摘要:

1975 277The Uncatalysed and Mercury(i1)- and Thallium(iii)-catalysed Eliminationof Chloride f rorn the p- Amido-p-chloro- bis[tetra-ammi neco balt ( III)]ComplexBy Siew-Wan Foong, Julian D. Edwards, Roger S. Taylor, and A. Geoffrey Sykes,’ Department of InorganicThe kinetics of the mercury(ii)- and thallium(i ti)-catalysed elimination of chloride from the title complex (I),equation (i). have been studied in aqueous HCIO, solution, / = 2 . 0 ~ (NaCIO,). Both reactions are independentand Structural Chemistry, The University, Leeds LS2 9JT( 1 ) (II) (rnIof [H+] in the range 0.5-2.0~. With mercury(ii) there is a less than first-order dependence on [Hg2+] whichcannot be satisfactorily accounted for in terms of a 1 : 1 adduct. Instead the dependence can be explained byconsidering Hg2+-catalysed conversion of (11) into (111) and applying the steady-state approximation for (11).From the treatment (at 25 ‘C), k, = 3.02 x s-l, AMl = 19-2 f 0.5 kcal mol-l, A S l = -5.7 f 1.8 cal K - lmol-l, and k,/k-l = 0.028.At 25 ‘C ratios kJk, = 1.57 x lo3 I mot-’ and k,/k2 = 1.3 x lo3 I mol-l aremeasures of the effectiveness of Hg2+ and HgCI+ as catalysts for (11) --+ (111). Thallium(iii) produces onlya mild catalytic effect, a first-order dependence on [TI3+] is observed, and k,/k, = 15 I mot-l is a measure of theeffectiveness of TI8+ as a catalyst. It is concluded that Ti3+ is a relatively weaker catalyst compared to Hg2+for (11) + (111) than for the catalysed aquation of mononuclear chloro-complexes.X-RAY diffraction and i.r.-spectral studies haveindicated the structure of the p-amido-p-chloro-bis [tetra-amminecobalt (III)] complex (I).Earlier WernerNH(1)assigned the p-aniido-[aquotetra-amminecobalt(~~~)]-[chlorotetra-amminecobalt ( III)] (hereafter aquo -chloro)structure (11) to this complex and this assignment hasL (n)persisted in the literature. Recently pH measure-ments and kinetic studies have indicated that there isretention of (I) rather than (11) as the dominant speciesin aqueous solution, prior to the conversion throughto the p-amido-phydroxo-bis[tetra-amminecobalt(~r~)]complex (I 11). The aquo-chloro complex (11) is formedas an intermediate, but there is no appreciable build-upof (11), and the latter has not as yet been isolated.Rate constants previously ascribed to the conversionof (11) into (111) therefore correspond to the conversionR.Barro, R. E. Marsh, and ’CV. P. Schaefer, Inorg. Chem.,a S.-W. Foong and A. G. Sykes, J.C.S. Dalton, 1974, 1463.A. Werner, Annalen, 1910, 375, 44.31. R. Hyde and A. G. Sykes, J.C.S. Dalton, 1974, 1583.1970, 9, 2131.of (I) into (111) which occurs in a stepwise manner[equations (I) and (2), ammonia ligands omitted]. In-formation regarding k, and k,/k-, has now been obtainedfrom studies on the mercury(I1)-catalysed reaction, anddetails of the uncatalysed reaction are reconsidered inthe light of the data presented. Intermediate formationof the p-amido-bislaquotetra-amminecobalt(1II)j com-plex, (IV), cannot be excluded in reaction (2), and wasCO(NH$~ I” [ INH2‘( N H3ILCo\H20 H20 0(IV)indeed one of the factors which prompted the presentstudy, since in the presence of mercury(I1) chloride is agood leaving group.Species (IV) has previously proveddifficult to identify because of the unfavourable equili-brium constant for formation when (111) is dissolved ine.g. 2lw-HC10,.Studies on the mercury(I1)-catalysed aquation ofchloride from mononuclear cobalt (111) and chromium-(111) ’ complexes have been numerous, and evidence forR. D. Mast, R f . B. Stevenson, and A. G. Sykes, J . Chcm. SOC.