6 Organometallic Chemistry Part (ii) Main-group Elements By K. SMITH Departmentof Chemistry University College of Swansea Swansea SA28PP 1 Introduction As in the previous Annual Report on main-group organometallic compounds,' B Si and As are included but not P Se and Te. In this the last Report from this author it seems appropriate to draw attention to a worrying trend in a nevertheless fascinating area of organometallic chemistry concerning the investigation of reactive intermediates and of novel structural types. A mechanistic rationalization accompanying a publication which is entirely justifi- able for say its synthetic utility causes little concern. On the other hand in main group organometallic chemistry in recent years there have appeared many papers whose very justijkation lies in the involvement of interesting intermediates or novel structural types but which contain little or no evidence for such species.Yet because these proposed species would indeed be highly interesting the publications cannot be ignored and warrant reference here. Fortunately not all of the publications concerning such species are of this type. An example is in the study of multiply-bonded Si compounds an area which has given some cause for concern in the past ' but which is now advancing. Three groups have independently studied the i.r. spectra of sila-alkenes isolated in Ar matrices,* clearly the outstanding method for investigation of such species. The Soviet group investi- gated 1,l-dimethylsila-ethene formed by pyrolysis of 1 1-dimethylsilacyclobutane whilst both American groups investigated 1,1,2-trimethyIsila-etheneproduced by photolysis of trimethylsilyldiazomethane.At the time of writing only the abstract of the Soviet paper has been seen but as far as can be ascertained from the three bands listed therein there appears to be fair agreement between the spectra of the two products. A band at 643 or 645cm-' may be assigned to an olefinic C-H deformation mode but no emphasis is placed on bands near 1410 cm-' the apparent region of the Si=C stretching mode according to the only previous report of an attempt to obtain an i.r. spectrum of a ~ila-alkene.~ It appears that the earlier assignment has now been withdrawn so it is especially interesting how accurately a K.Smith Ann. Reports (B),1975,72 136. (a) A. K. Mal'tsev V. N. Khabashesku and 0.M. Nefedov Izvest. Akad. Nauk S.S.S.R. Ser. khim. 1976,1193 (Chem.Abs. 1976,85,122861); (b) 0.L. Chapman C.-C. Chang J. Kolc M. E. Jung J. A. Lowe T. J. Barton and M. L. Tumey J. Amer. Chem. SOC.,1976 98 7844; (c) M. R. Chedekel M. Skoglund R. L. Kreeger and H. Schechter ibid. p. 7846. T. J. Barton and C. L. Mclntosh J.C.S. Chem. Comm. 1972,861. 121 122 K,Smith computer 'predicted' this earlier freq~ency.~ Also it is claimed5 that a band at 1315 cm-' in the i.r. spectrum of (l) if attributable to the Si=C mode would be inconsistent with a band around 1410 cm-' for free Si=C bonds. New approaches to sila-imines (2) via photolysis or pyrolysis of silyl and to substituted sila-alkenes (3) by photolysis of unsaturated disilanes (4; X =0 or CH2),7 have been reported.The species (3; R' = Me R2= Me3Si R3=Ph) shows unusual behaviour; in the absence of added trapping agents it forms a head-to-head rather than a head-to-tail dimer.7a SiMe,(-Fe(COI3SiMe R2Si=NR XSiR /R$Si=C HXR Si .Si(Rz)C \R3 \R3 (1) (2) (3) (4) Last year we highlighted' an apparently significant report of the production at low temperatures of a relatively stable silicenium ion (5). Fortunately some people working in the organosilicon field were not so easily misled and have shown' that the results obtained on reaction of (6) with Ph3C'CI0,- can be matched using other R3SiH species several of which have previously been widely believed nut to produce stable silicenium ions.After consideration of the lack of conductance by solutions of '(3,and examination of the spectral properties the conclusion is drawn that the supposed silicenium ion is merely a covalent perchlorate (or a tight ion-pair).' This case illustrates the dangers of drawing conclusions from a single piece of datum. Several highly interesting new Group I11 compounds have been claimed this year but again the data are not always conclusive. Amongst them are (7) (8),and (9),all supposedly produced by reactions of halogeno-derivatives with alkali metals. Compounds (7) and (8) were obtained by reaction of a suitably molten form of potassium with MeBCl in the presence of cycl~hexene.~ The authors admit the tentative nature of their structural assignments (based almost entirely upon accurate mass determinations of the supposed molecular ions which in one case did not even H.B. Schlegal S. Wolfe and K. Mislow J.C.S. Chem. Comm. 1975 246. H. Sakurai Y. Kamiyama and Y. Nakadaira J. Amer. Chem. SOC. 1976,98 7453. D. R. Parker and L. H. Sommer J. Amer. Gem. Soc. 1976,98,618; J. Organometallic Chem. 1976 110,c1. 7 (a)A. G. Brook and J. W. Harris J. Amer. Chem. Soc. 1976,98,381; (6)M. Ishikawa T. Fuchikami and M. Kumada J. Organometallic Chem.,1976. 