Are alkyl sulfate-based protic and aprotic ionic liquids stable with water and alcohols? A thermodynamic approach

Article abstract of DOI:10.1021/jp312241h

The knowledge of the chemical stability as a function of the temperature of ionic liquids (ILs) in the presence of other molecules such as water is crucial prior to developing any industrial application and process involving these novel materials. Fluid phase equilibria and density over a large range of temperature and composition can give basic information on IL purity and chemical stability. The IL scientific community requires accurate measurements accessed from reference data. In this work, the stability of different alkyl sulfate-based ILs in the presence of water and various alcohols (methanol, ethanol, 1-butanol, and 1-octanol) was investigated to understand their stability as a function of temperature up to 423.15 K over the hydrolysis and transesterification reactions, respectively. From this investigation, it was clear that methyl sulfate- and ethyl sulfate-based ILs are not stable in the presence of water, since hydrolysis of the methyl sulfate or ethyl sulfate anions to methanol or ethanol and hydrogenate anion is undoubtedly observed. Such observations could help to explain the differences observed for the physical properties published in the literature by various groups. Furthermore, it appears that a thermodynamic equilibrium process drives these hydrolysis reactions. In other words, these hydrolysis reactions are in fact reversible, providing the possibility to re-form the desired alkyl sulfate anions by a simple transesterification reaction between hydrogen sulfate-based ILs and the corresponding alcohol (methanol or ethanol). Additionally, butyl sulfate- and octyl sulfate-based ILs appear to follow this pattern but under more drastic conditions. In these systems, hydrolysis is observed in both cases after several months for temperatures up to 423 K in the presence of water. Therein, the partial miscibility of hydrogen sulfate-based ILs with long chain alcohols (1-butanol and 1-octanol) can help to explain the enhanced hydrolytic stability of the butyl sulfate- and octyl sulfate-based ILs compared with the methyl or ethyl sulfate systems. Additionally, rapid transesterification reactions are observed during liquid-liquid equilibrium studies as a function of temperature for binary systems of (hydrogen sulfate-based ionic liquids + 1-butanol) and of (hydrogen sulfate-based ionic liquids + 1-octanol). Finally, this atom-efficient catalyst-free transesterification reaction between hydrogen sulfate-based ILs and alcohol was then tested to provide a novel way to synthesize new ILs with various anion structures containing the alkyl sulfate group. ? 2013 American Chemical Society.

Full text of DOI:10.1021/jp312241h

Article
pubs.acs.org/JPCB
Are Alkyl Sulfate-Based Protic and Aprotic Ionic Liquids Stable with
Water and Alcohols? A Thermodynamic Approach
Johan Jacquemin,* Peter Goodrich,* Wei Jiang, David W. Rooney, and Christopher Hardacre
The School of Chemistry and Chemical Engineering/QUILL Research Centre, Queen’s University of Belfast, David Keir Building,
Stranmillis Road, Belfast BT9 5AG, Northern Ireland, U.K.
S
* Supporting Information
ABSTRACT: The knowledge of the chemical stability as a
function of the temperature of ionic liquids (ILs) in the presence
of other molecules such as water is crucial prior to developing any
industrial application and process involving these novel materials.
Fluid phase equilibria and density over a large range of
temperature and composition can give basic information on IL
purity and chemical stability. The IL scientific community requires
accurate measurements accessed from reference data. In this work,
the stability of different alkyl sulfate-based ILs in the presence of
water and various alcohols (methanol, ethanol, 1-butanol, and 1-octanol) was investigated to understand their stability as a
function of temperature up to 423.15 K over the hydrolysis and transesterification reactions, respectively. From this investigation,
it was clear that methyl sulfate- and ethyl sulfate-based ILs are not stable in the presence of water, since hydrolysis of the methyl
sulfate or ethyl sulfate anions to methanol or ethanol and hydrogenate anion is undoubtedly observed. Such observations could
help to explain the differences observed for the physical properties published in the literature by various groups. Furthermore, it
appears that a thermodynamic equilibrium process drives these hydrolysis reactions. In other words, these hydrolysis reactions
are in fact reversible, providing the possibility to re-form the desired alkyl sulfate anions by a simple transesterification reaction
between hydrogen sulfate-based ILs and the corresponding alcohol (methanol or ethanol). Additionally, butyl sulfate- and octyl
sulfate-based ILs appear to follow this pattern but under more drastic conditions. In these systems, hydrolysis is observed in both
cases after several months for temperatures up to 423 K in the presence of water. Therein, the partial miscibility of hydrogen
sulfate-based ILs with long chain alcohols (1-butanol and 1-octanol) can help to explain the enhanced hydrolytic stability of the
butyl sulfate- and octyl sulfate-based ILs compared with the methyl or ethyl sulfate systems. Additionally, rapid transesterification
reactions are observed during liquid−liquid equilibrium studies as a function of temperature for binary systems of (hydrogen
sulfate-based ionic liquids + 1-butanol) and of (hydrogen sulfate-based ionic liquids + 1-octanol). Finally, this atom-efficient
catalyst-free transesterification reaction between hydrogen sulfate-based ILs and alcohol was then tested to provide a novel way
to synthesize new ILs with various anion structures containing the alkyl sulfate group.
(trifluoromethylsulfonyl)imide ([NTf2]−) and the tris-
INTRODUCTION
■
(perfluoroalkyl)trifluorophosphate ([FAP]−) have been devel-
Due to their very low volatility under ambient conditions, ionic
liquids (ILs) are often classified as novel and green materials,1,2
oped. The stability of these anions is well-established and these
salts have found many applications in the fields of catalysis;
even if this classification does not take into account other
however, one major drawback is their expense.14 This has, in
properties such as toxicity and chemical or thermal stability.3 The
turn, led to the development of cheaper halide-free ILs, such as
knowledge, over a large temperature range, of the density of pure
ILs and their fluid phase equilibria in mixtures with other
molecular compounds reflects the purity and the chemical
those based on the alkyl sulfate, alkyl phosphate, and alkyl
carbonate anions. Of these, the alkyl sulfate ILs, of which the
ethyl analogue is industrially available, have received particular
stability of studied samples, and such information is crucial for
attention.15 Despite the potential applications provided, these
novel materials like ILs. It is very well documented by
ourselves4−6 and others7−11 that the presence of impurities like
ILs are also known to be hydrolytically unstable. Wasserscheid et
al. reported rapid hydrolysis of 1-ethyl-3-methylimidazolium
ethyl sulfate, [C1C2Im][EtSO4], by contaminating it artificially
with a large excess of water at 353 K.16 Ficke et al. have also
demonstrated that this IL decomposes to ethanol and 1-ethyl-3-
water and/or halogenated compounds dramatically changes the
physical properties of ILs. In that sense, the identification of the
impurities in the samples is of particular importance.
Furthermore, it was recently reported in the literature that
some ILs, for example, based on hexafluorophosphate ([PF6]−)
and tetrafluoroborate ([BF4]−) are not water-stable since they
hydrolyze easily in the presence of water.12,13 In this regard, more
stable hydrophilic fluorine-based anions such as bis-
Received: December 12, 2012
Revised: January 10, 2013
Published: January 15, 2013
© 2013 American Chemical Society
1938
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
methylimidazolium hydrogenosulfate, [C1C2Im][HSO4], over
extended time periods.17 This information is of great importance
in this field and must be taken into account in determining
reference data. Therein, our group recently published a critical
overview of our previously reported density studies,18 with more
than 1800 data available from the literature over temperature and
pressure ranges from 293 to 415 K and 0.1 to 40 MPa,
respectively.19 This investigation has clearly indicated again that
ethyl sulfate-based ILs are not water-stable and time-stable
compounds, since they hydrolyze at room temperature even in
presence of any trace amounts of moisture coming from the
atmosphere. Nevertheless, to date, limited information about the
reaction process between alkyl sulfate-based ILs with water is
available in the literature. It is well-known that reversible
hydrolysis reactions are often coupled with transesterifications
reactions. In that sense, Himmler et al.20 have reported a new
synthetic method to produce alkyl sulfate ionic liquids based on a
transesterification reaction between methyl sulfate or ethyl
sulfate ionic liquids and a long-chain alcohol (1-butanol, 1-
hexanol, 1-octanol, 2-methoxyethanol, 2-ethoxyethanol, 2-
butoxyethanol, diethyleneglycol monoethyl ether and diethyle-
neglycol monomethyl ether) in the presence of a Brønsted acid
catalyst (0.4 mol % methanesulfonic acid). More recently,
hydrolysis of methyl sulfate and ethyl sulfate was reported by
Ferguson et al.,21 from which it appears that a complete
hydrolysis reaction of methyl sulfate (ethyl sulfate) to the
hydrogen sulfate anion is obtained through the addition of 2
mol·L−1 of the methyl sulfate (ethyl sulfate) ILs in a aqueous
solution containing 40 mol % of sulfuric acid solution stirred at
373 K for 5 h. Nevertheless, to date, all the reported hydrolysis21
and transesterification20 reactions have been realized in the
presence of a Brønsted acid catalyst, which cannot explain the
low hydrolytic and transesterification stabilities of alkyl sulfate-
based ILs in the presence of only water or alcohols, respectively.
