摘要:

J. CHEM. SOC. DALTON TRANS. 1988 1501The Photochemistry of Cyclopentadienyl Platinum Carbonyl Dimers:Characterization of [P~,(~-CO)(Y~~-C,R,),] (R = H or Me) using Matrix Isolationand Fast Time-resolved Infrared SpectroscopyAndrew J. Dixon, Stephen Firth, Anthony Haynes, Martyn Poliakoff," and James J. Turner*Department of Chemistry, University of Nottingham, Nottingham NG7 2RDNeil M. BoagDepartment of Chemistry and Applied Chemistry, University of Safford, Salford M5 4 WTU.V. photolysis of [{Pt(CO) (q5-C5H5)}2], (1 a), or [{Pt(CO) (qs-C5Me5)},], (1 b), isolated in Ar, N,or CO matrices at 20 K causes ejection of a CO group and formation of [Pt,(p-CO) (q5-C5H5),],(2a), and [ Pt,(p-CO) (q5-C,Me,),], (2b), respectively. Isotopic enrichment with 13C0 shows thatthese photoproducts contain only one bridging CO group.Photolysis using plane-polarised lightdemonstrates that this CO group occupies a position bridging the Pt-Pt bond symmetrically. Both(2a) and (2b) have strong absorptions in the visible region. Irradiation with light corresponding tothese absorptions causes recombination with the photoejected CO t o regenerate (la) or (I b). Theunsaturated species, (2a) and (2b), have also been identified at room temperature in solution,using time-resolved i.r. spectroscopy. They are found to have lifetimes in excess of 2 ms underthese conditions, even in the presence of CO or CH3CN. Prolonged U.V. photolysis also causeshomolysis of the Pt-Pt bond of (1 a) or (1 b) in CO matrices, the resultant photoproduct being[Pt(CO),I~Recently the photochemical reactions of dinuclear transition-metal cyclopentadienyl carbonyl compounds, both in solutionand low-temperature matrices, have been the subject ofextensive study.'*2 The dimers [{ M(C0),(q5-C,H,)},] (M =Cr, Mo, or W) and [{M(CO),(q5-C,H,)},] (M = Fe or Ru),have received particular attention.It has been found that two primary photochemical processesoccur in these complexes; (i) dissociative loss of CO to give a co-ordinatively unsaturated dinuclear complex, e.g.equation (l),and (ii) homolytic cleavage of the metal-metal bond to givemononuclear metal-centred radicals, e.g. equation (2). In low-C{Fe(Co),(q5-C,Hs>}21 - 2CFe(Co)2(175-C5H5>I' (2)temperature matrices, however, ejection of a CO ligand isusually the sole observable process during the U.V.photolysis ofthese complexe~.~*~ Homolysis of the M-M bond is notobserved, because the restrictions of the matrix cage lead to theinstantaneous recombination of the bulky radicals. For both theiron and molybdenum complexes, flash photolysis 5 * 6 hasdemonstrated that the same dinuclear photoproducts areformed in room-temperature solution as in low-temperaturematrices. Time-resolved i.r. spectroscopy ' (a combination ofU.V. flash photolysis and fast4.r. detection) has recently proveda valuable tool in characterising these photoproducts insolution, since i.r. bands in the C-0 stretching region providemore structural information than the broad featurelesselectronic absorption bands of these compounds.The electronic structure of these dinuclear species isparticularly intriguing. Sometimes multiple M-M bonds areformed, as in the case of the Group 6 metals, where photolysis inhydrocarbon solvents leads to the formation of the triplybonded dimers [{M(CO),(q5-C,H,))2] (M = Cr, Mo, or W)with very short M-M bond lengths.8 The Fe dimer, [Fe,(p-CO)3(q-CsMes)2], produced by photolysis of [{ Fe(CO),(q-C,Me,)) ,I, is paramagnetic with a triplet ground state.'