The Typical Elements PART 11 Group III By A. J. CARTY 1 Boron Boron Hydrides and Borane Anions.- Theoretical Studies of Bonding and Structure. It is refreshing to find a paper in which theoretical calculations are used to predict G. Parry and R. J. Pulham J.C.S. Dalton 1975 2576. R. A. Andersen and G. E. Coates J.C.S. Dalton 1975 1244. A.J. Carty,R. H. Cragg and J. D. Smith chemical behaviour rather than to rationalize postfacto experimental observations. In an important contribution from Lipscomb's laboratory ground-state charge distributions derived from minimum basis set Slater orbital calculations using the PRDDO approximation have been employed to predict likely sites of electrophilic and nucleophilic attack in BgHlZ BI0Hl4 [BllH13]z- related boranes and car- baboranes.la Reactivity predictions are based on relative values of inner-shell eigenvalues Mulliken atomic and group charges and in addition sums of popula-tions over the several highest occupied MO's. Such calculations can be an invaluable aid in rationalizing reaction sequences or devising routes to boron hydride deriva- tives although it must be remembered that orbital symmetry correlations are ignored. In the past similar calculations have proved of significant value in decaborane( 14) chemistry. For BgH'Z electrophilic attack is predicted to occur preferentially at B-2 while B-7 is most susceptible to nucleophilic substitution. For [B8Hl3]- electrophilic substitution should occur in the order B-4 >B-2 =B-1>B-3>B-7 and a facile loss of H- from either B-4 or B-5 should yield BgHI2.In several cases [e.g. in BgH1 (C symmetry) where charge criteria indicate B-3 as a site for nucleophilic attack but eigenvalue differences do not differentiate between B-3 and B-7 sites] an unequivocal distinction between two possible sites of attack cannot be made. Here experimental results will be particularly significant in refining the model. The same paper analyses structural interrelationships between several Bg-B hydride species on the basis of detailed examination of localized molecular orbitals (LMO's). Topological and LMO descriptions generally correlate quite well. In the context of ab initio LMO theory for boron hydrides it is significant that experimental confirmation of the validity of this approach has been obtained from measurements of the Compton profile of decaborane( 14)? There is excellent agreement between the experimental and calculated electron momentum distribu- tions in this molecule showing that the wavefunctions used in constructing LMO's are 'good'.Ab initio MO calculations on seventeen small boron compounds have been carried out for comparison of geometries bonding preferences and stabilities with isoelectronic carbocations. These calculations lend quantitative support to many intuitive predictions. Thus for example substitution of H in BH by a w-donor group X (NH, OH or F) stabilizes BHzX relative to BH by 53-58 kcalmol-' while hyperconjugation in H,BCH gives only 12 kcal mol-' of stabilization energy and adduct formation H,N,BH 21 kcal mol-'.Similarly H2NBH2 and H2BOH are predictably planar with B-N and B-0 multiple bonds and with non-rigid rota- tional barriers about B-N and B-0 bonds of 29 and 144 kcal mol-' respectively. Geometries of boron compounds and isoelectronic carbocations (e.g. HB=CH and the classical vinyl cation CH,=& are generally similar but in terms of bonding boron is a stronger 0-donor and weaker w-acceptor than positively charged carbon. This comparison is best represented by the major resonance structures (1)and (2). Where experimental data are available for comparison the 6-3lG* calculations appear to give a better overall level of agreement than previous calculations. * (a)J. H. Hall jun. D. A. Dixon D. A. Kleier T. A. Halgren L.D. Brown and W. N. Lipscomb,J. Amer. Chern. SOC.,1975,97,4202; (b)I. R. Epstein P. Pattison M. G.H. Wallbridge and M. J. Cooper J.C.S. Chern. Cornm. 1975,567;(c) J. D. Hill P. von R. Schleyer and J. A. Pople J. Amer. Chern.Soc. 1975 97,3402; (d)E. L. Muetterties E. L. Hoel C. G. Salentine and M. F. Hawthorne Znorg. Chem. 1975 14 950; (e)E. L. Muetterties and B. F. Beier Bull Soc. Chim. belges. 1975 &I 397. The Typical Elements The borane anions [BnH,I2- present a unique opportunity for the study of rearrangement mechanisms and dynamics among the various classes of stereochemi-cally non-rigid molecules. These ions incorporate two nuclei ('H and "B) which are excellent n.m.r. probes and in atomic and electronic terms are relatively simple molecules lending themselves to sophisticated MO calculations.Several experimen- tal observations'd concerning rearrangement barriers in these ions have been clarified via extended Huckel MO calculations." Key features of polytopal isomer- ism for the family [B,H,]'- are (i) [B8H8]2- dodecahedra1 in the solid state fluxional in solution with rearrangement occurring between square-antiprismatic and bicapped trigonal-prismatic isomers; (ii) [B lH1 1]2- subject to rapid intramolecular rearrangement in solution even at low temperatures; (iii) [B7H7]'- a rigid pentagonal-bipyramidal structure in solution over a wide temperature range; (iv) [B,H,]'- a much larger rearrangement barrier than for [B8H8]'- being essen- tially rigid up to 200 "C. These apparently conflicting facts can be rationalized by consideration of MO energies in the various polyhedral forms of the ions.For [B8H8I2- calculations indicate a small energy gap between the highest occupied and lowest unoccupied MO's in all of the possible DZd,D4d,and c2u forms. Facile rearrangement of all three forms can occur via normal vibrational modes. By contrast the gap is large for D5h[B7H7]2- and this anion cannot rearrange via the degenerate C,,isomer. The most likely intermediate (C4, [B,H,]'-) in the re- arrangement of ground-state [B,H,]'- is degenerate. This may explain the apparently anomalously high barrier to interconversion in this species. However for [BllHll]*- the presence of a degeneracy in the C5uform suggests that the lower- symmetry C isomer may be a more reasonable transition state.It will be interesting to see whether predictions of fluxionality in co-ordination compounds ML7 ML8 and ML based on generalization of the borane anion rearrangement processes are borne out in subsequent work; there is evidence that ML8 complexes have very low rearrangement barriers. Incorporation of heteroatoms (carbon or metal atoms) into a closoborane framework might be expected to complicate rearrangement processes. This appears to be the case for metallocarbaboranes (see below). Structure of Boranes. The new boron hydride tetradecaborane(20) synthesized via reaction of excess octaborane( 12) with potassium nonahydrohexaborate in ether followed by treatment with HCl at -78 "C is a yellow crystalline solid for which X-ray data collected at -164 "C indicate structure (3).'" The molecule consists of two fragments fused at the B-7-B-12 positions with their open faces cis to one another.This molecule thus joins a growing number of boranes with structures built up by fusing two different borane fragments across a common edge. Examples include B13H19 B16H20 n-BI8Hz2 and i-B18H2'. The hydride B14H18 recently (a) J. C. Huffman,D. C. Moody and R. Schaeffer,J. Amer. Chem. Soc. 1975,97,1621; (b)S. Hermanek K. Felter J. Plesek L. J. Todd and A. R. Garber Znorg. Chem. 1975 14,2250; (c)J. D. Weiser D. C. Moody J. C. Huffman R. L. Hilderbrandt and R. Schaeffer,J. Amer. Chem. Soc. 1975,97 1074. A.J. Carty,R.H. Cragg and J. D.Smith characterized by ''B and 'H n.m.r.may be of this type,2b with decaborane and hexaborane frameworks fused across one edge. The existence of other hydrides with decaborane fragments linked to lower polyhedra seems possible. Interesting