Degradation Diagnosis of LIBs Using Electrochemical Impedance Spectroscopy after Correcting the SOC of a Cathode through dV/dQ Curve Analysis

Abstract

Introduction Electrochemical impedance spectroscopy (EIS) is an advantageous technique for analyzing the performance of lithium-ion batteries (LIBs) in an efficient and a non-destructive manner. Typically, when performing EIS for LIBs, battery conditions are adjusted by constant current/constant voltage charging to a fixed voltage or by discharging at a fixed capacity from the fully charged state. However, LIBs exhibit reduced capacity retention because of the side reactions, which consume electric charge, leading to a difference in the state of charge (SOC) of a cathode and anode. Therefore, as the difference in the SOC of a cathode and anode increases in a degraded battery (in addition to the degradation of each electrode), it becomes impossible to match the SOCs of each electrode in the degraded cell to those of the initial cell by only adjusting the state of the battery through a particular process. Hence, it is difficult to distinguish the degradation in the performance of the electrode from the difference in the SOCs of the initial and degraded electrodes.In this study, we developed a non-destructive diagnostic method for evaluating the cathode degradation degree using an AC impedance technique in a fixed SOC of a cathode using a quasi-three-electrode voltage curve obtained via dV/dQ curve analysis. The proposed method relies on the LIB degradation mechanism (primarily slippage of the reaction regions between the cathode and anode) [1]. Experimental Cycling tests were conducted on a commercially available lithium-ion cell (18650-type, 1.4 Ah). The cathode and anode consisted of LiNi0.5Co0.2Mn0.3O2 + LiMn2O4 (78:22 wt.%) and graphite as active materials, respectively. Cycling tests were conducted for 480 cycles at one charge/discharge rate (C/3), one SOC range (100–0%), and three temperatures (0, 25, and 45 °C). The battery performance was measured in low-current (C/20) charge/discharge tests at 25°C before and after each cycling test. The dV/dQ curves were calculated from the low-current discharge curves. EIS was performed at 25 °C with an alternating current excitation of ±5 mV over a frequency range of 100 mHz to 20 kHz. Results and discussion The capacity retentions of the cells, measured at C/20 at 25 °C, after 480 cycles at 0, 25, and 45 °C were 91, 89, and 83%, respectively. The capacity retentions of the cathodes after 480 cycles were estimated, using dV/dQ curve analysis, to be 94, 93, and 92% at 0, 25, and 45 °C, respectively—the variation in the cathodes was found to be lower than that in the cells. Figure 1 shows the AC impedance spectra of the cell before and after the cycling test. For a cell voltage adjusted to 3.755 V (Figure 1 (a)), there was a significant difference in the charge transfer resistance of the cathode (R cathode) in the tested cell at each temperature. The SOCs of the cathode before and after the cycling test at 0 , 25, and 45 °C for adjusted the cell voltage of 3.755 V were estimated, using the quasi-three-electrode voltage curve, to be 44, 34, 36, and 41%, respectively. In contrast, when the SOC of a cathode was adjusted to 44% using the quasi-three-electrode voltage curve, the difference in the R'cathode of the tested cell at each temperature decreased (Figure 1 (b)). These results suggest that the R cathode difference, observed when the cell voltages were adjusted to a constant value (3.755 V), is caused by the difference in the SOC of the cathode. Additionally, when the SOC of a cathode is fixed to a constant value (44%), the EIS spectra suggest that the cathode deteriorates during cycling because of the temperature-independent mechanical degradation caused by the expansion/contraction of the particles of the cathode. Therefore, an electrode must have the same SOC when comparing the degree of degradation of the electrodes using EIS, and the results provide considerable insights into the degradation evaluation of LIBs. Acknowledgement We thank M. Myojin, M. Nakajima, and H. Hata for their help in the life testing of LIBs at Japan Automobile Research Institute (JARI). This study was partially supported by the New Energy Promotion Council (NEPC). Reference [1] K. Ando, et al. J. Power Sources. 390 (2018) 278. Figure 1