Anion Induced Ads’orption of Cadmium(l1) on Mercury from Iodide and Bromide Media Fred C. Anson and Donald J. Barclay
Gates and Crellin Laboratories of Chemistry, California Institute of Technology, Pasadena, Gal$ 91109 A detailed study of the adsorption of cadmium(l1) on mercury electrodes from iodide and bromide media is reported. The technique of double-potential step chronocoulometry was employed with a computer for both data acquisition and analysis. Significant adsorption of cadmium(l1) from iodide and bromide media is demonstrated. The extent of adsorption correlates best with the amount of adsorbed halide rather than with the electrode potential or the electronic charge on the metal. On the basis of the very small changes in the charge on the electrode which are produced by the adsorption of cadmium(ll), and from the relationships between the amount of adsorption and the bulk halide ion concentrations, it is concluded that neutral cadmium(l1) species are adsorbed. Other instances of anion-induced adsorption of metal ions are examined in the light of the present results.
METALIONS OF GROUPIIb and the heavier metal ions of Groups IIIb, IVb, and Vb of the periodic table are often adsorbed on mercury electrodes in the presence of halide or pseudo-halide anions. Previously studied examples include cadmium(I1) ( I ) , lead(I1) and mercury(I1) ( 2 ) from iodide media, and cadmium(I1) (3), zinc(I1) (4, 5), lead(I1) (2, 6), thallium(1) (7), and indium(II1) (8) from thiocyanate media. In all cases, the amount of adsorbed metal ion varies in a complex way as the potential and anion concentrations are changed, and in no instance has the nature of the adsorbed species and the factors which contribute to its adsorption been unambiguously determined, although some stimulating speculations have been made (8). This paper reports a study of the adsorption of cadmium(I1) with more extensive data on the potential and halide concentration dependences than have previously been obtained. The use of a computer-based digital data acquisition and analysis system provided great quantities of data with heretofore unattainable precision and accuracy. Features which were hidden in the scatter of earlier data revealed themselves and it became possible to develop a consistent scheme which rationalizes much of the existing data on anion induced adsorption of metal ions. EXPERIMENTAL
The amount of adsorbed cadmium(I1) was determined by means of the double-potential step chronocoulometric technique, the theory and practice of which have been presented (1) G. W. O’Dom and R. W. Murray, ANAL.CHEM.,39,51 (1967). (2) R. W. Murray and D. J. Gross, ibid., 38, 392 (1966). (3) F. C. Anson, J. H. Christie, and R. A. Osteryoung, J. Electroanal. C/7em., 13, 343 (1967). (4) P. Farrington and F. C. Anson, Calif. Inst. Tech., Pasadena, Calif., unpublished data, 1965. (5) R. A. Osteryoung and J. H. Christie, J. Phys. Chem., 71, 1348 (1967). (6) J. H. Christie and F. C. Anson, Calif. Inst. Tech., Pasadena, Calif., unpublished data, 1965. (7) G. Lauer and R. A. Osteryoung, North American-Rockwell Science Center, Thousand Oaks, Calif., unpublished data, 1965. (8) G. W. O’Dom and R. W. Murray, J. Electroanal. Chem., 16, 327 (1968). (9) J. H. Christie, R. A. Osteryoung, and F. C. Anson, ibid., 13, 236 (1967).
in previous communications (3, 9). The experiments were performed with the digital computer-controlled data acquisition system which has been described in detail (IO). The computer was programmed to step the potential of a hanging mercury drop electrode from an initial potential, E,, where no current was flowing to a final potential, E,, where the diffusion-limited reduction of cadmium(I1) occurred. At a selected later time, r , the potential was stepped back to Ei. The measured parameter was the total charge passed at any time after the first potential step. The value of the total charge, Q, was sampled once every millisecond and the corresponding value of Q was stored as a digital data point (a 12-bit binary number) in the magnetic memory of the computer. One hundred such data points were obtained between the initiation of the experiment and time r , and an additional one hundred data points were obtained between time r and the end of the experiment. Thus, in each experiment data were taken for exactly 200 msec. After the data had been collected they were analyzed immediately by the computer according to a programmed least squares fit of the data to the theoretical equations of double-potential step chronocoulometry for the case of adsorbed reactant and non-adsorbed product (9). That is, the data points for the period 0 < t < r were fit to the equation and the data points for t > r were fit to the equation where
e
+
Q(t > r ) = So0 Zb T ) ~ ’ ~ tlIz.