117 C58. * J. B. Lambert and H. Sun J. Amer. Chem. SOC. 1976,98,5611; T.J. Barton A. K. Hovland and C. R. Tully ibid. p. 5695. S. M. van der Kerk J. Boersma and G. J. M. van der Kerk Tetrahedron Letters 1976,4765. Organometallic Chemistry -Part (ii) Main -group Elements agree with a classical molecular weight determination) but give no other justification for the claim to have generated 'methylborylene' (MeB:) the subject of the publication! The evidence'' for (9),produced by the action of K on BukAlCl comes from the production of DZ on treatment with MeOD [equation (l)].It is however possible to think of less obvious ways of obtaining Dz.BU~~AI-AIBU:MeOD) 4Bu'D +2AI(OMe)3+D2 (1) It would be fascinating if the structures (7) (8) and (9) were genuine for they would open up new areas for exploration but given only the evidence presented there must be considerable doubt. Indeed a study'' of the related reaction between BuZBCI and Na-K alloy did not allow a definitive structure to be assigned to the immediate product despite the availability of more data than are given for (7) (8) and (9).The data were at least consistent with an oligomeric borohydride (lo) which has little in common with (7) (8) or (9). With this in mind and in order to encourage definitive work aimed at genuine identification of the reaction products we predict that the compounds (7)-(9) will prove not to be those formed in the reactions described. To the first person to prove this prediction wrong one year's subscription to Annual Reports will gladly be provided.* OBMe BMe BU;AI-AIBU~L n K+-+BHB~~CHP++ (7) (8) (9) (10) Having commented at length on this worrying trend it is proper to proceed to a discussion of other highlights from the 1976 literature.There are several articles of general interest including reviews of the organic chemistry of metal vapours,12 and of metal a-hydrocarbyl~,'~ and a report of a method for estimating the pyrophoricity of metal alkyl~.'~ Volume 43 of Pure and Applied Chemistry contains useful reviews by Makosza (two-phase carbanion reactions) McKillop (TIIII nitrate in organic synthesis) Stork (kinetic enolates) Trost (novel alkylations) Kagan (asymmetric hydrosilylation) and others." 2 Group1 Lithium.-Reviews of the synthetic utility of 2-0xazolines~~ and of nucleophilic acylating agents1' have appeared. Various substituent groups direct lithiation of the aromatic nucleus to the 2-position. Studies of the position of lithiation of a series of 4-substituted methoxyben- lo H.Hoberg and S. Krause Angew. Chem. Internat. Edn. 1976,15,694. K. Smith and K. Swaminathan J.C.S. Dalton 1976 2297. l2 K. J. Klabunde Accounts Chem.Res. 1975,8,393. l3 P. J. Davidson M. F. Lappert and R. Pearce Chem. Rev. 1976,76,219. l4 W. L. Mudry D. C. Burleson D. B. Malpass and S. C. Watson J. fire Flammability 1975,478 (Chem. Abs. 1976,84 105680). Is M. Makosza Pure Appl. Chem. 1975,43,439; A. McKillop ibid. p. 463; G. Stork ibid.,p. 553; B. M. Trost ibid. p. 563; H. B. Kagan ibid. p. 401. l6 A. I. Meyers and E. D. Mihelich Angew. Gem. Internat. Edn. 1976,15 270. l7 0.W. Lever Tetrahedron 1976,32 1943. * Write directly to Dr. Smith -ed. 124 K. Smith zenes indicate that regioselective metallation still occurs and that the groups S02NRMe CONRMe (R= H or Me) and CH2NMe2 take priority over OMe whereas CH2CH2NMe2 NMe2 CF3 and F have less powerful ortho -directing abilities than 0Me.l' At low temperatures it is possible to prepare and use aryl-lithium compounds containing the CN group.'' Lithiated dimethylhydrazones are advocated as useful enolate equivalents (Scheme 1).20 When derived from unsymmetrical ketones they are lithiated highly regioselectively to yield anions which react readily with suitable electrophiles and can easily be converted into a-ketocarbonium ion equivalents.Dimethylhydrazones derived from aldehydes behave like the ketone derivatives unless there is branching at the a-carbon when removal of the aldehydic proton and formation of a nitrile may take over.With an optically active hydrazone the same method gives a-branched optically active ketones with very respectable optical purities.2' /NMe2 NMe2 / N N If II '' R3R'CH-C-CHR4 R3R'CHCN +(R2 1H)R3R'CH-c-R2 (R2= CH2R4f I Li /NMe2 NMe2 / 0 N N II II II R3R'CH-C-CHR4R5 ,V R3R'CH-C-CHR4R5 R3R1CH-C-CHR4 I SMe Reagents i LiNPr'z-HMPA-PhH; ii LiNPr'2-THF; iii (MeQ2; iv R'I; v NaI04-H20. Scheme 1 Dimethylthiopivalamide can be lithiated on one of the NMe groups22 to give a reagent which behaves as a synthetic equivalent to MeNHCH2-. Doubly deproton- ated nitroalkanes are also useful species.23 Purely aliphatic nitroalkanes lose both protons from the a-position and on treatment with electrophiles (E+),then water yield a-substituted products RCH(E)N02 whereas p -arylnitroalkanes lose one proton from each of the a-and @-positions to give species which behave as super-enamine derivatives reacting to give p -substituted products.Since nitro- compounds can readily be converted into amines or ketones the synthetic potential is clear. D. W. Slocum and C. A. Jennings J. Org. Chern. 