Lithium (Li)-metal batteries (LMBs) have been extensively investigated owing to the higher capacity of metallic Li anode compared to graphite employed in the current Li-ion batteries (LIBs). However, the surface stability of the Li electrode has not yet been fully addressed, attributed to the imperfect solid electrolyte interphase (SEI) layer and electrolyte solution. Many efforts have focused on seeking the pertinent electrolytes forming stable inorganic SEI layers on Li. For example, fluorinated ether solvents coordinate Li+ ion weakly, leading to Li+ - anion ion pairing and the formation of anion-derived SEI.[1] In addition, the utilization of diluent, which does not solvate Li+ but forms locally concentrated anions around Li+, further led to ion-pairing.[2] These approaches are promising to create (electro)chemically and mechanically robust SEI layers and have demonstrated improved cell cycling performance. Nonetheless, the weakly coordinating Li+ system has suffered from low ionic conductivity. In addition, the specific condition for the formation of ideal SEI composites, such as LiF, Li2S, Li3N, and Li2O, as the form of fully reduced anion (e.g., bis(fluorosulfonyl)imide (FSI-), has not been fully addressed.Here, we present the understanding of weakly solvating Li+ conditions correlated with the SEI composites and diffusivity of the electrolyte solution. We used 1 M lithium bis(fluorosulfonyl)imide (LiFSI) and fluorinated 1,4-dimethoxylbutane (FDMB) as a main weakly coordinating solvent and added diluent either tetrafluoroethyl-2,2,3,3-tetrafluoropropyl ether (TTE) or bis (2,2,2-trifluoroethyl) ether (BTFE). Comparing FDMB and FDMB and TTE (1/1 v/v) co-solution, 1 M LiFSI in FDMB and BTFE (1/1 v/v) electrolyte solution exhibited the best cycling performances in both Li|Li coin cells and Li|NCM811 pouch cells (>100 cycles). Notably, the BTFE-involving electrolyte formed the ideal LiF, Li2S, and Li3N composites in the SEI layers, whereas other electrolytes created partially decomposed FSI species and formed more organic-rich composites arising from the solvent decomposition. The desired SEI layer from the BTFE-containing electrolyte offered insignificant Li corrosion and the lowest presence of dead Li residue. In addition, the thinnest interphasial layer on NMC811 was observed, leading to no cathode crack after cycling.NMR analysis revealed that the addition of diluents, both BTFE and TTE, led to an increased amount of aggregated ion pairing, causing the formation of the anion-derived SEI layer. However, it may not explain the different degrees of anion reduction, thus forming different SEI composites. We address this disparity from ion diffusivity. BTFE has the lowest viscosity (0.7 cP at 25 oC) among FDMB (1.40 cP) and TTE (1.40 cP) and resulted in forming a small ionic cluster with a hydrodynamic radius of 0.49 pm, measured by NMR analysis. It contrasted with the large-size TTE, enhancing the solution viscosity despite no Li+ solvation and forming a large cluster with a hydrodynamic radius of 0.65 pm. Considering the Stoke-Einstein equation, the smaller hydrodynamic radius and low viscosity increase the Li+ and FSI- diffusion coefficients, supported by measured ionic conductivity and voltage hysteresis in Li|Li symmetric cells. In this presentation, I will discuss the details of the BTFE and TTE roles and the correlation between ion diffusivity and the SEI composite.
Read full abstract