The Impact of Energy Filtering on Fluctuation Electron Microscopy

Abstract Fluctuation electron microscopy (FEM) analyzes intensity fluctuations within diffraction patterns in order to draw conclusions regarding the structure of amorphous materials by calculating the normalized variance V(k,R). Ideally, such experiments only evaluate elastically diffracted electrons. However, an undesired inelastic background intensity is always present and degrades the FEM data. Energy filtered FEM experiments were performed on amorphous germanium created by self-ion implantation. FEM data were acquired in a transmission electron microscope at 60 and 300 kV with different electron doses as well as varying energy filter slit widths at two sample thicknesses. Generally, the measurements reveal that energy filtering greatly improves FEM data at both beam energies and sample thicknesses by removing a certain amount of the inelastic background intensity in the diffraction patterns. The narrower the energy filter, the larger the normalized variance. This brings energy filtered FEM data closer to the normalized variance determined by simulations under idealized conditions. Furthermore, preliminary results indicate that the medium range order length scale extracted from the pair-persistence analysis used in FEM is strongly affected by energy filtering.