Theory of core excitons

Abstract

The observation of core excitons with binding energies much larger than those of the valence excitons in the same material has posed a long-standing theoretical problem1. Here we review our proposed solution to this problem2 and show that Frenkel excitons and Wannier excitons can coexist naturally in a single material. A primary-edge core exciton is created by exciting a core electron to an orbital with energy in the forbidden gap below the lowest conduction band edge. Such an exciton may be either a Frenkel exciton or a Wannier exciton, depending on the strength of the static core-hole potential in the central-cell. Frenkel resonances may occur above the primary gap even when the primary-gap excitons are extended effective-mass Wannier states. We have predicted the major chemical trends in the binding energies of core Frenkel excitons in zincblende semiconductors, including the dependences on: (i) the host, (ii) the site to which the core hole is attached, (iii) the angular momentum or irreducible representation of the electron orbital, and (iv) the atom or impurity excited in the core transition. Furthermore, we have exhibited how a surface or interface alters these results3,4 and how the presence of a ‘spectator’ impurity adjacent to the core-excited atom5 affects the Frenkel exciton binding energy.