Transferable Drude-Based Polarizable Force Field Framework for Predictive Modeling of Battery Electrolytes

Predictive design of next-generation batteries hinges on the atomistic resolution of ion solvation structures that dictate interfacial stability and bulk transport. Capturing these features with quantitative fidelity remains challenging due to strong electrostatic coupling, many-body polarization, and competitive solvation within multicomponent environments. Conventional fixed-charge force fields, constrained by mean-field electrostatic approximations, fail to capture environment-dependent electronic responses across diverse chemical and concentration regimes. Here, we present a transferable polarizable force field framework based on the Drude oscillator model, employing a unified parameterization strategy systematically validated across neat and multicomponent systems. Extensive molecular dynamics simulations reproduce structural and dynamical observables with accuracy comparable to or exceeding state-of-the-art machine learning potentials and the commercial polarizable force field. The framework identifies a coherent set of mechanisms across chemically distinct electrolytes. Specifically, it reveals polarization-driven reorganization of ion solvation structures across concentration regimes, resolves competitive coordination in mixed ethers, elucidates energetic drivers of solvation preferences in cyclic and linear carbonates, and captures polarization-stabilized ion aggregation in fluorosulfonyl-based electrolytes while simultaneously recovering physically realistic transport dynamics. These findings position the proposed framework as a physically rigorous, computationally efficient, and open-source foundation for predictive electrolyte modeling across the vast chemical space.