Vacancy-interstitial pair production via electron-hole recombination in halide crystals

Abstract

The nature of the electronic states through which band-to-band excitations or excitons evolve into $F-H$ pairs is central to an understanding of defect production in halide crystals. The present paper develops a recent empirical model in which a self-trapped exciton becomes an $F-H$ pair through a relaxation process that preserves the $\sigma$ bond between the two halide ions on which the hole is localized. The model is shown to be consistent with relative energies of the initial and final states, and the potential surfaces which govern motion from one configuration to another have been investigated in terms of diabatic correlation rules. The defect state $F\left(1s\right)+H\left({\sigma }_u\right)$ lies on a potential surface which correlates with higher levels of free and self-trapped excitons and crosses the surface corresponding to the lowest exciton state. Possible bottlenecks and associated metastable intermediate states are discussed with reference to observations from time-resolved spectroscopy. The initial ionic motion between self-trapped exciton and $F\left(1s\right)+H\left({\sigma }_u\right)$ is not necessarily confined to a close-packed direction in alkali halides. The close $F-H$ pairs observed in alkaline-earth fluorides are cited as examples of initial vacancy formation involving rotation as well as translation of an excited halide pair, and an analogous reaction path in alkali halides is suggested.