528 J.C.S. DaltonReaction Mechanisms of Metal-Metal Bonded Carbonyls. Part X1.lReactions of Nonacarbonyl(tripheny1phosphine)dirhenium and Octa-carbonyl bis(tripheny1phosphine)dirheniumBy David G. DeWit, J. Paul Fawcett, and Anthony Poe,' Erindale College and the Department of Chemistry,The kinetics of the reversible reaction [Re,(CO)8(PPh3)] + PPh, =G== [Re,(CO)8(PPh,)2] + CO indecalin have been studied in each: direction and activation parameters obtained. The kinetic behaviour is quitesimple and is consistent with a ligand-dissociative mechanism. However, it i s also consistent with metal-migration and homolytic-fission mechanisms, The reaction of [Re2(CO)9( PPh,)] with carbon monoxide has quitedifferent activation parameters from the reaction with triphenylphosphine and these reactions cannot, therefore, bothgo via simple rate-determining homolytic fission.Reaction of [Re,(CO) PPh,),] with PPh, leads to mononuclearcarbonylphosphine products and is half order in [Re,(CO),(PPh,),]. This reaction most probably does go viareversible homolytic fission.University of Toronto, Mississauga, Ontario L5L 1 C6, CanadaReactions with oxygen, nitrogen monoxide, and iodine are also described.REACTIONS of the decacarbonyls [Mn,(CO),], [Tc,(CO),,,],[Re,(CO),,], and [MnRe(CO),,], and of their substitutedderivatives, have been shown to proceed according tocomplicated kinetic paths and several mechanisms hav1976 529been proposed, ideas changing as more extensive datawere obtained.l-, Thermal decomposition of [Mn,-(CO),,], in decalin and under an inert atmosphere, is halforder in [complex], while decomposition of [Mn,(CO),]and [MnRe(CO),] under air changes from being halforder in [complex] at high concentrations to first order atlow concentrations.ly2f The effect of carbon monoxideon some of these reactions showed that a simple CO-dis-sociative mechanism could not be operating, but theresults were consistent with an initial reversible homoly-tic-fission pr0cess.l The reaction of [(Mn(CO),(PPh,>>Jwith triphenyl phosphite increases in rate to a limitingvalue with increasing values of [P(OPh),] .26 Althoughthis was initially thought 2b to indicate reversible dis-sociation of triphenylphosphine as the initial step, theinfluence of added PPh, on reactions proceeding at lessthan the limiting rate is not in accord with this,3 andthese slower reactions are, in fact, half order in [complex].*As part of an investigation of the various mechanisticpaths followed by these relatively quite simple metal-metal bonded carbonyls and their derivatives, we havestudied some reactions of the complexes [Re,(CO),-(PPkJI and [(Re(CO),(PPhJI2I*EXPERIMENTAL AND RESULTSDecacarbonyldirhenium (Strem Chemicals, Inc.) was usedas received.Triphenylphosphine and triphenyl phosphite(B.D.H. Chemicals, Ltd.), respectively, were recrystallizedfrom hexane and distilled under reduced pressure (3 mmHg)before use. * 4-Ethyl-2,6,7-trioxa-l-phosphabicyclo[2.2.2]-octane (etpb) (Strem Chemicals, Inc.) was recrystallized fromhexane. Decalin (J.T. Baker, analysed reagent), xylene(B.D.H. AnalaR), and cyclohexane (Fisher Scientific Co.,,4 .C. S. certified) were dried over molecular sieves. Argon(99.998%) and carbon monoxide were ' Linde ' SpecialtyGases from Union Carbide of Canada, Ltd.Octacarbonylbis(tripheny1phosphine)dirhenium was pre-pared in low yield by the published method,6 but improvedyields were obtained when the reaction time was extendedand a greater excess of ligand was used. Decacarbonyl-dirhenium (1.80 g, 2.76 mmol) and triphenylphosphine (2.16g, 8.25 mmol) in xylene (80 cm3) were heated under refluxfor 20 h. The pale yellow solution was cooled, and colour-less crystals separated overnight. These were filtered off,washed with hexane (7 x 15 cm3), and dried in vacuo, yield2.30g (75%) (Found: C, 47.6; H, 2.55.