Evidence for small polaron formation leading to intrinsic photoexcited charge trapping in α-Fe_2O_3

Abstract

Hematite has long been a promising photoanode material for photoelectrochem. water splitting due to its suitable bandgap (1.9-2.2 eV), low cost, and photostability. However, photoexcited carriers with short diffusion lengths and short recombination times, thought to be caused by ultrafast trapping, have limited hematite's conversion efficiency to a third of its theor. value. Herein we discuss evidence that the ultrafast trapping of photoexcited carriers in hematite (α-Fe_2O_3) occurs by small polaron formation. Using visible pump-extreme UV (XUV) probe spectroscopy we observe a pulse-width limited shift in charge d. from the oxygen center to the iron center through a change in the multiplet splitting of the Fe M_(2,3)-edge, also understood as a change in oxidn. state of the iron center from Fe^(3+) to Fe^(2+). Small polaron formation then starts on a sub-100 fs time scale, as evidenced by a breaking of the Fe 3p core level's degeneracy in the XUV transition. This spectral signature continues increasing in amplitude up to 2-3 ps. The obsd. small polaron localization timescales match that commonly attributed to mid-gap or surface trap states in hematite. However, the pre-edge absorption or bleach expected for mid-gap or surface trap states in the XUV spectrum is not obsd. Small polaron formation would lead to a decrease in mobility of photoexcited carriers since conduction must occur by thermally activated hopping. Accordingly, we observe that the probability of photoexcited charge trapping with increasing excitation energy matches trends in the photoconversion efficiency with increasing excitation energy, particularly in the t2g conduction band. This presentation provides insight into the role of small polaron formation in photoexcited charge trapping in hematite, and more generally into the charge trapping process in transition metal oxide catalysts thought to be limited by small-polaron conduction.