Lithium Dissolution and SEI Formation on Lithium Metal Anodes: Electrolyte and Surface Effects

Abstract

The lithium metal battery (LMB) rises as one of the most promising energy storage alternatives primarily due to the lithium metal’s high theoretical specific capacity (3860 mAh/g vs. 372 mAh/g for graphite) and high achievable operation voltage [1, 2]. However, and regardless of the coupling cathode electrode, one of the main challenges before the practical application of LMBs comes down to engineering electrolyte formulations compatible with lithium metal to eliminate dendrite formation and the uncontrolled loss of electrolyte and active lithium by forming a robust, ionic-conductive, and electron insulating, solid electrolyte interphase (SEI).Any improvement of the electrolyte/lithium metal interface passes through a thorough understanding of the reaction mechanisms behind the electrolyte degradation and the subsequence formation of the SEI. Characterizing this interface via experimental techniques is a challenge because of its delicate nature and simultaneous presence of liquid and solid phases. Here, we successfully narrowed this gap via computational modeling by using a recently introduced hybrid ab initio and reactive molecular dynamics (HAIR) scheme that uses in sequence the ab-initio molecular dynamics (AIMD) and the reactive force field (ReaxFF) methods to extend the time window achievable with the AIMD method alone. We performed a series of HAIR calculations with multiple electrolytes formulations on a lithium metal slab (Li(100)) with varying electrolyte/anode (E/A) ratios to evaluate the impact of the lithium salt concentration, type of solvent, presence of a diluent, and lithium thickness on the rate of electrolyte depletion, lithium dissolution, and SEI morphology. It is found that increasing salt concentration in a LiFSI/DMC electrolyte from 1 to 5 and then to 10 M yields significant changes in SEI morphology. The 1 M electrolyte led to DMC degradation into -O-CH3 with no significant LiO formation. In contrast, the 5 and 10 M formulations evolved into an SEI structure dominated by continuous and stratified LiO and LiF phases.Changes in the SEI morphology based on the E/A ratio for the 10 M LiFSI/DMC electrolyte revealed the existence of an intermediate E/A ratio leading to a more compact and more inorganic-rich SEI; the formed LiO phase density was higher than that in other ensembles representative of dry and flooded electrolyte operation conditions. Our third set of calculations focused on the effect of solvent chemistry. The 5 M LiFSI/DME electrolyte led to the formation of a LiO phase surrounded by some residual DME solvent; in this wetted SEI, the residual solvent provides diffusion channels for Li+ ions. Finally, we tested the impact of adding 1,1,2,2-tetrafluoroethylene 2,2,3,3-tetrafluoropropyl ether (TTE) as a diluent into the electrolyte using a 1.3 M LIFSI:DMC:TTE (1:1.2:3 molar ratio) formulation. The TTE diluent showed decomposition and helped grow a LiF-dominated SEI film with a lower LiO/LiF phase ratio than the equivalent 5.0 M LiFSI/DMC formulation with no electrolyte. These results form the basis for a deeper understanding of the SEI formation mechanisms and provide a design guideline for SEI films based on the lithium salt concentration, relative electrolyte/electrode content, and presence of a diluent.