[{Re(CO),(PPh,)),]requires C, 47.1; H, 2.70%). 1.r. spectrum in decalin:1 980w(sh) and 1960 cm-l (B 32 500 dm3 mol-l cm-l) cf.spectrum in cyclohexane,s 1 988w and 1 965vs cm-l. U.v.-visible spectrum in hexane: 340 (4) and 300 (10) nm.(Numbers in parentheses are relative absorption coefficients.)Nonacarbonyl(tripheny1phosphine)dirhenium was pre-* 1 mmHg w 13.6 x 9.8 Pa, 1 G = 10-4T, 1 atm = 101 325Pa.Part X, J. P. Fawcett, A. J. Poe, and K. R. Sharma, J .Amer. Chem. SOC., in the press.(a) D. Hopgood and A. J. Po&, Chem. Comm., 1966, 831;( b ) H. Wawersik and F. Basolo, Inorg. Chim. Acta, 1969, 3, 113;(c) L. I. R. Haines, D. Hopgood, and A. J. Poe, J .Chem. Soc. ( A ) ,1968, 421; ( d ) L. I. B. Haines and A. J. Poe, ibid., 1969, 2826;(e) D. DeWit, J. P. Fawcett, A. J. Poe, and M. V. Twigg, Co-ordination Chem. Rev., 1972, 8, 81; (f) J. P. Fawcett, A. J. Poe,and M. V. Twigg, J . Organometallic Chem., 1973,51, C17; ( g ) J. P.Fawcett, A. J. Po&, and M. V. Twigg, J.C.S. Chem. Comm., 1973,267.pared by reaction of [{ Re(CO),(PPh,) j2] with carbon mon-oxide, A stream of carbon monoxide was passed through asolution of [{Re(CO),(PPh,)),] (0.60 g, 0.54 mmol) in xylene(150 cm3) a t 130 "C for ca. 30 h, after which the productmixture was shown by i.r. spectroscopy to have reachedequilibrium. Solvent was removed under reduced pressureand the complex [Re,(CO),(PPh,)] was separated fromunchanged bis(phosphine) complex and traces of decacar-bony1 by chromatography on a foil-wrapped column ofsilica gel (100-200 mesh) made up in hexane.The deca-carbonyl was eluted with hexane, the monophosphine com-plex with hexane-toluene (2 : 1 v/v), and unchanged bis-(phosphine) complex with toluene. Colourless crystalswere isolated from the hexane-toluene eluate, washed with alittle hexane, and dried in vacuo, yield 0.18 g (38%) (Found :C, 36.7; €3, 1.85. [Re,(CO),(PPh,)] requires C, 36.6; H,1.70%). 1.r. spectrum in decalin: 2 106 (2 SOO), 2 034(1 800), 1 998 (27 500), 1973w(sh), 1 967 (5 400), and 1 941cm-1 (E 5 900 dm3 mol-1 cm-1). U.v.-visible spectrum incyclohexane: 320 (8) and 288 (10) nm.Kinetic runs were carried out in foil-wrapped reactiontubes, as previously des~ribed,~*~ and monitored by i.r.spectroscopy with a Perkin-Elmer 257 spectrophotometer.Reaction of [{ Re(CO),(PPh,) ),I with Carbon Monoxide.-Because of the relative insolubility of the complex in decalinat room temperature, suspensions in deoxygenated decalinwere heated at 100 "C in vacuo for ca.5 min, when clearsolutions were obtained. The space above the solutions inthe reaction tubes was then filled with carbon monoxide, orcarbon monoxide-nitrogen mixtures, before commencingthe kinetic runs. Judging from i.r. spectra of the solutions,this procedure had no discernible effect on the complexwhich is quite stable in solution at 100 "C. The disappear-ance of the complex was monitored by observing thedecrease in absorbance of its C-0 stretching band a t 1 960cm-l. Bands due to [Re,(CO),(PPh,)] grew cleanly through-out the reaction. In the absence of added triphenylphos-phine, the reaction went to completion, the band at 1960cm-l disappearing into the side of the 1 967 cm-l band of theproduct.First-order rate plots were linear for at leastthree half-lives. The rateconstants were independent of the mole fraction of carbonmonoxide down to ca. 0.1. Activation parameters wereobtained by an unweighted least-squares treatment of thedependence of log (K,b,./T) on 1/T, only data for yco > 0.1being used (yco is the mole fraction of CO in CO-N, mix-tures). The quoted uncertainties are standard deviationscorrected for the number of degrees of freedom.