Understanding and Suppressing Gas Evolution in Lithium Metal Batteries with Ether-Based Electrolytes

Understanding and Suppressing Gas Evolution in Lithium Metal Batteries with Ether-Based Electrolytes | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Understanding and Suppressing Gas Evolution in Lithium Metal Batteries with Ether-Based Electrolytes Hansen Wang, Yuchun Wang, Samantha Kung, Ziman Cai, Juanjuan Sun, and 9 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6034057/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Understanding and suppressing gas evolution in lithium secondary batteries are critical to optimizing battery performance and ensuring safe operation. 1 , 2 However, no systematic investigations of gas evolution in ether-based lithium metal batteries (LMBs) have been conducted despite the enticing prospects of LMBs for achieving ultrahigh energy density. 3 – 5 In this work, gas generation in ether electrolyte-based LMBs was quantified and the underlying redox mechanisms were elucidated. Through studying cathode and anode half-cells, it was determined that CO 2 and CO gas were generated at the cathode and CH 4 gas at the anode. Notably, CO 2 and CO were not observed in the full cell as they were consumed at the anode, reacting with lithium to produce solid Li compounds such as Li 2 CO 3 . CH 4 generated at the anode is the major contributor of gas generated in the full cell, though its evolution during cycling is not immediate and occurs after an onset point. The total gas volume generated increases dramatically with increasing temperature and decreasing electrolyte concentration. Based on these findings, electrolyte engineering and anode surface activation strategies were explored to control CH 4 production and hence overall gas evolution. In particular, the anode activation approach resulted in increased Li nucleation sites and improved Li deposition morphology, leading to significantly suppressed interfacial reactions, thus delaying the onset of gas evolution by 800% and increasing the cycling life by 400%. Achieving these improvements without altering the electrolyte formulation demonstrates the potential broad applicability of anode activation across various electrolyte chemistries. The performance enhancements beyond merely suppressing gas generation advances the prospects of safer and higher-performing LMBs. Physical sciences/Chemistry/Energy Physical sciences/Chemistry/Electrochemistry/Batteries Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Lithium (Li) metal anodes, due to their high specific capacity of 3860 mAh/g, 3–5 are ideal for the development of high energy battery systems that could enable electric aviation, such as electric vertical take-off and landing (eVTOL) aircraft, and long-range electric vehicles. 5 – 8 Therefore, enhancing the cycling performance of lithium metal batteries (LMBs) has been a subject of constant interest and research. 9 – 17 However, gas evolution during LMB operation remains poorly understood, especially for recently developed ether-based electrolytes, which are more compatible with Li metal than conventional electrolytes due to their intrinsic cathodic stability. 1 , 2 , 9 , 18 – 22 Uncontrolled gas evolution can rupture pouch cells and cause unwanted venting in prismatic and cylindrical cells, leading to serious safety risks such as electrolyte leakage and the release of toxic and flammable gases. Operation at elevated temperatures could further exacerbate gas evolution and the accompanying hazards. 2 , 23 In particular, during the high-rate operations of eVTOLs, cell temperatures above 50°C can develop. 24 , 25 Simulations of high energy lithium-ion CATL prismatic cells indicate that temperatures increase with the discharging rate, reaching 42°C at 4 C discharge and 62°C at 8 C (Supplementary Information, Fig. S1 ). This temperature increase is expected to be aggravated at the pack level in eVTOLs due to more limited heat dissipation and the minimal thermal management necessitated by weight constraints. 26 , 27 Therefore, it is critical to understand and suppress gas evolution in LMBs at elevated temperatures (e.g., 45–70°C) to ensure safe operation under practical working conditions. Here, we present the first study to quantify gas evolution under operando conditions and elucidate its mechanism in LMBs with high concentration ether-based electrolytes (HCE) and lithium nickel manganese cobalt oxide (NMC811) cathodes. After cycling at 45–70°C, the dominant gas component observed was methane (CH 4 ) gas generated at the anode. Moderate amounts of carbon monoxide (CO) and carbon dioxide (CO 2 ) gas were generated at the cathode and solidified through cross-talk reactions with the anode and thus not detected in the full cell. Minimal gas was detected during initial cycling of the full cell and gas evolution dominated by CH 4 generation only began after a variable number of cycles. The onset of gas evolution shifted forward in time both when increasing the cycling temperature and decreasing the electrolyte concentration. The underlying gas generation pathways at the electrode-electrolyte interfaces were analyzed using first-principles simulations. Experimental and computational results suggest that the onset of CH 4 evolution is linked to the accumulation of the methyllithium (LiCH 3 ) precursor, a decomposition intermediate of the 1,2-dimethoxyethane (DME) solvent. According to the experimental results, preventing CH 4 generation following DME decomposition is key to addressing gas evolution in DME-based LMBs. A localized high concentration electrolyte (LHCE) and a fluorinated-ether electrolyte (FEE) were tested as alternative ether electrolytes, as they are unlikely to support CH 4 formation. While these formulations successfully suppressed gas evolution, they also significantly reduced rate performance, which is detrimental for high-power applications in electric aviation. To overcome this, we propose a novel approach to activate the lithium anode by pre-stripping the surface to form sites conducive to more homogeneous Li deposition. This approach resulted in remarkable suppression of the CH 4 -forming interfacial chemical reactions with minimal impact on the rate performance. The onset of gas generation was delayed by almost 800% and the cycle life was increased by over 400%. Additionally, consumption of the Li salt and DME through interfacial reactions was reduced by 82.61% and 82.46%, respectively. By providing a comprehensive mechanistic understanding of and viable solutions for suppressing gas evolution in ether-based LMBs, this work paves the way for LMB commercialization, offering long cycle life, high energy density, and exceptional safety characteristics. Overview of Gas Evolution in LMB Gas evolution was investigated for the Li||NMC single layer stack (SLS) pouch cell shown schematically in Fig. 1 a. The electrolyte used was 4 M lithium bis(fluorosulfonyl)imide (LiFSI) in DME. Figure 1 b shows the voltage versus time plot of the Li||NMC full cell cycled at 60 ℃ between 2.8 and 4.3 V with 0.2 C charge and 0.5 C discharge rates. Cell volume changes (ΔVolume in Fig. 1 b) were tracked by an Archimedes in situ gas analyzer. 28 Following a small initial volume increase and a gradual volume decline (between ~ 500–3000 min), an abrupt and dramatic volume increase was observed after an onset point of 5 cycles (~ 3000 min). After 28 cycles, ~ 9 mL gas was produced, accompanied by significant swelling of the cell (Fig. S2). The volume increase occurred on top of minor oscillations that track the reversible cell volume changes during charging and discharging, matching the changes in the voltage. Systematic investigation of the full cell was supported by studies of the anode and cathode half-cells, which independently revealed the gas evolution characteristics at each electrode. In this study, anode and cathode half-cells are defined as Li||lithiated LFP (Li||LFP) and delithiated LFP||NMC cells (d-LFP||NMC) (Fig. 1 a, “half-cell test systems”), respectively. The lithium iron phosphate (LFP, LiFePO 4 ) counter electrodes (Fig. 1 a, “inert electrodes”) are designated as “inert” in the context of this study since they do not generate gas when cycled from an SOC of 0 to 100% at 60 ℃ for over 30 cycles (Fig. S3c). Shown in Fig. 1 c are the normalized volume fractions of gases generated in the full cell, anode half-cell, and cathode half-cell. The total volumes were obtained from ex situ measurements and the proportions of the constituent gas components were determined from gas chromatography (GC) results. The total volumes were normalized by the design capacity and cycling time (see Methods) to facilitate comparison. The results for the anode half-cell closely aligned with that of the full cell, with both producing a normalized gas volume of ~ 0.3 mL/Ahžh and consisting of methane (CH 4 ) as the dominant (~ 97%) component (Tables S2 and S5). Trace components include ethane (C 2 H 6 ), ethylene (C 2 H 4 ), acetylene (C 2 H 2 ), propylene (C 3 H 6 ), propane (C 3 H 8 ), hydrogen (H 2 ), and oxygen (O 2 ). In contrast, the normalized gas volume for the cathode half-cell was ~ 0.1 mL/Ahžh, dominated by carbon monoxide (CO, ~ 43%) and carbon dioxide (CO 2 , ~ 56%) (Table S3), both of which were absent in the full cell. A cross-talk phenomenon (Fig. 1 d) was proposed to account for the discrepancy in CO and CO 2 content. When CO and CO 2 produced at the NMC cathode were deposited as lithium compounds such as lithium carbonate (Li 2 CO 3 ) 29 – 31 at the Li anode in the Li||NMC full cell, the two gases could not be detected. The resulting Li 2 CO 3 cross-talk product in particular could be readily detected using titration-mass spectroscopy (T-MS), 32 , 33 with sulfuric acid (H 2 SO 4 ) as the titrant (Fig. 1 e). The strong CO 2 signal from the Cu electrode (blue curve in Fig. 1 e, equivalent to 0.035 mmol or 2.567 mg Li 2 CO 3 ) in a Cu||NMC cell (representing a Li||NMC full cell) indicates substantial cross-talk (top of Fig. 1 d). Cu was used in place of Li for safety reasons (see methods). In contrast, for d-LFP||NMC (i.e., the cathode half-cell), in which CO and CO 2 production was established (Fig. 1 c), no cross-talk conversion into Li 2 CO 3 occurred (bottom of Fig. 1 d), thus resulting in no signal (gray curve in Fig. 1 e). To establish a baseline, fresh Cu and delithiated LFP electrodes were also titrated, and no CO 2 signals were observed from these electrodes (Figs. S4c and S4d). The features in the in situ volume plot for the Li||NMC full cell (Fig. 1 b) illustrate the time-dependence of the aforementioned processes. First, in situ measurements for the cathode half-cell reveal that gas evolution at the cathode occurs during the first cycle and levels off within the next cycle (Fig. S5, green curve). Similarly in the full cell (Fig. 1 b, magenta curve), following a relatively small initial increase in volume—indicating CO and CO 2 gas production—during the first cycle, the volume levels off for several cycles. Notably, the onset of cathode gas evolution for DME on NMC811 occurs at approximately 4.18 V (vs. Li/Li + ), which is comparable to that for carbonate solvents on NMC811 34 . The ensuing gradual decrease in volume shown in Fig. 1 b may reflect the consumption of CO and CO 2 at the anode via cross-talk, which was not observed in the cathode half-cell (Fig. S5). Some of the cathode gases may also dissolve into the electrolyte at the elevated experimental temperatures, but the effects are expected to be minor since the volume remains relatively constant in the cathode half-cell (Fig. S5). At about 5 cycles in (~ 3000 min), gas production (primarily CH 4 ) is anode-driven and steeply rises. Gas Evolution at Cathode Figure 2 a shows the normalized volume fractions of gases produced at 60°C in d-LFP||NMC cathode half-cells when the LiFSI/DME electrolyte concentration was varied between 2–5 M. The volume of gas produced decreases with increasing LiFSI concentration. Holding the concentration constant at 4 M LiFSI/DME (Fig. 2 b), the gas volume increases with increasing temperature over a range of 45–70°C. For all conditions, CO and CO 2 comprise of ~ 98% of the gas produced (Tables S3 and S4). The reaction mechanisms for CO and CO 2 formation were investigated by first-principles simulation. The production of CO and CO 2 both result from DME oxidation at the NMC cathode, which follows two exergonic pathways (Fig. 2 c). In both pathways, DME first adsorbs onto the NMC surface via interactions between the DME oxygen (O) and the transition metal (TM) atoms. Following the dark green pathway, after two successive dehydrogenations of a methyl end group, a DME C–O bond is cleaved to give an O = CH· radical, which then further loses a hydrogen (H) to a surface O atom to yield CO. The production of CO 2 (light green pathway) also begins with the dehydrogenation of a methyl end group. The methyl C then becomes associated with a surface O, leading to the extraction of the O atom from the NMC lattice. This O is incorporated into the DME structure while another H is lost to the surface. A C–O bond then breaks to give an O = C–O· radical. A final dehydrogenation step yields CO 2 . The decrease in the normalized gas volume (mL/Ahžh) with increasing LiFSI concentration shown in Fig. 2 a can be attributed to several factors. First, decreasing the DME (reactant) concentration reduces the reaction rate, thus slowing gas production. The corresponding increase in the FSI – concentration promotes the formation of a passivating cathode electrolyte interphase (CEI) at the NMC surface, which hinders oxidation reactions. 35 , 36 The polycrystalline NMC particles in a 5 M sample also exhibited less microcracking than for a lower electrolyte concentration (compare Figs. S6b and S6c). Since microcracks can form when interfacial side reactions erode the crystal surface, 37 more severe microcracking at a lower concentration may be indicative of an increased rate of DME reaction with the cathode surface, which then leads to more rapid gas production. A corresponding increase in the cathode thickness (Fig. S7) was also observed, which may result from reaction-induced deformations such as microcracking. In contrast, the cathode thickness remains essentially unchanged from a fresh NMC electrode for the higher concentration (5 M) sample. The increase in normalized volume with temperature (Fig. 2 b) suggests the reaction is thermally activated, meaning the reaction rate increases with a rise in kinetic energy and the frequency of molecular collisions at higher temperatures. For a given concentration, a higher rate of DME decomposition results in a relatively larger volume of CO and CO 2 produced during the same experimental time window. Gas Evolution at Anode The dominant gas species in Li||LFP anode half-cells with varying electrolyte concentrations (2–5 M LiFSI/DME, Fig. 3 a) and temperatures (45–70°C, Fig. 3 b) is consistently CH 4 (~ 96–98%, Tables S5 and S6). The time dependence of the anode gas evolution is shown in Figs. 3 c and 3 d, with arrows indicating the onset times at which volume surges were observed. At the lowest 2 M concentration, the onset of gas evolution begins earliest (Fig. 3 c, after 2 cycles), consistent with the most rapid gas generation and hence largest normalized volume of gas produced (Fig. 3 a, ~ 0.6 mL/Ahžh). At the highest 5 M concentration, no onset point was observed (Fig. 3 c) and no gas was detected over the experimental time window (Fig. 3 a). The results in Fig. 3 d similarly complement Fig. 3 b, with earlier onset points at higher temperatures corresponding to larger normalized gas volumes produced. To gain insights into the cell conditions before and after the onset point, the electrolyte content was measured at various cycles for a 4 M sample at 60°C using our recently developed electrolyte consumption analysis method. 33 As shown in Fig. 3 e, the masses of both LiFSI (circles) and DME (squares) decrease with cycle number. The consumption of LiFSI is more rapid (steeper slope), leading to an overall decrease in the electrolyte concentration (triangles). A linear decrease in both LiFSI and DME over 15 cycles suggests that the rate of electrolyte consumption at the anode is unchanged before and after the onset point (at cycle 4). However, the results in Figs. 3 c and 3 d suggest that electrolyte decomposition does not produce gaseous species before the onset point but results in gaseous species afterwards. To understand this disparity surrounding the onset point, the DME decomposition pathways were calculated using first-principles simulation, as illustrated in Figs. 3 g-i. In Step 1 (Fig. 3 g), DME adsorbs onto the Li surface. In the less exergonic pathway (pink, Δ G = − 2.69 eV), one of the C–O bonds in the ethane backbone is cleaved, resulting in two surface-bound species, a methoxy (Li–OCH 3 ) and a methoxyethyl (Li–C 2 H 4 OCH 3 , highlighted in green) structure. In the second pathway (dark pink, Δ G = − 3.58 eV), the cleavage of the methoxy C–O bond generates a methyl group, which reacts with surface Li to form methyllithium (LiCH 3 , highlighted in light orange). LiCH 3 can theoretically react with another DME, as shown in Step 2A (Fig. 3 h), to produce CH 4 gas, the major gas species observed. However, the depletion of an additional DME molecule in this reaction is expected to alter the DME consumption rate at the onset point, contrary to the linear decrease over cycle number observed experimentally (Fig. 3 e). Alternatively, as shown in Step 2B (Fig. 3 i), LiCH 3 can react with the methoxyethyl (Li–C 2 H 4 OCH 3 ) also produced in Step 1 to form CH 4 gas. This reaction does not involve any additional DME. Furthermore, the calculated activation energy of this pathway is 20% lower (0.88 vs. 1.14 eV). Thus, both theoretical and empirical results suggest that Step 2A does not occur, and only Step 2B occurs, with the two highlighted products from Step 1 reacting to produce CH 4 . To determine the trigger for CH 4 gas production and thus the onset point, a Li||LFP cell with 4 M LiFSI/DME was cycled at 60°C with 3.6 mmol/L of externally added LiCH 3 (compared to LiCH 3 formed in situ during cycling). The onset of gas evolution shifted forward, from cycle 4 to cycle 3 (Fig. 3 f), suggesting that when LiCH 3 was abundant, the threshold LiCH 3 concentration was more rapidly reached and hence Step 2B proceeded earlier. In the presence of a large excess of LiCH 3 , gas production began immediately (Fig. S8a). The linear consumption of DME and the reaction rate dependence on LiCH 3 concentration provide evidence for the pathway involving only Steps 1 and 2B. In other words, nongaseous Li–C 2 H 4 OCH 3 and LiCH 3 are first produced (Step 1, Fig. 3 g). Then, once the LiCH 3 concentration accumulates to a threshold value, Step 2B (Fig. 3 i) occurs, triggering the production of CH 4 gas, which manifests as the onset point. It should be noted that while LiCH 3 produced at the anode can react with oxidation species that cross-talk from the cathode, 29 , 30 , 38 , 39 such as CO 2 , 40 these pathways have limited effects on the experimental results. Comparison of gas evolution in Li||NMC (containing oxidation species, Fig. S9b) and Li||LFP (no oxidation species, Fig. S9a) cells cycled at 60°C with 4 M LiFSI/DME revealed that the onset point of gas evolution is only delayed by about one cycle in Li||NMC. In the Li||LFP cell, all LiCH 3 produced in Step 1 contributes to reaching the threshold level for CH 4 production. In contrast, in the Li||NMC cell, some LiCH 3 reacts with the oxidation species, delaying the buildup of LiCH 3 to reach threshold levels and hence produce CH 4 gas. Nonetheless, these reactions only delay the onset by one cycle, with minimal impact on the timing of the onset point. Considering the anode half-cell only, the increasing delay in the onset point of CH 4 production with increasing LiFSI/DME concentration is the result of several factors, including the reduced concentration of DME and changes in the solvation structures. Simulation results (Fig. S10) for the Step 1 DME reduction at the anode show that the activation energies for conversion into LiCH 3 or LiOCH 3 intermediates (Fig. 3 g, dark pink and pink pathways, respectively) are both higher for higher concentrations. Specifically, the activation energy for LiCH 3 formation increases from 0.69 to 0.88 eV (Figs. S10a-b), and for LiOCH 3 formation, from 0.58 to 0.75 eV (Figs. S10c-d). Thus, the rate of gas production and the gas volume decreases with increasing concentration. Note that for both pathways, the FSI – anion is found in the first Li solvation shell at higher concentrations (Figs. S10b and S10d) and not in lower concentrations (Figs. S10a and S10c). The shifting of the onset points to earlier times with increasing temperature at a given electrolyte concentration is, like the case for the cathode, a result of increased energy and frequency of molecular collisions leading to increased reaction rate. In this case, the accelerated reaction is the decomposition of DME to produce CH 4 . Electrolyte Engineering to Suppress Gas Evolution Given that gas evolution stems primarily from CH 4 production at the anode, inhibiting the CH 4 reaction pathway should significantly reduce full cell gas generation. Furthermore, in practical applications, cathode gases can be efficiently removed after a formation cycle since they are only generated during the first cycle. In a cell treated in this manner, the effects of cross-talk can be avoided altogether and only anode gas production necessitates treatment. When using the standard 4 M LiFSI/DME electrolyte (Fig. 4 a), referred to as a high concentration electrolyte (HCE) due to its high concentration of the Li salt, CH 4 is derived from the terminal methyl group of the DME solvent (Figs. 3 g-i). After only 20 cycles at 60°C, significant swelling of the Li||NMC cell was observed (see side and top views in Fig. 4 a). As mentioned above, the formation of the Li–C 2 H 4 OCH 3 and LiCH 3 intermediates that react to produce CH 4 is less kinetically favorable (higher activation energy) at higher concentrations of LiFSI/DME. While it is not pragmatic to indefinitely increase the LiFSI/DME concentration due to the associated costs of LiFSI and impractically high viscosities, a localized high concentration electrolyte (LHCE, Fig. 4 b) can elevate the local LiFSI concentration, thus increasing the relative proportion of DME solvation structures that correspond to those with higher activation barriers, without actually adding significantly more LiFSI. The non-Li solvating diluent used for the LHCE shown in Fig. 4 b is 1,1,2,2-tetrafluoroethyl-2,2,3,3-tetrafluoropropyl ether (TTE). 