About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
2026 Technical Division Student Poster Contest
|
| Presentation Title |
SPG-8: Machine Learning–Guided Design of SEI Architectures to Mitigate Lithium Dendrite Formation |
| Author(s) |
Arjun Kulathuvayal, Yanqing Su |
| On-Site Speaker (Planned) |
Arjun Kulathuvayal |
| Abstract Scope |
Lithium dendrite formation remains a critical challenge for the safety and longevity of lithium-ion batteries (LIBs). Dendrites often originate from the Solid Electrolyte Interphase (SEI) formed on electrode, where heterogeneous Li-ion diffusion channels with varying activation energies promote localized ion accumulation and non-uniform growth. In this work, we present a machine learning (ML) framework that integrates density functional theory (DFT) energetics with experimental electrochemical parameters to uncover the complex relationships between SEI composition, morphology, and Li-ion transport. Our framework extracts descriptors from multicomponent SEI models, considering atomic arrangements, exposed surface orientations, and current density, to predict site-specific activation energies and dendrite growth probability. Beyond predictive capability, the algorithm suggests optimized SEI architectures that redistribute low-barrier diffusion channels more uniformly, mitigating dendrite initiation. This approach enables the rational design of artificial SEIs, advancing the development of safer, more stable next-generation LIBs. |
| Proceedings Inclusion? |
Undecided |
| Keywords |
Machine Learning, Modeling and Simulation, Energy Conversion and Storage |