1976,41,3653. l9 W. E. Parham and L. D. Jones J. Org. Chern. 1976,41 1187. 2o (a) E. J. Corey and D. Enders Tetrahedron Letters 1976 3; (6) T. Cuvigny J. F. Le Borgne M. Larcheveque and H. Normant Synthesis 1976,237; (c) E. J. Corey and S. Knapp TefrahedronLetters 1976,4687. D. Enders and H. Eichenauer,Angew. Chern. Internat. Edn. 1976,15,549. 22 D. Seebach and W. Lubosch Angew. Chem. Internat.Edn. 1976,15,313. 23 D. Seebach and F. Lehr Angew. Chern.Internat. Edn. 1976 15 505; R. Henning F. Lehr and D. Seebach Helv. Chim.Acta 1976,59 2213. Organometallic Chemistry -Part (ii) Main -group Elements Compound (11) prepared by reaction of (12) with Bu"Li is a latent quinone ~arbanion.~~ Useful 1,2-epoxyalkyl-lithium reagents (13) can be obtainedz5 by lithiation of the corresponding epoxides provided that the group X is a second-row substituent group attached via Si P or S. Yet another synthetically useful reagent type is represented by (14; R'= Rz=Me X = 0; R' =Et R2=n-pentyl or Ph X =S) which reacts with ketones to give chain-extended a-substituted a@ -unsaturated aldehydes such as (15) from reaction with cyclohexanone.26 oBr Me0 OMe QLi Me0 OMe R' X Li>Llc R' R2xwy1D Li C H O XR' Me0 OMe Me0 OMe (1 1) (12) (13) (14) (15) Selenium-substituted organolithium reagents have been extensively The reagents are prepared by the action of Bu"Li on gem -diselenyl compounds and can be alkylated hydroxyalkylated acylated or silylated in the same way as the sulphur analogues.The selenium function can be removed reductively by oxidative elimination or in the case of 2-hydroxy-compounds by elimination to form an oxiran so the synthetic potential is great. Scheme 2 illustrates the preparation of an unsaturated ester. PhSe Li PhSe C0,Me 4 90% 60% Reagents i PhSeH; ii BunLi;iii ClCOzMe; iv H202-THF. Scheme 2 Lithiated thioacetal derivatives also continue to provide interesting new applica- tions.For example (16) which does not react with tosylates has been shown to react efficiently with benzenesulphonate esters of primary alcohols thus broadening the scope for alkylation.28 Also whilst compounds of type (17; R = alkyl) cannot be obtained because of elimination of ROLi (17; R = Li) is readily obtained by the (16) (17) (18) 24 M. J. Manning P. W. Raynolds and J. S. Swenton,J. Amer. Chem. Soc. 1976 98 5008. 25 J. J. Eisch and J. E. Galle J. Amer. Chem. SOC. 1976 98 4646; J. Organometallic Chem. 1976,121 ClO. 26 I. Vlattas L. D. Vecchia and A. 0.Lee J. Amer. Chem. Soc. 1976,98 2008; C. N. Skold Synthetic Comm. 1976,6 119. 27 J. N. Denis W. Durnont and A. Krief Tetrahedron Letters 1976,453; D. Van Ende and A. Krief ibid.p. 457; M. Sevrin D. Van Ende and A. Krief ibid. p. 2643; W. Durnont and A. Krief Angew. Chem. Zn&mat. Edn. 1976 15 161. 28 D. Seebach and E.-M. Wilka Synthesis 1976,476. 126 K.Smith action of 2 mol of Bu"Li on (18). The dianion reacts at carbon with electrophiles such as aldehydes.29 1-Lithio- 1 1-bis(pheny1thio)alkanes have previously not been recommended as acyl carbanion equivalents because of their low reactivity towards alkyl halides but they react readily with carbonyl compounds and an interesting new acid-catalysed reductive rearrangement reaction (Scheme 3) gives the same ketone that would be provided by simple deprotection of the alkyl halide Reagents i R1COR2; ii CF3C02H; iii R*CHBrR2; iv H20-HgC12-MeCN. Scheme 3 There have been two significant developments in the use of carbamoyl-lithium reagents (19).Thus when R' and/or R2are methoxymethyl groups the compounds are the synthetic equivalents of (19; R' and/or R2= H) thus providing a direct route to primary and secondary a -hydroxyamides and related Bases such as 0 R' II / Li-C-N LiNPri are the usual deprotonating agents used in the formation of compounds of type (19; R' =R2=alkyl) but provided that the correct solvent is used Bu'Li is also effective and this yields the reagent free from any secondary amine which might subsequently give The necessary solvent is a mixture of THF diethyl ether and pentane (4 :4 :1). A similar solvent mixture though in different propor- tions (4 :1:l) allows the conversion of terminal bromoalkenes into alkenyl-lithium compounds using Bu'Li a procedure which has not previously been developed in any general way.33 Both solvent mixtures are described as 'Trapp's solvent' the confu- sion arising because different original publications by Kobrich and Trapp refer to different proportions of these solvents.Perhaps it would be appropriate to generalize the term 'Trapp's solvent' to include any mixture of the three individual solvents and always to accompany the name by a statement of the proportions involved. The anion from 3-methylpentadienyl-lithium exists substantially in the conforma- tion (20) and is a useful reagent for introducing a diene unit common to several terpenoid species.34 Thus by reaction with R'COR2 it gives (21).29 H. Paulsen K. Roden V. Sinnwell and W. Koebernick Angew. Chem. Internat. Edn. 1976,15,439. 