As density data is often used as a probe for the purity of the
investigated materials, the density of seven pure ILs {1-ethyl-3-
methylimidazolium hydrogen sulfate, [C1C2Im][HSO4]; 1-
ethyl-3-methylimidazolium methyl sulfate, [C1C2Im][MeSO4];
1-ethyl-3-methylimidazolium ethyl sulfate, [C1C2Im][EtSO4]; 1-
ethyl-3-methylimidazolium octyl sulfate, [C1C2Im][OcSO4]; 1-
butyl-3-methylimidazolium methyl sulfate, [C1C4Im][MeSO4];
tributylmethylammonium methyl sulfate, [N1444][MeSO4]; and
1-butyl-1-methylpyrrolidinium methyl sulfate, [C1C4Pyrro]-
[MeSO4]} was measured as a function of temperature from
(293 to 363) K at atmospheric pressure to first confirm previous
data published by our group for [C1C2Im][EtSO4]4,5,18,22 and to
evaluate these data with those already published by others.23−32
Second, an attempt was performed to estimate from the density
data the effective molar volume of each unknown ion by using the
methodology already published by our group,5 i.e., to extend as
the function of temperature the observed differences highlighted
by Costa et al.24 in the effective molar volume of the hydrogen
sulfate anion in comparison with those calculated in the case of
aprotic alkyl sulfate ILs. Third, the mutual solubility of
[C1C2Im][HSO4] with linear alcohols was investigated by
measuring the clear and cloud point temperatures using a
standard nephelometric method, and the chemical stability of
each studied mixture was then checked through NMR
measurements. Hydrolysis and transesterification reactions
between the studied ILs with water and with alcohols,
respectively, were then investigated by using three types of
reactoropen and sealed with or without headspacein order
to study the chemical and thermal stabilities of each binary
mixture with the temperature. Such stabilities were then verified
by using NMR spectroscopy. Finally, the hydrolysis/trans-
esterification reactions between alkyl sulfate anions and water/
alcohol were then investigated by using G3 Gaussian calculations
of each species involved as a function of temperature. These
simulations were then compared with the experimental results.
MATERIALS AND METHODS
■
Materials. Chemicals. The reagents used for the synthesis of
ILs were supplied by Aldrich [tributylamine (99%), 1-
butylpyrrolidine (99%), 1-butylimidazole (99%), 1-methylimi-
dazole (99%), dimethyl sulfate (99%), diethyl sulfate (99%),
benzyl alcohol (99%), butylamine (99%), N,N-diisopropylami-
noethanol (99%) and 2,2,3,3,4,4,4-heptafluoro-1-butanol
(98%)]. Toluene (HPLC grade) and ethyl acetate (HPLC
grade) were supplied by Riedel-de-Haen. All chemicals and
reagents were used without any further purification. 1-Ethyl-3-
methylimidazolium hydrogen sulfate and the corresponding
octyl sulfate were kindly supplied by BASF and Merck,
respectively, and were used after drying under high vacuum (1
Pa) at 333 K overnight.
General Procedure for the Synthesis of Alkyl Sulfate-Based
ILs. All ILs were prepared according to the following procedure:
dialkyl sulfate (0.2 mol) dissolved in anhydrous toluene (50 mL)
was added slowly dropwise to a precooled solution of amine (0.2
mol) in anhydrous toluene (150 mL). The mixture was
continually cooled in an ice-bath under nitrogen with care
being taken to maintain the reaction temperature below 298 K.
After complete addition of the dialkyl sulfate, the reaction
mixture was then stirred at room temperature for 4 h. The upper
organic phase of the resulting mixture was decanted, and the
lower IL phase was washed with ethyl acetate (3 × 20 mL). After
washing, the remaining ethyl acetate and any residual water was
removed by heating under reduced pressure. All ILs were
obtained with more than 98% of purity. Their structures were
confirmed by comparison of 1H NMR available in the literature
(see Supporting Information, Figures S1−S6).23−26
1-Ethyl-3-methylimidazolium Methyl Sulfate, [C1C2Im]-
1
[MeSO4]. H NMR (300 MHz, DMSO capillary, ppm): 9.38
(s, 1H, H-2), 7.57 (s, 2H, H-4 and H-5), 4.31 (q, J = 7.2 Hz, 2H,
NCH2CH3), 4.01 (s, 3H, NCH3), 3.70 (s, 3H, OCH3), 1.55 (t, J =
7.2 Hz, 3H, NCH2CH3).
1-Butyl-3-methylimidazolium Methyl Sulfate, [C1C4Im]-
1
[MeSO4]. H NMR (300 MHz, DMSO capillary, ppm): 9.06
(s, 1H, H-2), 7.77 (s, 1H, H-4), 7.68 (s, 1H, H-5), 4.14 (q, J = 7.4
Hz, 2 H, NCH2CH2), 3.87 (s, 3H, NCH3), 3.42 (s, 3H, OCH3),
1.67 (m, 2H, NCH2CH2), 1.09 (m, 2H, N(CH2)2CH2), 0.64 (t, J
= 7.1 Hz, 3H, N(CH2)3CH3).
1-Butyl-1-methylpyrrolidinium Methyl Sulfate, [C1C4pyrro]-
[MeSO4]. 1H NMR (300 MHz, DMSO capillary, ppm, δ): 3.66
(s, 3H, OCH3), 3.39 (m, 4H, NCH2), 3.23 (m, 2H, NCH2), 2.93
(s, 3H, NCH3), 2.11 (m, 4H, NCH2CH2), 1.66 (m, 2H,
NCH2CH2), 1.29 (sextuplet, J = 7.4 Hz, 2H, N(CH2)2CH2), 0.85
(t, 3H, J = 7.4 Hz, N(CH2)3CH3).
Tributylmethylammonium Methyl Sulfate, [N1444][MeSO4].
1H NMR (300 MHz, DMSO, ppm, δ): 3.39 (s, 3H, OCH3), 3.19
(m, 6H, NCH2), 2.96 (s, 3H, NCH3), 1.59 (m, 6H, NCH2CH2),
1.32 (q, J = 7.5 Hz, 6H, N(CH2)2CH2), 0.95 (t, J = 7.3 Hz, 9H,
N(CH2)3CH3).
1-Ethyl-3-methylimidazolium Ethyl Sulfate, [C1C2Im]-
1
[EtSO4]. H NMR (300 MHz, DMSO capillary, ppm): 9.39 (s,
1H, H-2), 7.73 (s, 2H, H-4 and H-5), 4.34 (q, J = 7.4 Hz, 2H,
NCH2CH3), 4.12 (q, J = 7.0 Hz, 2H, OCH2CH3), 4.06 (s, 3H,
1939
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
Table 1. Structure, Source, Molecular Purity, Molar Mass (MIL) of Each Selected Dried ILs Sample
a
b
Purity estimated by the supplier. Purity estimated by NMR analysis (see Supporting Information)
NCH3), 1.79 (t, J = 7.33 Hz, 3H, NCH2CH3), 1.17 (t, J = 6.9 Hz,
3H, OCH2CH3).
with the temperature. The temperature was measured using a
100 Ω platinum resistance thermometer (precision of 0.1 K
and accuracy of 0.5 K). On the basis of a statistical analysis of
the present data, the uncertainty of the cloud point is estimated as
to be close to 0.5 K.
After treating the ILs for 15 h at room temperature under
vacuum (lower than 1 Pa), the sample was considered as dried
and was then stored under nitrogen atmosphere to avoid water
contamination from atmosphere. Each IL was analyzed for water
content using Coulometric Karl Fischer titration (Mettler
Toledo DL31) prior to all measurements. These measurements
revealed very low levels of water, always lower than 250 ppm, i.e.,
lower than 0.025 wt %, except in the case of the [C1C2Im]-
[HSO4], where the water content in this IL was found to be close
to 2000 ppm, i.e. ∼0.2 wt %). The structure, source, and
molecular purity of each IL studied are reported in Table 1.
