Studies in contact potentials the condensation of potassium and sodium on tungsten

Abstract

Recent work on the adsorption of gases has emphasized both the importance and the difficulty of ensuring the cleanliness of metal surfaces. Methods which involve the "flashing" of a filament of the metal thus present several advantages over the more usual experimental arrangements. With such a filament several physical methods may be employed for investigating the effects of adsorbed films thereon, and a study of the accommodation coefficient (J. K. Roberts 1935) has already yielded much new information. Of the three electrical methods of examination of the film-covered surface, the thermionic work function, the photoelectric work function, and the contact potential, the last is to be preferred, for many films are mobile even at low temperatures, thus precluding use of the thermionic method. In others the photoelectric threshold is so far in the ultra-violet as to render its measurement difficult. From the contact potential the dipole moment of the chemi-adsorbed complex may be obtained. Its value and any variation with the packing density of the film-forming material are required both as a basis for evaluating the true form of the adsorption isotherm as well as for obtaining further insight into the mechanism of chemical reactivity at such surfaces. The measurement of contact potentials by studying the infra-saturation currents in a thermionic tube originated in the work of Richardson and Robertson (1922). Oatley (1936) (cf. also Mönch 1928, 1930) obtained values for the contact potential difference of molybdenum, zinc and platinum against tungsten in quite satisfactory agreement with the best thermionic determinations of the work function; but, in general, the results obtained by contact-potential measurements are not always in satisfactory agreement with thermionic and photoelectric measurements, particularly when dealing with films. Although part of the discrepancy may be due, as Compton and Langmuir (1930) have pointed out, to the fact that in in-homogeneous surfaces the contact-potential method measures the mean work function of the surface, while the other two methods measure the minimum work function, it would be well to survey the experimental technique to choose a procedure which will avoid the more usual difficulties.