# Fileset

[advpub_23-00142.pdf](https://mdr.nims.go.jp/filesets/2bd10aa7-d8f5-42d1-8384-1a1bb508a33f/download)

## Creator

Fumisato OZAWA, Kazuki Koyama, Daiki IWASAKI, Shota AZUMA, [Akihiro NOMURA](https://orcid.org/0000-0001-5012-4739), Morihiro SAITO

## Rights

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

## Other metadata

[The Effect of Li Ion and Anion Supply on Li Dissolution/Deposition Behavior in LiNO3 Electrolyte Solutions for Li-Air Batteries](https://mdr.nims.go.jp/datasets/d6702d35-0ce3-48b4-b576-d3ba83ebdcd7)

## Fulltext

untitledArticle Electrochemistry, (in press) 1–7The Effect of Supply Rate of Li Ion and Anion on Li Dissolution/Deposition Behaviorin LiNO3 Electrolyte Solutions for Li-Air BatteriesFumisato OZAWA,a,*,§ Kazuki KOYAMA,a Daiki IWASAKI,a,§ Shota AZUMA,a,§Akihiro NOMURA,b,§ and Morihiro SAITOa,*,§a Faculty of Science and Technology, Seikei University, 3-3-1 Kichijoji-Kitamachi, Musashino, Tokyo 180-8633, Japanb Center for Green Research on Energy and Environmental Materials, National Institute for Materials Science (NIMS),1-1 Namiki, Tsukuba, Ibaraki 305-044, Japan* Corresponding authors: fumisato-ozawa@st.seikei.ac.jp (F. O.), mosaito@st.seikei.ac.jp (M. S.)ABSTRACTAlthough Li-air batteries (LAB) have a high theoretical energy density (3500Whkg−1), further developments are required to overcome theirpractical limitations. Regarding the Li-metal negative electrode (NE), we have previously reported on the reversibility of the Li dissolution/deposition reaction by using Li|Li symmetric cells with a tetraglyme (G4)-based electrolytic solution. Particularly, in the 1.0M (= mol L−1)LiNO3/G4 electrolyte under an O2 atmosphere, a Li2O protective layer is efficiently formed on the Li-metal electrode at a current density of0.40mA cm−2, and Li dendrite formation is suppressed. In the present study, we expanded the test conditions (current densities up to2.0mA cm−2 and temperatures of 10 to 50 °C) to clarify the dissolution/deposition behavior of the Li-metal NE. The effects of twoelectrolyte solutions, namely LiTFSI/G4 and LiNO3/G4, on the Li-metal NE were evaluated based on cyclical testing using Li|Li symmetriccells under an O2 atmosphere. The NEs were also examined by scanning electron microscopy and X-ray photoelectron spectroscopy. Theresults indicated that not only LiNO3 salt but also the supply of Li and nitrate ions at the Li electrode surface are critical factors in LABperformance.© The Author(s) 2023. Published by ECSJ. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 License (CC BY,http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse of the work in any medium provided the original work is properly cited. [DOI:10.5796/electrochemistry.23-00142].Keywords : Li Metal Anode, Li Dissolution/deposition, LiNO3 Electrolyte Solution, Li-air Battery (LAB)1. IntroductionLithium-air batteries (LAB) have attracted much attentionbecause of their theoretical energy density, reaching as high as3500Whkg¹1.1 However, all the components of LAB cells,specifically the air electrodes (positive electrode, PE), lithiumelectrodes (negative electrode, NE), and electrolyte solutions, havesome drawbacks and must be developed further. In addition to itsrole as the active material in PEs, oxygen dissolves into theelectrolyte solution and diffuses to the NE to react with the surfaceof Li-metal in LABs employing nonaqueous electrolyte solutions,forming a passivation film composed of Li2O.2–4 We previouslystudied the effect of oxygen on the lithium NE (Li NE) in threetetraglyme (G4)-based electrolyte solutions with different electrolytesalts: lithium trifluoromethanesulfonate (LiSO3CF3; LiOTf ), lithiumbis(trifluoromethylsulfonyl) imide (LiN(SO2CF3)2; LiTFSI), andlithium bis(fluorosulfonyl) imide (LiN(SO2F)2; LiFSI).5 The Li«Lisymmetric cells with different electrolyte solutions performeddifferently under an argon or oxygen atmosphere. Notably, the Lideposition/dissolution reaction was improved by the introduction ofO2 for all electrolytes.Lithium nitrate (LiNO3) is considered a useful Li salt forstabilizing the interface between organic electrolyte solutions andLi-metal electrodes,6–8 and it has been applied to stabilize the Li NEsurface in LABs.9–12 LiNO3 not only reacts and forms an oxide filmon the