+ (t -
= r112
(2)
The values of SI and sb resulting from the least squares fit are the “forward slope” and “reverse slope,” respectively. The values of Z, and Ib are the “forward intercept” and “reverse intercept,” respectively. The amount of adsorbed cadmium(II), 2FFcd, was calculated from the difference between the forward and reverse intercepts after applying the necessary small correction to the reverse intercept to take account of the fact that Equation 2 is only approximately linear in e when there is significant reactant adsorption (9). The major assumption involved in this method of measuring 2Frcd is that the electrode surface experiences the identical environment for times greater than r as it did before the application of the first potential step-i.e., the values of the surface concentrations of cadmium(I1) and halide ion, the amount of adsorption of these species, and, therefore, the electronic charge on the electrode are all assumed to have returned to their initial values. To assure that this condition had been achieved in the present experiments the first 10 of the 100 reverse data points were not included in the least squares analysis of the reverse step. This 10-msec “wait” was more than adequate for the readsorption of the cadmium(11) species, even at the lowest cadmium(I1) concentration (0.2mM) as the following calculation shows: If all the cadmium(I1) reaching the electrode surface by anodic generation and diffusion from the bulk of the solution is assumed to readsorb until the equilibrium value of adsorbed cadmium, req,is reached, it follows that the time needed for the surface amount to attain re,is given by: (10) G. Lauer, R. Abel, and F. C. Anson, ANAL.CHEM.,39, 765 (1967). VOL. 40, NO. 12, OCTOBER 1968
1791
Table I. Iodide-Induced Adsorption of Cadmium [Cd(Zl)],mM: 0.2 - E i , mV us.
SCE 250 300 350 400 450 500 550
6.8 12.7 21.8 27.4 28.9 28.4 25.7
450 500 550
8.3 20.8 29.8 32.8 34.1 34.0
350 400 450 500 550
14.6 24.8 29.2 31.9 32.4
450 500 550 600 650
8.0 13.4 17.7 18.2 17.1
450 500 550 600 650
2.6 7.0 10.4 12.1 13.5
300 3 50 400
0.4 0.8 2mcd, N c m Z Iodide concentration: 0.01M 7.6 11.2 18.6 25.9 29.4 33.2 31.9 34.1 32.4 33.6 30.9 31.3 27.4 27.2 Iodide concentration: 0.05M 11.o 17.5 28.6 33.7 35.7 38.9 35.6 39.3 36.6 37.8 35.6 36.3 Iodide concentration: 0.10M 22.2 27.5 32.4 35.6 34.3 36.6 35.6 37.3 34.8 35.4 Iodide concentration: 0.50M 11.9 15.8 19.3 24.0 22.7 26.2 22.8 26.3 22.5 24.8 Iodide concentration: 1.OM 5.3 4.1 11.5 13.5 15.8 19.6 17.1 22.2 18.2 21.0
