3108
Vol. 85
ALBERTHAIMAND HENRYTAUBE
method in being applicable only so long as the plateau region lasts : both are based on the conservation of mass and the fact that transport occurs only by sedimentation in a region where bc/br = 0. Unfortunately there seems to be little gain in accuracy when it is used with interference’j rather than schlieren optics. 2 6 8 2 7 I n addition to being precise, the present method has the advantage that the measurements and computations are simple. To find Ac it is necessary only to record the positions of a few fringes a t either end of the column and then to count the fringes in between.4 A New Method for Measuring Molecular Weights during the Transient Period.-Since the slope of the plot in Fig. 1 gives s2/D, it could be used to find s / D (and therefore ,If) if s were known. This is illustrated in Table 11, where predicted values of s (eq. 1) and known values of e, p , and b In y/dc have been combined with the present measurements of s2/Dto give &If. The
(inspection of the boundary shape also provides a check on the homogeneity of the material) and then to measure (Ac/co) us. t in experiments a t lower rotor speed where a freely sedimenting boundary is not formed. Then the results from the two experiments are combined to give the molecular weight. Factors Controlling the Time Needed to Reach Equilibrium.-With the aid of eq. 6 we can now derive expressions for two effects noted experimentally in an earlier study4of the approach to sedimentation equilibrium. First Ac/co is almost independent of column height while the plateau region lasts. When we conipare two experiments (1 and 2 ) at the same time and rotor speed but with different column heights, we find from eq. 6
The column height H enters only in a small correction term. There is however a definite dependence of Ac VALUES FOR THE on the position of the mid-point 1. I n principle one could hold 1 constant while varying H , but in routine experiments 7 would be likely to change when II is co, g./IOO ml. M (obsd.)/M” changed. 3.619 0.994 f 0.004 Secondly it was noted4 that the rate of approach to 4.498 0.998 f ,009 sedimentation equilibrium is nearly independent of 5.420 1.003 j= ,004 u2 during the existence of the plateau region as well as 5.579 1.001 f ,007 in later stages of the experiment. The departure from a Calculated from eq. 7 using known values for the sedimentaequilibrium may be measured by the parameter E , tion coefficient: in this case, the ones given by the extended Svedberg equation and the diffusion data of Gosting and M ~ r r i s . ~ which is defined as TABLE I1 MOLECULAR WEIGHT O F SUCROSE OBTAINED BY A NEW METHOD
The experimental uncertainty indicated for M (obsd.) is that of finding the slope of eq. 7.”
agreement is quite good. This new method might be a good way of making use of the potential accuracy of interference optics in finding molecular weights from the transient period, something which has not yet been accomplished with the Archibald method (cf. Richards and Schachman3). It could be useful with proteins of large molecular weight which take a long time to reach sedimentation equilibrium because of low diffusion coefficientsZ0 The procedure would be to find s from the rate of movement of a boundary a t high rotor speed ( 2 7 ) K. E. Van Holde, J . Phys. Chem., 63, 1574 (1959).
[CONTRIRUTION FROM
THE
E 5
1 - Ac/Ac(eq)
(12)
When Ac(eq) is expressed by eq. 9 and Ac by eq. 6, and these are substituted into eq. 11, one obtains a simple expression for E which is valid in the first part of the experiment, while the plateau region lasts. Here E = 1 - 4(Dt/~H~)’/2[ f1 ( H w Z s / 4 ) ( ~ t / D ) ’+ / 2 . . ] (13) u2 appears only in the minor correction term and e depends chiefly on ( D / H 2 ) ,as is also true after the plateau region vanishes.4 , 2 0 Acknowledgment.-We wish to thank Dr. T. Kotaka for his discussions of the theory.
DEPARTMENT O F CHEMISTRY, STANFORD UNIVERSITY, STANFORD, CALIF.]