structural results relevant to the problem of boron-substituent bond- ing in substituted boranes have been obtained. Thus the postulate that m-bonding interactions between fluorine (2p,) silicon (3p and 3d,) and the boron cage may be responsible for the known stability sequences 2-FB,H8> 1-FB5H8 and 1-SiH,B,H >2-SiH,B,H8 has been supported by extended Huckel MO calculations. (The 1-SiH,-isomer is favoured over the 2-isomer by ca. 0.4 eV.) In direct contrast whereas isomerization studies show a stability order 2-MeB,H8 >1-MeB,H, calcu- lations yield a negligible r-bond order for the B-C bond; hence B-C m-bonding is an unlikely source of stabilization.It is comforting to know that these results are borne out by subsequent electron diffraction results.2c The Si-B bond length in l-SiH,B,H [1.98l(5) A] is indeed significantly shorter than the corresponding length in the 2-isomer [2.006(4)A] while the B-C bond lengths in 1-MeB,H8 [1.595(5) A] and 2-MeB5H [1.592(5) A] are essentially identical. Synthesis and Structures of Borune Anions. The Bronsted acidity of boron hydrides is now well established. Abstraction of a bridging proton followed by insertion of an electrophile into the 'bare' boron-boron bond results in polyhedral expansion and is a useful synthetic method for borane anions or indeed metalloboranes.The nido carbaborane anion [2,3-C2B,H,]- undergoes quite analogous reaction sequences. Recent Annual Reports have illustrated the scope of these insertions. An important paper pertaining to properties and structures of borane anions derived in this way has now appeared. Thus B,H, can be deprotonated by potassium hydride or ammonia according to equations (1) and (2). R. J. Remmel H. D. Johnson jun. I. S. Jaworiwsky and S. G. Shore J. Amer. Chem. Soc. 1975,97 5395. The Typical Elements The competing hydrogen abstraction and cleavage implied in the latter reaction are notable and perhaps more general than previously suspected (cf. Annual Reports 1971-73). The anion [B4H9]-is dynamic at room temperature on the n.m.r.time-scale. The structure represented in terms of Lipscomb's fractional three- centre B-B-B bonds is shown in (4). More importantly the same paper describes H' 'H HHHH (4) (5) the anions [B5H12]- [B6H1J and [B7H12]-,produced via addition of BH3 to the corresponding species [B4H9]- [BSHS]- and [B6H9]-. The triphenylmethylphos- phonium salts of [B5HI2]-and [B6Hll]-are white solids stable at room temperature for short periods. "B and 'H n.m.r. spectra show that [B5H12]-is fluxional above -90 "C but a static structure is apparent at -135 "C. Topologically this ion has been represented as (5) and is a member of a new class of boron hydride species the hypho-boranes containing 2n +8 skeletal electrons. The hypho (Greek for net) structure can be compared with the increasingly open frameworks closo (2n +2) (e.g.[Bl2Hl2I2-),nido (2n+4) (e.g. B5H9),and arachno (2n+6) (e.g. B5Hil) as the number of skeletal electron pairs increases. The phosphine adducts B5H9(PMeJ2 described in last year's report (p. 181) and B6H10(PMe3)2are also members of the hypho-boranes. The structure of the latter not available at the time of writing is apparently an open fragment of the equatorial belt of an icosahedron. Shore and co-workers3 also give details of their high-yield synthesis of B5Hll and [equations (3) and (4)]briefly mentioned in the 197 1Annual Report. Finally it has become clear from recent work that the relative acidities (Bronsted) of boron hydrides of a given class (e.g. nido or arachno) increase with increase in framework size.-110°C KB,H, + HCl 'BSH11 + H + KCl (3) KB,H, + HCl B,H,2 + KCI (4) Metalloboranes.-An excellent review of metalloborane chemistry published in 1974,4" together with last year's Report gives perspective to this rapidly expanding (a) N. N. Greenwood and I. M. Ward Chem.SOC.Rev. 1974,3,231;(b)B. P. Sullivan R. N. Leyden and M. F. Hawthorne J. Arner. Chem. Soc. 1975,97 455; (c)R. N. Leyden and M. F. Hawthorne J.C.S. Chem. Comrn. 1975,310;(d)K. Wade Chem. inBritain 1975,11,177; (e)K. Wade Adv. Inorg. Chern. Radiochern. 1975,18 in the press; (f) J. T. Gill and S. J. Lippard Inorg. Chern. 1975,14 751. A.J. arty R. H. Cragg and J.D. Smith field of research. The major advance this year has been the synthesis of several closo-nickelaboranes derived from [B9HI2]- [B10H10]2- and [B 1H13]2-.4b,4c Syntnetic methods generally involve reduction of an anion with Na-Hg in the presence of nickelocene or for derivatives of [B10H10]2- direct reaction of the anion with [q-C,H,Ni(CO)] or [(q-C5H5)3Ni2]2+.Of the four compounds characterized two Bu:N[( q -C,H,)-1-Ni( B ,H 1)] and [(7-C5H5)2-1,2-Ni2( B loHlo)] have while Me4N[(q-C,H,)-2-Ni-(B9H9)] icosahedral ~tr~cture~~* and Me4"( q-C5H5)-1-Ni-(B9H9)] have bicapped square-antiprismatic structures with the nickel atoms in equatorial and apical positions respectively.& By analogy with related metallo- carbaboranes (see below) it might be expected that Me,N[(q-C,H,)- l-Ni-(B9H9)] would thermally rearrange to the 2-isomer where the metal has a higher-co-ordinate position.Rearrangement does occur photochemically and thermally but in the opposite sense indicating that not only co-ordination number but also the charge densities on the boron atoms in the cage may be pertinent to the stability sequence. The existence of these polyhedra is predictable from electron-counting For example the hypothetical anion [BloH,,]b- contains 13 skeletal electron pairs suitable for a 12-vertex polyhedron with two sites vacant. Incorporation of two formally [q-C5H5Ni]3' units (a neutral q-C,H,Ni fragment contributes three elec- trons to a cluster) would be expected to produce a closo-icosahedron. Similarly the anion [(B11H11)Ni(C5H5)]- is derived from the nido-species with one vacant site in an icosahedron.Clearly the existence of a vast number of metalloboranes can be predicted using electron-counting rules and unquestionably there will be major developments in this area paralleling the burgeoning chemistry of metallocarba- boranes. Finally attention is drawn to a conceptually useful correlation4' between the structures of nido-metalloboranes and the boron hydride species which can be derived from the complexed hydroborate ion by addition or elimination of H' or BH;. For example [B3H8]- formally derived from B4HIo by elimination of BH,' forms a complex [Cu(PPh3)2B,H8] with a structure quite analogous to B4H10. Similarly the molecule forms a complex [Fe(CO),(B6Hl0)] with a structure derived from that of [B6H11]+ by removal of the proton bridging a basal B-B bond.Whether this model has general predictive utility remains to be seen. Metal Complexes of the Tetrahydroborate Ion.-It has been known for many years that the 'H n.m.r. spectra of metal tetrahydroborate complexes provide little structural information since the bridge and terminal protons appear to be magneti- cally equivalent as a result of rapid intramolecular rearrangements. Structural assignments have thus been based mainly on vibrational spectroscopy and where possible X-ray analyses. Two groups have now independently observed separate resonances for bridge and terminal hydrogens in complexes containing bidentate [BH4]- ions. Marks and Kolb5" had previously observed that paramagnetism induced sufficient energy separation between exchanging proton sites in [(C,H,),UBH,] that the n.m.r.coalescence point could be approached at low temperatures. Subsequently a vibrational analysis of the paramagnetic molecule s(~)T. J. Marks and J. R. Kolb J. Amer. Chem. Soc.,1975,97,27; (b) T. J. Marks and W. J. Kennelly J. Amer. Chem. Soc. 1975,97,1439; (c) H. D. Empsall E. Mentzer and B. L. Shaw J.C.S. Chern.Cornrn. 