Phase behavior and crystalline phases of ionic liquid-lithium salt mixtures with 1-alkyl-3-methylimidazolium salts

Article abstract of DOI:10.1021/cm902691v

The thermal phase behavior of 1-alkyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide (IM_{10 R}TFSI where R = 1, 2, or 4 for methyl, ethyl or butyl, respectively) ionic liquid binary mixtures with LiTFSI have been investigated as models for electrolytes for lithium batteries. Diverse phase behavior is found with significant variations noted from similar mixtures in which the imidazolium cations are replaced with N-alkyl-N- methylpyrrolidinium cations. The crystal structure for a (1-x) IM ₁₀₁TFSI-(x) LiTFSI (x = 0.50) (or 1/1 IM₁₀₁TFSI/LiTFSI) phase is reported to further clarify the molecular level interactions occurring in these binary salt mixtures.

Full text of DOI:10.1021/cm902691v

Chem. Mater. 2010, 22, 1203–1208 1203
DOI:10.1021/cm902691v
Phase Behavior and Crystalline Phases of Ionic Liquid-Lithium Salt
Mixtures with 1-Alkyl-3-methylimidazolium Salts†
Qian Zhou, Kendall Fitzgerald, Paul D. Boyle, and Wesley A. Henderson*
Ionic Liquid & Electrolytes for Energy Technologies (ILEET ) Laboratory, Department of
Chemical & Biomolecular Engineering, North Carolina State University, Raleigh, North Carolina 27695
Received August 31, 2009
The thermal phase behavior of 1-alkyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide
(IM10RTFSI where R = 1, 2, or 4 for methyl, ethyl or butyl, respectively) ionic liquid binary mix-
tures with LiTFSI have been investigated as models for electrolytes for lithium batteries. Diverse
phase behavior is found with significant variations noted from similar mixtures in which the
imidazolium cations are replaced with N-alkyl-N-methylpyrrolidinium cations. The crystal structure
for a (1-x) IM101TFSI-(x) LiTFSI (x = 0.50) (or 1/1 IM101TFSI/LiTFSI) phase is reported to
further clarify the molecular level interactions occurring in these binary salt mixtures.
Introduction
TFSI- anion (IM10RTFSI with R = 1, 2, or 4 for methyl,
ethyl, or butyl, respectively)
Ionic liquids (ILs) may ultimately prove to be an
indispensable component of advanced electrolytes for
the lithium batteries intended for plug-in hybrid electric
vehicle (PHEV) and electric vehicle (EV) traction. To
reduce the cost, raise the energy density and improve the
safety of large-format lithium battery packs for PHEVs
and EVs, there is an increasing emphasis on the use of
high-voltage, high-energy electrodes requiring electro-
lytes that are stable above 4.3 V (vs Li/Liþ) and have
high thermal stability. The general formulation for cur-
rent state-of-the-art electrolytes, which consist of a mix-
ture of ethylene carbonate (EC) and a linear carbonate
such as diethyl carbonate (DEC) with LiPF6,1 used in
small Li-ion batteries has varied only marginally over the
past 20 years and is inadequate for high-voltage electrode
materials. ILs are one of the few classes of materials
available that may have a suitably high electrochemical
and thermal stability for these new batteries. At present,
however, there is only limited information available
regarding the phase behavior of IL-LiX mixtures2-7
despite the tremendous insight into electrolyte properties
such information provides.
These ILs were not selected with the intent of optimiz-
ing practical battery electrolytes as the imidazolium ca-
tions are well-known to have an unsuitably low
electrochemical stability (both cathodic and anodic).8,9
These ILs and the analogous N-alkyl-N-methylpyrroli-
dinium (PY1RTFSI) ILs, however, are the most widely
studied ILs for which the greatest amount of informa-
tion regarding their properties is available.2,3,10-21 It is,
(8) Matsumoto, H.; Sakaebe, H.; Tatsumi, K. J. Power Sources 2005,
146, 45.
(9) Matsumoto, H.; Sakaebe, H.; Tatsumi, K.; Kikuta, M.; Ishiko, E.;
Kono, M. J. Power Sources 2006, 160, 1308.