(3) The largest calculated value of t,, was 8.9 msec for 0.2mM 0.95M KNOl solution. cadmium(I1) in the 0.05M KI It was also essential to ensure that the iodide ion concentration returned to its initial value at the electrode surface before reverse data points were taken. That 10 msec sufficed was established in blank experiments with cadmium-free iodide solutions. To make sure that adsorption equilibrium was attained at the beginning of the experiment each newly extruded mercury drop was exposed to the stirred solution for 30 sec before the first potential step. No change in the measured parameters resulted if the waiting time was varied from 20-60 sec. It was necessary to measure the charge on the electrode, qm, at each of the initial potentials. This was accomplished for the iodide solutions by integrating the current flowing into a dropping mercury electrode potentiostated at Et and (11). determining the intercept of a plot of Q / t 2 / *us. This experiment was also performed by means of the computercontrolled data acquisition and analysis system which yielded values of qm that were in excellent agreement with those of Grahame for the nitrate-free 1M iodide solutions. For the bromide solutions, qm was evaluated from the experimental values of Q d , l .(the difference in qm between Eiand E,) by assuming that the charge on the electrode at the final potential
+
(11) G. Lauer and R. A. Osteryoung, ANALCHEM., 39,1866 (1967). 1792
ANALYTICAL CHEMISTRY
1.2
0.8 4m
4 l n
pC/cm2 11.3 29.2 37.2 37.1 34.6 32.6 28.6
54.1 30.3 21.3 15.4 11.8 8.2 3.7
8.0 9.2 4.5 2.1 -0.4 -2.5 -1.0
19.0 36.6 40.3 39.4 38.8 37.0
46.4 26.3 19.6 15.3 12.2 10.0
10.1 11.7 4.8 1.5 -0.7 -2.7
26.9 37.3 38.6 38.3 37.6
31.0 20.9 16.3 13.6 10.5
11.8 7.3 2.3 -0.4 -1.9
17.2 25.2 28.0 26.9 26.1
22.5 16.4 12.6 9.2 4.6
5.2 3.6 0.8 -1.0 -1.2
5.0 16.2 20.1 22.5 22.7
29.9 20.4 15.4 11.2 7.4
1.6 2.3 1.5 0.4 -0.4
E,, was the same as that given by Lawrence, Parsons, and Payne (12) for the same bromide salt activity. All solutions were prepared from reagent grade salts and triply distilled water and were deoxygenated with prepurified nitrogen. The ionic strength of these solutions was maintained at unity with potassium nitrate. The temperature was 25 f 2 "C. When the temperature dependence of 2FI'cd was being studied temperature control was obtained by circulating thermostated water through a jacketed cell. The reference electrode was maintained at room temperature (25 f 2 "C)in these experiments. The electrode was a commercially available hanging mercury drop electrode (Brinkman Instruments, Inc.). Cells were of conventional design. All potentials are referred to the saturated calomel electrode (SCE). RESULTS AND DISCUSSION
The values obtained for the amount of adsorbed cadmium(II) as a function of &, qm, halide concentration, and cadmium(I1) concentration are summarized in Tables I and 11. The values of qm reported are for the 0.8mM cadmium(I1) solutions, The final column in this table gives Aqn, which is the change in the charge on the electrode, at each initial potential, Et, brought about by the adsorption of cadmium-
(12) J. Lawrence, R. Parsons, and R. Payne, J. Electroanal. Chem., 16,193 (1968).