@Reaction in the presence of added PPh, was slower andproduced equilibrium mixtures of the mono- and bis-(phos-phine) complexes. Equilibrium constants for replacementof PPh, by carbon monoxide were calculated from theintensities of the bands at equilibrium, the known intensitiesof the bands of the pure complexes, and the known solu-bility of carbon monoxide from 20 to 75 "C.Correctionsfor the difference in temperature were made by taking intoaccount only the changing vapour pressure of decalin, theJ. P. Fawcett, Ph.D. Thesis, London University, 1973.J. P. Fawcett, R. A. Jackson, and A. J. Poe, J.C.S. Chem.M. Freni, D. Giusto, and P. Romiti, J . Inorg. Nuclear Chem.,P. W. Jolly and F. G. A. Stone, J . Chem. SOC., 1965, 5259.M. Basato and A.J. Po&, J.C.S. DaZton, 1974, 456.a M. Basato, J. P. Fawcett, and A. J. Poe, J.C.S. DaltonA. J. Poe and M. V. Twigg, J.C.S. Dalton, 1974, 1860.Rate data are shown in Table 1.Comm., 1975, 733.1967, 29, 761.1974, 1350530 J.C.S. DaltonTABLE 1Kinetic data for the reaction of [{Re(CO),(PPh,)},] withcarbon monoxide in decalin. [Re,(CO),(PPh,) Jo =3.4 x mol dm-,1 03[PPh3] /00l"C yco * rnol dm-3 lo5 koh./s-l150 1.0 0 4461 .o 0 3931.0 0 430140 1.0 0 1251.0 0 1161.0 0 127130 1.0 0 39.41.0 0 38.41.0 0 42.5120 1 .o 0 11.51.0 0 11.21.0 0 11.61.0 0 11.5140 0.74 0 1260.50 0 1220.25 0 1200.13 0 1220.063 0 1040.024 0 76.51 .o 2.06 1131.0 5.22 91.61.0 11.6 72.31.0 21.9 55.61.0 34.8 42.61.0 45.8 50.01 .o 47.7 45.11 .o 62.5 37.01.0 0.99 9.851.0 2.18 9.751.0 6.26 9.631 .o 10.1 9.741 .o 10.1 8.071.0 20.3 6.311.0 41.4 5.64AH: = 38.8 4 0.5 kcal mol-l (162.2 f 2.1 kJ mol-I),AS: = 21.5 f 1.2 cal K-l mol-l (89.9 6.3 J K-l mol-l),and o(kobs.) = f6.5%.* Mole fraction of carbon monoxide in CO-N, mixtures.heat of solution having been shown to be very sma11.3**Thus the concentrations of carbon monoxide were estimatedto be 4 x lo-, mol dm-3 at 140 "C and 4.6 x lo-, mol dm-3at 120 "C.The data for 103[PPh3] = 21.9, 34.8, 45.8, and47.7 mol dm-, at 140 "C led to an average value K = 23,with a mean deviation of f3. Those for 103[PPh,] = 20.3and 41.4 mol dm-3 a t 120 "C gave K = 20 & 3.*Reaction of [{ Re(CO),(PPh,) ),I with TriphenyZ Phosphile.-This reaction proceeds smoothly with eventual loss of allthe [{Re(CO),(PPh,)),], and the growth of i.r.bands at2 072vw, 1 996w(sh), 1 976vs, and 1 925w cm-l. The rateswere measured by observing the decreasing absorbance ofthe reactant complex a t 1960 cm-1, and first-order rateplots were linear for 2-3 half-lives. Rate data are givenin Table 2. Two runs were also made using etpb whenbands appeared a t 2 080w, 2 022w, 1981vs, and 1936mcm-l. The initial product of the reaction with 1.1 x lo-,mol dm-3 triphenyl phosphite was allowed to react further at140 "C when a strong band appeared a t 1 982 cm-l, the first-order rate constant for the appearance of this product beingca. 6 x s-l. It appears, therefore, that replacement ofthe triphenylphosphine ligands occurs in two stages by wayof the mixed phosphinephosphite complex.A few runs were made with added PPh, when an equilib-* K is the equilibrium constant of the reaction [(Re(CO),-lo F.Nyman, Chem. and I n d . , 1965, 604.