Silicon is a promising high capacity anode material for Li-ion batteries (LIBs)¹. Compared to graphite, silicon has ten times higher theoretical capacity, due to the ability to form a Li-rich alloy. The challenge with silicon is the large volume expansion during lithiation causing the solid-electrolyte interphase (SEI) to crack, exposing fresh electrode surface, which again leads to the formation of more SEI. This continuous formation of SEI during operation consumes lithium and results in very thick SEI. Therefore, there is a need for an electrolyte composition that creates a strong and/or flexible SEI that can withstand silicon’s volume expansion during operation.

One way to manipulate the reactions at the interface between electrode and electrolyte during operation is to change the lithium salt. Philippe et al.² evaluated the performance of the salt lithium bis(fluorosulfonyl)imide (LiFSI) compared to LiPF₆ for anodes of nano-sized silicon, and found that LiFSI resulted in better performance. The proposed reason for improved performance was that the fluorination of the silicon particles surface was avoided. Also, LiFSI has a lower susceptibility towards hydrolysis compared to LiPF₆. Hence, it does not form HF in the presence of trace amounts of water.

In this work, we present the performance of LiFSI as electrolyte salt for LIBs based on micron-sized silicon as anode material. More specifically, the silicon based anodes used in this work are made of 60 wt% Si (Silgrain®, e-Si 400, a commercially available battery grade silicon from Elkem), with an average particle size of 3 µm, 10 wt% graphite (Timcal, KS6L), 15 wt% carbon black (Timcal, C-Nergy C65, CB) and 15 wt% Na-CMC binder (Sigma Aldrich Mw ~90000). Slurries were cast onto dendritic copper foil. The electrodes were cycled in 2016 or 2032 coin cells. The reference electrolyte composition is 1M LiFSI in EC:PC:DMC (1:1:3) + 5 wt% FEC and 1 wt% VC. Furthermore, the effect of increasing concentrations of salt will be evaluated. The work also includes optimization of the amount of the additive FEC.

Half cell tests were conducted with circular Li foil as counter electrode, while full cell tests were conducted with layered oxides, such as NMC or NCA, as cathode. When running full cell experiments the failure mechanism and SEI at the silicon anode is different from half cell experiments³. This due to less cycleable lithium in the cell. The difference of the SEI for half cell and full cell configuration is also investigated in this work.

The results presented here include both electrochemical performance and post mortem characterization. The post mortem characterization include FTIR and cross sectional analysis of cycled silicon electrodes, in order to evaluate the SEI formed when LiFSI is the electrolyte salt. Cross sectional analysis is performed by focused ion beam in combination with scanning electron microscopy and energy-dispersive X-ray spectroscopy. The results are compared to the performance of the same electrodes with LiPF₆ as electrolyte salt.

1. Ling and J. R. Dahn (2007) Journal of the Electrochemical Society, 154 (3), A156-A161.
2. Philippe, R. Dedryvère, M. Gorgoi, H. Rensmo, D. Gonbeau and K. Edström (2013) Journal of the American Chemical Society, 135, 9829-9842.
3. Dupré, P. Moreau, E. De Vito, L. Quazuguel, M. Boniface, A. Bordes, C. Rudisch, P. Bayle-Guillemaud and D. Guyomard (2016) Chemistry of Materials, 28, 2557-2572.