Physical Chemistry

Ya. I. Tur’yan and Yu. S. Gorodetskii

An Oscillographic Investigation of Oxides Formed on a Nickel Anode during the Electrochemical Evolution of Oxygen

(Presented by Academician A. N. Frumkin, 21 V 1957)

To elucidate the mechanism of the oxygen overvoltage, it is important to know the nature and concentration of the oxides formed on a smooth nickel anode in an alkaline solution.

A number of authors have shown that oxygen is evolved on a Ni anode covered with 1–2 layers \(^{(1,2)}\) or 5 layers \(^{(3)}\) of \(\mathrm{Ni_2O_3}\) \((\beta\text{-}\mathrm{NiOOH})\) \(^{(4-6)}\). In addition, according to some authors \(^{(2,4)}\), the anode also contains the higher oxide \(\mathrm{NiO_2}\). Despite the important role of this oxide in the process of electrochemical oxygen evolution \(^{(7,3)}\), it has not yet been established in what quantity it is formed on a smooth Ni anode. Moreover, the literature contains no information on the amount of oxides formed on a smooth Ni anode in the region of high (industrial) current densities \((>0.1 \mathrm{\,a/cm^2})\).

We developed a method for the oscillographic investigation of a smooth Ni anode that made it possible, along with the amount of stable oxide, to determine the amount of the higher unstable oxide. The essence of this method was that the vibrator of the loop oscillograph \(O\) (Fig. 1), recording the discharge curve, was connected into the circuit of the electrode under investigation even before the polarizing current was switched off. The vibrator was selected with such sensitivity that switching it on changed the magnitude of the polarizing current only very slightly. After the polarizing current was switched off, the “complete” discharge curve was recorded, including the reduction of the higher unstable oxide. In another experiment, when the vibrator was connected some time after the polarizing current had been switched off, an “incomplete” discharge curve was recorded, corresponding to the reduction of the stable oxides. From the difference between the areas of the “complete” and “incomplete” discharge curves, the amount of the higher unstable oxide was found.

Fig. 1. Electrical circuit and cell

The work was carried out on an 8-loop MPO-2 oscillograph with photographic recording. The spectrally pure wire Ni anode \(A\) under investigation (Fig. 1) (three electrodes were studied: \(S_1 = 0.155 \mathrm{\,cm^2}\), \(S_2 = 0.21 \mathrm{\,cm^2}\), and \(S_3 = 0.333 \mathrm{\,cm^2}\)) was discharged in a pair with a nonpolarizing iron electrode \(Zh\), taken from an ordinary alkaline accumulator. The electrolyte was electrochemically purified \(7.5\,N\) KOH. The anodic space was separated from the cathodic space by a glass diaphragm \(D\). To impart a stable surface to the Ni anode, the electrode was previously subjected to the action of a pulsating current. The quantity measured after this by the alternating-current method \(^{(8)}\)

the ratio of the true surface to the apparent one was ~5.1. Since the potential of the iron electrode did not change during the recording of the discharge curve, the potential of the Ni anode during discharge was determined from the known resistance \(R_3\) (it was chosen to be many times greater than the internal resistance of the cell) and the magnitude of the discharge current. The potential found in this way was very close to that measured by the potentiometric method.

Fig. 2. “Complete” discharge curve

The time scale on the oscillograms was obtained with the aid of a 50-cycle alternating current. The density of the polarizing current was varied in the range \(0.1—2\ \mathrm{A/cm^2}\); the density of the discharge current was \(0.009—0.02\ \mathrm{A/cm^2}\).

Since rather thick oxide layers were formed on the electrode, especially during prolonged anodic polarization, not all of the oxygen had time, during the recording of the discharge curve, to diffuse to the surface of the anode. Therefore, for complete reduction of the oxygen, after the first discharge curve it was necessary to record several more, with intervals of 3–5 min between recordings.

The approximate appearance of the “complete” discharge curve—the first and subsequent ones, up to the final reduction of the oxides—is shown in Fig. 2. The upper horizontal segment on the first curve corresponds to the polarizing current still being switched on. The point of current drop is obtained at the moment the polarizing current is switched off; then follows the middle horizontal segment, after which the current falls to the lower horizontal segment. Since the discharging cell is essentially a micro-alkaline accumulator, the lower segment of current drop almost to zero corresponds to the presence of NiO on the surface. The middle horizontal segment, located at potentials from \(+0.70\) to \(+0.40\ \mathrm{V}\) (relative to the normal hydrogen electrode), corresponds to the reduction of \(\mathrm{NiO_2}\) and \(\mathrm{Ni_2O_3}\). Like other authors \((^{1,3})\), we did not observe a separate plateau for the reduction of \(\mathrm{NiO_2}\), which can be explained by the closeness of the potentials \(\mathrm{NiO_2/NiO}\) and \(\mathrm{Ni_2O_3/NiO}\). The current drop from the middle segment to the lower one proceeds relatively slowly, which is possibly connected with the formation of intermediate oxides of the type \(\mathrm{Ni_3O_4}\) \((^{2,4-6})\), although we, like other authors \((^{1,3,9})\), were unable to detect an intermediate plateau.