.''Although the cyclopentadienyl complexes of nickel andplatinum, [{ M(CO)(q '-C,H,)),], were originally prepared byFischer et a!.' ' twenty five years ago, their photochemistry hasbeen relatively neglected.The two compounds have markedlydifferent structures. In the Ni complex, both CO groups bridgethe metal-metal bond.', By contrast [{Pt(CO)(q5-C5H,)}2],(la), and the recently synthesised complex [{ Pt(C0)(q5-C,Me,)),], (lb), have two terminal CO group^,^^''^ with therelative orientation of the two Pt-CO bonds reminiscent of thetwo 0-H bonds in H,O,.0T5C\\ J t - C5"5(la)Preliminary investigations haveshown that photolysis of [{Ni-(p-CO)(q5-C,H,)),] in poly vinyl chloride matrices at 77 Kcauses dissociative loss of CO to form [Ni,(p-CO)(q5-C,H,),].2 The analogous dinuclear Pt species, [Pt,(p-CO)(q ,-C5H5),], (2a), with a single CO bridge has been proposed as anintermediate in the thermal reaction of (la) with acetylenes,such as PhCCPh.' Matrix-isolation spectroscopy offers thepossibility of characterising such a species.In this paper we report the results of our investigation into thephotochemistry of [{Pt(CO)(q5-C,H,)),], (la) and[(Pt(CO)(q5-C,Me,))2], (lb).Our experiments, in low-temperature matrices and in solution at room temperature(using time-resolved i.r. spectroscopy) have yielded detailedinformation about the structure of the intermediates, (2a) and(2b), formed on photolysis of these dimers.ExperimentalThe matrix isolation apparatus, Air Products Displex CS202,and photolysis equipment, Philips HPK 125-W mediu1502 J. CHEM. SOC. DALTON TRANS. 19882 rh/nmFigure 1. The electronic absorption spectrum of [{ Pt(CO)(q 5-C5H5)}2], (la) mol dm-3, I-cm pathlength), in cyclohexane, atroom temperaturepressure Hg arc, have been described previously.' 5,16 Filtersused for photolysis were a NiSO, (400 g dm-3)-CoS04 (200g dm-3) aqueous solution (4-cm pathlength, band-pass, 23&345 nm) and a 290 nm (& 10 nm) Balzers interference filter.Alli.r. spectra of matrix-isolated species were obtained using aNicolet MX-3600 FT-IR interferometer and model 1280 datasystem (32 K data collect, 256 K Fourier transform, i.e. 0.7 cm-'resolution). U.v.-visible spectra were recorded on a Perkin-Elmer Lambda 5 spectrophotometer with a model 3600 datastation.Matrices were prepared by 'slow spray-on' techniques,' withthe sample of [{Pt(CO)(q'-CSRs)},] (R = H or Me) beingheated to increase the rate of vaporisation. The concentration ofsample in the matrix was varied both by changing the tempera-ture used for vaporisation of [{Pt(CO)(q'-C,R,)),] and byaltering the rate of deposition of the matrix gas.Temperaturesused for sample vaporisation were typically 50 "C for [{ Pt(C0)-(q5-C5H5)),] (la) and 65 "C for [{Pt(CO)(q'-C,Me,)},] (lb).The results reported here were surprisingly insensitive to theinitial concentration of [{ Pt(C0)(q5-C,R,)},] in the matrix,even when the concentration was quite high, ca. 1:500. Thetemperature of the CsI matrix window was typically 17-20 K.The principles and practice of polarised photochemistry havebeen described previously.' 6*1 7 3 1 ' The experiments involvephotolysing the matrix-isolated sample with plane-polarisedlight. 1.r. and u.v.-visible spectra are recorded through apolariser with its plane of polarisation either parallel orperpendicular to the plane of the photolysing light.Comparisonof the two spectra reveals whether absorption bands of thematrix-isolated species show any linear dichroism. The detailsof recording polarised i.r. spectra with the Nicolet MX-3600have been given elsewhere. ' 6,1'The time-resolved infrared spectroscopy apparatus '3 ' uses apulsed-u.v. excimer laser (XeCl, 308 nm) as the U.V. photolysissource and a continuous wave CO i.r. laser, tunable in steps of 4cm-' between 1 700 and 2 000 cm-', to monitor the absorptionat a particular i.r. frequency. 1.r. spectra are built up 'point bypoint' by repeating the U.V. flash photolysis with the i.r.monitoring laser tuned to a different wavenumber for each U.V.shot.As used here the time resolution of the apparatus was ca.[{ Pt(CO)(rl '-C5H5))21 (la), [{ Pt(CO)(*rl 5-C5Mes)} 2 1 (1b)2 ps.and I3CO-enriched [{Pt(CO)(q'-C,Me,)},] were prepared byt I Y x 3 ir1815-0.10- '2150 2030 2000V/ cm-'Figure 2. 1.r. absorption spectra of [{Pt(CO)(q5-C,H,)),]. (la). (a)After deposition in an argon matrix at 20 K; the band shown in brokenlines is due to matrix-isolated molecular CO and the band marked * isdue to an oligomeric species, formed by incomplete isolation of (la).