Dec 3, 2024
3:15pm - 3:45pm
Hynes, Level 3, Room 300
Anna Hankin1,George Creasey1,Arend Moelich2,John Rodriguez Acosta1,Thomas Shalvey3,Dora Alicia Garcia-Osorio3,Jon Major3,Alexander Cowan3,Craig McGregor2,Andreas Kafizas1
Imperial College London1,Stellenbosch University2,University of Liverpool3
Anna Hankin1,George Creasey1,Arend Moelich2,John Rodriguez Acosta1,Thomas Shalvey3,Dora Alicia Garcia-Osorio3,Jon Major3,Alexander Cowan3,Craig McGregor2,Andreas Kafizas1
Imperial College London1,Stellenbosch University2,University of Liverpool3
We have designed, built and characterized a prototype photoelectrochemical demonstration system capable of splitting water into hydrogen and oxygen using only photon energies. The reactor was operated during Mar-May 2024 at Stellenbosch University (33.93° S, 18.86° E), while mounted on a 2-axis tracking platform. Light was directed laterally both into the (photo)cathode and photoanode compartments, which were separated by an ion-permeable membrane. Double-sided irradiation was achieved by two methods that were compared with each other: (i) using mirrors (Ag-coated mirror for the (photo)cathode side and Al-coated mirror for the photoanode side) and (ii) linear Fresnel lenses coupled with stepped Al waveguides. The latter irradiation method delivered light, concentrated by a factor of up to 4, though theoretical simulations show that through design improvement the concentration factor could ultimately reach ≈ 15.<br/><br/>The reactor was operated in two modes:<br/>Photoelectrochemical (PEC), utilising an FTO|WO<sub>3</sub>|BiVO<sub>4</sub>|NiFeO<sub>x</sub> photoanode and a FTO|Au|Sb<sub>2</sub>Se<sub>3</sub>|CdS|TiO<sub>2</sub>|Pt photocathode<sup>[2]</sup>;<br/>PV-assisted photoelectrochemical (PV-PEC), utilizing the same FTO|WO<sub>3</sub>|BiVO<sub>4</sub>|NiFeO<sub>x</sub> photoanode, Ni cathode and an externally mounted c-Si PV.<br/><br/>In mode I, a pH gradient was employed to assist water splitting, with a pH ≈ 0.8 aqueous catholyte comprising 0.1 M H<sub>2</sub>SO<sub>4</sub> and anolyte comprising 1 M H<sub>3</sub>BO<sub>3</sub> + 1 M NaOH at pH ≈ 9.3. In mode II, both electrolytes were 1 M H<sub>3</sub>BO<sub>3</sub> + 1 M NaOH. A cation-permeable membrane, Nafion<sup>TM</sup> 115, was utilized in all experiments. The reactor was operated in batch recycle mode. The areas of the (photo)electrodes and the PV were all 30 cm<sup>2</sup>.<br/><br/>We observed that our bismuth vanadate (BVO) photoanodes usually degraded within hours, for which we propose two reasons. Firstly, when coupled with c-Si PV, the potential of the photoanodes was observed to increase into the dark current regions under low irradiance. While the c-Si PV is able to generate a significant photocurrent even on cloudy days, the bismuth vanadate photoanode is unable to match this through its own electron-hole generation. When anode potentials exceeded ≈1.1 V (RHE), the photoanode is thought to have degraded through oxidation of the bismuth; the degradation was irreversible. Secondly, the photoanodes degraded equally quickly under concentrated irradiance; we are currently investigating whether this was caused by overheating, high flux of bubbles or both. It is currently unclear whether the issue is with the adhesion of the WO<sub>3</sub> layer to FTO or due to the bismuth film itself.<br/><br/>I shall discuss the experimental results from reactor testing, the performance under various modes of irradiation, and the effects of electrode materials, geometries and relative configurations within the reactor on its design, overall performance and further scale-up, as well as the future role of photoelectrochemical systems in energy storage.