Effect of Cation on Physical Properties and CO2 Solubility for Phosphonium-Based Ionic Liquids with 2-Cyanopyrrolide Anions

Article abstract of DOI:10.1021/acs.jpcb.5b05733

A series of tetraalkylphosphonium 2-cyanopyrrolide ([P_nnnn][2-CNPyr]) ionic liquids (ILs) were prepared to investigate the effect of cation size on physical properties and CO₂ solubility. Each IL was synthesized in our laboratory and characterized by NMR spectroscopy. Their physical properties, including density, viscosity, and ionic conductivity, were determined as a function of temperature and fit to empirical equations. The density gradually increased with decreasing cation size, while the viscosity decreased noticeably. In addition, the [P_nnnn][2-CNPyr] ILs with large cations exhibited relatively low degrees of ionicity based on analysis of the Walden plots. This implies the presence of extensive ion pairing or formation of aggregates resulting from van der Waals interactions between the long hydrocarbon substituents. The CO₂ solubility in each IL was measured at 22 °C using a volumetric method. While the anion is typically known to be predominantly responsible for the CO₂ capture reaction, the [P_nnnn][2-CNPyr] ILs with shorter alkyl chains on the cations exhibited slightly stronger CO₂ binding ability than the ILs with longer alkyl chains. We attribute this to the difference in entropy of reaction, as well as the variation in the relative degree of ionicity.

Full text of DOI:10.1021/acs.jpcb.5b05733

Article
pubs.acs.org/JPCB
Effect of Cation on Physical Properties and CO Solubility for
2
Phosphonium-Based Ionic Liquids with 2‑Cyanopyrrolide Anions
Samuel Seo, M. Aruni DeSilva, Han Xia, and Joan F. Brennecke*
Department of Chemical and Biomolecular Engineering, University of Notre Dame, Notre Dame, Indiana 46556, United States
*
S Supporting Information
ABSTRACT: A series of tetraalkylphosphonium 2-cyanopyrrolide
[Pnnnn][2-CNPyr]) ionic liquids (ILs) were prepared to investigate the
(
effect of cation size on physical properties and CO solubility. Each IL was
2
synthesized in our laboratory and characterized by NMR spectroscopy.
Their physical properties, including density, viscosity, and ionic
conductivity, were determined as a function of temperature and fit to
empirical equations. The density gradually increased with decreasing cation
size, while the viscosity decreased noticeably. In addition, the [Pnnnn][2-
CNPyr] ILs with large cations exhibited relatively low degrees of ionicity
based on analysis of the Walden plots. This implies the presence of
extensive ion pairing or formation of aggregates resulting from van der
Waals interactions between the long hydrocarbon substituents. The CO2
solubility in each IL was measured at 22 °C using a volumetric method.
While the anion is typically known to be predominantly responsible for the
CO capture reaction, the [P ][2-CNPyr] ILs with shorter alkyl chains on the cations exhibited slightly stronger CO binding
2
nnnn
2
ability than the ILs with longer alkyl chains. We attribute this to the difference in entropy of reaction, as well as the variation in
the relative degree of ionicity.
INTRODUCTION
points or glass transition temperatures regardless of the anion
with which it is paired. However, its large molecular weight is
disadvantageous in terms of volumetric and gravimetric CO2
uptake ratio. Utilizing the tetraalkylphosphonium cation allows
one to easily tune the size of the cation (and IL) by adjusting
the length of the hydrocarbon substituents. In this regard, our
group recently showed that phase-change ionic liquids (PCILs)
■
Ionic liquids (ILs) are defined as salts with melting points
below 100 °C. The unusual properties of ILs, including
extremely low vapor pressures, high thermal stabilities, and
wide liquid ranges, make them excellent replacements for the
current aqueous amine solutions used in postcombustion CO2
1
capture from flue gas. In addition to such physical properties,
can be used for postcombustion CO capture. PCILs are ILs
2
wide tunability allows one to tune the structure of the
component ions. For instance, by tethering an amine
functionality to the anion,
designed to react chemically with CO even at low partial
pressures. Unfortunately, most of these task-specific ILs suffer
from high viscosity and slow kinetics for CO2 capture,
especially when they react with CO . Recently, our group
proposed the use of aprotic heterocyclic anions (AHAs), which
can react reversibly and in a 1:1 stoichiometric ratio with CO
16
that turn from solid to liquid upon reaction with CO2 and
take advantage of IL tunability: shortening the hydrocarbon
substituents on the tetraalkylphosphonium cation effectively
increases the melting point, which is a key physical parameter
2
−6
7
8
cation, or both, ILs can be
2
17
that made the PCIL process thermodynamically feasible. In
this study, a series of tetraalkylphosphonium ([Pnnnn]+, n =
number of carbons in each hydrocarbon substituent, where the
n values do not have to be equal) ILs were synthesized and
characterized to investigate the effect of cation size (by
adjusting the alkyl chain length) on physicochemical properties
9
2
2
4
,10
with readily tunable enthalpy of reaction.
Unlike other ILs
that react with CO , the viscosity of these AHA ILs with
2
such as viscosity, density, ionic conductivity, and CO solubility.
tetraalkylphosphonium cations remain unchanged upon re-
2
11
Each cation is paired with a 2-cyanopyrrolide anion, [2-
action with CO2.
ILs with phosphonium cations are generally known for their
−
CNPyr] : this anion was chosen based on its high CO
2
12−14
15
capacity, moderate enthalpy of reaction (ΔH° ) with CO ,
high thermal stability
and ease of large-scale production.
rxn
2
10
excellent thermal stability, as well as fast CO -reaction
kinetics. Owing to these favorable properties, Brown et al.
In this context, ILs with a phosphonium cation are great
candidates for absorbents used in postcombustion CO capture
2
18
19
2
processes. Especially, trihexyl(tetradecyl)phosphonium
+
(
[P66614] ) is one of the most commonly adopted cations in
Received: June 16, 2015
Revised: August 3, 2015
Published: August 12, 2015
the development of ILs for CO capture applications, as its
bulkiness and asymmetricity generally induce very low melting
2
©
2015 American Chemical Society
11807
DOI: 10.1021/acs.jpcb.5b05733
J. Phys. Chem. B 2015, 119, 11807−11814

The Journal of Physical Chemistry B
Article
Table 1. Chemical Structures and Nomenclatures of [Pnnnn][2-CNPyr] ILs
a
P2222] = tetraethylphosphonium, [P ] = triethyl(butyl)phosphonium, [P2228]+ = triethyl(octyl)phosphonium, [P
+
+
]+ = triethyl(dodecyl)-
= trihexyl(tetradecyl)-
[
2
2
2
4
2
2
2
1
2
+
+
+
phosphonium, [P44412
]
= tributyl(dodecyl)phosphonium, [P44418
]
= tributyl(octadecyl)phosphonium, [P66614
]
phosphonium).
recently reported enhanced gravimetric CO capacity and
viscosity with [2-CNPyr] with different cations, namely
tetraalkylammonium and tetraalkylphosphonium, and their
on imidazolium to form imidazolium-2-carboxylate when the
cation is paired with a fairly basic anion (e.g., acetate or
2
−
25−30
AHA).
