CONGRESSES AND CONFERENCES

CONFERENCE ON SECONDARY ELECTRON EMISSION AND THE PHOTOELECTRIC EFFECT

1. On January 20–22, 1939, in Leningrad, at the Scientists’ Club in Lesnoye, sessions were held of the Conference on Secondary Electron Emission and the Photoelectric Effect, convened by the Department of Experimental Physics of the Engineering-Physics Faculty of the Leningrad Industrial Institute on the initiative of Prof. P. S. Tartakovsky.

At this conference more than two dozen reports and communications were presented by workers from various laboratories in Leningrad, Moscow, and Kiev. The number of reports reveals growth by several times in comparison with similar conferences held in 1936 and 1937. This growth proves even more significant if one compares the subject matter of the reports at the present and preceding conferences, since the present conference was devoted entirely to fundamental questions, and questions of an applied and technological nature were touched upon only in passing. All this testifies to the broad development of scientific activity in our country, and since the overwhelming majority of reports belonged to young scientific workers, it also testifies to the fact that ever new young cadres are coming into science.

Significant progress in comparison with previous years is also apparent in another respect. Whereas formerly at such conferences reports presented exclusively the results of experimental investigations, which was entirely natural, since broad interest in secondary emission and the photoelectric effect arose on the basis of practical requirements, here a considerable number of works of a theoretical character were already reported. Likewise, whereas previously the objects of investigations on secondary emission were exclusively complex surfaces, as they present practical interest, now, although complex surfaces remain at the center of attention, a considerable interest in the study of pure metals has also appeared. Thus there is evidently a substantial expansion of the field of research and a considerably fuller coverage of the problem.

The conference showed that in this field, in any case as far as can be judged from the foreign scientific periodicals, Soviet science holds a leading place both with respect to the excellent scope of work and in the sense of the variety and breadth of the investigations. At the same time it is necessary to note certain shortcomings in the organization of the work. These include the often excessively simple—not to say primitive—experimental technique, as well as the insufficient technical equipment of some laboratories. In particular, let us note that, as a rule, in investigations of the photoelectric effect the workers do not use monochromators, owing to the absence of such instruments. It is hardly necessary to prove that such a situation is extremely unsatisfactory and inevitably affects the quality and productivity of the work. It seems to us that this circumstance should attract serious attention from the organizations responsible for the appropriate equipment, with a view to making spectral apparatus and sensitive electrical measuring instruments more accessible to our laboratories, both as regards the possibility of acquiring them, as ...

and in terms of cost. It should not be forgotten that abroad all this apparatus is available in abundance and at comparatively accessible prices, which has no small significance for the foundations of the work, so that our industry, while successfully mastering incomparably more complex production processes, lags behind in this area.

Finally, the experience of this conference showed that the pace of work of our scientific journals, which sometimes print articles 8–9 months after their receipt by the editorial office, greatly lags behind the pace of work of laboratories. This circumstance does not allow the scientific press to fulfill its organizing role. This is manifested with particular clarity in the analysis of the reports on the photoeffect made at the conference. The point is that, with the exception of reports of a review nature, all reports on questions of the photoeffect concerned antimony–cesium cathodes, which is understandable if one takes into account the great fundamental and practical interest that these cathodes represent.

Meanwhile, of the three communications fully devoted to Sb—Cs photocathodes, only one constituted a substantial step forward. The remaining two represented little more than a description of their properties based on newer or less successful experiments, mainly amounting to a repetition of what had already been done earlier. There is no doubt that if the works had appeared in print in a shorter time, this would have made it possible to organize the work of all laboratories much more fruitfully.

II. The introductory report by S. Yu. Lukyanov was devoted mainly to the classification of photocathodes and emitters according to their structure, to a brief review of experimental data, and to comparison with the results of theory. At the end, a number of remarks were made on the question of the Sb—Cs photocathode, among which should be noted the indication of the high value of the quantum yield in the region of the selective maximum, reaching 30%.