Experimental Methods. Density Measurements. The
experimental apparatus used to measure the densities was a U-
shape vibrating-tube densitometer (Anton Parr 4500) operating
in a static mode and was calibrated as previously described by our
group.5 The effect of viscosity on the density data as a function of
temperature was then taken into account by using viscosity data
already available in the literature for each pure IL.23−25
Liquid−Liquid Equilibria. The (liquid + liquid) miscibility gap
was measured using a standard nephelometric method.22,33 In
this technique, the miscibility is determined by finding visually
the clear and cloud points temperatures at which a phase change
occurs in a binary mixture of known composition (gravimetri-
cally determined). The relative uncertainty on the mass fraction
is estimated to be less than 2 × 10−4. All binary mixtures were
placed in a thermostated equilibrium cell. The clear point
temperature was then obtained by increasing the temperature at
which the mixture becomes entirely homogeneous. The cloud
point temperatures were then visually determined by decreasing
the temperature until the first droplet of the second phase
appeared. Different heating and cooling cycles have been
measured to quantify the difference in the read-outs during
each cycle, in order to investigate the stability of each mixture
Monitoring of Hydrolysis and Transesterification Reac-
tions. Chemical and thermal stabilities of the (IL + water) and
(IL + n-alcohol) binary mixtures were investigated as a function
of the temperature up to 423 K by using three different types of
reactors. Open and sealed with and without headspace reactor
types were selected in order to investigate the effect of the
presence or absence of the vapor phase on their stabilities. The
stabilities were analyzed by studying each sample without any
1
purification by H NMR and 13C NMR in the presence of a
sealed capillary tube containing deuterated D2O or DMSO using
a Bruker AVANCE 300 MHz NMR. The limits of detection for
NMR when using a deuterated solvent capillary tube are
approximately 2%.34
RESULTS AND DISCUSSION
■
Volumetric Properties of Pure ILs. Density measurements
of the pure ILs were carried out at temperatures ranging from
293.15 to 363.15 K at 0.1 MPa. The melting point of the pure
[N1444][MeSO4] has been reported as close to 335.15 K,35 but
we observed that this IL remains as a subcooled liquid for several
hours at room temperature (≈293 K). This behavior has been
already reported in the case of other alkyl sulfate-based
ILs,16,36,37 providing then the possibility to measure their
thermodynamic properties in this subcooled state. Herein, the
liquid density of [N1444][MeSO4] has been measured from 293
K, i.e., above its melting point temperature. The experimental
data for all the pure ILs are reported in Table 2 and shown in
Figure 1.
For the pure ILs, the density values typically vary from 1.05 to
1.37 g·cm−3 at 293 K and from 1.01 to 1.33 g·cm−3 at 363 K. As
1940
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
Table 2. Experimental Densities (ρ) and Molar Volumes (Vm) of Dried ILs as a Function of Temperature at Atmospheric Pressure
[C1C2Im][HSO4]
Vm/cm3·mol−1
[C1C2Im][MeSO4]
[C1C2Im][EtSO4]
Vm/cm3·mol−1
T/K
ρ/g·cm−3
ρ/g·cm−3
Vm/cm3·mol−1
ρ/g·cm−3
293.15
298.15
303.15
313.15
323.15
333.15
343.15
353.15
363.15
1.3689
1.3659
1.3628
1.3568
1.3508
1.3447
1.3387
1.3326
1.3266
152.12
152.46
152.80
153.48
154.16
154.86
155.55
156.27
156.97
1.2895
1.2861
1.2826
1.2758
1.2691
1.2624
1.2558
1.2493
1.2429
172.36
172.82
173.29
174.21
175.13
176.06
176.99
177.91
178.82
1.2411
1.2376
190.39
190.92
191.47
192.51
193.57
194.64
195.70
196.78
197.86
1.2341
1.2274
1.2207
1.2140
1.2074
1.2008
1.1942
[C1C2Im][OcSO4]
[C1C4Im][MeSO4]
T/K
ρ/g·cm−3
Vm/cm3·mol−1
ρ/g·cm−3
Vm/cm3·mol−1
293.15
298.15
303.15
313.15
323.15
333.15
343.15
353.15
363.15
1.0987
1.0953
1.0920
1.0854
1.0789
1.0724
1.0659
1.0595
1.0532
291.66
292.57
293.45
295.24
297.02
298.82
300.64
302.45
304.26
1.2144
1.2110
1.2077
1.2010
1.1944
1.1879
1.1814
1.1749
1.1685
206.12
206.70
207.27
208.42
209.57
210.72
211.88
213.05
214.22
[C1C4Pyrro][MeSO4]
Vm/cm3·mol−1
[N1444][MeSO4]
T/K
ρ/g·cm−3
ρ/g·cm−3
Vm/cm3·mol−1
293.15
298.15
303.15
313.15
323.15
333.15
343.15
353.15
363.15
1.1699
1.1670
1.1638
1.1575
1.1517
1.1455
1.1395
1.1337
1.1279
216.56
217.10
217.70
218.88
219.99
221.18
222.34
223.48
224.63
1.0591
1.0562
1.0531
1.0470
1.0409
1.0347
1.0286
1.0225
1.0165
294.10
294.91
295.77
297.50
299.24
301.03
302.82
304.63
306.42
[C1C2Im]+. These trends are in good agreement with the
variation in density with the IL structure previously reported in
the literature.38,39 The variation of the density with temperature
for the ILs studied is illustrated in Figure 1. The polynomial
equation used to fit this variation with temperature was of the
form
ρIL(T) = a + b(T) + c(T)2
(1)
with the parameters a, b, and c, and standard deviations σρ are
given in Table 3.
Molar volumes, Vm, of dried ionic liquids were then calculated
via eq 2
MIL
ρIL
Vm,IL
=
(2)
□
Figure 1. Experimental densities of dried (Δ) [N1444][MeSO4], ( )
◆
[C1C4Pyrro][MeSO4], (∇) [C1C4Im][MeSO4], ( ) [C1C2Im]-
where MIL and ρIL are the molar mass and the density of the
studied ionic liquid, respectively.
◇
)
○
●
[HSO4], ( ) [C1C2Im][MeSO4], ( ) [C1C2Im][EtSO4], and (
[C1C2Im][OcSO4] as a function of the temperature at 0.1 MPa. The
The calculated molar volumes are reported in Table 2 as
function of temperature up to 363.15 K. As expected, the IL
molar volumes increase with the temperature and the alkyl chain
length on the cation. This result is in good agreement with results
already available in the literature for different IL structures.38,39
Furthermore, from the work of Rebelo and co-workers,40 who
reported that the excess molar volumes of IL mixtures can be
considered as ideal (e.g., VE ≈ 0 cm3·mol−1), our group
lines correspond to the fit of the data by eq 1 reported in Table 3.
expected, the density is found to decrease with increasing the
alkyl chain length on both the alkyl sulfate anion and on the
imidazolium cation. Furthermore, Figure 1 also shows that the
density of the ILs based on the [MeSO4]− anion increased in the
+
following order: [N1444
]
< [C1C4Pyrro]+ < [C1C4Im]+
<
1941
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
Table 3. Correlation Parameters a, b, and c and Standard Deviation, σρ, of Eq 1 for the Density of Dried ILs as a Function of
Temperature Determined from Measurements between 293.15 and 363.15 K
a
ionic liquid
a/g·cm−3
104 × b/g·cm−3·K−1
107 × c/g·cm−3·K−2
104 × σρ /g·cm−3
[C1C2Im][HSO4]
[C1C2Im][MeSO4]
[C1C2Im][EtSO4]
[C1C2Im][OcSO4]
[C1C4Im][MeSO4]
[C1C4Pyrro][MeSO4]
[N1444][MeSO4]
1.54609
1.52627
1.46283
1.31622
1.43255
1.38160
1.23977
−6.045
−9.218
−8.278
−8.167
−8.278
−8.188
−6.203
0.00280
3.894
2.429
2.543
2.448
3.308
0.1457
0.34
0.27
0.56
0.33
0.28
1.1
0.66
a
σρ = (∑(ρei xp − ρicalc)2/n − v)0.5, where n is the number of experimental points and ν is the number of adjustable parameters.
Table 4. Parameters Associated with Eq 4 Used To Predict the Molar Volume of Selected ILs as the Function of Temperature and
0.1 MPa from the Effective Molar Volume of Ions Available in the Literature,5,41 along with the Relative Absolute Average
Deviations (RAAD; see eq 5) between the Calculated Molar Volumes and Those Presented during This Work (Table 2)
Effective Molar Volumes/cm3·mol−1
ion
M/g·mol−1
C0/cm3·mol−1
102 × C1/cm3·mol−1·K−1
105 × C2/cm3·mol−1·K−2
5
5
[C1C2Im]+
[C1C4Im]+
111.17
139.22
142.26
200.38
200.38
100.02
134.26
145.03
222.13
222.53
+6.955
+8.890
+10.00
+1.428
+14.38
+2.996
+4.304
−6.600
+340.9
+26.73
5
[C1C4Pyrro]+
5
+
[N1444
[N1444
]
]
a
+
a
[HSO4]−
97.07
111.10
125.12
209.28
52.442
72.397
90.681
191.40
−0.211
+2.692
+3.158
+13.66
+0.110
−16.82
−6.914
−31.62
5
[MeSO4]−
5
[EtSO4]−
[OcSO4]−
5
Comparison between Calculated and Experimental Molar Volumes from 293 to 363 K at 0.1 MPa
ionic liquid
100 × RAAD
ionic liquid
100 × RAAD
b
c
[C1C2Im][HSO4]
0.002
0.271
0.212
0.267
[C1C4Im][MeSO4]
0.083
0.141
0.483
0.005
c
c
[C1C2Im][MeSO4]
[C1C4Pyrro][MeSO4]
c
c
[C1C2Im][EtSO4]
[N1444][MeSO4]
c
b
[C1C2Im][OcSO4]
[N1444][MeSO4]
a
b
c
Calculated during this work. Calculated using new parameters presenting into this work (correlation). Calculated using original parameters5
(prediction). Bold text is used to highlight large deviation between experimental molar volumes and those calculated using original parameters.5
formulated a group contribution model (GCM) that was able to
predict, within an accuracy close to 0.5%, the volumetric
properties of pure ILs as a function of the temperature5 and the
pressure41 based on the following equations
ρ
− ρ
GCM
100
N
exp
100 × RAAD =
∑
ρGCM
(5)
with N being the total number of data used, ρGCM. the density
values calculated by our GCM,5 and ρexp the corresponding value
obtained from our experimental work. The RAAD can be
regarded as a measure of the accuracy of our GCM.
