Taguchi method in experimental procedures focused on corrosion process of positive current collector in lithium-ion batteries

Abstract

Planning the design and discussing the results are routine tasks carried out daily by researchers all over the world. This study is focused on making these stages of research time- and cost-efficient without affecting the reliability of the data. Following the principles of the Taguchi method brings such a result. This approach ensures the quality of research by optimisation of the experimental procedure. In the presented work, the Taguchi method solves an exemplary research problem of the corrosion process of aluminium current collector in lithium-ion batteries. Operating factors selected for the optimisation include temperature (25, 35, 45 °C), the volume ratio of ethylene carbonate to diethyl carbonate in the solvent (1:1, 1:2, 1:4), and type of lithium salt (LiPF6, LiTFSI, LiBOB). Using three operating factors with three levels of variability requires data assembly in an orthogonal array L9 with nine experimental runs in total. In this study, the Taguchi approach is validated by performing a full-factorial procedure of twenty-seven individual experiments. The response of tested systems is analysed by cyclic voltammetry and electrochemical impedance spectroscopy, which enables the evaluation of corrosion parameters in each experimental run. The assessed parameters indicate the intensity of the aluminium corrosion depending on the operating conditions. These conditions cause unwanted variance, which accelerates corrosion and lowers the life expectancy of lithium-ion batteries. The Taguchi signal-to-noise-ratio analysis reveals that the most significant variation is caused by the lithium salt in the electrolyte (S/N LiBOB ≥ LiPF6 > LiTFSI). Temperature and alkyl carbonates play a secondary role. The Taguchi prediction method is also consistent with the full-factorial approach, which warrants reliability.