Role of Dissolved Gas in Ionic Liquid Electrolytes for SecondaryLithium Metal BatteriesJohanna K. Stark,† Yi Ding,‡ and Paul A. Kohl*,†

†Georgia Institute of Technology, School of Chemical and Biomolecular Engineering, Atlanta, Georgia 30332-0100, United States‡U.S. Army RDECOM-TARDEC, AMSRD-TAR-R, MS 159, 6501 East 11th Street, Warran, Michigan 48397-5000, United States

ABSTRACT: Lithium metal is a potential anode in lithium batteries;however, high cycling efficiency has not been achieved. The effect ofdissolved gas on the reversibility of a Li/Li+ electrode in an ionic liquidelectrolyte was investigated. The ionic liquid electrolyte was saturated withargon, nitrogen, oxygen, air, or carbon dioxide, and the Coulombic efficiencyfor the reduction and reoxidation of lithium metal was measured. Secondaryion mass spectroscopy (SIMS) was used to determine the thickness andcomposition of the solid electrolyte interphase (SEI) layer formed on thelithium metal. The Coulombic cycling efficiency was highest with oxygen,nitrogen, and air. The Coulombic efficiency was significantly lower withdissolved carbon dioxide. SIMS revealed that the SEI layer was made up ofprimarily LiF regardless of the gas used. The thickness of the SEI layer wasthe main determinant of cycling efficiency. A thinner SEI, observed withnitrogen and oxygen, lead to the highest Coulombic efficiency, whereas carbon dioxide produced a thick SEI that appeared toinhibit lithium deposition.

■ INTRODUCTION

The ubiquity of mobile electronics has spurred interest in high-energy density secondary batteries. Cell phones, laptops, andmusic players gain their strength from being portable devices,but long-lasting batteries with good cycleability must bedeveloped to support the increasing need for energy.Currently, these devices use lithium-ion batteries composed

of a graphite anode and metal oxide cathode. Lithium, being thethird lightest element, is already synonymous with high-energydensity batteries. Its low molecular weight makes it the lightestactive material for batteries. The structures and materials thatsupport shuttling lithium ions (battery cycling) such as theseparator, electrolyte, and cathode and anode superstructurescontribute most of the weight of the battery. To eliminate thisbulk, researchers have looked at higher energy densityelectrodes such as silicon and lithium metal anodes. Lithiummetal anodes present an interesting opportunity because theycompletely eliminate the anode superstructure material. Theactive material, extremely light lithium metal, is used directly asthe anode. The lithium anode could also be potentially cycledat higher rates because there would be no diffusion limitationwithin the anode; the active material is perpetually on thesurface. This lithium metal anode represents the upper boundfor lithium anode energy density.A lithium metal anode faces two principal challenges: the

formation of a stable solid electrolyte interface (SEI) and theability to electrodeposit lithium nondendritically. When lithiumis electrodeposited, as during battery charging, it tends to formneedle-like structures, also called dendrites. These protrusionsare a concern for batteries because they can short-circuit the

two electrodes and cause the battery to fail. Work has focusedon physically containing the anode with a hard polymer orceramic separator.1 This can prevent dendrites from shortingthe cell but does not solve the problem on a fundamental level.Additives have been shown to alter plating in lithium, includingthe appearance of dendrites, but a smooth deposit has not beenachieved.Because of lithium’s reactivity, the formation of a stable SEI

is imperative to prevent the irreversible reaction betweenlithium and the electrolyte. On a graphite anode, the SEI issupported on the graphite surface, whereas the lithium ionsintercalate into the bulk. The SEI on graphite is usually formedwith the help of additives within the electrolyte. Cycliccarbonates, such as ethylene carbonate and vinylene carbo-nate,2−7 are thought to undergo reduction and ring-opening onthe surface of the electrode.8 The SEI on the surface of thelithium metal must form directly on the active material, which ismore difficult than a host-type anode due to the dissolution ofmetal surface upon oxidation during cycling. Vinylenecarbonate and other additives have been successfully used asSEI forming agents but have not completely solved the lithiumanode stability problem.9−11

In this study, dissolved gases were investigated for their rolein forming the SEI on the metal surface. Lithium batteries canbe assembled in the absence of ambient water vapor (i.e., dryrooms), where the electrolyte is saturated with air.7,11−13 Gases

