Advanced hard carbon materials for practical applications of sodium-ion batteries developed by combined experimental, computational, and data analysis approaches
Published in Progress in Materials Science, 2025
Role: Second author; led the data-engineering and machine-learning components of the study.
Contribution: Developed the literature-data collection software and the associated data-processing pipeline; organized and standardized precursor, synthesis, structural, and electrochemical variables from published studies; and performed data cleaning, missing-data imputation, machine-learning modelling, and SHAP-based model interpretation.
Research Summary
This study combines experimental findings, computational insights, and literature-derived data to investigate the processing–structure–performance relationships of hard-carbon anodes for sodium-ion batteries. The data-driven component converts heterogeneous information reported across the literature into a structured dataset covering precursor characteristics, synthesis conditions, microstructural descriptors, initial Coulombic efficiency, and reversible capacity.
Random Forest models were used to analyse nonlinear relationships among these variables, while SHAP-based interpretation was employed to evaluate the direction and relative importance of individual features. The resulting analysis provides interpretable insights into how precursor selection, thermal treatment, pore structure, defect characteristics, and related material properties are associated with sodium-storage performance.
Beyond the hard-carbon system, the workflow demonstrates how literature-data collection, missing-data treatment, machine learning, and model interpretation can be combined to support the analysis of other energy-storage materials.
Recommended citation: Sun, Z.; Liu, H.; Li, W.; Zhang, N.; Zhu, S.; Chen, B.; He, F.; Zhao, N.; He, C. (2025). "Advanced hard carbon materials for practical applications of sodium-ion batteries developed by combined experimental, computational, and data analysis approaches." Progress in Materials Science 149, 101401.
Download Paper