Recently, Gohndrone et al. discovered that the
physical property trends and limited CO uptake data are
consistent with the more comprehensive results and analysis
presented in this study.
AHAs (and possibly other anions with similar basicities) are
also capable of abstracting an acidic α-hydrogen atom of
phosphonium cations to produce an ylide in the presence of
2
31
Since ILs are composed entirely of ions, the IL system may
be envisioned as a pool of free-flowing ions that are rapidly
associating and dissociating. However, most ILs do not follow
this ideal behavior due to strong interactions between two or
more ions, and this may lead to formation of ion pairs or ion
CO
somewhat different CO
series of [Pnnnn][2-CNPyr] ILs, and we attribute this to changes
2
at elevated temperature. In this study, we show that
uptake behavior was observed for a
2
in the entropy of the reaction (ΔS°xn) of the anion with CO
r
2
caused by the cation.
2
0
aggregations. Recent molecular simulation work done by Yee
et al. concluded that ILs do not behave like conventional alkali
halide salts, and the ions of ILs can remain associated in
MATERIALS AND METHODS
■
Chemicals and Synthesis. All ILs discussed in this study
21
aqueous solutions, even under highly diluted conditions. In
this regard, ion pairs or aggregates, if sufficiently long-lived, can
appear neutral, and thus they are not able to contribute to
consist of the 2-cyanopyrrolide anion paired with various
tetraalkylphosphonium cations. The chemical structure of each
ion is shown in Table 1. Trihexyl(tetradecyl)phosphonium 2-
cyanopyrrolide ([P66614][2-CNPyr]) was prepared by a two-
2
2
conduction. The degree to which ions are freely moving, as
opposed to existing as ion pairs or aggregates, is frequently
referred to as the “ionicity.” While Coulombic interactions are
predominant in ILs, the van der Waals interactions must not be
ignored, particularly when the cations include large hydro-
4
step synthesis protocol described in our previous work. First,
SBR LG NC(OH) ion-exchange resin or Amberlite IRN78
(
Sigma-Aldrich) in methanol (ACS grade, Fischer Scientific)
was used to produce trihexyl(tetradecyl)phosphonium hydrox-
ide ([P66614][OH]) from trihexyl(tetradecyl)phosphonium
bromide ([P66614][Br], 97% purity, Sigma-Aldrich). The resin
was then washed with methanol three times at room
temperature. After filtration of the resin, [P66614][OH] and
the anion precursor, pyrrole-2-carbonitrile (98%, Alfa Aeser),
were mixed in equimolar quantities, and an acid−base reaction
produced the desired IL. The reaction lasted 12 h with stirring
at room temperature. Excess methanol was separated by rotary
evaporator at 50 °C.
+
carbons (e.g., [P66614] ). Therefore, in this study, we explore the
relationship between the relative ionicity and the size of the IL
(
or cation). A Walden plot (log[molar conductivity] vs
−1
log[viscosity ]) approach, as suggested by Angell and co-
2
0
workers, is one way of qualitatively assessing the ionicity. ILs
with lower ionicity will exhibit substantially lower ionic
conductivity than other ILs with similar viscosities, and this
approach allows us to speculate on the relative ion conduction
behavior in the [Pnnnn][2-CNPyr] ILs in this study.
We also examine the effect of cation size on CO solubility.
2
For the preparation of ILs with tetraalkylphosphonium
cations containing shorter alkyl substituents, a three-step
procedure was adopted: the phosphonium halide precursor
was prepared in the first step, followed by the anion exchange
2
Since Gurkan et al. first reported the equimolar CO uptake
2
using anion-functionalized ILs under ambient conditions, the
majority of our research efforts and those of others who have
adopted the idea of functionalizing the anion to react with CO2
have focused on tuning and modifying the anion moiety. While
the choice of cation should be predominantly responsible for
and acid−base neutralization (see Figure 1). Tributyl(C )-
n
phosphonium bromide ([P4 ][Br]) and triethyl(C )-
44n
n
phosphonium bromide ([P222n][Br]) were prepared via the
alkylation of tributylphosphine (99%, VWR) or triethylphos-
phine (99%, Sigma-Aldrich), respectively. Corresponding
the physical properties of these ILs, the role of cation in CO
2
2
3
capture has been often neglected. On the other hand, Xu et al.
24
and Holloc
́
zki et al. reported that the basicity of ILs can be
trialkylphosphine and alkyl bromides, C H2n+1Br (>98%,
n
remarkably affected by the choice of the cation, as the degree of
cation−anion interactions can enhance or reduce the basicity of
the functionalized ion. It should be noted that several
researchers have shown that a dialkylimidazolium cation
Sigma-Aldrich) were reacted in acetonitrile and refluxed with
stirring at 60 °C. 1-Iodooctadecane, C H I (95%, Sigma-
18
37
Aldrich), was used instead of the alkyl bromide for the
preparation of tributyl(octadecyl)phosphonium iodide
([P44418][I]). The alkylation step was performed under
nitrogen atmosphere because of the sensitivity of the phosphine
(
which is often considered inert when paired with a
nonreacting anion) even has the potential to react with a
32
CO molecule via deprotonation of the acidic C(2) hydrogen
moiety toward oxygen in the air. Residual acetonitrile and
2
1
1808
DOI: 10.1021/acs.jpcb.5b05733
J. Phys. Chem. B 2015, 119, 11807−11814

The Journal of Physical Chemistry B
Article
heating from 10 to 70 °C. The uncertainty is approximately
±
5% above 100 cP and ±10% below 100 cP.