The Reactions of Iodopentaamminecobalt (111)with Various “Two-Electron” Oxidizing Agents BY ALBERTHAIM’A N D HENRYTAUBE RECEIVED APRIL2, 1963 The reactions of Co(NH3)51++with CIZand Brz result in quantitative yields of C O ( N H ~ ) ~ C Iand + + Co( X H 3 ) ~ B r A + ,respectively. Similarly, the reactions of C O ( N H ~ ) ~ Band ~ + +Cr(?;H3)5Br++ (in 0.10 ,21 CI-) with Clz produce quantitative yields of Co( SH3)5ClC+ and Cr( S H a ) & l + + , respectively. The reactions of Co( ?;H;)gB r + + with HOBr and HOC1 and of C O ( S H B ) J + +with HOBr, ICI, 03,CH3C03H, &08-z, HSC;-, and H2OZ yield Co(iYH3)jOHzT3quantitatively. 01* tracer studies on the reactions of C0(”8)51+: with HzOz and O3 indicate that 31 and 5.570, respectively, of the oxygen in the C O ( I ~ H ~ ) ~ Oproduct H ~ + ~ IS derived from the oxidizing agent. The reaction of Co( IYH3)sI++with H202when H + and H2OZare in excess obeys the rate law k3(Co( SH3)51 C)(H20~)(TH +) during approximately the first two half-lives, but this phase is terminated by a sharp decrease in ( e o ( h H 3 ) d +). Iodine and iodate ion strongly accelerate the rate of disappearance of Co( K H 3 ) J C + . The rate of reaction of Co(B&)61++ and HzOZis independent of (C1-) but as (C1-) is increased Co( SH3)sC1+Cbecomes an increasingly important product. Co( iVH3)5C1++is also formed in high yields when C0(”3)51++ reacts with 0 3 or S Z O ~in- ~the presence of CI-. The mechanisms of the reactions are discussed, H and they feature rearrangements of the type ( h’H3)5CoIOH + 3 to ( NH3)5CoOI+ 3 and ( NH3)aCoICI+ 3 to ( NH3)sCoClIf3, followed in each case by loss of I(1) (as HOI, for example). +
+
The behavior of the iodopentaamminecobalt(II1) ion toward various “one-electron” oxidation-reduction reagents has been reported recently.’ I t was demonstrated that with some of these reagents (hydroxyl and (1) A . Haim and H. T aube, J . A m . Chenz. SOC.,86, 495 (1963).
methyl radicals, iodine atoms) reduction of the Co(II1) center occurs, whereas with other reagents (ceric and cobaltic ions) the oxidation state of the Co(II1) center is preserved. In the present paper we report the results obtained in a study of the reactions of Co(NH3)b-
~
~
~
IODOPENTAAMMINECOBALT(III)
Oct. 20, 1963
I ++ with various "two-electron" oxidizing agents. Two important features have emerged from this investigation. First, the oxidation-reduction changes that occur in the bound iodide ion do not involve any net change in the oxidation state of the Co(II1) center and, in all instances, the cobalt-containing products are Co(II1) pentaammine complexes. Second, i t has been possible to show in some cases that the reduced form of the oxidizing agent is retained'in the coordination sphere of the Co(II1) product. The very efficient capture.of ligands when oxidizing agents act on I- or Br- bound in the coordination sphere of Co(II1) which we have observed in the present work has no precedent in the numerous reactions so far reported for Co(II1) complexes. Experimental Materials.-Hypochlorous and hypobromous acids were prepared by treating an excess of saturated chlorine or bromine water at 0' with silver perchlorate-perchloric acid solution. The silver halide formed was filtered out and nitrogen was bubbled through the resulting solutions t o remove the excess chlorine or bromine. These solutions were standardized iodonietrically . Peroxyacetic acid solutions were prepared from glacial acetic acid and 30% hydrogen peroxide.