Dynamic solvation in room-temperature ionic liquids

Article abstract of DOI:10.1021/jp0376828

The dynamic solvation of the fluorescent probe, coumarin 153, is measured in five room-temperature ionic liquids using different experimental techniques and methods of data analysis. With time-resolved stimulated-emission and time-correlated single-photon counting techniques, it is found that the solvation is comprised of an initial rapid component of a??55 ps. In all the solvents, half or more of the solvation is completed within 100 ps. The remainder of the solvation occurs on a much longer time scale. The emission spectra of coumarin 153 are nearly superimposable at all temperatures in a given solvent unless they are obtained using the supercooled liquid, suggesting that the solvents have an essentially glassy nature. The physical origin of the two components is discussed in terms of the polarizability of the organic cation for the faster one and the relative diffusional motion of the cations and the anions for the slower one. A comparison of the solvation response functions obtained from single-wavelength and from spectral-reconstruction measurements is provided. Preliminary fluorescence-upconversion measurements are presented against which the appropriateness of the single-wavelength method for constructing solvation correlation functions and the use of stimulated-emission measurements is considered. These measurements are consistent with the trends mentioned above, but a comparison indicates that the presence of one or more excited states distorts the stimulated-emission kinetics such that they do not perfectly reproduce the spontaneous emission data. Fluorescence-upconversion results indicate an initial solvation component on the order of a??7 ps.

Full text of DOI:10.1021/jp0376828

J. Phys. Chem. B 2004, 108, 10245-10255
10245
Dynamic Solvation in Room-Temperature Ionic Liquids†
P. K. Chowdhury, M. Halder, L. Sanders, T. Calhoun, J. L. Anderson, D. W. Armstrong,
X. Song, and J. W. Petrich*
Department of Chemistry, Iowa State UniVersity, Ames, Iowa 50011
ReceiVed: December 2, 2003; In Final Form: April 20, 2004
The dynamic solvation of the fluorescent probe, coumarin 153, is measured in five room-temperature ionic
liquids using different experimental techniques and methods of data analysis. With time-resolved stimulated-
emission and time-correlated single-photon counting techniques, it is found that the solvation is comprised of
an initial rapid component of ∼55 ps. In all the solvents, half or more of the solvation is completed within
1
1
00 ps. The remainder of the solvation occurs on a much longer time scale. The emission spectra of coumarin
53 are nearly superimposable at all temperatures in a given solvent unless they are obtained using the
supercooled liquid, suggesting that the solvents have an essentially glassy nature. The physical origin of the
two components is discussed in terms of the polarizability of the organic cation for the faster one and the
relative diffusional motion of the cations and the anions for the slower one. A comparison of the solvation
response functions obtained from single-wavelength and from spectral-reconstruction measurements is provided.
Preliminary fluorescence-upconversion measurements are presented against which the appropriateness of the
single-wavelength method for constructing solvation correlation functions and the use of stimulated-emission
measurements is considered. These measurements are consistent with the trends mentioned above, but a
comparison indicates that the presence of one or more excited states distorts the stimulated-emission kinetics
such that they do not perfectly reproduce the spontaneous emission data. Fluorescence-upconversion results
indicate an initial solvation component on the order of ∼7 ps.
Introduction
is based on a linear free-energy relationship. The solute retention
factor is determined chromatographically for a set of probe
Room-temperature ionic liquids (RTILs) are becoming an
increasingly rich area of study.1,2 They have been used as novel
solutes, and a multiple linear regression analysis relates it to a
set of five parameters that characterize the solvent.2
3
-16
solvent systems for organic synthesis,
for liquid-liquid
extraction,17,18
19
Solvent properties can dramatically influence the rate of
chemical reactions. Assuming equilibrium solvation along the
reaction coordinate, many of these effects can indeed be
explained by classical thermodynamics, as the construction of
the above linear free-energy relationships attempts. A thermo-
dynamic approach is valid, however, only if the motion of the
solvent molecules is very fast compared with the motion along
the reaction coordinate, so that the solvent is always in
equilibrium with the solute. In particular, the solvent relaxation
time scales influence the dynamics of electron transfer, proton
transfer, and other charge-transfer reactions by exerting a time-
dependent dielectric friction. In these cases, reaction rates may
in electrochemical studies, and as ultralow
volatility liquid matrixes for matrix-assisted laser desorption/
ionization mass spectrometry.20 RTILs have other properties that
make their application in chemical systems attractive. Some are
immiscible with water and nonpolar organic solvents. They are
stable to temperatures in excess of 300 °C, but they have
negligible vapor pressures, thus making them “green” solvents
by reducing environmental levels of volatile organic carbons.
Their viscosities can easily be varied by changing their cationic
or anionic constituents.
Most ionic liquids are said to have similar polarities, close
to those of short-chain alcohols.
2
1-24
The solvatochromic effect
2
9,30
2
1
22
be limited by the rate of solvent relaxation.
A well-
of Reichardt’s dye and Nile Red as well as of fluorescent
2
3,24
established method to measure time-dependent solvation is by
means of time-resolved fluorescence spectroscopy. After short-
pulse excitation, the fluorescence spectrum of a probe solute
red shifts in time as the surrounding solvent re-equilibrates to
the new, excited-state charge distribution. This time-dependent
fluorescence shift provides a direct measure of the kinetics of
solvation occurring at the microscopic level relevant to chemical
probes
and the Rohrschneider-McReynolds gas-liquid
25
chromatography (GLC) method have been used to characterize
ionic liquids by obtaining a general, one-dimensional polarity-
based parameter. This approach has not been successful for
RTILs because they all fall within the same narrow range of
values.21-24
Yet, two different ionic liquids that have essentially
identical “polarity,” as measured by such methods, can produce
very different results when used as solvents for organic reactions,
GLC, or extractions. Most recently, we have used the solvation
parameter model developed by Abraham, which has been
2
9
reactions.
Before the availability of RTILs, transient solvation by ions
was studied by Huppert and co-workers using molten salts3
and by Maroncelli and co-workers using nonaqueous solutions
1,32
developed to characterize either liquid- or gas-phase interactions
33
between solute molecules and liquid phases.26-28 This model
of dissolved ions. The dynamic aspects of solvation by RTILs
are becoming an object of experimental and theoretical studies.
The dynamic solvation and the reorientational behavior of
several probe molecules in different ionic liquids have recently
†
Part of the special issue “Gerald Small Festschrift”.
*
To whom correspondence may be addressed. E-mail: jwp@iastate.edu.
1
0.1021/jp0376828 CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/02/2004

10246 J. Phys. Chem. B, Vol. 108, No. 29, 2004
Chowdhury et al.
+
-
+
-
+
-
+
-]), formed
Figure 1. The solvation probe, coumarin 153, and 4 RTILs, ([BMIM ][Cl ], [BMIM ][BF
4
], [BMIM ][PF
2
6
], and [BMIM ][NTf
2
-
-
from the 1-butyl-3-methylimidazolium cation and the four anions indicated ([NTf
2
] is (CF
+
3
][TFA ]); butylimidazole and methylimidazole.
