STUDIES ON THE PHYSICS OF DIELECTRICS AND ELECTRICAL INSULATION CONDUCTED IN TOMSK BEFORE 1957

A. A. VOROB'EV

Introduction

Before the organization of the Siberian Physicotechnical Institute (SFTI) in Tomsk, Prof. V. D. Kuznetsov and his coworkers conducted scientific investigations of the mechanical properties and crystallization of ionic dielectrics, and of the electrical properties of solid and liquid dielectrics and semiconductors. After the organization of SFTI, under the direction of Prof. V. D. Kuznetsov and Prof. M. A. Bol'shanina, extensive investigations of the mechanical properties of solids were developed.

At SFTI in 1930, Prof. P. S. Tartakovskii and his coworkers began theoretical and experimental studies of electronic phenomena in dielectrics and semiconductors. The development of the theory of the mechanical properties of ionic structures, their stability, and other problems in the theory of the solid state, begun by Prof. P. S. Tartakovskii, was continued under the direction of V. A. Zhdanov.

In the thirties at SFTI, Prof. V. N. Kessenikh began studies of the properties of dielectrics in high-frequency fields, and the development of research methods and installations; Docent A. M. Venderovich studied electrical conductivity and ionic processes in solid dielectrics; Docent M. I. Korsunskii studied the action of electron beams and X-rays on dielectrics, and others.

Beginning in 1935, K. A. Vodop'yanov organized at SFTI a laboratory of dielectric losses, and A. A. Vorob'ev, together with A. K. Krasin, organized a high-voltage laboratory; in these laboratories, studies of insulating materials and constructions were also developed.

Beginning in 1939, A. A. Vorob'ev initiated scientific studies of dielectrics and insulating materials and constructions at the Tomsk Polytechnic Institute. Since 1954, under the direction of N. I. Vorob'eva, studies of the properties of insulating materials have been conducted in the Department of Insulation and Cable Engineering, while P. A. Savintsev, in the Department of Physics, has conducted studies of the physical properties of insulating crystals.

Scientific investigations of dielectrics and electrical insulation developed still further at TPI beginning in 1957, after the organization of the laboratory of dielectrics, semiconductors, and electronics.

An important role in the development of scientific research was played by scientific seminars on dielectrics, which had already been held regularly at SFTI in the thirties, and at TPI since 1950.

Electronic Transitions in Dielectrics

Energy Scheme of Electronic Levels

(A. A. Vorob’ev, N. K. Kalabukhov, I. K. Kurshev, V. M. Kudryavtseva, V. F. Poddubnyi, L. M. Reverdatto, P. S. Tartakovskii, N. N. Fedenev, B. M. Fishelev, and others).

In the 1930s, photoelectric phenomena in crystals of sulfur and alkali-halide salts were investigated at the SFTI, and a scheme of electronic levels was constructed for the latter. The results of these studies were described in P. S. Tartakovskii’s monograph The Internal Photoeffect in Dielectrics.

S p a r k l i n g c r y s t a l s. In 1935 E. K. Zavadovskaya showed that gamma radiation causes luminescence in crystals of quartz, fluorite, rock salt, sylvine, and calcite. Thus, a crystal counter for hard radiation was constructed, which has now become widely used.

Fluorescence in Strong Electric Fields

(A. V. Astafurov, A. A. Vorob’ev, V. D. Kuchin)

It proved impossible to detect experimentally any broadening of electronic bands in crystals under the action of strong electric fields. Luminescence of alkali-halide salt crystals in strong fields was detected. In ice crystals with a large number of impurities, strong luminescence is observed in an electric field.

Effect of Radiation on Dielectrics

(A. A. Vorob’ev, A. F. Kalganov, B. A. Kononov, V. A. Moskalev, A. D. Shchelokov).

Many crystals, under the action of fast electrons with energies of several MeV and of hard radiation, and after its cessation, glow with visible light characteristic of them. Under electron bombardment, rock salt is colored brown, and sylvine violet. The density of the coloration increases with the number of electrons that have bombarded the crystal. To obtain sharp images of the electron beam, coloration to a blackening density equal to 2 is necessary; this occurs if up to \(10^{14}\) electrons fall on the crystal. Crystals grown from the melt become colored at smaller doses than natural rock salt.

Crystals fluoresce after being irradiated with electrons.

The mass absorption coefficients of X-rays with an effective wavelength of 0.22 angstrom in alkali-halide single crystals are as follows (in square cm per gram):

The mass absorption coefficient has no linear relationship with density or molecular weight. As these quantities increase, the mass absorption coefficient increases. With increasing energy of the crystal lattice, the mass absorption coefficient generally decreases.

The degree of absorption of X-rays in crystals is determined by the absorption constant, which is independent of the wavelength of the absorbed rays over a large interval of wavelengths.

The concept has been introduced of the effective number of electrons in a molecule of a compound (with respect to X-ray absorption), which determines the degree of absorption of X-rays and is practically equal to the additively obtained number of electrons in the molecule.

The absorption constant for the wavelength interval 0.3–0.03 angstrom is determined through the effective number of electrons in the molecule to the power 2.58.

The mass absorption coefficient for a given wavelength is proportional to the absorption constant. The proportionality coefficient depends on the wavelength to the power 2.8.

The linear coefficient of X-ray absorption in crystals of solid solutions, as a function of composition, approximately obeys the law of additivity. The existing slight positive nonadditivity near the equimolar composition is in agreement with the negative nonadditivity for the lattice constant.

