Liquid-liquid equilibria for [C8mim][NTf2] + thiophene + 2,2,4-trimethylpentane or + toluene

Article abstract of DOI:10.1021/je800071q

Tie-line compositions for type II systems of 1-methyl-3-octylimidazolium bis[trifluoromethylsulfonyl]imide + thiophene + 2,2,4-trimethylpentane or + toluene have been determined at 298.15 K and atmospheric pressure. Solute distribution coefficient and selectivity values have also been determined. The experimental data were correlated with the NRTL and UNIQUAC equations. The best results were found with the UNIQUAC equation. The NRTL equation could not adequately correlate the system with toluene.

Full text of DOI:10.1021/je800071q

1750
J. Chem. Eng. Data 2008, 53, 1750–1755
Liquid-Liquid Equilibria for [C8mim][NTf2] + Thiophene +
2,2,4-Trimethylpentane or + Toluene
Luisa Alonso, Alberto Arce,* Mar´ıa Francisco, and Ana Soto
Department of Chemical Engineering, University of Santiago de Compostela, E-15782 Santiago de Compostela, Spain
Tie-line compositions for type II systems of 1-methyl-3-octylimidazolium bis[trifluoromethylsulfo-
nyl]imide + thiophene + 2,2,4-trimethylpentane or + toluene have been determined at 298.15 K and
atmospheric pressure. Solute distribution coefficient and selectivity values have also been determined.
The experimental data were correlated with the NRTL and UNIQUAC equations. The best results were
found with the UNIQUAC equation. The NRTL equation could not adequately correlate the system
with toluene.
liquid-liquid equilibrium (LLE) of sulfur compounds with
hydrocarbons and solvent. As part of our study8–10 about the
liquid-liquid equilibrium of thiophene with hydrocarbons
and 1-methyl-3-octylimidazolium bis[trifluoromethylsulfo-
nyl]imide ([C8mim][NTf2]) IL, to analyze the viability of
using this IL for gasoline desulfurization, in this work phase
diagrams at T ) 298.15 K for [C8mim][NTf2] + thiophene
+ 2,2,4-trimethylpentane and [C8mim][NTf2] + thiophene
+ toluene ternary systems were determined. Also, thio-
phene distribution ratio coefficients and selectivities were
determined. The experimental data were correlated using the
NRTL and UNIQUAC equations.
Introduction
The products of the Fluid Catalytic Cracking (FCC) tend
to contain sulfur impurities even though about half of the
sulfur compounds are converted to hydrogen sulfide during
the cracking process. The distribution of sulfur in the cracking
products is dependent on a large number of factors including
feed, type of catalyst, presence of additives, conversion, and
other operating conditions, but in any case, a certain
proportion of the sulfur tends to enter the light or heavy
gasoline fractions.
With the increase of environmental regulations applied to
petroleum derivates, the sulfur content of the products has
to be reduced in response to concerns about the emissions
of sulfur oxides and other sulfur compounds to the air
following combustion processes. The conventional process
to remove sulfur in fuels is known as catalytic hydrodes-
ulfurization (HDS). Among the various sulfur compounds
present in fuel oil, thiophene (TP), benzothiophene (BT),
dibenzothiophene (DBT), and its derivatives are the most
resistant to hydrogenation and require the use of a modified
catalyst and drastic reaction conditions, making deep des-
ulfurization to become an expensive process. Despite recent
sulfur reduction technologies,1–4 there continues the need for
effective ways to reduce the sulfur content of gasoline and
diesel fuels. Extractive desulfurization (EDS) is considered
to be one of the promising new methods for deep desulfu-
rization of fuel oil.5–7 Compared to conventional HDS, EDS
can be carried out under very mild conditions, at room
temperature, and under atmospheric pressure.
Experimental
Materials. 2,2,4-Trimethylpentane, C8H18,(Fluka, mass
fraction > 99.5 %), toluene, C7H8 (Sigma-Aldrich, mass frac-
tion > 99.5 %), and thiophene, C4H4S (Aldrich, mass fraction
> 99.5 %), were used as received from the supplier without
further purification. Gas chromatography (GC) analysis did
not detect any appreciable peaks of impurities.