1975.861. The Typical Elements [(C5H,),VBH,]56 indicated considerable covalency in the V-H(BH,) bonds. At -90 "C the fluxionality of this molecule was arrested; resonances due to bridge and teminal hydrogens were evident below this temperature. A free energy of activation of 7.6 f0.3 kcal mol-' compares with a barrier of ca. 5.0* 0.6 kcal mol-' estimated for [(C,H,),UBH,].The small size of these barriers emphasizes the facility of the intramolecular rearrangement which may well involve a bidentate-terdentate con-version. In view of the above results the iridium and rhodium complexes [IrH2(BH4)L2] (L=PBu',Me PBu PBu\Bu" or PBuiPh) prepared from [IrHCl,L,] by treatment with sodium borohydride are remarkable.," These com- pounds all exhibit typical i.r. spectra for doubly bridged metal-BH bonding but are non-fluxional in solution with resonances at ca. 6 = -6.5 for bridging [IrH2B] and at S =6.8-7.8 for terminal BH protons. Perhaps the bulky phosphines prevent the attachment of a third BH hydrogen in the intermediate [IrH,BH] species necessary for the bridge-terminal exchange. It should be noted that despite the obvious difference in structural types the terminal protons in static [(C,H,),VBH,] and [IrH,(BH,)L,] resonate at lower field than the bridging hydrogens.Carbaboranes and Metallocrubaboranes.-This is one of the most active areas of inorganic research and a great deal of interesting work is published each year. A small number of topics have been singled out for specific mention. Intramolecular Rearrangements. The facile intramolecular rearrangement of borane carbaborane and metallocarbaborane molecules is amply documented. Mechanistic understanding of these processes many of which have no precedent in other areas of chemistry is however in its infancy. An important paper which complements earlier work by Hawthorne and co-workers on large may provide an initial basis for rationalizing thermal isomerization processes for small metallocarbaboranes.6' Thermolysis of [1,2,4-(r)-CSH5)CoC2B3H5] produced no isomerization to the 1,2,3- or 1,2,6-~ystems.~' The seven-vertex cage [1,2,3-(q-C,H,)CoC,B,&] however isomerized in high yield at 400 "C to the 1,2,4-isomer.Rearrangement of [1,7,2,3-(q-C,H,),Co2C2B3H5] occurred stepwise yielding [1,7,2,4-(~-C5H,),Co2C2B3H5] via the intermediate 1,2,4,5- and 1,2,3,5-isomers as shown in Scheme 1. While it should be remembered that the isolation of inter-mediates does not prove a reaction mechanism the formation of all three sequential products in this reaction can be accomodated if (a)atoms permutate by a triangular rotation on the polyhedron surface [as opposed to a less likely diamond-square- diamond (dsd) mechanism] (6) carbon atoms may not move from a low to a high (apex) co-ordination position (c) carbon atoms will not decrease their mutual separations (d) rotation of B2C or B2Co triangles is preferred and (e) the metal atom prefers an apex position provided that rules (6)-(d) are followed.It is clear at least for [(q-C,H,),Co,~B,H,] that cobalt is not restricted to positions of high co-ordination number (cf. 1,2,4,5- and 1,7,2,3-isomers) although for cages with 26 boron atoms such a restriction has generally been {Note however (a)D. F. Dustin W. J. Evans C. J. Jones R. J. Wiersema H. Gong S. Chan and M. F. Hawthorne J. Amer. Chem. Soc. 1974 96 3085; (b) M. F. Hawthorne K. P. Callahan and R. J. Wiersema Tetrahedron 1974,30,1795;(c)V.R. Miller and R. N. Grimes J. Amer. Chem. SOC.,1975,97,4213; (d) G. D. Mercer M. Tribo and F. R. Scholer Inorg. Chem. 1975,14,764;(e) C. G. Salentine and M. F. Hawthorne J. Amer. Chem. Soc. 1975,97,6382; cr)B. J. Meneghelli and R. W. Rudolph Inorg. Chern. 1975,14 1429. A.J. Carty,R. H. Cragg and J. D.Smith 5 Scheme 1 that [~,~,~-(~-C,H,)CO(CM~~)~B&~] is converted into [1,2,3-(q-C,H,)Co(CMe,)B,H,] via [10,2,3-(q-CSH,)Co(CMe2)B8H,]6d and that [2,10,1-(q- C,H,),CoNiCB,H,] and [6,10 1-(q-CSH5)2CoNiCB7H8]6e contain nickel atoms in low-co-ordinate vertices.} Grimes" suggests that while the driving force for the 1,7,2,3+1,7,2,4 conversion is the separation of carbon atoms strong cobalt-cobalt bonding lowers the activation energy for apex+equatorial migration of cobalt thus allowing a facile reaction path via a series of trigonal rotations.Pyrolysis of the and eight-vertex cobaltacarbaboranes [3,1,7-(q-C5H,)CoC2BsH7][3,5,1,7-(q-C,H,),Co,C,B,H,] resulted in considerable cage fragmentation (boron and cobalt transfer) but no isomerization to the [3,1,8-(q-CSH,),CoC2BsH7] or [3,5,2,8-(q- C,H,),Co,C,B,H,] isomers predicted if the number of Co-C bonds was to be minimized. If the structures of these compounds have been correctly assigned (no X-ray data are available) these results may imply that the thermal stabilities of the 3,1,7-monocobalt and 3,5,2,8-dicobalt cages are higher than those of any other isomer. Alternatively there may be a kinetic barrier to interconversion.Conversion of the nine-vertex cage [1,7,5,6-(q-CSH5)2C02C2BSH7], in which two cobalt atoms are adjacent into the isomer [1,8,5,6-(q-C,H,),Co,C2BsH7] produced an equilibrium the first such example for closo-metallocarbaboranes. The AH for the 1,8,5,6+1,7,5,6 process is -2 kcal mol-' perhaps indicating a fine balance between competing steric and electrostatic repulsion effects on the one hand and strong cobalt-cobalt bonding on the other. The Typical Elements / 450"c 450 "C -95 % BH .C ,1L 111 IV Scheme 2 A second important paper this year on polyhedral rearrangements deals with the synthesis and thermal isomerization of the mixed-metal bimetallocarbaborane [(C,H,),CoNiCB,H,] (Scheme 2).6e The presence of a direct nickel-cobalt bond in the thermally most stable [6,9,l-(~-CSH,),CoNiCB,H,l isomer IV and preference of nickel for a high-co-ordinate position in IV was established although nickel atoms in I and I11 occupy low-co-ordinate sites.Rearrangement processes were discussed in terms of the dsd mechanism (cf. ref. 6c). This year's work thus establishes clearly that in metallocarbaboranes of inter- mediate stability the presence of metal atoms in low-co-ordinate positions is not precluded. Furthermore strong metal-metal bonds are possible and may play an important role in determining relative stabilities for other di- and polymetallocar- baboranes. Rearrangements of metallocarbaboranes incorporating second- and third-row transition elements will be interesting in this regard.It is also clear that a definitive experiment to distinguish between the alternative dsd and trigonal- rotational mechanisms for metallocarbaborane interconversion has not yet been described. A. J.Carty R. H. Cragg and J.D. Smith The reduction of closo-carbaborane cages with subsequent cage-opening and insertion of a metal ion into the vacant site is a valuable synthetic route to metallocarbaborane complexes. The mechanisms of ring-opening are generally obscure. Evidence is however accumulating that two-electron reduction of closo-carbaboranes occurs via conversion of deltahedra to nido counterparts. Thus extended Huckel MO calculations confirm that D3,,[C2B3H5lz- should not be stable and that trigonal-bipyramidal isomers are energetically inaccessible.6f However square-pyramidal isomers of [GB,H,]’-have appreciably higher stabilities with the stability order trans -basal-basal >cis-basal-basal >apical-basal.Accessible and symmetry-allowed interconversion pathways between these three isomers are avail- able (Figure 1).Addition of two protons to give C,B,H changes the relative stability ordering to apical-basal >trans-basal-basal >cis-basal-basal such that the two hydrogens can be accomodated in bridging positions on non-trigonal faces of the apical-basal isomer. The known isomer of C,B,H (isoelectronic with [C,B,H,]’-) has this stereochemistry. For the six-vertex carbaborane anions [C,B,H$ derived from CzB4& all pentagonal-bipyramidal isomers are more stable than octahedral species.Hence open pyramidal forms of [CzB3H,l2- and [CzB,H,]2- appear favoured over their closo analogues. The rearrangements closo =nido and -2e nido 7arachno provide a rationale for polyhedral expansion. Metal fragments capable of accomplishing these conversions can be predicted.4d.