35 For this LHCE system, no gas was generated for over 150 cycles at 60°C, i.e., the pouch cell remains flat (Fig. 4 b). Another option is to replace the terminal methyl CH 3 groups in DME with fluorinated groups, eliminating the CH 4 source altogether. For example, the fluorinated ether electrolyte (FEE), 2-[2-(2,2-difluoroethoxy)ethoxy]-1,1,1-trifluoroethane (F5DEE, Fig. 4 c), 41 terminates with difluoromethyl CF 2 and trifluoromethyl CF 3 groups. Since the critical intermediate LiCH 3 cannot be formed from F5DEE, no CH 4 should be generated. As observed in Fig. 4 c, no gas was generated for over 120 cycles at 60°C. Besides a reduction in gas generation, both LHCE and FEE exhibited improved capacity retention (Fig. 4 d), with cycle lives of about 300 cycles for LHCE (magenta) and 120 cycles for FEE (blue) at 60°C. However, despite these favorable results, the LHCE and FEE systems exhibit poor rate performances (Fig. 4 e) compared to the original HCE. This likely originates from their much lower ionic conductivities (Fig. 4 f). These limitations significantly hinder their practical application in high power applications such as eVTOL, where maintaining high rate performance is not only critical for fulfilling take-off and landing requirements, but also the capacity to respond to emergency situations and guarantee safety. 24 , 27 , 42 Thus, while the two electrolytes show promise, their poor rate performances render them inadequate for real-world applications. A more robust method was needed to simultaneously satisfy gas evolution and rate requirements. To achieve this, we turn our attention from the electrolyte to the other contributor in the anode-interfacial reactions, the anode itself. Anode Activation to Suppress Gas Evolution The thick native passivation film that forms on the surface of fresh, untreated Li metal, shown schematically as a brown film on the gray Li surface in Fig. 5 a, is lithiophobic. Thus, metallic Li (pale green, Fig. 5 b), deposits only from the more lithiophilic defects (mauve stripes in Figs. 5 a-b) in this passivation film. Due to the limited number of defect sites, metallic Li tends to grow continuously from these sites, forming dendrites. 43 , 44 Dendritic Li, with their large surface areas, are highly susceptible to interfacial chemical reactions, leading to substantial electrolyte consumption, severe gas generation, and poor cycling performance. However, if the native Li anode surface is activated via stripping (Fig. 5 c), it is possible to obtain an extended array of pits in which fresh, metallic Li is exposed. 43 , 45 A thin passivation layer (dark green, Fig. 5 c) forms in these newly created pits through reactions between the metallic Li and the electrolyte; these newly formed surfaces remain relatively lithiophilic and conducive to Li deposition. Given the abundance of pits, and thus Li deposition sites, Li deposits in a more two-dimensional manner across the surface, in particle or globule form (Fig. 5 d), instead of three-dimensionally away from the anode surface. 43 , 44 In other words, the activated anode is less susceptible to dendrite formation. The morphologies of untreated and activated anodes in Li||NMC cells with 4 M LiFSI/DME cycled at 60°C were investigated using SEM imaging. The untreated anode surface (Fig. 5 e) yielded roughly four nucleation sites (Fig. 5 f) within an approximately 0.2 mm 2 region. Shown in Fig. 5 f is Li deposition following an initial charge to 10% SOC. In comparison, for a similar sized region, the activated, or pre-stripped, anode contains about 80 pits (Fig. 5 g), 20 times the number of potential Li nucleation sites. These pits were confirmed to be Li nucleation sites as Li deposited within them upon initially charging the activated anode to 10% SOC (Fig. 5 h). When Li deposition was allowed to continue to 100% SOC, differences in the morphology of the deposited Li became apparent. For the untreated anode, Li metal grew continuously from the sparse defect sites, forming narrow and long dendrites (Fig. 5 i, 100% SOC). After 8 cycles, the cross-sectional image of the anode (Fig. 5 j, 0% SOC) revealed a surface covered with a porous solid electrolyte interphase (SEI) layer, also known as the residual SEI (rSEI), 8 , 46 with a thickness of close to 30 µm. On the activated anode, a comparable amount of Li deposition resulted in large, globular Li particles formed from the numerous nucleation sites (Fig. 5 k, 100% SOC), a distinctly different morphology compared to the elongated dendrites. The rSEI formed on the activated anode (Fig. 5 l), with a thickness of less than 1 µm, is 30 times thinner than that formed on the untreated anode. In addition to the differences in surface morphology and SEI thicknesses, there were also significant changes in the battery performance and gas evolution characteristics. The lower specific area of the deposited Li particles for the activated anode, compared to that of the Li dendrites on the untreated anode, resulted in fewer interfacial reactions, thus impacting electrolyte consumption, gas generation, and ultimately, cycling performance. For the same concentration of electrolyte (4 M LiFSI/DME) at the same temperature (60°C), the LiFSI and DME contents of the activated sample decreased much less rapidly with cycling compared to the untreated sample. This is shown in Fig. 5 m, in which the slopes of the linear fits to the masses for the activated sample (blue) are less steep relative to the untreated sample (magenta). The overall concentration (Fig. 5 n) of the electrolyte in the activated sample also only moderately decreased from 4 M to 3.9 M after 15 cycles, compared to a more drastic decrease to 2.9 M for the untreated sample. The consequences of fewer interfacial reactions, as evidenced by the reduced electrolyte consumption in the activated anode sample, include improved cycling performance and delayed gas generation. Figure 5 o shows that the untreated sample (magenta) had only undergone 30 cycles when the capacity plunged from an initial capacity of roughly 150 mAh to below 100 mAh. In contrast, the activated sample (blue) completed close to 160 cycles before the capacity finally decayed to 100 mAh, which is an extension of the cycle life by over a factor of five. Concurrently, gas evolution was delayed by at least a factor of nine. The onset point of the untreated sample (Fig. 5 p, magenta) occurs after 5 cycles, followed by abrupt and substantial gas generation. However, for the activated sample (blue), gas evolution begins gradually after 45 cycles. By solely controlling the morphology of Li deposition on the anode surface to reduce the surface area of deposited Li, interfacial reactions were successfully curbed, enabling improvements in various indicators of battery performance. Without changing the electrolyte system nor the cycling conditions, mitigated electrolyte consumption, better cycling performance, and delayed gas generation were all achieved. In other words, anode activation offers a universal means of enhancing the safety and cycling of lithium metal batteries. To fully explore the potential of the method, additional studies with other electrolytes and cycling conditions are warranted. Improving battery performance involves a diverse toolbox of strategies, and synergistic utilization of this form of anode activation with rationally designed materials and protocols, even other means of anode modification, stands to fully harness the potential of lithium metal batteries. Conclusion In this work, gas evolution in ether-based lithium metal batteries (LMBs) was quantified and its mechanistic underpinnings were elucidated through experimental and computational studies. Based on the finding that the primary source of gas generation is DME decomposition at the anode and with insights from the corresponding simulated mechanisms, two strategies targeting the solvent-anode interfacial reductions were implemented. Reduced gas evolution and longer cycle lives were obtained with both electrolyte engineering and anode activation strategies, but with a significant decline in rate performance for the alternative electrolytes tested. However, for the anode activation method, stellar rate performance was still possible. These results highlight the importance of developing a fundamental understanding of the problem (i.e., mechanism of gas evolution) and evaluating the pertinent engineering solutions against industrially relevant metrics using state-of-the-art battery prototypes. LMB research in recent years often focused on enhancing the cycle life 9–17 without holistic assessment of the product characteristics required for practical applications. Aviation batteries, such as those used in electric vertical take-off and landing (eVTOL) aircraft, must simultaneously exhibit excellent rate performance under room temperature and minimal gas generation at elevated temperatures to support the demands of take-off and landing operations with high reliability. 24,25,47 Thus, an improvement in cell safety (e.g., dendrite and gas generation suppression) without compromising on cell performance for the anode activation approach using 4 M LiFSI/DME electrolyte presents a significant step toward achieving such a goal. While the anode activation approach offers notable benefits, there remains the concern that it could potentially damage the cathode’s structural integrity, as Li anode stripping requires the cathode to accept the excess Li ions and overlithiate. Future research directions may include exploring synthesizing Li-deficient NMC cathodes or using sulfur cathodes, which can accommodate more Li ions without excessive strain. Furthermore, other strategies and electrolyte designs should also be evaluated against the standards laid out here. Only through such a product-oriented research methodology, can academic efforts serve to advance the viability of LMB commercialization, ultimately encouraging electrification efforts and reducing the societal carbon footprint. Methods Materials and Cells LiFSI (99.99%, Kaixin Co., Ltd), DME (99.95%, Guotai Co., Ltd), TTE (99.5%, Aladdin Co., Ltd), and F5DEE (99%, synthesized with the assistance of Sikang Chemical Co., Ltd.) were used for the electrolyte solutions. 2, 4 and 5 M LiFSI in DME (denoted as HCE), LiFSI-1.2DME-3TTE (by mole ratio, denoted as LHCE), and 1.2 M LiFSI in F5DEE (denoted as FEE) were prepared and injected (2.1 g/Ah for all samples) into corresponding dry cells in a dry room with a relative humidity of <1%. Cell assembly and teardowns were also performed in a dry room. The Li anodes consisted of 13 μm thick Cu foils with 50 μm of Li on either one side or both sides. The NMC used is the 811 type (LiNi 0.8 Mn 0.1 Co 0.1 O 2 , mono- and poly-crystalline mix, Rongbai Co., Ltd). Prior to cycling, samples were set to 0.4 MPa initial pressure using a custom-designed clamp and equilibrated at the desired temperature (45, 60, and 70 °C) for 2 h. Li||NMC and Li||LFP The Li||NMC full cells and the Li||LFP anode half-cells were fabricated in a CATL pilot line with double-sided NMC and LFP electrodes, with design capacities of 140 mAh (cathode areal capacity of 3.37 mAh/cm 2 ) and 150 mAh (3.61 mAh/cm 2 ), respectively. After electrolyte injection, they were cycled at 0.2 C (charge)-0.5 C (discharge) between 0-100% SOC (28–70 mA, 2.8–4.3 V for Li||NMC and 30–75 mA, 2.8–3.65 V for Li||LFP) without any formation cycles following a conventional constant current-constant voltage (CC-CV) protocol with a cutoff current of 0.1 C (14 mA for Li||NMC and 15 mA for Li||LFP). The battery cycler used was a Neware CT-4008Tn-5V6A-S1 (Neware Technology Ltd., Shenzhen, China). d-LFP||NMC Preparation of the d-LFP||NMC cathode half-cell involves multiple steps. First, a Li||LFP cell with 4 M LiFSI/DME electrolyte was fabricated using a Li anode and a single-sided LFP electrode with a design capacity of 75 mAh (3.61 mAh/cm 2 ). This cell was subjected to 1 cycle at 0.1 C-0.1 C and then charged at 0.1 C for 10 hours to produce a “Li vacancy” (specifically, 70 mAh) while keeping the cell voltage within a voltage plateau. The delithiated LFP electrode was then extracted, soaked for 10 min and rinsed in DME, and used as the “anode” for the d-LFP||NMC cathode half-cell. The NMC electrode was designed with a capacity of 40 mAh (1.92 mAh/cm 2 ) to enable cycling at the plateau voltage of LFP. The d-LFP||NMC cells were injected with the appropriate concentration of LiFSI/DME electrolyte solutions and cycled at 0.2 C-0.5 C (CC-CV with a cutoff current of 0.1 C) between 0-100% SOC (-0.63 to 0.87 V vs. E LFP plateau = 3.43 V) without any formation cycles. d-LFP||LFP Fabrication of the d-LFP||LFP cell (used for testing LFP inertness) also first involved construction and cycling of a Li||LFP cell as described above for the d-LFP||NMC cell. The delithiated LFP electrode, with a design capacity of 75 mAh (3.61 mAh/cm 2 ), was then used as the “anode” for the d-LFP||LFP cell. The other LFP electrode serving as the “cathode” is also single-sided, but designed with a capacity of 60 mAh. After injection of 4 M LiFSI/DME, the d-LFP||LFP cell was cycled at 0.2 C-0.5 C between 0-100% SOC (-0.63 to 0.22 V vs. E LFP plateau = 3.43 V) for 30 cycles at 60°C. Cu||NMC and Cu||Li The Cu||NMC and Cu||Li cells used for T-MS experiments were prepared with 13 µm thick Cu foils, double-sided NMC electrodes (design capacity of 140 mAh), and 13 µm thick Cu foils with 50 µm of Li on each side. Cu||NMC cells were cycled for 10 cycles at 60°C at 28 mA (charge)-100 mA (discharge) (CC-CV with a cutoff current of 14 mA) between 2.8–4.3 V. After that, a deep discharge was conducted to strip the remaining Li (7 mA to 2.8 V, 3.5 mA to 2.8 V, and 1 mA to 2.8 V, with 5 min rest between each step). Cu||Li cells were cycled for 10 cycles at 60°C. During the first cycle, 200 mAh of Li was deposited on the Cu electrode at 28 mA (charge) and 125 mAh of Li was stripped at 100 mA. For the following cycles, 125 mAh of Li was deposited and stripped at 28 mA and 100 mA, respectively. Finally, a deep discharge was conducted to strip the remaining Li from the Cu electrode (7 mA to -1 V, 3.5 mA to -1 V, and 1 mA to -1 V, with 5 min rest between each step). Gas Volume Determination Changes in the cell volume were determined using the Archimedes principle. 28 A fine wire was used to suspend a sample cell secured with metal clamps in a container of fluid. As the cell volume changes, the volume of fluid the cell displaces changes, thus resulting in changes in the buoyant force acting on the cell and ultimately the overall weight of the fluid-filled container. For ex situ measurements (performed at 25°C), a beaker with 1000 ml of filtered water (18.2 MΩ∙cm at 25°C, Thermo Scientific Barnstead NANOpure Water Purification System) was first tared on an analytical balance (Shimadzu, AUW200D). After submerging a suspended cell, the beaker system was weighed again. The volume of the immersed cell was calculated as the additional mass measured on the balance divided by the density of water (i.e., the volume of displaced water). The gas volume generated during cycling was determined by conducting measurements before and after cycling. Note, the cell tabs were insulated with tape to prevent short-circuiting. Normalized gas volumes were calculated as: $$\:{\text{V}}_{\text{N}}\text{=}\frac{{\text{V}}_{\text{total}}}{{\text{C}}_{\text{d}}\text{×t}}$$ where V N is the normalized gas volume (mL/Ahžh), V total is the total gas volume (mL), C d is the design capacity (Ah), and t is the total cycling time (h). In situ measurements were performed using an in situ battery gassing volume analyzer (Initial Energy Science & Technology, GVM2200). In this setup, the cell was submerged in silicone oil and volume changes were recorded in real time as the cells were cycled using a Neware BTS-5V1mA testing system (Neware Technology Ltd., Shenzhen, China). Cells tested at 45°C, 60°C and 70 ℃ were equilibrated for 4 h prior to testing. Gas Component Determination After a cell was cycled, the generated gases were sampled by puncturing the aluminum-plastic pouch with a syringe. The collected gas was injected into an Agilent 8890 gas chromatograph (GC) system with an inlet temperature of 120°C and a 10:1 split ratio. The argon carrier gas flow rate was set to 8 mL/min. Two columns were used in series: a 5 Å molecular sieve column (CP-Molsieve 5 Å; 25 m, 0.53 mm, 50 µm) and a PLOT column (HP-PLOT Q PT; 30 m, 0.53 mm, 40 µm). Following the columns are a thermal conductivity detector (TCD, Agilent) and a flame ionization detector (FID, Agilent). This set of columns and detectors enables clean detection of permanent gases (H 2 , O 2 , N 2 , CO, CO 2 ) and light hydrocarbons (CH 4 , C 2 H 6 , C 2 H 4 ) in a single injection. The TCD sensitivity was optimal at an operation temperature of 220°C, a makeup flow rate of 5 mL/min, and a reference flow rate of 30 mL/min. A custom gas mixture of known composition was used to determine the retention times for the gaseous compounds of interest and to perform signal calibration. The gas mixture was obtained from Jining Xieli Special Gas Co., Ltd and contains, by volume percent, 8.99% carbon dioxide, 8.98% carbon monoxide, 9.02% ethane, 8.93% ethylene, 0.497% acetylene, 9.02% propane, 8.99% propylene, 9.24% hydrogen, 0.633% oxygen, 8.99% methane, and 26.71% nitrogen. Cross-Talk Validation: Indirect CO 2 Detection To address why CO 2 was detected in the cathode half-cell but not the full cell, Li 2 CO 3 , the expected cross-talk product of CO 2 and the Li surface, was probed using titration-mass spectroscopy (T-MS). The reaction of Li 2 CO 3 with the H 2 SO 4 titrant is as follows: Li 2 CO 3 + H 2 SO 4 → Li 2 SO 4 + H 2 O་CO 2 ↑ In other words, if CO 2 produced at the cathode transfers to the anode to produce solid Li 2 CO 3 (i.e., cross-talk), the resulting Li 2 CO 3 on the extracted anode would react with H 2 SO 4 to produce CO 2 gas, which can then be collected and measured by MS (Shanghai LingLu Instruments Co., Ltd., QAS 100). In cells where CO 2 is produced (as evidenced by GC) but no cross-talk occurs (e.g., d-LFP||NMC cell), no Li 2 CO 3 would form at the anode, and hence no CO 2 would be detected by MS. As a precaution, since Li reacts violently with H 2 SO 4 , the system with cross-talk was modeled with Cu||NMC (“full cell”), using Cu instead of Li. The system without cross-talk was modeled with d-LFP||NMC (cathode half-cell). Furthermore, deposited Li was stripped from the Cu and LFP electrodes prior to cell disassembly. The extracted electrode was titrated with 10 M H 2 SO 4 . A cycled Cu||Li cell, in which CO 2 was not produced during cycling, was tested as a control sample (Fig. S4b). Fresh Cu and d-LFP electrodes were also tested to check for background signals Figs. S4c and S4d). A calibration equation for Li 2 CO 3 was obtained through titration of a Li 2 CO 3 standard sample (99.5%, Macklin Co., Ltd). Known masses of Li 2 CO 3 were titrated and the resulting CO 2 signals were integrated, yielding a calibration curve (Fig. S4a). Thus, for a given value of an integrated CO 2 signal, the corresponding mass of Li 2 CO 3 could be obtained. Electrolyte Consumption Analysis Electrolyte consumption analysis was performed to quantify the loss of LiFSI and DME with cycling. LiFSI was quantified using ion chromatography (IC) and DME by gas chromatography (GC). Both LiFSI and DME were extracted from the cell using an extraction agent. Hence, the complete method was termed extraction-gas and ion chromatography (E-G&IC). 