ISSN:0306-0012    

DOI:10.1039/CS99019FX005

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

ISSN 0306-001 2 CSRVBR 19(2) 83-195 (1990) Chemical Society Reviews Vol19 No 2 1990 Page The Mechanisms of Nucleophilic Substitution in Aliphatic compounds By Alan R. Katritzky, F.R.S. and Bogumil E. Brycki 83 NYHOLM LECTURE Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, and Dimer Chains By Robin J. H. Ckrk, F.R.S. 107 Stereoelectronic Origins of the Intrinsic Barrier to SN2 Reactions By Ikchoon Lee 133 Synthesis and Chemistry of Acyl Silanes By Philip C. Bulman Page, Sukhbinder S. Klair, and Step hen Rosent ha1 147 The Royal Society of ChemistryCambridge Chemical Society Reviews EDITORIAL BOARD Dr. M. J. Blandamer Professor H. W. Kroto F.R.S. (Chairman) Dr. A. R. Butler Professor J.A. McCleverty Professor B. T. Golding Professor S. M. Roberts Professor M. Green Professor B. H. Robinson Editor: Mr. K. J. Wilkinson Chemical Society Reviews (ISSN 0306-0012) is published quarterly and comprises approximately 20 articles (ca. 500 pp) per annum. Articles of three types appear: (a) personalized accounts of their own contributions by recognized authorities; (b) in-depth articles covering the state of the art of the subject under review; (c) introductory reviews of new topics, suitable for non-specialist readers. The texts of the lectures given by the Society’s named lecturers are also published in Chemical Society Reviews. Although the majority of articles are intended to be specially commissioned, the Society is always prepared to consider offers of articles for publication.In such cases a short synopsis, rather than the completed article, should be submitted to the Managing Editor, Books and Reviews Section, The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF. Members of the Royal Society of Chemistry may subscribe to Chemical Society Reviews at E22.00 per annum; they should place their orders on the Annual Subscription renewal forms in the usual way. All other orders accompanied with payment should be sent directly to The Royal Society of Chemistry, The Distribution Centre, Blackhorse Road, Letch- worth, Herts. SG6 1HN England. 1990 annual subscription rate U.K. E64.00, E.E.C. (x U.K.) E71, Rest of World g74.00, U.S.A.$144.00. Air freight and mailing in the USA. by Publications Expediting Inc., 200 Meacham Avenue, Elmont, New York 11003. USA. Postmaster: Send address changes to Chemical Society Reviews, Publications Expediting Inc., 200 Meacham Avenue, Elmont, New York 11003. Second class postage is paid at Jamaica, New York 1143 1. All other despatches outside the U.K. by Bulk Airmail within Europe, Accelerated Surface Post outside Europe. 0The Royal Society of Chemistry, 1990 All Rights Reserved No part of this book may be reproduced or transmitted in any form or by any means -graphic , electronic, in eluding p ho to copying, recording, taping ,or information storage and retrieval systems -without written permission from The Royal Society of Chemistry Published by The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF. Printed by Clays Ltd, St Ives plc

ISSN:0306-0012    

DOI:10.1039/CS99019FP007

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

~ COSHH IN LABORATORIES The Control of Substances Hazardous to Health Regulations 1988 (COSHH) forms the most signifi- cant legislation on occupational health in the UK since the Health and Safety at Work etc. Act was passed in 1974. This booklet gives practical guid- ance on the application of these regulations in labor- atories. The booklet was prepared by an expert Working Party of the Health, Safety and Environment Commit- tee (HSEC) of the Royal Society of Chemistry. It takes account of both the wide variety of laboratories that exist and the special problems that the implementa- tion of COSHH may cause in some of them. Broader questions of health and safety in chemical laboratories are dealt with in the Society's related publication 'Safe Practices in the Chemical Labora- tories'.Users are strongly recommended to read the two publications in conjunction. Contents Introduction; Assessment and control of risks to health; Maintenance of control measures; Monitoring; Health Surveil lance; Information , instruction and train- ing; Records and documentation of assessments; Glossary. Softcover 17 pp ISBN 0 85186 319 1 (1989)Price C7.50 SAFE PRACTICES IN CHEMICAL LABORATORIES This booklet is the successor to the Society’s ’Guide to Safe Practices in Chemical Laboratories’. Like its predecessor the new booklet points out relevant statutory requirements and provides general guid- ance on which specific in-house procedures can be based. The new booklet contains a Foreword by HM Chief inspector of Factories.Safe Practices in Chemical Laboratories takes ac- count of recent technical and legislative develop- ments affecting health and safety in chemical laboratories. In particular the Control of Substances Hazardous to Health Regulations 1988 (COSHH) will have profound implications for many laboratories and users are strongly recommended to read the new booklet in conjunction with the Society’s publications COSH H in Laboratories’ Brief Contents Acknowledgement; Foreword; Introduction; Organi- sation for Safety; Hazards; Design; Operation; Legis- lation and Bibliography. Softcover approx 50 pp Price f10.00 ROYAL SOCIETYOF CHEMISTRY 6 hfortTlallon SerVlCeS ISBN 0 85186 309 4 (1 989) POLLUTION: CAUSES, EFFECTS AND CONTROL 2nd Edition Edited by R.M.Harrison, University of€ssex The second edition of Pollution: Causes, Effects and Control has been considerably updated and ex-panded, reflecting the great changes that have taken place since the first edition was published in 1983. It contains two new chapters dealing with radioactive pollution and the chemistry and pollution of the sta- tosphere, reflecting the importance now attached to these areas, and gives each author greater space in which to cover his topic. The treatment is essentially introductory, although some aspects are covered in greater depth. The contributions combine to give a broad overview, touching on most of the important areas and delving deeper into many of them.The environment is now high on the political agenda, illustrating the need for authoritative scientific infor- mation. This book is a must for graduates and under- graduates with an interest in environmental and pollution research. Softcover 394 pp ISBN0 85186 2837 Price €29.50 March 1990 RayALSOCIETY m CHEMISTRY Mormatwxl !hVK= SUPERVISION OF TECHNICAL STAFF: AN INTRODUCTION FOR LINE SUPERVISORS Edited by R. Weston, Leicester Polytechnic D.C. Norton, Ex-Chief Technician, Brornley College of TechnologyM. Grimshaw, North fastSurrey College of Technology This unique book forms an introduction to supervisoryskills for line supervisors employed in scientific, educational, medical and industrial laboratories.Unlike other publications on supervision it is written specifically for supervisors working in laboratories and concentrates on the specific skills associated with the control of staff in scientific laboratories. The authors have considerable experience as laboratory supervisors and in teaching technical staff. Brief Contents Organization The Role of the Supervisor within the Laboratory LeadershipOrganization, Planning, and the Technical Supervisor Motivation Recruitment and Selection Salaries and Grading Induction and Monitoring of Staff TrainingCounselling and Discipline Industrial Relations: the Supervisor and the Trade Unions Health and Safety The Law and the Supervisor The Supervisor and New Technology Softcover Approx 242pp ISBN 085186 423 6 October 1989 Price: f15.95ROYAL SOCIETY OF CHEMISTRY lnformatm servwes

