Interface stabilization via lithium for promoted performance of ...30852...2 cathodes and cycled...

The Journalof Chemical Physics ARTICLE scitation.org/journal/jcp

Interface stabilization via lithiumbis(fluorosulfonyl)imide additive as a keyfor promoted performance of graphite∥LiCoO2pouch cell under −20 ○C

Cite as: J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280Submitted: 2 January 2020 • Accepted: 12 February 2020 •Published Online: 4 March 2020

Hieu Quang Pham,1,a) Gyeong Jun Chung,1 Jisoo Han,1 Eui-Hyung Hwang,2 Young-Gil Kwon,2and Seung-Wan Song1,b)

AFFILIATIONS1Department of Chemical Engineering and Applied Chemistry, Chungnam National University, Daejeon 34134, South Korea2Lichem Co., Ltd., Geumsan 32715, South Korea

Note: This paper is part of the JCP Special Topic on Interfacial Structure and Dynamics for Electrochemical Energy Storage.a)Present address: Electrochemistry Laboratory, Paul Scherrer Institute, 5232 Villigen, Switzerland.b)Author to whom correspondence should be addressed: [email protected]. Tel.: +82-42-821-7008. Fax: +82-42-822-6637

ABSTRACTThe effects of lithium bis(fluorosulfonyl)imide, Li[N(SO2F)2] (LiFSI), as an additive on the low-temperature performance of graphite∥LiCoO2pouch cells are investigated. The cell, which includes 0.2M LiFSI salt additive in the 1M lithium hexafluorophosphate (LiPF6)-based conven-tional electrolyte, outperforms the one without additive under −20 ○C and high charge cutoff voltage of 4.3 V, delivering higher dischargecapacity and promoted rate performance and cycling stability with the reduced change in interfacial resistance. Surface analysis results onthe cycled LiCoO2 cathodes and cycled graphite anodes extracted from the cells provide evidence that a LiFSI-induced improvement of high-voltage cycling stability at low temperature originates from the formation of a less resistive solid electrolyte interphase layer, which containsplenty of LiFSI-derived organic compounds mixed with inorganics that passivate and protect the surface of the cathode and anode from fur-ther electrolyte decomposition and promotes Li+ ion-transport kinetics despite the low temperature, inhibiting Li metal-plating at the anode.The results demonstrate the beneficial effects of the LiFSI additive on the performance of a lithium-ion battery for use in battery-poweredelectric vehicles and energy storage systems in cold climates and regions.

Published under license by AIP Publishing. https://doi.org/10.1063/1.5144280., s

I. INTRODUCTION

Nowadays, lithium (Li)-ion batteries (LIBs) are ubiquitouslyutilized as power sources for various electronic devices, electricvehicles (EV), and stationary energy storage systems (ESS). State-of-the art LIB technology, however, still shows a drastic decreasein energy and power at low temperatures below 0 ○C1,2 and inspecial environments such as high latitude, cold weather, andaerospace. Much less attention has been paid to the phenomenaand performance improvement of LIBs at low temperatures thanhigh temperatures, despite the exposure of commercialized EVand ESS to changing temperatures outdoors. The low-temperatureperformance of LIB is hampered by multiple factors including the

reduced Li+-ion conductivity in the cathode/anode, the electrolyte,the solid electrolyte interphase (SEI) layer, and a phase changeof high melting-point (mp) electrolyte component from liquid tosolid.14 A commercial LIB includes a conventional carbonate-basedorganic liquid electrolyte that is typically composed of lithium hex-afluorophosphate (LiPF6) salt, ethylene carbonate (EC), and lin-ear carbonate solvents such as dimethyl carbonate (DMC), diethylcarbonate (DEC), and/or ethyl methyl carbonate (EMC). EC is ahigh melting-point (mp) molecule (mp = 36 ○C), and the mp ofEMC is relatively low (mp = −14.5 ○C). Despite the high ionic con-ductivity of LiPF6 salt, its disadvantage is thermal instability andreactivity to moisture or protic species releasing a toxic, highlyaggressive, and undesirable gases, including HF, which is a safety

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-1

Published under license by AIP Publishing

https://scitation.org/journal/jcphttps://doi.org/10.1063/1.5144280https://www.scitation.org/action/showCitFormats?type=show&doi=10.1063/1.5144280https://crossmark.crossref.org/dialog/?doi=10.1063/1.5144280&domain=pdf&date_stamp=2020-March-4https://doi.org/10.1063/1.5144280https://orcid.org/0000-0001-8501-5815https://orcid.org/0000-0003-4208-206Xmailto:[email protected]://doi.org/10.1063/1.5144280


issue.5 Importantly, a major issue in low-temperature LIB opera-tion is the occurrence of severe Li metal-plating and its dendriticgrowth at the graphite anode due to the sluggish diffusion kineticsof Li+-intercalation, which is different from that at room temper-ature. Unless otherwise being charged at quite a low rate such as0.02C taking 50 h, Li metal-plating occurs by the sluggish kineticsof Li+-diffusion in the electrolyte as well as at both surfaces andbulk of the graphite anode during a sub-freezing charge, which ispermanent and cannot be removed with cycling, leading to a sud-den performance failure and safety hazard.5 In the perspectives ofelectrolyte technology, various approaches have been suggested toimprove LIB performance, including the blending of several solventsto ternary and quaternary electrolytes4,6–8 for lowering the mp andthe use of functional electrolyte additives3,9 and mixed lithium saltsor lithium salt as an additive, most of which, however, focus onthe performance improvement under room and elevated tempera-tures.2 A few literature studies on the development of new lithiumsalts such as lithium bis(oxalate)borate (LiBOB) or lithium diflu-oro(oxalato)borate (LiDFOB) for low-temperature operation of LIBsare available.10 For example, 0.9M LiDFOB:0.1M lithium tetraflu-oroborate (LiBF4) mixed salts in EC:dimethyl sulfite (DMS):EMCmixed solvents (1:1:3 by volume) showed an improved cycling sta-bility at −20 ○C than the LiPF6-based electrolyte.11 Recently, lithiumbis(fluorosulfonyl)imide (LiFSI) attracts considerable attention as anew salt for both Li-ion and Li metal batteries, despite high cost.This salt shows an improved thermal and electrochemical stability,higher ionic conductivity, and a higher Li transference number com-pared to conventional LiPF6.5,12,13 LiFSI does not hydrolyze in thepresence of water to release HF, showing the improved chemicaldurability and electrochemical performance of various half-cells ofLiCoO2,13 graphite,14–16 and Si,17 and full-cells of graphite∥LiCoO213and graphite∥LiNi1/3Mn1/3Co1/3O212 when being used as main saltor mixed salt under room and elevated temperatures. Wang et al.reported that the function of LiFSI as a mixed salt is limited whenbeing charged to high voltages up to 4.45 V.12,18 Studies of the useand effects of LiFSI as an additive on the low-temperature perfor-mance of high-voltage LiCoO2-based LIB have yet to be reportedand clarified.