(pph3))zl i- co [Rez(CO)s~(PPha)l -k pph3rium mixture of complexes was formed. The spectra of thetwo complexes were not as well separated as those for thereaction with carbon monoxide, and equilibrium constantscould not be calculated from the spectra of the productmixtures.TABLE 2Kinetic data for the reaction of [(Re(CO),(PPh,) ),I with[Re,(CO),- triphenyl phosphite in decalin a t 140 "C.(PPh3),I0 = 3.4 x lo-, rnol dm-3103[P(OPh),]/mol dm-, 103[PPh3]/mol dm-3 lo5 kobs.ls-l3.38 0 11510.4 0 11919.3 0 12331.4 0 12643.2 0 12693.5 0 12310.4 1.10 10310.4 11.9 47.910.4 26.9 29.810.4 44.7 20.646.5 * 0 125135 * 0 127* Reaction with etpb.Reaction of [{ Re(CO),(PPh3) ),I with TriphenyZphosphine.-This reaction occurred at conveniently measurable rates indecalin a t 160-180°C.1.r. bands grew and decayedduring the reaction and a careful analysis of the changes ofintensity with time enabled the bands to be grouped accord-ing to common behaviour within each group. The clearestbehaviour was shown by three products, the major one, (I),having a single band a t 1 931 cm-1. This can be attributedto the bis(axia1) form of the trigonal-bipyramidal complex[Re(CO),(PPh,) ,] which has been isolated and partlycharacterized [v(CO) at 1930 cm-l in ben~ene1.l~ Twoother products, (11) and (111), showed initial growth untilca.65% of the reactant complex was lost, after which theirbands decreased in intensity. The band wavenumbers are2 009s, 1 924s, and 1908s cm-1, and 2 074w, 1984m, and1970s cm-l, respectively, the main difference between thebehaviour of the two products being the relatively greaterintensity of the bands due to (11) as [PPh,] is increased.Product (11) has bands very close to those of the axial-equatorial trigonal-bipyramidal complexes [Re(CO) PMe-Ph,),] and [Re(CO),(PMe,Ph),] (2 005m, 1 928s, and 1 912scm-l, and 2 OOOm, 1 925s, and 1 910s cm-l, respectively).llProduct (111) has a similar pattern of bands to ax-[Fe(CO),-(PPh,)] (2 052w, 1980w, and 1947s cm-1)12 and this,coupled with the greater amount formed a t lower phosphineconcentrations, enables (111) to be formulated tentatively asax-[Re(CO),(PPh,)].Depending on the conditions, twoother products were sometimes formed in smaller and lessreproducible amounts [(IV), 1 960m and 1 860s cm-l; (V),2 038w and 1 990s cm-l] and these may be additional isomersof mononuclear phosphine complexes.The e.s.r. spectrum of a decalin solution containing almostpure product (I) was measured for us by Dr. A. Bassindalewith a Varian EPR 4 spectrometer. Scans were carried outover 1 000-4 000 G ranges, centred on 3 313 G, with amodulation amplitude of 16 G. A rather broad signal wasobtained and fine structure due to hyperfine splitting was notvery clearly defined due to an unfavourable signal : noiseratio.However, five of the six bands expected for an odd11 E. Singleton, J. T. Moelwyn-Hughes, and A. W. B. Garner,J . Organometallic Chem., 1970, 21, 449; J. T. Moelwyn-Hughes,A. W. B. Garner, and N. Gordon, ibid., 1971, 26, 373.12 D. J . Darensbourg, Inorg. Chim. Ada, 1970, 4, 6971976 531electron coupled to a nucleus of spin # were discernible andthis supports the conclusion that mononuclear rhenium(0)species are present.The rate of loss of [{Re(CO),(PPh,)},} (followed by moni-toring the band a t 1 960 cm-1) seemed to be relatively rapidover the first 10% of reaction. The reaction subsequentlyshowed first-order rate plots, linear for 1-2 half-lives, afterwhich