Electrochimica Acta 52 (2006) 1398–1408

Ternary polymer electrolytes with 1-methylimidazole based

ionic liquids and aprotic solvents
Jakub Reiter a,∗,1 , Jiřı́ Vondrák a , Jiřı́ Michálek b,c , Zdeněk Mička d
aInstitute of Inorganic Chemistry, Academy of Sciences, 250 68 Řež near Prague, Czech Republic
b
Institute of Macromolecular Chemistry, Academy of Sciences, 162 06 Prague, Czech Republic
c Centre for Cell Therapy and Tissue Repair, Charles University, 150 18 Prague, Czech Republic
d Department of Inorganic Chemistry, Faculty of Science, Charles University, 128 40 Prague, Czech Republic

Received 9 February 2006; received in revised form 24 July 2006; accepted 27 July 2006
Available online 24 August 2006

Abstract
New polymer gel electrolytes containing ionic liquids were developed for modern chemical power sources—supercapacitors and lithium-ion
batteries. Ternary systems polymer–ionic liquid–aprotic solvent as well as materials containing also lithium salts (LiClO4 or LiPF6 ) were prepared
by direct, thermally initiated polymerisation. Poly(2-ethoxyethyl methacrylate) PEOEMA was combined with various ionic liquids based on
1-methylimidazole. Only 1-butyl-3-methylimidazolium hexafluorophosphate BMIPF6 formed a homogenous and slightly translucent polymer
electrolyte, where aprotic solvents—propylene carbonate and ethylene carbonates were used as plasticisers. Materials were studied using the
electrochemical and thermogravimetric methods and exhibit high ionic conductivity up to 0.94 mS cm−1 at 25 ◦ C together with high electrochemical
stability: the accessible potential window on the glassy carbon was found ca. 4.3 V. Prepared non-volatile materials are long-term and thermally
stable up to 150 ◦ C.
© 2006 Elsevier Ltd. All rights reserved.

Keywords: Polymer gel electrolyte; Ionic liquids; Glassy carbon; Lithium-ion batteries; 2-Ethoxyethyl methacrylate

1. Introduction
Recent technology of modern chemical power sources deals
with several limiting factors of further development. Commer-
cialised lithium-ion batteries and supercapacitors usually con-
tain liquid organic compounds and their mixtures, which serve
as a solvent for electrolytes (e.g. lithium or alkylammonium
salts) [1–3]. This situation brings threat of possible leakage
or evaporation of the solvent, what leads not only to decrease
of the battery capacity, but is also environmentally hazardous.
Commonly used organic solvents like aliphatic carbonates, ace-
tonitrile, tetrahydrofuran or various fluorinated compounds are
volatile and flammable. Therefore a new kind of non-flammable
and environmentally friendly solvents is required for large size
lithium-ion batteries and supercapacitors possibly used in elec-
tric or hybrid vehicles.
∗ Corresponding author. Tel.: +420 266172198; fax: +420 220941502.
E-mail address: reiter@iic.cas.cz (J. Reiter).
1 ISE member.
At the present time, many solid and gel polymer electrolytes
are being developed with the aim to substitute the liquid elec-
trolytes in chemical power sources. First attempts were done by
M.B. Armand in late 1970s, when the first generation of polymer
electrolytes based on poly(ethylene oxide) containing lithium
perchlorate was introduced [4,5]. Various polymers, organic sol-
vents and salts were later combined and studied [6–10]. Used
materials have to fulfil several requirements: reasonable ionic
conductivity, high chemical and electrochemical stability, espe-
cially towards oxidation and hydrolysis and have to be available
with low content of impurities and water. Aliphatic carbonates
(e.g. propylene, ethylene or dimethyl carbonate) are recently
used as solvents immobilised in acrylate-based polymers or co-
polymers due to their sufficient electrochemical stability and
low toxicity, when poly(methyl methacrylate) is widely used in
the dental praxis. These materials were studied for the first time
by Iijima et al. [11], than by Bohnke et al. [12] and also in our
laboratory [13–15].
Room temperature ionic liquids (RTILs, ILs), also called
room temperature molten salts (RTMS) are promising organic
compounds to be used in modern chemical power sources

