Lawrence Berkeley National Laboratory

The criteria dictating the performance of gas diffusion electrodes (GDEs) for CO2 electrolysis are not well understood, due to the complex and highly coupled relationships of the underlying physical and chemical phenomena. In addition, a number of key performance indicators (KPIs) have been identified (e.g., Faradaic efficiency, reaction selectivity, single-pass conversion, and productivity), and optimizing for any single metric often leads to inherent tradeoffs. Consequently, much recent work has focused on understanding how operational and architectural parameters control GDE performance. The porous catalyst supports and diffusion media in these devices are critical components as they mediate the transport and reactive processes. The microstructure of these components influences the balance between the electrochemical surface area, which dictates CO2 consumption, and mass transfer of the aqueous and gaseous species. Traditional porous media is often spatially homogeneous and provides limited opportunity to tailor this balance. In contrast, novel advanced manufacturing methods have now enabled researchers to create electrodes with variable porosity that can be tuned for optimal performance. To explore the impact of locally altering the porous media structure, we leverage previous modeling work by Weng et al. to allow for spatially varying porosity. Further, we couple the forward simulation to adjoint based optimization algorithms to determine optimal porosity distribution in the diffusion media and catalyst layer of a GDE. The cost functions for the optimization are derived from the previously mentioned KPIs. Finally, the performance of the resultant spatially varying porosity GDEs is compared to the performance of homogenous porosity GDEs, and we identify key features of the porosity distribution leading to improved performance. Weng, Lien-Chun, Alexis T. Bell, and Adam Z. Weber. "Modeling gas-diffusion electrodes for CO2 reduction." Physical Chemistry Chemical Physics 20.25 (2018): 16973-16984. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. LLNL-ABS-830082