33 1. Preparation of Extraction Agent An internal standard, 1,2-diethoxyethane (DEE, 99.9%, Aladdin Co., Ltd), was added to correct for matrix effects and sample loss during preparation for GC analysis. The internal standard was added to the sample along with diethylene glycol dimethyl ether (diglyme, 99.99%, Aladdin Co., Ltd), in the form of an extraction agent, which was prepared by dissolving 4 g of DEE in diglyme, and making up the solution to 200 mL in a volumetric flask. The extraction agent was sealed and stored for later use. 2. Electrolyte Extraction A small cut was made in the cycled cell inside an argon-filled glovebox (H 2 O < 0.1 ppm, O 2 < 0.1 ppm). After injecting 5 mL of the extraction agent, the cell was immediately re-sealed and gently shaken. The cell was allowed to rest under room temperature for 7 days for thorough extraction. The liquid content was then extracted from the cell and syringe filtered for further processing. 3. Ion Chromatography (IC) for Quantifying LiFSI Six standard LiFSI solutions with 0, 25, 50, 100, 150, and 200 mg LiFSI were prepared by dissolving the respective amounts in 1000 mL of deionized water. 5 mL aliquots were analyzed with IC (Dionex Aquion RFIC, Thermo Scientific) to obtain the LiFSI peak area for each corresponding concentration. Linear regression was performed between the peak areas and LiFSI concentrations, forming a calibration curve (Fig. S11a). Samples extracted from the cell were diluted with deionized water using a dilution factor of 200. A 5 mL aliquot of the diluted solution was measured with IC. The concentration of the diluted sample ( \(\:{c}_{\text{LiFSI}}\) ) was determined from the measured peak area ( \(\:{S}_{\text{LiFSI}}\) ): $$\:{S}_{\text{LiFSI}}={k}_{\text{LiFSI}}\times\:{c}_{\text{LiFSI}}$$ where \(\:{k}_{\text{LiFSI}}\) is the slope of the calibration curve. To obtain the absolute mass of LiFSI ( \(\:{m}_{\text{LiFSI}}\) ) remaining in the cycled cell, the concentration of the diluted sample was multiplied by the dilution factor (DF = 200) and the volume of the extracted solution: $$\:{m}_{\text{LiFSI}}={c}_{\text{LiFSI}}\times\:200\times\:5.245\:\text{mL}$$ where 5.245 mL comes from 5 mL for the extraction agent and 0.245 mL for 0.3 g of 4 M LiFSI/DME. 4. Gas Chromatography (GC) for Quantifying DME A standard solution was prepared by first adding 0.3 g of fresh electrolyte (4 M LiFSI/DME) to 5 mL of extraction agent. This solution was then diluted with diglyme with DF = 5. A 1.5 mL aliquot was measured with GC (Nexis GC-2030, Shimadzu). A response factor \(\:{f}_{\text{DME}}\) was calculated for DME using the known masses of DME and DEE ( \(\:{m}_{\text{DME}}\) , \(\:{m}_{\text{DEE}}\) ) in the standard solution and the peak areas extracted from the GC measurement ( \(\:{S}_{\text{DME}}\) , \(\:{S}_{\text{DEE}}\) ): $$\:{f}_{\text{DME}}=\frac{{m}_{\text{DME}}/{S}_{\text{DME}}}{{m}_{\text{DEE}}/{S}_{\text{DEE}}}$$ Samples extracted from the cell were diluted with diglyme with DF = 5. A 1.5 mL aliquot of the diluted solution was measured with GC. To obtain the absolute mass of DME ( \(\:{m}_{\text{DME, exp}}\) ) remaining in the cycled cell, the ratio of the obtained peak areas ( \(\:{S}_{\text{DME, exp}}\) , \(\:{S}_{\text{DEE, exp}}\) ) were multiplied by the mass of the DEE added ( \(\:{m}_{\text{DEE, exp}}\) ) in the form of the extraction agent, and the response factor ( \(\:{f}_{\text{DME}}\) ): $$\:{m}_{\text{DME, exp}}={m}_{\text{DEE, exp}}\times\:{f}_{\text{DME}}\times\:\frac{{S}_{\text{DME, exp}}}{{S}_{\text{DEE, exp}}}$$ First-Principles Simulation Density functional theory (DFT) calculations were performed with the Vienna ab initio simulation package (VASP). 48 , 49 The generalized gradient approximation and Perdew-Burke-Ernzerhof functional within the DFT framework were applied to determine the electron exchange-correlation energies. 50 Transition metals were treated using the DFT + U augmented approach with U values of 4, 4.4, and 5 eV for Mn, Co, and Ni, respectively. 51 Weak interactions were accounted for with the DFT + D3 method with dispersion correction. All calculations were spin-polarized with a plane wave cutoff energy of 520 eV. Surface calculations were performed with a 2 x 2 x 1 Monkhorst-Pack k-point scheme. A 15 Å vacuum layer was introduced to avoid interactions between periodic slabs. A five-layer slab of lithium (110) was used to investigate the DME reduction processes and a charged slab of NMC811 was used to investigate the DME oxidation processes. Geometric structural optimization was performed until the force on all atoms was less than 0.02 eV/Å, with an energy convergence criterion set to less than 10 − 5 eV/atom. Transition states were located using the climbing image nudged elastic band (CI-NEB) 52 and dimer methods 53 , with only one imaginary vibrational frequency being identified for each transition state along the reaction coordinate. The ground state molecular and ion geometries were optimized using the DFT method at the B3LYP/6-311G+(d, p) level. All DFT calculations were performed using the Gaussian 09 software package. The MD simulations were conducted using the GROMACS 2018 program. 54 The molecules and ions were described using the optimized potentials for liquid simulations all-atom (OPLS-AA) force field. 55 The partial charges on the atoms of the solvent were computed by fitting the molecular restrained electrostatic potential (RESP) at the atomic centers using the B3LYP/aug-cc-pVDZ basis set. 56 The simulation boxes were cubic with a side length of approximately 4 nm and contained different concentrations of LiFSI/DME, LiFSI-1.2DME-3TTE and 1.2 M LiFSI/F5DEE. During the simulations, the temperature was maintained at 333 K using a Nosé-Hoover thermostat with a relaxation time of 0.2 ps, and the pressure was controlled at 1 bar using a Parrinello-Rahman barostat with a relaxation time of 2.0 ps. From a total simulation duration of 40 ns, the last 20 ns were utilized for analysis. Anode Activation Li||NMC cells with 4 M LiFSI/DME were used for investigating the effects of anode activation. To activate the anode, the cell was discharged to 1.5 V at 25 ℃ using progressively smaller C-rates: 1 C (140 mA), 0.8 C (112 mA), 0.6 C (84 mA), 0.4 C (56 mA), 0.2 C (28 mA) and 0.1 C (14 mA), with a 5 min rest between each step. To determine the optimal cut-off voltage (i.e., 1.5 V), a series of potentiostatic discharge experiments was performed. The anode activated using 1.5 V was observed to simultaneously support good cycling performance and minimize gas evolution. Results for all voltages tested (1.3–2.1 V) are shown in Fig. S12, along with comments on why voltages both above and below 1.5 V were not suitable. Activated and non-activated samples were subjected to three test conditions at 60 ℃: charging to 10% SOC at 0.2 C (28 mA), charging to 100% SOC at 0.2 C (28 mA), and discharging to 0% SOC after 8 cycles (0.2 C-0.5 C, 28 mA-70 mA). SEM Characterization Scanning electron micrographs of the anode samples were acquired with an ultrahigh resolution cold-field emission scanning electron microscope (SEM, Hitachi SU8600). Prior to imaging, anodes extracted from cycled cells were washed and dried thrice in a dry room, with a 10 min DME soak each time. To prepare a cross-sectional sample, a long, thin strip of washed anode was quenched in liquid nitrogen for 60 min, then swiftly split on a stainless-steel blade to expose a clean cross-section on each of the split pieces. Extracted cathode samples were washed and dried in the same manner as the anode. A small piece of the sample was cut and transferred to an ion beam cross-section polisher (CP, Hitachi ArBlade 5000) under dry atmosphere and polished at room temperature under inert conditions to expose a clean cross-section for imaging. Ionic Conductivities Ionic conductivities were measured using a benchtop conductivity meter (LeiCi, DDSJ-318). Declarations Author Information Corresponding Authors Hansen Wang https://orcid.org/0000-0002-6738-1659 Xiaonan Luo https://orcid.org/0000-0002-0555-5481 Chuying Ouyang https://orcid.org/0000-0001-8891-1682 Authors Yuchun Wang https://orcid.org/0009-0000-6141-7318 Samantha T. Hung https://orcid.org/0000-0001-9448-0962 Acknowledgements X. L. acknowledges support from Fujian Provincial Natural Science Foundation of China under Grant No. 2024J08378. H. W. acknowledges support from the National Natural Science Foundation of China under Grant No. 22209085, and the support from Fujian Provincial Natural Science Foundation of China under Grant No. 2024J010048. C. O. acknowledges support from the National Natural Science Foundation of China under Grant No. 51962010. This work is also supported by funds from Contemporary Amperex Technology Co., Limited (CATL). We would like to extend our gratitude to Xiangen Cao and other team members at Sikang Chemical Co., Ltd. for their invaluable assistance in the synthesis of F5DEE. Author contributions Y.W., X.L., H.W., N.L. and C.O. conceived the idea and designed experiments. Y.C., S.T.H. and Z.C. performed the experiments and analyzed the data. J.S., L.B., E.W. and J.L. performed first-principles simulations. H.L. facilitated the quantitative measurements on gas generations. S.T.H., Y.W. co-wrote the manuscript with revisions from U.U., X.L., H.W., N.L. and C.O. Competing interests All authors are employed by Contemporary Amperex Technology Co., Limited (CATL). Data availability All data supporting the findings of this study are included in the Article and its Supplementary Information. References Aiken, C. P. et al. A Survey of In Situ Gas Evolution during High Voltage Formation in Li-Ion Pouch Cells. J. Electrochem. Soc. 162 , A760 (2015). Liu, P. et al. Revealing Lithium Battery Gas Generation for Safer Practical Applications. Adv. Funct. Mater. 32 , 2208586 (2022). Tarascon, J.-M. & Armand, M. Issues and challenges facing rechargeable lithium batteries. Nature 414 , 359–367 (2001). Lin, D., Liu, Y. & Cui, Y. Reviving the lithium metal anode for high-energy batteries. Nat. Nanotechnol. 12 , 194–206 (2017). Albertus, P., Babinec, S., Litzelman, S. & Newman, A. Status and challenges in enabling the lithium metal electrode for high-energy and low-cost rechargeable batteries. Nat. Energy 3 , 16–21 (2018). Chen, S., Dai, F. & Cai, M. Opportunities and Challenges of High-Energy Lithium Metal Batteries for Electric Vehicle Applications. ACS Energy Lett. 5 , 3140–3151 (2020). Yang, X.-G., Liu, T., Ge, S., Rountree, E. & Wang, C.-Y. Challenges and key requirements of batteries for electric vertical takeoff and landing aircraft. Joule 5 , 1644–1659 (2021). Wang, H. et al. Rate-Dependent Failure Mechanisms and Mitigating Strategies of Anode-Free Lithium Metal Batteries. ACS Appl. Mater. Interfaces 15 , 12967–12975 (2023). Qian, J. et al. High rate and stable cycling of lithium metal anode. Nat. Commun. 6 , 6362 (2015). Fan, X. et al. Highly fluorinated interphases enable high-voltage Li-metal batteries. Chem 4 , 174–185 (2018). Suo, L. et al. Fluorine-donating electrolytes enable highly reversible 5-V-class Li metal batteries. Proc. Natl. Acad. Sci. 115 , 1156–1161 (2018). Liu, J. et al. Pathways for practical high-energy long-cycling lithium metal batteries. Nat. Energy 4 , 180–186 (2019). Weber, R. et al. Long cycle life and dendrite-free lithium morphology in anode-free lithium pouch cells enabled by a dual-salt liquid electrolyte. Nat. Energy 4 , 683–689 (2019). Yu, Z. et al. Molecular design for electrolyte solvents enabling energy-dense and long-cycling lithium metal batteries. Nat. Energy 5 , 526–533 (2020). Chen, Y. et al. Steric Effect Tuned Ion Solvation Enabling Stable Cycling of High-Voltage Lithium Metal Battery. J. Am. Chem. Soc. 143 , 18703–18713 (2021). Wang, H. et al. Dual‐Solvent Li‐Ion Solvation Enables High‐Performance Li‐Metal Batteries. Adv. Mater. 33 , 2008619 (2021). Wang, H. et al. Liquid electrolyte: The nexus of practical lithium metal batteries. Joule 6 , 588–616 (2022). Li, Z. et al. Non-polar ether-based electrolyte solutions for stable high-voltage non-aqueous lithium metal batteries. Nat. Commun. 14 , 868 (2023). Park, M. S. et al. A highly reversible lithium metal anode. Sci. Rep. 4 , 3815 (2014). Jie, Y., Ren, X., Cao, R., Cai, W. & Jiao, S. Advanced Liquid Electrolytes for Rechargeable Li Metal Batteries. Adv. Funct. Mater. 30 , 1910777 (2020). Gan, L. et al. Comparative study of thermal stability of lithium metal anode in carbonate and ether based electrolytes. J. Power Sources 551 , 232182 (2022). Xiang, Y. et al. Gas induced formation of inactive Li in rechargeable lithium metal batteries. Nat. Commun. 14 , 177 (2023). Nie, Q. et al. Research Progress of Liquid Electrolytes for Lithium Metal Batteries at High Temperatures. Small 19 , 2302690 (2023). Clarke, M. A. & Alonso, J. J. Forecasting the Operational Lifetime of Battery-Powered Electric Aircraft. J. Aircr. 60 , 47–55 (2023). Bills, A. et al. A battery dataset for electric vertical takeoff and landing aircraft. Sci. Data 10 , 344 (2023). Zhao, C., Mazo, J. R. & Verstraete, D. Optimisation of a liquid cooling system for eVTOL aircraft: Impact of sizing mission and battery size. Appl. Therm. Eng. 246 , 122988 (2024). Wu, Z., Lian, W., Chen, B. & Zheng, C. Research on battery heat generation characteristics and thermal management system applied to a typical eVTOL. Appl. Therm. Eng. 257 , 124187 (2024). Aiken, C. P. et al. An apparatus for the study of in situ gas evolution in Li-ion pouch cells. J. Electrochem. Soc. 161 , A1548 (2014). Ellis, L. D. et al. Quantifying, Understanding and Evaluating the Effects of Gas Consumption in Lithium-Ion Cells. J. Electrochem. Soc. 164 , A3518 (2017). Bharti, A., Deb, D., Achutharao, G. & Bhattacharyya, A. J. CO 2 Crossover to the Li Anode and Its Implications on the Solid Electrolyte Interphase Composition in a Rechargeable Li–CO 2 Battery. J. Phys. Chem. C 128 , 11543–11551 (2024). Cotton, F. A., Wilkinson, G. & Gaus, P. L. Basic inorganic chemistry . (Wiley India, 2004). Zhang, H. et al. Titration Mass Spectroscopy (TMS): A Quantitative Analytical Technology for Rechargeable Batteries. Nano Lett. 22 , 9972–9981 (2022). Wang, H. et al. Quantifying Interfacial Reactions in Lithium Metal Batteries for a New Paradigm of Long-Cycling Electrolyte Designs. Preprint at https://doi.org/10.21203/rs.3.rs-3855314/v1 (2024). Jung, R., Metzger, M., Maglia, F., Stinner, C. & Gasteiger, H. A. Chemical versus Electrochemical Electrolyte Oxidation on NMC111, NMC622, NMC811, LNMO, and Conductive Carbon. J. Phys. Chem. Lett. 8 , 4820–4825 (2017). Ren, X. et al. Enabling High-Voltage Lithium-Metal Batteries under Practical Conditions. Joule 3 , 1662–1676 (2019). Ren, X. et al. High-Concentration Ether Electrolytes for Stable High-Voltage Lithium Metal Batteries. ACS Energy Lett. 4 , 896–902 (2019). Han, D., Weng, J., Zhang, X., Tong, Q. & Zhu, M. Review—Revealing the Intercrystalline Cracking Mechanism of NCM and Some Regulating Strategies. J. Electrochem. Soc. 169 , 040512 (2022). Metzger, M., Strehle, B., Solchenbach, S. & Gasteiger, H. A. Origin of H 2 Evolution in LIBs: H 2 O Reduction vs. Electrolyte Oxidation. J. Electrochem. Soc. 163 , A798–A809 (2016). Harris, O. C., Lee, S. E., Lees, C. & Tang, M. Review: mechanisms and consequences of chemical cross-talk in advanced Li-ion batteries. J. Phys. Energy 2 , 032002 (2020). Kaufmann, E., Sieber, S. & Schleyer, P. V. R. Carboxylation of lithium compounds: ab initio mechanisms. J. Am. Chem. Soc. 111 , 4005–4008 (1989). Yu, Z. et al. Rational solvent molecule tuning for high-performance lithium metal battery electrolytes. Nat. Energy 7 , 94–106 (2022). Courtin, C., Mahseredjian, A., Dewald, A. J., Drela, M. & Hansman, J. A Performance Comparison of eSTOL and eVTOL Aircraft. in AIAA AVIATION 2021 FORUM (American Institute of Aeronautics and Astronautics, 2021). doi:10.2514/6.2021-3220. Huang, Y., Pan, R., Rehnlund, D., Wang, Z. & Nyholm, L. First‐Cycle Oxidative Generation of Lithium Nucleation Sites Stabilizes Lithium‐Metal Electrodes. Adv. Energy Mater. 11 , 2003674 (2021). Frenck, L., Sethi, G. K., Maslyn, J. A. & Balsara, N. P. Factors That Control the Formation of Dendrites and Other Morphologies on Lithium Metal Anodes. Front. Energy Res. 7 , 115 (2019). Rehnlund, D., Ihrfors, C., Maibach, J. & Nyholm, L. Dendrite-free lithium electrode cycling via controlled nucleation in low LiPF6 concentration electrolytes. Mater. Today 21 , 1010–1018 (2018). Wang, H. et al. Efficient Lithium Metal Cycling over a Wide Range of Pressures from an Anion-Derived Solid-Electrolyte Interphase Framework. ACS Energy Lett. 6 , 816–825 (2021). He, J. et al. Key technologies and upgrade strategies for eVTOL aircraft energy storage systems. J. Energy Storage 103 , 114402 (2024). Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54 , 11169–11186 (1996). Kresse, G. & Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 6 , 15–50 (1996). Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 77 , 3865–3868 (1996). Grimme, S., Antony, J., Ehrlich, S. & Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 132 , 154104 (2010). Henkelman, G., Uberuaga, B. P. & Jónsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 113 , 9901–9904 (2000). Zhang, H., Qiu, L. & Hu, D. Finite‐Temperature Dimer Method for Finding Saddle Points on Free Energy Surfaces. J. Comput. Chem. 40 , 1701–1706 (2019). Abraham, M. J. et al. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 1–2 , 19–25 (2015). Kaminski, G. A., Friesner, R. A., Tirado-Rives, J. & Jorgensen, W. L. Evaluation and Reparametrization of the OPLS-AA Force Field for Proteins via Comparison with Accurate Quantum Chemical Calculations on Peptides. J. Phys. Chem. B 105 , 6474–6487 (2001). Sambasivarao, S. V. & Acevedo, O. Development of OPLS-AA Force Field Parameters for 68 Unique Ionic Liquids. J. Chem. Theory Comput. 