DETERMINATION OF TELLURIUM IN INDIUM ANTIMONIDE CRYSTALS BY FILM POLAROGRAPHY WITH ACCUMULATION

V. F. Gridaev, A. A. Kalinin, A. G. Stromberg

Selenium and tellurium are important donor impurities in semiconductor compounds. The properties of the semiconductor vary with the content and distribution of the impurity. For the purpose of monitoring semiconductor compounds for selenium and tellurium content, various polarographic methods of analysis using a mercury dropping electrode and a stationary electrode exist [1–2].

We are not aware of literature data on the polarographic reduction of selenium (IV) and tellurium (IV) at a solid electrode to zero valence. We have established that at graphite and platinum electrodes selenium (IV) and tellurium (IV) are reduced to zero valence. We propose a method for determining microquantities of selenium and tellurium based on preliminary concentration of these elements on a solid electrode in the form of atoms of zero valence at the potential of the limiting current, followed by recording of the current of electro-dissolution of selenium and tellurium atoms under cathodic or anodic polarization. The process of concentrating and electro-dissolving atoms of selenium and tellurium of zero valence is carried out in electrolytes of different composition (the electrolyte-exchange technique is used [3]). Concentration of selenium (IV) and tellurium (IV) on a solid electrode in the elemental state has advantages over concentration of selenium (IV) and tellurium (IV) in the form of a sparingly soluble compound, since in this case the limiting sensitivity of the method is not restricted by the solubility product of the sparingly soluble compound.

Apparatus and reagents

The polarographic investigations were carried out in a conventional electrolytic cell with inserted cups. The cathode was a platinum wire (sealed into glass), \(d = 0.5\) mm; \(\ell = 5\) mm; or gra-

graphite electrode \(d = 2\) mm. The anode was a graphite rod, \(d = 5\) mm; \(l = 50\) mm. Oxygen was removed from the solution with argon or nitrogen. The polarograph was of the OH-101 type. Standard solutions of selenium and tellurium were prepared by dissolving elemental \(Se(Te)\) in a mixture of HCl and \(HNO_3\) (3:1) at \(t = 50^\circ C\). Solutions of the indifferent electrolyte were prepared with tridistilled water. All numerical values of potentials in the article, except in specially stipulated cases, are given relative to an auxiliary graphite electrode, whose potential in \(2N\) HCl relative to the saturated calomel electrode is \(-0.6\) V.

Experimental Part

Preliminary concentration of \(Se(Te)\) was carried out in \(2N\) HCl. The electrolysis potential for selenium was \(-1\) V. For cathodic electro-dissolution of elemental selenium and tellurium after preliminary concentration, the indifferent electrolyte selected was \(0.05N\) KOH \((NaOH)\). The potentials of the cathodic peaks of selenium and tellurium were \(-0.9\) V and \(-1.1\) V, respectively. The mechanism of dissolution of selenium and tellurium in alkaline solutions under cathodic polarization of the electrode may be represented by the following equations:

\[ Se^0(Te) + 2e \longrightarrow Se^{-2}(Te^{-2}), \]

\[ Se^{-2}(Te^{-2}) + 2K^+ \rightleftarrows K_2Se(K_2Te). \]

