Structural and functional analysis of surface film on li anode in vinylene carbonate-containing electrolyte

Article abstract of DOI:10.1149/1.1798411

The lithium cycling efficiencies of the lithium anode in the ethylene carbonate (EC)-based electrolytes were improved by adding vinylene carbonate (VC) to the electrolyte. We analyzed the surface films of deposited lithium on a nickel substrate in a VC-containing electrolyte with scanning electron microscopy, Fourier transform infrared spectroscopy, two-dimensional nuclear magnetic resonance, gel permeation chromatography, and X-ray photoelectron spectroscopy. The corresponding surface films comprise various polymeric species including poly-(vinylene carbonate) [poly-(VC)], oligomeric VC, and a ring-opened polymer of VC. Furthermore, the surface film of carbon double bonds (C = C-O) and lithium carboxylate (RCOOLi) as reduction products of VC were formed on deposited lithium. These structures of the surface film on the lithium anode were similar to those on the graphite anode. At elevated temperatures, the VC-containing electrolyte led to the formation of surface films comprising poly(VC). The VC-derived polymeric surface film, which exhibited gel-like morphology, could prevent the deleterious reaction which occurs between deposited lithium and the electrolyte, resulting in an enhanced lithium cycling efficiency. 2004 The Electrochemical Society.

Full text of DOI:10.1149/1.1798411

A1778
Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
0013-4651/2004/151͑11͒/A1778/11/$7.00 © The Electrochemical Society, Inc.
Structural and Functional Analysis of Surface Film on Li
Anode in Vinylene Carbonate-Containing Electrolyte
Hitoshi Ota,a, ,z Yuuichi Sakata,a Yumiko Otake,a Kunihisa Shima,a Makoto Ue,a,
*
*
and Jun-ichi Yamakib,
*
aMitsubishi Chemical Group Science and Technology Research Center, Incorporated, Ibaraki 300-0332,
Japan
bInstitute for Materials Chemistry and Engineering, Kyushu University, Kasuga 816-8580, Japan
The lithium cycling efficiencies of the lithium anode in the ethylene carbonate ͑EC͒-based electrolytes were improved by adding
vinylene carbonate ͑VC͒ to the electrolyte. We analyzed the surface films of deposited lithium on a nickel substrate in a
VC-containing electrolyte with scanning electron microscopy, Fourier transform infrared spectroscopy, two-dimensional nuclear
magnetic resonance, gel permeation chromatography, and X-ray photoelectron spectroscopy. The corresponding surface films
comprise various polymeric species including poly-͑vinylene carbonate͒ ͓poly-͑VC͔͒, oligomeric VC, and a ring-opened polymer
of VC. Furthermore, the surface film of carbon double bonds (C ϭ C-O) and lithium carboxylate ͑RCOOLi͒ as reduction products
of VC were formed on deposited lithium. These structures of the surface film on the lithium anode were similar to those on the
graphite anode. At elevated temperatures, the VC-containing electrolyte led to the formation of surface films comprising poly-
͑VC͒. The VC-derived polymeric surface film, which exhibited gel-like morphology, could prevent the deleterious reaction which
occurs between deposited lithium and the electrolyte, resulting in an enhanced lithium cycling efficiency.
© 2004 The Electrochemical Society. ͓DOI: 10.1149/1.1798411͔ All rights reserved.
Manuscript submitted September 8, 2003; revised manuscript April 29, 2004. Available electronically October 4, 2004.
A large number of studies have been conducted on rechargeable
since they adhere to the lithium metal anode, resulting in an im-
proved lithium cycling efficiency.17 Fluoroethylene carbonate has
also been proposed as an additive for improving the lithium cycling
performance of the lithium anode.28 Additives such as ethylene
sulfite and VC, which are known to prevent the decomposition of
propylene carbonate ͑PC͒ on graphite anodes, have so far shown no
improvements in the cycling performance.18 The addition of VC to
the electrolytes has been proposed in order to improve the perfor-
mance of lithium batteries which use graphite anodes.27 It is known
that treating the electrodes to a stacked pressure also influences the
lithium cycling performance of the lithium anode.28 Moreover, the
lithium cycling efficiencies are influenced by the temperature con-
ditions during deposition and dissolution.29-32
Recently, we reported that the lithium cycling efficiency for de-
posited lithium on nickel substrates in VC-containing electrolytes
can be improved if the electrodes are pressurized.33 Lithium cycling
efficiencies in EC ϩ dimethyl carbonate ͑DMC͒ ͑1:1͒ electrolytes
containing either LiPF6 , LiBF4 , LiN͑SO2CF3)2 ͑LiTFSI͒, or Li-
BETI at 25 and 50°C increased following the addition of VC. This is
in stark contrast to the same reactions at lower temperatures ͑0°C͒,
where the deposited lithium exhibits a thick dendrite-like morphol-
ogy on the anode surface, causing a high film resistance, and ulti-
mately a decrease in the lithium cycling efficiency. At elevated tem-
peratures ͑50°C͒, the surface films are much thinner than those
formed at 0°C, and exhibit a particulate morphology.
Generally, the electrochemical behavior of lithium electrodes in
organic electrolytes is controlled by the surface chemistry of the
lithium metal. The reduction of electrolyte components on the
lithium surface results in the formation of insoluble species, which
influence the overall effectiveness of the lithium batteries. Conse-
quently, surface characterization of the lithium electrodes plays an
important role in determining the main factors which effect the cy-
cling efficiency. Aurbach et al. have argued that the solid electrolyte
interphase ͑SEI͒ layers on lithium in an EC-based electrolyte consist
of (CH2OCO2Li)2 as the dominant constituent by Fourier transform
infrared ͑FTIR͒ spectroscopy.34 Moreover, their group investigated
whether various inorganic and organic lithium compounds, such as
lithium carbonate (Li2CO3), lithium oxide (Li2O), lithium hydride
͑LiOH͒, lithium fluoride ͑LiF͒, alkyl carbonate species (ROCO2Li)
to reduce alkyl carbonate, alkoxy species ͑ROLi͒ to reduce esters,
and alkyl carboxylate species (RCO2Li) to esters, are detected in the
surface film. Kanamura et al. reported that SEI layers deposited on
lithium in a nonaqueous electrolyte by adding fluoric acid ͑HF͒ con-
sist of an LiF/Li2O bilayer, based on X-ray photoelectron spectros-
batteries which use lithium metal anodes, because of lithium’s high
theoretical specific capacity ͑3860 Ah kgϪ1͒. The main areas of
research have focused primarily on investigating several key aspects
including low lithium cycling efficiency, the occurrence of deleteri-
ous dendrite-like lithium deposits on the anode, and general safety
concerns.1 These problems have been widely believed to result
from the reaction of freshly deposited lithium with the electrolyte
components.
