Improving electrochemical properties of room temperature ionic liquid (RTIL) based electrolyte for Li-ion batteries

Article abstract of DOI:10.1016/j.electacta.2010.04.041

Room temperature ionic liquids (RTILs) with high safety characteristic usually have high viscosity and melting point, which is adverse for the application of RTIL-based electrolytes in Li-ion batteries. In this investigation, a promising RTIL, i.e. PP13TFSI consisting of N-methyl-N-propylpiperidinium (PP13) cation and bis(trifluoromethanesulfonyl)imide (TFSI) anion is synthesized. The effect of the content of Li salt in the electrolytes containing PP13TFSI and LiTFSI on the ionic conductivity and cell performance is investigated. The electrolyte of 0.3 mol kg^-1 LiTFSI/PP13TFSI is recommended for its higher lithium transference number and discharge capacity in the LiCoO₂/Li cell than other electrolytes. In addition, it is found that, by introducing 20% diethyl carbonate (DEC) as a co-solvent into pure RTIL electrolyte, the rate capability and low-temperature performance of the LiCoO₂/Li cells are improved obviously, without sacrificing its safety characteristics. It suggests that a component with low viscosity and melting point, i.e. DEC, is necessary to effectively overcome the shortcomings of RTIL for the application in Li-ion batteries.

Full text of DOI:10.1016/j.electacta.2010.04.041

Electrochimica Acta 55 (2010) 5204–5209
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Improving electrochemical properties of room temperature ionic liquid (RTIL)
based electrolyte for Li-ion batteries
H.F. Xianga,b, B. Yina, H. Wanga, H.W. Lina, X.W. Geb, S. Xiec, C.H. Chena,∗
a CAS Key Laboratory of Materials for Energy Conversion, Department of Materials Science and Engineering, University of Science and Technology of China, Anhui, Hefei 230026, China
b Department of Polymer Science and Engineering, University of Science and Technology of China, Anhui, Hefei 230026, China
c Novolyte Technologies, Jiangsu, Suzhou 215123, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Room temperature ionic liquids (RTILs) with high safety characteristic usually have high viscosity and
melting point, which is adverse for the application of RTIL-based electrolytes in Li-ion batteries. In this
investigation, a promising RTIL, i.e. PP13TFSI consisting of N-methyl-N-propylpiperidinium (PP13) cation
and bis(trifluoromethanesulfonyl)imide (TFSI) anion is synthesized. The effect of the content of Li salt in
the electrolytes containing PP13TFSI and LiTFSI on the ionic conductivity and cell performance is investi-
gated. The electrolyte of 0.3 mol kg−1 LiTFSI/PP13TFSI is recommended for its higher lithium transference
number and discharge capacity in the LiCoO2/Li cell than other electrolytes. In addition, it is found that,
by introducing 20% diethyl carbonate (DEC) as a co-solvent into pure RTIL electrolyte, the rate capabil-
ity and low-temperature performance of the LiCoO2/Li cells are improved obviously, without sacrificing
its safety characteristics. It suggests that a component with low viscosity and melting point, i.e. DEC, is
necessary to effectively overcome the shortcomings of RTIL for the application in Li-ion batteries.
© 2010 Elsevier Ltd. All rights reserved.
Received 5 January 2010
Received in revised form 13 April 2010
Accepted 13 April 2010
Available online 18 April 2010
Keywords:
Room temperature ionic liquid
Electrolyte
Safety
Li-ion battery
1. Introduction
Sakaebe proposed a novel RTIL, N-methyl-N-propylpipe-
ridinium bis(trifluoromethanesulfonyl)imide (PP13TFSI) as the sol-
vent of electrolyte, which exhibited the electrochemical stability
up to lithium reduction potential (<0 V versus Li/Li+) and excel-
lent properties in Li/LiCoO2 cells [2,5–9]. After that, a lot of RTILs
with similar structure with PP13TFSI were investigated as the
solvent for Li-ion batteries [10,12,13]. In addition to the excel-
lent safety characteristics, they also showed good compatibility
with lithium electrode and cycling performance in the batteries.