ISSN:0306-0012    

DOI:10.1039/CS99019BP009

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Chern. SOC.Rev., 1990,19,83-105 The Mechanisms of Nucleophilic Substitution in Aliphatic Compounds * By Alan R.Katritzky, F.R.S. DEPARTMENT OF CHEMISTRY, UNIVERSITY OF FLORIDA, GAINESVILLE, FL 32601, USA Bogumil E. Brycki DEPARTMENT OF CHEMISTRY, ADAM MICKIEWICZ UNIVERSITY, 60780 POZNAN. POLAND 1 Introduction The following article reviews the insight into nucleophilic substitution mechanism that has been made available from studies conducted in non-polar solvents and in the gas phase. It argues for the existence of discrete alternative mechanisms often acting in competition. 2 Nucleophilic Substitution and Charge In the whole of organic chemistry there is no reaction more important than the replacement by a nucleophile of a leaving group attached to an aliphatic carbon atom.’ In its most general form this reaction involves the conversion of a nucleophile and a substrate into a product and a leaving group (equation 1).Nucleophile + R-X* -Nucleophile-R + X* (l)? In neutral substrates, such as alkyl halides (l),the leaving group is negatively charged [e.g.,halide anion (2) or arylsulphonate anion]. R-Br Br- R‘ R-N +-R” I I R”’ R’ N-R“ I I R“’ neutral anionic cationic neutral substrate leaving group substrate leaving group (1) (2) (3) (4) In cationic substrates, the leaving group is neutral: e.g., a tertiary amine from a quaternary ammonium cation [cj (3), (4)], a sulphide from a sulphonium salt, an ether from an oxonium salt. Furthermore, acid catalysis can convert a neutral substrate into a cationic substrate: many reactions of alcohols are of this type (equation 2), and the leaving group is effectively a neutral water molecule.* For more detailed accounts of our own work in this area see reference 31. t Throughout this article, when there are various possibilities of charge signs, as for X here, charges are omitted and asterisk is used instead. J. March, ‘Advanced Organic Chemistry’, 3rd edn.; Wiley, New York, 1985; (a) pp. 265--268; (6) pp. 259-265. The Mechanisms of Nucleophilic Substitution in Aliphatic Compounds R-OH % R-6H2 -R+ + OH2 alcohol protonatedalcohol neutral water (5) (6) (7) Nucleophiles are either anionic (e.g., OH-, OR-, halide anion, RS-, RSOz-, etc.) or neutral (e.g., H20, ROH, amines of all types).Thus, from the point of view of charge type, there are four main classes of nucleophilic substitutions, as first depicted by Ingold in equations 3-6: NU-+ RX-R-NU + X-(3) NU + RX-R-NU+ + X-(4) NU-+ RX+-R-Nu + X (5) Nu + RX+ -R-Nu+ + X (6) It can be seen that in equations 3 and 6 charge is conserved, while in equation 4 charge is created, and in equation 5 charge is destroyed. These distinctions are of fundamental importance, as will become apparent. 3 Unimolecular and Bimolecular Modes of Nucleophilic Substitution Ingold first saw clearly that not all nucleophilic substitutions occur by the same mechanism. He distinguished between the bimolecular mode of reaction, where the rate IS first order both in the nucleophile and in the substrate concentration, and the unimolecular mode, where the rate depends only on the concentration of the substrate and is independent of that of the nucleophile.Ingold interpreted2 these two alternative types of kinetic behaviour, respectively, as the SN2 direct displacement (equation 7), and as the SN~type mechanism, where the first, effectively irreversible step is followed by fast reaction of the resulting carbocation with nucleophile (equation 8). R-X* %R+ + X*, followed by R+ + Y* %R-Y* (8) In the 1950s Winstein showed4 convincingly that the sN1 mechanism was more complex than had been postulated by Ingold. Working with neutral C K Ingold, 'Structure and Mechanism in Organic Chemistry', (a) 1st edn ,Cornell University Press, New York, 1953,pp 306418, (b) 2nd edn, Cornell University Press, New York, 1969,pp 418-610 C K Ingold and E Rothstein, J Chem SOC,1928,1217-1221 (a) S Winstein and D Trifan, J Am Chem SOC,1952, 74, 1154-1160, (b) S Winstein and G C Robinson, J Am Chem SOC,1958, 80, 169-181, (c) S Winstein, P E Klinedmst, Jr, and G C Robinson, J Am Chem SOC,1961,83,885-895 Katritzky and Brycki substrates, he obtained evidence that the ionization took place in several stages and he distinguished between ‘intimate ion pairs’ (IIP) and ‘solvent-separated ion pairs’ (SSIP) as two distinct types of discrete intermediate before the stage of a free carbocation was reached (Scheme 1). R-x2% R+X-A R+//x---+R+ + x-ki IIP SSIPI I I“ul ko “UI k2 “ull “ul PRODUCTS Scheme 1 In the 1960s, Sneen pointed out’ that these intimate ion pairs should also be able to react in the bimolecular mode with a nucleophile (equation 10).Although Sneen’s postulate that all sN2 reactions involve intimate ion pairs was later shown to be untrue,6 the mechanism must be considered as a possible alternative in favourable cases. Sneen mechanism R-x-@-, R+ . . . X---3~,R-N~ intimate rate ion pair determining Another potential complication is the participation in the substrate of an adjacent substituent in the unimolecular bond-breaking phase of the SN~ rea~tion.~Carbocations readily undergo rearrangement (equation 1l), and the postulate is that such rearrangement could begin before the bond-breaking was complete, and might accelerate it.7 As shown in Scheme 2, the substrate (9) can undergo direct sN2 reaction to give (8), or form the intimate ion pair (11) which in turn could form unrearranged product (8) or rearrange to carbonium ion (12).In the participation hypothesis, carbonium ion (12) would be formed directly from (9). RR/CH&R” -R~HCHR’R” (1 1) Use of available experimental evidence to distinguish between these alternative possible pathways of Scheme 2 has been controversial. Direct sN2 displacement [(9) -(8), Scheme 21 by solvent as nucleophile to yield unrearranged product, (a) R A Sneen, Acc Chem Res, 1973,6,46-53, (b)R A Sneen and J W Larsen, J Am Chem SOC, 1969,91,362-366, (c) R A Sneen and J W Larsen, J Am Chem SOC,1969,91,6031-6035 6(a)T W Bentley and P v R Schleyer, J Am Chem SOC,1976, 98, 7658-7666, (6) F L Schadt, T W Bentley, and P v R Schleyer, J Am Chem SOC,1976,98, 7667-7674, (c) T W Bentley and P v R Schleyer, Adv Phys Org Chem, 1977,14, 1-67 ’(a) S Winstein and H Marshall, J Am Chem SOC,1952, 74, 112k1126, (b) A Streitwieser, ‘Solvolytic Displacement Reactions’, McGraw-Hill, New York, 1962 85 The Mechanisms of Nucleophihc Substitution in Alzphatic Compounds Scheme 2 Solvolysis processes for primary systems (Reproduced from reference 39 with permission) was believed by Winstein 7a to dominate normally, with occasional competitive first-order anchimerically assisted heterolysis [(9) ----,(12)], that could be rate- determining and was followed by fast formation of rearranged product [(12) -(13), Scheme 21.Anchimeric assistance by H or Me transfer, was rejected by later workers and the results interpreted in terms of path (9) -(8), path (9) --+ (11) -(12) -(13), and path (9) (11) -(8) of Scheme 2.8-’0 However, controversy remains l1 and, in particular, evidence from secondary kinetic isotope effects favours participation.’ 2-’4 More recently, it has been shown that electron transfer processes can also be involved in nucleophilic substitution in general,I5 and substitution reactions on charged substrates in particular.16 4 Symbolic Representation of Nucleophilic Substitution Reaction Mechanisms The use of poorly defined terms has caused confusion about reaction mechanisms in chemistry.The system introduced by Ingold suffers from having to serve both as a phenomenological description of the observed features of reactions (substitu- tion, elimination) and as a statement of the mechanism of the reaction V J Shiner, Jr , ‘Isotope Effects In Chemical Reactions’, ed C J Collins and N S Bowman, Van Nostrand, Reinhold, New York, 1970, p 90 J E Nordlander, S P Jindal, P v R Schleyer, R C Fort, Jr ,J J Harper, and R D Nicholas, J Am Chem SOL,1966,88,4475-4484 lo W M Schubert and W L Henson, J Am Chem SOC,1971,93,6299-6301 J M Harris, Prog Phys Org Chem, 1974,11,89--173 l2 T Ando, H Yamataka, H Morisaki, J Yamawaki, J Kuramochi, and Y Yukawa, J Am Chem SOC, 1981,103,43-34 l3 V J Shiner, Jr and R C Seib, Tetrahedron Lett, 1979,20, 123 126 l4 V J Shiner, Jr and J J Tai, J Am Chem SOC,1981,103,436442 (u) I P Beletskaya and V N Drozd, Russ Chem Rev (Eng Trunsl), 1979, 48, 431448, (b) M Chanon and M L Tobe, Angew Chem ,Int Ed Engl, 1982,21,1-23 l6 A R Katntzky, In ‘Substituent Effects in Radical Chemistry’, ed H G Viehe, Reidel Publishing Co, 1986, pp 347-360 Katritzky and Brycki (molecularity, concertedness, electronic characteristics). The Ingold nomenclature is also sometimes ambiguous in its interpretations of mechanisms, perhaps most noticeably for &2-&1 reactions in solvolysis and bimolecular substitutions; some quite different mechanisms fall under the same designation.Moreover, the original Ingold system has been overburdened with non-systematic modification for 30 years.A new, clear descriptive nomenclature of reaction mechanism is that recom- mended by IUPAC which deals directly with the basic currency of molecular change: bond making and bond breaking.” This system describes the most important properties of reaction mechanisms, e.g. the number of steps in the reaction, the sequence of these steps and their nature, including significant diffusion steps. The formation of a new bond during the transformation of one molecular structure to another is symbolized by ‘A’ (association or attachment) and, conversely, the bond breaking components are symbolized by ‘D’(dissociation or detachment). These ‘A’ and ‘D’ symbols are referred to as ‘primitive changes’. When the changes take place in separate reaction steps, they are punctuated by a ‘+’.A symbol ‘*’is used instead of the ‘+’ to designate an intermediate which is of such a short lifetime that it reacts in a step faster than diffusion but slower than a molecular vibration. Sets of non-punctuated ‘A’ and ‘D’symbols corre- spond to ‘elementary reactions’. The subscript ‘N’is used to designate bond forma- tion to a nucleophile or bond scission with loss of a nucleofuge (i.e.leaving group). There is also provision for describing homolytic and cyclic mechanisms and for extra-mechanistic information, including the class of transformation, the nature of the substitution, and the occurrence of catalysis, using easily pronounced terms. According to the above rules, the mechanisms of nucleophilic substitution at saturated carbon atoms can be described as follows: ANDN Concerted; one step.DN*AN Stepwise; short-lived intermediate, e.g. ion pair. Components are not dif-fusionally equilibrated with the bulk solvent. DN + AN Stepwise; the intermediate can diffuse thorough the solvent. These correspond to the Ingold sN2, the ion-pair, and the S,l mechanisms, re- spectively. 5 Why Are Different Mechanisms Followed Much of the experimental work on reaction mechanisms has been concerned with fitting reactions into the Ingold scheme, or other schemes, such as Winstein’s classification of ion-pair intermediates in solvolysis reactions. In comparison there has been comparatively little inquiry into the questions of why a reaction should follow one mechanism rather than another under a particular l7 (a) R.D. Guthrie, Pure. App. Chem., 1989, 61, 23-56; (b) R. Guthrie and W. P. Jencks, Acc. Chem. Res., 1989,22,343-349. The Mechanisms of Nucleophihc Substitution in Aliphatic Compounds set of experimental conditions and what is the nature of the transition from one mechanism to another as reactants or conditions are changed The answers to these questions are determined in large part by the reaction intermediates whether intermediates are formed at all, what are their structures, and what are their properties Many of the controversies regarding mechanisms come from uncertainty or disagreement about the existence and behaviour of intermediates An intermediate can be defined as a species with a lifetime that is longer than that of a molecular vibration of s, ie a species that has barriers to breakdown both to reactants and to products l8 Since the distinction between mechanisms of chemical reactions in solution are mainly concerned with the sequence in which reactants are assembled and dispersed in relation to the bond-making and bond-breaking steps, the choice of reaction mechanism is dictated by the lifetimes of intermediates that may be formed in a reaction A reaction can proceed through a stepwise, monomolecular reaction mech- anism when the intermediate has a significantly long lifetime in the solvent, but no lifetime when it is in direct contact with a stronger nucleophilic reagent in the solution and this reagent is favourably oriented Conversely, a concerted, bimolecular reaction mechanism, will take place only when such an intermediate does not exist, ie this species is not stable for a few vibration frequencies and there is no barrier for its collapse Concerted reactions with a solute and stepwise reactions with solvent can coexist in the same solution The existence of a solvent-equilibrated intermediate can be diagnosed in several ways (a) by trapping the intermediate, (b) by demonstrating a constant partitioning ratio to products of a common intermediate that is derived from different reactants, (c) by observing a change in rate-limiting step with increasing concentration of reactant or catalyst, (d) by demonstrating a diffusion-controlled reaction with the intermediate * When the lifetime of the intermediate becomes shorter, the reaction may proceed through an enforced preassociation mechanism in which a molecule that will react in a second step (or process) forms an encounter complex with the other reactant (or reactants) before the first step (or process) takes place A concerted mechanism also requires preassociation of the reactants and can be considered as a special case of a preassociation mechanism The reaction coordinate+nergy diagram of Figure 1 provides a convenient way to illustrate the above points The diagram, which has wings added to describe diffusional steps, is drawn with the carbon-nucleophile bond order along the abscissa, the carbon-leaving group bond order on the ordinate, and Gibbs energy indicated in the third dimension by the contour lines, in accordance with the usual convention for such diagrams, most of the energy contour lines are omitted for clarity The fully stepwise mechanism with a solvent-equilibrated carbocation intermediate proceeds through the pathway on the outside of the W P Jencks, Acc Chem Res, 1980,13, 161-169 l9 W P Jencks, Chem Soc Rev 1981 10,345 375 Katritzky and Brycki C...Nu.C '.* NU^ Figure 1 Reaction coordinate diagram for nucleophilic substitution on carbon with wings for the transport steps. The contour lines are omitted except for the lower left corner (Reproduced from reference 18 with permission) diagram (dotted lines), a reaction with an ion-pair in a weakly nucleophilic solvent enters the centre of the diagram at the upper right, and a preassociation mechanism can proceed through the intermediate in the central box, if it exists, or through concerted pathways 2 and 3 with varying amounts of cationic character in the transition state.The above discussion may be illustrated by a study of 1-phenylethyl systems.20-22 The lifetimes of carbocation intermediates were estimated by diffusion-controlled trapping with azide, extrapolating these lifetimes to less stable cations and taking account of the nucleophilic reactivity of added nucleophiles. It was shown that the appearance of concerted bimolecular reactions occurs only when the cation is predicted to have no lifetime in the presence of the nucleophile and the change from stepwise to a concerted mechanism occurs when the intermediate ceases to have a lifetime in the presence of nucleophile.6 Behaviour at Mechanistic Borderlines A longtime source of controversy has been the interpretation of behaviour at mechanistic borderlines. Although data is available for many 'borderline' 'O J. P. Richard and W. P. Jencks, J. Am. Chem. SOC.,1984,106,1383-1396. J. P. Richard and W. P. Jencks, J. Am. Chem. SOC.,1982,104,56894691. 22 J. P. Richard and W. P. Jencks, J. Am. Chem. SOC.,1982,104,4691 -4692. The Mechanisms of Nucleophilic Substitution in Aliphatic Compounds reactions, the interpretation of the results is often made uncertain by the assump- tions and corrections that have to be applied when treating the experimental observations For example, in reactions involving anionic nucleophiles, correc- tions have to be applied for salt effects and, in some cases, for incomplete dissociation of the salt used to supply the nucleophile An important aspect of the controversy has been whether borderline behaviour represents simply the availability of two alternative reaction paths in competition, or a gradual transition from one pathway to another In the first interpretati~n,~~ the dominant path under one set of conditions becomes less important as another mechanism becomes more important In the second interpretation,’ there exists a ‘spectrum of merging mechanisms’, with a character that is intermediate between sN1 and sN2 modes It has frequently been suggested that a clear-cut distinction between reaction mechanisms is impossible because, for example, there is a gradual transformation of an &2 into an sN1 mechanism with no sharp borderline as the transition state develops more carbocation character 24 Uncertainty on this point complicates the teaching of this important subject and is the source of much confusion at all levels However, a clear distinction can be made if the classification of mechanism is based upon the lifetime of intermediates rather than the character of the transition state The lifetime of intermediates permits a fairly sharp qualitative distinction between mechanisms, whereas the character of the transition state or the degree of assistance by the solvent in a reaction gives only a quantitative description with no sharp boundaries l8 The term ‘spectrum of merging mechanisms’ is widely used in the literature This implies that there is no clear distinction between mechanisms But an intermediate either exists or does not exist, with a fairly sharp distinction between these possibilities, so that a reaction either proceeds in one step, or two steps, or more steps It cannot proceed in one and a half steps Use of terms such as ‘merging mechanism’ has contributed to much of the confusion in this field and should be discouraged 7 Gas Phase Investigations In addition to the structure of the substrate, and the nature of the nucleophile and leaving group, the overall rates of sN1 (DN+ AN)and sN2(ANDN) reactions are strongly medium-dependent Sometimes, solvent effects can completely change relative nucleophilicities, as has been shown for halide anions in protic and aprotic solvents2’ Reaction in the gas phase, in the absence of the complicating solvent effects, can shed light on nucleophilicities, leaving group abilities, and steric effects sN2 (ANDN)type reactions have been extensively studied in the gas phase 23 (a) P Caspieri and E R Swart, J Chem SOC,1961, 4342-4347 (b) A Fava, A Iliceto, and A Ceccon, Tetrahedron Lett, 1963,4685-692 24 S Winstein, E Grunwald, and H W Jones, J Am Chem SOC,1951,73,2700--2705 25 S Winstein, L G Savedoff, S Smith, I D R Stevens, and J S Gall, Tetrahedron Lett, 1960, 1(9), 24-30 Katritzky and Brycki (1L 1 (15 1 (16 1 (17) Scheme 3 Fragmentation of alkyl-pyridinium cations to pyridine and carbocation where sN2 (AN&) reactions of anionic nucleophiles with neutral substrates can be faster than in solution due to preferential solvation of reactants compared to the transition state.26 The experimental results were interpreted by a model in which the collision complex and a complex of the leaving entities were separated by an energy barrier.Such reactions proceed in the gas phase with inversion of configuration at the carbon atom, just as in solution.26 Anions with a localized charge are better gas-phase nucleophiles than those with a delocalized charge: thus benzyl anions are poor nucleophiles despite their large methyl cation affinitie~.~~ In solution, polarizable nucleophiles are better than non-polarizable ones because they can respond better to demand for charge reorganization.The gas-phase studies 26 show just the opposite behaviour and indicate that the higher nucleophilicity of polarizable anions in solution is a consequence of stronger solvation of the anions with more concentrated charge. Relative nucleophilicities are similar in the gas phase and in dipolar aprotic solvents but very different from those found in protic solvents. However, leaving group abilities are similar for all three media.26 Interestingly, gas-phase studies of N-alkylpyridinium cations showed that they do indeed undergo fragmentation to neutral pyridine and the carbocation in the gas phase, similarly to their reactions in non-polar solvents (Scheme 3).28 In addition, many compounds which by beta-elimination can form a strain-free olefin, choose a second pathway to yield protonated pyridine and this olefin (Scheme 4).29 The kinetic energy required for dissociation (the ‘appearance energy’) was determined by measuring the percentage of collisionally activated dissociation (CAD) as a function of the relative kinetic energy.Since reliable thermochemical data are not available for comparison with experimentally derived appearance energies, theoretically calculated energies of pyridine +-R bond cleavage were applied. Heats of formation were calculated for the pyridinium ions Py+-R, for 26 W.N. Olmstead and J. I. Brauman, J. Am. Chem. Soc., 1977,99,42194228. 27 D. K. Bohme and L. B. Young, J. Am. Chem. SOC.,1970,92,7354-7358. 28 (a) C. H. Watson, G. Baykut, Z. Mowafy, A. R. Katritzky, and J. P. Eyler, Anal. Instrument., 1988, 17, 155-162; (6) A. R. Katritzky, C. H. Watson, Z. Dega-Szafran, and J. R. Eyler, J. Am. Chem. Soc., 1990,112,2471-2478. 29 A. R. Katritzky, C. H. Watson, Z. Dega-Szafran, and J. R. Eyler, J. Am. Chem. SOC.,1990,112,2479-2484. + +0 -o+>c=c, / N+Q=Q II4 I \CAI/H y+ I /c-c< / c\ \cO-c’eH-. H 0 \ pyridine, and for the alkyl cations R+ using the AM1 method.30 From these AHf values the theoretical heats of dissociation AAHf for the process of Schemes 3 and 4 were calculated using equation 12: For those alkyl-pyridinium cations where the carbocation cannot, or is unlikely to, undergo any rearrangement to a more stable system the appearance potentials are either higher or within 3 kcal mol-’ of the calculated AAHf values from equation 12 (methyl, allyl, adamantyl, 1-benzotriazolylmethyl, cinnamyl, diphenylmethyl, or triphenylmethyl pyridinium cations: see Figure 2).By contrast, for many other N-substituted pyridinium cations, the appearance potentials were found to be considerably less than those that would be required to produce an unrearranged carbocation. However, in all such cases it was found that rearrangement pathways were available to the carbocation that would reduce its heat of formation so that the appearance potential would provide enough energy.Thus the low appearance potential found for phenylthiomethyl cation (23) (Figure 2) indicates that rearrangement of (23) occurs through (24) almost all the way to (25) with this AAHf of 22 kcal mol-’ closest to the appearance potential of 32 kcal mol-’. + 6‘ Comparison of the calculated AAHf values with the appearance potentials indicates that for P-phenylethyl cations not only is their dissociation to unrearranged carbocations precluded, but also that rearrangement of (26) to 30 M. J. S. Dewar, E. G. Zoebisch, E. F. Healy, and J. J. P. Stewart, J. Am. Chem. SOC.,1985, 107, 3902-3910. Katritzky and Brycki CH2=CHCHz+ p’”: -a YU 10-10-1 I 1 I 0 20 40 60 80 100 A AHf [kcal mol-’1 Figure 2 Plot of experimentally estimated appearance energies (AE) against calculated heats of formation (AAHf) of R from N-substituted pyridinium salts: unrearranged cations + (0);cations with successively more extensive rearrangement (A).[Values from reJ 28(b)] spiro (27) during the dissociation process by anchimeric assistance does not provide enough energy.The low appearance potentials suggest that alterna- tive rearrangement of (26) to (28) must occur to a significant extent during the dissociation and possibly that further rearrangement of (28) to (29) occurs. The dissociation processes of N-alkylpyridinium cations in the gas phase invoke the Py R+ion-molecule pair as an important intermediate in which non-stabilized R+can rearrange to a more stable structure.8 The Role of the Solvent The investigation of the detailed mechanism of nucleophilic substitutions of neutral substrates in solution is particularly difficult because of the charge that is created in the transition state of an SN~(DN + AN)type reaction (equation 13). Because of this charge creation, the SN~(DN + AN) reaction mode is neither expected nor found for neutral substrates in non-polar solvents. Moreover, in polar media, the solvent is invariably a potential nucleophile and it may be difficult to disentangle whether such a solvent is behaving simply as a medium of dielectric constant sufficiently high to allow the charge creation in an SN~(DN + The Mechanisms of Nucleophilic Substitution in Allphatic Compounds AN)type process (equation 13), or whether it is behaving as a nucleophile in a sN2 (ANDN)type mechanism.R-X -R6+ X6-(transition state) (13) R-X+ -R6+---X6+(transition state) (14) Such difficulties of interpretation are far less for cationic substrates, because charge is spread in the transition state rather than created: contrast equation 14 with equation 13. Thus sN1 (DN+ AN)type reactions of cationic substrates are expected to, and do, occur in non-nucleophilic solvents of low dielectric constant.31 Their behaviour in such solvents can be followed and extrapolated through media of increasingly greater polarity. 9 Investigation of Cationic Substrates Surprisingly, until quite recently very little mechanistic work has appeared on the use of cationic substrates to investigate the mechanisms of nucleophilic substitu- tion at sp3 carbon centres, the most notable exception being the t-butyldimethyl- sulphonium cation.32 A study of the mechanistic aspects of nucleophilic substitution at saturated carbon atoms where a neutral heterocyclic species was the leaving group commenced in one of our research groups in 1978.Originally this work was motivated by the need for understanding a reaction of considerable synthetic potential,33 but it soon became evident that it could lead to a deeper under- standing of the mechanism of nucleophilic substitution in general.The second part of this review summarizes the conclusions from this work.The first fundamental question to be resolved was whether the different mechanisms remained distinct at borderlines or whether they merged into each other. Positive evidence will first be presented for the occurrence of distinct mechanisms in competition and the detailed behaviour of primary, secondary, and tertiary substrates will then be reviewed. 10 Positive Evidence for the Occurrence of Distinct sNl(DN + AN) and sN2 (AN&) Reactions Plots of substrate rates us. nucleophile concentration, under pseudo-first-order conditions, gave straight lines for a variety of primary and secondary alkyl substrates with a range of nucleophiles in several non-polar and non-nucleophilic solvent^.^'.^^ However, there was a highly significant difference in the behaviour 31 (a) A R Katritzky and G Musumarra, Chem Soc Rev, 1984, 13, 47-68, (b) A R Katritzky, K Sakizadeh, and G Musumarra, Heterocycles, 1985, 23, 1765-1813, (c) A R Katritzky and B E Brycki, J Phys Org Chem, 1988,1, 1 20 32 (a) C G Swain, L E Kaiser, and T E C Knee, J Am Chem SOC,1958, 80, 4092-4094, (b) D N Kevill, W A Kamil, and S W Anderson, Tetrahedron Lett, 1982,23,46354638 33 (a) A R Katritzky, Tetrahedron, 1980, 36, 679-699, (b) A R Katritzky and C M Marson, Angew Chem ,Int Ed Engl ,1984,23,420--429 34 A R Katntzky, G Musumarra, K Sakizadeh, S M M El-Shafie, and B Jovanovic, Tetrahedron Lett, 1980,21,2697-2699 Kutritzky and Brycki 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Nu(rno1 I-') Figure 3 Nucleophilic substitutions by simultaneous sN1 (DN+ AN) and sN2 (ANDN) reactions: kobsfor l-isopropyl-2,4,6-tr@henylpyridinium cation (30) (1.6 x M) plotted vs.nucleophilic concentration (chlorobenzene solution, 100 "C)(Reproduced from reference 34 with permission) I 4---' Ph 10( 0.0 0.1 0.2 0.3 0.4 NU (mol I-') Figure 4 Rate variation with N-substituent: kobsfor reactions of N-substituted 2,4-diphenyl- 5,6-dihydrobenzo[h]quinolinum cations (36),(38), (40), (41) (6.4 x M) with piperidine in chlorobenzene at 100 "C (Reproduced from reference 3 1 with permission) of the two types of alkyl substrates. The plots for secondary alkyl groups showed a positive intercept which was invariant with the nature of the nucleophile, as seen for example in Figure 3 for the 1-isopropyl-2,4,6-triphenylpyridiniumcation (30).34By contrast, the lines for primary alkyl substrates generally did not show such intercepts: in Figure 4this contrasting behaviour is illustrated for a series of benzo[h]quinolinum cations (36),(38), (40),and (41).The Mechanisms of Nucleophilic Substitution in Aliphatic Compounds Positive intercepts, which are invariant with the nature of the nucleophile, have also been found for many reactions in aqueous solution, including the reactions of methoxymethoxy-2,4-dinitrobenzenewith neutral nu~leophiles,~~ the reactions of N-(methoxymethy1)-N,N-dimethyl-m-nitroaniliniumcation with anionic nucleo- philes and N-(methoxymethy1)-N,N-dimethyl-p-bromoaniliniumcation with neu- tral nucleophiles 36 as well as reactions of 1-(4-nitrophenyl)ethyl chloride with anionic nu~leophiles.'~ In these cases, the positive intercepts could be caused by concerted reactions of substrates with water.However, the positive intercepts in the non-polar, non-nucleophilic solvents clearly indicate that unimolecular (DN+ AN)* and bimolecular reactions (ANDN) proceed simultaneously and independently for the secondary alkyl substrates. Much further evidence, part of which is summarized in the following sections, supports this conclusion. 11 Nucleophilic Substitution with primary-Alkyl Substrates Primary-alkyl substrates have a variety of reaction pathways open to them as seen in Scheme 1. We shall see that they normally react by a bimolecular sN2 (ANDN)mode although SN~(DN+ AN) type reactions can occur under some conditions favouring such behaviour.In addition to the evidence of Figure 4, where no intercepts were found for various substrates reacting with the same nucleophile (piperidine) further evidence for the bimolecular sN2 (AN&) mode is given in Figure 5. This shows separate plots of kobsfor three different nucleophiles (piperidine, morpholine, and pyridine) at various concentrations reacting with l-benzyl-2,4,6-triphenyl-pyridinium (31) at 100"C in chlorobenzene solution.34 Rates for each nucleophile plot as a straight line which passes through the origin, showing that the reaction is first order in nucleophile. The second-order rate constant is proportional to the slope of the line in the plot, and as expected is greatest for piperidine, less for morpholine, and by companson very much smaller for pyridine, a much less powerful n~cleophile.~~ A distinction between the two possible alternative types of sN2 reactions, i.e., classical on the substrate and Sneen-type on the intimate ion-molecule pair, was reached from pressure experiments.The classical SN~(ANDN)reaction rate is enhanced by pressure (i.e.,the AVt is negative), because the two reactants will be pushed closer together.37 However, for a Sneen-type SN2 (DN*AN)reaction on an intimate ion-molecule pair, a large positive AVs is expected because the pre- equilibrium (equation 10) will be pushed to the left by increasing pressure.The * The products of solvolysis reactions in non-nucleophihc solvents in the absence of nucleophile have been elucidated (see ref 47) N-Alkylpyridinium tetrafluoroborates solvolyse in the absence of nucleophiles in chlorobenzene to give products of alkylation both of the solvent and of the pyridine leaving group In nitrobenzene, only the leaving group is alkylated 3s G A, Craze, A J Kirby, and R J Osborne, J Chem SOC,Perkrn Trans 2, 1973,357-363 B L Knier and W P Jencks, J Am Chem SOC,1980,102,6789-6798 37 (a)T Asano and W J Le Noble, Chem Rev, 1978,78,407489,(b)M Okamoto, M Sasaki, and J Osugi, Rev Phys Chem Jpn ,1977,47,3343 Katritzky and Brycki 120 100 -80 c v) -12 60* LOa0 20 0 0.0 0.1 0.2 0.3 0.L NU (mol 1-l) Figure 5 Nucleophilic substitution by SN2(AN&) reaction only; kobsfor 1-benzyl-2,4,6-triphenylpyridinium cation (31) (1.6 x lC3 M) plotted vs.nucleophilic concentration (chlorobenzene solution, 100 "C)(Reproduced from reference 34 with permission) Ph Ph Ph&\ R \ (30) R z isopropyl (36)R: methyl (L2) R benzyl (31) R :benzyl (37)R= isopropyl (43 1 R :n -propyl (32) R=p-methoxybenzyl (38)R.n- butyl (441 R :n-pentyl (33) R~s-butyl (39)R: S-butyl (45) R :n-octyl (31) R = cyclopentyl (40)R:ally1 (16)R isobutyl 2(351 R: cyclohexyl (41)R= benzyl (47) R :neopentyl second stage of the reaction (see equation 10) will possess a negative AV*, but its magnitude should be smaller than that for the pre-equilibrium; thus overall the reaction rate is expected to decrease with increasing pressure.The rate of the reaction of l-p-methoxybenzyl-2,4,6-triphenylpyridiniumperchlorate (32) with piperidine clearly decreases with pressure indicating the intimate ion-molecule pair mechanism (Figure 6). Most interestingly, the reaction rate of the N-benzylpentacyclic derivative (42)with piperidine initially decreases with increas- ing pressure, but then passes through a minimum and starts to increase (Figure 7). This indicates that the reaction at normal and fairly low pressures is via the intimate ion-molecule pair (&*AN) but as the pressure increases, a competing The Mechanisms of Nucleophilic Substitution in Aliphatic Compounds -9.5 In k -10.0 LO 80 120 P(MPa) Figure 6 Reaction via intimate ion-molecule pairs reaction of l-p-methoxybenzyl-2,4,6-triphenylpyridinium perchlorate (32) (2 0 x M) with piperidine (0 1 M) at 30 "C in chlorobenzene solution at varying pressures (Reproduced from reference 38 with permission) I' 50 100 150 P( MPa) Figure 7 Competing reactions at high pressure the pseudo-first-order rate constant for the reaction of N-benzyl-S,6,8,9-tetrahydro-l-phenyl-bis-benzo~a,h)acridiniumtetra~uoroborate(42) (2 0 x M)with piperidine (2 0 x 10 M) at 30 "C in chlorobenzene as a function of pressure(Reproduced from reference 38 with permission) reaction by the classical sN2 (AN&) process gradually takes over 38 We conclude that both the alternative modes of SN2-type behaviour can occur For primary alkyl substrates, SN~(DN + AN) behaviour should be favoured by the most active leaving groups and by multiple P-branching of the alkyl group which hinders &2 We therefore studied the series of N-primary-alkyl acridinium "A R Katntzky, K Sakizadeh, B Gabnelson, and W J Le Noble, J Am Chem Soc 1984, 106 1879-1 880 Katritzky and Brycki ions (43)-(45).Compounds (43)-(49, with a straight chain alkyl group, sol- volyse in deuteriated methanol (CH30D) and in deuteriated acetic acid (CH3- C02D) to give mixtures of normal [(9) -(8)] and rearranged [(9) -(13)] methyl ethers and acetate esters, re~pectively.~~ None of the solvolysis pro- ducts contain deuterium and hence none are formed via olefin intermediates [(lo), Scheme 21.The rearranged products [(13), Scheme 21 must arise by an SN~ type mechanism, either by route (9) -(11) ---(12) --+ (13) (Scheme 2) or by the anchimerically assisted path (9) (13). Supporting the former (12) -route is the fact that the absolute rates disclose no evidence for rate-enhancing anchimeric assistance when P-phenyl or P-methoxy groups are present. The unrearranged products are likely to be formed by an sN2 (ANDN)type mechanism. In methanol the individual rates for the production of unrearranged products from the N-alkylacridinium ions decrease dramatically in the order: n-alkyl >> i-butyl >> neopentyl This parallels the behaviour of the corresponding tosylates, and is in agreement with the classical SN~(ANDN)path (9) -(8) of Scheme 2.Acetolysis of the N-isobutylacridinium ion (46) involves an important non- olefinic pathway yielding both isobutyl and sec-butyl acetate, and acetolysis in CH3C02D of the neopentyl derivative (47) yields undeuterated neopentyl acetate [cf. (8)] as well as deuterated tert-pentyl acetate [cf. (13), Scheme 21. The rates for the formation of unrearranged products for all the alkyl groups (n-alkyl, isobutyl, and neopentyl) are constant within a factor of -4; this cannot be reconciled with path (9) -(8), but is just what is expected for the ionization path (9) --+ (1 1) --+ (8), i.e., predissociation to an ion-molecule pair (cf. equation 10). For the solvolysis of the N-alkylacridinium cations in acetic acid, the rates for the formation of rearranged products (other than by way of an elimination reaction) show that the ratio of migration to direct substitution is nearly constant over the series for hydrogen migration (n-Pr, n-Pent, n-Oct), and again nearly constant over the series for methyl migration (i-Bu, neo-Pent).As the ion- molecule pair (11) is an intermediate in the formation of the unrearranged products by path (1 1) --+ (8), this supports path (1 1)(9) --(12)-(9) -(13), with the intimate ion pair (or ion-molecule pair) (11) a common intermediate for the formation of both the unrearranged and the rearranged products. By contrast, where the direct substitution occurs by path (9) -(8), i.e., as deduced for the N-alkylacridinium ions in MeOH and for the tosylates in EtOH and AcOH, such constancy of ratios is neither expected nor observed.The similarity between the ratios of the rate of hydrogen or methyl migration in an alkyl tosylate in CFSCOOH (at 75 "C) with those for the corresponding 39 A. R. Katritzky, Z. Dega-Szafran, M. L. Lopez-Rodriguez, and R. W. King, J. Am. Chem. Soc., 1984, 106,5577-5585. The Mechanisms of Nucleophilic Substitution in Aliphatic Compounds h c VI-20 --!n 0 -Phul n 0 .4 "I 0.0 0.2 0.4 Piperid ine (mol I-') Figure 8 Rate variation with N-substituent: kobsjfor reactions of N-substituted-2,4,6-triphenylpyridinium cations (33), (34), (35) (1.6 x 10-M)with piperidine in chlorobenzenes at 100 "C (Reproduced from reference 40 with permission) N-alkylpyridinium ion in AcOH (at 150 "C) suggests that a similar mechanism by path (9) +(11) --+ (12)-(13)operates for the tosylates.2b 12 Nucleophilic Substitution with secondary-Alkyl Substrates s-Alkyl substrates usually react by a combination of SN~(DN+ AN) and SN2 (ANDN)modes. Plots for the reactions of l-isopropy1-2,4,6-triphenylpyridinium cation (30) with piperidine, morpholine, and pyridine as nucleophiles in chlorobenzene were shown in Figure 3.The straight lines do not pass through the origin, but give the same significant intercept at zero nucleophile concentra- tion. Alongside the second order SN2 (AN&) component there is thus a first order SN~(DN+ AN)component, which is independent both of the amount and of the nature of the added n~cleophile.~~ Similar behaviour is shown for other secondary alkyl substrates.In Figure 8 (monocyclic series) and Figure 9 (tricyclic series) the nucleophile is kept constant, as piperidine, but the N-substituent is varied (isopropyl, secondary butyl, cyclopentyl, and cycl~hexyl.)~~ The activation entropies for the SN~(DN+ AN) reaction mode are less negative than those for sN2 (AN&) reactions:' and this pattern is found4' for the individual components of several reactions depicted in Figures 8 and 9. The SN~component can be either classical (DN+ AN),involving free car- bonium ions, or take place by fast capture of ion-molecule pairs formed in the 40 A. R. Katritzky, K.Sakizadeh, Y. X. Ou, B. Jovanovic, G. Musumarra, F. P. Ballistreri, and R. Crupi, J. Chem. SOC.,Perkin Trans. 2,1983,1427-1434. 41 S. R. Hartshorn, 'Aliphatic Nucleophilic Substitution,' Cambridge University Press, London, 1973, p. 81. Katritzky and Brycki 500--s -Bu 400300- @ph/rJI/ o! I I, J 0.0 0.2 0.4 NU (mol I-’) Figure 9 Rate variation with N-substituent:kobsfor reactions of N-substituted-2,4-diphenyl-5,6-dihydrobenzo[h]quinolinumcations (37), (39)(6.4 x lC5M)with piperidine in chloro-benzene at 100 OC (Reproduced from reference 40 with permission) rate-determining step (DN*AN).These modes are distinguished by the sensitivity of rates and products to the solvent type and to added nucleophile:40 (i) In non-nucleophilic solvents (chlorobenzene, acetonitrile, chloroform), with added nucleophile (piperidine) there is no rate dependence on nucleophile concentration and the products show no rearrangement of the carbon skeleton, i.e., Winstein SN1 (ion-molecule pair) mechanism applies.(ii) Solvolysis in weakly nucleophilic solvents (1,1,1,3,3,3-hexafluoropropan-2-01, trifluoroacetic acid, 2,2,2-trifluoroethanol) gives partially rearranged products in the absence of added nucleophile, i.e.,a classical SN~(DN+ AN)mechanism. (iii) In the presence of nucleophile (morpholine) the products of solvolysis in 1,1,1,3,3,3-hexafluoropropan-2-o1are not rearranged although the rates are unaffected by the nucleophile concentration. This proves that rearrangement occurs after the formation of the carbocation, and is prevented by trapping an intermediate after the transition state by the added nucleophile.A Winstein SNl mechanism must apply as no free carbocations are involved and the reaction is first order. (iv) Solvolysis in nucleophilic solvents (pentanol, acetic acid) gives unrear-ranged products, i.e., classical SN~is excluded by analogy and a Winstein SN1 (ion-molecule pair) mechanism is implied. Clearly under all these conditions an SN~type mechanism is involved. In (i) and (iii) the added nucleophile is able to intercept the incipient carbocation before rearrangement. This indicates the formation of an intimate molecular-ion pair in a rate determining stage. Arguments related to the variation of rates with solvent polarity parameters 101 The Mechanisms of Nucleophilic Substitution in Aliphatic Compounds Ph _I) Ph Ph (49) Ib) rPh Ph Me\*APh I OAc Complete inversion (571 Scheme 5 Evidence for intimate ion-molecule pair (Reproduced from reference 43 with permission) indicate42 that in nucleophilic solvents (pentanol or acetic acid), in addition to the SN~mechanism, a competitive sN2 mechanism occurs, where the acetic 42 A.R. Katritzky, M. L. Lopez-Rodriguez, and J. Marquet, J. Chem. SOC.,Perkin Trans. 2, 1984, 349- 354. Katritzky and Brycki acid or pentanol molecules are acting not simply as solvent but also as nucleo- philes. Additional evidence is available for the formation of intimate ion-molecule pairs from secondary alkyl substrates. N-a-Methylallyl-2,4,6-triphenylpyridinium cation (49) rearranges in solution to the y-methylallyl isomer (51) via an intimate ion-molecule pair (50) (Scheme 5a) because the 2,4,6-triphenylpyridine-free molecule is stable and unreactive and if formed from (49) could not give (51).43 Furthermore, optically active a-phenylethylamine and 2,4,6-triphenylpyrylium cation in acetic acid solvent give a reactive pyridinium ion which immediately forms the corresponding acetate with complete inversion (Scheme 5b); this demonstrates that no free carbocation PhCH(Me) is formed which would result + in ra~emization.~~ I R (58) R: adamantyl (59) Rz t-butyl (60)R 1-methyl-1 -phenylethyl The above evidence shows that secondary alkyl substrates can undergo nucleophilic substitution by the sN2 (AN&) and by both of the two types of SN~ reaction; that via the free carbonium ion (DN+AN), and that involving an intimate ion-molecule pair (DN*AN) as intermediate.13 Nucleophilic Substitution with tertiary-Alkyl Substrates Here the pattern is much more simple: SN1 (DN +AN) type mechanism only. 1-( l-Adamantyl)-(58), l-t-butyl-(59), and 1-(1-methyl-1-phenylethyl) pyridi- nium cations (60) solvolyse at rates independent of solvent polarity, of solvent electrophilicity, or of solvent nu~leophilicity.~~ In Figure 10, the solvent polarity parameter of Dimroth, &,45 measures the overall solvation ability.The ET scale corresponds to a linear combination of solvent dipolarity, II*,and hydrogen bond donor acidity, a (a in turn corresponds to solvent electrophilicity 46). These cationic substrates (solid points) give rates which vary less with the substrate structure than do those of the corresponding compounds with anionic leaving groups (open points), as clearly illustrated in Figure 9. This is due to the fact that charge is created in the transition state of the latter class of compounds, but not 43 A. R. Katritzky, Y. X. Ou, and G. Musumarra, J. Chem. SOC.,Perkin Trans 2,1983, 1449-1454. 44 A. R. Katritzky and B. Brycki, J. Am. Chem. SOC.,1986,108,7295-7299. 45 C. Reichardt, Angew. Chem.,Int. Ed. Engl., 1979,18,98-110. 46 R. W. Taft and M. J.Kamlet, J. Chem. SOC.,Perkin Trans. 2, 1979, 1723-1729. 47 A. R. Katritzky, C. M. Marson, J. L. Chen, F. Saczewski, and R. W. King, J. Chem. SOC.,Perkin Trans. 2, 1986, 1331-1337. 103 The Mechanisms of Nucleophilic Substitution in Aliphatic Compounds 4L2 + O:2 40 Dtoxane t-hu~~In-BuOH I -PrOH Figure 10 Plots against ET of logarithms of observed rate constants for the solvolysis in various solvents of (a) 1-(1-methyl-1-phenylethy1)pyridiniumperchlorate (60) at 80 "C, (b)cumyl chloride at 25 "C, (c) 1-t-butylpyridiniwn perchlorate (59) at 180 "C, (d) t-butyl chloride at 180 "C,(e) 1-(1-adamanty1)pyridiniumperchlorate (58) at 190 "C,(f)1-adamantyltosylate at 190 "C,and (g) 1-adarnantyl chloride at 50 "C (Reproduced from reference 45 with permission) in the former, where charge is only dispersed.Thus, halide solvolysis rates show a large sensitivity to solvent polarity. The rates for the cationic substrates (58)-(60) are unaffected by pH change, and by the presence of nucleophiles. Hence, these t-alkylpyridinium cations solvolyse by an sN1 (DN+ AN)type mechanism. There is no evidence for any participation of the solvent. 14 General Conclusions We have presented evidence for the occurrence or non-occurrence of four mechanisms of nucleophilic substitution in solution (Table). Katritzky and Brycki Table Mechanisms found for different classes of substrate Mechanism Substrate primary-alkyl s-alkyl t-alkyl Classical sN1 yes one or Classical SN~ yes Winstein sN1 yes yes one or Yes }both no Sneen sN2 Yes }both no These mechanism types remain distinct without merging at mechanistic borderlines.We find no evidence for rate enhancing anchimeric assistance in p-branched n-alkyl groups in solution. In the present review, we have not included evidence for electron transfer mechanisms; this can occur with certain substrates and nucleophiles, and has been summarized in part elsewhere.I6 These conclusions apply to cationic substrates; for neutral substrates direct evidence is much more difficult to obtain, but it is tempting to extrapolate all these conclusions to all substrate types. Acknowledgements. We thank many friends and colleagues for discussions (at times quite vigorous), particularly Drs. M. A. Battiste, J. A. Deyrup, W. P. Jencks, and J. A. Zoltewicz.