1-Methylimidazole was mixed with an excess of 1-chlorooc-
tane to produce [C8mim][Cl]. This chloride was mixed with
Li[NTf2] salt solvent in deionized water, thus obtaining the
[C8min][NTf2] by ion metathesis in a similar manner to that
reported in a previous paper.10 After washing, the purification
of the ionic liquid was completed by heating it under high
vacuum for at least 24 h (1 mbar, 80 °C). The water contain of
[C8min][NTf2] was measured using Karl Fisher titration and
resulted in 21 ppm. Chloride concentration was 413 ppm,
measured by means of capillary electrophoresis. The ionic liquid
was analyzed by 1H NMR and 13C NMR spectroscopy to
confirm the absence of any major impurities.
In Table 1, the experimental density, refractive index,
dynamic viscosity, and water content of pure components
are compared with values published by other authors.11–14
Apparatus and Procedure. Water contents were measured
using a Karl Fischer titration method in a Metrohm 737 KF
coulometer. Densities were measured with an Anton Paar
DMA 5000 densimeter automatically corrected for the
viscosity correction associated with this densimeter. The
uncertainty in the measurement was ( 10-5 g · cm-3. Refrac-
tive indices were measured in an ATAGO RX-5000 refrac-
The EDS technique is very efficient with economical
operation conditions, with the key question being to find an
effective solvent. Ionic liquids (ILs) are emerging solvents
because they have negligible vapor pressure, are nontoxic,
are thermally and chemically stable, and are not expensive
for commercial applications. ILs seem to be more competitive
that conventional solvents considering the fact that they are
environmentally benign and designable for desired properties.
The design of extraction processes to accomplish the
removal of sulfur compounds requires the knowledge of the
* Correspondingauthor.Tel.:-34981563100ext.16790.Fax:+34981528050.
E-mail address: eqaaarce@usc.es.
10.1021/je800071q CCC: $40.75
2008 American Chemical Society
Published on Web 06/26/2008

Journal of Chemical & Engineering Data, Vol. 53, No. 8, 2008 1751
Table 1. Water Mass Fraction (w), Density (G), Refractive Index (nD), and Dynamic Viscosity (η) of the Pure Components at 298.15 K and
Atmospheric Pressure
F/g·cm-3
nD
η/mPa·s
component
C8H18
C7H8
C4H4S
[C8mim][NTf2]
CAS number
106 ·w
exptl
lit.
exptl
lit.
exptl
lit.
540-84-1
108-88-3
110-02-1
178631-04-4
4.8
174
13.5
21
0.68784
0.86220
1.05859
1.31978
0.6878211
0.8621912
1.0588412
1.3113
1.38921
1.49393
1.52530
1.43270
1.3889812
1.4941312
1.5257212
Not found
0.871
0.577
0.612
90.37
0.88611
0.552512
0.61312
90.014
Table 2. Operating Conditions for the Gas Chromatography Instrument
[C8mim][NTf2] + C4H4S + C8H18
[C8mim][NTf2] + C4H4S + C7H8
column
HP-FFAP polyethylene glycol TPA
HP-5 5 % phenyl methyl siloxane
(25 m × 200 µm × 0.3 µm)
(30 m × 320 µm × 0.25 µm)
detector type
detector temperature
injector temperature
carrier gas
flow rate
T oven
TCD
FID
523.15 K
523.15 K
helium
523.15 K
583.15 K
helium
1 mL ·min-1
1 mL·min-1
-1
-1
-1
ramp 100 K·min
ramp 8 K·min
ramp 50 K·min
6.5 min at 353.15 K
isothermal for 6 min
8 413.15 K,
3.5 min at 313.15 K
isothermal for 6 min
8 333.15 K
8 433 K,
Table 3. Compositions of Experimental Tie-Lines, Solute
Distribution Ratios ꢀ, and Selectivities S for [C8mim][NTf2] (1) +
C4H4S (2) + C8H18 (3) at 298.15 K
tometer with a Heto Therm thermostat to maintain the
temperature constant. The uncertainty in the refractive index
measurement is ( 4 · 10-5. The kinematic viscosity was
determined by a micro Ubbelohde viscometer technique. The
capillaries are calibrated and credited by the company. Flow
time measurements were performed by Lauda Processor
Viscosity System PVS1. The temperature of the viscometer
was kept constant using a Lauda clear view thermostat D 20
KP with a through-flow cooler DLK 10. The dynamic
viscosities were calculated from densities with an estimated
uncertainty of ( 0.5 %.
