- pReactivity of cyclopentadienyl and indenyl alkylcarbonyliron(n) complexes.q5 Species as key intermediates in the migratory insertion of carbonmonoxideMoira Allevi, Mauro Bassetti,* Claudio Lo Sterzo and Donato MontiCentro C. N. R. di Studio sui Meccanismi di Reazione, c/o Dipartirnento di Chirnica,Universita’ ‘La Sapienza ’, 00185 Rorna, ItalyThe reactions of the complexes [Fe(q’-L)(CO),R] (L = C,H5 or C9H7, R = Me or CHMe,) with PR’,(PMe,Ph, PMePh, and PPh,) have been studied in toluene and tetrahydrofuran (thf) by UV/VIS and IRspectroscopy. The products of alkyl migratory insertion [Fe(q ’-C,H,)(CO)(COR)(PR‘,)] are formed uia anassociative mechanism, involving (i) a rapid pre-equilibrium ( K ) between complex and phosphine and (ii)rate-determining alkyl migration (k,), in analogy to the mechanism previously established for [Fe(q ’-C,H,)(CO),R].In the intermediate molecular complexes {[Fe(q’-L)(CO),R], phosphine) the compounds[Fe(q5-C,H,)(CO),R] bind to phosphine less efficiently than do the indenyl analogues and react one order ofmagnitude more slowly. Reaction rates are similar in thf and in toluene, in agreement with the proposal thatthe alkyl migration is not solvent assisted.Transition-metal complexes having different spectator ligandsof the cyclopentadienyl family (L) may exhibit changes in therate of associative reactions. ’ Remarkable effects were reportedfor ligand substitution in rhodium(1) complexes.2 For instance,[Rh(q5-C,H7)(CO),] (C9H, = indenyl) reacts lo8 times fasterwith PPh, than does [Rh(q5-C,H,)(CO),].This phenomenonhas been described as the indenyl ligand effect, and interpretedin terms of the different stabilities of the corresponding trihaptointermediates [Rh(q3-L)(C0),(PPh,)] or transition states. 2bAlthough the migratory insertion of carbon monoxide is areaction studied in great detail,3 relatively little attention hasbeen given to the effects of ancillary ligands which are not partof the migration step.4 In particular, effects in complexes ofdifferent pentahapto ligands have been described only in a fewcases. Facile carbonylation reactions have been reported for[Fe(q 5-C,H7)(PPh,)(CO)Me] and related complexes, and for[Fe(q5-C,H ,)(CO),(CH(OSiMe,)R)].6 A kinetic study on thereaction induced by tributylphosphine and triphenyl phosphitein molybdenum complexes [MO(~’-L)(CO)~M~] (L = C,H,or C,H,) has shown that the indenyl system is more reactivein both an associative and a nucleophile-independentmechanism.We have studied the CO migratory insertion in alkylcarbonyl(indenyl) iron(I1) complexes and shown that the reactionproceeds through an associative pathway (Scheme 1, L =C,H,; R = CHMe, or Me; PR’, = alkyl, aryl or mixed alkyl-aryl phosphine).8 Although cases of assistance by solvent orby the entering ligand in the migratory insertion have beenpresented.’ our studies have demonstrated that the reaction ofthe indenyl system is induced by the incoming phosphine, andthat the key intermediate is a molecular complex of the tworeactants, changing into product by carbonxarbon and iron-phosphorus couplings.This situation is in contrast with theproposed mechanism for migratory insertion, which describesthe process in terms of a two-step sequence characterized by (i)Scheme 1alkyl migration forming a co-ordinatively unsaturated orsolvent-saturated acyl species, and (ii) attack of phosphine toyield the final acyl We have also proposed thatthe intermediates of Scheme 1 are electron-donor-acceptorcomplexes, in which the structural features of the interactingmolecules are largely preserved, the indenyl ligand bindingthrough an q5 mode.8bWe have extended our studies to the cyclopentadienylcomplexes in order to compare the reactivity and mechanism ofindenyl- and cyclopentadienyl-iron(I1) species.The reactions ofiron cyclopentadienyl complexes [Fe(q’-C,H,)(CO),R] (R =Me, Et or Pr‘) with phosphines in polar solvents have beenpreviously described. ’ The alkyl to acyl migratory insertionhas also been studied in cationic complexes [Fe(q5-C,H5)-(CO)(PR,)(Me)] + and evidence has been presented that thereaction involves transient 1 9-electron seven-co-ordinate com-plexes [Fe(q’-C,H,)(CO)(PR,)(X)Me]+ (X is the incomingligand).4b*’Experiment a1The reactions and characterization of starting materialsand products have been reported previously.8*’ 1 9 1 3 3 1 4 Thephosphines were commercially available and used as received.Toluene was distilled under argon from sodium, and tetra-hydrofuran from potassium, in the presence of benzophenone.The methods used for kinetic measurements and for dataanalysis are described in detail in ref.8(a). Solutions of the ironcomplex and phosphine were prepared under argon, usingSchlenk-line techniques, in 1 cm quartz cells for UV/VIS, or 0.5mm CaF, cells for Fourier-transform IR measurements. Initialconcentrations of iron complexes were ca. lop4 (UVjVIS) andca. 0.01 mol dm-, (FTIR). Pseudo-first-order rate constants(kobs) were obtained by fitting the exponential dependence ofthe absorbance us. time data using a non-linear least-squaresregression program, which also provides the absorbance aftercompletion, A *, for slower reactions. Duplications of singlekinetic runs were reproducible to within 6%. The uncertaintiesof the reaction parameters K and k , are reported as standarddeviations, obtained from non-linear least-squares calculations.Activation parameters were obtained by an unweighted linearleast-squares analysis of the dependence of ln(k,/ T ) on 1 /T.J.Chem. SOC., Dalton Trans., 1996, Pages 3527-3531 352Table 1[Fe(q5-L)(CO),R] and [Fe(q 5-L)(CO)(COR)(PR'3)] complexesCarbonyl stretching frequencies ( ? 0.2 m-', hexane) forComplex L PR', ij(CO)/cm-'1 C5H5 * 2014.61961.2C5H5* PMePh, 1924.12 C9H7a9b 2011.71957.9C9H7"9' PMePh, 1923.7C9H7u7b PMe,Ph 1920.5C9H7 PBu", 1919.43 C5H5'qd 2006.41952.14 C9H7b9c 2003.71950.2C5H5' PPh, 1921.3C5H5' PMePh, 1921.9C5H5' PMe,Ph 1917.2C9H7' PPh, 1920.8C9H7' PMePh, 1919.4C9H7' PMe,Ph 1916.4* R = Me.Ref. 8. ' R = CHMe,. Ref. 14.F(COR)/cm-'160716171615162016101606160516141613161 1ResultsDuring our study on the reaction of [Fe(q5-C9H7)(CO),Me]with phosphines in toluene we observed a very slow reaction of[Fe(q5-C5H5)(CO),Me] 1 with PMe,Ph and PMePh, and acomplicated kinetic behaviour that did not allow a meaning-ful comparison of rate and mechanism for indenyl andcyclopentadienyl species. We have now observed morereproducible and first-order behaviour for the reaction carriedout in tetrahydrofuran (thf). The reactions of complex 1 and