(A), 1969, 937.C. Bifano and I<. G. Linck, Iizovg. Chewz., 1968, 7, 908 andreferences therein.J. H. Espenson and J. P. Rirk, Inovg. Chmz., 1965, 4, 527278S N ~ mechanisms has been a d d ~ c e d .~ , ~ Comparisons ofthe relative effectiveness of mercury(I1) and thallium(II1)in catalysing the aquation of cobalt (111) complexes havebeen rnade.l0S1l The kinetics of the mercury@)-catalysed aquations of cis-[Co(en),Cl&+ 6~10 and cis-[Cr(H,O),CLj+ l2 are considered relevant since a less thanfirst-order dependence on mercury@) is observed in bothcases.RESULTSThe UncataZysed Reaction.-Spectra of (I) and (111) givecross-over points at 462 and 535 nm. For the uncatalysedreaction good isosbestic points are observed at 45 "C. A t<25 "C there is some slight movement (ca. 5 nm) of the535 nm isosbestic to lower wavelengths. Possible explan-ations are that either a side reaction is effective or thatsome (11) and/or (IV) is present.Since at 25 "C first-order plots at 554, 500, and 435 nm are linear to 85%completion, for the interconversion of (I) and (111) in bothdirections and rate constants determined at differentwavelengths are in good agreement, we do not consider thiseffect to be of prime importance. Moreover Vz+ has beenused as a sensitive probe for (11), since the latter is expectedto yield [CO(NH,),(H,O)]~+ as the product of the first stageof red~ction,~ and reaction of [Co(NH,),(HZ0)]3+ with V2+would then be observed as a second stage at 490 nm.TABLE 1Pseudo-first-order rate constants kobs for the Hg2+-catalysedelimination of chloride from [(NH,),Coy(NH,,Cl)*Co(NH3)J4+, (I) --t. (111), I = 2*O~(NaC10,)Temp. [H+]OC M-30.035-040.025.0 0.500.601.001-501-501.961.501-601-501.601.601.601.501-601.501.601-601.501.501.601-501.501-50103[Complex]0.660.620.650.470.620.660.671.260.643.381.880-630.670.640.660.660.640.660.630.650-670.660.631M102[Hg2+)1.331.331-331-331.331-332.05.06.010.010.01-332.04.010.01-332.04.010.01-332.04.010.0M104kOb8S-111.011.811-311.611.011.814.820.321.4 a25.223-617.923.430.340.126.733.545.866.087.248.181-2103.3I = 2*oM(Lic104).Chloride salt of complex used: 6%0 Chloride salt of complexSuch experiments indicate no significant formation of[Co(NH3),(H,0)]3+, and at no time during the conversionof (I) into (111) a t 25 "C is there >6% of (11) present.Similar experiments on solutions of (I) in 0.4~-HCl and of(111) in 0*4~-HC1 also failed to give evidence for build-upof (11) during 3 h at 25 OC.The Mercury(Ii)-cataZysed Reaction.-Kinetic runs were8 F.A. Posey and H. Taube, J . Amer. Chem. SOL, 1957, 79,* J. P. Birk and C. M. Ingerman, Inorg. Chem., 1972,11, 3019.of mercury(I1) present as HgCI+.used; 7.6% of mercury(I1) present as HgCl+.225.J.C.S. Daltonfollowed by monitoring absorbance (O.D.) changes at the554 nm maximum for (I) (E = 165 1 mol-l cm-l) with[Hgn] in the range (1.3-10-0) x 10-2~i. At this wave-length (111) has an absorption coefficient E = 128 1 mol-lcm-l. Final absorbances (O.D.,) were measured after7-8 half-lives.No evidence for any build-up of (11) or (IV)was obtained and first-order plots of log (O.D.t - O.D.,)against time were linear to at least 85% completion. Rateconstants Kobs (s-l) evaluated from the gradients of such[Hg2 1 /MFIGURE 1 The variation of hoba for the conversion of (I) into(111) with concentration of mercury(I1) at 26 OC, I = %Ow(NaClO,)0 