30 P. Blatcher J. I. Grayson and S. Warren J.C.S. Chem. Comm. 1976,547. 3' U. Schollkopf and H. Beckhaus Angew. Chem. Internat. Edn. 1976,15,293. 32 K. Smith and K. Swaminathan J.C.S. Chem. Comm. 1976,387. 33 H. Neumann and D. Seebach Tetrahedron Letters 1976,4839. 34 S. R. Wilson K. M. Jernberg and D. T. Mao J. Org. Chem. 1976,41 3209. Organometallic Chemistry-Part (ii) Main -group Elements (20) (21) Two new methods for estimation of organolithium compounds have been reported.35 Both involve a single titration and one uses readily available diphenylacetic acid as the sub~trate-indicator.~'~ Theoretical-mechanistic studies of organolithium compounds lag far behind their synthetic applications but some are worthy of mention.Ab initio calculations for example reveal that the energy differences between planar and tetrahedral geomet- ries of methane derivatives drop considerably on replacement of hydrogens by highly electropositive substituents such as Li.36 If a narrowing of the energy difference merely accords with naive expectations when the ground-state energies of both geometries are substantially raised it is less obvious that cross-over should occur. Yet 1,l-dilithiocyclopropaneis predicted to be slightly more stable in planar than in tetrahedral Of course the relevance of these gas-phase calculations to the condensed phase where carbon attached to Li seldom has a co-ordination number of four must be questionable.The interesting result reported last year,' that more ketone and less tertiary alcohol are produced the longer a reaction between a lithium carboxylate and an alkyl-lithium reagent is allowed to proceed before quenching is not matched in the reaction of the carboxylic acid itself with the organolithium ~eagent.~' The only species not present in the previous reaction mixture is the acid (apart from alkane which can probably be ignored) so the explanation must involve a further reaction of this acid with one of the reaction intermediates. Equation (2) would seem to accommodate all the data though the authors of the publication propose a much more complicated scheme which is also consistent with the data. Sodium Potassium,Rubidium and Caesium.-The power of Me3SiCH2K as a metallating agent has been dem~nstrated.~~ 3 Group11 Magnesium.-1,3-Benzodithiolium salts formed by reaction of 2-alkoxy- 1,3- benzodithioles with Ph3C'C104- react with Grignard reagents to give 2-substituted 1,3-benzodithioIe~,~~ which were previously available only by much more tedious routes.Furthermore since these thioacetals can be hydrolytically cleaved to give aldehydes this provides a new procedure for formylation of Grignard reagents. 35 J. Kuyper and K. Vrieze J.C.S.Chem. Comm. 1976 64; W. G. Kofron and L. M. Baclawski J. Org. Chem. 1976,41. 1879. 36 J. B. Collins,J. D. Dill E. D. Jemmis Y. Apeloig P. von R.Schleyer,R. Seeger and J. A. Pople,J. Amer. Chem. Soc. 1976,98,5419. 37 R. Levine and M.J. Karten J. Org. Chem. 1976,41 1176. 38 J. Hartmann and M. Schlosser Helv. Chim. Acta 1976 59,453. 39 I. Degani and R. Fochi J.C.S. Perkin I 1976 1886. 128 K.Smith It is interesting that both 1,2- and 1,3-monothiodiketones react with Grignard reagents exclusively at the thiocarbonyl group and by thiophilic The reaction of the latter type gives a new access to cyclopropane derivatives (Scheme 4).40b Reagents i RMgX; ii H30+ Scheme 4 The reaction outlined in Scheme 5 though giving only moderate (ca. 50%) chemical yields of p -substituted aldehydes does so with very high optical induction (ca. 95%).4' R' *CHO '-,A NZ'H(Bu')CO2Bu' ii' iii b R1EH(R2)CH2CH0 * Reagents i BucCH(NH,)C02But; ii R2MgBr; iii H3O+. Scheme 5 Use of a reagent prepared by addition of EtMgI to a highly hindered phenol in a relatively non-polar solvent mixture (CH2C1,-diethyl ether) provides the first gen- eral procedure for incorporation of Mg into metal-free chlorophyll derivative~.~~ The puzzling origin of the secondary alcohol reduction products formed during reactions of ketones with MeMgBr has been traced to magnesium hydride formed during preparation of the Grignard reagent.43 The hydrogen of the hydride origi- nates in the ether solvent and the amount of hydride formed is dependent upon the physical shape and size of the Mg pieces The formation of pinacols in the same reactions is a result of a single-electron transfer (SET)mechanism induced by transition-metal impurities in the Mg used to make the Grignard reagent but for neopentyl Grignards SET mechanisms may play an integral part in uncatalysed reactions.44 Zinc Cadmium and Mercury.-Slurries of finely divided Cd or Zn formed by codeposition of the appropriate metal vapour with a solvent are active enough to form organometallic compounds (presumed to be metal dialkyls) by reaction with simple alkyl halides.45 Although co-ordinating solvents give better results reasona- ble yields can also be obtained in hexane.Respectable yields of a-40 (a) P. Metzner J. Vialle and A. Vibet Tetrahedron Letters 1976,4295;(b)J. L. Burgot J. Masson P. Metzner and J. Vialle ibid. p. 4297. 