Photoemission from Clean and Cesium-Covered Nickel Surfaces

Publisher Website
Google Scholar
Abstract
Photoemission energy distributions (PED) have been measured for clean and cesium-covered single-crystal Ni(111) surfaces. The surface condition was monitored during the experiments with a low-energy electron-diffraction (LEED) apparatus. Photoemission results for the clean surface are similar to those previously reported for polycrystalline material, except that minor structure resulting from direct transitions was observed, and a much higher work function of 5.42\pm\else\textpm\fi{}0.04 eV was obtained for the (111) surface than has been reported for polycrystalline surfaces. The relative intensity of the anomalous peak at -4.5 eV with respect to the upper edge of the PED was found to vary with surface condition. With monolayer cesium coverage, very strong peaks are observed in the PED at -4.0 and -5.4 eV. These peaks appear when cesium is applied in excess of that required to lower the work function to its minimum value. A LEED pattern indicating ordering over twice the Ni lattice spacing is correlated with the cesium-related peaks. We argue that a fixed energy-loss mechanism is operating that causes structure to be displaced in the external PED by the amount of the fixed energy loss. We present evidence that this loss does not occur during transport of a photoexcited electron across the surface region, but that it accompanies the optical absorption process. It is suggested that peaks at -4.5 eV on clean Ni and at -4.0 and -5.4 eV on cesium-covered Ni result from the simultaneous excitation of interband transitions and a surface plasmon of 4.0-4.5 eV. Further experiments are suggested to check this hypothesis. For cesium coverages in excess of about three monolayers, a shoulder appears in the PED at -2.8 eV that we interpret as resulting from the excitation of the bulk plasmon in cesium.