STUDY OF THE BEHAVIOR OF ULTRAMICROCONCENTRATIONS OF INDIUM AND CADMIUM IN THE NONAQUEOUS MIXTURE SiCl₄ (6.3 mol %)—n-C₃H₇OH BY AMALGAM POLAROGRAPHY WITH ACCUMULATION

Yu. A. Karbainov, A. G. Stromberg

(Presented by the Scientific-Methodological Seminar of the Faculty of Chemical Technology)

Earlier [1] we showed that high-purity silicon tetrachloride in a mixture with n-propyl alcohol forms a system with a maximum electrical conductivity of \(12.86 \cdot 10^{-3}\ \Omega^{-1}\,\text{cm}^{-1}\) at 6.3 mol % \(SiCl_4\). Such a value of the electrical conductivity fully satisfies the requirements imposed on polarographic analysis. It was also shown that the tendency of silicon tetrachloride toward hydrolysis in this system at the conductivity maximum decreases sharply. Thus, for example, when a mixture of \(SiCl_4\) (6.5 mol %)—\(n\text{-}C_3H_7OH\) was stored for several days, no turbidity of the solution and no change in its electrical conductivity were observed. Our investigations [2] on the behavior of ultramicroconcentrations of antimony, bismuth, and tin in the system \(SiCl_4\) (6.3 mol %)—\(n\text{-}C_3H_7OH\) showed that this system is suitable for the accelerated determination of \(10^{-6} \div 10^{-7}\%\) impurities in high-purity \(SiCl_4\) by the method of amalgam polarography with accumulation. At the same time, the duration of such an analysis is shorter than, for example, that of chemical-spectral analysis [3]. The present work is devoted to a comparative study of the behavior of ultramicroconcentrations of indium and cadmium in the nonaqueous mixture \(SiCl_4\) (6.3 mol %) \(n\text{-}C_3H_7OH\) and in aqueous solutions, and also to clarifying the possibility of determining indium and cadmium in the nonaqueous mixture and the maximum sensitivity that can be achieved thereby. Measurements were carried out on a self-recording polarograph of the firm “Orion,” Hungary. Mercury at the bottom of the electrolyzer served as the anode in investigations in the nonaqueous mixture 6.3 mol % \(SiCl_4\)—\(n\text{-}C_3H_7OH\), and a saturated calomel electrode in investigations in aqueous solutions. The cathode was a hanging mercury drop, obtained by electrolysis of a saturated solution of mercurous nitrate \((1+)\) on a platinum contact (sealed into a glass tube) with a current of 30 mA for 1 minute. The radius of the mercury drop was then equal to 0.05 cm. In the work an electrolyzer of conventional design [13] was used, into the bottom of which a platinum contact had been sealed. To prepare the initial solutions, salts \(Zn(NO_3)_3\) and \(CdCl_2\) of “chemically pure for analysis” grade were used. Solutions of indium and cadmium were prepared in n-propyl alcohol (0.01 M solutions). The investigations were carried out using solutions with concentrations of \(10^{-4} \div 10^{-6}\ M\). These solutions were prepared by successive dilution of the initial 0.01 M solutions. Silicon tetrachloride and n-propyl alcohol were used in “especially pure” grade. No additional dehydration of \(n\text{-}C_3H_7OH\) was carried out [2].

Because of the proximity of the hydrogen discharge potential in the nonaqueous mixture SiCl₄ (6.3 mol %)—n-C₃H₇OH to the dissolution potential of indium and cadmium (−0.53; −0.50; −0.62 V), and, in connection with this, the large value of the residual current, the determination of indium and cadmium proved difficult. However, upon partial neutralization of the background with pyridine (0.5 ml of C₅H₅N per 5 ml of nonaqueous mixture), the hydrogen discharge potential shifts from −0.53 to −0.85 V, which makes possible quantitative determinations of such electronegative elements as indium and cadmium. Earlier [2] we suggested that, when silicon tetrachloride is mixed with n-propyl alcohol at room temperature, the following reaction occurs:

\[ \mathrm{SiCl_4 + n\text{-}C_3H_7OH \rightleftarrows [SiCl_m(C_3H_7OH)_n]^{(4-m)+} + (4-m)Cl^-} \]

However, it is known from the literature [6] that an esterification reaction of silicon tetrachloride with n-propyl alcohol may also take place:

\[ \mathrm{SiCl_4 + 4\,n\text{-}C_3H_7OH \rightleftarrows Si(OC_3H_7)_4 + 4HCl.} \]

