Cupric oxide as low-cost photovoltaic material : a combined DFT and experimental study
Date of Issue2015
School of Materials Science and Engineering
Due to low cost and suitable band gap value, cupric oxide (CuO) is of great interest for large-scale photovoltaic application. However, CuO is usually p-type and n-type CuO is not available. The goal of this study is to understand the behaviour of various dopants and codopants that may affect the electrical conductivity of CuO, and to search for a suitable method to make n-type CuO by using first-principles density functional theory calculations followed by experimental studies. Our DFT calculations identify that copper vacancy can lead to p-type CuO, while oxygen vacancy is a deep donor. A systematic study of several potential p-type and n-type dopants in CuO indicates that Li and Na are shallow acceptors and their formation energies are low in oxygen rich environment. However, it is found that n-type conduction is relatively difficult to be achieved by donors, as most donors have deep donor levels in the band gap and/or high formation energies. Hf and Zr have the shallowest donor levels ~ 0.2 eV below the conduction band minimum, but their formation energies are relatively high, indicating low dopant solubility in CuO. Then, we performed DFT study of donor-acceptor (n-p) codoping approach to obtain n-type CuO. Our results show that nitrogen codoping can slightly improve the donor levels of Zr and In by forming shallower n-type complexes, but the codopant formation energies are so high that it is hard to realize them in experiments. However, it is found that Li codoping with Al and Ga can be achieved easily, because Li codoping can improve the solubility of Al/Ga in CuO. The corresponding n-type defect complexes have shallower donor levels than those of single Al and Ga donor doping by around 0.1 eV, and their formation energies are reasonably low to act as an efficient codopant. Moreover, Li codoping with both Al and Ga produce an empty impurity band just below the host conduction band minimum, which can reduce the donor ionization energy at high codoping concentrations. Our first-principle calculations suggest that Li codoping can benefit n-type CuO. To test some of the predictions, experimental studies have been conducted on Li and Zr doped CuO. Our results have shown that Li doped CuO is p-type conductive with the measured resistivity enhanced by more than two orders of magnitude. However, Zr doped CuO is also found to be p-type, because intrinsic copper vacancies is not totally compensated by Zr dopant due to its low solubility in CuO. Overall, our experimental results are consistent with the theoretical predictions.