Search

Surface-diffusion process of atoms is one of the important factors for understanding various types of surface-related phenomena such as thin-film growth and catalysis reaction[l]. Development of field ion microscope (FIM) was a great breakthrough for the study of the process in atomic level. The diffusion process of Rh and Re atoms on tungsten surfaces was quantitatively studied, and the barrier potential (Ed) and the pre-exponential factor (D0) were successfully determined[2]. Recent development of scanning tunneling microscope (STM) has also enabled us to measure directly the moving of Pb and Sb-atoms on Ge and Si (111) surfaces[3, 4]. The time-resolution of these methods is, however, not sufficient to trace the diffusion process in detail.In previous papers, we reported the first experiment of tracing the diffusion process of tungsten(W)-atoms on terrace-regions of MgO (001) surfaces in the time-resolution of 1/60 s.using dynamic high resolution electron microscopy[5,6]. The values Ed=0.18 eV and Do=1.5x10−14 cm2/s were successfully determined from the video-tape images. In the present study, we have observed diffusing atoms along the surface-steps and studied the phenomena quantitatively.

In a conventional alignment for high-resolution electron microscopy, the specimen point imaged at the viewing-screen center is made dispersion-free against a voltage fluctuation by adjusting the incident beam direction using the beam deflector. For high-resolution works the voltage-center alignment is important, since this alignment reduces the chromatic aberration. On the other hand, the coma-free alignment is also indispensable for high-resolution electron microscopy. This is because even a small misalignment of the incident beam direction induces wave aberrations and affects the appearance of high resolution electron micrographs. Some alignment procedures which cancel out the coma by changing the incident beam direction have been proposed. Most recently, the effect of a three-fold astigmatism on the coma-free alignment has been revealed, and new algorithms of coma-free alignment have been proposed.However, the voltage-center and the coma-free alignments as well as the current-center alignment in general do not coincide to each other because of beam deflection due to a leakage field within the objective lens, even if the main magnetic-field of the objective lens is rotationally symmetric. Since all the proposed procedures for the coma-free alignment also use the same beam deflector above the objective lens that is used for the voltage-center alignment, the coma-free alignment is only attained at the sacrifice of the voltage-center alignment.

Any attempt to study the relationship between the electronic and atomic struc ture of isolated defects in semiconductors by high resolution electron microscopy must deal with the following difficulties: (i) The limited point reso lution of modern TEM instruments, which has fallen by about 1Å from 3.8Å to 2.8Å (with tilt) over the last decade. This is still not sufficient to resolve the individual atomic columns in any semiconductor. (ii) The fundamental difficulties in obtaining both high spatial resolution structural information and high energy resolution spectroscopic data from the same, isolated, defect. (iii) The considerable difficulties in extracting chemical, or atomic number information from electron scattering and imaging experiments with high spatial resolution. Among other problems, the separation of composition variation effects from those of thickness is an important problem. Some of our recent approaches to these problems are outlined below.

The technique of surrounding virus particles with a neutralised electron dense stain was described at the Fourth International Congress on Electron Microscopy, Berlin 1958 (see Home & Brenner, 1960, p. 625). For many years the negative staining technique in one form or another, has been applied to a wide range of biological materials. However, the full potential of the method has only recently been explored following the development and applications of optical diffraction and computer image analytical techniques to electron micrographs (cf. De Hosier & Klug, 1968; Markham 1968; Crowther et al., 1970; Home & Markham, 1973; Klug & Berger, 1974; Crowther & Klug, 1975). These image processing procedures have allowed a more precise and quantitative approach to be made concerning the interpretation, measurement and reconstruction of repeating features in certain biological systems.

Elucidation of the 3-D molecular structure of actin has been a challenge to protein crystallo-graphers and electron microscopists ever since the pivotal role of this 43-kDa protein in muscle contraction - and later in cellular motility and cytoskeletal architecture - has been recognized. In its functional form actin is polymerized into filaments (F-actin) which are made of two linear strands of actin subunits helically twisted around each other with a pitch of about 72 nm and an axial stagger of 2.75 nm. In most in vivo situations F-actin is complexed with stoichiometric amounts of accessory proteins - e.g. tropomyosin, troponin, crosslinking or capping proteins - via specific binding sites the actin molecule has for these proteins. While ultimately one would like to reveal a near-atomic resolution model of the actin filament, as a first step towards this goal one would be quite happy to obtain a high resolution electron density map of the actin molecule which, in turn, requires the availability of large and well ordered 3-D crystals.