of[Fe(q5-C,H,)(CO),Me] 2 with PMePh, and PPh, proceed tothe formation of the corresponding acyl products, according toequation (1). The isopropyl complex [Fe(q5-C,H,)(CO),-thf50 OC[Fe(q5-L)(C0),Me] + PR', -(CHMe,)] 3 also reacts smoothly with PMe,Ph, PMePh, andPPh, in tetrahydrofuran or in toluene even at roomtemperature to give the corresponding acyl complexes [Fe(q 5 -C5H5)(C0)(COCHMe,)(PR1,)3 in high yields.The migratoryaptitude of the isopropyl group was well suited for akinetic study of both complex 3 1 1 7 1 3 and [Fe(q5-C,H,)-(CO),(CHMe,)] 4.8b*'4 Infrared data for alkyl and acylcomplexes are given in Table 1. The kinetics, carried outunder pseudo-first-order conditions using an excess ofphosphine, were studied by monitoring the product formationin the visible region (40W20 nm), or the disappearance ofthe carbonyl stretching bands of the alkyl complexes in theinfrared. The reaction of 3 with PPh, in toluene was studiedin the range 20-60 "C.The dependence of the rate of reaction ( I ) on theconcentration of phosphine is shown in Fig.1 for [Fe(q5-C,H,)(CO),Me], in Fig. 2 in the reciprocal form l/kobs us.l/[phosphine], in Fig. 3 on a logarithmic scale for bothcomplexes 1 and 2, and in Fig. 4 for 3 and 4. The experimentalpoints are fitted by equation (2), which is derived from theKkZCLl1 + K[L] kobs =mechanism in Scheme 1, yielding equilibrium constants K (dm3mol-') for adduct formation and rate constants k, (s-') for alkylmigration, or by equation (3) for Fig. 2. The y intercepts in the4.03.0 - I z0" $ 2.0u)0F1 .o0 0.2 0.4 0.6 0.8 1 .o[phosphine]/mol dmW3Fig. 1 Plots of kobs us. phosphine concentration for the reactions of[Fe(q5-C5H5)(CO),Me] with PPh, and PMePh, in tetrahydrofuran at50 OC86z .c'2.3*O 4z In20 2 4 6 8[phosphinel-I /dm3 mol-'Fig.2C5HdC0)zMelPlots of Ilkob, us. l/[phosphine] for the reaction of [Fe(q5-1 o41 o4[phosphine]/mol dme3Fig. 3 Plots of kobs us. phosphine concentration for the reactions of(a) [Fe(q5-C5H5)(C0),Me] and (b) [Fe(q 5-C9H,)(CO),Me] with PPh,(0) and PMePh, (0) in tetrahydrofuran at 50 "C1 1 1(3)graph represent the reciprocal values of the limiting rate at[phosphine] = a, i.e. the rate constants k,.The observed rate constants, kobs (s-'), are given in Tables 2and 3 for the reactions in tetrahydrofuran, and in Table 4 forthe reactions in toluene. Proton NMR or FTIR spectra takenduring reaction of 1 or 3, or immediately after mixing, showedonly absorptions due to the starting material and product, nointermediates being detected.3528 J.Chem. SOC., Dalton Trans., 1996, Pages 3527-353Table 2tetrahydrofuran at 50.0 "C *Observed rate constants, kobs, for the reaction of [Fe(q5-C,H,)(CO),Me] 1 and [Fe(q5-CgH,)(CO),Me] 2 with PPh, and PMePh, in1 2CPPh,I/ CPMePh:I/mol dm-, lo6 kobs/s-l mol dm- 1 o6 kobs/S-'0.189 1.48 0.130 1.420.277 1.77 0.253 2.420.463 2.39 0.483 3.060.595 2.57 0.693 3.690.896 3.09 0.976 3.80* [FeL(CO),Me] = 3 x 10-,-8 x rnol dm-,. h = 400-425 nm.