3
SO
2
)
N ); one RTIL formed from the triethylammonium
-
cation and trifluoroacetate ([NHEt
been investigated.34-39 These workers have shown that the
solvation dynamics is biphasic. A molecular dynamics simula-
tion by Shim et al.40 has been interpreted in terms of the faster
component corresponding to diffusional motion of the anion
and the slower component corresponding to collective motion
correlated single-photon counting and the early events by
transient absorption (stimulated emission), the latter of which
employed a particular method of constructing the solvation
correlation function using a single probe wavelength. Although
these methods produced results obeying reasonable trends, it
became apparent that neither measurement of stimulated-
emission nor single-wavelength analysis was adequate. The third
part of the discussion provides a comparison with results from
upconversion measurements that offer a more traditional and
direct measure of spontaneous emission. Salient points are
summarized in the conclusions section.
41
of the anion and the cation. In addition, Hyun et al. and Giraud
et al.42 have used the optical Kerr effect to study the low-
-
1
frequency vibrational motions (<200 cm ) of these solvents.
In this article, we address the dynamic aspects of solvation
by RTILs using Stokes shift data of the fluorescent probe,
coumarin 153 (Figure 1), from the sub-picosecond to the
nanosecond time regimes. We investigate four RTILs based on
+
the 1-butyl-3-methylimidazolium cation, BMIM , and either the
Materials and Methods
-
-
-
-
Cl , BF4 , PF6 , or (CF3SO2)2N anions (Figure 1). They are
+
-
+
-
+
-
typically referred to as [BMIM ][Cl ], [BMIM ][BF4 ],
BMIM ][PF6 ], and [BMIM ][NTf2 ], respectively. We com-
RTILs. 1-Butyl-3-methylimidazolium chloride [BMIM ][Cl ]
is produced by refluxing equimolar amounts of 1-methylimi-
dazole with 1-chlorobutane at 70 °C for 72 h. The resulting
+
-
+
-
[
pare these RTILs with the organic subunits, butylimidazole and
+
-
+
-
methylimidazole. A fifth ionic liquid studied is [NHEt3 ][TFA ],
triethylammonium trifluoroacetate. Major conclusions are that
the rapid initial phase of solvation arises from the cation and
that the simulations of Shim et al.40 may not provide an
appropriate comparison for large organic fluorescent solvation
probes, such as coumarin 153. In the course of this work,
comparisons were made between different methods of con-
structing the solvation correlation function and of obtaining the
solvation data itself. Consequently, this work addresses aspects
of methodology nearly as much as the problem of solvation by
RTILs.
[BMIM ][Cl ] is washed with ethyl acetate and dried under
vacuum. 1-Butyl-3-methylimidazolium tetrafluoroborate [BM-
+
-
+
-
IM ][BF4 ] is produced by mixing 10 g of [BMIM ][Cl ]
(0.057 mol) in ∼100 mL of acetone. Sodium tetrafluoroborate
(6.29 g (0.057 mol)) is added and stirred for 24 h. The resulting
sodium chloride is filtered off and the acetone removed under
vacuum. 1-Butyl-3-methylimidazolium hexafluorophosphate
+
-
[BMIM ][PF6 ] is produced by dissolving 164.49 g (0.94 mol)
+
-
of [BMIM ][Cl ] in ∼250 mL of water and reacting an
equimolar amount (0.94 mol) of hexafluorophosphoric acid
+
-
(HPF6) and stirring for 24 h. The resulting [BMIM ][PF6 ] is
washed with water until the washings are no longer acidic. The
IL is then dried under vacuum. Hexafluorophosphoric acid is
corrosive, toxic, and should be handled with care. 1-Butyl-3-
methylimidazolium bis[(trifluoromethyl)-sulfonyl]imide, [BM-
The plan of the article is as follows. After a description of
the various experimental and data-treatment methods, we present
the results. It became apparent during the course of the
investigation that they depend sometimes rather significantly
on the type of experiment or data treatment used. The discussion
section is separated into three parts. The first is a short
consideration of the steady-state spectra of the RTILs. The
second deals with the early stages of this work, where the long-
time behavior (up to 4 ns) of RTILs was studied by time-
+
-
IM ][NTf2 ], is produced by dissolving 34.0 g (0.118 mol) of
N-lithiotrifluoromethanesulfonimide in ∼150 mL of water and
+
-
mixing with 20.67 g (0.118 mol) of [BMIM ][Cl ] also
dissolved in ∼150 mL water. The mixture is stirred for 12 h.
The aqueous portion is removed and the resulting ionic liquid

Dynamic Solvation in Room-Temperature Ionic Liquids
J. Phys. Chem. B, Vol. 108, No. 29, 2004 10247
washed with water and then dried under vacuum. All RTILs
infinity to zero instead of from zero to positive infinity to avoid
interference from transitions to higher-lying excited states.
Taking these limits of integration is permitted as long as there
is mirror image symmetry between the emission and excitation
spectra. Time-resolved spectra are treated similarly, but the
emission data first must be continued down to zero intensity,
since it is impractical to obtain time-resolved and spectrally
resolved data out into the wings of the emission spectrum. This
continuation is done simply by connecting the last two points
of the available data with straight lines that extend to zero. These
new curves consisting of the data points and the straight lines
are interpolated in the same way as the steady-state excitation
curve. The value of λ is sensitive to the way in which the
spectrum is continued. We have used both straight-line and log-
normal continuations but have opted for the simplest. The errors
introduced in the continuation are effectively canceled when
time-dependent λs are compared with each other (as opposed
to the steady-state spectrum) as they are in Table 1, which
furnishes λ at 100 and 4000 ps. The rest of the process proceeds
as above. All data manipulations were performed with Microcal
Origin 7.0.
Time-Resolved Measurements. The laser source for the
time-correlated single-photon counting measurements was a
homemade mode-locked Ti-sapphire laser, tunable from 780
to 900 nm with a repetition rate of 82 MHz. The fundamental
from the Ti-sapphire oscillator was modulated by a Pockels
cell (Model 350-160, Conoptics Inc) to reduce the repetition
rate to about 8.8 MHz and was subsequently frequency doubled
by focusing tightly into a 0.4-mm BBO crystal. The resulting
blue light, which had a central wavelength of 425 nm, provided
the excitation source. The fluorescence decays were collected
at the magic angle (polarization of 54.7° with respect to the
vertical). Emission was collected through a single monochro-
mator (ISA H10) fitted with a slit having an 8-nm band pass.
A half-wave plate before a vertical polarizer ensured the
polarization of the excitation light. The instrument-response
function of the apparatus had a full-width-at-half-maximum
+
-
produced using the [BMIM ][Cl ] salt are subjected to the silver
nitrate test to ensure no chloride impurities remain in the
+
-
samples. [NHEt3 ][TFA ] is prepared by adding a slight molar
excess of trifluoroacetic acid dropwise to triethylamine in a 50-
mL round-bottom flask. The mixture is then heated at 40 °C
and stirred for 2 h. Excess trifluoroacetic acid is removed under
vacuum and the remaining ionic liquid dried through a silica
gel column and stored under P2O5.
Spectroscopic Measurements and Data Analysis. Butyl-
imidazole (>98% purity) and methylimidazole (>99% purity)
were obtained from Aldrich (St. Louis, MO) and dried over
molecular sieves (type 4A) before use. During spectroscopic
measurements, the quartz cuvettes were kept tightly sealed so
as to prevent moisture from being absorbed by the ionic liquids.