5 , 1038–1050 (2009). Additional Declarations Yes there is potential Competing Interest. All authors are employed by Contemporary Amperex Technology Co., Limited (CATL). Supplementary Files Supplementaryinformation.docx Supporting Information Simulated cell temperatures for different discharge rates, images of a swollen cell due to gas generation, d-LFP||LFP cycling results, additional T-MS results, in situ measurements for a d-LFP||NMC cathode half-cell, cross-sectional SEM images and thickness measurements for NMC electrodes, in situ measurements for cells with additional LiCH 3 , in situ cell volume measurements for Li||LFP and Li||NMC cells, simulated activation energies and corresponding solvation structures at different electrolyte concentrations for the first step of DME reduction at the anode, electrolyte consumption analysis for Li||NMC and Li||LFP cells, anode activation using different voltages, and simulated radial distribution functions and cumulative radial distribution functions for different electrolyte systems. Tabulated results for T-MS, gas composition (GC), gas volumes, and reaction energy profile for CH 4 generation. Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6034057","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":420876196,"identity":"62800c1f-193f-4218-aea0-82b553f4b5a4","order_by":0,"name":"Hansen Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABA0lEQVRIiWNgGAWjYJACZgjF2Pjwg4ENmHGASC3Mh40lKtJAWhqI1cKWJsFz5jCYiVeLwfGzh18X1Nyx55fuMTaQbDtvt7b9MNCWGptonFrO5KVZzzj2LHHmnDOGDwrbbidvO5MI1HIsLbcBhxazAzlmxjxshxMMbuSAbLmdbHYAqIWx4TBuLeffALX8O2xvfyPHTIK37Vyy2fmHBLQADX/M23aYcYNEGsj7B+zMbhCwxf7GGzNm3r7DiTNuJIMCOTnB7AbQlgQ8fpHszzH+zPPtsD3/jERQVNrZm51Pf/jgQ40NTi1AwCaBzEsEq0zArRwEmD+guBS/4lEwCkbBKBiJAAAy5WlDtcl79gAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-6738-1659","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":true,"prefix":"","firstName":"Hansen","middleName":"","lastName":"Wang","suffix":""},{"id":420876197,"identity":"87ed3f92-9ddb-437f-b965-cd5a63b4470b","order_by":1,"name":"Yuchun Wang","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Yuchun","middleName":"","lastName":"Wang","suffix":""},{"id":420876198,"identity":"6a8defd9-b12e-4f07-a1c6-372a79b8e2c5","order_by":2,"name":"Samantha Kung","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Samantha","middleName":"","lastName":"Kung","suffix":""},{"id":420876199,"identity":"6556fc5c-85d5-4e90-badb-4061f2ccbeba","order_by":3,"name":"Ziman Cai","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Ziman","middleName":"","lastName":"Cai","suffix":""},{"id":420876200,"identity":"b4e33928-5513-485b-a07d-e74469592fde","order_by":4,"name":"Juanjuan Sun","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Juanjuan","middleName":"","lastName":"Sun","suffix":""},{"id":420876201,"identity":"497da07c-e378-4a32-9eea-23515da3f322","order_by":5,"name":"Haoran Li","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Haoran","middleName":"","lastName":"Li","suffix":""},{"id":420876202,"identity":"2ef0b0a4-56e1-42ba-9c56-9f92a6eaaf6d","order_by":6,"name":"Jinding Liang","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Jinding","middleName":"","lastName":"Liang","suffix":""},{"id":420876203,"identity":"7dbb7473-5c75-41cf-8c3e-1f06161b23c7","order_by":7,"name":"Lu Bai","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Lu","middleName":"","lastName":"Bai","suffix":""},{"id":420876204,"identity":"cc75c4fe-04cc-4b72-ae0f-58da6bd6e398","order_by":8,"name":"Erxiao Wu","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Erxiao","middleName":"","lastName":"Wu","suffix":""},{"id":420876205,"identity":"1c818cb9-6b96-493f-82db-f2d8e4c3e4ca","order_by":9,"name":"Ulderico Ulissi","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Ulderico","middleName":"","lastName":"Ulissi","suffix":""},{"id":420876206,"identity":"d022d75d-d160-4c33-9052-3ea9b67926d1","order_by":10,"name":"Xiaolin Yan","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Xiaolin","middleName":"","lastName":"Yan","suffix":""},{"id":420876207,"identity":"bcd684e1-2b0b-4512-b8fc-25916375fc8b","order_by":11,"name":"Xiaonan Luo","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Xiaonan","middleName":"","lastName":"Luo","suffix":""},{"id":420876208,"identity":"19ba35c2-5b4b-4997-a6ed-c18555d13a56","order_by":12,"name":"Na Liu","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Na","middleName":"","lastName":"Liu","suffix":""},{"id":420876209,"identity":"ef6276ab-12ff-4753-8c01-43a993beac1b","order_by":13,"name":"Chuying Ouyang","email":"","orcid":"","institution":"Contemporary Amperex Technology Co., Limited (CATL)","correspondingAuthor":false,"prefix":"","firstName":"Chuying","middleName":"","lastName":"Ouyang","suffix":""}],"badges":[],"createdAt":"2025-02-15 03:30:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6034057/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6034057/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":77311534,"identity":"3765610e-78cf-418d-a25f-c4696dbe8d19","added_by":"auto","created_at":"2025-02-27 10:01:16","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":10759654,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eInvestigating and quantifying gas evolution in LMB. a, \u003c/strong\u003eSchematic of Li||NMC full cell, LFP counter electrodes, Li||LFP anode half-cell, and d-LFP||NMC cathode half-cell systems used to study gas evolution. \u003cstrong\u003eb\u003c/strong\u003e, Voltage (black) and change in cell volume (magenta) over time for a Li||NMC full cell with 4 M LiFSI/DME electrolyte cycled at 60 °C. \u003cstrong\u003ec\u003c/strong\u003e, Normalized volumes of gas components generated in Li||NMC, Li||LFP, and d-LFP||NMC cells. \u003cstrong\u003ed\u003c/strong\u003e, Schematic showing the presence or absence of cross-talk phenomenon occurring in cells where CO and CO\u003csub\u003e2\u003c/sub\u003e are produced at the NMC surface. \u003cstrong\u003ee,\u003c/strong\u003e Schematic and CO\u003csub\u003e2\u003c/sub\u003e signal over time for the titration-mass spectroscopy experiment used to test for the presence of and to quantify the amount of one cross-talk product, Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e.\u003c/p\u003e","description":"","filename":"Picture1.png","url":"https://assets-eu.researchsquare.com/files/rs-6034057/v1/1f40bdad2abdb4950cbf8580.png"},{"id":77311537,"identity":"db31cc9a-4a19-4eac-96d8-a9e54331c554","added_by":"auto","created_at":"2025-02-27 10:01:16","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":3341572,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGas evolution in a d-LFP||NMC cathode half-cell. \u003c/strong\u003eNormalized gas volumes and gas component distributions \u003cstrong\u003ea, \u003c/strong\u003eat\u003cstrong\u003e \u003c/strong\u003edifferent concentrations (2, 4, 5 M) of LiFSI and\u003cstrong\u003e b,\u003c/strong\u003e at different temperatures (45, 60, 70 °C).\u003cstrong\u003e c, \u003c/strong\u003eSimulated pathways for the oxidation of DME into CO and CO\u003csub\u003e2\u003c/sub\u003e.\u003c/p\u003e","description":"","filename":"Picture2.png","url":"https://assets-eu.researchsquare.com/files/rs-6034057/v1/9d66812fc453d0d35e861f47.png"},{"id":77311535,"identity":"933cd18b-6a93-4b8e-bd39-c378c1acb97e","added_by":"auto","created_at":"2025-02-27 10:01:16","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":8814327,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGas evolution in a Li||LFP anode half-cell. (a, b)\u003c/strong\u003e Gas volumes and gas component distributions \u003cstrong\u003ea, \u003c/strong\u003eat\u003cstrong\u003e \u003c/strong\u003edifferent concentrations (2, 4, 5 M) of LiFSI and\u003cstrong\u003e b,\u003c/strong\u003e at different temperatures (45, 60, 70 °C).\u003cstrong\u003e (c, d) \u003c/strong\u003eCorresponding\u003cstrong\u003e \u003c/strong\u003echanges in cell volume \u003cstrong\u003ec, \u003c/strong\u003eat\u003cstrong\u003e \u003c/strong\u003edifferent concentrations (2, 4, 5 M) of LiFSI and\u003cstrong\u003e d,\u003c/strong\u003e at different temperatures (45, 60, 70 °C).\u003cstrong\u003e e,\u003c/strong\u003e Electrolyte consumption results showing the decrease in the masses of DME (black squares) and LiFSI (black circles), as well as the overall decrease in the LiFSI/DME concentration (magenta triangles) over number of cycles. \u003cstrong\u003ef, \u003c/strong\u003eChange in cell volume over time for a Li||LFP cell with 4 M LiFSI/DME electrolyte cycled at 60 °C with and without added LiCH\u003csub\u003e3\u003c/sub\u003e. \u003cstrong\u003e(g-i)\u003c/strong\u003e Simulated pathways for the formation of CH\u003csub\u003e4\u003c/sub\u003e at the anode. \u003cstrong\u003eg\u003c/strong\u003e, Step 1 involves C-O bond cleavages that result in two key intermediates (highlighted in green and light orange). There are two possible subsequent pathways: \u003cstrong\u003eh\u003c/strong\u003e, Step 2A, in which LiCH\u003csub\u003e3\u003c/sub\u003e reacts with another DME to produce CH\u003csub\u003e4\u003c/sub\u003e, and \u003cstrong\u003ei,\u003c/strong\u003e Step 2B, in which the two intermediates from Step 1 react to produce CH\u003csub\u003e4\u003c/sub\u003e.\u003c/p\u003e","description":"","filename":"Picture3.png","url":"https://assets-eu.researchsquare.com/files/rs-6034057/v1/85eb357d180c4bf3da7c8e8d.png"},{"id":77312880,"identity":"4e79e30e-f3e6-4bf1-9bf7-4a8276b0834c","added_by":"auto","created_at":"2025-02-27 10:09:17","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":11206639,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSchematic and properties of different electrolyte systems.\u003c/strong\u003e \u003cstrong\u003ea, \u003c/strong\u003eFor the HCE (4 M LiFSI/DME) system, there are, on average, 1.62 DME molecules coordinated to a Li\u003csup\u003e+\u003c/sup\u003e ion, and these DME molecules react to form CH\u003csub\u003e4\u003c/sub\u003e, resulting in significant gas evolution, as observed in a Li||NMC cell after 20 cycles at 60 °C. \u003cstrong\u003eb, \u003c/strong\u003eFor the LHCE (LiFSI-1.2DME-3TTE) system, there are, on average, 0.99 DME molecules coordinated to a Li\u003csup\u003e+\u003c/sup\u003e ion. The change in the solvation structure results in significantly reduced CH\u003csub\u003e4\u003c/sub\u003e production, and hence no gas evolution was observed after 150 cycles at 60 °C. \u003cstrong\u003ec,\u003c/strong\u003e For the FEE (1.2 M LiFSI/F5DEE) system, since the F5DEE molecules do not contain methyl groups, they do not react to form CH\u003csub\u003e4\u003c/sub\u003e; no gas evolution was observed after 120 cycles at 60 °C. \u003cstrong\u003ed, \u003c/strong\u003eCycling performance at 60 °C. \u003cstrong\u003ee,\u003c/strong\u003e Rate performance at 25 °C. \u003cstrong\u003ef,\u003c/strong\u003e Ionic conductivity values at 25 °C.\u003c/p\u003e","description":"","filename":"Picture4.png","url":"https://assets-eu.researchsquare.com/files/rs-6034057/v1/95c05dd4d752748ef8eb2989.png"},{"id":77311541,"identity":"687b4076-957e-4f29-99b5-5d7e526803bd","added_by":"auto","created_at":"2025-02-27 10:01:16","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":23857319,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLi deposition morphology and performance of untreated and activated Li anodes. (a-d)\u003c/strong\u003e Schematic showing Li deposition on\u003cstrong\u003e \u003c/strong\u003euntreated and activated Li anode surfaces. \u003cstrong\u003ea,\u003c/strong\u003e Untreated Li is covered by a native passivation film. \u003cstrong\u003eb,\u003c/strong\u003e Li dendrites grow from defect sites on the passivated surface. \u003cstrong\u003ec,\u003c/strong\u003e Activation of the Li anode by pre-stripping the surface to expose fresh Li, forming many pits. \u003cstrong\u003ed,\u003c/strong\u003e Li deposits within these pits in a more controlled manner. \u003cstrong\u003e(e-l)\u003c/strong\u003e SEM micrographs of untreated and activated Li anode surfaces at various SOC. \u003cstrong\u003ee, \u003c/strong\u003eUntreated Li anode. \u003cstrong\u003ef,\u003c/strong\u003e Untreated anode with Li deposited at defect sites after an initial charge to 10% SOC. \u003cstrong\u003eg,\u003c/strong\u003e Pre-stripped (activated) anode with pits. \u003cstrong\u003eh,\u003c/strong\u003e Activated anode with Li deposited in the pits after an initial charge to 10% SOC. \u003cstrong\u003ei,\u003c/strong\u003e Untreated anode with Li dendrites covering the surface after an initial charge to 100% SOC. \u003cstrong\u003ej,\u003c/strong\u003e Cross-section of the untreated anode at 0% SOC after 8 cycles. The top layer is the rSEI. \u003cstrong\u003ek,\u003c/strong\u003e Activated anode with Li particles covering the surface after an initial charge to 100% SOC. \u003cstrong\u003el,\u003c/strong\u003e Cross-section of the activated anode at 0% SOC after 8 cycles. The thin, top layer is the rSEI. \u003cstrong\u003em,\u003c/strong\u003e Consumption of LiFSI (circles) and DME (squares) and \u003cstrong\u003en, \u003c/strong\u003edecrease in LiFSI/DME concentration in Li||NMC cells at 60 °C with activated (blue) and untreated (magenta) Li anodes and corresponding \u003cstrong\u003eo,\u003c/strong\u003e cycling performance and \u003cstrong\u003ep,\u003c/strong\u003e changes in cell volumes.\u003c/p\u003e","description":"","filename":"Picture5.png","url":"https://assets-eu.researchsquare.com/files/rs-6034057/v1/0d8dd3b4bc4e782ba90bed6f.png"},{"id":77314781,"identity":"7a860d45-a65c-48c0-8631-4ccb1fee9b24","added_by":"auto","created_at":"2025-02-27 10:25:49","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":60048155,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6034057/v1/3ddabad3-f1b7-47df-a27e-02dab9ecbc01.pdf"},{"id":77311536,"identity":"048084f0-9138-4e0b-8d79-96f9e833d9c8","added_by":"auto","created_at":"2025-02-27 10:01:16","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":15827601,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupporting Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSimulated cell temperatures for different discharge rates, images of a swollen cell due to gas generation, d-LFP||LFP cycling results, additional T-MS results, \u003cem\u003ein situ \u003c/em\u003emeasurements for a d-LFP||NMC cathode half-cell, cross-sectional SEM images and thickness measurements for NMC electrodes, \u003cem\u003ein situ \u003c/em\u003emeasurements for cells with additional LiCH\u003csub\u003e3\u003c/sub\u003e,\u003cem\u003e in situ\u003c/em\u003e cell volume measurements for Li||LFP and Li||NMC cells, simulated activation energies and corresponding solvation structures at different electrolyte concentrations for the first step of DME reduction at the anode, electrolyte consumption analysis for Li||NMC and Li||LFP cells, anode activation using different voltages, and simulated radial distribution functions and cumulative radial distribution functions for different electrolyte systems. Tabulated results for T-MS, gas composition (GC), gas volumes, and reaction energy profile for CH\u003csub\u003e4\u003c/sub\u003e generation.\u003c/p\u003e","description":"","filename":"Supplementaryinformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-6034057/v1/0724c66d17af8b7d5a5a82ce.docx"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nAll authors are employed by Contemporary Amperex Technology Co., Limited (CATL).","formattedTitle":"Understanding and Suppressing Gas Evolution in Lithium Metal Batteries with Ether-Based Electrolytes","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLithium (Li) metal anodes, due to their high specific capacity of 3860 mAh/g,\u003csup\u003e3\u0026ndash;5\u003c/sup\u003e are ideal for the development of high energy battery systems that could enable electric aviation, such as electric vertical take-off and landing (eVTOL) aircraft, and long-range electric vehicles.\u003csup\u003e\u003cspan additionalcitationids=\"CR6 CR7\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e Therefore, enhancing the cycling performance of lithium metal batteries (LMBs) has been a subject of constant interest and research.\u003csup\u003e\u003cspan additionalcitationids=\"CR10 CR11 CR12 CR13 CR14 CR15 CR16\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e However, gas evolution during LMB operation remains poorly understood, especially for recently developed ether-based electrolytes, which are more compatible with Li metal than conventional electrolytes due to their intrinsic cathodic stability.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan additionalcitationids=\"CR19 CR20 CR21\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e Uncontrolled gas evolution can rupture pouch cells and cause unwanted venting in prismatic and cylindrical cells, leading to serious safety risks such as electrolyte leakage and the release of toxic and flammable gases. Operation at elevated temperatures could further exacerbate gas evolution and the accompanying hazards.\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e In particular, during the high-rate operations of eVTOLs, cell temperatures above 50\u0026deg;C can develop.\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e Simulations of high energy lithium-ion CATL prismatic cells indicate that temperatures increase with the discharging rate, reaching 42\u0026deg;C at 4 C discharge and 62\u0026deg;C at 8 C (Supplementary Information, Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). This temperature increase is expected to be aggravated at the pack level in eVTOLs due to more limited heat dissipation and the minimal thermal management necessitated by weight constraints.\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e Therefore, it is critical to understand and suppress gas evolution in LMBs at elevated temperatures (e.g., 45\u0026ndash;70\u0026deg;C) to ensure safe operation under practical working conditions.\u003c/p\u003e \u003cp\u003eHere, we present the first study to quantify gas evolution under operando conditions and elucidate its mechanism in LMBs with high concentration ether-based electrolytes (HCE) and lithium nickel manganese cobalt oxide (NMC811) cathodes. After cycling at 45\u0026ndash;70\u0026deg;C, the dominant gas component observed was methane (CH\u003csub\u003e4\u003c/sub\u003e) gas generated at the anode. Moderate amounts of carbon monoxide (CO) and carbon dioxide (CO\u003csub\u003e2\u003c/sub\u003e) gas were generated at the cathode and solidified through cross-talk reactions with the anode and thus not detected in the full cell.\u003c/p\u003e \u003cp\u003eMinimal gas was detected during initial cycling of the full cell and gas evolution dominated by CH\u003csub\u003e4\u003c/sub\u003e generation only began after a variable number of cycles. The onset of gas evolution shifted forward in time both when increasing the cycling temperature and decreasing the electrolyte concentration. The underlying gas generation pathways at the electrode-electrolyte interfaces were analyzed using first-principles simulations. Experimental and computational results suggest that the onset of CH\u003csub\u003e4\u003c/sub\u003e evolution is linked to the accumulation of the methyllithium (LiCH\u003csub\u003e3\u003c/sub\u003e) precursor, a decomposition intermediate of the 1,2-dimethoxyethane (DME) solvent.