Herein, we present a LiFSI additive-assisted graphite∥LiCoO2high-voltage pouch full-cell that achieves a Li metal-plating-free LIBunder −20 ○C, in contrast to the occurrence of Li metal-plating inthe cell with a commercial LiPF6-based electrolyte under the sametest condition. Furthermore, systematic studies of the effects of theLiFSI additive on the interfacial phenomena of the graphite anodeand the LiCoO2 cathode were conducted utilizing ex situ surface-sensitive attenuated total reflectance (ATR) FTIR spectroscopy andx-ray photoelectron spectroscopy (XPS). A basic understanding onthe interfacial reaction behavior among cathode, anode, and LiFSIadditive-containing electrolyte and its correlation with the enhancedlow-temperature cycling performance are discussed, which wouldpave the way for a wider use of LIB-powered EV and ESS.

II. EXPERIMENTAL SECTIONA. Electrochemistry

The anodes were prepared by coating the slurry with the com-position of 85 wt. % natural graphite active material (BTR), 5 wt. %

carbon black (super-P, Timcal), and 10 wt. % polyvinylidene fluo-ride (PVdF; Aldrich) onto a copper foil. The cathodes, whose slurrywas composed of 80 wt. % LiCoO2 active material (supplied froman industry), 10 wt. % carbon black (super-P, Timcal), and 10 wt. %PVdF (Aldrich), were coated on an Al foil. Active mass loading ofthe LiCoO2 cathode was ≈4 mg cm−2, and the N/P ratio was 1.1.The coated anodes and cathodes were dried at 110 ○C for 12 h undervacuum prior to battery assembly.

The ionic conductivity of electrolytes, which are 1M LiPF6/EC:EMC (3:7 volume ratio, Panax E-Tec) as the base electrolyte(henceforth, Base EL) with various concentrations of LiFSI (bat-tery grade, >99.99% purity, Lichem, Ltd.) from 0M to 0.5M result-ing in the total concentration of Li+ ion in the electrolyte from1.0M to 1.5M, was measured using a conductivity meter (MettlerToledo S230) at room temperature under Ar atmosphere. The con-ductivity meter was calibrated with the standard solution of 1413 μScm−1 conductivity before the measurement. We determined 0.2M(according to ≈3.07 wt. %) as the optimized concentration of LiFSIsalt type-additive in maintaining a high enough ionic conductivityof the electrolyte [Fig. 1(a)].

Linear sweep voltammetry (LSV) of electrolytes was conductedto evaluate the anodic stability of electrolytes at a scan rate of 1 mVs−1 from an open-circuit potential of ≈3.4 to 6.0 V vs Li/Li+ forcharging of full-cells to 4.3 V (corresponding to 4.35 V vs Li/Li+)using a three-electrode cell that included LiCoO2 working electrode,lithium metal counter electrode, lithium metal reference electrode,and base electrolyte without and with 0.2M LiFSI additive.

FIG. 1. (a) Changes in the room temperature ionic conductivity of the electrolyteas a function of the concentration of LiFSI added to the base electrolyte (Base EL)of 1M LiPF6/EC:EMC (3:7 volume ratio) and (b) linear sweep voltammograms ofthree-electrode cells including the LiCoO2 working electrode and Base EL withoutand with 0.2M LiFSI additive.

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-2


https://scitation.org/journal/jcp


The dimension of the LiCoO2 cathode for pouch cells was20 mm × 20 mm, and that of the graphite anode was 22 mm ×22 mm. The pouch cells were 60 mm long, 60 mm wide, and 3.5 mmthick. An Al tab was welded to the cathode, and a Ni tab was weldedto the anode for connecting to external circuits. Pouch cells includedthe separator (Celgard C210) and were filled with 1 ml of base elec-trolyte without and with 0.2M LiFSI additive in a dry room with acontrolled dew point of less than −45 ○C. After electrolyte filling,the pouch cells were vacuum sealed with a compact vacuum sealer.The resultant graphite∥LiCoO2 pouch full-cells were cycled under−20 ○C for 50 cycles at a rate of 0.2C (28 mA g−1) in the constanttemperature chamber (TH-KE-065, JEIO Tech) after a formationcycle between 3.0 V and 4.3 V at 0.1C (14 mA g−1) at room tem-perature, using a multichannel battery cycler (WBCS3000, Won-ATech). 1C in this work is defined as 140 mA g−1, which is the currentrequired to charge the LiCoO2-based battery in one hour at roomtemperature. AC impedance measurement was conducted at −20 ○Cafter the 1st and 50th cycles using an impedance spectroscopic ana-lyzer (VSP SP-150, Bio-Logic) in the frequency range from 100 kHzto 10 mHz with an amplitude of 10 mV. The rate capability of thepouch cells was tested under −20 ○C by first charging to 4.3 V at 0.1Cand then discharge to 3.0 V at variable rates ranged from 0.1C to 5C.

B. CharacterizationTo monitor a possible Li metal-plating, particle morphology

changes of graphite anodes with cycling were analyzed using fieldemission scanning electron microscopy (FE SEM; FEI Sirion) at 10kV. Crystal structural changes of graphite anodes and LiCoO2 cath-odes were examined using ex situ x-ray diffraction (XRD) analysis.Those were mounted in a tightly sealed sample holder to avoid atmo-spheric contamination and measured from 10○ to 80○ at a scan rateof 1○ min−1 at 40 kV and 40 mA using a powder x-ray diffractometer(Rigaku D/MAX-2200) equipped with Ni-filtered Cu–Kα radiation.The composition of the SEI layer of graphite anodes and LiCoO2cathodes was analyzed by ex situ ATR FTIR spectroscopy using aNicolet 6700 FTIR spectrometer equipped with a MCT detector. Thecycled anodes and cathodes in the pouch cells were washed withDMC for the removal of the residual electrolyte followed by dryingin the glove box at room temperature. They were directly mountedon the tightly closed ATR unit in the glove box to avoid atmosphericcontamination during transportation to a dry nitrogen-purged sam-ple compartment of the IR instrument as well as during IR mea-surement. The spectra were collected with 512 scans and a spectralresolution of 4 cm−1. XPS surface analysis was conducted using XPS(Thermo, MultiLab 2000) with an Al Kα X-ray source at 15 kV. Highresolution spectra were obtained at a power of 150 W under a basepressure of 5 × 10−10 mbar with a pass energy of 100 eV. The cycledanodes and cathodes after washing with DMC were transferred fromthe glove box to the XPS chamber using a vacuum-sealed containerwithout exposure to the air. The spot size was about 500 μm, and thepass energy was 30 eV. Binding energy was corrected based on the C1s level at 284.5 eV.