Diffraction patterns of crystals having long period superlattices are described by their matrix lattice reflections, M = ha* + kb* + lc*, and a set of satellite reflections S appearing around each matrix reflection, which are characterized by S = m1a* + m2b* + m3c*. For usual superstructures the m's become rational numbers but for certain superstructures they are irrational. The latter are called the incommensurate superstructures. Crystal structure analysis by the conventional diffraction methods is based on perfectly periodic structures and generally cannot be applied to the crystals having the irrational superstructure periodicities. This paper deals with the study of such crystals using the high resolution lattice imaging method and the types of information obtained are discussed. It was intended that the study not only supplement the information on statistically averaged structures obtained from diffraction data but also provide further understanding of the incommensurate structures.

High resolution electron microscopy images of crystals are obtained by making some beams, diffracted by the crystal, interfere with the transmitted beam. For some experimental conditions, to be carefully defined for each specimen, the image can be shown to represent atomic columns; thus information in the atomic range is obtained, though it must be stressed that interpretation of these structure images must be done carefully as artefacts can give effects not linked with the crystal structure. In the last 3 years, this technique has been applied to the study of surfaces where it has given some striking results.

The recent proliferation of the so-called intermediate-voltage HREMs has led to a vast expansion in the ranges of materials which are nowadays being examined on the atomic scale. Impressive-looking micrographs containing detail as fine as 1A are comparatively easy to obtain. However, relating this image information in a quantitative manner to particular features of the object can be tedious and time-consuming even when the imaging conditions (eg. objective lens focus, illumination angle, focal spread) and the specimen parameters (eg. thickness, orientation, composition) are well-known. The presence of residual astigmatism and/or incident beam misalignment leads to spurious image features which can seriously compromise this process of interpretation, but it is difficult, even for skilled operators, to adjust these variables with sufficient precision that their effects are negligible.

The study of ordered alloys by means of conventional X-ray diffraction methods is hampered by the fact that even single crystals of such alloys are invariably fragmented in symmetry related domains; unscrambling of such combined patterns into mono domain patterns is usually not straightforward. In electron diffraction it is on the other hand often possible to select a single domain on the image and obtain a mono-domain diffraction pattern.The use of suitable high resolution imaging techniques allows often to derive a trial structure, for which the image and the diffraction pattern can be calculated and compared with the observations. It is our intention to demonstrate how the combined use of both techniques allows the determination of crystal structures of rather complicated alloys in the gold-manganese and gold-magnesium systems.

Image processing can be considered as an attempt to aid image interpretation and to improve image quality by using additional information not contained in the image, or to make optimal use of redundant information contained in the image. Since high resolution electron micrographs contain a very high amount of noise from substrate, electron statistics and photographic grain, the separation of the signal from the noise part of the image is one of the most important problems. An operation that produces an output image with a higher signal-to-noise ratio than the input image may appear as witchcraft but is in fact feasible through clever use of a-priori information such as the knowledge of object and noise statistics.

The era of atomic-resolution electron microscopy has finally arrived. In virtually all inorganic materials, including oxides, metals, semiconductors and ceramics, it is possible to image individual atomic columns in low-index zone-axis projections. A whole host of important materials’ problems involving defects and departures from nonstoichiometry on the atomic scale are waiting to be tackled by the new generation of intermediate voltage (300-400keV) electron microscopes. In this review, some existing problems and limitations associated with imaging inorganic materials are briefly discussed. The more immediate problems encountered with organic and biological materials are considered elsewhere.Microscope resolution. It is less than a decade since the state-of-the-art, commercially available TEM was a 200kV instrument with a spherical aberration coefficient of 1.2mm, and an interpretable resolution limit (ie. first zero crossover of the contrast transfer function) of 2.5A.

The strength of in situ electron microscopy lies in its ability to observe directly material changes which are pertinent to bulk processes. The most rigorous experiments employ a purpose-built specimen holder to simulate specific testing conditions (e.g. heating, cooling, straining, environment), with the associated microstructural changes deduced from appropriate micrographs, diffraction patterns or video-recordings. The influence of the electron microscope itself must always be taken into account (e.g. thin-foil specimen, large electron flux, high electron energy). However when artifacts are overcome, remarkable insight into the mechanisms of the behavior of solids can be achieved. The purpose of this article is to review our work in extending this capability into the high-resolution regime, so that atomic reactions can be followed. In the right circumstances this allows straightforward interpretation of atomic-scale phenomena.Our studies have employed a Philips EM430ST, 300 kV transmission electron microscope (TEM) with about 0.2 nm resolution. To-date we have carried out only heating experiments, utilizing the standard side-entry, single-tilt specimen holder (model number PW 6592), mainly on reactions at semiconductor interfaces. The zone axis orientation necessary for high-resolution TEM is obtained by careful sectioning of the substrate crystals and judicious positioning in the heating holder. The imaging conditions (including drift) can be optimized at a temperature slightly below that required for the changes of interest, so that recording can be initiated immediately on ramping up.