The ionic conductivity was determined by electrochemical
impedance spectroscopy (EIS) (Solatron SI 1260 impedance/
gain-phase analyzer and Solartron 1287 electrochemical
interface). The cells were purchased from Materials Mates,
and they consist of two Pt electrodes. Prior to the conductivity
measurements, the cell constant for each cell was determined
using diluted KCl standard solutions (from Palintest) as
references. To minimize the exposure to atmospheric moisture
Figure 1. Schematic of [P222n][2-CNPyr] synthesis procedure. The
synthesis of [P444n][AHA] ILs was started via alkylation of
tributylphosphine.
and CO , the IL samples were loaded into the conductivity cell
2
in the nitrogen glovebox. The temperature was controlled from
1
0 up to 70 °C in a Binder refrigerated incubator KB53 (E3.1).
alkyl halide resulting from the alkylation step were removed by
washing several times with hexane. The corresponding
About an hour was allowed for the sample to reach the desired
temperature prior to each measurement. The uncertainty in the
conductivities is approximately ±3%.
tetraalkylphosphonium hydroxide ([P444n][OH] or [P222n -
]
[
OH]) solutions were obtained by anion exchange, as described
A Mettler-Toledo differential scanning calorimeter (DSC)
above. The tetraalkylphosphonium hydroxide solutions were
then reacted with an equimolar amount of the heterocyclic
precursor (pyrrole-2-carbonitrile) at room temperature to
produce [P444n][2-CNPyr] or [P222n][2-CNPyr] along with
water. All ILs were dried at 60 °C under reduced pressure for at
least 48 h before use. A Brinkman 831 Karl Fischer coulometer
was used to determine water content, and all samples contained
(model DSC822e) was used to determine melting points (T
and glass transition temperatures (T ). The samples were
placed in a Mettler-Toledo standard 40 μL aluminum crucible
with lid, and heated under N from −120 to 110 °C at a heating
rate of 10 °C min . The onset point and midpoint were used
to determine T and T , respectively. The uncertainties in the
reported T and T values are approximately ±1 °C.
m
)
g
2
−1
m
g
m
g
1
lower than 1000 ppm water. H NMR (Varian INOVA-500)
The thermal stability of the ILs was determined with a
thermal gravimetric analyzer (TGA) (Mettler Toledo TGA/
SDTA 851e/SF/1100 °C). The nitrogen gas flow rate was
controlled by a Gilmont Industries PMR1-010750 flowmeter at
3
1
and P NMR (Varian INOVA-300) spectroscopy with
deuterated DMSO-d (99.9 atom % D, Sigma-Aldrich) as a
6
solvent were used for IL analysis. Furthermore, bromide
content was analyzed with a DIONEX DX-500 ion chromato-
graph was used to determine the bromide content, which was
below 20 ppm in all cases. The purity of all ILs in this study is
estimated to be greater than 98 mol % based on NMR
−1
50 mL min . A 20 mg aliquot of sample in a 40 μL aluminum
−1
crucible (without lid) was heated at 10 °C min , after being
initially dried in situ at 110 °C for 45 min. The decomposition
temperature (T ) in this study is the intersection of the
dec
1
spectroscopy. The H NMR spectral assignments can be found
baseline weight (after the drying step) and the tangent of the
weight versus temperature curve once decomposition starts.
The TGA has an accuracy of ±0.25 °C; however, when one
must manually identify the tangent point this increases the
uncertainty to ±2 °C. Both the TGA and the DSC use Mettler-
Toledo STARe software 12.10.
in the Supporting Information.
Characterization Methods. An oscillating U-tube densi-
tometer (Anton Paar model DMA 4500) was used to measure
the densities under ambient pressure. The temperature of the
densitometer was controlled from 10 to 70 °C with a precision
of ±0.01 °C. Considering the purity of ILs (>98%), the
uncertainty in the density measurements is approximately ±1 ×
CO Solubility Measurements. The CO solubility in
2
2
each IL was measured at room temperature (22 °C) using a
volumetric method with a custom-built instrument. The
−3
−3
10
g cm .
The viscosity measurements were carried out with an ATS
instrument consists of two main parts, a CO reservoir and a
2
Rheosystems Viscoanalyzer, which operates with a cone-and-
plate spindle. To prevent the absorption of atmospheric
reaction cell with known volumes (177 and 127 mL,
respectively). The temperature in each system was determined
using a type K thermocouple (Omega HH501DK, ±0.5 °C
uncertainty). The pressure in the reservoir and the reaction
vessel was monitored with a model CM dial pressure gauge (±2
mbar uncertainty) and a Heise model PM digital pressure
moisture and CO into the ILs, the sample chamber had
2
nitrogen gas (Praxair grade 4.8, 99.998%, H O < 3 ppm)
2
flowing over the sample throughout the measurements. The
measurements were performed at controlled temperatures with
Table 2. Thermal Properties and Parameters for Density, Viscosity, and Conductivity Data
thermal property
density
viscosity
conductivity
Tm
Tg
Tdec
a
b
ηo
α
Tg
σo
B
Tg
g cm−
3
10−4 g cm−3 K−1
cP
K
K
mS cm−1
K
K
cation
K
K
K
a
[
[
[
[
[
[
[
P2222]+
282, 292
312
na
576
578
567
570
582
588
589
0.389
0.390
0.042
0.065
0.185
0.177
0.180
593
583
195
199
167
172
181
183
175
P2224 +
]
193
196
1.151
1.122
1.115
1.084
1.075
1.073
−5.48
−5.58
−5.67
−5.68
−5.70
−5.78
1397
802
1331
154
72
777
812
180
181
164
197
196
196
+
a
P2228
]
na
1110
1063
897
P22212]+
215, 233
na
a
1165
743
P44412 +
]
na
a
200
+
a
P44418
]
]
278
na
920
730
+
a
P66614
na
196
904
60
741
a
na = a melting point or glass transition temperature does not exist for this IL.
1
1809
DOI: 10.1021/acs.jpcb.5b05733
J. Phys. Chem. B 2015, 119, 11807−11814

The Journal of Physical Chemistry B
Article
indicator (±0.5 mbar uncertainty), respectively. Approximately
indicated by the similar slopes of the linear fits. The densities
of all ILs at room temperature are less than that of water,
ranging in values from 0.903 to 0.990 g cm− for [P66614][2-
CNPyr] and [P2224][2-CNPyr], respectively. The density
2
g of pure IL sample was transferred to the reaction cell under
3
a nitrogen environment. Then the reaction cell was evacuated
using a vacuum pump (Welch Gem 8890), and CO was fed
2
+
from the reservoir into the reaction cell. The amount of CO2
that entered the reaction cell was calculated based on the
pressure drop in the reservoir. Upon the activation of the stir
bar, the pressure of the reaction cell decreased rapidly,
confirming the absorption of CO2 gas into the IL. The
temperature and pressure in the reaction cell were logged until
the pressure no longer changed (vapor−liquid equilibrium).
increased with decreasing cation alkyl chain length ([P66614]
+
+
+
+
+
< [P44418] < [P
] < [P
] < [P ] < [P ] ) without
44412
22212 2228 2224
exception. This trend is consistent with previously studied
imidazolium salts, in which an increase in the alkyl chain length
on the imidazolium ring systematically reduced the den-
34−38
sity.
Unlike trends observed for viscosity and conductivity,
the density is rather insensitive to the symmetricity of the
cation, and is mainly dependent on the total number of carbons
The CO solubility in the IL at each pressure was determined
2
39
from the total pressure drop using the ideal gas law with the
in the alkyl substituents. The variation in the densities of the
[Pnnnn][2-CNPyr] ILs is much larger than that of various
3
3
Lee−Kesler correlation. The uncertainty in the CO solubility
2
4
measurement is ±0.02 mol CO /IL.