% The unreacted hydrogen peroxide was titrated with Ce( IV) and the peroxyacetic acid was standardized iodometrically . Car0 acid was prepared by heating sodium peroxydisulfate in 5 A' perchloric acid.3 A11 other chemicals were reagent grade. Ordinary distilled water was used in all experiments. Preparation of Complexes.-[Co( NH3)jI](C104)P,[Co(NH3);Br](C104)~,[Co(SH3)sOH~](Cloa)s, and [Co(SH3)jC1]C12 were synthesized as described earlier.' Since some experiments required a very precise knowledge of the extinction coefficients of the chloro and aquo complexes a t various wave lengths, [Co(KH~)~OH P] [clo4)3 and [Co(NH~)jCllClawere purified by recrystallizing them four times from aqueous solution. [Cr( S H ~ ) ~ O H P ] ( ' SH,SOa was prepared according t o the procedure described by M ~ r i . Treatment ~ with hydrochloric or hydrobromic acid yielded [Cr(XH3)bCII C ~and P [Cr(SH3)aBrlBr?, respectively. Stoichiometric Experiments.-Solutions of the comDlex ion under study and of bther reagents required were pipet'ted into volumetric flasks. The oxidizing agent was then added (03and Cln as gases and other oxidizing agents in solution) and a n y volume defect was made u p by adding solvent. After the reactions had proceeded t o completion, the solutions were treated in the following manner. For the experiments with Clp, BrP, and 0 3 , the excess oxidant was removed by bubbling nitrogen through the solutions. For the experiments with HOC1 and HORr, the excess oxidant was reduced with sodium bromide and the bromine formed was removed by a stream of nitrogen. For the experiments with IC], HPO?,Sp0,-2, HSOj-, and CHsC03H, the iodine formed was extracted with carbon tetrachloride and the aqueous layer was centrifuged t o clarify i t . After treating the product solutions as described, aliquots were withdrawn and examined spectrophotometrically t o determine the nature of the cobalt species present in solution. Kinetics of the Co( NH3)J++-H?Op Reaction.-Solutions of Co( r u " ~ ) j I + + a n dall other reagents except the HZOPwere pipetted into a volumetric flask which was placed in a constant temperature bath at 25 f 0.1'. After temperature equilibrium had been reached, the HPOpwas added, and t h e solution was made up t o volume and rapidly transferred to a spectrophotometric cell. The cell was placed in the thermostated (25 2~ 0.1') cell holder of a Cary spectrophotometer, and a recording of optical density t ~ s .time a t the desired wave length was obtained. -411 experirneiita were performed a t 388 mp; a t this wave length Co( NH3)JfC C O ( N H ~ ) ~ O Hand ~ + I~p ,have extinction coefficients of 2700, 11, and 100, respectively. In all experiments the concentration of Co(SH3)d was less than 1% that of HpO?. Pseudo-first-order rate constants kl were obtained f r o m the slopes of plots of log (D,- D , ) v s . time; D,and D m are the optical densities of the solution a t 1 and after reaction is complete, respectively; D , could not be measured experimentally because even after all the Co( S H 3 ) J + +has reacted, the optical density at 388 mp continues t o decrease because Hp02 oxidizes iodine t o iodate. Therefore, values of D, were calculated from the initial concentration of the C Q ( S H ~ ) ~ Ireactant +' and the extinction coefficients of the Co( SH3)50Hp+3and IPproducts assuming that the stoichiometry of the reaction is given by eq. 1. 2H 2Co(XH3jjI + + HpOp = ~ C O ( S H ~ ) ~ O H 12 ~ (1) +~ +
+
+
+
+
+
( 2 ) W. C . Smit. Rer. frat'. ( h i m . , 49, 075 (1930). (3) I M Kolthoff and I . K . Miller, J . A m . Chrm. S O L . 73, , 3055 (1951) (1) T. Moeller, I n ~ g.Syn , 6 , 131 (1957).
WITH
OXIDIZINGAGENTS
3 1O!)
