Water-based synthesis of hydrophobic ionic liquids for high-energy electrochemical devices

Article abstract of DOI:10.1016/j.electacta.2013.02.082

In this work is described an innovative synthesis route for hydrophobic ionic liquids (ILs) composed of N-methyl-N-Alkylpyrrolidinium (or piperidinium) or imidazolium or tetralkylammonium cations and (perfluoroalkylsulfonyl)imide, ((C_nF_2n+

Full text of DOI:10.1016/j.electacta.2013.02.082

Electrochimica Acta 96 (2013) 124–133
Contents lists available at SciVerse ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Water-based synthesis of hydrophobic ionic liquids for high-energy
electrochemical devices
Maria Montaninoa, Fabrizio Alessandrinia, Stefano Passerinib,∗, Giovanni Battista Appetecchia,∗∗
a ENEA, Agency for NewTechnologies, Energy and Sustainable Economic Development, UTRINN-IFC, Via Anguillarese 301, Rome 00123, Italy
b University of Muenster, Institute of Physical Chemistry, MEET Battery Research Centre, Correns Strasse 46, D-48149 Muenster, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
In this work is described an innovative synthesis route for hydrophobic ionic liquids (ILs) composed
of N-methyl-N-alkylpyrrolidinium (or piperidinium) or imidazolium or tetralkylammonium cations and
(perfluoroalkylsulfonyl)imide, ((CnF2n+1SO2)(CmF2m+1SO2)N−), anions. This synthesis does not require the
use of any environmental unfriendly solvent such as acetone, acetonitrile or halogen-containing com-
pounds, which is not welcome in industrial applications. Only water is used as the process solvent
throughout the entire process. In addition, the commonly used iodine-containing reagents were replaced
by the cheaper, more chemically stable and less toxic bromine-containing compounds. A particular care
was devoted to the development of the purification route, which is especially important for ILs to be used
in high-energy electrochemical devices such as high voltage supercapacitors and lithium batteries. The
effect of the reaction temperature, the time and the stoichiometry in the various steps of the synthesis
have been investigated in detail. This novel procedure allowed obtaining ultrapure (>99.9 wt.%), clear,
colourless, inodorous ILs with an overall yield above 92 wt.% and moisture content below 1 ppm. NMR
measurements were run to confirm the chemical structure whereas elemental analysis and electrochem-
ical tests were performed to check the purity of the synthesized ILs.
Received 14 September 2012
Received in revised form 15 February 2013
Accepted 15 February 2013
Available online 24 February 2013
Keywords:
Ionic liquid
Water-based synthesis
High purity
Lithium
Recycling
© 2013 Elsevier Ltd. All rights reserved.
1. Introduction
butyl), cation and the bis(trifluoromethanesulfonyl)imide, TFSI−,
or bis(fluorosulfonyl)imide, FSI−, have been showing very good
Room temperature ionic liquids (RTILs) are a very interest-
ing new class of room temperature fluids. The main advantages
with respect to organic solvents can be: non-flammability, negligi-
ble vapour pressure, high chemical, electrochemical and thermal
stability, high ionic conductivity and, in some cases, hydropho-
cycling reversibility with respect to lithium [14] and graphite
[15–18] anodes, and LiCoO2 cathodes [19]. Particularly, PYR14FSI-
LiTFSI mixtures have been recently employed as electrolytes
in Li4Ti5O12/LiFePO4 lithium-ion cells, which have displayed
very good cycling performance [20–22]. Also, incorporation of
PYR1ATFSI into PEO-based electrolytes was found to enhance the
room temperature ionic conductivity above 10−4 S cm−1, thus
allowing to overcome the conduction drawback of solvent-free
polymer electrolytes [23,24]. For instance, PEO-LiTFSI-PYR1ATFSI
ternary electrolyte systems were tested at low-medium tempera-
tures (from 20 to 40 ◦C) in dry, all-solid-state, Li/LiFePO4 polymer
batteries, which delivered large capacity with high reversibility and
very good cycling performance [22–25].
However, ionic liquids in the purity level required for elec-
trochemical applications are not widely available commercially.
Therefore, the electrochemistry community has devoted large
efforts for obtaining highly pure ionic liquid materials. Previous
work by MacFarlane et al. focused on the purification of ionic liquids
[26,27]. The most stringent requirement in designing the synthe-
sis of the hydrophobic ionic liquids is the solvent restriction. In
our laboratory it was previously developed a synthetic route for
PYR1ATFSI ionic liquids, which has involved the use of acetone
in the purification step of the PYR1AI precursor [28]. It is known,
bicity. Therefore, RTILs have attracted
a large attention for
use as “green” solvents for chemical reactions, bi-phasic catal-
ysis, chemical synthesis, separations [1–5]. Also, RTILs have
been extensively investigated for applications as advanced high-
temperature lubricants and transfer fluids in solar thermal energy
systems. More recently, they are under wide investigation as
electrolytes (or electrolyte components) in the place of haz-
ardous and volatile organic compounds for the realization of
highly safe electrochemical devices including rechargeable lithium
batteries, fuel cells, double-layer capacitors, hybrid supercapa-
citors, photoelectrochemical cells [6–22]. Ionic liquid electrolytes
based on the N-methyl-N-alkylpyrrolidinium, PYR1A+ (A = propyl or
∗
Corresponding author. Tel.: +49 251 8336725; fax: +49 2518336797.
Corresponding author.
E-mail addresses: stefano.passerini@uni-muenster.de (S. Passerini),
∗∗
gianni.appetecchi@enea.it (G.B. Appetecchi).
0013-4686/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.electacta.2013.02.082

M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
125
exchange deionizer. Typically, ionic liquid batches ranging from
100 g to 250 g were prepared.
A one litre, glass, sealed reactor was consecutively used for the
synthesis of the precursor (e.g., N-butyl-N-methylpyrrolidinium
bromide, PYR14Br) and the ionic liquid, this avoiding to expose the
reagent materials to the external. The operating temperature of the
reactor was controlled using an oil bath connected to the reactor.
The filtration steps were performed using a Millipore vacuum fil-
ter system through hydrophilic, polyamide membranes whose pore
size was lower than 0.2 ␮m. An oil-free pump (<1 mTorr) was used
to generate vacuum.
The separation of the rinsing fluid (water phase) from the ionic
liquid was performed by aspiration generated by a water pump. The
water residual was removed using a Heidolph LaboRota 4000 rotary
evaporator connected with a KNF oil-free pump. The final drying
step of the ionic liquid was performed in a dry room (R.H. < 0.1%)
using a Büchi B-585 glass vacuum oven connected with a Leybold
PT-70 turbo-molecular pump (10−6 mTorr). The IL material was
dried at 20 ◦C for 2 h, then at 60 ◦C for additional 2 h and, succes-
sively, at 120 ◦C for at least 18 h. Finally, the ionic liquid was stored
in vacuum-sealed glass tubes.