The report of Prof. P. V. Timofeev contained an exposition of the views of the de Boer school on the mechanism of emission of oxygen–silver–cesium photocathodes, as well as a number of considerations in favor of the assumption that Sb—Cs cathodes should also be assigned to the type of ordinary complex cathodes. The second part of the report was devoted to the secondary emission of complex surfaces, about which further discussion will be given below.

B. I. Dyatlovitskaya reported on work with Sb—Cs cathodes at the Physical Institute of the Academy of Sciences of the USSR. According to the measurements carried out there, the specific resistance of these layers is of the order of \(1 \Omega\cdot\text{cm}\). From the temperature dependence of the resistance, the work of ionization of the structural elements of the layer was determined to be \(0.3\,V\). Nevertheless, an internal photoeffect in the region of long waves was not detected. In the discussion it was pointed out that the absence of an internal photoeffect should be attributed to the conditions of the experiment (low voltage). According to the data of this work, fatigue is absent in Sb—Cs cathodes.

Yu. I. Lunkova, in her communication, indicated that Sb—Cs cathodes exhibit fatigue, and also presented data on the volt-ampere characteristics at various light fluxes. The essence of these results amounts to the fact that as the light flux increases (in actuality, its density, i.e. the illumination), the saturation voltage increases.

N. S. Khlebnikov, on the question of the structure of antimony–cesium layers, reported that direct determination of the ratio of the content of Sb and Cs gives a ratio of the number of Sb atoms to the number of Cs atoms equal to \(1:3.1\)—\(1:3.2\). This agrees with the known values for the negative valence of Sb (2 and 3) and shows that this layer is a compound of Sb with Cs. The excess Cs must, however, be attributed, in the determination error, both to the presence of free cesium in the volume of the layer and to the film of Cs adsorbed on its surface.

Of special interest was, unfortunately, the brief communication of S. S. Prilezhaeva, who reported on her work, in which the role of the surface of an antimony–cesium cathode was thoroughly investigated, which [[unclear: continuation cut off at bottom of page]]

photoemission. Investigation of the relation between the degree of cesium coverage of the surface, its work function, and the integral sensitivity, as well as the study of the energy distribution of the photoelectrons, led the author to the idea that the properties of this photocathode are entirely determined by the surface.

In the discussion it was pointed out that this view encounters certain difficulties, in particular it is not in agreement with the high value of the quantum yield. Later, in a private conversation, S. S. Prilezhaev informed the author of these lines that, while attaching exceptional importance to the conditions at the surface, he allows for a volume character of the photoeffect and considers that the distribution of electrons in the layer is not Fermi-like.

Thus it may be considered that Sb—Cs photocathodes have as their basis the lattice structure SbCs₃, in which free cesium is present, and that on the surface there is an adsorbed cesium film; the photoeffect has a volume character; apparently, cathode fatigue is sometimes observed, the mechanism of which is not yet sufficiently clear.

III. A lively discussion developed on the question of general conceptions of secondary emission, because on this question two entirely different points of view were expressed. The first of them, developed by N. D. Morgulis and N. S. Khlebnikov, claims to provide a fairly broad generalization so as to make it possible to regard a certain qualitative scheme of the theory of the phenomenon in its full scope. As is known, the views of these authors are based on consideration of the processes of formation of secondary electrons in the emitting layer, their motion in the bulk of the emitter, and the process of replenishment of emitted electrons, the principal physical factor characterizing a material with respect to secondary emission being the concentration of conduction electrons. From this, in particular, it follows that large coefficients of secondary emission can be observed (under appropriate conditions) only in semiconductors and dielectrics and never in pure metals. The qualitative conclusions and predictions made on the basis of this point of view are in complete agreement with the existing experimental material. Thus, for example (as was reported by N. S. Khlebnikov), for electronic semiconductors there is observed a course of the temperature dependence distinguished by the presence of a maximum at a certain temperature, whereas for ionic semiconductors (Vul’skii’s data) this dependence is expressed by a curve with saturation. On the other hand, N. D. Morgulis gave an approximate calculation of the energy losses incurred by secondary electrons as a result of various types of interaction with the lattice and with the free electrons of the emitter, which showed that the most significant are losses due to Coulomb interaction.