*
*
Vm(T*, p) = Vc (T*, p) + Va (T*, p)
(3)
2
*
V (T*, p) =
C(T*, p)i
∑
ion
i
Furthermore, by using the same methodology as reported
previously,5 new parameters that are able to determine the
effective molar volume of the [HSO4]− anion were then
calculated and also are reported in Table 4. For each of the
studied ILs, the accuracy of our GCM is better than 0.3%, except
in the case of [N1444][MeSO4], where a deviation close to 0.5% is
observed. The effective molar volume of the [N1444]+ has been
calculated previously5 for a temperature range from 297 to 333 K
(4)
i=0
where Vc* and V*a are the cation and anion effective molar
volumes, respectively, at the discrete temperature T* = T −
298.15 K, and Ci is the effective molar volume group contribution
parameters of each ion described by our group previously as a
function of the discrete temperature, T * at 0.1 MPa.5,41
In our original paper,5 except for the case of the [HSO4]−
anion, the effective molar volumes of all ions involved during this
work were already defined at 0.1 MPa as a function of
temperature. Herein, in order to test the accuracy of our
model, the volumetric properties of each of the IL studies were
predicted. These comparisons are presented in Table 4 together
with the percent relative absolute average deviation (100 ×
RAAD) between density values reported, herein, and those
calculated by our GCM,5 as a function of temperature up to
363.15 K.
using a density data set published by Anthony et al.42 for the IL
+
[N1444][NTf2]. Herein, the effective molar volume of [N1444
]
from 293 to 363 K was first recalculated by using the density data
set in the case of [N1444][MeSO4] reported in Table 2 with eqs 3
+
and 4. New effective molar volume fitting parameters of [N1444
]
are reported into Table 4 as a function of temperature up to 363
K and 0.1 MPa. Using this new set of parameters, a deviation up
to 0.3% is then obtained by comparing experimental42 and
calculated (using our GCM5) molar volumes for [N1444][NTf2]
1942
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
from 297 to 333 K at 0.1 MPa. Such deviations can be used to
identify the accuracy of these fitting parameters.
deviations within the experimental uncertainty. Similarly, the
density data set published by Ficke et al.31 in the case of
[C1C2Im][HSO4] is in excellent agreement with our (RAAD
close to 0.03%). In the case of [C1C2Im][MeSO4], two other
density data sets were compared with our data.31,32 For this IL,
deviations up to 0.14% and 3.54% are observed as a function of
temperature by comparing our density data set with those
provided by Ficke et al.31 and by Shekaari et al.32 In each case,
published density data are higher than ours. Even if an accuracy
of 0.14% can be explained by taking into account each
experimental uncertainty, in the case of the density of the
[C1C2Im][MeSO4] published by Shekaari et al.32 an extremely
high deviation, up to 3.54%, is observed with our data set. This
observation is consistent with that made by Costa et al.24 In the
case of the [C1C2Im][OcSO4], two density data sets are
available,24,29 and in each case, excellent agreement with our
data is observed. The density of [C1C4Im][MeSO4] as a function
of temperature is also available in the literature; the density
reported herein is in good agreement with data published by
Pereiro et al.23 (RAAD ≈ 0.08%) and by Krumelan et al.27
(RAAD ≈ 0.09%) for temperatures up to 343 K, as well as density
As the density of all the selected ILs, except in the case of
[N1444][MeSO4], was already described by other groups in the
literature, it was then possible to compare our reported
experimental data as a function of temperature.23−32 Further-
more, it was then also possible to re-evaluate the accuracy of our
previously published data in the case of [C1C2Im]-
[EtSO4].4,5,18,22 These comparisons are presented in Table 5
Table 5. Relative Absolute Average Deviations RAAD (see eq
6) between the Densities Presented in This Work and Those
of the Literature4,5,18,22−32
a
ionic liquid
Nb data
T/K
100 × RAAD
ref
[C1C2Im][HSO4]
71
6
293−363
293−343
293−363
293−343
298−328
293−352
293−363
293−352
293−343
293−363
293−343
293−333
293−363
293−333
293−343
293−308
298−308
298−343
−
0.008
0.025
0.003
0.141
3.54
24
31
24
31
32
4
[C1C2Im][MeSO4]
[C1C2Im][EtSO4]
71
6
4
7
0.266
0.119
0.265
0.253
0.021
0.008
0.067
0.023
0.061
0.082
0.089
0.251
0.014
−
data published at 298.15 K by Doman
́
ska et al.43 within a
8
5
7
18
22
30
26
29
24
29
23
27
28
25
−
deviation up to 0.1%. Nevertheless, in the case of [C1C4Im]-
[MeSO4], the density data presented herein are systematically
higher (RAAD ≈ 0.25%) than those published by Navia et al.28
Finally, in the case of [C1C2Im][EtSO4], excellent agreement is
observed with the density data set provided by Gomez et al.26
(RAAD ≈ 0.01%), Russina et al.29 (RAAD ≈ 0.07%), Jacquemin
et al.5 (RAAD ≈ 0.12%), and de Castro et al.30 (RAAD ≈ 0.02%)
as a function of temperature up to 363 K. The density data sets
published by our group in different publications4,18,22 are higher
than data presented herein. As already mentioned by our
group,19 this density difference can be explained by a hydrolysis
reaction that occurs over time between the ethyl sulfate anion
and water. Therefore, we recommend not to use density data sets
of [C1C2Im][EtSO4] published by our group in refs 4, 18, and 22
but to reference data reported herein or in refs 5, 19, 26, 29, and
30. Similarly, the large deviation observed between the density
data set presented herein with that published by Shekaari et al.32
in the case of [C1C2Im][MeSO4] can be also explained by a
hydrolysis reaction between the methyl sulfate anion and water.
Nevertheless, to date very little data on such a reaction for IL
media are available in the literature; for this reason, our group has
then decided to investigate this problem.
11
71
11
b
[C1C2Im][OcSO4]
[C1C4Im][MeSO4]
71
b
11
8
9
[C1C4Pyrro][MeSO4]
[N1444][MeSO4]
10
−
a
Each reported temperature range corresponds to the T interval where
b
each comparison was realized. Calculated from fitting parameters for
the linear temperature dependence of density from 293.15 to 333.15
K.29 Bold text is used to highlight RAAD values higher than the
expected uncertainty on density data.
together with the percent RAAD between density values
reported herein and those available in the literature,4518,22−32
for temperature ranging from 293 to 363 K at atmospheric
pressure. In each case, the reported percent RAAD was calculated
from the prior evaluation of each individual relative deviation
(RD) reported in this temperature range between data from the
literature with those presented herein according to the following
equation
Thermal Hydrolytic Stability of Alkyl Sulfate-Based ILs.
As the thermal hydrolytic stability of ILs is a key parameter that
determines their possible future applications, we studied the
stability of [C1C2Im][EtSO4] in the presence of 2−20 mol equiv
of water at 353 K, as reported in Table 6. Hydrolysis experiments
were performed with 0.01 mol of IL and 0.02−0.20 mol of water
in Teflon-lined autoclaves (20 mL capacity). A comparison of the
density of the IL−H2O mixtures measured in each case at 298.15
K and 0.1 MPa before (ρi) and after treatment (ρm) showed no
significant change, even after keeping the mixture at 353 K for 48
ρlit. − ρ
100
N
exp
100
N
100 × RAAD =
=
|RD|
∑
∑
ρ
exp
(6)
where N is the total number of data used, ρlit. the density values
from the literature, and ρexp the corresponding value obtained
from our experimental work. The RAAD can be regarded as a
measure of the accuracy of our experimental data and can be used
as a probe of the purity of each IL.
