8 Organometallic Compounds By D. J. CARDIN Department of Chemistry Trinity College Dublin University Dublin 2 K. R. DIXON Department of Chemistry University of Victoria Victoria 6.C. Canada 1 Introduction This chapter deals first with main-group and early transition-metal (Groups 111-VI) compounds (D.J.C.) followed by metal carbonyls and the later transition-metal derivatives (K.R.D.). The first part contains a section of shorter topics which were felt to be of wide general interest (in which the order of presentation is by the Group of the periodic table to which the appropriate metal belongs) and a selection of review articles and books. In the main the coverage is for the year 1975 but the second part covers specifically November 1974 to early November 1975.2 Reactions of Metal Atoms Conceptually one of the simplest and most attractive methods for the synthesis of organometallic compounds is the direct reaction of metal atoms with appropriate ligands (see Chapter 6 p. 120). This subject has received much attention recently and various aspects have been reviewed by Skell,'" Timms,lb Koerner von Gustorf,* and Klabunde3 (for matrix-isolation work with metal carbonyls see Section 20). At present all the transition metals can be used in this type of synthesis many in gram quantities. Sandwich complexes have been made for chromium,44 titanium,' and even tungsten.4b The use of halogen-substituted arenes leads to some unexpected effect^.^**^ Thus for vanadium whereas monosubstitution by chloro- fluoro- or trifluoromethyl-groups gives greater yields than with non-substituted benzene disubstitution by the same groups gives reduced or zero yields.Yields were not so (a) P. S. Skell and M. J. McGlinchey Angew. Chem. Internat. Edn. 1975,14 195; (b) P. L. Timms ibid. p. 273. * E. A. Koerner von Gustorf 0.Jaenicke 0.Wolfbeis and C. R. Eady Angew. Chem. Internat. Edn. 197514 278. K. J. Klabunde Angew. Chem. Internat. Edn. 1975 14 287. (a)P. S. Skell D. L. Williams-Smith and M. J. McGlinchey J. Amer. Chem. SOC.,1973,95,3337;(b)M. P.Silvon E. M. Van Dam and P. S. Skell ibid. 1974,96,1945;(c)P. S. Skell and L. K. Wolf ibid. 1972 94 7919. R. Middleton J. R. Hull S. K. Simpson,C. H.Tomlinson andP. L.Timms J.C.S. Dalton 1973,120. V. Graves and J.J. Lagowski Abstracts 165th A.C.S. Meeting Dallas Texas 1973(Paper No. INOR 52). F. W. J. Benfield M. L. H. Green J. S. Ogden and D. Young J.C.S. Chem. Comm. 1973,866. (a) K. J. Klabunde and H. F. Efner J. Fluorine Chem. 1974 4 115; (b) K. J. Klabunde and H. F. Efner Inorg. Chem. 1975 14 789. 179 D.J. Cardin and K. R. Dixon sensitive with chromium and the bis(trifluoromethyl)benzenechromium(O) com-plexes are remarkably air stable.8b Interest also continues in the reactions of the vapours of main-group metals with organic molecules. Lithium vapour is known to react with chloro~arbons~” giving polylithio-organic and with carbon va~our,~’ compounds. The reaction of excess lithium vapour with alkenes has been studied and the products characterized by hydrolysis (H20and D20),affording alkenes and alkanes identified by n.m.r.g.l.c. and mass spectrometry or by conversion into poly(trimethylsily1) derivatives. Poly-lithioalkenes predominate in the products and addition across the double bonds is relatively uncommon.9c Although the reactions of aluminium vapour with olefins afforded no isolable aluminium D20 hydrolysis products were interpreted in terms of intermediate a-bonded alkylaluminiums. More recently the reaction with ethylene has provided evidence of a complex involving donation from the n-orbitals of the olefin based on e.s.r. spectra of the matrix-isolated complex. lo Complexes are also formed with gallium and indium atoms. The equilibrium between the carbene and ‘carbenoid’ species produced by reaction of lithium atoms with carbon tetrachloride is known to lie well to the left of C1,CLi $ CCI + LiCl (1) equation (l).’la Careful analysis of i.r.data of the matrix products with the metals ‘Li Na K or Cs has revealed a further type of carbenoid species (1).’lb 3 Carbene and Carbyne Complexes In spite of the preparation of numerous carbene complexes of the transition metals,12 until recently no methylene complexes had been obtained although they had been postulated as reaction intermediate^'^"-' and complexes of secondary carbenes (CHR)are well The synthesis of the first methylene complex is shown in equation (2).14” The bis(cyclopentadieny1)methylene complex decomposes slowly in WCP) (Cp’)Me Ph,C + BF,-[Ta(Cp)(Cp’)Me,I+BF,-*Ta(Cp)(Cp’)Me(CH,) (2) Cp = q5-CsH5;Cp’ = q5-CsH4R R = H or Me; base = Me,PCH, LiN(SiMe,), or NaOMe (a)C.Chung and R. J. Lagow J.C.S. Chem. Comm. 1972 1078; (b)L. A. Shimp and R. J. Lagow J. Amer. Gem. SOC.,1973,95,1343;(c)J.A. Morrison C. Chung and R. J. Lagow ibid. 1975,97,5015. lo P. H. Kasai and D. McLeod J Amer. Chem. Soc.,1975,97 5607. l1 (a)D. F. Hoeg D. 1. Lusk and A. L. Crumbliss J. Amer. Chem. SOC.,1965,87,4147,and refs. therein; (b)D. A. Hatzenbuhler L. Andrews and F. A. Casey ibid. 1975,97 187. l2 D. J. Cardin B. Qtinkaya M. J. Doyle and M. F. Lappert Chem. SOC.Reu. 1973,2,99;M. F. Lappert J. OrganometallicChem. 1975 100 139. l3 (a)M. L. H. Green M. Ishaq and R. N. Whiteley J. Chem.Soc. (A),1967,1508; (b)M.R. Collier B. M. Kingston and M.F. Lappert Chem. Comm. 1970,1498;(c) N. J. Cooper and M. L. H. Green J.C.S. Chem. Comm. 1974,761; (d)B. Cetinkaya M. F. Lappert G. M. McLaughlin and K. Turner J.C.S. Dalton 1974 1591; (e)R. R. Schrock J. Amer. Chem. SOC.,1974 % 6796. 1 (a)R. R. Schrock J. Amer. Chem. SOC.,1975,97,6577;(b)L. J. Guggenberger and R. R. Schrock ibid. p. 6578; (c)L. J. Guggenberger and R. R. Schrock ibid. p. 2935. OrganometaI I ic Compounds deuteriobenzene solution forming 0.5 mol of Ta(Cp),Me(CH,CH,) and presumably Ta(Cp),Me which can be trapped as Ta(Cp),Me(CO) under carbon monoxide. With CD31 the methylene complex yields CH,D and the two isomers of Ta(Cp),(CH2CD2)I presumably uia Ta(Cp),Me(CH,CD,)I. This and other reac- tions show nucleophilic properties for the co-ordinated methylene group.The crystal structure of the complex reveals two eclipsed cyclopentadienyl rings related by a mirror plane containing the C-Ta-C bonds of the CH,-Ta-CH sy~tem.'~' By contrast with the majority of carbene complexes,12 the methylene group can n-bond only with the metal atom and the Ta-CH distance [2.206(10)A] is considerably shorter than the Ta-CH (ca. 2.25A) of the same compound but longer than the formal triple bond [1.76(2) A] of Ta(CH,CMe,),CCMe,,Li(NN'-dimethylpiperazine). 14c This 7r-bonding is supported for the cyclopen-tadienyl(methylcyclopentadieny1)methyleneanalogue for which n.m.r. measure-ments indicate a rotational barrier about the Ta-CH bond in excess of 2 1kcal mol-' . Insertion reactions into M-Ccarb bonds are not common and few have been reported since the curious reaction of PhSeH with carbenes of the Group VIA metals was rep~rted."~ The insertion of aminoacetylenes equation (3) has now been described.15' ,*7NEt2 (OC),CrC(OMe)Me + HC_CNEt,-(OC),Cr -c (3) \ CH II C(0Me)Me Carbyne complexes which were originally prepared from carbene species using boron halidesl5" have now been obtained by a variety of routes. {A number of C-chlorocarbene complexes such as [Cr(CO),(CClNMe,)] [Mn(CO),-(CClNMe),)]' or [Rh(CO)Cl,(CClNMe,)] which may be intermediates in the carbyne synthesis have been prepared from (Me,NCCl,)'Cl-and Na',[Cr(C0),I2- Na'[Mn(CO)5]- or Rh' ~pecies."~}. Thus whereas chromium carbenes with hydroxyl substituents react with carbodi-imide forming new carbene complexes [equation (4)],'5e tungsten analogues give dinuclear carbyne derivatives [equation (5)].Whereas it has been known for some time that electron-rich olefins such as [:CN(Me)CH,CH,NMe], are sources of metal carbene complexes,12 it has now been shown that under milder conditions this olefin and [Cr(CO),(nor- bornadiene)] give the N,N-bonded [Cr(CO) (olefin)] complex.'5f The effect on structure and bonding of heteroatom substituents on the Ccarb atom of carbene complexes has been well documented but corresponding information on 1s (a)E. 0.Fischer and V. Kiener Angew. Chem. Internat. Edn. 1967,6,961; (b)K. H. Dotz and C. G. Kreiter J. Organornetallic Chem. 1975,99,309; (c)E. 0.Fischer C. G. Kreiter J. Miiller G. Huttner and H.Lorenz Agnew. Gem. Infernat. Edn. 1973,12,564;(d) A. J. Hartshorn M. F. Lappert and K. Turner J.C.S. Chem. Comm. 1975 929; (e) E. 0.Fischer K. Weiss and C. G. Kreiter Chem. Ber. 1975,107,3554; cf) B. atinkaya P. B. Hitchcock M. F. Lappert and P. L. he J.C.S. Chem. Comm. 1975 683; (g)E. 0.Fischer G. Kreis F. R. Kreissl W. Kalbfus and E. Winkler J. Organometallic Chem. 1974,65,113; (h)E. 0.Fischer G. Huttner W. Kleine and A. Frank Angew. Chem. Internat. Edn. 1975,14 760; (i) E. 0.Fischer and V. Schubert J. Organometallic Chem. 1975,100 59. D.J. Cardin and K.R.Dixon carbyne derivatives is new. Diethylaminocarbyne complexes of tungstenlSg and chromium have been obtained and the crystal structure of the latter has been While the Cr=C distance of trans-[Et,NC=Cr(Br)(CO),]-[1.720(10) A] is not significantly greater than that found in the analogous methylcar- byne c~mplex,'~' the C-N distance [1.294(12) A] is indicative of considerable double-bond character.4 a-and f3-Eliminations Of the various elimination processes involved in the decomposition of metal alkyls the a-process is relatively little documented and was not established for homoleptic derivatives of the early transition elements. 16" However evidence had been obtained for such a mechanism occurring in tungsten alkyl~,'~" supported by D-labelling experiments. The a-abstraction process has now also been demon- strated in a homoleptic species [equation (6)],also supported by labelling studies.16b Ta(CH,CMe,),Cl + 2LiCH,CMe -+ [Ta(CH,CMe,),] -+Ta(CH,CMe,),CHCMe (not (6) isolated) It appears from data so far available'," that intramolecular a-abstraction will occur most easily when the metal atom is sterically crowded.The isolation of stable Pr',Cr shows that neither this last impression nor the view that alkyls with p-hydrogen atoms will be unstable has general validity. The mechanism proposed'6C [equations (7) and (S)] was suggested by the increase of yield observed upon irradiation. Analogous compounds replacing isopropyl with methyl ethyl or t-butyl CrCl + 3Pr'MgBr -P [CrPr',] !% Pr'. + [CrPr',] (7) (not isolated) Pr'. + CrPri3 + ~r~r' (8) groups could not be obtained and the stability of the isopropyl species was attributed to the stability of the radicals as well as to steric crowding.l6 (a)P. J. Davidson M. F. Lappert and R. Pearce Accounts Chem. Res. 1974,7,209;(b) R. R. Schrock J. Amer. Chem. SOC.,1974 96 6796; (c)J. Miiller and W. Holzinger Angew. Chem. Znternat. Edn. 1975 14 760. Organometallic Compounds 183 5 Organo-lanthanides and -actinides An area which attracted much attention during 1975 is that of organo-lanthanide and -actinide species. Several review articles a~peared,'~+~ and a number of interesting publications will be described. The allyl compounds Cp2MC3H5 appear to be n-bonded when M is a lanthanide (Sm Er or H0),17= whereas a crystal structure of Cp,UCH,C(Me)CH shows that the 2-methylallyl group is c+-bonded;17' cr-bonding had previously been suggested for the allyl analogue17g on the basis of i.r.spectroscopy and indeed the new lanthanide species show a band at 1533cm-' associated with the C-C stretching mode of .rr-ally1 groups. The organometallic chemistry of the lanthanides has been mainly restricted to wbonding ligand~,'~~" but alkyl and aryl derivatives have now been prepared by the organolithium route for Gd Er and Yb [equation (9)]. Although air- and moisture-sensitive the new Cp2LnC1 + RLi + Cp2LnR + LiCl (R = Ph or Me) (9) compounds show thermal stability decomposing only above 130"C under Ar.l7" The unusual temperature dependence of magnetic susceptibility might provide evidence for considerable covalency in the new species and is not observed for example in the v-bonded complexes. The reaction of trimethylsilylmethyl-lithium with uranium tetrachloride afforded solvates of Li,U(CH,SiMe,), the first uranium compound with more than a single metal-carbon u-bond.17' Vibrational spectra of the product of the exchange reaction shown in equation (10) indicate a triple hydrogen bridge. The "B decoupled 'H n.m.r. spectrum of Cp,UBH exhibits at Cp,UBH + R,B + Cp,UH,BR + RzBH (R = Et or Ph) (10) low temperatures collapse of the BH multiplet although the slow exchange limit could not be This is the first time for any metal borohydride species that slowing of the bridge/ terminal hydrogen rearrangement could be observed. Bis(cyc1o-octatetraeny1)uraniumand related molecules are highly air-sensitive. An air-stable 'uranocene' derivative has been obtained by using 1,3,5,7-tetraphenylcyclo-octatetraene dianion and uranium tetrach10ride.l~" The stability is presumably the result of steric blocking of the metal atom and consequently suggests that oxidative attack (0,)takes place at U and not at the ring.6 Cyclopentadienyl and Cyclo-octatetraenyl Compounds Current interest in the bis(cyclopentadieny1)metal derivatives of the Group VIA metals undoubtedly springs from the discovery of the remarkable reactions with 17 (a)M. Tsutsui N. Ely and A. E. Gebala Ann. New York Acad. Sci. 1974,239,160;(b)M. Tsutsui U.S. Ntis AD-A Rep. 1975 No. 008871 [Govt. Report Announce. Index (U.S.) 1975 75 631; (c) T. J. Marks J. Orgunometullic Chem. 1975 95 301; (d) N. S. Vyazankin R. N. Shchelokov and 0.A. Kruglaya Metody Elem.