(b) After 90 min filtered U.V. irradiation (23CL-345 nm); filled-in bandsare due to molecular CO and (2a) (see text). [The bands due tomolecular CO observed on deposition, spectrum (a), have a slightlydifferent wavenumber from that of the photoejected CO, spectrum (b).This is probably due to weak interactions with trace H 2 0 impuritiesduring deposition (H.Dubost and L. Abouaf-Marguin, Chem. Phys.Lett., 1972,17,269.)] ( c ) 1.r. difference spectrum showing the effects of90min photolysis with visible light (>400 nm), i.e. spectrum beforephotolysis minus spectrum after. Negative peaks are due to CO and(2a) destroyed by the photolysis while positive peaks are due tothe regeneration of (la). Note the absorbance scale expansion factors inthe different regions of the spectrastandard literature methods.' 3 . 1 4 Cyclohexane (BDH Aristargrade), matrix gases (Ar and N,, Messer-Griesheim; COY BOCresearch grade) were all used without further purification.Results and DiscussionU.V.Photolysis of [{Pt(CO)(q s-C,H,))2], (la).-The elect-ronic absorption spectrum of [{Pt(CO)(q5-CsH,)},], (la),shown in Figure 1, is more structured than those of manyrelated dinuclear transition-metal complexes. By analogy withprevious studies on similar the two bands of lowestenergy (arrowed) are assigned to the dn---*o* and o+o*transitions. It has also been suggested that the absorptions athigher energy are associated with transitions involvingM-(CO)n* charge transfer. Since the n* orbitals are M-J. CHEM. SOC. DALTON TRANS. 1988antibonding, 1 , 2 0 excitation of these charge-transfer transitionsmight be expected to lead to cleavage of an M-CO bond andformation of [Pt,(p-CO)(q ,-C,H 5)2].The i.r. spectrum of (la) dissolved in hydrocarbon solvents atroom temperature (not illustrated) is consistent with the X-raystructure, having two bands in the terminal v(C0) region andnone in the bridging v(C0) region.The i.r. spectrum of (la)isolated in an argon matrix at 20 K, Figure 2(a), is identical tothe solution spectrum, apart from a small shift in wavenumberand a slight matrix splitting of one band. In the matrixspectrum, there is an additional weak band near 2 150 cm-' dueto traces of unco-ordinated molecular CO in the matrix,probably originating from slight decomposition of [{ Pt(CO)(q 5 -C,H,)J,] under vacuum. U.V. irradiation (230-345 nm) of thematrix leads to a decrease in intensity of i.r. bands of (la) andthe growth of two new absorptions* [filled-in bands in Figure2(6)].The lower wavenumber absorption is near 1 815 cm-',where bridging carbonyl groups normally absorb. The othernew band, also filled in, is much less intense and is due touncomplexed CO in the matrix. This suggests that U.V.photolysis is causing dissociative loss of CO from (la) to forman unsaturated dinuclear intermediate, @a), with a bridging COgroup; see equation (3).During U.V. photolysis the matrix develops a striking pinkcolouration, easily visible to the naked eye after only a fewminutes irradiation. The electronic absorption spectrum of thematrix after U.V. photolysis reveals a strong visible absorptionwith peaks at 470 and 522 nm not present before. These u.v.-visible bands grow in intensity at a rate, which exactly parallelsthe growth of the i.r.band due to the bridging CO group of (Za),see Figure 3. Thus it is reasonable to assume that the u.v.-visiblebands are also associated with (Za).Subsequent irradiation of the matrix with visible light( > 400 nm), corresponding to excitation of these absorptionsof (2a) causes the destruction of (2a) and regeneration of thestarting material, (la). This effect can be clearly seen from thei.r. difference spectrum shown in Figure 2(c). There arenegative peaks, indicating a decrease in the amounts of both(2a) and molecular CO, and positive peaks showing the re-generation of (la). The visible absorption bands of (2a) (470and 522 nm) behave similarly to the bridging v(C0) i.r. band(not illustrated).Thus, irradiation with visible light promotesthe recombination of (2a) with CO, equation (4).[Pt,(p-C0)(q5-C5H5),] + CO 5Similar changes in the i.r. and u.v.