III. RESULTS AND DISCUSSIONA. Ionic conductivity and linear sweep voltammetry

Figure 1(a) shows the change in the room temperature ionicconductivity of the electrolyte as a function of the concentration of

LiFSI that was added to 1M LiPF6/EC:EMC. As increasing the con-centration of LiFSI from 0M to 0.5M, corresponding to the totalconcentration of Li+-ion in the electrolyte from 1.0M to 1.5M, theionic conductivity of the electrolyte slightly decreases from 9.12 mScm−1 of base electrolyte of 1M LiPF6/EC:EMC. The addition of LiFSIbeyond 0.3M (total concentration of Li+ ion >1.3M) causes a drasticreduction in the ionic conductivity below 8.5 mS cm−1. This is thesimilar tendency to the electrolyte of LiPF6 only that was reportedin the literature;19,20 the maximum ionic conductivity is reached on1M, from which ionic conductivity decreases as increasing the con-centration of LiPF6 (>1.2M). Despite higher ionic conductivity ofLiFSI than LiPF6 at the fixed same concentration, the reduction inthe ionic conductivity with the concentration of LiFSI is due to anincrease in the viscosity and the salt association as described by Dinget al.20 To keep the ionic conductivity high enough and to empha-size the effectiveness LiFSI, we determined 0.2M [Fig. 1(a)] as theoptimized concentration of LiFSI additive.

The anodic stability of the base electrolyte without and with0.2M LiFSI additive was comparatively tested utilizing LSV withthree-electrode cells including the LiCoO2 working electrode. Bothelectrolytes [Fig. 1(b)] exhibit two anodic peak currents at ≈4.1 Vand ≈4.5–4.6 V, which are associated with Co oxidation from Co3+to Co4+ along with electrochemical electrolyte oxidation. A largeanodic current rise from 5.2 V is due to the oxidative decompo-sition of a whole electrolyte. A detailed look at the voltage regionof 4.0–4.2 V in the inset presents the anodic peak current of theLiFSI additive-containing electrolyte at 4.086 V, which is slightlylower than 4.115 V of the base electrolyte. The LiFSI additive is elec-trochemically oxidized prior to LiPF6 probably contributing to theformation of the SEI layer at the cathode and anode. As a reference,LSV was conducted on Al foil (not shown) with LiFSI additive; theanodic peak current rise of the LiFSI additive-containing electrolyteis observed from 3.97 V, which is slightly earlier than ≈4.2 V of thebase electrolyte only. No corrosion of Al is observed.

B. Electrochemical cycling performanceof graphite∥LiCoO2 pouch cells under −20 ○C

Figure 2 compares the electrochemical cycling performance ofgraphite∥LiCoO2 pouch cells under −20 ○C in the base electrolyte of1M LiPF6/EC:EMC without and with 0.2M LiFSI additive between3.0 V and 4.3 V at 0.2C. In the first charge curves of voltage pro-files [Figs. 2(a) and 2(b)], both cells exhibit two voltage plateaus{anodic peaks in dQ/dV plots [Figs. 2(a′) and 2(b′)]} at ≈3.9 and≈4.0 V that are attributed to the oxidation of Co3+ to Co4+ byLi+-deintercalation. In the subsequent discharge curves, two vagueplateaus {cathodic peaks in dQ/dV plots [Figs. 2(a′) and 2(b′)]} at≈3.67 and ≈3.95 V are due to the reduction of Co4+ to Co3+ byLi+-intercalation.21,22 In the base electrolyte, initial charge and dis-charge capacities of 66 mA h g−1 and 64 mA h g−1 correspond tothe initial coulombic efficiency (ICE) of 97% from which capacitiesdecrease gradually with cycling [Figs. 2(a), 2(c), and 2(d)] result-ing in the capacity retention of 72% after 50 cycles [Figs. 2(a) and2(c)]. The loss of structural resolution in dQ/dV plots [Fig. 2(a′)]and a new cathodic peak at ≈3.476 V that appears at the 50th cyclein the base electrolyte indicate the degradation of the LiCoO2 crystalstructure by a phase transition from the hexagonal to a monoclinicphase.23 Strain might form and accumulate by repetitive structural

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-3




FIG. 2. Voltage profiles ofgraphite∥LiCoO2 pouch cells between3.0 V and 4.3 V at 0.2C under −20 ○Cin (a) Base EL of 1M LiPF6/EC:EMCwithout and (b) with 0.2M LiFSI additiveand (a′) and (b′) their differentialcapacity plots, (c) cycling performance,and (d) coulombic efficiency.

stress during Li+-deintercalation and -intercalation processes24,25

under higher charge cutoff voltage 4.3 V than conventional 4.2 V,as reported in the earlier literature.23 On the other hand, with theuse of the LiFSI additive [Figs. 2(b), 2(c), and 2(d)], cycling perfor-mance is dramatically improved; initial charge and discharge capac-ities increase to 91 mA h g−1 and 90 mA h g−1, respectively, with anincrease in the ICE to ≈99%. In addition, a higher capacity retentionof 92% after 50 cycles and Coulombic efficiencies than those with thebase electrolyte are achieved [Figs. 2(c) and 2(d)].

The impedance spectral evolution during cycling is a relativemeasure of an electrolyte-dependent increase in internal interfacialresistance of the cells. The semicircle in the high-frequency regionafter the first cycle, attributed to the summed resistance of the SEIformation at the cathode and anode,26,27 is smaller upon the useof the LiFSI additive [Fig. 3(b)] than that of the base electrolyte[Fig. 3(a)]. After the 50th cycle with LiFSI additive [Fig. 3(b)], the

semicircles tend to maintain, in contrast to a large increase in thecell with the base electrolyte [Fig. 3(a)] with cycling. The LiFSIadditive might form a less resistive SEI layer at the surface of thecathode and anode and reduces the impedance rise, effectively sta-bilizing the cathode/anode–electrolyte interfaces. This is correlatedwith the higher capacity and improved cycling stability, as observedin Figs. 2(c) and 2(d).