0013-4686/$ – see front matter © 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2006.07.043
 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408
[16–18]. They attracted the attention of many researchers
due to their excellent properties: high ion content and high
conductivity, liquidity in wide temperature range, high ther-
mal stability (up to 400 ◦ C) and low toxicity. Especially for
lithium-ion batteries their non-volatility and non-flammability
is highly appreciated. However the ionic liquids are known
since 1914 (cit. [19]), recently the interest for them increased
in 1992, when new compounds with hydrolytically stable ions
were introduced by Wilkes and Zaworotko [20]. Generally
the ionic liquids can be used in organic synthesis as recy-
clable “green solvents” [21,22] and they can also serve as
conductors in electrochromic devices [23]. Nowadays several
ionic liquids were introduced for liquid electrolytes based on
aliphatic tetraalkylammonium [24] or aromatic pyrrolidinium
and isoquinolinium cations [25]. To improve the electrochem-
ical stability, the organic cation is combined with bulky anion
such as tetrafluoroborate, bis((trifluoromethyl)sulfonyl)imide or
hexafluorophosphate [26]. Also novel, “inversed” ionic liquids
with lithium cation and borate-based anions were also reported
[27].
Our recent research is based on previously reported gel elec-
trolytes based on poly(ethyl methacrylate) PEMA and poly(2-
ethoxyethyl methacrylate) PEOEMA polymers prepared by UV
initiated polymerisation. Reasonably high ionic conductivity, up
to 0.23 mS cm−1 , was reached in the case of lithium perchlo-
rate solution embedded in the cross-linked PEMA or PEOEMA
polymer matrix [28,29]. Both systems exhibited wide potential
window (over 4.2 V), and the cross-linkage of the polymer net-
work in the PEMA–PC–LiClO4 system showed positive effect
on the ionic conductivity [28].
This paper introduces new conducting polymer electrolytes
with embedded ionic liquids. Our work was aimed at devel-
opment of highly conductive solid electrolyte with suitable
properties for lithium-ion batteries and supercapacitors. New,
defined way of preparation together with excellent chemical
and electrochemical long-term stability of used compounds
offers combination of acrylate-based polymer electrolytes with
ionic liquids based on 1-methylimidazole. Ternary systems
polymer-IL-aprotic solvent (propylene carbonate or mixture
of propylene carbonate–ethylene carbonate 50:50, mol%) and
quaternary systems polymer–IL–aprotic solvent–lithium salt
(LiPF6 or LiClO4 ) were prepared and studied. Materials were
investigated from the electrochemical point of view as well as
they were characterised by using thermogravimetric analysis.
Composition of all samples was optimised to obtain samples
with sufficient mechanical and electrochemical properties.
Ternary systems polymer–IL–aprotic solvent can be combined
with high-surface carbon materials and used in supercapacitors.
Systems polymer–IL–aprotic solvent–lithium salt systems
are developed for lithium-ion batteries, when good ionic
conductivity and high content of lithium salt is utilized and
further improvement of conductivity is reached by employing
fourth phase, aprotic organic solvent with high boiling point
and wide accessible potential window (PC or PC–EC).
Various hydrophilic as well as hydrophobic ionic liq-
uids were synthesized by Bonhôte et al. [30] and com-
bined with polymers by Fuller et al. [31] and Lewandowski
1399
and Swiderska [32]. Materials based on poly(acrylonitrile),
poly(ethylene oxide), poly(vinylalcohol) or poly(vinylidene
fluoride)–hexafluoropropylene copolymer were prepared by the
casting technique, when particular polymers swollen in an
organic solvent were mixed with an ionic liquid with conse-
quent evaporation of the solvent. Our method of preparation
is based on mixing of the ionic liquid with monomer and the
cross-linking agent and subsequent direct polymerisation and
formation of cross-linked methacrylate polymer. This method
is combining advantages of the exactly defined and uniform
composition on contrary to materials prepared by the casting
technique. Thermally initiated polymerisation is carried out by
using 2,2 -azobis(isobutyronitrile) initiator that is active already
at temperature 65 ◦ C and finally produces aprotic compound and
nitrogen on contrary to previously used dibenzoylperoxide.
The compatibility of highly polar ionic liquid with gener-
ally low polar monomer or polymer is the crucial problem in
the sample preparation. Methyl methacrylate and styrene were
found to be non-compatible with 1-ethyl-3-methylimidazolium
tetrafluoroborate or 1-butylpyridinium tetrafluoroborate due to
the phase-to-phase separation [33]. Noda and Watanabe suc-
ceed in forming polymer electrolyte with an ionic liquid when
using hydrophilic 2-hydroxyethyl methacrylate [33], however
this monomer can not be used in aprotic systems for lithium-ion
batteries or supercapacitors due to limited electrochemical sta-
bility caused by presence of free hydroxyl groups. Morita used
more polar aprotic poly(ethylene glycol)-methacrylate)matrix
[34]. Solution of this compatibility problem is in using low-
volatile aprotic solvent in combination with a hydrophobic, low
polar ionic liquid such as 1-butyl-3-methylimidazolium hexaflu-
orophosphate.
2. Experimental
2.1. Materials
Butylbromide (Sigma–Aldrich) was distilled prior to the
ionic liquid synthesis. 1-Methylimidazole (Sigma–Aldrich)
was used as received. Monomer, 2-ethoxyethyl methacry-
late (EOEMA) was obtained from Sigma–Aldrich and puri-
fied by distillation under the reduced pressure. Cross-linking
agent, ethylene dimethacrylate (EDMA; Sigma–Aldrich)
was used as received. The polymerisation initiator, 2,2 -
azobis(isobutyronitrile) (AIBN) was purchased from Fluka and
recrystallised from chloroform. All monomers and the initiator
were stored at 4 ◦ C before use.
Propylene carbonate (Sigma–Aldrich, >99.7%, water con-
tent <0.005%) and ethylene carbonate (Sigma–Aldrich) were
stored under molecular sieves (3A pellets, Sigma–Aldrich). Ace-
tonitrile (Merck), chloroform (Lachema, Czech Republic) and
hexane (Lachema, Czech Republic) were used as received.
Lithium perchlorate (Merck) was dehydrated in vacuum
at 110 ◦ C for 48 h and than stored in a desiccator. Ammo-
nium tetrafluoroborate (Sigma–Aldrich), lithium hexafluo-
rophosphate and potassium hexafluorophosphate (both Merck)
were used as received. LiPF6 was stored under nitrogen.
1400 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408
2.2. Synthesis of ionic liquids
Two ionic liquids were prepared: 1-butyl-3-methyl-
imidazolium tetrafluoroborate (BMIBF4 ) and 1-butyl-3-methyl-
imidazolium hexafluorophosphate (BMIPF6 ). Synthetic work
followed previously published procedures [18,30,32]. General
method of preparation is based on two-step synthesis, when
we modified the published procedures to improve the purity
of prepared ionic liquids (the conversion into tetrafluorobo-
rate or hexafluorophosphate with 100% yield is required). In
the first step, 1-butyl-3-methylimidazolium bromide (BMIBr)
is prepared and than its anion is substituted by tetrafluorobo-
rate or hexafluorophosphate. Prepared ionic liquids were stored
under molecular sieves in a desiccator.
2.2.1. BMIBr
A total of 118 g of butylbromide (freshly distilled) was added
dropwise to a solution of 27.5 g of 1-methylimidazole in 270 ml
of chloroform. The solution was stirred for 3 h at room tempera-
ture, then refluxed for 3 h at 70 ◦ C and left overnight. Chloroform
was distilled off and the remaining salt BMIBr was washed five
times with 100 ml portions of hexane and finally dried at 70 ◦ C
under vacuum. Yield: 59.4 g of yellow-brown, highly viscous
liquid (81%).
2.2.2. BMIBF4
The substitution of bromide anion by tetrafluoroborate was
performed in an acetonitrile solution, where ammonium bromide
is not soluble. However, solubility of ammonium tetrafluorob-
orate in acetonitrile is also low. Our first experiments using
published procedures showed non-stoichiometric reaction and
contamination of prepared ionic liquid with BMIBr. This result
led to the modification of previously published preparation pro-
cedures [29,30].
A solution of 15.17 g BMIBr in 300 ml of acetonitrile was
placed in a round bottom flask with a stirrer and a Soxhlet extrac-
tor, where a cartridge with 7.256 g of powdered ammonium
tetrafluoroborate was fixed. NH4 BF4 was continuously extracted
for 12 h with acetonitrile into BMIBr solution, where white pre-
cipitate of NH4 Br appeared. Described procedure guaranteed
100% conversion to BMIBF4 . After extraction of NH4 BF4 into
solution the mixture was stirred for 12 h at room temperature.
The white precipitate of NH4 Br was filtered off and acetonitrile
was distilled off under vacuum at 60 ◦ C; the remaining liquid
was dried at 80 ◦ C under vacuum. Yield: 15.4 g of pale yellow
liquid (98%).
2.2.3. BMIPF6
The substitution of bromide anion by hexafluorophosphate
was performed in water, where BMIPF6 is not soluble and forms
hydrophobic globules at the bottom of the flask. The amount of
13.62 g of KPF6 in 150 ml of water was added into a solution of
16.2 g BMIBr in 70 ml of distilled water and stirred for 12 h at