The tie-line data were determined by analysis of the two
layers of a heterogeneous mixture. Mixtures with composi-
tions inside the immiscible region of the systems were
introduced into 30 mL glass wall-jacketed vessels, with
magnetic stirrers, and closed. Previous experiments showed
that equilibrium was established after about 5 h of stirring,
to get a good contact between both phases, and overnight to
settle down. Then, one sample of each layer was withdrawn
using syringes. The complete process was carried out at
constant temperature of 298.15 K using a thermostatic bath
(SELECTA 6000382). The uncertainty in the temperature
measurement is ( 0.002 K.
Samples of both liquid phases were analyzed by GC using
an internal standard method. IL composition was calculated
by difference. All weighing to prepare samples for analysis
calibration was carried out in a Mettler Toledo AT 261
balance with an uncertainty of ( 10-4 g. The uncertainty in
the mole fractions of the prepared mixtures was estimated
to be ( 2 · 10-4. The GC used was an HP 6890 Series
equipped with a thermal conductivity detector (TCD) and
the other with a flame ionization detector (FID), a capillary
column, and an empty precolumn to protect the column and
collect the ionic liquid that could not be retained by the liner.
The GC operating conditions are given in Table 2. To check
the uncertainty in the determination of mole fraction com-
positions, eight samples were prepared by weight, and
compositions were calculated using the calibration curves,
with the greatest deviations found being ( 0.005.
IL-rich phase
hydrocarbon-rich phase
x1
x2
x3
x′1
x′2
x′3
ꢀ
S
0.774 0.000 0.226
0.660 0.135 0.205
0.572 0.237 0.191
0.441 0.393 0.166
0.317 0.525 0.158
0.220 0.640 0.140
0.163 0.716 0.121
0.135 0.758 0.107
0.106 0.804 0.090
0.074 0.890 0.036
0.065 0.935 0.000
0.001
0.003
0.002
0.006
0.000
0.001
0.002
0.000
0.000
0.000
0.000
0.000
0.044
0.089
0.193
0.328
0.482
0.639
0.740
0.835
0.957
1.000
0.999
0.953
0.909
0.801
0.672
0.517
0.359
0.260
0.165
0.043
0.000
-
-
3.04 14.17
2.67 12.71
2.04 9.82
1.60 6.79
1.33 4.89
1.12 3.32
1.02 2.49
0.96 1.78
0.93 1.12
0.94
-
Table 4. Compositions of Experimental Tie-Lines, Solute
Distribution Ratios ꢀ, and Selectivities S for [C8mim][NTf2] (1) +
C4H4S (2) + C7H8 (3) at 298.15 K
IL-rich phase
hydrocarbon-rich phase
x1
x2
x3
x′1
x′2
x′3
ꢀ
S
0.113
0.110
0.097
0.095
0.091
0.087
0.086
0.080
0.078
0.065
0.000
0.069
0.246
0.154
0.341
0.425
0.515
0.654
0.765
0.935
0.887
0.821
0.657
0.751
0.568
0.488
0.399
0.266
0.157
0.000
0.000
0.001
0.003
0.000
0.002
0.000
0.000
0.001
0.004
0.000
0.000
0.074
0.257
0.159
0.381
0.475
0.571
0.727
0.834
1.000
1.000
0.925
0.740
0.841
0.617
0.525
0.429
0.272
0.162
0.000
-
-
0.93
0.96
0.97
0.89
0.90
0.90
0.90
0.92
0.94
1.05
1.08
1.09
0.97
0.96
0.97
0.92
0.95
-
[C8mim][NTf2] (1) + C4H4S (2) + C7H8 (3) ternary systems
are reported in Tables 3 and 4, respectively. Values of solute
distribution ratios (ꢀ) and selectivities (S) are also shown in
those tables. These parameters are defined as follows
x2
ꢀ )
(1)
(2)
x′2
x2 · x′3
x′3 · x3
S )
Results and Discussion
where superscript ′ indicates the hydrocarbon-rich phase and
no superscript the IL-rich phase.