Figure 4 displays the rate capability of pouch cells in the baseelectrolyte without and with 0.2M LiFSI additive between 3.0 V and4.3 V under −20 ○C at different rates from 0.1C to 5C. Overall,with LiFSI additive, much higher capacities are delivered at all rates[Figs. 4(b) and 4(c)]; discharge capacities at 0.1C, 0.2C, 0.5C, 1C, 2C,and 5C are 106 mA h g−1, 103 mA h g−1, 97 mA h g−1, 89 mA h g−1,69 mA h g−1, and 17 mA h g−1, respectively. However, the baseelectrolyte [Figs. 4(a) and 4(c)] produces somewhat lower dischargecapacities of 79 mA h g−1, 74 mA h g−1, 68 mA h g−1, 61 mA h g−1,

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-4




FIG. 3. AC impedance spectral evolution of graphite∥LiCoO2 pouch cells after the1st and 50th cycles between 3.0 V and 4.3 V at 0.2C under −20 ○C in (a) Base ELof 1M LiPF6/EC:EMC without and (b) with 0.2M LiFSI additive.

48 mA h g−1, and 15 mA h g−1 at the same rates, respectively. Adischarge capacity of 69 mA h g−1 delivered at 2C with LiFSI addi-tive is 67% of the capacity delivered at 0.2C, which is higher thanjust 48 mA h g−1 in the base electrolyte that is 65% of the capacity

FIG. 4. Discharge curves of graphite∥LiCoO2 pouch cells between 3.0 V and 4.3 Vat different rates of 0.1–5C under −20 ○C in (a) Base EL without and (b) with 0.2MLiFSI additive and (c) their rate capability depending on the electrolyte.

at 0.2C. When the discharge rate is back to 0.1C [Fig. 4(c)], the cellwith LiFSI additive still achieves 96 mA h g−1 that is 91% retentionof 106 mA h g−1 delivered at the initial 0.1C. In the case of base elec-trolyte [Fig. 4(c)], the discharge capacity upon the return to 0.1C isjust 65 mA h g−1, corresponding to the 82% retention of 79 mA hg−1 delivered at the initial 0.1C. The rate capability is influenced bymultiple factors such as Li+-ion conductivity of the electrolyte, Li+-ion diffusion kinetics to both the surface and bulk of the graphiteanode and LiCoO2 cathode, and interfacial resistance that dependson the properties of the surface protective SEI layer at the graphiteanode. In particular, as Li+-ion transport kinetics is predicted to besluggish at low temperatures, rate-capability would be inferior aslowering the battery operation temperature. Despite the relativelylower ionic conductivity of the LiFSI additive-containing 1.2M elec-trolyte [Fig. 1(a)] than the 1M base electrolyte, the improved rateof capability and stability of the LiFSI additive-assisted pouch cellunder −20 ○C imparts the benefits of lower interfacial resistance bythe formation of the less resistive SEI layer at the cathode/anode,13,14

as observed in impedance spectra [Fig. 3(b)].

C. LiFSI additive-derived inhibition of Li metal-platingat graphite anode

The photographs of graphite anodes obtained after cycling inthe base electrolyte without and with 0.2M LiFSI additive under−20 ○C are displayed in Fig. 5. The cycled anode with the baseelectrolyte [Fig. 5(b)] clearly shows the formation of light-coloredLi metal deposits, compared to a clean and smooth surface of thepristine anode [Fig. 5(a)]. Continuation of low-temperature cycling

FIG. 5. Photographs and SEM images, respectively, of [(a) and (a′)] pristineand cycled graphite anodes obtained from graphite∥LiCoO2 pouch cells between3.0 V and 4.3 V at 0.2C under −20 ○C in Base EL [(b) and (b′)] without and[(c) and (c′)] with 0.2M LiFSI additive.

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-5




without a death of the pouch cell, despite the occurrence of Li metal-plating, might be due to neither Li-dendrites growth nor penetra-tion through the separator. Safety hazard through shorting is yetto be occurred but lower capacities and inferior performance aswell as impedance rise should originate from this Li metal-platingphenomenon at the graphite anode and blocked Li+-ion interca-lation into interlayers. The corresponding SEM image [Fig. 5(b′)]clearly shows the presence of aggregated Li metal particles, but nolong-shaped Li-dendrites. On the other hand, the anode cycled withLiFSI additive displays a well-maintained smooth and Li metal-plating-free surface morphology [Fig. 5(c′)], and overall remainedparticle morphology of pristine [Fig. 5(a′)]. It ensures the role ofthe LiFSI additive in inhibiting the occurrence of Li metal-platingand thus improving low-temperature battery performance (Figs. 2and 4).

D. Bulk structural changes of graphite anodeand LiCoO2 cathode

XRD analyses were conducted to examine the bulk structuralchanges in graphite anodes and LiCoO2 cathodes with cycling with-out and with LiFSI additive under −20 ○C. First, Fig. 6(A) shows thecomparison of powder XRD patterns of pristine and cycled graphiteanodes. After cycling, in common, the peak of interlayer spacing,d002, of pristine (3.3511 Å) shifts to a lower 2θ region along with thereduction of peak intensity as well as peak broadening, particularly,at a higher degree in base electrolyte only (3.3534 Å) [Fig. 6(A′–b)]

FIG. 6. Power XRD patterns of (A) the graphite anode and the reflection of the(A′) (002) plane and (B) the LiCoO2 cathode and the reflections of (B′) (003) and(B′′) (104) planes obtained from graphite∥LiCoO2 pouch cells between 3.0 V and4.3 V at 0.2C under −20 ○C, (a) pristine and cycled anodes and cathodes in BaseEL (b) without and (c) with 0.2M LiFSI additive.

than the one with LiFSI additive (3.3512 Å) [Fig. 6(A′–c)]. The resul-tant crystallite size along the c-axis (Lc) [Fig. 6(A)], calculated fromthe patterns, decreases from 128.13 nm of pristine to 107.89 nmwith LiFSI additive to 103.49 nm with the base electrolyte withoutadditive. This indicates that the c-dimension of the graphite crystal-lite and the number of graphene layers decrease by the loss of bothcrystallinity and structural order, particularly more in the base elec-trolyte. It, in turn, reveals that the use of the LiFSI additive providesa better maintenance of the bulk structure of the graphite anode withcycling.