At room temperature this reaction proceeds very slowly. But when SiCl₄ and n-C₃H₇OH are mixed, the mixture heats up, and a noticeable amount of hydrochloric acid may form in it as a result of acceleration of this reaction at elevated temperature. The hydrogen chloride formed increases the acidity of the solution. When pyridine is added, the dissolved hydrogen chloride is bound by it, as a result of which the acidity of the solution decreases and the hydrogen discharge potential shifts in the more negative direction. The liquid mixture was prepared as follows: 0.5 ml of SiCl₄ was mixed with 4.5 ml of n-C₃H₇OH (the reaction mass partially heated up in the process). After cooling, 0.5 ml of C₅H₅N, previously thoroughly dehydrated, was added dropwise to this mixture with stirring. The prepared alcoholic solution of SiCl₄ can be stored for a long time without decomposition.

Oxygen, as before, was removed from the nonaqueous mixture by passing a stream of nitrogen through it for 1 hour. The nitrogen was first purified from traces of moisture and oxygen by passing it through a tube heated to 180°, filled with copper turnings, and also through a pyrogallol solution and a solution of n-propyl alcohol (to saturate the nitrogen with alcohol vapors). After each polarogram was recorded, additional dissolution of the metal from the drop was carried out with stirring for 2 min at potentials lower than the dissolution potentials of the corresponding metals. All potential values in the figures and text are given relative to mercury at the bottom of the electrolyzer in studies in the nonaqueous mixture, and relative to the saturated calomel electrode in studies in aqueous solutions.

Table 1 gives the characteristics of these waves: the width and effective width of the half-waves of indium and cadmium. The effective width of the half-waves was calculated by the formula [9]

\[ \frac{q}{l}=\frac{S_{\mathrm{eff}}}{W}, \]

where \(q\) is the area under the wave, in coulombs; \(l\) is the height of this wave, in amperes; \(\omega\) is the rate of change of potential, V/sec.

The \(l\)—\(c\) data show that there is a direct proportional relationship between the height of the anodic wave (\(l\)) and the concentration of metal ions in the solution.

The maximum sensitivity of the determination of indium and cadmium in the nonaqueous solvent (if as the minimum wave height after which—

…for which quantitative determinations are possible, if 5 mm of the instrument scale is taken (at the given sensitivity), are respectively equal to \(3\cdot10^{-7}\) mol/l and \(1\cdot10^{-7}\) mol/l, which, recalculated for pure \(\mathrm{SiCl_4}\), taking into account its content in the mixture (6.3 mol.%), amounts to \(2.5\cdot10^{-5}\%\) and \(8\cdot10^{-6}\%\). The maximum sensitivity in the determination of indium and cadmium in the case of aqueous solutions is respectively equal to: \(5\cdot10^{-8}\) mol/l, \(1\cdot10^{-7}\) mol/l.

Table 1

Values of the width and effective width of the half-peak of indium and cadmium

A comparison of the sensitivities for indium and cadmium in aqueous solutions and in the nonaqueous solvent shows that the slope coefficients for \(\mathrm{Cd}(2+)\) on the plots “\(I-c_1\)” and “\(I-\tau_{\mathrm{e}}\)” in aqueous and nonaqueous backgrounds are almost identical; however, for \(\mathrm{In}(3+)\) the slope coefficients on the plots “\(I-c_1\)” and “\(I-\tau_{\mathrm{e}}\)” in the aqueous background \(0.1\,N\,\mathrm{NaF}\) and in the nonaqueous solvent differ by approximately a factor of 3. Such a difference in the behavior of \(\mathrm{In}(3+)\) in the indicated backgrounds may be explained by three reasons. First, by differences in the mechanisms of anodic oxidation of indium from the amalgam [7]:

\[ \mathrm{In}\to \mathrm{In}(3+) + 3(e); \]

\[ \mathrm{In}\to \mathrm{In}(+) + (e). \]