The concentration of lithium cation (Li+) and bromide anion
(Br−) in the ionic liquid sample was checked by a Spectro ARCOS
ICP-OES (Spectro Analytical Instruments) instrument with axial
plasma view. A standard Fassel type torch (No. 75160526, Spectro
Analytical Instruments) was employed. The sample was introduced
using a peristaltic pump system with a cross flow nebulizer and a
double-pass spray chamber (Scott type). Elements were simulta-
neously detected at different individual emission lines.
The water content was measured using the standard Karl Fisher
method. The titrations were performed by an automatic Karl Fisher
coulometer titrator (Mettler Toledo DL32) located inside the dry
room. The Karl Fisher titrant was a one-component (Hydranal
34836 Coulomat AG) reagent provided from Aldrich.
The chemical structure of the ionic liquid was checked by
NMR measurements using an Avance III spectrometer working
at 200.13 MHz on 1H (Bruker, Rheinstetten). The spectra were
recorded with a broadband probe BBFO (Bruker, Rheinstetten).
Deuterated DMSO, D6 (>99%, Aldrich), was used as the solvent.
The peak assignment was made on the basis of the chemical shifts
(expressed in parts per million, ppm) and peak integrals with DMSO
(1H NMR ı = 2.49 ppm) signal as the reference.
The electrochemical stability window of the raw ionic liquid
and its mixtures with a lithium salt was evaluated by linear sweep
voltammetries (LSVs) at 5 mV s−1. A sealed, three-electrode, glass
micro-cell, described in details previously [31], was used for the
LSV tests. The cell was loaded with a small amount of the ionic
liquid sample (about 0.5 ml). A glass-sealed, platinum working
electrode (active area = 0.78 mm2) and a platinum foil counter elec-
trode (about 0.5 cm2) were used. The reference electrode was a
silver wire immersed in a 0.01 M solution of AgCF3SO3 (Aldrich) in
PYR14TFSI, separated from the cell compartment with a fine glass
frit [31]. High purity argon (3 ppm water and 2 ppm oxygen) was
flown over the ionic liquid sample under investigation for 30 min
before the start of the test. The gas flow was continued during the
experiment. Separate LSV tests were carried out to determine the
cathodic and anodic electrochemical stability limits. The measure-
ments were performed scanning the cell potential from the open
circuit potential (OCP) towards more negative (cathodic limit) or
positive (anodic limit) voltages. Clean electrodes were used for each
LSV test, which was performed at least twice on different fresh sam-
ples in order to confirm the results obtained. Cyclic voltammetry
(CV) experiments were run with the same cell structure under the
analogous operative conditions to investigate both the presence
of irreversible reactions associated to impurities and the lithium
stripping/plating process. The potential values are given vs. the
Scheme 1. Chemical structure of the PYR14TFSI ionic liquid.
however, that acetone is an undesirable solvent for industrial appli-
cations since its high volatility. In addition, tests carried out in our
laboratory have indicated that acetone traces remaining in ionic
liquids can form unwelcome, yellowish compounds with unpleas-
ant odour when in contact with the materials (i.e., activated carbon
and alumina) used for the purification of the final product, even at
moderate temperatures (>50 ◦C).
Successively, we have proposed an improved procedure for
the synthesis of hydrophobic room temperature ionic liquids,
composed of PYR1A+ or piperidinium (PIP1A+) cations and (perfluo-
roalkylsulfonyl)imide (CnF2n+1SO2)(CmF
2m+1SO2)N−), PFSI, anions,
which involves ethyl acetate (EtAc) and water as the solvents [29].
EtAc and H2O are safer and more environmentally friend than ace-
tone and, in particular, than acetonitrile and halogen-containing
solvents generally used for IL synthesis processes [11,12,30].
In the present work we have further optimized the previous
synthesis route. Only water is used throughout the entire process.
In addition, alkyl iodine reagents were replaced by cheaper, more
chemically stable and less toxic alkyl bromine compounds. A par-
ticular care was devoted in the purification route of the chemical
reagents and precursors. The effect of the reaction temperature,
the time and the stoichiometry in the various steps of the synthe-
sis has been investigated in detail. NMR measurements were run
to confirm the chemical structure whereas elemental analysis and
electrochemical tests were performed to check the purity of the
synthesized ionic liquid material. The results, referred to the syn-
thesis procedure route of the PYR14TFSI ionic liquid (Scheme 1), are
reported in the present paper.
2. Experimental
N-methylpyrrolidine (PYR1, Acros, 98 wt.%) and 1-bromobutane
(1-Br-But, Aldrich, 99 wt.%) were previously purified through acti-
vated carbon (Aldrich, Darco-G60) and alumina (Aldrich, acidic,
Brockmann I) whereas lithium bis(trifluoromethanesulfonyl) imide
salt, LiTFSI (3 M, 99.9 wt. %), activated carbon and alumina were
used as received. H2O was deionized by a Millipore ion resin

126
M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
Fig. 1. Dependence of the PYR14Br (precursor) synthesis yield as a function of the
processing time at different temperatures and LiTFSI excess amounts.
Fig. 2. Yield of the PYR14Br precursor purification step as
water/(AC + Al2O3) weight ratio used for the rinsing of the purifying materials. The
Al2O3/AC and (AC + Al2O3)/PYR14Br weight ratios are also indicated.
a function of the
Ag/Ag+ redox couple. The LSV and CV measurements were per-
formed (in the dry room) at 20 ◦C using a Schlumberger (Solartron)
Electrochemical Interface (model 1287) controlled by a software
developed at ENEA.
3.1.1. Purification of chemicals
In order to enhance the purity of the resulting ionic liquid, a
purification step was run on the neat reagents before the synthetic
route. PYR1 and 1-Br-But were previously purified through acti-
vated carbon (AC) and alumina (Al2O3) whereas LiTFSI, provided
with a purity higher than 99.9%, was used as received.
Activated carbon (0.30 g and 0.15 g of AC per gram of PYR1 and
1-Br-But, respectively) and alumina (0.30 g and 0.15 g of Al2O3
per gram of PYR1 and 1-Br-But, respectively) were loaded in a
glass reactor containing PYR1 or 1-Br-But. The resulting slurry was
stirred or ball-milled for 12–20 h at room temperature. Then, AC
and Al2O3 were separated from the liquid fraction (containing the
purified reagent) by vacuum filtration (oil-free pump) with PTFE fil-
ter (pore size < 0.2 ␮m). Finally, the liquid fraction, constituted by
clear and slightly coloured PYR1 or 1-Br-But, was collected for the
synthesis of PYR14TFSI. No solvents were added during the purifi-
cation of the reagents. A fraction lower than 10 wt.% of the two
reagents was physically trapped into the purifying materials. This
fraction could be recovered by rinsing the separated AC and Al2O3
with water (PYR1) or water and ethanol (1Br-But).
