Semiconductor Research in the Soviet Union

A. F. Ioffe

In 1927 the first solid alternating-current rectifiers made of copper oxide appeared. A short time later it was found that similarly prepared plates of copper oxide produce an electric current when illuminated. These facts drew the attention of physicists to the distinctive properties of semiconductors, which are the underlying cause both of these phenomena and of a number of others that emerged in the course of research.

At the same time, the study of semiconductors began in the Soviet Union. The initiators of these investigations were K. D. Sinelnikov and I. V. Kurchatov, and in Ukraine—A. G. Goldman. Systematic work on the comprehensive study of semiconductors was undertaken at the very end of the 1920s at the Leningrad Physico-Technical Institute and in Kiev at the Institute of Physics of the Academy of Sciences of the Ukrainian SSR. From the very beginning, theory and practice, physical laboratories and factories, were closely linked. This integration was carried out through seven congresses held in Leningrad, Kiev, and Odessa. As a result of the investigations conducted, both the physical understanding of processes in semiconductors and the production of solid rectifiers—first from copper oxide, later from selenium—were improved. B. V. Kurchatov and Yu. A. Dunaev developed a new type of powerful rectifier made of cuprous sulfide, which rectified currents per unit area 100 times stronger than oxide and selenium rectifiers. True, the energy losses in these rectifiers were somewhat higher; in addition, difficulties were encountered when connecting them in series in high-voltage units.

Photocells underwent considerable development. Yu. P. Maslakovets and B. M. Kolomiets in Leningrad, and T. Geikhman in Kiev, created new types of cells based on thallium sulfide and silver sulfide. Per lumen of light from an electric lamp these photocells yielded from 3,000 to 10,000 microamperes, whereas in photocells made of copper oxide and selenium the currents did not exceed 400 microamperes per lumen. In direct sunlight it was possible to obtain electric power with an efficiency of up to 1%, instead of 0.05% for the foreign devices then in existence. Thus, for the first time, the question of solar generators of electric power arose.

Alongside these practical results, and in close connection with them, research work also developed. The main problem was the properties of the boundary blocking layer that caused rectification and the photoelectromotive force.

The most important results of the prewar work of Soviet physicists in this direction were: V. P. Zhuze’s investigations of an artificial blocking layer made of sprayed silicon dioxide (quartz) on copper oxide; they made it possible to establish the optimal thickness of the blocking layer. Zhuze’s experiments have much in common with the work of O. V. Losev, who observed the influence

of a quartz layer on carborundum. Losev also has indisputable merits in the application of semiconductors for detecting and generating high-frequency radio waves.

A number of attempts were made by B. I. Davydov, D. I. Blokhintsev, and S. I. Pekar to create a theory of the blocking layer. Among them were theories that have retained their significance to the present day, and even the beginnings of ideas whose importance became clear only much later.

In parallel with the theory, experimental work was also carried out. A. V. Ioffe confirmed, on the basis of extensive experimental material, the validity of the diffusion theory of the blocking layer and at the same time its inadequacy for quantitative conclusions. It turned out that although the properties of the blocking layer at the boundary with a metal do lead to rectification, it is, however, far from being of the same order as that observed in technical rectifiers. A. V. Ioffe showed that the required scale of rectification is obtained only at the boundary between two semiconductors with different types of conductivity, at the boundary between $p$- and $n$-type semiconductors. This conclusion was later confirmed by examples of rectifiers made of selenium and cuprous oxide and opened the way to their improvement.

It was precisely these boundary conditions, transferred during the war by American physicists into the interior of a single crystal of germanium and silicon, that marked the beginning of a new stage in semiconductor radio engineering.

A major achievement of Soviet physicists was the discovery by I. K. Kikoin and M. M. Noskov of the photomagnetic effect and the correct explanation of it given by I. K. Kikoin. The study of the photomagnetic effect still arouses the close interest of physicists today.