~-[Pph,l/ CPMePh,l/mol dm-, 10' kobs/s-' mol dm-, 10' kobs/s-l0.122 1 S O 0.129 3.070.274 2.92 0.253 4.870.444 3.82 0.371 5.850.563 4.33 0.537 6.800.690 4.63 0.693 7.50Table 340.0 "C *Observed rate constants, kobs, for the reaction of [Fe(q5-CsHs)(CO),(CHMe,)J 3 with PMe,Ph, PMePh, and PPh, in tetrahydrofuran atCPMePh21/ CPMe,Phl/mOl dm-, kobs/S-l rnol dm-, kobslS-l mol dm-, kobs/S-l[Pph,l/0.0247 1.05 x lo-' 0.0267 2.56 x 0.035 6.27 x lo-'0.0632 2.87 x 0.040 3.31 x 0.068 1.18 x0.130 6.20 x 0.066 5.91 x 0.254 2.33 x lo40.343 1.31 x lo" 0.231 1.28 x lo" 0.639 2.98 x 10"0.262 1.01 x lo4 0.131 9.87 x lo-' 0.413 2.62 x 10-40.424 1.37 x lo4 0.304 1.66 x 10-40.512 1.68 x 104* 420 nm.Table 4 Observed rate constants, kobs, for the reaction of [Fe(qS-C5H5)(C0),(CHMe2)J 3 with PMe,Ph, PMePh, and PPh, in toluene at 40.0 OCa0.9750.8000.6540.4590.3680.3290.2630.2050.1640.1293.02 x lo4 0.4472.84 x 0.3782.42 x lo4 0.2551.95 x 0.1731.79 x 10" 0.0841.59 x 10"1.31 x 10" 0.3781.11 x lo-" 0.2551.02 x lo4 0.1737.70 x 0.0841.49 x 10-561.18 10-561.03 x 10-561.37 x6.58 x1.17 x8.85 x 1 0 - 4 c6.81 x 10-4"3.71 x 10-4'0.654 2.24 x lo4 0.8060.490 2.08 x 0.5370.327 1.83 x lo4 0.3760.163 1.32 x lo4 0.2180.0654 8.72 x lo-' 0.1090.0326 6.25 x lo-' 0.04360.0164 3.43 x lo-' 0.02140.0109h = 360-400 nm; [3] = 1.5 x 10-,-9.2 x mol dm-3.At 20.0 "C. At 60.0 "C.2.56 x 10-42.43 x 1042.09 x 10-41.63 x 10-41.21 x 1046.80 x lo-'3.30 x lo-'1.60 x lo-'Table 5 Equilibrium ( K ) and rate (k,) constants for the reaction of [Fe(q5-L)(CO),R] with different phosphines in tetrahydrofuran at 50.0 "C[Fe(tls-C5H5)(CO)z.Mel CWrl -C,H ,)(CO),Mel CFe(l15-C5Hs)(CO),(CHMe,)l *Kldm3 K/dm3 K/dm3Phosphine mol-' k,/s-' mol-' k,/s-' mol-' kZ/s-'PPh, 2.5 k 0.3 (4.4 f 0.2) x 2.0 _+ 0.2 (8.1 f 0.4) x 1.2 & 0.3 (4.3 f 0.8) xPMePh, 3.3 f 0.5 (5.1 f 0.3) x lop6 3.0 f 0.2 (1.1 k 0.1) x 3.0 f 0.7 (3.4 k 0.5) xPMe,Ph 6.6 k 0.6 (3.6 k 0.1) x* At 40 "C.DiscussionWith regard to simple ligand effects on the ground state of themetal complexes, Table 1 shows that substitution of C5H5 forthe indenyl ligand yields a shift of the carbonyl stretching bands(vco) to lower frequencies due to stronger x-back bonding fromiron to carbon, as the result of greater electron donation from Lto iron.The same effect is observed on replacement of methylby isopropyl in both indenyl and cyclopentadienyl complexes.l4The reactions of [Fe(q5-C,H,)(CO),Me] in tetrahydrofuranare not first order with respect to the phosphine concentrationand exhibit saturation behaviour (Fig. 1). The limiting values ofreactivity for increasing concentration of PMePh, and PPh,appear t o diverge indicating a dependence on the nature of thephosphine (Fig. 2). Although the selectivity is not as high asthat previously observed for the reactions of [Fe(q '-C,H,)(CO),Me] in toluene,8u nonetheless both the saturationbehaviour and different limiting reactivities suggest that themigratory insertion reactions of the cyclopentadienyl complex[Fe(q 5-C5H5)(CO)2Me] proceed according to the mechanismrepresented in Scheme 2, l 5 as in the case of the correspondingindenyl complexes.The parameters are reported in Table 5 forthe reactions in tetrahydrofuran and in Table 6 for the reactionsin toluene. The activation parameters for the alkyl migrationstep (k2) of the reaction of [Fe(q 5-C5H5)(C0)2(CHMe2)] 3with PPh, in toluene are A# = 98 k 13 kJ mol-' and ASt =0 _+ 33 J K-' mol-'.Since the variations in limiting rate with phosphine areJ. Chem. Soc., Dalton Trans., 1996, Pages 3527-3531 352Table 6 Equilibrium ( K ) and rate (k,) constants for the reaction of [Fe(q5-L)(CO),(CHMez)] with different phosphines in toluene at 40.0 "C[Fe(rl '-C5H5 )(co) 2 (CHMez )I CFe(rl 5-C9H7)(C0)2 (CHMez 11K/dm3 K/dm3Phosphine mol-' k,/s-' mol- ' k,/s-'PMe,Ph 7.4 k 1.0 (2.6 k 0.2) x 23.8 k 1.5 (3.4 k 0.1) x lo-,PMePh, 5.5 k 0.5 (3.1 k 0.1) x 9.8 k 0.9 (3.6 k 0.1) xPPh, 1.6 k 0.2 (3.0 _+ 0.2) x lo-,PPh, 1.2 k 0.1 (5.5 k 0.2) x 1.5 k 0.2 (6.0 k 0.5) x lo-,PPh, 5.6 k 0.5 (2.0 ? 