摘要:
1976 2187Kinetics of Nucleophilic Attack on Co-ordinated Organic Moeities. Part3.t Effect of Solvent on Addition of Pentane-2,4-dione to Complexes ofCyclic DienylsBy C. A. Mansfield and L. A. P. Kane-Maguire,' Department of Chemistry, University College Cardiff, CardiffThe kinetics of addition of pentane-2.4-dione (Hpd) to [Fe(C,H,) (CO),] [BF,] and related dienyl salts have beenstudied in the solvents nitromethane, acetone, and dichloromethane. Comparison with previous data in acetonitrileshows that the reactions are not sensitive to solvent changes, and indicate the general rate law -d[R]/dt = k[R]-[Hpd] (R = dienyl cation). These solvent effects, together with the effects of added acid, added base. anddeuteriation of Hpd, are explained in terms of a common mechanism involving rapid pre-equilibrium dissociation ofHpd to give the pd- carbanion which then adds directly to the dienyl rings.The analogous reaction of [Ru-(C6H1) (CO),] [BF,] with Hpd is considerably slower.CF1 1XLA PREVIOUS paper 1 in this series reported the kinetics ofaddition of p-diketones to cyclic dienyl iron cations inacetonitrile as solvent [equation (1); A = H or OMe].The results supported earlier suggestions that thereactions proceed via direct addition to the organic ring.We proposed a detailed mechanism involving a rapidpre-equilibrium in which the p-diketone dissociated togive a small concentration of reactive carbanion , whichthen attacked the dienyl ring in a rapid but rate-determining manner. The reactions of [Fe(C,H,) (Co),]-[BF4] (I), [Fe(C6H60Me)(CO),][BF4] (2), and somerelated cations with pentane-2,4-dione (Hpd) have nowbeen examined in greater detail in order to throwfurther light on the mechanims of the additions.Inparticular, the effects of (i) variation in solvent, (ii) acid,(iii) base, and (iv) deuteriation have been examined.EXPERIMENTALPt'efiaration of Complexes.-The salts [Fe(C6H7) (Co),]-[BF,] (3), and [Ru(C,H,)(CO),][BFJ (4) were prepared bypublished methods3-5 and their purity was checked byanalysis (Found: C, 35.6; H, 2.5. Calc. for C,H,BF,FeO,:C, 35.1; H, 2.3. Found: C, 35.9; H, 2.5. Calc. forC,oH,BF4Fe0,: C, 35.8; H, 2.7. Found: C, 37.4; H,2.9. Calc. for C,,H,BF,FeO,: C, 37.5; H, 2.8. Found:C, 31.1; H, 2.4. Calc. for C,H,BF,O,Ru: C, 30.8; H,Tricarbonyl( 2-methoxycyclohexadienyl) iron tetraphenyl-PF4I (l), [Fe(C6H60Me) (c0)31[BF41 (2)J [F~(C~HS)(CO)~I-2.6%).borate.The complex [Fe(C,H,OMe) (CO),][BFJ was dis-solved in a minimum volume of water and an excess ofsodium tetraphenylborate (also dissolved in water) wasadded. On cooling in ice a cream precipitate separated,which was filtered off and washed with water. Theprecipitate was dried in a vacuum desiccator (yield ca.95%) (Found: C, 71.7; H, 5.3. Calc. for C,,H,,BFeO,:C, 71.9; H, 5.3%).7 Part 3, K. M. Al-Ksthumi and L. A. Kane-Blaguire, J.0t)ganometallic Chem., 1975, 102, C4.L. A. P. Kane-Maguire, J . Chetn. SOC. ( A ) , 1971, 1602.a I. U. Khand, P. L. Pauson, and W. E. Watts, J . Chem. SOC.3 A.J. Birch, P. E. Cross, J. Lewis, D. A. White, and S. B.(C), 1969, 2024.Wild, J. Chem. SOC. (A), 1968, 332.All the solvents were distilled under nitrogen and storedover molecular sieves. Their water contents were deter-mined by the Karl-Fischer method.6 Typical values wereMeCN (0.003), MeNO, (0.05), and CH,Cl, (0.008~0 w/w).The solvents were deoxygenated by passage of a stream ofnitrogen for 30 min immediately prior t o use.Pentane-2,cdione (Hpd) was distilled under nitrogenimmediately prior t o use. Deuteriated pentane-2,4-dione(MeCOCD,COMe) was prepared by shaking Hpd (10 cm3)with deuterium oxide (10 cm3) for 1 h, and extracting withsodium-dried diethyl ether. The extract was dried overanhydrous magnesium sulphate and the ether removed.The degree of deuteriation was determined by 1H n.ni.r.tobe cu. 80%.Kinetic Studies.