Ultrafast dynamics in aromatic cation based ionic liquids: A femtosecond raman-induced kerr effect spectroscopic study

Article abstract of DOI:10.1246/bcsj.20160085

We studied the ultrafast dynamics of 40 aromatic cation based ionic liquids (ILs) by means of femtosecond Ramaninduced Kerr effect spectroscopy. The low-frequency Kerr spectra (ca. 0.3700 cm1) of the ILs were obtained from the Kerr transients by Fourier-Transform deconvolution analysis. The low-frequency Kerr spectra in the frequency range less than 200 cm^-1 coming mainly from the intermolecular vibrations for the ILs were discussed in terms of (i) anion dependence, (ii) imidazolium cations vs. pyridinium cations, (iii) alkyl group dependence, and (iv) effect of methylation in aromatic cations. Several liquid properties, such as density, viscosity, electrical conductivity, and surface tension, of the present sample ILs at 293K were also estimated in this study. We clarified that the aromatic cation based ILs show a different relation of the first moment of the low-frequency spectral band to the bulk liquid parameter, which is the square root of surface tension divided by liquid density, from aprotic molecular liquids. The slope of the first moment to the bulk parameter for the aromatic cation based ILs is gentler than that for aprotic molecular liquids.

Full text of DOI:10.1246/bcsj.20160085

Ultrafast Dynamics in Aromatic Cation Based Ionic Liquids:
A Femtosecond Raman-Induced Kerr Effect Spectroscopic Study
Hideaki Shirota,*1,2 Shohei Kakinuma, Kotaro Takahashi, Akito Tago,
1
1
1
2
3
Hocheon Jeong, and Tomotsumi Fujisawa
1Department of Nanomaterial Science, Graduate School of Advanced Integration Science, Chiba University,
-33 Yayoi, Inage-ku, Chiba 263-8522
1
2Department of Chemistry, Faculty of Science, Chiba University, 1-33 Yayoi, Inage-ku, Chiba 263-8522
3Department of Chemistry and Applied Chemistry, Graduate School of Science and Engineering, Saga University,
Honjo, Saga 840-8502
E-mail: shirota@faculty.chiba-u.jp
Received: March 14, 2016; Accepted: June 4, 2016; Web Released: June 13, 2016
Hideaki Shirota
Hideaki Shirota received his Ph.D. from the Graduate University for Advanced Studies in 1998 under the
supervision of Prof. Keitaro Yoshihara. His academic career started as a Research Associate at the
University of Tokyo in 1996 in advance of receiving his Ph.D. He then worked at the State University of New
Jersey, Rutgers and the University of Tokyo as a Postdoctoral Associate and a Research Associate. In 2006,
he joined Chiba University, as an Associate Professor of Chemistry. His current research interests include
molecular spectroscopy, laser spectroscopy, molecular dynamics in condensed phases, and reaction
dynamics in solutions.
Abstract
fer, proton transfer, isomerization, protein folding, and phase
transition dynamics, studies on the ultrafast dynamics in liquids
and solutions are very important to understand these processes
in detail and at molecular level. Accordingly, the intermolec-
ular dynamics in liquids and solutions are of long interest in
We studied the ultrafast dynamics of 40 aromatic cation
based ionic liquids (ILs) by means of femtosecond Raman-
induced Kerr effect spectroscopy. The low-frequency Kerr
spectra (ca. 0.3­700 cm ) of the ILs were obtained from the
Kerr transients by Fourier-transform deconvolution analysis.
The low-frequency Kerr spectra in the frequency range less
than 200 cm coming mainly from the intermolecular vibra-
tions for the ILs were discussed in terms of (i) anion depen-
dence, (ii) imidazolium cations vs. pyridinium cations, (iii)
alkyl group dependence, and (iv) effect of methylation in
aromatic cations. Several liquid properties, such as density,
viscosity, electrical conductivity, and surface tension, of the
present sample ILs at 293 K were also estimated in this study.
We clarified that the aromatic cation based ILs show a different
relation of the first moment of the low-frequency spectral band
to the bulk liquid parameter, which is the square root of sur-
face tension divided by liquid density, from aprotic molecular
liquids. The slope of the first moment to the bulk parameter for
the aromatic cation based ILs is gentler than that for aprotic
molecular liquids.
¹
1
1
chemistry, biology, and physics.
Progress in laser technology has made it possible to use
femtosecond lasers as a light source for molecular spectros-
¹
1
2
­4
copy. To probe liquid dynamics, a wide variety of spectro-
scopic techniques using femtosecond lasers including low-
¹
1
frequency (less than ca. 200 cm ) spectroscopies such as
femtosecond Raman-induced Kerr effect spectroscopy (fs-
RIKES) and terahertz time domain spectroscopy (THz-TDS)
have been developed. Molecular motions in the low-frequency
region for solutions and liquids include intermolecular dynam-
ics such as intermolecular vibrational dynamics and collective
orientational dynamics. As well as conventional Raman and FT-
IR spectroscopies, fs-RIKES and THz-TDS are complementary
spectroscopic techniques but they are specialized for the low-
¹1
frequency region below 200 cm . There are several review and
instructive articles on the ultrafast dynamics in liquids and
5­16
17,18
solutions studied by fs-RIKES
and THz-TDS.
fs-RIKES probes the polarizability response function that
is described in terms of the polarizability time-correlation
function:
1
. Introduction
Because the intermolecular dynamics of solvents influences
the barrier-crossing processes in chemical reactions, biological
processes, and relaxations in solutions, such as electron trans-
d
RRIKESðtÞ / h¡ð0Þ¡ðtÞi
ð1Þ
dt
1106 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

where ¡ is the collective polarizability. Because fs-RIKES is
a polarization-controlled third-order nonlinear spectroscopy, it
can detect isotropic and anisotropic (depolarized) signals, as
Using low-frequency vibrational spectroscopies based on
modern time-domain spectroscopic techniques such as fs-
3
0,59,60
61­67
RIKES
state Raman
and THz-TDS,
as well as traditional steady-
6
8­70
and far-IR spectroscopies,71­75 the micro-
well as other Raman active signals, by controlling the polari-
zations of pump and probe lights and analyzer.1
9­22
The iso-
scopic structures, interactions, and dynamics of different kinds
of ILs have been investigated. Among these spectroscopic tech-
niques, fs-RIKES, which has the advantage of access to the
1
9,23,24
tropic and anisotropic responses can be described by,
1
d
¹1
RIsoðtÞ ¼ ½RzzzzðtÞ þ 2RyyzzðtÞꢀ /
h¡Isoð0Þ¡IsoðtÞi
ð2Þ
lowest frequency region that is less than 1 cm , has been
extensively used to investigate ILs. The first fs-RIKES study of
imidazolium-based ILs was reported by Quitevis and co-
3
dt
1
d
RAnisoðtÞ ¼ ½RzzzzðtÞ ꢁ RyyzzðtÞꢀ /
h¡Anisoð0Þ¡AnisoðtÞi ð3Þ
2
dt
7
6
workers in 2002. So far, various kinds of ILs, e.g., imid-
6
5,66,77­89
86,89,90
66,86,89,91
where Rijkl(t) is the description of the third-order response and
¡Iso and ¡Aniso are the collective polarizabilities for the iso-
tropic and anisotropic components. Common and traditional
polarization condition for fs-RIKES measurements is the aniso-
tropic condition (eq 3). In contrast to fs-RIKES, THz-TDS
probes a dipole moment time-correlation function:
azolium-,
ammonium-,
pyridinium-,
pyrrolidinium-,
6
6,67,86,92,93
92
and phosphonium-based ILs were
investigated using fs-RIKES by several groups. Currently
dicationic imidazolium-based ILs were also investigated by
8
8,94
fs-RIKES.
Furthermore, fs-RIKES has also been applied
9
5,96
for not only neat ILs, but also IL­IL mixtures
molecular liquid mixtures.
and IL­
9
7­106
d
RTHz-TDSðtÞ / h®ð0Þ®ðtÞi
ð4Þ
Although the low-frequency spectra of ILs have been actively
studied, it is still not enough to fully understand the natures of
low-frequency spectra or intermolecular vibrational dynamics
in ILs. Previously, we collected the low-frequency spectra of as
dt
where ® is the dipole moment. Thus the THz-TDS (and
dielectric) spectrum is related to the solvation dynamics
2
5­27
12
measured by the time-dependent fluorescence Stokes shift
that has been extensively used in the study of ionic liquid (IL)
many as 40 aprotic molecular liquids measured by fs-RIKES.
1
2
In the previous work, we clarified that aromatic molecular
liquids, except for hexafluorobenzene, show bimodal line
shapes in the low-frequency spectra. We also found the linear
2
8­37
dynamics
that is also probed by fs-RIKES.
ILs are purely composed of ions, but they are molten at am-
bient temperatures (or often defined as below 373 K, that is, the
boiling point of water). Therefore ILs possess both the natures
of salt and liquid, and they are referred to as Janus materials.
Indeed, the earliest “exact” or “real” IL, ethylammonium nitrate
relationship between the first moment M of the low-frequency
1
spectral band and bulk parameters which are the square root of
1
/2
surface tension divided by density (£/µ) and the square root
1/2
of surface tension divided by formula weight (£/FW) . It is
thus no doubt that the large number of data gives better and
deeper understanding in science. Therefore, like the previous
(
the melting point is 285 K), was reported by Walden as long
3
8
ago as 1914 (but it may be ethanolammonium nitrate whose
melting point is 325­328 K reported by Gabriel and Weiner
1
2
study on aprotic molecular liquids, we have collected the low-
frequency spectra of 40 aromatic cation based ILs measured by
fs-RIKES. As far as we know, there is no report of this kind of
database of low-frequency spectra for aromatic cation based ILs
of as many as or even more than 40 samples. The purpose of this
study includes not only collecting the low-frequency spectra of
40 aromatic cation based ILs, but also overviewing the low-
frequency spectra of aromatic cation based ILs, understanding
the general and specific features of the spectra dependent on
the constituent ion species, and trying to find/understand the
relation between the intermolecular vibrational band and the
bulk properties. Thus, some bulk properties, such as density µ,
shear viscosity ©, electrical conductivity ·, and surface tension
£ of the ILs were also measured to investigate a correlation with
the low-frequency spectra in this study.
3
9
in 1888 if the definition of the melting point of less than
73 K for IL is applicable rather than real room temperature).
3
After this discovery, however, ILs did not receive much atten-
tion for a while. A milestone of research on ILs is probably the
report on water- and air-stable imidazolium salts by Wilkes and
4
0
Zawarotko in 1992. Since then, ILs have become more popu-
lar research targets in chemistry, physics, materials engineer-
ing, biochemical engineering, and so on.4
1­53
The reason why ILs are getting more attention recently is
probably their unique properties.41 Of course, the low melting
points for the ILs themselves are a purely interesting subject
in science. The low (or negligible) volatility, and thus low
flammability, for these liquids at ambient conditions is also
fascinating for safe and eco-friendly materials and solvents. The
high electrical conductivity is an attractive property for appli-
cation to electrolytes in batteries. The high dissolving power for
a wide variety of solutes can provide good solvents. In addi-
2
. Experiments
Most ILs used in this study were commercially available, but
5
9,87,107,108
tion, ILs show microsegregation structure that is not observed
some of them were synthesized in our laboratory.
The
in simple and conventional solvents.5
4­58
These unique features
sample ILs and their abbreviations are listed in Table 1. Some
ILs obtained from companies which showed brownish color
were purified by activated charcoal treatment that is commonly
and properties of ILs often come from the amphiphilic nature of
cations and the complicated and subtle balance of intermolec-
ular interactions and forces. It is thus important to investigate
the molecular-level aspects including microscopic structure and
interaction of ILs. It is expected that the low-frequency vibra-
tional spectroscopic techniques are very useful to investigate
such microscopic structure, interaction, and dynamics in ILs.
4
1
used in acetonitrile. The newly synthesized ILs in our labo-
1
ratory for this study were confirmed by H NMR and elemental
analysis. In the elemental analysis, the estimated values of
carbon, hydrogen, and nitrogen for the ILs agreed with their
calculated values within «0.4%, the criteria for the Journal of
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1107