-Org. Khim.1974,2,905[Nauka Moscow]; (e)M. Tsutsui and N. Ely J. Amer. Chem. Soc. 1975,97,3551; v) G. W. Halstead E. C. Baker and K. N. Raymond ibid. p. 3049; (g)F. A. Cotton J. W. Faller and A. Musco Znorg. Chem. 1967,6,179,and refs. therein; (h)See for example K. 0.Hodgson F. Mares D. F. Starks and A. Streitwieser J.Amer. Chem. Soc. 1973,95,8650;(j)H. Gysling and M. Tsutsui Adv. Organometallic Chem. 1970 9 361; (k) M. Tsutsui and N. M. Ely J. Amer. Chem. Soc. 1975 97 1280; (I) R. Anderson E. Carnowa-Guzman K. Mertis E. Sigurdson and G. Wilkinson J. Organometallic Chem. 1975,99 C19; (m)T. J. Marks and J. R. Kolb J.Amer. Chem.Soc. 1975,97,27;(n)A. Streitwieser and R. Walker,J. Organometallic Chem. 1975,97,C41. D.J. Cardin and K.R.Dixon C-H bonds.'8a The Mo and W compounds are relatively reactive compared with the Cr but only the W species inserts into aromatic C-H bonds.Reactivity with CO also varies and the adduct Cp,Cr(CO) is formed reversibly under carbon monoxide in contrast to the known stable Mo and W analogues.'8b Not surprisingly the com- plexes (C,H,)(C,H,)M(CO) are stable for all three metals but tungsten is unique in forming CP~W(CO)~ a 20-electron complex. The carbonyl i.r. stretches exhibit no variations attributable to 'back-bonding' effects which are presumably not significant in determining reactivity differences. These are discussed subsequently in terms of structure and bonding. 18' New q2-nitrile complexes of molybdenocene have been obtained some of which can be reduced to yield ammonia uia isolable iminium intermediates [equation (11)].18d These adducts differ from earlier related r NH2 1' Cp,Mo(CF,CN) HC'-Hzo k2M0(C1)C (1 1) II CF CliNaBH,-OH ~ 1 NH + Cp2MoH + unidentified hydrocarbons molecules which either proved to be a-bonded'8e or had labile nitrile molecules.'8f (Details of aza-ally1 derivatives of Mo and W are described in Section 22.) The dicarbonyl derivative of titanocene is now more easily accessible,18g and its crystal structure has been reported.lsh By contrast with titanocene dichloride the cyclo- pentadienyl rings are eclipsed.The Ti-C(0) bond length [2.030(11) A] is the first structural information available for a Group IVA metal carbonyl. Cyclo-octatetraene (cot) complexes of titanium have been prepared by a new route and (Reproduced by permission from J.Organometallic Chem. 1975,92 329) l8 (a)K. Elmitt M. L. H. Green R. A. Forder I. Jefferson and K. Prout J.C.S. Chem. Comm. 1974,747; (b) K. L. Tang Wong and H. H. Brintzinger J. Amer. Chem. SOC.,1975,97,5143; (c)H. H. Brintzinger L. L. Lohr and K. L. Tang Wong ibid. p. 5146; (d)J. L. Thomas ibid. p. 5943; (e)J. G. Dunn and D. Edwards Chem. Comm. 1971 482 and refs. therein; (f) W. J. Bland R. D. W. Kemmitt and R. D. Moore J.C.S. Dalton 1972 1292; (g) H. Alt and M. D. Rausch J. Amer. Chem. Soc. 1974,96,5936; (h)J. L. Atwood K. Stone H. G. Alt D. C. Hrncir and M. D. Rausch J. Organometallic Chem. 1975 96 C4;0')H. R. Van der Wal F. Overzet H. 0.Van Oven J. L. de Boer H. J. De Liefde-Meijer and F.Jellinek J. Organometallic Chem. 1975,92,329;(k)J. Knol A. Westerhof H. 0.Van Oven and H. J. De Liefde-Meijer J. Organometallic Chem. 1975,% 257. Organometallic Compounds their structures determined [(2) and (3)].'*j The cot ligands in (2) lie approximately perpendicular to the body diagonals of the hexahedron (distorted cube). In both structures the cot rings assume an umbrella-like shape with the hydrogen atoms bent in towards the titanium atoms. Related cyclo-octatetraenetitanium derivatives undergo formal oxidation with iodine [equations (12) and (13)] forming iodides whose crystals are both thermally- and air-stable (although they decompose rapidly with water).lgk (q8-cot)(qS-Cp)Ti+ +I2 (q8-cot)(qS-Cp)TiI (q8-cot)(q5-Cp)Ti+ +I2 -P (q8-cot)(qS-Cp)TiI 7 Hydrometallations Insertion reactions involving zirconium-carbon and -hydrogen (hydrozirconation) bonds not only provide potentially useful synthetic procedures but are mechanisti- cally unusual.Thus the hydrozirconation of olefins by contrast with e.g. hydrofor-mylation gives an alkylzirconium species having the metal attached to the least -or + -P Cp,(Cl)Zr (14) -or hindered carbon atom even if internal olefins are used1'= [equation (14)]. The alkyls thus formed undergo carbon monoxide insertion even faster than their titanium analogues 19b although dibenzylzirconocene fails to react with CO under forcing conditions.19c The resulting acyl compounds can be cleaved to yield the correspond- ing aldehyde acid or ester.These cleavage reactions are of mechanistic interest since the protic cleavage of many transition metal-carbon bonds is believed to occur by oxidative addition followed by reduction involving the loss of R,C-H species a process clearly not likely with docomplexes. This has been examined with optically active carbon substituents and found to occur with retention of configuration at C Cp,(CI)ZrCHDCHDBu' + Br2 benzene BrCHDCHDBu' (15) for the reaction in equation ( 15).19d (Halogen cleavage of alkyl-iron'" and -cobaltlgf complexes has been studied and found to occur with inversion at the C-centre.) The structure of the transition state remains in doubt but it is possible that the electronic configuration (16-electron) may favour a frontside attack at the Zr-C bond.'9d Sulphur dioxide also inserts into the C-Zr c+-bond (of the same asymmetric compound) with retention of configuration whereas reaction with CpFe(CO),CHDCHDBu' took place with > 95% inversion.19g Insertion into the l9 (a)C. A. Bertelo and J. Schwartz J. Amer. Chem. Sac. 1975,97,228; (6)G. Fachinetti and C. Floriani J. Orgunometaflic Chem. 1974,71 C5;(c)G. Fachinetti and C. Floriani J.C.S. Chem. Comm. 1972 654; (d)J. A. Labinger D. W. Hart W. E. Seibert and J. Schwartz J. Amer. Chem. Soc. 1975,97,3851; (e)P. L. Bock D. J. Baschetto J. R. Rasmussen J. P. Demers and G. M. Whitesides ibid. 1974 96 2814; (f) F. R. Jensen V. Madan and D. H. Buchanan ibid. 1971,93,5283; (g)See G. M. Whitesides and D. J. Boschetto ibid. 1971,93,1529; (h)D.W. Hart T. F. Blackburn and J. Schwartz ibid.,1975 97 679. 186 D.J. Cardin and K. R. Dixon Zr-H bond of Cp,ClZrH also occurs with disubstituted acetylenes. When unsym- metrical acetylenes are employed a mixture of products is obtained initially [equa- tion (16)] dependent on the bulk of the s~bstit~enfs.~~" The initial product mixture c1 / Cp,(Cl)ZrH + R'CECR' +Cp2Zr / R' (16). H 'R2 + slowly (at room temperature) rearranges to a composition of higher regioselectivity a process catalysed by the initial zirconium hydride. Since pure vinylzirconium derivatives undergo no such rearrangement whereas it is known to occur for the alkyls the process is envisaged as taking place through the alkyl species ZrR'CH-CHR'Zr formed by double hydrozirconation of the acetylene.8 Sixteen-electron Compounds It has often been remarked that in their organometallic compounds the transition elements 'obey' the effective atomic number rule (especially in the middle of the series) with surprising frequency sometimes adopting unexpected structures. Two recent examples are shown in (4)20a and (5).20b Perhaps more surprising are the * ._-/Mo\\PPhMe, PhMe2PMe2PhP (4) (5) well-known pyrazolylborato-complexes of molybdenum. Sixteen-electron com- plexes of Mo or W were prepared from CpM(Cl)(CO) and acetylenes RC,R (R =Me CF, or C0,Me) in which the (three) carbonyl groups were displaced by two acetylene The complex [Ph,B(pz),](CO),MoC,H,Me unlike related compounds,21 has 3-2-methylallyl carbonyl and pyrazolyl N-atoms as Mo ligands without additional metal co-ordination making it a 16-electron species.9 Silacyclopropanes Silacyclopropanes have long been known as intermediates in thermolyses and photolyses of organosilicon compounds. For example the photolysis of 2-phenyl-heptamethyltrisilane afforded a silacyclopropane which was trapped with 20 (a)J. Muller and H. Menig J. Orgunometullic Chem. 1975,96,83;(b)R.Mason K. M. Thomas and G. A. Heath J. Orgunometullic Chem. 1975,90 195; (c)J. L. Davidson M. Green D. W. A. Sharp F. G. A. Stone and A. J. Welch J.C.S. Chem. Comnt. 1974 706. 21 F. A. Cotton T. LaCour and A. G. Stannislowski J. Amer. Chem. SOC. 1974,% 754. Organometallic Compounds 187 methano1.22a*b The products of photolysis of the same trisilane with dimethyl- butadiene (Scheme 1) provide evidence for the formation of a 1,2-adduct (silacyclo- propane) (6),which undergoes further rearrangement either to give (7) which could have resulted from direct 1,4 attack or to give (8).The isolation of (8) is best 1. .Me r6 1 MePhHSiH,C \SZ w / Me-Si-Ph (8) Ph I (7) OMe (9) Scheme 1 explained by further reaction of (6),whose existence is also supported by trapping with methanol after irradiation affording (9).’” The thermolysis of phenyl- trimethylsilyldiazomethane which affords a benzosilacyclopentene had previously been reported to proceed via a silacyclopropane intermediate but this route has now been ruled out and it seems likely that the intermediates are the phenylcarbene and the cycloheptatrienylidene derived therefrom.22d The first stable silacyclopropanes were isolated in 1972 and proved to be highly reactive compared with larger rings containing silicon.22e Hexamethylsilacyclopropane has been synthesized [equation (17)] and is also stable under inert atmosphere but reacts with the protic species H20,NH3,or alcohols giving ring-opened products having Si-0 or Si-N bonds.*” Me,Si(CMe,Br) + Mg-THF +! / Si (17) ‘3 The new compound provides a low-temperature route to dimethylsilene.The latter has been identified by several techniques including trapping by a number of silanes. The previously prepared silacyclopropanes did not afford silenes on thermolysis but gave dimeric and polymeric species.22e.g MO calculations on 7-siladispiro[2,0,2 llheptane suggest increased Si-C bond strength due to hypercon- jugative effects and further that d-o-hyperconjugation may also contribute sig- nificantly to the strengthening of the Si-C bond in the (unknown) cyclopropylidenesilanes.’’h The tetraphenyl-substituted boracyclopropenyl anion has been proposed in reactions following the photolysis of sodium tetraphenylborate in the presence of tolan.22i 22 (a)M.Ishikawa M. Ishigino and M. Kumada J. Orgunomefullic Chem. 1973,49 C71; (b)M. Ishikawa and M. Kumada ibid. 1974,81 C3; (c)M. Ishikawa F. Ohi and M. Kumada ibid. 1975,85 C23; (d)J. J. Barton J. A. Kilgour R. R. Gallici A. J. Rothschild J. Slutsky A. D. Wolf and M. Jones J. Amer. Chem.Soc. 1975,97,658; (e)R. L. Labert and D.Seyferth ibid. 1972,94,9246;(f)D. Seyferth and D. C. Annarelli ibid. 1975,97,2273;(g) D. Seyferth and D. C. Annarelli ibid. 1975,97,7162; (h)P. D. Mollere and K. Hoffman ibid. 1975,97 3680; (j)J. J. Eisch K. Tameo and R. J. Wilcsek ibid. 1975 97. 895. D.J. Cardin and K. R. Dixon 10 Main-group Organometallic Ions Although siliconium ions (five-co-ordinate) have been described as intermediates and attempts have been made to trap silicenium ions (three-co-ordinate) the existence of the latter had not been established prior to 1975. The ion (10;M = Si) has now been trapped following reaction between the silepin (10; M = SiH) and an equimolar quantity of triphenylcarbenium perchlorate in CH,Cl at ca. -50 0C.230 The reaction produces a coloured solution which reacts with NaBD,-diglyme or NaBH,-H,O-dioxan to yield the silepin (Si-D) or a mixture of this (Si-H) and the silanol respectively.The substituted ferrocenylsilicenium ion has also been n N Me2 (10) Another interesting group of organometallic ions consists of boron-substituted carbanions. These have been suggested as intermediates in a number of reactions including the base-induced deboronation of gem-diboryl species which has been employed in the synthesis of C=C compounds from C=O ones. The ion (11) [equation (IS)] has been isolated and characterized (including n.m.r. data) and shown to react with a variety of carbonyl compounds affording substituted ethylene~.,~' 11 Double-bonded Silicon The pyrolysis of silacyclobutanes and production of intermediates containing double bonds to silicon were reported by Gusel'nikov and colleagues in 1966,24" and a short review of this area has now been published by the same group.24b The subject continues to excite interest concerning both the mechanistic details and reactivity and the identification of new Si=E bonds.TWOgroups have reported studies on 23 (a) J. Y. Corey J. Amer. Chem. SOC.,1975 97 3237; (b) J. Y. Corey D. Gust and K. Mislow J. Organometallic Chem. 1975 101,C7; (c) D. S. Matteson and L. A. Magelee ibid. 1975,93 21. 24 (a)N. S. Nametkin. V. M. Vdovin L. E. Gusel'nikov and V. I. Zav'yalov Zzvest. Akad. Nauk S.S.S.R. Ser. khim. 1966,589; (b)L. E. Gusel'nikov N. S. Nametkin and V. M. Vdovin Accounts Chem. Res. 1975,8 18; (c)T.J. Barton G. Marquardt and J. A. Kilgour J. Organornetulfic Chem. 1975,85,317; (d)C. M. Golino R. D. Bush P. On and L. H. Sommer J. Amer. Chem. SOC.,1975,97 1957; (e)C. M. Golino R. D. Bush and L. H. Sommer ibid. 1975 97 7371; (f) L. H. Sommer and J. McLick J. Organometallic Chem. 1975 101 171; (g) Yu. A. Ustynyuk P. I. Zakharov A. A. Azizov G. A. Shchembelov and I. P. Gloviozov ibid. 1975 96 195. Organometallic Compounds thermolysis of 2-substituted 1,l-dimethylsilacyclobutanes and conclude that C-C rather than Si-C bond cleavage OCCUTS.~~~,~ The products from both cleavage mechanisms are shown in Scheme 2 and the predominance of products formed via path b shows C-C rupture while the magnitude of the substituent effect makes the alternative (Si-C) bond breakage an unlikely initial RCH + Me,% -II II a CH CH2 RCH +CH *I 11 II \ R bSiMe -3 SiMe CH Scheme 2 The unsubstituted silaethene appears to be generated in the pyrolysis (560 "C N2 flow system) of silacyclobutane.It polymerizes in the absence of traps; reactions with a number of reagents are presented in Scheme 3.24e HZSi3 + H,Si=CH + high mol. wt. polymer /2sio)3 1 \ SiH2 -CH2 Ph2Co \N CISiH,CH,SiCI \ SiMe Ph,C=CH + Ph2CH2 ? I I I SiMe /0 49 % '% CH,SiH,CH,CN \ O-SiMe 3% 18 % Scheme 3 Evidence for the first Si=S intermediate comes from the reaction of thioben-zophenone with pyrolysis products of 1,l-disubstituted silacyclobutanes the prop- osed reaction sequence is given in Scheme 4.24f Me2Si3-+[CH =SiMe2] Ph2C=S b [Ph,C!