Films of metals and inorganic compounds, deposited upon organic supporting foils by vaporization in high vacuum, have been investigated by electron diffraction by the transmission method. Mean thickness of each of these films has been calculated from the total amount of vaporized material, distance of the film from the source, and calibration of the apparatus based upon weighing the material making up a known area of film. Films of the following substances have been studied: Au, CsI, ${\mathrm{Sb}}_2$${\mathrm{O}}_3$, Pd, Cu, CuCl, Ni, Ca${\mathrm{F}}_2$, Al, Mg, NaF and Be. Most of these films have been extremely thin, two-thirds of them less than 30A and many less than a single layer of atoms or molecules. The masking effect of the superposed diffraction pattern from the supporting foil is so slight that satisfactory patterns can be obtained from films of mean thickness as small as one layer of atoms if their atomic number is as large or larger than that of copper (29); in the case of beryllium (4) films as thick as 50A are required. Diffraction patterns obtained are characteristic of three-dimensional crystals. The ionic compounds studied produce films made up of rather large crystals, 100A or more on a side, even in the very thinnest films. Metal films are, in general, made up of smaller crystals than films of ionic compounds. The observations prove that atoms and molecules are able to move over the surface of the supporting foil, and that they are drawn together into crystals of considerable size. Crystals in metal films have been randomly oriented, except for slight orientation in aluminum films; crystals of some ionic compounds have been found to possess a strong preference for some particular orientation, and others have been randomly oriented. No correlation has been observed between average shape of crystal and degree of orientation. Intensity anomalies have been observed in diffraction patterns from calcium fluoride and cuprous chloride. Applications of the method of this investigation to study of corrosion and chemical change are suggested.

• Received 8 May 1939

DOI:https://doi.org/10.1103/PhysRev.56.58