The considerable dependence of the linear coefficient of X-ray absorption in solid solutions on their composition may be used to determine composition from absorption.

The absorption constant of solid solutions is formed additively from the absorption constants of the components. The linearity of the mass absorption coefficient as a function of the absorption constant is preserved for solid solutions as well.

By determining the mass absorption coefficients of a constant lattice, one can find the dependence of the number of particles in the elementary cell on composition.

At room temperatures, the conductivity of crystals of alkali-halide salts under irradiation by X-rays increases by a factor of 10–100. With increasing temperature, the effect of the increase diminishes and disappears at temperatures of 80–100 degrees Celsius.

An exception is the lithium fluoride crystal, whose conductivity under irradiation and at low temperatures decreases slightly.

Natural rock salt increases its conductivity during irradiation by one thousand to ten thousand times more strongly than a single crystal artificially grown from the melt.

At constant temperature, the activation energy increases with increasing energy of the crystal lattice and decreases with increasing mass coefficient of X-ray absorption in the crystal.

In measurements of weak currents in crystals of alkali-halide salts irradiated by X-rays, a potential difference arises between the measuring electrode and ground which is not connected with the conductivity of the crystal and is due to the phenomenon of the external photoelectric effect from the measuring electrode. The electrode (platinum), insulated from ground and irradiated by X-rays, becomes positively charged up to 3 volts relative to ground (photo-e.m.f.).

A system of two electrodes in a vacuum, one of which is insulated from ground and irradiated by electromagnetic radiation, while the other is grounded and not irradiated, constitutes a new type of source of e.m.f., operating at the expense of the energy of radiation quanta.

The magnitude of the photo-e.m.f. does not depend on the intensity of irradiation, but is determined by the maximum energy of the quantum (in nonhomogeneous radiation)

... and is equal to the maximum energy of the quantum minus the work function of the photoelectrons, divided by the charge of the electrode (under irradiation with X-rays—tens and hundreds of kV).

The internal resistance of the source of photo-emf (under irradiation with X-rays) is proportional to the quantum energy and inversely proportional to the intensity of irradiation.

The short-circuit power of the source of photo-emf is determined by the product of the maximum energy of the photoelectron and the number of photoelectrons emitted per second, i.e., it is proportional to the maximum quantum energy and the intensity of irradiation, and under irradiation with X-rays reaches microwatts.

When the source of photo-emf is closed through a resistance, some of the photoelectrons lying within a certain energy interval, whose upper boundary is the maximum energy of the photoelectrons, reach the grounded electrode, forming the current of the internal circuit of the source; and from the grounded electrode, through the load, the same number of electrons per second flows in, forming the current of the external circuit. The potential of the insulated electrode decreases even at loads of tens of megohms. As the load resistance is decreased, the effective energy interval widens owing to the displacement of the lower boundary of the interval toward lower photoelectron energies. The current in the circuit in this case increases, the voltage across the external load decreases, and the power released in the load rises to a maximum at a certain resistance; with a further decrease in the load resistance the power falls to zero (under short-circuit conditions).

The current in the circuit, the voltage, and the power in the load are expressed mathematically through the distribution function of the photoelectrons over energies. Determining the latter constitutes the principal difficulty in calculations for particular cases.

The possibility has been demonstrated of the practical implementation of direct transformation of the energy of electromagnetic radiation into electrical energy, based on the phenomenon of the external photoelectric effect. For example, the gamma radiation of atomic reactors can be used to obtain high voltages with the aid of appropriate “photobatteries.”

In electrometric measurements under irradiation conditions it is necessary to take into account the presence of the photo-emf arising at the electrodes.

The Influence of Various Factors on the Electrical Strength of Dielectrics

The Influence of Electrode Material on the Electrical Strength

(A. A. Vorob’ev, A. K. Krasin)

The electrical strength of dielectrics in thin layers in homogeneous and nonhomogeneous fields does not depend on the electrode material. With a dielectric thickness on the order of 10 mm, the electrical strength is lower when a cathode made of a metal with a smaller electron work function is used.

In a homogeneous electric field, the breakdown voltage of dielectrics varies linearly as a function of thickness. In nonhomogeneous electric fields, the breakdown voltage under static and impulse voltages increases with the thickness of the dielectric layer more slowly than according to a linear law.

In the region of thermal breakdown of rock-salt crystals, a linear dependence of the breakdown voltage on the specimen thickness is observed over a large interval of thicknesses.

Crystallographic Directionality of Discharges

(A. V. Astafurov, A. A. Vorob’ev, G. A. Vorob’ev, V. D. Kuchin,
K. K. Sonchik)

With negative polarity of the point, a discharge in alkali-halide crystals propagates predominantly along the direction of the cube edge, whereas with positive polarity it propagates predominantly along the diagonal of a face and the diagonal of the cube. The presence of mechanical stresses and damage within the volume of the crystal disturbs the discharge paths.

The maximum electric strength of rock salt in a uniform field, at a thickness on the order of half a millimeter, is greatest along the cube edge, smaller along the face diagonal, and smallest along the cube diagonal. The average electric strength does not depend on the crystallographic direction. The electric strength along the cube edge is 10 percent higher than the strength along the cube diagonal.

The breakdown voltage with a negative point and impulse voltage along the face diagonal is 20 percent lower than along the cube edge. With a positive point and impulse voltage, the breakdown voltages along the three principal crystallographic directions are the same.

