Deep Learning-Based Local Pore Modeling and Gas Mass Transfer Simulation in Electrochemical CO2 Reduction Catalyst Layers

The electrochemical <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{CO}_{2}$</tex> reduction reaction (CO2RR) is a promising approach to combat climate change, with the catalyst layer (CL) of membrane electrode assemblies playing a crucial role in both the reaction and mass transfer. Traditional methods have limitations, as experimental measurements of the internal environment are difficult, and most simulation models treat the CL as a whole, ignoring its local pore structure. This work thus develops a local pore model of the CL using deep learning techniques to examine the effects of different ionomer contents on gas mass transfer. Results align well with existing studies: the CL with an I/C ratio of 0.2 has the highest gas diffusion distance (679.53 nm), indicating optimal mass transfer; the CL with an I/C ratio of 0.6 has the largest silver catalyst surface area, reflecting the highest number of active sites; and the CL with an I/C ratio of 0.4 shows the highest CO concentration in the high-concentration region, suggesting a balance of mass transfer and reaction efficiency. This model could serve as a useful tool and pathway to offer valuable insights for optimizing CL design and then improving the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{CO}_{2}$</tex> RR performance.