Intrinsic<i>n</i>-type versus<i>p</i>-type doping asymmetry and the defect physics of ZnO

TLDR

ZnO exemplifies materials that can be doped via native defects in only one direction, either n‑type or p‑type. The study aims to explain the n‑type versus p‑type doping asymmetry in ZnO by investigating its intrinsic defect physics and the role of n‑type impur.

Abstract

ZnO typifies a class of materials that can be doped via native defects in only one way: either n type or p type. We explain this asymmetry in ZnO via a study of its intrinsic defect physics, including ${\mathrm{Zn}}_{\mathrm{O}},$ ${\mathrm{Zn}}_{i},$ ${\mathrm{V}}_{\mathrm{O}},$ ${\mathrm{O}}_{i},$ and ${V}_{\mathrm{Zn}}$ and n-type impurity dopants, Al and F. We find that ZnO is n type at Zn-rich conditions. This is because (i) the Zn interstitial, ${\mathrm{Zn}}_{i},$ is a shallow donor, supplying electrons; (ii) its formation enthalpy is low for both Zn-rich and O-rich conditions, so this defect is abundant; and (iii) the native defects that could compensate the n-type doping effect of ${\mathrm{Zn}}_{i}$ (interstitial O, ${\mathrm{O}}_{i},$ and Zn vacancy, ${V}_{\mathrm{Zn}}),$ have high formation enthalpies for Zn-rich conditions, so these ``electron killers'' are not abundant. We find that ZnO cannot be doped p type via native defects $({\mathrm{O}}_{i},{V}_{\mathrm{Zn}})$ despite the fact that they are shallow (i.e., supplying holes at room temperature). This is because at both Zn-rich and O-rich conditions, the defects that could compensate p-type doping ${(V}_{\mathrm{O}}{,\mathrm{}\mathrm{Zn}}_{i},{\mathrm{Zn}}_{\mathrm{O}})$ have low formation enthalpies so these ``hole killers'' form readily. Furthermore, we identify electron-hole radiative recombination at the ${V}_{\mathrm{O}}$ center as the source of the green luminescence. In contrast, a large structural relaxation of the same center upon double hole capture leads to slow electron-hole recombination (either radiative or nonradiative) responsible for the slow decay of photoconductivity.