Transition metal oxides are promising photoelectrode materials for solar-to-fuel conversion applications. However, their performance is limited by the low carrier mobility (especially electron mobility) due to the formation of small polarons. Recent experimental studies have shown improved carrier mobility and conductivity by atomic doping; however the underlying mechanism is not understood. A fundamental atomistic-level understanding of the effects on small polaron transport is critical to future material design with high conductivity. In this talk, we will discuss the effect of small polaron formation on optical and carrier transport properties of transition metal oxides from first-principles calculations.

First, we resolve the conflicting findings that have been reported on the optical gap of a well-known catalysis Co₃O₄ as an example[1]. We confirm that the formation of small hole polarons significantly influences the optical absorption spectra and introduces extra spectroscopic signature below the intrinsic band gap, leading to a 0.8 eV transition that is often misinterpreted as the band edge that defines the fundamental gap.

Then we discuss the formation of small polarons' effect on carrier concentration, by resolving the controversy of nature of "shallow" or "deep" impurities of intrinsic oxygen vacancies in BiVO₄ as an example[2], i.e. how to unify different experiments with the correct definition of ionization energy in polaronic oxides. We further discuss why certain dopants can have very low optimal concentrations (or very early doping bottleneck) in polaronic oxides such as Fe₂O₃, through a novel "electric-multipole" clustering between dopants and polarons[3]. These multipoles can be very stable at room temperature and are difficult to fully ionize compared to separate dopants, and thus they are detrimental to carrier concentration improvement. This allows us to uncover mysteries of the doping bottleneck in hematite and provide guidance for optimizing doping and carrier conductivity in polaronic oxides toward highly efficient energy conversion applications. In addition, we show the importance of synthesis condition such as synthesis temperature and oxygen partial pressure on dopant and polaron concentrations, and how to optimize the synthesis condition based on theoretical predictions.

At the end, we show different theoretical models for polaron mobility calculations from a macroscopic dielectric continuum picture with an example of spin polarons in CuO[4] and a microscopic polaron hopping picture by combining generalized Landau-Zener theory and kinetic Monte-Carlo samplings for doped oxides[5].

Our first-principles calculations provide important insights and suggest design principles for optimal optical and transport properties of polaronic oxides.