Doping engineering of p-type ZnO

Abstract

The formation energies of defects in ZnO taken from recently published first-principles calculations are used to simulate the incorporation and compensation of group-V impurities. A detailed analysis is carried out for the specific case of nitrogen-doped ZnO as a function of doping temperature, O/Zn stoichiometric ratio, and dopant activity. This study shows that net p-type doping of ZnO:N in thermodynamic equilibrium conditions is rather limited (∼10 16 cm −3) and involves a delicate balance between the achievement of high dopant solubility (on the Zn-rich side) and low compensation by donor defects (on the O-rich side). Efficient doping appears possible in various alternative ways: simultaneous incorporation of hydrogen as an interim compensating donor to be subsequently eliminated; use of a high activity dopant in the external phase (atomic N, NO) while preventing the formation of (N 2) O compensating donors; planar doping for separate optimization of nitrogen incorporation and a defect-free crystal. The other group-V impurities (P, As) introduce deep acceptor levels when substituted for O, but have been predicted to form a single acceptor complex in the antisite position, like [ As Zn–2V Zn], with both low formation and ionization energies. This leads to a moderate p-type behaviour. Comparison of the present theoretical modelling with experiments shows that accounting for the wide range of reported hole concentrations (10 16–10 19 cm −3) necessarily implies some departure from equilibrium.