[P66614][AHA] ILs from our previous paper, and this is
2
primarily due to significantly larger variation in the free volume
RESULTS AND DISCUSSION
(
and molecular weight) among different phosphonium cations
■
Thermal Properties. The DSC results are listed in Table 2,
along with the thermal stabilities of the [Pnnnn][2-CNPyr] ILs
obtained from the TGA. No endothermic peak, corresponding
in this study. The molar volume increased with increasing
cation size, as shown in Figure S1 in the Supporting
Information. In addition, the effective size of the ions were
estimated by calculating van der Waals volumes, as well as by
optimizing the ion structure using Density Functional Theory
(DFT) calculations. From the comparison between the molar
volume from the density (the total volume occupied by an IL
system) and van der Waals/DFT volume (sum of volume
occupied by each ion), it is clear that the ILs with larger cations
have significantly greater free volume (see Table S5 and Figure
S2 for details).
+
+
to a melting point was observed for [P ] , [P ] , and
2
228
44412
+
[
P66614] ILs in the cycle of the measurements. On the other
+
+
hand, ILs with [P2222] or [P ] cation showed distinct
2224
melting transition(s) around room temperature, due to a closer
packing of the ions facilitated by small, symmetric cations. For
the ILs that exhibited glass transition temperatures, the phase
transition occurred in the range of −70 °C. The TGA results
confirm that most of the [P ][2-CNPyr] ILs have excellent
nnnn
thermal stability around or above 300 °C at a scan rate of 10 °C
Viscosity. The temperature dependence of the viscosity for
the seven 2-cyanopyrrolide ILs is presented in Figure 3 and
−
1
19
min . As previously reported by Brown et al., the
decomposition temperature, T , seems to be independent of
dec
the alkyl-chain length on the cation. We have previously shown
that the thermal stability of the AHA-based ILs depends more
strongly on the choice of the anion species (i.e., the basicity of
4
the anion).
Density. The densities of the [Pnnnn][2-CNPyr] ILs as a
function of temperature are shown in Figure 2 and given in
Figure 3. Viscosity of [Pnnnn][2-CNPyr] ILs as a function of
1
0
temperature. (a) Values from Gurkan et al.
given in tabular form in the Supporting Information. The two
ILs with the shortest alkyl chains, [P2222][2-CNPyr] and
[
P2224][2-CNPyr], both have melting points near or slightly
above room temperature, resulting from their highly symmetric
cations; however, both remained subcooled liquids during the
course of the viscosity measurements. The viscosity increased
in the following order: [P ] < [P ] < [P ] < [P22212]
< [P66614] ≅ [P
the length of the alkyl-chain and the viscosity is consistent with
Figure 2. Density of [Pnnnn][2-CNPyr] as a function of temperature.
10
(
a) Values from Gurkan et al.
+
+
+
+
2
222
2224
2228
+
+
+
tabular form in the Supporting Information. As observed in
Figure 2, the densities decrease linearly with increasing
temperature, and the best-fit parameters for the linear
dependence of density with temperature (ρ = a + bT) are
listed in Table 2. The density of [P2222][2-CNPyr] was not
measured, anticipating the solidification of the sample in the
densitometer. All the [Pnnnn][2-CNPyr] ILs in this study had
similar volume expansion with increasing temperature, as
] < [P
] . Such relationship between
44412
44418
the trends exhibited by previously reported phosphonium-
19,39
34
40
based,
imidazolium-based ILs, and alkylimidazoles. It is
apparent that the viscosity is strongly correlated with the length
of the n-alkyl substituents on the cation: the viscosity of the
+
+
[P2222] IL is five times lower than the [P
] IL at room
44418
temperature. While the hydrogen-bond strength between [2-
1
1810
DOI: 10.1021/acs.jpcb.5b05733
J. Phys. Chem. B 2015, 119, 11807−11814

The Journal of Physical Chemistry B
Article
CNPyr] and [Pnnnn]+ (between the nucleophilic site of the
anion and the proton on the α-carbon of the cation) is expected
to be similar for all of the ILs, the strength of van der Waals
forces increases gradually with an elongation of the alkyl chains
on the phosphonium cation, eventually increasing the friction
and slowing down the dynamics of the ions. In addition,
Tsunashima et al. attributed the decrease in viscosity upon
shortening of the alkyl substituents to smaller Stokes radii of
−
create static images of anion/cation association/dissociation or
aggregation. Obviously, ionic conductivity and viscosity are
transport properties so they depend on the dynamics of the
system. Thus, as mentioned in the Introduction, we interpret
ionicity more in terms of how long-lived the association is
between the cation and anion pair, or between aggregates of
ILs.
As shown in Figure 5, all [Pnnnn][2-CNPyr] ILs fall below the
ideal line as expected. While the Walden plot can only be used
4
1
the cations. [P44418][2-CNPyr] has higher viscosity than
[
(
P66614][2-CNPyr] despite a slightly lower molecular weight
MW = 546.9 and 574.9 g mol , respectively). The
−
1
temperature dependency of viscosity may be described by a
42
Vogel−Fulcher−Tammann (VFT) equation:
⎛
⎝
⎞
g ⎠
α
⎜
⎟
⎟
η = η exp⎜
o
T − T
(1)
where η = viscosity in cP, T = absolute temperature, T = glass
g
transition temperature in absolute units, and η and α are
o
adjustable parameters. In this study, Tg was used as an
additional fitting parameter instead of applying experimentally
measured values since several ILs lack T values. The dotted
g
lines in Figure 3 represent the VFT model fit using the
parameters summarized in Table 2.
Conductivity. The conductivities of [Pnnnn][2-CNPyr] ILs
were measured from 10 to 70 °C, and the results are shown in
Figure 4 and in tabular form in the Supporting Information.
Figure 5. Walden plot of [Pnnnn][2-CNPyr] ILs.
as a qualitative tool for assessing relative ionicity, it is apparent
that the distance from the ideal line increases systematically
with increasing size of the phosphonium cation. ILs with
smaller cations have lower viscosities, and they have relatively
higher conductivities for these viscosities. On the other hand,
for ILs with larger cations, the cations and anions must be less
dissociated (i.e., have longer cation/anion contact times) as
they exhibit lower conductivity for their particular viscosity
20
range. Per the description given by Angell and co-workers,
P2224][2-CNPyr] would be considered a relatively “good” IL,
[
as opposed to [P66614][2-CNPyr] being a “poor” IL in
comparison to the ILs with smaller cations. We believe this
occurs primarily because a cation with longer hydrocarbon
substituents experiences more extensive van der Waals
interactions, leading to decreased mobility of the ions that
predominantly governs the ionic conductivity. Similar trends
h a v e b e e n o b s e r v e d f o r i m i d a z o l i u m b i s -
Figure 4. Conductivity of [Pnnnn][2-CNPyr] ILs as a function of
temperature.
