On the Luminescence of Liquid and Solid Dielectrics in Strong Electric Fields

A. A. Vorob’ev and V. D. Kuchin

Modern theories of high-voltage electrical conductivity and electrical breakdown of solid dielectrics proceed from the assumption that the concentration of electrons in the conduction band increases at high field strength. The increase in the concentration of electrons in the conduction band may occur by impact, electrostatic, and thermal ionization, facilitated by a strong electric field. The reverse transition of electrons to the ground zone or to levels of inhomogeneities may be accompanied by the emission of energy [1]. The emission of semiconductor luminophores in a strong electric field is known, although it has been little studied [2, 3, 4, 5].

The observed blue and ultraviolet radiation of semiconductors under the action of high voltage has been attributed to discharges in air inclusions or along the surface of the semiconductor. G. Destriau [2] showed that the glow of a discharge in an air bubble located in a phosphor differs in color from the glow of the phosphor itself. When a phosphor is excited by an alternating voltage, the maximum brightness of the dielectric glow does not coincide with the maximum brightness of the glow of the gas discharge or with the voltage maximum.

D. Curie [3, 4] found that electroluminescence is associated with the displacement of electrons in the conduction band under the action of a strong electric field.

F. Matossi [5] extended to the phenomenon of electroluminescence the scheme of the mechanism by which luminescence arises in transitions of electrons between the filled band, the conduction band, and trapping levels.

P. Gudden and Pohl [6] observed an enhancement of the luminescence of zinc sulfide in a strong electric field. Electrons accelerated by an electric field \(E\) accumulate, over the mean free path \(\lambda\), an energy \(\omega = eE\lambda\) sufficient to liberate electrons from traps. Experiments by J. Mattler and D. Curie [7] with ZnSCu showed that an electric field of \(42\ \text{kV/cm}\) liberates electrons with a binding energy of \(0.75\ \text{eV}\).

G. Destriau [8] established that, in an alternating field with a strength of \(0.1\)—\(1.0\ \text{MV/cm}\) at a frequency of \(10\)—\(10000\ \text{Hz}\), sulfides, tungstates, silicates, germanates, and other luminophores exhibit a glow in visible light. Some luminophores increase the brightness of phosphorescence in an electric field after preliminary irradiation with ultraviolet radiation.

The dielectric constant of the electroluminescent phosphor \(\mathrm{ZnS}_{0.8}\mathrm{Se}_{0.2}:\mathrm{Cu}\) changes considerably when it is illuminated and when the magnitude of the electric-field strength applied to the specimen is varied. In tests in the dark, no change in the dielectric constant as a function of field strength was detected [9].

When a voltage of negative polarity was applied to a green carborundum crystal near the point supplying the voltage, a well-defined green luminous spot with an area of about \(1\ \mathrm{mm}^2\) was formed. Applying a voltage of positive polarity to the specimen after removal of the negative-polarity voltage causes the green luminescence to disappear and a yellow luminescence to appear in the crystal. With increasing applied voltage of negative polarity, bright bluish-white spots appear, and the current-voltage characteristic changes: the current begins to increase much more rapidly than the change in voltage [10].

On the basis of studies of absorption spectra and photoconductivity in crystals of alkali-halide salts, it is believed that the transition of electrons from the conduction band to the fundamental band can be accompanied by the emission of quanta with energies up to \(10\ \mathrm{eV}\) [6]. Observation of the emission of conduction electrons in dielectrics when they are captured by traps and pass into other states of lower energy in a strong electric field could help in understanding the physical phenomena preceding breakdown.

Fig. 1. 1—chamber with rock-salt windows; 2—spherical electrodes; 3—photographic plate.

Fig. 2. The designations are the same as in Fig. 1.

We carried out experiments to observe the luminescence of liquid dielectrics and their surfaces in strong electric fields. With the electrodes arranged as shown in Fig. 1, the surface of transformer oil began to glow faintly with white light when the calculated field strength at the boundary of the two media—transformer oil and air—reached \(13\ \mathrm{kV/cm}\). When an alternating voltage was applied to the electrodes, the surface of the oil began to oscillate strongly, and drops of oil reached the surface of the upper electrode, where they were drawn out in the direction of the field and coronaed intensely. The charge from the drops, spreading over the surface of the oil, produced a weak luminescence of it. Breakdown of the air gap occurred at a calculated field strength at the boundary of the two media equal to \(16.6\ \mathrm{kV/cm}\).

