Benchmarking Classical Molecular Dynamics Simulations for Computational Screening of Lithium Polymer Electrolytes

Polymer electrolytes may play a crucial role in the development of safe, efficient energy-dense batteries thanks to their unique ability to facilitate ion transport while maintaining structural stability. However, experimental discovery is limited by the complexity of synthesizing and testing new monomer and polymer chemistries. In this study, we benchmark the ability of molecular dynamics (MD) simulations with Class 1 force fields to model the transport and structural properties of polymer electrolytes in a high-throughput screening setting. By systematically comparing simulation results with experimental data for 19 polymers, we evaluate the effect of simulation choices in predicting key transport properties. In particular, we evaluate convergence of diffusivities and conductivities as a function of simulation length, and how inaccuracies in modeling polymer glass-transition temperature carry over to ion transport properties. The results highlight both the strengths and limitations of affordable high-throughput MD simulations for these complex systems, providing insights into the optimization of MD simulations for polymer electrolyte research, and recommendations for modeling choices with optimal cost-accuracy trade-offs. Furthermore, we perform in-depth transport and structural property analysis across the polymer space to gain insights into the design of new polymer electrolytes.