Articles techniques et présentations

Electrochemical CO2 reduction provides a route to convert CO2 into valuable chemicals using renewable electricity, but practical deployment requires stable, scalable, and inexpensive cell components. One major limitation is the anode, as many high-performance CO2 electrolyzers still rely on costly and scarce Ir-based oxygen evolution reaction (OER) catalysts. Nickel foam is an attractive non-noble-metal alternative because it is conductive, porous, inexpensive, and stable under alkaline OER conditions. However, strongly alkaline electrolytes promote CO2 capture as carbonate/bicarbonate, lowering carbon efficiency. Near-neutral KHCO3 electrolytes can reduce this carbonate penalty, but Ni stability becomes more challenging because the local anode environment can become substantially more acidic than the bulk electrolyte. This work uses COMSOL Multiphysics® modeling to identify operating conditions that stabilize porous Ni foam by controlling the local Ni/electrolyte surface pH. A two-dimensional representative pore-scale model was developed in COMSOL Multiphysics® to resolve the electrolyte-filled pore space within porous Ni foam. The model employs Tertiary Current Distribution module to couple dilute-solution Nernst-Planck species transport, electroneutrality, homogeneous bicarbonate/carbonate buffer reactions, and OER-driven proton generation at the Ni/electrolyte interface. The solid Ni phase was treated as electronically conductive and equipotential, while convection inside the pore was neglected to isolate diffusion, migration, and buffer-mediated pH regulation. The main model output is the area-averaged local surface pH, which is used as a stability-relevant descriptor for Ni dissolution risk. COMSOL Multiphysics® was used to solve the coupled finite-element problem, and a fully automated Python-based workflow was developed to update model inputs, run simulations, and extract outputs without manual intervention. To accelerate screening, three COMSOL simulations were run concurrently. The automated sweep covered 15 current densities from 50 to 300 mA cm-2 and 20 KHCO3 concentrations from 0.02 to 2 M. No COMSOL Application Library model was used, the geometry, governing equations, boundary conditions, automation workflow, and post-processing procedure were developed specifically for this study. The simulations show that local pH is controlled by the balance between OER-generated H+ and bicarbonate/carbonate buffering. Low KHCO3 concentrations cause strong interfacial acidification and increased Ni dissolution, while higher buffer concentrations maintain a less acidic surface and suppress Ni leaching. Time-dependent simulations further show that continuous MEA operation causes stronger acidification because of diffusion-limited buffer replenishment, whereas on/off operation enables partial pH recovery during rest periods. Guided by these modeling insights, the optimized near-neutral CO2 electrolysis system operated for more than 500 h while maintaining a combined C1 Faradaic efficiency toward formate and CO of approximately 70 – 80%.