Machine-learning-assisted energy evolution analysis of lysine oligomers reveals ε-polylysine as an electrolyte additive for stable aqueous zinc batteries

Aqueous zinc-ion batteries (AZIBs) are promising for large-scale energy storage owing to their intrinsic safety and low cost; however, their practical operation is restricted by hydrogen evolution, corrosion, byproduct accumulation, and dendritic Zn growth at the Zn anode. Considering the structural complexity of polymer additives and the difficulty of direct high-accuracy calculations, in this work, an integrated “oligomer-based density functional theory (DFT) calculation-machine-learning- assisted energy evolution analysis-experimental validation” framework for polylysine chain-length-dependent energetics was developed. Lysine oligomers were used as model systems to calculate the molecular energy, additive-water-binding energy, and adsorption energy on Zn surfaces, which were further correlated with the degree of polymerization, functional-group density, charge distribution, and molecular size. The analysis revealed a nonlinear chain length–dependent energy evolution of polylysine, whereas ε-polylysine (ε-PL) lies within a relatively stable energy-response region. Owing to its chain-like multidentate structure, ε-PL synergistically reconstructs the hydrogen-bonding network of water and protects the Zn interface. The experimental results demonstrate that ε-PL facilitates interfacial Zn 2+ kinetics involving ion transport, partial desolvation, and charge-transfer processes, while forming an N/O-rich interfacial environment that suppresses parasitic reactions and nonuniform Zn deposition. Consequently, the ε-PL-modified electrolyte enabled Zn||Cu cells to cycle for 2000 cycles with a Coulombic efficiency of 99.2%, enabled Zn||Zn symmetric cells to operate for more than 4000 h, and enabled Zn||AlVO-NMP full cells to retain 70.63% capacity after 6000 cycles. This study provides a new strategy for the machine learning-assisted design of polymer electrolyte additives for AZIBs.