-visible spectra areobserved upon U.V. photolysis of the pentamethyl complex [{Pt-(C0)(q5-C5Me,)>,], (lb), isolated in an argon matrix at 20 K.A photoproduct, (2b), is formed with a bridging v(C0) band at1 772.0 cm-' and a visible absorption at 540 nm, again giving thematrix a bright pink colour. Excitation of (2b) with visible lightagain reverses the photochemical reaction, reforming (1 b),* There is also a broad band, marked with an asterisk, which grows inthe terminal v(C0) region during U.V. photolysis. This must be due tosome oligomeric species, since it is much more prominent in experimentswhere the concentration of (la) was deliberately high.1-0ac g 0.50n a015030.1 gu)0-0450400 500 600 1805 1785h/nm v/ cm - 1Figure 3.Visible and i.r. absorption bands showing the progressiveformation of [Pt,(p-CO)(q5-C5H5)2] (2a) after 0, 40, 80, and 120 minfiltered U.V. photolysis of (la), isolated in a CO matrix (the arrowindicates increasing photolysis time). U.v.-visible and i.r. spectra wererecorded on the same sample. (Similar results were obtained for Arand N, matrices.) Note that the relative intensities of the i.r. and visiblebands remain constant, indicating that both absorptions belong to thesame species, (2a)albeit rather more slowly than in the case of the C,H,compound.These results are in agreement with the photo-chemistry shown in equation (5). Formation of the same species,(2a) and (Zb), was observed in N, and CO matrices onphotolysis of (la) and (lb) respectively. The i.r. and u.v.-visibledata for these experiments are summarised in Table 1.Identification of Intermediates-The results described in theprevious section suggest that the most likely structures for (2a)and (2b) are [Pt,(p-CO)(q5-C,H5),] and [Pt2(p-CO)(q5-C,Me,),]. Since these conclusions involve interpreting therelatively weak i.r. absorption due to photoejected CO, we nowstrengthen the case by using ' 3 C 0 isotopic enrichment to provethat (2b) contains only one CO group.The photolysis experiments were repeated with samples 50%enriched with 13C0.If the photoproduct contained a singlebridging CO group, one would expect two v(C0) bands ofalmost equal intensity, corresponding to the p-l2C0 and p-13C0 groups. If, however, the product contained two bridgingCO groups, the arguments of Darling and Ogden2' wouldpredict four v(C0) bands (3 strong, 1 weak) under theseconditions.Figure 4(a) shows the spectrum obtained after U.V. photolysisof (lb), with 50% 13C0 enrichment, isolated in an N, matrix at20 K. Two v(C0) bands of equal intensity are observed. Thesecan be assigned to the two isotopomers of a photoproduct withone bridging CO group. The band at higher frequency, due to[Pt,(p-'2CO)(q5-C5Me,),], has the same half-width, 8.5 cm-',as in experiments without isotopic enrichment, see Figure 4 (6).A doubly bridged structure, [{ Pt(p-CO)(q '-C5H5)}J, couldonly give rise to the spectra in Figure 4 if the interaction forceconstant between the two CO groups were zero.This seemsmost improbable in view of the non-zero interaction constantsfound in related molecules such as [(Ni(p-CO)(q ,-CsH,>} 21504 J. CHEM. SOC. DALTON TRANS. 1988~~~~ ~ ~Table 1. Wavenumbers (cm-') of i.r. bands and wavelengths (nm) of u.v.-visible bands of (la), (lb), and their photoproducts in different matrices at 20 KMatrix gasAr N2I f-1.r. U.v.-visible2 017.0* 2601999.5* 3251 992.1 2901 970.3 3601 806.1 4745251 770.0* 540co1.r. U.v.