Formation and Cycling of Microscale Silicon Anodes at Elevated Temperatures with FEC-Based and EC-Based Electrolytes

Abstract

Silicon is a promising candidate to replace graphite (372 mAh g-1) as anode material in lithium-ion batteries (LIBs) due to its high theoretical gravimetric capacity of 3589 mAh g-1.1 Graphite is already partially substituted by silicon in commercial battery applications in order to increase the energy density of LIBs.2 However, a complete transition to silicon is still hindered by the extreme volume expansion of up to 300% upon complete electrochemical lithiation to Li15Si4.3 This causes particle cracking, electrical contact loss of active material, and the continuous (re-)formation of the solid electrolyte interphase (SEI) on the newly formed surface, leading to a continuous loss of cyclable lithium and a fast capacity fading of the cell.One approach to mitigate Si particle cracking is the partial lithiation of microscale silicon, limiting the capacity to ~1200 mAh g-1, i.e., one-third of the theoretical capacity, enabling a cycle-life of more than 200 cycles before reaching 80% state-of-health (SOH).4 While electronic disconnection of lithiated silicon contributes to the overall cyclable lithium losses, the continuous formation of the SEI remains one of the major challenges that limit the lifetime.5 Fluoroethylene carbonate (FEC) has been shown to be one of the most suitable additives for silicon-containing cells to ensure the formation of a suitable SEI and comparably stable cycling performance.6,7 However, especially at elevated temperatures, FEC decomposes to form vinylene carbonate and hydrogen fluoride, which is detrimental to the SEI.8 This study investigated temperature-dependent cycle-life stability of microscale-silicon dominant anodes with EC-based and FEC-based electrolytes at room and elevated temperature. The electrodes consist of 70% silicon (Wacker Chemie AG) with an LiPAA binder (8%) and conductive carbon (2%). Additionally, graphite (8% KS6L, Imerys) was added to enhance the electronic conductivity, whereby within the operating voltage range of the anode, only a minor capacity of the graphite is used. As electrolytes, either 1 M LiPF6 in EC:DEC 1:2 v:v (EC-based) or 1 M LiPF6 in FEC:DEC 2:8 v:v (FEC-based) were used. Full-cells were cycled with a nickel-rich transition metal oxide cathode active material (LiNi0.83Co0.12Mn0.05O2 (NCM831205), BASF SE) based cathode at 25 °C and 45 °C. As shown in Figure 1, cycling in coin cells revealed the expected decreased cycle-life of cells with the FEC-based electrolyte at 45 °C (light blue) compared to 25 °C (dark blue).8 However, cycling cells with the FEC-free, EC-based electrolyte (green) at elevated temperatures, showed the opposite trend. The cycle-life until 80% SOH increased by 110 cycles (+100%), surpassing that of the cells with the FEC-based electrolyte at 45 °C. The charge-averaged mean voltages for both charge and discharge were investigated to identify the main failure mechanisms. This allowed us to differentiate between the loss of cyclable lithium and internal cell resistance build-up.To better understand the contributions of anode and cathode to capacity fading, we assembled Swagelok® T-Cells with the same cell chemistry and cycled them analogously. While the resistance of the cathode remained nearly constant over the whole course of cycling, the resistance of the anode showed an increase with decreasing cell capacity. Furthermore, online electrochemical mass spectrometry (OEMS) was performed to determine the gases evolved during formation with the EC-based electrolyte at 25 °C and 45 °C. This allowed us to gain insights into the overall SEI formation mechanism and to differentiate the formation of suitable SEI components.