A complete ab initio thermodynamic and kinetic catalogue of the defect chemistry of hematite α-Fe2O3, its cation diffusion, and sample donor dopants.

This paper studies comprehensively the defect chemistry of and cation diffusion in α-Fe2O3. Defect formation energies and migration barriers are calculated using density functional theory with a theoretically calibrated Hubbard U correction. The established model shows a good agreement with experimental off-stoichiometry and cation diffusivities available in the literature. At any temperature, and are the predominant ionic defects in hematite at the two extremes of oxygen partial pressure (pO2) range, reducing and oxidizing, respectively. Between these two extremes, an intrinsic electronic regime exists where small polaronic electrons and holes are the dominant charge carriers. The calculated migration barriers show that Fe ions favor the diffusion along the 〈111〉 direction in the primitive cell through an interstitial crowdion-like mechanism. Our model suggests that cation diffusion in hematite is mainly controlled by the migration of , while may contribute to cation diffusion at extremely low pO2. Our analysis in the presence of two sample donor dopants Ti and Sn indicates that high temperature annealing at T > 1100 K is needed to prepare n-type hematite at ambient pO2, consistently with prior experimental findings. Alternatively, annealing at lower temperatures requires much lower pO2 to avoid compensating the donors with Fe vacancies. A synergistic comparison of our theoretical model and the experimental results on Ti-doped hematite led us to propose that free electrons and small polarons coexist and both contribute to n-type conductivity. Our validated model of defective hematite is a foundation to study hematite in applications such as corrosion and water splitting.

Tuning metal oxide defect chemistry by thermochemical quenching.

The low-temperature defect chemistry of monoclinic and tetragonal ZrO2 and hematite Fe2O3 is studied in the non-equilibrium state of thermochemical quenching; that is, rapid cooling starting from a certain high temperature and oxygen chemical potential. This non-equilibrium state is of great interest because many metal oxides are used at low temperatures below their growth temperatures. This paper addresses the importance of considering this non-equilibrium state rather than applying equilibrium thermodynamics as commonly used when studying point defects from first principles. Based on point defect formation energies calculated previously using density functional theory, we compare the type of dominant defects at equilibrium to those at a quenched state originating from a certain initial growth temperature and oxygen partial pressure. The comparison is facilitated by casting the dominant defects in a dominance diagram on the temperature - oxygen partial pressure plane. We consider two scenarios to model the quenched state. In the first, only electronic defects equilibrate whereas all ionic defects are frozen. Whereas, in the second, electronic defects and interstitials are allowed to equilibrate under the assumption of mobile interstitials at low temperatures. We find that new ionic charge compensation modes can appear on the dominance diagram after quenching. Additionally, purely ionic charge compensation modes consisting of vacancies and/or interstitials expand in the dominance diagram at the expense of purely electronic compensation modes. For the ZrO2 phases, we argue that scenario 2 is more realistic and leads to difficulty in achieving n-type doping by thermochemical quenching. For Fe2O3, and regardless of the quenching scenario, iron vacancies occupy a wider zone of domination, which limits the performance of this oxide as a water splitting photoanode. Our study shows that by controlling the growth thermochemical conditions, it is possible to tune the Fermi level of oxides over a considerable range within the band gap by quenching. This provides an extra tool to tune the electric conductivity of metal oxides beyond traditional extrinsic doping. This work indicates that non-equilibrium thermodynamic analysis is necessary to understand and control defect chemistry at low temperatures.