ISSN:0306-0012    

DOI:10.1039/CS9901900083

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Chern. Soc. Reu., 1990,19,107-131 NYHOLM LECTURE * Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, and Dimer Chains By Robin J. H. Clark, F.R.S. CHRISTOPHER INGOLD LABORATORIES, UNIVERSITY COLLEGE LONDON, 20 GORDON STREET, LONDON WClH OAJ 1 Introduction This lectureship, founded in 1973 to commemorate the name of Sir Ronald Nyholm, is intended in alternate years to be concerned with subjects of interest to inorganic chemists and to have regard for his wide international interests. Little restriction is imposed by the former condition; nor, for an antipodean with a penchant for travel, does the latter. Early studies at Canterbury University College of the University of New Zealand with W. S. Metcalf on diffusion controlled reactions, at the University of Otago with W.S. Fyfe on the effects of pressure on the electrical conductivity of weak electrolytes, and at University College London with Ron Nyholm on the chemistry of the early transition elements, led to my own research interests having a very wide and international base. Complete changes of research area early in one’s career, disruptive as they are in the narrowest sense, nevertheless are an enriching experience, the potential benefits of which are too often overlooked by graduates in the UK today. Ron Nyholm is remembered with warmth and real affection by all who knew him,’ and he is also remembered for several of his classic statements regarding inorganic chemistry. His definition of the subject, delivered in his inaugural lecture at University College London on 1 March 1956,* was as follows: ‘The integrated study of the formation, composition, structure, and reactions of the chemical elements and their compounds, excepting most of those of carbon.’ Over a sherry or two, the word ‘most’ would readily become displaced by ‘some’.He was also perceptive in recognizing the paramount need for equipping Chemistry Departments with all relevant apparatus if they were to carry out first- class research. To quote: ’ ‘Without this costly capital investment in research equipment, university depart- ments must be prepared to limit their objectives to relatively neglected fields away from the broad front of modern scientific advancement and to forfeit * Delivered at the Nyholm Symposium, University College London, on 6 December 1989, and on other occasions at Lancaster, Bristol, Christchurch, Wellington, Dundee, Manchester, and Glasgow.D. P. Craig, Biographical Memoirs of Fellows of the Royal Society, 1972,445. R. S. Nyholm, Inaugural Lecture, ‘The Renaissance of Inorganic Chemistry’, University College London, 1 March 1956. Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc. Figure 1 The structure ofthe [Re2Cl8I2-ion their place at the frontiers of knowledge to those with the necessary instruments of attack.’ These comments remain as true today as they were when first made in 1956. Ron Nyholm was responsible for initiating the systematic exploitation of a range of methods for the study of structure and properties of inorganic compounds. But his contribution to the subject was greater than implied by this comment alone.Joseph Chatt summed the situation up well when, in his Memorial Lecture of 1972, he said: ‘It was the infectious enthusiasm and forceful lecturing of Nyholm which prompted the renaissance of Inorganic Chemistry in the UK during the 1950’s.’ According to my own observations, the only matters to which Ron Nyholm took exception were those to do with (a) people who spilled o-phenylene- bisdimethylarsine, the immensely versatile chelating ligand with which he made his name and which, in 1958, took three months to synthesize (b) late arrival of research students in the laboratory (c) derogatory references to Australia, and (d) students making a mockery of examinations.One of Ron Nyholm’s greatest interests in 1971 was the topic of metal-metal bonding and, accordingly, I have based the Nyholm lecture on or around this topic. His interest began, in association with Jack Fergusson, with the synthesis of the red complex CSR~C~~.~,~ On account of its diamagnetism it was initially thought that this complex might prove to contain the first example of a spin- paired (e4) tetrahedral ion. Although this did not prove to be the case, nevertheless the complex was found in 1963 to be a trimer Cs3[Re3C112] possessing an extremely interesting structure. The triangle of rhenium atoms is the key feature in which three rhenium-rhenium bonds of bond order two are impli~it.~-~ Shortly after the discovery of this complex another substance with the same empirical formula was isolated as blue crystals.This proved to be Cs2[Re2C18], a J Chatt, Plenary Lecture, ‘Ronald Sydney Nyholm’, International Conference on Coordination Chemistry, Toronto, 1972 J Fergusson, PhD thesis, University of London 1960 W T Robinson, J E Fergusson, and B R Penfold, Proc Chem Soc, 1963, 116 J A Bertrand, F A Cotton, and W A Dollase, J Am Chem SOC,1963,85,1349’J A Bertrand, F A Cotton, and W A Dollase, lnorg Chem , 1963,2, 1166 Clark complex involving the highly symmetric (D4h) anion [Re2C18I2 -(Figure l).8*9 Although I was not at this stage directly involved, research in this area was clearly a matter of great general interest.The recognition that ions of this sort must be held together by rhenium-rhenium bonds alone (formally quadruple) prompted many further studies and developments in the chemistry of metal-metal bonded complexes." It was not until the early 1970's that I took a direct interest in metal-metal bonded complexes, and that came about through the realization that they were ideal subjects for electronic, infrared, Raman, and, in particular, resonance Raman studies.' ',I2 Such knowledge was clearly of critical importance to the understanding of the electronic and vibrational spectra of the ions and thus for the proper understanding of the bonding. In these contexts, the [M2Xg]"-ions were of pivotal importance.Subsequently, important questions arose in the spectroscopy of the analogous M2(carboxylate)4 complexes. Both subjects are addressed in this lecture, which is then developed into the areas of linear-chain and dimer-chain chemistry, structure, and spectroscopy. 2 Metal-Metal Bonded Complexes A. Spectroscopy of [M2X8IR-Ions.-The general philosophy of the research on metal-metal bonded complexes at University College London in the 1970's revolved around the realization that it was essential to be able to assign the electronic and vibrational spectra of structurally and chemically simple metal- metal bonded complexes before it was worthwhile paying attention to these matters for more complicated metal-metal bonded complexes. With this in mind, it was important to demonstrate that very intense RR spectra of the [Mo2C18I4-, [Mo2Br8I4-, [Re2C1812-, and [Re2Br8I2- ions could be obtained at resonance with the lowest electronic bands of these Such spectra (A-term RR spectra)12 took the form of long progressions in the v1 mode, the ReRe stretching mode, implying (a) that the resonant electronic transition is electric dipole allowed and (b) that on transition from the ground to this excited state the ions suffer a substantial change to the metal-metal bond length.These results allowed the resonant electronic transition to be assigned to the 'Azu +-6* +--6 transition since, on excitation to the 6* state, the metal-metal bond order would be reduced from four to three with consequential elongation of the metal-metal bond.Some years later it proved possible to elaborate on these studies in a very effective way. First, the synthesis of the remaining two ions [Re2F812- and [Re21g12-of the set of four structurally similar [Re2Xsl2-ions was F. A. Cotton, N. F. Curtis, B. F. G. Johnson, and W. R. Robinson, Inorg. Chem., 1965,4,326. F. A. Cotton and C. B. Harris, Inorg. Chern., 1965,4330. lo F. A. Cotton and R. A. Walton, 'Multiple Bonds between Metal Atoms', Wiley, New York,1982. l1 R. J. H. Clark in 'Advances in Infrared and Raman Spectroscopy', Heyden, London, 1975, p. 143. '' R. J. H. Clark and T. J. Dines, Angew. Chem., In?. Ed., 1986,25, 131. l3 R. J. H. Clark and M. L. Franks, J. Am. Chem. SOL'.,1975,97,2691. l4 R. J. H. Clark and M.L. Franks, J. Am. Chem. SOL'.,1976,98,2763. Is R. J. H. Clark and N. R. DUrso, J. Am. Chem. SOC.,1978,100,3088. Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc. Second, the development of new lasing dyes in the blue (stilbene 1 and 3) and red (LD 700) together with the availability of UV lines (Ar' +,Kr2+) permitted relatively easy coverage of the wide excitation range of 330-800 nm. Thus it became possible to excite within the contour of each electronic band of each ion; in this way the nature of each excited state of each ion could be probed. Resonance Raman spectra of each ion, taken at resonance with the electronic band of lowest wavenumber in each case (Figure 2), display long progressions (Figure 3) in v1, the symmetric ReRe stretching mode (318, 275, 276, and 257 cm-' for X = F, Cl, Br, and I, respectively).This clearly indicates the similar nature of the lowest electronic transition of each ion and, by comparison with the earlier work, the assignment of the lowest band to the 6* -6, lAZu +-'A1, transition. This conclusion is substantiated by the observation that the depolariza- tion ratio of the v1 bands of [Re2F8]'-, [RezCl~]~-and [RezBr~]'- are all 4at resonance, a situation which can only obtain if the resonant electronic transaction is z-polarized (consistent with its assignment to a 'Azu t--'A1, transition in the point group D4h). Irradiation within the contour of the second electronic transition of each ion produces entirely different RR spectra in each case.These spectra (Figure 4) are characterized by resonance enhancement to bands attributed to the v2 mode and its overtones, where vz is the totally symmetric metal-halogen stretching mode (624, 362, 211, and 152 cm-' for X = F, C1, Br, and I, respectively). Thus the principal structural change undergone by each ion on excitation is, in this case, along the metal-halogen coordinate, a result in keeping with that expected in consequence of a non-bonding-to-antibondinghalogen-to-metal charge-transfer transition. The assignment of the second strong band in the electronic spectra of the ions to the b1,(6*) f--(x)eg(n),'E, +'Alg,transition follows naturally. This assignment is confirmed by the fact that the measured depolariza- tion ratio of the vz band of the [Re~Cl8]~- and [ReZBr8l2- ions at resonance is i,a situation which can only obtain if the resonant electronic transition is xy polarized.This is precisely the polarization required (in D4h) for the 'EU+ 'A1, transition. RR spectra obtained at resonance with the third electronic band of each ion lead to resonance enhancement to both the v1 and vz bands (and their overtones) to comparable extents. Although no firm assignment of the resonant electronic band can be made in this case, the results are consistent with the assignment eg(n*)t-e,(n). This transition has been shown, by XaSCF calculations 19-z1 to be between metal-based orbitals with a considerable amount l6 W. Preetz and L. Rudzik, Angew.Chem., 1979,91, 159. W. Preetz, G. Peters, and L. Rudzik, 2.Naturforsch., Ted B, 1979,34, 1240. H. D. Glicksman and R. A. Walton, Inorg. Chem., 1978,17,3197. I9A. P. Mortola, J. W. Moskowitz, N. Rosch, C. D. Cowman, and H. B. Gray, Chem. Phys. Letters, 1975,32,283. 2o F. A. Cotton, J. MoI. Struct., 1980,59,97. 21 R. J. H. Clark and M. J. Stead, in 'Inorganic Chemistry Towards the 21st Century', ACS Symposium Series, No. 211 ed. M. H. Chisholm, 1983, p. 235. Clark [Re, Br, 12-30 25 20 15 10 Wavenumber / lo3cm-' Figure 2 Electronic spectra of the complexes [(n-C4H9)4N]2[Re2X8] at -14 K m the region of their 6* -6, 6* -n(X), and x* -x(?) transitions (Cs[BF4], KCl, KBr, and CsI discs for X = F, C1, Br, or I, respectively)21*22 of halogen character, and thus structural changes along both metal-metal and metal-halogen bond lengths would be expected, as implied by the RR results.RR spectra of the sort described21*22 thus demonstrate that the nature of and assignments for electronic bands of structurally simple inorganic molecules can be made in a new and very convincing way. With some confidence, then, one could proceed to the study of larger, more structurally complex molecules. 22 R. J. H. Clark and M. J. Stead, Inorg. Chem, 1983,22,1214. Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc. [Re~Cl~l~'h,=647.1 nm 1 I Re2&, 12-A, = 752.5 nm 1000 500 Wavenumber / cm-l Figure 3 Resonance Raman spectra of the complexes [(~-C~H~)~N]Z[R~ZX~]at -80 K in the region of 6 transitions21'22their 6* -B.Rhodium-Rhodium Stretching Frequencies in Dirhodium Tetracarboxy1ates.- The second problem in the area of metal-metal stretching frequencies to which attention will be drawn is the (until 1986) controversial one of the assignment of rhodium-rhodium stretching frequencies, v(RhRh), in dirhodium tetracarboxylate complexes. The initial indications of San Filippo and Sniad~ch,*~ subsequently supported by Kharitonov et and Kireeva ef al., 25 were in favour of the range 150-170 cm-' for v(RhRh) in a variety of complexes of the type 23 J. San Filippo and H. J. Sniadoch, Inorg. Chem., 1973, 12,2326. 24 Y. Y. Kharitonov, G. Ya. Mazo, and N. A. Knyazeva, Russ. J.Inorg. Chem. (Engl. Transl.), 1970, 15, 739. 25 I. K. Kireeva, G. Ya. Mazo, and R. N. Shchelekov, Russ.J. Inorg. Chem. (Engl. Transl.), 1979,24, 220. Clark [Re21e 12- h0=676.4 nm 1 1000 500 0 Wavenumber / cm-' Figure 4 Resonance Raman spectra of the complexes [(n-C4H9)4N]2[Re2Xs] at -80 K in the region of their 6* -n(X) transitions21*22 Rh2(02CR)4L2 (L = axial ligand). On the other hand Ketteringham and Oldham,26who also studied a range of complexes of this sort, favoured the range 288-351 cm-' for v(RhRh). Raman and extensive electronic spectral analyses by Miskowski et ~1.~~on Rh2(02CCH3)4(H20)2 and Li2Rh2(02CCH3)4C12.8H20 were based on the lower value for v(RhRh) and were given some theoretical basis as a result of molecular mechanics calculations.28 Despite this wealth of research extending over 16 years there nevertheless appeared to be no $rm evidence one way or the other regarding the correct wavenumber region of v(RhRh) in dirhodium tetracarboxylate complexes.It thus seemed highly desirable that detailed electronic, infrared, Raman, and resonance Raman studies of 26 A. P. Ketteringham and C. J. Oldham, J. Chem. Soc., Dalton Trans., 1973, 1067.''V. M. Miskowski, W. P. Schaefer, B. Sadeghi, B. D. Santarsiero, and H. B. Gray, Inorg. Chem., 1984, 23, 1154. J. C. A. Boeyens, F. A. Cotton, and S. Han, Inorg. Chem., 1985,24, 1750. Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc. Resonance Raman Spectra of (a) Rh,(02CCH,14(PPh3)2, (bl Rh2('e02CCH3)4(PPh312and (cl Rh2(02CCD314(PPh3)2 I *h.-v) c +J t-4 360 340 320 300 280 Wavenumber / cm-' Figure 5 Resonance Raman spectrum of (a) Rh2(160zCCH3)4(PPh3)2, (b) Rhz-('802CCH3)4(PPh3)2,and (c) Rh2('602CCD3)4(PPh3)z as KCI discs at ca.80 K (10= 363.8 nm). The v1 bands are truncated in each case 33*34 Rh2(02CCH3)4(PPh3)2 as its l60,l8O, and CD3 variants should be undertaken in order to resolve this problem definitively in one case at least. The choice of triphenylphosphine as axial ligand is advantageous for resonance Raman studies since the electric-dipole-allowed metal-based O-D* transition of this complex (and of related ones) occurs in the near UV region and is accessible with ATZ+ 363.8 nm excitation.The HOMO is considered to be the ulg (in D4h) orbital with mainly Rh-Rh d,-d, bonding but also some Rh-P do-n antibonding character. Raman spectra excited at resonance with electric dipole allowed transitions are usually much simpler than off-resonance spectra. Under these A-term conditions, it is only bands attributable to totally symmetric modes which are enhanced, together with bands attributable to overtones and combination tones involving the enhanced bands."*12 The resonance Raman spectra of Rh2(02CCH3)4(PPh& (Figure 5) convinc-ingly demonstrate in the following way that v(RhRh) occurs at ca. 289 cm-'. Any structural change attendant upon cr -cr* excitation would be expected to occur primarily along the RhRh bond rather than along the virtually orthogonal Rho bonds.Hence it is v(RhRh) and its overtones which are expected to undergo the greatest intensification in a Raman spectrum taken at resonance with the cr -cr* transition. The results indicate that it is the 289 cm-' band which undergoes the resonance enhancement, and is thus to be assigned to vl, v(RhRh). No band in the vicinity of 170 cm-' appears in the spectra at all. Although v(RhP) might also be expected to be enhanced at resonance with the cr-cr* transition it has not been positively identified. It could not be identified with VI since this moves from 289.3 cm-I to higher wavenumbers for Clark the analogous AsPh3 and SbPh3 derivatives (297 and 307 cm-l, respectively), i.e. in the opposite direction to that expected on both mass and Rh-M bond-length grounds [for the complexes Rh2(02CCH3)4.(MPh3)Z7 r(Rh-M) increases in the order RhP, 2.477 A, <RhAs, 2.576 A, <RhSb, 2.732 A].29930The axial nature of the resonant transition is confirmed by the fact that the depolarization ratio of the v1 band is 4 at resonance.The resonance Raman spectra of the l8O and CD3 analogues of Rhz(02- CCH3)4(PPh3)2 confirm this assignment convincingly. The most resonance-enhanced band, vl, occurs at 289.3, 289.1, and 287.5 cm-' for the l60, '*O, and CD3 derivatives, respectively, indicating the virtual independence of v1 of the equatorial modes, which is the situation to be expected for v(RhRh) since this coordinate is nearly orthogonal to v(Rh0). On the other hand the two bands at 338.4 and 321 cm-' in the resonance Raman spectrum of the l60complex are shifted by 6.4 and 7 cm-', respectively, to lower wavenumbers by l80 substitution, consistent with the behaviour expected for modes in which the Rho stretching coordinate dominates (Figure 5).On deuteration, these modes move by rather more than this (12-14 cm-'). The combination of resonance Raman and isotopic work thus demonstrates conclusively that, for Rh2(02CCH3)4(PPh3)2, v(RhRh) is at 289.3 cm-'. Since the RhRh bond length in this complex is 2.4505(2) A,29 longer than in the analogous AsPh3 [2.427(1) A],30 SbPh3 [2.421(4) A],30 and H20 [2.3855(5)A] 31 complexes, the implication is (in view of the expected reciprocal relation- ship between bond length and bond stretching frequency)32 that v(RhRh) in these complexes must all lie above 289.3 cm-', as indeed is found to be the case.33-35 It is surprising at first sight that v(RhRh) has as high a value as it does in view of the fact that the Rh-Rh bond (between the two d7 ions) is only single. The high value is thought to arise from the significant contribution to the RhRh restoring force brought about by the four chelating acetate groups (primarily via the four OCO bending coordinates, which are coupled only in second order to the RhRh stretching coordinate).The results obtained provided a firm basis for making vibrational band assignments for other dirhodium tetracarboxylates, viz. Rh2(02CR)4(PPh3)2, R = H, CH3, C2H5, or C3H7,36 and for Rh2(02CCH3)4- L2, L = H2037 or S(CHZP~)~.~~ 29G.G.Christoph, J. Halpern, G. P. Khare, Y. B. Koh, and C. Romanowski, Inorg. Chem., 1981, 20, 3029. 'OR. J. H. Clark, A. J. Hempleman, H. M. Dawes, M. B. Hursthouse, and C. D. Flint, J. Chem. Soc., Dalton Trans., 1985, 1775. '' F. A. Cotton, B. G. DeBoer, M. D. LaPrade, J. R. Pipal, and D. A. Ucko, Acta Crystallogr., Sect. B, 1971,27,1664. 32 R. J. H. Clark and C. S. Williams, Inorg. Chem., 1965,4,350. 33 R. J. H. Clark, A. J. Hempleman, and C. D. Flint, J. Am. Chem. Soc., 1986,108,518. 34 R. J. H. Clark and A. J. Hempleman, Inorg. Chem., 1988,27,2225. 35 R. J. H. Clark and A. J. Hempleman, Inorg. Chem., 1989,28,746. 36 R. J.H. Clark and A. J. Hempleman, Inorg. Chem., 1989,28,92.37 R. J. H. Clark and A. J. Hempleman, Croat. Chem. Acta, 1988,61,313. R. J. H. Clark and A. J. Hempleman, J. Mof.Struct., 1989,197, 105. Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc For certain of these complexes, eg Rh2(02CCH3)4(MPh3)2 M = P, As, or Sb, the geometric changes attendant upon excitation from the ground (IA1,) to the first excited (‘A2,) state have been calculated 39 using the time-dependent theory of Tannor and Heller 40 Yet further resonance Raman studies of complexes of the same general structure have included those on tetracetamidates of the type Rh2(CH3-CONH)d(MPh3)2 and Rh2(CF3CONH)d(MPh3)2, M = P, As, or Sb, the tetra- acetate to tetraacetamidate change for analogous complexes brings about an increase of ca 0 04 8, in the Rh-Rh bond length and a ca 15 cm-’ decrease in v(RhRh) 41 Similar studies of the diosmium(r~~) complexes Os2(02CCH3)4C12, 0s2(02- CCD3)4C12, 0~2(02CCH~Cl)~C12, Osz(02CC2H5)4C12, and Os2(02CC3H7)4C12 have led to the identification of bands at 236228,393-256, and 31 1-292 cm-’ to v(OsOs), v(OsO), and v(OsCl), respectively 42 43 These studies are illustrative of the ways in which v(MM) in metal-metal bonded dimers has been identified conclusively Indeed, there is now a sufficiently firm basis for the assignments of v(MM) that corresponding assignments for bigger, especially linear-chain, complexes can now be made with some confidence 3 One-Dimensional Linear-Chain Complexes of Platinum, Palladium, and Nickel of the Wolffram’s Red Type A.Survey of the Structural Types, Physical Properties, and Spectroscopy.- Potassium tetracyanoplatinate, KzPt(CN)4Bro 30 3H20 (KCP) and its analogues are, on account of their being one-dimensional conductors, the best studied linear chain complexes of platinum 44 However, several other kinds of linear-chain complex have been the subject of much recent attention In particular halogen- bridged linear-chain complexes of the platinum group, despite their having much lower conductivities than KCP, have proved to be of immense interest Two types of such chain complex will be discussed, uzz the Wolffram’s red (WR) type, where WR = [Pt”(C2H5NH2)4][Pt’V(C2H5NH2)4C12]C144H20, and the type formed from the barrel-shaped complex ion [Pt2(pop)4I4-, where pop = diphosphite = H2P~05~-Complexes of the WR sort [PtL4][PtL4X2]Y4, have chain structures as indicated below LL LL LL LL..3~*..*c1-pt 4V -C~...~$~....CI-..L c, ... 4\ 41 L/\ L/\LL LL L L where L, a neutral equatorial ligand, may be an amine such as NH3, CH3NH2, 39 K -S Shin, R J H Clark, and J I Zink J Am Chem SOC 1989, 111,4244 40 D Tannor and E Heller, J Phys Chem ,1982,77 202 41 S P Best, P Chandley, R J H Clark S McCarthy, M B Hursthouse, and P A Bates J Chem SOC,Dalton Trans, 1989,581 42 R J H Clark, A J Hempleman and D A Tocher, J Am Chem SOC 1988 110 5968 43 R J H Clark and A J Hempleman, J Chem SOC ,Dalton Trans 1988 2601 Clark or C2H5NH2 and X = C1, Br, or I. Bidentate ligands LL can also form complexes of this general structural type, where LL may be 1,2-diaminoethane (en), 1,2-diaminopropane (pn), 1,3-diaminopropane (tn), etc.Moreover, both terdentate ligands, LLL, such as diethylenetriamine (dien) and N-methyldiethyle- netriamine (Medien), and quadridentate ligands, LLLL, such as 1,4,8,1 l-tetra- azacyclotetradecane (cyclam) will also form complexes of this general type.45 WR consists of Pt" and PtIV units, each of which bears a +2 charge, and hence four anions are required for electroneutrality. In addition to this type of complex, however, chain complexes of the same basic structure can also be formed in which the equatorial amine ligands are replaced by one, two, three, or four halide ions, with the result that the overall charge must be balanced by two, zero, two, or four counter ions, respectively, e.g.as in [Pt"(Medien)I][Pt'V(Medien)13]12, [Pt11(en)C12][Pt'v(en)C14], K2[Pt"(NH3)C13][Pt1V(NH3)C15].2H20, and K4[Pt1114][Pt1V16], respectively. The counter ions may be C1-, Br-, I-, C104, BFT, NO;, K', or Cs+ etc.;they play an important role, often along with water of crystallization, in helping to hold the chains together by hydrogen bonding. A summary of the different types of M"/MtVlinear-chain-complexes of the +2, +2 type now known is given in Figure 6. The intense colours of linear-chain complexes, which differ markedly from those of their constituent platinum(r1) and (IV) entities, are caused by intervalence (IVCT) transitions Pt" -PtIV, which are polarized in the chain direction; they occur in the regions 25 000-18 200 cm-' (X = Cl), 23 600-14 300 cm-' (X = Br), and 20 600-7 500 cm-' (X = I).Extensive research has established that the shorter the Pt". * PttV chain distance, the lower in wavenumber is the intervalence transition. This is consistent with the idea that shortening of the chain repeat unit is primarily brought about by a shortening of the Pt" . -.X rather than the Pt"-X distance, resulting in more nearly central bridging by the chain halide. Central halide bridging is known also to be critical in bringing about increased chain conductivity. In addition to IVCT transitions, the complexes display luminescence with a large Stokes shift.46 The Raman spectra of halogen-bridged mixed-valence complexes of platinum at resonance with the Pt" -Pt" intervalence band are very intense.In particular, the band attributed to the symmetric X-PtIV-X chain stretching mode (vl), together with those attributed to long overtone progressions vlvl, are tremendously enhanced at resonance. The overtone progressions reach as far as 17vl in some cases (Figure 7), implying a very substantial (0.054.10 I$) change in Pt'"-Cl bond length on changing from the ground to the intervalence 1.12.47 this structural change constitutes, in effect, a reverse Peierls 44 M. B. Robin and P. Day, in 'Advances in Inorganic and Radiochemistry', ed. H. J. Emeleus and A. G. Sharpe, Academic Press, New York, Vol. 10, 1967, p. 247. 45 R. J.H. Clark, Chem. SOC.Rev., 1984, 13, 219. R. J. H. Clark, in 'Advances in Infrared and Raman Spectroscopy', ed. R. J. H. Clark and R. E. Hester, Wiley/Heyden, Chichester, Vol. 11, 1984, p. 95; R. I. Buckley and R. J. H. Clark, Coord. Chem. Rev., 1985,65,167. 46 H. Tanino and K. Kobayashi, J. Phys. SOC.Japan, 1983,52, 1446. 47 R. J. H. Clark, M. Kurmoo, H. J. Keller, B. Kepler, and U. Traeger, J. Chem. SOC.,Dalton Trans., 1980,2498. 117 Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc. ' L = NH,, CH,NH,, C,H,NH, LL = NnN N N hNN F N N N en tn pn btn chxn LLLL = N-ttN-HN-KN 2, 2, 2 -tet N-HN-CHN-H.N 2,3, 2 -tet n W (: :I f\ U r::I N+ N-H N +It N 3, 2, 3-tet N -tfCNSft N+ 3, 3, 3 -t et N c:: 11n'1L)aneN4 (13)ane~, W (15)ane N, Y = CIO,, BF4, PF,, X 1 I 1 3000 2000 1000 0 Wavenumber /cm-' Figure 7 Resonance Raman spectrum of [Pt(pn)2][Pt(pn)2Brz][Cu3Br5]2, pn = 1,2-pro-panediamine, in a KBr disc at ca.80 K with 568.2 nm excitation 47 Clark transition. The relationship between Peierls energy gaps and lattice distortions has been discussed by Whangbo and Fo~hee.~~ The v1 values (for vo at resonance with the intervalence transition) range from 309.1-297.8 cm-' (X = Cl), 175.7-172.0 cm-' (X = Br), and 122.3-114.2 cm-' (X = I). Recent synthetic, spectroscopic, and structural work has led to the characterization of structurally analogous palladium complexes as well as mixed- metal derivatives of the sort [Ni"(en)2][Pt'v(en)2C12][C104]4, [Pd"(en)2]-[Pt'v(en)2C12] [C104]4, and analogous 1,2-diaminopropane comple~es.~~-~ ' Doubt existed for some years as to whether mixed-valence nickel(r1,rv) complexes of the same structural type Characterization of such complexes had been complicated by difficulties experienced in growing single crystals and also by crystallographic disorder.However, their characterization is now certain on the basis of careful synthetic, conductivity, electronic, and photoelectron spectroscopy, EXAFS, X-ray structural, magnetic, and other work, in particular that of Toriumi, Yamashita et al.J5-60Such complexes are black, the deep colour being attributed to the Ni" -Ni" intervalence transition centred on ca. 15 000 cm-' for a chloride-bridged species.The wavenumbers of the intervalence band maxima of complexes of given equatorial ligands vary in the order C1 > Br > I, and Pd"/Pt'v > Ni"/Pt'" > Pt"/Pt'v > Pd"/Pd'v > Ni"/Ni'V.45*57360The result implies that the valence electrons of linear-chain M".'v complexes are most delocalized for those of Ni"/NitV, an implication in agreement with the fact that the (pellet) electrical conductivity of linear-chain Nill/Ni'V complexes (o(300 K) -lW7 R-' cm-') is much greater than those of analogous linear-chain complexes of platinum or palladium with the same bridging halogen atoms. Nevertheless, the chain conductivity even of Ni"/Ni" complexes is low, a matter which is clearly linked to the fact that the bridging atom is not centrally placed.The conductivity of such complexes is well represented by the expression o(T) = ooexp( -AE/kT); thus for [Ni(en)z][- Ni(en)2C12]C14, the o(300 K) -6 x 1C8 R-' cm-' and AE (the thermal activation energy) -0.5 eV.56 Speculation as to whether or not a symmetrically bridged chain complex might be stable was recently resolved in the affirmative. 