ABSTRACTS

ELECTRONIC ANALYSIS OF SURFACE STRUCTURES

The study of all kinds of structures may be carried out both with the aid of soft (approximately up to 1000 V) and with the aid of hard (considerably above 1000 V) cathode rays. The use of hard rays was at first limited to the study of bulk structures, since at the comparatively great depth of penetration of fast electrons (of the order of \(10^{-6}\) cm and higher) the diffraction pattern from surface layers is lost alongside the considerably stronger effect from several hundred atomic planes. The study of surface structures, in the field of which the names of Davisson and Germer and Rupp are pioneers, was for a long time conducted only with the aid of soft cathode rays. True, this latter method is convenient in that investigations with reflected electrons do not require the preparation of thin sheets or powders, which is unavoidable when working with transmitted hard electrons. However, it turns out that the study of surface structures can also be conducted with the aid of reflected fast electrons; for this purpose it is necessary to make them fall at a very small angle upon the surface under study.

Such experiments were first carried out by the same G. P. Thomson,\(^2\) to whom we owe the discovery of electron interference (1928).

This new method makes it possible to obtain good diffraction patterns from many of the most varied substances, for example from evaporated metals or from those whose surface has undergone oxidation or some other chemical change. The chief condition for success is that the surface be sufficiently flat.

The apparatus devised by Thomson is intended chiefly for work with single crystals, in which case the diffraction pattern should consist of only a small number of spots. This is indeed what is obtained, for example, when electrons are reflected from the cleavage plane of a rock-salt crystal. The black and white lines which also appear in the diagrams are explained by the existence of double scattering of the electrons (Kikuchi). But some experiments with metal single crystals led to completely unexpected results. In the case of reflection of electrons from a simple crystallographic face, there is obtained an entire system of regularly arranged spots, such as ought to be obtained in the case of a monoatomic layer situated normal to the cathode beam. Moreover, it should represent, as it were, an inverse image of the actual arrangement of atoms in the crystal, for each spot in the diagram corresponds to one row of atoms in the crystal; the entire atomic plane behaves like two intersecting optical gratings. In this respect the agreement between the diagram and theoretical conclusions extends to the smallest details. It must be supposed, however, that such a monoatomic plane, situated at a right angle to the main surface, would be mechanically impossible.

Therefore Thomson believes that the surface of a crystal is not ideally flat, but is covered with pits and pyramids, just as is obtained in etching. In such a case the reason for the appearance of this remarkable diffraction pattern could consist, for example, in the fact that, owing to the very small inclination of the ray, on its path it must meet the vertices of several pyramids, which are distributed over different places of the surface and which have different heights. In other words, diffraction occurs from single-atomic layers, but from ones situated in different planes. Such a pattern is obtained, for example, when electrons are reflected from the cubic face of a silver crystal.

In the case of films of fused platinum, besides diagrams corresponding to its normal structure, diagrams of anomalous types are sometimes obtained. For example, in one case straight lines were obtained which proved unstable and disappeared under the action of cathode rays. Judging from the diagrams, in this case one has to deal with the presence on the surface of small pieces of some crystalline structure, where the atoms are arranged in layers parallel to the surface of the film, each individual layer being built chaotically. It has not yet been possible to decide whether these layers are composed of platinum atoms or of some gas. Another anomalous structure is indicated by the presence in the diagrams of additional rings, which compel one to think of some connection between the structure and the catalytic activity of platinum.