\u003c/p\u003e \u003cp\u003eAccording to the experimental results, preventing CH\u003csub\u003e4\u003c/sub\u003e generation following DME decomposition is key to addressing gas evolution in DME-based LMBs. A localized high concentration electrolyte (LHCE) and a fluorinated-ether electrolyte (FEE) were tested as alternative ether electrolytes, as they are unlikely to support CH\u003csub\u003e4\u003c/sub\u003e formation. While these formulations successfully suppressed gas evolution, they also significantly reduced rate performance, which is detrimental for high-power applications in electric aviation. To overcome this, we propose a novel approach to activate the lithium anode by pre-stripping the surface to form sites conducive to more homogeneous Li deposition. This approach resulted in remarkable suppression of the CH\u003csub\u003e4\u003c/sub\u003e-forming interfacial chemical reactions with minimal impact on the rate performance. The onset of gas generation was delayed by almost 800% and the cycle life was increased by over 400%. Additionally, consumption of the Li salt and DME through interfacial reactions was reduced by 82.61% and 82.46%, respectively.\u003c/p\u003e \u003cp\u003eBy providing a comprehensive mechanistic understanding of and viable solutions for suppressing gas evolution in ether-based LMBs, this work paves the way for LMB commercialization, offering long cycle life, high energy density, and exceptional safety characteristics.\u003c/p\u003e\n\u003ch3\u003eOverview of Gas Evolution in LMB\u003c/h3\u003e\n\u003cp\u003eGas evolution was investigated for the Li||NMC single layer stack (SLS) pouch cell shown schematically in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea. The electrolyte used was 4 M lithium bis(fluorosulfonyl)imide (LiFSI) in DME. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb shows the voltage versus time plot of the Li||NMC full cell cycled at 60 ℃ between 2.8 and 4.3 V with 0.2 C charge and 0.5 C discharge rates. Cell volume changes (ΔVolume in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb) were tracked by an Archimedes \u003cem\u003ein situ\u003c/em\u003e gas analyzer.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e Following a small initial volume increase and a gradual volume decline (between ~\u0026thinsp;500\u0026ndash;3000 min), an abrupt and dramatic volume increase was observed after an onset point of 5 cycles (~\u0026thinsp;3000 min). After 28 cycles, ~\u0026thinsp;9 mL gas was produced, accompanied by significant swelling of the cell (Fig. S2). The volume increase occurred on top of minor oscillations that track the reversible cell volume changes during charging and discharging, matching the changes in the voltage.\u003c/p\u003e \u003cp\u003eSystematic investigation of the full cell was supported by studies of the anode and cathode half-cells, which independently revealed the gas evolution characteristics at each electrode. In this study, anode and cathode half-cells are defined as Li||lithiated LFP (Li||LFP) and delithiated LFP||NMC cells (d-LFP||NMC) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, \u0026ldquo;half-cell test systems\u0026rdquo;), respectively. The lithium iron phosphate (LFP, LiFePO\u003csub\u003e4\u003c/sub\u003e) counter electrodes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, \u0026ldquo;inert electrodes\u0026rdquo;) are designated as \u0026ldquo;inert\u0026rdquo; in the context of this study since they do not generate gas when cycled from an SOC of 0 to 100% at 60 ℃ for over 30 cycles (Fig. S3c). Shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec are the normalized volume fractions of gases generated in the full cell, anode half-cell, and cathode half-cell. The total volumes were obtained from \u003cem\u003eex situ\u003c/em\u003e measurements and the proportions of the constituent gas components were determined from gas chromatography (GC) results. The total volumes were normalized by the design capacity and cycling time (see Methods) to facilitate comparison. The results for the anode half-cell closely aligned with that of the full cell, with both producing a normalized gas volume of ~\u0026thinsp;0.3 mL/Ah\u0026#158;h and consisting of methane (CH\u003csub\u003e4\u003c/sub\u003e) as the dominant (~\u0026thinsp;97%) component (Tables S2 and S5). Trace components include ethane (C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e6\u003c/sub\u003e), ethylene (C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e4\u003c/sub\u003e), acetylene (C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e2\u003c/sub\u003e), propylene (C\u003csub\u003e3\u003c/sub\u003eH\u003csub\u003e6\u003c/sub\u003e), propane (C\u003csub\u003e3\u003c/sub\u003eH\u003csub\u003e8\u003c/sub\u003e), hydrogen (H\u003csub\u003e2\u003c/sub\u003e), and oxygen (O\u003csub\u003e2\u003c/sub\u003e). In contrast, the normalized gas volume for the cathode half-cell was ~\u0026thinsp;0.1 mL/Ah\u0026#158;h, dominated by carbon monoxide (CO, ~\u0026thinsp;43%) and carbon dioxide (CO\u003csub\u003e2\u003c/sub\u003e, ~\u0026thinsp;56%) (Table S3), both of which were absent in the full cell.\u003c/p\u003e \u003cp\u003eA cross-talk phenomenon (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed) was proposed to account for the discrepancy in CO and CO\u003csub\u003e2\u003c/sub\u003e content. When CO and CO\u003csub\u003e2\u003c/sub\u003e produced at the NMC cathode were deposited as lithium compounds such as lithium carbonate (Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e)\u003csup\u003e\u003cspan additionalcitationids=\"CR30\" citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e at the Li anode in the Li||NMC full cell, the two gases could not be detected. The resulting Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e cross-talk product in particular could be readily detected using titration-mass spectroscopy (T-MS),\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e with sulfuric acid (H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e) as the titrant (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee). The strong CO\u003csub\u003e2\u003c/sub\u003e signal from the Cu electrode (blue curve in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee, equivalent to 0.035 mmol or 2.567 mg Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e) in a Cu||NMC cell (representing a Li||NMC full cell) indicates substantial cross-talk (top of Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). Cu was used in place of Li for safety reasons (see methods). In contrast, for d-LFP||NMC (i.e., the cathode half-cell), in which CO and CO\u003csub\u003e2\u003c/sub\u003e production was established (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec), no cross-talk conversion into Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e occurred (bottom of Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed), thus resulting in no signal (gray curve in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee). To establish a baseline, fresh Cu and delithiated LFP electrodes were also titrated, and no CO\u003csub\u003e2\u003c/sub\u003e signals were observed from these electrodes (Figs. S4c and S4d).\u003c/p\u003e \u003cp\u003eThe features in the \u003cem\u003ein situ\u003c/em\u003e volume plot for the Li||NMC full cell (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb) illustrate the time-dependence of the aforementioned processes. First, \u003cem\u003ein situ\u003c/em\u003e measurements for the cathode half-cell reveal that gas evolution at the cathode occurs during the first cycle and levels off within the next cycle (Fig. S5, green curve). Similarly in the full cell (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, magenta curve), following a relatively small initial increase in volume\u0026mdash;indicating CO and CO\u003csub\u003e2\u003c/sub\u003e gas production\u0026mdash;during the first cycle, the volume levels off for several cycles. Notably, the onset of cathode gas evolution for DME on NMC811 occurs at approximately 4.18 V (vs. Li/Li\u003csup\u003e+\u003c/sup\u003e), which is comparable to that for carbonate solvents on NMC811\u003csup\u003e34\u003c/sup\u003e. The ensuing gradual decrease in volume shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb may reflect the consumption of CO and CO\u003csub\u003e2\u003c/sub\u003e at the anode via cross-talk, which was not observed in the cathode half-cell (Fig. S5). Some of the cathode gases may also dissolve into the electrolyte at the elevated experimental temperatures, but the effects are expected to be minor since the volume remains relatively constant in the cathode half-cell (Fig. S5). At about 5 cycles in (~\u0026thinsp;3000 min), gas production (primarily CH\u003csub\u003e4\u003c/sub\u003e) is anode-driven and steeply rises.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eGas Evolution at Cathode\u003c/h2\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea shows the normalized volume fractions of gases produced at 60\u0026deg;C in d-LFP||NMC cathode half-cells when the LiFSI/DME electrolyte concentration was varied between 2\u0026ndash;5 M. The volume of gas produced decreases with increasing LiFSI concentration. Holding the concentration constant at 4 M LiFSI/DME (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb), the gas volume increases with increasing temperature over a range of 45\u0026ndash;70\u0026deg;C. For all conditions, CO and CO\u003csub\u003e2\u003c/sub\u003e comprise of ~\u0026thinsp;98% of the gas produced (Tables S3 and S4).\u003c/p\u003e \u003cp\u003eThe reaction mechanisms for CO and CO\u003csub\u003e2\u003c/sub\u003e formation were investigated by first-principles simulation. The production of CO and CO\u003csub\u003e2\u003c/sub\u003e both result from DME oxidation at the NMC cathode, which follows two exergonic pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). In both pathways, DME first adsorbs onto the NMC surface via interactions between the DME oxygen (O) and the transition metal (TM) atoms. Following the dark green pathway, after two successive dehydrogenations of a methyl end group, a DME C\u0026ndash;O bond is cleaved to give an O\u0026thinsp;=\u0026thinsp;CH\u0026middot; radical, which then further loses a hydrogen (H) to a surface O atom to yield CO. The production of CO\u003csub\u003e2\u003c/sub\u003e (light green pathway) also begins with the dehydrogenation of a methyl end group. The methyl C then becomes associated with a surface O, leading to the extraction of the O atom from the NMC lattice. This O is incorporated into the DME structure while another H is lost to the surface. A C\u0026ndash;O bond then breaks to give an O\u0026thinsp;=\u0026thinsp;C\u0026ndash;O\u0026middot; radical. A final dehydrogenation step yields CO\u003csub\u003e2\u003c/sub\u003e.\u003c/p\u003e \u003cp\u003eThe decrease in the normalized gas volume (mL/Ah\u0026#158;h) with increasing LiFSI concentration shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea can be attributed to several factors. First, decreasing the DME (reactant) concentration reduces the reaction rate, thus slowing gas production. The corresponding increase in the FSI\u003csup\u003e\u0026ndash;\u003c/sup\u003e concentration promotes the formation of a passivating cathode electrolyte interphase (CEI) at the NMC surface, which hinders oxidation reactions.\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e The polycrystalline NMC particles in a 5 M sample also exhibited less microcracking than for a lower electrolyte concentration (compare Figs. S6b and S6c). Since microcracks can form when interfacial side reactions erode the crystal surface,\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e more severe microcracking at a lower concentration may be indicative of an increased rate of DME reaction with the cathode surface, which then leads to more rapid gas production. A corresponding increase in the cathode thickness (Fig. S7) was also observed, which may result from reaction-induced deformations such as microcracking. In contrast, the cathode thickness remains essentially unchanged from a fresh NMC electrode for the higher concentration (5 M) sample.\u003c/p\u003e \u003cp\u003eThe increase in normalized volume with temperature (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb) suggests the reaction is thermally activated, meaning the reaction rate increases with a rise in kinetic energy and the frequency of molecular collisions at higher temperatures. For a given concentration, a higher rate of DME decomposition results in a relatively larger volume of CO and CO\u003csub\u003e2\u003c/sub\u003e produced during the same experimental time window.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eGas Evolution at Anode\u003c/h3\u003e\n\u003cp\u003eThe dominant gas species in Li||LFP anode half-cells with varying electrolyte concentrations (2\u0026ndash;5 M LiFSI/DME, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea) and temperatures (45\u0026ndash;70\u0026deg;C, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb) is consistently CH\u003csub\u003e4\u003c/sub\u003e (~\u0026thinsp;96\u0026ndash;98%, Tables S5 and S6). The time dependence of the anode gas evolution is shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed, with arrows indicating the onset times at which volume surges were observed. At the lowest 2 M concentration, the onset of gas evolution begins earliest (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec, after 2 cycles), consistent with the most rapid gas generation and hence largest normalized volume of gas produced (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, ~\u0026thinsp;0.6 mL/Ah\u0026#158;h). At the highest 5 M concentration, no onset point was observed (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec) and no gas was detected over the experimental time window (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). The results in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed similarly complement Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb, with earlier onset points at higher temperatures corresponding to larger normalized gas volumes produced.\u003c/p\u003e \u003cp\u003eTo gain insights into the cell conditions before and after the onset point, the electrolyte content was measured at various cycles for a 4 M sample at 60\u0026deg;C using our recently developed electrolyte consumption analysis method.\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee, the masses of both LiFSI (circles) and DME (squares) decrease with cycle number. The consumption of LiFSI is more rapid (steeper slope), leading to an overall decrease in the electrolyte concentration (triangles). A linear decrease in both LiFSI and DME over 15 cycles suggests that the rate of electrolyte consumption at the anode is unchanged before and after the onset point (at cycle 4). However, the results in Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed suggest that electrolyte decomposition does not produce gaseous species before the onset point but results in gaseous species afterwards. To understand this disparity surrounding the onset point, the DME decomposition pathways were calculated using first-principles simulation, as illustrated in Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg-i.\u003c/p\u003e \u003cp\u003eIn Step 1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg), DME adsorbs onto the Li surface. In the less exergonic pathway (pink, Δ\u003cem\u003eG\u003c/em\u003e = \u0026minus;\u0026thinsp;2.69 eV), one of the C\u0026ndash;O bonds in the ethane backbone is cleaved, resulting in two surface-bound species, a methoxy (Li\u0026ndash;OCH\u003csub\u003e3\u003c/sub\u003e) and a methoxyethyl (Li\u0026ndash;C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e4\u003c/sub\u003eOCH\u003csub\u003e3\u003c/sub\u003e, highlighted in green) structure. In the second pathway (dark pink, Δ\u003cem\u003eG\u003c/em\u003e = \u0026minus;\u0026thinsp;3.58 eV), the cleavage of the methoxy C\u0026ndash;O bond generates a methyl group, which reacts with surface Li to form methyllithium (LiCH\u003csub\u003e3\u003c/sub\u003e, highlighted in light orange). LiCH\u003csub\u003e3\u003c/sub\u003e can theoretically react with another DME, as shown in Step 2A (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eh), to produce CH\u003csub\u003e4\u003c/sub\u003e gas, the major gas species observed. However, the depletion of an additional DME molecule in this reaction is expected to alter the DME consumption rate at the onset point, contrary to the linear decrease over cycle number observed experimentally (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee). Alternatively, as shown in Step 2B (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ei), LiCH\u003csub\u003e3\u003c/sub\u003e can react with the methoxyethyl (Li\u0026ndash;C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e4\u003c/sub\u003eOCH\u003csub\u003e3\u003c/sub\u003e) also produced in Step 1 to form CH\u003csub\u003e4\u003c/sub\u003e gas. This reaction does not involve any additional DME. Furthermore, the calculated activation energy of this pathway is 20% lower (0.88 vs. 1.14 eV). Thus, both theoretical and empirical results suggest that Step 2A does not occur, and only Step 2B occurs, with the two highlighted products from Step 1 reacting to produce CH\u003csub\u003e4\u003c/sub\u003e.\u003c/p\u003e \u003cp\u003eTo determine the trigger for CH\u003csub\u003e4\u003c/sub\u003e gas production and thus the onset point, a Li||LFP cell with 4 M LiFSI/DME was cycled at 60\u0026deg;C with 3.6 mmol/L of externally added LiCH\u003csub\u003e3\u003c/sub\u003e (compared to LiCH\u003csub\u003e3\u003c/sub\u003e formed \u003cem\u003ein situ\u003c/em\u003e during cycling). The onset of gas evolution shifted forward, from cycle 4 to cycle 3 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef), suggesting that when LiCH\u003csub\u003e3\u003c/sub\u003e was abundant, the threshold LiCH\u003csub\u003e3\u003c/sub\u003e concentration was more rapidly reached and hence Step 2B proceeded earlier. In the presence of a large excess of LiCH\u003csub\u003e3\u003c/sub\u003e, gas production began immediately (Fig. S8a). The linear consumption of DME and the reaction rate dependence on LiCH\u003csub\u003e3\u003c/sub\u003e concentration provide evidence for the pathway involving only Steps 1 and 2B. In other words, nongaseous Li\u0026ndash;C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e4\u003c/sub\u003eOCH\u003csub\u003e3\u003c/sub\u003e and LiCH\u003csub\u003e3\u003c/sub\u003e are first produced (Step 1, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg). Then, once the LiCH\u003csub\u003e3\u003c/sub\u003e concentration accumulates to a threshold value, Step 2B (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ei) occurs, triggering the production of CH\u003csub\u003e4\u003c/sub\u003e gas, which manifests as the onset point.\u003c/p\u003e \u003cp\u003eIt should be noted that while LiCH\u003csub\u003e3\u003c/sub\u003e produced at the anode can react with oxidation species that cross-talk from the cathode,\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e such as CO\u003csub\u003e2\u003c/sub\u003e,\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e these pathways have limited effects on the experimental results. Comparison of gas evolution in Li||NMC (containing oxidation species, Fig. S9b) and Li||LFP (no oxidation species, Fig. S9a) cells cycled at 60\u0026deg;C with 4 M LiFSI/DME revealed that the onset point of gas evolution is only delayed by about one cycle in Li||NMC. In the Li||LFP cell, all LiCH\u003csub\u003e3\u003c/sub\u003e produced in Step 1 contributes to reaching the threshold level for CH\u003csub\u003e4\u003c/sub\u003e production. In contrast, in the Li||NMC cell, some LiCH\u003csub\u003e3\u003c/sub\u003e reacts with the oxidation species, delaying the buildup of LiCH\u003csub\u003e3\u003c/sub\u003e to reach threshold levels and hence produce CH\u003csub\u003e4\u003c/sub\u003e gas. Nonetheless, these reactions only delay the onset by one cycle, with minimal impact on the timing of the onset point.