room temperature. The arising phase of BMIPF6 was removed
and washed eight times with 150 ml portions of water to remove
NH4 Br. The absence of bromide anions was confirmed using
the test with AgNO3 . Remaining water was distilled off under
vacuum at 60 ◦ C and the remaining viscous liquid was dried at
80 ◦ C under vacuum. Yield: 11 g of yellow liquid (52%).
2.3. Preparation of polymer electrolytes
Two systems were prepared: ternary system polymer–IL–
aprotic solvent and quaternary system polymer–IL–aprotic
solvent–lithium salt. These two principal combinations were
prepared from particular compounds using thermally initiated
polymerisation.
The initial mixture was prepared by mixing of these
compounds: monomer (EOEMA), cross-linking agent EDMA
(0.3 mol% of monomer) and the polymerisation initiator AIBN
(1 mol% of monomers). In the case of ternary systems, pure
propylene carbonate or propylene carbonate–ethylene carbonate
(50:50, mol%) mixture was used. For the quaternary systems
0.75 M solutions of LiClO4 or LiPF6 in particular solvent or
solvent mixture were used. After mixing with a liquid ionic the
whole mixture was bubbled by nitrogen for 8 min and poured
into a Petri dish (5 cm diameter), weighed and placed for 2 h
into an oven preheated at 85 ◦ C. After removal from the oven, the
samples were cooled slowly in an empty desiccator and weighed
again. Considering the composition of the initial mixture, the
difference in the weight before and after polymerisation pro-
cess is caused only by the monomer partial evaporation. This
effect is taken into account in designing the initial ratio of used
chemicals, however the evaporation of monomer did not exceed
3 wt.%. The composition of sample is expressed in molar per-
centage.
2.4. Methods of investigation
Potentiogalvanostat PGSTAT 10 (Eco Chemie, The Nether-
lands) was used for the electrochemical measurements including
FRA-2 (Frequency Response Analyser) module for impedance
measurements. The electrochemical measurement of liquid sys-
tems was performed with a glassy carbon electrode (Metroohm;
2.0 mm diameter) as the working, platinum as the counter and
PMMA–Cd–Cd2+ as the reference electrode. The conductiv-
ity was measured by the impedance spectroscopy (also called
ac impedance measurements) offering information about mate-
rial conductivity, electrode–polymer electrolyte boundary and
electrochemical behaviour of immobilised species in solid sam-
ples under various conditions (material composition, electrode
potential, temperature, age) [35,36]. Conductivities of prepared
ionic liquids were measured at 25 ◦ C using conductivity cell
(Jenway, platinum electrodes, cell constant S = 1.00 ± 0.01).
Temperature dependent conductivity measurements were per-
formed in the temperature range from 0 to 80 ◦ C using ther-
mostated bath (precision of the temperature ± 1 ◦ C). The same
procedure was used for the measurement of resistivity of solid
samples in the temperature range from −70 to 80 ◦ C. Here
a two-electrode arrangement was used, when a slice of gel
(2 cm × 2 cm, ca. 1.00 mm thickness) was sandwiched between
two parallel stainless steel electrodes. The sample thickness
was measured by using Mahr high-precision thickness gauge
(precision ± 0.01 mm). In both cases single potential impedance
 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408
spectra were measured in the frequency range from 200 kHz to
1 Hz at low temperatures and from 100 kHz to 10 Hz at tem-
peratures above 0 ◦ C. Obtained spectrum was analysed by the
EcoChemie Autolab software producing the values of the equiv-
alent circuit elements. The evaluated value of ohmic resistance
was converted in the value of specific resistivity or conductivity.
New cell for solid-state electrochemical measurements was
developed by Vondrák et al. [38,39] and has been used for inves-
tigation of prepared polymer electrolytes [39]. A glassy carbon
served as the working and counter electrode (both 6.1 mm diam-
eter) together with the PMMA–Cd–Cd2+ solid-state reference
electrode [15,39,40]. All potentials in the paper are related to this
Cd/Cd2+ system (E(Cd/Cd2+ ) = 2.66 V versus Li/Li+ ). The sur-
face of working and counter electrode was polished by abrasives
(0.3 alumina, Metroohm) and soft cloth after each measurement.
Before and after the voltammetrical measurement a single poten-
tial impedance spectrum was recorded using the same three-
electrode arrangement with Cd/Cd2+ reference electrode. The
frequency range was from 100 kHz to 10 Hz and the potential of
the working electrode was 0 V versus Cd/Cd2+ , where no elec-
trochemical reaction occurs. Except the temperature-dependent
measurements, all experiments were performed at 25 ◦ C.
The simultaneous TGA-DTA measurement was taken in air
and argon at the heating rate of 5 ◦ C min−1 . Experiments were
performed with a Simultaneous Thermal Analysis Netzsch STA
409 (Germany).
3. Results and discussion
Our recent experiments showed, that 1-ethyl-3-methyl-
imidazolium, 1-propyl-3-methylimidazolium, and 1-butyl-3-
methylimidazolium tetrafluoroborates are not compatible with
PEOEMA–PC polymer electrolytes, however they are soluble
in the EOEMA–PC initial mixture. Phase separation and the
ionic liquid exclusion were observed during the polymerisation
process.
Systems PEOEMA–PC–BMIPF6 and PEOEMA–PC–EC–
BMIPF6 were found to be compatible and prepared samples
were transparent to slightly translucent foils of moder-
ate elasticity. Required foils can be easily cut out. The
polymer–PC–BMIPF6 samples were somewhat sticky, what
can be advantage in an electrochemical device designing, where
a good contact between the electrolyte and the electrode is
required. However no residual traces were left after the sample
removal either from the metal electrode or Petri dish glass. No
visible changes of mechanical properties were observed within
several weeks, when stored either on air or in an inert, nitrogen
atmosphere.
3.1. Conductivity measurements
The specific ionic conductivities (σ) of prepared ionic liq-
uids were investigated as functions of the anion and the ionic
liquid–aprotic solvent ratio. The information on each system
is summarised in Table 1. Substitution of bromide anion with
weakly bonding PF6 − leads to the increase of conductivity from
0.26 to 1.50 mS cm−1 due to the strong decrease of the ionic liq-
1401
Table 1
Specific conductivities (at 25 ◦ C) of ionic liquids solutions in aprotic solvent
and corresponding neat ionic liquids
Composition (mol%) σ (mS cm−1 )
Solvent Ionic liquid
– 100% BMIBr 0.26
– 100% BMIPF6 1.50
– 100% BMIBF4 2.92
80% PC 20% BMIPF6 9.37
40% PC, 40% EC 20% BMIPF6 8.79
80% PC 20% BMIBF4 11.07
40% PC, 40% EC 20% BMIBF4 14.05
uid viscosity. For neat BMIPF6 , the obtained values are in good
agreement with the data by Balducci (1.58 mS cm−1 ; cit. [41]),
however Fuller published higher conductivity 1.8 mS cm−1 at
22–23 ◦ C measured by the ac impedance technique [31]. Dis-
solving of the ionic liquid in aprotic solvent results in increase
of conductivity due to the decrease of viscosity and decreas-
ing ion–ion association. Dissolving of BMIPF6 in propylene
carbonate causes almost five-time increase of conductivity (see
Table 1): the ionic conductivities for BMIPF6 and BMIPF6 –PC
(20–80 mol%) are 1.50 and 9.37 mS cm−1 at 25 ◦ C, respectively.
Table 1 also introduces similar effect for the BMIBF4 ionic liq-
uid and its solutions in PC and PC/EC. It is expected, that the
ionic liquid concentration decrease will change the equilibrium
(R1 R2 Im+ )solv + (X− )solv ↔ R1 R2 Im+ · · ·X−
to the left and thus leads to increase of conductivity.
This positive effect was observed in our previous inves-
tigation of PEOEMA–PC–LiClO4 systems [28] and also by
Lewandowski and Świderska [32] and Tokuda et al. [42].
No remarkable difference was found in comparing solutions
of BMIPF6 either in pure propylene carbonate or propy-
lene carbonate–ethylene carbonate mixture. Following these
results, the molar ratio BMIPF6 –PC (20:80, mol%) and
BMIPF6 –PC–EC (20:40:40, mol%) was used also in the case
of polymer gel electrolytes. Lemordant and others showed, that
higher content of the ionic liquid is not connected with the
improvement of the solution conductivity [21,43].
Fig. 1 presents relationship between the ionic conductivity of
BMIPF6 and BMIPF6 –PC liquids and temperature, when data
are plotted in Arrhenius coordinates (specific conductivity is
plotted as a decadic logarithm). In both cases linear, simple
Arrhenius behaviour over temperature range from 0 to 80 ◦ C
can be seen (correlation coefficient R2 > 0.995 in both cases).
Solution of BMIPF6 in PC exhibits higher conductivity than the
BMIPF6 itself in the whole range of temperatures. Obtained data
can be fitted with the Arrhenius equation in logarithmic form:
A
log10 (σ) = +B (1)
T
Similarly, the Arrhenius activation energy for conduction EA is
accessible from the parameter A using the formula (2). Both Eqs.
(1) and (2) are based on the exponential form σ = exp(−EA /RT).
The conductivity activation energy corresponds to a slope in
the Arrhenius coordinates (see Fig. 1) and explains, how the
1402 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408

