孙世刚院士领衔！厦门大学乔羽/程俊/邹业国最新JACS丨机器学习筛选外壳层共溶剂稳定锌负极SEI！

TL;DR - A JACS study used machine-learning molecular dynamics to screen co-solvents for aqueous zinc batteries, identifying DMAC as an additive that stabilizes the zinc anode interface. The resulting electrolyte improved zinc reversibility and enabled long-lived full cells.

• MLMD screened 28 co-solvents roughly 10,000× faster than AIMD and selected DMAC.
• DMAC restructures hydrogen bonding without strongly coordinating Zn²⁺, weakening Zn²⁺–H₂O interactions and promoting desolvation.
• Enhanced Zn²⁺–anion interactions produce a uniform ZnO/ZnF₂-rich SEI while suppressing hydrogen evolution and dendritic deposition.
• Zn∥Cu cells maintained 99.3% Coulombic efficiency for 950 cycles, while Zn∥I₂ cells cycled stably for 12,000 cycles.