March 20, 1964
KINETICSOF REDUCTION OF SOMECOBALT(III) COMPLEXES
[CONTRIBUTIONF R O M
THE
DEPARTMENT O F CHEMISTRY, UNIVERSITY
OF
1019
CHICAGO, CHICAGO 37, ILLISOIS]
Kinetics of the Reduction of Some Cobalt(II1) Complexes by Chromium(II), Vanadium(II), and Europium(I1) BY JOHN P. CANDLIN, JACK H A L P E R NA,N ~D DAVIDL. TRIMM RECEIVED OCTOBER 28, 1963 Kinetic measurements are reported on the reduction of a variety of pentaamminecobalt(II1) complexes by CrZtag, \‘2+aq, Eu2-,,, and Cr(bipy)32’ (bipy = 2,2’-bipyridine). The resulting reactivity patterns are examined from the standpoint of providing a possible basis for distinguishing between inner- and outer-sphere electron-transfer mechanisms. The variation in rate, with V 2 + as reductant, was found t o be much smaller than with Cr2* and to exhibit sufficient similarity t o the variation in rate observed with Cr(bipy)a2+(as well as with the rates of electrochemical reduction of the corresponding cobalt(II1) complexes a t a dropping mercury electrode) t o suggest t h a t both V 2 + and Cr( bipy)g2+ react through outer-sphere mechanisms. These correlations also contribute t o the confirmation of an important prediction of the Marcus theory of electron transfer. The reactivity of Eu*+toward the halopentaamminecobalt(111) complexes was found, in contrast t o the other reductants, to decrease in the order Co(NH3)jF2+> Co(NH3)jC12t> C O ( ~ Y H , ) ~> BC ~O ~ +( N H ~ ) ~ I ~ - , a trend which is interpreted in terms of an inner-sphere mechanism. The rates of reaction of Cr2+,V 2 + , and E u Z *with Co( SHZ)bP04 all exhibited a marked p H dependence which is interpreted in terms of the protonation equilibria of the complex.
Introduction
The present investigation was undertaken with a view t o further examining and extending this approach. Previous investigations, notably by Taube and his A limitation on such studies is imposed by the fact co-workers, have revealed t h a t the reduction of cobaltthat many of the reactions of interest are too fast for (111) complexes of the type C O ( N H ~ ) ~ X(where ~-~ conventional kinetic measurements. I n many cases X-” may be any one of a large number of ligands, it was found, however, t h a t the rates could be cone.g., H20, C1-, Br-, N3-, SCN-, Sod2-, etc.) may proveniently measured with a stopped-flow apparatus ceed either by outer- or inner-sphere m e ~ h a n i s m , ~ - ~and most of the results reported below were obtained with X acting as an inner-sphere bridge for electron in this way. transfer, in the latter case. With Crs+aq or Co(CN)b3- as reductant, evidence for an inner-sphere Experimental mechanism was deduced, in a number of cases, from the Materials.-Solutions of chromium( 11) perchlorate were preobservation t h a t electron transfer is accompanied by pared from the corresponding chromium( 111) salt, both electrothe transfer of X to the coordination shell of the newly lytically and by reduction with zinc amalgam, the two procedures giving consistent results. Chromium(II1) perchlorate was formed (and substitution-inert) CrIII(H20)bX or CoIIIprepared by reduction of an acidified solution of sodium dichro(CNjbX ion, r e s p e ~ t i v e l y . ~ ,With ~ C r ( b i p ~ ) 3 ~as + mate with hydrogen peroxide.1° Solutions of Cr(bipy)aClOd reductant on the other hand, it has been concluded’ were prepared by addition of aqueous chromium(I1) perthat electron transfer proceeds through an outerchlorate t o a solution containing an excess of bipy. Vanadium(11) perchlorate was prepared by reduction of a solution sphere mechanism to yield C r ( b i ~ y ) , ~ + . of vanadium(1V) perchlorate with zinc amalgam. EuroFor many other reductants, for example the aquo pium( 11) perchlorate was prepared by dissolving europium( 11) ions V 2 + and E u 2 + , which yield substitution-labile carbonatell in aqueous perchloric acid. Sodium perchlorate was prepared by neutralizing reagent grade perchloric acid with soproducts on oxidation, the mechanistic criteria