Table II. Bromide-Induced Adsorption of Cadmium [CdUOl, mM. 0.2 -Et, mV us. SCE
150 200 250 300 350 400
450 500
2.8 2.9 2.5 2.0 1.6 1.3 0.7 0.5
500
4.9 5.6 5.7 5.6 4.9 4.0 3.0
200 250 300 350 400 450 500
4.6 5.9 6.3 6.3 6.2 5.3 4.5
200 250 300 350
2.8 4.3 4.9 5.6
200 250 300 350 400
450
400
5.5
450 500 550
4.9 4.5 3.9
250 300 350 400 450 500
3.2 3.9 4.4 4.7 4.5 4.3
0.4
0.8
2Frcdl &/cm2 Bromide concentration: 0.10M 5.3 8.0 5.8 8.3 5.5 7.7 4.5 6.8 3.9 5.8 3.2 4.8 2.6 4.1 1.9 3.2 Bromide concentration: 0.30M 8.0 12.0 8.7 13.3 8.7 13.3 8.3 12.4 7.6 11.5 6.4 10.2 4.7 8.5 Bromide concentration: 0.5M 7.2 11.0 9.0 13.1 9.5 13.9 9.5 13.8 8.7 13.2 7.9 12.0 6.4 10.5 Bromide concentration: 0.7M 5.6 8.1 7.4 10.7 8.3 12.3 8.6 12.7 8.4 12.7 7.9 12.0 7.3 10.5 Bromide concentration: 1.OM 5.7 7.2 6.1 9.5 6.7 10.3 7.0 10.4 6.7 10.7 6.2 9.9
(11). The limits of E* which could be employed were set on the anodic side by the oxidation of mercury and on the cathodic side by the onset of cadmium(I1) reduction. The standard deviations in the intercepts of the least squares lines were typically about 0.1 NC/cm2 and the values obtained for the amounts of adsorbed cadmium(I1) were reproducible within ca. 1 %. Where they overlap, our results for r C d in iodide media are in substantial agreement with those published by O’Dom and Murray ( I ) except at the lower iodide concentrations where the earlier data, obtained in less conducting, KN08-free solutions, were very likely degraded by the unavoidable presence of uncompensated resistance in the cell (13). Dependence of 2 F r C d on Et and qm. It is particdarly important to know which electrical variable, potential or charge, if either, to hold constant when comparing amounts (13) G. Lauer and R. A. Osteryoung, ANAL. CHEM.,38, 1106 (1966).
0.8
1.2
Asm
4m
pC/cm2 8.4 9.5 9.5 8.5 7.4 6.1 4.4 3.6
23.4 17.3 13.3 10.0 7.3 5.8 4.5 1.1
14.0 15.5 15.3 14.6 13.3 11.9 9.8
20.9 16.8 13.0 10.5 8.4 6.2 3.6
12.4 14.0 15.0 15.0 13.9 13.3 10.7
24.9 19.7 16.0 13.1 10.3 7.7
...
29.5 21.5 17.9 14.0 11.8 9.9 6.9
13.2 15.4 15.5 15.7 14.0 12.5
...
8.2 10.1 11.2 11.7 12.6 11.8
5.5
3.0 1.6 1.4 1.6 1.4 0.7 -0.5
0.8
...
*..
... ... ... 1.7 1.4 1.5 1.5 0.6 0.6 -0.2 1.5 1..0 1.1 0.7 -0.4 -0.5
e . .
23.5 19.0 15.1 12.2 10.2 7.9
0.8
...
1.7 0.5
0.8 1.o 0.4 0.3
of adsorption under different conditions (8). Figures 1 and 2 show the dependence of WrCd on Et and qm, respectively, for one concentration of cadmium(I1) and varying halide concentration. The WrCd us. Et plots are much more complex and unsymmetrical than the corresponding plots us. qm. Similar behavior, though not so pronounced, was observed by Bockris, Devanathan, and Muller (14) in a study on the adsorption of n-butanol on mercury. Parsons (15) has suggested that the charge on the metal, rather than the potential, should be used as the independent variable in adsorption studies because at constant charge, electrode-particle interactions should be approximately constant. However, it can be seen from Figure 2 that the charge on the metal corresponding to maximum adsorption becomes less positive as the concentration of halide is increased. At a given charge (14) J. O’M. Bockris, M. A. V. Devanathan, and K. Muller, Proc. Roy. SOC.,A274,55 (1963). (15) R. Parsons, J. Electroanal. Chem., 7 , 136 (1964). VOL 40, NO. 12, OCTOBER 1968
1793
435
I
l
I
I
i
I
I
I
I i
% I
L;
IO
' 0
20
30
40
50
60
70
qm pc/Cm2 100
200
300
400
?OO
600
700