Received: January 4, 2013Revised: February 19, 2013Published: February 22, 2013

Article

pubs.acs.org/JPCC

© 2013 American Chemical Society 4980 dx.doi.org/10.1021/jp4001303 | J. Phys. Chem. C 2013, 117, 4980−4985

dissolve in the electrolyte in small amounts and can thus beused to affect the SEI in controlled ways. This effect has beenstudied for organic electrolytes such as propylene carbonate14

and ethylene carbonate/diethyl carbonate.12 Aurbach et al. alsostudied the effect of water and oxygen contamination on theSEI in dioxolane and γ-butyrolacetone.15,16 No such study hasbeen conducted for ionic liquid electrolytes, which interactdifferently with gases than their organic counterparts. Inaddition, the effect of other conditions, such as nitrogen, argon,and under vacuum are studied here.Ionic liquids are interesting because they are stable over a

wide potential range, making them good candidates for futurehigh-voltage batteries. Ionic liquids are also advantageous interms of battery safety because they have near-zero vaporpressure and are generally nonflammable. Ionic liquids used todeposit lithium metal include pyrolidinium, pyrridinium, andquaternary ammonium ionic liquids.6,17,18 Trimethylbutylam-monium bis(triflouromethanesulfonyl)imide (N1114-Tf2N)ionic liquid is used in this study because it has shown reversiblelithium deposition.19−21 Ionic liquids have also receivedattention as carbon-dioxide capture mediums. In particular,researchers have investigated imidazolium-based ionic liquidsfor CO2 capture because of the high solubility of CO2. Datacomparing CO2, N2, and O2 solubility in ionic liquid have beenreported.22−26 In battery applications, soluble gases can be usedto deliver small amounts of potentially SEI forming reactants.N2 and O2 can both react with lithium and be potentialcomponents of the SEI. The effects of dissolved gas on thelithium metal SEI and cycling efficiency are investigated in thisstudy.

■ EXPERIMENTAL METHODSA liquid electrolyte (compared with an ion-conducting solidelectrolyte) was chosen for this study because intimate contactis required between the lithium metal and the electrolyte duringthe changing surface area of lithium deposition and oxidation.Trimethylbutylammonium bis(triflouromethanesulfonyl)imide(N1114-Tf2N, 99%, Iolitec) (Figure 1) and lithium bis-

(triflouromethanesulfonyl)imide (LiTf2N, 99%, Acros) weremixed to give a 1 M Li+ electrolyte. The electrolyte was driedunder vacuum for at least 15 h before use.A beaker-cell with connections for two electrodes and a gas

port was constructed for maintaining the ionic liquid undersaturated conditions with each of the gases during theelectrochemical cycling experiments. The ionic liquid wastaken from vacuum line into an argon-filled glovebox andtransferred to the modified beaker cell for the experiment. Thegas port allowed for bubbling of a test gas before starting theelectrochemical experiments. During the electrochemicalexperiments, the head space above the ionic liquid was

maintained with the same test gas to avoid contaminationfrom the ambient air. A second cell of similar design was usedto for the experiments under vacuum. The vacuum cell hadground glass joints and vacuum-tight electrical feedthroughs.The cell was leak-tested with a helium lead detector before use.A stainless-steel type 316 working electrode and lithium

counter/reference electrode were used for the electrochemicalcycling experiments. First, excess metal (1 C/cm2) wasdeposited on the stainless steel at 0.1 mA/cm2. Then, 10% ofthe initial lithium capacity was cycled at 0.1 mA/cm2 on eachcycle until the lithium was exhausted during the oxidation partof the cycle. That is, 0.1 C/cm2 of lithium was deposited; then,0.1 C/cm2 of lithium was oxidized during each cycle. Thepotential was recorded during the constant current reductionand oxidation portions of the cycle, as shown in Figure 2. When

there was no longer adequate lithium metal for the oxidationstep, the potential rose sharply and the experiment wasterminated. The number of cycles was recorded, and the totalcharge for deposition and oxidation was used to give an averageCoulombic efficiency for the lithium deposition and strippingprocess. This cycling method avoided stripping the lithiummetal from anode to the bare stainless-steel substrate duringeach cycle. If the lithium was stripped completely from thestainless-steel substrate, the SEI layer would likely be destroyedor at least reformed on each cycle. Also, the iterative cyclingmethod more closely mimics a real battery that would containexcess active material.Time-of-flight secondary ion mass spectroscopy (ToF SIMS)

was used for depth profiling and chemical analysis of the SEI.The same experimental setup was used to prepare the ToFSIMS samples as previously described: 1 C/cm2 lithiumdeposition on a 316 stainless-steel electrode. The sample wasgently washed in DMC in the glovebox before SIMS analysis.Samples were analyzed with an ION-TOF5 SIMS usingbismuth as the primary ion. Depth profiling was done bysputtering the samples using 500 keV cesium ions over a 500 ×500 μm area. The analysis area was 150 × 150 μm within thelarger sputtered area.

