Solvatochromic probe behavior within ternary room-temperature ionic liquid 1-butyl-3-methylimidazolium hexafluorophosphate + ethanol + water solutions

Article abstract of DOI:10.1021/jp0276754

The solvatochromic probe behavior within ternary BMIMPF₆ + ethanol + water systems were investigated using four solvatochromic probes: pyrene, 1,3-bis(1-pyrenyl)propane, 1-pyrenecarboxaldehyde, and Reichardt's betaine dye (BMIMPF₆ [1-butyl-3-methylimidazolium hexafluorophosphate] is one of the most popular room-temperature ionic liquids [RTIL]). A simplified preferential solvation model based on the weighted mole fraction probe response shows that the pyrene cybotactic region is rich in BMIMPF₆ compared to the bulk. However, the 1,3-bis(1-pyrenyl)propane and 1-pyrenecarboxaldehyde solvation environments appear to be ethanol enriched. Although the aldehyde functional moiety on 1-pyrenecarboxaldehyde may explain an ethanol-rich cybotactic region, a significant decrease in bulk viscosity (more than that predicted from additivity) of BMIMPF₆ upon aqueous-ethanol addition may result in an increased intramolecular excimer formation efficiency of 1,3-bis(1-pyrenyl)propane. E_T(30) values in ternary BMIMPF₆ + ethanol + water solutions are higher than those observed in neat BMIMPF₆ or aqueous ethanol, suggesting a local solvation environment rich in water. The results strongly suggest that the physicochemical properties of this RTIL can be modulated in a controlled fashion by adding appropriate amounts of aqueous ethanol.

Full text of DOI:10.1021/jp0276754

13532
J. Phys. Chem. B 2003, 107, 13532-13539
Solvatochromic Probe Behavior within Ternary Room-Temperature Ionic Liquid
1-Butyl-3-methylimidazolium Hexafluorophosphate + Ethanol + Water Solutions
Kristin A. Fletcher and Siddharth Pandey*
Department of Chemistry, New Mexico Institute of Mining and Technology, Socorro, New Mexico 87801
ReceiVed: December 10, 2002; In Final Form: September 19, 2003
The solvatochromic probe behavior within ternary BMIMPF6 + ethanol + water systems were investigated
using four solvatochromic probes: pyrene, 1,3-bis(1-pyrenyl)propane, 1-pyrenecarboxaldehyde, and Reichardt’s
betaine dye (BMIMPF6 [1-butyl-3-methylimidazolium hexafluorophosphate] is one of the most popular room-
temperature ionic liquids [RTIL]). A simplified preferential solvation model based on the weighted mole
fraction probe response shows that the pyrene cybotactic region is rich in BMIMPF6 compared to the bulk.
However, the 1,3-bis(1-pyrenyl)propane and 1-pyrenecarboxaldehyde solvation environments appear to be
ethanol enriched. Although the aldehyde functional moiety on 1-pyrenecarboxaldehyde may explain an ethanol-
rich cybotactic region, a significant decrease in bulk viscosity (more than that predicted from additivity) of
BMIMPF6 upon aqueous-ethanol addition may result in an increased intramolecular excimer formation
efficiency of 1,3-bis(1-pyrenyl)propane. ET(30) values in ternary BMIMPF6 + ethanol + water solutions are
higher than those observed in neat BMIMPF6 or aqueous ethanol, suggesting a local solvation environment
rich in water. The results strongly suggest that the physicochemical properties of this RTIL can be modulated
in a controlled fashion by adding appropriate amounts of aqueous ethanol.
Introduction
removing the water impurity is a major concern.19-26 Another
factor that can impose limitations on the use of the binary
BMIMPF6 + water system in many applications is the fact that
the solubility of water in BMIMPF6 is very low.19-26 Ethanol
is the next choice as a “green” cosolvent that can be effectively
used to “tune” the physicochemical properties of BMIMPF6 and
other RTILs.19-21 However, although higher than water, the
solubility of ethanol in BMIMPF6 is also shown to be rather
limited. Surprisingly, a binary mixture of ethanol + water was
observed to be completely miscible in BMIMPF6 in the
composition range Xethanol ∼ 0.5-0.9.19-21 This suggests that
an ethanol-water mixture of appropriate composition can be
used as a “green” cosolvent to effectively modify and tailor
BMIMPF6 physicochemical properties for specific analytical
applications. To better understand any effect added ethanol-
water mixtures may have on BMIMPF6, we have investigated
the behavior of several solvatochromic probes having different
functionalities when solubilized in the ternary BMIMPF6 +
ethanol + water system.
Studying solute-solvent and solvent-solvent interactions by
means of solvatochromic probes is both simple and convenient.
Solvatochromic probe studies offer direct information on solvent
properties such as dipolarity, hydrogen bond donating/accepting
capabilities, etc., as experienced by the probe in its cybotactic
region (a cybotactic region may be defined as the volume around
a solute molecule in which the ordering of the solvent molecules
has been influenced by the solute, including both the first
solvation shell and the transition region).27 The study of
physicochemical properties that depend on solute-solvent and
solvent-solvent interactions is much more complex in mixed
solvent systems than in pure solvents.28 On one hand, the solute
can be preferentially solvated by any of the solvents present in
the mixture; on the other, solvent-solvent interactions can
strongly affect solute-solvent interactions. Preferential solvation
may arise whenever the bulk mole fraction solvent composition
One of the biggest problems posed to the chemical industry
is to continuously deal with the fact that all chemical plants
rely heavily on toxic, hazardous, and flammable organic
solvents. Newly rediscovered (in their modified forms) room-
temperature ionic liquids (RTILs), with no measurable vapor
pressure, can be used as replacements for select organic solvents.