Weak luminescence of the oil–air interface was observed when the field strength reached \(28\ \mathrm{kV/cm}\) with the electrodes arranged as shown in Fig. 2. With increasing field strength the luminescence intensifies. No luminescence of the oil in the interelectrode gap was observed in this case. A photographic plate placed as shown in Fig. 2 was exposed at a calculated field strength of \(3.3\ \mathrm{kV/cm}\). The sensitivity of the plate was 130 units according to GOST. Corona of the electrodes was not observed in these experiments. The weak luminescence of the oil surface and the exposure of the photographic plate occur owing to the spreading of charge over the surface. In a nonuniform field (point opposite a sphere) in transformer oil

a corona appears at a field strength of 2.6 kV/cm. Corona discharge from the electrode in oil is accompanied by a pinkish glow of the region adjacent to the corona electrode. With pulses and with constant voltage of positive and negative polarities, at field strengths up to breakdown, no glow was observed in transformer oil, xylol, alcohol, or water, although corona discharge in a nonuniform field did occur.

We attempted to detect electroluminescence and electrophotoluminescence of samples of polystyrene, Plexiglas, and crystals of NaCl, KCl, and KBr in strong fields.

The sample was prepared as follows. In a piece of dielectric (1) (in Fig. 3) measuring \(55\times55\times110\) mm, two holes (2) were drilled: one with a diameter of 8–12 mm and a continuation of this hole (3) with a diameter of 0.8–0.12 mm, ending in a hemisphere (4). Into the recess (3) was poured a saturated solution of a salt in water or alcohol, which served as one of the electrodes. The second electrode was a graphitized or metallized plane (5). Voltage was supplied to electrode (3) through a conductor (6), which covered the inner surface of a glass tube up to 400 mm high and 8–11 mm in diameter (7), inserted into the sample. Under conditions of good sealing between the walls of the sample and the tube, a liquid electrolyte or mercury could serve as such a conductor. The space (8) between the walls of the glass tube and the walls of the sample was filled with TiO\(_2\) powder mixed with castor oil. Coating the edges of the hole (9) with such a paste considerably increased the corona voltage of these sections. Conductor (10) was connected to the high-voltage source.

Fig. 3. Construction of the sample.

The construction of the sample shown in Fig. 1 made it possible to obtain a maximum electric-field strength for NaCl from 1 to 1.63 MV/cm and had a high corona voltage (up to 100 kV). The maximum field strength at the tip of the point was determined from the formula:

\[ E=\frac{U}{r\varepsilon}\ \text{kV/cm}, \]

where \(U\) is the applied voltage,
\(\varepsilon\) is the dielectric constant of rock salt, and
\(r\) is the radius of the sphere, measured with the aid of a microscope with an ocular scale.

The maximum field strength during experiments with rock-salt crystals reached values from 0.7 to 1.3 MV/cm. The sample under investigation was placed in transformer oil, xylol, or air. The glow of the surface of liquid dielectrics in strong electric fields forced us to abandon observation of the radiation of solid dielectrics when they were immersed in liquid. Therefore, in our experiments the sample was in air. Constant, alternating, or pulsed (0.5/50 μsec) voltage was used. We attempted to detect the expected radiation of the sample visually, photographically, or with a quantum counter.

In a strong uniform and nonuniform field, luminescence of NaCl, KCl, and KBr crystals was observed in visible light, exposing a photographic film in which the specimen was wrapped. The sensitive photographic film was applied to the surface of the specimen or placed at various distances from it (Fig. 4). To record X-radiation, the photographic film was placed in a cassette opaque to visible light. The possibility of exposure of the photographic film due to corona discharge from the lead wires, the specimen, and the liquid surface, or due to leakage current over the specimen surface, was excluded. The intensification of the visible luminescence of crystals with \(F\)-centers and impurities excludes the use of photographic recording of the emission from impurity dielectrics, expected by us, in a strong electric field.