4
3
Ionic conductivities of the ILs are inversely proportional to
their viscosities since ion mobility for conduction is
predominantly restricted by the viscosity: the increasing order
of ionic conductivity shown in Figure 4 is the reverse of the
viscosity trend presented in Figure 3. The VFT equation was
(trifluoromethanesulfonyl)amide and tri-n-butylalkylphos-
35
phonium-based ILs. Furthermore, an anion residing in close
proximity to the center of a cation may be “trapped” by the
long alkyl chains of neighboring phosphonium cations as
hydrocarbons align easily with respect to one another and form
44
also used for modeling the conductivity data (η and α in eq 1
aggregates or clusters to some extent. Such extensive ion
paring or aggregation lowers an ion’s ability as a charge carrier,
which may contribute to decreased ionic mobility and lower
conductivities relative to their viscosities. In general, the slope
of each line representing different ILs is approximately unity,
indicating that the conductivity−fluidity relationship in all ILs
remains roughly constant with increasing temperature. In an
attempt to provide explicit allowance for differences in ion sizes,
we also constructed an adjusted Walden plot (see Figure S3), as
o
−
1
are replaced by σ [mS cm ] and −B [K], respectively), and
o
Table 2 shows the best-fit parameters.
Walden Plot. The interdependence of viscosity and ionic
conductivity was analyzed using a Walden plot. A Walden plot
emphasizes differences in conductivity of various ILs as a
function of viscosity. Deviations from the “ideal” line, which
runs from corner to corner and is constructed from data for
dilute KCl aqueous solutions (generally considered to be fully
dissociated ions), is a qualitative measure of the ionicity of an
22
suggested by MacFarlane et al. While the relative distance
between each IL appears to be slightly smaller after the size
correction, the trend remains the same.
2
0,22
IL.
The further an IL falls below this “ideal” line, the lower
its ionicity. Unfortunately, the concept of “ionicity” tends to
1
1811
DOI: 10.1021/acs.jpcb.5b05733
J. Phys. Chem. B 2015, 119, 11807−11814

The Journal of Physical Chemistry B
Article
CO Solubility. The CO solubilities in ILs with function-
Table 3. [Pnnnn][2-CNPyr] Molecular Weight, CO Solubility
2
2
2
alized anions are primarily dependent on the enthalpy of
(PCO2 = 0.15 bar), Equilibrium Constants, and Standard
Entropy of Reaction at 22 °C
reaction (ΔH° or the strength of anion−CO interaction),
rxn
2
4
between a reactive anion and a CO molecule, and the
2
contribution from the cation is often considered negligible. Our
molecular
weight
CO
solubility
2
a
b
−
(PCO2 = 0.15 bar)
Keq
ΔS°
rxn
previous work has shown that the [2-CNPyr] anion is capable
10
of reacting with CO in an equimolar ratio, as depicted in
mole ratio
mole CO2
per IL
gravimetric
2
ratio
Scheme 1.
g mol−1
−1
unitless J mol−1 K−1
cation
mmol g
[
[
[
[
[
[
P2224]+
266.4
322.5
378.6
462.8
546.9
575.0
0.80
0.80
0.73
0.72
0.64
0.62
3.00
2.45
1.92
1.59
1.17
1.08
43.3
40.1
24.4
26.1
15.3
13.7
−121
−122
−124
−125
−129
−131
Scheme 1. Complexation Mechanism of CO with [Pnnnn][2-
CNPyr]
2
P2228]+
P22212]+
P44412]+
P44418]+
P66614]+
a
Henry’s constants and C were fixed at 62 bar and 0.92, respectively.
ΔH° was fixed at −45 kJ mol .
3
b
−1
rxn
PCO2/H
1 − PCO
K ·PCO2·C
The CO equilibrium isotherms for all of the ILs investigated
here were obtained at room temperature; [P2222][2-CNPyr]
eq
3
2
z =
+
2/H
1 + K ·P
eq CO2
(2)
solidified shortly after exposure to CO , and is not included in
2
−
Figure 6. Since all ILs in this study have the same [2-CNPyr]
where Z is the molar ratio of CO absorbed per IL, P is the
2 CO
2
CO pressure in bar, H is the Henry’s law constant in bar, K is
2
eq
the reaction equilibrium constant, and C is the fraction of
3
active IL (≈ 1). The Henry’s law constant (H) is a measure of
the physical solubility in ILs and is fixed at 62 bar, an
experimentally measured value for [P66614][2-CNPyr] at 22
10,44
°
C.
While we would expect the Henry’s law constants to be
marginally higher (i.e., lower physical solubility) for the ILs
with shorter alkyl chains on the cation because of reduced free
volume, the physical solubility is so low at these low partial
pressures of CO that the value of the Henry’s law constant
2
does not noticeably affect the fit of the data. In fact, the physical
solubility is expected to contribute only 1−2% of the total CO
2
uptake at 1 bar. C has also been fixed at 0.92, as all the
3
isotherms tend to level off around 0.92 mol CO /IL. Using the
2
Levenberg−Marquardt method, the reaction equilibrium
constant, K , was then fit to the experimental CO solubility
eq
2
Figure 6. CO solubility in [Pnnnn][2-CNPyr] at 22 °C. (a) Values
from Gurkan et al.
data for each IL, and the resulting values are summarized in
Table 3. The model fits the data very well for all of the ILs, as
shown in Figure 6.
2
1
0
Since the ΔH° between IL and CO is primarily determined
rxn
2
by the reaction between the anion and a CO molecule, as
2
anion, the CO binding mechanism (see Scheme 1) and,
2
depicted in Scheme 1, all ILs in this study should lend
−
1
therefore, the CO solubility, are expected to be similar. As
2
themselves to similar ΔH° (= −45 kJ mol , experimentally
rxn
10
shown in Figure 6, however, the CO solubility in [Pnnnn][2-
determined value reported by Gurkan et al. ). On the other
2
CNPyr] ILs varies with the variation in the n-alkyl substituents
on the phosphonium cation: the smaller the cation is, the
higher is the CO2 solubility. Especially in terms of the
hand, it is possible that the entropy of reaction (ΔS° ) may
rxn
vary for the different [P ][2-CNPyr] ILs, mainly due to
nnnn
different degrees of reorganization that must take place for the
anion to react with CO . We attribute this to the conformation
gravimetric ratio of CO uptake, the ILs with small cations have
2
2
of the cation, in terms of how it is associated with the anion and
how long-lived that association may be. Using a general
remarkably higher CO solubility owing to their low molecular
2
weight. At 0.15 bar (typical CO2 partial pressure in
thermodynamic relation ΔG° = −RT ln(K ) = ΔH°
-
rxn
eq
rxn
postcombustion flue gas), the gravimetric CO solubility
2
−1
TΔS
r
xn, where R is gas constant, the parameter ΔS°xn can be
r
increases from 1.08 to 3.00 mmol g as the cation is changed
from [P66614] to [P ] (see Table 3). Note that if the
determined from known values of K and ΔH° (rearranging
+
+
eq
rxn
2224
the above equation gives ΔS° = R ln(K ) + ΔH° /T). In
rxn
eq
rxn
physical dissolution of CO was significant, which is not the
2
other words, ΔS° can be regarded as a dependent variable that
rxn
case at these low partial pressures (see below), one would
determines the equilibrium shift between the CO -binding
2
expect the solubility to be higher in the ILs with the longer
reaction, assuming ΔH° is constant for a given anion. As
rxn
4
5−47
chains.
shown in Table 3, it is clear that the magnitude of ΔS°
rxn
The lines that go through the experimental data in Figure 6
increases (in absolute values) with increasing size of the cation.
−
1
−1
are obtained using a Langmuir-type model (eq 2) described in
ΔS° of −131 J mol
K
for [P66614][2-CNPyr] is in excellent
rxn
4
,48
our previous work:
agreement with the experimental value obtained from temper-
1
1812
DOI: 10.1021/acs.jpcb.5b05733
J. Phys. Chem. B 2015, 119, 11807−11814

The Journal of Physical Chemistry B
Article
1
0
ature-dependent CO solubility data. The increased magni-
exhibited by ILs with smaller cations, and we attribute this to a
smaller entropy change for the reaction, which is related to the
higher ionicity (less association and aggregation, as well as
shorter interaction times) of the ILs with smaller cations. This
suggests that the choice of cation has a nontrivial effect on the
2
tude of ΔS° for larger cations might be interpreted as greater
rxn
molecular reorientation during the CO capture reaction, which
2
eventually shifts the equilibrium toward the reactant side
compared to the cases in which the cations (therefore,
magnitude of ΔS° ) are relatively smaller. The increasing free
CO solubility in addition to being primarily responsible for the
physical properties. While the anion is still primarily affiliated
rxn
2
volume with increasing cation size, as described in Figure S2,
also corroborates such argument.
with the CO binding reaction, it was possible to achieve
2
Interpreting the CO uptake curves in this way, it is clear that
the standard entropy change for the reaction of the anion with
significantly lower viscosity and remarkably higher CO2
solubility in gravimetric ratio simply by reducing the size of
2
CO is related to the “ionicity” of the various [Pnnnn][2-CNPyr]
the cation, and this offers great potential for CO separation
using ILs.
2
2
ILs. As discussed in the previous section, the qualitative
measure of ionicity for [Pnnnn][2-CNPyr] ILs from the Walden
plot indicated that the ILs with smaller cations are less
associated (i.e., shorter-lived interactions) compared to the ILs
with larger cations. While the strength of the primary
intermolecular interaction between the alpha hydrogen of the
cation and the nucleophilic site of the anion should remain the
same for all ILs in this study, the degree of relative ionicity (ion
association/longevity of the interactions which could include IL
aggregates) varies significantly with the extent of van der Waals
interactions. ILs with relatively low ionicity are likely to have a
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jpcb.5b05733.
■
*
S
Details on the NMR characterization for each IL
synthesized; density, viscosity, conductivity, CO sol-
2
ubility, and molar volume data in tabular form. (PDF)
AUTHOR INFORMATION
Corresponding Author
*Tel: (574) 631-5847. Fax: (574) 631-8366. E-mail: jfb@nd.
reduced chance of reacting with a CO molecule because of the
2
■
presence of ion pairs or clusters. In addition, anions are
attracted to positive charge density, which is positioned at the
center of the phosphonium cation, to take advantage of
favorable Coulombic interactions. This could cause anions to
be “trapped” by the extensive alkyl chains, particularly when the
alkyl substituents on the phosphonium cation is sufficiently
edu.
Notes
The authors declare no competing financial interest.
44
large relative to the size of the anion. The anions that are
confined by long alkyl chains may interact more closely and
strongly with the pairing cation, reducing the reactivity of the
anion toward CO2.
ACKNOWLEDGMENTS
This material is based upon work supported by the Department
of Energy under Award Number DE-FC26-07NT43091
■
(
(
National Energy Technology Laboratory) and AR0000094
Advanced Research Projects Agency−Energy. We thank
We recognize that an alternative interpretation of the CO2
uptake curves would be that interaction between the cation and
anion affects the enthalpy of the reaction between the anion and
Megan Thedwall for density measurements.
CO . While entirely reasonable, we currently favor the
2
REFERENCES
interpretation in terms of the entropy of reaction because of
its consistency with the viscosity, conductivity, and “ionicity”
data, as well as the CO2 uptakes. We note that we,
unfortunately, are not able to experimentally measure the
enthalpy of reaction for the ILs with different cations to the
accuracy that would be needed to distinguish between these
■
(
1) Brennecke, J. F.; Gurkan, B. E. Ionic Liquids for CO Capture
and Emission Reduction. J. Phys. Chem. Lett. 2010, 1, 3459−3464.
2) Gurkan, B. E.; de la Fuente, J. C.; Mindrup, E. M.; Ficke, L. E.;
2
(
Goodrich, B. F.; Price, E. A.; Schneider, W. F.; Brennecke, J. F.
Equimolar CO Absorption by Anion-Functionalized Ionic Liquids. J.
2
Am. Chem. Soc. 2010, 132, 2116−2117.
two interpretations. This is because at a fixed value of ΔS° , a
rxn
(3) Goodrich, B. F.; de la Fuente, J. C.; Gurkan, B. E.; Zadigian, D. J.;
Price, E. A.; Huang, Y.; Brennecke, J. F. Experimental Measurements
of Amine-Functionalized Anion-Tethered Ionic Liquids with Carbon
Dioxide. Ind. Eng. Chem. Res. 2011, 50, 111−118.
−
1
range of just 3 kJ mol in ΔH° would be needed to explain
rxn
the variation in CO uptake shown in Figure 6.
2
(
4) Seo, S.; Quiroz-Guzman, M.; DeSilva, M. A.; Lee, T. B.; Huang,
Y.; Goodrich, B. F.; Schneider, W. F.; Brennecke, J. F. Chemically
Tunable Ionic Liquids with Aprotic Heterocyclic Anion (AHA) for
CO Capture. J. Phys. Chem. B 2014, 118, 5740−5751.
(5) Wang, C.; Luo, X.; Luo, H.; Jiang, D.-E.; Li, H.; Dai, S. Tuning
CONCLUSION
In an attempt to develop ILs with both high CO solubility and
low viscosity, and in order to examine the effect of the cation
■
2
2
size on physical properties and CO solubilities, a series of
2
tetraalkylphosphonium 2-cyanopyrrolide ([Pnnnn][2-CNPyr])
ILs were synthesized and characterized. This work shows that
the physical properties of [Pnnnn][2-CNPyr] ILs can be readily
tuned by adjusting the size of the hydrocarbon substituents on
the cation. Especially, the viscosity was reduced significantly by
the Basicity of Ionic Liquids for Equimolar CO Capture. Angew.
