WAYS OF INCREASING THE SENSITIVITY OF THE METHOD OF AMALGAM POLAROGRAPHY WITH ACCUMULATION

A. A. Kaplin, L. F. Trushina

(Presented by the scientific-methodological seminar of the KhTF)

The rapid growth of a number of new branches of modern technology that use materials with trace contents of impurities leads to the fact that, with each year, the requirements increase both for methods of purifying such materials and for methods of analyzing them.

Despite the fact that control methods, as a rule, lag in their development behind the achievements of science and technology in the field of deep purification of substances, at the present time the problem of determining \(10^{-5}\)—\(10^{-6}\%\) impurities may be considered practically solved thanks to successes in the development of a number of methods: radioactivation, mass- and chemical-spectral, amalgam-polarographic, luminescent, and others.

Such an abundance of methods for trace analysis indicates that there is no universal method which, with the necessary sensitivity and reliability, using comparatively small sample weights and simple and reliable apparatus, would make it possible to determine a considerable number of impurities simultaneously and rapidly.

With respect to further increasing the sensitivity of trace determination, the various methods of analysis are in an unequal position. One of the most promising methods for solving the principal problem of analytical chemistry—the determination of \(10^{-7}\)—\(10^{-9}\%\) and smaller amounts of impurities in high-purity materials—is the method of amalgam polarography with accumulation.

In the first publications on the APN method (1957—1960), the obtaining of anodic waves of certain elements at concentrations of \(10^{-5}\)—\(10^{-7}\) mole/l was reported. Studies aimed at increasing the sensitivity of the method of amalgam polarography with accumulation developed for some time in two directions. One of them was an attempt to combine certain promising variants of the polarographic method, in particular oscillographic and square-wave methods (the latter, according to Barker [1], without accumulation permits determinations down to \(10^{-7}\) mole/l), with the principle of accumulation. However, as carefully and comprehensively designed experiments by Kaplan [2] showed, a significant increase in the sensitivity of determination when using square-wave polarography with accumulation cannot be achieved, owing to the nonlinear dependence of the depth of the anodic wave on the accumulation time (in the absence of noticeable depletion of the solution) and to considerable interference in recording anodic waves at the maximum sensitivities of the polarograph recording device. The sensitivity of determination

$10^{-8}$—$10^{-9}$ mol/liter was achieved when oscillopolarography was used both in recording current–potential curves [3] and curves

\[ \left(\frac{dE}{dt}-E\right) \]

[4].

A more substantial increase in sensitivity has not yet been achieved either with the use of oscillographic, radio-frequency, or other variants of polarography, mainly because of the increase in capacitive current at high rates of change of the voltage applied to the cell. Nevertheless, further investigations in this direction (for example, using a TsLA vector polarograph) are of considerable interest.

Another line of research aimed at increasing the sensitivity of the method of amalgam polarography with accumulation is associated with the development of a variant of the method in which a linearly varying voltage (usually at a rate of $10^{-3}$—$10^{-2}$ V/sec) is used for recording the anodic peaks.

The increase in sensitivity of the APN method obtained in the first studies, compared with classical polarography with a dropping mercury electrode, is due to concentration of the metal in the mercury drop and, consequently, to an increase in the signal-to-noise ratio. A further increase in the sensitivity of the method was associated with the study of factors affecting the height of the anodic peak at a constant concentration of metal ions in solution. Most often, the dependence of the height of the anodic peak on the electrolysis time was studied.

The linear character of this dependence (in the absence of noticeable depletion of the solution), as was shown later [7], allowed a number of authors to increase the sensitivity of determination somewhat by means of 2–3 hours of preliminary electrolysis, which is hardly convenient from a practical point of view. Some increase in sensitivity was achieved by carrying out the analysis at higher accumulation potentials, by reducing the volume of the working solution from 50–20 ml to 2–5 ml, and by increasing the number of revolutions of the magnetic stirrer (to 600–800 rpm) in combination with intensive stirring of the solution by a gas jet.