The properties of excited states of a crystal—excitons, discovered theoretically by Ya. I. Frenkel in 1931 and subsequently confirmed and put to numerous uses in postwar work by Soviet physicists—are of great significance and attract growing interest. Thus, V. P. Zhuze and S. M. Ryvkin, as well as V. E. Lashkarev, showed their role as the primary act of the photoelectric effect in semiconductors. E. F. Gross and N. A. Karryev succeeded in detecting excitons in the absorption spectra of cuprous oxide; later E. F. Gross and his co-workers generalized and refined these results, creating a new direction in the optics of solids. Finally, E. D. Devyatkova and A. F. Ioffe were able to substantiate the participation of excitons in the transfer of thermal energy, which proves the ability of excitons to diffuse in the fundamental lattice of the crystal. A. I. Anselm developed the theory of their mobility.

No less important are the polarons introduced by S. I. Pekar, following an idea of Academician L. D. Landau—free electrons in a semiconductor surrounded by an atmosphere of dielectric polarization. The limits of applicability of Pekar’s theory and the criteria for the existence of polarons are still being discussed today, but their role in the electrical properties of ionic semiconductors is beyond doubt. The development of Pekar’s theory led him to a quantitative theory of color centers in alkali-halide crystals. The predictions of Pekar’s theory are confirmed experimentally. Great successes in the study of the related photographic process have been achieved by E. A. Kirillov in Odessa. New facts were discovered by M. A. Savostyanova and S. A. Artsybyshev.

An extensive new field of knowledge was created by the work of P. P. Kobeko and Academician I. V. Kurchatov on seignette-electricity, or, as it is called in the West, ferroelectricity.

These investigations acquired great practical importance when B. M. Vul and I. M. Goldman discovered the seignette-electric properties of barium titanate. Barium titanate became not only a material for high-

frequency capacitors, but also opened up new possibilities for sound reproduction. The possibilities of this material, discovered by B. M. Vul and A. V. Rzhanov, as a piezoelectric and a reversible converter of mechanical and electrostatic energy are still far from exhausted. A series of further ferroelectrics with various properties was prepared and studied by G. A. Smolenskii.

The doctrine of semiconductors was significantly enriched by A. R. Regel’s studies of liquid semiconductors, which showed that semiconductor properties are not connected with the strict periodicity of the structure of crystals and that what is decisive is short-range order, i.e., the chemical bonds of an atom with its nearest neighbors. Regel also studied semiconductor processes occurring in the liquid state. N. A. Goryunova and B. T. Kolomiets expanded the range of semiconductor materials by creating a series of glassy semiconductors.

A significant contribution to the doctrine of semiconductors was made by the studies of Academician V. E. Lashkarev and his collaborators on ambipolar diffusion of charges, on the photoelectric effect, and on barrier layers in semiconductors.

Much that was new was provided by investigations of the thermal conductivity of semiconductors carried out by A. F. Ioffe, A. V. Ioffe, E. D. Devyatkova, and P. V. Gul’tyaev. A connection was established with atomic weight and the character of the chemical bond, the share of electrons in thermal conductivity, and the role of bipolar diffusion. The theory of the latter phenomenon had been developed as early as 1940 by B. I. Davydov and I. M. Shmushkevich. The study of the influence of foreign impurities on thermal conductivity and on the mobility of electrons introduced essential features into the understanding of the scattering processes of phonons and electrons. A detailed study of impurities in lead telluride led T. L. Koval’chik and Yu. P. Maslakovets to a deeper understanding of their interaction with the crystal lattice.

Studies by L. S. Stil’bans and S. V. Airapetyants introduced new features into the picture of the motion of free charges in a semiconductor. It turned out that a disturbance of periodicity in the arrangement of positive ions in the crystal lattice sharply reduces the mobility of negative electrons, without having any noticeable effect on the motion of positive holes; and, conversely, a disturbance of order for negative ions affects only the mobility of holes, without changing the mobility of electrons.