surface of Li-metal NE but also works as a redox mediator.The bifunctional effect of O2 and LiNO3 on the surface of the Limetal has recently been reported.13,14 A 1.0M (= mol L¹1) LiNO3/N,N-dimethylacetamide electrolyte solution was used, and theresults suggested that the coexistence of O2 and LiNO3 ensuredstable cyclability while reducing the change in impedance. Recently,we reported that LiNO3 in a G4-based electrolyte solution showedhigh cycle performance under an O2 atmosphere, compared withLiOTf and LiTFSI.15 These differences are attributed to the chemicalinteractions between the Li-metal electrode and the anions, namelyOTf¹, TFSI¹, and NO3¹, in the G4 solvent and oxygen atmosphere,producing dissolved oxygen and its active intermediate products.However, it was not determined how LiNO3 in a G4-basedelectrolyte solution affected the Li NE in practical applications,assuming rapid charge/discharge and low or high temperatures. Thefactors that affect the supply of Li+ to the Li electrode surface,such as the Li+ ion concentration, solvent viscosity, degree ofdissociation, and ionic conductivity in the electrolyte, are importantto investigate for efficient Li dissolution and deposition.Herein, we applied wider test conditions (current densities from0.20 to 2.0mAcm¹2 and temperatures of 10 to 50 °C) to clarify thedissolution/deposition behavior of the Li electrode. We comparedthe effects of two lithium salts, LiTFSI and LiNO3, on Li NEperformance in G4-based electrolyte solutions by performingdischarge/charge cycle tests using Li«Li symmetric cells undercontrolled current densities and temperatures. We also conductedscanning electron microscopy energy-dispersive X-ray spectroscopy(SEM-EDS) and X-ray photoelectron spectroscopy (XPS) analysesof the deposits on the NE.2. Materials and Methods2.1 Preparation of electrolytesLiNO3 (Sigma-Aldrich, ²99.99%) were dried in a vacuum ovenat 110 °C overnight before dissolution. LiTFSI (Kishida Chemical,²99.9%) and dried LiNO3 were handled in an Ar-filled glove box§ECSJ Active MemberF. Ozawa orcid.org/0009-0008-7214-0885ElectrochemistryThe Electrochemical Society of JapanAdvance Publication by J-STAGEhttps://doi.org/10.5796/electrochemistry.23-00142https://doi.org/10.50892/data.electrochemistry.25448005Received: December 12, 2023Accepted: March 16, 2024Published online: March 22, 2024Issued: to be determined.1UNCORRECTEDPROOF 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566https://orcid.org/0009-0008-7214-0885http://creativecommons.org/licenses/by/4.0/https://doi.org/10.5796/electrochemistry.23-00142http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/https://orcid.org/0009-0008-7214-0885https://orcid.org/0009-0008-7214-0885https://orcid.org/0009-0008-7214-0885https://doi.org/10.5796/electrochemistry.23-00142https://doi.org/10.5796/electrochemistry.23-00142https://doi.org/10.50892/data.electrochemistry.25448005https://doi.org/10.50892/data.electrochemistry.25448005(Miwa, MDB-1BK-NT1). Each Li salts were dissolved in tetra-ethylene glycol dimethyl ether (G4, Nippon Nyukazai, <10 ppmH2O) at the concentration of 1.0M. The H2O contents of theelectrolyte solutions were quantified to be under 100 ppm usinga Karl Fisher titration apparatus (CA-31, Mitsubishi ChemicalAnalytech). The ionic conductivity, viscosity, and density of eachelectrolyte solution at 25–60 °C were ascertained using a pH/ionmeter (SevenExcellence S500, Mettler Toledo) and a rolling-ballviscometer (Lovis2000ME, Anton Paar), respectively.2.2 Electrochemical measurementsLi«Li symmetric cells were assembled using Li metal foil(thickness: 0.5mm, Honjo Metal) as both electrodes, a separator(Celgard 2400, Celgard), and each electrolyte in an argon filledglove box. The Li«Li symmetric cells were used for Li dissolution/deposition cycle tests under an O2 atmosphere at 10, 30, and 50 °C.The O2 atmosphere was prepared by purging Ar out of theassembled cells using high-purity O2 gas (G1 grade, >99.99995%)injected at 200mLmin¹1 for 15 s. The applied current density was0.20 to 2.0mAcm¹2 at a maximum dissolution/deposition capacityof 0.50mAh cm¹2 in the voltage range between ¹2.0 and 2.0V for15 cycles.2.3 Electrode analysisAfter 15 cycles, the Li-metal electrodes were picked out the Li«Lisymmetric cells and rinsed with fresh