Table 1. Abbreviations and supplier sources of ionic liquids used in this study
Entry Number
Abbreviation
Name
Source
1
2
3
4
5
6
7
8
9
[C2MIm][NTf2]
[C2MIm][BF4]
[C2MIm][OTf]
[C2MIm][NPf2]
[C2MIm][TCNB]
[C2MIm][FAP]
[C2MIm][EtSO4]
[C3MIm][NTf2]
[C3MIm][BF4]
[C4MIm][NTf2]
[C4MIm][BF4]
[C4MIm][OTf]
[C4MIm][NPf2]
[C4MIm][PF6]
[C4MIm]I
[C4MIm][DCA]
[C4MIm][NO3]
[C4MIm][SCN]
[C4MIm][NF2]
[C4DMIm][NTf2]
[C6MIm][NTf2]
[C6MIm][BF4]
[C6MIm][PF6]
[C8MIm][NTf2]
[C8MIm][BF4]
[C10MIm][NTf2]
[C10MIm][BF4]
[SiMIm][NTf2]
[C4Py][NTf2]
1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide
1-ethyl-3-methylimidazolium tetrafluoroborate
1-ethyl-3-methylimidazolium trifluoromethanesulfonate
1-ethyl-3-methylimidazolium bis(pentafluoroethylsulfonyl)amide
1-ethyl-3-methylimidazolium tetracyanoborate
1-ethyl-3-methylimidazolium tris(pentafluoroethyl)trifluorophosphate
1-ethyl-3-methylimidazolium ethylsulfate
1-methyl-3-propylimidazolium bis(trifluoromethylsulfonyl)amide
1-methyl-3-propylimidazolium tetrafluoroborate
1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide
1-butyl-3-methylimidazolium tetrafluoroborate
1-butyl-3-methylimidazolium trifluoromethanesulfonate
1-butyl-3-methylimidazolium bis(pentafluoroethylsulfonyl)amide
1-butyl-3-methylimidazolium hexafluorophosphate
1-butyl-3-methylimidazolium iodide
Ref. 107
Ref. 108
This work
Ref. 107
Merck
Merck
Iolitec
Ref. 107
Ref. 107
Ref. 87
Ref. 107
Ref. 87
This work
Fluka
10
1
1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Kanto
Merck
1-butyl-3-methylimidazolium dicyanamide
1-butyl-3-methylimidazolium nitrate
1-butyl-3-methylimidazolium thiocyanate
Ref. 107
a)
Aldrich
1-butyl-3-methylimidazolium bis(fluorosulfonyl)amide
1-butyl-2,3-dimethylimidazolium bis(trifluoromethylsulfonyl)amide
1-hexyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide
1-hexyl-3-methylimidazolium tetrafluoroborate
1-hexyl-3-methylimidazolium hexafluorophosphate
1-methyl-3-octylimidazolium bis(trifluoromethylsulfonyl)amide
1-methyl-3-octylimidazolium tetrafluoroborate
1-decyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide
1-decyl-3-methylimidazolium tetrafluoroborate
1-methyl-3-trimethylsilylmethylimidazolium bis(trifluoromethylsulfonyl)amide
1-butylpyridinium bis(trifluoromethylsulfonyl)amide
1-butylpyridinium tetrafluoroborate
1-butyl-2-methylpyridinium bis(trifluoromethylsulfonyl)amide
1-butyl-3-methylpyridinium bis(trifluoromethylsulfonyl)amide
1-butyl-3-methylpyridinium tetrafluoroborate
1-butyl-4-methylpyridinium bis(trifluoromethylsulfonyl)amide
1-butyl-4-methylpyridinium tetrafluoroborate
1-hexylpyridinium bis(trifluoromethylsulfonyl)amide
1-hexylpyridinium tetrafluoroborate
1-octylpyridinium bis(trifluoromethylsulfonyl)amide
1-decylpyridinium bis(trifluoromethylsulfonyl)amide
1-dodecylpyridinium bis(trifluoromethylsulfonyl)amide
Kanto
Iolitec
Ref. 107
Ref. 108
This work
Ref. 108
Iolitec
Ref. 108
Iolitec
Ref. 59
Ref. 86
Iolitec
[C4Py][BF4]
a)
[1C42MPy][NTf2]
[1C43MPy][NTf2]
[1C43MPy][BF4]
[1C44MPy][NTf2]
[1C44MPy][BF4]
[C6Py][NTf2]
[C6Py][BF4]
[C8Py][NTf2]
[C10Py][NTf2]
[C12Py][NTf2]
Iolitec
Iolitec
Iolitec
Iolitec
a)
a)
a)
a)
Iolitec
This work
Iolitec
This work
This work
This work
a) Purified by activated charcoal treatment.
Organic Chemistry published by the American Chemical Soci-
ety. The details of the preparation procedures and analytical
data of the synthesized ILs for this study are summarized in
Supporting Information. The sample ILs were dried in a vac-
uum oven at 313 K for more than 36 h before experiments.
µ of the ILs were obtained at 293.0 « 0.5 K using a 2 mL
volumetric flask. © of the ILs were measured using a recipro-
cating electromagnetic piston viscometer (Cambridge Viscos-
ity, ViscoLab 4100) equipped with a circulating water bath
AQ-300). The values of the water contents of the ILs are sum-
marized in Table 2.
The details of the fs optical heterodyne detected (OHD)
RIKES setup used in this study were reported in previous
1
2,109
publications.
Optical heterodyne detection is well used
in nonlinear spectroscopy including fs-RIKES to enhance the
3
,4
signal. In the current fs-OHD-RIKES setup, the light source
was a titanium sapphire laser (KMLabs, Griffin) pumped by a
110
Nd:YVO diode laser (Spectra Physics, Millennia Pro 5sJ).
4
(
Yamato, BB300) at 293.0 « 0.2 K. · of the ILs were mea-
The output power of the titanium sapphire laser was approx-
imately 390­440 mW. The temporal response, which is the
cross-correlation between the pump and probe pulses measured
using a 200 ¯m thick KDP crystal (type I), of the fs-RIKES
setup was typically ca. 36 fs (full-width at half-maximum,
FWHM). In the fs-RIKES experiments, scans with a high time
sured by an electrical conductivity meter (Mettler Toledo, S470
SevenExcellence) at 293.0 « 0.5 K. £ of the samples were
measured using a duNouy tensiometer (Yoshida Seisakusho) at
2
93.0 « 0.5 K. The water contents of the ILs were estimated
from Karl Fischer titrations using a coulometer (Hiranuma,
1108 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1109

1110 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

resolution of 3072 points at 0.5 ¯m per step that corresponds to
.34 fs per step were made for a short time window (ca. 10.2
3
ps). Longer time window transients (ca. 300 ps) of 2000 points
at 25.0 ¯m per step that correspond to 167 fs per step were also
measured. A pure heterodyne signal was obtained by recording
scans for both plus and minus ca. 1.5° rotations of the input
probe beam polarization using a quarter wave plate; these
signals were then combined to eliminate the residual homodyne
signal. The rotations by the quarter wave plate were done with
monitoring the intensity of the local oscillator. The numbers of
scans for the short and long time window transients were 3 and
5
in each polarization, respectively. The ILs were injected into
a quartz cell (Tosoh Quartz, 3 mm optical path length) via a
.2 ¯m or 0.02 ¯m Anotop filter (Whatman). During the fs-
0
RIKES experiments the temperatures of the samples were kept
at 293.0 « 0.2 K by a laboratory-built temperature controller
based on a Peltier temperature controller (VICS, VPE35-5-
2
0TS).
Ab initio quantum chemistry calculations were performed
for some ion species to obtain the stable structures and the
111
Raman spectra using the Gaussian 09 program suite. The
B3LYP/6-31+G(d,p) level of theory was used for the calcu-
lations in this study.
3
. Data Analysis
The Kerr transients of the ILs obtained by fs-RIKES were
analyzed by the Fourier-transform deconvolution method to
provide their low-frequency spectra, because the Kerr transients
include many underdamped and overdamped motions and thus
they are overlapped in time domain and complicated. The
method of the Fourier-transform deconvolution analysis for
Kerr transients was extensively developed by Lotshaw and
1
12,113
McMorrow,
and the details of the analysis procedure used
in this study were reported elsewhere.5
9,60
Figure 1. (a) Logarithmic plots of the normalized Kerr
transient (red curve) together with its triexponential func-
tion (blue curve) and (b) short time window transient from
Figure 1a shows the logarithmic plots for the Kerr transi-
ent of [C MIm][OTf] with the time window of 300 ps and
2
Figure 1b shows the short time range from ¹0.5 to 3.0 ps. The
intensity of the Kerr transient is normalized at the intensity at
t = 0 that is the electronic response. Figure 1a also shows the
fit to the Kerr transient from 3 to 300 ps by a triexponential
function,
¹0.5 to 3.0 ps for [C MIm][OTf]. (c) Fourier-transform
Kerr spectra: entire spectrum (black curve); picosecond
overdamped relaxation component (blue curve); entire
spectrum minus picosecond overdamped relaxation com-
ponent (red curve).
2
3
X
a0 þ
ai expðꢁt=¸iÞ
ð5Þ
slow relaxation component in the frequency region as low as
¹
1
i¼1
ca. 0.6 cm (or probably even low frequency region) can be
9
4
where a is the offset amplitude, a is the amplitude of the i-th
discussed adequately. The Fourier-transform Kerr spectrum
0
i
component, and ¸ is the time constant of the i-th component. In
of [C MIm][OTf] is shown in Figure 1c. This figure also shows
i
2
this study, a triexponential function was used for all the Kerr
transients. The fit parameters obtained for all the present 40 ILs
are given in Supporting Information. In fact, the time scale of
the decompositions of the spectra: the overdamped compo-
nent (triexponential fit excluded the fast component: a +
0
a exp(¹t/¸ ) + a exp(¹t/¸ )) and the spectrum produced
2
2
3
3
the slowest relaxation component for ILs is typically nano-
by the subtraction of the overdamped relaxation component.
Because the relaxation time constants of the first exponential
components (a exp(¹t/¸ )) were a few picoseconds, which is
competitive to the dephasing time in intramolecular vibrations
and the characteristic time in collision-induced intermolecular
motions, this component was included in the low-frequency
spectra for all the present ILs (Supporting Information).
The low-frequency Kerr spectra after subtractions of the
slow overdamped relaxation components were further analyzed
for their line shapes. The primary purpose of the line shape
seconds though it depends on the IL.6
5­67,114­116
The present
fs-RIKES apparatus whose light source is a titanium sapphire
oscillator, not an amplifier, and the delay stage is the length of
1
1
5
cm, can obtain reliable data in the time range of less than
109
approximately 300 ps. Thus the slow relaxation components
over subnanoseconds are treated as the amplitude a in this
0
study. Note that even if the slow relaxation component cannot
be captured completely in the present setup, the line shape
of the low-frequency spectrum without the component of the
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1111

analyses of the low-frequency spectra includes (i) distinguish-
ing clear intramolecular vibrational modes from low-frequency
broad bands, (ii) estimations of the first moments (M ) of
1
spectra which excluded clear intramolecular vibrational modes,
and (iii) representing the low-frequency spectra numerically.
There are two major line shape models for low-frequency
spectra of liquids and solutions measured by fs-RIKES. One
117
is the sum of Ohmic (or Bucaro­Litovitz) and antisymme-
118
trized Gaussian functions, and the other is the multimode
Brownian oscillator model.1
19,120
There are small differences in
the line shape fits especially in the high-frequency side of the
1
2,109
low-frequency spectra of aromatic molecular liquids,
but
5
9,60
82,91
both the former
and latter models
seem to work equally
well for ILs. In this study, we used the former model. The
Ohmic component is given by,
2
X
IO;ið½Þ ¼
aO;i½ expðꢁ½=½O;iÞ;
ð6Þ
i¼1
where aO,i and ½O,i are the amplitude and characteristic
frequency parameters for the i-th Ohmic function, respectively.
The antisymmetrized Gaussian component is1
18
ꢀ
ꢁꢁ
ꢂ
n
2
X
2ð½ ꢁ ½G;iÞ
IG;ið½Þ ¼
aG;i exp
ꢀ
½G;i2
Figure 2. Line shape analysis results for the low-frequency
Kerr spectra of (a) [C4MIm][DCA] and (b) [C4Py][NTf2].
Black dots denote the experimental data, red lines denote
the entire fits, blue areas denote the Ohmic functions, green
areas denote the antisymmetrized Gaussian functions,
brown areas denote the Lorentz functions, and red areas
denote the sums of the Ohmic and antisymmetrized
Gaussian functions.
i¼1
ꢁ
ꢂꢃ
2
ꢁ
2ð½ þ ½G;iÞ
ꢁ
aG;i exp
;
ð7Þ
ꢀ
½G;i2
where a , ½ , and ¦½ are the amplitude, characteristic
G,i
G,i
G,i
frequency, and bandwidth parameters for the i-th antisymme-
trized Gaussian function, respectively. For a clear intramolec-
ular vibrational mode, a Lorentzian function,
aL
2
ILð½Þ ¼
;
2
ð8Þ
ð½ ꢁ ½LÞ þ ꢀ½L
correlation can contribute to the spectral intensity sometimes
positively and sometimes negatively not only for molecular
where a , ½ , and ¦½ are the amplitude, peak frequency,
L
L
L
1
21­126
127­129
and bandwidth parameters, respectively, has been used to
fit an intramolecular vibrational band. Figure 2 shows the
line shape analysis results for the low-frequency spectra of
liquids,
but also ILs.
The above line shape analysis
never models such a contribution in the low-frequency spec-
trum. Exceptions are fixed intermolecular systems, such as
hydrogen-bonded complex systems including 7-azaindole
[
C MIm][DCA] and [C Py][NTf ], as examples. All the
4 4 2
1
30,131
Fourier-transform Kerr spectra of the present ILs and the line
shape analysis results are summarized together with the nor-
malized Kerr transients and their triexponential fits in Support-
dimer that show clear intermolecular vibrational modes.
However, the present ILs are not such rigid intermolecular
systems and thus are not this type of research target. Nonethe-
less, as mentioned above, the line shape analysis is useful
for the purposes of resolving clear intramolecular vibrational
ing Information. The M values of the low-frequency spectra
1
for the ILs are listed in Table 2. M is defined as,
1
Z
ꢄ Z
modes from broad spectra, estimating M of spectrum, and
representing the low-frequency spectra numerically.
1
M1 ¼ ½Ið½Þd½
Ið½Þd½
ð9Þ
4
. Results and Discussion
where I(½) is the frequency-dependent intensity of the spectra
from which the contributions of the picosecond overdamped
relaxation process and clear intramolecular vibrational bands
are subtracted (red areas in Figure 2, for example). The inte-
4.1 Bulk Properties. 4.1.1 Density µ: Table 2 lists the µ
values of the ILs at 293 K measured in this study. The values of
µ for all the ILs at 293 K and/or near temperatures were report-
ed in other literature. In Table 2, we can find reports on the µ of
the ILs that agree with that estimated in this study within « ca.
¹
1
gration was made from 0 to 1000 cm .
We should notice an important issue in the line shape
analysis of the low-frequency spectra of solutions and liquids
including ILs. The simple assignments of intermolecular vibra-
tional modes in the low-frequency spectra of liquids based on
the line shape analysis are often beyond safe discussion: the
nature of intermolecular vibrations that is strongly coupled
between them is neglected. It is known that the intermolecular
vibrational motions are coupled with each other and their cross
5
9,85,87,107,132­153
147
1­2%,
except for [1C 2MPy][NTf ],
if we
4
2
take the temperature effect on the µ in ILs into considera-
tion. Note that the µ values of ILs slightly decrease with
1
54
increasing temperature. As far as we know, there is only one
1
47
report on the µ of [1C 2MPy][NTf ] by Bittner et al. The
4
2
difference in the µ of [1C 2MPy][NTf ] between the value
4
2
estimated in this study and that reported by Bittner et al. is
1112 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