■ RESULTS AND DISCUSSIONIonic liquids absorb ambient gas from the surroundings but canbe degassed by placing them in vacuum. The absorbed gas canbe observed as bubbles leaving the liquid. In this work, the ionicliquid was saturated with specific gases to determine their effect

Figure 1. Structure of N1114-Tf2N ionic liquid. The corresponding salt,Li Tf2N, was used in the electrolyte.

Figure 2. Typical voltage profile recorded under a CO2 atmosphere.Initially, an excess of lithium was deposited. Then, the cell was cycleduntil the excess was depleted. An overall efficiency was calculated bysumming the combined cathodic and anodic charge.

The Journal of Physical Chemistry C Article

dx.doi.org/10.1021/jp4001303 | J. Phys. Chem. C 2013, 117, 4980−49854981

on SEI formation and the Coulombic efficiency of Li cycling.The gases evaluated include argon, oxygen, nitrogen, dry air,and carbon dioxide.A 1 M Li+ ionic liquid electrolyte was bubbled with nitrogen

gas for 15 min, 3 h, and 6 h, followed by measurement of thecycling efficiency. After the initial bubbling period, the gas ofinterest was maintained in the head space of the cell so that theelectrolyte would be blanketed with the proper gas during theexperiment. The electrolyte was allowed to settle for 10 minbefore the cycling experiment began. The Coulombicefficiencies calculated from these experiments are shown inFigure 3.

The Coulombic efficiency using argon gas, 75%, was used asa reference. Argon was chosen as a reference because it is notreactive with lithium metal and is commonly used ingloveboxes. After 15 min of nitrogen bubbling, the Coulombicefficiency increased slightly compared with the argon reference.A longer nitrogen bubbling time, 3 h, led to higher Coulombicefficiency values; however, no further improvements wereobserved by extending bubbling time beyond 3 h. These resultsindicate that the time required for the electrolyte to saturateand equilibrate with the ambient is ∼3 h. Each subsequentexperiment was started with the cell assembled in the argonglovebox, and the same saturation time of 3 h was used for eachgas.Figure 4 shows the Coulombic efficiency for lithium with

different gases bubbled for 3 h prior to the experiment. Thevacuum experiment was conducted as a control because little orno gas should be present in the system. The experiments undervacuum and argon yielded similar Coulombic efficiencies. Thesimilarity between argon and vacuum confirms that argon hasessentially no effect on the Coulombic efficiency. In these cases,the SEI is formed solely from ionic liquid components. Anincrease in Coulombic efficiency was seen when the cell waspurged with oxygen or nitrogen. Dry air, being mainly a mixtureof nitrogen and oxygen, also showed an improvement; however,dry air also contains a small amount of carbon dioxide. Whenthe cell was purged with carbon dioxide, a pronounced decreasein Coulombic efficiency was observed. The potential curves forall cycling experiments were similar, showing an overpotentialof 50−60 mV versuss Li/Li+. This indicates no major changesin conductivity as a result of the gas treatment. SEMobservation of deposited samples showed no major differences

in the surface structure of the lithium. Samples werecharacterized by mossy and dendritic growth, as expectedfrom previous work.27,28 The main difference between thesamples here should be the SEI layer formed to passivate thelithium surface. ToF SIMS was used to sputter into the sample,and an elemental depth profile of the SEI and sample wascreated.Figure 5 shows a typical depth profile of a lithium sample

deposited under argon. Normalized counts are plotted versus

the sputtering time, which is proportional to sample depth,assuming uniform sputtering rate between the different layerson the lithium surface. The sensitivity factor for each specieswas not determined, so an exact mole fraction of each materialcannot be calculated. The SEI is clearly observed as a layer ofLiF on the surface extending to ∼60 s depth, where depth isexpressed as a sputtering time. During the course of theexperiment, Li+ and Li2

+ signals increase, indicating that bulklithium has been reached. Li2

+ represents a lithium cluster peakthat appears when the bulk Li deposit has been reached. Thismakes it a good indicator for the SEI thickness. The Li2O peakwas not tracked because it overlaps with a prominent surfacespecies, distorting the results. Instead, the LiO peak is shown.LiO is a natural byproduct of Li2O during ion bombardmentand shows the expected trends. Peaks for Li3N and LiOH werealso tracked and do not appear in significant quantities. Whensaturated with argon, the SEI is is primarily composed of a LiFlayer on the surface.

