Review of Papers Presented at the Meetings of the Section “Photoelectric Phenomena in Semiconductors”

S. M. Ryvkin

The study of photoelectric and optical properties occupies an important place in the overall complex of investigations of the properties of semiconductors. This determined the interest in the work of the photoelectric section, which became evident during the days of the conference. More than thirty papers were submitted for presentation in the section. Owing to the overloaded program, only some of them could be heard at the meetings. The meetings were usually attended by a large number of participants.

Before proceeding to a review of the papers delivered or submitted for presentation in the section, we shall try briefly to consider the main trends currently manifest in the development of photoelectric studies of semiconductors.

An important distinguishing feature of work in the postwar period is the transition to carrying out a comprehensive experimental study of photoconductivity, including a separate investigation of the phenomenological parameters that determine photoconductivity (such as “quantum yield,” “lifetime,” etc.). In other words, the study of stationary photoconductivity, characteristic of earlier work, is increasingly being replaced by the study of the kinetics of photoconductivity.

It is precisely this circumstance that has made it possible in a number of cases to discover and interpret new and interesting regularities of photoconductivity.

Another, no less important, feature is connected with the broadening of studies beyond the circle of purely photoelectric phenomena, owing to the use, in addition to light, of other radiations.

Such broadening is justified not only by the demands of practice, but also by the deep analogy between the phenomena of ionization arising under the action of different radiations. It can be shown that, from the point of view of their behavior in the crystal lattice, current carriers released by light and by other radiations differ only slightly. This is explained by the following considerations.

Processes that arise during ionization in semiconductors may be divided into two categories: processes of ionization proper, i.e., the formation of free current carriers, and processes of motion and recombination of the released carriers. The former are determined by the interaction of the ionizing radiation with the substance and are characterized by constants of the type of the absorption (or “attenuation”) coefficient and the so-called “yield”; the latter depend on the interaction of the “nonequilibrium” current carriers created during ionization with the substance and are characterized by the relaxation time (mobility) and lifetime.

If the character of the processes of ionization proper is specific for each type of ionizing radiation, then the subsequent processes of motion and recombination of carriers in most cases practically do not depend on the radiation by which these carriers have been released. This is explained by the very strong interaction of the current carriers with the lattice of the substance. As a result of this interaction, current carriers that arise in regions as a result of ionization with energies greater than the mean energy of thermal motion and, generally speaking, different for different radiations, in very short times (in comparison with the lifetime) are “cooled” as a result of collisions with the lattice and become indistinguishable from “equilibrium” current carriers released by thermal ionization.

Thereafter, i.e., during the main part of the time of their free existence in the band, such carriers possess normal values of the mobility and recombination coefficient and do not exhibit any specific features connected with the character of the radiation that produced the ionization. This often makes it possible to use practically identical methods of investigation and interpretation of the results in studying conductivity induced as a result of the action of radiations with particle or quantum energies differing by millions of times. In this connection, the comparatively narrow field of investigation of the photoelectric properties of semiconductors is at the present time increasingly turning into the broad field of investigation of internal ionization in semiconductors under the action of diverse radiations. However, even this broad field can no longer fully encompass the entire aggregate of investigations being carried out that are related in the methods of experimentation and interpretation of results and, consequently, that constitute in essence a single scientific field.

Indeed, whereas until recent years all methods of producing nonequilibrium carriers in a semiconductor (if cases of very large electric fields are excluded) reduced to internal ionization, in recent times the possibility has been clearly recognized and widely used of producing nonequilibrium carriers by means of injection of “minority”* nonequilibrium carriers with the aid of a rectifying

* “Minority” carriers are those whose sign is opposite to the sign of the semiconductor’s conductivity (for example, holes in an electronic semiconductor).

contact (especially effective at the contact of two semiconductors with different signs of conductivity).

In atomic semiconductors with high mobility and, consequently, a long diffusion-mixing length of the carriers, this makes it possible to increase the concentration of carriers in a sufficiently large volume of the semiconductor near the injecting contact*).

The nature of the behavior of the injected current carriers and of the carriers released during ionization is the same (the same considerations concerning strong interaction with the lattice, discussed above, are applicable here). Thus injection appears as a new variety of method: the creation in a semiconductor of nonequilibrium current carriers.

Taking this new possibility into account, an even broader field is in fact taking shape—the field of investigating nonstationary electronic processes in semiconductors, or, in other words, investigating the behavior of nonequilibrium current carriers in semiconductors.

The expansion of the field of photoelectric investigations through the use of various radiations and injection, and the use of methods for the comprehensive study of induced conductivity, have extraordinarily enriched experimental possibilities and have led to the obtaining of new results important from both the scientific and the applied points of view. In this connection it is necessary, first of all, to mention the broad development of investigations of the behavior of “minority” current carriers, which determine the operation of the most important semiconductor devices (diodes, triodes, photodiodes, etc.).

In contrast to the “majority” carriers, the “minority” ones have the feature that their diffusion or drift in an electric field does not lead to the appearance of volume charges in the semiconductor. This is due to the fact that the charge of minority carriers, wherever they may appear, is rapidly compensated by the corresponding redistribution of the charge of the majority carriers. As a result, the diffusion of nonequilibrium minority carriers occurs as the diffusion of neutral particles, i.e., it propagates over very large distances determined by the mean diffusion-mixing length**) (whereas the diffusion of majority carriers propagates

*) The use of an appropriately directed electric field makes it possible to increase to an even greater extent the volume with an increased concentration of carriers by means of their drift.

**) Let us note that as early as 1935, in a work devoted to photoelectromotive forces in cuprite crystals, A. F. and A. V. Ioffe emphasized the need to take into account, in explaining the results they obtained, large values for the mixing length of photoelectrons.

In 1948 V. E. Lashkarev, analyzing the results of the study of photo-emf and longitudinal photoconductivity, for the first time pointed out the principal features of the diffusion and drift of minority carriers. At present the specific properties of minority carriers are widely investigated and used in explaining the mechanism of action of semiconductor devices.

in semiconductors over considerably smaller distances, determined by the screening length).

The behavior of minority carriers in an electric field is also distinctive. On the one hand, the electric field acts on them as on charged particles, i.e., it causes their drift. On the other hand, the rapidly established redistribution of the concentration of majority carriers ensures neutrality at any place where minority carriers have appeared as a result of drift. This circumstance, as well as the possibility of easily injecting and detecting minority carriers, led to the use of the most direct method for measuring the mobility of current carriers from the drift time between two points of the specimen (alongside the long-known method based on measuring conductivity and the steady Hall effect).

From the experimental point of view, the simplicity of determining the concentration of minority carriers by measuring the blocking current through a rectifying contact placed in the region where it is desired to determine this concentration is of great importance.