Measurement of Currents before Breakdown

(G. A. Andreev, A. K. Krasin)

At a specified high voltage close to the breakdown voltage, the current through a dielectric increases with time. A direct dependence is observed between the logarithm of the prebreakdown current and temperature for alkali-halide crystals. In the temperature range of 20–180 degrees, the straight line has a larger slope than in the temperature range of 180–440 degrees.

Over a wide temperature interval (from 20 to 440 degrees), the current as a function of the relative electric-field strength in the specimen varies exponentially. In the temperature range of 20–180 degrees, the prebreakdown current does not depend on the thickness of the specimens. At temperatures above 200 degrees, the prebreakdown current depends on the thickness of the specimen. As the thickness of the specimen increases, the current increases in absolute magnitude. At temperatures above 200 degrees, there is a linear dependence of the prebreakdown current on the energy of the crystal lattice.

Electrical Strength and Thermochemical Characteristics of Dielectrics

(A. A. Vorob’ev, G. A. Vorob’ev, E. K. Zavadovskaya, V. D. Kuchin,
A. F. Kalganov)

The electrical strength of ionic crystals increases with increasing electron binding energy, the melting temperature of the compounds, the heat of formation, and the lattice energy. The electrical strength of gaseous and liquid dielectrics increases with increasing forces of bonding between their molecules.

Dependence of the Breakdown Voltage of Solid Dielectrics on the Polarity of the Electrodes

(A. V. Astafurov, A. A. Vorob’ev, G. A. Vorob’ev, V. D. Kuchin,
E. K. Zavadovskaya)

A specimen form with a conical recess has been developed for breakdown testing in a nonuniform field, in which discharges in the surrounding medium are eliminated and, consequently, so is its influence on the test result.

The breakdown voltage of solid dielectrics, as well as of gaseous ones, is lower with a positive point than with a negative one. This phenomenon was observed for crystalline dielectrics and glasses at layer thicknesses from several tenths of a millimeter to twenty millimeters.

Dependence of the Electric Strength of Ionic Crystals on Temperature

(G. A. Andreev, A. A. Vorob’ev, V. D. Kuchin)

The electric strength of ionic crystals, for exposures from tenths of a microsecond to several seconds, in the low-temperature region, increases linearly with increasing temperature at a rate on the order of \(1.75\ \text{kV/cm deg}\).

In the region of positive temperatures above \(50\text{—}100^\circ\text{C}\), for crystals of alkali-halide salts an exponential decrease of the breakdown voltage with temperature is observed. In the temperature range from \(20^\circ\) to \(440^\circ\text{C}\), among crystals of alkali-halide salts, compounds with a lower lattice energy have a lower breakdown voltage.

At exposures on the order of hundredths of a microsecond, the electric strength of alkali-halide crystals does not depend on temperature. At temperatures above \(180\text{—}200^\circ\text{C}\), for alkali-halide crystals the breakdown voltage decreases exponentially with increasing temperature, which satisfies V. A. Fock’s theory of thermal breakdown.

The breakdown voltage of crystals during breakdown in a nonuniform field, on pulses of both polarities with a front length of several microseconds, does not depend on temperature.

At all temperatures and durations of voltage application, the electric strength is the higher, the greater the energy of the crystal lattice of the dielectric. For the region of thermal breakdown (\(200\text{—}450^\circ\text{C}\)), there is a linear dependence of electric strength on crystal-lattice energy.

Participation of Electrons in the Breakdown of Solid Dielectrics

(A. A. Vorob’ev, G. A. Vorob’ev, K. K. Sonchik)

Discharges in partially colored crystals of rock salt on pulses proceed predominantly in the part with the greater concentration of electrons. On pulses with exposures on the order of hundredths of a microsecond, photoconductivity and the presence of color centers do not affect the electric strength of crystals.

With an increase in exposure in the interval from tenths of a microsecond to constant voltage, i.e., within the limits of the volt-time characteristic where the influence of space charges on electric strength is manifested, the influence of color centers and photoconductivity on breakdown is observed. The electric strength of colored crystals is lower than that of uncolored ones.

Breakdown under photoconductivity is accompanied by a decrease in the strength of colored crystals.

During breakdown in a nonuniform field under pulse voltage and negative polarity of the point, the breakdown voltage decreases with coloration of the crystals and photoconductivity, as in a uniform field, but the difference in breakdown values is smaller than in a uniform field. With positive polarity of the point, a slight increase in breakdown voltage is observed for colored specimens broken down in darkness.

The delay time of the discharge in colored crystals of rock salt ...

salts punctured under illumination and uncolored salts at the same value of the electric-field strength are identical.

Impulse Strength of Solid Dielectrics

(A. V. Astafurov, A. A. Vorob’ev, G. A. Vorob’ev, V. D. Kuchin,
K. M. Kevroleva, M. A. Mel’nikov, K. K. Sonchik, A. M. Trubitsyn,
A. M. Venderovich, E. K. Zavadovskaya)

The breakdown voltage of solid dielectrics decreases with an increase in the number of acting impulses. Therefore, at TPI a method was developed for single-impulse breakdown tests on the front of a voltage wave and under the action of a wave with a sloping front.

The volt–time characteristics of various solid dielectrics—ionic single crystals, crystalline organic materials, glass, and paraffin—over a wide range of thicknesses from several hundredths of a millimeter to 20 mm, in homogeneous and nonhomogeneous fields, retain the same form.

As the duration of voltage application is reduced from several seconds to hundred-thousandths of a second, the breakdown voltage remains constant. Reducing the exposure to tenths of a microsecond at room temperature is accompanied by a decrease in the breakdown voltage. A further reduction of the exposure is accompanied by an increase in the breakdown voltage, which is a consequence of discharge lag.

