STUDY OF THE CONDITIONS FOR PHOTOMETRIC DETERMINATION OF TRACE SILICON WITH A SENSITIVITY OF \(10^{-5}\)—\(10^{-6}\%\)

G. A. Detkova, Yu. L. Lel’chuk

(Presented by the scientific seminar of the Department of Analytical Chemistry)

The requirements for the sensitivity of methods for determining microimpurities in high-purity materials are continuously increasing. Traces of silicon are usually determined photometrically in the form of silicomolybdenum blue after its extraction with one or another organic extractant \([1—5]\). However, the sensitivity of these methods with a sample weight of the substance being analyzed of 1 g does not exceed \(10^{-4}\%\).

In silicomolybdenum blue, 1 atom of silicon corresponds to 12 atoms of molybdenum, and the weight content of the latter is 41 times greater than that of silicon. We have studied the conditions for the quantitative separation of molybdenum from silicomolybdenum blue and the indirect photometric determination of silicon from the separated molybdenum after converting it into a rhodanide complex. This same approach was previously used by us \([6—8]\) to increase the sensitivity of determining trace phosphorus in high-purity tin.

As a result of the investigation, a procedure was developed for the photometric determination of trace silicon in high-purity tin with a sensitivity of \(10^{-5}\)—\(10^{-6}\%\).

The method is based on decomposition of a weighed portion of tin with a mixture of hydrochloric and hydrobromic acids and bromine, removal of tin and the arsenic contained in it by triple evaporation with hydrochloric acid, conversion of silicon into silicomolybdenum blue, extraction of it with isoamyl alcohol, destruction of the blue, and photometric measurement of the liberated molybdenum in the form of a rhodanide complex.

The conversion factor from molybdenum to silicon is 0.0244. The main operations were carried out in polyethylene, fluoroplastic, and platinum vessels. Decomposition of the weighed portion and evaporation were performed under a lamp in an organic-glass box. All reagents were of chemically pure grade. Hydrochloric acid and ammonium hydroxide hydrate were purified by isothermal distillation at ordinary temperature in a polyethylene apparatus; hydrobromic acid, by distillation in a fluoroplastic distillation apparatus; bromine, by fourfold distillation in a quartz apparatus. Sodium hydroxide was prepared from sodium and ultrapure water in a polyethylene vessel. Ammonium molybdate was purified from traces of phosphorus, arsenic, and silicon by the method previously described by us \([9]\).

In the work, aqueous solutions were used of sulfuric acid, 2 and 10 N; hydrochloric acid, 3.6 and 9 N; sodium hydroxide, 20%; ammonium molybdate, 10%; potassium rhodanide, 50%; copper sulfate, 0.04%; citric—

acid, 50%; stannous chloride, 0.5%, in 2N hydrochloric acid. The ammonium buffer solution was prepared by mixing equal volumes of 0.1N solutions of ammonia and ammonium chloride.

The initial standard silicon solution with a titer of 0.5 mg/ml was prepared from chemically pure sodium silicate. The titer was established by the gravimetric method [10]. Working standard silicon solutions with titers of 10 and 1 µg/ml were prepared on the day of use by the corresponding dilution of the initial solution with water.

Course of analysis. To a tin sample within 1 g in a fluoroplastic cup are added 2 ml of 9N hydrochloric acid, 2.5 ml of hydrobromic acid, and, carefully dropwise, 2 ml of bromine. After the tin has dissolved, the solution is evaporated at 90–100°C. Then 1 ml portions of 6N hydrochloric acid are added twice, each time evaporating to dryness. The dry residue is dissolved with heating in 10 drops of caustic soda; the solution is transferred with 5–6 ml of water into a polyethylene beaker, neutraliz—

Table 1

Verification of the accuracy of the method on standard silicon solutions

Table 2

Results of parallel determinations of silicon
in high-purity tin

—ized with sulfuric acid, 0.75 ml of ammonium molybdate solution is added, the mixture is heated for 5 min, 1 ml of citric acid solution is added, and it is again heated for 5 min. The solution is cooled, 1.5 ml of 10N sulfuric acid is added, it is transferred into a separatory funnel with a capacity of 20–25 ml, and the silicon complex is extracted with two 2-ml portions of isoamyl alcohol.

