First-principles Investigation of Small Polarons in Metal Oxides

An important factor limiting the conductivity is the interaction of the charge carrier with polar phonon modes. Such a phonon-dressed charge carrier is called polaron. The strength of the electron-phonon (el-ph) interaction determines the localization of the polaron, which in turn e.g. defines its characteristic temperature dependence for the charge-carrier mobility. We focus on metal oxides with strong el-ph coupling, where small polarons are formed. Density-functional theory is often used for calculating properties of polarons. However, there are two challenges: sensitivity of the calculated properties to the errors in exchange-correlation (XC) treatment and finite-size effects in supercell calculations. In this work, we develop an approach that addresses both challenges. The polaron properties are obtained using a modified neutral potential-energy surface (PES). By changing the fraction of exact exchange in the hybrid HSE functional we show that the modified PES model significantly reduces the dependence of the polaron properties on the XC functional. Based on Pekar's potential for the long-range el-ph coupling, we derive the proper elastic long-range behavior of the polaron and a finite-size correction for the polaron properties. These findings are proofed by an extensively test for rock salt MgO and rutile TiO2. Finally, the approach is used to investigate the influence of the crystal structure on the polaron properties for rutile and anatase TiO2, as well as for the monoclinic β- and orthorhombic κ-phase of Ga2O3. While in rutile TiO2 only small electron polarons are stable, only small hole polarons are found in anatase. Further, small hole polarons exist in both Ga2O3 polymorphs but have significantly different binding energies. Thus, we conclude that growing crystals of the same material but with different structure can be used to manipulate conductivity and charge-carrier mobility