Fig. 4. 1—specimen, 2—photographic film, 3—light-tight chamber, 4—insulating liquid, 5—photographic film, 6—light-tight cassette with photographic film.

With the aid of photographic film we did not detect X-radiation from solid dielectrics in strong fields. To record the quanta of the expected electroluminescence of dielectrics, a quantum counter [8, 9] was used, having a lower sensitivity limit of about \(3000 \text{ Å}\). The sensitivity of the quantum counter decreased with decreasing wavelength. It did not count radiation from corona in air.

E. K. Zavadovskaya [12], using a quantum counter, established weak ultraviolet radiation of crystals caused by the action of \(\gamma\)-radiation. V. A. Kudryavtseva and others [13], using a quantum counter, studied the ultraviolet fluorescence of X-rayed rock salt.

To observe the expected radiation of dielectrics in a strong electric field in the ultraviolet region of the spectrum, we used a counter and an arrangement of type B. Table 1 presents the relative transmission coefficient for various materials. The source of ultraviolet radiation was a mercury-quartz lamp and a PRK-4 quartz lamp with a spectrum from 2400 to 6500 Å.

Table 1

Because of the considerable absorption of ultraviolet radiation by various materials, observations of the electroluminescence of solid dielectrics were carried out in air.

To observe the radiation of crystals, the quantum counter was placed at a distance of 32 mm or less from the specimen. The readings of the counter were determined at various voltages on the specimen, and also under simultaneous illumination of it with visible or ultraviolet light. Illumination was carried out with a beam parallel or perpendicular to the axis of the counter tube. The dependence of the counter reading on time was plotted for various experimental conditions. Taking into account the afterglow of crystals containing \(F\)-centers under the action of the field, heating, and illumination, determination of the counter readings under illumination of the specimen and a given field strength on it was alternated with determination of the counter readings in the absence of the crystal, etc. At field strengths up to 88% of breakdown, with the aid of the quantum counter we did not observe radiation of solid dielectrics caused by the presence of a high-strength electric field.

In the theory of luminescence, transitions of electrons from the conduction band to the ground band or to trapping levels by means of nonradiative transitions are known. It is assumed that in this case the electron transfers its energy to the lattice in collisions with phonons in relatively small portions—less than 0.1 eV. E. I. Adirovich [14] indicates that in nonluminescent crystals the probability of nonradiative electronic transitions exceeds that of transitions accompanied by radiation. The results of our experiments on observing the luminescence of crystals in strong fields are in agreement with the concepts of the theory of nonradiative electronic transitions.

Taking this opportunity, we express our gratitude to Docent F. I. Vergunas and Docent E. K. Zavadovskaya, whose advice in setting up the experiments and consultations in discussing the questions touched upon in the work were valuable to us.

REFERENCES

1. A. S. Zingerman. UFN, 46, 4, 450, 1952.
2. G. Destriau. J. Phys. Rad. 14, 307, 1953.
3. D. Curie. J. Phys. Rad. 14, 510, 1953.
4. D. Curie. J. Phys. Rad. 14, 672, 1953.
5. F. Matussi. Phys. Rev. 92, 846, 1953; 99, 1100, 1955.
6. P. Gudden. Leichtelektrische Erscheinungen, Berlin, 1928.
7. D. Curie. Arbeitstagung Festkörperphysik, 119, 1952.
8. Destriau. Phyl. Mag. 38, 700, 1947.
9. S. Roberts. JOSA 43, 7, 590, 1953.
10. K. Lehovec, C. A. Accardo and E. Jamgochian. Phys. Rev. 89, 20, 1953.
11. A. A. Vorob’ev. Proceedings of the Siberian Physico-Technical Institute, 3, 2, 3, 1935.
12. E. K. Zavadovskaya. Proceedings of the Siberian Physico-Technical Institute, 4, 3, 98, 1936.
13. V. A. Kudryavtseva. ZhETF, 4, 710, 1934.
14. E. I. Adirovich. Some Problems in the Theory of Crystal Luminescence, GTTI, 1951.

Tomsk Polytechnic Institute.