2
Chem., Int. Ed. 2011, 50, 4918−4922.
(6) Wang, C.; Luo, H.; Li, H.; Zhu, X.; Yu, B.; Dai, S. Tuning the
Physicochemical Properties of Diverse Phenolic Ionic Liquids for
Equimolar CO Capture by the Substituent on the Anion. Chem. - Eur.
2
−
J. 2012, 18, 2153−2160.
pairing the [2-CNPyr] anion with a smaller tetraalkylphos-
(
7) Bates, E. D.; Mayton, R. D.; Ntai, I.; Davis, J. H. CO Capture by
2
phonium cation (viscosity at room temperature decreases by 5-
fold when [P44418]+ was replaced with [P2222] ). The [Pnnnn][2-
CNPyr] ILs with a larger phosphonium cation showed a
relatively low degree of ionicity compared to those with smaller
cation due to strong van der Waals interactions that lead to ion
a Task-Specific Ionic Liquid. J. Am. Chem. Soc. 2002, 124, 926−927.
+
(8) Zhang, Y.; Zhang, S.; Lu, X.; Zhou, Q.; Fan, W.; Zhang, X. Dual
Amino-Functionalised Phosphonium Ionic Liquids for CO Capture.
2
Chem. - Eur. J. 2009, 15, 3003−3011.
(9) Gutowski, K. E.; Maginn, E. J. Amine-Functionalized Task-
Specific Ionic Liquids: a Mechanistic Explanation for the Dramatic
pairing or formation of aggregates. Higher CO solubilities were
2
1
1813
DOI: 10.1021/acs.jpcb.5b05733
J. Phys. Chem. B 2015, 119, 11807−11814

The Journal of Physical Chemistry B
Article
Increase in Viscosity Upon Complexation with CO From Molecular
(29) Besnard, M.; Cabaco
̧
, M. I.; Chav
́
ez, F. V.; Pinaud, N. On the
2
Simulation. J. Am. Chem. Soc. 2008, 130, 14690−14704.
Spontaneous Carboxylation of 1-Butyl-3-Methylimidazolium Acetate
by Carbon Dioxide. Chem. Commun. 2012, 48, 1245−1247.
(30) Seo, S.; DeSilva, M. A.; Brennecke, J. F. Physical Properties and
(
10) Gurkan, B.; Goodrich, B. F.; Mindrup, E. M.; Ficke, L. E.;
Massel, M.; Seo, S.; Senftle, T. P.; Wu, H.; Glaser, M. F.; Shah, J. K.;
et al. Molecular Design of High Capacity, Low Viscosity, Chemically
CO
with Aprotic Heterocyclic Anions. J. Phys. Chem. B 2014, 118, 14870−
4879.
31) Gohndrone, T. R.; Lee, B. T.; DeSilva, M. A.; Quiroz-Guzman,
Reaction Pathway of 1-Ethyl-3-Methylimidazolium Ionic Liquids
2
Tunable Ionic Liquids for CO Capture. J. Phys. Chem. Lett. 2010, 1,
2
1
(
3
(
494−3499.
11) Wu, H.; Shah, J. K.; Tenney, C. M.; Rosch, T. W.; Maginn, E. J.
M.; Schneider, W. F.; Brennecke, J. F. Competing Reactions of CO2
With Cations and Anions in Azolide Ionic Liquids. ChemSusChem
Structure and Dynamics of Neat and CO -Reacted Ionic Liquid
2
Tetrabutylphosphonium 2-Cyanopyrrolide. Ind. Eng. Chem. Res. 2011,
2
(
014, 7, 1970−1975.
5
(
0, 8983−8993.
32) Hamilton, P. A.; Murrells, T. P. Kinetics and Mechanism of the
12) Xie, W.; Xie, R.; Pan, W.-P.; Hunter, D.; Koene, B.; Tan, L.-S.;
3
4
Reactions of PH With O( P) and N( S) Atoms. J. Chem. Soc., Faraday
Vaia, R. Thermal Stability of Quaternary Phosphonium Modified
3
Trans. 2 1985, 81, 1531−1541.
Montmorillonites. Chem. Mater. 2002, 14, 4837−4845.
(33) Lee, B. I.; Kesler, M. G. A Generalized Thermodynamic
(
13) Wooster, T. J.; Johanson, K. M.; Fraser, K. J.; MacFarlane, D. R.;
Correlation Based on Three-Parameter Corresponding States. AIChE
Scott, J. L. Thermal Degradation of Cyano Containing Ionic Liquids.
J. 1975, 21, 510−527.
Green Chem. 2006, 8, 691−696.
(34) Tokuda, H.; Hayamizu, K.; Ishii, K.; Susan, M.; Watanabe, M.
(
14) Cassity, C. G.; Mirjafari, A.; Mobarrez, N.; Strickland, K. J.;
Physicochemical Properties and Structures of Room Temperature
Ionic Liquids. 2. Variation of Alkyl Chain Length in Imidazolium
Cation. J. Phys. Chem. B 2005, 109, 6103−6110.
O’Brien, R. A.; Davis, J. H. Ionic Liquids of Superior Thermal Stability.
Chem. Commun. 2013, 49, 7590−7592.
(
15) Bradaric, C. J.; Downard, A.; Kennedy, C.; Robertson, A. J.;
(35) Yoshii, K.; Yamaji, K.; Tsuda, T.; Tsunashima, K.; Yoshida, H.;
Zhou, Y. Industrial Preparation of Phosphonium Ionic Liquids. Green
Chem. 2003, 5, 143−152.
Ozaki, M.; Kuwabata, S. Physicochemical Properties of Tri-N-
Butylalkylphosphonium Cation-Based Room-Temperature Ionic
Liquids. J. Phys. Chem. B 2013, 117, 15051−15059.
(
16) Seo, S.; Simoni, L. D.; Ma, M.; DeSilva, M. A.; Huang, Y.;
Stadtherr, M. A.; Brennecke, J. F. Phase-Change Ionic Liquids for
(36) Fredlake, C. P.; Crosthwaite, J. M.; Hert, D. G.; Aki, S. N. V. K.;
Postcombustion CO Capture. Energy Fuels 2014, 28, 5968−5977.
2
Brennecke, J. F. Thermophysical Properties of Imidazolium-Based
(
17) Eisinger, R. S.; Keller, G. E., II. Process for CO Capture Using
2
Ionic Liquids. J. Chem. Eng. Data 2004, 49, 954−964.
Ionic Liquid That Exhibits Phase Change. Energy Fuels 2014, 28,
070−7078.
