On the widths of dislocation images in X-ray topography under low-absorption conditions

Detailed measurements have been made of dislocation image widths in X-ray topographs taken under conditions of low absorption (product of linear absorption coefficient and crystal thickness, μt = 0.15) with a silicon specimen and Mo Kα radiation. These have been compared with two theoretical models, the `diffracting-zone' model involving computation of misorientation gradient contours, and the direct-image `mosaic' model using calculations of effective misorientation contours. Neither model completely explains the image widths. The diffracting-zone model, though theoretically more attractive, gives a very poor correlation, probably because of (a) severe photographic limitations and (b) the uncertainty in the reconstruction of the intermediary image in translation topographs. For a given reflection the direct-image mosaic model gives a reasonably good prediction of image widths measured at half peak height. However, the widths in different reflections are not self-consistent; in particular, images in lower-order reflections are significantly wider than those predicted by the model. This implies that contributions from interbranch scattering (giving an intermediary image component) are more important in lower-order reflections, and that such contributions are proportional to the widths obtained from misorientation (rather than misorientation gradient) contours.