the gradients increased.Half-order rate plots of( A , - Am)* against time were linear for more than 2 half-lives, apart from the initial, somewhat faster, 10% ofreaction. The dependence of the subsequent rates on theconcentration of complex was determined, the rates beingmeasured by multiplying the concentrations by apparentfirst-order rate constants obtained from plots of log(At -Am) against time. The results are shown in Figure 1 where5.5 1FIGURE 1 Concentration dependence of the initial rates ofreaction of [(Re(CO),(PPh,)},] with triphenylphosphine.Gradient of line = 0.5a straight line of gradient 0.5 passes very close to theexperimental points. Concentrations were corrected forexpansion of the de~a1in.l~ Values of the half-order rateconstants are in Table 3, together with the activation para-meters calculated from data a t [Re,(CO),(PPh,),], = 3 x10-4 mol dm'-3.Other Reactions of [{Re(CO),(PPh,) Id.-Reaction withoxygen was complete within 5 min a t 140 "C.Measurablerates were obtained at 70-90 OC, but the results were veryerratic although the growth of bands a t 2025m, 2 OlOs,1 958m, and 1 900w(br) cm-1 was always observed. Reac-tion at 140 "C was completely inhibited by 0.2 mol dm-,PPh,. Reaction a t 140 "C with ca. 1 : 1 mixtures of oxygenand carbon monoxide followed essentially the same path asthe reaction under pure carbon monoxide, [Re,(CO),(PPh,)]being formed in 75% yield before decomposing.Simple thermal decomposition in the absence of addedreagents was rapid, at 180 "C, until ca.80% completion ofreaction. The complex [Re,(CO),(PPh,)] was formed asthe major product together with some (I), (11), and (111),but this mixture then underwent slow decomposition, theintensities of all the i.r. bands decreasing and no new onesappearing. The presence of 2.7 x mol dm-3 PPh,caused the fast reaction to proceed to only 20% completion,after which slow reaction occurred to form (I) and some (IV).At 140 "C, in the absence of added triphenylphosphine,[Re,(CO),(PPh,)] was formed in 50% yield and this reactionhas a first-order rate constant of ca. 3 x s-l, which isca. 25% of the rate with carbon monoxide. A t 140 "C thecomplex reacted smoothly under 1 atm of nitrogen monoxide,[Re,(CO),(PPh,)] being formed in ca.25% yield by 65% ofreaction, together with some (11), after which general de-composition occurred slowly. A good first-order rate plotwas obtained for loss of [{Re(CO),(PPh,)},], the value (1 xlo-, s-l) being close to that for reaction with carbon mon-oxide.TABLE 3Kinetic data for reaction of [{ Re(CO),(PPh,)),] withtriphenylphosphine in decalin at 180 "C105[Re,(CO),-(PPh,)z]o/mol dm-, 103[PPhs]/mol dm-, 106 k,,b,./mol* dm-4 s-l0.796 22.4 2.963.58 a 22.4 2.889.31 22.4 3.3425.7 b 22.4 2.9965.1 22.4 3.4588.0 e 22.4 3.53148.0 22.4 3.7430.0 21.6 2.9930.0 78.5 2.8430.0 134 2.6330.0 b 279 2.8930.0 b*e 20.8 1.3230.0 b*f 21.2 0.56AH: = 30.6 f 1.4 kcal mol-1 (128.0 -& 5.9 kJ mol-l),AS! = -17.3 f 3.2 cal K-l mol-1 (-72.4 & 13.4 J K-l mol-l),and G(kob8.) = 3 ~ 7 % .2.0 mm cell pathlength.1.0 mm pathlength. 0.20mm pathlength. d 0.11 mm pathlength. 6 At 170 "C. f At160 "C.Finally, reaction with iodine was too rapid to followconventionally at room temperature, the products appearingto be cis- and trans-[Re(CO),I(PPh,)] with i.r. bands at2 103n1, 2 023m, 2 006s, and 1958m ern-', and 1 998s(sh)and 1 990s cm-l, respectively, the assignments being madeby comparison with the spectra of the bromo-analogues.6At 90 "C the trans isomer reacted smoothly to form the cisisomer, together with some other unidentified products,with a half-life of ca. 30 min.Reaction of [Re,(CO),(PPh,)] with TriphenyZphosphine.