Quantitative modelling of defect processes in ionic crystals

This paper is concerned with the accurate prediction of defect properties and processes by quantum methods, with emphasis on ionic systems. Such approaches are of especial value for situations where direct experiment is difficult because of timescale or of extremes of temperature or pressure. In most cases one has to solve an electronic structure problem coupled with a technique for including polarisation and distortion of the surrounding solid. Such polarisation terms can be large and even dominant, and usual simplistic approaches are insufficient. I discuss several aspects, all including use of empirical potentials, namely (i) strategies for combining electronic structure and distortion, (ii) classical calculations of quantum energies, and (iii) cases where one needs quantum dynamics. Examples are given for work expoiting codes developed at Harwell.