Monitoring the State of Health (SOH) of Green Batteries (GreenBat)

Abstract

Here we focus on the study of state of health monitoring of secondary batteries with advanced “green” formulations. European recycle map on battery energy storage plans to replace positive electrodes consisting of cobalt (Co) and fluorinated (F) graphite-based negative electrodes constituents with environmentally sustainable alternatives. Recently, "green formulations" was validated; in particular, recent trends suggest the use of electrodes based on silicon (Si), Co-free manganites (LiMnO2) aqueous soluble binders and fluorine-free electrolytes [1]. Our aims are to describe the health and the degradation of the cells with electrochemical techniques and to describe the chemical and morphological changes occurring during the battery cycling, through spectroscopies (Raman and IR) operating in parallel with impedance analysis. These experimental data will be integrated together in a multiblock dataset, which will constitute the basis for chemometric processing. The purpose of this chemometric modelling is to correlate SOH evolution to the chemical and physical state of the battery [1]. Here we illustrate the analysis of SOH of two different pouch cells formulation constituted by LTO (Lithium-Titanate) as anode, LFP (Lithium-Iron-Phosphate) as cathode and the LP30 electrolyte (EC:DMC 1:1 and LiPF6 1M); the second formulation is constituted by a thin film layer of commercial Li metal (purity of 99.7%) as anode, LFP as cathode and an ethers based formulation as electrolyte. After formation cycles (CC-CV), the cells have been submitted to aging tests. The different operating conditions are chosen with an Experimental Design (ED) and supported by a PCA analysis. The PCA analysis it was used to choose the operating parameters for the subsequent aging experiments. Is illustrated also the analysis of SOH of a coin cell formulation consisted by a LRLO (Lithium Rich Layer Oxide) as anode, commercial Li metal layer as cathode and LP30 additive with a ionic liquid (IL) as electrolyte. Turning to the chemometric modelling, as a starting point we verified the ability of PLS regression applied to the voltage profiles during charge cycles to estimate the SOH of a benchmark dataset by Lin et al. [1] References [1] M. Lin,D. Wu, Journal of Power Sources, 2022, 518, 230774. Figure 1