50 100[ H g / M-lFIGURE 2 The dependence of (kobs - knncat) for the conversion of(I) into (111) on the concentration of mercury(@ as a functionof temperature, I = 2.0hi(NaC104)plots (Table 1) were independent of [H+] in the range0.6-2.0~ (the first acid dissociation constant of HgZf is2.8 x 10-4 mol 1-1 at 25 "C in 3~-NaClO,,l~ and [HgII] istherefore equivalent to [Hg2+]), but gave a less than first-order dependence on [Hgz+f3 (Figure 1).Plots of (k&s -kmcat)-l against [Hg2+]-1 at each temperature are linear(Figure 2) where Kuncat is the rate constant for uncatalysedconversion of (I) into (111).The dependence on mercury(I1) can therefore be sum-10 S.-W. Foong, B. Kipling, and A. G. Sykes, J . Chem. SOC. ( A ) ,l1 S. C. Chan and S. F. Chan, Austral. J . Chem., 1973,26, 1235.12 J. P. Birk, Inorg. Chem., 1970, 9, 735.13 I. ANberg, Acta Chem. Scand., 1962, 16, 887.1971, 1181975marised by equation (3), or by the alternative expression(3)1 1 -- - 1(kobs - Amcat) a[Hg2+] $- 5(4). Values of a and b a t temperatures 25-40 "C deter-(4)mined from (3) using a linear least-squares programme withweighting factor (hob: - kuncat) 2 are listed in Table 2.TABLE 2Summary of a and b values at different temperatures forthe Hg2+-catalysed elimination of chloride from[(NH8),Co*p(NHz,C1)*C~(NH,)~*f, (I) + (111), I =%OM (NaC10,)Temp.10% 103b a"C 1 mol-l s-l S-126.0 12.4 & 0.2 2.94 0.0636.0 26.6 f 0.6 8.26 f 0.2240.0 33.0 & 1.1 16.4 -j= 0.9As defined in (3).Absorbance (O.D.,) values a t 554 nm obtained by extra-polating first-order plots t o the start of the reaction werewithin experimental error as calculated for (I).The effect of small amounts of chloride, [Cl-] < [HgII],was also investigated. Rate constants obtained axe shownin Table 3. Concentrations of HgCl+ were calculated30.0 19-6 & 0.5 4-79 & 0.10TABLE 3The effect of chloride on the mercury(I1)-catalysed elimin-ation of chloride from [ (NH,),Co*p(NH,,Cl) *Co(NH3)J4+a t 25 "C, I = 2.0~(NaClO,)103[C1-] a [H+] 103[Hg2+] 103[HgCl+J lO%obsM M M M S-10 1-60 13.3 0 11-32.6 1.00 10.8 2.36 10.72.6 1.96 10.8 2.39 11.04.6 1.50 9.2 3.69 10-66.7 1-50 7.1 5.61 10.6-0 Total chloride present.b Mercury(I1) not present as HgC1+assuming an equilibrium constant of 0.38 a t 25 "C for (5) .6*7or HgCl,. e Calculated for equilibration (6).2HgCl+ *, Hg2+ + HgCl, (5)Concentrations of HgCl, were less than 5% of the totalmercury(II), and since catalysis by HgCl, is small in othersimilar studies,' contributions from a pathway involvingthis species were assumed to be negligible. From the dataobtained it is concluded that HgCP and Hg2+ are aboutequally effective in catalysing the conversion of (I) into(111) [equations (1) and (2)].Thalliuun(Ir1) -catalysed Reaction.-The catalysis bythallium(m), concentrations in the range (0.9-10.7) x~ O - , M , was also followed at 554 nm, I = %OM (NaC10,).Plots of log (0.D.t - O.D.,) against time were linear to a tleast 75% reaction, where O.D., was the absorbance after7-8 half-lives of reaction. Pseudo-first-order rate con-stants, k'obs, are listed in Table 4.There was no appreciabledependence on the concentration of hydrogen ions, andsince the acid dissociation constant of TP+ is 7 x 10-2mol 1-l a t 25 "C in 3~-NaC10,'~ i t is concluded that T13+and TlOH2+ have about the same catalytic effect.A279TABLE 4Pseudo-first-order rate constants, VObS, for the thallium (111)-catalysed elimination of chloride from [(NH,),Co=(NaC10,)p,(NHz,C1)*C~(NH,)]", (I) + (111), I = 2 . 