behaviour in solutions in particular solvents. This conclusion

Fig. 1. Arrhenius plot of ionic conductivities for (1) neat BMIPF6 and (2)
BMIPF6 –PC (20:80, mol%); temperature range from 0 to 80 ◦ C.
conductivity is influenced by the temperature:
EA = −2.303AR
The activation energies for liquid systems, BMIPF6 and
BMIPF6 –PC are 37.9 and 18.0 kJ mol−1 respectively. In agree-
ment with observations of Balducci, who compared neat
BMIPF6 and BMIBF4 to the tetraethylammonium tetrafuorobo-
rate solution in PC, the systems containing propylene carbonate
exhibits lower activation energy values and therefore their con-
ductivity is more resistant against temperature influence [41].
On contrary, the systems containing high concentration of ions
such as neat ionic liquids are more sensitive concerning influ-
ence of temperature on the conductivity and ion–ion association
and possibly also on viscosity.
Polymer gel electrolytes can be described as a system of
immobilised microscopic liquid phase (ionic liquid and apro-
tic solvent) in the cross-linked polymer network. Qualita-
tive electrochemical behaviour of immobilised inorganic or
organic species in the polymer system is corresponding to their
is supported by our previous experiments, when electrochemi-
cal behaviour of ferrocene and other transition metal complexes
was described in PMMA–PC polymer electrolytes [39]. Pres-
ence of the polymer is not influencing qualitative behaviour of
immobilised species, but it strongly affects the mobility and
therefore conductivity of prepared samples. The diffusion coef-
ficient of ferrocene in liquid PC and PMMA–PC polymer gel
electrolyte is 7 × 10−6 and 2 × 10−9 cm2 s−1 , respectively [39].
The effect of polymerisation is however not so strong in the
case of mobility of solvated ions, when the difference between
conductivity of the liquid and polymer gel electrolyte is one
order approximately. Beside the restriction of the ion mobility
the conductivity decrease is also caused by decrease of the ion
concentration. The conductivity of the PMMA–PC gel contain-
ing no ionic compound is not exceeding 2.5 × 10−7 S cm−1 at
20 ◦ C (cit. [15]), where very low concentration of the polymeri-
sation initiator products can be present.
Specific conductivities of some representative PEOEMA–
PC–BMIPF6 , PEOEMA–PC/EC–BMIPF6 gel electrolytes
together with the samples containing also lithium salts such as
LiClO4 or LiPF6 are summarised in Table 2. Used lithium salts
are widely used in the technology of lithium batteries and also
in the electrochromic devices.
(2) The composition of the gel electrolyte with the highest molar
content of BMIPF6 was found to be PEOEMA–PC–BMIPF6
50:40:10, mol% (53.3:27.5:19.2, wt.%) and further increase of
BMIPF6 content led to the phase-to-phase separation and het-
erogeneous samples with ionic liquid excluded on the sample
surface were formed. As was mentioned in the previous text,
the ratio BMIPF6 –PC or BMIPF6 –PC–EC in gel was close to
10:40 and 10:20:20 (mol%), respectively, following the high
conductivity values of corresponding liquid systems. Suitable
PEOEMA content was found to be 50 mol%, what ensures good
mechanical strength and material stability.
In the ternary systems, the highest conductivity value
0.50 mS cm−1 was reached with 10 mol% content of BMIPF6 .
In the set of polymer electrolytes with 8 mol% content of
BMIPF6 positive influence of the increasing microscopic liq-
uid phase (propylene carbonate) was found similar to previ-
ously described polymer gel systems [28]. The increase of the
PC content from 32 to 37 mol% caused 2.1-time increase of

Table 2
Specific conductivities (at 25 ◦ C) of prepared polymer electrolytes containing ionic liquids and lithium salts (system PEOEMA–PC–LiClO4 is introduced for
comparison from [28])
Composition (mol%) σ (mS cm−1 )