3. Results and discussion
The synthesis here proposed was developed considering the
use of water as the only solvent. No other compounds were used
throughout the synthesis. Work was devoted in determining the
minimum excess of reagents and the reaction conditions (temper-
ature and time) necessary to optimize the yield, thus producing the
minimum amount of waste. Butyl bromine was used in the place
of the more expensive, more oxygen- and light-sensitive, and more
toxic butyl iodine [29]. Particular care was taken in the purification
route, which was carried out on both the chemical reagents (with
the exception of LiTFSI) and the precursor (instead of the final ionic
liquid), e.g., before the addition of the most expensive chemical
(LiTFSI). This allowed to reduce the final cost of the ionic liquid and
avoid the use of any organic solvent for the ionic liquid dilution.
Finally, the possibility of chemical recycling was investigated. The
procedure route was also optimized for the use of a single reactor.
In the following the results of the investigation performed to
optimize the synthesis of PYR14TFSI, starting from PYR1, 1-Br-But
and LiTFSI, are reported. The effect of the processing temperature
and time, as well as the weight ratio of reagents and the amount
of purifying materials and solvents, was investigated. The results
are illustrated in Figs. 1–4. The weight of the reagents as well as
the purification materials and deionized water used in the overall
synthesis process of a 250-g batch is reported in Table 1. However,
the same procedure has been successfully carried out for the syn-
3.1.2. Synthesis of PYR14Br precursor
The precursor, PYR14Br, was synthesized from PYR1 and 1-Br-
But through the following reaction performed in deionized water:
PYR1(aqueous) + 1 − Br → PYR14Br(aqueous)
(1)
+
+
thesis of several other ionic liquids composed of PYR1A or PIP1A
or imidazolium or tetraalkylammonium cations and PFSI anions.
3.1. Synthetic route of the PYR14TFSI ionic liquid
The synthesis of the hydrophobic PYR14TFSI ionic liquid through
the aqueous route was performed through five iterative steps: (i)
purification of the chemicals; (ii) synthesis of the PYR14Br precur-
sor; (iii) purification of the PYR14Br precursor; (iv) synthesis of the
PYR14TFSI ionic liquid; (v) rinsing of the PYR14TFSI ionic liquid. The
overall synthesis process is schematized in Fig. 5 (panels from A
through D).
Fig. 3. Yield vs. processing time dependence of the Br−/TFSI− anion exchange step
at different LiTFSI excess amounts (vs. the stoichiometric amount). T = 20 ◦C.

M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
127
Fig. 4. Summary of the yields and losses detected through the entire PYR14TFSI aqueous synthesis route. The data given for each synthesis step are referred to the expected
yield for that step. The overall process yield is reported in the last panel.
PYR1 density = 0.819 g cm−3) whereas the upper one was mainly
a PYR1/water mixture. The two phases (typically, clear and slightly
coloured) were mixed (by magnetic stirring) at suitable temper-
ature and time (Fig. 1). The progressive disappearance of the
lower phase (and the simultaneous increase of the upper one) has
given a clear indication of the ongoing formation of the PYR14Br
Purified PYR1 was previously dissolved in deionized water
(PYR1/H2O volume ratio equal to 1/1) and loaded into the glass reac-
tor. Then, the appropriate amount of purified 1-Br-But was added.
The latter chemical is immiscible with water and, therefore, a liquid
phase separation was observed. The lower phase was mostly com-
posed by the heavier 1-Br-But (1-Br-But density = 1.276 g cm−3
;
Table 1
Weight of all chemical used in the overall aqueous synthesis route for a 250 g PYR14TFSI ionic liquid batch.
Chemicals
Weight/g
Notes
(i) Purification of chemicals
PYR1
1-Br-But
Activated carbon
Alumina
Purified PYR1(product)
Purified 1-Br-But (product)
56.0
88.8
30.1
30.1
54.9
87.9
Impurity content: 2 wt.%
Impurity content: 1 wt.%
AC/PYR1 = 0.30 AC/1-Br-But = 0.15
Al2O3/PYR1 = 0.30 Al2O3/1-Br-But = 0.15
(ii) Synthesis of the PYR14Br precursor
Purified PYR1
Purified 1-Br-But
54.9
87.9
67.0
0.5 wt.% PYR1 excess
PYR1/1-Br-But weight ratio = 0.62
PYR1/H2O volume ratio = 1/1
Step yield: about 100 mol.%
H2O
PYR14Br in aqueous solution
142.5
(about 260 g of PYR14Br per 100 g of PYR1)
Process yield: about 100 mol.%
(iii) Purification of the PYR14Br precursor
PYR14Br (aqueous)
H2O
Activated carbon
Alumina
142.5
67.0
85.5
128.3
641.0
139.5
From the synthesis step
Extra dilution to reduce viscosity
60 g of carbon per 100 g of PYR14Br
90 g of alumina per 100 g of PYR14Br
Extraction of PYR14Br from the filtered solids
Step yield: about 98 mol.%
H2O
Purified PYR14Br in aqueous solution
Process yield: about 98 mol.%
(254 g of PYR14Br per 100 g of PYR1).
The final solution contains 708 ml of H2O.
(iv) Synthesis of the PYR14TFSI ionic liquid
Purified PYR14Br(in aqueous solution)
LiTFSI (99.9 wt.%)
139.5
185.5
From the purification step
3 wt.% LiTFSI excess
(133 g of LiTFSI per 100 g of PYR14Br)
Step yield: 94.3 mol.%
PYR14TFSI
250
(about 180 g of PYR14TFSI per 100 g of PYR14Br)
Process yield: 92.3 mol.%
(455g of PYR14TFSI per 100 g of PYR1)
(v) Rinsing of the PYR14TFSI ionic liquid
PYR14TFSI
H2O
Purified PYR14TFSI
250
From the synthesis step
H2O/PYR14TFSI volume ratio = 1/1 for each single rinse
Step yield: about 100 mol.%
186(5×)
250
Process yield: 92.3 mol.%
(455 g of PYR14TFSI per 100 g of PYR1)

128
M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
Fig. 5. Schematic representation of the synthesis process of the PYR14TFSI, ionic liquid. Panel A: purification of chemicals. Panel B: synthesis of the PYR14Br precursor. Panel
C: purification of PYR14Br. Panel D: synthesis and rinsing of PYR14TFSI. Thick arrows indicate solids or slurries, thin arrows indicate liquids.
precursor (soluble in water and, therefore, in the upper phase). The
completion of the reaction (1) has been easily detected by com-
plete disappearance of the lower phase, e.g., a transparent, aqueous
solution of PYR14Br was obtained as final product. Although reac-
tion (1) could be performed by directly mixing the two reagents
in a reactor, water (which is a good solvent for PYR1 but not for
1-Br-But) was used as diluent to promote the formation of the sol-
uble precursor. It is to note that PYR14Br (as well as butyl bromine)
is much less sensitive to light and oxygen than PYR14I [29] and,
therefore, it may be easily handled without particular precau-
tions.