-visible2 01 5.2 *1996.91987.8 2951 966.4 3 641 996.9 * 4785281 759.6* 545and [(q5-C,H,)(CO)Fe(p-CO),Ni(q5-C,H,)I.12*22723 Theobserved and calculated wavenumber data for these isotopicexperiments are summarised in Table 2.We have now established that U.V.photolysis of (la) and (lb)generates monocarbonyl species (2a) and (2b) and haveeliminated the possibility that the photoproducts contain twoor more bridging CO groups, which give rise to only one v(C0)absorption due to some geometrical quirk of the structure, e.g. aplanar [Pt2(p-C0),] moiety. There are, however, a number ofpossible structures for (2a) and (2b), two of which (A) and (B)are illustrated. Structure (A) with a symmetrically bridging COgroup contains the previously unknown (formal) Pt=Pt doublebond, and could in principle be singlet or triplet; (B) possesses asemi-bridging CO group, potentially a four-electron donor. Wenow show how photolysis with plane-polarised light can elimin-ate structure (B).Table 2.Data for isotopic enrichment experiments; observed andcalculated wavenumbers (cm-') of the v(C0) bands for the isotopomersof (lb) and (2b) isolated in an N2 matrix at 20 KSpecies Observed Calculated(lb) CPt2(12C0)2(~5-CsMes)21 1970.3 1 970.4"1 992.1 1991.6"[Pt2(' 2CO)(1 3C0)(rl s-CsMe5)2] I 930.8 1 930.8",b1982.8 1 983.2".bCPt2(' 3C0),(15-C5Mes)21 1922.8 1 922.6"lb1943.2 1 943.4a,b(2b) CPt,(CI-I 2CO)(rlS-CsMes)*l 1 770.0 1 770.0'1 730.6'" Calculated with the use of the C-0 stretching force constant k =1 585.39 N m-I, and interaction constant ki = 17.03 N m-'. To achievethe best fit the reduced mass of the I3CO group was treated as a variablein these calculations.' Calculated with the use of the C-0 stretchingforce constant k = 1 265.6 N m-'.CPt2(CI-' 3co)(rl 5-CsMes)21 1 727.60IIC/ /Pt =Pt/ \ T5 - C5R5 T5 - C5R 5Pt - Pt/ \ q 5 - C5R T5- C5R5Photolysis using Plane Polarised Light.-This technique,briefly described in the Experimental section, has been valuablefor studying the photochemistry of dinuclear metal carbonyls inlow-temperature matrices. 8,24 Using this method, Dunkin etaI.24 were able to show that [Mn,(CO),] contained a semi-bridging CO group.In our experiment, an N, matrix containing (lb) wasphotolysed with plane polarised light (290 nm) and i.r. and U.V.spectra were recorded with polarisers parallel and perpendicularto the plane of the photolysing light, see Figure 5.The analysisof these spectra involves two stages: (i) the linear dichroism inthe bands of (lb) is used to establish that direction of thephotoactive transition moment which lies along the axis of thePt-Pt bond and (ii) given this orientation the presence ofdichroism in the i.r. band of (2b) can be used to distinguishbetween a ketonic and a semi-bridging CO group; a ketonic-bridging CO group [as in (A)] should show linear dichroismwhile a semi-bridging CO group, lying at approximately 45" tothe Pt-Pt bond [as in (B)] should not.*We first consider the bands of those molecules of (lb) whichremain intact after polarised photolysis. In such circumstancesthe photoactive U.V.absorption will always display dichroismperpendicular to the plane of polarisation of the photolysinglight. By contrast, both i.r. bands of (lb) exhibit dichroismparallel to this plane [Figure 5(a)]. Thus, the directions of thetransition moments of the v(C0) modes and the photoactiveU.V. transition must be orthogonal. Given the structure of (lb)(as already seen), the symmetric and antisymmetric v(C0)modes have dipole moment changes in a plane, almostperpendicular to the axis of the Pt-Pt bond. The photoactivetransition moment must therefore be parallel to the Pt-Pt bond.The U.V. absorption of (lb) at 360 nm was also observed to bedichroic, as shown in Figure 5(b), such that its transitionmoment must be parallel to that of the photoactive transition(290 nm), i.e.lying along the Pt-Pt bond. This is consistent with*This argument assumes no rotation of the Pt-Pt bond within thematrix cage on ejection of CO, which is reasonable considering the sizeof the moleculeJ. CHEM. SOC. DALTON TRANS.9.020 i8 - 50,010 -1988*rm 0 18-5 CK; + 0.01201800 1775 1750 1725v/ cm -1Figure 4. 