Isotope Effects in the Electromigration of Ions in Molten Tin Halides

I. A. Tikhomirov, S. I. Golyshev

(Presented by the scientific seminar of the Faculty of Physics and Engineering)

At present, electromigration in molten salts has been used to enrich the isotopes of a large number of elements [1]. Experimental data show that the isotope effect in electromigration depends to a considerable degree on various factors and, above all, on the temperature and chemical composition of the salt. A study of these dependences makes it possible to understand more fully the nature of isotope effects in electromigration. The present work was undertaken with the aim of investigating isotope effects for tin in molten halides.

In the work, reagents of the grades SnCl₂, SnBr₂, and SnI₂, “chemically pure,” were used. Each salt was kept in a desiccator over P₂O₅ for several days, then slowly heated in a Pyrex test tube to the melting temperature; the melt was evacuated and held for 2–3 hours at a pressure of 0.1 mm Hg, gradually raising the temperature to 580–600°C, until gas bubbles ceased to be evolved from the melt. In this process, intensive sublimation of the salts was observed only when heating powders that had not yet been subjected to melting, especially SnBr₂. The remelted salts did not sublime noticeably and melted at temperatures coinciding with the literature data [2].

The other salts and eutectic mixtures used in the work, LiCl + KCl and LiBr + KBr, were subjected to the same preliminary treatment. The prepared salts were stored in a desiccator over P₂O₅.

The general scheme of the apparatus is given in Fig. 1. It consists of the following principal elements: cell 1, quartz furnace 2, direct-current source 3, coulometer 4, voltage regulator 5, temperature controller 6 together with thermocouple 7 and millivoltmeter 8, instruments for monitoring the voltage and current in the electrolysis circuit and in the furnace-supply circuit. In Fig. 2 the cell for enrichment of the light isotopes of tin is shown separately. The cell is a tube 9 made of Pyrex glass, 160 mm long and 4 mm in diameter, filled with Pyrex-glass powder with a grain diameter of 0.2–0.3 mm. This packing was held in the tube by two fused-in Pyrex filters with a diameter of 8 mm. The cell was placed in a tubular furnace made of transparent quartz, to which power was supplied from voltage regulator 5. The required temperature interval was maintained with an accuracy of ±5°C and was additionally monitored by a millivoltmeter, by which the emf of a chromel–alumel thermocouple placed in the furnace was measured.

Power supply 3 was used to carry out the process ...

Fig. 1. General schematic of the apparatus

electrolysis. The electrolysis current was kept constant throughout the entire experiment with an accuracy of $\pm 5$ mA. The amount of electricity passed was measured with a copper coulometer. A carborundum (silite) element—a heater for electric furnaces—served as the cathode. A channel was drilled inside the silite rod through which, in some experiments, gaseous $\mathrm{Cl_2}$ or $\mathrm{Br_2}$ was supplied for the regeneration of the alkali metals deposited at the cathode. A silite or graphite rod served as the anode. Before the start of an experiment, the separating tube of the cell was filled with the molten salt. For example, in experiments with chlorides, the cell as a whole was placed in a furnace, and at a temperature of about $580^\circ\mathrm{C}$ the molten salt $\mathrm{SnCl_2}$ was introduced into its wide limb, so that the salt filled the separating tube and rose $1$–$2$ mm above the upper filter. Then, by introducing the eutectic melt $\mathrm{LiCl}+\mathrm{KCl}$ into the cathode compartment, the level of the $\mathrm{SnCl_2}$ melt in the separating tube was lowered to $10$–$15$ mm below the filter. The electrolysis current was switched on after the anode compartment had been filled with the molten salt $\mathrm{ZnCl_2}$.