Noticeable structural changes are observed on the cathodes, asshown in Fig. 6(B). The XRD pattern of the pristine LiCoO2 cath-ode [Fig. 6(B–a)] indicates a well-defined α-NaFeO2-type hexagonal(R3̄m) layered structure with the a and c lattice parameters of 2.8742Å and 13.9410 Å, respectively, and the c/a ratio of 4.8504 and unitcell volume of 99.73 Å3. After cycling, in common, the coexistence oftwo hexagonal phases (H1 and H2) is found and their correspond-ing split peaks are indexed in Figs. 6(B′) and 6(B′′). The H2 phaseis known to form by the phase transition from H1 to H2 duringthe charge process and to have the same hexagonal crystal structure(CdCl2 type) to H1 but a larger c lattice parameter.23 The presenceof H2 phase, despite after 50 cycles under −20 ○C that was endedas discharged, indicates that the bulk structure of the LiCoO2 didnot reversibly transform back to H1.28,29 Simultaneously observedafter cycling in common is the shift of the (003) peak of the H1phase to a lower 2θ region [Fig. 6(B′–b) and Fig. 6(B′–c)], represent-ing the lattice expansion along the c-axis, particularly, at a higherdegree in base electrolyte [Fig. 6(B′–b)]. As a result, cell volume[Fig. 6(B–b)] increases at a higher degree in base electrolyte.In addition, peaks of the H1 phase [Figs. 6(B′), 6(B′′–b), and6(B′′–c)] are broadened and weakened, particularly, at a higherdegree on the cathode cycled in the base electrolyte [Figs. 6(B′–b)and 6(B′′–b)], compared to those of pristine [Figs. 6(B′–a) and6(B′′–a)] and the one cycled with LiFSI additive [Figs. 6(B′–c) and6(B′′–c)]. In turn, this is the clear evidence that the bulk crys-tal structure of the LiCoO2 cathode is better maintained duringcycling with LiFSI additive, representing a higher reaction reversibil-ity, which is consistent with the greater maintenance of structuralresolution in dQ/dV plots [Fig. 2(b′)] and improved cycling perfor-mance [Figs. 2(c) and 2(d)]. It is curious how the LiFSI electrolyteadditive induces to maintain better the bulk structure of LiCoO2 aswell as the graphite anode so dramatically under −20 ○C. Because, ingeneral, the electrolyte and additive should be first in contact withthe anode and cathode at their interfaces and the electrochemicalreaction begins at the interface (i.e., the surface of the anode andcathode), surface chemistry perspective can give an insight into theroles of the LiFSI additive in this −20 ○C LIB. We, thus, conductedthe surface characterization of anodes and cathodes using surfacesensitive analytical tools.

E. Surface chemistry1. SEI characterization of graphite anodes

Figure 7 compares the ATR FTIR spectra of pristine andcycled graphite anodes under −20 ○C. All cycled anodes [Figs. 7(b)and 7(c)] show peaks at 2970–2910 cm−1 and 1460–1413 cm−1

attributed to methyl (CH3−−) and methylene (−−CH2−−) groupsof alkyl functionalities that come from newly formed sur-

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-6




FIG. 7. ATR FTIR spectral comparison of (a) pristine and cycled graphite anodesof graphite∥LiCoO2 pouch cells between 3.0 V and 4.3 V at 0.2C under −20 ○C inBase EL (b) without and (c) with 0.2M LiFSI additive.

face compounds in addition to the PVdF binder of pristine[Fig. 7(a)].30–32 Also shown [Figs. 7(b) and 7(c)] are peaks at1740−1720 cm−1 attributed to νsym(C==O) of ester (RCO2R′; R,R′ =alkyl group), which are confirmed with the peaks at 1270 cm−1 and1160−1150 cm−1 due to ν(C−−O−−C)asym and ν(C−−O−−C)sym.30–34

New peaks at 1660−1580 cm−1 are attributed to νsym(C==O) of

alkyl carbonate salt (ROCO2Li) and carboxylate salt (RCO2Li).30,33

A prominent peak near 1470 cm−1 is due to the ν(CO3)sym ofLi2CO3, along with δ(CO3) at 850 cm−1. Peaks at 1320 cm−1,1040 cm−1, and 964 cm−1 are due to ν(P==O), ν(P−−O−−C)asym,and ν(P−−O−−C), respectively, of organic phosphorus fluorides (O==PF3−y(OR)y).30,32,33 Also observed are PF-containing species at 930and 839−828 cm−1, formed by LiPF6 decomposition. What is dis-tinguished with the use of the LiFSI additive [Fig. 7(c)] fromthe base electrolyte [Fig. 7(b)] is relatively stronger absorbance oforganic compounds of ester and alkyl carbonate salt implying theirhigher concentration, whereas reduced intensity of Li2CO3 due tolower concentration. In addition, new features at 1210−1180 cm−1attributed to ν(S==O) of (RO)2SO and −−SO2N−− functionalitiesmight be derived from LiFSI decomposition products.

IR analysis results on cycled graphite anodes reveal that theLiFSI additive alters the SEI composition and concentration (i.e.,SEI thickness and/or surface coverage). With the use of the LiFSIadditive, the SEI layer is composed of a plenty of organic com-pounds including ester, alkyl carbonate salt, and carboxylate salt andorganic phosphorus-fluoride compound along with LiFSI decom-position products passivating better the surface of the graphiteanode; the SEI formed in the base electrolyte includes Li2CO3 andLiPF6 decomposition products as well-known major species5,35 andorganic compounds at a relatively lower concentration.

Figure 8 shows XPS spectral changes on the surface elementsof the graphite anode with cycling. In the C 1s spectra, after cycling

FIG. 8. XPS spectral comparison of surface elements of (a) pristine and cycled graphite anodes in graphite∥LiCoO2 pouch cells between 3.0 V and 4.3 V at 0.2C under−20 ○C in Base EL of 1M LiPF6/EC:EMC (b) without and (c) with 0.2M LiFSI additive.

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-7




[Figs. 8(b) and 8(c)], new peaks of C−−O (286.2 eV), C==O (288.3 eV),and CO3 (290 eV) containing species are commonly observed,30,32,36

in addition to C−−C/C−−H from carbon black, and those at 291 eVand 285.6 eV of C−−F and C−−H from the PVdF binder32 of pris-tine [Fig. 8(a)]. Those might be of organic compounds such asester, alkyl carbonate salt, and carboxylate salt that we observedin IR spectra (Fig. 7). Their relatively stronger intensity upon theuse of LiFSI additive [Fig. 8(c)] is associated with higher concen-tration of organic compounds, consistent IR observation (Fig. 7).In addition, still strongly remained features of carbon black andthe PVdF binder of pristine on the cycled anode in the base elec-trolyte [Fig. 8(b)] contrary to their vague features with LiFSI addi-tive indicate a lower surface coverage by the SEI species in the baseelectrolyte.