ꢀ
(10) Lassegues, J.-C.; Grondin, J.; Talaga, D. Phys. Chem. Chem. Phys.
2006, 8, 5629.
(11) Duluard, S.; Grondin, J.; Bruneel, J.-L.; Pianet, I.; Grelard, A.;
ꢁ
Campet, G.; Delville, M.-H.; Lassegues, J.-C. J. Raman Spectrosc.
ꢀ
2008, 39, 627.
(12) Umebayashi, Y.; Mitsugi, T.; Fukuda, S.; Fujimori, T.; Fujii, K.;
Kanzaki, R.; Takeuchi, M.; Ishiguro, S. J. Phys. Chem. B 2007, 111,
13028.
This study focuses on mixtures of lithium bis-
(trifluoromethanesulfonyl)imide (LiTFSI) with ILs con-
sisting of 1-alkyl-3-methylimidazolium cations and the
ꢀ
(13) Lassegues, J.-C.; Grondin, J.; Aupetit, C.; Johansson, P. J. Phys.
Chem. A 2009, 113, 305.
(14) Borodin, O.; Smith, G. D.; Henderson, W. A. J. Phys. Chem. B
2006, 110, 16879.
† Accepted as part of the 2010 “Materials Chemistry of Energy Conversion
Special Issue”.
*Corresponding author. E-mail: whender@ncsu.edu.
(15) Umebayashi, Y.; Yamaguchi, T.; Fukuda, S.; Mitsugi, T.; Takeuchi,
M.; Fujii, K.; Ishiguro, S. Anal. Sci. 2008, 24, 1297.
(16) Fujii, K.; Fujimori, T.; Takamuka, T.; Kanzaki, R.; Umebayashi,
Y.; Ishiguro, S. J. Phys. Chem. B 2006, 110, 8179.
(1) Xu, K. Chem. Rev. 2004, 104, 4303.
ꢀ
(17) Lassegues, J.-C.; Grondin, J.; Holomb, R.; Johansson, P. J. Raman
(2) Henderson, W. A.; Passerini, S. Chem. Mater. 2004, 16, 2881.
(3) Matsumoto, K.; Hagiwara, R.; Tamada, O. Solid State Sci. 2006, 8,
1103.
Spectrosc. 2007, 38, 551.
(18) Fujii, K.; Soejima, Y.; Kyoshoin, Y.; Fukuda, S.; Kanzaki, R.;
Umebayashi, Y.; Yamaguchi, T.; Ishiguro, S.; Takamuku, T.
J. Phys. Chem. B 2008, 112, 4329.
(4) Burba, C. M.; Rocher, N. M.; Frech, R.; Powell, D. R. J. Phys.
Chem. B 2008, 112, 2991.
(5) Lerner, H.-W.; Bolte, M. Acta Crystallogr., Sect. E 2003, 59, m625.
(6) MacFarlane, D. R.; Forsyth, M. Adv. Mater. 2001, 13, 957.
(7) MacFarlane, D. R.; Huang, J.; Forsyth, M. Nature 1999, 402, 792.
(19) Fukuda, S.; Takeuchi, M.; Fujii, K.; Kanzaki, R.; Takamuku, T.;
Chiba, K.; Yamamoto, H.; Umebayashi, Y.; Ishiguro, S. J. Mol.
Struct. 2008, 143, 2.
r
2009 American Chemical Society
Published on Web 12/04/2009
pubs.acs.org/cm

1204 Chem. Mater., Vol. 22, No. 3, 2010
Zhou et al.
therefore, quite informative to compare the phase beha-
vior of IM10RTFSI-LiTFSI mixtures with that of
PY1RTFSI-LiTFSI mixtures.2 Note that although aprotic
solvent-LiTFSI electrolytes are known to corrode Al
current collectors,22-26 there is evidence to suggest that
such corrosion by the TFSI- anion in ILs is greatly
reduced or inhibited.27-29
calibrated with cyclohexane (solid-solid phase transition
at -87.06 °C, melt transition at 6.54 °C) and indium (melt
transition at 156.60 °C). Hermetically sealed Al pans were
prepared in the glovebox. Typically, sample pans were slowly
cooled (5 °C min-1) to -150 °C and then heated (5 °C min-1) to
100 °C or higher (determined by the composition of the sample).
