(237c) Modeling Hydrogen Crossover in Proton Exchange Membrane Water Electrolyzers with and without Recombination Catalyst Interlayers

Low-temperature proton exchange membrane water electrolyzers (PEMWEs) are poised to undergo rapid expansion in the coming decades for clean hydrogen production. Hydrogen has been hailed as the ‘Swiss Army knife’ of decarbonization as it can help mitigate carbon dioxide emissions from ammonia production, metal refining, and heavy-duty vehicle transportation. Proliferation of clean hydrogen production necessitates large reductions in iridium loadings in the anode while also operating at higher current density values. The latter goal can be achieved by adopting thinner proton exchange membranes (PEMs), but the use of thinner membranes exacerbates hydrogen crossover – which leads to unacceptable levels of hydrogen in the anode effluent poising safety concerns (i.e., above the lower explosion limit of 4 vol% hydrogen in oxygen). Notably, the hydrogen content in the anode effluent is high when operating the PEMWE at low current density values (0.1 to 0.5 A cm^-2). A PEMWE will sometimes throttle down and operate at low current density because of the intermittent nature of renewable energy sources which are used to power the electrolyzer. Recombination catalyst (RC) interlayers, located within the PEM, can consume hydrogen and mitigate the hydrogen concentration in the anode effluent. This Talk presents our continuum-level modeling approach for hydrogen crossover in PEMWEs with polymeric membranes containing RC interlayers. The model accurately captures the hydrogen content in the anode effluent as a function of the PEMWE current density, cathode back pressure, membrane thickness, anode backpressure, RC location, and RC loading. The modeling results, as well as experimental observations in the literature, suggest that RC interlayer effectiveness is limited by oxygen availability at the RC interlayer. The low oxygen concentration at the interlayer relative to hydrogen arises from the lower solubility of oxygen in hydrated PEMs and the lower diffusivity of oxygen in hydrated PEMs. The Talk concludes with future directions to improve the model as well as suggestions for future experiments to aid the accuracy of the model.