# Fileset

[Energy Tech - 2025 - Nomura - Rational Choice of Amide‐Based Electrolytes Toward High‐Power Rechargeable Lithium‐Air.pdf](https://mdr.nims.go.jp/filesets/628367a9-8952-452c-89db-4bb0017d03f7/download)

## Creator

[Akihiro Nomura](https://orcid.org/0000-0001-5012-4739), [Shota Azuma](https://orcid.org/0000-0003-1209-3075), Fumisato Ozawa, Morihiro Saito

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Rational choice of amide-based electrolytes toward high-power rechargeable lithium-air batteries](https://mdr.nims.go.jp/datasets/96f37ad0-95b7-44c2-8a16-5f5872a66799)

## Fulltext

Rational Choice of Amide‐Based Electrolytes Toward High‐Power Rechargeable Lithium‐Air BatteriesRational Choice of Amide-Based Electrolytes TowardHigh-Power Rechargeable Lithium-Air BatteriesAkihiro Nomura,* Shota Azuma, Fumisato Ozawa, and Morihiro Saito1. IntroductionThe growing demand for high-energy-density energy storage hasdriven research into lithium–air batteries (LABs), which generateelectricity through the aerobic oxidation of lithium, offering a hightheoretical energy density of 3,500Whkg�1(including oxygen mass).[1] Recent advance-ments in LAB materials and cell assemblyhave led to the development of LAB cellswith Ah-class capacities and energy densitiesranging from 500 to 1000Whkg�1.[2–5]Despite these achievements, high-energy-density LABs are hindered by extremelylow power output. Although their capacityper unit weight is high, limited dischargecurrent rates, typically less than0.5mA cm�2 per unit electrode area, resultin a power density of 100Wkg�1 or less,which remains lower than that of lead-acidbatteries. Besides, the sluggish oxygenreduction reaction (ORR) requires a highoxygen partial pressure during discharge.Under normal air conditions with a ≈21%oxygen gas concentration, discharge perfor-mance deteriorates.[6] These factors severelylimit the commercialization of LABs, mak-ing the enhancement of their power outputan urgent challenge.Because battery power is directly linkedto the discharge current rate, researchers have strived to enhancedischarge current capabilities by supporting ORR catalysts on thecathode surface[7,8] or introducing ORR redox mediators into theelectrolyte.[9] However, their practical applicability in practicalLAB systems remains controversial, as they require substantialmaterial loading to achieve catalytic effects, which increases cellweight and compromises the high-energy-density nature ofLABs. Alternatively, some studies have improved the dischargeperformance through cathode design and electrolyte composi-tion, including the open-cell system in air batteries that requireoxygen gas exchange.[10–14] Cathode pore architecture signifi-cantly affects the discharge performance. A high surface areacathode provides an extensive ORR surface, enabling efficientcurrent generation. The distribution of cathode pores ensuresa smooth supply of Liþ and oxygen to the ORR surface, facilitat-ing battery reactions and enhancing cell capacities.[10–12]Electrolytes also contribute to improved rate capability by facili-tating reactant supply. It has been reported that the oxygen sol-ubility and diffusion in electrolytes positively influence dischargeperformance.[5,13,14] Given the typically low solubility of oxygen (afew ppm) in most liquid electrolyte solvents, the oxygen supply ismore critical than Liþ transfer in the LABs. By utilizing highlyporous, high surface area cathodes combined with low-viscosityamide-based electrolytes that facilitate fast oxygen diffusion, werecently demonstrated a discharge current rate of 5.0mAh cm�2in LAB cells under dry air conditions.[15] This resulted in aA. Nomura, S. AzumaResearch Center for Energy and Environmental MaterialsNational Institute for Materials Science1-1 Namiki, Tsukuba, Ibaraki 305-0044, JapanE-mail: NOMURA.Akihiro@nims.go.jpS. Azuma, F. Ozawa, M. SaitoDepartment of Science and TechnologyFaculty of Science and TechnologySeikei University3-3-1 Kichijoji-Kitamachi, Musashino, Tokyo 180-8633, JapanS. AzumaNational Institute of TechnologyTokyo College1220-2 Kunugida-machi, Hachioji, Tokyo 193-0997, JapanThe ORCID identification number(s) for the author(s) of this articlecan be found under https://doi.org/10.1002/ente.202500556.© 2025 The Author(s). Energy Technology published by Wiley-VCHGmbH. This is an open access article under the terms of the CreativeCommons Attribution License, which permits use, distribution andreproduction in any medium, provided the original work is properly cited.DOI: 10.1002/ente.202500556Lithium–air batteries (LABs) are a promising technology for high-energy-densitybattery storage. However, their open-cell structure for oxygen exchange leads toelectrolyte evaporation, which limits cycling performance under ambient con-ditions. Herein, volatile amide-based electrolytes for LABs using gravimetricanalysis are evaluated. The cell weight change during discharge–charge cyclesconfirms the two-electron oxygen reduction/evolution reactions while alsorevealing that electrolyte evaporation correlates with the solvent vapor pressure.This behavior significantly compromises the cycle performance of low-viscosityamide electrolyte cells. Despite this, rate-dependent cycling experiments dem-onstrate the superior cyclability of the low-viscosity amide electrolyte cells at highcurrent rates (0.8 mA cm�2 or higher), conditions under which cells with aconventional tetraethylene glycol dimethyl ether (TEG)-based LAB electrolyte fail.Scanning electron microscopy and X-ray diffraction analyses show that these cellsexhibit improved rechargeability at high-rate cycles, with discharge productmorphology changing to a more easily decomposable form. This electrolytedesign strategy marks a significant advancement toward developing high-power,high-energy rechargeable LABs.RESEARCH ARTICLEwww.entechnol.deEnergy Technol. 2025, 2500556 2500556 (1 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbHmailto:NOMURA.Akihiro@nims.go.jphttps://doi.org/10.1002/ente.202500556http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/http://www.entechnol.dehttp://crossmark.crossref.org/dialog/?doi=10.1002%2Fente.202500556&domain=pdf&date_stamp=2025-09-07discharge capacity of 3.5 mAh cm�2, nearly twice the capacity oflithium-ion batteries (LiBs, ≈2mAh cm�2), encouraging thedevelopment of a “true” LAB capable of operating in ambient air.For rechargeable LABs, achieving good cyclability relies on oxi-dative tolerance against active oxygen species and electrochemi-cal inertness within the battery voltage range. Since the discoveryof the reversible deposition and decomposition of lithium perox-ide (Li2O2) on a carbon cathode with ether-based electro-lytes,[16,17] tetraethylene glycol dimethyl ether (TEG)-containinglithium bis(trifluoromethanesulfonyl)imide (LiTFSI) has becomea standard electrolyte in many LAB studies.[3–13] The high boilingpoint of TEG (275 °C) effectively suppresses electrolyte volatiliza-tion in the semi-open cell system, supporting long-termdischarge–charge cycle operations under continuous gas flow.However, the high viscosity of TEG-based electrolytes limitstheir current rate capabilities.