摘要:

2398 J.C.S. DaltonCarbon43 Nuclear Magnetic Resonance Spectra and Mechanism ofBridge-Terminal Carbonyl Exchange in Di-p-carbonyl-bis[carbonyl(q-cyclopentadienyl)iron] (Fe-Fe) [((q-C5H5) Fe(C0)2},] ; cd-Di-p-carbonyl-f-carbsnyl-ae-di(q-cyclopentadieny1)-6-(triethyl -phosphite)di-iron(&-Fe)[(~-C,#,)2Fe,(CO)3P(OEt)3],t and some Related ComplexesBy Daniel C. Harris, Edward Rosenberg, and John D. Roberts,* The Gates and Crellin Laboratories ofChemistry, California Institute of Technology, Pasadena, California 91 109, U.S.A.A mechanism involving carbonyl-bridge breaking, rotation about the Fe-Fe bond, and bridge reformation is shownto account qualitatively for changes in the carbonyl region of the 13C n.m.r. spectrum of the complex [(cp)(OC)-Fe(p-CO)zFe(cp){P(OEt),)] and quantitatively for [(cp) (OC)Fe(p-CO),Fe(CO) (cp)] (cp = 3-cyclopentadienyl).The activation energy for this process, 49.0 f 4 kJ m0l-l (1 1.7 f 1 kcal mol-l), is close to that reported for cis-trans-isomerization of the cp groups, in accord with this mechanism.Variable-temperature 13C n.m.r. spectra ofthe complexes [ ( c ~ ) ( O C ) R U ( ~ - C O ) ~ R U ( C O ) ( C ~ ) ] and [(cp)(OC)Fe(p-CO),Ni(cp)] are also reported.- - - -THE variable-temperature 13C n.m.r. spectrum of thecomplex [(cp)(OC)Fk(p-CO),Fe(cp)L] (I; L = CO, cp =q-cyclopentadienyl) has been interpreted in terms ofboth bridge-terminal and cis-trans-carbonyl exchange-lMechanisms have been recently published by Adams andCott0n,~>3 which account for the fact that bridge-terminal exchange in the trans-isomer of complex (I) ismuch more facile than that in the cis-isomer.Carbonyl-bridge opening, followed by bridge closure, with norotation of the non-bridged intermediate [equation (l)]accounts for bridge-terminal exchange in the trans-isomer. The cis-isomer, however, cannot undergobridge-terminal exchange without simultaneous rotationabout the metal-metal bond [equation ( Z ) ] .Before the Adams and Cotton papers appeared, wei Throughout the paper, only the name of the cis-isomer is0. A. Gansow, A. R. Burke, and W. D. Vernon, J . Amer.given in full.Chem. SOC., 1972, 94, 2550.initiated an investigation of carbonyl exchange in thecomplex [(~p)(OC)~e(p-CO)~&+e(cp)L] [I; L = P(OEt),]for which equation (1) is not applicable, because thephosphite ligand cannot reside in a bridging position.It was expected, therefore, that the trans-isomer wouldnot exhibit bridge-terminal averaging independent ofcis-trans-isomerization.We also sought to delineatethe kinetics of exchange from the change in linewidthsof carbonyl absorptions of complex (I; L = CO) as afunction of temperature. Finally, we undertook briefinvestigations of the complexes [(cp) (OC)Ku(p-CO),Ru-The results obtained are in full agreement with thoseof Adams and Cotton with different ligands.- - (CO) (cp)] , (11) J and [(cp) (OC) Fe(p-CO),Ni(cp)l, (111).R. D. Adams and F. A. Cotton, I?.torg. Chim. Acta, 1973, 7,153; J . Awev. Chem. SOC., 1973, 95, 6589.3 The same mechanisms were developed independently in thislaboratory, cf.D. C . Harris, Ph.D. Thesis, California Institute ofTechnology, 1973, pp. 404-4081974a2399Ltrans- (I) L =CO, P(O€t),,or P(OMe)3a,f- bridge closure.a,t- bridge openingCis-( I) L d O , P( EOt)&or P( OM e )30 c'70c cL JLEXPERIMENTAL - The complex [(cp) (OC)Fe(y-CO),Fe(cp)L] (I; L = CO)was obtained from a commercial source and ciystallizedfrom ethanol. The method of Haines and DuPreez4 wasused t o prepare the complexes [I; L = P(OMe), andP(OEt),]. The complex [(cp) (OC)Ru(p-CO),Ru(CO)(cp)],(11) , was prepared by a procedure similar to Manning's 5*and sublimed twice, while [(cp) (OC)Fe(p-CO),Ni(cp)], (111) ,was obtained by the method of Tilney-Bassett,' chromato-graphed on neutral alumina with 4 : 1 light petroleum (b.p.30-60 OC)-benzene, and crystallized from 6 : 1 lightpetroleum-benzene.13C0 Enrichnzeizt.