The basis for the intensification of work on increasing the sensitivity of the method was certain successes achieved in recent years in the development of the theory of the method. The conclusion of Vinogradova and Vasil’eva [5], as well as of Shain and Lewinson [6], concerning the uniform distribution of the metal in the mercury drop before recording the anodic polarogram made it possible to calculate the mean concentration of the metal in the amalgam and the height of the anodic peak ($I$); it is assumed that the latter is proportional to the concentration of the metal in the amalgam. On the basis of works [5, 6], Stromberg derived expressions for the dependence of the height of the anodic peak on a number of factors [7]^1):

\[ I_a = a \cdot \gamma, \tag{1} \]

where

\[ a = 3K_2 \cdot C_i^0 \cdot V \cdot \frac{1}{r}, \tag{2} \]

\[ \gamma = 1 - e^{-b}, \tag{3} \]

\[ b = \frac{4\pi}{zF} \cdot K_1 \cdot r^2 \cdot \frac{\tau}{V}. \tag{4} \]

^1) Later, an analogous expression was derived by another route in the work of E. N. Vinogradova and L. N. Vasil’eva [8].

\(C_1^0\) — concentration of ions in the solution, g-ion/l;
\(V\) — volume of the solution, ml;
\(r\) — radius of the drop, cm;
\(\tau\) — electrolysis time, sec;
\(K_2\) — anodic-peak constant, a·cm\(^2\)·g-atom\(^{-1}\);
\(K_1\) — electrolysis constant,\(^1\) a·cm·g-ion\(^{-1}\);
\(\gamma\) — degree of depletion of the solution.

It follows from expression (1) that the maximum peak depth can be attained only at definite ratios of \(V\), \(\tau\), and \(S\), determined by the dimensionless parameter \(b\). Let us consider two limiting cases. For values of the parameter \(b < 0.1\) (depletion of the solution is practically absent; less than 5% of the ions present in the solution are deposited on the drop),

\[ I_a \cong a \cdot b = \frac{12\pi}{zF}\cdot K_1 \cdot K_2 \cdot C_1^0 \cdot \tau \cdot r . \tag{6} \]

The anodic peak current is proportional to the constants \(K_1\), \(K_2\), to the electrolysis time and to the radius of the mercury drop (for small \(\tau\) and \(r\)), and does not depend on the volume of the solution. If, however, \(b > 3\) (practically complete, \(>95\%\), depletion of the solution), then

\[ I_a = a = 3K_2 \cdot C_1^0 \cdot V \cdot \frac{1}{2}. \tag{7} \]

The anodic peak current is directly proportional to the volume of the solution (for small \(V\)), to the anodic-current constant, inversely proportional to the radius of the mercury drop (for large \(r\)), and does not depend on the electrolysis time (for large \(\tau\)).

For intermediate values of the parameter \(b\) \((0.1 < b < 3)\), the dependence of the anodic-peak depth on the indicated factors is described by expression (1). In this case the value \(I_a\) (and consequently also the sensitivity of the determination) is governed not by the absolute magnitude of each of the factors, but by the value of \(b\) (see formula (4)), which depends on the numerical value of \(V\), \(\tau\), \(r\), and \(K_1\), characterizing the intensity of stirring of the solution and, in the general case, depending on the electrolysis potential [9]. Thus (from expression (3)), to each value of the electrode radius (at \(K_1 = \mathrm{const}\)) there corresponds a definite ratio \(\tau/V\), at which the required degree of depletion of the solution is achieved.

At constant \(r\) and \(K_1\), it is necessary to increase the electrolysis time and decrease the volume of the solution until the value of the parameter \(b\) is not greater than three (or \(b = 1.26\)), which corresponds to a degree of solution depletion \(\gamma = 0.71\).

Experimental verification of the relations derived, in particular the study of the dependence of \(I_a\) on \(V\), \(r\), \(\tau\), \(\varphi_n\), showed satisfactory agreement between theory and experiment. The relations given above served as the basis for further theoretical studies on increasing the sensitivity of the APA method.

\(^1\) \(K_1\) is determined from the expression

\[ K_1 = \frac{g}{\tau C_1^0 S}, \tag{5} \]

where \(g\) is the area under the anodic peak, coulomb; \(S\) is the electrode surface;

\[ g = q \cdot i \cdot t; \tag{5a} \]

\(q\) — area under the peak, cm\(^2\); \(i\) — sensitivity of the polarograph, a/cm; \(t\) — time for 1 cm of the scale to pass through the recording device, sec/cm.