The emergence of new possibilities for experimentation with nonequilibrium minority carriers led, in particular, to an expansion of the study of recombination processes both in the bulk and on the surface of semiconductors.

It should be noted that the existing theory of recombination still does not make it possible to predict the coefficients characterizing the probability of an elementary recombination event, although interesting attempts to develop the theory in this direction are being made (E. I. Adirovich, S. I. Pekar). On the other hand, the phenomenological theory of recombination has by now achieved serious successes. Here, under the assumption that recombination proceeds through so-called “recombination centers,” it has been possible to establish a connection between the character of the filling of energy states by electrons and the intensity of recombination (Shockley and Read). This makes it possible, if the energy structure of the electron spectrum in a given substance is known, to determine experimentally, on the basis of measurements of the mean lifetime, the true recombination cross section at centers of various nature.

Of considerable interest is the study of recombination, begun already by O. V. Losev, by means of investigating the spectra of recombination luminescence. Experimental work in these directions has been developed rather extensively for germanium and silicon. But there is no doubt that, as the energy spectrum and basic properties of certain other semiconductors are established with the same completeness as has been done for germanium, experimental study of the true recombination cross sections will become possible in other substances as well.

A special place in the study of semiconductors is occupied by investigations connected with the exciton hypothesis. This line of work, which has

great fundamental importance, has developed over recent years in connection with the study of the photoelectric and optical properties of a number of semiconductors.

A comprehensive study of the photoconductivity of cuprous oxide and cadmium sulfide (V. P. Zhuse and S. M. Ryvkin, V. E. Lashkarev and G. A. Fedorus) led to the conclusion that impurities play an essential role in the photoelectric processes in these semiconductors. However, this assumption was in contradiction with the circumstance that the study of photoelectric properties was carried out using light corresponding to the intrinsic absorption band, determined not by impurities but by the basic crystal lattice. The contradiction between the “impurity” character of the photoeffect and the “intrinsic” character of absorption led to the hypothesis of an essential role of energy-migration processes, by means of which the energy absorbed by the crystal lattice is transferred to impurity centers. It was natural to suppose that the carriers of the migrating energy are excited states of the lattice—excitons, the concept of which was introduced by Ya. I. Frenkel. Thus arose ideas about the exciton mechanism of the internal photoeffect. The exciton mechanism is also used to explain certain regularities of the external photoeffect (Apker and Taft). Subsequently, searches for exciton absorption in cuprous oxide and other semiconductors led Gross and his collaborators to the discovery of extremely interesting features in the absorption spectra of these substances, consisting in the existence, near the edge of the intrinsic band, of narrow absorption lines whose frequencies in a number of cases fit definite series laws.

Although all these interesting facts cannot yet serve as unambiguous proof of the determining role of excitons in the phenomena of the photoeffect and absorption, it is nevertheless beyond doubt that the introduction of exciton concepts leads to their most natural explanation. The development of experimental investigations of the photoelectric and optical properties of semiconductors, with the aim of refining and developing the exciton hypothesis, has already led to the discovery of many new and interesting facts. In this sense, the role of exciton concepts has proved very fruitful.

Alongside the more profound study of the mechanism of internal ionization and the behavior of nonequilibrium current carriers, the last decade has been characterized by intensified searches for new semiconductor substances sensitive to light and ionizing radiation. Among the rather large number of semiconductors discovered and presently being studied that possess photoconductivity, the following three groups of substances have acquired the greatest interest from both the practical and theoretical points of view:

1) the so-called substances of the PbS group (PbS, PbSe, PbTe);
2) substances of the CdS group (CdS, CdSe, CdTe);

3) substances with a pronounced covalent type of bond, crystallizing in diamond or diamond-like structures (Ge, Si, \(A^{III}B^{V}\)*) etc.).

The practical significance of photoconductors of the PbS group is determined by their very high sensitivity in the visible and especially in the infrared (i.r.) regions of the spectrum. Lead sulfide photoresistors are manufactured by industry in large quantities and are used in automation, infrared technology, sound cinema, etc.

The mechanism of photoconductivity in substances of the PbS group is distinguished by certain special features, which has led to the appearance of a new explanation of their photoelectric properties on the basis of the so-called “barrier” theories of photoconductivity, in principle different from the classical “concentration” theory. From this point of view, the investigation of photoconductors of the PbS group is also of considerable theoretical interest.

Cadmium sulfide, like lead sulfide, began to be intensively investigated in the postwar years (after the development of an effective method for obtaining single-crystal specimens of this material). At the same time, a whole range of peculiar photoelectric and optical phenomena was immediately discovered in CdS, arousing great interest. The sensitivity of cadmium sulfide to the ionizing action of radiation is very high, and it should be noted that CdS possesses a remarkable “universality,” changing its conductivity under the action of the most varied radiations: visible light, X-rays, \(\gamma\)-rays, and fast charged particles. This has already made possible the practical use of this substance as photoresistors (single- and polycrystalline), detectors in dosimetric apparatus, crystal counters, etc.

Investigations of substances crystallizing in diamond and diamond-like structures, and above all of germanium and silicon, have determined the principal success of semiconductor physics achieved in recent years.

The intensity of the investigation of these semiconductors, determined mainly by their practical importance, led in a short time to the obtaining of a large amount of new and interesting data. The success achieved was to a considerable extent connected with the development of effective methods of purification and control of the composition of these materials. No smaller role was played by the circumstance that the conclusions of the existing theory of semiconductors and, in particular, of the “band” theory and the theory of mobility proved fully applicable to germanium and silicon (whereas for the previously studied semiconductors with a strongly pronounced ionic type of bond the theory becomes considerably more complicated and the existing “band” theory and theory of mobility are apparently unsuitable). All this determined the obtain—

*) \(A^{III}B^{V}\) is a general conventional notation for compounds of elements of the third and fifth groups of the Mendeleev table.

tion on germanium and silicon, yielding reproducible and comparatively easily interpreted results.

The regularities characterizing the photoelectric and optical properties of germanium and silicon are also relatively simple. Thus, for example, one should note the agreement, for germanium, of the values of the forbidden-band width and of the ionization energy of impurity centers, as determined from photoelectric, optical, and electrical measurements, the quantum yield being equal to unity, etc.

Important results have been obtained in the study of the optical properties of germanium and silicon. Here one must note the experimental proof of the existence, alongside “direct” phototransitions between bands (without a change in the wave vector \(k\)), of so-called “indirect” transitions, which occur with the simultaneous emission or absorption of a phonon. The absorption by “free” current carriers is being widely studied; in this absorption, phototransitions take place within a single band. The fulfillment of the conservation laws in such phototransitions is ensured by interaction with the crystal lattice. Absorption by current carriers has a considerable magnitude and is readily observed in the region of relative transparency of germanium, outside its intrinsic absorption band. Since the concentration of current carriers in germanium can easily be varied over considerable limits by means of injection of minority carriers, there appears the possibility of practical use of “injection” absorption for modulating the intensity of light beams by means of electrical signals.