-This reaction was followed by observing the decreasingabsorbance of the i.r.band a t 1 995 cm-1 due to the nona-carbonyl. Reactions went to completion to form thecomplex [{ Re(CO),(PPh,) },I. First-order rate plots werelinear for ca. 3 half-lives. The data are shown in Table 4,together with the activation parameters, the high precisionof the data being concordant with the excellent linearity ofthe rate plots.One reaction of this complex was followed in decalin underan atmosphere of oxygen at 160 "C. A good linear first-order rate plot was obtained by observing the loss of react-ant complex, the rate constant being only ca. 20% slowerthan that for reaction with PPh,. Product i.r. bands wereobserved to grow at 2 068w, 2 012s, and 1 972m cm-l untilthe reaction was ca. 80% complete when they began todecrease in intensity.Reaction of [Rez(C0),(PPh,)] with Carbon Monoxide.-This reaction was followed by measuring the decreasingl3 W.F. Seyer and C. H. Davenport, J. Amer. Chem. SOC.,1941, 68, 2425532 J.C.S. Daltonabsorbance of the band at 1 9 4 1 cm-1, due to the nona-carbonyl, and first-order rate plots were linear for 1-2 half-lives. The yield of product decacarbonyl decreased from70 to 55% as the temperature increased, but quite goodrate plots were also obtained over ca. 1 half-life when thetheoretical value for A , was used ( E 7 070 dm3 mol-1 cm-1 a t2 070 cm-'),ld and the rate constants were in good agreementwith those from loss of reactant complex. The data areshown in Table 5 .TABLE 4Kinetic data for the reaction of [Re,(CO),(PPh,)] with PPh,in decalin.[Re,(CO),(PPh,)], z 4 x 10-4 rnol dm-,[PPh,] = 1.99 x lo-, mol dm-, (except where indi-cated)lo5 kobs./S-lec1 "C r * \130140 15.7, 15.6, 16.6150 44.0, 44.3, 46.0160 118, 114, 1165.29, 5.28, 5.44, 5.44, 5.33," 5.53bAH! = 34.8 f 0.2 kcalmol-1 (145.5 f 1.3 kJ mol-I), AS! =7.7 4 0.3 cal K-l mol-l (32.2 f 2.9 J K-l mol-l), and o(kobs.) =& 3.0%.a [PPh,] = 4.96 x 10-3mol dm-3. [PPh,] = 1.02 x 10-2mol dm-3.TABLE 5Kinetic data for the reaction of [Re,(CO),(PPh,)] withcarbon monoxide in decalin. [Re,(CO),(PPh,)], M4 x mol dm-,, yco = 1lo5 kob#./S-lfJC/"C a b150 2.00 1.88160 5.94 5.47170 14.5 15.9170 15.4 14.9170 13.9180 38.1 41.5180 42.0 40.4AH! = 37.6 f 0.6 kcal mol-1 (157.3 & 2.5 kJ mol-I),AS* = 8.0 & 1.3 cal K-' mol-1 (33.5 f 5.4 J K-l mole'), andFrom growth of [Re,-D(kobs.) = f6.6%.(C0)IJ.' yco = 0.27.a From loss of [Re,(CO),(PPh,)].DISCUSSIONThe general behaviour of the reversible interchangereactions of the complex [(Re(CO),(PPh,)},] is entirelyconsistent with the dissociative mechanism shown inequations (1) and (2) [L = CO or P(0Ph)J which leadsto rate equation (3). At [PPh,] = 0 the reaction goes toklk- 1[@e(CO),(PPh3))21 * [Re,(CO),(PPh,)l + PPh3 (1)k,k-,[Re,(CO)t@%)l -I- L H CRe2(CO),(PPh3)LI (2)kobs. = ( ( ~ ~ ~ ~ [ L I / ~ - I . [ P P ~ ~ I ) + k-2)/(1 + (k,[LI /k-,[PPh,I) 1 (3)completion, k-, can be neglected, and kobs. = k,. At aconstant value of [L], kobs.should decrease with increas-ing values of [PPh,]. Rearrangement of equation (3)leads to (4) so that plots of kobs. against ( k , - kobs.)/kobs. = k-2 (k2[L]/k-,) ((k1 - kobs.)