In some cases, it was necessary to hold or cycle the samples at
various subambient temperatures in the instrument prior to the
measurements to ensure complete crystallization. Some of the
sample pans were also stored in a freezer prior to analysis and
then transferred into the instrument quickly (to aid in crystal-
lization for slowly nucleating samples).
Experimental Section
Materials. 1-Methylimidazole (99%), iodomethane (99%),
iodoethane (99%), and 1-iodobutane (99%) were purchased
from Aldrich and used as received. LiTFSI was purchased from
3M. The LiTFSI was dried under a vacuum at 120 °C for 12 h
before use.
X-ray Structural Determination. Single crystals of the of the
1/1 (x = 0.50) IM101TFSI/LiTFSI phase formed in the x =
0.20 sample vial after storage at room temperature. A crystal
was mounted on a nylon loop with a small amount of Paratone
N oil. All X-ray measurements were made on a Bruker-Nonius
Kappa Axis X8 Apex2 diffractometer at a temperature of
-163 °C. Data collection was carried out using Mo KR
radiation. The unit cell dimensions were determined from a
symmetry constrained fit of 9889 reflections with 4.56° < 2θ <
76.28°. The data collection strategy was a number of ω and
j scans that collected data up to 76.58° (2θ). The frame
integration was performed using SAINT.30 The resulting raw
data was scaled and absorption corrected using a multiscan
averaging of symmetry equivalent data using SADABS.31 The
structure was solved by direct methods using the SIR92
program.32 All non-hydrogen atoms were obtained from the
initial solution. The hydrogen atoms were introduced at idea-
lized positions and were allowed to ride on the parent atom.
The structural model was fit to the data using full matrix least-
squares based on F2. The calculated structure factors included
corrections for anomalous dispersion from the usual tabula-
tion. The structure was refined using the XL program from
SHELXTL.33 Crystallographic figures were drawn using
Mercury 2.2 software.
Sample Preparation. The 1-alkyl-3-methylimidazolium salts
were prepared by combining 1-methylimidazole with a stoichio-
metric amount of alkyliodide in ethyl acetate (Aldrich). The
resulting white or yellow crystalline IM10RI salts were repeatedly
washed using ethyl acetate. The IM10RI salts were then dissolved
in deionized water, stoichiometric amounts of LiTFSI dissolved
in deionized water were added, and the mixtures were stirred.
The aqueous phase with dissolved LiI was removed and the
remaining ILs were washed 8 times with deionized water to
remove reagents and contaminants from the IM10RTFSI salts.
The final aqueous layers were removed and the salts were heated
while stirring on a hot plate for several hours at 120-140 °C to
remove most of the residual water. Activated carbon (Darco-
G60, Aldrich) was added and the mixtures were stirred on a hot
plate at 140 °C for 48 h. The mixtures were then cooled to room
temperature; ethyl acetate was added to dilute the solutions, and
the solutions were then filtered through an activated alumina
(acidic, Brockmann I, Aldrich) column. The ethyl acetate was
then removed by a rotary evaporator and the salts were dried
under high vacuum at 120 °C overnight and then at 140 °C for
another 8 h. After purification and drying, the ILs were clear,
colorless liquids at room temperature. The materials were stored
in hermetically sealed bottles in a N2 glovebox (water concen-
tration <0.5 ppm).
Results and Discussion
DSC heating traces for (1-x) IM101TFSI-(x) LiTFSI
mixtures are shown in Figure 1a. Adding LiTFSI initially
slightly lowers the melting point (Tm) of the samples from
that of the neat IM101TFSI salt (Tm 24 °C). The data
indicate that new mixed-salt crystalline phases exist for
x = 0.50 and what appears to be a 0.75 composition
(corresponding to 1/1 and 1/3 IM101TFSI/LiTFSI
phases, respectively). 1/1 crystalline phases have been
previously reported for IM102CF3SO3/LiCF3SO3,
Et3NHCF3SO3/LiCF3SO3, and IM102AlCl4/NaAlCl4
mixtures.4,5,34 In all of these structures, each anion is
coordinated to two Liþ cations. Each Liþ cation, in turn,
is coordinated by four different anions (4-fold co-
(1-x)IM10RTFSI-(x) LiTFSI mixtures were prepared in the
glovebox by combining appropriate amounts of the IM10RTFSI
and LiTFSI salts in vials and then heated on a hot plate while
stirring to form homogeneous mixtures. The resulting IL-LiX
mixtures were clear and colorless with a moisture content below
10 ppm (tested by Karl Fischer titration). The materials were
stored in hermetically sealed glass vials in the glovebox.