-Complexes [I; L = CO or P(OEt),],(11), and (111), were stirred in benzene solutions under 90%13C-enriched carbon monoxide for 1 day at 45-50 "C.Enrichments were 14, 33, 48, and 367; respectively.Thepercent enrichment was determined from heights of them/e 28 (I2C0) and 29 (13CO) peaks in the 70 eV massspectrum. All complexes other than (I; L = CO) reactedwith CO to produce products in addition to enrichedstarting material. Thus, complex [I; L = P(OEt),]R. J. Haines and A. L. DuPreez, Inorg. Chem., 1969,8, 1459.A. R. Manning and P. A. McArdle, J . Chem. SOG. (A), 1970,-I 12125.d lproduced a mixture of 21% (I; L = CO) and 79% [II;L = P(OEt),].This mixture was converted t o pure[I; L = P(OEt),] by stirring i t with excess of triethylphosphite [the mole ratio of triethyl phosphite t o (I;L = CO) was 5 : 11 in benzene at 60 "C in vacuo with periodicremoval of the evolved CO by freeze-thaw degassing on avacuum line. The purity was readily checked by IH n.m.r.spectroscopy (Table 1). Complex (11) was freed fromunidentified ruthenium carbonyls produced by enrichmentby crystallization from toluene. The carbonylation by-products of complex (111) included (I; L = CO) and agreen volatile solid, presumably (cp)&. No attempt wasmade to remove complex (I; L = CO) from (111) afterenrichment.N.M.R. Spectra.-Fourier-transform 13C n.m.r.spectrawere recorded on our modified Bruker-Varian DFS-60spectrometer operating at 15.09 MHz with deuteriumlock and proton-noise decoupling. Chemical shifts are allin p.p.m. downfield of internal tetramethylsilane (tms) .Temperatures were measured with a thermocouple orthermometer and were generally constant t o &2 "C. 1.r.An apparently superior synthesis has recently been pub-lished: A. P. Humphries and s. A. R. Knox, J.C.S. Chem. Comm.,1973, 326.J . F. Tilney-Bassett, J. Chem. Soc., 1963,47842400 J.C.S. Daltonspectra were obtained xvith a Perkin-Elmer 225 spcctro-meter. All reactions were carried out in atmospheres ofnitrogen or argon. All solids, except (111), could behandled in the air without apparent harm, but were storedunder a nitrogen atmosphere in a refrigerator.Solventsfor the 1i.m.r. spectra were dried over Linde 4A molecularsieves and distilled in vacuo into the n.m.r. tubes. Solventsfor other purposes were deoxygenated by bubbling a streamof nitrogen through them. Unless otherwise stated, alln.m.r. samples contained 0~08-0~10~-Cr(acac), (acac =2,4-pentanedionate) (crystallized from benzene-hexane,n1.p. 212-213 "C) as relaxation reagent and were sealedunder a N, atmosphere.lH Chemical shifts of all groups in complexes (I) wereat significantly higher fields in benzene and toluene than innon-aromatic solvents (see Table 1). The 13C shifts fortriethyl phosphite are given in Table 2. Complex [I;L = P(OMe),] exhibited bridge and terminal carbonylcarbon absorptions of the predominant isomer a t 282.0TABLE 1complexes (I)Chemical shifts (p.p.m.) and coupling data (J/Hz) forChemicalshiftJJ Solvent cpl cp2 J(31P-1HZ)CO MeC0,H 4.89MeCN 4.83CD2C12 4-77c6c6 4.31CDC1, 4.66 4.51 0.9CD2C1, 4-67 4.52 1(CD,),CO 4.67 4-56 1.0P(OEt), CH,C12 4-66 4.50 1.2c6c6 4.45 4.35 1.0C6H6Me, 4.48 4.40 1-1P(OMe), Me,C02H 4.67 4.63 1.5C6H6 4.46 4.36Chemicalshifts for LCH, CH,3-503.533-523.503.303.97 1.183.86 1.023.87 1-01For P(OEt), in benzene, 6(CH3) 1.12 and 6(CH,) 3.81 p.p.m.[J(lH-lH) 7,J(31P-CH,) 7 Hz].TABLE 2Triethyl phosphite 13C n.m.r.data aChemical shift/p.p.m. J(3lP- W)/Hz57.9 17.1 11.6 4.8Compound CH2 CH, CH, CH316.2 4 6 [I; L = P(OEt),] 60.8[I; L = P(OEt),] 60.5 16 4 6P(OW3in CH,Cl,in tolueneChemical shifts arereferred to tms as zero.The triethyl phosphite sample con-sisted of triethyl phosphite (2 cm3), CDCI, (1 cm3), and tms(0.2 cm3).( J NN 19 Hz) and 215.4 p.p.ni., respectively. Cyclopenta-dienyl carbon atoms appeared a t 87.3 and 85.8 p,p.m., andthe methyl carbons a t 51.4 p.p.m. These data are for 1 : 1Cl,C(D)C(D)Cl,-toluene solutions, 0 . 