Figure 3. Dependence of carbon number of alkyl group
of the cations on the density µ for [CnMIm][NTf2] (open
circles), [CnMIm][BF4] (open squares), and [CnPy][NTf2]
Figure 4. Semi-logarithmic plots of viscosity © vs. carbon
number of alkyl group for [CnMIm][NTf2] (open circles),
[CnMIm][BF4] (open squares), and [CnPy][NTf2] (filled
circles). Linear fits in the plots (exponential fits) are also
shown.
(
filled circles). Linear fits are also shown.
as large as 3.5%, but the µ value of [1C 2MPy][NTf ] esti-
and/or its near temperatures were also reported. As seen in
Table 2, we can find similar values of the © for the present ILs
4
2
mated in this study is quite similar to those of the isomeric
cation based ILs, [1C 3MPy][NTf ] and [1C 4MPy][NTf ] (see
1
32
within approximately «20% in a handbook
literature
and reported
if the tem-
4
2
4
2
5
9,85,87,107,133,135,136,138,140­143,146­152,155­165
Table 2). The values of the molar volume Vm estimated from
the µ and FW (Vm = FW/µ) for the ILs are also summarized in
Table 2.
perature effect is accounted for. Note that the temperature
effects on the © values of ILs are quite large compared to
common molecular liquids: Increasing temperature lowers © in
ILs as in most liquids (again the temperature effect on © is
As seen in Table 2, the µ values of the ILs are variable with
the different anion species. On the basis of the data of the ILs
+
+
154
of [C MIm] and [C MIm] cations, the tendency of the µ with
much larger in ILs than in molecular liquids).
2
4
variation of the anion species is found as,
It is well-known that the © of IL depends strongly on the
anion species, as seen in Table 2. The anion species depend-
ence of © appears to change with the counter cations for the
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
½
FAPꢀ > ½NPf2ꢀ > I > ½NTf2ꢀ > ½PF6ꢀ ꢂ ½NF2ꢀ
ꢁ
ꢁ
ꢁ
ꢁ
>
½OTfꢀ > ½BF4ꢀ ꢂ ½NO3ꢀ ꢂ ½EtSO4ꢀ
½SCNꢀ ꢂ ½DCAꢀ > ½TCNBꢀ :
+
+
present ILs: Even the [C2MIm] - and [C4MIm] -based ILs
ꢁ
ꢁ
ꢁ
>
show different trends (42.7 cP for [C MIm][BF ] and 54.4 cP
2
4
Overall, the anions having the larger number of fluorine atoms
provide the more dense ILs.
Alkyl group dependence of the µ for the [C MIm][NTf ],
for [C MIm][OTf], but 116 cP for [C MIm][BF ] and 104 cP
2 4 4
for [C MIm][OTf]). The following tendency of © with the
4
+
variation of anion species is found for the present [C MIm] -
n
2
4
[
C MIm][BF ], and [C Py][NTf ] is shown in Figure 3. Previ-
based ILs.
n
4
n
2
ously, the linear dependence of the carbon number of the alkyl
group on the µ for [C MIm][BF ] and [C MIm][PF ] was
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
I > ½PF6ꢀ > ½NO3ꢀ > ½NPf2ꢀ > ½BF4ꢀ ꢂ ½OTfꢀ
n
4
n
6
ꢁ
ꢁ
ꢁ
ꢁ
>
½SCNꢀ > ½NTf2ꢀ > ½NF2ꢀ ꢂ ½DCAꢀ :
reported by Seddon and coworkers.1 It is clear from the figure
54
+
that the µ values of the [C MIm][NTf ] are quite similar to
For the [C MIm] -based ILs, the tendency of the © with the
n
2
2
those for the [C Py][NTf ] when we compare the ILs having
variation of anion species is following:
n
2
the same alkyl group (and also the alkyl group dependence
of µ). This is most likely attributed to the structural similarity:
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
½
EtSO4ꢀ > ½NPf2ꢀ > ½FAPꢀ > ½OTfꢀ > ½BF4ꢀ
ꢁ
ꢁ
>
½NTf2ꢀ > ½TCNBꢀ :
(
i) the molecular weights of the two cation moieties are quite
+
+
similar (­MIm : 82.1 and ­Py : 79.1) and (ii) the both are
aromatic. Furthermore, the µ values for the three series of the
ILs, [CnMIm][NTf2], [CnMIm][BF4], and [CnPy][NTf2], de-
crease almost monotonically with increasing the alkyl groups
of the cations: µ = 1.557 ¹ 0.02954n for the [C MIm][NTf ];
The © values of the ILs increase with the longer alkyl
group of the cations. Figure 4 shows the semi-logarithmic
plots of © vs. the carbon number of the alkyl group for the
[C MIm][NTf ], [C MIm][BF ], and [C Py][NTf ]. As dis-
played in Figure 4, the © value is roughly exponentially pro-
portional to the number of carbon of alkyl group n: log(©) =
1.457 + 0.07404n for the [C MIm][NTf ]; log(©) = 1.463 +
n
2
n
4
n
2
n
2
µ = 1.313 ¹ 0.02641n for the [C MIm][BF ]; µ = 1.543 ¹
n
4
0
.02545n for the [C Py][NTf ]. This indicates that the intrinsic
n 2
n
2
density µ (n = 0) is determined by the combination of the
cation and anion, but the alkyl group length dependent feature
0.1450n for the [C MIm][BF ]; log(©) = 1.628 + 0.06392n for
n 4
the [C Py][NTf ]. The slope of the plots for the [C MIm][BF ]
n
2
n
4
is mainly due to the increment ratio of the CH volume to the
is steeper than those for the [C MIm][NTf ] and [C Py][NTf ],
n 2 n 2
2
sum of the anion and cation volumes.
and the slopes for the [CnMIm][NTf2] and [CnPy][NTf2] are
quite similar. Therefore, the alkyl group length dependence of
the © values for the ILs is varied with the anion species, but the
difference between imidazolium and pyridinium cations is not
significant.
4
.1.2 Viscosity ©: As well as µ, © is probably one of the
41
most well characterized physical properties of ILs. The ©
values of the present ILs at 293 K measured in this study are
summarized in Table 2. The © values of all the ILs at 293 K
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1113

It is clearly shown in Table 2 that the methylation on the
aromatic cations has a significant influence on © although it is
not straightforward. The © value of [C MIm][NTf ] at 293 K is
4
2
approximately doubled by the methylation ([C DMIm][NTf ]).
4
2
This feature is actually known.1 The pyridinium-based ILs
show more complicated methylation effect on ©. The © for
the pyridinium-based ILs slightly increases by the methylation
in most cases, as the © of [1C 4MPy][BF ] is higher than that of
66
4
4
[
C Py][BF ]. However, it is noteworthy that [1C 4MPy][NTf ]
4 4 4 2
is slightly less viscous than [C Py][NTf ]. This indicates that
4
2
the methylation on the aromatic cations not only modifies the
cation’s nature itself but can also change the intermolecular
nature of cation and anion appreciably.
4
.1.3 Electrical Conductivity ·: The · value is probably
one of the most important physical properties of ILs, because
the use of ILs as electrolyte in capacitors, batteries, solar cells,
and so forth, is a major industrial application. The · values of
the present ILs at 293 K measured in this study are tabulated in
Table 2. The · values of most of those ILs can be found in the
literature, and some reported data at 293 K and/or near tem-
peratures are in good agreement with the present data within
approximately «20% if we take the temperature effect into
4
1,135,140,141,143,150,152,155,157,159,161,162,167­176
account.
As well as
Figure 5. Plots of (a) electrical conductivity · vs. inverse
¹1
the temperature effect on ©, the · values of ILs are quite tem-
perature dependent: the · value increases with increasing tem-
of viscosity © and (b) molar electrical conductivity Λ vs.
¹1
©
for the present ILs. The linear fits are also shown.
perature. According to our survey, the · values of [C MIm]-
4
[
[
NO ], [SiMIm][NTf ], [1C 2MPy][NTf ], [1C 4MPy][NTf ],
The trends for the Λ are better correlated to those for the ©
rather than those for the ·. Figure 5 shows the plots of · vs.
1/© and Λ vs. 1/©. It is obvious that the Λ is more clearly
inversely proportional to the © than the · is. It is reported that
3
2
4
2
4
2
C Py][BF ], [C Py][NTf ], and [C Py][NTf ] studied here
6
4
10
2
12
2
are the first report.
When we look at the · values of [C MIm] - and [C MIm] -
based ILs, they are variable with the different anion species,
as the © values discussed above. In addition, the change in the
+
+
2
4
1
77
the ionicity is an important factor for electrical conductivity.
Angell and coworkers pointed out that the different fragilities
1
78
·
for the ILs by an exchange of the anion species looks
show the different correlations in the Walden plots. While a
few data points are scattered from the correlation between · vs.
1/© and Λ vs. 1/© in the present results (Figure 5), which are
likely coming from such natures of cation and anion combi-
nations and liquids, the electrical conductivities of the present
ILs at 293 K roughly obey the Walden rule overall.
dependent on the cation, as well as the © values. The tendency
of the · with the exchange of the anion species for the present
+
[C MIm] -based ILs is as follows:
4
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
½
DCAꢀ > ½NF2ꢀ > ½SCNꢀ > ½BF4ꢀ ꢂ ½NTf2ꢀ
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
>
½OTfꢀ > ½NO3ꢀ > ½NPf2ꢀ ꢂ ½PF6ꢀ > I :
It is also clear from the data for [CnMIm][NTf2], [CnMIm]-
[BF ], and [C Py][NTf ] in Table 2 that the · value depends
+
The following tendency is for the [C MIm] -based ILs.
2
4
n
2
greatly on the alkyl group of the cations. Watanabe and co-
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
½
BF4ꢀ ꢂ ½TCNBꢀ > ½NTf2ꢀ ꢂ ½OTfꢀ > ½FAPꢀ
workers reported that the · value decreases with the longer
ꢁ
ꢁ
>
½EtSO4ꢀ ꢂ ½NPf2ꢀ :
179
alkyl group of cation in ILs.
They pointed out that the
These orders in the · are roughly related to the opposite trends
variation in the alkyl chain length causes change in not only the
ionic concentration, but also the van der Waals interactions.
Figure 6 shows the semi-logarithmic plots of · vs. carbon
number of alkyl group n, as well as the semi-logarithmic
plots of Λ vs. n, for the [C MIm][NTf ], [C MIm][BF ], and
in the ©, though that correlation is not perfect. In particular, the
¹
¹
ILs with the anions of [NTf ] and [NPf ] show the lower ·
2
2
values than the values expected from the ©.
We also estimated the molar electrical conductivity Λ from
the · and V (Λ = ·/V ). The Λ values of the present ILs are
n
2
n
4
[C Py][NTf ]. As seen in the figure, the · and Λ values do
m
m
n
2
also listed in Table 2. The trend of the Λ with the replacement
not simply express an exponential feature unlike the ©. Instead,
+
180­183
of the anion species for the present [C MIm] -based ILs is the
the Vogel­Tammann­Fulcher function,
which is used to
4
following:
express the temperature dependent © of liquids including ILs,
fits well to each series.
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
½
DCAꢀ > ½NF2ꢀ > ½NTf2ꢀ > ½SCNꢀ > ½BF4ꢀ
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
· ¼ · expðB=ðn ꢁ n ÞÞ
ð10Þ
>
½OTfꢀ > ½NPf2ꢀ > ½NO3ꢀ > ½PF6ꢀ > I :
0
0
+
The following trend is for the [C MIm] -based ILs.
where · is the limiting conductivity, B is a constant that could
2
0
be related to a parameter of the carbon number sensitivity to ·,
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
½
TCNBꢀ > ½BF4ꢀ > ½NTf2ꢀ > ½OTfꢀ ꢂ ½FAPꢀ
and n is a constant that could be the critical carbon number of
0
ꢁ
ꢁ
>
½NPf2ꢀ > ½EtSO4ꢀ :
the alkyl group that the conductivity diverges at. Also,
1114 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

Figure 7. Plots of surface tension £ vs. carbon number of
alkyl group for [CnMIm][NTf2] (open circles), [CnMIm]-
[BF4] (open squares), and [CnPy][NTf2] (filled circles).
[
NTf ], were already reported. As seen in Table 2, we can find
2
similar values of the £ for the present ILs within approximately
«
20% in reported literature when the temperature effect is taken
5
9,87,107,134,136­139,142,145,147,150,152,155,184­189
into consideration.
Note that the £ values of ILs decrease with increasing
1
84,186,187
temperature.
It is clear from Table 2 that the £ of the ILs is strongly
influenced by the anion species. When we compare the £
Figure 6. Semi-logarithmic plots of (a) electrical conduc-
tivity · vs. carbon number of alkyl group and (b) molar
electrical conductivity Λ vs. n for [CnMIm][NTf2] (open
circles), [CnMIm][BF4] (open squares), and [CnPy][NTf2]
+
+
values of the [C MIm] - and [C MIm] -based ILs in Table 2,
2
4
it is found that the tendency of the £ for the anion species of the
aromatic cation based ILs is as follows:
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
I
> ½NO3ꢀ ꢂ ½SCNꢀ > ½DCAꢀ > ½BF4ꢀ ꢂ ½NF2ꢀ
(
filled circles). Fits by the Vogel­Tammann­Fulcher
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
equation are also shown.
ꢂ ½PF6ꢀ > ½TCNBꢀ ꢂ ½EtSO4ꢀ > ½OTfꢀ > ½NTf2ꢀ
ꢁ
ꢁ
>
½FAPꢀ > ½NPf2ꢀ :
Table 3. Fit parameters for the alkyl group dependence of
electrical conductivity and molar electrical conductivity for
It seems that £ depends on the ionic volume of the anion
species. We will see the relation between the £ and V for the
[
CnMIm][NTf2], [CnMIm][BF4], and [CnPy][NTf2]
m
ILs (vide infra).
Besides the anion effect, the length of the alkyl group for the
·
0
Λ
0
B
n
0
B¤
0
n ¤
¹1
2
¹1
/
mS cm
0.103 16.1 ¹6.56
0.097 9.61 ¹2.50
0.090 15.4 ¹6.33
/Scm mol
1
-alkyl-3-methylimidazolium and 1-alkylpyridinium cations,
+
+
[C
[C
[C
n
MIm][NTf
MIm][BF
Py][NTf
2
]
0.094
0.055
0.093
8.95 ¹4.72
5.42 ¹1.45
7.12 ¹3.60
[C MIm] and [C Py] , affects £ too. For example, Kolbeck
n n
n
n
4
]
et al. reported that the £ of [CnMIm][NTf2] at 298 K decreases
to the value of 29.5 mN m with the longer alkyl group up to
¹1
2
]
octyl, but the longer alkyl groups than octyl do not influence
1
90
greatly the £ value.
Figure 7 shows the carbon number
0
0
ꢀ
¼ ꢀ0 expðB =ðn ꢁ n0 ÞÞ
ð11Þ
dependences of £ for the [C MIm][NTf ], [C MIm][BF ], and
n
2
n
4
where Λ is the limiting molar conductivity, B¤ is a constant that
[C Py][NTf ]. As shown in the figure, £ decreases with the
n 2
0
could be related to a parameter of the carbon number sensitivity
longer alkyl group of the cations for the [C MIm][BF ] within
n 4
to Λ, and n ¤ is a constant that could be the critical carbon
number of the alkyl group. Note that the fits using eqs 10 and
the ethyl group and the decyl group. On the other hand, the
[CnMIm][NTf2] and [CnPy][NTf2] show that the £ becomes
low with increasing length of the alkyl group up to the octyl
0
1
1 to the · and Λ data here are just empirical, not based on a
theory or an authority. The fit parameters for · (eq 10) and Λ
eq 11) are summarized in Table 3. Therefore the alkyl group
dependence of the · and Λ for the [C MIm][NTf ] is quite
¹
1
group but it becomes a constant at ca. 30.5 mN m with further
lengthening the alkyl group. In addition, the carbon number
dependence of £ for the [C MIm][NTf ] is almost the same as
(
n
2
n
2
similar to that of the [C Py][NTf ]. These results suggest that
that for the [C Py][NTf ]. This indicates that the anion species
n 2
n
2
the alkyl group dependence of the · and Λ is critically influ-
enced by the anion species, but not much affected by the cation
moiety.
critically affects the alkyl group dependence of the £ for the
imidazolium- and pyridinium-based ILs, but the difference in
the cations, imidazolium and pyridinium, is not crucial for the
alkyl group dependence of £.
4
.1.4 Surface Tension £: The £ values of the present ILs at
2
93 K measured in this study are listed in Table 2. The £ values
Maroncelli and coworkers previously reported a single rela-
at 293 K and/or the near temperatures for the present ILs, except
for [C MIm][NF ], [C Py][BF ], [C Py][NTf ], and [C Py]-
tion between £ and V in alkali halides at their melting points
and ILs at room temperature. Figure 8 shows the plots of £
m
1
91
4
2
6
4
10
2
12
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1115