[15] In contrast, amide-based elec-trolytes, which also enable reversible Li2O2 deposition anddecomposition,[18–22] have a lower viscosity compared to TEG-based electrolytes, presenting a potential breakthrough forhigh-power rechargeable LABs. Additionally, amides areexpected to exhibit greater tolerance to active oxygen species thanethers,[23] which could enhance improved cyclability in LABs.However, the narrow voltage window and high volatility ofamide-based electrolytes limit their use in open-cell LAB sys-tems. Although lithium nitrate (LiNO3) or lithium nitrite(LiNO2) as supporting salts improves the compatibility of amidesas electrolyte solvents by forming a high-quality solid electrolyteinterphase (SEI) on the Li anode,[22] they do not address the vol-atility issue. Electrolyte evaporation is particularly problematicunder lean electrolyte conditions, which is necessary for achiev-ing high energy and power densities in LABs.[24] To avoid solventvolatilization, some studies have used closed-cell systems filledwith pure oxygen,[25] but this approach does not align with thepractical need for lightweight batteries that operate in normalair environments. Therefore, the rational design of amide-basedelectrolytes and cell structures, considering evaporation behaviorin open-cell systems, is urgently needed.This study investigates electrolytes based on seven amide sol-vents with different vapor pressures: N,N-dimethylformamide(DMF), N,N-dimethylacetamide (DMA), N,N-diethylacetamide(DEA), N-methylpyrrolidone (NMP), N-ethylpyrrolidone (NEP),N-methylpiperidone (NMPi), and N-methylcaprolactam (NMC)(Figure 1). The discharge–charge cycle behaviors of cells withthese amide-based electrolytes were examined. These electrolyteswere incorporated into a stack-type LAB cell, a semi-open cell sys-tem that enables integration into a large-capacity battery cell.[2,3]This setup simulated solvent evaporation during the discharge–charge cycle operation. A gravimetric analysis system, whichrecords continuous weight changes during battery operation,[26]showed that electrolyte evaporation is correlated with thevapor pressure of each solvent. While the weight during dis-charge/charge confirmed the 2e�/O2 oxygen reduction/evolutionreactions (ORR/OER) across all amide-based electrolyte cells,high-viscosity amide electrolytes like NMC-based electrolytesexhibited poor cycling performance due to excessively high over-potentials. In contrast, low-viscosity electrolytes, such as DMA-based electrolytes, performed poorly as well due to their highvapor pressures, which led to fast electrolyte dissipation.However, under high-rate cycling conditions, the issue ofelectrolyte volatilization is minimized. The open-cell structurealso helps mitigate the evaporation issue. The DEA-based elec-trolyte strikes an optimal balance between viscosity and evapora-tion, demonstrating superior cyclability at high-rate cycleconditions of 0.8mA cm�2 or more. In contrast, cells withTEG-based electrolytes failed to complete discharge–chargecycles, highlighting the advantages of low-viscosity amide-basedelectrolytes for high-current operations. Scanning electronmicroscopy (SEM) and X-ray diffraction (XRD) analyses of cath-odes after discharge revealed that high-rate conditions altered thedischarge product morphology, forming a film-like structure thatenhanced the rechargeability of the amide-based electrolyte cells.These findings contribute to the development of high-power andhigh-energy rechargeable LABs.2. Results and Discussion2.1. Electrolyte SolventTo investigate the discharge–charge cycle behavior of the LABcells in an open system, seven amides with varying viscositiesand boiling points were used as electrolyte solvents (Figure 1).TEG was also employed as a conventional LAB electrolyte solventused elsewhere.[3–13] The vapor pressure of liquid solvents is typ-ically measured by determining the gas-phase pressure in a her-metically sealed chamber (static method).[27] However, solventswith low vapor pressures, especially TEG with its extremely lowvolatility, are difficult to measure using this approach. To moreaccurately assess the low vapor pressures of various electrolytesolvents, the vaporization rate was measured directly by monitor-ing the continuous weight loss of a stainless-steel dish containingFigure 1. Chemical structure of amides and TEG used as electrolytesolvents in this study.www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (2 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.dethe liquid sample (Figure S2, Supporting Information). Thesedata were then used in the Hertz–Knudsen equation.π ¼ Arvð2πRT=MwÞ1=2 (1)where P is the vapor pressure, A is a constant, rv is the vaporiza-tion rate per unit area,Mw is the solvent’s molecular weight, R isthe gas constant, and T is the absolute temperature.[28] The con-stant A was adjusted to yield a value of P= 19.8mm Hg fordeionized water at room temperature (22 °C).[29] This methodsimplifies the “transpiration method” typically used for materialswith low vapor pressures, where volatilization speed is measuredwith a thermogravimetric apparatus to determine vapor pres-sure.[30] The measured P values, along with the Mw and viscosity(η) values, are listed in Table 1. As expected, solvents withhigher boiling points showed lower P. The value for TEG(1.5� 10�3 mmHg) closely matches the catalog value of0.25 Pa (1.9� 10�3 mmHg) at 25 °C, confirming the validity ofthe vapor pressure measurement in this study.The highest occupied molecular orbital and the lowest unoc-cupied molecular orbital (HOMO-LUMO) levels of each solventmolecules were calculated to assess the redox stability of the elec-trolyte solvents, as summarized in Figure 2a. The results indicatethat the TEG solvent exhibits good redox stability, with a widerenergy gap than compared to the other amides. The higherHOMO levels of amides suggest instability against larger over-potential during charging, while their lower LUMO levels indi-cate poor reductive tolerance. Both negatively affect the cycleperformance when amides are used as LAB electrolyte solvents.However, amide-based electrolytes offer superior cycle perform-ances under high-rate conditions, as discussed later. Figure 2bpresents an electrostatic potential map showing the size andcharge distribution of each solvent molecule. Larger moleculesinduce higher viscosities, consistent with the viscosity data inTable 1. The higher viscosity of DMF (0.94 mPa·s) than DMA(0.92 mPa·s), despite DMF having a smaller Mw, can be attrib-uted to greater polarization and stronger molecular interaction,namely, hydrogen bonding. Similarly, TEG has a lower viscosity(3.25 mPa·s) than NMC (5.10 mPa·s), despite TEG having nearlytwice the Mw of NMC due to the lower polarization of the TEGmolecule. The straight-chain structure of TEG can also contrib-ute to its relatively low viscosity, resulting from its shear-thinningbehavior. However, the significantly higher viscosity of TEG sol-vents compared to low Mw amides limits high-power perfor-mance in LABs, as discussed later.In amide solvents, LiNO3 or LiNO2 were dissolved as support-ing salts. Amides are generally unstable with lithium anodes dueto their low LUMO levels and lack of reductive tolerance.However, LiNO3 and LiNO2 salts enable the use of amides asLAB electrolyte solvents by forming a thin Li2O layer on the lith-ium anode, acting as a high-quality SEI and preventing amidereduction.[20,22,31,32] The NO3� anion in LiNO3 salt electrolytespassivates the Li surface (NO3�þ 2Li ! NO2�þ Li2O), prevent-ing direct contact between amide molecules and Li. The Li2Olayer produced by the NO3� under oxygen atomosphere is quitethin and uniform, successfully preventing needle-like dendritegrowth and surface lithium pulverization throughout the lithiumdissolution/deposition cycles.