The minimum in the course of the dependence of breakdown voltage on exposure decreases as the temperature is lowered. For crystals of alkali-halide compounds in the region of negative temperatures, the minimum on the volt–time characteristic is absent, which indicates that the minimum is due to the action of volume charges.

Ionic space charge increases the measured value of the electric strength of dielectrics, whereas electronic space charge lowers it.

Ionic space charge, in the dependence of electric strength on the duration of voltage application at positive temperatures, causes the formation of a “bucket.” In the breakdown of crystalline dielectrics by impulses with a front length of several microseconds, the electric strength is 15–25% lower than in breakdown under direct voltage.

For annealed and unannealed crystals, the general form of the volt–second characteristic is the same. However, for annealed single crystals, the increase in electric strength occurs at exposures not of the order of tens of microseconds, but of hundreds of microseconds. In annealed single crystals the minimum of the volt–time characteristic shifts toward larger exposures.

The time of discharge formation in solid dielectrics, defined as the discharge lag time under the condition that statistical lag is absent, is equal to several hundredths of a microsecond. The time of discharge formation in solid dielectrics increases with increasing thickness. The average velocity of discharge development, defined as the ratio of the thickness of the dielectric layer along the breakdown path to the breakdown-formation time, is of the order of tens of kilometers per second. It is less than the discharge velocity in gases by one to two orders of magnitude; the discharge velocity increases as the lattice energy decreases.

With positive polarity of the point, the velocity of discharge development in solid dielectrics is greater than with negative polarity by a factor of 3–6.

As the temperature is raised from minus one hundred degrees to 60–80 degrees, the average velocity of discharge development in crystals increases. At temperatures above 60–80 degrees it decreases.

In dielectrics with a layer thickness of tens of millimeters, the average

the rate of propagation of the discharge is greater than in layers with a thickness of hundredths of a millimeter. Just as in gases, the average rate of development of a discharge in solid dielectrics increases with increasing overvoltage.

With a positive point, the average rate of development of a discharge in solid dielectrics is 3–6 times greater than with a negative one.

Influence of Molecules of Water of Crystallization on the Physicochemical Properties of Crystalline Hydrates

(K. A. Vodopyanov, A. A. Vorob'ev, A. P. Izergin, K. M. Kevroleva,
P. A. Savintsev, M. P. Tonkonogov, N. M. Torbin)

Molecules of water of crystallization and constitutional water, entering the lattice of crystalline hydrates, increase the heat of formation and the partial pressure.

Molecules of water of crystallization reduce the electrical strength, the hardness in mutual grinding of crystals, the refractive index, the mass absorption coefficient of X-rays, and the energy of the crystalline lattice of the compound.

The electrical conductivity of crystalline hydrates increases at the dehydration temperature of the latter.

Ionic crystals containing polar molecules in their composition have a relaxation character of dielectric losses, caused by the orientation of the polar molecules.

The relaxation mechanism of dielectric losses in ionic polar crystals (talc, gypsum, copper sulfate, etc.) manifests itself in the form of frequency and temperature maxima of the dielectric-loss angle, which shift toward high frequencies with increasing temperature.

The temperature-frequency maxima of the dielectric-loss angle of crystalline hydrates are caused by the orientation of polar molecules of water of crystallization. The structure of the crystalline hydrate determines the number of temperature and frequency maxima of the loss angle, while their position and magnitude are determined by the activation energy and the number of weakly bound polar molecules.

Toward a Theory of Electrical Breakdown of Solid Dielectrics

(A. A. Vorob'ev, E. K. Zavadovskaya, V. D. Kuchin)

At the moment of breakdown the mobility of electrons is minimal, and their interaction with the lattice is maximal. The energy accumulated by an electron is expended on destruction of the crystal lattice and on overcoming the interaction of ions. From the condition of proportionality between the electron energy and the lattice energy, and from the change in electron mobility, a calculation gives a linear dependence between the electrical strength of the crystal and the lattice energy, in satisfactory agreement with experiment.

The electrical strength of homogeneous dielectrics does not depend on mechanical load. Compaction of a dielectric under the action of a load is accompanied by an increase in its mechanical strength, while disruption of the continuity of the dielectric is accompanied by a decrease in electrical strength.

High electric fields in dielectrics, sustained under the action of short pulses, cause the appearance of mechanical stresses exceeding their mechanical strength. This led to the expla-

... to breakdown on short pulses as a mechanical destruction of dielectrics.

The phenomenon of breakdown of solid dielectrics appears to be a complex process. At high temperature and long exposure, thermal destruction of the dielectric develops predominantly. Ionization processes, complicated by the action of space charges and leading to destruction of the dielectric, determine the third mode of breakdown of solid dielectrics. Each of the three indicated modes of destruction of dielectrics in an electric field is not opposed to the others and operates within a certain range of temperatures and exposures.

Systematization of the Properties of Ionic Crystals

(A. A. Vorob'ev)

The properties of ionic crystals are associated with the energy of the crystal lattice and vary as shown in the table.

Properties of ionic crystals and the energy of the crystal lattice.

The table provides indications of the directions for searches for ionic dielectrics with the necessary physical properties.

Properties of Dielectrics and Molecular Concentration

(P. A. Savintsev)

The physicochemical properties of ionic crystals are compared with molecular concentration, defined as the ratio of the density of the crystal to its molecular weight. With increasing molecular concentration, Brinell hardness, scratch hardness, hardness in mutual grinding, hardness by damped oscillations, drilling hardness, elastic moduli, yield pressure, melting and boiling temperatures, heat of formation of the solid compound, heat of fusion and sublimation, and Debye temperature increase; the coefficient of linear expansion, heat capacity, electrical conductivity, dielectric losses, and refractive index of light decrease.

