Revisiting nuclear quantum effects on the mechanism of electrochemical reduction of CO at Cu(111) surface.

Abstract

The initial hydrogenation step of *CO on Cu(111) dictates product selectivity in electrochemical CO2 reduction, yet its mechanism remains controversial. Here, at an acidic aqueous Cu(111) interface, we systematically compute the free energy profiles for *CHO and *COH formation via multiple pathways, including direct *H transfer, water-assisted shuttling, and proton-coupled electron transfer (PCET), using machine learning force fields to enable statistical sampling within a grand canonical constrained path integral hybrid Monte Carlo framework that incorporates constant electrode potential, explicit solvation, and nuclear quantum effects (NQEs). At the investigated potential of U = -0.4 V vs SHE and 300 K, *COH does not remain a stable intermediate within the present explicit-solvent Cu(111) model and spontaneously dissociates into *CO and solvated H3O+. For *CHO formation, solvated protons are identified as the primary hydrogen source, predominantly proceeding via the PCET pathway. Surface-adsorbed *H atoms can also reduce *CO, but only through a two-step water-assisted shuttling mechanism rather than concerted transfer. Incorporating NQEs, the activation free energy of the PCET step exhibits a 30% reduction (by 0.12 eV) relative to the quantum barrier, indicating a non-negligible difference in the estimated reaction rate. This work highlights the synergistic importance of explicit solvation, constant-potential conditions, and NQEs in modeling electrochemical interfaces.