18) Gurkan, B. E.; Gohndrone, T. R.; McCready, M. J.; Brennecke,
(37) Fang, D. W.; Guan, W.; Tong, J.; Wang, Z. W.; Yang, J. Z. Study
7
(
on Physicochemical Properties of Ionic Liquids Based on Alanine
[
C mim][Ala] (n= 2,3,4,5,6). J. Phys. Chem. B 2008, 112, 7499−7505.
n
J. F. Reaction Kinetics of CO Absorption in to Phosphonium Based
2
(38) Tariq, M.; Forte, P. A. S.; Gomes, M. F. C.; Lopes, J. N. C.;
Anion-Functionalized Ionic Liquids. Phys. Chem. Chem. Phys. 2013, 15,
Rebelo, L. P. N. Densities and Refractive Indices of Imidazolium- and
Phosphonium-Based Ionic Liquids: Effect of Temperature, Alkyl
Chain Length, and Anion. J. Chem. Thermodyn. 2009, 41, 790−796.
(39) Kagimoto, J.; Taguchi, S.; Fukumoto, K.; Ohno, H. Hydro-
phobic and Low-Density Amino Acid Ionic Liquids. J. Mol. Liq. 2010,
7
(
796−7811.
19) Brown, P.; Gurkan, B. E.; Hatton, T. A. Enhanced Gravimetric
CO Capacity and Viscosity for Ionic Liquids with Cyanopyrrolide
2
Anion. AIChE J. 2015, 61, 2280−2285.
(
20) Xu, W.; Cooper, E. I.; Angell, C. A. Ionic Liquids: Ion
1
53, 133−138.
Mobilities, Glass Temperatures, and Fragilities. J. Phys. Chem. B 2003,
07, 6170−6178.
21) Yee, P.; Shah, J. K.; Maginn, E. J. State of Hydrophobic and
(40) Shannon, M. S.; Bara, J. E. Properties of Alkylimidazoles as
1
(
Solvents for CO Capture and Comparisons to Imidazolium-Based
2
Ionic Liquids. Ind. Eng. Chem. Res. 2011, 50, 8665−8677.
(41) Tsunashima, K.; Kawabata, A.; Matsumiya, M. Low Viscous and
Highly Conductive Phosphonium Ionic Liquids Based on Bis-
(Fluorosulfonyl)Amide Anion as Potential Electrolytes. Electrochem.
Commun. 2011, 13, 178−181.
(42) Vogel, H. The Temperature Dependence Law of the Viscosity
of Fluids. Phys. Z. 1921, 22, 645−646.
(43) Ueno, K.; Tokuda, H.; Watanabe, M. Ionicity in Ionic Liquids:
Correlation with Ionic Structure and Physicochemical Properties. Phys.
Chem. Chem. Phys. 2010, 12, 1649−1658.
(44) Fraser, K. J.; Izgorodina, E. I.; Forsyth, M.; Scott, J. L.;
MacFarlane, D. R. Liquids Intermediate Between “Molecular” and
Hydrophilic Ionic Liquids in Aqueous Solutions: Are the Ions Fully
Dissociated? J. Phys. Chem. B 2013, 117, 12556−12566.
(
22) MacFarlane, D. R.; Forsyth, M.; Izgorodina, E. I.; Abbott, A. P.;
Annat, G.; Fraser, K. On the Concept of Ionicity in Ionic Liquids. Phys.
Chem. Chem. Phys. 2009, 11, 4962−4967.
(
23) Xu, D.; Yang, Q.; Su, B.; Bao, Z.; Ren, Q.; Xing, H. Enhancing
the Basicity of Ionic Liquids by Tuning the Cation−Anion Interaction
Strength and via the Anion-Tethered Strategy. J. Phys. Chem. B 2014,
1
(
18, 1071−1079.
24) Holloczki, O.; Kelemen, Z.; Ko
́
̈
nczo
̈
l, L.; Szieberth, D.; Nyulas
́
zi,
L.; Stark, A.; Kirchner, B. Significant Cation Effects in Carbon
“
Ionic” Liquids: Liquid Ion Pairs? Chem. Commun. 2007, 0, 3817−
819.
45) Aki, S. N. V. K.; Mellein, B. R.; Saurer, E. M.; Brennecke, J. F.
Dioxide-Ionic Liquid Systems. ChemPhysChem 2013, 14, 315−320.
3
(
(
25) Rodríguez, H.; Gurau, G.; Holbrey, J. D.; Rogers, R. D. Reaction
of Elemental Chalcogens with Imidazolium Acetates to Yield
Imidazole-2-Chalcogenones: Direct Evidence for Ionic Liquids as
Proto-Carbenes. Chem. Commun. 2011, 47, 3222−3224.
High-Pressure Phase Behavior of Carbon Dioxide with Imidazolium-
Based Ionic Liquids. J. Phys. Chem. B 2004, 108, 20355−20365.
(46) Shariati, A.; Peters, C. J. High-Pressure Phase Behavior of
(
26) Gurau, G.; Rodríguez, H.; Kelley, S. P.; Janiczek, P.; Kalb, R. S.;
Systems with Ionic Liquids: II. the Binary System Carbon Dioxide+1-
Ethyl-3-Methylimidazolium Hexafluorophosphate. J. Supercrit. Fluids
Rogers, R. D. Demonstration of Chemisorption of Carbon Dioxide in
1,3-Dialkylimidazolium Acetate Ionic Liquids. Angew. Chem., Int. Ed.
2
(
004, 29, 43−48.
2
(
011, 50, 12024−12026.
47) Yunus, N. M.; Mutalib, M. I. A.; Man, Z.; Bustam, M. A.;
27) Shiflett, M. B.; Elliott, B. A.; Lustig, S. R.; Sabesan, S.; Kelkar, M.
Murugesan, T. Solubility of CO in Pyridinium Based Ionic Liquids.
2
S.; Yokozeki, A. Phase Behavior of CO In Room-Temperature Ionic
2
Chem. Eng. J. 2012, 189, 94−100.
Liquid 1-Ethyl-3-Ethylimidazolium Acetate. ChemPhysChem 2012, 13,
(48) Goodrich, B. F.; de la Fuente, J. C.; Gurkan, B. E.; Lopez, Z. K.;
1
(
806−1817.
Price, E. A.; Huang, Y.; Brennecke, J. F. Effect of Water and
Temperature on Absorption of CO By Amine-Functionalized Anion-
Tethered Ionic Liquids. J. Phys. Chem. B 2011, 115, 9140−9150.
28) Cabac o̧ , M. I.; Besnard, M.; Danten, Y.; Coutinho, J. A. P.
2
Carbon Dioxide in 1-Butyl-3-Methylimidazolium Acetate. I. Unusual
Solubility Investigated by Raman Spectroscopy and DFT Calculations.
J. Phys. Chem. A 2012, 116, 1605−1620.
1
1814
DOI: 10.1021/acs.jpcb.5b05733
J. Phys. Chem. B 2015, 119, 11807−11814