Figure 8. Plots of surface tension £ vs. molar volume Vm
for aromatic cation based ILs. Exponential fit is also
shown. Empirical correlation for molten salts including ILs
reported by Maroncelli and coworkers is also shown.1
91
vs. Vm for the present ILs. In the figure, the empirical corre-
1
91
lation curve reported by Maroncelli and coworkers
the fit to the present data by an exponential function (£ =
78.52 exp(¹0.0124V ) + 28.705) are also given. When we
and
1
m
carefully look at Figure 8, it is found that the present data
depart slightly from the empirical correlation reported by
Maroncelli and coworkers especially in the smaller V region:
m
the empirical correlation underestimates the £ values for com-
pact ILs. This is probably because the £ data for the alkali
halides, which are smaller Vm than the ILs, are at the melting
Figure 9. Low-frequency Kerr spectra of (a) [C4MIm]I,
b) [C4MIm][BF4], and (c) [C4MIm][PF6]. Spherical top
anions are compared.
(
191
points but those for the ILs are at 298 K. Because the £ in
most liquids decreases with increasing temperature,1 the £
values in the ILs with smaller Vm are likely underestimated in
the single correlation for alkali halides and ILs.1
32
91
ence, (ii) imidazolium cations vs. pyridinium cations, (iii) alkyl
group dependence, and (iv) effect of methylation in aromatic
cations. The correlation between the low-frequency spectral
band and the bulk parameter will be discussed after this section.
4.2.2 Anion Dependence: It is well-known that the line
4.2 Ultrafast Dynamics: Low-Frequency Spectra. 4.2.1
General Features of Aromatic Cation Based ILs: Prior to
discussing each subject regarding the low-frequency Kerr spec-
tra of the aromatic cation based ILs in this section, let us sum-
marize some unique spectral features of aromatic cation based
shape of the low-frequency Kerr spectrum in ILs depends on
6
0
¹
ILs. First, the Kerr intensities due to the nuclear response
relative to the signal intensities of electronic responses are much
higher for aromatic cation based ILs than nonaromatic cation-
based ILs when the same anions are compared. Second, the
the anion species. For example, [NTf ] anion was compared to
2
¹
78,96
¹ 78,82,92,96
¹ 83
spherical top anions, Br ,
[PF6] ,
and [BF4] ,
¹
82,96
¹
and a simpler anion [OTf] ,
center elements, P, As, and Sb, were compared, and some
systematically differently branched anions, [SCN] , [DCA] ,
[C(CN) ] , [OTf] , [NTf ] , and tris(trifluoromethylsulfonyl)-
3 2
methide [CTf ] , were studied. Here, we show the low-
[XF6] anions with different
8
5
¹1
¹
¹
spectral density in the frequency region higher than 50 cm
¹
¹
¹
is large for aromatic cation based ILs because of the ring
librational motion. The maxima in the low-frequency spectra
for aromatic cation based ILs usually appear at approximately
¹
87
3
frequency spectra of the ILs with 13 different anions (the
¹1
+
+
8
0­100 cm . Conversely, the spectral intensity responsible for
[C MIm] - and [C MIm] -based ILs).
2 4
¹
¹
cation motion is very small for nonaromatic cation based ILs.
These differences in the intensities of the Kerr transients and the
First of all, we compare the spherical top anions, I , [BF4] ,
¹
+
and [PF ] , in the [C MIm] -based ILs, as shown in
6
4
low-frequency spectra between aromatic and nonaromatic ILs
Figure 9. It is well-known that the Kerr signals of spherical
top molecular liquids such as carbon tetrachloride are tiny
compared to aromatic molecules in the typical fs-RIKES which
are commonly observed in molecular liquids as well.1
92­195
The
third one is unique for ILs. There is a correlation between the
characteristic frequency of low-frequency spectral band (M )
uses the polarization geometry for the anisotropy detec-
1
and the bulk parameter ((£/µ)1 ) for aprotic molecular liquids
whether they are aromatic or nonaromatic, but the correlations
seem to be different between aromatic and nonaromatic cation
based ILs, although the number of data is not extensive.
Keeping these general features for aromatic cation based ILs
in mind, we see the following four topics of the low-frequency
Kerr spectra for aromatic cation based ILs: (i) anion depend-
/2
tion.
109,130,131,196,197
As well as such spherical top molecular
liquids, it is expected that the contributions of spherical top
anions to the low-frequency Kerr spectra of aromatic cation-
based ILs are small compared to the cations. Nonetheless, it is
clear from Figure 9 that the line shapes of the low-frequency
Kerr spectra of [C MIm]I, [C MIm][BF ], and [C MIm][PF ]
4
4
4
4
6
are quite different, especially in the low-frequency region
1116 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

¹1
below ca. 50 cm : The intensity in this low-frequency region
¹1
relative to the high-frequency region at ca. 100 cm is small in
¹
1
[
[
C MIm]I but that is close to the intensity at ca. 100 cm in
4
C MIm][BF ].
4
4
In a neat aromatic molecular liquid benzene, it was point-
ed out that the translational motion (interaction-induced or
collision-induced motion) appears in the low-frequency region
¹
1
below 50 cm but the reorientational motion (libration) con-
tributes to the intermolecular vibrational spectrum entirely,
especially the spectral density in the high-frequency region
¹
1
(
ca. 75­200 cm ), by the molecular dynamics (MD) simulation
121
work of Ryu and Stratt. If this consideration is also appli-
cable for these aromatic cation based ILs, the spherical top
anions significantly influence the spectral density in the low-
frequency region and the translational motion (and probably the
coupling motion with the cation as well) arising from the anion
¹
is expected to be suppressed with the order of the masses: I >
¹
¹
[
PF ] > [BF ] . Previously, we compared the low-frequency
6 4
spectra of [C MIm][PF ], [C MIm][AsF ], and [C MIm]-
4
6
4
6
4
8
5
[
5
SbF6]. The intensity in the low-frequency region below
0 cm relative to that in the high-frequency region at ca. 100
cm decreases with the order of [PF ] > [AsF ] > [SbF ] .
¹
1
¹
1
¹
¹
¹
6
6
6
In the present study, however, the low-frequency Kerr spectra of
C MIm]I, [C MIm][BF ], and [C MIm][PF ] do not show such
[
4
4
4
4
6
a clear correlation with the formula weights (FW) or size of their
anions. As shown in the low-frequency Kerr spectra in Figure 9,
¹
1
the intensity at the low-frequency region below 50 cm relative
¹1
to the high-frequency region at ca. 100 cm increases with the
¹
¹
¹
order of I < [BF ] µ [PF ] but FW of the anions changes
4
6
¹
¹1
¹
¹1
differently as [BF ] (86.80 g mol ) < I (126.90 g mol ) <
PF6] (144.96 g mol ) and so does the ion radii (I (220 ¡) <
BF ] (230 ¡) < [PF ] (245 ¡)). Rather, the relative inten-
4
¹
¹1
¹
[
[
¹
¹
198
4
6
sities in this low-frequency region changes in parallel with the
fluidity properties (© and ·, see Table 2). This might imply that
the intermolecular interaction is a dominant factor for the spec-
tral intensity in the low-frequency region or translational motion
Figure 10. Low-frequency Kerr spectra of 1-butyl-3-meth-
ylimidazolium-based ILs with the anions of (a) bis-
(and probably coupling motion too) compared to the effects of
(
fluorosulfonyl)amides and bis(perfluoromethylsulfonyl)-
the mass and volume. Note that spherical top molecules do not
show depolarized Raman intensities arising from their libra-
tions. This is an example that shows how complicated the line
shape of the low-frequency Kerr spectra of ILs are: even the ILs
consisting of spherical top anions do not show a simple trend.
Secondly, the ILs with the series of the bis(fluorosulfonyl)-
amide and bis(perfluoroalkylsulfonyl)amide anions, i.e.,
amides, (b) [C4MIm][NTf2], and (c) [C4MIm][NPf2]. (d)
Spectrum of 1-butyl-3-methylimidazolium trifluorometha-
nesulfonate [C4MIm][OTf] is also shown as a reference.
¹
¹
of the [NTf ] or [CTf ] anion and/or the coupling of trans-
2
3
8
7
lational and reorientational motions previously. The transla-
tional motions in molecular liquids and ILs often appear in the
low-frequency region below 50 cm . Thus the translational
motion can influence the spectral density in the low-frequency
region. In polymer solutions, on the other hand, the spectral
intensity in this low-frequency region is lower for the heavier
¹
¹
¹
[
NF ] , [NTf ] , and [NPf ] , are compared, as shown in
2 2 2
¹
1
Figure 10. In the figure, the spectrum of [C MIm][OTf] is also
4
shown for a comparison. The clear characteristic peaks of
the spectra for [C MIm][NF ], [C MIm][NTf ], and [C MIm]-
4
2
4
2
4
¹
1
[
3
NPf ] are observed at ca. 10­30 cm , but the peak at ca. 10­
0 cm is not clear in [C MIm][OTf]. Indeed, this unique
2
¹
1
199­201
polymer and oligomers than the lighter monomer.
4
spectral feature is not observed in the ILs with the other anions,
except for [FAP] (Supporting Information).
Furthermore, the intensity in the low-frequency region below
¹
¹1
20 cm of the low-frequency Kerr spectra of [C MIm]-based
4
Previously, we compared the spectra of [C MIm][OTf],
ILs with hexafluoropnictogenate anions becomes smaller with
4
¹
¹
¹ 85
[
C MIm][NTf ], and 1-butyl-3-methylimidazolium tris(trifluoro-
the substitution from [PF6] to [AsF6] and [SbF ] . MD
4
2
6
methylsulfonyl)methide ([C4MIm][CTf3]), and found that the
simulation results of [C4MIm][PF6], [C4MIm][AsF6], and
[C MIm][SbF ] indicated that the translational motion is clear
¹
1
peaks at ca. 20 cm
were there in [C MIm][NTf ] and
4 2
4
6
[
C MIm][CTf ], but that of [C MIm][OTf] shows no clear
in [C MIm][PF ] but not active in the other two ILs with heav-
4 6
4
3
4
¹
1 87
128
peak at ca. 20 cm . We assigned the characteristic spectral
peaks for [C MIm][NTf ] and [C MIm][CTf ] to the libration
ier anions. However, the trend observed in [C MIm][OTf],
4
[C MIm][NTf ], and [C MIm][CTf ] is opposite to these
4
2
4
3
4
2
4
3
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1117

Figure 11. Low-frequency Kerr spectra of (a) [C4MIm]-
DCA], (b) [C4MIm][NO3], and (c) [C4MIm][SCN]. Sim-
[
ple symmetric anions (C2v, D3h, C¨v) are compared.
reported cases. Thus the peak intensities at the low-frequency
¹
1
region at ca. 20 cm for [C MIm][NTf ] and [C MIm][CTf ]
4
2
4
3
were assigned to the anion’s librations and/or the coupling of
translational and reorientational motions.
Figure 12. Low-frequency Kerr spectra of (a) [C2MIm]-
[TCNB] and (b) [C2MIm][FAP]. Low-frequency spectra of
ILs with unique anions are shown as examples. Calculated
Raman spectra of (c) [TCNB] and (d) [FAP] based on
the B3LYP/6-31G+(d,p) level of theory are also shown.
In the same way as the previous study of [C MIm][NTf ]
4
2
87
and [C MIm][CTf ], we attribute the characteristic peaks at
ca. 10­30 cm for [C MIm][NF ] and [C MIm][NPf ] to the
4
3
¹
1
4
2
4
2
¹
¹
anion’s libration and/or the coupling of translational and
reorientational motions, not mainly due to the translational
motion. This is because the chemical structures of anions are
quite similar. It is further found from the comparison of
contribute to the low-frequency spectra of [C MIm][DCA],
4
[
C MIm][NF ], [C MIm][NTf ], and [C MIm][NPf ] shown in
[C MIm][NO ], and [C MIm][SCN], and the line shapes of
4
2
4
2
4
2
4
3
4
Figure 10 that the peak in the low-frequency region shifts to
the lower frequency with the larger anion: [C MIm][NF ] (ca.
their low-frequency spectra would be different. However, this
is not the case. As seen in Table 2, the three ILs have similar
FW values (Table 2) and their bulk properties such as µ and £,
which are related to the M1 of the low-frequency spectrum
(vide infra), are also similar. Thus, it is not very surprising that
the three ILs share the close similarity in the line shapes of their
low-frequency spectra.
4
2
¹
1
¹1
2
1 cm ) > [C MIm][NTf ] (ca. 18 cm ) > [C MIm][NPf ]
4 2 4 2
¹
1
(ca. 14 cm ). This is rather reasonable, because the order of
the peak frequencies of the three ILs is the inverse of the orders
of the masses and the volumes of the anions.
Figure 11 displays the low-frequency Kerr spectra of the
+
[
C MIm] -based ILs with relatively simple and highly sym-
Figures 12a and 12b displays the low-frequency Kerr spectra
4
¹
¹
¹
+
metrized anions: [DCA] (C ), [NO ] (D ), and [SCN]
C¨v). There is a strong intramolecular band at ca. 182 cm
of the [C MIm] -based ILs with relatively complicated anions:
2
v
3
3h
2
¹
¹1
¹
(
[TCNB] and [FAP] . In contrast to the simple anions that
consist of relatively small number of light atoms as shown in
Figures 9 and 11, the line shapes of the low-frequency Kerr
spectra of [C MIm][TCNB] and [C MIm][FAP] are compli-
in [C MIm][DCA], but the low-frequency bands, which are
4
mainly attributed to the intermolecular vibrations for the three
¹
ILs, are quite similar. If we consider the fact that [DCA] ,
2
2
¹
¹
[
NO ] , and [SCN] anions are not spherical top and their
cated and unique: [C MIm][TCNB] shows the strong bands at
2
3
¹1
librational motions are most likely active because of their
non-zero polarizability anisotropy volumes, they could fairly
ca. 120 and 145 cm and [C MIm][FAP] shows a bimodal line
2
shape. Figures 12c and 12d show the calculated Raman spectra
1118 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