The use of LiAsF6 as a solute has sparked considerable interest
due to its excellent lithium cycling efficiency.2-6 However, the tox-
icity of LiAsF6 limits the practical application of this material as a
suitable electrolyte. The use of imide salts such as LiN͑SO2C2F5)2
͑LiBETI͒ have been proposed to yield higher lithium cycling effi-
ciencies and a more enhanced cycling performance.7-9 Further, the
use of lithium organoborate with a salicylic ligand has also demon-
strated a certain success in improving the lithium cycling perfor-
mance in the ethylene carbonate ͑EC͒ and dimethoxyethane ͑DME͒
solvent system.10 The ether compounds such as 2-methyl-
tetrahydrofuran ͑2-MeTHF͒,2-5 1,3-dioxolane ͑DN͒,6 tetrahydropy-
ran ͑THP͒,8 and 1,3-dioxane ͑DOX͒,9 which have good reductive
stability, are effective solvents for the lithium metal rechargeable
batteries. Another key area of research generating much interest
concerns the incorporation of additives into the electrolyte. Many
attempts have proposed that the addition of inorganic
compounds,11-16 organic compounds,11,17,18 surfactants,19-21 and
gaseous compounds22-25 to the electrolyte improves the lithium cy-
cling efficiency and cycling performance of the lithium metal re-
chargeable batteries. Inorganic additives such as AlI3 result in a
Li-Al alloy interface with a high conductivity.13 The addition of
hydrofluoric acid ͑HF͒ to electrolytes results in a smoothly depos-
ited lithium morphology on the anode surface and a much enhanced
cycling performance.14-16,26 The presence of dissolved CO2 within
the electrolyte has also proven effective for enhancing the cycling
performance of lithium batteries,22-25 forming a Li2CO3 film on the
lithium anode.23 The incorporation of organic additives such as
2-methylfuran within the electrolyte also results in an enhanced cy-
cling performance due to the surface film formation on reaction with
the deposited lithium.17 The addition of aromatic additives such as
benzene and toluene induces an isolation effect with the electrolyte
* Electrochemical Society Active Member.
z E-mail: ota.hitoshi@mp.m-kagaku.co.jp
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
A1779
(CH2OCO2Li)2 , polyethylene oxide, and lithium ethylene dicarbon-
ate containing an oxyethylene, based on two-dimensional nuclear
magnetic resonance ͑2D NMR͒ spectroscopy.35
In this study, we examined the structure and function of the sur-
face films of deposited lithium on nickel substrates in the VC-
containing electrolytes by scanning electron microscopy ͑SEM͒,
FTIR, two-dimensional nuclear magnetic resonance ͑2D NMR͒,
gel permeation chromatography ͑GPC͒, XPS, and temperature
programmed desorption or decomposition-mass spectrometry
͑TPD-MS͒.
Experimental
Evaluation of lithium cycling efficiency and sample prepara-
tion.—The lithium cycling efficiency was evaluated using sealed
coin-type cells with
a charge-discharge apparatus ͑Nagano,
BTS2004W͒ in a constant temperature oven maintained within
Ϯ0.5°C. Charging and discharging were performed at 0, 25, and
50°C. The cells were prepared in an argon glove box. The working
electrode was a 1.25 cm diam nickel foil ͑99.7%, Nilaco͒. The
counter electrode and the reference electrode were made of lithium
foil ͑Ͼ99.8%, Honjo Metal͒ with the same diameter. The electro-
lytes ͑Sol-Rite, Mitsubishi Chemical Corp.͒ were EC ϩ DMC sol-
vent ͑volume ratio of 1:1͒ containing 1 mol dmϪ3 LiPF6 , as the
solute. The vinylene carbonate ͑VC, Mitsubishi Chemical Corp.͒
content was 2 wt %. Additionally, VC-containing 1 mol dmϪ3 LiPF6
was used as the electrolyte. The VC contained no stabilizer. The
water content in the electrolytes was analyzed to be Ͻ10 ppm by
Karl-Fischer titration. The electrodes were separated by a polyeth-
ylene separator and held together by a stainless steel disk with a
spring. Cells were assembled in an argon-filled glove box. The
lithium deposition and dissolution current density were 0.6 mA/cm2,
and lithium of 0.5 C/cm2 was deposited. The cut-off potential was
1.5 V for the lithium dissolution. The cell was dismantled in the
argon-filled glove box. The analytical samples were transported to
instruments using inactive transfer vessels. All analytical samples,
except those for XPS, were prepared at 0.6 mA/cm2 ͑current den-
sity͒ and 0.5 C/cm2 ͑plating charge͒. Samples for XPS analysis were
prepared at 0.6 mA/cm2 ͑current density͒ and 1.6 C/cm2 ͑plating
charge͒.
Evolved gas analysis.—The evolved gas in the cell was analyzed
by gas chromatography ͑GC͒. Separation of the degradation prod-
ucts in the evolved gases was obtained using a plot-Q column and
molecular sieve column. The GC column was programmed at the
constant temperature of 70°C.
Microscopic observations.—After the tenth charge-discharge
test, the working electrode was removed from the coin cell and
rinsed by highly purified dimethoxyethane ͑F-DME, battery grade,
Mitsubishi Chemical Corp.͒ with a water content below 5 ppm in the
argon-filled glove box in order to remove the electrolyte. After dry-
ing under vacuum for 30 min at room temperature, the working
electrode was moved for surface observation by using a transfer
vessel. Cross-sectional morphology of lithium on the working elec-
trode was observed by using SEM ͑JEOL, JEM-6300F͒.
Nuclear magnetic resonance spectroscopy (NMR).—All NMR
spectra were obtained using a NMR spectrometer ͑JEOL JNM-
1
GSX400͒ operating at 400 MHz for the H nucleus and 100 MHz
for the 13C nucleus. The surface film of deposited lithium on the
nickel substrate after the 50th deposition/dissolution cycle was ex-
tracted with dimethyl sulfoxide (DMSO)-d6 in the argon-filled
glove box. The dissolved solution with DMSO was used as the
NMR measurement solutions. All spectra were recorded at room
temperature, and DMSO was used as the internal standard. Further
details may be found elsewhere.35
Figure 1. Lithium cycling efficiencies for deposition and dissolution of
lithium on Ni anode in ͑a͒ 1 mol dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒, ͑b͒ 1 mol
dmϪ3 LiPF6 /EC ϩ DMC (1:1) ϩ 2 wt % VC, and ͑c͒ 1 mol dmϪ3
LiPF6 /VC at various temperatures: ͑1͒ 0, ͑2͒ 25, and ͑3͒ 50°C.
copy ͑XPS͒.14-16 Recently, Ota et al. reported the organic surface
films of deposited lithium on nickel substrates in the EC-THF elec-
trolyte consist of lithium etoxides, lithium ethylene dicarbonate
Gel permeation chromatography (GPC).—The number and
weight average molecular weights (Mn and Mw) of surface films
dissolved with DMSO were determined using a GPC ͑TOSOH-
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1780
Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
Figure 2. Cross-sectional SEM im-
ages of surface film on Ni anode in ͑1͒
1 mol dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒,
͑2͒ 1 mol dmϪ3 LiPF6 /EC ϩ DMC
͑1:1͒ ϩ 2 wt % VC, and ͑3͒ 1 mol
dmϪ3 LiPF6 /VC at 0, 25, and 50°C.