The temperature-dependent measurements were carried out in
a thermoelectric temperature-controlled cuvette holder (Quantum
Northwest, WA), permitting regulation over the range -40 to
1
00 °C. All ionic liquids were carefully dried before optical
measurements were performed. Silica gel of 60-200 mesh is
activated at 150 °C. A small silica gel column is prepared using
a Pasteur pipet, and the pure ionic liquid is added to the column.
Depending on the viscosity of the ionic liquid, a bulb is used
to force the ionic liquid through the silica gel column. Once
the ionic liquid emerges from the column, it is stored in a
desiccator under P2O5 until further use. The RTILs were
subsequently dried over molecular sieves: Type 4A, GRADE
5
14, 8-12 Mesh, and with effective pore size of 4 Å.
Steady-State Measurements. Steady-state excitation and
emission spectra were recorded with a SPEX Fluoromax with
a 4-nm band pass and were corrected for detector response. A
1-cm path length quartz cuvette was used for the measurements.
The steady-state spectra can be used to compute the reorganiza-
tion energy, λ, by means of43
∞
∫
dυ[σ
a
(υ) - σ
f
(υ)]υ
(υ)]
0
λ ) p
(1)
∞
(fwhm) of 80 ps. A cuvette of 1-cm path length was used for
∫
dυ[σ
a
(υ) + σ
f
0
the time-resolved measurements of C153 in the different
solvents. To construct the time-resolved spectra, a series of
decays (∼3000 counts in the peak channel) were collected over
as much of the fluorescence spectrum as possible and were fit
to a maximum of three exponentials, yielding fits with ø-squared
values of ∼1. Transient spectra were reconstructed from these
fits by normalizing to the steady-state spectra of the samples
The σa,f are the absorption (or excitation) and emission spectra,
respectively, on a wavenumber scale. The reorganization energy
is widely used as a measure of the strength of interactions
between a chromophore and its surrounding dielectric media
in solvation dynamics studies. It is usually taken as half of the
Stokes shift. This estimation is accurate if the excitation and
emission spectra are Gaussian, but it becomes unreliable if they
are not. The actual computation of λ is accomplished by first
manipulating the emission and excitation spectra to permit their
addition and subtraction. This requires normalized spectra
consisting of equally spaced points. We interpolate and renor-
2
9
according to the equation
S0(λ)
S(λ,t) ) D(λ,t)
(2)
∞
∫
D(λ,t) dt
0
-
1
malize them so as to obtain spectra having 20-cm spacing
between each point and then shift the crossing point of the two
D(λ,t) is the wavelength-resolved fluorescence decay; S0(λ) is
the steady-state emission intensity at a given wavelength. We
have employed the traditional approach of fitting the time-
resolved spectra to a log-normal line-shape function, from which
we extract the peak frequency, ν(t), as a function of time. The
solvation dynamics were described by the normalized correlation
-
1
curves so that it lies at 0 cm . The spectral baselines are then
corrected by subtracting the lowest intensity and renormalizing.
This manipulation is motivated by the low-intensity emission
near 800 nm and questions concerning the utility of the
correction factors of our fluorimeter in this region. In any case,
baseline subtraction is minor and changes the final result by
approximately 1%. An appropriate number of zeros is added to
the high-energy end of the emission spectrum and to the low-
energy end of the excitation spectrum so that the curves can be
added and subtracted along their entire breadths. λ may now
be calculated according to eq 1. In practice, however, the
integration is more conveniently performed from negative
2
9
function as follows
ν(t) - ν(∞)
ν(0) - ν(∞)
C(t) )
(3)
ν(0) is the frequency at zero time, that is, immediately after
excitation. ν(∞) is the frequency at “infinite time,” the maximum
of the steady state fluorescence spectrum. The decays used to

10248 J. Phys. Chem. B, Vol. 108, No. 29, 2004
Chowdhury et al.
TABLE 1: Spectral Characteristics of Coumarin 153 in Some RTILs
〈ν〉absd
〈ν〉emd
λe
λ
f,g
100ps
(cm-1)
λ
f,g
4000ps
〈τ〉SRh
(ns)
〈τ〉SWi
(ns)
melting
point (°C)
η
(cP)
ionic liquida
b
(cm-1)
(cm-1)
(cm-1)
(cm-1
)
+
-
c
[
[
[
BMIM ][Cl ] (30 °C)
65
65
-8
11000
334
371
23490
23540
24080
1.4 ( 0.3
18570
18090
18420
2294
2582
2592
1940
2285
2015
2112
2627
2313
2.0
0.53
1.0
1.7
0.33
0.38
+
-
c
c
b
BMIM ][Cl ] (70 °C)
+
-
BMIM ][PF
6
] (20 °C)
j
3
1
3
6
12
k
.8 ( 0.4
l
.35 ( 0.02
s
.47 ( 0.20
+
-] (20 °C)
-] (20 °C)
c
[
BMIM ][BF
4
-82
-4
154
23670
18350
18530
2481
2607
1968
1965
2203
2180
0.46
0.72
0.31
0.17
b
m
2
2
33
1.44
n
.13
+
b
[
BMIM ][NTf
2
52
24110
0
0
0
0
.28 ( 0.04°
p
.48 ( 0.07
q
.56 ( 0.08
r
.38 ( 0.06
butylimidazole (20 °C)
methylimidazole (20 °C)
< 50
< 50
26100
24040
24000
19450
18700
18330
2836
2622
2672
2451
2263
2484
2764
2344
2480
0.42
0.070
0.050
0.27
0.041
0.025
+
-
[
NHEt ][TFA ] (20 °C)
3
a
+
-
57
For abbreviations, see the caption to Figure 1. Although the melting point of [BMIM ][Cl ] is ∼65 °C, it usually takes several days at room
temperature before it solidifies. From ref 57. From ref 52. d 〈V〉 ) (∫ νI(ν) dν/∫ I(ν) dν), computed using 70% of the emission and excitation
b
c
∞
0
∞
0
spectra in order to exclude the contributions from absorption to states higher than S
1
in energy. e Computed from eq 1. See text and ref 43. The
f
reorganization energies at 100 and 4000 ps are computed as discussed in the Experimental Section. Note that, because of the limited data set used
to obtain the time-resolved spectra, the λ at 4000 ps deviate from the steady-state values, for which in most cases we would expect them to be
g
approximately the same. The fractional amount of solvation at 100 ps, for example, can be determined by f100ps ) (λ100ps - λ
i
/λ4000ps - λ
i
), where
by proposing
Average solvation times obtained from the spectral-reconstruction
λ
i
is the intramolecular contribution to the reorganization energy. Maroncelli and co-workers have addressed the estimation of λ
methods to compute the “zero-time” spectrum arising only from solvation.
i
58
h
data, computed from the fit parameters given in the caption to Figure 4. The comparison of the average solvation times is made using only data
i
accumulated from time-correlated single-photon counting data. Errors in the solvation times are ∼(15%. Average solvation times obtained from
j
the single-wavelength data. The average time is computed from the spliced picosecond and nanosecond experiments. At 25 °C using
3
7
k
37
l
38
m
4
-aminophthalimide (4-AP).
dimethylaminonaphthalene (Prodan).
is computed from parameters obtained from a stretched exponential fit.