48 M.-H. Whangbo and M. J. Foshee, Inorg. Chem., 1981,20,113. 49 R. J. H. Clark and V. B. Croud, Inorg. Chem., 1984,24,588; 1986,25,1751. 'O R. J. H. Clark, V. B. Croud, and R. J. Wills, Inorg. Chem., 1988,27,2096.'' R. J. H. Clark, V. B. Croud, R. J. Wills, P. A. Bates, H. M. Dawes, and M. B. Hursthouse, Acta Cryst., Sect. B, 1989,45, 147. ''M. Yamashita, Y. Nonaka, S.Kida, Y. Hamaue, and R. Aoki, Inorg. Chim. Acfa, 1981,52,43. 53 G. C. Papavassiliou and D. Layek, Z. Naturforsch., Teil B, 1982,37, 1406. "D. A. Cooper, S. J. Higgins, and W. Levason, J. Chem. SOC., Dalton Trans., 1983,2131. 55 M. Yamashita and T. Ito, Inorg. Chim. Acta, 1984,87, L5; M. Yarnashita and I. Murase, Inorg. Chim. Acta, 1985,97, L43. 56 M. Yarnashita, I. Murase, I. Ikemoto, and T. Ito, Chem. Lett., 1985, 1133. "Y. Wada, T. Mitani, M. Yamashita, and T. Koda, J. Phys. SOC.Japan, 1985, 54, 3143; M. Yamashita, I. Murase, T. Ito, Y. Wada, T. Mitani, and I. Ikernoto, Bull. Chem. SOC.Japan, 1985,58,2336.''K. Toriumi, T. Kanao, Y. Urnetsu, A. Ohyoshi, M. Yarnashita, and T. Ito, J. Coord. Chem., 1988, 19, 209. 59 J. Evans, J. T. Gauntlett, and W.Levason, Inorg. Chem., 1988,27,4523and references therein. 60K. Toriumi, Y. Wada, T. Mitani, S. Bandow, M. Yamashita, and Y. Fujii, J. Am. Chem. SOC.,1989, 111, 2341; Y. Wada, T. Mitani, K. Toriurni, and M. Yamashita, J. Phys. SOC.Japan, 1989,58, 3013. Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chams, etc Figure 8 X-ray crystal structure of [Ni(R,R-chxn)2Br]Br2 showing the environment of the bromide ions The dashed lines indicate the orientation of the hydrogen bonds6' This matter and two others of great contemporary interest are now discussed in more detail B. Symmetrically Bridged Chain Complexes of Nickel.-Nickel complexes of the sort [Ni(R,R-~hxn)2Cl]Cl2 (red) and [Ni(R,R-chxn)2Br]BrZ (black) can be formed by oxidation of Ni(R,R-chxn)2X2 with C12/N2 or Br2/N2, respectively, in 2-methoxyethanol to form semi-conducting complexes in which the bridging group is centrally positioned 6o The low temperature (-151 "C) structure, with very small thermal ellipsoids, thus provides no evidence for the expected Peierls distortion of the linear chain Presumably, exceptionally effective hydrogen bonding prevents this distortion from taking place Indeed the complexes have in effect two- rather than one-dimensional character on account of strong hydrogen bonding not only in the chain direction but also in one of the two directions perpendicular to the chains (Figure 8) Central bridging has the effect of increasing the chain conductivity by ca lo3 Conductivity and other properties of these cyclohexanediamine complexes are Clark Table 1 Properties of symmetric chain nickel complexes Activation Complex o/R-' cm-' energylev Energy gaplev [Ni( R,R-chxn)2CI)]Cl~ 1~3 0.19 1.9 [Ni(R,R-~hxn)zBr]Br2 2 x 1W2 0.11 1.28 Table 2 Variants ofNickel(n1) Complexes Discrete Ni"' complex: [Ni([14]-aneN4)C12][C104],[14]-aneN4 = cyclam (or 2 x en) Ni" complex coordinated to cation radicals: [NiClz(diars)2]Cl, diars = a-phenylene bis-dimeth ylarsine Ni"/Ni'" 1D-chain complex: [Ni(en)2][Ni(en)2C12][C104]4,en = 1,2-diaminoethane Symmetrical 1D-chain Ni"' complex: [Ni(R,R-chxn)tBr]Br~, chxn = cyclohexanediamine = [Ni(Hd~g)~]Partially oxidized Ni" 1D-complex: Ni(Hd~g)~1 515r a NiZs2 complex involving I; given in Table 1.Despite the fact that much consideration has been given to the matter, the mechanism for the conductivity of all such chain complexes is not yet clear. Raman spectroscopy of the bromo complex shows no clear evidence of a v1 band (or of its overtone bands), consistent with the behaviour expected for a centrally bridged halide.61 The bromo complex is nearly diamagnetic, which implies that there is a strong antiferromagnetic coupling of the unpaired electrons (s = +) in the d,z orbitals located on each nickel atom. It is worth noting that there are many different sorts of apparent nickel(iii) complexes, in only some of which the metal is genuinely in this oxidation state (Table 2). The most obvious are the trans-octahedral discrete species [Ni([ 14)- aneN4)Cl2][C1O4] and [Ni(en)2C12][C104].Others apparently of this sort, e.g. [Ni(diars)2Clz]Cl, are thought likely to contain nickel@) coordinated to a cation radical derived from o-phenylenebisdimethylarsine; however, this conclusion seems uncertain in the case of the analogous diamine.52*59 The nickel(r1,rv) complexes are, of course, variants on the nickel(m) formulation, which is only strictly realized for the two complexes for which it is established that the bridging atom is centrally positioned. The final variant, Ni(Hdpg)zI, where H2dpg = diphenylglyoxime, is not a complex of nickel(ii1) but one of Ni2'2 since the counterion is not I- but C. Copper(rr)/Platinum(rv) Linear-Chain Complexes.-It is interesting to investig- ate the possibility of inserting other metals of the first transition series into the site of the bivalent metal in linear-chain complexes.In this context the K. Toriumi, H. Okamoto, T. Mitani, S. Bandow, M. Yamashita, Y. Fujii, R. J. H. Clark, D. J. Michael, A. J. Edward, D. Watkin, M. Kurmoo, and P. Day, Mol. Crysf.Liq. Crysf.,1990, in press. 62 M. Cowie, A. Gleizes, G. W. Grynkewich, D. W. Kalina, M. S. McClure, R. P. Scaringe, R. C. Teitelbaum, S. L. Ruby, J. A. Ibers, C. R. Kannewurf, and T. J. Marks, J. Am. Chern. SOC.,1979, 101, 2921. Synthesis Structure and Spectroscopy of Metal-Metal Dimers Linear Chains etc following reaction between copper(I1) and platinum(1v) complexes has been reported 63 [Cu(en)2]Cl~ 2H20 + [Pt(en)2C12]Cl2 + 4Na[C104] [C~(en),][Pt(en)zC12][ClO~]~ red needles The complex is isomorphous with [M(en)2][M(en)2Clz][ClO& M = Pd or Pt, X = C1 or Br, and thus the bridging chlorine atom (as is usually the case for such complexes) is positionally disordered at two sites equidistant from the midpoint between the two metal atom sites and with an occupancy of 0 5 The Cu" and Pt" ions are also disordered at the metal ion sites and thus cannot be distinguished from one another Nevertheless, the structure has been established and found to have r(Cu"-Cl) = 3 081 A and r(PtIv-C1) = 2 313 A Difficulties arising from partial substitution of Cu" by Pt" during preparation 63-leading to a reported Pt Cu ratio of 1 11 0 89-can be overcome by rapid rather than slow growth of crystrlls 64 The magnetic susceptibility (C = 0 415 emu mol-' K) of the complex obeys the Curie-Weiss law (0 = -1 2 K) and the ESR spectrum (gl = 2 048, gll = 2 167, 11 relating to the chain direction) implies that there is no significant magnetic interaction between the unpaired electrons occupying the dX2-,,2 orbitals on the copper (d9) ions which are -10 8 A apart along the chains 63 This situation is thus in contrast with that for Ni"' chain complex (vide supra) in which the unpaired electrons occupy d,z orbitals and interact so strongly as to render the complex antiferromagnetic, almost diamagnetic Contrary to the earlier report,63 the lack of dichroism and resonance Raman progressions involving the v(Cl-Pt'"-Cl) mode using visible excitation suggests that the intervalence transition in this complex is in the ultra-violet, the red colour being caused by local transitions on the copper(I1) moieties 64 Several attempts have been made to introduce other heavy metals into the chains but to date the only success in this respect is with gold, the complex [AuxPt, x(en)21]S04 3H20, x = 003, having been prepared 65 D.Structure to the v1 Band of Linear-Chain Complexes.-One of the more surprising discoveries made in 1983 was that the wavenumber of v1, vSy,(XPtX), in linear-chain complexes appears, under conditions of only moderate resolution, to increase as that of the exciting line (vo) increases 66 This apparent dispersion in v1 is very evident when seen against the backdrop of bands attributed to amine modes, which show virtually no dispersion For any particular set of complexes, the apparent shift in v1 increases in the order C1 < Br c I, and is largest for those in which the bond length ratio r(Pt'v-X)/r(Ptll-X) is nearest to 63 H Oshio K Toriumi S Bandow K Miyagawa and S Kurita J Chem SOC Dalton Trans 1990 1013 64R J H Clark D J Michael andM Yamashita to be published 6s N Matsushita N Kojima T Ban and I Tsujikawa J Phys SOCJapan 1987 56 3808 66 R J H Clark and M Kurmoo J Chem SOC Farada.v Trans 2 1983 79 519 122 Clark unity (usually iodine-chain complexes) and in which the Pt" .-* Pt" distance is least (brought about by strong hydrogen-bonding between the amines of different chains and the counter ion).The magnitude of the apparent dispersion to v1 can be as much as 20 cm-'. Careful, high resolution studies of these complexes as single crystals at low temperatures and by use of low laser powers have now shown that the dispersion in v1 is only apparent; it is caused by the fact that the v1 band consists of many, closely spaced, components, those of higher wavenumber being resonance enhanced with excitation lines of higher wavenumber. If wider slits or more power is used, only the overall contour of the v1 band is detected, which accordingly disperses as the relative intensities of its components ~hange.~~-~' The origin of the components to v1 is still a matter for debate. At the moment, the most likely explanation is that each component arises from a chain segment of different correlation length (defined as the length over which the oxidation states alternate 11 and IV regularly and without defects of any sort).Each segment has slightly different wavenumbers for vl, its IVCT band maximum, and its excitation profile maximum, from the corresponding values for other segments. Since shorter segments in conjugated polyenes are known to give rise to higher v1 values and to higher values for the absorption band maxima, it is presumed that these same features will hold true for linear-chain metal complexes. This explanation of the apparent dispersion in v1 thus requires postulation of the existence of a certain small distribution of discrete but relatively short correlation lengths within each chain; each segment (or small group of segments) can thus be made, in turn, to come into resonance as vo is changed, leading (where the resolution is moderate) to the apparent dispersion of the overall v1 band contour with change of vo.Somewhat analogous behaviour has also been observed for certain bands (e.g. v(C=C) and v(C-C) at -1450 and 1060 cm-', respectively) of trans-poly-acetylene, (CH), and (CD),.71-75 These bands likewise exhibit an apparent dependence of their wavenumber on vo, an observation which has been attributed in this case also to the existence of segments within each chain of different correlation lengths. Such segments are likewise expected to possess different wavenumbers for v1, the absorption band maxima, and excitation profile maxima.Change of vo will selectively enhance different components to v1 leading, if the components are not resolved, to dispersion in the overall v1 band contour with change of VO. Why segments of only certain lengths should be 67 R. J. H. Clark and V. B. Croud, J. Phys. C: Solid State Phys., 1986,19,3467. M. Tanaka and S. Kurita, J. Phys. C: Solid State Phys., 1986,19,3019. 69 S. D. Conradson, R. F. Dallinger, B. I. Swanson, R. J. H. Clark, and V. B. Croud, Chem. Phys. Lett., 1987,135,463. 70 R. J. H. Clark and D. J. Michael, J. Mof.Struct., 1988, 189, 173. 71 H. Kumany and P. Kroll, J. Raman Spectrosc., 1986, 17,89. 72 Z. Vardeny, J. Ornstein, and G. L. Baker, Phys. Rev. Lett., 1983,50,2032.73 Z. Vardeny, E. Ehrenfreund, 0.Brafman, and B. Horovitz, Phys. Reu. Lett., 1983,51,2326. 74 R. Tiziani, G.P. Brivio, and E. Mulazzi, Phys. Rev. B, 1985,31, 3019. 75 A. W. Tarr and W. Siebrand, J. Raman Spectrosc., 1989,20,209. 123 Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc formed, however, is far from clear, a consideration which undermines confidence in the proposed explanation for the dispersion (both for (CH), as well as for Wolffram's red salts) The better understanding of the nature and properties of the segments giving rise to the discrete components to vl-thought to be linked closely to the properties of solitons and polarons 76-79-is important to the understanding of the structures, spectral properties, and conductivity mechanisms displayed by both Wolffram's red type complexes as well as trans-polyacetylene Indeed, there are many similarities between these two, otherwise disparate, systems, hinging around the fact that the former has a centre of inversion at each metal atom while the latter has one at the middle of each double and single bond 8o Since in each case there is two-fold degeneracy the possibility that each might support the existence of solitons (domain walls) seems likely Thus, the valence bond struc- tures X-PtIV-X Pt" X-PtIV-X Pt" and Pt" X-PP-X Pt" X-PtIV-X are equivalent in an infinite chain, translating from one to the other by a concerted movement of the X-atoms away from Pt" towards Pt" with an interchange of oxidation states Likewise -CH=CH-CH=CH-and =CH-CH=CH-CH= are equivalent and may be interconverted by a concerted lengthening and shortening of the C-C bonds accompanied by a transfer of electron density between the bonds The soliton and polaron analogies between trans-poly- acetylene and Wolffram's red salts are illustrated in Figure 9 Apart from defects occurring as intrinsic features in crystals of Wolffram's red salts, as grown, defects may also be generated uza impurity doping, high pres- sures,8182 or photoexcitation Increased concentrations of defects lead to much enhanced chain conductivities Thus, iodine doping of [Pt(en)z][Pt(en>212]- [C1O4I4 leads to an increase in the chain conductivity of this complex of over lo7 The absorbance of two mid-gap bands (at ca 165 and 198 eV)83 and the ESR intensity in dilute doped crystals increase with dopant concentration Thus low levels of halogen doping have the same effects on optical absorption l6 N Kuroda, M Sakai, and Y Nishina, Phys Rev Lett, 1987,58,2122 77 S Kurita, M Haruki, and K Miyagawa, J Phys Soc Japan, 1988,57,1789 78 Y Onadera, J Phys SOLJapan, 1987,56,250 79 J T Gammel, R J Donohue, A R Bishop, and B I Swanson, to be published 8o P Day, Discussion comment, Phil Trans Roy SOL Lond, 1985, A314, 144 P Day, in 'Organic and Inorganic Low-Dimensional Crystalline Materials', ed P Delhaes and M Drillon, NATO AS1 Series B Physics, 1987, vol 168, p 33 See also R J H Clark and V B Croud, idem, p 341 B I Swanson, M A Stroud S D Conradson, and M H Zietlow Solid State Comm ,1988,65, 1405 82 M A Stroud, H G Drickamer, M H Zietlow, H B Gray, and B I Swanson, J Am Chem Soc, 1989,111,66 83 S Kuritd M Haruki, and K Miyagawa, J Phys Soc Japan, 1988,57,1789 \* \/ Clark 0 charged localised\fi/ / / soliton --.---~----.charged partly delocalised soliton----0 ------------__--(actual bo!d.*-.-*'* A alternation -0.06 A) / / / neutral soliton Pt4' Pt2' Pt4+ Pt" Pt" Pt*+ Pt4+ Pt3+ Pt4+ Pt3+ defect (polaron) P P in Wolffram's red Pt4+ Pt4' PtZ+ Pt4+ Pt2+ PtZ' Pt4' PtZ+ Stacking fault S' S' (soliton) in Wolffram's red Figure 9 Soliton andpolaron analogies between trans-(CH), and Wolffram's red spectra and ESR spectra as does photoexcitation using radiation with its electric vector parallel to the chain axis at a wavenumber greater than or equal to that of the IVCT transition.The origin of the defects is thought to be the same in each case. Doping-induced states are located on single chains and have both charge and spin, suggesting that they are polar on^.'^ Photoinduced defects occur as a result of separation of the electron-hole pairs created by the IVCT transition. Extensive theoretical discussions of the potential energy surfaces of Wolffram's red type complexes have been given by Nasu and T~yazawa~~,*~ and more recently, by Prassides, Day et al." Both the resonance Raman band profiles as well as those of the intense emission at about half the bandgap have been successfully modelled.The reader is referred elsewhere for a discussion of this treatment.81 Many other studies of these intriguing complexes are currently under study, particularly related to the full understanding of their luminescence and its dependence on pressure.88 4 Diphosphite Complexes The synthesis of two barrel-shaped ions [Pt2(H2P205)4I4-and [Pt2(H2P205)4C12]4-,89.90 the properties of which (for X = C1) are given in Figure 10, posed the intriguing question as to whether or not the ions could be induced to co-crystallize to form a stacked chain complex directly analogous to Wolffram's red. Here the ligand bridging the platinum atoms is the diphosphite 84 M. Haruki and S. Kurita, J.Phys. Rev. B.,1989,39,5706. 85 K. Nasu, and Y.Toyazawa, J. Phys. SOL..Japan,1982,51,2098. 86 K.Nasu, J. Phys. SOL..Japan,1983,52, 3865; ibid., 1984,53, 427; ibid., 1985,54, 1933. 87 K. Prassides, P. N. Schatz, K. Y. Wong, and P. Day, J. Phys. Chem., 1986,90,5588. 88 H. Tanino, N. Koshizuka, K. Kobayashi, M. Yamashita, and K. Hoh, J. Phys. SOC.Japan, 1985, 54, 483. 89 R. P. Sperline, M. K. Dickson, and D. M. Roundhill, J.Chem. SOL..,Chem. Cornmun., 1977,62. 90 M. A. Filomena Das Remedios Pinto, P. J. Sadler, S. Neidle, M. R. Sanderson, and A. Subbiah, J. Chem. SOC.,Chem. Cornmun., 1980,13. Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc. 4-r 7 4-L J r (PtPtl = 2 925 A r(PtPt1= 2 695 A 3 (Pt Pt I = 115 cm-' 3( Pt Pt) = 158 cm-' 452 nm (c=110 1 'AI~-~A~~ 345nm (E = 8200) 'A'~-'E, ,'A~, du*-P, d,* , du -dd,r 367nm (c=34,500)'AIg-'A2, 282 nm (E =48,8001 'Alg-lAlu d, -d,r J (PtCI = 305 cm-' Figure 10 Spectroscopic and structural properties of the complex ions [Pt2(H2P205)4I4-and [Pt2(H2P205)4C12I4-HO HO \ POP = Figure 11 Diphosphite ligands x-PtGLXn Figure 12 Schematic representation of the chain structure of K~[P~z(HzP~O~)~C~].~H~O95 H2P205'-(pop, Figure 11).This synthesis has, indeed, proved to be po~sible,~'-~~ as indicated schematically in Figure 12 (with pop as the ligand). A more detailed view of one repeat unit along the chain is given in Figure 13, which also shows the two equivalent positions for the bridging chlorine atoms along the chain. The relationship between the chains, which are an (AABCCB), structure, is shown in Figure 14.'' Such complexes have chain conductivities some six orders of magnitude greater than that of Wolffram's red salts with the same bridging halogen atom.The conductivity of K4[Pt2(H2P205)4C1].3H20 is thermally activated, with a very low activation energy of -0.08 eV. The complexes crystallize as golden metallic needles in which the colours are largely determined by IVCT transitions in the 10 000-20 000cm-' region. The Raman spectrum of K4[Pt2(H2PzO5)4Cl].3H20 at resonance with the 91 C -M Che, W P Schaefer, H B Gray, M K Dickson, P Stein, and D M Roundhill, J Am Chem Soc, 1982,104,4253 92 R J H Clark and M Kurmoo, J Chem Soc ,Dalton Trans, 1985,579 "M Kurmoo and R J H Clark, Inorg Chem ,1985,24,4420 94 R J H Clark, M Kurmoo, H M Dawes, and M B Hursthouse, Inorg Chem ,1986,25,409 95 L G Butler, M H Zietlow, C -M Che, W P Schaefer, S Sndhar, P J Grunthaner, B I Swanson, R J H Clark, and H B Gray, J Am Chem Soc, 1988,110,1155 Clark Figure 13 Model of the proposed chain repeat unit 94 in K4[Pt2(HzP205)4C1].3HzO 2 441A -2 969 A 2 605 A 7 _.2 299 A IC-J 93A -i 0510AfJ .-CI Figure 14 (AABCCB),chain structure o~K~[P~z(H~PzO~)~CI]3H20at 22 K intervalence band is dominated by a band at 291 cm-' (ho = 647.1 nm) which is assigned to the symmetric Pt-C1 stretching mode of the chain.93 Its value is a little lower, on account of bridging, than that found (305 cm-') for the discrete di- platinum(II1) species, [Pt2(H2P205)4Cl2l4-.Six-membered progressions in the 291 cm-' band are detected in the resonance Raman spectrum implying (a) that the chloride ion is not centrally bridging and (b) that the principal structural change oxidation state is-as for Wolffram's red complexes-substantial, and along the Pt-Cl coordinate. The close similarities between the PtPt and PtCl stretching frequencies of the isolated units and ofthe units as they occur in the chain suggest that the platinum ion valences approximate for this complex in the ground state to . . . II,II. . . . . III,III . (Table 3). 2:)undergone by the ion on excitation to the intervalence (24: Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers, Linear Chains, etc Table 3 Selected structural and vibrational data for diphosphite complexes platinum- platinum data v(PtPt)/ PtPt Bond PtPt Complex r(PtPt)/A cm ' Order Oxidation State K4[Pt2(H~P205)4] 2Hz0 2 925 115 0 232 K~[P~~(HzP~O~)~CI] 2 969 119 03H20" 22 (22 K) 2 685 152 1 333 K4[Pt2(HzP205)4Br] 3H20b 2 781 117 05-6 25 -6, 25-6 (19 K) 122 05+6 25 + 6, 25+6 K4[Ptz(HzPz05)4C12] 2Hz0 2 695 158 1 3,3 K4[Pt2(H2P205)4Br~]2H20 2 723 133 1 393 a (AABCCB).structure Symmetrically modified (AAB). structure The X-ray crystallography of the analogous chain bromide was originally interpreted to imply that the bromine atom is centrally placed along the chain This situation would be inconsistent with the expected Peierls distortion of a symmetric linear chain, and also with the observed Raman activity of v(PtX), though it would be consistent with relatively high chain conductivity However, neither the chain bromide nor the chain iodide yield progressions in v(PtX), as found for the chain chloride, at resonance with the intervalence band This implies that the bridging atom is much nearer to the central position in these cases than is the case for the chain chloride, instead a progression in v(PtPt) is observed Thus for the chain bromide or iodide, the platinum atoms are close to, but not exactly at, the +2 5 oxidation state The results are summarized in Table 3, in which the oxidation states of the platinum atoms in [Pt2(H2Pz05)4Br]4- are represented as 2 5 -6,2 5 -6,2 5 + 6,2 5 + 6, where 6 might be as low as 0 1 These conclusions are in accord with those drawn on the basis of low temperature X-ray diffraction studies These had, of necessity, to be carried out at 20 K in order for the thermal ellipsoids to be sufficiently small to demonstrate that the two equivalent positions of the bromine atoms were in fact 0 1 A from the central position 95 Much other work on further characterization of these complexes is in hand, particularly concerned with defining the nature of defect states in the chains and with investigating the consequences of applying pressures (4 GPa) to the complexes The latter has the effect of reversing the Peierls transition, ie the complex reverts towards the symmetric linear chain,96 as made evident by the red shift (750-1000 cm-' per GPa) in the IVCT transition (d:.Pt",Pt"-d:* Pt"',Pt"') (cf for Wolffram's red, the red shift of the IVCT transition is 1600 cm per GPa) Much effort has also been applied towards the synthesis of analogous linear- 96 B I Swanson M A Stroud, S D Conradson, and M H Zietlow, Sofid State Commun, 1988 65 1405 Clark chain complexes involving pcp rather than [pcp = HOP(O)CH2P(O)OH, Figure 11.1 However, complexes of this ligand have proved to be very difficult to isolate pure and in crystallographically characterizable form, although there is little doubt that chain complexes can be formed from the established [Pt2(pcp)4I4- and [Pt2(pcp)4C12I4- moieties.'* 5 Dithioacetate Complexes The diphosphite complexes described above are closely similar in both structures and properties to the dithioacetate complexes of platinum and nickel of the sort M2(CH3CS2)41, M = Pt or Ni These complexes are likewise semiconductors with essentially equal chain M-I bond distances (2.975 and 2.981 A for M = Pt, 2.928 and 2.940 A for M = Ni) and a Pt-Pt distance of 2.677 A99,'00(which is about 0.1 A shorter than in platinum metal).The complexes are thus close to being symmetrically bridged chains involving Pt2.5 alone. The 'intervalence' bands arising from interaction between the adjacent M2(CH3CS2)4 moieties occur in the 6000-8000 cm-' region with cmax-lo4 M-' cm-'. The conductivity of a compressed powder of the material at 300 K is 7 x Q-' cm-' ,a value which is diminished from the true value by interparticle contact resistance and by averaging over all orientations.The temperature dependence of the conductivity (77-300 K) is such that the activation energy is only 0.05 eV (-400 cm-I). The conductivity of a single crystal of PtZ(CHsCS2)41 at atmospheric pressure is 2 k'cm-', a value which increases to 10 Q-' cm-' at 7 GPa."' Evaporated thin films (-500 A) of this complex have, moreover, been prepared in YUCUO and found to possess a conductivity of 0.2 Q-' cm-', a possibly significant finding for the electronics in- dustry. Extensive infrared, Raman and resonance Raman studies of these and related sulphur-bonded complexes have recently been completed.' 027'03 The Raman spectra show strong resonance enhancement for excitation within the contours of the electronic bands in the visible region of the spectrum, but they have not yet been probed within the contour of the IVCT transition.97 R. J. H. Clark and C. M. Pout, unpublished work. C. King, R. A. Auerbach, F. R. Fronczek, and D. M. Roundhill, Inorg. Chem., 1986,108,5626. 99 C. Bellito, A. Flamino, L. Gastaldi, and L. Scarramuzza,Inorg. Chem., 1983,22,444. loo C. Bellito, G. Dessy, and V. Fares, Mol. Cryst. Liq. Cryst., 1985, 120, 381; Inorg. Chem., 1985, 24, 28 15. lo' I. Shirotani, Y. Inagaki, and M. Yamashita, 'Physics and Chemistry of Organic Superconductors', ISSP, 1989. R. J. H. Clark and J. R. Walton, Inorg. Chim. Acta, 1987, 129, 163. lo3 R. J. H. Clark and J. R. Walton, J.Chem. Soc.,Dalton Trans., 1987, 1535. Synthesis, Structure, and Spectroscopy of Metal-Metal Dimers Linear Chains etc Figure 15 Structure of the chain of (from the left) Reihlens green [Pt(C2H5NH2)4]-[Pt(CzH5NH2)4Br~]Br44H20, K4[Ptz(HzPz05)4Br] 3H20, and KCP,KZPt(CN), Bro 30 -3Hz0 6 Conclusion Both the degree of enhancement of the band assigned to vl, the symmetric X-Pt”’-X stretch, as well as the length of the overtone progressions in v1 observed at resonance with the intervalence band of linear-chain complexes are related to the extent of structural change attendant upon excitation to the intervalence state Spectacularly detailed band progressions are observed for these complexes, yielding an immense amount of valuable spectroscopic informa- tion Of particular interest is the comparison of Wolffram’s red type complexes with diphosphite complexes (Table 4) The latter have chain conductivities some lo6 times greater than the former, and this can be understood to be a consequence of the more nearly central bridging of their chain chlorine or bromine atoms In this context chain nickel complexes-especially those with central bridging halogen atoms-are potentially very important since they involve a cheaper metal and have comparatively high chain conductivities It is, however, worth concluding by putting the chain conductivities of halogen-bridged com- plexes into perspective (Figure 15) The conductivity of KCP (0= lo3 C2-l cm I) is lo6 times greater than that of K4[Pt2(H2P205)4Br] 3H20 which, in Clark Table 4 Analogy between WolfSram's red salts and chain pop complexes: platinum-halogen (PtX) data (X = C1 or Br, etn = ethylamine) 2.26 2.299'}0.142 10.8, 1:/.51 2'441'.'09} 0.142 3.13 2.951 vl, v,ym(PtCl)/cm-l 312.3 29 1 (to v1 = 16 at res.) (to u1 = 6 at res.) r(PtBr)/A 2.46} 3.12 o.66 2*58}2.78 0.20 vl, v,,,(PtBr)/cm-' 177.0 210 (to u1 = 11 at res.) 223 (to u1 = 1 only) '7' Assumes equivalent (a)or two different (b)PtPt bond lengths turn, is a further lo6 times greater than that of Reihlen's green (the bromo equivalent of Wolffram's red).All three complexes are bromides, but the first involves direct metal-metal bonds and short (2.89 A) metal-metal distances. Those of the bromine-bridged species are much longer than this (5.36 and 5.58 A, respectively) and this feature is clearly responsible for their much lower chain conductivities.In effect, the bridged complexes have easy access to lattice distortions which operate to generate band gaps and so prevent high con-ductivity. The principal concern of this work has been to attempt to understand the relationships between structure, spectroscopy, conductivity, and bonding in chain complexes likely to be of interest as new materials with desirable properties. Acknowledgements. The author is indebted to a number of co-workers for their contributions to this work, in particular Drs. M. Kurmoo, V. B. Croud, C. M. Pout, A. J. Hempleman, M. Stead, A. R. Khokhar, M. Yamashita, and Mr. D. J. Michael; he also thanks Johnson Matthey plc for the loan of chemicals, and the ULIRS, SERC, and NATO for support.

