Comments on „HRTEM-Bildkontrastsimulation von Strukturen mit Punktdefekten in speziellen Lagen“

Publisher Website
Google Scholar
Abstract
We wish to point out that papers by Rahman (1988) and Rahman and Weichert (1990) contain a basic error in concept. In these papers, the authors seek to shorten the amount of computation necessary for a simulation of a high-resolution transmission electron microscope (HRTEM) image of a structure containing a defect, using as an example an oxygen vacancy in mullite. Unfortunately, the method used leads to errors in the simulated image. The usual method of computing an image of a non-periodic structure is to construct a “supercell” consisting of a number of perfect unit cells surrounding the cell(s) containing the defect (e. g. Iijima and O'Keefe, 1978). The computation is then carried out using the supercell as the new unit cell for the defect structure. Note that the new defect structure is periodic with the periodicity of the chosen supercell, and thus the number, n , of original unit cells making up the supercell must be chosen large enough not only to sample diffraction space sufficiently finely, but also to keep the images of adjacent defects from interfering. Such interferences can arise if electrons scattered from non-perfect regions near to the defect in one supercell contribute to the image of the defect in an adjacent supercell. In addition, if a smooth “periodic continuation” of the structure at the supercell boundary is not possible, the area immediately surrounding the defect of interest must be far enough away from the unsmooth boundary to avoid interferences between the image of the defect and that of the (probably nonphysical) boundary.

This publication has 2 references indexed in Scilit: