126
NOTES
Vol. 63
KINETICS OF ELECTRON TRANSFER BETWEEN COBALTOUS AND COBALTIC AMMINES BY EVAN APPELMAN, MICHAEL ANBARAND
8ome of the fastest systems the reaction was followed to 9701, exchange. Samples for spectroanalysis were quenched in hydrochloric acid and their absorption spectra measured between 3200 and 10,000 8.,in a Beckman Model D U spectrophotomHENRY TAUBE eter.
George Herbert Jones Chemical Laboratory, University o f Chicago, Chicago, Illinois Received Auguat 8, 1968
I n ammoniacal cobaltous solutions there exists a labile equilibrium among all possible species CO(NHa)n++ and Co(NH3),0H+ in which n ranges from zero through six.l These species exchange electrons only very slowly with the hexammine cobaltic ion12presumably due to the absence of a low energy mechanism by which the electron can be transferred through the proton shell surrounding the cobaltic ion. If an acidopeiitammine cobaltic ion is used instead of the hexammine, the acido group ought t o provide such a transfer mechanism by forming the symmetrical bridged activated complex3 (NH&,CO*XCO(NH~)~. Exchange has been reported t o take place between cobaltous species and aquopentammine cobaltic ion in strongly ammoniacal solution14and we have uiidertaken a detailed study of the reaction.
Experimental Aquopentammine cobaltic perchlorate was prepared as described elsewhere6 and a Com(I1) solution of high specific activity was obtained from the Health Physics group of the University of Chicago. All other chemicals were commercial products of reagent grade. To prevent air oxidation of the cobaltous ammines, most reactions were carried out in rubber-ca ped serum bottles under a static nitrogen stmosphere. gamples for radioanalysis were withdrawn by a hypodermic syringe and needle and were delivered into a nearly saturated solution of mercuric chloride in concentrated hydrochloric acid cooled to its freezing point. This quenched the exchange and precipitated the cobalt(111) as Co(NH3)sHzOHgCls.6 The precipitate was filtered, washed with alcohol and ether, airdried, mounted on an aluminum card and counted in an endwindow, methane-flow p-proportional counter. Corrections were made for variations in self-absorption among the samples. I n a few preliminary experiments a different procedure was used. Mixtures were outgassed and sealed under vacuum. For analysis they were frozen, broken open and acidified, thawed and saturated with ammonium thiocyanate. The cobalt(I1) was then extracted into an ethyl ether-amyl alcohol mixture as the thiocyanate complex,B and the remaining aqueous solution of cobalt(II1) was counted in a glass-jacketed, thin-walled Geiger counter. Tests indicated separation-induced exchange to be negligible in either procedure. I n all cases the radioactive tracer was added in the cobaltous form. All reaction mixtures were protected from exposure to light and were discarded if any precipitate was observed. Blanks were run to determine the extent of oxidation by residual oxygen and the appropriate corrections, amounting to 5--10'% of the sample activity, were made. Reactions were followed a t least to 50% exchange in all hut the very slowest systems, in a few of which the exchange was only 2&3070 completed at the end of the run. In (1) J . Bjerrum, "Metal Ammine Formation in Aqueous Solution," P. Haase and Son, Copenhagen, Denmark, 1941, pp. 286-295. (2) W. B . Lewis, C. D. Coryell and J. W. Irvine, J . Chem. Sac., 8386
(1949).
(3) H. Taube and H. Myers, J. A m . Chem. Soc.. 1 6 , 2103 (1954). (4) K. McCallum, "Brookhaven Conf. on Isotopic Exchange Reactions and Chemical Kinetics" (Chem. Conf. No. 21, BNL-C-8, (19481, p. 128. (5) A. C . Rutenberg and H. Taube, J . Chem. Phys., 30, 826 (1952). (6) J . Flagg, J . A m . Chem. Sac., 63, 557 (1941).
TABLE I" Expt. no.