The effect of the temperature, time and PYR1 excess (vs. the
stoichiometric amount of 1-Br-But) on the yield of the precursor
synthesis process was investigated. Several trail tests were car-
ried out under different operative conditions. After each test, the
liquid compound was massively removed through vacuum distil-
lation (by rotary evaporator at 90 ◦C for 3 h) to obtain a white,
crystalline precipitate of PYR14Br. Then, the solid precursor was
dried (in the dry room) under vacuum at 90 ◦C overnight and, suc-
cessively, weighted. The results are summarized in Fig. 1. The data
clearly show that the processing temperature largely affects the
kinetic of the PYR14Br synthetic route. At room temperature (20 ◦C),

M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
129
prolonged reaction times (24 h) were required to increase the yield
up to reasonable values (70 wt.%) with a 5 wt.% PYR1 excess. At
higher temperatures (≥60 ◦C) more than 80 wt.% of the stoichio-
metric amount of PYR14Br was obtained in relatively short times.
At 60 ◦C (5 wt.% PYR1 excess) yield values equal to 85 wt.% and
approaching 100 wt.% were achieved after 1 and 2 h process, respec-
tively. At 70 ◦C about 80 wt.% and 100 wt.% of the stoichiometric
PYR14Br amount were obtained after 15 and 25 min process, respec-
tively. In addition, a PYR1 excess ten times lower (0.5 wt.%) was
used, this allowing to reduce the cost for chemicals and purification.
Higher processing temperatures were not investigated for safety
reasons (e.g., to avoid to get close to the boiling temperature of the
compounds). Therefore, the operative conditions to synthesize the
precursor were fixed at 70 ◦C (temperature) and 25 min (reaction
time) with a PYR1 excess equal to 0.5 wt.%. (e.g., allowing to drive
the synthesis process of the PYR14Br precursor with a 100% yield).
It is to note that the reaction of PYR1 with bromoalkyl compounds
exhibits slower kinetics with respect to the iodoalkyl reagents [29]
(e.g., the removal of bromide from haloalkyl compounds is slower
with respect to iodide [32,33]), thus leading to higher process tem-
peratures. However, this results in enhanced safety of the synthesis
route of the precursor, especially when using haloalkyl compounds
with long (haloesane and longer) or short (haloethane and shorter)
chain since the reaction with PYR1 is much more exothermic (e.g.,
the removal of halide from long chain haloalkyl compounds is eas-
ier since the higher stability of the carbocation [32,34] whereas
the reduced steric hindrance of the carbocation from short chain
haloalkyl compounds results in faster addition to the nitrogen of
the pyrrolidinium ring) [29,33].
turned from uncoloured to slightly yellowish. The liquid fractions,
collecting upon rinsing, were vacuum distilled (in rotary evapora-
tor at 90 ◦C for 3 h) to remove most of H2O (from the precursor).
Then, the (white) PYR14Br precursor was dried under vacuum (by a
oil-free pump) at 90 ◦C overnight and, successively, weighted. The
results are displayed in Fig. 2, which reports the yield, in terms
of the expected PYR14Br stoichiometric amount, as a function of
the water/(AC + Al2O3) weight ratio used for rinsing the purifying
materials. Both Al2O3/AC and (AC + Al2O3)/PYR14Br weight ratios
were fixed to 1.5. The first data point, about 70 mol.%, represents
the fraction of precursor directly obtained by the vacuum filtra-
tion of the AC + Al2O3 slurry. The rinsing of the purifying materials
allowed to enhance the yield of the purification step up to about
97 mol.% for a water/(AC + Al2O3) weight ratio of 3.0. Repeated tests
performed on different size batches (the precursor was fully dried
and weighed) indicated a reproducible yield of the purification test
close to 97 mol.%. No relevant amount of PYR14Br was recovered
for further water addition (Fig. 2) whereas a more extensive rins-
ing led to a slightly yellowish colouration of the liquid fraction, this
indicating the partial release of impurities by carbon and alumina.
Therefore, the overall amount of deionized water used for the rins-
ing step was equal to 641 g whereas the final volume of the aqueous
solution, containing approximately 140 g of PYR14Br precursor, was
about 770 ml.
3.1.4. Synthesis of the PYR14TFSI ionic liquid
PYR14TFSI was synthesized from aqueous PYR14Br and LiTFSI by
the following reaction:
PYR14Br(aqueous) + LiTFSI(solid) → PYR14TFSI(liquid)
3.1.3. Purification of the PYR14Br precursor
+ LiBr(aqueous)
(2)
The purification of the precursor was performed through AC and
Al2O3 which were found able to trap impurities [29]. The suitable
amount of purifying materials (Table 1) was loaded into the reac-
tor containing the aqueous PYR14Br solution (previously diluted
with deionized water to reduce the viscosity) obtained from the
synthesis step of the precursor. The resulting slurry was continu-
ously stirred at room temperature (20 ◦C) for 3–4 h. Then, AC and
Al2O3 were separated from the liquid fraction by vacuum filtration
(performed using cellulose filter paper). The liquid fraction (consti-
tuted by a clear and colourless aqueous solution of PYR14Br) was
collected for the synthesis of PYR14TFSI whereas the solid fraction
(still loaded in the filtering glassware) was rinsed (Fig. 5) in situ with
deionized water to remove PYR14Br trapped through AC and Al2O3.
The liquid fraction from the rinsing step (constituted by a clear and
colourless diluted aqueous solution of PYR14Br) was combined with
the previous one (from the filtering step).
The amount of AC and Al2O3 used to purify the PYR14Br pre-
cursor as well the temperature and the stirring time was selected
on the basis of an iterative process in which the amount of the
purifying materials was increased until to obtain a complete dis-
coloration of the PYR14Br/H2O solution. The optimum C/PYR14Br
and Al2O3/PYR14Br weight ratios were found to be 0.6 and 0.9,
respectively (Table 1). Higher weight ratios are not recommended
because the fraction of precursor obtained during the first filtration
is reduced and larger amounts of deionized water are required for
the further extraction of PYR14Br. Therefore, the purification step
was run at room temperature (20 ◦C), e.g., no relevant temperature
effect was observed, with stirring time ranging from 3 to 4 h. The
operative conditions selected as above were found to give good
results for all precursors synthesized.