In O 1s spectra, no peak is observed on pristine. A broadpeak on cycled anodes [Figs. 8(b) and 8(c)] is, thus, due to oxygen-containing species of the SEI layer. A stronger peak intensity uponthe use of the LiFSI additive implies the presence of higher concen-tration of oxygen-containing species such as ester, alkyl carbonatesalt, carboxylate salt, consistent with C 1s spectra. In the Li 1s spectra,the concentration of Li2CO3 is relatively lower with LiFSI additive[Fig. 8(c)] consistent with the IR spectrum [Fig. 7(c)], as well asLiF at 685 eV that is consistent with F 1s spectra. Electrically resis-tive LiF might contribute to the increase in the interfacial resistance[Fig. 3(a)]. The surface feature of plated Li metal is vaguely detectedprobably due to its formation in the limited area [Fig. 5(b)]. In P 2pspectra, while OPF3−y(OR)y/LixPFyOz type compounds mainly formupon the use of the LiFSI additive [Fig. 8(c)], phosphates/phosphitesare also equal major species in the base electrolyte [Fig. 8(b)]. Theuse of the LiFSI additive [Fig. 8(c)] induces to form its trace in S 2pand N 1s spectra; a peak of the −−SOy−− bond17,36–38 at 169 eV in theS 2p spectrum and a peak of the −−N−− bond at 400.8 eV39,40 in theN 1s spectrum are attributable to FSI reductive decomposition prod-ucts, consistent with IR observation. The data reveal that while theLi metal-plating-free [Figs. 5(c) and 5(c′)] surface of the cycledgraphite anode with LiFSI additive is covered with a plenty oforganic compounds (i.e., soft matters) along with LiFSI decompo-sition products, inorganic species including Li2CO3 and LiF are ata relatively higher concentration despite lower surface coverage inthe base electrolyte, along with the occurrence of Li metal-plating[Figs. 5(b) and 5(b′)], as illustrated in Scheme 1(a). A possible metal-dissolution from the cathode followed by metal-deposition at theanode surface as a result of high-voltage charge and cathode degra-dation was checked with the Co 2p spectra, but no signature for

the deposited Co species is observed. Given that improved low-temperature cycling performance (Fig. 2) is achieved by the use ofthe LiFSI additive, the effective surface passivation of the graphiteanode with the LiFSI-derived stable and less resistive and softer SEIlayer is believed to be the key in inhibiting the Li metal-plating.

2. SEI characterization of LiCoO2 cathodesATR FTIR spectra of pristine and cycled LiCoO2 cathodes of

the pouch cell operated under −20 ○C are compared in Fig. 9. Pris-tine [Fig. 9(a)] exhibits major peaks at 1400 cm−1, 1170 cm−1, and1072 cm−1, characteristic of the PVdF binder.30–32,34 The spectra ofcycled cathodes [Figs. 9(b) and 9(c)] show in common the peaks at2970−2910 cm−1 and 1470−1410 cm−1 due to methyl (CH3−−) andmethylene (−−CH2−−) groups of alkyl functionality from new sur-face compounds,30,33 in addition to the PVdF binder. Both cycledcathodes present a new peak near 1730−1710 cm−1, characteristicof νsym(C==O) from ester (RCO2R′), which is confirmed with thepresence of ν(C−−O−−C)asym and ν(C−−O−−C)sym at 1280 cm−1 and1170 cm−1.30–34,41 Also, in common, observed at 1650−1640 cm−1and 1580 cm−1 are attributed to alkyl carbonate salt ROCO2M (M= Li/Co) and carboxylate salt RCO2M,30,32,33 which are at a relativelyhigher concentration upon the use of the LiFSI additive [Fig. 9(c)].A tiny peak of Li2CO3 near 1472 cm−1 33,34 is observed on thecycled cathode with the base electrolyte [Fig. 9(b)], whereas it isabsent upon the use of the LiFSI additive. Peaks at 1324 cm−1 and1050 cm−1, attributed to ν(P==O) and ν(P−−O−−C)asym, respectively,of organic phosphorus fluorides POF3−y(OR)y.30,33,35 A strong peakat 931 cm−1 together with those at 823 and 798−744 cm−1 is bythe presence of ν(P−−F) from different PF-containing species suchas LixPFy, LixPOFy, and organic phosphorus fluorides, which are ata relatively higher concentration in the base electrolyte [Fig. 9(b)],similar to the case of the graphite anode. The trace of ν(S==O)of (RO)2SO and −−SO2N−− functionalities of LiFSI decompositionproducts [Fig. 9(c)] at 1210 cm−1 and 1170 cm−1 34 is observed forthe cycled cathode with LiFSI additive.

The IR analysis results of cycled LiCoO2 cathodes reveal thatwhile the LiFSI additive-containing electrolyte induces to formorganic compounds at a relatively higher concentration along withLiFSI decomposition products as the SEI species, lower concen-tration of organic compounds and higher concentration of LiPF6-derived inorganic compounds are present at the surface of thecathode cycled in the base electrolyte.

Figures 10(a)–10(c) compare the XPS spectra of pristine andcycled LiCoO2 cathodes without and with LiFSI additive. The Li

SCHEME 1. Schematic representation ofinterfacial phenomena occurring in thegraphite∥LiCoO2 pouch cells between3.0 V and 4.3 V at 0.2C under −20 ○Cin Base EL of 1M LiPF6/EC:EMC (a)without and (b) with 0.2M LiFSI additive.

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-8




FIG. 9. ATR FTIR spectral comparison of (a) pristine and cycled LiCoO2 cathodesin graphite∥LiCoO2 pouch cells between 3.0 V and 4.3 V at 0.2C under −20 ○C inBase EL of 1M LiPF6/EC:EMC (b) without and (c) with 0.2M LiFSI additive.

1s spectrum of the pristine cathode [Fig. 10(a)] shows the peakof lattice Li−−O but after cycling; the cathode surface is fully cov-ered with newly produced Li-containing surface compounds despiteweak peak intensity. The Co 2p spectrum of pristine [Fig. 10(a)]presents two peaks at ≈780 eV and ≈795 eV due to the 2p3/2 and2p1/2 of Co−−O bonds, respectively.24,42 With cycling, peaks becomeunclear to analyze, probably due to the full surface coverage with

electrolyte decomposition products. The O 1s spectrum of pristine[Fig. 10(a)] has two peaks, characteristic of lattice oxygen (Co−−O) at530 eV and surface oxygen at 531.7 eV.24,39,40,42 Although both peakssignificantly weaken with cycling in both electrolytes at the growthof new strong and broad peaks of C−−O/P−−O (533.1–533.7 eV) andC==O (532.3–532.6 eV) [Figs. 10(b) and 10(c)]36,39,43 of new sur-face species, the peak of lattice oxygen is better maintained uponthe use of the LiFSI additive, implying improved structural mainte-nance of the LiCoO2 cathode. In the C 1s spectra, new peaks of C−−O(286.2 eV), C==O (288.3 eV), and CO3 (290 eV)30,32,36 are observedon both cycled LiCoO2 cathodes [Figs. 10(b) and 10(c)], besides thepeaks of C−−H/C−−C and C−−F (PVdF) of pristine [Fig. 10(a)]. Lit-tle difference is seen, regardless of the use of the LiFSI additive. Aclear difference without and with LiFSI additive is shown in theP 2p spectra; LiPF6-derived organic phosphorus fluoride/LixPOyFzcompounds and LiF are present at both cycled cathodes, but thoseform at a relatively higher concentration in the base electrolyte[Fig. 10(b)], consistent with the F 1s spectrum and IR data. In the F1s spectra, both cycled cathodes show new peaks near 686.7 eV and684.9 eV, attributable to OPF3−y(OR)y/LixPFyOz and LiF, respec-tively, in addition to the peak of the PVdF binder at 688.2 eV ofpristine.30,32,39,40 The S 2p and N 1s spectra exhibit features of LiFSIdecomposition products on the cathode cycled with LiFSI additive[Figs. 10(c)],17,37–40 similar to the case of the graphite anode.

