Calendar aging of lithium-ion batteries is a function of cell voltage and temperature. The most common aging mechanism is the formation of SEI by passivating lithium ions on the surface of the graphite anode leading to a lower amount of active lithium und with this to a lower usable capacity of the cell. Additionally, towards higher voltages, an effect called cathode lithiation is proposed in e.g. by Hartmann et al. [1]. For this aging effect, the conductive salt LiPF6 is decomposed at the cathode and the lithium from the decomposed salt is inserted into the cathode. The inserted lithium is not passivated and is actively useable. In this way, the cathode lithiation, on the one hand, destroys the conductive salt and, on the other hand, increases the usable capacity of the full cell at the same time. In summary, there is a negative effect on usable capacity by SEI formation and a positive effect on capacity by cathode lithiation.
Comparing the aging mechanisms to the capacity loss rate, the anode overhang effect must be subtracted or excluded. We excluded the anode overhang effect by ignoring the initial transient part and evaluating only the slope of the following almost linear part of the capacity loss trends I_(C_loss ). Accordingly, the capacity loss rate is the difference of the SEI loss current and the cathode lithiation current:
I_(C_loss )=I_SEI-I_CL
In order to measure I_SEI and I_CL, we used the float current analysis. Therefore, the current is measured precisely to keep the full cell voltage constant [2]. As for the capacity loss trends, we ignore the initial transient part stemming from anode overhang effect and evaluate only the following steady-state of the float current [3]. As these aging effects contribute to the float currents only if they cause significant voltage decay of the full cell in OCV mode, the ratio of the anode-to-full-cell voltage slope and the cathode-to-full-cell voltage slope must be considered, as aging occurs on one electrode while recharging affects both [3].
For the float current analysis, typically 8 cells with different float voltage are used. After the decay of transient-phase at 30°C, the cells are then characterized using temperature steps from 5 to 50°C in 5K steps waiting 1 day at each temperature to equalize entropy-induced currents.
In this presentation, we show the optimization to link capacity loss rate to float currents at 30°C for different cathode materials and graphite anodes. Furthermore, we introduce a model to expand the results measured at 30°C to the range of 5-50°C of the temperature step results of the float current analysis. Finally, we obtain 3D plots of the SEI losses, losses due to cathode lithiation and capacity loss rates for all temperatures although the capacity losses were only measured at 30°C. The results are shown in the attached figure.
The proposed float current analysis is therefore a promising method to enhance the characterization of calendar aging.