Although IQis oxidized t o IOa-, the calculation of D, o n the basis of eq. 1 is a sufficiently good approximation. During t h e time when useful kinetic data can be obtained D , is onlv 2.2.5'70 of D, and SS-96%c of the IZcorresponding t o eq. 1 is f o r k e d . Stoichiometry of the Co(NH3)J++-Hp0?Reaction in t h e Presence of CI-.-Solutions of Co(SH3)61++,HCIO,, and HC1 of known concentrations were pipetted into volumetric flasks. The desire'd quantity of H202 was then added and any volume defect was made by adding solvent. After the reaction had proceeded t o completion, the iodine was extracted with carbon tetrachloride, the aqueous layer was centrifuged, and the optical densities at 507 and 550 mp were measured. At 507 mp, CO(SH3)sC1C+and C O ( X H ~ ) ~ O Hhave Z ' ~ an isosbestic point with extinction coefficient 45.2, and therefore this wave length was useful in establishing that the chloro and aquo complexes were the only two cobalt-containing products. Furthermore, any large ( > 2 5 ) difference between experimental and calculated optical density a t 507 m p was adopted as a criterion t o reject experiments (7 out of 44 experiments were rejected on this basis). At 550 m p the extinction coefficients of Co(SHa)jCl-.+ and Co(NH3)jOH2+3are 47.2 and 20.8, respectively; thus this wave length was useful in calculating theconcentration of the Co( NHa16Cl++ product. 0'8 Tracer Experiments.-Co( XH3)Jf+ was dissolved in 0lxenriched water containing the perchloric acid. The oxidizing agent was added and after reaction had proceeded to completion the solution was cooled t o 0' and [Co(nlHg)jOHg]Br3was precipitated by addition of an excess of concentrated H B r . The removal of the bound water and subsequent equilibration with COP were performed as described by Posey and Taube.5 'Ihe mass spectrometric analyses were performed by the courtesy of Dr. John W . Harbaugh of the Geology Department. Spectrophotometric Measurements.-Absorption spectra were obtained with a Cary Model 14 recording spectrophotometer fitted with a thermostated cell holder and cell compartment. All experiments with Co( SH3)sI++were performed in semidarkness t o avoid photochemical decomposition. The results obtained for the reactions of Co(SH3)jIf+ have been corrected for the amount of [Co( S H 3 ) s O H ~C104)3 ]( originally present in the [Co(SHa)51](C104)apreparation.% All stoichiometric experiments were performed a t room temperature.
Results In Table I we present a summary of the experiments designed to determine the identity of the cobalt(II1) products formed in the reactions of C O ( N H ~ +) +~,I C O ( N H ~ ) ~ B ~and + + , Cr(NH3)6Br++ with various oxidants. Columns 5 and 6 of Table I list the positions of the maxima and the optical densities a t the maxima, respectively, of the solutions resulting from the various reactions. Column '7 lists the optical densities calculated for quantitative formation of the products indicated in the last column. The excellent agreement between the values in columns 6 and 7 , and also between the values in column 5 , with those expected for the formation of the corresponding products, demonstrates that the major products of the reactions are indeed the ones listed in the last column of Table I . In addition to the wave lengths listed in column 5, other wave lengths a t which the various possible reaction products display a larger difference in extinction coefficients were examined, and i t is estimated that the reaction products indicated in Table I are formed in >95% yield, except where otherwise indicated. T h e agreement referred to above shows that Co(I1) is not an important reaction product. In addition, the absence of Co(I1) was demonstrated