In the indifferent electrolyte of \(2N\) hydrochloric acid, after preliminary accumulation at a potential of \(-1.2\) V, anodic peaks of tellurium were obtained with the aid of anodic polarization of the electrode. The potential of the anodic peak of tellurium was \(-0.25\) V. The dependence of the height of the anodic and cathodic peaks of selenium and tellurium on the concentration of their ions in a \(2N\) hydrochloric acid solution is linear (Fig. 1). A linear dependence of the height of the cathodic peak of \(Se(Te)\) and of the anodic peak of tellurium is observed in the accumulation-time interval of 1–20 min. The dependence of the height of the cathodic and anodic peaks of selenium and tellurium on the concentration potential passes through maxima; the decrease in peak height with increasing electrolysis potential is due to the evolution on the electrode of \(H_2Te\), \(H_2Se\) (Fig. 2). The detection limit of the cathodic peaks of selenium is \(n \cdot 10^{-8}\) mol/L, and for tellurium it is \(n \cdot 10^{-6}\) mol/L. The detection limit of the anodic peaks of tellurium at an instrument sensitivity of \(2 \cdot 10^{-9}\) A/mm is \(1 \cdot 10^{-8}\) mol/L. The possibility of cathodic and anodic electro-dissolution of selenium and tellurium from the surface is of interest both for physicochemical studies and for selecting optimal conditions for eliminating the interfering influence of other elements.

Fig. 1. Dependence of the peak height of \(Se\) and \(Te\) on concentration.

Fig. 2. Effect of the electrolysis potential on the height of the selenium and tellurium peaks.

To study the layer-by-layer distribution of tellurium in \(InSb\) compounds doped with tellurium, a procedure was developed for determining tellurium in the presence of indium and antimony. When choosing the concentration potential for tellurium, equal to \(-1.1\) V, indium is not reduced at the electrode. For the purpose of anodic dissolution of the antimony formed with

tellurium on the surface of the electrode, a precipitate of an intermetallic compound is formed; this takes 2–5 min, with holding at \((\varphi_{в} = -0.5 \text{ to } -0.6\ \mathrm{V})\). The holding time is determined by the amount of mercury reduced on the electrode.

Procedure for analyzing \(InSb\) powder doped with tellurium. A 0.002 g sample is dissolved in 1 ml of \(HNO_3\) in a quartz beaker at \(t = 90^\circ\mathrm{C}\). Evaporation is carried out to moist salts. To the moist residue, 5 ml of \(5N\) HCl is added, and the residue is dissolved by heating. After dissolution of the residue, the beaker is cooled at room temperature and polarographed. The electrolysis potential is \(-1.1\ \mathrm{V}\), \(\varphi_{п} = -0.3\ \mathrm{V}\). The average tellurium content in the analyzed \(InSb\) powders was \((0.15 \pm 0.01)\%\) and \((0.12 \pm 0.001)\%\). The initial data on the doping of the powders in wt. % Te were respectively 0.29% and 0.018%. Thus, the final degree of doping does not correspond to the specified value.

Table I

Results of layer-by-layer analysis of a crystal (634)

The procedure was used in the analysis of the layer-by-layer distribution of tellurium in \(InSb\) crystals doped with tellurium. The layers were removed mechanically. The layer was removed with a cutter; the layer thickness was checked with a dial micrometer. The error in measuring the layer thickness was \(\pm 15\%\) for a layer of \(10\ \mu\) and \(\pm 30\%\) for a thickness of \(100\ \mu\). Table I gives data on the results of layer-by-layer analysis of an \(InSb\) crystal for Te content; layer numbers 3, 5, 7, 9 correspond to a thickness of \(100\ \mu\); 1, 2, 4, 6, 8—to a thickness of \(10\ \mu\). For analysis of a \(100\ \mu\) layer, several samples weighing 0.002–0.003 g are taken and the average tellurium content in the layer is determined. It follows from Table I that the distribution of tellurium along the length of the crystal is uniform.

Conclusions

A method is proposed for determining microquantities of selenium and tellurium by the FPA method. A procedure has been developed for layer-by-layer analysis of a crystal

$InSb$, doped with tellurium.

References

1. B. Ya. Kaplan, A. S. Rezakova. ZhAKh, XXI, No. 10, 1268, 1966.
2. E. G. Chikrizova and L. S. Kopanskaya. ZhAKh, XXIII, No. 3, 394, 1968.
3. A. A. Kaplan, M. S. Zakharov, A. G. Stromberg. Methods for the Analysis of Especially Pure Reagents and Preparations. IREA, Moscow, 1963.