

240

Si(EtO),

+

100°C

HSi(OEt),

(96%)

(2%)

Carbene complexes of rhodium, e.g., RhCl(COD):&-N(Ph)CH,CH,hPh

and

[(PhMe,P),Rh(CO):dl-N(Me)CH,CH,hMe] +Br-, catalyze the hydrosilylation of 2,3dimethylbutadiene giving 1,Cadduct as the major product together with the 1,2-adduct7. A rhodium(II1) complex, Rh(acac),, also catalyzes the hydrosilylation of 1,3-dienes giving 1,4-adduct predominantly8. The regioselectivity is opposite to that of Rh(PPh,),Cl, i.e., the same as that of palladium complexes, giving 2-methylbut-2-enyltriethoxysilane in 89% yield in the hydrosilylation of isoprene with HSi(OEt),. A chromium(0) complex, Cr(CO),, is an excellent photocatalyst (hv: 300-800 nm) for the hydrosilylation of 1,3-dienes'. The reaction proceeds smoothly at RT to give the 1,Cadducts exclusively in quantitative yield:

+ HSiK,

+ HSiMe,

hv

hv

Cr(C0)6

Me,Si (90%)

(10%) SiMe,

(f 1

(I. OJIMA)

1. S. Takahashi, T. Shibano, H. Kojima, H. Hagihara, Organomet. Chem. Syn.,I , 193 (1970/1971). 2. V. P. Yurlev, I. M. Salimareeva, G . A. Tolstikov, 0. Zh. Zhebarov, Zh. Obshch. Khim., 45,955 (1975); J . Gen. Chem. USSR,45,945 (1975). 3. J. Rejhon, J. HetflejS, Collect. Czech. Chem. Commun., 40, 3190 (1975). 4. M. Ciipka, J. HetflejS, Collect. Czech. Chem. Commun., 40, 2073 (1975). 5. A. J. Cornish, M. F. Lappert, T. A. Nile, J . Organomet. Chem., I36,73 (1977). 6. 1. Ojima, M. Kumagai, J . Organomet. Chem., 157, 359 (1978). 7. J. E. Hill, T. A. Nile, J . Organomet. Chem., 137, 293 (1977). 8. A. J. Cornish, M. F. Lappert, J . Organomet. Chem., 172, 153 (1979). 9. M. S. Wrighton, M. A. Schroeder, J . Am. Chem. Soc., 96,6235 (1974).

14.4.4 Hydrosilylationof Carbonyl Compounds The hydrosilylation of carbon-heteroatom multiple bonds had received little attention until it was found in 1972 that Rh(PPh,),Cl is an extremely effective catalyst for the hydrosilylation of carbonyl compounds'. This is a new and unique reduction method since the resulting silicon-oxygen bond can easily be hydrolyzed. Other transition metal complexes including platinum2, and r h ~ d i u m ~also - ~ have good catalytic activity in the selective and asymmetric hydrosilylation of carbonyl compounds8-' (I. OJIMA)


14.4. Addition Reactions 14.4.4 Hydrosilylationof Carbonyl Compounds

240

Si(EtO),

+

100°C

HSi(OEt),

(96%)

(2%)

Carbene complexes of rhodium, e.g., RhCl(COD):&-N(Ph)CH,CH,hPh

and

[(PhMe,P),Rh(CO):dl-N(Me)CH,CH,hMe] +Br-, catalyze the hydrosilylation of 2,3dimethylbutadiene giving 1,Cadduct as the major product together with the 1,2-adduct7. A rhodium(II1) complex, Rh(acac),, also catalyzes the hydrosilylation of 1,3-dienes giving 1,4-adduct predominantly8. The regioselectivity is opposite to that of Rh(PPh,),Cl, i.e., the same as that of palladium complexes, giving 2-methylbut-2-enyltriethoxysilane in 89% yield in the hydrosilylation of isoprene with HSi(OEt),. A chromium(0) complex, Cr(CO),, is an excellent photocatalyst (hv: 300-800 nm) for the hydrosilylation of 1,3-dienes'. The reaction proceeds smoothly at RT to give the 1,Cadducts exclusively in quantitative yield:

+ HSiK,

+ HSiMe,

hv

hv

Cr(C0)6

Me,Si (90%)

(10%) SiMe,

(f 1

(I. OJIMA)

1. S. Takahashi, T. Shibano, H. Kojima, H. Hagihara, Organomet. Chem. Syn.,I , 193 (1970/1971). 2. V. P. Yurlev, I. M. Salimareeva, G . A. Tolstikov, 0. Zh. Zhebarov, Zh. Obshch. Khim., 45,955 (1975); J . Gen. Chem. USSR,45,945 (1975). 3. J. Rejhon, J. HetflejS, Collect. Czech. Chem. Commun., 40, 3190 (1975). 4. M. Ciipka, J. HetflejS, Collect. Czech. Chem. Commun., 40, 2073 (1975). 5. A. J. Cornish, M. F. Lappert, T. A. Nile, J . Organomet. Chem., I36,73 (1977). 6. 1. Ojima, M. Kumagai, J . Organomet. Chem., 157, 359 (1978). 7. J. E. Hill, T. A. Nile, J . Organomet. Chem., 137, 293 (1977). 8. A. J. Cornish, M. F. Lappert, J . Organomet. Chem., 172, 153 (1979). 9. M. S. Wrighton, M. A. Schroeder, J . Am. Chem. Soc., 96,6235 (1974).

14.4.4 Hydrosilylationof Carbonyl Compounds The hydrosilylation of carbon-heteroatom multiple bonds had received little attention until it was found in 1972 that Rh(PPh,),Cl is an extremely effective catalyst for the hydrosilylation of carbonyl compounds'. This is a new and unique reduction method since the resulting silicon-oxygen bond can easily be hydrolyzed. Other transition metal complexes including platinum2, and r h ~ d i u m ~also - ~ have good catalytic activity in the selective and asymmetric hydrosilylation of carbonyl compounds8-' (I. OJIMA)

14.4. Addition Reactions 14.4.4 Hydrosilylationof Carbonyl Compounds 14.4.4.1. By Transition Metal Catalysts

241

1. I. Ojima, M. Nihonyanagi, Y. Nagai, J . Cham. Soc., Chem. Commun., 938 (1972); I. Ojima, T. Kogure, M. Nihonyanagi, Y. Nagai, Bull. Chem. SOC.Jpn., 45, 3506 (1972). 2. K. Yamamoto, T. Hayashi, M. Kumada, J. Organomet. Chem., 46, C65 (1972). 3. R. J. P. Corriu, J. J. E. Moreau, J . Chem. Soc., Chem. Commun.,38 (1973). 4. C. Eaborn, K. Odell, A. Pidcock, J . Organomet. Chem., 63,93 (1973). 5 . M. Bottrill, M. Green, J . Organomet. Chem., 111, C6 (1976). 6. T. E. Paxson, M. F. Hawthorne, J . Am. Chem. Soc., 96,4674 (1974). 7. L. I. Zakharkin, T. B. Agakhanova, Izv. Akad. Nauk SSSR, Ser. Khim., 2151 (1978). 8. I. Ojima, in Organorransition-Metal Chemistry, Y. Ishii, M. Tsutsui, eds., Plenum Press, New York, 1975, pp. 255-264. 9. H. B. Kagan, Pure Appl. Chem., 43,401 (1975). 10. I. Ojima, K. Yamamoto, M. Kumada, in Aspects of Homogeneous Catalysis, Vol. 3, R. Ugo, ed., D. Reidel, Dordrecht, Holland, 1977, pp. 189-228. 11. I. Ojima, in Fundamental Research in Homogeneous Catalysis, 11, Y. Ishii and M. Tsutsui, eds., Plenum Press, New York, 1978, pp. 181-206.

14.4.4.1. By Transition Metal Catalysts

The hydrosilylation of carbonyl compounds can be effected by various catalytic systems such as UV light, metal halides, e.g., NiCl,, GaCl,, InCl,, ZnCl,, KF, CsF, and H2PtC16*6H,0’: R’ \C=O+HSiX3 RZ’

cat. -+

R’

\ CH- OSiX,

R2’

Nickel metal catalysts give mixtures of the corresponding silyl ether and silyl enol ether. The former is produced via hydrosilylation, while the latter is produced via dehydrogenative silylation’. The reaction catalyzed by zinc chloride proceeds under drastic conditions, and the product of aldehydes disproportionate^^. The reaction of a$-unsaturated carbonyl compounds with H,PtC16 proceeds by 1,4-addition4, while coupling is also observed on using metallic Ni as catalyst’. The use of metal halide catalysts is restricted to mono-hydrosilanes because disproportionation of polyhydrosilanes is also promoted by these catalysts5. It has been shown that Rh(PPh,),Cl, (v3-C,H5)Rh[P(OMe),],, hydridorhodiumcarboranes, Ru(PPh3),C1,, and [Pt(PR,)C1,I2 can catalyze the hydrosilylation of ketones and aldehydes under mild conditions without serious side reactions. Among them, Rh(PPh,),Cl is the most effective. Dihydrosilanes and trihydrosilanes are more reactive than monohydrosilanes. Typical results are shown in Table 1. Although Rh(PPh,),Cl is an excellent catalyst for the decarbonylation of aldehydes, and thus olefinic substrates with aldehyde moieties cannot be used for the hydrogenation, the hydrosilylation of aldehydes using the same catalyst proceeds smoothly without such a side reaction6. The active catalyst species is Rh(PPh,),Cl, and its oxidative addition adduct, Rh(PPh3),(SiEt3)(H)C1, is isolated6. Similarly, the active species for Ru(PPh,),Cl, and ( v3-C3H5)Rh[P(OMe),], are Ru(PPh,),(H)CI4 and (Rh[P(OMe),],H},, respectively. The main side reaction of hydrosilylation catalyzed by transition metal complexes is dehydrogenative silylation: >CH-C-

II

0

+ HSiX, %

>CH-CH-+>C=C-

I

OSiX,

I

+ H,

OSiX,

(b)


14.4. Addition Reactions 14.4.4 Hydrosilylationof Carbonyl Compounds 14.4.4.1. By Transition Metal Catalysts

241

1. I. Ojima, M. Nihonyanagi, Y. Nagai, J . Cham. Soc., Chem. Commun., 938 (1972); I. Ojima, T. Kogure, M. Nihonyanagi, Y. Nagai, Bull. Chem. SOC.Jpn., 45, 3506 (1972). 2. K. Yamamoto, T. Hayashi, M. Kumada, J. Organomet. Chem., 46, C65 (1972). 3. R. J. P. Corriu, J. J. E. Moreau, J . Chem. Soc., Chem. Commun.,38 (1973). 4. C. Eaborn, K. Odell, A. Pidcock, J . Organomet. Chem., 63,93 (1973). 5 . M. Bottrill, M. Green, J . Organomet. Chem., 111, C6 (1976). 6. T. E. Paxson, M. F. Hawthorne, J . Am. Chem. Soc., 96,4674 (1974). 7. L. I. Zakharkin, T. B. Agakhanova, Izv. Akad. Nauk SSSR, Ser. Khim., 2151 (1978). 8. I. Ojima, in Organorransition-Metal Chemistry, Y. Ishii, M. Tsutsui, eds., Plenum Press, New York, 1975, pp. 255-264. 9. H. B. Kagan, Pure Appl. Chem., 43,401 (1975). 10. I. Ojima, K. Yamamoto, M. Kumada, in Aspects of Homogeneous Catalysis, Vol. 3, R. Ugo, ed., D. Reidel, Dordrecht, Holland, 1977, pp. 189-228. 11. I. Ojima, in Fundamental Research in Homogeneous Catalysis, 11, Y. Ishii and M. Tsutsui, eds., Plenum Press, New York, 1978, pp. 181-206.

14.4.4.1. By Transition Metal Catalysts

The hydrosilylation of carbonyl compounds can be effected by various catalytic systems such as UV light, metal halides, e.g., NiCl,, GaCl,, InCl,, ZnCl,, KF, CsF, and H2PtC16*6H,0’: R’ \C=O+HSiX3 RZ’

cat. -+

R’

\ CH- OSiX,

R2’

Nickel metal catalysts give mixtures of the corresponding silyl ether and silyl enol ether. The former is produced via hydrosilylation, while the latter is produced via dehydrogenative silylation’. The reaction catalyzed by zinc chloride proceeds under drastic conditions, and the product of aldehydes disproportionate^^. The reaction of a$-unsaturated carbonyl compounds with H,PtC16 proceeds by 1,4-addition4, while coupling is also observed on using metallic Ni as catalyst’. The use of metal halide catalysts is restricted to mono-hydrosilanes because disproportionation of polyhydrosilanes is also promoted by these catalysts5. It has been shown that Rh(PPh,),Cl, (v3-C,H5)Rh[P(OMe),],, hydridorhodiumcarboranes, Ru(PPh3),C1,, and [Pt(PR,)C1,I2 can catalyze the hydrosilylation of ketones and aldehydes under mild conditions without serious side reactions. Among them, Rh(PPh,),Cl is the most effective. Dihydrosilanes and trihydrosilanes are more reactive than monohydrosilanes. Typical results are shown in Table 1. Although Rh(PPh,),Cl is an excellent catalyst for the decarbonylation of aldehydes, and thus olefinic substrates with aldehyde moieties cannot be used for the hydrogenation, the hydrosilylation of aldehydes using the same catalyst proceeds smoothly without such a side reaction6. The active catalyst species is Rh(PPh,),Cl, and its oxidative addition adduct, Rh(PPh3),(SiEt3)(H)C1, is isolated6. Similarly, the active species for Ru(PPh,),Cl, and ( v3-C3H5)Rh[P(OMe),], are Ru(PPh,),(H)CI4 and (Rh[P(OMe),],H},, respectively. The main side reaction of hydrosilylation catalyzed by transition metal complexes is dehydrogenative silylation: >CH-C-

II

0

+ HSiX, %

>CH-CH-+>C=C-

I

OSiX,

I

+ H,

OSiX,

(b)

242

14.4. Addition Reactions 14.4.4 Hydrosilylation of Carbonyl Compounds 14.4.4.1. By Transition Metal Catalysts

TABLE1. HYDROSILYLATION OF TYPICAL KETONES AND ALDEHYDES CATALYZED BY Rh(PPh,),CI

~

Carbonyl Compound

Hydrosilane

PhCOMe MeCOMe Cyclohexanone PhCHO PhCHO PhCOBu' MeCOMe PhCOPh "WHO 'BuCOMe Cyclohexanone Cyclohexanone(2) Cyclohexanone(3) MeCOCN PhCOCOPh( 1) MeCOCOMe

HSiEt, HSiEt, HSiEt, HSiEt, HSiMe,Ph HSiMe,Et HSiPh, H,SiEtz H,SiMePh H2SiPh, H,SiPh, H,SiPh( 1) H,SiPh( 1) HSiEt, HSiEt3(2) HSiEt,

~~

Product

Yield(%)

Reference

PhMeCHOSiEt, Me,CHOSiEt, c-C,H, ,OSiEt, PhCH,OSiEt, PhCH,OSiMe,Ph Ph('Bu)CHOSiMe,Et Me,CHOSiPh, Ph,CHOSiHEt, "BuOSiHMePh "Bu(Me)CHOSiHPh, C-CsH, ,OSiHPh, (c-C,H, ,O),SiHPh (c-C,H, ,O),SiPh Me(CN)CHOSiEt, [Ph(Et,SiO)CH-1, MeCH(OSiEt,)COMe

97 95 98 95 94 96 98 95 98 97 98 98 98 88 83 85

6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

Although Rh(PPh,),Cl exhibits high selectivity for the hydrosilylation, dehydrogenative silylation occurs on using ketones bearing electron-withdrawing groups on the a-carbon, such as methyl acetoacetate and benzoylacetonitrile. With acetylacetone exclusive dehydrogenative silylation takes place6: RCOCH,X

X

+ HSiEt,

Rh(PPh,),CI

= CN, COMe, COOMe

' RCHCHZX + RC =CHX I

OSiEt,

I

OSiEt,

(c)

When benzoyl cyanide is employed, coupling predominates over hydrosilylation 1,2-bis(triethylsiloxy)ethane in 50% yield. Howwhich gives 1,2-dicyano-l,2-diphenylever, normal hydrosilylation proceeds on using PdClz as catalyst6. (I. OJIMA)

1. E. Lukevics, Z. V. Belyakova, M. G.Pomerantseva, M. G . Voronkov, in Organometallic Chemistry Review, Journal of Organometallic Chemistry Library, Vol.5, D. Seyferth, A. G. Davies, E. 0. Fischer, J. F. Normant, 0. A. Reutov, eds., Elsevier Scientific Publ. Co., Amsterdam, 1977, pp. 1-179. 2. E. Frainnet, Pure Appl. Chem., 19,489 (1965). 3. For example, R. Calas, E. Frainnet, J. Bonastre, C . R. Hebd. Seances, Acad. Sci., 251, 2987 (1960). 4. S. I. Sadykh-Zade, A. D. Petrov, Zh. Obshch. Khim., 29,3194 (1959); J . Gen. Chem. USSR, 29, 3159 (1959); A. D. Petrov, S . I. Sadykh-Zade, Dok. Akad. Nauk SSSR, 121, 119 (1958). 5 . M. Gilman, D. H. Miles, J . Org. Chem., 23, 326 (1958). 6. I. Ojima, M. Nihonyanagi, T. Kogure, M. Kumagai, S. Horiuchi, K. Nakatsugawa,J. Organomet. Chem., 94,449 (1975).


14.4. Addition Reactions 14.4.4 Hydrosilylation of Carbonyl Compounds 14.4.4.2. Stereoselective and Chemoselective Reactions

243

14.4.4.2. Stereoselective and Chemoselective Reactions

Stereoselective hydrosilylation of terpene ketones such as camphor and menthone catalyzed by Rh(PPh,),Cl followed by hydrolysis produces different stereochemistry from reductions using metal hydrides'.,. Bulky hydrosilanes favor the production of the more stable alcohols:

HSiR,

hOsiR3 6 +

Rh(PPh3)3C1,

H

H

OSiR,

Isobomeol

Bomeol Bomeol (%)

Isobomeol (%)

H,SiEt, H,SiPh, HSiEt,

(a)

9

91

13

21

70

30

This is unusual since in the reduction of monocyclic and bicyclic ketones, bulkier hydroboranes produce the less stable of the two possible alcohols in accordance with steric approach control3. This suggests that the transition state for the catalytic hydrosilylation cannot be accommodated by a simple four-centered process. The mechanism must take into account the intermediacy of an a-siloxyalkylrhodium complex, which forms according to the soft-hard concept and must be characteristic of the ketone hydrosilylation. Studies on the stereoselective hydrosilylation of substituted cyclohexanones were performed4s5.The trend observed in the camphor case was confirmed in the hydrosilylation of 4-t-butylcyclohexanone with RhCl(PPh,), and RuCl,(PPh,),, viz., the bulkier the hydrosilane, the larger the amount of more stable equatorial isomer (eq) relative to the axial isomer (ax): Stereoselectivities (ax/eq) of 11/89 with RhCl(PPh,), and 5/95 with RuCl,(PPh,), were achieved using HSiEt,4". OSiEt,

'H&

[Rh] HSiEt3 or [Ru]

+ &OSiEt3

H (b)

Equilibrium is proposed between the two a-siloxyalkyl-Rh(II1) intermediates via silyl enol ether and dihydrido-Rh(II1) ~ o m p l e xFor . ~ other substituted cyclohexanones, the influence of the hydrosilanes' bulk on the stereoselectivity is observed only for sterically hindered 2-substituted cyclohe~anones~~. High regio- and stereo-controlled synthesis of 1,3-diols is effected through the intramolecular hydrosilylation of chiral allylic and homoallylic alcohols followed by oxidative cleavage of the carbon-silicon bond. This method consists of (1) the silylation of

244


the hydroxyl group of a chiral allylic or homoallylic alcohol with (HSiMe,),NH followed by intramolecular hydrosilylation catalyzed by H,PtCl, or RhCl(PPh,), and (2) the oxidative cleavage of the resulting cyclic silyl ether with hydrogen peroxide and a base'. OH

OH

I

base

r\h

OH

OH

I

H,O,

(1) (HSiMe,),NH (2) RhCl(PPh,),

'O

O H >99% d.e.

(a)

(4

+EL(++

Me$ -0

OSiMe,Bu'

OSiMe,Bu'

2,3-erythro 10/1 3,4-erythro 13/1 Highly stereoselective hydrosilane reduction of a-methyl-P-keto amides has been developed using HSiMe,Ph/TASF [TASF = tris(dimethy1amino)sulfonium difluorotrimethylsilicate] and HSiMe,Ph/CF,COOH combinations6. The HSiMe,Ph/TASF combination gives threo product with >98% purity when the keto amide has a P-aryl substituent, while the HSiMe,Ph/CF,COOH combination gives erythro product with >98% purity in most cases.

Ph

NMe,

NMe,

HSiMe,Ph TASF

wc: . _

'

CF,COOH HSiMe,Ph

ooc

Ph V N M e , threo:erythro = >99: 1

'

#NMe,

(f)

threo:erythro = 1 :99

The hydrosilylation of a,P-unsaturated ketones and aldehydes using monohydrosilanes catalyzed by Rh(PPh,),Cl proceeds by 1,Caddition, while dihydrosilanes or trihydrosilanes undergo 1,2-addition to carbonyl f u n c t i o n a l i t i e ~ ~ * ~ ~ ' ~ .

14.4. Addition Reactions 14.4.4 Hydrosilylationof Carbonyl Compounds 14.4.4.2. Stereoselectiveand Chemoselective Reactions

--c HSiR,

R' R2C= C(R3)- C -R4

II

0

245

C -R4

R' R2CH-C(R3)=

I

Rh(PPh ) C1

H,SiR,

OSiRJ

R1R2C=C(R3)-CH- R4

(€9

I

OSMR,

Typical results are shown in Table 1. Since the resulting silyl enol ethers and allylic silyl ethers can readily be converted by hydrolysis to saturated carbonyl compounds and allylic alcohols, respectively, these

TABLE1. HYDROSILYLATION OF (Y,P-UNSATURATEDCARBONYL COMPOUNDS CATALYZED BY Rh(PPh,),Cl Carbonyl Compound Me,C=CHCOMe Me,C=CHCOMe Me,C=CHCOMe MeCH=CHCOPh MeCH=CHCHO MeCH=CHCHO PhCH=CHCHO

Hydrosilane

Yield(%)

Reference

94 98 95 90 95 91 96

16 16 13 16 16 16 16

HSiEt,

92

16

H,SiEt,

100

13

100

13

OSiMe,Et

90

13

H,SiPh,

ff OSiHPh2

97

13

HSiMe,Et

Me,C=C(OEt)OSiMe;?Et CH,=C( Me)CH(OEt)OSiMqEt

85.5 4.5

15

HSiEt, HSiMe,Ph H,SiPh HSiEt, HSiEt, HSiPh, HSiEt,

H,SiPh,

HSiMe,Et

CH,=C(Me)CO,Et

Product(s) Me,CHCH=C(Me)OSiEt, Me,CHCH=C(Me)OSiMe;?Ph Me,C=CHCH(Me)OSiH,Ph EtCH=C(Ph)OSiEt, EtCH=CHOSiEt, EtCH=CHOSiPh, PhCH,CH=CHOSiEt,

v

@

OSiHPh,

Y

246


reactions furnish a unique method for the selective reduction of a,P-unsaturated carbonyl compounds'~*.a$-Unsaturated carbonyl compounds containing an isolated double bond also undergo selective reduction without any isomerization nor reduction at the isolated double bond'.' e.g.:

(97%) 0°C

MeOH

(97%)

It is known that (1) the selectivity of 1,Zaddition using trihydrosilanes is higher than that using dihydrosilanes, and (2) the selectivity of dihydrosilane addition to the carbonyl group is increased by the use of a nonpolar solvent and by increasing the dihydrosilane concentration'. The limits to the applicability of this selective reduction method are shown by the hydrosilylation of PhCH=CHCOPh or 1,3-diphenylprop-2-en-1-one, benzylideneacetophenone, or cinnamoylbenzene, which proceeds only by 1,6addition even when dihydrosilanes and trihydrosilanes are e m p l ~ y e d ' ~ However, ~'~. selective 1,2-addition is achieved by using CsF as catalyst": H,SiPhz

PhCH=CHCOPh CsF, 250c: Hzo

9

PhCH=CHCH(OH)Ph (95% yield, 95% selectivity)

(j)

+

Nickel(0) complexes such as Ni(COD),, 2PPh3 Ni(COD),, and Ni[P(OEt),], catalyze the selective 1 ,Zaddition of the monohydrosilanes, triethoxysilane and trimethoxysilane to crotonaldehyde12: CH,CH=CHCHO

+ HSi(OEt),

Ni(CoD)z 20"c

CH,CH=CHCH,OSi(OEt), (80%)

(k)

However, trialkylsilane gives a mixture of 1,Cadduct (major) and 1,Zadduct (minor) using the same catalyst. The deuteriosilane-Rh(PPh,),Cl combination can serve as convenient reagent for regiospecific deuteration, which leads either to the 1-d-allylic alcohol or to the 3-d-ketone or aldehyde',:

14.4. Addition Reactions 14.4.4 Hydrosilylation of Carbonyl Compounds 14.4.4.3. Asymmetric Synthesis

247

(1)

OSiDPh,

OH-

OH

D

D

The hydrosilylations of a,P-unsaturated esters such as acrylates or methacrylates catalyzed by H,PtC16.6H,0" or Rh(PPh3)3C1'4*'5are complicated, i.e., the selectivity of the reaction is markedly dependent on both the substituents of the substrate and the structure of hydrosilane. Nevertheless, high selectivities are achieved by proper choice of reactant^'^. (I. OJIMA) 1. I. Ojima, M. Nihonyanagi, Y. Nagai, Bull. Chem. SOC.Jpn., 45, 3722 (1972). 2. I. Ojima, in Chemistry of Organosilicon Compounds, Ch. 25, S . Patai, Z. Rappaport, ed., Wiley, New York, 1989, p. 1479. 3. H. C. Brown, V. Barma, J . Am. Chem. Soc., 88,2871 (1966). 4. M. F. Semmelhack, R. N. Misra, J . Org. Chem., 47, 2469 (1982). (b) J. Ishiyama, Y. Senda, I. Shinoda, S. Imaizumi, Bull. Chem. SOC.Jpn., 52, 2353 (1979). 5. K. Tamao, T. Nakajima, R. Sumiya, H. Arai, N. Higuchi, Y. Ito,J. Am. Chem. SOC., 108,6090 (1986). 6. M. Fujita, T. Hiyama, J . Am. Chem. SOC.,107, 8294 (1985). 7. I. Ojima, T. Kogure, Y. Nagai, Tetrahedron Lett., 5035 (1972). 8 . M. Fieser, L. F. Fieser, eds., Reagents for Organic Synthesis, Vol. 4, Wiley, New York, 1974, p. 562. 9. V. Z. Sharf, L. Kh. Freidlin, I. S. Shekoyan, V. N. Krutii, Izv. Akad. Nauk SSSR, Ser. Khim., 1087 (1977). 10. T. Kogure, Ph.D. Thesis, Tohoku University, 1979. 11. J. Boyer, R. J. P. Corriu, R. Perz, C. Reye, J . Organomet. Chem., 172, 143 (1979). 12. M. F. Lappert, T. A. Nile, J . Organomet. Chem., 102,543 (1975). 13. I. Ojima, T. Kogure, Organometallics, 1, 1390 (1982). 14. E. Yoshii, Y. Kobayashi, T. Koizumi, T. Oribe, Chem. Pharm. Bull., 22, 2767 (1974). 15. I. Ojima, M. Kumagai, J . Organomet. Chem., 111,43 (1976). 16. I. Ojima, M. Nihonyanagi, T. Kogure, M. Kumagai, S. Horiuchi, K. Nakatsugawa, J . Organomet. Chem., 94,449 (1975).

14.4.4.3. Asymmetric Synthesis

Extensive studies of the asymmetric hydrosilylation of prochiral ketones have been performed using rhodium(1) complexes with chiral phosphine ligandsl and nitrogen ligand~'-~,although the first studies of asymmetric hydrosilylation of ketones used chiral phosphine-platinum complexes6*'.

R'

'C=O / R2

+ HSiR,

cat?

R' * 'CH-OSiR,

/

RZ

H+or OH-

R' %H-OH / R2

(a)



247

(1)

OSiDPh,

OH-

OH

D

D

The hydrosilylations of a,P-unsaturated esters such as acrylates or methacrylates catalyzed by H,PtC16.6H,0" or Rh(PPh3)3C1'4*'5are complicated, i.e., the selectivity of the reaction is markedly dependent on both the substituents of the substrate and the structure of hydrosilane. Nevertheless, high selectivities are achieved by proper choice of reactant^'^. (I. OJIMA) 1. I. Ojima, M. Nihonyanagi, Y. Nagai, Bull. Chem. SOC.Jpn., 45, 3722 (1972). 2. I. Ojima, in Chemistry of Organosilicon Compounds, Ch. 25, S . Patai, Z. Rappaport, ed., Wiley, New York, 1989, p. 1479. 3. H. C. Brown, V. Barma, J . Am. Chem. Soc., 88,2871 (1966). 4. M. F. Semmelhack, R. N. Misra, J . Org. Chem., 47, 2469 (1982). (b) J. Ishiyama, Y. Senda, I. Shinoda, S. Imaizumi, Bull. Chem. SOC.Jpn., 52, 2353 (1979). 5. K. Tamao, T. Nakajima, R. Sumiya, H. Arai, N. Higuchi, Y. Ito,J. Am. Chem. SOC., 108,6090 (1986). 6. M. Fujita, T. Hiyama, J . Am. Chem. SOC.,107, 8294 (1985). 7. I. Ojima, T. Kogure, Y. Nagai, Tetrahedron Lett., 5035 (1972). 8 . M. Fieser, L. F. Fieser, eds., Reagents for Organic Synthesis, Vol. 4, Wiley, New York, 1974, p. 562. 9. V. Z. Sharf, L. Kh. Freidlin, I. S. Shekoyan, V. N. Krutii, Izv. Akad. Nauk SSSR, Ser. Khim., 1087 (1977). 10. T. Kogure, Ph.D. Thesis, Tohoku University, 1979. 11. J. Boyer, R. J. P. Corriu, R. Perz, C. Reye, J . Organomet. Chem., 172, 143 (1979). 12. M. F. Lappert, T. A. Nile, J . Organomet. Chem., 102,543 (1975). 13. I. Ojima, T. Kogure, Organometallics, 1, 1390 (1982). 14. E. Yoshii, Y. Kobayashi, T. Koizumi, T. Oribe, Chem. Pharm. Bull., 22, 2767 (1974). 15. I. Ojima, M. Kumagai, J . Organomet. Chem., 111,43 (1976). 16. I. Ojima, M. Nihonyanagi, T. Kogure, M. Kumagai, S. Horiuchi, K. Nakatsugawa, J . Organomet. Chem., 94,449 (1975).

14.4.4.3. Asymmetric Synthesis

Extensive studies of the asymmetric hydrosilylation of prochiral ketones have been performed using rhodium(1) complexes with chiral phosphine ligandsl and nitrogen ligand~'-~,although the first studies of asymmetric hydrosilylation of ketones used chiral phosphine-platinum complexes6*'.

R'

'C=O / R2

+ HSiR,

cat?

R' * 'CH-OSiR,

/

RZ

H+or OH-

R' %H-OH / R2

(a)

248


The catalysts used in asymmetric hydrosilylation include rhodium(1) complexes of type L*,Rh(S)Cl, prepared in situ from [Rh(olefin),Cl], and chiral p h ~ s p h i n e s ~ or -'~ chiral nitrogen ligand~'-~,cationic rhodium complexes of type [L*,RhH,S,] +C104- l 5 (S = solvent molecule), and rhodium complexes attached to organic9J4J5or polymer supports. Rhodium(II1) complexes, L*,RhCl,, were prepared from the reaction of chiral bis(oxazoliny1)pyridineligands (pybox) with RhC1,.3H,05. Many chiral ligands have been developed18 for the asymmetric hydrogenation of prochiral olefins, which are also effective for the asymmetric hydrosilylation of prochiral ketones. Among those ligands, benzylmethylphenylphosphine (BMPP)19 and 2,3-O-isopropylidene-2,3-dihydroxy-l,4-bis(diphenylphosphino)butane (DIOP, 1)'' are most popular. The best results (97.6% e.e.) for the asymmetric reduction of acetophenone have been achieved with 2-(2-pyridyl)-4-carbethoxy-1,3-thiazolidine (2)3and the highest enantioselectivity (99% e.e.) was achieved in the reaction of 1-tetralone catalyzed by a rhodium(II1) complex with 2,6-bis[4'-(S)-isopropyloxazolin-2'-yl]pyridine[(S,S)-ip-pybox, 315.

1

2

3

It has been shown in the asymmetric hydrosilylation catalyzed by rhodium complexes that both configuration and optical yield of the alcohol products depend on the structure of hydrosilanes used. Typical results are shown in Table 1. Copper(1) complexes with chiral diphosphine ligands were used as catalysts for the asymmetric reduction of acetophenone; optical induction was in the range 10-40% e.e.'l. A chiral hydrosilane, pinanylmethylsilane, was employed for the asymmetric hydrosilylation of 2-octanone catalyzed by RhCl(PPh,),. The reaction followed by hydrolysis gave 2-octanol with 25.7% e.e.', The DIOP-rhodium(1) complex attached to organic polymers9915,e.g., polystyrene resin and poly(methy1 vinyl alcohol), exhibits good catalytic activity as a chiral catalyst comparable to the corresponding homogeneous catalyst. In contrast, the rhodium(1) complexes anchored on inorganic supports display only a low effi~iency'~,'~. Studies show that the steric requirements for a match of the chiral ligand, a hydrosilane and a ketone are of definite importance in bringing about effective asymmetric induction. The a-siloxyalkylrhodium complex plays a key role in the catalytic cycle for the asymmetric hydrosilylation of prochiral The intermediacy is suggested in the hydrosilylation of cyclic terpene ketones catalyzed by Rh(PPh,),Cl%, supported by spin trapping experiments in which spin adducts such as 425*26are detected27.

'SiHPhNp" 4

H,SiPh,

H,SiPh,

(ip-pybox)-Rh"'

(ip-pybox)--Rh"'

MeCH(0H)hex-n

OH

I

PhCH(0H)Me PhCH(0H)Me PhCH(0H)Me PhCH(0H)Me PhCH(0H)Et PhCH(0H)Et PhCH(0H)ff PhCH(0H)Pf PhCH(0H)Bu' PhCH(0H)Bu' PhCH(0H)Bu' PhCH(OH)CH,NMe, PhCH(OH)(CH,),NMe, 'BuCH(OH)CH,OCOPh PhCH(OH)CH,Cl p-MeOC,H,CH(OH)Ph

Product

Optical Yieldb(%)

*L,,-RhN, L,Rh(S)CI; L,-Rh+, [L,,F&H,S,]+ClO;. Substrate/rhodium = 200-2000. bValues in the parentheses are the optical yields on using polymer-supported rhodium catalyst. R or S in parentheses stands for the configuration of the product. W F A , (R)- ~~~meth yl-1 -[(s)-2-(di me thyl pho sp h~o)fe ~e nyl]e thyl~in~.

MeCO-hex-n

0

p

,

+

(MPFA),-Rh" (2F--RhN [( - )BMPP],-RhN ( +)DIOP-R~~ [( - )BMPPIz-RhN [( - )BMPP],-RhN [( + )BMPP],-RhN [( )BMPP],-RhN [( + )BMPP],-Rh+ ( - )DIOP-R~~ ( - )DIOP-R~~ ( - )DIOP-RhN ( - )DIOP-RhN ( - )DIOP-RhN (ip-pybox)--Rh'"

+ )DIOP-RhN

[( - )BMPP],-RhN

HSiMqPh H,SiPhNp" HzSiPhNpa H,SiPh, HSiMqPh HzSiPhNpu H,SiEt, HSiMe,Ph HSiMe,Et HSiM%Ph HSiMqPh H,SiEt, H,SiEt, H,SiPhNp" H,SiPhNp" H,SiPhNp" H,SiPh,

PhCOMe PhCOMe PhCOMe PhCOMe PhCOEt PhCOEt PhCOPf PhCOPf PhCOBu' PhCOBu' PhCOBu' PhCOCH, NMe PhCO(CH,),NMe, 'BuCOCH,OCOPh PhCOCH,CI p-MeOC&COPh (

Catalysta

Hydrosilane

Prochiral Ketone

TABLE1. TYPICAL RESULTS ON ASYMMETRIC REDUCTION OF PRoCHIRAL KETONES VIA HYDROSILYLATION

5

5

4 4 4 17 1 1 20b 20b 20b 5

4

8 3 4 6

4 5

Reference

250


The steps involved are (I) oxidative addition of hydrosilane to the rhodium(1) complex (5) to give 6;(ii) insertion of the carbonyl into the resulting silicon-rhodium bond of 6 to form diastereomeric a-siloxyalkylrhodium hydride intermediate 8, and (3) formation of an optically active silyl ether of sec-alcohol by reductive elimination:

H

I

Rh(P;)(S)Cl -k 5 H-si E

6

G sSi-Rh(P;)Cl

P* = chiral phosphine; S = solvent H

H

I ‘C-Rh(P:)Cl \*

a SiO’

,

8

/

SiO

I

.\

7

s Si,-Rh(P;)Cl f

Si-shift

‘\\

’.o&

--+H+

P hI( ? ! H C O O h

(i)

OH 21% e.e., S = (-)DIOP, 37% e.e., R = (+)DIOP, 60% e.e., S = (+)DIOP, 77% e.e., S = PPh,,

The marked increase in optical yield with pyruvates compared with simple prochiral ketones can be ascribed to a ligand effect of the ester moiety in the key intermediate or transition state. Further support of this hypothesis comes from the results for asymmetric hydrosilylation of le~ulinates~~. The hydrosilylation of levulinates followed by acid solvolysis affords 4-methyl- ybutyrolactone with 8 0 4 4 % e.e. through the silyl ether of 4-hydroxypentanoates, e.g., CH,CO(CH2)2COOBui

H,SiPhNp" [( +)DIOP]Rh(S)Cl

>

CH3~H(CH,),COOBu'

I*

TsOH MeOH

OSiHPhND"

Supporting evidence is obtained in the reaction of ethyl levulenate with H,SiPh, catatalyzed by the (ip-pybox)-Rh(II1) complex, which gives the corresponding ( 0 4 hydroxypentanoate with 95% e.e. in 91% yield5. As to the structure of the catalytic species in the hydrosilylation of keto esters, an interesting difference between neutral and cationic rhodium(1) complexes was disclosed35.Use of a cationic complex, DIOP-Rh +,in the hydrosilylation of n-propyl pyruvate caused a decrease in the optical yield (41% e.e.) compared with that obtained by using the corresponding neutral complex, DIOP-RhN. This shows that the chlorine ligand of the neutral catalyst plays a significant role in the induction of asymmetry. However, such a difference in optical yield was not observed in the case of isobutyl levulinate: DIOP-RhN,85% e.e.; DIOP-Rh+, 84% e.e. To accommodate the results, two different modes for the ligand effect of the ester moiety in the key intermediate, 9 and 10, were

?R

9

10


253

Asymmetric hydrosilylation of a-ketoacylamino esters catalyzed by DIOP-RhN complexes and RhC1(PPh3), followed by methanolysis gave a-hydroxyacylamino esters which are depsipeptide building blocks3'. Asymmetric induction by the chiral catalysts predominates over that by the chiral center in the substrate. A relatively large simple asymmetric induction was observed on using achiral catalyst, RhCl(PPh3),.

Y

A COOMe MeCOCONH

*

%2 MeCHCONH MeOH 1

Y

ACOOMe

(k)

OH (+)DIOP-RhN 68% d.e. (S, S) (-)DIOP-RhN 72% d.e. (R, S) RhCl(PPh,), 42% d.e. (S, S)

PhCOCONH

COOMe

%

MeOH

PhEHCONH I

OH

XPh

COOMe

(1)

(+)DIOP-RhN 82% d.e. (S, S) (-)DIOP-RhN 42% d.e. (R, S) RhC1(PPh3) 56% d.e. (S,S) Asymmetric synthesis at a prochiral silicon center in catalytic asymmetric reactions has been effected in the hydrosilylation of ketones with dihydrosilanes using rhodium(1) complexes as catalyst^'^*^^. Addition of certain dihydrosilanes, H2SiR'R2 (R' # Rz), to ketones in the presence of chiral rhodium complexes gives silyl ethers in an optically active form associated with the silicon atom, which are converted into optically active monohydrosilanes by the action of Grignard reagents, R3MgX (R3 # R', R2):

+

Et2C=o

[( +)DIOP]Rh(S)Cl.

MeMgBr

,Et2CHOH + HSiMePhNp"

(m)

(46% e.e., R)

H,SiPhNp"

With prochiral ketones asymmetric induction at both silicon and carbon centers takes place6a:

+

PhCoEt H,SiPhNp"

[( +)DIOP]Rh(S)Cl.

MeMgBr

* * ,PhEtCHOH + HSiMePhNp"

(n)

(56% e.e., S) (36% e.e., R)

With chiral ketones, asymmetric induction takes place at a silicon center39even with an achiral catalyst such as Rh(PPh,),Cl, which is an application of the stereoselective hydrosilylation of cyclic terpene ketones (see 14.4.4.2). Double asymmetric induction is also effective and achieves a high optical yield2?

254

14.4. Addition Reactions 14.4.4 Hydrosilylation of Carbonyl Compounds 14.4.4.3. Asymmetric Synthesis ( - IMenthone

+

H,SiPhNp"

[( +)DIOP]Rh(S)Cl

Neomenthol (isomer ratio = 86/14)

*

+

HSiEtPhNp" (82% e.e., R)

(0)

(I. OJIMA)

1. (a) I. Ojima, in The Chemistry of Organosilicon Compounds, S. Patai, Z. Rappoport, eds., Wiley, New York, 1989, Chap. 25, pp. 1479-1526. (b) I. Ojima, K. Yamamoto, M. Kumada, in Aspects of Homogeneous Catalysis, Vol. 3, R. Ugo, ed., D. Reidel, Dordrecht, Holland, 1977, pp. 189-228. 2. H. Brunner, Angew. Chem. Int. Ed. Engl., 22, 897 (1983). 3. (a) H. Brunner, R. Becker, G . Riepl, Organometallics, 3, 1354 (1984). (b) H. Brunner, G . Riepl, H. Weitzer, Angew. Chem. Int. Ed. Engl., 22, 331 (1983). 4. (a) H. Brunner, G. Riepl, Angew. Chem. Int. Ed. Engl., 21, 377 (1982). (b) H. Brunner, B. Reiter, G . Riepl, Chem. Ber., 117, 1330(1984). (c) H. Brunner, A. F. M. M. Rahman, Chem. Ber., 117,710(1984). (d)H.Brunner,H. Weber,Chem. Ber., 118,3380(1985).(e)H.Brunner, G. Riepl, Inorg. Chim. Acta, 112, 65 (1986). 5. (a) H. Nishiyama, H. Sakaguchi, T. Nakamura, M. Horihata, M. Kondo, K. Itoh, Organometallics, 8, 846 (1989). (b) H. Nishiyama, M. Kondo, T. Nakamura, K. Itoh, Organometallics, 10, 500 (1991). 6. K. Yamamoto, T. Hayashi, M. Kumada, J. Organometal. Chem., 46, C65 (1972). 7. T. Hayashi, K. Yamamoto, M. Kumada, J . Organometal. Chem., 112, 253 (1976). 8. (a) I. Ojima, T. Kogure, M. Kumagai, S . Horiuchi, T. Sato, J . Organometal. Chem., 122, 83 (1976). (b) I. Ojima, T. Kogure, Y. Nagai, Chem. Lett., 541 (1973). (c) I. Ojima, Y. Nagai, Chem. Lett., 223 (1974). 9. W. Dumont, J.-C. Poulin, T.-P. Dang, H. B. Kagan, J . Am. Chem. SOC.,95, 8295 (1973). 10. (a) R. J. P. Corriu, J. J. E. Moreau, J . Organometal. Chem., 85, 19 (1975); (b) R. J. P. Corriu, J. J. E. Moreau, J . Organometal. Chem., 64, C51 (1974). 11. (a) J. Benes, J. Hetflejs, Coll. Czech. Chem. Commun.,41,2264 (1976). (b) J. Benes, J. Hetflejs, Proc. 6th Int. Cong. Catal. 2 , 1034 (1976). 12. (a) T. Hayashi, K. Yamamoto, M. Kumada, Tetrahedron Lett., 4405 (1974). (b) T. Hayashi, T. Mise, M. Fukushima, M. Kagotani, N. Nagashima, Y. Hamada, A. Matsumoto, S. Kawakami, M. Konishi, K. Yamamoto, M. Kumada, Bull. Chem. SOC.Jpn., 53, 1138 (1980). 13. K. Tamao, H. Yamamoto, H. Matsumoto, N. Miyake, T. Hayashi, M. Kumada, Tetrahedron Lett., 1389 (1977). 14. J.-C. Poulin, W. Dumont, T.-P. Dang, H. B. Kagan, C . R. Acad. Sci. Paris, Ser. C . , 227, 41 (1973). 15. T. Masuda, J. K. Stille, J . Am. Chem. Soc., 100, 268 (1978). 16. M. Capka, CON.Czech. Chem. Commun., 42, 3410 (1977). 17. (a) I. Kolb, M. Cemy, J. Hetflejs, React. Kinet. Catal. Lett., 7, 199 (1977). (b) H. B. Kagan. T. Yamagishi, J. C. Motte, Israel J . Chem., 17, 274 (1978). 18. D. Valentine, Jr., J. W. Scott, Synthesis, 329 (1978). 19. (a) L. Homer, H. Winkler, A. Rapp, A. Mentrup, H. Hoffman, D. Beck, Tetrahedron Lett., 161 (1961). (b) K. Naumann, G. Zon, and K. Mislow, J. Am. Chem. SOC.,91, 7012 (1969). 20. H. B. Kagan, T.-P. Dang, J . Am. Chem. Soc., 94,6429 (1972). 21. H. Brunner, W. Miehling, J . Organornetal. Chem., 275, C17 (1984). 22. D. Wang, T. H. Chang, Tetrahedron Lett., 24, 1573 (1983). 23. T. Hayashi, K. Yamamoto, K. Kasuga, H. Omizu, M. Kumada, J . Organomet. Chem., 113, 127 (1976); K. Yamamoto, T. Hayashi, M. Kumada, J . Organomet. Chem., 54, C45 (1973). 24. I. Ojima, M. Nihonyanagi, Y. Nagai, Bull. Chem. SOC.Jpn., 45, 3722 (1972). 25. J. F. Peyronel, H. B. Kagan, Nouv. J . Chim., 2,211 (1978). 26. (a) H. B. Kagan, J. F. Peyronel, T. Yamagishi, in Advances in Chemistry Series, No. 173, R. B. King, ed., American Chemical Society, Washington, D. C., 1979, pp. 55-66. (b) J. F. Peyronel, 4057 (1980). J. C. Fiaud, H. B. Kagan, J . Chem. Res. (M), 27. I. Ojima, T. Kogure, Organometallics, 1 , 1390 (1982). 28. T. Hayashi, K. Yamamoto, M. Kumada, Tetrahedron Lett., 3 (1975). 29. I. Ojima, T. Kogure, Y. Nagai, Chem. Lett., 985 (1975); T. Kogure, I. Ojima, Abstr. 24th Sympos. Organomet. Chem. Jpn., 115B (1976).

255

14.4. Addition Reactions 14.4.5 Hydrosilylationof Carbon-Nitrogen Double Bonds 14.4.5.1. Of lmines

30. T. Kogure, Ph.D. Thesis, Tohoku University, 1979. 31. M. Kobayashi, T. Koyama, K. Ogura, S. Seto, F. J. Ritter, I. E. M. Bruggemann-Rotgans, J . Am. Chem. Soc., 102,6602 (1980). 32. K. Yamamoto, K. Tsuruoka, J. Tsuji, Chem. Lett., 1 1 15 (1977). 33. I. Ojima, T. Kogure, M. Kumagai, J. Org. Chem., 42, 1671 (1977). 34. I. Ojima, T. Kogure, Y. Nagai, Tetrahedron Lett., 1889 (1974). 35. I. Ojima, Pure Appl. Chem., 56,99 (1984). 36. I. Ojima, K. Hirai, in Asymmetric Synthesis, Vol. 5 , J. D. Morrison, ed., Academic Press, New York, 1985, pp. 104-125. 37. I. Ojima, T. Tanaka, T. Kogure, Chem. Lett., 823 (1981). 38. T. Hayashi, K. Yamamoto, M. Kumada, Tetrahedron Lett., 331 (1974). 39. (a) R. J. P. Comu, J. J. E. Moreau, J . Organomet. Chem., 91, C27 (1975); (b) R. J. P. Comu, J. J. E. Moreau, Nouv.J . Chim., 1, 71 (1977).

14.4.5 Hydrosilylation of Carbon-Nitrogen Double Bonds The hydrosilylation of carbon-nitrogen double bonds has attracted only limited attention. Metal salts such as ZnCl,, NiCl,, and PtC1, catalyze the hydrosilylation of Nsubstituted imines'** and an N-trimethylsilyl ketenimine3. The former reaction requires drastic conditions (- 150°C) and gives complex results, but Rh(PPh,),Cl is an excellent catalyst for the reaction to give the corresponding N-silylamines in high yield under mild conditions3. Asymmetric reduction of prochiral imines is also successfully performed4*'. Hydrosilylations of isocyanates7 and carbodiimide~~.'are effectively promoted by Rh(PPh3),C1 or PdC1,. (I. OJIMA)

1. E. Frainnet, A. Bazouin, R. Calas, C. R. Hebd Seances Acad. Sci. Paris, 257, 1304 (1963). 2. K. A. Andrianov, V. I. Sidorov, M. I. Filimonova, Dokl. Akud. Nauk SSSR, 220, 349 (1975). 3. M. Fieser, L. F. Fieser, eds., Reagents for Organic Synthesis, Vol. 5, Wiley, New York, 1975, p. 739. 4. (a) N. Langlois, T.-P. Dang, H. B. Kagan, Tetrahedron Lett., 2475 (1973). (b) H. B. Kagan, N. Langlois, T.-P. Dang, J . Organometal. Chem., 90, 353 (1975). 5. R. Becker, H. Brunner, S. Mahboobi, W. Wiegrebe, Angew. Chem. Int. Ed. Engl., 24, 995 (1985). 6. I. Ojima, S. Inaba, Y. Nagai, J . Organomet. Chem., 72, C11 (1974). 7. I. Ojima, S. Inaba, J . Organomet. Chem., 140,97 (1977).

14.4.5.1. Of lmlnes

The hydrosilylation of imines giving N-silylamines is synthetically valuable, not only because it provides a convenient route to N-silylamines, which are known as useful reagents for organic synthesis, but also because it is a new and effective method for the reduction of imines': R1RZC=N-R3

+ HSiR, 5 R'R2CH-N-R3 I

H,q R'RZCHNHR3

(a)

SiR, The reaction in the presence of Rh(PPh3),C1 or PdCl, proceeds under much milder conditions and in higher than the ZnC1,-catalyzed addition of triethylsilane to benzylideneaniline. Dihydrosilanes react more smoothly than monohydrosilanes. Based


255





14.4.5.1. Of lmlnes



H,q R'RZCHNHR3

(a)



255





14.4.5.1. Of lmlnes



H,q R'RZCHNHR3

(a)


256


on the reaction with dihydrosilanes, the catalytic activities are Rh(PPh3),C1 > Rh(PPh,),(CO)Cl > Py,RhCl(DMF)BH, (Py = pyridine; DMF = dimethylformamide) > [(1,5-hexadiene)RhC1I2 > [(1,5-cyc1ooctadiene)RhC1], > PdCl, > Pd(PPh3),C12. Chloroplatinic acid cannot be used because of many side reactions. The silylamine thus produced is desilylated by methanol to the corresponding amine or is converted to the corresponding amide by the reaction with acyl chloride liberating chlorosilane, e.g.: j = g N H C H 2 P(96%) h @NcH2ph

5OoC

SiHEt,

NCHzPh

(b)

I

MeCOCl

COMe (95%) Asymmetric reduction of ketenimines to optically active secondary amines via hydrosilylation is achieved by using [( )DIOP]Rh(S)Cl as catalyst:

+

\OMe (39% e.e., S)

-0Me

The optical yield depends considerably on the reaction temperature, i.e., the lower temperature increases the stereoselectivity?

24°C: 50% e.e., S 2°C: 65% e.e., S (-)DIOP-RhN

Ph

(CF3CO)zO

H

.

I

COCF3 64% e.e.

(e)

The hydrosilylation of oximes with H,SiPh, and RhCl(PPh,), gives the corresponding N-silylamines via N-siloxyimine intermediate^^^. The reactions using DIOP-RhN catalysts followed by hydrolysis give optically active amines with low to moderate optical purities7.

14.4. Addition Reactions 14.4.5 Hydrosilylation of Carbon-Nitrogen Double Bonds 14.4.5.2. Of Isocyanates, Carbodiimides, and Nitriles

Ph)fBut N %OH

HC

(-)DIOP-RhN

--Kz

257

"Y

(f)

NHZ 36% e.e.

Using camphor oxime as a substrate, a carbon-carbon bond cleavage occurred to give an optically pure aminoethylcyclopentanone in addition to the reduction products, bornylamines8. The complexes RhCl(PPh,), (with H,SiPh2), K[PtCl,C,H,] (with HSiMeCl,), and PtO,.H,O (with HSiMeCl,) are effective as catalysts.

N-oH

H,SiPh, >-----, RhCl(PPh,), MeOH 38% yield

4%

15%

81% (I. OJIMA)

1. 2. 3. 4. 5.

6. 7. 8.

I. Ojima, S . Inaba, J . Organornet. Chem., I40,97 (1977). E. Frainnet, A. Bazouin, R. Calas, C . R . Hebd. Seances Acad. Sci. Paris, 257, 1304 (1963). I. Ojima, S. Inaba, Y. Nagai, J . Organornet. Chern., 72, C11 (1974). (a) I. Ojima in The Chemistry of Organosilicon Compounds, S.Patai, Z . Rapoport eds., Wiley, 1989; Chapter 25, pp. 1479-1526. (b) I. Ojima, K. Yamamoto, M. Kumada, in Aspects of HomogeneousCatalysis, Vol. 3, R. Ugo, ed., D. Reidel, Dordrecht, Holland, 1977; pp. 189-228. (a) N. Langlois, T.-P. Dang, H. B. Kagan, Tetrahedron Lett., 2475 (1973). (b) H. B. Kagan, N. Langlois, T.-P. Dang, J . Organometal. Chem., 90, 353 (1975). R. Becker, H. Brunner, S.Mahboobi, W. Wiegrebe, Angew. Chern. Int. Ed. Engl., 24,995 (1985). (a) H. Brunner, R. Becker, Angew. Chem. Int. Ed. Engl., 23, 222 (1985). (b) H. Brunner, R. Becker, S. Gauder, Organornetallics, 5 , 7 3 9 (1986). H. Brunner, R. Becker, Angew. Chem. h t . Ed. Engl., 24, 703 (1985).

14.4.5.2. Of Isocyanates, Carbodiimides, and Nitriies

The hydrosilylation of isocyanates with HSiEt, catalyzed by PdCl, gives either

N-triethylsilylformamides1 or C-triethylsilylamides 2 depending on the structure of iso-

cyanates, viz., the reaction of aryl isocyanates gives 1as sole product via normal addition of the silane to the C=N bond of isocyanates, while alkyl isocyanates gives 2 exclusively via reverse addition'.,:



Ph)fBut N %OH

HC

(-)DIOP-RhN

--Kz

257

"Y

(f)

NHZ 36% e.e.

Using camphor oxime as a substrate, a carbon-carbon bond cleavage occurred to give an optically pure aminoethylcyclopentanone in addition to the reduction products, bornylamines8. The complexes RhCl(PPh,), (with H,SiPh2), K[PtCl,C,H,] (with HSiMeCl,), and PtO,.H,O (with HSiMeCl,) are effective as catalysts.

N-oH

H,SiPh, >-----, RhCl(PPh,), MeOH 38% yield

4%

15%

81% (I. OJIMA)

1. 2. 3. 4. 5.

6. 7. 8.

I. Ojima, S . Inaba, J . Organornet. Chem., I40,97 (1977). E. Frainnet, A. Bazouin, R. Calas, C . R . Hebd. Seances Acad. Sci. Paris, 257, 1304 (1963). I. Ojima, S. Inaba, Y. Nagai, J . Organornet. Chern., 72, C11 (1974). (a) I. Ojima in The Chemistry of Organosilicon Compounds, S.Patai, Z . Rapoport eds., Wiley, 1989; Chapter 25, pp. 1479-1526. (b) I. Ojima, K. Yamamoto, M. Kumada, in Aspects of HomogeneousCatalysis, Vol. 3, R. Ugo, ed., D. Reidel, Dordrecht, Holland, 1977; pp. 189-228. (a) N. Langlois, T.-P. Dang, H. B. Kagan, Tetrahedron Lett., 2475 (1973). (b) H. B. Kagan, N. Langlois, T.-P. Dang, J . Organometal. Chem., 90, 353 (1975). R. Becker, H. Brunner, S.Mahboobi, W. Wiegrebe, Angew. Chern. Int. Ed. Engl., 24,995 (1985). (a) H. Brunner, R. Becker, Angew. Chem. Int. Ed. Engl., 23, 222 (1985). (b) H. Brunner, R. Becker, S. Gauder, Organornetallics, 5 , 7 3 9 (1986). H. Brunner, R. Becker, Angew. Chem. h t . Ed. Engl., 24, 703 (1985).

14.4.5.2. Of Isocyanates, Carbodiimides, and Nitriies

The hydrosilylation of isocyanates with HSiEt, catalyzed by PdCl, gives either

N-triethylsilylformamides1 or C-triethylsilylamides 2 depending on the structure of iso-

cyanates, viz., the reaction of aryl isocyanates gives 1as sole product via normal addition of the silane to the C=N bond of isocyanates, while alkyl isocyanates gives 2 exclusively via reverse addition'.,:

258


R-"=C=O

+

HSiEt,

PdCI,

C

-lEI

SiEt,

I

R-NHCHO

R-N-CHO (R = 1"yI) R-NH--0-SiEt, (R = alkyl)

R,COCI

(a)

R-YCHO I

COR'

3

R = a-naphthyl, p-chlorophenyl, phenyl, cyclohexyl, n-butyl Palladium on carbon exhibits similar catalytic activity, whereas Rh(PPh,),Cl does not promote the reaction. The resulting 1 or 2 is desilylated by methanol giving N-substituted formamide quantitatively, while the reaction with acyl chloride proceeds smoothly to give N-acylformamide. Hydrosilylation of carbodiimides proceeds only at 140-200°C in the presence of PdCl, or Rh(PPh,),Cl to give N-silylformamidines 3. The N-silylformamidine thus obtained reacts with methanol and acyl chlorides to afford the corresponding formamidines and N-acyl formamidines, respectively2":

-c: 3

R-

R"C0CI

R = isopropyl, cyclohexyl

R-

NH-CH=N-R ( b)

N-CH=N-R

I

COR"

The hydrosilylation of aromatic nitriles is promoted by Co2(CO)* to give the corresponding NJ-bis(trimethylsily1)benzylamines in moderate to excellent yields4. Functional substituents on a phenyl nucleus such as MeO, C1, Me2N-, and MeOCO-groups, are tolerant of the reaction and do not affect the reaction. M e 0 C - N

HSiMe3/C0 CO,(CO),

9

Me-@CH,N(SiMe,),

(c)

(91%) HSiMe3/C0

Me2N-@CH2N(SiMe,),

(d)

(73%) (I. OJIMA)

14.4. Addition Reactions 14.4.6. Hydrocyanation of Olefins and Dienes 1. 2. 3. 4.

259

I. Ojima, S. Inaba, Y. Nagai, Tetrahedron Lett., 4363 (1973). I. Ojima, S. Inaba, J . Organornet. Chern., 140,97 (1977). I. Ojima, S. Inaba, Y. Nagai, J . Organornet. Chern., 72, C11 (1974). T. Murai, T. Sakane, S. Kato, Tetrahedron Letr., 26,5145 (1985).

14.4.6. Hydrocyanation of Olef ins and Dienes Organic nitriles as a class are of significant practical utility since they are precursors to numerous other groups of organic compounds. By hydrolysis or alcoholysis, carboxylic acids or esters are obtained, also both useful as organic intermediates. Hydrogenation affords primary amines, which are useful themselves and as precursors to still other organic compounds, e.g., higher amines, isocyanates, amine salts, and amides. Nitriles have frequently been the object of preparative studies. Several methods have been explored and developed for their preparation, but none is more appealing for its simplicity than reaction of hydrogen cyanide'.z with unsaturated organics3.This reaction, olefin hydrocyanation, has been the subject of studies since the 1950s. The facility with which reaction proceeds depends on the reactivity of the olefin:

"'Hl;+

HCN

RZ

- "'ft""l R,

H

(a)

CN

Olefins activated by appropriate substituents react readily with HCN without a catalyst, but even conjugated olefins containing adjacent a,P-unsaturated activating substituents4, e.g., C=O, COOR, require use of cyanide ion, CN- to facilitate reaction'. Nonactivated olefins fail to react even under strenuous conditions with cyanide anion catalysis. Due to this lack of reactivity coupled with the inherent desirability of the products, much research has focused on developing catalysts for the hydrocyanation of these nonactivated olefins. This has led to nickel, palladium, copper, and cobalt-based catalysts effective at 25125°C with or without a solvent. Most were developed for the hydrocyanation of unactivated olefins, but many are equally applicable for other olefins. For example, much work has been reported on butadiene hydrocyanation employing all of the catalysts mentioned above except palladium. A multistep synthetic method has been reported for accomplishing hydrocyanation of various types of olefins that utilizes an organic isonitrile instead of HCN'. The olefin is first treated with a stoichiometric quantity of Cp,Zr(H)Cl(Cp = $C,H,), followed by successive additions of t-butyl isocyanide and iodine. Trimethylsilyl isocyanide was also used as the CN source. The remainder of 14.4.6 deals with the four major hydrocyanation catalyst groups described in the literature to date. (E. S. BROWN, B. D. DOMBEK) 1. W. Nagata, M. Yoshiska, Organic Reactions, Vol. 25, Wiley, New York, 1977, p. 255. 2. Caution. Hydrogen cyanide is very toxic and should be used only in a well-ventilated hood. User should wear heavy gloves and avoid all contact with skin or clothing. 3. To avoid handling of hydrogen cyanide directly, cyanohydrinshave been employed as a source of HCN under reaction conditions. See, for example, a disclosure of this technique employing acetone cyanohydrin and either nickel, palladium, copper or cobalt catalysts: W. C. Drinkard, Jr., U.S. Pat. 3,655,723 (1972); Chern.Absrr., 77, 4986p (1972).


14.4. Addition Reactions 14.4.6. Hydrocyanation of Olefins and Dienes 1. 2. 3. 4.

259

I. Ojima, S. Inaba, Y. Nagai, Tetrahedron Lett., 4363 (1973). I. Ojima, S. Inaba, J . Organornet. Chern., 140,97 (1977). I. Ojima, S. Inaba, Y. Nagai, J . Organornet. Chern., 72, C11 (1974). T. Murai, T. Sakane, S. Kato, Tetrahedron Letr., 26,5145 (1985).

14.4.6. Hydrocyanation of Olef ins and Dienes Organic nitriles as a class are of significant practical utility since they are precursors to numerous other groups of organic compounds. By hydrolysis or alcoholysis, carboxylic acids or esters are obtained, also both useful as organic intermediates. Hydrogenation affords primary amines, which are useful themselves and as precursors to still other organic compounds, e.g., higher amines, isocyanates, amine salts, and amides. Nitriles have frequently been the object of preparative studies. Several methods have been explored and developed for their preparation, but none is more appealing for its simplicity than reaction of hydrogen cyanide'.z with unsaturated organics3.This reaction, olefin hydrocyanation, has been the subject of studies since the 1950s. The facility with which reaction proceeds depends on the reactivity of the olefin:

"'Hl;+

HCN

RZ

- "'ft""l R,

H

(a)

CN

Olefins activated by appropriate substituents react readily with HCN without a catalyst, but even conjugated olefins containing adjacent a,P-unsaturated activating substituents4, e.g., C=O, COOR, require use of cyanide ion, CN- to facilitate reaction'. Nonactivated olefins fail to react even under strenuous conditions with cyanide anion catalysis. Due to this lack of reactivity coupled with the inherent desirability of the products, much research has focused on developing catalysts for the hydrocyanation of these nonactivated olefins. This has led to nickel, palladium, copper, and cobalt-based catalysts effective at 25125°C with or without a solvent. Most were developed for the hydrocyanation of unactivated olefins, but many are equally applicable for other olefins. For example, much work has been reported on butadiene hydrocyanation employing all of the catalysts mentioned above except palladium. A multistep synthetic method has been reported for accomplishing hydrocyanation of various types of olefins that utilizes an organic isonitrile instead of HCN'. The olefin is first treated with a stoichiometric quantity of Cp,Zr(H)Cl(Cp = $C,H,), followed by successive additions of t-butyl isocyanide and iodine. Trimethylsilyl isocyanide was also used as the CN source. The remainder of 14.4.6 deals with the four major hydrocyanation catalyst groups described in the literature to date. (E. S. BROWN, B. D. DOMBEK) 1. W. Nagata, M. Yoshiska, Organic Reactions, Vol. 25, Wiley, New York, 1977, p. 255. 2. Caution. Hydrogen cyanide is very toxic and should be used only in a well-ventilated hood. User should wear heavy gloves and avoid all contact with skin or clothing. 3. To avoid handling of hydrogen cyanide directly, cyanohydrinshave been employed as a source of HCN under reaction conditions. See, for example, a disclosure of this technique employing acetone cyanohydrin and either nickel, palladium, copper or cobalt catalysts: W. C. Drinkard, Jr., U.S. Pat. 3,655,723 (1972); Chern.Absrr., 77, 4986p (1972).

260

14.4. Addition Reactions 14.4.6. Hydrocyanationof Olefins and Dienes 14.4.6.1. By Nickel Catalysts

4. Use of heterogeneous catalysts containing alkali or alkaline earth salts for the gas-phase hydrocyanation of activated olefins, primarily methyl acrylate, has been reported: F. A. Pesa, A. M. Graham, U.S.Patent 4,367,179 (1983); Chem. Abstr., 98, 1976228 (1983). 5. S. L. Buchwald, S. J. LaMaire, Tetrahedron Lett., 28, 295 (1987).

14.4.6.1. By Nickel Catalysts

Interest in the hydrocyanation of nonactivated olefins with nickel catalysts arose from the discovery that finely divided nickel' or nickel cyanide' on inert supports gives higher yields of nitrile products at less severe reaction conditions than do cobalt or copper heterogeneous catalysts3. Use of nickel carbonyl to add 1 mol of HCN to 1,3-butadiene may be the first example of hydrocyanation by a homogeneous nickel catalysf. That work also recorded the important observation that substantial improvement in nitrile product yield results from conducting the reaction in the presence of (C,H,),P or (C,H,),As. This work led to extensive studies to develop effective nickel hydrocyanation catalysts. Virtually all subsequent developments have focused on finding the most effective nickel complex and the identification and application of promoters to improve catalyst efficiency and life5. The extensive work now done confirms zerovalent phosphine or phosphite complexes as the most effective nickel catalysts tested (see Table 1). Further improvements have been gained through the use of more bulky ligands6v7. Hydrocyanation usually leads to anti-Markovnikov addition:

R

+ HCN

- '\

CN

(For electrophilic attack, Markovnikov addition is that in which the positive portion of the reagent adds to the least substituted carbon atom of the double bond undergoing reaction.) This may result from a steric preference for the least-substituted metal alkyl intermediate formed by insertion of olefin into the metal hydride bond8v9. Vinylarenes comprise an exception, where interaction of nickel with the aromatic ring stabilizes the precursor of the branched nitrile, leading primarily to a Markovnikov addition produ~t~.'~. Although anti-Markovnikov addition is favored for many olefins, significant amounts of branched nitrile products are often observed. This can be because of rapid olefin isomerization prior to hydrocyanation. For example, it is estimated that 85% of the 2-methylhexenenitrile formed in hexene- 1 hydrocyanation results from direct hydrocyanation of hexene-2 and only 15% from Markovnikov addition to hexene-18. Lewis acids improve the hydrocyanation results with nickel. Two classes of promoters most frequently mentioned are the metal halides typified by ZnC1,6.8 and A1C137.9s11 and a variety of boron compo~nds~~''. These promoters work by binding to the nitrogen atom of coordinated cyanide, reducing the electron density on nicke15s9.The Lewis acid promoters can enhance rates, improve yields (particularly when employed in hydrocyanation of nonactivated olefins), and afford substantial improvements in catalyst life. Some, in particular bulky boron compounds, increase the selectivity to a linear nitrile product by binding to coordinated cyanide and shifting equilibria to favor a less sterically crowded linear alkyl nickel intermediate. With appropriate use of promoters, product yields are often greater than 90%,and catalyst turnover numbers (mole product per mole catalyst) are in the hundreds.


260

14.4. Addition Reactions 14.4.6. Hydrocyanationof Olefins and Dienes 14.4.6.1. By Nickel Catalysts

4. Use of heterogeneous catalysts containing alkali or alkaline earth salts for the gas-phase hydrocyanation of activated olefins, primarily methyl acrylate, has been reported: F. A. Pesa, A. M. Graham, U.S.Patent 4,367,179 (1983); Chem. Abstr., 98, 1976228 (1983). 5. S. L. Buchwald, S. J. LaMaire, Tetrahedron Lett., 28, 295 (1987).

14.4.6.1. By Nickel Catalysts

Interest in the hydrocyanation of nonactivated olefins with nickel catalysts arose from the discovery that finely divided nickel' or nickel cyanide' on inert supports gives higher yields of nitrile products at less severe reaction conditions than do cobalt or copper heterogeneous catalysts3. Use of nickel carbonyl to add 1 mol of HCN to 1,3-butadiene may be the first example of hydrocyanation by a homogeneous nickel catalysf. That work also recorded the important observation that substantial improvement in nitrile product yield results from conducting the reaction in the presence of (C,H,),P or (C,H,),As. This work led to extensive studies to develop effective nickel hydrocyanation catalysts. Virtually all subsequent developments have focused on finding the most effective nickel complex and the identification and application of promoters to improve catalyst efficiency and life5. The extensive work now done confirms zerovalent phosphine or phosphite complexes as the most effective nickel catalysts tested (see Table 1). Further improvements have been gained through the use of more bulky ligands6v7. Hydrocyanation usually leads to anti-Markovnikov addition:

R

+ HCN

- '\

CN

(For electrophilic attack, Markovnikov addition is that in which the positive portion of the reagent adds to the least substituted carbon atom of the double bond undergoing reaction.) This may result from a steric preference for the least-substituted metal alkyl intermediate formed by insertion of olefin into the metal hydride bond8v9. Vinylarenes comprise an exception, where interaction of nickel with the aromatic ring stabilizes the precursor of the branched nitrile, leading primarily to a Markovnikov addition produ~t~.'~. Although anti-Markovnikov addition is favored for many olefins, significant amounts of branched nitrile products are often observed. This can be because of rapid olefin isomerization prior to hydrocyanation. For example, it is estimated that 85% of the 2-methylhexenenitrile formed in hexene- 1 hydrocyanation results from direct hydrocyanation of hexene-2 and only 15% from Markovnikov addition to hexene-18. Lewis acids improve the hydrocyanation results with nickel. Two classes of promoters most frequently mentioned are the metal halides typified by ZnC1,6.8 and A1C137.9s11 and a variety of boron compo~nds~~''. These promoters work by binding to the nitrogen atom of coordinated cyanide, reducing the electron density on nicke15s9.The Lewis acid promoters can enhance rates, improve yields (particularly when employed in hydrocyanation of nonactivated olefins), and afford substantial improvements in catalyst life. Some, in particular bulky boron compounds, increase the selectivity to a linear nitrile product by binding to coordinated cyanide and shifting equilibria to favor a less sterically crowded linear alkyl nickel intermediate. With appropriate use of promoters, product yields are often greater than 90%,and catalyst turnover numbers (mole product per mole catalyst) are in the hundreds.

Ni[P(O-p-tolyl),],, excess ligand, ZnCI,

Ni[P(O-p-tolyl),],, excess ligand

~

NiZ+ or Ni(0) on a support [Ni(CN),I4 [Ni(CN),I4 Ni[P(O-p-tolyl),],, excess ligand, metal salts or boron compounds Ni[P(O-p-t~lyl)~],, excess ligand, AlCI, Ni[P(O-p-tolyl),],, excess ligand, B(C,H,), Ni[P(O-p-tolyl),],, excess Iigand, B(C&15),

Catalyst c2H4

CH,O

p

(C,H,)CHCH,

CH,:CHCH,

Hexene-1

CH,:CHCOOC,H, CH,:CHCN Substituted norbornenes

Olefin

ProducP

CH30m

c

I

CH3

I

Heptenenitrile (70) Branched nitriles (30) CH3CH,CH2CN (89) CH,CH(CN)CH, (1 1) (C,H,)CH,CH,CN (33) (67) (C~H~)CH(CN)CH3) CH3

CH,CH,CN NCCH,CH,COOC,H, NCCH,CH,CN Substituted cyanonorbomanes

TABLE1. TYPICAL HYDROCYANATION REACTIONS OF OLEFINS CATALYZED BY NICKEL

N

~

68

93

b

b

b

b

40

-95 62

Yield (%)

~~

(continued)

10

10

5

5

8

16 7

1 16

Reference

2

2

n

n

P

m m

\Dm

"t 6

i=

e

/

/

a

5

0

/

m

e

I I

I

I

m

-

I

I

-

I

+2%

rl \

i

I

ii

\

\

264

14.4. Addition Reactions 14.4.6. Hydrocyanation of Olefins and Dienes 14.4.6.1. By Nickel Catalysts

The hydrocyanation of conjugated (activated) olefins occurs with greater facility than for those not activated. Strained ring olefins are intermediate in their behavior. Conjugated olefins give q3 (n-allyl) intermediates by addition to a nickel hydride that are more stable than the q' intermediates from monoenes. Polyenes are also easily hydrocyanated, particularly under conditions that induce isomerization to conjugated dienes. Addition of acids of the proper pK, to form nickel hydrides (responsible for initial olefin activation and olefin isomerization) reportedly enhances the rate of hydrocyanation of dienes and polyenes13. Even 1,6pentadiene and 13-hexadiene give products resulting from hydrocyanation of conjugated dienes formed by isomerization' However, 1,7octadiene gave a 22% yield of the linear dinitrile product, sebacic dinitrile. Hydrocyanation of 1,3-butadiene to a mixture of pentenenitriles and 2-methyl-3butenenitrile (and, in a second step, their reaction with hydrogen cyanide to produce adiponitrile) has received special attention because of its commercial application5.

'.

CN As expected, conjugate addition of HCN predominates leading to 3-pentenenitrile as the major product of the first step, proceeding through a nickel .rr-ally1 intermediate. Some 13-addition does occur, and at least some of that by Markovnikov addition leading to 2-methyl-3-butenenitrile. Isomerization of 3-pentenenitrile to 4-pentenenitrile probably occurs, enhancing the quantity of this product produced by 1,2-addition. The second process step is more demanding, since the olefinic bond in the pentenenitriles is isolated and unactivated. Lewis acid promoters are particularly beneficial, allowing both double bond isomerization of the pentenenitriles to the desired 4-pentenenitrile and promoting the selective addition of HCN to this isomer:

A chelating diphosphite prepared from biphenol and PCl, provides a very stable Ni(0) complex that catalyzes the hydrocyanantion of butadiene without excess ligand14. Although the stability of this catalyst is enhanced, the amount of butadiene dimerization byproducts is significant. A related nickel catalyst prepared using a chiral chelating diphosphite based on R-2,2'-binaphthol provides enantioselectivity in the hydrocyanation of n~rbornene'~. The major product, R-exo-2-cyanonorbornane,was obtained in an enantiomeric excess of up to 38%. Although application of nickel hydrocyanation catalysts to the production of adiponitrile has been extensively studied, much remains to be learned about other applications of these catalysts. (E. S. BROWN, B. D. DOMBEK) 1. D. D. Davis, L. S. Scott, U.S. Patent 3,278,575 (1966); Chem. Abstr., 66, 5 5 0 6 8 ~ (1967). 2. G . C. Monroe, Jr., G . N. Hammer, U.S. Patent 3,297,742 (1967); Chem. Abstr., 66, 46117j (1967). 3. J. W. Teter, U.S. Patent 2,385,741 (1945); Chem. Abstr., 40, 590 (1946). 4. P. Arthur, Jr., B. C. Pratt, U.S. Patent 2,571,099 (1951); Chem. Absrr., 46, 3068a (1952). 5 . C. A. Tolman, R. J. McKinney, W. C. Seidel, J. D. Druliner, W. R. Stevens, Adv. Cutul., 33, 1 (1985), and references therein. 6. Y.T. Chia, W. C. Drinkard, Jr., E. N. Squire, U.S. Patent 3,766,237 (1973); Chem. Absrr., 80, 703732 (1974).

14.4. Addition Reactions 14.4.6. Hydrocyanationof Olefins and Dienes 14.4.6.2. By Palladium Catalysts

265

7. C. M. King, W. C. Seidel and C. A. Tolman, U.S.Patent 3,925,445 (1975); Chem. Abstr., 84, 8892111(1976). 8. B. W. Taylor and H. E. Swift, J . Card.,26, 254 (1972), and references therein. 9. C. A. Tolman, W. C. Seidel, J. D. Druliner, J. P. Domaille, Organometullics,3, 33 (1984), and references therein. 10. W. A. Nugent, R. J. McKinney, J . Org. Chem., 50,5370 (1985). 11. W. Keim, A. Behr, H.-0. Liihr, J. Weisser,J. Caful., 78, 209 (1982). 12. C. M. King and M. T. Musser, U.S.Patent 3,864,380 (1975); Chem. Abstr., 82, 155421e(1975). 13. H. E. Bryndza, U.S.Patent 4,810,815 (1989); Chem. Absrr., I l l , 97927d (1989). 14. M. J. Baker, K. N. Harrison, A. G. Orpen, P. G. Pringle, G. Shaw, J . Chem. SOC., Chem. Commun., 803 (1991). 15. M. J. Baker, P. G. Pringle, J . Chem. Soc., Chem. Commun., 1292 (1991). 16. G. R. Coraor and W. Z. Heldt, U.S.Patent 2,904,581 (1959); Chem. Abstr., 54,4393f (1960). 17. W. C. Seidel and C. A. Tolman, U.S. Patent 3,850,973 (1974); Chem. Abstr., 82, 97704m (1975).

14.4.6.2. By Palladium Catalysts Olefin hydrocyanation using palladium catalysts has been less well studied than with nickel. Nevertheless, zerovalent complexes of palladium, particulrly triarylphosphite complexes, hydrocyanate a wide range of olefins in useful yields (see Table 1). Early work reported the merit of excess phosphorus ligand to promote the reaction', and further paralleling the observations with nickel, Lewis acids have been used to improve catalytic activity2. However, addition of ZnC1, fails to improve nitrile product ~ i e l d ~Asym.~. metric induction in hydrocyanation results in optical yields of -30% in the synthesis of exo-Zcyanonorbomane using the chiral ligand DIOP3,and studies on the stereochemistry of HCN and DCN addition to terminal alkenes and a substituted cyclohexene with the same catalyst have been reported5. Synthesis of adiponitrile from pentenenitrile has been reported using tetrakis(triarylphosphite)palladium(O~,but no report of 1,3-butadiene hydrocyanation by

TABLE1. TYPICAL HYDROCYANATION REACTIONS OF OLEFINS CATALYZED BY PALLADIUM Catalyst

Olefin

Pd[P(OR),],, excess Norbornene and P(OR), substituted derivatives Pd[P(OR),],, excess CzH4 P(OR)? Pd[P(OR),],, excess CH,:CHSi(OC,H,), P(OR), Pd[P(OR),],, metal (C,H,)CH:CH, salt, boron compound Pd[P(OR),],, metal 3-Pentenenitrile salt, boron compound Pd(DIOP)b Norbornene

Product Exo-2-cyanonorbornane

Yield (%) Reference 25-80

1

CH,CH,CN

a

1

NCCHzCHzSi(OCzH,)z

a

1

(C&,)CH(CN)CH3

a

2

Adiponitrile

20

2

Exo-2-cyanonorbornane

53

3,4

aData not available. bDIOP, 2,3-o-isopropylidene-2,3-dihydroxy1,4-bis(diphenyIphosphino)butane.


14.4. Addition Reactions 14.4.6. Hydrocyanationof Olefins and Dienes 14.4.6.2. By Palladium Catalysts

265

7. C. M. King, W. C. Seidel and C. A. Tolman, U.S.Patent 3,925,445 (1975); Chem. Abstr., 84, 8892111(1976). 8. B. W. Taylor and H. E. Swift, J . Card.,26, 254 (1972), and references therein. 9. C. A. Tolman, W. C. Seidel, J. D. Druliner, J. P. Domaille, Organometullics,3, 33 (1984), and references therein. 10. W. A. Nugent, R. J. McKinney, J . Org. Chem., 50,5370 (1985). 11. W. Keim, A. Behr, H.-0. Liihr, J. Weisser,J. Caful., 78, 209 (1982). 12. C. M. King and M. T. Musser, U.S.Patent 3,864,380 (1975); Chem. Abstr., 82, 155421e(1975). 13. H. E. Bryndza, U.S.Patent 4,810,815 (1989); Chem. Absrr., I l l , 97927d (1989). 14. M. J. Baker, K. N. Harrison, A. G. Orpen, P. G. Pringle, G. Shaw, J . Chem. SOC., Chem. Commun., 803 (1991). 15. M. J. Baker, P. G. Pringle, J . Chem. Soc., Chem. Commun., 1292 (1991). 16. G. R. Coraor and W. Z. Heldt, U.S.Patent 2,904,581 (1959); Chem. Abstr., 54,4393f (1960). 17. W. C. Seidel and C. A. Tolman, U.S. Patent 3,850,973 (1974); Chem. Abstr., 82, 97704m (1975).

14.4.6.2. By Palladium Catalysts Olefin hydrocyanation using palladium catalysts has been less well studied than with nickel. Nevertheless, zerovalent complexes of palladium, particulrly triarylphosphite complexes, hydrocyanate a wide range of olefins in useful yields (see Table 1). Early work reported the merit of excess phosphorus ligand to promote the reaction', and further paralleling the observations with nickel, Lewis acids have been used to improve catalytic activity2. However, addition of ZnC1, fails to improve nitrile product ~ i e l d ~Asym.~. metric induction in hydrocyanation results in optical yields of -30% in the synthesis of exo-Zcyanonorbomane using the chiral ligand DIOP3,and studies on the stereochemistry of HCN and DCN addition to terminal alkenes and a substituted cyclohexene with the same catalyst have been reported5. Synthesis of adiponitrile from pentenenitrile has been reported using tetrakis(triarylphosphite)palladium(O~,but no report of 1,3-butadiene hydrocyanation by

TABLE1. TYPICAL HYDROCYANATION REACTIONS OF OLEFINS CATALYZED BY PALLADIUM Catalyst

Olefin

Pd[P(OR),],, excess Norbornene and P(OR), substituted derivatives Pd[P(OR),],, excess CzH4 P(OR)? Pd[P(OR),],, excess CH,:CHSi(OC,H,), P(OR), Pd[P(OR),],, metal (C,H,)CH:CH, salt, boron compound Pd[P(OR),],, metal 3-Pentenenitrile salt, boron compound Pd(DIOP)b Norbornene

Product Exo-2-cyanonorbornane

Yield (%) Reference 25-80

1

CH,CH,CN

a

1

NCCHzCHzSi(OCzH,)z

a

1

(C&,)CH(CN)CH3

a

2

Adiponitrile

20

2

Exo-2-cyanonorbornane

53

3,4

aData not available. bDIOP, 2,3-o-isopropylidene-2,3-dihydroxy1,4-bis(diphenyIphosphino)butane.

266

14.4. Addition Reactions 14.4.6. Hydrocyanation of Olefins and Dienes 14.4.6.3. By Copper Catalysts

palladium appears to be recorded yet. Hydrocyanation of norbomadiene and dicyclopentadiene were reported in an early study'. Much is yet to be learned about palladium catalysts in olefin and diene hydrocyanation.

(E.S. BROWN, B. D. DOMBEK) 1. E. S . Brown, Inorganic Synthesis by Metal Carbonyls, Vol. 11, T. Wender, P. Pino, eds., Wiley, New York, 1977, pp. 655-672. 2. W. C. Drinkard, Jr., R. V. Lindsey, Jr., German Offen. 1,806,098 (1969): Chem. Abstr.. 71. 49343 (1969). 3. P. S. Elmes, W. R. Jackson,J. Am. Chem. Soc., 101,6128 (1979). 4. P. S. Elmes, W. R. Jackson,Ann. N.Y. Acad. Sci., 333,225 (1980). 5. W. R. Jackson, C. G. Lovel, Aust. J. Chem., 35,2053 (1982).

14.4.6.3. By Copper Catalysts

Copper(1) chloride was first disclosed as a catalyst for hydrocyanation of conjugated dienes in 1947l. The process employed an aqueous copper(1) chloride solution containing, in addition, NH4Cl, Cu metal powder, and HC1, presumably to maintain a high level of Cu+ ion in solution. The continuous reaction, conducted by flowing the diene and HCN over the copper solution at 200°C, leads only to modest nitrile yields (20-30%) even when stoichiometric quantities of copper(1) chloride are employed. With anhydrous copper(1) chloride as catalyst2 at high catalyst loadings (1 mol/mol C4H6) nitrile yields of 30% were achieved, but catalytic quantities of Cu(1) (1 mo1/50 mol C4H6)gave much lower nitrile yields (

3

C2H5

H$cH3

C2H5

(M. ORCHIN)

1. D. Dingham, D. E. Webster, F. B. Wells, J . Chem. SOC.Dalton, 1519 (1974). 14.5.1.2.3. By Platinum Catalysts.

As with other hydrido transition metal isomerization catalysts, the hydridoplatinums are also good homogeneous hydrogenation catalysts. Polyunsaturated fatty acids can be hydrogenated with [PtH(SnCl,)(PPh,),], e.g., but in the absence of molecular H, this hydride can catalyze isomerization. Thus, 1,5-~yclooctadieneundergoes stepwise migration of the double bond to the conjugated species'. When Pt-olefin complexes such as PtCl,(olefin) are treated with a second olefin, replacement of the coordinated olefin by the incoming olefin does not result in either double bond migration or (Z), (E) isomerization of the displaced olefin,. However, the olefin complexes when treated at low temperature with a nucleophile such as pyridine (py) or a secondary amine undergo conversion to a u-complex3 by a stereospecific trunsprocess4, i.e., truns-addition and truns-elimination. Treatment of truns-[PtCl,(Z)ethylene-l,Zd,)](py)] with excess py at - 15°C gives reversible formation of the carbon a-bonded complex with slow release of pure (Z)-ethylene-1,2-d2. The sequence of reactions, monitored by 'H NMR and IR, is5

275

14.5. Olefin Transformations 14.5.1. lsomerization 14.5.1.3. Skeletal Rearrangement

However, vinyl ethers do undergo Pt-catalyzed (E),(Z)isomerization6.In this special case it is possible that a trigonal stabilized carbocation intermediate is formed:

Etox!; H

(b)

(M. ORCHIN)

1. H. A. Tayim, J. C. Bailar, J. Am. Chem. Soc., 89, 3420 (1967). 2. H. B. Jonassen, W. B. Kirsch, J. Am. Chem. SOC., 79, 1279 (1957); J. Joy, M. Orchin, J. Am. Chem. SOC., 81,310 (1959). 3. P. Kaplan, M. Orchin, J. Am. Chem. SOC., 90,4175 (1968). 4. A. Panunizi, A. DeRenzi, G. Paiaro, J. Am. Chem. Soc., 92, 3488 (1970). 5. F. Pesa, M. Orchin, J. Organomet. Chem., 108, 135 (1976). 6. P. Busse, F. Pesa, M. Orchin, J. Organomet. Chem., 104, 229 (1977).

14.5.13. Skeletal Rearrangement

The skeletal rearrangement of 2,3-dimethyl- 1,Cpentadiene

is brought about by treatment of the diene with NiC12(PR3), in the presence of (i-Bu),AlCl. The active catalyst is a nickel hydride, perhaps having the structure [NiCl(H)(PR3),], where the hydrogen comes from the conversion of a Ni-isobutyl group to isobutylene'. The reaction involves fragmentation of the diene followed by recombination of the .rr-ally1 and ethylene fragments; a feature demonstrated by the thermal rearrangement' of 2-methyl-l,4-pentadienel-d, achieved in chlorobenzene for 2 h at 25°C (NiH represents the Ni hydride complex shown above):


275

14.5. Olefin Transformations 14.5.1. lsomerization 14.5.1.3. Skeletal Rearrangement

However, vinyl ethers do undergo Pt-catalyzed (E),(Z)isomerization6.In this special case it is possible that a trigonal stabilized carbocation intermediate is formed:

Etox!; H

(b)

(M. ORCHIN)

1. H. A. Tayim, J. C. Bailar, J. Am. Chem. Soc., 89, 3420 (1967). 2. H. B. Jonassen, W. B. Kirsch, J. Am. Chem. SOC., 79, 1279 (1957); J. Joy, M. Orchin, J. Am. Chem. SOC., 81,310 (1959). 3. P. Kaplan, M. Orchin, J. Am. Chem. SOC., 90,4175 (1968). 4. A. Panunizi, A. DeRenzi, G. Paiaro, J. Am. Chem. Soc., 92, 3488 (1970). 5. F. Pesa, M. Orchin, J. Organomet. Chem., 108, 135 (1976). 6. P. Busse, F. Pesa, M. Orchin, J. Organomet. Chem., 104, 229 (1977).

14.5.13. Skeletal Rearrangement

The skeletal rearrangement of 2,3-dimethyl- 1,Cpentadiene

is brought about by treatment of the diene with NiC12(PR3), in the presence of (i-Bu),AlCl. The active catalyst is a nickel hydride, perhaps having the structure [NiCl(H)(PR3),], where the hydrogen comes from the conversion of a Ni-isobutyl group to isobutylene'. The reaction involves fragmentation of the diene followed by recombination of the .rr-ally1 and ethylene fragments; a feature demonstrated by the thermal rearrangement' of 2-methyl-l,4-pentadienel-d, achieved in chlorobenzene for 2 h at 25°C (NiH represents the Ni hydride complex shown above):

276

14.5. Olefin Transformations 14.5.2. Olefin Dimerizationand Oligomerization 14.5.2.1. Introduction

%D Ni

-+

x D + N i H

(b)

Rearrangement of 2,3-dimethyl- 1,Cpentadiene to its isomer [equation (a)] can be rationalized by assuming fragmentation into dimethallyl and ethylene fragments followed by recombination:

A further example of skeletal rearrangement is provided by the equilibrium rearrangement3 of acylcobalt complexes: (CH~)~CHC(O)CO(CO)~ FCH~CH~CH~C(O)CO(CO)~

(4

The mechanism4 involves the elimination of and readdition of HCo(CO), as shown in Scheme 1. Starting with either acyl isomer, a 50% yield of the mixed aldehydes is obtained in addition to the mixture of isomeric acylcobalt compounds. Formation of aldehyde and olefin from the acylcobalt carbonyl is thus a disproportionation reaction that competes with the skeletal rearrangement.

(M.ORCHIN)

1. R. G. Miller, P. A. Pinke, R. D.Stauffer, H. J. Golden, D.J. Baker, J . Am. Chem. SOC.,96,421 1

(1974). 2. H. J. Golden, D. J. Baker, R. G . Miller, J . Am. Chem. SOC., 96, 4235 (1974). 3. Y. Takegami, C. Yokokawa, H. Masada, Y. Okuda, Bull. Chem. SOC. Jpn., 37, 1190 (1964); Y. Takegami, Y. Watanabe, H. Masada, T. Mitsudo, Bull. Chem. SOC.Jpn., 42,206 (1969). 4. W. Rupilius, M. Orchin, J . Org. Chem., 37,936 (1972).

14.5.2. Olefin Dimerizationand Oligomerization 14.5.2.1. Introduction

Catalytic dimerization, oligomerization, and cooligomerizati n of monoolefins and diolefins are important recent areas of research both in academe and in industry. Investigation of these reactions has greatly promoted development of homogeneous catalysis. Valuable intermediates for important industrial processes such as hydroformylation, alkylation, methathesis, etc. may be obtained by oligomerization of low-molecular-weight olefins and diolefins, which are produced in large quantities by petroleum refining processes. Propylene and butylene are selectively dimerized or codimerized to either a hexene


276

14.5. Olefin Transformations 14.5.2. Olefin Dimerizationand Oligomerization 14.5.2.1. Introduction

%D Ni

-+

x D + N i H

(b)

Rearrangement of 2,3-dimethyl- 1,Cpentadiene to its isomer [equation (a)] can be rationalized by assuming fragmentation into dimethallyl and ethylene fragments followed by recombination:

A further example of skeletal rearrangement is provided by the equilibrium rearrangement3 of acylcobalt complexes: (CH~)~CHC(O)CO(CO)~ FCH~CH~CH~C(O)CO(CO)~

(4

The mechanism4 involves the elimination of and readdition of HCo(CO), as shown in Scheme 1. Starting with either acyl isomer, a 50% yield of the mixed aldehydes is obtained in addition to the mixture of isomeric acylcobalt compounds. Formation of aldehyde and olefin from the acylcobalt carbonyl is thus a disproportionation reaction that competes with the skeletal rearrangement.

(M.ORCHIN)

1. R. G. Miller, P. A. Pinke, R. D.Stauffer, H. J. Golden, D.J. Baker, J . Am. Chem. SOC.,96,421 1

(1974). 2. H. J. Golden, D. J. Baker, R. G . Miller, J . Am. Chem. SOC., 96, 4235 (1974). 3. Y. Takegami, C. Yokokawa, H. Masada, Y. Okuda, Bull. Chem. SOC. Jpn., 37, 1190 (1964); Y. Takegami, Y. Watanabe, H. Masada, T. Mitsudo, Bull. Chem. SOC.Jpn., 42,206 (1969). 4. W. Rupilius, M. Orchin, J . Org. Chem., 37,936 (1972).

14.5.2. Olefin Dimerizationand Oligomerization 14.5.2.1. Introduction

Catalytic dimerization, oligomerization, and cooligomerizati n of monoolefins and diolefins are important recent areas of research both in academe and in industry. Investigation of these reactions has greatly promoted development of homogeneous catalysis. Valuable intermediates for important industrial processes such as hydroformylation, alkylation, methathesis, etc. may be obtained by oligomerization of low-molecular-weight olefins and diolefins, which are produced in large quantities by petroleum refining processes. Propylene and butylene are selectively dimerized or codimerized to either a hexene

277

14.5. Olefin Transformations 14.5.2. Olefin Dimerization and Oligomerization 14.5.2.1. Introduction

C

c- c- c- co-

1kC0

c- c- c- co-

11 1lC0

Co(cO),

I

c-c-co-co(co),

1

2

II

c l\-co I c-c- co- Co(cO),

3

Co(cO),

I I

5

c-c-c-co(co),

L o r3

0

II

6

7

YOr4 c o I

C-C-C-H

+

c-c=c

1

C-C-Co(CO),

C-CTC HCo(CO),

C-C-C-C-H

It

4

1I

+

Scheme 1.

c-c=c

gasoline blending component improving the octane number or to a heptene oxoalcohol feed stock by the IFP Dimersol process’.*. Oligomerization of monoolefins as part of the “Shell Higher Olefin Process affords higher linear a-olefins, which are useful intermediates for detergents, lubricating oils, and plasticizers. Further industrial syntheses with considerable significance are the production of 1,Chexadiene by cooligomerization of C,H, with butadiene and of 1,5-~yclooctadieneor 1,5,9-~yclododecatriene by cyclodimerization or cyclotrimerization, respectively, of butadiene7. Olefins and diolefins are readily dimerized or oligomerized by many catalysts involving almost all of the transition metals. Monoolefins normally afford open-chain linear or branched dimers or oligomers. Cyclodimers or cyclooligomers are formed from monoolefins only in some special cases. The oligomerization of 1,3-diolefins may lead to a large diversity of open-chain or cyclic dimers or oligomers, which are sometimes formed together. Using specific catalysts it is possible to a large extent to direct the course and to control the selectivity of these reactions. In addition to dimerization and oligomerization, many metal-catalyzed cooligomerizations are known, which can occur either between two different monoolefins, two different diolefins, or between a monoolefin and a diolefin. These reactions, which may lead to various open-chair or cyclic codimers or cooligomers, are not treated in detail in this section.

278

8

14.5.2. Olefin Dimerizationand Oli omerization 14.5.2.2. Linear Dimerization and ligomerizationof Monoolefins 14.5.2.2.1. Mechanistic Aspects.

Several review articles have appeared recently that give extensive information either on the whole subject or on some aspects of this field8-I4 including cooligomerization reactionslO*l‘,I4. (W. KAMINSKY, R. KRAMOLOWSKY)

1. Y. Chauvin, J. F. Gaillard, D. V. Quang, J. W. Andrews, Chem. Ind., 375 (1974). 2. Y. Chauvin, J. F. Gaillard, J. Leonard, P. Bonnifay, J. W. Andrews, Hydrocarbon Process. 110 (1982). 3. W. Keim, Angew. Chem. Int. Ed. Engl., 29,235 (1990). 4. M. Peukert, W. Keim, Organometallics, 2, 594 (1983). 5 . M. Peukert, W. Keim, J. Mol. Catal., 22, 289 (1984) ref. 4-7. 6. E. R. Freitas, C. R. Gum, Chem. Eng. Prog., 75, 73 (1979). 7. G . Wilke,Angew. Chem. Int. Ed. Engl., 27, 185 (1988). 8. J. Skupihska, Chem. Rev., 91,613 (1991). 9. S. Muthukumaru Pillai, M. Ravindranathan, S. Sivaram, Chem. Rev., 86, 353 (1986). 10. W. Keim, A. Behr, M. Roper, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 11. P. W. Jolly, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, pp. 615, 664, 671. 12. B. Bogdanovik, Adv. Organomet. Chem., 17, 105 (1979). 13. J. Tsuji, Adv. Organomet. Chem., 17, 141 (1979). 14. A. C. L. Su, Adv. Organomet. Chem., 17, 269 (1979).

