Effects of Electrolyte Additives on Unwanted Lithium Plating in Lithium Ion Cells
Qianqian Liu,Remi Petibon,Jeff R Dahn,Chunyu Du
DOI: https://doi.org/10.1149/ma2017-02/4/322
2017-01-01
ECS Meeting Abstracts
Abstract:Many electrolyte additives have been shown to promote longer lifetime lithium ion cells by forming protective films on the electrodes [1]. However, the additives used usually lead to higher negative electrode resistance which can increase the likelihood of unwanted lithium plating on the graphite negative electrode especially when charging at high rates and/or low temperatures [2,3]. In this work, effects of different electrolyte additives on unwanted lithium plating were studied in Li[Ni1/3Co1/3Mn1/3O2(NMC111)/graphite pouch cells. The relation between the negative electrode impedance and the onset current for unwanted lithium plating was also studied. Experiments were performed using charge-discharge cycling with different charge rates between 2.8 V - 4.1 V at 20°C. Unwanted lithium plating was detected by the onset of rapid capacity loss during cycling and post-mortem analysis of the negative electrode. Figure 1a shows that rapid capacity loss above the blue dashed line is caused by unwanted lithium plating and cells with the addition of 2% VC, PES211 or 2% TAP demonstrate obvious lithium plating at lower C-rates. Figure 1b shows that the negative electrode impedance depends strongly on the choice of electrolyte additives and the temperature. The negative electrode impedance increases with the addition of 2% VC, PES211 or 2% TAP, suggesting a negative correlation between the negative electrode impedance and the onset current for unwanted lithium plating. Figure 1c shows the calculated current for unwanted lithium plating (Iu) by the expression: Iu = 0.080 V x S/Rnegative, where 0.080 V is the overpotential needed for lithium plating, S is the geometric electrode surface area in a full cell and Rnegative is the area specific negative electrode resistance obtained from negative/negative symmetric cells shown in Figure 1b. Figure 1c shows that the prediction of this simple rule-of-thumb relation agrees well with the trends observed in Figure 1a under conditions where Rnegative is the dominant factor for anode polarization. This simple rule of thumb can be used by those searching for electrolyte additives that simultaneously lead to enhanced life time and the ability to charge at high rate. References: 1. K. Xu, Chem. Rev., 114, 11503–11618 (2014). 2. J.-P. Jones, M. C. Smart, F. C. Krause, B. V. Ratnakumar, and E. J. Brandon, ECS Trans., 75, 1–11 (2017). 3. Q. Q. Liu, R. Petibon, C. Y. Du, and J. R. Dahn, J. Electrochem. Soc., 164, A1173-A1183 (2017). Figure 1a) The capacity loss measured from the C/20 cycles before and after high rate cycling (350 hours of high rate cycling at the C-rates indicated) at 20°C for cells with the baseline electrolyte (1M LiPF6 EC:EMC(3:7)), 2% vinylene carbonate (VC), a ternary additive blend of 2% propene sultone (PES), 1% ethylene sulfate (DTD) and 1% tris(-trimethyl-silyl)-phosphite (TTSPi) (called PES211) and 2% triallyl phosphate (TAP); b) the area-specific Nyquist plots of negative electrode symmetric coin cell impedance divided by two at 20°C and 10°C; c) the measured onset current for unwanted lithium plating for cells with different electrolytes during cycling at 20°C and 10°C as well as the calculated result plotted versus the negative electrode area specific resistance, Rnegative from symmetric cells. Figure 1