From expression (1), in a somewhat transformed form,

\[ I_a = K_2 \cdot Q^{1)} \cdot \frac{S}{V} \cdot \gamma \tag{8} \]

it follows that the depth of the anodic peak is proportional to the ratio of the electrode surface \((S)\) to its volume \((V)\), and, consequently, by increasing the ratio \(\frac{S}{V}\), one can increase the sensitivity of the determination [28]. The most suitable electrode for this purpose is one in the form of a thin (2–10 micron) mercury film deposited on a silver substrate, usually a wire 0.8–1 mm thick and 10–15 mm long [28, 29]. An essential advantage of the film amalgamated electrode is also the possibility of obtaining narrower peaks on it, owing to the decrease in the time required for the metal to leave the electrode. The possibility of determining impurities at concentrations of \(10^{-10} - 10^{-11}\) mole/liter on a film electrode has been shown theoretically. The possibility of increasing sensitivity by using a film electrode, however, is somewhat reduced because of the decrease in the value of the constant \(K_2\) (in comparison with a mercury drop) and the increase in the residual current on an electrode with a large surface when peaks are recorded at the maximum sensitivities of the polarograph [12].

A. G. Stromberg derived a theoretical expression for the minimally determinable \((a_*, \%)\), by the anodic stripping polarography method, content of impurities in high-purity materials [13]

\[ a_* = \frac{100 A}{zF} \cdot \frac{\nu \cdot \delta}{\omega} \cdot \frac{hm}{g} \cdot \frac{i_*}{\gamma_*}, \tag{9} \]

\[ C^0_{1*} = \frac{10g}{AV} \cdot a_*, \tag{10} \]

where \(A\) is the atomic weight; \(\delta\) is the width of the anodic half-wave, volt; \(\nu\) is a factor depending on the form of the mercury electrode \((\nu = 1.2 \div 1.4)\); \(\omega\) is the rate of change of potential, volt per sec; \(h_m\) is the minimum depth of the anodic peak that can be measured with the specified accuracy, mm; \(g\) is the weighed amount of material, g.

Thus, the sensitivity of the determination increases with an increase in the degree of exhaustion \((\gamma)\), with narrowing of the anodic-peak width \((\delta)\), with an increase in the rate of change of potential \((\omega)^2\), and with an increase in the weighed amount of material \((g)\). At the same time, the quantity \(a_*\) is explicitly independent of \(V\), \(S\), \(K_1\), and the residual-current density \((j, \text{A}/\text{cm}^2)\). But the optimum value of the degree of exhaustion \(\gamma\) is related to the indicated parameters by expressions (3, 4). The optimum value of the electrode surface is determined from the condition that the minimum depth of the anodic peak \((I_m, \text{A})\) must be related to the residual-current density \(j = p \cdot C_y \cdot \omega \, (\text{A}/\text{cm}^2)\) by the relations

\[ I_m = h_m \cdot i_* = m \cdot S_* \cdot p \cdot C_y \cdot \omega = m \cdot S \cdot j_{\text{res}}, \tag{11} \]

where \(m\) is a factor whose value is determined by the requirement of measuring the depth of the anodic tooth with a specified accuracy (for example, \(\pm 10\%\) relative); the value of \(m\) can be calculated from error theory or determined experimentally; \(p\) is a factor showing how many times the residual-current density is greater than the capacitive-current density—

\(^{1)}\) \(Q\) is the amount of metal deposited on the film, g-atom. The sign (*) denotes the optimum value of the quantity.

\(^{2)}\) At the same time the interference—the capacitive current—increases.

... (\(C_y \cdot m\)); \(C_y = 20\ \mu\mathrm{F}/\mathrm{cm}^2\) is the average capacitance of the double electric layer. Then

\[ S_*=\frac{h_m}{m\cdot p\cdot C_y\cdot \omega}\cdot i_* . \tag{12} \]

The optimum volume of solution from which, on the optimum surface of the electrode during the optimum electrolysis time, the optimum degree of depletion is reached (for example, \(b_* = 1\); \(\gamma_* = 0.63\)) is determined from expression (4):

\[ V_*=\frac{K_1^*\cdot S_*\cdot \tau_*}{b_*\cdot z\cdot F}. \tag{12} \]

If \(V_*=\mathrm{const}\), then

\[ K_1^*=\frac{b_*\cdot z\cdot F}{S_*\cdot \tau_*}. \tag{14} \]