RTILs are organic salts composed of anions and cations that
are in the liquid state at ambient conditions. The new-generation
RTILs have the potential to act as environmentally benign
solvent media for many industrially important chemical
processes.1-3 Recently, these novel RTILs have shown promise
toward important applications such as synthesis, catalysis,
polymerization, separation, and extraction processes.1-14 Some
of the recent investigations of RTILs as novel solvent systems
have been a part of this Journal.15-18
To expand the utility of RTILs and better tune their
physicochemical properties, researchers have started to focus
on RTIL-based mixed solvent systems. A part of the effort
toward this end has been in the area where interesting RTILs
are combined with other “green” solvents to tailor the physi-
cochemical properties of the RTIL of interest in a favorable
fashion.19-23 Often, to increase the efficiency of a process (e.g.,
separation, extraction, synthesis, etc.), one wishes to “tune” a
solvent or solvent mixture by addition of cosolvents. It is
beneficial in many ways to understand how added cosolvents
(or impurities) affect the physical properties of RTILs.
In our previous investigations, we examined the affect of
added water and ethanol, respectively,23 on the behavior of a
variety of solutes when dissolved in a popular RTIL 1-butyl-
3-methylimidazolium hexafluorophosphate (BMIMPF6). RTIL
BMIMPF6, however, is known to be hygroscopic in nature and
* Corresponding author. Phone: (505) 835-6032. Fax: (505) 835-5364.
E-mail: pandey@nmt.edu.
10.1021/jp0276754 CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/08/2003

Probe Behavior in BMIMPF6 + Ethanol + Water Solutions
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13533
Figure 1. Chemical structures and the corresponding solvatochromic responses of the probes used in this study: Py, pyrene; 1-PyCHO,
1-pyrenecarboxaldehyde; 1,3-BPP, 1,3-bis(1-pyrenyl)propane; ET(30), Reichardt’s dye. Solid and broken lines represent probe responses in neat
BMIMPF6 and ethanol, respectively.
is different from the solvation microsphere solvent composition.
The response of spectroscopic probes is dependent upon the
composition of the solvation microsphere and therefore provides
a convenient means to measure the extent of preferential
solvation.
Experimental Section
Chemicals and Reagents. Pyrene (99%) was obtained from
AccuStandard, Inc., and 1-pyrenecarboxaldehyde (99%), Rei-
chardt’s dye, and 4-nitroaniline were obtained from Aldrich
Chemical Co. 1,3-Bis(1-pyrenyl)propane was purchased from
Molecular Probes, Inc., and N,N-diethyl-4-nitroaniline was
purchased from Frinton Laboratories. All materials were of the
highest purity available and were used as received. Stock
solutions were prepared in ethanol and stored in precleaned
amber glass vials at ∼4 °C. BMIMPF6 (electrochemical grade,
99.99+%) was stored under argon and used as received from
Covalent Associates. Karl-Fisher titrations showed no detectable
presence of water in freshly purchased BMIMPF6. Alternatively,
BMIMCl can be prepared by reaction of equimolar amounts of
1-methylimidazole and chlorobutane in a round-bottomed flask
fitted with a reflux condenser by heating and stirring at 70 °C
for 48-72 h. The resulting viscous liquid was cooled to room
temperature followed by three washings with ethyl acetate. The
remaining ethyl acetate was removed by heating to 70 °C under
vacuum. To prepare BMIMPF6, hexafluorophosphoric acid was
slowly added to a mixture of BMIMCl in water. After stirring
for ∼12 h, the upper acidic aqueous layer was decanted and
the lower ionic liquid layer was washed with water several times
until the washings were no longer acidic. Thus obtained
BMIMPF6 was then heated under vacuum at 70 °C to remove
any excess water.20,21 Spectroscopic grade high-purity ethanol
was obtained from Fisher Scientific and used as received.
Doubly distilled deionized water was obtained from a Millipore,
Milli-Q Academic water purification system.
It is important to mention that many research groups have
used a variety of solvatochromic probes to gain information on
the cybotactic region that these probes encounter when solu-
bilized in different RTILs. In one of our earlier studies, we used
pyrene, Reichardt’s betaine dye, 1-pyrenecarboxaldehyde, Nile
Red, and dansylamide to probe neat BMIMPF6 at ambient
conditions.29 Muldoon et al. have also utilized Reichardt’s dye
along with [Cu(acac)(tmen)][X] (acac ) acetylacetone, tmen
) N,N,N′,N′-tetramethylethylenediamine, X ) [BPh4]G or
[ClO4]G) to investigate the solute-solvent interactions in many
of the neat 1,3-dialkylimidazolium cation-based RTILs.25 Car-
michael and Seddon have performed polarity studies of some
neat 1-alkyl-3-methylimidazolium-based RTILs with the sol-
vatochromic dye Nile Red.26 Recently, Baker et al. have
investigated the effects of temperature and added carbon dioxide
on the cybotactic region surrounding three fluorescent probes,
pyrene, PRODAN (6-propionyl-2-(N,N-dimethylamino)naph-
thalene), and BTBP (N,N′-bis(2,5-di-tert-butylphenyl)-3,4,9,-
10-perylenedicarboximide), dissolved in neat BMIMPF6.30 In
another study, the same group has reported on the temperature-
dependent microscopic solvent properties of dry and wet
BMIMPF6 using ET(30) and Kamlet-Taft solvent polarity
scales.31 Most recently, these authors have used fluorescence
correlation spectroscopy to determine the influence of water on
the diffusion coefficients of several fluorescent probes dissolved
within BMIMPF6.32
Methods. Samples for spectroscopic studies were prepared
as follows: appropriate aliquots of solvatochromic probe stock
solutions were transferred into 1 cm2 quartz cuvettes and
evaporated under argon. Ethanol-water mixtures of appropriate
compositions were prepared separately, and required amounts
were added along with the appropriate amounts of the fresh
sample of RTIL BMIMPF6 directly to the cuvette, mixed
thoroughly, and allowed to equilibrate for sufficient time in a
In this paper, we present the behavior of four different
solvatochromic probes dissolved in different compositions of
the ternary BMIMPF6 + ethanol + water solutions: pyrene,
1-pyrenecarboxaldehyde, 1,3-bis(1-pyrenyl)propane, and Rei-
chardt’s betaine dye (see Figure 1). It is important to note that
all compositions studied are in the complete miscibility region
of the ternary phase diagram.19-21

13534 J. Phys. Chem. B, Vol. 107, No. 48, 2003
Fletcher and Pandey
Figure 2. Pyrene II/IIII (10 µM) in ternary BMIMPF6 + ethanol + water: (A) BMIMPF6 + 0.55 mole fraction ethanol in water; (B) BMIMPF6
+ 0.65 mole fraction ethanol in water; (C) BMIMPF6 + 0.75 mole fraction ethanol in water; (D) BMIMPF6 + 0.85 mole fraction ethanol in water,
respectively. Key: (open circles, dotted lines) experimentally measured values; (filled circles, solid lines) predicted additive values.
