The Spinel/Alumina Phase Boundary
The formation of spinel during solid-state reactions between two oxides of the type A0 (e.g. Mg0) and B203 (e.g. A1203), has been extensively studied both from a theoretical viewpoint and an experimental one. The present paper will illustrate the structural aspects of the study of the spinel/sesquioxide interface. It has recently been shown by Carter and Schmalzried, that, when Co0 and A1203 react to form Co-Al spinel, the {111} oxygen planes in the spinel do not lie parallel to the (0001) oxygen planes in the parent alumina even though the oxygen ions are almost close-packed in both planes. The small rotation which is present implies that the mechanism whereby the alumina is transformed to spinel is not simply the glide of either isolated, or bundles of, Shockley partial-like transformation dislocations as had previously been assumed, but rather involves a new defect which causes a rotation of the oxygen sublattice.