Validity of Monte Carlo calculations for the interpretation of x-ray profiles in stem

Recently developed techniques involving the combination of fine electron probes and detectors suitable for the collection of data from various signals emitted by specimens promise microchemical analysis on a small scale. One such combination is that of scanning transmission electron microscopy (STEM) and energy-dispersive x-ray spectroscopy (EDS). A very important question is the limit of spatial resolution of this combination of techniques. This paper deals with questions about the spatial resolution of such an analysis. A fine electron probe at the top (electron entrance) surface of a foil broadens as the elctrons go through the foil. Elastic scattering accounts for most of this broadening. Thus, although the beam is initially small (25 A), beam spreading broadens the lateral extent of the beam. The implication is then that the spatial resolution of the analysis may be considerably coarsened. For this reason, a method is needed of calculating beam spreading. One of the simplest of such calculations is the single scattering model proposed by Goldstein et al. It is expected that such a model best describes thin foils. For thicker samples, a more sophisticated calculation would be desirable. Such a model would have to take plural scattering into account. A very flexible approach to the problem of simulating plural scattering is that afforded by Monte Carlo techniques. In this model, many electron trajectories are simulated via computer calculations. Decisions such as the electron path length between atomic collisions and the change of electron direction following a collision are influenced by values generated by a pseudo-random number generator