The optimum minimally determined concentration of the metal in the amalgam after electrolysis, to which there corresponds the minimum anodic peak for the given sensitivity of the polarograph (height \(h\)), is determined by the relation (for a mercury drop)

\[ C_2^*=\frac{10^3}{K_2}\cdot m\cdot j . \tag{15} \]

Thus, for a hanging mercury drop electrode characterized by the constant \(K_2\), and for a given background characterized by the coefficient \(m\) and the residual current \((j)\), the condition for determining the minimum concentration in the solution is the attainment of a definite minimum concentration of metal atoms in the amalgam by selecting optimum conditions [14]. The selection of optimum conditions using a film electrode reduces to attaining in the film the minimum amount of metal necessary for recording the anodic peak at the chosen sensitivity of the polarograph; in this case (according to expression 8)

\[ I_{\min}=K_2\cdot Q_{\min}\cdot \gamma^* \tag{16} \]

and at \(\gamma^* = 1\)

\[ I_{\min}=K_2\cdot Q_{\min}^* . \]

The optimum \(Q_{\min}\) is determined by the expression

\[ Q_{\min}^*=\frac{1}{K_2}\cdot m\cdot j . \tag{17} \]

The optimum surface of the film electrode, the volume of solution, and the degree of depletion are also determined from expressions (12, 13). Let us make a numerical estimate of \(a_*\), \(C_{1*}^0\), \(S_*\), \(K_1^*\) for a hanging mercury drop electrode at different polarograph sensitivities and for the following numerical values of quantities, suitable for practice: \(g=1\) 2; \(\delta=0.04\ \mathrm{V}\); \(\omega=400\ \mathrm{mV}/\mathrm{min}=7\cdot 10^{-3}\ \mathrm{V}/\mathrm{sec}\); \(v=1.2\); \(A=100\); \(z=2\); \(h_m=10\ \mathrm{mm}\); \(m=1\); \(\tau=30\ \mathrm{min}=1800\ \mathrm{sec}\). We shall assume that the residual-current density is equal to the capacitive-current density. Then \(j=C_y\cdot \omega=1.4\cdot 10^{-7}\ \mathrm{A}/\mathrm{cm}^2\). Let us consider two variants. In the first, we shall take the volume of solution as constant (\(V=1\ \mathrm{ml}\)) and, in order to obtain the necessary degree of depletion for achieving the sensitivity \(a_*\) on the surface \(S_*\), we shall vary the value of the constant \(K_1^*\) (see expressions 9, 10, 12, 14). In the other variant we shall take as constant the value of the constant \(K_1^*\) (for example, \(K_1^*=10^4\ \mathrm{A}\cdot\mathrm{cm}\cdot\mathrm{g\text{-}ion}^{-1}\)) and shall achieve the necessary degree of depletion with the chosen electrode by decreas-

by increasing the volume of the working solution into which the sample being analyzed is transferred (see expressions 9, 10, 12, 13). The results are given in Table 1.

It follows from the table and formula (15) that if \(K_1\) is increased by an order of magnitude \((K_1 = 1.6 \cdot 10^{+4})\) and the influence of the residual current is reduced by a factor of 10, then the minimum determinable impurity content can be reduced by two orders of magnitude, i.e., about \(10^{-10}\%\) impurity can be determined at \(S = 7 \cdot 10^{-3}\ \text{cm}^2\), \(i = 10^{-11}\ \text{a}/\text{mm}\), with the other conditions specified in the numerical example unchanged.

Table 1

\(V = 1\ \text{ml}\)

\(K_1^* - 1.6 \cdot 10^{+3}\ \text{a}\cdot\text{cm}/\text{g-ion}^{-1}\)

\(K_1^* = 1.6 \cdot 10^{5}\ \text{a}\cdot\text{cm}/\text{g-ion}^{-1}\)

When the conditions are changed in comparison with those specified in the numerical example, the sensitivity of the method may deteriorate appreciably. Thus, for example, under ordinary analytical conditions: \(K_1 = 0.4 \cdot 10^{-4}\); \(S = 0.01\ \text{cm}^2\); \(g = 0.2\ \text{g}\); \(\delta = 0.12\ \text{g}\); \(V = 2\ \text{ml}\); \(\tau = 30\ \text{min}\); \(j = 10 C_y\); \(\omega = 1.4 \cdot 10^{-6}\ \text{a}/\text{cm}^2\). We shall take the other conditions \((A, z, \omega, \nu, h_m, m)\) to be the same as in the numerical example. Calculation shows that the minimum determinable impurity content under these conditions is \(4 \cdot 10^{-6}\%\), i.e., almost three orders of magnitude below the optimal values \((0.6 \cdot 10^{-8}\%)\).

