A fuzzy logic approach to image analysis of HREM micrographs of III–V compounds

Abstract

High resolution electron microscopy can within certain limits provide quantitative information on morphology and composition of crystalline materials. In layered structures of III-V semiconductor compounds changes resulting from interdiffusion phenomena across interfaces may result in image contrast changes because of composition sensitive electron scattering. A new approach for extracting the corresponding information from cross-section images of such interfaces is introduced. Local variations of similarity, a measure of comparability with respect to previously defined templates of known material composition on both sides of the interface, are recorded by image analysis. On the basis of the well-worked outfuzzy set theory variations of the underlying chemical composition are evaluated by applying fuzzy logic criteria of inference to image regions of about 1 nm x 1 nm in size.With comparatively little computational effort this new approach can be used to localize regions of significant composition changes, i.e. edge detection, and to determine the composition across the interface region, i.e. chemical mapping, respectively. The method is introduced and its reliability demonstrated by means of simulated contrast patterns of GaAs/GaAs, _XP, as well as of experimental images of AlAs/Al,Ga, _.As.

Fuzzy logic approaches to the analysis of HREM images of III–V compounds

“…For practical applications of this FL rule base, the shape, the base size and the intersections of the actual membership functions have to reproduce the similarity/composition characteristics of the III–V system under investigation. The quasilinear case of (100)‐orientated GaAs 1− x P x has been analysed using triangular fuzzy sets of similarity (Hillebrand et al ., 1996b) for varying chemical gradients in the group V sublattice. In this paper, both quasilinear and highly nonlinear similarity/composition characteristics of Al x Ga 1− x As are studied in detail using trapezoidal and triangular membership functions of similarity (for experimental details see Fig.…”

“…1 present the full-range normalized similarity/composition ᭧ 1998 The Royal Microscopical Society, Journal of Microscopy, 190, 61-72 characteristics j nrm ¼ j/j max in the parameter windows for t ¼ t 0 Ϯ 1·5 nm. The possibilities of studying diffusion phenomena in the group V sublattice by HREM image analysis are treated extensively elsewhere (Hillebrand et al, 1996b). As Fig.…”

“…In interpreting simulated HREM patterns of interfaces one should consider that the crystal cell sizes and the cell positions in the image field are exactly known. The effect of lateral cell/template misalignments is discussed elsewhere (Hillebrand et al ., 1995a, 1996b). Practical aspects of QUANTITEM, also with respect to the crystal cell detection in experimental images, are discussed by Kisielowski et al .…”

“…Practical problems of template fit, caused by distorted HREM images of strained crystals, are discussed elsewhere (Hillebrand et al ., 1996b). It proved to be complicated to derive templates representative of the two materials in the corresponding images.…”

HREM Image Analysis of III-V Heterostructures Based on Neural Networks

“…Respective questions of non-linear effects and the difference in the information content of (110)-and (100)-projected crystals have been discussed elsewhere [7,8]. It should be noted that alternative fuzzy logic approaches [9,10] have also been elaborated and successfully applied to determine the local sublattice composition in III±V micrographs. In [11], specimen thickness and defocus were determined by comparing experimental and simulated high resolution images of Nb bicrystals.…”