Machine Learning Driven Optimization of Electrolyte for Highly Reversible Zn‐Air Batteries with Superior Long‐Term Cycling Performance

Abstract Aqueous alkaline Zn‐air batteries (ZABs) have garnered widespread attention due to their high energy density and safety, however, the poor electrochemical reversibility of Zn and low battery round‐trip efficiency strongly limit their further development. The manipulation of an intricate microscopic balance among anode/electrolyte/cathode, to enhance the performance of ZABs, critically relies on the formula of electrolytes. Herein, the Bayesian optimization approach is employed to achieve the effective design of optimal compositions of multicomponent electrolytes, resulting in the remarkable enhancement of ZAB performance. Notably, ethylene glycol has been successfully employed as both electrolyte additive and fuel, playing key roles in changing the reaction pathways of ZABs, especially the storage form of discharge products from ZnO deposition on the anode to Zn 2+ ‐based hybrid particle colloids in the electrolyte. As a result, the as‐obtained novel ZABs can deliver superior battery reversibility and stability (1700 h at 2 mA cm −2 and 1400 h at 20 mA cm −2 ), greatly improved round‐trip efficiency as high as 76.3%, and even continuous discharge until complete Zn anode depletion. This work has demonstrated enormous potential for long‐term energy storage applications and holds promise for bringing new opportunities to the development of ZABs.