LiTFSI was added, e.g., in slight excess with respect to the
stoichiometric amount (Fig. 3), to the aqueous PYR14Br solution
obtained from the purification step, resulting in the dissolution
of the former immediately followed by its reaction with PYR14Br
(through anion exchange which replaces Br− with TFSI−) to form
hydrophobic PYR14TFSI and hydrophilic LiBr. The rapid formation
of two liquid phases clearly indicated that the anion exchange reac-
+
tion, driven by the intrinsic hydrophobicity of both the PYR14
cation and the TFSI− anion, proceeded quickly. For instance, both
PYR14+ and TFSI−, in which the charge is well shielded by hydropho-
bic groups (PYR14+) or extensively delocalized (TFSI−), do not easily
form hydrogen bonds with water molecules and tend to separate
from the aqueous phase forming a second (denser) liquid phase.
Therefore, the disappearance itself of the (ionic liquid) PYR14TFSI
product from the aqueous solution drove the exchange reaction to
completion. Then, the two liquid phases were vigorously stirred at
room temperature to facilitate the anion exchange reaction. After
a selected time (Fig. 3) the stirring was interrupted and the phase
separation took place in a few minutes. The upper phase (clear and
colourless) was mostly composed of water, lithium bromide (LiBr)
and LiTFSI excess whereas the lower one (uncoloured and generally
slightly cloudy) was mostly constituted of PYR14TFSI ionic liquid
with lithium salts (i.e., LiBr and LiTFSI) and traces of water.
The dependence of the process yield vs. exchange reaction time,
for different LiTFSI excess amounts with respect to the stoichio-
metric amount, is reported in Fig. 3. After 10 min about 91% of the
expected stoichiometric amount of PYR14TFSI was obtained with
an excess of LiTFSI equal to 1 wt.%. After 30 min, the yield of the
anion exchange process levelled to 92 mol.% whereas no practical
increase of the yield was observed at longer process times (1 h). An
increase of the LiTFSI excess up to 3 wt.% allowed to obtain a yield
equal to 94 mol.% even after 1 min reaction whereas no relevant
increase was detected at longer process time (1 h). Previous work
The water amount used for the rinsing step of AC and Al2O3
needs to be accurately dosed in order to recover most of the pre-
cursor without extracting the impurities trapped. Therefore, the
solid fraction (AC and Al2O3) was progressively treated in situ
(rinsed) with fixed amounts of water until the extracted solution

130
M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
[29] has demonstrated that further increases of the LiTFSI excess
led to a decrease of the anion exchange yield. This behaviour is due
to the fact that, even if PYR14TFSI is insoluble in water (at room
temperature), its anion (TFSI−) bounds very strongly to Li+ cations
(from LiBr and LiTFSI excess) [35,36]. The strong coordination of
lithium ions by water molecules and TFSI− anions is the driving
force causing the partial dissolution of PYR14TFSI in the aqueous
phase. The higher is the amount of lithium in the aqueous phase
the lower is the yield of the anion exchange process. These results
suggested the use of a very limited excess of LiTFSI, which, in addi-
tion, is the most expensive reagent used in the ionic liquid synthesis
process. Therefore, the anion exchange reaction time was limited
up to 2–3 min whereas the LiTFSI excess was fixed equal to 3 wt.%.
of the carbon and alumina waste (which is fully recycled) and of
the aqueous phase waste (from the anion exchange).
The precursor trapped on the surface of the purifying materials
was recovered through extensive rinsing (of the carbon and alu-
mina waste) with deionized water. A fixed amount of AC and Al2O3
(loaded in a glass reactor) was quickly stirred with H2O for 1 h at
room temperature. The H2O/(AC and Al2O3) weight ratio was fixed
equal to 2 for each rinsing step. Then, the solid fraction (AC and
Al2O3) was separated from the aqueous phase by vacuum filtration
(performed using polyamide filter having a pore size lower than
0.2 ␮m) whereas the liquid fraction (clear and almost uncoloured
diluted aqueous solution of PYR14Br) was separately collected. The
rinsing steps on the purifying materials were iteratively repeated
until no trace of PYR14Br (revealed from the bromide anion) was
detected in the aqueous phase. The presence of Br− in the aque-
ous phase upon each rinsing step was checked with AgNO3. For
instance, 1.0 ml of the aqueous phase was added to 1.0 ml of 0.1 N
AgNO3 in water solution. A pale-yellow, solid precipitate indicated
formation of AgBr. After a few rinsing steps, just a turning of the
resulting solution from clear to opaque was observed. Therefore,
the solution was centrifuged for a few minutes to better evidence
the AgBr precipitate. An overall amount equal to 1 litre of H2O per
100 g of carbon and alumina was required for a complete removal
of PYR14Br, e.g., no AgBr precipitate was observed in the aqueous
phase upon the last rinsing step, this resulting in an estimated Br−
concentration lower than 1 ng dm−3. The PYR14Br precursor, recov-
ered from the purifying materials, was separated from the aqueous
liquid fractions by vacuum distillation (90 ◦C). The above-described
route, schematized in panel A of Fig. 6, allowed recovering about
100% of the precursor lost in the purification step. Finally, AC and
Al2O3 from the rinsing steps were vacuum annealed at 200 ◦C for
3 h to fully remove impurities.
The ionic liquid lost during the anion exchange step was recov-
ered by vacuum distillation (90 ◦C) of the aqueous phase waste (the
vapor pressure of LiBr, LiTFSI and PYR14TFSI is not measurable).
After water removal, a slurry composed of ionic liquid and a white
precipitate of LiBr (mostly undissolved in PYR14TFSI at room tem-
perature) was obtained, which was, then, vacuum filtered (using
Teflon filter having a pore size lower than 0.2 ␮m) to separate the
lithium salt. Successively, the ionic liquid was rinsed (10 min stir-
ring at room temperature) a few times with deionized water to
remove the fraction of LiBr (and LiTFSI) dissolved in PYR14TFSI. The
H2O/PYR14TFSI volume ratio was fixed equal to 1/1 in each rinsing
step. As well as reported above for the rinsing of the purifying mate-
rials, the presence of Br− in the aqueous phase (removed by vacuum
aspiration upon each rinsing step) was checked with 0.1 N AgNO3
in water. After six consecutive rinsing steps, no practical amount
of bromide was detected. This route (panel B of Fig. 6) allowed to
recover practically the 100% of the ionic liquid (and LiBr) lost in the
anion exchange step.