G4 solvent to remove the Lisalts in an Ar-filled glove box. After drying under vacuum, theelectrodes were placed in a transfer vessel to avoid exposure to air.The morphology and composition of the deposited films wereevaluated by SEM-EDS (JSM-7800F, JEOL). Cross-section samplesof Li-metal electrodes were treated using an Ar ion beam in a JEOLcross polisher IB-09020CP at an accelerating voltage of 8 kV for 3 hat 0 °C.In addition, the chemical bonding state and elemental composi-tion of the Li electrode surface after Li dissolution/deposition wereexamined by XPS (ULVAC PHI, VersaProbe II). After 15 cycles, theLi-metal electrode was rinsed with G4, dried sufficiently, placed in atransfer vessel, and transported to the main chamber for analysis,without exposure to the atmosphere. In the XPS measurements, theacceleration voltage and emission current of a monochromatic AlKAX-ray (1484.6 eV) were set at 15 kV and 3mA, respectively. Peakfits and baseline correction of all spectra were performed usingOrigin Pro Peak Analyzer. The binding energy scale was calibratedusing the hydrocarbon C 1s peak at 285.0 eV.3. Results and Discussion3.1 Cycle tests at 0.50mAhcm−2Figure 1 shows dissolution/deposition cycle test results for thecells using 1.0M LiTFSI/G4 electrolyte solutions. The graphsexhibited relatively low overvoltages, which suggests a stablesurface state of the Li electrode. At a lower current density(0.20mAcm¹2), the voltage decreased up to the 15th cycle, whichwas likely caused by the sufficient stability of the surface film, andthe continuous gradual decrease of the voltage thereafter is ascribedto the increased effective surface area caused by dendrite deposition(Fig. 1A). In contrast, at a higher current density (0.40 to2.0mAcm¹2), the voltage significantly changed from the first cycle,and unstable polarization curves were obtained (Figs. 1B–1E).These results suggested that the Li2O layer protecting the Li-metalelectrode protection from O2 could not withstand the current densityand the exposed Li metal reacted with the solvent and anion,resulting in the growth of Li dendrites. Moreover, the overvoltagedecreased when the operating temperature was raised to 10, 30, and50 °C, suggesting that the diffusion rate of Li+ ions in the electrolytewas improved and the charge transfer resistance during the reactionwas reduced (Figs. S1A–S1C). The correlation was also suggestedfrom the physical properties of the electrolyte, as the Li+ diffusioncoefficient and Li ion transference number increased as thetemperature rose (Table S1). At 10 and 30 °C, the flatness of thepolarization curve collapsed from the first cycle (Figs. S1A–S1B),implying that the solvent decomposed. In addition, the overvoltagegradually decreased, suggesting the formation of Li dendrites duringthe dissolution and deposition cycle, which increases the specificsurface area of the Li electrode. The flat part of the polarizationcurve was depressed even at 50 °C, indicating that Li dendritesformed in the LiTFSI/G4 electrolyte under all operating temper-atures (Fig. S1C). The potential fluctuation was suppressed byraising the operating temperature, suggesting that electrolytedecomposition was suppressed.Figure 2 shows that cells with the 1.0M LiNO3/G4 electrolytesolution exhibited larger overpotentials after the first cycle, mainlybecause this electrolyte has the lowest ionic conductivity,0.22mS cm¹1. The overvoltage gradually increased with additionalFigure 1. Polarization curves of the Li«Li symmetric cells tested at 0.40mAh cm¹2 and 30 °C using the LiTFSI electrolyte and differentcurrent densities: (A) 0.20, (B) 0.40, (C) 0.60, (D) 1.0, and (E) 2.0mAcm¹2.Electrochemistry, (in press) 1–72UNCORRECTEDPROOF 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566cycles, suggesting an accumulation of Li deposited on the surface(Fig. 2A). The flat polarization curves indicate that no dendriticgrowth occurred.16 However, the overvoltage also increased for 0.4and 0.6mAcm¹2 (Figs. 2B–2C). Above 1.0mAcm¹2, the flatnessof the polarization curve was lost, suggesting that electrolytedecomposition occurred (Figs. 2D–2E). Moreover, the overvoltagedecreased for increasing operating temperatures, similar to LiTFSI/G4, suggesting that the diffusion rate of Li+ ions in the electrolytewas