Although PP13TFSI as well as its analogue has been investigated
widely, there are still some puzzles for the application in the Li-ion
batteries. For example, they usually had relatively high viscosity
and melting point, which caused poor rate capability and low-
temperature performance [5,10,14]. Also, the ionic conductivity of
the binary LiTFSI-RTIL electrolyte systems is much more sensitive to
the salt content, since the viscosity of RTILs is much higher than the
commonly used carbonate-based electrolytes. However, the salt
content in the binary LiTFSI-PP13TFSI electrolyte systems is usually
fixed at 0.4 mol/kg or 10 wt.%, and few papers report the influence
of the salt content on the physicochemical and cell performances
of the electrolytes [8,10].
With fossil energy decreased and exhausted, renewable energy
(i.e. wind and solar energy) is expected to play a leading role in
the future. Meanwhile, the energy storage and converse devices
are necessary for storing the electric power because of the unsta-
ble output from the renewable energy like wind power or sunshine.
Li-ion batteries have many advantages, such as higher energy den-
sity, longer cycling lifetime and lower self-discharge rate than other
rechargeable batteries, so that they can play an important role
on the electric power storage [1]. However, there are some issues
restricting the development of the large-scale Li-ion batteries such
as batteries for electric vehicles (EV) and hybrid electric vehicles
(HEV), even though the small-scale Li-ion batteries have been suc-
cessfully put into practice for portable electronic products. One of
the biggest barriers is the safety concern, which is mainly owing
to volatile and flammable organic solvents used in the electrolyte.
Recently, room temperature ionic liquids (RTILs) have been exten-
sively studied as new solvents in electrochemical batteries and
supercapacitors due to their extremely low vapor pressure and
good flame resistance [2–14]. The unique properties make RTILs
attractive candidates for the electrolytes of Li-ion batteries with
excellent safety characteristics.
In this paper, we first investigate the influence of the content
of Li salt in the binary LiTFSI-PP13TFSI electrolyte on ionic conduc-
tivity and cell performance, and then diethyl carbonate (DEC) as
a co-solvent with low viscosity and melting point is introduced
into the viscous RTIL electrolyte to improve the rate capability
and low-temperature performance of the RTIL-based electrolyte
∗
Corresponding author. Tel.: +86 551 3606971.
E-mail address: cchchen@ustc.edu.cn (C.H. Chen).
0013-4686/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2010.04.041

H.F. Xiang et al. / Electrochimica Acta 55 (2010) 5204–5209
5205
in the LiCoO2/Li cells. This approach provides an effective route to
overcome the shortcomings of pure RTIL-based electrolyte for the
application in Li-ion batteries.
2. Experimental
2.1. Synthesis of LiTFSI-RTIL electrolytes
N-methyl-N-propylpiperidinium bis(trifluoromethanesulfonyl)
imide (PP13TFSI) was prepared by stirring two aqueous solutions of
PP13-Br and LiTFSI (Ferro, battery grade) at room temperature for
12 h. The amount of LiTFSI was slightly in excess versus that of PP13-
Br. PP13-Br was obtained by adding propylbromide (Sinopharm
Chemical Reagent Co., Ltd, as received) to N-methylpiperidine
(Aldrich, as received) in acetonitrile at 70 ◦C with stirring for 24 h.
The resulting RTIL PP13TFSI was extracted by CH2Cl2 and then
washed with water. Finally, the RTIL was dried in vacuum at 80 ◦C
for 48 h, followed by a storage in an argon-filled glove box (Mbraun
Labmaster 130). The chemical structure and the purity of PP13TFSI
were confirmed by 1H NMR (Bruker, 300 M) and thermogravimetric
(TG) measurement (Shimadzu).
Fig. 1. Schematic illustration of the molecular structure of cation and anion from
the room temperature ionic liquid (PP13TFSI).