The other point of view was formulated by P. V. Timofeev in the second part of his memorable report. His views, which claimed no more than to explain the high secondary emission of complex cathodes, were this time reduced to the proposition that the high values of \(\sigma\) are due to three causes: 1) the existence of potential barriers that form at the boundary between particles of metal and particles of dielectric and, possibly, in cracks; 2) the existence of channels (cracks) in the cathode, into which electrons can be knocked out and then extracted outward by an additional field arising as a consequence of 3) the formation of charges on the cathode surface (similar to the Malter effect). As is known, earlier P. V. Timofeev attributed all properties of complex surfaces only to factor (1). A characteristic feature of all three propositions put forward is the absence, noted in the discussion, of any direct experimental proof of their presumed influence of the required sign on secondary emission. The presence of potential barriers within a layer, as follows in particular from the statement of P. V. Timofeev himself in the discussion, should be an obstacle to high secondary emission, since these barriers are more effective for scattering slow electrons than fast ones.

The criticism by P. V. Timofeev regarding the first of the points of view mentioned this time was based on the experiments of Katz, who found that slow electrons (10–15 V) pass through thin metallic films with less scattering than fast ones (≈100 V). But even leaving aside the complex doubts as to the sufficiently continuous structure of the thin films with which Katz worked, no conclusion at all follows from these experiments that slow electrons (with energies of the order of a volt) do not lose energy through Coulomb interaction with conduction electrons. Moreover, Katz’s data (obtained when a beam passes through a film) cannot be directly applied to ordinary cases of secondary emission, since in this latter case, in addition to energy losses, one must also bear in mind processes connected with a change in the initial angular distribution of the secondary electrons produced.

Among the other communications in the present section, mention should be made of the papers by Kubetskii, Pyatnitskii, and Frimer.

L. A. Kubetskii, in his report, proposed that high secondary emission can be obtained by using emitters composed, like cesium, of one of the alkali metals, one of the heavy metals of Group I of the periodic system (Cu, Ag, Au), and one of the following elements of Group VI: O, S, Se, Te (which gives, obviously, 60 combinations), and then described the technology and properties of one such combination (Cu—S—Cs).

A. I. Pyatnitskii reported on work investigating the dependence of the secondary-emission coefficient on the density of the primary beam. In this work it was established that, with an increase in the density of the primary beam, the secondary-emission coefficient falls. If, however, the collector voltage is raised, then \(\sigma\) also increases. Pyatnitskii explained this by “changes in the properties of the surface,” without specifying what changes these are.

A. I. Frimer reported the results of measuring the coefficient of secondary emission and the energy distribution of secondary electrons for copper oxide. In the case of reduced copper oxide, \(\sigma_{\max}\) does not exceed 1.5 (\(V_p = 400\)—500 V); an oxide layer in which there is free oxygen gives higher secondary emission. The energy distribution of the electrons shows no substantial differences from the distribution for metals. In the discussion it was noted that the method of obtaining the layer (\(\mathrm{Cu_2O}\)) was not irreproachable, and therefore it remained unclear to what extent the properties described were specifically those of copper oxide.

IV. In connection with the special position which, among other substances, semiconductors and dielectrics occupy in the phenomena of the photoelectric effect and secondary emission, three reports were devoted to the theory of these substances.

The report by P. S. Tartakovskii was of a survey character and contained an exposition of various points of view (those of the schools of Pohl, Tartakovskii, Frenkel, etc.) on the motion of charges inside these substances. The speaker noted that the interpretation of phenomena in the spirit of any single point of view is, as a rule, not sufficiently exhaustive, and pointed to the possibility of synthesizing the point of view developed in Pohl’s works (the basis of which is the idea of “electron shift”) with the point of view of energy zones.

