Data Mining Accelerates Layered Transition Metal Oxide Cathode Doping Design Optimization
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Currently, the doping of layered transition metal oxide anode materials in sodianion batteries has been extensively studied, but quantitative guidance for the design of doping elements is lacking. To address this problem, this paper combines high-throughput literature data mining with multivariate machine learning methods to construct a database containing key descriptors and electrochemical properties of doping elements. Random forests, gradient boosted decision trees, and multilayer perceptrons are used to reveal complex nonlinear relationships between doping parameters and material properties. The main findings include: the oxidation state of dopants is positively correlated with capacity retention; The electronnegativity is kept in the medium range of 1.3 ~ 1.6, which can significantly enhance the cycle stability. There is a particularly strong negative correlation between ion radius and 5C rate performance with a correlation coefficient of -0.64, indicating that the doping of smaller size elements can effectively optimize Na⁺ transport. Finally, interpretability analysis tools such as SHAP and PDP are used to verify the accuracy of the model and the synergistic mechanism between the doping parameters. This study provides a data-driven theoretical basis and optimization approach for the doping design of layered oxide cathodes, and promotes the rational development of high-performance cathode materials for sodium-ion batteries.
Fields of Study & Subjects
Machine Learning in Materials Science
Advancements in Battery Materials
Battery Modeling and Management
Publication details
- Journal / source
- IEEE International Conference on Power, Electronics and Green Energy (ICPEGE)
- Pages
- 19–24
- Publication date
- 28 January 2026
- Language
- EN