1.r. spectra showing how 13C0 enrichment establishes thenumber of bridging CO groups in (2b). (a) The bands observed after30 min filtered U.V. photolysis of (lb), 50% enriched with 13C0 andisolated in a N, matrix. The two bands are assigned to [Pt,(p-'2CO)(q5-C,Me,),] and [Pt,(p-'3C0)(q5-CsMes)z]. (b) The spect-rum observed when an unenriched sample of (Ib) was photolysedunder similar conditions. Note that the bands due to l2C0 haveidentical half-widths in both experimentsthe assignment of the 360 nm band to a dx+o* or o-m*transition.It is also clear from Figure 5(a) that the i.r. band dueto the bridging CO group of (2b) shows significant dichroism ina direction perpendicular to the plane of the photolysing light.Such dichroism is inconsistent with a semi-bridging CO groupand therefore (B) can be rejected as a possible structure for (2b).The visible absorption band of (2b) at 540 nm, Figure 5(b), alsoshows dichroism, but parallel to the plane of the photolysinglight indicating that the electronic transition moment must liealong the Pt-Pt bond.The Electronic. Ground State of [Pt,(p-C0)(q5-C5H5),].-Itis much more dificult to determine the details of structure (A).The isoelectronic species [Fe,(p-CO),(q 5-C5Me5),] is a triplet,but this is because, by symmetry, the highest occupiedmolecular orbital is a doubly degenerate x* orbital.' We haveattempted to determine whether (2a) and (2b) are paramagneticby magnetic circular dichroism (m.c.d.). * Unfortunately theresults of experiments using the m.c.d.matrix apparatus 25,26 atthe University of East Anglia have so far been inconclusive.Both (2a) and (2b) are surprisingly unreactive towardsreagents added to the matrix. Thus, exactly the samephotochemistry occurs in N, matrices as in Ar matrices. Thereis no evidence for formation of any other products correspond-ing to the dinitrogen complexes, [Re,(CO),(N,)], which areobserved when [Re,(CO),,] is photolysed in N, matrices.27Such behaviour would be consistent with (2a) and (2b) havingtriplet ground states, since reaction with N, and similar ligands* Magnetic circular dichroism measurements on matrix isolated (2b)would appear to be an ideal method to determine its electronic groundstate, since a paramagnetic species should exhibit a temperature-dependence in its m.c.d.spectrum, whereas that of a diamagnetic speciesshould be independent of temperature.261505/ A A2040 1960 ' ' 1770V/cm -'' 500 600A / nmFigure 5. Spectra illustrating the dichroism generated by 60 h photo-lysis of (lb) in an N, matrix with plane-polarised light (290 nm). (a)1.r. difference spectrum ( A , minus All); bands are assigned to (Ib) and(2b), as labelled.(b) Superimposed u.v.-visible spectra; the spectrarecorded with the polariser perpendicular to the plane of photolysis,A,, have been filled in to distinguish them from A , l . U.v.-visible and i.r.spectra were recorded on the same samplewould be formally spin-forbidden. This lack of reactivity in (2a)and (2b) persists up to room temperature.[Pt,(p-C0)(q5-C5Me,),] in Solution at Room Temperature.-Time-resolved i.r. spectroscopy (t.r.i.r.) has already proved to bea successful method for detecting dinuclear intermediates inhydrocarbon solution at ambient temperature^.^,' 9.28 U.V. flashphotolysis of (lb) in cyclohexane solution generates anintermediate with a single i.r. absorption centred at 1 765 cm-',Figure 6.This wavenumber is very close to the absorptionobserved for (2b) in low-temperature matrices (see above) andcan be reasonably assigned to the same species. The lifetime of(2b) in these room-temperature experiments was surprisinglylong for an unsaturated species. The kinetic trace shown in theinset in Figure 6 illustrates that no significant decay of the i.r.absorption of (2b) was observable nearly 2 ms after the laserpulse. Similar traces were obtained even when CO oracetonitrile had been added to the solution, indicating that onthis time-scale the lifetime of (2b) is unaffected by the presence ofthese potential reactants. T.r.i.r. experiments with (la) revealeda transient species with an absorption at 1 802 cm-', very closeto that of (2a) isolated in a matrix (see Table 1).(It was notfeasible to heat the samples in these time-resolved experimentsbecause of their thermal sensitivity in solution.) Experiments todetermine more accurately the lifetimes of (2a) and (2b) insolution are in progress.In all previous t.r.i.r. studies on dinuclear metal carbonyls1506Ir 0.06rt - . . . . I9 .