Study and Optimization of the Formation Step of Next Generation Li-Ion Cells Using Design of Experiments (DOE) Methodology

The importance of the formation step in the Li-ion cell assembly process is often underestimated both in terms of economic impact due to its elevated cost 1 (Left Figure 1, with more than one third of cell finishing cost), but also in terms of electrochemical performance influence mainly driven by the Solid-Electrolyte-Interphase (SEI) formation during this first charge of the Li-Ion cell after its activation by electrolyte 2. In this study concerning the optimization of the formation step, we will use the design of experiments (DOE) methodology as this approach is a powerful tool that allows evaluating the effect of numerous parameters (including their linear, quadratic or interaction terms) on selected responses of interest with a limited number of trials when the experimental resources (including testing time, number of cells...) are limited. The formation step is generally divided in several sub-steps that includes many parameters 3,4 with some of them illustrated on the right figure 1: different current densities during the successive constant current steps, the different temperatures associated for each of these sub-steps that can also be of different durations and/or associated with limitations in states of charge. Additionally, specific electrolyte additives maybe used to improve the efficiency and quality of the formation step especially to modify the chemistry of the SEI.... In our case, we present here a first approach using a screening design that is a specific type of DOE particularly adapted when the number of factors is elevated. Screening designs especially allow precising the first order input factors on the selected responses of interest, in our case Initial Coulombic Effciciency (ICE), resistance, discharge capacity; cyclability... The experimental data are generated thanks to the use of industrial dry ( i.e. unactivated) NMC532/Graphite pouch-cells of 0.435 Ah. These dry cells allowed us including electrolyte factors in the DOE. Considering 8 different factors, a total of 17 trials were defined in the DOE that was set with MINITAB software. The full analysis of the DOE will be presented with the main effects for each response combined with the Pareto diagrams. These results will give general trends to define better conditions for the formation step and improve the cell performances. In a next step, we plan to pursue this approach with a surface response design on the selected first order factors, to enable a quantification of the performances and an optimization of the formation step. The study will also be completed by post-mortem analysis such as XPS studies to better correlate the formation step parameters with the SEI characteristics (chemistry, thickness...). Boston Consulting Report (2018). E Peled, S. Menkin. SEI: Past, Present and Future. Electrochem. Soc. 164 (7), 2017, A1703 -C. Chiang, M.-S. Wu, J.-C. Lin. A Novel Dual-Current Formation Process for Advanced Lithium-Ion Batteries. Electrochem. Solid State Letters 8 (8), 2005 , A423 S. Zhang, K. Xu, T. R. Jow. Optimization of the Forming Conditions of the Solid-State Interface in the Li-Ion Batteries. J. Power Sources 130 (1-2) 2004 , 281–285 Figure 1