On the Question of Separating Interfering Metals in the Photocolorimetric Determination of Small Amounts of Cobalt in Ores with β-Nitroso-α-naphthol

Yu. L. Lelchuk, L. L. Skripova

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

Earlier we showed that an alcoholic solution of β-nitroso-α-naphthol can be successfully used for the photocolorimetric determination of thousandths and ten-thousandths of a percent of cobalt in ores after precipitation of interfering metals with cupferron, extraction of the cupferrates with chloroform, and destruction of the excess cupferron by evaporation with \(\mathrm{HNO_3}\) and \(\mathrm{HClO_4}\) [1].

In our experiments we used Boyland’s data on the spectrophotometric determination of small amounts of cobalt in biological materials [2], Rooney’s procedure for the spectrophotometric determination of cobalt in cast iron and steel [3], and the materials of Peshkova and Bochkova on the comparative characteristics of colorimetric methods for determining cobalt with β-nitroso-α-naphthol and α-nitroso-β-naphthol [4].

Our experiments showed that preliminary separation of interfering metals is needed mainly in order to eliminate the harmful influence of iron on the determination of small amounts of cobalt by this method. The complexes of the other metals with β-nitroso-α-naphthol do not interfere with the determination of cobalt, since they are completely destroyed when the benzene extract is treated with acetic acid (1 : 5) and a normal alkali solution.

The inconveniences associated with the use of chloroform as an extractant for cupferrates and with removing excess cupferron by evaporation with \(\mathrm{HNO_3}\) and \(\mathrm{HClO_4}\) in mass analyses led us to continue our search for a more convenient and less time-consuming method of removing interfering metals. The results of these investigations are given below.

The first experiments were carried out by us to test the possibility of removing interfering metals by boiling the solution with sodium acetate before adding β-nitroso-α-naphthol, as was once proposed by Mallory [5] in the visual colorimetric determination of cobalt with nitroso-R salt, and by Shcherbov [6] in its photocolorimetric determination in ores with the same reagent.

As is known, under the action of sodium acetate during boiling, iron precipitates in the form of basic salts, copper in the form of copper oxide, nickel in the form of \(\mathrm{Ni(OH)_2}\), while cobalt remains in solution in the form of the complex \([ \mathrm{Co(CH_3COO)_6} ]'''\) [7].

Ammonia was added to the solution under investigation until turbidity of iron hydroxide hydrate appeared, sulfuric acid (1:1) was then added dropwise until the precipitate dissolved, and then 10 ml of a 50% sodium acetate solution was added. The solution was heated, and boiling was continued until complete coagulation of the precipitate. The subsequent determination of cobalt was continued according to the procedure described by us earlier [1].

Our experiments did not give positive results either when the precipitate was separated by filtration before the addition of β-nitroso-α-naphthol or without its separation, as is usually practiced in the determination of cobalt in ores with nitroso-R-salt [7]. In four parallel determinations of cobalt by this method in 20 ml of an ore solution containing 15.6 γ of cobalt, the optical density in a 50-mm cuvette with a green light filter varied from 0.372 to 0.521 instead of 0.390, which we had obtained repeatedly in constructing the calibration graph. We obtained the same negative result when we tried to replace the process of destroying the excess cupferron by evaporation with HNO₃ and HClO₄ with oxidation by ammonium persulfate, according to the method proposed by Smirnov-Averin, Krotov, and Sokolov [8] for removing ethylenediaminetetraacetic acid from solutions.

We obtained quite satisfactory data only when, instead of cupferron, we used sodium fluoride as the precipitant, which precipitates iron in the form of a white finely crystalline precipitate of composition 5NaF·2FeF₃ [9] and aluminum in the form of the compound 11NaF·4AlF₃ [10]. Such a substitution considerably accelerates, simplifies, and reduces the cost of determining small amounts of cobalt with β-nitroso-α-naphthol. In this case, instead of 25 ml of a 6% cupferron solution, 30 ml of chloroform, 5 ml of concentrated nitric acid, and 3 ml of concentrated hydrochloric acid, only 3–4 g of sodium fluoride are consumed for each cobalt determination. In addition, the need for a lengthy process of destroying cupferron residues is eliminated.

On the basis of our experiments, we recommend the following procedure for determining small amounts of cobalt in nonferrous-metal ores when the cobalt content is in the thousandths of a percent.

A weighed portion of finely ground ore, 1–2 g, is treated with 15–20 ml of concentrated hydrochloric acid in a 200–250 ml beaker and heated on a sand bath for 5–10 minutes. Then 1–2 g of sodium fluoride is added, the mixture is thoroughly stirred, 5 ml of concentrated nitric acid is poured in, and the mixture is evaporated to a thick consistency. After this, 10 ml of sulfuric acid (1:1) is added and evaporation is continued until dense fumes of sulfuric anhydride appear. After the solution has cooled, 10 ml of water is added and evaporation is repeated. Then 50 ml of water is poured in, the mixture is boiled until the sulfates dissolve, cooled, and neutralized with ammonia until a brown turbidity of Fe(OH)₃ appears. Thereafter the turbidity is dissolved with a few drops of hydrochloric acid, the solution is heated, and 70–80 ml of a heated 4% sodium fluoride solution is added with stirring. After 8–10 minutes the white precipitate that has formed is filtered off and washed with water, collecting the filtrate and wash waters in a 200- or 250-ml volumetric flask; the solution is brought to the mark, mixed, and a 25–50 ml portion is taken with a pipette for the determination of cobalt.