0.1) xa Ref.8(b). At 60.0 "C. At 20.0 "C.[phosphine]/mol dm4Fig. 4with PMe,Ph (0) and PMePh, (0) in toluene at 40 "CPlots of kobs vs. phosphine concentration for the reactions of (a)CWrl 5-C5H5)(CO),(CHMe,)I and (b) [Fe(r15-C,H7)(CO),(CHMe,)l- MeScheme 2relatively small, the alternative of a slow initial migration stepcoupled with competition between rapid reversal fromintermediate to starting material and rapid further reactionwith phosphine to give the product should not be e x ~ l u d e d . ~ " ~On the other hand, when the reactions of complex 3 are carriedout in tetrahydrofuran as well as in toluene, the similar values ofkobs and of the reaction parameters give further support to theproposal that the process is not a solvent-assisted formation ofan acyl complex, which should exhibit observable rate effects inthe two solvents.The same holds for the indenyl methylcomplex 2: K = 2.0 5 0.2 dm3 mol-l, k2 = (8.1 2 0.4) x lop5s-' in tetrahydrofuran; K = 3.5 k 0.8 dm3 mol-', k , =(5.7 k 0.6) x s-' in toluene (50 oC).*a The associativemechanism observed here has been also reported for thereaction of the (pentamethylcyclopentadienyl)iridium(IrI) com-plex [Ir(q 5-C5Me5)Cl(CO)Me]. 'We have not found spectroscopic evidence for the adduct([Fe(q'-L)(CO),R], phosphine}, which is supposed to ac-cumulate during the reaction.As previously discussed in thecase of [Fe(q5-C,H7)(CO),R],8 the intermediates may beregarded as molecular complexes in which the structuralfeatures of the interacting species are essentially unperturbedwith respect to the separated reactants.17 In line with thesuggestion that stronger association between the iron complexand the phosphine is favoured by higher electron density at themetal centre,8b we observe here that the cyclopentadienylcomplexes have smaller values of K than those of thecorresponding indenyl species (Tables 5 and 6). A strongerelectron-donor ability of C,H, us. C,H, toward the metalfragment has been observed by photoelectron spectroscopy inrhodium(1) Complexes,' is shown in Table 1 by the shift of thecarbonyl stretching frequencies in the infrared, and causesfaster dissociation of PPh, from ruthenium(rr) complexes[Ru(q5-L>C1(PPh,),].' This property can therefore affectground states, as well as intermediate and transition states, inwhich the spectator ligand binds the metal in a pentahaptofashion, as it does in the starting complex.The effect of the pentahapto ligands on the rate is shown inFig.3 for the reactions of the methyl complexes 1 and 2 intetrahydrofuran at 50 OC, and in Fig. 4 for the reactions of theisopropyl complexes 3 and 4 in toluene at 40 "C. In the region ofsaturation, the less-reactive methyl complexes and the morereactive isopropyl complexes display a rate difference, C,H7us. C5H5, of about twenty and ten times respectively.Thisdifference is, for the migratory insertion