ISSN:0306-0012    

DOI:10.1039/CS9901900107

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Chem. SOC.Rev., 1990,19,133-145 Stereoelectronic Origins of the Intrinsic Barrier to SN2 Reactions By Ikchoon Lee DEPARTMENT OF CHEMISTRY, INHA UNIVERSITY, INCHON 402-751, KOREA 1 Introduction The concept of the intrinsic barrier, which can be determined by means of identity reaction barriers, is very useful in allowing a better understand- ing of factors influencing the reactivity and thermodynamic effects on rates. In this work, the origin of the intrinsic barrier and its significance in the transition state structures of sN2 processes are reviewed. The Marcus expression, equation 1, gives the kinetic barrier (AG*) in terms of intrinsic (AG8) and thermodynamic (AGO) barriers,' and provides a simple AGf = AG$ + AGO12 + (AG0)'/(16AG$) (1) picture of how kinetic and thermodynamic substituent effects combine to affect the overall barrier to a group-transfer reaction.Although the Marcus equation was originally derived for electron-transfer reactions in solution,'" it has been shown to apply to various processes including group transfers, such as methyl- transfer reactions (equation 2).le-' Studies of gas-phase &2 reactions have provided a direct means of determining experimentally the intrinsic barriers involved in the gas-phase methyl-transfer reactions.'e-f.2 Alternatively, several elegant studies by Lewis et aL3 involving solution-phase methyl-transfer reactions have provided kinetic data from which the intrinsic barriers can also be determined for solution-phase reactions.[The (a) R A Marcus, Annu Rev Phys Chem, 1964, 15, 155, (b) R A Marcus, J Phys Chem, 1968, 72, 891, (c) A 0 Cohen and R A Marcus, hid, 1968,72,4249,(d)W J Albery and M M Kreevoy, Adv Phys Org Chem ,1978,16,87, (e)M J Pellerite and J I Brauman, J Am Chem Soc , 1980,102,5993,cf) M J Pellerite and J I Brauman, [bid, 1983, 105, 2672, (g)E S Lewis and D D Hu, ibid, 1984, 106, 3992, (h) E S Lewis, J Phys Chem, 1986, 90, 3756, (I) J R Murdoch, J Am Chem Soc, 1983, 105,2660,b)E Grunwald, ibzd, 1985, 107, 125 (a) W N Olmstead and J I Brauman, J Am Chem SOC,1977, 99, 4219, (b) J A Dodd and J I Brauman, ibrd, 1984,106,5356, (c)C -C Han, J A Dodd, and J I Brauman, J Phys Chem ,1986,90, 47 1 (a) E S Lewis and S J Kukes, J Am Chem Soc, 1979, 101,417, (b)E S Lewis, T I Yousaf, and T A Douglas, ibid, 1987,109,2152, (c)T I Yousaf and E S Lewis, ibid, 1987,109,6137 133 Stereoelectronic Origins of the Intrinsic Barrier to SN~Reactions H 1 X-+ CH3X X-*CH,X X----C?--X-= XCH, *X-.XCH, + X-/\HH Reaction coordinate Figure 1 Double-well potential-energy surface for the gas-phase identity reaction x-+ CHjX eXCH3 + X-free-energy form of equation 1 is normally used for solution-phase reactions, but the application of the potential-energy form (1’) AE’ = AE$ i-AE0/2 -+ (AE0)’/16AE$ (1’) is more intuitive and appropriate for gas-phase and theoretical studies. Thus the free-energy (AGa) or the potential-energy form (AE8) will be used for the appropriate cases. For solution-phase reactions, the work terms, which are largely made up of solvent-solute interactions, should be included in equation 1.However, they are independent of structure throughout a series of related reactions and hence if the solvent is constant the work terms can be dropped from the Marcus equation as shown in equation 1.3 Theoretically the intrinsic barriers can be obtained by MO cal~ulation,~from the energy difference between the transition state (TS) and the ion-molecule cluster, i.e., the height of the central barrier in Figure 1. For an identity-exchange reaction, X = Y in equation 2, AEO = 0 so that the activation barrier becomes equal to the intrinsic barrier, AE* = AE8. Comparison of experimental and theoretical intrinsic barriers for some of the identity-exchange reactions in Table 1 shows that agreement between the two is only moderate.However, this could be attributed to the difficulties in determining experimental values and the inadequacy of the use of the 4-31G basis set for this type of calcula- tion.6 ‘(a) S Wolfe, D J Mitchell, and H B Schlegel, J Am Chem Soc, 1981, 103, 7692, 7694, (6) J D Mitchell, Ph D Thesis, Queen’s University, 1981, (c) D J Mitchell, H B Schlegel, S S Shaik, and S Wolfe, Can J Chem , 1985,63, 1642, (d)I Lee, C K Kim, and C H Song, Bull Korean Chem Soc , 1986,7,391 Errors in AE& are reported to be within a few kcal/rnol.lrF Keil and R Ahlrichs, J Am Chem Soc , 1976,98,4787 Lee Table 1 Comparison of experimental (gas-phase) and theoretical (4-31G level) values of AE& (kcal/mol) for X-+ CH3X XCH3 + X-AE& Expt.Theoret. HCC 41.3 50.4 CN 35.0 43.8 NH2 26.6 23.5 CHjO 24.2 15.6 F 26.2 11.7 c1 10.2 5.5 NC -28.5 CH3 -28.5 OH -21.2 OF -18.8 OOH -18.5 HI Various interpretations of the significance of the intrinsic barrier have been offered: (i) Brauman et al. correlated the gas-phase identity-exchange barriers, AE& with the methyl-cation affinity Em, of the nucleophile defined as the heterolytic bond dissociation energy of the CH3-X bond. CH3X---+ CH: + X-AHo = Emc(X-)= Do(CH3-X) -Eea(X) + E,(CH3) where Do is the bond dissociation energy, Ees(X) is the electron affinity of the X radical and Ei(CH3) is the methyl-radical ionization potential. They interpreted the correlation as a consequence of charge separation in the trigonal-bipyramidal penta-coordinate (TBP-SC) TS, (l), of the exchange reaction.(ii) Wolfe et a1.F correlated their AEgX values calculated using the 4-31G basis set with the deformation energies, AEd,r, of CH3X in the TBP-5C TS formation, especially with the C-X stretching energies AER. Although these rationalizations are reasonable, they seem unsatisfactory since they deal with deformation energies of CH3X only and all X's are included indiscriminately into a single correlation. Alternatively, the intrinsic barrier in the SN2 type processes at a carbon centre has been shown to have stereoelectronic origins. Stereoelectronic Origins of the Intrinsic Barrier to SN~Reactions Table 2 Empiricalparameters a,, and b,, from equation 3 over the Periodic Table H 1 2 3 4 H 118 1.46 a,, 1.74 182 1 96 1 1.85 2.06 2.15 2.28 2 2.36 2 45 2.62 3 2.57 2 71 4 2.87 blJ H 0.58 0.56 0.64 0.66 0.72 1 0.55 0.63 0.64 0.70 2 0.70 0.73 0.79 3 0 74 0.81 4 0.87 2 Steric Origin Before proceeding to elaborate on the origins of the intrinsic barriers involved in the identity-exchange reaction, equation 2 with X = Y, let us first consider the correlation between the bond length, r, and the force constant, F, for the sym- metric vibrational mode of the C-X bond in the TBP-SC TS, (1).It has been shown by Badger’ that the empirical correlation shown in equation 3 holds between bond lengths and force constants of diatomic molecules, where constants r = a,, -b,, log I; (3) a,, and b,, depend only on the rows (i and J) of the periodic table for the two atoms being bonded, i.e., one atom is in row i and the other in row j.The relationship (equation 3) has been extended to loose, Lennard-Jones, noble-gas two-atom clusters, demonstrating a striking continuity of bonding between the noble-gas clusters and the usual strong chemical bonds.’ The empirical parameters a,, and b,, in equation 3, including those for noble-gas clusters, are summarized in Table 2. The a,, values are seen to increase slowly and smoothly as the valence shell of the atoms expands gradually along with the increase in the number of the rows of the periodic table.The parameter a,, is regarded as a standard bond length for a related family of bonds between two atoms belonging to i-th andj-th rows at unit force constant, F = 1, i.e.,8d ’(a) R M Badger, J Chem Phys, 1933,2, 128, (b)R M Badger, Phys Rev, 1935,48,284, (c)J Waser and L Pauling, J Chem Phys, 1950,18,618, (d)R M Badger, J Chem SOC, 1934,3,710* (a) H S Johnston, ‘Gas Phase Reaction Rate Theory’, The Ronald Press, New York, 1966, (b)T L Cottrell, ‘The Strength of Chemical Bonds’, Butterworths, London, 1958, (c) D R Herschbach and V W Laurie, J Chem Phys, 1961,35,458, (d)H S Johnston, J Am Chem SOC,1964,86,1643 136 Lee Table 3 Average non-polar covalent radii, rij,over the Periodic Table 1 2 3 4 1 1.72 2.07 2.24 2.43 2 2.42 2.59 2.78 3 2.76 2.95 4 3.14 whereas the parameter bij, represents the degree of a bond length change in a tenfold increase of the stretching force constant: On the other hand, averaging the non-polar covalent bond radii (rj) of the representative elements within the i-th row, median (or average) nonpolar covalent bond radii, 7ij, between two atoms in the i-th andj-th rows are obtained by adding the two average covalent radii of elements, Yij = Fi + Yj, as presented in Table 3.9 The Yij values vary slowly and smoothly over the periodic table as was found for the aij value; in fact the Yij values are within +_ 10%of the empirical aij values.This was taken as an indication that the aij values reflect the median covalent bond radii for the family of bonds between one atom in i-th and the other inj-th row.Since equation 3 was originally derived empirically for diatomic molecules, it was of interest to see if one can arrive at the same correlation theoretically. The MO calculations at the 4-316 and 6-31G* levels" have indeed shown that the relationship (equation 3) holds for various diatomics which were used in the empirical correlation for 1-1, i.e., i = 1 andj = 1, series. The parameters all and bll obtained theoretically are 1.80 and 0.53 (4-31G)," and 1.80 and 0.51 (6- 31G*),1° which agree quite well with the empirical values of all = 1.85 and bl = 0.55, respectively. Moreover, it has been shown MO theoretically using the 4-316 basis set that Badgers rule, equation 3, also holds for the ion-molecule cluster, X----CHJX, as well as for the TBP-5C TS, (l)." Using ten anionic nucleophiles (X-) of the first row elements, a satisfactory linear correlation (r = 0.970) between r and log F (equation 3) for C-X bonds in (1) was obtained with all = 2.48 and bll = 1.10," which are much greater than the empirical parameters (all = 1.85 and bl1 = 0.55) for the 1-1 series of diatomic molecules.The results of higher level ab initio computations," however, afforded the following conclusions: R. T. Sanderson, 'Chemical Bonds and Bond Energy', Academic Press, New York, 1971. lo J. K. Cho, Ph.D. Thesis, Inha University, 1989. l1 I. Lee, J. K. Cho, and C. H. Song, J. Chem. Soc., Furaduy Trans. 2,1988,84,1177.Stereoelectronic Origins of the Intrinsic Barrier to SN~Reactions (i) For anionic TS, (l), incorporation of diffuse function" and correlation effect lowers both parameters, a,, and b,,. (ii) The one-heavy-atom nucleophiles (X-= F-, OH-, NH2-, and CH3-) provide excellent correlations between the bond length r and the force constant F at all levels of sophistication, and best mimic the diatomic series. (iii) The best values of all (2.19) and bll (0.69) obtained with the four one- heavy-atom nucleophiles at the MP2/6-31 + G*//6-31 + G* level of computation10 suggested that the carbon centre in the TBP-SC TS, (l), has a covalent bond radius corresponding to a 3rd row element; this should originate in the drastic extension of the C-X bonds which is required in order to accommodate five ligands in the rather small sized first row element C.Dewar et have pointed out that the drastic bond extension in forming the SN2 TS is mainly of a steric origin, i.e., the C-X bonds must be extended in the TS in order to accommodate five ligands sterzcally at the small central carbon atom.14 In the identity reactions (X = Y in equation 2) the energy needed to form a TBP-SC TS constitutes the intrinsic barrier, AE&, since the reactions are thermoneutral; the expansion of the covalent bond radius from that of the 1st row to that corresponding to the 3rd row requires energy for the reaction centre, carbon, so that the origin of this barrier lies in the expansion of the covalent bond radius.I5 The median covalent bond length of the central carbon in the TBP-5C TS corresponding to that of a 3rd row element (r13 = a13 = ro) therefore means that substantial median activation barrier, AE&, is required for the reaction series (2) with X = Y.Thus, the expansion of covalent bond radius to a median value of ro provides a median intrinsic barrier, AE&, for the reaction series, which originates really in the steric repulsion between the two (entering and leaving) groups (X) bonding at the relatively small-sized carbon atom in a 5-coordinated TS.14,15 This was also clear in a recent theoretical study of energy barrier involved in the SN2 reactions by Sand et a1.;l5"they found that the activation barrier originates primarily from the long bond length in the TS.There is also an interesting approach to the energy barrier which is also concerned with the bond stretching in the TS.ISb 3 Electronic Origin Since a median covalent bond length ro provides a median intrinsic barrier, AE&, for a family of closely related reactions, a small variation, Ar = rx -ro, from the median bond length ro for the individual nucleophile (or LG) X leads to a small variation in the intrinsic barrier, 6AE& = AEZX -AE$o. It has been l2 (a) T Clark, J Chandrasekhar, G W Spitznagel, and P v R Schleyer, J Comput Chem, 1983, 4, 294, (b) M J Frisch, J A Pople, and J S Binkley, J Chem Phys, 1984, 80, 3265, (c) Z Latajka and S Schemer, Chem Phys Lett, 1984,105,435 l3 (a) C Moller and M S Plesset, Phys Rev, 1934, 46, 618, (b) J S Binkley and J A Pople, Int J Quant Chem ,1975,9,229 l4 F Carrion and M J S Dewar, J Am Chem Soc, 1984,106,3531 l5 (a) P Sand, J Bergnman, and E Lindholm, J Phys Chem, 1988, 92, 2039, (b) S S Shaik, J Am Chem Soc ,1988,110, 1127 Lee ‘0 ‘x r: r > ro G rx ‘0 r (bl rx -c ‘0 Figure 2 Coupling of the two harmonic oscillators.Oscillators, B and C, cross oscillator A at rx and rj, giving barriers, AE&, between them shown that this variation within a series originates in the electronic state ofthe nucleophile X. A simple harmonic oscillator treatment of the variation in the intrinsic barrier within a reaction series (Figure 2) gave equation 4.16 6AE& = 1/2 F(rx -ro)’ (4) According to this equation the intrinsic barrier to the identity SN~reaction 2 with a particular X (= Y) can be represented by a point on a parabolic curve for a harmonic oscillator, A, in Figure 2. If the stretching, rx, in the TBP-SC TS, (l), is greater than ro, the barrier will be represented as a point on the right-hand branch of the curve; this intrinsic barrier originates in the stretching or expanding beyond ro (E region). Alternatively if rx < ro, then the barrier will be represented by a point on the left-hand branch of the curve; the intrinsic energy in this case corresponds to the compression energy (C region).However, near the l6 I. Lee, J. Chem. SOC.,Perkin Trans. 2,1989,943. Stereoelectronic Origins of the Intrinsic Barrier to sN2Reactions I rx < ro rx "0 region E reg I on--ri00 -I I rO 'X Figure 3 A parabolic curve representing three regions Two points on each region of the curve correspond to the two crossing points given in Figures 2(a) and 2(b) respectively bottom of the well, ie, rx ro (B region), the relation 4will be more accurate N since the harmonic oscillator model of the potential-energy function is better applicable near the bottom of the well In fact, application of equation 4to all members of X within a series tacitly assumes that the force constants Fx are nearly invariant within a particular reaction series i e , Thus, if for a series, rx values were greater than ro, the intrinsic barriers AEjfx, will be given by points in the E region of a curve similar to that in Figure 3 and will form an approximately straight line provided the range covered by 6AE& or Ar is sufficiently small, i e ,equation 5 will apply and have a positive slope, a l6 Similarly, if for a series rx < ro, the AE& values will be linearly related to Ar with a negative slope a in the C region When, however, rx values are not much different from ro, rx N ro, ie, in the B region, no such linearity will be expected but parabolic behaviour will be observed in accordance with equation 4 Therefore, for any series of identity-exchange reactions, the intrinsic barriers should belong to one of the regions, E, C, or B, provided the range covered by the series is sufficiently small Applications of the simple harmonic oscillator model of the intrinsic barrier to series of symmetrical methyl-transfer reactions, re X = Y in equation 2 have provided evidence in support of the electronic origin of the variations in barrier from a median value, AE2 l6 The identity sN2reactions 2(X = Y)with anionic nucleophiles (or nucleofuges) X-of the first-row elements, X = F, OH etc have been studied both experi- mentally2 and theoretically The ab initio MO theoretical results with the 4-31G basis set on rx, Fx, and AEjfx given in Table 4have shown that plots of AE& Lee Table 4 The distance, rx/A, force constants, Fx/mdyne A-', of the C-X bond in the TS, (l),and the intrinsic barrier, AE&lkcal mol-', calculated using 4-31G basis set for reaction: X -+ CHJXIXCH3 + X-(mdyne A-' = 10' N m-'; 1 kcal = 4.184 kJ) X- rX Fx AEZX CH3 2.161 1.998 - CCH 2.124 1.913 50.4 CN 2.112 2.417 43.8 NC 2.0 14 2.702 28.5 NH2 2.008 2.795 - OF 1.932 3.194 18.8 OCH3 1.924 3.232 23.5 OOH 1.920 3.281 18.5 OH 1.909 2.317 21.2 F 1.827 2.674 11.7 uersus rx are linear with a positive slope, a > 0 in equation 5, and correlation coefficient of 0.973.Thus the series belongs to the E region covering a relatively wide range, &AE& -40 kcal/mol.Analysis involving changes in rx and Fx indicated that the total contribution from the increase in YX to the increase in the intrinsic barrier, AEZX, is more than 3 times greater than that from the change (decrease) in Fx, and hence 6AEzx is relatively insensitive to the variation in Fx, supporting the invariance assumption of F used in equation 4.Moreover, the reactivity follows the order of leaving group (LG) ability; a better LG has a lower barrier and hence is more reactive. Thus the degree of C-X bond breaking is substantially greater than that of X-C bond formation so that the TBP-5C TS has a loose structure with a positive charge on the reaction centre carbon, as in (1). In addition, a worse LG has a higher LUMO, o&, and in order to lower the LUMO so that a smaller inter-frontier level gap (AEFMO)is obtained for a greater charge-transfer stabilization in the TS,17 a greater degree of bond breaking is required as the reaction progresses, which results in a greater deformation energy of the TS giving a net increase in the intrinsic barrier.A second application involved solution-phase methyl-transfer reactions be- tween substituted (X) benzenesulphonates, reaction 6.Ig,' The normalized Bronsted coefficients, PN = 0.37 and pL = -0.63,were used to assign +0.25 electronic charge unit (e.u.) at the reaction centre carbon and -0.63 e.u. on both sulphonate oxygens.' Here again a better LG with a more electron-withdrawing substituent gave a lower barrier, AG&, indicating that the reactivity trend is in the order of LG ability. Thus the TBP-5C TS for this reaction series with X "(a) N. D. Epiotis, W. R. Cherry, S. S. Shaik, R. L. Yates, and F. Bernardi, 'Structural Theory of Organic Chemistry', Top.Curr. Chem., 1977,70, Part 1; (b)A. Pros and S. S. Shaik, J. Am. Chem. Soc., 1981,103,3702; (c) K. Fukui, H. Fujimoto, and S. Yamabe, J. Phys. Chem., 1972,76,232. l8 This is true only when the normalized coefficients are used. See also: E. S. Lewis, J. Phys. Org. Chem., 1990,3, 1. Stereoelectronic Origins of the Intrinsic Barrier to SN~Reactions varying from p-CH30to 3,4-dichloro is loose and bond breaking greatly exceeds bond formation. Moreover the range of GAG;fx covered in this series was very small with only GAG& = 0.9 kcal/mol. Therefore, GAG&, should be linearly related to rx, by equation 5, with a positive slope, a > 0. On the other hand, plots of AGgx uersus Hammett substituent constants cr gave an excellent linear correlation (equation 7) with negative slope and correlation coefficient of 0.993. Comparison of equations 5 and 7 thus led to rx = ho + ro Ar = ho h<O where h is a negative constant.This relation shows that the C-X bond stretch beyond ro, Ar = rx -ro, in the TS is linearly related to o,and hence the variation of the intrinsic barrier, GAGZx, is related to the electron-donating or -withdrawing power of the substituent in the nucleophile. This clearly shows that 6AGgx is of an electronic origin. Solution-phase phenacyl-group transfer reactions, (equation 9),3c indicated that the reaction centre carbon is, in this case, negatively (-0.48 e.u.) charged in the TBP-5C TS, and the reactivity trend follows the order of nucleophilicity.This reactivity trend as well as the large negative charge on the reaction centre carbon in the TS is exactly opposite to those found for the examples belonging to the E region, and indicates that in this reaction series bond formation is ahead of bond breaking and the TS is relatively tight so that the bond length rx is shorter than ro, rx < ro. Thus the reaction series belongs to the C region, in which a linear correlation (equation 5) between AG& and Ar should hold with a negative slope, 01 < 0. The range covered by 6AGZx was small, ca. 1.10 kcal/mol. Alternatively, plots of AG$x versus cr gave a good linearities with a positive slope and a corre- lation coefficient of 0.988, equation 10. Comparison of equations 5 and 10 yielded again equation 8 which is exactly the same relation as that obtained for reactions belonging to the E region.Solution-phase methyl-transfer reactions between thiophenoxides, equation 1lY3O is reported to give negligibly small (negative) charge, (-0.08 e.u.) on the Lee Nucleophile I 'XY 'YZ I Leaving (NX) group (LZ) Subst rate (RY1 Scheme 1 o,and R,are substituent and reaction centre, respectively, where i = X, Y,or Z reaction centre carbon in the TS; this reaction shows no trend in the reactivity uersus nucleophilicity or LG ability as a whole so that there is no linearity be- tween 6AGgx and Ar. This series in fact belongs to the B region, for which the equa- tion 4 applies well. Moreover the G& values were not linearly related to 0but were a quadratic function of 0,equation 12, which suggested the same quadratic correlation between 6AGZx and Ar as equation 4.Comparison of equations 4 and 12 again led to equation 8. Thus the relation (equation 8) is a fundamental one for the thermoneutral reactions, AGO = 0 in equation 1, i.e., the intrinsic-controlled reaction series; ''>'' for a more electron-withdrawing substituent in the LG (6a~> 0), i.e., a better LG, a greater degree of bond formation (Arxy < 0), is obtained whereas for a more electron-donating substituent (60x < 0) in the nucleophile, i.e., a stronger nucleophile, a greater degree of bond breaking (Aryz > 0) is obtained, equation 13 (see Scheme 1). These predictions of the TS variation are shown to be consistent with those by the quantum-mechanical (QM) m0de1.'~~'~ It has been shown that for a thermodynamic-controlled reaction l9 series, i.e., 6AG" % 6AG& and hence 6AGf a 6AGo in equation 1, the sign of h in equation 8 reverses to positive, h > 0, which results in the relationships of equation 14.The TS variation can be predicted by the potential energy surface (PES) diagram,4e~'~*~~Figure 4,for this type of reaction series. These relations indicate that a more electron-withdrawing substituent in the LG (60~> 0) gives a greater bond distance TXY, i.e., less bond formation, Arxy > 0. This corresponds to a TS l9 (a) A. Pross and S. S. Shaik, J. Am. Chem. SOC.,1981, 103, 3702; (b) I. Lee and C. H. Song, Bull.Korean Chem. SOC.,1986,7, 186. 2o (a)I. Lee and H. S. Seo, Bull. Korean Chem. SOC.,1986,7,448; (6) I. Lee, C. S. Shim, S. Y. Chung, and H. W. Lee, J. Chem. SOC.,Perkin Trans. 2, 1988, 975; (c) I. Lee, Y. H. Choi, and H. W. Lee, J. Chem. SOC.,Perkin Trans. 2, 1988, 1537; (d) For this type of series, the reactivity-selectivity principle is known to apply in general: E. Buncel and H. Wilson, J. Chem. Educ., 1987, 64, 475 and references cited therein. Stereoelectronic Origins of the Intrinsic Barrier to &2 Reactions XN + R++LZ-XNR+ + LZ-bond breaking ( fyz increase; A‘yz ’0) XN + RLZ bond format ion XN+-R-LZ (rxydecrease; Arxy < 0) Figure 4 Potential energy surface diagram of an associative SN~reaction. X and Z representsubstituents in the nucleophile and leaving group, respectively shift from point 0 toward C as a result of Hammond (0-B) and anti- Hammond (or Thornton) shifts (0-A) in the PES diagram.A more electron-donating substituent in the nucleophile (60x < 0) gives a less bond breaking i.e., a smaller ryz (Aryz < 0),which corresponds to a TS shift to point E as a result of the sum of two vectors OB and OD. These TS variations of the PES model are shown to apply to the thermodynamically controlled reaction series. l9 Arxy = a’oZ Aryz = b’ ox a’,b‘ > 0 The important features can be recapitulated as follows. (i) The major component of the intrinsic barrier originates in the steric effect of the two, entering and leaving, groups; this component provides a median barrier height, AE&, at a median bond length, ro, in the TBP-5C TS, for a particular family of closely related reactions. (ii) The minor component of the intrinsic barrier originates in the electronic state of the reactants, nucleophile or leaving group and/or substrate. This component determines small variations in the intrinsic barrier, GAEgx from the median value, AE2, due to small variations in the bond length from the median length, Ar = rx -ro, within a reaction series.(iii) The variations, 6AE$x, can be grouped in three categories: expansion Lee (rx > rJ, compression (rx < ro), and borderline (rx z ro) regions of a harmonic oscillator potential energy curve, when 6AEgx covered is relatively small.(iv) There is a fundamental correlation between the bond length change Ar in the TS and the Hammett’s constant CJ of a substituent in the nucleophile or LG i.e., Ar = ho with a negative constant h for the thermoneutral reactions, the reactivities of which are controlled by the intrinsic barrier (intrinsic-controlled reaction series). 4 Concluding Remarks We have dealt here with applications to only the most simple and clear-cut cases of thermoneutral &2 reactions. However, in reality non-identity reactions with X # Y in equation 2 will be more general. Although the TS will not be of the symmetrical TBP-SC type, (l), classification of such reactions may still be possible into the intrinsic- and thermodynamic-controlled reaction series in many cases, so that the effects of substituents on the degree of bond formation and cleavage (equations 13 and 14) can be very useful in characterizing the TS.In this respect, the use of experimentally determinable cross-interaction constants 22 pxy and PYZ are helpful, since changes in the logarithms of the magnitude of such constants (Scheme 1) are similarly related to substituent constants with opposite slopes, where the constants A and B are now positive for the intrinsic-controlled series, whereas they are negative for the thermodynamic-controlled reaction series. Acknowledgements. This work was supported by the Ministry of Education and the Korea Science and Engineering Foundation. 21 (a) R. A. More O’Ferrall, J. Chem. SOC.,B, 1970,274; (b) W. P. Jencks, Chem. Rev., 1985,85,511. 22 (a) I. Lee and S. C. Sohn, J. Chem. SOC.,Chem. Commun., 1986, 1055; (b) I. Lee, H. Y. Kim, and H. K. Kang, J. Chem. SOC.,Chem. Commun., 1987, 1216; (c) I. Lee, H. K. Kang, and H. W. Lee, J. Am. Chem. SOC.,1987,109, 7472; (d) I. Lee, H. Y. Kim, H. K. Kang, and H. W. Lee, J. Org. Chem., 1988, 53,2678; (e)I. Lee, C. S. Shim, S. Y.Chung, H. Y. Kim, and H. W. Lee, J. Chem. SOC.,Perkin Trans. 2, 1988,1919; (f)I. Lee, H. Y. Kim, H. W. Lee, and 1. C. Kim, J. Phys. Org. Chem., 1989,2,35.