-E, mV, E S C E Figure 1. Potential dependence of Cd(I1) adsorption induced by iodide (dashed lines) and bromide (full lines). [Cd(II)] = 0.8mM Halide concentrations, M : x, 1.0; 0,O.S; 0,O.l; A,O.OS; 0,0.01 Iodide: Bromide: x, 1.0; 0,0.7; 0 , O . S ; A, 0.3; 0,O.l on the metal there is more adsorbed halide on the electrode the higher the halide concentration, so that this behavior suggests that the amount of adsorbed halide is influencing the adsorption of the cadmium(I1) species. The amount of adsorbed halide on the electrode was estimated from Grahame's data (16) for iodide and Lawrence, Parsons, and Payne's data (12) for bromide at the corresponding concentrations of halide, and at the charge on the metal in the presence of adsorbed cadmium(I1). Because these data were not obtained at constant ionic strength, the amount of adsorbed halide was estimated at corresponding salt concentrations rather than corresponding halide ion concentrations in accord with a recent suggestion by Dutkiewicz and Parsons (17). The resulting values for adsorbed halide are only approximations because of the neglect of any coadsorption of nitrate anions. This source of error will be most serious for the bromide-nitrate mixtures at low bromide concentrations. The amount of adsorbed cadmium(I1) at a given charge on the metal was then plotted against the amount of halide estimated to be on the electrode at the same charge, and the result is shown in Figure 3 for the different halide concentrations. Within the estimated error (+ 2 pC/cm*) in the determination of the adsorbed halide these plots are parallel and it is therefore postulated, on the basis of this symmetry, that the cadmium(I1) species which is adsorbed is bound to the electrode surface by means of strong interaction with already adsorbed halide. (16) D. C.Grahame, J. Amer. Chem. Soc., 80,4201 (1958). (17) E. Dutkiewicz and R. Parsons, J. Electroanal. Chem., 11, 100 (1966).
1794
ANALYTICAL CHEMISTRY
Figure 2. Charge dependence of Cd(I1) adsorption induced by iodide (dashed lines) and bromide (full lines). [Cd(II)] = 0.8mM Halide concentrations, M : Iodide: x, 1.0; 0,O.S; 0,O.l; A,O.OS; 0,O.Ol Bromide: x, 1.0 0,0.7; 0 , O . S ; A,O.3; 0,O.l
40 35 30 N
E 25-
a. 20
-
15
-
IO
-
TI
LL
cu
5-
01 0
I
IO
I
20
I
30
I
40
I
50
I
60
J
70
F ~ H A L ~ 1Dpc/cm2 E
Figure 3. Dependence of Cd(I1) adsorption upon adsorbed iodide (dashed lines) and bromide (full lines). [Cd(II)] = 0.8mM Halide concentrations, M : Iodide: x, 1.0; 0,O.S; O,O.l;A, 0.05; 0,0.01 Bromide: x, 1.0; 0 , 0.7; 0,O.S; A , 0 . 3 ; 0,O.l At the most positive charges data were not available for I'haiide at all halide concentrations
The most interesting prominent features of the data in Tables I and I1 and Figures 2 and 3 are the maxima in the plots of 2Frcd 1;s. adsorbed halide (and charge on the electrode) and us. the bulk concentration of halide. The decrease in adsorption at the lowest values of qm seems surely to be caused by the decrease in the amount of adsorbed halide at these charges. Similar behavior has been observed in several other studies (3-6), the most striking example being zinc ion in thiocyanate media where the zinc adsorption is completely eliminated when the charge on the electrode is made negative enough to desorb essentially all of the thiocyanate (4, 5). The decrease in cadmium adsorption at higher positive charges, where halide adsorption is increasing rapidly, is most easily rationalized in terms of a geometric “squeezing out” of the cadmium(I1) species when the adsorbed halide becomes so closely packed that the adsorbing cadmium(I1)-halide species is sterically prevented from associating with the adsorbed halide. The flatter maxima in bromide media probably reflects the somewhat smaller size of bromide compared to iodide. Dependence of 2FI’cd on Temperature. The effect of temperature on the cadmium(I1) adsorption was measured in a 0.90M KN03-0.10M KI solution containing 0.4mM cadmium(I1). The results are summarized in Table 111. These data show that thetempe rature coefficient is dependent upon the initial electrode potential. At most potentials raising the temperature decreases the cadmium(I1) adsorption as would be expected both because of the decrease in the amount of specifically adsorbed iodide which is inducing the adsorption as well as the likelihood that the association reaction is itself temperature dependent. However, at the most anodic potential, where