INVESTIGATION OF THE ANISOTROPY OF THE ELECTRICAL STRENGTH OF ALKALI-HALIDE CRYSTALS

L. T. MURASHKO

(Presented by the scientific seminar of the Scientific Research Institute of High Voltage)

In studying electrical breakdown in model dielectrics, namely alkali-halide crystals, it was found that the breakdown paths follow definite crystallographic directions [1, 2, 3]. This gave grounds to suppose that the electrical strength of crystals depends on the crystallographic direction, i.e., that anisotropy of electrical strength exists.

Many works, both theoretical and experimental, have been devoted to the study of the anisotropy of electrical strength [4, 5, 6, 7, 10]. Interest in this phenomenon is understandable, since its solution would provide material by which one could judge the correctness of certain theories of breakdown [9, 10].

However, despite the fact that the crystallographic directionality of the discharge in alkali-halide crystals had long been established [1, 2, 3] and a number of works [9, 10] theoretically predicted anisotropy of electrical strength, the assumption was expressed that electrical strength does not depend on crystallographic direction. This was promoted by the experiments of A. A. Vorob'ev, Hippel, and Davisson, which did not reveal any difference in electrical strengths along different crystallographic directions.

Only recently have reports appeared concerning the anisotropy of the electrical strength of ionic crystals in a nonuniform field [11]. However, it should be pointed out that although the data presented in [11] indicate that discharges in NaCl most frequently occur along the [110] direction, the electrical strength in this case of field application is not the smallest, which is contradictory.

Our task was to establish the anisotropy of electrical strength in alkali-halide crystals in a uniform field. For this purpose we used the method [7] developed by us earlier, which made it possible to visually monitor the quality of the specimens. The experiment was carried out with NaCl, KCl, and KBr crystals.

NaCl specimens were obtained from natural rock salt, while KCl and KBr specimens were obtained from crystals grown by the Kyropoulos method. Specimens for investigation along the [100] direction were cleaved from the crystal in the usual manner. Specimens for investigation along the [110] and [111] directions were cut on a special machine.

The studies were carried out with a pulsed voltage of duration 0.5/2000 μsec; the voltage was applied to the specimen in steps; the overvoltage did not exceed 5%. The change in voltage across the specimen was recorded by means of an oscillograph on photographic film.

The measurements showed that, in the investigated specimens of the alkali-halide salts NaCl, KCl, and KBr, anisotropy of the electrical strength is observed.

The anisotropy of the electrical strength of crystals decreases with decreasing thickness, and at a thickness of 3 μm no changes in electrical strength along different crystallographic directions are observed.

Specimens made from crystals having the same type of crystal lattice but different chemical composition have different types of anisotropy.

At room temperature, for rock salt we have the following relation (Fig. 1):

\[ E_{[110]} < E_{[111]} < E_{[100]}; \]

for KBr crystals (Fig. 2):

\[ E_{[111]} < E_{[100]} < E_{[110]}; \]

For KCl specimens (Fig. 3):

\[ E_{[110]} < E_{[100]} < E_{[111]}. \]

Fig. 1. Dependence of the electrical strength of NaCl on crystallographic direction.
1 — [100]; 2 — [111]; 3 — [110].

Fig. 2. Dependence of the electrical strength of KBr on crystallographic direction.

As can be seen from the results presented, in crystals having the same structure the conditions for the development of an electrical discharge along different crystallographic directions are different.

All this indicates that, in order to determine and explain the type of anisotropy, Hipple’s scheme [9] is insufficient, since it considers only the change in the potential barrier created by the lattice ions when an electron moves along different crystallographic directions, whereas the character of its change is the same for all alkali-halide crystals.

The results for NaCl and KCl specimens agree with calculations of the “preferred” direction of electrical breakdown of alkali-halide salts, carried out by Callen and Offenbacher [10], who calculated scattering probabilities as a function of electron energy for

different directions of electron motion in the crystal lattice, based on the assumption that conduction electrons interact with longitudinal vibrations of the lattice corresponding to the optical branch of the dispersion curve.

The results for KBr do not agree with their calculations.

Thus, on the basis of the experimental data obtained, it may be said that the electrical strength of ionic crystals is anisotropic.

The observed anisotropy of the electrical strength of ionic crystals can be attributed to the fact that all specimens were carefully prepared for the experiment. Preparing them under a microscope made it possible to reject specimens with macrodefects, and to position the electrodes so that the applied field corresponded precisely to the chosen crystallographic direction.

Fig. 3. Dependence of the electrical strength of KCl on crystallographic direction.

Of great importance for reducing scatter is the examination of specimens after breakdown. Very often specimens break down not in the direction of the applied field, but along a “weaker” one.

In such cases the breakdown voltage proves to be lower, which introduces a significant change both in the value of the electrical strength and in the character of the anisotropy. The electrical strength of such specimens must be assigned to the direction along which the breakdown channel developed.

Only such an approach, in our view, can lead to the correct determination of the dependence of electrical strength on crystallographic direction.

In addition, further investigations showed that the anisotropy is manifested more clearly when an electrolyte is used as the electrodes.

REFERENCES

1. A. F. Val’ter, L. D. Inge, ZhTF 71, 627, 1931.
2. A. Hippel, Zs. f. Phys. 67, 707, 1931.
3. J. Lass, Zs. f. Phys. 69, 313, 1931.
4. M. E. Caspari, Phys. Rev. 98, 1679, 1955.
5. A. A. Vorob’ev, E. K. Zavadovskaya, Electrical Strength of Solid Dielectrics, 1956.
6. I. W. Davisson, Phys. Rev. 70, 685, 1946.
7. I. W. Davisson, Hippel A., Phys. Rev. 57, 1956, 1940.
8. G. A. Vorob’ev, V. A. Kostrygin, L. T. Murashko, PTE 5, 1961.
9. A. Hippel, Zs. f. Phys. 75, 145, 1932.
10. E. L. Offenbacher, H. B. Callen, Phys. Rev. 90, 3, 1953.
11. R. Cooper, D. T. Crossart, A. A. Wallace, Pros. Phys. Soc. 70, 4463, 1957.