Totrtlb NHa
Free NH3
Co++C X 104
ROH++d X 103
Rate X 104
M/hr.
k X 10-6, hr,-I, 144-4
1' 0 . 2 3 0.163 0.31 42 1.32 0.87 2' .52 5.6 .72 ,167 1.42 42 1 30 .70 3' .34 .75 21 ,164 4h .35 5.5 .70 ,174 .62 79 3.1 .88 5 .33 .158 .85 42 6; .35 .171 .66 42 3.0 .75 7 .69 .47 ,0153 43 10.6 .71 8i .69 .47 .0153 42 10.5 .71 .62 2.4 9 .167 .052 16.5 41 loi .167 .052 16.5 40 .37 1.5 11 ,092 .020 88 39 . 3 9 36 12' .095 ,020 88 36 .22 22 13' .33 .150 .48 43 .75 .48 1.52 $65 14'" .33 .154 .63 43 5.8 1.19 15l .33 .lGl 1.07 42 13.0 1.82 16" .33 .165 1.41 41 " All concentrations are moles/liter (ilf). The bulk electrolyte was a mixture of NaN03 and NH4N03 such that the total ionic strength was 2.28, with c104-present a t concentrat,ions between 0.13 and 0.35 M. Unless otherwise specified, temp. = 35.5", NH4+ = 0.94 M , total Co(I1) = 0.0535 M , and total Co(II1) = 0.043 M. Total NH3 = free NH3 NH3 complexed by Co(I1). Distribution of Co(T1) species was calculated from equilibrium data on Co(11) ammines' by neglecting formation of hydroxocobaltous ions and assuming that substitution of Na+ did not alter the equilibria. The mean number of "3's bonded to a Co++ varied from 4.1 a t the highest NHI concentrations to 1.4 a t the lowest. R = Co(NH3):+++. Correction has been made for the small fraction ot Co(1II) remaining as RHzO+++,using the acid dissociation constants of NH4+and RHzO+++inthesemedia.' e Co(I1) = 0.0214M. Co(II1) = f Co(I1) = 0.107 M . 0 Co(II1). = 0.0210 M . 0.080M. NHd+ = 1.83M. 15.65'. '" 25.0'. 44.4". 55.2".
+
'
Results and Discussion After a week a t 45", exchange in a mixture 0.01 M in acid and 0.04 M each in Co(I1) and aquopent,ammine Co(II1) has proceeded less than 0.4% toward completion. This agrees with the absence of exchange in the CO++-CO(T\TH~)~C~++ system.6 I n ammoniacal solutioiis the reaction followed the exponential equation for homogeneous exchange.' The principd results appear in Table I. I n addition, compatible results were obtained in the few experiments in which the Co(IT1) was radioassayed in solution, and an experiment in an all-perchlorate medium indicated nitrate ion t o have no great specific effect on the reaction. After exchange in a mixture similar to expt. 7 had proceeded to 80% of completion, the absorption spectrum of an acidified sample could be quantitatively matched by summing the spectrum of a solution containing cobaltous ion with that of a solution containing aquopentammiiie cobaltic ion, indicating that no net reaction had taken place. The values of "k" in the last column of the table (7) G. Friedlander and J . Kennedy, "Nuclear and Radiochemistry," John Wiley and Sons, Inc., New York, N. Y., 1955, p. 316.
L
NOTES
,Jan., 1959
127
At 35.5", and with amnioiiia greater than 0.15 M, exchange rate = ~ ( C ~ + + ) ( C O ( K H ~ ) ~ O H + ~(A) ) ( N H ~k') ~ = 1.6 (M)--l(hr.)-l, AH*' = 15.6 kcal., aiid AS+.' = -23.5 e.u. At 45" k' becomes 2.2, in sharp From the first five experiments we conclude that contrast to the corresponding constant for the the reaction is first order in Co++ and in CO(NHB)~-CO(NH~)~++-CO(NH&+++ exchange, t o which OH++. Comparison of 8 with 7, and of 6 with 5, an upper limit of 0.003 (M)-l(hr.)-l has been set shows that at ammonia concentrations above 0.16 a t this temperature.2 M the rate is independent of acidity. Comparison Acknowledgments.-We gratefully acknowledge of 7 and 8 with 1-6 shows that between ammonia the extensive aid and encouragement given by Dr. concentrations of 0.16 and 0.47 M the reaction is closely fifth order in ammonia. At lower am- John W. Cahn during the early part of this work, monia concentrations (expts. 9-12) the reaction and we wish to thank Professor Nathan Sugarman is much more rapid than is predicted by rate law for providing our proportional counter. This (A), and for such conditions paths of lower order research was supported by the Office of Naval Rein ammonia presumably are involved. I n the search under Contract N 6-ori-02026. range 0.02 to 0.05 A4 ammonia, the order with respect to ammonia is between one and two. In THERMODYNAMICS OF THE MERCURY this range an inverse dependence on acidity of REDUCTION OF TITANIUM somewhat Iess than first order is indicated. TETRAFLUORIDE A plot of In ( k / T ) versus 1/T for expts. 