Into a 150–200 ml separatory funnel are poured 20 ml of a 2 M sodium acetate solution and 1 ml of concentrated hydrochloric acid; the mixture is stirred, an aliquot portion of the ore solution and 10 ml of a 0.5% solution of β-nitroso-α-naphthol in methyl alcohol are added. The mixture is then stirred thoroughly for 30 sec and, after standing for five minutes, is shaken with 20 ml of benzene for another 30 sec. After separation of the phases, the aqueous solution is drained off and the benzene layer is washed successively with 10 ml

water, 10 ml of hydrochloric acid (1:5), 10 ml of water, 10 ml of normal NaOH, and finally another 10 ml of water. The benzene layer is then filtered through a dry 5 cm ashless white-ribbon filter into a 50 ml volumetric flask; the precipitate on the filter is washed with benzene, the benzene is made up to the mark, mixed, and the optical density of the solution is measured in a 50 or 30 mm cuvette with a green light filter. The comparison solution is a blank solution prepared by adding all the reagents, but without cobalt.

When the cobalt content is in tenths of a percent, the ore sample is increased to 5 g and a larger aliquot portion of the solution is taken for each determination.

Preliminary experiments showed that, for precipitation, crystalline sodium fluoride may also be used instead of a sodium fluoride solution. In this case, with vigorous stirring, 3–4 g of NaF are added to the heated test solution; the solution is filtered through a dense dry filter, and the precipitate is washed, collecting the filtrate and wash waters in a 100 ml volumetric flask. It is then made up to the mark, mixed, and for each determination 30–40 ml of solution are taken when the cobalt content is within the range \(10^{-4}\%\), and 10 ml when the cobalt content is within the range \(10^{-3}\%\). For treating the solution with β-nitroso-α-naphthol, extracting the complex, and preparing it for photocolorimetry, a 100 ml separatory funnel is sufficient.

However, when dry sodium fluoride was used, we often obtained somewhat low results, and filtration of the precipitate proceeded much more slowly. Therefore preference should be given to precipitating iron with a heated 4% NaF solution, although in this case the solution is considerably diluted.

Table 1 gives data obtained by us in the determination of cobalt by the above-described procedure in two ore samples from the Norilsk Combine. Table 2 gives the results of checking the accuracy of the proposed procedure by the method of additions.

Table 1

Results of parallel determinations of cobalt in ores of the Norilsk Combine with β-nitroso-α-naphthol

Table 2

Checking the accuracy of the proposed procedure by the method of additions

As is evident from the data in Tables 1 and 2, the reproducibility of parallel determinations and the agreement of the results with the actual cobalt content are entirely satisfactory.

Conclusions

The previously proposed procedure for determining small amounts of cobalt in ores with β-nitroso-α-naphthol has been improved. In the new procedure, the lengthy and inconvenient process, in routine analyses, of removing interfering metals with cupferron is replaced by precipitation of iron and aluminum with sodium fluoride. As a result, the need for prolonged evaporation of the solution with HNO₃ and HClO₄ to destroy the excess cupferron is eliminated, and instead of 25 ml of a 6% cupferron solution, 30 ml of chloroform, 5 ml of concentrated nitric acid, and 3 ml of concentrated hydrochloric acid, only 3–4 g of sodium fluoride are consumed for each determination.

Literature

1. Lel’chuk Yu. L., Skrinova L. L., Kristalev P. V. Izv. Sib. otdel. AN SSSR, 11, 63, 1960.
2. Bouland E., The Analyst, 71, 230, 1946.
3. Roony R. C. Metallurgia, 348, 205–208, 1958.
4. Peshkova V. M., Bochkova V. M. Trudy komissii po analiticheskoi khimii, 7, 125, 1956.
5. Malyuga D. P. Zhurnal analit. khimii, 1, 176, 1946; 11, 233, 1947.
6. Shcherbov D. P. Zavodskaya laboratoriya, 15, 1399, 1949.
7. Fainberg S. Yu. Analiz rud tsvetnykh metallov, Goskhimizdat, 360, 1953.
8. Smirnov-Averin A. P., Krot N. N., Sokolov A. B. Zhurnal analit. khimii, 13, 280, 1958.
9. Tananaev I. V., Deichman E. N. Zavodskaya laboratoriya, 12, 30, 1946.
10. Tananaev I. V., Lel’chuk Yu. L. Zhurnal analit. khimii, 2, 94, 1947.