For the CV test, the voltage range was set at 3.0–0 V, and for the
LSV, the voltage was limited to 6.0 V.
The thermal stability of different electrolytes with charged
LiCoO2 electrode was evaluated by using a Calvet-type calorime-
ter (Setaram C80). LiCoO2/Li cells with different electrolytes were
charged to 4.2 V after three formation cycles, followed by being
disassembled in the glove box. The charged electrodes were rinsed
in dimethyl carbonate (DMC) and then dried in vacuum. Finally,
a mixture of 20 mg electrolyte supplied and 20 mg charged elec-
trode materials scraped from the aluminum foil was placed in a
high-pressure stainless steel vessel with dry argon atmosphere.
The measurement was performed in the temperature range from
ambient temperature to 300 ◦C at a 0.2 ◦C min−1 heating rate.
The binary LiTFSI-RTIL electrolytes were prepared with different
concentrations, i.e. 0.1, 0.2, 0.3 and 0.4 mol kg−1 in the glove box.
Another two electrolytes containing diethyl carbonate (DEC) were
prepared by adding LiTFSI into the mixture of RTIL and DEC; and
the two electrolytes have 20 wt.% and 40 wt.% DEC (compared to the
weight of RTIL) with the same amount of the salt corresponding to
3. Results and discussion
0.4 mol kg−1
.
3.1. Structure and thermal behavior of PP13TFSI
2.2. Thermal, electrical and electrochemical characterization of
LiTFSI-RTIL electrolytes
The structure of PP13TFSI (Fig. 1) was confirmed by 1H NMR
spectra (CDCl3, chemical shift, ppm relative to TMS): 3.4 (3H, 1),
1.1 (3H, 4), 3.8 (2H, 2), 3.6–3.7(4H, 5,9), 1.6–1.9(8H, 3,6,7,8). Fig. 2
shows the TG/DTA curves of the RTIL synthesized, which confirms
the expected high thermal stability. From the TG curve, the RTIL
only exhibits a decomposition process starting at 342 ◦C and ending
at 484 ◦C, with ∼100% weight loss. The DTA curve also indicates only
one endothermic decomposition process with a peak at 453 ◦C, and
the heat generation is −153.5 J g−1. The TG/DTA results also validate
that the RTIL that we have synthesized has the high purity.
The melting point of RTIL was measured by differential scan-
ning calorimetry (DSC, Shimadzu) in the temperature range from
−60 ◦C to 150 ◦C at the heating rate of 10 ◦C min−1. The ionic con-
ductivities of RTIL and the electrolyte solutions were measured
over a wide temperature range of −30 to 90 ◦C using a Model DDS-
307A conductometer (Shanghai Precision & Scientific Instrument
Co. Ltd., China). Each conductivity measurement at a given tem-
perature was carried out after holding the electrolyte solution at
the temperature for 30 min in order to reach full thermal equilibra-
tion. The temperature of the electrolyte solution and subsequent
cell testing at low temperatures was controlled by a WD4005
low-temperature chamber (Shanghai Experimental Equipment Co.
Ltd., China). The lithium transference number was determined
cursorily by a dc polarization method with a CHI604 Electro-
chemical Workstation [15]. By applying a small dc pulse (0.08 V
here) to a symmetrical Li/electrolyte/Li cell and measuring the
initial current, I0, and the steady-state current, Iss, which flow
through the cell, the lithium ion transference number, tLi+ , is sim-
ply given by the equation: tLi+ = Iss/I0 [16]. The self-extinguishing
time (SET) was measured to evaluate the flammability of the elec-
trolytes, and the detailed procedures were described elsewhere
[17,18].