¹
¹
of [TCNB] and [FAP] anions based on ab initio quantum
chemistry calculations. As seen in the figure, the strong bands
¹
1
at 124 and 147 cm for [C MIm][TCNB] are assigned to the
2
anion’s intramolecular vibrational modes based on the calcu-
lated Raman spectrum. The first band includes two modes
(NC­B­CN twisting and scissoring) and the latter one is triple
degeneracy mode (combination of rocking and wagging or
¹
scissoring and scissoring). On the other hand, [FAP] anion
shows many Raman active intramolecular modes though their
intensities are not that strong. It may be realized from
Figure 12 that the low-frequency Kerr spectrum of [C MIm]-
2
¹1
[FAP] shows the strong intensity at ca. 10 cm despite the
absence of a strong Raman active intramolecular mode in the
anion. As discussed above, the ILs with bis(fluorosulfonyl)-
amide and bis(perfluoroalkylsulfonyl)amide anions display
similar bimodal spectral shapes. Like the case of the ILs with
bis(fluorosulfonyl)amide and bis(perfluoroalkylsulfonyl)amide
¹
1
anions, the strong intensity at ca. 10 cm for [C MIm][FAP] is
2
attributable to the intermolecular vibration, such as libration, of
the anion.
4
.2.3 Imidazolium Cations vs. Pyridinium Cations: So
far, most aromatic ILs studied by fs-RIKES are imidazolium-
based ILs. Indeed, only four pyridinium-based ILs, 1-meth-
90
oxyethylpyridinium dicyanamide, 1-butylpyridinium bis(tri-
fluoromethylsulfonyl)amide,86 1-cyclohexylmethylpyridinium
89
bis(trifluoromethylsulfonyl)amide, and 1-benzylpyridinium
bis(trifluoromethylsulfonyl)amide89 were studied for their low-
frequency vibrational dynamics by fs-RIKES, as far as we
know. In this section, we compare the low-frequency spectra
between imidazolium- and pyridinium-based ILs to clarify the
difference of the two major types of aromatic cations.
Figure 13 displays the low-frequency Kerr spectra of
[C MIm][BF ], [C Py][BF ], [C MIm][NTf ], and [C Py]-
4 4 4 4 10 2 10
[
NTf ]. As seen in the figure, the line shapes of the low-fre-
2
quency Kerr spectra of the imidazolium and pyridinium cation-
based ILs with the same anions and the same alkyl groups are
quite different. Namely, the strong spectral densities at ca. 80­
Figure 13. Low-frequency Kerr spectra of (a) [C4Py][BF4],
b) [C4MIm][BF4], (c) [C10Py][NTf2], and (d) [C10MIm]-
[NTf2]. Pyridinium and imidazolium cations are compared.
(
¹
1
1
00 cm are found in the pyridinium-based ILs compared to
the imidazolium-based ILs. This characteristic difference in the
low-frequency Kerr spectra between pyridinium- and imidazo-
lium-based ILs is also found in the other comparisons, e.g.,
uted to the larger polarizability anisotropy of the pyridinium
cations than the respective methylimidazolium cations. In
addition, the methyl group likely influences the imidazolium
ring libration with the asymmetry and bulkiness in the case of
1-alkyl-3-methylimidazolium based ILs. We will discuss this in
the later section (effects of methylation in aromatic cations).
4.2.4 Alkyl Group Dependence: There are several reports
on the alkyl group dependence (or comparisons of different
alkyl groups) of cations in ILs on the low-frequency Kerr
[C Py][BF ] vs. [C MIm][BF ], [C Py][NTf ] vs. [C MIm]-
6 4 6 4 8 2 8
[
NTf ], etc. (see Supporting Information). A similar feature was
2
¹
also reported in [NTf2] salts of 1-benzyl-3-methylimidazo-
lium, 1-benzylpyridinium, 1-cyclohexylmethyl-3-methylimida-
zolium, and 1-cyclohexylmethylpyridinium cations.89 Because
the large spectral density in the high-frequency region (>ca.
¹
1
5
0 cm ) for the broad low-frequency Kerr spectra of imida-
zolium- and pyridinium-based ILs arises mainly from the
aromatic ring libration, it is plausible that the difference in the
spectra between the pyridinium- and imidazolium-based ILs
is attributed to the difference in the librational motions of the
¹
1 76,82,88
spectra less than ca. 200 cm
.
However, only the series
of 1-alkyl-3-methylimidazolium bis(trifluoromethylsulfonyl)-
amides [C MIm][NTf ] were investigated for the alkyl group
n
2
pyridinium and imidazolium rings. For example, the polar-
dependence of the low-frequency spectra so far. In this study,
we compare three different series: [C MIm][NTf ], [C MIm]-
+
izability anisotropy of [C Py] estimated from the quantum
4
n
2
n
chemistry calculations based on the B3LYP/6-31+G(d,p) level
[BF4], and [CnPy][NTf2]. We focus on the [CnMIm][BF4] and
[C Py][NTf ] series in this section, because they are reported
+
of theory is about 10% larger than that of [C MIm] (Support-
4
n
2
ing Information). The larger spectral intensity in the Kerr
spectra for the pyridinium-based ILs than the imidazolium-
based ILs having same anions and alkyl groups can be attrib-
here for the first time.
Figure 14 shows the low-frequency spectra with the fre-
quency range of 0­300 cm for the [C MIm][BF ] and [C Py]-
¹
1
n
4
n
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1119

Figure 14. Low-frequency Kerr spectra of (a) [C2MIm][BF4], (b) [C3MIm][BF4], (c) [C4MIm][BF4], (d) [C6MIm][BF4], (e)
C8MIm][BF4], (f) [C10MIm][BF4], (g) [C4Py][NTf2], (h) [C6Py][NTf2], (i) [C8Py][NTf2], (j) [C10Py][NTf2], and (k) [C12Py][NTf2].
[
Alkyl group dependence is shown.
[
NTf ] series. Because of the simpler anion’s structure, let us
to the hexyl group, but is increasing by the substitution from
the hexyl group to the decyl group.
2
take a look at the low-frequency spectra of the [C MIm][BF ]
n
4
series first. As seen in Figure 14a, there is a clear intra-
The low-frequency spectrum of [C2MIm][BF4] inspires us
to think that the intramolecular vibrational modes of the
¹1
molecular vibrational mode at 155 cm in [C MIm][BF ].
2
4
¹
1
Further, the spectral intensity at the low-frequency ca. 20 cm
relative to that at the high-frequency ca. 70­100 cm
cations influence the low-frequency spectra. Figure 15 shows
¹
1
+
is
the calculated Raman spectra of the optimized [C MIm] and
n
+
decreasing with alkyl group substitution from the ethyl group
[C Py] cations based on ab initio quantum chemistry calcu-
n
1120 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

respectively. The tendency in the frequency shifts observed
in the present experiments is almost consistent with that of the
ab initio quantum chemistry calculations. We also note that the
spectral area of the bending mode in [C MIm][BF ] that shows
2
4
the clearest bending band is about 15% of the entire broad
¹
1
spectrum from 0 to 200 cm based on the line shape analysis.
Accordingly, the variation of the spectral line shape for the
[
C MIm][BF ] with different alkyl groups of the cations is
n 4
partially attributed to the intramolecular bending modes
between alkyl groups and imidazolium ring. However, the
contribution of the intramolecular vibrational mode to the
broad low-frequency spectrum is more ambiguous in both the
imidazolium- and pyridinium-based ILs whose cation has a
longer alkyl group, as shown in Figure 14. This is probably
because a longer alkyl group possibly forms a wider variety of
conformers that gives a broadened band due to the different
conformers.
As displayed in Figure 15, the bending mode between the
+
alkyl group and pyridinium ring in the [C Py] also depends on
n
¹1
+
¹1
+
the alkyl group: 78 cm for [C Py] ; 53 cm for [C Py] ; 40
4
6
¹
1
+
¹1
+
Figure 15. Calculated Raman spectra of (a) [CnMIm]+
n = 2 (red), 3 (blue), 4 (green), 6 (light blue), 8 (brown),
and 10 (black)) and (b) [CnPy] (n = 4 (green), 6 (light
blue), 8 (brown), and 10 (black)) based on the B3LYP/6-
cm for [C Py] ; 30 cm for [C Py] . Unlike the [C MIm]-
8 10 n
[
BF ] series, it is hard to find the difference among the low-
4
(
+
frequency spectra for the [CnPy][NTf2] series (Figures 14g­14k)
because of the strong intensity of the intermolecular vibrational
band relative to the intramolecular vibrational modes. However,
the quantum chemistry calculation results shown in Figure 15b
suggest that the intramolecular vibrational modes of the
pyridinium cations exist in the low-frequency region from 0
31+G(d,p) level of theory. The bending mode between
aromatic rings and alkyl groups shifts to the lower fre-
quency with the longer alkyl group.
¹
1
to 150 cm though they are buried in the broad intermolecular
vibrational bands.
lations (B3LYP/6-31+G(d,p)). With the coexistence of trans
and gauche conformations of the butyl group of a 1-butyl-3-
methylimidazolium-based IL,2 it is helpful to understand a
general feature of the alkyl group length dependence of the
bending mode of the optimized (all trans alkyl group) cations.
It is found from the quantum chemistry calculation results
that the bending modes between the alkyl groups and the
imidazolium ring depend on the length (or the weight)
Therefore, as shown in Figure 14, the small difference in the
low-frequency spectral shapes of the ILs is likely attributed to
the intramolecular vibrational motions in large part. It should
also be noted that the intensities of the low-frequency bands
mainly coming from intermolecular contributions (ring libra-
tions) of pyridinium-based ILs are larger than those of the
comparable imidazolium-based ILs, as discussed in “imidazo-
lium cations vs. pyridinium cations”. This makes ambiguous
the effect of alkyl group bending mode on the low-frequency
spectrum.
4.2.5 Effect of Methylation in Aromatic Cations: In this
section, we look at the effect of methylation of the cations on
the low-frequency spectra. Because the aromatic ring libration
contributes to the low-frequency Kerr spectra substantially, it
is not surprising that the low-frequency Kerr spectra of the
aromatic cation based ILs are varied by a methylation in the
aromatic cations. In addition, it is known that the proton at 2
position of 1-alkyl-3-methylimidazolium cation is acidic and
the substitution of the proton by methyl group (1-alkyl-2,3-
dimethylimidazolium) affects physical properties, especially ©
02
¹
1
+
¹1
89
of the alkyl group: 136 cm for the [C MIm] ; 104 cm
2
+
¹1
+
¹1
for the [C3MIm] ; 85 cm for the [C4MIm] ; 53 cm for
the [C MIm] ; 39 cm for the [C MIm] ; 27 cm for the
+
¹1
+
¹1
6
8
+
[
C MIm] . Ab initio quantum chemistry calculation results
10
+
+
of [C MIm] and [C MIm] whose conformations of the end
4
6
methyl groups of the alkyl groups are gauche also show that the
bending mode shifts to the lower frequency with increasing the
+
¹1
length of alkyl group (gauche [C MIm] : 99 cm and gauche
4
+
¹1
[
C MIm] : 59 cm ). Therefore, it is plausible that the longer
6
alkyl groups of the cations provide the lower frequencies in the
bending modes, whether they are trans or gauche.
Although it is difficult to exactly distinguish the intra-
molecular vibrational modes from the broad and strong inter-
molecular vibrational bands, except for [C MIm][BF ], it is
1
66,186
and thus · (see Table 2), of the ILs.
We thus believe that
2
4
weakly recognizable from the low-frequency spectra of the
C MIm][BF ] (Figures 14a­14f) that the intramolecular vibra-
tional band shifts to lower frequency with increasing alkyl
group length of the imidazolium ring. The crude frequencies
it is important to study the methylation effect in the aromatic
rings of the aromatic cation based ILs on the low-frequency
Kerr spectra. Actually, Wynne and coworkers compared the
low-frequency Kerr spectra between [C4MIm][NTf2] and
[
n
4
8
2
estimated from the gentle peaks or the shoulder (in [C MIm]-
[C DMIm][NTf ] in 2003. We further explore the methyl-
4 2
3
[
BF ]) of the spectra are approximately 155, 102, 83, 69, 55,
ation effect on the low-frequency Kerr spectra of aromatic
cation based ILs, that is, the comparisons of [1C 2MPy][NTf ],
4
¹
1
and 39 cm
for [C MIm][BF ], [C MIm][BF ], [C MIm]-
2 4 3 4 4
4
2
[
BF ], [C MIm][BF ], [C MIm][BF ], and [C MIm][BF ],
[1C 3MPy][NTf ], [1C 4MPy][NTf ], and [C Py][NTf ] in
4 2 4 2 4 2
4
6
4
8
4
10
4
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1121