[31,32] The consumed NO3� anioncan be regenerated in an oxygen atmosphere (NO2� þ 1/2O2! NO3�). In LiNO2 salt electrolytes under oxygen, the NO3�Table 1. Characteristics of the electrolyte solvents.Solvents Molecular weight,Mw/gmol�1Boilingpoint/ °Ca)Viscosity,η/mPa sVapor pressure,P/mmHgc)DMF 73.1 153 0.94 3.5DMA 87.1 165 0.92 1.8DEA 115.2 185 1.34 6.4� 10�1NMP 99.1 202 1.63 3.4� 10�1NEP 113.2 218 1.87 2.0� 10�1NMPi 113.2 250 2.87 1.3� 10�1NMC 127.2 108b) 5.10 7.2� 10�2TEG 222.3 275 3.25 1.5� 10�3a)Catalogue values; b)Boiling point at 6 mm Hg; c)Estimated by measuring thevolatilization rate of each solvent (Figure S2, Supporting Information).Figure 2. HOMO–LUMO levels a) and electrostatic potential mappingb) of the solvent molecules in this study.www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (3 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.deanion, which coexists with NO2�, also contributes to the forma-tion of a thin Li2O layer, making amides compatible as LAB elec-trolyte solvents. The ionic conductivities and viscosities of theprepared electrolytes are listed in Table S1, SupportingInformation. Dissolving LiNO3 and LiNO2 salts increases solventviscosity due to enhancedmolecular interactions between the sol-vent and dissociated salt. This effect is more pronounced forLiNO3, which has a higher solvation energy and undergoesgreater ionic dissociation than LiNO2.[22] The Walden plot ofthe prepared electrolytes (Figure S3, Supporting Information)shows almost identical salt dissociation degrees across all amidesolvents, but TEG exhibited an ionic conductivity one order ofmagnitude lower than that of NMC- or NMPi-based electrolytes.Due to the lower salt dissociation in TEG, LAB cells with TEG-NO3 and TEG-NO2 electrolytes failed to show discharge–chargebehavior under the tested cycle conditions. Therefore, TEG-TFSI,a conventional electrolyte used in many LAB studies,[3–13] wasemployed instead.2.2. Cycle PerformanceDischarge–charge cycle experiments were conducted on LABcells with amide-based electrolytes. Stack-type LAB cells wereemployed (see schematic Figure S1, Supporting Information),where oxygen gas is exchanged through the gas diffusion layer(GDL) cross-section rather than from the plane surface. Thisdesign enables multiple electrode stackings, achieving Ah-classcell capacities while ensuring an efficient oxygen flow path dur-ing discharge and charge.[2,3] A highly porous, high surface areacarbon nanotubes (CNT) sheet (Brunauer–Emmett–Teller (BET)surface area of 910m2 g�1) was used as the cathode to achievehigh-rate performance.[12] However, the semi-open-cell structureinevitably allows some electrolyte evaporation, which becomesmore prominent with high P solvent electrolytes. Figure 3a,bshows the voltage profiles of LAB cells with DMA-, DEA-, andNMC-based electrolytes, along with the cell weight change(W ) profiles (results for other amide-based electrolyte cells areFigure 3. a-c) Cell weight change, W, and voltage profiles of the LAB cells of DMA-, DEA-, NMC-, and TEG-based electrolytes. Additional W and voltageprofiles were shown in Supporting Information (Figure S4). d) Cycle number of the LAB cells under the cycle condition of 0.8mA� 5 h dischargesand charges.www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (4 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.dein Figure S4, Supporting Information). A 48-hour rest period wasapplied before the 0.8 mA� 5 h discharge–charge cycles to fullyequilibrate the cells with the oxygen environment. TheW profilesin Figure 3 present continuous weight loss immediately after cellassembly (t= 0 h) due to electrolyte solvent evaporation, alongwith the weight changes observed during ORR/OER in dis-charge/charge cycles. The weight loss due to solvent evaporationfollowed the order of the solvent’s boiling point and was thusrelated to the solvent P magnitude. Due to the rapid electrolytedissipation, the cells with DMF-based electrolytes could notcomplete discharge–charge cycles (Figure S4, SupportingInformation). The voltage fluctuation in DMF electrolyte cells after≈24 h indicates the loss of cell contact due to significant DMF sol-vent evaporation and alterations in electrolyte composition.The cells with DMA-based electrolytes, which exhibited thenext fastest evaporation, were able to perform discharge–chargecycles. However, these cells showed voltage drops after the 4th or5th charge, likely due to internal short circuits. This behavior iscommonly observed in lean electrolyte conditions, where inho-mogeneous lithium plating leads to Li dendrite growth andmicro-short circuits.[24] As a result of charging failures,DMA-based electrolyte cells had a short cycle life of approxi-mately six cycles. In contrast, the W profiles demonstrate suc-cessful suppression of electrolyte evaporation in NMC-basedelectrolyte cells. However, high overpotentials during both dis-charging and charging limited the number of cycles to 3–6.The high viscosity of NMC-based electrolytes hinders smoothoxygen diffusion and Liþ transfer, increasing overpotentialsand causing battery material deterioration, which leads to rapidloss of rechargeability. In addition, the high-viscosity electrolytesmay poorly wet the cathode pores, which could also contribute tothe short cycle life. However, this was not the case because thecell capacitance immediately after the cell assembly (0.15 F,Figure S5, Supporting Information) was close to that of theCNT-based cathode (0.16 F, 40 F g�1 per CNT weight[12]), sug-gesting that the electrolyte had fully wetted the electrolyte.The balance between solvent volatility, which causes cell contactloss, and solvent viscosity, which impedes reactant supply, makesDEA-based electrolytes the most efficient for LABs, enabling13-cycle runs in the stack-type LAB cell system. The DEA-basedelectrolyte cell was also able to run even under a dry air conditionwith a dew point of�60 to�50 °C, demonstrating the same evap-oration behavior and cycle number as that under pure oxygencondition (Figure S6, Supporting Information). This result posi-tively suggests that the cell system can be applied to develop“true” LABs that can be used in atmospheric air. Althoughthe atmospheric oxygen concentration (≈21%) slightly reducesthe discharge voltage, the decay calculates the energy density lossof no more than 4%. The same cycle number indicates that thecharge failure due to the electrolyte dissipation limits the cyclelife. Though, the cycle life was shorter than that of the TEG-TFSI cell (Figure 3c), which achieved 17 cycles and demonstratedthe most successful evaporation suppression.Many studies have reported LAB cells achieving over a hun-dred cycles with TEG-based electrolytes.[33–35] However, most ofthese cycles do not meet the practical energy requirements ofmodern batteries. The cycle condition we investigated here(0.8mA� 5 h) delivers an energy density of ≈150 Wh kg�1 tothe stack-type LAB cell (≈70mg). This condition canbe considereda minimum benchmark for surpassing the current LiB technol-ogy, which offers an energy density of 100-200 Wh kg�1.[36] Inthis context, the TEG-TFSI electrolyte cell, which achieves17cycles, demonstrates competitiveperformancecompared topre-vious high-energy-density LAB reports,[2–4] though extending itscycle life remains a significant challenge. Figure 3d summarizesthe cycling performance of stack-type LAB cells, showing thatDEA-based electrolyte cells exhibit the best cyclability amongamide-based electrolytes. The graph also highlights the superiorcyclability of the LiNO2 salt electrolytes, which achieve ≈4 morecycles than LiNO3 salt electrolyte cells. This improved perfor-mance is attributed to the redox mediator (RM) effect of theNO2� anion.