0 ~ -Temp. [H+lOC M25.0 0.31.01.91.01.61.430.0 1.41.435.0 0.51.41.40.71.41.41.41.440.0 1.41.41.4similar result was1 O4 [Complex]6-066-065-586.196-196.406.226.416.226-416.166.366-38M24.825.824-819.6;6.026.021 O*[TlIII]M0.90.90.95.36.310-75.010.03.33.33.35.05.06.010.010.03.35.010.01 owohe5-10.930.860.801-341-452.102.343.252.892.793.043.613-803-244.624-894.676-337-70obtained in the studies with cis-[C~(en),Cl,]~.~~ A t temperatures in the range 2 6 - 4 0 "Ca linear dependence on [Tlrrl] was observed (Figure 3) and'O trlcI 10.05 0.10[TIn1] /MFIGURE 3 The dependence of k b h for the conversion of (I) into(111) on the concentration of thallium(II1) as a function oftemperature, I = 2-0~(NaC10~from the gradients second-order rate constants, kn, forthe [Tlm]-dependent path were obtained (Table 5 ) .Activation parameters, A H ~ T ~ = 14.8 f 1.2 kcal mol-l,AStT~ = -22-4 f 4 cal K-l mol-I, were determined usinga standard least-squares treatment with weighting accord-ing to the number of data points a t each temperature.Other Studies.-Although a linear dependence on [Hg2+]has been demonstrated for the mercury(I1)-catalysedaquation of [CO(NH,),CI]~+,~ the activation parametershave not previously been determined.The following runswere therefore carried out with [Hg"] = 1-33 x 1 0 - 3 ~ ,[H+] = 1.5w, I = 2 . 0 ~ (NaClO,). Rate constants obtainedwere = 20-2 (25.0 "C), 32.8 (30.1 "C), and 55.0(36.0 "C) 1 mol-l s-l. Activation parameters determined byl4 G. Biedermann, Arkiv. Kemi, 1963,5, 441; Rec. Trav. cham.,1966, 75, 716; T. E. Rogers and G. M. Waind, Trans. FaradaySOC., 1961, 57, 1360280the same procedure as for K H ~ above are AH$ = 18.4 kcalmol-1 and ASS = -0.3 cal K-1 mol-l.DISCUSSIONThe less than first-order dependence on [Hgxl] in themercury(I1)-catalysed conversion of (I) into (111) issimilar to that observed in the Hg2+-catalysed aquationsof the first chloride from cis-[Co(en),ClJ+ 6~10 and cis-[Cr(HzO),Cl.J+.12 The explanation for such a depend-ence in the latter cases was the formation of significantamounts of mercury(Ir)-cobalt(IIx) and -chromium(III)adducts.It is possible therefore that a similar adductis formed in the present study. The dependence on[TlIT in the thallium(II1)-catalysed reaction wasstrictly first order for the range of concentrationsinvestigated.The conversion of (I) into (111) occurs in two stages[equations (1) and (2)]. Both stages, since they involvecleavage of a metal-chloride bond, may be influencedby the presence of Hg2+. We consider possible adductswhich might be formed during the conversion of (I) into(111), and which could therefore give rise to the observeddependence on [Hg2+].If Hg2+ functions as a catalyst for (2) alone, (V) and/or(VI) might be formed. By analogy with the Hg2+-IY) iu,catalysed aquation of [Co(NH,),C1]2+ and other mono-nuclear cobalt (111) complexes containing a singlechloride,6 no build-up of (V) would be anticipated, and aJ.C.S.Daltonimportant, this must be occurring with species (I) andwill have the structure (VIII). Formation of (VIII)would be very rapid, and would constitute a pre-equilibrium step. However the existence of (VIII) isperhaps even less likely than that of (VI) and (VII).There seems no logical reason why (VIII) containing achloride bonded to three metal centres should be stable,when stable adducts are not formed between, forexample, [CO(NH,),C~]~+ and Hg2+.6It is felt that the above reasoning does not favourformation