Diffuse scattering, size effect and alloy disorder in ternary and quaternary III–V compounds

Influence of the static atomic displacement on atomic resolution Z-contrast imaging

“…The InGaAs alloy was chosen as a case material to investigate the effect of the SDs upon HAADF images because, in this material, the SDs have already been studied in detail by extended x-ray absorption fine structure, 20 electron diffraction, 21 and diffraction contrast TEM imaging. 16,22 For simulation purposes, we first generate alloy supercells and then compute the HAADF images.…”

“…16,22 For simulation purposes, we first generate alloy supercells and then compute the HAADF images. To generate crystal supercells with SDs, we use the valence force field model 21,[23][24][25] which is known to reproduce accurately the SD field in semiconductor alloys. 21 In the specific case of TEM, this model already proved successful as a basis for simulating the effects of the SDs on diffraction patterns, 16 channelled-electron-beam-induced x-ray emission, 26 diffraction contrast imaging ͑allowing to reproduce the fine structure of the two-beam dark field images of III-V alloys 21 ͒, and structure factor calculation.…”

“…To generate crystal supercells with SDs, we use the valence force field model 21,[23][24][25] which is known to reproduce accurately the SD field in semiconductor alloys. 21 In the specific case of TEM, this model already proved successful as a basis for simulating the effects of the SDs on diffraction patterns, 16 channelled-electron-beam-induced x-ray emission, 26 diffraction contrast imaging ͑allowing to reproduce the fine structure of the two-beam dark field images of III-V alloys 21 ͒, and structure factor calculation. 22,27 The HAADF images were calculated by multislice simulations in the frozen-phonon approach; 28 this method allows the most accurate simulation of the Z contrast.…”

Response to `About the diffuse absences and the diffuse planes due to the atomic size effect'

“…This creates an absence of diffuse intensity along rows connecting Bragg peaks. We believe this interpretation of the apparent diffuse absences in the two material examples of our paper (where scattering contrast was minimal) is the most informative.Admittedly, Glas et al (1990) have computed the diffuse intensity exactly from their model so aspects of both the Warren terms and this symmetric component must be present. However, as we pointed out in our paper, when the scattering contrast is large, the asymmetric Warren size-effect intensity dominates and when the scattring contrast diminishes, the diffuse intensity is composed mainly of symmetric components associated with mean-squared pair displacements.…”

“…Admittedly, Glas et al (1990) have computed the diffuse intensity exactly from their model so aspects of both the Warren terms and this symmetric component must be present. However, as we pointed out in our paper, when the scattering contrast is large, the asymmetric Warren size-effect intensity dominates and when the scattring contrast diminishes, the diffuse intensity is composed mainly of symmetric components associated with mean-squared pair displacements.…”