[22] During charging, the NO2� anion is oxidizedto NO2 at a redox potential of 3.5 V (NO2� ! NO2þ e�, 3.5 Vvs Li/Liþ), prior to the direct oxidation of Li2O2 (Li2O2 ! 2Liþþ O2þ 2e�). The NO2 molecule then chemically oxidizes theLi2O2 discharge product, reducing itself back to NO2�(Li2O2þ 2NO2! 2Liþ þ O2þ 2NO2�). This reaction preventsthe charging voltage from exceeding the NO2� redox potential(3.5 V), thus avoiding the oxidative decomposition of batterymaterials and extending the cycle life of LAB cells. These resultshighlight the advantages of using LiNO2 as an electrolyte sup-porting salt with RM functionality.The observed decrease in W throughout the battery test indi-cates an increase in salt concentration due to solvent dissipation,which decreases the electrolyte vapor pressure according toRaoult’s law. This behavior is more pronounced in high P solventelectrolytes, such as DMF- or DMA-based electrolytes, and is par-ticularly evident in LiNO2 salt electrolytes than in the LiNO3 elec-trolyte. This is because LiNO2 has a lower solvation energy thanLiNO3.[22] Although both salts are fully dissolved in amide sol-vents at 1.0 M, the lower dissociation degree of LiNO2 resultsin higher volatility compared to LiNO3 electrolytes dissolved inthe same solvent. To investigate the evaporation behavior ofthe stack-type LAB cells, the evaporation rate (revap) was deter-mined from the slope of the W profile immediately before thestart of the discharge–charge cycle (t= 48 h). For the high P sol-vent electrolytes (DMF-, DMA-, and DEA-based electrolyte cells),the revap was derived from the slope of the W profile atW=�3.2 mg, corresponding to the 10 wt% decrease in electro-lyte solvent. This adjustment eliminates the influence of salt con-densation and ensures a fair evaluation of the evaporationbehavior. The obtained revap values are plotted in Figure 4 againstthe PMw1/2 of each solvent, which correlates with the pure sol-vent volatilization rate (rv) measured previously (Figure S2,Supporting Information). The graph shows a nonlinear relation-ship between revap and PMw1/2 (or rv), expressed as.revap ∝ ðPMw1=2Þα (2)where α is 0.62, derived from the linear fitting of the data points(black dotted line). An α value less than one indicates that the cellgeometry and the small pores in the cathode/separator impedesolvent evaporation,[37] suggesting that electrolyte retention canbe enhanced by optimizing cell structures and cathode porearchitectures. For instance, applying a wrap film with a 1mmpinhole to a stainless-steel cathode cell with perforated holessignificantly suppressed revap, yielding an α value of 0.49 (graydotted line).www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (5 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.deInterestingly, wrapping with a pinhole significantly extendedthe cycle life of the high P amide electrolyte cell (Figure S7,Supporting Information). The DMA-NO2 and DEA-NO2 cells,which completed 6 and 13 cycles in Figure 2, respectively,increased to 26 and 18 cycles when a pinhole wrap was appliedto the cathode case lid. In contrast, the wrapping had minimalimpact on the cyclability of the TEG-TFSI cell, suggesting thatevaporation was not a problem for the TEG-based electrolyte.However, the intrinsic instability of the TEG electrolyte limitsits cyclability, as previously reported.[16] The DMF-NO2 cell, evenwith a pinhole wrap, failed to function. While DMF-based elec-trolytes are known to enable rechargeability in LABs,[38] their rel-atively fast oxidative decomposition likely rendered the DMFelectrolyte nonfunctional after an initial 48-hour rest period.During this period, the DMF-NO3 and DMF-NO2 cells voltagequickly dropped to 2.8 V within ≈5 h of assembling (Figure S4and S7, Supporting Information), indicating self-discharge dueto the spontaneous oxidation of DMF. It should be also notedthat hermetically sealed LAB cells with no headspace couldnot discharge because they could not intake oxygen. This high-lights the challenge of packaging cells to prevent electrolyte evap-oration while ensuring adequate oxygen inhalation andexhalation for large-capacity battery integration.The weight changes of cells during discharge and charge rep-resent the ORR/OER process in LAB. The time derivatives of W(dW/dt) were derived, with profiles shown in Figure 5a for thefirst three cycles of cells with TEG-TFSI, DEA-NO3, and DEA-NO2 electrolyte (profiles for other electrolyte cells are inFigure S8, Supporting Information). The revap values werededucted from dW/dt to evaluate discharge and charge behaviorwithout considering solvent evaporation. The dW/dt profiles fol-low the red dotted lines (þ0.478mg h�1) during discharge at anapplied current of 0.8 mA, accurately representing the 2e�/O2ORR process of LAB (2Liþ þ O2þ 2e� ! Li2O2). The chargingbehavior, however, varies with electrolyte solvent and salt. First,the TEG-TFSI electrolyte cell showed a decrease in gas evolutionFigure 4. Evaporation rate of the electrolyte solvents from the LAB cells(revap) plotted against the PMw1/2 value of each electrolyte solvent. The filledand open circles represent the data points for LiNO3 and LiNO2 salt electro-lytes, respectively. The black-filled triangle shows the data point for TEG-TFSI electrolyte. The black dotted line represents the linear fitting of the datapoints, deriving 0.62 as the slope of the line, α. The� andþ symbols denotethe revap for, DMF-NO2, DMA-NO2, DEA-NO2 (�), and TEG-TFSI (þ) elec-trolyte cells with a pinhole wrap film, providing a linear fitting line (gray dot-ted line) with a slope of α = 0.49.Figure 5. a) The dW/dt-revap and voltage profiles of TEG-TFSI (black, filled circle), DEA-NO3 (red, filled circle), and DEA-NO2 (red, open circle) electrolytecells for the first three cycle runs. The profiles were derived from the weight profiles shown in Figure 3. The pale blue and red boxes show the regionsof discharge and charge, respectively. The red dotted lines in the dW/dt-revap profiles show the weight increase/decrease rates by 2e�/O2 ORR/OER(�0.478mg h�1 at 0.8mA). b) LSV anodic scan profiles of TEG-TFSI (black straight line), DEA-NO3 (red straight line), DEA-NO2 (red dotted line), andDMF-NO3 (gray straight line) electrolyte cells from 2.95 V to 4.50 V. The superimposed shows the enlarged profiles of the dotted square region.www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (6 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.demidway through charging (≈8, ≈20, and ≈32 h), then exceededthe 2e�/O2 OER line (�0.478mg h�1) toward the end of thecharge (≈10, ≈22, and ≈34 h). This is attributed to a decreasein O2 evolution, followed by CO2 evolution with the charge, indi-cating incomplete O2 recovery in TEG-based electrolyte cells, asconfirmed by online differential electrochemical mass spectrom-etry (DEMS).