Abstract
CO dimerization is a key step in the electrochemical reduction of CO(2) to multicarbon (C(2+)) products at low overpotentials. Although Cu(100) is uniquely active for this process, its performance remains limited, and the mechanisms behind improved activity and selectivity through alloying are not fully understood. Here, we combine machine-learning screening with constant-potential density functional theory simulations to systematically investigate CO dimerization on dilute CuM(100) alloys. p-block metals, particularly Al and Ga, make the reaction exothermic and lower the activation barrier relative to pure Cu, with Al showing the highest activity. Charge analysis along the reaction path reveals that electron donation from these heteroatoms stabilizes the CO dimer intermediate, enabling efficient C-C coupling under operating conditions. This behavior is captured by a strong linear correlation between reaction energy and excess surface charge at fixed potential, introducing a physically grounded descriptor that integrates covalent and electrostatic contributions to reaction energetics. Our findings reveal that excess surface charge can serve as a practical reactivity descriptor that directly correlates with C-C coupling activity and guides the rational design of more efficient CO(2)RR electrocatalysts.