At 25 °C using coumarin 102.
At 25 °C using coumarin 153.
At 20 °C using 6-propionyl-2-
3
5
n
34 o
At room temperature using coumarin 153.
At 25 °C using coumarin 153. The average solvation time
3
9
p
36
q
36 r
At 20 °C using coumarin 153.
At 20 °C using Prodan.
At 20 °C
3
6
s
53
using 4-AP.
At 298 K using Prodan.
7-49
construct the time-resolved emission spectra were typically
collected over a range of wavelengths from 470 to 610 nm at
intervals of 10 nm; unless otherwise indicated, a total of 15
transients were used to construct the time-resolved emission
spectra, from which the C(t) values were obtained.
wavelength solvent-correlation function, CSW(t).4
As a result,
we measured stimulated-emission kinetics at 480 and 560 nm
and typically collected over a full scale of 100 ps.
To have a picture of the solvation over the entire time scale,
starting from “time zero” (as defined by the 130-fs pulses of
the transient-absorption apparatus) to about 4 ns, the longest
time scale we investigated with our photon-counting system,
we spliced the data from the two experiments together and
constructed the single-wavelength solvation response function
at the probe wavelength of 480 nm. The decay curves from the
stimulated emission and the photon-counting measurements were
first fit to multiexponential functions. The data points for the
first 100 ps were taken from the fitted stimulated-emission
kinetics and from 100 ps to 4 ns, from the fitted photon-counting
experiments. The spliced curves were normalized according to
Transient-Absorption (Stimulated-Emission) Measure-
ments. The instrument function of our time-correlated single-
photon counting system has a fwhm of 80 ps. To investigate
more rapid phenomena, better time resolution was required. This
was provided by a homemade regeneratively amplified Ti-
4
4,45
sapphire system,
providing 130-fs pulses, which we used
to perform pump-probe transient stimulated-emission measure-
+
-
ments. Samples were excited at 407 nm. ([BMIM ][Cl ] was
not investigated with this system because our rotating sample
holder is not temperature controlled and because the solvent
slowly solidifies at room temperature.)
4
7
the following equation
The use of stimulated emission to measure solvation requires
special conditions, notably that excited-state (and ground-state)
absorption does not contaminate the signal. Such a situation
should not be assumed to be the norm. Ernsting and co-
∞
tot
F′(ν,t) ) F(ν,t) exp{+K t}
(4)
∞
tot
where F(ν,t) is the spliced emission curve and exp{+K t} is
46
workers have studied the time-dependent spectra of coumarin
53. Their results suggested that the spectral regions near 480
and 560 nm are free of absorbing species. This was fortuitous
the population factor. For our normalization, we used the inverse
of the average fluorescence lifetime (1/〈τF〉), obtained from
measuring the fluorescence decay while collecting light from
most of the emission band, to approximate Ktot. An average
lifetime is employed since the fluorescence decay is not well
described by a single exponential. Once the population factor
was taken into account, the normalized single-wavelength
solvation-response curve, CSW(t), was constructed
1
47
because Gardecki and Maroncelli indicated that 480 and 560
nm could be used for coumarin 153 to construct single-
wavelength solvation correlation functions. Such a situation
seemed to be optimal experimentally since laboratories are more
typically equipped to perform transient absorption than fluo-
rescence-upconversion experiments and because a single-
wavelength analysis is obviously not as time-consuming as one
requiring data collection over a range of wavelengths. Several
research groups have discussed and employed the single-
F′(νSW,t) - F′(νSW,∞)
CSW(t) )
(5)
F′(νSW,0) - F′(νSW,∞)

Dynamic Solvation in Room-Temperature Ionic Liquids
J. Phys. Chem. B, Vol. 108, No. 29, 2004 10249
Figure 3. Representative wavelength-resolved decays for C153
+
-
[
BMIM ][NTf ] at 490 and 600 nm. A typical instrument function
2
profile is included. The trace at 600 nm reveals the growth of a solvent-
relaxed state (which is absent in the profile at 490 nm). The decays
used to construct the time-resolved emission spectra were typically
collected over a range of wavelengths from 470 to 610 nm at intervals
of 10 nm; a total of fifteen time-resolved fluorescence traces were used
to construct the time-resolved emission spectra, from which the C(t)
values were obtained.
SR-400 gated photon counter operated in CW mode with a
threshold level of -100 mV. This signal was fed into a boxcar
averager. A part of the blue pulse train was used to normalize
pump-beam fluctuations. A translation stage (Compumotor) with
a resolution of 0.06 mm/step was used to delay the exciting
pulses and a computer with an interfacing card from Keithley
Metrabyte (DAS 800) was used for driving the motor. The
instrument-response function was obtained by collecting the
cross-correlation function of the blue and red pulses; the
resulting third harmonic intensity was plotted against delay time.
The cross-correlation functions typically have a fwhm of ∼1
ps. This instrument response is a little over 3 times as broad as
Figure 2. Excitation and emission spectra of coumarin 153 in
+
-
-1
[
2
[
(
BMIM ][PF
6
] at 90 °C (solid) λ ) 2662 cm , 25 °C (dotted) λ )
-
1
-1
592 cm , and -10 °C supercooled (dashed) λ ) 2272 cm ;
+
-
-1
BMIM ][Cl ] at 70 °C (solid) λ ) 2582 cm and 30 °C supercooled
1
-
+
-
2
dashed) λ ) 2294 cm ; [BMIM ][NTf ] at 90 °C (solid) λ ) 2649
50,51
-1
-1
that obtained with our unamplified system.
We attribute this
cm , 25 °C (dotted) λ ) 2607 cm , and -10 °C supercooled (dashed)
-
1
λ ) 2466 cm . All samples were excited at 420 nm. For excitation
to the absence of compensating prisms after frequency doubling,
the presence of the rotating sample cell, and perhaps a nonideal
optical geometry, which nevertheless permits the facile inter-
change between pump-probe transient absorption and fluores-
cence-upconversion measurements. All curves were fit and
deconvoluted from the instrument function using an iterative
convolute and compare least-squares algorithm.
spectra, the emission monochromator was at 600 nm.
Fluorescence-Upconversion Measurements. The system
used for these measurement is that of the amplified Ti-sapphire
referred to above for the stimulated-emission experiments. The
fundamental output from the amplifier (815 nm) is doubled by
a type-I LBO crystal (2 mm). The frequency-doubled blue pulses
(407 nm) are separated from the fundamental by a dielectric
Results
mirror coated for 400 nm and are focused onto a rotating cell
containing the sample using a 5-cm convex lens. The remaining
fundamental was used as the gate to upconvert emission.