3.2. Electrochemical stability of PP13TSFI
The electrochemical stability of PP13TSFI is investigated by the
voltammetry, including CV and LSV shown in Figs. 3 and 4. As shown
in Fig. 3, the negligible electrochemical behavior with the peak cur-
CR2032 coin cells were used for measuring the electrochemical
properties and cell performances. A positive electrode consisting of
84 wt.% LiCoO2, 8 wt.% acetylene black and 8 wt.% PVDF were made
on aluminum foils, with the mass loading of 1.5 mg cm−2 active
material. All cells were assembled in the glove box and then gal-
vanostatically cycled on a multi-channel battery cycler (Neware
BTS2300, Shenzhen). For all the LiCoO2/Li cells, the cutoff voltages
were set at 2.8 and 4.2 V. Cyclic voltammetry (CV) and linear sweep
voltammetry (LSV) (CHI604 Electrochemical Workstation) were
used to study the electrochemical stability of the RTIL-based elec-
trolyte in the stainless steel (SS)/Li cells at a scan rate of 0.2 mV s−1
.
Fig. 2. TG/DTA curves of PP13TFSI synthesized. The heating rate is 10 ◦C min−1
.

5206
H.F. Xiang et al. / Electrochimica Acta 55 (2010) 5204–5209
Fig. 3. Cathodic cyclic voltammetry of 0.4 mol kg−1 LiTFSI/PP13TFSI electrolyte in
the stainless steel (SS)/Li cell. The scan rate is 0.2 mV s−1
Fig. 5. Temperature dependence of ionic conductivity of electrolytes with different
concentration of lithium salt. The inserted graph shows conductivity variation with
temperature following VTF equation, and R is correlation coefficient for linear fitting.
.
rent density lower than 0.01 mA cm−2 indicates that PP13TFSI has
good cathodic stability in the range of 3–0 V, which is highly advan-
tageous compared with the imidazolium-based ionic liquids, such
as 1-ethyl-3-methyl imidazolium (EMI)-TFSI, with high cathodic
potential of 1.1 V versus Li+/Li [2]. Thus, PP13TFSI can be used in
the battery with metallic lithium anode, where EMI-based RTIL
cannot be used. Also, the obviously anodic decomposition poten-
tial of PP13TFSI is about 5.5 V (in Fig. 4), which suggests that
PP13TFSI has a wide electrochemical window for use in the Li-ion
batteries.
equation is as follows:
ꢀ
ꢁ
−Ea/R
T − T0
ꢀ = ꢀ0 exp
In the VTF equation, Ea is the energy of activation, ꢀ0 is the
pre-exponential factor, T0 is the theoretical glass transition tem-
perature and T is the absolute temperature. For each sample, the
T0 is determined by fitting the temperature-dependent conductiv-
ity data to the VTF equation for the best linearity relationship. Ea
is determined from the linear slope. The deduced values of T0 and
Ea are listed in Table 1. Though the pure RTIL has the highest ionic
conductivity, it cannot be used solely as the electrolyte in the Li-ion
batteries and the addition of LiTFSI is necessary for Li+ conduction
that has more direct and important effect on the cell performance.
Thus, the lithium transference number was cursorily determined
by a dc polarization method at room temperature. From Table 1,
it can be found that T0, which indirectly reflects on the viscos-
ity of the electrolyte systems, increases with the concentration of
lithium salt. With LiTFSI added into the RTIL, the ionic concentration
increases. However, owing to the interaction between the anion
(TFSI−) and cations (Li+ and PP13+), the higher viscosity makes
the ionic motion and transfer more difficult, which results in the
ionic conductivity of the electrolytes reduced rather than raised.
Ea is a parameter reflecting on the temperature dependence of
ionic conductivity of electrolytes. With the concentration of LiTFSI
increasing, Ea firstly ascends for the increased ionic concentration,
and then descends because the increasing viscosity becomes the
more important factor. The lithium transference number, tLi+ , has
the similar variation as Ea. When the lithium salt is added into the
RTIL, the free Li+ appears and becomes more and more with the
concentration increasing, so tLi+ increases. But when the amount
of Li+ exceeds a certain value, the excessive Li+ along with the
anion constitutes ion pairs, which results in the increase of viscos-
ity of electrolyte systems. So, tLi+ descends after the concentration
3.3. Electrical properties of LiTFSI-PP13TFSI electrolytes
It was reported that PP13TFSI had a high viscosity of 151 mPa s
at 25 ◦C, and after addition of lithium salt, i.e. LiTFSI, the electrolyte
become more viscous [14]. High viscosity of RTIL-based electrolytes
could have an adverse effect on the ionic mobility and thus the
ionic conductivity of the electrolytes. Hence, here we investigate
the effect of concentration of lithium salt on the ionic conductivity
and lithium ion mobility in the binary LiTFSI-PP13TFSI electrolyte.