To the combined extracts add 1 ml of 2N sulfuric acid, 0.2 ml of stannous chloride solution, and shake for 30 sec. After separation of the layers, the aqueous layer is drained off. The isoamyl extract is washed 4 times with 2N sulfuric acid, the stem of the funnel is rinsed with bidistillate, and the silicomolybdenum blue is re-extracted with three portions of ammonia buffer, 1.5 ml each. The re-extract is collected in a 25 ml volumetric flask, 1 ml of sodium hydroxide is added, and the solution is boiled for 5 min to destroy the blue and liberate molybdenum.

To the cooled solution add 9 ml of 6N hydrochloric acid, 0.5 ml of 0.04% copper sulfate solution, 5 ml of potassium thiocyanate, and 3N hydrochloric acid to the mark, and mix. After 15 min the optical density of the solution is measured in cuvettes with a 5 cm path length using a blue light filter with a transmission maximum at 453 mµ. The blank solution is prepared in the same way as in constructing the calibration curve. A blank experiment is carried through all stages of the analysis; its optical density is subtracted from the optical density of the solution being analyzed. The weight content of silicon in tin is found by multiplying the weight of molybdenum by the conversion factor 0.0244.

Construction of the calibration curve. Into 5 volumetric flasks of 25 ml capacity are introduced 5, 10, 15, 20, and 25 μg of molybdenum, respectively, 1 ml of sodium hydroxide, 9 ml of 6N hydrochloric acid, 0.5 ml of copper sulfate solution, 5 ml of potassium thiocyanate, and 3N hydrochloric acid to the mark, and the contents are mixed. After 15 min the optical density is measured on an FEK-N-57 in cuvettes with a 5 cm path length using a blue light filter with a transmission maximum at 453 mµ. The blank solution is prepared in a 50 ml flask by adding double the volume of all reagents, except for the standard molybdenum solution.

Conclusions

A procedure has been developed for determining traces of silicon in high-purity tin with a sensitivity of \(10^{-5}\)—\(10^{-6}\%\) for a one-gram tin sample.

The duration of the analysis is 5 hours. The relative error is within \(\pm 15\%\).

LITERATURE

1. Colorimetric (photometric) methods for determining nonmetals. Translated from English, edited by A. I. Busev. Izd. IL, 34–68, Moscow, 1963.

2. R. I. Alekseev. Quantitative separation of anions of phosphoric, arsenic, and silicic acids by means of selective extraction. Zavodskaya laboratoriya, 2–3, 122 (1945).

3. N. A. Filippova, L. I. Kuznetsova. Colorimetric determination of small amounts of phosphorus, arsenic, and silicon in copper. Zavodskaya laboratoriya, No. 5, 536–546 (1950).

4. G. N. Harrison. Photometric determination of traces of silicon in nickel. Metallurgia, 66, No. 398, 300 (1962).

5. Sono Kin’ya, Watanabe Hiroto, Mitsukami Yoshimiti. Extraction-photometric determination of ultramicroamounts of silicon in water. Referativnyi zhurnal. Khimiya, 20 G, 124 (1963).

6. V. B. Sokolovich, Yu. L. Lel’chuk, G. A. Detkova. Reports at the All-Union Conference on the Production and Analysis of Substances of Special Purity. Gorky, 57, 1963.

7. Yu. L. Lel’chuk, V. B. Sokolovich, G. A. Detkova, I. P. Onufrienok, Z. N. Tereshkova, V. A. Ivashina. Reports at the All-Union Conference on Analytical Chemistry of Semiconductor Substances. Kishinev, 47, 1965.

8. Yu. L. Lel’chuk, V. B. Sokolovich, G. A. Detkova. Methods for analyzing high-purity substances. Academy of Sciences of the USSR, “Nauka” Publishing House, 349, Moscow, 1965.

9. Yu. L. Lel’chuk, V. B. Sokolovich, G. A. Detkova. Preparation and analysis of substances of special purity. “Nauka” Publishing House, 268, Moscow, 1966.

10. V. F. Hillebrand et al. Practical guide to inorganic analysis. Goskhimizdat, 691, Moscow, 1957.