-All the reactions were studied in thedark under nitrogen using the i.r. technique previouslydescribed-l In general, the disappearance of the longest-wavelength carbonyl band of the original dienyl salt atca. 2 110 cm-1 was monitored. A large excess of pentane-2,4-dione was always employed, and pseudo-first-order rateconstants were calculated from the gradients of plots oflog At against time, where At = absorbance at time f ofthe reaction solution at the chosen wavelength.For the runs at constant pH, analytical grade sulphuricacid was added t o the solvents acetonitrile, nitromethane,and acetone prior t o mixing with ( 1) or (2). With dichloro-methane, toluene-p-sulphonic acid was employed due to theimmiscibility of H,SO, with this solvent.Conductivity Studies.-Solutions of fluoroboric acid(40%) of the appropriate concentrations were made up infreshly distilled nitromethane or acetonitrile.The amountof water introduced in this manner was never greater than0.3%. Conductivities were measured at 25 f 0.1 "C usinga Pye-Unicam type E7566/3 conductivity bridge.Keto-EnoZ Equilibrium Values.-The keto-enol equili-brium concentrations for Hpd in various (deuteriated)solvents were calculated from the relative intensities oftheir respective methyl peaks in the lH n.m.r. spectrum.Reaction of Tricarbonyl[2-methoxy-5-(pentane-2', 4'-dionyZ)-cycZohexa-l,3-diene]iron with A cid.-A small amount of theM.A. Hasmi, S. D. Munro, and P. L. Pauson. J . Chem. SOC.( A ) , 1967, 240.R. G. Cowles, Ph.D. Thesis, University College, London,1969.A. Vogel, ' Quantitative Inorganic Analysis,' 2nd edn.,Longmans, 1951, p. 6982188 J.C.S. Daltoncomplex1 was dissolved in acetonitrile and exhibited i.r.bands at 2 040 and 1980 cm-1. Addition of one drop ofconcentrated H,SO, to the solution resulted in immediateloss of these peaks and the appearance of bands due to[Fe(C,H,OMe)(CO)J+, i.e. at 2 110 and 2 060 cm-l.RESULTSLinear first-order kinetics were generally obtained for40-50y0 completion of reaction in nitromethane and for atleast 75% completion of reaction in dichloromethane.These deviations from linearity are analysed in the Discus-sion.Kinetic data for the addition of pentane-2,4-dione tocations (1) and (2) are summarised in Tables 1 and 2. AllTABLE 1Kinetic data for the addition of pentane-2,kdione to[Fe(C,H,) (CO),][BFJ in nitromethane35.1 5.4035.2 5.6240.1 8.5050.2 17.154.9 23.0TABLE 2Kinetic data for addition of Hpd to[Fe(C,H,OMe) (CO),][BF,] in various solventsed"C 1 05kob8./S-1[HPdl - 0, 1O5k0bO.Solvent mol dm-3 "C S-1CH,Cl,MeNO, 1 .oo1 .oo1 .oo1.000.800.500.201.001 .oo1 .oo1 .oo0.601 .oo1.00Me,CO 1 .oo40.145.050.659.959.959.959.924.529.234.839.634.829.340.149.60.791.021.682.852.101.260.6750.531.142.053.850.962.314.167.03the rate constants are the average of at least duplicatedeterminations (reproducibility f 5% in MeNO,, andf 10% in CH,Cl,).The poorer reproducibility in dichloro-methane may arise from evaporation problems, since someof the runs were performed near the boiling point of thesolvent (40 "C). The results with acetone are subject toconsiderable error since the first-order kinetic plots showedan unusual acceleration in rates after ca. 30-50% com-pletion of reaction. This acceleration may be