Abstract
The diffusional coherency strain theory and its implications for chemically induced migration of liquid films and grain boundaries are reviewed and critically examined. The critical experiments in metals and oxides show the validity of this theory. The observed characteristics of this phenomenon: misfit dislocations at the initial boundary location; migration direction; faceting; dependence on grain boundary structure, on boundary curvature, and on temperature are also consistent with the coherency strain theory. Other theories are also described, but are mostly inconsistent with the observations, and there is very little experimental evidence for the validity of any of them. There was experimental evidence that the coherency strain theory is also valid for discontinuous precipitation. The significance of lattice diffusion at small distances in the order of the lattice spacing is examined. The related phenomena of diffusion induced recrystallisation and surface corrugation are also described. Finally, some examples of these phenomena in the practical materials are presented.