Fig. 5 shows the temperature dependence of the ionic conductivity
of electrolytes with different concentrations of the lithium salt. It
can be easily found that, with increasing the concentration of the
lithium salt, the ionic conductivity gradually decreases, which is
mainly because a significant increase of viscosity after introduc-
ing lithium salt depresses the motion and transfer of free ions in
the electrolyte systems. In addition, the conductivity variation with
temperature of all samples follows the Vogel–Tamman–Fulcher
(VTF) behavior as shown in the inserted graph of Fig. 5. The VTF
Table 1
Data from the VTF equation of temperature dependence of ionic conductivity of
electrolytes with different concentration of lithium salt, and results of lithium trans-
ference number.
Concentration (mol kg−1
)
Pure
0.1
0.2
0.3
0.4
T0 (K)
185
5.18
N/A
185
5.53
0.35
185
5.60
0.41
200
4.59
0.46
205
4.22
0.31
Ea (kJ mol−1
)
Fig. 4. Anodic linear sweep voltammetry of 0.4 mol kg−1 LiTFSI/PP13TFSI electrolyte
tLi
+
in the SS/Li cell. The scan rate is 0.2 mV s−1
.

H.F. Xiang et al. / Electrochimica Acta 55 (2010) 5204–5209
5207
Fig. 6. First voltage profiles (a) and cycling performance (b) of the LiCoO2/Li cells with different electrolytes: (a) 0.2 mol kg−1 LiTFSI/RTIL, (b) 0.3 mol kg−1 LiTFSI/RTIL, (c)
0.4 mol kg−1 LiTFSI/RTIL. The current rate is set at 1/10 C.
Table 2
from the RTIL, is decreased so that Ea is reduced to a certain
extent.
Data from the VTF equation of temperature dependence of ionic conductivity of
electrolytes with different amounts of DEC.
High viscosity of RTIL usually brings poor rate capabil-
ity of the Li-ion batteries. Some groups replaced TFSI− with
bis(fluorosulfonyl)imide anion (FSI−) and obtained the RTILs with
low viscosity, but the improvement of rate capability was still lim-
ited [14,19]. For example, at 2 C discharge rate, the LiCoO2/Li cell
with PP13FSI had only 40% capacity retention of the discharge
capacity at 0.1 C [13]. So, we think it is difficult to solve the prob-
lem of poor rate capability only by designing the structure of RTILs
owing to their intrinsical flaw. The experience of using mixed
solvents in the state-of-the-art electrolyte can give a solution.
Nakahara prepared a nonflammable LiPF6 based electrolyte with
50% PP13TFSI and 50% (EC-DMC-EMC) carbonate mixture, which
exhibited higher reversible capacity at room temperature, signif-
icantly worse capacity retention at lower temperature and better
rate performance than the electrolyte without RTIL [20]. Herein, we
manage toimprove the rate capabilityandlow-temperature perfor-
mance by incorporating DEC with low viscosity and melting point
into PP13TFSI. Fig. 8 shows the rate capability of the LiCoO2/Li cells
with different electrolytes. When the discharge rate is below C/2,
there is slight difference between discharge capacities of the cells
containing three electrolytes. However, a significant capacity loss
is found at C/2, for the electrolyte of 0.4 mol kg−1 LiTFSI/RTIL. And
the discharge capacity is about 30 mAh g−1 at 1 C, which is only 23%
of the discharge capacity at 0.1 C. However, after 20% DEC is added,
excellent rate capability is obtained, that is 89% at 1 C and 78% at 2 C
LiTFSI/RTIL
LiTFSI/RTIL + 20% DEC
LiTFSI/RTIL + 40% DEC
T0 (K)
205
4.22
160
5.25
160
4.40
Ea (kJ mol−1
)
of lithium salt exceeds a certain extend. Here the electrolyte with
0.3 mol kg−1 LiTFSI has the highest Li+ transference number.