of a Hg2+ adduct.As an alternative explan-ation of the [Hg2+]-dependence we apply steady-statekinetics for (11) in the reaction sequence (1) and (2). Suchan approximation is perfectly feasible since Hg2+ has avery strong catalytic effect on the aquation of chloridefrom [CO(NH,),C~]~+,~ and the catalysed elimination ofchloride from (11) is therefore expected to be rapid.The various equations derived on this basis would seemto provide a satisfactory explanation of the experimentaldata.If the steady-state approximation applies for (11) inthe absence of added Hg2+, the uncatalysed rate constant,kuncat, is given by equation (6). In the presence ofHg2+ an additional reaction, ('i), is included.Thestrictly first-order dependence on [Hg2+] would be overall observed rate constant, kobs, is nouexpected. The adduct (VI) would not be particularlystable either, since water is known to be reluctant tofunction as a bridging ligand.16 Rapid conversion of(VI) into the more stable species (VII) would give rise(WIto an inverse hydrogen-ion dependence. No [H']dependence was observed over the concentration range0.5-2.0~, and it is therefore unlikely that either (VI)or (VII) can be formed at concentrations sufficientlyhigh enough to give rise to a less than first-order depend-ence on catalyst. If the formation of a stable adduct isrk1@2 + MHg2+l)kobs = (k-, + k , + k3[Hg2+])given by (8),(8)and, from (6), (8) can be written in the form (9).Valuesof a and b (Table 2) can accordingly be expressed as inTABLE 5Values of ATl, equation (8), for the thallium(I1r)-catalysedelimination of chloride from [(NH,),Coy(NH,,Cl)*Co(NH,),]4+, (I) ----t (111), I = 2.O~(NaC10,)Temp. 103kT1"C 1 n i o l - ' ~ - ~1.16 0.06 2530 1-99 f 0.0636 2.70 & 0.0840 4.18 & 0.1415 D. L. Toppcn and R. G. Linck, Iitovg. Chein., 1971, 10, 26361975 281(10) and (11). From (6) and (11) k, and k,/k-, can bek1k3k-1 a =(k-1 + k2Yevaluated, (12) and (13). Furthermore a term kngk1= b + kunmt (12)k2Ik-1 = kuncat/b (13)may be defined as klk,l(k-l + k,) which is similar inform to the expression for kumat, (6), and kT1 (see below).This ratio can be obtained from (14).These various(14)functions are listed in Table 6. Activation parametersfor k, and KHg, computed by a least-squares treatmentTABLE 6Summary of data for the uncatalysed and Hg"-catalysedeliminations of chloride from [ (NH,),Coy(NH,,Cl)Co(NH,) J4+, (I) -w (111), I = %OM(N~CIO,)26-0 12.7 & 0.5 0.81 3.02 0-06 0.02830.0 20.1 f 0.9 1-29 4.92 f 0.12 0.02735.0 26-1 & 1.3 2.06 8.47 f 0.20 0.02540.0 33.5 5 3.0 3.40 15.6 & 0.5 0.022kHg = klk3/(k1 + A?). ' /tuncat = k1k2/(k-, + k,) fromreference 6. Calculated using equation (12). d Calculatedusing equation (1 3).of data, weighted according to the number of runs ateach temperature, are given in Table 7.TABLE 7A comparison of rate constants and activation parametersfor the uncatalysed (Kuncat), Hg2'(KHg), HgCl+(kH,l),and TlrI1(kn) catalysed eliminations of chloride [(I) --+(111)] and for the conversion (I) + (11) (K,) at I =2-Ohl(NaC1O4)A H : A SRate constant (25 "C) kcal mol-l cal mol-l K-1= 12.7 x 1 mol-l s-l 11.5 0.8 -24.1 f 2.6k ~ m = 10.6 x lo-, 1 mol-1 s-1TI = 1.16 x 1 mol-l s-l 14.8 &- 1-2 -22.0 4.0knncat = o m x 10-4 S-10 16.2 f 0.6 -23.3 f 2.0 aA, = 3.02 x 10-3 S-1 19.2 f 0.5 -6.7 & 1.8Data from reference 5.A similar treatment to that used above, for the Hg2+catalysis in the presence of free chloride reveals a valueof the function we define as kRsCl = k1k4/(k-l + k2)where k4 is the rate constant for HgCl+-catalysedelimination of chloride from (11).A t 25 "C, the valueobtained for kHgCl = 0.106 1 mol-l s-l is very similar toKQ = 0-127 1 mol-l s-l.Similar relative values of kHgand kHgcl have been noted previously in the Hg2+-catalysed aquation of mononuclear complexes.6J0For the thallium(II1)-catalysed reaction, the rateconstant k,, which is analogous to k,, may be definedand substituted for k, in equation (8). Since kIobsgives a first-order dependence on [TlIII] it is concludedthat (k-l + k,) > k,[TPY. The experimentally deter-mined quantity kT1 is then given by equation (15) andis again of similar form to kancat, equation (6).Although it has not been possible to determine theindividual values of k3, k,, and k,, various ratios,kHg/koncat, kHgCi/kuncatr and kTi/kuncat, provide ~ ~ f o m a t i o nregarding k3/k2, k,/k,, and k5/k2.These values dongwith related (activation) enthalpic and entropic differ-ences between the catalysed and uncatalysed paths arecompared with related data for [Co(NH,),Cll2+ in Table 8.TABLE 8A summary of the effectiveness of HgZt, HgClt, and TPfas catalysts for the elimination of chloride from[Co( NH,),CI] 2+ and [ (NH,),Coy( NH,,CI) -Co(NH,) 4]*+ a t25 "C and I = 2-0M(XaC104)AfAS $\ 5 -,-- I Ratio rate constants A(AHr)Catalyst 1 mol-l kcal mol-l mol-lReaction of [ ( NH3),Co*p( NH,Cl) CO( NH3)4l4+Hg2+HgCl+Reaction of [CO(NH,),C~]~+ CHg2+ kHg/kllncat = 8.7 x 10' - 4.0 + 8.7TP+ kTl/kuncat = 5.2 x lo3 - 3.5 + 6.3k , p , = 1.57 x 103 -4.7 f 1.3 -0.8 f 5Tl3+ k,/k2 = 0.15 x lo2 -1.4 f 1.7 $0.9 f 6k4/k2 = 1.30 x lo3HgClf kHCl/kuncat = 9.3 X lo4a Correspond to differences in activation parameters forcatalysed and uncatalysed reactions.6 This work. C Un-catalysed reaction ; rate constant extrapolated from data forT13f-catalysed reaction, reference 10. Activation parametersfrom A. W. Adamson and F. Basolo, Acia Chenz. S c a d , 1966,9, 1261.In addition to yielding information on the relativecatalytic efficiencies of Hg2+, HgCli, and TlrIr, thisstudy has shed new light on the mechanism of the un-catalysed reaction. Now that (I) rather than (11) isknown to be predominant in solution and k, and k,/k-,have been determined, some reappraisal of the un-catalysed reaction is called for. Rate constants previ-ously reported for the conversion of (11) into (111) infact correspond to the overall reaction (I) into (111),and to a first approximation are given by klk,/k-l (since(Table 6) as compared to the experimentally observedrate constant of 0.81 x Activation para-meters for the latter case are AH$ = 16.2 & 0.5 kcalmol-1, A S = -23.3 & 2.0 cal K-l mol-l.It is esti-mated that k,/k-, < 0.1,4 which means that for theintramolecular conversion (11) into (111) at 25 "C andI = %OM (NaClO,), k, > 8.4 x s-l. This is atleast 300 times faster than the intermolecular aquationof chloride from [Co(NH3),C1I2+, k = 2.3 x lo* s-1,I = %OM (LiCIO,).lo Ligand substitution of mono-nuclear l6 and bridge cleavage and formation processesin binuclear l7 complexes of cobalt(II1) are believed tobe S N 1 .It is concluded therefore that there is somel6 A. Haim, Inorg. Chem., 1970, 9, 426.l7 R. S. Taylor and A. G. Sykes, Inovg. Chem., 1974, 13, 2624.k-l >> k2). At 25 "C, I = 2*0M, klk2/k-1 = 0.84 x lo4 s - 282 J.C.S. Daltonlabilisation of the chloride and aquo-ligands in (11)which can be attributed to the presence of the amido-bridge. Previously, from an