It is especially necessary to emphasize the possibility of using germanium, silicon, and certain other substances for constructing photocells. This question is of great practical importance, and we shall dwell on it in more detail.

Semiconductor barrier-layer photocells have long been known, and some of their types (selenium, silver sulfide, etc.) find practical application.

In a barrier-layer photocell there occurs a direct conversion of light energy into electrical energy, and therefore it would be natural to use them as energy converters. However, the efficiency of such conversion in the “old” types of photocells is too low. This has led to their being used, if at all, not as energy converters but as signal converters (light signals into electrical ones). However, even this use of photocells is hampered by their considerable inertia, relatively low sensitivity, and certain other shortcomings.

If the occurrence of barrier-layer photo-emf requires only the presence of a blocking layer bordering on a photosensitive substance, then in order to obtain a perfect barrier-layer converter it is necessary to satisfy certain additional conditions. These conditions are realized in new semiconductor photocells—germanium and silicon ones—in which the photo-emf arises upon illumination of a region near

from the electron-hole transition. As a result of ionization produced by such illumination, in the corresponding regions there appear nonequilibrium minority carriers, which are ejected by the contact field of the \(n\)-\(p\) junction into the neighboring semiconductor.

The potential difference of the semiconductors forming the \(n\)-\(p\) junction, arising as a result of this, leads to the appearance of a current of majority carriers through the \(n\)-\(p\) junction, compensating the current of nonequilibrium minority carriers. To the stationary state there corresponds a certain potential difference at the \(n\)-\(p\) junction—the photo-emf. Analysis shows that both the magnitude of the stationary photo-emf and the efficiency are the greater, the smaller the saturation current of the junction (the current in the blocking direction), i.e. the smaller the concentration of equilibrium minority carriers.

When using photoelements as energy converters it is also necessary that the resistance of the bulk of the semiconductors connected in series with the \(n\)-\(p\) junction be as small as possible. Consequently, the concentration of majority carriers in them must be large.

A small concentration of equilibrium minority carriers and a large concentration of majority carriers correspond to a close position of the Fermi level to the edge of the “majority” band.

Donor and acceptor levels in germanium, silicon, and compounds of elements of the 3rd and 5th groups (\(\mathrm{A}^{\mathrm{III}}\mathrm{B}^{\mathrm{V}}\)) are located very close to the band edge, and, consequently, with a sufficiently large number of them the Fermi level will also be close to this edge. Thus, from the indicated point of view, such substances as germanium, silicon, etc., should be especially advantageous for manufacturing photoelements.

The efficiency of photoelements is also determined by that fraction of nonequilibrium minority carriers, liberated by light, which manages during its lifetime to diffuse to the \(n\)-\(p\) junction. Consequently, for manufacturing photoelements one should use a material with the greatest possible diffusion length. And from this point of view, semiconductors of the germanium type, where the diffusion length reaches millimeters, prove to be the most suitable. Let us note, finally, that, since in practice the question can be only of converting solar energy, obtaining significant efficiencies is possible only when the width of the semiconductor forbidden band is matched with the characteristics of the spectral distribution of energy in the solar spectrum. Analysis shows that in the case of silicon such matching proves to be close to optimal. Thus, in silicon, essentially all the basic conditions for obtaining high efficiencies are combined. Silicon solar energy converters constructed by the present time possess efficiencies of the order of \(10\%\).

With the appearance of highly efficient germanium and silicon photoelements, apparently, there arises the interesting possibility of converting the nuclear energy of radioactive decay directly

into electrical energy. Indeed, by using as the ionizing radiation, instead of light, the beta radiation of some long-lived isotope (for example, \(Sr^{90}\), whose half-life is \(\sim 20\) years), one could obtain a voltage source with a long service life.

In this connection, for the most effective conversion of energy possible, it is necessary to take into account a distinctive feature of photocells used as energy converters, namely the increase in efficiency with increasing intensity of the radiation being converted. This leads to the necessity of using the radioactive isotope in a sufficiently concentrated form.

At present it has been shown that the prospects for converting the decay energy of \(Sr^{90}\) are apparently slight, because of the formation, under the action of high-energy \(\beta\)-particles, of defects in the germanium lattice*). However, the practical use of the described method of energy conversion may become possible for isotopes with a \(\beta\)-spectrum boundary not exceeding the threshold for defect formation.

Along with energy conversion, new photocells made of germanium (photodiodes) can be used successfully as converters of light signals into electrical signals.

The large diffusion length in germanium and certain other favorable features make it possible to obtain a current sensitivity of germanium photodiodes an order of magnitude greater (\(30000\ \mu\mathrm{A}/\mathrm{lm}\)) than that of the best barrier-layer photocells of the old types. However, the chief advantage of germanium photodiodes used as signal converters is determined by the possibility of their operation in a mode not characteristic of the old types of photocells, namely with an external voltage applied in the reverse direction.

In such a “photodiode” mode it becomes possible to obtain photocurrents of the same magnitude as with a short-circuited photocell, but with a large load resistance present in the circuit. In this case the sensitivity “in volts per lumen,” which plays the principal role in most cases of photocell use, increases immeasurably. In germanium photodiodes, high sensitivity is accompanied by low inertia (time constant \(\sim 10\ \mu\mathrm{sec}\)). All this determines the possibility of broad use of photodiodes for signal conversion in many fields of technology (for example, in sound motion pictures, in computing mechanisms, automation, etc.).

At present the problem arises of using new materials for the construction of photodiodes. The use of semiconductors with a small band gap should lead to the creation of

*) In addition to ionization, \(\beta\)-particles with energies above \(\sim 0.6\ \mathrm{MeV}\) are capable of knocking Ge atoms out of their positions at the lattice sites, which leads to damage of the device.

photodiodes sensitive to the longer-wavelength part of the spectrum. The development of photodiodes with small dark currents and with maximum sensitivity in the visible and ultraviolet regions of the spectrum requires a transition from semiconductors to dielectrics. In connection with this, the problem of introducing impurities into dielectrics and obtaining $n$-$p$ junctions in them becomes important.

From the brief survey given above it is evident that investigations of photoelectric phenomena in semiconductors (or, more precisely, investigations of nonstationary electronic processes in them) are being carried out very intensively and are marked by the same general upsurge that characterizes the development of all semiconductor physics in recent years.

To some extent this was also reflected in the character of the work of the photoelectric section of the conference, the number of papers presented there considerably exceeding the “physical” possibilities for hearing and discussing them. The content of many papers was devoted to the most urgent and important problems.

Below we attempt briefly to set forth the main content of the majority of the works reported or submitted for presentation at the photoelectric section*), without, of course, claiming the completeness and thoroughness incompatible with the limited scope of the present article.

