Contrast in High Resolution Images of Crystals

Abstract

An attempt is made to understand experimentally observed high resolution electron microscope images. Iijima has obtained images that are good representations of the crystal potentials for a resolution of about 3Å using a defocus corresponding to the “optimum defocus condition”. This is to be expected for a thin phase object. However, Iijima's images are obtained from crystals having thicknesses of 100Å and greater for which the maximum phase change is more than 10 radians, and so the thin phase object approximation is not expected to be valid. These images must therefore be explained by other means.Many beam calculations have been made on thick crystals using a trial crystals structure as shown in Fig. 1. A metal atom sits at the centre of each octahedral site with oxygen atoms at the corners. This structure contains blocks of 3 x 3 octahedra of the type found in disordered samples of Ti2Nb10O29 as shown in Fig. 2.