for experiments 1, 2, 6, and 8 by examining the absorption of the corresponding solutions in 9.6 M HCl a t 098 mp (where Co(I1) hds a maximum with extinction coefficient 522). Since no absorption could be detected, it is concluded that less than 0.570 Co(I1) is formed in these reactions. Experiments 2, 10, and 1.1 of Table I show that + + C O ( N H ~ ) ~ Bwith ~++ the reaction of C O ( N H ~ ) ~ Iand Clz yield C O ( N H ~ ) ~ C I +even + , when as in 0.010 AI H i the hydrolysis of Cls is appreciable. In contrast, the reaction of Cr(NH3)5Br++with C12 in 1.0 *If H + yields principally the Cr(h"3)50H2+3 ion (cf. expt. 15). However, i n 1.0 -If H + and 0.10 -11C1- ( c f . expt. 16) ( 5 ) F. A . Posey and H. Taube, J . A m Chem. Soc., 79, 255 (1957)
Vol. 85
ALBERTHAIMAND HENRYTAUBE
31 10
TABLE I SPECTROPHOTOMETRIC IDENTIFICATION O F THE c O ( 111) PRODUCTS
Cr( NH3)5Br
++
Expt
1 2
Complex, c O ( "3)sI
M X 108 ++,3 , 3 0 ++,3 . 2 0
FORMED B Y THE REACTION$ OF VARIOUSOXIDANTS
M
( H +I, M
Br2, 0,050 Clz, bubbling
0.010 0.010
Oxidant,
c O ( "3)6I++,
Co( NH3)5Br++,A N D
WITH
hnax,
mp
Dmax
DIU,,, calcd.*
Co(II1) product
550' 532" 363e 492' 345' 492' 345' 492' 345' 492' 345' 492h 345h 492' 345J 492h 345* 532e 363e 532e
0.170 0 . 176d Co(XH3)5Br++ c O ( "3)6I ,157 . 160d CO(1\"3)6C1++ ,155 151d CO(NH3)J ++, 1 .05 3 HOBr, 0.020 1.0 ,490 ,500 CO(NH~)~OHZ+~ 460 ,470 Co(NH3)51++, 1 . 1 4 4 ICI, 0.10 1.0 544 ,541 CO(N H ~ ) ~ O H Z + ~ ,525 ,509 CO(NH3 )&I+ +, 9 . 0 0 5 CH3C03H, 0.065 0.010 420 ,425 CO(N H ~ ) ~ O H Z + ~ ,400 ,396 cO( "3)51++, 0 . 8 9 6 03,bubbling 0.010 ,427 ,422 CO(NH3 )60Hz +3 ,401 ,396 7 CO(NH3 )a1 +,5.36 &0*-2, 0.20 1.o ,254 CO(NH3)60Hz+3 ,250 ,230 ,239 CO(NH3 j51 ++,5 . 4 8 HzOz, 0.50 8 1.0 ,258 ,260 CO(NH3)50Hz ,243 ,240 CO(3" jaI ++, 6 . 5 0 9 HS06-, 0.020 1.0 ,308 CO(NH3)sOHzi3 ,310 ,290 ,300 Co( NH3 j5Br++, 5.40' 10 Clz, bubbling 0,010 ,278 CO(NH3)5C1++ .277 ,256 ,261 11 Co( NH3J5Br,5.20 Clz, bubbling 1.o ,265 ,267 CO(NH3)bCI + + 3ae ,251 ,252 C O ( N H ~ ) ~0B. 9~1, 12 HOBr, 0.020 1.0 492' ,426 ,431 CO(NH3)60HzC3 345' ,405 ,395 Co( NHa)5Br,3 . 0 4 HOC], 0.020 13 1.o 492' ,140 ,144 CO(NH~)~OHZ+~ 345' ,135 ,139 14 CO(XH3)5Cl++, 6 . 2 5 HOC1, 0.020 1.o 532" . 322i ... ,326 363e ,302' ,303 Cr( NH2)6Br++,5 . 9 0 Clz, bubbling 15 492',k 1.0 ,191 ,190 9% Cr(NH3)5Cl++ 3673,k ,188 ,187 91% C ~ ( N H ~ ) ~ O H Z + ~ Clz, bubbling' Cr( NH3)5Br++,2.86 16 1.0 513k ,510 ,512 Cr( NH3)6C1++ 376k ,538 ,540 a The reactions are complete on mixing except in experiments 7 (ti/* 3 min. a t 0.05 M &Os-2),8 (rate reported on elsewhere in this paper), and 14 (no reaction in 10 min.). Optical density calculated for formation of the product indicated in the last column. A,, for C O ( N H ~ ) ~ Bis~ +550 + mp. Corrected for the C O ( N H & O H Z +originally ~ present in the [ C O ( N H ~ ) ~ I ] ( C Ipreparation. O~)~ e A ,, for Co(NH3)5CI++are 532 and 363 mp. Amsx for Co(NH&OH2+3 are 492 and 345 mp. 0 Amax for C O ( N H ~ ) ~ O C O C Hare ~++ 510 and Amax for Co(NH3)5S04+are 515 and 355 mp. 353 mp. Calculated for no reaction. 1 Xma, for Cr(NH3)50HZ+3 are 480 and 361 mp. Amax for Cr(NH3)5Cl++ are 512 and 375 mp. 0.10 M CI- present. +
+'
{
'
the dominant reaction product becomes Cr(NH3)5CI++. The iodometric titer of the solutions resulting from experiments 1, 2, 10, and 11 of Table I was determined after the excess chlorine or bromine had been removed. For the reactions of Co(NH3)51++, the titer corresponded t o oxidation of I - to I03-.6,7 In contrast, 'the solutions derived from the experiments with Co(NHs)gBr++ did not display any oxidizing capacity after the volatile oxidizing agents were removed. The stoichiometries of these reactions are therefore described by eq. 2 and 3.
