1972 715Chloride-catalysed Oxidation of Thallium(i) by Manganese(ii1) in AqueousPerchlorateBy D. R. Rosseinsky * and R. J. Hill, Department of Chemistry, The University, Exeter EX4 4QDManganese(ll1). in contrast with cobalt(iri), does not react with thallium(i), except when manganese dioxide.from decomposition, or added chloride, are present. The observed rates in the presence of chloride are consistentwith ' reversible ' reaction of MnCI2+ with CI- to give CI2-, which oxidises TI1 slowly and TIIL rapidly. All T Pappears as TIC14-. Curvature of rate plots at >70--80% reaction may be due to the presence of unreactive poly-nuclear MnTI1, in perhaps <lo% amounts. With an assumed association constant for MnCI2+ of 13.5 I mol-l, therate constant for MnCI2+ + CI- is 1 .I 6 I mot-' s-l.MANGANESE(III) is oxidised by mechanisms differentfrom those of cobalt(II1) when the latter is the fasteroxidant.With cobalt(II1)-mercury(1) CoIII + Hg;+ israte deterrnining,l whereas the manganese(II1)-mercury-(I) reaction involves a multiple-sequence mechanism.2By contrast, in the mechanistically identical oxidationsof iron(11) ,394 manganese(II1) is oxidised faster thancobalt (111). Since the oxidation of Hg22+ involves two-electron transfer it was of interest to investigate anothertwo-electron reductant, thallium(I), with which thecobalt(m) reaction is CoIII + TP. Contrary to ex-pectation, manganese(II1) was found to be unreactivewith pure thallium(1) in ca. ~M-HC~O,, even at 80 "C.This observation, confirming that of Davies,6 holds despitethe favourable 25 "C standard electrode potentials E"of 1.56 V for manganese(III)-(II) and 1.23 V for thal-lium(III)-(I) ; this emphasises the contrast with co-balt(rr1).Oxidation of thallium(1) was found to proceedat an observable rate either after prolonged heatinghad resulted in visible decomposition to MnO, (seeExperimental section) or when chloride was present.Addition of chloride promoted reaction at measurablerates at 25 "C. A comprehensive kinetic investigationwas undertaken as follows, the manganese@) beingfollowed by amperometric titration with iron@).EXPERIMENTALThe manganese(I1r) (2-4 x 10-3~) was prepared by slowelectrolysis of manganese(r1) on platinum in 6~-HC10, atlow (ca. 10 pA ern-,) current density, in a largely vainattempt to preclude polynuclear manganese(II1). Theevidence for polynuclear manganese(II1) is tenuous butpersistent and such species may exceed 10% of totaloxidant.All kinetic samples were ' aged ', i.e. at approxi-mate equilibrium with respect to polynuclear species, toprevent ageing taking place during sometimes protractedruns. Manganese(II1) reaction samples with acidity 4~were allowed to stand prior to reaction for periods in therange 3-36 h, with no kinetic consequences.against iron(II), andkinetic determinations of manganese(m), we used a cellwith a rotating platinum electrode, a perchlorate bridge,and a 1 O-2hI-chloride calomel reference electrode. Theiron(r1) was run into the MnIII solution, the iron(I1) diffusioncurrent being measured on either a Sangamo-WestonD.R. Rosseinsky and W. C. E. Higginson, J . Chem. Soc.,1960, 31.D. R. Rosseinsky, J . Chem. SOL, 1963, 1181.L. E. Bennett and J. C. Sheppard, J . Phys. Chem., 1962,For amperometric standardisation66, 1275.microammeter, for the higher currents, or a calibrated110 Scalamp galvanometer, for the lower. Manganese(I1)(from G. Frederick Smith) was standardised by EDTAtitration, as was the excess of manganese(I1) stabilising(with respect to disproportionation 2) the manganese(II1) ,after reduction of the latter with hydroxylamine. Thallouscarbonate (B.D.H.) was dissolved in HClO, and the TlClO,thus obtained was several times recrystallised from water.The solution was standardised against KBrO, in HC1.Thallic perchlorate (G.Frederick Smith) in solution,was determined, after SO, reduction, with bromate, ordirectly by oxidation of iodide and