‘ -370 -374 -378 4 Rearrangement co-ordinate Figure Energy as a function of geometry for [C2B3H,]2-. The solid line represents the intercon- version of isomeric forms of [CzB3H5]2-via symmetry-allowed arcing movements of the designated atoms. The broken line designates interconversion via a dsd mechanism (Reproducedby permission from Inorg. Chem. 1975,14,1429) The Typical Elements Synthesis. There has been very little information relating to the mechanism of formation of carbaboranes.The kinetics of formation of derivatives of 1,2-dicarba- closo-dodecaborane(12) from B10H12(Me2S) and acetylenes RCGCH show that the rate determining process is attack by the acetylene on B10H12(Me2S) forming the adduct BloHl2(Me2S)(RC-CH). Subsequent loss of a molecule of Me,S from B-9 and co-ordination of the acetylne to B-9 adjacent to B-10 yields a geometry favourable to closo-carbaborane f~rmation.'~ Importantly isolable BloHl,L species synthesized by other workers are not intermediates and hence not identical to the active species BloHl,L formed in situ by reversible dissocation of BloH12L2. Clearly positive identification of these active species which may also be implicated in [B,oHlo]2- formation is desirable.The general synthetic methods namely cage expansion by carbaborane reduction in the presence of metal halide and sodium cyclopentadienide direct insertion of metal fragments (e.g. d" complexes) into a polyhedral cage and the polyhedral subrogation reaction have been described in detail in previous Annual Reports and recent survey^.^^*^ These methods have been extended this year. Only a few novel examples can be mentioned. Until this year the only v-hydrocarbon ligand which had been extensively utilized as a capping ligand in metallocarbaboranes was CSH5-. Derivatives with 778-Cg&7d and q6-CloHg6' have now been synthesized. Thus reaction of [(C,H,)TiCl] with Na,C,B,H, followed by treatment with Et,N' yielded [Et,N][3,1,2-(q8-C,H,)TiC,B,Hll].Reaction with H202gave [3 1,2-(q8-C,H,)Ti~B,Hll].7d Reaction of [2,1-(q5-C5H5)CoCBloHll]sodium with naphthalide followed by treatment with C5H5- and Co" gave orange diamagnetic [2,1-(q6-CloH,)CoCBloHll], formulated as (6) {cf.[(C,,H,)Cr(CO),J) and the first metallocarbaborane with a neutral q6-arene ligand.6" There are other metallocar- baborane complexes purported to contain qz-and q4-arene ligands in the recent literature but these must be considered highly speculative. In view of the interest and (a)W. E. Hill F. A. Johnson and R W. Novak Inorg. Chem. 1975,14,1244; (b)M.F.Hawthorne,J. OrganometaNicChem.,1975,100,97;(c)F.G.A. Stone J. OrganometallicChem. 1975,100,257;(d)C. G.Salentine and M. F. Hawthorne J.C.S.Chem. Comm. 1975 848; (e) C.G.Salentine and M. F. Hawthorne,J. Amer. Chem.SOC.,1975,97,426;(f) F.Y.Yo,C. E. Strouse,K. P. Callahan C. B. Knobler and M. F. Hawthorne,J. Amer. Chem. Soc. 1975,97,428;(g)P.L.Timms,Angew. Chem. Internut. Edn. 1975,14,2/3;(h) E.L.Hoe1 and M. F. Hawthorne J. Amer. Chem. Soc. 1975,97,6388. A.J.Carty,R. H. Cragg and J.D. Smith controversy surrounding the structures of 'titanocenes' and related complexes the characterization of [Me4NI2[{ 1,6-C2B10H10Me2}2Ti] as a 14-electron titanacar- baborane is From the few studies presently available it appears that metallocarbaborane complexes of the early transition metals are considerably more stable than the &responding cyclopentadienyl complexes. Whether these metallo- carbaboranes undergo the facile thermal rearrangements typical of the cyclopen- tadienyl titanium complexes remain to be established.The use of metal vapours in the cage expansion of carbaboranes is a potentially useful but to date little used synthetic method. Thus nickel vapour reacts with the closo-carbaborane C2B9H11 to give the known complex [(C2B9H 1)2Ni] derived from the nido-anion [C2B9Hll]2-.7g A variety of B-a-carbaboranyl complexes of iridium have been obtained by oxidative addition of terminal B-H bonds to iridium@ ~pecies'~ [equations (5)and (611. [Ir(C8H14),C1] + 6 1-PMe,-1,2-C,B,,Hll -+2[IrL,Cl] (5) (L) [IrL,Cl] + [IrL,( 1-PMe,-1,2-C2B,,,H ,)HCl] (6) (7) The complex (7) has a cis-octahedral stereochemistry at iridium with three phos- phorous atoms cis to the hydride and the chloride trans.The carbaborane cage is probably attached to the metal via a boron atom in the 3- or 6-position. Intermolecu- lar oxidative addition is also possible. Thus reaction of 1,2-C2B10H12 with [(Ph,P),IrCl] gave small amounts of 3-[(PPhJ2IrHC1]- 1 2-C2BloHl Substitution on the cage was at the 3,6-sites. These reactions thus complement the methods discussed in last year's Report for synthesizing B-a-carbaboranyl complexes. Structures of Carbaboranes and Metallocarbaboranes. With the availability of auto- mated diffractometers X-ray analysis has become a rapid analytical tool not least in the field of metallocarbaboranes. Structures are too numerous to list. We mention only a few important structural trends.X-ray data are now available for a wide variety of twelve-vertex metallocar- baboranes in which the metal d-electron configurations vary from d2to d9.Electron-rich d8 and d9 complexes e.g. [( 1,2-C2B9H11)2Ni"]2-,have 'slipped' sandwich structures with the distortion largest in the d9 cases. For electron-deficient metal- locarbaboranes of which the 15-electron complex Cs[{C2B9H9Me2},Cr] is an exam- ple a symmetrical structure with long metal-ring distances is observed. Long metal-ring bonds are also evident in the 14-electron 13-vertex cage complex [Me4FJ12[( 1,6-~BloHloMe2)2Ti]. An alternative explanation of the slippage in the electron-rich species has now been proposed.8" Basically comparison with the MO energy-level diagram of ferrocene is useful.In the bis-dicarbollide systems the orbitals corresponding to the 2el MO's of ferrocene are antibonding with respect to (a)P. A. Wegner Inorg. Chem.,1975,14,212;(b)W. T. Robinson and R. N. Grimes,Znorp. Chem.. 1975 14,3056;(c)G. K. BarKer M. Green J. L. Spencer,F. G. A. Stone B. F. Taylor and A. J. Welch J.C.S. Chem. Comm.,1975,804;(d)K. P. Callahan W. J. Evans F. Y. Lo,C. E. Strouse and M. F. Hawthorne J. Amer. Chem. Soc. 1975,97 296. The Typical Elements both the metal-ring interactions and the icosahedral cage framework. In d8and d9 complexes two and three electrons occupy these antibonding MO’s. Distortion and eventually cage-opening (closojnido) ensue. The process is analogous to that occurring on reduction of CZ~SO-C~B~~H~~ systems.Although the author mentions that similar distortions do not occur for electron-rich [(C,H,),M] complexes it should be pointed that there are some unusual features in the structure of gaseous nickelocene (a 20-electron complex) in particular that the Ni-C bonds may not all be equivalent. The structure of [2-Me-1,7,2,4-(q-C5H,),Co2GB3H4]consists of a triple-decked sandwich analogous to the [(T-C,H~)~N~~]+ ion. The planar [C2B3H,I4- ring system in 1,7,2,3- and 1,7,2,4-[(q-C,H,),Co2C2B3H5] is isoelectronic and isostructural with [C,H,]-. Development of metallocene-like polymers based on [C2B3H,l4- thus seem possible.86 P(21’) P(31’) (8) A.J. Carty,R. H. Cragg,andJ. D.Smith The reactions of cZoso-2,4-~B5H7 and doso-1,6-C2BsHlo with [Pt(styrene) (PEt3),] and [Pt(C,H,,)(PMe,),] respectively afforded the novel complexes (8) and (9)." Compound (8) has a closed nine-atom bimetallocarbaborane structure but illustrates that M-M bonds in