Electrolyte development for Li-ion batteries is a multi-objective challenge that balances oxidative stability and cathode electrolyte interphase (CEI) formation with reduction stability, solid–electrolyte interphase (SEI) formation, transport, and safety. We present a scripted, theory-guided screening workflow that couples Rational Discovery Toolkit ( RDKit) descriptors with automated DFT calculations in Psi4 to produce interpretable molecular proxies and a transparent, multi-criteria ranking. Sixteen representative solvents and additives spanning carbonates, sulfones, phosphates, ethers, and nitriles are evaluated. Quantum descriptors, such as HOMO, LUMO, and dipole moment, form the basis of the composite scoring function, which is combined using z-score standardization and auditable weights. Cheminformatics features, such as cLogP and TPSA, are computed and reported for interpretability but deliberately excluded from the score to prevent collinearity-driven distortion of the ranking. The resulting maps reproduce known trends, such as the effectiveness of EC and FEC, while highlighting adiponitrile as a stability-oriented candidate. The framework is explicitly scoped to bulk-solvent stability ranking; the low scores of well-known additives such as VC reflect the intentional design objective rather than a failure of the method. A separate additive-evaluation mode that incorporates decomposition-product thermodynamics and reduction-onset selectivity is identified as a key next step. Because all plots, tables, and workbooks are generated automatically, the workflow enables rapid iteration as new constraints, such as viscosity or cost, become available. Finally, a targeted literature benchmark aligns the computed ranking with established experimental behaviour and clarifies where formulation context dominates.
Munnangi et al. (Fri,) studied this question.