1
h. The observed hydrolytic stability was also confirmed by H
NMR. This behavior was surprising given that previous reports
had highlighted extremely rapid hydrolysis of the same IL, albeit,
in 100 mol equiv of water at 368 K.20
As reactions conducted at 353 K showed no evidence of
degradation, increases in reaction temperature, water content,
and time were investigated, the details of which are highlighted in
Table 7. In general, increases in temperature, water content, and
reaction times resulted in higher degrees of degradation for
According to the Table 5, density data sets of [C1C2Im]-
[HSO4], [C1C2Im][MeSO4], [C1C2Im][OcSO4] published by
Costa et al.24 as well as those published in the case of
[C1C4Pyrro][MeSO4] by Gonzalez et al.25 are in excellent
agreement with ours (RAAD close to 0.01% in each case), with
1943
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
reported in Table 7, entry 11. When this IL was dried under
vacuum overnight at 333 K, as reported in Figure S8 of the
Supporting Information, a large difference of viscosity is first
observed between pure [C1C2Im][EtSO4] sample and that
Table 6. Density Data of [C1C2Im][EtSO4] + H2O Mixtures
Measured at 298.15 K and 0.1 MPa before (ρi) and after (ρm)
a
Treatment
entry
mole ratio IL:H2O
t/h
ρi/g·cm−3
ρm/g·cm−3
1
measured after vacuum treatment. Furthermore, H NMR and
density measurements at 298.15 K (ρ = 1.3647 g·cm−3) were
almost identical to those reported for [C1C2Im][HSO4] in the
literature,24,31 as well as that of an authentic sample of the same
IL purchased from BASF measured in our laboratory (see Table
2). Further evidence for a thermodynamic equilibrium was
observed when a similar reaction conducted in a sealed vessel
with no headspace resulted in 27% conversion to the equivalent
[HSO4]−-based IL after 4 h. Thereafter, no further change was
observed upon heating for a further 24 h.44
1
2
3
4
5
6
7
1:2
1
1
1.2020
1.1670
1.1670
1.1670
1.1670
1.1237
1.0809
1.2024
1.1660
1.1664
1.1669
1.1674
1.1235
1.0807
1:5
1:5
3
1:5
24
48
1
1:5
1:10
1:20
1
a
For example, [C1C2Im][EtSO4] heated at 353 K with various molar
equivalents of water. Reactions were conducted in a sealed autoclave
with headspace.
A range of other ILs was also tested for thermal hydrolytic
stability at 423 K with 5 mol equiv of water, as reported in Table
8. All reactions were conducted in Teflon-sealed reactors. For the
methyl sulfate-based ILs (Table 8, entries 1−4) significant
conversions were observed, which were independent of the
choice of cation (see Figures S9−S12 in the Supporting
Information). Unsurprisingly, and as reported in Table 8 (entries
5−7), increasing the alkyl chain length of the sulfate anion
resulted in an increase in thermohydrolytic stability. Despite the
increase in stability, the octyl sulfate anion showed up to 32%
degradation under the experimental conditions after 2 months
(Table 8, entry 8), as shown in Figure S13 of the Supporting
Information. This unexpected result is in fact in contradiction
with all conclusions reported into the literature regarding the
hydrolytic stability of this anion.16
Transesterification Reaction of Hydrogen Sulfate-
Based ILs with Alcohols. On the basis of the knowledge that
alkyl sulfate ILs show different hydrolytic stabilities, we focused
our study on the stability of [C1C2Im][HSO4] IL in the presence
of methanol and ethanol. Experiments were again conducted in a
sealed vessel without a headspace using 2.5 mol equiv of each
alcohol at 423 K, the results of which are detailed in Table 9.
From the table, it can be concluded that the [HSO4]− IL is not
stable in the presence of either methanol and ethanol (see
Figures S14−S16 of the Supporting Information). This result is
in fact in good agreement with the thermodynamic equilibrium
observed during the investigation of the thermal hydrolytic
stability of [C1C2Im][EtSO4].
The thermal stability of the [C1C2Im][HSO4] was also
examined in the presence of other alcohols with longer chain
lengths such as 1-butanol or 1-octanol. Unexpectedly, we
observed that the [C1C2Im][HSO4] + 1-butanol and [C1C2Im]-
[HSO4] + 1-octanol binary mixtures are only partially miscible at
room temperature. From 1H NMR measurements,34 at 293.15 K
the 1-butanol mole fraction in the [C1C2Im][HSO4]-rich phase
is x1‑butanol = 0.375, while the [C1C2Im][HSO4] mole fraction is
close to 0.048 in the 1-butanol-rich phase. Similarly, at 293.15 K
the 1-octanol mole fraction in the [C1C2Im][HSO4]-rich phase
is x1‑octanol = 0.004, while the [C1C2Im][HSO4] mole fraction is
close to 0.005 in the 1-octanol-rich phase. In other words, the
immiscibility region of [C1C2Im][HSO4] + n-alcohol increases
with the alkyl chain length, which can be linked to the increase of
the van der Waals force in the solution. Such information can also
be associated with the increase of thermal hydrolytic stability
observed with the increase of the alkyl chain length on the alkyl
sulfate anion. To investigate the stability of the [C1C2Im]-
[HSO4] IL with 1-butanol and 1-octanol, the liquid + liquid
miscibility gap was measured in both cases using a standard
nephelometric method.22,33 Clear and cloud point temperatures
Table 7. Thermal Hydrolytic Conversion of
[C1C2Im][EtSO4] to [C1C2Im][HSO4] at a Range of
Temperatures with Varying Mole Ratios of Water
a
b
entry
T/K
mole ratio IL:H2O
t/h
% conv
1
2
3
4
5
6
7
8
9
10
383
383
403
403
423
423
423
423
423
423
1:2
1:5
1:2
1:5
1:2
1:5
1:1
1:2
1:5
1:10
18
18
18
18
18
18
72
72
72
72
0
0
5
16
13
31
0
72
87
87
c
c
c
c
c
11
423
1:10
18
>98
a
Reactions were conducted in sealed autoclave with headspace.
Conversion measured by H NMR. Reaction conducted in an open
b
c
1
vessel.
1
[C1C2Im][EtSO4], as observed by H NMR and exemplified,
herein, in Figure S7 of the Supporting Information. For example,
an increase in temperature from 383 to 403 to 423 K with double
the molar amount of water present resulted in small increases in
degradation from 0% to 5% to 13%, respectively (see Table 7
entries 1, 3, and 5). Although these conversions appear to be low,
even small amounts of degradation can significantly impact
thermophysical properties.19 Similar trends in degradation were
also observed for analogous reactions conducted using 5 mol
equiv of water (see Table 7 entries 2, 4, and 6). Significant
increases in degradation were observed by 1H NMR (see Figure
2) for reactions conducted at 423 K, with up to 87% degradation
observed after 72 h in the presence of 5 or 10 mol equiv of water,
as reported in Table 7 (entries 9 and 10). Interestingly, for
reactions conducted in sealed vessels with a headspace, the
volume of water added was critical for high conversions. For
example, in an equimolar mixture of IL and H2O, no conversion
was observed at 423 K, probably due to the majority of the water
being in the headspace and, therefore, unable to react, as reported
in Table 7 (entry 7). Moreover, the small increases in conversion
from 72% to 87% on increasing the water content from 2 to 5 to
10 mol (Table 7, entries 8−10) suggests that an equilibrium is
present between the headspace phase and the reactant liquid
phase.
In this regard, when the thermal hydrolysis of [C1C2Im]-
[EtSO4] was conducted in an open vessel, almost complete
conversion was observed due to the removal of ethanol, as
1944
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
Figure 2. 1H NMR spectrum showing the hydrolysis reaction between [C1C2Im][EtSO4] and 10 mol equiv of water at 423 K, forming ethanol and
[C1C2Im][HSO4].
K. Similarly, the UCST was reached at x1‑octanol = 0.5 and 443.15
K in the case of the [C1C2Im][HSO4] + 1-octanol mixture.
Surprisingly, in both cases large differences are observed between
clear and cloud point temperatures (up to 82 K) in the case of the
[C1C2Im][HSO4] + 1-octanol binary mixture. In the case of
[C1C2Im][HSO4] + 1-butanol, no cloud point is then observed
by decreasing the temperature down to room temperature
(293.15 K) for each individual composition studied during this
work (still a one-phase system). In the case of [C1C2Im][HSO4]
+ 1-octanol, no cloud point is then observed after maintaining
the temperature at 433.15 K for 5 h for each individual
composition investigated herein. In other words, rapid trans-
esterification is observed between [C1C2Im][HSO4] and 1-
butanol and between [C1C2Im][HSO4] and 1-octanol. In both
Table 8. Thermal Hydrolytic Conversion of Various Alkyl
Sulfate-Based ILs after Treating the Mixture Containing 1:5
Mol Ratios of IL:H2O at 423 K over 18 h (entry 8, after 2
a
months)
b
entry
cation
anion
% conv
1
2
3
4
5
6
7
[C1C2Im]+
[C1C4Im]+
[MeSO4]−
53
51
49
26
51
28
<2
[C1C4Pyrro]+
+
[N1444
]
[C1C2Im]+
[MeSO4]−
[EtSO4]−
[OctSO4]−
c
c
c
8
[OctSO4]
32
a
Reactions were conducted in sealed autoclave with headspace.
Conversion measured by H NMR. After 2 months.