metallocarbaboranes can be very weak.The Pt-Pt bond length of 3.051(4) A is barely a bonding distance. In (9) an unusual exocyclic Pt-Pt bond is present which can readily be removed to generate nido-[8,8- ((Me3P),}-7,8 lO-CPtCB,H,,]. Despite much effort it does not yet seem possible to predict the course of these d10insertions. It is also becoming obvious (cf. refs. 6c and 6e) that metal-metal bonds in metallocarbaboranes can vary enormously in strength. The frequent occurrence of unusual structures in the products derived from insertion of electron-rich metal fragments into carbaborane cages should not mask the utility of electron-counting rules for rationalizing the basic polyhedral geomet- ries of metallocarbaboranes.For example the structure of [(q-CsH5)2Fe2~B6Hs] is that of a capped tricapped trigonal prism and differs from the idealized bicapped square antiprism of other ten-vertex borane species. However this molecule does not have the 22 electrons required for skeletal bonding according to the 2n +2 With 20 electrons only [(q-CSH5),Fe2C2B6Hs] should be derived from a nine-vertex polyhedron as is observed.8d Boracarbocycles and "heir Metal Complexes.-The search for carbocycles with aromaticity as predicted by the Huckel 4n +2 rule together with the notable stabilization of reactive molecules which can often be achieved by attachment to a suitable transition-metal fragment has motivated much research on metal-hydrocarbon v-complexes.The isoelectronic character of CH and BH- and C-C and B-N units is well recognized and has been exploited in the developing chemistry of carbaboranes and borazines. Substitution of B-H- or B-N for C-H fragments in (4n +2)v carbocycles gives rise to boracarbocycles (10)-(15) or borazacarbo- cycles [e.g. (16)-(17)]. Although derivatives of a few of these are known [e.g. pentaphenylborole corresponding to the polyene of (12Jsa and (14) (15) (16) (17) (a)J. J. J. Eisch H. K. Hota and S. Kozima J. Amer. Chem. SOC.,1969,91,4575; (b)M. F. Lappert in 'The Chemistry of Boron and its Compounds' ed. E. L.Muetterties Wiley New York 1968 Ch. 6; (c) A. J. Ashe tert. E. Myers P. Shu T. V. Lehmann and J. Bastide J. Amer. Chem. SOC.,1975,97,6865;(d) G. E. Herberich and H. J. Becker Angew. Chem. Infernat.Edn. 1975,14,184; (e)G. E. Herberich G. Greiss and H. F. Heil Angew. Chem. Zntentat.Edn. 1970,9,805;cr)R. N. Leyden and M. F. Hawthorne Znorg.Chem. 1975,14,2018; (g) G. E. Herberich H. J. Becker and G. Greiss Chem. Ber. 1974,107 3780; (h)J. J. Eisch and J. E. Galle J. Amer. Chem. SOC.,1975,97,4436. The Typical Elements borazaronaphthalenes related to (16)"] synthesis of other boracarbocycles has been very limited. Recently two have characterized phenyborinate (boraben- zene) anions by two different routes (Scheme 3). Ashe and co-workers have also made alkyl- and bromo-substituted anions.The anions are pyrophoric. Originally the phenylborinate anion was trapped as the [(q5-C5H5)(q6-C,H5BPh)Co]', [(7'-C,H,)( q6-C,H,BPh)Co] or [(q5-C,H,Ph),Co] complexes by ring expansion of q5-C5H5 rings of cobaltocene with phenylboron dichloride." A dicarbollylphenylborinate-cobaltcomplex [3,1,2-{ 1-C6H5( q6-C5BH5)}CoC2B9H has been prepared in analogous fashion.'' For the synthesis of other transition- metal derivatives ligand transfer from the cobalt complexes can be used (cf. transfer of cyclobutadienes) but prior generation of the anion would seem preferable. The 6~-aromatic system (13) can be compared to the cyclopentadienide ion. Indeed the air-stable bis(b0rabenzene)iron complexe~~~~~ appear to resemble ferrocene quite closely.In particular the chemical isomer shift 6 is identical (0.72 mm s-') to that of ferrocene indicating that differences in 0-,T-,and S-components to the bonding if any compensate one another so that no net change in s-electron density at the nucleus is apparent. A comparison of S and A values with those of the known isoelectronic [(v6-C6H&Fe]2' species would be rewarding. The bis(boraben- zene)iron complexes also undergo electrophilic substitution; a monoacetyl deriva- tive has been characterized.'" Ferrocene is however approximately four times more reactive than bis-( 1-methylborabenzene)iron. Although the borabenzene ring is bis-( 1-methoxyborinato)cobalt is asymmetrically bonded to the metal atom with substantially longer Co-B than Co-C bond lengths it must be remembered that the cobalt complex is formally a 19 electron species; a distortion might therefore be expected.An X-ray structure of the corresponding iron complex will be interesting. In view of the stability of these systems it seems clear that attempts to generate complexes derived from other boracarbocycles may well be feasible. In this context the recent synthesis of heptaphenylborepin (18) by a suprafacial sigmatropic re- arrangement of heptaphenyl-7-borabicyclo[2,2,llheptadiene followed by a disrota- tory ring-opening of the bicyclic intermediate is of intere~t.'~ The borepin (18) with six .rr-electrons may well be planar and aromatic; spectra of (18)compare well with spectra of the heptaphenyltropenium ion. Nevertheless corroborative X-ray data seem necessary especially in view of the propensity of other seven-membered heterocycles to undergo valence tautomerism.Diene triene or boratrienyl com- plexes of (18) seem possible. The electron deficiency of boron could also be satisfied by a metal lone pair with interesting sterochemical consequences. 0qpo B-Ref. 9c Bu/\Bu RI I R Ref. 9d Scheme 3 A.J. Carty,R.H. Cragg and J. D.Smith Boron-Carbon m-Bonding in Viny1boranes.-Last year's Report (p. 193) men- tioned briefly evidence for B-C v-bonding in vinylboranes. Spectroscopic support for electron delocalization in vinylboranes has now been summarized.'oa Unfort- unately previous assessments of B-C bond shortening have been hampered by a lack of accurate structural data.The gas-phase structure of trivinylborane has now been determined."' The molecule is dynamic having a planar BC skeleton but with extrfme thermal motion of the vinyl groups. The B-C bond length of 1.558(3) A is shorter at the 4a level than the B-C bond length in triphenylborane [1.577(5) A av]. Comparison of these distances is valid since both compounds have Bsp2-Csp2 links. The inference of B-C 7r-bonding appears justified although the extent of bond shortening per vinylgroup is slight. Carbon-13 n.m.r. data have been interpreted as indicating that electron drift from carbon to boron is maximized in the monovinylboranes despite the presence of competing w-donors on boron."" In view of the predicted strongly mesomeric v-donor properties of fluorine in species such as H2BF," it seems unusual that B-C w-bonding should fall off in the sequence F2B(C2H3)>FB(C,H,) >B(C,H,),.Indeed ab initio studies show that the absolute charge on C of the vinylboranes F,B(C,H,),- decreases with increased vinyla- tion."' Delocalization of vinyl 7r-density is thus greatest in trivinylborane. It is unfortunate that structural data for F,B(C2H3)'Od are not sufficiently accurate to allow a comparison with B(C,H,),. Comparison of calculations for vinyl- and methyl-boranes shows that a vinyl group donates more than twice as much w-density to boron in F,B(GH,) FB(C,H,), or B(C2H3)3 than a methyl group in F2BMe FBMe, or BMe,. Hence B-C (pm-pm)interactions in vinylboranes are considerably more important than hyperconjugation in methylboranes.Boron Halides.