Do we really need configuration interaction theory to understand the negative vacancy in silicon?

Abstract

In its negative charge state, the lattice vacancy V − in silicon experiences a C 2v Jahn–Teller distortion. In an LCAO picture, the unpaired electron of b 1 symmetry gives rise to a spin density that is concentrated in a (1 1 ̄ 0) plane with a nodal (1 1 0) plane. Hence, the contact hyperfine (hf) interactions with all ligand nuclei in this plane should be zero. The order of magnitude of the nonzero albeit small contact interactions found experimentally was explained by configuration interaction (CI) theory. This raises the question whether the vacancy can be described without the incorporation of CI theory. We have performed an ab initio calculation using the standard local spin density approximation (LSDA) for the exchange-correlation interaction. We obtain the order of magnitude of the hf interaction with the nuclei in the (1 1 0) plane, if we just impose the C 2v symmetry on the unrelaxed V Si − defect state. When taking into account the properly relaxed ligand coordinates, we obtain quantitative agreement with experimental data. Thus for V Si − in silicon, correlation effects are important and are treated satisfactorily within the LSDA.