About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Fundamentals of Sustainable Metallurgy and Processing of Materials
|
| Presentation Title |
Kinetic Model Selection and Process Optimization for LIB Recycling Using Reinforcement Learning |
| Author(s) |
Irem Topsakal, Irmak Sargin |
| On-Site Speaker (Planned) |
Irem Topsakal |
| Abstract Scope |
Recycling of lithium from end-of-life lithium-ion batteries (LIBs) is critical for sustainable resource management. Conventionally, hydrometallurgical leaching is a key step in this process, but optimizing conditions for maximum recovery remains challenging because a quantitative understanding of process kinetics has yet to be achieved. This study presents a reinforcement learning (RL) model designed to both optimize leaching parameters and identify the most suitable combinations of kinetic models. A digital twin of the leaching process allows the RL agent to interact with and learn from the model environment using data collected from scientific literature. By optimizing both the recovery and the choice of kinetic model, the framework adjusts effectively to changes in reaction behavior and process conditions. This data-efficient, adaptive approach reduces reliance on experimental trial-and-error and supports the development of autonomous leaching operations. The results demonstrate the potential of AI-driven tools to enhance the sustainability of LIB recycling processes. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Recycling and Secondary Recovery, Hydrometallurgy, Machine Learning |