improved and the charge transfer resistance during the reactionwas reduced (Figs. S2A–S2C). In fact, although LiNO3/G4 electro-lyte had a lower ionic conductivity than LiTFSI/G4, but had thediffusion coefficient of Li+ equal to or higher than that of LiTFSI/G4 (Table S1). This is owing to the solvation structure of Li+ in theelectrolyte solutions. In the case of LiTFSI salt electrolyte, relativelylarge amount of Li+ was dissociated form TFSI¹ anion comparedwith LiNO3 salt one and the Li+ solvated by G4 solvent diffused.Therefore, the diffusion radius was relatively large. On the otherhand, LiNO3 salt was poorly dissociated in the G4-based electrolyteand the contact ion pairs of Li+-NO3¹ diffused with a rather smalldiffusion radius. Conversely, at the low temperature of 10 °C, theflatness of the polarization curve was poor, and the overvoltagegradually decreased (Fig. S2A). This also suggests that Li dendritesformed, increasing the specific surface area of the Li electrodeduring dissolution and deposition. Nevertheless, the overvoltage wasnot reduced to the extent observed for the LiTFSI electrolyte,indicating that solvent decomposition was not observed. Thus, theLiNO3/G4 electrolyte affects solvent decomposition even at 10 °C.At 30 and 50 °C, the flatness of the polarization curve wasmaintained, and the change in the overvoltage with additional cycleswas small (Figs. S2B–S2C). Therefore, the diffusion rate of Li ionsincreased at higher operating temperatures, and Li ions diffused tothe electrode surface more quickly, suppressing the formation of Lidendrites. Since the diffusion coefficient of solvent and Li+ increaseas the temperature increases, it is assumed that the diffusion ofnitrate anions also increased. Even if the Li2O protective layer waspartially broken and the Li-metal electrode was exposed during Lidissolution and deposition, the electrode immediately reacted withnitrate anions to form a solid-electrolyte interphase (SEI) film. TheSEI formation can reduce the uneven spots and suppress Li dendriteformation.The differences in performance between cells using LiTFSI andLiNO3 electrolytes demonstrated the effect of the Li2O protectivelayer formed on the Li NE surface by the reaction between the Limetal and NO3¹.3.2 Surface analysis of the electrode after 15 cycles at0.50mAhcm−2Figure 3 shows SEM images of the Li NE surfaces after 15cycles in the 1.0M LiTFSI/G4 and 1.0M LiNO3/G4 electrolytesolutions. A larger dendrite surface area was observed for LiTFSI/G4 (Figs. 3A–3C) than for LiNO3/G4 (Figs. 3D–3E). At0.20mAcm¹2 in LiTFSI/G4, flat deposits covered the wholesurface, and Li dendrites were observed (Fig. 3A). A similarmorphology was observed for 0.60 to 2.0mAcm¹2, but with areasof exposed Li metal, indicating that the deposits were fragile(Figs. 3B–3C). As the current density increased, the decompositionof the electrolyte progressed, resulting in the deposition of fine Lidendrites and decomposition products.17 These results are alsoconsistent with the Li dissolution/deposition results shown inFig. 1, which suggests the continuous deposition of dendrites anddecomposition products. In addition, the Li dendrites becomethicker at higher operating temperatures (Figs. S1D–S1F). Becausethe deposits easily peeled off, the anion and solvent were likelydecomposed. A different morphology was observed for Li NEsin LiNO3/G4 (Figs. 3D–3F), revealing dense deposition withoutdendrite formation, even above 0.40mAcm¹2. As the currentdensity increased, numerous granular depositions were observed,which consist of electrolyte decomposition products, such asLi2CO3, and agglomerated Li2O particles that peeled off thesurface.18 Furthermore, the granular deposits were present on theelectrode surface at 10 °C, suggesting that the diffusion rate of Liions was slower than the rate of Li dissolution and deposition, andthe electrolyte decomposed (Fig. S2D). In contrast, at 30 and 50 °C,Li dendrites and granular deposits did not form on the electrodesurface (Figs. S2E and S2F). In LiNO3, no easy-to-exfoliatedeposits or no Li dendrites were observed. The LiNO3/G4electrolyte effectively suppressed Li dendrite growth even at 10 °Cand suppressed electrolyte decomposition to some extent.Figure 4 shows cross-sectional SEM images of Li NEs after 