3.1.5. Rinsing of the PYR14TFSI ionic liquid
The strong coordination of Li+ and TFSI− is also the reason for
the presence of lithium salts (i.e., unreacted LiTFSI and LiBr) in the
ionic liquid phase. Therefore, after removal of the upper aqueous
phase, the PYR14TFSI ionic liquid was rinsed five consecutive times
with deionized water to remove water-soluble salts (e.g., LiBr and
LiTFSI excess) and impurities. The water/ionic liquid volume ratio
was fixed to 1/1 for each rinsing. In previous work [29] we have
demonstrated that this rinsing route allows reducing the Li+ con-
centration in the ionic liquid below 3 ␮g dm−3 (2 ppb). The rinsed
PYR14TFSI was vacuum filtered (over polyamide filter having pore
size < 0.2 ␮m) to remove solid residual (mostly AC and Al2O3 parti-
cles from the purification step) and, then, vacuum distilled (rotary
evaporator) at 90 ◦C for 2–3 h (to remove most of water). Succes-
sively, PYR14TFSI was dried (within the dry room) at 20 ◦C for 1 h,
then at 60 ◦C for 3 h and, finally, at 120 ◦C for 18 h (the progressive
heating avoids fast bubbling of the residual water, this resulting in
ionic liquid spying out off the glass container). The ionic liquid was
stored in vacuum-sealed glass tubes within the dry room.
In Fig. 4 are summarized the yield and loss values, reported in
mol.%, of the entire synthesis process as well as of each step. About
100% of the expected stoichiometric amount was obtained from
the PYR14Br precursor synthesis using a very small PYR1 excess
(0.5 wt.%). The purification step allowed to minimize the PYR14Br
loss (2.1 mol.%); more than 68 mol.% of precursor was obtained
directly through the first filtration whereas about 30 mol.% was
recovered through carbon and alumina rinsing. The anion exchange
process showed a yield of 94.3 mol.% using a moderate excess
(3 wt.%) of LiTFSI. The loss is due to the enhanced ionic force of
the aqueous phase due to the presence of LiBr (1:1 mole ratio with
respect to the ionic liquid), which favours the ionic liquid disso-
lution in water [29]. Conversely, no detectable ionic liquid loss
was observed upon the (following) rinsing step of PYR14TFSI. To
summarize, above 92 mol.% of the expected stoichiometric amount
of PYR14TFSI was obtained through the entire aqueous synthetic
route, this value resulting higher than the overall yield detected in
ethyl acetate solvent (86 mol.%) [29]. More important, the aqueous
route (with respect to the organic one) allowed to reduce the loss of
the expensive fluorinated anion (TFSI) to about one third (5.7 mol.%
vs. 16.3 mol.%).
3.3. Analysis of the PYR14TFSI ionic liquid
The aqueous route described above allowed synthesizing clear
and colourless PYR14TFSI ionic liquid with water content lower
than 1 ppm. However, we have observed that glass tubes (in which
are generally housed ionic liquids) are capable to deliver mois-
ture to ionic liquid material. To demonstrate this issue, we have
dried a PYR14TFSI sample, synthesized in external environment
and housed in a glass tube, following the protocol described above
(within the dry room). After the drying step, the humidity con-
tent of the ionic liquid sample was checked, resulting below 1 ppm.
Then, the glass tube containing PYR14TFSI was immediately closed,
vacuum-sealed in a pouch envelope (previously evacuated for 1 h)
and kept in the dry room. The moisture content of the ionic liq-
uid sample was periodically checked. The results (not shown here)
The overall synthesis route of the PYR14TFSI ionic liquid is
schematized in Fig. 5 whereas the weight of all chemicals used is
reported in details in Table 1.
3.2. Recycling of chemicals
The ionic liquid losses through the aqueous synthesis route were
localized in the precursor purification (2.1 mol.% of PYR14Br trapped
on the surface of carbon and alumina) and anion exchange (about
5.7 mol.% of PYR14TFSI was lost by dissolution in water) steps. How-
ever, these fractions may be recovered by appropriate processing

M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
131
Fig. 6. Schematic representation of the recycling route of PYR14Br lost during the purification step (panel A) and PYR14TFSI lost during the anion exchange step. The recycling
of the purifying materials waste and processing water was also investigated. Thick arrows indicate solids or slurries, thin arrows indicate liquids.
showed a progressive increase of the humidity amount with the
storage time, achieving 30 ppm and 180 ppm after a storage period
of 48 days and 20 months, respectively. This suggests a continu-
ous release of water from the glass tube, which, therefore, has to
be carefully dried and stored in a controlled environment before
loading with ionic liquid.
NMR measurements were performed to verify the correct struc-
ture of the synthesized ionic liquid. The results, reported in Fig. 7,
indicate that the PYR14TFSI material was correctly prepared. Ele-
mental analysis has revealed an extremely low content of lithium
cation and bromide anion, e.g., below 1 ppb (Li+) and 0.8 ppm (Br−),
respectively
of linear sweep voltammetries (LSVs) performed on PYR14TFSI
obtained from aqueous route (solid line) to investigate the anodic
and the cathodic electrochemical stability at 20 ◦C. The anodic cur-
rent flow observed at voltages above 2.0 V (vs. Ag/Ag+) corresponds
to the TFSI− anion oxidation whereas the cathodic one recorded at
-3.8 (vs. Ag/Ag+) is related to the PYR14+ cation reduction. No other
features were observed during the anodic and cathodic scans, thus
excluding the presence of impurities which are oxidized or reduced
within the anodic and cathodic electrochemical stability limits of
the ionic liquid, showing an ESW exceeding 5 V. On the other hand,
an analogous ionic liquid sample from organic solvent route [29]
(dotted line) exhibits some current flow during the cathodic scan
above -1.7 V (vs. Ag/Ag+), i.e., well above the extensive decom-
position (reduction) of the PYR14TFSI material. The nature of the
processes taking place is not well known but it is to be related with
the reduction of impurities. Therefore, the aqueous route allows
obtaining ionic liquids with higher purity.