1
cases, H NMR and 13C NMR spectra confirm the trans-
b
c
1
esterification reaction between the hydrogen sulfate and the
alcohol (see Figures S17 and 18 of the Supporting Information,
respectively). After reaction, the product from the equimolar
mixtures of [C1C2Im][HSO4] with 1-butanol or of [C1C2Im]-
[HSO4] with 1-octanol was dried under vacuum overnight at 433
K. The density of each product was then measured at 298.15 K
and 0.1 MPa, giving values of 1.1760 and 1.0895 g·cm−3 in the
case of the [C1C2Im][HSO4] + 1-butanol and [C1C2Im][HSO4]
+ 1-octanol binary mixtures, respectively. In other words, each
product has NMR spectra and density at 298.15 K close to those
observed in the case of [C1C2Im][BuSO4]24 and of [C1C2Im]-
[OcSO4],24,29 respectively. In both cases, rapid transesterifica-
tion is observed. These results mirror those obtained in the
hydrolysis experiments, that the octyl sulfate-based IL is more
Table 9. Reaction of [C1C2Im][HSO4] with 2.5 Mol Equiv of
Alcohol at 423 K over 18 h
a
b
entry
ROH
% conv
1
2
MeOH
EtOH
51
51
a
Reaction were conducted in sealed autoclave without headspace.
Conversion measured by H NMR.
b
1
were systematically investigated as reported in the Figure 3. In
the case of the [C1C2Im][HSO4] + 1-butanol binary mixture,
during the first heating cycle, the upper critical solution
temperature (UCST) is observed at x1‑butanol = 0.6 and 317.15
1945
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
■
Figure 3. “LLE coexistence curves” demonstrating the transesterification reaction between (a) [C1C2Im][HSO4] + 1-butanol, where each represents
the first observed clear point at a composition expressed in IL mole fraction xIL (no cloud point is then observed after cooling the temperature down to
●
○
293 K); (b) [C1C2Im][HSO4] + 1-octanol, where each or represents the first observed clear and cloud points at a composition expressed in IL mole
fraction xIL, respectively. Each dashed line is for guidance only.
Table 10. Estimated Thermochemistry Properties Using G3 Calculations for Each Individual Compound Involved during This
Work as a Function of Temperature from 298.15 to 423.15 K
compd
G3(0K)
G3 energy
G3 enthalpy
G3 free energy
ZPE
T = 298.15 K
water
−76.38204
−76.37920
−76.37826
−76.39963
0.02052
0.02655
0.04940
0.07680
0.13142
0.24035
0.05408
0.08112
0.13553
0.24414
[HSO4]−
methanol
ethanol
−699.50093
−115.62922
−154.90475
−233.44605
−390.53036
−738.75680
−778.03307
−856.57538
−1013.66658
−699.49600
−115.62587
−154.90039
−233.43904
−390.51778
−738.75064
−778.02557
−856.56505
−1013.65069
−699.49505
−115.62492
−154.89944
−233.43810
−390.51683
−738.74970
−778.02462
−856.56411
−1013.64974
−699.52954
−115.65195
−154.93017
−233.47632
−390.56994
−738.78727
−778.06624
−856.61341
−1013.71066
1-butanol
1-octanol
[MeSO4]−
[EtSO4]−
[BuSO4]−
[OcSO4]−
T = 383.15 K
water
−76.38204
−76.37838
−76.37717
−76.40587
0.02052
0.02653
0.04940
0.07681
0.13142
0.24035
0.05408
0.08112
0.13553
0.24414
[HSO4]−
methanol
ethanol
−699.50096
−115.62922
−154.90475
−233.44605
−390.53036
−738.75680
−778.03308
−856.57538
−1013.66658
−699.49368
−115.62455
−154.89830
−233.43541
−390.51105
−738.74763
−778.02176
−856.55970
−1013.64222
−699.49246
−115.62334
−154.89709
−233.43420
−390.50984
−738.74641
−778.02055
−856.55849
−1013.64101
−699.53986
−115.65985
−154.93921
−233.48770
−390.58596
−738.79843
−778.07870
−856.62817
−1013.73022
1-butanol
1-octanol
[MeSO4]−
[EtSO4]−
[BuSO4]−
[OcSO4]−
T = 423.15 K
water
−76.38204
−76.37799
−76.37665
−76.40890
0.02052
0.02653
0.04940
0.07681
0.13142
0.24035
0.05408
0.08112
0.13553
0.24414
[HSO4]−
methanol
ethanol
−699.50096
−115.62922
−154.90475
−233.44605
−390.53036
−738.75680
−778.03308
−856.57538
−1013.66658
−699.49247
−115.62386
−154.89717
−233.43343
−390.50735
−738.74604
−778.01974
−856.55682
−1013.63761
−699.49113
−115.62252
−154.89583
−233.43209
−390.50601
−738.74470
−778.01840
−856.55548
−1013.63627
−699.54487
−115.66370
−154.94368
−233.49339
−390.59410
−738.80395
−778.08488
−856.63559
−1013.74000
1-butanol
1-octanol
[MeSO4]−
[EtSO4]−
[BuSO4]−
[OcSO4]−
hydrolytically stable than the corresponding methyl or ethyl
sulfate-based ILs. In other words, hydrogen sulfate-based ILs are
unstable in presence of alcohols, which can be seen as an
advantage for the separation of alcohol in a liquid mixture as well
as the starting point for the formation of new family of alkyl
sulfate anions. For example, perfluoro alcohol (2,2,3,3,4,4,4-
heptafluoro-1-butanol) and an aromatic ring alcohol (benzyl
alcohol) both reacted at 383 K to form the corresponding alkyl
and aryl sulfate anion (see Figures S19−S20 of the Supporting
Information). An amino alcohol (N,N-diisopropylaminoetha-
1946
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
Table 11. Hydrolysis Equilibrium Constants (KT) and Transesterification Equilibrium Constants (KT−1) Calculated by Using Eqs
7−10 with Values Reported in Table 10 as a Function of Temperature from 298.15 to 4223.15 K
T = 298.15 K
T = 383.15 K
T = 423.15 K
−1
−1
−1
anion
KT
KT
KT
KT
KT
KT
[MeSO4]−
[EtSO4]−
[BuSO4]−
[OcSO4]−
0.00323
309.73
681.43
2015.0
94862
0.0226
44.20
93.61
209.7
4801
0.0411
24.33
49.63
104.3
0.00147
0.0107
0.0201
0.000496
0.0000105
0.00477
0.000208
0.00959
0.000603
1658
nol) was also reacted with [C1C2Im][HSO4], rapidly forming a
highly viscous material at 293K (see Figure S21 in the Supporting
Information). A similar rapid reaction was also observed when
[C1C2Im][HSO4] was reacted with butylamine, forming a solid
material. In both of these cases, the exact structure of the new IL
is currently under investigation (see Figure S22 in the Supporting
Information).
and eqs 7−10, hydrolysis equilibrium constants (KT) and
transesterification equilibrium constants (KT−1) were then
deduced for all reactions between water and an [alkyl sulfate]−
(methyl, ethyl, butyl, or octyl sulfate) anion and are reported,
herein, in Table 11. From the Tables 10 and 11, and as expected
from the experimental results, it appears that each investigated
reaction is driven by a thermodynamic equilibrium. Furthermore,
it can be observed that the hydrolysis reaction described in eq 7
increases by increasing the temperature. In the case of the ethyl
sulfate hydrolysis reaction, KT increases up to 20 times by
increasing the temperature from 298.15 to 423.15 K; a similar
observation is made for each alkyl sulfate investigated during this
work, as shown in Table 11. On the basis of results reported in
Tables 10 and 11, it appears also that the hydrolysis reaction is
more favorable in the case of the anion having a short chain
length such as the methyl and ethyl sulfate, where KT values are
close to 0.04 and 0.02 at 423 K, respectively. In the case of the
hydrolysis reaction involving the ethyl sulfate-based ILs, the
calculated KT value is in agreement (same order of magnitude)
with that observed experimentally (experimental KT ≈ 0.07) by
using a sealed vessel without a headspace.44 It appears, finally,
that the hydrolysis thermodynamic equilibrium constant
decreases, as expected from experimental value, with the alkyl
chain length on the alkyl sulfate anion, a contrario of the
transesterification equilibrium constants KT−1. For alkyl sulfate
anions containing a long chain length, such as found in butyl or
octyl sulfate, KT values are 1 or 2 orders of magnitude lower than
those observed in the case of methyl and ethyl sulfate, which can
explain why hydrolysis products are obtained after a few months
of reaction in the case of the octyl sulfate anion. Additionally, and
as observed experimentally, a favorable transesterification
reaction is predicted during these calculations by mixing 1-
octanol (or 1-butanol) with [HSO4]− anion, as its equilibrium
constants KT−1 is close to 1658 (or 104) at 423 K.
