.png){kind=icon}
.png){kind=icon}
Return
Molecular dynamics and machine learning framework for predicting ion transport and mechanical properties of ionic liquid@polyvinylidene fluoride gel polymer electrolyte
.png){kind=icon}
PRE
AI
K
DOI：10.1016/j.jiec.2025.11.034.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Advancing high-performance gel polymer electrolytes (GPEs) is critical for the development of safe, efficient, and mechanically robust lithium-ion batteries (LIBs). In this work, we present a synergistic framework that integrates molecular dynamics (MD) simulations with machine learning (ML) to predict and rationalize the ionic conductivity and mechanical properties of ionic liquid@polyvinylidene fluoride (IL@PVDF)-based GPEs. A comprehensive analysis of MD-derived descriptors, including the lithium-ion diffusion coefficient, density, Pugh's ratio, and elastic moduli, reveals the underlying structure-property relationships. Benchmarking five ML regressors using log-transformed conductivity data identifies Extreme Gradient Boosting (XGB) as the most accurate, generalizable, and robust predictor. SHapley Additive exPlanations (SHAP) analysis provides mechanistic interpretability, highlighting the dominant roles of ion diffusion and matrix flexibility in governing conductivity. Furthermore, SHAP interaction plots uncover nonlinear synergies between ion mobility and ductility, demonstrating that transport and mechanical performance can be simultaneously optimized. Collectively, this MD-ML framework offers a predictive, interpretable, and scalable strategy to accelerate the rational design of nextgeneration IL-based GPEs, thus providing actionable guidance for high-throughput screening and advanced electrolyte engineering in LIBs.
Keywords:
Gel polymer electrolytes
Molecular dynamics
Machine learning
Lithium-ion batteries
Journal
IF:
6
Papers:
8.6K
Citations:
3.2W
Organization
Cited Papers