Figure 3. Coulombic efficiency plotted for argon and different gasnitrogen bubbling times.

Figure 4. Coulombic efficiency for a lithium anode in ionic liquidbubbled with different gases.

Figure 5. SIMS depth profile of sample deposited under an argonatmosphere.

The Journal of Physical Chemistry C Article

dx.doi.org/10.1021/jp4001303 | J. Phys. Chem. C 2013, 117, 4980−49854982

Depth profiles of samples when the ionic liquid was saturatedwith dry air, oxygen, and nitrogen are shown in Figure 6. Eachprofile shows a dominate LiF peak at the surface, with Li2increasing after a decrease in LiF. Surprisingly, Li3N and LiOwere absent from all profiles in Figure 6, even though they arepossible reaction products produced between lithium andnitrogen or oxygen. The fact that these reaction products aresoluble in the electrolyte and simply do not become part of theSEI is not ruled out, but chemically, the SEI appears to remain alayer of LiF regardless of absorbed gas. This is a sharpdeparture from previous results by Momma et al. and Younesiet al., who conducted similar experiments in organic electro-lytes.12,14 Specifically, Momma et al. observed Li3N in the SEIwhen conducting experiments in dry air; however, no Li3N orits fragments were observed in several iterations of ourexperiments.12

The lack of reaction with oxygen and nitrogen is consistentwith the fact that lithium reacts quickly with humid air but onlyslowly with dry air. The thermogravimetric analysis byMarkowitz et al. shows no weight gain under dry gas, andthus no reaction was observed with lithium metal.29,30 Inaddition to the slow reaction with nitrogen or oxygen under dryconditions, the gases are present in the ionic liquid at relativelysmall concentrations compared with the concentration of theionic liquid components. Thus, LiF formation due to reactionbetween lithium and the ionic liquid is favored. The SEIcomposed primarily of LiF layer would then provide surfacepassivation.There appears to be a significant difference is in the thickness

of the SEI layer formed depending on the different ambientgases used. Under argon, the thickness of the SEI correspondedto 60 s sputtering time and the Coulombic efficiency was 75%.For oxygen and nitrogen, the coubombic efficiency was 86%and the thickness of the SEI corresponded to 20 s of sputteringtime. Finally, under dry air, and Coulombic efficiency was 83%and it took 150 s to sputter through the SEI. The Coulombicefficiency appears to correlate with the thickness, except for thedry air case. Dry air has a Coulombic efficiency near nitrogenand oxygen, but an SEI thickness close to the argon sample.The apparent inconsistency with dry air may be explained by

the presence of carbon dioxide. Carbon dioxide was quitedetrimental to the Coulombic efficiency. A depth profile of theSEI formed in carbon dioxide ambient, Figure 7, has a very

thick layer of LiF, requiring 250 s sputtering time to reach thelithium. In addition, there was no rise of Li2 as seen in otherprofiles. When the experiment was conducted in carbon dioxideambient, bulk lithium deposition was inhibited and a largeamount of LiF was produced.Of the gases tested, carbon dioxide formed the thickest SEI.

According to Condemarin et al., Henry’s constant for carbondioxide is 60 atm, whereas oxygen and nitrogen were notdetectable.22 No data were found for argon solubility. Tf2N-based ionic liquids have the highest affinity for CO2 regardlessof cation.26,31 Separation of the anion and cation was observedat high CO2 pressures, but ambient pressures showed noseparation.32 The association of CO2 with Tf2N

− seems toinhibit lithium ion reduction and allow direct electrolytereduction and degradation. The salt, LiTf2N, could also beaffected by the presence of carbon dioxide.In an attempt to quantify the amount of SEI formed on each

sample, two methods were used to estimate the SEI thickness.

Figure 6. SIMS depth profiles for lithium samples deposited under nitrogen, oxygen, and dry air.

Figure 7. SIMS depth profile of sample deposited in carbon dioxideambient.