5 and 13-16 shows a slight positive curvature a t the high BYJOHN N. BLOCHER, JR.A N D ELTON H. HALL temperature end, giving a AH* which rises from 5.1 Contribution from Battelle Mernorial Institute, Columbus,Ohio kcal. a t the four lower temperatures to 7.5 kcal. Received August 8 , 1958 between the two highest. From the lower value As part of a program to determine the thermowe calculate a AS* of -31.5 e.u. Thus, although a t the highest ammonia concen- dynamic properties of the titanium halides, the trations exchange may proceed primarily through authors have studied the mercury reduction of the symmetrical activated complex (NHJ&o*- titanium tetrafluoride by the technique previously O H C O ( N H ~ ) ~ . lowering +~, of the ammonia con- employed with the bromide system. The reagents centration brings out actiQated complexes contain- were triple-distilled mercury and titanium tetraing less than 5 ammonia molecules for each Co++- fluoride twice sublimed in vacuo as described Symmetrical activated complexes can be formulated earlier. Two independent determinations of the equilibusing Iess than five ammonia complexes for each Co++. However, these would require multiple rium pressure as a function of temperature gave the bridging and bridging through the ammonia, and data shown in Fig. 1 which are represented by the in the limit would still require 2 ammonia mole- equation of the full line cules for each Co++. Such bridging seems very log Patrn= -5574/T - 1.86 log T + 14.62 (1) unlikely, and in any case the observation that the order of the reaction with respect t o ammonia con- The coefficient of the log T term (ACp/R) was evalucentrations does fall below two is not accommodated ated from an estimate of AC, = -3.7 for the reacby such mechanisms. An alternative suggestion tion '/z HgzFds) TiFds) = T i F k ) H d l ) (2) is that unsymmetrical activated complexes play a role. Taking account also of the observation that to which the equilibrium pressures were tentatively at low ammonia the reaction rate increases with assigned. For the calculation of ACp, the heat alkalinity, the reaction may be represented by the capacity a t 298.2"K. of Hg(1) = 6.65 cal./mole/ parallel reactions deg. was taken from Circular 500, National Bureau of Standards, and that of TiF4(g) = 19.8 cal./mole/ ki"9" -+ deg. was calculated from the estimated vibrational C O * ( N H ~ ) ~ O H ~Co(NHa)60Hf+ ~-~ +-frequencies in reference 2. The heat capacities of k- Im,n HgzFz(s) = 22.5 cal./mole/deg. and TiF,(s) = 18.9 C O * ( N H ~ ) ~ O H ~ + ICo(NHa),++ '-~ (B) cal./mole/deg. were estimated by comparison of with n equal to zero or one and m taking all values published value^^-^ for the fluorides and chlofrom zero through 5. Since under tbe conditions rides of Al, Mg, Ti and Hg. of our experiments Co(NH&OH++ is stable with From equation 1 are calculated respect to the lower ammine complex,l the mechAPz9s.z = 11.84 kcal./mole anism requires that Co(I1) bring about the rapid AHOZQB.~ = 24.4 kcal./mole (compared t o rate of electron transfer) transformaAh'oz~s.2 = 42.1 e.u. tion of the lower ammines to the pentammine complex. The direct test of this feature of the mech- Although X-ray diffraction patterns of the reaction anism has not yet been made, but there is no evi- product showed only HgzFz aiid some poorly de(1) E. H. Hall and J . 31. Blochcr, J r . , J . A'lPclrochem. Roc., 106, 40 dence from preparative procedures that the re(1958). actions in question are rapid enough. (2) E. H. Hall, J. M . Blocher, Jr., and I. E. Campbcll, ibkl., 106, 275 For the renction: Co++ 5NHa = CO(N&)~++, (1958). K e q = 4.79 X loW5in our media a t 35.5" with (3) Circular 500, National Bureau of Standards. (4) F. D. Rossini, et al.. "Properties of Titanium Compounds and AH = -10.5 kcal. and A S = -8.0 e.u. We may Related Substances," O N R Report ACR-17, Office of Naval Research then rewrite rate law (A) (195G). were calculated from the rate law
+
+
+
+
*
+
exchange rate = ~ ' ( C O ( N H ~ ) ~ + + ) ( C ~ ( N H ~ ) ~(C) OH++)
(5) IC. K. Kelly, U. S. Bureau of Mines Bulletin 47A (1849).