Defect relaxation energy as a criterion for doping engineering based on high-throughput computation: Application to co-doping CaTiSiO5

Abstract

CaTiSiO5 (titanite or sphene) is a potential linear dielectric material with high dielectric constant for high temperature stability multilayer ceramic capacitor applications. To suppress the formation of α phase CaTiSiO5, which has low temperature stability, co-doping is considered to be an effective approach. Here, we propose to use defect relaxation energy as a screening criterion to quickly identify the potential co-doping elements. The co-doping at the cation sites employing the donor–acceptor pairs is studied. One hundred and twenty element pairs at nine site combinations are considered. Based on first-principles calculations, our high-throughput screening not only identifies those dopant pairs that have been explored experimentally and demonstrated feasible, but also predicts new candidates. In particular, we experimentally synthesized CaTiSiO5 with (Nb, Na) co-doping at Ti–Ca sites (Ca1−xNaxTi1−xNbxSiO5). When x = 0.02, the sample achieves a high room-temperature dielectric constant (εr = 53) and exhibits excellent temperature stability in the range of −55 to 300 °C with the temperature coefficient of capacitance less than ±170 ppm/°C. It is expected that the proposed quick screening approach can be readily applied to the doping engineering of other inorganic nonmetallic functional materials.