A brief communication by E. M. Tsenter concerned a theoretical estimate of the effective cross section for the production of secondary electrons (regarded as the “transfer” of an electron from the band of occupied levels into the conduction band) in the case of a dielectric, both pure and containing inclusions. From the data presented it follows that the effective cross section proves to be of one order for electrons of inclusions and for electrons of the lattice itself. Applied to the secondary emission of complex surfaces, this means that, since the concentration of inclusions is very low (\(\sim 10^{-6}\)), the source of the secondary electrons produced must basically be the dielectric itself.

In his report N. P. Pisarenko drew the attention of the conference to those improvements in the Bloch theory of crystals whose necessity arose as a result of the most recent investigations, in particu-

these data on the temperature dependence of the spectral distribution of sensitivity (the position of the long-wave limit) of cuprite and silver chloride.

V. Of the reports concerning secondary emission of pure metals, one contained an account of experimental investigation (secondary emission of silver as a function of the structure of the metal), and two were devoted to the theory of secondary emission.

In the work reported by Morozov, jump-like changes in the secondary emission of silver were established at the transitions from the \(\beta\)- to the \(\gamma\)-modification, and also during melting.

The theories of secondary emission of metals proposed by A. Vyatskin and A. E. Kadyshevich differed very substantially from one another.

Vyatskin quite justifiably pointed out a number of shortcomings in Fröhlich’s theory. But at the same time he himself constructed his theory on a very precarious foundation (which was subjected to well-deserved criticism in the discussion), namely on the assumption that the secondary emission of metals is a purely surface effect. One consequence of this was the absence of maxima on the curves \(\delta=f(V_p)\) obtained by calculation, which contradicts experiment.

Kadyshevich’s theory was constructed as a theory of a volume effect. Here, in contrast to Vyatskin, a Sommerfeld model of the metal and Born’s method were used; the process of occurrence was treated as paired Coulomb collisions between the primary electron and the electrons of the metal, and scattering of the secondary electrons arising within the bulk of the emitter was taken into account. The presence of a maximum on the curves \(\delta=f(V_p)\) follows directly from the distinction in the lengths of the paths of the primary and the emerging secondary electrons. Thus the scheme of this theory does not differ from the qualitative scheme proposed by Khlebnikov and Morgulis.

In the very lively discussion of the last two reports, several persons (Davydov, Timofeev) expressed the opinion that a conclusion is still premature as to which of the two viewpoints is correct; the completed theory of the phenomenon should represent a compromise between both points of view, and the role of the surface (as follows from all the available experimental material) must be taken into account as a certain correction to the volume treatment of the problem.

VI. In another group of reports (Yu. M. Kushnir, P. P. Borzyak, A. A. Ravdel), phenomena were considered that arise under the simultaneous action on the emitter of two exciting agents (light and an electron beam, two electron beams of different velocities). The aim of these works was to verify the experiments of Losev and Shpakov, who had found an absence of additive summation of the emission currents caused by each of the factors. However, all the experimental data are as yet quite far from giving a definite answer even to the question of the existence of these effects, and opinions in the discussion were divided.

VII. As became clear in the discussion, the agenda of the conference, for one reason or another, did not include communications on works of direct and unquestionable interest and significance. Among these should be included, for example, the work of N. L. Yasnopolsky on the theory of secondary emission, the further investigations of M. M. Vudynsky on the secondary emission of dielectrics, and others. Thus this conference, while demonstrating the significant progress in this field, nevertheless does not constitute a complete reflection of the entire volume of work being carried out on these questions in the Union.

In conclusion, we wish to emphasize the timeliness of the initiative of Prof. P. S. Tartakovsky and the benefit brought by the broad exchange of opinions—circumstances that were gratefully noted at the concluding session of the conference.

N. Khlebnikov, Moscow