Dependence of experimentally determined inelastic mean free paths of electrons on the measurement geometry

The inelastic mean free path of electrons (IMFP) can be derived from the ratio of elastic peak intensities measured for the studied sample and a reference material. This method is frequently described with the acronym EPES (elastic peak electron spectroscopy). The corresponding calculations should be based on a realistic theoretical model of electron transport in solids. If such a model is available and if the elastically backscattered intensities are measured correctly, the resulting IMFP values should be independent of the experimental configuration used to measure the intensity ratios. In this paper, the role of experimental geometry is systematically checked for elements with a wide range of atomic numbers: Si, Cu, Ag, and Au. A pronounced dependence of the derived IMFPs on emission angle, alpha, has been observed in the range 5 degrees less than or equal to alpha less than or equal to 20 degrees. This dependence becomes noticeably weaker for larger emission angles, 25 degrees less than or equal to alpha less than or equal to 45 degrees. The scatter of the IMFPs corresponding to the range 25 degrees less than or equal to alpha less than or equal to 45 degrees seems to decrease with increasing energy, especially above 200 eV. The observed effects are discussed in terms of the validity of the theoretical model and possible instrumental problems. An effective Monte Carlo strategy has been proposed for the needed EPES calculations which made the present computations manageable since calculations of over 600 calibration curves were necessary for processing the experimental data. (C) 1998 Elsevier Science B.V. All rights reserved.