HLC2000͒ instrument equipped with two PL gel mixed-C columns
and a refractive index ͑RI͒ detector. Samples were eluted with a
dimethylformamide ͑DMF͒ solution containing 10 mmol/L lithium
bromide ͑LiBr͒ at a flow rate of 1.0 mL/min at 50°C. The GPC was
calibrated against the following polyethylene oxide ͑PEO͒ standards
(Mw 460,000; 124,700; 58,400; 29,600; 11840; 6450; 4120; and
1,900͒.
Results and Discussion
Lithium cycling efficiencies and cross-sectional SEM
morphology studies of surface films.—Figure 1 compares the lithium
cycling efficiencies for lithium deposited on a nickel substrate in the
presence of three different electrolytes in LiPF6 /EC ϩ DMC ͑1:1͒,
LiPF6 /EC ϩ DMC (1:1) ϩ 2 wt % VC, and LiPF6 /VC at 0, 25,
and 50°C. In a previous study, we observed that the cycling efficien-
cies of deposited lithium on a nickel substrate in an EC ϩ DMC
͑1:1͒ electrolyte containing either LiPF6 , LiBF4 , LiTFSI, or Li-
BETI at both 25 and 50°C, improved in the presence of VC when
the electrodes were pressurized.33 This was in stark contrast to the
results obtained at 0°C, where the lithium cycling efficiencies
showed no improvement on addition of VC. In this study, we found
that the pressurized electrode cell in the LiPF6 /VC electrolyte ex-
hibited satisfactory cycling efficiencies at low temperature. For the
lithium metal anode, it is common knowledge that the coulombic
efficiency is strongly dependent on the morphology of the deposited
lithium.36 Figure 2 shows cross-sectional SEM images of the surface
film after the tenth lithium dissolution at various temperatures. The-
ses morphologies are dead lithium because SEM observations are
carried out at a lithium dissolution state. In the VC and VC-free
LiPF6 /EC ϩ DMC ͑1:1͒ electrolytes at low temperature, the depos-
ited lithium exhibits a dendrite-like morphology which is much thin-
ner than that obtained at 25°C, resulting in a subsequent decrease in
the lithium cycling efficiency ͑Fig. 1͒. The dendritic lithium can be
isolated from the substrate by conducting repeated deposition and
dissolution.37 The isolated lithium becomes electrochemically inac-
tive. However, the surface film increases when repeating the
deposition/dissolution of lithium due to the high reactivity of iso-
lated lithium. The dendrite-like morphology cause the formation of
thicker films, although it has little or no influence on the initial
lithium cycling efficiency, leading to an overall decrease in the final
cycling efficiency. In the LiPF6 /VC electrolyte at low temperatures,
the deposited lithium exhibits a particulate morphology, which pre-
sumably might result in an improved cycling efficiency. At higher
X-ray photoelectron spectroscopy (XPS).—XPS ͑ULVAC-
PHI5700ci͒ was used to obtain information about the elementary
chemical state and depth profile distribution for the surface film. The
surface film of deposited lithium on the nickel substrate was rinsed
using F-DME. Monochromatic Al-K radiation ͑1486.6 eV͒, which
operated at a power of 350 W ͑14 kV͒, was applied as the X-ray
source. All XPS measurements were done on the sample after the
first deposition/dissolution. Sputtering of the sample surface was
performed with an argon ion gun under an accelerating voltage of 3
kV. All the samples were measured after 0, 1, 5, 15, and 45 min of
Arϩ ion sputtering.
Fourier transform infrared spectroscopy (FTIR).—The organic
components of the surface film of deposited lithium on nickel sub-
strate after the tenth deposition/dissolution cycle were characterized
by FTIR ͑Nicolet Magna500͒ equipped with a TGS detector. Spectra
were obtained using a resolution of 4 cmϪ1 and averaged over 32
scans. The measurement was carried out by the attenuated total re-
flectance ͑ATR͒ method. The surface film was rinsed using F-DME
in the argon-filled glove box.
Temperature programmed desorption or decomposition—Mass
spectrometry.—The thermal stability of the surface film of deposited
lithium on the nickel substrate after the tenth deposition/dissolution
cycle was evaluated by TPD-MS ͑ANELVA AGS-7000͒ equipped
with a quadrupole mass spectrometer. The carrier gas was helium of
60 mL/min flow rate. The heating rate was 10°C/min. The surface
film was rinsed using F-DME in the argon-filled glove box.
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
A1781
Figure 4. Evolved gas analysis after the 50th deposition and dissolution
cycle in 1 mol dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒, 1 mol dmϪ3 LiPF6 /EC
ϩ DMC (1:1) ϩ 2 wt % VC, and 1 mol dmϪ3 LiPF6 /VC at various tem-
perature: ͑1͒ 0, ͑2͒ 25, and ͑3͒ 50°C.
Figure 3. Evolved gas analysis during deposition and dissolution at various
cycles in ͑1͒ 1 mol dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒, ͑2͒ 1 mol dmϪ3
LiPF6 /EC ϩ DMC (1:1) ϩ 2 wt % VC, and ͑3͒ 1 mol dmϪ3 LiPF6 /VC at
25°C.
deposited lithium exhibit low reactivity with the electrolyte because
gel-like morphologies become low surface areas. The surface film in
the VC-containing electrolytes can prevent the deleterious reaction
between deposited lithium and the electrolyte at an elevated
temperature.
temperatures ͑25 and 50°C͒, the cycling efficiencies were typically
greater than those obtained for the same reaction conducted in VC-
free electrolytes. The corresponding surface films ͑at 25 and 50°C͒
obtained from reactions in both LiPF6 /VC and VC-containing
LiPF6 /EC ϩ DMC electrolytes, were markedly thinner than those
formed at lower temperatures and exhibited gel-like morphologies
containing particulate forms, not dendrite-like morphologies. These
morphologies are largely attributed to the surface film of VC-
derived polymeric species. In the VC-containing electrolytes,
Evolved gas analysis during lithium deposition and dissolu-
tion.—It is common knowledge that gaseous products are generated
by electrochemical side reactions between the deposited lithium and
the electrolyte. The evolved gases provide important chemical infor-
mation relating to the surface film formation. Figure 3 shows results
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1782
Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
Figure 5. 1H-NMR spectrum of dissolved solution with DMSO-d6 of sur-
face film of deposited lithium on nickel substrate in
1
mol dmϪ3
LiPF6 /EC ϩ DMC ͑1:1͒ at 25°C.