1 6 ~ in complex [I;L = P(OMe),].Complex [I; L = P(OEt),].-To test the proposedmechanism, both cis- and trans-isomers should be present(1 All spectra were run at 30 f 2 "C.RESULTS AND DISCUSSIONin solution. From the work of Haiiies and DuPreez,*it is clear that both are present in cyclohexane, but anassignment of which isomer gives rise to which i.r.bandshas never been made. In cyclohexane, we found two+ 12? "c+72 "CA 1 +32 "Ci1 -35°C6 1p.p. m.FIGURE 1 Variable-temperature Fourier-transform I3C n.m.r.spectrum of a 0 . 1 3 ~ solution of complex [I; L = P(OEt),](33% enriched in WO) with O.OSM-Cr(acac), in toluene. Thelower five traces represent 2 000 transients and the upper two4 000. The 238.8 p.p.m. absorption belongs to an irreversiblyformed thermal-decomposition product, probably (I ; L = CO) ,which remains when the sample is cooled to 52 "Cstrong bands in the terminal metal-carbonyl stretchingregion at 1964 and 1944 crn-l, logically taken toindicate the presence of both cis- and trans-isomers inthis solution. In toluene, the bands were much broaderand overlapped, yielding a maximum a t 1955 and ashoulder at 1940 cm-l.In dichloromethane, a secondisomer was not evident from the i.r. spectrum, ther1974 2401being a nearly symmetrical peak at 1 952 cm-l. Finally,the solid (KBr pellet) appeared t o contain only one ofthe two isomers, the small splitting of the bands being asolid-state effect also seen in the individual isomers ofcomplex (I; L = CO).*The low-temperature 13C n.m.r. spectrum of complex[I; L = P(OEt),] in toluene (Figure 1) and dichloro-which must be under the cis-absorptions. This smalltrans-signal coalesced with the cis-signals at higher tem-perature. It is to be noted that phosphorus-carboncoupling was reduced when exchange was rapid amongstthe carbonyl positions.The coalescence behaviour isfully consistent with the bond-rotation mechanism[equation (2)]. Bridge and terminal absorptions of the+48"C* 282.2J=23 HZ87.6 86.3-32 'C 89-5 \fL .1.__1__. Bridge CO Terminal CO ~--~.&%%%* 2 1 CPS/ppm.P'IGURE 2 Variable-temperature Fourier-transform 13C n.m.r. spectrum of a 0 . 1 7 ~ solution of complex [I; L = P(OEt),j(33% enriched in WO) with O.OSM-Cr(acac), in dichloromethane. Each trace represents 2 000 transientsmethane (Figure 2) is very revealing. (We confirmedthat no further changes took place in the spectrumbelow the lowest temperatures shown in Figures 1 and2.) In the terminal-carbonyl region (ca. 215 p.p.m.),boih solutions exhibited two isomers, The ratio of thelatter was 4-4 (hl) : 1 in toluene and 5.9 ( 3 1 ) : 1 in di-chloromethane.As we expect the cis-isomer to bemore polar than the trans-isomer, there should be morecis-isomer in dichloromethane than in toluene (or cyclo-hexane). On this basis, we believe the predominantisomer in all three solvents to be the Inthe bridging-carbonyl region, the isomers producedoverlapping absorptions near 280 p.p.m. u(31P-13C)22 Hz]. In the cp region, the cis-isomer gave two peaksat 87.6 and 86.3 p.p.m. (Figure 3) with no observable31P-13C coupling. A smaller peak at 89.5 p.p.m. is prob-ably clue to the tram-isomer, the other absorption of* 1.r. spcctra of solutions of complex [I; L = P(OhIe),] weresimilar to those of [I; I. = P(OEt),], so we believe the cis-isomeris dominant herc also : (cyclohexane) 4 1 964s, 1 944m, and 1 750s;(chlorofomi) 2 056X11, 2 OlOm, 1 063s, 1 773m, and 1 731s cm-l.The solid niay contain the trans-isomer because the KBr pelletspectrum showed bands a t 1983vw.1933vs, 1 8 8 7 ~ . 1770m,1 733vs, and 1 707sh, ni cm-l. The I<Br pellet of complex1 766w, 1 739vs, 1 727vs, and 1704w cm-l) did not allow us toassign a configuration for i t in the solid state, but the differentwavenumbera of the terminal-carbonyl bands of complexes [I ;1, :: P ( O l k ) , and P(OEt),] suggest that they possess differentconfigurations in the solid state.