which dium carbonate. Other chemicals were of reagent grade. have been invoked in the above cases are inapplicable The chromium(I1) and europium( 11) solutions were analyzed and it becomes necessary to invoke other, less direct, by adding an aliquot of the solution t o a deoxygenated iron(II1) criteria to infer the mechanism of electron transfer. solution and back-titrating the iron( 11) produced with chromium( Y I ) , Vanadium( 11)was titrated with deoxygenated potasOne such approach is through the study of the desium permanganate in the presence of sulfuric acid . l o C r ( b i ~ y ) ~ ~ + pendence of the rate for a given reductant and a series was determined spectrophotometrically ( E S ~ ? . &n,p 4.34 X l o 3 of oxidants CoIII(NH3)bX, in which the ligand X is Jw-1, e467 m p 3.95 x 103 A L - ~ ) . varied, on the nature of the latter. It is reasonable All the cobalt( 111) complexes used as oxidants were prepared using previously reported procedures. ROH2( C101)3 (where to expect t h a t the pattern of this dependence will differ R = C O ( N H , ) ~was ) generated by treating a solution of RCO3as a consequence of whether X acts as an inner-sphere (NO,) with perchloric acid.12 RCI( CI04)2,l3 RBr(C104)?,14 bridging ligand or whether the reactions proceed by an RF(ClO4)2,’2 RNOa( CI04k;” RS04(CIO4),l5R S C S ( C104)?,l6RS2outer-sphere mechanism. I n the latter case, it may 03(clo4),17ROAc( C101)2, and R(H-maleato)(C104)r,’*were prepared by standard procedures, described in the references cited. further be anticipated on the basis of theoretical Separation from residual aquo complex was effected either by ion considerations formulated by Marcus8 that, providing exchange or recrystallization, and the identity and purity of the the electrostatic work terms are similar (or sufficiently products confirmed analytically and spectrophotometrically. small to be neglected), the reactivity pattern toward a RI(CIOa)zwas prepared by the method of Haim and Taubelgwith care being taken t o minimize product hydrolysis. R S z ( Clod)? series of such oxidants will be the same for different was prepared in the dark by the method of Linhard and F l y g a ~ e . ~ ~ reductants. This criterion was invoked by Zwickel Both the latter complexes were found to he light-sensitive and and Taubeg to infer, on the basis of the rather limited their reactions were studied in blackened vessels. RPOI was data hitherto available, that the reduction of CoV 2 + (in contrast (NH3)50H23+and C O ( N H ~ ) ~ Cby I~+ (10) A . I. Vogel, “Quantitative Inorganic Analysis,” John 1Viley and Sons, h7ew York, N . Y , 1961. to C r 2+) proceeds through an outer-sphere mechanism. (1) T h i s research was supported by t h e h’ational Science Foundation through Grant GP654 ( 2 ) Alfred P S l o a n Foundation Research Fellow. (3) &I T a u b e , A d v o n Inorg Chem Radiochem., 1, 1 (1959) ( 4 ) J. Halpern, Quart Reu. (London), 16, 207 (1961). (.5i R T . M. Fraser, Reo Pure A p p ! . C h e m . , 11, 64 (1961). (6) J P Candlin, J. Halpern, and S. N a k a m u r a , J A m . Chem. S O L . ,86, 2517 (1963) ( 7 ) A hl. Zwickel and H T a u b e , D ~ s c u s s i o n sF o r a d a y Soc., 29, 42 (1960). ( 8 ) R A Marcus, J . P h y s . Chem., 67, 8R3 (1963). (9) A Zwickel and H . T a u b e . J . A m Chem. Soc., 83, 793 (1961).
(11) R . A. Cooley and D . M. Yost, Inorg. S y n , 2 , 69 (1946) (12) F. Basolo and R K M u r m a n n , ibid.. 4, 171 (1953) (13) W. A. Hynes, L. K. Yanowski, and M. Shuller, J . A m Chem .So< 60, 3053 (1938). (14) H . Diehl, H Clark, a n d H . H. Willards, I n o r g S y n , 1, 186 (1939) (15) S. M. Jorgensen, J . Prakl Chem., 31, 268 (1886) (16) A Werner and H Muller, Z nnorg a i ! g e m Chem , 2 2 , 102 (1900). (17) P. R . R a y , J . Indian Chem. Soc., 4 , 64 (1927) (18) D . K . Sebera, P h . D . Dissertation, University of Chicago, 1960 (19) A Haim and H T a u b e , J A m Chem Soc , 8 6 , 495 (1863). (20) M . Linhard and H . Flygave, Z anorg. allgem Chem , 262, 328 (1960).