Thermal Measurements. Thermal measurements were per-
formed using a TA Instruments Q2000 differential scanning
calorimeter (DSC) with liquid N2 cooling. The instrument was
(20) Umebayashi, Y.; Mitsugi, T.; Fujii, K.; Seki, S.; Chiba, K.;
ꢁ
Yamamoto, H.; Canongia Lopes, J. N.; Padua, A. A. H.; Takeuchi,
M.; Kanzaki, R.; Ishiguro, S. J. Phys. Chem. B 2009, 113, 4338.
ꢁ
(21) Canongia Lopes, J. N.; Shimizu, K.; Padua, A. A. H.; Umebayashi,
ordination) resulting in polymeric [Liþ (anion)2]n
Y.; Fukuda, S.; Fujii, K.; Ishiguro, S. J. Phys. Chem. B 2008, 112,
1465.
(22) Zhang, X.; Devine, T. M. J. Electrochem. Soc. 2006, 153, B375.
(23) Behl, W. K.; Plichta, E. J. J. Power Sources 1998, 72, 132.
(24) Morita, M.; Shibata, T.; Yoshimoto, N.; Ishikawa, M. Electro-
chim. Acta 2002, 47, 2787.
3 3 3
chains. For electrostatic neutrality, the organic cations
(either IM102þ or Et3NHþ cations) surround these ionic
(25) Kanamura, K.; Umegaki, T.; Shiraishi, S.; Ohashi, M.; Takehara,
Z. J. Electrochem. Soc. 2002, 149, A185.
(26) Zhang, S. S.; Jow, T. R. J. Power Sources 2002, 109, 458.
(27) Garcia, B.; Armand, M. J. Power Sources 2004, 132, 206.
(28) Peng, C.; Yang, L.; Zhang, Z.; Tachibana, K.; Yang, Y. J. Power
Sources 2007, 173, 510.
(30) SAINT version 2008.6; Bruker-Nonius: Madison, WI, 2008.
(31) SADABS version 2008.4; Bruker-Nonius: Madison, WI, 2008.
(32) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.;
Burla, M. C.; Polidori, G.; Camalli, M. J. Appl. Crystallogr.
1994, 27, 435.
(33) SHELXTL version 6.12; Bruker-AXS: Madison, WI.
(34) Boon, J. A.; Carlin, R. T.; Elias, A. M.; Wilkes, J. S. J. Chem.
Crystallogr. 1995, 25, 57.
(29) Peng, C.; Yang, L.; Zhang, Z.; Tachibana, K.; Yang, Y.; Zhao, S.
Electrochim. Acta 2008, 53, 4764.

Article
Chem. Mater., Vol. 22, No. 3, 2010 1205
Figure 1. DSC heating traces (5 °C min-1) of (1-x) IM10RTFSI-(x) LiTFSI mixtures (a) R = 1, (b) R = 2, and (c) R = 4 (peak and Tg temperatures
noted).
chains and have a single hydrogen bond to the anions.
The crystal structure of the 1/1 (x = 0.50) IM101TFSI/
LiTFSI phase (Table 1 and Figure 2) bears some resem-
is coordinated through a hydrogen bond to one anion
oxygen, whereas the cation C4 and C5 hydrogens are both
hydrogen bonded to another anion oxygen. One of these
hydrogen also may have an additional hydrogen bond to
another oxygen from a third anion (Figure 4). Note that
oxygen atoms from addition anions are also closely
positioned over the positively charged ring nitrogen
atoms (not shown in Figure 4), adding further stability
to this manner of ion packing. It is notable that all of the
blance to these structures. Polymeric [Liþ (TFSI-)2]n
3 3 3
chains are present with each anion coordinated to two Liþ
cations. The difference is that the Liþ cations now have 6-
þ
fold coordination (Figure 3). The IM101 cations are
located in rows between these ionic polymer chains. The
hydrogen bound to the carbon at position C2 of the cation

1206 Chem. Mater., Vol. 22, No. 3, 2010
Zhou et al.