-The extent of B-X (p,-p,) bonding in the boron halides has intrigued inorganic chemists for years and is directly pertinent to discussion of relative Lewis acidities. An attractive approach to this problem would be compari- son of structural and spectroscopic properties of the mixed halides BXAXZ- where competitive B-X' and B-X2 p,-p interactions might be expected. Unfortunately the pure mixed halides are not isolable. Despite this an elegant study"" has recently shown that from a statistical equilibrium mixture of BCl,F,- (n = 0-3) photo- electron and microwave spectra of individual species can be obtained. For example for BClF2 IP's of 12.85 (4b,) 13.00 (2b1) 15.1 (5a1) 16.93 (3b2 a2),and 18.35 (lb,) eV were experimentally observed.Ionization from 4a1 (calc. 18.58 eV) and (a)L. W. Hall J. D. Odom and P. D. Ellis J. Amer. Chem. Soc. 1975 97 4257; (b) A. Foord B. Beagley W. Reade and I. A. Steer J. Mol. Structure 1975 24 131; (c) N. J. Fitzpatrick and N. J. Mathews J. Orgunometuific Chem. 1975 94 1; (d)J. R. Durig R. 0.Carter and J. D. Odom Znorg. Chem. 1975 13,701. The Typical Elements 2b2(calc. 19.32 eV) wasnot detected. AB-Clbondlengthof 1.71(1) AandanFBF angle of 116.6(1)0 were derived from the microwave spectrum of this compound. The B-Cl bond has 14% ?r-character; ab initio calculations agree with this estimate and indicate that the B2p,-C13pT overlap population is greater than B2p,-F2pT. This result agrees with earlier EHMO calculations on the trihalides BX where n-charge transfer from halogen to boron decreased in the order BI > BBr > BCl > BF although the c7-charge drift X+B dominates contributions to the overall bond polarity."' This opinion is not shared universally however.The a priori prediction of a more compatible p,-p, overlap in the B-F bond has usually led to the text book inference of greatest 7-bond order in the fluoride. This view still persists.10a Nevertheless the order of n-back-donation from ligand to boron B-N > B-S = B-I > B-0 =r B-Br >B-Cl> B-F deduced earlier seems realistic and can be compared with B-N > B-0 > B-F calculated for H,BX species." While halogen bridging is common for the heavier Group I11 metal halides it is rare in boron chemistry.At very low temperatures (-155 "C)1 1mixtures of BF3 and [Bu",N]'[BF,]-give 19F n.m.r. spectra consisting of a high-field doublet (JFBF 95f10 Hz)and a broad low-field resonance with area ratios of 1 6 consistent with the presence of the very labile single-fluorine-bridged species [B2F7]- the first time this ion has been observed in The Boron-Phosphorus Bond.-The nature of the co-ordinate bond between boron and phosphorus in simple co-ordination compounds has been a source of considerable controversy. Some of the more puzzling aspects which appear to have largely defied logical explanation to date are the variations in P-B bond lengths which accompany changes in the substituents on boron and phosphorus together with some rather unusual stability sequences.Thus H3P,BH3 and F,P,BH3 have P-B bond lengths [1.937(5) and 1.836(6) A] near the extremes for borane (BH,) adducts yet these compounds are considerably less stable than HF2P,BH3 [B-P 1.832(9)A] Me,P,BH [B-P 1.901(7)813 or HMe2P,BH [B-P 1.906(6) A]. The P-B bond lengths of 1.921(7)w and 1.84(2) 8 recently determined for H3P,BH312a and MePF2,BH3l2* respectively the former a compound of low dis- sociative stability shed no further light on this matter. The most that can be said at present is that for BH adducts the B-P bond lengths lie in two groups with B-P bonds in fluqrophosphine adducts ca. 0.06A shorter than in the phosphine and methylphosphine series. For boron trihalide complexes the situation is clearer. Crystal structures have been determined for Me,P,BX3 (X= C1 Br or I) this year.'2c Boron-phos horus bond lengths are 1.957(5) 8 (X= Cl) 1.924(12) 8 (X= Br) and 1.918(15) 8 (X= I).There is a significantly shorter B-P bond in the chloride complex compared with the bromide or iodide. Thus the pattern of donor-acceptor bond strengths and (a)H. W. Kroto M. F. Lappert M. Maier J. B. Pedley and M. Vidal J.C.S. Chem. Comm. 1975,810; (b)M. F. Lappert M. R. Litzow J. B. Pedley P. N. K. Riley and A. Tweedale J. Chem. Soc. (A) 1968 3105; (c)M. F. Lappert M. R. Litzow J. B. Pedley P. N. K. Riley T. R. Spalding and A. Tweedale J. Chem. Soc. (A),1970,2320; (d)J. S. Hartman and P. Stilbs J.C.S. Chem. Comm. 1975,566. l2 (a) J. D. Odom V. F. Kalasinsky and J. R. Durig Inorg. Chem.1975,14,2837; (6)R. A. Cresswell R. A. Elzaro and R. H. Schwendeman Inorg. Chem. 1975,14,2256;(c)D. L. Black and R. C. Taylor Acta Cryst.,1975 B31 11 16; (d)D. C. Mente and J. L. Mills Inorg. Chem. 1975,14,1862;(e)P. Cassoux R. L. Kuczkowski P. S. Bryan and R. C. Taylor Inorg. Chem. 1975,14,126;cf) P. M. Kuznesof F. B. T. Pessine R. E. Burns and D. F. Shriver Inorg. Chim. Acta 1975 14 271. A.J. Carty,R.H. Cragg and J. D. Smith possibly Lewis acidities evident in the Me,N,BX and MeCN,BX series is followed with BI LBBr >BCl,. This order receives direct support from gas-phase calorimetric measurements.'2d Furthermore the B-P bond length in Me,P,BH [1.901(7) A] is shorter than in any of the halide complexes Me,P,BX whereas for Me,N,BX (X = H C1 NF) the distances are virtually identical.'" The B-N bond length in Me,N,BH [1.638(10) A] is longer than in Me,N,BX (X =Br or I).Thus the greater affinity of BH3 than the boron halides towards phosphine donors predicted chemically seems substantiated. CNDO-2D calculations of dipole. moments for Me,H,-,E and Me,H,-,E,BH (E =N or P; x = 0-3) provide evi- dence for substantial differences in B-N and B-P bonds in these simple adducts.12f Co-ordination of NH to BH results in charge transfer primarily between N-bound and H-bound hydrogens (0.33e) but for phosphine complexes the transfer is largely from phosphorus to boron (0.27e). Whereas the B-N bonding MO remains essentially unchanged on methyl substitution (46% covalent character in the B-N bond with essentially sp2.' hybridization at nitrogen) substantial changes in hybridi- zation at phosphorus (from ~p'.~ in PH,,BH to spl.' in Me,P,HB,) occur on formation of B-P bonds.The B-P bond has considerably more covalent character ((51%) and a higher s-character. It also appears from these calculations that distortion of the lone-pair electrons of Me,P and PH by BH is more favourable for the former ligand. The Boron-Nitrogen Bond.-Research in boron-nitrogen chemistry is focused to a large extent on compounds where B-N n-bonding plays a major role. This year synthetic methods have been developed for several molecules whose ground-state properties should be interesting. Thus the first 1,3,2,4-diazaboretidineshave been synthesized [equation (7)].13" X I R XR B /\ R,N-P=NR + BX3 + \/N-P-N I \ +RN \/NR (7) I R BX2 P I Y Convenient routes to dialkylaminohydridophenoxyboraneshave been described.13' Restricted rotation about B-N bonds in these molecules is expected and experi- mentally observable by means of n.m.r.Aminodifluoroborane H2NBF2 has been characterized as a volatile product of the purolysis of H3N,BF3 at 185°C.'3c Polycyclic borazines (19) have been prepared from thioborane~.'~~ X = 0,NH or NMe n = 2 or 3 (19) l3 (a) E. Niecke and W. Bitter Angew. Chm. Internat. Edn. 1975 14 56; (b) R. A. Kovar and G. G. Waldvogie Inorg. Chem. 1975 14 2239; (c) E. F. Rothgery H. A. McGee jun.,and S. Pusatcioglu Znorg. Chem. 1975 14 2236; (d) R. H. Cragg and A. F. Weston J.C.S. Dalton 1975 1961; (e) A.Serafini and J. F. Labarre J. Mol. Structure 1975,26 129; (f) D. T. Haworth and V. M. Scherr J. Inorg. Nuclear Chem. 1975,37 2010; (g)A. DeStefano and R. F. Porter Inorg. Chem. 1975 14,2882. The Typical Elements As regards cr-and n-components to bonding between boron and first-period elements borazine B3N31& and boroxine B303H3 are key compounds. The influence of n-delocalization on the electronic structures of these planar six- membered ring compounds has been examined by ab inifio SCF-LCAO-MO calculations.13c In borazine there is a formal a-charge transfer from HB and B towards nitrogen (1.02e) and a much smaller n-charge transfer from N to B (0.35e). Thus nitrogen is more negatively charged. For boroxine the a-transfer from H and B to 0is 1.14e and there is a reverse .rr-transfer of only 0.26e.Hence the direction of polarity of the B-0 bond is similar to that of the B-N bond the n-overlap population of the B-0 bond is less and the total charge transfer along the ring bonds is much larger in boroxine. In fluorinated boroxines the total ring population is decreased. Electron withdrawal through the cr-system is not counterbalanced by F2p,-B2pT bonding.13' These calculations thus support the inferences made for these molecules on the basis of their chemical reactivities. In this context the likely structures (20) and (2 1) for the protonated borazine cation (cf.protonated benzene) have the proton located at a site close to nitr~gen.'