THE EFFECT OF TIN DOPING ON α‑Fe2O3 PHOTOANODES FOR HYDROGEN PRODUCTION


Christopher D. Bohn, Alec A. Talin, Veronika A. Szalai

The sun delivers more than 120,000 TW of power to the Earth’s surface – nearly 10,000 times the current global consumption rate of 15 TW. 1 The viability of solar energy relies on the development of efficient and affordable energy storage solutions to meet stable supply requirements.  Splitting water to form separate streams of H2 and O2 is one environmentally benign possibility.

Hematite, α-Fe2O3, has shown potential as a photoanode material because (i) it has a bandgap between 1.9 eV and 2.2 eV that maximizes absorption from the solar spectrum,2 (ii) it is stable in an aqueous environment under operating conditions,3 and (iii) it is abundant and affordable. Several challenges remain, however, including overcoming hematite’s low electrical mobility of 0.01-0.1 cm2V-1s-1,4,5,6 short hole diffusion length7,8 of 2 nm - 4 nm, and the thermodynamic requirement of an applied overpotential. Some shortcomings can be overcome by doping and, recently, it has been shown that annealing α-Fe2O3 on fluorine-doped tin oxide (FTO) coated soda lime glass at 800 °C gives markedly enhanced photocurrents for porous films9 as well as  nanowires.10,11 This enhancement has been attributed to Sn n-type doping of the α-Fe2O3 phase.  To design improved photoanodes for water splitting, a quantitative understanding of Sn incorporation is necessary. 

In the present paper, we produced sputtered films of α-Fe2O3 to permit facile characterization of the Sn incorporation upon annealing.  The sputtered samples were characterized with secondary ion mass spectrometry, which permitted the determination of the Sn doping as a function of distance from the FTO interface for the first time. Additionally, the effect of heat treatment on morphology was investigated with TEM and XRD and highlighted changes in crystallinity. Finally, the incorporation of Sn was shown to dramatically reduce the specific resistivity of the film by 3 to 6 orders of magnitude. Real change in photocurrent by orders of magnitude is most likely a result of this change in specific resistance upon annealing. Using the determined dopant density, further steps towards the rational design of photoanodes from inexpensive and earth-abundant α-Fe2O3 for storing solar energy as hydrogen should be possible.