14.5.2.2. Linear Dimerizatlon and Oiigomerization of Monoolefins 14.5.2.2.1. Mechanistic Aspects.

Mainly two alternative mechanisms are discussed for the linear dimerization and oligomerization of monoolefins catalyzed by transition metal systems: an insertion-elimination mechanism via a metal hydride (alkyl) species’-5 and metallacycle The metal hydride mechanism is the most generally accepted one for the majority of oligomerization reactions. It has been assumed that the products are formed by way of “insertion” of the olefin into the M-H bond followed by one (dimerization) or several successive (oligomerization) alkyl migratory “insertions” into an M-C bond (propagation step) and finally @elimination reconstituting the metal hydride species. As an example, Scheme 1 shows this catalytic cycle (in simplified form) for the C,H4 dimerization. The metal hydride species may be generated in various ways, e.g., by oxidative addition of HX, by reduction of a metal halide with NaBH,, or by /3-elimination from a metal-alkyl complex. For an unsymmetrically substituted monoolefin dual modes of “insertion” into the M-H bond and an M-C bond are possible (Scheme 2). This can lead to various linear or branched dimers and oligomers. Regioselectivity in these reactions depends in a complicated manner on the metal, the ligands applied, and the reaction conditions. Furthermore, most of the transition metal systems also catalyze isomerization of olefins and therefore the primarily formed products may be converted to other isomers. Recently, other pathways where the C-C bond is postulated to form via metallacyclopentanes (Scheme 3) have been established for the catalytic dimerization of ethylene and higher a-olefins by zirconium6, tantalum7, and nickel8 complexes. At least for the tantalum systems, the mechanisms of conversion of the metallacyclopentane to the product olefin differ significantly with the specific catalyst. With some variations, the metal-


278

8

14.5.2. Olefin Dimerizationand Oli omerization 14.5.2.2. Linear Dimerization and ligomerizationof Monoolefins 14.5.2.2.1. Mechanistic Aspects.

Several review articles have appeared recently that give extensive information either on the whole subject or on some aspects of this field8-I4 including cooligomerization reactionslO*l‘,I4. (W. KAMINSKY, R. KRAMOLOWSKY)

1. Y. Chauvin, J. F. Gaillard, D. V. Quang, J. W. Andrews, Chem. Ind., 375 (1974). 2. Y. Chauvin, J. F. Gaillard, J. Leonard, P. Bonnifay, J. W. Andrews, Hydrocarbon Process. 110 (1982). 3. W. Keim, Angew. Chem. Int. Ed. Engl., 29,235 (1990). 4. M. Peukert, W. Keim, Organometallics, 2, 594 (1983). 5 . M. Peukert, W. Keim, J. Mol. Catal., 22, 289 (1984) ref. 4-7. 6. E. R. Freitas, C. R. Gum, Chem. Eng. Prog., 75, 73 (1979). 7. G . Wilke,Angew. Chem. Int. Ed. Engl., 27, 185 (1988). 8. J. Skupihska, Chem. Rev., 91,613 (1991). 9. S. Muthukumaru Pillai, M. Ravindranathan, S. Sivaram, Chem. Rev., 86, 353 (1986). 10. W. Keim, A. Behr, M. Roper, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 11. P. W. Jolly, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, pp. 615, 664, 671. 12. B. Bogdanovik, Adv. Organomet. Chem., 17, 105 (1979). 13. J. Tsuji, Adv. Organomet. Chem., 17, 141 (1979). 14. A. C. L. Su, Adv. Organomet. Chem., 17, 269 (1979).

14.5.2.2. Linear Dimerizatlon and Oiigomerization of Monoolefins 14.5.2.2.1. Mechanistic Aspects.

Mainly two alternative mechanisms are discussed for the linear dimerization and oligomerization of monoolefins catalyzed by transition metal systems: an insertion-elimination mechanism via a metal hydride (alkyl) species’-5 and metallacycle The metal hydride mechanism is the most generally accepted one for the majority of oligomerization reactions. It has been assumed that the products are formed by way of “insertion” of the olefin into the M-H bond followed by one (dimerization) or several successive (oligomerization) alkyl migratory “insertions” into an M-C bond (propagation step) and finally @elimination reconstituting the metal hydride species. As an example, Scheme 1 shows this catalytic cycle (in simplified form) for the C,H4 dimerization. The metal hydride species may be generated in various ways, e.g., by oxidative addition of HX, by reduction of a metal halide with NaBH,, or by /3-elimination from a metal-alkyl complex. For an unsymmetrically substituted monoolefin dual modes of “insertion” into the M-H bond and an M-C bond are possible (Scheme 2). This can lead to various linear or branched dimers and oligomers. Regioselectivity in these reactions depends in a complicated manner on the metal, the ligands applied, and the reaction conditions. Furthermore, most of the transition metal systems also catalyze isomerization of olefins and therefore the primarily formed products may be converted to other isomers. Recently, other pathways where the C-C bond is postulated to form via metallacyclopentanes (Scheme 3) have been established for the catalytic dimerization of ethylene and higher a-olefins by zirconium6, tantalum7, and nickel8 complexes. At least for the tantalum systems, the mechanisms of conversion of the metallacyclopentane to the product olefin differ significantly with the specific catalyst. With some variations, the metal-

8

279

14.5.2. Olefin Dimerization and Oli omerization 14.5.2.2. Linear Dimerization and ligomerization of Monoolefins 14.5.2.2.1. Mechanistic Aspects.

Scheme 1.

R

I 1 + (M-C-C--) I 1 H

+ \c=c /

R

\

I;'

+ (M-C-C--)

I

H

I I

1 '

'c=c /

l l

+ (M-C-C-C--)

+

/

R

I (M-C--J I

l l

/ \

R

I ; ' l I l l

+ (M-C-C-C--J

Scheme 2.

HZC =CHCZH, Scheme 3. lacyclopentane mechanism may well be a valid alternative to the hydride mechanism7. Several other mechanisms have been p o s t ~ l a t e d ~ - ' ~ . (W. KAMINSKY, R. KRAMOLOWSKY) 1. G. Henrici-OlivC, S.OlivC, Coordination and Carulysis, Verlag Chemie, Weinheim, New York,

1977, p. 186.

280

14.5.2. Olefin Dirnerizationand Oli ornerization 14.5.2.2. Linear Dimerization and 81i omerization of Monoolefins 14.5.2.2.2. By Group 4 and 5 Metal 8atalysts.

2. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Fink, Principles and Applications of Organotransition Metal Chemistry, University Science Books, Mill Valley, CA, 1987, p. 578. 3. U.Miiller, W. Keim, C. Kruger, P. Betz, Angew. Chem. Int. Ed. Engl., 28, 1011 (1989). 4. N. M. Doherty, J. E. Bercaw, J . Am. Chem. SOC., 107,2670 (1985). 5 . H. Lehmkuhl, Pure Appl. Chem., 58,495 (1986). 6. S. Datta, M. B. Fischer, S. S . Wreford, J . Organomet. Chem., 188, 353 (1980). 7. J. D. Fellmann, R. R. Schrock, G. A. Rupprecht, J . Am. Chem. SOC.,103,5752 (1981). 8. R. H. Grubbs, A. Miyashita, J . Am. Chem. SOC., 100,7416 (1978). 9. J. Skupidska, Chem. Rev., 91, 613 (1991). 10. S. Muthukumaru Pillai, M. Ravindranathan, S. Sivaram, Chem. Rev.,86, 353 (1986). 11. H. W. Turner, R. R. Schrock,J. Am. Chem. SOC., 104,2331 (1982). 12. Ph. 0. Nubel, Th. L. Brown, J . Am. Chem. Soc., 106, 3474 (1984). 14.5.2.2.2. By Group 4 and 5 Metal Catalysts.

Apart from nickel complexes (see 14.2.2.3), titanium compounds are the most often used catalysts for olefin oligomerizations. Catalysts composed of either a Ti(1V) or a Zr(1V) compound, which is activated by an organoaluminum compound, (Ziegler-Natta type) are commonly used for olefin polymerization. The same type of catalysts can yield dimers or oligomers from ethylene and higher a-olefins by changes in the ligands, the ratio of the components, and the reaction conditions'. Comprehensive lists of catalyst systems and the products formed in these reactions are compiled in refs. 2 and 3. Selected examples are given in Table 1. Ethylene is polymerized by Ti(OR),/AIR, at a molar ratio AIR,/Ti(OR),>20, whereas dimers are formed at a molar ratio AIR,/Ti(OR),< 10'. Dimerization selectivity may be quite high (>90%); the dimer fraction consists exclusively of 1-butene in acTABLE1. SOMETYPICAL GROUP4 AND 5 METALCATALYSTS FOR LINEAR OLICOMERIZATION OF MONOOLEFINS Olefin Ethylene

Catalyst System M(OR),/AIEt," TiCl,/Et,AI,CI, TiCl4/A1Cl,Et3 - ,,/t-BuOH TiCl(OEt),/AICl,Et ZrCl4/AlClEb(Al/Zr-4) [Zr(ri'-C4H,),(dmpe)Ib

I("C)

Main Products

Ref.

60-90 10 - 20 - 20 120 70

1-Butene (60-99%) C,-C,, a-olefins C,-C,, a-olefins C,-C,, a-olefins C,,-C,, a-olefins H,C=CEt, H,C=CHCH(Me)Et 1-Butene

1, 3

25 Propylene RCH=CH,d R'CH=CH,e

Ti(OR)jkEt, TiCl,/AlEt,/RCl TiCl,Me/AlCl,Me

60-90 40 - 70

[Ta(~5-C5Me5)C12(C,Hlo)l 50

'M = Ti,Zr. bdmpe, 1,2-bis(dirnethylphosphino)ethane. 'n = 0-2;R = Et,CH,CMe,. dR = Et,Pr,Bu,Pent,Hex. 'R' = H, Me, Pr,Hex, CH,CMe,, CH,CHMe2.

+

Methylpentenes C, oligomers (65%) Oligomers of the type H,C=CRR' Head-to-head and tail-to-tail dimers

2

4

5 2 6

7 1

2

1

6


280

14.5.2. Olefin Dirnerizationand Oli ornerization 14.5.2.2. Linear Dimerization and 81i omerization of Monoolefins 14.5.2.2.2. By Group 4 and 5 Metal 8atalysts.

2. J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Fink, Principles and Applications of Organotransition Metal Chemistry, University Science Books, Mill Valley, CA, 1987, p. 578. 3. U.Miiller, W. Keim, C. Kruger, P. Betz, Angew. Chem. Int. Ed. Engl., 28, 1011 (1989). 4. N. M. Doherty, J. E. Bercaw, J . Am. Chem. SOC., 107,2670 (1985). 5 . H. Lehmkuhl, Pure Appl. Chem., 58,495 (1986). 6. S. Datta, M. B. Fischer, S. S . Wreford, J . Organomet. Chem., 188, 353 (1980). 7. J. D. Fellmann, R. R. Schrock, G. A. Rupprecht, J . Am. Chem. SOC.,103,5752 (1981). 8. R. H. Grubbs, A. Miyashita, J . Am. Chem. SOC., 100,7416 (1978). 9. J. Skupidska, Chem. Rev., 91, 613 (1991). 10. S. Muthukumaru Pillai, M. Ravindranathan, S. Sivaram, Chem. Rev.,86, 353 (1986). 11. H. W. Turner, R. R. Schrock,J. Am. Chem. SOC., 104,2331 (1982). 12. Ph. 0. Nubel, Th. L. Brown, J . Am. Chem. Soc., 106, 3474 (1984). 14.5.2.2.2. By Group 4 and 5 Metal Catalysts.

Apart from nickel complexes (see 14.2.2.3), titanium compounds are the most often used catalysts for olefin oligomerizations. Catalysts composed of either a Ti(1V) or a Zr(1V) compound, which is activated by an organoaluminum compound, (Ziegler-Natta type) are commonly used for olefin polymerization. The same type of catalysts can yield dimers or oligomers from ethylene and higher a-olefins by changes in the ligands, the ratio of the components, and the reaction conditions'. Comprehensive lists of catalyst systems and the products formed in these reactions are compiled in refs. 2 and 3. Selected examples are given in Table 1. Ethylene is polymerized by Ti(OR),/AIR, at a molar ratio AIR,/Ti(OR),>20, whereas dimers are formed at a molar ratio AIR,/Ti(OR),< 10'. Dimerization selectivity may be quite high (>90%); the dimer fraction consists exclusively of 1-butene in acTABLE1. SOMETYPICAL GROUP4 AND 5 METALCATALYSTS FOR LINEAR OLICOMERIZATION OF MONOOLEFINS Olefin Ethylene

Catalyst System M(OR),/AIEt," TiCl,/Et,AI,CI, TiCl4/A1Cl,Et3 - ,,/t-BuOH TiCl(OEt),/AICl,Et ZrCl4/AlClEb(Al/Zr-4) [Zr(ri'-C4H,),(dmpe)Ib

I("C)

Main Products

Ref.

60-90 10 - 20 - 20 120 70

1-Butene (60-99%) C,-C,, a-olefins C,-C,, a-olefins C,-C,, a-olefins C,,-C,, a-olefins H,C=CEt, H,C=CHCH(Me)Et 1-Butene

1, 3

25 Propylene RCH=CH,d R'CH=CH,e

Ti(OR)jkEt, TiCl,/AlEt,/RCl TiCl,Me/AlCl,Me

60-90 40 - 70

[Ta(~5-C5Me5)C12(C,Hlo)l 50

'M = Ti,Zr. bdmpe, 1,2-bis(dirnethylphosphino)ethane. 'n = 0-2;R = Et,CH,CMe,. dR = Et,Pr,Bu,Pent,Hex. 'R' = H, Me, Pr,Hex, CH,CMe,, CH,CHMe2.

+

Methylpentenes C, oligomers (65%) Oligomers of the type H,C=CRR' Head-to-head and tail-to-tail dimers

2

4

5 2 6

7 1

2

1

6

14.5.2. Olefin Dirnerization and Oli ornerization 14.5.2.2. Linear Dimerization and 8li ornerization of Monoolefins 14.5.2.2.2. By Group 4 and 5 Metal tatalysts.

28 1

cordance with the low isomerization activity of Ti complexes. The influence of reaction conditions on ethylene oligomerization by the systems TiC14/A1C1, - ,Et, has been thoroughly investigated2. Catalyst efficiency is improved at Al/Ti ratios >2. Increase in the reaction P increases the molecular weight of the oligomers and inhibits cooligomerization of ethylene with the oligomers formed. An increase in reaction T increases the molecular weight of the oligomers and the amount of branched a-olefins. Donor ligands such as ketones, amines, nitriles, or phosphines added to this system increase the catalyst selectivity to linear a-olefins2. By variations in the composition of this system and in the reaction conditions, the oligomerization of ethylene may also be directed to yield either linear C4-C30 a-olefins4 or branched C4-C2, olefins with vinylidene groups‘ as main products. Propene and higher a-olefins also may be dimerized or oligomerized by these catalysts. Generally, reactivity is much lower than that of ethylene and decreases in the order ethylene >> propylene > l-butene > l-hexene > l-octene > l-decene2. Also the selectivity is lower and mainly branched dimers or oligomers are formed’. By a chiral zirconocene such as (S)-[1,1’-ethylenebis(4,5,6,7-tetrahydro-l-indenyl)zirconium-bis[O-acetyl-(R)-mandelate](1) activated by methylaluminoxane (2), propene, or 1-butene is asymmetrically oligomerized to optically active oligomers’.

2

1 The established mechanism for the oligomerizations catalyzed by Ziegler-Natta type Ti(Zr) systems involves formation of a C-C bond by “insertion” of the olefin into a metal-alkyl bond. Quite different “metallacycle mechanisms” have been established for olefin dimerizations catalyzed by low-valent Ti, Zr6, and Ta7q9 systems. Thus, C,H4 is dimerized with excellent selectivity to l-butene by [Ta(CHCMe,),, (H2C=CH2), - ,R,(PMe3),] (n=0-2) via a tantalacyclopentane complexg. The complexes [Ta( q5- C,Me,)Cl,(olefin)] selectively dimerize ethylene to l-butene and also higher a-olefins to a mixture of the tail-to-tail (tt) and the head-to-tail (ht) dimer. This is rationalized by the occurrence of the two metallacycles 3a and 3b (Scheme 1). With increasing size of R the product composition switches over from virtually all tt dimer (98% when R = Me) to exclusively ht dimer (100% when R = CH2CMe,)9. (W. KAMINSKY, R. KRAMOLOWSKY) 1. G. Lefebvre, Y. Chauvin, in Aspects of Homogeneous Catalysis, Vol. 1, R. Ugo, ed., C. Man-

fredi, Milan, 1970, p. 108. 2. J. Skupifiska, Chem. Rev., 91,613 (1991).

282

14.5.2. Olefin Dimerization and Oli omerization 14.5.2.2. Linear Dirnerization and dligomerization of Monoolefins 14.5.2.2.3. By Nickel Catalysts.

3a

R

3b Scheme 1. 3. S. Muthukumaru Pillai, M. Ravindranathan, S. Sivaram, Chem. Rev., 86, 353 (1986). 4. A. W. Langer, Jr., J. Macromol. Sci. (A), 4,775 (1970). 5. G. Henrici-Olivt, S . Olivt, Angew. Chem. Int. Ed. Engl., 9,243 (1970). 6. S . Datta, M. B. Fischer, S. S . Wreford, J. Organomet. Chem., 188, 353 (1980). 7. J. D. Fellmann, R. R. Schrock, G. A. Rupprecht, J. Am. Chem. SOC.,103, 5752 (1981). 8. W. Kaminsky, A. Ahlers, N. Moller-Lindenhof,Angew. Chem. Int. Ed. Engl., 28, 1216 (1989). 9. S. J. McLain, J. Sancho, R. R. Schrock, J. Am. Chem. SOC.,102,5610 (1980).

14.5.2.2.3. By Nickel Catalysts.

Nickel complexes are by far the most numerous and the most extensively studied’-6 of the catalysts for oligomerization of monoolefins. Particular interest in Ni catalysts stems from their usually very high activity and the possibility to direct the course and to control the selectivity of the reactions by “tailoring the catalyst”’. Extensive lists of catalysts and products formed in these reactions are compiled in refs. 5 and 6. Some selected examples of catalyst systems are given in Table 1. The catalyst systems may roughly be divided into two groups: (1) Ziegler-Natta type catalysts composed of a Ni(I1) salt or complex activated by a Lewis acid (frequently an organoaluminum compound) or a Bronsted acid and often modified by a P-donor [PR, or P(OR),] and (2) one-component systems, most of them composed of a Ni(I1) chelate complex with a Ni-C bond that are active without a Lewis acid or reducing agent as cocatalyst. As the nickel component of the Ziegler-Natta type systems, Ni(I1) carboxylates, P-diketonates, P-dithiodiketonates, and phosphine complexes are frequently used, but organonickel compounds [NiXRL] or Ni(0) derivatives have also been thoroughly The neutral q3-allyl complexes2 [(Nix(T/~-C,H,)}~] and [NiX(q3-C,H,)(PR,)] are extremely active catalysts in the presence of Lewis acids. Also, cationic q3-allyl complexes [Ni( q3-C3Hs)L2]+PF6 (L=P-donor) have been tested’s8. Many experimental results strongly indicate the presence of a common active species which is probably a short lived nickel h ~ d r i d e l - The ~ . precise way in which this species is formed depends on the nature of the catalyst precursor. Circumstantial evidence has led to the square-planar structures 1 and 2a or 2b (PR, cis or trans to H) for the Ni-H species of donor-free and donor-modified systems, respectively’.


282

14.5.2. Olefin Dimerization and Oli omerization 14.5.2.2. Linear Dirnerization and dligomerization of Monoolefins 14.5.2.2.3. By Nickel Catalysts.

3a

R

3b Scheme 1. 3. S. Muthukumaru Pillai, M. Ravindranathan, S. Sivaram, Chem. Rev., 86, 353 (1986). 4. A. W. Langer, Jr., J. Macromol. Sci. (A), 4,775 (1970). 5. G. Henrici-Olivt, S . Olivt, Angew. Chem. Int. Ed. Engl., 9,243 (1970). 6. S . Datta, M. B. Fischer, S. S . Wreford, J. Organomet. Chem., 188, 353 (1980). 7. J. D. Fellmann, R. R. Schrock, G. A. Rupprecht, J. Am. Chem. SOC.,103, 5752 (1981). 8. W. Kaminsky, A. Ahlers, N. Moller-Lindenhof,Angew. Chem. Int. Ed. Engl., 28, 1216 (1989). 9. S. J. McLain, J. Sancho, R. R. Schrock, J. Am. Chem. SOC.,102,5610 (1980).


Nickel complexes are by far the most numerous and the most extensively studied’-6 of the catalysts for oligomerization of monoolefins. Particular interest in Ni catalysts stems from their usually very high activity and the possibility to direct the course and to control the selectivity of the reactions by “tailoring the catalyst”’. Extensive lists of catalysts and products formed in these reactions are compiled in refs. 5 and 6. Some selected examples of catalyst systems are given in Table 1. The catalyst systems may roughly be divided into two groups: (1) Ziegler-Natta type catalysts composed of a Ni(I1) salt or complex activated by a Lewis acid (frequently an organoaluminum compound) or a Bronsted acid and often modified by a P-donor [PR, or P(OR),] and (2) one-component systems, most of them composed of a Ni(I1) chelate complex with a Ni-C bond that are active without a Lewis acid or reducing agent as cocatalyst. As the nickel component of the Ziegler-Natta type systems, Ni(I1) carboxylates, P-diketonates, P-dithiodiketonates, and phosphine complexes are frequently used, but organonickel compounds [NiXRL] or Ni(0) derivatives have also been thoroughly The neutral q3-allyl complexes2 [(Nix(T/~-C,H,)}~] and [NiX(q3-C,H,)(PR,)] are extremely active catalysts in the presence of Lewis acids. Also, cationic q3-allyl complexes [Ni( q3-C3Hs)L2]+PF6 (L=P-donor) have been tested’s8. Many experimental results strongly indicate the presence of a common active species which is probably a short lived nickel h ~ d r i d e l - The ~ . precise way in which this species is formed depends on the nature of the catalyst precursor. Circumstantial evidence has led to the square-planar structures 1 and 2a or 2b (PR, cis or trans to H) for the Ni-H species of donor-free and donor-modified systems, respectively’.

14.5.2. Olefin Dimerization and Oli omerization 14.5.2.2. Linear Dimerization and 8ligomerization of Monoolefins 14.5.2.2.3. By Nickel Catalysts.

283

TABLE1. SOME TYPICALNICKELCATALYSTS FOR OLIGOMERIZATION OF MONOOLEFINS Olefin Ethylene

T (“C)

Catalyst System Nix, or [Ni(acac),]/AIClEt, or A1,C13Et3/PR3 “ix(.r73-c3H,)1(pR3)1/ A1 C1 Et “iX(Ar)(PR3),l/BF3 “i(.r75-C,H5)z1 “i{ CH,(CH,), t (PPh3)31 NiX,/NaBH,/POa [Ni(acac),]/Al(OEt)Et, [Ni(acac),]/AlClEt, or Al,C13Et3/PR3 “ix(.r73-c3H,)(pR3)I/ A1,Cl3ET3 [NiCl2(PBu3),]/A1Cl,Et [Ni(cod),]/OOd

- 20

,

Propylene

1-Butene H,C=CHR

0 200 0 75 40 0-30

- 20 20 70

[Ni(l:4-5-q3-C,H,,)(hfacac)]

40

[Ni(acac),]/Al(OEt)EtJ SnEt,

Main Products 2-Butene

+ hexenes

2-Butene 2-Butene + hexenes 1-Butene 1-Butene + cyclobutane C,-C,, linear a-olefins Linear hexenes Methylpentenesb or 2.3dimethylbutenes‘ methylpentenesb or 2.3dimethylbutenes‘ Isooctenes Highly linear octenes Highly linear dimers and trimers Mainly linear dimers

Ref.

1-5 1-4

1-4 3 3,4 3 3, 4 1-4 1-4 3 3, 5 3 3

‘PO = R,PCH,CO,-, R,PC(O)NPh-.

bFor R = Ph. CForR = i-F’r, Cy. dOO = hfa, hfacac.

1 2a 2b The Ni-catalyzed reactions of monoolefins generally yield dimer/oligomer mixtures; the primary products frequently are further isomerized. However, the selectivity relative to dimerization versus oligomerization andtor formation of a particular dimer may be significantly influenced by reaction conditions, primarily by the choice of the PR, donor and the molar ratio Ni/PR,. As one-component systems, particularly the chelate Ni(I1) complexes of the general type 3-5 are very active and highly selective catalysts for olefin olig~merization~.’.

3

4

284

14.5.2. Olefin Dimerizationand Oli omerization 14.5.2.2. Linear Dimerization and 8ligomerization of Monoolefins 14.5.2.2.3.By Nickel Catalysts.

The activity of these catalysts and the geometric distribution of the olefins formed can be modified by adding phosphines. Also for this type of catalysts, Ni-H species have been formulated as common active specie^^*^,^. Homogeneous catalysts dissolved in halogenated or aromatic hydrocarbons are most often used; however supported systems obtained by depositing a nickel species on silica, alumina, or a polymer containing functional groups have also been developed3-,. 1. Oligomerization of Ethylene. The most active catalysts for ethylene oligomerization are [ { NiX(q3-C,H,)),]/AlCl,Et,-,/PR, systems'-4. Phosphine-free catalysts of this type convert H2C=CH, in C6H,Cl at - 20°C mainly to dimers with high proportions of 2-butenes. With PR,-modified systems the degree of dimerization increases with decreasing steric demand of PR,. Thus, with PMe, dimers are formed in high yield (>98%) whereas with Pr-Bu, (ratio Ni:Pt-Bu3=l:4) polyethylene is obtained quantitatively. Quite long-lived catalysts exhibiting enzyme-like activities are derived from the air-stable dithio nickel(I1) chelates'O'l' 6 activated by Et2AlC1. R'

i 3

6 Compounds 3-5 are very effective for ethylene oligomerization, they have activities up to 10,000 mol ethylene/mol catalyst3. The PO chelate complexes 4 and 5 and related systems are effective in the industrial SHOP (Shell Higher Olefin P r o ~ e s s ) ' ~ reaction. -'~ In this process, oligomerization of ethylene can be directed to yield C,-C,, oligomers, which exhibit high linearity (>99%) and an a-olefin content >95%. Activating the sulfonated complex (5; R3=Ph, R4=S03Na) with aluminum alkoxides AlEt,-,(OR), (n=l-3) significantly increases the oligomerization activity by a factor 20- 1O0l6. The nickelacyclopentane [Ni{CH,(CH,),CH, ](PPh3)3] is an interesting catalyst since it selectively dimerizes CpH4 to a mixture of 1-butene and cyclobutane'. The complex [Ni($-C,H,),] dimerizes C2H4mainly to 1-butene,.

2. Oligomerization of Propylene. In principle, the same types of catalysts are used for propylene oligomerization as described for C2H4 o l i g ~ m e r i z a t i o n ~ . ~ ~ ~ . Propylene is mainly dimerized to a mixture of isomeric C,-olefins in varying ratios depending on the catalyst systems (Table l)'-,. As is shown in Scheme 1, for propylene as an unsymmetric olefin the possibility of dual migration (addition via N i - C , or Ni-C,) exists conceivably both for the first and for the second step of the dimerizations. Hence, six primary products (neglecting cisltrans isomers) may be formed in principle, which can further isomerize under favorable reaction conditions2. However, the regioselectivity of the propylene dimerization can be effectively controlled by proper choice of the Ni complex and/or the PR, donor. Thus, high amounts of n-hexenes (up to 80% in the dimer fraction) may be obtained by catalysts with NO or 00 chelating ligands, e.g., [Ni(acac>,], combined with electron-withdrawing systems such as Al(0Et)EtZ. Addition of PR, generally increases activity of the catalyst and inhibits formation of oligomers higher than dimers. The composition of the dimer mixture is greatly influenced

8

14.5.2. Olefin Dirnerization and Oli ornerization 14.5.2.2. Linear Dirnerization and ligornerization of Monoolefins 14.5.2.2.3. By Nickel Catalysts.

u uI uI uI I v, u uII sx

u uI uI I U II p uI

t X

I

g I

-u

I

U

U

I

uI uI

U

W--u

I W

-+

I

u I U-U

6

I

u

W

uI

I II u

U

II

V

+

u I U II W

+

285

286

14.5.2. Olefin Dirnerization and Oli ornerization 14.5.2.2. Linear Dirnerization and 8ligomerization of Monoolefins 14.5.2.2.3. By Nickel Catalysts.

by the particular PR, ligand, the directing effect of which can be attributed mainly to steric reasons. Table 2 shows that the phosphines exert their directing influence during the second step of addition. With increasing steric demand of PR,, the mode of insertion changes from predominantly Ni+C, type addition to almost exclusive N i - C , type addition. The first step likely is not influenced by phosphines; the main product is the isopropyl species formed by Ni+C, type addition. Only with the very bulky phosphine P(i-Pr)(t-Bu),, is the direction of this addition significantly altered to favor the sterically less-demanding n-propyl species.

3. Higher a-Olefins. The rate of oligomerization of monoolefins generally decreases in the 0rdeS9~ethylene >> propylene > 1-butene > 1-hexene > cycloalkene > M e H C X H M e . However, most catalyst systems give mainly branched oligomers. Catalysts of type I containing 00 chelate ligands are effective catalysts for the linear oligomerization of higher a-olefins. Using (3;R'=R2=CF,), 1-butene can be dimerized in 82% yield to linear o c t e n e ~ ' The ~ . bulky olefin neohexene can be converted selectively into linear dimersI8 with the same catalyst activated by BF,.

4. Asymmetric Codimerizations. The use of Ni catalysts modified by an optically active phosphine in catalytic asymmetric syntheses, e.g., of the type [ { NiX(q3-C3H5)J2]/ Al,Cl,Et,/PR,, is imp~rtant'~. Phosphines bearing the optical activity in the substituents such as 7 are most effective. Particularly high optical yields occur in the codimerization of C2H, with a strained cyclic olefin such as norbornene [equation (a)]:

7

TABLE2. EFFECTOF TERTIARY PHOSPHINE LIGANDS ON THE DIMERIZATION OF P R O P Y L E N E ~ ~ ~ Tertiary Phosphine

-

PPh, PMe, P(n-Bu), P(i-Pr),(t-Bu) P(i-Pr)(t-Bu),

Zfiexenes

methylpentenes

(%)

(%)

19.8 21.6 9.9 7.1 0.1 0.6

76.0 73.9 80.3 69.6 19.0 70.1

~.3-Dimethylbutenes 4.2 4.5 9.8 23.3 80.9 29.1

Direction of addition' P 9 25:75 15:85 18:82 15:85 69:31

'From Ref. 2. bCatalyst system: [I NiX(q3-C3Hs)t,I/AI,CI,Et,/PR,; - 20°C. 'p = %Ni+C,/%Ni+C, for the first step: q = %Ni+CI/%Ni+C2 for the second step: p and q are calculated from the product distribution.

18:82 1585 34:66 97:3 98.2

14.5.2. Olefin Dimerization and OH omerization 14.5.2.2. Linear Dimerization and 8ligornerization of Monoolefins 14.5.2.2.4. By Other Group 8-1 0 Metal Catalysts.

287

Thus, in the presence of 7, an optical purity of 30% at 10°C has been obtained for 8. The optical purity increases significantly on lowering the temperature, and a value of 80.6% is observed' at - 97°C. (W. KAMINSKY, R. KRAMOLOWSKY)

1. P. W. Jolly, G. Wilke, The Organic Chemistry of Nickel, Vol. 11, Academic Press, New York, 1975, p. 6. 2. B. Bogdanovi6,Adv. Organomet. Chem., 17, 105 (1979). 3. W. Keim, A. Behr, M. Rbper, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 4. P. W. Jolly, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G . A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 615. 5 . J. Skupihska, Chem. Rev., 91,613 (1991). 6. S . Muthukumaru Pillai, M. Ravindranathan, S . Sivaram, Chem. Rev., 86, 353 (1986). 7. J.4. Brunet, A. Sivade, I. Tkatchenko,J . Mol. Catal., 50, 291 (1989). 8. E. Balbolov, M. Mitkova, K. Kurtev, J.-P. Gehrke, R. Taube, J. Organomet. Chem., 352, 247 (1988). 9. U. Miiller, W. Keim, C. Kriiger, P. Betz, Angew. Chem. Int. Ed. Engl., 28, 1011 (1989). 10. St. J. Brown, A. F. Masters, J . Organomet. Chem., 367, 371 (1989). 11. K. J. Cavell, A. F. Masters, J . Chem. Res. (S),72 (1983). 12. W. Keim, Angew. Chem. Int. Ed. Engl., 29,235 (1990). 13. M. Peukert, W. Keim, Organometallics, 2,594 (1983). 14. M. Peukert, W. Keim, J . Mol. Catal., 22, 289 (1984) ref. 4-7. 15. E. R. Freitas, C. R. Gum, Chem. Eng. Prog., 75.73 (1979). 16. Y. V. Kissin, J . Polym. Sci., Part A: Polym. Chem., 27, 147 (1989). 17. D. L. Beach, J. E. Bozik, Ch.-Y. Wu, Y. V. Kissin, J . Mol. Catal., 34, 345 (1986). 18. A. Behr, V. Falbe, U. Freudenberg, W. Keim, Isr. J. Chem., 27, 277 (1986). 19. B. BogdanoviC, B. Henc, A. Lbsler, B. Meister, H. Pauling, G. Wilke, Angew. Chem. Int. Ed. Engl., 12,954 (1973). 14.5.2.2.4. By Other Group 8-10 Metal Catalysts.

All other group 8-10 metals have been shown to constitute catalytic systems for oligomerization of monoolefins, although these catalysts have not quite reached the importance of nickel catalysts in this field'-5. Some typical examples are given in Table 1.

1. Rhodium Catalysts. The catalytical behavior of Rh complexes in oligomerizations has been extensively studied'-3. A wide variety of olefins such as a-olefins, chlorinated olefins, styrene, and acrylates (but not acrylonitrile) are dimerized by RhC1,.3H20. The reactions are camed out at T from 20 to 200°C and are generally not very dependent on the solvent, although small amounts of alcohols lead to increased activity'. A detailed insertion mechanism based on an anionic Rh(II1)-H-intermediate has been suggested3 but a carbene mechanism" and a metallacycle pathway" have also been discussed. A general characteristic of Rh catalysts is their considerably greater activity for codimerization than for homodimerization' . Thus, C2H4 and butadiene are codimerized with excellent selectivity (up to 80%) to trans- 1,Chexadiene using RhC13.3H20 in C2H,0H as precatalyst12. The high selectivity toward 1,4-hexadiene has been explained with the formation of a relatively stable ~~-crotyl-Rh(III) species as the key intermediate. The cis to trans ratio depends on the nature of added donor ligands. The reaction forms the basis for the industrial production of trans-l,4-hexadiene, which is used as comonomer in ethylene-propylene-diene synthetic rubber.

2. Ruthenium Catalysts. The catalytic activity of Ru complexes for olefin oligomerization is considerably lower than that of Rh systems2S6.Therefore, reactions that are


14.5.2. Olefin Dimerization and OH omerization 14.5.2.2. Linear Dimerization and 8ligornerization of Monoolefins 14.5.2.2.4. By Other Group 8-1 0 Metal Catalysts.

287

Thus, in the presence of 7, an optical purity of 30% at 10°C has been obtained for 8. The optical purity increases significantly on lowering the temperature, and a value of 80.6% is observed' at - 97°C. (W. KAMINSKY, R. KRAMOLOWSKY)

1. P. W. Jolly, G. Wilke, The Organic Chemistry of Nickel, Vol. 11, Academic Press, New York, 1975, p. 6. 2. B. Bogdanovi6,Adv. Organomet. Chem., 17, 105 (1979). 3. W. Keim, A. Behr, M. Rbper, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 4. P. W. Jolly, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G . A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 615. 5 . J. Skupihska, Chem. Rev., 91,613 (1991). 6. S . Muthukumaru Pillai, M. Ravindranathan, S . Sivaram, Chem. Rev., 86, 353 (1986). 7. J.4. Brunet, A. Sivade, I. Tkatchenko,J . Mol. Catal., 50, 291 (1989). 8. E. Balbolov, M. Mitkova, K. Kurtev, J.-P. Gehrke, R. Taube, J. Organomet. Chem., 352, 247 (1988). 9. U. Miiller, W. Keim, C. Kriiger, P. Betz, Angew. Chem. Int. Ed. Engl., 28, 1011 (1989). 10. St. J. Brown, A. F. Masters, J . Organomet. Chem., 367, 371 (1989). 11. K. J. Cavell, A. F. Masters, J . Chem. Res. (S),72 (1983). 12. W. Keim, Angew. Chem. Int. Ed. Engl., 29,235 (1990). 13. M. Peukert, W. Keim, Organometallics, 2,594 (1983). 14. M. Peukert, W. Keim, J . Mol. Catal., 22, 289 (1984) ref. 4-7. 15. E. R. Freitas, C. R. Gum, Chem. Eng. Prog., 75.73 (1979). 16. Y. V. Kissin, J . Polym. Sci., Part A: Polym. Chem., 27, 147 (1989). 17. D. L. Beach, J. E. Bozik, Ch.-Y. Wu, Y. V. Kissin, J . Mol. Catal., 34, 345 (1986). 18. A. Behr, V. Falbe, U. Freudenberg, W. Keim, Isr. J. Chem., 27, 277 (1986). 19. B. BogdanoviC, B. Henc, A. Lbsler, B. Meister, H. Pauling, G. Wilke, Angew. Chem. Int. Ed. Engl., 12,954 (1973). 14.5.2.2.4. By Other Group 8-10 Metal Catalysts.

All other group 8-10 metals have been shown to constitute catalytic systems for oligomerization of monoolefins, although these catalysts have not quite reached the importance of nickel catalysts in this field'-5. Some typical examples are given in Table 1.

1. Rhodium Catalysts. The catalytical behavior of Rh complexes in oligomerizations has been extensively studied'-3. A wide variety of olefins such as a-olefins, chlorinated olefins, styrene, and acrylates (but not acrylonitrile) are dimerized by RhC1,.3H20. The reactions are camed out at T from 20 to 200°C and are generally not very dependent on the solvent, although small amounts of alcohols lead to increased activity'. A detailed insertion mechanism based on an anionic Rh(II1)-H-intermediate has been suggested3 but a carbene mechanism" and a metallacycle pathway" have also been discussed. A general characteristic of Rh catalysts is their considerably greater activity for codimerization than for homodimerization' . Thus, C2H4 and butadiene are codimerized with excellent selectivity (up to 80%) to trans- 1,Chexadiene using RhC13.3H20 in C2H,0H as precatalyst12. The high selectivity toward 1,4-hexadiene has been explained with the formation of a relatively stable ~~-crotyl-Rh(III) species as the key intermediate. The cis to trans ratio depends on the nature of added donor ligands. The reaction forms the basis for the industrial production of trans-l,4-hexadiene, which is used as comonomer in ethylene-propylene-diene synthetic rubber.

2. Ruthenium Catalysts. The catalytic activity of Ru complexes for olefin oligomerization is considerably lower than that of Rh systems2S6.Therefore, reactions that are

288

14.5.2. Olefin Dimerization and Oli omerization 14.5.2.2. Linear Dimerization and 8ligomerizationof Monoolefins 14.5.2.2.4. By Other Group 8-1 0 Metal Catalysts.

TABLE1. SOMEGROUP8-10 METALCATALYSTS (OTHER THANNi) FOR OLIGOMERIZATION OF MONOLEFINS Olefin Ethylene

Propylene

HzC=C(CH~)Z H2C=CHCN H,C=CHCO,R CH3CH=CHC0,R

Catalyst System

T ("C)

RhC13.3HZO [PdC12(PhCN)zIc [FeHZ(dmpe),ld

"cocatalysts: HCI, UAIH,, EtOH. bCocatalysts:HAlEh, HAl(octyl),; NR, = ~ - 0 , cCocatalysts: AgBF,, benzoquinone. LJ ddmpe,1,2-bis(dimethylphosphino)ethane.

Ref.

2-Butene + hexenes 12 2-Butene 12 2-Butene 1-3 2-Butene 12 Methylpentenes 12 2-Methyl- 1-pentene 6 Methylpentenes 12 n-hexenes Branched C9-C,, oligomers 5 50-100 n-Hexenes 12 9 20 Branched dimers I30 1,4-Dicyanobutenes l,2J propionitrile, adiponitrile 140 RO,CCH=CH(CHZ)2CO,R 1,2,7 50-80 R02CCH=CH(CHz)zCO2R 1,23 20/hv Dimers (95% yield) 9

[C~(acac),]/Al(OEt),Et~-~ 30 tCoH(Nz)(PPh3)31 25 RhC13~3H20/cocatalysta 50 PdCl,/cocatalyst" 50-100 [CoH(Nz)(PPh3)31 25 [Co(aca~),]/cod/Ph~PNR~~25 RhCly3HzO 50 [RuX(V~-C~H~)(CO),I [PdCl,(PhCN),] [WMeCN)&BFJz RuC1,.3H20/H2

Main Products

+

I

~3

usually carried out in an alcohol require more extreme conditions (100-150°C). The dimerization of acrylates is catalyzed by Ru(II1) compounds such as RuC13.3H,0 in the presence of CH30H7-'3. The Ru(0) complex [Ru(r76-C6H6)(H,C=CHC0,CH3)] activated by sodium naphthalenide selectively catalyzes the dimerization of methyl acrylate to dimethyl he~enedionate'~. In the presence of H,, RuCl, also dimerizes H,C=CHCN to a mixture of 1,6dicyanobutenes and adiponitril. An Ru-H intermediate is suggested by the requirement of H, for this reaction. 3. Palladium Catalysts. In general, Pd catalysts have considerably lower activities than Ni catalysts in olefin olig~merization~. Propylene is selectively dimerized to up to 95% linear hexenes by the catalyst [Pd(acac),]/EtzAIC1/PR, in 1,2-di~hloroethane~. The heterogeneous Pd(I1) catalyst system, prepared by treatment of sepiolite with [Pd(v3-C,H,),], exhibits high activity in the styrene dimerization to give exclusively (E)1,3-diphenyi-l - b ~ t e n e ' ~ . (W. KAMINSKY, R. KRAMOLOWSKY)

1. G. Lefebvre, Y. Chauvin, in Aspects of Homogeneous Catalysis, R. Ugo, ed., Vol. 1, Car10 Manfredi, Milan, 1970, p. 108. 2. W. Keim, A. Behr, M. Raper, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 3. R. Cramer, Acc. Chem. Res., I, 186 (1968). 4. S . Muthukumaru Pillai, M. Ravindranathan, S. Sivaram, Chem. Rev., 86, 353 (1986). 5. J. Skupifiska, Chem. Rev., 91,613 (1991). 6. F. Petit, H. Masotti, G. F'feiffer, G. Buono, J. Organomet. Chem., 244, 273 (1983).

14.5. Olefin Transformations 14.5.2. Olefin Dimerization and Oligomerization 14.5.2.3. Cyclodimerizationand Cyclooligomerizationof Monoolefins

289

~~~

7. M. Hidai, A. Misano, in Aspects of Homogeneous Catalysis, R. Ugo, ed., Vol. 2, D. Reidel, Dordrecht, 1974, p. 159. 8. G. Oehme, H. Pracejus, J . Organomet. Chem., 320, C56 (1987). 9. S. Komiya, N. Oyasato, T. Furukawa, Bull. Chem. SOC.Jpn., 62, 4078 (1989). 10. K. J. Ivin, J. J. Roonev, C. D. Stewart, M. L. H. Green, R. Mahtab, J . Chem. Soc., Chem. Comrnun., 604 (1978). 11. J. D. Fellmann, R. R. Schrock, G. A. Rupprecht, J . Am. Chem. Soc., 103, 5752 (1981). 12. A. C. L. Su. Adv. Oraanomet. Chem., 17.269 (1979). 13. R. J. McKinney, Or&nometallics, 5, 1752 (1986). 14. K. Kaneda, T. Kiriyama, T. Hiraoka, T. Imanaka, J. Mol. Catal., 48, 343 (1988). '

14.5.2.3. Cyclodlrnerizatlonand Cyclooligornerizationof Monoolefins

-

Simple monoolefins generally yield open-chain dimers or oligomers by metal-catalyzed oligomerization. Only the nickelacyclopentane 1 catalytically dimerizes' ethylene to a mixture of cyclobutane and l-butene: (Ph,P),NI()

1

CHz=CH,

toluene

In addition, propylene and acrylonitrile can be converted to cyclic dimers with the aid of 1 but in stoichiometric reactions'. In contrast, cyclodimers and cyclooligomers are obtained from strained monoolefins by oligomerizations with Ni(0) and other Group 8- 10 metal catalyst^^-^. Methylenecyclopropane is dimerized by [Ni(cod),] to a mixture of the cyclobutane derivative 2 and the cyclopentane derivative 3 (Scheme 1). By modifying the catalyst with phosphines, the cyclodimerization is suppressed and a mixture of trimers forms in which, depending on the nature of the phosphine, the linear trimer 4 or the cyclotrimer 5 A metallacycle mechanism has been developed for these reactions. The intermediate A may react further with methylenecyclopropane to form the metallacyclic precursors to the

4

Scheme 1.

5


14.5. Olefin Transformations 14.5.2. Olefin Dimerization and Oligomerization 14.5.2.3. Cyclodimerizationand Cyclooligomerizationof Monoolefins

289

~~~

7. M. Hidai, A. Misano, in Aspects of Homogeneous Catalysis, R. Ugo, ed., Vol. 2, D. Reidel, Dordrecht, 1974, p. 159. 8. G. Oehme, H. Pracejus, J . Organomet. Chem., 320, C56 (1987). 9. S. Komiya, N. Oyasato, T. Furukawa, Bull. Chem. SOC.Jpn., 62, 4078 (1989). 10. K. J. Ivin, J. J. Roonev, C. D. Stewart, M. L. H. Green, R. Mahtab, J . Chem. Soc., Chem. Comrnun., 604 (1978). 11. J. D. Fellmann, R. R. Schrock, G. A. Rupprecht, J . Am. Chem. Soc., 103, 5752 (1981). 12. A. C. L. Su. Adv. Oraanomet. Chem., 17.269 (1979). 13. R. J. McKinney, Or&nometallics, 5, 1752 (1986). 14. K. Kaneda, T. Kiriyama, T. Hiraoka, T. Imanaka, J. Mol. Catal., 48, 343 (1988). '

14.5.2.3. Cyclodlrnerizatlonand Cyclooligornerizationof Monoolefins

-

Simple monoolefins generally yield open-chain dimers or oligomers by metal-catalyzed oligomerization. Only the nickelacyclopentane 1 catalytically dimerizes' ethylene to a mixture of cyclobutane and l-butene: (Ph,P),NI()

1

CHz=CH,

toluene

In addition, propylene and acrylonitrile can be converted to cyclic dimers with the aid of 1 but in stoichiometric reactions'. In contrast, cyclodimers and cyclooligomers are obtained from strained monoolefins by oligomerizations with Ni(0) and other Group 8- 10 metal catalyst^^-^. Methylenecyclopropane is dimerized by [Ni(cod),] to a mixture of the cyclobutane derivative 2 and the cyclopentane derivative 3 (Scheme 1). By modifying the catalyst with phosphines, the cyclodimerization is suppressed and a mixture of trimers forms in which, depending on the nature of the phosphine, the linear trimer 4 or the cyclotrimer 5 A metallacycle mechanism has been developed for these reactions. The intermediate A may react further with methylenecyclopropane to form the metallacyclic precursors to the

4

Scheme 1.

5

290

14.5.2. Olefin Dimerizationand Oli omerization 14.5.2.4. Linear Dimerization and dligomerization of 1,3-Diolefins 14.5.2.4.1. By Palladium Catalysts.

linear or cyclotrimers. Monosubstituted cyclopropenes like 6 can be cyclodimerized by palladium catalysts such as PdX, (X=Cl, Br, I, NO,), [(PdCl(~3-C,H,)),] or [PdCl,(PhCN),] to a mixture of the head-to-head (7) and the head-to-tail derivative (8)4.

6

7

8

The 3,3-disubstituted cyclopropenes (9; R=CH,, OCH,) are cyclodimerized by Pd(0) catalysts such as Pd(dibenzy1idene acetone), selectively to the cyclodimers 10 ( 7 5 8 0 % ~ i e l d ) ~ Using . ~ . p(i-P19~ modified Pd(0) catalysts, the cyclopropene (9; R=CH,) can be quantitatively cyclotrimerized to the a-truns-trishomobenzene derivative 11.

% 'R 9

R

10

'VR 11

(W. KAMINSKY, R. KRAMOLOWSKY)

1. R. H. Grubbs, A. Miyashita, J . Am. Chem. Soc., 100,7416 (1978). 2. P. W. Jolly, G . Wilke, The Organic Chemistry of Nickel, Vol. 11, Academic Press, New York, 1975, p. 39. 3. P. W. Jolly, in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G . A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 633. 4. L. J. Kricka, A. Ledwith, Synthesis, 539 (1974). 5 . P. Binger, J. McMeeking, U. Schuchardt, Chem. Ber., 113,2372 (1980). 6. P. Binger, B. Biedenbach, Chem. Ber., 120, 601 (1987).

14.5.2.4. Linear Dimerization and Oligomerizatlon of 1,3-Diolefins 14.5.2.4.1. By Palladium Catalysts.

The catalytic behavior of Pd in 1,3-diolefin oligomerizations is quite different from that of Ni. Cyclooligomerization, as in the case of Ni catalysis, is normally not observed. Pd-catalyzed reactions of 1,3-diolefins are characterized by linear oligomerizations with H-migrations yielding, in the absence of nucleophiles, linear dimers or trimer~l-~. Another typical feature of Pd catalysis is the facile formation of dimeric telomers with incorporation of various nucleophiles (alcohols, carboxylic acids, amine~)'-~. A variety of Pd(0) complexes and also of Pd(I1) complexes, which are more easily accessible and may be reduced in situ to Pd(O), have been examined1-,. Some selected examples are given in Table 1. 1. Oligomerization of Butadiene. Butadiene is dimerized by Pd(0) complexes with readily available coordination sites, e.g., [Pd(PPh,),L] (L = maleic anhydride) at


290

14.5.2. Olefin Dimerizationand Oli omerization 14.5.2.4. Linear Dimerization and dligomerization of 1,3-Diolefins 14.5.2.4.1. By Palladium Catalysts.

linear or cyclotrimers. Monosubstituted cyclopropenes like 6 can be cyclodimerized by palladium catalysts such as PdX, (X=Cl, Br, I, NO,), [(PdCl(~3-C,H,)),] or [PdCl,(PhCN),] to a mixture of the head-to-head (7) and the head-to-tail derivative (8)4.

6

7

8

The 3,3-disubstituted cyclopropenes (9; R=CH,, OCH,) are cyclodimerized by Pd(0) catalysts such as Pd(dibenzy1idene acetone), selectively to the cyclodimers 10 ( 7 5 8 0 % ~ i e l d ) ~ Using . ~ . p(i-P19~ modified Pd(0) catalysts, the cyclopropene (9; R=CH,) can be quantitatively cyclotrimerized to the a-truns-trishomobenzene derivative 11.

% 'R 9

R

10

'VR 11


1. R. H. Grubbs, A. Miyashita, J . Am. Chem. Soc., 100,7416 (1978). 2. P. W. Jolly, G . Wilke, The Organic Chemistry of Nickel, Vol. 11, Academic Press, New York, 1975, p. 39. 3. P. W. Jolly, in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G . A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 633. 4. L. J. Kricka, A. Ledwith, Synthesis, 539 (1974). 5 . P. Binger, J. McMeeking, U. Schuchardt, Chem. Ber., 113,2372 (1980). 6. P. Binger, B. Biedenbach, Chem. Ber., 120, 601 (1987).

14.5.2.4. Linear Dimerization and Oligomerizatlon of 1,3-Diolefins 14.5.2.4.1. By Palladium Catalysts.

The catalytic behavior of Pd in 1,3-diolefin oligomerizations is quite different from that of Ni. Cyclooligomerization, as in the case of Ni catalysis, is normally not observed. Pd-catalyzed reactions of 1,3-diolefins are characterized by linear oligomerizations with H-migrations yielding, in the absence of nucleophiles, linear dimers or trimer~l-~. Another typical feature of Pd catalysis is the facile formation of dimeric telomers with incorporation of various nucleophiles (alcohols, carboxylic acids, amine~)'-~. A variety of Pd(0) complexes and also of Pd(I1) complexes, which are more easily accessible and may be reduced in situ to Pd(O), have been examined1-,. Some selected examples are given in Table 1. 1. Oligomerization of Butadiene. Butadiene is dimerized by Pd(0) complexes with readily available coordination sites, e.g., [Pd(PPh,),L] (L = maleic anhydride) at

14.5.2. Olefin Dimerization and Oli omerization 14.5.2.4. Linear Dimerization and 8ligomerization of 1,3-DioIefins 14.5.2.4.1. By Palladium Catalysts.

291

TABLE1. SOMETYPICAL PALLADIUM CATALYSTS FOR LINEAR OLIGOMERIZATION

OF

1 ,$DIOLEFINS

1,3-Diolefin Butadiene

Isoprene

Catalyst System [Pd(PPh3),LIa [Pd(PR3),1/COzb IP~(OAC)(V~--C~H,) 121 [Pd(OAc),] /HC02H/NEt3 [PdCl,(PPh,),]/NaOPr [Pd(PPh3),LJa [Pd(acac),]/PPh3/AC [~P~(OAC)(V~--C~H,) ),I/ P(o-tOlyl)JNEtg'HCO,H

T ("C)

Main Products

Ref.

75 110 50 50 80 110 65 25

1,3,7-Octatriene 2,4,6-0ctatriene 1,3,6,10-Dodecatetraene 1,6-0ctadiene 1,7-0ctadiene 2,7-Dimethyl-l,3,7-octatriene 2,7-Dimethyl-1,3,7-octatriene 2,6-Dimethyl-1,7-0ctadiene 3,7-Dimethyl-1,6-0ctadiene

1-3 1-3 1-3 1,2S 1

1,3 1 1

'L = maleic anhydride, benzoquinone. bn = 2,3; R = Et, Cy, Ph. 'A = rn-MeOC,H,CHO.

100-120°C in aprotic solvents such as acetone or, more smoothly, in i-PrOH, with 80% yield to 1,3,7-octatrienes (1). The presence of CO, greatly enhances the catalytic activity of [Pd(PR,),] ( n = 2-4) and causes isomerization of 1 to 2,4,6-octatrienes. In formic acid, reductive dimerization proceeds and, depending on catalyst and cocatalyst, either 1,6- or 1,7-octadiene is formed highly selecti~ely',~*~. In the absence of phosphines, certain q3-allylpalladium complexes catalyze the oligomerization of butadiene to linear trimers' [equation (a)].

\

79% selectivity (a) 30% conversion

Addition of PR3 suppresses trimerization; linear dimers are obtained.

2. Oligomerization of Isoprene. Isoprene may be oligomerized by Pd(0) catalysts to yield'^^,^ a mixture of four isomeric linear dimers 2-5. The distribution of isomers is highly sensitive to catalyst, solvent and reaction conditions. Lower T affords higher selectivity'. The tail-to-tail dimer 2 is obtained in high yield (75%) and excellent selectivity (98%) with the aid of [Pd(PPh,),L] (L = maleic anhydride). Phenol as an additive for the catalyst system [PdBr,(dppe)]/NaOPh (dppe = 1,2-bis(diphenylphosphino)ethane) greatly enhances the activity and changes the regioselectivity. Depending on the molar ratio pheno1:isoprene varying amounts of 2-5 are produced. At a 1:30 ratio, the tail-to-tail dimer 2 is formed almost exclusively, while with a 1:3 ratio the head-to-tail dimer 5 prevails. A ratio 1:1.5 leads to head-to-head dimerization yielding 3 and 4 as main products

'.

292

8

14.5.2. Olefin Dimerization and Oli omerization 14.5.2.4. Linear Dimerizationand ligomerizationof 1,3-Diolefins 14.5.2.4.2. By Nickel Catalysts.

2

3

4

5 (W. KAMINSKY, R. KRAMOLOWSKY)

J. Tsuji, Adv. Organomet. Chem., 17,141 (1979). 5. Tsuji, Acc. Chem. Rex, 6, 8 (1973). R. Baker, Chem. Rev., 73,487 (1973). B. M. Trost, T. R. Verhoeven, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 799. 5 . C. U.Pittmann, Jr., R. M. Hanes, J. J. Yang, J . Mol. Catal., 15, 377 (1982).

1. 2. 3. 4.


In the presence of suitable cocatalysts such as alcohols, phenols, or secondary amines, 1,3-diolefins are oligomerized to linear dimers or trimers by the same “nickelligand” systems, which are effective for cyclooligomerization (see 14.5.2.5.1). Reactions may be accompanied by telomerization. Typical examples are given in Table 1. Butadiene is mainly dimerized to n-octatrienes the structures of which depend highly upon the particular ligand and c~catalyst’-~. Thus, by the catalysts given in equation (a) or equation (b), respectively, butadiene is converted almost quantitatively either to 1,3-trans, 6-trans-octatriene or to 2-trans, 4-trans, 6-trans-octatriene:

[Ni(acac),]/PBu,

80% yield

AIEt,/CH,CH,NH

(b)

TABLE1. SOME TYPICALNICKELCATALYSTS FOR LINEAROLIGOMERIZATION OF 1,%DIOLEFINS 1.3-Diolefin Butadiene

Catalyst System

[NiBr,(PPh,),]/NaBH,/EtOH

[Ni(acac),]/PBu3/A1Et,/

CH,CH,NH

[Ni(PPh,),]/EtOH

[Ni(173-C,H,)(PPh3),lPF,/

Isoprene

i-PrOH [NiCI,(PPh,),]/NaBH,/ EtMgBr [Ni(acac),]/PBu,/AIEt,/

CH,CH,~H

[Ni(l13-C3H,),]/As(i-Pr)~

T (“C)

Main Products

Ref.

100

1,3,6-Octatriene 2,4,6-0ctatriene

1-3 23

25 80

1,3,7-0ctatriene 1,3,6-0ctatriene

1-3

60

2,6-Dimethyl-l,3,6-octatriene 1-3

100

2,6-Dimethyl-2,4,6-octatriene 2,3

100

50

Linear trimers

4

1-3


292

8

14.5.2. Olefin Dimerization and Oli omerization 14.5.2.4. Linear Dimerizationand ligomerizationof 1,3-Diolefins 14.5.2.4.2. By Nickel Catalysts.

2

3

4


J. Tsuji, Adv. Organomet. Chem., 17,141 (1979). 5. Tsuji, Acc. Chem. Rex, 6, 8 (1973). R. Baker, Chem. Rev., 73,487 (1973). B. M. Trost, T. R. Verhoeven, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 799. 5 . C. U.Pittmann, Jr., R. M. Hanes, J. J. Yang, J . Mol. Catal., 15, 377 (1982).

1. 2. 3. 4.


In the presence of suitable cocatalysts such as alcohols, phenols, or secondary amines, 1,3-diolefins are oligomerized to linear dimers or trimers by the same “nickelligand” systems, which are effective for cyclooligomerization (see 14.5.2.5.1). Reactions may be accompanied by telomerization. Typical examples are given in Table 1. Butadiene is mainly dimerized to n-octatrienes the structures of which depend highly upon the particular ligand and c~catalyst’-~. Thus, by the catalysts given in equation (a) or equation (b), respectively, butadiene is converted almost quantitatively either to 1,3-trans, 6-trans-octatriene or to 2-trans, 4-trans, 6-trans-octatriene:

[Ni(acac),]/PBu,

80% yield

AIEt,/CH,CH,NH

(b)

TABLE1. SOME TYPICALNICKELCATALYSTS FOR LINEAROLIGOMERIZATION OF 1,%DIOLEFINS 1.3-Diolefin Butadiene

Catalyst System

[NiBr,(PPh,),]/NaBH,/EtOH

[Ni(acac),]/PBu3/A1Et,/

CH,CH,NH

[Ni(PPh,),]/EtOH

[Ni(173-C,H,)(PPh3),lPF,/

Isoprene

i-PrOH [NiCI,(PPh,),]/NaBH,/ EtMgBr [Ni(acac),]/PBu,/AIEt,/

CH,CH,~H

[Ni(l13-C3H,),]/As(i-Pr)~

T (“C)

Main Products

Ref.

100

1,3,6-Octatriene 2,4,6-0ctatriene

1-3 23

25 80

1,3,7-0ctatriene 1,3,6-0ctatriene

1-3

60

2,6-Dimethyl-l,3,6-octatriene 1-3

100

2,6-Dimethyl-2,4,6-octatriene 2,3

100

50

Linear trimers

4

1-3

8

14.5.2. Olefin Dimerizationand Oli omerization 14.5.2.4. Linear Dimerization and ligomerization of 1,3-Diolefins 14.5.2.4.3. By Other Transition Metal Catalysts.

293

Ni(0) catalysts such as [Ni(cod),] modified by P-methyl substituted 1,3,2oxazaph~spholidine~ or aminophosphinite ligands63’ Ph,POCHR’CHR2NHR3 dimerize butadiene to mainly 1,3,6-octatrienes. Using the catalyst system [Ni(acac),]/ P(NMe,),/LiBu/MeOH, butadiene is dimerized with 90% selectivity to a mixture of 1,3,6- and 1,3,7-0ctatrienes~~~. The cationic catalyst [Ni(~3-C,H4Me)(PPh,)2]PF6 cyclodimerizes butadiene in aprotic solvents whereas in alcohols such as MeOH and i-PrOH and by addition of NaOMe, mainly a mixture of n-octatrienes is obtained4. Isoprene may be dimerized by certain Ni catalysts to methylsubstituted octatrienes or heptatrienes2s3.Linear trimers such as 1 and 2 form from isoprene by Ni-L- systems with AsR, ligand~l-~.

1


1. P. W. Jolly, G. Wilke, The Organic Chemistry of Nickel, Vol. 11, Academic Press, New York, 1975, p. 185. 2. W. Keim, A. Behr, M. Roper, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 3. P. W. Jolly, in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 671. 4. P. Grenouillet, D. Neibecker, I. Tkatchenko, J . Organornet. Chem., 243, 213 (1983). 5. W. J. Richter, J . Mol. Catal., 34, 145 (1986). 6. P. Denis, A. Jean, J. F. Crcizy, A. Mortreux, F. Petit, J . Am. Chem. SOC., 112, 1292 (1990). 7. P. Denis, A. Mortreux, F. Petit, G. Buono, G. Pfeiffer, J . Org. Chem., 49, 5274 (1984). 8. J. Beger, Ch. Duschek, H. Fullbier, W. Gaube, J . Prakt. Chem., 316,26 (1974). G. H. Fullbier, W. Gaube, B. Leuner, Wiss. Z . Ernst-Moritz-Arndt-Univ. Greifswald. Math.-nat. wiss. Reihe, 35, 34 (1986). 14.5.2.4.3. By Other Transition Metal Catalysts.

Linear oligomerization of 1,3-diolefins may be accomplished by a variety of transition-metal systems other than Ni and Pd. Linear dimers are formed mainly, the structure of which depends to some extent upon the metal. Typical examples are given in Table 1. Butadiene is dimerized by Co catalysts mainly to methylsubstituted heptatrienes whereas n-octatrienes are obtained by Fe and Zr catalysts. By Group 4 and 5 catalysts, isoprene is generally converted to head-to-tail dimers such as 2,6-dimethyl- 1,3,6octatriene~~,~, Depending on the nature of R, the Cr catalysts [Cr(RN=CH-CH=NR),] dimerize isoprene either in a tail-to-tail or in a head-to-tail fashion, e.g., with R-CH(i-Pr), the tt-dimer 2,7-dimethyl-2,4,6-octatriene is obtained with 80% selectivity”. (W. KAMINSKY, R. KRAMOLOWSKY)

1. R. Baker, Chem. Rev., 73,487 (1973). 2. W. Keim, A. Behr, M. Roper, in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 3. H. tom Dieck, H. Muller, J . Organornet. Chem., 221, C7 (1981). 4. H. tom Dieck, A. Kinzel, Angew. Chem. Int. Ed. Engl., 18, 324 (1979). 5. H. Fullbier, K. Erler, W. Gaube, Wiss. Z. Ernst-Moritz-Arndt-Univ. Greifswald, Math.-nat.wiss. Reihe, 35, 39 (1986).


8

14.5.2. Olefin Dimerizationand Oli omerization 14.5.2.4. Linear Dimerization and ligomerization of 1,3-Diolefins 14.5.2.4.3. By Other Transition Metal Catalysts.

293

Ni(0) catalysts such as [Ni(cod),] modified by P-methyl substituted 1,3,2oxazaph~spholidine~ or aminophosphinite ligands63’ Ph,POCHR’CHR2NHR3 dimerize butadiene to mainly 1,3,6-octatrienes. Using the catalyst system [Ni(acac),]/ P(NMe,),/LiBu/MeOH, butadiene is dimerized with 90% selectivity to a mixture of 1,3,6- and 1,3,7-0ctatrienes~~~. The cationic catalyst [Ni(~3-C,H4Me)(PPh,)2]PF6 cyclodimerizes butadiene in aprotic solvents whereas in alcohols such as MeOH and i-PrOH and by addition of NaOMe, mainly a mixture of n-octatrienes is obtained4. Isoprene may be dimerized by certain Ni catalysts to methylsubstituted octatrienes or heptatrienes2s3.Linear trimers such as 1 and 2 form from isoprene by Ni-L- systems with AsR, ligand~l-~.

1


1. P. W. Jolly, G. Wilke, The Organic Chemistry of Nickel, Vol. 11, Academic Press, New York, 1975, p. 185. 2. W. Keim, A. Behr, M. Roper, in Comprehensive Organometallic Chemistry, G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 3. P. W. Jolly, in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 671. 4. P. Grenouillet, D. Neibecker, I. Tkatchenko, J . Organornet. Chem., 243, 213 (1983). 5. W. J. Richter, J . Mol. Catal., 34, 145 (1986). 6. P. Denis, A. Jean, J. F. Crcizy, A. Mortreux, F. Petit, J . Am. Chem. SOC., 112, 1292 (1990). 7. P. Denis, A. Mortreux, F. Petit, G. Buono, G. Pfeiffer, J . Org. Chem., 49, 5274 (1984). 8. J. Beger, Ch. Duschek, H. Fullbier, W. Gaube, J . Prakt. Chem., 316,26 (1974). G. H. Fullbier, W. Gaube, B. Leuner, Wiss. Z . Ernst-Moritz-Arndt-Univ. Greifswald. Math.-nat. wiss. Reihe, 35, 34 (1986). 14.5.2.4.3. By Other Transition Metal Catalysts.

Linear oligomerization of 1,3-diolefins may be accomplished by a variety of transition-metal systems other than Ni and Pd. Linear dimers are formed mainly, the structure of which depends to some extent upon the metal. Typical examples are given in Table 1. Butadiene is dimerized by Co catalysts mainly to methylsubstituted heptatrienes whereas n-octatrienes are obtained by Fe and Zr catalysts. By Group 4 and 5 catalysts, isoprene is generally converted to head-to-tail dimers such as 2,6-dimethyl- 1,3,6octatriene~~,~, Depending on the nature of R, the Cr catalysts [Cr(RN=CH-CH=NR),] dimerize isoprene either in a tail-to-tail or in a head-to-tail fashion, e.g., with R-CH(i-Pr), the tt-dimer 2,7-dimethyl-2,4,6-octatriene is obtained with 80% selectivity”. (W. KAMINSKY, R. KRAMOLOWSKY)

1. R. Baker, Chem. Rev., 73,487 (1973). 2. W. Keim, A. Behr, M. Roper, in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 3. H. tom Dieck, H. Muller, J . Organornet. Chem., 221, C7 (1981). 4. H. tom Dieck, A. Kinzel, Angew. Chem. Int. Ed. Engl., 18, 324 (1979). 5. H. Fullbier, K. Erler, W. Gaube, Wiss. Z. Ernst-Moritz-Arndt-Univ. Greifswald, Math.-nat.wiss. Reihe, 35, 39 (1986).

294

14.5.2. Olefin Dirnerization and Oligomerization 14.5.2.5. C clodimerizationand Cyclooligornerization of 1,3-Diolefins 14.5.2.5.1. b y Nickel Catalysts.

TABLE1. SOME TRANSITION METALCATALYSTS (OTHER THANNi AND Pd) FOR LINEAR OLIGOMERIZATION OF 1 ,3-DIOLEFINS 1,3-Diolefin Butadiene

Isoprene

Catalyst System [CO~(CO)~I/AIE~~ [Co(q3-C3H,),l FeC13/PPh,/A1Et, [Zr(v3-C3H5),(cot)l TiC14/P(OPh),/A1Et3 [Ti(6b),l/AlEt: Z~CI,Y,/AIC~E~~ Hf(OBu),/AlClEt, [Cr(RN=CH-CH=NR)JC

T (“C) 40 20 13 65 120 50 130 120

Main Products

Ref.

5-Methyl-1,3,6-heptatriene 5-Methyl-1,3,6-heptatriene 1,3,6-0ctatriene 1,3,6-0ctatriene 2,6-Dimethyl- 1,3,6-0ctatriene 2,6-Dimethyl- 1,3,6-0ctatriene 2,7-Dimethyl- 1,3,6-0ctatriene 2,6-Dimethyl- 1,3,6-octatriene 2,7-Dimethyl-2,4,6-octatriene

‘00 = [X(C,H40)21z- (X=O, S). by = p-toluenesulfonate. ‘R = CH(i-Pr),.

14.5.2.5. Cyclodlmerlzatlonand Cyclooligomerlzatlonof 1,&Diolefins 14.5.2.5.1. By Nickel Catalysts.

The Ni-catalyzed cyclooligomerization of 1,3-diolefins is important for preparative applications and mechanistic insights. The reactions form the basis for syntheses of medium and large ring compounds which are useful starting materials for further organic syntheses. In general, Ni(0) catalysts in which one coordination site is occupied by an electron donor (‘ ‘nickel-ligand catalysts”’) cyclodimerize 1,3-diolefins whereas “ligand free” Ni(0) systems’ lead to cyclotrimers or, occasionally, to higher cyclooligomers’-3.

1. Cyclodimerization of 1,3-Diolefins. The “nickel-ligand” catalysts effective in the cyclodimerization of 1,3-diolefins are composed of preformed complexes such as [Ni(CO),L,-,] or [Ni(cod),]/L [L = PR,, P(OR),] or are prepared in situ by reducing a Ni(I1) compound such as [Ni(acac),] in the presence of L and the di~lefinl-~. Organoaluminum compounds are most commonly used reducing agents, but various other systems have also been investigated’,*. Butadiene is cyclodimerized by these catalysts predominantly to cod and vch, but dvcb and vmcp are also formed under some condition^'-^. The rate of conversion and the selectivity for cyclodimerization are highest with a Ni:L ratio of 1:1, lower Ni:L ratios favoring cyclotrimerization.

The catalytic activity of the Ni(0)-L system, the yield of cyclodimers and the ratio of cod to vch are highly dependent on the nature of L (Table l)’-,. If P-donors are used as L, the degree of cyclodimerization is influenced mainly by the steric properties of L; particularly small as well as very large ligands, e.g., P(OMe), and P(t-Bu)(i-Pr),, favor cyclotrimerization. With bulky phosphites such as P(OC,H4-o-Ph),, the selectivity to


294

14.5.2. Olefin Dirnerization and Oligomerization 14.5.2.5. C clodimerizationand Cyclooligornerization of 1,3-Diolefins 14.5.2.5.1. b y Nickel Catalysts.

TABLE1. SOME TRANSITION METALCATALYSTS (OTHER THANNi AND Pd) FOR LINEAR OLIGOMERIZATION OF 1 ,3-DIOLEFINS 1,3-Diolefin Butadiene

Isoprene

Catalyst System [CO~(CO)~I/AIE~~ [Co(q3-C3H,),l FeC13/PPh,/A1Et, [Zr(v3-C3H5),(cot)l TiC14/P(OPh),/A1Et3 [Ti(6b),l/AlEt: Z~CI,Y,/AIC~E~~ Hf(OBu),/AlClEt, [Cr(RN=CH-CH=NR)JC

T (“C) 40 20 13 65 120 50 130 120

Main Products

Ref.

5-Methyl-1,3,6-heptatriene 5-Methyl-1,3,6-heptatriene 1,3,6-0ctatriene 1,3,6-0ctatriene 2,6-Dimethyl- 1,3,6-0ctatriene 2,6-Dimethyl- 1,3,6-0ctatriene 2,7-Dimethyl- 1,3,6-0ctatriene 2,6-Dimethyl- 1,3,6-octatriene 2,7-Dimethyl-2,4,6-octatriene

‘00 = [X(C,H40)21z- (X=O, S). by = p-toluenesulfonate. ‘R = CH(i-Pr),.

14.5.2.5. Cyclodlmerlzatlonand Cyclooligomerlzatlonof 1,&Diolefins 14.5.2.5.1. By Nickel Catalysts.

The Ni-catalyzed cyclooligomerization of 1,3-diolefins is important for preparative applications and mechanistic insights. The reactions form the basis for syntheses of medium and large ring compounds which are useful starting materials for further organic syntheses. In general, Ni(0) catalysts in which one coordination site is occupied by an electron donor (‘ ‘nickel-ligand catalysts”’) cyclodimerize 1,3-diolefins whereas “ligand free” Ni(0) systems’ lead to cyclotrimers or, occasionally, to higher cyclooligomers’-3.

1. Cyclodimerization of 1,3-Diolefins. The “nickel-ligand” catalysts effective in the cyclodimerization of 1,3-diolefins are composed of preformed complexes such as [Ni(CO),L,-,] or [Ni(cod),]/L [L = PR,, P(OR),] or are prepared in situ by reducing a Ni(I1) compound such as [Ni(acac),] in the presence of L and the di~lefinl-~. Organoaluminum compounds are most commonly used reducing agents, but various other systems have also been investigated’,*. Butadiene is cyclodimerized by these catalysts predominantly to cod and vch, but dvcb and vmcp are also formed under some condition^'-^. The rate of conversion and the selectivity for cyclodimerization are highest with a Ni:L ratio of 1:1, lower Ni:L ratios favoring cyclotrimerization.

The catalytic activity of the Ni(0)-L system, the yield of cyclodimers and the ratio of cod to vch are highly dependent on the nature of L (Table l)’-,. If P-donors are used as L, the degree of cyclodimerization is influenced mainly by the steric properties of L; particularly small as well as very large ligands, e.g., P(OMe), and P(t-Bu)(i-Pr),, favor cyclotrimerization. With bulky phosphites such as P(OC,H4-o-Ph),, the selectivity to

14.5.2. Olefin Dimerization and Oligomerization 14.5.2.5. C clodimerization and Cyclooligomerizationof 1,3-Diolefins 14.5.2.5.1. b y Nickel Catalysts.

295

TABLE1. PRODUCTS FROM THE NI-LIGANDCATALYZED CYCLOOLIGOMERIZATION OF

BUTADIENE5’8

eb

Ligand P(t-Bu)(i-Pr), P( i-P& PPr, PPh, P(OMe), P(OC~H,-O-P~), P(OPh),

cod

Xb

[“I

[cm-’I

167 160 139 145 107 152 128

2.0 3.1 4.9 12.9 23.4 28.9 29.2

[%I

28.2 38.9 41.7 63.2 20.5 94.5 79.4

vch

cdt‘

[%]

[%I

14.7 23.0 21.0 20.7 15.4 2.4 7.4

49.6 23.6 12.9 14.8 59.8 1.5 12.2

Misc.“ [%]

Tumover6-e

7.5 14.5 24.4 1.3 4.3 1.6 1 .o

180 780 100

a[Ni(cod),]:ligand = 1:l; T = 60°C. bf3,,y= cone angle and electronic parameter, respectively, after Tolman’. ‘cdt, cyclcdodecatrienes. dunidentified dimers and oligomers. eTurnover in g butadiene(gNi)-’ h-’; [Ni(cod),]:ligand = 1:1, T = 80°C, 3 h, 1 bar.

cyclodimers is 98%. The distribution of cod and vch is controlled to a large extent by electronic effects. Formation of cod is increased with decreasing donor strength of the P-ligand4. Whereas the yield of vch never exceeds 50% under these conditions, the selectivity for cod can reach 95%. In the presence of P(OR), ligands, dvcb may be isolated in yields up to 40% if the reaction is carried out at 40°C and terminated at maximum 85% conversion to prevent thermic or catalytic rearrangement (Cope) to cod’,,. In the presence of methanol, butadiene is converted by various Ni(0) catalyst systems’.’ such as [Ni(CO),(PPh,),], [NiBr,(PBu,),]/BuLi, or [Ni(cod),]/threophos (1)*with high selectivity (up to 95%) to vmcp.

Ph,POCH,

CH3 H I N-PPh,

\ITcH

PhzPO H

3

1 An optical induction of up to 24% enantiomeric excess for the formation of the chiral dimer vch from butadiene can be obtained using optically active 1,3,2-dioxaphospholanes L in the catalyst system’ [Ni(cod),]/L. Substituted 1,3-dienes are far less reactive than butadiene, the rate of reaction decreasing in the order:

w>>-> /b/>>

+

whereas 2,4-hexadiene does not react’. Isoprene is cyclodimerized by Ni(0)-L catalysts to a mixture of methylsubstituted cod, vch, and dvcb derivatives, the distribution of which depends on the ligand used and

296

14.5.2. Olefin Dimerization and Oligomerization 14.5.2.5. Cyclodimerization and Cyclooligomerizationof 1,3-Diolefins 14.5.2.5.1. By Nickel Catalysts.

reaction conditions. The main product in the presence of PPh, or NEt, is 1,5- or 2,5dimethyl-cod, respectively'. Using fluoroalkylphosphites, high selectivities for cyclodimers (up to 97%) are obtained, with 1,4-dimethyl-vch as main product (60%)". ~~~~

2. Cyclotrimerization of 1,3-Diolefins. The cyclotrimerization of 1,3-diolefins is possible using "ligand-free"' Ni(0) systems. These catalysts may consist of preformed Ni(0) complexes, e.g., [Ni(cod),] or [Ni(H,C=CHCN),], or are prepared in situ by reducing a Ni(I1) compound such as [Ni(acac),] with a reducing agent, generally an organoaluminum compound, in the presence of the 1,3-di0lefin'.~. Butadiene is converted by these catalysts with high selectivity to a mixture of three (2-4) of the possible four isomers of 1,5,9-~yclododecatriene; the all-trans-isomer 2 is the main product.

2, all-trans

3, trans-trans-cis

4, trans -cis -cis

The all-cis isomer has never been detected. By addition of pyridine, formation of the trans-trans-cis isomer 3 is suppressed in favor of 2. Alkylsubstituted butadienes are far less reactive than butadiene. The cyclotrimerization of isoprene to trimethylcyclododecatriene isomers is favored by Ni(0) catalysts modified by phoshites with low basicity and low steric demand'.

3. Mechanistic Considerations. The Ni-catalyzed cyclooligomerizations of 1,3-diolefins are one of the mechanistically best understood reactions in transition metal catalysis'. For example, with butadiene, a mechanism (in simplified form) for the formation of the cyclic dimers and trimers is shown in Scheme 1. All the cyclooligomers result from the key intermediate 5 with 7'-allyl, ~,-allylcoordination, which is formed by oxidative coupling of two Ni-bonded butadiene molecules. In the absence of donor ligands, a further butadiene is inserted to give the bis(~~-allyl) intermediate 6.The cyclotrimer is formed from 6 by reductive coupling and displacement of the complexed cyclotrimer with butadiene. By addition of electron-donor ligands L, cyclotrimerization is prevented. The intermediate 5 may react to give vch or rearrange to the bis(q3-allyl) complexes 7,8 as precursors for the formation of dvcb and cod. The reductive elimination steps presumably proceed through [ NiL(~',T1-C8H12)) intermediates'. (W. KAMINSKY, R. KRAMOLOWSKY)

1. P. W. Jolly, G. Wilke, The Organic Chemistry of Nickel, Vol. 11, Academic Press, New York, 1975, p. 133. 2. P. W. Jolly, in Comprehensive Organometallic Chemistry,G . Wilkinson, F. G. A. Stone, E. W. Abel, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 671. 3. P. Heimbach, Angew. Chem. Int. Ed. Engl., 12,975 (1973). 4. P. Heimbach, in Aspects of Homogeneous Catalysis, R. Ugo, ed., Vol. 2, D. Reidel, Dordrecht, 1974, p. 81. 5. P. Heimbach, J. Kluth, H. Schenkluhn, B. Weimann, Angew. Chem. Int. Ed. Engl., 19, 569, 570 (1980). 6. W. Brenner, P. Heimbach, H. Hey, E. W. Mliller, G. Wilke, Liebigs Ann. Chem., 727, 161 (1969).

14.5.2. Olefin Dimerizationand Oligomerization 14.5.2.5. Cyclodimerizationand Cyclooligornerizationof 1,&Diolefins 14.5.2.5.2. By Other Transition Metal Catalysts.

297

I

(L-Nil

+ 2/ -

- (L-Ni)

-(L-Ni)

-(L-Ni)

Scheme 1. 7. C. A. Tolmann, Chem. Rev., 77,313 (1977). 8. H. Masotti, G . Peiffer, C. Siv, E. Joblet, R. Phan Tan Luu, Bull. Soc. Chim. Belg., 98, 191 (1989). 9. W. J. Richter, J . Mol. Catal., 13,210 (1981); 18, 145 (1983). 10. P. W. N. M. van Leeuwen, C. F. Roobeek, Tetrahedron, 37, 1973 (1981).

14.5.2.5.2. By Other Transition Metal Catalysts.

In addition to Ni catalysts, systems involving Ti, Cr, Mn, Fe, and Co are effective for cyclooligomerization of 1,3-diolefins1.Some selected examples are given in Table 1. Catalysts prepared from a Ti(1V) salt and an alkylaluminum compound (ZieglerNatta type) convert butadiene with high selectivities and yields of 80-90% to 1,5,9cyclododecatrienes consisting mainly of the trans-trans-cis isomer with small amounts of the all-trans isomer'*2.The activity and selectivity of these catalysts systems is improved by adding substances such as NaCl, which can complex with the aluminum chloride2. Using nitrosyliron catalysts such as [Fe(NO),(CO),] or [ { FeCl(NO)2)2]/reductant, butadiene is cyclodimerized with high conversions and extraordinary specificity (up to 99.5%)to 4-vinyl-1-cyclohexene'. With Fe(0) catalysts of the type [FeCl,(dad)]/RMgX (dad = R'N=CR2-CR2=NR1) butadiene can be converted almost quantitatively to cyclodimers. The formation of vch and cod depends markedly on the nature of the group R' and can be directed to 98% cod or 80% vch3v4. With chiral dad-ligands the cyclodimerization of butadiene to 4-vinyl- 1-cyclohexene occurs with an enantiomeric excess up to 62%4. Butadiene is selectively (89%)cyclodimerized to cod5 using the cationic Ru catalyst [Ru($-Me,C,T)( v4-C4H,)]OTf (Otf = trifluoromethanesulfonate) in THF. (W. KAMINSKY, R. KRAMOLOWSKY)


14.5.2. Olefin Dimerizationand Oligomerization 14.5.2.5. Cyclodimerizationand Cyclooligornerizationof 1,&Diolefins 14.5.2.5.2. By Other Transition Metal Catalysts.

297

I

(L-Nil

+ 2/ -

- (L-Ni)

-(L-Ni)

-(L-Ni)

Scheme 1. 7. C. A. Tolmann, Chem. Rev., 77,313 (1977). 8. H. Masotti, G . Peiffer, C. Siv, E. Joblet, R. Phan Tan Luu, Bull. Soc. Chim. Belg., 98, 191 (1989). 9. W. J. Richter, J . Mol. Catal., 13,210 (1981); 18, 145 (1983). 10. P. W. N. M. van Leeuwen, C. F. Roobeek, Tetrahedron, 37, 1973 (1981).

14.5.2.5.2. By Other Transition Metal Catalysts.

In addition to Ni catalysts, systems involving Ti, Cr, Mn, Fe, and Co are effective for cyclooligomerization of 1,3-diolefins1.Some selected examples are given in Table 1. Catalysts prepared from a Ti(1V) salt and an alkylaluminum compound (ZieglerNatta type) convert butadiene with high selectivities and yields of 80-90% to 1,5,9cyclododecatrienes consisting mainly of the trans-trans-cis isomer with small amounts of the all-trans isomer'*2.The activity and selectivity of these catalysts systems is improved by adding substances such as NaCl, which can complex with the aluminum chloride2. Using nitrosyliron catalysts such as [Fe(NO),(CO),] or [ { FeCl(NO)2)2]/reductant, butadiene is cyclodimerized with high conversions and extraordinary specificity (up to 99.5%)to 4-vinyl-1-cyclohexene'. With Fe(0) catalysts of the type [FeCl,(dad)]/RMgX (dad = R'N=CR2-CR2=NR1) butadiene can be converted almost quantitatively to cyclodimers. The formation of vch and cod depends markedly on the nature of the group R' and can be directed to 98% cod or 80% vch3v4. With chiral dad-ligands the cyclodimerization of butadiene to 4-vinyl- 1-cyclohexene occurs with an enantiomeric excess up to 62%4. Butadiene is selectively (89%)cyclodimerized to cod5 using the cationic Ru catalyst [Ru($-Me,C,T)( v4-C4H,)]OTf (Otf = trifluoromethanesulfonate) in THF. (W. KAMINSKY, R. KRAMOLOWSKY)

14.5. Olefin Transformations 14.5.3. Olefin Polymerization 14.5.3.1. Introduction

298

TABLE1. SOMETYPICAL TRANSITION METALCATALYSTS (OTHER THANNI) FOR CYCLOOLlGOMERlZATION OF 1,3-DIOLEFINSa 1,3-Diolefin Butadiene

Isoprene

Catalyst System

T ("C)

TiC1,/AIEt,/Et,NCH,CHzOH TiCl,/AlClEt,

70 40

[Cr(acac),]/AIEt,/cocatalystb

70

[Mn(acac),]/A1Et3/dppec [ { FeCl(NO), ) ,]/electrolysis [Fe(~3-C3H,)(Co)z(NO)l [Fe(acac),]/AlEt,/bipy TiCl,.dioxane/AIClEt, Ti(Ot-Bu),/A1,C1,Et3/AlH3 [ [ FeCI(NO),},]/reductantd

120 25

[Fe(acac),]/AlEt,/bipy

100

100 50 25 90

Main Products 4-Vinylc yclohexene

1,5,9-~yclododecatriene(trans, trans, cis) 1,5,9-cyclododecatriene(trans, trans, cis)

4-Vin y lcyclohexene

4-Vinylcyclohexene 4-Vinylcyclohexene 1,5-Cyclooctadiene 2,4-Dimethyl-4-vinylcyclohexene Trimethyl-1,5,9-~yclododecatriene 1,4(2,4)-Dimethyl-4vinylcyclohexene 1,5(2,5)-Dimethyl-1,5cyclooctadiene

aFrom ref. 1 . bCocatalyst = PPhJI,; t-BuCI.

'dppe, 1,2-bis(diphenyIphosphino)ethane. dReductant = Zn,electrolysis.

1. W. Keim, A. Behr, M. Roper in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G . A. Stone, E. W. Abed, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 2. R. Baker, Chem. Rev., 73, 487 (1973). 3. H. tom Dieck, J. Dietrich, J. Ehlers, U.Schacht, Chem.-lng. Tech., 61, 832 (1989). 4. H. tom Dieck, J. Dietrich, Angew. Chem. lnt. Ed. Engl., 24,781 (1985). 5. K. Masuda, K. Nakano, T. Fukahori, H. Nagashima, K. Itoh, J . Organomet. Chem., 428, C21 (1992).

14.5.3. Olefin Polymerization 14.5.3.1. Introduction The polymerization of olefins at organometallic compounds is one of the most important industrial catalyses. Ethylene and other a-olefins are polymerized by transitionmetal compounds, with cocatalysts of a main-group metal alkyl, halide or oxide, e.g., aluminoxane, magnesia, or silica. Catalysts of this type operate at P between 1 and 50 atm and at T between 20 and 200°C in solvents, in the liquid monomer, or in the gas phase. Approximately 19 X lo6 tons of polyethylene and 14 X lo6 tons of polypropylene are produced per year with the aid of such catalysts. Discoveries in 195 1 I (Phillips process) and 1954 (Ziegler)' of catalysts based on transition metals opened up new possibilities, e.g., to produce high-density polyethylene (HDPE), linear low-density polyethylene (LLDPE) or propylene diene rubber (EPDM). The range of application of organometallic catalysts expanded significantly due to the findings (Natta) that propylene and higher aolefins can be polymerized by stereoselectivity .


14.5. Olefin Transformations 14.5.3. Olefin Polymerization 14.5.3.1. Introduction

298

TABLE1. SOMETYPICAL TRANSITION METALCATALYSTS (OTHER THANNI) FOR CYCLOOLlGOMERlZATION OF 1,3-DIOLEFINSa 1,3-Diolefin Butadiene

Isoprene

Catalyst System

T ("C)

TiC1,/AIEt,/Et,NCH,CHzOH TiCl,/AlClEt,

70 40

[Cr(acac),]/AIEt,/cocatalystb

70

[Mn(acac),]/A1Et3/dppec [ { FeCl(NO), ) ,]/electrolysis [Fe(~3-C3H,)(Co)z(NO)l [Fe(acac),]/AlEt,/bipy TiCl,.dioxane/AIClEt, Ti(Ot-Bu),/A1,C1,Et3/AlH3 [ [ FeCI(NO),},]/reductantd

120 25

[Fe(acac),]/AlEt,/bipy

100

100 50 25 90

Main Products 4-Vinylc yclohexene

1,5,9-~yclododecatriene(trans, trans, cis) 1,5,9-cyclododecatriene(trans, trans, cis)

4-Vin y lcyclohexene

4-Vinylcyclohexene 4-Vinylcyclohexene 1,5-Cyclooctadiene 2,4-Dimethyl-4-vinylcyclohexene Trimethyl-1,5,9-~yclododecatriene 1,4(2,4)-Dimethyl-4vinylcyclohexene 1,5(2,5)-Dimethyl-1,5cyclooctadiene

aFrom ref. 1 . bCocatalyst = PPhJI,; t-BuCI.

'dppe, 1,2-bis(diphenyIphosphino)ethane. dReductant = Zn,electrolysis.

1. W. Keim, A. Behr, M. Roper in Comprehensive Organometallic Chemistry, G . Wilkinson, F. G . A. Stone, E. W. Abed, eds., Vol. 8, Pergamon Press, Oxford, 1982, p. 371. 2. R. Baker, Chem. Rev., 73, 487 (1973). 3. H. tom Dieck, J. Dietrich, J. Ehlers, U.Schacht, Chem.-lng. Tech., 61, 832 (1989). 4. H. tom Dieck, J. Dietrich, Angew. Chem. lnt. Ed. Engl., 24,781 (1985). 5. K. Masuda, K. Nakano, T. Fukahori, H. Nagashima, K. Itoh, J . Organomet. Chem., 428, C21 (1992).

14.5.3. Olefin Polymerization 14.5.3.1. Introduction The polymerization of olefins at organometallic compounds is one of the most important industrial catalyses. Ethylene and other a-olefins are polymerized by transitionmetal compounds, with cocatalysts of a main-group metal alkyl, halide or oxide, e.g., aluminoxane, magnesia, or silica. Catalysts of this type operate at P between 1 and 50 atm and at T between 20 and 200°C in solvents, in the liquid monomer, or in the gas phase. Approximately 19 X lo6 tons of polyethylene and 14 X lo6 tons of polypropylene are produced per year with the aid of such catalysts. Discoveries in 195 1 I (Phillips process) and 1954 (Ziegler)' of catalysts based on transition metals opened up new possibilities, e.g., to produce high-density polyethylene (HDPE), linear low-density polyethylene (LLDPE) or propylene diene rubber (EPDM). The range of application of organometallic catalysts expanded significantly due to the findings (Natta) that propylene and higher aolefins can be polymerized by stereoselectivity .

14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.1. By Titanium Catalysts.

The polyolefins produced by transition metal catalysts are characterized by the absence of large amounts of long- or short-chain branching, which causes variability in density, crystallinity, and melting points. Most catalysts used are heterogeneous but some homogeneous systems are known. A two-step mechanism for catalysis is widely accepted: (1) adsorption of the monomer, which may be activated by the configuration established in this step, and (2) insertion of the activated monomer into a metal-carbon bond. Several articles and books have been published in the last few years that give detailed information on this field4-14. Although olefin polymerization is applied on a broad technical scale, the nature of the active center and many reaction mechanisms even 40 years after its discovery are incompletely understood". (W. KAMINSKY, R. KRAMOLOWSKY) 1. 2. 3. 4.

5. 6. 7. 8. 9. 10. 11.

12. 13. 14.

15.

J. P. Hogan, Polym. Prepr., 10, 240 (1969). K. Ziegler, B. Holzkamp, H. Martin, H. Breil, Angew. Chem., 67, 541 (1955). G. Natta, J. Polym. Sci., 16, 143 (1955). T. Keii, Kinetics of Ziegler-Natta Polymerization, Kodanska, Tokyo, and Chapman & Hall, London, 1972. J. C. W. Chien, ed., Coordination Polymerization. Academic Press, New York, 1975. A. C. Caunt, Catalysis, Vol. 1, C. Kemball, ed., Chemistry Society, London, 1977, p. 234. J. Boor, Ziegler-Natta Catalysts and Polymerizations. Academic Press, New York, 1979. H. Sinn, W. Kaminsky, Adv. Organomet. Chem., 18, 99 (1980). R. P. Quirk, ed., Transition Metal Catalyzed Polymerizations, MMI Press Symposium Series, Vol. 4, Parts A & B, Harwood, New York, 1983. P. Pino, B. Rotzinger, Makromol. Chem. Suppl., 7,41 (1984). I. Pasquon, U. Giannini, in Catalysis, Science and Technology, J. R. Anderson, M.Boudart, eds., Springer-Verlag, Berlin, 1984, p. 65. Y. V. Kissin, Isospecific Polymerization of Olefins, Springer-Verlag, Berlin, 1985. T. Keii, K. Soga, eds., Studies in Surfnce Science and Catalysis, Vol. 25, Elsevier-Kodansha, Tokyo, 1986. W. Kaminsky, H. Sinn, eds., Transition Metals and Organometallics as Catalysts for Olefin Polymerization. Springer-Verlag, Berlin, 1988. R. B. Seymour, T. Cheng, eds., History of Polyolefins, D. Reidel, Dordrecht, 1985.

14.5.3.2. Ethylene Polymerization 14.5.3.2.1. By Titanium Catalysts.

The first generation of catalysts based on titanium tetrachloride or different modifications of titanium trichloride and ethyl-, isobutyl-, isoprenyl-, or chloride containing aluminum compounds, Ziegler catalysts, (Table 1) was characterized by low polymerization activity. Kinetic studies and applications of various methods have helped to define the nature of the active center'-4. In all systems that contain at least one C1 atom per titanium-aluminum system, there is rapid complex formation, which can be recognized by color intensification (Scheme 1). The complex 1 is heterogeneous and is dark olive in color. The next step is the alkylation at the transition metal 2. Fixation of the ethylene takes place at the free coordination site (0) of the titanium'~*.Insertion into the Ti-C bond follows. High-density polyethylene (HDPE) is formed. Scheme 2 shows this chain propagation and the following chain termination steps (a). Hydrogenation terminates the chains, forming saturated polymers and metal-hydride, the latter adding olefin to reactivate the catalyst center (b). As a consequence, the MW of the polymers is decreased.



The polyolefins produced by transition metal catalysts are characterized by the absence of large amounts of long- or short-chain branching, which causes variability in density, crystallinity, and melting points. Most catalysts used are heterogeneous but some homogeneous systems are known. A two-step mechanism for catalysis is widely accepted: (1) adsorption of the monomer, which may be activated by the configuration established in this step, and (2) insertion of the activated monomer into a metal-carbon bond. Several articles and books have been published in the last few years that give detailed information on this field4-14. Although olefin polymerization is applied on a broad technical scale, the nature of the active center and many reaction mechanisms even 40 years after its discovery are incompletely understood". (W. KAMINSKY, R. KRAMOLOWSKY) 1. 2. 3. 4.

5. 6. 7. 8. 9. 10. 11.

12. 13. 14.

15.

J. P. Hogan, Polym. Prepr., 10, 240 (1969). K. Ziegler, B. Holzkamp, H. Martin, H. Breil, Angew. Chem., 67, 541 (1955). G. Natta, J. Polym. Sci., 16, 143 (1955). T. Keii, Kinetics of Ziegler-Natta Polymerization, Kodanska, Tokyo, and Chapman & Hall, London, 1972. J. C. W. Chien, ed., Coordination Polymerization. Academic Press, New York, 1975. A. C. Caunt, Catalysis, Vol. 1, C. Kemball, ed., Chemistry Society, London, 1977, p. 234. J. Boor, Ziegler-Natta Catalysts and Polymerizations. Academic Press, New York, 1979. H. Sinn, W. Kaminsky, Adv. Organomet. Chem., 18, 99 (1980). R. P. Quirk, ed., Transition Metal Catalyzed Polymerizations, MMI Press Symposium Series, Vol. 4, Parts A & B, Harwood, New York, 1983. P. Pino, B. Rotzinger, Makromol. Chem. Suppl., 7,41 (1984). I. Pasquon, U. Giannini, in Catalysis, Science and Technology, J. R. Anderson, M.Boudart, eds., Springer-Verlag, Berlin, 1984, p. 65. Y. V. Kissin, Isospecific Polymerization of Olefins, Springer-Verlag, Berlin, 1985. T. Keii, K. Soga, eds., Studies in Surfnce Science and Catalysis, Vol. 25, Elsevier-Kodansha, Tokyo, 1986. W. Kaminsky, H. Sinn, eds., Transition Metals and Organometallics as Catalysts for Olefin Polymerization. Springer-Verlag, Berlin, 1988. R. B. Seymour, T. Cheng, eds., History of Polyolefins, D. Reidel, Dordrecht, 1985.

14.5.3.2. Ethylene Polymerization 14.5.3.2.1. By Titanium Catalysts.

The first generation of catalysts based on titanium tetrachloride or different modifications of titanium trichloride and ethyl-, isobutyl-, isoprenyl-, or chloride containing aluminum compounds, Ziegler catalysts, (Table 1) was characterized by low polymerization activity. Kinetic studies and applications of various methods have helped to define the nature of the active center'-4. In all systems that contain at least one C1 atom per titanium-aluminum system, there is rapid complex formation, which can be recognized by color intensification (Scheme 1). The complex 1 is heterogeneous and is dark olive in color. The next step is the alkylation at the transition metal 2. Fixation of the ethylene takes place at the free coordination site (0) of the titanium'~*.Insertion into the Ti-C bond follows. High-density polyethylene (HDPE) is formed. Scheme 2 shows this chain propagation and the following chain termination steps (a). Hydrogenation terminates the chains, forming saturated polymers and metal-hydride, the latter adding olefin to reactivate the catalyst center (b). As a consequence, the MW of the polymers is decreased.

T* mol of active centers.

TiC14/MgC1,/A1Et,

TiC14/Mg(O-C,H,),/Al(i-butyl),

Second generation Heterogeneous TiCl,/Mg(butyI)Br/Al/Et, TiC14/Mg(R)C1/AIEt,

First generation Heterogeneous TiCl,/ AlEt, TiC14/ Al(i-butyI), TiCl,/ AlEt,CI Homogeneous CfiTiCI,/AIMqCI Cp,TiCI,/AIEt,Cl Cp,TiMe,/AIMe,/H,O Ti(benzyl),/ AlEt,Cl

Catalyst

1.600 250,000 20,000 500,000

40

200 50 3,500

60

5 10

Activity (kg PE/mol TLh)

0.70

0.6 0.36

0.02 0.02 0.3

0.005

0.03

0.01

Efficiency (mol C*/mol Ti)”

50 50

75

30 15

49 30 50

50 70 70

20 20

70 50

60

Temperature (“C)

42 70

Activation Energy (W m o U

TABLE1. POLYMERIZATION RATESOF HOMOGENEOUS AND HETEROGENEOUS ZIEGLER-CATALYSTS FOR ETHYLENE

2 5 1 6

3 3 1 4

1 2

1

Ref.


