Aspects of Reduced Performance of Proton-Exchange-Membrane Water Electrolysis Via Continuum Modeling

Abstract

Proton-exchange-membrane water electrolysis (PEMWE) is a technology, where hydrogen is electrochemically produced from water. Efficient industrial-scale production of hydrogen via PEMWE is contingent upon developing systems that are sufficiently resilient to operate under non-ideal operating conditions. Mathematical models can provide insight into the underlying physical mechanisms and inform the development of mitigating strategies. In the current work, a cell-level non-isothermal continuum model that is based on concentrated-solution theory that accounts for multiple ionic mobile species in the membrane and ionomer, energy transport, and fluid flow through porous media, was used to investigate two challenges to PEMWE becoming a viable technology, specifically contamination and hydrogen crossover. In cationic contamination, dissolved salts in the reactant water supplant protons in the membrane and ionomer and significantly increase the applied potentials required for operation. Hydrogen crossover is a concern especially in low-current operational modes, which can arise due to PEMWE powered by variable and cyclical renewable energy sources and can result in the production of combustible gas mixtures.For cation-contaminated cells, simulation results suggest that the uptake of the contaminant cation increased as a function of current density until local conditions in the cathode catalyst layer (cCL) allowed the contaminant desorption to occur. The model suggests that the increased cell potential is primarily due to proton activity losses in the hydrogen evolution reaction within the cCL and that ionic conductivity losses are insufficient to cause the performance losses observed experimentally. Agreement with experimental data lends the model as a possible method or predictive models for PEMWE cell recovery methods. Observations of hydrogen crossover simulations suggest several aspects of hydrogen crossover, which include that localized pressures within the catalyst layers can enhance crossover, bubble-nucleation kinetics limit the exhaust of product gases, and convective fluxes due to electroomostic drag are negligible. An improved understanding of the crossover mechanisms can inform the development of low-crossover designs and mitigation strategies. Figure 1