Atomic Resolution Electron Microscopes are now producing useful results in many fields of science and technology. This success was obtained not only by the improvement of resolution of TEM but also through the developments of theories and experiments of diffraction crystallography, image formation and recording technique over the past 44 years. Boersch (1946-47) discussed the image of atoms and crystals by the phase object approximation; the image contrast is due to the phase shift of electron waves passing through them. Scherzer (1949) discussed the effect of the phase shift by the electron lens and proposed the phase contrast transfer function. He pointed out that the key to observe images of single atoms is by contrast enhancement, which might be possible by dark field images if resolution is improved. This proposal was attempted by improving the resolution using tilted-beam dark fields (Th atoms, Hg & Pt atoms 1971). Crystal lattice fringes were observed by Menter in 1956 who discussed the contrast by kinematical theory. The interpretation of lattice images by dynamical theories were carried out by Hashimoto et al. (Bethe theory) and Cowley (Multi-slice theory) in 1958-60 and noted the correct position of atoms is in neither black fringe regions nor white ones when the images are taken in the Bragg reflecting position. However Miyake et al in 1964, showed that black or white contrast peaks appear at atom positions if the crystal is in symmetry position. Two dimensional lattice structure images were first photographed in 1970-71, which stimulated strongly the application of electron microscopy to materials science, crystallography and engineering at many laboratories such as the National Center at Berkeley. For the identification of correct position of atoms, two types of image contrast calculation have been proposed (e.g. multi-slice 1972, Bethe method 1975). The partially coherent theory originally developed in optics was introduced into the contrast calculation in 1979-80. Around this time, many observations of defect structures have been carried out, some of which are shown in Table 1. In situ observation of moving atoms and atom columns in molecules and crystals have been carried out in 1978 by using a TV system, which enable us to see the transition phenomena with a speed of 1/30 or 1/60 seconds. More recently atomic imaging at 1.8 A-1.6 A played an important role in the structure research of new materials and phenomena such as superconducting materials (metallic A15, 1989, ceramic high Tc, 1988), metal ceramic interface (Nb/Al2O3 1984), superstructure (GaAs/AlAs/GaAs 1985) quasi crystals (1987) etc. Many subsequent observations are presented in this congress. In spite of these developments, there are still some problems to be solved, e.g. imaging of atoms in the correct positions and the identification of different kinds of atoms in the materials especially with unknown crystal structure.
Read full abstract