The motion of phonons, however, is disturbed to an equal degree by a distortion of periodicity or by the introduction of foreign ions of either sign. A. V. Ioffe and A. F. Ioffe showed that the relative decrease in thermal conductivity \(\frac{\Delta \chi}{\chi}\), caused by the introduction of \(N\) foreign atoms into a medium of \(N_0\) lattice atoms, is determined by the quantity

\[ \frac{\Delta \chi}{\chi} = - \frac{N}{N_0} \frac{\lambda_0}{a} S, \]

where \(\lambda_0\) is the mean free path of phonons in the pure substance, \(a\) is the lattice constant, and \(S\) is the scattering cross section expressed in areas of an elementary cell. It also turned out that replacement of half of all atoms of a given type by others reduces the mean free path to atomic dimensions, as in amorphous materials.

Significant results were obtained by S. G. Kalashnikov and B. I. Boltaks on problems of the diffusion, solubility, and recombination of impurities in germanium, and by B. M. Vul in his investigations of the mechanism of breakdown of barrier layers.

An outstanding event in Soviet physics was the discovery in 1942 by E. K. Zavoiskii of paramagnetic resonance, theoretically predicted as early as 1923 by Ya. G. Dorfman. Since then, paramagnetic resonance

became one of the most essential means for the study of matter. The cyclotron resonance observed later had also been predicted in advance by Ya. G. Dorfman as a diamagnetic resonance.

Surface electron levels, introduced by Academician I. E. Tamm and S. P. Shubin in the consideration of photoelectric phenomena, acquired equally great importance.

A new direction in the theory of the solid state is represented by the theory developed by S. V. Vonsovsky, which takes into account the interaction of electrons as a single indivisible system.

Another very promising direction in theoretical physics is represented by the work of A. G. Samoilovich and his collaborators, proceeding from the picture of chemical bonds in a crystal, and from the study of the magnetic and thermoelectric properties of semiconductors.

F. F. Vol’kenshtein and S. Z. Roginsky showed that the phenomenon of catalysis is in many cases due to electronic processes in the surface layer of the catalyst, and developed an electronic theory of catalysis that aroused wide interest.

An independent field of research by Soviet physicists is constituted by the thermoelectric properties of semiconductors and the practical results based on them.

These studies began long before the Patriotic War and by 1940 had already led to the possibility of obtaining currents of up to 100 amperes and a coefficient of efficiency for the conversion of thermal energy into electrical energy of up to 3%, which was ten times greater than the efficiency of metallic thermoelements. The theory of thermoelectric generators, refrigerators, and heaters was published by A. F. Ioffe in 1950. Since then a group of physicists (Yu. P. Maslakovets, A. N. Voronin, L. S. Stil’bans, and others), under the direction of A. F. Ioffe, has developed both the theory of semiconductor thermoelements and their various applications. In seven years, significant successes have been achieved and numerous thermoelectric devices have been developed, some of which have already gone into production.

There are grounds to believe that at present the level of the results achieved in this question exceeds what has been created abroad.

As for the most important radio-engineering applications of semiconductors in our time, the initiative here belonged to the USA. Soviet physicists succeeded in mastering the technology of germanium radio devices and, together with industry, in organizing their production. This is the merit of physicists at the Physico-Technical Institute in Leningrad, of the scientific schools of B. M. Vul and S. G. Kalashnikov in Moscow, and of V. E. Lashkarev in Kiev. Further development passed into the hands of branch institutes, with which physicists maintain constant contact. It must be acknowledged, however, that in the production of semiconductor radio devices Soviet industry has still not reached the level of the USA.

The situation is considerably better with the so-called thermistors, or thermal resistances. Not only in laboratories but also in production there are instruments developed by B. T. Kolomiets jointly with Sheftelem, which are not inferior in their sensitivity and in the stability of their readings to the best foreign models. Alongside them, the production of photoresistances has been organized; at a number of factories these are used for purposes of production automation and for the prevention of accidents and breakdowns.

V. G. Karmanov developed specimens of microthermistors up to fractions of a millimeter in size, with low heat capacity and, in connection with this, a very short settling period. Microthermistors have been used with great success

...are used to measure the temperature of plant leaves and the skin of animals. They make it possible to record the course of temperature over time with a lag of less than 1 sec. With their aid, periodic processes in the life of plants and new features in the physiology of respiration and transpiration have been discovered.

A. F. Chudnovskii and M. A. Kaganov, on the basis of semiconductors, created a large series of measuring instruments for agricultural purposes.