ISSN:0306-0012    

DOI:10.1039/CS9901900133

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Chem. SOC.Rev., 1990,19,147-195 Synthesis and Chemistry of Acyl Silanes By Philip C. Bulman Page, Sukhbinder S. Klair, and Stephen Rosenthal ROBERT ROBINSON LABORATORIES, DEPARTMENT OF CHEMISTRY, UNIVERSITY OF LIVERPOOL, P.O. BOX 147, LIVERPOOL L69 3BX 1 Introduction Acyl silanes exhibit unusual spectroscopic behaviour and possess interesting chemistry. In general they enjoy poor stability, particularly towards basic conditions and light. Early work, perhaps spurred by the green or yellow-green colour of simple acyl silanes, concentrated on spectroscopic properties, and indeed the structure and bonding of acyl silanes is a subject still worthy of discussion in the literature. Crystallographic studies have only added to the argument. More recently the chemistry of acyl silanes has been investigated resulting both in valuable new reactions and in improved methods of synthesis of acyl silanes of several types.’ This survey covers the major developments in all areas of acyl silane chemistry since the first isolation in 1968, beginning with a simple descriptive spectroscopy section; readers are referred to publications cited in the text for a more thorough treatment of the structure and bonding aspects of acyl silane chemistry.2 Structure and Spectroscopy of Acyl Silanes A. Infrared Spectroscopy.-Early work on acyl silanes 2-4 mainly concerned their interesting and unusual spectroscopic properties. For example, the alkyl acyl silane (1) and the aryl acyl silane (2) have carbonyl stretching absorption frequencies at ca.1645 cm-’ and 1620 cm-’ respectively in their infrared spectra (Table 1, entries 6 and 7). This lowering of the carbonyl stretching frequency in the infrared spectrum relative to simple ketones is usually explained as an inductive effect: electron A. Ricci and A. Degl’Innocenti, Synthesis, 1989,647. A. G. Brook, J. Am. Chem. SOC.,1957, 79, 4373; A. G. Brook, M. A. Quigley, G. J. D. Peddle, N. V. Schwartz, and C. M. Warner, J. Am. Chem. SOC.,1960,82,5102. V. Bazant, V. Chvalovsky, and J. Rathousky, in ‘Organosilicon Compounds’, Academic Press, New York, 1965; V. Chvalovsky, in ‘Handbook of Organosilicon Compounds-Advances Since 1961’, Marcel Dekker, New York, 1974. A. G. Brook and R. J. Mauris, J. Am. Chem. SOC.,1957,79,971; A.G. Brook and G. J. D. Peddle, Can. J. Chem., 1963,41,2351. Synthesis and Chemistry of Acyl Silanes release from the silicon atom towards the carbonyl group should favour polarization of the carbonyl group, and so resonance structure (3) should be expected to have a more significant contribution than in a simple ketone, thus weakening the bond.' 0 0--RASIR', R 0kSIR', The position of the carbonyl absorption of acyl silanes is not significantly altered by changing the type of group attached to the silicon atom, similar absorptions being observed for trimethyl, triphenyl, and tris-(para-substituted) phenylsilanes, the last showing a linear correlation of Hammett constants. Further, whereas ketone carbonyl absorptions shift to lower frequency on increasing the polarity of the medium in which they are measured, the position of acyl silane carbonyl absorptions is relatively independent of medium polarity.6 The large 0 inductive effect is caused by the low electronegativity of silicon (1.8) relative to carbon (2.5), and by the larger mass of silicon.Ground state mesomeric effects acting between silicon d orbitals and the carbonyl group [n (n-d) bonding] would be expected to give rise to absorption at higher frequency. However, differences in electronegativity producing significant inductive release of electron density towards the carbonyl group would be expected to lead to a decrease in the frequency, as is indeed the 0 Me,SI ASiMe, (4) Acyl silanes and acyl germanes exhibit lower frequency absorptions than a-silyl ketones, which in turn absorb at lower frequency than p-silyl ketones.p-Silyl ketones absorb at similar frequency to their carbon analogues (Table 1, entries 2-49. This pattern is consistent with a strong inductive release of electrons, an effect which is less apparent in the a-metallo ketones, where it must operate through an extra methylene group, and which has no apparent effect through two methylene groups. The strong inductive release in acyl silanes is approxi- mately additive as shown by the pink-coloured bis-(trimethylsilyl) ketone (4), where the infrared stretching frequency is shifted by approximately 80 cm-' to lower frequency compared with pivaloyl trimethylsilane (Table 1, entries 32 and 33)? 'A G Brook, Adu Organomel Chem ,1968,7,96 A G Brook, R Kiviskik, and G E Le Grow, Can J Chem ,1965,43,1175 Page, Klair, and Rosenthal Table 1 Infrared C--Oabsorption of selected acyl metalloids and related compounds Entry Compound v(M=C) v(M=Si) v(M=Ge) v(M=Sn) ReJ Ph3MCOCH3 1645 1669 1670 5 Ph3MCOPh 1692 1618 1629 1627 5,7 Ph3MCHzCOPh 1698 1667 1661 5 Ph3MCH2CHzCOPh 1692 1692 5 PhMezMCOPh 1620 597 Me3MCOPh 1675 1620 1629 597 Me3MCOCH3 1710 1645 5, 7 Bu'Me2MCOCH 1630,1640 8 Me3MCO(CH&CH3 1700 1640 7,9, 10 (Me3MCOCHzCH2)2 1640 11 Me3MCOCH2Ph 1635 1660 1656 12 Et3MCOCH2CH3 1640 13 Prn3MCOCH2CH3 1640 13 Bu'MezMCOCH2CH 1630,1640 8, 13 BU'M~~MCO(CH~)~CH~ 1630 8 Me3MCOCHCl(CHz)3CH3 1650 14 Me3MCOCHBr(CH&CH3 1645,1652 7, 14 Me3MCOCH=CH2 1635,1641 15,16 Me3MCOCH=CHCH3 1642 1644 1648 12 Me3MCOC(CH3)=CH2 1630 16 PhMe2MCOC(CH3)=CH2 1620 17 M~JMCOCH=C(CH~)~ 1640 1645 1640 12 Me3MCOCH=CH(CH2)3CH3 1620 18-2 1 ButMe2MCOCC1=C(CH3)2 1640 17 (2 1 a) 1640 11 (21b), (21c) 1610 11 (2W 1620 11 (21e),X = 0 1590 22 (21e),X = S 1580 22 (21e),X = NMe 1575 22 Bu'MezMCOC=CCH3 1605 15 Me3MCOCMe3 1636 5 Me3MCOMMe3 1556,1570 23 Ph3MCOMPh3 1558 + 1592 1616 5 Me3MCOCOCH3 1713 + 1658 7, 15 Me3MCOCOPh 1680 + 1610 7 Me3MCOCO2Me 1660 24 PhMezMCOCO2Me 1660 24 Me3MCONEt2 1620 1560 25 (Me3Si)3MCOzH 1630 26 (Me&) MCO ?Me 1655 26 (Me3Si)3MCOAd * 1620 1640 27,28 Me(Me3Si)zMCOAd * 1613 29 Bu1(Me3Si)~MC0Ad 1622 29* Ph(Me3Si)zMCOAd* 1609 29 * Ad adamantyl 149 Synthesis and Chemistry of Acyl Silanes B.13CNMR Spectroscopy.-In comparison with the corresponding ketones, the 13C signals of the carbonyl groups in acyl silanes are dramatically shifted downfield (Table 2).30-31"Carbonyl groups in acyl silanes have chemical shifts differing by between ca.25 and 100 ppm from those of the analogous ketones; '* the effect is approximately additive (Table 2, entry 16). 13C studies also expose some interesting resonance features of acyl silanes. The carbonyl group of an alkyl phenyl ketone (e.g. Ph-CO-But) displays a 13C chemical shift close to its aliphatic analogue (i.e. Me-CO-But); however, the difference between the two corresponding silicon species (e.g.Ph-CO-SiMe3 and CH3-CO-SiMe3) is a little more marked. Benzoyltrimethylsilane (2) exhibits a carbonyl shift ca. 11-14 ppm upfield of that displayed by acetyltrimethylsilane (1) (Table 2, entries 2 and 5). This could be ascribed to participation of resonance structures such as (5) which, by reducing the amount of positive charge on the carbonyl carbon atom, increase its shielding, thus displacing the chemical shift upfield relative to the aliphatic analogue where such an effect cannot operate. 'P C B Page and S Rosenthal, Tetrahedron, 1990, 46, 2573, P C B Page and S Rosenthal, unpublished data I Kuwajima and J Enda, J Am Chem SOC,1985,107,5495 G Zweifel and J A Miller, Synthesis, 1981, 288 lo C Heathcock and J Lampe, J Org Chem, 1983,48,4330 l1 A Capperucci, A Degl'Innocenti, C Faggi, A Ricci, P Dembech, and G Seconi, J Org Chem, 1988, 53,3612 l2 J A Soderquist and A Hassner, J Am Chem SOC,1980, 102, 1577, D I Gasking and G H Whitham, J Chem SOC,Perkin Trans 1, 1985,409 l3 D Schinzer, Synthesis, 1989, 179 l4 I Kuwajirna, T Sato, K Matsumoto, and T Abe, BuN Chem SOCJpn, 1984, 57, 2167, T Sato, T Abe, and I Kuwajima, Tetrahedron Lett, 1978,259 l5 H J Reich, M J Kelly, R E Olson, and R C Holtan, Tetrahedron, 1983,39,949, H J Reich and M J Kelly, J Am Chem SOC,1982,104,1119 R L Danheiser, D M Fink, K Okano, Y -M Tsai, and S W Szczepanski, J Org Chem, 1985, 50, 5393 l7 H J Reich, E K Eisenhart, R E Olson, and M J Kelly, J Am Chem SOC,1986, 108, 7791, H J Reich and E K Eisenhart, J Org Chem ,1984,49,5282 I8 R Mantione and Y Leroux, Tetrahedron Lett, 1971,12,591 l9 R G Visser, L Brandsma, and H J T Bos, Tetrahedron Lett, 1981,22,2827 2o K J H Kruithof and G W Klumpp, Tetrahedron Lett, 1982,23,3101 21 J C Clinet and G Linstrumelle, Tetrahedron Lett, 1980,21,3987 22 A Ricci, A Degl'Innocenti, S Chimichi, M Fiorenza, G Rossini, and H J Bestmann, J Org Chem , 1985,50,130 23 A Ricci, M Fiorenza, A Degl'Innocenti, G Seconi, P Dernbech, K Witzgall, and H J Bestmann, Angew Chem ,Int Ed Engl, 1985,24,1068 24 A Sekiguchi, Y Kabe, and W Ando, Tetrahedron Lett, 1979,20,871 25 G J D Peddle and R W Walsingham, J Chem SOC ,Chem Commun ,1969,462 26 A G Brook and L Yau, J Organomet Chem ,1984,271,9 27 A G Brook, J W Harris, J Lennon, and M El Sheikh, J Am Chem SOC, 1979, 101, 83, A G Brook, S C Nyburg, W F Reynolds, Y C Poon, Y -M Chang, J -S Lee, and J -P Picard, J Am Chem SOC ,1979, 101,6750, A G Brook, S C Nyburg, F Abdesaken, B Gutekunst, G Gutekunst, R K M R Kallury, Y C Poon, Y -M Chang, and W Wong-Ng, J Am Chem SOC,1982,104,5667 28 A G Brook, F Abdesaken, and H Sollradl, J Organomet Chem ,1986,299,9 29 K M Baines, A G Brook, R R Ford, P D Lickiss, A K Saxena, W J Chatterton, J F Sawyer, and B A Benharn, OrganometaIIics,1989,8,693 30 E M Dexheimer, G L Buell, and C le Croix, Spectroscopy Lett, 1978,11,751 31 (a) F Bernardi, L Lunazzi, A Ricci, G Seconi, and G Tonachini, Tetrahedron, 1986,42, 3607, (b)S Fliszar, G Cardinal, and M T Beraldin, J Am Chem SOC ,1982,104,5287 Page, Hair, and Rosenthal Table 2 3C Carbonyl group chemical sh fts of selected acyl silanes and related compounds Entry Compound G(M4i) 6(M=C) Re$ 1 Me3MCOCMe3 249.0 215.1 31a 2 Me3MCOCH3 244.3, 247.6 2 10.4 31a, 32 3 Me3MCO(CH2)5CH3 25 1.2 7 4 (1 12) 246.7 33 5 Me3MCOPh 233.6, 237.5 207.8,209.1 31a, 32 6 Me3MCOCHZH2 236.7, 237.9 15,16 7 Me3MCOC(CH3)=CH2 237.3 16 8 Me3MCOCH=CH(CH2)&H3 236.4 16 9 (2W 236.0 11 10 (21e),X = 0 220.7 22 11 (21e), X = S 223.1 22 12 (21e),X = NMe 220.0 22 13 Bu'Me2MCOC=CCH 225.7 15 14 Me3MCOCHBr(CH&CH3 234.8 7 15 Me3MCOCHBr(CH2)3C1 235.2 7 16 Me3MCOSiMe3 318.2,318.8 249.0 23,31a 17 Me3M COCO(CHZ)~CH~ 235.1 7 18 Me3MCOCOPh 220.4 7 19 Me3M COCOCH 235.5 15 20 Ph3MCOCH3 240.1 34 21 (Me3Si)~McoPh 233.8 34 22 (Me3Si)3MCOC(CH3)3 244.6 34 C.'H NMR Spectroscopy.--'H NMR data indicate that protons attached to the a carbon atom of acyl silanes and acyl germanes are generally somewhat deshielded relative to their carbon analogues (Table 3). This property no doubt results from differences in electronegativity and magnetic anisotropy. Interest- ingly, a,P-unsaturated acyl silanes appear to be an exception to this rule (entry 16). D. 29SiNMR Spectroscopy.-The 29SiNMR properties of acyl silanes 30 indicate that the acyl group has a moderate shielding effect upon the silicon atom compared with the situation in tetramethylsilane similar in magnitude to that observed in vinyl trimethylsilane (Table 4). 32 G. A. Olah, A.L. Berrier, L. D. Field, and G. K. Surya Prakash, J. Am. Chem. Soc., 1982,104, 1349. "B. Frei and M. E. Scheller, Helv. Chim. Acta, 1984,67, 1734. 34 A. G. Brook, F. Abdesaken, G. Gutekunst, and N. Plavac, Organometallics, 1982,1,994. Synthesis and Chemistry of Acyl Silunes Table 3 'H Chemical shifts at a-carbon atom of selected acyl metalloids Entry Compound 6(M=C) F(M=Si) G(M=Ge) G(M=Sn) Ref: 1 Me3MCOCHzPh 3.77 3.73 3.80 12 2 Me3MCOCH3 2.07 2.1 8,2.20 5, 7 3 Ph3MCOCH3 2.01 2.30 2.38 5 4 Bu'MezMCOCH3 2.32 8 5 Me3MCOCH2(CHz)3CH3 2.33 2.50 7,9, 10 6 Me3MCOCH(CH2CH3)2 2.50 8 7 Bu'MezMCOCHzCH3 2.64 8, 13 8 Et3MCOCHzCH 3 2.60 13 9 Pr"3MCOCHzCH3 2.60 13 10 Me3MCOCHPhCHzCH3 3.86 8 11 Me3MCOCOCH3 2.03 7, 15 12 Me3MCOCOCH2(CH&CH3 2.55 7 13 M~~MCOCOCHZ(CHZ)ZCH~ 2.58 7 14 Me3MCOCHCl(CHz)3CH3 4.1 1 14 15 Me3MCOCHBr(CH2)3CH3 4.39,4.33 7, 14 16 Me3MCOCH=CHz 6.88 6.28,6.38 7, 15, 16,35 17 Me3MCOCH=CHCH3 6.78 6.62 6.39 7, 12 18 Me3MCOCH=C(CH3)2 6.56 6.32 6.40 12 The nature of the groups attached to the silicon moiety also affects the chemical shift of the silicon atom (Table 4), although it is difficult to see any reliable pattern except in the cases of acyl tris-(trimethylsilyl) silanes and acyl alkyl bis-(trimethylsilyl) silanes (entries 1 1-23). E.X-Ray Diffraction Studies.-Single-crystal X-ray analyses of both acetyl triphenylsilane 37 and acetyl triphenylgermane 38a have been completed.In both cases the three phenyl groups and the acetyl group are located tetrahedrally around the metalloid centre. For the acyl silane, whereas the phenyl-Si bond length is 1.864 A, the Si-CO (acetyl) bond length is significantly longer, at 1.926 A; a typical value for a saturated carbon-silicon single bond would be 1.84- 1.87 A. However, the C=O bond length, which might have been expected to be abnormally long (reflecting the unusually long wavelength carbonyl absorption and the enhanced basicity) was found to be 1.21 &---approximately the same as that normally found in ketones. The bond angles involving the carbonyl group were approximately the theoretical 120" and the CH3-CO bond length of 1.50 8, was normal. The structure of acetyl triphenylgermane is very similar to its silicon analogue and shows a lengthening of the Ge-C (acetyl) bond by 0.066 8, compared with the Ge-Ph bond.Trotter ascribes the lengthening of the metalloid-carbon (acetyl bond) to a contribution to the structure not only of canonical forms (6) and (7), whose carbon analogues contribute to the structure of ordinary ketones, but also (8) in ''I. Naito, A. Kanishita, and T. Yonenitsu, Bull. Chem. SOC.Jpn., 1976,49, 339. Page, Klair, and Rosenthal Table 4 29Si Chemical shifts of selected acyl silanes and related compounds En try Compound wi)/PPm Ref. 1 Me3SiCOMe -10.1 32 2 Ph 3SiCOMe -30.4 34 3 (Me3Si)& -135.5 34 4 Ph3SiCOPh -28.3 34 5 Et3SiCOPh -28.3 34 6 Me3SiCOPh -7.4, -15.1 32,34 7 Me3SiCOSiMe3 -14.4 23 8 (21e),X = 0 -6.3 22 8 (21e),X = S -6.9 22 9 (21e),X = NMe -8.8 22 10 Me3SiCHXH2 -7.6 36 11 (Me3Si)3GeCO-Ad * -5.24 28 12 (Me3Si)&iCO-Ad * -78.8 34 13 (Me3 Si)3SiCO-Ad * -11.5 34 14 Me(Me3Si)zSiCO-Ad* -50.38 29 *15 M~(M~JS~)~S~CO-A~ -13.72 29 16 But( Me $3) 2SiCO- Ad * -32.24 29 *17 BU'(M~~S~)~S~CO-A~ -15.48 29 18 Ph(Me3Si)zSiCO-Ad* -44.75 29 19 Ph(Me3Si)2SiCO-Ad* -13.45 29 20 (Me3Si)3SiC02H -73.84 26 21 (Me3Si)3SiC02H -6.15 26 22 (Me3 Si) SiCOz Me -74.45 26 23 (Me&)3SiCOzMe -6.34 26 * Ad = adamantyl which there is no formal bond between the metalloid atom and the acetyl carbon atom, a structure considered possible as a contributing resonance form because of the considerable differences in electronegativity between the metalloids and ~arbon.~'.~*'The acetyl triphenylgermane molecule is therefore represented by Trotter as (9), a structure said to represent both the basicity of the ketone oxygen atom and the long germanium-arbonyl carbon bond.37 0 0-0' E.Lippmaa, M. Magi, V. Chvalovsk'y, and J. Schraml, CON. Czech. Chem. Commun., 1977,42,318, 37 P. C. Chieh and J. Trotter, J. Chem. SOC.,1969,1778. 38 (a) R. W. Harrison and J. Trotter, J. Chem. Soc., 1968, 258; (b) F. Agolini, S. Klemenko, I. G. Csizmadia, and K. Yates, Spectrochim. Acta, 1968, 24a, 169; (c) G. J. Brealey and M. Kasha, J. Am. Chem. SOC.,1955, 77, 4462; (d) R. West, R. H. Baney, and D. L. Powell, J. Am. Chem. SOC.,1960,82, 6269; (e)N.A. Matwiyoff and R. S. Drago, J. Organomet. Chem., 1965,3,393. 153 Synthesis and Chemistry of Acyl Szlanes F. Ultraviolet and Visible Spectroscopy.-Acyl silanes display most unusual ultraviolet and visible spectral characteristics (Table 5) 39 44 Several acyl metalloids, including acetyl trimethylsilane and acetyl trimethylgermane, have been studied by photoelectron spectro~copy,~' and a range of ab znitio and semiempirical calculations have been carried out 41 42 Acyl silanes show absorp- tions due to n --x* and 71 __+ x* transitions, as do simple ketones However, the n x* excitation of alkyl acyl silanes occurs around 370 nm, a shift of about 100 nm to longer wavelength compared with the analogous carbon compound pinacolone (vmax279 nm) This translates into a ca 25 kcal mol-' lowering in the energy of the n n* transition Comparison of extinction coefficients of acetyl trimethylsilane and pinacolone (E = 126 and 21 respectively) shows that absorption for the silicon derivative is far more intense In addition, considerable fine structure is observed, usually consisting of three main bands, sometimes with two additional shoulders at lower wavelengths The persistence of this vibrational structure in polar solvents, an effect not commonly observed in ketones, is not clearly understood In aryl acyl silanes, which may be lime green in colour, and in a,P-unsaturated acyl silanes, the n -n* transition occurs around 420 nm, again shifted to longer wavelength by about 100 nm than in the corresponding carbon compounds In the yellow a-carboxyacyl silanes the transition appears at 455 nm, and in a-ketoacyl silanes around 520 nm, accounting for the deep crimson colour of this species, bis-(trimethylsilyl) ketone (4) is also pink-coloured The x --+ x* transition for the carbonyl group in aryl acyl silanes produces a fairly intense absorption band in the range 250-260 nm The position and extinction coefficient of such transitions does not vary greatly, regardless of substituents on the silyl group, although slight red shifts are observed in polar solvents as would be expected for conjugated carbonyl group transitions As is seen for ketones, a third absorption is observed in the region 185-195 nm For acyl silanes with aromatic substituents this could either be ascribed to a primary benzene band or to a second x --x* transition, whereas for other acyl silanes this transition (195 nm, E = 4200 for acetyl trimethylsilane) is presumably the latter In general the nature of the groups (other than acyl) attached to the silicon atom has little effect on the energies of n-x* and x---+x* transitions 3 Synthesis of Acyl Silanes A.Simple Acyl Manes.-(1) Formyl Silanes Formyl silanes have been investigated only briefly, due primarily to their instability After several years of speculation 39 R West, J Organomel Chem ,1965,3,314 40 D F Harnish and R West, Inorg Chem , 1963,2,1082 41 B G Ramsey, A G Brook, A R Bassindale, and H Bock, J Organomet Chem ,1974,74, C41 42 E B Nadler, Z Rappoport, E Arad, Y Apeloig, J Am Chem SOC,1987,109,7873 43 L E Orgel, in 'Volatile Silicon Compounds', ed E A V Ebsworth, Pergamon Press, Oxford, 1963, p 81 44 K Yates and F Agolini, Can J Chem ,1966,44,2229 154 Page, Klair, and Rosenthal Table5 UVl Visible absorption of selected metalloids and related compounds Entry Compound 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 31 32 33 34 35 36 Me3CCOMe Me3SiCOMe Et 3GeCOMe Me3SiCO(CH2)4CH3 Me3SiCO(CH2)&H3 Me3SiCOCHBr(CH2)4CH3 (1 12) Ph3CCOPh Ph3SiCOPh Ph3GeCOPh PhsSnCOPh Me3SiCOPh Me3GeCOPh Me3SiCOCHzPh Me3GeCOCHzPh Me3SnCOCHzPh Me3SiCOCH=CH2 Me3SiCOCH=CHCH3 Me3GeCOCH=CHCH3 Me3SnCOCH=CHCH3 Me3SiCOCH=C(CH& Me3GeCOCH=C( CH 3)2 Me3SnCOCH=C(CH& Me3SiCOC(CH3)=CH2 Me3SiCOCH=CH(CH2)&H3 Me3SiCOC&CH3 Bu'Me2SiCOCO2Me Me3SiCOC02Et PhMezSiCOCO2Me Me3SiCOCOCH3 Me3SiCOCO(CH2)4CH3 Me3SiCOCO(CH2)3CH3 Me3SiCOCOPh Me3SiCONEt2 BU'(M~~S~)~S~CO-A~* Ph(Me3Si)zSiCO-Ad* 278 (1 5) 372 (126) 365 (173) 367 (129) 357 (131) 374 (142) 372 (16) 329 (299) 424 (292) 417 (210) 435 425 (1 27) 412 (120) 377 (113) 366 (1 24) 379 434 (96) 424 (98) 416 (109) 432 439 (120) 432 (140) 453 425 (77) 424 (68) 420 (1 70) 455 (loo), 227 (254) 455 (97), 230 (388) 455 (213), 279 (620) 535 (99) 530 (96) 528 (98) 518 (117) 264 (270) 380 (200) 380 (200) Ref: 5 5 5 9 7 7 33 251 (11 600) 5 257 (16 200) 5 254 (16 200) 5 5 250 (14 500) 5, 7 252 (10 700) 5 12 12 12 213 (8630) 15 224 (10 300) 7, 12 258 (1 1 300) 12 258 12 249 (12 700) 12 12 12 222 (5900) 16 225 (10 358) 16 227 (7450) 22 24 24 24 15 7 7 275 (6500) 7 25 29 29 about these specie^,^^^^ Ireland and Norbeck obtained evidence for formyl trimethylsilane MeSSiCHO by a trap with a Wittig reagent following Swern oxidation of trimethylsilyl methanol [see Scheme 9,Section ~A(v)].~~*~~Subse-45 L.H. Sommer, D. L. Bailey, G. M. Goldberg, C. E. Buck, T. S. Bye, F. J. Evans, and F. C. Whitmore, J. Am. Chem. SOC.,1954,76, 1613. 46 A. J. Mancuso, D. S.Brownfain, and D.Swern, J. Org. Chem., 1979,44,4148. 47 R.E. Ireland and D. W. Norbeck, J. Org. Chem., 1985,50,2198. 155 Synthesis and Chemistry of Acyl Silanes CI I Cpgr?o(10) SIMe, quently Tilley identified formyl trimethylsilane by NMR spectroscopy as the product of reaction of the zirconium q 2-sila-acyl complex (q 5-C5H5)2Zr-(q2-COSiMe3)Cl (10) with hydrochloric More recently, Tilley was able to prepare, isolate, and characterize the first stable formyl silane, formyl tris-(trimethylsily1)silane (Me3Si)3SiCH0 (1 l), from the zirconium q 2-sila-acyl complex (12).49 This complex was prepared in good yield by reaction of silyl zirconium species (13) with carbon monoxide at 100 psi. Treatment of complex (12) with anhydrous hydrogen chloride at -78 "C in toluene solution gave formyl silane (ll), which was stable under an inert atmosphere and showed typical carbonyl group reactivity (Scheme la).Formyl silanes, formed zn sztu, have been used to prepare a,P-acetylenic acyl silanes by lithium acetylide addition and Swern ~xidation.~' co- CI I 100 psi Cpcp*Zr?o -78 OC +rt Scheme la (ii) Hydrolysis of Thioacetals. The most general synthesis of acyl silanes is based on dithiane methodology, first investigated by Brook 51 and Corey 52 in the late 1960's. This route is outlined in Scheme lb. The major drawback of this synthesis lies in the final deprotection step; ease of hydrolysis of the silyl 1,3-dithiane (14) with retention of the silicon moiety is very dependent on the nature of R and on the size of the groups attached to silicon.In general, the larger these groups are, the more successful is the final deprotection. Initial methods of hydrolysis involved the use of mercury(I1) salts, and although this process has been improved by the use of such reagents as chloramine T 48 B K Carnpion,J Falk, and T D Tilley, J Am Chem SOC,1987,109,2049 49 F H Elmer, H -G Woo, and T D Tilley, J Am Chem Soc , 1988,110,313 50 R J Linderrnan and Y Suhr, J Org Chem ,1988,53,31569 51 A G Brook, J M Duff, P F Jones, and N R Davis, J Am Chem Soc, 1967,89,4431 52 E J Corey, D Seebach, and R Freedman, J Am Chem SOC,1967,89,434 Page, Klair, and Rosenthal various 0fi nBuLi* fl -A'yS X,SiC1 R ASIR', R R SIR', (14) Scheme 1b h~drate,~~-~~this step still remains a significant problem, production of the corresponding aldehyde accounting for lO-lOO% of the product mi~ture.~ To circumvent this difficulty a number of more readily hydrolysed thioacetals have been utilized, for example, lithio bis-(methylthio) methanes (15) (Scheme 2).57 Some interesting cyclic acyl silanes such as (105) have been prepared using the dithiane method.54 ___) SIMe, SMe 70% 0SMe-Scheme 2 0-Trimethylsilyl hemithioacetals (16) have been used in the synthesis of bis-(trimethylsilyl) ketone (4) as hydrolysis of these compounds occurs under very mild condition^.^^ The route is outlined in Scheme 3; oxidation of methyl tris- (trimethylsilyl) methyl thioether (17) with rneta-chloroperbenzoic acid gives the unstable sulphoxide species (18), which readily undergoes a sila-Pummerer rearrangement, perhaps due to relief of steric compression, to give the inter- mediate hemithioacetal (16)." This hemithioacetal reacts with the Pummerer intermediate to give a 1:1 mixture of bis-(trimethylsilyl) ketone (4)and the thioketal (19).However, this route suffers from difficulties in the purification of the acyl silane product and in preparation of the highly functionalized starting materials (17). The sila-Pummerer rearrangement of bis-(trimethylsilyl) phenyl- selenenyl phenyl methane, induced by hydrogen peroxide, has also been used to prepare benzoyl trimethylsilane in 46% yield.60 A much better synthesis of acyl silanes is based on methoxy phenyl- thio trialkylsilyl methane (20), which acts as an a-silyl acyl anion equivalent 53 D.W. Emerson and H. Wynberg, Tetrahedron Lett., 1971, 12, 3445; M. E. Scheller, G. Iwasaki, and B. Frei, Helv. Chim. Acta, 1986,69, 1378. 54 A. G. Brook and H. W. Kucera, J. Organomel. Chem., 1975,87,263. 55 H. J. Reich, J. J. Rusek, and R. E. Olson, J. Am. Chem. Soc., 1979, 101,2225. 56 R. L. Danheiser and D. M. Fink, Tetrahedron Lert., 1985,26,2509.''R. Biirstinghaus and D. Seebach, Chem. Ber., 1977,110,841.'* A. Ricci, A. Degl'lnnocenti, M. Ancillotti, G. Seconi, and P. Dembech, Tetrahedron Lett., 1986, 27, 5985. 59 D. J. Ager, Chem. Rev., 1982,11,493. 6o H. J. Reich and S. K. Shah, J. Org. Chem., 1977,42, 1773.Synthesis and Chemistry of Acyl Silanes 0 Me,SI ASlMe3 (Me,SI),C/' s+N CH, (4) + Me,Si TicH3 OSIMe, (Me$I)&+SCH, -Me,SI (16) S+ 'CH, Me,SI (19) Scheme 3 (Scheme 4) 61 This synthesis is successful for a wide range of aliphatic R groups and provides an excellent overall yield However, the synthesis of aromatic acyl silanes by this method would clearly be difficult or impossible, requiring aromatic nucleophilic substitution (20) Scheme 4 (in) Silyl Metallic Species One of the earliest syntheses of an acyl silane involved the reaction of triphenylsilyl potassium with benzoyl chloride at low temperature (Scheme 5a) 0 Ph3SIK + Ph PhA!3Ph3 Scheme 5a The very low yield of benzoyl triphenylsilane (6%) and the similar yields obtained from the reactions of silyl lithium reagents with acetyl chloride62 to give acetyl triphenylsilane demonstrate that this IS not a very useful method for the synthesis of simple acyl silanes However, this method has proved successful 61 T Mandai, M Yamaguchi, Y Nakayama, J Otera, and M Kawada, Tetrahedron Lett, 1985, 26, 2675 62 D Wittenberg and H Gilman, J Am Chew Soc ,1958,80,4529 Page, Klair, and Rosenthal for the preparation of acyl tris-(trimethylsilyl) silanes and their derivative^.^^.