3.4. Electrochemical performance of LiCoO2/Li cells with
LiTFSI-PP13TFSI electrolytes
Fig. 6 shows the electrochemical performance of LiCoO2/Li
cells with three electrolytes containing different concentrations of
LiTFSI. The current density is 0.02 mA cm−2, which is equivalent to
1/10 C. It can be seen that all the cells have a high coulombic effi-
ciency of about 95% (Fig. 6a) and good cycling performance (Fig. 6b).
Obviously, the cell with the electrolyte containing 0.3 mol kg−1
LiTFSI has the highest discharge capacity of 146 mAh g−1, which is
coincident with the highest lithium transference number. Although
the electrolyte of 0.3 mol kg−1 LiTFSI/RTIL has lower ionic con-
ductivity than the electrolyte of 0.2 mol kg−1 LiTFSI/RTIL, and
lower concentration of LiTFSI than the electrolyte of 0.4 mol kg−1
LiTFSI/RTIL, it has the best overall performance and thus is recom-
mended to replace 0.4 mol kg−1 or 10 wt.% usually used in previous
papers.
3.5. Effects of DEC addition on the LiTFSI-PP13TFSI electrolytes
PP13TFSI has a good electrochemical stability, but its high vis-
cosity (151 mPa s) and high melting point (8.7 ◦C, reported by
Sakaebe [5]) brings poor rate capability and low-temperature per-
formance in the Li-ion batteries. Thus, a co-solvent with low
viscosity and melting point, DEC, is introduced into the RTIL to
improve the rate capability and low-temperature performance of
the relative cells. Fig. 7 shows the temperature dependence of
ionic conductivity of electrolytes containing 0%, 20% and 40% DEC.
Obviously, the ionic conductivity is substantially improved after
the introduction of DEC. Also, conductivity variation with tem-
perature of all samples follows the VTF behavior in the inserted
graph of Fig. 7. And the corresponding Ea and T0 based on the
VTF equation are all listed in Table 2. After introduction of DEC,
T0 is sharply reduced, which suggests that the viscosity of the
electrolyte system should obviously be decreased. The decreased
viscosity brings the greater Ea due to the introduction of 20% DEC,
but after adding more DEC, the ionic density, especially contributed
Fig. 7. Temperature dependence of ionic conductivity of electrolytes with different
amounts of DEC. The inserted graph shows conductivity variation with temperature
following VTF equation, and R is correlation coefficient for linear fitting.

5208
H.F. Xiang et al. / Electrochimica Acta 55 (2010) 5204–5209
Fig. 8. Rate capability of the LiCoO2/Li cells with different electrolytes: 0.4 mol kg−1
LiTFSI/RTIL (close square), 0.4 mol kg−1 LiTFSI/(RTIL + 20%DEC) (open circle) and
0.4 mol kg−1 LiTFSI/(RTIL + 40%DEC) (open triangle). All the cells are charged to 4.2 V
at 1/10 C and then discharged to 2.8 V at different rates.
Fig. 10. Discharge capacities of LiCoO2/Li cells with the electrolytes under differ-
ent temperatures. All the cells are charged to 4.2 V at room temperature and then
discharged to 2.8 V under different temperatures. The current rate is 1/10 C.
at −10 ◦C is 102 mAh g−1, which is about 72% of the capacity at
room temperature. Therefore, the low-temperature performance
of the cell with the PP13TFSI-based electrolyte can be significantly
improved by the introduction of 20% DEC.
compared with the discharge capacity at 0.1 C, respectively. There-
fore, we can conclude that a small amount of DEC (i.e. 20%) can
efficiently improve the rate capability of the RTIL electrolyte in the
Li-ion batteries. Definitely, it is beneficial to introduce more DEC
(i.e. 40%) for further improving the rate capability, but the gain is
quite narrow.