-; CH,-SiMe 1 -I CH J/ S Me,S/ 'SiMe, p,A,h + [Me,Si=S] -+ \/ S Scheme 4 No isolable Si=C species has yet been described but the synthetic chemist is challenged by the suggestion based on CND0/2 calculations that 6,6-dimethyl-6- silafulvene is potentially such a stable 12 Silatranes and Stannatranes The discovery of neurophysiological effects in tricyclic esters having the structure (12; M =Si) has provided impetus for further chemical studies.Numerous syntheses D.J.Cardin and K.R. Dixon of substituted derivatives of the silicon ('silatrane') compounds have been pub- li~hed,~~",~ as has a new route equation (19) to the tin analogues ('~tannatranes').~~~ The tin compounds also have five-co-ordinate metal although n.m.r.evidence has been interpreted in terms of equilibrium between four- and five-co-ordinate species in s01ution.~~~ RSn(OEt) + N(CH,CH,OH) + RSn(OCH,CH,),N + 3EtOH (19) The acid-catalysed solvolysis of 1-organosilatranes has been studied. It is first- order in silatrane and probably goes by initial protonation of the nitrogen with breaking of the Si-N bond but the protonation is almost complete at the transition state.25d &>O'M-0 "IOR (1 2) 13 YIides The chemistry of ylides in an inorganic context has been developed over the past few years and the chemistry of P As and S ylides and in particular the work of Schmidbaur's group has been reviewed.26" Relatively few arsonium ylides have been isolated and the first example which is not stabilized by carbonyl groups was recently prepared [equation (20)].26b This thermally unstable compound gives on (Ph,AsMe)Br + NaNH Ph,As=CH f NaBr + NH3 (20) decomposition triphenylarsine polymethylene and ethylene among the products.The 'J(CH) coupling (compared with the arsonium cation) suggests that little rehybridization occurs on the CH -+CH2 change i.e. that the carbon atom in the methylide has pseudo-tetrahedral geometry. The same ylide has also been obtained using NaH and found to react with a number of acyl halides with transylidation.26c3d Interest in [Ph3P=N=PPh3]+X- and in the di-ylide Ph,P=C=PPh, stemmed from the unusual physical properties the PNP angle is dependent on the anion and the ylide is triboluminescent probably as a result of a delicate balance of energy levels associated with the angular deformation.The methyl analogue can be prepared as shown in equations (21)-(23).26' The di-ylide is a very air-sensitive material a strong base and a stronger nucleophile than mono-ylides.26' 25 (a)M. G. Voronkov V. M. D'yakov and L. I. Gubanova Zhur. obshcheiKhim. 1975,45,1901,1902 1903 1904 1905; (b) E. Popowski M. Michalik and H. Kelling J. Orgunomefulfic Chem. 1975 88 157; (c)M. Zeldin and J. Ochs ibid. 1975,86,369;(d)A. Daneshrad C. Eaborn R. Eidenschenk and D. R. M. Walton ibid. 1975,90 139. 26 (a)H. Schmidbaur Accounts Chem. Res. 1975,8 62; (b)Y. Yamamoto and H. Schmidbaur J.C.S. Chem. Comm. 1975,668; (c)P. S. Kendurkar and R. S. Tewari J.OrganometallicChem. 1975,102 141; (d)P. S. Kendurkar and R. S. Tewari ibid. p. 173; (e)0.Gasser and H. Schmidbaur J. Amer. Chem. SOC.,1975,97,6281; ('j) H. Schmidbaur,H. J. Fuller and F. H. Kohler J. Orgunometullic Chem. 1975,99,353;(g) H. Schmidbaur and R. Franke Znorg. Chim. Actu 1975,13,85; (h)H. Schmidbaur and R. Wolfgang Chem. Ber. 1975,108,2659; 6)E. Kurras U. Rosenthal H. Mennenger G. Oehme and G. Engelhardt 2.Chem. 1974,14 160; (k) E. Kurras H. Mennenger G. Oehme U. Rosenthal and G. Engelhardt J. Organometallic Chem. 1975,84 C13. Organometallic Compounds 2Me,P=CH + Me,PF2 -+ Me,P=CHPMe,F + (Me,P)F (21) or Me,P=CHSiMe + Me,PF + Me,P=CHPMe,F + Me,SiF (22) Me,P=CHPMe,F + NaH + Me,P=C=PMe + H + NaF (23) Reaction of phosphorus ylides with trialkyl derivatives of Ga In or Tl,has been found to yield phosphonium betaine structures e.g.Me3P'-CH,-GaMe3.26f The use of ylides in the preparation of metal alkyls is well known. Recent examples are shown in equations (24)-(26) including extension to the relatively novel arsenic ylides.L6g*h rAU -7 Me,PAuCl + Me,P=CH + Me,P PMe LAuJ Me,As=CHSiMe + CuCl -+ CU I -+ Me,AsAgCl + Me,As=CH Ag/-AsMe,\ I Me As An interesting related reaction is the synthesis of homoleptic alkyls of chromium26i and using phosphonium salts and alkyl metal anions [equation (27)]. The molybdenum species are less air-sensitive and reactive and react with Li,Mo,Me + 4(Me,P)CI + Mo,[(CH,),PMe,] + 8MeH + 4LiCl (27) acetic acid to give molybdenum(r1) acetate.The probable structure involves bridging alkyl ligands in an arrangement similar to the acetate. 14 Chiral Metal Centres The synthesis of organometallic compounds with chiral metal centres continues to receive attention particularly for main-group elements. Many of the syntheses use classical methods involving separation of diastereomeric pairs but the hydrosilyla- tion of (-)-menthone or (+)-camphor using Wilkinson's catalyst or an analogue having optically active phosphine ligands has been reported.27" The alkoxysilanes are obtained in up to 82% optical purity and can be converted into alkyl silanes using 27 (a)R. J. P. Corriu and J. J. E. Moreau J. Organometaffic Chem. 1975,91 C27; (b)R. J. P. Corriu F. Larcher and G. Royl J. Organometaffic Chem.1975 102 C25; (c) J. Tirouflet A. Dormond J. C. Leblanc and F. Le Moigne Tetrahedron Letters 1373 257; (d) F. Le Moigne A. Dormond J. C. Leblanc C. Moise and J. Tirouflet J. Organometuffic Chem. 1973,54 C13; (e)C. Moise J. C. Leblanc and J. Tirouflet J. Amer. Chem. Soc. 1975,97,6272; (f) J. C. Leblanc C. Moise and T. Bounthakna Compt. rend. 1974,278,973; (g) G. Simonneaux,A. Meyer and G. Jaouen J.C.S. Chern. Comm. 1975 69. 192 D.J. Cardin and K. R. Dixon Grignard reagents with retention of configuration. The first optically active bifunctional silanes in which the Si atom is the only chiral centre a-naphthylferrocenylfluorosilane and a-naphthylferrocenylchlorofluorosilane,have been prepared. Fluorine was introduced by reaction of menthoxysilanes with BF3 which goes with inversion of configuration; chlorine was attached with retention using PdC1 and ~ilanes.~~’ Relatively few early transition-metal compounds with chirality at the metal have been reported partly because the mechanism and steric course of reactions are not always straightforward particularly at octahedral metals.Full details are now available of the first resolved titanocene derivative of which preliminary reports have a~peared.,~‘,~ The preparation is outlined in Scheme 5,27eand is based on the separation of diastereoisomers. The second chiral centre was removed by HC1 cleavage which has been establishedz7’ to be selective and stereospecific. CpTiC1 + Cp’ + CpCp’TiC1 C6F5MgBr + ( f)-CpCp’Ti(C6F,)C1 ( -)-S-2-phenyl-propan-1-01 CP\ ,C& I HCI-benzene CpCp’Ti(C,F,)OCH,CHMePh /Ti\ CP‘ CI Scheme 5 The first optically active chromium(0) species enantiomeric at Cr have been prepared from optically pure 1-S-[l-CO2Me-2-Me-(q6-C,H,)]Cr(CO), by sub- stitution of carbonyl by CS and (PhO),P followed by removal of the initial chirality by reduction with LiAlH,-AlCl,.The preparation of chiral arenechromium tricar- bonyls offers the possibility of catalytic asymmetric 15 Metal-Carbon Bond Strength Data Thermochemical data on alkyl and aryl derivatives of transition metals and derived strengths of v-C bonds do not in general support the once widely held view that such bonds are weak and $an only be sustained by suitable ligand combinations. Mean bond-energy terms for the bonds M-C in homoleptic metal alkyls (M =Ti Zr or Hf) with bulky ligands R range from 44 kcal mol-’ for Ti R =neopentyl to 75 kcal mol-’ for Zr R = CH,SiMe,.Zr forms stronger bonds with C (by ca. 15%) than Ti.28a It is interesting (i) that whereas bond strengths decrease upon descending a group of typical elements (e.g. Ge>Sn>Pb) the reverse trend is found in a transition-metal group (e.g.,Ti<< Zr <Hf) and (ii) that steric effects influence bond strengths (Me3CCH2 <Me,SiCH,). Microcalorimetric measurements have been 28 (a)M. F. Lappert D. S. Patil and J. B. Pedley 3.C.S.Chem. Comm. 1975,830; (b)F. A. Adedigi D. L. S. Brown J. A. Connor M. L. hung I. M. Paz-Andrade and H. A. Skinner J. Organometallic Chem. 1975,97,221; (c)J.Tamas G.Czira A. Mal’tsev andO. M. Nefedov MagyarKim. FolyWrat 1974,80 439 (Chem. Abs. 1975,82,24 962j); (d) R. A. Burnham and S. R. Stobart J. Organometallic Chem. 1975,86 C45. Organometallic Compounds 193 made on arenechromium tricarbonyl compounds and bond enthalpy contributions derived.286 These decrease as the donor power of the arene decreases from 55(4) kcal mol-' for the arene-metal bond of the hexamethylbenzene complex to 38(5)kcal mol-' for the corresponding bond of the chlorobenzene derivative. Mass spectrometric data were used to determine the dissociation energy of the Ge-C bond in Cl,GeMe which was found to be 66 kcal mo1-'.28c Metal-metal bond dissociation energies have also been obtained by the mass spectrometer technique for the species Me3M'M2(CO), as follows D(M'-M2)/kcal mol-' Ge-Co 73.8; Si-Co 64.6; Si-Re 71.5; Ge-Re 73.8; and Sn-Re 85.3.28d 16 Oleh Metathesis The remarkable catalytic disproportionation of olefins continues to attract consider- able attention and two reviews of the process have ap~eared.~~~'~ The question of mechanism in this process is one of the most fascinating aspects and several have been advanced including cyclobutane-metal complexes and metallocyclic species in the transition In 1972 a mechanism for the dismutation of electron-rich olefins was advanced which involved one-carbon transfers and a metal-carbene was identified as an intermediate.29c However the unusual nature of the olefins and the low activation energy of the process for other olefins left it an open question as to whether the mechanism would have wider connotation.Two groups have reported findings which do lend support to this type of mechanism. The distribution of deuterium in the products of metathesis of [1,l,8,8-2H,]octa- 1,7-diene with octa- l,7-diene using the catalyst systems WC1,-BunLi-benzene and PhWC1,-AlC1 is consistent (agreement is particularly good for the latter catalyst) with the single-C transfer idea.29d Similar conclusions have also been reached from determination of product ratios with time in an experiment in which cyclo-octene trans -but-2-ene and trans-oct-4-ene disproportionated with a Mo-A1 catalyst.29' The authors also point out that the formation of catenanes in the disproportionation reaction can be simply accounted for by cyclization of a terminal carbene with an internal double bond.Further experimental data will be needed before it is possible to say whether the dismutation of alkynes is likely to occur through carbyne intermediates. The photochemical generation of an olefin dismutation catalyst requiring no co-catalyst has been Now photolysis of W(CO) in CCI has been found to give Cl,W(CO), which catalyses the reaction upon thermal or photochemical activati01-1.~~~ A brief description of the olefin metathesis reaction interpreted as a synchronous process has been given.29i 17 Metal-Metal Bonds and Clusters The chemistry of compounds containing intermetallic bonds is once again an area of great activity. (For metal carbonyl clusters see Section 20; see also Chapter 7 29 (a)R.J. Haines and G. J. Leigh Chem. SOC.Rev. 1975,4 155; (b)J. C. Mol and J. A. Moulijn Adu. Catalysis 1975,24 131; (c)D. J. Cardin M. J. Doyle and M. F. Lappert J.C.S. Chem. Comm. 1972 927; (d)R. H. Grubbs P.L. Burk and D. D. Carr J. Amer. Chem. SOC.,1975,97,3265; (e)T. J. Katz and J. McGinnis ibid. 1975,1592; cf) A. A. Agapiou and E. McNelis J.C.S. Chem. Comm. 1975 187; (g) P.Krausz F. Gamier and J. E. Dubois J.Amer. Chem. SOC., 1975,97,437 (h)A. A. Agapiou and E. McNelis,J. Organometallic Chem. 1976,99 C47; (j)F. D. Mango Coordination Chem. Rev. 1975,15 142. D.J. Cardin and K.R.Dixon pp. 154 and 161.) The chemistry of systems having quadruple and and other metal- metal bonds of high order has been re~iewed.~'" A number of bimetallic and metal cluster compounds particularly with metals of Main Groups I1and I11 have bridging carbonyls of novel geometry.The crystallographic identification of a carbonyl group bridging two Mn atoms with both C and 0interacting with one of the metal centres was noted for the first time (for details see In certain situations the spectroscopic identification of bridging carbonyls may not be simple. Reaction of trimethylaluminium or dimethylaluminium hydride with CpW(CO),H gives [CpW(CO),AlMe212. This compound contains rings of 12 atoms (WCOAIOC), each aluminium having two methyl substituents and the tungsten atoms an additional carbonyl and the cyclopentadienyl group. The rings are easily cleaved by protic species HX affording the tungsten hydride and (Me,AlX), while donors trimethylamine and diethyl ether yield the adducts of the AlW monomer The structure of the ring compound is unusual in that two of the carbonyl bridges are very close to hear (173-176").In the reaction of trimethylgallium with the same tungsten h ydride the compound CpW(CO),GaMe is formed,30d and this has been shown to have a Ga-W bond.30' In hot hydrocarbon solution decomposition to yield [CpW(CO),],GaMe and [CpW(CO),],Ga occurs neither of which like [CpW(CO),],In,30f shows i.r. bands due to bridging carbonyl ligands. The four-metal gallium system has three Ga-W bonds lying in a plane the molecule (apart from the cyclopentadienyl rings) having roughly C3h Magnesium resembles aluminium in forming several compounds linked to a transition metal via a Mg-0-C-M bridge and e.g.the crystal structure of (py)4Mg[Mo(CO)3Cp]2 has established this type of bonding in a trans-arrangement about approximately octahedral magnesium.3og The conditions favourable for the formation of M-C-0-Mg bonds have now been examined and several useful synthetic approaches described.,Oh The solubility of the magnesium compounds in hydrocar- bon solvents makes them potentially very useful in metallation reactions where polar solvents are not suitable. It has been remarked that lithium is capable of stabilizing unusual transition- metal compounds notable among which are polynuclear species such as Li,Cr2Me,,4THF,30' Li6Ni2N2(C6&)2,2Et20,30k and L~,MO,C,~H,,.