The calculations presented show that, in principle, by the APM method one can determine \(10^{-13}\%\) by weight and less of impurities, and the sensitivity limit of the method is limited by the attainable values of instrument sensitivity, the electrolysis constant \(K_1\), difficulties in working with microvolumes, and interference when using high sensitivities of the polarograph. An instrument sensitivity of \(10^{-12}\)–\(10^{-14*}\ \text{a}/\text{mm}\) is apparently attainable with the present development of radio engineering. It should also be noted that the use of higher polarograph sensitivities leads to an increase in various kinds of interference (instrumental, capacitive current, dissolved oxygen, impurities reduced at the electrode but not forming an amalgam, etc.) and therefore is associated with considerable practical difficulties. Of great importance for increasing the sensitivity of the APM method, as follows from the calculations and relations presented, is an increase in the constant \(K_1\). Let us consider some variants of increasing \(K_1\).

* Of the polarographs currently in use, the greatest sensitivity is possessed by the LP-60 \((i_M = 3.6 \cdot 10^{-11}\ \text{a}/\text{mm})\).

With maximum stirring of the solution (0.5–6 ml) by a magnetic stirrer (600–2000 rpm) or by means of a nitrogen jet (2 l/min), the numerical value of the constant \(K_1\) \((a\cdot cm\cdot g\text{-ion}^{-1})\) is respectively equal to \((0.2-1.8)\cdot 10^3\,g\) [15]. A somewhat higher value of \(K_1\) is obtained with simultaneous stirring of the solution by an electrode vibrating at sonic frequency (50 Hz) and by gaseous nitrogen: \((1 \div 1.5)\cdot 10^3\,g\) [16]. With the combined application of all three effects, it is possible to attain an electrolysis constant of \((60-80)\cdot 10^3\,g\).

A considerable increase in the electrolysis constant \(K_1\) \([(15 \div 20)\cdot 10^3 g]\) was obtained when using a circulation cell, owing to the high (up to 1 m/sec) linear velocity of motion of the solution being analyzed in the narrow gap between the electrode and the capillary walls [17]. The circulation cell, being simple in operation, makes it possible in 10–15 minutes to achieve practically complete depletion in 2–3 ml of the solution being analyzed.

A certain increase in the constant \(K_1\) \([(4 \div 5)\cdot 10^3 g]\) can also be achieved when using electrodes rotating at high linear velocity, especially an electrode with an end face directed perpendicular to the stream of solution flowing past [23]. In addition to increasing the constant \(K_1\) by increasing the intensity of delivery of the substance to the electrode, procedures have been proposed that make it possible to increase the degree of depletion of the solution during the accumulation process, and to record the anodic peaks on the initial surface of the electrode in order to reduce the magnitude of the residual current. The effect of increasing the delivery of substance to the electrode is characterized by the effective electrolysis constant.

\[ K_{\mathrm{eff}} = K_1 \cdot \frac{S_1}{S}, \tag{15} \]

where \(K_1\) is the actual electrolysis constant.

Promising in this direction is the effect of ammonium amalgam (AAA) proposed by Nazarov [20], which consists in carrying out accumulation at considerable negative potentials in ammoniacal solutions; as a result of the formation of ammonium amalgam, the size of the mercury drop increases, and consequently the effective electrolysis constant increases. When the potential is lowered, intense evolution of gaseous products from the amalgam occurs, and before the registration of the anodic wave the drop assumes its initial size. On the basis of theoretical calculations [19–21], confirmed by experimental data [19–22], it has been shown that it is possible to increase \(K_{\mathrm{eff}}\) by approximately an order of magnitude.

A significantly greater increase in \(K_{\mathrm{eff}}\) is given by combining the ammonium-amalgam effect with carrying out electrolysis at the boiling temperature of the solution [22]. With an increase in temperature to the boiling point, the diffusion coefficients of ions in the solution and of atoms in the amalgam increase. In addition, in the boiling solution the near-electrode layer is vigorously stirred. According to Karbainov, the value of the constant \(K_1\) is \((25 \div 30)\times 10^3\,g\).