Figure 7. 2D 1H-13
C HMQC spectrum of dissolved solution with
DMSO-d6 of surface film of deposited lithium on nickel substrate in 1 mol
dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒ at 25°C.
of the evolved gas analysis following lithium deposition and disso-
lution in a Li/Ni cell at 25°C. Electrochemical decomposition of the
solvent results in the formation of ethylene (C2H4), carbon monox-
ide ͑CO͒, methane (CH4), and carbon dioxide (CO2), all of which
were detected in the EC ϩ DMC ͑1:1͒ electrolyte. These gas spe-
cies would be related to the formation of the surface film as in the
following equations38
the anodes are substituted for graphite, the initial cycling efficiency
in EC-based electrolytes is typically around 90%, the value which
improves over a period of time to yield efficiencies much closer to
100%. Unlike lithium anodes, the surface films formed on graphite
can prevent reductive decomposition of the electrolyte. Interestingly,
the presence of VC in EC-based electrolytes was found to reduce the
formation of C2H4 , and CH4 , as shown in Fig. 3-2. However, the
same conditions also resulted in a concomitant increase in the for-
mation of CO2 , with increasing cycling. These results suggest that
the presence of VC decomposition products within the surface film
is enough to suppress the reduction of EC-DMC. These findings are
echoed in Fig. 3-3, where the LiPF6 /VC electrolyte produced a
large amount of CO2 gas and a much smaller volume of CO. Figure
4 compares the evolved gas analysis following deposition and dis-
solution cycling ͑50 times͒ at 0, 25, and 50°C. In the LiPF6 /EC
ϩ DMC (1:1) ϩ 2% VC and LiPF6 /VC electrolytes, the volume
of CO2 increased with increasing temperature. The volume of CO2
formed in electrolytes containing VC additives is quantitatively
larger than in VC-free electrolytes, implying that the evolution of
CO2 is directly related to the electrochemical decomposition of VC.
Similarly, the lithium-graphite half-cell in the LiPF6 /EC ϩ DMC
͑1:1͒ with 2% VC and LiPF6 /VC electrolytes also generated CO2
gas at the anode.38 We conjecture that the quantity of CO2 produced
corresponds directly with the presence of VC decomposition prod-
ucts within the surface film.
2EC ϩ 2eϪ ϩ 2Liϩ
EC ϩ 2eϪ ϩ 2Liϩ
→
CH OCO Li͒ ↓ ϩ C H ↑
͓1͔
͓2͔
͓3͔
͓4͔
͓5͔
͑
2
2
2
2
4
→
CH OLi͒ ↓ ϩ CO↑
͑
2
2
DMC ϩ eϪ ϩ Liϩ → CH3OCO2Li↓ ϩ CH•3
DMC ϩ 2eϪ ϩ 2Liϩ → 2CH3OLi↓ ϩ CO↑
CH•3 ϩ H• → CH4↑
The volume of these gases was found to increase with repeated
deposition and dissolution. For the lithium metal anode, the lithium
cycling efficiency cannot reach 100%. Since the reaction between
the deposited lithium and the electrolytes occurs with cycling, the
volume of reductive gaseous species is also found to increase. When
NMR analysis of the VC-derived surface films.—The 1H, 13C,
2D (1H-13C) HMQC ͑heteronuclear multiple quantum coherence͒
and 2D (1H-13C) HMBC ͑heteronuclear multiple bond coherence͒
NMR spectra of the surface film of deposited lithium on nickel
substrate in LiPF6 /EC ϩ DMC ͑1:1͒ are shown in Fig. 5 to 8. From
the NMR results, we confirmed the presence of five components
within the LiPF6 /EC ϩ DMC ͑1:1͒ electrolyte, and four compo-
nents within the LiPF /VC electrolyte 1H-NMR spectrum ͑Fig. 9͒
͓
6
and 13C-NMR spectrum ͑Fig. 10͔͒. The chemical structures of the
surface films formed in both electrolyte solvent systems are shown
1
in Fig. 11 and 12, respectively. The signals for all H and 13C NMR
spectra have been assigned accordingly and summarized in Table I.
The HMQC technique provides direct ͑one-bond͒ heteronuclear
͑e.g., C-H͒ correlations. In contrast, the HMBC technique allows the
identification of proton-carbon connectivities through multiple bond
couplings ͑typically from two to three bonds͒.
Figure 6. 13C-NMR spectrum of dissolved solution with DMSO-d6 of sur-
face film of deposited lithium on nickel substrate in
LiPF6 /EC ϩ DMC ͑1:1͒ at 25°C.
1
mol dmϪ3
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
A1783
Figure 8. 2D 1H-13C HMBC spec-
trum of dissolved solution with
DMSO-d6 of surface of deposited
lithium on nickel substrate in 1 mol
dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒ at
25°C.
In the 1H NMR spectrum ͑Fig. 5͒, the triplet at 1.06 ppm and
quartet at 3.77 ppm have been assigned to H-1 and H-2 in
CH3CH2OϪ ͑compound 1͒, respectively. The signal at 3.50 ppm has
been assigned to the methylene group -(CH2O)- of polyethylene
oxide ͑PEO, compound 2͒. In the HMQC spectrum ͑Fig. 7͒, the
signal at 3.5 ppm, associated with H-3 -(CH2O)- ͑compound 2͒, has
Figure 9. 1H-NMR spectrum of dissolved solution with DMSO-d6 of sur-
face film formed of deposited lithium on nickel substrate in 1 mol dmϪ3
LiPF6 /VC at 25°C.
Figure 10. 13C-NMR spectrum of dissolved solution with DMSO-d6 of
surface film formed of deposited lithium on nickel substrate in 1 mol dmϪ3
LiPF6 /VC at 25°C.
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1784
Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
Table I. 13C NMR and 1H chemical shifts of surface film.
1H chem shift
13C chem shift
Assignment
͑ppm͒
͑ppm͒
Compound 1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
1.1
3.8
3.5
15.50
59.40
69.90
156.40
63.40
70.00
63.40
156.40
51.70
157.10
75.50
152.20
100.60
70.30
153.70
82.50
79.30
71.30
104.70
98.80
70.30
153.70
153.70
82.40
79.30
71.30
101.90
95.40
70.30
153.70
153.70
Compound 2
Compound 3
3.8
3.5
3.8
Compound 4
Figure 11. The structure of EC-DMC derived surface film obtained from
NMR spectroscopy.
Compound 5
Compound 6
Compound 7
3.3
5.3
been shown to correspond with the ͑C-3͒ carbon signal at 69.9 ppm.