the iodide adsorption has increased to the point that the cadmium(I1) is being “squeezed out” the temperature coefficient becomes much smaller and may even change sign. This behavior is in accord with our interpretation of the origin of the maxima in plots of 2FrCd 1;s. potential or electrode charge. Dependence of 2FI’Cd on Halide Concentration, So long as the electrode is not saturated with respect to cadmium adsorption the dependence of 2FI’cd on the bulk halide concentration will reflect the changes in the bulk concentrations of the various cadmium-halide complexes as well as in the amount of adsorbed halide. Figure 4 shows a plot of 2FI’cd 1;s. bulk iodide concentration at constant values of adsorbed iodide in order to hold constant the second of these two effects. The maximum in this curve should then provide information on the bulk cadmium-iodide complex which is the most strongly adsorbed. To test this, the concentrations of the various cadmium-iodide complexes were calculated from the rather wide range of stability constants that have been reported for this system (18). Irrespective of the set of stability constants used, the cadmium(I1) species which corresponded best to the observed maximum in adsorption was CdIa. The data in Table I clearly indicate that the relationship between I’Cd and Cd(I1) concentration is nonlinear at the concentrations employed. The apparent failure of the 2FI’cd curve to decay as rapidly as the azcurve at higher iodide concentrations may merely reflect the fact that the maximum in the 2FI’cd curve would be very much higher if the electrode surface were not nearly saturated (see Figure 4) In the case of adsorption of cadmium(I1) from bromide,
.
(18) L. G. Sillen and A. E. Martell, “Stability Constants,” Special Publication No. 17, The Chemical Society, London, 1964.
C
5
0
Figure 4. a, (the fraction of cadmium present as CdI,) and 2FrCd at a constant rI- of 35 pC/cm2 as a function of the iodide concentration. [Cd(II)] = 0.8mM The plotted values of 2FrCd were read from the smooth curves connecting the measured points in Figure 3. The stability constants used to calculate the LY values are from (18,19)
the maximum adsorption occurs at higher bromide concentrations than was true with iodide (0.5 us. 0.05M). This behavior is in qualitative agreement with the existence of the neutral bromide complex at higher bromide concentrations. The scatter in reported bromide complex stability constants does not allow a more quantitative analysis. An apparently good correlation of the adsorption with the concentration of a particular bulk species is a necessary but not a sufficient basis for drawing a firm conclusion about the identity of the adsorbing species. The scatter in the values of the reported equilibrium constants and the necessary neglect of activity coefficients in our calculations are two reasons for being hesitant. However, compelling evidence in support of CdI2 and CdBr2 as the adsorbing species was obtained by examining the way in which the adsorption of the cadmium(I1) species changed the electronic charge on the electrode at constant potential. Effect of Adsorption on qm. The results presented in Tables I and I1 show that the change in the charge on the electrode, Aqm, brought about by the adsorption of the cadmium(I1)halide complex is small when compared with the amount of cadmium(I1) which is adsorbed except at the most anodic potentials. For example, the adsorption of 39.3 pC/cm2 of cadmium-iodide complex at -450 mV in 0.05M KI-0.8mM Cd(I1) alters qm by only 1.5 pC/cm2. According to standard double layer theory in the presence of specific adsorption (20) (19) A. M. Golub, Ukrain. Khim. Zhur., 19,205 (1953). (20) Delahay, P., “Double Layer and Electrode Kinetics,” Interscience, New York, 1965, Chapter 4.
-Ei,mV SCE 350
Table III. Temperature Dependence of 2FI’cd in O.1M KI; [Cd(II)I = 0.4mM T O C : 10 21 31 2FrCd (pC/cma)
VS.