These results make it possible to regard molecular concentration as a physical quantity characterizing the bond between the particles of the lattice. Molecular concentration increases with an increase in the energy of the lattice.

Properties of Dielectric Alloys

(V. E. Avericheva, K. A. Vodop’yanov, A. A. Vorob’ev, A. F. Gorodetskii, E. K. Zavadovskaya, M. S. Ivankina, A. N. Kislina, M. A. Krivov, G. I. Potakhova, P. A. Sazhinsev, D. D. Saratovkin, A. M. Trubitsyn)

The proposition of Academician N. S. Kurnakov concerning the correspondence between mechanical properties and heat of formation has been extended to the electrical properties of solid solutions. It has been shown that, as the heat of formation of single crystals of solid solutions increases, their stability, hardness, and electric strength decrease; dielectric losses, electrical conductivity, etc., increase.

Electrical conductivity, electric strength, and dielectric losses at high frequency in single crystals of solid binary solutions of alkali-halide salts depend on composition. In the region of equal content of the components and the greatest lattice disturbance, there are observed a maximum of losses, a maximum of heat of formation and of the coefficient of linear expansion, of electrical conductivity, and a minimum of hardness under mutual polishing and of electric strength.

Single crystals of solid solutions, after being grown, have considerable defectiveness, and the number of ions in the elementary cell is less than eight. With storage time, during normalization, the defectiveness of the single crystals decreases, and the number of ions in the elementary cell approaches eight. Upon normalization of the single crystals and improvement of their structure, microhardness and electric strength increase, while electrical conductivity, heat of formation, coefficient of linear expansion, etc., decrease.

A comparison of the physical properties of solid solutions of alkali-halide salts with molecular concentration shows that, with increasing molecular concentration, the electric strength, melting temperature of the solid solutions, Debye temperature, and relative hardness under mutual grinding and drilling increase, while the heat of formation, coefficient of linear expansion, electrical conductivity, microhardness, and Brinell hardness decrease.

A new physical phenomenon has been discovered and studied: the lowering of the melting temperature at the point of contact of crystals capable of forming eutectics and solid solutions with a descending liquidus line. In the contact of dissimilar alkali-halide crystals, the melting temperature is lowered by tens and hundreds of degrees in comparison with the melting temperature of the crystals in contact. It has been shown that the low melting point of a eutectic is not a property of their special structure or definite chemical composition.

Studies of Dielectric Losses in High-Frequency Fields

On the Mechanism of Dielectric Losses in Ionic Nonpolar Crystals

(K. A. Vodop’yanov, A. A. Vorob’ev, and G. I. Potakhova)

Dielectrics with a more perfect structure and greater bond energy between structural elements, like pure liquids that contain no polar molecules, have smaller dielectric losses.

Measurement of the dielectric loss angle in alkali-halide crystals at high frequency \((10^5—10^8\ \text{cps})\) showed that, at the same temperature and frequency, the loss angle increases with an increase in the lattice constant and decreases with an increase in the melting temperature, heat of formation, heat of sublimation, and molecular weight of the crystal.

The mechanism of dielectric losses in ionic crystals may be represented as follows. Under the influence of an external action (electrical, thermal, etc.), the number of weakly bound ions in the crystal, having different activation energies, increases. A small number of ions possessing a high activation energy, under the influence of an electric field in the crystal, undergo translational motion, producing a through conduction current. The greater part of the ions, which have a lower activation energy, under the action of an alternating electric field will undergo a return-translational displacement within a limited volume, giving rise to dielectric losses of a relaxation character. The smaller the energy of the crystal lattice, the greater will be the number of weakly bound ions giving rise to dielectric losses in it. Therefore the dielectric loss angle is a characteristic of crystals along with their other physicochemical characteristics.

Dielectric losses in inorganic glasses

(K. A. Vodop’yanov, R. M. Kessenikh, M. A. Krivov, V. A. Presnov, A. A. Khar’kov)

In pure silicate and borate glasses, dielectric losses are due to vibrations of structural elements with small amplitude, independent of changes in temperature \((20—320^\circ\text{C})\) and in the frequency of the electric field \((10^5—10^8\ \text{cps})\).

For two-component glasses, an increase in the loss angle with increasing concentration of the impurity ion is characteristic. Moreover, monovalent ions cause a more appreciable increase in the loss angle than divalent ions. At low impurity concentrations the loss angle does not depend on frequency. Analysis of the temperature-frequency dependence of the loss angle shows that dielectric losses are due to relaxation displacement of impurity ions.

In the case of high impurity concentrations, up to 30%, ohmic dielectric losses appear in two-component silicate and borate glasses.

It has been experimentally proved that dielectric losses in simple glasses, as in alkali-halide crystals, are determined by the chemical composition and by the strength of the chemical compound.

In the study of three-component silicate and borate glasses it was found that there exists a functional dependence between the loss angle and the energy of dissociation of the ions, on the basis of which it is possible to determine the concentration of impurity ions corresponding to the minimum of the dielectric loss angle. It is shown that dielectric losses in ternary glasses differ from ohmic losses and are associated with processes of a relaxation character.

The regularities found in the characteristics of dielectric losses in simple glasses made it possible to relate the physicochemical properties of multicomponent glasses to their dielectric losses. This made it possible to prepare new grades of glass possessing high dielectric properties.