~~' In such structures the interactions of the organic groups with the lithium atoms may be more significant than the base-lithium and it is noteworthy that in a neutron and X-ray diffraction study of LiBMe the bridging of lithium atoms by both linear Li-CH,-B and double and triple H-bridges is found.30m The authors note the well-known reactivity of C-H when p to early main-group or transition metals and point out that electron-deficient structures can become co-ordinatively saturated by the interaction of alkyl H atoms with metals.In this respect lithium shows strong Lewis 30 (a)F. A. Cotton Chem. SOC.Rev. 1975,4,27;(b)R. Colton C. J. Commons and A. F. Hoskins J.C.S. Chem. Comm. 1975 363; (c) A. J. Conway G. J. Gainsford R. R. Schrieke and J. D. Smith J.C.S. Dalton 1975 2499; (d) A.J. Conway P. B. Hitchcock and J. D. Smith ibid. p. 1945; (e) Personal communication to the authors of ref. 30d by J. P. Oliver; (f) A. T. T. Hsieh and M. J. Mays J Orgunometallic Chem. 1972,37,9; (g)S. W. Ulmer P. M. Sharstad J. M. Burlitch and R. E. Hughes J. Amer. Chem. SOC.,1973,95,4469;(h)G. B. McVicker Inorg. Chem.,1975,142087;(j)F. Mein and K. Schmiedenknecht J. Organometuffic Chem. 1966,6,45,and refs. therein; (k)K. Jonas Angew. Chem. Zntemat.Edn. 1973,12,997171)C. Prout and M. L. H. Green J.C.S. Chem. Comm. 1973,259; (m)W. E. Rhine G. Stucky and S. W. Pederson J. Amer. Chem. SOC. 1975 97 6401; (n) F. Armitage 'Inorganic Rings and Cages' E. Arnold London 1972 and refs. therein. Organometallic Compounds 195 acidity towards the weakly basic hydrogen atoms of hydrocarbon chains and in the structure investigated each lithium atom has 10 hydrogen atoms within its first co-ordination sphere [6 at 2.234(10) A and 4 at 2.115(8) A]?0m The tetrahedral P can be substituted by various groups affording cluster com- pounds of which those structurally examined are edge-substituted species (see Chapter 6 p.121).30" 18 Shorter Topics The structure of the dilithium derivative of the hexatriene dianion involves a planar geometry for four hydrogens and carbons of the anion and lithium interactions with four of the C atoms.31 Two structural studies of bis(cyclopentadieny1)magnesium have appeared; the gas-phase electron diffraction data indicate eclipsed cyclopen- tadienyl rings but do not completely exclude the staggered while X-ray data for the crystal at least show a staggered ring a~rangement.~~' Mic-rowave examination of CpBeH and isotopically labelled species shows a dipole moment of 2.08(1) D and a Be-H distance of 1.32(1) A the molecule having the expected CSu The existence of Grignard analogues for strontium and barium has been suggested in the literature and THF adducts have now been reported obtained from the finely divided metal and alkyl iodides at -78 "C.The compounds have low solubility and thermal stability but are solvolysed with pro- duction of alkane using protic solvents and afford low yields of tertiary alcohols on treatment with The reaction of indene with calcium bis(tetraethy1)ala- nate gives an indenylaluminium compound (LAlEt,)Ca(AlEt,) (L= indenyl) which disproportionates easily losing Et3A1.336 Perfluoroalkyl iodides react with calcium amalgam to give the perfluoroalkylcalcium iodides as evidenced by their reactions with carbonyl compounds which in some cases proceed in good yield suggesting that the reagents may have potential value as perfluoroalkylating species.33c When the Grignard reagents RMgX (R = 3-phenyl-1-propyl or 1-phenyl-2-propyl) are irradiated in the Mg-C chromophore region (ca.254 nm) terminal olefins and HMgX are formed the latter disproportionating. The elimination reaction is 99% (or more) a p-elimination process established by D-labelling Several structural studies of tin(r1) compounds appeared during 1975 mostly of compounds with co-ordination number for Sn of 4 or more.An interesting com- pound is (C61-&)Sn(AlC1,),C6H, in which the tin atom has approximate pentagonal- bipyramidal co-ordination one axial position being the centre of a benzene ring making appropriate the description 'r-benzene complex of tin(11)'.~~ N.m.r. studies of the dynamic equilibrium in Cp,M(BH,) (M = Zr or Hf) and Cp,Zr(H)BH have revealed a new type of borohydride exchange in which hyd- rogens of the cyclopentadienyl ring exchange with those of the BH4 group. The 31 S. K. Akora R. B. Bates W. A. Beavers and R. S. Cutler J. Amer. Chem. SOC. 1975,97,6271. 32 (a)A. Haaland J. Lusztyk J. Brunvon and K. B. Starowieyski J. Orgunometullic Chem. 1975,85,279; (b) W. Biinder and E. Weiss ibid. 1975 92 1; (c) T.C. Bartke A. Bjoerseth A. Haaland K. M. Marstokk and H. Moelendal ibid. 1975 85 271. 33 (a)B. G. Gowenlock W. E. Lindsell and B. Singh J. Orgunometullic Chem. 1975,101 C37 (b)L. I. Zakharkin Y. S. Zavizion and L. L. Ivanov Zhur. obshchei Khim. 1975,45,1900;(c)G.Santini M. Le Blanc and J. G. Riess J.C.S.Chem. Comm. 1975 678; (d) B. 0.Wagner and G. S. Hammond J. Organometallic Chem. 1975 85 1. 34 P. F. Rodesiler T. Auel and E. L. Amma J. Amer. Chem. Soc. 1975,97 7405. 196 D.J. Cardin and K. R. Dixon predominantly unimolecular process is thought to occur via q '-C,H,M carbene (or ylide) intermediates. It also occurs but more slowly in the solid phase for Cp2Zr(BD,),.35" Reactions between Cp2ZrC12 and Et3A1 have been reported to yield a range of alkyl aluminium and zirconium species including one with a 16-membered ring containing four Al and two Zr atoms.Structural evidence based on n.m.r. spectra is Monomeric paramagnetic carbonyls of vanadium are a unique group and to date no dimeric carbonyl derivatives can be regarded as established. However the chemistry of these species has been little studied owing to the lack of a reactive and suitable starting material. Photosubstitution of the carbonyl anion [v(co)6]-by o-phenylenebis(dimethylarsine) (diars) affords the thermally stable anion [V(CO),(diars)]- which affords the first alkyl or 3-allyl carbonyl derivatives of V on treatment with methyl iodide or ally1 chloride re~pectively.~~" The ion Cp2V+ which is isoelectronic with the monomeric form of titanocene has been isolated from aqueous media in the form of adducts [VCp2L]' (L= H20 acetone or pyridine).The .rr-basicity of these species is shown by their ready conversion into the known cation [CP~V(CO)~]+ and also into isocyanide or phosphine complexes.366 The preparation of phenylvanadium dichloride by reaction of diphenyl-lithium with vanadium(1v) chloride has been reported but the solid appears to be a co-ordination p01ymer.~~~ Niobium(v) chloride reacts with the cyclo-octatetraene anion to yield M'[Nb(cot),]- the first cyclo-octatetraene complex of the Group VA metals. A crystal study shows a trigonal arrangement of the ligands with two q3-bonded and one q4-bonded although the molecule is fluxional in solution. Reactions with phosphines and hydrogen afforded no characterizable vanadium compounds although with CO salts of the [V(CO),]- anion were is~lated.~'" A new low- pressure synthesis of Cp,NbH from Cp,NbCI has been reported to give yields up to 55% after hydrolysis with aqueous NaOH.The insertion of acetylenes into the Nb-H bond and the reactions of the vinyls so formed are described.37b The reversible addition of carbon monoxide (2 mol) across a Mo-Mo bond is the first reported example of this type of process [equation (28)]. The forward reaction (C,Me5)(CO),Mo~Mo(CO),(C,Me,)+ 2CO -+ (C,Me,)(CO),Mo-Mo(CO),(C5Me5) (28) can be induced by visible light or thermall~,~~" while there is precedent for the The reverse reaction which is also photochemical probably involves cleavage of the Mo-Mo bond and photolysis of the hexacarbonyl species in carbon tetrachloride yields the species (q5-C5Me5)Mo(CO)3CI.38Q Phosphines or phos- phites also react with the tetracarbonyl complex (with C,HJ contrary to earlier 35 (a)T.J. Marks and J. R. Kolb J. Amer. Chem. SOC. 1975 97 2397; (b)W. Kaminsky and H. Sinn Annalen 1975 429; (c)W. Kaminsky and H. J. Vollmer ibid.,p. 438. 36 (a)J. E. Ellis and R. A. Faltynek J. Orgunometallic Chem. 1975 93 205; (b)G. Fachinetti and C. Floriani J.C.S. Chem. Comm. 1975,578; (c)S. Schroeder A. Lachowitz and K. H. Thiele Z. anorg. Chem. 1975,415 104. 3' (a)L. J. Guggenberger and R. R. Schrock J. Amer. Chem. SOC. 1975,97,6693; (b)J. A. Labinger and J. Schwartz ibid. p. 1596. 38 (a)D. S. Ginley and M.S. Wrighton J. Amer. Chem. SOC.,1975,97,3533;(b)P. Hackett P. S. O'Neill and A. R. Manning J.C.S. Dalton 1974 1625; (c)R. J. Klinger W. Butler and M. D. Curtis J. Amer. Chem. SOC.,1975,97,3535;(d)T. Ito and A. Yamamoto J.C.S. Dalton 1975,1398;(e)D. L. Lewis and S. J. Lippard J. Amer. Chem. SOC. 1975,97 2697. Organometallic Compounds 197 Acetylenes react forming bridged dimolybdenum derivatives. The X-ray structure determination of the tetracarbonyl compound with unsubstituted cyclopentadienes confirms the triple-bonded formulation MorMo = 2.448( 1)A.38c (For further photochemistry of metal carbonyls see Section 20.) The reactions of Mo(C,H,)(dppe) with a number of reagents have been examined among which the contrasting thermal {to give cis-[Mo(CO,(dppe),]} and photochemical (to give an uncharacterized CO adduct) behaviour with carbon dioxide is The structure of [(Bu'NC),Mo]+PF,- shows the same co-ordination polyhedron as had been established for the iodohexakis(iso-cyanide)molybdenum(iI) cation i.e.a monocapped trigonal There is much evidence to suggest that tungstenocene is an intermediate in the reactions (both thermal and photochemical) of a number of species Cp2WX (X includes H2 CO and Cl,) and is highly reactive for example inserting into the C-H bonds of mesitylene or p-~ylene.~~" The photo-induced insertion of tungsten into methanol has been shown to give both Cp,W(H)OMe and Cp,WMe(OMe) probably via the insertion of tungstenocene into the 0-H and C-H bonds of methanol re~pectively.~~" Alkyl- and acyl-pentacarbonyltungsten anions have been obtained by two routes treatment of the corresponding halide anions with organolithium reagent and photolysis of the appropriate neutral species (OC),WCOR.39b Reactions with HCl trityl tetrafluoroborate CO or PPh are reported for some of the anions.The reaction of dibenzylmagnesium with tungsten tetrachloride affords tetrabenzyltung~ten.~~" Other reports on homoleptic metal alkyls and organometal- lic compounds in high oxidation states (including tungsten) are covered in Section 22. 19 Books and Reviews Main-Group and Early Transition Elements Among many review articles on the subject matter of this Chapter the following are among the most directly relevant. Matteson's book on organometallic reaction mechanisms of the non-transition elements presents a critical review of its subject including the elements B and Si but not P.40 Other books include works on organometallic reaction^,^^ and organometallic compounds of the transition ele- ments and related aspects of catalysis.42A new edition of 'Baiant' has appeared reviewing the Si literature since 196 1.43 The organometallic chemistry of the main-group elements has been reviewed,44 and articles dealing with aspects of this area cover synthesis and reactions of organo-lithium reagents derived from weakly acidic C-H and 3y (a)L.Farrugia and M. L. H. Green J.C.S. Chem. Comm. 1975,416 and refs. therein; (6)C. P. Casey S. W. Polichnowski and R. L. Anderson J. Amer. Chem. Soc. 1975,97,7375; (c) K.H. Thiele A. Russek R. Opitz B. Mohai and W. Brueser Z. anorg. Chem. 1975,412 11. 40 D. S. Matteson 'Organometallic Reaction Mechanisms of Non-transition Elements' Academic Press New York 1974. 41 'Organometallic Reactions' ed. E. I. Becker and M. Tsutsui Wiley New York 1975. 42 B. L. Shaw and N. L. Tucker 'Organo-transition Metal Compounds and Related Aspects of Catalysis' Pergamon New York 1975. 43 V. Baiant J. Hetflejs V. Chvalovsky J. Joklik 0.Kruchna J. Rathousky and J. Schraml 'Handbookof Organo-silicon Compounds. Advances since 1961' Vol. 1 Dekker New York 1975. 44 J. D. Smith and D. R. M. Walton Adu. Organornetallic Chem. 1975,13,453. 45 D. Ivanov G. Vasilev and I. Panaiotiev Synthesis 1975 83. 198 D.J. Cardin and K.R. Dixon cy~lopolyarsines.~~ In the series edited by Nesmayanov and Kocheskov the organic chemistry of Tc,~’ Ta,48 lanthanides,”‘ Nb,49 Hf,” V,’ and ZrS2 has been described. The Journal of Organometallic Chemistry annual review series continues. Organometallic compounds with bonds between transition metals and elements of Group IIIB have also been s~rveyed.’~ 20 Metal Carbonyls Matrix Isolation of Radical Species other Metal Carbonyl Transients.-Interest in paramagnetic carbonyl complexes and in unstable carbonyls in general has acceler- ated in recent years with the development of methods for treating metal atoms with various substrates (see also Section 2) and also with the development of carbonyl photochemistry. Mn(CO) in particular has been of interest to many research groups.54 Originally suggested as a precursor of [Mn(CO),]’ in the mass spectrum of pyrolysed Mn,(CO),, Mn(CO) had been proposed as an intermediate in a number of chemical reactions of Mn,(CO),, and an e.s.r.spectrum obtained from 350 nm photolysis of Mn,(CO), in THF had been assigned to the pentacarbonyl. A report this year5’= disputes this e.s.r. assignment and suggests that the six-line spectrum is actually due to an Mn“ species [MII(THF)~],+. The spectrum is remarkable because of its very narrow hyperfine lines. An alternati~e~~’ approach to the problem via homolysis of the M-C bond of [RMn(CO),] (R =Me or PhCH,) has yielded e.s.r. evidence for the generation of Mn(CO),. U.V. irradiation in the presence of the spin trap nitrosodurene gives e.s.r.spectra of the nitroxides ArN(O)Mn(CO) and ArN(0)R (Ar =2,3,5,6-Me4C,H) but the spectrum of the free Mn(CO) radical has still not been [Similarly photolysis of vitamin BI2 coenzyme 5’-deoxyadenosylcobalamin or ethylcobalamin in aqueous medium in presence of Bu‘NO as spin trap has afforded the 5’-deoxyadenosyl(Bu‘)NO or Et(Bu‘)NO re~pectively.