How Local Reaction and Process Conditions Influence CO2 Reduction to Multicarbon Products on Copper Gas-Diffusion Electrodes

Abstract

Electrochemical CO2 reduction is one technology being investigated to convert waste CO2 into value-added products due to the ability for the process to be performed under room-temperature conditions using intermittent electricity. Within the last 2 years, a number of new reports have moved to both gas-diffusion electrode (GDE) architectures and membrane electrode assemblies has increased greatly.Through these new architectures, a substantial amount of CO2 reduction studies have been performed using copper catalysts. The wide range of CO2 reduction products formed has further been shown to vary with catalyst, electrolyte, current density and even CO concentration. The latter parameter has been shown to play a critical role in CO2 reduction selectivity/activity as shown through Au-Cu tandem catalysis (Lum and Ager, EES, 2018; Morales-Guio et al, N. Cat., 2018) and by varying CO2:CO feedstocks (Wang et al. N. Nano., 2019). What is often-overlooked, however, is that a copper GDE will, by itself, produce a substantial amount of CO during CO2 reduction at current densities >300 mA/cm2.Understanding the role of CO(aq) in CO2 reduction is slowly being understood through various mechanistic studies, as well as some modelling work (Wu et al, JECS, 2015; Weng et al, PCCP, 2018), but as of yet there does not exist a tool for researchers to quickly take experimental data and predict the local concentrations of species and products within a GDE configuration. Such a tool would be invaluable for providing an often-overlooked aspect for analyzing experimental data.Here in this work, we begin to distill many of the complexities of GDE operation down to a singular crucial factor, the local reaction environment. Specifically, we use advanced mass transport models to predict both CO2 and product concentrations throughout the gaseous GDE and liquid catalyst layer using data from experiments on controlled Cu nanostructures. Of primary novelty, using experimental data (Faradaic efficiency and gas channel composition) we can understand transport to and from the catalyst’s surface for a variety of current densities and electrolytes, giving us the steady-state concentration of important species (CO2, CO, OH-, C2H4) at the electrode’s surface.Here we experimentally show that despite large differences in product selectivity of CO2 reduction at lower current densities (<200 mA/cm2) in 1M KHCO3, KCl and KOH, once a current density of 300 mA/cm2 is applied, we find that the selectivity, activity, local environment and potential required to reduce CO2 on a copper electrode are identical. All electrolytes were found to produce C2H4 selectivities of 50-55%. We then show that the concentration of CO(aq) within the catalyst layer is 2-3 times that of a CO-saturated electrolyte. Given the recent demonstration that even small amounts of CO can influence ethylene formation (Wang et al. N. Nano, 2019), we conclude that the presence of by-product CO already plays a strong role in the high observed ethylene selectivity in all previous GDE studies on copper. Extending our results further, we discuss what these findings mean for future studies that aim to maximize either single-pass conversion efficiencies or move to larger cell sizes.In short, we have shown with experiments and modeling that process conditions and the local reaction environment for CO2 reduction on copper electrodes in a gas-diffusion layer configuration are complex, but can be predicted. Without the field taking mass transport of copper GDE’s into consideration when analyzing experimental results, we feel that a critical piece of information for describing selectivity and activity is missing.