An ab initio study of the origin of p-type doping in ZnO using group-V elements

Abstract

Zinc oxide is a transparent semiconductor with a direct wide band-gap 3.4 eV and large exciton binding-energy of 60 meV, that combine to make ZnO a promising material for possible applications such as optoelectronic devices, lasers and light emitting diodes. Recently, the difficulty in obtaining high quality p-type ZnO has attracted much attention. Considerable effort has been made to obtain p-type ZnO by doping with the group- V elements N, P, As, and Sb, with the anticipation of replacing oxygen atoms in the ZnO lattice. However, experimentally these dopants can produce both p-type and n-type conductivity. Here the results of first principles density functional theory calculations performed using the AIMPRO code are presented. By evaluating the relative energies of substitution on the oxygen and zinc sub-lattices, it is possible to predict the most likely forms of doping centres that might be achieved depending both upon the dopant species and whether the ZnO is grown under oxygen or zinc rich conditions. As a general trend, it is found that dopants tend to be stabilised in environments where covalent bonds with oxygen can be formed, such as substitution on the zinc sub-lattice. The doping properties of the group-V elements can be best understood by not considering the dopant atoms individually, but as a part of an atomic group such as phosphate and nitrate ions either substituting for host atoms, or lying in interstitial sites. The preferential formation of dopant-oxygen bonds leads to a revision of the zinc-vacancy based model for p-type doping (such as P-(VZn)2 complexes) to structures involving interstitial oxygen.