~~‘] The characterization of Mn(CO) has now been achieved by matrix-isolation technique~.’~ Using an Mn:CO ratio of 1:lo4or less condensed in an argon matrix the principal species is Mn(CO), and at higher Mn:Co ratios the dinuclear Mn,(CO) species are obtained. 1.r. spectra of the pentacarbonyl indicate a square-pyramidal structure and this observation completes the characterization of the series of pentacarbonyls of the first transition series from vanadium to iron (see Table I p.220).54An alternative trigonal-bipyramidal fcrm of Cr(CO) has been claimed to exist but it now seems likely that this is in~orrect.’~ Thus there is a clear 46 L. R. Smith and J. L. Mills J. Organometallic Chem. 1975 84 1. 47 A. A. Ioganson K. N. Anisimov and N.E. Kolobova Metody. Elem.-Org. Khim. 1974 2,851. 48 A. A. Pasinskii Metody Elem.-Org. Khim. 1974 1,453. 49 A. A. Pasinskii Metody Elem.-Org. Khim. 1974 1 434. so E. M. Brainina Metody Elem.-Org. Khim. 1974,1 373. s1 A. A. Pasinskii Metody Elem.-Org. Khim. 1974,1 389. 52 E. M. Brainina Metody E1em.-Org. Khim. 1974 1 320. 53 A. T. T. Hsieh Znorg. Chim. Acta 1975 14 87. 54 H. Huber E.P. Kundig G. A. Ozin and A. J. Poe J. Amer. Chem. SOC.,1975,97,308,and references therein. 5s (a)A. Hudson M. F. Lappert and B. K. Nicholson J. Organometallic Chem. 1975,92 C11; (b) A. Hudson M. F. Lappert P. W. Lednor and B. K. Nicholson J.C.S.Chem. Comm. 1974,966; (c)K. N. Joblin A. W. Johnson M. F. Lappert and B. K. Nicholson ibid. p. 441. 56 J. D. Black and P. S. Braterman J. Amer. Chem. Soc. 1975,97 2908. Organometallic Compounds 199 correlation of stereochemistry with electron configuration and this is expected in terms of the generally accepted MO sequences for trigonal-bipyramidal and square- pyramidal complaes. Assuming low-spin configurations the 15-and 17-electron trigonal-bipyramidal forms should be subject to first-order Jahn-Tdler distortion towards a square-pyramidal form and the 16-electron complexes are stable to second-order Jahn-Teller distortions only in the square-pyramidal configuration.The 18-electron species should exhibit second-order instability in both D3hand C, forms and this is considered to account for the well-known facile interchange of axial and equatorial CO groups.54 Similar conclusions regarding the relationship between electron configuration and geometry in M(CO) species have been reached as part of a wide-ranging theoretical study of M(CO), M(CO), and M(CO) carbonyl frag- ments using extended Huckel calculation^.^^ Among the many topics covered by this study is a rationalization of the preferences of Fe(CO) and Cr(CO) fragments for bonding with conjugated and non-conjugated dienes respectively.In other matrix-isolation studies the range of binary carbonyls has been extended along the transition series to copper the species characterized being Cu(CO),- and cu,(co)6. The latter is the dinuclear carbonyl which would be predicted to follow Ni( CO),. 58 Photochemistry.-Metal-centred radicals of the Mn(CO) and related types may also be generated in photochemical processes. Of the large amount of reported work on photochemistry of metal carbonyl~,'~ relatively little is concerned with homolysis of metal-metal bonds to yield metal-centred radicals. For example previous studies on photolytic reactions of M2(CO), species (M=Mn or Re) have yielded both simple substitution products such as Mn,(CO),(PPh,) from Mn,(CO), and PPh3 and mononuclear products resulting from cleavage of the M-M bond such as ReCl(CO) from Re,(CO), and CCl,.These reports were qualitative in nature and quantitative studies giving definitive evidence for the primary photochemical proces- ses were not available. A series of studies reported this year6' presents evidence that the primary process in photochemical reactions of M,(CO), (M =Mn or Re) Mn2(CO)9(PPh3) MnZ(CO)8(PPh3)2 MnRe(CO)lO [(q5-C5HS)M(Co)312 (M =Mo or W) and (OC),M'M2(CO),(v5-C5H5) (M'=Mn or Re and M2=Mo or W) is homolytic cleavage of the metal-metal bond. Results for the [(~5-C,H,)M(CO)3] complexes are typical. These molecules have an intense fairly narrow near-u.v. absorption (ca. 20 000 cm-') which has no analogue in (q5-C,H,)M(CO),C1 com-plexes and is assigned to a one-electron transition which is essentially u +u*with respect to the M-M bond.Photolysis results in excitation of this transition with consequent bond homolysis. The principal evidence is (i) photolysis in CCl solution proceeds according to equation (29) where n is very near 2.0 and the chloride is the hv CCI (T 5-C~H5)2M2(C0)6-+ n (T '-C5HS)M(CQ3C1 (29) sole photoproduct; (ii) photolysis in the presence of Ph3CC1 gives e.s.r. evidence for formation of Ph3C* and in the presence of PhCH,Cl the products are PhCH,CH,Ph and (qS-C,H,)M(CO),C1; and (iii) photolysis of mixtures of [(r)5-C5Hs)M1(CO)3]2 57 M. Elian and R. Hoffmann Znorg. Chem. 1975 14 1058. 5R H. Huber E. P. Kundig M. Moskovits and G. A. Ozin J.Amer. Chem. Soc. 1975,97 2097. 59 M. Wrighton Chem. Rev. 1974,74 401 and references therein. 6o M. S. Wrighton and D. S. Ginley J. Amer. Chem. Soc. 1975,97 2065 4246 4908. D.J. Cardin and K. R.Dixon with Mi(CO), (M’ =Mo or W M’ = Mn or Re) gives all four M’-M2 species (q5-C5H5)M’(CO),-M’(CO)5 in high yields based on the disappearance of homonuclear species. The 17-electron intermediates generated in these photolytic reactions may be compared with the well-known and important catalytic species [Co(CN),I3-. Reac- tions of the latter with alkyl halides (RX) are known to proceed by radical pathways giving both [Co(CN),RI3- and [Co(CN),XI3-. The difference between this process and the reactions of M(CO) (M = Mn or Re) or (q5-C,H5)M(CO) (M = Mo or W) with PhCH2Cl to yield PhCH,CH,Ph and M(CO),Cl or (q5-C,H,)M(CO),Cl is attributed to the very low steady-state concentration of metal radicals in the latter experiments.This renders coupling of metal and organic radicals an unlikely process.6o The comparison with [Co(CN),]’- is strengthened by the discovery that Re(CO), generated by irradiation of Re,(CO), at 311nm is capable of activating molecular hydrogen. The proposed mechanism involves dissociation of CO from Re(CO), oxidative addition of H to the resulting Re(CO), species and subsequent reaction to ReH(CO), H,Re,(CO), and H,Re3(C0)l,.61 The apparent conflict between the above processes requiring M-M bond homolysis as the primary photochemical process and previous reports of heterolysis and substitution reactions is considered to be due to the possibility of secondary thermal reactions.For example Scheme 6 is proposed6’ to account for photosub- stitution in Mn,(CO)lo. An independent flash photolysis study6’ of [(q5-C,H,)Mo(CO),] is slightly at variance with these suggestions in that two inter- mediates were detected both of which react by independent thermal processes to ::32[Mn(CO),]%Mn,(CO), * 2[Mn(CO),PPh,] + 2CO Mn(CO) +‘Mn(C p3\ ),PPh3 r/A Mn2(C0)8(PPh3)2 Mn,(CO),PPh Scheme 6 regenerate the starting complex. One of these is assigned as (q’-C,H,)Mo(CO) and the other as (q5-C,H,),Mo,(CO), but it is not certain that the latter is a primary photoproduct.62 However the possibility of substitution via homolysis has been further supported by studies of the reactions of [Mn(CO),(PPh,)] with P(OPh) and of ReH(CO) with PPh3.63 Both reactions proceed by initial generation of radicals M(CO),L (M = Re L = CO; M = Mn L =PPh,) followed by associative ligand exchange at the reactive intermediate.This type of process was previously unknown for simple substitution reactions which had been considered to proceed by straightforward dissociative or associative pathways involving the substrate species and no radical intermediate^.^^ Fluxional Processes.-The possibility of ‘carbonyl scrambling’ as an important class of fluxional process was first pointed out in 1972 by Bullitt Cotton and Marks as a 61 J. L. Hughey C. R. Bock and T. J. Meyer J. Amer. Chem. Soc. 1975,97,4440. 62 B.H. Byers and T. L. Brown J. Amer. Chem. Soc. 1975,97,3260. 63 J. P. Fawcett R. A. Jackson and A. J. P& J.C.S. Chem. Comm. 1975 733; B. H. Byers and T. L. Brown J. Amer. Chem. SOC.,1975,97 947. Organometallic Compounds 20 1 result of studies on the carbonyl-bridged complex (q5-CSH5)2Fe2(C0)4. Since that time the ready availability of I3C n.m.r. has led to a rapid expansion of knowledge in this area the majority of studies being devoted to bridge-terminal interchanges involving migration of CO groups from one metal atom to another particularly in dinuclear species. Typical of these studies are the observations for (77’-CSHS)2Fe,(CO)3[P(OPh)3](13) and the similar results obtained by an independent group on the analogous triethyl phosphite complex.64 In the solid state the triphenvl phosphite complex has structure (13) but in solution i.r.and ‘H and 13C n.m.r. spectra show rapid interconversion between isomers having cis and trans arrangements of the cyclopentadienyl groups and simultaneous exchange of car- bonyls between bridging and terminal sites. The generally accepted mechanism for processes of this type involves two essential steps (i) concerted opening and closing of pairs of ligand bridges and (ii) hindered internal rotations in the non-bridged tautomers. If the above mechanism is correct then cis-trans isomer interconversion and bridge-terminal CO exchange in (13) should occur at the same rate. This is observed experimentally and the above mechanism is thus well established although no rationalization for the necessity of concerted pairwise bridge opening and closing has been pre~ented.~~ A survey of the stereochemistry of the compounds [Fe(q’- C,H,)(CO),Y] where Y can be a univalent group or may be a group capable of bridging to the iron has now shown that the geometry in these species is consistently close to octahedral.This observation implies a rigidity in the local stereochemistry at iron and one consequence of this is that the tautomer interconversion mechanism for [Fe($-C,H5)(CO)2]2 and related species must involve simultaneous making or breaking of two carbonyl bridge ~ystems.~’ The mechanism of CO transfer in systems such as (q5-CSH5)zRh2(C0)3 which involve only a single carbonyl bridge is less well established. Intermediates involving a triple CO bridge have been prop- osed but the facile CO site exchange in [(q5-C5H5)2Rh2(C0)2{P(OPh)3}], where such an intermediate requires a phosphite bridge renders this process unlikely.Moreover the 31P n.m.r. spectrum of the phosphite complex is temperature invariant indicating that the phosphite remains co-ordinated to one rhodium atom. The authors favour a mechanism involving formation of a new CO bridge synchron- ous with the breaking of the existing bridge.66 A fundamentally different type of carbonyl scrambling process has been demon- strated by a I3C n.m.r. study of (14).67Several previous experiments had indicated that in species having non-equivalent CO groups within M(CO) sets site exchange 64 F. A. Cotton L. Kruczynski and A.J. White Inorg. Chem. 1974,13,1402 and references therein; D. C. Harris E. Rosenberg and J. D. Roberts J.C.S. Dalton 1974 2398. 6s J. R. Miller and F. S. Stephens J.C.S. Dalton 1975 833. 66 J. Evans B. F. G. Johnson J. Lewis and T. W. Matheson J.C.S. Chem. Comm. 1975,576. 67 F. A. Cotton D. L. Hunter and P. Lahuerta Inorg. Chem. 1975 14 511 and references therein. D.J. Cardin and K. R.Dixon (14) by a process equivalent to rotation of the M(CO) fragment was a possibility. For the acenapthylene complex (14) this process may be unambiguously demonstrated even though no other fluxional process (e.g.‘ring whizzing’) is occurring. Between -60 and +45”C 13C n.m.r. spectra show that three CO groups a b and b’ are interchanging but there is no interchange with the groups c and c’.The process is essentially rotation of the Fe(CO) unit and the more normal bridge-terminal mechanism is prevented by the unusually long Fe-Fe bond (2.77 A). The complex Fe,(CO),(cycloheptatriene) also has an exceptionally long Fe-Fe bond (2.87 A) and is believed to represent another example of the same type of process although in this case definite proof is lacking.67 As noted above the majority of the CO scrambling processes studied in detail have involved dinuclear complexes. Observations on trinuclear complexes were limited to the prediction and confirmation that the barrier to total CO scrambling in Fe,(CO), is <5 kcal mol-l i.e. the system is still fluxional at -150 0C.68 Variable- temperature 13C n.m.r.of M3(CO)12 (M = Fe Ru,or 0s) and related complexes has now shown that several different scrambling processes can Fe,(CO), has the solid-state structure (15) and it is therefore probable that scrambling occurs via pair-wise bridge-terminal interchange of the type discussed above. This type of mechanism is also required for [HFe,(CO),,]- (16) since the unique bridging CO is not involved in the interchange. However an alternative 0 OCGCO p\ oc&c:o oc co OC -M ICO\,Lco oc /-\ oc co co (15) 68 F. A. Cotton and D. L. Hunter Inorg. Chim. Acta 1974 11,L9. Organometallic Compounds mechanism which is essentially a rotation of an M(CO) group similar to that discussed above for M(CO) groups is required by evidence on [Ru,(C~),,(N~),] (17).This species exhibits three I3C n.m.r. signals with relative intensity 4:4:2 indicating that localized exchange of axial and equatorial CO is occurring at the Ru(CO) group. Since the Ru(CO) groups are rigid CO-bridged intermediates are clearly not co OC\ / Similar fluxional processes have been observed for tetranuclear species. For example in H,FeRu,(CO), (18)at least three carbonyl exchange processes can be distinguished the most rapid (ca. -70 "C) localized at Fe the next (ca. -45 "C) localized at the three Ru atoms and the last general over all metal centres at +95 0C.70 13C N.m.r. studies have also prompted a re-opening of the long-standing argument as to the solution structure of Co,(CO),,. At -100°C three equal- intensity resonances corresponding to one bridging and two terminal environments are observed suggesting a structure with DZdsymmetry rather than the accepted structure with C3 In contrast the 13C n.m.r.spectrum of Co,(CO),,[P(OMe),] at -82 "C shows bridging terminal CO groups in a ratio of 3:8 as expected for a C, structure with one of the basal terminal CO groups replaced by phosphite. The differences are most probably due to the difficulties of interpreting the. relative intensities of n.m.r. lines affected by rapid s9Coquadiupolar re~axation.'