2

cl\ C1’

.P1+ E t \ A1/C1\

l-1

‘C1

2 C1’

Et

I \C1/ c1

’

\Cl/

\Et

301

/Et A1 ‘Et

1

n

Scheme 1. Complexation and alkylexchange in the titanium system. Decrease in MW is also a consequence of labilization of the metalorganic bond brought about by the nature of the alkyl group, an equilibrium that leads to polymers having a vinyl endgroup (c). The resulting hydride reacts in the same manner as the transition metal alkyl compound. Most of the homogeneous Ziegler-type catalysts have been preferentially investigated in order to understand the elementary steps of the polymerization, which is simplified in soluble systems. Bis(cyclopentadienyl)titanium(IV) compounds, which are soluble in aromatic hydrocarbons could be used together with aluminum alkyls to give Ziegler-catalysts3.As to the kinetics of polymerization and to side reactions of the catalyst components, this system is probably the best understood. It has not been used in a technical process because of the low activity and short life of systems that contain chloride (see Table 1).

3 The complex 3 analogous to the heterogeneous system (Schemes 1 and 2) cannot be isolated; it has been identified only spectroscopically. The soluble Ti-A1 complexes may have an octahedral structure relative to the coordination sphere of the Ti atoms’. Polymerization of a-olefins aided by homogeneous and highly alkylated ZieglerNatta catalysts overlaps a series of reactions that involves the formation of inactive compounds and thus give an idea of what happens during inactivation of heterogeneous systems (Scheme 3). A continued olefin insertion into the metal-carbon bond of the Ti complex, alkyl exchange and hydrogen transfer reactions are observed. In the case of

302


Chain Propagation

Chain Termination (a) by p-elimination with H-transfer to monomer /CH,-CH M j >CH,---H CH,

/ CH,=CH

/R

I

-

M

+ \ CH,-

R

CH3

(b) by hydrogenation

/R CHZ-CH, M/i j /H H

-

CH3-

+

R / CH2

M-H

-

(c) by p-elimination forming hydride CH, M/'CH-R '%

/

CH,=CH-R

+

M-H

H Scheme 2. Mechanism of ethylene polymerization (M

=

transition metal).

the Ti compound, the Ti atom in the (metal)-CH2-CH,-(meta1)structureis reduced by separation of C,H,. The Ti(II1)complexesthat form are inactive for CzH4 polymerization. Subsequent research' showed, at least for the primarily formed complex, a tetrahedral structure for the Ti. Because only 1% of the Ti complexes form active centers it is reasonable that only 1% of the complexes have an octahedral configuration. The efficiency of the Ti can increase by X 10 to X 100 in the supported catalyst, if, e.g., 5 wt% TiCl, is mixed with milled MgC1, (Table 1)'031'.It is not known why Mg compounds cause this high increase in activity. The similarity in ionic radii, 70 pm for Ti4+ and 66 pm for Mgz+, could play a part. It is clear that by heterogenization with MgCl, the Ti atoms stay far enough apart so that side reactions involving inactive species like those in the homogeneous systems are suppressed. With this second generation catalyst it is possible to produce polyethylene with: 1. high activity; catalyst can remain in the polymer, 2. regulation of the MW,

-

14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.2. By Vanadium Catalysts. Insertion

R mtT CH2- CH,

HExchange

/r------4 - 7 mt ; H Imt'

CH2-CH2

I

;LCH3 _ - -- _CH, _ _4 J_ _ I

,CH2-CH,, Reduction

mt

mt

mt

-

303

mt -CH2- CH,-

R

FH2CH2\ mt' +

mt

CH, -CH, CH,=

mt

CH2 mt

= Metal valence

Scheme 3. Reactions with homogeneous catalysts.

3. control over the MW distribution, 4. control over the copolymerization, and 5. production of polymer particles of high bulk density. By copolymerization of C2H, with 1-butene, 1-hexene, or other 1-olefins, linear low-density polyethylene (LLDPE) is obtained with a density of 0.92 to 0.93 g cmP3 instead of 0.95-0.97 g c m P 3in homopolyethylene. The number of CH, groups per 1000 carbon atoms is 5-10 instead of 1-2 in pure polyethylene. (W. KAMINSKY, R. KRAMOLOWSKY) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

K. H. Reichert,Angew.Makromol. Chem., 94, 1 (1981). K. Soga, T. Shiono, Y. Doi, J . Chem. SOC. Chem. Commun., 840 (1984). W. P. Long, D. S.Breslow, Liebigs Ann. Chem., 463 (1975). J. C. W. Chien, J. T. T. Hsieh, J . Polym. Sci. Polym. Chem. Ed., 14, 1915 (1976). V. A. Zakharov, G. D. Bukatov, N. B. Chumalvskil, Y. L. Ermakov,Kinet. Katal., 18, 1 (1980). C. L. Bohm, J . Appl. Polym. Sci., 29 279 (1984). P. Cossee, J . Catal., 3, 80 (1964). G. Henrici-Olive, Angew. Chem. Int. Ed. Engl., 10, 776 (1971). G. Fink, R. Rottler, Angew. Makromol. Chem., 94, 25 (1981). K. Soga, M. Ohgizawa, T. Shiono, D. Lee, Macromol., 24, 1699 (1991). M. Nowakowska, K. Bosowska, Makrornol. Chem., 193,889 (1992).

14.5.3.2.2. By Vanadium Catalysts.

Other well known homogeneous Ziegler-Natta catalysts include derivatives of V and organoaluminum compounds'-4. VC14/A1C1Et2/anisole VO( a~ac)~/AlClEt~/activator VCl,/Al(i-Bu), VO(OR),/AlCl(i-Bu)2 VOCl,/AlCl,-,R,


-

14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.2. By Vanadium Catalysts. Insertion

R mtT CH2- CH,

HExchange

/r------4 - 7 mt ; H Imt'

CH2-CH2

I

;LCH3 _ - -- _CH, _ _4 J_ _ I

,CH2-CH,, Reduction

mt

mt

mt

-

303

mt -CH2- CH,-

R

FH2CH2\ mt' +

mt

CH, -CH, CH,=

mt

CH2 mt

= Metal valence

Scheme 3. Reactions with homogeneous catalysts.

3. control over the MW distribution, 4. control over the copolymerization, and 5. production of polymer particles of high bulk density. By copolymerization of C2H, with 1-butene, 1-hexene, or other 1-olefins, linear low-density polyethylene (LLDPE) is obtained with a density of 0.92 to 0.93 g cmP3 instead of 0.95-0.97 g c m P 3in homopolyethylene. The number of CH, groups per 1000 carbon atoms is 5-10 instead of 1-2 in pure polyethylene. (W. KAMINSKY, R. KRAMOLOWSKY) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

K. H. Reichert,Angew.Makromol. Chem., 94, 1 (1981). K. Soga, T. Shiono, Y. Doi, J . Chem. SOC. Chem. Commun., 840 (1984). W. P. Long, D. S.Breslow, Liebigs Ann. Chem., 463 (1975). J. C. W. Chien, J. T. T. Hsieh, J . Polym. Sci. Polym. Chem. Ed., 14, 1915 (1976). V. A. Zakharov, G. D. Bukatov, N. B. Chumalvskil, Y. L. Ermakov,Kinet. Katal., 18, 1 (1980). C. L. Bohm, J . Appl. Polym. Sci., 29 279 (1984). P. Cossee, J . Catal., 3, 80 (1964). G. Henrici-Olive, Angew. Chem. Int. Ed. Engl., 10, 776 (1971). G. Fink, R. Rottler, Angew. Makromol. Chem., 94, 25 (1981). K. Soga, M. Ohgizawa, T. Shiono, D. Lee, Macromol., 24, 1699 (1991). M. Nowakowska, K. Bosowska, Makrornol. Chem., 193,889 (1992).

14.5.3.2.2. By Vanadium Catalysts.

Other well known homogeneous Ziegler-Natta catalysts include derivatives of V and organoaluminum compounds'-4. VC14/A1C1Et2/anisole VO( a~ac)~/AlClEt~/activator VCl,/Al(i-Bu), VO(OR),/AlCl(i-Bu)2 VOCl,/AlCl,-,R,

304

14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.3. By Homogeneous Zirconium Catalysts.

In contrast to the catalysts containing Ti that produce polymers with a broad MW distribution of 5 to 30, the compounds containing V produce polyethylene with a narrow MW distribution of 2-4. The V systems are suited to polymerization of higher a-olefins and for copolymerization. Therefore, these systems are used technically to make a rubber (EPDM type) by copolymerization of CZH4, propylene, and as diene ethylidene norbornene (ENB). These catalysts are initially very active because of the presence of V(III), which seems to be the active oxidation form. However, some of them lose activity by reduction after a short polymerization time. They can be reactivated by weak oxidizing agents (activator) like chlorobenzene. The active centers for polymerization by the catalyst system VCl,/Al(i-Bu), are considered to be alkylated entities, such as VCl,R, on the surface of VC1, crystals5. Chain initiation is believed to arise from the following sequence of reactions: VCl,

+ AlR,

c1 --+

\ \Al / R / \R/ \R

c1

(fast)

+AIR,

L

VCl2R Polyethylene

t-

+ AlZR5Cl

(slow)

Ethylene

The optimum ratio between V and A1 compounds is 1:10. (W. KAMINSKY, R. KRAMOLOWSKY) 1 . W. L. Carrick, J . Am. Chem. SOC., 80,6455 (1958). 2. H. Ernde, Angew. Makromol. Chem.,60, 1 (1977). 3. N. Kashiwa, T. Tsutsui, Makromol. Chem.,Rapid Comm., 4,491 (1983). 4. T. Nozaki, J. C. W. Chien, J . Polym. Sci., Polym. Chem. Ed., 29, 1807 (1991). 5. P. J. T. Tait, in Developments in Polymerization, Vol. 2, R. N. Haward ed., Applied Science Publishers, London, 1979, p. 81. 14.5.3.2.3. By Homogeneous Zirconium Catalysts.

If zirconocene compounds or other metallocenes are mixed with aluminoxanes (produced by the controlled reaction of H,O and AIR,) in hydrocarbon solvents, soluble catalysts are formed with extremely high polymerization activities'.'. The obtainable polymerization activities of the C,H4 polymerization using the catalyst system Zr(~5-CSH5)2Me2/methylal~minoxane are shown in Table 1. These figures imply that insertion of an CZH4 molecule (turnover time) takes only 5 X s, assuming that every Zr atom forms an active center. Kinetic measurements show that approximately 70% of all Zr atoms-at higher temperatures the amount is greater-form active polymerization centers. The higher the degree of oligomerization of the aluminoxane, the more active are the catalysts. They possess a polymerization activity that lasts for days. The MW of polyethylene can be varied widely by changing the polymerization temperature between 20 and 110°C, by variation of the concentration of Cp,ZrCl,, or by


304

14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.3. By Homogeneous Zirconium Catalysts.

In contrast to the catalysts containing Ti that produce polymers with a broad MW distribution of 5 to 30, the compounds containing V produce polyethylene with a narrow MW distribution of 2-4. The V systems are suited to polymerization of higher a-olefins and for copolymerization. Therefore, these systems are used technically to make a rubber (EPDM type) by copolymerization of CZH4, propylene, and as diene ethylidene norbornene (ENB). These catalysts are initially very active because of the presence of V(III), which seems to be the active oxidation form. However, some of them lose activity by reduction after a short polymerization time. They can be reactivated by weak oxidizing agents (activator) like chlorobenzene. The active centers for polymerization by the catalyst system VCl,/Al(i-Bu), are considered to be alkylated entities, such as VCl,R, on the surface of VC1, crystals5. Chain initiation is believed to arise from the following sequence of reactions: VCl,

+ AlR,

c1 --+

\ \Al / R / \R/ \R

c1

(fast)

+AIR,

L

VCl2R Polyethylene

t-

+ AlZR5Cl

(slow)

Ethylene

The optimum ratio between V and A1 compounds is 1:10. (W. KAMINSKY, R. KRAMOLOWSKY) 1 . W. L. Carrick, J . Am. Chem. SOC., 80,6455 (1958). 2. H. Ernde, Angew. Makromol. Chem.,60, 1 (1977). 3. N. Kashiwa, T. Tsutsui, Makromol. Chem.,Rapid Comm., 4,491 (1983). 4. T. Nozaki, J. C. W. Chien, J . Polym. Sci., Polym. Chem. Ed., 29, 1807 (1991). 5. P. J. T. Tait, in Developments in Polymerization, Vol. 2, R. N. Haward ed., Applied Science Publishers, London, 1979, p. 81. 14.5.3.2.3. By Homogeneous Zirconium Catalysts.

If zirconocene compounds or other metallocenes are mixed with aluminoxanes (produced by the controlled reaction of H,O and AIR,) in hydrocarbon solvents, soluble catalysts are formed with extremely high polymerization activities'.'. The obtainable polymerization activities of the C,H4 polymerization using the catalyst system Zr(~5-CSH5)2Me2/methylal~minoxane are shown in Table 1. These figures imply that insertion of an CZH4 molecule (turnover time) takes only 5 X s, assuming that every Zr atom forms an active center. Kinetic measurements show that approximately 70% of all Zr atoms-at higher temperatures the amount is greater-form active polymerization centers. The higher the degree of oligomerization of the aluminoxane, the more active are the catalysts. They possess a polymerization activity that lasts for days. The MW of polyethylene can be varied widely by changing the polymerization temperature between 20 and 110°C, by variation of the concentration of Cp,ZrCl,, or by

14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.4. By Chromium Catalysts.

305

TABLE1. POLYMERIZATION OF ETHYLENE WITH HOMOGENEOUS ZIEGLER-CATALYSTS AND ALUMINOXANE (lo-)

Transition Metal (mol/liter) Cp,TiMe,, Cp,TiCI,, Cp2ZrMe2,3X cp,zrci,, 10-7 Cp,HfMe,, Cp,ZrcI,, 10 - 6 Cp,ZrMe,,

MOL/LITER) AS & C A T A L Y S T IN

Cocatalyst Methylaluminoxane Methylaluminoxane Methylaluminoxane Methylaluminoxane Methylaluminoxane Ethylaluminoxane Tetraisobutyl Dialuminoxane

TOLUENE

Activity (g PE/g Zrh.bar)

Temperature ("C)

500 90,000 9,000 1,000,000 60,000 23,000 175,000

20 20 20

70 70 60 70

addition of small amounts of H,. At RT polyethylene is formed with a molecular weight of 500,000, whereas at temperatures above 100°C it is only ca. 50,000. These catalyst systems also copolymerize C,H, with long-chain a-olefins, e.g., propylene, 1-butene, and 1-hexene. Through variations in the monomer behavior, the desired proportion of olefin/ethylene in the copolymer can be regulated. By incorporating 1-hexene into the polyethylene matrix, polymers (LLDPE) with densities between 0.90 and 0.98 g/ml form; properties of high-pressure polyethylene are also obtainable. The copolymerization of C,H, together with a diolefin, e.g., 1,7-octadiene, as comonomer produces a cross-linked polyethylene with rubber elasticity, insoluble in all organic solvents. With the aid of homogeneous, highly active Ziegler-Natta catalysts it is possible to polymerize olefins in the presence of starch, cellulose, or lignin. The active centers of the catalysts are fixed on the surface of the starch grains, and these centers are covered by polyethylene after polymerization. As cocatalyst also fluorinated boranes can be used3. (W. KAMINSKY, R. KRAMOLOWSKY)

1. W. Kaminsky, M. Miri, H. Sinn, R. Woldt, Makromol. Chem. Rapid Commun., 4,417 (1983). 2. W. Kaminsky, H. Liiker, Makromol. Chem. Rapid Commun., 5, 225 (1983). 3. J. C. W. Chien, W. M. Tsai, M. D. Rausch, J . Am. Chem. Soc., 113, 8570 (1991).

14.5.3.2.4. By Chromium Catalysts.

To the Cr catalysts belong the alkyl-free Phillips catalysts and the Cr($-ally1)3 system, which produces highly linear polyethylene. The most widely investigated Phillips catalyst is prepared by impregnating a silicaalumina or a pure silica support with 1-5 wt% CrO,'. High surface area supports are used at about 400 to 600 m2/g. The catalyst is dried at 400 to 800°C and treated with CO to reduce the Cr(V1) to Cr(I1). The Phillips catalyst is specific for polymerization of C,H, to high-density polyethylene. The reaction between silica and CrO, is


14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.4. By Chromium Catalysts.

305

TABLE1. POLYMERIZATION OF ETHYLENE WITH HOMOGENEOUS ZIEGLER-CATALYSTS AND ALUMINOXANE (lo-)

Transition Metal (mol/liter) Cp,TiMe,, Cp,TiCI,, Cp2ZrMe2,3X cp,zrci,, 10-7 Cp,HfMe,, Cp,ZrcI,, 10 - 6 Cp,ZrMe,,

MOL/LITER) AS & C A T A L Y S T IN

Cocatalyst Methylaluminoxane Methylaluminoxane Methylaluminoxane Methylaluminoxane Methylaluminoxane Ethylaluminoxane Tetraisobutyl Dialuminoxane

TOLUENE

Activity (g PE/g Zrh.bar)

Temperature ("C)

500 90,000 9,000 1,000,000 60,000 23,000 175,000

20 20 20

70 70 60 70

addition of small amounts of H,. At RT polyethylene is formed with a molecular weight of 500,000, whereas at temperatures above 100°C it is only ca. 50,000. These catalyst systems also copolymerize C,H, with long-chain a-olefins, e.g., propylene, 1-butene, and 1-hexene. Through variations in the monomer behavior, the desired proportion of olefin/ethylene in the copolymer can be regulated. By incorporating 1-hexene into the polyethylene matrix, polymers (LLDPE) with densities between 0.90 and 0.98 g/ml form; properties of high-pressure polyethylene are also obtainable. The copolymerization of C,H, together with a diolefin, e.g., 1,7-octadiene, as comonomer produces a cross-linked polyethylene with rubber elasticity, insoluble in all organic solvents. With the aid of homogeneous, highly active Ziegler-Natta catalysts it is possible to polymerize olefins in the presence of starch, cellulose, or lignin. The active centers of the catalysts are fixed on the surface of the starch grains, and these centers are covered by polyethylene after polymerization. As cocatalyst also fluorinated boranes can be used3. (W. KAMINSKY, R. KRAMOLOWSKY)

1. W. Kaminsky, M. Miri, H. Sinn, R. Woldt, Makromol. Chem. Rapid Commun., 4,417 (1983). 2. W. Kaminsky, H. Liiker, Makromol. Chem. Rapid Commun., 5, 225 (1983). 3. J. C. W. Chien, W. M. Tsai, M. D. Rausch, J . Am. Chem. Soc., 113, 8570 (1991).

14.5.3.2.4. By Chromium Catalysts.

To the Cr catalysts belong the alkyl-free Phillips catalysts and the Cr($-ally1)3 system, which produces highly linear polyethylene. The most widely investigated Phillips catalyst is prepared by impregnating a silicaalumina or a pure silica support with 1-5 wt% CrO,'. High surface area supports are used at about 400 to 600 m2/g. The catalyst is dried at 400 to 800°C and treated with CO to reduce the Cr(V1) to Cr(I1). The Phillips catalyst is specific for polymerization of C,H, to high-density polyethylene. The reaction between silica and CrO, is

306

14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.5. By Nickel-Ylid Catalyst.

OH

I -Si-O-SiI

OH

I I

+ CrO,

-+

0

0

o/

\o

I -Si-O-SiI

I

+ H,O

(a)

The Cr(V1) is reduced by C2H4 or CO to an active, blue Cr(I1) complex. However, only 5% of the total Cr atoms are active centers. The activity reaches 100,000 kg polyethylene per mol Cr/h or per 2 kg of catalyst. The Cr(q3-allyl), is another Cr catalyst. The 7-ally1 bond is not the active center’ but when C2H4 is complexed to the metal center it converts the 7-ally1bond to a a-bond.

\ /CH,\ Cr

CH

\

’

+ CH,=CH,

F H 2 \

Cr ‘CH,

CH= CH,

( b)

//

CH, The polymerization temperature is 80°C; however, the activity is low. (W. KAMINSKY, R. KRAMOLOWSKY)

1. H. R. Sailors, J. P. Hogan, J . Macromol. Sci. Chem., A l , 1377 (1981). 2. D. G. H. Ballard, E. Jones, T. Modinger, A. J. Pioli, Mukromol. Chem., 148, 175 (1971). 3. M. P. McDaniel, J. Cafal.,67, 71 (1981). 14.5.3.2.5. By Nlckel-Ylld Catalyst.

Some bis(y1id) Ni catalysts are remarkably active for C,H, polymerization’J. The ylid can be synthesized by reaction of Ni(0) complexes and phosphines: (COD),Ni

Ph,

,Ph

COD = cyclooctadiene For the one-component catalyst is is possible to use solvents of various polarities. There is good activity even in THF or acetone. The best solvents are CH,Cl, or hexane. If the hydrogen next to the oxygen in the ylid is replaced by larger groups, the activity increases and reaches at l0-atm C2H4 pressure and 100°C about 50,000 mol of reacted C2H4 per mol of Ni3. The MW ranges up to lo6. The branches are higher than in polyethylene prepared with Ziegler catalyst; up to 58 CH, groups per 1,000 C atoms were found in oligomers. (W. KAMINSKY, R. KRAMOLOWSKY)

1. W. Keirn, in CaralyficPolymerizafion ofOlefins,T. Keii, K. Soga, eds., Kodanska Tokyo, 1986, p. 201. 2. K. A. Ostoja Starzewski, J. Witte, Angew. Chem. Inf. Ed. Engl., 26, 63 (1987).


306

14.5.3. Olefin Polymerization 14.5.3.2. Ethylene Polymerization 14.5.3.2.5. By Nickel-Ylid Catalyst.

OH

I -Si-O-SiI

OH

I I

+ CrO,

-+

0

0

o/

\o

I -Si-O-SiI

I

+ H,O

(a)

The Cr(V1) is reduced by C2H4 or CO to an active, blue Cr(I1) complex. However, only 5% of the total Cr atoms are active centers. The activity reaches 100,000 kg polyethylene per mol Cr/h or per 2 kg of catalyst. The Cr(q3-allyl), is another Cr catalyst. The 7-ally1 bond is not the active center’ but when C2H4 is complexed to the metal center it converts the 7-ally1bond to a a-bond.

\ /CH,\ Cr

CH

\

’

+ CH,=CH,

F H 2 \

Cr ‘CH,

CH= CH,

( b)

//

CH, The polymerization temperature is 80°C; however, the activity is low. (W. KAMINSKY, R. KRAMOLOWSKY)

1. H. R. Sailors, J. P. Hogan, J . Macromol. Sci. Chem., A l , 1377 (1981). 2. D. G. H. Ballard, E. Jones, T. Modinger, A. J. Pioli, Mukromol. Chem., 148, 175 (1971). 3. M. P. McDaniel, J. Cafal.,67, 71 (1981). 14.5.3.2.5. By Nlckel-Ylld Catalyst.

Some bis(y1id) Ni catalysts are remarkably active for C,H, polymerization’J. The ylid can be synthesized by reaction of Ni(0) complexes and phosphines: (COD),Ni

Ph,

,Ph

COD = cyclooctadiene For the one-component catalyst is is possible to use solvents of various polarities. There is good activity even in THF or acetone. The best solvents are CH,Cl, or hexane. If the hydrogen next to the oxygen in the ylid is replaced by larger groups, the activity increases and reaches at l0-atm C2H4 pressure and 100°C about 50,000 mol of reacted C2H4 per mol of Ni3. The MW ranges up to lo6. The branches are higher than in polyethylene prepared with Ziegler catalyst; up to 58 CH, groups per 1,000 C atoms were found in oligomers. (W. KAMINSKY, R. KRAMOLOWSKY)

1. W. Keirn, in CaralyficPolymerizafion ofOlefins,T. Keii, K. Soga, eds., Kodanska Tokyo, 1986, p. 201. 2. K. A. Ostoja Starzewski, J. Witte, Angew. Chem. Inf. Ed. Engl., 26, 63 (1987).

14.5. Olefin Transformations 14.5.3. Olefin Polymerization 14.5.3.3. Propylene Polymerization

307

3. K. A. Ostoja Starzewski, J. Witte, in Transition Metal Catalyzed Polymerizations, Ziegler-Natta and Metathesis Polymerizations, R. P. Quirk, ed., Cambridge Univ. Press, New York, 1988.

14.5.3.3. Propylene Polymerization

Transition metal catalysts can polymerize propylene and other higher a-olefins in a stereoregular way to isotactic structures (l),syndiotactic structures (2), or to the statistical atactic structure (3).

1

2

R

R

R

R

3 Industrially used catalysts for the propylene polymerization are based mainly on Ti compounds (Table 1)'s2, but there are some systems that use MoO,/CaO pretreated with H, and CrO, supported on A1,03/Si0, or metallocene/aluminoxane systems. (W. KAMINSKY, R. KRAMOLOWSKY)

1. P. Pino, R. Miilhaupt, Angew. Chem. Int. Ed. Engl., 19, 857 (1980). 2. A. Yamamoto, T. Yamamoto, Macromol. Rev.,13, 161 (1978). 3. B. L. Goodall, Transition Metal Catalyzed Polymerization, MMI Symposium Series, Vol. 4, R. P. Quirk, ed., 1983, p. 355. 4. E. A. Youngman, J. Boor, J. Polym. Sci. B 3, 577 (1965).

TABLE1. DIFFERENT TYPESOF TITANIUM CATALYSTS FOR THE POLYMERIZATION OF PROPYLENE~

Activity (g PP/g Ti.h.bar) ~

TiCI,/AlEt, (1:3) MgCl,/TiCl,/AlEt, MgCl,/TiCl,/AlEt,/LB P-TiCI,/AlEt,Cl/LB, TiCI,/MgCI,/Al(iC,H,), P-TiC1,/AICl,/AlEt,C1/LB3 TiCI,/LiAl,H,/NaF TiCI,/NEt,/H,

~

~~

30 3,840 755 99 3,800 520 70 50

"LB,, ethyl benzoate; LB,, methyl methacrylate; LB,, diisoamyl ether.

Stereoselectivity (% isotactic) ~

~~

Ref.

~

27 21 91 95

5 3 3

98 90 93

6 1

21

1 1

4


14.5. Olefin Transformations 14.5.3. Olefin Polymerization 14.5.3.3. Propylene Polymerization

307

3. K. A. Ostoja Starzewski, J. Witte, in Transition Metal Catalyzed Polymerizations, Ziegler-Natta and Metathesis Polymerizations, R. P. Quirk, ed., Cambridge Univ. Press, New York, 1988.

14.5.3.3. Propylene Polymerization

Transition metal catalysts can polymerize propylene and other higher a-olefins in a stereoregular way to isotactic structures (l),syndiotactic structures (2), or to the statistical atactic structure (3).

1

2

R

R

R

R

3 Industrially used catalysts for the propylene polymerization are based mainly on Ti compounds (Table 1)'s2, but there are some systems that use MoO,/CaO pretreated with H, and CrO, supported on A1,03/Si0, or metallocene/aluminoxane systems. (W. KAMINSKY, R. KRAMOLOWSKY)

1. P. Pino, R. Miilhaupt, Angew. Chem. Int. Ed. Engl., 19, 857 (1980). 2. A. Yamamoto, T. Yamamoto, Macromol. Rev.,13, 161 (1978). 3. B. L. Goodall, Transition Metal Catalyzed Polymerization, MMI Symposium Series, Vol. 4, R. P. Quirk, ed., 1983, p. 355. 4. E. A. Youngman, J. Boor, J. Polym. Sci. B 3, 577 (1965).

TABLE1. DIFFERENT TYPESOF TITANIUM CATALYSTS FOR THE POLYMERIZATION OF PROPYLENE~

Activity (g PP/g Ti.h.bar) ~

TiCI,/AlEt, (1:3) MgCl,/TiCl,/AlEt, MgCl,/TiCl,/AlEt,/LB P-TiCI,/AlEt,Cl/LB, TiCI,/MgCI,/Al(iC,H,), P-TiC1,/AICl,/AlEt,C1/LB3 TiCI,/LiAl,H,/NaF TiCI,/NEt,/H,

~

~~

30 3,840 755 99 3,800 520 70 50

"LB,, ethyl benzoate; LB,, methyl methacrylate; LB,, diisoamyl ether.

Stereoselectivity (% isotactic) ~

~~

Ref.

~

27 21 91 95

5 3 3

98 90 93

6 1

21

1 1

4

308

14.5.3. Olefin Polymerization 14.5.3.3. Propylene Polymerization 14.5.3.3.1. Heterogeneous Catalysts.

5. G. Natta, P. Pino, G. Mazzanti, P. Longi, Gazz. Chim. Iral., 87, 570 (1957). 6. J. C. W. Chien, Preparations and Properties of Stereoregular Polymers, R. W.

delli, eds., D. Reidel, Dordrecht, 1980, p. 115.

Lenz, F. Ciar-

14.5.3.3.1. Heterogeneous Catalysts.

Highly active catalysts are obtained by ball milling MgCl, with TiCl, to form electron donor complexes such as TiCl, ethyl benzoate'. The crystal size of the milled MgC1, is 3-10 mm. In order to combine high activity with good stereoselectivity, the cocatalyst used is a modified or complexed trialkylaluminoxane with electron donors such as ethyl anisate, ethyl toluate, or ethyl benzoate. Typically, the pale yellow solid catalyst, contains 1-2 wt% Ti and 5-20 wt% ethyl benzoate. Polymerization activity reaches 1,500 kg polypropylene/g Ti, with an isotactic index of 96%. It is often assumed that the electron donors improve stereoselectivity by selectively complexing atactic sites on the catalyst surface2. This may be the mechanism with simple electron donors such as tertiary amines. However, in contrast, esters undergo irreversible reactions with AIR, under polymerization conditions (Scheme 1). Reactions are complex and consecutive steps involving alkylation, reduction, and elimination can lead to many products. Although third components, e.g., esters, are very effective selectivity control agents, they also greatly depress catalyst activity and increase catalyst decay rate (Table 1, 14.5.3.3). Stereoregulation in stereospecific polymerization of propylene can arise from interactions between entering monomer and the growing chain, between monomer and transition metal, or from both factors,. Because of the complete regioselectivity (only head-to-tail enchainments), the stereospecificity in the polymerization of a-olefins to isotactic polymers indicates that OCH, I

OCH, I

I (C2H5)2Al-O-C-C2H,

I

C2H5

(C2H5),A1-0 -C -C2H5

I

0C2H5

Scheme 1. The alkylation of ethyl anisate with aluminumalkyls.


308

14.5.3. Olefin Polymerization 14.5.3.3. Propylene Polymerization 14.5.3.3.1. Heterogeneous Catalysts.

5. G. Natta, P. Pino, G. Mazzanti, P. Longi, Gazz. Chim. Iral., 87, 570 (1957). 6. J. C. W. Chien, Preparations and Properties of Stereoregular Polymers, R. W.

delli, eds., D. Reidel, Dordrecht, 1980, p. 115.

Lenz, F. Ciar-

14.5.3.3.1. Heterogeneous Catalysts.

Highly active catalysts are obtained by ball milling MgCl, with TiCl, to form electron donor complexes such as TiCl, ethyl benzoate'. The crystal size of the milled MgC1, is 3-10 mm. In order to combine high activity with good stereoselectivity, the cocatalyst used is a modified or complexed trialkylaluminoxane with electron donors such as ethyl anisate, ethyl toluate, or ethyl benzoate. Typically, the pale yellow solid catalyst, contains 1-2 wt% Ti and 5-20 wt% ethyl benzoate. Polymerization activity reaches 1,500 kg polypropylene/g Ti, with an isotactic index of 96%. It is often assumed that the electron donors improve stereoselectivity by selectively complexing atactic sites on the catalyst surface2. This may be the mechanism with simple electron donors such as tertiary amines. However, in contrast, esters undergo irreversible reactions with AIR, under polymerization conditions (Scheme 1). Reactions are complex and consecutive steps involving alkylation, reduction, and elimination can lead to many products. Although third components, e.g., esters, are very effective selectivity control agents, they also greatly depress catalyst activity and increase catalyst decay rate (Table 1, 14.5.3.3). Stereoregulation in stereospecific polymerization of propylene can arise from interactions between entering monomer and the growing chain, between monomer and transition metal, or from both factors,. Because of the complete regioselectivity (only head-to-tail enchainments), the stereospecificity in the polymerization of a-olefins to isotactic polymers indicates that OCH, I

OCH, I

I (C2H5)2Al-O-C-C2H,

I

C2H5

(C2H5),A1-0 -C -C2H5

I

0C2H5

Scheme 1. The alkylation of ethyl anisate with aluminumalkyls.

14.5.3. Olefin Polymerization 14.5.3.3. Propylene Polymerization 14.5.3.3.2. Homogeneous Catalysts.

309

the active center discriminates sharply between the two prochiral faces of the a-olefin. Thus, the catalyst system must possess one or more chirality centers (Scheme 2). According to this scheme, the position of diastereomeric and rotameric equilibria in the Tcomplexes and activation energy for the insertion reactions of the .rr-complexed olefin could cause the large regioselectivity and enantioface discrimination necessary for synthesis of the stereoregular poly-cr-olefin. 1. B. L. Goodall, Transition Metal Catalyzed Polymerizafion, MMI Symposium Series, Vol. 4, R. P. Quirk, ed., 1983, p. 355. 2. P. Galli, L. Luciani, G. Cecchin, Angew. Makromol. Chem., 94, 63 (1981). 3. A. Zambelli, P. Locatelli, M. C . Sacchi, I. Tritto, Macromol., 15, 831 (1982). 14.5.3.3.2. Homogeneous Catalysts.

With V-based catalysts, e.g., VC1,/A1C1(C,H5),/aniso, polymers are obtained that consist of syndiotactic stereoblocks and stereo-irregular blocks. Interesting changes occur with T . Chiral titanocenes, zirconocenes, and hafnocenes in combination with methylaluminoxane [Al(CH,)-0],, can lead to highly isotactic propylene. Nonchiral metallocenes like (Cp),ZrCl, or other similar compounds produce only pure atactic polypropylenes. Molecular mass of 590,000 for atactic polypropylenes can be achieved by low polymerization TI. The activities of these hydrocarbon soluble catalysts are extremely high. Different structures of polypropylenes are obtained when the T-bonded ligand of the transition metal is varied (Fig. 1). With no other catalyst can atactic, isotactic, stereoblock, isoblock, and syndiotactic polypropylene of such purity be produced. The first chiral bridged zirconocene used as an isospecific polymerization catalyst was the racemic mixture of ethylenebis(4J ,6,7-tetrahydro-1-indenyl)ZrC12-4. Other chiral metallocenes have dimethylsilyl bridges, indenyl, or substituted cyclopentadienyl rings5-’.

M = Zr, Hf; R, = CH,, (CH,),, (CH,), The propene units are bonded mainly head to tail. A small part (0.7%)is incorporated in 1,3 enchainment*. MW of the polypropylene depends strongly on reaction T. At T between - 20 and 60°C MW range from 300,000 to 15009.Hafnocenes give polypropylenes with a higher MW”. Stereoblock and isoblock polypropylenes were obtained by bis(neomenthy1) ZrC1, or dimethyl silyl(cyclopentadienylindenyl)ZrC12. High yields and high activities for the polymerization of syndiotactic polypropylene were obtained when using metallocene-methylaluminoxane catalysts. One +bonded ligand of the Zr or Hf has to be a sterically demanding fluorenyl. The compound isopropyl (cyclopentadienyl-fluorenyl)Zr(1V)- or -Hf(IV)Cl, together with aluminoxane give high yields of syndiotactic polypropylenes’ I .



309

the active center discriminates sharply between the two prochiral faces of the a-olefin. Thus, the catalyst system must possess one or more chirality centers (Scheme 2). According to this scheme, the position of diastereomeric and rotameric equilibria in the Tcomplexes and activation energy for the insertion reactions of the .rr-complexed olefin could cause the large regioselectivity and enantioface discrimination necessary for synthesis of the stereoregular poly-cr-olefin. 1. B. L. Goodall, Transition Metal Catalyzed Polymerizafion, MMI Symposium Series, Vol. 4, R. P. Quirk, ed., 1983, p. 355. 2. P. Galli, L. Luciani, G. Cecchin, Angew. Makromol. Chem., 94, 63 (1981). 3. A. Zambelli, P. Locatelli, M. C . Sacchi, I. Tritto, Macromol., 15, 831 (1982). 14.5.3.3.2. Homogeneous Catalysts.

With V-based catalysts, e.g., VC1,/A1C1(C,H5),/aniso, polymers are obtained that consist of syndiotactic stereoblocks and stereo-irregular blocks. Interesting changes occur with T . Chiral titanocenes, zirconocenes, and hafnocenes in combination with methylaluminoxane [Al(CH,)-0],, can lead to highly isotactic propylene. Nonchiral metallocenes like (Cp),ZrCl, or other similar compounds produce only pure atactic polypropylenes. Molecular mass of 590,000 for atactic polypropylenes can be achieved by low polymerization TI. The activities of these hydrocarbon soluble catalysts are extremely high. Different structures of polypropylenes are obtained when the T-bonded ligand of the transition metal is varied (Fig. 1). With no other catalyst can atactic, isotactic, stereoblock, isoblock, and syndiotactic polypropylene of such purity be produced. The first chiral bridged zirconocene used as an isospecific polymerization catalyst was the racemic mixture of ethylenebis(4J ,6,7-tetrahydro-1-indenyl)ZrC12-4. Other chiral metallocenes have dimethylsilyl bridges, indenyl, or substituted cyclopentadienyl rings5-’.

M = Zr, Hf; R, = CH,, (CH,),, (CH,), The propene units are bonded mainly head to tail. A small part (0.7%)is incorporated in 1,3 enchainment*. MW of the polypropylene depends strongly on reaction T. At T between - 20 and 60°C MW range from 300,000 to 15009.Hafnocenes give polypropylenes with a higher MW”. Stereoblock and isoblock polypropylenes were obtained by bis(neomenthy1) ZrC1, or dimethyl silyl(cyclopentadienylindenyl)ZrC12. High yields and high activities for the polymerization of syndiotactic polypropylene were obtained when using metallocene-methylaluminoxane catalysts. One +bonded ligand of the Zr or Hf has to be a sterically demanding fluorenyl. The compound isopropyl (cyclopentadienyl-fluorenyl)Zr(1V)- or -Hf(IV)Cl, together with aluminoxane give high yields of syndiotactic polypropylenes’ I .


-

a ' I

I 2 u I $ z-u-u I 2 u $u-u--2 1 I E

a '

I

i

m I 9 3-u-" I

;"

I 9

2-u-u

I

El

f-.1 h

. .Y

X

a '

I E

-

I

gl

Y

I

a '

I S 2-u-u

I

iI

z-u-u $ I

iI E

3. h

Y

Scheme 2. Possible diastereomeric wcomplexes and insertion reactions leading to isotactic or syndiotactic polypropylene:P,,, polymer chain; [MI, metal atom with ligands X and Y.

31 0


atactic

< l1111l11111111l111

isotactic syndiotactic

31 1

2

1

stereo block is0 block

k

l

Figure 1. Different structures of polypropylene.

The stereosequences indicate site stereochemical control with chain migratory insertions, which result in site isomerization and occasional reversal in diastereoface selectivity. With the zirconocene, the activity was found to be 56 kg PP/mmol Zr. Even cyclic olefins such as cyclobutene, cyclopentene, or norbornene could be polymerized with the chiral catalysts to high melting polymers (mp > 400°C)'2-'4. (W. KAMINSKY, R. KRAMOLOWSKY) 1. W. Kaminsky, in History of Polyolefins, R. B. Seymour, T. Cheng eds., Reidel, New York, 1986. 2. J. A. Ewen, J . Am. Chem. SOC., 106,6355 (1984). 3. F. R. W. P. Wild, M. Wasiucionek, G . Huttner, H. H. Brintzinger, J . Organomet. Chem., 288, 63 (1985). 4. W. Kaminsky, K. Kiilper, S. Niedoba, Makromol. Chem. Macromol Symp., 3, 377 (1986). 5. W. A. Hemnann, J. Rohrmann, E. Herdtweck, W. Spaleck, A. Winter, Angew. Chem. Int. Ed. Engl., 28, 1511 (1989). 6. T. Mise, S.Miya, H. Yamazaki, Chem. Lett. 1853 (1989). 7. W. Spaleck, M. Antberg, V. Dolle, R. Klein, J. Rohrmann, A. Winter, New J . Chem., 14, 499 (1990). 8 . K. Soga, T. Shiono, M. S. Takemura, W. Kaminsky, Makromol. Chem. Rapid Comrnun., 8, 305 (1987). 9. B. Rieger, J. C. W. Chien, Polym. Bull., 21, 159 (1989). 10. J. A. Ewen, L. Haspeslagh, J. L. Atwood, H. Zhang, J . Am. Chem. SOC.,109, 6544 (1987). 11. J. A. Ewen, R. L. Jones, A. Razavi, J. P. Ferrara, J . Am. Chem. SOC., 110,6255 (1988). 12. W. Kaminsky, R. Spiehl, Makromol. Chem., 190, 515 (1989). 13. W. Kaminsky, N. Moller-Lindenhof, Bull. SOC. Chim. Belg., 99, 103 (1990). 14. W. Kaminsky, A. Bark, M. Arndt, Makromol. Chem. Macromol. Symp., 47, 83 (1991).


312

14.5.3. Olefin Polymerizations 14.5.3.4. Butadiene Polymerization 14.5.3.4.2. By Cobalt and Nickel Catalysts.

14.5.3.4. Butadiene Polymerlzation 14.5.3.4.1. By Titanium Catalysts.

Polymerization of butadiene can produce three basic polymer structures: -CH,\ CH,=CH-CH=CH, butadiene

?

/CH,-

H/c=c\H

A

(a)

cis- 1,4-polybutadiene

/H

-CH'\

(b) H/c=c\cH,-

trans- 1,Cpolybutadiene -CH-CH2-

I

CH= CH,

(c)

1,2-polybutadiene From the 1,2-polybutadiene, analogous to polypropylene, there are three structurally different polymers: the isotactic, syndiotactic, and atactic form. All have been isolated'.'. With TiC1, and AlEt,, mainly 1,4-polybutadiene is formed. When TiCl, is replaced by Ti14, the 1,4-structure content increases to 90-94%. For Ziegler-type catalysts based on Ti there are examples of monodentate trans- or bidentate cis-coordination of the diolefin to the Ti which, respectively, gives a trans- or cis-configuration in the polymer3v4. The halide y-TiCl,, which offers only one coordination site at the active center, promotes formation of trans-l,4-polymers; in contrast P-TiCl,, which provides more sites, favors formation of a mixture of homo-cis and homotrans polymers596.Butadiene, isoprene, and 2,3-dimethylbutadiene can be treated in the same manner. Rare earth catalysts, e.g., Nd compounds, are also used*. (W. KAMINSKY, R. KRAMOLOWSKY) 1. A. Zambelli, J . Polym. Sci. Part A,, I , 403 (1963). 2. H. Sinn, W. Kaminsky, Adv. Organomet. Chem., 18, 99 (1980). 3. P. Cossee, in Stereochemistry of Macromolecules, D. Eethey, ed., Vol. 1, Decker, New York, 1967, p. 145. 4. E. J. Adman, J . Catal., 5 , 178 (1966). 5. R. T. K. Baker, P. S . Harries, R. J. Waite, J . Polym. Sci. Polym. Lett. Ed., 11, 45 (1973). 6. S . Lin, Q. Wu, L. Sun, in Catalytic Olefin Polymerization, T. Keii, K. Soga, eds., Kodansha, Tokyo, 1990, p. 245. 7. L. Porri, M. C. Gallazzi, S . Destri, A. Bolognesi, in Transition Metal Catalyzed Polymerization, MMI Symposium Series, Vol. 4, R. P. Quirk, ed., 1983, p. 555. 8. N. Nagata, T. Kobatake, H. Watanabe, A. Ueda, A. Yoshioka. Rubber Chem. Technol., 60,837 (1987).

14.5.3.4.2. By Cobalt and Nickel Catalysts.

Polymerization of butadiene by CoCl, uses AIR,, AlXR,, or AlX,R as alkylaluminum cocatalyst to give high cis-l,4-p0lyrner'-~.If the Cl/Al ratio is greater than 1


312

14.5.3. Olefin Polymerizations 14.5.3.4. Butadiene Polymerization 14.5.3.4.2. By Cobalt and Nickel Catalysts.

14.5.3.4. Butadiene Polymerlzation 14.5.3.4.1. By Titanium Catalysts.

Polymerization of butadiene can produce three basic polymer structures: -CH,\ CH,=CH-CH=CH, butadiene

?

/CH,-

H/c=c\H

A

(a)

cis- 1,4-polybutadiene

/H

-CH'\

(b) H/c=c\cH,-

trans- 1,Cpolybutadiene -CH-CH2-

I

CH= CH,

(c)

1,2-polybutadiene From the 1,2-polybutadiene, analogous to polypropylene, there are three structurally different polymers: the isotactic, syndiotactic, and atactic form. All have been isolated'.'. With TiC1, and AlEt,, mainly 1,4-polybutadiene is formed. When TiCl, is replaced by Ti14, the 1,4-structure content increases to 90-94%. For Ziegler-type catalysts based on Ti there are examples of monodentate trans- or bidentate cis-coordination of the diolefin to the Ti which, respectively, gives a trans- or cis-configuration in the polymer3v4. The halide y-TiCl,, which offers only one coordination site at the active center, promotes formation of trans-l,4-polymers; in contrast P-TiCl,, which provides more sites, favors formation of a mixture of homo-cis and homotrans polymers596.Butadiene, isoprene, and 2,3-dimethylbutadiene can be treated in the same manner. Rare earth catalysts, e.g., Nd compounds, are also used*. (W. KAMINSKY, R. KRAMOLOWSKY) 1. A. Zambelli, J . Polym. Sci. Part A,, I , 403 (1963). 2. H. Sinn, W. Kaminsky, Adv. Organomet. Chem., 18, 99 (1980). 3. P. Cossee, in Stereochemistry of Macromolecules, D. Eethey, ed., Vol. 1, Decker, New York, 1967, p. 145. 4. E. J. Adman, J . Catal., 5 , 178 (1966). 5. R. T. K. Baker, P. S . Harries, R. J. Waite, J . Polym. Sci. Polym. Lett. Ed., 11, 45 (1973). 6. S . Lin, Q. Wu, L. Sun, in Catalytic Olefin Polymerization, T. Keii, K. Soga, eds., Kodansha, Tokyo, 1990, p. 245. 7. L. Porri, M. C. Gallazzi, S . Destri, A. Bolognesi, in Transition Metal Catalyzed Polymerization, MMI Symposium Series, Vol. 4, R. P. Quirk, ed., 1983, p. 555. 8. N. Nagata, T. Kobatake, H. Watanabe, A. Ueda, A. Yoshioka. Rubber Chem. Technol., 60,837 (1987).

14.5.3.4.2. By Cobalt and Nickel Catalysts.

Polymerization of butadiene by CoCl, uses AIR,, AlXR,, or AlX,R as alkylaluminum cocatalyst to give high cis-l,4-p0lyrner'-~.If the Cl/Al ratio is greater than 1

14.5.3. Olefin Polymerizations 14.5.3.4. Butadiene Polymerization 14.5.3.4.3. By Lithium.

313

and the Al/Co ratio more than 100, the catalyst is soluble in hydrocarbons. Thus, catalysts prepared from CoCl,/R,Al appear heterogeneous whereas those from CoCl,/(CH,), ,5A1 C1, or Co(I1)octoate are homogeneous. Even though CoCl, forms a stable blue solution in methyl aluminum sesquichloride, polymerization in the presence of butadiene is colorless and gives an optically clear polybutadiene that analyzed as 99.8% cis-1,4 structure4. CoCl, is also soluble in other aluminum alkyls but the solutions are unstable and the CoCl, is rapidly reduced. A mechanism is proposed for the polymerization that involves two metal alkyl units associated with a Co atom differing in Lewis acidity such as AlClR, AlC1,R. The second growth steps involve 1,4-addition of the coordinated butadiene to a butadienyl Co. The Co atom acts as carrier of the cis-coordinated butadiene and confers increased polarizability on the butadiene molecule. Polybutadiene containing a high content of trans- 1,4 structure has been prepared with [(q3-allyl)NiX], (X = C1, Br, I). Since [(q3-allyl)NiX], exists in the dimer state, it breaks into the monomer state only after coordination with butadiene. Butadiene can be polymerized with (17,-allyl) Ni compounds without base metal alkyls being present5v6. NMR studies show that the complex carrying the growing chain in the butadiene polymerization with bis[(q3-ally1)nickel-trifluoracetate] is predominantly in the form of a binuclear syn-( q3-ally1)nickel complex. With [(q3-allyl)NiI], polybutadiene is formed with 97% truns-l,4 microstructure’.

+


1 . S . E. Home, in Transition Metal Catalyzed Polymerization, MMI Symposium Series, Vol. 4, R. P. Quirk, ed., 1983, p. 527. 2. M. Gippin, Rubber Chem. Technol., 39,508 (1966). 3. F. Borg-Visse, F. Dawans, E. Marechal, J . Polym. Sci., Polym. Chem. Ed., 18, 2491 (1980). 4. K. Soga, K. Yamamoto, Polym. Bull., 6, 263 (1982). 5. E. Kobayashi, J. Furukawa, M. Ochiai, T. Tsujimoto, Euro. Polym. J., 19, 871 (1983). 6. Y. G. Li, G. Yu, J . Mucromol. Sci. Chem. Part A , 26,405 (1989). 7. B. A. Dolgoplosk, S . I. Beilein, G. M. Chemenko, J . Polym. Sci., Polym. Chem. Ed., 11, 2569 (1973).

14.5.3.4.3. By Lithium.

It is necessary to distinguish between (1) initiation-insertion of the monomer into the L i - C bond of the metal alkyl and formation of a alkenyl metal and (2) propagationthe continuing addition of monomer to the growing center, which leads to an ever lengthening polymer chain. Since no termination takes place, the term “living polymers”’ is used. Extremely high MW, up to several million and very narrow MW distributions can be obtained. The microstructure of the polymer is influenced by the reaction conditions generally as2 (1) 1,Zstructures increase markedly with greater polarity of the solvent, and (2) the amount of lP-structure obtained from polymerization in a hydrocarbon solvent increase with decreasing initiator concentration. For the n-BuLi/butadiene/heptane system, a yield of less than 10% 1,2-structures and a maximum of 60% 1,4-structures in the trans-configuration is typically obtained. Using high monomer concentrations (10 mol/liter) and low initiator concentrations ( lop5mol/liter) the trans-configuration is lowered to 20%. To explain the kinetic results, association of the organolithium plays an important role and leads to a model in which the influence of the concentration on stereoselectivity is described3g4.The microstructure can be varied desirably by addition of complexing reagents5-’. In the t-BuLi/buta-


14.5.3. Olefin Polymerizations 14.5.3.4. Butadiene Polymerization 14.5.3.4.3. By Lithium.

313

and the Al/Co ratio more than 100, the catalyst is soluble in hydrocarbons. Thus, catalysts prepared from CoCl,/R,Al appear heterogeneous whereas those from CoCl,/(CH,), ,5A1 C1, or Co(I1)octoate are homogeneous. Even though CoCl, forms a stable blue solution in methyl aluminum sesquichloride, polymerization in the presence of butadiene is colorless and gives an optically clear polybutadiene that analyzed as 99.8% cis-1,4 structure4. CoCl, is also soluble in other aluminum alkyls but the solutions are unstable and the CoCl, is rapidly reduced. A mechanism is proposed for the polymerization that involves two metal alkyl units associated with a Co atom differing in Lewis acidity such as AlClR, AlC1,R. The second growth steps involve 1,4-addition of the coordinated butadiene to a butadienyl Co. The Co atom acts as carrier of the cis-coordinated butadiene and confers increased polarizability on the butadiene molecule. Polybutadiene containing a high content of trans- 1,4 structure has been prepared with [(q3-allyl)NiX], (X = C1, Br, I). Since [(q3-allyl)NiX], exists in the dimer state, it breaks into the monomer state only after coordination with butadiene. Butadiene can be polymerized with (17,-allyl) Ni compounds without base metal alkyls being present5v6. NMR studies show that the complex carrying the growing chain in the butadiene polymerization with bis[(q3-ally1)nickel-trifluoracetate] is predominantly in the form of a binuclear syn-( q3-ally1)nickel complex. With [(q3-allyl)NiI], polybutadiene is formed with 97% truns-l,4 microstructure’.

+


1 . S . E. Home, in Transition Metal Catalyzed Polymerization, MMI Symposium Series, Vol. 4, R. P. Quirk, ed., 1983, p. 527. 2. M. Gippin, Rubber Chem. Technol., 39,508 (1966). 3. F. Borg-Visse, F. Dawans, E. Marechal, J . Polym. Sci., Polym. Chem. Ed., 18, 2491 (1980). 4. K. Soga, K. Yamamoto, Polym. Bull., 6, 263 (1982). 5. E. Kobayashi, J. Furukawa, M. Ochiai, T. Tsujimoto, Euro. Polym. J., 19, 871 (1983). 6. Y. G. Li, G. Yu, J . Mucromol. Sci. Chem. Part A , 26,405 (1989). 7. B. A. Dolgoplosk, S . I. Beilein, G. M. Chemenko, J . Polym. Sci., Polym. Chem. Ed., 11, 2569 (1973).

14.5.3.4.3. By Lithium.

It is necessary to distinguish between (1) initiation-insertion of the monomer into the L i - C bond of the metal alkyl and formation of a alkenyl metal and (2) propagationthe continuing addition of monomer to the growing center, which leads to an ever lengthening polymer chain. Since no termination takes place, the term “living polymers”’ is used. Extremely high MW, up to several million and very narrow MW distributions can be obtained. The microstructure of the polymer is influenced by the reaction conditions generally as2 (1) 1,Zstructures increase markedly with greater polarity of the solvent, and (2) the amount of lP-structure obtained from polymerization in a hydrocarbon solvent increase with decreasing initiator concentration. For the n-BuLi/butadiene/heptane system, a yield of less than 10% 1,2-structures and a maximum of 60% 1,4-structures in the trans-configuration is typically obtained. Using high monomer concentrations (10 mol/liter) and low initiator concentrations ( lop5mol/liter) the trans-configuration is lowered to 20%. To explain the kinetic results, association of the organolithium plays an important role and leads to a model in which the influence of the concentration on stereoselectivity is described3g4.The microstructure can be varied desirably by addition of complexing reagents5-’. In the t-BuLi/buta-

314

14.5 Olefin Transformations 14.5.4. Olefin Metathesis

diene/cyclopentane system, addition of 1,2-dipiperidinoethane raises the yield of 1,2polybutadiene from about 15 to 99%8. Functional groups can be attached to the living polymer chains by choosing suitable termination reagents, resulting in linear or star-shaped molecules. With bi- or multifunctional organolithium initiators’, a greater variety of polymer products can be synthesized, e.g., well-defined polymer networks. (W. KAMINSKY, R. KRAMOLOWSKY) 1. 2. 3. 4.

5.

6. 7. 8.

9.

M. Szwarc, Adv. Polym. Sci., 49, 7 (1983). S. Bywater, Adv. Polym. Sci., 4 , 6 (1966). W. Gebert, J. Hinz, H. Sinn,Makromol. Chem., 144, 97 (1971). D. J. Worsfold, S. Bywater, Macromolecules, 1 1 , 582 (1978). S. Raynal, J. Macrornol. Sci., A 19, 1049 (1983). D. H. Richards, M. S . Stewart, Polymer, 24, 883 (1983). G. Jin, W. Wang, J. Polym. Muter, 6 , 31 (1989). D. J. Worsfold, S. Bywater, F. Schue, J. Sledz, V. Marti-Collet, Makromol. Chem., Rapid Cornmun., 3, 239 (1982). P. Lutz,E. Franta, P. Rempp, Polymer, 23, 1953 (1982).

14.5.4. Olefin Metathesis Material for this section is projected for publication in a final Supplementary unit of the Inorganic Reactions and Methods series.


14.6. Carbon Monoxide Reactions 14.6.1. Introduction Carbon monoxide is a peculiar molecule in many respects, and its reactivity cannot be properly understood unless some of its ground-state physical and bonding properties are taken into consideration’. Some of the relevant physical properties of carbon monoxide are displayed in Table 1. Carbon monoxide is the dinuclear molecule with the highest dissociation energy; its dipole moment is, on the other hand, small, partially explaining its low degree of reactivity. The C-0 bond distance is 112.8 pm, corresponding to bond order three, with an MO electronic configuration (a,)’ (a,*)’ (a,)’

(r2.

(rJ2

+

Carbon monoxide can be regarded as containing carbon in the oxidation state 11, and therefore intermediate between CO, ( + IV) and the element. The molecule can undergo both oxidations and reductions to C,-C,, organic compounds (vide infra). The electronic structure of carbon monoxide on valence bond terms can be visualized as I and 11: ICE01

IC=O \

(1)

(11)

/

In I, the electron pair base character of CO can arise from donation of the lone pair on carbon, while its acid character can arise from utilization of an empty p-orbital on carbon. In MO terms the r-acidity of CO originates from the interaction of filled orbitals of appropriate symmetry with empty, antibonding orbitals, ( r y *or) ( rz*), of CO. Despite lone pairs located formally at both the carbon and the oxygen ends, carbon monoxide behaves as a base only at carbon in metal c a r b o n y l ~However, ~ ~ ~ . spectroscopic evidence points to the existence of linkage isomerism for carbon monoxide in Au(CO)(OC) obtained by cocondensation of gold vapors and CO at approximately 10K4. When coordinated to transition metals, the electronic properties are drastically modified, and a change of reactivity is expected accordingly. The reactions known for carbon monoxide can be classified under the following headings: 1. 2. 3. 4.

oxidation and disproportionation (14.6.1.1.), reduction (14.6.1.2.), base-catalyzed reactions (14.6.1.3.), acid-catalyzed reactions (14.6.1.4.),

TABLE1. SOME PROPERTIES

OF CARBON

Dissociation energy (kJ/mol) Ionization potential (!d/mol) Dipole moment (0) C-0 bond distance

MONOXIDE 1070 1352 0.112 112.8 pm

31 5

316

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1.1. Oxidation and Disproportionationof Carbon Monoxide

5 . reactions with transition metals (14.6.1.4.), 6. coordinative addition (14.6. lS.), 7. chemisorption (14.6.1.7.) and 8. metal-catalyzed reactions (14.6.1.8.). (F. CALDERAZZO)

1. 2. 3. 4.

M. Orchin, I. Wender, in Catalysis, Vol. 5, P. H. Emmett, ed., Reinhold, New York, 1957, p. 1. J. Ladell, B. Post, I. Fankuchen, Acta Crystallogr., 5, 795 (1952). J. Donohue, A. Caron, Acta Crystallogr., 17, 663 (1964). D. McIntosh, G . A. Ozin, Inorg. Chem., 16, 51 (1977).

14.6.1.l.Oxidation and Disproportionation of Carbon Monoxide

Some oxidations of CO are1-,.

co + c1, AHSmK

= -109.7 kJ

co +

55 0,

AG298K = -257.1 kJ

-

COC1,

(a)

co,

Although thermodynamically favored, these reactions are slow at ordinary temperature, and they require the use of a catalyst, e.g., phosgene can be obtained already at 40°C when active coal is used as a catalyst in equation (a). The Au-catalyzed oxidation of CO to phosgene has been carried out4 under exclusion of light, to avoid any possible interference with the established fact that the CO-C1, mixture reacts at ordinary temperature under irradiation in the visible region. With Br,, less than 10% of COBr, is present at equilibrium, the attainment of equilibrium is slow and is not accelerated by irradiation. Diiodine (I,) does not react with CO at ordinary temperature under UV irradiation. Combustion of CO to CO, can be carried out at temperatures as low as 50°C by using catalytic systems based on Pt or Pd. Carbon monoxide disproportionates to C and CO,:

+ co,

2co-c

(c)

AGZggK= - 120.1 kJ/mol

Although the reaction is thermodynamically favored, it is slow under ordinary conditions of temperature and pressure; over heterogeneous catalysts based on Fe, Co, and Ni, the disproportionation reaction proceeds at reasonable rates between 400 and 600°C. Carbon monoxide can be oxidized to CO, by metal oxides, such as those of iron and o ~ m i u m ~ . ~ : FeO

+ CO

oso, + 9 co

--

+ CO, Os(CO), + 4 CO, Fe

(4 (el


316

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1.1. Oxidation and Disproportionationof Carbon Monoxide

5 . reactions with transition metals (14.6.1.4.), 6. coordinative addition (14.6. lS.), 7. chemisorption (14.6.1.7.) and 8. metal-catalyzed reactions (14.6.1.8.). (F. CALDERAZZO)

1. 2. 3. 4.

M. Orchin, I. Wender, in Catalysis, Vol. 5, P. H. Emmett, ed., Reinhold, New York, 1957, p. 1. J. Ladell, B. Post, I. Fankuchen, Acta Crystallogr., 5, 795 (1952). J. Donohue, A. Caron, Acta Crystallogr., 17, 663 (1964). D. McIntosh, G . A. Ozin, Inorg. Chem., 16, 51 (1977).

14.6.1.l.Oxidation and Disproportionation of Carbon Monoxide

Some oxidations of CO are1-,.

co + c1, AHSmK

= -109.7 kJ

co +

55 0,

AG298K = -257.1 kJ

-

COC1,

(a)

co,

Although thermodynamically favored, these reactions are slow at ordinary temperature, and they require the use of a catalyst, e.g., phosgene can be obtained already at 40°C when active coal is used as a catalyst in equation (a). The Au-catalyzed oxidation of CO to phosgene has been carried out4 under exclusion of light, to avoid any possible interference with the established fact that the CO-C1, mixture reacts at ordinary temperature under irradiation in the visible region. With Br,, less than 10% of COBr, is present at equilibrium, the attainment of equilibrium is slow and is not accelerated by irradiation. Diiodine (I,) does not react with CO at ordinary temperature under UV irradiation. Combustion of CO to CO, can be carried out at temperatures as low as 50°C by using catalytic systems based on Pt or Pd. Carbon monoxide disproportionates to C and CO,:

+ co,

2co-c

(c)

AGZggK= - 120.1 kJ/mol

Although the reaction is thermodynamically favored, it is slow under ordinary conditions of temperature and pressure; over heterogeneous catalysts based on Fe, Co, and Ni, the disproportionation reaction proceeds at reasonable rates between 400 and 600°C. Carbon monoxide can be oxidized to CO, by metal oxides, such as those of iron and o ~ m i u m ~ . ~ : FeO

+ CO

oso, + 9 co

--

+ CO, Os(CO), + 4 CO, Fe

(4 (el

317

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1.2. Reductions of Carbon Monoxide

Equation (e) is one of the basic reactions occurring during the metallurgical processing of iron. Also, CO can be chlorinated by gold(II1) chloride at ordinary temperature and pressure7? AuC13

+ 2 CO

Au(C0)Cl

+ COCl,

(f)

These reactions are believed to involve activation of CO by the metal and they will be discussed again in 14.6.1.5. (F. CALDERAZZO)

1. Kirk-Othmer Encyclopedia of Chemical Technology, Vol. 4, 3rd ed., Kirk-Othmer Wiley-Interscience, New York, 1978, p. 772. 2. Kirk-Othmer Encyclopedia of Chemical Technology, 2nd ed., Kirk-Othmer Wiley-Interscience,

New York, 1964, p. 424. 3. Gmelins Handbuch der Anorganischen Chemie, Kohlenstoff, Teil C2, Verlag Chemie, Weinheim, 1972. 4. F. Calderazzo, D. Belli Dell’Amico, Inorg. Chem., 21, 3639 (1982). 5 . W. Hieber, H. Stallmann, Z. Elektrochem., 49,288 (1943). 6. F. Calderazzo, F. L’Eplattenier,Inorg. Chem., 6, 1220 (1967). 7. D. Belli Dell’Amico, F. Calderazzo, Gazz. Chim. Ital., 103, 1099 (1973). 8. D. Belli Dell’Amico, F. Calderazzo, F. Marchetti, J . Chem. SOC.,Dalton Trans., 1829 (1976). 14.6.1.2. Reductions of Carbon Monoxide

Carbon monoxide undergoes reduction with electropositive metals such as alkali and alkaline earth metals to yield products originally known as alkali carbonyls’, whose nature was later understood in terms of reduction with simultaneous formation of carbon-carbon bonds (a reductive coupling):

n CO

+ n e-

-+

[(CO),Jne-

(a)

The products with n = 2 and 6 are described. The first reaction of this type (1834) treated carbon monoxide with molten potassium’: 6 CO

+6K

C,5K,jO6

(b)

then with dilute hydrochloric acid to yield hexahydroxybenzene: 6 H+

+ C6K6O.5

--+

-

C6H60.5

+

6 H+

(c)

Liquid NH3 solutions3 of K with CO give4 the potassium salt of the unstable acetylen-dio15 2 CO

+ 2K

[OCrCO]’- 2K’

(4

The product is a yellow solid, extremely sensitive to air and moisture, characterized by X-ray powder methods to give the following distances (pm): K-0, 267 and 278; K-K, 365; C-0, 128; C-C, 121. Treatment with water results in the formation of glycolic acid3. The salt obtained from molten cesium and CO also gives glycolic acid on hydrolysis6,’: [OC=CO]’-

+ 3 HZO

-

2 OH-

+ HOH, C-COOH

(el


317

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1.2. Reductions of Carbon Monoxide

Equation (e) is one of the basic reactions occurring during the metallurgical processing of iron. Also, CO can be chlorinated by gold(II1) chloride at ordinary temperature and pressure7? AuC13

+ 2 CO

Au(C0)Cl

+ COCl,

(f)

These reactions are believed to involve activation of CO by the metal and they will be discussed again in 14.6.1.5. (F. CALDERAZZO)

1. Kirk-Othmer Encyclopedia of Chemical Technology, Vol. 4, 3rd ed., Kirk-Othmer Wiley-Interscience, New York, 1978, p. 772. 2. Kirk-Othmer Encyclopedia of Chemical Technology, 2nd ed., Kirk-Othmer Wiley-Interscience,

New York, 1964, p. 424. 3. Gmelins Handbuch der Anorganischen Chemie, Kohlenstoff, Teil C2, Verlag Chemie, Weinheim, 1972. 4. F. Calderazzo, D. Belli Dell’Amico, Inorg. Chem., 21, 3639 (1982). 5 . W. Hieber, H. Stallmann, Z. Elektrochem., 49,288 (1943). 6. F. Calderazzo, F. L’Eplattenier,Inorg. Chem., 6, 1220 (1967). 7. D. Belli Dell’Amico, F. Calderazzo, Gazz. Chim. Ital., 103, 1099 (1973). 8. D. Belli Dell’Amico, F. Calderazzo, F. Marchetti, J . Chem. SOC.,Dalton Trans., 1829 (1976). 14.6.1.2. Reductions of Carbon Monoxide

Carbon monoxide undergoes reduction with electropositive metals such as alkali and alkaline earth metals to yield products originally known as alkali carbonyls’, whose nature was later understood in terms of reduction with simultaneous formation of carbon-carbon bonds (a reductive coupling):

n CO

+ n e-

-+

[(CO),Jne-

(a)

The products with n = 2 and 6 are described. The first reaction of this type (1834) treated carbon monoxide with molten potassium’: 6 CO

+6K

C,5K,jO6

(b)

then with dilute hydrochloric acid to yield hexahydroxybenzene: 6 H+

+ C6K6O.5

--+

-

C6H60.5

+

6 H+

(c)

Liquid NH3 solutions3 of K with CO give4 the potassium salt of the unstable acetylen-dio15 2 CO

+ 2K

[OCrCO]’- 2K’

(4

The product is a yellow solid, extremely sensitive to air and moisture, characterized by X-ray powder methods to give the following distances (pm): K-0, 267 and 278; K-K, 365; C-0, 128; C-C, 121. Treatment with water results in the formation of glycolic acid3. The salt obtained from molten cesium and CO also gives glycolic acid on hydrolysis6,’: [OC=CO]’-

+ 3 HZO

-

2 OH-

+ HOH, C-COOH

(el

318

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1.3. Base-Catalyzed Reactions of Carbon Monoxide

Carbon monoxide at superatmospheric pressure can be reduced electrochemicall? in a specially designed9, high-pressure electrolytic cell. The reaction yields the conjugate base of dihydroxycyclobutenedione (squaric acid):

4CO+2e-

-

-

at RT by using the following electrochemical system: (Anode)Al or MglCO(50-300 a m ) , aprotic solvent, Bu,NBr(stainless steel (cathode) The reaction mixture is evaporated to dryness under reduced pressure and the squaric acid obtained in 60% yield with respect to absorbed CO after treatment of the solid residue with dilute hydrochloric acid. Reduction of CO has been further confirmed"; reduction to telomerized radical anions on activated magnesium or calcium oxides has been reported". Reductive coupling of coordinated CO on Nb12,Ta", or V13 complexes gives a coordinated bis(trimethylsi1oxy)ethyne ligand. Carbon monoxide reacts at atmospheric pressure with the radical anions of condenses aromatic hydrocarbons in ether solvents, e.g., the sodium-naphthalene-tetrahydrofuran system, is used for promoting the reduction of CO',, to give products with new carbon-carbon bonds. Among the products with the naphthalene radical anion, dihydronaphthalene dicarboxylic acids and oxalic acid are isolated after oxidation with air and treatment with water. (F. CALDERAZZO)

1. M. Orchin, I. Wender, in Catalysis, Vol. 5, P. H. Emmett, ed., Reinhold, New York, 1957, p. 1. 2. J. Liebig, Ann. Chem. Pharm., 11, 82 (1834). 3. A. Joannis, C . R . Hebd. Stances. Acad. Sci., 158, 874 (1914). 4. E. Weiss, W. Buchner, Helv. Chim. Acta, 46, 1121 (1963). 5. W. Buchner, Helv. Chim. Acta, 48, 1229 (1965). 6. L. Hackspill, L. A. van Altena, C. R. Hebd. Stances Acad. Sci., 206, 1818 (1938). 7. W. Buchner, Chem. Ber., 98, 31 18 (1965). 8. G . Silvestri, S. Gambino, G . Filardo, M. Guainazzi, R. Ercoli, Gazz. Chim. Ital., 102, 818

(1972). 9. M. Guainazzi, G. Silvestri, S.Gambino, G . Filardo, J . Chem. SOC.,Dalton Trans., 927 (1972). 10. P. W. Lednor, P. C. Versloot, J . Chem. SOC. Chem. Commun., 284 (1983). 11. R. M. Moms, K. J. Klabunde, J . Am. Chem. SOC., 105,2633 (1983). 12. P. A. Bianconi, R. N. Vrtis, Ch. P. Rao, I. D. Williams, M. P. Engeler, S.J. Lippard, Organometallics, 6 , 1968 (1977). 13. J. D. Protasiewicz, S.J. Lippard, J . Am. Chem. SOC., 113, 6564 (1991). 14. W. Biichner, Chem. Ber., 99, 1485 (1966). 14.6.1.3. Base-Catalyzed Reactions of Carbon Monoxide

The insertions of CO into the 0-H or the N-H bonds of alcohols and amines are base catalyzed and lead to derivatives of formic acid:


318

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1.3. Base-Catalyzed Reactions of Carbon Monoxide

Carbon monoxide at superatmospheric pressure can be reduced electrochemicall? in a specially designed9, high-pressure electrolytic cell. The reaction yields the conjugate base of dihydroxycyclobutenedione (squaric acid):

4CO+2e-

-

-

at RT by using the following electrochemical system: (Anode)Al or MglCO(50-300 a m ) , aprotic solvent, Bu,NBr(stainless steel (cathode) The reaction mixture is evaporated to dryness under reduced pressure and the squaric acid obtained in 60% yield with respect to absorbed CO after treatment of the solid residue with dilute hydrochloric acid. Reduction of CO has been further confirmed"; reduction to telomerized radical anions on activated magnesium or calcium oxides has been reported". Reductive coupling of coordinated CO on Nb12,Ta", or V13 complexes gives a coordinated bis(trimethylsi1oxy)ethyne ligand. Carbon monoxide reacts at atmospheric pressure with the radical anions of condenses aromatic hydrocarbons in ether solvents, e.g., the sodium-naphthalene-tetrahydrofuran system, is used for promoting the reduction of CO',, to give products with new carbon-carbon bonds. Among the products with the naphthalene radical anion, dihydronaphthalene dicarboxylic acids and oxalic acid are isolated after oxidation with air and treatment with water. (F. CALDERAZZO)

1. M. Orchin, I. Wender, in Catalysis, Vol. 5, P. H. Emmett, ed., Reinhold, New York, 1957, p. 1. 2. J. Liebig, Ann. Chem. Pharm., 11, 82 (1834). 3. A. Joannis, C . R . Hebd. Stances. Acad. Sci., 158, 874 (1914). 4. E. Weiss, W. Buchner, Helv. Chim. Acta, 46, 1121 (1963). 5. W. Buchner, Helv. Chim. Acta, 48, 1229 (1965). 6. L. Hackspill, L. A. van Altena, C. R. Hebd. Stances Acad. Sci., 206, 1818 (1938). 7. W. Buchner, Chem. Ber., 98, 31 18 (1965). 8. G . Silvestri, S. Gambino, G . Filardo, M. Guainazzi, R. Ercoli, Gazz. Chim. Ital., 102, 818

(1972). 9. M. Guainazzi, G. Silvestri, S.Gambino, G . Filardo, J . Chem. SOC.,Dalton Trans., 927 (1972). 10. P. W. Lednor, P. C. Versloot, J . Chem. SOC. Chem. Commun., 284 (1983). 11. R. M. Moms, K. J. Klabunde, J . Am. Chem. SOC., 105,2633 (1983). 12. P. A. Bianconi, R. N. Vrtis, Ch. P. Rao, I. D. Williams, M. P. Engeler, S.J. Lippard, Organometallics, 6 , 1968 (1977). 13. J. D. Protasiewicz, S.J. Lippard, J . Am. Chem. SOC., 113, 6564 (1991). 14. W. Biichner, Chem. Ber., 99, 1485 (1966). 14.6.1.3. Base-Catalyzed Reactions of Carbon Monoxide

The insertions of CO into the 0-H or the N-H bonds of alcohols and amines are base catalyzed and lead to derivatives of formic acid:

--

14.6. Carbon Monoxide Reactions 14.6.1. introduction 14.6.1-4. Acid-Catalyzed Reactions of Carbon Monoxide

+ ROH CO + R,NH CO

319

HCOOR

(a)

HCONR,

(b)

These reactions are catalyzed by NaOH and NaOMe, respectively, and occur at 170- 190°C (formate esters) and at 60-130°C (dimethylformamide) at superatmospheric pressure of CO in the range 5-20 atm. N,N-Dimethylformamide is produced industrially' either from the direct reaction (b) or by treatment with Me,NH of the formate ester obtained by the carbonylation reaction (a). The rate of the reaction of MeOH with CO catalyzed by NaOMe3 depends on the concentration of the MeO- ion and on the pressure of CO to the first power. The rates are higher for secondary and tertiary than for primary alcohols. The mechanism is:

[

C CO M-C-H M-H Although this reaction presumably plays an important role in several catalytic processes involving CO and H,6,', and although several formyl complexes of transition metals are known*-", no well-established examples of formyl complexes as obtained by the direct insertion (e) are k n o ~ n ~ ~ J ~ . Lithium borohydride does not react with carbon monoxide at RT in tetrahydr~furan'~.

+

(F. CALDERAZZO) 1. T. H. Coffield, J. Kozikowski, R. D. Closson, J . Org. Chem., 22, 598 (1957). 2. A. Wojcicki, Adv. Organomet. Chem., 11, 87 (1973); F. Calderazzo, Angew. Chem., Int. Ed. Enos., 16, 299 (1977). 3. W. Hieber, W. Beck, E. Lindner, Z. Naturforsch. TeilB, 16, 229 (1961). 4. M. Orchin, I. Wender, in Catalysis, Vol. 5, P. H. Emmett, ed., Reinhold, New York, 1957, p. 1. 5. F. G. Fischer, 0. Stoffers, Ann. Chem., 500, 253 (1930). 6. C. Masters, Adv. Organomet. Chem., 17, 61 (1979). 7. D. R. Fahey, J . Am. Chem. SOC., 103, 136 (1981). 8. J. P. Collman, S.R. Winter, J . Am. Chem. Soc., 95,4089 (1973). 9. C. P. Casey, S. M. Neumann, J . Am. Chern. SOC.,100,2544 (1978). 10. J. A. Gladysz, W. Tarn, J . Am. Chem. SOC.,100,2545 (1978). 11. M. R. Churchill, H. J. Wasserman, J . Chem. SOC.Chem. Commun., 274 (1981). 12. See, however, B. B. Wayland, B. A. Woods, J . Chem. SOC., Chem. Commun., 700 (1981). 13. R. Guilard, K. M. Kadish, Chem. Rev., 88, 1121 (1988), and references therein. 14. M. W. Rathke, H. C. Brown, J . Am. Chem. Soc., 88,2606 (1966).

14.6.1.8. Chemlsorptlon of Carbon Monoxide

By interaction with metallic or nonmetallic (active coal adsorbs 9 ml of CO/g of material at 15OC) surfaces, higher concentrations of CO can normally be detected at the

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1.8. Chernisorption of Carbon Monoxide

325

surface of the solid with respect to the inner sites. Since for surfaces (e.g., transition metals such as Fe, Ni, Co, Pd, Ir, W, and others) the heats of CO adsorption are 105-167/kT/m01'~~,which are comparable to most of the metal-CO bond energies for molecular metal carbonyl complexes3, this phenomenon is chemisorption. Studies of CO chemisorption on metals or metal oxides surfaces are essential for the understanding of heterogeneous chemical processes involving CO, and because of the high dissociation energy of CO, it is often possible to study the chemisorption phenomena on surfaces, prior to dissociation of the molecule. Infrared spectroscopy is used4 to study the CO chemisorbed on metal surfaces: infrared C-0 stretches between 2100 and 1700 cm-' are usually observed, which is the same range of values for molecular metal carbonyl complexes. Terminal and bridging types of chemisorbed CO are present on surfaces as shown in (1)-(5). 0

0

I

c

II

c-0 I I M-M

I

M

M-M

(4) (5) All binding modes (1)-(5) are found in molecular carbonyl complexes by X-ray'. The initial chr:misorption of CO on a single crystal surface of Pt is dissociative at steps and kinks, followed by an associative adsorption at terraces?

-

+

co-c 0 The chemisorbed 0 atoms also react with CO to give gaseous CO,: 0

+ co

CO,

(a)

(b) Work on metal carbonyl clusters' adds to the understanding of chemisorption phenomena of carbon monoxide on metals. With early nontransitional metals and with metals of group 11 having completely filled d-orbitals (Cu,Ag,Au), chemisorption of CO has some peculiar features. Carbon monoxide is rapidly adsorbed on Ca, Sr, and Ba below O°C8. It is possible that these strongly reducing rnetals induce some reduction of CO of the type described in 14.6.1.2. Titanium and Zr are covered by C09-12 close to unity, and the heat of adsorption is the highest among transition elements. Not much is known about the nature of CO chemisorbed on these metals, but with the molecular counterparts: (1) no binary metal carbonyls of Ti, Zr, and Hf are known except in the -11 oxidation state"-", i.e., [M(CO)6]2, (2) these metals in the oxidation state +I1 bind CO strongly in the M( q5 - C,Me,),(CO), (M=Ti,Zr,Hf) c~mplexes'~, characterized by low wavenumber values of the CO stretches and (3) these metals have a high affinity for oxygen, as evidenced, e.g., by the structure of the acyl complexes of TiI4 and Zr15, which have the arrangement shown in 6, with the q2-bonded acyl group:

326

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1 -9.Metal-CatalyzedReactions of Carbon Monoxide

(6)

Group 11 metals8 show only one C-0 stretching vibration at 2105 (Cu), 2160 (Ag), and 21 10 (Au) cm- after the surfaces are exposed to CO. Again, when comparing molecular complexes of these metals, Cu gives carbonyl complexes in its + I oxidation state, (see 14.6.1.6), but only one example of a well-defined carbonyl complex of Ag is known16 and Au forms the carbonyl halo complexes Au(CO)C~'~, and Au(CO)Br'* with an infrared absorption around 2160 cm-', higher than in free CO.

'

(F. CALDERAZZO)

1. 2. 3. 4. 5. 6. 7.

8.

9. 10. 11. 12. 13. 14. 15. 16. 17.

18.

G. Ertl, Gazz. Chim. ltal., 109, 217 (1979), and references therein. R. Ugo, Catal. Rev.-Sci. Eng., 11, 225 (1975). J. A. Connor, Top. Curr. Chem., 71,71 (1977). R. R. Ford, A h . Catal., 21, 51 (1970).

E. L. Muetterties, T. N. Rhodin, E. Band, C. F. Brucker, W. R. Pretzer, Chem. Rev., 79, 91 (1979). Unpublished data by Y.Iwasawa, R. Mason, Textor M., and G . A. Somorjai, quoted in W. L. Jolly, Top. Curr. Chem., 71; 149 (1977). P. Chini, J . Organomet. Chem., 200, 37 (1980), and references therein. B. M. W. Trapnell, Proc. SOC.,Ser. A, 218, 566 (1953); S . Wagener, J . Phys. Chem., 60, 567 (1956). S . Wagener, J . Phys. Chem., 61, 267 (1957). K. M. Chi, S . R. Frerichs, S . B. Wilson, J. E. Ellis, Angew. Chem., Int. Ed. Engl., 26, 1190 (1987). S . B. Wilson, J. E. Ellis, J . Am. Chem. SOC.,110, 303 (1988). J. E. Ellis, K. M. Chi, J . Am. Chem. SOC.,112, 6022 (1990). D. J. Sikora, M. D. Rausch, R. D. Rogers, J. L. Atwood, J . Am. Chem. Soc., 103, 1265 (1981). G. Fachinetti, C. Floriani, H. Stoeckli-Evans, J . Chem. SOC., Dalton Trans., 2297 (1977). G. Fachinetti, C. Floriani, F. Marchetti, S. Merlino, J . Chem. Soc., Chem. Commun., 522 (1976). P. K. Hurlburt, 0. P. Anderson, S . H. S t r a w , J. Am. Chem. Soc., 113, 6277 (1991). D. Belli Dell'Amico, F. Calderazzo, F. Marchetti, J . Chem. Soc., Dalton Trans., 1829 (1976), and references therein. D. Belli Dell'Amico, F. Calderazzo, P. Robino, A. Segre, J . Chem. Soc., Dalton Trans., 3017 ( 1991).

14.6.1.9. Metal-Catalyzed Reactions of Carbon Monoxide

Carbon monoxide becomes a versatile reagent when it is activated by transition metals, giving a large number of both homogeneous and heterogeneous reactions. In the former, reactions occur at metal-coordinated CO; in the latter, chemisorbed CO is usually responsible for the observed catalytic reaction. - The known metal-catalyzed reactions of carbon monoxide can be classified as 1. reactions by which CO is reduced by dihydrogen, 2. additions of CO across double bonds, usually in combination with a hydrogencontaining substrate. 3. reactions in which CO acts as a reducing agent, and 4. insertion reactions.


326

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1 -9.Metal-CatalyzedReactions of Carbon Monoxide

(6)

Group 11 metals8 show only one C-0 stretching vibration at 2105 (Cu), 2160 (Ag), and 21 10 (Au) cm- after the surfaces are exposed to CO. Again, when comparing molecular complexes of these metals, Cu gives carbonyl complexes in its + I oxidation state, (see 14.6.1.6), but only one example of a well-defined carbonyl complex of Ag is known16 and Au forms the carbonyl halo complexes Au(CO)C~'~, and Au(CO)Br'* with an infrared absorption around 2160 cm-', higher than in free CO.

'

(F. CALDERAZZO)

1. 2. 3. 4. 5. 6. 7.

8.

9. 10. 11. 12. 13. 14. 15. 16. 17.

18.

G. Ertl, Gazz. Chim. ltal., 109, 217 (1979), and references therein. R. Ugo, Catal. Rev.-Sci. Eng., 11, 225 (1975). J. A. Connor, Top. Curr. Chem., 71,71 (1977). R. R. Ford, A h . Catal., 21, 51 (1970).

E. L. Muetterties, T. N. Rhodin, E. Band, C. F. Brucker, W. R. Pretzer, Chem. Rev., 79, 91 (1979). Unpublished data by Y.Iwasawa, R. Mason, Textor M., and G . A. Somorjai, quoted in W. L. Jolly, Top. Curr. Chem., 71; 149 (1977). P. Chini, J . Organomet. Chem., 200, 37 (1980), and references therein. B. M. W. Trapnell, Proc. SOC.,Ser. A, 218, 566 (1953); S . Wagener, J . Phys. Chem., 60, 567 (1956). S . Wagener, J . Phys. Chem., 61, 267 (1957). K. M. Chi, S . R. Frerichs, S . B. Wilson, J. E. Ellis, Angew. Chem., Int. Ed. Engl., 26, 1190 (1987). S . B. Wilson, J. E. Ellis, J . Am. Chem. SOC.,110, 303 (1988). J. E. Ellis, K. M. Chi, J . Am. Chem. SOC.,112, 6022 (1990). D. J. Sikora, M. D. Rausch, R. D. Rogers, J. L. Atwood, J . Am. Chem. Soc., 103, 1265 (1981). G. Fachinetti, C. Floriani, H. Stoeckli-Evans, J . Chem. SOC., Dalton Trans., 2297 (1977). G. Fachinetti, C. Floriani, F. Marchetti, S. Merlino, J . Chem. Soc., Chem. Commun., 522 (1976). P. K. Hurlburt, 0. P. Anderson, S . H. S t r a w , J. Am. Chem. Soc., 113, 6277 (1991). D. Belli Dell'Amico, F. Calderazzo, F. Marchetti, J . Chem. Soc., Dalton Trans., 1829 (1976), and references therein. D. Belli Dell'Amico, F. Calderazzo, P. Robino, A. Segre, J . Chem. Soc., Dalton Trans., 3017 ( 1991).

14.6.1.9. Metal-Catalyzed Reactions of Carbon Monoxide

Carbon monoxide becomes a versatile reagent when it is activated by transition metals, giving a large number of both homogeneous and heterogeneous reactions. In the former, reactions occur at metal-coordinated CO; in the latter, chemisorbed CO is usually responsible for the observed catalytic reaction. - The known metal-catalyzed reactions of carbon monoxide can be classified as 1. reactions by which CO is reduced by dihydrogen, 2. additions of CO across double bonds, usually in combination with a hydrogencontaining substrate. 3. reactions in which CO acts as a reducing agent, and 4. insertion reactions.

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1.9. Metal-Catalyzed Reactions of Carbon Monoxide

327

Only the general types of reactions will be presented here, since these subjects will be treated in greater detail in other sections. Reductions of CO by dihydrogen. Among these reactions, the methanol ~ynthesisl-~:

CO

+ 2 H,

+CH30H

the Fischer-Tropsch synthesis, typified by4,?

CO

+ H2

CnH2n+2+ RCHZOH

+ RCHO +

(a)

* * *

(b)

the homologation r e a ~ t i o n ~ - ~ :

CH,OH

+ CO + 2 H2 *CH3CH2OH + H2O

(c)

and the ethylene glycol synthesis”:

2 CO

+ 3 H2 +CHzOH-CH20H

can be mentioned. Reactions (b)-(d) are new carbon-carbon bond-forming systems. Reactions (a) and (b) are industrially relevant and are usually carried out over heterogeneous catalysts. In contrast, reactions (c) and (d) are catalyzed by homogeneous systems based on cobalt and rhodium, respectively. Concerning the reactions of type 2, addition of CO across double bonds can be exemplified by the hydroformylation reaction’

’:

the preparation of carboxylic acids by carbonylation of olefinsI2:

or acetylenic derivativesi3:

+

+

2 HCeCH

+ 2CO

HC=CH 2CO 2Hz0 HOOCCH2CH,COOH the synthesis of hydr~quinone’~ in the presence of iron and ruthenium catalysts: +2H2

[Rul +

the synthesis of esters from CO and alcohols’2:

the synthesis of anilides from CO and aniline”:

p-HO-C,H,-OH

(g)

(h)

328

14.6. Carbon Monoxide Reactions 14.6.1. Introduction 14.6.1-9.Metal-CatalyzedReactions of Carbon Monoxide

In reactions of type 3, CO acts as a reducing agent. These reactions include the water gas shift (WGS) reaction by which dihydrogen is produced by reduction of water:

CO

+ H2O * CO, + H,

(k)

The reaction is thermodynamically favored (AGOz9, = -28.6 !d/mol), and can be carried out either heterogeneou~ly’~ or homogeneously’6. Another reaction of type 3 is the formation of sym-dialkylureas from primary amines and CO catalyzed by manganese carbonyl complexes”:

CO

+ 2 RNH,

-

CO(NHR),

+ H2

(1)

The reaction, which is the dehydrogenative carbonylation of amines, involves carbamoyl derivatives as reactive intermediates’8. Carbon monoxide can also be used to reduce nitrobenzene with a catalytic homogeneous system based on ruthenium”, in combination with molecular hydrogen:

C6H5N02

+ 2 CO + H2

2 CO,

+ CSHsNH,

(m)

The formation of aryl or alkyl isocyanates by reaction of nitro derivatives with CO catalyzed by transition elements” is another example of the reductive capability of CO:

RNO,

+ 3 CO

RNCO

+ 2 CO,

(n)

The reductions of metal oxides and metal chlorides to metal carbonyl derivatives [see equations (e)-(g) of 14.6.1.11 are further examples. Among metal-catalyzed reactions of type 4 is the acetic acid synthesis from methanol and CO in the presence of cobalt” or rhodium” catalysts:

CH,OH

+ CO

-

CH3COOH

(0)

(F. CALDERAZZO)

1. K. Weissennel and H. J. Ape, Industrial Organic Chemistry, Verlag Chemie, Weinheim, 1978, p. 27. 2. J. W. Rathke, H. M. Feder, J . Am. Chem. SOC., 100, 3623 (1978). 3. J. S . Bradley, J . Am. Chem. SOC., 101,7419 (1979). 4. C. Masters, Adv. Organomet. Chem., 17, 61 (1979). 5 . G. Henrici Olivt, S . Olivt, Angew. Chem., Int. Ed. Engl., 15, 136 (1976). 6. I. Wender, R. A. Friedel, M. Orchin, Science, 113, 206 (1951). 7. I. Wender, Catal. Rev., 14, 108 (1976). 8. G. Braca, G. Sbrana, G. Valentini, G. Andrich, G. Gregorio, J . Am. Chem. SOC., 100, 6238 (1978). 9. T. Mizoroki, T. Matsumoto, A. Ozaki, Bull. Chem. SOC. Jpn., 52, 479 (1979). 10. R. L. Pruett, Science, 211, 11 (1981), and references therein. 11. P. Pino, F. Piacenti, M. Bianchi, in Organic Syntheses via Metal Carbonyls, Vol. 2, I. Wender, P. Pino, eds., Wiley, New York, 1977, p. 43. 12. P. Pino, F. Piacenti, M. Bianchi, in Organic Reactions via Metal Carbonyls, Vol. 2, I. Wender, P. Pino, eds., Wiley, New York, 1977, p. 233. 13. P. Pino and G. Braca, in Organic Syntheses via Metal Carbonyls, Vol. 2, P. Pino, I. Wender, eds., Wiley, New York, 1977, p. 419. 14. P. Pino, C. Braca, G. Sbrana, A. Cuccuru, Chem. Ind. (London), 1732 (1968). 15. C. L. Thomas, Catalytic Processes and Proven Catalysis, Academic Press, New York, 1970. 16. P. C. Ford,Acc. Chem. Res., 14, 31 (1981). 17. F . Calderazzo, Inorg. Chem., 4,293 (1965). 18. B. D. Dombek, R. J. Angelici, J . Organomef. Chem., 134, 203 (1977). 19. F. L’Eplattenier, P. Matthys, F. Calderazzo, Inorg. Chem., 9, 342 (1970).

14.6. Carbon Monoxide Reactions 14.6.2. Metal Carbonyls Important in Catalysis

329

20. F. J. Weigert,J. Org. Chem., 38, 1316 (1973). 21. K. Weissemel, H. J. Arpe, Industrial Organic Chemistry, Verlag Chemie, Weinheim, 1978, p. 156. 22. D. Forster, J . Chem. Soc., Dalton Trans., 1639 (1979), and references therein.

14.6.2. Metal Carbonyls Important in Catalysis Stable binary metal carbonyls, negatively charged or uncharged, exist for all the elements of the 3d, 4d, and 5d transition series from Group 4 to Group 10, with the exception of palladium. As shown in Table 1, some of the elements (Ti, Zr, Hf, Nb, Ta, and Pt) have anionic carbonylmetalates only and no stable uncharged derivatives have so far been reported. Of the known metal carbonyls, only those of Group 6, Group 7 (with the exclusion of technetium), iron, ruthenium, cobalt and nickel have been used much in catalytic processes. Metal carbonyl preparations can be divided into three main categories: (1) dry methods, i.e., reactions occurring in the absence of any inert or reactive liquid, ( 2 ) wet methods, i.e., reactions occurring in the presence of an inert or reactive liquid, and (3) matrixisolation techniques (see 14.6.1.5), by which a vaporized metal reacts with CO usually at temperatures around 20K. The latter low T method has been widely used recently for unstable metal carbonyls, their presence being usually established by IR spectroscopy. Examples are Cu(CO),', Al(C02, Ti(CO);, and Ag(C0);. We deal here only with the conventional methods (1) and (2), which are used almost exclusively for the preparation of thermally stable metal carbonyls. Since metal carbonyls contain the metal in an unusually low oxidation state (zero or lower), their preparation invariably involves a reducing system since the starting materials are common salts of the metal in higher oxidation states. Only with Ni(CO), is the direct combination of CO with the metal a practical method or preparation. Reducing systems commonly used are as follows: 1. CO itself, especially toward convalent metal oxides, such as oxides of osmium (OsO,), technetium, and rhenium (M,O,). The corresponding oxidation product is CO,. 2. Molecular H, with CO, as with Co,(CO), prepared from Co(I1) salts. 3. Strongly electropositive metals, such as Na or Mg or the Na-K alloy in the presence of CO, usually in ether solvent. Sometimes, addition of an electron transfer agent has been emphasized, such as pyridine, naphthalene, or benzophenone, which also allows the reduction to be carried out homogeneously. 4. Alkylating agents, such as aluminum alkyls, as in the case of the preparation of Mn,(CO),, from Mn(I1) salts. Reduction of the metal cation occurs via homolytic cleavage of the metal-alkyl bond in the M-R intermediate formed by the exchange reaction (a): M-X

+ RM-R

-

+M-R

R'

+ X-

+ M'

(a) (b)

In all of the above systems (1-4) reduction can be regarded to occur stepwise. Carbon monoxide may become coordinated as reduction proceeds with partial loss of the original anionic ligands and the oxidation state of the cation decreases. The oxidation state at which the metal-CO bond acquires some stability is a function of the metal.



329

20. F. J. Weigert,J. Org. Chem., 38, 1316 (1973). 21. K. Weissemel, H. J. Arpe, Industrial Organic Chemistry, Verlag Chemie, Weinheim, 1978, p. 156. 22. D. Forster, J . Chem. Soc., Dalton Trans., 1639 (1979), and references therein.

14.6.2. Metal Carbonyls Important in Catalysis Stable binary metal carbonyls, negatively charged or uncharged, exist for all the elements of the 3d, 4d, and 5d transition series from Group 4 to Group 10, with the exception of palladium. As shown in Table 1, some of the elements (Ti, Zr, Hf, Nb, Ta, and Pt) have anionic carbonylmetalates only and no stable uncharged derivatives have so far been reported. Of the known metal carbonyls, only those of Group 6, Group 7 (with the exclusion of technetium), iron, ruthenium, cobalt and nickel have been used much in catalytic processes. Metal carbonyl preparations can be divided into three main categories: (1) dry methods, i.e., reactions occurring in the absence of any inert or reactive liquid, ( 2 ) wet methods, i.e., reactions occurring in the presence of an inert or reactive liquid, and (3) matrixisolation techniques (see 14.6.1.5), by which a vaporized metal reacts with CO usually at temperatures around 20K. The latter low T method has been widely used recently for unstable metal carbonyls, their presence being usually established by IR spectroscopy. Examples are Cu(CO),', Al(C02, Ti(CO);, and Ag(C0);. We deal here only with the conventional methods (1) and (2), which are used almost exclusively for the preparation of thermally stable metal carbonyls. Since metal carbonyls contain the metal in an unusually low oxidation state (zero or lower), their preparation invariably involves a reducing system since the starting materials are common salts of the metal in higher oxidation states. Only with Ni(CO), is the direct combination of CO with the metal a practical method or preparation. Reducing systems commonly used are as follows: 1. CO itself, especially toward convalent metal oxides, such as oxides of osmium (OsO,), technetium, and rhenium (M,O,). The corresponding oxidation product is CO,. 2. Molecular H, with CO, as with Co,(CO), prepared from Co(I1) salts. 3. Strongly electropositive metals, such as Na or Mg or the Na-K alloy in the presence of CO, usually in ether solvent. Sometimes, addition of an electron transfer agent has been emphasized, such as pyridine, naphthalene, or benzophenone, which also allows the reduction to be carried out homogeneously. 4. Alkylating agents, such as aluminum alkyls, as in the case of the preparation of Mn,(CO),, from Mn(I1) salts. Reduction of the metal cation occurs via homolytic cleavage of the metal-alkyl bond in the M-R intermediate formed by the exchange reaction (a): M-X

+ RM-R

-

+M-R

R'

+ X-

+ M'

(a) (b)

In all of the above systems (1-4) reduction can be regarded to occur stepwise. Carbon monoxide may become coordinated as reduction proceeds with partial loss of the original anionic ligands and the oxidation state of the cation decreases. The oxidation state at which the metal-CO bond acquires some stability is a function of the metal.

Hf(C0)z

Group 4

22

[Ta(CO),] -26,28

Group 5

w(c0)6M [W(C0),]2 - 32

Group 6

Re2(CO),038

Group 7

OS(CO),~~,~' O%(C0),48 0s3(c0)i?

Group 8

Group 9

Group 10

"Nonexistent compounds can be identified by considering that the substances in the table have been listed in the following order of priority: uncharged mononuclear, anionic mononuclear, uncharged plynuclear, anionic polynuclear.