[26] Conversely, the DEA-NO3 electrolyte cellfollowed the 2e�/O2 OER line (�0.478mg h�1), indicatingimproved O2 evolution and reduced CO2 evolution. This demon-strates the improved rechargeability of the DEA-based electrolyte,providing better stability against attack by active oxygen species.However, the dW/dt profile of the DEA-NO3 cell showed signs ofCO2 evolution at the end of the second or third charge (≈22 and≈34 h), suggesting limited stability against active oxygen. ThedW/dt profile of the DEA-NO2 cell closely mirrored that of theDEA-NO3 cell but showed an earlier reduction in O2 evolutionnear the end of the first charging cycle (≈10 h). This indicatesthe oxidation of the NO2� anion to the NO3� anion(NO2�þ 1/2O2!NO3�), likely accelerated in the highly oxidativeenvironment during charging. The O2 evolved from Li2O2 oxida-tion may have partly been consumed in the oxidation of NO2�anion. Although NO2� is the primary anion in LiNO2 electrolyte,ion chromatographic analysis revealed that the NO3� anionbecame dominant after several discharge–charge cycles, finallybeing the same NO2�/NO3� anion composition with LiNO3 saltelectrolyte.[22] The conversion of NO2� to NO3� is further sup-ported by the disappearance of the RM effect of the NO2� anion.The DEA-NO2 electrolyte cell exhibited a charging voltage plateauof ≈3.5 V during the first charge, but this plateau vanished by the3rd or 4th charge, following the same voltage profile as the DEA-NO3 cell. This explains the longer cycle life of LiNO2 salt electrolytecells compared to LiNO3 cells. Increasing the NO2� anion concen-tration and suppressing the NO2� to NO3� conversion couldenhance cyclability in the LiNO2 salt electrolyte cells.The oxidative degradation of LAB electrolytes results from twoprimary factors: active oxygen attack on the electrolyte moleculesand electrochemical oxidation within the charging potential win-dow. During the charging process, excess active oxygen producedby Li2O2 oxidation deteriorates both the electrolytes and electro-des, particularly in TEG-based electrolyte cells that cause intenseCO2 evolution. The oxygen radical attacks the glyme chain carbonto form alkyl peroxides, which undergo a cascade reaction anal-ogous to combustion to eventually produce CO2, H2O, and car-bonates.[39] A similar oxidation mechanism has been proposedfor amide molecules by the oxygen radical attacking the carbonylcarbon or alkyl carbons adjacent to the nitrogen, but the higheractivation energy and free energy than that of glyme indicate thegreater tolerance of amides to active oxygen attack.[40,41] Instead,the higher HOMO levels of the amides pose oxidative instabilityby the high-voltage electrode. To distinguish between these oxi-dation factors, linear sweep voltammetry (LSV) anodic scanswere performed on amide- and TEG-based electrolyte cells afterassembly. The scan profiles, as shown in Figure 5b, reveal asharp increase in oxidation currents at >4.1 V for the DEA-NO3, DEA-NO2, and DMF-NO3 cells, while the TEG-TFSI cellshows a minor oxidation current at >4.3 V. These results alignwith the expected oxidative stabilities, as indicated by the HOMOlevels of the amides and TEG in Figure 2a. The intense oxidationcurrent peak at 3.6 V for the DEA-NO2 cell is attributed to theoxidation of NO2� anion to NO2 (NO2�!NO2þ e�), with a con-version yield of 46% from the peak area capacity (1.4 C). As thecells contained no discharge product (Li2O2) to decompose, theoxidation currents stem from the electrochemical conversion ofthe electrolytes. LSV profiles indicated that the enhanced rechar-geability of the amide-based electrolyte cells results from theirimproved stability against active oxygens attack. These low-viscosity amide electrolytes enable LAB cells to charge with sup-pressed overpotentials by promoting efficient oxygen and Liþtransport, thus enhancing rechargeability. However, amide-based electrolyte cells experience rapid loss of cyclability oncethe charging voltage exceeds 4.1 V due to the lack of electrochem-ical oxidation stability. In fact, amide electrolyte cells undergooxidative decomposition even at the beginning of the anodicscan. The enlarged scan profiles in Figure 5b show an oxidationcurrent of ≈21 μA for the amide-based electrolyte cells, higherthan the 16 μA observed for the TEG-TFSI cell. Given that the0.1mV s�1 scan rate for cells with a CNT sheet cathode(0.16 F, 40 F g�1 per CNT weight[12]) produces a capacitor currentof 16 μA (gray dotted line), the enlarged profile confirms that theTEG-TFSI cell does not undergo oxidative decomposition nearthe open-circuit voltage (OCV). In contrast, the amide electrolytesundergo electrochemical oxidation of ≈5 μA. This spontaneousoxidation of amide-based electrolytes contributes to voltage decayduring the initial 48-hour rest period, particularly for the DMF-based electrolyte cells, which exhibit the highest fluidity (or thelowest viscosity) among the electrolytes studied here.2.3. Cycle Performance at High Current RatesThe cycle test on the LAB stack cells demonstrated the superiorityof TEG-TFSI as an LAB electrolyte. However, the dW/dt profilesuggests limited stability against oxidative degradation. Amongamide-based electrolytes, DEA-NO2 exhibits the best cyclabilitydue to its low solvent viscosity and minimal solvent dissipationduring cycling. Additionally, the RM effect of the NO2� anion fur-ther enhances cyclability. Although the cycle condition current rateof 0.8mA (0.4mA cm�2 per electrode area) exceeds that of LABstudies reported elsewhere,[33–35] it still fails to meet the practicalbattery power requirements. The cell mass of ≈70mg, which canbe reduced to ≈40mg by using a thinner Li foil anode and thedecreasing electrolyte volume, barely achieves a power densityof 50W kg�1 at 0.8mA. This is significantly lower than the>103W kg�1 delivered by LiB technology.[36] Improving powerperformance is a critical challenge for LAB, prompting furtherinvestigation into cycle performance at higher current rates.Figure 6a,b presents the W and voltage profiles of the DEA-NO2 and TEG-TFSI cells at current rates of 1.6, 2.4, and3.2mA (0.8, 1.2, and 1.6mA cm�2 per electrode area, respec-tively) with a fixed capacity of 4 mAh (2mAh cm�2). The initialrest time before the cycling experiment was reduced to 12 h toretain as much electrolyte as possible. Figure 6c summarizesthe number of cycles achieved at each cycle rate. When the cur-rent rate was increased to 1.6 mA, the DEA-NO2 cell achieved 17cycles, demonstrating improved cyclability under high-rate con-ditions in the shorter experimental times. This improvement isattributed to the high-rate capability of low-viscosity amide elec-trolytes, which reduce internal resistance by enhancing oxygenwww.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (7 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.dediffusion, thus enabling discharge at high current.[15] The DEA-NO2 cell was rechargeable at a rate of 2.4 mA for 16 cycles.However, at 3.2mA, the cycle number decreased to six, withincreasing W behavior during the cycle test, indicating chargingfailure. The cell reached the cutoff voltage of 4.5 V before fullycharging, and DEA-based electrolytes undergo significant oxida-tion at voltages >4.1 V, accelerating cell degradation. Reducingthe charge current rate only will improve cyclability, even underthe high-rate cycling conditions. The high-rate cycling experi-ment also revealed the TEG-TFSI electrolyte limitations.The TEG-TFSI cell exhibited only one cycle at 1.6 mA(0.8mA cm�2) and failed to operate at 2.4mA (1.2 mA cm�2),indicating its suitability only for extremely low-power devices.Figure 6d,e shows the dW/dt profiles for the DEA-NO2 andTEG-TFSI cells during their first discharge and charge cycles.The DEA-NO2 cells follow the 2e�/O2 lines (red dotted lines),indicating ideal ORR/OER at all cycling current rates. Thedecrease in gas evolution after a charge of≈3mAh representsthe NO2� anion oxidation to NO3�, as previously discussed.In contrast, the TEG-TFSI cell fails to maintain a stable