15cycles at 0.40mAcm¹2 in the 1.0M LiTFSI/G4 and 1.0M LiNO3/G4 electrolyte solutions. Approximately 30µm-thick sediments andlocally 70µm-thick pit-shaped sediments were observed forLiTFSI/G4 (Figs. 4A–4B), suggesting that Li dissolution/deposi-tion progressed locally owing to the formation of a non-uniform SEIFigure 2. Polarization curves of the Li«Li symmetric cells tested at 0.40mAh cm¹2 and 30 °C using the LiNO3 electrolyte and differentcurrent densities: (A) 0.20, (B) 0.40, (C) 0.60, (D) 1.0, and (E) 2.0mAcm¹2.Electrochemistry, (in press) 1–73UNCORRECTEDPROOF 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566layer. In contrast, thin sediment was observed for LiNO3/G4(Figs. 4C–4D), suggesting that Li dissolution/deposition progresseduniformly owing to the formation of a uniform SEI layer.19,20Figure 5 shows the SEM-EDS mapping of the O KA and C KAsignals. Tables S2 and S3 summarize the results for the Li electrodesurface after 15 cycles at different current densities in LiTFSI/G4and LiNO3/G4, respectively. In LiTFSI/G4, the sediment containssubstantial amounts of F and S, and the elemental ratio F : S isdistributed at 3 : 1. At the interface between the Li metal and thesediment, C and O accumulate on the Li-metal side, whereas it doesnot appear on the Li-metal surface, suggesting that the solventdecompose on the Li-metal surface (Figs. 5A–5C). However, the Lidendrite grows away from the Li-metal surface. These resultssuggested that Li dissolution and deposition at 0.40mAcm¹2progressed without a uniform SEI layer and the Li electrode surfacewas not completely protected. On the other hands, the oxygendistribution was uniform in LiNO3/G4, suggesting that the Li2Olayer is spread uniformly (Figs. 5D–5F). In addition, the mapping ofC consistent with O, suggesting that Li2CO3 was also deposited. Tocompare of the LiTFSI, the ratio of oxygen was large, indicating thatthe ratio of Li2CO3 and solvent decomposition products was smaller.Figure 6 shows the XPS spectra of Li NE surfaces after 15 cyclesin 1.0M LiTFSI/G4 at different current density. Analysis of theC 1s peak revealed solvent decomposition peaks such as C-O(286.9 eV) and C=O (287.7 eV). A peak for Li2CO3 (289.7 eV) wasalso found, which was consistent with the polarization curve andEDS mapping, showing that electrolyte decomposition occurred.This was consistent with the observation of peaks such as Li(54.5 eV) and Li2CO3 (55.4 eV) in Li1s spectra. Furthermore, wefound that the Li2O layer formed by O2 remained even at0.40mAcm¹2. However, the suppression of Li dendrite growthcould not be confirmed from the SEM image. From these results, Lidissolution and deposition at a high rate in a TFSI electrolytesolution does not have the effect of suppressing Li dendrite growthusing the Li2O layer formed by O2 and the decomposition of theelectrolyte solution. The samples operated at 0.40 and 0.60mAcm¹2both exhibited C 1s peaks, which appeared to be C-F (291.6 eV),indicating the decomposition of electrolyte anions.Figure S3 shows the XPS spectra of Li NE surfaces after 15cycles at 10, 30, and 50 °C. The C 1s spectra at each temperaturerevealed signals of solvent decomposition, such as the peaks forC-O and C=O. The signal of Li2CO3 decreased as the temperatureincreased from 10 °C, showing that higher operating temperaturesimproved Li-ion diffusion and suppressed electrolyte decomposi-tion. In addition, the suppression of Li dendrites could not beconfirmed from SEM images, suggesting that the LiTFSI/G4electrolyte did not suppress the growth of Li dendrites using theLi2O layer formed by O2.Figure 7 shows the XPS spectra of Li NE surfaces after 15 cyclesin 1.0M LiNO3/G4 at different current density. Considering that thepeaks derived from Li2CO3 are strong in C 1s and Li 1s, electrolytedecomposition occurred, and many decomposition products weredeposited on the Li NE surface. However, the flatness of thepolarization curve suggests that electrolyte decomposition was notserious and the formation of Li2CO3 suppressed excessive electro-lyte decomposition. In addition, the peaks derived from C=O andC-O were smaller than the peaks derived from Li2CO3, implying thatsolvent decomposition was suppressed. For the Li 1s spectra,no peak derived from Li2O (55.7 eV) was observed at 0.2–0.6mAcm¹2, indicating