In general, all the papers presented may, in their content, be assigned to the following two broad divisions:

1. Photoconductivity and optical properties.
2. Photoelectromotive forces.

In addition, a number of papers were devoted to the external photoeffect from semiconductors and to certain methodological questions. In presenting the content of the papers we shall adhere to the order determined by this, very conventional, classification.

In a large number of the works heard, concepts connected with excitons were used in interpreting the results. Two papers (E. F. Gross—A. D. Kaplyansky and V. E. Lashkarev—Yu. I. Karkhanin) were directly devoted to the investigation of their properties.

E. F. Gross and A. D. Kaplyansky reported new results of an investigation of the structure of the absorption spectra of semiconductors. It is known that a detailed study of these spectra has in recent years led E. F. Gross and his collaborators to the discovery of line structure in crystals of $\mathrm{Cu_2O}$, CdS, $\mathrm{PbJ_2}$, $\mathrm{HgJ_2}$,

*) Unfortunately, the interesting communications of Dr. Tauz and Prof. Rompe could not be reported because the author did not have the texts of the papers at his disposal.

CdJ\(_2\). At the same time it was found that in a large number of other crystals such a structure is absent.

The authors of the report set themselves the goal of clarifying the connection between the presence or absence of a line structure in the spectra and the structure of the crystal lattice. To this end, the absorption spectra of two modifications—red (tetragonal) and yellow (rhombic) mercuric iodide—were studied in detail. These spectra proved to be sharply different: for the yellow modification, near the edge of the fundamental band a monotonic increase of absorption with decreasing wavelength was found, without any discrete structure, whereas for the red modification a complex spectrum was observed, characterized by the presence of fine lines, bands, and absorption “steps.” A series of experiments showed that the structure observed in the red modification is not connected with stoichiometric excesses of mercury or iodine and, apparently, cannot be explained by lattice defects. In this connection the authors conclude that it is related to the excitation of excitons.

The difference between the absorption spectra of the yellow and red modifications testifies in this case to the substantial influence of the lattice structure on the exciton levels of the crystals.

This is also confirmed by the results of investigation of the absorption spectra of mercuric bromide, HgBr\(_2\). The rhombic lattice of HgBr\(_2\) is isomorphic with the lattice of yellow HgJ\(_2\), and, as it turned out, the absorption spectrum of HgBr\(_2\), like the spectrum of yellow HgJ\(_2\), is devoid of any structure. The presence in the red modification of HgJ\(_2\), in addition to narrow lines near the edge of the absorption band, of comparatively broad bands (100–150 Å) in its depth is explained by the authors, in accordance with the theory of Dykman and Pekar, as arising along with “nonpolarizing” and “polarizing” excitons.

In addition to absorption, the authors also investigated the photoconductivity of red and yellow HgJ\(_2\). The “exciton” mechanism of photoconductivity has been discussed repeatedly in a number of studies. Recently Gross and Belle showed experimentally that, among a large group of substances, only those in whose spectra lines of exciton absorption are found possess clearly expressed photoelectric sensitivity. The authors note that the indicated parallelism between the presence of an exciton absorption spectrum and photoconductivity also holds for the two modifications of mercuric iodide: the presence of structure in the spectrum of the red modification is accompanied by high photoconductivity; the absence of this structure in the yellow modification—by its negligible photosensitivity.

Experimental investigations connected with the exciton hypothesis have in recent years been carried out chiefly as studies of the fine structure of absorption spectra. Meanwhile, when the concept of excitons was applied to the interpretation of data on the internal and external photoeffect, the property of the exciton to be a carrier of migrating energy was used mainly. Hence arises

there is understandable interest in questions concerning the study of exciton diffusion and the determination of the quantitative characteristics of this diffusion. In the report by V. E. Lashkarev and Yu. I. Karkhanin, interesting data relating to this question were presented.

The authors showed that the infrared luminescence of cuprous oxide does not depend on the electric field applied to the specimen and rapidly removing from it all nonequilibrium nonbasic carriers (in cuprous oxide—photoelectrons). On this basis one may conclude that the luminescence must be connected with currentless excitations—excitons. On the other hand, it was found that, when a film of water is present on the surface of the specimen, the luminescence intensity drops sharply with decreasing wavelength of the exciting light (i.e., with decreasing depth of its penetration into the specimen). The authors believe that water on the surface of the specimen promotes radiationless annihilation of excitons. In this case, knowing the dependence of the luminescence intensity on the wavelength and the penetration depth of the light, one can calculate the mean length of the diffusional displacement of excitons. It proved to be approximately equal to one micron. Taking, according to N. A. Tolstoi and N. I. Tkachuk, the exciton lifetime as \(\sim 10^{-6}\) sec, the authors obtain for the diffusion coefficient \(D \simeq 10^{-2}\ \mathrm{cm}^2/\mathrm{sec}\). This value is much smaller than the diffusion coefficient for current carriers in cuprous oxide, which the authors explain by the presence of trapping levels for the exciton.

In the work of Yu. I. Gritsenko and V. E. Lashkarev it was shown that, in cuprous-oxide specimens of low conductivity, obtained by annealing \(\mathrm{Cu}_2\mathrm{O}\) in vacuum under conditions in which reduction of copper takes place, in addition to the usual “short-time” component of photoconductivity (with a time constant \(\sim 10^{-4}\) sec), there arises a “long-time” component, characterized by times \(\sim 10^{-1}—10^{-3}\) sec.

An experimental comprehensive investigation of the temperature dependence of the time constant and of the “yield” of the short-time component confirmed the results obtained earlier by other authors in the study of the photoconductivity of “unreduced” cuprous oxide. A detailed investigation of the long-time photoconductivity showed that it is determined by two components with times \(\tau_2 \sim 5 \cdot 10^{-3}\) sec and \(\tau_3 \sim 5 \cdot 10^{-2}\) sec. In this case the times \(\tau_2\) and \(\tau_3\) proved to be practically independent of temperature.

To explain this fact it is assumed that photoelectrons are rapidly localized at certain centers, after which their recombination with dark holes occurs in the form of two successive transitions: a certain transition whose probability does not depend on temperature and is comparatively small (consequently, the time constant for such transitions is large), and a transition corresponding to the final recombination of the photoelectron with a hole, whose probability is determined by the concentration of holes and, conse-

therefore, depends on temperature. Since the time constant of the first process is assumed to be greater at all temperatures than that of the second, it is precisely this process that determines the observed recombination time, which is independent of temperature.

As a possible model, the first transition is associated with the tunneling transition of an electron from one local center to another, located sufficiently nearby. It is assumed here that recombination of an electron located at the first center with a hole is forbidden, while at the second, on the contrary, it is allowed. The authors believe that, since the appearance of the long-time component is due to reduction of copper, the “long-time” localization centers of photoelectrons include oxygen vacancies in the Cu₂O lattice in their structure.