Fluorescence was collected by an LMH-10x microscopic
objective (OFR Precision Optical Products) coated for near UV
transmission. The gate and the emission are focused by a quartz
lens (12 cm) onto a type-I 0.4-mm BBO crystal (MgF2 coated,
cut at 31°, and mounted by Quantum Technology, Inc). The
polarization of both the gate and excitation source was controlled
with a set of zero-order half-wave plates for 800 and 400 nm,
respectively. The upconverted signal is then directed into an
H10 (8 nm/mm) monochromator (Jobin Yvon/Spex Instruments
S. A. Group) with a 5-cm convex lens coupled to a Hamamatsu
R 980 PMT equipped with a UG11 UV-pass filter and operated
at maximum sensitivity. The PMT output was amplified in two
stages (total by a factor of 25, 5 for each stage) by a Stanford
research Systems SR-445 DC-300 MHz amplifier with input
terminated at 500 Ω and was carefully calibrated after a long
Table 1 summarizes most of the results obtained in this study.
(Upconversion results will be discussed separately.) Figure 2
presents excitation and emission spectra of coumarin 153 in
the three ionic liquids for which we were able to obtain spectra
above and below their melting points (while the ionic liquids
still remained liquid). The significant feature is that the emission
spectra of coumarin 153 are superimposable at all temperatures
in a given solvent unless the spectra are obtained using the
supercooled liquid.
Figure 3 presents two typical wavelength-resolved decay
+
-
profiles of coumarin 153 in [BMIM ][NTf2 ]. Figure 4 gives
the solvation correlation functions, CSR(t), constructed from the
time-correlated single-photon counting data using the spectral
reconstruction method.
The single-wavelength solvent-response functions, CSW(t),
obtained at 480 nm from combining the stimulated-emission
and the photon-counting results are shown in Figure 5. These
results are direct measurements of the early rapid solvation
(1-2 h) warm-up. Photon arrival events were registered with

10250 J. Phys. Chem. B, Vol. 108, No. 29, 2004
Chowdhury et al.
+
-
Figure 4. CSR(t) curves and corresponding fits for C153 in (A) [BMIM ][NTf
2
] (O) [C(t) ) 0.72 exp(-t/350 ps) + 0.28 exp(-t/1670 ps)];
+
-
+
-
4
[
BMIM ][PF ] (b) [C(t) ) 0.36 exp(-t/360 ps) + 0.64 exp(-t/1360 ps)]; [BMIM ][BF ] (2) [C(t) ) 0.62 exp(-t/60 ps) + 0.38 exp(-t/1100
6
+
-
+
-
ps)]; [NHEt
0.76 exp(-t/2530 ps)]; [BMIM ][Cl ] at 70 °C (b) [C(t) ) 0.63 exp(-t/120 ps) + 0.37 exp(-t/1290 ps)]; [BMIM ][NTf
2) [C(t) ) exp(-t/70 ps)].
3
][TFA ] (0) [C(t) ) 0.58 exp(-t/120 ps) + 0.42 exp(-t/940 ps)] and (B) [BMIM ][Cl ] at 30 °C (9) [C(t) ) 0.24 exp(-t/270 ps)
+
-
+
-
2
+
(
] (O); butylimidazole
+
-
+
-
process in these ionic liquids, at least 50% of which is completed
in the first 100 ps. In general, this method reports a fast solvation
component of 40-70 ps, which is considerably longer than the
the cases of [BMIM ][Cl ] at 70 °C, [BMIM ][BF4 ], and
+
-
butylimidazole. It is rather poor in the cases of [BMIM ][PF6 ]
+
-
and [BMIM ][NTf2 ] (not shown). The agreement in general
appears poorer upon comparison of the average solvation times
in Table 1 than when visually inspecting the form of the
correlation functions. This is because the curves fail to
superimpose in the long-time region, where even low-amplitude
slow dynamics can contribute considerably to the average. Even
in this long-time region, however, the form of the two correlation
functions may remain similar, as can be seen by comparing the
37,39
<
5 ps suggested by Maroncelli and co-workers
based on
their estimation of the resolution of their time-correlated single-
photon counting apparatus. The significance and accuracy of
our result will be discussed in more detail below.
A comparison of the solvation correlation functions obtained
from the single-wavelength and the spectral-reconstruction
methods is given in Figure 6 and in Table 1 in order to provide
a check on the appropriateness of using the stimulated-emission
data to measure the initial rapid phase of solvation. The
agreement between the two methods is often quite good, as in
+
-
+
-
data for [BMIM ][Cl ] at 70 °C and [BMIM ][BF4 ].
Fluorescence-upconversion measurements provide an inde-
pendent gauge of the time scale for this initial rapid process.

Dynamic Solvation in Room-Temperature Ionic Liquids
J. Phys. Chem. B, Vol. 108, No. 29, 2004 10251
Whereas stimulated-emission (i.e., transient-absorption) mea-
surements are sensitive to both emitting and absorbing species,
fluorescence measurements report only on spontaneous emission
and thus provide a more straightforward means of probing the
solvation behavior of interest. Upconversion data are provided
+
-
in Figures 7-9 for [BMIM ][PF6 ], butylimidazole, and
methanol. For each of the solvents, the transients obtained using
the two methods are compared at 480 and 560 nm, the
wavelengths suggested for use in single-wavelength analysis
47
for coumarin 153. For each solvent, there are deviations arising
from excited-state absorbance at these wavelengths, which
conspire to distort the time scale for solvation.
Discussion
Steady-State Spectra. Steady-state spectra taken as a function
of temperature are presented in Figure 2 for three RTILs. Spectra
obtained at temperatures above the melting point are nearly
superimposable. These three solvents can be supercooled easily.
Once the temperature is lowered below the melting point,
Figure 5. CSW(t) for C153 obtained at 480 nm at room temperature.
The plots are stimulated-emission kinetics obtained over 100 ps with
1
30-fs pulses and, where necessary, time-correlated single-photon
+
-
crystallization occurs slowly. In the case of [BMIM ][Cl ], it
can occur over a period of several days. Upon supercooling, a
significant change in the spectra is observed, which is manifest
essentially in their position. There is less solvent relaxation,
which we have quantified by means of the reorganization
counting data obtained over 4 ns with an 80-ps instrument function.
The two traces were spliced together at 100 ps in order to present a
continuous trace of the solvation dynamics. Global fits to the CSW(t)
are summarized in Table 2. The CSW(t) for methanol is presented as a
“
control” experiment, yielding an average solvation time of 3.5 ps. That
+
the fast solvation components of the BMIM liquids resemble that of
butylimidazole and that the fast component is seen to become more
rapid as the solvent is changed to methylimidazole or to methanol
suggests that the ∼55-ps event in the ionic liquids is a plausible result.
See, however, the correlation functions obtained from fluorescence-
upconversion data in Figures 7, 8, and 10.
+
-
energy, λ (Table 1 and Figure 2). In the case of [BMIM ][Cl ],
-
1
-1
λ decreases from 2582 cm at 70 °C to 2294 cm for the
supercooled liquid at 30 °C. This shift suggests that the solvent
+
-
is essentially “glassy” in the supercooled state. For [BMIM ][Cl ],
supercooling increases the average solvation time by about a
Figure 6. A comparison of the solvation correlation functions obtained from the single-wavelength (solid lines), CSW480nm(t), and the spectral-
reconstruction (dashed lines), CSR(t), methods. All data were obtained from time-correlated single-photon counting experiments. A complete comparison
+
-
+
-
+
-
6
is given in Tables 1 and 2. The agreement ranges from very good for [BMIM ][Cl ] at 70 °C and [BMIM ][BF ] to poor for [BMIM ][PF ].