Figure 16. Low-frequency Kerr spectra of (a) [C4DMIm]-
NTf2] and (b) [C4MIm][NTf2]. Methylation effect in the
[
imidazolium cation on the low-frequency spectrum is
shown.
addition to the comparison between [C DMIm][NTf ] and
4
2
[C MIm][NTf ] in this study.
4 2
Figure 16 compares the low-frequency spectra between
C DMIm][NTf ] and [C MIm][NTf ]. The difference in the
[
4
2
4
2
spectra is not large, but the amplitude of the spectral density in
¹
1
the high-frequency region (ca. 90 cm ) relative to the low-
¹1
frequency region (ca. 10­20 cm ) becomes large by methyl-
ation. This feature was already reported by Wynne and co-
8
2
workers. Figure 17 shows further comparisons of the low-
frequency Kerr spectra of [1C 2MPy][NTf ], [1C 3MPy]-
4
2
4
[
NTf ], and [1C 4MPy][NTf ] together with [C Py][NTf ]. In
2 4 2 4 2
Figures 16 and 17, we clearly see several noticeable points: (i)
the spectral line shape of [1C 3MPy][NTf ] is apparently
Figure 17. Low-frequency Kerr spectra of (a) [1C42MPy]-
4
2
[
(
NTf2], (b) [1C43MPy][NTf2], (c) [1C44MPy][NTf2], and
d) [C4Py][NTf2]. Methylation effects of positions in the
different from those of [1C 2MPy][NTf ], [1C 4MPy][NTf ],
4
2
4
2
and [C Py][NTf ]; (ii) the spectral shape for [1C 3MPy][NTf ]
4
2
4
2
pyridinium cations are shown.
is closer to that of [C4MIm][NTf2] instead of the other
pyridinium-based ILs; (iii) the spectra of [1C 2MPy][NTf ],
4
2
[
1C 4MPy][NTf ], and [C Py][NTf ] are similar but not exact-
down the ring libration less than that at 3 position because of
the steric effect for the libration along near the butyl group.
This is also well supported by the trend of the fastest rotational
4
2
4
2
ly the same (e.g., the peaks in the high-frequency region above
5
¹
1
¹1
0 cm are ca. 95.4, 81.6, and 89.6 cm , respectively).
As discussed above, the low-frequency Kerr spectra, espe-
+
+
constants (3.69 GHz for [C Py] , 3.35 GHz for [1C 4MPy] ,
4
4
¹
1
+
cially in the high-frequency region above 50 cm , of aromatic
cation based ILs arise mainly from the aromatic ring librations.
The peaks in this high-frequency region for [C4DMIm]-
2.35 GHz for [1C 3MPy] , see Supporting Information). On
4
the other hand, [1C 2MPy][NTf ] shows the highest frequency
4
2
of the peak in the high-frequency region among these
[
NTf ], [C MIm][NTf ], [1C 2MPy][NTf ], [1C 3MPy][NTf ],
pyridinium-based ILs, despite that the fastest rotational con-
2
4
2
4
2
4
2
+
[1C 4MPy][NTf ], and [C Py][NTf ] are ca. 89.2, 79.8, 95.4,
stant for [1C 2MPy] is 2.18 GHz (Supporting Information),
4
2
4
2
4
¹
1
7
8.6, 81.6, and 89.6 cm , respectively. Because the long butyl
being smaller than those for the other three cations. This feature
is also similar to the case of [C DMIm][NTf ] compared to
group bonded with the cation rings, the librations of the rings
responsible for the low-frequency spectra likely occur about
the axis along near the butyl group direction. Therefore, it is
expected that the libration of an aromatic ring slows down
with the methylation in the aromatic rings. When we com-
pare the spectra of [1C 3MPy][NTf ], [1C 4MPy][NTf ], and
4
2
[C MIm][NTf ]. It is speculated that the pyridinium ring of
4
2
+
+
[1C 2MPy] and the imidazolium ring of [C DMIm] cannot
4
4
fully librate about the axis near the butyl group direction
because of the steric hindrance of the methyl group to the butyl
group that provides a steeper potential and this causes the high-
frequency librations. The bulk parameters, © and £, for these
ILs should be also noted. As shown in Table 2, both © and £
for [1C 2MPy][NTf ] are larger than those for [C Py][NTf ],
4
2
4
2
[
C Py][NTf ], the trend of the peak frequencies can be rea-
4 2
sonably explained by this consideration. Of course, the methyl
group at the 4 position of the pyridinium ring most likely slows
4
2
4
2
1122 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

Figure 19. Plots of the first moment M1 of low-frequency
band vs. bulk parameter (£/µ)1 for aromatic cations (red
circles). Linear fit is also shown by red line. For com-
parison, the relation for aprotic molecular liquids is also
shown by black line.1
/2
2,110,203
similar vibrational spectral shapes. Such a behavior was
observed in 1-methoxyethylpyridinium dicyanamide and a
9
0
1
:1 mixture of methoxyethylbenzene and dicyanomethane.
In addition to the librations, we need to take the effects of the
intramolecular vibrational modes on the low-frequency spec-
tra into consideration. Figure 18 shows the calculated Raman
+
+
+
+
spectra of [C DMIm] , [C MIm] , [1C 2MPy] , [1C 3MPy] ,
[
4
+
4
+
4
4
1C 4MPy] , and [C Py] cations by ab initio quantum chem-
4 4
istry calculations based on the B3LYP/6-31+G(d,p) level of
theory. It is clear from Figure 18 that there are intramolecular
¹
1
vibrational modes in the frequency range of at 50­100 cm .
The strong modes are the bending modes between the butyl
group and aromatic rings with different coordinate axes. The
frequencies of the bands are not very different among the
6
cations, and they do not correlate with the line shapes of
the experimental low-frequency Kerr spectra. Accordingly, the
methylation effects on the low-frequency spectra for the present
pyridinium- and imidazolium-based ILs are mostly due to the
ring librations, but not the intramolecular vibrational modes.
4
.3 Relation between Low-Frequency Spectrum and Bulk
Parameters. Understanding/clarifying the relationship
+
between the microscopic aspects such as the intermolecular
vibrational band and the bulk parameters in liquids is a major
issue and challenge in chemistry. Previously, we reported data
for the intermolecular vibrational dynamics of 40 aprotic
Figure 18. Calculated Raman spectra of (a) [C4DMIm] ,
+
+
+
(
[
b) [C4MIm] , (c) [1C42MPy] , (d) [1C43MPy] , (e)
1C44MPy] , and (f) [C4Py] based on the B3LYP/6-
1+G(d,p) level of theory.
+
+
3
1
2
12
molecular liquids studied by fs-RIKES. In that study, we
found correlations between the first moment (M1) of the low-
frequency Kerr spectrum and the bulk parameters such as the
square root of the surface tension (£) divided by the liquid
[
[
£
[
1C43MPy][NTf2], and [1C44MPy][NTf2], but [C4Py][NTf2],
1C 3MPy][NTf ], and [1C 4MPy][NTf ] show similar © and
. © and £ for [C DMIm][NTf ] are also larger than those for
4 2
C MIm][NTf ]. Thus, the stronger intermolecular interactions
for [1C 2MPy][NTf ] and [C DMIm][NTf ] compared to their
comparable ILs can also provide the higher frequencies of the
librations of the aromatic rings.
4
2
4
2
pffiffiffiffiffiffiffiffi
pffiffiffiffiffiffiffiffiffiffiffiffiffi
density (µ) and FW: £=µ and £=FW. The idea of the plots
was based on a simple consideration of the intermolecular
vibrations in analogy with the harmonic oscillator, but we used
£ and µ instead of the force constant and reduced mass in the
harmonic oscillator model because of its intermolecular nature
and liquid state. Accordingly, the relation between the micro-
scopic (low-frequency spectrum) and bulk (£ and µ) quantities,
if there is one, implies a kind of scaling relation between them.
We previously suggested from the data of 20 aromatic cation-
based ILs and 10 nonaromatic cation based ILs that the corre-
4
2
4
2
4
2
Regarding the similarity in the spectral shapes for [C MIm]-
4
[
NTf2] and [1C43MPy][NTf2], it is not clear in the origin and it
might be coincidence. However, the geometrical structures
positions of methyl group and butyl group at aromatic rings)
of the two cations are similar. This implies that the aromatic
cations having the similar geometrical structures provide the
(
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1123

lations were very different between aromatic and nonaromatic
imidazolium and pyridinium cations provide the similar bulk
properties of ILs, but the bulk properties of ILs vary signifi-
cantly with the anion species. On the basis of the results of the
6
0
cation based ILs. In this section, we see the correlation
pffiffiffiffiffiffiffiffi
between M and £=µ for the present aromatic cation based
1
+
+
ILs of as many as 40 samples.
Figure 19 shows the plots of M vs. £=µ for the data of
bulk properties of the ILs for [C MIm] and [C MIm] cations,
2 4
pffiffiffiffiffiffiffiffi
we found the following trends:
Density: [FAP] > [NPf ] > I > [NTf ] > [PF ] µ [NF ] >
2 2 6 2
1
¹
¹
¹
¹
¹
¹
the present 40 aromatic ILs together with the linear fit. The
correlation for aprotic molecular liquids based on the data of
¹
¹
¹
¹
¹
¹
[OTf] > [BF4] µ [NO3] µ [EtSO4] > [SCN] µ [DCA] >
1
2,110,203
¹
4
9 liquids
is also shown as a reference. As clearly seen
[TCNB] .
+
¹
¹
¹
in Figure 19, the correlation for the aromatic ILs is obviously
different from that for the aprotic molecular liquids. The M1
values for most aromatic ILs studied here are higher than the
values expected from the correlation for the aprotic molecular
liquids. In addition, the slope in the correlation of the aromatic
ILs is much gentler than that for the aprotic molecular liquids.
Note that the data of the correlation for the aprotic molecular
liquids include both aromatic and nonaromatic molecular
liquids. It is found in Figure 19 that the feature of the present
correlation for the aromatic ILs based on the 40 samples is
quite similar to that of the previous correlation on the basis of
Viscosity ([C MIm] -based ILs): I > [PF ] > [NO ] >
4 6 3
¹
¹
¹
¹
¹
¹
[NPf ] > [BF ] µ [OTf] > [SCN] > [NTf ] > [NF ] µ
2
4
2
2
¹
[DCA] .
Viscosity ([C MIm] -based ILs): [EtSO ] > [NPf ] >
+
¹
¹
2
¹
4
2
¹
¹
¹
¹
[FAP] > [OTf] > [BF ] > [NTf ] > [TCNB] .
4
2
+
¹
Electrical conductivity ([C MIm] -based ILs): [DCA] >
4
¹
¹
¹
¹
¹
¹
[NF2] > [SCN] > [BF4] µ [NTf2] > [OTf] > [NO3] >
¹
¹
¹
[NPf ] µ [PF ] > I .
2
6
+
¹
Electrical conductivity ([C MIm] -based ILs): [BF ] µ
2
4
¹
¹
¹
¹
¹
¹
[TCNB] > [NTf ] µ [OTf] > [FAP] > [EtSO ] µ [NPf ] .
Surface tension: I > [NO ] µ [SCN] > [DCA] > [BF ] µ
[NF ] µ [PF ] > [TCNB] µ [EtSO ] > [OTf] > [NTf ] >
2 6 4 2
2
4
2
¹
¹
¹
¹
¹
3
4
6
0
¹
¹
¹
¹
¹
¹
2
0 aromatic ILs.
The weak bulk parameter dependence of M in the aromatic
¹
¹
[FAP] > [NPf ] .
1
2
ILs compared to molecular liquids and nonaromatic ILs was
discussed in a previous report.60 In that report, we attributed
We also found a good correlation between the molar
electrical conductivity and the inverse of the viscosity for the
present 40 ILs.
60
this moderate bulk parameter dependence of M to the more
1
heterogeneous structure and the stronger signal of a charged
aromatic moiety. Our scenario is the following. Aromatic ILs
In the second part, the low-frequency Kerr spectra of the
aromatic cation based ILs are discussed. Because we exper-
imentally obtained low-frequency Kerr spectra of as many as
40 aromatic cation based ILs in this study, we focused on
several topics, that is, (i) anion dependence; (ii) imidazolium
cations vs. pyridinium cations, (iii) alkyl group dependence,
and (iv) effect of methylation in aromatic cations. In the first
subsection, anion dependence, we looked at the low-frequency
spectra of several ILs with different anions. The line shapes of
the low-frequency spectra of the ILs vary significantly with the
different anion species. Even in the simple spherical top anions,
are more segregated than nonaromatic ILs when the same anion
and the same alkyl group of cations are compared.5
7,204
The
charged aromatic moieties, such as imidazolium and pyridi-
nium rings, have stronger signals than the charged nonaromatic
parts, such as pyrrolidinium and ammonium groups. As a
result, fs-RIKES probes more local information on the ionic
region in aromatic ILs but more delocalized and homogeneous
information in nonaromatic ILs. Of course, the neutral molec-
ular liquids studied for the correlation that is rather simple
do not show such a segregation structure. Therefore, the low-
frequency intermolecular vibrational band is weakly dependent
on the bulk parameter in the case of aromatic ILs, but it is
strongly dependent in the cases of aprotic molecular liquids and
nonaromatic ILs. The present study supports this scenario
strongly in terms of aromatic cation based ILs because of the
¹
¹
¹
e.g., I , [BF ] , [PF ] , the ILs show distinct low-frequency
4
6
spectra. The ILs with different bis(fluorosulfonyl)amide and
¹
¹
bis(perfluoroalkylsulfonlyl)amide anions ([NF ] , [NTf ] , and
2
2
¹
¹1
[NPf2] ) show a unique peak at ca. 10­30 cm that depends
slightly on the anion species. We also picked up some other ILs
that show general and unique spectral shapes. In the second
subsection, imidazolium cations vs. pyridinium cations, we
showed the difference in the low-frequency spectra between the
6
0
much larger number of data than the previous report. We
are planning to study nonaromatic cation based ILs, such as
pyrrolidinium-, ammonium-, and phosphonium-based ILs, to
imidazolium- and pyridinium-based ILs. When we compared
pffiffiffiffiffiffiffiffi
+
see the correlation between M and £=µ for nonaromatic
the low-frequency Kerr spectra between the [C MIm] and
1
n
+
cation based ILs in the near future.
[C Py] cation-based ILs having the same anions and alkyl
n
groups, the pyridinium-based ILs showed the stronger intensity
5
. Summary
¹1
in the high-frequency region (>50 cm ) that is attributed to
In this study, we reported data on the ultrafast dynamics of
0 aromatic cation based ILs measured by fs-RIKES, as well as
the ring libration than the imidazolium-based ILs. This is an
indication that the pyridinium ring libration is more active
in depolarized Raman signal than the imidazolium ring libra-
tion. The alkyl group dependence of the cations on the low-
frequency spectra was discussed in the third subsection, alkyl
group dependence. The line shapes of the low-frequency
spectra of both the [CnMIm] - and [CnPy] -based ILs depend
on the alkyl group of the cations. On the basis of the quantum
chemistry calculation results, we attribute the alkyl group
dependence of the low-frequency spectra to the intramolecular
vibrational modes (bending modes between the alkyl groups
4
some bulk properties including density, viscosity, electrical
conductivity, and surface tension, of the ILs. First we showed
and discussed the results of the bulk properties of the aromatic
cation based ILs. From the results of the densities, viscosities,
electrical conductivities, and surface tensions for the [CnMIm]-
+
+
[
NTf ], [C MIm][BF ], and [C Py][NTf ], we found that the
2 n 4 n 2
alkyl group dependences of these bulk properties of the
C MIm][NTf ] and [C Py][NTf ] are quite similar, but that
[
n
2
n
2
of the [C MIm][BF ] is distinct. The results indicate that the
n
4
1124 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