^^'^^ The tris-(trimethylsilyl) silyl (‘sisyl’) lithium reagent is formed by deprotonation of tris-(trimethylsilyl) silane using an alkyl lithium reagent; alkyl bis-(trimethyl- silyl) silyl lithium species have been prepared by a number of methods (Scheme 5b).29 Some of these silyl lithium reagents, which are yellow to orange to brown in colour, are remarkably stable; for example phenyl bis-(trimethylsilyl) silyl lithium forms a pale yellow solid ‘ate’ complex Ph(Me3Si)zSiLi.THF which could be recrystallized.tBuSiCl MeLi/ Et20 Me3SICt / Lt (~e~Si)~Si’Bu brown solution -(Mefii):BuSILI thf, 18 days, TMEDA. thf rt, 79% 2 days, 50 OC PhMgBr MeLi/Et20 -(Me3SI)3SIPh -(&Sl)phSILl.3THF(Me3SI)3SIBr rt, 95% thf,48h, (recrystallised) pale yellow rt, 81% KCOCI (Me3SI),RSILI P pentanes, 0 OC Scheme 5b The tris-(trimethylsily1)- and alkyl bis-(trimethylsily1)- silyl lithium reagents react cleanly with acyl chlorides, typically at 0 “C, to give the corresponding acyl silanes in up to ca.85% yields. Tris-(trimethylsilyl) silyl lithium, which also forms an ‘ate’ complex with three molecules of THF, has also been shown to react with carbon dioxide to give the interesting crystalline tris-(trimethylsilyl) silane carboxylic acid after acidic work-up in 85% yield.26 The methyl ester and several silyl esters were prepared from this acid by conventional means (Scheme 5c). A less reactive silyl metal species may be prepared by the addition of copper(1) salts to form a silyl cuprate. Lithium bis-(triphenylsilyl) cuprate reacts with a variety of acyl chlorides to give the corresponding acyl silanes in moderate to good yields (Scheme 6a).64 Dilithium bis-(trimethylsilyl) cyano cuprate Liz(Me3- ‘’A.G. Brook and K. M. Baines, Adv. Organornet. Chern., 1986, 25, 1; G. Raabe and J. Michl, Chern. Rev., 1985, 85, 419; ‘The Chemistry of Organic Silicon Compounds’, ed. S. Patai and Z. Rappoport, Wiley, 1989. 64 N. Duffaut, J. Dunogds, C. Biran, R. Calas, and J. Gerval,J. Organornet.Chern., 1978,161, C23 Synthesis and Chemistry of Acyl Silanes 1.alkyllithium -(Me3SIhSIC02R(MqSl)3SlH * (Me$l)3SlC02H 2.co, 3. HCl R = CH,Ph3Si,Me3Si, (Me3Si)3Si Scheme 5c Si)2CuCN was found to be particularly effective for the preparation of sterically hindered acyl silanes (Scheme 6b); this reaction appears to be a good general preparative method.’ ’ R Yield Me >90% Ph3SILI cul;RCOCl w Et >90% Bu 75% Ph <20% Scheme 6a Scheme 6b Recent work by Kang utilizes the low ionic character of the aluminium-silicon bond to develop probably the most general and effective synthesis of simple acyl silanes that has been reported in the literature to date, successful for aliphatic, aromatic, heteroaromatic, and even cyclopropyl acyl ~ilanes.~’ Acyl chlorides are treated with lithium tetrakis-(trimethylsilyl) aluminium or lithium methyl tris- (tnmethylsilyl) aluminium in the presence of a cuprous cyanide catalyst to give the acyl silanes in truly excellent yields after work-up (Scheme 7).65 J Kang, J H Lee, K S Kim, J U Jeong, and C Pyun, Tetrahedron Lett, 1987,28,3261 Page, Klair, and Rosenthal LIMeAI(SiMe3), *. CuCN, Et20 typically up to cu. 90%~ Or LIAI(SIMe,), 2.Rcm1, thf R 3. H+ Scheme 7 (iv) Palladium-catalysed Coupling.-A wide range of aromatic and heteroaromatic acyl silanes has been synthesised via transition metal catalysed coupling (e.g. 8).22766Scheme This synthesis is very successful for a variety of substrates, including furyl, thienyl, and pyrryl, and also for aryl acyl silanes containing electron-withdrawing groups, which are otherwise difficult to prepare. However the method cannot be employed for the synthesis of aliphatic acyl silanes66 and is therefore complementary to some of the methods described above. 0 0 (n-ally1 PdCl), V C I + Me,SiSiMe, P(OEt)3, toluene, A (21e),X=O,S, NMe Scheme 8 (v) Oxidation Reactions. Among the most common routes for the synthesis of acyl silanes are those based on oxidation methodology.Since the oxidation of secondary alcohols to ketones by any of a wide variety of oxidizing agents is commonly a successful reaction, it might be expected that the corresponding oxidation of a-hydroxy silanes would be useful for acyl silane ~ynthesis.~ However, this method is often far from straightforward. Several oxidizing systems have been examined,67 but the most satisfactory method for oxidation of a-hydroxy silanes to acyl silanes is based on the Swern oxidation and is shown in Scheme 9.46,47,50 DMSO, (COCl)2, -60 OC, or 0 4 -SIMe, DMSO, (CF,CO),O, -78 "C R A sIM3 Scheme 9 If chromic acid oxidation reagents are used for this transformation the major products often arise from silicon-carbon bond cleavage, as illustrated in Scheme 10,' although such a reaction has been used successfully for the preparation of cyclopropyl acyl silanes (Section 3E).68 66 K.Yamamoto, S. Suzukr, and J. Tsuji, Tetrahedron Lett., 1980,21, 1653. 67 A. G. Brook and J. Pierce, J. Organomet. Chem., 1965,30,2566. 68 M. E. Scheller and B. Frei, Helv. Chim. Acta, 1986, 69, 44; M. E. Scheller and B. Frei, Helv. Chim. Acta, 1985,68,44. 161 Synthesis and Chemistry of AcyI SiIanes H20 : 3 Scheme 10 However, this problem of silicon-carbon bond cleavage has been cleverly utilized in an acyl silane synthesis by incorporation of two silicon moieties (Scheme 11). Scheme1 1 Benzoyl silanes and substituted benzoyl silanes have been produced from the corresponding a,a-dibromobenzyl silanes by oxidation with silver acetate in acetone-ethanol-water mixture4 or silica gel 69 in good yields (Scheme 12).The very first synthesis of an acyl silane was accomplished using this meth~d,~ but it is clearly restricted to benzyl and ally1 silanes where geminal dihalides can easily be procured. Brook has been able to prepare the diacyl silane (20a)4 and Corriu has prepared several interesting cyclic acyl silanes such as (21)" using this ap- proach. 0 NBS PhASIPh2CI - Br Br AgOAc PhKSlPh20H 0 Various acyl silanes have been synthesized by the photosensitized oxygenation of silyl diazo compounds using meso-tetraphenyl porphine as the sensitizer (Scheme 13).24 Although this synthesis is successful for aromatic acyl silanes and the 69 A Degl'lnnocenti, D R M Walton, G Seconi, G Pirazzini, and A Ricci, Tetrahedron Lett, 1980, 21, 3927 'O R Corriu and J Masse, J Organomet Chem ,1970,22,321 Page, Klair, and Rosenthal Table 6 Acyl silane preparation from silyl diazo compounds 24 Entry Silyl diazo compound Acyl silane Yield/% 1 2 4 5 6 3 Me3SiC(Nz)C02Et Me3SiC(N2)C02Me Me&iC(Nz)Ph PhMe2SiC(N2)Ph Ph3SiC(N2)Ph PhMezSiC(N2)COzMe Me3SiCOC02Et Me3SiCOCOzMe Me3SiCOPh PhMezSiCOPh Ph SiCOPh PhMe2SiCOC02Me 40 42 42 51 64 43 - N R'R22SI SIR'R22 + Hg(N2CC02dh R'R2,SI 'CO@ Scheme 13 interesting yellow-coloured a-carboxyacyl silanes, as shown in Table 6, it is not general, and purification difficulties are often encountered in the synthesis of aliphatic and other acyl silanes, resulting in low overall yields.A successful alternative method for acyl silane synthesis involving silyl diazo intermediates utilizes trimethylsilyl diazomethane (Scheme 14).7 The lithiated RX,thf, MCPBA, Me,SI A LI -780c W3SI ph (22) (23) buffer pH 7.6 Scheme 14 derivative (22) smoothly reacts with alkyl halides in THF solution in good yields to give a-trimethylsilyl diazoalkanes (23). Cleavage of the diazo moiety is effected in benzene solution using meta-chloroperbenzoic acid oxidation in the presence of a phosphate buffer (pH 7.6) to prevent side reactions. This gives access to a wide variety of acyl silanes in yields as high as 71%. Once again, however, aromatic acyl silanes cannot be readily prepared using this route since an aromatic nucleophilic substitution reaction would be required.Aromatic acyl silanes and the unstable pink bis-(trimethylsilyl) ketone (4) have been prepared in good yields oia the oxidation of phosphonium ylids (Scheme 15).23 (vi) Hydroboration-Oxidation of Alkynyl Silanes. Hydroboration-oxidation of alkynyl silanes is a superior method for producing a variety of substituted acyl silanes from readily available starting rnaterial~.~'.~~Although the T. Aoyama and T. Shioiri, Tetrahedron Lett., 1986,27,2005.''A. Hassner and J. Soderquist, J. Organomet. Chem., 1977,131, C1. G. Zweifel and S. J. Backlund, J. Am. Chem. Soc., 1977,99,3184. '3 Synthesis and Chemistry of Acyl Silunes Scheme 15 hydroboration-oxidation procedures initially studied gave only moderate yields of acyl silanes, the sequence was modified by Zweifel to produce an excellent and general one-pot synthesis of acyl silanes which proceeds in high yields (Scheme 16).' In the opinion of the present authors, this method is the best for the preparation of simple alkyl acyl silanes.One limitation of this scheme is that the yield of acyl silane is reduced if the ;.oup R is very bulky; furthermore, aryl acyl silanes obviously cannot be prepared using this method. The use of alkaline peroxide in the oxidation step must be avoided as the acyl silane product is converted into silyl ester by this re-agent. BH,.Me2S R- SiMe, =- (Rq B Me,NO1 ca 80% \ '3 Scheme 16 (vii) Rearrangement of Sifyfoxycarbenes.Certain acyl silanes may be obtained by the pyrolytic rearrangement of silyloxycarbenes (24) derived from a-ketoesters (25) (Scheme 17).74 Scheme 17 Page, Klair, and Rosenthal The generality of this approach is restricted, for example by possible difficulty in the preparation of the requisite a-ketoesters (25), and by the fact that the pyrolysis only takes place in high yield when R = Ph.(viii) Enolate Methodology. The trimethylsilyl enol ethers of acyl silanes (27) have been prepared using a variety of routes and are obvious precursors to acyl silanes, since simple hydrolysis of the silyl enol ethers would be expected to liberate the corresponding acyl silanes in excellent yields.Acyl imidazoles (26) react via a silyl acyloin reaction under somewhat unpleasant conditions to give the corresponding silyl enol ethers (27) in moderate yields (Scheme 18).’ Mg/Me3SiC1 HMPA,7O0C (27)(26) Scheme 18 A possible mechanism for this reaction is outlined in Scheme 19. OSIMe3 AN AN ANA N -_t ANAN u Lf u Scheme 19 The lithium alkoxides of bis-(trimethylsilyl) alkan-1-01s (28) react with benzo- phenone to produce the silyl enol ethers of acyl silanes in good yields (Scheme 20).76 (28) Scheme 20 l4 A. G. Brook, J. W. Harris, and A. R. Bassindale, J. Organomet. Chem., 1975,99,379. lS P. Bourgeois, J. Dunogues, N. Duffaut, and P. Lapouyade, J. Organomet. Chem., 1974,80, C25.l6 I. Kuwajima, M. Arai, and T. Sato, J. Am. Chem. SOC.,1977,99,4181. 165 Synthesis and Chemistry of Acyl Silanes The parent bis-(trimethylsilyl) alkan-1-01s were synthesized in reasonable yields by hydrolysis of the geminal disilane carbinol silyl ethers (29),77978 which in turn were synthesized from the corresponding esters using a silyl acyloin reaction which, ironically, proceeds via formation of an acyl silane intermediate (Scheme 21).77 /-Me,SiOR Na, Me3SiC1 4 SIMe, SiMe:, (29) Scheme 21 Picard has reported alternative, rather more direct, approaches to acyl silanes and to silyl enol ethers (27) by similar reductive silylation of substituted benzoates and of a$-dihalo-a$-unsaturated acyl chlorides respectively using trimethyl chlorosilane-magnesium-HMPA.79 A much more general synthesis of silyl enol ethers (27) is based on the reductive cleavage of the carbon-sulphur bond of the silyl enol ether of a thioester using sodium metal and chlorotrimethylsilane, again in a silyl acyloin reaction (Scheme 22).80*81 80-90% 7040% (27) Scheme 22 Another efficient synthesis of silyl enol ethers (27) utilizes an intramolecular 1,2-silicon shift in a-silyl acyl lithium species (29a), prepared from an a-lithiosilane 77 I Kuwajima, N Minami, T Abe, and T Sato, Bull Chem Soc Jpn ,1978,51,2391 78 I Kuwajima, T Sato, N Minami, and T Abe, Tetrahedron Lett, 1976,17, 1591 79 J -P Picard, R Calas, J Dunogues, and N Duffaut, J Organomet Chem , 1971, 26, 183, J -P Picard, R Calas, J Dunogues, N Duffaut, J Gerval, and P Lapouyade, J Org Chem, 1979,44,420, J -P Picard, A Ekouya, J Dunogues, N Duffaut, and R Calas, J Organomet Chem, 1975, 93, 51, P Bourgeois, J Organomet Chem ,1972,76, C1 I Kuwajima, M Kato, and T Sato, J Chem SOC,Chem Commun ,1978,478 T Cohen and J R Matz, J Am Chem SOC,1980,102,6900 Page, Klair, and Rosenthal 0 ItCO, Et20 -Scheme 23 (Scheme 23).82 This method is very simple to carry out in the laboratory and produces the silyl enol ethers in good yield with high isomeric purity (usually E isomer >90% of mixture).Methyl enol ethers of acyl silanes have been prepared in good yield by the silylation of vinyl lithium reagents derived from methyl enol ethers (Scheme 24).83Reaction conditions were found to be critical; best results were obtained by the low temperature addition of an excess of t-butyllithium in THF solution followed by gradual temperature increase to 0 "C, followed in turn by addition of chlorotrimethylsilane at -78 "C.Scheme 24 (ix) Silylation of Metalloaldimines. Several acyl silanes have been prepared by the silylation of metalloaldimines followed by hydrolysis (Scheme 25).84*85 Scheme 25 One limitation of this scheme is the ready decomposition of the aldimine (30) to give aldehyde in addition to acyl silane in approximately equal amounts (Scheme 26): * 82 S. Murai, I. Ryu, J. Iriguchi, and N. Sonoda, J. Am. Chem. SOC.,1984,106,2440. 83 J. A. Soderquist and G. J.-H. Hsu, Organometallics, 1982, 1, 830; E.M. Dexheimer and L. Spialter, J. Organomet. Chem., 1976,107,229. 84 P. Bourgeois, J. Organomet. Chem., 1974,76, C1. 85 G. E. Niznik, W. H. Morrison, and H. M. Walborsky, J. Org. Chem., 1974,39,600. Synthesis and Chemistry of Acyf Silanes 0 R ASIMe,"R' R ASIMe, (30) 0 ca. 1:l Scheme 26 (x) Silylation of Acyf Metallic Species. Perhaps the most direct method of synthesizing an acyl silane is by reaction of an acyl lithium with a silicon electrophile as shown in Scheme 27.86,87 Although this method is successful for a variety of alkyl acyl silanes in moderate yields, low temperatures must be used in order to avoid destruction of the products as indicated, and the method is not suitable for aryl acyl silanes., Ll N,N-Diethyl carbamyl trimethylsilane (3 1) has been prepared by the reaction of bis-(trimethylsilyl) sulphide with bis-N,N-diethyl carbamyl mercury (Scheme 28).25 B. a-Haloacyl !Manes.-The most obvious procedure for the synthesis of a-haloacyl silanes is direct bromination of silyl enol ethers (27) using bromine at low temperature (Scheme 29a).14,88 While this route can be successful, it does 86 D Seyferth and R M Weinstem, J Am Chem SOC, 1982, 104,5534 D Seyferth and R C HUI,Organometallm, 1984,3, 327 I Kuwajima, T Abe, and N Minaml, Chem Lett, 1976,993 Page, Klair, and Rosenthal I Br ca. 80% Scheme 29a suffer from the general sensitivity of both the starting materials and the products. Electrophilic halogenation of alkyl enol ethers of acyl t-butyldimethylsilanes, prepared by deprotonation and silylation of vinyl ethers,83 has also been used (Scheme 29b).89 a-Iodoacyl silanes were prepared by the same authors by treatment of a-bromoacyl silanes with sodium iodide in acetone.1.tBuLi OEt NBS orNCS 0 2. tBuMezSiCl A~i-fsu %O/MeCN* Scheme 29b There is also one report in the literature that treatment of a series of bis- (trimethylsilyl) alkan-1-01s (32) with N-bromosuccinimide in carbon tetrachloride gives the corresponding a-haloacyl silanes in moderate yields (Scheme 30).88 Scheme 30 The best method to date for preparing a-haloacyl silanes is based on bromination of an enol borinate (33) at OOC, a route which proceeds in good yield and involves no sensitive intermediates.This method offers a convenient one-pot synthesis of a-haloacyl silanes from readily available starting materials as the intermediate enol borinates are very easily prepared from silyl acetylenes (Scheme 31a).90791 (33) 50-60%overall Scheme 31a 89 J. S. Nowick and R. L. Danheiser, J. Org. Chem., 1989,54,2798. 90 P. C. B. Page and S. Rosenthal, Tetrahedron Lett., 1986,21,5421. 91 J. Hoozand 3. N. Bridson, Can. J. Chem., 1972,50,2387. Synthesis and Chemistry of Acyl Silanes Recent work has shown that a-bromoacyl silanes may be isolated in variable yields from the reaction of magnesium bromide etherate with 2-phenylsulphonyl- 2-trimethylsilyl oxiranes (Scheme 31 b) 920 0 sIMe, + other products SO2Ph 2 -79% + RFiMe3R Br Scheme 31b C.a-Ketoacyl Si1anes.-a-Ketoacyl silanes are a deep rich crimson colour and are particularly sensitive to light The a-ketoacyl silanes (34) were first synthesized in 1983 uza allene methodology as shown in Scheme 32,'' although the yellow- coloured a-carboxyacyl silanes were already known at that time 24 [Section 3A(v)] Thus, isomerization of the propargyl ether (35) with potassium t-butoxide at 70 "C gave the allene (36a) in excellent yield 92b Lithiation of allene (36a) with butyl-lithium at -78 "C and subsequent reaction with a chlorosilane to give (36b), followed by oxidative work-up with meta-chloroperbenzoic acid (presumably proceeding vzu an epoxide) provided the unstable a-ketoacyl silanes (34) in moderate yields (35) Scheme 32 A much more general and very simple synthesis requiring a minimum of laboratory manipulation utilizes a Swern oxidation of the corresponding diols (37) to give the a-ketoacylsilanes directly in useful yields (Scheme 33) 93 Purification in this case was accomplished in the dark by chromatography at -78 OC or distillation C T Hewkin and R F W Jackson, Tetrahedron Lett, 1990,31, 1877 92b S Hoff, L Brandsma, and J F Arens, Red Trav Chim Pays-Bas, 1968,87,916 93 P C B Page and S Rosenthal, Tetrahedron Lett, 1986,27,2527 Page, Hair, and Rosenthal HO OH DIBAL --R -R~R-SIMe, SIMe, J-tSlMe3Et20 (CH&”JO (37) j2em/swemoxidation 40-50% 0 Scheme 33 D.a#-Unsaturated Acyl Silanes.-&Unsaturated acyl silanes, which are yellow or yellow-green in colour, are less sensitive than the other types of acyl silane discussed above and have been synthesized by several different routes.(i) Hydroboration-Oxidation of Enynes. +Unsaturated acyl silanes have been prepared by hydroboration methodology, similar to that used in the synthesis of aliphatic acyl silanes (uide supra) (Scheme 34).72 This synthesis is somewhat unsatisfactory due to the difficulty of synthesizing functionalized enynes such as (38). BHC12.Et20; Me3N0.2H20ur: MeOH/Me3N (38) Scheme 34 (ii) Oxidation of Allylic Carbinols. A superior synthesis of a$-unsaturated acyl silanes is based on the Swern oxidation of a-hydroxy allyl silanes (39) (Scheme 35)? R~O “BuLi, Me,SiCl, -78 “C;H * .+,,R‘ ‘BuLi, -78 +-33 “C oxidation SIMe3 R’ R‘ (39) Scheme 35 This simple two-step synthesis hinges on the Wittig rearrangement 94,95 as illustrated in Scheme 36, and is successful on a large scale.The metallation of allyl silyl ethers (40) generates a rapidly interconverting mixture of two organometallic species (41) and (42). Although alkylation of this mixture of organometallic derivatives generally proceeds at the C-3 position [via (4l)], 94 W. C. Still and T. L. Macdonald, J. Am. Chem. SOC.,1974,96, 5561. 95 A. Wright and R. West, J. Am. Chem. SOC.,1974,%, 3214. 171 Synthesis and Chemistry of Acyl Silanes HR~O "BuLi, Me,SiCl, -78 "C; R' R' (39) Scheme 36 'hard' electrophiles such as protons react predominantly at the oxygen atom of the alkoxide intermediates (42) leading to formation of the desired cx-hydroxy ally1 silanes.(iii) Enolate Methodology. A number of a$-unsaturated acyl silanes have been prepared via silyl enol ethers of acyl silanes as shown below (Scheme 37).96 Addition of phenyl sulphenyl chloride to the silyl enol ether (43) with subsequent elimination of chlorotrimethyl silane gave the a-thioacyl silane (44). Oxidation of the sulphide (44) to the sulphoxide (45) with meta-chloroperbenzoic acid, followed by in situ elimination of benzenesulphenic acid produced the a$-unsaturated acyl silane in good yield. +Unsaturated acylsilanes have also been prepared in a stereospecific manner SIMe, MCPBA R+ SIMe,~ R SPh PhS+, 0-0 96 N.Minami, T. Abe, and I. Kuwajima, J. Organomet. Chem., 1978,145, C1. 172 Page, Hair, and Rosenthal by an interesting aldol-Peterson reaction sequence uia intermediate (46) (Scheme 3 8). p9 LDA, RX-w3si4s1Me3 _____) Y3SdSiW RCHO -oysik&sLDA; (47) R sim3/ mainly E isomer; R' 4 S i Y 3 8O-90% yields R Scheme 38 Although this synthesis is reasonably successful for a variety of acyl silanes, it is rather lengthy and requires initial preparation of the sensitive acyl silanes (47). (iv) Hydrolysis ofsilyl Dienes and Silyl Allenes. One of the simplest methods for preparation of an +unsaturated acyl silane is by hydrolysis of a silyl diene (48) (Scheme 39).' * However, preparation of more highly functionalized dienes would necessarily involve difficult and lengthy synthesis and this method is therefore most useful for simple substrates.(48) 60-70% yields Scheme 39 Perhaps the most versatile synthesis of a,P-unsaturated acyl silanes involves the use of allene methodology as developed by a number of group^,'^.'^-^^ indeed the first example of an a$-unsaturated acyl silane was prepared by such a route l8 as was the only example of an allenic acyl silane (from a l-trimethylsilyl- 1-trimethylsilyloxy- 1,2,3-alkatriene).I9 Reich uses ethoxyethyl allenyl ethers (35) as precursors to a$-unsaturated acyl silanes, as the relatively large and polar protecting group gives them much better handling characteristics than simpler analogues (Scheme 40). Allenes such as (49) are also excellent precursors to a variety of x-substituted a$-unsaturated acyl silanes in good yields, as shown in Scheme 41.15 As a further development of this work the synthesis of acetylenic acyl silanes (50) has been achieved as shown in Scheme 42a.15 Oxidation of the selenium substituted allenyl ethers (51) with meta-chloroperbenzoic acid at -78 "C gave ''J.A. Miller and G Zweifel, J. Am. Chem. SOC.,1981, 103,6217. 98 D. J. Ager, Synthesis,1984, 384 Synthesis and Chemistry of Acyl Silanes Scheme 40 R (49) Br2/CH2C1,or Phsecl/CH2C12, or S02C1,/ thf R/8/4nU, X X = Br, PhSe, C1 60-80% Scheme 41 pLA:'OEt R' MCPBA 0 1OEt Scheme 42a the corresponding unstable selenoxides which underwent an in situ [2,3] sigmatropic shift to give acetals (52).Loss of selenenyl ester on work-up gave the acetylenic acyl silanes (50) in approximately 50% yields. A limited number of functionalized acyl silanes have been prepared by the use of 1,3-dithianes as described above; 45*99 however, this route is unfavourable for the synthesis of +unsaturated acyl silanes.I6 Analogous acetylenic 1,3-dioxanes 99 R. L. Danheiser and D. M. Fink, Tetrahedron Lett., 1985,26,2513. Page, Klair, and Rosenthal have however been used as precursors to a$-acetylenic acyl silanes,loo as have formyl silanes [Section 3A(i)].'' (v) Horner-Emrnons Reactions. Danheiser has successfully employed the Horner- Emmons reaction of a-phosphonoacyl silanes to prepared a$-unsaturated acyl silanes in 5697% yields (Scheme 42b).89 The a-phosphonoacyl silane inter- mediates, prepared from a-iodoacyl silane (53a) via the Arbuzov reaction, were shown to undergo enolate alkylation, for example using potassium t-butoxide and methyl iodide; the alkylated products also underwent Horner-Emmons re-action. 0 Scheme 42b E.Cyclopropyl Acyl Si1anes.-Cyclopropyl acyl silanes were first prepared from a$-unsaturated acyl silanes by treatment with diazomethane followed by vapour-phase pyrolysis of the intermediate pyrazoline derivatives (Scheme 63, section 4D)." They are cleaved or rearranged by acid under more mild conditions than their carbon analogues."' In one alternative route, Frei has successfully employed the Wittig rearrange- ment of allylic silyl ethers followed by Simmons-Smith cyclopropanation and Collins oxidation to produce cyclopropyl acyl silanes in 1&85% yields (Scheme 42~).~*Particular success was achieved using geraniol as the substrate.01. CqIp Zn-Hg II 2.AcOH 18% Scheme 42c Nakajima has shown that cyclopropyl acyl silane (53b) may be prepared by reaction of 1-lithio- 1-trimethylsilyl cyclopropanes with dichloromethyl methyl loo K. J. H. Kruithof, R. F. Schmitz, and G. W. Klumpp, J. Chem. Soc., Chem. Commun., 1983,239; K. J. H. Kruithof, R. F. Schmitz, and G. W. Klumpp, Tetrahedron, 1983,39,3073. lo' T. Nakajima, H. Miyaji, M. Segi,and S. Suga, Chem. Lett., 1986, 181. Synthesis and Chemistry of AcyI Silanes ether at low temperature in THF solution, a reaction which is said to involve a carbene intermediate and a 1,2-silicon shift (Scheme 42d).Io2 0 (53b)Scheme 42d A much superior route, successful as a general synthetic method for acyl silanes, involves treatment of acid chlorides with lithium tetrakis-(trimethylsilyl) aluminium or lithium methyl tris-(trimethylsilyl) aluminium and cuprous cyanide [Scheme 7, Section 3A(ii)].65 For example, cyclopropyl acyl silane (53b) was formed in 89%yield using this procedure.Recent work by Danheiser '03 has explored cyclopropyl acyl silane generation from a-haloacyl silanes via McCoy reactions (Scheme 42e) and via sulphur ylids (Scheme 42f). Ylid species such as (53c) were found to be stable in aprotic solvents in the presence of lithium salts and were used effectively for the cyclopropanation of a,P-unsaturated aldehydes. 75-89% Scheme 42e 00 1.Me2S, rt, 24h Bra SIM~:BU 2.NaOH, H20-CH$l2 SiMe,'Bu* u2s'd (53dScheme 42f 4 Reactions of Acylsilanes A. Simple Acylsi1anes.-Acyl silanes, although sensitive to light and to basic media, frequently behave as typical ketones when treated with a wide variety of reagents (Scheme 43).5*'04-'07 However, they also exhibit abnormal behaviour involving rearrangements leading to silicon-oxygen bond formation, especially when treated with nucleo- philic reagenk5v5 5,104, 'O5 Acyl silanes are extremely sensitive to basic conditions; for example, alcoholic solutions of benzoyl triphenyl silane containing a trace of aqueous base lo2 T Nakajima, H Miyaji, M Segi, and S Suga, Chem Left,1986,177 J S Nowick and R L Danheiser, Tetrahedron, 1988,44,4113 lo4 I Fleming, in 'Comprehensive Organic Chemistry', Vol 3, ed N Jones, Pergamon Press, Oxford, 1979,647 lo' P D Magnus, T Sarkar, and S DUJUriC, in 'Comprehensive Organometallic Chemistry', Vol 7, Pergamon Press, Oxford, 1982,631 E W Colvin, in 'Silicon in Organic Synthesis', Butterworths, London, 1981 lo' E W Colvin, in 'Silicon Reagents in Organic Synthesis', Academic Press, London, 1988 Page, Klair, and Rosenthal 0 LMH4 Ph1PhKSlW3 77% SIMe, 0 ASIPh, TsNHNH2 h 78% 0 NNH,N2H4 h PhK SIPh, 60% PhASiPh, LDA, Me,SiCl Jslhb3 82% Scheme 43 rapidly produce triphenylsilanol and benzaldehyde.'9' At least three mechanisms are conceivable for this reaction: SN2 displacement at the silicon atom [Path A]; nucleophilic attack at the carbonyl carbon atom followed by Brook rearrange- ment [Path B], initially to give a hemiacetal (54); or alternatively, nucleophilic attack at the silicon atom to form a pentacoordinate silicon anionic intermediate (59, followed by migration of the nucleophile to the carbonyl group [Path C] and subsequent Brook rearrangement as described for [Path B] (Scheme 44). 