moisture-free environment. All solutions were prepared on an
analytical balance with (0.1 mg accuracy. Steady-state emission
experiments were performed with a PTI QuantaMaster Model
C-60/2000 L-format scanning spectrofluorometer with a 75 W
xenon arc lamp as the excitation source and single-grating
monochromators as wavelength selection devices. All emission
spectra were corrected for emission monochromator response
and were background subtracted using appropriate blanks.
Absorption spectra were recorded on an Agilent Hewlett-
Packard 8453 photodiode array spectrophotometer.
fraction average of the probe’s spectral response in pure solvents,
R0i , is given by36-39
Radditive
)
X R0
(1)
∑
i
i
i
In the case of preferential solvation, however, the measured (or
observed) mole fraction (Yi) composition may differ substantially
from the bulk. Acree and co-workers have shown that in the
case of pyrene, II/IIII in a ternary solution of components A-C,
becomes40
Results and Discussion
II
[YAII,A + YBII,B + YCII,C]
Pyrene. Pyrene is one of the most widely used neutral
fluorescence probes.33-35 The pyrene solvent polarity scale is
defined as the II/IIII emission intensity ratio, where band I
corresponds to a S1(V ) 0) f S0(V ) 0) transition and band III
is a S1(V ) 0) f S0(V ) 1) transition.35 II/IIII increases with
increasing solvent polarity. It is important to mention here that
we observed a pyrene II/IIII in neat BMIMPF6 that was
)
(2)
IIII
[YAIIII,A + YBIIII,B + YCIIII,C
]
Pyrene band I and III emission intensities are measured in the
three pure solvents under identical conditions, and pyrene II/IIII
values for an additive ternary mixture are calculated using the
bulk mole fractions, Xi, for each component. Data depicted as
solid lines in each panel of Figure 2 represent this calculated
probe behavior. A careful examination of Figure 2 reveals that
(1) Experimental pyrene II/IIII values deviate significantly from
predicted additive values. (2) Although experimental pyrene II/
IIII values in aqueous ethanol are lower than predicted (the
significantly higher than that measured in neat ethanol (II/IIII
)
1.37 ( 0.01 in ethanol).29 The only solvents with higher pyrene
II/IIII than observed in BMIMPF6 were acetonitrile, dimethyl
sulfoxide, and neat water, and among these three solvents only
neat water showed a significant difference in the measured
pyrene II/IIII (1.96 for water versus 1.84 for BMIMPF6).29
Our experimental pyrene II/IIII (open circle, dotted lines) at
ambient conditions is presented in Figure 2 (panels A-D show
pyrene II/IIII for BMIMPF6 + 0.55, 0.65, 0.75, and 0.85 mole
fractions of ethanol in water, respectively). The data clearly
show higher pyrene II/IIII in the higher BMIMPF6 mole fraction
region. At higher ethanol compositions where BMIMPF6 and
water mole fractions are low, as expected, the observed pyrene
II/IIII values are low. However, when the BMIMPF6 concentra-
tion is low, increasing the water content in ethanol results in
higher experimental pyrene II/IIII. To assess the extent of solute-
solvent and solvent-solvent interactions present in a multi-
component system, a first-order spectroscopic approach assumes
an idealized situation where solvent-solvent interactions can
be neglected. Here, the additive spectral response, Radditive, a
weighted measured (or observed) mole fraction, Xi, or volume
difference between experimental and predicted pyrene I
I/IIII in
aqueous ethanol decreases as the mole fraction of ethanol
increases), the addition of BMIMPF6 to the aqueous-ethanol
mixture clearly shows increased experimentally observed pyrene
II/IIII values as compared to predicted additive values. (3) Except
at extremely low BMIMPF6 concentrations, the differences
between experimental and predicted pyrene II/IIII values are
significant over the entire composition range.
Though our aqueous-ethanol pyrene II/IIII data suggest
preferential solvation of pyrene by the less polar ethanol, it is
clearly not the case in the ternary BMIMPF6 + ethanol + water
system. The fact that experimental pyrene II/IIII values are higher
than predicted on the addition of BMIMPF6 to aqueous ethanol
may imply the possibility of favorable interaction(s) between
pyrene and BMIMPF6 rather than between pyrene and ethanol.
It can be suggested that a lack of the presence of any polar

Probe Behavior in BMIMPF6 + Ethanol + Water Solutions
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13535
Figure 3. 1,3-Bis(1-pyrenyl)propane IM/IE (10 µM) in ternary BMIMPF6 + ethanol + water: (A) BMIMPF6 + 0.55 mole fraction ethanol in
water; (B) BMIMPF6 + 0.65 mole fraction ethanol in water; (C) BMIMPF6 + 0.75 mole fraction ethanol in water; (D) BMIMPF6 + 0.85 mole
fraction ethanol in water, respectively. Key: (open circles, dotted lines) experimentally measured values; (filled circles, solid lines) predicted
additive values.
functionalities within the pyrene molecular architecture may
force pyrene moieties to have a solvation sphere constituted
primarily of BMIMPF6. Further, these observations may also
suggest the presence of interactions between the π-electron cloud
on pyrene with the 1-butyl-3-methylimidazolium cation moiety.
However, examination of the data presented in Figure 2 clearly
indicates experimentally observed pyrene II/IIII values that are
due to fluorescence from an intramolecular excimer.44 In a low-
viscous solvent, the two pyrenes easily fold together to form
an intramolecular excimer. As the microviscosity in the cy-
botactic region increases, the efficiency of the excimer formation
decreases and a corresponding reduction in the intensity of the
excimer band (band maximum ca. 450-500 nm) is observed.