400
450 500 550
40
19 30.8 35.8 37.8 38.3
22.7 31.6 35.5 36.8 36.9
21.9 30.1 33.7 34.9 34.4
VOL 40, NO. 12, OCTOBER 1968
22.0 29.4 32.6 32.5 32.6 1795
Table IV. Calculated Changes in the Rate of the Irreversible Reduction of Cr(&o)~'+ in 1M NaI Produced by the Hypothetical Adsorption of Various Cadmium Iodide Complexes 1M NaI, 0.05M Cr(ClO&, 0.01M HC104, l.OmM Cd(NO& E = -650 mV US. SCE; 2Frcd = 22 pc/Cm2 qm, pC/cm2 41-, pClcm2 qomplx,&/cm2 %, wC/cm2 -42," mV A b , mV (i/io)oalcdb -22.6 72.4 0 (CdIz) 0 1 7.4 - 30 -11.6 44.8 +11 (CdI+) 7.4 - 30 27.6 0.07 $22 (Cd+2) -0.6 7.4 - 30 0.3 72.1 0.001 7.4 - 30 -11 (cdI~-) -33.6 90.5 -18.1 5.7 -22 (CdId-') -44.6 7.4 - 30 105 -32.6 23 a
$z =
(4)
51.3 sinh-1- zq mV;
11.74
10
= caled
exp A*F [
RT
(i/io).,ba 0.65 0.65 0.65 0.65 0.65
(an - Z)]
10-2.5(A&/59.6). b
i/iois the ratio of the (mass transfer-corrected)Cr3+ reduction currents in the presence and absence of cadmium.
4m =
-(a
+ q2-7
(4)
where q1 is the charge carried by specifically adsorbed ions and q2-' is the ionic charge contained in the diffuse layer. The fact that such small values of Aqm are produced by the cad-
mium(I1) adsorption suggests that one of three possibilities may be the case: a) The adsorbing cadmium(IIkha1ide species is uncharged and it is adsorbed without major modification in the amount of initially adsorbed halide (or, therefore, of the ionic charge in the diffuse layer); b) an anionic cadmium(I1) complex is adsorbed which displaces an amount of the initially adsorbed halide bearing an approximately equivalent charge; c) a charged cadmium(I1) species is adsorbed without major modification in the amount of initially adsorbed halide thus producing a significant change in the ionic charge in the diffuse layer. Case (b) does not seem reasonable because the small values of Aqm persist even at relatively low total coverages of the electrode where, in fact, the cadmium adsorption is increased by additional halide adsorption (see Tables I and I1 at low bulk halide concentrations and the most negative potentials). Cases (a) and (c) should be experimentally distinguishable because the adsorption of cadmium according to scheme (a) would lead to no significant change in &, the potential drop across the diffuse layer, while scheme (c) would necessarily result in a large change in 42. We sought to examine the effect of cadmium adsorption on the +2 potential by measuring the rates of suitable irreversible electrode reactions in the halide-supporting electrolytes in the presence and absence of cadmium(I1). Although a number of electrode reactions were examined the only reaction that behaved irreversibly enough in iodide media at potentials anodic of the cadmium(I1) reduction to give useful data was the reduction of Cr(Hz0)~~'to C r ( H ~ 0 ) 6 ~ + in 1 M NaI adjusted to pH 2 with HC1O4. The half-wave potential of this couple in this media is -840 mV and the most cathodic potential at which the reduction of cadmium(I1) in 1 M NaI was still negligible was -650 mV. Therefore, the rate of the Cr(H@)63f reduction was measured polarographically using a 0.05M concentration of Cr(H20)6'+ in order to obtain substantial currents at -650 mV. The addition of 1mM cadmium(I1) produced only a small change in rate which was much less than would have resulted if the 22 pC/cm2 of absorbed cadmium-iodide species was even singly charged (see Table 1.V). The reduction of 1mM CO(NH&~+in 0.1M NaBr, 0.5M Na2SO4 at pH = 2 was examined with and without 1.OmM cadmium(I1) being present. The adsorption of the cad1796
ANALYTICAL CHEMISTRY
