Monte Carlo strategies for simulations of electron backscattering from surfaces

Reliable algorithms for simulation of electron trajectories in solids are needed for quantification of Auger-electron spectroscopy, particularly for the determination of the backscattering factor (BF). The computational schemes for this purpose should be universal, i.e. applicable to any solid for energies down to 50 eV. In addition, the algorithms should ideally require simple input parameters, e.g. the stoichiometry of the solid and the experimental configuration. Previous Monte Carlo simulations have often been based on the continuous slowing down approximation (CSDA), a relatively simple and convenient approach. However, this approach requires information on the electron stopping power for a given solid. General expressions for the stopping power are briefly reviewed and discussed. We report results of BF calculations for Cu M3VV, Ag M5VV, and Au N67VV Auger transitions from Monte Carlo simulations based on the CSDA in which stopping powers from optical data and from Tougaard's two-parameter universal inelastic-scattering cross section were employed. The resulting BFs, for primary energies between 500 eV and 10 keV, differed by up to 12% (at an energy of 10 keV). BFs from the Monte Carlo algorithm based on the CSDA were compared with similar results from a more sophisticated code involving simulations of individual elastic- and inelastic-scattering events. Both algorithms yielded equivalent BFs for the same Auger transitions with primary energies exceeding 1000 eV; at lower energies, deviations up to 10% occurred because of an approximation in the evaluation of an integral. Copyright (C) 2005 John Wiley & Sons, Ltd.