due to aconcomitant acid-catalysed aldol condensation. The datain Table 2 indicate a linear dependence of bobs. on [HpdJfor its reaction with [Fe(C,H,OMe) (CO),][BF,] in bothnitromethane and dichloromethane as solvent.Activation enthalpies for each reaction were calculatedfrom the gradients of plots of log kobs.against 1/T, and arecollected in Table 3. The gradients were estimated byleast-squares analysis, the errors quoted being the standarderrors of estimation. Entropies of activation were cal-culated by standard methods using kobs. data obtained with[Hpd] = 1.00 mol dm-3.The effect of added acid on bobs. for addition of Hpd tocations (1) and (2) is shown in Table 4 and Figure 1. In allfour solvents (acetonitrile, nitromethane, acetone, anddichloromethane) this resulted in a significant decrease inrate. On the other hand, addition of the base 2,6-dimethyl-pyridine caused an instantaneous reaction in all solvents.FIGURE 1 Dependence of hob& on [H+] for the addition ofHpd t o [Fe(C,H,)(CO),][BFJ i n acetonitrile at 50.4 OCTABLE 3Comparison of rates and activation parameters (1 cal =4.184 J) for addition of Hpd (1.00 mol dm-3) to dienylcations in various solventslO5hOb,. (40 "C) AH* A SComplex Solvent S-1 kcal mol-1 cal K-1 mol-lMeCN 5.6 16.3 f 0.2 -26(') MeNO, 8.5 14.0 f 0.7 -33MeCN a 0.62 17.1 & 0.5 -28MeNO, 0.79 13.5 f- 0.6 -41Me,CO 4.2 10.9 f 0.4 -52CH,Cl, 3.7 23.7 f 0.9 - 3(2)4.53b 15.6 f 0.3 -33 MeCN aMeNO, 3.94 a(3)(4) MeNO, 0.8a Ref.1. At 61.0 "C. C Upper limit.TABLE 4Effect of added H,SO, on k&s. for addition of Hpd(1.00 mol dm-3) to cyclic dienyl salts--- [H2S04] 105[H+] lO5kob,.Complex Solvent mol dm-3 mol dm-3 s-lMeCN 0 0 15.00.016 2.75 3.970.027 3.72 2.180.040 4.4 1.150.050 5.00 1.17MeNO,b 0 17.10.027 cMe,CO 0 4.160.027 0.74MeNO,d 0 0.790.027CH,Cl, d p c 0 4.000.027 1.2( 1)(2)ca Calculated using the relation p H = &pK, - *loge, andconsidering only dissociation of the first proton [pK, 7.3 (ref.9)]. b At 50.4 "C.No reaction. d At 40.0 "C. Toluene-p-sulphonic acid was added.TABLE 5Effect of deuteriation on k&s. for addition of Hpd(2.00 mol dm-3) to [Fe(C,H,)(CO),][BFJ at 25 "CSolvent Me,CO MeCOCD,COMe R E / ~ DMeCN 3.92 4.60 0.85CH,Cl, 1.70 1.87 0.961 05kobs./~-1r 1976 2189Deuteriation of the Hpd nucleophile had little effect onkobs. for addition to [Fe(C,H,OMe) (CO),][BF,] in eitheracetonitrile or dichloromethane as solvent (Table 5).Figure 2 shows the variation in equivalent conductance(A) of fluoroboric acid with concentration (c) in the solventsacetonitrile and nitromethane. The proportions of keto-and enol tautomers of Hpd present at equilibrium in varioussolvents are summarised in Table 6.Attempts t o extend the range of solvents were hamperedby solubility problems or reaction of the iron substrateswith the solvent (methanol, ethanol, or dimethyl sulpho-xide).The complex [Ru(C,H,)(CO),][BF,] (4) could not bestudied in acetonitrile or nitromethane because slowdecomposition occurred to give an unidentified product (s) .However, since addition of 1.00 mol dm-3 Hpd caused nosignificant increase in the rate of disappearance of (4), anupper rate limit could be estimated for the reaction of (4)with Hpd (Table 3).DISCUSSIONIn our previous investigation of the addition [equ-ation (l)] of pentane-2,4-dione to cyclic dienyl salts(R[BF,]) in acetonitrile solvent the mechanism inequations (2) and (3) was proposed.This mechanismsolvents investigated (acetonitrile, nitromethane, ace-tone, and dichloromethane). After considering theFIGURE 2 Variation of equivalent