Dielectric Losses in Steatite Ceramic Materials

(V. M. Belousov, K. A. Vodop’yanov, A. P. Vyatkin, M. A. Krivov)

In studying the dielectric properties of steatite ceramic materials as a function of their firing temperature and gaseous medium, regularities were obtained for the temperature–frequency behavior of the dielectric-loss angle that are characteristic of the relaxation mechanism of dielectric losses. It was established that, in order to obtain a high-quality insulating material, an appropriate composition of components and manufacturing conditions are necessary so that, as a result of the final firing, good sintering of the body occurs, whose basis is the crystalline phase. In both the crystalline and amorphous phases there must be structural elements with high activation energy that do not give dielectric losses of a relaxation character.

Dielectric Properties of Mica from East Siberian Deposits

(K. A. Vodop’yanov, I. G. Vorozhtsova, A. P. Izergin)

The nature of the dielectric losses of muscovite mica in the high-frequency range differs from ohmic losses when the electric field is perpendicular to the plane of cleavage. Depending on frequency and temperature, spotted mica may exhibit dielectric losses of both ohmic and relaxation character.

The ohmic character of dielectric losses in spotted mica appears when the electric field acts parallel to the plane of cleavage.

The dielectric-loss angle of muscovite mica depends on the type and concentration of mineralogical inclusions and crystallization water. Calcination of spotted muscovite at 600° C, when crystallization water begins to be removed from the mica itself, leads to the appearance of a temperature maximum in the loss angle. The appearance of the maximum in the loss angle is associated with the weakening of bonds of polar molecules of crystallization water in the muscovite lattice as a result of thermal action and with the emergence of the possibility of their orientation under the action of an electric field.

The dielectric-loss angle for phlogopite mica is determined by the quantitative content of crystallization water and has a maximum in the range of audio frequencies. In the temperature interval 15–200° C the loss angle has a maximum. With increasing frequency, the maximum of the loss angle shifts toward higher temperatures.

The dielectric losses of phlogopite mica are explained on the basis of ideas concerning the possibility of orientation of polar molecules of crystallization water.

The study of the dielectric properties of muscovite and phlogopite mica yielded practical recommendations on the rational use of mica resources and on expanding the field of application of low-grade mica.

Works on the Crystallization of Ionic Dielectrics

(M. A. Bolshanina, E. P. Volochneva, K. M. Kevroleva, A. M. Kuzmin, M. V. Maksimov, I. Ya. Melik-Gaikazyan; D. D. Saratovkin, P. S. Sasim, S. A. Stroitelev).

The study of the crystallization mechanism showed:

a) An increase in the number of crystallization centers when radioactive radiation acts on the solution.

b) The independence of the growth rate of potassium alum crystals from the viscosity of the solution. This proves that crystal growth is governed by the phenomenon of diffusion, and not by convection.

c) The rate of dissolution of rock-salt crystals is the same for all faces, whereas the rate of growth differs.

The influence of co-dissolved impurities on crystallization from aqueous solutions of ammonium chloride and sodium chloride, potassium iodide, and potassium alum was studied. It was shown that, in order to change the form of crystals, not only the presence of impurities is required, but also their chemical action on the crystallizing substance.

The works are summarized in the monograph Physics of the Solid State, vol. I, published in 1937 and expanded and republished in 1954 under the title Crystals and Crystallization.

The study of dendritic and skeletal forms of crystals at the Department of Physics of TPI showed that impurities chemically interacting with the crystallizing substance change the faceting of dendrites and the decay time of dendrites.

The works on the study of dendritic forms are summarized in the monograph Dendritic Crystallization, published in 1953 and 1957.

At the Department of Crystallography and Mineralogy of TPI, using crystallization from aqueous solutions of magnesium and sodium sulfates in the presence of various additives as an example, a criterion was found for identifying effective impurities that facilitate the growth of technically valuable crystals. This criterion is the indication of a change in solubility. It was shown that co-dissolved substances that increase the solubility of the principal salt of the solution lead to the elimination of defects in complex crystals.

By the planetary method, dense, regularly faceted crystals of magnesium, sodium, and copper sulfates and crystals of potassium chloride, bromide, and iodide are grown. It was shown that the solubility criterion makes it possible to select impurities that facilitate the growth of these crystals.

Mechanical Properties of Ionic Dielectrics

(K. A. Ambroz, A. E. Abramovich, M. A. Bolshanin, N. A. Bolshanina,
N. A. Bessonov, A. A. Vorob’ev, A. A. Vorob’eva, Yu. A. Grdina,
V. Ya. Zlenko, V. D. Kuznetsov, N. F. Kunin, V. P. Kutsepalenko,
A. N. Kislina, V. M. Kudryavtseva, A. A. Kurbatov, E. V. Lavrent’eva,
A. F. Naumov, V. V. Pozdnyakov, N. D. Popov, K. V. Savitskii,
N. A. Sementsov, P. A. Savintsev, A. M. Trubitsyn).

The mechanical properties of ionic crystals have been most fully studied on rock-salt crystals. It has been established that, at the first measurement, the values of the Young’s modulus of rock salt vary within rather wide limits. After several successive hardenings, by means of relaxation or flow of the material, the values of the Young’s modulus up to the proportional limit become constant and, beginning from this limit, gradually decrease. With an increase in the Young’s modulus of the crystals, the yield point and the strength limit of rock salt increase.