~~ Work on CO scrambling in larger clusters has also been reported this year.' R~~(CO)~~ undergoes rapid is not fluxional at +70 "C whereas [R~~(CO),S]~- complete carbonyl scrambling even at -70 "C. The structure of the anion is derived from the neutral species by removal of one face-bridging carbonyl and this change is considered to make available intermediates involving minimal changes in symmetry and bonding.Another anion [Rh7(CO),,l3- is intermediate between the other two species being rigid at -70 "C but undergoing partial CO exchange at +25 0C.72 BridgingCarbonyl Structures.-The fundamental terminal and symmetrically bridg- ing modes of carbonyl bonding have been known for a very long time and more 69 A. Forster B. F. G. Johnson J. Lewis T. W. Matheson B. H. Robinson and W. G. Jackson J.C.S. Chem. Comm. 1974,1042. 70 L. Milone S. &me E. W. Randall and E. Rosenberg J.C.S. Chem. Comm. 1975,452. 71 (a)J. Evans B. F. G. Johnson J. Lewis and T. W. Matheson J.Amer. Chem. SOC.,1975,97 1245; (6) M. A. Cohen D. R. Kidd and T. L. Brown J. Amer. Gem. Soc. 1975,97,4408. 72 B.T. Heaton A. D. C. Towl P. Chini A. Fumagelli D. J. A. McCaffrey and S. Martinengo J.C.S. Chem. Comm. 1975,523. D.J. Cardin and K. R. Dixon h Ph recently two additional types of co-ordination have been shown to be relatively common. Co-ordinated carbonyls may form bridges to main-group or other transi- tion elements via oxygen bonding. Moreover as reported last year there exists a complete range of carbon-bridged types varying from the normal symmetrical bridge to the very unsymmetrical bridges exemplified by (19). There are also examples where only one of a pair of bridging carbonyls is asymmetric and others where a single asymmetrical carbonyl is the only bridging group.73 The product of reaction of Mn,(CO), with Ph,PCH,PPh (dpm) has the molecular formula Mn,(CO),(dpm), and appears to represent an important new type of carbonyl bridging stru~ture.'~ The manganese atoms are bridged by the dpm groups so that the two manganese and four phosphorus atoms are almost coplanar.Approximately perpendicular to this plane is a plane containing the carbonyl groups [see (20)]. The Mn-Mn bond is 2.934A comparable with that in Mn,(CO)l, and the bridging carbonyl is apparently bound by a rather long u-bond to Mn( 1) [Mn(l)-C is 1.93 compared with an average 1.69 A for the terminal Mn-C bonds] and via its .rr-electrons to Mn(2) [Mn(2)-C is 2.01 8 and Mn(2)-0 is 2.20 A].'" Bonding to Mn(2) is thus similar to olefin or acetylene bonding and may be compared to structures such as (21).75 Cluster Carbonyl Structures.-One of the interesting features of cluster chemistry is the intermediate position which these complexes occupy between molecular species and metal lattices.In the long term this relation could lead to important results in the field of catalysis. However most of the clusters studied to date have an essential difference from metallic lattices in that the metal atoms are arranged in one or other of the common co-ordination polyhedra and the centre of the polyhedron is unoccupied i.e. there is no metal atom co-ordinated solely by other metals. Exceptions to this statement such as the [Pt(SnCl3),l3- ion are fairly normal co-ordination complexes and the environment of the central metal ion bears little relation to that of a metal in a metallic lattice.However in recent years X-ray diffraction studies on gold clusters have been reported in which a polyhedron of gold atoms has another gold atom at its centre.76 The geometries of both [Au,{P(p- MeC,H,),},][PF,] and [Au l13{P(p-Fc6H4)3}] are derived from a centred icosahed- ron in the former case by removal of four atoms constituting an equatorial rectangle 73 F. A. Cotton and J. M. Troup J. Amer. Chem. Soc. 1974,96 1233 5070 and references therein. 74 R. Colton C. J. Commons and B. J. Hoskins J.C.S. Chem. Comm. 1975,363; C.J. Commons and B. J. Hoskins Austral. J. Chem. 1975,28 1663; R. Colton and C. J. Commons Austral. J. Chem. 1975,28 1673.75 €3. A. Patel R. G.Fischer,A. J. Carty D. V. Naik and G. J. Palenik,J. Olganometallic Chem.,1973,60 c49. Organometallic Compounds 205 and in the latter by replacement of one triangular face by a single gold atom.76 [Rh6(CO),,C]2- is the first metal cluster with a trigonal-prismatic arrangement of metal atoms and its oxidation with Fe3+ has yielded several interesting produ~ts.~' [Rh,(CO),,C] is related to the trigonal prism by capping one rectangular face and bridging one base edge. [Rh15(C0)&]- has the more complex geometry (22) in which the metal atoms form a centred tetracapped pentagonal prism and is the first example apart from the gold clusters above of a completely encapsulated metal atom. The mean distance from the central Rh to its 12 near neighbors is 2.908,.The analogy between this complex and a metal lattice is spoiled by the two carbide carbons which lie at 2.06 A from the centre Rh.77 However in the structure (23) of the [Rh13(C0)24H3]2- ion derived from reaction of [Rh12(C0)30]2- with hydrogen and studied by X-ray diffraction7' of its [(Ph,P),N]' salt the arrangement of rhodium atoms is essentially that found in hexagonal close-packing. The 13rhodium atoms are located in three nearly parallel layers making a cluster of D3,, idealized symmetry and the central metal atom is surrounded by 12 metal atoms at mean distances of 2.81 8,. Comparison with the cubic close-packed distance of 2.69 8 in Rh metal indicates that the electron density is higher in the (22) (23) The first example of a homonuclear trigonal-bipyramidal metal cluster has been Reduction of Ni(CO) by alkali metals in THF gives [Ni6(C0),,]'- and [Ni,(C0)12]2- and X-ray diffraction study of the latter as its [(Ph,P),N]' salt reveals a trigonal-bipyramidal cluster of Ni atoms with three bridging and three terminal carbonyls symmetrically disposed in the equatorial plane and three terminal carbonyls on each axial Ni.79 Platinum(1) and Palladium(i).-These are rare oxidation states the former being previously represented only bJ' JPtcl(PPh3)2], [Pt2C14(C0)zI2- [Pt,C12(Ph2PCH2PPh,)2] and [(Ph3P)2 t.S.Pt(CO)(PPh,)] and the latter by [Pd,Cl,(Bu'NC),] [Pd2Cl,(C0)2]2- an ill defined species [Pd(C6&)(H,0)(C10,)], and some very unusual organo- bridged compounds [Pd,(C,H,)(PPh,),I] (A) and 76 P.L. Bellon M. Manassero and M. Sansoni J.C.S. Dalton 1972 1481; P. L. Bellon F. Cariati M. Manassero L. Naldini and M. Sansoni J.C.S. Chem. Comm. 1971 1423. 77 V. G. Albano M. Sansoni P. Chini S. Martinengo and D. Strumolo J.C.S. Dalton 1975 305 and references therein. 78 V. G. Albano A. Ceriotti P. Chini G. Ciani S. Martinengo and W. J. Anker J.C.S. Chem. Comm. 1975,859. 79 G. Longini P. Chini L. D. Lower and L. F. Dahl J. Amer. Chem. Soc. 1975,97 5034. 206 D.J Cardin and K. R. Dixon ,c-c w' CC 0 CI cI t CI-Pt -Pt -Cl I I Pd - c 0 CI Br (25) [Pd~,C17(C6H6)],(B). X-Ray structure studies were available only on the sulphur- bridged platinum compound and the organo-bridged palladium compounds the latter having structures analogous to (24) except for the replacement of the C,H and Br bridges by allyl and iodide in (A) and by two benzene bridges in (B).Moreover all of the known compounds appear to be ligand bridged with the observed diamagnetism indicating some metal-metal bonding.80*8' An X-ray struc- ture determination carried out this year has confirmed the original suggestions (made in 1973 by Goggin and Goodfellow and based mainly on vibrational spectra) that the [Pt2C14(C0)2]2-ion contains an unsupported Pt -Pt bond and can exhibit an unusual form of isomerism. The anion was studied as its [NPr,]+ salt and is of the basic structural type shown in (25). Co-ordination about each platinum is distorted square-planar and the two planes are twisted about the Pt-Pt bond to give a transoid configuration with the dihedral angle between the two Pt-Cl groups being 120".The complex studied was one of the two isomers originally isolated; the other is thought to have the related cisoid structure. The 120"angle is considered to be a compromise between repulsions due to filled interaxial d -orbitals (optimum angle 135") and interligand repulsions (optimum angle 90°).80 The closely related pal- ladium(1) ion [Pd2(CNCH3)6]2+ has a similar structure and is thus the first Pd' dimer without bridging ligands in the solid state. Unlike the platinum complex the two square planes are almost perpendicular (dihedral angle 86.2"),a situation akin to the Ni' complex [Ni,(CN),l2- (dihedral angle 82").*' The metal-metal bonds in both palladium and platinum complexes are among the shortest known (2.531 A and 2.584 A respectively).X-Ray structure determinations of [Pd2(p-Br)(p-CSHS)(PPri)2] (24)82 and [Pd,(p-C,H,)(p-C,H,)(PPh,),l (C)"' demonstrate that the unusual type of bridging discussed above for allyl groups or benzene in complexes (A) and (B) can be extended to the cyclopentadienyl group. Both structures are of the type (24)except that in (C) the bridging bromide is replaced by a bridging methallyl group. The Pd-Pd bond lengths are 2.609 % and 2.679 % for (24) and (C) respectively which may be compared with the values of 2.686 A and 2.58 A reported for (A) and (B) respectively. In (24) the cyclopentadienyl ring has four C-C distances approxi- mately equal (1.46-1.52 A) but the bond parallel to the Pd-Pd axis is much shorter (1.33 A) and the ring may be regarded as an allyl plus alkene group.SO A. Modinos and P. Woodward J.C.S.Dalton 1975 1516 and references therein. D. J. Doonan A. L. Balch S. Z. Goldberg R. Eisenberg and J. S. Miller J. Amer. Cfwm. Soc. 1975,97 1961. 82 A. Ducruix H. Felkin C. Pascard and C. K. Turner J.C.S. Chem. Comm. 1975 615. 83 H. Werner D. Tune G. Parker C. Kruger and D. J. Brauer Angew. Chem. Infernat. Edn. 1975,14 185. Organometallic Compounds 207 21 Metal Hydrides Paramagnetic hydrido-complexes are still very unusual species. [ReHX,(PPh,),(acac)] (X=C1 or I) has been isolated from reaction of [ReH,(PPh,),(acac)] with CCl or I, and several salts of the type [CoHL,]X [L =P(OEt),Ph P(OMe),Ph or P(OPh),; X =PF or BF,] have been obtained by reaction of trityl salts with CoHL,.Details of the preparation and characterization of the first paramagnetic hydrides of Fe' and Fe'" have now been rep~rted.~ Mild oxidation of [FeHCl(diphos),] (diphos =Ph,PCH,CH,PPh,) with AgClO or a trityl salt yields [FeHCl(diphos),]X (X =ClO or BF,) and reduction of the same substrate by powdered sodium in benzene gives [FeH(diphos),]. The Fe' species may also be obtained by a similar reduction of [FeH(diphos),]BPh,. The Fe' and Fe"' species have magnetic moments 1.8 and 2.16 BM respectively consistent with a low-spin configuration and a single unpaired electron as found in the [CoHL,]+ ions. Presence of hydride is confirmed by evolution of hydrogen in reactions with HC1 and by oxidation and reduction reactions to known Fe" hydride complexes.Although they deteriorate rapidly in solution the new hydride complexes are stable to air for several hours in the solid state and are thus considerably more stable than the [CoHL,]' Although Fe' is a well-established oxidation state in organometal- lic complexes such as [Fe,(C0),(q'-C5H5),] it is still rare in derivatives of nitrogen and oxygen donors. The macrocyclic ligand (26) forms iron complexes which can be reduced by controlled-potential electrolysis to purple Fe' derivative^.^^ Electrolysis in MeCN solution on the first reduction plateau (-1.2 V) gives [Fe'(tetraene-N,)]+ [tetraene-N =(26)] isolated as its [CF,SO,]- salt and repeated scanning of the second reduction plateau (-1.6 V) gives [Fe'H(tetraene-N,)(MeCN)] v(Fe-H) = 1890 cm-' via hydrogen abstraction from the supporting electrolyte Bu",NBF,; reaction with CCl yields CHCl,.In addition to this unusual hydride stable alkyl or aryl derivatives of Fe' are isolated by reduction of [Fe"(tetraene-NJCl]' by lithium alkyls or aryls to yield [FeR(tetraene-N,)] R =Me or Ph. All of these Fe' species have magnetic moments 2.1-2.3 BM and their e.s.r. spectra are typical of low-spin d7 ~ysterns.~' A hydrido-complex of Fe' is. also formed when FeC1 is reduced by NaBH in butan-1-01 in the presence of the triphosphine ligand MeC(CH,PPh,) (P,). Addi-tion of [Bu",N]PF permits isolation of the complex [Fe2H3(P3),]PF6 which has been studied by X-ray diffraction on the CH,Cl solvate.The structure is as shown in (27) 84 M. Gargano P. Giannoccaro M. Rossi G. Vasapollo and A. Sacco J.C.S. Dalton 1975 9 and references therein. 