Modelling and Optimisation of Battery Electrode Slurry Preparation Process by Extrusion

EPSRC Research Area: Energy

Mixing is the first controllable process in battery manufacturing, where the characteristics of the slurry and the associated coated electrodes could be defined and optimised. Compared to batch mixing, the extrusion mixing technique has the ability to mix electrode active material with binders and additives while using significantly less solvent, or even no solvent at all. This not only reduces the environmental impact of the solvent-based electrodes (currently a big share in the current battery market), but also facilitates continuous mixing practice and quality assurance. It is less prone to process uncertainties and operator errors compared to batch mixing and offers better flexibility for real-time control and optimisation.
The research and technological gap here is that the impact of the extrusion process factors (e.g. feed rate, temperature, pressure) on the slurry, electrode, and finished cell characteristics is not very clear. Because the technology is much younger (compared to batch mixing) in the battery domain, and includes too many control factors, systematic methods for controlling/optimising the factor combinations for a desired characteristic are not available, and the existing trial-and-error approach is very time- and resource-consuming. Therefore, the research questions to be answered in relation to the identified gap here is: "What is the impact of the extrusion mixing process key factors in combination with the equipment structure (e.g. screw configuration, dimensions) on the electrochemical, mechanical and structural characteristics of the slurry, electrodes, and final battery cells?", and "How to use the information for optimisation purposes?".
For these questions to be answered, a model-based process identification and control methodology will be adopted. By developing a customised (to the battery manufacturing) physics-based model, the mechanics of the continuum media will be simulated. This would facilitate the evaluation of free surface flows during the extrusion process and reveal the interaction/correlation of the key process variables with the slurry and electrode characteristics. By allowing the integration of sensors for continuous measurements of key factors such as temperature and pressure.
Experimental data from an existing project will be used and complementary experiments in the EIC's Twin-Screw Extruder Equipment will run for validation purposes. For benchmarking purposes, validations will be aimed on the commercial material, with the potential expansion to other technologies such as solid-state batteries.
The model will then be investigated for being integrated into a slurry-electrode-cell optimisation and control framework. leveraging the surrogate data-driven (AI) modelling approach. While the physics-based models might struggle to directly perform in the optimisation loops due to computational complexity, a surrogate version is expected to be integrable more efficiently.