\u003c/p\u003e \u003cp\u003eConsidering the anode half-cell only, the increasing delay in the onset point of CH\u003csub\u003e4\u003c/sub\u003e production with increasing LiFSI/DME concentration is the result of several factors, including the reduced concentration of DME and changes in the solvation structures. Simulation results (Fig. S10) for the Step 1 DME reduction at the anode show that the activation energies for conversion into LiCH\u003csub\u003e3\u003c/sub\u003e or LiOCH\u003csub\u003e3\u003c/sub\u003e intermediates (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg, dark pink and pink pathways, respectively) are both higher for higher concentrations. Specifically, the activation energy for LiCH\u003csub\u003e3\u003c/sub\u003e formation increases from 0.69 to 0.88 eV (Figs. S10a-b), and for LiOCH\u003csub\u003e3\u003c/sub\u003e formation, from 0.58 to 0.75 eV (Figs. S10c-d). Thus, the rate of gas production and the gas volume decreases with increasing concentration. Note that for both pathways, the FSI\u003csup\u003e\u0026ndash;\u003c/sup\u003e anion is found in the first Li solvation shell at higher concentrations (Figs. S10b and S10d) and not in lower concentrations (Figs. S10a and S10c).\u003c/p\u003e \u003cp\u003eThe shifting of the onset points to earlier times with increasing temperature at a given electrolyte concentration is, like the case for the cathode, a result of increased energy and frequency of molecular collisions leading to increased reaction rate. In this case, the accelerated reaction is the decomposition of DME to produce CH\u003csub\u003e4\u003c/sub\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eElectrolyte Engineering to Suppress Gas Evolution\u003c/h3\u003e\n\u003cp\u003eGiven that gas evolution stems primarily from CH\u003csub\u003e4\u003c/sub\u003e production at the anode, inhibiting the CH\u003csub\u003e4\u003c/sub\u003e reaction pathway should significantly reduce full cell gas generation. Furthermore, in practical applications, cathode gases can be efficiently removed after a formation cycle since they are only generated during the first cycle. In a cell treated in this manner, the effects of cross-talk can be avoided altogether and only anode gas production necessitates treatment.\u003c/p\u003e \u003cp\u003eWhen using the standard 4 M LiFSI/DME electrolyte (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea), referred to as a high concentration electrolyte (HCE) due to its high concentration of the Li salt, CH\u003csub\u003e4\u003c/sub\u003e is derived from the terminal methyl group of the DME solvent (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg-i). After only 20 cycles at 60\u0026deg;C, significant swelling of the Li||NMC cell was observed (see side and top views in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea).\u003c/p\u003e \u003cp\u003eAs mentioned above, the formation of the Li\u0026ndash;C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e4\u003c/sub\u003eOCH\u003csub\u003e3\u003c/sub\u003e and LiCH\u003csub\u003e3\u003c/sub\u003e intermediates that react to produce CH\u003csub\u003e4\u003c/sub\u003e is less kinetically favorable (higher activation energy) at higher concentrations of LiFSI/DME. While it is not pragmatic to indefinitely increase the LiFSI/DME concentration due to the associated costs of LiFSI and impractically high viscosities, a localized high concentration electrolyte (LHCE, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb) can elevate the local LiFSI concentration, thus increasing the relative proportion of DME solvation structures that correspond to those with higher activation barriers, without actually adding significantly more LiFSI. The non-Li solvating diluent used for the LHCE shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb is 1,1,2,2-tetrafluoroethyl-2,2,3,3-tetrafluoropropyl ether (TTE).\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e For this LHCE system, no gas was generated for over 150 cycles at 60\u0026deg;C, i.e., the pouch cell remains flat (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb).\u003c/p\u003e \u003cp\u003eAnother option is to replace the terminal methyl CH\u003csub\u003e3\u003c/sub\u003e groups in DME with fluorinated groups, eliminating the CH\u003csub\u003e4\u003c/sub\u003e source altogether. For example, the fluorinated ether electrolyte (FEE), 2-[2-(2,2-difluoroethoxy)ethoxy]-1,1,1-trifluoroethane (F5DEE, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec),\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e terminates with difluoromethyl CF\u003csub\u003e2\u003c/sub\u003e and trifluoromethyl CF\u003csub\u003e3\u003c/sub\u003e groups. Since the critical intermediate LiCH\u003csub\u003e3\u003c/sub\u003e cannot be formed from F5DEE, no CH\u003csub\u003e4\u003c/sub\u003e should be generated. As observed in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec, no gas was generated for over 120 cycles at 60\u0026deg;C.\u003c/p\u003e \u003cp\u003eBesides a reduction in gas generation, both LHCE and FEE exhibited improved capacity retention (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed), with cycle lives of about 300 cycles for LHCE (magenta) and 120 cycles for FEE (blue) at 60\u0026deg;C. However, despite these favorable results, the LHCE and FEE systems exhibit poor rate performances (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee) compared to the original HCE. This likely originates from their much lower ionic conductivities (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef). These limitations significantly hinder their practical application in high power applications such as eVTOL, where maintaining high rate performance is not only critical for fulfilling take-off and landing requirements, but also the capacity to respond to emergency situations and guarantee safety.\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e Thus, while the two electrolytes show promise, their poor rate performances render them inadequate for real-world applications. A more robust method was needed to simultaneously satisfy gas evolution and rate requirements. To achieve this, we turn our attention from the electrolyte to the other contributor in the anode-interfacial reactions, the anode itself.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eAnode Activation to Suppress Gas Evolution\u003c/h3\u003e\n\u003cp\u003eThe thick native passivation film that forms on the surface of fresh, untreated Li metal, shown schematically as a brown film on the gray Li surface in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, is lithiophobic. Thus, metallic Li (pale green, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb), deposits only from the more lithiophilic defects (mauve stripes in Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea-b) in this passivation film. Due to the limited number of defect sites, metallic Li tends to grow continuously from these sites, forming dendrites.\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e,\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e Dendritic Li, with their large surface areas, are highly susceptible to interfacial chemical reactions, leading to substantial electrolyte consumption, severe gas generation, and poor cycling performance. However, if the native Li anode surface is activated via stripping (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec), it is possible to obtain an extended array of pits in which fresh, metallic Li is exposed.\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e,\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e A thin passivation layer (dark green, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec) forms in these newly created pits through reactions between the metallic Li and the electrolyte; these newly formed surfaces remain relatively lithiophilic and conducive to Li deposition. Given the abundance of pits, and thus Li deposition sites, Li deposits in a more two-dimensional manner across the surface, in particle or globule form (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed), instead of three-dimensionally away from the anode surface.\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e,\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e In other words, the activated anode is less susceptible to dendrite formation.\u003c/p\u003e \u003cp\u003eThe morphologies of untreated and activated anodes in Li||NMC cells with 4 M LiFSI/DME cycled at 60\u0026deg;C were investigated using SEM imaging. The untreated anode surface (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee) yielded roughly four nucleation sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef) within an approximately 0.2 mm\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e region. Shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef is Li deposition following an initial charge to 10% SOC. In comparison, for a similar sized region, the activated, or pre-stripped, anode contains about 80 pits (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg), 20 times the number of potential Li nucleation sites. These pits were confirmed to be Li nucleation sites as Li deposited within them upon initially charging the activated anode to 10% SOC (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eh).\u003c/p\u003e \u003cp\u003eWhen Li deposition was allowed to continue to 100% SOC, differences in the morphology of the deposited Li became apparent. For the untreated anode, Li metal grew continuously from the sparse defect sites, forming narrow and long dendrites (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ei, 100% SOC). After 8 cycles, the cross-sectional image of the anode (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ej, 0% SOC) revealed a surface covered with a porous solid electrolyte interphase (SEI) layer, also known as the residual SEI (rSEI),\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e with a thickness of close to 30 \u0026micro;m. On the activated anode, a comparable amount of Li deposition resulted in large, globular Li particles formed from the numerous nucleation sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ek, 100% SOC), a distinctly different morphology compared to the elongated dendrites. The rSEI formed on the activated anode (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003el), with a thickness of less than 1 \u0026micro;m, is 30 times thinner than that formed on the untreated anode.\u003c/p\u003e \u003cp\u003eIn addition to the differences in surface morphology and SEI thicknesses, there were also significant changes in the battery performance and gas evolution characteristics. The lower specific area of the deposited Li particles for the activated anode, compared to that of the Li dendrites on the untreated anode, resulted in fewer interfacial reactions, thus impacting electrolyte consumption, gas generation, and ultimately, cycling performance.\u003c/p\u003e \u003cp\u003eFor the same concentration of electrolyte (4 M LiFSI/DME) at the same temperature (60\u0026deg;C), the LiFSI and DME contents of the activated sample decreased much less rapidly with cycling compared to the untreated sample. This is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003em, in which the slopes of the linear fits to the masses for the activated sample (blue) are less steep relative to the untreated sample (magenta). The overall concentration (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003en) of the electrolyte in the activated sample also only moderately decreased from 4 M to 3.9 M after 15 cycles, compared to a more drastic decrease to 2.9 M for the untreated sample.\u003c/p\u003e \u003cp\u003eThe consequences of fewer interfacial reactions, as evidenced by the reduced electrolyte consumption in the activated anode sample, include improved cycling performance and delayed gas generation. Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eo shows that the untreated sample (magenta) had only undergone 30 cycles when the capacity plunged from an initial capacity of roughly 150 mAh to below 100 mAh. In contrast, the activated sample (blue) completed close to 160 cycles before the capacity finally decayed to 100 mAh, which is an extension of the cycle life by over a factor of five. Concurrently, gas evolution was delayed by at least a factor of nine. The onset point of the untreated sample (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ep, magenta) occurs after 5 cycles, followed by abrupt and substantial gas generation. However, for the activated sample (blue), gas evolution begins gradually after 45 cycles.\u003c/p\u003e \u003cp\u003eBy solely controlling the morphology of Li deposition on the anode surface to reduce the surface area of deposited Li, interfacial reactions were successfully curbed, enabling improvements in various indicators of battery performance. Without changing the electrolyte system nor the cycling conditions, mitigated electrolyte consumption, better cycling performance, and delayed gas generation were all achieved. In other words, anode activation offers a universal means of enhancing the safety and cycling of lithium metal batteries. To fully explore the potential of the method, additional studies with other electrolytes and cycling conditions are warranted. Improving battery performance involves a diverse toolbox of strategies, and synergistic utilization of this form of anode activation with rationally designed materials and protocols, even other means of anode modification, stands to fully harness the potential of lithium metal batteries.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn this work, gas evolution in ether-based lithium metal batteries (LMBs) was quantified and its mechanistic underpinnings were elucidated through experimental and computational studies. Based on the finding that the primary source of gas generation is DME decomposition at the anode and with insights from the corresponding simulated mechanisms, two strategies targeting the solvent-anode interfacial reductions were implemented. Reduced gas evolution and longer cycle lives were obtained with both electrolyte engineering and anode activation strategies, but with a significant decline in rate performance for the alternative electrolytes tested. However, for the anode activation method, stellar rate performance was still possible.\u003c/p\u003e\n\u003cp\u003eThese results highlight the importance of developing a fundamental understanding of the problem (i.e., mechanism of gas evolution) and evaluating the pertinent engineering solutions against industrially relevant metrics using state-of-the-art battery prototypes. LMB research in recent years often focused on enhancing the cycle life\u003csup\u003e9–17\u003c/sup\u003e without holistic assessment of the product characteristics required for practical applications. Aviation batteries, such as those used in electric vertical take-off and landing (eVTOL) aircraft, must simultaneously exhibit excellent rate performance under room temperature and minimal gas generation at elevated temperatures to support the demands of take-off and landing operations with high reliability.\u003csup\u003e24,25,47\u003c/sup\u003e Thus, an improvement in cell safety (e.g., dendrite and gas generation suppression) without compromising on cell performance for the anode activation approach using 4 M LiFSI/DME electrolyte presents a significant step toward achieving such a goal.\u003c/p\u003e\n\u003cp\u003eWhile the anode activation approach offers notable benefits, there remains the concern that it could potentially damage the cathode’s structural integrity, as Li anode stripping requires the cathode to accept the excess Li ions and overlithiate. Future research directions may include exploring synthesizing Li-deficient NMC cathodes or using sulfur cathodes, which can accommodate more Li ions without excessive strain. Furthermore, other strategies and electrolyte designs should also be evaluated against the standards laid out here. Only through such a product-oriented research methodology, can academic efforts serve to advance the viability of LMB commercialization, ultimately encouraging electrification efforts and reducing the societal carbon footprint.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eMaterials and Cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLiFSI (99.99%, Kaixin Co., Ltd), DME (99.95%, Guotai Co., Ltd), TTE (99.5%, Aladdin Co., Ltd), and F5DEE (99%, synthesized with the assistance of Sikang Chemical Co., Ltd.) were used for the electrolyte solutions. 2, 4 and 5 M LiFSI in DME (denoted as HCE), LiFSI-1.2DME-3TTE (by mole ratio, denoted as LHCE), and 1.2 M LiFSI in F5DEE (denoted as FEE) were prepared and injected (2.1 g/Ah for all samples) into corresponding dry cells in a dry room with a relative humidity of \u0026lt;1%. Cell assembly and teardowns were also performed in a dry room. The Li anodes consisted of 13 μm thick Cu foils with 50 μm of Li on either one side or both sides. The NMC used is the 811 type (LiNi\u003csub\u003e0.8\u003c/sub\u003eMn\u003csub\u003e0.1\u003c/sub\u003eCo\u003csub\u003e0.1\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e, mono- and poly-crystalline mix, Rongbai Co., Ltd). Prior to cycling, samples were set to 0.4 MPa initial pressure using a custom-designed clamp and equilibrated at the desired temperature (45, 60, and 70 °C) for 2 h.\u003c/p\u003e\n\u003ch3\u003eLi||NMC and Li||LFP\u003c/h3\u003e\n\u003cp\u003eThe Li||NMC full cells and the Li||LFP anode half-cells were fabricated in a CATL pilot line with double-sided NMC and LFP electrodes, with design capacities of 140 mAh (cathode areal capacity of 3.37 mAh/cm\u003csup\u003e2\u003c/sup\u003e) and 150 mAh (3.61 mAh/cm\u003csup\u003e2\u003c/sup\u003e), respectively. After electrolyte injection, they were cycled at 0.2 C (charge)-0.5 C (discharge) between 0-100% SOC (28\u0026ndash;70 mA, 2.8\u0026ndash;4.3 V for Li||NMC and 30\u0026ndash;75 mA, 2.8\u0026ndash;3.65 V for Li||LFP) without any formation cycles following a conventional constant current-constant voltage (CC-CV) protocol with a cutoff current of 0.1 C (14 mA for Li||NMC and 15 mA for Li||LFP). The battery cycler used was a Neware CT-4008Tn-5V6A-S1 (Neware Technology Ltd., Shenzhen, China).\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003ed-LFP||NMC\u003c/h2\u003e \u003cp\u003ePreparation of the d-LFP||NMC cathode half-cell involves multiple steps. First, a Li||LFP cell with 4 M LiFSI/DME electrolyte was fabricated using a Li anode and a single-sided LFP electrode with a design capacity of 75 mAh (3.61 mAh/cm\u003csup\u003e2\u003c/sup\u003e). This cell was subjected to 1 cycle at 0.1 C-0.1 C and then charged at 0.1 C for 10 hours to produce a \u0026ldquo;Li vacancy\u0026rdquo; (specifically, 70 mAh) while keeping the cell voltage within a voltage plateau. The delithiated LFP electrode was then extracted, soaked for 10 min and rinsed in DME, and used as the \u0026ldquo;anode\u0026rdquo; for the d-LFP||NMC cathode half-cell. The NMC electrode was designed with a capacity of 40 mAh (1.92 mAh/cm\u003csup\u003e2\u003c/sup\u003e) to enable cycling at the plateau voltage of LFP. The d-LFP||NMC cells were injected with the appropriate concentration of LiFSI/DME electrolyte solutions and cycled at 0.2 C-0.5 C (CC-CV with a cutoff current of 0.1 C) between 0-100% SOC (-0.63 to 0.87 V vs. E\u003csub\u003eLFP plateau\u003c/sub\u003e = 3.43 V) without any formation cycles.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003ed-LFP||LFP\u003c/h2\u003e \u003cp\u003eFabrication of the d-LFP||LFP cell (used for testing LFP inertness) also first involved construction and cycling of a Li||LFP cell as described above for the d-LFP||NMC cell. The delithiated LFP electrode, with a design capacity of 75 mAh (3.61 mAh/cm\u003csup\u003e2\u003c/sup\u003e), was then used as the \u0026ldquo;anode\u0026rdquo; for the d-LFP||LFP cell. The other LFP electrode serving as the \u0026ldquo;cathode\u0026rdquo; is also single-sided, but designed with a capacity of 60 mAh. After injection of 4 M LiFSI/DME, the d-LFP||LFP cell was cycled at 0.2 C-0.5 C between 0-100% SOC (-0.63 to 0.22 V vs. E\u003csub\u003eLFP plateau\u003c/sub\u003e = 3.43 V) for 30 cycles at 60\u0026deg;C.