Table 1. Crystal and Refinement Data
Different anions are coordinated differently, but each
anion has either three or all four of the oxygens atoms
coordinated to Liþ cations, thereby creating the planar
sheets. The sheets (composed of Liþ cations and TFSI-
anions) result in the formaþtion of cavities in which reside
two uncoordinated IM102 cations (see the Supporting
Information).
For the (1-x) IM104TFSI-(x) LiTFSI mixtures, it was
not possible to crystallize the samples when x g 0.15 using
either slower cooling/heating rates (1 °C min-1) or by
annealing the mixtures for extended periods of time at low
temperature, despite extensive efforts to do so. Only a
fraction of the neat IM104TFSI phase could be crystal-
lized for dilute samples with x < 0.15 (Figure 1c). Thus,
no crystalline phase forms (up to a composition of x =
0.67) that contains the LiTFSI salt. The Tg of the x = 0.67
mixture is remarkably low for a sample composed of
2/3 lithium salt.
structure
[IM101TFSI]1[LiTFSI]1
chemical formula
fw
cryst syst
C9H9F12LiN4O8S4
664.38
triclinic
space group
temperature (K)
P1
110
˚
a (A)
5.3707(2)
11.7032(5)
17.9205(7)
88.6792(18)
84.198(2)
82.7165(17)
1111.51(8)
2
˚
b (A)
˚
c (A)
R (deg)
β (deg)
γ (deg)
3
˚
V (A )
Z
F
calcd (g cm-3
)
1.985
0.575
μ (mm-1
)
crystsize (mm3)
F(000)
0.38 ꢀ 0.33 ꢀ 0.18
660
76.58
2θmax (deg)
N (Rint
)
51556 (0.0248)
12123
0.0297, 0.0766
0.0411, 0.0829
1.005
N [I > 2σ(I)]
R1a, wR2b [I > 2σ(I)]
R1a, wR2b (all data)
GOFc
A comparison of the phase behavior of the IM10RTFSI
and PY1RTFSI salt mixtures with LiTFSI shows both
similarities and differences in behavior. For example,
there is no evidence that the IM10RTFSI salts form a 2/1
(x = 0.33) IL/LiTFSI phase, which is a predominant
feature of the PY1RTFSI mixtures.2 The reason for this is
unknown at present. There are no structural models
(crystal structures) known for such a composition to aid
in clarifying this. The IM102TFSI salt does, however,
form the 1/2 phase (x = 0.67), which is also a common
feature of the PY1RTFSI mixtures.2 The known structure
for this phase shows that cavities form in the
-3
˚
Δemin,max (e A
)
-0.369, 0.647
a R1 = Σ Fo| - |Fc /Σ|Fo|. b wR2 = [Σ[w(Fo2 - Fc2)2]/ Σ[w(Fo2)2]]1/2
c GOF = [Σ[w(Fo2 - Fc2)2]/(n - p)]1/2
.
.
anions have the C1 conformation,35-39 which is also
found for the anion conformations in crystalline IM101
-
TFSI.38 In the 1/1 phase with LiTFSI, this conformation
is favored as it enables the 6-fold coordination of
the anions to the Liþ cations (and IM101 hydrogen
þ
bonding), whereas in the neat IM101TFSI structure, this
conformation enables a favorable lattice of hydrogen
bonds to form between the cations and anions.
Liþ TFSI- anion planar lattice that hold two organic
3 3 3
The Tm of IM102TFSI is either -17 or -10 °C - the IL
is polymorphic with two different crystalline phases able
to form depending upon the salt0s thermal history.