Monomer-to-excimer emission intensity ratios (IM/IE) for 1,3-
bis(1-pyrenyl)propane dissolved in ternary BMIMPF6 + ethanol
+ water is presented in Figure 3 (panels A-D present 1,3-bis-
(1-pyrenyl)propane IM/IE for BMIMPF6 + 0.55, 0.65, 0.75, and
0.85 mole fraction of ethanol in water, respectively). A close
inspection of Figure 3 reveals that, as expected, the IM/IE, or
the microviscosity of the cybotactic region, decrease as the
concentrations of the less viscous components in the mixture,
ethanol and water, increase. We observed similar behavior from
this probe dissolved in binary BMIMPF6 + ethanol and
BMIMPF6 + water.23 A modification of eq. 2 (i.e., replacing I1
with IM, and IIII with IE) allows us to calculate the additive IM/
IE at various compositions of the ternary mixture. The predicted
additive probe responses IM/IE are conveniently represented in
Figure 3. It is important to mention here that the solubility of
1,3-bis(1-pyrenyl)propane in neat water is extremely low, and
therefore it is impossible for us to measure any meaningful IM/
IE in neat water. As a consequence, the additive 1,3-bis(1-
pyrenyl)propane IM/IE in ternary BMIMPF6 + ethanol + water
are calculated using IM/IE for binary aqueous ethanol and neat
BMIMPF6, respectively.
A careful examination of Figure 3 clearly implies that our
measured IM/IE does not show the predicted additive behavior
upon addition of RTIL BMIMPF6 to the aqueous-ethanol
solution. For all aqueous-ethanol compositions investigated, the
departure of 1,3-bis(1-pyrenyl)propane IM/IE from additivity first
increases and then decreases as the concentration of the
BMIMPF6 is increased. It seems that the deviation from
additivity is considerably significant in the BMIMPF6 mole
fraction range ∼0.30 to ∼0.80. Further, lower experimental 1,3-
bis(1-pyrenyl)propane IM/IE than predicted from additivity
initially suggests that the solvation sphere immediately sur-
rounding 1,3-bis(1-pyrenyl)propane is enriched in low-viscosity
component(s) of the mixture, i.e., either ethanol, water, or both,
higher than that observed in neat BMIMPF6 (pyrene II/IIII
)
1.84 in neat BMIMPF6). In our system, water is the only solvent
in which the experimentally observed pyrene II/IIII is higher than
1.84 (pyrene II/IIII ) 1.96 in neat water). Considering the
molecular architecture and the associated hydrophobic nature
of pyrene combined with its extremely low solubility in water
(<1 µM), it is inconceivable to imagine a solvation microsphere
composition surrounding excited-state pyrene that is rich in
water. Further, the aqueous-ethanol results clearly indicate a
situation quite contrary to this.
In a recent investigation using ATR and transmission IR
spectroscopic techniques, Kazarian et al. have investigated the
state of water within many RTILs, including BMIMPF6.41 Their
results indicated that molecules of water are present in the “free”
(not self-aggregated) state, bound via hydrogen-bonding with
PF6-. Further, they also concluded that most of the water
molecules exist in symmetric 1:2 type hydrogen-bonded com-
plexes: PF6-...HOH...PF6-. A possible explanation of experi-
mental pyrene II/IIII values in BMIMPF6 + ethanol + water
that are higher than those observed in neat BMIMPF6 could be
due to the favorable interaction(s) between the probe and the
RTIL; however, the presence of a hydrogen-bonded complex
of the anion of the RTIL with water may increase the water
composition within the cybotactic region of the excited-state
pyrene probe. This will result in a higher experimentally
observed pyrene II/IIII than observed in neat BMIMPF6.
1,3-Bis(1-pyrenyl)propane. A change in the viscosity of the
immediate microenvironment surrounding a probe molecule is
effectively manifested through 1,3-bis(1-pyrenyl)propane steady-
state emission spectra.42,43 It is well-established that, in addition
to a usual structured monomer fluorescence band, the emission
spectra of these compounds exhibit a broad and structureless
band with maximum intensity in the vicinity of 450-500 nm

13536 J. Phys. Chem. B, Vol. 107, No. 48, 2003
Fletcher and Pandey
Figure 4. 1-Pyrenecarboxaldehyde (10 µM) lowest energy emission maxima in ternary BMIMPF6 + ethanol + water: (A) BMIMPF6 + 0.55 mole
fraction ethanol in water; (B) BMIMPF6 + 0.65 mole fraction ethanol in water; (C) BMIMPF6 + 0.75 mole fraction ethanol in water; (D) BMIMPF6
+ 0.85 mole fraction ethanol in water, respectively. Key: (open circles, dotted lines) experimentally measured values; (filled circles, solid lines)
predicted additive values.
in comparison to the bulk-phase composition (viscosity of
BMIMPF6 is reported to be 312 cP at 303 K).45 This is in
contrast to what is observed and suggested in the case of pyrene
(vide infra). It is important to mention here that it is well-known
that the addition of even small quantities of organic solvents to
RTILs results in a significant decrease in the bulk viscosity of
the medium.22 It is possible that the observed changes in the
IM/IE arise principally from the decrease in the bulk viscosity
of the BMIMPF6 + ethanol + water solutions as compared to
neat BMIMPF6. This decrease in bulk viscosity of BMIMPF6
on addition of aqueous ethanol can be envisioned as a
consequence of solvent-solvent interactions, and as a result,
the increased efficiency of intramolecular excimer formation
may not be due to some specific solute-solvent interactions.
To further substantiate this reasoning, we measured the bulk
viscosities of the BMIMPF6 + water system at 27.6 °C in a
separate preliminary investigation (a detailed report of these
experiments and their interpretation will appear in a separate
publication soon). For 0.03, 0.08, 0.19, and 0.26 mole fraction
of water in BMIMPF6 (0.26 mole fraction is close to the
solubility limit of water in BMIMPF623a), the experimentally
measured viscosities were significantly lower than those cal-
culated from additivity (∼8%, 30%, 98%, and 114% lower for
0.03, 0.08, 0.19, and 0.26 mole fraction added water, respec-
tively). It is clear that this dramatic reduction in bulk viscosity
(as compared to that obtained from additivity) results in
increased intramolecular excimer formation efficiency in 1,3-
bis(1-pyrenyl)propane.