5d

Valence shell

TABLE1. A SELECTION OF KNOWNUNCHARGED AND ANIONIC BINARY METALCARBONYLS STABLE AT ORDINARY TEMPERATURE AND PRESSURE^

0

w

0

330 14.6. Carbon Monoxide Reactions 14.6.2. Metal Carbonyls Important in Catalysis


331

Carbonyl complexes are known in oxidation states as high as IV, as with the [Pt(CO)Cl,]- anion5, and as low as -1V and -111 [M(C0),l4, M=Cr,Mo,W and ([V(C0),]3 -637. The intermediacy of metal carbonyl complexes in some catalytic processes is well established. Specific examples are presented in 14.6. Here, only some general aspects of the problem are discussed. Often the species having a key role in the catalytic cycle is not recognized as a chemically defined substance. Its structural characterization is based on indirect evidence only. Kinetic studies are of paramount importance; in addition, very sensitive experimental methods, such as ESR spectroscopy, can detect catalytically relevant species at low concentrations. However, since an efficient catalyst should be present in a reaction mixture at steady-state and small concentrations, its physical detection can be difficult. Several main prerequisites are needed by a catalyst precursor for it to produce an efficient catalyst.

1. It should bind the reactive substrate at one given stage in the reaction sequence. 2. The chemical bond to the substrate should not be strong enough to cause irreversibility. This would cause the reaction sequence to stop at one of the minima in the potential energy curve of the reaction. 3. If the reaction sequence at one stage involves the collision between two molecules, e.g., A with B, two possibilities generally exist for a homogeneous system. Only one of the two molecules will bind the metal center; attack by the other will be intermolecular, or both substances will become loosely bonded at the metal center. The latter possibility assumes that the coordination positions occupied by A and B are cis, if the intermediate complex has a square planar or a pseudooctahedral geometry. For carbonylation reactions involving metal carbonyls as catalyst precursors, the best fit to requirements 1 and 2 apparently is given by 4d transition elements. It is known that when activation of CO is required, carbonyl complexes of 4d metals are better catalyst precursors than their 3d and 5d congeners. Pertinent to this point are the following experimental observations: 1. The hydroformylation reaction (see 14.6.3) is about lo4 faster with rhodium than with

2. 3.

4.

5.

cobalt8. Thus, contamination of the high pressure reaction containers by rhodium is one of the main sources of erratic results in laboratory scale experiments. The reduction of nitrobenzene to aniline by CO and H, is more effectively carried out with ruthenium than with iron or osmiumg. The reduction of nitro aromatic compounds to isocyanates by CO is preferably carried out with Pd, in the presence of a Lewis acid as promoter.'0*''. Recent observation^'^^'^ that carbon monoxide bonded to Pd,(CO),Cl, has a very high Pc0 (slightly above 2160 cm-', with a small dependence on the solvent) indicate the degree of wback donation from the metal to the carbonyl ligand is small, if any. The activation energy for the M(12C0)6-14C0isotopic exchange process is lower for molybdenum than for chromium and tungsten14-". The kinetics of CO displacement from Cp,M(CO), [see equation (c)] depend on the metal in the following order C O < R ~ > I ~ ' ~ * ' ~ :

+

L _.) Cp,M(CO)L Cp,M(CO), M=Co,Rh,Ir; L = PPh,

+ CO

(c)

332


Even though the 4d elements appear to be the best candidates for catalytic carbonylation reactions, it should be considered that the electronic and steric properties of the metal in a metal-ligand system ML, can be vastly modified by appropriate changes in the ligand environment. This matter, of considerable current interest, requires a better knowledge of the metal-to-ligand bond parameters than are presently available. The reactivity of metal carbonyls can be classified into two main categories: (1) reactions in which the oxidation state of the metal changes, and (2) reactions in which the oxidation state of the metal is not modified. Within these two headings, there are the following subtypes.

la. Reductions, predominantly by alkali metals, e.g., reduction of Mn,(CO),, by Na in tetrahydrofuran: Mn,(CO),o

+ 2 Na thf_ 2 Na [Mn(CO),]

(4

Ib. Oxidations, especially by halogens, e.g., reaction of Mn,(CO)lo with X, to give the corresponding halo-carbonyl complex: Mn,(CO),,

+ X, -+

2 Mn(CO)5X X = C1,Br

-

(el

lc. Disproportionations, usually promoted by a Lewis base B, e.g.:

3 Co,(CO),

+ mB

2 [COB,] [Co(CO),],

-

+ 8 CO

(f)

2a. Substitutions, by which a carbonyl group is substituted by another neutral ligand that contributes two electrons to the EAN of the central metal cation, as Cr(CO),

+ C,H,N

Cr(CO),C,H,N

+ CO

(€9

A special case of substitution reaction is the isotopic exchange between coordinated lZCOand external 13C0 or 14C0.In the former case the exchange can be monitored by IR spectroscopy, whereas in the latter it is more appropriate to use radioactive counting methods. (F. CALDERAZZO)

H. Huber, E. P. Kundig, M. Moskovits, G. A. Ozin, J . Am. Chem. Soc., 97, 2007 (1975). A. J. Hinchcliffe, J. S. Ogden, D. D.Oswald, J . Chem. Soc., Chem. Comrnun., 338 (1972). R. Busby, W. Klotzbucher, G. A. Ozin, Inorg. Chem., 16, 822 (1977). D. McIntosh, G . A. Ozin, J . Am. Chem. Soc., 98, 3167 (1976). C. Crocker, P. L. Goggin, R. J. Goodfellow,J. Chem. SOC., Chem. Commun., 1056 (1978). J. E. Ellis, C. P. Pamell, G. P. Hagen, J . Am. Chem. Soc., 100, 3605 (1978). J. E. Ellis, K. L. Fjare, T. G . Hayes, J. Am. Chem. Soc., 103, 6100 (1981). B. Heil, L. Mark6, Chem. Ber., 101,2209 (1968). F. L'Eplattenier, P. Matthys, F. Calderazzo, Inorg. Chem., 9, 342 (1970). W. B. Hardy, R. P. Bennett, Tetrahedron Lert., 961 (1967). F. J. Weigert,J. Org. Chem., 38, 1316 (1973). F. Calderazzo, D. Belli Dell'Amico, Inorg. Chem., 20, 1310 (1981). F. Calderazzo, J. Organometul Chem., 400, 303 (1990). G. Pajaro, F. Calderazzo, R. Ercoli, Gazz. Chim. Ztal., 93, 1486 (1960). 15. G. Cetini, 0. Gambino, Arti Acc. Sci. Torino, 97, 757 (1963). 16. G . Cetini, 0. Gambino, Arti Acc. Sci. Torino, 97, 1189 (1963). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.


333

17. F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, Vol. 1, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1967, p. 175. 18. F. Basolo, R. Pearson, Mechanisms of Inorganic Reactions, 2nd ed., Wiley-Interscience, New York, 1967, p. 576. 19. J. W. Freeman, F. Basolo, Organometallics, 10, 256 (1991). 20. K. M. Chi, S. R. Frerichs, S. B. Wilson, J. E. Ellis, J . Am. Chem. SOC., 110, 303 (1988). 21. K. M. Chi, S. R. Frerichs, S. B. Wilson, J. E. Ellis, Angew. Chem., Int. Ed. Engl., 26, 1190 (1987). 22. J. E. Ellis, K. M. Chi, J. Am. Chem. Soc., 112, 6022 (1990). 23. G. Natta, R. Ercoli, F. Calderazzo, A. Alberola, P. Corradini, G. Allegra, Atfi Acc. Naz. Lincei, CI. Sci. Mat. Fis.Nut., [8], 27, 107 (1959). 24. R. Ercoli, F. Calderazzo, A. Alberola, J . Am. Chem. Soc., 82, 2965 (1960). 25. F. Calderazzo, G. Pampaloni, D. Vitali, Gazz. Chim. Ital., I I l , 455 (1981). 26. R. P. M. Werner, H. E. Podall, Chem. Ind. (London), 144 (1961). 27. R. P. M. Werner, A. H. Filbey, S . A. Manastyrskyi, Inorg. Chem., 2, 298 (1964). 28. J. E. Ellis, A. Davison, Inorg. Synth., 16, 68 (1976). 29. F. Calderazzo, G. Pampaloni, G. Pelizzi, J . Organometal. Chem. 233, C41 (1982). 30. G. Natta, R. Ercoli, F. Calderazzo, A. Rabizzoni, J . Am. Chem. Soc., 79, 3611 (1957), and references therein. For recent X-ray and neutron diffraction analyses, see B. Rees, A. Mitschler, J . Am. Chem. SOC.,98, 7918 (1976); A. Whitaker, J. W. Jeffery, Acta Crystallogr., 23, 977 (1967). 31. E. Lindner, H. Behrens, D. Uhlig, Z. Nuturforsch., 18b, 276 (1973). 32. J. E. Ellis, G. P. Hagen, J . Am. Chem. Soc., 96,7825 (1974). 33. L. Mond, H. Hirtz, M. D. Cowap, J. Chem. SOC.,798 (1910); Z. Anorg. Allg. Chem., 68, 207 (1910). 34. A. Job, J. Rouvillois, Compt. Rend., 187, 564 (1928). 35. E. 0. Brimm, M. A. Lynch, W. J. Sesny, J. Am. Chem. SOC., 76, 3831 (1954). For an X-ray diffraction study, see: L. F. Dahl, R. E. Rundle, Acta Crysfallogr.,16, 419 (1963). 36. W. Hieber, C. Herget, Angew. Chem., 73, 579 (1961). 37. J. C. Hileman, D. K. Huggins, H. D. Kaesz, J . Am. Chem. Soc., 83, 2953 (1961). For an Xray diffractometric study, see: D. Wallach, Acta Crysfallogr.,I S , 1058 (1962). 38. W. Hieber, H. Fuchs, Z . Anorg. Allg. Chem., 248 256 (1941). 39. M. Berthelot, Compt. Rend., 112, 1343 (1891). 40. L. Mond, F. Quincke, J . Chem. SOC.,604 (1891). For an X-ray diffractometric study, see J. Donohue, A. Caron, Acta Crystallogr., 17, 663 (1964). 41. J. Dewar, H. 0. Jones, Proc. R. SOC.(London), 76A, 558 (1905). For a recent X-ray diffractometric study, see F. A. Cotton, J. M. Troup, J . Chem. SOC.,Dalton Trans., 800 (1974). 42. J. Dewar, H. 0. Jones, Proc. R. SOC. (London), 79A, 66 (1907). For a recent reinvestigation of the crystal structure, see F. A. Cotton, J. M. Troup, J . Chem. Soc., 96, 4155 (1974), and references therein. 43. W. Hieber, G. Brendel, Z. Anorg. Allg. Chem., 289, 338 (1957). 44. W. F. Edgell, M. T. Yang, B. J. Bulkin, R. Bayer, N. Koizumi, J . Am. Chem. Soc., 87, 3080 (1965). 45. F. Y. K. Lo, G. Longcni, P. Chini, L. D. Lower, L. F. Dahl, J . Am. Chem. Soc., 102, 7691 (1980). 46. W. Manchot, W. J. Manchot, Z . Anorg. Allg. Chem., 226, 385 (1936). 47. F. Calderazzo, F. L’Eplattenier, Inorg. Chem., 6, 1220 (1967). 48. J. R. Moss, W. A. G. Graham, J . Chem. SOC.,Dalton Trans., 95 (1977). 49. M. I. Bruce, F. G. A. Stone, Angew. Chem., 80, 460 (1968) and references therein. 50. R. Mason, A. I. M. Rae, J . Chem. Soc., A, 778 (1968). 51. W. Hieber, H. Stallmann, Z. Elekrrochern., 49, 288 (1943). 52. C. W. Bradford, R. S . Nyholm, J . Chem. SOC., Chem. Commun., 384 (1967). For an X-ray diffractometric study, see E. R. Corey, L. F. Dahl, Inorg. Chem., 1, 521 (1962). 53. W. Hieber, U. Teller, Z. Anorg. Allg. Chem., 249, 43 (1942). 54. I. Wender, H. W. Stemberg, M. Orchin, J . Am. Chem. SOC., 74, 1216 (1952). 55. R. A. Friedel, I. Wender, S. L. Shufler, H. W. Stemberg, J . Am. Chem. SOC., 77, 3951 (1955). 56. W. F. Edgell, M. T. Yang, N. Koizumi, J . Am. Chem. Soc., 87,2563 (1965). 57. L. Mond, H. Hirtz, M. D. Cowap, J . Chem. SOC.,798 (1910).

334

14.6.2. Metal Carbon Is Important in Catalysis 14.6.2.1. Chromium, holybdenum, and Tungsten Carbonyls 14.6.2.1.1. Preparation of the Hexacarbonyls M(CO), (M=Cr,Mo,W).

58. G. G. Sumner, H. P. Klug, L. E. Alexander, Acta Cryst., 17, 732 (1964). 59. For X-ray diffractometric studies, see P. Corradini, J . Chem. Phys., 31, 1676 (1959); C. H. Wei, L. F. Dahl, J . Am. Chem. Soc., 88, 1821 (1966). 60. V. Albano, P. Chini, V. Scatturin, J . Organornetal. Chem., 15,423 (1968). 61. P. Chini, S. Martinengo, Inorg. Chim. Acta, 3, 21 (1969). 62. H. Lagally, Z. Anorg. Allg. Chem., 251,96 (1943). 63. W. Beck, K. Lottes, Chem. Ber., 94, 2578 (1961). For an X-ray diffractometric study, see C. H. Wei, G . R. Wilkes, L. F. Dahl, J . Am. Chem. Soc., 89,4792 (1967). 64. E. R. Corey, L. F. Dahl, W. Beck, J . Am. Chem. SOC., 85, 1202 (1963), and references therein. 65. A. Fumagalli, T. F. Koetzle, F. Takusagawa, P. Chini, S. Martinengo, B. T. Heaton, J . Am. Chem. SOC., 102, 1740 (1980). 66. W. Hieber, H. Lagally, Z. Anorg. Allg. Chem., 245, 321 (1940). For an X-ray diffractometric study, see M. R. Churchill, J. P. Hutchinson, Inorg. Chem., 17, 3528 (1978). 67. L. Mond, C. Langer, F. Quincke, J . Chem. Soc., 749 (1890). 68. J. C. Calabrese, L. F. Dahl, A. Cavalieri, P. Chini, G . Longoni, S. Martinengo, J. Am. Chem. Soc., 96,2616 (1974). 69. J. C. Calabrese, L. F. Dahl, P. Chini, G . Longoni, S. Martinengo,J . Am. Chem. Soc., 96,2614 (1974). 70. G. Longoni, P. Chini, J . Am. Chem. Soc., 98, 7225 (1976). 71. D. M. Washecheck, E. J. Wucherer, L. F. Dahl, A. Ceriotti, G. Longoni, M. Manassero, M. Sansoni, P. Chini, J . Am. Chem. Soc., 101, 61 10 (1979).

14.6.2.1. Chromium, Molybdenum, and Tungsten Carbonyls 14.6.2.1 .l.Preparation of the HexacarbonylsM(CO), (M=Cr,Mo,W).

Formation of the metal-CO bond for the three metals of Group 6 requires one of the following wet methods. Hexacarbonylchrorniurn(O),Cr(CO),,a colorless solid sublimable at 60-80°C under reduced pressure, forms from reaction of anhydrous CrCl, with PhMgBr and CO at atmospheric pressure in diethyl ether as solvent'. Improvements to the original method have been 67% yields were reported when the PhMgBr/CrCl, molar ratio was 7.5, the CO pressure was above atmospheric, and the T was maintained between - 4 and 10"C3. Although the reaction stoichiometry can be represented by equation (a), CrX,

+ 3 PhMgBr + 6 Co

-

Cr(C0,

+ 3 MgXBr + 3 Ph'

(a)

Cr(CO), is not present in the final reaction mixture. Hydrolysis yields the Cr(CO),, which is recovered by steam distillation and further purified by sublimation. Cr(CO),Ph, as an intermediate has been proposed5, which upon Cr-Ph bond cleavage would yield the Cr(CO), precursor prior to hydrolysis. Several possibilities about the Cr(CO), precursor exist, e.g., that the final carbonylation product could be the anionic species arising from either the insertion reaction (b) or the nucleophilic attack by the Grignard reagent on a carbonyl group of Cr(CO), according to equation (c)? [Cr(CO),Ph]Cr(CO),

+ CO

+ Ph-

--

[Cr(CO),COPh]-

(b)

[Cr(CO),COPh]-

(c)

The fact that Cr(CO), reportedly7ss yields [Cr(CO),COR] - by reaction with lithium alkyls, support this hypothesis. By treatment with protons7, aldehydes and Cr(CO), were detected.


334

14.6.2. Metal Carbon Is Important in Catalysis 14.6.2.1. Chromium, holybdenum, and Tungsten Carbonyls 14.6.2.1.1. Preparation of the Hexacarbonyls M(CO), (M=Cr,Mo,W).

58. G. G. Sumner, H. P. Klug, L. E. Alexander, Acta Cryst., 17, 732 (1964). 59. For X-ray diffractometric studies, see P. Corradini, J . Chem. Phys., 31, 1676 (1959); C. H. Wei, L. F. Dahl, J . Am. Chem. Soc., 88, 1821 (1966). 60. V. Albano, P. Chini, V. Scatturin, J . Organornetal. Chem., 15,423 (1968). 61. P. Chini, S. Martinengo, Inorg. Chim. Acta, 3, 21 (1969). 62. H. Lagally, Z. Anorg. Allg. Chem., 251,96 (1943). 63. W. Beck, K. Lottes, Chem. Ber., 94, 2578 (1961). For an X-ray diffractometric study, see C. H. Wei, G . R. Wilkes, L. F. Dahl, J . Am. Chem. Soc., 89,4792 (1967). 64. E. R. Corey, L. F. Dahl, W. Beck, J . Am. Chem. SOC., 85, 1202 (1963), and references therein. 65. A. Fumagalli, T. F. Koetzle, F. Takusagawa, P. Chini, S. Martinengo, B. T. Heaton, J . Am. Chem. SOC., 102, 1740 (1980). 66. W. Hieber, H. Lagally, Z. Anorg. Allg. Chem., 245, 321 (1940). For an X-ray diffractometric study, see M. R. Churchill, J. P. Hutchinson, Inorg. Chem., 17, 3528 (1978). 67. L. Mond, C. Langer, F. Quincke, J . Chem. Soc., 749 (1890). 68. J. C. Calabrese, L. F. Dahl, A. Cavalieri, P. Chini, G . Longoni, S. Martinengo, J. Am. Chem. Soc., 96,2616 (1974). 69. J. C. Calabrese, L. F. Dahl, P. Chini, G . Longoni, S. Martinengo,J . Am. Chem. Soc., 96,2614 (1974). 70. G. Longoni, P. Chini, J . Am. Chem. Soc., 98, 7225 (1976). 71. D. M. Washecheck, E. J. Wucherer, L. F. Dahl, A. Ceriotti, G. Longoni, M. Manassero, M. Sansoni, P. Chini, J . Am. Chem. Soc., 101, 61 10 (1979).

14.6.2.1. Chromium, Molybdenum, and Tungsten Carbonyls 14.6.2.1 .l.Preparation of the HexacarbonylsM(CO), (M=Cr,Mo,W).

Formation of the metal-CO bond for the three metals of Group 6 requires one of the following wet methods. Hexacarbonylchrorniurn(O),Cr(CO),,a colorless solid sublimable at 60-80°C under reduced pressure, forms from reaction of anhydrous CrCl, with PhMgBr and CO at atmospheric pressure in diethyl ether as solvent'. Improvements to the original method have been 67% yields were reported when the PhMgBr/CrCl, molar ratio was 7.5, the CO pressure was above atmospheric, and the T was maintained between - 4 and 10"C3. Although the reaction stoichiometry can be represented by equation (a), CrX,

+ 3 PhMgBr + 6 Co

-

Cr(C0,

+ 3 MgXBr + 3 Ph'

(a)

Cr(CO), is not present in the final reaction mixture. Hydrolysis yields the Cr(CO),, which is recovered by steam distillation and further purified by sublimation. Cr(CO),Ph, as an intermediate has been proposed5, which upon Cr-Ph bond cleavage would yield the Cr(CO), precursor prior to hydrolysis. Several possibilities about the Cr(CO), precursor exist, e.g., that the final carbonylation product could be the anionic species arising from either the insertion reaction (b) or the nucleophilic attack by the Grignard reagent on a carbonyl group of Cr(CO), according to equation (c)? [Cr(CO),Ph]Cr(CO),

+ CO

+ Ph-

--

[Cr(CO),COPh]-

(b)

[Cr(CO),COPh]-

(c)

The fact that Cr(CO), reportedly7ss yields [Cr(CO),COR] - by reaction with lithium alkyls, support this hypothesis. By treatment with protons7, aldehydes and Cr(CO), were detected.

14.6.2. Metal Carbon Is Important in Catalysis 14.6.2.1. Chromium, holybdenum, and Tungsten Carbonyls 14.6.2.1-1. Preparation of the Hexacarbonyls M(C0)8 (M=Cr,Mo,W).

335

Another synthesis of Cr(CO), uses the Mg-Zn-pyridine reducing system in the presence of CO under pressureg, starting from easily available Cr(II1) salts: 2 CrX,

+

12 CO

+ 3 Mg pyridine

3 MgXz + 2 Cr(CO),

(d)

It appears that pyridine plays an important r6le in the reduction; an unstable pyridine radical anion may form in steady-state concentrations during the reaction, which is then responsible for the electron transfer process in the homogeneous phase. Sodium and pyridine react"-13 to form solutions whose dark blue color was attrib~ted'~ to the 4,4'dipyridyl radical anion. Magnesium reacts with pyridine to form a blue solid, presumably the radical anion of pyridine or of its dehydrogenated coupling productg. Similarly reduction of CrX, with sodium metal in tetrahydr~furan'~ in the presence of benzophenone yields Cr(CO), (59%). LiAlH, in diethyl ether also reduces CrCl, to Cr(CO), (65% yield)',. Electrochemical reduction of Cr(II1) salts has been carried out in a thick-walled electrolytic cell under CO pre~sure'~ in pyridine solvent, in the cell system: Anode(A1) 11 CrCl, (pyridine), Bu4NBr, CO, pyridine 11 Cathode Best yields (83%) were obtained using preformed CrCl, (pyridine),. Hexacarbonyls of chromium, molybdenum and tungsten form (yields of 92,76, and 92%, respectively) by treatment of the corresponding anhydrous metal halide with AlEt, and CO in diethyl ether". Details of this method, as applied to Cr(CO),, are a~ailable'~. Hexacarbonyls of molybdenum(0) and tungsten(O), colorless solids with properties similar to those of Cr(CO),, are more easily prepared than Cr(CO),. Zinc or Fe powders reduce molybdenum and tungsten halides in the presence of C020. Mo(CO), and W(CO), were obtained in 46 and 70% yields, respectively, using superatmospheric pressures of

co.

Hexacarbonylmolybdenum (80% yield) is obtained by reducing MoC1, with Cu-A1 (Devarda) alloy at a CO pressure of 150 atm at 100°C21.Excellent yields of W(CO), (86% of sublimed product) are obtained by operating under the more drastic conditions of 300 atm at 14O0CZ1. (F. CALDERAZZO)

1. A. Job, A. Cassal, Compt. Rend., 183, 392 (1926). 2. W. Hieber, E. Romberg, Z. Anorg. Allg. Chem.,221, 321 (1935). 3. B. B. Owen, J. English, H. G. Cassidy, C. V. Dundon, J . Am. Chem. Soc., 69, 1723 (1947). 4. B. B. Owen, J. English, H. G. Cassidy, C. V. Dundon, Inorg. Syn., 3, 156 (1950). 5. H. Zeiss, in Organometallic Chemistry, H. Zeiss, ed., Reinhold, New York, 1960, p. 380. 6. F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1967, p. 30. 7. E. 0. Fischer, A. Massbol, Chem. Ber., 100,2445 (1967). 8 . M. Y. Darensbourg, D. J. Darensbourg, Inorg. Chem., 9, 32 (1970). 9. R. Ercoli, F. Calderazzo, G. Bemardi, Gazz. Chim. Ital., 89, 809 (1959). 10. B. Emmert, Chem. Ber., 47,2598 (1914). 11. B. Emmert, Chem. Ber., 49, 1060 (1916). 12. B. Emmert, P. Buchert, Chem. Ber., 54,204 (1921). 13. R. L. Ward, J . Am. Chem. SOC., 83, 3623 (1961). 14. C. D. Schmulbach, C. C. Hinckley, D. Wasmund, J . Am. Chem. SOC., 90,6600 (1968). 15. R. D. Closson, L. R. Buzbee, G. G. Ecke,J. Am. Chem. SOC., 80, 6167 (1958). 16. A. N. Nesmeyanov, K. N. Anisimov, V. L. Volkov, A. E. Fridenberg, E. P. Mikheev, A. V. Medvedeva, Zh. Neorg. Khim., 4 , 1827 (1959).

336

14.6.2. Metal Carbon Is Important in Catalysis 14.6.2.1. Chromium, hol bdenum, and Tungsten Carbon Is 14.6.2.1.2. Reactions of &exacarbonyls of Cr, Mo, and

d

17. M. Guainazzi, G. Silvestri, S. Gambino, G . Filardo, J. Chem. SOC., Dalton Trans.,927 (1972). 18. H. E. Podall, J. H. Dunn, H. Shapiro, J . Am. Chem. Soc., 82, 1325 (1960). 19. G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, Verlag, Stuttgart, 1981, p. 1816. 20. K. N. Anisimov, A. N. Nesmeyanov, Dokl. Akad. Nauk SSSR,26,57 (1940). 21. G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, Verlag, Stuttgart, 1981, p. 1821. 14.6.2.1.2. Reactions of Hexacarbonyisof Cr, Mo, and W.

Reduction of the Group 6 metal hexacarbonyls produces the corresponding carbonylmetalate anions 1-41*2.Reduction of the hexacarbonyls with alkali metals in liquid NH, yields the metal carbonyl anions (1): [M(co),12-

[M(CO)4142

1

M(CO),

[M2(C0),,l2 3

[M3(cO)i4l24

+ 2 Na -+ NaJM(CO),] + CO

-

(a)

Reaction of the pentacarbonylchromate( - 11) anion with water yields dihydrogen3:

2 [Cr(CO),]2-

+ 3 H,O

[Cr,H(CO),,]-

+ H, + 3 OH-

(b)

Since two CO groups of the hexacarbonyl are replaced to form the [M(C0),l4- anions, these highly reduced anions are prepared4 by reducing diamino-substituted tetracarbonyl derivatives with sodium in liquid NH,: M(CO),(TMDA)

+ 4 Na

-

Na4[M(CO),]

+ TMDA

TMDA = N,N,N',N'-tetramethylethylenediamine

(c)

The tetracarbonyl anions (2) are characterized by one main carbonyl stretching vibration around 1500 cm-'. The dinuclear dianions (3) are obtained by reduction of the hexacarbonyls with NaBH, in liquid NH,. For Cr the reaction can be represented 2 M(CO),

-

+ 2 NaBH, + 6 NH,

Na,[M,(CO),,]

+ 2 B(NH,), + 2 CO + 7 H,

(d)

Trinuclear anions (4) of Cr and Mo are obtained by treating the hexacarbonyls with NaBH, at the tetrahydrofuran reflux':

3 M(CO),

+ 2 NaBH,

Na,[M,(CO),,]

+ 4 CO + H, + B2H6

(e)

The oxidation state of the metal can be increased by several methods, the simplest being treatment with halogens. Cr(CO), is unaffected by bromine and iodine8, but is attacked by C1, to give CrCl,, carbon monoxide, and some phosgene. Mo(CO), and W(CO), yield the corresponding triiodides (MI,) by reaction with I, at 105 and 120°C, respectivelyg. However, MO(CO), and W(CO), give M(CO),Br2 with Br, at about - 78°C in dichloromethanelO.ll: M(C0)6

+ Br,

-

2 CO (M-Mo,W)

+ M(C0)4Br,

(f)

Under UV irradiation, Mo(CO), and W(CO), with I, in n-hexane yield the corresponding diiodotetracarbony 1s"


336

14.6.2. Metal Carbon Is Important in Catalysis 14.6.2.1. Chromium, hol bdenum, and Tungsten Carbon Is 14.6.2.1.2. Reactions of &exacarbonyls of Cr, Mo, and

d

17. M. Guainazzi, G. Silvestri, S. Gambino, G . Filardo, J. Chem. SOC., Dalton Trans.,927 (1972). 18. H. E. Podall, J. H. Dunn, H. Shapiro, J . Am. Chem. Soc., 82, 1325 (1960). 19. G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, Verlag, Stuttgart, 1981, p. 1816. 20. K. N. Anisimov, A. N. Nesmeyanov, Dokl. Akad. Nauk SSSR,26,57 (1940). 21. G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, Verlag, Stuttgart, 1981, p. 1821. 14.6.2.1.2. Reactions of Hexacarbonyisof Cr, Mo, and W.

Reduction of the Group 6 metal hexacarbonyls produces the corresponding carbonylmetalate anions 1-41*2.Reduction of the hexacarbonyls with alkali metals in liquid NH, yields the metal carbonyl anions (1): [M(co),12-

[M(CO)4142

1

M(CO),

[M2(C0),,l2 3

[M3(cO)i4l24

+ 2 Na -+ NaJM(CO),] + CO

-

(a)

Reaction of the pentacarbonylchromate( - 11) anion with water yields dihydrogen3:

2 [Cr(CO),]2-

+ 3 H,O

[Cr,H(CO),,]-

+ H, + 3 OH-

(b)

Since two CO groups of the hexacarbonyl are replaced to form the [M(C0),l4- anions, these highly reduced anions are prepared4 by reducing diamino-substituted tetracarbonyl derivatives with sodium in liquid NH,: M(CO),(TMDA)

+ 4 Na

-

Na4[M(CO),]

+ TMDA

TMDA = N,N,N',N'-tetramethylethylenediamine

(c)

The tetracarbonyl anions (2) are characterized by one main carbonyl stretching vibration around 1500 cm-'. The dinuclear dianions (3) are obtained by reduction of the hexacarbonyls with NaBH, in liquid NH,. For Cr the reaction can be represented 2 M(CO),

-

+ 2 NaBH, + 6 NH,

Na,[M,(CO),,]

+ 2 B(NH,), + 2 CO + 7 H,

(d)

Trinuclear anions (4) of Cr and Mo are obtained by treating the hexacarbonyls with NaBH, at the tetrahydrofuran reflux':

3 M(CO),

+ 2 NaBH,

Na,[M,(CO),,]

+ 4 CO + H, + B2H6

(e)

The oxidation state of the metal can be increased by several methods, the simplest being treatment with halogens. Cr(CO), is unaffected by bromine and iodine8, but is attacked by C1, to give CrCl,, carbon monoxide, and some phosgene. Mo(CO), and W(CO), yield the corresponding triiodides (MI,) by reaction with I, at 105 and 120°C, respectivelyg. However, MO(CO), and W(CO), give M(CO),Br2 with Br, at about - 78°C in dichloromethanelO.ll: M(C0)6

+ Br,

-

2 CO (M-Mo,W)

+ M(C0)4Br,

(f)

Under UV irradiation, Mo(CO), and W(CO), with I, in n-hexane yield the corresponding diiodotetracarbony 1s"

t

14.6.2. Metal Carbon Is Important in Catalysis 14.6.2.1. Chromium, 01 bdenum, and Tungsten Carbon Is 14.6.2.1.2. Reactions of hexacarbonyls of Cr, Mo, and

M(CO),

+ I,

-

2 CO

---

+ M(CO),I,

d

337

(g)

Oxidation to the +0.5 state occurs on reaction of M(CO), (M = Mo, W) with SiI, or HgI,, re~pectively’~: MO(CO), w(c0)6

SiI,

HgI,

WO(CO),II,

(h)

[w(co),II~

6)

InterThe deep blue Cr(CO),I forms by oxidizing the [Cr,(C0),,I2- anion with mediate formation of the red Cr,(CO),,I is likely, since the latter compound can be isolated using the reaction stoichiometry of equation (j): [Cr,(CO)l,]2[Cr,(C0)l,]2-

+ 2 I, + 1.5 I,

2 Cr(CO),I Cr,(CO),,I

+ 2 I+ 2 I-

(j)

(k)

The Group 6 metal hexacarbonyls are also oxidized by several organic substances that contain active protons, presumably by evolution of H, or by hydrogen transfer to the excess of organic substrate, e.g., in the formation of P-diketonato complexes of Cr(III)’, and Mo(III)16: M(CO),

+ 3(RCO),CH,

-

% H,

+ M [(RCO),CH], + 6 CO

(1)

Substitution reactions are well documented for the three Group 6 metal hexacarbonyls. Since they are low-spin d6 systems, reactivity is normally lower than that of other metal carbonyls. To promote substitution reactions, it is necessary to use elevated temperatures or uv irradiation. The hexacarbonyls usually undergo substitution via a dissociative mechanism, unless direct attack of a nucleophile at coordinated carbonyl groups occurs”. Usually Mo(CO), is more reactive than its two congeners. This is shown, inter alia, by kinetic studies on the isotopic exchange reaction between M( “CO), and 14CO’8,19. The higher kinetic lability of Mo(CO), may effect some of the reductions and oxidations, when the electron transfer process is accompanied by loss of carbonyl groups, e.g., equations (a), (e)-(g), and (1). Several examples of substitution processes can be described. Amine-substituted carbony1 complexes have been obtained. By thermal treatment, mixtures of mono-, bi-, and trisubstituted products are usually formed, whereas UV irradiation better yields the monosubstituted compound”. The reaction mixture composition is a function of both temperature and CO partial pressure: M(CO),

+ nAmine

-I heat

M(CO),

+ Amine

hv

M(CO),(Amine) M(CO),(Amine), M(CO),(Amine),

+ CO + 2 CO + 3 CO

(m)

M(CO),(Amine)

+ CO

(n)

Substitution of carbonyl groups in Cr(CO),, Mo(CO),, and W(CO), is not limited to nitrogen donors; there are many examples of carbonyl substitutions by molecules containing phosphorus, arsenic, antimony, and bismuth and Group 16 (oxygen, chalcogens) donor atomsz1. The labile [Cr(CO),H] - ”and [Cr(CO),X] - 2 3 (X=Cl,Br,I) ions are substitution products of hydrido and halo ligand reactions:

338

14.6.2. Metal Carbonyls Im ortant in Catalysis 14.6.2.2. Manganese and &hsnium Carbonyls 14.6.2.2.1. Preparation of the Metal Carbonyls.

M(CO),

+ X- +[M(CO),X]- + CO

(0)

Substitution of CO by ligands containing unsaturated carbon-carbon bonds is shown by the substitution of CO by the cyclopentadienyl a n i ~ n ' ~ * 'and ~ by aromatic hydrocarbons26-28:

(F.CALDERAZZO)

E. Lindner, H. Behrens, D. Uhlig, Z. Naturforsch., lab, 276 (1973). H. Behrens, R. Weber, Z. Anorg. Allg. Chem., 291, 122 (1957). H. Behrens, W. Klek, Z. Anorg. Allg. Chem., 292, 151 (1957). J. E. Ellis, C. P. Parnell, G. P. Hagen, J . Am. Chem. Soc., 100, 3605 (1978). H. Behrens, W. Haag, Z. Naturforsch., 14b, 600 (1959). H. Behrens, W. Haag, Chem. Ber., 94,312 (1961). H. Behrens, W. Haag, Chem. Ber., 94, 320 (1961). W. Hieber, E. Romberg, Z. Anorg. Allg. Chem., 221,321 (1935). C. Djordjevit, R. S . Nyholm, C. S . Pande, M. H. B. Stiddard, J. Chem. Soc., A, 16 (1966). R. Colton, I. B. Tomkins, Aust. J . Chem., 19, 1519 (1966). J. A. Bowden, R. Colton, Aust. J . Chem., 21,2657 (1968). R. Colton, C. J. Rix, Aust. J . Chem., 22, 305 (1969). G. Schmid, R. Boese, E. Welz, Chem. Ber., 108,260 (1975). H. Behrens, R. Schwab, Z. Naturjorsch., 19b, 768 (1964); see also H. Behrens, D. Herrmann, Z. Anorg. Allg. Chem., 351,225 (1967). 15. T. G. Dunne, F. A. Cotton, Inorg. Chem., 2, 263 (1963). 16. M. L. Larson, F. W. Moore, Inorg. Chem., I, 856 (1962). 17. F. Basolo, R. Pearson, Mechanisms of Inorganic Reactions, 2nd ed., Wiley-Interscience, New York, 1967, p. 526. 18. G. Pajaro, F. Calderazzo, R. Ercoli, Gazz. Chim. Ital., 93, 1486 (1960). 19. F. Calderazzo, R. Ercoli, G. Natta, in Organic Synthesis via Metal Carbonyls, Vol. 1, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1967, p. 175. 20. W. Hieber, W. Abeck, H. K. Platzer, Z. Anorg. Allg. Chem., 280, 252 (1955). 21. F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, Vol. 1, I. Wender, P. Pino, eds. Wiley-Interscience, New York, 1967, p. 166. 22. W. Hieber, W. Abeck, H. K. Platzer, Z. Anorg. Allg. Chem., 280, 241 (1955). 23. E. W. Abel, I. S . Butler, I. G. Reid, J . Chem. SOC., 2068 (1963). 24. T. S. Piper, G. Wilkinson, Naturwissenschaften, 42, 625 (1955). 25. E. 0. Fischer, W. Hafner, H. 0. Stahl, Z. Anorg. Allg. Chem., 282,47 (1955). 26. G. Natta, R. Ercoli, F. Calderazzo, Chim. Ind. (Milano),40, 287 (1958). 27. E. 0. Fischer, K. Oefele, H. Essler, W. FrGhlich, J. P. Mortensen, W. Semmninger, Z. Naturforsch.. 13b. 458 (1958). 28. B.Nicholls,'M. C: Whiting, J . Chem. SOC.,551 (1959). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

14.6.2.2. Manganese and Rhenium Carbonyls 14.6.2.2.1. Preparation of the Metal Carbonyls.

Decacarbonyldimanganese(O), Mn,(CO),,, a golden-yellow solid, (mp 154OC)' was first obtained by reaction of MnI, with Mg, Cu, and CuI, at room temperature in diethyl ether under CO at 200 atm P. Yields are low (-1%). The most frequently used reducing agents for carbonylation of Mn(I1) salts are (1) Na metal or Na in the presence of benzophenone or (2) aluminum alkyls. The sodium-benzophenone' reduction occurs at superatmospheric CO pressures (200-700 atm)


338

14.6.2. Metal Carbonyls Im ortant in Catalysis 14.6.2.2. Manganese and &hsnium Carbonyls 14.6.2.2.1. Preparation of the Metal Carbonyls.

M(CO),

+ X- +[M(CO),X]- + CO

(0)

Substitution of CO by ligands containing unsaturated carbon-carbon bonds is shown by the substitution of CO by the cyclopentadienyl a n i ~ n ' ~ * 'and ~ by aromatic hydrocarbons26-28:

(F.CALDERAZZO)

E. Lindner, H. Behrens, D. Uhlig, Z. Naturforsch., lab, 276 (1973). H. Behrens, R. Weber, Z. Anorg. Allg. Chem., 291, 122 (1957). H. Behrens, W. Klek, Z. Anorg. Allg. Chem., 292, 151 (1957). J. E. Ellis, C. P. Parnell, G. P. Hagen, J . Am. Chem. Soc., 100, 3605 (1978). H. Behrens, W. Haag, Z. Naturforsch., 14b, 600 (1959). H. Behrens, W. Haag, Chem. Ber., 94,312 (1961). H. Behrens, W. Haag, Chem. Ber., 94, 320 (1961). W. Hieber, E. Romberg, Z. Anorg. Allg. Chem., 221,321 (1935). C. Djordjevit, R. S . Nyholm, C. S . Pande, M. H. B. Stiddard, J. Chem. Soc., A, 16 (1966). R. Colton, I. B. Tomkins, Aust. J . Chem., 19, 1519 (1966). J. A. Bowden, R. Colton, Aust. J . Chem., 21,2657 (1968). R. Colton, C. J. Rix, Aust. J . Chem., 22, 305 (1969). G. Schmid, R. Boese, E. Welz, Chem. Ber., 108,260 (1975). H. Behrens, R. Schwab, Z. Naturjorsch., 19b, 768 (1964); see also H. Behrens, D. Herrmann, Z. Anorg. Allg. Chem., 351,225 (1967). 15. T. G. Dunne, F. A. Cotton, Inorg. Chem., 2, 263 (1963). 16. M. L. Larson, F. W. Moore, Inorg. Chem., I, 856 (1962). 17. F. Basolo, R. Pearson, Mechanisms of Inorganic Reactions, 2nd ed., Wiley-Interscience, New York, 1967, p. 526. 18. G. Pajaro, F. Calderazzo, R. Ercoli, Gazz. Chim. Ital., 93, 1486 (1960). 19. F. Calderazzo, R. Ercoli, G. Natta, in Organic Synthesis via Metal Carbonyls, Vol. 1, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1967, p. 175. 20. W. Hieber, W. Abeck, H. K. Platzer, Z. Anorg. Allg. Chem., 280, 252 (1955). 21. F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, Vol. 1, I. Wender, P. Pino, eds. Wiley-Interscience, New York, 1967, p. 166. 22. W. Hieber, W. Abeck, H. K. Platzer, Z. Anorg. Allg. Chem., 280, 241 (1955). 23. E. W. Abel, I. S . Butler, I. G. Reid, J . Chem. SOC., 2068 (1963). 24. T. S. Piper, G. Wilkinson, Naturwissenschaften, 42, 625 (1955). 25. E. 0. Fischer, W. Hafner, H. 0. Stahl, Z. Anorg. Allg. Chem., 282,47 (1955). 26. G. Natta, R. Ercoli, F. Calderazzo, Chim. Ind. (Milano),40, 287 (1958). 27. E. 0. Fischer, K. Oefele, H. Essler, W. FrGhlich, J. P. Mortensen, W. Semmninger, Z. Naturforsch.. 13b. 458 (1958). 28. B.Nicholls,'M. C: Whiting, J . Chem. SOC.,551 (1959). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

14.6.2.2. Manganese and Rhenium Carbonyls 14.6.2.2.1. Preparation of the Metal Carbonyls.

Decacarbonyldimanganese(O), Mn,(CO),,, a golden-yellow solid, (mp 154OC)' was first obtained by reaction of MnI, with Mg, Cu, and CuI, at room temperature in diethyl ether under CO at 200 atm P. Yields are low (-1%). The most frequently used reducing agents for carbonylation of Mn(I1) salts are (1) Na metal or Na in the presence of benzophenone or (2) aluminum alkyls. The sodium-benzophenone' reduction occurs at superatmospheric CO pressures (200-700 atm)

339

14.6.2. Metal Carbonyls Im ortant in Catalysis 14.6.2.2. Manganese and F! henium Carbonyls 14.6.2.2.1. Preparation of the Metal Carbonyls.

2 MnX,

+ 6 Na +

10 CO

-

2 NaMn(CO),

+ 4 NaX

(a)

at 65-200°C in tetrahydrofuran solvent. The [Mn(CO),]- anion is the final product, which then requires hydrolysis for its conversion to Mn,(CO),,. The latter is recovered [Mn(CO),]-

+ H,O

@ MnH(CO),

+ OH-

(b)

from the resulting crude mixture by steam distillation and purified by sublimation under reduced pressure. If the acid dissociation constant of MnH(CO), - in aqueous solution3 is 0.8 X the constant for the hydrolysis reaction (b) is -1.2 X The formation of Mn,(CO),o is probably due to adventitious oxygen or to thermal decomposition of the pentacarbonyl hydride during the steam distillation procedure. Sodium metal reduction of (~5-methylcyclopentadienyl)tricarbonylmanganese(I) has been reported4s5. Again, an acidification in aqueous solution is r e q ~ i r e dto ~ ?obtain ~ MnH(CO), and Mn,(CO),,. This method is described in detail elsewhere5. Aluminum alkyls are also frequently used for the preparation of Mn,(CO),,*. Using an AIR,/Mn mixture (molar ratio 4: 1) and under selected conditions, yields up to 60% 6 MnX,

+ 4 AIR, + 30 CO

-

3 Mn,(CO),,

+ 4 AlX, +

12 R'

(c)

were obtained. The yield of Mn,(CO),, increases with the covalent character of the Mn(I1) salt employed (e.g., halides, with their high lattice energies, are worse than the isopropoxide). Ether solvents usually give best results and the AIR,/Mn molar ratio should be larger than expected by the stoichiometry of equation (c); best results are obtained with 7.2-9.0: 1 ratios. Although hydrolysis was employed in the original work', this was subsequently showng to be unnecessary. A simplified procedure has been reportedg; a solution of the less dangerous iso-Bu,Al in diisopropyl ether is sucked into an autoclave already containing anhydrous manganous acetate, followed by pressurization with CO (125 atm) and heating at 60°C (3 h) and 140°C (20 h). After cooling and venting residual gases, the reaction mixture is collected under nitrogen and cooled to - 50°C. Crystalline Mn,(CO),, is recovered by filtration and purified by sublimation (50"C/O. 1 torr; yield 50%). Decacarbonyldirhenium(O), Re,(CO),,, a colorless solid (mp 177"C), was first obtained" nearly quantitatively by treatment of Re,07 with CO at 350 atm of pressure at 250°C for 16 h without solvent (dry method): Re,07

+

17 CO

-

7 CO,

+ Re,(CO),,

(4

The Re,(CO),, has also been obtained from reduction of rhenium halides with Na in tetrahydrofuran at 130°C under 250-280 atm CO pressure". However, with alkali metal used for the reduction, the [Re(CO),] - anion or more complicated cluster anions form and hydrolysis is required to sort the carbonyl out of the reaction mixture (70% yield). Although poor yields of Re,(CO)lo from KRe0, were originally reported'', KReO, is carbonylated to Re,(CO),, in good yields in the presence of Cu powder as the reducing agentI2. Moreover, NH,ReO, was carbonylated to Re,(CO),, in up to 69% yield without any additional reducing agent (at 290°C and 180 atm CO pre~sure)'~. 2 Re

+ 7 H,O, + 2 KOH 2 KReO, + 17 CO

--

+ 8 H,O Re,(CO),o + 7 CO, + K,O 2 KReO,

(el (f)

(F. CALDERAZZO)

340

14.6.2. Metal Carbonyls Im ortant in Catalysis 14.6.2.2. Manganese and I$henium Carbonyls 14.6.2.2.2. Reactions of the Carbonyls of Mn and Re.

E. 0. Brimm, M. A. Lynch, W. J. Sesny,J. Am. Chem. SOC., 76, 3831 (1954). R. D. Closson, L. R. Buzbee, G. G. Ecke, J . Am. Chem. SOC., 80,6167 (1958). W. Hieber, G. Wagner, Z. Naturforsch., 12b, 478 (1957). R. B. King, J. C. Stokes, T. F. Korenowski, J . Organometal. Chem., 1 1 , 641 (1968). 5 . G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stutt-

1. 2. 3. 4. 6. 7. 8. 9.

10.

11. 12. 13.

gart, 1981, p. 1823. M. I. Bruce, F. G. A. Stone. Angew. Chem., 80,460 (1968). R. Mason, A. I. M. Rae, J . Chem. SOC., A, 778 (1968). H. E. Podall, J. H. Dunn, H. Shapiro, J . Am. Chem. SOC., 82, 1325 (1960). F. Calderazzo, lnorg. Chem., 4 , 293 (1965). W. Huber, H. Fuchs, Z. Anorg. Allg. Chem., 248, 256 (1941). A. Davison, J. A. McCleverty, G. Wilkinson, J . Chem. SOC., 1133 (1963). G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1826 and references therein. F. Calderazzo, U. Mazzi, G. Pampaloni, R. Poli, F. Tisato, P. F. Zanazzi, Gazz. Chim. Zfal., 119, 241 (1989).

14.6.2.2.2. Reactions of the Carbonyls of Mn and Re.

-

Reduction of Mn,(CO),, and Re,(CO),, with alkali metals yields the corresponding carbonyl metalate anions [M(CO),]- (M = Mn,Re):

+ 2 Na

M,(CO),,

2 NaM(CO),

(a)

The reduction of Mn,(CO),, proceeds smoothly under an inert atmosphere, usually in tetrahydrofuran as solvent, yielding only the corresponding pentacarbonylmanganate( - I) anion's'. Synthesis of the [Re(CO),] - anion is more critical and the product is frequently contaminated by other carbonylrhenates. Reduction of Re,(CO),, by Na/Hg in dimethyl ethe? yields pure Na[Re(CO),], whereas in tetrahydrofuran the solvated species Na[Re(CO),] n(thf), is obtained. Aqueous alkaline solutions of [Mn(CO),] - are stable under inert atmospheres. The pentacarbonylmanganate( - I) anion readily undergoes reaction with electrophilic reagents such as dilute H3P0,'v2 with HCl in nonaqueous media to produce water-free MnH(C0),4, and with alkyl and acyl halides5:

-

[Mn(CO),][Mn(CO),][Mn(CO),]-

+ Hf + RX

+ RCOX

--

-

MnH(CO),

+ RMn(CO), RCOMn(CO), + XX-

(b) (c) (4

Depending on R, several alkyl and acyl pentacarbonylmanganese(1) derivatives are interconvertible with C06.

+ CO I RCOMn(CO),

RMn(CO),

-

(el

Oxidation of the manganese and rhenium carbonyls by halogens occurs: M,(CO),,

+ X,

2 MX(CO),

(f)

X=Cl,Br; M=Mn7, Re8-"

X=I; M=Re1's'2 Reactions in equation (f) occur smoothly at room T (Br,) or lower (Cl,), normally in halogenated solvents. MnI(CO), is better prepared by I, oxidation of [Mn(CO),]- I 3 - l 5 :


340

14.6.2. Metal Carbonyls Im ortant in Catalysis 14.6.2.2. Manganese and I$henium Carbonyls 14.6.2.2.2. Reactions of the Carbonyls of Mn and Re.

E. 0. Brimm, M. A. Lynch, W. J. Sesny,J. Am. Chem. SOC., 76, 3831 (1954). R. D. Closson, L. R. Buzbee, G. G. Ecke, J . Am. Chem. SOC., 80,6167 (1958). W. Hieber, G. Wagner, Z. Naturforsch., 12b, 478 (1957). R. B. King, J. C. Stokes, T. F. Korenowski, J . Organometal. Chem., 1 1 , 641 (1968). 5 . G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stutt-

1. 2. 3. 4. 6. 7. 8. 9.

10.

11. 12. 13.

gart, 1981, p. 1823. M. I. Bruce, F. G. A. Stone. Angew. Chem., 80,460 (1968). R. Mason, A. I. M. Rae, J . Chem. SOC., A, 778 (1968). H. E. Podall, J. H. Dunn, H. Shapiro, J . Am. Chem. SOC., 82, 1325 (1960). F. Calderazzo, lnorg. Chem., 4 , 293 (1965). W. Huber, H. Fuchs, Z. Anorg. Allg. Chem., 248, 256 (1941). A. Davison, J. A. McCleverty, G. Wilkinson, J . Chem. SOC., 1133 (1963). G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1826 and references therein. F. Calderazzo, U. Mazzi, G. Pampaloni, R. Poli, F. Tisato, P. F. Zanazzi, Gazz. Chim. Zfal., 119, 241 (1989).

14.6.2.2.2. Reactions of the Carbonyls of Mn and Re.

-

Reduction of Mn,(CO),, and Re,(CO),, with alkali metals yields the corresponding carbonyl metalate anions [M(CO),]- (M = Mn,Re):

+ 2 Na

M,(CO),,

2 NaM(CO),

(a)

The reduction of Mn,(CO),, proceeds smoothly under an inert atmosphere, usually in tetrahydrofuran as solvent, yielding only the corresponding pentacarbonylmanganate( - I) anion's'. Synthesis of the [Re(CO),] - anion is more critical and the product is frequently contaminated by other carbonylrhenates. Reduction of Re,(CO),, by Na/Hg in dimethyl ethe? yields pure Na[Re(CO),], whereas in tetrahydrofuran the solvated species Na[Re(CO),] n(thf), is obtained. Aqueous alkaline solutions of [Mn(CO),] - are stable under inert atmospheres. The pentacarbonylmanganate( - I) anion readily undergoes reaction with electrophilic reagents such as dilute H3P0,'v2 with HCl in nonaqueous media to produce water-free MnH(C0),4, and with alkyl and acyl halides5:

-

[Mn(CO),][Mn(CO),][Mn(CO),]-

+ Hf + RX

+ RCOX

--

-

MnH(CO),

+ RMn(CO), RCOMn(CO), + XX-

(b) (c) (4

Depending on R, several alkyl and acyl pentacarbonylmanganese(1) derivatives are interconvertible with C06.

+ CO I RCOMn(CO),

RMn(CO),

-

(el

Oxidation of the manganese and rhenium carbonyls by halogens occurs: M,(CO),,

+ X,

2 MX(CO),

(f)

X=Cl,Br; M=Mn7, Re8-"

X=I; M=Re1's'2 Reactions in equation (f) occur smoothly at room T (Br,) or lower (Cl,), normally in halogenated solvents. MnI(CO), is better prepared by I, oxidation of [Mn(CO),]- I 3 - l 5 :

-

14.6.2. Metal Carbonyls Im ortant in Catalysis 14.6.2.2.Manganese and Ifhenium Carbonyls 14.6.2.2.2. Reactions of the Carbonyls of Mn and Re.

[Mn(CO),]-

+ I,

I-

+ MnI(C0)

34 1

(g)

The Re,(CO),, is oxidized by I, under UV irradiation": The pentacarbonyl halides of Mn(1) and Re(1) lose CO to give the corresponding dimeric tetracarbonyls:

2 MnX(CO),

~ =M2X2 2 (CO),

+ 2 CO

(h)

Decacarbonyldimanganese(0) does not show a tendency to give disproportionation reactions, compared to V(CO), and CO,(CO)~.Only relatively strong nitrogen bases, such as pyridine, ethylene diamine, and piperidine, promote disproportionation of Mn,(CO),, 16,17:

+ 2 nB

3 Mn,(CO),,

-

2 [MnB,] [Mn(CO),],

+

10 CO

(9

Reaction (i) occurs at 50-120°C. At room T, neat "BuNH, and Mn,(CO),, react',, with disproportionation to Mn(1) and Mn( - I): Mn,(CO),,

+ RNH, -+

[Mn(CO),(NH R)1+ [Mn(CO),]-

(3

This reaction yieldsI8 a carbamoyl complex of Mn(1): Mn,(CO),,

+ 3 BuNH, I

cis-Mn(CO),(NH,Bu)(CONHBu)

+ [Mn(CO),]- + BuNH,+

(k)

Decacarbonyldirhenium(0) with Lewis bases usually does not disproportionate, presumably because of the low accessibility of the Re(I1) oxidation state. The two carbonyls of manganese and rhenium frequently undergo carbonyl substitution by Lewis bases. With tertiary phospines, L, thermal or photochemical reactions yield mono- or bisubstituted products, (1)and (2), r e s p e ~ t i v e l y ' ~ -Substitution ~~. occurs at the axial positions, Mn,(CO),,

-c

M%(Co)gL (1) Mn2(C0)8L2

L = PR,

0

0

0

(1) (2) normally, with staggered CO gro~ps'~-'~. However, monodentate tertiary arsines (R,As) can substitute at the equatorial positions27, the most favorable positions statistically, presumably due to the longer Mn-As and As-C bonds and to the lower steric hindrance and/or repulsion among the equatorial ligands. At elevated T (in boiling xylene), Mn,(CO),, and AsR, form the diamagnetic Mn(CO),(AsR,),, assumed to be (3)28,29:

342

14.6.2. Metal Carbonyls Irn ortant in Catalysis 14.6.2.2. Manganese and I$henium Carbonyls 14.6.2.2.2. Reactions of the Carbonyls of Mn and Re.

Mn,(CO),,

+ 2 ASR,

-

Mn,(CO),(AsR,),

+ 2 CO + 2 R'

( 4

R R \ / AS /-\ (OC),Mn Mn(OC), \ / AS / \ R R

-

The reaction is like that of Mn,(CO),, with TePh,, which forms dimeric diamagnetic Mn,( CO),(TePh);': Mn,(CO),,

+ 2 TePh,

Mn,(CO), (TePh),

+ Ph, + 2 CO

(4

(F. CALDERAZZO) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

W. Hieber, G . Wagner, Z. Naturjorsch., 12b, 478 (1957). R. B. King, F. G. A. Stone, Inorg. Syn., 7, 198 (1963). W. Beck, W. Hieber, G . Braun, Z . Anorg. Allg. Chem., 308, 23 (1961). F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, Vol. 1, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1967, p. 190. R. D. Closson, J. Kozikowski, T. H. Coffield, J . Org. Chem., 22, 598 (1957). T. H. Coffield, J. Kozikowski, R. D. Closson, J. Org. Chem., 22, 598 (1957). E. W. Abel, G. Wilkinson, J. Chem. SOC., 1501 (1959). E. W. Abel, G. B. Hargreaves, G. Wilkinson, J . Chem. Soc., 3149 (1958). G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1951. D. Vitali, F. Calderazzo, Gazz. Chim. Ital., 102, 587 (1972). K. Moedntzer, Synth. Inorg. Met. Org. Chem., I , 63 (1971). G. Brauer. Handbuch der PraDarativen Anornanischen Chemie, Vol. 3, F. Enke Verlag, - Stuttgart, 1981, p. 1952. W. Schropp, Ph.D. Thesis, Technische Hochschule, Munchen, 1960. W. Hieber. K. Wollman, Chem. Ber., 94, 305 (1961). M. H. Quick, R. J. Angelici, Inorg. Syn., 19, 161 (1979). W. Hieber, W. Beck, G. Zeitler, Angew. Chem., 73, 364 (1961). W. Hieber, W. Schropp, Z . Naturjorsch., ISb, 271 (1960). B. D. Dombek, R. J. Angelici, J. Organometal. Chem., 134,203 (1977). R. S. Nyholm, D. V. R. Rao, Proc. Chem. SOC., 130 (1959). W. Hieber, W. Freyer, Chem. Ber., 92, 1765 (1959). W. Hieber, W. Freyer, Chem. Ber., 93,462 (1960). A. G. Osbome, M. H. B. Stiddard, J. Chem. SOC., 634 (1964). R. B. King, T. F. Korenowski, J. ORganometal. Chem., 17,95 (1969). A. S.Kasenally, R. S.Nyholm, D. J. Parker, M. H. B. Stiddard, 0. J. R. Rodder, H. M. Powell, Chem. Ind. (London),2097 (1965). M. J. Bennett, R. Mason, J. Chem. SOC., A , , 75 (1968). M. Laing, E. Singleton, R. Reimann, J . Organometal. Chem., 56, C21 (1973). M. Laing, T. Ashworth, P. Sommerville, E. Singleton, R. Reimann, J . Chem. Soc., Chem. Commun., 1251 (1972). R. F. Lambert, Chem. Ind. (London),830 (1961). H. Ashton, B. Brady, A. R. Manning, J. Organometal. Chem., 221, 71 (1981). W. Hieber, T. Kruck, Chem. Ber., 95,2027 (1962).


343

14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.3. Iron and Ruthenium Carbonyls 14.6.2.3.1. Preparation of the Metal Carbonyls. 14.6.2.3. Iron and Ruthenlum Carbonyls 14.6.2.3.1. Preparation of the Metal Carbonyls.

Pentacarbonyliron(O), Fe(CO),, a yellow liquid (mp - 20.5"C) is formed quantitatively when Fel, and CO at 200°C and 200 atm are allowed to react in the presence of Cu or Ag powders as halogen acceptors'; in 40-50% yields when an NH, solution of Fe(I1) is treated with CO at 80°C for 16 h at 115 atm':

2 H,O

--

+ Fez+ + 6 CO

Fe(CO),

+ C0:- + 4 Hf

(a)

Enneacarbonyldiiron(O),Fe,(CO),, a golden-yellow solid, slightly soluble in most common organic solvents but slowly decomposing in tetrahydrofuran, is prepared by irradiation of an acetic acid solution of Fe(CO),,-?

2 Fe(CO),

hv

CO

+ Fe,(CO),

(b)

Dodecacarbonyltriiron(O), Fe,(CO),,, a black-green solid, sublimable at about 60°C under reduced pressure, decomposing at 14O-15O0C, forms by oxidation of the [FeH(CO),]- anion by CU(II)~ or by air. A quantitative yield of Fe,(CO),, is obtained by oxidizing the anion with MnO;-,. This method is used most frequently. Fe(CO),

3 [FeH(CO),]-

+ 2 OH-

+ 3 MnO, + 9 H +

+ C0:- + H + + 3 MnZf + 6 H,O

[FeH(CO),]-

(c)

Fe,(CO),,

(d)

Pentacarbonylruthenium(O), Ru(CO),, a colorless liquid (mp - 22"C), soluble in hydrocarbons, was originally prepared" from RuI, and CO (450 atm) at 170"C, but later was prepared in 50% yield by reaction of tris(acetylacetonato)ruthenium(III) with CO/H, (2:1, 200 atm) at 180°C": Ru(C5H,OZ),

+ 5 CO +

1.5 H,

-

3 C5H80,

+ Ru(CO),

(e)

Enneacarbonyldiruthenium(O), Ru,(CO),, unlike its Fe and 0 s analogs, is extremely unstable at RT. It cannot be isolated; by UV irradiation of Ru(CO), it is observed as an intermediate prior to its decomposition to Ru,(CO),, 12. Dodecacarbonyltriruthenium(O), Ru,(CO),,, an orange-red solid, decomposing at about 150°C in air, can be obtained by (1) Zn reduction of RuCl, in methanol or ethanol under CO pressures of 10 atm13-15, or (2) reaction of Ru(C,H,O,), with CO/H, in methanoli6. Preliminary preparation of the acetylacetonato Ru(II1) complex can be avoided by contacting RuCl, with NaC,H,O, in high pressure vessel in the presence of a CO/H, mixture',:

3 Ru(C,H,O,),

+

12 CO

+ 4.5 H,

-

Ru,(CO),,

+ 9 C,H,O,

(f)

Synthesis of Ru,(CO),, using atmospheric pressure CO involves first treating RuCl,.xH,O with CO in refluxing 2-ethoxyethanol to give a halo-carbonyl complex of ruthenium, which is then reduced by Zn to the product ( 7 5 4 0 % yield).', (F. CALDERAZZO) 1 . W. Hieber, H. Behrens, U. Teller, Z. Anorg. Allg. Chem., 249, 26 (1942). 2. W. Reppe, Ann. Chem., 582, 116 (1953).

344

14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.3. Iron and Ruthenium Carbonyls 14.6.2.3.2. Reactions of the Carbonyls of Iron and Ruthenium.

3. E. Speyer, H. Wolf, Chem. Ber, 60, 1424 (1927). 4. G. Eyber, Z. Phys. Chem., 144A, 1 (1929). 5 . G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1827. 6. H. Freundlich, E. J. Cuy, Chem. Ber., 56, 2264 (1923). 7. W. Hieber, G. Brendel, Z. Anorg. Allg. Chem., 289, 324 (1957). 8. R. B. King, F. G. A. Stone, Inorg. Syn., 7, 193 (1963). 9. G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1828. 10. W. Manchot, W. J. Manchot, Z . Anorg. Allg. Chem., 226, 385 (1936). 11. F. Calderazzo, R. L'Eplattenier,Inorg. Chem., 6, 1220 (1967). 12. J. R. Moss, W. A. G. Graham, J. Chem. SOC.,Dalton Trans., 95, (1977). 13. M. I. Bruce, F. G. A. Stone,J. Chem. SOC.,A, 1238 (1967). 14. G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1831. 15. J. L. Dawes, J. D. Holmes, Inorg. Nucl. Chem., Lett., 7, 847 (1971). 16. G. Braca, G. Sbrana, P. Pino, Chim. Ind. (Milan), 50, 121 (1968),

14.6.2.3.2. Reactions of the Carbonyls of lron and Ruthenium.

Reductions of iron carbonyls have been extensively studied; the carbonyl anions 1-4, which contain iron in a lower oxidation state, have been reported: [Fe(Co)41ZColorless

[Fe2(Co),lZOrange (1) (2) [Fe3(CO)l 11,[Fe4(C0)i3I2Red Brown (3) (4) Mononuclear anion (1) is obtained by Na metal reduction of Fe(CO), in liquid NH3's2 or in tetrahydrofuran3p4(in the latter case, sodium amalgam is preferred): Fe(CO),

+ 2 Na

-

+

Na,[Fe(CO),] CO (a) For preparative purposes, the use of sodium-benzophenone has also been reported5; alternatively the reduction of the Ph,P-substituted product can be used6:

+

K(OMe),BH

Fe(C0),PPh3 > K,[Fe(CO),] PPh, (b) The [Fe(C0),J2- anion is extremely sensitive to air and moisture. It is readily hydrolyzed, followed by loss of H:v8: [Fe(CO),]'-

+ H,O

2 [FeH(CO),]-

-

I[FeH(CO),]H,

+

+ OH-

(c)

[Fe2(Co)81z-

(4 Anion (1) is a strong nucleophile. Reaction (e) has been reported5; H+ attack on the anionic acyl derivative gives aldehydes, similar to the already encounteredg reaction of [Cr(CO),COR] - anions with protons: [Fe(CO),COR][Fe(Co)41z-

+-I

(el RCOX

[Fe(CO),COR]-


344


3. E. Speyer, H. Wolf, Chem. Ber, 60, 1424 (1927). 4. G. Eyber, Z. Phys. Chem., 144A, 1 (1929). 5 . G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1827. 6. H. Freundlich, E. J. Cuy, Chem. Ber., 56, 2264 (1923). 7. W. Hieber, G. Brendel, Z. Anorg. Allg. Chem., 289, 324 (1957). 8. R. B. King, F. G. A. Stone, Inorg. Syn., 7, 193 (1963). 9. G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1828. 10. W. Manchot, W. J. Manchot, Z . Anorg. Allg. Chem., 226, 385 (1936). 11. F. Calderazzo, R. L'Eplattenier,Inorg. Chem., 6, 1220 (1967). 12. J. R. Moss, W. A. G. Graham, J. Chem. SOC.,Dalton Trans., 95, (1977). 13. M. I. Bruce, F. G. A. Stone,J. Chem. SOC.,A, 1238 (1967). 14. G. Brauer, Handbuch der Praparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1831. 15. J. L. Dawes, J. D. Holmes, Inorg. Nucl. Chem., Lett., 7, 847 (1971). 16. G. Braca, G. Sbrana, P. Pino, Chim. Ind. (Milan), 50, 121 (1968),

14.6.2.3.2. Reactions of the Carbonyls of lron and Ruthenium.

Reductions of iron carbonyls have been extensively studied; the carbonyl anions 1-4, which contain iron in a lower oxidation state, have been reported: [Fe(Co)41ZColorless

[Fe2(Co),lZOrange (2)

(1) [Fe3(CO)l 11,Red

[Fe4(C0)i3I2Brown (3) (4) Mononuclear anion (1) is obtained by Na metal reduction of Fe(CO), in liquid NH3's2 or in tetrahydrofuran3p4(in the latter case, sodium amalgam is preferred): Fe(CO),

+ 2 Na

-

+

Na,[Fe(CO),] CO (a) For preparative purposes, the use of sodium-benzophenone has also been reported5; alternatively the reduction of the Ph,P-substituted product can be used6:

+

K(OMe),BH

Fe(C0),PPh3 > K,[Fe(CO),] PPh, (b) The [Fe(C0),J2- anion is extremely sensitive to air and moisture. It is readily hydrolyzed, followed by loss of H:v8: [Fe(CO),]'-

+ H,O

2 [FeH(CO),]-

-

I[FeH(CO),]H,

+

+ OH-

(c)

[Fe2(Co)81z-

(4 Anion (1) is a strong nucleophile. Reaction (e) has been reported5; H+ attack on the anionic acyl derivative gives aldehydes, similar to the already encounteredg reaction of [Cr(CO),COR] - anions with protons: [Fe(CO),COR][Fe(Co)41z-

+-I

(el RCOX

[Fe(CO),COR]-

-

14.6.2. Metal Carbonyis important in Catalysis 14.6.2.3. lron and Ruthenium Carbonyls 14.6.2.3.2. Reactions of the Carbonyls of Iron and Ruthenium.

+ H+

[Fe(CO),COR]-

RCHO

345

(f)

Proton attack on [Fe(C0),l2- gives the hydride FeH2(CO)10*'1,a light-yellow substance (m.p. - 70°C), which decomposes rapidly at RT. The [Fe(C0),l2- anion, isoelectronic with Ni(CO),, has a distorted tetrahedral structure, depending on the c o u n t e r c a t i ~ n ~ ~ * ~ ~ . The [Fez(C0),lz- anion is best prepared by Na/Hg reduction of Fe(CO), under UV irradiationi4:

-

2 Fe(CO),

hv

+ 2 Na

Na,[Fe,(CO),]

+ 2 CO

(g)

The neutral iron carbonyls, [Fe(CO),, Fe,(CO),, and Fe,(C0)12] in reactions with aqueous alkali undergo reduction^'^*'^: Fe,(CO),

+ 4 OH-

[Fe,(CO),-J-

+ C0:- + 2 HzO

( m = l , n = 5 ; m = 2 , n = 9 ; m=3,n= 12)

(h)

Reductions of ruthenium carbonyls have been studied less. Reduction of Ru,(CO),, by Na in liquid NH, yields17 the [Ru(C0),lz- anion, probably in admixture with [RuH(CO),] - . Acidification of the anion with dilute yields the thermally unstable RuH,(CO),. The analogous OsH,(CO), is a rather thermally stable compound" showing that the thermal stability of the MH,(CO), complexes increases drastically from Fe and Ru to 0 s ; Fe = Ru ((0s. Ligand substitution Ru,(CO),, with NaBH, occurs in tetrahydrofuran": Ru,(CO)i,

+ H-

thf

CO

+ [ R u ~H(CO)ii]-

(0

Reaction of Ru,(CO),, dissolved in tetrahydrofuran with saturated aqueous KOH gives good yields of the hexanuclear cluster anion [Ru6(C0)16]2- ' O :

2 RU,(CO)i,

4 OH-

* [RU6(CO)i6]'- + 7 CO + CO:- + 2 HzO

(j)

Typical oxidations of iron carbonyls are shown by reactions with halogens. In aqueous solution with C1, or Br,, decomposition to Fe(II1) is obtained due to the large hydration energy of the cation. However, under controlled conditions and in nonaqueous solvents, the dihalogenotetracarbonyl complexes of Fe(I1) form: Fe(CO),

+ X,

-+ Fe(CO),X, (X = Cl,Br,I)

+

CO

Reaction (k) occurs through intermediate formation of thermally unstable 1:1 adducts formulated" as Fe(CO),.X,. The adducts may be halo-acyl derivatives of Fe(II), Fe(CO)4X(COX).22 Oxidation of Fe(CO), to Fe(I1) with simultaneous loss of the CO groups occurs by reaction with substances containing active protons, such as a~etylacetone'~or Schiff basesz4:

+ 2 C5H80, Fe(CO), + H,Salen

Fe(CO),

--

+ Fe(C,H70,), + H, FeSalen + 5 CO + H, 5 CO

[Salen = conjugate base of N,"-ethylenebis(salicylideneimine)]

(1)

(m)

346


The behavior of Ru(0) toward acetylacetone is remarkably different,,. Here the intermediate cis-bis(acety1acetonato)dicarbonylruthenium(I1) is sufficiently stable to be isolated.

'/3 R U ~ ( C O ) ~+, 2 C5HsO2 -+ H2

+ 2 CO + Ru(CO),(C,H,O,),

(n)

Both R u ( C O ) , ~and ~ R u ~ ( C react ~ ) with ~ ~halogens ~ ~ below 0°C to give the corresponding tetracarbonyldihalo complexes: Ru(CO), Ru3(C0),,

-

+ X2 +Ru(CO),X, + CO

+ 3 X,

(0)

3 Ru(CO),X,

(PI

The tetracarbonyl derivatives rapidly decompose at RT to the dimeric halide-bridged tricarbonyl complexes with loss of C026,27:

2 Ru(C0)4X,

-

0 C I/CO Ru

X

oc\ I / x \ Ru oc/I'x'I\co C 0

+2co

(9)

X

These complexes also can be obtained from RU,(CO)~,rection with CHX,(X = Cl,Br)28. Other oxidations of iron carbonyls have been reported. Fe(CO), and o-phthalonitrile react in dimethylformamide (dmf) to yield a carbonyl phthalocyaninatoiron(II), [FePc] derivati~e~~.~'.

4 O-C6H4(CN)2 + Fe(CO),

+ dmf

-

-

FePc(dmf)CO

+ 4 co

Dialkyldisulfides with Fe,(CO),, undergo cleavage of the S-S Fe2(C0)6(SR),, an SR-bridged dimer of Fe(I1) with a metal-metal 2 Fe,(CO),,

+ 3 S,R,

3 Fe, (CO), (SR),

(4

bond to form

+ 6 CO

6)

Disproportionation reactions are common for iron carbonyls, especially with Lewis bases that contain nitrogen and oxygen donor atoms. Fe(CO), reacts with pyridine under UV i r r a d i a t i ~ n ~ ~ , ~ ~ : 5 Fe(CO),

+ 6 C,H,N

-+

[Fe(C,H,N),]

[Fe,(CO),,]

+

12 CO

(0

The other carbonylferrate anions, (1)-(3), have been obtained by reaction of Fe(CO), with nitrogen bases; the anion formed is a function of experimental condition^^^-^^. The primary reaction products are of interest; it has been suggested35936 that amine.Fe(CO), adducts form prior to the disproportionation reaction. The IR spectra of products arising from reaction between Fe(CO), and piperidine show [FeH(CO),] - is presen?'. More recently41, it has been shown that only primary and secondary amines react with Fe(CO),. Reaction with primary amines might yield carbamoyl derivatives as primary products: Fe(CO),

+ 2 RNH,

RNH3+

+

[Fe(CO), (CONHR)]-

(u)

Substitution reactions are common for carbonyls of iron, particularly with substances that contain C-C unsaturated bonds, with Lewis bases containing Group V (P,


-

347

As, Sb, Bi) donor atoms, and with isocyanides. Reaction with butadiene to give the pale yellow tricarbonyl product (m.p. 19°C) is t y p i ~ a l ~ ~ . ~ ~ : Fe(CO),

+ C4H,

Fe(CO,C,H,

+ 2 CO

(v)

Cyclobutadiene was predictedu to be stabilized by complexation to transition metals; the prediction was confirmed45by isolation of cyclobutadienetricarbonyliron(O),a yellow crystalline solid (m.p. 26"C, b.p. 68-7OoC), obtained by dehalogenation of 3,4-cisdichloroctacyclobutene:

clD c1

Pentacarbonyliron and Fe,(CO), react with monodentate tertiary phosphines, arsines, and stibines to give substitution products of the general formula Fe(CO),-,(ER,),. Usually substitution does not proceed beyond n = 2: Fe(CO),

+ n ER,

-

Fe(CO),-, (ER,),

+ n CO

(x)

Similar substitution reactions with isocyanides occur, which seldom proceed beyond disubstitution, even under drastic condition^^^. Fe(CO),

+ n CNR

Fe(CO),-, (CNR),

(n 5 2)

+ n CO

(Y)

(F. CALDERAZZO)

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

12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

H. Behrens, Z . Naturforsch., 7b, 321 (1952). H. Behrens, R. Weber, Z . Anorg. Allg. Chem., 281, 190 (1955). W. Hieber, 0. Vohler, G. Braun, Z . Naturforsch., 13b, 192 (1958). W. Hieber, G. Braun, Z . Naturforsch.. 14b, 132 (1959). J. P. Collman, R. G. Finke, J. N. Cawse, J. I. Brauman, J. Am. Chem. SOC.,99,2515 (1977). Y. S. Chen, J. E. Ellis, J . Am. Chem. SOC., 104, 1141 (1982), and references therein. H. W. Sternberg, R. Markby, I. Wender, J . Am. Chem. SOC., 78,5704 (1956). H. W. Sternberg, R. Markby, I. Wender, J . Am. Chem. SOC., 79,6116 (1957). E. 0. Fischer, A. Massbiil, Chem. Ber., 100, 2445 (1967). W. Hieber, F. Leutert, Chem. Ber., 64, 2832 (1931) E. A. McNeil, F. R. Scholer, J . Am. Chem. SOC.,99, 6243 (1977). A recent electron diffraction study. H. B. Chin, R. Bau, J . Am. Chem. SOC., 98,2434 (1976). R. G. Teller, R. G. Finke, J. P. Collman, H. B. Chin, R. Bau, J . Am. Chem. SOC.,99, 1104 (1977). J. K. Ruff, Inorg. Chem., 7, 1818 (1968). W. Hieber, F. Leutert, E. Schmidt, Z. Anorg. Allg. Chem., 204, 145 (1932). W. Hieber, G. Brendel, Z . Anorg. Allg. Chem., 289, 338 (1957). A, 2162 (1968). J. D. Cotton, M. I. Bruce, F. G. A. Stone, J . Chem. SOC., F. L'Eplattenier, F. Calderazzo, Inorg. Chem., 6, 2092 (1967). B. F. G. Johnson, J. Lewis, P. R. Raithby, G. Suss, J . Chem. SOC., Dalton Trans., 1356 (1979). P. F. Jackson, B., F. G. Johnson, J. Lewis, M. McPartlin, W. J. H. Nelson, J. Chem. SOC., Chem. Communl., 735 (1979). W. Hieber, G. Bader, Z . Anorg. Allg. Chem., 190, 193 (1930).

340 22. 23. 24. 25. 26. 27. 28. 29.

14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.4. Cobalt Carbonyls 14.6.2.4.1. Preparation of Cobalt Carbonyls.

K. Noack, J . Organometal. Chem., 13,411 (1968). J. W. Fitch, J. J. Lagowski, Inorg. Chem., 4,910 (1965) F. Calderazzo, C. Floriani, R. Henzi, F. L’Eplattenier, J . Chem. Soc., A, 1378 (1969). F. Calderazzo, F. L’Eplattenier, Inorg. Chem., 6, 1220 (1967). B. F. G . Johnson, R. D. Johnson, J. Lewis, J . Chem. Soc., A, (1969). A. Trovati, A. Araneo, P. Uguagliati, F. Zingales, Inorg. Chem., 9, 671 (1970). G. Braca, G. Sbrana, E. Benedetti, P. Pino, Chim. Ind. (Milan),51, 1245 (1969). F. Calderazzo, D. Vitali, G. Pampaloni, I. Collamati, J . Chem. Soc., Chem. Commun., 221

(1979). 30. F. Calderazzo, G. Pampaloni, D. Vitali, G. Pelizzi, I. Collamati, S. Frediani, A. M. Serra, J . Organometal. Chem., 191, 217 (1980). 31. R. B. King, J . Am. Chem. Soc., 84,2460 (1962). 32. R. B. King, M. B. Bisnette, Inorg. Chem., 4 , 1663 (1965). 33. W. Hieber, R. Werner, Chem. Ber., 90,286 (1957). 34. W. Hieber, E. H. Schubert, 2. Anorg. Allg. Chem., 338, 37 (1965). 35. W. Hieber, J. G. Floss, Chem. Ber., 90, 1617 (1957). 36. W. Hieber, N. Kahlen, Chem. Ber., 91,2223 (1958). 37. W. Hieber, A. Lipp, Chem. Ber., 92,2075 (1959). 38. W. Hieber, R. Werner, Chem. Ber., 90, 1116 (1957). 39. W. Hieber, N. Kahlen, Chem. Ber., 91,2234 (1958). 40. W. F. Edgell, M. T. Yang, B. J. Buekin, R. Bayer, N. Kaizumi, J . Am. Chem. Soc., 87, 3080 (1965). 41. J. R. Miller, B. D. Podd, M. 0. Sanchez, J . Chem. Soc., Dalton Trans., 1461 (1980). 42. H. Reihlen, A. Gruhl, G. Hessling, 0. Pfrengle, Ann. Chem., 482, 161 (1930). 43. B. F. Hallam, P. L. Pauson, J . Chem. Soc., 642 (1958). 44. H. C. Longuet-Higgins, L. E. Orgel, J . Chem. SOC., 1969 (1956). 45. G. F. Emerson, L. Watts, R. Pettit, J . Am. Chem. Soc., 87, 131 (1965). 46. F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, Vol. 1, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1967, p. 120.

14.6.2.4. Cobalt Carbonyls 14.6.2.4.1. Preparatlon of Cobalt Carbonyls.

Several methods for preparation of Co,(CO),, an orange crystalline solid (m.p. 51-52°C with decomposition) have been reported. The most commonly used method is the treatment of Co(I1) salts, normally anhydrous acetate’,’ or carbonate3 with a CO/H, mixture. Reactions are carried out at about 200 atm pressure and 130°C. Anhydrous Co(I1) acetate can be avoided if acetic anhydride is used as reaction medium. 2 Co(CH3COO)z 2 CoCO,

CO,(CO)~+ 4 CH3COOH + 8 CO + 2 H, + 8 CO + 2 H, --+ 2 CO, + 4 H,O + Co,(CO),

(a) (b)

An aluminum-alkyl method has been used4 for preparation of Co,(CO),; using 3:l (mole ratio) AlR,/Co(II), Co,(CO), is the main product. Using a 1:l reagent ratio, Co(CO),(COC,H,), isolated from the reaction mixture as the PPh, substitution product, Co(CO),(COC,H,)(PPh,), was the main product (in this case AlEt, was used as alkylating agent). Synthesis of Co,(CO), from soluble Co(I1) salts in hydrocarbon solvents has been studied5s6.The presence of preformed Co,(CO), increases the reaction rate. Lewis bases also enhance reaction rates, e.g., addition of acetone to a toluene solution of Co(II)-2ethylhexanoate under super atmospheric CO/H, pressure allows the carbonylation re-


340 22. 23. 24. 25. 26. 27. 28. 29.

14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.4. Cobalt Carbonyls 14.6.2.4.1. Preparation of Cobalt Carbonyls.

K. Noack, J . Organometal. Chem., 13,411 (1968). J. W. Fitch, J. J. Lagowski, Inorg. Chem., 4,910 (1965) F. Calderazzo, C. Floriani, R. Henzi, F. L’Eplattenier, J . Chem. Soc., A, 1378 (1969). F. Calderazzo, F. L’Eplattenier, Inorg. Chem., 6, 1220 (1967). B. F. G . Johnson, R. D. Johnson, J. Lewis, J . Chem. Soc., A, (1969). A. Trovati, A. Araneo, P. Uguagliati, F. Zingales, Inorg. Chem., 9, 671 (1970). G. Braca, G. Sbrana, E. Benedetti, P. Pino, Chim. Ind. (Milan),51, 1245 (1969). F. Calderazzo, D. Vitali, G. Pampaloni, I. Collamati, J . Chem. Soc., Chem. Commun., 221

(1979). 30. F. Calderazzo, G. Pampaloni, D. Vitali, G. Pelizzi, I. Collamati, S. Frediani, A. M. Serra, J . Organometal. Chem., 191, 217 (1980). 31. R. B. King, J . Am. Chem. Soc., 84,2460 (1962). 32. R. B. King, M. B. Bisnette, Inorg. Chem., 4 , 1663 (1965). 33. W. Hieber, R. Werner, Chem. Ber., 90,286 (1957). 34. W. Hieber, E. H. Schubert, 2. Anorg. Allg. Chem., 338, 37 (1965). 35. W. Hieber, J. G. Floss, Chem. Ber., 90, 1617 (1957). 36. W. Hieber, N. Kahlen, Chem. Ber., 91,2223 (1958). 37. W. Hieber, A. Lipp, Chem. Ber., 92,2075 (1959). 38. W. Hieber, R. Werner, Chem. Ber., 90, 1116 (1957). 39. W. Hieber, N. Kahlen, Chem. Ber., 91,2234 (1958). 40. W. F. Edgell, M. T. Yang, B. J. Buekin, R. Bayer, N. Kaizumi, J . Am. Chem. Soc., 87, 3080 (1965). 41. J. R. Miller, B. D. Podd, M. 0. Sanchez, J . Chem. Soc., Dalton Trans., 1461 (1980). 42. H. Reihlen, A. Gruhl, G. Hessling, 0. Pfrengle, Ann. Chem., 482, 161 (1930). 43. B. F. Hallam, P. L. Pauson, J . Chem. Soc., 642 (1958). 44. H. C. Longuet-Higgins, L. E. Orgel, J . Chem. SOC., 1969 (1956). 45. G. F. Emerson, L. Watts, R. Pettit, J . Am. Chem. Soc., 87, 131 (1965). 46. F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, Vol. 1, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1967, p. 120.

14.6.2.4. Cobalt Carbonyls 14.6.2.4.1. Preparatlon of Cobalt Carbonyls.

Several methods for preparation of Co,(CO),, an orange crystalline solid (m.p. 51-52°C with decomposition) have been reported. The most commonly used method is the treatment of Co(I1) salts, normally anhydrous acetate’,’ or carbonate3 with a CO/H, mixture. Reactions are carried out at about 200 atm pressure and 130°C. Anhydrous Co(I1) acetate can be avoided if acetic anhydride is used as reaction medium. 2 Co(CH3COO)z 2 CoCO,

CO,(CO)~+ 4 CH3COOH + 8 CO + 2 H, + 8 CO + 2 H, --+ 2 CO, + 4 H,O + Co,(CO),

(a) (b)

An aluminum-alkyl method has been used4 for preparation of Co,(CO),; using 3:l (mole ratio) AlR,/Co(II), Co,(CO), is the main product. Using a 1:l reagent ratio, Co(CO),(COC,H,), isolated from the reaction mixture as the PPh, substitution product, Co(CO),(COC,H,)(PPh,), was the main product (in this case AlEt, was used as alkylating agent). Synthesis of Co,(CO), from soluble Co(I1) salts in hydrocarbon solvents has been studied5s6.The presence of preformed Co,(CO), increases the reaction rate. Lewis bases also enhance reaction rates, e.g., addition of acetone to a toluene solution of Co(II)-2ethylhexanoate under super atmospheric CO/H, pressure allows the carbonylation re-

14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.4. Cobalt Carbonyls 14.6.2.4.2. Reactions of Cobalt Carbonyls.

349

action to be carried out at RT. Both Co,(CO), and [Co(acetone),][Co(CO),], form. The role of preformed Co,(CO), and Lewis base in this reaction could be explained as Co2(CO),

2 CoH(CO), co[co(co),l,

+ H,

+ COX, + 4 co

-

2 CoH(CO), 2 HX

(c)

+ CO[CO(CO),],

% co, (CO),

(4 (el

The Co,(CO), can be prepared at atmospheric pressure and at RT by reducing anhydrous CoI, with a stoichiometric amount of Zn powder in the presence of CO in toluene containing small amounts of t-butyl alcohol (yields -80%)':

2 CoI,

+ 2 Zn + 8 CO

+ 2 ZnI,

CO,(CO)~

(f)

Excess zinc should be avoided since the mixed-metal carbonyl Zn[Co(CO),], forms instead in good yields. Dodecacarbonyltetracobalt(O), Co,(CO),,, a black crystalline substance slightly soluble in hydrocarbons, is prepared by thermal decomposition of Co2(CO), without solvent' or, better, in a hydrocarbon solvent?.". The kinetics of CO,(CO),, formation from Co,(CO),

2 Co,(CO),

CO,(CO),~

+ 4 CO

(g)

have been reinvestigated". The thermodynamics of reaction in the melt', and in hexane ~olution'~ [equation (g) have been studied. The AW of Co,(CO)8 to Co4(COl2conversion is positive, i.e., reaction is endothermic (No= 125 kJ/mol-'). Preparation of Co,(CO),, has also been rep~rted'~: using the carbonyl redistribution reaction that occurs when soluble Co(I1) salts react with Co,(CO), and H, at 30°C for 8 h. For bis(acetylacetonato)cobalt(II), reaction stoichiometry can be represented as

2 Co(C,H,O,),

+ 3 CO,(CO)~+ 2 H,

-

4 C,H,02

+ 2 CO,(CO),,

(h)

(F. CALDERAZZO)

1. P. Szab6, L. Mark6, G. Bor, Chem. Techn. (Berlin),13, 549 (1961). 2. G. Brauer, Handbuch der Priparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1833. 3. I. Wender, H. W. Stemberg, S . Metlin, M. Orchin, Inorg. Syn., 5 , 190 (1957). 4. P. Szabb, L. Markb, J . Organometal. Chem., 3, 364 (1965). 5. P. Chini, Chim. Ind. (Milan),4 2 , 133 (1960). 6. P. Chini, Chim. Ind. (Milan),4 2 , 137 (1960). 7. P. Chini, M. C. Malatesta, A. Cavalieri, Chim. Ind. (Milan), 55, 120 (1973). 8. L. Mond, H. Hirtz, M. D. Cowap, J . Chem. SOC., 790 (1910). 9. R. A. Friedel, I. Wender, S. L. Shufler, H. W. Stemberg, J . Am. Chem. SOC.,77,3951 (1955). 10. G. Natta, R. Ercoli, S. Castellano, Chim. Ind. (Milan),37, 6 (1955). 11. M. F. Mirbach, A. Saw, A. M. Krings, M. J. Mirbach, J . Organometal. Chem.,205,229 (1981). 12. R. Ercoli, F. Barbieri-Herrnitte, Atti Accad. Naz. Lincei, Cl. Sci. Fis. Mat. Nat. Rend., 16, 249 (1954). 13. G . Bor, U. K. Dietler, J . Organometal. Chem., 191, 295 (1980). 14. R. Ercoli, P. Chini, M. Massi Mauri, Chim. Ind. (Milan),4 1 , 132 (1959). 14.6.2.4.2. Reactions of Cobalt Carbonyis.

The Co(CO),- anion is readily obtained by Na/Hg reduction of Co,(CO), in inert solvent':



349

action to be carried out at RT. Both Co,(CO), and [Co(acetone),][Co(CO),], form. The role of preformed Co,(CO), and Lewis base in this reaction could be explained as Co2(CO),

2 CoH(CO), co[co(co),l,

+ H,

+ COX, + 4 co

-

2 CoH(CO), 2 HX

(c)

+ CO[CO(CO),],

% co, (CO),

(4 (el

The Co,(CO), can be prepared at atmospheric pressure and at RT by reducing anhydrous CoI, with a stoichiometric amount of Zn powder in the presence of CO in toluene containing small amounts of t-butyl alcohol (yields -80%)':

2 CoI,

+ 2 Zn + 8 CO

+ 2 ZnI,

CO,(CO)~

(f)

Excess zinc should be avoided since the mixed-metal carbonyl Zn[Co(CO),], forms instead in good yields. Dodecacarbonyltetracobalt(O), Co,(CO),,, a black crystalline substance slightly soluble in hydrocarbons, is prepared by thermal decomposition of Co2(CO), without solvent' or, better, in a hydrocarbon solvent?.". The kinetics of CO,(CO),, formation from Co,(CO),

2 Co,(CO),

CO,(CO),~

+ 4 CO

(g)

have been reinvestigated". The thermodynamics of reaction in the melt', and in hexane ~olution'~ [equation (g) have been studied. The AW of Co,(CO)8 to Co4(COl2conversion is positive, i.e., reaction is endothermic (No= 125 kJ/mol-'). Preparation of Co,(CO),, has also been rep~rted'~: using the carbonyl redistribution reaction that occurs when soluble Co(I1) salts react with Co,(CO), and H, at 30°C for 8 h. For bis(acetylacetonato)cobalt(II), reaction stoichiometry can be represented as

2 Co(C,H,O,),

+ 3 CO,(CO)~+ 2 H,

-

4 C,H,02

+ 2 CO,(CO),,

(h)

(F. CALDERAZZO)

1. P. Szab6, L. Mark6, G. Bor, Chem. Techn. (Berlin),13, 549 (1961). 2. G. Brauer, Handbuch der Priparativen Anorganischen Chemie, Vol. 3, F. Enke Verlag, Stuttgart, 1981, p. 1833. 3. I. Wender, H. W. Stemberg, S . Metlin, M. Orchin, Inorg. Syn., 5 , 190 (1957). 4. P. Szabb, L. Markb, J . Organometal. Chem., 3, 364 (1965). 5. P. Chini, Chim. Ind. (Milan),4 2 , 133 (1960). 6. P. Chini, Chim. Ind. (Milan),4 2 , 137 (1960). 7. P. Chini, M. C. Malatesta, A. Cavalieri, Chim. Ind. (Milan), 55, 120 (1973). 8. L. Mond, H. Hirtz, M. D. Cowap, J . Chem. SOC., 790 (1910). 9. R. A. Friedel, I. Wender, S. L. Shufler, H. W. Stemberg, J . Am. Chem. SOC.,77,3951 (1955). 10. G. Natta, R. Ercoli, S. Castellano, Chim. Ind. (Milan),37, 6 (1955). 11. M. F. Mirbach, A. Saw, A. M. Krings, M. J. Mirbach, J . Organometal. Chem.,205,229 (1981). 12. R. Ercoli, F. Barbieri-Herrnitte, Atti Accad. Naz. Lincei, Cl. Sci. Fis. Mat. Nat. Rend., 16, 249 (1954). 13. G . Bor, U. K. Dietler, J . Organometal. Chem., 191, 295 (1980). 14. R. Ercoli, P. Chini, M. Massi Mauri, Chim. Ind. (Milan),4 1 , 132 (1959). 14.6.2.4.2. Reactions of Cobalt Carbonyis.

The Co(CO),- anion is readily obtained by Na/Hg reduction of Co,(CO), in inert solvent':

350

-


+ 2 Na

Co,(CO),

2 Na[Co(CO),]

(a)

The reduction occurs via the intermediate formation of H~[CO(CO),],~~~. The [Co(CO),]- anion can also be directly prepared in aqueous solution from Co(I1) salts and CO in the presence of a reducing agent. With S202-, the yield of [Co(CO),]- is substantially quantitative4, as demonstrated by precipitation of the mercury derivative Hg[Co(CO)4lz. Related to the alkali metal reduction of Co,(CO),, it is observed5 that the trinuclear cluster anion [Co,(C0),,] - forms from Li[Co(CO),] and Co,(CO), in di-n-butyl ether: Co,(CO),

+ Li[Co(CO),]

C 2 CO

+ Li[Co,(CO),,]

(b)

Only with Li as countercation is the equilibrium in equation (b) shifted far right. This explains why lithium reduction of Co,(CO), in diethyl ether gives the [Co,(CO),,]anion6. Both Co,(CO), and Co,(CO),, are oxidized by air even in the solid state, and by halogens. With the latter, complete decomposition to CO and Co(I1) halides occurs. Reaction of cobalt carbonyl derivatives with I, in pyridine is frequently used for analyses by gas volumetry of CO evolved. Octacarbonyldicobalt, Co,(CO),, is attacked by organic substances that contain active protons and converted to the corresponding Co(I1) salts', as in reactions for Fe(CO), [equations (1) and (m), 14.6.2.3.21. Included here is reaction of Co,(CO), with cyclopentadiene' to give dicarbonylcyclopentadienylcobalt(1) as a dark red oil (b.p. 75"C/22 torr): Co,(CO),

+ 2 C,H,

-

2 Co(CO),C,H,

+ H, + 4 CO

(c)

Disproportionation reactions are common for Co,(CO), and CO,(CO),,. They occur with numerous Lewis bases. With Co,(CO), such reactions usually occur according to one of the following stoichiometries, depending on the Lewis base:

--

+ * [CO(CO)~-,B,][CO(CO),] + (n [CoB,][Co(CO),] + 4 CO Co,(CO), + nB 2 [CoB,][Co(CO),], + 8 CO 3 Co2(CO), + nB C O ~ ( C O ) ~ nB

1) CO

(4 (el (f)

Reaction in equation (d) occurs normally with tertiary phosphine~~-'~, reaction in equation (e) occurs with is~cyanides'~~'~, whereas reactions in equation (f) take place with Lewis bases that contain nitrogen and oxygen donor Reactions with tertiary phosphines exhibit disproportionation [equation (g)], followed by the neutral substitution product: Co,(CO),

+ 2 PR,

--

+ CO

[Co(CO), (PR,),I[Co(CO),]

[co(co),(PR,)z)[Co(CO),l

[Co(CO),PR,lz

-

+ CO

The dimeric tertiary-phosphine substituted compounds undergo a Co-Co and Na/Hg reduction in tetrahydrofuran": [Co(CO),PR,],

+ 2 Na

2 Na[Co(CO),PR,]

(€9 (h)

bond cleavage

(i)

Disproportionation reactions with monodentate, bidentate, and tridentate nitrogen bases [equation (f)] have been studied. It has been suggested', and later confirmed" in

14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.5. Nickel Carbonyls 14.6.2.5.1. Preparation of Tetracarbonylnickel(0).

351

a study carried out using alcohols as Lewis bases that the primary product of reaction is [Co(CO), -nB,][Co(CO)4]. Reactions in equations (d)-(f) would take place via a common mechanism of initial heterolytic cleavage of the Co-Co bond in Co2(CO),. Only with tertiary phosphines would the cationic carbonyl complex be stable enough to be isolated. Substitution reactions of Co,(CO), and CO,(CO),~occur with substances that contain C-C unsaturated bonds. Reaction of Co,(CO), with diphenylacetylene20-22yields the dimeric hexacarbonyl derivative:

+

+

Co,(CO), C,Ph, +2 CO Co2(CO),C2Ph, (j) This compound has only terminal CO groups; each Co atom has distorted pseudooctahedral geometry and the acetylenic carbon-carbon bond is nearly perpendicular to the Co-Co bond. (F. CALDERAZZO)

1. I. Wender, H. W. Stemberg, M. Orchin, J. Am. Chem. Soc., 74,1216 F(1952). 2. S. V. Dighe, M. Orchin, Inorg. Chem., I , 965 (1962). 3. R. B. King, J . Inorg. Nucl. Chem., 25, 1296 (1963). 4. W. Hieber, E. 0. Fischer, E. Bockly, Z . Anorg. Allg. Chem., 269,308 (1952). 5. G. Fachinetti, J . Chem. SOC.,Chem. Commun., 396 (1979). 6. S. A. Fieldhouse, B. H. Freeland, C. D. M. Mann, R. J. O'Brien, J . Chem. SOC.,Chem. Commun., 181 (1970). 7. F. Calderazzo, C. Floriani, R. Henzi, F. L'Eplattenier, J . Chem. SOC.,A, 1378 (1969). 8. T.S.Piper, F. A. Cotton, G. Wilkinson, J . Inorg. Nucl. Chem., 1 , 165 (1955). 9. W. Reppe, W. J. Schweckendiek, Ann. Chem., 560,104 (1948). 10. A. Sacco, Ann. Chim. (Rome), 43,495 (1953);Chem. Absrr., 48,5012 (1954). 11. A.Sacco, M. Freni, J . Inorg. Nucl. Chem., 8,566(1958). 12. W.Hieber, W. Freyer, Chem. Ber., 91,1230 (1958). 13. A. Sacco, Gazz. Chim. Ital., 83,632 (1953). 14. W.Hieber, J. Sedmeier, Chem. Ber., 87,789 (1954). 15. H. W. Stemberg, I. Wender, R. A. Friedel, M. Orchin, J . Am. Chem. SOC., 75,3148(1953). 16. W. Hieber, J. Sedlmeier, Chem. Ber., 87,25 (1954). 17. W. Hieber, R. Wiesboeck, Chem. Ber., 91, 1146 (1958). 18. W. Hieber, E. Lindner, Z. Naturforsch., I6b,137 (1961). 19. E. R. Tucci, B. H. Gwynn, J . Am. Chem. SOC.,86,4838(1964). 20. H. W. Stemberg, H. Greenfield, R. A. Friedel, J. H. Wotiz, R. Markby, I. Wender, J . Am. Chem. SOC.,76,1457 (1954). 21. H. Greenfield, H. W. Stemberg, R. A. Friedel; J. H. Wotiz, R. Markby, I. Wender, J . Am. Chem. SOC.,78,120 (1956). 22. W. G. Sly,J . Am. Chem. SOC., 81,18 (1959). 14.6.2.5. Nickel Carbonyls 14.6.2.5.1. Preparation of Tetracarbonylnickel(0).