The methylene signal at 3.8 ppm, assigned to the ͑H-5͒ proton of
(CH2OCO2Li)2 , ͑compound 3͒, shows a correlation with the carbon
signal ͑C-5͒ at 63.4 ppm. In the HMBC spectrum ͑Fig. 8͒, the same
methylene protons ͑H-5͒ are associated with the carbon ͑C-5͒ signal
at 63.4 ppm, which implies that compound 3 has a contrastive struc-
ture since the CH2 group exhibits a relationship between H-5 and
C-5. Additionally, the proton ͑H-5͒ is shown to have a correlation
with the carbonate carbon ͑C-4͒ at 156.4 ppm ͑Fig. 8͒. From FTIR
evidence, Aurbach et al. proposed that (CH2OCO2Li)2 was formed
following the reduction of EC.5,34 The NMR results support the
presence of (CH2OCO2Li)2 within the surface film. In the HMQC
spectrum ͑Fig. 7͒, the triplets of protons ͑H-6 and H-7͒ at 3.48
and 3.84 ppm correspond to the carbons ͑C-6 and C-7͒ at 70
and 63.4 ppm, respectively. As well as its association with ͑C-6͒
͑Fig. 8͒, H-6 was also shown to correlate with the carbon ͑156.4
ppm͒ associated with protons H-7 and H-8. On the basis of the
NMR and HMBC results, we conclude that this component ͑com-
pound 4͒ has a carbonate-like structure with ether units within the
chain ͑Fig. 11͒. C-4 and C-8 in compound 4 show the same chemi-
cal shift. Therefore, the number ͑n͒ of compound 4 may be zero for
the EC ϩ DMC electrolyte containing LiPF6 . Compounds 1-4 were
also isolated from EC ϩ THF ͑1:1͒ electrolytes containing LiPF6
and LiBETI following the reaction in the Li/Ni cells.35 In addition,
the same reaction conducted in an EC ϩ DMC ͑1:1͒ electrolyte,
generated compounds 1-4, together with a new species, lithium me-
thyl carbonate (CH3OCO2Li) here after referred to as compound 5.
Compounds 1-4 and 5 were generated following the reduction of EC
and DMC, respectively. The chemical structure of compound 5 was
determined using both HMQC and HMBC. In the HMQC spectrum
͑Fig. 8͒, the proton ͑H-9͒ at 3.32 ppm was found to correlate with
the carbon ͑C-9͒ at 51.7 ppm. In the HMBC spectrum ͑Fig. 10͒,
these same protons ͑H-9͒ were also shown to form a cross peak with
the carbonate carbon ͑C-10͒ at 157.1 ppm. Compound 5 would be
related to the formation of the surface film due to reductive decom-
position of DMC.
5.9
4.22, 4.56
Compound 8
5.2
5.5
4.03, 4.25
5.7
6.1
4.25, 4.63
Compound 9
16
Ј
17
Ј
18
Ј
19
Ј
20
Ј
21
Ј
22
Ј
23
Ј
5.2
5.5
3.96, 4.22
5.6
6.0
4.25, 4.63
lithium on nickel substrate in LiPF6 /VC are comparable with those
previously reported on graphite. The 1H signal ͑H-11͒ at 5.1-5.5
ppm ͑Fig. 9͒, hereafter denoted as poly-͑VC͒, exhibits a character-
istic broadness commonly associated with polymeric species. The
corresponding 13C signals ͑Fig. 10͒ at ͑C-11͒ 75.5 ppm and ͑C-12͒
152.2 ppm have chemical shifts characteristic of the ring structure
͑compound 6͒ depicted in Fig. 11. The 1H signals associated with
͑H-14͒ at 4.22 and 4.56 ppm, and the signal of ͑H-13͒ at 5.89 ppm
were shown to attach directly to the carbons ͑C-14͒ at 70.3 ppm and
͑C-13͒ at 100.6 ppm in the HMQC spectrum. Based on these chemi-
cal shifts, the signals were assigned a partial structure of O-C-O.
The broadening of both proton ͑H-13,H-14͒ and carbon ͑C-13,C-14͒
signals suggests compound 7 is a polymeric species, as depicted in
1
Fig. 12. From the HMQC spectrum, the H signals ͑H-16͒ at 5.19
ppm, ͑H-17͒ 5.45 ppm, ͑H-18͒ 4.03 ppm and 4.25 ppm, ͑H-19͒ 5.70
ppm, ͑H-20͒ 6.06 ppm, and ͑H-21͒ 4.25 ppm and 4.63 ppm, were
found to have a close relationship with the following carbons: ͑C-
16͒ at 82.5 ppm, ͑C-17͒ at 79.3 ppm, ͑C-18͒ at 71.3 ppm, ͑C-19͒ at
104.7 ppm, ͑C-20͒ at 98.8 ppm and ͑C-21͒ at 70.3 ppm. Compounds
8 and 9 have been assigned as structural isomers, since their respec-
Recently, our group reported on the formation of surface films on
graphite anodes in LiPF6 /VC.38 The present 1H ͑Fig. 9͒ and
13C-NMR ͑Fig. 10͒ results suggest the surface films of deposited
1
tive H and 13C signals have chemical shifts that are slightly offset
from those of the corresponding isomer.
It has been known that DMSO is a reactive solvent with lithium.
If surface film formation between DMSO and lithium occurs, CD3
unit in DMSO-d6 would be found by 13C-NMR spectra. However, a
CD3 unit related to DMSO-d6 was not detected as shown in Fig. 6.
This result indicates that there is no mainly DMSO-related com-
pound. Furthermore, we confirmed the reactivity between the
lithium and VC/LiPF6 solution containing DMSO as shown in Fig.
13. Figure 13 demonstrates 1H-NMR spectra of surface film ͑Fig.
13-1͒ in LiPF6 /VC, solution after the 50th deposition/dissolution in
LiPF6 /VC ͑Fig. 13-2͒ and VC/LiPF6 solution containing DMSO
and lithium ͑Fig. 13-3͒. In 1H-NMR spectra in Fig. 13-2 and Fig.
Figure 12. The structure of VC-derived surface film obtained from NMR
spectroscopy.
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
A1785
Table II. The number and weight average molecular weights
„Mn and Mw… of surface film on nickel substrate by GPC. „Cal-
culated above Mw 1000.…
Electrolyte
Mn
Mw
1,600
2,100
3,100
1,900
2,400
6,700
LiPF6 /EC ϩ DMC ͑1:1͒
LiPF6 /EC ϩ DMC (1:1) ϩ 2% VC
LiPF6 /VC
trolyte yielded a molecular weight of Mw 1900 ͑Table II͒. This value
is consistent with the formation of an oligomeric species ͑compound
2 and compound 4 in Fig. 11͒ as previously established in the NMR
studies above. The dissolved solution of the surface film formed in
LiPF6 /VC had a lower retention time compared to that in the VC-
free electrolyte, as shown in Fig. 14͑c͒. This result clearly indicates
that the surface film contains polymeric species as shown in Fig. 12.