Carrying out electrolysis at the boiling temperature of the solution (with a reflux condenser) makes it possible to obtain a value of the electrolysis constant up to \((6 \div 8)\cdot 10^3\,g\). With the combined use of the ammonium-amalgam effect and boiling of the solution, an effective electrolysis constant up to \((25 \div 30)\cdot 10^3\,g\) was achieved [22].

An effective electrolysis constant of the order of \((20 \div 30)\cdot 10^3\,g\) is obtained when using the method of stepwise accumulation [23]. Electrolysis is carried out in a comparatively large volume of solution on an elect—

kind with a comparatively large surface. Then the electrode is transferred to a much smaller volume of solution, anodic dissolution of the metal from the electrode is carried out; then repeated accumulation and recording of the peak are performed on an electrode with a small surface, optimal for the given conditions.

Thus, in a relatively short time, by using a number of techniques it has been possible to increase the electrolysis constant by several orders of magnitude.

As noted above, an increase in the sensitivity of the ASV method can be achieved by decreasing the volume of the solution being analyzed (at a constant sample weight) in combination with the selection of optimal polarographic conditions. Studies in this direction involve the use of ultramicroanalysis techniques and the development of the necessary apparatus. For volumes of \(0.03\)–\(0.3\) ml, a design has already been developed for a microcell with comparatively intensive delivery of substance to the electrode, \(K_1 = (3—4)\cdot 10^3\) s \([27]\).

The sensitivity achieved in the microcell is \(5\cdot 10^{-10}\) mole/l for bismuth \([23]\).

It should be noted that increasing the constant \(K_1\) does not completely eliminate the need to work with microvolumes in the ASV method. Thus, if the constant \(K_1\) is increased by a factor of 100 (see Table 1) and is, for example, \(1.6\cdot 10^5\), then the required optimal volume of solution at \(i_* = 10^{-11}\) A/mm and \(S_* = 7\cdot 10^{-4}\) cm\(^2\) with \(1\cdot 10^{-2}\) ml will increase to 1 ml. However, to achieve a sensitivity of \(10^{-14}\%\), it is necessary to transfer the analyzed sample into a volume of \(1\cdot 10^{-3}\) ml instead of \(1\cdot 10^{-5}\) ml at \(K_1 = 1.6\cdot 10^3\) A·cm·g-ion\(^{-1}\). But in the ASV method, unlike other methods, there is no need to reduce the solution volume below the optimum determined by expression (13). This is due to the fact that there is the possibility of internal accumulation of the substance from the analyzed solution on the electrode over a certain period of time.

Of great importance for increasing the sensitivity of determination is reducing the magnitude of the residual current \(i\). The investigations carried out \([25]\) showed that, through more thorough purification of the nitrogen passed through the solution, the residual current can be reduced by a factor of 5–10 and, consequently, the sensitivity of determination can be increased. Satisfactory purification of nitrogen was obtained by passing nitrogen through copper shavings heated to \(450\)—\(500^\circ\) and a solution of vanadium(II) sulfate with zinc amalgam at the bottom of the vessel. A significant decrease in the residual current is achieved with the aid of certain polarograph designs (square-wave, vector polarograph).

A decrease in the magnitude of the residual current may also be associated with more careful preparation of the electrode with the metallic contact (for example, preliminary amalgamation of the platinum contact \([26]\)) and with the choice of the supporting electrolyte.

Thus, further investigations aimed at increasing sensitivity are connected above all with the theoretical and experimental study of the factors affecting the depth of the anodic peak, with the selection of optimal polarographic conditions, the mastering of considerably higher sensitivities of the polarograph, further increase of the constant \(K_1\), the application of ultramicroanalysis techniques in the ASV method, and the development of the corresponding apparatus and working procedures.

The practical attainment of the theoretical sensitivity of determination is limited by a number of factors. Thus, the transition to ever higher sensitivities of polarographs is associated with an increase in the negative influence of the residual current, which causes an increase in the derivative of the residual current with respect to potential. This leads to an increase in the theoretically calcu-

calculated \(h_{\min}\), i.e., toward a worsening of the sensitivity of determination [24]. The absolute value of the residual current depends on the magnitude of the capacitive current, the oxygen reduction current, and organic impurities that do not form an amalgam, on the nature of the background, adsorption and desorption of ions on the electrode surface, the rate of change of the potential and, probably, on other factors whose nature has not yet been established. An increase in \(h_{\min\,\mathrm{theor}}\) is caused by background fluctuations associated both with oscillations of the residual current during recording of the peak and with instrumental interference. Thus, attainment of the theoretical sensitivities of determination is connected with purification of the solution from oxygen and impurities, cutoff of the capacitive current, etc.

