Sub Angstrom imaging of dislocation core structures: How well areexperiments comparable with theory?

During the past 50 years Transmission Electron Microscopy (TEM) has evolved from an imaging tool to a quantitative method that approaches the ultimate goal of understanding the atomic structure of materials atom by atom in three dimensions both experimentally and theoretically. Today's TEM abilities are tested in the special case of a Ga terminated 30 degree partial dislocation in GaAs:Be where it is shown that a combination of high-resolution phase contrast imaging, Scanning TEM, and local Electron Energy Loss Spectroscopy allows for a complete analysis of dislocation cores and associated stacking faults. We find that it is already possible to locate atom column positions with picometer precision in directly interpretable images of the projected crystal structure and that chemically different elements can already be identified together with their local electronic structure. In terms of theory, the experimental results can be quantitatively compared with ab initio electronic structure total energy calculations. By combining elasticity theory methods with atomic theory an equivalent crystal volume can be addressed. Therefore, it is already feasible to merge experiments and theory on a picometer length scale. While current experiments require the utilization of different, specialized instruments it is foreseeable that the rapid improvement of electron optical elements will soon generate a next generation of microscopes with the ability to image and analyze single atoms in one instrument with deep sub Angstrom spatial resolution and an energy resolution better than 100 meV