RTIL is a promising candidate for the electrolyte of Li-ion bat-
teries, because its advantages of nonflammability and low vapor
pressure are favorable for the safety characteristics of Li-ion bat-
teries. It is definitely a compromise that the way by adding a
flammable and volatile organic solvent, i.e. DEC, into PP13TFSI is
used to improve the rate capability and low-temperature perfor-
mance of the RTIL-based electrolyte. In our flammability test, all
the three electrolytes with 0, 20% and 40% DEC are validated to be
nonflammable, because they cannot be ignited, meaning their SET
is all measured to be 0.0 s. Fig. 11 shows the results of C80 calorime-
try for the charged LiCoO2 cathode materials and electrolytes. All
the three samples have a wide exothermic band at 200–275 ◦C,
with two peaks at 225 ◦C and 250 ◦C. There is negligible difference
between the three electrolytes on the exothermic peak at 225 ◦C,
but obvious differences between the peaks at 250 ◦C can be found.
Compared with the sample without DEC, only a tiny increase in
heat generation (denoted by the area of the exothermic peak) is
found for the sample containing 20% DEC, but the heat flow and
heat generation of the sample containing 40% DEC are increased
sharply. In addition, from the consideration of vapor pressure, the
electrolyte containing 40% DEC is also unfavorable. Therefore, the
electrolyte containing 20% DEC is recommended for good safety
characteristics.
Fig. 9 shows the DSC spectra of pure RTIL and the mixtures of
RTIL and DEC. An endothermic peak at 10.5 ◦C is corresponding to
the melting point of PP13TFSI. After 20% DEC is added into the RTIL,
the endothermic peak moves to −18.7 ◦C, and an exothermic peak
at −35 ◦C is attributed to a process of crystallization. When 40%
DEC is added, the endothermic peak corresponding to the melting
point disappears above −50 ◦C, but at the range of 100–150 ◦C, an
endothermic process related to the evaporation of DEC appears dis-
tinctly. The results of these DSC spectra indicate that introducing
DEC can reduce the melting point of the system containing RTIL,
but too much DEC brings a high vapor pressure which destroys the
safety characteristic. Fig. 10 shows the comparison results between
low-temperature performances of the cells with the electrolyte of
0.4 mol kg−1 LiTFSI/RTIL and the electrolyte containing 20% DEC.
For the cell with the electrolyte of 0.4 mol kg−1 LiTFSI/RTIL, the dis-
charge capacity is 119 mAh g−1 at 10 ◦C, which is equal to about 87%
of the discharge capacity at room temperature, and at a tempera-
ture lower than 10 ◦C, the electrolyte is frozen and the cell cannot
be discharged smoothly. As for the electrolyte containing 20% DEC,
the capacity utilization accessed at 10 ◦C and even 0 ◦C keeps a
high level of above 130 mAh g−1, and also the discharge capacity
Fig. 11. C80 heat-flow profiles of the charged Li0.5CoO2 with three kinds of elec-
Fig. 9. DSC spectra of pure RTIL and the mixtures of RTIL and DEC. The scanning rate
trolytes. The heating rate is set at 0.2 ◦C min−1
.
is 10 ◦C min−1
.

H.F. Xiang et al. / Electrochimica Acta 55 (2010) 5204–5209
5209
4. Conclusions
References
[1] J.-M. Tarascon, M. Armand, Nature 414 (2001) 359.
[2] H. Sakaebe, H. Matsumoto, Electrochem. Commun. 5 (2003) 594.
[3] H. Nakagawa, S. Izuchi, K. Kuwana, T. Nukuda, Y. Aihara, J. Electrochem. Soc.
150 (2003) A695.