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eCu||NMC and Cu||Li\u003c/h2\u003e \u003cp\u003eThe Cu||NMC and Cu||Li cells used for T-MS experiments were prepared with 13 \u0026micro;m thick Cu foils, double-sided NMC electrodes (design capacity of 140 mAh), and 13 \u0026micro;m thick Cu foils with 50 \u0026micro;m of Li on each side. Cu||NMC cells were cycled for 10 cycles at 60\u0026deg;C at 28 mA (charge)-100 mA (discharge) (CC-CV with a cutoff current of 14 mA) between 2.8\u0026ndash;4.3 V. After that, a deep discharge was conducted to strip the remaining Li (7 mA to 2.8 V, 3.5 mA to 2.8 V, and 1 mA to 2.8 V, with 5 min rest between each step). Cu||Li cells were cycled for 10 cycles at 60\u0026deg;C. During the first cycle, 200 mAh of Li was deposited on the Cu electrode at 28 mA (charge) and 125 mAh of Li was stripped at 100 mA. For the following cycles, 125 mAh of Li was deposited and stripped at 28 mA and 100 mA, respectively. Finally, a deep discharge was conducted to strip the remaining Li from the Cu electrode (7 mA to -1 V, 3.5 mA to -1 V, and 1 mA to -1 V, with 5 min rest between each step).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eGas Volume Determination\u003c/h2\u003e \u003cp\u003eChanges in the cell volume were determined using the Archimedes principle.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e A fine wire was used to suspend a sample cell secured with metal clamps in a container of fluid. As the cell volume changes, the volume of fluid the cell displaces changes, thus resulting in changes in the buoyant force acting on the cell and ultimately the overall weight of the fluid-filled container.\u003c/p\u003e \u003cp\u003eFor \u003cem\u003eex situ\u003c/em\u003e measurements (performed at 25\u0026deg;C), a beaker with 1000 ml of filtered water (18.2 MΩ∙cm at 25\u0026deg;C, Thermo Scientific Barnstead NANOpure Water Purification System) was first tared on an analytical balance (Shimadzu, AUW200D). After submerging a suspended cell, the beaker system was weighed again. The volume of the immersed cell was calculated as the additional mass measured on the balance divided by the density of water (i.e., the volume of displaced water). The gas volume generated during cycling was determined by conducting measurements before and after cycling. Note, the cell tabs were insulated with tape to prevent short-circuiting. Normalized gas volumes were calculated as:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:{\\text{V}}_{\\text{N}}\\text{=}\\frac{{\\text{V}}_{\\text{total}}}{{\\text{C}}_{\\text{d}}\\text{\u0026times;t}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere V\u003csub\u003eN\u003c/sub\u003e is the normalized gas volume (mL/Ah\u0026#158;h), V\u003csub\u003etotal\u003c/sub\u003e is the total gas volume (mL), C\u003csub\u003ed\u003c/sub\u003e is the design capacity (Ah), and \u003cem\u003et\u003c/em\u003e is the total cycling time (h).\u003c/p\u003e \u003cp\u003e \u003cem\u003eIn situ\u003c/em\u003e measurements were performed using an \u003cem\u003ein situ\u003c/em\u003e battery gassing volume analyzer (Initial Energy Science \u0026amp; Technology, GVM2200). In this setup, the cell was submerged in silicone oil and volume changes were recorded in real time as the cells were cycled using a Neware BTS-5V1mA testing system (Neware Technology Ltd., Shenzhen, China). Cells tested at 45\u0026deg;C, 60\u0026deg;C and 70 ℃ were equilibrated for 4 h prior to testing.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eGas Component Determination\u003c/h2\u003e \u003cp\u003eAfter a cell was cycled, the generated gases were sampled by puncturing the aluminum-plastic pouch with a syringe. The collected gas was injected into an Agilent 8890 gas chromatograph (GC) system with an inlet temperature of 120\u0026deg;C and a 10:1 split ratio. The argon carrier gas flow rate was set to 8 mL/min. Two columns were used in series: a 5 \u0026Aring; molecular sieve column (CP-Molsieve 5 \u0026Aring;; 25 m, 0.53 mm, 50 \u0026micro;m) and a PLOT column (HP-PLOT Q PT; 30 m, 0.53 mm, 40 \u0026micro;m). Following the columns are a thermal conductivity detector (TCD, Agilent) and a flame ionization detector (FID, Agilent). This set of columns and detectors enables clean detection of permanent gases (H\u003csub\u003e2\u003c/sub\u003e, O\u003csub\u003e2\u003c/sub\u003e, N\u003csub\u003e2\u003c/sub\u003e, CO, CO\u003csub\u003e2\u003c/sub\u003e) and light hydrocarbons (CH\u003csub\u003e4\u003c/sub\u003e, C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e6\u003c/sub\u003e, C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e4\u003c/sub\u003e) in a single injection. The TCD sensitivity was optimal at an operation temperature of 220\u0026deg;C, a makeup flow rate of 5 mL/min, and a reference flow rate of 30 mL/min.\u003c/p\u003e \u003cp\u003eA custom gas mixture of known composition was used to determine the retention times for the gaseous compounds of interest and to perform signal calibration. The gas mixture was obtained from Jining Xieli Special Gas Co., Ltd and contains, by volume percent, 8.99% carbon dioxide, 8.98% carbon monoxide, 9.02% ethane, 8.93% ethylene, 0.497% acetylene, 9.02% propane, 8.99% propylene, 9.24% hydrogen, 0.633% oxygen, 8.99% methane, and 26.71% nitrogen.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eCross-Talk Validation: Indirect CO\u003csub\u003e2\u003c/sub\u003e Detection\u003c/h2\u003e \u003cp\u003eTo address why CO\u003csub\u003e2\u003c/sub\u003e was detected in the cathode half-cell but not the full cell, Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e, the expected cross-talk product of CO\u003csub\u003e2\u003c/sub\u003e and the Li surface, was probed using titration-mass spectroscopy (T-MS). The reaction of Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e with the H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e titrant is as follows:\u003c/p\u003e \u003cp\u003eLi\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e \u0026rarr; Li\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;H\u003csub\u003e2\u003c/sub\u003eO་CO\u003csub\u003e2\u003c/sub\u003e \u0026uarr;\u003c/p\u003e \u003cp\u003eIn other words, if CO\u003csub\u003e2\u003c/sub\u003e produced at the cathode transfers to the anode to produce solid Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e (i.e., cross-talk), the resulting Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e on the extracted anode would react with H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e to produce CO\u003csub\u003e2\u003c/sub\u003e gas, which can then be collected and measured by MS (Shanghai LingLu Instruments Co., Ltd., QAS 100). In cells where CO\u003csub\u003e2\u003c/sub\u003e is produced (as evidenced by GC) but no cross-talk occurs (e.g., d-LFP||NMC cell), no Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e would form at the anode, and hence no CO\u003csub\u003e2\u003c/sub\u003e would be detected by MS.\u003c/p\u003e \u003cp\u003eAs a precaution, since Li reacts violently with H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e, the system with cross-talk was modeled with Cu||NMC (\u0026ldquo;full cell\u0026rdquo;), using Cu instead of Li. The system without cross-talk was modeled with d-LFP||NMC (cathode half-cell). Furthermore, deposited Li was stripped from the Cu and LFP electrodes prior to cell disassembly. The extracted electrode was titrated with 10 M H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e. A cycled Cu||Li cell, in which CO\u003csub\u003e2\u003c/sub\u003e was not produced during cycling, was tested as a control sample (Fig. S4b). Fresh Cu and d-LFP electrodes were also tested to check for background signals Figs. S4c and S4d).\u003c/p\u003e \u003cp\u003eA calibration equation for Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e was obtained through titration of a Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e standard sample (99.5%, Macklin Co., Ltd). Known masses of Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e were titrated and the resulting CO\u003csub\u003e2\u003c/sub\u003e signals were integrated, yielding a calibration curve (Fig. S4a). Thus, for a given value of an integrated CO\u003csub\u003e2\u003c/sub\u003e signal, the corresponding mass of Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e could be obtained.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eElectrolyte Consumption Analysis\u003c/h2\u003e \u003cp\u003eElectrolyte consumption analysis was performed to quantify the loss of LiFSI and DME with cycling. LiFSI was quantified using ion chromatography (IC) and DME by gas chromatography (GC). Both LiFSI and DME were extracted from the cell using an extraction agent. Hence, the complete method was termed extraction-gas and ion chromatography (E-G\u0026amp;IC).\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e1. Preparation of Extraction Agent\u003c/h2\u003e \u003cp\u003eAn internal standard, 1,2-diethoxyethane (DEE, 99.9%, Aladdin Co., Ltd), was added to correct for matrix effects and sample loss during preparation for GC analysis. The internal standard was added to the sample along with diethylene glycol dimethyl ether (diglyme, 99.99%, Aladdin Co., Ltd), in the form of an extraction agent, which was prepared by dissolving 4 g of DEE in diglyme, and making up the solution to 200 mL in a volumetric flask. The extraction agent was sealed and stored for later use.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e2. Electrolyte Extraction\u003c/h2\u003e \u003cp\u003eA small cut was made in the cycled cell inside an argon-filled glovebox (H\u003csub\u003e2\u003c/sub\u003eO\u0026thinsp;\u0026lt;\u0026thinsp;0.1 ppm, O\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.1 ppm). After injecting 5 mL of the extraction agent, the cell was immediately re-sealed and gently shaken. The cell was allowed to rest under room temperature for 7 days for thorough extraction. The liquid content was then extracted from the cell and syringe filtered for further processing.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e3. Ion Chromatography (IC) for Quantifying LiFSI\u003c/h2\u003e \u003cp\u003eSix standard LiFSI solutions with 0, 25, 50, 100, 150, and 200 mg LiFSI were prepared by dissolving the respective amounts in 1000 mL of deionized water. 5 mL aliquots were analyzed with IC (Dionex Aquion RFIC, Thermo Scientific) to obtain the LiFSI peak area for each corresponding concentration. Linear regression was performed between the peak areas and LiFSI concentrations, forming a calibration curve (Fig. S11a).\u003c/p\u003e \u003cp\u003eSamples extracted from the cell were diluted with deionized water using a dilution factor of 200. A 5 mL aliquot of the diluted solution was measured with IC. The concentration of the diluted sample (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{LiFSI}}\\)\u003c/span\u003e\u003c/span\u003e) was determined from the measured peak area (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{\\text{LiFSI}}\\)\u003c/span\u003e\u003c/span\u003e):\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:{S}_{\\text{LiFSI}}={k}_{\\text{LiFSI}}\\times\\:{c}_{\\text{LiFSI}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}_{\\text{LiFSI}}\\)\u003c/span\u003e\u003c/span\u003e is the slope of the calibration curve. To obtain the absolute mass of LiFSI (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{m}_{\\text{LiFSI}}\\)\u003c/span\u003e\u003c/span\u003e) remaining in the cycled cell, the concentration of the diluted sample was multiplied by the dilution factor (DF\u0026thinsp;=\u0026thinsp;200) and the volume of the extracted solution:\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\:{m}_{\\text{LiFSI}}={c}_{\\text{LiFSI}}\\times\\:200\\times\\:5.245\\:\\text{mL}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere 5.245 mL comes from 5 mL for the extraction agent and 0.245 mL for 0.3 g of 4 M LiFSI/DME.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e4. Gas Chromatography (GC) for Quantifying DME\u003c/h2\u003e \u003cp\u003eA standard solution was prepared by first adding 0.3 g of fresh electrolyte (4 M LiFSI/DME) to 5 mL of extraction agent. This solution was then diluted with diglyme with DF\u0026thinsp;=\u0026thinsp;5. A 1.5 mL aliquot was measured with GC (Nexis GC-2030, Shimadzu). A response factor \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{f}_{\\text{DME}}\\)\u003c/span\u003e\u003c/span\u003e was calculated for DME using the known masses of DME and DEE (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{m}_{\\text{DME}}\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{m}_{\\text{DEE}}\\)\u003c/span\u003e\u003c/span\u003e) in the standard solution and the peak areas extracted from the GC measurement (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{\\text{DME}}\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{\\text{DEE}}\\)\u003c/span\u003e\u003c/span\u003e):\u003cdiv id=\"Equd\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equd\" name=\"EquationSource\"\u003e\n$$\\:{f}_{\\text{DME}}=\\frac{{m}_{\\text{DME}}/{S}_{\\text{DME}}}{{m}_{\\text{DEE}}/{S}_{\\text{DEE}}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eSamples extracted from the cell were diluted with diglyme with DF\u0026thinsp;=\u0026thinsp;5. A 1.5 mL aliquot of the diluted solution was measured with GC. To obtain the absolute mass of DME (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{m}_{\\text{DME, exp}}\\)\u003c/span\u003e\u003c/span\u003e) remaining in the cycled cell, the ratio of the obtained peak areas (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{\\text{DME, exp}}\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{\\text{DEE, exp}}\\)\u003c/span\u003e\u003c/span\u003e) were multiplied by the mass of the DEE added (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{m}_{\\text{DEE, exp}}\\)\u003c/span\u003e\u003c/span\u003e) in the form of the extraction agent, and the response factor (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{f}_{\\text{DME}}\\)\u003c/span\u003e\u003c/span\u003e):\u003cdiv id=\"Eque\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Eque\" name=\"EquationSource\"\u003e\n$$\\:{m}_{\\text{DME, exp}}={m}_{\\text{DEE, exp}}\\times\\:{f}_{\\text{DME}}\\times\\:\\frac{{S}_{\\text{DME, exp}}}{{S}_{\\text{DEE, exp}}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eFirst-Principles Simulation\u003c/h2\u003e \u003cp\u003eDensity functional theory (DFT) calculations were performed with the Vienna \u003cem\u003eab initio\u003c/em\u003e simulation package (VASP).\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e,\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e The generalized gradient approximation and Perdew-Burke-Ernzerhof functional within the DFT framework were applied to determine the electron exchange-correlation energies.\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e Transition metals were treated using the DFT\u0026thinsp;+\u0026thinsp;U augmented approach with U values of 4, 4.4, and 5 eV for Mn, Co, and Ni, respectively.\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e Weak interactions were accounted for with the DFT\u0026thinsp;+\u0026thinsp;D3 method with dispersion correction. All calculations were spin-polarized with a plane wave cutoff energy of 520 eV. Surface calculations were performed with a 2 x 2 x 1 Monkhorst-Pack k-point scheme. A 15 \u0026Aring; vacuum layer was introduced to avoid interactions between periodic slabs.\u003c/p\u003e \u003cp\u003eA five-layer slab of lithium (110) was used to investigate the DME reduction processes and a charged slab of NMC811 was used to investigate the DME oxidation processes. Geometric structural optimization was performed until the force on all atoms was less than 0.02 eV/\u0026Aring;, with an energy convergence criterion set to less than 10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e\u0026nbsp;eV/atom. Transition states were located using the climbing image nudged elastic band (CI-NEB)\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e and dimer methods\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e, with only one imaginary vibrational frequency being identified for each transition state along the reaction coordinate.\u003c/p\u003e \u003cp\u003eThe ground state molecular and ion geometries were optimized using the DFT method at the B3LYP/6-311G+(d, p) level. All DFT calculations were performed using the Gaussian 09 software package. The MD simulations were conducted using the GROMACS 2018 program.\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e The molecules and ions were described using the optimized potentials for liquid simulations all-atom (OPLS-AA) force field.\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e The partial charges on the atoms of the solvent were computed by fitting the molecular restrained electrostatic potential (RESP) at the atomic centers using the B3LYP/aug-cc-pVDZ basis set.\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e The simulation boxes were cubic with a side length of approximately 4 nm and contained different concentrations of LiFSI/DME, LiFSI-1.2DME-3TTE and 1.2 M LiFSI/F5DEE. During the simulations, the temperature was maintained at 333 K using a Nos\u0026eacute;-Hoover thermostat with a relaxation time of 0.2 ps, and the pressure was controlled at 1 bar using a Parrinello-Rahman barostat with a relaxation time of 2.0 ps. From a total simulation duration of 40 ns, the last 20 ns were utilized for analysis.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eAnode Activation\u003c/h2\u003e \u003cp\u003eLi||NMC cells with 4 M LiFSI/DME were used for investigating the effects of anode activation. To activate the anode, the cell was discharged to 1.5 V at 25 ℃ using progressively smaller C-rates: 1 C (140 mA), 0.8 C (112 mA), 0.6 C (84 mA), 0.4 C (56 mA), 0.2 C (28 mA) and 0.1 C (14 mA), with a 5 min rest between each step. To determine the optimal cut-off voltage (i.e., 1.5 V), a series of potentiostatic discharge experiments was performed. The anode activated using 1.5 V was observed to simultaneously support good cycling performance and minimize gas evolution. Results for all voltages tested (1.3\u0026ndash;2.1 V) are shown in Fig. S12, along with comments on why voltages both above and below 1.5 V were not suitable.\u003c/p\u003e \u003cp\u003eActivated and non-activated samples were subjected to three test conditions at 60 ℃: charging to 10% SOC at 0.2 C (28 mA), charging to 100% SOC at 0.2 C (28 mA), and discharging to 0% SOC after 8 cycles (0.2 C-0.5 C, 28 mA-70 mA).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eSEM Characterization\u003c/h2\u003e \u003cp\u003eScanning electron micrographs of the anode samples were acquired with an ultrahigh resolution cold-field emission scanning electron microscope (SEM, Hitachi SU8600). Prior to imaging, anodes extracted from cycled cells were washed and dried thrice in a dry room, with a 10 min DME soak each time. To prepare a cross-sectional sample, a long, thin strip of washed anode was quenched in liquid nitrogen for 60 min, then swiftly split on a stainless-steel blade to expose a clean cross-section on each of the split pieces.\u003c/p\u003e \u003cp\u003eExtracted cathode samples were washed and dried in the same manner as the anode. A small piece of the sample was cut and transferred to an ion beam cross-section polisher (CP, Hitachi ArBlade 5000) under dry atmosphere and polished at room temperature under inert conditions to expose a clean cross-section for imaging.\u003c/p\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003ch2\u003eIonic Conductivities\u003c/h2\u003e \u003cp\u003eIonic conductivities were measured using a benchtop conductivity meter (LeiCi, DDSJ-318).