Refinement of rigid shift component normal to a Σ5 grain boundary in rutile by quantitative HREM
In contrast to the macroscopic parameters of a grain boundary, its microscopic parameters are difficult to determine with sufficient accuracy to make critical comparisons between models and experimental observations. Whereas the axis and angle of misorientation and the orientation of the boundary plane are easily measured from diffraction patterns or high resolution images, the rigid body shift and the localized atomic relaxation are far more complicated to determine and are more sensitive to image or sample artifacts. Yet it is only these microscopic parameters that differentiate the behavior of the same crystallographic interface in different materials. While the angle of misorientation and the plane of inclination are fixed macroscopically, the rigid body shift and localized atomic relaxations depend directly on the type of bonding that characterizes a given material.Several methods for the measurement of rigid shifts at interfaces have been given recently. Their accuracy is limited by instrumental parameters such as beam tilt, sample tilt, residual two-and threefold astigmatism, specimen noise or twist components.