The optical yield point of rock salt has a quite definite value after annealing of the crystal. A reduction in the annealing time is accompanied by a lowering of the yield point.

To determine the hardness of brittle bodies, methods of drilling, damped oscillations, mutual grinding, abrasive grinding, and an improved splitting method have been proposed.

It was found that crystals grown from solutions have lower hardness than crystals grown from the melt.

By the grinding method it was found that the hardness of different faces of a rock-salt crystal is different. The octahedron face has the greatest hardness, and the cube face the least.

The frictional force in grinding along different faces of the crystal does not change if the pressing force and the wear area remain constant.

The particles of powders obtained when grinding different faces of rock salt under identical conditions have identical linear dimensions.

The amount of heat released when grinding different faces of rock salt under identical conditions does not depend on the name of the face.

In determining hardness by the scratching method on the cube face of rock salt, hardness rosettes were obtained, showing the dependence of hardness on the direction of motion of the scratching point. The hardness has its greatest value when the point moves parallel to the edge of the cube, and its smallest value when it moves at an angle of 45° to the edge of the cube.

It was found that, in scratching, only a small part of the work is expended on the comminution of the rock-salt crystal, i.e., on increasing the surface energy; the greater part of the work is spent on elastic and plastic deformations, on friction, and is converted into heat.

The Brinell hardness decreases when rock salt is heated. The dependence is linear within the temperature range from room temperature to 400° in the case of the static method of loading the specimen. Under dynamic loading of rock salt, the Brinell hardness decreases nonlinearly with temperature.

Rock salt colored in sodium vapor and by X-rays has an increased proportional limit and yield point; the Young’s modulus changes little in this case.

The hardness, as determined by drilling and mutual grinding of polycrystals of solid solutions of alkali-halide salts, changes with composition along curves having a maximum in the region of intermediate concentrations.

The hardness, as determined by mutual grinding and drilling of single crystals of solid solutions of alkali-halide salts, changes with composition along curves having a minimum in the region of intermediate concentrations.

In eutectic systems of alkali-halide salts, the lowest hardness by mutual grinding occurs for eutectic compositions.

The connection with production of the Tomsk scientists engaged in dielectrics and insulation was many-sided. It included mainly the scientific development of problems of dielectrics of interest to enterprises and other scientific institutions in Moscow, Leningrad, Kharkov, Novosibirsk, Irkutsk, Norilsk, Saratov, Kemerovo, Stalinsk, and others; expert examinations and consultations; lectures at enterprises; personnel training; assistance teams for enterprises; publication of the scientists’ works; and so on.

Some examples of ties with production are given below.

Properties of insulating materials at low temperatures

(Z. F. Bozhko, N. B. Bogdanova, A. A. Vorob'ev, V. A. Vorob'ev,
N. I. Vorob'ev, V. S. Dmitrevsky, K. A. Dukov, E. K. Zavadovskaya,
A. K. Krasin, R. M. Kessenikh, N. A. Prikhodko, N. A. Snegirev,
A. M. Trubitsyn, M. N. Treskina)

The density of transformer oils increases linearly as the temperature is lowered. The viscosity of oils increases exponentially as the temperature is lowered. It has been established that, in some grades of oils, at

At low temperatures, a structure is formed that leads to a considerable increase in the viscosity of oils. Stirring the oil is accompanied by destruction of the structure and a decrease in the viscosity of the oil.

When passing through zero degrees and when the dissolved water freezes out, the electric strength of the oil increases abruptly. A further lowering of temperature is accompanied by an increase in the electric strength of the oil. The electric strength of paper materials impregnated with oil, and the voltage of surface discharges, also increase. The electric strength of casting compounds increases as the temperature is lowered. In the temperature region of cracking, the electric strength of casting compounds decreases abruptly.

The electric strength of porcelain at low temperatures down to 90–150° remains constant.

The electrical conductivity of oil, paper insulation, varnished fabric, and certain other technical dielectrics decreases with a lowering of temperature below zero according to the usual exponential law.

A study of the dielectric losses of vacuum-tight ceramics (of the VK-92 type) showed that the loss angle at a frequency of 1 Mc does not change in the temperature interval \((+20) \div (-70^\circ\mathrm{C})\).

The electric strength of polyethylene, when the temperature is lowered from \(+80^\circ\) to \(-100^\circ\mathrm{C}\), increases from 140 to 260 kV/mm. A smaller temperature dependence is observed for polystyrene. Fluoroplastic 4 almost does not increase its strength when the temperature is lowered. The increase in the electric strength of polyethylene and polystyrene with decreasing temperature agrees with the data of Oakes.

A study of the dielectric losses and permittivity of thermoreactive compounds at a frequency of 1 kc in the temperature interval \(-60 \div +120^\circ\mathrm{C}\) revealed that in the temperature range \(+20 \div 40^\circ\mathrm{C}\) there is a dipole maximum of the loss angle. The dielectric permittivity varies with temperature along a curve characteristic of dipolar dielectrics.

Improvement of the Properties of Transformer Oils by the Addition of Coumarone Resins

(A. A. Vorob’ev, N. P. Prikhod’ko)

Dissolving a small amount of coumarone resins in insulating oils improves their stability in operation and storage, prevents the formation of structural viscosity at low temperature, lowers hygroscopicity, and improves many other service properties of the oils.

Voltage Distribution along Insulation

(V. S. Dmitrevskii)

A method has been developed for measuring the distribution of voltage over the surface of insulation immersed in a liquid. The measurement error does not exceed 10–15 percent of the measured quantity.

