Abstract
A new method of experimental attack upon the problem of the structural dependence of the work functions of the metals is described. The method utilizes the results of recent electron diffraction studies which show that many cubic lattice metals when thermally vaporized and condensed on heated cleavages of rocksalt form films which are in effect single crystals with a (100) plane parallel to NaCl (100). With silver as a typical metal of this class, there has been developed a technique of preparing an Ag(rocksalt) film in the measuring tube itself and of then determining its contact difference of potential against (1) polycrystalline silver films deposited on glass at room temperature, and (2) a reference metal of known work function, barium. The silver was fractionally distilled and only the middle fractions deposited on the glass and rocksalt targets, the barium fractionally and multiply distilled and its discarded fractions used to getter the tubes in which the measurements were made. The contact p.d. between Ag(rocksalt) and Ag(glass) has been found to range from 0.12 v in a preliminary measurement to 0.32±0.03 v in final measurements made after the technique of preparing the Ag(rocksalt) films had been perfected. The contact p.d. for Ag(glass) —Ba is found to be 1.95 v, in close agreement with our earlier results at liquid-air temperature. Work function values computed from these measurements and the work function of Ba, 2.52 ev, are 4.47 ev for Ag(glass) and 4.79 ev for Ag(rocksalt). Since two independent electron diffraction investigations, both especially complete for silver, indicate that Ag(rocksalt) films formed under the conditions of the present experiments invariably approximate closely to the ideal (100) structure, and since the contact potential method measures the average work function of a surface without weighting vestigial regions of imperfect symmetry, our Ag(rocksalt) work function value is identified tentatively with the work function of Ag(100). Our Ag(100) value is in good agreement with Farnsworth and Winch's recent photoelectric measurements on a massive single crystal but Winch's earlier value of 4.74 ev for the work function of a heat-treated silver wire is probably characteristic of the crystal faces of preferred orientation developed by thermal etching rather than of randomly oriented polycrystalline silver. The method developed in the present work for measuring Ag(100) appears to be directly applicable to the (100) orientations of Ni, Cu, Au, Pd and Al.