Tetracarbonylnickel(O), Ni(CO),, a colorless liquid (b.p. 42.1"C, map. - 19.3"C), miscible in hydrocarbon solvents, reported in 18901 was the first metal carbonyl to be described. It was prepared by treating activated Ni metal with CO under mild conditions:

-

Ni(C0)4(1) (a) Ni,,) + 4 CO,,, Essentially quantitative yields of Ni(CO), are obtained2 when NiL, is allowed to react at 250°C with CO at 200 atm CO pressure in the presence of Cu or Ag powders as halogen acceptor.


14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.5. Nickel Carbonyls 14.6.2.5.1. Preparation of Tetracarbonylnickel(0).

351

a study carried out using alcohols as Lewis bases that the primary product of reaction is [Co(CO), -nB,][Co(CO)4]. Reactions in equations (d)-(f) would take place via a common mechanism of initial heterolytic cleavage of the Co-Co bond in Co2(CO),. Only with tertiary phosphines would the cationic carbonyl complex be stable enough to be isolated. Substitution reactions of Co,(CO), and CO,(CO),~occur with substances that contain C-C unsaturated bonds. Reaction of Co,(CO), with diphenylacetylene20-22yields the dimeric hexacarbonyl derivative:

+

+

Co,(CO), C,Ph, +2 CO Co2(CO),C2Ph, (j) This compound has only terminal CO groups; each Co atom has distorted pseudooctahedral geometry and the acetylenic carbon-carbon bond is nearly perpendicular to the Co-Co bond. (F. CALDERAZZO)

1. I. Wender, H. W. Stemberg, M. Orchin, J. Am. Chem. Soc., 74,1216 F(1952). 2. S. V. Dighe, M. Orchin, Inorg. Chem., I , 965 (1962). 3. R. B. King, J . Inorg. Nucl. Chem., 25, 1296 (1963). 4. W. Hieber, E. 0. Fischer, E. Bockly, Z . Anorg. Allg. Chem., 269,308 (1952). 5. G. Fachinetti, J . Chem. SOC.,Chem. Commun., 396 (1979). 6. S. A. Fieldhouse, B. H. Freeland, C. D. M. Mann, R. J. O'Brien, J . Chem. SOC.,Chem. Commun., 181 (1970). 7. F. Calderazzo, C. Floriani, R. Henzi, F. L'Eplattenier, J . Chem. SOC.,A, 1378 (1969). 8. T.S.Piper, F. A. Cotton, G. Wilkinson, J . Inorg. Nucl. Chem., 1 , 165 (1955). 9. W. Reppe, W. J. Schweckendiek, Ann. Chem., 560,104 (1948). 10. A. Sacco, Ann. Chim. (Rome), 43,495 (1953);Chem. Absrr., 48,5012 (1954). 11. A.Sacco, M. Freni, J . Inorg. Nucl. Chem., 8,566(1958). 12. W.Hieber, W. Freyer, Chem. Ber., 91,1230 (1958). 13. A. Sacco, Gazz. Chim. Ital., 83,632 (1953). 14. W.Hieber, J. Sedmeier, Chem. Ber., 87,789 (1954). 15. H. W. Stemberg, I. Wender, R. A. Friedel, M. Orchin, J . Am. Chem. SOC., 75,3148(1953). 16. W. Hieber, J. Sedlmeier, Chem. Ber., 87,25 (1954). 17. W. Hieber, R. Wiesboeck, Chem. Ber., 91, 1146 (1958). 18. W. Hieber, E. Lindner, Z. Naturforsch., I6b,137 (1961). 19. E. R. Tucci, B. H. Gwynn, J . Am. Chem. SOC.,86,4838(1964). 20. H. W. Stemberg, H. Greenfield, R. A. Friedel, J. H. Wotiz, R. Markby, I. Wender, J . Am. Chem. SOC.,76,1457 (1954). 21. H. Greenfield, H. W. Stemberg, R. A. Friedel; J. H. Wotiz, R. Markby, I. Wender, J . Am. Chem. SOC.,78,120 (1956). 22. W. G. Sly,J . Am. Chem. SOC., 81,18 (1959). 14.6.2.5. Nickel Carbonyls 14.6.2.5.1. Preparation of Tetracarbonylnickel(0).

Tetracarbonylnickel(O), Ni(CO),, a colorless liquid (b.p. 42.1"C, map. - 19.3"C), miscible in hydrocarbon solvents, reported in 18901 was the first metal carbonyl to be described. It was prepared by treating activated Ni metal with CO under mild conditions:

-

Ni(C0)4(1) (a) Ni,,) + 4 CO,,, Essentially quantitative yields of Ni(CO), are obtained2 when NiL, is allowed to react at 250°C with CO at 200 atm CO pressure in the presence of Cu or Ag powders as halogen acceptor.

14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.5. Nickel Carbon Is 14.6.2.5.2. Reactions oYTetracarbonylnicke1.

352

Nickel sulfide is quantitatively converted to Ni(CO), in alkaline aqueous solution3: NiS

+ 5 CO + 4 OH-

+Ni(CO),

+ Cog- + 2 H,O + S 2 -

(b)

(F. CALDERAZZO) 1. L. Mond, C. Langer, F. Quincke, J . Chem. SOC., 749 (1890). 2. W.Hieber, H. Behrens, U. Teller, Z. Anorg. Allgem. Chem., 249, 26 (1942). 3. H. Behrens, E. Eisenmann, Z. Anorg. Allg. Chem., 278, 155 (1955). 14.6.2.5.2. Reactions of Tetracarbonylnlckel.

Tetracarbonylnickel(O),a very toxic chemical, is rather reactive. It is oxidized by atmospheric 0, in the gas phase and it reacts with halogenated organic compounds to form the corresponding Ni(I1) halides’.2. Reduction of Ni(CO),, although extensively investigated3-’, has not produced conclusive results. The carbonylnickelate anions reported are (1)-(7). ~Ni2(co)61z“i5(CO)91 (4)

“i3(CO>~12-

“4(co)’12-

(1) (2) “i5(CO~1,l2 “is(C0)1,l2(5)

-

(3) “is(C0),’l2 (7)

(6)

-

Anion (5) forms by reduction of Ni(CO), with Na-naphthalene or Na-anthracene’ in tetrahydrofuran: 5 Ni(CO),

+ 2e-

[Ni5(CO)l,]2-

+ 8 CO

(a)

Anion (6) is best obtained in a redox condensation reaction [equation (b)] by reducing Ni(C0) with Na in the presence of anthracene and in boiling diethyl ether. Reaction in equation (b) is favored due to the removal of CO from the reaction mixture: [Nis(CO)l,]2-

+ Ni(CO), e [Ni6(co)1,]2- + 4 co

(b)

Based on comparison of IR spectra, [Ni4(C0),l2- might be identical to ~Ni6(CO)12~z-. Anion (7) forms” by a redox condensation involving Ni(CO), and [Ni6(co)1,]2-:

3 Ni(CO),

+

[Ni6(C0),,l2-

[Nis(CO)1812-

+ 6 CO

(c)

Tetracarbonylnickel undergoes disproportionation with nitrogen Lewis bases“. Relatively weak bases such as pyridine initially give unstable substitution products, which then disproportionate to Ni(I1) and Ni( - I):

3 Ni(CO),(C,H,N)

+ 3 C,H,N e [Ni(C,H,N),][Ni,(CO),]

+ 3 co

(d)

Addition of Ni(CO),(C,H,N) converts the dinuclear anion into polynuclear carbonylnickelates. Tetracarbonylnickel(0) undergoes substitution reactions by a dissociative process”: Ni(CO), INi(CO), Ni(CO),

+ CO (slow)

+ L -+ Ni(CO), L (fast)

(el (f)


14.6.2. Metal Carbonyls Important in Catalysis 14.6.2.5. Nickel Carbon Is 14.6.2.5.2. Reactions oYTetracarbonylnicke1.

352

Nickel sulfide is quantitatively converted to Ni(CO), in alkaline aqueous solution3: NiS

+ 5 CO + 4 OH-

+Ni(CO),

+ Cog- + 2 H,O + S 2 -

(b)

(F. CALDERAZZO) 1. L. Mond, C. Langer, F. Quincke, J . Chem. SOC., 749 (1890). 2. W.Hieber, H. Behrens, U. Teller, Z. Anorg. Allgem. Chem., 249, 26 (1942). 3. H. Behrens, E. Eisenmann, Z. Anorg. Allg. Chem., 278, 155 (1955). 14.6.2.5.2. Reactions of Tetracarbonylnlckel.

Tetracarbonylnickel(O),a very toxic chemical, is rather reactive. It is oxidized by atmospheric 0, in the gas phase and it reacts with halogenated organic compounds to form the corresponding Ni(I1) halides’.2. Reduction of Ni(CO),, although extensively investigated3-’, has not produced conclusive results. The carbonylnickelate anions reported are (1)-(7). ~Ni2(co)61z-

“i3(CO>~12“4(co)’12(1) (2) (3) “i5(CO)91 “i5(CO~1,l2 “is(C0)1,l2“is(C0),’l2 (4)

(5)

-

(6)

-

(7)

Anion (5) forms by reduction of Ni(CO), with Na-naphthalene or Na-anthracene’ in tetrahydrofuran: 5 Ni(CO),

+ 2e-

[Ni5(CO)l,]2-

+ 8 CO

(a)

Anion (6) is best obtained in a redox condensation reaction [equation (b)] by reducing Ni(C0) with Na in the presence of anthracene and in boiling diethyl ether. Reaction in equation (b) is favored due to the removal of CO from the reaction mixture: [Nis(CO)l,]2-

+ Ni(CO), e [Ni6(co)1,]2- + 4 co

(b)

Based on comparison of IR spectra, [Ni4(C0),l2- might be identical to ~Ni6(CO)12~z-. Anion (7) forms” by a redox condensation involving Ni(CO), and [Ni6(co)1,]2-:

3 Ni(CO),

+

[Ni6(C0),,l2-

[Nis(CO)1812-

+ 6 CO

(c)

Tetracarbonylnickel undergoes disproportionation with nitrogen Lewis bases“. Relatively weak bases such as pyridine initially give unstable substitution products, which then disproportionate to Ni(I1) and Ni( - I):

3 Ni(CO),(C,H,N)

+ 3 C,H,N e [Ni(C,H,N),][Ni,(CO),]

+ 3 co

(d)

Addition of Ni(CO),(C,H,N) converts the dinuclear anion into polynuclear carbonylnickelates. Tetracarbonylnickel(0) undergoes substitution reactions by a dissociative process”: Ni(CO), INi(CO), Ni(CO),

+ CO (slow)

+ L -+ Ni(CO), L (fast)

(el (f)

14.6. Carbon Monoxide Reactions 14.6.3. Hydroformylationof Olefins

353

Phenylisocyanide, completely substitutes to Ni(CNR),I3-l6: Ni(CO),

+ 4 CNPh +4 CO + Ni(CNPh),

-

(g)

Partial substitution occurs with other isocyanides. Tertiary phosphines, arsines, and stibines also partially displace the carbonyl groups Ni(C0),'7-22: Ni(CO),

+ n PR,

Ni(CO),-.

(PR,),

+ n CO

0)

(F. CALDERAZZO)

1. H. Reihlen, A. Gruhl, G. Hessling, Ann. Chem., 472, 268 (1929). 2. I. D. Webb, G. T. Borcherdt, J . Am. Chem. SOC.,73,2654 (19510. 3. H. Behrens, F. Lohofer, Z. Naturforsch., 86, 691 (1953). 4. H. Behrens, F. Lohofer, Chem. Ber., 94, 1391 (1961). 5. H. Behrens, H. Zizlsperger, R. Rauch, Chem. Ber., 94, 1497 (1961). 6. W. Hieber, W. Kroder, E. Zahn, 2.Naturforsch., 15b, 325 (1960). 7. W. Hieber, J. Ellermann, Z. Naturforsch., 18b, 595 (1963). 8. H. W. Stemberg, R. Markby, I. Wender, J . Am. Chem. Soc., 82, 3638 (1960). 9. G. Longoni, P. Chini, A. Cavalieri, Inorg. Chem., 15, 3025 (1976). 10. G. Longoni, P. Chini, Inorg. Chem., 15, 3029 (1976). 11. W. Hieber, J. Ellermann, E. Zahn, Z. Naturforsch., 18b, 589 (1963). 12. F. Basolo, A. Wojcicki, J . Am. Chem. SOC., 83, 520 (1961). 13. F. Klages, K. Monkemeyer, Naturwissenschaflen, 37, 210 (1950). 14. W. Hieber, Z. Naturforsch., 5b, 129 (1950). 15. W. Hieber, E. Bockly, Z. Anorg. A&. Chem., 262, 344 (1950). 16. M. Bigorgne, J . Organometal. Chem., 1, 101 (1963). 17. L. S . Meriwether, M. L. Fiene, J . Am. Chem. Soc., 81, 4200 (1959). 18. J. D. Rose, F. S. Statham, J . Chem. SOC., 69 (1950). 19. G. Wilkinson, J . Am. Chem. SOC.,73, 5502 (1951). 20. A. B. Burg, W. Mahler, J . Am. Chem. SOC.,80, 2334 (1958). 21. R. J. Clark, E. 0. Brimm, Inorg. Chem., 4,651 (1965). 22. W. Reppe, W. J. Schweckendiek, Ann. Chem., 560, 104 (1948).

14.6.3. Hydroformylationof Olef ins Hydroformylation involves addition of the elements of H, and CO' added to an olefinic substrate:

The reaction was discovered in Germany prior to World War I1 during an investigation of the effect of olefins on the Fischer-Tropsch synthesis of hydrocarbon^^^^. It is also called the 0x0-synthesis, particularly in patents and other commercial literature. Hydroformylation will not proceed in the absence of a catalyst, but proceeds readily in the presence of certain group VIII metals, particularly the Co-Rh-Ir triad. Cobalt, present in the original Fischer-Tropsch process was employed for >20 yrs prior to the introduction of Rh (vide infra). Hydroformylation, as practiced in the early days with Co catalysis, presented formidable requirements of high P, containment of explosive H,, toxic CO and toxic and unstable metal carbonyls. Since it provided a direct route for converting inexpensive


14.6. Carbon Monoxide Reactions 14.6.3. Hydroformylationof Olefins

353

Phenylisocyanide, completely substitutes to Ni(CNR),I3-l6: Ni(CO),

+ 4 CNPh +4 CO + Ni(CNPh),

-

(g)

Partial substitution occurs with other isocyanides. Tertiary phosphines, arsines, and stibines also partially displace the carbonyl groups Ni(C0),'7-22: Ni(CO),

+ n PR,

Ni(CO),-.

(PR,),

+ n CO

0)

(F. CALDERAZZO)

1. H. Reihlen, A. Gruhl, G. Hessling, Ann. Chem., 472, 268 (1929). 2. I. D. Webb, G. T. Borcherdt, J . Am. Chem. SOC.,73,2654 (19510. 3. H. Behrens, F. Lohofer, Z. Naturforsch., 86, 691 (1953). 4. H. Behrens, F. Lohofer, Chem. Ber., 94, 1391 (1961). 5. H. Behrens, H. Zizlsperger, R. Rauch, Chem. Ber., 94, 1497 (1961). 6. W. Hieber, W. Kroder, E. Zahn, 2.Naturforsch., 15b, 325 (1960). 7. W. Hieber, J. Ellermann, Z. Naturforsch., 18b, 595 (1963). 8. H. W. Stemberg, R. Markby, I. Wender, J . Am. Chem. Soc., 82, 3638 (1960). 9. G. Longoni, P. Chini, A. Cavalieri, Inorg. Chem., 15, 3025 (1976). 10. G. Longoni, P. Chini, Inorg. Chem., 15, 3029 (1976). 11. W. Hieber, J. Ellermann, E. Zahn, Z. Naturforsch., 18b, 589 (1963). 12. F. Basolo, A. Wojcicki, J . Am. Chem. SOC., 83, 520 (1961). 13. F. Klages, K. Monkemeyer, Naturwissenschaflen, 37, 210 (1950). 14. W. Hieber, Z. Naturforsch., 5b, 129 (1950). 15. W. Hieber, E. Bockly, Z. Anorg. A&. Chem., 262, 344 (1950). 16. M. Bigorgne, J . Organometal. Chem., 1, 101 (1963). 17. L. S . Meriwether, M. L. Fiene, J . Am. Chem. Soc., 81, 4200 (1959). 18. J. D. Rose, F. S. Statham, J . Chem. SOC., 69 (1950). 19. G. Wilkinson, J . Am. Chem. SOC.,73, 5502 (1951). 20. A. B. Burg, W. Mahler, J . Am. Chem. SOC.,80, 2334 (1958). 21. R. J. Clark, E. 0. Brimm, Inorg. Chem., 4,651 (1965). 22. W. Reppe, W. J. Schweckendiek, Ann. Chem., 560, 104 (1948).

14.6.3. Hydroformylationof Olef ins Hydroformylation involves addition of the elements of H, and CO' added to an olefinic substrate:

The reaction was discovered in Germany prior to World War I1 during an investigation of the effect of olefins on the Fischer-Tropsch synthesis of hydrocarbon^^^^. It is also called the 0x0-synthesis, particularly in patents and other commercial literature. Hydroformylation will not proceed in the absence of a catalyst, but proceeds readily in the presence of certain group VIII metals, particularly the Co-Rh-Ir triad. Cobalt, present in the original Fischer-Tropsch process was employed for >20 yrs prior to the introduction of Rh (vide infra). Hydroformylation, as practiced in the early days with Co catalysis, presented formidable requirements of high P, containment of explosive H,, toxic CO and toxic and unstable metal carbonyls. Since it provided a direct route for converting inexpensive

354

14.6. Carbon Monoxide Reactions 14.6.3. Hydroform lation of Olefins 14.6.3.1. by Cobai Catalysts

olefins into valuable oxygenated building blocks, widespread industrial usage occurred throughout Europe, Japan and the United States. Therefore, most of the older literature consists of patents and industrial technical publications. The search for new catalysts which could effect hydroformylation under milder conditions and with higher yields of the desired aldehyde resulted in new processes which utilize Rh as the group VIII metal catalyst. Therefore, the great majority of both journal and patent literature in the last decade has been devoted to Rh catalysis. Smaller efforts have involved Ru and, more recently, Pt.

(E. BILLIG, R. L. PRUETT)

1. H. Adkins, G. Krsek, J . Am. Chem. SOC., 71,3051 (1949). 2. R. L. Pruett, Adv. Organometal. Chem., 17, 1 (1979). 3. 0. Roelen, German Patent 849,548 (1938); U. S. Patent 2,327,066 (1944); Chem. Abstr., 38, 550, (1944).

14.6.3.1. by Cobalt Catalysts

Formation of a new C-C bond by the hydroformylation reaction proceeds through an organometallic intermediate formed from cobalt hydrocarbonyl which is regenerated in the aldehyde-forming stage. The mechanism for the formation of propionaldehyde from ethylene'.' is illustrated in Scheme 1. HCo(CO), +co 1I -co

CH3CHZCHO

II

CH, =CH,

i

CH,CH,COCo(CO),

CO

Scheme 1.

The catalyst, HCo(CO),, is generated under the high pressure conditions of the reaction. The catalyst precursor added to the reactor may be Co,(CO),, which is preferable, or a soluble cobalt salt. The new bond is formed by the transformation of the alkyl cobalt tetracarbonyl to the acyl cobalt tricarbonyl. Details of this step are not known, but reaction probably proceeds either by a three-center intermediate or by direct migration of the alkyl group to one of the ligated CO groups. The mechanism in Scheme 1 explains the chemistry involved; however, the last step, the product forming hydrogenolysis has been controversial. Thus, an alternate pathway has been suggested394,which involves cleavage by cobalt hydrocarbonyl to form


354


olefins into valuable oxygenated building blocks, widespread industrial usage occurred throughout Europe, Japan and the United States. Therefore, most of the older literature consists of patents and industrial technical publications. The search for new catalysts which could effect hydroformylation under milder conditions and with higher yields of the desired aldehyde resulted in new processes which utilize Rh as the group VIII metal catalyst. Therefore, the great majority of both journal and patent literature in the last decade has been devoted to Rh catalysis. Smaller efforts have involved Ru and, more recently, Pt.

(E. BILLIG, R. L. PRUETT)

1. H. Adkins, G. Krsek, J . Am. Chem. SOC., 71,3051 (1949). 2. R. L. Pruett, Adv. Organometal. Chem., 17, 1 (1979). 3. 0. Roelen, German Patent 849,548 (1938); U. S. Patent 2,327,066 (1944); Chem. Abstr., 38, 550, (1944).


Formation of a new C-C bond by the hydroformylation reaction proceeds through an organometallic intermediate formed from cobalt hydrocarbonyl which is regenerated in the aldehyde-forming stage. The mechanism for the formation of propionaldehyde from ethylene'.' is illustrated in Scheme 1. HCo(CO),

+co 1I -co

CH3CHZCHO

II

CH, =CH,

i

CH,CH,COCo(CO),

CO

Scheme 1.

The catalyst, HCo(CO),, is generated under the high pressure conditions of the reaction. The catalyst precursor added to the reactor may be Co,(CO),, which is preferable, or a soluble cobalt salt. The new bond is formed by the transformation of the alkyl cobalt tetracarbonyl to the acyl cobalt tricarbonyl. Details of this step are not known, but reaction probably proceeds either by a three-center intermediate or by direct migration of the alkyl group to one of the ligated CO groups. The mechanism in Scheme 1 explains the chemistry involved; however, the last step, the product forming hydrogenolysis has been controversial. Thus, an alternate pathway has been suggested394,which involves cleavage by cobalt hydrocarbonyl to form

14.6. Carbon Monoxide Reactions 14.6.3. Hydroform lation of Olefins 14.6.3.1. by Cobaz Catalysts

355

aldehyde. Each step in the mechanism is thought to be reversible except the final product forming one. The reverse of this reaction is so slow that it is generally neglected5. The reaction kinetics include first-order dependence on olefin and Co. The rate increases with increasing H, and decreases with decreasing CO partial pressure6. Thus, it seems that reaction should proceed as well at low P as at high. However, the cobalt hydrocarbonyl is stable only under high P of CO and H, at the T necessary for practical rates'. Commercial processes operate at >140°C and >24 MPa. These processes require rigid precautions owing to the pressure,flammability and toxicity of the gases and the toxicity of the cobalt hydrocarbonyl. This latter compound is volatile and unstable at ambient T or above in the absence of stabilizing pressure. Cobalt metal will plate out on vessel walls after depressurizing. Cobalt can be removed from the product by extraction with aqueous acid8. A major interest for those practicing hydroformylation syntheses is the selectivity to the product desired. The factors which affect the yield of a specific aldehyde are: (1) the structure of the olefinic substrate (a-olefin or internal olefin, branching, cyclic), (2) the isomers formed during the reaction (directly, with concomitant isomerization), (3) the effects of functional groups, and (4)the subsequent reactions of the product aldehyde. Reaction rate depends on the structure of the hydrocarbon olefin (see Table 1). Internal olefins are less reactive than terminal olefins. Branching in remote locations has little effect, whereas branching at one of the olefinic carbons reduces the reactivity by another order of magnitude. In addition, the product derives from isomerized olefin. Essentially no quaternary aldehydes are formed by hydrof~rmylation~. For example, the product from 2,3-dimethyl-2-butene is 3,4-dimethyl~aleraldehyde~~. Isomers formed during the hydroformylation reaction are of tremendous importance. Much effort has been directed towards controlling the product distribution of the aldehyde mixture obtained from simple olefins. Using 1-hexene, all possible isomeric aldehydes are produced involving even originally saturated carbons' ',12. Olefin isomerization proceeds through an intermediate alkyl. Hydride shifts along the chain occur without complete elimination and readdition of metal hydride-olefin moieties. A high partial pressure of CO and low T favor linear aldehyde formation, whereas a low partial pressure of CO and higher T result in more branched aldehyde^'^. The isomer ratio obtained is mainly determined by the rate of interconversion of the n-butyryl- and iso-butyryl cobalt tetra-

TABLE1. RELATIVE RATESOF HYDROFORMYLATION OF VARIOUS OLEFIN~O Olefin

Relative Reaction Ratea,b

1 -hexene 2- hexene 4-methyl- 1-pentene 2-methyl- 1-pentene 2,3-dimethyl-2-butene cyclopentene cyclohexene

100 27.3 97.1

11.1 2.0

33.8 8.8

aAll rates are based on an arbitrary rate of 100 for I-hexene. bolefin (0.5 mole); methylcyclohexane (65 mL) as solvent; 2.8 g. CO,(CO)~;1 10°C; 23.3 MPa pressure: 50/50 CO/H,

356

14.6. Carbon Monoxide Reactions 14.6.3. H droform lation of Olefins 14.6.3.1.i y Cobai Catalysts

carbonyl intermediates which depends critically on the CO partial pressure and T. At low CO partial pressure conditions, 0.25 MPa and 11O"C, the rate of acyl isomerization is competitive with acyl reductive elimination to aldehyde. The n/iso isomer product ratio of 1.6:l reflects the equilibrium isomer ratio of the precursor butyrylcobalt tetracarbonyls. At high CO partial pressures, e.g. 9 MPa and llO"C, acyl isomerization is almost completely suppressed. The 4.4: 1 n:iso ratio reflects the reductive elimination of an acylcobalt isomer ratio mixture containing principally n-butyrylcobalt tetracarbonyl, the kinetically favored acyl',. Cobalt clusters may be invovled in hydroformylation catalysis and concomitant olefin isomerization. Thus, the 46-electron complex, HCo,(CO),, shown to be in equilibrium with HCo(CO), and Co(CO), under catalytic reaction condition^'^, is a potent olefin isomerization catalyd6. Similarly, HCo3(CO), and HCo(CO), reacts stoichi~metrically'~ with 3,3-dimethylbutene at - 15°C. to give 4,4-dimethypentanal where the hydrogen of the formyl group has come from HCo(CO), and the P-hydrogen is derived from the HCo,(CO),. Only a few well-documented trends in the effects of functional groups attached to an olefinic carbon will be mentioned here. More specific questions are discussed in reviews' Olefins substituted by aryl groups, e.g., styrene or more especially a-methylstyrene, form predminantly hydrogenated products rather than aldehydes or alcohols". There is CIDNP NMR evidence that the 0x0 reaction of styrene and related olefins proceeds via a radical mechanism". Olefins substituted with carbonyl groups likewise hydrogenate readilyz3.Butadiene is a catalyst poison, owing to the formation of an 73-allylcobalt tricarbonylZ4.Vinyl ethers attack the formyl group of the carbon a to the oxygen. Substitution at the P-position is difficult, even under forcing conditions. A dramatic example is provided by substituted Sdihydr~pyrans'~.Dihydropyran reacts with CO and H, at 180°C and 20 MPa with Co,(CO), as catalyst to give high yields of the a-hydroxymethyl derivative with C 10% substitution in the P-position:

,-'-'.

()+CO+2H2

-

(a) CH,OH

Hydroformylation of 2,6-dimethyldihydropyran,by contrast, occurs only at 220°C and then with low yields of 2,6-dimethyl-3-hydroxymethylpyran,contrary to the general rule and vinyl ether preference for a-carbon s u b s t i t u t i ~ n ~ ~ Total - ~ ~ .reactivity decreases significantly. Hydroformylation of unsaturated carbohydrates containing the dihydropyran structural unit again results in substitution at the a-carbon. Further hydrogenation to the alcohol occurs, even at 125°C. This unexplained tendency for sequential hydrogenation is marked; only modest yields of the intermediate aldehyde are obtained. Since aldehydes are reactive, subsequent reactions occur to varying extent under the conditions necessary for cobalt-catalyzed hydroformylations, e.g. 10% of the aldehyde is further hydrogenated to alcohol'8. This product forms acetals with the parent aldehyde which makes separation and purification difficult. Aldol condensations occur, particularly under basic conditions. This property has been used to form higher aldehydes and alcohols directly in a single step under hydroformylation conditionsz8,the so-called Aldox process.

-


357

Other reactions are alkane formation by hydrogenation, ketone formation (especially with ethylenez9), ester formation through hydrogen transfer and formate ester synthesis. An improved catalyst system in which one CO ligand of CoH(CO), is substituted with a trialkylphosphine ligand3', was disclosed by Shell workers in the early 1960s. With this catalyst, which is more thermally stable than the unsubstituted cobalt carbonyl, reaction proceeds at 140-190°C with 3-7 MPa of CO and H,. Additionally, mostly linear aldehydes are obtained from linear terminal and internal olefins. This remarkable result arises from the high preference for the terminal addition to an a-olefin, and the isomerization of the olefinic position which occurs simultaneously with hydroformylation. Unfortunately, the substitution of trialkylphosphine ligand has some drawbacks. The more stable trialkylphosphine substituted catalyst has a lower activity. It also has a higher tendency for hydrogenation, and as a consequence, alcohols instead of aldehydes are the more usual products. Competing hydrogenation of the substrate olefin to alkane may be significant. However, the advantages of the modified cobalt systems are important, and large industrial usage has followed, particularly for the preparation of detergent-range alcohols". There are several reports of bimetallic catalysts of Co,(CO), being used in conjunction with R U ~ ( C O )these ~ ~ ; have improved hydroformylation activity over that via either metal a l ~ n e ~ ~ . ~ ' . (E. BILLIG, R. L. PRUETT) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

D. S. Breslow, R. F. Heck, Chem. Ind. (London),467 (1960. R. F. Heck, D. S. Breslow, J . Am. Chem. Soc., 83,4023 (1961). N. H. Alemdaroglu, J. M. L. Penninger, E. Oltay, Monats. Chem., 197, 1043 (1976). W. D. Jones, R. G. Bergman, J . Am. Chem. Soc., 101,5447 (1979). C. Kim, Y. Matsui, M. Orchin, J. Organometal. Chem., 279, 159 (1985). G. Natta, R. Ercoli, S. Castellano, Chim. Ind. (Milan),37, 6 (1955). J. Berty, E. Oltay, Chem. Tech. (Leipzig),9, 283 (1957). R. Kummer, H. J. Nienberg, H. Hohenschutz, M. Strohmeyer, Adv. Chem. Ser., 132, (1974). A. J. M. Keulemans, A. Kwantes, T. V. Bard, Recl. Trav. Chim. Pays-Bas, 67, 298 (1948). R. L. Pruett, Ann. N.Y.Acad. Sci., 295, 239 (1977). F. Piancenti, M. Bianchi, P. Frediani, U. Matteoli, J . Chem. SOC., Chem. Commun., 137, 789

(1976). 12. M. Bianchi, F. Piacenti, P. Frediani, U. Matteoli, J . Organometal. Chem., 137, (1977). 13. F. Piacenti, P. Pino, R. Lazzeroni, M. Bianchi, J . Chem. SOC. (C), 488 (1966). 14. M. S. Borovikov, I. Koviics, F. UngvBry, A. Sisak, L. Mark6, Organometallics, 112, 1576 (1992). 15. G. Fachinetti, L. Balocchi, F. Sacco, M. Venturini, Angew. Chem., Int. Ed. Engl., 20, 204 (1981). 16. G. Fachinetti, A. Stefani, Angew. Chem., Int. Engl. Ed. 21, 925 (1982). 17. P. Bradamante, A. Stefani, G. Fachinetti, J . Organometal. Chem., 266, 303 (1984). 18. R. L. Pruett, Adv. Organometal. Chem., 17, 1 (1979). 19. P. Pino, F. Piacenti, M. Bianchi, Organic Synthesis via Metal Carbonyls, I. Wender, P. Pino, eds., Vol. 2, John Wiley and Sons, New York, 1977. 20. J. Falbe, New Syntheses With Carbon Monoxide, Springer Verlag, Berlin, 1980. 21. D. M. Rudkovskii, N. S. Imyanitov, Zh. Prikl. Khim, 2719 (1962). 22. T. M. Bockman, J. F. Garst, R. B. King, L. Marko, F. Ungvary, J . Organometal. Chem., 279, 165 (1985). 23. R. W. Goetz, M. Orchin, J . Am. Chem. SOC.,85, 2782 (1963). 24. D. W. Moore, H. B. Jonassen, T. B. Joyner, Chem. Ind. (London), 1304 (1960). 25. J. Falbe, F. Corte, Chem. Ber., 97, 1104 (1964). 26. A. Rosenthal, D. Abson, Can. J . Chem., 42, 1811 (1964). 27. A. Rosenthal, D. Abson, T. D. Field, H. J. Koch, R. A. J. Mitchell, Can. J . Chem., 45, 1525 (1967).

14.6. Carbon Monoxide Reactions 14.6.3. H droformylation of Olefins 14.6.3.2. &y Rhodium Catalysts 28. 29. 30. 31. 32.

C. R. Greene, U S . Pat. 3,278,612, (1966); Chem. Absrr., 6746 (1960). E. Naragon, A. Millendorf, J. Vergilio, U.S. Pat. 2,699,453 (1955); Chem. Absrr. 1893, (1956). L. H. Slaugh, R. D. Mullineaux, J . Organometal. Chem., 13, 469, (1968). M. Hidai, A. Fukuoka,Y. Koyasu, Y. Uchida, J . Mol. Caral., 35, 29, (1986). Y.Ishii, M. Sato, H. Matsuzaka, M. Hidai, J . Mol. Caral., 54, L13, (1989).


Catalysis of hydroformylation by soluble rhodium complexes, particularly with phosphine ligands, provides a selective and mild route to aldehydes. Extensive reviews compare catalysis to the older cobalt systems'-3. Non-activated carbon-carbon double bonds react catalytically and rapidly with H, and CO at -80°C and a few atmospheres pressure4 ( < I mPa). The milder conditions are attractive for the synthesis of aldehydes without high pressure equipment and for minimizing competitive but undesirable reactions. Among the latter are partial hydrogenation to alcohols which form acetals, aldol condensation products, ketone formation and hydrogenation to alkane. Hydroformylation proceeds in the absence of solvents, but better rates are obtained in polar solvents5. Triphenylphosphine ligands increase the rate6 to a P/Rh of 14. At high ligand/Rh ratios, the rate declines, but the selectivity to the linear aldehyde increases497.Thus, the catalyst choice depends on whether the objective is high rate or isomeric selectivity. Several soluble rhodium species are generally acceptable, and best results under mild conditions are obtained with non-halide precursors6, e.g. HRh(CO)(PPh,),. The substrate olefin structure has a marked effect on the reaction rate with a HRh(CO)(PPh,), catalyst8 (see Table 2). a,P-Unsaturated aldehydes are unreactive due to the formation of a stable complex. However, this lack of reactivity can be overcome by converting the aldehyde group to an acetal. The commercially important olefin, acrolein, can be hydroformylated easily as its cyclic acetal. Hydrolysis and hydrogenation following the Rh catalyzed hydroformylation, produces butanediol and regenerates the diol used for acetal formation (Scheme 1)9,10. TABLE2. RELATIVE RATESOF HYDROFORMYLATION OF VARIOUS OLEFINS

Olefin

Relative Reaction Rate

ally1 alcohol

2.01 1.23 1.15 1.oo 0.21 0.07 0.03

styrene o-allylphenol 1-heptene vinyl acetate

cyclooctene cis-2-heptene 2-methyl-1-pentene

0.02 ~

~~

~

Tatalyst, HRh(CO)(PPh,),, 2.5 mmole; solvent, benzene: olefin concentration, 1.0 M, 25°C.; 1/1 CO/H, at 500 m m pressure bRates are relative to 1-heptane


14.6. Carbon Monoxide Reactions 14.6.3. H droformylation of Olefins 14.6.3.2. &y Rhodium Catalysts 28. 29. 30. 31. 32.

C. R. Greene, U S . Pat. 3,278,612, (1966); Chem. Absrr., 6746 (1960). E. Naragon, A. Millendorf, J. Vergilio, U.S. Pat. 2,699,453 (1955); Chem. Absrr. 1893, (1956). L. H. Slaugh, R. D. Mullineaux, J . Organometal. Chem., 13, 469, (1968). M. Hidai, A. Fukuoka,Y. Koyasu, Y. Uchida, J . Mol. Caral., 35, 29, (1986). Y.Ishii, M. Sato, H. Matsuzaka, M. Hidai, J . Mol. Caral., 54, L13, (1989).


Catalysis of hydroformylation by soluble rhodium complexes, particularly with phosphine ligands, provides a selective and mild route to aldehydes. Extensive reviews compare catalysis to the older cobalt systems'-3. Non-activated carbon-carbon double bonds react catalytically and rapidly with H, and CO at -80°C and a few atmospheres pressure4 ( < I mPa). The milder conditions are attractive for the synthesis of aldehydes without high pressure equipment and for minimizing competitive but undesirable reactions. Among the latter are partial hydrogenation to alcohols which form acetals, aldol condensation products, ketone formation and hydrogenation to alkane. Hydroformylation proceeds in the absence of solvents, but better rates are obtained in polar solvents5. Triphenylphosphine ligands increase the rate6 to a P/Rh of 14. At high ligand/Rh ratios, the rate declines, but the selectivity to the linear aldehyde increases497.Thus, the catalyst choice depends on whether the objective is high rate or isomeric selectivity. Several soluble rhodium species are generally acceptable, and best results under mild conditions are obtained with non-halide precursors6, e.g. HRh(CO)(PPh,),. The substrate olefin structure has a marked effect on the reaction rate with a HRh(CO)(PPh,), catalyst8 (see Table 2). a,P-Unsaturated aldehydes are unreactive due to the formation of a stable complex. However, this lack of reactivity can be overcome by converting the aldehyde group to an acetal. The commercially important olefin, acrolein, can be hydroformylated easily as its cyclic acetal. Hydrolysis and hydrogenation following the Rh catalyzed hydroformylation, produces butanediol and regenerates the diol used for acetal formation (Scheme 1)9,10. TABLE2. RELATIVE RATESOF HYDROFORMYLATION OF VARIOUS OLEFINS

Olefin

Relative Reaction Rate

ally1 alcohol

2.01 1.23 1.15 1.oo 0.21 0.07 0.03

styrene o-allylphenol 1-heptene vinyl acetate

cyclooctene cis-2-heptene 2-methyl-1-pentene

0.02 ~

~~

~

Tatalyst, HRh(CO)(PPh,),, 2.5 mmole; solvent, benzene: olefin concentration, 1.0 M, 25°C.; 1/1 CO/H, at 500 m m pressure bRates are relative to 1-heptane

14.6. Carbon Monoxide Reactions 14.6.3. H droformylation of Olefins 14.6.3.2. L y Rhodium Catalysts

359

Scheme 1. Butanediol has also been produced by hydroformylation of allyl alcohol' ',I2, a very reactive olefin, which, unless the conditions are mild, undergoes isomerization and other undesirable side reactions. The sensitivity of allyl alcohol as a substrate is overcome by converting it initially to allyl t-butyl ether. After hydroformylation and hydrogenation, the t-butyl group is removed by acid cleavage to give butanediol and t-butanol which is re~ycled'~. Aryl olefins and a$-unsaturated esters and ketones, which are hydrogenated with standard Co hydroformylation catalysts, may be hydroformylated in high yields with Rh as catalyst14. The examples above, demonstrate that a great variety of functionalized olefins can be hydrofonnylated with Rh, but each substrate must be approached individually when selecting a proper T, P and ligand. Product recovery is simplified over that with Co catalysts owing to the high selectivity to aldehyde (low amounts of complicating byproducts) and to the stability and the low vapor P of the catalyst and triphenylphosphine ligand. In most cases, the product aldehyde can be distilled away from residual catalyst and ligand. In a typical example, acetophenone is used as a ~ o l v e n t ~ .Soluble '~. Rh complex is added in an amount sufficient to produce a final concentration of 100-500 ppm, calculated as the metal. Excess phosphine is added, together with the olefinic substrate. The reaction is conducted at 70-120°C and about 1 MPa pressure of synthesis gas with a composition of either 1/1 or 3/1 H,/CO. Preferentially, the pressure is held constant by continuous introduction of the stoichiometric 1/ 1 HJCO until absorption ceases. Gas-phase hydroformylations over heterogeneous catalysts have been studied for ethylene and propylene. Various catalysts have been examined. These include ligand anchored catalyst, where the ligand is a diphenylphosphino-substituteddivinyl benzene cross-linked polystyrene. A hybrid liquid/heterogeneous catalyst system consisting of the catalyst dissolved in a supported liquid phase (SLPC)'6*'7gives high reaction rates, an especially attractive feature usually associated with homogeneous systems. Product recovery is substantially simpler with heterogenized Rh catalysts.

