Search

Conditions for the intensity enhancement of a Bragg reflection are investigated through convergent-beam (CB) reflection electron diffraction experiments. This intensity enhancement is of great interest for reflection electron microscopy of surfaces. Comparison of CB reflection diffraction patterns with CB transmission diffraction patterns shows a very similar enhancement which can be explained by Bloch waves. The observed surface diffraction parabolae are closely related to Kikuchi envelopes. The intensity enhancement can be interpreted as a channelling phenomenon.

Interference fringes obtained between overlapping diffraction orders in the Kossel-Mollenstedt pattern from a pair of crossed single crystals have been used to demonstrate the degree of coherence across the incident illumination. The experiment also demonstrates the principle of formation of this type of image in the transmission scanning microscope.

For the examination of extended specimens such as ribbons of serial sections, thin metal films and some replicas in the electron microscope, it is desirable that the field of view should not be obstructed by grid bars. Suitable supporting substrates must satisfy five essential conditions. a) They should be easily prepared and mounted over holes of at least 0.5 mm diameter. b) Under intense electron bombardment they should not contribute to specimen shrinkage or movement. c) They must not break under the strains imposed in picking up floating specimens and subsequently drying them. d) Their area density should be very low. e) They must be clean and smooth.

The work on the observation of moire patterns from overlapping metal films carried out by the authors (1, 2) has already been described in part (3). In this contribution a few of the more important aspects in relation to the application of the techniques to studying dislocations are considered in more detail.

The fringe patterns which occur in electron micrographs of overlapping thin crystals were first observed almost simultaneously but independently in 1951 by Mitsubishi, Nagasaki and Uyeda (1), by Hillier (2) and by one of us (J. L. F.) (2) and since then have been the subject of many investigations. In 1956 the present authors (3) showed that in the case of pairs of sufficiently thin crystals rotationally disoriented through a small angle e it should be possible to observe a complex of fringes, i. e., a moire pattern, which is almost identical with the Patterson distribution of structure analysis. This paper describes the physical processes involved in the formation of moire patterns and the possibility of observing patterns more directly related to the structure of the crystals than the Patterson function.

It has been demonstrated that specimen contamination in convergent-beam diffraction operation can be prevented by maintaining the specimen temperature between -110 °C and -165 °C, without the use of especially high or clean vacuum conditions. At these temperatures, surface migration of molecules causing contamination is evidently inhibited. Precautions to prevent deposition from the vapour phase both before and after cooling are also required.

Stereo imaging is not new to electron microscopy. Von Ardenne, who first published transmission pairs nearly forty hears ago, himself refers to a patent application by Ruska in 1934. In the early days of the electron microscope von Ardenne employed a pair of magnetic lenses to view untilted specimens but soon opted for the now standard technique of tilting the specimen with respect to the beam.In the shadow electron microscope stereo images can, of course, be obtained by tilting the specimen between micrographs. This obvious method suffers from the disadvantage that the magnification is very sensitive to small changes in specimen height which accompany tilting in the less sophisticated stages and it is also time consuming. A more convenient method is provided by horizontally displacing the specimen between micrographs. The specimen is not tilted and the technique is both simple and rapid, stereo pairs being obtained in less than thirty seconds.

Oblique illumination, although accompanied by distortion, provides a simple method of obtaining stereoscopic images in a shadow electron microscope. It is not necessary to tilt the specimen.

A method of performing selected-area diffraction in the shadow electron microscope is described. The procedure is simple and the diffracting region is accurately located.

An accurate and reproducible determination of the magnification in point-projection or electron-shadow microscopy is obtainable by using as a reference the objective lens current required for infinite magnification. The magnification at any other objective lens current is then determined as a function of the difference between that current and the current at infinite magnification. It is generally a simple matter to choose a reproducible current difference. A replica of a crossed grating with line spacing of 0.463 mu m was used as a test specimen in experiments with a converted Siemens Elmiskop 1. The relationship, magnification=8720/n, where 12.8 n mA is the current difference, was found to hold to better than 3%.