conductance of HBF, withconcentration in the aprotic solvents MeCN (0) and MeNO,( 0 )general influence of solvent on reaction (l), the specificeffects of added acid, added base, and deuteriation willbe discussed in turn.envisages rapid pre-equilibrium dissociation of Hpd togenerate the reactive pd- carbanion, which then adds tothe dienyl complexes in a rapid but rate-determiningfashion. The rate equation for such a scheme is givenin (a), and was supported by the decrease in kobs.observed on addition of acid.However, in less-polarsolvents such as dichloromethane Ka( = kd/h+) for the(4)k k Rate =A *[Hpd][Fe]kd [ H+]= kobs. [ Fe]acid-dissociation step (2) is expected to be very small.'In such solvents an alternative mechanism [equations (5)and (S)] involving direct addition of undissociated HpdIn.uence of Solvent on Reaction (l).-Variation insolvent is seen (Table 3) to have little effect on the size ofTABLE 6Keto-enol equilibrium concentrations of Hpd in varioussolventsKeto (%) En01 (%) Solvent16.0 84.038 62 CD,CNCD,CN (+H+) 38 62CD,C1 17 83HPdCD,COCD, 26 74kobs. for reaction (1).changing from acetonitrile to dichloromethane.There is only a six-fold increase onWhileto the dienyl cations might be anticipated.It is inter-esting therefore that the present results confirm theoperation of the mechanism in (2) and (3) in each of the2nd edn., Wiley, New York, 1961, ch. 7.this small solvent dependence might be thought to lendsupport to the mechanism in (5) and (6) in which therate-determining step involves a neutral reactant, it isK, value anticipated for step (2) in dichloromethane 7 A. A. Frost and R. G. Pearson, ' Kinetics and Mechanism,' not inconsistent with processes (2) and (3). Th2190 J.C.S. Daltoncompared to acetonitrile is expected to be counter-balanced by a more rapid k for step (3).7Solvation of anions is generally poor in the aproticsolvents employed here.8p9 This lack of anion stabilis-ation by solvation leads many neutral acids to dissoci-ation in acetonitrile as shown in equation (7).9 TheHA + solvent == (solvent)H+ + AH .. . A- (7)anion A- achieves stabilisation via hydrogen bonding(homoconjugation) to the parent acid. Unfortunately,the extent of homoconjugation has not been measuredfor Hpd in any of the solvents employed here. If itwere complete, the nucleophile in step (3) of the mechan-ism in (2) and (3) would be the homoconjugate speciesHpd . . . pd-. This would lead to the modified rateequation (8). However, the first-order dependence ofRate = K,k[HpdI2[Fe]/[H+]kobs. on [Hpd] for the solvents acetonitrile,l nitromethane(Table 2), and dichloromethane (Table 2) argues againstsignificant homoconjugation in reaction (1).MeN02 oMe2C0-30 - 60AS* / cal K-' rnol-'FIGURE 3 Correlation of AHt with AS: for the reaction ofpentane-2,4-dione with [Fe(C,H,OMe) (CO),] [BF,]Hydrogen-1 n.m.r.measurements show that Hpdexists largely in the enol form in each of the solventsacetonitrile, acetone, and chloroform (Table 6). Theenol : keto ratio in dichloromethane is expected to besimilar to that in chloroform (ca. 83%). These resultsare in keeping with earlier observations lo that enol : ketoratios generally increase with decreasing hydrogenbonding capacity of the solvent. The presence of acidhas no significant effect on the position of the tautomericequilibrium (Table 6).The significance of the observed activation parametersfor reaction (1) in various solvents (Table 3) is difficult toassess in terms of the mechanism in (2) and (3) since theyare composite values, i.e.AHobs.f = AH2* + AH,$ and* Since small amounts of water are reported to greatly increaseconductivity values in nitromethane, i t is not known whetherthese observations represent an intrinsic property of the solvent.Asobs.f = AS2* + AS,I. Nevertheless, it is interestingthat a plot of AHobs.f against AsobsJ for reaction of Hpdwith cation (2) in the various solvents is linear (Figure 3),indicating an isokinetic relation. This observation isconsistent with a commonmechanism with similar roles forthe solvent in each case. Fortuitous linearity seems un-likely in view of the wide range of activation parametersobserved.(However, it should be noted that there areliterature examples l1 of irregular plots of AH1 againstAS1 despite constancy of mechanism in a range ofsolvents.) The similar kobs. values in each solvent areseen to arise from the compensating influences of theenthalpy and entropy contributions, e.g. the loweractivation enthalpies for cations (1) and (2) in nitro-methane compared to acetonitrile are almost exactlybalanced by more negative entropies of activation.Injluence of Added Acid.-A sharp decrease in kobs.is observed in all solvents on addition of acid (Table 4).This behaviour provides strong support for the mechan-ism in (2) and (3), and eliminates processes (5) and (6)which should be acid independent.Quantitative sup-port for the mechanism in (2) and (3) comes from theapproximate linearity of a plot of kobs. against l/[H+] forthe reaction of [Fe(C,H,)(CO)d[BF,] with Hpd in Hf-MeCN (Table 4, Figure 1). A further alternativemechanism [equations (5) and (6a)] in which step (6) isconsidered to be reversible (with equilibrium constantK') is also inconsistent with acid retardation, since thiswould lead to the rate equation (9).The effect of added acid is most dramatic with nitro-methane as solvent, where reaction (1) is effectivelystopped. Conductivity studies suggest that this resultmay arise from the more extensive dissociation of acidsin nitromethane compared with acetonitrile or dichloro-methane, leading to a higher effective free [H+].Theshapes of the curves of A against c) for fluoroboric acid(Figure 2) show that HBF, is a much stronger acid inthe nitromethane employed than in acetonitrile." Thepresent data are consistent with the A,, value of 196.8previously found l2 for HPF, in acetonitrile. Althoughconductivity values for HBF, in dichloromethane werenot measured, it seems reasonable to assume that theacid will be almost completely undissociated in thissolvent. This would account for the fact that pseudo-first-order rate plots are linear for at least 75% com-pletion of reaction (1) in dichloromethane. The earliercurvature generally observed in the more polar solvents(ca. 40% completion of reaction) is expected since HBF,is produced during reaction (1).The possibility that the effect of added acid arises fromthe presence of added electrolytes is eliminated by the* A.J. Parker, Chem. Rev., 1969, 69, 1.lo R. G. Pearson and 0. P. Anderson, Inorg. Chem., 1970,9, 39.l1 H. P. Bennett0 and E. F. Caldin, J . Chem. SOC. (A), 1971,l2 C. J. Janz and S. S. Danyluk, Chem. Rev., 1960,60, 209.S . F. Coetzee, Progr. Phys. Org. Chem., 1967, 4, 45.22071976 2191observation that added sodium tetraphenylborate (0.027mol dm-3) makes no measurable difference to the rateat which Hpd reacts with [Fe(C,H,OMe) (CO)J[BPh,].Efect of Added Base.