Leed, auger, and electron energy loss studies of Ni epitaxial growth on Cu(100)

Abstract

Abstract Nickel was epitaxially deposited onto a clean, flat Cu(100) surface. Low energy electron diffraction I ( E ) curves were recorded for 0.6, 1.1, and 2.7 monolayer (ML) Ni coverage. Multilayer relaxation was considered in theoretical calculations, which were compared with experiment by means of the R | ΔE | factor. The estimated relaxations of the first and second interlayer spacings are estimated to be − 2% and + 1.5% for clean Cu(100), − 2% and − 1.5% for 1 ML Ni coverage, relative to the bulk Cu interlayer spacing of 1.81 A, and −1% and 0% for 3 ML Ni coverage, relative to the bulk Ni spacing of 1.76 A. Decreasing the surface Debye temperature of the Ni layer to 268 K from the bulk value of 440 K improves the agreement between theory and experiment. The optimum inner potential values are − 9 and − 10 eV for clean Cu(100) and Ni on Cu(100), respectively. Auger electron spectroscopy was used to determine the thickness of the Ni films, and LEED indicates layer-by-layer growth until about 4 layers, when the LEED spots begin to spread, indicating island formation. Electron energy loss spectra were obtained with primary electron energies of 150 and 300 eV. The 3p core ionization transition was clearly observed after 0.5 ML Ni coverage. Peaks at 3.8 and 7.5 eV for clean Cu are ascribed to interband transitions, and shift to higher energy with Ni coverage. Peaks at 10 and 16 eV for clean Cu (ascribed to an interband transition and a surface plasmon, respectively) disappear with Ni coverage. Bulk plasmon peaks at 19 and 27 eV remain unshifted with Ni coverage. The effect of 0.9 and 1.3 ML Ni coverage of Cu(100) on the chemisorption of Co and oxygen was also studied. The behavior of the surface towards oxygen chemisorption was similar to that of the pure Ni surface. For a large exposure of oxygen (50 L and more) the EEL and Auger spectra are very similar to those observed for NiO. In the case of CO, for submonolayer Ni coverage, the surface shows a more Cu-like behavior, while for larger Ni coverage (a monolayer and more) there is a great similarity with the behavior of the pure Ni(100) surface.