thiosulphate titrationof the iodine thus liberated. Recrystallised AnalaRNaClO,, was determined as solute by evaporation to drynessand weighing of the residue. AnalaR HC10, and NaClwere employed, and the water was twice distilled, thesecond time from permanganate.Kinetics samples were cooled to 0 "C to quench reaction.The analysis was tested at 0 "C over the range 1.5 x5 x 1 0 - 6 ~ in manganese(II1) in the presence of all speciesconstituting the reaction solution; in all cases the deter-mination was within 1% of the nominal. Negligible re-action was shown to occur at 0 "C in 20 min, well in excessof the few minutes required for titration.No loss of MnlI1was detected in 12 h at 25 "C either in a solution of 5 x lo-,-MnIII, O-lM-MnII, 0=05~-Cl-, and ~M-HC~O,, or in one of5 x 10-4~-Mn111, O.lh.r-MnIr, O.O~M-T~I, and ~M-HC~O,.The kinetic experiments were conducted at ionic strength3 . 4 5 ~ at 25 f 0.02 "C. It was shown that rates were in-dependent of whether Tll or C1- were added to initiatereaction.The stoicheiometry of the chloride-catalysed reaction,and also that of MnO, with TlI, were established by appro-priate analyses. The MnO, (B.D.H.) was dried at 150 "Cfor 2 h. Known mmol amounts were added to Tll in3~-HC10, and after agitation of the suspension for 4-6 h atSO "C, residual Tll was determined with bromate.Thestoicheiometry (MnO, consumed)/(Tlx consumed) was1.05 4 0.02 (despite the small E" difference at 25 "C, 1.25and 1.23 V for MnIVJ1 and T1*I1J).RESULTS AND DISCUSSIONThe stoicheiometry in the presence of chloride, MnIrrconsumed per TlIII produced, was 2.03-2.08, which,barring the unlikely coincidence of identical losses ofTlIII and MnIII by reaction with solvent, confirms theD. R. Rosseinsky and M. J. Nicol, Trans. Faraduy Soc.,I<. G. Ashurst and W. C . E. Higginson, J . Chem. SOL, 1956,G. Davies, personal communication.1968, 64, 2410.343716 J.C.S. DaltonThis function has been plotted for 10-fold changes in[MnIIqo (and so ca. 100-fold in [MnIII]), 50-fold in TP,12-fold in C1-, and in 1--3~-HC10,.Four separatestock solutions of manganese(II1) were employed. Theplots for three of these stock solutions were linear toca. 70 or SOYo reaction, but for one the curvature startedsooner, usually between 65 and 70%. This deviation isdiscussed below, but we emphasise the linearity of themajor part of the runs, as illustrated in the Figure fortwo runs of virtually identical composition, but onecomprising the more deviant stock manganese(II1). InTable 1, when k' (depending on [TlI] and [MnII]) isconstant, then klK(l - k') from the linear slope remainsreaction to be almost exclusively 2Mn111 + TP+2MnII + TPII. Different determinations showed MnIIIconsumed per T1I consumed to be 2 & 0.04 indicatingno side-reaction of MnIII. Free chlorine was neverdetect able.First, initial rates were examined in the form of slopesof log [MnIII] against time, t.In nearly all runs [MnIII]was the only variable, all other participants being insufficient excess. (Two runs do not quite comply,[TlI] being slightly low, but deviations of the rateconstants were insignificant .) The initial data showedan order close to 2 in C1-, ca. 1 in MnIII, <1 in TP, and alow, apparently inverse, order in MnII. Some of thesefeatures resemble those in the chloride-catalysedoxidation of thallium(1) by cerium(~v),~ in which theeffective oxidant is C1,-. Following Duke and Borchers'scheme for cerium(1v) as oxidant, we tested the follow-ing mechanism. Before elaboration, the roles of thechloro-complexes of TlIII have to be reckoned.Thespecies 8 9 9 TlCk(3-n)+, with n = 1 4 and possibly 6,are established. However, when the ratio [Cl-]/[TPII] >4.5 the stability constants are such that virtually all ofthe TllI1 is TlC14-, which simplifies our task enormously.We have accordingly arranged for the ratio always tobe >4*5 (usually >4.5).The mechanism to be tested is :KMn3+ + C1- 'T- MnCP+ (rapid)MnCP+ + C1- -% Mn2+ + C1,- rTll + C1,- _k"_ TFI + 2C1-TPI + C1,- TlIII + 2Cl-TllI1 + 4C1- + TlC14- (rapid and virtually complete)Here [CI-1, = [CI-1, - 4[TPITIt. Participation of TlIIin a variety of reactions has been listed by Benson.10The stationary state assumption for this and the otherintermediate, C12-, gives the rate law (1). [MnII] andd[MnlII] ---dt -[TlI], but not [Cl-1, being constant in each experiment, wewrite the parenthesised factor as (1 - k'), and obtaindxldt = k"x(a + 2 ~ ) ~ by putting [TPII], = 0, [Mn1I1Ib =x , and [Cl-1, - 2[Mn1II], = a, with k" = klK(l - k').Hence on integration :([Cl-1, - 2[Mn111],)-2 In [MnIII]-l( [Cl-1, +2[MnIII] - 2[Mn11f],) - (([Cl-1, + 2[MnIII] -2[MnIr1],) ([Cl-1, - 2[Mn1Ill0)>-l = k"tF.R. Duke and C. E. Borchers, J . Amer. Chem. SOC., 1963,* M. J. M. Woods, P. K. Gallagher, 2. 2. Hugus, jun., and75, 6186.E. L. King, Inorg. Chem., 1964, 8, 1313.I 09/'.0/a a,#f - / B+3001 I I I0 100t l m i nPlots of function F of the integrated rate equation against time,t, for the two different stock solutions indicated in Table 1.F is the function ([Cl-] - 2[Mn111]o)-1 In [MnII1]-l([Cl-]o + 2[Mn111] - 2[Mn1I1J0) - ([Cl-1, + 2[Mn111] - 2[Mn1I1l0).Linearity: A, 80% reaction; B, 60% reactionsatisfactorily constant with extensive change of [MnIII]and [Cl-1, underlining the close correspondence of theobservations to the assumed mechanism.Values ofk,K(l - k') are given in the second column of the Table.Where expected (i.e. at fixed [HCIO,]) they are constant,to ca. 10% a t worst and to 3% on average.1 - k' = (2k,/k2)/((2k3/k2) + [Mn2+]/[T11]],several guesses of (2k3/k2) were made and the constancyof klK checked, for reactions with differing [TlI] and[MnII], for each choice. The result is shown in Table 2,where 2k,/k2 = 30 & 5 is clearly indicated. This valuewas used to obtain the quoted klK results in the thirdcolumn of Table 1.Using an estimate 11*12 of K =13-5 & 0.5 1 mol-I, from a fit of rate data for the MnII-catalysed oxidation of oxalic acid by bromine in thepresence of chloride (ionic strength 2 ~ , 25 "C), oneobtains, for k,, 1-16 & 0.05 1 mol-l s-l, the standarddeviation of the 36 individual values (from Table 1) beingquoted.SinceT. G. Spiro, Inorg. Chem., 1965, 4, 731.tion,' McGraw-Hill, London, 1968, pp. 110, 126.lo D. Benson, 'Mechanisms of Inorganic Reactions in Solu-l1 H. Taube, J . Amer. Chem. SOC., 1948, 70, 1216, 3928.l2 G. Davies and K. Kustin, Inorg. Chem., 1969, 8, 11961972 717The deviations from linearity in the rate plots, perhapsascribable to error in MnIII analyses (systematic over-estimates of 10-20(30 accounting for most of thedeviation) are more likely the result of slight, irreversiblepolymer-MnIII formation since there is the clear differ-ence in extents of linearity between different stocks.The requisite unreactive nature of the polymer and itspersistence is possibly a consequence of resonance, forterms of total oxidant, [MnIII].There is no simpleexplanation for this, since the extensive hydrolysis ofMnIII gives appreciable amounts of both MnS+ andMnOH2+. If only the former reacts, the apparent orderin [H+] would be 0-5-0.6, just less than the valuequoted. The only other hydrolysed species is TFIIwhich, however, as TlCl,-, would be essentially un-affected by the [H+] changes.Further marginal effectsTABLE 1Variation of rate constants (I2 mol-2 s-l) with composition, for four stock manganese(II1) solutionsComposition *As footnote *t1 mM-MnlI1, 1 OmM-C1-50m~-TlI6m~-TlI3rn~-TlIlmM-Tll10rnM-MnIIZM-HClO,IM-HClO,As footnote *7.5mM-Cl-7.5mM-Cl-7.5m~-C1-, 5rn~-TlI10mM-Cl-50rn~-Cl-, 10mM-TllI12mM-Cl-7*5mM-C1-, 2M-H+7,5m~-Cl-, 1.6~-H+As footnote *1mM-Mn1I17-5m~-Cl-, 1mM-MnIZ11 OmM-C1-45m~-Cl-, 10mM-Tllllk,K(1 - k')12.312.614.49.727.223-838.335.2814.6For ~M-HC~O,, mean:12-511.812.111.712.911-39.06.44For ~M-HC~O,, mean:13.111.112.111.112.88.55Mean :k,K16.416.715.415.115.216.615.111.115.816.715-716.114-315-617.215.112.08.615.617.414.816-114.817.016-07.04Composition *As footnote *As footnote *?O.lmM-MnIIIIrnM-MnIII0.25m~-Mn~I~2rn~-MnI~I1 OmM-C1-7.5rnM-Cl-7.5mM-Cl-7.5mM-Cl-3.75mM-Cl-7.5mM-C1-, ImM-TlI7.5mM-Cl-, 3m~-TlI6m~-TlI12rn~-MnII20m~-MnII200rn~-Mn~~klK(1 - R') k,K10.6 14.113.3 17.811.4 15.211.8 15.711.9 15.911.7 15.611.7 15.611.4 15.212.2 16.311.1 14.811.1 14.83.78 16-46.39 13.59-31 14.514.7 15.312.8 13.69.72 16.2Mean: 15.3For 3M-HC104, mean for all four stock MnIn solutions(& av.dev.): k,K = 16-61 & 0.17* 0.50m~-Initial MnIII, 5m~-Cl-, 100mM-MnI1, 10mM-TlI. ~M-HC~O,, unless otherwise stated : 25 "C, 3 . 4 5 ~ ionic strength ;(1 - k') as from Table 2.7 Plotted in Figure.example MnIII-MnIII .f--)- MnIr-MnIV, all three oxidationstates being viable in solution and hence of accessibleenergy in the indicated wavefunction combination of themulti-ion species. We do not pursue the kinetic com-plication quantitatively both because of its irrepro-ducibility and also because of the further possibility ofTABLE 2Relative sums of deviations from mean K,K for variouschoices of 2k,/k, *as activity coefficient changes with composition (possiblymore marked with both cationic and anionic reactants),or slight increases in amount of unreactive polymer atlower acidity, also require consideration. We do not,however, consider that the [H+] dependence casts anydoubt on the mechanism favoured.The assumption implicit hitherto, that only a smallfraction of MnIII is present as MnC12+, is not quite exact,introducing a maximum error in one case of 7% and anaverage error of 3%.The chloro-complexes of Tll and2ksIk2 20 25 30 35 40Residuals 6-25 5-60 5.50 5.80 6.25Chosen 2 k , / k , 30 & 5differ from the values given in the footnote.* From the 12 reactions in Table 1 where [MnII] and [TlI]the formation of mixed TP-TPII chloro-complexesduring the run. These are known in the solid l3 but arenot detected in solution by Raman spectroscopy.14They may still be present at concentrations < l O - z ~ ,which may be sufficient to affect kinetic results.The order of reaction in [H+] (Table l), from 1~ to3 ~ , is apparently 0.7-0.8 if the rate law is written in13 H.Remy, ' Treatise on Inorganic Chemistry,' vol. I,l4 A. McAuley, Co-ordination Chem. Rev., 1970, 5, 245.Elsevier, London, 1956, p. 385.D DMnII may be assumed weak, and so neglected.The species C1,- has been invoked in the reaction ofH,O, with 0, in chloride solution, and it presumablyarises also in the MnII-catalysed oxidation of oxalic acidby chlorine,ll as well as in the CeIV-TF-C1- reaction. Thehalogen species X2- is now quite widely established asa reactive intermediate in redox reactions.l4Our mechanism differs from that for CeIV-TP-C1- inthat the rate for the latter is independent of [TlI], andseveral Cerl chloro-complexes are involved ; the ratelaws thus differ.Silver(1) catalyses the (chloride free) oxidation ofthallium(1) by cerium(1v),15 but not by manganese(m),l5 W. C. E. Higginson, D. R. Rosseinsky, J. B. Stead, andA. G. Sykes, Discuss. Faraday SOC., 1960, 29, 49718 J.C.S. Daltoneven in 0 - 1 ~ concentration ; our observation accordswith that of Davies6 The difference does not depend onoxidation strength, E" values being 1.56 V (MnIII911)and 1.5 V (CerVJI1) .16 Further trial catalysts for the MnIIIreaction, CoII and Curl in the absence of C1-, were alsowithout effect.The reaction of CoIII with TF in the presence ofl6 L. G. SillCn and A. E. Martell, ' Stability Constants,' Clzem.Soc. Special Publ. No. 17, 1964.chloride can be confidently predicted to be fast sinccobalt(II1) reacts rapidly with C1- to give Cl,, via Cl,-.lThis result would not conform with the MnIII-CoIIpattern which we have remarked; it is noteworthy thaa clear first order in H+ is asserted for the CoIII-Cl-reaction.li[1/1267 Rcccived, 23vd July, 1971l7 B. SrArnkovA, J. Zjlka, and J. Doleial, J . ElectvoanalytCkem. Interfacial Electvochem., 1971, 30, 185