When the surface of a metal is subjected to polishing, then, according to Boalby’s views, under the influence of friction the metal flows and then hardens as a supercooled liquid, as a result of which the surface of the metal becomes amorphous. To investigate this phenomenon French made use of Thomson’s structural tube. The experiment was carried out with small pieces of copper and silver, which were first examined after etching and then after polishing. In the first case sharp diffraction rings were obtained, situated in exact correspondence with the known atomic distances in the lattices of copper and silver. When the tube was comparatively roughly polished (diamond paper, down to No. 0000) with benzene lubricant to protect the surface of the metal from emery dust being rubbed into it, rings were obtained corresponding to the same atomic distances, but now in a blurred form, similar to what is obtained in the case of electron diffraction by gas molecules. Since polycrystals with very small grain size give sharp diffraction rings and, at the same time, as the grain size decreases the sharpness of the rings does not diminish, the result of the indicated type of rubbing is very simply explained by a decrease in the initial grain size. With further polishing by means of burnt magnesia and water as lubricant, the mirror surface of the metals gives only two broad diffraction rings, situated moreover in different places than the former rings. In French’s opinion, the change in the surface structure occurs here at the expense of the shifting of the metal atoms and the formation of their random arrangement, in any case one different from their regular arrangement in the cubic lattice. And since the atoms can approach one another, in the final analysis, only to a definite distance, which will be the predominant one, this half-ordered arrangement will serve as a diffraction lattice. Calculation shows that, instead of the smallest distances by which atoms in the lattices of copper \((2.54 \ \text{Å})\), silver \((2.88 \ \text{Å})\), and iron \((2.78 \ \text{Å})\) are separated from one another, in the case of a polished surface we have, respectively, \(2.59 \ \text{Å}\), \(2.72 \ \text{Å}\), and \(2.67 \ \text{Å}\). Exactly the same rings are also given by iron, silver, and copper in the case of their polishing with crocus and water and subsequent washing, in different cases, with absolute alcohol, benzol, or water. In-

It is interesting that when crocus is used without water, a Debye diagram of a normal crystal is obtained together with additional rings that the crocus itself gives. As observations with an immersion microscope show, the crocus particles do indeed penetrate into the surface of the metal. Thus French believes that Beilby’s theory is confirmed by these experiments.

Exactly the same results with polished metals were obtained also in Kirchner’s laboratory in Germany. Kirchner, however, believes that in order to explain these results there is no need to assume that the crystals constituting the surface change their size so greatly during polishing. He gives a simpler explanation. The surface of a polycrystalline metal gives sharp diffraction rings because it consists of small pieces that are sufficiently thin for electrons to pass through them. This means that, in this respect, Kirchner adheres to Thomson’s views. The polishing process itself, in his opinion, comes down to the leveling of these pieces and, consequently, to a gradual decrease in the resolving power of the lattice in accordance with the decrease in its width.

In other words, the degree of sharpness of the diffraction rings says nothing about the actual size of the crystals, but serves only as an indication of the degree of leveling of the surface. At the highest stage of polishing, when the protrusions on the surface completely disappear, there is true reflection from the perfectly plane surface of the polycrystal. At the same time, refraction of the cathode rays also occurs; the result of this is a considerable change and blurring of the interference maxima, and in their place the disappearance of rings that had previously arisen from the crystallographic planes (III) and (002).

In connection with these views, the new observations made by Thomson and his co-workers are of great interest.^5 By means of electron diffraction they examined films obtained by cathodic sputtering of platinum onto a glass surface. As noted above, such films give the most varied diffraction patterns, one of which corresponds to platinum in its normal state. But in the case of a film obtained by sputtering the metal in oxygen at a pressure of not less than 0.04 mm and at a voltage not exceeding 1150 V, and which is a good catalyst in the reaction between hydrogen and oxygen, diffraction rings are obtained that are quite similar to those obtained for polished metals by French. However, after the platinum loses its catalytic activity—which occurs, for example, as a result of its prolonged use as a catalyst, and also upon simple heating in vacuum—these diffuse rings are replaced by sharp rings characteristic of normal platinum. It can hardly be assumed that these diffuse rings are caused by a completely plane platinum surface, for such a surface could not be catalytically active. It is more probable that, with this replacement of diffuse rings by sharp rings, the surface becomes smoother rather than rougher, i.e. that here there is an improvement in the conditions for true reflection. Thomson believes that these diffuse rings are caused by some substance located in a state of extreme fine dispersion and, in addition, perhaps containing gas molecules.

