Low-Voltage Operation and Lithium Bis(fluorosulfonyl)imide Electrolyte Salt Enable Long Li-Ion Cell Lifetimes at 85 °C

Abstract

LiFePO4/graphite (LFP), Li[Ni0.5Mn0.3Co0.2]O2/graphite (NMC3.8V, balanced for 3.8 V cut-off), and Li[Ni0.83Mn0.06Co0.11]O2/graphite (Ni83, balanced for 4.06 V cut-off) cells were tested at 85°C. Three strategies were used to improve cell lifetime for all positive electrode materials at 85°C. First, low-voltage operation (< 4.0 V) was used to limit the parasitic reactions at the positive electrode. Second, LiFSI (lithium bis(trifluoromethanesulfonyl)imide) was used as the electrolyte salt for its superior thermal stability over LiPF6 (lithium hexafluorophosphate). The low-voltage operation avoids the aluminum corrosion seen at higher voltages with LiFSI. NMC3.8V cells were operated at 6C charge and 6C discharge without issue for 2500 cycles and then moved to room temperature where normal operation was obtained. Finally, dimethyl-2,5-dioxahexane carboxylate (DMOHC) was used as a sole electrolyte solvent or mixed with dimethyl carbonate. μ-XRF data showed no detectable levels of transition metal deposition on the negative electrode of Ni83 and LFP cells, and DMOHC cells showed less gassing after testing compared to EC-based electrolytes. We found incredible capacity retention and cycle life for Ni83 and NMC3.8V cells using DMOHC and LiFSI at 70°C and at 85°C in tests that ran for more than 6 and 3 months (and are still running), respectively.

Alkyl Dicarbonates, Common Electrolyte Degradation Products, Can Enable Long-Lived Li-Ion Cells at High Temperatures

“…The loadings reported here are higher than previously reported data at 85 °C for LFP cells using an EC-based electrolyte. 16 However, here the LFP cell cycled for ∼3500 h, compared to ∼500 h in the previous study. A couple of studies show that there is a correlation between iron loadings on the anode and cycling temperature.…”

“…However, compared to previously reported results at 85 °C for the same cell type, these loadings of 0.1-0.2 μg cm −2 are somewhat larger but so close to the limit of detection, the significance is limited. 16 The main difference in the previous report is that the cells that went for μ-XRF analysis used an electrolyte that contained no DMOHC or used a DMOHC blend, whereas here all the cells used the same 1 M LiFSI DMOHC electrolyte. These loadings are much lower compared to NMC532 cells that cycled at 70 °C to an upper voltage limit of 4.3 V. 15 In that case, no DMOHC was used, and transition metal loadings were found to be on the order of 1-2 μg cm −2 for one example where a cell used DMC in its electrolyte.…”

“…Data for this cell was previously reported when the cell had far fewer cycles and had operated for much less time. 16 These two cells show little to no gas generation. Figure 5d shows that the capacity retention of cells using LiPF 6 with or without additives is much inferior to that of NMC3.8 V cells using LiFSI.…”

“…First, cells with various electrolyte or electrode variations can be screened much faster than at ambient conditions. 15,16 second, it has been suggested that there is evidence of an Arrhenius relationship between cell lifetime and temperature, and thus cycling cells at higher temperatures may provide insight into cell lifetimes at ambient conditions assuming the failure mechanisms at high temperatures are the same at ambient conditions. 15,[17][18][19][20] Analysis of NMC811/graphite cells cycled (unpublished work by the authors) at 70 °C using LiPF 6 salt in their electrolytes showed up to 10% DMOHC produced in their electrolytes.…”

Beneficial Effects of Oxide-Based Additives on Li-and Mn-rich Cathode Active Materials

“…On the other hand, the sieved SG3 graphite has a higher first-cycle capacity of 402 ± 4 mAh g −1 and a lower coulombic efficiency of 85.8 ± 0.1%, which one would expect based on its ∼4-fold higher BET surface area. 62,63,65 Its reversible capacity of 345 ± 4 mAh g −1 , however, is similar to the unsieved SG3 graphite.…”