* R. I:. Bryan, P. T. Greene, RI. J. Newlands, and D. S. Field,J. Chriii. SOC. ( A ) , 1970, 3068.rr; r, -= ~ ( o ~ t ) , ] (z omw, z O O ~ W , 1 957vs, i 95ovs, 1 906w,cis- and trans-isomers coalesced simultaneously with nodiscernible exchange occurring in one isomer before theother.+L5 "C hBridge Averaged Terminal+32 "C+17 'Cn-17 "C- 3 4 "CA_Bridge Averaged Terminal6 / p.p.m.FIGURE 3 Calculated l3C n.1n.r. spectra as a function of exchangerates for complex (I; L = CO). Each curve was calculatedusing the rate constants of Table 3 a t the indicated temperatures The same reasoning led to the first successful assignment ofthe isomers of complex (I; L == CO) ; A. I<. Manning, J . Chain.Soc. ( A ) , 1968, 1319J.C.S. DaltonCol.rzpZex (I; L = CO).-The spectrum of complex(I; L = CO) published by Gansow et aZ. was interpretedas fo1lows.l At -85 "C, the cis- and trans-isomers areboth static on the n.m.r.time scale. Each exhibitssingle absorptions for the bridge- and terminal-carbonylgroups but the signals accidentally overlap so that onlyone absorption is observed in each region. At -73 "C,the less-abundant trans-isomer exhibits bridge-terminalexchange, producing an averaged signal between thebridge and terminal signals of the cis-isomer. Thisaveraged signal sharpens as the temperature is raiseduntil, at -35 "C, all three signals begin to broaden andcoalesce near -12 "C. In the high-temperature limit,only a single sharp absorption is observed in the bridge-terminal averaged position. The process stated toaccount for coalescence of the cis-isomer carbonylsignals is either bridge-terminal exchange within theisomer or cis-trans-isomerization.We investigated spectra of complex (I; L = CO)using dichloromethane as solvent and our results are inagreement with those reported with one exception:namely, the low-temperature absorption due to averagedcarbonyl groups of the trans-isomer was not preciselyTABLE 3Linewidths of carbonyl absorptions of compIex (I;L = CO) and rate constants for exchangeLinewidth/Hzt , "C cis a Averaged k , c1s-l k , e/s-l- 70 9 12 8 13- 57 10 14 9 15-51 13 23 16.5 18- 44 18 32 24 26- 34 38 61 47 65- 17 d d 490 490- 1; ca. 130 7 400 7 400- 32 38 20 600 20 50039 24 33 500 33 500-- 45 20 40 000 40 000+51 15 5s 000 68 000-- 5s 10 100 000 100 000a Average value for bridge and terminal signals.Un-certainty is f l Hz.Only at -34 "C did the two linewidthsdiifer by more than 1.0 Hz (bridge -41, terminal -36 Hz).a Below - 17 "C this is the width of the single trans-absorption.Above - 17 "C this is the width of the only signal in the car-bony1 region. C Assuming intrinsic linewidths in the absence ofexchange to be 2.0 Hz, these rate contants reproduce theobserved linewidths within 5 0 . 5 Hz. d Too broad to measure.Rates estimated from an Arrhenius plot of the other data.midway between static carbonyl signals of the cis-isomer. At -73 "C, the averaged trans-signal was at2424 p.p.m., while bridge and terminal signals of thecis-isomer appeared a t 272.9 and 211-0 p.p.m. Thisleads us to suggest that both bridge-terminal andcis-trans-isomerization occur rapidly at room tem-* We verified that 0.08M-Cr(acac), does not affect carbonyllinewidths at -62 "C.t Line-shape calculations were made with program DNMR3written by D. A. Kleier and G. Binsch, Quantum ChemistryProgram Exchange, Indiana University, 1969. We are gratefulto Drs. F. A. L. Anet and W. Larson for helpful discussion con-cerning these calculations.From -72 to -34 "C, trans : cis = 0.25 : 1, within experi-mental error estimated at f0-03.perature, because, if there were only separate bridge-terminal exchange for the cis- and trans-isomers (with-out cis-trans-isomerization) , one would