Experimental LLE Data. The compositions of the experi-
mental tie-lines at the temperature of 298.15 K for
Correlation of LLE Data. The NRT15 and UNIQUAC16
equations were used to fit the experimental data, assigning a
[C8mim][NTf2] (1)
+ C4H4S (2) + C8H18 (3) and

1752 Journal of Chemical & Engineering Data, Vol. 53, No. 8, 2008
Table 5. UNIQUAC Structural (Volume and Area) Parameters17–19
2
γˆ´
F )
min
(x - xˆ )2 + Q P2 + ln S∞ꢀ∞
∑ ∑∑
∑
b
ijk
ijk
n
r
q
(
)
[
]
γˆS∞
k
i
j
thiophene
2,2,4-trimethylpentane
toluene
2.8569
5.8463
3.9228
13.800
2.140
5.008
2.968
9.310
(4)
where xijk is the experimental mole fraction of component i in
phase j on tie-line k; xˆijk is the value calculated, using the
parameters being optimized, for the corresponding end of a tie-
line that “min” makes coincide as closely as possible with the
experimental tie-line using the Nelder-Mead method. ꢀ∞ is the
solute molar distribution ratio at infinite dilution, and γˆS∞ and
γˆ′S∞ represent the solute (thiophene) activity coefficients
calculated at infinite dilution in IL and hydrocarbon phases,
respectively. Q ) 10-10 for eq 4.
[C8mim][NTf2]
Table 6. Binary Interaction Parameters (∆gij, ∆gji) and (∆uij, ∆uji)
and Root Mean Square Deviations (F, ∆ꢀ) for the NRTL and
UNIQUAC Correlation of [C8mim][NTf2] (1) + C4H4S (2) + C8H18
(3) at 298.15 K
components
parameters
model
NRTL
(R ) 0.3)
rmsd
i-j
∆uij/J·mol-1 ∆uji/J·mol-1
F
0.7293
1-2
1-3
2-3
1-2
1-3
2-3
-7659
2253
741.6
-1489
-1297
-340.4
19173
11051
-1594
3248
∆ꢀ 10.2
The goodness of the fit was quantified by the residual function
F and by the mean error of the solute distribution ratio, ∆ꢀ,
defined as
UNIQUAC
F
∆ꢀ
0.5620
4.40
3759
1367
Table 7. Binary Interaction Parameters (∆gij, ∆gji) and (∆uij, ∆uji)
and rmsd (F, ∆ꢀ) for the NRTL and UNIQUAC Correlation of
[C8mim][NTf2] (1) + C4H4S (2) + C7H8 (3) at 298.15 K
components
parameters
model
NRTL
(R ) 0.3)
rmsd
i-j
∆uij/J·mol-1 ∆uji/J·mol-1
F
∆ꢀ
14.4682
29.90
1-2
1-3
2-3
1-2
1-3
2-3
-1653
-2106
300.9
13183
12832
242.0
UNIQUAC
F
∆ꢀ
1.6702
3.70
-499.6
-1163
-117.8
1826
2545
76.80
Table 8. Binary Interaction Parameters (∆gij, ∆gji) and (∆uij, ∆uji)
and rmsd (F, ∆ꢀ) for the NRTL and UNIQUAC Correlation of
[C8mim][NTf2] (1) + C4H4S (2) + C7H8 (3) at 298.15 Ka
components
parameters
model
NRTL
(R ) 0.3)
rmsd
i-j
∆uij/J·mol-1 ∆uji/J·mol-1
F
8.1977
10.90
1.14
1.4306
4.30
1-2
1-3
2-3
1-2
1-3
2-3
-2862
-2688
1237
12698
12157
1384
∆ꢀ
ꢀ∞
F
∆ꢀ
ꢀ∞
UNIQUAC
-796.5
-1238
-154.8
2192
2685
222.8
Figure 1. Liquid-liquid equilibria for the ternary system [C8mim][NTf2]
(1) + C4H4S (2) + C8H18 (3) at 298.15 K: b, solid line, experimental
tie-lines; O, short dash, tie-lines correlated by means of the NRTL equation
with R ) 0.3.