Thermodynamic Equilibrium between Hydrolysis and
Transesterification Reactions. The standard procedure for
estimating thermochemistry values from Gaussian has been
used,45 by using the optimized ideal gas phase structure of all
species as a function of temperature. Herein this was performed
using HF within Gaussian version 3.0,46 by means of using the 6-
31G(d) basis set. The resultant optimized structure for each
molecule is subsequently then used to calculate the thermo-
chemistry values using the G3 method,45 from which different
outputs like the G3(0K), the total electronic energy as defined by
the compound model, including the zero-point energy (ZPE)
scaled by the factor defined in the model (ε0 + εZPE); the G3
energy, the total electronic energy plus the internal thermal
energy, which already includes the ZPE (ε0 + Etot); the G3
enthalpy, the sum of the total electronic energy and thermal
enthalpies (ε0 + Hcorr); and the G3 free energy, the sum of the
total electronic energy and thermal free energies (ε0 + Gcorr), are
obtained for each specie involved in the following hydrolysis/
transesterification reaction
T
[alkyl sulfate]− + water ↔ [hydrogen sulfate]− + alcohol
KT
(7)
where [alkyl sulfate]− is the methyl, ethyl, butyl, or octyl sulfate
anion, the alcohol structure is then obtained from the
corresponding [alkyl sulfate]− anion, T is the studied temper-
ature set as equal to 298.15, 383.15, and 423.15 K, and KT is the
equilibrium constant in the case of the hydrolysis reaction,
calculated according to the following equations
CONCLUSION
■
Δ H°(T) =
(ε + H
)
−
(ε + H
)
∑
∑
∑
r
0
corr products
0
corr reactants
During this work the volumetric properties of a series of alkyl
sulfate ILs as a function of temperature are reported. From this
study we extended our group contribution models to explain
some anomalies in density measurements compared with those
reported previously in the literature. This was explained by the
hydrolytic instability of aprotic alkyl sulfate anion-based ILs and
prompted us to further investigate their hydrolytic stability.
Therein, the role of the cation, i.e., imidazolium, ammonium, or
pyrrolidinium, played a minimal role in the IL stability with
water. However, the anion played a crucial role in determining
the rate of hydrolysis of the IL, with longer chain alkyl sulfate
anion systems such as 1-octyl and 1-butyl showing increased
stability compared to methyl and ethyl sulfate-based ILs. Despite
the increased stability of the octyl sulfate system, degradation to
the corresponding 1-octanol and hydrogen sulfate IL was
observed, albeit, after a few months at 423 K. This instability
(8)
Δ G°(T) =
(ε + G
)
−
(ε + G
)
∑
r
0
corr products
0
corr reactants
(9)
⎛
⎞
ΔrG°(T)
K = exp −
⎜
⎟
⎠
T
⎝
RT
(10)
where ΔrH°(T) is the enthalpy of reaction and ΔrG°(T) is the
enthalpy of reaction calculated at the temperature T, R is the
universal gas constant (R = 1.9859 cal·mol−1·K−1), and KT is the
hydrolysis equilibrium constant, from which the transesterifica-
−1
tion equilibrium constant can be calculated as equal to KT
.
By using this methodology, the energy of each individual
species was investigated during this work and reported at 298.15,
383.15 and 423.15 K in Table 10. By using data from this table
1947
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
[bmim][PF6] and [bmim][NTf2]. J. Chem. Eng. Data 2006, 51,
1856−1859.
(9) Carvalho, P. J.; Regueira, T.; Santos, L.M.N.B.F.; Fernandez, J.;
Coutinho, J. A. P. Effect of water on the viscosities and densities of 1-
butyl-3-methylimidazolium dicyanamide and 1-butyl-3-methylimidazo-
lium tricyanomethane at atmospheric pressure. J. Chem. Eng. Data 2010,
55, 645−652.
(10) Spohr, H. V.; Patey, G. N. The influence of water on the structural
and transport properties of model ionic liquids. J. Chem. Phys. 2010, 132,
234510−1−13.
(11) Goodrich, B. F.; De La Fuente, J. C.; Gurkan, B. E.; Lopez, Z. K.;
Price, E. A.; Huang, Y.; Brennecke, J. F. Effect of water and temperature
on absorption of CO2 by amine-functionalized anion-tethered ionic
liquids. J. Phys. Chem. B 2011, 115, 9140−9150.
(12) Swatloski, R. P.; Holbrey, J. D.; Rogers, R. D. Ionic liquids are not
always green: Hydrolysis of 1-butyl-3-methylimidazolium hexafluor-
ophosphate. Green Chem. 2003, 5, 361−363.
could help to explain the biodegradability of this alkyl sulfate IL
family.
The observed hydrolytic instability is thermodynamically
driven, and therefore, competition between hydrolysis and
transesterification is observed. The balance between the two
reactions depends on the alkyl sulfate chain length, affinity, and
mutual solubility of the component mixtures. For example,
mixtures of [C1C2Im][HSO4] and alcohols react to form the
corresponding alkyl sulfate and water, with rapid synthesis of
butyl and octyl sulfate ILs being observed when the mixture is
heated above the UCST. The partial miscibility of hydrogen
sulfate and 1-octanol can help to explain the increased stability of
ocyl sulfate ionic liquids. This type of reaction has the potential to
be employed for separation of alcohols as well as forming novel
alkyl sulfate-based ionic liquids using aromatic alcohols,
perfluoro alcohols, amino alcohols, and amines without the use
of a catalyst.
(13) Freire, M. G.; Neves, C. M. S. S.; Marrucho, I. M.; Coutinho, J. A.
P.; Fernandes, A. M. Hydrolysis of tetrafluoroborate and hexafluor-
ophosphate counter ions in imidazolium-based ionic liquids. J. Phys.
Chem. A 2010, 114, 3744−3749.
(14) Parvulescu, V. I.; Hardacre, C. Catalysis in ionic liquids. Chem.
̂
Rev. 2007, 107, 2615−2665.
ASSOCIATED CONTENT
■
S
* Supporting Information
1H and 13C NMR spectra recorded during this work are provided
in Figures S1−S6 and S7−S23 in the case of pure ILs and of their
reactions with water or alcohol, respectively. This material is
available free of charge via the Internet at http://pubs.acs.org
(15) Eβer, J.; Wasserscheid, P.; Jess, A. Deep desulfurization of oil
refinery streams by extraction with ionic liquids. Green. Chem. 2004, 6,
316−322.
(16) Wasserscheid, P.; Van Hal, R.; Bosmann, A. 1-n-Butyl-3-
̈
methylimidazolium ([bmim]) octylsulfateAn even ‘greener’ ionic
AUTHOR INFORMATION
Corresponding Author
*E-mail: johan.jacquemin@qub.ac.uk (J.J.); p.goodrich@qub.ac.
uk (P.G.).
■
liquid. Green Chem. 2002, 4, 400−404.
(17) Ficke, L. E.; Rodriguez, H.; Brennecke, J. F. Heat capacities and
excess enthalpies of 1-ethyl-3-methylimidazolium-based ionic liquids
and water. J. Chem. Eng. Data 2008, 53, 2112−2119.
(18) Jacquemin, J.; Husson, P.; Majer, V.; Cibulka, I. High-pressure
volumetric properties of imidazolium-based ionic liquids: Effect of the
anion. J. Chem. Eng. Data 2007, 52, 2204−2211.
Notes
The authors declare no competing financial interest.
(19) Jacquemin, J.; Husson, P. Comments and additional work on
“High-pressure volumetric properties of imidazolium-based ionic
liquidsEffect of the anion” [J. Chem. Eng. Data 2007, 52, 2204−
2211]. J. Chem. Eng. Data 2012, 57, 2409−2414.
ACKNOWLEDGMENTS
■
The authors would like to thank Dr. Alfonso Pensado from
Wilhelm-Ostwald Institute of Physical and Theoretical Chem-
istry of the University of Leipzig for his useful suggestions
regarding Gaussian calculations.
(20) Himmler, S.; Hormann, S.; van Hal, R.; Schulz, P.; Wassercheid, P.
̈
Transesterification of methylsulfate and ethylsulfate ionic liquidsAn
environmentally benign way to synthesize long-chain and functionalized
alkylsulfate ionic liquids. Green Chem. 2006, 8, 887−894.
REFERENCES
■
(21) Ferguson, J. L.; Holbrey, J. D.; Ng, S.; Plechkova, N. V.; Seddon,
K. R.; Tomaszowska, A. A.; Wassell, D. F. A greener, halide-free
approach to ionic liquid synthesis. Pure Appl. Chem. 2012, 84, 723−744.
(22) Wang, S.; Jacquemin, J.; Husson, P.; Hardacre, C.; Costa Gomes,
M. F. Liquid−liquid miscibility and volumetric properties of aqueous
solutions of ionic liquids as a function of temperature. J. Chem.
Thermodyn. 2009, 41, 1206−1214.
(1) Esperanca, J.; Lopes, J. N. C.; Tariq, M.; Santos, L.; Magee, J. W.;
Rebelo, L. P. N. Volatility of aprotic ionic liquidsA review. J. Chem.
Eng. Data 2010, 55, 3−12.
(2) Earle, M. J.; Esperanca, J.; Gilea, M. A.; Lopes, J. N. C.; Rebelo, L. P.
N.; Magee, J. W.; Seddon, K. R.; Widegren, J. A. The distillation and
volatility of ionic liquids. Nature 2006, 439, 831−834.
(3) Siedlecka, E. M.; Czerwicka, M.; Stolte, S.; Stepnowski, P. Stability
of ionic liquids in application conditions. Curr. Org. Chem. 2011, 15,
1974−1991.