85 M. C. Rakowski and D. H. Busch J. Arner. Chern. Soc. 1975,97,2570. D.J. Cardin and K. R. Dixon and differs from the Fe' species discussed above in being diamagnetic due to a metal-metal interaction.86 The other interesting structural feature is the trihydrido bridge; only two examples seem to have been reported previously namely [Ir2H3(q5- C5H.J2]+ and [Re,H3(C0)6]-.87 The product of reaction of Co(BF,) with the arsenic analogue (A,) of P under similar conditions has also been studied by X-ray diffraction. [Co,H,(A,),]BPh has the same structure as (27) but is paramagnetic with peff=3.17 BM for the dimeric entity.The metal-metal bond lengths are very similar (Fe-Fe = 2.332 82 and Co-Co =2.377 82) and unusually short. Both this fact and the magnetic differences can be rationalized using an MO scheme proposed previously for confacial bioctahedral complexes. This predicts a bond order of three for the Fe complex and two for the Co complex with two unpaired electrons in degenerate anti-bonding orbitals in the latter case.86 The bulky ligand tricyclohexylphosphine has been used to stabilize a number of unusual hydrido species. An addition this year is the paramagnetic hydride [~H(B~){P(C6Hl1),},], formed by reaction ofCoC1 and P(C6Hll) with NaBH in toluene-ethanol solution. Magnetic susceptibility (pea=2.15 BM) and e.s.r.meas- urements are typical of low-spin Co" with v(Co-H) = 1797 cm-'. X-Ray diffrac- tion shows structure (28) with co-ordinated BH completing a distorted square- pyramidal geometry about cobalt. The Co-H length (1.34 & appears unusually short but this conclusion is tentative in view of the high standard deviation (0.09 A). The complex is an active catalyst for hydrogenation and isomerization of olefins.88 22 Organometallic Compounds One-carbonLigands.-Alkyls (see ref. 16a). A brief report suggests that reaction of manganese(I1) chloride with a trimethylsilylmethyl-metal derivative or a related reagent gives the crystalline alkyls Mn(CH,R) (R =SiMe, CMe, or CMe,Ph) which are thermally stable to over 100°C in contrast to the ready detonation of MnMe, the only previously known binary alkyl for this oxidation state.Oxidation of the new alkyls by molecular oxygen gives green products which may be Mn(CH,R) species analogous to the tetranorbornyl Mn'" compounds reported previously. Lithium salts of the alkylate anions [MnMe,12- [COR,]~- (R =Me or CH,SiMe,) and [U(CH2SiMe3)6]2- have also been ~btained.~' Alkyls of rhenium are even more difficult to obtain than those of manganese except of course for carbonyls and 86 P. Dapporto S. Midollini and L. Sacconi Inorg. Chem. 1975,14 1643. 87 C. White A. J. Oliver and P. M. Maitlis J.C.S. Dulton 1973 1901 and references therein. M. Nakajima H. Moriyama A. Kobayashi T. Saito and Y. Sasaki J.C.S. Chem. Comm. 1975 80. 89 R. Andersen E.Carmona-Guzman K. Mertis E. Sigurdson and G. Wilkinson J. Organometallic Chem. 1975,99 C19. Organometallic Compounds 209 related species. [ReOMe,] was reported only last year and the full papers have now been p~blished.'~ Reactions of [ReOCl,(PPh,),] or [ReOCl,] with LiMe in diethyl ether both give [ReOMe,] the former in 70% and the latter in 20% yield. In the former case traces of 0 are necessary for high yields; 0,does not affect the yield in the latter reaction but it is evident that both processes are more complex than simple methylation. [ReO(CH,SiMe,),] and [Re2O3(CH2SiMe3),] have also been prepared by reaction of Me,SiCH,MgCl with [ReOCl,(PPh,),]; the former is air stable in contrast to [ReOMe,]. Both oxotetra-alkyl species are paramagnetic and e.s.r.electronic absorption and i.r. spectra are consistent with a square-pyramidal ~tructure.~~ WMe was first prepared in 1972 and as reported last year TaMe is (with ReMe, vide infra) the only other example of a homoleptic methyl in an oxidation state greater than four. Preparation of WMe by interaction of LiMe and WCl is a rather complex reaction and it now appears that some of the difficulties are due to the fact that adventitious oxygen is required." Reaction of WCl with trimethylaluminium at -70 "C is a better procedure for preparing WMe, and it is important to note that the compound is potentially explosive. The analogous reaction of AlEt with [ReOMe,] yields the new green paramagnetic ReMe, which is reasonably stable at 25 "C. A stable methyl of an even higher oxidation state [Rev1IO2Me3] is obtained by oxidation of [ReOMe,] with nitric oxide," a reaction which must finally dispel the idea that alkyls are stabilized only in low oxidation states.Insertion Reactions.-The importance of the /3 -hydride elimination process as a factor in destabilizing transition-metal alkyl complexes has been noted.'6n There is still very little information available on the intimate mechanism of the reaction but the reverse process hydride insertion has been studied in more detail especially for platinum complexes. For the general reaction (30) two principal mechanisms have trans-[PtHXQ,]"' + alkene trans-[PtX(alkyl)Q,]"+ (30) been discussed and a paper this year9* summarizes the existing data and proposes a unified reaction scheme (Scheme 7).Originally the reaction was assumed to involve insertion insert ion ?l 11 trans-Q,PtHX + un [Q,PtHX(un)] E [Q,PtHX(un)] etc. (29) (294 11 11 X-+ [trans-Q,PtH(un)]+ X-+ [cis-Q,PtH(un)]+ 11 (30a) insertion un = unsaturated ligand Scheme7 90 K. Mertis D. H. Williamson and G. Wilkinson J.C.S. Dalton 1975,607;J. F. Gibson K. Mertis and G. Wilkinson J.C.S. Dalton 1975 1093. 91 L. Galyer K. Mertis and G. Wilkinson J. Organomefallic Chem. 1975,88 C37. D.J. Cardin and K. R. Dixon associative alkene co-ordination to give a five-co-ordinate intermediate (29) followed by rearrangement to the square-planar insertion product. Complexes similar to (29) have been isolated e.g. [PtH(CN){C2(CN)4}(PEt3)2] and [PtClMe(C,(CF,),}(AsMe,Ph)] but only the methyl complex has been shown to undergo a subsequent insertion reaction.Recent kinetic data have tended to favour four-co-ordinate intermediates (30) and trans-[PtH(GH,)(PEt,),]+ has been observed by n.m.r. spectroscopy in the reaction of trun~-[PtH(acetone)(PEt,)~]+with C2H4 and isolated as its [BPh,]- salt from the reaction of trans-[PtH(NO,)(PEt,),] with C2H4. In the unified reaction Scheme 7 the five-co-ordinate intermediate (29) may be either a transition state or a bona fide intermediate. Weak ligands (NO3- acetone MeOH etc.) favour (30) [i.e. (29) is then a transition state] and stronger ligands (29) although in the latter case definition of the necessary ligand properties is not clear-cut.The intermediates (29a) and (30a) represent changes of stereochemistry which may be necessary before insertion can occur.92 For (29) this presents little difficulty since stereochemical interchange is facile in five-co-ordinate systems but in (30) there is a problem presented by the observed trans geometry of the intermediates and the necessity for adjacent hydride and C,H positions prior to insertion. This could be resolved either by an X-assisted trans +cis isomerization (Scheme 7) or by an associative process involving an additional C;H4 The trans -B cis isomerization route is suggested by the analogous reaction of truns-[PtHX(PEt,),] with alkynes which always results in a vinyl with Pt and H in mutually cis positions.94 However this mechanism is clearly not possible for ethylene insertion into the Pt-H bond of trans-[PtHCI(P-P)L where P-P is the diphosphine (31) which spans puns positions in square-planar complexes (32).95 The problem has also been discussed regarding the insertion (31) reactions of trans-[PtMe(q-allene)(PMe,Ph),]+ to cis-[Pt(~-2-methylallyl)-(PMe,Ph),]'.The relation to the above insertions is clear and the reaction shows first-order kinetics. A five-co-ordinate transition state is considered unlikely since addition of anionic or neutral donors or excess allene suppresses the insertion. A slight dependence of AH+ on the counter-ion SbF6- -BF,-> PF6- suggests that the reactive species is a tight ion pair but it seems that the chief role of cationic intermediates in promoting Pt-H and Pt-C insertions is activation of the unsatu- rated hydrocarbon rather than stabilization of five-co-ordinate intermediate^.^^ The 92 H.C. Clark C. R. Jablonski and C. W. Wong Znorg. Chem. 1975,14 1332 and references therein. 93 H. C. Clark C. R. Jablonski J. Halpern,A. Mantovani,andT. A. Wed Znorg. Chem. 1974,13,1541. g4 H. C. Clark and C. S. Wong J. Organometallic Chem. 1975,92 C31. 95 G. Bracher P. S. Pregosin and L. M. Venanzi Angew. Chem. Internat. Edn.,1975,14,563. 96 M. H. Chisholm and W. S. Johns Inorg. Chem. 1975,14 1189. Organometallic Compounds 211 possibility of achieving adjacent hydride and ethylene positions via tetrahedral intermediates remains essentially unexplored. Oxidative Addition.-This is another reaction of central importance in organometallic mechanisrn~.~~ As reported last year use of the spin trap Bu'NO has provided evidence for a non-chain radical process in the oxidative addition of alkyl halides to [Pt(PPh,),] and stereochemical studies on similar reactions of [Pd(PPh,),] have suggested an SN2type of attack by the metal at the carbon centre.The latter work also indicated a need for caution in that addition of Bu'NO to the reaction of PhMeCHBr with [Pd(PPh,),] produced Bu'(PhCH,)NO* even though the reaction probably does not normally proceed by a radical mechanism (this is because Pd" alkyls are significantly less stable than Pt" analogues and react with Bu'NO). However CIDNP evidence has now confirmed that radicals are involved in oxidative - SN2 Pto + RX R-Pt"-X % 7 Pt'-XR.1 PtI-X + R' R2 q0, RX X-Pt"-X + R' chain process R-Pt"-X H-PP-X Scheme 8 addition of isopropyl iodide to [Pt(PEt3),]. Moreover this reaction in common with several others involving reactive alkyl halides is more complex than a simple addition to form trans-[Pt(Pri)I(PEt,)J; other products are trans-[PtHI(PEt,),] trans-[PtI,(PEt,),] propene propane and 2,3-dimethylbutane. Scheme 8 is prop- osed to unify these and other observations. The initially formed radical pair (Pt'-X R.)can collapse to the regular adduct or diffusively separate. Subsequent reaction of the separated radicals depends on the reactivity of the alkyl halide; very reactive ones form dihalide and organic radicals and less reactive ones initiate a chain process.97 The initial radical pair formation may be preceded by one-electron transfer.The reaction of M(PEt,) (M = Pt or Ni) with tetracyanoethylene (TCNE) generates the TCNE radical anion detected by e.s.r. but the corresponding M' species is not In contrast reductive elimination reactions in two cases studied this year appear to occur by concerted intramolecular processes. Ther- molysis of cis-[PtAr,L,] {Ar= Ph or 4-MeC6H,) = (PPh3)* [P(4-MeC6HJ3], Ph2PGH4PPh2 or Ph,PCH,PPh,} generated Ar2 quantitatively and without isomerization whereas radical pathways or ortho-metallation should produce ArH or isomerized aryk9' Moreover kinetic studies on reductive elimination of ethane 97 A. V. Kramer and J. A. Osborn J. Amer. Chem. Soc.1974,96,7832 and references therein. 98 I. H. Elson D. G. Morrell and J. K. Kochi J. OrganometallicChem. 1975,84 C7. 99 P. S. Braterman R. J. Cross and G. B. Young J.C.S. Chern. Cornrn. 1975,627. 212 D.J.Cardin and K. R. Dixon from fac-[PtClMe,(PMe,Ph),] suggest a concerted elimination process from a five- co-ordinate intermediate formed by dissociation of tertiary phosphine. The inter- mediate resembles an ethane complex of platinum(I1) and the process is similar to that previously proposed for reductive elimination from Au'" species.'oo Dialkylaminomethyl and Dialkylphosphinomethyi Ligands.-The ligating proper- ties of unsaturated species containing heteroatoms are a potentially rich field of chemistry. A structure containing q2-bonded [Me,N=CH,]' a formal three- electron donor was first suggested in 197 1 for the complex [CuCl(Me,NCH,)]Br.pis type of co-ordination has been confirmed by X-ray study of (CH2CH2NCH2)Mn(CO), which has Mn-N = 1.98 A Mn-C= 2.09 A and C-N = 1.45 A the last being an essentially typical single bond length. A wide range of (q2-R,CNH,)M(CO) (M = Mn or Re) complexes has now been reported from elimination of RiSnBr between R;NCH2SnR and M(CO),Br."' A more direct approachlo2 uses reaction of [Me,NCH,]I with Na[M(C0),(q5-C,H,)] (M = Mo or W) to synthesize the 0-bonded derivatives M(CO),(q'-C,H,)(q '-CH,NMe,). Wheo M = Mo the 0-bonded complex transforms under reflux in light petrol to the n-bonded derivative Mo(C0),(q5-C,H,)(q2-CH2NMe2), a reaction reminiscenf of the familiar CT -P T ally1 interchange.The change is from one- to three-electron donor in both cases. Curiously the expected magnetic inequivalence of N-methyl groups and methylene protons is not observed in the n.m.r. spectrum and this situation is also found for the seven-co-ordinate derivatives MI(CO),(NCMe)( q2-CH,NMe,) obtained by reaction of [Me,NCH,]I with M(CO),(NCMe)j. In addition to the q ' and q2derivatives an unusual reaction of [Me,NCH2]I with [Fe(C0),I2- or [Cr(CO),]'-yields the known carbene complexes Fe(CO),(CHNMe,) and Cr(CO),(CHNMe,) presumably by hydrogen abstraction from one MCH,NMe group to give NMe3.'02 Analogous phosphino-derivatives have also been reported. '03 Reaction of CoCl(PMe,) with the ylide Me,PCH2 gives a very air-sensitive red product formu- lated as Co(q2-CH2PMe2)(PMe3) on the basis of 'H and 31P n.m.r.Interestingly in view of the above observations on q2-CH,NMe, the compound is fluxional and assignment of definite phosphorus positions in the co-ordination sphere is not possible. The q2-CH2PMe2 system can also arise by an unusual oxidative addition reaction. Treatment of FeCl,(PMe,) with Mg and PMe in THF yields FeH(q2- CH,PMe,)(PMe,) from oxidative addition of a trimethylphosphine to the Fe centre. It is suggested that equilibrium formation of Fe(PMe,) may be involved in reactions of this species. lo3 X-Ray confirmation of the q2-CR,PR possibility has come from a rather unexpected quarter. trans-[PtCl,(PPr",),] reacts with 1-lithium 2-phenyl-1,2-dicarbaclosododecaborane to give a product containing a 0-carbaboranyl ligand and an internally metallated q 2-P(Pr"2)CHEt ligand (cited in ref.103). Two Carbon Ligands.