* . Ii0 < P p1800 1 750 1700v/cm-1Figure 6. The transient i.r. absorption band observed 250 ps after U.V.laser flash photolysis of (lb) in cyclohexane solution at room temper-ature. (Experimental points are shown bold and lighter points areinterpolated by computer.”) The inset illustrates the kinetic traceobtained at 1764.7 cm-’ showing no decay of the transient nearly 2ms after the laser pulsesignals due to metal-centred free radicals, formed by homolysisof the M-M bond have been observed.28 It was thereforesurprising that we did not observe any signals attributable to[Pt(CO)(C5R5)] radicals.Possibly the v(C0) absorptions of[Pt(CO)(C,R,)] lie outside the wavenumber range of ourpresent t.r.i.r. spectrometer. However, kinetic traces due to thev(C0) absorptions of (la) and (lb) (not illustrated) showed noevidence for regeneration of the dimers from any otherphotoproducts such as radicals. It is possible that for photolysisat 308 nm, the quantum yield for bond homolysis is very muchlower than that for CO loss. Under the conditions of ourexperiment, therefore, any signals due to [Pt(CO)(C,R,)] maybe too weak to be detected.Evidence for Photochemical Cleavage of the Pt-Pt Bond.-Despite our t.r.i.r.results, there is considerable chemicalevidence that homolysis of the Pt-Pt bond can occur. Forexample, U.V. photolysis of a mixture of (la) and (lb) inhydrocarbon solution leads to the slow formation of a‘crossover’ product,29 equation (6). There is also evidence for[(Pt(CO)(rl 5-C5H5))21 -I- [(Pt(CO)(rl 5-C5Me5))21 a2[( Pt(CO)(rl 5-C5Me5)(rl 5-C5H5)Pt(CO)I (6)bond homolysis in pure CO matrices, where chemical reactionwith CO circumvents the problems of recombination within thematrix cage. Filtered U.V. irradiation (23&345 nm) of (la) in aCO matrix produces results broadly similar to those observedin Ar matrices; the i.r. bands of (la) decrease in intensity and theband of (2a) (filled in) grows in, see Figure 7(a) and (6).Thereare, however, two weak absorptions in the terminal C-0stretching region, arrowed in Figure 7(6), which are notobserved in Ar matrices. There is a striking increase in theintensity of these arrowed bands after further photolysis of thematrix using unfiltered u.v.-visible light, Figure 7(c). The samebands are also observed, at exactly the same wavenumbers,when (lb) is photolysed with unfiltered U.V. light. This indicatesthat the photoproduct does not contain any C,R, ligands sinceone would expect a significant shift in wavenumber between thespecies where R = H and R = Me. [There is a shift of ca. 30cm-’ between the v(C0) bands in (la) and (lb).] Thus thepossibility of mononuclear species such as [Pt(CO)(q 5-C5R5)],[Pt(CO),(q3-C5R5)], or [Pt(CO),(o-C,R,)] is eliminated.However, the wavenumbers of the bands that we have0.60.300.E0.3al Ve ea 0v) n0- 60.3C( a 1I I2 050 1990x 4 -( b )J. CHEM. SOC. DALTON TRANS. 1988V / cm - 11810 1750Figure 7. 1.r. spectra illustrating the cleavage of the Pt-Pt bond in(la). (a) After deposition of (la) in a CO matrix at 20K. The bandmarked * is due to a trace impurity of [Ni(CO),] in the CO matrix. (h)After 120 min filtered U.V. photolysis; the filled-in band is due to (2a).(c) After 90 min unfiltered U.V. photolysis; note the striking growth ofthe arrowed bands due to [Pt(CO),]observed, 2 054.0 and 2 047.6 cm-’, correspond well withreported values for matrix-isolated [Pt(CO),], generated byco-condensation of Pt atoms and CO;30 the bands are due tothe t, v(C0) stretching mode, split into two components by siteeffects in the solid CO matrix.Under similar conditions [Pt(CO),] is generated morerapidly from (la) than from (lb), indicating that the quantumyield for formation of [Pt(CO),] is larger for (la) than for (lb).Although our experiments do not reveal the precise mechanismfor the formation of [Pt(CO),], it is clear that photolysis cancause cleavage of the Pt-Pt bond in both (la) and (lb).ConclusionsThe results presented here have shown that photolysis of[(Pt(CO)(q5-C,H5)),] leads to loss of CO and formation of arelatively stable intermediate, [Pt,(p-CO)(qS-C5H5),], with asingle bridging CO group.The fact that a CO group is bridginga bond between two third-row metal atoms makes aninteresting contrast to [Re,(CO),], which does not have anJ. CHEM. SOC. DALTON TRANS. 