4

10252 J. Phys. Chem. B, Vol. 108, No. 29, 2004
Chowdhury et al.
+
-] at different wavelengths. ∆A(t) is the kinetic trace for the pump-
Figure 7. Normalized upconversion traces, I(t), for C153 in [BMIM ][PF
6
probe stimulated-emission experiment at the corresponding wavelength. All upconversion traces are fitted globally with time constants of 5, 210,
and 5100 ps.
Figure 8. Normalized upconversion traces, I(t), for C153 in butylimidazole at different wavelengths. ∆A(t) is the kinetic trace for the pump-probe
stimulated-emission experiment at the corresponding wavelength. All upconversion traces are fitted globally with the following time constants: 28,
1
02, and 4700 ps.
factor of 4. An interesting aspect of the spectra is that the change
in reorganization energy between the normal (90 °C) and the
Stimulated-Emission, Time-Correlated Single-Photon
Counting, and Single-Wavelength Analysis. Inspection of
Table 1 indicates that there is a considerable disparity in the
values of the average solvation times obtained from different
laboratories using the spectral reconstruction method. Reason-
+
-
supercooledliquids(-10°C)ismuchsmallerfor[BMIM ][NTf2 ]
+
-
+
-
than for [BMIM ][Cl ] and [BMIM ][PF6 ]: ∆λ ) 183, 288,
and 390 cm , respectively (see the caption to Figure 2). That
the change is the smallest for [BMIM ][NTf2 ] may be
attributed to it having the lowest viscosity of the three solvents;
upon supercooling, it is immobilized to a lesser degree than
-
1
+
-
+
-
able agreement is obtained between our result for [BMIM ][PF6 ]
and those results of Maroncelli and co-workers using 4-AP and
3
7
coumarin 102. There is also reasonable agreement between
+
-
+
-
+
-
[
BMIM ][Cl ] or [BMIM ][PF6 ]. We were unable to perform
our result for [BMIM ][NTf2 ] and those of Karmakar and
-
36
the same temperature-dependent analysis with the BF4 solvent
Samanta using coumarin 153 and Prodan. Some of the
because it melts at -82 °C and our apparatus did not permit us
disparity may be attributed to the number of counts collected
in the fluorescence decay, which affects the signal-to-noise ratio,
and to the full-scale time window used, which determines how
accurately the long-time component is determined. In our case,
we use ∼4 ns, which is smaller than that used by other groups.
3
7,39
to obtain such a temperature. Maroncelli and co-workers
have recently shown that ionic liquids conform to standard glass
behavior by fitting their kinetic data to stretched exponentials
and by showing that it varies linearly with viscosity.

Dynamic Solvation in Room-Temperature Ionic Liquids
J. Phys. Chem. B, Vol. 108, No. 29, 2004 10253
Figure 9. Normalized upconversion traces, I(t), for C153 in MeOH at 480 and 560 nm along with the stimulated-emission traces, ∆A(t), obtained
at the same wavelengths.
TABLE 2: Global Fitting of the Initial Rapid Phase of
A more significant origin of the discrepancies may arise from
C
SW(t) of Some RTILsa
the preparation and purity of the ionic liquids themselves. Trace
amounts of water and chloride impurities are known to have
ionic liquid
η (cP)b
a
1
τ
1
(ps)
2
τ (ps)
52
large effects on the viscosities and densities of RTILs. Notably,
[BMIM
+
][BF4
-
] (20 °C)
15452
233
0.39
55
1000
53
57
Bright and co-workers report that the average solvation time
[NHEt3+][TFA ] (20 °C)
-
0.61
0.62
55
55
650
+
-
of Prodan in [BMIM ][PF6 ] decreases from 6.5 to 3.9 ns as
the water content is increased from <50 ppm to 1.8 wt %. Since
the focus of this article is not to obtain a quantitatiVe comparison
of the long-time behavior of the RTILs, these differences do
not cause excessive concern. They should, however, give an
indication to future workers of the spectrum of results that have
been obtained with an eye to ensuring reproducibility when
examining the glassiness of the RTILs.
+
-
52
[
BMIM ][PF
6
] (20 °C)
371
2250
57
3
52
12
+
-] (20 °C)
57
[
BMIM ][NTf
2
0.73
0.98
0.80
0.35
55
55
8.5
0.32
1500
120
butylimidazole (20 °C)
methylimidazole (20 °C)
methanol (20 °C)
< 50
< 50
190
5.2
0.59
a
Csw(t) obtained from 480-nm stimulated-emission data as described
in the text are fit to a sum of exponentials, C (t) ) a exp(-t/τ ) + a
sw
1
1
2
Of more immediate interest in this work is the initial rapid
exp(-t/τ
2
), where a
1
+ a ) 1. In all cases (except the last two solvents),
2
the rapid-solvation component is fixed at 55 ps. Given the 100-ps full
phase of solvation in the ionic liquids. Several previous
+
-
studies34-39
scale used in most cases (for [NHEt ][TFA ], a 400-ps scale was used),
3
have suggested, and our results indicate (Figures 4
the longer-lived components should not be regarded as accurate
and 5), that more than 50% of solvation relaxation is very rapid.
The origin of these fast solvent fluctuations is not clear. We
have hypothesized that the fast relaxations result from the
organic cations. To test this idea, we attempted to compare the
solvation dynamics of butylimidazole with those of the four
ionic liquids bearing the BMIM cation. A comparison of
estimations of the slower solvation phase. The Csw(t) are displayed in
Figure 5. b We have no viscosity information for [NHEt +
-
3
][TFA ] and
have not been able to obtain any for butylimidazole or methylimidazole
from either the literature or the supplier.
by a rapid component of 8.5 ps. The time constant and average
solvation time for methanol, as measured by fluorescence
+
-
[
NHEt3 ][TFA ] with triethylamine was, unfortunately, not
5
5
possible because triethylamine quenches excited-state coumarin
techniques, are 2.3 and 5.0 ps, respectively. Our CSW(t) for
coumarin 153 in methanol (Figure 5) obtained from stimulated-
emission kinetics is described by an average solvation time of
3.5 ps. This suggests that ionic liquids based on the methylimi-
dazolium cation will have much faster initial solvation com-
ponents owing to the shortening of the aliphatic chain from butyl
to methyl, just as the solvation by the alcohols is faster for
methanol than for butanol. A more direct test of this hypothesis
will require measurements of the solvation times of different
RTILs at the same viscosity as well as a comparison with ionic
liquids based on other cations.
54
by electron transfer. We attempted other comparisons with
RTILs formed from butylpyridinium and pyridinium organic
-
-
cations with BF4 and NTf2 . These also presented complica-
tions in the excited-state kinetics that prevented a further
comparison of solvation dynamics of the RTILs with their
organic counterparts.
The organic solvent, butylimidazole, has a solvation time of
between 40 and 70 ps (Figures 4 and 5, Tables 1 and 2).
