Solar energy harvesting using e.g. photovoltaic modules needs to be coupled to energy storage, because of the diurnal and intermittent nature of solar energy. Hydrogen is a candidate for such chemical energy storage, because of its facile oxidation in fuel cells. Splitting liquid water to produce hydrogen (and oxygen) using solar energy requires a minimum of 1.48 eV under isothermal conditions. However, material restrictions, low energy conversion efficiencies and little focus on scalability have inhibited development and deployment of such technology.

Thermodynamically less energy intensive processes, such as hydrogen sulfide splitting that requires a minimum of only 0.27 eV, remain largely as a concept for lack of suitable semiconductors, though its feasibility has been reported [1]. Hydrodesulfurisation processes in the world’s oil refineries use hydrogen to remove sulfur from oil products as (highly toxic and corrosive) H₂S, which has to be oxidised subsequently to elemental sulfur at a scale of 64 Mt per year by Claus processes. Using hydrogen sulfide to produce hydrogen and polysulfide ions (S_n^-2) with solar energy could achieve the same result without the need to produce hydrogen from fossil fuels effectively emitting CO₂.

The feasibility of photo-assisted splitting of (aqueous absorbed) hydrogen sulfide ions in alkaline solutions has been demonstrated using Sn^IV-doped α-Fe₂O₃ photo-anodes fabricated on Ti substrates by spray pyrolysis [2]. As shown in Figure 1, these exhibited hydrogen sulfide oxidation at lower electrode potentials than for oxidation of hydroxide ions and appeared stable, despite thermodynamic predictions that iron sulfide should form, at least under dark conditions. Photo-current densities were stable over 75 hours. The average stoichiometry of polysulfide (S_n^2-) was found to be n = 2.63 after 30 hours of photo-electrolysis. However, S₅^2- is predicted to be the predominant species at the surface of the photo-anode as a product of photo-oxidation. Applied potentials needed to be < 1 V (RHE), to avoid the (reversible) formation of inhibiting elemental sulfur on the photo-anode.

Photo-electrochemical impedance spectroscopy, open circuit potential at high intensities and chrono-amperometry were used to determine the flat band potential and incident photon-to-current efficiencies (IPCEs) of the Sn^IV-doped α-Fe₂O₃ photo-anodes. The time dependence of polysulfide concentrations was also measured using two photoelectrochemical reactors with 0.18 cm² and 10 cm² photo-anodes. A kinetic model was developed to describe the photo-oxidation of hydrogen sulfide ions in alkaline conditions, together with a numerical model using Comsol Multiphysics to analyse and predict the performance of photo-anodes in photo-electrochemical reactors.

Despite its success for HS^- oxidation, H₂ production using an α-Fe₂O₃ photo-anode requires an electrical bias to be applied. This is necessary to depress the potential corresponding to the energy of its conduction band edge below that of the equilibrium potential of the water / hydrogen couple.

References

[1] R.O. Shamim, I. Dincer, G. Naterer, Thermodynamic analysis of solar-based photocatalytic hydrogen sulphide dissociation for hydrogen production, International Journal of Hydrogen Energy, 39 (2014) 15342-15351.

[2] C.K. Ong, S. Dennison, S. Fearn, K. Hellgardt, G.H. Kelsall, Behaviour of Titanium-based Fe₂O₃ Photo-Anodes in Photo-Electrochemical Reactors for Water Splitting, Electrochimica Acta, 125 (2014) 266-274.