Depending on the composition of the salt, experiments were carried out with the following electrolytic chains:

1. $\quad +\mathrm{C}\,|\,\mathrm{ZnCl_2}\,|\,\mathrm{SnCl_2}\,|\,\mathrm{Sn}\,-$

2. $\quad +\mathrm{C}\,|\,\mathrm{ZnCl_2}\,|\,\mathrm{SnI_2}\,|\,\mathrm{Sn}\,-$

3. $\quad +\mathrm{C}\,|\,\mathrm{ZnCl_2}\,|\,\mathrm{SnCl_2}\,|\,\mathrm{LiCl}+\mathrm{KCl}\,|\,\mathrm{PbCl_2}\,|\,\mathrm{Pb}\,|\,-$

4. $\quad +\text{silite}\,|\,\mathrm{ZnBr_2}\,|\,\mathrm{SnBr_2}\,|\,\mathrm{LiBr}+\mathrm{KBr}\,|\,(\text{silite}+\mathrm{Br_2})\,|\,-$

5. $\quad +\text{silite}\,|\,\mathrm{ZnCl_2}\,|\,\mathrm{SnCl_2}\,|\,\mathrm{LiCl}+\mathrm{KCl}\,|\,(\text{silite}+\mathrm{Cl_2})\,|\,-$

Chains 1 and 2 were used for enrichment of the heavy isotopes of tin, and chains 3, 4, and 5 for enrichment of the light isotopes of tin.

In all experiments, the salt-interface boundaries in the separating tube formed 1.0–1.5 hours after the electrolysis current was switched on and co-

Table 1

Table 2

were maintained throughout the entire experiment. The conditions under which the experiments were carried out are summarized in Tables 1 and 2.

Table 1 characterizes the main parameters of the experiments on the separation of tin isotopes during electromigration of ions in the salt SnCl₂, and Table 2, in the salts SnBr₂ and SnI₂.

After the electrolysis current was switched off, the cooled separating tube was cut into pieces 15–20 mm long, starting from the salt boundary. Each such sample was subjected to chemical and mass-spectrometric analysis. The salt was dissolved in a small amount of water acidified with HCl. The total amount of salt in the sample was determined by difference: the weight of the sample minus the weight of the packing and of the piece of glass tube. The tin content in the sample was then found by iodometric titration.

In some samples adjacent to the boundary with neighboring salts, the total chlorine content was additionally determined argentometrically. Isotopic analysis was carried out on an MI-1305 mass spectrometer with a furnace ion source and automatic recording of mass spectra by means of an EPP-09 electronic potentiometer. After processing the mass-spectrometric analysis data, the results were summarized in Tables 3 and 4.

Fig. 2. Electrolytic cell for enrichment of the light isotopes of tin

Table 3 gives the results of analyses for SnCl₂, and Table 4 for SnBr₂ and SnI₂.

Table 3

Tin isotope content in atomic % in SnCl₂ samples with altered isotopic composition. The samples are numbered from the SnCl₂|LiCl+KCl boundary in experiments 1 and 2 and from the SnCl₂|ZnCl₂ boundary in experiments 3 and 4. Each result is the mean of 8–10 measurements; the root-mean-square error of an individual measurement is ±2%

Table 4

Tin isotope content in atomic % in SnBr₂ and SnI₂ samples after electrolysis. The samples are numbered, respectively, from the SnBr₂|LiBr+KBr and SnI₂|ZnI₂ boundaries. The root-mean-square error of an individual measurement on the mass spectrometer is 2%; each result is the mean of 8–10 measurements

The relative difference in the mobilities of the isotopic tin ions was calculated by the formula [3]

\[ \frac{\Delta U}{U} = \left( \frac{{}^{i}N}{{}^{i}N_{0}} - \frac{{}^{\kappa}N}{{}^{\kappa}N_{0}} \right) \cdot \frac{FN_{\mathrm{Cl}}}{Q}. \]

Here \({}^{i}N\) and \({}^{\kappa}N\) are the total amounts of the \(i\)-th and \(\kappa\)-th isotopes, in g, for samples whose isotopic composition changed after electrolysis in comparison with the initial salt. \({}^{i}N_{0}\) and \({}^{\kappa}N_{0}\) are the total amounts of the \(i\)-th and \(\kappa\)-th isoto-

... in g in the sample before electrolysis. \(N_{\mathrm{Cl}}\) is the number of gram-equivalents of chlorine in the sample. \(Q\) is the amount of electricity passed. \(F\) is the Faraday constant.