P. P. Bryodzhinas and M. P. Mikalkevichus presented in their communication some data on the photoelectric and optical properties of polycrystalline and amorphous layers of triselenide and trisulfide of antimony. These materials, possessing appreciable photosensitivity at low dark conductivity, are of considerable interest in connection with the possibility of using them as photoconducting screens of transmitting television tubes (see below the report by Ya. A. Oksman).

The authors note the coincidence of the energies of thermal, photoelectric, and optical (by absorption) activation, as well as the coincidence of the thermal activation energies for layers and bulk specimens of these materials. The latter circumstance does not correspond to the assumption of a substantial influence of intercrystalline interlayers on the properties of the layers studied. At the long-wavelength edge of the intrinsic absorption band, the authors discovered an additional band of strong absorption, which narrows with decreasing temperature and which they associate with the excitation of excitons.

The report by A. V. Airapetyants and S. M. Ryvkin was devoted to a theoretical and experimental consideration of certain questions of pulsed ionization in semiconductors. It was shown that the mechanism of pulse formation in semiconductor crystal counters differs substantially from what occurs in insulators according to the well-known scheme of Gudden and Pohl.

If the lifetime of nonequilibrium current carriers arising during ionization is sufficiently long, then the entire process of pulse formation can be divided into two periods: 1) the period of “primary” current, characterized by the absence of diffusion-drift equilibrium, and 2) the period of “through” current, flowing after this equilibrium has been established. If the amount of charge that passes during the second period is greater than during the first, then a “semiconductor” mechanism is realized, and the charge in the pulse may exceed the total charge released during ionization. In the work, the conditions under which the “semiconductor” mechanism is realized were established, and an experimental verification of the main propositions of the theory was carried out on сер-

cadmium-sulfide α-counters. The report presented data on semiconductor α-counters developed on the basis of single crystals of cadmium sulfide and germanium electron-hole junctions.

B. T. Kolomiets, A. O. Olesk, and S. G. Pratusevich reported on their investigations of the influence of impurities of a number of metals on the electrical and photoelectric properties of pressed polycrystalline specimens of cadmium sulfide. Specimens of this type possess very high sensitivity to light and at present are widely used as photoresistors. The introduction of impurities makes it possible to improve their properties and, in this connection, serves as a technological method in the manufacture of these photoresistors.

The experimental work carried out by the authors showed that, among the metals used as impurities (Li, Sn, In, Tl, Bi, Au, Cu, Ag, Sm, Ga), only copper and silver have a substantial influence on the properties of the specimens. The introduction of these impurities greatly decreases the conductivity. The authors associate a possible explanation of this fact with the replacement of divalent cadmium atoms in the CdS lattice by monovalent copper (or silver), and with the consequent formation of acceptor centers. Then, since the conductivity of ordinary CdS is electronic, the appearance of acceptors leads to partial “compensation” of donor centers and, consequently, to a decrease in conductivity. The introduction of copper and silver leads to the appearance of a new maximum and to an increase in the magnitude of photoconductivity. A new maximum also appears in the spectral distribution of absorption. When the amount of impurity is varied, the character of the dependence of photoconductivity on light intensity may change from “sublinear” to “superlinear.” Of interest is the fact that photoconductivity is quenched in the region of the impurity “copper” maximum upon additional introduction of impurities Fe, Co, and Ni, which, as is known, also quench luminescence. All these features of photoconductivity, as the authors note, have not yet found a satisfactory interpretation.

N. A. Tolstoi, in his report, emphasized the importance of taking into account the influence of the exciting light on recombination processes. Such an influence follows from the experimentally observed anomaly—the absence of direct proportionality between the intensity of the exciting light and the initial rate of relaxation of photoconductivity from its stationary value (i.e., the rate at the moment when the exciting light is switched off). Taking into account the influence of illumination on recombination processes makes it possible, in principle, to explain also the fact of a strong deviation of the ratio of the areas above the “growth” curve and below the “decay” curve of photoconductivity and luminescence from that predicted by theory. The concretization of general considerations on the influence of light on recombination led the author to consideration of a definite variant of a two-step mechanism of phototransitions, with the aid of which it is possible to explain a number of experimental facts discovered by him in the investigation of polycrystalline cadmium-sulfide photoresis-

of resistances activated by copper (see the report by Kolomyets, Olesk, and Pratusevich), and, in particular, two fundamental anomalies discussed above. In trying to remain within the framework of a two-stage mechanism and in explaining the “normal” photoconductivity of CdS, N. A. Tolstoi considers the possibility of photoconductivity via local levels. This assumption makes it possible to explain the increase with temperature of the “normal” photoconductivity of CdS (Cu, Fe) by an increase in the mobility of electrons along local levels. The difficulty that always arises when one attempts to use the concept of an “impurity band”—the necessity of a very large impurity concentration—is removed by the peculiarities of the object investigated by the author. Indeed, as N. A. Tolstoi notes, the properties of fine-crystalline pressed CdS specimens may to a considerable extent be determined by the behavior of the surfaces of the individual crystallites. The impurities, too, are probably located in the near-surface layers. In this connection, the true concentration of impurities in those regions of the specimen which determine the principal photoelectric properties may be sufficient for the formation of “impurity bands,” manifested especially effectively when the temperature is raised.

D. V. Chepur reported on the very high sensitivity to light and X-rays of single crystals and polycrystalline layers of red (tetragonal) mercuric iodide. The relaxation curves of the photoconductivity of HgJ$_2$ are characterized by the presence of two components with somewhat different time constants. These constants depend on illumination and temperature. In the absorption spectrum of red HgJ$_2$, the author found a maximum at about 600 mμ, which coincides with the photoconductivity maximum discovered earlier by Pucheiko. The presence or absence of the maximum is determined by technological conditions, and the author is inclined to attribute it to superstoichiometric iodine. As the temperature is lowered, the maximum narrows and shifts toward shorter wavelengths. In this connection, the supposition is expressed that this impurity maximum is identical with the absorption line at 533 mμ observed by Gross and Kaplyanskii at 77° K. However, against this identification there apparently testify the data reported by Gross and Kaplyanskii on the position of the line in the spectrum at 20° C.

K. K. Demidov described a study of the kinetics of photoconductivity in silver chloride (in the wavelength region 450–550 mμ), which led to a value of the time constant of the process of the order of 10$^{-2}$ sec. Illumination with inactive light (from the region 600 to 900 mμ) led to a certain decrease in the time constant.

Two reports at one of the section meetings were devoted to the photoelectric and optical properties of a new class of semiconductors—organic compounds.

