Computational discovery of p-type transparent oxide semiconductors using hydrogen descriptor

The ultimate transparent electronic devices require complementary and symmetrical pairs of n-type and p-type transparent semiconductors. While several n-type transparent oxide semiconductors like InGaZnO and ZnO are available and being used in consumer electronics, there are practically no p-type oxides that are comparable to the n-type counterpart in spite of tremendous efforts to discover them. Recently, high-throughput screening with the density functional theory calculations attempted to identify candidate p-type transparent oxides, but none of suggested materials was verified experimentally, implying need for a better theoretical predictor. Here, we propose a highly reliable and computationally efficient descriptor for p-type dopability-the hydrogen impurity energy. We show that the hydrogen descriptor can distinguish well-known p-type and n-type oxides. Using the hydrogen descriptor, we screen most binary oxides and a selected pool of ternary compounds that covers Sn2+ -bearing and Cu1+-bearing oxides as well as oxychalcogenides. As a result, we suggest La2O2Te and CuLiO as promising p-type oxides.