Polymer Solvent Ionic liquid Lithium salt

50% PEOEMA 40% PC 10% BMIPF6 – 0.50

55% PEOEMA 37% PC 8% BMIPF6 – 0.38
60% PEOEMA 32% PC 8% BMIPF6 – 0.18
50% PEOEMA 20% PC, 20% EC 10% BMIPF6 – 0.35
50% PEOEMA 40% PC 9.8% BMIPF6 1.8% LiPF6 0.61
49.2% PEOEMA 19.6% PC, 19.6% EC 9.8% BMIPF6 1.8% LiPF6 0.94
48.4% PEOEMA 38.8% PC 9.7% BMIPF6 3.1% LiClO4 0.61
47.9% PEOEMA 19.6% PC, 19.6% EC 10.3% BMIPF6 2.6% LiClO4 0.64
41.2% PEOEMA 37.5% PC – 21.3% LiClO4 0.23
 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408
the conductivity (see Table 2). Co-immobilisation of lithium
hexafluorophosphate or lithium perchlorate into PEOEMA
electrolytes improves the conductivity up to 0.94 mS cm−1
in the case of PEOEMA–PC/EC–BMIPF6 –LiPF6 electrolyte
(49.2:19.6:19.6:9.8:1.8, mol%), what is the highest reached
value among the studied polymer systems (see Table 2).
Three functions of present propylene carbonate or propylene
carbonate–ethylene carbonate mixture in the polymer electrolyte
can be introduced. However the initial monomer and ionic liq-
uid are miscible, the system PEOEMA–BMIPF6 was found to
be non-compatible and the attempt to form a binary electrolyte
led to the phase-to-phase separation. Aprotic cyclic carbonates
therefore serve as plasticisers and moderators between the poly-
mer and the ionic liquid, which are compounds of different
polarity. Similar function was used in the second-generation
polymer gel electrolytes in early 1980s, when low volatile sol-
vents known from the technology of plastic materials were used
[1,4,5].
The second reason for PC and PC–EC co-immobilisation
in polymer electrolyte is in their positive influence on the
material conductivity. The impedance measurements of the
liquid systems showed, the conductivity of BMIPF6 –PC
solution is ca. five-time higher can be reached in compar-
ison with the neat BMIPF6 . Similar effect was observed
by Lewandowski and Świderska in the case of solid elec-
trolytes based on poly(acrylonitrile) and BMIPF6 , when
conductivity 3 × 10−4 mS cm−1 was reached with mentioned
binary system. Co-immobilisation of sulpholane in the
sample led to ca. four-order increase of conductivity [32].
Caillon–Caravanier introduces the model of a gel electrolyte
(here diacrylate–butyrolactone–LiTFSI) as a system of immo-
bilised conductive phase (aprotic solvent), in a polymer, where
the increasing content of the solvent leads to the improvement
of conductivity due to decreasing ion–ion association [46]. In
agreement with the literature and our previous experiments, the
polymer phase does not bring any remarkable contribute to the
material overall conductivity, the motion of ions is located in
the immobilised microscopic liquid phase.
Third reason is important mainly for the application in the
lithium-ion batteries. Here, organic carbonates serve also as SEI
forming agents on the anodes [44,45]. Therefore also dimethyl
or diethyl carbonate can be immobilised in the polymer gel elec-
trolyte.
In the case of acrylate-based polymer electrolytes with
ionic liquids is the relationship between conductivity and
temperature more complex due to the presence of polymer.
Fig. 2 presents the Arrhenius plot for PEOEMA–PC–BMIPF6
and PEOEMA–PC/EC–BMIPF6 –LiPF6 polymer electrolytes,
which exhibit the highest conductivity among the gel samples.
The plot of the specific conductivity consists of two linear parts
with different slopes (see Fig. 2). The point of transformation
occurs in the temperature range from −25 to −30 ◦ C, where
the polymer structure is changed from the elastomeric (less
organised) to crystalline form (more organised). This process
is associated with remarkable increase of resistivity. This effect
is common for all acrylate-based polymer electrolytes and was
observed in other similar systems [14,15,28].
1403
Fig. 2. Arrhenius plot of ionic conductivities for (1) PEOEMA–PC–BMIPF6
(50:40:10, mol%; empty dots) and (2) PEOEMA–PC/EC–BMIPF6 –LiPF6
(49.2:19.6:19.6:9.8:1.8, mol%; full dots) polymer electrolytes; temperature
range from −70 to 80 ◦ C.
Both PEOEMA–PC–BMIPF6 and PEOEMA–PC/EC–
BMIPF6 –LiPF6 samples exhibit high ionic conductivity above
−25 ◦ C. The values for 60 ◦ C are 1.40 and 2.80 mS cm−1 ,
respectively. The temperature region around 60 ◦ C is of interest
for electric vehicle applications involving supercapacitors,
lithium-ion batteries and also fuel cells (all systems promising
as the auxiliary power sources).
Higher conductivity values of PEOEMA–PC/EC–BMIPF6 –
LiPF6 samples were observed above −15 ◦ C. Under this temper-
ature the conductivity of sample without lithium salt is higher,
when the positive contribution of present LiPF6 is restricted
probably due to increasing viscosity of the microscopic PC–EC
liquid phase. Also higher ion concentration can lead to an
increase of the ion association.
Similarly to the liquid systems, at temperatures over
0 ◦ C the conductivity can be expressed with the Arrhe-
nius formula (1). Calculated apparent activation ener-
gies for PEOEMA–PC–BMIPF6 and PEOEMA–PC/EC–
BMIPF6 –LiPF6 are similar, 22.5 and 24.6 kJ mol−1 , respec-
tively. The EA values for the polymer systems are higher than
EA of BMIPF6 –PC solution (see text above). The activation
energy value is mainly depending on the type of salt, e.g.
PEOEMA–PC–LiClO4 (41.2:37.5:21.3, mol%) polymer elec-
trolyte exhibits higher EA value 58.3 kJ mol−1 (cit. [28]). Beside
the polymer structure transformation point at ca. −25 ◦ C another
area of interest is to be found below −55 ◦ C (freezing point of
propylene carbonate). In this temperature region the conductiv-
ity of all acrylate-based samples is rapidly decreasing to values
10−8 –10−9 S cm−1 and the material is becoming almost non-
conductive due to solidification of propylene carbonate.
Impedance spectroscopy was successfully used for the
description of the polymer electrolyte behaviour under vari-
1404 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408
Fig. 3. Impedance spectra of PEOEMA–PC–BMIPF6 (50:40:10, mol%) mea-
sured at 21 ◦ C (), 8 ◦ C (䊉) and −9 ◦ C () with the corresponding equivalent
circuit.
ous conditions. Obtained impedance spectra can be expressed
using the equivalent circuits according to the Boucamp nota-
tion [36] and usually are plotted as Nyquist plots of real and
imaginary part of the complex impedance. Similar explanation
of the obtained equivalent circuits and corresponding processes
in the investigated polymer electrolytes was already described
in literature [35,37,47]. Impedance spectra measured at temper-
atures above the structural transformation of the polymer (above
ca. −25 ◦ C) can be easily described as the material bulk resis-
tance R in a serial connection with a constant phase element Q
(RQ equivalent circuit according to the Boucamp notation). The
resistance value is used for the calculation of specific resistivity
and conductivity; the Q element parameters (capacity Y0 and
phase angle Θ) are describing the boundary between the block-
ing electrode and inserted sample. Fig. 3 is presenting impedance
spectra measured at 21, 8, and −9 ◦ C. At low frequency, the
impedance plot should be a straight line parallel with the imagi-
nary axis, but the double-layer at the blocking electrodes causes