dischargevoltage plateau at 1.6mA (0.8 mA cm�2), and its dW/dt profileduring charging indicates insufficient oxygen evolution. Thisis attributed to the high viscosity of the TEG-TFSI electrolyte,which hinders smooth oxygen exchange. These results confirmthat low-viscosity amide-based electrolytes are essential for high-power rechargeable LABs. However, their volatility poses a chal-lenge for cell packaging. It is crucial to maintain fast oxygen flowwhile suppressing electrolyte dissipation. Although volatilizationcan be suppressed by gelling the liquid electrolyte or replacing itwith solid Liþ-conductive electrolytes,[42–44] achieving high cur-rent rate capability remains difficult. Because the cell’s geometricdesign plays an important role in electrolyte retention, studyingthe gas flow around the cell structure would provide a new strat-egy for developing high-power, long-life rechargeable LABs.Figure 6. a,b) Cell weight change, W, and voltage profiles of the DEA-NO2 (a) and TEG-TFSI (b) electrolyte cells at the cycle conditions of 1.6, 2.4, and3.2mA (0.8, 1.2, and 1.6 mA cm�2) current rates. The cycle capacity was fixed to 4.0 mAh (2.0 mAh cm�2). c) Cycle number of the LAB cells at differentcurrent rates of 0.8, 1.6, 2.4, and 3.2 mA (0.4, 0.8, 1.2, and 1.6 mA cm�2). d,e) The dW/dt-revap and voltage profiles of DEA-NO2 (d) and TEG-TFSI (e)electrolyte cells for the first cycle runs. The red dotted lines in the dW/dt-revap profiles represent the weight increase/decrease by 2e�/O2 ORR/OER at eachcurrent rate.www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (8 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.deTo further investigate the mechanism behind high-ratecyclability, the discharge products deposited on the CNT cathodewere analyzed. Figure 7 a,b shows SEM images of the cathodesafter an 8mAh (4mAh cm�2) discharge in DEA-NO2 cells underlow- (0.4 mA (a)) and high-rate (2.4 mA (b)) conditions. Thecorresponding energy-dispersive spectroscopy (EDS) oxygen (O)elemental maps are superimposed on the SEM images (additionalSEM images at different magnifications are shown in Figure S9,Supporting Information). Under the low-rate condition (a), dis-crete particles, ≈200 nm in diameter, precipitated on the CNTbundles, which were covered with O. This suggests two pathwaysfor discharge product formation, that is, the solution and surfacereduction pathways.[45] The ORR in LAB occurs as followsLiþþO2� þ e� ! LiO2� (3)LiO2� ! LiO2 (4a)2LiO2 ! Li2O2� þ O2 (4b)LiO2� þ Liþþ e� ! Li2O2� (5)where * denotes the species adsorbed on the cathode surface.First, one-electron oxygen reduction occurs at the cathode surface(3), followed by two pathways. The one-electron reduction interme-diate (LiO2*) is released from the cathode surface (4a) andundergoes disproportionation (4b), forming highly crystalline par-ticles, typically in a toroidal shape, on the cathode surface. This isknown as the “solution” pathway. Alternatively, the LiO2* interme-diate remains on the surface, further accepting one-electron reduc-tion (5), forming a thin Li2O2 film known as “surface” reduction.The SEM image under the low-rate condition (a) suggests a mixedreduction pathway, forming both discrete particles and thin filmswith different discharge product morphologies. In contrast, theSEM image under high-rate conditions (b) likely favors surfacereduction, as no distinct particles precipitate, and the CNT bundlesare uniformly covered with a thin Li2O2 film. XRD analysis of theCNT cathodes after discharge (Figure 7c) supports this trend.While both spectra confirm that Li2O2 is the main discharge prod-uct, the low-rate condition slightly sharpens the Li2O2 reflectionpeaks, indicating higher crystallinity. The low-rate spectrum alsoshows a small peak at 2thete= 33.5°, between the 100 and 101reflections of Li2O2. Although we could not precisely assign thispeak, it suggests partial evolution of the Li2O 111 reflection, likelydue to Li2O2 disproportionation (2Li2O2 ! 2Li2OþO2).[46]Though Li2O2 disproportionation typically occurs at>250 °C, traceimpurities might promote the reaction even at room temperature.The morphology of the discharge product influences therechargeability of the subsequent charge. Figure 7d comparestwo charge profiles after a 4.0mAh discharge, with dischargeFigure 7. a,b) SEM images of CNT sheet cathodes after 8 mAh (4mAh cm�2) discharges at discharge rates of 0.4 mA (a) and 2.4 mA (b). The insets showthe corresponding EDS oxygen (O) element mappings. The yellow dotted circles in (a) indicate the toroid particles deposition on CNT bundle surface.c) XRD profiles of CNT sheet cathodes after 8 mAh discharges at 0.4 mA (black line) and 2.4 mA (gray line). The inset shows the enlarged XRD profiles ofthe dotted square region. The numbers in the graph indicate the miller indices for the Li2O2 crystal diffraction. The * and # symbols denote the reflectionsfrom a stainless-steel plate (sample holder) and GDL, respectively. d) Discharge and charge profiles of LAB cells at a charge rate of 0.8 mA after the 4 mAh(2mAh cm�2) discharges at 0.8 mA (solid red line) and 4.0mA (dotted red line). The electrolyte was 1.0 M LiNO3 dissolved in an equal volume mixture ofDEA and DMA. e) Schematic illustrations of ORR near the carbon cathode surface at low (left) and high (right) discharge current rates. The low dischargerate rather produces Li2O2 solid particles due to the slow ORR kinetics allowing free LiO2 and its disproportionation, while the high discharge rate tends toproduce surface filmy Li2O2 deposit due to the fast ORR.www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (9 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.derates of 0.8mA and 4.0 mA, while the charge rate remained fixedat 0.8mA. LiNO3 salt was used to minimize the RM effect andcompare charge behavior, while DMA was added to the DEAsolvent to reduce internal resistance and facilitate a high dis-charge rate of 4.0 mA (2.0mA cm�2). The cells discharged atboth 0.8 mA and 4.0 mA current rates, maintaining stable voltageplateaus of 2.75 V and 2.60 V, respectively. Although the chargerate was constant, the cell discharging at 4.0 mA exhibited alower charge voltage throughout the charging process, indicatingimproved rechargeability under high-rate discharge. This effectis associated with the formation of highly decomposable Li2O2.The charge profile of the low-rate discharge cell exhibited a 4.3 Vplateau near the end of the charge (>≈3.5mAh), indicating thatsevere electrolyte oxidation occurred in this high-voltage region,as shown in the LSV scan in Figure 5b (>4.1 V). The Li2O2 par-ticles deposited during the low-rate discharge should have lostelectrical contact without being fully decomposed. Figure 7eillustrates the ORR process under low- and high-rate dischargeconditions. At low discharge rates, the solvated LiO2 intermediatein the electrolyte gradually forms Li2O2 particles, however, theseparticles are less likely to decompose during subsequent charg-ing due to distance from the cathode surface and the high crys-tallinity. At higher discharge rates, the reduction of LiO2* occursat the cathode surface, driven by a relative shortage of surfaceoxygen to accept cathode electrons. This results in a thinner dis-charge product, which is more easily decomposed due toincreased contact with the cathode surface. Furthermore,although Li2O2 is typically insulating, non-crystalline amorphousLi2O2 has been shown to be conductive, enhancing rechargeabil-ity by improving the electrical conductivity of the discharge prod-uct.