and imidazolium ring) of the cations. We also showed the
methylation effects of the cations on the low-frequency spectra
in effect of methylation in aromatic cations.
In the last part, we discussed the relation between the low-
frequency Kerr spectrum and bulk properties. We confirmed
Chem. Soc. Jpn. 2009, 82, 1347.
13 B. J. Loughnane, R. A. Farrer, A. Scodinu, T. Reilly, J. T.
Fourkas, J. Phys. Chem. B 2000, 104, 5421.
1
1
4
5
R. A. Farrer, J. T. Fourkas, Acc. Chem. Res. 2003, 36, 605.
N. T. Hunt, A. A. Jaye, S. R. Meech, Phys. Chem. Chem.
Phys. 2007, 9, 2167.
K. Polok, W. Gadomski, B. Ratajska-Gadomska, Rev. Sci.
Instrum. 2015, 86, 103109.
C. A. Schmuttenmaer, Chem. Rev. 2004, 104, 1759.
18 J. B. Baxter, G. W. Guglietta, Anal. Chem. 2011, 83, 4342.
C. J. Fecko, J. D. Eaves, A. Tokmakoff, J. Chem. Phys.
002, 117, 1139.
J. D. Eaves, C. J. Fecko, A. L. Stevens, P. Peng, A.
Tokmakoff, Chem. Phys. Lett. 2003, 376, 20.
P. P. Wiewiór, H. Shirota, E. W. Castner, Jr., J. Chem. Phys.
002, 116, 4643.
I. A. Heisler, S. R. Meech, J. Phys. Chem. B 2008, 112,
that the first moment M of the low-frequency spectral band
correlates to the bulk parameter £=µ for the present aromatic
cation based ILs. The correlation for the aromatic cation based
ILs is much gentler than that for aprotic molecular liquids. The
1
pffiffiffiffiffiffiffiffi
1
6
1
7
pffiffiffiffiffiffiffiffi
difference in the correlation of M and £=µ between the
1
1
9
aromatic cation based ILs and aprotic molecular liquids is
likely attributed to the differences in the microstructure and
probing space scale: the Kerr spectra of aromatic cation based
ILs are mostly coming from the ionic region via the strongly
Kerr active aromatic moieties; on the other hand the spectra
of the aprotic molecular liquids include rather homogeneous
and thus delocalized information on the liquids. As a result, the
low-frequency Kerr spectra of the aromatic cation based ILs are
less sensitive to the bulk properties than the aprotic molecular
liquids.
2
2
0
21
2
2
2
12976.
23 B. J. Berne, R. Pecora, Dynamic Light Scattering: With
Applications to Chemistry, Biology, and Physics, Dover Publica-
tions, Inc., Mineola, NY, 2000.
2
4
A. Tokmakoff, J. Chem. Phys. 1996, 105, 1.
This work was supported in part by Grant-in-Aid (No.
5K05377) by the Japan Society for the Promotion of Science
HS). TF appreciates Saga University for the Research Support
25 P. F. Barbara, W. Jarzeba, Adv. Photochem. 2007, 15, 1.
2
2
6
7
M. Maroncelli, J. Mol. Liq. 1993, 57, 1.
M. L. Horng, J. A. Gardecki, A. Papazyan, M. Maroncelli,
1
(
J. Phys. Chem. 1995, 99, 17311.
A. Samanta, J. Phys. Chem. B 2006, 110, 13704.
29 A. Samanta, J. Phys. Chem. Lett. 2010, 1, 1557.
E. W. Castner, Jr., J. F. Wishart, H. Shirota, Acc. Chem. Res.
007, 40, 1217.
Program for Young Researchers (2015).
2
8
Supporting Information
3
0
Data of the Kerr transients and Fourier-transform Kerr
spectra and lists of the fit parameters for the Kerr transients and
the Kerr spectra of the ILs are summarized in Supporting
Information. Synthesis and purification procedures for some
ILs synthesized in this study are also described in it. Ab initio
quantum chemistry calculation results are summarized, as well.
This material is available on http://dx.doi.org/10.1246/bcsj.
2
3
3
1
2
Y. Nagasawa, J. Photochem. Photobiol., C 2011, 12, 31.
J. A. Ingram, R. S. Moog, N. Ito, R. Biswas, M. Maroncelli,
J. Phys. Chem. B 2003, 107, 5926.
N. Ito, S. Arzhantsev, M. Heitz, M. Maroncelli, J. Phys.
Chem. B 2004, 108, 5771.
X.-X. Zhang, M. Liang, N. P. Ernsting, M. Maroncelli,
J. Phys. Chem. B 2013, 117, 4291.
X.-X. Zhang, J. Breffke, N. P. Ernsting, M. Maroncelli,
3
3
3
4
2
0160085.
3
5
References
Phys. Chem. Chem. Phys. 2015, 17, 12949.
36 Y. Kimura, M. Fukuda, K. Suda, M. Terazima, J. Phys.
Chem. B 2010, 114, 11847.
37 H. K. Kashyap, R. Biswas, J. Phys. Chem. B 2008, 112,
12431.
38 P. Walden, Bull. Acad. Imper. Sci. (St. Petersburg) 1914,
405.
39 S. Gabriel, J. Weiner, Ber. Dtsch. Chem. Ges. 1888, 21,
2669.
1
A. Nitzan, Chemical Dynamics in Condensed Phases:
Relaxation, Transfer and Reactions in Condensed Molecular
Systems, Oxford University Press, Oxford, 2006.
2
G. R. Fleming, Chemical Applications of Ultrafast Spec-
troscopy, Oxford University Press, New York, 1986.
M. D. Levenson, S. S. Kano, Introduction to Nonlinear
Laser Spectroscopy, Revised ed., Academic Press, New York,
3
1
988.
4
S. Mukamel, Principles of Nonlinear Optical Spectroscopy,
40 J. S. Wilkes, M. J. Zaworotko, J. Chem. Soc., Chem.
Commun. 1992, 965.
Oxford University Press, New York, 1995.
D. McMorrow, W. T. Lotshaw, G. A. Kenney-Wallace,
IEEE J. Quantum Electron. 1988, 24, 443.
W. T. Lotshaw, D. McMorrow, N. Thantu, J. S. Melinger,
R. Kitchenham, J. Raman Spectrosc. 1995, 26, 571.
5
41 Ionic Liquids in Synthesis, 2nd ed., ed. by P. Wasserscheid,
T. Welton, Wiley-VCH, Weinheim, 2008.
42 Electrochemical Aspects of Ionic Liquids, ed. by H. Ohno,
Wiley-Interscience, Hoboken, 2005. doi:10.1002/0471762512.
43 J. S. Wilkes, J. Mol. Catal. A: Chem. 2004, 214, 11.
44 K. R. Seddon, J. Chem. Technol. Biotechnol. 1997, 68, 351.
45 J. H. Davis, Jr., Chem. Lett. 2004, 33, 1072.
46 T. Welton, Coord. Chem. Rev. 2004, 248, 2459.
47 N. V. Plechkova, K. R. Seddon, Chem. Soc. Rev. 2008, 37,
123.
6
7
8
R. Righini, Science 1993, 262, 1386.
S. Kinoshita, Y. Kai, T. Ariyoshi, Y. Shimada, Int. J. Mod.
Phys. B 1996, 10, 1229.
9
0
E. W. Castner, Jr., M. Maroncelli, J. Mol. Liq. 1998, 77, 1.
1
N. A. Smith, S. R. Meech, Int. Rev. Phys. Chem. 2002, 21,
Q. Zhong, J. T. Fourkas, J. Phys. Chem. B 2008, 112,
H. Shirota, T. Fujisawa, H. Fukazawa, K. Nishikawa, Bull.
7
1
5.
1
1
48 H. Ohno, K. Fukumoto, Electrochemistry 2008, 76, 16.
49 H. Weingärtner, Angew. Chem., Int. Ed. Engl. 2008, 47,
654.
5529.
1
2
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1125

5
0
M. Armand, F. Endres, D. R. MacFarlane, H. Ohno, B.
Bartsch, E. L. Quitevis, Chem. Phys. Lett. 2010, 497, 37.
82 G. Giraud, C. M. Gordon, I. R. Dunkin, K. Wynne,
J. Chem. Phys. 2003, 119, 464.
83 H. Shirota, E. W. Castner, Jr., J. Phys. Chem. B 2005, 109,
21576.
Scrosati, Nat. Mater. 2009, 8, 621.
5
5
1
2
J. F. Wishart, Energy Environ. Sci. 2009, 2, 956.
E. W. Castner, Jr., J. F. Wishart, J. Chem. Phys. 2010, 132,
1
20901.
5
5
5
3
4
5
Q. Zhang, J. M. Shreeve, Chem. Rev. 2014, 114, 10527.
Y. Wang, G. A. Voth, J. Am. Chem. Soc. 2005, 127, 12192.
J. N. A. C. Lopes, A. A. H. Pádua, J. Phys. Chem. B 2006,
84 H. Shirota, J. F. Wishart, E. W. Castner, Jr., J. Phys. Chem.
B 2007, 111, 4819.
85 H. Shirota, K. Nishikawa, T. Ishida, J. Phys. Chem. B 2009,
113, 9831.
1
10, 3330.
5
29.
6
S. Shigeto, H. Hamaguchi, Chem. Phys. Lett. 2006, 427,
86 T. Fujisawa, K. Nishikawa, H. Shirota, J. Chem. Phys.
2009, 131, 244519.
3
5
7
A. Triolo, O. Russina, H.-J. Bleif, E. Di Cola, J. Phys.
87 H. Fukazawa, T. Ishida, H. Shirota, J. Phys. Chem. B 2011,
115, 4621.
88 H. Shirota, T. Ishida, J. Phys. Chem. B 2011, 115, 10860.
89 H. Shirota, H. Matsuzaki, S. Ramati, J. F. Wishart, J. Phys.
Chem. B 2015, 119, 9173.
Chem. B 2007, 111, 4641.
J. C. Araque, J. J. Hettige, C. J. Margulis, J. Phys. Chem. B
015, 119, 12727.
H. Shirota, H. Fukazawa, in Ionic Liquids: Theory, Proper-
ties, New Approaches, ed. by A. Kokorin, InTech, Rijeka, Croatia,
011, Chap. 9, pp. 201­224. doi:10.5772/14817.
5
8
2
5
9
90 H. Shirota, E. W. Castner, Jr., J. Phys. Chem. A 2005, 109,
9388.
2
6
6
0
1
H. Shirota, ChemPhysChem 2012, 13, 1638.
M. L. T. Asaki, A. Redondo, T. A. Zawodzinski, A. J.
91 H. Shirota, A. M. Funston, J. F. Wishart, E. W. Castner, Jr.,
J. Chem. Phys. 2005, 122, 184512.
Taylor, J. Chem. Phys. 2002, 116, 10377.
K. Yamamoto, M. Tani, M. Hangyo, J. Phys. Chem. B
007, 111, 4854.
M. Koeberg, C.-C. Wu, D. Kim, M. Bonn, Chem. Phys.
Lett. 2007, 439, 60.
A. Chakraborty, T. Inagaki, M. Banno, T. Mochida, K.
Tominaga, J. Phys. Chem. A 2011, 115, 1313.
D. A. Turton, J. Hunger, A. Stoppa, G. Hefter, A. Thoman,
M. Walther, R. Buchner, K. Wynne, J. Am. Chem. Soc. 2009, 131,
1140.
92 H. Shirota, H. Fukazawa, T. Fujisawa, J. F. Wishart,
J. Phys. Chem. B 2010, 114, 9400.
93 D. A. Turton, T. Sonnleitner, A. Ortner, M. Walther, G.
Hefter, K. R. Seddon, S. Stana, N. V. Plechkova, R. Buchner, K.
Wynne, Faraday Discuss. 2012, 154, 145.
94 H. Shirota, S. Kakinuma, J. Phys. Chem. B 2015, 119,
9835.
95 D. Xiao, J. R. Rajian, S. Li, R. A. Bartsch, E. L. Quitevis,
J. Phys. Chem. B 2006, 110, 16174.
96 D. Xiao, J. R. Rajian, L. G. Hines, Jr., S. Li, R. A. Bartsch,
E. L. Quitevis, J. Phys. Chem. B 2008, 112, 13316.
97 D. Xiao, L. G. Hines, Jr., R. A. Bartsch, E. L. Quitevis,
J. Phys. Chem. B 2009, 113, 4544.
98 P. Yang, G. A. Voth, D. Xiao, L. G. Hines, Jr., R. A.
Bartsch, E. L. Quitevis, J. Chem. Phys. 2011, 135, 034502.
99 F. Bardak, D. Xiao, L. G. Hines, Jr., P. Son, R. A. Bartsch,
E. L. Quitevis, P. Yang, G. A. Voth, ChemPhysChem 2012, 13,
1687.
6
2
2
6
3
6
4
6
5
1
6
6
T. Sonnleitner, D. A. Turton, S. Waselikowski, J. Hunger,
A. Stoppa, M. Walther, K. Wynne, R. Buchner, J. Mol. Liq. 2014,
92, 19.
T. Sonnleitner, D. A. Turton, G. Hefter, A. Ortner, S.
1
6
7
Waselikowski, M. Walther, K. Wynne, R. Buchner, J. Phys. Chem.
B 2015, 119, 8826.
6
8
K. Iwata, H. Okajima, S. Saha, H. Hamaguchi, Acc. Chem.
Res. 2007, 40, 1174.
69
70
71
M. C. C. Ribeiro, J. Chem. Phys. 2010, 133, 024503.
M. C. C. Ribeiro, J. Chem. Phys. 2011, 134, 244507.
K. Fumino, A. Wulf, R. Ludwig, Angew. Chem., Int. Ed.
100 L. Xue, G. Tamas, E. Gurung, E. L. Quitevis, J. Chem.
Phys. 2014, 140, 164512.
101 R. M. Lynden-Bell, L. Xue, G. Tamas, E. L. Quitevis,
J. Chem. Phys. 2014, 141, 044506.
Engl. 2008, 47, 3830.
K. Fumino, A. Wulf, R. Ludwig, Angew. Chem., Int. Ed.
Engl. 2008, 47, 8731.
K. Fumino, A. Wulf, S. P. Verevkin, A. Heintz, R. Ludwig,
ChemPhysChem 2010, 11, 1623.
A. Wulf, K. Fumino, R. Ludwig, Angew. Chem., Int. Ed.
Engl. 2010, 49, 449.
K. Fumino, S. Reimann, R. Ludwig, Phys. Chem. Chem.
Phys. 2014, 16, 21903.
B.-R. Hyun, S. V. Dzyuba, R. A. Bartsch, E. L. Quitevis,
J. Phys. Chem. A 2002, 106, 7579.
J. R. Rajian, S. Li, R. A. Bartsch, E. L. Quitevis, Chem.
Phys. Lett. 2004, 393, 372.
D. Xiao, J. R. Rajian, A. Cady, S. Li, R. A. Bartsch, E. L.
Quitevis, J. Phys. Chem. B 2007, 111, 4669.
D. Xiao, L. G. Hines, Jr., S. Li, R. A. Bartsch, E. L.
Quitevis, O. Russina, A. Triolo, J. Phys. Chem. B 2009, 113, 6426.
O. Russina, A. Triolo, L. Gontrani, R. Caminiti, D. Xiao,
7
2
102 L. Xue, G. Tamas, R. P. Matthews, A. J. Stone, P. A. Hunt,
E. L. Quitevis, R. M. Lynden-Bell, Phys. Chem. Chem. Phys.
2015, 17, 9973.
103 L. Xue, G. Tamas, E. L. Quitevis, ACS Sustainable Chem.
Eng. 2016, 4, 514.
104 H. Shirota, R. Biswas, J. Phys. Chem. B 2012, 116, 13765.
105 H. Shirota, J. Phys. Chem. B 2013, 117, 7985.
106 H. Shirota, S. Kakinuma, Y. Itoyama, T. Umecky, T.
Takamuku, J. Phys. Chem. B 2016, 120, 513.
107 H. Shirota, T. Mandai, H. Fukazawa, T. Kato, J. Chem.
Eng. Data 2011, 56, 2453.
7
3
7
4
7
5
7
6
7
7
108 T. Masaki, K. Nishikawa, H. Shirota, J. Phys. Chem. B
2010, 114, 6323.
7
8
109 H. Shirota, J. Chem. Phys. 2005, 122, 044514.
110 H. Shirota, J. Phys. Chem. A 2011, 115, 14262.
111 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P.
Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg,
M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida,
7
9
8
0
L. G. Hines, Jr., R. A. Bartsch, E. L. Quitevis, N. Plechkova, K. R.
Seddon, J. Phys.: Condens. Matter 2009, 21, 424121.
8
1
D. Xiao, L. G. Hines, Jr., M. W. Holtz, K. Song, R. A.
1126 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan

T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. J. A.
Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E.
Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S.
Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.
Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.
Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G.
Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S.
Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz,
J. Cioslowski, D. J. Fox, Gaussian 09, Revision A.02, Gaussian,
Inc., Wallingford CT, 2009.
141 H. Tokuda, S. Tsuzuki, Md. A. B. H. Susan, K. Hayamizu,
M. Watanabe, J. Phys. Chem. B 2006, 110, 19593.
142 J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D.
Willauer, G. A. Broker, R. D. Rogers, Green Chem. 2001, 3, 156.
143 A. Nazet, S. Sokolov, T. Sonnleitner, T. Makino, M.
Kanakubo, R. Buchner, J. Chem. Eng. Data 2015, 60, 2400.
144 A. Kobayashi, K. Osawa, M. Terazima, Y. Kimura, Phys.
Chem. Chem. Phys. 2012, 14, 13676.
145 J. Wei, T. Ma, X. Ma, W. Guan, Q. Liu, J. Yang, RSC Adv.
2014, 4, 30725.
146 D. Almantariotis, T. Gefflaut, A. A. H. Pádua, J.-Y. Coxam,
M. F. C. Gomes, J. Phys. Chem. B 2010, 114, 3608.
147 B. Bittner, R. J. Wrobel, E. Milchert, J. Chem. Thermodyn.
2012, 55, 159.
112 D. McMorrow, W. T. Lotshaw, Chem. Phys. Lett. 1990,
1
1
1
2
74, 85.
113 D. McMorrow, W. T. Lotshaw, J. Phys. Chem. 1991, 95,
148 N. M. Yunus, M. I. A. Mutalib, Z. Man, M. A. Bustam, T.
Murugesan, J. Chem. Thermodyn. 2010, 42, 491.
149 I. Bandrés, F. M. Royo, I. Gascón, M. Castro, C. Lafuente,
J. Phys. Chem. B 2010, 114, 3601.
0395.
114 H. Cang, J. Li, M. D. Fayer, J. Chem. Phys. 2003, 119,
3017.
115 J. Li, I. Wang, K. Fruchey, M. D. Fayer, J. Phys. Chem. A
150 Q.-G. Zhang, Y. Wei, S.-S. Sun, C. Wang, M. Yang, Q.-S.
Liu, Y.-A. Gao, J. Chem. Eng. Data 2012, 57, 2185.
151 L. G. Sánchez, J. R. Espel, F. Onink, G. W. Meindersma,
A. B. de Haan, J. Chem. Eng. Data 2009, 54, 2803.
152 Q.-S. Liu, M. Yang, P.-P. Li, S.-S. Sun, U. Welz-Biermann,
Z.-C. Tan, Q.-G. Zhang, J. Chem. Eng. Data 2011, 56, 4094.
153 D. Tomida, S. Kenmochi, K. Qiao, T. Tsukada, C.
Yokoyama, Fluid Phase Equilib. 2013, 340, 31.
154 K. R. Seddon, A. Stark, M.-J. Torres, ACS Symp. Ser. 2002,
819, 34.
006, 110, 10384.
1
16 M. D. Fayer, Chem. Phys. Lett. 2014, 616­617, 259.
117 J. A. Bucaro, T. A. Litovitz, J. Chem. Phys. 1971, 54, 3846.
118 Y. J. Chang, E. W. Castner, Jr., J. Chem. Phys. 1993, 99,
7
289.
1
1
1
19 Y. Tanimura, S. Mukamel, J. Chem. Phys. 1993, 99, 9496.
20 Y. Nagata, T. Hasegawa, Y. Tanimura, J. Chem. Phys. 2006,
24, 194504.
1
1
1
21 S. Ryu, R. M. Stratt, J. Phys. Chem. B 2004, 108, 6782.
22 G. Tao, R. M. Stratt, J. Phys. Chem. B 2006, 110, 976.
23 M. D. Elola, B. M. Ladanyi, J. Chem. Phys. 2005, 122,
155 Z.-B. Zhou, H. Matsumoto, K. Tatsumi, ChemPhysChem
2005, 6, 1324.
2
1
1
1
1
1
24506.
156 M. G. Freire, A. R. R. Teles, M. A. A. Rocha, B. Schröder,
C. M. S. S. Neves, P. J. Carvalho, D. V. Evtuguin, L. M. N. B. F.
Santos, J. A. P. Coutinho, J. Chem. Eng. Data 2011, 56, 4813.
157 G. P. Pandey, S. A. Hashmi, Bull. Mater. Sci. 2013, 36, 729.
158 G. B. Dutt, J. Phys. Chem. B 2010, 114, 8971.
159 C. Daguenet, P. J. Dyson, I. Krossing, A. Oleinikova, J.
Slattery, C. Wakai, H. Weingärtner, J. Phys. Chem. B 2006, 110,
12682.
160 A. L. Sturlaugson, K. S. Fruchey, M. D. Fayer, J. Phys.
Chem. B 2012, 116, 1777.
161 K. Nakamura, T. Shikata, ChemPhysChem 2010, 11, 285.
162 S. Coleman, R. Byrne, N. Alhashimy, K. J. Fraser, D. R.
MacFarlane, D. Diamond, Phys. Chem. Chem. Phys. 2010, 12,
7009.
163 C. Hardacre, J. D. Holbrey, S. P. Katdare, K. R. Seddon,
Green Chem. 2002, 4, 143.
164 S. Tiwari, A. Kumar, J. Phys. Chem. A 2012, 116, 1191.
165 K. R. Harris, M. Kanakubo, L. A. Woolf, J. Chem. Eng.
Data 2007, 52, 1080.
166 E. I. Izgorodina, R. Maganti, V. Armel, P. M. Dean, J. M.
Pringle, K. R. Seddon, D. R. MacFarlane, J. Phys. Chem. B 2011,
115, 14688.
24 M. D. Elola, B. M. Ladanyi, J. Phys. Chem. B 2006, 110,
5525.
25 S. M. Vechi, M. S. Skaf, J. Chem. Phys. 2005, 123, 154507.
26 H. Torii, M. Tasumi, J. Phys. Chem. A 2000, 104, 4174.
27 Z. Hu, X. Huang, H. V. R. Annapureddy, C. J. Margulis,
J. Phys. Chem. B 2008, 112, 7837.
28 T. Ishida, K. Nishikawa, H. Shirota, J. Phys. Chem. B 2009,
13, 9840.
1
1
1
1
1
29 T. Ishida, H. Shirota, J. Phys. Chem. B 2013, 117, 1136.
30 T. Kato, H. Shirota, J. Chem. Phys. 2011, 134, 164504.
31 H. Shirota, T. Fukuda, T. Kato, J. Phys. Chem. B 2013, 117,
1
6196.
32 D. R. Lide, CRC Handbook of Chemistry and Physics, 89th
ed., ed. by D. R. Lide, CRC Press, Boca Raton, 2008.
1
1
1
33 D. Song, J. Chen, J. Chem. Eng. Data 2014, 59, 257.
34 W. Martino, J. Fernandez de la Mora, Y. Yoshida, G. Saito,
J. Wilkes, Green Chem. 2006, 8, 390.
35 G. B. Appetecchi, M. Montanino, M. Carewska, M.
Moreno, F. Alessandrini, S. Passerini, Electrochim. Acta 2011,
6, 1300.
36 T. Koller, M. H. Rausch, P. S. Schulz, M. Berger, P.
1
5
1
Wasserscheid, I. G. Economou, A. Leipertz, A. P. Fröba, J. Chem.
Eng. Data 2012, 57, 828.
167 Y.-H. Yu, A. N. Soriano, M.-H. Li, J. Chem. Thermodyn.
2009, 41, 103.
1
37 Q.-S. Liu, J. Tong, Z.-C. Tan, U. Welz-Biermann, J.-Z.
Yang, J. Chem. Eng. Data 2010, 55, 2586.
38 E. Gómez, B. González, N. Calvar, E. Tojo, A. Domínguez,
J. Chem. Eng. Data 2006, 51, 2096.
39 G. V. S. M. Carrera, C. A. M. Afonso, L. C. Branco,
J. Chem. Eng. Data 2010, 55, 609.
40 S. Katsuta, Y. Shiozawa, K. Imai, Y. Kudo, Y. Takeda,
168 N. V. Ignat’ev, U. Welz-Biermann, A. Kucheryna, G.
Bissky, H. Willner, J. Fluorine Chem. 2005, 126, 1150.
169 J. Vila, P. Ginés, J. M. Pico, C. Franjo, E. Jiménez, L. M.
Varela, O. Cabeza, Fluid Phase Equilib. 2006, 242, 141.
170 S. Bulut, P. Eiden, W. Beichel, J. M. Slattery, T. F.
Beyersdorff, T. J. S. Schubert, I. Krossing, ChemPhysChem 2011,
12, 2296.
1
1
1
J. Chem. Eng. Data 2010, 55, 1588.
171 T. Nishida, Y. Tashiro, M. Yamamoto, J. Fluorine Chem.
Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan | 1127

2
1
003, 120, 135.
72 J. Leys, R. N. Rajesh, P. C. Menon, C. Glorieux, S.
188 P. D. Haller, L. C. Bradley, M. Gupta, Langmuir 2013, 29,
11640.
Longuemart, P. Nockemann, M. Pellens, K. Binnemans, J. Chem.
Phys. 2010, 133, 034503.
189 M. García-Mardones, P. Cea, I. Gascón, C. Lafuente,
J. Chem. Thermodyn. 2014, 78, 234.
1
73 O. Zech, A. Stoppa, R. Buchner, W. Kunz, J. Chem. Eng.
Data 2010, 55, 1774.
74 J. Leys, M. Wübbenhorst, C. P. Menon, R. Rajesh, J.
190 C. Kolbeck, J. Lehmann, K. R. J. Lovelock, T. Cremer, N.
Paape, P. Wasserscheid, A. P. Fröba, F. Maier, H.-P. Steinrück,
J. Phys. Chem. B 2010, 114, 17025.
1
Thoen, C. Glorieux, P. Nockemann, B. Thijs, K. Binnemans, S.
Longuemart, J. Chem. Phys. 2008, 128, 064509.
191 H. Jin, B. O’Hare, J. Dong, S. Arzhantsev, G. A. Baker,
J. F. Wishart, A. J. Benesi, M. Maroncelli, J. Phys. Chem. B 2008,
112, 81.
192 S. Kakinuma, H. Shirota, J. Phys. Chem. B 2015, 119,
4713.
1
75 I. Bandrés, D. F. Montaño, I. Gascón, P. Cea, C. Lafuente,
Electrochim. Acta 2010, 55, 2252.
76 T. Tu, X. Bao, W. Assenmacher, H. Peterlik, J. Daniels,
1
K. H. Dötz, Chem.®Eur. J. 2009, 15, 1853.
193 P. Cong, H. P. Deuel, J. D. Simon, Chem. Phys. Lett. 1995,
240, 72.
1
77 H. Tokuda, K. Hayamizu, K. Ishii, Md. A. B. H. Susan, M.
Watanabe, J. Phys. Chem. B 2004, 108, 16593.
78 W. Xu, E. I. Cooper, C. A. Angell, J. Phys. Chem. B 2003,
07, 6170.
79 H. Tokuda, K. Hayamizu, K. Ishii, M. Susan, M. Watanabe,
J. Phys. Chem. B 2005, 109, 6103.
194 Y. J. Chang, E. W. Castner, Jr., J. Phys. Chem. 1996, 100,
3330.
195 Q. Zhong, J. T. Fourkas, J. Phys. Chem. B 2008, 112, 8656.
196 N. T. Hunt, S. R. Meech, Chem. Phys. Lett. 2003, 378, 195.
197 N. T. Hunt, A. A. Jaye, A. Hellman, S. R. Meech, J. Phys.
Chem. B 2004, 108, 100.
1
1
1
1
1
80 H. Vogel, Phys. Z. 1921, 22, 645.
81 G. Tammann, W. Hesse, Z. Anorg. Allg. Chem. 1926, 156,
198 Y. Marcus, Ionic Properties, Marcel Dekker, New York,
1997.
2
45.
1
1
1
82 G. S. Fulcher, J. Am. Ceram. Soc. 1925, 8, 339.
83 G. W. Scherer, J. Am. Ceram. Soc. 1992, 75, 1060.
84 P. J. Carvalho, M. G. Freire, I. M. Marrucho, A. J.
199 H. Shirota, E. W. Castner, Jr., J. Am. Chem. Soc. 2001, 123,
12877.
200 H. Shirota, E. W. Castner, Jr., J. Chem. Phys. 2006, 125,
034904.
201 H. Shirota, H. Ushiyama, J. Phys. Chem. B 2008, 112,
13542.
202 H. Okajima, H. Hamaguchi, Chem. Lett. 2011, 40, 1308.
203 H. Shirota, T. Kato, J. Phys. Chem. A 2011, 115, 8797.
204 A. Triolo, O. Russina, B. Fazio, G. B. Appetecchi, M.
Carewska, S. Passerini, J. Chem. Phys. 2009, 130, 164521.
Queimada, J. A. P. Coutinho, J. Chem. Eng. Data 2008, 53, 1346.
85 D. Garoz, C. Bueno, C. Larriba, S. Castro, I. Romero-Sanz,
J. F. de la Mora, Y. Yoshida, G. Saito, J. Appl. Phys. 2007, 102,
64913.
86 P. J. Carvalho, C. M. S. S. Neves, J. A. P. Coutinho,
J. Chem. Eng. Data 2010, 55, 3807.
87 J. Klomfar, M. Součková, J. Pátek, J. Chem. Eng. Data
009, 54, 1389.
1
0
1
1
2
1128 | Bull. Chem. Soc. Jpn. 2016, 89, 1106–1128 | doi:10.1246/bcsj.20160085
© 2016 The Chemical Society of Japan