the curvature represented by a phase angle. The phase angle,
which is representing the difference from the ideal capacitance
behaviour (Θ = 90), is decreasing with decreasing temperature.
However the phase angles are always higher than in previously
reported measurements with PEOEMA–PC–LiClO4 gel elec-
trolytes [28], especially at temperatures above 30 ◦ C.
At low temperatures the impedance measurements were per-
formed in broader frequency range from 200 kHz to 1 Hz. Fig. 4
shows spectra measured at −43, −49, and −54 ◦ C. The cir-
cuit can be described with a formula (RP C)Q according to the
Boucamp notation. The resistance RP represents the bulk mate-
rial resistance. The capacity element C describes the dielectric
behaviour of the solidified liquid phase (the viscosity of the
PC microscopic liquid phase is strongly increasing) and pos-
sibly also the bulk polarization of immobile polymer chains.
Here, the system behaves like a capacitor and resistor in par-
Fig. 4. Impedance spectra of PEOEMA–PC–BMIPF6 (50:40:10, mol%) mea-
sured at −43 ◦ C (), −49 ◦ C (䊉) and −54 ◦ C () with corresponding equivalent
circuit.
allel. At higher temperatures (above −10 ◦ C) the contribution
of the capacity element is negligible due to absence of solidi-
fied PC phase and the equivalent circuit can be described as in
Fig. 3. Similar conclusion was done by Baskaran in the case
of poly(vinyl acetate)–DMF–LiClO4 polymer electrolytes [35]
and also by Marzantowicz investigating influence of tempera-
ture on the conductivity of PEO–LiTFSI electrolytes [54]. In
the temperature region from ca. −30 to −10 ◦ C the behaviour
of the polymer gel can be described as transient between the
two limiting states. The phase angle Θ is decreasing with the
decreasing temperature (see Figs. 3 and 4). On contrary to
PEOEMA–PC–LiClO4 gel electrolytes, impedance measure-
ments of PEOEMA–PC–BMIPF6 gels showed, that the con-
tribution of the capacity element appears at lower temperatures,
what is probably caused by the presence of BMIPF6 forming
supercooled liquids, even down to the temperatures of solid CO2 .
However Fuller determined the melting point of BMIPF6 to be
∼10 ◦ C (cit. [31]).
3.2. Cyclic voltammetry
The main advantage of using ionic liquids in modern chem-
ical power sources is in their higher electrochemical stability
in comparison with aqueous and several non-aqueous systems.
The accessible stability window of BMIBr on the glassy carbon
electrode was found to be ca. 3.2 V. The electrochemical stability
of BMIBr is mainly limited at positive potentials, where bro-
mide anion is oxidised. Substitution of the bromide anion with
weakly complexing anions such as tetrafluoroborate or hexaflu-
orophosphate remarkably enlarges the potential window up to
4.6–4.7 V. This modification allows the application of prepared
ionic liquids in the technology of high-voltage (4 V) lithium-
ion batteries. Fig. 5 illustrates the electrochemical stability of
 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408
Fig. 5. Cyclic voltammograms of BMIBr, BMIPF6 and BMIPF6 –PC (20:80,
mol%) (scan rate 50 mV s−1 ; working electrode: glassy carbon; auxiliary elec-
trode: platinum; reference electrode: PMMA–Cd–Cd2+ ).
three different systems measured by cyclic voltammetry on the
glassy carbon electrode: BMIBr, BMIPF6 and 20 mol% solu-
tion of BMIPF6 in propylene carbonate. From the measured
voltammograms, limiting reduction (ERED ) and oxidation (EOX )
potentials can be obtained, when defined as the potential where
the current density reached 1 mA cm−2 (cit. [16]).
Concerning the BMIPF6 ionic liquid, the anodic oxidation
limit at 2.8–2.9 V is assigned to decomposition of PF6 − anion,
when formation of PF5 and F− was proposed [43]. Voltammet-
rical measurements showed that combination of BMIPF6 with
PC or PC–EC solvents does not restrict the accessible potential
window, however ca. 0.2–0.4 V decrease of the cathodic poten-
tial was introduced by Lewandowski and Galinski [18]. The
anodic limit of prepared electrolytes was improved up to 2.8
versus Cd/Cd2+ (5.5 V versus Li/Li+ ). In agreement with liter-
ature [49–53] partial oxidation of carbonates at high potentials
above 2.5 V was observed, close to the anodic limit. The cathodic
limit for the BMIPF6 , BMIPF6 –PC and BMIPF6 –PC/EC liq-
uids was found at ca. −1.8 V. It is believed that reduction of
1,3-dialkylimidazolium cation is related to the H(2) acidity on
carbon C(2) and a difference was previously found between imi-
dazolium cations [30,48]. Within mentioned potential range sev-
eral irreversible waves were found: traces of oxygen and possible
water are reduced at potentials around −0.1 and −1.5 V. Due to
their low current density (wave height less than 0.1 mA cm−2 )
can be omitted as negligible. The low concentration of oxygen
and water is also obvious from very low currents of the anodic
wave around 1.5 V, when the products of oxygen and water
reduction are reoxidized in agreement with literature [49–53]
and our previous observations [28].
Electrochemical characterisation of prepared PEOEMA–
PC–BMIPF6 and PEOEMA–PC/EC–BMIPF6 electrolytes is
1405
Fig. 6. Cyclic voltammetry of PEOEMA–PC–BMIPF6 , PEOEMA–PC/EC–
BMIPF6 –LiClO4, and PEOEMA–PC/EC–BMIPF6 –LiPF6 polymer electrolytes
(scan rate 10 mV s−1 ; working and auxiliary electrode: glassy carbon; reference
electrode: PMMA–Cd–Cd2+ ).
more complex. Fig. 6 shows cyclic voltammograms of ternary
PEOEMA–PC–BMIPF6 as well as PEOEMA–PC–BMIPF6 –
LiPF6 and PEOEMA–PC/EC–BMIPF6 –LiPF6 electrolytes con-
taining lithium hexafluorophosphate. Due to presence of more
chemical compounds in comparison with liquid systems two
main differences appear. The irreversible cathodic wave under
−1 V of the water reduction is broader and cannot be neglected.
However the current densities are still under the 1 mA cm−2 limit
(ca. 0.7–0.8 mA cm−2 ), the application of materials in power
sources in future will require discussion of modification of the
polymer electrolyte preparation in an inert atmosphere. Simi-
larly to our previous experiments with PEOEMA–PC–LiClO4
electrolytes, the wave of oxygen reduction is strongly reduced
in further cycling due to Li2 O and Li2 O2 conductive layer for-
mation [28,51,52].
Comparison of voltammograms of the liquid and PEOEMA
polymer electrolytes shows that also the anodic waves at poten-
tials above 2.5 V are noticeably higher. As they appear in the case
of gel electrolytes containing lithium salts, they can be assigned
to impurities in used salts. Summary of cathodic and anodic lim-
its of studied electrochemical systems is summarised in Table 3.
As it can be seen, the immobilisation of liquid electrolytes in
cross-linked PEOEMA polymer does not principally restrict the
accessible electrochemical window more than by 0.3 V on the
cathodic side due to reduction of oxygen and water moisture
absorbed in the gel during the preparation.
Impedance measurements done before and after voltammet-
rical investigation showed in all prepared polymer electrolytes,
that no non-conducting layer is formed on the surface of the elec-
trode within several cyclic voltammograms. Impedance spectra
were measured in the three-electrode arrangement (see Section
1406 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408

