Prospects for high resolution electron microscopy

Abstract

The resolving power of the electron microscope which is required for a direct read out of atomic structures in real specimens is considerably smaller than the average interatomic spacing. This is due to the overlapping of the more or less defocused projection images of those atoms which do not lie in the plane of optimum focus. Thus, with the best present-day microscopes, which have a resolving power of about 0.2 nm and an accelerating voltage of about 100 kV, atom positions can be observed only in monolayer specimens. If more than one layer is present the intensity distribution is only a measure of the projection image density. Because of the objective aperture being directly proportional to the electron wavelength and inversely proportional to the resolving power the depth of focus strongly depends on both these quantities. Approximation formulae for limit of resolution imaging conditions have been derived. They show that in order to obtain a direct read out of the atomic structure of a specimen of only 2 to 3 nm thickness a mega volt electron microscope should be used, and the required resolving power is as low as a fraction of the average atomic spacing. If the resolving power is improved for microscopes in the 100 kV range the depth of focus will decrease so rapidly that the optimum specimen thickness finally turns out to be less than 1 nm. A small amount of chromatic aberration can produce an effective enlargement of the depth of focus for accelerating voltages U < 500 kV, but this gain is purchased by an order of magnitude increase of the background intensity. This in turn deteriorates the signal: noise ratio, and the electron dose required to transmit the image information is considerably larger with U < 500 kV voltages than in the megavolt range, leading to correspondingly larger radiation damage.