Surface Science 162 (1985) 965-970 965 North-Holland, Amsterdam SLOW-ELECTRON-ENERGY-LOSS SPECTRA AND NON-DIPOLE TRANSITIONS IN NICKEL J. CAZAUX and A.G. NASSIOPOULOS Laboratotre de Spectroscopie Electrontque, Unwersttb de Reims, UER Sctences, BP 347, F-51062 Reims, France Received 1 April 1985; accepted for publication 29 April 1985 Electron-energy-loss spectroscopy in the reflection mode has been used to study the surface of polycrystalline nickel. Spectra corresponding to M2. 3 and M I core losses are presented, obtained with impact energies ranging from 0.5 to 2.5 keV. The M I core loss is mainly attributed to dipole-forbidden transitions. The ratio of the intensities of the M~ to M2. 3 core losses is found to decrease when the impact energy E o is increased. This is attributed to the decrease of the momentum transfer of the inelastic event at high primary energies which, in turn, causes a decrease of the non-dipole transition probability relative to the probability of dipole transitions. The analytical variation of /(M I )//(M2.3) is suggested. I. Introduction Slow-electron-energy-loss spectroscopy (SEELS) in the reflection mode has recently emerged as a powerful tool for surface microanalysis [1] as well as for surface physics and chemistry [2]. From the point of view of surface physics and chemistry this technique has been proved very useful, beause a SEELS spectrum presents very rich features (core + plasmon losses, near-edge struc- tures, oscillations of the EXAFS type). It has been applied recently on Ni at various primary-beam energies in order to deduce the nearest-neighbour dis- tances from the oscillatory part of the spectrum [3,4]. Also in ref. [5] the shift in the energy postion of the M2. 3 core edges in Ni and Ti when the impact energy E 0 is decreased, is presented. In this paper we mainly study the variation of the relative intensities of the M t and M2. 3 core losses of nickel when the impact energy E u is changed. This variation is explained by the fact that the most probable empty final states in Ni have a "d" character, so that transitions from initial M2. 3 states (2p) toward these empty d states are dipole transitions; but when the initial state is of the s type (M~ edge) the correspond- ing transitions are non-dipole transitions. The band structure calculations of Ni confirm the well-known high density of d states just above the Fermi level but also shows less intense and less localised s and p partial density of states [61. 0039-6028/85/$03.30 '~?,Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)