360

14.6. Carbon Monoxide Reactions 14.6.3. Hydroformylation of Olefins 14.6.3.2. by Rhodium Catalysts ~

~~~

The synthesis of asymmetric products from prochiral olefins using chiral phosphorus ligands is extremely useful for hydrogenations”. Similar advances in hydroformylation have been much slower, because of the ease of racemization of the various aldehyde products. For example, the most common ligand employed in asymmetric hydrogenation, DIOP [2,3-O-isopropylidene-2,3-dihydroxy-1,4-bis(diphenylphosphino)-butane] which gives an 81% optical yield in the hydrogenation of a-acetamido cinnamic acid18, yields hydratropaldehyde from the hydroformylation of styrene” in only 25% ee (ee = enantiomeric excess). Recent progress has been madez0-22,particularly with Pt catalysts (see below). Binuclear forms and higher aggregates of Rh have been implicated in hydroformylation catalysis. While it is difficult to totally exclude the possibility that mononuclear fragments formed from polynuclear precursors under hydroformylation conditions are the true catalysts in any given instance, there is increasing evidence suggesting binuclear and higher forms of rhodium do in fact participate catalytically in the 0x0 reaction. A kinetic study of the hydroformylation of cyclohexene with Rh,(CO),, as catalyst precursor at 125°C and 3.4 MPa gave a rate expressionz3:

consistent with the view that no fragmentation of the Rh,(CO),, to lower nuclearity or mononuclear species occurs. A binuclear complex, Rh,(nbd),(eLTTP)’ (nbd = norbomadiene, eLTTP = (Et,PCH,CH,)(Ph)PCH,P(Ph)(CH,CH,PEt,), showed remarkably high reactivity and regioselectivity with a-olefins. The binuclear integrity of this catalyst precursor is forced by the binucleating tetraphosphine ligand. The higher reactivity and normal to branched aldehyde product ratios evidenced by this catalyst system over that obtained with conventional mononuclear catalysts may result from homobimetallic c~operativity’~. In contrast, a wide variety of di- and tritertiary phosphines, e.g. Ph,P(CH,),PPh,, CH,C(CH,PPh,),, enhance rate and selectivity when employed at less than molar equivalent concentrations to Rh. However, at equivalent or higher ligand concentrations, rate is strongly suppressed and eventually quenchedz5. 2,2’-Bis((diphenylphosphino)methyl)- 1,l ‘-biphenyl (BISBI), which forms a rather unusual nine-membered chelate complex with Rh at ligand/Rh ratios greater than one, is less active but gave n-/ibutyraldehyde ratios significantly greater than a Rh triphenylphosphine controlz6. Thus, at 125°C and 1.9 MPa, a BISBI catalyst 1.4mM in Rh and with a 2.8/1 ligand/Rh ratio gave a rate of 5 gmols/liter hour and a n/i-butyraldehyde ratio of 25/1. In contrast, the Rh triphenylphosphine catalyst which was 0.65 mM in Rh, and with a ligand to Rh ratio of 124/1, gave a rate of 8.3 gmoles/liter/hour and an isomer ratio of 2.4/1 under the same T and P conditions. Bis phosphites are unique hydroformylation ligands because they give high rates and high n-/branched isomer ratios with both terminal and internal olefinic substrates. These ligands are effective at ligand to Rh ratios greater than unity. For example, a catalyst containing bis-phosphite 1, 1.2 mM in Rh and with a 4/1 ratio of 1 to Rh, at 70”C, with a butene-1 substrate, gave a rate of 3 gmoles/liter/hour and a 68/1 n- to branched C5 aldehyde ratio.” With trans-butene-2 at lOO”C, the catalyst containing 1, 3.8 mM n Rh and a 4/1 ligand 1 to Rh ratio, gave a rate of 1.5 gmoles liter/hour and +

14.6. Carbon Monoxide Reactions 14.6.3. Hydroformylationof Olefins 14.6.3.2. by Rhodium Catalysts

361

an n-/branched C5 aldehyde ratio of 16/lZ7.Various extremely reactive, highly regioselective bis-phosphite promoted catalysts have been described. For example, bisphosphite 228yields a catalyst giving a 53/1 n- to i-butraldehyde ratio at 74°C.

[CgH19

*o$P\ot&

OMe /O

/ 0

Me0

I:$

g > p \ ; B U

/ 0

t-Bu

OMe t-Bu

t-Bu

1

M

e

O

g

\ Me0

0'"0

&-

t-Bu

OMe

Bu-t

%o\p/o

2 3 The employment of bulky phosphite ligand~'~-~', e.g., tri(o-t-butylpheny1)phosphite, yields high reactivity catalysts for the hydrofomylation of very low reactivity P-alkyl-a-olefins, e.g. isoprene. Diorganoph~sphites~' such as 3 overcome many of the problems associated with the employment of simple triarylphosphites under hydroformylation reaction conditions such as reaction of the phosphite with the product aldehyde, trans-esterification and organometallic degradation reactions. (E. BILLIG, R. L. PRUETT) 1. R. L. Pruett, Adv. Organometal. Chem., 17, 1 (1979). 2. P. Pino, F. Piacenti, M. Bianchi, in Organic Synthesis Via Metal Carbonyls, Vol. 2., I. Wender, P. Pino, eds., John Wiley and Sons, New York, 1977, p. 43. 3. R. L. Pruett, J. Chem. Ed., 63, 196 (1986). 4. R. L Pruett, J. A. Smith, J . Org. Chem., 34, 327 (1969). 5. J. H. Craddock, A. Hershman, F. E. Paulik, J. F. Roth, Ind. Eng. Chem., Prod. Res. Dev., 8, (1969). 6. K. L. Olivier, F. B. Booth, Hydrocarbon Process., 49, 112 (1969). 7. R. L. Pruett, J. A. Smith, U.S. Patent 3,527,809 (1970); Chem. Abstr. 90819, (1969). 8 . C. K. Brown, G. Wilkinson, J . Chem. SOC. (A), 1392 (1970). 9. C. C. Cumbo, K. K. Bhatia, U.S. Patent 3,929,809, (1975); Chem Absrr., 85, 121185 (1976). 10. C. C. Cumbo, K. K. Bhatia, U.S. Patent 3,963,755, (1976); Chem Abstr., 85, 1601 10 (1976). 1 1 . W. E. Smith, U.S. Patent 4,224,255 (1980).

14.6. Carbon Monoxide Reactions 14.6.3. Hydroformylationof Olefins 14.6.3.3. by Ruthenium Catalysts 12. C. U. Pittman, Jr., A. Hirao, C. Jones, R. M. Hanes, Q.Ng, Ann. N.Y. Acad. Sci., 241, 15 (1977). 13. Davey-McKee, European Pat. Appln. 18, 163, (1980). 14. J. Falbe, New Syntheses With Carbon Monoxide (Springer Verlag, Berlin, 1980). 15. C. U. Pittman, Jr., W. D. Honnick, J . Org. Chem., 45, 2132 (1980). 16. P. R. Rony, J . Catal., 14, 142 (1969). 17. L. A. Genitsen, W. Klut, M. H. Vreugdenhil, J . Mol. Catal., 9 , 265 (1980). 18. B. R. James, Adv. Organometal. Chem., 17, 319 (1979). 19. P. Pino, G. Consiglio, in Fundamental Research in Homogeneous Catalysis, Vol. 3, M. Tsutsui ed., Plenum Press, New York, NY, 1979, p. 519. 20. R. Noroyi, M. Kitamura, Mod. Synth. Methods, 5 , 115, (1989). 21. H. Brunner, Synthesis, 645, (1988). 22. H. B. H-Kagan, Bull SOC. Chim. Fr., 846, (1988). 23. N. Rosas, C. Marquez, H. Hernandez, R. Gomez, J . Mol. Catal., 48,59, (1988). 24. M. Broussard, S. Train, W. Peng, B. Juma, S . Laneman, G. G. Stanley, Abstracts 204th ACS National Meeting, Washington, D.C., Aug. 23-28, 1992, INOR 0047. 25. A. R. Sanger, J. Mol. Catal., 3,221 (1977/78). 26. T. J. Devon, G . W. Phillips, T. A. Puckette, J. L. Stavinohs, J. J. Vanderbilt, U.S. Pat. 4,694,109, (1989). 27. E. Billig, A. G . Abatjoglou, D. R. Bryant, U.S. Pat., 4,748,261, (1988). 28. E. Billig, A. G . Abatjoglou, D. R. Bryant, U.S. Pat., 4,769,498, (1988). 29. P. W. M. N. van Leeuwen, C. F. Roobeek, J . Organomet. Chem., 258,4343 (1982). 30. T. Jongsma, G . Challa, P. W. N. M. van Leeuwen, J . Organomet. Chem., 421, 121 (1991). 31. E. Billig, A. G . Abatjoglou, D. R. Bryant, R. E. Murray, J. M. Maher, U.S. Pat., 4,737,588, (1988).

14.6.3.3. by Ruthenium Catalysts Ruthenium has long been discussed as an effective metal for hydroformylation', but its activity is lower than that of Co and much less than Rh.

Rh

Co

Ru

Mn

103-104

1

10-2

10-4

Fe 10-6

Cr,Mo,W -0

For monomeric Ru carbonyl triphenylphosphine species as catalyst', optimum conditions are 120°C, 10 MPa of CO and H, with benzene as solvent. For 1-hexene, 86% conversion results after 20 hr (2,000 ppm Ru) with 99% selectivity to aldehydes. The ratio of linear to branched aldehydes is 2.4:l. At ?150"C, reduced conversions occur, owing to formation of inactive Ru complexes. High H, partial pressures increase the reaction rate, but hydrogenation to alkane results. Excess triphenylphosphine gives improved selectivity for linear aldehyde, but at the expense of a drastic decrease in rate. In what may be another example of true cluster catalysis, [HRu,(CO),,]-, shows CO catalytic activity for the hydroformylation of ethylene and p r ~ p y l e n e ~ -Solvent, ~. partial P, and T are important variables. In monoglyme, at 80"C, and starting partial of C,H6, C O and H, of 0.034, 0.022 and 0.011 MPa, respectively, the catalyst turnover number (mols product/mols catalyst) was 34.3 and the n- to i-butyraldehyde ratio was 49.4/1. In acetonitrile solvent, all other things being equal, the turnover number dropped to 25.7 and the isomer ratio decreased to 12.1/1. (E. BILLIG, R. L. PRUETT) 1. 2. 3. 4. 5.

D. Evans, J. A. Osborn, F. H. Jardine, G. Wilkinson, Nature, 203, 1203 (1965). R. A. Sanchez-Delgado,J. S . Bradley, G . Wilkinson, J . Chem. SOC.,Dalton Trans., 399 (1976). G. Suess-Fink, G. F. Schmidt, J . Mol. Catal., 4 2 , 361-6 (1987). G. Suess-Fink, J . Organometal. Chem., 193, C20 (1980). G. Suess-Fink, G. Herrmann, J . Chem. SOC. Chem. Commun., 735 (1985).


14.6. Carbon Monoxide Reactions 14.6.3. Hydroformylationof Olefins 14.6.3.3. by Ruthenium Catalysts 12. C. U. Pittman, Jr., A. Hirao, C. Jones, R. M. Hanes, Q.Ng, Ann. N.Y. Acad. Sci., 241, 15 (1977). 13. Davey-McKee, European Pat. Appln. 18, 163, (1980). 14. J. Falbe, New Syntheses With Carbon Monoxide (Springer Verlag, Berlin, 1980). 15. C. U. Pittman, Jr., W. D. Honnick, J . Org. Chem., 45, 2132 (1980). 16. P. R. Rony, J . Catal., 14, 142 (1969). 17. L. A. Genitsen, W. Klut, M. H. Vreugdenhil, J . Mol. Catal., 9 , 265 (1980). 18. B. R. James, Adv. Organometal. Chem., 17, 319 (1979). 19. P. Pino, G. Consiglio, in Fundamental Research in Homogeneous Catalysis, Vol. 3, M. Tsutsui ed., Plenum Press, New York, NY, 1979, p. 519. 20. R. Noroyi, M. Kitamura, Mod. Synth. Methods, 5 , 115, (1989). 21. H. Brunner, Synthesis, 645, (1988). 22. H. B. H-Kagan, Bull SOC. Chim. Fr., 846, (1988). 23. N. Rosas, C. Marquez, H. Hernandez, R. Gomez, J . Mol. Catal., 48,59, (1988). 24. M. Broussard, S. Train, W. Peng, B. Juma, S . Laneman, G. G. Stanley, Abstracts 204th ACS National Meeting, Washington, D.C., Aug. 23-28, 1992, INOR 0047. 25. A. R. Sanger, J. Mol. Catal., 3,221 (1977/78). 26. T. J. Devon, G . W. Phillips, T. A. Puckette, J. L. Stavinohs, J. J. Vanderbilt, U.S. Pat. 4,694,109, (1989). 27. E. Billig, A. G . Abatjoglou, D. R. Bryant, U.S. Pat., 4,748,261, (1988). 28. E. Billig, A. G . Abatjoglou, D. R. Bryant, U.S. Pat., 4,769,498, (1988). 29. P. W. M. N. van Leeuwen, C. F. Roobeek, J . Organomet. Chem., 258,4343 (1982). 30. T. Jongsma, G . Challa, P. W. N. M. van Leeuwen, J . Organomet. Chem., 421, 121 (1991). 31. E. Billig, A. G . Abatjoglou, D. R. Bryant, R. E. Murray, J. M. Maher, U.S. Pat., 4,737,588, (1988).

14.6.3.3. by Ruthenium Catalysts Ruthenium has long been discussed as an effective metal for hydroformylation', but its activity is lower than that of Co and much less than Rh.

Rh

Co

Ru

Mn

103-104

1

10-2

10-4

Fe 10-6

Cr,Mo,W -0

For monomeric Ru carbonyl triphenylphosphine species as catalyst', optimum conditions are 120°C, 10 MPa of CO and H, with benzene as solvent. For 1-hexene, 86% conversion results after 20 hr (2,000 ppm Ru) with 99% selectivity to aldehydes. The ratio of linear to branched aldehydes is 2.4:l. At ?150"C, reduced conversions occur, owing to formation of inactive Ru complexes. High H, partial pressures increase the reaction rate, but hydrogenation to alkane results. Excess triphenylphosphine gives improved selectivity for linear aldehyde, but at the expense of a drastic decrease in rate. In what may be another example of true cluster catalysis, [HRu,(CO),,]-, shows CO catalytic activity for the hydroformylation of ethylene and p r ~ p y l e n e ~ -Solvent, ~. partial P, and T are important variables. In monoglyme, at 80"C, and starting partial of C,H6, C O and H, of 0.034, 0.022 and 0.011 MPa, respectively, the catalyst turnover number (mols product/mols catalyst) was 34.3 and the n- to i-butyraldehyde ratio was 49.4/1. In acetonitrile solvent, all other things being equal, the turnover number dropped to 25.7 and the isomer ratio decreased to 12.1/1. (E. BILLIG, R. L. PRUETT) 1. 2. 3. 4. 5.

D. Evans, J. A. Osborn, F. H. Jardine, G. Wilkinson, Nature, 203, 1203 (1965). R. A. Sanchez-Delgado,J. S . Bradley, G . Wilkinson, J . Chem. SOC.,Dalton Trans., 399 (1976). G. Suess-Fink, G. F. Schmidt, J . Mol. Catal., 4 2 , 361-6 (1987). G. Suess-Fink, J . Organometal. Chem., 193, C20 (1980). G. Suess-Fink, G. Herrmann, J . Chem. SOC. Chem. Commun., 735 (1985).


14.6. Carbon Monoxide Reactions 14.6.3. H droformylation of Olefins 14.6.3.4. i y Platinum Catalysts

363

14.6.3.4. by Platinum Catalysts

Platinum catalysis of hydroformylation has seen remarkable progress in the last twenty years. The metal has extremely low or nonexistent activity. A phosphine complex, [Ph3P)3PtH]+PF6-, is a low activity catalyst at 150°C and 10 MPa of H,/CO’. A phosphine platinum(I1) iodide, (Ph,P),PtI,, is somewhat bette?, giving 89% conversion of olefin to aldehydes at 180OC and 25 MPa pressure in 1 hr in dimethylformamide solvent. Selectivity and activity is significantly increased with a Pt catalyst which utilizes both phosphine and Sn(I1) halide ligands. High specificity to the linear aldehyde from 1pentene or 1-heptene is realized from HPtSnC13(CO)(PPh3)3which is active at 100°C and 20 MPa and produces 95% n-hexanal from 1-pentene. The Sn:P:Pt ratio is significant as well as solvent polarity”. Intermediate polarity solvents, such as acetophenone and methyl isobutyl ketone, which solvate catalyst intermediates but do not form stable and unreactive discrete complexes, are best. The catalyst ligand ratio of 5Sn:2P:lPt is optimum with methyl isobutyl ketone as solvent and 1-heptene as substrate. Reaction occurs at 66°C and 10 MPa of H, and CO to give 85% yield of aldehyde, of which 90% was linear. The yield of hydrogenated product, heptane, and isomerized product, internal heptenes, is 2.7 and 3.6%, respectively. A further improvement in Pt catalysis5 is claimed from Sn(I1) halide and phosphine ligands using rigid bidentate ligands, e.g. 1,2-bis(diphenylphosphinomethyl)cyclobutane. High rates are disclosed with a product containing 99% linear aldehyde. However, a pressure of 10 MPa CO/H, is still required. There has been considerable academic and industrial effort to produce chiral, nonsteroidal anti-inflammatory agents such as Naproxenm and Ibuprofen@ via the stereoselective hydrofonnylation of vinyl aromatics. Such a process requires not only a high enatiomeric excess (i.e. optical yield) of the branched isomer but also a high branched to normal (b/n) regioselectivity:

Ar-CH=CH,+CO+H,

cat.* -+

CHO

Ar branched

(a)

normal

High optical yields (-70-88% ee) have been reported using PtCl(SnCl,)BPPM, where BPPM is the chiral ligand (2S,4S)-N-(tert-butoxycarbonyl)-4-(diphenylphosphino)-2-[(diphenylphosphino)-methyl]pyrollidine6. Even higher enantiomeric excesses have been obtained (>96%) by trapping the aldehyde with triethylorthoformate. However, rates and b/n ratios with these catalysts are low (e.g. 0.5-3.3:l). The dibenzophospholyl analogs’ of BPPM gave the highest b/n ratios, 3.3:1, of any of these ligands and were examined with a variety of aryl vinyl substrates. Generally, optical yields were low because of in situ racemization while the b/n ratios depended on the aromatic vinyl substituent. When the hydrofonnylation was conducted in the presence of triethylorthoformate, enantiomerically pure acetals were obtained. The BPPM catalysts were only effective for the hydroformylation of vinyl aromatic compounds and were ineffective with prochiral ally1 compounds and enol ethers. Molecular modelling studies of styrene hydroformylation’ employing an augmented Dreiding force field, suggest the importance of steric factors in favoring branched over normal regioselectivity with L,Pt(CO)X catalysts. These studies indicate a pronounced

14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarboxylation of Olefins ~

~

~

~

_

_

p-stacking interaction between the a-methylstyryl intermediate (leading to branched product) and an aromatic ring of the phosphine ligand which is absent in the phenethyl intermediate leading to the straight chain product, e.g. when pentene is substituted for styrene where such p-interactions are precluded, normal over branched regioselectivity is favored. High enantioselectivity has been achieved in the hydroformylation of a vinylidene ester, dimethyl itaconate, employing PtCl(SnCl,)[R,R')-DIOP] as catalystg. (R)(MeO,C)CH,CH(CO,Me)CH~CHOwas obtained in >82% optical yield. Under the conditions employed, a competitive hydrogenation occurred to give dimethyl methylsuccinate in 51% enantiomeric excess. Asymmetric induction in both hydroformylation and hydrogenation was lower with other vinylidene esters.

(E.BILLIG, R. L. PRUETT)

1. J. J. Mrowca, U.S.Patent 3,876,672 (1975); Chem. Abstr., 83, 30432, (1975). 2. G. A. Rowe, British Patent 1,368,434 (1974); Chem. Abstr. 142590, (19750. 3. C. Hsu, M. Orchin, J . Am. Chem. SOC.,97, (1975). 4. I. Schwager, J. F. Knifton, J . Catal., 45, 256 (1976). 5 . I. Ogata, Y. Kawabata, M. Tanaka, T. Hayashi, U.S.Pat. 4,229,381, (1980). 6. G. Parrinnello, J. K. Stille, J . Am. Chem. SOC.,109, 7122 (1987). 7. J. K. Stille, H. Su, G . Parinello, L. S . Hegedus, Organornetallics, 10, 1183 (1991). 8. L. A. Castonguay, A. K. Rappt, C. J. Casewit, J. Am. Chem. SOC., 113, 7117 (1991). 9. L. Kollar, G . Consiglio, P. Pino, J . Organometal. Chem., 330, 305 (1987).

14.6.4. Hydrocarboxylationof Olefins Hydrocarboxylation refers to the formation of carboxylic acids by addition of the elements of CO and H 2 0 to an olefin: RCH=CH,

+ CO + H,O

RCH2CH2C0,H

+ RCHC0,H

(4

CH3 As with hydroformylation, a mixture of isomeric acids is usually formed. Much effort has been directed toward obtaining regioselective processes for both the linear and the branched products. Carbonylation of olefins in the presence of alcohols to give esters is called hydroesterification. Similarly, olefin carbonylation in the presence of carboxylic acids yields acid anhydrides. Both hydroesterification and acid anhydride formation by olefin carbonylation are covered in section 14.6.4. Other carbonylation variations, including the use of acetylenic substrates, thiols and amines as hydrogen sources and the carbonylation of allylic halides are not discussed. Several excellent reviews of hydrocarboxylation and carbonylation of olefins'*2*2b have appeared. Early hydrocarboxylation-hydroesterificationliterature deal^^-^ largely with Ni and Co as activating metals, but during the last three decades the noble group VIII metals, especially Pd, Pt, Rh and Ir, have been studied. Similarly, the use of pyridine promoted Co catalysts has been optimized. This section will not include references to metals of lesser or more specialized activity, such as Fe, Ru and Cu(I), nor strong acid catalysis, nor oxidative carbonylation of alkanes. The mechanism of hydrocarboxylation is thought to proceed through initial metalhydride bond formation. For example, this may occur by the reaction of HX (X =

_

_

_


14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarboxylation of Olefins ~

~

~

~

_

_

p-stacking interaction between the a-methylstyryl intermediate (leading to branched product) and an aromatic ring of the phosphine ligand which is absent in the phenethyl intermediate leading to the straight chain product, e.g. when pentene is substituted for styrene where such p-interactions are precluded, normal over branched regioselectivity is favored. High enantioselectivity has been achieved in the hydroformylation of a vinylidene ester, dimethyl itaconate, employing PtCl(SnCl,)[R,R')-DIOP] as catalystg. (R)(MeO,C)CH,CH(CO,Me)CH~CHOwas obtained in >82% optical yield. Under the conditions employed, a competitive hydrogenation occurred to give dimethyl methylsuccinate in 51% enantiomeric excess. Asymmetric induction in both hydroformylation and hydrogenation was lower with other vinylidene esters.

(E.BILLIG, R. L. PRUETT)

1. J. J. Mrowca, U.S.Patent 3,876,672 (1975); Chem. Abstr., 83, 30432, (1975). 2. G. A. Rowe, British Patent 1,368,434 (1974); Chem. Abstr. 142590, (19750. 3. C. Hsu, M. Orchin, J . Am. Chem. SOC.,97, (1975). 4. I. Schwager, J. F. Knifton, J . Catal., 45, 256 (1976). 5 . I. Ogata, Y. Kawabata, M. Tanaka, T. Hayashi, U.S.Pat. 4,229,381, (1980). 6. G. Parrinnello, J. K. Stille, J . Am. Chem. SOC.,109, 7122 (1987). 7. J. K. Stille, H. Su, G . Parinello, L. S . Hegedus, Organornetallics, 10, 1183 (1991). 8. L. A. Castonguay, A. K. Rappt, C. J. Casewit, J. Am. Chem. SOC., 113, 7117 (1991). 9. L. Kollar, G . Consiglio, P. Pino, J . Organometal. Chem., 330, 305 (1987).

14.6.4. Hydrocarboxylationof Olefins Hydrocarboxylation refers to the formation of carboxylic acids by addition of the elements of CO and H 2 0 to an olefin: RCH=CH,

+ CO + H,O

RCH2CH2C0,H

+ RCHC0,H

(4

CH3 As with hydroformylation, a mixture of isomeric acids is usually formed. Much effort has been directed toward obtaining regioselective processes for both the linear and the branched products. Carbonylation of olefins in the presence of alcohols to give esters is called hydroesterification. Similarly, olefin carbonylation in the presence of carboxylic acids yields acid anhydrides. Both hydroesterification and acid anhydride formation by olefin carbonylation are covered in section 14.6.4. Other carbonylation variations, including the use of acetylenic substrates, thiols and amines as hydrogen sources and the carbonylation of allylic halides are not discussed. Several excellent reviews of hydrocarboxylation and carbonylation of olefins'*2*2b have appeared. Early hydrocarboxylation-hydroesterificationliterature deal^^-^ largely with Ni and Co as activating metals, but during the last three decades the noble group VIII metals, especially Pd, Pt, Rh and Ir, have been studied. Similarly, the use of pyridine promoted Co catalysts has been optimized. This section will not include references to metals of lesser or more specialized activity, such as Fe, Ru and Cu(I), nor strong acid catalysis, nor oxidative carbonylation of alkanes. The mechanism of hydrocarboxylation is thought to proceed through initial metalhydride bond formation. For example, this may occur by the reaction of HX (X =

_

_

_

14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarbox lation of Olefins 14.6.4.1. by Cobalt atalysts

365

z

halogen) with Ni(C0)46, or HI with a soluble Rh species', or by H,O with Co,(CO),. The complete reaction sequence with a Ni catalyst is (linear isomer only) shown in Scheme 1.

H20,2 CO

I

>

RCH,CH,CONi(CO),X

co

HNi(CO),X RCH= CH,

RCH,CH,Ni(CO),X Scheme 1. (E. BILLIG, R. L. PRUETT)

1. A. Mullen, New Syntheses with Carbon Monoxide, J. Falbe, ed., Springer Verlag, Berlin, 1980, p. 257. This book and the following referenced chapter provide excellent and current reviews of the topic. 2. (a) P. Pino, F. Piacenti, M. Bianchi, Organic Syntheses Via Metal Carbonyls, I. Wender, P. Pino, eds., Vol. 2, John Wiley & Sons, New York, N.Y., 1977, p. 43; (b) D. P. Riley, Mech. Inorg. Organometal. React., 5, 335 (1988). 3. W. Reppe, H. Kraper, German Patent 863,194 (1953); Chem. Abstr., 48, 1425b (1953). 4. W. Reppe, H. KriSper, German Patent 862,748 (1953); Chem. Abstr., 48, 10059h (1953). 5. G. Natta, P. Pino, Chim. Ind. (Milan),31, 109 (1949). 6. R. F. Heck,J.Am. Chem. SOC.,85, 2013 (1963). 7. D. E. Moms, G. V. Johnson, Sympos. Rhodium Homogen. Catal., Veszprem, Hungary (1978).


Cobalt is the catalyst most widely employed for hydrocarboxylations and hydroesterifications. The strong similarity to hydrofonnylation is shown in the catalyst and in the conditions of T and P. In general the three process rates are in the order: hydroformylation>hydrocarboxylation>hydroesterification. Like hydrofonnylation, the high P of CO required, the instability and toxicity of cobalt carbonyl or hydrocarbonyl and the difficulty of catalyst-product separation detract from the overall attractiveness of this reaction'. Hydrogen donors for hydroesterification may be primary or secondary alcohols, cyclohexanol, phenol or polyols. Linear or branched primary alcohols react similarly; secondary alcohols are less active and tertiary alcohols are not suitable'. The reactivity of various olefins has been compared3. As with iodide promoted Ni catalysts4, iodide promoted Co catalysts convert ethylene, H,O and CO to propionic acid in high efficiency'. The use of CO containing 3% H, in the hydroesterification reaction is standard', suggesting that a cobalt hydrocarbonyl is the active catalyst species. The reaction sequence involves olefin insertion into the Co-H bond, (carbonyl insertion) to give an acyl complex and cleavage with alcohol assisted by the pyridine promoter?


14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarbox lation of Olefins 14.6.4.1. by Cobalt atalysts

365

z

halogen) with Ni(C0)46, or HI with a soluble Rh species', or by H,O with Co,(CO),. The complete reaction sequence with a Ni catalyst is (linear isomer only) shown in Scheme 1.

H20,2 CO

I

>

RCH,CH,CONi(CO),X

co

HNi(CO),X RCH= CH,

RCH,CH,Ni(CO),X Scheme 1. (E. BILLIG, R. L. PRUETT)

1. A. Mullen, New Syntheses with Carbon Monoxide, J. Falbe, ed., Springer Verlag, Berlin, 1980, p. 257. This book and the following referenced chapter provide excellent and current reviews of the topic. 2. (a) P. Pino, F. Piacenti, M. Bianchi, Organic Syntheses Via Metal Carbonyls, I. Wender, P. Pino, eds., Vol. 2, John Wiley & Sons, New York, N.Y., 1977, p. 43; (b) D. P. Riley, Mech. Inorg. Organometal. React., 5, 335 (1988). 3. W. Reppe, H. Kraper, German Patent 863,194 (1953); Chem. Abstr., 48, 1425b (1953). 4. W. Reppe, H. KriSper, German Patent 862,748 (1953); Chem. Abstr., 48, 10059h (1953). 5. G. Natta, P. Pino, Chim. Ind. (Milan),31, 109 (1949). 6. R. F. Heck,J.Am. Chem. SOC.,85, 2013 (1963). 7. D. E. Moms, G. V. Johnson, Sympos. Rhodium Homogen. Catal., Veszprem, Hungary (1978).


Cobalt is the catalyst most widely employed for hydrocarboxylations and hydroesterifications. The strong similarity to hydrofonnylation is shown in the catalyst and in the conditions of T and P. In general the three process rates are in the order: hydroformylation>hydrocarboxylation>hydroesterification. Like hydrofonnylation, the high P of CO required, the instability and toxicity of cobalt carbonyl or hydrocarbonyl and the difficulty of catalyst-product separation detract from the overall attractiveness of this reaction'. Hydrogen donors for hydroesterification may be primary or secondary alcohols, cyclohexanol, phenol or polyols. Linear or branched primary alcohols react similarly; secondary alcohols are less active and tertiary alcohols are not suitable'. The reactivity of various olefins has been compared3. As with iodide promoted Ni catalysts4, iodide promoted Co catalysts convert ethylene, H,O and CO to propionic acid in high efficiency'. The use of CO containing 3% H, in the hydroesterification reaction is standard', suggesting that a cobalt hydrocarbonyl is the active catalyst species. The reaction sequence involves olefin insertion into the Co-H bond, (carbonyl insertion) to give an acyl complex and cleavage with alcohol assisted by the pyridine promoter?

366

14.6. Carbon Monoxide Reactions 14.6.4. H drocarbox lation of Olefins 14.6.4.1. i y Cobalt Zatalysts

Hydroesterification with Co catalysts most commonly utilize pyridine or alkyl pyridines such as y-picoline as cocatalysts. In a massive screening study, Co with pyridine prornotors5, unsubstituted in the ortho positions, gave the highest proportion of linear esters with either 1-octene 01 a mixture of isomeric internal n-dodecenes. The promoter comparison5 was conducted at 160°C and 16.0 MPa for octene and 170°C and 18.0 MPa for the internal dodecenes, in both cases in an excess of CH,OH. The maximum rate for 1-octene conversion occurred at a 15-35:l pyridine: Co ratio. At lower ratios substantial conversions to aldehydes and acetals occurred. In the absence of pyridine, there is nearly complete isomerization of 1-octene to internal olefins7, with only 12% ester formation. Under optimum conditions of 35:l pyridine:Co, a 98% selectivity to ester with 85% methyl n-nonanoate was obtained from 1-octene. With linear, internally-unsaturated dodecenes', the ratio of y-pico1ine:Co of 5- 10:1 is critically important for rate, linearity of product and selectivity of ester. Optimum results with linear, internally-unsaturated dodecenes', 95% selectivity to ester of 75% linearity, were obtained with a 5-1O:l y-pico1ine:cobalt ratio. However, the choice of optimum reaction conditions with internal dodecenes is a compromise since good isomerization activity (for continuous generation of equilibrium quantities of a-olefin) requires a low N:Co ratio; good product rates a medium N:Co ratio; and high selectivity to straight chain product, a high N:Co ratio'. Diols and polyols have also been used effectively as proton sources in hydroesterification. For example', ethylene glycol, 1-octene, Co,(CO),, and in the molar ratio 0.05:0.15:0.03:0.12 in diethyl ether slution at 80°C and 0.4 MPa after 1.5 hr yielded entirely ethylene glycol di-nonanoate. The high ratio of catalyst to olefin may account for the effectiveness under mild conditions. Hydroesterification of butadiene with CH,OH employing a Co catalyst can proceed in a stepwise fashion, initially to methyl 3-penteneoate and subsequently, under more rigorous conditions, to dimethyl adipate. Thus hydroesterification of an olefin stream containing 44% butadiene (remainder mostly butenes) with CH,OH, employing Co,(CO), catalyst and an isoquinoline promoter, gave 98% methyl pent-3-enoate" after 2.5 hours at 120°C under 60 MPa CO. Treatment of the isolated, unsaturated ester with more CO and CH,OH in the presence of Co,(CO), and pyridine promoter at 170°C and 15 MPa gave mostly dimethyl adipate". Other conjugated dienes are less reactive than butadiene; reactivity decreases in the order: butadiene > isoprene > piperylene > 2,4-hexadiene > 2,3-dimethyl-l,3butadiene' '. Olefins substituted with a cyano group at a remote location carbonylate under conditions similar to an unsubstituted olefin. Thus, heating 3-pentenenitrile in CH,OH at 160°C under 20 MPa CO for 4 hr employing a cobalt acetate and a pyridine promotor, yielded methyl 8-cyanovalerate (98% linear) at 40% conver~ion'~. Similar results under similar conditions are obtained with H,C=CHCH(CH,)CN'4. However, acrylonitrile reacts under milder conditions (125"C, 14 MPa) with ethanol and CO in the presence of Co,(CO), and a-picoline to give 95% conversion to the a-substituted cyanoester'': CH,=CHCN

+ CO + C,H,OH

C,H,OOCCHCN

I

Hydroesterification of methyl acrylate with CH,OH employing a Co,(CO), catalyst at 110°C and 16 MPa, gives dimethyl succinate in 75% yield, along with 3% of dimethyl

14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarboxylation of Olefins 14.6.4.2. by Rhodium and Iridium Catalysts

367

methylmalonate'6. Crotonic acid gives a mixture of glutaric acid and methyl succinic acid with water at 150°C and 20 MPa CO. When a hydroxyl group is present in the olefin, lactonization can O C C U ~ ' a~ ,re'~, action which is quite solvent dependent:

Thus, ally1 alcohol gives a 60% yield of y-butyrolactone in acetonitrilelE, but only 2% in benzene17. (E. BILLIG, R. L. PRUETT)

1. Two excellent reviews of this topic are: (a) New Syntheses with Carbon Monoxide, J. Falbe, ed., Springer Verlag, Berlin, 1980; and (b) P. Pino, F. Piacenti, M. Bianchi, Organic Syntheses Via Metal Carbonyls, I. Wender, P. Pino, eds., Vol. 2, John Wiley & Sons, New York, N.Y., 1977, p. 43. The former contains numerous patent references and discusses industrial implications; the latter draws many correlations and includes much mechanistic interpretation. 2. R. Ercoli, Chim. Ind. (Milan), 37, 1029 (1955). 3. R. F. Heck, J . Am. Chem. SOC.,85,2013 (1963). 4. A. Hershman, D. Forster, U.S. Patent, 3,944,604 (1976); Chem. Abstr., 83, 27587 (1975). 5 . P. Hofmann, K. Kosswig, W. Schaefer, Ind. Eng. Chem., Prod. Res. Dev., 19, 330 (1980). 6. M. Katsnel'sm, Zh. Prikl. Khim., 47(1), 155 (1974). 7. R. L. Shubkin, U.S. Patent 3,644,443 (1972); Chem. Abstr., 76, P99122 (1972). 8. See also European Pat. Appln. 17,051 (1980). 9. J. Nakayama, A. Matsudo, K. Bandoh, Japan Kokai, 78/34, 711 (1978); Chem. Abstr., 89, 108176~(1978). 10. R. Kummer, H. W. Schneider, K. Schwirten, Ger. Offen. 2,630,086 (1978); Chem. Abstr., 88, 104688~(1978). 11. R. Kummer, H. W. Schneider, R. Platz, P. Magnussen, F. Weiss, U.S. Patent 4,171,451 (1979); Chem. Abstr., 90, P22343 (1979). 12. N. M. Bogoradovskaia, N. S. Imyanitov, D. M. Rudkovskii, Zh. Prikl. Khim., 39,2807 (1966), as reported in Ref. lb. 13. D. Y. Waddan, D. Wright, British Patent 1,497,046 (1978); Chem. Abstr., 89, P23784 (1978). 14. R. Fischer, H. M. Weitz, Ger. Offen. 2,648,004 (1978); Chem. Abstr., 89, 42485e (1978). 15. M. El-Chahawi, U. Prange, H. Richtzenhain, W. Vogt, Ger. Offen. 2,639,327 (1978); Chem. Abstr., 89, 23785m (1978). 16. A. Matsuda, Bull. Chem. SOC.Japan, 42,571 (1969). 17. J. Falbe, H. J. Schulze-Steinen, F. Korte, Chem. Ber., 98, 886 (1965). 18. A. Matsuda, Bull. Chem. SOC.Japan, 41, 1876 (1968).

14.6.4.2. by Rhodium and Iridium Catalysts

Hydrocarboxylation and hydroesterification of olefins by Rh and Ir catalysts are considered here together since the catalysis schemes are similar and the two metals re treated as one unit in many patents. As with Co, Rh and Ir catalyze the olefin carbonylation reactions of hydrocarboxylation, hydroesterification and acid anhydride formation. Rhodium' or 13 complexes and iodide promoters with H,O as the hydrogen source yields a mixture of linear and branched carbocylic acids; the branched isomer predominates. Many soluble complexes, such as Id3, Ir,(CO)4Brz, Rh(PPh,),(CO)Cl or Ir[(C,H,),P](CO)I can be utilized as a solution in a carboxylic acid solvent. The iodide source can be HI or any material which


14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarboxylation of Olefins 14.6.4.2. by Rhodium and Iridium Catalysts

367

methylmalonate'6. Crotonic acid gives a mixture of glutaric acid and methyl succinic acid with water at 150°C and 20 MPa CO. When a hydroxyl group is present in the olefin, lactonization can O C C U ~ ' a~ ,re'~, action which is quite solvent dependent:

Thus, ally1 alcohol gives a 60% yield of y-butyrolactone in acetonitrilelE, but only 2% in benzene17. (E. BILLIG, R. L. PRUETT)

1. Two excellent reviews of this topic are: (a) New Syntheses with Carbon Monoxide, J. Falbe, ed., Springer Verlag, Berlin, 1980; and (b) P. Pino, F. Piacenti, M. Bianchi, Organic Syntheses Via Metal Carbonyls, I. Wender, P. Pino, eds., Vol. 2, John Wiley & Sons, New York, N.Y., 1977, p. 43. The former contains numerous patent references and discusses industrial implications; the latter draws many correlations and includes much mechanistic interpretation. 2. R. Ercoli, Chim. Ind. (Milan), 37, 1029 (1955). 3. R. F. Heck, J . Am. Chem. SOC.,85,2013 (1963). 4. A. Hershman, D. Forster, U.S. Patent, 3,944,604 (1976); Chem. Abstr., 83, 27587 (1975). 5 . P. Hofmann, K. Kosswig, W. Schaefer, Ind. Eng. Chem., Prod. Res. Dev., 19, 330 (1980). 6. M. Katsnel'sm, Zh. Prikl. Khim., 47(1), 155 (1974). 7. R. L. Shubkin, U.S. Patent 3,644,443 (1972); Chem. Abstr., 76, P99122 (1972). 8. See also European Pat. Appln. 17,051 (1980). 9. J. Nakayama, A. Matsudo, K. Bandoh, Japan Kokai, 78/34, 711 (1978); Chem. Abstr., 89, 108176~(1978). 10. R. Kummer, H. W. Schneider, K. Schwirten, Ger. Offen. 2,630,086 (1978); Chem. Abstr., 88, 104688~(1978). 11. R. Kummer, H. W. Schneider, R. Platz, P. Magnussen, F. Weiss, U.S. Patent 4,171,451 (1979); Chem. Abstr., 90, P22343 (1979). 12. N. M. Bogoradovskaia, N. S. Imyanitov, D. M. Rudkovskii, Zh. Prikl. Khim., 39,2807 (1966), as reported in Ref. lb. 13. D. Y. Waddan, D. Wright, British Patent 1,497,046 (1978); Chem. Abstr., 89, P23784 (1978). 14. R. Fischer, H. M. Weitz, Ger. Offen. 2,648,004 (1978); Chem. Abstr., 89, 42485e (1978). 15. M. El-Chahawi, U. Prange, H. Richtzenhain, W. Vogt, Ger. Offen. 2,639,327 (1978); Chem. Abstr., 89, 23785m (1978). 16. A. Matsuda, Bull. Chem. SOC.Japan, 42,571 (1969). 17. J. Falbe, H. J. Schulze-Steinen, F. Korte, Chem. Ber., 98, 886 (1965). 18. A. Matsuda, Bull. Chem. SOC.Japan, 41, 1876 (1968).

14.6.4.2. by Rhodium and Iridium Catalysts

Hydrocarboxylation and hydroesterification of olefins by Rh and Ir catalysts are considered here together since the catalysis schemes are similar and the two metals re treated as one unit in many patents. As with Co, Rh and Ir catalyze the olefin carbonylation reactions of hydrocarboxylation, hydroesterification and acid anhydride formation. Rhodium' or 13 complexes and iodide promoters with H,O as the hydrogen source yields a mixture of linear and branched carbocylic acids; the branched isomer predominates. Many soluble complexes, such as Id3, Ir,(CO)4Brz, Rh(PPh,),(CO)Cl or Ir[(C,H,),P](CO)I can be utilized as a solution in a carboxylic acid solvent. The iodide source can be HI or any material which

368

14.6. Carbon Monoxide Reactions 14.6.4. H drocarboxylation of Olefins 14.6.4.3. t y Palladium and Platinum Catalysts

generates HI in situ, e.g., alkyl iodides, elemental I, or acetyl iodide. Aqueous HI is a convenient source of both iodide and water for the reaction. Typically, Ir(PPh,),(CO)Cl in acetic acid and aqueous HI reacts with propylene and CO at 175°C and 4.8 MPa pressure (1.3 MPa of CO) to give butyric acids in a 5: 1 i-/n- ratio in 99% selectivity (Ir) or 1.6:lratio (Rh). Catalysis proceeds by oxidative addition of the HI to the Rh complex, followed by insertion of the olefin (ethylene) into the H-Rh bond3 (Scheme 1):

Scheme 1 With Ir as catalyst, special control of reaction parameters are required to maintain a sufficient concentration of the most active catalyst species5. A nonhalide Rh system with tributylphosphine as ligand converts ethylene to methyl propionate4. Thus, a solution of rhodium dicarbonyl acetylacetonate and tributylphosphine in CH,OH at 175°C and 4.7 MPa, gave -30% methyl propionate after about 4 hr. The selectivity to methyl propionate was 80%, the principal by-product being diethylketone. The optimum tributy1phosphine:rhodium ratio is in the range 2: 1-5: 1. Iridium with tributylphosphine like Co and Ru, is inferior to Rh as an ethylene hydroesterification catalyst. Palladium acetate gives no reaction. Propylene and other higher olefins react extremely slowly, even under optimum Rh tributylphosphine catalytic conditions. Iridium complexes containing triphenylphosphine, e.g., HIr(CO),(PPh,),, in propionic acid catalyze ethylene carbonylation to propionic anhydride6. Reaction occurs at a reasonable rate at 195°C and 5 Pa of CO/ethylene pressure. The corresponding Rh complexes are ineffective. The reaction is inoperative with higher olefins, even propylene. Iodide promoted Rh or Ir catalysts are effective for forming propionic anhydride from ethylene, and propionic acid, but the reaction (175"C, 5 MPa) deactivates r a ~ i d l y ; ~ . After one hr of reaction, the rate falls to only 2% of the original amount. However, this rate can be restored by addition of 5% H, to the baseous mixture. The high rate is sustained by maintaining a minimal level of H,. (E. BILLIG, R. L. PRUETT)

1. J. H. Craddock, A. Hershman, F. E. Paulik, J. F. Roth, U.S. Patent 3,579,552 (1971); Chem. Abstr., 72, P110811 (1970). 2. J. H. Craddock, A. Hershman, F. E. Paulik, J. F. Roth, U.S. Patent 3,579,551 (1971); Chem. Absrr., 72, P110811 (1970).

14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarboxylationof Olefins 14.6.4.3. by Palladium and Platinum Catalysts

369

3. D.E.Moms,US.Patent 3,917,677 (1975); Chem. Abstr., 80,70359 (1974). 4. D.E.Moms,G.V. Johnson, Sympos. Rhodium Homogen. Catal., Veszprem, Hungary (1978). 5. D. Forster, A. Hershman, D. E. Moms,U.S. Patent 4,000,170 (1976); Chem. Abstr., 85, 32431 (1976). 6. D.E.Moms,US.Patent 3,944,603 (1976); Chem. Abstr., 84, 179686 (1976). 7. D. Forster, A. Hershman, F. E. Paulik, U.S. Patent 3,852,346 (1974); Chem. Abstr., 81, 19316 (1974).

14.6.4.3. by Palladium and Platinum Catalysts

Hydrocarboxylation and hydroesterification reactions are catalyzed effectively by a variety of Pd and Pt complexes. Palladium and Pt are considered together in patents, since the conditions for reaction are similar and many principles of ligand modification apply to both. Considerable progress in the application of solvent, ligand and promoter effects, have led to milder reaction conditions. Functional olefins have been successfully carbonylated and process regioselectivity has been controlled. Ester synthesis by olefin carbonylation has been known for three decades. For example, Pd(I1) chloride in ethanol containing 15% HC1' at 80°C and 10 MPa, slowly converts a mixture of CO and ethylene to ethyl propionate. The by-products, obtained in small amounts, are ethyl P-ethoxypropionate and ethyl y-ketocaproate. Vinyl chloride gives ethyl propionate and ethyl chloropropionate. Terminal olefins of ZC, chain length give a mixture of linear and a-methyl acid esters. A platinum-tin salt couple, e.g. HPtCl,.nSnCl,, is an effective catalyst for olefin carbonylation'. Oxygenated solvents such as ketones or dimethoxyethane, are preferred, giving a high proportion (-85%) of the linear ester. Typical process conditions are 90°C, 15 MPa CO, and 1 mole-% of H,PtCl,~5SnC12. Isomerization to unreactive internal olefins is the only significant side reaction. A high catalyst concentration, e.g. 1 mole-% of H,PtCl, based on olefin, for high rates of conversion. Certain phosphine and arsines are effective promoters with both Pt and Pd catalysts4x5,with the triphenylarsine ligands being particularly useful. Preferred platinumcontaining catalyst couples are [(C6H5),As],PtC1,~SnC1,, [(C,H5),C1As],PtCl,~SnC12and [(C,H,O),P],PtCl,~SnCl,. A typical reaction composition of CH,OH olefin and Pt catalyst, (Ph,As),PtCl,.lO SnCl,, in a methyl isobutylketone solvent, at a [Pt]:[ 1-heptene]:[CH,OH] ratio of 1:100:740, at 80°C and 24 MPa CO after six hr, gave an 86% yield of methyl octanoates (93% linear) at 94% conversion of 1-heptene. In contrast, a triphenylphosphite promoter [(C,H,O),P],PtCl,~ 10SnC1, under the same conditions gives only 34% conversion and 28% yield, albeit with high regioselectivity (98% linear). 3-Methyl- 1-pentene and even 4-methyl- 1-pentene gives the linear isomer almost exclusively (arsine ligand). Methanol is the best hydrogen source. Water and phenol are useful but 2-propanol gives low conversions. The optimum Pt-Sn ratio for high conversions is 1:lO. Steric factors in the olefin substrate can severely retard carbonylation with the platinum catalysts. For example olefins such as 2-decene, cyclohexene or 2,4,4-trimethyl-lpentene are unreactive. Moderately polar solvents are preferred, e.g., methyl isobutylketone or dimethoxyethane. N,N-Dimethylformamide inhibits carbonylation by forming a stable Pt adduct incapable of olefin complexation. The most effective catalyst combinations for Pd5 are: [(C6H5),],PdC1,~nSnC12; [(p-CH,-C,H4)3P],PdC~z~n~n~~2; [(p-CH3OC6H,),P],PdC~,~n~nC~, and [(C,&)$]z-


14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarboxylationof Olefins 14.6.4.3. by Palladium and Platinum Catalysts

369

3. D.E.Moms,US.Patent 3,917,677 (1975); Chem. Abstr., 80,70359 (1974). 4. D.E.Moms,G.V. Johnson, Sympos. Rhodium Homogen. Catal., Veszprem, Hungary (1978). 5. D. Forster, A. Hershman, D. E. Moms,U.S. Patent 4,000,170 (1976); Chem. Abstr., 85, 32431 (1976). 6. D.E.Moms,US.Patent 3,944,603 (1976); Chem. Abstr., 84, 179686 (1976). 7. D. Forster, A. Hershman, F. E. Paulik, U.S. Patent 3,852,346 (1974); Chem. Abstr., 81, 19316 (1974).

14.6.4.3. by Palladium and Platinum Catalysts

Hydrocarboxylation and hydroesterification reactions are catalyzed effectively by a variety of Pd and Pt complexes. Palladium and Pt are considered together in patents, since the conditions for reaction are similar and many principles of ligand modification apply to both. Considerable progress in the application of solvent, ligand and promoter effects, have led to milder reaction conditions. Functional olefins have been successfully carbonylated and process regioselectivity has been controlled. Ester synthesis by olefin carbonylation has been known for three decades. For example, Pd(I1) chloride in ethanol containing 15% HC1' at 80°C and 10 MPa, slowly converts a mixture of CO and ethylene to ethyl propionate. The by-products, obtained in small amounts, are ethyl P-ethoxypropionate and ethyl y-ketocaproate. Vinyl chloride gives ethyl propionate and ethyl chloropropionate. Terminal olefins of ZC, chain length give a mixture of linear and a-methyl acid esters. A platinum-tin salt couple, e.g. HPtCl,.nSnCl,, is an effective catalyst for olefin carbonylation'. Oxygenated solvents such as ketones or dimethoxyethane, are preferred, giving a high proportion (-85%) of the linear ester. Typical process conditions are 90°C, 15 MPa CO, and 1 mole-% of H,PtCl,~5SnC12. Isomerization to unreactive internal olefins is the only significant side reaction. A high catalyst concentration, e.g. 1 mole-% of H,PtCl, based on olefin, for high rates of conversion. Certain phosphine and arsines are effective promoters with both Pt and Pd catalysts4x5,with the triphenylarsine ligands being particularly useful. Preferred platinumcontaining catalyst couples are [(C6H5),As],PtC1,~SnC1,, [(C,H5),C1As],PtCl,~SnC12and [(C,H,O),P],PtCl,~SnCl,. A typical reaction composition of CH,OH olefin and Pt catalyst, (Ph,As),PtCl,.lO SnCl,, in a methyl isobutylketone solvent, at a [Pt]:[ 1-heptene]:[CH,OH] ratio of 1:100:740, at 80°C and 24 MPa CO after six hr, gave an 86% yield of methyl octanoates (93% linear) at 94% conversion of 1-heptene. In contrast, a triphenylphosphite promoter [(C,H,O),P],PtCl,~ 10SnC1, under the same conditions gives only 34% conversion and 28% yield, albeit with high regioselectivity (98% linear). 3-Methyl- 1-pentene and even 4-methyl- 1-pentene gives the linear isomer almost exclusively (arsine ligand). Methanol is the best hydrogen source. Water and phenol are useful but 2-propanol gives low conversions. The optimum Pt-Sn ratio for high conversions is 1:lO. Steric factors in the olefin substrate can severely retard carbonylation with the platinum catalysts. For example olefins such as 2-decene, cyclohexene or 2,4,4-trimethyl-lpentene are unreactive. Moderately polar solvents are preferred, e.g., methyl isobutylketone or dimethoxyethane. N,N-Dimethylformamide inhibits carbonylation by forming a stable Pt adduct incapable of olefin complexation. The most effective catalyst combinations for Pd5 are: [(C6H5),],PdC1,~nSnC12; [(p-CH,-C,H4)3P],PdC~z~n~n~~2; [(p-CH3OC6H,),P],PdC~,~n~nC~, and [(C,&)$]z-

370

14.6. Carbon Monoxide Reactions 14.6.4. H drocarboxylation of Olefins 14.6.4.3. i y Palladium and Platinum Catalysts

PdCl,-nGeCl,, all of which are effective at lower CO partial P than the Pt catalysts. The optimum rate is achieved at a 5 : 1 Sn:Pd ratio (n = 5 ) . 2-Olefins are also carbonylated, although at a slower rate than 1-olefins. Similarly, cis-Zheptene is five times more active than trans-2-heptene. Ligand steric effects play a prominent role with the phosphine modified Pd catalysts. For example, tris(p-methoxypheny1)-phosphine gives acceptable rates, whereas the catalyst employing the bulkier, tris(o-methoxypheny1)-phosphine, has only 5% of the activity. The promoters, SnCl, or GeC1, with phosphine modified Pd catalysts, increase the regioselectivity of the reaction to the linear ester. With (Ph,P),PdCl, alone, for example, about equal quantities of linear and branched isomers are obtained5.In a typical example, 1-heptene was converted into methyl octanoate with an ester yield of 76% and 87% linear product at 96% conversion of the 1-heptene substrate. The reaction was conducted in methyl isobutylketone at 80°C and 13.6 MPa CO partial P with (Ph3P),PdC1,.10 SnCl, as catalyst. The initial [ 1-heptene], [Pd]:[I-heptene]: [CH,OH] ratios were 1:100:740. Dramatic solvent effects are observed on the regioselectivity of propylene carbonylation to butyl n- and iso-butyrate6. In butanol, for example, a (Ph,P),PdC12 catalyst and excess Ph3P ligand, at 100°C and under 9 MPa of CO partial P, a 2: 1 ratio of normal to branched butyl butyrates are obtained. In methyl ethylketone, however, 1:3 ratio of the n- to i-butyrate esters are obtained. Anhydrides are produced by the carbonylation of olefins when carried out in the presence of carboxylic acids'. For example, Pd acetate, 2-pyridyldiphenylphosphine and p-toluenesulfonic acid in the ratio 0.1:3:5 in a propionic anhydride solvent containing propionic acid (50 mL and 10 mL, respectively), when treated with ethylene (2 MPa) and CO (3 MPa) at 105°C for 1 hr produces 1340 moles of propionic anhydride per mole of Pd per hr with 76% propionic acid conver~ion'~. Palladium catalyzed hydroesterification reactions of a-olefins can be effectively carried out in molten Et,NSnC1,899. For example, a PdCl,(PPh,),.lO[(C,H,),Nl[SnC1,] catalyst composition containing 1-octene and ethanol in the ratio [alkene]:[ethanol]:[Pd] of 100:200:1, at 85°C and 10 MPa CO gives a clear, homogeneous liquid phase. After 8 hr, nonanoate ester yields consistently exceed 80 mole % with product linearity >83%. The major by-products are internal olefins from competing isomerization. On cooling, the quaternary salt precipitates, facilitating separation. Palladium catalysts are at least an order of magnitude more reactive than their Pt counterparts. Similarly, internal olefins are approximately ten times less reactive than aolefins with the Pd catalysts. The Pd catalyst can be recycled but loss of activity occurs slowly and is evident after four cycles. Hydrogen peroxide regenerates the activity". The carbonylation of butadiene, readily available from petroleum sources, yields a dicarbonylated adduct and dimer-carbonylated product, both desirable from a commercial standpoint. Palladium-catalyzed reactions yield monocarbonylated:

, /+

CO + ROH

or dimer-carbonylated products' I : 2,/+CO+ROH

-

-

-C02R

(a)

\

CO,R

(b)

The one-step dimerization-carbonylation utilizes a halide-free Pd complex suspended in a suitable alcohol12,but a combination of tertiary monodentate phosphorus(II1) ligands and tertiary amines gives increased yields, increased solubility, more linear ester and

14.6. Carbon Monoxide Reactions 14.6.4. Hydrocarboxylation of Olefins 14.6.4.3. by Palladium and Platinum Catalysts

371

longer catalyst life". The best combination involves halide-free Pd salts with phosphines of pKa >8 in tertiary amines with dissociation constants which give 80 mole % yields and >90% selectivity to nonadienoate esters. The competing reaction is dimerization of butadiene to a noncarbonylated product, i.e., vinylcyclohexene or linear octatrienes. Ligands of similar size but increasing basicity, for a given series of palladium acetate tertiary monodentate phosphines solubilized in isoquinoline, lead to improved nondienoate ester yields in the order: P(OPh,), 99% from CH,OH to acetyl group, is also markedly increased. The catalyst system consists of a soluble Rh compound and an iodide source such as CH,I. The rhodium compound may be e.g. RhC1,.3 H20, Rh203, [Rh(CO),Cl],, [Ph,As][Rh(CO),I,], (Ph,P),RhCl, [Rh(COD)Cl],, but not complexes which contain certain bifunctional ligands such as [Rh(COD)(a,a'-bipyridyl)] +PF,- or [RhCl(COD)], Ph,PCH2CH2PPh,3,4.Normally, the charged rhodium species is converted to the anion, [Rh(CO),I,)] - , which is thought to be the catalyst precursor2. The commonly accepted reaction mechanism is similar to that for the cobalt catayst.

+

CH,OH CH,I

+ [Rh(CO),I,]

-

[CH,Rh(CO),I,][CH,C(O)Rh(CO)I,]CH,C(O)I

--

+ HI -+ CH,I + H 2 0

+ CO

+ H,O

(a)

[CH,Rh(CO),I,]-

(b)

[CH,C(O)Rh(CO)I,]-

(c)

+ [Rh(CO)JJ CH,COOH + HI

(4

--

CH,C(O)I

(e)

The [Rh(CO),I,] - ion is the only species observed via in situ FTIR under normal catalytic conditions'. The slow step in the catalytic cycle is oxidative addition of CHJ to [Rh(CO),I,]-, which is consistent with the observed reaction rate being first order in [Rh] and [CH,II2. If the reaction is carried out in neat CH,I, the methyl species [CH,Rh(CO),I,]can be observed by FTIR or NMR6. The acetyl complex [CH,C(O)Rh(CO)I,] - has been obtained by reaction of [Rh(CO),I,] - with CH,17. The reaction is zero order in CH,OH and independent of the partial pressure of CO at >200 KPa33'. Methanol carbonylation functions better in a polar solvent. The reaction product, acetic acid, serves well as a s o l ~ e n t ~Water * ' ~ . has a beneficial effect and also effects the methyl acetate:acetic acid composition of the productg. In acetophenone the formation of (CH,),O is greatly diminished over that produced when CH,OH alone is used as solvent'l. Reaction rate is independent of CO and is also insensitive to H,. Thus, CO feedstocks containing various amounts of H, are suitable'. In the absence of added ligands, the rhodium CH,I catalyst is a very poor homologation catalyst". Normally, with a standard rhodium-CH,I catalyst the H,O content in the reactor is quite high (up to 50% by weight) in order to shift the methyl acetate to acetic acid.

TABLE1. REACTION CONDITIONS FOR Co AND RHCATALYZED CARBONYLATION OF METHANOL

co [M](mol./l.) t("C) rate, Mhr- ' P(MPa)

0.1 230

1.o

70

Rh 0.001

180

10.0

3

14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.1. Carbon lation of Alcohols 14.6.5.13. by Otter Metal Catalysts.

375

Removal of the H 2 0 in the product recovery step is very expensive. Carbonylation of CH,OH in near anhydrous conditions can be achieved with ionic iodide prornoter~'~''~. For example, a catalyst consisting of Rh and LiI carbonylates a mixture of CH,OH, CH,I, and methyl acetate to acetic acid at about the same rate as the standard rhodium CH,I catalyst, but only with about one-eighth of the H,O present. Low H,O content in the reaction product leads to significant reduction in the cost of distilling acetic acid. Carbonylation of CH,OH is important because of the strong commercial interest in acetic acid, but other alcohols may also be e m p l ~ y e d ' ~Butyric . and valeric acids are obtained from the carbonylation of n/i-propanol and n/i-butanol16, octanoic acid is produced by carbonylation of heptanol', adipic acid by the double carbonylation of butanediol', phenylenediacetic acid from p-xylene-c~,c~'-diol'~ and phenylacetic acid from benzyl alcohol''. However, carbonylation of ethylene glycol produces propionic acid18. The carbonylation of CH,OH to acetic acid is also catalyzed by Rh on support^'^-'^. Several studies have involved Rh cation-exchanged zeolites, which operate at 150-200°C and atmospheric pressure with CH,I as promote?'. The kinetics are similar to the homogeneous reaction, i.e., zero-order in CH,OH and CO and first-order in CH,I. For NaX zeolites ion exchanged with [Rh(NH,),CI]Cl, and activated by heating to 400°C, under N, for 30 minutes, an activity similar to homogeneous systems is claimed at 100 KPa of 25:1:0.12 CO:CH,OH:CH,IZ2. (R. W. WEGMAN)

1. 2. 3. 4. 5.

6.

7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

F. E. Paulik, J. F. Roth, J . Chem. SOC., Chem. Commun., 1578 (1968). D. Forster, Adv. Organometal. Chem., 17, 255 (1979). J. F. Roth, J. H. Craddock, A. Hershman, F. E. Paulik, Chemtech., 600 (1971). D. Brodski, E. Leclere, B. Denise, G. Pannetier, Bull. SOC. Chim. Fr. 61 (1976). D. E. Moms, H. B. Tinker, Chem. Techno/. 554, (1972). A. Haynes, B. E. Mann, D. J. Gulliver, G. E. Moms, P. M. Maitlis, J . Am. Chem. Soc., 113, 8569, (1991). D. Forster, J. Am. Chem. Soc., 98, 846, (1976). H. Hjortkjaer, J. W. Jensen, Znd. Eng. Chem., Prod. Res. Dev., 15 (1). 46 (1976). F. E. Paulik, A. Hershman, W. R. Knox, J. F. Roth, U.S. Patent 3,769,329 (1973); Chem. Abstr., 71, P12573 (1969). D. Brodski, B. Denise, G. Pannetier, J . Mol. Car., 2, 149 (1977). T. Matsumoto, K. Mori, T. Mizoroki, A. Ozaki, Bull. Chem. SOC.Japan, 50 (9), 2337 (1977). H. Dumas, J. Levisalles, H. Rudler, J . Organometal. Chem., 177, 239 (1979). M. A. Murphy, B. L. Smith, G . P. Torrence, A. Aguilo, J . Organomet Chem.,303,257, (1986). B. L. Smith, G. P. Torrence, M. A. Murphy, A. Aguilo, J . Mol. Cat., 39, 115, (1987). D. Forster, T. W. Dekleva, J . Chem. Ed., 63, 204, (1986). S. B. Dake, R. V. Chaudhari, J . Mol. Cat., 35, 119, (1986). D. E. Moms, Res. Discl., 14,037, Dec. 1975, p. 27. F. E. Paulik, A. Hershman, W. R. Knox, J. F. Roth, U.S. Patent 3,813,428 (1974); Chem. Abstr., 72, P110807 (1970). R. G. Schultz, P. D. Montgomery, J . Catal., 13, 105 (1969). K. K. Robinson, JA. Hershman, J. H. Craddock, J. F. Roth, J . Catal., 27, 389 (1972). M. S. Jarrell, B. C. Gates, J . Catal., 40, 255 (1975). M. S . Scurrell, R. F. Howe, J . Mol. Cat., 7, 535 (1980).

14.6.5.1 3.by Other Metal Catalysts.

Most group VIII metals will, to some extent, catalyze carbonylation. Early work utilized Ni and iodine as catalyst. Using metal iodides on silica as carbonylation catalysts at 175-23OoC, -27 MPa, and a 2 hr residence time shows the reactivity order Ni > Co


14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.1. Carbon lation of Alcohols 14.6.5.13. by Otter Metal Catalysts.

375

Removal of the H 2 0 in the product recovery step is very expensive. Carbonylation of CH,OH in near anhydrous conditions can be achieved with ionic iodide prornoter~'~''~. For example, a catalyst consisting of Rh and LiI carbonylates a mixture of CH,OH, CH,I, and methyl acetate to acetic acid at about the same rate as the standard rhodium CH,I catalyst, but only with about one-eighth of the H,O present. Low H,O content in the reaction product leads to significant reduction in the cost of distilling acetic acid. Carbonylation of CH,OH is important because of the strong commercial interest in acetic acid, but other alcohols may also be e m p l ~ y e d ' ~Butyric . and valeric acids are obtained from the carbonylation of n/i-propanol and n/i-butanol16, octanoic acid is produced by carbonylation of heptanol', adipic acid by the double carbonylation of butanediol', phenylenediacetic acid from p-xylene-c~,c~'-diol'~ and phenylacetic acid from benzyl alcohol''. However, carbonylation of ethylene glycol produces propionic acid18. The carbonylation of CH,OH to acetic acid is also catalyzed by Rh on support^'^-'^. Several studies have involved Rh cation-exchanged zeolites, which operate at 150-200°C and atmospheric pressure with CH,I as promote?'. The kinetics are similar to the homogeneous reaction, i.e., zero-order in CH,OH and CO and first-order in CH,I. For NaX zeolites ion exchanged with [Rh(NH,),CI]Cl, and activated by heating to 400°C, under N, for 30 minutes, an activity similar to homogeneous systems is claimed at 100 KPa of 25:1:0.12 CO:CH,OH:CH,IZ2. (R. W. WEGMAN)

1. 2. 3. 4. 5.

6.

7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

F. E. Paulik, J. F. Roth, J . Chem. SOC., Chem. Commun., 1578 (1968). D. Forster, Adv. Organometal. Chem., 17, 255 (1979). J. F. Roth, J. H. Craddock, A. Hershman, F. E. Paulik, Chemtech., 600 (1971). D. Brodski, E. Leclere, B. Denise, G. Pannetier, Bull. SOC. Chim. Fr. 61 (1976). D. E. Moms, H. B. Tinker, Chem. Techno/. 554, (1972). A. Haynes, B. E. Mann, D. J. Gulliver, G. E. Moms, P. M. Maitlis, J . Am. Chem. Soc., 113, 8569, (1991). D. Forster, J. Am. Chem. Soc., 98, 846, (1976). H. Hjortkjaer, J. W. Jensen, Znd. Eng. Chem., Prod. Res. Dev., 15 (1). 46 (1976). F. E. Paulik, A. Hershman, W. R. Knox, J. F. Roth, U.S. Patent 3,769,329 (1973); Chem. Abstr., 71, P12573 (1969). D. Brodski, B. Denise, G. Pannetier, J . Mol. Car., 2, 149 (1977). T. Matsumoto, K. Mori, T. Mizoroki, A. Ozaki, Bull. Chem. SOC.Japan, 50 (9), 2337 (1977). H. Dumas, J. Levisalles, H. Rudler, J . Organometal. Chem., 177, 239 (1979). M. A. Murphy, B. L. Smith, G . P. Torrence, A. Aguilo, J . Organomet Chem.,303,257, (1986). B. L. Smith, G. P. Torrence, M. A. Murphy, A. Aguilo, J . Mol. Cat., 39, 115, (1987). D. Forster, T. W. Dekleva, J . Chem. Ed., 63, 204, (1986). S. B. Dake, R. V. Chaudhari, J . Mol. Cat., 35, 119, (1986). D. E. Moms, Res. Discl., 14,037, Dec. 1975, p. 27. F. E. Paulik, A. Hershman, W. R. Knox, J. F. Roth, U.S. Patent 3,813,428 (1974); Chem. Abstr., 72, P110807 (1970). R. G. Schultz, P. D. Montgomery, J . Catal., 13, 105 (1969). K. K. Robinson, JA. Hershman, J. H. Craddock, J. F. Roth, J . Catal., 27, 389 (1972). M. S. Jarrell, B. C. Gates, J . Catal., 40, 255 (1975). M. S . Scurrell, R. F. Howe, J . Mol. Cat., 7, 535 (1980).

14.6.5.1 3.by Other Metal Catalysts.

Most group VIII metals will, to some extent, catalyze carbonylation. Early work utilized Ni and iodine as catalyst. Using metal iodides on silica as carbonylation catalysts at 175-23OoC, -27 MPa, and a 2 hr residence time shows the reactivity order Ni > Co

376

14.6. Carbon Monoxide Reactions 14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.2. lsornerization of Formates

> Fe'. With CH,I and various inorganic and organic promoters, a Ni based catalyst carbonylates CH,OH at 150-170°C and 2-7 MPa2. Iridium complexes catalyze carbonylation of CH,OH to acetic acid, also with an iodide promote?. The reaction rate relative to Rh is much slower. The steps in the reaction sequence are similar to those for Rh, but the kinetics are more ~ o m p l e x ~ - ~ . Complex interactions involve H20, the form of the iodide promoter and CO pressure6,'. For example, at high concentration of I- ion, the rate increases with increasing pressure. At low I - levels and low H,O concentration, the reaction rate is inversely dependent on CO pressure. Catalyst species under these different reaction conditions include Ir(CO)31, IrH(CO)21z(OHz),[Ir(CH,)(CO),I,] - and [IrH(CO),I,] -. In acetophenone solvent at 175°C and 3 MPa, the reaction is first-order in CH,OH and independent of CH31 at concentrations in which the 1:Ir ratio is >205. Under some conditions, the water gas shift reaction becomes important; careful control is necessary for high efficiencies to acetic acid. (R. W. WEGMAN)

1. 2. 3. 4. 5. 6. 7.

K. Bhattacharyya, S. Sourirajan, J . Appl. Chem., 9, 126 (1959). J. Gauthier-Lafaye, R. Perron, Methanol and Carbonylation, Editions Technip, Paris, 1987. F. E. Paulik, J. F. Roth, J . Chem. SOC., Chem. Commun., 1578 (1968). D. Brodzki, B. Denise, G. Pannetier, J . Mol. Cat., 2, 149 (1977). T. Matsumoto, T. Mizoroki, A. Ozaki, J . Catal., 51, 96 (1978). D. Forster, Adv. Organomet. Chem., 17, 255 (1979). D. Forster, J . Chem. SOC.,Dalton Trans., 1639 (1979).

14.6.5.2. lsomerization of Formates

Isomerization of formates:

-

RC(O)OH, (a) HC(0)OR forms carboxylic acids in chemistry that is quite similar to carbonylation of alcohols. The reaction is catalyzed by group VIII metals such as Co, Rh, Ir, Pd, or Isomerization of methyl formate to acetic acid is a well-known; reports in the patent literature date back to 1929. With a Co-iodide catalyst the reaction is carried out at 160" and 10.5 MPa CO'. The selectivity to acetic acid is >95%. The best reported productivities are obtained with a Rh-LiI catalyst'. In this case, the reaction is carried out at 180°C and 2.75 MPa with 99% conversion and near quantitative yield of acetic acid. The mechanism of the reaction involves initial cleavage of methyl formate by LiI. CH,I, obtained in the cleavage reaction, is carbonylated to acetyl iodide via the same catalytic chemistry observed in CH,OH carbonylation. The key to making acetic acid is that the mixed anhydride CH,C(O)OC(O)H is unstable and thermally decomposes to acetic acid and CO at the reaction conditions. HC(O)OCH,

+ LiI

+ CO + Rh catalyst CH,C(O)I + HC(0)OLi

CH,I

CH,C(O)OC(O)H

--

CH31 + HC(0)OLi

(b)

CH,C(O)I

(c)

-

CH,C(O)OC(O)H CH,COOH

+ CO

+ LiI

(dl (el

(R. W. WEGMAN)


376

14.6. Carbon Monoxide Reactions 14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.2. lsornerization of Formates

> Fe'. With CH,I and various inorganic and organic promoters, a Ni based catalyst carbonylates CH,OH at 150-170°C and 2-7 MPa2. Iridium complexes catalyze carbonylation of CH,OH to acetic acid, also with an iodide promote?. The reaction rate relative to Rh is much slower. The steps in the reaction sequence are similar to those for Rh, but the kinetics are more ~ o m p l e x ~ - ~ . Complex interactions involve H20, the form of the iodide promoter and CO pressure6,'. For example, at high concentration of I- ion, the rate increases with increasing pressure. At low I - levels and low H,O concentration, the reaction rate is inversely dependent on CO pressure. Catalyst species under these different reaction conditions include Ir(CO)31, IrH(CO)21z(OHz),[Ir(CH,)(CO),I,] - and [IrH(CO),I,] -. In acetophenone solvent at 175°C and 3 MPa, the reaction is first-order in CH,OH and independent of CH31 at concentrations in which the 1:Ir ratio is >205. Under some conditions, the water gas shift reaction becomes important; careful control is necessary for high efficiencies to acetic acid. (R. W. WEGMAN)

1. 2. 3. 4. 5. 6. 7.

K. Bhattacharyya, S. Sourirajan, J . Appl. Chem., 9, 126 (1959). J. Gauthier-Lafaye, R. Perron, Methanol and Carbonylation, Editions Technip, Paris, 1987. F. E. Paulik, J. F. Roth, J . Chem. SOC., Chem. Commun., 1578 (1968). D. Brodzki, B. Denise, G. Pannetier, J . Mol. Cat., 2, 149 (1977). T. Matsumoto, T. Mizoroki, A. Ozaki, J . Catal., 51, 96 (1978). D. Forster, Adv. Organomet. Chem., 17, 255 (1979). D. Forster, J . Chem. SOC.,Dalton Trans., 1639 (1979).

14.6.5.2. lsomerization of Formates

Isomerization of formates:

-

RC(O)OH, (a) HC(0)OR forms carboxylic acids in chemistry that is quite similar to carbonylation of alcohols. The reaction is catalyzed by group VIII metals such as Co, Rh, Ir, Pd, or Isomerization of methyl formate to acetic acid is a well-known; reports in the patent literature date back to 1929. With a Co-iodide catalyst the reaction is carried out at 160" and 10.5 MPa CO'. The selectivity to acetic acid is >95%. The best reported productivities are obtained with a Rh-LiI catalyst'. In this case, the reaction is carried out at 180°C and 2.75 MPa with 99% conversion and near quantitative yield of acetic acid. The mechanism of the reaction involves initial cleavage of methyl formate by LiI. CH,I, obtained in the cleavage reaction, is carbonylated to acetyl iodide via the same catalytic chemistry observed in CH,OH carbonylation. The key to making acetic acid is that the mixed anhydride CH,C(O)OC(O)H is unstable and thermally decomposes to acetic acid and CO at the reaction conditions. HC(O)OCH,

+ LiI

+ CO + Rh catalyst CH,C(O)I + HC(0)OLi

CH,I

CH,C(O)OC(O)H

--

CH31 + HC(0)OLi

(b)

CH,C(O)I

(c)

-

CH,C(O)OC(O)H CH,COOH

+ CO

+ LiI

(dl (el

(R. W. WEGMAN)

14.6. Carbon Monoxide Reactions 14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.4. Reductive Carbonylation of Alcohols 1. 2. 3. 4. 5.

377

M. Roeper, E. Elvenvoll, M. Luetgendorf, Erdol. Kohle. Erdgas. Petrochem., 38, 38, (1985). D. J. Schreck, D. C. Busby, R. W. Wegman, J . Mol. Chem., 47, 117, (1988). R. L. Pruett, R. T. Kacmarick, Organometallics, 1, 1693, (1982). J. Gauthier-Lafaye, R. Perron, Methanol and Carbonylation, Editions Technip, Paris, 1987. G . Jenner, E. M. Nahmed, J . Organomet. Chem., 407, 135, (1991).

14.6.5.3. Carbonylation of Esters

-

Carbonylation of an ester with group VIII metal catalyst leads to the anhydride': RC(0)OR + CO RC(O)OC(O)R (a) Cobalt and Ni catalyze the carbonylation of strained cyclic esters such as /3-propiolactone and b u t y r ~ l a c t o n ePalladium ~~~. chloride carbonylates esters of ally1 alcohol to 3-butenoic anhydrides at 100" and 9.6 MPa4. The most studied ester carbonylation reaction is that of methyl acetate to acetic anhydride. This reaction is practiced commercially by E. Kodak'. The catalyst consists of Rh and a mixture of promoters that include Cr and CH316.For example, a mixture of 350 parts methyl acetate, 2.25 parts RhC1,.3H,O, 57 parts CH,I, and 17 parts Cr(CO), in 90 parts acetic acid was heated at 175°C under 2.5 MPa pressure of CO for 1 hr to give 54% acetic anhydride. The observed rate of acetic anhydride formation is 5 MHIat optimum operating conditions of 170-200" and 3.5 MPa C07. Marked improvements in rate and catalyst stability were made by the addition of lithium salts or Li18,9.In part, it is thought that LiI cleaves methyl acetate as shown in the highly simplified reaction mechanism: CH,C(O)OCH, LiI --+CH,I + CH,C(O)OLi (b)

+

CH31 + CO CH,C(O)I

+ Rh catalyst

+ CH,C(O)OLi

--

CH,C(O)I CH3C(0)OC(O)CH3

(c)

+ LiI

(4

The catalyst chemistry is similar to that already described for CH,OH carbonylation. Exceptions are the involvement of Li+ as a stabilizing counter ion for Rh anion catalyst of H, to help regenerate the active intermediates and the use of a small amount (4%) catalyst precursors. In the absence of H,, the complex [Rh(CO),I,] - forms and is catalytically inactive. H, reacts with [Rh(CO),I,] - to regenerate [Rh(CO),I,] -, the precursor to the active catalyst'. (R. W. WEGMAN)

1 . H. M. Colquhoun, D. J. Thompson, M. V. Twigg, Carbonylation: Direct Synthesis of Carbonyl Compounds, Plenum Press, New York, 1991. 2. Y. Mori, J. Tsuji, Bull, Chem. SOC.Jap., 42,77, (1969). 3. W. Reppe, H. Koper, H. J. Pistor, 0. Weissbarth, Justus Liebigs Ann. Chem., 582, 87, (1953). 4. J. Tsuji, J. Kiji, S. Imamura, M. Morikawa, J. Am. Chem. Soc., 86, 4350, (1964). 5 . H. W. Coover, R. C. Hart, Chem. Eng. Progr., 78,72, (1982). 6. N. Rizkalla, Belgium Patent 839,322 (1976); Chem. Abstr., 85, P159463 (1976). 7. G . Luft, M. Schrod, J. Mol. Cat., 20, 175, (1983). 8. S. W. Polichnowski, J . Chem. Ed., 63, 206, (1986). 9. A. Fulford, C. E. Hickey, P. M. Maitlis, J . Organomet. Chem., 398, 31 1, (1990).

14.6.5.4. Reductive Carbonylation of Alcohols

Reductive carbonylation refers to the chain growth of an alcohol by a CH, unit. For example, in the case of CH,OH, the products are acetaldehyde and/or ethanol:



377



-


+

CH31 + CO CH,C(O)I

+ Rh catalyst

+ CH,C(O)OLi

--


(c)

+ LiI

(4







377



-


+

CH31 + CO CH,C(O)I

+ Rh catalyst

+ CH,C(O)OLi

--


(c)

+ LiI

(4





378

14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.4. Reductive Carbonylation of Alcohols 14.6.5.4.1. by Cobalt Catalysts.

+ CO + H, CH30H + CO + 2Hz CH3OH

CH,C(O)H CH,CH,OH

+ H20 + HZO

(4 (b)

The reaction is catalyzed by most group VIII metals, including Rh, Ir, Mn, Cr, Os, Ru and Co'. (R. W. WEGMAN)

1. J. Falbe, New Syntheses wirh Carbon Monoxide, Springer Verlag, Berlin, 1980. 14.6.5.4.1. by Cobalt Catalysts.

Of the group VIII metal catalysts, Co has received the most attention. With Co,(CO),, at 180-185°C and CO:H, pressures of 35 MPa, CH,OH gives a product distribution of 38.8% ethanol, 4.7% n-propanol, 9.0% methyl acetate, 6.3% ethyl acetate, 8.5% methane and traces of acetaldehyde, methyl formate, propyl acetate and butanol'.'. At 160-180" and 21 MPa of CO:H,, tert-butanol gives a 63% yield of iso-amyl alcohol, and iso-propanol gives 11% of a mixture of n- and iso-butanol. n-Propanol reacts slowly at 180°C. Substituted benzyl alcohols undergo homologation at 185°C and 24 MPa of 2:l H,:CO, but hydrogenolysis to toluene is a significant side reaction3 The rate decreases with imcreasing electron-withdrawing abilities of the substituting group: p-OCH, >> p-CH, > m-CH, >p-tert-butyl > H > C1 > m-OCH, >> m-CF,. p-Methoxybenzyl alcohol gives 44% homologation product, 15% hydrocarbon and -34% high-boiling compounds which are aldol condensation products of the intermediate aldehyde. The electronic effects are striking; the rate of the p-methoxy derivative is lo4 that of the m-methoxy. Most work has used CH,OH as a feedstock, since it is readily obtained from synthesis gas. Addition of an iodide promoter, solvent, and a group V promoter markedly increases the rate and, under controlled conditions, the selectivity to acetaldehyde4s5.In general, the best acetaldehyde rates and selectivities, typically 3-5 Mhr- and 80-90%, are obtained with cobalt I,-PPh, in ether or polyether solvents at 170" and 34.5 MPa (H,:CO = 1 3 5 . Under forcing conditions, temperatures > 200", it is possible to hydrogenate acetaldehyde to ethanol, as it is formed, with a cobalt I,-PPh, catalyst. Yields of ethanol up to 70% have been reported4. Better results are obtained by adding Ru to the Co catalyst6. Here, Ru serves as a hydrogenation catalyst for the conversion of acetaldehyde to ethanol. For example, with a catalyst consisting of Co:Ru:I = 1:5:4, at 140°C and 24.1 MPa (H,:CO = 2:l) the ethanol selectivity is 86%7. The postulated mechanism for the reductive carbonylation reaction involves initial formation of CH,Co(CO), via reaction of HCo(CO), with CH31 or CH30H8: HCo(CO),

CH,Co(CO), CH,C(O)Co(CO), CH,C(O)H

--

+ CH31/CH30H +CH,Co(CO), + HI/H,O + CO + H,

+ H, + catalyst

-

CH,C(O)Co(CO), CH,C(O)H CH,CH,OH

+ HCo(C0)4

(a) (b) (c)

(4


378

14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.4. Reductive Carbonylation of Alcohols 14.6.5.4.1. by Cobalt Catalysts.

+ CO + H, CH30H + CO + 2Hz CH3OH

CH,C(O)H CH,CH,OH

+ H20 + HZO

(4 (b)

The reaction is catalyzed by most group VIII metals, including Rh, Ir, Mn, Cr, Os, Ru and Co'. (R. W. WEGMAN)

1. J. Falbe, New Syntheses wirh Carbon Monoxide, Springer Verlag, Berlin, 1980. 14.6.5.4.1. by Cobalt Catalysts.

Of the group VIII metal catalysts, Co has received the most attention. With Co,(CO),, at 180-185°C and CO:H, pressures of 35 MPa, CH,OH gives a product distribution of 38.8% ethanol, 4.7% n-propanol, 9.0% methyl acetate, 6.3% ethyl acetate, 8.5% methane and traces of acetaldehyde, methyl formate, propyl acetate and butanol'.'. At 160-180" and 21 MPa of CO:H,, tert-butanol gives a 63% yield of iso-amyl alcohol, and iso-propanol gives 11% of a mixture of n- and iso-butanol. n-Propanol reacts slowly at 180°C. Substituted benzyl alcohols undergo homologation at 185°C and 24 MPa of 2:l H,:CO, but hydrogenolysis to toluene is a significant side reaction3 The rate decreases with imcreasing electron-withdrawing abilities of the substituting group: p-OCH, >> p-CH, > m-CH, >p-tert-butyl > H > C1 > m-OCH, >> m-CF,. p-Methoxybenzyl alcohol gives 44% homologation product, 15% hydrocarbon and -34% high-boiling compounds which are aldol condensation products of the intermediate aldehyde. The electronic effects are striking; the rate of the p-methoxy derivative is lo4 that of the m-methoxy. Most work has used CH,OH as a feedstock, since it is readily obtained from synthesis gas. Addition of an iodide promoter, solvent, and a group V promoter markedly increases the rate and, under controlled conditions, the selectivity to acetaldehyde4s5.In general, the best acetaldehyde rates and selectivities, typically 3-5 Mhr- and 80-90%, are obtained with cobalt I,-PPh, in ether or polyether solvents at 170" and 34.5 MPa (H,:CO = 1 3 5 . Under forcing conditions, temperatures > 200", it is possible to hydrogenate acetaldehyde to ethanol, as it is formed, with a cobalt I,-PPh, catalyst. Yields of ethanol up to 70% have been reported4. Better results are obtained by adding Ru to the Co catalyst6. Here, Ru serves as a hydrogenation catalyst for the conversion of acetaldehyde to ethanol. For example, with a catalyst consisting of Co:Ru:I = 1:5:4, at 140°C and 24.1 MPa (H,:CO = 2:l) the ethanol selectivity is 86%7. The postulated mechanism for the reductive carbonylation reaction involves initial formation of CH,Co(CO), via reaction of HCo(CO), with CH31 or CH30H8: HCo(CO),

CH,Co(CO), CH,C(O)Co(CO), CH,C(O)H

--

+ CH31/CH30H +CH,Co(CO), + HI/H,O + CO + H,

+ H, + catalyst

-

CH,C(O)Co(CO), CH,C(O)H CH,CH,OH

+ HCo(C0)4

(a) (b) (c)

(4

14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.4. Reductive Carbonylation of Alcohols 14.6.5.4.2. by Other Metals.

379

The mechanism is highly simplified and neglects coordination of added group V ligands or iodide. The rate enhancement observed with added iodide is thought to be a result of the faster reaction of HCo(CO), with CH31relative to CH,OH. The principal by-product is methyl acetate, which results from the reaction of CH30H with the acyl intermediate CH,C(O)Co(CO),. (R. W. WEGMAN) I. Wender, R. Levine, M. Orchin, J . Am. Chem. Soc., 71,4160 (1949). I. Wender, R. A. Friedel, M. Orchin, Science, 113,206 (1951). I. Wender, H. Greenfield, S.Metlin, M. Orchin, J . Am. Chem. Soc., 74, 4079, (1952). W. R. Pretzer, T. P. Kobylinski, Annals. N.Y. Acad. Sci., 333, 58, (1980). R. W. Wegman, D. C. Busby, J. Mol. Chem., 32, 125, (1985). Pertinent patents include: U.S.Patent 3,285,948 (1966); Chem.Abstr., 66, P65072 (1967). Japan Kokai 77-133,914, Chem. Abstr., 88, 74045 (1978); 77-136,110, Chem. Abstr., 88, 74047 (1978); 77-136,111, Chem. Abstr., 88, 74046 (1978), U.S.Patent 4,133,966 (1979); Chem. Absrr., 90, 120998 (1979). 7. G. Jenner, P. Andrianary, J. Cat. 88,535, (1985). 8. Caralysis in Organic Chemistry, D. W. Slocum, W. H. Jones, eds., Academic Press, New York, 1980.

1. 2. 3. 4. 5. 6.

14.6.5.4.2. by Other Metals.

Rhodium, promoted only with CH31, homologates CH30H to ethanol at 110°C and 12 MPa, but only with very high (40:l) ratios of H, to CO'. Catalyst activity is low; in 5 hr 50% selectivity to ethanol or ethyl acetate is obtained. Even under these conditions which favor hydrogenation, methyl acetate is a significant component of the product. Below a 10:1 ratio of H, to CO very little ethanol is produced. It has been discovered recently that with diphosphine ligands Rh selectively generates acetaldehyde'. For example, with Ph,P(CH,),PPh,, acetaldehyde is produced with selectivities approaching 90% at 130°C and 6.8 MPa (H,:CO = 1:l). If Ru is employed as a cocatalyst, acetaldehyde is hydrogenated in situ to ethanol with the same high selectivity. The catalyst performs with good rates and selectivities at much lower T and P than previously demonstrated for traditional cobalt based catalysts. The marked change in selectivity to acetaldehyde may be due to stabilization of a Rh acyl intermediate by the diphosphine3. The postulated mechanism is similar to Rh catalyzed carbonylation of CH,OH (in the scheme (P-P) represents a coordinated diphosphine). Rh(CO)I(P-P) CH31 -+ CH,C(O)RhI,(P-P) (a)

+

CH,C(O)RhI,(P-P) HRhI,(P-P) HI

+ H, + CO

+ CH,OH

-

CH,C(O)H

_.)

+ HRhI,(P-P)

Rh(CO)I(P-P) CH,I

+ H,O

+ HI

(b) (c) (4

The acyl compound CH,C(O)Rhl,(P-P) is isolated nearly quantitatively at the end of reaction. Stabilization of the Rh acyl intermediate by the diphosphine and its subsequent hydrogenation leads to acetaldehyde3. An iron-amine catalyst homologates CH,OH as4: CH,OH 2CO H, __* CH,CH,OH CO, (el The reaction is carried out at 200°C and 31 MPa; the rate and selectivity are low. Manganese and ruthenium carbonyls, alone or together, are catalyst^^.^.

+

+

+

(R. W. WEGMAN)


14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.4. Reductive Carbonylation of Alcohols 14.6.5.4.2. by Other Metals.

379

The mechanism is highly simplified and neglects coordination of added group V ligands or iodide. The rate enhancement observed with added iodide is thought to be a result of the faster reaction of HCo(CO), with CH31relative to CH,OH. The principal by-product is methyl acetate, which results from the reaction of CH30H with the acyl intermediate CH,C(O)Co(CO),. (R. W. WEGMAN) I. Wender, R. Levine, M. Orchin, J . Am. Chem. Soc., 71,4160 (1949). I. Wender, R. A. Friedel, M. Orchin, Science, 113,206 (1951). I. Wender, H. Greenfield, S.Metlin, M. Orchin, J . Am. Chem. Soc., 74, 4079, (1952). W. R. Pretzer, T. P. Kobylinski, Annals. N.Y. Acad. Sci., 333, 58, (1980). R. W. Wegman, D. C. Busby, J. Mol. Chem., 32, 125, (1985). Pertinent patents include: U.S.Patent 3,285,948 (1966); Chem.Abstr., 66, P65072 (1967). Japan Kokai 77-133,914, Chem. Abstr., 88, 74045 (1978); 77-136,110, Chem. Abstr., 88, 74047 (1978); 77-136,111, Chem. Abstr., 88, 74046 (1978), U.S.Patent 4,133,966 (1979); Chem. Absrr., 90, 120998 (1979). 7. G. Jenner, P. Andrianary, J. Cat. 88,535, (1985). 8. Caralysis in Organic Chemistry, D. W. Slocum, W. H. Jones, eds., Academic Press, New York, 1980.

1. 2. 3. 4. 5. 6.

14.6.5.4.2. by Other Metals.

Rhodium, promoted only with CH31, homologates CH30H to ethanol at 110°C and 12 MPa, but only with very high (40:l) ratios of H, to CO'. Catalyst activity is low; in 5 hr 50% selectivity to ethanol or ethyl acetate is obtained. Even under these conditions which favor hydrogenation, methyl acetate is a significant component of the product. Below a 10:1 ratio of H, to CO very little ethanol is produced. It has been discovered recently that with diphosphine ligands Rh selectively generates acetaldehyde'. For example, with Ph,P(CH,),PPh,, acetaldehyde is produced with selectivities approaching 90% at 130°C and 6.8 MPa (H,:CO = 1:l). If Ru is employed as a cocatalyst, acetaldehyde is hydrogenated in situ to ethanol with the same high selectivity. The catalyst performs with good rates and selectivities at much lower T and P than previously demonstrated for traditional cobalt based catalysts. The marked change in selectivity to acetaldehyde may be due to stabilization of a Rh acyl intermediate by the diphosphine3. The postulated mechanism is similar to Rh catalyzed carbonylation of CH,OH (in the scheme (P-P) represents a coordinated diphosphine). Rh(CO)I(P-P) CH31 -+ CH,C(O)RhI,(P-P) (a)

+

CH,C(O)RhI,(P-P) HRhI,(P-P) HI

+ H, + CO

+ CH,OH

-

CH,C(O)H

_.)

+ HRhI,(P-P)

Rh(CO)I(P-P) CH,I

+ H,O

+ HI

(b) (c) (4

The acyl compound CH,C(O)Rhl,(P-P) is isolated nearly quantitatively at the end of reaction. Stabilization of the Rh acyl intermediate by the diphosphine and its subsequent hydrogenation leads to acetaldehyde3. An iron-amine catalyst homologates CH,OH as4: CH,OH 2CO H, __* CH,CH,OH CO, (el The reaction is carried out at 200°C and 31 MPa; the rate and selectivity are low. Manganese and ruthenium carbonyls, alone or together, are catalyst^^.^.

+

+

+

(R. W. WEGMAN)

380

14.6. Carbon Monoxide Reactions 14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.5. Reductive Carbonylation of Esters

1. H. Dumas, J. Levisalles, H. Rudler, J . Orgunornetul. Chem., 239, 117, (1979). 2. K. Moloy, R. W. Wegman J . Chem. SOC.,Chem. Commun., 820, (1988). 3. K. G . Moloy, R. W. Wegman, Orgunornetullics, 8, 2883, (1989). 4. M. J. Chen, H. M. Feder, J. W. Rathke, J . Am. Chem. SOC.,104,7346, (1982). 5 . Mitsubishi Gas, J 52073804 (1975). 6 . Mitsubishi Gas, GB 2087393 (1980).

14.6.5.5. Reductive Carbonylatlon of Esters

-

Reaction of a methyl ester with synthesis gas and a Co-based catalyst leads to acetaldehyde': RC(O)OCH, H, CO CH,C(O)H RC(0)OH (a) With methyl acetate at 200°C and 34.5 MPa and a Co-LiI-NPh, catalyst the acetaldehyde yield is nearly quantitative. In the reductive carbonylation of CH30H, the product mixture contains a wide variety of compounds including methyl acetate, acetic acid, 1,l-dimethoxyethane and copious amounts of H,O. Separating acetaldehyde from this mixture is difficult. In contrast, with a methyl ester feedstock, the product mixture is anhydrous and acetaldehyde is readily distilled from RC(0)OH. RC(0)OH is esterified in a separate step and recycled. The simplified reaction mechanism for methyl acetate is: CH31 CH,C(O)OLi CH,C(O)OCH, + LiI (b)

+

CH,I

+

+ CO + H, + cobalt catalyst HI + CH,C(O)OLi

+

--

+

CH,C(O)H

LiI

+ HI

+ CH3COOH

(c) (4

Since the carboxylate half of the ester is not involved in the catalytic chemistry with Co, any methyl ester which can be cleaved by LiI should show activity. Indeed, acetaldehyde and the corresponding carboxylic acid were obtained with methyl isobutyrate, dimethyl malonate, dimethyl succinate, methyl propionate, and dimethyl phthalate,. A lower operating P of 10.3 MPa is possible with a rhodium-cobalt-lithium-iodide catalyst'. In this case, the chemistry changes to: CO Rh/LiI catalyst CH,C(O)OC(O)CH, CH,C(O)OCH, (el

+

CH,C(O)OC(O)CH,

+

+ H, + Co catalyst

--

CH,C(O)H

+ CH,C(O)OH

(f)

Acetic anhydride is formed by Rh catalyzed carbonylation of methyl acetate followed by hydrogenation via the cobalt catalyst to acetaldehyde and acetic acid. The selectivity to acetaldehyde is 85%. Ethylidene diacetate is a by-product resulting from the reaction of acetaldehyde with acetic anhydride. A similar example involves a catalyst consisting of palladium(I1) acetate, CH31and trib~tylphosphine~. In this case, the acetaldehyde selectivity is 81% at 160"C, 1.7 MPa CO and 0.4 MPa H,. Ethyl acetate is obtained from methyl acetate if the reductive carbonylation is carried out with a catalyst capable of in situ hydrogenation of acetaldehyde to ethanol. The reaction sequence is: CH,C(O)OCH, H, CO CH,C(O)H CH,C(O)OH (g)

+

CH,C(O)H CH,CH,OH

+

+ H2

+ Ch,C(O)OH

--

-

+

CH,CH,OH CH3C(0)OCH,CH3

+ H,O

(h)

(9


380

14.6. Carbon Monoxide Reactions 14.6.5. Carbonylation and Reductive Carbonylation 14.6.5.5. Reductive Carbonylation of Esters

1. H. Dumas, J. Levisalles, H. Rudler, J . Orgunornetul. Chem., 239, 117, (1979). 2. K. Moloy, R. W. Wegman J . Chem. SOC.,Chem. Commun., 820, (1988). 3. K. G . Moloy, R. W. Wegman, Orgunornetullics, 8, 2883, (1989). 4. M. J. Chen, H. M. Feder, J. W. Rathke, J . Am. Chem. SOC.,104,7346, (1982). 5 . Mitsubishi Gas, J 52073804 (1975). 6 . Mitsubishi Gas, GB 2087393 (1980).

14.6.5.5. Reductive Carbonylatlon of Esters

-

Reaction of a methyl ester with synthesis gas and a Co-based catalyst leads to acetaldehyde': RC(O)OCH, H, CO CH,C(O)H RC(0)OH (a) With methyl acetate at 200°C and 34.5 MPa and a Co-LiI-NPh, catalyst the acetaldehyde yield is nearly quantitative. In the reductive carbonylation of CH30H, the product mixture contains a wide variety of compounds including methyl acetate, acetic acid, 1,l-dimethoxyethane and copious amounts of H,O. Separating acetaldehyde from this mixture is difficult. In contrast, with a methyl ester feedstock, the product mixture is anhydrous and acetaldehyde is readily distilled from RC(0)OH. RC(0)OH is esterified in a separate step and recycled. The simplified reaction mechanism for methyl acetate is: CH31 CH,C(O)OLi CH,C(O)OCH, + LiI (b)

+

CH,I

+

+ CO + H, + cobalt catalyst HI + CH,C(O)OLi

+

--

+

CH,C(O)H

LiI

+ HI

+ CH3COOH

(c) (4

Since the carboxylate half of the ester is not involved in the catalytic chemistry with Co, any methyl ester which can be cleaved by LiI should show activity. Indeed, acetaldehyde and the corresponding carboxylic acid were obtained with methyl isobutyrate, dimethyl malonate, dimethyl succinate, methyl propionate, and dimethyl phthalate,. A lower operating P of 10.3 MPa is possible with a rhodium-cobalt-lithium-iodide catalyst'. In this case, the chemistry changes to: CO Rh/LiI catalyst CH,C(O)OC(O)CH, CH,C(O)OCH, (el

+

CH,C(O)OC(O)CH,

+

+ H, + Co catalyst

--

CH,C(O)H

+ CH,C(O)OH

(f)

Acetic anhydride is formed by Rh catalyzed carbonylation of methyl acetate followed by hydrogenation via the cobalt catalyst to acetaldehyde and acetic acid. The selectivity to acetaldehyde is 85%. Ethylidene diacetate is a by-product resulting from the reaction of acetaldehyde with acetic anhydride. A similar example involves a catalyst consisting of palladium(I1) acetate, CH31and trib~tylphosphine~. In this case, the acetaldehyde selectivity is 81% at 160"C, 1.7 MPa CO and 0.4 MPa H,. Ethyl acetate is obtained from methyl acetate if the reductive carbonylation is carried out with a catalyst capable of in situ hydrogenation of acetaldehyde to ethanol. The reaction sequence is: CH,C(O)OCH, H, CO CH,C(O)H CH,C(O)OH (g)

+

CH,C(O)H CH,CH,OH

+

+ H2

+ Ch,C(O)OH

--

-

+

CH,CH,OH CH3C(0)OCH,CH3

+ H,O

(h)

(9

14.6. Carbon Monoxide Reactions 14.6.6. Oxidation and Reduction of CO

381

The catalysts require an Ruthenium, Rh-Ru, and Co-Ru catalysts have been iodide promoter; the reaction is typically carried out at >2OO0C and >30 MPa. (R. W. WEGMAN)

R. W. Wegman, D. C. Busby, J. Mol. Cat., 261, 39, (1987). R. W. Wegman, D. C . Busby, J. B. Letts, ACS Symp. Ser. 328, 125, (1987). R. V. Porcelli, British Patent Appln. 2,038,829A (1980). G. Braca, L. Paladini, G. Sbrana, G . Valentini, G . Andrich, S. Gregorior, Ind. Eng. Chem., Prod. Res. Dev., 115, 20, (1981). 5. E. Drent, J. Mol. Cat., 93, 37, (1986). 6. H. Kheradmand, A. Kiennemann, G. Jenner, J. Organomet. Chem., 251, 339, (1983).

1. 2. 3. 4.

14.6.6. Oxidation and Reduction of CO Carbon monoxide oxidation and reduction reactions invariably involve a change in the oxidation state of carbon. The only oxidation product of CO is CO, (or carbonate if carried out in base), and it is produced in a number of different reactions which rely on CO as a reducing agent. One of these is the water gas shift (WGS) reaction, equation (a). CO

+ HZO

COZ

+ H,

(a)

Reduction of CO to compounds containing C-H and/or C-C bonds has been actively studied because these reductions are important in the conversion of coal-derived CO into fuels and organic chemicals. Reactions in this class include methanation, MeOH synthesis, and Fischer-Tropsch (F-T) synthesis, e.g., equations (b)-(d); equation (d) yields a range of hydrocarbon and oxygenate (ROH and polyol) products.

+ 3 H3 CO + 2 H,

CO nCO

+ 2n H,

CH4

+ H,O

(b)

CH3OH (-CH,-)n

(c)

+ n H,O + oxygenates

(dl

Heterogeneous catalysts for equations (a)-(d) are known, and discussions of them can be found in many sources1-’. In this section, homogeneous reactions are presented in which both CO substrate and promoter complex exist in the solution phase. Reactions may or may not be catalytic with respect to the metal complex. For reductions of CO, the selectivity of product formation is an important consideration. Besides reactions in which H, is the reducing agent, there exist several reactions in which an active hydride (BH4-, BHEt,-, etc.) serves as the reductant. These are usually stoichiometric in metal complex. The following sections contain detailed discussions of CO oxidations and reductions in solution phase. (R. EISENBERG, C. KUBIAK) 1. Comprehensive discussions of CO/H, reactions, methanation and MeOH synthesis are given in Encyclopedia of Chemical Technology, R. E. Kirk, D. Othmer, eds., 2nd ed., Wiley-Interscience, New York, 1963-1970. For CO/H, reactions, see Vol. 4, p. 446; for methanation, see Vol. 13, p. 364; for MeOH synthesis, see Vol. 13, p. 370. 2. I. Wender, Catal. Rev.-Sci.Eng., 14, 97 (1976). 3. M. A. Vannice, Catal. Rev.-Sci. Eng., 14, 153 (1976).


14.6. Carbon Monoxide Reactions 14.6.6. Oxidation and Reduction of CO

381

The catalysts require an Ruthenium, Rh-Ru, and Co-Ru catalysts have been iodide promoter; the reaction is typically carried out at >2OO0C and >30 MPa. (R. W. WEGMAN)

R. W. Wegman, D. C. Busby, J. Mol. Cat., 261, 39, (1987). R. W. Wegman, D. C . Busby, J. B. Letts, ACS Symp. Ser. 328, 125, (1987). R. V. Porcelli, British Patent Appln. 2,038,829A (1980). G. Braca, L. Paladini, G. Sbrana, G . Valentini, G . Andrich, S. Gregorior, Ind. Eng. Chem., Prod. Res. Dev., 115, 20, (1981). 5. E. Drent, J. Mol. Cat., 93, 37, (1986). 6. H. Kheradmand, A. Kiennemann, G. Jenner, J. Organomet. Chem., 251, 339, (1983).

1. 2. 3. 4.

14.6.6. Oxidation and Reduction of CO Carbon monoxide oxidation and reduction reactions invariably involve a change in the oxidation state of carbon. The only oxidation product of CO is CO, (or carbonate if carried out in base), and it is produced in a number of different reactions which rely on CO as a reducing agent. One of these is the water gas shift (WGS) reaction, equation (a). CO

+ HZO

COZ

+ H,

(a)

Reduction of CO to compounds containing C-H and/or C-C bonds has been actively studied because these reductions are important in the conversion of coal-derived CO into fuels and organic chemicals. Reactions in this class include methanation, MeOH synthesis, and Fischer-Tropsch (F-T) synthesis, e.g., equations (b)-(d); equation (d) yields a range of hydrocarbon and oxygenate (ROH and polyol) products.

+ 3 H3 CO + 2 H,

CO nCO

+ 2n H,

CH4

+ H,O

(b)

CH3OH (-CH,-)n

(c)

+ n H,O + oxygenates

(dl

Heterogeneous catalysts for equations (a)-(d) are known, and discussions of them can be found in many sources1-’. In this section, homogeneous reactions are presented in which both CO substrate and promoter complex exist in the solution phase. Reactions may or may not be catalytic with respect to the metal complex. For reductions of CO, the selectivity of product formation is an important consideration. Besides reactions in which H, is the reducing agent, there exist several reactions in which an active hydride (BH4-, BHEt,-, etc.) serves as the reductant. These are usually stoichiometric in metal complex. The following sections contain detailed discussions of CO oxidations and reductions in solution phase. (R. EISENBERG, C. KUBIAK) 1. Comprehensive discussions of CO/H, reactions, methanation and MeOH synthesis are given in Encyclopedia of Chemical Technology, R. E. Kirk, D. Othmer, eds., 2nd ed., Wiley-Interscience, New York, 1963-1970. For CO/H, reactions, see Vol. 4, p. 446; for methanation, see Vol. 13, p. 364; for MeOH synthesis, see Vol. 13, p. 370. 2. I. Wender, Catal. Rev.-Sci.Eng., 14, 97 (1976). 3. M. A. Vannice, Catal. Rev.-Sci. Eng., 14, 153 (1976).

14.6. Carbon Monoxide Reactions 14.6.6. Oxidation and Reduction of CO 14.6.6.1. Oxidation

382

4. G. A. Mills, F. W. Steffgen, Catal. Rev., 8, 159 (1973). 5. H. H. Storch, H. Golumbic, R. B. Anderson, The Fischer-Tropsch and Related Syntheses. Wiley, New York, 1951. 6. H. Pichler, H. Schulz, Chew.-1ng.-Tech.,42, 1162 (1970). 7. Catalyst Handbook. Springer-Verlag, Berlin, 1970.

14.6.6.1. Oxidation

On a thermodynamic basis, CO is a stronger reducing agent than H,, as indicated by equation (a)'. CO,

+ 2 Hf + 2e-

CO

+ H,O E?(CO,/CO)

= -0.106 V vs NHE (pH = 0)

(a)

However, for kinetic considerations CO generally requires a transition metal catalyst for its potential as a reducing agent to be realized. Indeed, use of CO as a reductant with consequent formation of CO, is well documented in transition metal chemistry and represents one of the principal methods for synthesizing metal carbonyl complexes2. If the reaction is carried out in aqueous base, then carbonate or bicarbonate results as the CO oxidation product. Several representative reactions are ~ h o w n ~Further - ~ . examples are given in ref. 2. 2 CoC1,

+

12 KOH

+

11 CO

RhC13.xHZO

NiS

-

+ 3 K&03 + 4 KCl + 6 H,O [RhCl,(CO),]- + CO, Ni(CO), + C032- + Sz- + 2 H20

2 K[Co(CO),]

(b)

+ CO

(c)

+ 5 CO + 4 OH-

-+

(d)

Conversion of NO to N20 via equation (e) is another example of CO's reducing ability, but again the reaction requires a catalyst. One such catalyst system 2 NO

+ CO -+

N,O

+ CO,

(e)

is [RhCl,(CO),]- in aqueous acidic ethanol6; a second is based on PdC1,-CuCl, in 2 M HC17. Catalyst lifetime is limited in this case. Reduction of Se to H,Se using CO H,O is an unusual example of an uncatalyzed CO oxidation which occurs readily at 25"C/1 atm8. This reaction, equation (f), which may be important in developing hydrogenation cycles using CO H,O,

+

Se

+ CO + H,O

U

+

H2Se + CO,

may proceed via SeCO and HSeCOOH. Early examples of the CO oxidation in aqueous solution involve use of metal ions such as Hg(I1) or Ag(1) as oxidantsg. Oxidation of CO by Fe(CN):-, equation (g), is catalyzed by [Co(CO)(CN),(PEt,),] - in alkaline solution". CO

+ 2 Fe(CN)Z- + 3 OH-

HCO,-

+ 2 Fe(CN):-

+ H,O

(g)

A key intermediate is a cobalt hydroxycarbonyl species, Co-COOH, formed by OHattack on an activated, Co(II1)-coordinatedCO ligand.



382

4. G. A. Mills, F. W. Steffgen, Catal. Rev., 8, 159 (1973). 5. H. H. Storch, H. Golumbic, R. B. Anderson, The Fischer-Tropsch and Related Syntheses. Wiley, New York, 1951. 6. H. Pichler, H. Schulz, Chew.-1ng.-Tech.,42, 1162 (1970). 7. Catalyst Handbook. Springer-Verlag, Berlin, 1970.

14.6.6.1. Oxidation

On a thermodynamic basis, CO is a stronger reducing agent than H,, as indicated by equation (a)'. CO,

+ 2 Hf + 2e-

CO

+ H,O E?(CO,/CO)

= -0.106 V vs NHE (pH = 0)

(a)

However, for kinetic considerations CO generally requires a transition metal catalyst for its potential as a reducing agent to be realized. Indeed, use of CO as a reductant with consequent formation of CO, is well documented in transition metal chemistry and represents one of the principal methods for synthesizing metal carbonyl complexes2. If the reaction is carried out in aqueous base, then carbonate or bicarbonate results as the CO oxidation product. Several representative reactions are ~ h o w n ~Further - ~ . examples are given in ref. 2. 2 CoC1,

+

12 KOH

+

11 CO

RhC13.xHZO

NiS

-

+ 3 K&03 + 4 KCl + 6 H,O [RhCl,(CO),]- + CO, Ni(CO), + C032- + Sz- + 2 H20

2 K[Co(CO),]

(b)

+ CO

(c)

+ 5 CO + 4 OH-

-+

(d)

Conversion of NO to N20 via equation (e) is another example of CO's reducing ability, but again the reaction requires a catalyst. One such catalyst system 2 NO

+ CO -+

N,O

+ CO,

(e)

is [RhCl,(CO),]- in aqueous acidic ethanol6; a second is based on PdC1,-CuCl, in 2 M HC17. Catalyst lifetime is limited in this case. Reduction of Se to H,Se using CO H,O is an unusual example of an uncatalyzed CO oxidation which occurs readily at 25"C/1 atm8. This reaction, equation (f), which may be important in developing hydrogenation cycles using CO H,O,

+

Se

+ CO + H,O

U

+

H2Se + CO,

may proceed via SeCO and HSeCOOH. Early examples of the CO oxidation in aqueous solution involve use of metal ions such as Hg(I1) or Ag(1) as oxidantsg. Oxidation of CO by Fe(CN):-, equation (g), is catalyzed by [Co(CO)(CN),(PEt,),] - in alkaline solution". CO

+ 2 Fe(CN)Z- + 3 OH-

HCO,-

+ 2 Fe(CN):-

+ H,O

(g)

A key intermediate is a cobalt hydroxycarbonyl species, Co-COOH, formed by OHattack on an activated, Co(II1)-coordinatedCO ligand.


383

The importance of hydroxycarbonyl intermediates is well illustrated in a recent study of the stepwise oxidation of CO to CO, by binuclear ruthenium complexes”. Deprotonation of a diruthenium(1) aquo species yields a hydroxy intermediate which rearranges to an isolable hydroxycarbonyl complex, equation (h). Deprotonation of the hydroxycarbonyl with NEt, in dichloromethane results in a formally diruthenium(0) p-CO, complex, equation (i).

Et

The corresponding base form of a hydroxycarbonyl, obtained by deprotonation, is generally called a “rnetallocarboxylate.” Several other metallocarboxylates have now been isolated, e.g., addition of excess hydroxide to [(~5-C,H,)Fe(CO),(PPh,)l[BF4] affords the series of metallocarboxylates (q5-C,H,)Fe(CO)(PPh3)COO-M+ (M = Li, Na, K)”. Air oxidation of CO to CO, has been much studied in the context of automotive emission control. Studies of this reaction, equation (j), using heterogeneous catalysts have been r e ~ i e w e d ’ ~ , ’ ~ .

2 co

+ 0,

-

2c0,

(j)

14.6.6. Oxidation and Reduction of CO 14.6.6.2. In the Water Gas Shift Reaction 14.6.6.2.1. General Aspects.

384

Homogeneous catalysis of this reaction has also been reported using solutions of Rh6(CO),, at 80°C under 15-20 atm of the reactant gases in the proper stoichiometric ratioI5. Oxidation of CO is also important in fuel cell applications. By combining the half reaction for the CO,/CO couple, equation (a), with electrochemical reduction of 0,, fuel cells may achieve a maximum open circuit potential given by IEl(CO,/CO) EI(O,/H,O)I = 1.34l6.l7.In practice, electrocatalysts are required to lessen the normally high kinetic overpotentials for electrodic CO oxidation. An example of a CO/O, fuel cell which operates at the relatively low T of 80°C by employing [Rh(CO),Br,] - as the electrocatalyst and the RhlIm couple to mediate CO oxidation is shown below's.

&=-A

graphite anode

HZO Nafion' 117 cell separator

platinum cathode (R. EISENBERG, C.KUBIAK)

1. J. P. Collin, J. P. Sauvage, Coordination Chem. Rev., 93, 245 (1989); P. H. Rieger, Electrochemistry, Prentice-Hall, Englewood Cliffs, N.J., 1987, p. 452. 2. F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1968, p. 1. 3. (a) A. A. Blanchard, J. R. Rafter, W. B. Adams, J . Am. Chem. Soc., 56, 16 (1934); (b) G. W. Coleman, A. A. Blandhard, J . Am. Chem. Soc., 58, 2160 (1958). 4. (a) B. R. James, G. L. Rempel, F. T. T. Ng, J . Chem. SOC.A, 2454 (1969); (b) B. R. James, G. L. Rempel, J. Chem. SOC. A, 78 (1969). 5 . H. Behrens, E. Eisenmann, Z. Anorg, Allgem Chem., 278, 155 (1955). 6. C. D. Meyer, R. Eisenberg, J. Am. Chem. SOC., 98, 1364 (1976). 7. M. Kubota, K. J. Evans, C. A. Koerntgen, J. C. Marters, Jr., J . Am. Chem. Soc., 100, 342 ( 1978). 8. N. Sonoda, K. Kondo, K. Nagano, N. Kambe, F. Morimoto, Angew. Chem. lnt. Ed., 19, 308 (1980). 9. J. Halpem, Comm.lnorg. Chem., I , 3 (1981). 10. J. E. Bercaw, L.-Y. Goh, J. Halpern, J . Am. Chem. Soc., 94,6534 (1972). 11. J. S . Field, R. J. Haines, J. Sundermeyer, S. F. Woollam, J . Chem. Soc., Chern. Commun., 985 (1990). 12. D. H. Gibson, T . 4 . Ong, Y. Ming, Organometallics, 10, 950 (1991). 13. T. Engel, G. Ertl, Adv. Catal., 28, 2 (1979). 14. K. C. Taylor, Catal. Sci. Technol., 5, 119 (1984). 15. G . D. Mercer, W. B. Beaulieu, D. M. Roundhill, J. Am. Chem. Soc., 99, 6551 (1977). 16. J. 0. M. Bockris, S. Srinivasan, Fuel Cells: Their Electrochemistry. McGraw-Hill, New York, 1969. 17. W. Vielstich, Fuel Cells. Wiley-Interscience, New York, 1970. 18. J. Wu, C. P. Kubiak, J . Am. Chem. Soc., 105,7456 (1983).

14.6.6.2. In the Water Gas Shift Reaction 14.6.6.2.1. General Aspects.

Numerous systems for catalyzing the water gas shift (WGS) reaction homogeneously have been reported. These operate for the most part under milder conditions than those



384

Homogeneous catalysis of this reaction has also been reported using solutions of Rh6(CO),, at 80°C under 15-20 atm of the reactant gases in the proper stoichiometric ratioI5. Oxidation of CO is also important in fuel cell applications. By combining the half reaction for the CO,/CO couple, equation (a), with electrochemical reduction of 0,, fuel cells may achieve a maximum open circuit potential given by IEl(CO,/CO) EI(O,/H,O)I = 1.34l6.l7.In practice, electrocatalysts are required to lessen the normally high kinetic overpotentials for electrodic CO oxidation. An example of a CO/O, fuel cell which operates at the relatively low T of 80°C by employing [Rh(CO),Br,] - as the electrocatalyst and the RhlIm couple to mediate CO oxidation is shown below's.

&=-A

graphite anode

HZO Nafion' 117 cell separator

platinum cathode (R. EISENBERG, C.KUBIAK)