Neither of these phases undergoes a solid-solid phase
transition prior to melting.40 DSC heating traces for
(1-x)IM102TFSI-(x) LiTFSI mixtures (Figure 1b) indi-
cate that as the LiTFSI mole fraction (x) increases, the
crystallization process becomes more difficult, and no
crystalline phase was observed for 0.25 e x e 0.50
mixtures. The glass transition (Tg) of the amorphous
samples tends to increase with increasing fraction of
LiTFSI, as expected. For samples with high concentra-
tion (x > 0.50), however, a new crystalline peak for a 1/2
(x = 0.67) IM102TFSI/LiTFSI phase is observed. The
crystal structure of this phase has been previously re-
ported and a portion of this structure is shown in the
Supporting Information.3 The ions of this 1/2 phase form
two-dimensional planar sheets stacked together. The
Liþ cations have 5-fold coordinated by anion oxygens
(from either three or four different TFSI- anions).
cations that remain essentially uncoordinated. It is easy to
envision that the cations can be readily replaced with
other cations of similar size to create the same phase with
a different IL. Interestingly, the IM104TFSI salt does not
form this phase. One possible explanation is that the
þ
IM104 cations are too large to fit within the cavities,
but this does not explain why the IM101TFSI salt alsþo
does not form this phase unless it is because the IM101
cations are too small and leave too much void space in the
structure. An alternative, and more likely, explanation,
however, is that the coordination in the high LiTFSI
composition phase (assumed to be x = 0.75) is more
energetically favorable than that of the 1/2 phase. The
crystal structures are low energy structures formed
by optimizing all of the interactions (packing of ions,
electrostatics, hydrogen bonding, etc.), not just the
Liþ TFSI- coordination.
3 3 3
In addition, the 1/1 phase forms readily for the IM101
-
TFSI salt, but the IM102TFSI and IM104TFSI salts
remain amorphous instead of forming this phase. There
is some evidence that the PY1RTFSI-LiTFSI mixtures
also form a 1/1 phase, although this appears to be a
metastable phase for some of the ILs.2 Once again, this
variance in behavior may be an ion size/shape effect that
prevents the ions from packing together in the optimal
manner to form the “preferred” structures dictated
(35) Foropoulos, J., Jr.; DesMarteau, D. D. Inorg. Chem. 1984, 23, 3720.
(36) Arnaud, R.; Benrabah, D.; Sanchez, J.-Y. J. Phys. Chem. 1996, 100,
10882.
(37) Johansson, P.; Gejji, S. P.; Tegenfeldt, J.; Lindgren, J. Electrochim.
Acta 1998, 43, 1375.
(38) Holbrey, J. D.; Reichert, W. M.; Rogers, R. D. Dalton Trans. 2004,
2267.
(39) Rey, I.; Johansson, P.; Lindgren, J.; Lassegues, J. C.; Servant, L.
ꢀ
by optimized Liþ TFSI- coordination and other
3 3 3
J. Phys. Chem. A 1998, 102, 3249.
(40) Unpublished data.
energetic considerations. The amorphous phase may then

Article
Chem. Mater., Vol. 22, No. 3, 2010 1207
Figure 2. Ion packinginthe 1/1(x= 0.50) IM101TFSI/LiTFSIcrystallinephase viewedalongthe (a) b-and(b)a-axis (Li, purple;O, red;S, gold;C, gray;F,
green).
were also compared with a [Li(TFSI)4]3- complex. The
conclusion was the Liþ cations exist in these ILs as
[Li(TFSI)2]- complexes in which the anions predomi-
nantly have an equal mix of C1 and C2 conformations.
This type of complex, however, requires that the TFSI-
anions have two of their oxygen atoms coordinated to the
same Liþ cation. If one closely scrutinizes the coordination
in the 1/2 IM102TFSI/LiTFSI phase, such coordination is
found for anions with both C1 and C2 symmetry (see the
Supporting Information), although the anions are also
coordinated to other Liþ cations (as required for such a
Figure 3. Ion coordination in the polymeric [Liþ (TFSI-)2]n chains in
the 1/1 (x = 0.50) IM101TFSI/LiTFSI crystalline phase (Li, purple; O,
red; S, gold; C, gray; F, green).