/[PPh3]) (4)[PPhJ should be linear with gradient k,[L]/k-, and anintercept of k-,. The data for [L] = [CO] = 4.0 x10-3 mol dm-3 (yco = l), and for [L] = [P(OPh)J =1.04 x mol dm-3, at 144 "C are plotted in this wayin Figure 2, reasonably good straight lines being obtained.The value of k, was taken to be (12.3 & 0.1) x 10"' s-lwhich was the mean of 12 determinations of the limitingrate constants with L = CO or P(OPh),, the standarddeviation of an individual measurement being & 3%.The intercept of 2.0 x lo-, s-l for L = CO is in quitegood agreement with the independently measured valueof 1.6 x s-l for k-, from Table 4.No directlymeasured value of k-, is available for L = P(OPh),.'Y I 9 .5 10 15FIGURE 2 The influence of free triphenylphosphine on thekinetics of the reaction of [(Re(CO),(PPh,)}k) with carbonmonoxide (a) and with triphenyl phosphite (a), in decalin at140 "CThe kinetic data for L = CO at 140 "C lead to a valueof k,/k-, (i.e. k ~ ~ / k p p h , ) of 3.0 and K = 20, in excellentagreement with the value of 23 & 3 obtained from theequilibrium mixtures. The values of kobs. calculated byusing k, = 12.3 x lo4 s-l, k-, = 2.0 x lo4 s-l, and k,lk , = 3.0 have a mean deviation of &6% from the ex-perimental values. The data for [L] = [CO] = 4.6 x10-3 mol dm-3 at 120 "C are not as extensive or as welldistributed as those at 140 "C.However, by takingk, = 11.5 x s-l (from the data when [PPh,] = 0)and k-, = 1.7 x s-l (obtained by extrapolation to120 "C of the data in Table 4), we obtain k,/k-, = 4 and6, and K = 26 and 40, from the values of kobs. at 103-[PPh,] = 20.3 and 41.4 mol dm-3, respectively. Thesevalues are in moderate agreement with the value K =20 & 3 obtained from the equilibrium mixtures. Thevalues of kobs. calculated for the various values of [PPhJhave a mean deviation of &$yo from the experimentalvalues when k,/k-, was taken as 5. The data in Table 1976 533and Figure 2 for reaction with triphenyl phosphite leadto values of kp(Oph),/kpph, = 0.6 and K = 7 at 140 "C.The data for these reactions are, therefore, all quitewell represented by rate equation (5), this equationbeing characteristic of a reversible dissociative mechan-ism.The values kCO/kpph, = 3, and kCO/kp(OPh), = 5,are quite compatible with similar competition parametersfor other co-ordinatively unsaturated intermediate~.~J~It can be shown," however, that at least two othermechanisms for ligand interchange also lead to rateequation (5). One of these involves reversible homolyticfission as the first step, followed by a very rapidly attainedequilibrium in which the PPh, in [Re(CO),(PPh,)] isreplaced by L, the final stage being the combination of[Re(CO),L] and [Re(C0)4(PPh,)]. The other mechanisminvolves a reversible metal-migration process followedby associative stepwise displacement of PPh, by L, aswas proposed for reactions of the decacarbonyls ofmanganese and rhenium.2@ This type of intermediatehas also been proposed for some other reactions,16although direct kinetic evidence for such a mechanism islacking.A kinetic distinction between these variousmechanisms would only be possible if kobs. were to de-crease below the limiting value Kobe.(lim) as [L] isdecreased, even in the absence of free PPh,. Althoughthere is some evidence that kobs. does decrease in thisway (Tables 1 and 2), the deviation is not largeenough to make promising a detailed attempt to dis-tinguish these alternative mechanisms in this particulars ys tern.However, the reaction of [{Re(CO),(PPh,)}J with freePPh, clearly leads to a variety of mononuclear radicalspecies such as [Re(CO),(PPh,)d in various isomericforms, the most stable of which appears to be the bis-(axial) complex.The half-order kinetics (with respectto [Re2(CO),(PPh,),]} strongly