-General base catalysis isanticipated for a mechanism of the type in (2) and (3).As expected, the addition of 2,6-dimethylpyridine(0.4 mol dm-3) to a solution of Hpd (1.0 mol dm-3) andcation (2) in acetonitrile resulted in extremely rapidformation of a diene product.This reaction is notdue to addition by 2,6-dimethylpyridine, since only aslow reaction is observed between this base and (2) in theabsence of Hpd. A similar acceleration in rate by addedbase was observed for reaction (1) in dichloromethanesolvent, Interestingly, in nitromethane as solvent thereis an instantaneous reaction between (2) and 2,6-dimethylpyridine (in the absence of Hpd). This isthought to arise via proton abstraction from the solventgenerating CH,NO,- which then adds to (2). Such aprocess is only feasible in nitromethane (K, = 6 xmol dm-3 in water l3), since acetonitrile and dichloro-methane are extremely weak acids (K, ca.10-25 mol dm-3in water).Efect of Deuteriated Pentane-2,4-dione.-Experimentswith MeCOCD,COMe in both acetonitrile and dichloro-methane as solvent indicate that there is no primarykinetic-isotope effect for addition to [Fe(C,H,OMe)-(CO),][BF,] (Table 5). This might at first sight beregarded as evidence against the mechanism in (2) and(3) since step (2) involves cleavage of a C-H bond.Replacement of the two methylene hydrogens in Hpdby deuterium should decrease K,. For example, inenolisation of the keto-form of 3-methylpentane-2,4-dione the primary isotope effect in water solvent isk ~ / k ~ = 3.5.14 However, a smaller K, value will beexactly balanced in kobs. by the smaller resulting [H+][equation (4)]. This result is therefore fully consistentwith the mechanism in (2) and (3), although it is alsoanticipated for the alternative in (5) and (6) in whichrapid proton loss follows the rate-determining step.l3 D.J. Cram and G. S. Hammond, ‘ Organic Chemistry,’ 2ndl4 F. A. Long and D. Watson, J . Chem. SOL, 1958, 2019.edn., McGraw-Hill, 1964, p. 214.Conclusions.-All the kinetic data presented here areconsistent with a common mechanism [equations (2) and(3)] for reaction (1) in each of the solvents acetonitrile,nitromethane, acetone, and dichloromethane. Whilemost of the observations are also consistent with analternative path [equations (5) and (6)] involving rate-determining addition of neutral pentane-2,4-dione, thislatter mechanism is eliminated by the general observ-ation of acid retardation. A possible transition state isgiven below which assumes direct addition of the pd-carbanion to the dienyl ring, without prior attachmentI L 6’ -Meto the metal or other sites. Direct addition is supportedby stereochemical evidence 1 5 9 1 6 showing exo-addition ofpd- to cation (2).It is interesting that [Ru(C,H,) (CO)J[BF,] (4) reactsat least 10 times slower with Hpd than does its ironanalogue (1) (Table 3). Even larger rate differenceshave recently been observed l7 for addition of aromaticnucleophiles to (1) and (4). While these differences mayindicate a change in mechanism on varying the metal,they may be simply related to changes in electron densityor free valency at the dienyl rings. INDO molecular-orbital calculations are currently being made to deter-mine whether relative rates can be correlated with suchparameters.We thank the S.R.C. for the award of a studentship (toC. A. M.), and Dr. D. A. Sweigart for many helpful dis-cussions.[5/2494 Received, 22nd December, 19751l5 C. A. Mansfield, R. Davis, and L. A. P. Kane-Maguire, un-A. J. Birch, K. B. Chamberlain, and D. J. Thompson, J.C.S.17 L. A. P. Kane-Maguire and C. A. Mansfield, following paper.published work.Perkin I , 1973, 1900
ISSN:1477-9226
DOI:10.1039/DT9760002187
出版商:RSC
年代:1976
数据来源: RSC