1-Pyrenecarboxaldehyde. The spectral response of 1-pyrene-
carboxaldehyde is known to be dependent on the static dielectric
constant of the cybotactic region.46 Bredereck et al. were the
first to report on the solvent-dependent emission characteristics
of this probe.47 In nonpolar solvents, the fluorescence spectra
of 1-pyrenecarboxaldehyde are highly structured and fluores-
cence quantum yields are fairly low. However, as the polarity
of the surrounding medium is increased, the fluorescence
emitting state changes and it is manifested through a broad,
structureless, and moderately intense fluorescence behavior. The
fluorescence maxima of 1-pyrenecarboxaldehyde is known to
red-shift with increasing static dielectric constant of the sur-
rounding milieu.46,47 For this reason, 1-pyrenecarboxaldehyde
has been used as a probe of solvent static dielectric constant in
various isotropic and complex solubilizing media.
Figure 4 presents the lowest energy emission maxima of
1-pyrenecarboxaldehyde (ν in kK units) measured in ternary
BMIMPF6 + ethanol + water solutions (panels A-D present
1-pyrenecarboxaldehyde lowest energy emission maxima, ν, for
BMIMPF6 + 0.55, 0.65, 0.75, and 0.85 mole fraction of ethanol
in water, respectively). We have used eq 1 to determine the
extent of preferential solvation (or lack thereof) by using lowest
energy emission maxima of 1-pyrenecarboxaldehyde, ν in kK
units, for spectral response, R, as suggested in the litera-
ture.27,36,40 The calculated additive values are represented as
filled circles and solid lines whereas the experimentally
measured values are shown using open circles and dotted lines
in Figure 4. A close examination of data presented in Figure 4
reveals interesting probe behavior. The 1-pyrenecarboxaldehyde
probe response in aqueous-ethanol solutions strongly suggests
a preferential solvation by ethanol. The more interesting fact is
that there is no significant change in 1-pyrenecarboxaldehyde
lowest energy emission maxima upon addition of up to ∼0.35
mole fraction BMIMPF6 to the aqueous-ethanol solutions.
Further, as more BMIMPF6 is added to the aqueous-ethanol
solutions, ν for 1-pyrenecarboxaldehyde increases; nonetheless,
the measured values are lower than those predicted from
additivity. This overall behavior strongly suggests a preferential
solvation of the excited-state 1-pyrenecarboxaldehyde by ethan-
ol. It is interesting to observe, however, for very high BMIMPF6
concentrations the departure from additivity for this probe is
decreased. Our previous studies of 1-pyrenecarboxaldehyde
behavior within binary BMIMPF6 + ethanol and BMIMPF6 +
water suggested an augmented local concentration of ethanol
and water (with respect to the bulk), respectively,23 in the
solvation microsphere surrounding the excited-state probe
molecule. Although in our previous studies, 1-pyrenecarboxal-
dehyde probe behavior suggested the possibility of preferential
solvation by water when solubilized within the BMIMPF6 +
water system, the extent of preferential solvation by ethanol
was observed to be greater in the BMIMPF6 + ethanol solvent
system.23 We believe the presence of the polar aldehyde

Probe Behavior in BMIMPF6 + Ethanol + Water Solutions
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13537
Figure 5. ET(30) using lowest energy absorbance maxima of Reichardt’s betaine dye (100 µM) in ternary BMIMPF6 + ethanol + water: (A)
BMIMPF6 + 0.55 mole fraction ethanol in water; (B) BMIMPF6 + 0.65 mole fraction ethanol in water; (C) BMIMPF6 + 0.75 mole fraction
ethanol in water; (D) BMIMPF6 + 0.85 mole fraction ethanol in water, respectively. The corrected additive profiles are denoted in panels B and
D as short-dashed lines (according to eq 3). Key: (open circles, dotted lines) experimentally measured values; (filled circles, solid lines) predicted
additive values.
functional group may induce an enrichment of ethanol within
the cybotactic region. Solubility constraints may restrict a similar
affinity of this probe for water as the solubility of 1-pyrenecar-
boxaldehyde in water is significantly lower than that in ethanol.
Reichardt’s Betaine Dye. One of the most widely used
water revealed no preferential solvation, binary BMIMPF6 +
ethanol mixtures showed strong synergistic effects.23 For ternary
BMIMPF6 + ethanol + water, contrary to what was observed
for 1-pyrenecarboxaldehyde, experimental ET(30) increases
significantly as small amounts of BMIMPF6 are added to
aqueous ethanol and as aqueous ethanol is added to BMIMPF6,
the increase being more drastic for the former. Although the
predicted curve for additive solvation suggests a linear decrease
in ET(30) with increasing BMIMPF6 mole fraction, the experi-
mentally observed values rapidly increase to a maximum near
0.10 mole fraction BMIMPF6 falling off slowly thereafter.
Because aqueous-ethanol mixtures do not follow additivity, one
can generate a corrected additive curve for the ternary system
using the expression
empirical scales of solvent polarities is the ET(30) scale, where
abs
max
ET(30) (in kcal mol-1) ) 28591/λmabasx (in nm).48-50
λ
is the
maximum wavelength of the lowest energy, intramolecular
charge-transfer π-π* absorption band of the zwitterionic 2,6-
diphenyl-4-(2,4,6-triphenyl-N-pyridino)phenolate molecule. This
zwitterionic compound, also known as Reichardt’s betaine dye,
exhibits one of the larger observed solvatochromic effects of
any known organic molecule. Because of its zwitterionic nature,
solvatochromic probe behavior of Reichardt’s dye is strongly
affected by the hydrogen-bond donating acidity of the solvent;
hydrogen-bond donating solvents stabilize the ground state more
than the excited state.
