Accelerated Discovery of Solvation Structure Engineering for Stable Aqueous Rechargeable Zinc Batteries via Physics‐Guided Bayesian Active Learning

Aqueous rechargeable zinc batteries, despite advantages like safety and performance, struggle with water-based side reactions such as hydrogen evolution and corrosion. Regulating the solvation structure of Zn²⁺ is essential for stability. Introducing n-hexane, a nonpolar alkane, modifies Zn²⁺ coordination and stabilizes the Zn anode-electrolyte interface. The miscibility of n-hexane is improved through the formation of an oil-in-water macroemulsion with amphiphilic Zn(OTf)₂ and β-cyclodextrin. Macroemulsion stability is highly sensitive to component concentrations, requiring precise balance to ensure proper electrolyte function. However, designing multi-component electrolytes remains empirical. To address this, a Bayesian optimization framework is presented, incorporating physical relationships into machine learning to efficiently explore the design space. This approach rapidly identifies the critical concentration for macroemulsion stability, which is key for maintaining phase stability in the electrolyte. The optimized electrolyte maintains a low overpotential (30 mV) for over 1300 h in a Zn||Zn symmetric cell, with a current density of 1 mA cm^-2.