E. A. Kolenko and A. N. Voronin applied semiconductor thermoelements to the solution of many problems in biology, meteorology, vacuum production, to the stabilization of radio equipment, to increasing the sensitivity of radiant-energy receivers, to supplying radio equipment with electric power, to electric lighting, and to many other purposes.

In this brief survey I, of course, could not even touch upon many hundreds of scientific works carried out in the USSR on various questions of the theory of semiconductors, experimental investigations of their properties, and those applications which have not yet received practical realization. Among the latter are such original and promising problems as semiconductor catalysis in chemical production, powerful rectifiers for hundreds and thousands of amperes, solar generators of electric power based on semiconductor thermoelectric batteries, and thermoelectric refrigerators.

Considering as a whole the scientific work done in the Soviet Union on the problem of semiconductors, one may note that it developed at a rapid pace, like all aspects of life in our homeland, which has built socialism and is advancing with rapid strides toward the highest form of human society—communism.

Having been among the first to take up the investigation of semiconductors, Soviet physicists, right up to the beginning of the Great Patriotic War, stood in the front ranks of world science alongside Germany. During this period there were elucidated: the mechanism of rectification, the structure of the blocking layer, photoelectric and thermoelectric phenomena in semiconductors, the influence of strong electric fields; new types of rectifiers and photocells with high indices for that time were developed, and the first semiconductor thermoelements were created.

The Great Patriotic War of the Soviet Union against the Hitlerite aggressors and its consequences halted the development of semiconductor research for almost ten years. During these years the principal center of research and applications of semiconductors shifted to the USA. The problem of germanium and silicon as semiconductor materials, in which it proved possible to realize \(p\)—\(n\) junctions, came to the fore. With their aid devices for radio engineering were created—detectors, amplifiers, and generators of high-frequency oscillations. They required an unprecedented chemical purity of the starting materials, new methods of technology, and new research techniques. Soviet physicists joined in the solution of these problems considerably later, when germanium devices in the USA had already entered radio-engineering practice. Nevertheless, with the united participation of physicists and engineers, all difficulties were overcome. The production of germanium of the required qualities was established thanks chiefly to N. P. Sazhin and a number of physicists and chemists of academic institutes. We are still lagging behind in the production of ultrapure silicon and products made from it.

To assess the work that has been done, it should first of all be pointed out what significance it had for the development of solid-state physics and for

of technical progress, and the prospects opened up by our work for the near future.

From these most important standpoints one may note with satisfaction the close connection between physical research and the needs of production, and its positive role in the technical progress of our industry. The significant role of theory in the study of semiconductors should be pointed out: more than three dozen representatives of theoretical physics are actively participating in the common work. Among the many hundreds of published scientific works devoted to semiconductors, a prominent place is occupied by investigations of the fundamental problems of the doctrine of the solid state. Soviet works have introduced new features into this doctrine and have refuted some views that had already become established. The most important problems defining the physical theory of semiconductors, if not yet solved, have been posed by Soviet physicists and are being successfully resolved by them. Such technical problems are also being successfully carried out as the direct conversion of thermal and radiant energy into electrical energy, the production of cold without complex machines, and new effective means of catalysis in chemical production.

The problem of semiconductors arose in the decade of the Great October Socialist Revolution and developed under the conditions of socialist construction and the industrialization of the country. Having originated in one or two institutes, the study of semiconductors now embraces dozens of research institutes: in Leningrad, Moscow, Kiev, Odessa, Minsk, Vilnius, Saratov, Makhachkala and Baku, in Lviv, Rostov, Kazan, Gorky, Tashkent and other cities.

Several specialized plants are engaged in the production of semiconductor materials and products made from them. The Academy of Sciences of the USSR and the Leningrad Council of the National Economy have adopted decisions on the comprehensive development of semiconductor work.

All this may be regarded as favorable prerequisites for further progress in this field—progress that will deepen our understanding of natural phenomena and strengthen our mastery over them. The successful development of our efforts must lead to revolutionary advances in technology—to advances that can be compared only with the prospects of nuclear physics.