ON THE DETERMINATION OF COPPER IN ORES WITH GUARANTEED ACCURACY TO THE SECOND DECIMAL PLACE AT A CONTENT OF ABOUT 0.5%

V. B. Sokolovich, I. P. Onufrienok, B. N. Besprozvannykh

(Presented by Professor A. G. Stromberg, Doctor of Chemical Sciences)

Some industries, for the control of technological processes, require analytical methods of increased accuracy, guaranteeing the content of hundredths and thousandths of a percent of the components being determined. At the request of one of the mining-and-metallurgical combines, we investigated the possibility of determining copper in ores at a content of about 0.5%, with guaranteed accuracy to hundredths of a percent.

At low contents of the components being determined, increased accuracy of determinations is ensured by physicochemical methods [1]. In the present case, however, with a comparatively high copper content in the ore, the required accuracy of determinations could not be guaranteed by them, as was confirmed by verification of the photocolorimetric method used at the combine. Solutions with copper concentrations suitable for colorimetry can be obtained from the indicated ores either by reducing the original weighed portion of ore, which would increase the weighing error, or by using a small aliquot portion, which would give a large dilution error. Measuring thousandths “by eye” on the optical-density scale also reduces the accuracy of determinations.

Among chemical methods, for mass analyses it is most convenient to use volumetric methods, which are comparatively rapid and accurate. For determining copper, in our opinion, iodometric or iodofluoride methods, as well as rhodanide methods, should be used.

The iodofluoride method is known from the literature as the most rapid, accurate, and suitable for semimicrodetermination of copper. In the present case, however, the iodofluoride method did not give the desired results. The main difficulty in the work proved to be the absence of a sharp end point in titration and the rapid restoration of the starch color, which indicated a considerable reversibility of the process. In our opinion, the reverse course of the reaction could have been caused by:

1) a lowered pH of the medium by the time the reaction was completed (there are indications that the best value is pH = 3.5–4 [6]);

2) insufficiently strong binding of Fe³⁺ into a complex ion, which allowed iron, interacting with potassium iodide, to liberate free iodine;

3) an elevated pH of the medium at the beginning of the reaction, which prevented sufficiently vigorous interaction of copper with potassium iodide, and the reac-

tion did not have time to pass during the titration. This could also have been facilitated by the binding of \( \mathrm{Cu}^{2+} \) into a complex, adsorption or occlusion by the precipitate, etc.

To eliminate the indicated difficulties, various buffer mixtures were introduced into the reaction system; ammonium fluoride was partially replaced by phosphoric acid and its salts, tartaric acid and its salts; indicators were selected for controlling pH, etc.

Apparently, the decisive factor affecting the reaction conditions in the present case is the amount of ammonium fluoride introduced, which must be determined by the content in the ore not only of iron, but also of aluminum and silicic acid. An arbitrarily excessive introduction of ammonium fluoride into the reaction system should increase the pH, slow the interaction of copper with potassium iodide, and prolong the end of the reaction. The behavior of ammonium fluoride should be affected not only by the content of silicic acid in the ore, but also by its state in each individual sample, which must be determined by the completeness of decomposition of the insoluble residue, i.e., by the conditions of decomposition of the sample (temperature, time, etc.).

Thus, if for the determination of copper with ordinary accuracy the iodofluoride method is suitable, then for the purposes of an accurate determination of copper in ores—one guaranteeing hundredths of a percent—where its content is about 0.5%, the method cannot be used when the content of iron and silicic acid is high. Fluctuations in the consumption of thiosulfate for titration, amounting to only 0.04 ml (0.15 ml of 0.01 N solution), which are not reflected in results of ordinary accuracy, in the present case correspond to about 0.01% copper, making the results unacceptable.

When working by the rhodanide method, it seemed possible to obtain more accurate results, since this should be promoted by the decrease in the magnitude of the gram-equivalent of copper in the reactions of interaction of CuCNS with oxidizing agents (\(E_{\mathrm{Cu}} = 9.081\)). However, the rhodanide method with reduction of copper by hydroxylamine hydrochloride and a permanganatometric end point likewise did not ensure the required accuracy of the determinations.

The most accurate of the chemical methods—gravimetric analysis—could be used in the present case provided that the final product (the weighing form) was weighed on microchemical balances or that copper was isolated in a weighing form with a conversion factor of about 0.1. After discussing the gravimetric methods known from the literature, we settled on the precipitation of copper with Reinecke’s salt in the form of the complex \( \mathrm{Cu[Cr(NH_3)_2(CNS)_4]} \). Most metals encountered in ores do not interfere with the determination (exceptions are Hg, Ag, Te). The empirical conversion factor to copper is 0.1636 [2].