The purity of the ionic liquid has been investigated also by means
of electrochemical measurements. Fig. 8, panel A, reports the result
The addition of LiTFSI (previously dried under vacuum at 120 ◦C
for 24 h, PYR14+/Li+ mole ratio = 9/1) resulted in the appearance of
current features (panel B of Fig. 8) in the voltage (cathodic) region
ranging from −1.5 V to −3.3 V (vs. Ag/Ag+). In previous work [15]
we have demonstrated that these features are due to LiTFSI salt
impurities which act as catalyst for the reduction of the TFSI−
anion. Cathodic features, detected prior the lithium plating onto
platinum working electrodes, were observed in lithium salt liquid
electrolytes also by Compton et al. [37]. These impurities, how-
ever, are eliminated during the PYR14TFSI synthesis route (e.g.,
PYR14TFSI rinsing step) since no similar profiles were evidenced in
the cathodic stability window of the pure ionic liquid. It is impor-
tant to note, however, that lithium metal is plated on the platinum
working electrode, as indicated by the cathodic peak shown at
−3.8 V (vs. Ag/Ag+), prior the ionic liquid decomposition (reduc-
tion). At more cathodic potentials the current associated to the
Fig. 7. NMR spectrum of a PYR14TFSI ionic liquid sample synthesized through the
aqueous route. NMR chemical shifts are also reported.

132
M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
4. Conclusions
An innovative and improved route has been developed
to synthesize hydrophobic, ionic liquids composed of N-
methyl-N-alkylpyrrolidinium (or piperidinium) or imidazolium
or tetralkylammonium cations and (perfluoroalkylsulfonyl)imide,
((CnF2n+1SO2)(CmF
2m+1SO2)N−), anions. The preparation procedure
requires water as the only solvent and is suitable of lab-scale and
industrial applications. Particular care was devoted to the purifi-
cation steps and the recycling of chemicals was investigated. The
process allowed to obtain clear, colourless, inodorous, high purity
ionic liquid (>99.9 mol.%) with a water content below 1 ppm and
an overall yield above 92 mol.%. Electrochemical measurements
performed on three-electrode cells containing PYR14TFSI and its
mixtures with LiTFSI as the electrolyte have confirmed the high
purity of the ionic liquid by showing the possibility to reversibly
reduce lithium ions to lithium metal. This ionic liquid is viable for
the realization of next generation batteries based on lithium metal
anodes and elemental cathodes such as sulphur or air.
Acknowledgement
The authors wish to thank the financial support of the
European Commission within the FP6 STREP Projects ILHY-
POS (Contract TST4-CT-2005–518307) and ILLIBATT (Contract
NMP3-CT-2006-033181) and FP7 Project LABOHR (Contract FP7-
2010-GC-ELECTROCHEMICAL-STORAGE 265971).
Fig. 8. Linear sweep voltammetries (LSVs) of pure PYR14TFSI (panel A) synthesized
through the aqueous route (solid line) and VOC route (dotted line, from Ref. [29]).
In panel B are compared the LSVs of pure PYR14TFSI (aqueous route, solid line) and
its mixture with LiTFSI (PYR14+/Li+ = 9, dotted line). Magnifications of the LSVs are
offered in the inserts. Platinum was used as the working and the counter electrode,
respectively. The reference electrode was a silver wire immersed in a 0.01 M solution
of AgCF3SO3 in PYR14TFSI. Scan rate: 5 mV s−1. T = 20 ◦C. The potential is given vs. the
Ag/Ag+ redox couple.
References
[1] P. Wasserscheid, W. Keim, Ionic liquids - New “solutions” for transition metal
catalysis, Angewandte Chemie International Edition 39 (2000) 3772.
[2] R.D. Rogers, K.R. Seddon (Eds.), Ionic Liquids-Industrial Applications to Green
Chemistry;. ACS Symposium Series 818, Oxford University Press, Oxford, 2002.
[3] M.J. Earle, K.R. Seddon, Ionic liquids. Green solvents for the future, Pure and
Applied Chemistry 72 (2000) 1391.
[4] J.L. Anderson, J. Ding, T. Welton, D.W. Armstrong, Characterizing ionic liquids on
the basis of multiple solvation interactions, Journal of the American Chemical
Society 124 (2002) 14247.
[5] J. Dupont, R.F. de Souza, P.A.Z. Suarez, Ionic liquid (molten salt) phase
organometallic catalysis, Chemical Reviews 102 (2002) 3667.
[6] A.I. Bhatt, I. May, V.A. Volkovich, M.E. Hetherington, B. Lewin, R.C. Thied, N.
Ertok, Group 15 quaternary alkyl bistriflimides: ionic liquids with potential
application in electropositive metal deposition and as supporting electrolytes,
Journal of the Chemical Society, Dalton Transactions 24 (2002) 4532.
[7] S. Panozzo, M. Armand, O. Stephan, Light-emitting electrochemical cells using
a molten delocalized salt, Applied Physics Letters 80 (2002) 679.
[8] P. Wang, S.M. Zakeeruddin, I. Exnar, M. Gratzel, High efficiency dye-sensitized
nanocrystalline solar cells based on ionic liquid polymer gel electrolyte, Chem-
ical Communications 22 (2002) 2972.
lithium plating process assumed the limiting current value until
+
the extensive reduction of the PYR14 cations took place at −4.5 V
(vs. Ag/Ag+). If the cathodic limit was set above such a value, the
lithium plating process was found to be reversible as shown by
the cyclic voltammetry presented in Fig. 9. Such a good lithium
plating-stripping reversibility makes the LiTFSI-PYR14TFSI mixture
an appealing electrolyte for high safety lithium batteries.
[9] J. Fuller, A.C. Breda, R.T. Carlin, Ionic liquid-polymer gel electrolytes from
hydrophilic and hydrophobic ionic liquids, Journal of Electroanalytical Chem-
istry 459 (1998) 29.
[10] H. Nakagawa, S. Izuchi, K. Kunawa, T. Nukuda, Y. Aihara, Liquid and polymer gel
electrolytes for lithium batteries composed of room-temperature molten salt
doped by lithium salt, Journal of the Electrochemical Society 150 (2003) A695.
[11] H.
Sakaebe,
H.
Matsumoto,
N-Methyl-N-propylpiperidinium
bis(trifluoromethanesulfonyl) imide (PP 13-TFSI) – novel electrolyte base for
Li battery, Electrochemistry Communications 5 (2003) 594.
[12] A. Noda, M.A.B.H. Susan, K. Kudo, S. Mitsushima, K. Hayamizu, M. Watanabe,
Bronsted acid–base ionic liquids as proton-conducting nonaqueous elec-
trolytes, Journal of Physical Chemistry B 107 (2003) 4024.
[13] A. Balducci, R. Dugas, P.L. Taberna, P. Simon, D. Plée, M. Mastragostino, S.
Passerini, High temperature carbon–carbon supercapacitor using ionic liquid
as electrolyte, Journal of Power Sources 20 (2007) 922.
[14] G.-T. Kim, G.B. Appetecchi, M. Montanino, F. Alessandrini, S. Passerini, Long-
term cyclability of lithium metal electrodes in ionic liquid-based electrolytes
at room temperature, ECS Transactions 25 (2010) 127.
