Chemistry Letters Vol.37, No.3 (2008)
369
Figure 1b compares the variations of ꢁ and ꢂ of EFEC and
DEC single solutions with respect to the molar concentration of
100
90
4
2
0
EFEC
DEC
LiPF6 (cLiPF ) at 25 ꢁC. ꢁ increases to some extent with increas-
6
ing the concentration, because the number of charge carrier per
unit volume increases. The ꢂ and mass density of the electrolytic
solution increase rapidly with increasing the concentration.
Therefore, ꢁ adversely decreases in high-concentration ranges.
ꢁ of the EFEC solution was about the same as that of the
DEC counterpart even at ca. 1 mol dmꢂ3 of LiPF6. The finding
indicates that high permittivity rather than high viscosity is a
dominant factor governing ꢁ of the EFEC solution below
1 mol dmꢂ3 of LiPF6 and that the high viscosity plays a major
role in determining ꢁ at higher concentrations. The threshold
concentration (1 mol dmꢂ3), below which monofluorination
also caused the increase in ꢁ, was higher than those observed
for FMMC/DMC (ca. 0.3 mol dmꢂ3) and FEMC/EMC (ca.
0.7 mol dmꢂ3). Thus, ꢁ of the EFEC solutions is higher than that
of the DEC counterparts over wide temperature and concentra-
tion ranges.
The introduction of a fluorine atom into a DEC molecule
increases the electric dipole moment but may decrease the
electron-pair donicity of oxygen atoms in the –OCOO– moiety.
The low electron-pair donicity results in the decreased solvation
to Liþ and, consequently, in the decreased degree of ionic
dissociation. Accordingly, both the ꢁ of the EFEC solutions at
high concentrations and the solubility of the electrolyte become
lower than in the cases of DEC. It is known that lithium salts
form ion-pair dimers and quadrupoles in DEC and DMC.1a Such
noncharged species may also be formed in EFEC.
LSV was carried out to investigate electrochemical voltage
windows of chain carbonate solutions. Figure 2a shows linear
potential sweep voltammograms obtained with a Pt electrode
for EFEC and DEC single solutions at a scan rate of 5 mV sꢂ1
at 25 ꢁC. These solutions contain 1 mol dmꢂ3 LiPF6. The
anodic stability of EFEC was higher than that of DEC, while
the reductive-decomposition potentials were uncertain. A
shoulder peak observed at ca. 5.8 V can be ascribed to trace
amounts of degradation products of LiPF6.
An alkaline metal ion (Lewis acid) such as Liþ has an
influence on the reduction of organic compounds in aprotic
solvents.7 The metal ion forms ion-pairs with reduced organic
compounds, and the formation shifts peaks to positive potentials
to greater or lesser extents. This influence is known to be pro-
nounced in protophobic solvents such as propylene carbonate
and acetonitrile. The reduction of chain carbonates can also in-
volve the formation of the ion-pairs and their decomposition
on a lithium anode.
80
EC–EFEC
EC–DEC
(a)
(b)
70
0
2
4
6
0
10
20
30
40
50
E / V vs. Li | Li+
Cycle number
Figure 2. (a) Linear potential sweep voltammograms obtained
with a Pt electrode for EFEC and DEC single solutions at a scan
rate of 5 mV sꢂ1 at 25 ꢁC. These electrolytic solutions contain
1 mol dmꢂ3 LiPF6. (b) Evolution of cycling efficiency of Li/
LiCoO2 coin cells at 25 ꢁC with respect to the cycle number.
Electrolytic solutions: EC–EFEC and EC–DEC equimolar bina-
ry solutions containing 1 mol dmꢂ3 LiPF6. The coin cells were
charged in a constant current (0.5 C) mode and further charged
to keep constant voltage (4.2 V). Total charging time was set
to 5 h. The coin cells were then discharged to 3.0 V at the
constant current (0.5 C).
The use of the EC–EFEC binary solution improved the
cycling efficiency over the range of high cycle numbers.
Degradation products of the EC–EFEC binary solution by
reduction on a lithium anode can form a passive film containing
fluorine compounds. Although the components have not
been identified, they would show low electron-pair donicity.
Therefore, Liþ ion may readily be permeable into the surface
film. The use of EFEC can decrease the interfacial resistance
between the electrode and the electrolytic solution.
In conclusion, EFEC is a candidate for prominent co-
solvents for lithium batteries.
References and Notes
1
a) L. A. Dominey, in Lithium Batteries, ed. by G. Pistoia,
Elsevier, Amsterdam, 1994, Chap. 4. b) D. Aurbach, A.
Schechter, in Lithium Batteries, Science and Technology,
ed. by G.-A. Nazri, G. Pistoia, Kluwer Academic Publishers,
Boston, 2004, Chap. 18.
M. C. Smart, B. V. Ratnakumar, V. S. Ryan-Mowrey, S.
Surampudi, G. K. S. Prakash, J. Hu, I. Cheung, J. Power
a) M. Takehara, S. Watanabe, N. Nanbu, M. Ue, Y. Sasaki,
1205. c) N. Nanbu, M. Takehara, S. Watanabe, M. Ue, Y.
2
3
We assembled 2025-type coin cells to evaluate the perform-
ance by a charge–discharge test. Figure 2b shows the evolution
of cycling efficiency with respect to the cycle number at 25 ꢁC.
The cycling efficiency stands for the capacity ratio of discharge
to charge. The cycling efficiencies observed were 98% and
95–96% for EC–EFEC and EC–DEC equimolar binary solu-
tions, respectively, after the 2nd cycle. The cycling efficiency
at the first charge/discharge cycle was somewhat lower than that
above the 2nd cycle, and the irreversible capacity was observed
at the first cycle. This is because the electric charge is consumed
for forming the solid electrolyte interphase (SEI). The irreversi-
ble capacity in the EC–EFEC system was lower than that in the
EC–DEC system.
4
5
M. Takehara, N. Tsukimori, N. Nanbu, M. Ue, Y. Sasaki,
Electrochemistry 2003, 71, 1201.
GC-MS (EI) (m=z): 45 CH3CH2Oþ (60), 47 CH2FCH2
þ
(69), 63 CH2FCH2Oþ (23), 91 CH2FCH2OCOþ (11).
1H NMR (CDCl3, 500.00 MHz): ꢃ 1.60 (t, 3H), 4.51 (q,
2H), 4.66 (m, 2H), 4.93 (m, 2H). 13C NMR (CDCl3,
125.65 MHz): ꢃ 13.69 (s), 64.10 (s), 66.72 (s), 81.63 (d,
1JCF ¼ 167 Hz), 155.23 (s).
6
7
M. Takehara, R. Ebara, N. Nanbu, M. Ue, Y. Sasaki,
Electrochemistry 2003, 71, 1172.
K. Izutsu, in Electrochemistry in Nonaqueous Solutioins,
Wiley-VCH, Weinheim, 2002, Chap. 8.
Tsukimori, Naoko