The magnitude of the mass effect was calculated by the formula:

\[ \mu=\frac{\Delta U}{U}\bigg/\frac{\Delta m}{m}, \]

where \(\frac{\Delta m}{m}\) is the relative mass difference, and \(\frac{\Delta U}{U}\) is the relative difference in the mobilities of the \(i\)-th and \(k\)-th isotopes. For one and the same experiment, the quantities \(\frac{\Delta U}{U}\) and \(\mu\) were found for the isotope pairs 112—119, 116—119, 117—119, 118—119, 120—119, 122—119, 124—119. Tables 1 and 2 give the arithmetic mean values of \(\mu\) for these isotope pairs. Comparing the values of \(\mu\) for \(\mathrm{SnCl_2}\), \(\mathrm{SnBr_2}\), and \(\mathrm{SnI_2}\) at a temperature of \(470^\circ\mathrm{C}\), it is easy to see that the regularity mentioned above—the increase in the values of \(\mu\)—is well fulfilled for tin isotopes. Thus, the “rule for halides” [4] is also confirmed in the case of tin salts. This dependence most accurately falls on a single curve when the values of \(\mu\) and \(\frac{r_k}{r_a}\) are compared, where \(r_k\) and \(r_a\) are the radius of the cation and the radius of the anion. The increase in \(\mu\) with a decrease in the ratio \(\frac{r_k}{r_a}\) is shown in Fig. 3. A number of authors associate the decrease in \(\mu\) with an increase in the ratio

Fig. 3. Mass effect for divalent cations in molten halides

\[ \frac{m_k}{m_a}, \]

where \(m_k\) is the mass of the cation, and \(m_a\) is the mass of the anion.

However, such a dependence is not observed for every series of cations. Apparently, in this case the decisive role in the mass effect is played by the structure of the molten salts, or even by the structure of the given salt in the crystalline state. With comparable atomic weights

metals forming salts with a layered lattice have larger values of \(\mu\) in comparison with metals forming salts with a coordination lattice. Klemm [5] proposed a method of “leveling” the isotopic composition of the separated element in each enriched sample, assuming that the enrichment of a given isotope is proportional to the difference between its mass \({}^{i}m\) and the mean mass number of the element. Following this procedure, it can be shown that, in the case considered, the quantity

\[ \ln \frac{{}^{i}\Gamma_n}{{}^{i}\Gamma_0} \]

is a linear function of

\[ \frac{{}^{i}m}{m_0} - 1, \]

where

\[ m_0 = \sum_i {}^{i}\Gamma_0\,{}^{i}m_0, \]

\({}^{i}\Gamma_n\) and \({}^{i}\Gamma_0\) are the contents of isotope \(i\), in atomic percent, respectively in the enriched sample and in the natural isotopic mixture. In Fig. 4, for each of the enriched samples, the averaged straight lines calculated from the formula are shown:

\[ \ln \frac{{}^{i}\Gamma_n}{{}^{i}\Gamma_0} = \frac{ \sum_i {}^{i}\Gamma_0 \left( \frac{{}^{i}m}{m_0} - 1 \right) \ln \frac{{}^{i}\Gamma_n}{{}^{i}\Gamma_n} }{ \sum_i \left( \frac{{}^{i}m}{m_0} - 1 \right)^2 } \cdot \left( \frac{{}^{i}m}{m_0} - 1 \right). \]

Fig. 4. Comparison of the results of mass-spectrometric analysis with the “leveled” isotopic composition of enriched samples:
a — enrichment of the light isotopes of tin in \(\mathrm{SnBr_2}\),
b — enrichment of the heavy isotopes of tin in \(\mathrm{SnI_2}\).

The experimental points correspond to the samples:

\[ \odot \;-\; \text{No. 1}, \qquad \triangle \;-\; \text{No. 2}, \]

\[ \square \;-\; \text{No. 3}, \qquad \oplus \;-\; \text{No. 4}. \]

It is evident that the experimental points found from the data of Table 2 agree quite well with their “leveled” values. This “leveling” can serve as a criterion for the reliability of the results of the mass-spectrometric analysis of the isotopic composition of the separated element in enriched samples.

REFERENCES

1. M. Gatliih, A. Roy, Electrochemical Technology, 2, 85–89 (1964).
2. Handbook of the Chemist, Khimiya Publishing House, Leningrad–Moscow, vols. 1, 2, p. 227, 1964.
3. Klemm, Z. Naturforschung, 1a, 252–7 (1946).
4. A. Klemm, J. Chim. phys. et phys-chim. biol., 60, No. 1, 2, 237–44 (1963).
5. A. Klemm, H. Hintenberger, W. Seelmann-Eggebert, Z. Naturforschung, 3a, 172–6 (1948).