[15] G.B. Appetecchi, M. Montanino, A. Balducci, S.F. Lux, M. Winter, S. Passerini,
Lithium insertion in graphite from ternary ionic liquid-lithium salt electrolytes
I. Electrochemical characterization of the electrolytes, Journal of Power Sources
192 (2009) 599.
[16] S.F. Lux, M. Schmuck, G.B. Appetecchi, S. Passerini, M. Winter, A. Balducci,
Lithium insertion in graphite from ternary ionic liquid-lithium salt electrolytes:
II. Evaluation of specific capacity and cycling efficiency and stability at room
temperature, Journal of Power Sources 192 (2009) 606.
Fig. 9. Lithium/plating process of a PYR14TFSI-LiTFSI (PYR14+/Li+ = 9/1) mixture on
a platinum working electrode at 20 ◦C. Platinum was also used as the counter elec-
trode. The reference electrode was a silver wire immersed in a 0.01 M solution of
AgCF3SO3 in PYR14TFSI. Scan rate: 5 mV s−1. T = 20 ◦C. The potential is given vs. the
Ag/Ag+ redox couple.

M. Montanino et al. / Electrochimica Acta 96 (2013) 124–133
133
[17] A. Guerfi, S. Duchesne, Y. Kobayashi, A. Vijh, K. Zaghib, LiFePO4 and graphite
electrodes with ionic liquids based on bis(fluorosulfonyl)imide (FSI)(-) for Li-
ion batteries, Journal of Power Sources 175 (2008) 866.
[28] W.A. Henderson, S. Passerini, Phase behavior of ionic liquid-LiX mixtures:
pyrrolidinium cations and TFSI-anions, Chemistry of Materials 16 (2004)
2881.
[18] T. Sugimoto, Y. Atsumi, M. Kikuta, E. Ishiko, M. Kono, M. Ishikawa, Ionic liquid
electrolyte systems based on bis(fluorosulfonyl)imide for lithium-ion batteries,
Journal of Power Sources 189 (2009) 802.
[29] G.B. Appetecchi, C. Tizzani, S. Scaccia, F. Alessandrini, S. Passerini, Synthesis
of hydrophobic ionic liquids for electrochemical applications, Journal of the
Electrochemical Society 153 (2006) A1685.
[19] H. Matsumoto, H. Sakaebe, K. Tatsumi, M. Kikuta, E. Ishiko, M. Kono,
Fast cycling of Li/LiCoO2 cell with low-viscosity ionic liquids based on
bis(fluorosulfonyl)imide [FSI], Journal of Power Sources 160 (2006) 1308.
[20] G.T. Kim, S.S. Jeong, M. Joost, E. Rocca, M. Winter, S. Passerini, A. Balducci, UV
cross-linked, lithium-conducting ternary polymer electrolytes containing ionic
liquids, Journal of Power Sources 195 (2010) 6130.
[21] S.F. Lux, F. Schappacher, A. Balducci, S. Passerini, M. Winter, Low Cost,
Environmentally Benign Binders for Lithium-Ion Batteries, Journal of the Elec-
trochemical Society 157 (2010) A320.
[22] G.-T. Kim, S.S. Jeong, M.-Z. Xue, A. Balducci, M. Winter, S. Passerini, F. Alessan-
drini, G.B. Appetecchi, Development of ionic liquid-based lithium battery
prototypes, Journal of Power Sources 199 (2012) 239.
[23] J.-H. Shin, W.A. Henderson, G.B. Appetecchi, F. Alessandrini, S. Passerini, Recent
developments in the ENEA lithium metal battery project, Electrochimica Acta
50 (2005) 3859.
[30] D.R. MacFarlane, P. Meakin, J. Sun, N. Amini, M. Forsyth, Pyrrolidinium imides:
a new family of molten salts and conductive plastic crystal phases, Journal of
Physical Chemistry 103 (1999) 4164.
[31] S. Randstrom, M. Montanino, G.B. Appetecchi, C. Lagergren, A. Moreno,
S. Passerini, Effect of water and oxygen traces on the cathodic stability
of N-alkyl-N-methylpyrrolidinium bis(trifluoromethanesulfonyl)imide, Elec-
trochimica Acta 53 (2008) 6397.
[32] N.L. Allinger, M.P. Cava, D.C. De Jongh, C.R. Johnson, N.A. Lebel, C.L. Stevens,
Organic Chemistry, 2nd edition, Worth Publishers Inc. Editor, New York, 1976.
[33] G.B. Appetecchi, M. Montanino, D. Zane, M. Carewska, F. Alessandrini,
S. Passerini, Effect of the alkyl group on the synthesis and the electro-
chemical properties of N-alkyl-N-methyl-pyrrolidinium bis(trifluoromethane-
sulfonyl)imide ionic liquids, Electrochimica Acta 54 (2009) 1325.
[34] R.T. Morrison, R.N. Boyd, Organic Chemistry, 3rd edition, Allyn and Bacon Inc.
Editors, Boston, 1973.
[24] G.-T. Kim, G.B. Appetecchi, F. Alessandrini, S. Passerini, Solvent-free, PYR1ATFSI
ionic liquid-based ternary polymer electrolyte systems I. Electrochemical char-
acterization, Journal of Power Sources 171 (2007) 861.
[25] G.B. Appetecchi, G.T. Kim, M. Montanino, F. Alessandrini, S. Passerini, Room
temperature lithium polymer batteries based on ionic liquids, Journal of Power
Sources 196 (2011) 6703.
[26] B.R. Clare, P.M. Bayley, A.S. Best, M. Forsyth, D.R. MacFarlane, Purification or
contamination? The effect of sorbents on ionic liquids, Chemical Communica-
tions 23 (2008) 2689.
[27] B. Clare, A. Sirwardana, D.R. MacFarlane, Synthesis, Purification and Character-
ization of Ionic Liquids, Topics in Current Chemistry 290 (2009) 1.
[35] M. Castriota, T. Caruso, R.G. Agostino, E. Cazzanelli, W.A. Henderson, S. Passerini,
Raman investigation of the ionic liquid N-methyl-N-propylpyrrolidinium
bis(trifluoromethanesulfonyl) imide and its mixture with LiN(SO2CF3)2, Journal
of Physical Chemistry A 192 (2005) 92.
[36] I. Nicotera, C. Oliviero, W.A. Henderson, G.B. Appetecchi, S. Passerini, NMR
investigation of ionic liquid-LiX mixtures: pyrrolidinium cations and TFSI−
anions, Journal of Physical Chemistry B 109 (2005) 2218.
[37] C.A. Paddon, R.G. Compton, Kinetics and thermodynamics of the Li/Li+ cou-
ple in tetrahydrofuran at low temperatures (195–295 K), Journal of Physical
Chemistry C 111 (2007) 9016.