Butylimidazole has a very low viscosity, and amplitudes of the
fast components of the solvation times for the ionic liquid
counterparts seem to scale with viscosity, as indicated by Figure
On the other hand, another interpretation of the origin of the
initial fast component has resulted from a recent computer
5
and Table 2. The latter presents the results of a global fitting
4
0
analysis to the CSW(t) functions (except where indicated). In
this procedure, the CSW(t) were fit to two decaying exponential
components where the smallest time constant was fixed to agree
roughly with that of butylimidazole, 55 ps. The amplitudes of
the two components and the time constant of the longer were
permitted to vary. A 55-ps component was present in all the
fits. Such a value seems reasonable given that the time constant
and the average solvation time for butanol, obtained by
simulation. Shim et al. place a dipolar excitation on a model
diatomic solute in a RTIL. They observe fast initial relaxation
apparently arising from translations of the anions. Given the
small size of the model solute and the large changes in net
charge, the motions of the resulting concentrated anionic charges
may dominate the relaxation. In the experimental solvation
dynamics studies of organic dye molecules such as coumarin
153 in RTILs, the charge distribution change is scattered across
the whole probe molecule, which is most likely well solvated
by the cations given the hydrophobic nature of C153 in the
ground state. We suggest, therefore, that the most possible
5
5
fluorescence measurements, are 47 and 63 ps, respectively.
Another point of comparison is provided by the CSW(t) of
methylimidazole (Figure 5 and Table 2), which is dominated

10254 J. Phys. Chem. B, Vol. 108, No. 29, 2004
Chowdhury et al.
studies is corroborated. A more comprehensive comparison is,
however, required and is currently being undertaken.
Conclusions
Our results are generally consistent with those of previous
3
4-39
workers,
but there are significant discrepancies in the
average solvation times reported for some ionic liquids, most
likely arising from the determination of the longer-lived
dynamics. Solvation times obtained from the spectral-resolution
and the single-wavelength methods are compared. Our stimulated-
emission experiments with sub-picosecond time resolution are
the first to probe directly the initial rapid solvation component
that had been suggested in the earlier studies. They yield a time
constant in the range of 40-70 ps (Figure 5, Table 2).
Nevertheless, fluorescence-upconversion measurements indicate
that the presence of excited-state absorption increases this time
by approximately a factor of 8 from its value as obtained by
monitoring spontaneous emission directly. Consequently, while
stimulated-emission measurements can be useful in indicating
general trends, direct measurements of spontaneous emission
and use of spectral reconstruction methods are required for
quantitative work. These results lead one to inquire into the
nature and the number of the excited states contributing to this
absorption since their presence could profoundly affect the
interpretation of solvation dynamics data. Finally, a comparison
of the solvation times in the 1-butyl-3-methylimidazolium ionic
liquids with that of butylimidazole itself (and methylimidazole)
leads us to consider the role of the polarizability of the cationic
partner in giving rise to the initial rapid solvation component.
+
-]
6
Figure 10. Solvation correlation functions, CSR(t), for [BMIM ][PF
solid line) and butylimidazole (dashed line). The correlation functions
(
+
-
6
], 0.14 exp-
are well described by the following forms: [BMIM ][PF
-t/7 ps) + 0.86 exp(-t/160 ps); butylimidazole, 0.09 exp(-t/7 ps) +
.91 exp(-t/72 ps). It is likely that the relative amplitude of the short
(
0
component in these correlation functions is reduced with respect to
those obtained from stimulated-emission measurements owing to the
poorer time resolution of ∼1 ps.
scenario is that the cation motions dominate the contribution
of the initial fast relaxation upon excitation. To elucidate the
nature of the fast relaxations, a realistic model of C153 in RTILs
will be needed. Znamenskiy and Kobrak have recently per-
formed molecular dynamics simulations of the dye betaine-30
in [BMIM ][PF6 ]. They obtain radial distribution functions
indicating the proximity of the 1-butyl-3-methylimidazolium ring
to the dye, which is consistent with our experimental observa-
tions for coumarin.
Analysis Based upon Fluorescence-Upconversion Results.
Finally, we wish to comment on the relative merits of using
stimulated-emission measurements, single-wavelength construc-
tion of C(t), and direct fluorescence measurements. The
comparison provided here, in particular the data presented in
Figures 6-9, indicate that while the stimulated-emission
measurements provide a good qualitative picture of the dynam-
ics, they deviate considerably from the fluorescence-upconver-
sion measurements. We have already noted the discrepancies
in the sub-nanosecond and nanosecond time regimes using
single-wavelength and “complete” spectral data from photon-
counting measurements (Figure 6). The fluorescence-upcon-
version measurements permit us to make this comparison in
the picosecond time regime. Examination of the traces at 480
and 560 nm indicates that there is at least one absorbing species
that contributes to the stimulated-emission signal. In the absence
of absorption, the two techniques should give identical results.
+
- 56
Acknowledgment. D.W.A. was supported by NIH Grant
RO1 GM53825-08. We thank Dr. Ranjan Das for technical
assistance in collecting the fluorescence-upconversion data.
Edward Castner and Mark Maroncelli provided stimulating
comments. We thank the latter for sharing his Kerr-gated
fluorescence data with us before publication.
References and Notes
(
1) Seddon, K. R. Nature (Materials) 2003, 2, 363.
(2) Anderson, J. L.; Ding, J.; Welton, T.; Armstrong, D. W. J. Am.
Chem. Soc 2002, 124, 14247.
3) Wilkes, J. S.; Zaworotko, M. J. J. Chem. Soc., Chem. Commun.
1992, 965.
(
(
4) Adams, C. J.; Earle, M. J.; Roberts, G.; Seddon, K. R. Chem.
Commun. 1998, 2097.
(5) Earle, M. J.; McCormac, P. B.; Seddon, K. R. Chem. Commun.
(Cambridge) 1998, 2245.
(6) Dyson, P. J.; Ellis, D. J.; Parker, D. G.; Welton, T. Chem. Commun.
Cambridge) 1999, 25.
(
(
(
7) Leadbeater, N. E.; Torenius, H. M. J. Org. Chem. 2002, 67, 3145.
8) Mann, B. E.; Guzman, M. H. Inorg. Chim. Acta 2002, 330, 143.
+
-
But in fact, for [BMIM ][PF6 ] and butylimidazole, excited-
state absorption causes the stimulated-emission kinetics at 480
and 560 nm to decay and rise, respectively, slower than those
of their spontaneous-emission counterparts. This translates into
an overestimation of the time scale for solvation by a factor of
(9) Wasserscheid, P.; Keim, W. Angew. Chem., Int. Ed. 2000, 39, 3772.
10) Reynolds, J. L.; Erdner, K. R.; Jones, P. B. Org. Lett. 2002, 4,
17.
(
9
(
11) Welton, T. Chem. ReV. 1999, 99, 2071.
(12) Nara, S. J.; Harjani, J. R.; Salunkhe, M. M. Tetrahedron Lett. 2002,
43, 2979.
(
(
13) Yao, Q. Org. Lett. 2002, 4, 2197.
14) Fletcher, K. A.; Pandey, S.; Storey, I. K.; Hendricks, A. E.; Pandey,
∼
8. Deviations are also apparent for methanol. In particular,
while the fluorescence-upconversion trace in methanol decays
essentially to zero, the stimulated-emission trace levels off at
about 40 ps to a steady-state value of about 20% of the initial
signal intensity, again indicating the presence of one or more
absorbing states.