examination of proton-ation constants for bridging ligands, e.g. oxalate andph~sphate,l**~~ it has been concluded that two cobalt(II1)atoms bonded to the same ligand have approximatelythe same effect as a single proton. The amido-bridgemight therefore be regarded as exhibiting behaviourintermediate between those of NH,- and NH,. At theSame time it must be noted that the NH,-bridge has nofree lone-pairs of electrons.The labilisation of ammonialigands by the hydroxo-bridge in (111) has been detectedin earlier work.2o The observation that there is nobuild-up of the bisaquo-complex (IV) when (111) ismade up in ~ M - H C ~ O , ~ ~ may therefore be attributableto the labilisation of the aquo-ligands by the amido-bridge. Further discussion of these effects is to befound in ref. 22. No evidence was obtained for build-upof (IV) in the present study, nor is formation of (IV)evident in the uncatalysed reaction. Its formationfrom (11) would require intermolecular substitution ofC1- by solvent H,O. Intramolecular processes aregenerally preferred in this type of reaction, even thoughin the conversion of (11) into (111) a transitory p-aquo-species must be formed.The relative catalytic effects of Hg2+, HgCl+, andT13+ are much smaller in the conversion of (I) into(111) than in the aquation of [Co(NH,),C1]2+ (Table 8).However the order of effectiveness is the same Hg2+ -HgC1+ > TP+.Chan and Chad1 have recently observed that second-order rate constants for Hg2+ and TP+-catalysedaquations of mononuclear complexes containing a singlechloride give enthalpies of activation, AH$, which fora particular complex are the same within experimentalerror, and conclude that differences in rate are residentin the entropy term.Activation parameters have nowbeen determined for the Hg,+-catalysed aquation of[Co(NH3),C1]2+ and the same point may be made(Table 8).Interesting though this observation isfurther comment is called for because second-orderrate constants are in fact composite terms, and catalysispresumably occurs by a stepwise mechanism [equations(16) and (17)]. According to this mechanism observedK (rapid)ColII-C1+ catalyst @ CouI-C1-catalyst (16)(17)kCoIII-C1-catalyst + CoIII + C1-catalystsecond-order rate constants correspond to KK, andactivation enthalpies contain contributions from boththe pre-equilibrium (16) and the rate-determining step(17). It is not possible to comment on the values ofAH for (16) directly. However enthalpy changes forthe formation of HgCl+ (-5.8 kcal mol-l) and T1C12+I* K. L. Scott, M. Green, and 4 . G. Sykes, J. Chem. SOC. ( A ) ,1971, 3651.19 J.D. Edwards, S.-W. Foong, and A. G. Sykes, J.C.S.Dalton, 1973, 829.2o R. S. Taylor and A. G. Sykes, J . Chern. SOC. (A), 1971, 1426.21 A. G. Sykes and R. S. Taylor, J . Chem. SOC. (A), 1970, 1424.(-5.45 kcal mol-l) 24 at 3 . 0 ~ ionic strength are about thesame. For the overall enthalpies to be identical, AH$values for reaction (17) with Hg2+ and TP+ as catalystsmust also be the same. This is perhaps reasonablesince (17) corresponds to the cleavage of a cobalt(rI1)-chloride bond and may therefore be independent of theidentity of the catalyst. However for the binuclearcomplexes enthalpies of activation are different forHg2+ and TP+. The unfavourable enthalpy of activationfor TP+ is reflected in a much reduced catalytic effectcompared to Hg2+.Previously10 we expressed thebelief that two factors explained the effectiveness ofcatalysts, HgCl+ - Hg2+ > TP+ > T1Cl2+. The firstwas the affinity of the catalyst for chloride and thesecond was the charge on the catalyst. The results ofthis study would seem to reinforce this belief in thatwhile Hg2+ and TP+ have similar affinities for chloride(viz. similar equilibrium constants for formation ofHgC1+ and T1C12+), Hg2+ is a more effective catalyst.For a comparison of effects in mononuclear and bi-nuclear complexes (Table 8) it is perhaps more reasonableto consider the charge product in the activated complexrather than the charge on the catalyst. With thisconsideration in mind, the decrease in efficiency ofHg2+, and the much more marked decrease with TP+for binuclear as opposed to mononuclear complexes canbe rationalised.EXPERIMENTALSamples of the perchlorate salts of the complexes pamido-y-chloro-bis[tetra-amminecobalt (III)], [ (NH,)4Co*p(NH,,C1)*Co(NH,)J (C104),,H20, and p-amido-p-hydroxo-bis[tetra-(C104)4,H@, were prepared as described previously.26Solutions of mercury(1r) perchlorate were obtained in thefollowing manner.Mercury(1r) perchlorate (50 g, G. F.Smith) was dissolved in O.S~-perchloric acid solution(100 ml). The solution was exchanged onto an AmberliteIR-l20(H) resin column (35 cm, 2-cm diam.). The columnwas washed with water until the pH was ca. 5.5, washedwith O.OS~-perchloric acid to remove any 1+ ions, andfinally the Hg2+ was eluted with ZM-perchloric acid.Theconcentration of mercury(I1) was determined by titrationwith a freshly prepared standard 0.h-ammonium thio-cyanate solution using ferric alum as indicator. Thehydrogen ion concentration was determined after removalof the Hg2+ using the same cation-exchange resin.Thallium(I1r) perchlorate hexahydrate (G. F. Smith) ishygroscopic and was dried and stored in a desiccator overP,O,. A weighed amount of the solid was dissolved in5M-perchloric acid (10 ml), and diluted to 25 ml withdistilled water. It was necessary initially to use 5M-perchloric acid to ensure that all the solid was dissolved.Subsequent dilutions were carried out carefully withmechanical stirring to avoid formation of insoluble Tl,O,in regions of low acidity. The concentrated solutions were22 J. D. Edwards, K. Wieghardt, and A. G. Sykes, J.C.S.Dalton, t o be published.33 R. Arnek, Arkiv. Kemi, 1965, 24, 831.24 M. J. M. Woods, P. K. Gallagher, 2.2. Hugus, jun., and E. L.King, Inorg. Chem., 1964, 3, 1313.25 S.-W. Foong, R. D. Mast, M. B. Stevenson, and A. G. Sykes,J . Chem. SOC. (A), 1971, 1266.amminecobalt( III)] , [ (NH,)&3-p(NH2,OH) *CO (NH3)4]1975standaxdised by addition of an excess of potassium iodideand titration with sodium thiosulphate. To prevent aerialoxidation of the potassium iodide in acidic solutions, solidsodium carbonate was added to the thallium(II1) solutionto displace air by carbon dioxide.AnalaR grade perchloric acid (72 %), sodium perchlorate(B.D.H.), and sodium chloride were used as requiredwithout further purification. The ionic strength in thepresent study was adjusted with NaClO,, which haspreviously been shown to give satisfactory agreement withruns in which LiC10, was used.25 The sodium perchloratewas shown to contain 0.02% of chloride, by titration withsilver nitrate. Such an amount is negligible as far as thisstudy is concerned.S. W. F. is grateful to the University of Leeds and J. D. E.to S.R.C. for post-graduate studentships.[2/2810 Received, 14th December, 1972

ISSN:1477-9226    

DOI:10.1039/DT9750000277

出版商:RSC    

年代:1975

数据来源: RSC