No reaction occurs between C O ( N H ~ ) ~ Cand ~ + +HOCl in 10 min. a t room temperature (expt. 14). Finally, the reaction between Co(NH3),I++ and IC1 (expt. 4) yields only C O ( N H ~ ) ~ O H Z + ~ . Kinetics of the Co ("3)61 ++-HzOzReaction.-This reaction was studied in detail and proved to be fairly complicated. Although only some general features of the reaction have been uncovered, they appear to us to be of such interest as t o justify a report a t the present time. First, as already indicated, the only cobalt-containing product (in the absence of anions other than C104-) is C O ( N H ~ ) ~ O H ~With + ~ . regard to C O ( X H ~ ) S I ++ + 3x2 + 3Hz0 = Co(NH3)5X++ 1 0 3 the iodine-containing products, Iz is a reaction product, 5X- + 6 H + ( X = C1, Br) (2) but under the experimental conditions (high H + and l/zBrz (3) Co(NHs)sBr+++ '/2C12 = Co(NH,),Cl++ HzOz concentrations), 1 2 is ultimately oxidized to 103-.* I t is noteworthy that the reactions of C O ( N H ~ ) ~ I + +The other features of the reaction are best considered with CH3C03H(expt. 5 ) , SzOs-2(expt. 7 ) , and HS05in relation to the kinetic studies summarized in Table (expt. 9) do not yield acetato or sulfato complexes but I1 and Fig. 1. The pseudo-first-order rate constants CO(NH~)SOHZ as+the ~ only cobalt-containing reaction kl listed in column 7 of Table I1 were obtained from the product. It is not surprising that the reactions of O3 initial slopes in plots of log (Dt - D m )v s . time. I t was (expt. 6) and H202 (expt. 8) with Co(NH3),1++ yield necessary to use initial slopes because the plots for all C O ( N H ~ ) ~ O H ~additional +~; insight into these rekinetic experiments displayed deviations from linearity actions is provided by O'*-tracer experiments which between the second and third h a l f - l i ~ e s . ~This is are described below. The reactions of HOBr with ( 8 ) For a review of work on t h e 1 0 3 - , 1 2 , H 2 0 ~system see J. H . Baxendale ~ + + 3, 12) and of Co(NH3)sI + + and C O ( N H ~ ) ~ B (expt. in "Advances in Catalysis," Vol. I V , Academic Press, Inc., New York, N . Y . , 1952, p. 31. HOCl with Co(NH3)5Br++ (expt. 13) are rapid and (9) T h e deviations from a simple pseudo-first-order behavior were Co(r\"3)50H~+~is the final product in all instances. dramatically displayed in the optical densities v s . time curves obtained
+
+
( 6 ) This reaction has been exploited t o analyze iodine in Co(NHa)aI++ salts (7) See R G. Yalman, J . A m . Chem SOC.,71, 1842 (1953), and ref 1
with the spectrophotometer. Such curves are concave upward f o r 3-4 halflives when suddenly a n inflection point occurs and t h e optical density rapidly decreases.
KINETICS OF Expt
3111
IODOPENTAAMMINECOBAI ~(111)WITH OXIDIZING AGENTS
Oct. 20, 1963
(Co("3)aI ++), M X 10'
THE
TABLE I1 CO( iYH3)51++-H202 REACTION; IONIC STRENGTH, 0.94; TEMPERATURE, 25" (H+),' M
(HZOP),
M
(C17
(Id,
I
M
M
x
k3, M - 2 min.-l
kl, min. -1
10'
1 2 3 4 5 6
5.11 0.935 0.097 ... ... 0.0262 0.288 5.20 ,935 ,096 ... 5.0 ,0254 282 ... ... ,0198 ,218 0.68 ,935 ,097 7.20 ,935 19 ... ... ,0604 ,340 4.86 ,935 .38 ... ... ,113 ,318 4.24 935 .36 ... ... ,110 328 7 .38 ... ... ,110 ,310 4.86 935 8 4.86 ,935 .38 ... 10.0 ,104 ,293 9 4.86 ,935 .38 ... 0 . 2Ob 102 ,292 10 0.50 ,935 .36 ... ... 079 236 4.80 ,0935 .38 ... ... ,0045 ,126 1lC 12d 4.35 ,935 .36 ... ... ,077 ,229 4.90 ,935 ,97 ... ... ,265 ,294 13 14e 4.98 ,0748 .97 .. ... .017 .24 5.20 ,937 096 0.0124 ... ,020 .22 15 ... 