An Electronically Coupled Interfacial Energy Descriptor for Predicting Coulombic Efficiency in Aqueous Zinc Batteries

Solid-liquid interfaces dictate the performance and longevity of aqueous batteries. While rationally designed electrolyte additives offer a potent strategy to stabilize these interfaces, the complex interactions between additives and metal surfaces are difficult to isolate experimentally, hindering the predictive design of optimal additives. Here, we introduce an electronically coupled interfacial energy descriptor (ΔG) that quantitatively captures additive-metal interactions. By unifying the energetic contributions of additive adsorption and interfacial charge transfer at the zinc surface into a single metric, ΔG provides a fundamental measure of interfacial stability. Integrating this descriptor with intrinsic molecular properties, we develop a machine learning framework that accurately predicts Coulombic efficiency across different additive families, successfully identifying pentaerythritol (PTT) as a highly effective candidate with a minimized ΔG value. Formulated with PTT additive, the ZnSO4 electrolyte enables highly reversible zinc plating and stripping with an average Coulombic efficiency of 99.71%, sustaining stable cycling for over 1000 h, outperforming additives predicted by intrinsic molecular descriptors with a 2-fold longer cycle life. The resulting Zn-I2 full cells also sustain stable cycling under high-temperature and lean electrolyte conditions. Mechanistically, additives with minimized ΔG values preferentially accumulate at the interface to expel water molecules, promoting the formation of a robust solid electrolyte interphase composed of PTT- and anion-derived species. By redefining interfacial interactions through electronic-energetic coupling, this work provides a physically interpretable data-driven framework for additives screening in aqueous Zinc-ion batteries.