The difference in the number of electrons participating in the anodic process of dissolution of the indium amalgam may cause a threefold difference in the slope coefficients on the plots “\(I-c_1\)” or “\(I-\tau_{\mathrm{e}}\)” in different backgrounds. Determination of the number of electrons was carried out by the method of “standard” areas [12]. In this case, cathodic accumulation of indium was carried out in a background of \(1\,M\,\mathrm{NaCl}\), \(0.1\,N\,\mathrm{HClO_4}\), \(\mathrm{NaClO_4}\) [11], and anodic dissolution was carried out in one of three backgrounds: \(1\,M\,\mathrm{NaCl}\), \(0.1\,M\,\mathrm{HClO_4}\), \(\mathrm{NaClO_4}\), \(\mu=1.42\); \(0.1\,N\,\mathrm{NaF}\), or \(\mathrm{SiCl_4}\) (6.3 mol.%)—\(n\)-\(\mathrm{C_3H_7OH}\). It is known that [11] the anodic process of indium in a background of \(1\,M\,\mathrm{NaCl}\), \(0.1\,N\,\mathrm{HClO_4}\), \(\mathrm{NaClO_4}\) proceeds with the participation of 3 electrons. Comparison of the quantity of electricity expended on the anodic dissolution of indium in the above-indicated backgrounds and calculated from the area under the anodic peak (Table 2) shows that the anodic process of indium in backgrounds of \(0.1\,N\,\mathrm{NaF}\) and \(\mathrm{SiCl_4}\) (6.3 mol.%)—\(n\)-\(\mathrm{C_3H_7OH}\) also proceeds with the participation of three electrons. Thus, the fact indicated above for indium cannot be explained by a difference in the number of electrons participating in aqueous and nonaqueous backgrounds.

Second, the difference in the slope coefficients may be associated with a difference in the diffusion coefficients of indium in different sol…

solvents. For this purpose, cathodic waves recorded with a dropping mercury electrode for In(3+) on the indicated backgrounds under identical conditions were compared. The height of the indium wave on the nonaqueous background SiCl₄ (6.3 mol %)—n-C₃H₇OH proved to be approximately 2.4 times smaller than the height of the cathodic indium wave on the aqueous background 0.1 N NaF (16.8; 40.5 μA). This difference is due primarily to the difference

Table 2

Comparative polarographic characteristics of indium
on aqueous and nonaqueous backgrounds

¹) \(q\) is the quantity of electricity (coulombs) expended for the anodic oxidation of indium on the indicated backgrounds.

in the diffusion coefficients of indium on these backgrounds, with the same number of electrons participating in the anodic dissolution of indium from the amalgam. Table 3 gives the values of the constants of the cathodic peaks

Table 3

Amalgam-polarographic characteristics of indium and cadmium

¹) \(K_1\) is the electrolysis constant,

\[ K_1=\frac{q}{SC_1\tau_{\mathrm{e}}}, \]

where \(q\) is the area under the peak, in coulombs; \(\tau_{\mathrm{e}}\) is the electrolysis time, in sec.;

²) \(K_2^{\text{st.}}\) is the anodic-peak constant under standard conditions,

\[ K_2^{\text{st.}}=K_2\frac{0.1}{\sqrt{W}}, \]

where

\[ K_2=\frac{I_2}{SC_2}; \]

³) \(K_3^{\text{st.}}\) is the cathodic-peak constant under standard conditions,

\[ K_3^{\text{st.}}=K_3\frac{0.1}{\sqrt{W}}, \]

where

\[ K_3=\frac{I_3}{SC_1}; \]

⁴) \(\gamma\) is the shape coefficient;

⁵) \(\gamma_{\text{p}}^{\text{st.}}\) is the stirring coefficient under standard conditions,

\[ \gamma_{\text{p}}^{\text{st.}}=\frac{500z}{K_3^{\text{st.}}}. \]

$In(3+)$ on the indicated supporting electrolytes [10]. Their difference (226, 114) is also due to the difference in diffusion coefficients with the same number of electrons $(z = n)$. The values of the diffusion coefficients calculated from the Ilkovič equation are given in Table 2.

The depth of the anodic peaks of indium and cadmium is directly proportional to the electrolysis time $(\tau_{\mathrm{e}})$. This is in agreement with the theory [4], since under these conditions the dimensionless parameter $b$ is small $(b < 0.1)$ and depletion of the solution is practically not observed.

