Role of Reservoir and Diffusion for Electrocatalytic CO₂ Reduction to Ethylene on Cu

Electrochemical CO₂ reduction (EC-CO₂R) on Cu is a promising approach to produce value-added-chemicals using renewable feedstocks. The ability of Cu to produce multi-carbon products during electrochemical CO₂ reduction is conventionally understood in terms of the binding energy of the crucial CO intermediate and the existence of active sites which facilitate C-C coupling. Yet the vast differences in activity and selectivity towards single and multi-carbon products produced by various Cu preparations suggest that this picture is incomplete. In this work, we find that the effective catalytic activity towards ethylene improves with a larger fraction of <b>less active</b> sites acting as reservoirs of *CO on the Cu surface. This surprising finding can be explained by the rapid supply of *CO from reservoir to active sites which increases the effective activity of active sites. Experimental evidence is provided by real-time measurements of mass flow exiting a gas diffusion electrode cell held at a fixed potential and subjected to different precursor feeds. When the precursor feed is switched from inert Ar to CO, CO reduction to C₂+ products begins promptly, as expected. However, when the precursor feed is switched from CO to Ar, CO reduction continues for several seconds (delay time) before transitioning to hydrogen evolution. Control experiments establish that some Cu preparations, most notably oxide-derived Cu, can maintain COR activity in the absence of the gas phase precursor due to reservoir sites on the Cu surface that bind CO but do not convert it. We introduce a microkinetic model which shows that the delay time originates from a less active *CO reservoir that exhibits fast diffusion to active sites on the surface of OD-Cu. Experiments find that delay time and *CO reservoir sizes are affected by the catalyst preparation, applied potential, and microenvironment. Notably, we associate large *CO reservoir surface coverages (maximally 86±5% at -1.52 V vs. SHE) with increased ethylene production. We also estimate that *CO can travel long distances (up to 10s of nm) prior to reaction. We propose a broader scope for EC-CO₂R catalyst design: in order to design high C₂+ activity, in addition to active sites, the role of less active sites in controlling the *CO availability must be considered.