Radditive
)
X R0 + ∆E (30)
(3)
∑
i
i
T
i
It has been shown by our group and many others that the
ET(30) in dry BMIMPF6 is similar to that observed in ethan-
ol.24,25,29,31 Figure 5 (panels A-D show ET(30) for BMIMPF6
+ 0.55, 0.65, 0.75, and 0.85 mole fraction of ethanol in water,
respectively) presents experimentally observed ET(30) for Rei-
chardt’s dye in ternary BMIMPF6 + ethanol + water solutions.
The additive Reichardt’s dye behavior in ternary BMIMPF6 +
ethanol + water mixtures was calculated using eq. 1. These
predicted ET(30) from additivity are presented as filled circles
and solid lines in Figure 5. Examination of the experimental
ET(30) in Figure 5 reveals anomalous probe behavior in the
ternary mixture. It is instructive to first consider the behavior
of Reichardt’s dye in aqueous-ethanol mixtures. We observe
that the measured ET(30) in aqueous ethanol, for the ethanol
mole fraction range considered here, are lower than that
predicted from additivity, the departure from additivity decreas-
ing with increased ethanol mole fraction. These observations
are in agreement with literature reports.51 Most of these
investigations explain such observations by suggesting the
presence of a “third” solvent environment formed as a result of
ethanol-water complexation. Though our previous investigation
of Reichardt’s dye solvation behavior in binary BMIMPF6 +
where ∆ET(30), the “excess polarity”, is the difference between
the measured ET(30) and that obtained from additivity for
aqueous ethanol at a given ethanol mole fraction. The so-
corrected curves for additive behavior for 0.65 and 0.85 mole
fraction ethanol are included in Figure 5B,D. In this way, it
becomes clear that Reichardt’s dye is preferentially solvated
by ethanol in ethanol-water mixtures whereas in BMIMPF6 +
ethanol + water mixtures preferential solvation by water occurs.
This remarkable result may have important implications for
chemical processing within RTIL + ethanol + water mixtures.
Kamlet-Taft Treatment. To further explore these observa-
tions, we performed Kamlet-Taft investigation of ternary
BMIMPF6 + ethanol + water system.48-50,52-54 In this treat-
ment, the solvent dipolarity/polarizability, π*, is estimated using
the response of N,N-diethyl-4-nitroaniline, a non-hydrogen bond
donor solute by using
π* ) 0.314(27.52 - νDENA,max
)
(4)
where νDENA,max is the position of the absorbance maximum of
N,N-diethyl-4-nitroaniline in kK units.50,52-54 The hydrogen

13538 J. Phys. Chem. B, Vol. 107, No. 48, 2003
Fletcher and Pandey
Figure 6. Dipolarity/polarizability (π*, open circles), hydrogen-bond donating acidity, HBD (R, open triangles), and hydrogen-bond accepting
basicity, HBA (â, open squares) in ternary BMIMPF6 + ethanol + water: (A) BMIMPF6 + 0.55 mole fraction ethanol in water; (B) BMIMPF6 +
0.65 mole fraction ethanol in water; (C) BMIMPF6 + 0.75 mole fraction ethanol in water; (D) BMIMPF6 + 0.85 mole fraction ethanol in water,
respectively.
bond accepting basicity (HBA), â, for solvents can be deter-
mined by using the enhanced solvatochromic shift of 4-nitro-
aniline relative to the homomorphic N,N-diethyl-4-nitroaniline
and using the expression:
(Figure 5) which, in part, depends on the dipolarity/polarizability
of the solubilizing medium.
The HBA (i.e., â) ability of a solvent depends on the
absorbance maxima of both N,N-diethyl-4-nitroaniline and
4-nitroaniline (eq 5). The â for ternary BMIMPF6 + ethanol +
water are shown (open squares) in Figure 6 (panels A-D show
â for BMIMPF6 + 0.55, 0.65, 0.75, and 0.85 mole fractions of
ethanol in water, respectively). As expected, on addition of
BMIMPF6 to aqueous ethanol, the HBA of the solvent mixture
decreases; the decrease is initially rapid but becomes more
gradual at higher BMIMPF6 mole fractions. Unlike ethanol and
water, BMIMPF6 lacks a genuine HBA moiety or site. PF6- is
known to have a compact structure possessing much weaker
HBA basicity in comparison to ethanol or water.41
â ) 0.358(31.10 - νNA,max) - 1.125π*
(5)
Here, νNA, max is the observed absorbance maxima for 4-nitro-
aniline in kK units.48-50,52-54 Finally, the hydrogen bond
donating acidity (HBD), R, of the solvent is calculated using
48-50,52-54
the ET(30) and π* values:
R ) 0.0649ET(30) - 2.03 - 0.72π*
(6)
For ternary BMIMPF6 + ethanol + water, measured π* are
presented (open circles) in Figure 6 (panels A-D show π* for
BMIMPF6 + 0.55, 0.65, 0.75, and 0.85 mole fractions of ethanol
in water, respectively). It is clear that as BMIMPF6 is added to
aqueous ethanol, π* increases rapidly, reaches a maxima, and
then gradually decreases to its value in neat BMIMPF6. It is
interesting to note that the position of the π* maxima shifts to
higher BMIMPF6 mole fraction as the concentration of ethanol
is increased. It is evident from the data that the dipolarity/
polarizability increases as BMIMPF6 is added to aqueous-
ethanol mixtures. The pyrene polarity scale (i.e., pyrene II/IIII)
is known, in part, to depend on the dipolarity/polarizability of
the solubilizing media. A comparison of data in Figure 2 and
Figure 6 reveals that pyrene II/IIII also increases rapidly as
BMIMPF6 is added to aqueous-ethanol mixtures. At higher
BMIMPF6 mole fractions, the change in pyrene II/IIII, for the
most part, is not as drastic. However, the behavior of 1-pyrene-
carboxaldehyde (Figure 4), which is largely dependent on the
static dielectric constant of the probe cybotactic region, does
not show any similarity with observed π* values of the medium.