[4] M. Egashira, M. Nakagawa, I. Watanabe, S. Okada, J. Yamaki, J. Power Sources
146 (2005) 685.
[5] H. Sakaebe, H. Matsumoto, K. Tatsumi, J. Power Sources 146 (2005) 693.
[6] K. Hayashi, Y. Nemoto, K. Akuto, Y. Sakurai, J. Power Sources 146 (2005)
689.
[7] M. Galinski, A. Lewandowski, I. Stepniak, Electrochim. Acta 51 (2006) 5567.
[8] J. Xu, J. Yang, Y. NuLi, J. Wang, Z. Zhang, J. Power Sources 160 (2006) 621.
[9] H. Sakaebe, H. Matsumoto, K. Tatsumi, Electrochim. Acta 53 (2007) 1048.
[10] A. Fernicola, F. Croce, B. Scrosati, T. Watanabe, H. Ohno, J. Power Sources 174
(2007) 342.
[11] V. Baranchugov, E. Markevich, E. Pollak, G. Salitra, D. Aurbach, Electrochem.
Commun. 9 (2007) 796.
[12] J.-H. Shin, W.A. Henderson, S. Scaccia, P.P. Prosini, S. Passerini, J. Power Sources
156 (2006) 560.
The RTIL-based electrolyte is an appealing candidate to replace
the flammable and volatile carbonate-based electrolyte in the Li-
ion batteries, owing to its excellent physicochemical properties,
i.e. nonflammability and negligible vapor pressure. As one of the
most promising RTILs, PP13TFSI is synthesized in this study. The
effects of the content of Li salt in the electrolytes containing
PP13TFSI and LiTFSI on the ionic conductivity and cell performance
are investigated. The electrolyte of 0.3 mol kg−1 LiTFSI/PP13TFSI
is recommended for its higher lithium transference number and
discharge capacity in the LiCoO2/Li cell than other electrolytes.
In addition, it is found that 20% DEC as a co-solvent introduced
into the RTIL-based electrolyte can improve the rate capability and
low-temperature performance, with a negligible damage to the
safety characteristics of the Li-ion batteries. It provides an effective
approach to overcome the shortcoming of RTIL for the application
in Li-ion batteries.
[13] S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, Y. Mita, N. Terada, P. Charest, A.
Guerfi, K. Zaghib, J. Phys. Chem. C 112 (2008) 16708.
[14] H. Matsumoto, H. Sakaebe, K. Tatsumi, M. Kikuta, E. Ishiko, M. Kono, J. Power
Sources 160 (2006) 1308.
[15] P.G. Bruce, C.A.J. Vincent, Electroanal. Chem. 225 (1987) 1.
[16] J.S. Zhao, L. Wang, X.M. He, C.R. Wan, C.Y. Jiang, J. Electrochem. Soc. 155 (2008)
A292–A296.
Acknowledgements
[17] H.F. Xiang, H.Y. Xu, Z.Z. Wang, C.H. Chen, J. Power Sources 173 (2007) 562.
[18] H.F. Xiang, Q.Y. Jin, C.H. Chen, X.W. Ge, S. Guo, J.H. Sun, J. Power Sources 174
(2007) 335.
[19] A. Guerfi, S. Duchesne, Y. Kobayashi, A. Vijh, K. Zaghib, J. Power Sources 175
(2008) 866.
This study was supported by National Science Foundation of
China (grant nos. 20971117 and 10979049) and Education Depart-
ment of Anhui Province (grant no. KJ2009A142). We are also
grateful to the Solar Energy Operation Plan of Academia Sinica. We
gratefully thank Prof. Sun Jinhua and Ms. Ping Ping in the State
Key Laboratory of Fire Science for their kind assistance on the C80
measurements.
[20] H. Nakagawa, Y. Fujino, S. Kozono, Y. Katayama, T. Nukuda, H. Sakaebe, H.
Matsumoto, K. Tatsumi, J. Power Sources 174 (2007) 1021.