1988 1507CO bridges. The change from terminal to bridging CO groups isreminiscent of the photolysis of [{Fe(CO),(qS-C,H,)),l 3,4and such changes may transpire to be a characteristic feature ofthe photochemistry of cyclopentadienyl compounds. It isperhaps surprising that the quantum yield for homolysis of thePt-Pt bond in (la) and (lb) should be so low, particularlybecause formation of mononuclear species is common in thesesystems. Perhaps the higher quantum yield for loss of CO ispartially offset by the low reactivity of the dinuclear inter-mediates.Tantalisingly, the electronic ground states of (2a)and (2b) remain unknown and it is hoped to carry out furtherexperiments to try to resolve this question.AcknowledgementsWe thank the S.E.R.C., the E.E.C. (Grant No. ST2*/00081), ThePaul Instrument Fund of the Royal Society, and the Donors ofthe Petroleum Research Fund, administered by the AmericanChemical Society for supporting this research. We also thankMr. J. G. Gamble, Dr. M. A. Healy, and Mr. J. W. Whalley fortheir help and advice.References1 T. J. Meyer and J. V. Caspar, Chem. Rev., 1985, 85, 187.2 R. H. Hooker, Ph.D. Thesis, University of Southampton, 1986.3 R. H. Hooker, K. A. Mahmoud, and A. J. Rest, J. Chem. SOC., Chem.4 A. F. Hepp, J.P. Blaha, C. Lewis, and M. S. Wrighton, Organo-5 J. L. Hughey, C. R. Bock, and T. J. Meyer, J. Am. Chem. SOC., 1975,6 J. V. Caspar and T. J. Meyer, J. Am. Chem. SOC., 1980, 102, 7794.7 M. Poliakoff and E. Weitz, Adv. Organomet. Chem., 1986, 25, 277.8 D. S. Ginley, C. R. Bock, and M. S. Wrighton, Inorg. Chzm. Acta,1977, 23, 85.9 J. P. Blaha, B. E. Bursten, J. C. Dewan, R. B. Frankel, C. LewisRandolf, B. A. Wilson, and M. S. Wrighton, J. Am. Chem. SOC., 1985,107,4561.Commun., 1983, 102.metallics, 1984, 3, 174.97, 4440.10 H. B. Abrahamson, M. C. Palazotto, C. L. Reichel. and M. S.Wrighton, J. Am. Chem. SOC., 1979,101,4123; C . Gianotti and G. J.Merle, J . Organomet. Chem., 1976, 105, 97.11 E. 0. Fischer, H. Schuster-Woldan, and K. Bittler, Z. Naturforsch.,Teil B, 1963, 18, 429.12 L. R. Byers and L. F. Dahl, Inorg. Chem., 1980, 19, 680.13 N. M. Boag, R. J. Goodfellow, M. Green, B. Hessner, J. A. K.Howard, and F. G. A. Stone, J. Chem. SOC., Dalton Trans., 1983,2585.14 N. M. Boag, Organometallics, in the press.15 S. P. Church, M. Poliakoff, J. A. Timney, and J. J. Turner, Inorg.16 S. C. Fletcher, Ph.D. Thesis, University of Nottingham, 1985.17 M. S. Baird, I. R. Dunkin, N. Hacker, M. Poliakoff, and J. J. Turner,J. Am. Chem. SOC., 1981, 103, 5190.18 S. C. Fletcher, M. Poliakoff, and J. J. Turner, Inorg. Chem., 1986,25,3597.19 A. J. Dixon, M. A. Healy, P. M. Hodges, B. D. Moore, M. Poliakoff,M. B. Simpson, J. J. Turner, and M. A. West, J. Chem. SOC., FaradayTrans. 2, 1986, 2083.Chem., 1983, 22, 3259.20 A. E. Stiegman and D. R. Tyler, Acc. Chem. Res., 1984, 17, 61.21 J. H. Darling and J. S. Ogden, J. Chem. SOC., Dalton Trans., 1972,2496.22 P. McArdle and A. R. Manning, J. Chem. SOC. A , 1971, 717.23 0. Crichton, A. J. Rest, and D. J. Taylor, J. Chem. SOC., Dalton24 I. R. Dunkin, P. Haerter, and C. J. Shields, J . Am. Chem. Soc., 1984,25 R. G. Graham, R. Grinter, D. R. Stern, and K. Timms, J. Phys. E: Sci.26 T. J. Barton, R. Grinter, A. J. Thomson, B. Davies, and M. Poliakoff,27 S. Firth, W. E. Klotzbuecher, M. Poliakoff, and J. J. Turner, Inorg.28 B. D. Moore, M. B. Simpson, M. Poliakoff, and J. J. Turner, J. Chem.29 N. M. Boag, unpublished work.30 E. P. Kundig, D. McIntosh, M. Moskovits, and G. A. Ozin, J. Am.Chem. Soc., 1973, 95, 7234.Trans., 1980, 167.106, 7248.Instrum., 1986, 19, 776.J. Chem. SOC., Chem. Commun., 1977, 841.Chem., 1987, 26, 3370.SOC., Chem. Commun., 1984, 972.Received 6th August 1987; Paper 7/ 146

ISSN:1477-9226    

DOI:10.1039/DT9880001501

出版商:RSC    

年代:1988

数据来源: RSC