In the report by A. N. Terenin some data were considered on the semiconductor properties of these compounds and, in particular,

dyes. A study of the conductivity and photoconductivity of dyes, carried out over a number of years by Vartanyan, made it possible to establish their semiconductor properties. Among dyes there occur both electronic and hole semiconductors. The temperature dependence of their conductivity has the exponential character typical of semiconductors. The speaker, together with E. K. Putseiko, discovered and investigated in detail an interesting effect of sensitization of the photo-emf, observed by means of the condenser method, in inorganic semiconductors when organic dyes are deposited on them.

The report analyzes in detail experimental data on the absorption spectra and photoconductivity of dyes. The speaker emphasized that the absorption spectra of microcrystalline layers and of isolated dye molecules differ substantially, which argues against the assumption of delocalization of electrons in the ground and excited states in the dye crystal. Consequently, these levels remain localized within the molecules. On the other hand, the presence of electronic photoconductivity in some dyes leads to the conclusion that there exist continuous bands of energy levels, providing the possibility of electron motion through the entire crystal. The contradiction is removed by the author with the aid of the assumption that the bands arise as a result of splitting of levels for which direct optical transitions are forbidden. In this case photoconductivity arises as a result of a two-step transition, namely a phototransition to an excited level of the dye molecule and a thermal transition (with release of energy) from the excited level of the molecule into the conduction band. Dark conductivity appears due to direct thermal transitions into the conduction band. Thus, the model considered by the author, along with explaining the coincidence of the photoconductivity spectrum of the crystal and the absorption spectrum of an individual molecule, also makes it possible to explain the observed excess, in a number of dyes, of the photoionization energy over the thermal ionization energy.

E. K. Putseiko reported on experiments in which sensitization of the photo-emf of certain inorganic semiconductors by organic dyes deposited on their surface was discovered and investigated. The sensitization manifests itself in the appearance, in the photo-emf spectrum, of maxima corresponding to the absorption spectra of the dyes. The sensitization effect of zinc oxide by chlorophyll and its analogs is especially strong. In this connection the author emphasized that it proved impossible to detect the intrinsic photoeffect of chlorophyll. The report presented data from detailed studies of the influence of thermal treatment of samples in air, which indicate the surface character of the sensitized photo-emf. The influence of certain vapors and gases on the intrinsic photoeffect of a number of oxide semiconductors (ZnO, PbO, HgO) was also investigated. The observed

The strong influence on the intrinsic photo-emf of aging or heating in vapors and gases also indicates the surface nature of the observed phenomenon.

Recently, semiconductors with a pronounced covalent type of bonding, crystallizing in a diamond-type structure, have acquired great importance. However, the number of such semiconductors is limited to a few elements of group IV of the Mendeleev table (silicon, germanium, α-tin). In this connection, intensive attempts are being made to find new semiconductors, possessing properties similar to those of germanium and silicon, among binary compounds. Naturally, attention is directed first of all to binary compounds crystallizing in a ZnS-type structure, which in character is closest to the diamond structure.

In the paper by N. A. Goryunova, V. S. Grigorieva, B. M. Konovalenko, and S. M. Ryvkin, results are presented of a study of the spectral distribution of photoconductivity of certain compounds with a defective ZnS-type structure, in particular gallium and indium chalcogenides, as well as some other binary compounds of Ga and In with elements of group VI, belonging to other crystal structures. The relative sensitivities of the compounds studied and the width of the forbidden band (from the long-wavelength boundary of the photoeffect) were determined. In a number of chalcogenides a regular change in the width of the forbidden band was found, which is connected with a change in the degree of ionicity of the compounds.

One of the new and very promising applications of photosensitive semiconductors is their use as photoconducting screens in transmitting television tubes. The simplicity, small dimensions, and certain other advantages of such tubes, which have received the name “vidicon,” make it possible to count on their broad introduction, in particular in industrial television, etc. Of special interest is the possibility, by using different semiconductors, of carrying out television in different regions of the spectrum.

In the report by Ya. A. Oxman the basic requirements imposed on photoconducting semiconductor layers when they are used in a “vidicon” were formulated. It is noted that, along with high photosensitivity, the layers must possess a large specific resistance (not less than \(10^{11}\ \mathrm{ohm\cdot cm}\)) in order that the intrinsic time of the layer

\[ \left(\frac{\varepsilon \rho}{4\pi}\cdot 10^{-12}\ \mathrm{sec}\right) \]

exceed the frame time

\[ \left(\frac{1}{25}\ \mathrm{sec}\right). \]

The speaker studied the current-voltage characteristics of thin amorphous layers of \(\mathrm{Sb_2S_3}\)—a material used in “vidicons”—under conditions close to those of their operation in tubes. In use-

in the method described, the layer was deposited on a rigid substrate covered with a translucent layer of metal. The other side of the layer was charged by an electron beam to a potential that could be measured. The current through the layer was measured with a galvanometer. The author found that deviations from Ohm’s law in the layers under study begin at fields of \(10^4\) V/cm, and that the dependence of the conductivity \(\sigma\) on the field strength \(E\) follows the formula of Ya. I. Frenkel

\[ \sigma \sim e^{\mathrm{const}\cdot \sqrt{E}}. \]

It was shown here that the constant in the exponent is the same for the dark and illuminated specimens. This led the author to the conclusion that in thin amorphous layers of \(\mathrm{Sb_2S_3}\) the field changes not the concentration of carriers, but their mobility.

B. T. Kolomiets and S. G. Pratusevich reported on a new industrial type of photoresistor developed by them on the basis of pressed cadmium selenide powder. It is known that in recent years cadmium sulfide photoresistors (polycrystalline and single-crystal) have appeared, possessing enormous sensitivity of the order of an ampere per lumen. This sensitivity is approximately equal to the sensitivity of photomultipliers, but, whereas the limiting current of a photomultiplier does not exceed \(\sim 100\,\mu\mathrm{A}\), cadmium sulfide photoresistors withstand currents up to several milliamperes. This makes it possible to use them widely together with ordinary relays for the automation of a number of production processes.

The new cadmium selenide photoresistors are another order of magnitude more sensitive than cadmium sulfide ones. The spectral region of their sensitivity extends from \(0.5\) to \(1.2\,\mu\), with a maximum at \(0.8\,\mu\). The new photoresistors will undoubtedly find wide application in technology.

N. A. Goryunova and B. T. Kolomiets described the study of the spectral distribution of photoconductivity in a large number of complex semiconductor systems.