(4) Jacquemin, J.; Husson, P.; Padua, A. A. H.; Majer, V. Density and
viscosity of several pure and water-saturated ionic liquids. Green Chem.
2006, 8, 172−180.
(5) Jacquemin, J.; Ge, R.; Nancarrow, P.; Rooney, D. W.; Costa
Gomes, M. F.; Padua, A. A. H.; Hardacre, C. Prediction of ionic liquid
properties. I. Volumetric properties as a function of temperature at 0.1
MPa. J. Chem. Eng. Data 2008, 53, 716−726.
(6) Husson, P.; Pison, L.; Jacquemin, J.; Gomes, M. F. C. Influence of
water on the carbon dioxide absorption by 1-ethyl-3-methylimidazolium
bis(trifluoromethylsulfonyl)amide. Fluid Phase Equilib. 2010, 294, 98−
104.
(23) Pereiro, A. B.; Verdia, P.; Tojo, E.; Rodriguez, A. Physical
Properties of 1-butyl-3-methylimidazolium methyl sulfate as a function
of temperature. J. Chem. Eng. Data 2007, 52, 377−380.
(24) Costa, A. J. L.; Esperanca̧ , J. M. S. S.; Marrucho, I. M.; Rebelo, L. P.
N. Densities and viscosities of 1-ethyl-3-methylimidazolium n-alkyl
sulfates. J. Chem. Eng. Data 2011, 56, 3433−3441.
(25) Gonzalez, B.; Gomez, E.; Dominguez, A.; Vilas, M.; Tojo, E.
Physicochemical characterization of new sulfate ionic liquids. J. Chem.
Eng. Data 2011, 56, 14−20.
(26) Gomez, E.; Gonzalez, B.; Calvar, N.; Tojo, E.; Dominguez, A.
Physical properties of pure 1-ethyl-3-methylimidazolium ethylsulfate
and its binary mixtures with ethanol and water at several temperatures. J.
Chem. Eng. Data 2006, 51, 2096−2102.
(27) Kumelan, J.; Perez-Salado Kamps, A.; Tuma, D.; Maurer, G.
Solubility of CO2 in the Ionic Liquids [bmim][CH3SO4] and
[bmim][PF6]. J. Chem. Eng. Data 2006, 51, 1802−1807.
(28) Navia, P.; Troncoso, J.; Romani, L. Excess magnitudes for ionic
liquid binary mixtures with a common ion. J. Chem. Eng. Data 2007, 52,
1369−1374.
(7) Seddon, K. R.; Stark, A.; Torres, M. J. Influence of chloride, water,
and organic solvents on the physical properties of ionic liquids. Pure App.
Chem. 2000, 72, 2275−2287.
(8) Troncoso, J.; Cerdeirina, C. A.; Sanmamed, Y. A.; Romani, L.;
Rebelo, L. P. N. Thermodynamic properties of imidazolium-based ionic
liquids: Densities, heat capacities, and enthalpies of fusion of
1948
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949

The Journal of Physical Chemistry B
Article
(29) Russina, O.; Gontrani, L.; Fazio, B.; Lombardo, D.; Triolo, A.;
Caminiti, R. Selected chemical-physical properties and structural
heterogeneities in 1-ethyl-3-methylimidazolium alkyl-sulfate room
temperature ionic liquids. Chem. Phys. Lett. 2010, 493, 259−262.
(30) de Castro, C. A. N.; Langa, E.; Morais, A. L.; Lopes, M. L. M.;
(45) http://www.gaussian.com/g_whitepap/thermo.htm viewed on-
line on July 24, 2012.
(46) Gaussian View 3.0 2000−2003, Gaussian, Inc., Semichem, Inc.;
Pittsburgh, PA.
Lourenco
I.; Macatrao, M.; Esperanca
̧
, M. J.; Santos, F. J.; Santos, M. S. C.; Lopes, J. N. C.; Veiga, H.
, J. M. S. S.; Marques, C. S.; Rebelo, L. P.;
̧
̃
Afonso, C. A. Studies on the density, heat capacity, surface tension and
infinite dilution diffusion with the ionic liquids [C4mim][NTf2],
[C4mim][dca], [C2mim][EtOSO3] and [Aliquat][dca]. Fluid Phase
Equilib. 2010, 294, 157−179.
(31) Ficke, L. E.; Novak, R. R.; Brennecke, J. F. Thermodynamic and
thermophysical properties of ionic liquid + water systems. J. Chem. Eng.
Data 2010, 55, 4946−4950.
(32) Shekaari, H.; Armanfar, E. Apparent molar volumes and
expansivities of aqueous solutions of ionic liquids, 1-alkyl-3-methyl-
imidazolium alkyl sulfate at T = (298.15−328.15) K. Fluid Phase Equilib.
2011, 303, 120−125.
(33) Heintz, A.; Lehmann, J. K.; Wertz, C.; Jacquemin, J.
Thermodynamic properties of mixtures containing ionic liquids. 4.
LLE of binary mixtures of [C2MIM][NTf2] with propan-1-ol, butan-1-
ol, and pentan-1-ol and [C4MIM][NTf2] with cyclohexanol and 1,2-
hexanediol including studies of the influence of small amounts of water.
J. Chem. Eng. Data 2005, 50, 956−960.
(34) Ab Manan, N.; Atkins, M. P.; Jacquemin, J.; Hardacre, C.; Rooney,
D. W. Phase equilibria of binary and ternary systems containing ILs,
dodecane, and cyclohexanecarboxylic acid. Sep. Sci. Technol. 2012, 47,
312−324.
(35) Ramdin, M.; Vlugt, T. J. H.; de Loos, T. W. Solubility of CO2 in
the ionic liquids [TBMN][MeSO4] and [TBMP][MeSO4]. J. Chem.
Eng. Data 2012, 57, 2275−2280.
(36) Davila, M. J.; Aparicio, S.; Alcalde, R.; Garcia, B.; Leal, J. M. On the
properties of 1-butyl-3-methylimidazolium octylsulfate ionic liquid.
Green Chem. 2007, 9, 221−232.
(37) Jacquemin, J.; Husson, P.; Majer, V.; Padua, A. A. H.; Costa
Gomes, M. F. Thermophysical properties, low pressure solubilities and
thermodynamics of solvation of carbon dioxide and hydrogen in two
ionic liquids based on the alkylsulfate anion. Green Chem. 2008, 10,
944−950.
(38) Rooney, D. W.; Jacquemin, J.; Gardas, R. Thermophysical
properties of ionic liquids. Top. Curr. Chem. 2009, 290, 185−212.
(39) Dong, Q.; Muzny, C. D.; Kazakov, A.; Diky, V.; Magee, J. W.;
Widegren, J. A.; Chirico, R. D.; Marsh, K. N.; Frenkel, M. ILThermo: A
free-access web database for thermodynamic properties of ionic liquids.
J. Chem. Eng. Data 2007, 52, 1151−1159.
(40) Rebelo, L. P. N.; Najdanovic-Visak, V.; Gomes de Azevedo, R.;
Esperanca̧ , J. M. S. S.; Nunes da Ponte, M.; Guedes, H. J. R.; de Sousa, H.
C.; Szydlowski, J.; Canongia Lopes, J. N.; Cordeiro, T. C. Phase
behaviour and thermodynamic properties of ionic liquids, ionic liquid
mixtures, and ionic liquid solutions. In Ionic Liquids IIIA: Fundamentals,
Progress, Challenges, and OpportunitiesProperties and Structure; Rogers,
R. D., Seddon, K. R., Eds; ACS Symposium Series 901; American
Chemical Society: Washington, DC, 2005; Chapter 21, pp 270−291.
(41) Jacquemin, J.; Nancarrow, P.; Rooney, D. W.; Costa Gomes, M.
́
F.; Padua, A. A. H.; Hardacre, C. Prediction of ionic liquid properties. II.
Volumetric properties as a function of temperature at 0.1 MPa. J. Chem.
Eng. Data 2008, 53, 716−726.
(42) Anthony, J. L.; Anderson, J. L.; Maginn, E. J.; Brennecke, J. F.
Anion effects on gas solubility in ionic liquids. J. Phys. Chem. B 2005, 109,
6366−6374.
(43) Domanska, U.; Pobudkowska, A.; Wisniewska, A. Solubility and
excess molar properties of 1,3-dimethylimidazolium methylsulfate, or 1-
butyl-3-methylimidazolium methylsulfate, or 1-butyl-3-methylimidazo-
lium octylsulfate ionic liquids with n-alkanes and alcohols: Analysis in
terms of the PFP and FBT models. J. Solution Chem. 2006, 35, 311−334.
(44) [C1C2Im][EtSO4]:H2O molar ratio 1:5, 27% conversion after 4 h,
29% conversion after 8 h, 27% conversion after 18 h, and 27%
conversion after 24 h. All conversions were monitored by 1H NMR.
1949
dx.doi.org/10.1021/jp312241h | J. Phys. Chem. B 2013, 117, 1938−1949