Even though lithium metal has long been regarded as a promising active material for future lithium metal batteries (LMB), the native passivation layers covering pristine lithium metal (pLi) and its possible influence on the evaluation of the solid electrolyte interphase (SEI) has largely been disregarded by the battery research community.
The systematic study presented here utilizes a preformed SEI (pSEI) approach to evaluate the impact of the native passivation layer on vital SEI characteristics impacting the LMB full cell performance and safety. Furthermore, the utilized approach enabled an unclouded comparison of two well-known functional additive vinylene carbonate (VC) and fluoroethylene carbonate (FEC). Three organic carbonate-based electrolytes were used for this study: A baseline electrolyte (BE, 1.2 M LiPF6 in EC/EMC), the VC additive-containing electrolyte (AE-VC = BE + 5wt.%VC) and FEC additive-containing electrolyte (AE-FEC = BE +6.09 wt.%). The wt.% of the additives were chosen to ensure molar ratio equality between the additive containing electrolytes.
Electrochemical analysis was first performed by conducting stripping/plating experiments in six different electrode×electrolyte Li||Li symmetric cell setups. The pLi containing cell setups constantly generate higher overvoltage values compared to the pSEI containing analogues. Comparing the two considered functional additives, using AE-VC resulted in higher and inhomogeneous overvoltage profiles, whereas especially the double FEC containing cell setup (AE-FEC@Li×AE-FEC) low overvoltage values and the longest stripping/plating cell lifetime the performed study. Higher overvoltage values during the stripping/plating experiment can be an indication for inhomogeneous stripping/plating behavior.
This hypothesis was validated by post mortem optical analysis (incl. SEM): pLi containing cell setups revealed large areas of untouched native passivation surface, in combination with large patches of dendritic high surface area lithium (HSAL) and significant sideway growth over the stainless steel spacer. The pretreated electrodes, on the other hand, showed a more homogeneous stripping/plating behavior as well as less dendritic/sideway lithium growth.
Finally, in a NMC811||Li full cell setup, the influence of the considered functional additives and the native passivation layer on a full cell LMB was evaluated: The determined state of health and Coulombic efficiency data revealed both, the necessity of an adequate functional additive for LMB full cell performance as well as a 25 % lifetime increase by removing the native passivation layer and implementing a pSEI. Furthermore, FEC was shown to be vastly superior over VC for the application in future LMB systems.