~~ H H (20) (21) -Boron Carbides.-The history of the boron carbide crystal structure goes back 35 years to the initial description of a 15-atom unit cell containing a nearly regular icosahedron of boron atoms and a linear chain of supposedly three atoms linking icosahedra.Much later it was realized that the central atom of the three-atom chains was boron. This can be represented as (B12) [CBC]. The chemical composition B,C can then be attained by the average substitution of a carbon atom for boron in each icosahedron uiz. (B,,C)[CBC] a feature which was very recently revealed by careful X-ray work. With the ideal rhombohedral (B12)[CBC] and carbon-rich (BllC) [CBC] structures established the question remains as to structural modifications accompanying carbon depletion of the ideal B13C2 structure.The boron-carbon phase diagram indicates a single phase in the solubility range 9-20 atom YOcarbon. Several solutions to this problem have been suggested none of which corresponds to the structure of a boron-rich boron carbide determined this year.14 The crystal system of this carbide (containing 8k1atom O/O C) is rhombohedral (Rjrn)but unit cell parameters are significantly larger than for the material with 20% C. Icosahedra (B,2) are still present but one fourth of the linear [CBC] chains in the (B,,)[CBC] structure are replaced by planar [B4] groups. Both terminal and bridge atoms of these [B4] units have five-fold co-ordination. Thus the following changes may accompany decreases in carbon content for boron carbide l4 H. L.Yakel Actu Cryst.1975 B31,1695. A.J. Carty,R.H. Cragg,and J. D.Smith The extent to which the presence of increasing numbers of [B4] units in boron carbides contributes to the known solubility limits in the boron-carbon phase diagram remains to be deciphered. 2 Aluminium Gallium Indium,and Thallium Structure and Bonding in Aluminates.-Two particular studies are singled out for mention. Few inorganic compounds are as essential to our society as tricalcium aluminate 3Ca0,A1203 (C,A) a major component of Portland cement. Despite the efforts of numerous laboratories dating back to 1929 the structure of C,A remained unsolved until this year. The structure (22) finally elucidated by Mondal and Jeffrey"" after 12 attempts must stand as a monument in cement and aluminium- oxygen chemistry and consists of six A10 tetrahedra (AlhOI8) eight to a cell surrounding holes of radius 1.47 A with Ca2+ ions in distorted six-fold co-ordination holding the rings together.The presence of rather short Ca-0 contacts (2.26 A) and the observed compression of CaO octahedra may indicate that strain together with the availability of large holes in the lattice facilitates a rapid break-up of the structure on reaction with water to give the initial hydration product 2Ca0,A1203,8H20 and finally the hexahydrate 3CaO,AI2O3,6H,O. This structure should form the basis for an understanding of the effects of impurities on the reactivity of cement. The SCF-X method has been used for the first time to calculate the electronic structure of the aluminate ion [A1O4I5- for comparison with the isoelectronic and (u)P.Mondal and J. W. Jeffrey Actu Cryst.,1975 B31,689; (b)J. A. Tossell J. Amer. Chem. SOC.,1975 97,4840. The Typical Elements 115 isostructural [Mg0,I6- and [SiO,]"- ions.156 The calculated MO energies agree qualitatively with the separations of the Ks. (assigned to the 3t2 MO) and Ks (assigned to the 41 MO) peaks in the X-ray emission spectrum. The calculations also indicated that peaks at 9.3 and 11.1eV in the U.V. spectra of a natural phlogophite mica are due to the [AlO,]'- unit. Si3d-02p (u-and ?r-types) bonding does occur in the [SiO,]"- ion but it is relatively small in magnitude compared with Si3s-02p and Si3p-02p bonding. For [AlO,l5- the 51 MO is of similar energy to that in [S'iO,]".Thus a small A13d-02p bonding component may still be present. A sharp decrease in the strength of M3s-02p (M =Si Al or Mg) bonding is largely responsible for decreasing covalency and increasing instability of the tetrahedral clusters in the sequence [SiO,]"- >[AlO4I5->[Mg0,I6-. The Co-ordination Sphere of Aluminium in Solution.-The n.m.r. probes 'H 27Al 31 P and I3C have all been utilized to investigate the inner co-ordination sphere of A13+. Solvation numbers of 6 are the rule. However in ethanol [Al(EtOH)4]3+ is the dominant species. By utilizing both 27Aland ,'P n.m.r. Delpuech and co-workers'6 have demonstrated the existence of octahedral [AIL6],+ [L =(MeO),PO,(EtO),PO (MeO),MePO,(EtO),EtPO or (MeO),HPO] and the tetrahedral solvate [L = (Me,N),PO] in anhydrous nitromethane.Kinetic data are indicative of a dissociative SN1 ligand-exchange mechanism for [AIL6],+ while conversely an associative SN2 mechanism is applicable to [Al{(Me,N),P0},]3'. The activation energy for the associative process is dramatically smaller by ca. 12 kcal mol-' and ligand exchange on A13+ in (Me,N),PO is five orders of magnitude faster than in systems where A13+ is six-co-ordinate. These observations may have important implications for the synthetic and solution chemistry of A13+, especially if they can be generalized for other solvent systems. It is interesting that whereas activation enthalpies for exchange in [AlL6I3+ (L =DMF or DMSO) seem to indicate dissocia- tive pathways for [GaL6I3+ (L=H20 or DMF) associative mechanisms may be operative.Halide ion exchange on [GaCl,]- is also associative. Aluminium-Halogen Compounds.-The heats of reaction of Group I11 halides and organometallics with neutral ligands containing Group V and VI donor atoms have been measured over the past 20 years in an attempt to compare acceptor properties. Recently Wood and ~o-workers~~~ have determined the necessary heats of forma-tion crystal structures and lattice energies for M1[M2X,] salts (M' =Na or Cs M2=A1 or Ga X =C1 or Br) to calculate M2X3-X- donor-acceptor bond energies. (Average heats of dissociation DMzx3-x-(kcal mol-l) are 82 ([GaCl,]-) 87 ([AICl,]-) 75 ([GaBr,]-) and 80 ([AlBrJ). Hence the halide ions form stronger donor-acceptor bonds than the neutral ligands.An estimate of the InCl,-Cl- bond energy by a newly developed method suggests that towards the halide ions the order of acidities is InCI >AICI >GaCI,. These results can be compared with the relative acidity order BCl >AlCl >GaCl >InCl towards ethyl acetate as the reference base the order A1 >Ga >In for Ph3M2 towards pyridine and the common sequence l6 J. J. Delpeuch M. R. Khaddar A. A. Peguy and P. R. Rubini J. Amer. Chem. Soc. 1975,97 3373. l7 (a)R. C. Gearhart jun. J. D. Beck and R. H. Wood Inorg. Chem. 1975,14,2413;(b)G. K. Barker M. F. Lappert J. B. Pedley G.J. Sharp and N. P. C. Westwood J.C.S.Dalton 1975,1765; (c)J. L.Dehmer J. Berkowitz L. C. Cusachs and H. S. Aldrich J. Chem.Phys. 1974,61,594;(d)K. Wittel and R. Manne J.Chem. Phys. 1975,63 1322; (e)R. G. S. Pong R. A. Stachnik A. E. Shirk and J. S. Shirk J. Chem. Phys. 1975,63 1525. 