that Li2O was not exposed on the topsurface of the Li NE. In contrast, a peak of Li2O appeared at1.0mAcm¹2, suggesting that Li2O was partially exposed on the topsurface of the Li NE. These results indicate that in the 1.0M LiNO3/G4 electrolyte, Li2CO3 covered the Li NE surface during Lidissolution and deposition, even at a high rate above 0.60mAcm¹2.The Li NE in the 1.0M LiNO3/G4 electrolyte exhibited enhancedreversibility owing to the preferential reduction of LiNO3 and theformation of an inorganic-rich SEI.6,21,22 This observation can beattributed to the anions localized on the Li-metal surface, resultingfrom the reaction: 2Li + NO3¹ ¼ Li2O + NO2¹ on the Li-metalFigure 3. SEM images of Li NE surfaces after 15 cycles at 0.20, 0.60, and 1.0mAcm¹2 in different electrolyte solutions: (A–C) LiTFSI/G4and (D–F) LiNO3/G4.Figure 4. Cross-sectional SEM images of Li NEs after 15 cyclesat 0.40mAcm¹2 in different electrolyte solutions: (A–B) LiTFSI/G4 and (C–D) LiNO3/G4.Electrochemistry, (in press) 1–74UNCORRECTEDPROOF 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566surface. However, NO2¹ produced was smoothly oxidized back toNO3¹ by dissolved O2, and it has almost no effect on the NO3¹concentration localized on the Li NE surface.23 These phenomenasuppressed excessive electrolyte decomposition and Li dendritegrowth.Figure S4 shows the XPS spectra of Li NE surfaces after 15cycles at 10, 30, and 50 °C. Analysis of the C 1s spectra at eachtemperature revealed signals of solvent decomposition, such as thepeaks for C-O and C=O. In addition, a peak was observed forLi2CO3, consistent with the polarization curve, suggesting thatelectrolyte decomposition occurred. Moreover, Li2O was notdetected in the Li 1s spectra at 10 and 30 °C, indicating that Li2Owas not exposed on the top surface of the Li NE after 15 cycles.Figures S4D–S4F confirms that LiNO3/G4 electrolyte suppressedLi dendrite growth and electrolyte decomposition using the Li2Olayer formed by O2. These results indicate that electrolytedecomposition products, such as Li2CO3, were deposited on theLi2O layer. The signal of Li2O also appeared at 50 °C, suggestingthat higher operating temperatures can suppress electrolyte decom-position.3.3 Models of surface reactions and morphology dependingon the supply of ionsBased on the results, we modeled the surface reactions andmorphology on the Li NE, as shown in Fig. 8. Using 1.0M LiNO3/G4, the flatness of the polarization curve was maintained foroperation below 0.60mAcm¹2. The Li-ion diffusion was sufficient,and stable Li dissolution and deposition progressed. The oxygenderived from NO3¹ in the electrolyte was highly active andefficiently produced inorganic SEI components such as Li2O andLi2CO3. In contrast, the flatness of the polarization curve was lostabove 1.0mAcm¹2 because the ion reaction rate was insufficient,resulting in electrolyte decomposition. Thus, the diffusion of ions inthe electrolyte affected the Li dissolution and deposition behavior.Regarding the Li dissolution/deposition behavior at hightemperatures, the polarization curve shows that overvoltage wasreduced, and the electrolyte decomposition was suppressed. This isdue to improved Li-ion diffusion in the electrolyte, which allows Liions to quickly reach the Li electrode surface before the electrons areused in other reactions. Furthermore, in the LiNO3/G4 electrolyte,the diffusion of nitrate anions improved in addition to the diffusionFigure 5. SEM-EDS compositional mapping of Li NE surfaces after 15 cycles at 0.40mAcm¹2 in different electrolyte solutions:(A–C) LiTFSI/G4 and (D–F) LiNO3/G4.5055605055605055600.40 mA cm˗2 0.60 mA cm˗20.20 mA cm˗2C1sLi1sA B CD E FIntensity / a .u.Intensity / a.u.Intensity / a.u.Intensity / a.u.Intensity / a.u.Intensity / a.u.Binding energy / eV Binding energy / eVBinding energy / eVBinding energy / eV Binding energy / eVBinding energy / eV280284288292Li2CO3C-OC-CC=OC-F280284288292Li2CO3C-OC-CC=OC-HLi2CO3LiLi2CO3LiLi2CO3LiLi2CO3C=O C-OC-C280284288292Figure 6. XPS spectra of the Li NE surface after 15 cycles in 1.0M LiTFSI/G4. (A–C: C 1s and D–F: Li 1s)Electrochemistry, (in press) 1–75UNCORRECTEDPROOF 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566of Li ions, which allowed them to reach the Li electrode surfacemore