For the surface film formed in LiPF6 /EC ϩ DMC ͑1:1͒ in the pres-
ence of 2% VC, the sample elution occurred around 14.8 min (Mw
7,500͒ ͓Fig. 14͑b͔͒. It is sufficiently high to confirm the presence of
polymers within the surface film, although the molecular weight
obtained in the LiPF6 /EC ϩ DMC (1:1) ϩ 2% VC is lower than
that in the LiPF6 /VC electrolyte. The molecular weights of surface
films formed in EC ϩ DMC, EC ϩ DMC ϩ VC, and VC electro-
lytes gave Mw values of 1900, 2400, and 6700, and Mn values of
1600, 2100, and 3100, respectively, confirming that the increased
VC content leads to larger molecular weights. We consider that a
great part of the surface films in the EC-based electrolyte may con-
tain relatively small molecules. The NMR spectra of surface film in
the LiPF6 /VC electrolyte demonstrated polymeric species including
͓poly-͑VC͔͒, oligomeric VC, and a ring-opened polymer of VC.
These polymers would be detected by the GPC analysis.
Figure 13. 1H-NMR spectra of ͑a͒ surface film on nickel substrate in
LiPF6 /VC, ͑b͒ solution after the 50th deposition/dissolution in LiPF6 /VC
and, ͑c͒ solution after reacting lithium with DMSO in LiPF6 /VC for 24 h.
13-3, broad peaks related to the polymer could not be detected.
Hence, our experiment can ignore reactions of nucleophilic surface
species with DMSO. We strongly suggest that these polymers are
formed by the electrochemical process, not an impurity.
GPC molecular weight determination of surface films.—Gel per-
meation chromatography ͑GPC͒ has been used as a chromatographic
technique since the 1960s for polymer analysis, and is in common
use as a fast and reliable means of gathering molecular mass distri-
bution information. Figure 14 shows the GPC chromatograms of the
surface film of deposited lithium on the nickel substrate. In typical
chromatograms, very large molecules move through the column
quickly ͑species with lower retention time͒ and smaller molecules
are retained ͑species of higher retention time͒. When pure DMF
eluents were used, sample elution occurred typically before the ex-
clusion limit due to the presence of an ionic polymer. Consequently,
all molecular weight determination analyses were conducted in
DMF solutions containing 10 mM LiBr, the presence of which pre-
cluded the detection of distributions below Mw 1000. Table II shows
Mn and Mw of surface film on nickel substrate by GPC, where Mn
and Mw values were calculated from distributions greater than Mw
1000. The surface film formed in the LiPF6 /EC ϩ DMC ͑1:1͒ elec-
XPS analysis of surface films.—Figure 15 shows the C 1s and O
1s XPS spectra for the surface films of deposited lithium on a nickel
substrate at 25°C, after the first dissolution step. An XPS study can
obtain surface information of inorganic species in addition to or-
ganic species. For the surface film formed in the EC-DMC electro-
lyte, XPS confirms the existence of lithium alkyl carbonate ͑com-
pound 3 in Fig. 11͒ with C 1s signals occurring at 291 eV (Cគ O3),
288 eV ͑Cគ O͒, and 285.8 eV ͑Cគ CO͒ in the C 1s spectrum and O 1s
signals at 533.5 eV ͑COគ C͒ and 532.5 eV (C ϭ Oគ ). For the surface
film formed in the same electrolyte containing 2% VC, XPS exhibits
additional peaks within the C 1s and O 1s spectra at 534.4 and 291.8
eV, respectively, suggesting the surface film comprises poly-͑VC͒,
as well as a further C 1s signal ͑289 eV͒ corresponding to the pres-
ence of an ester ͑-Cគ OO͒ functionality. These peaks were also ob-
served in the surface film formed in LiPF6 /VC. Based on the NMR
results, the surface film of deposited lithium on nickel substrate in
the EC-based electrolyte have been shown to comprise polymer con-
stituents containing -CH2O- groups, which have been assigned to
those of PEO ͑compound 2 in Fig. 11͒. The XPS C 1s spectrum of
PEO is known to produce a peak at 286.2 eV corresponding to the
-CH2O- groups.35 We estimate from the C 1s spectrum in Fig. 15
that the signal at 286 eV can only be attributed to the presence of
polymer constituents containing -CH2O- groups ͑compound 4 in
Fig. 11͒. The additional peaks at 283.5 and 281.8 eV have been
tentatively assigned to those of carbide species such as RLi and
Li2C,20,21,25 that together with Li2O, exist deep within the surface
film. A comparison of the surface films measured before and after
Arϩ ion sputtering shows that the Li2O layer formed in the VC-
containing EC-DMC electrolyte ͑Fig. 15b͒ is thinner than that
formed in the VC-free electrolyte ͑Fig. 15a͒. It is difficult to assign
XPS spectra after Arϩ ion sputtering, which often causes decompo-
sition of the organic compounds within the sample. A large carbon
peak at 285 eV was observed in the C 1s spectrum of the surface
film formed in LiPF6 /VC, which we have tentatively assigned to
Figure 14. GPC chromatograms of the surface film of deposited lithium on
nickel substrate in ͑a͒ 1 mol dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒, ͑b͒ 1 mol dmϪ3
LiPF6 /EC ϩ DMC ͑1:1͒ ϩ2 wt % VC, and ͑c͒ 1 mol dmϪ3 LiPF6/VC at
25°C.
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1786
Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
Figure 15. C 1s and O 1s XPS spectra
of surface film of deposited lithium on
nickel substrate in ͑a͒ 1 mol dmϪ3
LiPF6 /EC ϩ DMC ͑1:1͒, ͑b͒ 1 mol
dmϪ3 LiPF6 /EC ϩ DMC (1:1) ϩ 2
wt
% VC, and ͑c͒ 1
mol dmϪ3
LiPF6 /VC.
polymeric and oligomeric forms of VC. The XPS measured atomic
concentrations for surface films obtained before Arϩ ion sputtering
are given in Table III. The content of P ͑from LiPF6) was deter-
mined to be the same for surface films formed in all electrolytes.
The F ͑from LiF͒, and Li content, in the VC-derived surface films
are typically lower than those detected in the VC-free surface film
particularly in LiPF6 /VC, which exhibits a considerably low Li
content. On the other hand, the carbon content in the VC-derived
surface films is significantly higher than in the VC-free surface
films. Lithium alkyl carbonate formed in EC-based electrolyte reacts
with HF to form LiF. We believe that the VC-containing electrolyte
suppresses the formation of LiF because VC-derived polymeric sur-
face films are difficult to react with HF.
surface films formed in EC-based electrolytes show peaks attributed
to lithium alkyl carbonate-like species ͑compound 3 and compound
4 in Fig. 11͒ ͓1647 cmϪ1 (␷as CϭO), 1400 cmϪ1 (␦CH
), 1317
,CH
3
2
cmϪ1 (␷s CϭO), 1074 cmϪ1 (␷C-O) and 820 cmϪ1 (␦ϪOCO )]. Aurbach
2
et al. assigned the peaks to be (CH2OCO2Li)2 , which is the major
EC reduction product on Li and Li-C electrodes.5,34 The FTIR spec-
tra of the surface film formed in all electrolytes exhibited a peak at
3780 cmϪ1, which we have assigned to the ␷ stretching mode of
OH
the hydroxyl group in lithium hydroxide ͑LiOH͒. For the surface
films formed in VC-containing electrolytes, FTIR spectra reveal spe-
cific peaks at 1817, 1147, 1080, and 758 cmϪ1, which we have
attributed to the presence of poly-͑VC͒ ͑compound 6 in Fig. 12͒.