116 A.J. Carty,R.H. Cragg,and J. D.Smith Al > B -Ga > In for bulky amines. Additional information on M-X p,,-p, bonding and Lewis acidities for monomeric Group I11 acceptors has been sought using p.e. ~pecfro~copy.'~~ For gaseous MCl (M = B Al Ga or In) the relative orbital energies increase in sequence u; e' e" a e' a;. The a; orbital which is primarily responsible for .rr-bonding in MX has energies indicative of greater M-X T-bonding in the boron compounds. These authors suggest that the baricentre of the first two IP's for the MX species provides an estimate of residual charge on the halogen atom and the relative .rr-density on the central metal atom.Thus A1 > B -Ga >> In if this is considered a measure of Lewis acidity. However the same criterion yields BCI >>BBr3 contrary to the usual order of affinities. As a general observa- tion changes in electronic and structural properties for both donor and acceptor occur on complexation and reliable orders of acidities cannot generally be expected from a single physical measurement. A further item of interest is the splitting observed in the e" ionization energies of the iodides. Although Lappert et al. favoured a second-order spin-orbit interaction as an explanation of this effect this has been rejected earlier.'" Instead it was proposed that GaI had a pyramidal ground state. A reinterpretation of the phenomenon has since appeared confirming the spin-orbit origin of the ~p1itting.l~~ Moreover in argon matrices all of the gallium halide monomers are planar D3h molecules as intuitively expe~ted.'~' Compounds with Aluminium-Hydrogen Bonds.-One notable development in the past five years has been the synthesis and characterization of a series of poly-(N-alkyliminoalanes).Several routes to these compounds are available e.g. 1 MAlH + RNH hydrocarbon* -(HAlNR) + 2H2 + MH A simpler direct method has been developed:18" 1 1 A1 + RNHZ + -(HAlNR) + -Hz (9) n 2 These polyiminoalanes have fascinating cage structures. A typical structure that of the alane adduct [HAINPr'],AlH is shown in (23).18* Finally mention should be made of an important study on the mechanism of hydroalumination of Group IV substituted alkynes.18c For trimethyl(pheny1- ethynyl)silane a kinetically controlled cis-hydroalumination by R,AlH is followed by a rapid isomerization to the trans-adduct.Carbon-silicon and/or carbon- aluminium p,,-d bonding as in (24) and (25)may promote cis-trans isomerization. Transition Metal-Gallium,-Indium and -Thallium Bonds.-The chemistry of compounds with transition metal-M (M = Ga In or Tl) bonds has been developed. Reaction of Me,Ga with [q-C5H5W(C0),H] gave [q-C5H5W(CO),]3Ga.1y~ The 18 (a)S. Cucinella A. Mazzei and G. Dozzi J. Orgunometullic Chem. 1975,84 C19; (b)G. Perego M. Cesari G. Del Piero A. Balducci and E. Cernia J. Orgunomefullic Chem.,1975,87,33;(c)J. J. Eisch and S. G. Rhee J. Amer.Chem.SOC.,1975,97,4673. 19 (a)A. J. Conway P. B. Hitchcock and J. D. Smith J.C.S. Dalton 1975 1945; (b)A. T. T. Hsieh Inorg. Chim. Acfu 1975,14,87;(c)H. J. Haupt. F. Neumann and H. Preut. J. Orgunometullic Chem. 1975,99 439; (d)S. G. Pedersen and W. R. Robinson Inorg. Chem. 1975,14,2360;(e)S. G. Pedersen and W. R. Robinson Inorg. Chem. 1975,14,2365. 117 The TypicalElements OAl OC ONOH (23) stereochemistry of gallium is as expected trigonal planar. This complex completes the series [q-C5H5W(C0)3]3M (M =Ga In or Tl).19b Interestingly the correspond- ing aluminium complex [q-C5H5W(C0),]A1,3THF does not contain Al-W bonds. Treatment of indium metal with [Re,(CO),,] in a bomb afforded [Re,(CO),(p- InRe(CO)5)21(cf. [Mn,(CO),{p-InMn(CO)5},1) and [Re4(Co>12{p3-InRe(c0),),1 having a tetracapped tetrahedral structure.19' There is an obvious analogy between the terminal transition-metal moieties in these species and bulky alkyl groups. Furthermore the instability usually associated with Tl' organometallics is apparent in the behaviour of Tl'-transition metal Thus reaction of Tl[Co(CO),] with a phosphine (L) causes disproportionation to thallium metal and Tl[Co(CO),L],. Only less basic phosphites yield Tl' derivatives. Thallium(1)-iron -vanadium or -chromium bonds are only stable when the corresponding transition- metal anion is weakly basic.lge Synthesis of Indium (111) and Indium (I) Compounds.-The development of a direct electrochemical synthesis of indium compounds warrants special mention.20a The method which can be used to produce neutral anionic or cationic complexes uses a cell with an indium anode a platinum cathode and a non-aqueous solvent system Ph SiMe Ph SiMe, \+ \+ / /c-c' // c-c, \ H AIR; (24) (25) 2o (a)J.J. Habeeb and D. G. Tuck J.C.S.Chem. Comm. 1975,808;(b) J. J. Habeeb and D. G. Tuck J.C.S. Dalton 1975 1815. 118 A.J. Carty,R.H. Cragg,and J. D.Smith (usually benzene-methanol). Applied voltages of 50-100 V at 20-100 mA for 1-3 h gave gram quantities of complexes. The facile synthesis of anhydrous InCl (cf.burning indium metal in an atmosphere of dry chlorine gas) and Et,N[InI,] (from Et4NI Iz and In) are notable. Apparently the method can be extended to the synthesis of transition-metal halide compounds e.g.CrCl and may therefore be applicable to organometallics. Developments will be awaited with interest. The co-ordination chemistry of indium(1) has been slow to develop owing principally to the intractability of potential percursors such as InCl InzO and In,S. This is in marked contrast to the isoelectronic tin(I1) species such as SnCl which have been extensively investigated both as Lewis acids and as useful synthetic reagents. Structural data for indium(1) compounds are very sparse when compared to those available for tin(r1). Habeeb and Tuck2" have used cyclopentadienylin- dium(1) as an organic solvent-soluble starting material. The reaction CSH,In + HX -P InX + C5H (10) yielded various indium(1) complexes including 4,4,4-trifluoro- 1-(thien-2- yl)butane-1,3-dionatoindium(1) quinolin-S-olatoindium(I) and 2-mercap-togentan-3-onatoindium(1).These indium(1) compounds e.g. quinoline-8-olatoindium(1) [In(qno)] may in their own right be useful starting materials for indium(II1) complexes. Thus acetylacetone yielded [In"'(qno)(acac),] and iodine [In(qno)I,]. The availability of soluble In' compounds such as [In(qno)] may be of considerable use in extending the range of compounds having indium-transition metal bonds via oxidative addition reactions (see above). Pentahalogenometallates(m).-New vibrational data for single crystals of the square-pyramidal ions [MCl,]'- (M = In or TI) together with low-temperature Raman and i.r. studies of polycrystalline samples allow a reassignment of vibrational spectra.21 In the isomorphous tetraethylammonium salts [MC151Z- ions reside on sites of Czsymmetry rather than C4as assumed in previous analyses based on X-ray data.Rerefinement of the earlier X-ray data in the space group P4 including anisotropic temperature factors gave an improved R value of 0.067. However the basic structural features of the molecules remain unchanged. Stereochemical Activity of the Thallium(I) Lone Pair.-Although the ions Ga' In+ and Tl' have ns2configurations there has been little structural evidence to indicate whether this pair of electrons is 'inert' or stereochemically active. X-ray studies of hexafluoroacetylacetonatothallium(I),22" (BU'~NCS~)TI,~~~ and TlzS522c have now shown unequivocally that the 6s' pair plays a major role in dictating the co- ordination geometry around the thallium atom.21 G. Joy A. P. Gaughan jun. I. Wharf D. F. Shriver and J. P. Dougherty Inorg. Chem.,1975,14,1795. 22 (a)S. Tachiyashiki H. Nakayama R. Kuroda S. Sato and Y. Saito Acru Cryst. 1975 B31,1483; (b) H. Pritzkow and P. Jennische Actu Chem. Scand. 1975 A29,60; (c)B. Leclerc and T. S. Kabre Acru Cryst. 1975 B31 1695.