We propose that an enrichment of the cybotactic solvation
microsphere of excited 1-pyrenecarboxaldehyde by ethanol (as
suggested earlier) would not allow this probe to report on the
bulk solvent characteristics. Similar trends between observed
ET(30) and π* confirm the behavior of Reichardt’s betaine dye
The HBD ability of a solvent system, R, depends on ET(30)
and π* (eq 6). Kamlet and Taft calculated that approximately
two-thirds of the shift in the absorbance maxima of the
Reichardt’s dye could be assigned directly to specific interac-
tions involving the phenoxide oxygen (see Figure 1).52 This
suggests that the ET(30) scale is largely, albeit not exclusively,
a measure of the HBD ability of the solvent system (vide supra).
In light of these facts, it is clear that, as the composition of the
ternary system is varied, any significant changes in the
BMIMPF6 + ethanol + water R values (and π*, to a lesser
extent) should be manifested through the observed ET(30). We
remind the reader that the cybotactic region experienced by
Reichardt’s dye in this solvent mixture may differ from those
observed by 4-nitroaniline and N,N-diethyl-4-nitroaniline due
to the possibility of differential preferential solvation. For the
ternary BMIMPF6 + ethanol + water, measured R are presented
(open triangle) in Figure 6 (panels A-D show R for BMIMPF6
+ 0.55, 0.65, 0.75, and 0.85 mole fractions of ethanol in water,
respectively). A careful examination of the R clearly shows a
sharp increase in HBD ability as very small amounts of
BMIMPF6 are added, followed by a gradual decrease as more
BMIMPF6 is added to aqueous-ethanol mixtures. A comparison
of ET(30) (Figure 5) with R (Figure 6) for the same compositions
of the ternary mixture demonstrates a very similar trend (refer
to eq 6). Therefore, the surprising nature of ET(30) (i.e., the

Probe Behavior in BMIMPF6 + Ethanol + Water Solutions
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13539
(9) Liu, F. C.; Abrams, M. C.; Baker, R. T.; Tumas, W. Chem.
Commun. 2002, 433.
(10) Wheeler, C.; West, K. N.; Liotta, C. L.; Eckert, C. A. Chem.
Commun. 2001, 887.
(11) Dzyuba, S. V.; Bartsch, R. A. Chem. Commun. 2001, 1466.
(12) Sheldon, R. Chem. Commun. 2001, 2399.
(13) Gordon, C. M. Appl. Catal. A: Gen. 2001, 222, 101.
(14) Wasserscheid, P.; Keim, W. Angew. Chem. Int. Ed. 2000, 39, 3773.
(15) Anthony, J. L.; Maginn, E. J.; Brennecke, J. F. J. Phys. Chem. B
2002, 106, 7315.
(16) Noda, A.; Hayamizu, K.; Watanabe, M. J. Phys. Chem. B 2001,
105, 4603.
deviation from additivity, vide infra, Figure 5) may be attributed
to the corresponding changes in the HBD acidity as well as the
dipolarity/polarizability of the system as the composition of the
ternary mixture is varied. It appears that the Kamlet-Taft
treatment is able to explain, in part, the experimentally observed
solvatochromic probe behavior (i.e., ET(30) and pyrene II/IIII)
for ternary BMIMPF6 + ethanol + water. A major part of the
current efforts in our group is focused on identifying and/or
developing preferential solvation models that may conform to
our experimental solvatochromic probe data within RTIL-based
multicomponent systems.
(17) Grodkowski, J.; Neta, P. J. Phys. Chem. B 2002, 106, 9030.
(18) Hyun, B. R.; Dzyuba, S. V.; Bartsch, R. A.; Quitevis, E. L. J. Phys.
Chem. A 2002, 106, 7579.
(19) Najdanovic-Visak, V.; Esperanca, J. M. S. S.; Rebelo, L. P. N.; da
Ponte, M. N.; Guedes, H. J. R.; Seddon, K. R.; Szydlowski, J. Phys. Chem.
Chem. Phys. 2002, 4, 1701.
(20) Swatloski, R. P.; Visser, A. E.; Reichert, W. M.; Broker, G. A.;
Farina, L. M.; Holbrey, J. D.; Rogers, R. D. Chem. Commun. 2001, 2070.
(21) Swatloski, R. P.; Visser, A. E.; Reichert, W. M.; Broker, G. A.;
Farina, L. M.; Holbrey, J. D.; Rogers, R. D. Green Chem. 2002, 4, 81.
(22) Seddon, K. R.; Stark, A.; Torres, M. J. Pure Appl. Chem. 2000,
72, 2275.
(23) (a) Fletcher, K. A.; Pandey, S. Appl. Spectrosc. 2002, 56, 266. (b)
Fletcher, K. A.; Pandey, S. Appl. Spectrosc. 2002, 56, 1498.
(24) Aki, S. N. V. K.; Brennecke, J. F.; Samanta, A. Chem. Commun.
2001, 413.
Conclusion
The deviations between experimentally observed and pre-
dicted additive behavior of the probes could arise not only from
specific solute-solvent interactions but also due the altered
properties of the solution due to any solvent-solvent interac-
tions. For most cases, the dipolarity of the ternary solution
appears to be greater than that predicted from the simple
solvation model used in the current study. Presumably, interac-
tions between pyrene and the imidazolium cation and/or
-
interactions between water molecules and the PF6 anion
(25) Muldoon, M. J.; Gordon, C. M.; Dunkin, I. R. J. Chem. Soc., Perkin
account for these observations. Model predictions indicate 1,3-
bis(1-pyrenyl)propane may be preferentially solvated by the
lower-viscosity components in solution; however, these observa-
tions seem to result from a reduction in the bulk viscosity of
BMIMPF6 due to the presence of low-viscosity cosolvents.
1-Pyrenecarboxaldehyde appears to be preferentially solvated
by ethanol; however, for XBMIMPF6 < ∼0.35, the observed
response of 1-pyrenecarboxaldehyde remains constant. As the
composition of BMIMPF6 increases past 0.35 mole fraction,
the observed response of 1-pyrenecarboxaldehyde is lower than
that obtained from additivity, indicating the static dielectric
constant of the cybotactic region surrounding this probe in the
ternary solvent system is lower than predicted. On the basis of
the response of Reichardt’s betaine dye, synergistic effects are
again observed upon the addition of aqueous ethanol to
BMIMPF6. The dipolarity/polarizability and hydrogen-bond
donating ability of the ternary solution first increase and then
gradually decrease as the composition of BMIMPF6 increases.
