Effect of ZnO grain boundaries on non-linearity: first-principles calculations

The non-linearity (NL) of monocrystal and polycrystal ZnO materials was studied on the basis of density functional theory. Firstly, wurtzite ZnO was selected as a candidate material for selector devices. As indicated by the calculated I-V curves on different ZnO surfaces, non-linearity and electrical conductivity varied among the (), (0001) and () directions. The calculating models for wurtzite ZnO in three directions were named SF1, SF2 and SF3, respectively. SF1 and SF2 showed poor non-linearity although they demonstrated high electrical conductivity. SF3 exhibited excellent NL, but its electrical conductivity was low. Polycrystal ZnO material was studied, and three grain boundary (GB) models, namely, ZnO∑7()[0001]GB (GB1), 45° rotating GB(GB2) and ∑1()[]GB (GB3), were constructed. The GB barrier was found between two ZnO grains by calculating the I-V curves of GB and analysing the electrostatic potential, which resulted in NL improvement of GB compared with SF. By comparing the electrostatic potential analysis, we found that the electrical conductivity of GB3 was significantly enhanced to greater extent than that of SF3 because the Fermi level moved towards the conduction band. Therefore, the corresponding structure of polycrystalline ZnO should have better performance compared with single crystal ZnO to meet the selector design requirements. This work may be instructive and valuable for the design and optimization of ZnO-based gate selectors.