smoothly, generating SEI films comprising Li2O and Li2CO3.This reduced the exposure of the Li metal and promoted dissolution/deposition while suppressing side reactions. In addition, operating athigh temperatures improved Li diffusion and suppressed the sidereactions even at high current densities that require more Li.However, the overvoltage increased when the diffusion of Li ionswas suppressed at low temperatures. In the LiNO3/G4 electrolyte,although the overvoltage increased compared with the test at 30 °C,the flatness of the polarization curve was maintained, indicating thatthe Li2O protective layer suppressed Li dendrite growth. In contrast,solvent decomposition was observed at the beginning of the cycletests in the LiTFSI/G4 electrolyte, and the electrode surface wascovered with solvent decomposition products. The Li dissolution/deposition rate is faster than the formation rate of the Li2Oprotective layer, resulting in solvent decomposition and non-uniformLi dissolution and deposition.4. ConclusionsTwo electrolyte solutions, LiTFSI/G4 and LiNO3/G4, werecompared regarding their effect on the Li NE during cycling Li«Lisymmetric cells under an O2 atmosphere. In addition to the cyclingperformance at different current densities and temperatures, themorphology and composition of the Li NE surface were examinedby SEM and XPS. Using the LiTFSI/G4, the polarization curveswere unstable and the NE surface was accompanied by Li dendritegrowth. Whereas using the LiNO3/G4, the polarization curves werestable and the NE surface did not exhibit Li dendrite formation.These differences were due to the chemical interactions between theLi-metal electrode and the anions, TFSI¹ and NO3¹, in the G4solvent and O2. Particularly, the O2 derived from NO3¹ in theLiNO3/G4 electrolyte was highly active and produced inorganic SEIcomponents such as Li2O and Li2CO3. Cycle tests under controlledcurrent densities and temperatures confirmed the stability of the Li-2802842882925055605055605055601.0 mA cm˗2Li1s0.60 mA cm˗20.20 mA cm˗2A B CD E FLiLi2CO3Intensity / a.u.Intensity / a.u.Intensity / a.u.Intensity / a.u.Intensity / a.u.Intensity / a.u.Binding energy / eV Binding energy / eVBinding energy / eVBinding energy / eV Binding energy / eVBinding energy / eV280284288292Li2CO3C-OC-CC1s280284288292Li2CO3C-OC-CC=OLi2CO3LiLi2CO3LiLi2OLi2CO3C-OC-CFigure 7. XPS spectra of the Li NE surface after 15 cycles in 1.0M LiNO3/G4. (A–C: C 1s and D–F: Li 1s)NormalSlow Rapid1.0 M LiNO3/ G4DendriteG4, TFSI-A li le decomp.DendriteG4, TFSI-Decomp.DendriteG4, TFSI-No decomp.1.0 M LiTFSI/ G4Li metal Li2O layer Decomposi on productsLi2CO3 layerG4, NO3-A li le decomp.No dendriteG4, NO3-A li le decomp.No dendriteG4, NO3-No decomp.No dendriteSupply of ionsFigure 8. Models of the surface reactions and morphologies that occur depending on the supply of ions.Electrochemistry, (in press) 1–76UNCORRECTEDPROOF 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566metal electrode in the LiNO3/G4 electrolyte solution. Our findingsdemonstrate that not only LiNO3 salt but also the supply of Liand nitrate ions at the NE surface are critical factors in LABperformance. Additionally, it is necessary to quickly form protectivefilms, such as Li2O or Li2CO3, with appropriate thicknesses.AcknowledgmentThis work was partly supported by the Japan Science andTechnology Project for the Next Generation Batteries Area inAdvanced Low Carbon Technology Research and Development(ALCA-SPRING JPMJAL1301) and the National Institute forMaterials Science (NIMS) Joint Research Hub Program. TheSEM-EDS and XPS measurements were performed at the NIMSBattery Research Platform.CRediT Authorship Contribution StatementFumisato Ozawa: Conceptualization (Equal), Project administration (Supporting),Writing – original draft (Lead)Kazuki Koyama: Data curation (Lead), Formal analysis (Lead)Daiki Iwasaki: Data curation (Supporting), Formal analysis (Supporting)Shota Azuma: Formal analysis (Supporting), Methodology (Supporting)Akihiro Nomura: Writing – review & editing (Supporting)Morihiro Saito: Conceptualization (Equal), Project administration (Lead), Writing –review & editing (Lead)Data Availability StatementThe data that support the findings of this study are openly available under the termsof the designated Creative Commons License in J-STAGE Data listed in D1 ofReferences.Conflict of InterestThe authors declare no conflict of interest in the manuscript.FundingJapan Science and Technology Agency: ALCA-SPRING JPMJAL1301National Institute for Materials Science: NIMS Joint Research Hub ProgramReferencesD1. F. Ozawa, K. Koyama, D. Iwasaki, S. Azuma, A. Nomura, and M. Saito, J-STAGEData, https://doi.org/10.50892/data.electrochemistry.25448005, (2024).1. P. G. Bruce, S. A. Freunberger, L. J. Hardwick, and J.