0.95
a Solute distribution ratio at infinite dilution, ꢀ∞, is fixed.
priori different values for the nonrandomness parameter R in
the NRTL equation: 0.1, 0.2, and 0.3. The structural parameters
for the application of the UNIQUAC equation were taken from
the literature17–19 and are shown in Table 5.
The binary interaction parameters for both equations were
calculated running a computer program designed by Sꢁrensen
and Arlt,20 which proceeds in two stages. In the first, ap-
proximate values of the binary interaction parameters are
obtained using an objective function that requires no initial
estimate of their values
F )
[(a′ - a ) ⁄ (a′ + a )]2 + Q ·
P2 (3)
n
∑∑
∑
n
a
ik
ik
ik
ik
k
i
where aik is the activity of component i on tie-line k. Superscript
′ indicates the hydrocarbon-rich phase, and no superscript the
IL-rich phase. In Pn, n ) 6 parameters being optimized and Q
) 10-6, the factor weighting this contribution to Fa. The effect
n
of the term Q∑ P2n is to minimize the risk of the Gibbs energy
surface having more than the required number of minima. In
the second stage, the parameter values found in the first stage
are used as initial values in minimizing the objective function
Figure 2. Liquid-liquid equilibria for the ternary system [C8mim][NTf2]
(1) + C4H4S (2) + C8H18 (3) at 298.15 K: b, solid line, experimental
tie-lines; ∆, long dash, tie-lines calculated from UNIQUAC.

Journal of Chemical & Engineering Data, Vol. 53, No. 8, 2008 1753
0.5
(xijk - xˆijk)2
6M
j
F ) 100 ·
min
(5)
(6)
∑ ∑∑
[
]
k
i
0.5
2
ˆ
((ꢀk - ꢀk) ⁄ ꢀk)
∆ꢀ ) 100 ·
∑
[
]
M
k
where M refers to the number of experimental tie-lines.
In this work, experimental data were correlated in two ways:
without defining an a priori value of ꢀ∞, the last term of eq 4
becoming zero, and specifying an optimal value of this
parameter. An appropriate given value of ꢀ∞ can improve the
fit at low solute concentrations. Here, the value of ꢀ∞ minimizing
the goodness-of-fit index ∆ꢀ was found by trial and error.
Tables 6 and 7 show the binary interaction parameters and
residuals for the NRTL and UNIQUAC correlations of ternary
data, without fixing an optimal value for the solute distribution
ratio at infinite dilution. For the NRTL model, the value of the
nonrandomness parameter R ) 0.3 was used because it gives
the best results. For the ternary system with toluene, when ꢀ∞
is fixed at its optimal value and the NRTL (R ) 0.3) equation
is used, the residuals F and ∆ꢀ decreases extensively. If
UNIQUAC equation is used, F decreases, but ∆ꢀ slightly
increases. Nevertheless, for the system with 2,2,4-trimethyl-
pentane, there is nearly no difference between fixing or not
fixing the value of ꢀ∞, and so, the results using the optimal
value for ꢀ∞ are not shown. Table 8 shows the values of root-
mean-square deviations for the UNIQUAC and NRTL (R )
0.3) correlations, by fixing the optimal value of the solute
distribution ratio at infinite dilution for the system with toluene.
Figure 1 shows the experimental and correlated NRTL (R )
0.3) tie-lines for the 1-methyl-3-octylimidazolium bis[trifluo-
romethylsulfonyl]imide + thiophene + 2,2,4-trimethylpentane
ternary system. A similar comparison between experimental and
correlated data is done in Figure 2 with the UNIQUAC equation.
Figures 3 and 4 show the experimental, NRTL, and UNIQUAC
correlated (ꢀ∞ fixed) data for bis[trifluoromethylsulfonyl]imide
+ thiophene + toluene. Despite that the values of ꢀ∞ fixed are
used, NRTL is incapable of adequately correlating this system.