of alkyliron(I1)complexes about the same as that displayed by molybdenumcomplexes [Mo(~~-L)(CO)~M~], in the same reaction withphosphorus donors, in hexane or tetrahydrofuran. The effectis small when compared to that of several orders of magnitudeobserved for carbonyl-substitution reactions of rhodium(I),2but also of iron(1r) 2o or molybdenum(II).21 At the moment, aconsistent 'indenyl effect' for migratory insertion has not beenobserved and, therefore, analogies regarding the mechanismand reaction intermediates (e.g. q species in associativeprocesses) for carbonyl substitution and migratory insertionshould not be drawn.In the chemistry of iron(u) several reactions of indenyl andcyclopentadienyl complexes are available for consideration.Shorter reaction times and lower temperatures have been usedfor carbonylation of [Fe(q 5-L)(PPh,)(CO)Me] when L =C9H7.5 About a 10-20 fold rate difference has been reported inthe epimerization of chiral indenyl and cyclopentadienyl methylcomplexes, which proceeds through a dissociative mechanism.22The rates of insertion of SO2 in [Fe(q5-L)(C0),(CH2Ph)]complexes are the same for L = C,H7 and C5H5 in a reactionwhich proceeds via an electrophilic attack on the benzyliccarbon atom by SO2.,, In studies concerning ethylene rotationit was found that replacement of C5H, by C9H7 in [Fe(q5-C 5 €3 5)(CO)2 (C2 H4)I +Or CFe(q 5-C5H5 )(CO)(C2HdSnR3 11(R = Me or Ph) has an almost negligible effect.24 Carbonylsubstitution, by rate-determining dissociation, in [Fe(q5-L)-(CO),X] occurs 600 times faster for L = C9H7 than for L =C5H5.20 Instead, substitution by P- and As-donor nucleophiles3530 J.Chem. SOC., Dalton Trans., 1996, Pages 3527-353in 19-electron radicals occurs by dissociation of CO which is lo3times faster in [Fe(q5-CsH5)(C0)3]+ than in the indenylcomplex (inverse indenyl effect).25 Among such a variety ofmechanisms and ligand effects, the only case in which indenylcomplexes react much faster than C5Hs analogues is thethermal carbonyl-substitution reaction, which proceeds througha dissociative mechanism. 2o This analysis of the literaturesuggests that q3 species of indenyliron(r1) complexes may notplay a significant role as intermediates.If it were so, rate effectswould be expected to be of several orders of magnitude inassociative reactions, due to the different stability of q3-C9H7and -C5Hs structures, or to be small in dissociative or otherprocesses, which is not the case. Complexes of iron@)containing q 3-L have been isolated and structurally character-ized,26 or observed spectroscopically.27 Therefore, it is ouropinion that ring-slipped species may be sufficiently stable toexist as complexes or to form in sideways equilibria, but are notnecessarily reaction intenriediates in the transformation ofiron(I1) complexes.References1 F. Basolo, Polyhedron, 1990, 9, 1503; see, for instance, C. Bonifaci,G. Carta, A. Ceccon, A. Gambaro, S.Santi and A. Venzo,Organometallics, 1996, 15, 1630.2 ( a ) P. Caddy, M. Green, E. O'Brien, L. E. Smart and P. Woodward,J. Chem. Soc., Dalton Trans., 1980, 962; (b) M. E. Rerek andF. Basolo, J. Am. Chem. SOC., 1984,106,5908.3 R. H. Crabtree, The Organometallic Chemistry of the TransitionMetals, 2nd edn., Wiley, New York, 1994.4 (a) D. Monti, M. Bassetti, J. Sunley, P. Ellis and P. M. Maitlis,Organometallics, 1991, 10, 4015; (b) A. Prock, W. P. Giering, J. E.Green, R. E. Meirowitz, S. L. Hoffman, D. C. Woska, M. Wilson,R. Chang, J. Chen, R. H. Magnuson and K. Eriks, Organometallics,1991,10, 3479; (c) M. Kubota, T. M. McClesky, R. K. Hayashi andC. G. Webb, J. Am. Chem. SOC., 1987, 109, 7569; (d) J. D. Cottonand H. A. Kimlin, J. Organomet.Chem., 1985, 294, 213; (e)G. Cardaci, G. Reichenbach and G. Bellachioma, Znorg. Chem., 1984,23,2936; (f) R. Berger, H. Schenkluhn and B. Weimann, TransitionMet. Chem., 1981,6, 272; ( g ) H. Berke and R. Hoffmann, J. Am.Chem. Soc., 1978,100,7224.5 T. C. Forschner and A. R. Cutler, Organometallics, 1985,4, 1247.6 R. D. Theys, R. M. Vargas and M. M. Hossain, Organometallics,7 A. J. Hart-Davis and R. J. Mawby, J. Chem. SOC. A, 1969,2403.1994, 13, 866.8 (a) D. Monti and M. Bassetti, J. Am. Chem. Soc., 1993,115,4658; (b)M. Bassetti, L. Mannina and D. Monti, Organometallics, 1994, 13,3293.9 M. J. Wax and R. G. Bergman, J. Am. Chem. SOC., 1981,103,7028;S . L. Webb, C. M. Giandomenico and J. Halpern, J. Am. Chem. Soc.,1986, 108, 345; B. D.Martin, K. E. Warner and J. R. Norton,J. Am. Chem. Soc., 1986, 108, 33; F. U. Axe and D. S. Marynick,Organometallics, 1987, 6, 572; T. L. Bent and J. D. Cotton,Organometallics, 1991, 10, 3 156.10 R. J. Mawby, F. Basolo and R. G. Pearson, J. Am. Chem. SOC.,1964, 86, 3994; I. S. Butler, F. Basolo and R. G. Pearson, Znorg.Chem., 1967,6,2074.11 M. Green and D. J. Westlake, J. Chem. Soc. A , 1971,367.12 M. J. Therien and W. C. Trogler, J. Am. Chem. SOC., 1987,109,5127.13 J. D. Cotton, G. T. Crisp and L. Latif, Inorg. Chim. Acta, 1981,47,171.14 L. Ambrosi, M. Bassetti, P. Buttiglieri, L. Mannina, D. Monti andG. Bocelli, J. Organomet. Chem., 1993,455, 167.15 W. P. Jenks, Catalysis in Chemistry and Enzymology, McGraw-Hill,New York, 1969.16 D. Monti, G. Frachey, M. Bassetti, A. Haynes, G. J. Sunley,P. M. Maitlis, A. Cantoni and G. Bocelli, Znorg. Chim. Acta,1995,240,485.17 R. S. Mulliken and W. P. Person, Molecular Complexes, Wiley-Interscience, New York, 1969.18 T. M. Frankcom, J. C. Green, A. Nagy, A. K. Kakkar andT. B. Marder, Organometallics, 1993,12, 3688.19 M. P. Gamasa, J. Gimeno, C. Gonzales-Bernardo, B. M. Martin-Vaca, D. Monti and M. Bassetti, Organometallics, 1996, 15, 302.20 D. J. Jones and R. J. Mawby, Znorg. Chim. Acta, 1972,6, 157.21 A. J. Hart-Davis, C. White and R. J. Mawby, Znorg. Chim. Acta,22 H. Brunner, K. Fisch, P. G. Jones and J. Salbeck, Angew. Chem.,23 S. E. Jacobson and A. Wojcicki, J. Am. Chem. SOC., 1973,95,6962.24 J. W. Faller, B. V. Johnson and C. D. Schaeffer, jun., J. Am. Chem.Soc., 1976,98, 1395.25 K. A. Pevear, M. M. Banaszak Holl, G. B. Carpenter, A. L. Rieger,P. H. Rieger and D. A. Sweigart, Organometallics, 1995, 14, 512.26 T. C. Forschner, A. R. Cutler and R. K. Kullnig, Organometallics,1987,6, 889.27 D. J. Fettes, R. Narayanaswamy and A. J. Rest, J. Chem. SOC.,Dalton Trans., 1981, 231 1; J. A. Belmont and M. S. Wrighton,Organometallics, 1986, 5, 1421; H. Ahmed, D. A. Brown, N. J.Fitzpatrick and W. K. Glass, J, Organomet. Chem., 1991,418, C14.1970,4,441.1989,101, 1558.Received 20th March 1996; Paper 6/01 933AJ. Chem. Soc., Dalton Trans., 1996, Pages 3527-3531 353