0205 ,227 16 ,124 5.06 ,930 ,097 5.0 ,023 .26 ,930 ,096 ,124 17 4.96 18 ,936 .38 ,0124 ... ,088 ,248 5.20 3.40 ,935 .36 ,036' ... ,081 ,240 19 5.10 ,935 .38 ,075 ... ,083 ,234 20 ,227 21 .38 ,125 ... ,081 5.20 ,935 22 5.00 ,935 .38 .125 5.0 ,082 ,230 .87 .50 ... ,337 282 5.21 1.23 23' .37 . 024i ... ,116 ,273' 24 4.57 0.81h,i .37 .12i ... ,292 ,288' 4 86 0 .9 J h , { 25 a Added as HC104 and HCI. b This is the Concentration of added Io3-. Ionic strength 0.094. Measured a t 460 r n p . e lonic strength adjusted with SaC104. f Added as NaC1. Ionic strength 1.23. Added as HClOa and HzS04. ' ( H + )and (SOa-') calculated using a value of 0.15 for the dissociation constant of HS04-. 2 After subtracting from the total rate that given by the function 4.;(CO( "3)jI I+)(HzO*)(H')(SOa-2). Q
exemplified for expt. 'iof Table I1 by the curve A in Fig. 1. The other experiments summarized in Fig. 1 were designed to uncover the factors that cause the deviations from first-order behavior. A comparison of curves A and B of Fig. 1 shows t h a t although addition of 1.0 X -if I? does not appreciably change the initial slope, i t brings about the rapid disappearance of C O ( N H ~ )++ ~ I a t shorter times. Similarly, expt. 1 and 2 of Table I1 show that addition of an amount of I? equal to ca. twice the iodine in the Co(NH3),I++ reactant does not appreciably change the value of k , (0.0262 and 0.(3254 min.-', respectively). Curves C and D of Fig. 1 serve to demonstrate that either addiM IO3- alone or 2 X M IO3- and tion of 2 X 1 X M Iz tremendously increases the rate of disappearance of Co(NH3),I++. As noted above, both iodine and iodate are formed in this reaction, and therefore t h e autocatalytic nature of the reaction is ascribed to the accumulation of these two products. Direct evidence bearing on the question of iodate formation was obtained from expt. 12 performed a t 460 mH; a t this wave length the extinction coefficients of Iz, Co("3)~I++, and C O ( K H ~ ) ~ O H :are , + ~ 746, 272, and 37.7, respectively. In this experiment, the optical density of the solution first increases, reaches a maximum a t 19 min., and then decreases, demonstrating t h a t Iz is both formed and consumed during the course of the reaction. At the maximum optical density, i t is estimated that 96"/c of the iodine content of the Co(NH3)bI + + consumed is present as Iz. Furthermore, the smaller rate constant a t 460 mpl0 as compared to those (10) T h e rate constant was obtained from the initial slope of a plot of log ( D m - 0 ) zrs. time ( D is the measured optical density a t 400 m r ; D m is the optical density calculated if the stoichiometry of the reaction were t h a t given by eq. 1 ) . I t must be noted t h a t the slope of log ( D m - D ) ws. time does ?tot give the rate of formation of 1 2 , since i t can be shown t h a t
Dm-D=2'C~(1-A)+ 2
governing the disappearance of CO("~)~I + + ( k l values of 0.077 and 0.108 min.-l, respectively) definitely show that during the first two half-lives, the stoichiometry of the reaction deviates from that represented 0.2 I
I
I
I
4
6
I
I
1
1
1
1
8
10
12
14
16
18
1
0.0
-0.2
-0.4 ri
8 -0.6 n I
+- -0.8
n
r-
u
1.0
-I
- 1.2 - 1.4 C
D 2
Time i n minutes. Fig. 1.-The reaction of Co(NH3)jI-T with H202; 4.86 X M Co(NHa)J++, 0.935 M H i , 0.38 'iA Hz02, 25": 0, no addition (curve .4);0 , 1.0 X M 1 2 added (curve B ) ; 8 , 2.0 X 121 I o 3 - added (curve C ) ; (3, 2.0 X M 1 0 3 - and 1.0X M I2 added (curve D ) .
by eq. 1. A measure of iodine formation as the reaction proceeds was obtained from the expression (see footnote 10 for definition of the symbols)
where e r 2 , f R I . and tI