In addition to this discussion, it is also interesting to cite the results of a study of the surfaces of ground metals obtained in Germany by Boas and Schmid,^6 who for this purpose used the X-ray method. The latter consisted in obtaining reflection diagrams with very soft rays. For the study they took aluminum, copper, zinc, cadmium, and antimony in the form of rods about 4 mm in diameter. First Laue reflection diagrams were taken from these crystals and then, after one or another degree of polishing—

polishing and polishing perpendicular to the direction of the rays, one could obtain Debye–Scherrer diagrams. In addition, radiographs were also obtained after various degrees of etching. As a result of these experiments the following became clear. For example, in the case of zinc an undamaged initial crystal gives only a single Laue spot. After grinding on a slowly rotating wheel, to a depth of approximately \(0.5\) mm, the Laue spot disappears and in its place two diffuse rings appear, belonging to planes \((10\text{–}15)\) and \((12\text{–}32)\). At certain points on these rings there are distinct maxima. Subsequent polishing does not change the character of the reflected diagrams. But after etching this polished ground surface, by about \(0.015\) mm, the rings disappear and, in the places of the maxima indicated above, strong monochromatic spots appear which, apparently, belong to layer lines (Schichtlinien). With further etching the latter become very small, and only gradually does the former Laue pattern begin to reappear. Boas and Schmid interpret all these diagrams as follows. On the surface of the ground layer there is a fine-grained layer in which, judging by the Debye–Scherrer rings, recrystallization has occurred. Behind this layer, whose thickness does not exceed \(0.015\) mm, there follows a zone in which regions of the lattice, displaced and deflected, are present simultaneously; here lies the basis for the existence of a deformation whose magnitude and direction along the layer lines can be determined. The thickness of this second zone, situated between the fine-grained layer and the undamaged crystal, is approximately \(0.16\) mm. In general, in contrast to the abrupt transition from the upper layer to the intermediate one, the transition to the main structure proceeds gradually.

The results obtained with other metals do not differ especially from these results. In almost all cases the thickness of the deformed layer is approximately ten times greater than the thickness of the fine-crystalline upper layer, which varies from \(0.005\) to \(0.030\) mm.

In conclusion to all these somewhat discrepant results, one should point to the work of Hamburger\(^7\) in Holland. On the basis of the results obtained by him and others, he believes that polishing is a very subtle process of abrasion, accompanied by smoothing of the surface relief (the formation of platelets or scales), and that the first step in this direction is already made during comparatively rough grinding, when plastic deformation of the surface layers also takes place. Careful polishing leads to dragging along the surface the smallest particles, whose size, depending on the hardness of the metal, extends from 3 to 30 atoms. Each time, in the region of such magnitudes, the properties of aggregates show extraordinary variety depending on the size of the particles. In any case, we are dealing here with a process that is not molecular in character. Nevertheless, some of the particles initially torn away fall into pits on the surface and are held there by cohesive forces. Naturally, therefore, the mechanism of polishing is always associated with the formation of a surface film, whose properties always differ from the properties of the “grain.” The existence of plastic deformation is confirmed by the fact that even such a hard body as quartz, when rubbed in one direction against a piece of cotton wool, begins to show an increase in surface anisotropy, which extends to a considerably greater depth than the molecular dimensions. Such anisotropy, of course, disappears upon heating.

Thus carefully polished surface films of crystals consist of platelets of aligned crystals of submicroscopic dimensions. These films usually have a transition to the grain through an intermediate layer of comparatively great thickness, where the structure shows a continuous course of change. In addition, the character of polished surfaces may prove very complex owing to their physical instability, the adsorption of foreign substances, [[unclear: word cut off at page bottom]]

…of the action of the polishing material (crocus, etc.) and of the existence of more or less distorted contact boundaries between the finest particles.

Literature

1. Davisson, “Waves or Electrons?” U. F. N., 8, 483, 1928; Mark and Wierl, “New Data on Electron Diffraction,” U. F. N., 10, 737, 1930; N. A. Shishakov, “Cathode Rays and Structural Analysis,” Socialist Reconstruction and Science, issue 1, p. 131, 1931.

2. G. P. Thomson, Proc. Roy. Soc., ser. A 128, 550, 1930. Nature, 129, 81, Jan. 16, 1932.

3. R. C. French, Nature, 129, 169, Jan. 30, 1932.

4. F. Kirchner, Nature, 129, 545, Apr. 9, 1932.

5. G. P. Thomson, C. A. Murison, N. Stuart, Nature, 129, 545, Apr. 9, 1932.

6. W. Boas and E. Schmid, Naturwiss., 20, 416, May 27, 1932.

7. L. Hamburger, Nature, 130, 435, Sept. 17, 1932.

N. Shishakov