The hydrogen-bond accepting ability decreases with increasing
BMIMPF6. The results of the current study indicate that one is
able to significantly alter the physicochemical properties of
BMIMPF6 by simply adding appropriate amounts of ethanol
and water.
Trans. 2 2001, 433.
(26) Carmichael, A. J.; Seddon, K. R. J. Phys. Org. Chem. 2000, 13,
591.
(27) (a) Marcus, Y. Chem. Soc. ReV. 1993, 409. (b) Reichardt, C.
SolVents and SolVent Effects in Organic Chemistry, 2nd, revised and enlarged
ed., VCH: New York, 1990; p 34 footnote.
(28) Roses, M.; Rafols, C.; Ortega, J.; Bosch, E. J. Chem. Soc., Perkins
Trans 2 1995, 1607.
(29) Fletcher, K. A.; Storey, I. K.; Hendricks, A. E.; Pandey, S.; Pandey,
S. Green Chem. 2001, 3, 210.
(30) Baker, S. N.; Baker, G. A.; Kane, M. A.; Bright, F. V. J. Phys.
Chem. B 2001, 105, 9663.
(31) Baker, S. N.; Baker, G. A.; Bright, F. V. Green Chem. 2002, 4,
165.
(32) Werner, J. H.; Baker, S. N.; Baker, G. A. Analyst 2003, 128, 786.
(33) Waris, R.; Acree, W. E., Jr.; Street, K. W., Jr. Analyst 1988, 113,
1465.
(34) Dong, D. C.; Winnik, M. A. Can. J. Chem. 1984, 62, 2560.
(35) Karpovich, D. S.; Blanchard, G. J. J. Phys. Chem. 1995, 99, 3951.
(36) Marcus, Y. J. Chem. Soc., Perkins Trans 2 1994, 1015.
(37) Chatterjee, P.; Bagchi, S. J. Chem. Soc., Faraday Trans. 1991, 87,
537.
(38) Bosch, E.; Roses, M. J. Chem. Soc., Faraday Trans. 1992, 88, 3541.
(39) Phillips, D. J.; Brennecke, J. F. Ind. Eng. Chem. Res. 1993, 32,
943.
(40) Acree, Jr., W. E.; Wilkins, D. C.; Tucker, S. A.; Griffin, J. M.;
Powell, J. R. J. Phys. Chem. 1994, 98, 2537 and references therein.
(41) Cammarata, L.; Kazarian, S. G.; Salter, P. A.; Welton, T. Phys.
Chem. Chem. Phys. 2001, 3, 5192.
(42) Snare, M. J.; Thistlethwaite, P. J.; Ghiggino, K. P. J. Am. Chem.
Soc. 1983, 105, 3328.
Acknowledgment. This work was supported in part by a
Sandia-University Research Program (SURP) grant. We thank
Rebecca A. Redden, Ashley E. Hendricks, and Isaiah K. Storey
for their help during the experiments. We also appreciate
suggestions and efforts from one of the reviewers involved
during the peer-review process of this manuscript.
(43) Zachariasse, K. A. Chem. Phys. Lett. 1978, 57, 429.
(44) Birks, J. B. Photophysics of Aromatic Molecules; Wiley-Inter-
science: New York, 1970.
(45) Suarez, P. A. Z.; Einloft, S.; Dullius, J. E. L.; de Souza, R. F.;
Dupont, J. J. Chim. Phys. Phys.-Chim. Biol. 1998, 95, 1626.
(46) Kalyanasundaram, K.; Thomas, J. K. J. Phys. Chem. 1977, 81, 2176.
(47) Bredereck, K.; Forster, Th.; Oenstein, H. G. Luminescence of
Inorganic and Organic Materials; Kallman, H. P., Spruch, G. M., Eds;
Wiley: New York, 1960.
(48) Reichardt, C. Chem. ReV. 1994, 94, 2319 and references therein.
(49) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry;
VCH: Weinheim, 1988.
(50) Kamlet, M. J.; Abboud, J. L.; Taft, R. W. J. Am. Chem. Soc. 1977,
99, 6027.
(51) (a) Tada, E. B.; Novaki, L. P.; El Seoud, O. A. J. Phys. Org. Chem.
2000, 13, 679. (b) Wu, Y. G.; Tabata, M.; Takamuku, T. J. Soln. Chem.
2002, 31, 381. (c) Marcus, Y. J. Chem. Soc., Perkins Trans. 2 1994, 1751.
(52) Taft, R. W.; Kamlet, M. J. J. Am. Chem. Soc. 1976, 98, 2886.
(53) Kamlet M. J.; Taft, R. W. J. Am. Chem. Soc. 1976, 98, 377.
(54) Kamlet, M. J.; Abboud, J. L.; Abraham, M. H.; Taft, R. W. J. Org.
Chem. 1983, 48, 2877.
References and Notes
(1) Seddon, K. R. J. Chem. Technol. Biotechnol. 1997, 68, 351.
(2) Freemantle, M.Chem. Eng. News 1998, 76, 32.
(3) Freemantle, M. Chem. Eng. News 2000, 78, 37.
(4) Welton, T. Chem. ReV. 1999, 99, 2071.
(5) Carmichael, A. J.; Haddleton, D. M.; Bon, S. A. F.; Seddon, K. R.
Chem. Commun. 2000, 1237.
(6) Visser, A. E.; Swatloski, R. P.; Reichert, W. M.; Griffin, S. T.;
Rogers, R. D. Ind. Eng. Chem. Res. 2000, 39, 3596.
(7) Fletcher, K. A.; Pandey, S.; Storey, I. K.; Hendricks, A. E.; Pandey,
S. Anal. Chim. Acta 2002, 453, 89.
(8) Visser, A. E.; Swatloski, R. P.; Reichert, W. M.; Mayton, R.; Sheff,
S.; Wierzbicki, A.; Davis, J. H.; Rogers, R. D. Chem. Commun. 2001, 135.