The FTIR spectrum of poly-͑VC͒, which was synthesized in order to
identify poly-͑VC͒, plotted in Fig. 16. The FTIR spectra for the
surface film in the VC-containing electrolyte at 50°C correspond
with the spectrum of poly-͑VC͒. Aurbach et al. reported that FTIR
spectrum of graphite electrodes cycled in the VC-containing electro-
FTIR analysis of the VC-derived surface films.—A comparison
of the FTIR spectra of surface films of deposited lithium on nickel
substrate after the tenth lithium deposition-dissolution cycle at 0, 25,
and 50°C, respectively, is shown in Fig. 16. The FTIR spectra of the
lyte at 60°C showed the broad peak around 1850 cmϪ1 39
assigned that the peak may belong to polycarbonate species. Accord-
.
They
ing to FTIR spectrum poly-͑VC͒, it has the peak around 1820 cmϪ1
.
Table III. Atomic concentration of surface film by XPS.
The broad peak around 1850 cmϪ1 may be related to poly-͑VC͒.
Furthermore, the FTIR spectra of VC-derived surface films exhibit
bands that are typical of a carboxylate (COOLi) functional group
Atomic concentration ͑%͒
Electrolyte
Li
C
O
F
P
Ni
Ϫ
Ϫ
͓1580 cmϪ1 (␷ ) and 1413 cmϪ1 (␷ )]. In addition to poly-
CO
CO
23.9 34.2 37.9 3.8 0.2 0.0
10.2 50.3 36.8 2.5 0.2 0.0
0.1 58.0 40.3 1.3 0.2 0.0
LiPF6 /EC ϩ DMC ͑1:1͒
LiPF6 /EC ϩ DMC (1:1) ϩ 2% VC
LiPF6 /VC
2
2
͑VC͒ and RCOOLi, FTIR has also detected evidence of carbon-
carbon double bonds (C ϭ C) ͓1620 cmϪ1 (␷CϭC) and 972 cmϪ1
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
A1787
Figure 17. Thermal stability of organic surface film of deposited lithium on
nickel substrate in ͑a͒ 1 mol dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒, ͑b͒ 1 mol dmϪ3
LiPF6 /EC ϩ DMC (1:1) ϩ 2 wt % VC, and ͑c͒ 1 mol dmϪ3 LiPF6 /VC at
25°C by TPD-MS.
(CH ϭ CHOCO2Li)2 . These compounds were also detected in the
surface film formed on graphite anodes.38 The concentration of
poly͑VC͒ and C ϭ C-O compounds within the surface film in-
creased with increasing temperature ͑Fig. 16-3͒. According to
evolved gas analysis, it was also observed that the VC-containing
electrolytes generate large amounts of CO2 at elevated temperatures,
thought to be related to the formation of C ϭ C-O functional
groups. We consider that CO2 would be related to decarbonation
(-C ϭ C-O-) via the reduction of VC. The reactivity between VC
and deposited lithium is found to increase with increasing tempera-
ture, implying that VC-derived surface films promote effective sur-
face film formation at elevated temperatures.
Thermal stability (TPD-MS) analysis of the VC derived surface
films.—TPD-MS profiles of the surface films are shown in Fig. 17.
The thermal stability of each surface film was evaluated by measur-
ing the decomposition gas pattern of CO2 (m/z 22͒ related to or-
ganic carbonate surface films. The surface film of deposited lithium
on nickel substrate in the LiPF6 /EC ϩ DMC ͑1:1͒ electrolyte was
shown to decompose around 130°C. While the surface films ob-
tained from the reaction in LiPF6 /EC ϩ DMC (1:1) ϩ 2% VC
and LiPF6 /VC decomposed at clearly higher temperatures, 150 and
300°C, respectively. As described in the analytical results above,
surface films comprising decomposed VC polymer components are
found to improve the thermal stability of the organic surface film.
When the thermal stability of the surface film is poor, it is also
expected to occur due to the reformation of the surface film bought
on by thermal decomposition at elevated temperatures. We propose,
therefore, that the lithium cycling performance in the VC-containing
electrolytes at elevated temperatures might be improved in surface
films that exhibit high thermal stabilities.
Figure 16. FTIR spectra of surface film of deposited lithium on nickel sub-
strate in ͑a͒ 1 mol dmϪ3 LiPF6 /EC ϩ DMC ͑1:1͒, ͑b͒ 1 mol dmϪ3
LiPF6 /EC ϩ DMC (1:1) ϩ 2 wt % VC, and ͑c͒ 1 mol dmϪ3 LiPF6 /VC at
various temperatures: ͑1͒ 0, ͑2͒ 25, and ͑3͒ 50°C. FTIR spectra of ͑d͒ are
poly-͑VC͒
Conclusions
We have used various analytical techniques to investigate the
surface of deposited lithium on a nickel substrate in LiPF6 /EC
ϩ DMC ͑1:1͒ without and with 2% VC additive and LiPF6 /VC.
Moreover, the evolved gas formed during the deposition and disso-
lution steps has also been analyzed in order to understand its effect
on surface film formation.
The presence of VC in EC-based electrolytes caused a decrease
in the volume of gases (C2H4 and CH4) produced due to the forma-
tion of the surface film from the EC-DMC solvent. The volume of
gas produced was found to increase with repeated lithium deposition
and dissolution cycling and reaction in VC-containing electrolytes at
(␦CϭC-O)] within the VC-derived surface films. Aurbach et al.
also reported that VC is reduced to species containing -OCO2Li
groups, part of which may also have C ϭ C double bonds.39
We estimate that the structure of the C ϭ C-O unit comprises
reduced VC decomposition products such as lithium vinylene
dicarbonate (CHOCO2Li)2 and lithium divinylene dicarbonate
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1788
Journal of The Electrochemical Society, 151 ͑11͒ A1778-A1788 ͑2004͒
13. M. Ishikawa, M. Morita, and Y. Matsuda, J. Power Sources, 68, 501 ͑1997͒.
14. K. Kanamura, S. Shiraishi, and Z. Takehara, J. Electrochem. Soc., 141, L108
͑1994͒.