As a rule, these systems are solid solutions of the components that form them. However, at certain ratios of the components both chemical compounds and eutectic mixtures are formed. Among the substances studied there are crystalline and glassy ones. In the work, photoconductivity was found and its spectral distribution studied in the systems: \(\mathrm{Tl_2Se — Sb_2Se_3}\), \(\mathrm{Tl_2S — Sb_2S_3}\), \(\mathrm{Bi_2S_3 — Sb_2S_3}\), \(\mathrm{Sb_2Se_3 — As_2Se_3}\), \(\mathrm{Tl_2Se — As_2Se_3}\), \(\mathrm{Tl_2Se \cdot As_2Se_3 — Tl_2Se \cdot As_2Te_3}\), \(\mathrm{Tl_2Se \cdot As_2Se_3 — Tl_2Se \cdot Sb_2Se_3}\), \(\mathrm{CdTe — ZnTe}\), \(\mathrm{CdSe — In_2Se_3}\), and others. By changing the ratio between the components of the system, the authors obtained a change in the spectral distribution over wide limits. At the same time, the absence of any “additivity” was demonstrated: the spectral distribution in a substance obtained by combining binary components may differ sharply from the distribution in these components. The authors believe that

Further investigation of the properties of ternary and other complex systems may lead to the possibility of deliberate control of their electrical and photoelectric properties.

In the report by A. I. Goryachev and K. A. Yumatov, the results of investigations of noise in lead-sulfide photoresistors are presented. It is known that the magnitude of the noise determines the sensitivity threshold of photoresistors and, consequently, the possibility of using them for detecting weak signals. The investigation, carried out under matched-load conditions, showed that prolonged “aging” of photoresistors leads to a decrease in the noise level and to an approximation of the dependence of the noise signal on voltage to a linear one. At the same time, the noise level is essentially determined by the magnitude of the electric field in the specimen. In the specimens investigated, the spectral density of the noise-voltage intensity proved to be inversely proportional to frequency. The authors conclude that the noises observed in PbS are a consequence of fluctuations of the dark resistance of the layer.

A large part of the section’s work was devoted to discussion of reports concerned with the study of photoelectromotive forces in semiconductors. These questions have not only theoretical but also great practical significance in connection with the use of barrier-layer photocells, photodiodes, etc.

V. E. Lashkarev and V. A. Romanov reported on the study of a new type of photo-emf. Investigations of photoelectromotive forces carried out over the last two decades had led to the conclusion that observation of a stationary photo-emf in a specimen with metallic electrodes is possible only in the presence of “blocking” or “antiblocking” layers at the electrodes, and under the condition that the illuminated region of the semiconductor is sufficiently close to these electrodes. It is known that photo-emf can also be observed when a region in the depth of the semiconductor is illuminated, but close to an electron-hole transition.

The authors described a new type of photoelectromotive force, which they called “bulk,” arising in a semiconductor devoid of \(n\)-\(p\) junctions and when a region remote from the electrodes is illuminated.

It was shown that the appearance of the “bulk” photo-emf is connected with the presence of inhomogeneity of the specimen in conductivity. Extension of the general theory of photoelectromotive forces, developed earlier by Lashkarev, to the case under consideration led the authors to the development of a theory of the “bulk” photo-emf. It was found that the magnitude of the photo-emf is proportional to the gradient of the specific resistance and to the square of the diffusion-displacement length. In the work an experimental verification of the basic relation for the bulk

photo-emf. For this purpose, the variation of the resistance and emf along an inhomogeneous germanium sample was recorded. The check led to qualitative and quantitative agreement of the experimental results with the theory. The authors note the possibility of using the bulk photo-emf effect to check the homogeneity of samples with respect to specific resistance. Such a method of homogeneity control may be especially effective for germanium and other materials with a large diffusion displacement length (since the magnitude of the bulk photo-emf is proportional to the square of this length).

It is interesting to note that an independent investigation of the bulk photo-emf effect was carried out at the Prague Institute of Technical Physics by Dr. Tauc, who gave brief information on the principal results of this work in the discussion following the report by V. E. Lashkarev and V. A. Romanov. Speaking in the same discussion, Prof. L. Sosnowski (Warsaw) emphasized that the photo-emf discovered by him, arising upon illumination of polycrystalline PbS and PbSe films, where the diffusion displacement length is small, can be explained only by the presence of very large impurity concentration gradients, which in essence lead to the formation of \(n\)-\(p\) junctions. In this sense the bulk photo-emfs in germanium and in PbS are opposite limiting cases, between which, however, there is the possibility of a continuous transition.

In the report presented by Zh. I. Alferov, B. M. Konovalenko, S. M. Ryvkin, V. M. Tuchkevich, and A. I. Uvarov, data were given on the design and principal characteristics of planar germanium photodiodes developed at the Leningrad Physico-Technical Institute of the Academy of Sciences of the USSR. The photodiodes have a sensitivity of \(\sim 30\,000\) \(\mu\mathrm{A}/\mathrm{lm}\), and their intrinsic time, determined by the diffusion time of nonequilibrium carriers from the place of generation to the \(n\)-\(p\) junction, in the design used was not more than \(\sim 10^{-5}\) sec. The work considered the mechanism of formation of the photovoltaic photo-emf at an \(n\)-\(p\) junction, as well as the equation of a photodiode operating with an applied external voltage. The question was analyzed and a general expression was obtained for the efficiency of photodiodes operating in the mode of a photovoltaic cell. An experimental check of the obtained relations was carried out for the “photovoltaic” and photodiode modes at room temperature and at the temperature of dry ice, with good agreement obtained with the predictions of the theory. The theoretically predicted dependence of the efficiency on illumination intensity was experimentally confirmed.

The report by V. S. Vavilov and L. S. Smirnov was devoted to the results of an experimental investigation of the question of the efficiency of conversion of light energy into electrical energy in germanium photocells. The authors analyzed in detail the question of the factors affecting the efficiency of energy conversion. Experimental accounting for all these factors made it possible to determine the coefficient,

characterizing the fraction of minority carriers that manage to pass through the \(n\)-\(p\) junction before recombination. In the photocells investigated this coefficient was \(0.5\)—\(0.7\). To increase it, the authors propose creating a reflecting barrier on the front surface, i.e., essentially an anti-blocking layer, whose existence should reduce recombination at the surface. To create the reflecting barrier, a semitransparent layer of a specially selected material must be deposited on the illuminated surface. The study showed that the increase in efficiency predicted by the theory with increasing power of the absorbed radiation occurs only up to a certain limit. This limit is determined by the presence of a finite resistance of the germanium bulk and of the contacts.

In the report by A. G. Goldman, some results are presented from an investigation of selenium photocells in the photodiode regime.

G. B. Abdullaev and M. A. Talibi reported on the sensitivity of \(n\)-\(p\) junctions, formed at the contact between selenium and cadmium sulfide (a selenium rectifier with a thin CdS layer under the upper electrode), to x-rays and gamma rays.