[47,48] This also accounts for the improved cyclability underhigh-rate cycling conditions.The discharge product morphology in LABs can beinfluenced by factors such as the rate-dependent Li2O2 mor-phology,[49] the acceptor/donor numbers (AN/DN) of the elec-trolyte,[45] traces of protic solvent molecules,[50] and cathodesurface chemistry.[51–54] Amide electrolytes, with approximatelytwice the DN of TEG,[15] tend to stabilize the LiO2 intermediateand promote the growth of discrete particles via the solutionpathway, in contrast to the TEG-TFSI electrolyte. However,high-rate discharge suppresses the solution pathway andencourages the deposition of Li2O2 as a thin film through sur-face reduction, even in amide electrolytes. Previous studieshave suggested that a thin-film discharge product hinders elec-tron transfer and causes premature discharge termination.[55]This, however, is not the case in this study, which uses a highlyporous CNT sheet cathode with a high surface area. The CNTcathode offers a large ORR surface and ensures rapid reactantsupply, preventing thin-film discharge products from blockingthe cathode pores and terminating the reaction. Therefore, aCNT cathode paired with a low-viscosity amide-based electrolytepresents a promising combination for developing high-energy,high-power LABs. However, electrolyte evaporation through theoxygen ventilation path is inevitable, necessitating careful con-sideration of cell packaging and operating conditions to fullyrealize the full potential of high-energy-density LABs. Thisstudy provides a rational design approach for LAB electrolytesaimed at high-power, rechargeable LABs.3. ConclusionTo clarify the rational choice of amide-based electrolytes, thisstudy investigated the discharge and charge cycle performanceof LAB cells in an open-cell structure. The gravimetric analysisdemonstrated the 2e�/O2 ORR/OER across all amide-based elec-trolyte cells, along with the continuous evaporation behavior ofthe electrolytes that correlates with the vapor pressure of the elec-trolyte solvents. The discharge–charge cycle experiment revealedthe superior cycle performance of the cells with low-viscosityamide-based electrolytes, such as DEA-NO2, at high currentrates, conditions that cells with TEG-based electrolyte were neverable to complete. This is due to the fast oxygen and Liþ transferpropensity of the low-viscosity amide electrolyte to facilitate highcurrent rate discharge. The study revealed that high-rate condi-tion improves the rechargeability of amide-based electrolyte cellsby forming a filmy Li2O2 discharge product that is easier todecompose during charging. However, low-viscosity amide elec-trolyte is highly volatile, thus open-cell structure that ensuresrapid oxygen gas supply while suppressing electrolyte evapora-tion is inevitable to achieve the superior cycle performance athigh rates. Developing such open-cell structures and exploringlow-viscosity electrolyte with low vapor pressure will improvethe cycle performance of LAB, paving the way for the develop-ment of high-power and high-energy rechargeable LABs.4. Experimental SectionMaterials: Lithium nitrate (LiNO3, 99.0%, Kishida Chemical), lithiumnitrite (LiNO2, n-hydrate form, Mitsuwa Chemicals), and LiTFSI(99.95%, Kishida Chemical) were dehydrated overnight at 110 °C in a vac-uum before being dissolved in solvents: DMF (super dehydrated grade,Fujifilm Wako Chemicals), DMA (super dehydrated grade, FujifilmWako Chemicals), DEA (>99.0%, Tokyo Chemical Industry), NMP (superdehydrated grade, Fujifilm Wako Chemicals), NEP (>98.0%, TokyoChemical Industry), NMPi (99%, Sigma-Aldrich), NMC (99%, Sigma-Aldrich), and TEG (lithium battery grade, Kishida Chemical) at a 1.0 Mconcentration. The resulting electrolytes were labeled as X–Y, where X indi-cates the solvent and Y refers to the salt anion. The H2O content,measured using a Karl Fischer moisture meter (CA-31, MitsubishiChemical Corp.), was ≈150 ppm across all electrolytes. Single-walledCNTs (ZEONANO SG101, obtained from Sigma-Aldrich) were dispersedin deionized water at a concentration of 0.1 wt% using an ultrasonichomogenizer (450D, Branson). The slurry was then filtered through car-bon paper (300 μm thick, Kureha Corp.) to obtain a CNT sheet cathode,which was supported on the carbon paper as a GDL. The CNT loading was2.1 mg cm�2, and the CNT layer thickness was ≈130 μm. The BET surfacearea was 910m2 g�1 per the CNT loading weight, determined by the nitro-gen adsorption isotherm at 77 K (3FLEX, Micromeritics).Cell Assembly and Battery Testing: A stack-type LAB cell (the schematicillustration was shown in Figure S1, Supporting Information) was fabri-cated using a comprising layer of lithium foil (ϕ16mm, 2.0 cm2 electrodearea, 200 μm thick, Honjo Metal), a porous polyolefin separator (20 μmthick, W-scope), and a CNT sheet cathode combined with a carbon paperGDL (ϕ16mm) between two stainless steel plates. The separator and cath-ode were immersed in 32 μL of electrolyte, matching the void volume ofthe cathode and separator to prevent leakage. The total mass of the batterymaterials, including GDL carbon paper, CNT cathode, separator, Li foil,and electrolyte, was ≈70mg. The assembled stack cell was sealed in aCR2032 coin cell configuration with a perforated mesh on the cathode side(50% aperture ratio, Hohsen Corp.). This design allows oxygen gas to flowthrough the gap between the hole mesh of the cathode case and the stain-less steel plates sandwiching the stack cell components (Figure S1,www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (10 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.deSupporting Information). The cell was placed in a 5 L air-tight chamberwith a continuous pure oxygen flow of 150mLmin�1. A homemade gravi-metric analysis system, set at room temperature, was used to monitor thecell weight and voltage profile during battery operation.[26] The systeminvolved a repeated weighing of the cell, connected to a battery-testingmachine (ECAD-1000, EC Frontier), using a high-precision weight module(AD-4212B-23, A&D Co., Ltd.). Before discharging and charging, the cellwas left in an OCV state for 48 h. The cell was then subjected to discharge–charge cycles at a constant current (0.8mA� 5 h) within the voltage rangeof 2.0-4.5 V. The number of cell cycles was determined by the number ofdischarges that provided a 4mAh (2mAh cm�2 per electrode area) capac-ity, specifically, those discharges that reached the 2.0 V cutoff voltageminus one.Characterization: The ionic conductivity, viscosity, and density of theelectrolytes were measured using a pH/ion meter (SevenExcellenceS500, Mettler Toledo) and a rolling ball viscometer (Lovis2000ME,Anton Paar), respectively. The HOMO–LUMO levels of the solvent mol-ecules were calculated using the MOPAC software with the PM7Hamiltonian. The conductor-like screening model (COSMO) wasemployed to reproduce solvent effects. LSV anodic scans were conductedwith an electrochemical potentiostatic system (SP-50e, Biologic) at a scanrate of 0.1 mV s�1 from 2.95 V to 4.50 V. The cathode morphology andstructures were analyzed using field-emission SEM (JSM-7800 F, JEOL,5 keV accelerating voltage) equipped with an EDS analyzer (X-MaxN,Oxford Instruments). XRD measurement was carried out on an X-ray dif-fractometer (SmartLab, Rigaku) using a CuKα source (λCuKα= 1.542 Å).Supporting InformationSupporting Information is available from the Wiley Online Library or fromthe author.AcknowledgementsThis study was partially supported by the JST Adaptable and SeamlessTechnology Transfer Program through Target-Driven Research andDevelopment (A-STEP; grant no. JPMJTM22AQ), JSPS KAKENHI (grantnos. 