As the dielectric permittivity of the solid dielectric increases, the voltage distribution under impulses becomes more uniform.

The field strength at the electrodes is directly proportional to the applied voltage up to the voltage at which surface discharges appear.

Operation of Electrical Apparatus and Machine Insulation at Low Temperatures

(A. A. Vorob’ev, N. I. Vorob’ev, E. K. Zavadovskaya, A. K. Krasin, V. P. Silinskii)

A study of the thermal regime of high-voltage transformers with solidified oil at a temperature of −50°C and their switching on under normal load and under a 160% load showed the absence of dangerous overheating, since the oil being heated quickly begins to perform its function as a heat conductor.

The increase in the viscosity of oils as the temperature is lowered is accompanied by an increase in the travel time of the rod of an oil circuit breaker. At high oil viscosity the travel time of the breaker rod becomes indeterminate, and its motion may stop. In oil circuit breakers intended for operation at low temperatures, it is necessary to use special low-viscosity oils or to provide for their heating.

A study of the frost resistance of the insulation of motors of a single series showed sufficient frost resistance of the insulation. The frost resistance of the insulation of motors of normal design proved to be higher than that of motors of moisture-resistant design.

Electrical Conductivity of Varnish Coatings

Measurements were made of the electrical conductivity of the varnish coatings of tubular arresters located in the open air under Siberian winter conditions as the temperature was lowered to −40°C.

The compressive strength of porcelain, determined on cubes with linear dimensions of 10 mm, has no regular relationship with porosity as measured under a microscope.

Preventive Testing of High-Voltage Insulation in the Kuzbass

(N. D. Bulatov, A. A. Vorob’ev, N. A. Prikhod’ko, R. M. Shevchuk)

For several years before the war, teams of SPTI personnel carried out scheduled preventive tests of high-voltage insulation at enterprises of the Kuzbass and the Tomsk Railway; in connection with this, new instruments and measurement methods were developed—for example, determination of defective high-voltage insulators in service by radio interference, a rod for measuring the distribution of voltage drop along an insulator or string, and others.

Preventive tests made it possible to detect weakened sections of insulation in good time and to prevent accidents. Tomsk scientists contributed to the organization of an overvoltage-protection and insulation-monitoring service in the Kuzbass.

Study of the Aging of High-Voltage Porcelain Insulators and Analysis of Accidents with Insulators in Service

(A. A. Vorob’ev, E. K. Zavadovskaya, N. N. Onchukov, V. N. Titov)

It has been established that microfractures develop over time in insulators in service. The greater the load falling on an insulator, the greater the porosity of the porcelain. With increasing porosity of the porcelain, dielectric losses at high voltage increase and the impulse electric strength of the insulator decreases.

Dependence of the electrical strength of porcelain on its porosity

To determine the porosity of insulator porcelain, a conditional characteristic was introduced: the length occupied by pores over a distance of 1 mm on the polished section.

Comparing the magnitude of the impulse breakdown voltages of insulators with the magnitude of their closed porosity, it was found that, as closed porosity in the insulators increases, their impulse breakdown voltage decreases sharply.

No systematic dependence between breakdown voltage and porosity was obtained when testing porcelain in specimens from 0.5 to 2 mm thick, since large pores that happened to be present significantly reduced the breakdown voltage of such weakened specimens.

Compressive strength of porcelain

The compressive strength of porcelain cubes varied within the range from 4100 to 5000 kg/cm², while the length of pores along the polished section varied from 127 to 200 microns per millimeter; moreover, both small and large values of porosity corresponded to high values of mechanical strength.

Application of impulse discharge for the destruction of rocks, the processing of materials, and the drilling of boreholes

(A. V. Astafurov, A. A. Vorob’ev, E. K. Zavadovskaya, I. I. Kalyatskii).

At the end of the 1940s, experiments were conducted on crushing coal and rocks by means of powerful high-voltage electrical impulses; these confirmed the economic feasibility (high efficiency) and technical possibility of breaking coal in open faces with the aid of powerful spark impulse discharges. It was found that, for purposes of destruction, it is best to use impulses on the order of a microsecond, at which the breakdown voltage of dielectrics at positive polarity is the lowest. The degree of crushing of the rock or coal being broken increases with increasing impulse energy in the discharge, etc.

Methods of combating icing of wires and cables

(N. B. Bogdanova, A. A. Vorob’ev and D. D. Saratovkin)

1. Destruction of ice by means of mechanical oscillations of the wire during vibration.

2. A physicochemical method of bringing ice into a mechanically unstable state, when the ice crust can fall off under its own weight. This is achieved by coating the wires with a grease containing substances that lower the freezing temperature of water.

3. Prevention of icing on wires during their corona discharge, when moisture droplets or crystals, having become charged in the corona sheath, are repelled from the wire and, under certain conditions, may not settle on its surface at all.

Conclusion

Tomsk scientists engaged in the study of current problems in the field of the physics of dielectrics and insulating materials, as well as the practice of their operation.

A varied and comprehensive study was carried out of the physical properties of dielectrics. Macroscopic and microscopic investigation of phenomena during the breakdown of solid dielectrics, electrical conductivity, di-

electrical losses, etc., expanded scientific knowledge about physical processes in insulation and contributed to the development of practically useful insulating materials.

In the works of Tomsk researchers, the systematization of the properties of dielectrics on the basis of the concept of the forces of interaction between structural elements received factual substantiation. The scientific direction concerning the relationship between the properties of dielectrics, their composition, and structure, which guided Tomsk scientists, is finding adherents in the USSR and abroad.