Non-Destructive Degradation Diagnostics of Lithium-Ion Cells Considering Aging-Related Changes in the Shape of the Half-Cell Open Circuit Potential Curve of Silicon-Graphite

Abstract

Analyzing changes in the quasi-stationary open circuit potential (OCP) curve during aging is a well-established method for the non-destructive diagnosis of degradation occurring in lithium-ion cells during calendar and cycle aging. It allows detecting and quantifying different degradation modes, especially loss of lithium inventory and loss of active material at a specific electrode, without the need of opening up the cells and performing post-mortem analysis. In order to obtain information on the degradation, the quasi-stationary OCP curve measured from an aged cell is reconstructed based on the half-cell OCP curves of the electrode materials. For this purpose, both half-cell OCP curves are linearly scaled and relatively shifted against each other until the difference between cathode and anode OCP fits the measured full-cell OCP. The parameters for the scaling and shifting of the half-cell OCP curves obtained during this optimization procedure are quantitative indicators of the degradation modes that have occurred in the full-cell. Usually, half-cell OCP curves obtained from pristine electrode material are used in this process and the shape of the half-cell OCP curves is assumed invariant during aging.While this assumption is probably justified for graphite and most established cathode materials, there are a number of recent publications describing a change in the shape of the OCP curve of blended silicon-graphite electrodes during cycle aging. This change is probably mainly caused by a decrease in the relative contribution of the silicon to the overall electrode capacity because of its faster degradation in comparison with the graphite. In this study, we investigate the influence of aging-induced changes in the shape of the OCP of silicon-graphite on the applicability of the method for non-destructive degradation analysis described above for cells containing silicon-graphite as anode material. We also present possible extensions of this method to consider changes on the half-cell level.In the scope of this study, we cycle commercial lithium-ion cells containing nickel-rich nickel manganese cobalt oxide (NMC-811) as cathode material and silicon-graphite as anode material until different degradation stages are reached. We then measure the quasi-stationary OCP of the full-cell and subsequently open the cells to extract samples from both electrodes. Afterwards, we measure the quasi-stationary OCP curve of the electrode samples by building coin-cells containing an electrode sample as cathode and lithium metal as anode.We reconstruct the full-cell OCP curves at different degradation stages by linear scaling and shifting of half-cell OCP curves both obtained from pristine electrodes and from the electrodes extracted from aged cells as shown in figure 1. This allows us to analyze whether the accuracy of the full-cell OCP reconstruction can be improved by using half-cell OCP curves of aged electrodes. We also present the changes in the quantitative results concerning the occurrence of different aging modes depending on whether pristine or aged half-cell OCP curves are used to reconstruct the full-cell OCP.In addition to this, we present an extended method for non-destructive degradation diagnostics that contains silicon-specific active material degradation as an additional aging mode. We analyze to which degree this method is capable of determining the loss of silicon-related capacity in the anode from full-cell OCP measurements.These results are relevant for both science and application as lithium-ion cells containing blended silicon-graphite as anode material are increasingly used due to their high specific capacity. Adapting the methods for aging diagnosis to this anode material is necessary to obtain valid results from experiments investigating the degradation of full-cells containing silicon-graphite anodes. Furthermore, expanding the existing models describing the aging-related change in full-cell OCP curves by considering changes in the shape of the half-cell OCP curves is a first step for the development of reliable algorithms for electrode-specific state of health estimation for cells containing silicon-graphite. Such algorithms could then be used on-board as part of the functionality of battery management systems.Figure 1: Reconstruction of the open circuit potential curve of a cycle aged full-cell based on both half-cell open circuit potential curves obtained from pristine and cycle aged electrodes Figure 1