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorresponding Authors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHansen Wang https://orcid.org/0000-0002-6738-1659\u003c/p\u003e\n\u003cp\u003eXiaonan Luo https://orcid.org/0000-0002-0555-5481\u003c/p\u003e\n\u003cp\u003eChuying Ouyang https://orcid.org/0000-0001-8891-1682\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYuchun Wang https://orcid.org/0009-0000-6141-7318\u003c/p\u003e\n\u003cp\u003eSamantha T. Hung https://orcid.org/0000-0001-9448-0962\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eX. L. acknowledges support from Fujian Provincial Natural Science Foundation of China under Grant No. 2024J08378. H. W. acknowledges support from the National Natural Science Foundation of China under Grant No. 22209085, and the support from Fujian Provincial Natural Science Foundation of China under Grant No. 2024J010048. C. O. acknowledges support from the National Natural Science Foundation of China under Grant No. 51962010. This work is also supported by funds from Contemporary Amperex Technology Co., Limited (CATL). We would like to extend our gratitude to Xiangen Cao and other team members at Sikang Chemical Co., Ltd. for their invaluable assistance in the synthesis of F5DEE.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eY.W., X.L., H.W., N.L. and C.O. conceived the idea and designed experiments. Y.C., S.T.H. and Z.C. performed the experiments and analyzed the data. J.S., L.B., E.W. and J.L. performed first-principles simulations. H.L. facilitated the quantitative measurements on gas generations. S.T.H., Y.W. co-wrote the manuscript with revisions from U.U., X.L., H.W., N.L. and C.O.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors are employed by Contemporary Amperex Technology Co., Limited (CATL).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data supporting the findings of this study are included in the Article and its Supplementary Information.\u003cstrong\u003e\u003cbr\u003e \u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAiken, C. P. \u003cem\u003eet al.\u003c/em\u003e A Survey of In Situ Gas Evolution during High Voltage Formation in Li-Ion Pouch Cells. \u003cem\u003eJ. Electrochem. Soc.\u003c/em\u003e \u003cstrong\u003e162\u003c/strong\u003e, A760 (2015).\u003c/li\u003e\n\u003cli\u003eLiu, P. \u003cem\u003eet al.\u003c/em\u003e Revealing Lithium Battery Gas Generation for Safer Practical Applications. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 2208586 (2022).\u003c/li\u003e\n\u003cli\u003eTarascon, J.-M. \u0026amp; Armand, M. Issues and challenges facing rechargeable lithium batteries. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e414\u003c/strong\u003e, 359\u0026ndash;367 (2001).\u003c/li\u003e\n\u003cli\u003eLin, D., Liu, Y. \u0026amp; Cui, Y. Reviving the lithium metal anode for high-energy batteries. \u003cem\u003eNat. Nanotechnol.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 194\u0026ndash;206 (2017).\u003c/li\u003e\n\u003cli\u003eAlbertus, P., Babinec, S., Litzelman, S. \u0026amp; Newman, A. Status and challenges in enabling the lithium metal electrode for high-energy and low-cost rechargeable batteries. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 16\u0026ndash;21 (2018).\u003c/li\u003e\n\u003cli\u003eChen, S., Dai, F. \u0026amp; Cai, M. Opportunities and Challenges of High-Energy Lithium Metal Batteries for Electric Vehicle Applications. \u003cem\u003eACS Energy Lett.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 3140\u0026ndash;3151 (2020).\u003c/li\u003e\n\u003cli\u003eYang, X.-G., Liu, T., Ge, S., Rountree, E. \u0026amp; Wang, C.-Y. Challenges and key requirements of batteries for electric vertical takeoff and landing aircraft. \u003cem\u003eJoule\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1644\u0026ndash;1659 (2021).\u003c/li\u003e\n\u003cli\u003eWang, H. \u003cem\u003eet al.\u003c/em\u003e Rate-Dependent Failure Mechanisms and Mitigating Strategies of Anode-Free Lithium Metal Batteries. \u003cem\u003eACS Appl. Mater. Interfaces\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 12967\u0026ndash;12975 (2023).\u003c/li\u003e\n\u003cli\u003eQian, J. \u003cem\u003eet al.\u003c/em\u003e High rate and stable cycling of lithium metal anode. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 6362 (2015).\u003c/li\u003e\n\u003cli\u003eFan, X. \u003cem\u003eet al.\u003c/em\u003e Highly fluorinated interphases enable high-voltage Li-metal batteries. \u003cem\u003eChem\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 174\u0026ndash;185 (2018).\u003c/li\u003e\n\u003cli\u003eSuo, L. \u003cem\u003eet al.\u003c/em\u003e Fluorine-donating electrolytes enable highly reversible 5-V-class Li metal batteries. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e115\u003c/strong\u003e, 1156\u0026ndash;1161 (2018).\u003c/li\u003e\n\u003cli\u003eLiu, J. \u003cem\u003eet al.\u003c/em\u003e Pathways for practical high-energy long-cycling lithium metal batteries. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 180\u0026ndash;186 (2019).\u003c/li\u003e\n\u003cli\u003eWeber, R. \u003cem\u003eet al.\u003c/em\u003e Long cycle life and dendrite-free lithium morphology in anode-free lithium pouch cells enabled by a dual-salt liquid electrolyte. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 683\u0026ndash;689 (2019).\u003c/li\u003e\n\u003cli\u003eYu, Z. \u003cem\u003eet al.\u003c/em\u003e Molecular design for electrolyte solvents enabling energy-dense and long-cycling lithium metal batteries. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 526\u0026ndash;533 (2020).\u003c/li\u003e\n\u003cli\u003eChen, Y. \u003cem\u003eet al.\u003c/em\u003e Steric Effect Tuned Ion Solvation Enabling Stable Cycling of High-Voltage Lithium Metal Battery. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e143\u003c/strong\u003e, 18703\u0026ndash;18713 (2021).\u003c/li\u003e\n\u003cli\u003eWang, H. \u003cem\u003eet al.\u003c/em\u003e Dual‐Solvent Li‐Ion Solvation Enables High‐Performance Li‐Metal Batteries. \u003cem\u003eAdv. Mater.\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 2008619 (2021).\u003c/li\u003e\n\u003cli\u003eWang, H. \u003cem\u003eet al.\u003c/em\u003e Liquid electrolyte: The nexus of practical lithium metal batteries. \u003cem\u003eJoule\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 588\u0026ndash;616 (2022).\u003c/li\u003e\n\u003cli\u003eLi, Z. \u003cem\u003eet al.\u003c/em\u003e Non-polar ether-based electrolyte solutions for stable high-voltage non-aqueous lithium metal batteries. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 868 (2023).\u003c/li\u003e\n\u003cli\u003ePark, M. S. \u003cem\u003eet al.\u003c/em\u003e A highly reversible lithium metal anode. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 3815 (2014).\u003c/li\u003e\n\u003cli\u003eJie, Y., Ren, X., Cao, R., Cai, W. \u0026amp; Jiao, S. Advanced Liquid Electrolytes for Rechargeable Li Metal Batteries. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 1910777 (2020).\u003c/li\u003e\n\u003cli\u003eGan, L. \u003cem\u003eet al.\u003c/em\u003e Comparative study of thermal stability of lithium metal anode in carbonate and ether based electrolytes. \u003cem\u003eJ. Power Sources\u003c/em\u003e \u003cstrong\u003e551\u003c/strong\u003e, 232182 (2022).\u003c/li\u003e\n\u003cli\u003eXiang, Y. \u003cem\u003eet al.\u003c/em\u003e Gas induced formation of inactive Li in rechargeable lithium metal batteries. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 177 (2023).\u003c/li\u003e\n\u003cli\u003eNie, Q. \u003cem\u003eet al.\u003c/em\u003e Research Progress of Liquid Electrolytes for Lithium Metal Batteries at High Temperatures. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 2302690 (2023).\u003c/li\u003e\n\u003cli\u003eClarke, M. A. \u0026amp; Alonso, J. J. Forecasting the Operational Lifetime of Battery-Powered Electric Aircraft. \u003cem\u003eJ. Aircr.\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 47\u0026ndash;55 (2023).\u003c/li\u003e\n\u003cli\u003eBills, A. \u003cem\u003eet al.\u003c/em\u003e A battery dataset for electric vertical takeoff and landing aircraft. \u003cem\u003eSci. Data\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 344 (2023).\u003c/li\u003e\n\u003cli\u003eZhao, C., Mazo, J. R. \u0026amp; Verstraete, D. Optimisation of a liquid cooling system for eVTOL aircraft: Impact of sizing mission and battery size. \u003cem\u003eAppl. Therm. Eng.\u003c/em\u003e \u003cstrong\u003e246\u003c/strong\u003e, 122988 (2024).\u003c/li\u003e\n\u003cli\u003eWu, Z., Lian, W., Chen, B. \u0026amp; Zheng, C. Research on battery heat generation characteristics and thermal management system applied to a typical eVTOL. \u003cem\u003eAppl. Therm. Eng.\u003c/em\u003e \u003cstrong\u003e257\u003c/strong\u003e, 124187 (2024).\u003c/li\u003e\n\u003cli\u003eAiken, C. P. \u003cem\u003eet al.\u003c/em\u003e An apparatus for the study of in situ gas evolution in Li-ion pouch cells. \u003cem\u003eJ. Electrochem. Soc.\u003c/em\u003e \u003cstrong\u003e161\u003c/strong\u003e, A1548 (2014).\u003c/li\u003e\n\u003cli\u003eEllis, L. D. \u003cem\u003eet al.\u003c/em\u003e Quantifying, Understanding and Evaluating the Effects of Gas Consumption in Lithium-Ion Cells. \u003cem\u003eJ. Electrochem. Soc.\u003c/em\u003e \u003cstrong\u003e164\u003c/strong\u003e, A3518 (2017).\u003c/li\u003e\n\u003cli\u003eBharti, A., Deb, D., Achutharao, G. \u0026amp; Bhattacharyya, A. J. CO \u003csub\u003e2\u003c/sub\u003e Crossover to the Li Anode and Its Implications on the Solid Electrolyte Interphase Composition in a Rechargeable Li\u0026ndash;CO \u003csub\u003e2\u003c/sub\u003e Battery. \u003cem\u003eJ. Phys. Chem. C\u003c/em\u003e \u003cstrong\u003e128\u003c/strong\u003e, 11543\u0026ndash;11551 (2024).\u003c/li\u003e\n\u003cli\u003eCotton, F. A., Wilkinson, G. \u0026amp; Gaus, P. L. \u003cem\u003eBasic inorganic chemistry\u003c/em\u003e. (Wiley India, 2004).\u003c/li\u003e\n\u003cli\u003eZhang, H. \u003cem\u003eet al.\u003c/em\u003e Titration Mass Spectroscopy (TMS): A Quantitative Analytical Technology for Rechargeable Batteries. \u003cem\u003eNano Lett.\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 9972\u0026ndash;9981 (2022).\u003c/li\u003e\n\u003cli\u003eWang, H. \u003cem\u003eet al.\u003c/em\u003e Quantifying Interfacial Reactions in Lithium Metal Batteries for a New Paradigm of Long-Cycling Electrolyte Designs. Preprint at https://doi.org/10.21203/rs.3.rs-3855314/v1 (2024).\u003c/li\u003e\n\u003cli\u003eJung, R., Metzger, M., Maglia, F., Stinner, C. \u0026amp; Gasteiger, H. A. Chemical versus Electrochemical Electrolyte Oxidation on NMC111, NMC622, NMC811, LNMO, and Conductive Carbon. \u003cem\u003eJ. Phys. Chem. Lett.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 4820\u0026ndash;4825 (2017).\u003c/li\u003e\n\u003cli\u003eRen, X. \u003cem\u003eet al.\u003c/em\u003e Enabling High-Voltage Lithium-Metal Batteries under Practical Conditions. \u003cem\u003eJoule\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 1662\u0026ndash;1676 (2019).\u003c/li\u003e\n\u003cli\u003eRen, X. \u003cem\u003eet al.\u003c/em\u003e High-Concentration Ether Electrolytes for Stable High-Voltage Lithium Metal Batteries. \u003cem\u003eACS Energy Lett.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 896\u0026ndash;902 (2019).\u003c/li\u003e\n\u003cli\u003eHan, D., Weng, J., Zhang, X., Tong, Q. \u0026amp; Zhu, M. Review\u0026mdash;Revealing the Intercrystalline Cracking Mechanism of NCM and Some Regulating Strategies. \u003cem\u003eJ. Electrochem. Soc.\u003c/em\u003e \u003cstrong\u003e169\u003c/strong\u003e, 040512 (2022).\u003c/li\u003e\n\u003cli\u003eMetzger, M., Strehle, B., Solchenbach, S. \u0026amp; Gasteiger, H. A. Origin of H \u003csub\u003e2\u003c/sub\u003e Evolution in LIBs: H \u003csub\u003e2\u003c/sub\u003e O Reduction vs. Electrolyte Oxidation. \u003cem\u003eJ. Electrochem. Soc.\u003c/em\u003e \u003cstrong\u003e163\u003c/strong\u003e, A798\u0026ndash;A809 (2016).\u003c/li\u003e\n\u003cli\u003eHarris, O. C., Lee, S. E., Lees, C. \u0026amp; Tang, M. Review: mechanisms and consequences of chemical cross-talk in advanced Li-ion batteries. \u003cem\u003eJ. Phys. Energy\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 032002 (2020).\u003c/li\u003e\n\u003cli\u003eKaufmann, E., Sieber, S. \u0026amp; Schleyer, P. V. R. Carboxylation of lithium compounds: ab initio mechanisms. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e111\u003c/strong\u003e, 4005\u0026ndash;4008 (1989).\u003c/li\u003e\n\u003cli\u003eYu, Z. \u003cem\u003eet al.\u003c/em\u003e Rational solvent molecule tuning for high-performance lithium metal battery electrolytes. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 94\u0026ndash;106 (2022).\u003c/li\u003e\n\u003cli\u003eCourtin, C., Mahseredjian, A., Dewald, A. J., Drela, M. \u0026amp; Hansman, J. A Performance Comparison of eSTOL and eVTOL Aircraft. in \u003cem\u003eAIAA AVIATION 2021 FORUM\u003c/em\u003e (American Institute of Aeronautics and Astronautics, 2021). doi:10.2514/6.2021-3220.\u003c/li\u003e\n\u003cli\u003eHuang, Y., Pan, R., Rehnlund, D., Wang, Z. \u0026amp; Nyholm, L. First‐Cycle Oxidative Generation of Lithium Nucleation Sites Stabilizes Lithium‐Metal Electrodes. \u003cem\u003eAdv. Energy Mater.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 2003674 (2021).\u003c/li\u003e\n\u003cli\u003eFrenck, L., Sethi, G. K., Maslyn, J. A. \u0026amp; Balsara, N. P. Factors That Control the Formation of Dendrites and Other Morphologies on Lithium Metal Anodes. \u003cem\u003eFront. Energy Res.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 115 (2019).\u003c/li\u003e\n\u003cli\u003eRehnlund, D., Ihrfors, C., Maibach, J. \u0026amp; Nyholm, L. Dendrite-free lithium electrode cycling via controlled nucleation in low LiPF6 concentration electrolytes. \u003cem\u003eMater. Today\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 1010\u0026ndash;1018 (2018).\u003c/li\u003e\n\u003cli\u003eWang, H. \u003cem\u003eet al.\u003c/em\u003e Efficient Lithium Metal Cycling over a Wide Range of Pressures from an Anion-Derived Solid-Electrolyte Interphase Framework. \u003cem\u003eACS Energy Lett.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 816\u0026ndash;825 (2021).\u003c/li\u003e\n\u003cli\u003eHe, J. \u003cem\u003eet al.\u003c/em\u003e Key technologies and upgrade strategies for eVTOL aircraft energy storage systems. \u003cem\u003eJ. Energy Storage\u003c/em\u003e \u003cstrong\u003e103\u003c/strong\u003e, 114402 (2024).\u003c/li\u003e\n\u003cli\u003eKresse, G. \u0026amp; Furthm\u0026uuml;ller, J. Efficient iterative schemes for \u003cem\u003eab initio\u003c/em\u003e total-energy calculations using a plane-wave basis set. \u003cem\u003ePhys. Rev. B\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 11169\u0026ndash;11186 (1996).\u003c/li\u003e\n\u003cli\u003eKresse, G. \u0026amp; Furthm\u0026uuml;ller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. \u003cem\u003eComput. Mater. Sci.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 15\u0026ndash;50 (1996).\u003c/li\u003e\n\u003cli\u003ePerdew, J. P., Burke, K. \u0026amp; Ernzerhof, M. Generalized Gradient Approximation Made Simple. \u003cem\u003ePhys. Rev. Lett.\u003c/em\u003e \u003cstrong\u003e77\u003c/strong\u003e, 3865\u0026ndash;3868 (1996).\u003c/li\u003e\n\u003cli\u003eGrimme, S., Antony, J., Ehrlich, S. \u0026amp; Krieg, H. A consistent and accurate \u003cem\u003eab initio\u003c/em\u003e parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e132\u003c/strong\u003e, 154104 (2010).\u003c/li\u003e\n\u003cli\u003eHenkelman, G., Uberuaga, B. P. \u0026amp; J\u0026oacute;nsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e113\u003c/strong\u003e, 9901\u0026ndash;9904 (2000).\u003c/li\u003e\n\u003cli\u003eZhang, H., Qiu, L. \u0026amp; Hu, D. Finite‐Temperature Dimer Method for Finding Saddle Points on Free Energy Surfaces. \u003cem\u003eJ. Comput. Chem.\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 1701\u0026ndash;1706 (2019).\u003c/li\u003e\n\u003cli\u003eAbraham, M. J. \u003cem\u003eet al.\u003c/em\u003e GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. \u003cem\u003eSoftwareX\u003c/em\u003e \u003cstrong\u003e1\u0026ndash;2\u003c/strong\u003e, 19\u0026ndash;25 (2015).\u003c/li\u003e\n\u003cli\u003eKaminski, G. A., Friesner, R. A., Tirado-Rives, J. \u0026amp; Jorgensen, W. L. Evaluation and Reparametrization of the OPLS-AA Force Field for Proteins via Comparison with Accurate Quantum Chemical Calculations on Peptides. \u003cem\u003eJ. Phys. Chem. B\u003c/em\u003e \u003cstrong\u003e105\u003c/strong\u003e, 6474\u0026ndash;6487 (2001).\u003c/li\u003e\n\u003cli\u003eSambasivarao, S. V. \u0026amp; Acevedo, O. Development of OPLS-AA Force Field Parameters for 68 Unique Ionic Liquids. \u003cem\u003eJ. Chem. Theory Comput.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1038\u0026ndash;1050 (2009).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6034057/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6034057/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eUnderstanding and suppressing gas evolution in lithium secondary batteries are critical to optimizing battery performance and ensuring safe operation.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e However, no systematic investigations of gas evolution in ether-based lithium metal batteries (LMBs) have been conducted despite the enticing prospects of LMBs for achieving ultrahigh energy density.\u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e In this work, gas generation in ether electrolyte-based LMBs was quantified and the underlying redox mechanisms were elucidated. Through studying cathode and anode half-cells, it was determined that CO\u003csub\u003e2\u003c/sub\u003e and CO gas were generated at the cathode and CH\u003csub\u003e4\u003c/sub\u003e gas at the anode. Notably, CO\u003csub\u003e2\u003c/sub\u003e and CO were not observed in the full cell as they were consumed at the anode, reacting with lithium to produce solid Li compounds such as Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e. CH\u003csub\u003e4\u003c/sub\u003e generated at the anode is the major contributor of gas generated in the full cell, though its evolution during cycling is not immediate and occurs after an onset point. The total gas volume generated increases dramatically with increasing temperature and decreasing electrolyte concentration. Based on these findings, electrolyte engineering and anode surface activation strategies were explored to control CH\u003csub\u003e4\u003c/sub\u003e production and hence overall gas evolution. In particular, the anode activation approach resulted in increased Li nucleation sites and improved Li deposition morphology, leading to significantly suppressed interfacial reactions, thus delaying the onset of gas evolution by 800% and increasing the cycling life by 400%. Achieving these improvements without altering the electrolyte formulation demonstrates the potential broad applicability of anode activation across various electrolyte chemistries. The performance enhancements beyond merely suppressing gas generation advances the prospects of safer and higher-performing LMBs.\u003c/p\u003e","manuscriptTitle":"Understanding and Suppressing Gas Evolution in Lithium Metal Batteries with Ether-Based Electrolytes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-27 10:01:11","doi":"10.21203/rs.3.rs-6034057/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-chemistry","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nchem","sideBox":"Learn more about [Nature Chemistry](http://www.nature.com/nchem/)","snPcode":"","submissionUrl":"","title":"Nature Chemistry","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"ea34c832-2e40-4eaf-be5c-19502fcc3ada","owner":[],"postedDate":"February 27th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":44864876,"name":"Physical sciences/Chemistry/Energy"},{"id":44864877,"name":"Physical sciences/Chemistry/Electrochemistry/Batteries"}],"tags":[],"updatedAt":"2026-05-08T10:00:38+00:00","versionOfRecord":[],"versionCreatedAt":"2025-02-27 10:01:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6034057","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6034057","identity":"rs-6034057","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}