24K08154 and 24K08585), and Japan Keirin Autorace (JKA) founda-tion (grant no. 2024M-380), and the National Institute for MaterialsScience (NIMS) Joint Research Hub Program. The authors thank HisaeUematsu for the technical assistance. Material characterization and cellassembly were performed using the NIMS Battery Research Platform.Conflict of InterestThe authors declare no conflict of interest.Data Availability StatementThe data that support the findings of this study are available from thecorresponding author upon reasonable request.Keywordsamide-based electrolytes, carbon nanotubes, high-energy-densitybatteries, high-power batteries, lithium, lithium–air batteries, oxygenReceived: March 21, 2025Revised: May 7, 2025Published online:[1] W. J. Kwak, Rosy, D. Sharon, C. Xia, H. Kim, L. R. Johnson,P. G. Bruce, L. F. Nazar, Y. K. Sun, A. A. Frimer, M. Noked,S. A. Freunberger, D. Aurbach, Chem. Rev. 2020, 120, 6626.[2] J. O. Park, M. Kim, J. H. Kim, K. H. Choi, H. C. Lee, W. Choi, S. B. Ma,D. Im, J. Power Sources 2019, 419, 112.[3] Y. Kubo, K. Ito, ECS Transactions 2014, 62, 129.[4] S. Q. Zhao, L. Zhang, G. N. Zhang, H. B. Sun, J. Y. Yang, S. G. Lu, J.Energy. Chem. 2020, 45, 74.[5] Z. Wen, Y. Liu, K. Li, S. Yang, H. Zhou, P. He, National Science Review2025, 12, nwaf059.[6] H. J. Kwon, H. C. Lee, J. Ko, I. S. Jung, H. C. Lee, H. Lee, M. Kim,D. J. Lee, H. Kim, T. Y. Kim, D. Im, J. Power Sources 2017, 364, 280.[7] C. Z. Yang, R. A. Wong, M. S. Hong, K. Yamanaka, T. Ohta,H. R. Byon, Nano Lett. 2016, 16, 2969.[8] X. Xin, K. Ito, Y. Kubo, Electrochim. Acta 2018, 261, 323.[9] X. Gao, Y. Chen, L. Johnson, P. G. Bruce, Nat. Mater. 2016, 15, 918.[10] S. B. Ma, D. J. Lee, V. Roev, D. Im, S. G. Doo, J. Power Sources 2013,244, 494.[11] K. Sakai, S. Iwamura, R. Sumida, I. Ogino, S. R. Mukai, ACS Omega2018, 3, 691.[12] A. Nomura, E. Mizuki, K. Ito, Y. Kubo, T. Yamagishi, M. Uejima,Electrochim. Acta 2021, 400, 139415.[13] O. Wijaya, P. Hartmann, R. Younesi, I. I. E. Markovits, A. Rinaldi,J. Janek, R. Yazami, J. Mater. Chem. A 2015, 3, 19061.[14] H. Wan, Q. Bai, Z. Peng, Y. Mao, Z. Liu, H. He, D. Wang, J. Xie,G. Wu, J. Mater. Chem. A 2017, 5, 24617.[15] A. Nomura, S. Azuma, F. Ozawa, M. Saito, J. Power Sources 2025, 633,236426.[16] S. A. Freunberger, Y. H. Chen, N. E. Drewett, L. J. Hardwick, F. Barde,P. G. Bruce, Angew. Chem. Int. Edit. 2011, 50, 8609.[17] C. Laoire, S. Mukerjee, E. J. Plichta, M. A. Hendrickson,K. M. Abraham, J. Electrochem. Soc. 2011, 158, A302.[18] J. Uddin, V. S. Bryantsev, V. Giordani, W. Walker, G. V. Chase,D. Addison, J. Phys. Chem. Lett. 2013, 4, 3760.[19] W. Walker, V. Giordani, J. Uddin, V. S. Bryantsev, G. V. Chase,D. Addison, J. Am. Chem. Soc. 2013, 135, 2076.[20] D. W. Kim, D. H. Lee, S. M. Ahn, D. Y. Kim, J. Suk, D. H. Choi,Y. Kang, J. Power Sources 2017, 347, 186.[21] C. Chen, L. Li, J. Su, C. Zhang, X. Chen, T. Huang, A. Yu, Electrochim.Acta 2017, 243, 357.[22] S. Azuma, M. Sano, I. Moro, F. Ozawa, M. Saito, A. Nomura,Electrochim. Acta 2024, 489, 144261.[23] V. S. Bryantsev, V. Giordani, W. Walker, M. Blanco, S. Zecevic,K. Sasaki, J. Uddin, D. Addison, G. V. Chase, J. Phys. Chem. A2011, 115, 12399.[24] H. Kwon, J. Baek, H.-T. Kim, Energy Storage Materials 2023, 55, 708.[25] K. Nishioka, M. Tanaka, T. Goto, R. Haas, A. Henss, S. Azuma,M. Saito, S. Matsuda, W. Yu, H. Nishihara, H. Fujimoto,M. Tobisu, Y. Mukouyama, S. Nakanishi, ACS Appl. Mater.Interfaces 2024, 16, 46259.[26] A. Nomura, K. Ito, D. Y. W. Yu, Y. Kubo, J. Power Sources 2024, 592,233924.[27] A. Smith, A. W. C. Menzies, J. Am. Chem. Soc. 1910, 32, 1412.[28] F. J. Wilkins, Nature 1930, 125, 236.[29] A. Wexler, J. Res. Natl. Bur. Stand., A Phys. Chem. 1976, 80A, 775.[30] R. Viswanathan, T. S. L. Narasimhan, S. Nalini, J. Phys. Chem. B 2009,113, 8362.[31] F. Ozawa, K. Koyama, D. Iwasaki, S. Azuma, A. Nomura, M. Saito,Electrochemistry 2024, 92, 047003.[32] M. Saito, T. Fujinami, M. Sohmiya, Y. Hayashi, H. Koyama, H. Otsuka,K. Ito, Y. Kubo, T. Horiba, J. Electrochem. Soc. 2021, 168, 010520.[33] Y. Zhu, H. Sun, T. Zhang, Q. Li, Z. Xue, M. Yu, J. Li, X. Wang,Langmuir 2024, 40, 23511.www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (11 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.de[34] Y. Pan, A. Hu, R. Xu, J. Chen, B. Yang, T. Li, K. Li, Y. Li, Z. W. Seh,J. Long, ACS Sustain. Chem. Eng. 2024, 12, 17177.[35] X. Li, G. Zhang, D. Zhang, R. Yang, H. Yu, X. Zhang, G. Lian, H. Hou,Z. Guo, C. Hou, X. Yang, F. Dang, Energy Environ. Sci. 2024, 17, 8198.[36] A. M. Stux, K. Swindler-Lyons, Batteries, Naval Research Laboratory,NRL/MR/6110—07-9073, 2007, https://apps.dtic.mil/sti/citations/ADA472272.[37] M. Tanaka, S. Inasawa, Colloids Surf. A Physicochem. Eng. Asp. 2023,672, 131693.[38] Y. Chen, S. A. Freunberger, Z. Peng, F. Bardé, P. G. Bruce, J. Am.Chem. Soc. 2012, 134, 7952.[39] S. Freunberger, Y. Chen, N. E. Drewett, L. J. Hardwick, F. Bardé,P. G. Bruce. Angew. Chem. Int. Ed. 2011, 50, 8609.[40] V. Bryanstev, V. Giordani, W. Walker, M. Blanco, S. Zecevic, K. Sasaki,J. Uddin, D. Addison, G. V. Chase, J. Phys. Chem. A 2011, 115, 12399.[41] D. Sharon, D. Hirsberg, M. Afri, A. Garsuch, A. A. Frimer, D. Aurbach,J. Phys. Chem. C 2014, 118, 15207.[42] J. Hassoun, F. Croce, M. Armand, B. Scrosati, Angew. Chem. Int. Ed.2011, 50, 2999.[43] M. Longo, M. Gandolfo, C. Francia, S. Bodoardo, M. Sangermano,J. Amici, Electrochim. Acta 2023, 466, 143026.[44] Q. Chen, W. Tang, M. Kuai, G. Gao, Y. Hou, Q. Huang, W. Fang,Y. Chen, L. Liu, Y. Wu, ACS Appl. Mater. Interfaces 2025, 17, 16245.[45] N. B. Aetukuri, B. D. McCloskey, J. M. Garcia, L. E. Krupp,V. Viswanathan, A. C. Luntz, Nat. Chem. 2015, 7, 50.[46] K. P. C. Yao, D. G. Kwabi, R. A. Quinlan, A. N. Mansour, A. Grimaud,Y.-L. Lee, Y.-C. Lu, Y. Shao-Horn, J. Electrochem. Soc. 2013, 160, A824.[47] B. Lee, J. Kim, G. Yoon, H.-D. Lim, I.-S. Choi, K. Kang, Chem. Mat.2015, 27, 8406.[48] S. T. Plunkett, H.-H. Wang, S. H. Park, Y. J. Lee, J. Cabana, K. Amine,S. Al-Hallaj, B. P. Chaplin, L. A. Curtiss, ACS Appl. Energy Mater. 2020,3, 12575.[49] R. R. Mitchell, B. M. Gallant, Y. Shao-Horn, C. V. Thompson, J. Phys.Chem. Lett. 2013, 4, 1060.[50] S. Wu, J. Tang, F. Li, X. Liu, H. Zhou, Chem. Commun. 2015, 51,16860.[51] M. Saito, Y. Tachikawa, T. Fujinami, K. Mikami, Y. Hayashi,H. Shiroishi, D. Streich, E. J. Berg, P. Novak, Electrochemistry 2018,86, 265.[52] A. Dutta, K. Ito, A. Nomura, Y. Kubo, Adv. Sci. 2020, 7, 2001660.[53] S. Azuma, M. Sano, I. Moro, F. Ozawa, M. Saito, A. Nomura, J. Phys.Chem. C 2023, 127, 7087.[54] S. Azuma, I. Moro, M. Sano, F. Ozawa, M. Saito, A. Nomura, J.Electrochem. Soc. 2024, 171, 100511.[55] J.-J. Xu, Z.-W. Chang, Y. Wang, D.-P. Liu, Y. Zhang, X.-B. Zhang, Adv.Mater. 2016, 28, 9620.www.advancedsciencenews.com www.entechnol.deEnergy Technol. 2025, 2500556 2500556 (12 of 12) © 2025 The Author(s). Energy Technology published by Wiley-VCH GmbH 21944296, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ente.202500556 by Akihiro Nomura - National Institute For , Wiley Online Library on [07/09/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.entechnol.de Rational Choice of Amide-Based Electrolytes Toward High-Power Rechargeable Lithium-Air Batteries 1. Introduction 2. Results and Discussion 2.1. Electrolyte Solvent 2.2. Cycle Performance 2.3. Cycle Performance at High Current Rates 3. Conclusion 4. Experimental Section