mium(I1) (cu. 7-8 pC/cm2) had no discernible effect on the rate of this irreversible reduction at potentials between - 100 and -400 mV. These results and the previously adduced dependence of the adsorption on the bulk halide concentration make it seem reasonable to propose that neutral Cd12 and CdBr2 are the primary adsorbed species. The reason why Cd12and CdBr2are adsorbed preferentially and the nature of the process by which the adsorbed halide ions induce the adsorption of the neutral cadmium-halide complexes may now be examined. Configuration of the Adsorbed Complex. There appear to be two possible configurations for the adsorbed cadmium halide species. One would result from the formation of covalent bonds between (in the case of iodide-induced adsorption) CdIz and the specifically adsorbed iodide ions to form trigonal (cd13-),d or tetrahedral (CdII2-),d species in the adsorbed state. The alternative possibility would not involve bond formation between cadmium and adsorbed iodide but rather an attraction between the Cd12and the electrode surface at sites not occupied by adsorbed iodide. Unfortunately, it is not possible to choose between these two possibilities on the basis of the observed magnitude of the adsorption because both lead to similar estimates of the amount of adsorption that would correspond to monolayer coverage of the electrode. Thus, 40-65 pC/cm2 of an adsorbed tetrahedral CdId+ species would result in a saturated electrode (the absolute saturation value is sensitive to the manner in which the packing of Cd14+ units is visualized) while an electrode with 35 pC/cm2 of adsorbed iodide and 40 pC/cm2 of adsorbed Cd12-Le., the linear molecule adsorbed parallel to the electrode surface-would be completely covered. Our data are compatible with both of these calculations ; however, we favor the configuration involving bond formation with the absorbed halide because it makes it much easier to understand the relationship between the cadmium adsorption and the amount of adsorbed halide (Figure 3). This configuration also rationalizes the fact that little adsorption of cadmium(I1) is observed from chloride media, even under conditions where there are substantial amounts of CdC12 in the solution and of specifically adsorbed chloride on the electrode. Covalent bond formation between the adsorbed anion (iodide, bromide, or chloride) and a cadmium complex involves the formation of a bridge between the electrode and the cadmium complex. As iodide and bromide are more polarizable ligands than chloride {for example, the complex ion [Hg-I-Hg]'+ has been identified in solution by Raman spectrometry while the corresponding Br and C1 complexes
Table V. Mole Ratio of Adsorbed Cadmium to Adsorbed Iodide. 650
Iodide conc., M 0.01
0.05
0.1
r+
450
400
350
16.5 0.82
22.5 0.69
26.5 0.63
31 .O 0.55
31.5 0.44
rr
...
28.0 0.65
30.5 0.62
34.0 0.58
38.5 0.50
..
rCd/2rI
...
rI
... ...
...
29 0.61
33 0.56
36.5 0.50
40 0.44
rr
25.5 0.49
32 0.41
36 0.36
40 0.30
46 0.34
...
rr
30 0.35
35 0.32
40 0.24
44
rCd/2rI a
- E , mV, L’S. SCE 500
... ...
rCd/2rI 1 .o
550
rCd/2rI
rCd/2rI 0.5
600
...
...
0.15
...
...
...
...
...
...
... ...
...
The cadmium concentration was 0.8mM. The ionic strength of all solutions was maintained at 1.0 with potassium nitrate.
* The units of rI are pC/cm2.
could not be observed (21)} one might expect more adsorption of cadmium(I1) from iodide or bromide media than from chloride. Similarly, thiocyanate, a good bridging ligand, would also be expected to induce strong adsorption of metal complexes, in agreement with experiment. Table V shows how the molar ratio of adsorbed cadmium to adsorbed iodide (Fcd/2r1) changes as the amount of adsorbed iodide is changed. Except at the lowest iodide coverages there is ample adsorbed iodide to permit the stoichiometric formation of CdIa+ on the electrode ( F c ~ / ~ I ’