S. Anal. Chim. Acta 2002, 453, 89.
(15) Grodkowski, J.; Neta, P. J. Phys. Chem. A 2002, 106, 5468.
(16) Handy, S. T.; Zhang, X. Org. Lett. 2001, 3, 233.
(
(
17) Huddleston, J. G.; Rogers, R. D. Chem. Commun. 1998, 1765.
18) Dai, S.; Ju, Y. H.; Barnes, C. E. J. Chem. Soc., Dalton Trans. 1999,
1
201.
(19) Dickinson, V. E.; Williams, M. E.; Hendrickson, S. M.; Masui,
H.; Murray, R. W. J. Am. Chem. Soc. 1999, 121, 613.
20) Armstrong, D. W.; Zhang, L.-K.; He, L.; Gross, M. L. Anal. Chem.
001, 73, 3679.
21) Muldoon, M. J.; Gordon, C. M.; Dunkin, I. R. J. Chem. Soc., Perkin
Trans. 2 2001, 433.
+
-
Figure 10 presents the CSR(t) for [BMIM ][PF6 ] and
butylimidazole obtained from the data in Figures 7 and 8,
respectively. These solvation correlation functions can both be
fit to the same initial rapid component of 7 ps, and as such, the
role of the organic cation suggested by the stimulated-emission
(
2
(

Dynamic Solvation in Room-Temperature Ionic Liquids
J. Phys. Chem. B, Vol. 108, No. 29, 2004 10255
(
22) Carmichael, A. J.; Seddon, K. R. J. Phys. Org. Chem. 2000, 13,
91.
23) Bonhote, P.; Dias, A.-P.; Papageorgiou, N.; Kalyanasundaram, K.;
Graetzel, M. Inorg. Chem. 1996, 35, 1168.
24) Aki, S. N. V. K.; Brennecke, J. F.; Samanta, A. Chem. Commun
Cambridge) 2001, 413.
(42) Giraud, G.; Gordon, C. M.; Dunkin, I. R.; Wynne, K. J. Chem.
Phys. 2003, 119, 464.
(43) Jordanides, X. J.; Lang, M. J.; Song, X. Y.; Fleming, G. R. J. Phys.
Chem. B 1999, 103, 7995.
(44) English, D. S.; Zhang, W.; Kraus, G. A.; Petrich, J. W. J. Am.
Chem. Soc. 1997, 119, 2980.
5
(
(
(
(
(
25) Armstrong, D. W.; He, L.; Liu, Y.-S. Anal. Chem. 1999, 71, 3873.
26) Abraham, M. H.; Whiting, G. S.; Doherty, R. M.; Shuely, W. J. J.
(45) English, D. S.; Das, K.; Zenner, J. M.; Zhang, W.; Kraus, G. A.;
Larock, R. C.; Petrich, J. W. J. Phys. Chem. A 1997, 101, 3235.
(46) Kovalenko, S. A.; Ruthmann, J.; Ernsting, N. P. Chem. Phys. Lett.
1997, 271, 40.
(47) Gardecki, J. A.; Maroncelli, M. J. Phys. Chem. A 1999, 103, 1187.
(48) Nagarajan, V.; Brearley, A. M.; Kang, T.-J.; Barbara, P. F. J. Chem.
Phys. 1987, 86, 3183.
Chromatogr., A 1990, 518, 329.
27) Abraham, M. H.; Whiting, G. S.; Andonian-Haftvan, J.; Steed, J.
W. J. Chromatogr., A 1991, 588, 361.
28) Abraham, M. H.; Whiting, G. S.; Doherty, R. M.; Shuely, W. J. J.
Chromatogr., A 1991, 587, 213.
(
(
(
29) Maroncelli, M.; Fleming, G. R. J. Chem. Phys. 1987, 86, 6221.
(49) Changenet-Barret, P.; Choma, C. T.; Gooding, E. F.; DeGrado, W.
F.; Hochstrasser, R. M. J. Phys. Chem. B 2000, 104, 9322.
(50) Das, K.; Smirnov, A. V.; Wen, J.; Miskovsky, P.; Petrich, J. W.
Photochem. Photobiol. 1999, 69, 633.
(51) Smirnov, A. V.; Das, K.; English, D. S.; Wan, Z.; Kraus, G. A.;
Petrich, J. W. J. Phys. Chem. A 1999, 103, 7949.
(52) Seddon, K. R.; Stark, A.; Torres, M.-J. Viscosity and Density of
1-Alkyl-3-methylimidazolium Ionic Liquids. In Clean SolVents: AlternatiVe
Media for Chemical Reactions and Processing; ACS Symposium Series;
American Chemical Society, Washington, DC, 2002; Vol. 819, p 34.
(53) Baker, S. N.; Baker, G. A.; Munson, C. A.; Chen, F.; Bukowski,
E. J.; Cartwright, A. N.; Bright, F. V. Ind. Eng. Chem. Res. 2003, 42, 6457.
(54) Castner, E. W., Jr.; Kennedy, D.; Cave, R. J. J. Phys. Chem. A
2000, 104, 2869.
(30) Vajda, S.; Jimenez, R.; Rosenthal, S. J.; Fidler, V.; Fleming, G.
R.; Castner, E. W. J. Chem. Soc., Faraday Trans. 1995, 91, 867.
31) Huppert, D.; Ittah, V.; Kosower, E. M. Chem. Phys. Lett. 1989,
59, 267.
(
1
(
(
(
(
(
(
32) Ittah, V.; Huppert, D. Chem. Phys. Lett. 1990, 173, 496.
33) Chapman, C. F.; Maroncelli, M. J. Phys. Chem. 1991, 95, 9095.
34) Karmakar, R.; Samanta, A. J. Phys. Chem. A 2002, 106, 4447.
35) Karmakar, R.; Samanta, A. J. Phys. Chem. A 2002, 106, 6670.
36) Karmakar, B.; Samanta, A. J. Phys. Chem. A 2003, 107, 7340.
37) Ingram, J. A.; Moog, R. S.; Ito, N.; Biswas, R.; Maroncelli, M. J.
Phys. Chem. B 2003, 107, 5926.
38) Chakrabarty, D.; Hazra, P.; Chakraborty, A.; Seth, D.; Sarkar, N.
Chem. Phys. Lett. 2003, 381, 697.
39) Arzhantsev, S.; Ito, N.; Heitz, M.; Maroncelli, M. Chem. Phys. Lett.
003, 381, 278.
40) Shim, Y.; Duan, J.; Choi, M. Y.; Kim, H. J. J. Chem. Phys. 2003,
19, 6411.
41) Hyun, B.-R.; Dzyuba, S. V.; Bartsch, R. A.; Quitevis, E. L. J. Phys.
Chem. A 2002, 106, 7579.
(
(
(55) Horng, M. L.; Gardecki, J.; Papazyan, A.; Maroncelli, M. J. Phys.
Chem. 1995, 99, 17311.
(56) Znamenskiy, V.; Kobrak, M. N. J. Phys. Chem. B 2004, 108, 1072.
(57) Carda-Broch, S.; Berthod, A.; Armstrong, D. W. Anal. Bioanal.
Chem. 2003, 375, 191.
2
(
1
(
(58) Fee, R. S.; Maroncelli, M. Chem. Phys. 1994, 183, 235.