41 S. Hashimoto S. Yamada and K. Koga J.Amer. Chem. SOC.,1976,98 7450. 42 H.-P. Isenring E. ass K. Smith H.Falk J.-L. Luisier and A. Eschenmoser Helv. Chim. Acra 1975 58 2357. 43 E. C. Ashby T. L. Wiesemann J. S. Bowers and J. T. Laemmle Tetrahedron Letters 1976 21. 44 E. C. Ashby J. D. Buhler I. G. hpp T. L. Wiesemann J. S. Bowers and J. T. Laemmle,J. Amer. Chem. Soc. 1976,98,6561. 45 T. 0.Murdock and K. J. Klabunde J. Org. Chem. 1976,41 1076. Organometallic Chemistry -Part (ii ) Main -group Elements 129 (alky1thio)carbonyl compounds have been on reaction of Reformatsky- type reagents with RSC1. Catalysed reactions of vinylmercuric chlorides continue to provide synthetically useful procedures. Thus treatment of these readily accessible sterically defined compounds with PdClz and LiCl (stoicheiometry 2 :1:4) in HMPA at 0 "C results in essentially quantitative coupling of the vinyl groups without loss of stereochemical definiti~n.~~ Furthermore application of previously established Pd-catalysed car- bonylation of vinylmercuric halides to compounds derived from propargylic alcohols provides a useful new approach to butenolides (Scheme 6).48 Unfortunately the preparation of the organomercurials in these cases proceeds in only moderate yield though the carbonylation is almost quantitative.Finally in the presence of A1Cl3 vinylmercury compounds react readily with acyl chlorides to give crp -unsaturated ketones in high yield.49 ,PR -CO/Li,PdCI ~ Hg CI R2 OH Scheme 6 Peroxymercuration of dienes with hydrogen peroxide results in good yields of Hg-substituted 1,2-dioxacycloalkanes such as (22) from penta- 1,4-diene.These can be converted by standard procedures into Hg-free products thus providing the first high-yielding approach to such compound^.^^ Isopropylidenecarbene (23) is produced by thermal decomposition of (24) at 150"C and reacts with alkenes of various types to give reasonable yields of the corresponding isopropylidenecyclo- propanes.51 The first fully documented example of a 1,4-oxymercuration product of a diene has been reported." It appears that the 1,4-oxymercuration reaction is reversible. 46 I. I. Lapkin G. G. Abashev and F. G. Saitkulova Zhur. org. Khim. 1976,12,976 (Chem.Abs. 1976 85,46 149). 47 R. C. Larock J. Org. Chem. 1976,41,2241. 48 R. C. Larock and B. Riefling Tetrahedron Letters 1976 4661.49 R. C. Larock and J. C. Bernhardt Tetrahedron L.ehi?rs 1976,3097. A. J. Bloodworth and M. E. Loveitt J.C.S. Chem. Comm. 1976,94. 51 D. Seyferth and D. Dagani J. Organometallic Chem. 1976,104 145. 52 A. J. Bloodworth M. G. Hutchings and A. J. Sotowicz J.C.S. Chem. Comm. 1976 578. 130 K. Smith 4 Group111 Boron.-A book on organoborane ~hemistry,~~ and reviews concerned with C-C bond formation using organoborane~,~~ and the chemistry of organoborate~,~~ cate~holborane,~~ have appeared. A comment on some novel types of organoboron compound was made in the Introduction. Other interesting new compounds include (25) (26) and (27). The production of (25a) involves the action of Meerwein’s salt on the known trimethylamine-cyanoborane adduct and it is converted into (25b) by hydrolysis with HCI.Compound (25b) is a boron analogue of an amino-acid and shows potential as an antitumour agent.57 Compounds (26) and (27) are prod~ced~~~,~ by reaction of Mn2(CO)lo with the appropriate boron ligand or an isomer and are of interest as members of a select group of ‘triple-decker-sandwich’ complexes. (25) a; X=NHEt (26) b;X=OH The main interest of organoboranes continues to be in their application as synthetic reagents and studies of alkynylborates in particular are still widespread. Two reports cite the synthesis of unsymmetrical dialkyldialkynylborates and their conversion into unsymmetrical conjugated di-ynes (Scheme 7).59 The essence of YMe I R;BYMe -L Li+[R:BCrCR2]-& R:BCrCR2 Y R2CrC.C=CR3 Li+[R:B(C=CR3)CrCR3]-Reagents i RZC=CLi; ii Y = 0,BF3,OEtz; Y = S no reagent necessary; iii R3C~CLi; iv 12.Scheme 7 both methods is the reaction of R:BYMe (Y=O or S) with an alkynyl-lithium reagent complexation of the initial dialkylalkynylborane by MeY- hindering its further reaction. For Y = 059a this complex must be treated with BF3 to cause decomplexation before the second alkynyl-lithium is added whereas for Y = S because the complex is weaker subsequent reaction with alkynyl-lithium occurs 53 T. Onak ‘Organoborane Chemistry’ Academic London 1975. 54 J. Weill-Raynal Synthesis 1976 633. 55 E. Negishi J. Organometallic Chem. 1976 108 281. 56 C. F. Lane and G. W. Kabalka Tetrahedron 1976,32,981. 57 B. F. Spielvogel L.Wojnowich M. K. Das A. T. McPhail and K. D. Hargrave J. Amer. Chem. Soc. 1976,98,5702. 58 (a)G. E. Herberich J. Hengesbach U. Kolle G. Huttner and A. Frank Angew. Chem. Internat. Edn. 1976 15,433; (b)W. Siebert and K. Kinberger ibid. p. 434. 