3 3
concentrated composition). The 1/1 (x = 0.50) IM101
-
TFSI/LiTFSI phase also has anions with two oxygen
atoms coordinating the same Liþ cation. A molecular
dynamics (MD) simulation of (1-x) PY1RTFSI-(x) LiTFSI
(x = 0.25, R = 1 and 3) at 227 °C, however, reported that
the probability that a single TFSI- anion contributes a
second oxygen atom to the coordination shell of the same
Liþ cation was <5%.14 The MD simulation work14 also
emphasized the formation of Liþ OdSdO Liþ co-
3 3 3
3 3 3
ordination that is present for all of the anions in the
known crystalline 1/1 and 1/2 phases (Figure 3 and the
Supporting Information). This form of coordination
facilitates the creation of polymeric aggregates. Such a
liquid structure might reasonably occur with short
aggregated ionic chains with anions having both C1 and
C2 conformations in which all of the anions have two
oxygens coordinated to two Liþ cations.
Given that the liquid phase is being considered, a
variety of forms of coordination may actually coexist.
At this time, therefore, it is not possible to fully confirm
the liquid phase Liþ cation coordination, but the interac-
tions between the cations and anions is what dictates the
ionic conductivity (mechanisms for Liþ cation transport)
and other physical properties for these electrolytes and
therefore is of considerable interest. It must be empha-
sized that the electrolyte components of an ideal battery
electrolyte should be unable to crystallize within the
battery operating temperature range as the capture of
þ
Figure 4. Space-filling model of anion hydrogen bonding to the IM101
cations in the 1/1 (x = 0.50) IM101TFSI/LiTFSI crystalline phase (Li,
purple; O, red; S, gold; C, gray; F, green).
be viewed as a “frustrated” phase in which many of the
ions in the frozen glass or supercooled liquid have the
same or similar coordination as found in the crystalline
phases of similar composition, but the entire ensemble of
the crystal structures cannot come together because of the
poor fit of the organic cations within such a structure.
It has been suggested that the addition of LiTFSI to
either IM102TFSI or IM104TFSI results in Liþ cations with
an anion solvation number of approximately two.10-12 A
spectroscopic analysis with DFT calculations of possible
[Li(TFSI)2]- complexes was conducted in which the
conformation of the anion was varied.13 These complexes

1208 Chem. Mater., Vol. 22, No. 3, 2010
Zhou et al.
the Liþ cations in a solid IL-LiX phase generally drasti-
cally reduces the conductivity of the electroactive Liþ
cations and thus the ability of the battery to operate.41
With this in mind, the phase information in panels b and c
in Figure 1 is intriguing, as concentrated electrolyte
mixtures may be prepared in which crystallization is
completely inhibited. If an explanation for why this
occurs is forthcoming, then such information may aid in
the optimization of future choices of ILs for battery
electrolytes.
crystallinity gaps in which it was not possible to crystallize
some or all of the mixtures. The reported crystal structure
of the 1/1 (x = 0.50) IM101TFSI/LiTFSI phase provides
crucial insight into ion coordination and packing, espe-
cially when compared to the known crystal structure of
the 1/2 (x = 0.67) IM102TFSI/LiTFSI phase. This in-
formation should aid in computational modeling of these
electrolytes, as well as deconvoluting wide variations in
the transport properties of the liquid phase of such
materials.
Acknowledgment. This material is based upon work sup-
ported by, or in part by, the U.S. Army Research Laboratory
and the U.S. Army Research Office under contract/grant
number W911NF-07-1-0556.
Conclusion
The thermal phase behavior of IM10RTFSI-LiTFSI
mixtures with imidazolium ILs are reported here and
compared with information known about the analogous
mixtures with pyrrolidinium ILs. Intriguing variations in
crystalline phases formed are noted, as is the presence of
Supporting Information Available: Experimental details and
schematic illustrations of crystal structures (PDF) and X-ray crystal-
lographic data (CIF) for [IM101TFSI]1[LiTFSI]1. This material is
available free of charge via the Internet at http://pubs.acs.org.
(41) Sakaebe, H.; Matsumoto, H. Electrochem. Commun. 2003, 5, 594.