0 0 H20p-OH _I 9--Ph tSiPh, Ph,SIOH + Ph Ph OSlPh3 OSIPh, 0 -Ph4--phAH -Ph OH OH 'OH (54) Ph3SIOH 0 -A0 ---Ph 4H CPhASc 'OH Ph SIPh3 PhASIP, I OH Ph,SIOHOH (55) Scheme 44 The 1,2-migration of silicon to an oxygen anion formed by nucleophilic addition to a carbonyl group'95,108 (Brook rearrangement) is a very common lo* A.G.Brook, Acc. Chem.Res., 1974,7 77. 177 Synthesis and Chemistry of Acyl Silanes pathway by which acyl silanes react when treated with nucleophiles, the major driving force probably being formation of the strong Si-0 bond. Brook has shown by using various optically active acyl silanes that this rearrangement usually occurs with retention of configuration at the silicon atom (vide infra).lo' The stereochemical course of the Brook rearrangement and that of a multitude of other nucleophilic additions to silicon atoms can be accounted for if a pentacovalent trigonal bipyramidal intermediate is involved in the substitution proces~.~~~~~'~An intermediate is of course possible in the silicon series, where it is not in the carbon series, because the empty d-orbitals on silicon are just about low enough in energy for bonding to be profitable.In the simplest case, inversion of configuration is observed when the intermediate is formed (56) --+ (57) and decomposes (57)---(58) without pseudorotation taking place (Scheme 45). V Y 6 I*,,,pcx %, i -Y -.("I"-"-*4slF I B X X The base-catalysed solvolysis of acyl silanes has been studied in detail by Ricci, who, following various kinetic measurements, has suggested that the probable reaction pathway involves direct attack of hydroxide ion at the carbonyl group (Scheme 44, Path B), the rate determining step being the migration of the trialkylsilyl group from carbon to oxygen.' '' The reaction of acyl silanes with alkoxide ions has been studied in great detail (Scheme 46)."*3' l3 Again, the reaction pathway was rationalized by invoking a OSIR2OR"SIR20R" -I-R' R' R RR Scheme 46 I Fleming, in 'Comprehensive Organic Chemistry', Vol 3, ed N Jones, Pergamon Press, Oxford, 1979,554 'Io R J P Corriu and C Gurein, Adv Organomet Chem ,1982,20,265 ''I D Rietropaolo, A Ricci, M Taddei, and M Fiorenza, J Organomet Chem , 1980,197,7 'I2 A G Brook and N V Schwartz, J Org Chem , 1962,27,2311 'I3 A G Brook, W Limburg, and T S D Vandersar, Can J Chem ,1978,56,2758 Page, Klair, and Rosenthal nucleophilic attack by alkoxide ion at the silicon atom of the ketone giving a pentacoordinate silicon anionic species (59), which can then suffer 1,2-migration of an alkyl group from the silicon atom to the carbonyl carbon atom to give the alkoxide ion (60).The intermediate (60) then undergoes a Brook rearrangement to yield the unsymmetrical dialkoxy silane (61), usually the major product. Other reaction products such as alcohol (62) and dialkoxy silane (63), arise from a transetherification reaction between the alkoxide ion and the unsymmetrical dialkoxy silane. A competing reaction, corresponding to nucleophilic displace- ment of the acyl group from the silicon atom, is also observed. This displacement reaction becomes favoured over rearrangement as the polarity of the solvent system increases (Scheme 47).53' l4 R' 'd'SiR3 R3SIOR" + R"-Scheme 47 Later elegant work by Brook l1 using t-butoxide ion and optically active acyl silanes led him to suggest that the cleavage products arise from direct attack of the butoxide ion at the carbonyl group followed by Brook rearrangement.Evidence for this proposal is outlined in Scheme 48.Should cleavage arise from (64) uia attack at silicon, then the t-butoxy silane (65) would be formed with overall inversion of configuration at silicon relative to (66). Reduction of (65) would lead, with retention of configuration, to the (+)-silane (67) (Path B), although experiments have shown that under the reaction conditions employed by Brook this reduction is at best very slow.Similar rearrangements of alkyl acyl silanes have been observed upon treatment with fluoride ion.' (66) (68) (69) Nap = I-naphthyl Scheme 48 E. D. Hughes and C. K. Ingold, J. Chem. SOC.,1935,244. P. C. B. Page, S. Rosenthal, and R. V. Williams, Tetrahedron Lett., 1987,28,4455. Synthesis and Chemistry of Acyl Silanes Conversely, were nucleophilic attack of the alkoxide ion at the carbonyl group of (66) occurring, then the ion formed, (68), should rearrange to (69) with retention of configuration at silicon. Reduction of (69) with lithium aluminium hydride would then produce (-)-naphthyl phenyl methyl silane (70). When the reaction was carried out only the (-)-silane (70) was isolated from the cleavage products of the reaction mixture, suggesting that cleavage occurred via nucleophilic attack of the alkoxide ion at the carbonyl group (Path A).No t-butoxysilane was detected among the products of reduction. Brook does not discuss the possibility of pseudorotation of (64), which could lead to the ether (65) being formed with overall retention of configuration at the silicon atom (see Scheme 45). However, the fact that no t-butoxysilane could be isolated from the reduction products suggests that cleavage probably does arise via path A. There is also the possibility of a migration of the t-butoxide group from the silicon atom to the carbonyl carbon atom as shown in Scheme 49, a sequence which cannot be disproved by these experiments. However, such a migration of the butoxide moiety in (64) seems unlikely as it requires the strong silicon-oxygen bond to be broken.0 0 II CH, I Scheme 49 Acyl silanes display a range of behaviour when treated with carbon nucleo- ’9’philes.’9’ O4 For example, on the one hand, when a variety of aryl acyl silanes (71) were reacted with an alkylidene phosphorane, none of the expected alkenes were obtained, and the only reaction products found were silyl enol ether (72) and triphenylphosphine (Scheme 50).’ ’6,1’ 0 Ph3P+-C%-Ar KS,R3 Ar4?iR3 -Ar r (711 (72)v+Ph3 Scheme 50 On the other hand, when alkyl acyl silanes (73a) were reacted with Wittig reagents only the normal olefinated vinyl silane products were isolated (Schemes 51a,b).’16v1 l7 Under soluble lithium salt conditions 2-vinyl silanes were produced ’‘‘ A.G. Brook and S. A. Fieldhouse, J. Organomet. Chem., 1967,10,235. ‘17 J. A. Soderquist and C. L. Anderson, Tetrahedron Lett., 1988, 29, 2425; J. A. Soderquist and C. L. Anderson, Tetrahedron Lett., 1988,29,2777. Page, Klair, and Rosenthal 0 0Ph3P+-C%-*&P+Ph, SIPh, SIPh3SiPh,(73a) Scheme 51a SIMe,-PhsPCH(CH*)&H=CH* LiI J J ;\ OAC (73b) Scheme 51b with very high selectivity; this route was used to prepare a true pheromone component of the sweet potato leaf folder moth (73b) (Scheme 51b).'17 Diazomethane reacts in a similar manner to Wittig reagents,' suggesting that if the substituent at the acyl group is alkyl, and hence carbanion-destabilizing, the rearrangement is inhibited and generally does not occur at all, or at best is very slow relative to the alternative 'normal' reaction pathway; however, when the substituent is aromatic, and therefore capable of stabilizing incipient carbanion formation as the silicon-carbon bond cleaves, rearrangement occurs readily, and silyl ether product is predominant or exclusive.Kuwajima has used the Brook rearrangement to great effect by using acyl silanes as homoenolate equivalents (Scheme 52).89' l8Vinyl Grignard reagents react with acyl silanes to give intermediates (74) which subsequently undergo Brook rearrangement to give the homoenolates (75).One important side-reaction is that (75)can undergo a 0 (75) (76) OSiMe, I.Kuwajima, T. Matsutani, and J. Enda, Tetrahedron Lett., 1984, 25, 5307. 181 Synthesis and Chemistry of Acyl Silanes 0 H (79) + 0--20 "C t1,2=90m Ph CHZPh (80) Scheme 53a further, irreversible, 1,4-silyl group shift to produce the enolate of a P-trimethylsilyl ketone (76). This side-reaction can be partially suppressed by keeping the temperature low, by using larger alkyl groups attached to the silicon atom, and by using magnesium instead of lithium enolates. The difference observed between lithium and magnesium homoenolates may be due to the less ionic character of the carbon-magnesium bond reducing the propensity for attack at the silicon atom to form (76). Sila-p-ionone (77a), an intermediate for sila-vitamin A synthesis, has been prepared via addition of an acetylenic Grignard reagent to a cyclic acyl silane.' '' Kuwajima was able to form cuprate reagents from (75) by using copper trimethylsilyl acetylide.These reagents perform various conjugate addition reactions with enones to provide 1,6-dicarbonyl compounds (77b) in good overall yields. a-Phenylthioacyl silanes (78) give silyl enol ethers with very high stereo-selectivity when reacted with organolithium reagents (Scheme 53a).12' The major diastereoisomer formed in the addition reaction is the erythro isomer (79); this is perhaps best rationalized by invoking a Felkin-Anh transition state (PhS group anti to attacking nucleophile).121 The alcohols rearrange uia a Brook migration with concerted expulsion of the phenylthiolate leaving group; because of the stereo- electronic demands of the reaction, the silyl group must be eclipsed with a hydro- gen atom during the carbon to oxygen migration in the major diastereoisomer (79) and with a benzyl moiety in the minor, less reactive diastereoisomer (80).Recently acyl silanes containing chiral centres at the a-carbon atom have been shown to undergo highly stereoselective addition of organolithium and Grignard reagents; the resulting cr-hydroxy silanes, which could be protiodesilylated with >99% retention of configuration, being formed in 39-89% yield and with diastereoselectivity up to > 100:1.'22 '19 R. Miinsted and U. Wannagat, J. Organomet. Chem., 1987,322, 11; R. Miinsted and U.Wannagat, Monatsh. Chem., 1985,116,693. lZo H. J. Reich, R. C. Holtan, and S. L. Borkowsky, J. Org. Chem., 1987,52,312. M. Cherest, H. Felkin, and N. Prudent, Tetrahedron Left.,1968, 2199; N. T. Anh and 0.Eisenstein, Now. J. Chim., 1977, 1,61; A. S. Cieplak, J. Am. Chem. SOC.,1981, 103,4540. lZ2 M. Nakada, Y. Urano, K. Susumu, and M. Ohno, J. Am. Chem. SOC.,1988,110,4826. Page, Hair, and Rosenthal Buynak has successfully exploited the steric bulk of the silyl group in some acyl silanes to effect asymmetric reduction of the carbonyl group using the Itsuno reagent 123 [a 2:1 complex of borane and (S)-C-l-2-amin0-3-methyl-l,1-diphenylbutan-1-01] (Scheme 53b). 124 Transformation of the resultant alcohols into the products via a thermal silicon-to-carbon migration of a phenyl group was found to occur stereospecifically to give products of high enantiomeric purity.12' OACItsuno OH Ac20, PYI reagent _____t E-RASiPh3 DM RxSIP,I95% * H202,KF !h R'Ph = R ASiPh,OAc R=Me, 50%(95%ee) Scheme 53b Acyl silanes react with bis-(trimethylsilyl) sulphide in the presence of a CoC12.6H20 catalyst to afford the corresponding thiocarbonyl derivatives (Scheme 53c).126 The reaction is mild and proceeds in good yields and, interestingly, is also applicable to aldehydes.Scheme 5% Acyl silanes have very recently been used by Schinzer as reagents for diastereoselective aldol condensations. ' The acyl silanes, prepared using the method of Soderquist,12 were treated with LDA followed by aldehydes to give aldol products in up to >20:1 selectivity in favour of the syn products.Product mixtures were analysed as the carboxylic acids (Scheme 53d). For R' =methyl no selectivity was observed. However, increasing steric bulk at the silicon atom gave superior results. Kuwajima has used acyl silanes as aldehyde equivalents in Lewis acid- catalysed aldol reactions between silyl enol ethers of acyl silanes and acetals. The resulting P-alkoxy acyl silanes were treated with tetrabutyl ammonium hydroxide or tetrabutyl ammonium fluoride to give the corresponding a$-unsaturated al-dehydes.l 123 S. Itsuno, M. Nakano, K. Miyazaki, H. Masuda, K. Ito, A. Hirao, and S. Nakahama, J. Chem. SOC., Perkin Trans. I, 1985,2039. 124 J. D. Buynak, J.B. Strickland, T. Hurd, and A. Phan, J. Chem. SOC.,Chem. Commun., 1989,89. 12' A. R. Bassindale, A. G. Brook, P. F. Jones, and J. M. Lennon, Can. J. Chem., 1975,53,332. 126 A. Ricci, A. Degl'Innocenti, A. Capperucci, and G. Reginato, J. Org. Chem., 1989,54, 19. T. Sato, M. Arai, and 1. Kuwajima, J. Am. Chem. SOC.,1977,99,5827. 183 Synthesis and Chemistry of Acyl Silanes JsiR13 2.R2CH01.LDA, -78 "C R2qc02H + R2 - - 3. H202 syn :anti R'=Et, "Pr, 'BuMe, R2=Ph 4:l t09:l R2='Pr 201 Scheme 53d Wilson has found that acyl silanes may be used as synthetic equivalents of sterically-hindered aldehydes.I2*Coupling of aldehydes such as (119) and (120) with 3-methylpentadienyl lithium gave rise to a mixture of regioisomers.However, when analogous acyl silanes (121) and (122) were used, only the conjugated isomers were formed (Scheme 53e). This selectivity is similar to that found in the reaction of sterically-hindered ketone (123). The silyl carbinol adducts are easily desilylated via a Brook rearrangement to give the correspond-ing alcohols. I(119),R=H (120),R=allyl I KH, HMPAt (121), R=H (1221, R=allyl Scheme 53e Adduct (124) underwent a highly diastereoselective intramolecular Diels-Alder reaction to give alcohol (125) after stereospecific desilylation (Scheme 53f). The authors comment that acyl silanes provide higher overall yields than aldehydes in these reactions as they are less prone to self-condensation and are alsopsim3OH OHpsi2=p%o" (124) (125) Scheme 53f 12'S.R.Wilson, M. S. Hague, and R.N.Misra, J. Org. Chem., 1982,47, 747. Page, Klair, and Rosenthal superior substrates in that the bulky silyl group may be used for stereocontrol of subsequent reactions. B. a-Haloacyl Manes.-a-Haloacyl silanes react with Grignard reagents in a rather unusual manner (Scheme 54).'4.'06*107 The expected alkoxide (8 1) is generated initially, and the reaction then proceeds through loss of chloride ion from the alkoxide and concomitant 1,Zrearrangement of the trimethylsilyl group to afford the product (82). a-Haloacyl silanes therefore behave as a-trimethylsilyl acylium ion (83) equivalents in this reaction. Initial nucleophilic attack at the carbonyl group of the acyl silane may be facilitated compared with a-chloro- ketones by 0--71interaction with the neighbouring silyl group, while the 1,2- rearrangement may be accelerated compared with the lithium analogue by the more polar character of the oxygen-magnesium bond.' 29,1 30 0 CI If the Grignard reagents possess P-hydrogen atoms, then P-hydroxyalkyl trimethylsilanes (84), rather than P-ketoalkyl trimethylsilanes, are formed with high diastereoselectivity (Scheme 55).'4.128 The hydroxy silanes can then undergo stereocontrolled elimination of trimethyl- silanol to give alkenes (Scheme 56).a-Haloacyl silanes have also been used as a-trimethylsilyl acylium ion equivalents in their reactions with enolates (Scheme 57).I3l R'CHZCHzMgX/ (major isomer) SIMe3 (84) Scheme 55 N.de Kimpe, P. Sulmon, and N. Schamp, Angew. Chem., Int. Ed. Engl., 1985,24,881. 130D. J. Cram and F. A. Ald Elhafez, J. Am. Chem. SOC.,1952,74,5828. 13' I. Kuwajima and K. Matsumoto, Tetrahedron Lett., 1979,4095. 185 Synthesis and Chemistry of Acyl Silanes KH BF3.Et20 R' R SIMe3 (EwScheme 56 Scheme 57 Reich has utilized the Brook rearrangement in a-halo-a$-unsaturated acyl silanes to synthesize a series of silyloxyallenes (85), from which were synthesized a number of sesquiterpenes (Scheme 58).' X Although Reich has also prepared silyloxyallenes by the alkylation of silyloxy- allenyl lithium reagents, he favours the above scheme as it is less sensitive to solvent effects and other experimental parameters. An outline of the synthesis of dehydrofukinone (86) which elegantly exemplifies this methodology is shown in Scheme 59.a-Iodoacyl silanes have been used as precursors of a-phosphonoacyl silanes, which underwent successful enolate alkylation, and from which were prepared a,P-unsaturated acyl silanes uia Horner-Emmons reaction [Section 3D(v)]; 89 a-haloacyl silanes have also been used as precursors of cyclopropyl acyl silanes via McCoy reactions and via conversion into sulphur ylids (Section 3.5).'03 C.a-Ketoacyl Manes.-a-Ketoacyl silanes are a deep rich red in colour and are especially sensitive, necessitating that purification and handling be carried out in the dark. Little chemistry of these interesting and very sensitive materials has been reported, although they have been implicated as intermediates in the Page, Klair, and Rosenthal PhS02-/ OSIMe2Ph CI JI MeLi; H,O 5steps Et2AlC1-P 0 (86) Scheme 59 oxidation of trimethylsilyl acetylenes with osmium tetroxide to give m-ketoesters (Scheme 60).132 R-SiMe3 Me,NO,OS0, _[ Rfis,m] OR-R& R'OH 0 0 Scheme 60 D.a#-Unsaturated Acylsi1anes.-a$-Unsaturated acyl silanes, yellow or yellow-green in colour, serve as highly reactive carboxylic acid equivalents in conjugate NaOH/ H202 / ;jT""SiMe, (87)R' Scheme 61 allylation reactions with ally1 silane derivatives (Scheme 6 1). a$-Unsaturated acyl silanes are much more electrophilic than the corresponding carboxylic acids 132 P C.B. Page and S. Rosenthal, Tetrahedron Lett., 1986,27, 1947. Synthesis and Chemistry of Acyl Silanes and esters due to the net destabilizing effect of trialkylsilyl groups on a-carbocationic centres in intermediates such as (87).32 They are oxidized to carboxylic acids by alkaline peroxide.97 a,P-Unsaturated acyl silanes combine with allenyl silanes in [3 + 21 and [3 + 31 annelation reactions to give five and six-membered carbocycles (Scheme 62a).68,89,99,101-103 TiC1, SIR', -78 OC Scheme 62a By manipulating the acyl silane trialkylsilyl group, the reaction temperature, and the nature of the acyl group, the course of these reactions may be controlled to produce either five or six-membered rings as desired. The reaction pathway is outlined in Scheme 62b. Regiospecific electrophilic substitution at C-3 of the allenyl silane provides a vinyl cation (88) which undergoes a 1,2-cationic trimethylsilyl shift to afford an isomeric vinyl cation (89).Cyclization then provides the cyclopentene (90). If R' is alkyl, R3Si is trimethylsilyl, and the reaction is carried out at elevated temperature, (90) can undergo a further ring expansion to give the six-membered carbocycle (9 l), which undergoes a second 1,2-cationic shift of the trimethylsilyl moiety to produce the cyclohexenone (92). If desired, this further transformation can be prevented simply by employing the less mobile t-butyldimethylsilyl acyl silanes, minimizing the reaction time, and maintaining the reaction temperature below -78 "C.1,3-Dipolar cycloaddition of diazomethane to acyl silane (93) afforded a pyrazoline derivative which was subjected to vapour phase pyrolysis to produce the first recorded cyclopropyl acyl silane (94) in 44% yield (Section 3E).99 Exposure of (94) to one equivalent of titanium tetrachloride in dichloromethane (-78 "C-+ 0 "C,1 h) provided the cyclobutanone (95) in 75% yield (Scheme 63). The Diels-Alder reactivity of a$-acetylenic and ethylenic acyl silanes is comparable to that of the related methyl ketones, and these reactions can be used to prepare other useful a,P-unsaturated acyl silanes. The acetylenic acyl silanes (96) and( 97) react with 2,3-dimethylbuta- 1,3-diene and 4-phenyloxazole to give Page, Klair, and Rosenthal (92) OTiCl3 me351 : +3x (90) (89)Scheme 62b (93) (94) (95) Scheme 63 Synthesis and Chemistry of Acyl Silanes (98) and (99) respectively, under conditions similar to those required for more conventional acetylenic dienophiles (Scheme 64). The cycloaddition of selenium-substituted a$-unsaturated acyl silane (100) with 2,3-dimethylbuta-l,3-dieneshows an unusual effect in that a significant portion of the hetero-Diels-Alder adduct (101) is formed.It seems that the phenyl selenide substituent is responsible for this unusual reactivity, since the vinyl acyl silane (102) gives only the expected regioisomer (103) (Scheme 65).15 (loo), X=PhSe (103) (101)(102),X=H X=PhSe, 85:15 Scheme 65 a-Substituted a,P-unsaturated acyl silanes such as (100) have also been used to synthesize a series of substituted dienes in excellent yield (Scheme 66).l5 0 n Scheme 66 E.Acyl Silanes as Acyl Anion Precursors.-Various aromatic and heterocyclic acyl trimethylsilanes have been used as acyl anion equivalents by treatment with fluoride ion (Scheme 67, path A).22*97,' 1591333134 X=O,S,NMe 133 C. H. Heathcock and D. Schinzer, Tetrahedron Lett., 1981,22,1881. '34 A. Degl'Innocenti, S. Pike, D. R.M. Walton, G. Seconi, A. Ricci, and M. Fiorenza, J. Chern. Soc., Chern. Cornmun., 1980,1201. 190 Page, Hair, and Rosenthal Provided that the acyl substitutent is electron-withdrawing and there are no aryl substituents on the silicon atom, the acyl anion can be trapped by a variety of electrophiles in good to moderate yield, indeed pentacoordinate silicon anionic species and acyl anions have both been detected in gas-phase reactions of acyl silanes with fluoride ion.' 35 One interesting variation is that the bis-(trimethyl- silyl) ketone (4) may act as a source of the dianion C02-in the presence of fluoride ion (Scheme 68).23 60% Scheme 68 An alternative rearrangement pathway may be observed when simple alkyl acyl silanes are used or when the silicon atom bears aryl substituents.This pathway is similar to that suggested, but not observed, by Brook for reaction of acyl silanes with alkoxide ions.112*"3 Addition of fluoride ion induces a migration of one of the alkyl or aryl groups attached to the silicon atom to the carbonyl carbon atom to give (104) followed by a Brook rearrangement giving a rearranged alcohol after protic work-up (Scheme 67, path B).'" The acyl anion reaction pathway may only be observed for these substrates at higher tempera- tures in the presence of acid.Both pathways may proceed uia a pentacoordinate silicon anionic species as a common intermediate. F. Photochemistry.-The interesting cyclic acyl silane 1,l-diphenyl silacyclohexa- none (105) has been found to be unstable towards light, especially in the presence of oxygen. Subsequently (105) was shown to undergo photooxidation promoted by ambient light to produce the silicon-containing lactone (106). The lactone was subsequently hydrolysed to give a &(hydroxysilyl) carboxylic acid (107) (Scheme 69).7 Compound (105) was stable in the presence of oxygen in the absence of light over long periods. The interesting photochemistry displayed by acyl silanes has been attributed to 13' C. H. de Puy, V. M. Bierbaum, R. Damrauer, and J. A. Soderquist, J. Am. Chem. SOC.,1985, 107, 3385. 191 Synthesis and Chemistry of Acyl Silanes the low-energy n -+ n* carbonyl group transition; subsequent work has shown that this photochemical oxidation is typical of other alkyl acyl silanes, but that aryl acyl silanes are inert.5 Acyl silanes also react with remarkable facility in alcoholic solution in the presence of near-visible radiation.'36 In the absence of base the reaction process Scheme 69 involves cleavage of the acyl-silicon bond to give a silyl ether and an acetal.Silanol and aldehyde may also be isolated. The probable mechanism, which was suggested following experiments involving an optically active acyl silane, is shown in Scheme 70.5 Quite apart from the above process is the near-quantitative formation of mixed acetal which occurs on photolysis of an alcoholic solution of an acyl silane containing trace amounts of base (typically ~yridine).~ This acetal is formed by the photochemical generation of a silyloxycarbene (108) from the acyl silane which inserts into the OH bond of a solvent molecule (Scheme 71). Scheme 71 Dalton has examined the above reaction in detail and has indeed found from various kinetic measurements that acetal formation occurs exclusively via reaction of the alcohol with an intermediate presumed to be the silyloxycarbene (108), formed from the acyl silane TI state.'37 Silyloxycarbenes are also formed on heating arylsilanes and have been trapped in an intramolecular fashion as shown in Scheme 72.13' A G Brook and J M Duff,J Am Chem SOC,1967,89,454 13' R A Bourque, P D Davis, and J C Dalton, J Am Chem SOC,1981,103,697 138 A R Bassindale,A G Brook, and J Harris, J Organomet Chem , 1975,90, C6 Page, Hair, and Rosenthal 0 R Scheme 72 Swenton attempted to utilize an intramolecular insertion reaction of silyloxycar- bene generated from an acyl silane as a route to benzocyclobutenols (109) (Scheme 73).'j9 However, the unstable compound (1 10) underwent ring opening (109) Scheme 73 and further rearrangement as shown above to give the aldehyde (1 11) in good yield.Upon n -n* excitation, the acyl silane (1 12) undergoes a Norrish type I1 reaction as the major pathway, involving hydrogen abstraction and fragmentation to give the acyl silane (113) and the diene (114) as the major products (Scheme 74).'02 The ketone (115) was shown to behave in an analogous manner. Additionally, acyl silane (1 12) showed more typical photochemical behaviour, undergoing rearrangement to the silyloxycarbene intermediate (1 16). Insertion of (116) into the OH bond of the enol (117) led to compound (118) (Scheme 75). 13' C Shih and J. S. Swenton, J Org Chem., 1982,41,2668. Synthesis and Chemistry of Acyl SiIanes These findings demonstrate that the silyloxycarbene (116) reacts preferentially by an intermolecular insertion into an OH bond rather than by an intramolecular OH ASitBuMe2 (117) Brook has shown that upon irradiation a wide variety of acyl tris-(trimethylsilyl) silanes and acyl alkyl bis-(trimethylsilyl) silanes undergo clean 1,3-rearrangements of silyl groups from silicon to oxygen to give silenes, many of which are remarkably stable and even recrystallizable (Scheme 76).26,27y29.63These silenes hv OSIMe, hv SiMeROSiMe, R' KSiR(SlMe3), R' ASIR(SIMB3) ReASiMe2 Scheme 76 194 Page, Klair, and Rosenthal may undergo head-to-head or head-to-tail [2 + 2) dimerization to give 1,2- or 1,3-disilacyclobutanes dependent upon the nature of the alkyl groups pre~ent.~' The structures of several 1,3-disilacyclobutanes have been determined by X-ray crystallography.140 5 Conclusion We hope that we have shown in the above discussion that early difficulties encountered in synthesizing acyl silanes have been surmounted to a large extent by the multitude of synthetic methods now available for their preparation. In recent years several reactions of acyl silanes, for example the Brook rearrange- ment, have been utilized to great effect in forming a variety of useful synthetic intermediates. As the synthesis of acyl silanes becomes more sophisticated, and as functionalized acylsilanes become more accessible, it seems likely that these interesting materials will find increasing use in organic chemistry. I4O K.M. Baines, A. G. Brook, P. D. Lickiss, and J. F. Sawyer, Organometallics, 1989,8, 709. 195

ISSN:0306-0012    

DOI:10.1039/CS9901900147

出版商:RSC    

年代:1990

数据来源: RSC