Lithium plating is one of the most critical ageing mechanisms of lithium ion batteries; it reduces the cell’s lifetime due to a loss of cyclable lithium and it reduces the safety due to the presence of metallic lithium in the cell. The deposition of metallic lithium occurs on the negative graphite electrode during charging preferentially at low temperatures, high currents, and high states of charge (SOC).
A novel approach enables the indirect detection of lithium plating through the analysis of voltage plateaus at discharge immediately after the charging process. Using the method of differential voltage analysis (DVA) [1] the charge amount corresponding to lithium stripping – as the reverse reaction to plating – is calculated [2].
A characterization of the cell behavior and a deeper analysis of the processes that take place at and after plating are performed using electrochemical impedance spectroscopy (EIS). Subsequent impedance spectra measured over a time span of one hour after metallic lithium was deposited on the negative electrode reveal fundamental changes in the polarization of the cell (see Fig. 1). After plating, the graphite electrode exhibits an impedance behavior which is typical for lithium metal electrodes [3]. The mid-frequency impedance arc increases with time (see Fig. 2) and approximates the size of the arc of the spectrum without plated lithium. The low polarization in this frequency range immediately after plating indicates the presence of new electrode surface area [4]. Whereas the increase in polarization over time reveals the formation of additional surface films. The approximation of the impedance arc over time towards the size of the arc without deposited lithium points to the development of a stable surface film. Nonetheless, the absence of the low frequency tail indicates that the electrode still behaves like a lithium metal electrode without solid state diffusion of lithium ions in graphite.
Further model-free analysis of the measured impedance spectra using the distribution of relaxation times (DRT) enables the identification of the underlying processes due to the relevant time constants [5]. It is shown that Li-ion migration through surface films is the rate-determining process in case of a plated graphite electrode.
References
[1]I. Bloom, A. N. Jansen, D. P. Abraham, J. Knuth, S. A. Jones, V. S. Battaglia, und G. L. Henriksen, J Power Sources, 139 (1–2) 2005, 295–303.
[2] M. Petzl und M. A. Danzer, J Power Sources, DOI 10.1016/j.jpowsour.2013.12.060, 2013.
[3] D. Aurbach, J Power Sources, 89 (2), 2000, 206–218.
[4] N. Schweikert, A. Hofmann, M. Schulz, M. Scheuermann, S. T. Boles, T. Hanemann, H. Hahn, und S. Indris, J Power Sources, 228, 2013, 237–243.
[5] J. P. Schmidt, T. Chrobak, M. Ender, J. Illig, D. Klotz, und E. Ivers-Tiffée, J Power Sources, 196 (12), 2001, 5342–5348.