XPS analysis results on cycled cathodes indicate that while theSEI layer formed in the base electrolyte includes a relatively higher

FIG. 10. XPS spectral comparison of surface elements of (a) pristine and cycled cathodes in graphite∥LiCoO2 pouch cells between 3.0 V and 4.3 V at 0.2C under −20 ○C inBase EL of 1M LiPF6/EC:EMC (b) without and (c) with 0.2M LiFSI additive.

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-9




concentration of OPF3−y(OR)y/LixPFyOz and LiF, LiFSI decompo-sition products are additionally present as the SEI species at thesurface of the LiCoO2 cathode cycled with LiFSI additive. Recollect-ing the earlier oxidative decomposition of LiFSI than LiPF6 that wasobserved in LSV [Fig. 1(b)], the LiFSI additive might undergo earliersacrificial decomposition in the early cycles forming the relativelysofter SEI layer that effectively passivates the surface of both LiCoO2cathode and graphite anode [Scheme 1(b)]. Thus, reduced chance ofLiPF6 decomposition and release of HF and gases5 are given to thepouch cell cycled with LiFSI additive under −20 ○C.

IV. CONCLUSIONSThe effects and roles of the LiFSI additive in promoting the

cycling performance of graphite∥LiCoO2 Li-ion pouch cells between3.0 V and 4.3 V at 0.2C under −20 ○C are investigated. System-atic surface chemistry studies reveal how the LiFSI additive influ-ences the composition, surface coverage, and stability of the SEIlayer formed at the graphite anode and LiCoO2 cathode. As theLiFSI additive undergoes earlier sacrificial decomposition duringearly cycles forming the surface-passivating SEI layer at both LiCoO2cathode and graphite anode, a reduced chance of LiPF6 decomposi-tion is given to the pouch cell cycled under −20 ○C. A robust, lessresistive, and softer SEI layer, which is enriched with organic com-pounds formed by carbonates, solvent decomposition along LiFSI-derived inorganic compounds, and reduced concentration of LiPF6decomposition products, contributes to an effective passivation ofthe surface of both cathode and anode, leading to the inhibitionof Li metal-plating at the graphite anode. The resultant data indi-cate that enhanced interfacial stability via the LiFSI additive plays akey role in dramatically increasing the capacity and promoting thecycling performance and rate capability at the reduced interfacialresistance despite under low temperature of−20 ○C and high-voltagecharge cutoff voltage of 4.3 V. We believe that the design and con-trol of interfacial chemistry of the anode–electrolyte and cathode–electrolyte in LIBs operated at low temperatures based on a basicunderstanding of interfacial phenomena are essential in enablinghigh battery performance and safety.

ACKNOWLEDGMENTSThis research was supported by the Korean Ministry of Trade,

Industry and Energy (Grant Nos. A0022-00725 and R0004645) andthe National Research Foundation grant funded by the KoreanMinistry of Science and ICT (Grant No. 2019R1A2C1084024).

REFERENCES1Y. Ji, Y. Zhang, and C.-Y. Wang, J. Electrochem. Soc. 160, A636 (2013).2S. S. Zhang, K. Xu, and T. R. Jow, Electrochem. Commun. 4, 928 (2002).3L. Liao, X. Cheng, Y. Ma, P. Zuo, W. Fang, G. Yin, and Y. Gao, Electrochim. Acta87, 466 (2013).4S. S. Zhang, K. Xu, and T. R. Jow, J. Power Sources 115, 137 (2003).5G. G. Eshetu, S. Grugeon, G. Gachot, D. Mathiron, M. Armand, and S. Laruelle,Electrochim. Acta 102, 133 (2013).6M. C. Smart, J. Electrochem. Soc. 146, 486 (1999).7M. C. Smart, B. V. Ratnakumar, V. S. Ryan-Mowrey, S. Surampudi, G. K.S. Prakash, J. Hu, and I. Cheung, J. Power Sources 119-121, 359 (2003).8M. C. Smart, B. V. Ratnakumar, and S. Surampudi, J. Electrochem. Soc. 149,A361 (2002).