It turned out that, when copper is extracted from a gram sample of ore containing about 0.5% copper, results satisfactory in accuracy can be obtained. The work, however, requires great care. When weighing the final product, containing about 0.0050 g of copper, discrepancies of 0.0005 g make the results almost unacceptable, since they amount to about 0.01%. The use of a two-gram sample increases the accuracy of the determinations, since 0.0005 g gives a discrepancy much smaller than 0.01%. A substantial drawback of the method is its laboriousness. After decomposition of the sample, 3–4 hours are required for the determination.

Experimental part

The ore was decomposed both according to the procedure adopted at the combine, using \( \mathrm{NH_4F} \), and according to the prescription recommended by Feinberg—a mixture of acids with an oxidizer [2]. In determining copper by the iodofluoride method after decomposition of the ore according to Feinberg, the results were low, whereas after decomposition with \( \mathrm{NH_4F} \) they were high (Table 1). On

results of gravimetric determinations of copper, the method of decomposing the ore had no appreciable effect (Table 2).

In the iodofluorimetric method, a 0.01 N sodium thiosulfate solution was used, its titer being established against electrolytic copper. For a gram sample of decomposed ore, neutralized with 25% ammonia, from 0.3 to 0.6 g of \(NH_4F\) and from 10 to 20 ml of 10% potassium iodide solution were consumed. The titration was completed in the presence of a 1% starch solution [6]. Some results are given in Table 1.

Table 1

Table 2

When working by the rhodanide method, the insoluble residue was separated; for the reduction of copper and iron, hydroxylamine hydrochloride was used in a weakly acidic (approximately 0.2 N) solution in an amount of 2–3 g per gram sample of ore. Cuprous thiocyanate was precipitated from the hot solution with a 10% ammonium thiocyanate solution [4]. As the oxidizing agent, a 0.1 N potassium permanganate solution was used, with which first, in an alkaline medium, \(CNS^-\) was titrated, and then in an acidic medium—\(Cu^+\) [5]. The results of some determinations are given in Table 3.

In the gravimetric determination of copper with its separation using Reinecke’s salt, the procedure was as follows.

Into a conical flask with a stopper, to 2 g of finely ground ore were added 1 g of NH₄F and 20 ml of HCl, sp. gr. 1.19. After boiling for 2–5 minutes, 10 ml of HNO₃, sp. gr. 1.40, was added to the same flask, and boiling was continued for 25–30 minutes. Then the stopper was removed, 20 ml of H₂SO₄ (1 : 1) was introduced into the flask, and everything was evaporated until dense white fumes of SO₃ appeared. After rinsing the walls of the flask, fuming was continued for several more minutes until the nitrogen oxides were completely removed. To the decomposed and cooled sample were added 50 ml of water and 10 ml of HCl (1 : 1), and the whole was heated until the sulfates dissolved. The insoluble residue was filtered off and washed with hot water until the reaction for iron with yellow prussiate of potash ceased. The completely clear solution, in a volume of 120–150 ml, was heated almost to boiling. Copper and iron were reduced with a 25-percent solution of K₂SnCl₄·2H₂O, which was added until the reaction mixture became decolorized, with a subsequent excess of 2 ml [2, 3].

Table 3

To the hot reduced solution, in small portions with continuous shaking, a 1-percent solution of Reinecke salt was added until the solution above the precipitate acquired a reddish color (for 10 mg of copper, about 10 ml of the salt solution is required). The reaction mixture was kept on a water bath for about 15 minutes; then, after cooling, the precipitate was filtered through a previously weighed crucible with a porous bottom No. 4. The precipitate was washed three times with a cold 5-percent HCl solution, five times with warm water, twice each with alcohol and ether, and dried at 110° to constant weight. The results of several determinations are given in Table 4.

Table 4

Conclusions

1. Copper, when its content in ores is about 0.5%, can be determined with guaranteed accuracy to the second digit by a gravimetric method based on the precipitation of copper with Reinecke salt from a two-gram sample.

2. The volumetric iodofluoride and rhodanide methods, as applied to the indicated ores, do not provide the required accuracy of determination.

References

1. Mukhina Z. S., Nikitina E. I., et al. Methods for the Analysis of Metals and Alloys. Oborongiz, 1959.

2. Feinberg S. Yu. Analysis of Nonferrous-Metal Ores. Metallurgizdat, 1953.

3. Analysis of Mineral Raw Materials. Goskhimizdat, 1956.

4. Berl-Lunge. Chemical-Technological Methods of Investigation, vol. II, Moscow—Leningrad, 1938.

5. Gillebrand W. F., Lendell G. E., et al. Practical Manual of Inorganic Analysis. Goskhimizdat, 1957.

6. Podchainova V. N. Methods for the Determination of Copper. Metallurgizdat, 1947.