59 (a)J. A. Sinclair and H. C. Brown J. Org. Chem. 1976,41 1078; (b)A. Pelter R. J. Hughes K. Smith and M. Tabata Tetrahedron Letters 1976,4385. Organometallic Chemistry -Part (ii ) Main -group Elements directly given sufficient time (ca. 1h at 20°C).596 Details6' of the previously reported alkylations of trialkylalkynylborates have appeared and are accompanied by an attempt using MIND0/3 and a6 initio calculations to obtain a mechanistic rationalization of the reaction course. The large calculated energy release accom- panying rearrangement is noteworthy.9,9-Dialkyl-9-boratabicyclo[3,3,l]nonanesare capable of great selectivity in reductions of carbonyl compounds and oxirans,61 and the rearrangement which accompanies reduction gives a useful access to bicyclo-octylboranes (Scheme 8).62 Scheme 8 Reactions of trialkylboranes with 1-methyloxyvinyllithium at -80 "C give pre- sumed intermediate dialkylvinylboranes which have opposite regiochemistry to those obtained by hydroboration of alkyne~.~~ Attempts to oxidize these inter- mediates to give methyl ketones are successful only when the alkyl groups of the original organoborane are unhindered other examples leading to mixtures of the ketones with tertiary alcohols resulting from a second rearrangement.Indeed in aqueous HCl the second rearrangement is quantitative and oxidation then gives the 1,l-dialkylethanol in good yield.63 This gives further evidence that many vinyl- boranes are very susceptible to rearrangement reactions. It has been known for several years that trialkylboranes add 1,4 to @-unsaturated carbonyl compounds by a radical mechanism. This year however several modifica- tions of this general behaviour have been recorded. For example alkenylboron compounds add to cisoid enones in refluxing THF apparently by a non-radical mechanism involving a cyclic transition and the outcome is a useful synthesis of $3 -unsaturated ketones which is even applicable to reactions using methyl vinyl ketone. Another exception is the 1,2-addition under vigorous conditions of triphenylborane to 2-methyla~rolein,~~ and copper organoborates prepared by successive addition of MeLi and CuBr to trialkylboranes allow extension of the conjugate addition process to reaction with acrylonitrile and other compounds.66 The reaction of trialkylboranes with iron(II1) azide is catalysed by HzOz and gives azidoalkanes in respectable yields based on transfer of a single alkyl group.67 The direct replacement of the boron of an organoborane by a carbon atom bearing an extra substituent which can be achieved using several reagents is one of the most 6o A.Pe1ter.T.W. Bentley,C. R. Harrison C. Subrahrnanyam and R. J. hub J.C.S.PerkinZ 1976,2419 Y.Yamarnoto H. Toi A. Sonoda and S.-I. Murahashi J. Amer. Gem. SOC. 1976,98 1965; J.C.S.Chem. Comm. 1976,672. G. W. Kramer and H. C. Brown J. Amer. Chem. Soc. 1976,98 1964. 63 A. B. Levy and S. J. Schwartz Tetrahedron Letters 1976,2201. 64 P. Jacob and H. C. Brown,J. Amer. Chem. Soc.,1976,98,7832. 6s R. Koster H.-J. Zimrnermann and W. Fenzl Annalen. 1976 1116. 66 N. Miyaura M. Itoh and A. Suzuki Tetrahedron Letters 1976 255. 67 A. Suzuki M. Ishidoya and M. Tabata Synthesis 1976 687. 132 K.Smith interesting of all reactions of organoboranes. So far there is no comparable replacement by an element other than carbon but transfer of two alkyl groups to nitrogen using a reagent which is prepared in situ from 2,4-dinitrophenoxyamine and Bu'OCI has now been reported.68 Application of the reaction to a perhyd- roboraphenalene results in a product which can be cyclized to the parent system of the coccinellidae alkaloids (Scheme 9).ClNH I Reagents i react at -78 +25 "C; ii H2@-OH-; iii acid or heat. Scheme 9 Indeed it is pleasing to note that there has been a significant increase in the use of organoboranes in syntheses of natural products or their model systems. For exam- ple a synthesis of epijuvabione and juvabione (28) uses carbonylation of an organoborane as its final whilst formation of the prostaglandin model (29) was accomplished by a modified I2 coupling reaction of the appropriate alkenylalkyl- dimeth~xyborate.~'A number of terpenoid species were obtained from hydrobora- tion products of protected gerani01.~' These included (30),prepared most efficiently using a cyanoborate reaction.C0,Me & GC,H., OSiMe,Bu' G O H Bu' H (28) (29) (30) The same publication7' also reported cyclization caused by treatment of diboryl compounds with AgN03 and other have reported formation of (E)-alkenes by treatment of 172-diboryl compounds (obtained by hydroboration of alkynes) with basic AgN03. Many of the better known reactions of organoboranes involve an intramolecular 1,2-shift which is mechanistically analogous to the Wagner-Meerwein rearrange- ment and as in that reaction the migrating centre retains its stereochemical config- uration during rearrangement. However not all non-radical reactions of 68 R. H. Mueller Tetrahedron Letters 1976 2925. 69 E. Negishi M. Sabanski J.-J. Katz and H.C. Brown Tetrahedron 1976,32,925. 70 D. A. Evans T. C. Crawford R.C. Thomas and J. A. Walker J. Org. am. 1976,41,3947. R. Murphy and R. H. Prager Austral. J. Chem. 1976,29,617. 