The Journal of Physical Chemistry C Article

dx.doi.org/10.1021/jp4001303 | J. Phys. Chem. C 2013, 117, 4980−49854983

First, the SIMS depth profiles were converted to depth,assuming the materials here sputter at the same rate as astandard material. The sputter rate of gold was experimentallymeasured under the same sputtering conditions as those usedin the lithium experiments and was found to be 5.2 nm/s. Goldwas chosen because it has a molecular-weight-to-density ratiosimilar to lithium, which is a determining factor in sputter rate.The time to sputter through the SEI was determined from theSIMS depth profile and converted to thickness. Second, the SEIthickness was calculated assuming that LIF formation is the solecause of the loss in Coulombic efficiency. The theoretical LiFthickness was calculated by assuming that all lost charge in thecycling experiments (i.e., all lost lithium) went to form LiFinstead of metallic Li. By assuming a bulk density for LiF, athickness can be calculated. The results of these two methodsare shown in Table 1. The SEI thickness correlates very well

with the loss of efficiency. A thinner SEI leads to a higherefficiency, lending credibility to the second method forcalculating SEI thickness.For the argon, nitrogen, and oxygen samples, the two

thickness approximations shown in Table 1 are in reasonableagreement. Given that lithium deposits are uneven by theirnature and the deposit itself is not evenly distributed, thecalculated and experimental thicknesses are similar. For carbondioxide, however, the second method, loss of Coulombicefficiency, gave a smaller value compared with the sputteringrate measurement. This could be due to a difference in densityof the LiF formed when carbon dioxide is present, which couldchange the sputter rate. Considering the lack of bulk lithiumformed, as evidenced by the low Li2 peak, it is more likely thathigher molecular weight reaction products are present in theSEI layer. Peaks for the fragments of the N1114 cation werefound throughout the carbon dioxide depth profile, suggestingan increased breakdown of the ionic liquid; however, thepresence of residual ionic liquid on the surface complicates theanalysis. Another possibility is that some of the LiF formed inthe carbon dioxide was due to direct reaction between lithiumand LiTf2N rather than the electrochemical reduction of thesalt. The association of Tf2N

− with carbon dioxide maydestabilize the salt, causing additional nonelectrochemical sidereactions.

■ CONCLUSIONSLithium metal deposition and reoxidation was studied underdifferent gas atmospheres in an attempt to evaluate the SEIformed. Cycling experiments were conducted to determine theefficiency of a lithium metal anode under these conditions.SIMS depth profiling was performed on samples depositedunder the same conditions to analyze the SEI layer formed.Nitrogen and oxygen atmospheres resulted in higherCoulombic efficiency compared with argon. An experimentunder vacuum confirmed that argon does not affect the SEI

formation. Carbon dioxide had a detrimental effect on cyclingefficiency.The chemical makeup of the SEI was similar regardless of

atmosphere, consisting of LiF. This is a departure from similarstudies in which composition changed to include more oxygenor nitrogen species. The fact that cycling efficiency increaseswith exposure to nitrogen, oxygen, and to a lesser degree airmakes this electrolyte promising for lithium-air batteries whenpaired with a water-rejecting membrane. A correlation betweenthe thickness of the SEI and cycling efficiency was observed,and thus the lower efficiency comes from the active materialand charge lost through the formation of the SEI. The SEI layerthickness in carbon dioxide ambient does not correlate as wellas the other gases. Carbon dioxide associates strongly with theTf2N

− and likely destabilizes the ionic liquid and salt, causingother side reactions.

■ AUTHOR INFORMATIONCorresponding Author*E-mail: paul.kohl@chbe.gatech.edu.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe gratefully acknowledge the financial support of the USArmy, contract US001-0000245070.