15. K. Kanamura, S. Shiraishi, and Z. Takehara, J. Electrochem. Soc., 143, 2187
͑1996͒.
elevated temperatures. The VC-containing electrolyte showed re-
markably high lithium cycling efficiency with increasing tempera-
ture. The following conclusions can be drawn from the various ana-
lytical techniques.
Based on NMR and GPC studies, the VC-derived surface films
were found to consist of the following polymeric species poly-
͑vinylene carbonate͒, poly-͑VC͒, oligomeric VC, and a ring-opened
polymeric form of VC.
XPS and FTIR studies confirmed the presence of carbon-carbon
double bonds (CϭC-O) and carboxylate ͑COOLi͒ functional groups
on the film surface.
The VC-containing LiPF6 electrolytes suppress the formation of
LiF ͑HF͒ due to the presence of polymeric VC decomposition prod-
ucts within the surface film.
The VC-containing electrolytes can easily form VC-derived sur-
face films at elevated temperatures. Deleterious reactions between
lithium and the electrolyte are suppressed by the presence of poly-
meric VC species within the surface film.
16. S. Shiraishi, K. Kanamura, and Z. Takehara, Langmuir, 13, 3542 ͑1997͒.
17. M. Morita, S. Aoki, and Y. Matsuda, Electrochim. Acta, 37, 119 ͑1992͒.
18. R. Mogi, M. Inaba, S.-K. Jeong, Y. Iriyama, T. Abe, and Z. Ogumi, J. Electrochem.
Soc., 149, A1578 ͑2002͒.
19. T. Hirai, I. Yoshimatsu, and Yamaki, J. Electrochem. Soc., 141, 2300 ͑1994͒.
20. M. Mori, Y. Naruoka, K. Naoi, and D. Fauteux, J. Electrochem. Soc., 145, 2340
͑1998͒.
21. K. Naoi, M. Mori, M. Inoue, T. Wakabayashi, and K. Yamauchi, J. Electrochem.
Soc., 147, 813 ͑2000͒.
22. E. Plichta, S. Slane, M. Uchiyama, M. Salomon, D. Chua, W. B. Ebner, and H. W.
Lin, J. Electrochem. Soc., 136, 1865 ͑1989͒.
23. D. Aurbach, Y. Gofer, M. Ben-Zion, and P. Aped, J. Electroanal. Chem., 339, 451
͑1992͒.
24. J. O. Besenhard, M. W. Wagner, M. Winter, A. D. Jannakoudakis, P. D. Jannak-
oudakis, and E. Theodoridou, J. Power Sources, 44, 413 ͑1993͒.
25. T. Osaka, T. Momma, Y. Matsumoto, and Y. Uchida, J. Electrochem. Soc., 144,
1709 ͑1997͒.
26. Z. Takehara, J. Power Sources, 68, 82 ͑1997͒.
Mitsubishi Chemical Group Science and Technology Center, Incorpo-
rated, assisted in meeting the publication costs of this article.
27. C. Jehoulet, P. Biensan, J. M. Bodet, M. Broussely, C. Moteau, and C. Tessier-
Lescourret, Abstract 135, p. 153, The Electrochemical Society and International
Society of Electrochemistry Meeting Abstracts, Vol. 97-2, Paris, France, Aug 31-
Sept 5, 1997.
References
1. J. Yamaki and S. Tobishima, in Handbook of Battery Materials, J. O. Besenhard,
Editor, Part III-3, Wiley-VCH, Weinheim ͑1999͒.
2. K. M. Abraham, J. S. Foos, and J. L. Goldman, J. Electrochem. Soc., 131, 2197
͑1984͒.
3. S. Tobishima and T. Okada, Electrochim. Acta, 30, 1715 ͑1985͒.
4. M. Arakawa, S. Tobishima, T. Hirai, and J. Yamaki, J. Electrochem. Soc., 133, 1527
͑1986͒.
5. D. Aurbach, A. Zaban, Y. Gofer, O. Abramson, and M. Ben-Zion, J. Electrochem.
Soc., 142, 687 ͑1995͒.
6. D. Aurbach, E. Zinigrad, H. Teller, Y. Cohen, G. Salitra, H. Yamin, P. Dan, and E.
Elster, J. Electrochem. Soc., 149, A1267 ͑2002͒.
7. K. Naoi, M. Mori, Y. Naruoka, W. M. Lamanna, and R. Atanasoski, J. Electrochem.
Soc., 146, 462 ͑1999͒.
8. X. Wang, Y. Yasukawa, and S. Mori, J. Electrochem. Soc., 146, 3992 ͑1999͒.
9. X. Wang, E. Yasukawa, and S. Kasuya, J. Electrochem. Soc., 147, 2421 ͑2000͒.
10. Y. Sasaki, M. Handa, K. Kurashima, T. Tonuma, and K. Usami, J. Electrochem.
Soc., 148, A999 ͑2001͒.
28. T. Hirai, I. Yoshimatsu, and J. Yamaki, J. Electrochem. Soc., 141, 611 ͑1994͒.
29. M. Ishikawa, Y. Takaki, M. Morita, and Y. Matsuda, J. Electrochem. Soc., 144, L90
͑1997͒.
30. M. Ishikawa, M. Kanemoto, and M. Morita, J. Power Sources, 81-82, 217 ͑1999͒.
31. R. Mogi, M. Inaba, Y. Iriyama, T. Abe, and Z. Ogumi, J. Electrochem. Soc., 149,
A385 ͑2002͒.
32. R. Mogi, M. Inaba, Y. Iriyama, T. Abe, and Z. Ogumi, J. Power Sources, 108, 163
͑2002͒.
33. H. Ota, K. Shima, M. Ue, and J. Yamaki, Electrochim. Acta, 49, 565 ͑2004͒.
34. D. Aurbach, Nonaqueous Electrochemistry, Marcel Dekker, New York ͑1999͒.
35. H. Ota, Y. Sakata, X. Wang, J. Sasahara, and E. Yasukawa, J. Electrochem. Soc.,
151, A437 ͑2004͒.
36. H. Ota, X. Wang, and E. Yasukawa, J. Electrochem. Soc., 151, A427 ͑2004͒.
37. M. Arakawa, Y. Nemoto, S. Tobishima, M. Ichimura, and J. Yamaki, J. Power
Sources, 43-44, 27 ͑1993͒.
38. H. Ota, Y. Sakata, A. Inoue, and S. Yamaguchi, J. Electrochem. Soc., 151, A1659
͑2004͒
11. Y. Matsuda, J. Power Sources, 43, 1 ͑1993͒.
39. D. Aurbach, K. Gamolsky, B. Markovsky, Y. Gofer, M. Schmidt, and U. Heider,
Electrochim. Acta, 47, 1423 ͑2002͒.
12. M. Ishikawa, S. Yoshitake, M. Morita, and Y. Matsuda, J. Electrochem. Soc., 141,
L159 ͑1994͒.
Downloaded on 2015-02-25 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).