expect twopeaks in the high-temperature limit. Our interpretationof the low-temperature spectrum differs from that ofthe earlier one only in that we do not believe the low-temperature limit is reached a t -85 "C.The observedspectrum a t this temperature corresponds to an inter-mediate rate of trans-isomer bridge-terminal exchangewhereby the signal is so broad that nothing is seen.The cyclopentadienyl absorptions for the cis- andtrans-isomers at -73 "C were 88-6 and 89.9 P.P.M., re-spectively. These coalesced below -44 "C and gave asingle sharp peak at 89-0 p.p.m. at 3-60 "C.Rates of cis-trans-isomerization and bridge-terminalexchange were determined by fitting measured carbonyllinewidths * to a series of theoretical linewidths corre-sponding to particular rate constants t (Table 3).Because spectra span only the region where bridge-terminal exchange of the trans-isomer is very rapid,only one averaged chemical shift was considered forthis isomer.In equation (3), (A) and (B) correspondto bridge and terminal carbonyl groups of the cis-isomer and (C) corresponds to those of the trans-isomer.Attempts to match linewidths at low temperature, withk, equal to zero, were not successful. Allowing k , andk, to vary, and using the measured ratio of cis- andtrans-isomers to determine k,,$ a unique fit was ob-tained below -17 "C where there are two observedlinewidths to fit with two rate constants. The valuesof k, and k, (Table 3) seem close enough to being equalto indicate essentially free rotation interconverting allthree of the staggered non-bridged species in equation(2). Reproducing one Linewidth with two rate constantsat high temperature did not yield a unique solution,but the low-temperature fit seemed to justify assumingequal values of k, and k,.With this assumption, thehigh-temperature fit was unique also. Simulated spectrausing the rate constants of Table 3 are shown in Figure 3.An Arrhenius plot of the experimental rate constantsTABLE 413C Chemical shifts a for complex (11)Solvcnt t/"C co CPC,H,Me + 37 212.8 88.2C,H,Me - 10 215.5 89.1C,H,Me - 47 219.2 90.3CHQ, $- 34 216-9 89.4CH,CI, - 42 222-6 90.8a With tms as zero, positive shifts are downfield.was linear and yielded an activation energy of 49.0 -J-- 4kJ mol-l (11.7 & 1 kcal mol-l), as compared to 16.7 &1.6 kcal mol-l calculated from cis-trans-isomerizatio1974followed by 1H n.m.r. spectroscopy.1° Significantdeviations from a linear Arrhenius plot occurred only atour lowest temperature, -70 "C, where the calculatedrate constants were too large. This could be due tobreakdown of the assumption of infinitely rapid bridge-terminal exchange of the trans-isomer and generalbroadening of all lines in the spectrum, includingsolvent, at this low temperature.Other CompZexes.-For complex (11) there was a singlesharp peak in the carbonyl region and one in the cpregion (Table 4). Both had significantly temperature-dependent chemical shifts, consistent with the rapidlyaveraging set of cis- and trans-bridged and non-bridgedisomers believed to be present in s o l ~ t i o n . ~ * ~ ~ Themixed complex (111) gave a single carbonyl n.m.r.absorption at 237.0 p.p.m. at +35 "C, and at 238.1p.p.m. at -62 "C in toluene, indicating rapid bridge-terminal exchange throughout this temperature range.Only one isomer has been detected in solutions of0e-N , /c\complex (111),11 but this does not preclude intermediacyof the other isomer in bridge-terminal exchange.We thank the National Science Foundation and thePublic Health Service for support.[4/034 Received, 8th Jnitztary, 19741lo J. G. Bullitt, F. A. Cotton, and T. J. Marks, Inorg. Clzem.,1972,11, 671 ; the authors state that their value of E, may be toolarge.l1 P. McArdleand A. R. Manning, J . Chem. Soc. ( A ) , 1971, i 1 7

ISSN:1477-9226    

DOI:10.1039/DT9740002398

出版商:RSC    

年代:1974

数据来源: RSC