Another factor that acquires great importance in connection with increasing the sensitivity of the APN method (as well as other methods of trace analysis) is the requirement that there be no contamination in the reagents and glassware, and the ever stricter requirements, due to this, on the cleanliness of the premises in which the experiment is conducted. Of special value is the improvement of old and the development of new techniques for purifying reagents, glassware, and apparatus for carrying out the experiment under sterile conditions.

Thus, the method of amalgam polarography with accumulation, by means of which, as stated above, it is possible to determine \(10^{-10}\)—\(10^{-12}\) wt.% and less of impurities in high-purity materials, is at present becoming one of the most sensitive methods, alongside mass-spectral and radioactivation methods.

REFERENCES

1. A. Barker. Analyst. IV, 30, 1952.
2. B. Ya. Kaplan. Zavod. lab., I, 29, 4, 1963.
3. S. B. Tsfasman. Zavod. lab., No. 1, 1964; Zavod. lab., No. 3, 1964.
4. R. Kalvoda. Zav. labor., 3, 1959.
5. L. V. Vasil’eva, E. N. Vinogradova. Zavod. labor. XXVII, 1079, 1961.
6. S. Shain, V. Levinson. Anal. Chem., 33, 187 (1961).
7. A. G. Stromberg. Izv. SO AN SSSR, 5, 76, 1962.
8. E. N. Vinogradova, L. V. Vasil’eva. Zavod. lab., 28.12.1962.
9. A. G. Stromberg, M. S. Zakharov. Coll. Methods for the Analysis of Chemical Reagents and High-Purity Preparations, issue 5–6, IREA, Moscow, 1962.
10. A. G. Stromberg, A. A. Kaplin. Zavod. labor., XXX, 525, 1964.
11. A. G. Stromberg, V. I. Kuleshov. Coll. Methods for the Analysis of Chemical Reagents and High-Purity Preparations, 5–6, IREA, Moscow, 1963.
12. A. G. Stromberg. Izv. TPI, vol. 128, 1964.
13. A. G. Stromberg. Zavod. labor., 13, XXXI, 10, 1175, 1965.
14. A. G. Stromberg, A. A. Kaplin. ZhAKh, XX, 10, 1043, 1966.
15. A. A. Kaplin, V. M. Pichugina. Present collection, p. 138. Dissertation, TPI, 1964.
16. E. K. Spirin. Zavod. labor. (in press).
17. E. K. Spirin. Abstract of Candidate’s dissertation, Tomsk, 1964.
18. V. I. Kuleshov. Zavod. labor. XXXII, 4, 1966.
19. Yu. A. Karbainov. Dissertation, TPI, Tomsk, 1966.
20. B. F. Nazarov. Dissertation, TPI, Tomsk, 1967.
21. Yu. A. Karbainov, A. G. Stromberg. Coll. Modern Methods for the Analysis of Materials. Metallurgiya, Moscow, 1967.
22. Yu. A. Karbainov, A. G. Stromberg. ZhAKh, XX, 8, 769, 1965.
23. A. A. Kaplin, L. Dissertation, TPI, Tomsk, 1966.
24. A. A. Kaplin. Izv. TPI, Tomsk, 167, 1967.
25. A. G. Stromberg, Yu. N. Zhikharev. Zavod. labor. (in press). XXXI, 10, 1965.
26. A. A. Kaplin, B. F. Nazarov. Coll. Preparation and Analysis of High-Purity Substances. Nauka, Moscow, 1966.
27. A. A. Kaplin, A. G. Stromberg. Present collection, p. 112.
28. V. A. Igolinsky. Dissertation, Tomsk Polytechnic Institute, Tomsk, 1963.
29. V. A. Igolinsky, A. G. Stromberg. Coll. Methods for the Analysis of Chemical Reagents and Preparations, 5–6, IREA, Moscow, 1963.
30. L. F. Trushina, M. S. Zakharov. Izvestiya TPI (in press).