■ REFERENCES(1) Hirai, T.; Yoshimatsu, I.; Yamaki, J. J. Electrochem. Soc 1994, 141,611−614.(2) Zhang, W.-J. J. Power Sources 2011, 196, 13−24.(3) Zhuang, G. V.; Xu, K.; Jow, T. R.; Ross, P. N. Electrochem. Solid-State Lett. 2004, 7, A224−A227.(4) Edstrom, K.; Herstedt, M.; Abraham, D. P. J. Power Sources 2006,153, 380−384.(5) Zheng, H.; Jiang, K.; Abe, T.; Ogumi, Z. Carbon 2006, 44, 203−210.(6) Lewandowski, A.; Swiderska-Mocek, A. J. Power Sources 2009,194, 601−609.(7) Xu, K.; Zhuang, G. V; Allen, J. L.; Lee, U.; Zhang, S. S.; Ross, P.N.; Jow, T. R. J. Phys. Chem. B 2006, 110, 7708−7719.(8) Wang, X. J.; Lee, H. S.; Li, H.; Yang, X. Q.; Huang, X. J.Electrochem. Commun. 2010, 12, 386−389.(9) Stark, J. K.; Ding, Y.; Kohl, P. a. J. Electrochem. Soc. 2011, 158,A1100−A1105.(10) Ota, H.; Sakata, Y.; Otake, Y.; Shima, K.; Ue, M.; Yamaki, J. J.Electrochem. Soc. 2004, 151, A1778−A1788.(11) Morigaki, K. J. Power Sources 2002, 104, 13−23.(12) Momma, T.; Nara, H.; Yamagami, S.; Tatsumi, C.; Osaka, T. J.Power Sources 2011, 196, 6483−6487.(13) Kim, G.-T.; Jeong, S. S.; Xue, M.-Z.; Balducci, a.; Winter, M.;Passerini, S.; Alessandrini, F.; Appetecchi, G. B. J. Power Sources 2012,199, 239−246.(14) Younesi, R.; Hahlin, M.; Roberts, M.; Edstrom, K. J. PowerSources 2013, 225, 40−45.(15) Aurbach, D.; Youngman, O.; Dan, P. Electrochim. Acta 1990, 35,639−655.(16) Aurbach, D. J. Electrochem. Soc. 1989, 136, 1611.(17) Borgel, V.; Markevich, E.; Aurbach, D.; Semrau, G.; Schmidt, M.J. Power Sources 2009, 189, 331−336.(18) Kim, K.; Lang, C. M.; Kohl, P. a. J. Electrochem. Soc. 2005, 152,E56.(19) Stark, J. K.; Ding, Y.; Kohl, P. a. J. Electrochem. Soc. 2011, 158,A1100.(20) Shimamura, O.; Yoshimoto, N.; Matsumoto, M.; Egashia, M.;Morita, M. J. Power Sources 2011, 196, 1586−1588.

Table 1. Comparison of SEI Thicknesseses Based on SIMSand Efficiency Calculations

Ar(75%)(nm)

N2/O2(85%)(nm)

CO2(44%)(nm)

thickness based on Au sputter rate 312 104 1300thickness calculated from Coulombicefficiency

255 152 570

The Journal of Physical Chemistry C Article

dx.doi.org/10.1021/jp4001303 | J. Phys. Chem. C 2013, 117, 4980−49854984

(21) Sakaebe, H.; Matsumoto, H.; Tatsumi, K. Electrochim. Acta2007, 53, 1048−1054.(22) Condemarin, R.; Scovazzo, P. Chem. Eng. J. 2009, 147, 51−57.(23) Ramdin, M.; Loos, T. W. De; Vlugt, T. J. H. Ind. Eng. Chem. Res.2012, 51, 8149−8177.(24) Blath, J.; Christ, M.; Deubler, N.; Hirth, T.; Schiestel, T. Chem.Eng. J. 2011, 172, 167−176.(25) Anthony, J. L.; Maginn, E. J.; Brennecke, J. F. J. Phys. Chem. B2002, 106, 7315−7320.(26) Anthony, J. L.; Anderson, J. L.; Maginn, E. J.; Brennecke, J. F. J.Phys. Chem. B 2005, 109, 6366−7634.(27) Gireaud, L.; Grugeon, S.; Laruelle, S.; Yrieix, B.; Tarascon, J.-M.Electrochem. Commun. 2006, 8, 1639−1649.(28) Lopez, C. M.; Vaughey, J. T.; Dees, D. W. J. Electrochem. Soc.2009, 156, A726−A729.(29) Markowitz, M.; Boryta, D. J. Chem. Eng. Data 1962, 7, 586−591.(30) Rhein, R. A. Lithium Combustion: A Review; Naval WeaponsCenter: China Lake, CA, 1990; pp 1−60(31) Sharma, P.; Park, S.; Do; Baek, I. H.; Park, K. T.; Yoon, Y. I.;Jeong, S. K. Fuel Process. Technol. 2012, 100, 55−62.(32) Shi, W.; Maginn, E. J. J. Phys. Chem. B 2008, 112, 2045−2055.

The Journal of Physical Chemistry C Article

dx.doi.org/10.1021/jp4001303 | J. Phys. Chem. C 2013, 117, 4980−49854985