1. J. P. Collin, J. P. Sauvage, Coordination Chem. Rev., 93, 245 (1989); P. H. Rieger, Electrochemistry, Prentice-Hall, Englewood Cliffs, N.J., 1987, p. 452. 2. F. Calderazzo, R. Ercoli, G. Natta, in Organic Syntheses via Metal Carbonyls, I. Wender, P. Pino, eds., Wiley-Interscience, New York, 1968, p. 1. 3. (a) A. A. Blanchard, J. R. Rafter, W. B. Adams, J . Am. Chem. Soc., 56, 16 (1934); (b) G. W. Coleman, A. A. Blandhard, J . Am. Chem. Soc., 58, 2160 (1958). 4. (a) B. R. James, G. L. Rempel, F. T. T. Ng, J . Chem. SOC.A, 2454 (1969); (b) B. R. James, G. L. Rempel, J. Chem. SOC. A, 78 (1969). 5 . H. Behrens, E. Eisenmann, Z. Anorg, Allgem Chem., 278, 155 (1955). 6. C. D. Meyer, R. Eisenberg, J. Am. Chem. SOC., 98, 1364 (1976). 7. M. Kubota, K. J. Evans, C. A. Koerntgen, J. C. Marters, Jr., J . Am. Chem. Soc., 100, 342 ( 1978). 8. N. Sonoda, K. Kondo, K. Nagano, N. Kambe, F. Morimoto, Angew. Chem. lnt. Ed., 19, 308 (1980). 9. J. Halpem, Comm.lnorg. Chem., I , 3 (1981). 10. J. E. Bercaw, L.-Y. Goh, J. Halpern, J . Am. Chem. Soc., 94,6534 (1972). 11. J. S . Field, R. J. Haines, J. Sundermeyer, S. F. Woollam, J . Chem. Soc., Chern. Commun., 985 (1990). 12. D. H. Gibson, T . 4 . Ong, Y. Ming, Organometallics, 10, 950 (1991). 13. T. Engel, G. Ertl, Adv. Catal., 28, 2 (1979). 14. K. C. Taylor, Catal. Sci. Technol., 5, 119 (1984). 15. G . D. Mercer, W. B. Beaulieu, D. M. Roundhill, J. Am. Chem. Soc., 99, 6551 (1977). 16. J. 0. M. Bockris, S. Srinivasan, Fuel Cells: Their Electrochemistry. McGraw-Hill, New York, 1969. 17. W. Vielstich, Fuel Cells. Wiley-Interscience, New York, 1970. 18. J. Wu, C. P. Kubiak, J . Am. Chem. Soc., 105,7456 (1983).

14.6.6.2. In the Water Gas Shift Reaction 14.6.6.2.1. General Aspects.

Numerous systems for catalyzing the water gas shift (WGS) reaction homogeneously have been reported. These operate for the most part under milder conditions than those


385

employed by known heterogeneous catalysts for the shift reaction. Various preparations of the latter use Co and Fe oxides [so-called high T (350°C) shift catalysts] or Cu/ZnO [low T (250°C) shift catalysts]'. The homogeneous shift catalysts are summarized in Table 1. Because of different operating conditions, direct comparison of relative reactivities of these catalysts is not possible. Required co-factors, solvent media, operating conditions, and approximate turnover rates are included where possible. Catalysis of the shift reaction by complexes in solution is based on equations (a) and (b).

M("-2)+

+ H+

M"+-H

H+

M"+

___z

+ H,

0)

The former corresponds to reduction of the metal center by CO with the two paths evolving CO, showing the stoichiometric equivalence of a reduced metal plus proton with a metal hydride. The latter reaction corresponds to proton reduction. Clearly, (a) and (b) are influenced by the relative redox ability of the metal, which in turn is affected by the ligand environment of the complex. Proton concentration will also have an effect; low pH favors reaction (b), while high pH promotes equation (a) through nucleophilic attack by OH- rather than by H,O on coordinated CO to yield the M-COOH species penultimate to CO, formation. It is not surprising that homogeneous WGS catalysts are in two categories-those which operate in acidic and those in basic media. Of the acid-based systems, the most active are the rhodium carbonyl iodide corn bin at ion^^-^, the PtC1,- /SnCl, preparation5, and the system based on the ruthenium carbonyl cluster catalyst precursors Ru,(CO),, and H4Ru4(C0),,637.The Rh carbonyl iodide system under more vigorous conditions ( 185"C, 23 atm') shows a catalytic rate of 400 turnovers/h. The catalyst systems in base are functional by virtue of the fact that bicarbonate produced from CO, OH- is unstable under the reaction conditions (T and medium), thus permitting the systems to be truly catalytic in base. The most active catalysts in basic media appear to be Ru,(CO),,, Rh,(CO),,, and RhCl, in aqueous picoline9*10, although the different conditions render direct comparisons impossible. There is an additional consideration in catalysis of the shift reaction in base which arises because of equilibrium (c).

+

CO

+ OH-

HCOO-

(c)

Catalysis by certain systems may actually proceed through the intermediacy of formate ion and its decomposition. This may be the case in at least the M(CO), (M = Cr, Mo, W) systems' ' * I 2 . A 1970 patent describes shift catalysis under more forcing conditions (56 atm, 200°C) using a variety of metal salts and organic amines as basesI3. (R. EISENBERG, C. KUBIAK) 1. 2. 3. 4.

Catalyst Handbook. Springer-Verlag, West Berlin, 1970. C.-H. Cheng, D. E. Hendriksen, R. Eisenberg, J . Am. Chem. Soc., 99, 2791 (1977). R. Eisenberg, C.-H. Cheng, U.S.Patent 4,107,076 (1978); Chem. Abst., 90, 57437k (1979). E. C. Baker, D. E. Hendriksen, R. Eisenberg, J . Am. Chem. Soc., 102, (1980).

WATERGASSHIFTREACTION

HI/HOAc HCI/HOAc H,SO,/diglyme KOH/ethoxyethanol Piperidine/ethoxy -ethanol Pyridine NMe,/THF NMe,/THF NMe,/THF NMe,/THF NaOH/n-BuOH KOH/MeOH KOH/MeOH KOH/MeOH Acetone Acetone PTSA/LiCl/n-PrOH H2O 4-F'icoline-H20

Prornoters/solvent

THE

CReactionnot necessarily catalytic in base. dEstimatedfrom data given for 18 h runs. 'dppm, bis(dipheny1phosphino)meth~e; PTSA, p-toluenesulfonic acid. 'Original number reported for 5 h experiment in error. ~dmpm.bis(dimethy1phosphino)meth~e.

bPressure of CO before heating to the reaction temperature.

"Estimatedfrom data given for 10 h runs.

RhI2(CO)2PtCL2-/SKI., Ru3(c0)12 Ru,(CO)i2 Ru3(CO),,/Fe(CO), H4Ru4(CO)I ,/Fe(CO)5 Rlb(c0),6 Ru3(co)12 1r4(C0)12 [BU4NI[R3(CO)61 Fe(CO), Cr(CO)6 Mo(CO), W(C0)6 PttP(i-W313 l"Et3)3 [Rh2(~--CO)(CO~(d~~m)2~ Pd,Cl,(dmprn),g RhCI,

Complex

TABLE1. HOMOGENEOUS CATALYSTS FOR 0.5 0.5 0.9 0.9 0.9 0.9 23.3 23.3 23.3 23.3 34.0b 27.2b 27.2b 27.2b 20.0b 20.0b 1.o 1.O 0.9

pco(atm)

88 100 100 100 100 150 100 150 125 180 180 160 160 153 100 90 77 100

90

T ("C)

~~

~~

~

1.0 1.O 1.4 -.I 1.3 0.5-1.3 170" 330" 30 70 83.3 63" 7.3" 9.9 19.2d 0.6d 0.Sf 1 4.1

Turnovers/h

2-4 5 67 14 6 6 15 15 15 15 16,17 11,12,16 11,12,16 11,12,16 18 18 19,20 21 9

Ref.

14.6.6. Oxidation and Reduction of CO 14.6.6.2. In the Water Gas Shift Reaction 14.6.6.2.2. Applications of the Reaction.

387

5. C.-H. Cheng, R. Eisenberg, J . Am. Chem. SOC., 100,5968 (1978). 6. P. C. Ford, R. G . Rinker, C. Ungermann, R. M. Laine, V. Landis, S. A. Moya, J. Am. Chem. SOC.,100,4595 (1978). 7. P. Yarrow, H. Cohen, C. Ungermann, D. Vandenberg, P. C. Ford, R. G. Rinker,J. Mol. Catul., 22, 239 (1983). 8. T. C. Singleton, L. J. Park, J. L. Price, D. Forster, ACS Petroleum Division Preprints, 24, 329 (1979). 9. A. J. Pardey, P. C. Ford, J . Mol. Catul., 53, 247 (1989). 10. B. S. L. Neto, K. H. Ford, A. Pardey, R. G. Rinker, P. C. Ford, Inorg. Chem., 30, 3837. 11. A. D. King, Jr., R. B. King, D. B. Yang, J. Am. Chem. Soc., 103,2699 (1981). 12. A. D. King, Jr., R. B. King, E. L. Sailers, 111, J . Am. Chem. SOC., 103, 1867 (1981). 13. D. M. Fenton, U.S. Patent 3,539,298 (1970); Chem. Abst., 94, 14642a (1971). 14. R. M. Laine, R. G. Rinker, P. C. Ford, J . Am. Chem. Soc., 99,252 (1977). 15. H. C. Kang, C. H. Mauldin, T. Cole, W. Slegeir, K. Cann, R. Pettit, J . Am. Chem. Soc., 99, 8323 (1977). 16. R. B. King, C. C. Frazier, R. M. Hanes, A. D. King, Jr.,J. Am. Chem. SOC., 100,2925 (1978). 17. A. D. King, Jr., R. B. King, D. B. Yang,J. Am. Chem. SOC., 102, 1028 (1980). 100, 3941 (1978). 18. T. Yoshida, Y. Ueda, S . 0tsuka.J. Am. Chem. SOC., 19. C. P. Kubiak, R. Eisenberg, J . Am. Chem. SOC., 102, 3637 (1980). 20. R. M. Laine, J . Am. Chem. SOC., 100, 6451 (1978). 21. M. L. Kullberg, C. P. Kubiak, C, Mol. Chem., I (1984). 14.6.6.2.2. Applications of the Reaction.

+

The use of CO H,O as an in situ source of H, via the shift reaction or as a source of reducing electrons with concomitant oxidation of CO to CO, has recently been explored with homogeneous catalyst solutions. Equations (a) and (b) are modifications (Reppe) of hydrofonnylation (see 14.6.3) and olefin hydrogenation, respectively. The most effective catalysts for equation (a) are Ru,(CO),,, Rh6(C0),6, and 1r4(CO),, in alkaline THF or MeOH. The first of these shows great selectivity in the formation of linear vs. branched With Rh,(CO),,, the aldehyde is reduced further to the alcohol. A different catalyst based on Co,(CO),/diphos in polar ether solvents has also been used to catalyze equation (a) with propylene as substrate3. R +2CO+H20

RyH + CO,

0

Catalysis of equations (a) and (b) occurs using the PtCl,,- /SnCl,/HCl-HOAc shift catalyst system when ethylene is present as the olefin substrate4. The rate of catalysis is similar to that for the shift reaction using this catalyst (1.3 turnovers/h at 88°C) but no shift catalysis is observed with this system as long as ethylene is present in the reaction vessel. The Fe(CO),/KOH/MeOH water gas shift catalyst system also catalyzes the Reppe reaction with ethylene. The overall rate of reaction and the selectivity toward production of propionaldehyde vs. propanol were found to be quite sensitive to the nature of the base5. Activated olefins such as a,P-unsaturated ketones have also been reduced by CO H,O using homogeneous catalysts6. The electrons released in the oxidation of CO to CO, can also reduce more oxidized organic substrates, e.g., nitrobenzene:

+


14.6.6. Oxidation and Reduction of CO 14.6.6.2. In the Water Gas Shift Reaction 14.6.6.2.2. Applications of the Reaction.

387

5. C.-H. Cheng, R. Eisenberg, J . Am. Chem. SOC., 100,5968 (1978). 6. P. C. Ford, R. G . Rinker, C. Ungermann, R. M. Laine, V. Landis, S. A. Moya, J. Am. Chem. SOC.,100,4595 (1978). 7. P. Yarrow, H. Cohen, C. Ungermann, D. Vandenberg, P. C. Ford, R. G. Rinker,J. Mol. Catul., 22, 239 (1983). 8. T. C. Singleton, L. J. Park, J. L. Price, D. Forster, ACS Petroleum Division Preprints, 24, 329 (1979). 9. A. J. Pardey, P. C. Ford, J . Mol. Catul., 53, 247 (1989). 10. B. S. L. Neto, K. H. Ford, A. Pardey, R. G. Rinker, P. C. Ford, Inorg. Chem., 30, 3837. 11. A. D. King, Jr., R. B. King, D. B. Yang, J. Am. Chem. Soc., 103,2699 (1981). 12. A. D. King, Jr., R. B. King, E. L. Sailers, 111, J . Am. Chem. SOC., 103, 1867 (1981). 13. D. M. Fenton, U.S. Patent 3,539,298 (1970); Chem. Abst., 94, 14642a (1971). 14. R. M. Laine, R. G. Rinker, P. C. Ford, J . Am. Chem. Soc., 99,252 (1977). 15. H. C. Kang, C. H. Mauldin, T. Cole, W. Slegeir, K. Cann, R. Pettit, J . Am. Chem. Soc., 99, 8323 (1977). 16. R. B. King, C. C. Frazier, R. M. Hanes, A. D. King, Jr.,J. Am. Chem. SOC., 100,2925 (1978). 17. A. D. King, Jr., R. B. King, D. B. Yang,J. Am. Chem. SOC., 102, 1028 (1980). 100, 3941 (1978). 18. T. Yoshida, Y. Ueda, S . 0tsuka.J. Am. Chem. SOC., 19. C. P. Kubiak, R. Eisenberg, J . Am. Chem. SOC., 102, 3637 (1980). 20. R. M. Laine, J . Am. Chem. SOC., 100, 6451 (1978). 21. M. L. Kullberg, C. P. Kubiak, C, Mol. Chem., I (1984). 14.6.6.2.2. Applications of the Reaction.

+

The use of CO H,O as an in situ source of H, via the shift reaction or as a source of reducing electrons with concomitant oxidation of CO to CO, has recently been explored with homogeneous catalyst solutions. Equations (a) and (b) are modifications (Reppe) of hydrofonnylation (see 14.6.3) and olefin hydrogenation, respectively. The most effective catalysts for equation (a) are Ru,(CO),,, Rh6(C0),6, and 1r4(CO),, in alkaline THF or MeOH. The first of these shows great selectivity in the formation of linear vs. branched With Rh,(CO),,, the aldehyde is reduced further to the alcohol. A different catalyst based on Co,(CO),/diphos in polar ether solvents has also been used to catalyze equation (a) with propylene as substrate3. R +2CO+H20

RyH + CO,

0

Catalysis of equations (a) and (b) occurs using the PtCl,,- /SnCl,/HCl-HOAc shift catalyst system when ethylene is present as the olefin substrate4. The rate of catalysis is similar to that for the shift reaction using this catalyst (1.3 turnovers/h at 88°C) but no shift catalysis is observed with this system as long as ethylene is present in the reaction vessel. The Fe(CO),/KOH/MeOH water gas shift catalyst system also catalyzes the Reppe reaction with ethylene. The overall rate of reaction and the selectivity toward production of propionaldehyde vs. propanol were found to be quite sensitive to the nature of the base5. Activated olefins such as a,P-unsaturated ketones have also been reduced by CO H,O using homogeneous catalysts6. The electrons released in the oxidation of CO to CO, can also reduce more oxidized organic substrates, e.g., nitrobenzene:

+

14.6.6. Oxidation and Reduction of CO 14.6.6.3. In Reduction 14.6.6.3.1. To Formyl.

388

No,

NH,

The most effective catalysts for this reaction are the noble metal cluster carbonyls listed in Table 1, 14.6.6.2.1,under conditions similar to those tabulated7. In all of the foregoing reactions involving CO oxidation, actual water gas shift catalysis is not observed. Instead, it appears that a metal hydride forms which acts as the reducing agent, thereby making it unnecessary to liberate H, in order to carry out the reduction. (R. EISENBERG, C. KUBIAK) 1. H. C. Kang, C. H. Mauldin, T. Cle, W. Slegeir, K, Cann, R. Pettit, J . Am. Chem. Soc., 99, 8323 (1977). 2. R. M. Laine, J . Am. Chem. SOC., 100,6451 (1978). 3. K. Murata, A. Metsuda, K. Bando, Y. Sugi, J . Chem. Soc., Chem. Commun., 785 (1979). 4. C.-H. Cheng, L. Kuritzkes, R. Eisenberg, J . Orgunometul. Chem., 190, C21 (1980). 5. R. Massoudi, J. H. Kim, R. B. King, A. D. King, Jr., J . Am. Chem. Soc., 109,7428 (1987). 6. T. Kitamura, N. Sakamoto, T. Joh, Chem. Lett., 379 (1973). 7. K. Cann, T. Cole, W. Slegeir, R. Pettit, J . Am. Chem. Soc., 100, 3969 (1978).

14.6.6.3. In Reduction

This section focuses on stoichiometric reductions of coordinated CO. The product complexes will contain formyl, methyl, hydroxymethyl, or methoxy ligands which originated as CO in the starting complexes. Whereas these complex products may be synthesized by alternative methods using more reduced C-containing starting materials, the key formation feature outlined here is reduction of coordinated carbonyl. These reactions, besides producing new and interesting organometallic compounds, also provide insight into how CO can be reduced catalytically to fuels and organic chemicals using yetundeveloped homogeneous catalysts. 14.6.6.3.1. To Formyl.

Hydride donors such as BHEt,- and BH(0-i-Pr),- ions react with metal carbonyls to generate formyl complexes in situ: HBR,-

+ L#(CO)

\

H

The comdexes have been isolated in moderate yield for several cases, ex., Fe(CHO)(CO),-, Fe(CHO)(CO),[P(OPh,)] -, and Re,(CO),(CHO)- I-,. A key property of these complexes is their ability to serve as hydride donors, thereby reversing equation (a) with regard to the carbonyl ligand. Two neutral formyl complexes have been synthesized and isolated via equation (a), beginning with the cationic carbonyl compounds, CpRe(NO)(CO)L+ where L = CO and PPh34*5.These compounds are intermediates in the BH,- reduction of coordinated


14.6.6. Oxidation and Reduction of CO 14.6.6.3. In Reduction 14.6.6.3.1. To Formyl.

388

No,

NH,

The most effective catalysts for this reaction are the noble metal cluster carbonyls listed in Table 1, 14.6.6.2.1,under conditions similar to those tabulated7. In all of the foregoing reactions involving CO oxidation, actual water gas shift catalysis is not observed. Instead, it appears that a metal hydride forms which acts as the reducing agent, thereby making it unnecessary to liberate H, in order to carry out the reduction. (R. EISENBERG, C. KUBIAK) 1. H. C. Kang, C. H. Mauldin, T. Cle, W. Slegeir, K, Cann, R. Pettit, J . Am. Chem. Soc., 99, 8323 (1977). 2. R. M. Laine, J . Am. Chem. SOC., 100,6451 (1978). 3. K. Murata, A. Metsuda, K. Bando, Y. Sugi, J . Chem. Soc., Chem. Commun., 785 (1979). 4. C.-H. Cheng, L. Kuritzkes, R. Eisenberg, J . Orgunometul. Chem., 190, C21 (1980). 5. R. Massoudi, J. H. Kim, R. B. King, A. D. King, Jr., J . Am. Chem. Soc., 109,7428 (1987). 6. T. Kitamura, N. Sakamoto, T. Joh, Chem. Lett., 379 (1973). 7. K. Cann, T. Cole, W. Slegeir, R. Pettit, J . Am. Chem. Soc., 100, 3969 (1978).

14.6.6.3. In Reduction

This section focuses on stoichiometric reductions of coordinated CO. The product complexes will contain formyl, methyl, hydroxymethyl, or methoxy ligands which originated as CO in the starting complexes. Whereas these complex products may be synthesized by alternative methods using more reduced C-containing starting materials, the key formation feature outlined here is reduction of coordinated carbonyl. These reactions, besides producing new and interesting organometallic compounds, also provide insight into how CO can be reduced catalytically to fuels and organic chemicals using yetundeveloped homogeneous catalysts. 14.6.6.3.1. To Formyl.

Hydride donors such as BHEt,- and BH(0-i-Pr),- ions react with metal carbonyls to generate formyl complexes in situ: HBR,-

+ L#(CO)

\

H

The comdexes have been isolated in moderate yield for several cases, ex., Fe(CHO)(CO),-, Fe(CHO)(CO),[P(OPh,)] -, and Re,(CO),(CHO)- I-,. A key property of these complexes is their ability to serve as hydride donors, thereby reversing equation (a) with regard to the carbonyl ligand. Two neutral formyl complexes have been synthesized and isolated via equation (a), beginning with the cationic carbonyl compounds, CpRe(NO)(CO)L+ where L = CO and PPh34*5.These compounds are intermediates in the BH,- reduction of coordinated

14.6.6. Oxidation and Reduction of CO 14.6.6.3. In Reduction 14.6.6.3.2. To Alkyl, Hydroxyalkyl, Alkoxyl Ligands

389

CO to CH,, and can be used as starting compounds in the further reduction of the formyl ligand. The L = CO derivative also undergoes dimerization and subsequent methanolysis via equation (b)6. 3

e e

e

MeOH

RP

'0

Re

+

Re (b)

The (octaethylporphyrin) rhodium dimer [Rh(OEP)], reacts in the presence of H, and CO to produce the formyl species, Rh(0EP)CHO'. The rhodium(I1) porphyrins are the only systems to date to react with H, and CO to produce formyls directly'. Alternatively, the metalloformyl complex Rh(0EP)CHO can be prepared as9: Rh(0EP)H

+ CO

+ H, + 2 CO + H,O + 3 CO

[Rh(OEP)], [Rh(OEP)],

-

Rh(0EP)CHO

(c)

2 Rh(0EP)CHO

(4

2 Rh(0EP)CHO

+ CO,

(e) Different mechanisms for production of the rhodium formyls are followed in different reaction media. In benzene solution where [Rh(OEP)], and the metalloradical [Rh(OEP)]' are present, a radical chain reaction is involved in CO activation and production of the formyl complex'0,' In pyridine solution where [Rh(OEP)]' disproportionates to [Rh(OEP)(py),] + and [Rh(OEP)I - , interaction of the metalloanion [Rh(OEP)J- with CO is believed to be the dominant pathway for CO activation and production of the metalloformyl species by protonation".

'.

(R. EISENBERG, C. KUBIAK) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

C. P. Casey, S. M. Neumann, J . Am. Chem. SOC., 98,5395 (1976). C. P. Casey, S. M. Neumann, J . Am. Chem. SOC., 100,2544 (1978). J. A. Gladysz, W. Tam, J . Am. Chem. SOC., 100,2545 (1978). C. P. Casey, M. A. Andrews, J. E. Ring, J . Am. Chem. SOC., 101, 741 (1979). W. Tam, W. K. Wong, J. A. Gladysz, J . Am. Chem. SOC.,101, 1589 (1979). C. P. Casey, M. A. Andrews, D. R. McAlister, J . Am. Chem. SOC., 101, 3371 (1979). B. B. Wayland, B. A. Woods, J . Chem. SOC., Chem. Commun., 700 (1981). B. B. Wayland, S. L. Van Voorhees, C. Wilker, Znorg. Chem., 25,4039 (1986). B. B. Wayland, B. A. Woods, R. Pierce, J . Chem. Chem. SOC.,104, 302 (1982). B. B. Wayland, K. J. Balkus, Jr., M. D. Famos, Organometallics, 8, 950 (1989). R. S. Paonessa, N. C. Thomas, J. Halpem, J . Am. Chem. SOC., 107,4333 (1985).

14.6.6.3.2. To Alkyl, Hydroxyalkyl, and Alkoxyl Ligands and Reductlve Coupllng.

Conversion of a coordinated CO to a coordinated alkyl using active hydride sources has been reported for [CpM(CO),(PPh,)]+ where M = Mo, W, and for



389

CO to CH,, and can be used as starting compounds in the further reduction of the formyl ligand. The L = CO derivative also undergoes dimerization and subsequent methanolysis via equation (b)6. 3

e e

e

MeOH

RP

'0

Re

+

Re (b)

The (octaethylporphyrin) rhodium dimer [Rh(OEP)], reacts in the presence of H, and CO to produce the formyl species, Rh(0EP)CHO'. The rhodium(I1) porphyrins are the only systems to date to react with H, and CO to produce formyls directly'. Alternatively, the metalloformyl complex Rh(0EP)CHO can be prepared as9: Rh(0EP)H

+ CO

+ H, + 2 CO + H,O + 3 CO

[Rh(OEP)], [Rh(OEP)],

-

Rh(0EP)CHO

(c)

2 Rh(0EP)CHO

(4

2 Rh(0EP)CHO

+ CO,

(e) Different mechanisms for production of the rhodium formyls are followed in different reaction media. In benzene solution where [Rh(OEP)], and the metalloradical [Rh(OEP)]' are present, a radical chain reaction is involved in CO activation and production of the formyl complex'0,' In pyridine solution where [Rh(OEP)]' disproportionates to [Rh(OEP)(py),] + and [Rh(OEP)I - , interaction of the metalloanion [Rh(OEP)J- with CO is believed to be the dominant pathway for CO activation and production of the metalloformyl species by protonation".

'.

(R. EISENBERG, C. KUBIAK) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

C. P. Casey, S. M. Neumann, J . Am. Chem. SOC., 98,5395 (1976). C. P. Casey, S. M. Neumann, J . Am. Chem. SOC., 100,2544 (1978). J. A. Gladysz, W. Tam, J . Am. Chem. SOC., 100,2545 (1978). C. P. Casey, M. A. Andrews, J. E. Ring, J . Am. Chem. SOC., 101, 741 (1979). W. Tam, W. K. Wong, J. A. Gladysz, J . Am. Chem. SOC.,101, 1589 (1979). C. P. Casey, M. A. Andrews, D. R. McAlister, J . Am. Chem. SOC., 101, 3371 (1979). B. B. Wayland, B. A. Woods, J . Chem. SOC., Chem. Commun., 700 (1981). B. B. Wayland, S. L. Van Voorhees, C. Wilker, Znorg. Chem., 25,4039 (1986). B. B. Wayland, B. A. Woods, R. Pierce, J . Chem. Chem. SOC.,104, 302 (1982). B. B. Wayland, K. J. Balkus, Jr., M. D. Famos, Organometallics, 8, 950 (1989). R. S. Paonessa, N. C. Thomas, J. Halpem, J . Am. Chem. SOC., 107,4333 (1985).

14.6.6.3.2. To Alkyl, Hydroxyalkyl, and Alkoxyl Ligands and Reductlve Coupllng.

Conversion of a coordinated CO to a coordinated alkyl using active hydride sources has been reported for [CpM(CO),(PPh,)]+ where M = Mo, W, and for

390


’*’.

~

[CpRe(CO),(NO)] The product complexes are neutral methyl derivatives obtained in 30-80% yield. Reductions probably proceed through the formyl species. It has been shown3 that reaction of the formyl complex CpRe(CO)(NO)(CHO) with BH,.THF, which may be formed by BH,- after initial hydride donation, yields the corresponding methyl complex CpRe(CO)(NO)(CH,) in 80% isolated yield. Reduction of M-CO+ to the hydroxymethyl derivative M-CH,OH using active hydrides has also been reported, and represents part of the reduction process which culminates in M-CH, formation. +

Rh(0EP)H

+ CO FRh(0EP)CHO

(a) Specifically, equation (a) is observed with -45% isolated yield3g4 and the Re-CH,OH product of (b) undergoes further reaction with BH, to form the corresponding Re-CH, system.

-I+ Re

4 j \c No

Re

Na[AIH,Et,]/THF or NaBH,/THF-H,O

0

( b)

0

Reduction of CO to a methoxy ligand is one of the cleanest examples of metal complex promoted reduction of CO by H,. Equation (c) occurs nearly quantitatively5s6. The starting dicarbonyl forms readily from the dinitrogen complex (CP*,ZrN,),N;.

+ 2H2

2 atm

R r . . . H

+ co

__j

100°C

K

O

C

H

3

The great stability of the methoxyhydride product of eq. (c) precludes incorporation of this reaction into a catalytic cycle, although irreversible hydrolysis of the product yields MeOH. The mechanism of the reaction has been studied extensively’. A reductive coupling of CO has been seen as is shown in eq. (d)’

jg -

x’H s’H Zi-*.Hco

> -50°C

J g r . . . o ‘c=c /

H

H

/

\

A key feature appears to be the hydridic nature of Cp* ,ZrH, and its ability to donate H- to coordinated CO similar to that observed for BHR, -.


391

Organoactinide complexes of Cp* show reactivity with CO analogous to that found for the Zr system outlined above, e.g., equation (e), which proceeds in 50% isolated yield'. \

/

*co v

25T

Since the starting material is formed via CO insertion, equation (e) represents a reductive coupling of four COs to a C, backbone. Reductive coupling of CO has been examined both experimentally and theoretically. In general, four factors promote reductive coupling: (1) short nonbonded distances between CO ligands, which often result from high coordination numbers; (2) a molecular geometry in which molecular orbitals localized on the CO ligands to be coupled are favorably aligned; (3) an electron rich metal center; and (4) Lewis acid coordination to the carbonyl oxygen These factors all appear to contribute to the reductive coupling of the CO ligands of the complex Ta(CO),(PMe2CH,CH2PMez)2Cl, equation (f)'O.

Ta(CO),(dmpe),Cl

1. Mg/HgCl,(C,Me,),ZCl, 2. 2 Me,SiCl -78°C THF

Ta(Me,SiOC=COSiMe,)(dmpe),Cl

(f)

(dmpe = PMe2CH,CH2PMe2)

Radical activation of CO by Rh(OEP), discussed in 14.6.6.3.1,with respect to production of formyls also plays a role in reductive coupling. The weak Rh-Rh bond of the octaethylporphyrin dimer [Rh(OEP)], inserts either one or two CO molecules affording metalloketone (0EP)Rh-C(0)-Rh(0EP) or metallo-a-diketone (0EP)Rh-C(0)-C(0)-Rh(0EP)products. Higher CO pressures (Pco > 12 atm) and lower T favor the formation of the metallo-a-diketone which exists in equilibrium with the metalloketone''. (R. EISENBERG, C. KUBIAK)

1. 2. 3. 4. 5.

P. M. Treichel, R. L. Shubkin, Inorg. Chem., 6 , 1328 (1967). R. P. Stewart, N. Okamoto, W. A. G. Graham, J . Organomet. Chem., 42, C32 (1972). W. A. G. Graham, J. R. Sweet, J . Organornetal. Chem., 173, C9 (1979). C. P. Casey, M. A. Andrews, D. R. McAlister, J . Am. Chem. Soc., 101, 3371 (1979). J. M. Manriquez, D. R. McAlister, R. D. Sanner, J. E. Bercaw, J . Am. Chem. Soc., 98, 6733

(1976). 6. J. M. Manriquez, D. R. McAlister, R. D. Sanner, J. E. Bercaw, J . Am. Chem. SOC., 100,2716 (1978). 7. J. E. Bercaw, P. T. Wolczanski, Acct. Chem. Res., 13, 121 (1980), and references therein. 8. P. J. Fagan, J. M. Manriquez, T. J. Marks, V. W. Day, S.H. Vollmer, C. S. Day, J . Am. Chem. Soc., 102,5393 (1980).

392

14.6.6. Oxidation and Reduction of CO 14.6.6.3. In Reduction 14.6.6.3.3. To Alcohols and Alkanes.

9. R. Hoffmann, C. N. Wilker, S. J. Lippard, J. L. Templeton, D. C. Brower, J . Am. Chem. SOC., 105, 146 (1983). 10. P. A. Bianconi, I. D. Williams, M. P. Engeler, S. J. Lippard, J . Am. Chem. Soc., 108, 31 1 (1986). 11. V. L. Coffin, W. Brennen, B. B. Wayland, J . Am. Chem. Soc., 110,6063 (1988). 14.6.6.3.3. To Alcohols and Alkanes.

The complementary role of a Lewis acid in promoting CO reduction is important. Various complexes have been characterized in which Lewis acids bind the oxygen atom of CO coordinated to a transition metal through carbon'-9. This is shown schematically as 1. + M-c~o, M=Co., 'A 'A

..

-

1

Often the Lewis acid is generated after H- is donated from the active hydride sources so frequently used to reduce CO, as in pany of the examples in 14.6.6.3.1 and 14.6.6.3.2. This section describes examples of CO reduction to ROH and RH in which the principal source of hydrogen is protons from solvent or acid, with a possible additional source from a hydride donor. The ability of Hf to serve as the Lewis acid in 1 is shown in the protonation of the bridging carbonyl in [HFe,(CO),,!; lo. A similar reaction occurs in the acid decomposition of [Fe4(C0),,l2-, Scheme 1 ,in which CH, is produced from one of the carbonyl ligands. The sequential acid induced rearrangement of the cluster framework of [Fe,(C0),,I2- to a butterfly structure with an $-CO ligand, 0-protonation of the T ~ CO, and cleavage of the C-0 bond is found to result ultimately in liberation of CH, derived from the original CO ligand. The electrons for reduction come from the Fe, cluster with ca. 0.56 molecules of CH, produced per cluster"^'*.

H -

H H+, e-

1-

CH,, Fez", CO.. H Scheme 1.

H -


392


9. R. Hoffmann, C. N. Wilker, S. J. Lippard, J. L. Templeton, D. C. Brower, J . Am. Chem. SOC., 105, 146 (1983). 10. P. A. Bianconi, I. D. Williams, M. P. Engeler, S. J. Lippard, J . Am. Chem. Soc., 108, 31 1 (1986). 11. V. L. Coffin, W. Brennen, B. B. Wayland, J . Am. Chem. Soc., 110,6063 (1988). 14.6.6.3.3. To Alcohols and Alkanes.

The complementary role of a Lewis acid in promoting CO reduction is important. Various complexes have been characterized in which Lewis acids bind the oxygen atom of CO coordinated to a transition metal through carbon'-9. This is shown schematically as 1. + M-c~o, M=Co., 'A 'A

..

-

1

Often the Lewis acid is generated after H- is donated from the active hydride sources so frequently used to reduce CO, as in pany of the examples in 14.6.6.3.1 and 14.6.6.3.2. This section describes examples of CO reduction to ROH and RH in which the principal source of hydrogen is protons from solvent or acid, with a possible additional source from a hydride donor. The ability of Hf to serve as the Lewis acid in 1 is shown in the protonation of the bridging carbonyl in [HFe,(CO),,!; lo. A similar reaction occurs in the acid decomposition of [Fe4(C0),,l2-, Scheme 1 ,in which CH, is produced from one of the carbonyl ligands. The sequential acid induced rearrangement of the cluster framework of [Fe,(C0),,I2- to a butterfly structure with an $-CO ligand, 0-protonation of the T ~ CO, and cleavage of the C-0 bond is found to result ultimately in liberation of CH, derived from the original CO ligand. The electrons for reduction come from the Fe, cluster with ca. 0.56 molecules of CH, produced per cluster"^'*.

H -

H H+, e-

1-

CH,, Fez", CO.. H Scheme 1.

H -


393

Reaction of Ru,(CO),, with excess AlH, in THF followed by acidification with 1

M H,SO, yields hydrocarbons (CH,, C,H,, etc.), corresponding to reduction of -30%

of the CO ligands originally present',. Similar results are obtained with M(CO), systems (M = Cr, Mo, W)',. Acid decomposition of Mg[CpFe(CO),].4THF also yields small amounts of hydrocarbon^'^. In both of these Lewis acid coordination of CO as in 1 is believed to play an important role. The Lewis acidity and tendency to form stable metal 0x0 complexes, characteristic of many of the early transition metals, is an important factor in the cleavage of CEO. The Ta complex (t-Bu,SiO),Ta reacts with CO to yield a tantalum 0x0 complex and a ditantalum dicarbide complex, equation (a)15.

2 (siloX),Ta

+ co

25°C

benzene (silox = r-Bu,SiO-)

+

(silox),Ta = 0 '/z (silox),Ta=C=C=Ta(silox),

(a)

Cleavage reactions may have important implications for the heterogeneously catalyzed Fischer-Tropsch CO red~ction'~-~', which is believed to involve dissociative adsorption to surface oxide and carbide. Reduction of CO to alcohols using DIBAH (i-Bu,AlH) and Cp,ZrCl, occurs and makes use of both the acidic properties of aluminum alkyls and the hydridic nature of Zr hydrides31.The reaction proceeds as in equation (b) via the intermediate 11; hydrolysis after CO uptake yields a mixture of C,-C, alcohols. The observation of isobutylene suggests that each AlH(i-Bu), donates more than one H to the ROH products.

q. B '

* H-Al(i-Bu), Zr- H 'H-Al(i-Bu),

\ /

c1

(b)

2

A

MeOH, EtOH, PrOH, n-BuOH, and /

(R. EISENBERG, C.KUBIAK)

1 . D. F. Shriver, J . Organomet. Chem., 94, 259 (1975); Chem. Ber., 8,419 (1972). A. E. Crease, P. Legzdins, J . Chem. Ed., 52,499 (1975). J. S. Kristoff, D. F. Shriver, Znorg. Chem., 13,499 (1974), and references therein. N. E. Kim, N. J. Nelson, D. F. Shriver, Znorg. Chim. Acta, 7 , 393 (1973). N. J. Nelson, N. E. Kime, D. F. Shriver, J . Am. Chem. Soc., 91, 5173 (1969). Sr. Agnes Alich, N. J. Nelson, D. F. Shriver, J . Chem. Soc., Chem. Commun., 254 (1971). J. M. Burlitch, R. B. Petersen, J . Organomet. Chem., 24, C65 (1970). J. C. Kotz and C. D. Tumipseed, J . Chem. Soc., Chem. Cornmun., 41 (1970). R. B. Petersen, J. J. Stezowski, C. Wan, J. M. Burlitch, R. E. Hughes, J . Am. Chem. Soc., 93,

2. 3. 4. 5. 6. 7. 8. 9.

3532 (1971). 10. H. A. Hodali, D. F. Shriver, C. A. Ammlung, J . Am. Chem. Soc., ZOO, 5239 (1978). 11. D. F. Shriver, M. J. Sailor, Acct. Chem. Res., 21, 374 (1988). 12. K. Whitmire, D. F. Shriver, J . Am. Chem. Soc., ZO2, 1456 (1980).

394

14.6. Carbon Monoxide Reactions 14.6.6. Oxidation and Reduction of CO 14.6.6.4. Reduction by H,

13. C. Masters, C. van der Woude, J. A. van Doom, J . Am. Chem. SOC., 101, 1633 (1979). 14. A. Wong, M. Harris, and J. D. Atwood, J . Am. Chem. Soc., 102,4529 (1980). 15. D. R. Neithamer, R. E. LaPointe, R. A. Wheeler, D. S. Richeson, G . D. Van Duyne, P. T. Wolczanski, J . Am. Chem. SOC., 111,9056 (1989). 16. Comprehensivediscussions of CO/H2 reactions, methanation and methanol synthesis are given

21. 22.

in Encyclopedia of Chemical Technology, R. E. Kirk, D. F. Othmer, eds., 2nd ed. WileyInterscience, New York, 1963. For CO/H, reactions, see Vol. 4, p. 446; for methanation, see Vol. 13, p. 364; for MeOH synthesis, see Vol. 13, p. 370. I. Wender, Catal. Rev.-Sci. Eng., 14, 97 (1976). M. A. Vannice, Catal. Rev.-Sci. Eng., 14, 153 (1976). G. A. Mills and F. W. Steffgen, Catalysis Rev., 8, 159 (1973). H. H. Storch, H. Golumbic, R. B. Anderson, The Fischer-Tropsch and Related Syntheses. Wiley, New York, 1951. H. Pichler, H. Schulz, Chem.-1ng.-Tech.,42, 1162 (1970). C . L. Thomas, Catalytic Processes and Proven Catalysts., Ch. 14, Academic Press, New York,

23. 24. 25. 26. 27. 28. 29. 30. 31.

W. J. Thomas, S . Portalski, Ind. Eng. Chem.,50, 967 (1958). M. L. Poutsma, L. F. Elek, P. A. Ibarbia, A. D. Risch, J. A. Rabo, J . Catul., 52, 157 (1978). V. M. Vlasenko, G. E. Yuzefovich, Russ. Chem. Rev. (Engl. Trans.), 39, 728 (1969). D. L. King, J. Catal., 51, 386 (1978). B. K. Nefedov, Y. T. Eidus, Rum. Chem. Rev. (Eng. Trans.), 34,272 (1965). C. D. Chang, W. H. Lang, A. J. Silvestri,J . Catul., 56, 268 (1979). G. Henrici-Olivb, S . Olivb, Angew. Chem., Int. Ed. Engl., 15, 136 (1976). Y. T. Eidus, Russ. Chem. Rev. (Engl. Trans.), 36, 338 (1967). L. I. Shoer, J. Schwartz, J. Am. Chem. Soc., 99,5831 (1977).

17. 18. 19. 20.

1970.

14.6.6.4. Reduction by H2 Catalyzed reductions of C O by H, have been long known, beginning with methanation, equation (a), in 19021.2. MeOH synthesis, equation (b) in 19233, and Fischer-Tropsch (F-T) synthesis, equation (c) in 19234.

+ H,O CO, + H, CH, + HZO C O + 3H2 n C O + 2nH2 F (-CH,-)n + nH,O + oxygenates CO

(a) (b) (c)

These reactions and their catalysts, which are of the heterogeneous type, have been discussed e x t e n s i ~ e l y ~ ~In~ -order ' ~ . to develop a highly selective catalyst for F-T synthesis and the desirability of running these reactions under milder conditions, attention has focused recently on catalyzing H2 reductions of C O homogeneously. This section outlines recent progress in the area. These studies, together with those outlined in previous sections (vide supra), provide insight into the mechanisms of C O reduction. For F-T synthesis there are currently two dominant, competing mechanistic proposals: the first involves stepwise C-H and C-C bond formation via complexed intermediates such as formyl prior to cleavage of the C-0 bond, while the second espouses initial C-0 bond cleavage with formation of a metal carbide intermediate prior to C-H and C-C bond formation. The latter, a dissociative mechanism, is now generally more widely accepted for hydrocarbon production. In homogeneously catalyzed reductions of C O to oxygenates (alcohols, glycols and esters) considerable evidence exists for the intermediacy of formyls and formaldehyde". Discussions of the similarities and differences of CO reduction by heterogeneous catalysts and organometallic com-


394

14.6. Carbon Monoxide Reactions 14.6.6. Oxidation and Reduction of CO 14.6.6.4. Reduction by H,

13. C. Masters, C. van der Woude, J. A. van Doom, J . Am. Chem. SOC., 101, 1633 (1979). 14. A. Wong, M. Harris, and J. D. Atwood, J . Am. Chem. Soc., 102,4529 (1980). 15. D. R. Neithamer, R. E. LaPointe, R. A. Wheeler, D. S. Richeson, G . D. Van Duyne, P. T. Wolczanski, J . Am. Chem. SOC., 111,9056 (1989). 16. Comprehensivediscussions of CO/H2 reactions, methanation and methanol synthesis are given

21. 22.

in Encyclopedia of Chemical Technology, R. E. Kirk, D. F. Othmer, eds., 2nd ed. WileyInterscience, New York, 1963. For CO/H, reactions, see Vol. 4, p. 446; for methanation, see Vol. 13, p. 364; for MeOH synthesis, see Vol. 13, p. 370. I. Wender, Catal. Rev.-Sci. Eng., 14, 97 (1976). M. A. Vannice, Catal. Rev.-Sci. Eng., 14, 153 (1976). G. A. Mills and F. W. Steffgen, Catalysis Rev., 8, 159 (1973). H. H. Storch, H. Golumbic, R. B. Anderson, The Fischer-Tropsch and Related Syntheses. Wiley, New York, 1951. H. Pichler, H. Schulz, Chem.-1ng.-Tech.,42, 1162 (1970). C . L. Thomas, Catalytic Processes and Proven Catalysts., Ch. 14, Academic Press, New York,

23. 24. 25. 26. 27. 28. 29. 30. 31.

W. J. Thomas, S . Portalski, Ind. Eng. Chem.,50, 967 (1958). M. L. Poutsma, L. F. Elek, P. A. Ibarbia, A. D. Risch, J. A. Rabo, J . Catul., 52, 157 (1978). V. M. Vlasenko, G. E. Yuzefovich, Russ. Chem. Rev. (Engl. Trans.), 39, 728 (1969). D. L. King, J. Catal., 51, 386 (1978). B. K. Nefedov, Y. T. Eidus, Rum. Chem. Rev. (Eng. Trans.), 34,272 (1965). C. D. Chang, W. H. Lang, A. J. Silvestri,J . Catul., 56, 268 (1979). G. Henrici-Olivb, S . Olivb, Angew. Chem., Int. Ed. Engl., 15, 136 (1976). Y. T. Eidus, Russ. Chem. Rev. (Engl. Trans.), 36, 338 (1967). L. I. Shoer, J. Schwartz, J. Am. Chem. Soc., 99,5831 (1977).

17. 18. 19. 20.

1970.

14.6.6.4. Reduction by H2 Catalyzed reductions of C O by H, have been long known, beginning with methanation, equation (a), in 19021.2. MeOH synthesis, equation (b) in 19233, and Fischer-Tropsch (F-T) synthesis, equation (c) in 19234.

+ H,O CO, + H, CH, + HZO C O + 3H2 n C O + 2nH2 F (-CH,-)n + nH,O + oxygenates CO

(a) (b) (c)

These reactions and their catalysts, which are of the heterogeneous type, have been discussed e x t e n s i ~ e l y ~ ~In~ -order ' ~ . to develop a highly selective catalyst for F-T synthesis and the desirability of running these reactions under milder conditions, attention has focused recently on catalyzing H2 reductions of C O homogeneously. This section outlines recent progress in the area. These studies, together with those outlined in previous sections (vide supra), provide insight into the mechanisms of C O reduction. For F-T synthesis there are currently two dominant, competing mechanistic proposals: the first involves stepwise C-H and C-C bond formation via complexed intermediates such as formyl prior to cleavage of the C-0 bond, while the second espouses initial C-0 bond cleavage with formation of a metal carbide intermediate prior to C-H and C-C bond formation. The latter, a dissociative mechanism, is now generally more widely accepted for hydrocarbon production. In homogeneously catalyzed reductions of C O to oxygenates (alcohols, glycols and esters) considerable evidence exists for the intermediacy of formyls and formaldehyde". Discussions of the similarities and differences of CO reduction by heterogeneous catalysts and organometallic com-

14.6. Carbon Monoxide Reactions 14.6.6. Oxidation and Reduction of CO 14.6.6.4. Reduction by H2

395

~

plexes in solution, as well as the studies of related model systems and their mechanistic implications have been reviewedZ0-’’. Production of oxygenates (alcohols, glycols and esters) from CO H, using solution complex catalysts has been reported for several systems (Table l). In all cases relatively high pressures are employed. The dominant solution species in the Co and Ru cases19,23-24 are mononuclear species, HCo(CO), and Ru(CO),, respectively, which are thought to be the active catalysts. Halide promoters significantly increase the activities and selectivities for ethylene glycol of the ruthenium catalystsz5. Halide promoters also increase activities for homologation of methanol to ethan01’~-’~. The examples of ethylene glycol formation using Rh catalyst^'^*^^-^^ are still the most important entries in Table 1 because they achieve relatively selective C-C bond formations (-70% of products are glycol). Whereas several cluster species are observed directly under catalytic conditions, the actual catalytic species are not known with certainty. Both [Rh,(CO),,]- and [Rh(CO),]- are the principal species observed by high P IR s p e c t r o s ~ o p yThere ~ ~ . is little evidence to suggest that metals other than Co, Rh, or Ru have significant activity for the homogeneously catalyzed reduction of CO19. However, solvents and promoters exert an enormous influence on catalyst activity and selectivity, to the extent that other metals cannot be excluded from consideration. Sufficient data exist to permit generalizations about the activities and selectivities of promoted and unpromoted Co, Rh, and Ru catalysts for the reduction of CO to oxygenates3,. Rhodium catalysts show the highest activity and selectivity for ethylene glycol production. Halide promoted ruthenium catalysts show comparable overall activities for CO reduction, but at the expense of selectivity. Unpromoted ruthenium catalysts are relatively unreactive and unselective. Cobalt catalysts have similar activities to unpromoted ruthenium catalysts, but are less stable overall. Hydrocarbon formation from CO H, has been described using metal carbonyl clusters such as Ir,(CO),, and Rh6(CO),6 in molten NaCI.2 AICI, at 180°C and 1.5 am3’. The medium is highly acidic, suggestive of Lewis acid activation of coordinated CO as in 1. The primary product appears to be C&, with secondary formation of CH,. The concomitant formation of H,O in this reaction ultimately deactivates the catalyst medium via hydrolysis. Other examples of hydrocarbon formation from synthesis gas are limited. In one study, an ‘‘immobilized homogeneous catalyst,” polystyrene-attached CpCo(CO),, is employed, leading to a Fischer-Tropsch type distribution of products (C, to Czoproducts in declining amounts) at 190°C and 5 atrn3,. The free complex, however, does not yield F-T products. Catalytic alkylation of benzene with CO + H2 has also been reported in which W(CO), AlCl, acts as the catalyst at 200”C34.The primary reaction product is ethylbenzene with smaller amounts of longer chain alkylbenzenes. Again, catalyst lifetime is limited. Numerous studies designed to model some aspect of the reduction of CO by H, have been described. Notable, in addition to those studies described in earlier sections, are the use of carbide clusters as a means of studying hydrocarbon formation after carbonyl C-0 bond the determination of surface bound meth~lene,~, M=CH,, as the means of propagating chain growth in hydrocarbon formation, and the use of supported complexes on metal oxides, zeolites, and reticulated resin^,^-^'. Currently, the challenge of developing homogeneous catalysts for selective reduction of CO by H, remains largely unanswered, but much has been learned about CO reduction, and the outlook remains promising.

+

+

+

(R. EISENBERG, C. KUBIAK)

260 260

"Ethanol and propanol also obtained as secondary products. Ru precursors also employed including Ru(acac), at higher P. 'Other Ru precursors also employed. dAlso MeOH, n-PrOH, n-BuOH, cthylcne glycol, CH,.

Diglyme or THF/ 2-hydroxypyridine Sulfolane Sulfolane 1000

lo00

[Rhi7(CO),&13 Rh(CO),(acac)'

~

240 210 230

R3m/12

N-Methy lpyrrolidonelKI

1020 410 >lo00

Ru3(c0)12 Ru,(CO),, [Ru~(CO),~I

230

340

Acetic acid

225-275

1300

THF

200 180

T ("C)

Ru3(CO,z. H8~4(co)iz RU3(CO)I2

300 200-300

p (am)

C,H, or dioxane Dioxane or trifluroethanol

Solvent/promoter

C%,(CO), HCo(CO)4

Complex

TABLE1. OXYGENATE PRODUCTION FROM CO/H, USINGSOLUBLE METALCOMPLEXES

Ethylene glycol, MeOH Ethylene glycol, MeOH

AcOMe, AcOCH,CH,OAc Ethylene glycol, MeOH, EtOH Etohd Ethylene glycol, MeOH

MeOH, HCOOMe" MeOH, HCOOMe, ethylene glycol MeOH, HCOOMe

products

31 31

25 19 28,29

27

24

23 30

Ref.

396 14.6. Carbon Monoxide Reactions 14.6.6. Oxidation and Reduction of CO 14.6.6.4. Reduction by H2

14.6. Carbon Monoxide Reactions 14.6.6. Oxidation and Reduction of CO 14.6.6.4. Reduction by H2

397

1. G . A. Mills, F. W. Steffgen, Catal. Rev., 8, 159 (1973). 2. P. Sabatier, J. B. Senderens, C. R. Hebd. Seances Acad. Sci., 134,514 (1902). 3. Comprehensive discussins of CO/H, reactions, methanation and methanol synthesis are given in Encyclopedia of Chemical Technology, R. E. Kirk, D. F. Othmer, eds., 2nd ed., WileyInterscience, New York, 1963. For CO/H, reactions, see Vol. 4, p. 446; for methanation, see Vol. 13, p. 364; for MeOH synthesis, see Vol. 13, p. 370. 4. F. Fischer, H. Tropsch, Chem. Ber., 56,2428 (1923); 59, 830,832, and 923 (1926). 5. I. Wender, Catal. Rev.-Sci. Eng., 14,97 (1976). 6. M. A. Vannice, Catal. Rev.-Sci. Eng., 14, 153 (1976). 7. G . A. Mills, F. W. Steffgen, Catal. Rev., 8, 159 (1973). 8. H. H. Storch, H. Golumbic, R. B. Anderson, The Fischer-Tropsch and Related Syntheses. Wiley, New York, 1951. 9. H. Pichler, H. Schulz, Chem.-1ng.-Tech.,42, 1162 (1970). 10. C. L. Thomas, Catalytic Processes and Proven Catalysts, Ch. 14., Academic Press, New York, 1970. 11. W. J. Thomas, S . Portalski, Ind. Eng. Chem.,50, 967 (1958). 12. M. L. Poutsma, L. F. Elek, P. A. Ibarbia, A. D. Risch, J. A. Rabo, J . Catal., 52, 157 (1978). 13. V. M. Vlasenko, G . E. Yuzefovich, Russ. Chem. Rev. (Engl. Trans.). 39, 728 (1969). 14. D. L. King, J . Catal., 51, 386 (1978). 15. B. K. Nefedov, Y. T. Eidus, Russ. Chem. Rev. (Engl. Transl.), 34,272 (1965). 16. C. D. Chang, W. H. Lang, A. J. Silvestri, J . Catal., 56, 268 (1979). 17. G. Henrici-Olive, S. Olivk, Angew. Chem., Int. Ed. Engl., 15, 136 (1976). 18. Y. T. Eidus, Rum. Chem. Rev., (Engl. Trans.), 36, 338 (1967). 19. B. D. Dombek, Adv. Catal., 32, 325 (1983). 20. C. Masters, Adv. Organometal. Chem., 17, 61 (1979). 21. E. L. Muetterties, J. Stein, Chem. Rev., 79, 479 (1979). 22. R. Eisenberg, D. E. Hendriksen, Adv. Catal., 28, 79 (1979). 23. J. W. Rathke, H. M. Feder, J . Am. Chem. SOC.,100, 3623 (1978). 24. J. S. Bradley, J . Am. Chem. Soc., 101, 7419 (1979). 25. B. D. Dombek, J . Am. Chem. Soc., 103,6508 (1981). 26. G . Braca, G . Sbrana, G . Valentine, G . Andrich, G . Gregorio, J . Am. Chem. SOC., 100, 6238 (1978). 27. B. D. Dombek, J . Am. Chem. SOC.,102,6855 (1980). 28. R. L. Pruett, Ann. N.Y. Acad. Sci., 295, 239 (1977). 29. R. L. Pruett, W. E. Walker, U.S. Patent No. 3,833,634 (1974); Chem. Abst., 79,78088k (1973). 30. J. W. Rathke, H. M. Feder, Catalysis of Organic Reactions, W. R. Moser, ed., Dekker, New York, 1981, p. 219. 31. L. J. Vidal, W. E. Walker, Inorg. Chem., 19, 896 (1980). 32. G. C. Demitras, E. L. Meutterties, J . Am. Chem. Soc., 99, 2796 (1977). 33. P. Perkins, K. P. C. Vollhardt, J . Am. Chem. Soc., 101, 3985 (1979). 34. G . Henrici-Olive, S. Olive, Angew. Chem. Int. Ed., 18, 77 (1979). 35. J. S. Bradley, G . B. Ansell, E. W. Hill, J . Am. Chem. Soc., 101,7417 (1979). 36. M. Tachikawa, E. L. Muetterties, J . Am. Chem. SOC.,102,4541 (1980). 37. M. Tachikawa, A. C. Sievert, E. L. Muetterties, M. R. Thompson, C. S . Day, V. W. Day, J . Am. Chem. Soc., 102, 1725 (1980). 38. R. C. Brady 111, R. Pettit, J . Am. Chem. Soc., 102, 6181 (1980). 39. A. K. Smith, A. Theolier, J. M. Basset, R. Ugo, D. Commereuc, Y. Chauvin, J . Am. Chem. SOC., 100,2590 (1978). 40. M. Ichikawa, J. Chem. Soc., Chem. Commun., 566 (1978). 41. M. Niwa, T. Iizuka, J. H. Lunsford, J . Chem. SOC.,Chem. Commun., 673 (1979).


14.7. Oxidation 14.7.1. Introduction Additional sections which deal with oxidation reactions catalyzed by inorganic compounds are projected for publication in a final Supplementary unit of the Inorganic Reactions and Methods series.

14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds Methane, n-butane, isobutane, and cyclohexane have strong C-H bonds (DC-H > 390 kJ mol- ')*,, and high oxidation potentials (IP > 9.8 e ~ )and ~ , vigorous conditions are required to oxidize these hydrocarbons with inorganic oxidants. On the other hand, toluene and o-, m-, and p-xylenes have weaker C-H bonds (Dc-H5 355 kJ mol- ')' and lower ionization potentials (IP < 8.8 eV)3, and can be oxidized more readily. The primary oxidation products are hydroperoxides, alcohols, ketones, and other oxygenated compounds containing the same carbon skeleton as the starting hydrocarbon. These can be described as products of C-H bond attack or H-atom abstraction. These compounds are oxidized more readily than the parent hydrocarbon, and secondary oxidation often occurs. Nevertheless, by a proper choice of catalysts and conditions many commercially important primary oxidation products are formed in good yields. (J. A. HOWARD)

1. S.W. Benson, Thermochemical Kinetics. Wiley, New York, 1976. 2. A. L. Castelhano, P. R. Mamott, D. Griller, J . Am. Chem. Soc., 103, 4262 (1981). 3. H. M. Rosenstock, K. Draxl, B. W. Steiner, J. T. Herron, J . Phys. Chem. Ref.Data, 6, Suppl. 1

(1977).

14.7.2.1. In Methane Oxldatlon

Complete oxidation of methane to carbon dioxide and water occurs when it is burned in air and controlled combustion is an extremely important source of thermal energy. Methane and 0, also form extremely dangerous mixtures in the gas phase and explosions are easily detonated. Explosions occur under oxygen-deficient conditions and CO is an important and toxic product. In addition to its use as a fuel, CH, is a raw material for the production of synthesis gas, a mixture of CO and H,, which can be used to make methanol'. This oxidation is achieved with steam which reacts with CH, at high T in the presence of a variety of catalysts to give hydrogen and either CO, or CO (see 1.2.8):

+ HZO CO + H,O

CH4 CH,

398

+ 2H,O

--

+ CO CO, + H,

(b)

4H,

(c)

3H,

+ CO,

(a)


14.7. Oxidation 14.7.1. Introduction Additional sections which deal with oxidation reactions catalyzed by inorganic compounds are projected for publication in a final Supplementary unit of the Inorganic Reactions and Methods series.

14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds Methane, n-butane, isobutane, and cyclohexane have strong C-H bonds (DC-H > 390 kJ mol- ')*,, and high oxidation potentials (IP > 9.8 e ~ )and ~ , vigorous conditions are required to oxidize these hydrocarbons with inorganic oxidants. On the other hand, toluene and o-, m-, and p-xylenes have weaker C-H bonds (Dc-H5 355 kJ mol- ')' and lower ionization potentials (IP < 8.8 eV)3, and can be oxidized more readily. The primary oxidation products are hydroperoxides, alcohols, ketones, and other oxygenated compounds containing the same carbon skeleton as the starting hydrocarbon. These can be described as products of C-H bond attack or H-atom abstraction. These compounds are oxidized more readily than the parent hydrocarbon, and secondary oxidation often occurs. Nevertheless, by a proper choice of catalysts and conditions many commercially important primary oxidation products are formed in good yields. (J. A. HOWARD)

1. S.W. Benson, Thermochemical Kinetics. Wiley, New York, 1976. 2. A. L. Castelhano, P. R. Mamott, D. Griller, J . Am. Chem. Soc., 103, 4262 (1981). 3. H. M. Rosenstock, K. Draxl, B. W. Steiner, J. T. Herron, J . Phys. Chem. Ref.Data, 6, Suppl. 1

(1977).

14.7.2.1. In Methane Oxldatlon

Complete oxidation of methane to carbon dioxide and water occurs when it is burned in air and controlled combustion is an extremely important source of thermal energy. Methane and 0, also form extremely dangerous mixtures in the gas phase and explosions are easily detonated. Explosions occur under oxygen-deficient conditions and CO is an important and toxic product. In addition to its use as a fuel, CH, is a raw material for the production of synthesis gas, a mixture of CO and H,, which can be used to make methanol'. This oxidation is achieved with steam which reacts with CH, at high T in the presence of a variety of catalysts to give hydrogen and either CO, or CO (see 1.2.8):

+ HZO CO + H,O

CH4 CH,

398

+ 2H,O

--

+ CO CO, + H,

(b)

4H,

(c)

3H,

+ CO,

(a)

14.7.1. Introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.2. In Butane Oxidation ~

399

~~

The catalysts used are invariably of the impregnated- or coprecipitated-nickel type. Complex oxide supports reduce the deposition of carbon on the catalyst and the formation of polymeric material on the catalyst surface. Although the proportions of CO and CO, depend on the reaction conditions, CH,, and H,O alone do not contain enough carbon for the subsequent synthesis of methanol. This deficiency is made up by adding CO, to the reaction mixture: 3CH4

+ C02 + 2H2O

4CO

+ 8H2

(4

In practice methane (natural gas), steam, and CO, are preheated and passed through externally heated tubes packed with promoted nickel catalyst. Exact control of the reaction is required in order to obtain the correct H,/CO ratio. The rate controlling step for reaction involves methane adsorption. Catalyst structure has a marked effect on the kinetics of the reaction. Thus, under certain conditions the rate of reaction over a Ni/Kieselguhr catalyst at 91 1 K is first order with respect to the partial pressure of CH, and independent of H,O and product partial P2, while for other nickel catalysts the rate depends on the partial P of H,O, H,, and C03. The reaction of CH, with 0, is catalyzed by a wide variety of transition metals and their oxides (e.g., Pt, Pd, Cu, Ag, Au, and V) on a variety of supports such as alumina, asbestos, silica-gel, Linde 4A molecular sieves, and pumice at 300-1000 K. The reaction products are mainly CO, and H,O, along with traces of formaldehyde. Overall rates of palladium-catalyzed oxidations obey - d [cH4I

dt

=

k[CH4]0.5[02]0

where k is the overall rate constant. Activation energies in flow systems are 96 kJ mol-', while lower values have been found in static systems4.The rate-controlling step involves a surface reaction between oxide ions and dissociatively adsorbed, linearly bound methylene radicals. The catalytic activity of transition metal oxides decreases: CO@, > Cu 2 NiO > MnO, > Cr203 > Fe,O, > ZnO > V,O, > TiO,. The rate-controlling step in these oxidations involves cleavage of an oxygen-catalyst bond. (J. A. HOWARD) 1. H. F. Woodward, Jr., Kirk-Othmer Encycl. Chem. Technol., 13, 370 (1967). 2. W. W. Akers, D. P. Camp, Am. Inst. Chem. Eng. J., 1,471, (1955). 3. I. M. Bodrov, L. 0. Apel'baum, M. Temkin, Kinet. Kutul., 5, 696 (1964); 8, 821 (1967); 9, 1065 (1968). 4. C. F. Cullis, D. E. Keene, D. L. Timm, Trans. Faruduy SOC., 67, 864 (1971).

14.7.2.2.

In Butane Oxidation

Butane like CH, bums in air to produce thermal energy and reacts with steam over nickel catalysts to give synthesis gas by a mechanism which involves initial breakdown of butane to CH,. The di-t-butylperoxide-initiatedliquid phase oxidation of n-butane at 275-400 K gives a variety of oxygen-containing compounds and can be described by'


14.7.1. Introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.2. In Butane Oxidation ~

399

~~

The catalysts used are invariably of the impregnated- or coprecipitated-nickel type. Complex oxide supports reduce the deposition of carbon on the catalyst and the formation of polymeric material on the catalyst surface. Although the proportions of CO and CO, depend on the reaction conditions, CH,, and H,O alone do not contain enough carbon for the subsequent synthesis of methanol. This deficiency is made up by adding CO, to the reaction mixture: 3CH4

+ C02 + 2H2O

4CO

+ 8H2

(4

In practice methane (natural gas), steam, and CO, are preheated and passed through externally heated tubes packed with promoted nickel catalyst. Exact control of the reaction is required in order to obtain the correct H,/CO ratio. The rate controlling step for reaction involves methane adsorption. Catalyst structure has a marked effect on the kinetics of the reaction. Thus, under certain conditions the rate of reaction over a Ni/Kieselguhr catalyst at 91 1 K is first order with respect to the partial pressure of CH, and independent of H,O and product partial P2, while for other nickel catalysts the rate depends on the partial P of H,O, H,, and C03. The reaction of CH, with 0, is catalyzed by a wide variety of transition metals and their oxides (e.g., Pt, Pd, Cu, Ag, Au, and V) on a variety of supports such as alumina, asbestos, silica-gel, Linde 4A molecular sieves, and pumice at 300-1000 K. The reaction products are mainly CO, and H,O, along with traces of formaldehyde. Overall rates of palladium-catalyzed oxidations obey - d [cH4I

dt

=

k[CH4]0.5[02]0

where k is the overall rate constant. Activation energies in flow systems are 96 kJ mol-', while lower values have been found in static systems4.The rate-controlling step involves a surface reaction between oxide ions and dissociatively adsorbed, linearly bound methylene radicals. The catalytic activity of transition metal oxides decreases: CO@, > Cu 2 NiO > MnO, > Cr203 > Fe,O, > ZnO > V,O, > TiO,. The rate-controlling step in these oxidations involves cleavage of an oxygen-catalyst bond. (J. A. HOWARD) 1. H. F. Woodward, Jr., Kirk-Othmer Encycl. Chem. Technol., 13, 370 (1967). 2. W. W. Akers, D. P. Camp, Am. Inst. Chem. Eng. J., 1,471, (1955). 3. I. M. Bodrov, L. 0. Apel'baum, M. Temkin, Kinet. Kutul., 5, 696 (1964); 8, 821 (1967); 9, 1065 (1968). 4. C. F. Cullis, D. E. Keene, D. L. Timm, Trans. Faruduy SOC., 67, 864 (1971).

14.7.2.2.

In Butane Oxidation

Butane like CH, bums in air to produce thermal energy and reacts with steam over nickel catalysts to give synthesis gas by a mechanism which involves initial breakdown of butane to CH,. The di-t-butylperoxide-initiatedliquid phase oxidation of n-butane at 275-400 K gives a variety of oxygen-containing compounds and can be described by'

400

14.7.1. Introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.2. In Butane Oxidation

t-BuOOBu-t t-BuO' n-BuH

+

sec-Bu'

+ 0,

+ n-BuH

sec-BuO,' 2sec-Bu0,'

--

2t-BuO' t-BuOH

+ sec-Bu'

c --

sec-BuO,' sec-BuOOH

+ n-BuH sec-BuO'

(b) (c) (4

+ 0, MeC(0)Et + sec-BuOH sec-BuOH + sec-Bu'

(g)

Et'

(h)

sec-Bu0,Bu-sec

sec-BuO'

+ sec-Bu'

(a)

2sec-BuO'

+ CH,CHO

(el (f)

sec-Butyl hydroperoxide is the major product at low rates of chain initiation and yields as high as 77% based on the amount of oxygen absorbed are obtained'. At higher rates of initiation and lower butane concentrations chain lengths are lower and products from the self-reaction of sec-butylperoxy radicals become more important. Nonterminating interactions of sec-BuO,' give sec-butoxy radicals, see equation (c) in 14.7.5.1, and these radicals can either abstract a H-atom from n-butane to give sec-butanol [equation (g)], or they can undergo p-scission to give acetaldehyde and ethyl radicals. The ethyl radicals formed in this reaction combine with oxygen to give ethylperoxy radicals, and these radicals can either abstract a hydrogen from n-butane or react with a secbutylperoxy radical:

+ 0,

Et'

+ n-BuH

EtO,' EtO,'

+ sec-BuO,'

&

c

EtO'

Et-0,Bu-sec

-

EtO,'

EtOOH

+ sec-Bu'

+ sec-BuO' + 0,

CH,CHO

(9 (j)

(k)

+ EtOH + MeC(0)Et + sec-BuOH + 0,

(1)

Further reactions of EtO' and oxidation of acetaldehyde and formaldehyde lead to the formation of ethanol, acetic acid, formic acid, and CO,. Alkylperoxy radicals are selective, and abstraction of a secondary H-atom is favored over abstraction of a primary H-atom'. Alkoxy radicals, on the other hand, are less selective and can abstract primary hydrogens3:

+ n-BuH n-Bu' + 0,

sec-BuO'

--

sec-BuOH n-BuO,'

+ n-Bu'

(m) (n)

Formation of the n-butylperoxy radical leads to n-butyl hydroperoxide, n-butanol, and butyraldehyde. In the gas phase chain lengths are shorter, and yields of the reaction products reflect the increased importance of self-reactions of sec-butylperoxy radicals', Metal ion-catalyzed oxidation of n-butane at 440-450 K and 50 atm is an important industrial process which gives acetic and formic acids along with low yields of methyl

14.7.1. introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.2. In Butane Oxidation

40 1

ethylketone and various alkyl acetates4. The ratio of the two acids depends on the type and selectivity of the catalyst. Cobalt(II1) acetate catalyst forms acetic over formic acid, while manganese(II1) acetate is less selective. These reactions are initiated by radicals produced by interaction of the metal ions with hydroperoxides present in the hydrocarbon as illustrated for cobalt?

-

+ Co2+ F= [se~-Bu0OHCo]~+ [se~-BuO0HCo]~+ sec-BuO' + HO- + Co3+ sec-BuOOH + Co3+ F= [sec-BuOOHCoI3+

sec-BuOOH

[se~-Bu0OHCo]~+

sec-BuO,'

+ H+ -t Co2+

(0)

(PI

(4) (4

There is, therefore, often an induction period before reaction commences. The overall rate increases autocatalytically until a limiting rate is reached. At this point the concentration of hydroperoxide available for initiation reaches a constant value. In many cases the steady state concentration of hydroperoxide is low, autoxidation rates are high and kinetic chain lengths short. Most of the hydrocarbon is consumed by reactions with other than peroxy radicals6. Oxidation of n-butane to acetic acid can be achieved by large concentrations of cobalt(II1) acetate at 383 K and a P of 24 atm7. The reaction is initiated by electron transfer from the substrate to the metal ion: C4H,,

+ Co(II1)

[C,H,,]+'

--

+

[C4HLO]+* Co(I1)

(s)

+ H+

(0

C4H;

Isobutane reacts with 0, by a free radical chain process similar to the one for nbutane except that cage combinations of t-butoxy radicals produced by terminating interactions of t-butylperoxy radicals gives di-t-butylperoxide*. t-Butoxy radicals which escape the solvent cage undergo &scission to give methyl radicals and acetone: 2t-BuOz'

-c -

t-BuO~BU-t t-BuO'

Me'

t-BuOOBu-t 2t-BuO'

+ Me2C0

+02

(u) (v) (w)

Chain lengths are considerably longer for isobutane than for n-butane because of the difference in the strengths of the tertiary and secondary C-H bonds; yields of tbutyl hydroperoxide are, therefore, high. The i-BuO,' radicals, formed by abstraction of a primary hydrogen from isobutane by t-BuO', play an important role in the autoxidation of this hydrocarbon'. Oxidation of isobutane with cobalt(II1) acetate is slower than oxidation of n-butane with this salt, and acetone, t-butyl alcohol, and methanol are the major products7. The lower reactivity of isobutane compared to n-butane is attributed to steric hindrance with the bulky, ligated metal ion. (J. A. HOWARD) 1. T. Mill, F. R. Mayo, H. Richardson, K. Irwin, D. L. Allara, J . Am. Chem. Soc., 94,6802 (1972). 2. J. A. Howard, J. H. B. Chenier, Can. J . Chem., 58, 2808 (1980).

402

14.7.1. Introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.3. In Cyclohexane Oxidation

3. D. G. Hendry, T. Mill, L. Piskiewicz, J. A. Howard, H. K. Eigenmann, J. Phys. Chem. Ref. Data, 3,937 (1974). 4. W. Dumas, W. Bulani, Oxidation of Petrochemicals: Chemistry and Technology, Wiley, New York, 1972. 5 . J. F. Black, J . Am. Chem., Soc., 100, 527 (1978). 6. C. Walling, J . Am. Chem., Soc., 91, 7590 (1969). 7. A. Onopchenko, J. G . D. Schultz, J . Org. Chem., 38, 909 (1973). 8. D. L. Allara, T. Mill, G . D. Hendry, F. R. Mayo, Adv. Chem. Ser., 76,40 (1968). 9. P. S. Nangia, S. W. Benson, Inr. J . Chem. Kinet., 12, 169 (1980).

14.7.2.3. In Cyclohexane Oxidation

Oxidation of cyclohexane with 0, can be described by a series of elementary reactions similar to the ones for n-butane. In this case, however, the hydrocarbon contains no primary hydrogens and three times as many secondary hydrogens as the acyclic alkane. Chain lengths are, therefore, longer and reaction products less numerous. Cyclohexyl hydroperoxide and cyclohexanol are the major propagation products while cyclohexanone and cyclohexanol are formed in termination:

h+o-()+o OOH

2

c

(a>

(b)

6 6

+

0 2

+02

@Cleavage of cyclohexyloxy radicals formed in reaction (d) gives 5formylpentyl radicals which undergo the propagation and termination reactions (8) and (h):


402


3. D. G. Hendry, T. Mill, L. Piskiewicz, J. A. Howard, H. K. Eigenmann, J. Phys. Chem. Ref. Data, 3,937 (1974). 4. W. Dumas, W. Bulani, Oxidation of Petrochemicals: Chemistry and Technology, Wiley, New York, 1972. 5 . J. F. Black, J . Am. Chem., Soc., 100, 527 (1978). 6. C. Walling, J . Am. Chem., Soc., 91, 7590 (1969). 7. A. Onopchenko, J. G . D. Schultz, J . Org. Chem., 38, 909 (1973). 8. D. L. Allara, T. Mill, G . D. Hendry, F. R. Mayo, Adv. Chem. Ser., 76,40 (1968). 9. P. S. Nangia, S. W. Benson, Inr. J . Chem. Kinet., 12, 169 (1980).

14.7.2.3. In Cyclohexane Oxidation

Oxidation of cyclohexane with 0, can be described by a series of elementary reactions similar to the ones for n-butane. In this case, however, the hydrocarbon contains no primary hydrogens and three times as many secondary hydrogens as the acyclic alkane. Chain lengths are, therefore, longer and reaction products less numerous. Cyclohexyl hydroperoxide and cyclohexanol are the major propagation products while cyclohexanone and cyclohexanol are formed in termination:

h+o-()+o OOH

2

c

(a>

(b)

6 6

+

0 2

+02

@Cleavage of cyclohexyloxy radicals formed in reaction (d) gives 5formylpentyl radicals which undergo the propagation and termination reactions (8) and (h):


H \ C(CH2),CH,02' OH

+ RH

-

H\C(CH2),CH200H OH 1

3

2

-

Secondary oxidation of 1, 2, and 3 leads to adipic acid, e.g.:

+ R'

OCH(CH,),CH,OOH

403

OCH(CH,), CHO

+

'OH

+ RH

(i)

Commercially, cyclohexane is oxidized to a mixture of cyclohexyl hydroperoxide, cyclohexanol, and cyclohexanone at 3.5-19 X 10, kP,and 400-450 K using a cobalt(I1) naphthanate catalyst'. The hydroperoxide in this mixture is then decomposed to cyclohexanol and cyclohexanone with a second cobalt catalyst. Adipic acid, the desired endproduct of cyclohexane oxidation, is produced by oxidation of the mixture of cyclohexanol and cyclohexanone with 40% HNO, containing 0.2% of a catalyst consisting of ammonium metavanadate and copper turnings. Oxidation of cyclohexane with cobalt(II1) acetate in acetic acid under nitrogen at 343 K gives mainly cyclohexyl acetate, 2-acetoxycyclohexanone and cyclohexylidene diacetate and traces of cyclohexanone, cyclohexanol, and bicyclohexy12.In the presence of O,, adipic acid is the main product. Reactivities of cyclohexane and cyclohexane-d,, are equal within experimental error, and loss of a proton does not occur in the ratecontrolling step. The mechanism for this oxidation has been rationalized in terms of an initial electron transfer, in which the species are in equilibrium, followed by loss of a proton from the radical cation:

0

+ Co(II1) +

+ Co(I1)

(k)

Manganese(II1) acetate oxidation of cyclohexane gives the same products as the Co(OAc), oxidation'. The selectivity of these two oxidants is, however, different. The former thermolyzes to give carboxymethyl radicals and the reaction products are formed by

404

0 -0 0 -0

14.7.1. Introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.4. in Toluene Oxidation

Mn(OAc),

Mn(OAc),

+ 'CHZCOOH

+ 'CH,COOH

(m)

+ CH,COOH

(n)

+ Mn(I1)

+ Mn(II1)

(0)

(J. A. HOWARD) 1 . W. Dumas, W. Bulani, Oxidation of Petrochemicals: Chemistry and Technology. Wiley, New

York, 1972. 2. A. Onopchenko, J. G. D. Schulz, J Org. Chem., 40, 3338 (1975). 14.7.2.4. In Toluene Oxidation

Toluene reacts slowly with 0, at T below -450 K to give benzyl hydroperoxide, benzyl alcohol, benzaldehyde, and dibenzyl peroxide by a free-radical chain process:

-

Production of free-radicals Initiation: 0, RO,' Propagation: R' ROOH RH RO,'

+

+

Termination: Ro2' -k Ro2' RO,' R'

+

-

+ R'

%} non radical products

where RH represents toluene and RO,' and R' the benzylperoxy and benzyl radical, respectively. The ratio of the reaction products depends on the T, chain length, and 0, pressure. At long chain lengths and high 0, pressures the rate of oxidation is given by

where k , and 2 4 are the rate-controlling propagation and termination rate constants, and

Ri is the rate of chain initiation.

Air oxidation of toluene in the presence of transition metal ions in their lower oxidation state gives benzoic acid. Initiation is achieved by a redox reaction between benzyl hydroperoxide and the metal ion. In addition, if all the metal ion is not associated with hydroperoxide, initiation can occur via reaction of the higher valency state of the metal ion with toluene':


404

0 -0 0 -0

14.7.1. Introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.4. in Toluene Oxidation

Mn(OAc),

Mn(OAc),

+ 'CHZCOOH

+ 'CH,COOH

(m)

+ CH,COOH

(n)

+ Mn(I1)

+ Mn(II1)

(0)

(J. A. HOWARD) 1 . W. Dumas, W. Bulani, Oxidation of Petrochemicals: Chemistry and Technology. Wiley, New

York, 1972. 2. A. Onopchenko, J. G. D. Schulz, J Org. Chem., 40, 3338 (1975). 14.7.2.4. In Toluene Oxidation

Toluene reacts slowly with 0, at T below -450 K to give benzyl hydroperoxide, benzyl alcohol, benzaldehyde, and dibenzyl peroxide by a free-radical chain process:

-

Production of free-radicals Initiation: 0, RO,' Propagation: R' ROOH RH RO,'

+

+

Termination: Ro2' -k Ro2' RO,' R'

+

-

+ R'

%} non radical products

where RH represents toluene and RO,' and R' the benzylperoxy and benzyl radical, respectively. The ratio of the reaction products depends on the T, chain length, and 0, pressure. At long chain lengths and high 0, pressures the rate of oxidation is given by

where k , and 2 4 are the rate-controlling propagation and termination rate constants, and

Ri is the rate of chain initiation.

Air oxidation of toluene in the presence of transition metal ions in their lower oxidation state gives benzoic acid. Initiation is achieved by a redox reaction between benzyl hydroperoxide and the metal ion. In addition, if all the metal ion is not associated with hydroperoxide, initiation can occur via reaction of the higher valency state of the metal ion with toluene':

14.7.1. Introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.4. In Toluene Oxidation

405

Benzaldehyde and benzyl alcohol are formed in the oxidation, but they are rapidly oxidized to benzoic acid: CHO

CHZOH

COOH I

The cobalt-catalyzed autoxidation of toluene in acetic acid at 363 K is accelerated by butan-2-one and benzaldehyde because peroxy radicals play a minor role in ratecontrolling propagation reactions*. High rates of autoxidation are also obtained in the presence of Br - because bromine atoms are important chain-propagating species'. Oxidation of toluene by manganese(II1) acetate in refluxing acetic acid in the absence of 0, yields three major products, benzyl acetate, the isomeric methyl benzyl acetates and the isomeric tolylacetic acids3:

6 0 6 CHZOAC

MNOAc),,

+

CH,

&I + \

CH,OAc

(9

\

CH,COOH

The mechanism accounting for the production of these three compounds is a nonchain free-radical process involving carboxymethyl radicals: Mn(OAc),

C

CHZOAC (-0Ac

6

HOAc 7 'CH,COOH + Mn(OAc),

6

6

H

k

H

I

CH,COOH

2

406

14.7.1. Introduction 14.7.2. Oxidation of Saturated Unactivated and Activated C-H Bonds 14.7.2.5. In Xylene Oxidation

+

The ratio of 1:2 3 = 0.3 represents the relative rate constants for &-hydrogen abstraction ( k , ) to that of nuclear addition (k2) by the 'CH,C(O)OH radical. Oxidation of toluene by lead tetraacetate in the absence of 0, gives benzyl acetate, methyl benzyl acetates, toluic acids, and xylenes4. Lead acetate pyrolyzes to give methyl radicals and carbon dioxide:

Methyl radicals either add to toluene to give xylenes, or abstract a H-atom from acetic acid or toluene. The products other than the xylenes are produced by the reactions shown in equation (j) (J. A. HOWARD)

1 . K. U. Ingold, Lipids and Their Oxidation, H. W. Schultz, E. A. Day, eds., Avi, Westport, CT, Chap. 5 (1963). 2. C. Walling, J . Am. Chem. SOC., 91,7590 (1969). 3. E. I. Heiba, R. M. Dessau, W. J. Koehl, Jr., J. Am. Chem. SOC.,91, 138 (1969). 4. E. I. Heiba, R. M. Dessau, W. J. Koehl, Jr., J . Am. Chem. SOC., 90, 1082 (1968).

14.7.2.5. In Xylene Oxidation

Autoxidation of 0-,m-,and p-xylenes at long chain lengths gives 0-,m-,and pmethyl-substituted benzyl hydroperoxides. Rates of autoxidation are more than twice the rate of autoxidation of toluene and depend on the position of the substituent'. Industrially, terephthalic acid is produced by the cobalt(II1) acetate, manganese(II1) acetate, or ammonium molybdate-catalyzed air oxidation of p-xylene in acetic acid'. Sodium bromide reduces the induction period and increases the rate of conversion to terephthalic acid. p-Xylene is initially oxidized to p-methylbenzyl hydroperoxide, and further oxidation gives p-methylbenzaldehyde and p-methylbenzoic acid:

+-o-+++---$) CH200H

CH3

CH3

CHO

CH3

COOH

CH3

COOH

COOH

(a)


406


+

The ratio of 1:2 3 = 0.3 represents the relative rate constants for &-hydrogen abstraction ( k , ) to that of nuclear addition (k2) by the 'CH,C(O)OH radical. Oxidation of toluene by lead tetraacetate in the absence of 0, gives benzyl acetate, methyl benzyl acetates, toluic acids, and xylenes4. Lead acetate pyrolyzes to give methyl radicals and carbon dioxide:

Methyl radicals either add to toluene to give xylenes, or abstract a H-atom from acetic acid or toluene. The products other than the xylenes are produced by the reactions shown in equation (j) (J. A. HOWARD)

1 . K. U. Ingold, Lipids and Their Oxidation, H. W. Schultz, E. A. Day, eds., Avi, Westport, CT, Chap. 5 (1963). 2. C. Walling, J . Am. Chem. SOC., 91,7590 (1969). 3. E. I. Heiba, R. M. Dessau, W. J. Koehl, Jr., J. Am. Chem. SOC.,91, 138 (1969). 4. E. I. Heiba, R. M. Dessau, W. J. Koehl, Jr., J . Am. Chem. SOC., 90, 1082 (1968).

14.7.2.5. In Xylene Oxidation

Autoxidation of 0-,m-,and p-xylenes at long chain lengths gives 0-,m-,and pmethyl-substituted benzyl hydroperoxides. Rates of autoxidation are more than twice the rate of autoxidation of toluene and depend on the position of the substituent'. Industrially, terephthalic acid is produced by the cobalt(II1) acetate, manganese(II1) acetate, or ammonium molybdate-catalyzed air oxidation of p-xylene in acetic acid'. Sodium bromide reduces the induction period and increases the rate of conversion to terephthalic acid. p-Xylene is initially oxidized to p-methylbenzyl hydroperoxide, and further oxidation gives p-methylbenzaldehyde and p-methylbenzoic acid:

+-o-+++---$) CH200H

CH3

CH3

CHO

CH3

COOH

CH3

COOH

COOH

(a)


407

Terephthalic acid is not formed until p-methylbenzaldehyde has been completely converted to p-methylbenzoic acid. Oxidation of o-xylene in the homogeneous liquid phase gives phthalic anhydride2:

a

II

COOH

COCH,

+CH30H

CH3

a:;

+H20

CH3

0

0

II

II

o,\ CH3

-a 0

0 + CH3OH + H2O

(4

1I

0

Catalysts for step (b) include salts of cobalt and manganese, for step (c) sulfuric acid and for step (d) cobalt and manganese salts of the C,-C,, fatty acids. Oxidation of the isomeric xylenes by manganese(II1) acetate in the absence of 0, gives methyl substituted benzyl acetates by a mechanism similar to the oxidation of toluene3. (J. A. HOWARD) 1. J. A. Howard, J. H. B. Chenier, J . Am. Chem. Soc., 95, 3054 (1973). 2. W. Dumas, W. Bulani, Oxidation of Petrochemicals: Chemistry and Technology, Wiley, New York, 1972. 3. E. I. Heiba, R. M. Dessau, W. J. Koehl, Jr., J . Am. Chem. Soc., 91, 138 (1969).


14.8. Bioinorganic Catalysis 14.8.1. Introduction Additional sections which deal with bioinorganic catalysis are projected for publication in a final Supplementary unit of the Inorganic Reactions and Methods series.

14.8.2. Cobalamin Reactions 14.8.2.1. Cobalamin Models 14.8.2.1 -1. Formatlon of the Cobalt-Carbon Bond.

Coenzyme B,,', methylcobalamin (Fig. l a and b) and closely related derivatives were the first organometallic compounds to be found in nature. Their complex structures, together with that of vitamin B,, (Fig. lc), were elucidated by X-ray crystallography. After isolation of the B,, coenzyme it was realized that the cobalt-carbon (Co-C) bond could be stabilized by other ligands. Many organocobalt compounds were subsequently synthesized and used as models for the alkylcobalamins. The most commonly used are cobaloximes and Costa complexes. The former has two dimethylglyoximate monoanions (dmgH) with a general chemical composition of alkylcobalt(II1) (dmgH),(base) (Q3, the latter has a diacetylmonoximeimino diacetylmonoximato imine propane monoanion [(DO) (DOH)pn] with a general chemical composition of alkylcob-

-NH2 OH OH a.R= H H H b. R = CH, c. R = CN CH,OH Figure 1. (a) Vitamin B,, coenzyme; (b) methylcobalamin; (c) vitamin B I Z . 408

14.8.2. Cobalamin Reactions 14.8.2.1. Cobalamin Models 14.8.2.1.1. Formation of the Cobalt-Carbon Bond.

409

alt(II1) [(DO) (DOH)pn]X (2)4? Molecular orbital calculations of the cobaloxime and cobalamin based on HMO and modified Wolf-Helmholtz MO techniques indicate that cobalamin and cobaloxime are somewhat similar, but that the cobalt atom in cobalamin has a smaller partial positive charge, which results in a stronger binding of axial ligands and a lower kinetic reactivity of c o b a l o ~ i m e s ~Nevertheless, *~. cobaloximes continue to be widely studied B,, model compound^^^^.

RCo(dmgH), B

RCo[(DO)(DOH)pn]B 2

RCo( salen)B 3

RCo(tim)B

1

R'

:R'

;

R"

a

R'

R'

R

R"

4

' R' R"

R'

R' = alkyl, R" = H R' = H, R" = aryl

RCo(Porp)B 5 Costa-type model compounds, however, have been shown to be a closer mimic of the B,, electrochemical behavior than the cobaloximes, or any other model compounds"-". Hence, they are used more widely as models in kinetic studies relevant to the mechanisms of catalysis mediated by B,,, which involve electron transfer p r o c e ~ s e s ' ~ -The ~ ~ .nonplanarity of the equatorial ligand of (DO) (DOH)pn, caused by propylene group p ~ c k e r ' ~ .also ' ~ , appears to mimic the distortion in the corrin ring.

41 0

14.8.2. Cobalamin Reactions 14.8.2.1. Cobalamin Models 14.8.2.1. I . Formation of the Cobalt-CarbonBond.

Imine bases (Schiff bases)” such as bis(salicyla1dehyde)ethylene diamine (salen) (3), are planar tetradentate ligands used less widely as models of Bl,. Tetraazamacrocyclic ligands, such as 2,3,9,10-tetramethyl-1,4,8,1l-tetraaza-cyclotetradeca-l,3,8,10tetraene (tim) (4)19-”and porphyrinsZ3(5) are similar. These compounds stabilize the three oxidation states of cobalt: 3, + 2, and 1, each of which can form a Co-C

+

+

bond. Reactions of Co(II1) with carbanions, Co(I1) with alkyl radicals, and Co(1) with electrophilic reagents give a Co-C bondz4:

+ RBizr + R’ BIZs + R + Biza

-

RBI2

(a)

RBl2

(b)

RBI2

(c)

Both cobaloxime(1) and cob(1)alamin have been described as very strong nucleop h i l e ~ Nucleophilicity ~ ~ ~ ~ ~ . (n) has been defined as nMeI = log(ky/kMeOH), where ky and kMeOH are the second-order specific rate constants for attack by a nucleophile Y and methanol, respectively, on the substrate CH,I at 25°C in methanolic solution. The observed values of n are 14.4 for cob(1)alamin and 14.3 for cobaloxime(1). For comparison, n for MeO- is 6.3.Because of their high nucleophilicity, these Co(1) complexes react with many alkylating agents, a commonly used method for the formation of the Co-C bond. The B,, coenzyme is prepared by reaction of B,2s with the tosylate 6 followed by removal of the protecting group [equation (d)]”.

X

ki

TsOCHz

1. *I,,

( N T N

N NH,

2. H,O+

~

E{TN (4

B1zCH2

NH2

6 Dicyanocobyrinic acid heptamethyl ester (Fig. 2) is useful because of its relative simplicity compared to cobalamins, and because of its high solubility in organic solvents. It is prepared by reaction of vitamin B,, with MeOH/H,SO, and then with KCNZ8.Like B,, and c o b a l o ~ i m e s ~it ~is, reduced to Co(1) with NaBH, and alkylated by alkyl halides”. Alkylcobaloximes can be prepared directly from Co(I1) chloride and dimethylglyoxime. However, more reproducible results and purer products are obtained when the preformed bromobis(dimethylglyoximato)(dimethyl sulfide)cobalt(III) is reduced with N,BH, and then alkylated3’. Dimethyl sulfide is easily exchanged, so other base-substituted cobaloximes may be prepared3’. Reaction of Co(1) with alkylating agents is generally described as a nucleophilic substitution (S,2)25:


41 1

0

OCH,

-

-

Figure 2. Dicyanocobyrinicacid heptamethyl ester.

[Co(dmgH),B]-

+ RX

z

B(dmgH),Co'-

-

. . . R . . . X'-

RCo(dmgH),B

+ X-

(e)

However, electron transfer has been observed with both Blzs and cobaloxime(I)31-39: [Co(dmgH),B]-

+ RX

X-

+ Co(dmgH),B + R' fast\ RCo(dmgH),B

-

Cobaloxime(1) and Blzsadd to acetylene^^'-^' yielding the tuted isomer or mixtures of both: HCECH

+ BlZs

0-

CH,=CHB,,

(f)

and @carbon substi(8)

Addition to electron deficient olefins gives predominantly the P-carbon substituted isomePO: CH,=CHCN

+ BlZs

B,,CH,CH,CN

(h)

Reduction of Co(I1) or Co(II1) to Co(1) in the presence of acid4, or reduction of Co(I1) with hydrogen at pH7 gives hydridocobalt c o m p l e x e ~ ~The - ~ ~Co(1) . and hydridocobalt species are in equilibrium: [Co(I)]-

+ H,O+

F H-[Co]

+ H,O

(0

Hydridocobalamin decomposes readily to B lZr and hydrogen4,; the stability of hydridocobaloximes depends on the axial base present. Strong .rr-acceptor bases stablize the complexes4'. Hydridotri-n-butylphosphinatocobaloximehas been characterized. Hydridopyridinatocobaloxime is unstable. In protic solvent at neutral to slightly acidic pH, hydridocobalt(II1) species react with alkyl halides to give alkyl Co(II1) species at a compatible rate to the corresponding Co(1) complex4'. The stereochemistry of addition to olefins and alkynes changes as the

412


pH is lowered from neutral to acidic because the reactive species changes from the Co(1) to Co(II1)-H complex. The hydrido complex gives only a-substituted alkylcobaloxime40~47. Hydridocobalamin is more reactive than hydridocobaloxime and reacts with unactivated olefins, alkyl halides, acyl halides and a ~ e t y l e n e ~ ~ Both - ~ O . primary and secondary alkyl cobalamin, previously inaccessible via BlZs,are now prepared from hydridoc~balamin~~. Cobaloxime(I1) and BlZr couple with alkyl radicals to form organocobalt

corn pound^^'^^^: Me,COOH

Me,CO'

+ HO' % Me,CO + BlZa + Me' % MeB,,

(j) Cobalt(I1) also abstracts a halogen atom from alkyl halides to form halocobalt compounds and an alkyl radical which reacts with CO(II)~,:

RX

--+

+ Co(dmgH,)B

--+

R

. . . X'- . . . Co'+

(dmgH),B XCo(dmgH),B R'

+

--

RCo(dmgH),B

(k) The reaction rate increases as the stability of the leaving organic radical increases. The bond dissociation energy and axial ligand are also important factors54.Since equimolar amounts of organo and halo complexes form and separation is often difficult, this synthetic method is employed ony when nopther alternative is available. However, halocobaloxime can sometimes be recycled by reduction to the Co(I1) state using Zn wool until all Co(I1) complex is converted to the corresponding o r g a n ~ c o b a l t ( I I I ) ~ ~ * ~ ~ . The mechanism for reaction of Blzrwith alkyl halides5' is shown in scheme 1. X = C1, Br

X = I

+ RX BIzr + R' XB,, + H,O BlZr

BIZr

+ RI

B,,*RI

+ BIZr

--

XB,,

+ R'

RBI,

--

B1,;RI RBI,

+ X+ BIZa + I -

Scheme 1.

Reaction of alkyl Grignard and other organometallic reagents with Co(II1) species was originally used for formation of the Co-C bond3, in particular for the preparation of alkylcobalamin analogueP. Bromopyridinatocobaloxime and B,, react with electron-rich olefins, e.g., vinyl ethers, t o form organocobalt compound^^^*^^. (L. Y. XIE, P. F. ROUSSI, D. H. DOLPHIN)

1. The skeleton of vitamin B,, (i.e., the porphyrin nucleus minus (2-20) is called corrin. The compound containing the corrin nucleus is called a corrinoid. The compound containing the cobalt atom and the standard side chains in the free acid form is called cobyrinic acid, but cobyric acid when the side chains are at positions a, b, c, d, e, g, are in the amide form. Cobyrinic acid substituted with D-1-amino-2-propanol at position f is called cobinic acid. The substituted cobyric acid is called cobinamide. Cobinic acid substituted with D ribofuranose-3phosphate at position 2 of the aminopropanol is called cobamic acid; the substituted cobinamide is called cobamide. Many B,, vitamins and derivatives in which the heterocyclic base is 5,6dimethylbenzimidazole are given the trivial name cobalamin; see D. Dolphin, ed., B,,, WileyInterscience, New York, 1982.

14.8.2. Cobalamin Reactions 14.8.2.1. Cobalamin Models Formation of the Cobalt-Carbon Bond. 14.8.2.1.l.

413

P. G. Lenhert, D. C. Hodgkin, Vitamin B,, and Intrinsic Factor, Enke-Verlag, Stuttgart, 1962. G. N. Schrauzer, J. Kohnol, Chem. Ber., 97, 3056 (1964). G. Costa, G. Mestroni, E. de Savorgnani, Inorg. Chim. Acta, 3, 323 (1969). G. Costa, G. Mestroni, G. Tauzer, J . Chem. SOC., Chem. Commun., 450 (1972). G. N. Schrauzer, L. P. Lee, J. W. Sibert, J . Am. Chem. SOC., 92,2997 (1970). J. Halpem, F. T. T. Ng, G. L. Rempel, J . Am. Chem. SOC., 101,7124 (1979). N. Bresciani-Pahor, M. Forcolin, L. G. Marzilli, L. Randaccio, M. F. Summers, P. J. Toscano, Coord. Chem. Rev.,63, 1 (1985). 9. L. Randaccio, N. Bresciani-Pahor, E. Zangrando, L. G. Marzilli, Chem. SOC. Rev.,18, 225

2. 3. 4. 5. 6. 7. 8.

(1989). 10. C. M. Elliott, E. Hershenhart, R. G. Finke, B. L. Smith, J . Am. Chem. SOC.,103,5558 (1981). 11. R. Seeber, R. Marassi, W. 0. Parker, Jr., G. Kelley, Inorg. Chim. Acta, 168, 127 (1990). 12. R. G. Finke, W. P. McMenna, D. A. Schiraldi, B. L. Smith, E. Pierpoint, J . Am. Chem. SOC., 105, 7592 (1983). 13. B. E. Daikh, J. E. Hutchison, N. E. Gray, B. L. Smith, T. J. R. Weakley, R. G. Finke, J . Am. Chem. SOC., 112,7830 (1990). 14. R. G. Finke, D. A. Schiraldi, J. Am. Chem. SOC.,105,7065 (1983). 15. B. E. Daikh, R. G. Finke,J. Am. Chem. Soc., 113,4160 (1991). 16. N. Bresciani-Pahor, L. Randaccio, E. Zangrando, lnorg. Chim. Acta, 168, 115 (1990). 17. W. 0. Parker, Jr., E. Zangrando, N. Bresciani-Pahor. R. A. Marzilli, L. Randaccio, L. G. Marzilli, Inorg. Chem., 27, 2170 (1988). 18. G. Costa, Pure Appl. Chem., 30, 335 (1972). 19. S . C. Jackels, K. Farmery, E. K. Barefield, N. J. Rose, D. H. Busch, Inorg. Chem., 11, 2893 (1972). 20. C. Y.Mok, J. F. Endicott, J . Am. Chem. SOC., 100, 123 (1978). 21. J. F. Endicott, J. Lillie, J. M. Kusza, B. S . Ramaswamy, W. G . Schmonsees, M. G. Simic, M. D. Glick, D. P. Rillema, J . Am. Chem. SOC., 99, 429 (1977). 22. A. Bakac, J. H. Espenson, Inorg. Chem., 28,4319 (1989). 23. M. K. Geno, J. Halpem, J . Am. Chem. SOC.,109, 1238 (1987). 24. B,,, = aquo or hydroxocobalamin(II1); Blzr = aquo or hydroxocobalamin(I1); Blzs = aquo or hydroxocobalamin(1). 25. G. N. Schrauzer, E. Deutsch, J . Am. Chem. Doc., 91, 3341 (1969). 26. G. N. Schrauzer, E. Deutsch, R. J. Windgassen, J . Am. Chem. SOC., 90,2441 (1968). 27. E. L. Smith, L. Mervyn, P. W. Muggleton, A. W. Johnson, N. Shaw, Ann. NYAcad. Sci., 112, 565 (1964). 28. D. Dolphin, D. J. Halko, R. B. Silverman, Inorg. Synth., 20, 134 (1980). 29. D. Dolphin, in Methods in Enzymology, Vol. 18, D. B. McCormick, L. D. Wright, ed., Academic Press, New York, 1971, p. 34. 30. J. Bulkowski, A. Cutler, D. Dolphin, R. B. Silverman, Inorg. Synth., 20, 127 (1980). 31. J. Schaffler, J. Rbtey, Angew. Chem., Int. Ed. Engl., 17, 845 (1978). 32. M. Tada, M. Okabe, Chem. Lett., 201 (1980). 33. R. Breslow, P. L. Khanna, J . Am. Chem. SOC.,98, 1297, 6765 (1976). 34. A. I. Scott, J. B. Hansen, S . K. Chung, J. Chem. SOC., Chem. Commun., 388 (1980). 35. M. Okabe, M. Tada, Bull. Chem. SOC.Jpn., 55, 1498 (1982). 36. M. Okabe, M. Tada, Chem. Lett., 831 (1980). 37. R. Scheffold, S . Abrecht, R. Orlinski, H. R. Ruf, P. Stamouli, 0. Tinembart, C. Walder, C. Weymoth, Pure Appl. Chem., 59,363 (1987). 38. H. Bhandal, V. F. Patel, G. Pattenden, J. J. Russell, J . Chem. SOC., Perkins Trans I , 2691 (1990). 39. V. F. Patel, G . Pattenden, J . Chem. SOC.,Perkins Trans. I , 2703 (1990). 40. G. N. Schrauzer, R. J. Windgassen, J . Am. Chem. SOC., 89, 1999 (1967). 41. M. D. Johnson, B. S . Meeks, J . Chem. SOC.B, 185 (1971). 42. K. N. V. Duong, A. Gaudemer, J . Organomet. Chem., 22,473 (1970). 43. G . N. Schrauzer, R. J. Holland, J . Am. Chem. Soc., 93,4060 (1971). 44. L. I. SimBndi, Z. Szeverbnyi, E. Bud6-ZBhonyi, lnorg. Nucl. Chem. Lett., 11, 773 (1975). 45. L. I. Simiindi, E. Bud6-Ziihonyi, Z. SzeverBnyi, Inorg. Nucl. Chem. Lett., 12,237 (1976). 46. L. I. Simhndi, E. Bud6-ZBhonyi, Z. Szeverhyi, S . Nbmeth, J . Chem. SOC.,Dalton Trans., 276 (1980).

414

14.8.2. Cobalamin Reactions 14.8.2.1. Cobalamin Models 14.8.2.1.2. Cleavage of the Cobalt-Carbon Bond.

47. G. N.Schrauzer, R. J. Holland, J . Am. Chem. SOC., 93, 1505 (1971). 48. J. H. Grate, G. N. Schrauzer, J . Am. Chem. Soc., 101, 4601 (1979). 49. A.Fischli, P. M. Muller, Helv. Chim. Acta, 63, 529 (1980). 50. S. M. Chemaly, J. M. Pratt, J . Chem. Soc., Dalton Trans., 595 (1984). 51. J. H.Espenson, A. H. Martin, J . Am. Chem. Soc., 99,5953(1977). 52. G. N.Schrauzer, J. W. Sibert, R. J. Windgassen, J . Am. Chem. Soc., 90,6681(1968). 53. J. Halpem, P. F. Phelan, J . Am. Chem. Soc., 94, 1881 (1972). 54. D. Dodd, M. D. Johnson, J . Organomet. Chem.,52, 1 (1973). 55. P. F. Roussi, D. A. Widdowson, J. Chem. SOC.,Perkins, Trans. I , 1025 (1982). 56. P. R. Roussi, D. A. Widdowson, J . Chem. SOC.,Chem. Commun., 810 (1979). 57. H.U.Blaser, J. Halpern, J . Am. Chem. Soc., 102, 1684 (1980). 58. F. Wagner, K.Bernhauer,Ann. NY Acad. Sci., 112,580(1964). 59. R. B. Silverman, D. Dolphin, J . Am. Chem. SOC.,95, 1686 (1973). 60. R. B. Silverman, D. Dolphin, J. Am. Chem. Soc., 96,7094 (1974). 14.8.2.1.2. Cleavage of the Cobalt-Carbon Bond.

The Co-C

bond can be cleaved either homolytically:

+ [Co"] or heterolytically + R + + [Co'lR-[Co] + R

+

(a) (b)

+ R[Co"']+ (c) (a) Hornolytic Cleavage. Cleavage of the Co-C bond can be achieved both thermally and photochemically. For most alkylocobalamins and alkylcobaloximes the Co-C bond dissociation energies (BDE) are between 84 and 142 kJ/mol, so homolysis occurs at relatively low temperat~rel-~. The Co-C bond in alkylcobalamins is generally weaker than in alkylcobaloximes.8~9Dissociation energies of the Co-C bond in coenzyme B 12 and methylcobalamin are 1334*7,'0and 155 kJ/mo15. A noticeably lower Co-C bond dissociation energy (BDE) for coenzyme B12, 109 kJ/mol has also been independently reported' '-13. The 23 kJ/mol difference was attributed to solvent viscosity effects' (pH 4.3 H,O vs ethylene glycol). However, the BDE of the coenzyme measured in H,O at pH 7.0 is 125 k J / m ~ l ' ~which , is only 6.3 kJ/mol lower than the value measured in ethylene g l y c 0 1 ~Adenine ~'~ and Co(II1) BlZawere observed as products in an independent experiment carried out in H,O at pH 4.3, indicative of a heterolysis The difference (23 kJ/mol) in the coenzyme Co-C BDE, as suggested, originated both from (1) the complication of a partial heterolytic cleavage of the Co-C bond which expectedly results in a lower activation energy as opposed to a homolytic cleavage, and (2) a different solvent viscosity". Alkylcobalamins and alkylcobaloximes are more stable than expected because of the recombination of the alkyl radical with the Co(I1) species. This is especially apparent with MeB,, and MeCo(dmgH),Py, since elimination of the phydrogen to form an olefin is not possible for the methyl radical. Radical scavengers increase the apparent rate of both thermolysis and photolysis by intercepting the back reaction. l6 The thermolysis rate depends mainly on the stability of the organic radical. Steric effects are also important; secondary cobalamins are thermally less stable than their primary counterparts. Products of the photolysis of alkylcobalt complexes depend on the organic radical and whether anaerobic or aerobic conditions are used. The B,, coenzyme forms the organic radical 1 which, in the absence of 0,, yields cyclic 2'' and in the presence of 0, oxidizes to 3a and 3b (Scheme 1)17,18


414


47. G. N.Schrauzer, R. J. Holland, J . Am. Chem. SOC., 93, 1505 (1971). 48. J. H. Grate, G. N. Schrauzer, J . Am. Chem. Soc., 101, 4601 (1979). 49. A.Fischli, P. M. Muller, Helv. Chim. Acta, 63, 529 (1980). 50. S. M. Chemaly, J. M. Pratt, J . Chem. Soc., Dalton Trans., 595 (1984). 51. J. H.Espenson, A. H. Martin, J . Am. Chem. Soc., 99,5953(1977). 52. G. N.Schrauzer, J. W. Sibert, R. J. Windgassen, J . Am. Chem. Soc., 90,6681(1968). 53. J. Halpem, P. F. Phelan, J . Am. Chem. Soc., 94, 1881 (1972). 54. D. Dodd, M. D. Johnson, J . Organomet. Chem.,52, 1 (1973). 55. P. F. Roussi, D. A. Widdowson, J. Chem. SOC.,Perkins, Trans. I , 1025 (1982). 56. P. R. Roussi, D. A. Widdowson, J . Chem. SOC.,Chem. Commun., 810 (1979). 57. H.U.Blaser, J. Halpern, J . Am. Chem. Soc., 102, 1684 (1980). 58. F. Wagner, K.Bernhauer,Ann. NY Acad. Sci., 112,580(1964). 59. R. B. Silverman, D. Dolphin, J . Am. Chem. SOC.,95, 1686 (1973). 60. R. B. Silverman, D. Dolphin, J. Am. Chem. Soc., 96,7094 (1974). 14.8.2.1.2. Cleavage of the Cobalt-Carbon Bond.

The Co-C

bond can be cleaved either homolytically:

+ [Co"] or heterolytically + R + + [Co'lR-[Co] + R

+

(a) (b)

+ R[Co"']+ (c) (a) Hornolytic Cleavage. Cleavage of the Co-C bond can be achieved both thermally and photochemically. For most alkylocobalamins and alkylcobaloximes the Co-C bond dissociation energies (BDE) are between 84 and 142 kJ/mol, so homolysis occurs at relatively low temperat~rel-~. The Co-C bond in alkylcobalamins is generally weaker than in alkylcobaloximes.8~9Dissociation energies of the Co-C bond in coenzyme B 12 and methylcobalamin are 1334*7,'0and 155 kJ/mo15. A noticeably lower Co-C bond dissociation energy (BDE) for coenzyme B12, 109 kJ/mol has also been independently reported' '-13. The 23 kJ/mol difference was attributed to solvent viscosity effects' (pH 4.3 H,O vs ethylene glycol). However, the BDE of the coenzyme measured in H,O at pH 7.0 is 125 k J / m ~ l ' ~which , is only 6.3 kJ/mol lower than the value measured in ethylene g l y c 0 1 ~Adenine ~'~ and Co(II1) BlZawere observed as products in an independent experiment carried out in H,O at pH 4.3, indicative of a heterolysis The difference (23 kJ/mol) in the coenzyme Co-C BDE, as suggested, originated both from (1) the complication of a partial heterolytic cleavage of the Co-C bond which expectedly results in a lower activation energy as opposed to a homolytic cleavage, and (2) a different solvent viscosity". Alkylcobalamins and alkylcobaloximes are more stable than expected because of the recombination of the alkyl radical with the Co(I1) species. This is especially apparent with MeB,, and MeCo(dmgH),Py, since elimination of the phydrogen to form an olefin is not possible for the methyl radical. Radical scavengers increase the apparent rate of both thermolysis and photolysis by intercepting the back reaction. l6 The thermolysis rate depends mainly on the stability of the organic radical. Steric effects are also important; secondary cobalamins are thermally less stable than their primary counterparts. Products of the photolysis of alkylcobalt complexes depend on the organic radical and whether anaerobic or aerobic conditions are used. The B,, coenzyme forms the organic radical 1 which, in the absence of 0,, yields cyclic 2'' and in the presence of 0, oxidizes to 3a and 3b (Scheme 1)17,18

14.8.2. Cobalamin Reactions 14.8.2.1. Cobalamin Models 14.8.2.1-2. Cleavage of the Cobalt-Carbon Bond.

41 5

Methyl radicals from methylcobaloxime and methylcobalamin abstract a hydrogen atom to give methane, or dimerize to form ethane. Often the radical scavenges hydrogen atoms and methyl groups originate from the ligands of the cobalt complexes'9320. With a P-hydrogen is in the alkyl group, homolytic fission generates an caged [Co(II) R ] pai?'. The Co(I1) complex abstracts a H atom to form an olefin and the hydridocobalt complex (Scheme 2)20-22.With radical scavenger such as O2is present, a peroxy alkyl radical is formed that reacts further with the Co(I1) complex (Scheme 2)'6*23-25. The Co-C bond can also be cleaved homolytically by radical acceptors such as C O ( I I ) ~ ~ , Cr(II)27328,Sn(II)29930, and A U ( I I I ) ~ ' ,The ~ ~ .mechanism is a homolytic substitution (S,2), although a redox S,2 mechanism cannot be ruled out. (b) Cleavage by Electrophiles. The Co-C bond of simple alkyl Co(II1) complexes resists acid c l e a ~ a g e ~The ~ , ~Co-C ~ . bond of alkylcobalt(II1) compounds, which have an oxygen substituent on the P-C position of the alkyl ligand, is more susceptible to acid and to solvent induced c l e a v a g e ~ ~P-Hydroxyethyl ~.~~. and P-methoxyethyl cobalamins, on addition of acid, give the same Blza and ethylene products36.The results from model compounds studies suggest strongly the formation of a "wcation" intermediate by H + electrophilic a t t a ~ k ~ ~ - ~ ~ .

+

OH OH I

I

2

1

NH,

NH2

a. X = CHO b. X = C02H

Scheme 1.

R'

Olefin

+ [Co-+H Scheme 2.

-

[Co"]

+ YzH2

B(dmgH),Co -OOR

41 6


The reaction of alkylcobalamins with metal electrophiles, such as Hg(II), has been studied extensively since MeB,, may be responsible for formation of highly toxic methylmercury(I1) compounds under environmental condition^^-^^. The reaction mechanism has been studied for both alkylcobaloximes""~45and a l k y l c ~ b a l a m i n s ~Reaction -~~. of alkylcobaloximes with Hg(I1) salts can be described as a bimolecular electrophilic substitution (S,2), which takes place with inversion of configurations0. The reaction of alkylcobalamins is more complicated, since Hg(I1) coordinates to 5,6-dimethylbenzimidazoleto form a base-off complex:

RB,,-Hg(II) (d) RBI2 + Hg(I1) The extent of equilibrium in equation (d) depends on the Hg(I1) salt and the anions present, since only the Hg(I1) ion complexes with the base47.48.Both base-on and baseoff complexes react with Hg(I1) salts, but the former reacts f a ~ t e p . ~ ~ . Similar reactions are observed with T1(III)45J1and Pd(II)52*53; Pd(1I) also coordinates 5,6-dimethylbenzimidazolebut kinetic studies show that the base-on complex is the only reactive speciess3. Platinum(1V) also reacts with alkylcobalamins but Pt(I1) is required as catalyst suggesting a redox m e ~ h a n i s m ~ ~ - ~ ~ ~ ~ ~ . (c) Nucleophilic Cleavage. Some P-substituted alkycobaloximes and alkylcobalamins react with bases to form olefins and a Co(1) complexs6: XCH~CHJCO]

+ OH-

FX C H S H ,

+ H2O + [CO]-

(e)

This side reaction, during the preparation of alkylcobalt complexes from Co(1) species, lowers the yield. wComplexes are formed between electron deficient olefins and Co(1) species under strongly basic conditions". P-Carboxyethylcobalt(II1)and P-methoxyethylcobalt(II1)compounds are stable in base, but P-hydroxyethylcobalt(II1) compounds react to form an aldehyde (or ketone) and CO(I)'~: HOCH,CH,B,, MeCH(0H) CH,Co(dmgH) ,Py

+ OH+ OH-

+ MeCHO +B,,, + MeCOMe

+

+ H,O

[Co(dmgH) ,Py)]-

(f) (g)

Epoxides are not formed, a reaction that does occur with the analogous P-hydroxyethyl halides. Reaction of coenzyme B,, with both OH- and CN- ions has been studied (Scheme 3)'7*59*60.Cyanide may attack the coenzyme from below to substitute the axial base. Attack from above, assisted by synchronous cleavage of the C ( 3 - 0 bond, may cleave the C o - C bond. This would explain the absence of reaction when the ring oxygen is substituted by a methylene group in th furanoside ring". Adenine and cyanocobalamin (initially in its dicyano form) along with cyanohydrin (4) are obtained from reaction of the coenzyme with the CN- ion17. The primary alkyl group of Co(II1) species can be transferred to a Co(1) species under basic conditions26.However, alkyl group transfer from Co(II1) to Co(I1) complexes is more commonly achieved where one electron transfer via a saturated carbon bridge may be i n ~ o l v e d ~ l - ~ ~ . Dealkylation of alkyl cobalt species by thiolate was suggested as a reductive cleavage process, involving reduction of alkylcobalamin with thiol to form a trans axial thiol c o m p l e ~ ~This . ~ ~accounts . for the low yield of thioether%. Carbon-13 and 31Pstudies


41 7

CH2=CHCH(OH)CH(OH)CH(OH)CN 4

(X=O),HO-

B12S + H2C

N no reaction Scheme 3.

NH2

show a strong pH dependence, indicating that thiolate anion is the nucleophile; a direct attack of Co-C bond by thiolate without prior coordination of the thiolate is The alkyl group is transferred to thiol yielding thioether and cob(I1)alamin. The higher alkylcobalamin reacts more slowly than the methyl derivative. Cleavage of the Co-C bond in carboxymethylcobalamin by thiols is more complex, involving both a nucleophilic attack by thiolate and a reductive cleavage of the Co-C bond. At low pH, cobalamin is in the base-off form, facile reductive cleavage gives acetate and cob(I1)alamin as products. At high pH (pH > 8), S-(carboxymeThe presence of base-on (pH > 8) thy1)mercaptoethanol and cob(I1)alamin and base-off (pH < 7) forms may be responsible for the different pH dependent reactivities68.

(L.Y. XIE, P. F. ROUSSI, D. H. DOLPHIN) 1. F. T. T. Ng, G . L. Rempel, J. Halpem, J . Am. Chem. SOC., 104,621 (1982); F. T. T. Ng, G . L. Rempel, J. Halpem, Inorg. Chim. Acta, 77, L165 (1983); J. Halpem, F. T. T. Ng, G . L. Rempel, J . Am. Chem. SOC., 101,7124 (1979). 2. D. Dodd, M. D. Johnson, J . Organomet. Chem., 52, 84 (1973). 3. G . N. Schrauzer, J. H. Grate, J . Am. Chem. SOC., 103,541 (1981). 4. B. P. Hay, R. G . Finke, J . Am. Chem. SOC., 109, 8012 (1987). 5. B. D. Martin, R. G . Finke, J. Am. Chem. SOC., 112,2419 (1990). 6. P. J. Toscano, A. L. Seligson, M. T. Curran, A. T. Skrobutt, D. C. Sennenberger, Inorg. Chem., 28, 166 (1989). 7. R. G . Finke, B. P. Hay, Inorg. Chem., 23, 3041 (1984). 8. P. G . Lenhert, D. C. Hodgkin, Vltamin BI2and Intrinsic Factor, Enke-Verlag, Stuttgart, 1962. 9. G . N. Schrauzer, J. Kohnle, Chem. Ber., 97,3056 (1964). 10. R. G . Finke, B. P. Hay, Polyhedron, 7, 1469 (1988). 11. M. K. Geno, J. Halpem, J. Chem. SOC., Chem. Commun., 1052 (1987). 12. J. Halpem, Science, 227, 869 (1985). 13. J. Halpem, S-H. Kim, T. W. Leung,J. Am. Chem. SOC., 106, 8317 (1984). 14. B. P. Hay, R. G . Finke, J . Am. Chem. SOC., 108, 4820 (1986). 15. R. G . Finke, in Molecular Mechanisms in Bioorganic Processes, C. Bleasdale, B. T.Golding, eds., Royal Society of Chemistry, Cambridge, 1990, p . 244. 16. J. Halpem, in B,,, Vol. 1, D. Dolphin, ed., Wiley-Interscience, New York, 1982, p. 501. 17. A. W. Johnson, N. Shaw, J . Chem. SOC., 4608 (1962). 18. H. P. C. Hogenkamp, J . Biol. Chem., 238,477 (1963).

41 8

14.8.2. Cobalarnin Reactions 14.8.2.1. Cobalamin Models 14.8.2.1.2.Cleavage of the Cobalt-Carbon Bond.