Table 3
Potential limits of prepared ionic liquids and polymer electrolytes on glassy
carbon (potentials are introduced vs. PMMA–Cd–Cd2+ )
Sample Cathodic limit (V) Anodic limit (V)

BMIBr −1.8 1.5

BMIPF6 −1.9 2.8
BMIPF6 –PC −1.8 2.8
BMIPF6 –PC–EC −1.8 2.9
PEOEMA–PC–BMIPF6 −1.6 2.8
PEOEMA–PC/EC–BMIPF6 −1.6 2.8
PEOEMA–PC–BMIPF6 –LiPF6 −1.5 2.8
PEOEMA–PC–BMIPF6 –LiClO4 −0.9 (−1.4)a 2.8
a Lower anodic limit value involves the effect of present water traces.

2.4) at the potential E = 0 V, where no electrochemical reaction

occurs.
The resistance of prepared electrolyte, represented by the
semicircle in measured impedance spectrum on Fig. 7 is not
changed significantly during the electrochemical treatment. The
same results were obtained in the case of liquid systems. Despite
of the facts described above, the purity of prepared polymer
electrolytes with ionic liquids seems to be satisfactory for the
technology of chemical power sources.
3.3. Thermogravimetric analysis
From the application point of view the thermal stability
is another principal question beside the electrochemical and
long-term stability and reasonable conductivity. Used ionic liq-
uids are thermally stable up to 300–400 ◦ C (cit. [53]). The
PEOEMA–PC–BMIPF6 sample exhibits thermal stability up
Fig. 8. Simultaneous TGA and DTA curves of PEOEMA–PC–BMIPF6
(50:40:10, mol%) gel electrolyte (5 ◦ C min−1 heating rate, temperature range
30–380 ◦ C; argon atmosphere).
to 150 ◦ C. The weight loss was found less than 0.7 wt.% up
to 90 ◦ C and ca. 4.5 wt.% under 150 ◦ C. These changes can be
explained as a partial evaporation of the immobilised solvent (PC
and EC), however the change is small and is corresponding with
low vapour pressure of both solvents (pPC = 130 Pa at 50 ◦ C).
Both in air and argon all samples start to decompose at tem-
peratures above ca. 160–170 ◦ C. Fig. 8 presents simultaneous
TGA and DTA analysis of PEOEMA–PC–BMIPF6 (50:40:10,
mol%) gel electrolyte. The weight loss is ca. 80 wt.% and is
caused by evaporation of solvent (boiling point of PC is 240 ◦ C
at 101.3 kPa) and decomposition of the PEOEMA polymer net-
work and the ionic liquid. Similar behaviour was observed in the
case of samples with lithium salts. Both processes are exother-
mic and correspond to two peaks on the DTA curve on Fig. 8.