9M. C. Smart, B. L. Lucht, S. Dalavi, F. C. Krause, and B. V. Ratnakumar,J. Electrochem. Soc. 159, A739 (2012).10S. S. Zhang, J. Power Sources 163, 713 (2007).11S. Li, X. Li, J. Liu, Z. Shang, and X. Cui, Ionics 21, 901 (2015).12D. Y. Wang, A. Xiao, L. Wells, and J. R. Dahn, J. Electrochem. Soc. 162, A169(2014).13H. B. Han, S. S. Zhou, D. J. Zhang, S. W. Feng, L. F. Li, K. Liu, W. F. Feng,J. Nie, H. Li, X. J. Huang, M. Armand, and Z. Bin Zhou, J. Power Sources 196,3623 (2011).14A. Guerfi, S. Duchesne, Y. Kobayashi, A. Vijh, and K. Zaghib, J. Power Sources175, 866 (2008).15G. G. Eshetu, T. Diemant, S. Grugeon, R. J. Behm, S. Laruelle, M. Armand, andS. Passerini, ACS Appl. Mater. Interfaces 8, 16087 (2016).16M. Nie and B. L. Lucht, J. Electrochem. Soc. 161, A1001 (2014).17B. Philippe, R. Dedryvère, M. Gorgoi, H. Rensmo, D. Gonbeau, and K. Edström,J. Am. Chem. Soc. 135, 9829 (2013).18K. Xu, Chem. Rev. 114, 11503 (2014).19A. Nyman, M. Behm, and G. Lindbergh, Electrochim. Acta 53, 6356 (2008).20M. S. Ding, K. Xu, S. S. Zhang, K. Amine, G. L. Henriksen, and T. R. Jow,J. Electrochem. Soc. 148, A1196 (2001).21M. V. Reddy, T. W. Jie, C. J. Jafta, K. I. Ozoemena, M. K. Mathe, A. S. Nair,S. S. Peng, M. S. Idris, G. Balakrishna, F. I. Ezema, and B. V. R. Chowdari,Electrochim. Acta 128, 192 (2014).22K. S. Tan, M. V. Reddy, G. V. S. Rao, and B. V. R. Chowdari, J. Power Sources147, 241 (2005).23J. N. Reimers and J. R. Dahn, J. Electrochem. Soc. 139, 2091 (1992).24Y.-C. Lu, A. N. Mansour, N. Yabuuchi, and Y. Shao-Horn, Chem. Mater. 21,4408 (2009).25B. Wu, J. Wang, J. Li, W. Lin, H. Hu, F. Wang, S. Zhao, C. Gan, and J. Zhao,Electrochim. Acta 209, 315 (2016).26X. Zuo, C. Fan, X. Xiao, J. Liu, and J. Nan, J. Power Sources 219, 94 (2012).27M. Zhao, X. Zuo, X. Ma, X. Xiao, L. Yu, and J. Nan, J. Power Sources 323, 29(2016).28T. Ohzuku and A. Ueda, J. Electrochem. Soc. 141, 2972 (1994).29L. Liu, L. Chen, X. Huang, X. Q. Yang, W.-S. Yoon, H. S. Lee, and J. McBreen,J. Electrochem. Soc. 151, A1344 (2004).30H. Q. Pham, K.-M. Nam, E.-H. Hwang, Y.-G. Kwon, H. M. Jung, and S.-W.Song, J. Electrochem. Soc. 161, A2002 (2014).31J.-W. Song, C. C. Nguyen, H. Choi, K.-H. Lee, K.-H. Han, Y.-J. Kim, S. Choy,and S.-W. Song, J. Electrochem. Soc. 158, A458 (2011).32H. Q. Pham, E.-H. Hwang, Y.-G. Kwon, and S.-W. Song, J. Power Sources 323,220 (2016).33Y.-M. Lee, K.-M. Nam, E.-H. Hwang, Y.-G. Kwon, D.-H. Kang, S.-S. Kim, andS.-W. Song, J. Phys. Chem. C 118, 10631 (2014).34G. Socrates, Infrared Characteristic Group Frequencies, Tables and Charts,2nd ed. (John Wiley & Sons, New York, 1994).35H. Q. Pham, B. J. Kim, H. Jo, S. Kang, S.-J. You, and S.-W. Song, J. Electrochem.Soc. 164, A3045 (2017).36R. A. Quinlan, Y.-C. Lu, Y. Shao-Horn, and A. N. Mansour, J. Electrochem. Soc.160, A669 (2013).37X. Zheng, T. Huang, Y. Pan, W. Wang, G. Fang, K. Ding, and M. Wu, J. PowerSources 319, 116 (2016).38X. Zheng, T. Huang, Y. Pan, W. Wang, G. Fang, and M. Wu, J. Power Sources293, 196 (2015).39See http://Srdata.Nist.Gov/Xps/VC for NIST X-Ray Photoelectron Spec-troscopy Database, 2003 Copyright (last accessed 2019).40C. D. Wagner, W. M. Riggs, L. E. Davis, and J. F. Moulder, Handbook ofX-Ray Photoelectron Spectroscopy, edited by G. E. Muilenberg (Perkin-ElmerCorp, Minnesota, 1979).41S.-W. Song, G. V. Zhuang, and P. N. Ross, J. Electrochem. Soc. 151, A1162(2004).42L. Dahéron, H. Martinez, R. Dedryvére, I. Baraille, M. Ménétrier, C. Denage,C. Delmas, and D. Gonbeau, J. Phys. Chem. C 113, 5843 (2009).43L. Baggetto, N. J. Dudney, and G. M. Veith, Electrochim. Acta 90, 135 (2013).

J. Chem. Phys. 152, 094709 (2020); doi: 10.1063/1.5144280 152, 094709-10

https://scitation.org/journal/jcphttps://doi.org/10.1149/2.047304jeshttps://doi.org/10.1016/s1388-2481(02)00490-3https://doi.org/10.1016/j.electacta.2012.09.083https://doi.org/10.1016/s0378-7753(02)00618-3https://doi.org/10.1016/j.electacta.2013.03.171https://doi.org/10.1149/1.1391633https://doi.org/10.1016/s0378-7753(03)00266-0https://doi.org/10.1149/1.1453407https://doi.org/10.1149/2.058206jeshttps://doi.org/10.1016/j.jpowsour.2006.09.040https://doi.org/10.1007/s11581-014-1275-0https://doi.org/10.1149/2.0821501jeshttps://doi.org/10.1016/j.jpowsour.2010.12.040https://doi.org/10.1016/j.jpowsour.2007.09.030https://doi.org/10.1021/acsami.6b04406https://doi.org/10.1149/2.054406jeshttps://doi.org/10.1021/ja403082shttps://doi.org/10.1021/cr500003whttps://doi.org/10.1016/j.electacta.2008.04.023https://doi.org/10.1149/1.1403730https://doi.org/10.1016/j.electacta.2013.10.192https://doi.org/10.1016/j.jpowsour.2005.01.019https://doi.org/10.1149/1.2221184https://doi.org/10.1021/cm900862vhttps://doi.org/10.1016/j.electacta.2016.05.085https://doi.org/10.1016/j.jpowsour.2012.07.026https://doi.org/10.1016/j.jpowsour.2016.05.052https://doi.org/10.1149/1.2059267https://doi.org/10.1149/1.1772781https://doi.org/10.1149/2.1141412jeshttps://doi.org/10.1149/1.3556638https://doi.org/10.1016/j.jpowsour.2016.05.038https://doi.org/10.1021/jp501670ghttps://doi.org/10.1149/2.0601713jeshttps://doi.org/10.1149/2.0601713jeshttps://doi.org/10.1149/2.069304jeshttps://doi.org/10.1016/j.jpowsour.2016.04.053https://doi.org/10.1016/j.jpowsour.2016.04.053https://doi.org/10.1016/j.jpowsour.2015.05.061http://Srdata.Nist.Gov/Xps/VChttps://doi.org/10.1149/1.1763771https://doi.org/10.1021/jp803266whttps://doi.org/10.1016/j.electacta.2012.11.120

Interface stabilization via lithium for promoted performance of ...30852...2 cathodes and cycled...

Documents

Transcript of Interface stabilization via lithium for promoted performance of ...30852...2 cathodes and cycled...