Between the depth of the anodic peak and the electrolysis potential $(\varphi_{\mathrm{e}})$ in the nonaqueous solution, a linear dependence is observed. The curves $I—\varphi_{\mathrm{e}}$ have a sharp drop at $\varphi_{\mathrm{e}} = -1.4$ V. The reason for this phenomenon is not yet clear. In the case of aqueous solutions, the dependence $I—\varphi_{\mathrm{e}}$ for indium and cadmium has the form of a polarogram. This dependence was studied in the potential interval $-0.7 \div -1.4$ V. In this potential interval, the dependence of the peak depth on the electrolysis potential obeys the theoretical relation [5] (for irreversible electrode processes):

\[ I = \frac{I^{\max}} {l \dfrac{\alpha}{b'}(\varphi - \varphi^{1/2}) + 1} \cdot \]

In appearance, the curves resemble a cathodic polarogram on a dropping mercury electrode, and the half-wave potential approximately coincides with the potential of half the height of the rise of the curve $I = f(\varphi_{\mathrm{e}})$. The values of the half-wave potentials in our case for indium and cadmium are, respectively: $-0.92$; $-0.87$ V. The values of the half-height potential of the curve $I = f(\varphi_{\mathrm{e}})$ for indium and cadmium are: $-1.0$; $-1.1$ V. The reason for this discrepancy is not yet clear. Because of the closeness of the deposition potentials of indium and cadmium ($-0.5$ and $0.62$ V), their simultaneous determination on the nonaqueous supporting electrolyte under consideration on a hanging mercury drop is possible when the ratio of the concentrations of indium and cadmium is not less than $1:3$. It has been established that indium and cadmium on the nonaqueous supporting electrolyte are oxidized irreversibly at the electrode: $\Delta \varphi_n^{\mathrm{ak}} In = -0.07$; $\Delta \varphi_n^{\mathrm{ak}} Cd = -0.09$ (V). The expected theoretical values are $\Delta \varphi_n^{\mathrm{ak}} In = -0.018$; $\Delta \varphi_n^{\mathrm{ak}} Cd = -0.027$ (V).

Table 3 gives the amalgam-polarographic characteristics of indium and cadmium on the supporting electrolytes studied. Their calculation is based on three formulas [8]:

\[ I_1 = K_1 S C_1; \]

\[ I_2 = K_2 S C_2; \]

\[ I_3 = K_3 S C_1. \]

Here $I_1$ is the electrolysis current; $I_2$ is the anodic-peak current; $I_3$ is the cathodic-peak current; $K_1$, $K_2$, $K_3$ are, respectively, their constants; $S$ is the surface of the mercury drop; $C_1$ is the concentration of metal ions in solution; $C_2$ is the concentration of metal atoms in the mercury amalgam.

The characteristics were calculated under standard conditions, which were taken as: $K_1^{\mathrm{st}} = 500z$, where $z$ is the valence of the element; $\gamma_n^{\mathrm{st}}$ is the concentration coefficient,

\[ \gamma_n^{\mathrm{st}} = \frac{c_2}{c_1} = 100; \]

the rate of change of potential

\[ W^{\mathrm{st}} = 0.01\ \text{V/sec};\quad t^\circ C = 25; \]

Conclusions

1. The dependence of the depth of anodic notches in the method of amalgam polarography with accumulation for indium and cadmium on the concentration of metal ions in solution, the electrolysis potential, and the electrolysis time was studied against the background of the nonaqueous solvent $\mathrm{SiCl_4}$—$n$-$\mathrm{C_3H_7OH}$ and aqueous supporting media.

2. The results obtained have been discussed.

References

1. Yu. A. Karbainov, A. G. Stromberg. Izv. TPI (in press).

2. Yu. A. Karbainov, A. G. Stromberg. Journal of Analytical Chemistry (in press).

3. A. F. Pchelintseva, N. A. Rakov, A. P. Smosareva. Zavodskaya laboratoriya, 28, 677, 1962.

4. A. G. Stromberg. Izv. SO AN SSSR, No. 5, 76, 1962.

5. M. S. Zakharov. Zavodskaya laboratoriya, No. 1, 14, 1964.

6. M. Bazhant, V. Khvalovskii, I. Ratouskii. Silikony, Goskhimizdat, p. 109, 1960.

7. L. F. Kozin, N. N. Tananaeva. Proceedings of the Institute of Chemical Sciences of the Kazakh SSR, Alma-Ata, 9, 143, 1962.

8. M. S. Zakharov, A. G. Stromberg. Journal of Analytical Chemistry (in press).

9. P. Delakhei. New Instruments and Methods in Electrochemistry, IL, 1957.

10. I. Randles, Analyst, 72, 301, 304, 1947.

11. T. A. Kryukova, S. I. Sinyakova, T. V. Aref’eva. Polarographic Analysis, Goskhimizdat, p. 676, 1959.

12. A. A. Kaplin, E. A. Zakharova. Izv. TPI (in press).

13. A. G. Stromberg, E. A. Stromberg. Zavodskaya laboratoriya, No. 3, 27, 1961.