72 K. Avasthi S. S. Ghosh and D. Devaprabhakara TetrahedronLetters 1976,4871. Organometallic Chemistry -Part (ii) Main -group Elements 133 organoboranes involve this characteristic intramolecular shift and consequently not all need involve retention of configuration. Exceptions have been noted in the base-induced halogenations of organoborane~,~~ and in the formation of cyclo- propanes by hydroboration-eliminationof allylic halides where inversion occurs at both From data obtained using the same techniques that have previously been said to indicate the aromaticity of borazaro- and related compounds Mikhailov has con- cluded that there is no evidence for ar~maticity.~' Careful attention to choice of model systems will obviously be required if a concensus view is to be established.Selective cyclic borylations of polyols and glycosides which were previously achieved with PhB(OH), can now be achieved with triethylborane-pivalic acid systems under mild condition^.^^ This should provide a valuable addition to the repertoire of protection methodology in the carbohydrate field. Aluminium Gallium Indium and Thallium.-The (E)-alkenyldialkylalanes which are obtained by hydroalumination of alkynes have received much attention as potentially important synthetic reagents. The general reaction involves transfer of the alkenyl unit intact to an organic electrophilic group [equation (3)] and several useful procedures have emerged.Reactive electrophiles such as ethyl chloroformate (givingtrans -a@-unsaturated esters) and chloromethyl ethyl ether (ally1 ethers) react readily at moderate temperatures and give good yields of whereas less reactive electrophiles such as alkyl halides and oxirans require prior conversion of the alane into an ate-complex by addition of an alkyl-lithi~m.~~ Reactions with aryl and alkenyl halides can be encouraged by addition of catalytic quantities of Ni and Pd complexes.79 New experimental approaches to oxythallation have been reviewed," and several significant developments have been reported. Undoubtedly the most exciting is the use of a reagent consisting of thallium(II1) nitrate (TIN) deposited on an acidic montmorillonite clay (K-10).81This reagent is apparently capable of all the known oxidative reactions of TTN generally under milder conditions often with better yields and always with an easier work-up procedure.For reactions in solution trimethyl orthoformate as solvent often gives better results than the previously used acidic media.82 73 (a)H. C. Brown N. R. Delue G. W. Kabalka and H. C. Hedgecock J. Amer. Chem. Soc.,1976.98 1290; (6) D. E. Bergbreiter and D. P. Rainville J. Organometallic Chem. 1976 121 19. 74 H. L. Goering and S. L. Trenbeath,J. Amer. Gem. Soc. 1976,98 5016. 75 B. M. Mikhailov and M. E. Kuimova J. Organometallic Chem. 1976 116 123. 76 R. Koster and W. V. Dahlhoff Annalen 1976 1925.77 G. Zweifel and R. A. Lynd Synthesis 1976,625 816. 78 (a)J. J. Eisch and G. A. Damasevitz,J. Org. am. 1976,41,2214;(b)S. Baba D. E. Van Horn and E. Negishi Tetrahedronhtters 1976,1927; (c)E. Negishi S. Baba and A. 0.King J.C.S. Chem. Comm. 1976 17. 79 E. Negishi and S. Baba J.C.S. Chem. Comm. 1976,596; J. Amer. Chem. Soc. 1976,98,6729. A. McKillop and E. C. Taylor Endeaoour 1976,3588. 81 E. C. Taylor C.-S. Chiang A. McKillop and J. F. White J. Amer. Chem. SOC. 1976 98 6750. 82 E. C. Taylor R. L. Robey K.-T. Liu B. Favre H. T. Bozimo R. A. Conley C.-S.Chiang A. McKillop and M. E. Ford J. Amer. Chem. SOC.,1976,98 3037. 134 K. Smith Dithalliation of a single aromatic ring has not previously been recorded but extended reaction of anisole with TI(OCOCF3) at room temperature gives dithal- liated Furthermore the thalliation reaction must be reversible because the monothalliation isomer distribution varies with time (Scheme 10).6 6 ,i OMe &1(0c0cF3)2 ~Tl(ococF3)2 ~ ~ \ \ \ \ n(ocmF3)Z Tl(OCOCF,) Reagents i Tl(OCOCF,), -25 "C; ii TI(0COCF3), 1 mol 20 "C;iii TI(0COCF3),,excess 20 "C 3 days. Scheme 10 5 GroupIV Silicon.-Advances in the field of multiply-bonded Si species and comments con- cerning silicenium ions have been presented in the Introduction. Following on recent success in the synthesis of silacyclopropanes come two reports of syntheses of silacyclopropenes (3 1).84 Both examples are formed by reaction of thermally generated Me2% with the appropriate alkyne and show surprising thermal stability under an inert atmosphere.The 29Si resonance of the ring Si atom of (31b) is over 100p.p.m. upfield from Me4% Amongst other interesting new organosilicon compounds prepared during 1976 are (32) and (33). The former is prepared by successive reaction of Me3SiCECLi with sulphur and Me3SiC1 and is a distillable It reacts exothermally with alcohols and amines losing one Me,Si unit in the process giving a-trimethylsilyl thioesters and thioamides which may be useful for further synthetic transformations. Compound (33) was prepareds6 from 1,8-dilithionaphthalene and Me2SiC12. Fluoride ion is a powerful nucleophile towards Si and can be used to increase the anionic activity of Si-bound groups. Thus when a tetra-alkylammonium fluoride is added to an alkynylsilane it behaves as a solution of R'C_C-R2,N'; addition of an aldehyde or ketone followed by aqueous work-up gives a good yield of the propar- gylic Similarly treatment of (34) with Me4N'F- gives a solution which SiMe SiMe R/-\R CI (31) a;R=Me (32) (33) (34) b; R = SiMe3 83 G.B. Deacon R. N. M. Smith and D. Tunaley J. Organometallic Chem. 1976 114 C1. 