Cyclic carbonate for highly stable cycling of high voltage lithium metal batteries

Abstract

The lithium metal battery (LMB) is one of the most promising next-generation battery systems due to its ultrahigh energy density. However, problematic dendrite formation and low Coulombic efficiency (CE) greatly limit its practical application. Carbonate electrolyte solvents are still indispensable for the operation of LMBs using a transition metal oxide cathode. We determined the impact of different cyclic carbonates, which actively participate in the formation of the solid-electrolyte interface (SEI), on the stable cycling of LMBs using a nickel-rich layered cathode LiNi_0.6Mn_0.2Co_0.2O₂ (NMC622). The substitution of fluorine atoms in the cyclic carbonate profoundly enhances the stability of the lithium metal anode while fluoroalkyl and alkoxy substituents are detrimental. Cyclic carbonate trans-difluoroethylene carbonate (DFEC) was identified as a novel SEI enabler on the lithium metal anode, facilitating the formation of a protective SEI with relatively high lithium fluoride content. A Li/NMC622 cell utilizing DFEC electrolyte solvent as SEI enabler displayed a capacity retention larger than 82% after 400 cycles and an average CE of 99.95%. In contrast, the cycling retention after 400 cycles for a Li/NMC622 cell using monofluoroethylene carbonate was only 31% with an average CE of 99.73%. Other fluoroalkyl or alkoxy cyclic carbonates do not provide improved stabilization of the lithium metal anode over ethylene carbonate. The fundamental studies in this work provide critical insight for the further development of advanced electrolytes in LMBs.

Introduction

Owing to their relatively high energy density, lithium-ion batteries (LIBs) have been extensively utilized in portable electronics. [1], [2], [3] However, the energy density of state-of-the-art LIBs is not sufficient to meet the application needs of electric vehicles. [4] The high-voltage lithium metal battery (LMB) is regarded as a highly promising energy storage system due to the ultrahigh theoretical specific capacity and extremely low electrochemical potential of the lithium metal anode. [5] Yet, the practical application of LMBs is still hindered by the extreme reactivity of lithium metal, [6] the low Coulombic efficiency (CE) induced by the instability of the lithium/electrolyte interface, [7] and the continuous growth of lithium dendrites, which can lead to severe safety issues. [8], [9] Recently, many researchers have been investigating methods for the stabilization of the lithium metal anode, including use of solid-state electrolytes, [10], [11], [12] robust protective layers on the lithium metal, [13], [14], [15], [16], [17] and various electrolyte salts, [18], [19] electrolyte additives, [20], [21], [22], [23] ionic liquids, [24], [25] electrolyte solvents, [26], [27], [28] concentrated electrolytes, [29], [30], [31], [32] and localized concentrated electrolytes. [33], [34] Among all these methods, re-designing the electrolyte has received the most attention due to its easy process and cost effectiveness, as well as the significantly higher conductivity of liquid electrolytes compared to their solid counterparts.

In 2015, Lu et al. investigated the failure mechanism of the lithium metal anode and conventional LIB electrolyte, i.e., lithium hexafluorophosphate (LiPF₆) dissolved in ethylene carbonate (EC) and linear carbonates such as dimethyl carbonate (DMC) and ethyl methyl carbonate (EMC). The results revealed the incompatibility of the LiPF₆-EC based electrolyte with lithium metal anode. [35] Moreover, EC was also found to be prone to oxidation at high voltage (4.4 V vs. Li/Li⁺), rendering it not suitable for nickel-rich layered cathodes. [36], [37], [38] While several researchers indicate that lithium bis(fluorosulfonyl)imide (LiFSI) is promising as the lithium salt for LMBs, [29], [30], [31], [32] there are concerns about the corrosion of the current collector by LiFSI at high voltage. [39], [40] Owing to its high dissociation constant in polar aprotic solvents, high conductivity, and ability to passivate the current collector, LiPF₆ is still a promising lithium salt for LMBs. Moreover, the complete absence of cyclic carbonates such as EC is not possible in LIBs due to their ability to enable the formation of a robust solid-electrolyte interface (SEI). While EC is not a qualified candidate for LMBs, Aurbach and his co-workers demonstrated that with the replacement of EC by monofluoroethylene carbonate (FEC), the stability of the lithium metal anode was greatly enhanced. [41], [42] Zhang et al. also introduced an FEC-based electrolyte that can enable highly stable cycling of LMBs using lithium iron phosphate as the cathode. [43] Xu and his coworkers demonstrated the advantage of using FEC-based carbonate electrolyte in high-voltage LMBs. [44] It is believed that the sacrificial reduction of FEC leads to the formation of a benign and robust SEI that contains lithium fluoride (LiF) on the lithium metal surface, mitigating the side reaction between electrolyte and the lithium metal. [27], [41], [42], [43], [44] Different researchers have also discovered independently the protection effect of LiF on lithium metal anode. [41], [45], [46] Several studies showed that other cyclic fluorinated carbonates such as trans-difluoroethylene carbonate (DFEC) and 4-(trifluoromethyl)-1,3-dioxolan-2-one (TFPC) displayed similar electrochemical properties to FEC, while some of the cyclic fluorinated carbonate-based electrolytes even outcompeted FEC in terms of electrochemical performance in LIBs. [47], [48], [49] Therefore, an in-depth understanding of the substituent effect in cyclic carbonates on the electrochemical performance of LMBs is critical to electrolyte optimization.

In this study, we designed and synthesized a variety of cyclic carbonates, including EC, FEC, DFEC, TFPC, 4-(2,2,3,3,4,4,5,5,5-nonafluoropentyl)-1,3-dioxolan-2-one (NFPEC), and 4-((2,2,3,3-tetrafluoropropoxy)methyl)-1,3-dioxolan-2-one (HFEEC) (Fig. 1). In cell tests, we then determined the electrochemical properties of electrolytes consisting of the above cyclic carbonates as SEI enablers, EMC as the thinning solvent, and LiPF₆ as the lithium salt. Although cyclic carbonates with fluoroalkyl or alkoxy substituents do not provide extra stability over EC, our results showed superior stability for the fluorine-substituted FEC- and DFEC-based electrolytes with the lithium metal anode. Moreover, the DFEC-based electrolyte clearly outperformed the FEC-based electrolyte in a Li/LiNi_0.6Mn_0.2Co_0.2O₂ (NMC622) cell, showing better high voltage compatibility. We found that the stabilization of the lithium metal anode originates from the formation of a compact SEI layer, which contains a relatively high level of lithium fluoride (LiF), through the reductive decomposition of FEC or DFEC. The fundamental findings in this work clearly illustrate the significance of the substituent effect in cyclic carbonates on the electrochemical performance of high-voltage LMBs, providing new insights on the development of advanced electrolytes. We report for the first time that the cyclic carbonate DFEC is a novel SEI enabler on the lithium metal anode, and that the robust SEI formed by the reductive decomposition of the DFEC dramatically stabilizes the cycling of lithium metal compared to conventional cyclic carbonate EC.