suggest a reversible fissionprocess of the type shown in equations (6) and (7). Thek sk-8[{Re(CO)4(PPh3)),1 * 2[Re(C0)4(PPh,)] (6)[Re(CO),(PPh,)l -k PPh, [Re(CO)3(PPh3)d (7)faster rate over the first 10% of reaction would then bedue to the setting up of the steady-state concentration of[Re(CO),(PPh,)]. Since the rates are independent of[PPhJ, reaction (7) cannot be a straightforward SN2* The derivation of rate equations for a variety of mechanismsfor substitution and oxidation reactions of metal-metal bondedcarbonyls will be described elsewhere.l4 D.J. Darensbourg and H. L. Condor, Inorg. Chem., 1974.13, 374; C . L. Hyde and D. J. Darensbourg, ibid., 1973,12, 1286.process, and the observed half-order rate constants aregiven by 0.5(K,/K,)*k,.1f This reaction is very muchslower than that with carbon monoxide (ca. 500 timesslower a t 160 "C and [Re,(CO),(PPh,),] = 3 x lo-*mol dm-3), but these experiments do not, unfortunately,lead to independent values for k6 for comparison withthose for k,.Reaction of oxygen with [Mn,(CO)lo] and [MnRe(CO),,]has provided convincing kinetic evidence that thesecomplexes undergo reversible thermal homolytic fission,and there are several reactions for which the very closesimilarity of the kinetic parameters for substitution andreaction with oxygen provides indirect evidence in sup-port of the homolytic fission mechanism for substitu-tion.l6 Direct evidence is also available for one sub-stitution reaction.4 This is not, unfortunately, the casehere since the reaction with oxygen is very much fasterthan the substitutions, although it is not kineticallyclean.The virtually complete suppression of thisreaction by CO or PPh, does imply that the oxidationand substitution reactions do not proceed by completelyindependent paths. The detailed significance of this isuncertain as is that of the thermal reactions under argonand with nitrogen monoxide. Although reaction ofNO with [{Mn(CO),(PPh,))J leads l7 to formation ofequimolar amounts of [Mn(CO),(NO) (PPh,)] and [Mn-(CO),(NO)], the absence of analogous products of reactionof the rhenium complex probably reflects the fact that nosimple carbonylnitrosyl complexes of rhenium have yetbeen characterized. It is interesting that the reactionsof both the bis- and rnono-(phosphine) complexes withoxygen lead to spectroscopically definable productsalthough the nature of these is quite unknown.Finally, the reactions of [Re,(CO),(PPh,)] with PPh,or CO cannot both be going by simple rate-determininghomolytic fission since the rate parameters should thenbe identical and they are clearly not (Tables 4 and 5 ) .The fact that reactions with oxygen and with PPh,proceed at about the same rate may signify that it isthese reactions that proceed by rate-determining homo-lytic fission, in which case, by microreversibility, so mustthe reaction of [{Re(CO),(PPh,)).J with CO. This issupported by the strong evidence for homolytic fission inthe reaction of [{Re(CO),(PPh,)),] with PPh,.We thank Erindale College and the National ResearchCouncil of Canada for support.[5/1206 Received, 30th Jaiite, 19751l5 P. F. Barrett and A. J. Poe, J . C h i n . SOC. ( A ) , 1968, 429;M. Basato and A. J. Poe, J.C.S. Dalton, 1974, 607.l6 A. J. Poe, D. M. Chowdhury, D. DeWit, J. P. Fawcett, andM. V. Twigg, Proc. 14th Internat. Conf. Co-ordination Chem.,Toronto, 1972, p. 120; R. A. Jackson and A. J. Poe, Proc. 16thInternat. Conf. Co-ordination Chsm., Dublin, 1974, pa er 3.20.l7 H. Wawersik and F. Basolo, Inorg. Cheun., 1967, 0, 1066