Cationic and anionic vacancies on the NiO(100) surface: DFT+U and hybrid functional density functional theory calculations

The electronic structure of oxygen and nickel vacancies at the surface of NiO(100) has been investigated theoretically by means of density functional theory (DFT) comparing plane wave density functional theory with the Hubbard correction DFT+U with atomic-orbital-hybrid-functional (20% exact-exchange) calculations. The two approaches provide a similar description of the nature of the oxygen vacancy. At variance with the same defect center in MgO, where two electrons are trapped in the vacancy, on NiO the charge is more delocalized, partly over the Ni ions around the vacancy, the rest either trapped in the vacancy or delocalized over other Ni ions. Concerning the nickel vacancy, both methods clearly show that the removal of a neutral Ni atom does not result in the oxidation of other nickel ions from Ni2+ to Ni3+ but rather in the formation of two holes in the O 2p valence band. However, the description is slightly different with the approaches, the hybrid functional indicates that the holes are essentially localized on the oxygen ions nearest to the vacancy, while they result much more delocalized from the DFT+U calculation. Comparison with the corresponding results, obtained with the two methods for the case of the Mg vacancy in MgO, suggests that the DFT+U approach does not adequately correct for the self-interaction of the unpaired electrons in this case. However, the overall picture that emerges clearly from the present calculations is that both defects affect the electronic structure in a much wider region in NiO than in MgO.