V. E. Kozhevin and V. E. Lashkarev described their investigation of the influence of external voltage, the type of “insulating” spacer, etc., on the magnitude and sign of the photo-emf measured by the capacitor method. The appearance of a photo-emf upon illumination of a completely homogeneous semiconductor placed in a capacitor is connected with the diffusion of photocurrent carriers from the front illuminated surface into the interior of the semiconductor. In the presence of photocurrent carriers of one sign, this leads to the appearance of a photo-emf whose sign is connected with the sign of the carriers and can serve to determine it. However, if photocarriers of both signs are present and blocking or anti-blocking layers exist on the semiconductor surface, with the corresponding band bending, then it is clear that the indicated method for determining the sign becomes unsuitable. In the latter case a barrier or “anti-barrier” photo-emf arises, and its sign will be determined not so much by the predominant sign of the photocurrent carriers as by the sign of the band bending. Theoretically, the question of the influence of band bending and surface states on the “capacitor” photo-emf in the case of “monopolar” and “bipolar” photoconductivity was previously considered in detail by V. E. Lashkarev. The report presents the results of an experimental verification of the considerations obtained in the theoretical treatment of the question. It turned out that, depending on the type of insulating spacers creating one or another band bending in close contact with certain semiconductors, the magnitude and sign of the capacitor photo-emf can change.

Detailed studies of the influence of the field on the capacitor photo-emf showed that the action of the field is observed only for spacers possessing appreciable conductivity, and the observed phenomena can

may in a number of cases be explained by the superposition of photo-emf and photoconductivity.

In the report by R. Ya. Berlagi, M. A. Rumsh, and L. P. Strakhova, the results were presented of a study of oxidized layers of lead sulfide capable, under illumination, of generating a photo-emf. A distributed photo-emf in PbS layers was first discovered by Sosnovsky, Starkevich, and Simpson in layers through which a direct current had previously been passed. In the work reported here, such preliminary treatment was not used. The authors showed that the directed asymmetry in the structure of the layers, necessary for the appearance of a photo-emf distributed along the entire layer, is in their case connected with the textured structure of the polycrystalline specimens: the crystallites in the layer grow during the sublimation of PbS predominantly in the direction of the molecular beam. The presence of crystallite growth directed along the direction of the molecular beam was proved by the authors by electron-diffraction and direct electron-microscopic investigations. The relation of the photo-emf to such a structure of the layers is illustrated by the dependence of the magnitude and sign of the photo-emf on the direction of the light beam. When this direction is changed, the sign of the photo-emf changes at the moment when the directions of the light and molecular beams coincide. The authors connect a possible explanation of the mechanism of the photo-emf either with an external photoelectric effect from the illuminated boundaries of crystallites, or with an internal photoelectric effect and the formation of a photo-emf as a result of diffusion or at surface \(n\)-\(p\) junctions. It is noted, however, that the mechanism of the external photoelectric effect is hardly possible, since it contradicts the data on the spectral distribution of the photo-emf.

The authors’ experiments apparently testify to the important role played by conditions at the boundary between crystallites in explaining the photoelectric properties of lead-sulfide photoresistors.

In the work of A. L. Rvachev and A. I. Andrievsky, the influence of the degree of reduction of cuprous oxide on the front-wall and back-wall barrier photo-emf of cuprous-oxide photocells was investigated. It was shown that the degree of reduction strongly affects the character of the spectral distribution of sensitivity. The existence, in the spectral distribution of the front-wall photo-emf, of infrared sensitivity is attributed by the authors to the presence of a copper-oxide film. This is proved, in particular, by the fact that in the absorption of cuprous oxide taken from photocells with infrared sensitivity, features appear that are inherent in the absorption of pure cupric oxide. The increase in relative sensitivity in the infrared region for the back-wall photo-emf is explained by the effect of “filtering” of light in the cuprous-oxide layer. The authors found a new maximum of spectral sensitivity of the back-wall photo-emf at \(655\ \mathrm{m}\mu\).

Two papers were devoted to the external photoelectric effect from semiconductors.

Yu. A. Shuba, studying the external photoelectric effect from CdS, found a sharp influence of illumination in the region of intrinsic absorption on such quantities as the contact potential, the quantum yield, and the energy of the emitted electrons. In the author’s experiments the quantum yield could increase by an order of magnitude, while the contact potential decreased by 0.2–0.5 eV. Illumination with long-wavelength light led to changes of the opposite character. It was found that the energy of electrons emitted from CdS does not depend on the energy of the exciting quanta, which the author relates to the exciton mechanism of the external photoelectric effect.

P. S. Popov reported on the results of an investigation of the external photoelectric effect from lead sulfide and copper selenide. As a result of the measurements, data were obtained for the distance \(\delta\) between the top of the filled band and the Fermi level, and also for the photoelectron work function; the thermionic work functions were calculated as well. The values of \(\delta\) varied from sample to sample, which is apparently connected with changes in impurity content. The values of the photoelectron work functions were \(\sim 4.2\) eV for PbS and \(\sim 4.0\) eV for CuSe.

In conclusion, we shall give a brief summary of two methodological papers.

In the paper by A. M. Bonch-Bruevich and Ya. I. Imass, the question was considered of the possibility of effectively reducing the inertia of semiconductor radiation receivers (for example, bolometers) by using a special correcting electrical circuit. In the proposed method, the reduction of inertia is accompanied by an increase in the noise level and, consequently, is achieved at the cost of lowering the threshold sensitivity of the receiver. Comparing the method of electrical correction with another known method for reducing the inertia of bolometers by increasing heat transfer, the authors show that the method of electrical correction, in addition to its simplicity and flexibility, in some cases possesses (for a fluctuation voltage spectral distribution decreasing with frequency) a higher value of threshold sensitivity.

In the paper presented by S. M. Ryvkin, the operating principle and construction of a two-disk mechanical modulator are described, making it possible to obtain light pulses with steep fronts, separated by long intervals of darkness. The modulator consists of two disks, rigidly coupled by a worm gear, with sector-shaped cutouts; the disks are placed in the path of the light beam and rotate at different speeds. For certain ratios between the parameters of the disks, the front of the light pulses is determined by the rapidly rotating “shaping” disk, and the time between pulses by the slowly rotating “triggering” disk. Data are given

constructed modulator, which produces light pulses of duration up to \(\sim 10^{-2}\) sec with a leading edge of \(\sim 10^{-5}\) sec and a dark interval of \(\sim 2\) sec. The modulator is also suitable for obtaining single light pulses with the indicated parameters.

The work carried out in recent years on the study of the photoelectric and optical properties of semiconductors, and partially reflected in the reports at the “photoelectric” section, testifies to the definite successes achieved in this field.

It is necessary, however, to emphasize that a number of important questions relating to the study of nonstationary electronic processes—and, in particular, questions of internal ionization in semiconductors under the action of radioactive radiation, recombination of carriers, the optics of atomic semiconductors, etc.—are still being developed with insufficient intensity. The conference resolutions emphasized the importance of considerably strengthening research in these directions. Equally evident is the need for a serious intensification of work in the field of the practical use of semiconductor devices as energy converters, signal converters, and detectors of various radiations.