-M. Tarascon, Nat. Mater.,11, 19 (2012).2. D. Aurbach, Y. Gofer, and J. Langzam, J. Electrochem. Soc., 136, 3198 (1989).3. D. Aurbach, M. Daroux, P. Faguy, and E. Yeager, J. Electroanal. Chem. InterfacialElectrochem., 297, 225 (1991).4. D. Aurbach and O. Chusid, J. Power Sources, 68, 463 (1997).5. M. Saito, T. Fujinami, S. Yamada, T. Ishikawa, H. Otsuka, K. Ito, and Y. Kubo,J. Electrochem. Soc., 164, A2872 (2017).6. C. Yan, Y.-X. Yao, X. Chen, X.-B. Cheng, X.-Q. Zhang, J.-Q. Huang, and Q.Zhang, Angew. Chem. Int. Ed., 57, 14055 (2018).7. H. Yang, X. Chen, N. Yao, N. Piao, Z. Wang, K. He, H. M. Cheng, and F. Li, ACSEnergy Lett., 6, 1413 (2021).8. Z. L. Brown, S. Heiskanen, and B. L. Lucht, J. Electrochem. Soc., 166, A2523(2019).9. J. Uddin, V. S. Bryantsev, V. Giordani, W. Walker, G. V. Chase, and D. Addison,J. Phys. Chem. Lett., 4, 3760 (2013).10. V. Giordani, W. Walker, V. S. Bryantsev, J. Uddin, G. V. Chase, and D. Addison,J. Electrochem. Soc., 160, A1544 (2013).11. W. Walker, V. Giordani, J. Uddin, V. S. Bryantsev, G. V. Chase, and D. Addison,J. Am. Chem. Soc., 135, 2076 (2013).12. Y. Hayashi, S. Yamada, T. Ishikawa, Y. Takamuki, M. Sohmiya, H. Otsuka, K. Ito,Y. Kubo, and M. Saito, J. Electrochem. Soc., 167, 020542 (2020).13. D. Sharon, D. Hirsberg, M. Afri, A. Garsuch, A. A. Frimer, and D. Aubach,J. Phys. Chem. C, 118, 15207 (2014).14. D. Sharon, D. Hirsberg, M. Afri, F. Chesneau, R. Lavi, A. A. Frimer, Y. K. Sun,and D. Aurbach, ACS Appl. Mater. Interfaces, 7, 16590 (2015).15. M. Saito, T. Fujinami, M. Sohmiya, Y. Hayashi, K. Koyama, H. Otsuka, K. Ito, Y.Kubo, and T. Horiba, J. Electrochem. Soc., 168, 010520 (2021).16. L. Carbone, D. D. Lecce, M. Gobet, S. Munoz, M. Devany, S. Greenbaum, andJ. Hassoun, ACS Appl. Mater. Interfaces, 9, 17085 (2017).17. M. Ue, H. Asahina, S. Matsuda, and K. Uosaki, RSC Adv., 10, 42971 (2020).18. K. Edström, M. Herstedt, and D. P. Abraham, J. Power Sources, 153, 380 (2006).19. F. Shi, A. Pei, D. T. Boyle, J. Xie, X. Yu, X. Zhang, and Y. Cui, Proc. Natl. Acad.Sci. U.S.A., 115, 8529 (2018).20. Y. Zhang, Y. Zhong, S. Liang, B. Wang, X. Chen, and H. Wang, ACS Mater. Lett.,1, 254 (2019).21. L. Chen, A. Lv, F. Guo, M. Wang, and S. Jiao, ACS Sustainable Chem. Eng., 8,706 (2020).22. Y. Liang, W. Wu, D. Li, H. Wu, C. Gao, Z. Chen, L. Ci, and J. Zhang, Adv. EnergyMater., 12, 2202493 (2022).23. X. Xin, K. Ito, A. Dutta, and Y. Kubo, Angew. Chem., Int. Ed., 57, 13206 (2018).Electrochemistry, (in press) 1–77UNCORRECTEDPROOF 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566https://doi.org/10.50892/data.electrochemistry.25448005https://doi.org/10.1038/nmat3191https://doi.org/10.1038/nmat3191https://doi.org/10.1149/1.2096425https://doi.org/10.1016/0022-0728(91)85370-5https://doi.org/10.1016/0022-0728(91)85370-5https://doi.org/10.1016/S0378-7753(97)02622-0https://doi.org/10.1149/2.0151713jeshttps://doi.org/10.1002/anie.201807034https://doi.org/10.1021/acsenergylett.1c00149https://doi.org/10.1021/acsenergylett.1c00149https://doi.org/10.1149/2.0991912jeshttps://doi.org/10.1149/2.0991912jeshttps://doi.org/10.1021/jz402025nhttps://doi.org/10.1149/2.097309jeshttps://doi.org/10.1021/ja311518shttps://doi.org/10.1149/1945-7111/ab6975https://doi.org/10.1021/jp506230vhttps://doi.org/10.1021/acsami.5b04145https://doi.org/10.1149/1945-7111/abd60dhttps://doi.org/10.1021/acsami.7b03235https://doi.org/10.1039/D0RA07924Chttps://doi.org/10.1016/j.jpowsour.2005.05.062https://doi.org/10.1073/pnas.1806878115https://doi.org/10.1073/pnas.1806878115https://doi.org/10.1021/acsmaterialslett.9b00167https://doi.org/10.1021/acsmaterialslett.9b00167https://doi.org/10.1021/acssuschemeng.9b06624https://doi.org/10.1021/acssuschemeng.9b06624https://doi.org/10.1002/aenm.202202493https://doi.org/10.1002/aenm.202202493https://doi.org/10.1002/anie.201808154