When do constant-charge and constant-potential computational simulations agree? A dipole-based framework for predicting electrochemical surface coverages

Surface coverage effects play a central role in electrocatalysis, yet they are still most often analyzed computationally using vacuum models that neglect proper treatment of electrolyte pH, applied potential, and solvation. Although constant-potential approaches can capture these effects explicitly, they remain computationally demanding, which motivates the continued use of simpler constant-charge simulations. Here, we combine constant-potential density functional theory within the grand canonical ensemble with conventional constant-charge calculations to determine aqueous surface coverages on the (111) facets of Ni, Pd, Pt, Cu, Ag, and Au. By directly comparing the two frameworks, we identify when the less expensive constant-charge approach reproduces constant-potential energetics and show that its reliability is governed by adsorption-induced shifts in the potential of zero charge. Across all systems considered, the potential of zero charge exhibits a universal linear correlation with the surface-normal dipole moment, enabling the dipole moment to serve as a simple and transferable descriptor for predicting such shifts and, in turn, for assessing whether an explicit constant-potential treatment is required. As a demonstration of the approach, phase-transition potentials obtained from grand canonical calculations are compared directly with experimental electrochemical measurements and found to be in good agreement. Together, these results establish a practical framework for connecting constant-charge and constant-potential simulations, enabling more efficient computational screening and mechanistic analysis by identifying cases in which constant-potential treatments can be avoided without compromising accuracy.