The quantity of electricity corresponding to the “complete” discharge curve depended very little on the magnitude of the polarizing current (in the range \(0.1—2\ \mathrm{A/cm^2}\)), but with increasing duration of polarization (similarly to the data of \((^3)\)), this quantity of current increased. Thus, if with 5-minute polarization the area of the “complete” discharge curve was \(7—8\ \mathrm{mQ/cm^2_{true}}\), then with 30–60-minute polarization the area increased to \(11—12\ \mathrm{mQ/cm^2_{true}}\).

The “incomplete” discharge curve (Fig. 3) had no upper horizontal segment, since the polarizing current was switched off before the vibrator was switched on.

tor. The area under the “incomplete” discharge curve was smaller than the area under the “complete” discharge curve. With a very brief “rest” (several seconds), this reduction in area increased with increasing duration of the “rest,” but, beginning with 5–10 sec and up to several minutes of “rest,” a fairly constant value of the reduction in area was observed, of the order of 0.8–1 mC/cm\(^2\) of true area. This value was practically independent neither of the density of the polarizing current (0.1–2 A/cm\(^2\)) nor of the polarization time. From comparison with the complete discharge curve it was found that the reduction

Fig. 3. “Incomplete” discharge curve

of the curve occurred at potentials from \(+0.70\) to \(+0.60\) V. This makes it possible to explain the observed reduction in area by spontaneous decomposition of the higher oxide during the “rest” of the anode*:

\[ 2\mathrm{NiO}_2 = \mathrm{Ni}_2\mathrm{O}_3 + {}^1/_2\mathrm{O}_2 . \tag{1} \]

From the magnitude of the reduction in area, the number of layers of \(\mathrm{NiO}_2\) was found to be \(\sim 3\). This confirms the conclusion made by one of us \((^{7,10})\), on the basis of indirect observations, concerning complete coverage of the surface of the Ni anode with the higher oxide in the region of high current densities.

The amount of stable \(\mathrm{Ni}_2\mathrm{O}_3\) oxide was found from the area of the “incomplete” discharge curve, recorded after 20 sec of “rest,” after subtracting the amount of \(\mathrm{Ni}_2\mathrm{O}_3\) formed according to reaction (1). The amount of \(\mathrm{Ni}_2\mathrm{O}_3\) already after 5-minute polarization is \(\sim 17\) layers, and after one hour \(\sim 30\) layers. Thus, in the region of high current densities a smooth nickel anode is covered by an oxide layer that is much thicker than that found in works \((^{1–3})\), carried out at comparatively low current densities.

The invariability of the amount of oxides with a change in current density indicates the possibility of electrochemical liberation of oxygen not only through the processes of formation and decomposition of oxides, which is also consistent with indirect observations in other works \((^{7,11})\).

Kishinev State University

Received
20 IV 1957

CITED LITERATURE

1. A. Hickling, J. E. Spice, Trans. Farad. Soc., 11–12, 762 (1947).
2. El. Wakkad, S. H. Emara, J. Chem. Soc., 1953, 3504.
3. Л. М. Елина, Т. И. Борисова, Ц. И. Залкинд, ЖФХ, 28, 785 (1954).
4. O. Glemser, J. Einhand, Zs. Elektrochem., 54, 302 (1950).
5. T. Seiyama, M. Abo, W. Sakai, J. Chem. Soc. Japan, Industr. Sect., 57, 343 (1954).
6. W. Feitknecht, H. R. Christen, H. Studer, Zs. Anorg. Chem., 283, 88 (1956).
7. В. Н. Фисейский, Я. И. Турьян, ЖФХ, 24, 567 (1950).
8. А. Н. Фрумкин, В. С. Баготский, З. А. Иофа, Б. Н. Кабанов, Кинетика электродных процессов, Moscow, 1952.
9. G. W. D. Briggs, E. Jones, W. F. K. Wynne-Jones, Trans. Farad. Soc., 51, 1433 (1955).
10. Я. И. Турьян, И. С. Гольденштейн, ЖПХ, 28, 379 (1956).
11. С. А. Тантман, П. Д. Луковцев, Тр. совещ. по электрохимии, Izd. AN SSSR, 1953, p. 504.

* With prolonged “rest” (more than 5 min), in a number of experiments a further irreproducible reduction in the area of the “incomplete” discharge curve was observed, which apparently is connected with reduction of \(\mathrm{Ni}_2\mathrm{O}_3\) by impurities in the electrolyte.