84 (a)R. T. Conlin and P. P. Gasper,J. Amer. Chem. SOC.,1976,98,3715;(b)D. Seyferth D. C. Annarelli and S. C. Vick ibid. p. 6382. 85 S. J. Harris and D. R. M. Walton J.C.S. Chem. Comm. 1976 1008. s6 L. S. Yang and H. Schechter J.C.S. Chem. Comm. 1976 775. 87 E. Nakamura and I. Kuwajima Angew. Chem. Zntemat. Edn. 1976,15,498. Organometallic Chemistry -Part (ii) Main -group Elements 135 behaves as a source of isopropylidenecarbene.88This must of course be generated in the presence of the desired trapping agent.Treatment of Si-substituted oxirans with &N'F-in DMSO provides a convenient method for desilylation of such Allylsilanes show potential as synthetic intermediates. Their reactions with electrophiles such as ketones and ketals are promoted by TiCL and the products are 4-hydroxy- or 4-alkoxy-alkenes produced with concomitant rearrangement of the allyl group.go Rearrangement also accompanies reactions of allyl silanes with peracids and phenylsulphenyl tetrafluoroborate which yield allyl alcohols and sulphides respe~tively,~' and with chlorosulphonyl isocyanate followed by pyridine which give py-unsaturated nit rile^.^' Me,SiLi adds in a conjugate manner to cyclohexenone and the product enolate can then be alkylated providing a convenient route to 2-alkyl-3-silylcyclo-hex an one^.^^ Trialkylsilyl alkali-metal derivatives can also be used to effect stereo- specific (trans-) deoxygenation of oxiran~.~~ Lithiated a-(trimethylsilyl)aldimines behave as a synthetic equivalent of the Wittig reagent Ph,P=C(R)CHO providing an effective approach to cup -unsaturated aldehyde^,^^ and reductive alkylation of alkynylsilanes as illustrated in Scheme 11 offers a useful stereoselective approach to vinyl~ilanes,~~ themselves interesting reagents.Me3Sg1 iii iv Me,Si R' R,SiC=CR' A b >=( (R= Me) B&AI H R2 H (R= Et) Et3Si Et3Si H MH iii iv * )=( R' BU~AI R' R2 Reagents i Bu';AIH/heptane-ether; ii BuiAlH/heptane; iii MeLi; iv R'I.Scheme 11 Germanium Tin and Lead.-Specific reductive ring-opening of the cyclopropane ring in vinylcyclopropanes is not normally an easy task but the two-stage process shown in Scheme 12 provides a useful approa~h.'~ Treatment of (35) with two moles of Bu"Li gives (36) a convenient reagent for synthesis of other heterocycle^.^^ The interesting compounds of type R2Sn where R 88 R. F. Cunico and Y.-K. Han J. Organometallic Chem. 1976 105 C29. 89 T. H. Chan P. W. K. Lau and M. P. Li Tetrahedron Letters 1976 2667. 90 A. Hosomi and H. Sakurai Tetrahedron Letters 1976 1295; A. Hosomi M. Endo and H. Sakurai Chem. Letters 1976 941. 91 M.J. Carter and I. Fleming J.C.S. Chem. Comm. 1976 679. 92 G. Deleris J. Dunogues and R. Calas J. Organometallic Chem. 1976 116 C45. 93 W. C. Still J. Org. Gem. 1976,41 3063. 94 P. B. Dervan and M. A. Shippey J. Amer. Chem. SOC.,1976,98 1265; M. T. Reetz and M. Plachky Synthesis 1976 199. 95 E. J. Corey D. Enders and M. G. Bock Tetrahedron Letters 1976,7. 96 K. Uchida K. Utimoto and H. Nozaki J. Org. Gem. 1976,41 2215. 97 M. Ratier and M. Pereyre Tetrahedron Letters 1976 2273. 98 G. Mark1 and P. Hofmeister Tetrahedron Letters 1976,3419. 136 K. Smith R3 R3 A R2 IR' R' R2 R3=H Me Scheme 12 se Sn Li Li Bl5 'Bu" (35) (36) is a bulky group such as bis(trimethylsilyl)methyl show a bent C-Sn-C unit in the crystalline and behave like other Snl' species in their reactivity towards alkyl and aryl halides."' Treatment of Bu2Bu'SnCI with Bu'O- radicals produces the less stable Bun* rather than the But* radical a phenomenon which is tentatively ascribed to steric constraints in a five-co-ordinate intermediate.lo' 6 GroupV The preparation and spectroscopic identification of bismabenzene (37) have been claimed,"* and this completes the series of Group V heterobenzene derivatives. Hetero-Cope rearrangements are involved in the formation of (38),the sole product of treatment of 4-hydroxyarsabenzene with ally1 bromide.lo3 Allylation at As is followed by the hetero-Cope rearrangement and this process is repeated before the third allylation gives the final product.'H N.m.r. spectra suggest that compounds of type (39; E = As or P) are better considered as ylides than as compounds possessing an aromatic sextet.lo4 (37) 9y D. E. Goldberg D. H. Harris M. F. Lappert and K. M. Thomas J.C.S. Gem. Comm. 1976 261. loo M. J. S. Gynane M. F. Lappert S. J. Miles and P. P. Power J.C.S. Chem. Comm. 1976 256. A. G. Davies B. Muggleton B. P. Roberts M.-W. Tse and J. N. Winter J. Organometallic Chem. 1976 118 289. l02 A. J. Ashe Tetrahedron Letters 1976 415. Io3 G. Mark1 and J. B. Rampal Angew. Chem. Internat. Edn. 1976 15 690. Io4 A. J. Ashe and T. W. Smith J. Amer. Chem. SOC.,1976 98 7861.