J . P. CANDLIN, J. HALPERN, ASD D. L. T R I h r h f
1020
plunger of the receiving syringe, a t the same time triggering the oscilloscope. The oscilloscope trace \T;E recorded with a I'(iI:troid camera. The reactions were generally followed by observing the decrease in absorbance of the cobalt(II1) complex a t the wave length o f maximum extitictioil i n tlie vicinity of 500 inp (see Table I ) . The reductant \va5 generally in large excess so that tlie observed kinetics were of pseudo-first order. Exceptions t o this were very fast reactions ( t i < 20 nisec.) where the initial coiicetitratiriiis of both reactants were kept 3 s low as possible (anti hence of t h e same order) t o slow do\vii the reaction; anti tlie reactioiis itiv!)lving Cr( bipy),Z* as retiuct:uit which were followed by observing the disappearance of the latter ( a t S62 n i b ) with the oxidant i n excess. The practical limits of the flow apparatus were 5 X 10 - 3 sec. < ti < 30 sec. Iieactioiis with a half-life greater than :3!1 sec. were folloxed spcctropliotoinetrically in a Cary Model 14 spectrophotometer, using a conventional thermostated ahsorption cell which was filled, following deos>-genatioti,by injecting tlie solutions froin thermostaterl syringes through a mixing chamber similar to t h a t of the stopped-flow apparatus. In the reactioiis involving Cr?+, Y?+, and ELI?-,the iiiitial oxidant concentration was geiierall>- in the range 10-' to 1 0 F .lf
TABLE I SPECTRAL DATA( R = Co( MI3)&) Complex
Amax m p
Co(en),3+ I RC12+ > IiRr't > KI". This is the reverse of that found for V", Cr(bipy),'+, and C O ( C N ) ~as ~ -reductants and of most other electron-transfer systems thus far studied including the Cr2+-CrX!+ and Cr'+-. Cr(NH3)6X'+ reaction^.'^ 2 4 T h e sequence for Eu'+ is difficult to reconcile with an outer-sphere inechanism and the only explanation for i t which we are able to suggest is t h a t i t reflects a contribution to the driving force for reaction derived from increasing thermodynamic stability of the europium(II1) halide product complex in going along the series Eu12-, EuRr+, EuCl", and ELIF'+. The implication of this is that these reactions proceed. in part a t least, by an innersphere mechanism. The reactivities of Cr2+, V 2 + , and Eu2+ toward the phosphato complexes all appear to follow the same KP04H33+ order R P 0 4 > RP04Hf > RP04H2?+ and indeed the rate constants for each complex do not vary much from one reductant t o another. This is considered as suggestive of, b u t hardly compelling evidence for, the operation of a common mechanism. Unfortunately corresponding rate constants for Cr(bipy)32Aare not available for comparison, since the reaction with this reductant appears to proceed through an indirect mechanism (see previous discussion) involving a rate-determining dissociation of C r ( b i p ~ )+,, ~ It has been suggested4 t h a t measurements of AS*, which should be sensitive t o the structure of the activated complex, might provide another basis for distinguishing between inner- and outer-sphere electrontransfer mechanisms. Neither earlier results nor those obtained in this study (Table 111) appear t o offer much encouragement in this direction. Most of the activation entropies are seen t o lie in the range -(25 i IO) e.u. The variation is probably not greatly outside the range of experimental uncertainty and no systematic trend is discernible. Measurements are presently under way in this laboratory of the volume of activation of some of these reactions, another kinetic parameter of potential interest from this standpoint. A% final comment would appear to be in order, emphasizing the lack of justification, within the context of the general approach adopted here, for assuming that the reactions of a given reductant, even with a series of closely related oxidants, all proceed through the same mechanism. Thus, it has recently been shown6 t h a t the reduction of pentaamminecobalt(II1) complexes by C O ( C N ) ~ may ~ - proceed either by outer- or inner-sphere mechanisms. A similar conclusion is indicated for Vst,convincing evidence having previously been advanced that the reduction of methylfumaratopentaamminecobalt(111)and related complexes by this reductant proceeds through an inner-sphere bridging mechanism involving remote attack by \'?+ on the conjugated ligand.29
-
0
0
Co(en)P
Vol. Mi
liOAc*-
IiSCS?'
1 i x ~ ~ 3 -R F Z -
1