4. Conclusions

Our research showed, that combination of a hydrophobic

Fig. 7. Single potential impedance spectra measured after the cell being assem-
bled (full dots) and after voltammetrical measurements of PEOEMA–PC–
BMIPF6 (empty dots). Inserted is the detail of measured high frequencies (exper-
imental conditions: three-electrode arrangement; measuring potential E = 0 V vs.
Cd/Cd2+ ).
ionic liquid, 1-butyl-3-methylimidazolium hexafluorophosphate
or tetrafluoroborate with aprotic solvents, propylene and ethy-
lene carbonate strongly increases the ionic conductivity up to
9.4 and 14.1 mS cm−1 in comparison with the neat ionic liquid.
Accessible potential range of the liquid electrolytes BMIPF6 –PC
and BMIPF6 –PC/EC on the glassy carbon was found to be
4.6–4.7 V.
New polymer gel electrolytes based on 2-ethoxyethyl
methacrylate containing embedded solutions of BMIPF6 in
cyclic carbonates (PC and PC/EC) were prepared and charac-
terised by using electrochemical (impedance and voltammetri-
cal) and thermogravimetric techniques. Beside ternary systems
PEOEMA–PC–BMIPF6 and PEOEMA–PC/EC–BMIPF6 , new
materials containing lithium salts LiClO4 or LiPF6 were pre-
pared and studied.
Prepared materials exhibit high ionic conductivity due to
the high content of BMIPF6 in the polymer gel structure
 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408
(up to 19.2 wt.%). Highest conductivity value 0.94 mS cm−1
was reached in the case of PEOEMA–PC/EC–BMIPF6 –LiPF6
(49.2:19.6:19.6:9.8:1.8, mol%). Influence of composition and
temperature was studied both on liquid and polymer gel elec-
trolytes including determination of the conductivity activation
energy values.
Voltammetrical measurements showed wide accessible
potential window of prepared polymer gel electrolytes based
on ionic liquids, 4.3–4.4 V on glassy carbon electrode. All
polymer electrolytes are thermally stable up to 150 ◦ C and
both mechanical and electrochemical properties are stable for
weeks.
Three main functions of present carbonates in the polymer
electrolytes were proposed. PC and EC serve as plasticisers and
co-solvents of a polar ionic liquid in an aprotic polymer. The
positive effect on the material ionic conductivity and uniform
SEI formation agents are also utilized. Obtained results allow
the application of prepared materials in modern chemical power
sources. The quaternary systems containing lithium salts will
be studied as solid electrolytes in lithium-ion batteries and the
ternary systems will be combined with carbon powders of high
surface area to form supercapacitors. Described systems can
be used with materials like Li4 Ti5 O12 as anodic material while
coupled with high voltage cathode such as LiMn2 O4 , LiNiO2 or
LiCoO2 .
Acknowledgements
This work was supported by the Academy of Sciences
(Research Plan AV0Z40320502), by the Grant Agency of the
Academy of Sciences (Grant No. B400320701), by the Ministry
of Environment (Grant No. SN/171/05) and by the Ministry of
Education (project MSMT LC523).
References
[1] M. Alamgir, K.M. Abraham, in: G. Pistoia (Ed.), Lithium Batteries: New
Materials, Developments and Perspectives, vol. 5, Elsevier Science, Ams-
terdam, 1994 (Chapter 3).
[2] D. Linden, T.B. Reddy (Eds.), Handbook of Batteries, McGraw-Hill, New
York, 2002.
[3] D. Aurbach, Y. Talyosef, B. Markovsky, E. Markevich, E. Zini-
grad, L. Asraf, J.S. Gnanaraj, H.J. Kim, Electrochim. Acta 50 (2004)
247.
[4] M.B. Armand, J.M. Chabagno, M.J. Duclot, in: P. Vashishta, J.N. Mundy,
G.K. Shenoy (Eds.), Proceedings of the International Conference On Fast
Ion Transport in Solids, Electrodes and Electrolytes, North-Holland, New
York, 1979.
[5] M.B. Armand, Solid State Ionics 9/10 (1983) 745.
[6] M.B. Armand, Adv. Mater. 2 (1990) 278.
[7] J.Y. Song, Y.Y. Wang, C.C. Wan, J. Power Sources 77 (1999) 183.
[8] Y. Matoba, S. Matsui, M. Tabuchi, T. Sakai, J. Power Sources 137 (2004)
284.
[9] G.B. Appetecchi, F. Croce, B. Scrosati, Electrochim. Acta 40 (1995)
991.
[10] C.S. Harris, T.G. Rukavina, Electrochim. Acta 40 (1995) 2315.
[11] T. Iijima, Y. Toyoguchi, N. Eda, Denki Kogagu 53 (1985) 619;
T. Iijima, Y. Toyoguchi, N. Eda, Chem. Abstr. 103 (1985) 180931f.
[12] O. Bohnke, C. Rousselot, P.A. Gillet, C. Truche, J. Electrochem. Soc. 139
(1992) 1862.
1407
[13] J. Vondrák, M. Sedlařı́ková, J. Reiter, T. Hodal, Electrochim. Acta 44 (1999)
3067.
[14] J. Vondrák, J. Reiter, J. Velická, M. Sedlařı́ková, Solid State Ionics 170
(2004) 79.
[15] J. Vondrák, J. Reiter, J. Velická, B. Klápště, M. Sedlařı́ková, J. Dvořák, J.
Power Sources 146 (2005) 436.
[16] T. Sato, T. Maruo, S. Marukane, K. Takagi, J. Power Sources 138 (2004)
253.
[17] B. Garcia, S. Lavallée, G. Perron, C. Michot, M. Armand, Electrochim.
Acta 49 (2004) 4583.
[18] A. Lewandowski, M. Galiński, J. Phys. Chem. Solids 65 (2004)
281.
[19] P. Walden, Bull. Acad. Imer. Sci. (St. Peterburg) 1914 (1800).
[20] J.S. Wilkes, M.J. Zaworotko, J. Chem. Soc., Chem. Commun. (1992)
965.
[21] P. Wasserscheid, T. Welton, Ionic Liquids in Synthesis, Wiley–VCH, Wein-
heim, 2003.
[22] T. Welton, Coord. Chem. Rev. 248 (2004) 2459.
[23] L. Wen, A.G. Fadeev, B. Qi, B.R. Mattes, J. Electrochem. Soc. 151 (2004)
H33.
[24] M. Galin, A. Chapoton, J.-C. Galin, J. Chem. Soc., Perkin Trans. 2 (1993)
545.
[25] M. Yoshizawa, A. Narita, H. Ohno, Aust. J. Chem. 57 (2004) 139.
[26] M.A. Klingshirn, S.K. Spear, R. Subramanian, J.D. Holbrey, J.G. Huddle-
ston, R.D. Rogers, Chem. Mater. 16 (2004) 3091.
[27] H. Shobukawa, H. Tokuda, S.I. Tabata, M. Watanabe, Electrochim. Acta
50 (2004) 305.
[28] J. Reiter, J. Michálek, J. Vondrák, D. Chmelı́ková, M. Přádný, Z. Mička, J.
Power Sources 158 (2006) 509.
[29] J. Reiter, J. Michálek, J. Vondrák, D. Chmelı́ková, M. Přádný, in: C. Delmas
(Ed.), International Meeting on Lithium Batteries, Biarritz, France, June
18–23, 2005. Meeting Abstracts (abstract 345).
[30] P. Bonhôte, A.-P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel,
Inorg. Chem. 35 (1996) 1168.
[31] J. Fuller, A.C. Breda, R.T. Carlin, J. Electroanal. Chem. 459 (1998)
29.
[32] A. Lewandowski, A. Świderska, Solid State Ionics 169 (2004) 21.
[33] A. Noda, M. Watanabe, Electrochim. Acta 45 (2000) 1265.
[34] M. Morita, T. Shirai, N. Yoshimoto, M. Ishikawa, J. Power Sources 139
(2005) 351.
[35] R. Baskaran, S. Selvasekarapandian, G. Hirankumar, M.S. Bhuvaneswari,
J. Power Sources 134 (2004) 235.
[36] B.A. Boukamp, Solid State Ionics 18/19 (1986) 136.
[37] J. Vondrák, M. Sedlařı́ková, J. Velická, B. Klápště, V. Novák, J. Reiter,
Electrochim. Acta 48 (2003) 1001.
[38] J. Vondrák, J. Velická, B. Klápště, in: V. Novák (Ed.), Proceedings of the
Sixth Advanced Batteries and Accumulators, Brno, Czech Republic, June
5–9, 2005, 2005, p. 75 (Book of Abstracts).
[39] J. Reiter, J. Vondrák, Z. Mička, Electrochim. Acta 50 (2005) 4469.
[40] J. Reiter, J. Vondrák, Z. Mička, Solid State Ionics, submitted for publication.
[41] A. Balducci, U. Bardi, S. Caporali, M. Mastagostino, F. Soavi, Electrochem.
Commun. 6 (2004) 566.
[42] H. Tokuda, S.J. Baek, M. Watanabe, Electrochemistry 73 (2005) 620.
[43] C.S. Stefan, B. Montigny, D. Lemordant, in: C. Delmas (Ed.), International
Meeting on Lithium Batteries. Meeting Abstracts, Biarritz, France, June
18–23, 2006 (abstract 423).
[44] M. Egashira, S. Okada, J. Yamaki, D.A. Dri, F. Bonadies, B. Scrosati, J.
Power Sources 138 (2004) 240.
[45] M. Egashira, M. Maika, I. Watanabe, S. Okada, J. Yamaki, J. Power Sources
146 (2005) 685.
[46] M. Caillon-Caravanier, B. Claude-Montigny, D. Lemordant, G. Bosser,
Solid State Ionics 149 (2002) 275.
[47] A.J. Bard, L.R. Faulkner, Electrochemical Methods—Fundamentals and
Applications, J. Wiley and Sons, New York, 2001.
[48] V.R. Koch, L.A. Dominey, C. Nanjundiah, M.J. Ondrechen, J. Electrochem.
Soc. 143 (1996) 798.
[49] B. Rash, E. Cattaneo, P. Novák, W. Vielstich, Electrochim. Acta 36 (1991)
1397.
1408 J. Reiter et al. / Electrochimica Acta 52 (2006) 1398–1408

[50] P. Novák, P.A. Christensen, T. Iwasita, W. Vielstich, J. Electroanal. Chem.
263 (1989) 37.
[51] D. Aurbach, H. Gottlieb, Electrochim. Acta 34 (1989) 141.
[52] D. Aurbach, M. Daroux, P. Faguy, E. Yeager, J. Electroanal. Chem. 297
(1991) 225.
[53] S.A. Forsyth, J.M. Pringle, D.R. MacFarlane, Aust. J. Chem. 57 (2004)
113.
[54] M. Marzantowicz, J.R. Dygas, F. Krok, A. Lasinska, Z. Florjanczyk, E.
Zygadlo-Monikowska, Electrochim. Acta 51 (2006) 1713.