METHOD FOR DETERMINING GOLD IN ROCKS AND NATURAL WATERS BY STRIPPING POLAROGRAPHY

N. A. KOLPAKOVA, A. G. STROMBERG, N. I. BELOUSOVA

One of the main tasks now facing analytical chemistry is the development of a simple, rapid, and highly sensitive method for determining nanogram quantities of gold in various objects.

The aim of this work was to develop a method for determining nanogram quantities of gold in ores and natural waters by film polarography with accumulation on a graphite electrode (FPE).

Until now, the principal method for determining gold in these objects has remained the spectral method with preliminary sorption of gold on some adsorbent \[1–4\]. The sensitivity of the methods did not exceed \(0.1\ \mu\text{g}/\text{l}\). A number of works \[5–9\] have been devoted to the study of the polarographic behavior of gold on a graphite electrode.

The main content of these works consists in selecting the most convenient supporting electrolyte for determining low concentrations of gold. The authors mainly propose nitric-acid and chloride supporting electrolytes. However, in the polarography of gold at concentrations less than \(10^{-7}\) g-ion/l, in these solutions we were unable to obtain sufficiently reproducible peaks. The gold peaks in this case increased not in proportion to the concentration and disappeared completely with time. We were unable to determine the cause of this behavior of gold in hydrochloric-acid and nitrate solutions.

Well-reproducible anodic peaks of gold were obtained against a potassium bromide supporting electrolyte. In this case the dependence of the gold peak current on the concentration of ions in solution is linear over a wide concentration range (Fig. 1).

Coulometrically it was found that the anodic peak of gold in bromide solutions is due to the ionization of gold to the monovalent state, which at more positive potentials passes into the trivalent state. It was not possible to record the anodic peak of trivalent gold in bromide solutions, since it is evidently masked by the discharge of the ions of the supporting electrolyte.

At low bromide-ion concentrations the diffusion branch of the anodic gold peak is poorly expressed (Fig. 2). This is probably connected with the fact that the transition of monovalent gold to trivalent gold in this case begins at more negative potentials.

It was also observed that, when the supporting electrolyte is acidified with hydrobromic acid, the anodic gold peak increases somewhat. As a result of the studies carried out, the most optimal supporting electrolyte proved to be \(1\ M\ \mathrm{KBr} + 0.03\ N\ \mathrm{HBr}\). Electrodissolution of gold in this supporting electrolyte occurs in the potential interval \(+0.5\)—\(+0.6\) V relative to the SCE. The magnitude

Fig. 1. Dependence of the gold electro-dissolution current on the concentration of gold in solution against the background of 1 M KBr + 0.03 N HBr. Experimental conditions: \(\tau = 10\) min, \(W = 200\) mV/min, \(\varphi_{\mathrm{e}} = -0.2\) V.

Fig. 2. Anodic polarization curves of gold at different concentrations of potassium bromide. Experimental conditions: 1—0.4 M KBr, 2—0.6 M KBr, 3—1 M KBr, \(W = 200\) mV/min, \(\tau = -10\) min, \(C_{\mathrm{Au}} = 5 \cdot 10^{-8}\) g-ion/L.

The maximum anodic current depends on the electrolysis potential. A shift of the potential into the cathodic region to \(-0.2\) V vs. SCE is accompanied by an increase in the electro-dissolution current. A further increase in cathodic polarization does not affect the character of the anodic polarograms. The determination of gold against the background of 1 M KBr + 0.03 N HBr is not interfered with by most of the more electronegative elements: mercury, silver, copper, lead, bismuth, and others.

On the basis of the studies carried out, the following procedure was developed for determining gold in ores, minerals, and natural waters. The apparatus and experimental procedure are analogous to those described previously [10].

Procedure of Analysis

A weighed portion of finely ground ore weighing 1–25 g¹ is roasted in a muffle furnace at 500–600°C. In this process carbon is completely burned off and iron is converted into an insoluble form. The sample is then dissolved in aqua regia and filtered through a filter, washing the precipitate with water. The resulting filtrate is evaporated to moist salts. A 100 ml sample of the natural water under analysis is evaporated to moist salts at 70–80°C. If the residue has a yellow or brown tint due to the presence of organic impurities, it is ignited in a muffle furnace at 500–600°C. The subsequent analysis of the ore and natural water is analogous. To the hot residue, 5–10 ml of supporting electrolyte (1 M KBr + 0.03 N HBr)² is added dropwise. The solution is boiled for better dissolution

¹ If a rock is being analyzed in which the gold is present in a dispersed state, a larger weighed portion should be taken (up to 25 g). In the analysis of minerals, a weighed portion of 1 g is sufficient.

² All reagents used in the work were purified electrochemically.

gold, purge with nitrogen, and polarograph. Electrolysis of gold is carried out at a potential of \(-0.2\) V for 3–10 min.

The method described above for determining gold is very conveniently combined with the extraction-polarographic method for determining silver and mercury that we developed earlier [10].

During extraction of silver and mercury with a dithizone solution, gold is also completely extracted. As was established [10], during re-extraction of mercury and silver, gold is practically not re-extracted and remains in the organic phase. The organic phase is decomposed at a temperature of 200–250°C, 5–10 ml of background electrolyte is added, and polarography is carried out under the conditions described above. Tables 1 and 2 give the results of analysis of some

Table 1

Results of determining gold by the PPN method in some natural waters

Experimental conditions: electrolysis time 1–20 min, rate of potential change 100–500 mV/min. Instrument sensitivity \(1\cdot10^{-10}\)—\(1\cdot10^{-9}\) A/mm

Table 2

Results of determining gold by the PPN method in ores

Experimental conditions: sample weight 1–25 g, electrolysis time 1–10 min, instrument sensitivity \(1\cdot10^{-10}\)—\(10^{-9}\) A/mm, \(W=100\)—500 mV/min.

in ores is \(10^{-7}\)—\(10^{-8}\%\), and in waters 0.01—0.002 μg/l. The relative error of determination does not exceed 20%. The analytical results obtained by us are in satisfactory agreement with the data of the chemical-spectral method according to the authors’ procedure [1].

Conclusions

A procedure has been developed for determining gold in rocks and natural waters by the stripping polarography method. The sensitivity of the procedure is 0.01—0.002 μg/l (\(10^{-7}\)—\(10^{-8}\%\)). With the developed procedure, mercury, silver, and gold can be analyzed simultaneously.

References

1. N. I. Brodskaya, I. P. Vychuzhanina, A. D. Miller. Zh. prikl. khim., 40, 4, 1967.

2. A. D. Miller, A. P. Karavaikova. VITR, series “Exchange of Experience,” 41, L., 1961.

3. I. Yu. Sokolov, M. M. Kopylova, E. I. Galitsina, T. S. Korchagina. In: “New Methods for Analyzing the Chemical Composition of Groundwaters,” Moscow, 1967.

4. Kh. Arkhipova, P. A. Prikhodko. Uzbek Geological Journal, Academy of Sciences of the USSR, 4, 1965.

5. E. Jacobs. Analyt. Chem., 35, 13, 1963.

6. Takayashi Vashimoti, Masao Arakawa, Tsudio Takeuchi. Talanta, 12, 2, 1965.

7. M. Tanaka. Japan. Analyst, 12, 897, 1963.

8. I. E. Harrar, F. B. Stephens. J. Elect. Chem., 3, 112, 1962.

9. V. B. Belyavskaya. Film Voltammetry of Metals. Dissertation, Donetsk, 1967.

10. M. A. Kolpakova, A. A. Kaplin, A. G. Stromberg, ZhAKh, 23, No. 5, p. 665, 1968.