-Aza-allyl Complexes and Related Species. Continuing the theme of ligating properties of hetero-unsaturated ligands we note that 2-aza-ally1 100 M. P. Brown R. J. Puddephatt and C. E. E. Upton J.C.S. Dalton 1974 2457. 101 E. W. Abel and R. J. Rowley J.C.S. Dalton 1975 1096 and references therein. 102 C. W. Fong and G. Wilkinson J.C.S. Dalton 1975 1100. 103 H. H. Karsch H. F. Klein and H. Schmidbaur Angew. Chem. Znternat. Edn. 1975 14 637 and references therein. Organometallic Compounds complexes have been known for several years. An unusual reaction of LiNCR2 (R =Ph or p-tolyl) with MC1(C0),(q5-C5H5) (M =Mo or W) yields M(CO),($- C5H5)(R2CNCR2) complexes by elimination of cyanate.An X-ray study of the complex with M =Mo and R =p-tolyl shows a bent (128") aza-allene structure with the CR2 planes mutually perpendicular but low-temperature n.m.r. studies suggest an asymmetric ?r-aza-ally1 structure in Extension of this work to 1,3- diaza-ally1 compounds has been achieved by reactions of RNC(R)NRLi with MC1(C0),(q5-C5H5) or with MnX(CO) (X =Cl,Br or I). In the latter case the initial product is a carbamoyl complex Mn(CO),{CO.NRC(R)=NR} which decar- boxylates under U.V. irradiation to Mn(CO),{RNC(R)=NR}. Co-ordination of the hetero-ligand is probably of a a,a-type analogous to a carboxylate rather than an ally1 group. lo4 Reactions of an cy -chloroenamine Me2C=C(NR,)C1 =Me or CH3\ /NRZ CH3\ /NRZ R R -c1-A /c=c\ ___) c=c -\N/ I CH3 c1 CH3/\M(CO),(CsHs)x CH3\ /c\ M(Co)y- 1(C5H5)x + 1 /c\c/ (C5H5)xM(CO),-CH3 II RR 0 \/ I-co .C' II 0 1-2co 4+H' H R H I CH3 R C=C=N +/ Scheme 9 2R =(CH,),] with a carbonyl anion [M(CO),J- (M =Mn or Re) [M(CO),(q'- C,H,)]-(M =Mo or W) [Co(CO),]- or [Fe(CO),(r]5-C,H,)]- reveal an exception- ally varied co-ordination chemistry (Scheme 9).Seven different modes of co-ordination are established by spectroscopic data including the first examples of 1-aza-ally1 and 2-azabutadiene complexes the former being an Mo species isolated as 104 T. Inglis and M. Kilner J.C.S. Dalton 1975 924 930 and references therein. D.J.Cardin and K. R. Dixon its PF6- salt and the latter an Mn compound formed by hydrogen migration in the ligand Phosphonium Betaine Complexes. (See also Section 13.) Reaction of (21) with triethyl phosphite results in nucleophilic attack at the a-carbon of the acetylide to yield a new type of two-carbon three-electron ligand. X-Ray diffraction studies of the product Fe2(PPh2)(Co),C{P(oEt),}CPh,show a structure with C-C and P-C bond lengths in the bridging group 1.34 and 1.74 A respectively. This indicates considerable multiple character in both bonds and a formulation intermediate between (33) and (34).lo6 (EtO),P\' Ph / Ph (33) (34) Tris(ethylene)plutinurn and Related Species. The synthetic potential and catalytic activity of labile olefin-substituted Mo complexes are well established.The many elegant studies using bis(cyc10-octa- 1,5-diene)nickel and cyclo-dodeca- 1,5,9- trienenickel are of particular interest and the tremendous potential of correspond- ing platinum complexes is obvious. Although bis(cyc1o-octa- 1,5-diene)platinum (35) has been obtained previously by U.V. irradiation of PtPr\(1,5-C8H,,) in the presence of 1,5-C8H12,a convenient and reliable synthesis was lacking. Moreover the other known Pto species such as Pt(PPh,) or Pt(C,H,)(PPh,) contain relatively non-labile tertiary phosphines. Thus the preparation of (35) in good yield by the treatment of Li2C8H8with PtC12(1,5-C,H12) in the presence of excess 1,5-C8H12 is an important advance and has already led to much novel chemistry (Scheme 10).lo7 The structure of Pt(C2H4)3 and related species is of especial interest since a theoretical study of Ni(C,H,) has predicted a trigonal-planar structure rather than one having alkene ligands perpendicular to the co-ordination plane.The only X-ray structure confirming this prediction was that of tris(bicyclo[2,2,l]heptene)nickel. This type of structure has now been confirmed for Pt(C,H,),(C,F,) easily derived from Pt(eH,) by a simple displacement reaction and also for tris-(bicyclo[2,2,l]heptene)palladium (D) and the corresponding platinum compound and Pt(GHJ2(PMe3). The last example was characterized by 13C n.m.r. studies (the others by X-ray diffraction) which show rotation of the co-ordinated ethylene at ambient temperature and a trigonal-planar structure at low temperatures.Bis(cyc1o- octa-1,s-diene)palladium (E)has also been obtained by a reaction analogous to the synthesis of (33 and reacts with C2H4 to produce a complex which is probably Pd(C.J34),.'07 Preparation of (D)and (E)on a gram scale has also been achieved by condensation of palladium atoms with the appropriate ligand at low temperatures (-196 and -120 "C respectively)."' lo5 R. B. King and K. C. Hodges J. Amer. Chem. Soc.,1975,97,2702. Io6 Y. S. Wong H. N. Paik P. C. Chieh and A. J. Carty J.C.S. Chem. Comm. 1975 309. lo' M. Green J. A. Howard J. L. Spencer and F. G. A. Stone J.C.S. Chem. Comm. 1975 3 449; M. Green J. A. K. Howard A. Laguna M. Murray J. L. Spencer and F. G. A. Stone J.C.S. Chem. Comm. 1975,451,and references therein.R. M. Atkins R. MacKenzie P. L. Timms and T. W. Turney J.C.S. Chem. Comm. 1975,764. Organometallic Compounds ,CF 0-C.. I 1 'CF, Pt-Pt /CF,CFCF, z Scheme 10 Formation and Cleavage of C-C Bonds at Metal Sites.-Linking reactions of unsaturated substrates at metal centres are of obvious importance in understanding many organometallic processes but established general reaction classes are rela- tively rare. A common approach is to utilize fluorocarbon species since this frequently stabilizes reactive intermediates. During the past several years oxidative linking such as the reaction of tricarbonyl(butadiene)iron with C2F4 to yield (36) has been shown to be a fairly general reaction. A series of papers published this year'09 shows that several other types of product are also obtained.Thus tricarbonyl(cyc1o- octa- 1,3-diene)iron with GF4 gives the ferracyclopentane (37) in addition to the expected product and tricarbonyl(cyc1ohexa- 1,3-diene)iron with CF,CFCF gives (38) in addition to products related to (36) and (37). Compound (37) is regarded as derived from the initial n-allylic insertion product by n +(+ ally1 interchange. Isolation of (38) together with results on the stereochemistry and position of linking in substituted dienes and fluoro-olefins suggests that the mechanism of the reaction is via endo attack by olefin on the diene to give an ionic intermediate such as (39). The alternative reaction path uia initial attack of fluoro-olefin on Fe cannot be excluded in some cases.Reactions essentially similar to the formation of (36) also occur between M(1,3-diene)(CO) (M = Fe or Ru) and hexafluoroacetone and between Ir(q3-RC3H4)(CO)L (L = PPh or AsPhJ and GF or CF,CrCCF,. In the Ir-GF case an initial adduct (40) is isolated and the linking reaction involves a second C,F molecule which forms [CH A CRCH,CF,CF,fr(q'-C,F,)(CO)L] when R = Me. However when R = H the unusual iridocycle (41) results.'o9 lo9 M. Bottrill R. Goddard M. Green R. P. Hughes M. K. Lloyd S. H. Taylor and P. Woodward J.C.S. Dalton 1975 1150; see also preceding papers in this series and references therein. D.J. Cardin and K. R. Dixon co (42) Extension of this work''' to reactions of hexafluorobut-2-yne. (HFB)with tricar- bonyl(butadiene)iron yields an insertion product analogous to (36) and thermolysis of this product under reflux in hexane gives (42).This overall reaction involves stepwise Diels-Alder addition of HFB to co-ordinated buta- 1,3-diene a result which suggests that previously reported examples of concerted 'forbidden' reactions at transition metals may also have stepwise mechanisms. HFB and tricarbonyl- M. Bottrill R. Goddard M. Green R. P. Hughes M. K. Lloyd B. Lewis and P. Woodward J.C.S. Chem. Comm. 1975,253;J. L. Davidson M. Green F. G. A. Stone and A. J. Welch ibid. p. 286; R. Davis M. Green and R. P. Hughes ibid.,p. 405. Organometallic Compounds 217 (cyclohexa- 1,3-diene)iron give the double insertion product (43) but with tricarbonyl(cyc1oheptatriene)iron a remarkable addition of two molecules of CF3CCCF3 occurs on the endo face of the triene to yield the quadricyclic ligand complex (44).Another unusual product is (45) {obtained by reaction of HFB with [Fe(CO),(q '-C5H5)],} a type of intermediate which has been frequently pos- tulated in metal-catalysed formation of cyclopentadienones and quinones.' lo H H (C0)3Fe' dF3 CF3 CF3 (44) (45) Another important class of reaction is transition-metal cleavage of C-C bonds of cyclic ligands. The classic example here is insertion of PtIV into cyclopropane ring systems and X-ray studies have established that a C-C bond is cleaved even though cyclopropane can be displaced by several donor ligands. '11 An interesting addition this year is the cleavage of trans-divinylcyclopropane by bis(ethy1ene)hexa- fluoroacetylacetonatorhodium(1)to give the bis(ally1) complex (46).'12 However examples for ring systems other than cyclopropane are still rare and the following are therefore of interest.In each case products have been characterized by X-ray diffraction. The final product from reaction of norbornadiene (nbd) with Rb(CO)16 is (47),derived from ring-opening of nbd and possibly a stabilized retro-Diels-Alder intermediate.lI3 Low-temperature n.m.r. spectra of {endo-(RO)C,Ar,}Pd(acac) 0 // (47) R = Me or Et Ar =Ph or p-FC6H4 demonstrate that the ring-closing step in the formation of cyclobutenyl derivatives from PhCCPh and PdCl in alcohols is readily reversible. Isomers (48)and (49) are both present in the ~olufion.''~ Reversible opening of a saturated C ring has been demonstrated in the conversion of (50) into (51).Both compounds contain two of the illustrated structural units joined by chloride bridges. The endo-phenyl analogue of (51) is present during the reaction R. D. Gillard M. Keeton R. Mason M. F. Pilbrow and D. R. Russell J. Organometallic Chem. 1971 33 247. *I2 N. W. Atcock J. M. Brown J. A. Conneely and J. J. Stofko jun. J.C.S. Chem. Comm. 1975 234. 113 J. A. J. Jarvis and R. Whyman J.C.S. Chem. Comm. 1975,562. 114 P. T. Cheng T. R. Jack C. J. May S. C. Nyburg and J. Powell J.C.S. Chem. Comm. 1975,369. D.J.Cardin and K. R. Dixon and has been isolated thus demonstrating that the ring-closure is readily rever~ible."~ Finally perhaps the most unusual of all is the insertion of Pto into an arene C-C bond.[Pt,(Bu'NC),] or trans-stilbenebis(trimethy1phosphine)platinum reacts with hexakis(trifluoromethy1)benzene to give (52; L =Bu'NC or PMe,) which represents a new type of possible intermediate in cyclo-oligomerization reactions of a1 kynes. l6 OR OR Ar Ar Ar 23 Catalytic Processes The 'heterogenizing' of homogeneous catalytic systems is an area of developing interest and has been reviewed re~ently."~ The objective of this research is to combine the ease of recovery of a heterogeneous catalytic system with the great activity selectivity and ease of mechanistic study of a homogeneous system. The most usual approach uses polystyrene cross-linked with divinylbenzene as a polymer support and ligating groups are introduced by substitution on the polymer chain.Papers published this year'18 provide a thorough investigation of a number of typical systems and describe the important new possibility of anchoring more than one type of catalyst to the same polymer chain thus enabling catalysis of sequential multistep reactions. Iron-catalysed bromination of polystyrene followed by reaction with LiPPh leads to the pura-PPh substituted derivative (53). Simple phosphine exchange reactions as shown in Scheme 11 yield the appropriate transition-metal substituted polymers (54)-(58). Among the reactions studied are cyclo-oligomerization of butadiene catalysed by (54)to yield (59) (60),and (61); hydrogenation of these products catalysed by (55) to yield the fully saturated analogues of (59)-(61); and hydrofor- mylation of (59) catalysed by (57) to yield (62) and (63).The product distributions 115 D. J. Mabbott P. M. Bailey and P. M. Maitlis J.C.S. Chem. Comm. 1975 521. 116 J. Browning M. Green A. Laguna L. E. Smart J. L. Spencer and F. G. A. Stone J.C.S.Chem. Comm. 1975,723. 117 J. C. Bailar jun. CatulysisReu. 1974,10 17. 118 c.U.Pittman jun. L.R. Smith and R. M. Hanes J. Amer. Chern. Soc. 1975,97,1742; C. U. Pittman jun. and L. R. Smith ibid. p. 1749. Organometallic Compounds Phl 'f"H(C0)(Ph3 P) (57) Ph 3-x Scheme 11 and responses to excess PPh3 temperature changes and changes in H2or CO pressures demonstrate the mechanistic similarity of the polymer-anchored systems to their homogeneous analogues.Rates are generally somewhat lower than those achieved with the homogeneous analogues probably owing to diffusion retardation in the anchored catalysts. D.J. Cardin and K. R. Dixon In addition to the above processes sequential cyclo-oligomerization of butadiene to (59),(60) and (61) followed by hydrogenation to the fully saturated compounds has been achieved using the catalyst (56) thus demonstrating the possibility of constructing a single catalyst to achieve a multistep conversion in a 'one pot' process. Sequential cyclo-oligomerization of butadiene followed by hydroformylation to (62) and (63) is achieved by catalyst (58)."* Table 1 Molecular structures and electronic configurations of some pentacarbonyl species No.of valence Stereo- shell chemistry * electrons S.P. 15 S.P. 16 S.P. 17 S.P. 17 T.B. 18 T.B. 18 (a) S.P.=square-pyramidal,T.P. =trigonal-bipyramidal.