19. G. N. Schrauzer, L. P. Lee, J. W. Sibert, J . Am. Chem. Soc., 92, 2997 (1970). 20. G. N. Schrauzer, J. W. Sibert, R. J. Windgassen, J . Am. Chem. Soc., 90, 6681 (1968). 21. D. A. Baldwin, E. A. Betterton, S . M. Chemaly, J. M. Pratt, J . Chem. Soc., Dalton Trans., 1613 (1985). 22. D. Dolphin, A. W. Johnson, R. Rodrigo, J . Chem. SOC., 3186 (1964). 23. J. Deniau, A. Gaudemer, J . Organomet. Chem., 191, C1 (1980). 24. V. F. Patel, G. Pattenden, J . Chem. Soc., Perkins Trans. I , 2703 (1990). 25. C. Bied-Charreton, A. Gaudemer, J . Organomet. Chem., 124, 199 (1977). 26. D. Dodd, M. D. Johnson, B. L. Lockmann, J . Am. Chem. Soc., 99, 3664 (1977). 27. J. H. Espenson, J. S . Shveima, J . Am. Chem. Soc., 95,4468 (1973). 28. J. H. Espenson, T. D. Sellers, Jr., J . Am. Chem. Soc., 96, 94 (1974). 29. L. J. Dijikes, W. P. Ridley, J. M. Wood, J . Am. Chem. Soc., 100, 1010 (1978). 30. Y. T. Fanchiang, J. M. Wood,J. Am. Chem. SOC.,103,5100 (1981). 31. J. M. Wood, in B,,, Vol. 2, D. Dolphin, ed., Wiley-Interscience, New York, 1982, p. 151. 32. J. M. Wood, Y. T. Fanchiang, in Vitamin B,,, Proceedings of the Third European Symposium, B. Bajalak, W. Friedrick, eds., Walter de Gruyter, New York, 1979, p. 539. 33. R. M. McAllister, J. H. Weber, J . Organomet. Chem., 77, 91 (1974). 34. V. E. Magnuson, J. H. Weber, J . Organomet. Chem., 92, 233 1975. 35. A. W. Johnson, N. Shaw, J . Chem. Soc., 4608 (1962). 36. H. P. C. Hogenkamp. J. E. Rush, C. A. Swenson, J . Biol. Chem., 240, 3641 (1965). 37. K. L. Brown, S . Ramaurthy, Organometallics, I , 413 (1982). 38. J. H. Espenson, D. Wang, Znorg. Chem., 18, 2853 (1979). 39. K. L. Brown, M. M. L. Chu, L. L. Ingraham, Biochemistry, 15, 1402 (1976). 40. J. M. Wood, F. S . Kennedy, C. G. Rosen, Nature (London), 220, 173 (19680. 41. J. M. Wood, Science, 183, 1049 (1974). 42. W. P. Ridley, L. J. Dijikes, J. M. Wood, Science, 197, 329 (1977). 43. S. Jensen, A. Jemelov, Nature (London), 223,753 (1969). 44. A. Adin, J. H. Espenson, J . Chem. Soc., Chem. Cornmun., 653 (1971). 45. P. Abley, E. R. Dockal, J. Halpem, J . Am. Chem. SOC.,95,3166 (1973). 46. R. E. DeSimone, M. W. Penley, L. Charbonneau, S . G. Smith, 3. M. Wood, H. A. 0. Hill, J. M. Pratt, S . Ridsdale, R. J. P. Williams, Biochim. Biophys. Acta, 304, 851 (1973). 47. C. W. V. Chu, W. D. Gruenwedel, Bioinorg. Chem., 7, 169 (1977). 48. P. J. Craig, S . F. Morton, J . Organomet. Chem., 145, 79 (1978). 49. G. C. Robinson, F. Nome, J. H. Fendler, J . Am. Chem. SOC., 99, 4969 (1977). 50. H. L. Fritz, J. H. Espenson, D. A. Williams, G. A. Molander, J . Am. Chem. Soc., 96, 2378 (1974). 51. G. Agnes, S. Bendle, H. A. 0. Hill, F. R. Williams, R. J. P. Williams, J . Chem. SOC., Chem. Cornmun., 850 (1971). 52. J. Y. Kim, H. Yamamoto, T. Kwan, Chem. Pharm. Bull., 23, 1091 (1975). 53. W. M. Scovell, J . Am. Chem. SOC., 96, 3451 (1974). 54. R. T. Taylor, M. L. Hanna, Bioinorg. Chem., 6, 281 (1976). 55. R. T. Taylor, M. L. Hanna, J . Environ. Sci. Health Part A , 11,201 (1976). 56. D. Dodd, M. D. Johnson, J . Organomet. Chem., 52,52 (1973). 57. G. N. Schrauzer, J. H. Weber, T. M. Beckham, J . Am. Chem. SOC., 92,7078 (1970). 58. G. N. Schrauzer, R. J. Windgassen, J . Am. Chem. Soc., 89, 143,4250 (1967). 59. S. S. Kenvar, T. A. Smith, R. H. Abeles, J . Biol. Chem., 245, 1169 (1970). 60. H. P. C. Hogenkamp, in B,,, vol. 1, D. Dolphin, ed., Wiley-Interscience, New York, 1982, p. 295. 61. F. J. Endicott, K. P. Balakrishnan, C-L. Wong, J . Am. Chem. Soc., 102,5519 (1980). 62. B. Kraiitler, M. Hughes, C. Caderas, Helv. Chim. Acta, 69, 1571 (1986). 63. D. Dodd, M. D. Johnson, J . Chem. Soc., Chem. Cornmun., 1371 (1971). 64. G. N. Schrauzer, J. A. Seck, T. M. Beckham, Bioinorg. Chem., 2,211 (19730. 65. D. W. Jacobsen, L. S . Troxell, K. L. Brown, Biochemistry, 23, 2017 (1984). 66. G. N. Schrauzer, E. A. Stadlbauer, Bioinorg. Chem., 3, 353 (1974). 67. H. P. C. Hogenkamp, G. T. Bratt, S . Sun, Biochemistry, 24,6428 (1985). 68. H. P. C. Hogenkamp, G. T. Bratt, A. T. Kotchevan, Biochemistry, 26,4723 (1987).


14.8. Bioinor anic Catalysis 14.8.2. Cobasamin Reactions 14.8.2.2. Cobalamin-CatalyzedEnzymatic Reactions

419

14.8.2.2. Cobalamin-Catalyzed Enzymatic Reactions

Cobalamin-catalyzed reactions are generally classified into two groups: methylcobalamin-dependent reactions (Table 1, entry 1 to 3) and coenzyme B,,-dependent rearrangements (Table 1, entry 4 to 11). The first group includes the biosynthesis of methionine from homocysteine, the reduction of CO, to acetic acid via an acetyl-CoA pathway, and the biosynthesis of CH, also via an acetyl-CoA pathway'. In the synthesis of methionine, cob(1)alamin is methylated by methyl tetrahydrofolate [MeH,THFZ (equation (a)], followed by the transfer of CH3+ from methylcobalamin to hom~cysteine'.~. HSCH,CH,CH(NH,)CO,H

+ MeH,THF

-

CH,SCH,CH,CH(NH,)CO,H

+ H,THF

(a)

The biosynthesis of acetate was believed to form a similar methyl corrinoid which transfer CH,- to CO, (carboxylation of methyl~obalamin)~. However a reduction pathway involving methylcorrinoid and CO via formation of an acetyl radical (carbonylation) has recently been proposed5. In this process, one molecule of CO, is reduced to the methyl level and appears as MeH,THF. Again methylcobalamin, formed by MeH,THF and BIzs, reacts with another CO (derived from CO,), via radical intermediates, followed by recombination of an acetyl radical and B,2r to give acetylc~balamin~-~. However, release of an acetyl group, via reaction of acetylcobalamin with a thiol to give a thioester, analogous to acetyl-CoA, is still not demonstrated. The role of cobalamin in the biosynthesis of CH, is more complex. Cobalamin serves as an electron donor for enzymatic reduction of methyl-CoM to CH.: Dithiothreitol or SnCl, reduces BlZato BlZr,which is known to disproportionate subsequently to Blza and BlZs.The latter then acts as a stronger reducing agent in the reduction. Inhibition of CH, production by propyl iodide is attributed to formation of Pr-Co species, which suppresses electron transfer*. Attempts have been made to mimic the above reactions under nonenzymatic conditions. Methylcobalamin methylates homocysteine, but the reaction is a free radical process". Also, MeH,THF does not methylate cobaloxime(1) or B 12s. N-Methylamines are also unreactive toward Co(1) species'. Tetraalkylammonium compounds alkylate Co(1) nucleophiles to form alkylcobalt compounds' it is possible that MeH,THF is protonated by the enzyme and this methylates the cobalamin. Methylation in low yield of cobalamin(1) by MeH,THF at low pH" has been disputed'. The second group of enzymatic reactions involves coenzyme Bl,. All but one (Table 1, entry 11) can be described as in equation (b).

';

H

X

I I -C1-C,I I

X

H

I I e C1-C,I I

Spectroscopic studies indicate homolytic cleavage of the enzyme bound coenzyme to a deoxyadenosyl radical (ACH,') and BIzras a common characteristic of these enzymatic reactions'. Isomerization, described above, involves an apparent intramolecular 1,Zshift of a hydrogen and an electronegative group (X = OH, NH,, Table 1, entry 4 to 6) or a carbon skeleton (Table 1, entry 7 to 10). However, in the nucleotide reductase system, coenzyme B,, has a unique role of radical initiator in a radical chain mechanism rather

420

14.8. Bioinor anic Catalysis 14.8.2. Cobafamin Reactions 14.8.2.2. Cobalamin-CatalyzedEnzymatic Reactions

TABLE1 (a) (b)

1. Synthesis of methionine 2. Synthesis of methane 3. Synthesis of acetate 4. Dioldehydrase, e.g.,

H OH &OH

/KH + H20 0

5. Ethanolamine ammonia lyase

6. Aminomutase, e.g.,

COZH

-

H

NH2

NH2

COZH He C NH O z H

H 2 N y N H 2 H NHZ

CO,H

H2N

H

H

NHZ

7. Methylmalonyl- CoA mutase

0 CoA- S

H

CH3

8. Glutamate mutase

H NHZ HO2C -CO,H

__*

CO,H H NH, HOzC \cOzH H CH3

9. a-Methyleneglutarate mutase

")$,

HOzC

H G C 0 2 H HozC H CH,

CO2H

10. Isobutyryl-CoA mutase

CH3

I H-C-CH3 I

CH3

I

H-CC-CH~COSCOA

COSCoA

Hg

1 1. Ribonucleotide reductase

+ R(SH)2

01$3H2C

Base

I

H

_*

0,$3HZC

+RS2 + H20 Base

B

14.8. Bioinor anic Catalysis 14.8.2. Coba amin Reactions 14.8.2.2. Cobalamin-CatalyzedEnzymatic Reactions

421

than acting as an intermediary hydrogen transfer Unlike the other 1,Zshift reaction, hydrogen derives from an excogenous thiol reductant. l 5 The migrating atom of the substrate is equilibrated with the two hydrogens of the 5'-carbon atom of deoxyadenosine (ACH, - ) bound to cobalt (Scheme 1). In the reaction in eq. b, ESR experiments show Co(I1) as an intermediate in the these transformation'. Blzr existence has also been confirmed by X-ray crystallography in the solid state16 and EXAFS spectroscopy in ~olution'~. B 12r contains 5-coordinate low-spin Co(I1). Structurally, the Co(I1) comn is strikingly similar to that of the coenzyme B1216917. The axial base of the coenzyme was thought previously to play a key role in Co-C bond weakening by pushing the comn ring upward. The bond dissociation energy (BDE) in coenzyme B,, is only 23 kJ/mol lower than in adenosylcobinamide (base-off form)I8. It is very likely that binding the coenzyme to the enzyme through the benzimidazole base increases the protein interactions both sterically and electronically, so that the bond is readily cleaved to give ACHz"8*'9. ACHZB,

--

+ Biz,. RH ACH3 + R' + Bizr ACH3 + P' + Bizr ACHZ' + PH + Bizr

ACHZ'

RH = substrate

ACH'B,,

PH = product Scheme 1.

The carbon skeleton rearrangements [Table 1, entry 7 to 10; equation (b), X = COSCoA, CHN+H3C0,-, C(=CH,)CO,- ,CH,COSCoA] are reversible. The isobutylCoA mutase reaction was discovered recently; little is yet known about this reaction". Many model systems have been developed for the methylmalonyl-CoA mutase reaction. With simple models, low yields of rearranged product are obtained".". When the thioester 1 reacts with BlZs,up to 70% of the rearranged product 2 form^'^*'^. A free radical pathway involving 3 and 4 was proposedz4.The key intermediate 3 was generated later by reaction of 1 with n-Bu,SnH, in the absence of coenzyme, and 4 was trapped by H'. Therefore 2 is the only product Spontaneous 1,Zmigration of the thioester group was first demonstrated in the model free radical26. Rearrangement in methylmalony-CoA mutase reaction may occur at the free radical stage. Although cobalt participation in the rearrangement cannot definitely be excluded, it seems unlikely". BrCH,C(COSEt)MeCO,Et 1 'CH,C(COSEt)MeCO,Et 3

EtSCOCH,CHMeCQEt 2 EtSCOCH,C'MeCO,Et 4

The principal role of coenzyme B,, in the rearrangement process appears to be as a free radical precursor, a role that depends on the weakness of the cobalt-carbon bond of coenzyme B1225. Both free organic radical and organocobalt pathways have been postulated (Scheme 2) in the rearrangement of 2-methyleneglutarate to methylitaconate. 1,ZMigration of an acrylate group in a free radical intermediate is precedented via a cyclopropyl methyl

422

14.8. Bioinor anic Catalysis 14.8.2. Cobafamin Reactions 14.8.2.2. Cobalarnin-Catalyzed Enzymatic Reactions ~~

radical intermediate27-28. Although cobalt participation cannot be excluded, the free radical mechanism appears more favorable. However, some model studies suggest a cobaltparticipation pathway (Scheme 3). 1-Methylbut-3-enylcobaloxime(5) rearranges to 2methylbut-3-encyclobaloxime(6)29*30, via 3-methylcyclopropylcarbinylcobaloxime, giving a 1:lO equilibrium mixture of compounds 5 and 6 . Similarly, the suggested intermediate 3-methylcyclopropylcarbinylcobaloximerearranges to 5 and 6 to give the same equilibrium mixture31. CH,=CHCH,CH(CH,)Co( dmgH),Py CHz=CHCH(CH3)CH,Co(dmgH),Py

6 5 Rearrangement of glutamic to methylaspartic acid is one of the four rearrangement reactions (Tahle 1, 7 to 10) catalyzed by coenzyme BIZ,which involves carbon skeleton rearrangements. Furthermore, the migrating group is a saturated carbon. Thermolysis of model compound 7 produces 8 and 9;no rearranged glutamic acid was detected3'.

H3N+YC0;

tHz

CHZ

II

(1) HO,C~HCH,CCO,H

HO'CCH-

CCO'H

s

I

HO'CCH-

Scheme 2.

+ 'CH,A

/

co

Scheme 3.

+ 'CHZA

CCO'H

14.8. Bioinor anic Catalysis 14.8.2. Cobasamin Reactions 14.8.2.2. Cobalamin-CatalyzedEnzymatic Reactions

(H3Nf)CH(C0,-)CH(CH3)C0,H

423

(H3N' )CH(CO, - )C(=CH,)CO,H

9

8

Experiments designed to mimic vitamin B holoenzyme by introducing some hydrophobic cavity have been initially successful toward elucidating this rearrangement mechanism. Hydrophobic cage effects and molecular aggregates favor formation of rearranged products, because in both cases the thermolysis-generated free radicals are ~tabilized,~-,~. Many attempts made to induce rearrangement of 10 (Scheme 4) with B,2s were unsuccessful. However, under free radical conditions (initiated by n-Bu,SnH) rearrangement occurs (Scheme 4),'. The above results suggest a free radical mechanism for the enzymatic reactions.

"IN

J3

",CO@

C0,Et

COZEt

Br

0%

G C H ,

n- Bu3SnH

silica

,

5 C H 3

COZEt

10

COZEt

Scheme 4. The dioldehydrase reaction involves formation of aldehyde from 1,2-diols (glycols). The ethanolamine ammonia lyase and aminomutase reactions are similar. The favored mechanism is shown in Scheme 5 . Like other coenzyme catalyzed rearrangements, homolytic cleavage of the Co-C bond of coenzyme B,, produces Blzr and ACH,' and isotopic scrambling of hydrogen at the C5' carbon of the deoxyadenosyl group confirms the involvement of this radical in hydrogen abstraction.

@+

HO OH

I

I

HO OH

ACH,

+ CH2CH2* I I

QT0+fZ 7 ACH,

CHzCH2

-O

-0

CH3CHO -H,O CH,CH(OH), -

T

I

@+

-0 Scheme 5.

ACH,

+ kH,CH(OH),

424

14.8. Bioinor anic Catalysis 14.8.2. Cobagamin Reactions 14.8.2.2. Cobalamin-CatalyzedEnzymatic Reactions

Recent studies have progressed toward understanding the mechanism of the rearrangement step. In a model system (11 in Scheme 6), the vicinal diol group, protected as its carbonate, is activated by MeO- in MeOH; rearrangement reaction yielded 100% Co(11) and 95% CH,CHO (Scheme 6)36. The added axial base 1,5,6-trimethylbenzimidazole drastically alters the observed products (Scheme 6). No CH,CHO or Co(II)CH,CHO, the supposed cobalt participation intermediate, was observed36. Recently a carbonate-protected form of a BIZ-bound l,Zdiol12 was prepared and activated as above. 0

K0

& 0

a * I1

CH30-/CH30H,

+ CH3CH0 + CH30COzCH, + CH,OCO;

1.o

11

6

0.95

0.5

0.5

80

+ CH3OCOZCH3+ HOCHzCHO

CH ,O-/CH ,OH

(CH,),Bz

(CH,),Bz

Scheme 6. Only Blzs and CH,OCO,CH, were detected. As expected from the model system (Scheme 7), no CH,CHO, BlZror formylcobalamin (key intermediate for cobalt participation) were detected37.These products are also obtained in a separate experiment using pulse radiolysis generated HOCH,C'H(OH) radicals in the presence of B 12:8339. The currently available enzymatic and chemical model supporting evidence strongly argues for a pathway involving nonparticipation of the cobalt. Cobalt participation can lead to a redox side reaction37. Rearrangement of an a-hydroxy radical occurs36*40*41.

14.8. Bioinorganic Catalysis 14.8.2. Cobalamin Reactions 1 4.8.2.2. Cobalamin-CatalyzedEnzymatic Reactions

K 0

0

+] -9 cH3"0xo"-

0 CH,O-/CH,OH

425

RO

--0

-0

-0

(R = CH,OCO-)

12

Co(I)--BiZ,

+ CH3OCO2CH3 + HOCHZCHO

0.9

0.95 trace Scheme 7. Computational studies41 of a-OH radicals indicate that protonation of the hydroxy group lowers significantly the barrier to OH migration (Scheme 8), and suggests a possible acidic protein binding site in the enzyme4'. Finally, a yet unproved radical chain mechanism has been proposed4'. OH OH

I

HzC-CHz

H20+

I

HZC-CHZ

-[

1

Ea = 71 kJ/mol

Ea = 33 kJ/mol

*CHz-CHz +OH2

/'.'\,

HZC----CHz

]*

+0h2

1

*CHz-CHz

Scheme 8. (L.Y. XIE, 2.F. ROUSSI, D. H. DOLPHIN)

1. B. T. Golding, in Comprehensive Organic Chemistry, Vol. 5, D. H. R. Barton, W. 0. Ollis, eds., Pergamon Press, Oxford, 1979, p. 549. 2. Where H,THF = tetrahydrofolic acid; MeH,THF = A@-methyltetrahydrofolicacid. 3. R. T. Taylor, in B,,, Vol. 2, D. Dolphin, ed., Wiley-Interscience, New York, 1982, p. 307. 4. L. G. Ljungdahl, H. G. Wood, in B,,, Vol. 2, D. Dolphin, ed., Wiley-Interscience, New York, 1982, p. 165. 5. B. Kraiitler, Helv. Chim. Acta, 67, 1053 (1984). 6. H. G. Wood, S. W. Ragadale, E. Pezacka, Trends Biochem. Sci., 11, 14 (1986). 7. J. G. Zeikus, R. Kerby, J. A. Krzycki, Science, 227, 1167 (1985). 8. D. Ankel-Fuchs, R. K. Thauer, Eur. J. Biochem., 156, 171 (1986). 9. G. Fuchs, FEMS Microbio. Letr., 39, 181 (1986). 10. D. Dodd, M. D. Johnson, J . Organomer. Chem., 52,58 (1973). 11. G . Costa, A. Puxedu, E. Reisenhofer, J. Chem. Soc., Dalton Trans., 2034 (1973). 12. H. Riidiger, Eur. J. Biochem.,21,264 (1971). 13. G. W. Ashley, G. Hams, J. Stubbe, J . Biol. Chem., 261, 3958 (1986). 14. J. Stubbe, Mol. Cell Biochem., 50, 25 (1983). 15. W. S. Beck, Am. J. Hematol., 34, 83 (1990). 16. B. Kraiitler, W. Keller, C. Kratkey, J. Am. Chem. Soc., 111, 8936 (1989). 17. I. Sagi, M. D. Wirt, E. Chen, S. Frisbie, M. R. Chance, J. Am. Chem. Soc., 112,8639 (1990). 18. B. P. Hay, R. G. Finke, J. Am. Chem. SOC., 109, 8012 (1987). 19. B. T. Golding, in B,,, Vol. 1, D. Dolphin, ed., Wiley-Interscience, New York, 1982, p. 543. 20. D. Gani, D. O'Hagan, K. Reynolds, J. A. Robinson, J. Chem. Soc., Chem. Commun., 1002 (1985); K. Reynolds, J. A. Robinson, J. Chem. Soc., Chem. Commun., 1831 (1985); K. Rey-

426

21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

14.8. Bioinor anic Catalysis 14.8.2. Cobafhin Reactions 14.8.2.3. Bioalkylation

nolds, D. Gani, J. A. Robinson,J. Chem. SOC.,Chem. Commun., 1334 (1986);G . Brendelberger, J. Retey, D. M. Ashworth, K. Reynolds, F. Willenbrock, J. A. Robinson, Angew. Chem. Intl. Ed. Engl., 27, 1089 (1988). G . Bidi'lngmaier,U. M. Kempe, T. Krebs, 3. RBtey, H. Flohr, Angew. Chem. Intl. Ed. Engl., 14. 822 (1975). D.'Dowd, M. Shapiro, J . Am. Chem. SOC., 98, 3724 (1976). A. I. Scott, K. Kang, J . Am. Chem. SOC., 99, 1997 (1977). A. I. Scott, J. B. Hansen, S. K. Chung, J . Chem. SOC.,Chem. Commun., 388 (1980). J. Halpem, Science, 227, 869 (1985). S. Wollowitz, J. Halpern, J . Am. Chem. SOC.,106, 8319 (1984). A. L. J. Beckwith, K. U. Ingold, in Rearrangements in the Ground State and Excited States, Vol. 1, P. de Mayo, ed., Academic Press, New York, 1980, p. 161. A. Effio, D. Griller, K. U. Ingold, A. L. J. Beckwith, A. K. Serelis, J . Am. Chem. SOC., 102, 1734 (1980). A. Bury, M. R. Ashcroft, M. D. Johnson, J . Am. Chem. SOC., 100, 3217 (1978). M. P. Atkins, B. T. Golding, A. Bury, M. D. Johnson, P. J. Sellars, J . Am. Chem. SOC., 102, 3630 (1980). M. P. Atkins, B. T. Golding, P. J. Sellars, J . Chem. SOC., Chem. Commun., 954 (1978). P. Dowd, S-C. Choi, F. Duah, C. Kaufman, Tetrahedron, 44, 2137 (1988). Y. Murakami, Y. Hisaeda, J. Kikuchi, T. Ohno, M. Suzuki, Y. Matsuda, Chem. Lett., 727 (1986). Y. Murakami, Y. Hisaeda, T. Ohno, Y. Matsuda, Chem. Lett., 731 (1986). Y. Murakami, Y. Hisaeda, T. Ohno, Chem. Lett., 1357 (1987). R. G . Finke, D. A. Schiraldi, J . Am. Chem. SOC., 105,7605 (1983). Y. Wang, R. G . Finke, Inorg. Chem., 28,983 (1989). H. Elroi, D. Meyerstein, J . Am. Chem. SOC., 100, 5540 (1978). W. A. Mulac, D. Meyerstein, J . Am. Chem. SOC., 104,4124 (1982). R. G . Finke, in Molecular Mechanisms in Bioorganic Processes, C. Bleasdale, B. T. Golding, eds., Royal Society of Chemistry, Cambridge, 1990, p. 244. B. T. Golding, L. Radom, J . Am. Chem. SOC., 98, 6331 (1976).

14.8.2.3. Bloalkylatlon Metals and metalloids undergo bioalkylation by microorganisms; however, other bioalkylations (i.e., bioethylation) have not been conclusively demonstrated. Biological methylation is a general process in living organisms', however, microorganisms (especially bacteria and fungi) play an important role in metal and metalloid transformations (including alkylations) in nature2. The bioconversion of inorganic metal and metalloid compounds to organometal and organometalloid forms is often accompanied by an increase in volatility and toxicity of the compounds. Alkylated forms of metals and metalloids are readily bioaccumulated in higher organisms owing to their lipophilicit? . The metals Hg, Sn, Pd, Pt, Au, and T, and the metalloids As, Se, Te, and S accept methyl carbanions from methylcobalamin (CH,B,,) in biological systems. The metals Pb, Cd, and Zn are not methylated owing to the extreme instability of their monoalkyl derivatives in aqueous systems, and CH3B,, does not transfer methyl groups to these elements4. Specific mechanisms of methylation will be discussed later.

(G.J. OLSON, F. E. BRINCKMAN) 1. J. S . Thayer, J . Chem. Educ., 50, 390 (1973). 2. J. Saxena, P. H. Howard, Adv. Appl. Microbiol., 21, 185 (1977); Chem. Abstr., 87,716j (1977). 3. J. S. Thayer, J . Organomet. Chem., 76, 265 (1974). 4. Y. T. Farchiang, W. P. Ridley, J. W. Wood, in Organometals and Organometalloids: Occurrence and Fate in the Environment, F. E . Brinckman, J. M. Bellama, eds., American Chemical Society Symposium Series, No. 82, Washington, DC, 1978, p. 65.


426

21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

14.8. Bioinor anic Catalysis 14.8.2. Cobafhin Reactions 14.8.2.3. Bioalkylation

nolds, D. Gani, J. A. Robinson,J. Chem. SOC.,Chem. Commun., 1334 (1986);G . Brendelberger, J. Retey, D. M. Ashworth, K. Reynolds, F. Willenbrock, J. A. Robinson, Angew. Chem. Intl. Ed. Engl., 27, 1089 (1988). G . Bidi'lngmaier,U. M. Kempe, T. Krebs, 3. RBtey, H. Flohr, Angew. Chem. Intl. Ed. Engl., 14. 822 (1975). D.'Dowd, M. Shapiro, J . Am. Chem. SOC., 98, 3724 (1976). A. I. Scott, K. Kang, J . Am. Chem. SOC., 99, 1997 (1977). A. I. Scott, J. B. Hansen, S. K. Chung, J . Chem. SOC.,Chem. Commun., 388 (1980). J. Halpem, Science, 227, 869 (1985). S. Wollowitz, J. Halpern, J . Am. Chem. SOC.,106, 8319 (1984). A. L. J. Beckwith, K. U. Ingold, in Rearrangements in the Ground State and Excited States, Vol. 1, P. de Mayo, ed., Academic Press, New York, 1980, p. 161. A. Effio, D. Griller, K. U. Ingold, A. L. J. Beckwith, A. K. Serelis, J . Am. Chem. SOC., 102, 1734 (1980). A. Bury, M. R. Ashcroft, M. D. Johnson, J . Am. Chem. SOC., 100, 3217 (1978). M. P. Atkins, B. T. Golding, A. Bury, M. D. Johnson, P. J. Sellars, J . Am. Chem. SOC., 102, 3630 (1980). M. P. Atkins, B. T. Golding, P. J. Sellars, J . Chem. SOC., Chem. Commun., 954 (1978). P. Dowd, S-C. Choi, F. Duah, C. Kaufman, Tetrahedron, 44, 2137 (1988). Y. Murakami, Y. Hisaeda, J. Kikuchi, T. Ohno, M. Suzuki, Y. Matsuda, Chem. Lett., 727 (1986). Y. Murakami, Y. Hisaeda, T. Ohno, Y. Matsuda, Chem. Lett., 731 (1986). Y. Murakami, Y. Hisaeda, T. Ohno, Chem. Lett., 1357 (1987). R. G . Finke, D. A. Schiraldi, J . Am. Chem. SOC., 105,7605 (1983). Y. Wang, R. G . Finke, Inorg. Chem., 28,983 (1989). H. Elroi, D. Meyerstein, J . Am. Chem. SOC., 100, 5540 (1978). W. A. Mulac, D. Meyerstein, J . Am. Chem. SOC., 104,4124 (1982). R. G . Finke, in Molecular Mechanisms in Bioorganic Processes, C. Bleasdale, B. T. Golding, eds., Royal Society of Chemistry, Cambridge, 1990, p. 244. B. T. Golding, L. Radom, J . Am. Chem. SOC., 98, 6331 (1976).

14.8.2.3. Bloalkylatlon Metals and metalloids undergo bioalkylation by microorganisms; however, other bioalkylations (i.e., bioethylation) have not been conclusively demonstrated. Biological methylation is a general process in living organisms', however, microorganisms (especially bacteria and fungi) play an important role in metal and metalloid transformations (including alkylations) in nature2. The bioconversion of inorganic metal and metalloid compounds to organometal and organometalloid forms is often accompanied by an increase in volatility and toxicity of the compounds. Alkylated forms of metals and metalloids are readily bioaccumulated in higher organisms owing to their lipophilicit? . The metals Hg, Sn, Pd, Pt, Au, and T, and the metalloids As, Se, Te, and S accept methyl carbanions from methylcobalamin (CH,B,,) in biological systems. The metals Pb, Cd, and Zn are not methylated owing to the extreme instability of their monoalkyl derivatives in aqueous systems, and CH3B,, does not transfer methyl groups to these elements4. Specific mechanisms of methylation will be discussed later.

(G.J. OLSON, F. E. BRINCKMAN) 1. J. S . Thayer, J . Chem. Educ., 50, 390 (1973). 2. J. Saxena, P. H. Howard, Adv. Appl. Microbiol., 21, 185 (1977); Chem. Abstr., 87,716j (1977). 3. J. S. Thayer, J . Organomet. Chem., 76, 265 (1974). 4. Y. T. Farchiang, W. P. Ridley, J. W. Wood, in Organometals and Organometalloids: Occurrence and Fate in the Environment, F. E . Brinckman, J. M. Bellama, eds., American Chemical Society Symposium Series, No. 82, Washington, DC, 1978, p. 65.


427

14.8.2. Cobalamin Reactions 14.8.2.3. Bioalkylation 14.8.2.3.1. Mercury. 14.8.2.3.1. Mercury.

Bioalkylation of Hg has received considerable attention owing to its suspected involvement in Minimata DiseaselV2,in which human poisonings in Japan and Sweden occurred after ingestion of fish and shellfish containing methylmercury. Methylmercury is a potent neurotoxin at low concentrations with a long biological half-life, and thus can be concentrated in food chains, even if present in small amounts in waters394. Inorganic Hg can be methylated by microorganisms in sediments4-’, creating an environmental hazard, but methylmercury can be degraded to Hg and CH, by sediment microorganisms”. Thus, the amount of methylmercury present in a sediment represents a balance between synthetic and degradative processes’ Microbial Hg-methylation also occurs in lake water8, rat intestines”, and soil^'^,'^. Microbial activity is probably responsible for methylmercury production observed in fish intestine^'^ and human feces16. Many species of bacteria methylate inorganic Hg in laboratory culture including Pseudomonas j l ~ o r e s c e n s ’ ~ Mycobacterium ~’~, phlei17, Escherichia c01i12*17,18, Aerobacter (Enterobacter) aer~genes’~,’’,Bacillus rnegateri~m‘~, and Clostridium cochlearium20. Species of Streptococcus‘2, Lactobacillus“, Bacteroides”, Bifidobacterium”, and Pseudomonas’ 7,21*22also methylate Hg. Extracts from a methanogenic bacterium synthesized methyl- and dimethylmercury from HgCl;,. Some capacity to methylate Hg may be a common property of aerobic bacterial7. Fungi, including Neurospora crassaZ4, Saccharomyces cerevi~iae’~, Scopulariopsis brevica~lis’~, and Aspergillus niger17, produce methylmercury from inorganic Hg. The biogenesis of ethylmercury is observedz5. Guppy, snail, elodea, and coontail take up phenylmercuric acetate and convert it primarily to Hg(I1) ions plus small amounts of ethyl- and methylmercury. Additional investigations into bioethylation of metals and metalloids are needed. In most pure culture and environmental studies of Hg methylation, HgCl, is used to demonstrate methylation. Additionally, however, HgI,, HgO, Hg(NO,),, Hg(CN),, Hg(SCN),, and Hg(CH,COO),, but not HgS are methylated by Clostridium cochlearium”. The sulfide must be converted to a soluble Hg2+ salt or HgO to be methylated by C. cochleariumz6. The low solubility of HgS suggests that in anaerobic sediments (i.e., H,S-containing environments) Hg might be effectively bound to S , preventing methylation. In aerobic sediments, however, methylmercury can be formed from HgS, although at a much lower rate than a soluble (HgC1,) mercury saltz7.Therefore, methylmercury could be formed in periodically aerobic sediments where some HgS oxidation may occur. When anaerobic conditions return to such a sediment, biogenic H,S may mobilize relatively nonvolatile methylmercury by reacting with methylmercury to form a volatile organosulfur derivative (CH,Hg),S, which decays to volatile dimethylmercury and HgS?

’.

2CH3Hgf

+ Sz-

- (CH,Hg),S

HgS &

+ (CH,),Hg t

(a)

Mercury methylation may also proceed abiologically in soils2’ or in aqueous environments3’.

(G.J. OLSON, F. E. BRINCKMAN) 1. K. Irukayama, in Advances in Water Pollution Research, Proc. 3rd Int. Conf., Vol. 3, Water Pollution Control Federation, Washington, DC,1967, p. 153. 2. T. I. Ishikawa, Y. Ikegaki, J. Water Pollut. Contr. Fed., 52, 1013 (1980); Chem. Abstr., 94, 70898p (1981). 3. Panel on Mercury, Coordinating Committee for Scientific and Technical Assessments of Environmental Pollutants, An Assessment of Mercury in the Environment, National Academy of Sciences, Washington, DC,1978; Chem. Abstr., 89, 64421h (1978).

428

14.8.2. Cobalamin Reactions 14.8.2.3. Bioalkylation 14.8.2.3.2. Arsenic.

4. A. Jemelbv, in Environmental Mercury Contamination, R. Hartung, B. D. Dittman, eds., Ann Arbor Science, Ann Arbor, MI, 1972, p. 174. 5. S. Jensen, A. Jemeltiv, Nature (London), 223, 753 (1969). 6. I. Berdicevsky, H. Shoyerman, S. Yannai, Environ. Res., 20, 325 (1979). 7. B. H. Olson, R. C. Cooper, Water Res., 10, 113 (1976). 8. A. Furutani, J. W. M. Rudd, Appl. Environ. Microbiol., 40, 770 (1980); Chem. Abstr., 94, 70891f (1981). 9. W. Blair, W. P. Iverson, F. E. Brinckman, Chemosphere, 3, 167 (1974); Chem. Abstr., 81, 101692e (1974). lO.,W. J. Spangler, J. L. Spigarelli, J. M. Rose, H. M. Miller, Science, 180, 192 (1973). 11. A. Kudo, H. Akagi, D. C. Mortimer, D. R. Miller, Nature (London),270,419 (1977). 12. I. Rowland, M. Davies, P. Grasso, Arch. Environ. Health, 32, 24 (1977); Chem. Abstr., 86, 115748t (1977). 13. W. F. Beckert, A. A. Moghissi, F. H. F. Au, E. W. Bretthauer, J. C. McFarlane, Nature (London), 249,674 (1974). 14. R. D. Rogers, J. Environ. Qual., 5,454 (1976). 15. J. W. M. Rudd, A. Furutani, M. A. Turner, Appl. Environ. Microbiol., 40, 777 (1980); Chem. Abstr., 93,216268m (1980). 16. T. Edwards, B. C. McBride, Nature (London), 253,462 (1975). 17. J. W. Vonk, K. Sijpesteijn,Ant. Van Leeuwenhoek, 39,505 (1973);Chem. Abstr., 79,123451m (1973). 18. D. G. Langley, J. Water Pollut. Contr. Fed., 45,44 (1973); Chem. Abstr., 78, 80477j (1973). 19. M. K. Hamdy, 0.R. Noyes, Appl. Microbiol., 30,424 (1975);Chem Abstr., 83,203622r(1975). 20. M. Yamada, K. Tonomura, J. Ferment. Technol., 50, 159 (1972); Chem. Abstr., 76, 124001q (1972). 21. C. W. Huey, Diss. Abstr., 37E, 2823 (1976). 22. S. Kitamura, Jpn. J. Hyg., 24, 132 (1969), cited in A. Jemelev, A. Martin, Ann. Rev. Microbiol., 29, 61 (1975). 23. J. M. Wood, F. S . Kennedy, C. G. Rosen, Nature (London), 220, 173 (1968). 24. L. Landner, Nature (London), 230,452 (1971). 25. S. C. Fang, Arch. Environ. Contam. Toxicol., 1 , 18 (1973). 26. M. Yamada, K. Tonomura, J. Ferment. Technol., 50, 901 (1972); Chem. Abstr., 78, 55219~ (1973). 27. T.Fagerstrbm, A. Jemelbv, Water Res., 5, 121 (1971). 28. P. J. Craig, P. D. Bartlett, Nature (London),275,635 (1978). 29. R. D. Rogers, J. Environ. Qual., 6,463 (1977); Chem. Abstr., 88, 21154k (1978). 30. H. Akagi, Y. Fujita, E. Takabatake, Chem. Lett., 171 (1975). 14.8.2.3.2. Arsenic.

As with methylmercury, organoarsenicals receive attention because of human poisoning. During the nineteenth century, domestic wallpapers containing As-based pigments poisoned humans. Many suggestions as to how toxic As products evolve from wallpaper were offered', but decades of speculation were ended when volatile As was determined to be produced by fungi growing on wallpaper which contained As-based pigments*. The toxic substance was later shown to be Me,As, produced by the fungus Scopulariopsis brevicaulis from A s 0 2 - , HS032- , methylarsonate, and dimethylarsinate'. Many species of fungi including Penicillium brevicaule, Aspergillus jischeri, A . sydowi (14 strains), and 10 strains or species of Scopulariopsis produce volatile As from AS,^,^, however, the species of volatile As is not identified. Microbial organoarsenical production is now demonstrated in a variety of microorganisms, e.g., Me2AsH is produced from As04, - by Methanobacterium strain M.o.H.~,Me,As forms from A s 0 2 - , HASO,'-, monomethylarsonate, and dimethylarsonate by C. humicola5. Arsenic methylation by Scopulariopsis brevicaulis, Candida humicola, and Gliocladium roseum also occurs6. Marine algae produce methylarsonate and dimethylar~inate~. A


428

14.8.2. Cobalamin Reactions 14.8.2.3. Bioalkylation 14.8.2.3.2. Arsenic.

4. A. Jemelbv, in Environmental Mercury Contamination, R. Hartung, B. D. Dittman, eds., Ann Arbor Science, Ann Arbor, MI, 1972, p. 174. 5. S. Jensen, A. Jemeltiv, Nature (London), 223, 753 (1969). 6. I. Berdicevsky, H. Shoyerman, S. Yannai, Environ. Res., 20, 325 (1979). 7. B. H. Olson, R. C. Cooper, Water Res., 10, 113 (1976). 8. A. Furutani, J. W. M. Rudd, Appl. Environ. Microbiol., 40, 770 (1980); Chem. Abstr., 94, 70891f (1981). 9. W. Blair, W. P. Iverson, F. E. Brinckman, Chemosphere, 3, 167 (1974); Chem. Abstr., 81, 101692e (1974). lO.,W. J. Spangler, J. L. Spigarelli, J. M. Rose, H. M. Miller, Science, 180, 192 (1973). 11. A. Kudo, H. Akagi, D. C. Mortimer, D. R. Miller, Nature (London),270,419 (1977). 12. I. Rowland, M. Davies, P. Grasso, Arch. Environ. Health, 32, 24 (1977); Chem. Abstr., 86, 115748t (1977). 13. W. F. Beckert, A. A. Moghissi, F. H. F. Au, E. W. Bretthauer, J. C. McFarlane, Nature (London), 249,674 (1974). 14. R. D. Rogers, J. Environ. Qual., 5,454 (1976). 15. J. W. M. Rudd, A. Furutani, M. A. Turner, Appl. Environ. Microbiol., 40, 777 (1980); Chem. Abstr., 93,216268m (1980). 16. T. Edwards, B. C. McBride, Nature (London), 253,462 (1975). 17. J. W. Vonk, K. Sijpesteijn,Ant. Van Leeuwenhoek, 39,505 (1973);Chem. Abstr., 79,123451m (1973). 18. D. G. Langley, J. Water Pollut. Contr. Fed., 45,44 (1973); Chem. Abstr., 78, 80477j (1973). 19. M. K. Hamdy, 0.R. Noyes, Appl. Microbiol., 30,424 (1975);Chem Abstr., 83,203622r(1975). 20. M. Yamada, K. Tonomura, J. Ferment. Technol., 50, 159 (1972); Chem. Abstr., 76, 124001q (1972). 21. C. W. Huey, Diss. Abstr., 37E, 2823 (1976). 22. S. Kitamura, Jpn. J. Hyg., 24, 132 (1969), cited in A. Jemelev, A. Martin, Ann. Rev. Microbiol., 29, 61 (1975). 23. J. M. Wood, F. S . Kennedy, C. G. Rosen, Nature (London), 220, 173 (1968). 24. L. Landner, Nature (London), 230,452 (1971). 25. S. C. Fang, Arch. Environ. Contam. Toxicol., 1 , 18 (1973). 26. M. Yamada, K. Tonomura, J. Ferment. Technol., 50, 901 (1972); Chem. Abstr., 78, 55219~ (1973). 27. T.Fagerstrbm, A. Jemelbv, Water Res., 5, 121 (1971). 28. P. J. Craig, P. D. Bartlett, Nature (London),275,635 (1978). 29. R. D. Rogers, J. Environ. Qual., 6,463 (1977); Chem. Abstr., 88, 21154k (1978). 30. H. Akagi, Y. Fujita, E. Takabatake, Chem. Lett., 171 (1975). 14.8.2.3.2. Arsenic.

As with methylmercury, organoarsenicals receive attention because of human poisoning. During the nineteenth century, domestic wallpapers containing As-based pigments poisoned humans. Many suggestions as to how toxic As products evolve from wallpaper were offered', but decades of speculation were ended when volatile As was determined to be produced by fungi growing on wallpaper which contained As-based pigments*. The toxic substance was later shown to be Me,As, produced by the fungus Scopulariopsis brevicaulis from A s 0 2 - , HS032- , methylarsonate, and dimethylarsinate'. Many species of fungi including Penicillium brevicaule, Aspergillus jischeri, A . sydowi (14 strains), and 10 strains or species of Scopulariopsis produce volatile As from AS,^,^, however, the species of volatile As is not identified. Microbial organoarsenical production is now demonstrated in a variety of microorganisms, e.g., Me2AsH is produced from As04, - by Methanobacterium strain M.o.H.~,Me,As forms from A s 0 2 - , HASO,'-, monomethylarsonate, and dimethylarsonate by C. humicola5. Arsenic methylation by Scopulariopsis brevicaulis, Candida humicola, and Gliocladium roseum also occurs6. Marine algae produce methylarsonate and dimethylar~inate~. A

14.8.2. Cobalamin Reactions 14.8.2.3. Bioalkylation 14.8.2.3.4. Selenium and Tellurium.

429

mixed microfloral community in pond sediments converts Me,As(O)H(OH) (cacodylic acid) to Me,As aerobically and anaerobically'. Reduction to methylarsines, not methylation to Me3As, may be the primary mechanism for gaseous As loss from soils since earlier investigations may have misidentified the arsenical gasg. (G. J. OLSON, F. E. BRINCKMAN)

F. Challenger, C. Higginbottom, L. Ellis, J . Chem. SOC., 95 (1933). B. Gosio, Ber. Deut. Chem. Gesell.,30, 1024 (1897). C. Thom, K. B. Raper, Science, 76, 548 (1932). B. C. McBride, R. S. Wolfe, Biochemistry, 10,4312 (1971). D. P. Cox, M. Alexander, Appl. Microbiol. 25,408 (1973); Chem. Abstr., 79, 27788f (1973). W. R. Cullen, C. L. Froese, A. Lui, B. C. McBride, D. J. Patmore, M. Reimer, J. Organomet. Chem., 139, 61 (1977). 7. M. 0. Andreae, D. Klumpp, Environ. Sci. Technol., 13, 738 (1979). 8. F. E. Brinckman, G. E. Pams, W. R. Blair, K. L. Jewett, W. P. Iverson, J. M. Bellama, Environ. Health Perspect., 19, 11 (1977); Chem. Abstr., 88, 1067f (1978). 9. C. N. Cheng, D. D. Focht, Appl. Environ. Microbiol., 38,494 (1979); Chem. Abstr., 92, 5418b (1980).

1. 2. 3. 4. 5. 6.

14.8.2.3.3. Lead.

Microbiological methylation of Pb(I1) salts and trimethyllead acetate gives Me,Pb'v2. Tetraalkyllead in rural air samples is attributed to production by extensive intertidal mud flats nearby3. However, biomethylation of Pb is not detected in bioreactors containing lead compounds and river sediment, sewage sludge, marine sediment, or in mixed cultures of methanogenic bacteria or Escherichia coli4, consequently, there is doubt that direct biological Pb(I1) methylation occurs in nature. Sulfide-induced chemical conversion of organic Pb(1V) salts to lead alkyls is possible. Methylcobalamin will not methylate Pb compounds5 and Me,Pb in anaerobic sediments is formed from trimethyllead acetate via formation of (Me,Pb),S, which decomposes to Me,Pb5. Thus, contribution of chemical vs biological Pb methylation in natural systems is unknown (but see 14.8.3.4). (G. J. OLSON, F. E. BRINCKMAN)

1. 2. 3. 4. 5.

P. T. S. Wong, Y. K. Chau, P. Luxon, Nature (London), 253,263 (1975). U. Schmidt, F. Huber, Nature (London), 259, 157 (1976). R. M. Hamson, D. P. H. Laxen, Nature (London),275,738 (1978). K. Reisinger, M. Stoeppler, H. W. Niirnberg, Nature (London), 291, 228 (1981). A. W. P. Jarvie, R. N. Markall, H. R. Potter, Nature (London), 255, 217 (1975).

14.8.2.3.4. Selenium and Tellurium.

Selenium and Te are biomethylated by Scopulariopsis brevicaulis'. Selenium biotransformations are important in the natural Se cycle. Penicillium isolated from sewage produces Me,Se from Se0;and Me,Te from TeCl,, H,Te03, and H,TeO,'. Me,Te is not formed if Se is not present, indicating that Se is required to induce some step in the biosynthesis of methyltellurium2. Soils with glucose and Na,Se03 formed Me,Se, suggesting methylation of Se by microorganisms is widespread3. In soils and sewage sludge selenite and Se are methylated, probably by microorganisms, to Me,Se, Me,Se,, and Me,Se04. Me,Se and Me,Se, and, in some cases, an unknown volatile selenium compound, probably Me,SeO;, are microbiologically produced from soils and sediments enriched with Na,Se03, Na,SO,, selenocysteine, selenourea, and seleno-~~-methionine~.



429



1. 2. 3. 4. 5. 6.

14.8.2.3.3. Lead.


1. 2. 3. 4. 5.






429



1. 2. 3. 4. 5. 6.

14.8.2.3.3. Lead.


1. 2. 3. 4. 5.




430

14.8.2. Cobalamin Reactions 14.8.2.3. Bioalkylation 14.8.2.3.6.Other Metals, Metalloids, and Nonmetals. ~

~

Species of Aeromonas, Flavobacrerium, and Pseudomonas grown in pure culture could methylate Na,SeO, to Me,Se and Me'Se,. Biosynthesis of volatile Me,Sez is a major metabolic pathway for detoxifying selenite in rat tissue6, microorganisms are not involved. (G. J. OLSON, F. E. BRINCKMAN)

1. F. Challenger, Chem. Rev., 36, 315 (1945). 2. R. W. Fleming, M. Alexander, Appl. Microbiol., 24, 424 (1972); Chem. Abstr., 77, 137210~ (1972). 3. A. J. Francis, J. M. Duxbury, M. Alexander, Appl. Microbiol., 28, 248 (1974); Chem. Abstr., 83,95526b (1975). 4. D. C. Reamer, W. H. Zoller, Science, 208, 500 (1980). 5. Y.K. Chau, P. T. S. Wong, B. A. Silverberg, P. L. Luxon, G. A. Bengert, Science, 192, 1130 (1976). 6. I. Rowland, M. Davies, P. Grasso, Arch. Environ. Healrh, 32, 24 (1977).

14.8.2.3.5.Tin.

Tetramethyltin and methylstannanes, Me,SnH4., ( n = 2,3) are formed by a species of Pseudomonas isolated from the Chesapeake Bay, and are detected in Chesapeake Bay waters'. Chesapeake Bay sediments contain microorganisms which produce trimethyland dimethyltin cations from tin(1V) chloride'. Me3SnOH is converted to Me4Sn in sediments from San Francisco Bay3, and lake waters and sediments are sites of volatile tin species formation4. (G. J. OLSON, F. E. BRINCKMAN)

1. J. A. Jackson, W. R. Blair, F. E. Brinckman, W. P. Iverson, Environ. Sci. Technol., 16, 110 (1982). 2. L. E. Hallas, J. C. Means, J. J. Cooney, Science, 215, 1505 (1982). 3. H. E. Guard, A. B. Cobet, W. M. Coleman, Science, 213,770 (1981). 4. Y.K. Chau, P. T. S. Wong, 0. Kramar, G. A. Bengert, Abstracts, 3rd Int. Conf. on the Organometallic and Coordination Chem. Germanium, Tin, and Lead, Dortmund, July, 1980, p. 30.

14.8.2.3.6.Other Metals, Metalloids, and Nonmetals.

Inorganic T1 is methylated in amended sediments in laboratory experiments, presumably by microorganisms'. No methylthallium compounds are found in natural waters. Although volatile, methylated S compounds are common in the en~ironment'-~, these are probably produced largely from decomposition of methionine or other organic sulfur compounds (reviewed in refs. 4,5). Few reports of methylation of inorganic S exist. The fungus Schizophyllum commune methylates simple inorganic S corn pound^^^^. This organism formed MeSH from glucose and SO4'- and also produced Me,S and MezSz from SO:-. The bacterium Pseudomonas aeruginosa also produced Me$, from SO4'-*. Eight species of marine algae produced the methylated sulfur compounds Me,S and Me'Se?. A preliminary report on microbial reaction with Cd indicated a strain of Pseudomonas produced trace amounts of a volatile Cd species from inorganic Cd(I1) in the presence of vitamin B,,, but provided no evidence for biomethylation of Cd". The prospect of Cd methylation has been considered by others".


430


~



14.8.2.3.5.Tin.






430


~



14.8.2.3.5.Tin.





B

14.8. Bioinor anic Catalysis 14.8.2. Coba amin Reactions 14.8.2.4. Biomethylation Mechanisms

43 1

Methylantimony compounds occur in natural waters under circumstances implying biogenesis ”.

(G.J. OLSON, F. E. BRINCKMAN) 1. F. Huber, in Organometals and Organometalloids: Occurrence and Fate in the Environment, F. E. Brinckman and J. M. Bellama, eds., American Chemical Society Symposium Series No. 82, Washington, DC, 1978, p. 65. 2. B. C. Nguyen, A. Gaudry, B. Bonsang, G. Lambert, Nature (London),275,637 (1978). 3. D. F. Adams, S. 0. Farwell, M. R. Pack, W. L. Bamesberger, J . Air Pollut. Contr. Assoc., 29, 380 (1979); Chem. Abstr., 91,43792f (1979). 4. J. M. Bremner, C. G. Steele, Adv. Microbial Ecol., 2, 155 (1978); Chem. Abstr., 90, 51124j (1979). 5 . S. H. Zinder, T. D. Brock, in Sulfur in the Environment, Part 11, Ecological Impacts, J. 0. Nriagu, ed., John Wiley, New York, 1978, p. 446. 6. J. H. Birkinshaw, W. P. K. Findley, R. A. Webb, Biochem. J., 36, 526 (1942). 7. F. Challenger, P. T. Charlton, J. Chem. SOC., 424 (1947). 8. R. A. Rasmussen, Tellus, 26, 254 (1974). 9. M. 0. Andreae, in Biogeochemistry of Ancient and Modern Environments, P. A. Trudinger, M. R. Walter, B. J. Ralph, eds., Canberra, Australia, 1980, p. 253. 10. C. W. Huey, F. E. Brinckman, W. P. Iverson, S. 0. Grim, in International Conference on Heavy Metals in the Environment, Abstracts, October 27-31, Toronto, Canada 1975. 11. J. W. Robinson, E. L. Kiesel, J. Environ. Sci. Health, A16, 341 (1981). 12. M. 0. Andreae, J. F. Asmode, P. Foster, L. Van’t dack, Anal. Chem., 53, 1766 (1981).

14.8.2.4. Biomethylation Mechanlsms

1. Mercury. In the biosynthesis of methylmercury by extracts from a methanogenic bacterium, electrophilic attack on the Co-C bond of methylcobalamin results in methyl transfer from Co3+ of methylcobalamin (CH3B12)to Hgz+ with the synthesis of methyl- and dimethylmercury’:

Transfer of the methyl carbanion from methylcobalamin to Hg2+ ion is nonenzymatic. Production of CH,B 12 is enzyme dependent. The initial methylation of Hg’+ proceeds at a rate 6000 times faster than the second methylation’. In addition to electrophilic attack involving the displacement of a carbanion, two other reactions result in Co-C bond breakage are proposed’. One involves displacement of a methyl radical from CH,B,,. The other is a redox switch mechanism which requires the presence of two different oxidation states of the metal.


B

14.8. Bioinor anic Catalysis 14.8.2. Coba amin Reactions 14.8.2.4. Biomethylation Mechanisms

43 1

Methylantimony compounds occur in natural waters under circumstances implying biogenesis ”.

(G.J. OLSON, F. E. BRINCKMAN) 1. F. Huber, in Organometals and Organometalloids: Occurrence and Fate in the Environment, F. E. Brinckman and J. M. Bellama, eds., American Chemical Society Symposium Series No. 82, Washington, DC, 1978, p. 65. 2. B. C. Nguyen, A. Gaudry, B. Bonsang, G. Lambert, Nature (London),275,637 (1978). 3. D. F. Adams, S. 0. Farwell, M. R. Pack, W. L. Bamesberger, J . Air Pollut. Contr. Assoc., 29, 380 (1979); Chem. Abstr., 91,43792f (1979). 4. J. M. Bremner, C. G. Steele, Adv. Microbial Ecol., 2, 155 (1978); Chem. Abstr., 90, 51124j (1979). 5 . S. H. Zinder, T. D. Brock, in Sulfur in the Environment, Part 11, Ecological Impacts, J. 0. Nriagu, ed., John Wiley, New York, 1978, p. 446. 6. J. H. Birkinshaw, W. P. K. Findley, R. A. Webb, Biochem. J., 36, 526 (1942). 7. F. Challenger, P. T. Charlton, J. Chem. SOC., 424 (1947). 8. R. A. Rasmussen, Tellus, 26, 254 (1974). 9. M. 0. Andreae, in Biogeochemistry of Ancient and Modern Environments, P. A. Trudinger, M. R. Walter, B. J. Ralph, eds., Canberra, Australia, 1980, p. 253. 10. C. W. Huey, F. E. Brinckman, W. P. Iverson, S. 0. Grim, in International Conference on Heavy Metals in the Environment, Abstracts, October 27-31, Toronto, Canada 1975. 11. J. W. Robinson, E. L. Kiesel, J. Environ. Sci. Health, A16, 341 (1981). 12. M. 0. Andreae, J. F. Asmode, P. Foster, L. Van’t dack, Anal. Chem., 53, 1766 (1981).

14.8.2.4. Biomethylation Mechanlsms

1. Mercury. In the biosynthesis of methylmercury by extracts from a methanogenic bacterium, electrophilic attack on the Co-C bond of methylcobalamin results in methyl transfer from Co3+ of methylcobalamin (CH3B12)to Hgz+ with the synthesis of methyl- and dimethylmercury’:

Transfer of the methyl carbanion from methylcobalamin to Hg2+ ion is nonenzymatic. Production of CH,B 12 is enzyme dependent. The initial methylation of Hg’+ proceeds at a rate 6000 times faster than the second methylation’. In addition to electrophilic attack involving the displacement of a carbanion, two other reactions result in Co-C bond breakage are proposed’. One involves displacement of a methyl radical from CH,B,,. The other is a redox switch mechanism which requires the presence of two different oxidation states of the metal.

432

14.8. Bioinor anic Catalysis 14.8.2. Cobasamin Reactions 14.8.2.4. Biomethylation Mechanisms

The fungus Neurospora crassa uses a different Hg methylation mechanism3. Neurospora methylates Hg by a process involving one or more steps of the methionine biosynthesis pathway. This methylation occurs by incorrect transfer of a methyl group to Hg, which had complexed with homocysteine in the cell. Normally, homocysteine (a methyl acceptor) would be methylated to methionine, however, a methylmercuryhomocysteine complex is formed. The methyl group is transferred to the Hg atom complexed to homocysteine, then methylmercury is cleaved: SH

Hg2+ +

S-Hg'

CH3Hg+

CH,

SH

+

I

I CH2 I

I

CHZ

CH,

H-C-NH,

H-C-NH,

I

I

I

I

CH, donor transmethylase

I

COOH homocysteine

COOH mercuryl-homocysteine

CH,

I

( b)

>

CH,

H-

C -NH,

I

I

COOH

The reaction requires a methyl donor such as choline or betaine and a transmethylase. Methylcobalamin is not involved in the metabolism of Neurospora.

2. Arsenic. Methyl carbonium ions from methionine, betaine, or choline are involved in certain As biomethylation~~: 0

0

II

:As -OH

+CH3

+

I H~C-AS-OH I

OH arsenious acid

methylarsonic acid

0 *CH3

II I

> H3C-As-OH

'CH,

CH3 cacodylic acid 0

It

(CH3)3A's (c> trimethylarsine oxide Recently other work supports this scheme in showing that methionine, S-adenosylmethionine, or a related methyl carbonium producing compound is the source of methyl groups in bioalkylation of As'. The species CH3B,, and CO,, but not S-CH,-tetrahydrofolate or serine, can serve as precursors for alkylarsine synthesis by Methanobacterium6.The formation of methane by Methanobacterium is inhibited by Se and Te which serve as methyl carbanion acceptors. 3. Recent developments. Not all biomethylations follow the classical carbanion-methylcobalamin scheme since indirect biomethylation by metabolic products occurs. Inorganic Hg is photochemically methylated in aqueous acetic acid solutions by sun-

14.8. Bioinorganic Catalysis 14.8.3. In Oxygen Transport

433

light or UV lamp irradiation'.'. MeI, produced by marine algae, especially in coastal watersg, methylates Pb(II)lo~l'and lead powder" possibly by an S,2 oxidative methylation mechanism. Methyl transfer from biogenic methyltin species to Hg occurs1', resulting in formation of methylmercury: Sn(1V)

Methyltin species

-

Methyltin species (biological)

+ Hgz+ +MeHg+

(abiotic)

(4

The extent of these potential methylation mechanisms in nature remains uncertain, as is the degree to which anthropogenic release of these potential methylating agents influences methylation in the environment. Controversy still exists over the occurrence of direct biomethylation of certain metals, such as PbI3 in nature.

(G.J. OLSON, F. E. BRINCKMAN)

1. 2. 3. 4. 5.

J. M. Wood, F. S . Kennedy, C. G . Rosen, Nature (London), 220, 173 (1968). J. M. Wood, Science, 183, 1048 (1974). L. Landner, Nature (London),230, 452 (1971). F. Challenger, Chem. Rev. 36, 315 (1945). W. R. Cullen, C. L. Froese, A. Lui, B. C. McBride, D. J. Patmore, M. Reimer, J . Organomet.

Chem., 139,61 (1977). B. C. McBride, R. S . Wolfe, Biochemistry, 10, 4312 (1971). H. Akagi, Y. Sakagomi, J . Hyg.Chem., 18,358 (1972). H. Akagi, Y. Fujita, E. Takabatake, J . Chem. Soc. Jpn., 1180 (1974). J. E. Lovelock, Nature (London), 256, 193 (1975). I. Ahmad, Y. K. Chau, P. T. S . Wong, A. J. Carty, L. Taylor, Nature (London), 287, 716 (1980). 11. A. W. P. Jarvie, A. P. Whitmore, Environ. Techol. Lett., 2, 197 (1981). 12. C. Huey, F. E. Brinckman, S . Grim, W. P. Iverson, in Proc. Int. Con$ Transport Persistent Chem. Aquatic Ecosys., A. S . W. DeFreitas, D. J. Kushner, S . U. Quadri, eds., National Research Council, Ottawa, Canada, 1974, p. 11-74, 13. K. Reisinger, M. Stoeppler, H. W. Nurnberg, Nature (London),291, 228 (1981).

6. 7. 8. 9. 10.

14.8.3. In Oxygen Transport Metal complexes are known that provide the delicate balance needed to form 0, adducts without the metal (M) or the ligands (L) being irreversibly oxidized. Such oxygen carriers are used in biological systems for the transport and storage of 0, by the reversible equilibria:

To be classified as an oxygen carrier, it is necessary that the reverse reaction, i.e., the dissociation of the 0, complex to give M(L) and 0,, be observable. This is achieved by lowering the partial pressure of 0,, by heating the complex, or by the addition of a ligand capable of replacing the bound 0,. The story of metal complexes able to bind 0, reversibly begins with the discovery in 1938 that certain 0,N-chelates (Schiff bases) of cobalt(I1) are oxygen carriers', to be joined in 1963 by Ir(PPh,),(CO)(Cl) which reversibly adds 0,'. Other low valent Pt


14.8. Bioinorganic Catalysis 14.8.3. In Oxygen Transport

433

light or UV lamp irradiation'.'. MeI, produced by marine algae, especially in coastal watersg, methylates Pb(II)lo~l'and lead powder" possibly by an S,2 oxidative methylation mechanism. Methyl transfer from biogenic methyltin species to Hg occurs1', resulting in formation of methylmercury: Sn(1V)

Methyltin species

-

Methyltin species (biological)

+ Hgz+ +MeHg+

(abiotic)

(4

The extent of these potential methylation mechanisms in nature remains uncertain, as is the degree to which anthropogenic release of these potential methylating agents influences methylation in the environment. Controversy still exists over the occurrence of direct biomethylation of certain metals, such as PbI3 in nature.

(G.J. OLSON, F. E. BRINCKMAN)

1. 2. 3. 4. 5.

J. M. Wood, F. S . Kennedy, C. G . Rosen, Nature (London), 220, 173 (1968). J. M. Wood, Science, 183, 1048 (1974). L. Landner, Nature (London),230, 452 (1971). F. Challenger, Chem. Rev. 36, 315 (1945). W. R. Cullen, C. L. Froese, A. Lui, B. C. McBride, D. J. Patmore, M. Reimer, J . Organomet.

Chem., 139,61 (1977). B. C. McBride, R. S . Wolfe, Biochemistry, 10, 4312 (1971). H. Akagi, Y. Sakagomi, J . Hyg.Chem., 18,358 (1972). H. Akagi, Y. Fujita, E. Takabatake, J . Chem. Soc. Jpn., 1180 (1974). J. E. Lovelock, Nature (London), 256, 193 (1975). I. Ahmad, Y. K. Chau, P. T. S . Wong, A. J. Carty, L. Taylor, Nature (London), 287, 716 (1980). 11. A. W. P. Jarvie, A. P. Whitmore, Environ. Techol. Lett., 2, 197 (1981). 12. C. Huey, F. E. Brinckman, S . Grim, W. P. Iverson, in Proc. Int. Con$ Transport Persistent Chem. Aquatic Ecosys., A. S . W. DeFreitas, D. J. Kushner, S . U. Quadri, eds., National Research Council, Ottawa, Canada, 1974, p. 11-74, 13. K. Reisinger, M. Stoeppler, H. W. Nurnberg, Nature (London),291, 228 (1981).

6. 7. 8. 9. 10.

14.8.3. In Oxygen Transport Metal complexes are known that provide the delicate balance needed to form 0, adducts without the metal (M) or the ligands (L) being irreversibly oxidized. Such oxygen carriers are used in biological systems for the transport and storage of 0, by the reversible equilibria:

To be classified as an oxygen carrier, it is necessary that the reverse reaction, i.e., the dissociation of the 0, complex to give M(L) and 0,, be observable. This is achieved by lowering the partial pressure of 0,, by heating the complex, or by the addition of a ligand capable of replacing the bound 0,. The story of metal complexes able to bind 0, reversibly begins with the discovery in 1938 that certain 0,N-chelates (Schiff bases) of cobalt(I1) are oxygen carriers', to be joined in 1963 by Ir(PPh,),(CO)(Cl) which reversibly adds 0,'. Other low valent Pt

14.8. Bioinorganic Catalysis 14.8.3. In Oxygen Transport 14.8.3.1. Nature of the Bound Dioxygen

434 ~~

~

metal complexes act as oxygen carriers. These systems, do not closely mimic the biological systems and are not included here. The cobalt(I1) chelates that reversibly add and release 0, like the hemoproteins are now joined by appropriate iron (11) and manganese(I1) porphyrins which behave as oxygen carrier^^-^. Examples of these and a copper(1) oxygen carrier are given in the sections that follow. (F. BASOLO)

T. Tsumaki, Bull. Chem. SOC.Jpn., 13,252 (1938). L. Vaska, Science, 140,809 (1963). J. S. Valentine, Chem. Rev., 73, 235 (1973). A. E. Martell, M. Calvin, Chemistry ofthe Metal Chelate Compounds. Prentice-Hall,Englewood Cliffs, NJ, 1952, p. 336. 5 . R. D. Jones, D. A. Summerville, F. Basolo, Chem. Rev., 79, 139 (1979).

1. 2. 3. 4.

14.8.3.1. Nature of the Bound Dioxygen

The formation of 0, metal complexes involves transfer of electron density from the metal complex to the 0, moiety, since complexes of Co(II), Fe(II), Mn(II), and Cu(I), which have higher oxidation states readily form 0, complexes, whereas complexes of Ni(II), Cu(II), and Zn(I1) with less stable higher oxidation states do not form 0, adducts. On this basis the 0, metal complex could be represented as [M(L)]*'[O,]". The structures of some 0, metal complexes are known, and the different types of complexes are classified as is shown in Table 1 (see 3.8.2.1.2). This classification is only a formalism, remembering that electron transfer from the metal complex to the bound 0, moiety is not complete. Thus superoxide-like does not mean superoxide ion, but the 0-0 stretching frequencies (vo'o-o, Table 1) and the

TABLE1. TYPESOF DIOXYGEN METALCOMPLEXES ~

Type Superoxide-like

~~

M/O, ratio 111

Structure

o.N 0

I

M Superoxide-like Peroxide-like

Ionic compounds

211

/0

\.

/M

0

Examplea

vo-o cm-'

Co(bzacen)(py)(O,) Fe(TpivPP)(1-MeIm)(O,)

1128 1159

[COLNHAO(OJI~+

1122 875

1145 842

Ligands: bzacen, N,"-ethylene bis(benzoy1acetoiminide); TpivPP, "pick fence parphyrin" diamen; 1 MeIm, 1 -methylimidazole; TPP, tetraphenylporphyrin. a


14.8. Bioinorganic Catalysis 14.8.3. In Oxygen Transport 14.8.3.1. Nature of the Bound Dioxygen

434 ~~

~

metal complexes act as oxygen carriers. These systems, do not closely mimic the biological systems and are not included here. The cobalt(I1) chelates that reversibly add and release 0, like the hemoproteins are now joined by appropriate iron (11) and manganese(I1) porphyrins which behave as oxygen carrier^^-^. Examples of these and a copper(1) oxygen carrier are given in the sections that follow. (F. BASOLO)

T. Tsumaki, Bull. Chem. SOC.Jpn., 13,252 (1938). L. Vaska, Science, 140,809 (1963). J. S. Valentine, Chem. Rev., 73, 235 (1973). A. E. Martell, M. Calvin, Chemistry ofthe Metal Chelate Compounds. Prentice-Hall,Englewood Cliffs, NJ, 1952, p. 336. 5 . R. D. Jones, D. A. Summerville, F. Basolo, Chem. Rev., 79, 139 (1979).

1. 2. 3. 4.

14.8.3.1. Nature of the Bound Dioxygen

The formation of 0, metal complexes involves transfer of electron density from the metal complex to the 0, moiety, since complexes of Co(II), Fe(II), Mn(II), and Cu(I), which have higher oxidation states readily form 0, complexes, whereas complexes of Ni(II), Cu(II), and Zn(I1) with less stable higher oxidation states do not form 0, adducts. On this basis the 0, metal complex could be represented as [M(L)]*'[O,]". The structures of some 0, metal complexes are known, and the different types of complexes are classified as is shown in Table 1 (see 3.8.2.1.2). This classification is only a formalism, remembering that electron transfer from the metal complex to the bound 0, moiety is not complete. Thus superoxide-like does not mean superoxide ion, but the 0-0 stretching frequencies (vo'o-o, Table 1) and the

TABLE1. TYPESOF DIOXYGEN METALCOMPLEXES ~

Type Superoxide-like

~~

M/O, ratio 111

Structure

o.N 0

I

M Superoxide-like Peroxide-like

Ionic compounds

211

/0

\.

/M

0

Examplea

vo-o cm-'

Co(bzacen)(py)(O,) Fe(TpivPP)(1-MeIm)(O,)

1128 1159

[COLNHAO(OJI~+

1122 875

1145 842

Ligands: bzacen, N,"-ethylene bis(benzoy1acetoiminide); TpivPP, "pick fence parphyrin" diamen; 1 MeIm, 1 -methylimidazole; TPP, tetraphenylporphyrin. a

14.8. Bioinorganic Catalysis 14.8.3. In Oxygen Transport 14.8.3.2. Natural Oxygen Carriers

435

0-0 bond distances of these complexes resemble more closely the values for superoxide ion (O,-) than for peroxide ion (O;-). The use of infrared is a convenient way to identify 0, metal complexes as superoxide-like (uo'o-o 1050-1 150cm- ') or peroxidelike 800-950 cm-'). (F. BASOLO)

14.8.3.2. Natural Oxygen Carriers

Biological oxygen carriers are of three main types: (1) the common heme-containing proteins such as hemoglobin (Hb) and myoglobin (Mb), and the nonheme-containing proteins; (2) the hemerythrins; and (3) the hemocyanins. The hemocyanins are Cu complexes', whereas the others are Fe complexes. There is also a V-containing protein, hemovanadin, capable of reversibly binding 0,. This is found in the blood cells of certain ascidians, it does not contain porphyrin but little is known about the nature of the V complex. Dioxygen combines reversibly with Hb and Mb in the blood and tissues by virtue of a heme [iron(II) porphyrin] prosthetic group. The proteins bind one 0, for each Fe(I1). For this class of respiratory pigments, the Fe(I1) is coordinated to the four core N atoms of the protoporphyrin (1).

CH;

-CH=

CH3-

.CH3 CH, CH,

\

COOH

\

/CH,

COOH

Protoporphyrin 1 The Fe(I1) porphyrin complex is embedded in a polypeptide chain of MW ca 17,000. A proximal-imidazole N coordinates to Fe in one of the axial positions, but a distal-imidazole is prevented from attaching itself to the opposite axial position. This means that Fe(I1) is five-coordinated, and that it can add 0, in the hydrophobic pocket provided by the globin chain (Fig. 1). Myoglobins are monomeric, whereas Hbs are tetrameric. The four hemes in Hb add oxygen in a cooperative fashion. This means that 0, adds to the first Fe(I1) less readiy than it does to the second Fe(I1) and most readily to the fourth Fe(I1). A trigger mech-

436

14.8. Bioinorganic Catalysis 14.8.3. In Oxygen Transport 14.8.3.3. Cobalt(l1) Complexes

Figure 1. Hemoprotein showing the large globin polypeptide chain, with one imidazole group (proximal) attached to Fe in heme and opposite it the unattached (distal) imidazole in the hydrophobic pocket. anism is suggested for this cooperativitp, but details of the process are not yet understood. (F. BASOLO)

1. W. P. J. Gaykema, W. G. J. Hol, J. M. Vereijken, N. M. Soeter, H. J. Bak, J. J. Beintema,Nature

(London),309, 23 (1984). 2. M. F. Perutz, Nature (London),228, 726 (1970); 237, 495 (1972).

14.8.3.3. Cobalt(l1) Complexes

Most cobalt(I1) complexes, under certain conditions, behave as 0, carriers (see 3.8.2.1.2).The solid phases isolated from reaction mixtures of cobalt(I1) complexes and 0, are dimeric bridged complexes of the type [(L)Co"'-O,2--C o1"(L)I4+ .The equilibria involved are

Co(L) Co(L)(O,)

+ 0, e Co(L)(O,)

+ Co(L) e (L)co-o,-co(L)

(a) (b)

Since the monomeric species in equation (a) relates more closely to Hb(0,) and Mb(O,), the complexes of the type Co(L)(O,) are isolated and Characterized'. The O2 Co complexes are examined by Ir, EPR, and X-ray, which establishes structures of the end-on bent-type, Co-0-0, (see the first entry in Table 1, 14.8.5.1). In many cases the base used is prepared by condensing two acetylacetonates with one ethylenediamine, to give the Co chelate structure:


436


Figure 1. Hemoprotein showing the large globin polypeptide chain, with one imidazole group (proximal) attached to Fe in heme and opposite it the unattached (distal) imidazole in the hydrophobic pocket. anism is suggested for this cooperativitp, but details of the process are not yet understood. (F. BASOLO)

1. W. P. J. Gaykema, W. G. J. Hol, J. M. Vereijken, N. M. Soeter, H. J. Bak, J. J. Beintema,Nature

(London),309, 23 (1984). 2. M. F. Perutz, Nature (London),228, 726 (1970); 237, 495 (1972).

14.8.3.3. Cobalt(l1) Complexes

Most cobalt(I1) complexes, under certain conditions, behave as 0, carriers (see 3.8.2.1.2).The solid phases isolated from reaction mixtures of cobalt(I1) complexes and 0, are dimeric bridged complexes of the type [(L)Co"'-O,2--C o1"(L)I4+ .The equilibria involved are

Co(L) Co(L)(O,)

+ 0, e Co(L)(O,)

+ Co(L) e (L)co-o,-co(L)

(a) (b)

Since the monomeric species in equation (a) relates more closely to Hb(0,) and Mb(O,), the complexes of the type Co(L)(O,) are isolated and Characterized'. The O2 Co complexes are examined by Ir, EPR, and X-ray, which establishes structures of the end-on bent-type, Co-0-0, (see the first entry in Table 1, 14.8.5.1). In many cases the base used is prepared by condensing two acetylacetonates with one ethylenediamine, to give the Co chelate structure:


437

Co(acacen)

A monodentate ligand coordinates in the fifth coordination, axial position and 0, adds to the remaining sixth position. To determine the effect of the axial ligand on the 0, affinity of the complex, the equilibrium constants, Ko2, are measured for: Co(SB)(L)

+ 0, 5 Co(SB)(L)(O,)

(c)

The data show a linear correlation of 0, affinity with the ease of oxidation of the cobalt(I1) complex (see Fig. 1). This result is in accord with electron transfer from the complex to the 0, moiety being involved in the formation of the complex. The successful preparation of monomeric Co-0, complexes suggests replacing the Fe(I1) with Co(I1) in Hb and Mb to give cobalt models of the hemoproteins. These models, known as coboglobins, behave similarly to the natural proteins, but the paramagnetic 0x0-coboglobins can be probed by EPR, whereas the corresponding diamagnetic Hb(0,) and Mb(0,) are EPR silent. Valuable information on the bonding and structure of these 0, complexes by studying their EPR ~ p e c t r a ~ , ~ . (F. BASOLO)

1. C. Floriani, F. Calderazzo, J . Chem. Soc., A, 946 (1969); A. L. Crumbliss, F. Basolo, Science, 164, 1168 (1969); J . Am. Chem. Soc., 92,55 (1970). 2. B. M. Hoffman, D. H. Petering, Proc. Natl. Acad. Sci. U.S.A., 67, 637 (1970). 3. M. J. Carter, D. P. Rillema, F. Basolo, J . Am. Chem. Soc., 96, 392 (1974).

I

z

0

I

-l-

31

-0.30

-0.35

I -0.40

I

-0.45

-0.50

I

-0.55

E,, volts

I -0.60

I

-0.65

I

I

-0.70

Figure 1. Comparison of oxygen uptake (log Ko2) at - 21°C for Co(benacen)B to the polarographic half-wave potentials (E,,,) for Co"-"'(benacen)B,; 1, PPh,; 2, CNpy; 3, py; 4,3,4-lutidine;5, piperidine; 6, sec-BuNH,; 7, 5-C1-N-Melm; 8, N-Melm; 9, i-BuNH,; 10, n-BuNH,. benacen, benzoylacetylacetonateion; PPh,, triphenylphosphine; py, pyridine; Bu, butyl (from ref. 3).


438

14.8. Bioinorganic Catalysis 14.8.3. In Oxygen Transport 14.8.3.4. Iron(l1) Complexes

14.8.3.4. Iron(l1) Complexes

Coordination chemists were unable to have synthetic Fe(I1) complexes carry 0,, although this is the method of choice in nature. In most cases the reaction of 0, with Fe(I1) complexes results in the irreversible formation of the stable p-0x0 dimer, Fe(II1)-0-Fe(II1): Fe(II)(L)

+

Fe(0,)

Fe(lL)(L)

(L)Fe(III)-0-Fe(III)(L)

(a)

Nature prevents this by isolating individual hemes in the globin polypeptide chain (see Fig. 1, 14.8.5.2), which means two Fe complexes cannot come near enough to form a p o x 0 bridge. In the laboratory three approaches are now successful in preventing formation of the p-0x0 dimer: (1) steric-such that dimerization is inhibited; (2) low temperature-in order that reactions leading to dimerization are negligibly slow; and (3) rigid surfaces-attachment of the Fe complex on a surface in a manner to prevent dimerization. The steric approach to the preparation of Fe complexes which carry 0, is the most elegant, and it is the approach most studied. The easiest porphyrin to prepare is tetraphenylporphyrin (1):

Q 0

(1) Tetraphenylporphyrin, TPP This porphyrin is not present in biological systems, but it is used extensively as a model because of its ease of synthesis. Derivatives of this porphyrin led to the discovery of O2 carriers of Fe(I1) complexes. Two of the most studied examples are the picket-fence' (Fig. 1) and the capped' (Fig. 2) porphyrins. Both these porphyrins are ortho-substituted TPP. The picket-fence porphyrin has t-butylamido group group sticking up on one side of the porphyrin plane, creating a picket-fence around the Fe. The steric bulk of these groups renders addition of large ligands (e.g., pyridine or 1-methyl imidazole) to this enclosed coordination site difficult, but small molecules such as 0,, CO, and NO readiy add. The capped porphyrin has one side of the porphyrin enclosed in acap of sufficient size to permit small molecules to enter and bind with Fe, but larger monodentate ligands cannot enter.


439

Figure 1. Structure of Fe(TpivPP)(L)(O,), the oxygen adduct of the “picket-fence’’ porphyrin. Double bonds are deleted for clarity.

When the open axial-coordination site of Fe is protected with a ligand, then these models function as oxygen carriers in a manner similar to Mb. A schematic representation is shown (Fig. 3). Note that unless the outside axial-site is protected, irreversible oxidation takes place. At temperatures below - 50°C in aprotic solvents, even Fe(TPP)(L), reversibly adds

0,:

Fe(TPP)(L),

+ 0, +Fe(TPP)(O)(O,) + L

(a)

Another way that Fe(TPP) is made to add 0, reversibly, even at RT is to isolate the complex by attaching it to the rigid surface of an imidazole-modified silica gel3 (Fig. 4). (F. BASOLO)

1. J. P. Collman, R. R. Gagne, T. R. Halbert, J. C. Marchon, C. A. Reed, J. Am. Chem. SOC., 95, 7863 (1973). 2. J. Almog, J: E. Baldwin, J. Huff,J. Am. Chem. Soc., 97, 227 (1975). 3. 0. Leal, D. L. Anderson, R. G . Bowman, F. Basolo, R. L. Burwell, Jr., J . Am. Chem. Soc., 97, 5125 (1975).


440

Figure 2. Structure of the ferrous “capped” porphyrin. Double bonds are deleted for clarity.

L(excess) ___,

a I L

I 0 I

(-7) Fe

I

L

Reversible oxygenation

Irreversible oxidation

e 7

Figure 3. Reversible and irreversible reaction of O2 with Fe(I1) capped porphyrin.

(IPG)Fe(TPP)

L

(IPG)Fe(TPP)(0,)

Figure 4. The oxygen carrying properties of an Fe(I1) complex attached to a rigid surface.


14.8. Bioinorganic Catalysis 14.8.3. In Oxygen Transport 14.8.3.5. Manganese(l1) and Copper(1) Complexes

441

14.8.3.5. Manganese(l1) and Copper(1) Complexes

Manganese(I1) complexes are irreversibly oxidized by 0,, and in some cases p-0x0 bridged complexes of the type Mn(II1)-0-Mn(II1) form, similar to that of Fe(I1) complexes. Using an approach for making oxygen carriers of Fe(I1) complexes, experiments at low temperature yield an oxygen carrier of a Mn(I1) porphyrin': Mn(TPP)(L)

+ 0,

Mn(TPP)(O,)

+L

(a)

Unlike the oxygen complexes of Co and Fe, which are end-on bent superoxide-like, the Mn complex has a side-on peroxide-like structure (see Table 1, 14.8.5.1). The natural oxygen carriers hemocyanins are Cu proteins, and much effort is spent in attempts to model these systems. Often Cu(I1) complexes do not react with O,, and Cu(1) complexes are irreversibly oxidized; however, some exceptions are known. One Cu(1) complex that carries oxygen is shown in Fig. 1. The delicate nature of the proper metal-ligand combination to give an 0, camer is well illustrated by this system. If the terminal ligand nitrogens in this quinquidentate ligand are imidazole group N, then 0, adds reversibly, whereas if these are nitrogens of two pyridine groups then 0, adds irreversibly. The oxygen-carrying properties involve the equilibrium':

Reversible 0, binding by a series of Cu(1) complexes, [Cu,(NnPY,)]'+, PY, = bis[2(2-pyridyl)ethylamine], has been described3. Very recently Cu(I1) intermediate complex participation has been demonstrated spectrally in tris(pyridy1)amine-Cu(1) complex reversible of 0;. (F. BASOLO)

1. C. J. Weschler, B. M. Hoffman, F. Basolo, J . Am. Chem. SOC., 97, 5279 (1975). 2. M. Simmons, C. L. Menill, L. J. Wilson, L. A. Bottomley, K. M. Kadish. J . Chem. SOC.Dalton Trans., 1827 (1980).

Figure 1. A Cu(1) complex that carries 0,.

442

14.8. Bioinorganic Catalysis 14.8.4. In Oxidases 14.8.4.1. In Cytochrome Oxidases

3. K. D. Karlin, M. S. Haka, R. W. Cruse, Y. Gultnek, J. Am. Chem. Soc., 107, 5828 (1985). 4. K. D. Karlin, N. Wei, B. Jung, S. Kaderli, A. D. Zuberbiihler, J . Am. Chem. Soc., 113, 5868 (1991).

14.8.4. In Oxidases 14.8.4.1. In Cytochrome Oxidases

The terminal oxidase in an energy-transducing, cytochrome-based electron-transport system maintains electron flow by coupling cytochrome oxidation to dioxygen (0,)reduction. Members of this protein class are referred to as cytochrome oxidases; they carry out 0,-binding and redox chemistry at transition metal-containing active sites. Although iron is the most commonly used metal and may occur as a protoheme or iron-chlorin species in the protein, this section is concerned only with mitochondria1 cytochrome oxidase, which contains 2 mol of Cn and 2 mol of heme a bound Fe per function unit. Biochemistry of the protein' will not be considered here, instead the focus will be on the structure of the metal centers, on the reactions they catalyze and on models for these centers. In the most widely accepted model for cytochrome oxidase, the metal components of the enzyme function in pairs. One of the iron atoms, designated cytochrome a, and one of the copper atoms, Cu,, are associated with cytochrome c oxidation; the second iron atom, denoted cytochrome u3, and its associated copper, Cua3,form the 0, reducing site2. Consistent with differentiation of the two hemes a in terms of function are the ligand-binding properties of a and a3: cytochrome a3 is susceptible to ligation by exogeneous ligands (e.g., [CNI-, CO, NO, N3-, S2->; cytochrome a is substitutionally inert. This behavior also extends to the two copper atoms; Cu,, will bind NO under certain conditions3 whereas exogeneous ligand binding to Cu, does not occur. Mitochondrial cytochrome oxidase is a multisubunit protein, and progress toward a functional separation of the metal components in their polypeptide environments progresses4. The Fe(II1) and Cu(I1) atoms of cytochrome oxidase are paramagnetic; moreover, the heme u chromophore (1) of cytochromes a and a35has a strong absorbance in the visible region. Purified preparations of the enzyme with excellent optical, magnetic, and solubility properties, have allowed sophisticated physical studies. As a consequence, a good understanding of the immediate environments of the metal ions, particularly of the iron atoms, has emerged. Cytochrome a is magnetically isolated in the purified enzyme and is low-spin and six-coordinated in both its Fe(I1) and Fe(II1) valence states. The ligand field parameters determined for Fe(II1) cytochrome a by EPR and the gross features of its MCD spectrum, including the unusual reversed A term in the Soret region, are reproduced well by simple, six-coordinated heme a model compounds in which imidazole ligands occupy the iron axial ligation sites6,'. Cu,, probably the most enigmatic of the oxidase metal centers, is considerably less well characterized. Like cytochrome a, it is magnetically isolated and contributes a strong EPR signal to the magnetic resonance spectrum of the oxidized enzyme. The spectral, characteristics of this signal are unusual: whereas the hyperfine splitting owing to the I = 3/2 Cu nucleus in simple, inorganic copper complexes is typically 0.015 to 0.020 cm-', and that in Type I blue copper proteins lie between 0.0035 and 0,009 cm-', for


442


3. K. D. Karlin, M. S. Haka, R. W. Cruse, Y. Gultnek, J. Am. Chem. Soc., 107, 5828 (1985). 4. K. D. Karlin, N. Wei, B. Jung, S. Kaderli, A. D. Zuberbiihler, J . Am. Chem. Soc., 113, 5868 (1991).

14.8.4. In Oxidases 14.8.4.1. In Cytochrome Oxidases

The terminal oxidase in an energy-transducing, cytochrome-based electron-transport system maintains electron flow by coupling cytochrome oxidation to dioxygen (0,)reduction. Members of this protein class are referred to as cytochrome oxidases; they carry out 0,-binding and redox chemistry at transition metal-containing active sites. Although iron is the most commonly used metal and may occur as a protoheme or iron-chlorin species in the protein, this section is concerned only with mitochondria1 cytochrome oxidase, which contains 2 mol of Cn and 2 mol of heme a bound Fe per function unit. Biochemistry of the protein' will not be considered here, instead the focus will be on the structure of the metal centers, on the reactions they catalyze and on models for these centers. In the most widely accepted model for cytochrome oxidase, the metal components of the enzyme function in pairs. One of the iron atoms, designated cytochrome a, and one of the copper atoms, Cu,, are associated with cytochrome c oxidation; the second iron atom, denoted cytochrome u3, and its associated copper, Cua3,form the 0, reducing site2. Consistent with differentiation of the two hemes a in terms of function are the ligand-binding properties of a and a3: cytochrome a3 is susceptible to ligation by exogeneous ligands (e.g., [CNI-, CO, NO, N3-, S2->; cytochrome a is substitutionally inert. This behavior also extends to the two copper atoms; Cu,, will bind NO under certain conditions3 whereas exogeneous ligand binding to Cu, does not occur. Mitochondrial cytochrome oxidase is a multisubunit protein, and progress toward a functional separation of the metal components in their polypeptide environments progresses4. The Fe(II1) and Cu(I1) atoms of cytochrome oxidase are paramagnetic; moreover, the heme u chromophore (1) of cytochromes a and a35has a strong absorbance in the visible region. Purified preparations of the enzyme with excellent optical, magnetic, and solubility properties, have allowed sophisticated physical studies. As a consequence, a good understanding of the immediate environments of the metal ions, particularly of the iron atoms, has emerged. Cytochrome a is magnetically isolated in the purified enzyme and is low-spin and six-coordinated in both its Fe(I1) and Fe(II1) valence states. The ligand field parameters determined for Fe(II1) cytochrome a by EPR and the gross features of its MCD spectrum, including the unusual reversed A term in the Soret region, are reproduced well by simple, six-coordinated heme a model compounds in which imidazole ligands occupy the iron axial ligation sites6,'. Cu,, probably the most enigmatic of the oxidase metal centers, is considerably less well characterized. Like cytochrome a, it is magnetically isolated and contributes a strong EPR signal to the magnetic resonance spectrum of the oxidized enzyme. The spectral, characteristics of this signal are unusual: whereas the hyperfine splitting owing to the I = 3/2 Cu nucleus in simple, inorganic copper complexes is typically 0.015 to 0.020 cm-', and that in Type I blue copper proteins lie between 0.0035 and 0,009 cm-', for


443

'.

Cu, the copper hyperfine is estimated to be less than 0.003 cm- ' Thus, the paramagnetic species responsible for the Cu, EPR signal may be better represented by a sulfur radical/Cu(I) ion formulation, RsaCu' -t '. Magnetic resonance"-" and EXAFS13results indicate that Cu, has sulfur and nitrogen ligands in its first coordination sphere, but the geometry and the electron distribution remain unclear.

HO-CH

CH3

I

I

I,\ I X

J

l

CH,

CH,

COOH

I COOH

\

Y

heme a 1 As opposed to cytochrome a and Cu,, neither cytochrome a3 nor Cu,, is EPR detectable in the oxidized enzyme. Thus the two metal centers are magnetically coupled so as to result in an EPR-silent binuclear center14; MCD and magnetic susceptibility measurements are corroboratory and indicate that high-spin Fe(II1) in cytochrome a3 is strongly antiferromagnetically coupled (12JI > 200 cm- ') to Cu(I1) Cu,, so that an S = 2 ground state result^'^*'^. If the iron of cytochrome a3 is forced low spin by cyanide binding, the exchange coupling between a3 iron and Cu=a3 survives although the magnitude of 12JI decreases to -40 cm-lI6 or less17. These observations, coupled with EPR data that indicate that one of the axial ligands to the iron of a3 is histidine'*.'', led to formulation of the active site in terms of an imidazolate bridge between Cu=a3 and the iron of cytochrome a3, the exchange coupling being mediated by the r-system of the intervening ligand". However, subsequent model compound work (see 14.8.6.1.1) and data on the ligand binding properties of Cu=a3and Fe,3,3~21~22 suggest that the two metals are arranged so that 0, acts as a transient bridge in the reduction process. An alternative ligand, B, serves as the bridging in the oxidized enzyme.23 A detailed model for the resting enzyme in which the bridging species is a sulfur atom derived from a polypeptide amino acid is shown in Fig. la13; oxygen intermediates as sequential bridging species during dioxygen reduction are shown in Fig. lb3. Taken together, these two proposals provide a reasonable hypothesis for events in the dioxygen binding and reducing site and provide testable models for futher physical and synthetic research.


444

BrqoH2 0 = T4-N< I heme plane

Rl

R2

A

HO--

I Fef3-NANH

I

---\

heme plane

“g5” CONFORMATION

‘OXYGENATED’CONFORMATION

B@

heme plane

OH2

R2

OH2

H’dt,

O-*

= 1 hr. (pH 7.4)

-Fr3 I

-

A

“g 12” CONFORMATION

heme plane ( b)

Figure 1. (a) Proposed model for the structure of the cytochrome a,-Cu,a, site in resting cytochrome oxidase (after ref. 13, with permission); (b) proposed models for the a,-Cu,a, site during turnover conditions (after ref. 3, with permission). (G.T. BABCOCK)

14.8.4. In Oxidases 14.8.4.1. In Cytochrome Oxidases 14.8.4.1. I . Reactions

445

1. R. A. Capaldi, in Membrane Proteins in Energy Transduction R. A. Capaldi, ed., Dekker, New York, 1979, p. 201. 2. B. G. Malstrom, Biochim. Biophys. Acta, 549, 281 (1979). 3. G. W. Brudvig, T. H. Stevens, R. H. Morse, S . I. Chan, Biochemistry, 20 3912 (1981). 4. D. B. Winter, W. J. Bruyninckx, F. G. Foulke, N. P. Grinich, H. S . Mason, J . Biol. Chem., 255, 11408 (1980). 5. W. S. Caughey, G. A. Smythe, D. H. O’Keefe, J. Maskasky, M. L. Smith,J . Biol. Chem., 250, 7602 (1975). 6. G. T. Babcock, J. Van Steelandt, G . Palmer, L. E. Vickery, I. Salmeen, Dev. Biochem., 5, 105 (1979). 7. T. Nozawa, Y. Orii, A. Kaito, T. Yamamoto, M. Hatano, in Developments in Biochemistry,

Vol.5 Cytochrome Oxidase, T. E. King, Y. Orii; B. Chance, K. Okunuki, eds., Elsevier, Amsterdam, 1979, p. 117. 8. T. VlnngLd, in Biological Applications of Electron Spin Resonance, H. M. Swartz, J. R. Bolton, D. C. Borg, eds., Wiley-Interscience, New York, 1972, p. 411. 9. J. Peisach, W. E. Blumberg, Arch. Biochem. Biophys., 165,691 (1974). 10. W. B. Mims, J. Peisach, R. W. Shaw, H. Beinert, J . Biol. Chem., 255, 6843 (1980). 11. B. M. Hoffman, J. E. Roberts, M. Swanson, S . W. Speck, E. Margoliash, Proc. Natl. Acad. Sci USA, 77, 1452 (1980). 12. W. Froncisz, C. P. Scholes, J. S. Hyde, H.-H. Wei, T. E. King, R. W. Shaw, H. Beinert, J . Biol. Chem., 254,7482 (1979). 13. L. Powers, B. Chance, Y. Ching, P. Angiolillo, Biophys. J., 34,465 (1981). 14. B. F. Van Gelder, H. Beinert, Biochim. Biophys. Acta, 189, 1 (1969). 15. G. T. Babcock, L. E. Vickery, G . Palmer,J. Biol. Chem., 251,7907 (1976). 16. M. F. Tweedle, L. J. Wilson, L. Garcia-Iniguez, G. T. Babcock, G. Palmer, J . Biol. Chem., 253, 8065 (1978).

A. J. Thomson, M. K. Johnson, C. Greenwood, P. E. Gooding, Biochem. J., 193,687 (1981). 18. M. F. J. Blokzijl-Homan, B. F. Van Gelder, Biochim. Biophys. Acta, 234,493 (1971). 19. T. H. Stevens, S . I. Chan, J . Biol. Chem., 256, 1069 (1981). 20. G. Palmer, G. T. Babcock, L. E. Vickery, Proc. Natl. Acad. Sci. USA, 73,2206 (1976). 21. W. E. Blumberg, J. Peisach, in Developments in Biochemistry, Vol. 5 , Cytochrome Oxidase, T. E. King, Y. Orii, B. Chance, K. Okunuki, eds., Elsevier, Amsterdam, 1970, p. 153. 22. J. 0. Alben, P. 0. Moh, F. G . Fiamingo, R. A. Altschuld, Proc. Natl. Acad. Sci USA, 78, 234 (1981). 23. G. T. Babcock, P. M. Callahan, M. R. Ondrias, I. Salmeen, Biochemistry, 20,959 (1981). 14.8.4.1.1. Reactions.

Oxidation of cytochrome c involves cytochrome a as the primary reaction site in a rapid biomolecular process ( k lo7 M-’s - l ) . Initial electron transfer is followed by electron redistribution within the oxidase and further oxidation of cytochrome c. Several enigmatic features process remain, e.g., cytochrome oxidase may function as a dime? with each monomer unit in the dimer having two cytochrome c binding sites.3 Despite these two sites, initial oxidation of c by oxidase results in reduction of substoichiometric amounts of a4, and indicates heretogeneity among the cytochrome a centers. Similarly, hetereogeneous behavior is observed in the redox chemistry of cytochrome a5, These observations suggest a model in which the redox and kinetic properties of the two cytochromes a in the dimer structure are differentiated, although unambiguous tests remain to be done. A second aspect of the cytochrome clcytochrome a reaction relates to the electron transfer distance. When the cytochrome c binding sites are occupied by the Fe(II1) species, the EPR parameters of neither a3+ nor c3+ are perturbed; this indicates that the distance between the paramagnets is greater than lOA. Both fluorescence6 and magnetic relaxation’ measurements support this conclusion. This observation makes difficult a simple outer-sphere electron transfer explanation for the high rates observed in the cy-

-


14.8.4. In Oxidases 14.8.4.1. In Cytochrome Oxidases 14.8.4.1. I . Reactions

445

1. R. A. Capaldi, in Membrane Proteins in Energy Transduction R. A. Capaldi, ed., Dekker, New York, 1979, p. 201. 2. B. G. Malstrom, Biochim. Biophys. Acta, 549, 281 (1979). 3. G. W. Brudvig, T. H. Stevens, R. H. Morse, S . I. Chan, Biochemistry, 20 3912 (1981). 4. D. B. Winter, W. J. Bruyninckx, F. G. Foulke, N. P. Grinich, H. S . Mason, J . Biol. Chem., 255, 11408 (1980). 5. W. S. Caughey, G. A. Smythe, D. H. O’Keefe, J. Maskasky, M. L. Smith,J . Biol. Chem., 250, 7602 (1975). 6. G. T. Babcock, J. Van Steelandt, G . Palmer, L. E. Vickery, I. Salmeen, Dev. Biochem., 5, 105 (1979). 7. T. Nozawa, Y. Orii, A. Kaito, T. Yamamoto, M. Hatano, in Developments in Biochemistry,

Vol.5 Cytochrome Oxidase, T. E. King, Y. Orii; B. Chance, K. Okunuki, eds., Elsevier, Amsterdam, 1979, p. 117. 8. T. VlnngLd, in Biological Applications of Electron Spin Resonance, H. M. Swartz, J. R. Bolton, D. C. Borg, eds., Wiley-Interscience, New York, 1972, p. 411. 9. J. Peisach, W. E. Blumberg, Arch. Biochem. Biophys., 165,691 (1974). 10. W. B. Mims, J. Peisach, R. W. Shaw, H. Beinert, J . Biol. Chem., 255, 6843 (1980). 11. B. M. Hoffman, J. E. Roberts, M. Swanson, S . W. Speck, E. Margoliash, Proc. Natl. Acad. Sci USA, 77, 1452 (1980). 12. W. Froncisz, C. P. Scholes, J. S. Hyde, H.-H. Wei, T. E. King, R. W. Shaw, H. Beinert, J . Biol. Chem., 254,7482 (1979). 13. L. Powers, B. Chance, Y. Ching, P. Angiolillo, Biophys. J., 34,465 (1981). 14. B. F. Van Gelder, H. Beinert, Biochim. Biophys. Acta, 189, 1 (1969). 15. G. T. Babcock, L. E. Vickery, G . Palmer,J. Biol. Chem., 251,7907 (1976). 16. M. F. Tweedle, L. J. Wilson, L. Garcia-Iniguez, G. T. Babcock, G. Palmer, J . Biol. Chem., 253, 8065 (1978).

A. J. Thomson, M. K. Johnson, C. Greenwood, P. E. Gooding, Biochem. J., 193,687 (1981). 18. M. F. J. Blokzijl-Homan, B. F. Van Gelder, Biochim. Biophys. Acta, 234,493 (1971). 19. T. H. Stevens, S . I. Chan, J . Biol. Chem., 256, 1069 (1981). 20. G. Palmer, G. T. Babcock, L. E. Vickery, Proc. Natl. Acad. Sci. USA, 73,2206 (1976). 21. W. E. Blumberg, J. Peisach, in Developments in Biochemistry, Vol. 5 , Cytochrome Oxidase, T. E. King, Y. Orii, B. Chance, K. Okunuki, eds., Elsevier, Amsterdam, 1970, p. 153. 22. J. 0. Alben, P. 0. Moh, F. G . Fiamingo, R. A. Altschuld, Proc. Natl. Acad. Sci USA, 78, 234 (1981). 23. G. T. Babcock, P. M. Callahan, M. R. Ondrias, I. Salmeen, Biochemistry, 20,959 (1981). 14.8.4.1.1. Reactions.

Oxidation of cytochrome c involves cytochrome a as the primary reaction site in a rapid biomolecular process ( k lo7 M-’s - l ) . Initial electron transfer is followed by electron redistribution within the oxidase and further oxidation of cytochrome c. Several enigmatic features process remain, e.g., cytochrome oxidase may function as a dime? with each monomer unit in the dimer having two cytochrome c binding sites.3 Despite these two sites, initial oxidation of c by oxidase results in reduction of substoichiometric amounts of a4, and indicates heretogeneity among the cytochrome a centers. Similarly, hetereogeneous behavior is observed in the redox chemistry of cytochrome a5, These observations suggest a model in which the redox and kinetic properties of the two cytochromes a in the dimer structure are differentiated, although unambiguous tests remain to be done. A second aspect of the cytochrome clcytochrome a reaction relates to the electron transfer distance. When the cytochrome c binding sites are occupied by the Fe(II1) species, the EPR parameters of neither a3+ nor c3+ are perturbed; this indicates that the distance between the paramagnets is greater than lOA. Both fluorescence6 and magnetic relaxation’ measurements support this conclusion. This observation makes difficult a simple outer-sphere electron transfer explanation for the high rates observed in the cy-

-

14.8.4. In Oxidases 14.8.4.1. In Cytochrome Oxidases 14.8.4.1 -2. Models.

446

tochrome c oxidation reaction; instead, a thermally activated tunneling mechanism is proposed.8 In addition to binding exogeneous ligands, the cytochrome a3 . . . Cu,a3 center catalyzes O2 reduction at rates that approach the diffusion limit under physiological conditions. However, antifreeze solvent systemsgand low temperature mixing techniques have been developed that overcome these high turnover rates". So far, results from the low temperature work show that 0, reduction by cytochrome oxidase proceeds through distinct intermediates with characteristic optical properties' '-13. Researchers agree that (1) the first intermediate involves a O,:Fe(II) cytochrome a3 complex analogous to oxyhemoglobin; (2) at some point in the reduction, 0, binds as well to Cu=a3; (3) intramolecular transfer to the Fea3 . . . 0, . . . Cu,a3 complex from Cu, and Fe, occurs at T well below - 50°C; and (4)at no time are partially reduced 0, intermediates released from the enzyme-active site. The latter observation rationalizes the low overvoltage for 0, reduction by cytochrome oxidase and explains the efficiency of the enzyme reaction. The detailed nature of the intermediate species, e.g., whether peroxy species, ferry1 iron, or organic redox centers are involved at some point in the catalysis, remains to be established. (G. T. BABCOCK) 1. B. G . Malstrom, Biochim. Biophys. Acta, 549, 281 (1979). 2. R. A. Capaldi, in Membrane Proteins in Energy Transduction, R. A. Capaldi, ed., Dekker, New York, 1979, p. 201. 3. S. Ferguson-Miller, D. L. Brautigan, E. Margoliash, J . Biol. Chem., 253, 149 (1978). 4. M. T. Wilson, C. Greenwood, M. Brunori, E. Antonini, Biochem. J., 147, 145 (1975). 5. R. P. Carithers, G . Palmer, J . Biol. Chem. 256, 7967 (1981). 6. J. M. Vanderkooi, R. Landsberg, G . S. Hayden, C. S. Owen, Eur. J. Biochem., 81,339 (1977). 7. T. Ohnishi, H. Blum, J. S. Leigh, Jr., J. C. Salerno, in Membrane Bioenergetics, C. P. Lee, G . Schatz, L. Emster, eds., Addison-Wesley, Reading, MA, 1979, p. 21. 8. B. Chance, C. Saronio, A. Waring, J. S . Leigh, Jr., Biochim. Biophys. Acta, 503, 37 (1978). 9. B. Chance, C. Saronio, J. S . Leigh, Jr., J . Biol. Chem., 250,9226 (1975). 10. G . M. Clore, L.-E. Andreasson, B. Karlsson, R. Aasa, B. G . Malmstrom, Biochern. J., 185, 139 (1980). 11. L. Powers, B. Chance, Y. Ching, P. Angiolillo, Biophys. J., 34,465 (1981). 12. M. Denis, Biochim. Biophys. Acta, 364, 30 (1981). 13. K. R. Carter, T. M. Anatlis, G. Palmer, N. S. Ferris, W. H. Woodruff, Proc. Natl. Acad. Sci USA, 78, 1652 (1981). 14.8.4.1.2. Models.

The iron of both cytochrome a and of a3 occur as the heme a complex (see 1 in 14.8.6.1) and can be successfully modeled. Iron(II1) and (11) bis-imidazole heme a complexes reproduce the optical, magnetic circular dichroism and EPR properties of cytochrome a in its 3 and 2 valence states, respectively'.2. Raman data indicate modifications at the periphery of heme a in the cytochrome a site which are not yet duplicated in vitro3. Reduced cytochrome a3, model compounds'*4 owing to the aqueous solvent systems used, only partially reproduce the spectroscopic properties of the in situ chromophore. Recently, an Fe(II), five-coordinate, high-spin heme a complex with a 2-methylimidazole as the axial ligand in aprotic solvents5 provides a suitable model for deoxy cytochrome a:+. Using low T techniques, analogous to those used in generating oxyprotoheme species as models for hemoglobin, oxy-heme a complexes are generated that mimic the spectral properties of oxy-cytochrome a:. The exchange-coupled Fe,,/Cu,, pair in the oxidized enzyme is the target of con-

+

+


14.8.4. In Oxidases 14.8.4.1. In Cytochrome Oxidases 14.8.4.1 -2. Models.

446

tochrome c oxidation reaction; instead, a thermally activated tunneling mechanism is proposed.8 In addition to binding exogeneous ligands, the cytochrome a3 . . . Cu,a3 center catalyzes O2 reduction at rates that approach the diffusion limit under physiological conditions. However, antifreeze solvent systemsgand low temperature mixing techniques have been developed that overcome these high turnover rates". So far, results from the low temperature work show that 0, reduction by cytochrome oxidase proceeds through distinct intermediates with characteristic optical properties' '-13. Researchers agree that (1) the first intermediate involves a O,:Fe(II) cytochrome a3 complex analogous to oxyhemoglobin; (2) at some point in the reduction, 0, binds as well to Cu=a3; (3) intramolecular transfer to the Fea3 . . . 0, . . . Cu,a3 complex from Cu, and Fe, occurs at T well below - 50°C; and (4)at no time are partially reduced 0, intermediates released from the enzyme-active site. The latter observation rationalizes the low overvoltage for 0, reduction by cytochrome oxidase and explains the efficiency of the enzyme reaction. The detailed nature of the intermediate species, e.g., whether peroxy species, ferry1 iron, or organic redox centers are involved at some point in the catalysis, remains to be established. (G. T. BABCOCK) 1. B. G . Malstrom, Biochim. Biophys. Acta, 549, 281 (1979). 2. R. A. Capaldi, in Membrane Proteins in Energy Transduction, R. A. Capaldi, ed., Dekker, New York, 1979, p. 201. 3. S. Ferguson-Miller, D. L. Brautigan, E. Margoliash, J . Biol. Chem., 253, 149 (1978). 4. M. T. Wilson, C. Greenwood, M. Brunori, E. Antonini, Biochem. J., 147, 145 (1975). 5. R. P. Carithers, G . Palmer, J . Biol. Chem. 256, 7967 (1981). 6. J. M. Vanderkooi, R. Landsberg, G . S. Hayden, C. S. Owen, Eur. J. Biochem., 81,339 (1977). 7. T. Ohnishi, H. Blum, J. S. Leigh, Jr., J. C. Salerno, in Membrane Bioenergetics, C. P. Lee, G . Schatz, L. Emster, eds., Addison-Wesley, Reading, MA, 1979, p. 21. 8. B. Chance, C. Saronio, A. Waring, J. S . Leigh, Jr., Biochim. Biophys. Acta, 503, 37 (1978). 9. B. Chance, C. Saronio, J. S . Leigh, Jr., J . Biol. Chem., 250,9226 (1975). 10. G . M. Clore, L.-E. Andreasson, B. Karlsson, R. Aasa, B. G . Malmstrom, Biochern. J., 185, 139 (1980). 11. L. Powers, B. Chance, Y. Ching, P. Angiolillo, Biophys. J., 34,465 (1981). 12. M. Denis, Biochim. Biophys. Acta, 364, 30 (1981). 13. K. R. Carter, T. M. Anatlis, G. Palmer, N. S. Ferris, W. H. Woodruff, Proc. Natl. Acad. Sci USA, 78, 1652 (1981). 14.8.4.1.2. Models.

The iron of both cytochrome a and of a3 occur as the heme a complex (see 1 in 14.8.6.1) and can be successfully modeled. Iron(II1) and (11) bis-imidazole heme a complexes reproduce the optical, magnetic circular dichroism and EPR properties of cytochrome a in its 3 and 2 valence states, respectively'.2. Raman data indicate modifications at the periphery of heme a in the cytochrome a site which are not yet duplicated in vitro3. Reduced cytochrome a3, model compounds'*4 owing to the aqueous solvent systems used, only partially reproduce the spectroscopic properties of the in situ chromophore. Recently, an Fe(II), five-coordinate, high-spin heme a complex with a 2-methylimidazole as the axial ligand in aprotic solvents5 provides a suitable model for deoxy cytochrome a:+. Using low T techniques, analogous to those used in generating oxyprotoheme species as models for hemoglobin, oxy-heme a complexes are generated that mimic the spectral properties of oxy-cytochrome a:. The exchange-coupled Fe,,/Cu,, pair in the oxidized enzyme is the target of con-

+

+

447

14.8.4. In Oxidases 14.8.4.1. In Cytochrome Oxidases 14.8.4.1-2. Models.

siderable synthetic effort’. Several imidazolate-bridged/binuclearcenters8 can be synthesized and characterized’-’*, including I”, but it appears that imidazolate cannot support an exchange coupling of the magnitude (1251 > 200 cm- ’) observed in the oxidase. Alternative models invoke c(-oxo‘~.’~ or ~ u l f u r as ’ ~ the bridging ligand. In response to these suggestions, 2, which has the average plane of the copper ligands parallel to that of the porphyrin planeI6, and 3, into which Fe, Cu, and potentially bridging ligands are inserted”, have been prepared. In the latter, the copper ligand plane is perpendicular to that of the porphyrin plane. It provides a more favorable geometry for magnetic coupling between the two metals, but only weak exchange is observed between CU” and high spin Fe3’ in these models. Another difficulty is that the iron of u3 in the exchangecoupled oxidized protein site is six-coordinate and in the porphyrin plane3, whereas the Fe in these models is five-coordinate. Considerable work remains before the structural requirements for a successful synthesis of the a3-Cu,, are elucidated.

0 1

(A) 0 0 = acacTTP = Tetraphenylporphyrin 2

n

oc I f

/

S

3

s\

Lco \

(G.T. BABCOCK)

1. G. T. Babcock, J. Van Steelandt, G. Palmer, L. E. Vickery, I. Salmeen, Dev. Biochem., 5, 105 (1979). 2. T.Nozawa, Y. Orii, A. Kaito, T. Yamamoto, M. Hatano, in Developments in Biochemistry, Vol. 5, T. E. King, Y. Orii, B. Chance, K. Okunuki, eds., Elsevier, Amsterdam, 1979, p. 117. 3. G. T. Babcock, P. M. Callahan, M. R. Ondrias, I. Salmeen, Biochemisfry,20, 939 (1981). 4. M. R. Lemberg, Physiol. Rev., 4 9 , 4 8 (1969). 5 . J. Van Steelandt-Frentrup,I. Salmeen, G. T.Babcock, J . Am. Chem. SOC.,103, 5981 (1981). 6. G. T. Babcock, C. K. Chang, FEBS Letts., 97, 358 (1979).

14.8. Bioinorganic Catalysis 14.8.4. In Oxidases 14.8.4.2. In Copper-Containing Oxidases

448 ~~

7. 8. 9. 10. 11. 12.

13. 14. 15. 16. 17.

~~

J. A. Ibers, R. H. Holm, Science, 209, 223 (1980). G.Palmer, G.T. Babcock, L. E. Vickery, POC.Natl. Acad. Sci. USA, 73,2206 (1976). R. N. Katz, G. Kolks, S. J. Lippard, Znorg. Chem., 19, 3845 (1980). R. H. Petty, B. R. Welch, L. J. Wilson, L. A. Bottomley, K. M. Kadish, J . Am. Chem. SOC., 102, 611 (1980). J. T. Landrum, C. A. Reed, K. Hatano, W. R. Scheidt, J. Am. Chem. SOC.,100, 3232 (1978). T. Prosperi, A. A. G. Tomlinson, J . Chem. SOC., Chem. Commun., 196 (1979). W. E. Blumberg, J. Peisach, in Developments in Biochemistry, Vol. 5 , Cytochrome Oxidase, T. E. King, Y.Orii, B. Chance, K. Okunuki, eds., Elsevier, Amsterdam, 1970, p. 153. C. A. Reed, J. T. Landrum, FEBS Let& 106, 265 (1979). L. Powers, B. Chance, Y.Ching, P. Angiolillo, Biophys. J., 34, 465 (1981). M. J. Gunter, L. N. Mander, K. S . Murray,J. Chem. SOC., Chem. Commun., 799 (1981). M. J. Gunter, L. N. Mander, K. S . Murray, P. E. Clark, J . Am. Chem. SOC., 103, 6784 (1981).

14.8.4.2. In Copper-Containing Oxidases

Protein-bound copper can carry out biological functions that range in complexity from straightforward electron transport activity through 0, storage and transport to 0, and substrate activation. Thus the coordination geometry of copper in its protein binding site is variable, being adapted to the specific catalytic activity of the protein. Several of these functions and structural adaptations are illustrated by copper oxidases, and, in particular, by the class of copper proteins referred to as "blue copper oxidases," which will be considered here'-5. The blue copper oxidases are similar to cytochrome oxidase in their ability to catalyze reduction of 0, to H,O. Catalysis is centered upon the protein-bound copper ions that can be differentiated into three classes according to their physical, chemical, and functional properties. They are designated Types 1,2, and 3 copper6. In the blue copper proteins (tree and fungal laccases, ceruloplasmin, ascorbate oxidase) these three classes of copper appear in varying amounts; the laccases contain the minimum amounts of each (one each of Types 1 and 2 and two Type 3 coppers). The presence of Type 1 copper accounts for the characteristc color of the blue copper oxidases. One of the distinguishing features of this class of copper is its relatively intense optical absorption near 600 nm ( E 2000-6000 M-' cm-'). Coupled with this unusually high visible extinction is the unusually small copper hyperfine interaction apparent in the EPR spectrum of Type 1 copper (0.0035 cm-

Inorganic Reactions and Methods- Reactions Catalyzed by Inorganic Compounds

Inorganic Reactions In Water

Mechanisms of Inorganic Reactions

Inorganic Reactions and Methods, The Formation of Bonds to Halogens (Part 2) (Inorganic Reactions and Methods)

Inorganic Reactions and Methods - Formation of Ceramics

Physical Inorganic Chemistry: Principles, Methods, and Reactions

Physical Inorganic Chemistry: Reactions, Processes, and Applications

Metal Catalyzed Cascade Reactions

Inorganic Reactions and Methods, Formation of Bonds to O, S, Se, Te, Po (Part 2) (Zuckerman: Inorganic Reactions and Methods)

Metal Catalyzed Cascade Reactions

Metal Catalyzed Cascade Reactions

Inorganic Reactions and Methods- Oligomerization and Polymerization Formation of Intercalation Compounds

Modern Rhodium-Catalyzed Organic Reactions

Modern Rhodium-Catalyzed Organic Reactions

Modern Rhodium-Catalyzed Organic Reactions

Metal-catalyzed cross-coupling reactions

Metal-catalyzed cross-coupling reactions

Metal-Catalyzed Cross-Coupling Reactions

Metal-Catalyzed Cross-Coupling Reactions

Electron transfer reactions: inorganic, organometallic, and biological applications

Molecular Modeling of Inorganic Compounds

Molecular Modeling of Inorganic Compounds

Inorganic Compounds with Unusual Properties

Spectroscopic Properties of Inorganic and Organometallic Compounds

Esterification: Methods, Reactions, and Applications

Transition-Metal-Catalyzed Reactions in Heterocyclic Synthesis

Esterification: Methods, Reactions, and Applications

Inorganic Reactions and Methods, Cumulative Index: Author and Subject Indexes (Volumes 1-18)

Inorganic Compounds with Unusual Properties II

Structural Chemistry of Inorganic Actinide Compounds

Structural Chemistry of Inorganic Actinide Compounds

Inorganic Reactions and Methods- Reactions Catalyzed by Inorganic Compounds

Inorganic Reactions In Water

Mechanisms of Inorganic Reactions

Inorganic Reactions and Methods, The Formation of Bonds to Halogens (Part 2) (Inorganic Reactions and Methods)

Inorganic Reactions and Methods - Formation of Ceramics

Physical Inorganic Chemistry: Principles, Methods, and Reactions

Physical Inorganic Chemistry: Reactions, Processes, and Applications

Metal Catalyzed Cascade Reactions

Inorganic Reactions and Methods, Formation of Bonds to O, S, Se, Te, Po (Part 2) (Zuckerman: Inorganic Reactions and Methods)

Metal Catalyzed Cascade Reactions

Metal Catalyzed Cascade Reactions

Inorganic Reactions and Methods- Oligomerization and Polymerization Formation of Intercalation Compounds

Modern Rhodium-Catalyzed Organic Reactions

Modern Rhodium-Catalyzed Organic Reactions

Modern Rhodium-Catalyzed Organic Reactions

Metal-catalyzed cross-coupling reactions

Metal-catalyzed cross-coupling reactions

Metal-Catalyzed Cross-Coupling Reactions

Metal-Catalyzed Cross-Coupling Reactions

Electron transfer reactions: inorganic, organometallic, and biological applications

Molecular Modeling of Inorganic Compounds

Molecular Modeling of Inorganic Compounds

Inorganic Compounds with Unusual Properties

Spectroscopic Properties of Inorganic and Organometallic Compounds

Esterification: Methods, Reactions, and Applications

Transition-Metal-Catalyzed Reactions in Heterocyclic Synthesis

Esterification: Methods, Reactions, and Applications

Inorganic Reactions and Methods, Cumulative Index: Author and Subject Indexes (Volumes 1-18)

Inorganic Compounds with Unusual Properties II

Structural Chemistry of Inorganic Actinide Compounds

Structural Chemistry of Inorganic Actinide Compounds

Recommend Documents