Figure 4. Liquid-liquid equilibria for the ternary system [C8mim][NTf2]
(1) + C4H4S (2) + C7H8 (3) at 298.15 K: b, solid line, experimental tie-
lines; ∆, long dash, tie-lines calculated from UNIQUAC using the optimal
value of the solute distribution ratio at infinite dilution.
Figure 5. Solute distribution ratio as a function of the mole fraction of
solute in the hydrocarbon-rich phase (x′2) for the [C8mim][NTf2] (1) +
C4H4S (2) + C8H18 (3) ternary system. b, exptl; -, NRTL; - - -, UNIQUAC
model.
Figures 5 and 6 show the correlated solute distribution ratios
and selectivities plotted against the mole fraction of solute in
the hydrocarbon-rich phase of the [C8mim][NTf2] + C4H4S +
C8H18 ternary system for comparison with experimental values.
Solute distribution ratios and selectivities are higher than one
at low thiophene concentrations (note that in the desulfurization
of fossil fuels the content of sulfur compounds is low, making
the ionic liquid an adequate solvent for the extraction of
thiophene from 2,2,4-trimethylpentane). Despite that NRTL and
UNIQUAC correlate relatively well these parameters, consider-
able deviations are found to low solute concentrations. Analo-
gously, Figures 7 and 8 present experimental and correlated (ꢀ∞
fixed) solute distribution ratios and selectivities for the
[C8mim][NTf2] + C4H4S + C7H8 ternary system. In this case,
both parameters are lower or very close to one, extracting the
IL important quantities of the aromatic compound. Correlation
of these parameters is not adequate, especially with the NRTL
equation, as was previously commented.
Figure 3. Liquid-liquid equilibria for the ternary system [C8mim][NTf2]
(1) + C4H4S (2) + C7H8 (3) at 298.15 K: b, solid line, experimental tie-
lines; O, short dash,tie-lines correlated by means of the NRTL equation
with R ) 0.3 using the optimal value of the solute distribution ratio at
infinite dilution.

1754 Journal of Chemical & Engineering Data, Vol. 53, No. 8, 2008
zolium bis[trifluoromethylsulfonyl]imide + thiophene + toluene
have been experimentally obtained at the temperature of 298.15
K and atmospheric pressure. Both systems are type II because
[C8mim][NTf2] is partially miscible with all other components.
Thiophene solubility in the [C8mim][NTf2] is high (94 %), and
[C8mim][NTf2] solubility in hydrocarbons is poor (values around
or below detection limits).
Due to the high selectivity values found for the system
with 2,2,4-trimethylpentane, separation of the sulfur com-
pound seems to be feasible, but low values of distribution
coefficients imply the use of large quantities of solvent.
Nonetheless, negligible vapor pressure of IL facilitates
solvent recovery without losses. However, low values of both
parameters found for toluene indicate that separation of
thiophene from toluene using the IL studied is not favorable
thermodynamically. The studies would have to focus on a
combined desulfurization and dearomatization.
Figure 6. Selectivity as a function of the mole fraction of solute in the
hydrocarbon-rich phase (x′2) for the system [C8mim][NTf2] (1) + C4H4S
(2) + C8H18 (3). b, exptl; -, NRTL; - - -, UNIQUAC model.
The experimental LLE data were correlated using NRTL and
UNIQUAC models. For both systems, the UNIQUAC model
gives the best results, NRTL being incapable of adequately
correlating the system with toluene.
Literature Cited
(1) Hai, M.; Mei, B. W.; Fu, Y. T. A new method for obtaining ultralow
sulfur diesel fuel via ultrasound assisted oxidative desulfurization. Fuel
2003, 82, 405–414.
(2) Schmitz, C.; Datsevitch, L.; Jess, A. Deep desulfurization of diesel
oil: kinetic studies and process-improvement by the use of a two-
phase reactor with pre-saturator. Chem. Eng. Sci. 2004, 59, 2821–
2829.
(3) Tailleur, R.; Galiasso, R. J.; Quenza, S.; Valencia, N. Catalyst for
ultra-low sulfur and aromatic diesel. Appl. Catal., A 2005, 282, 227–
235.
(4) Babich, I. V.; Moulijn, J. A. Science and technology of novel processes
for deep desulfurization of oil refinery streams: a review. Fuel 2003,
82, 607–631.
(5) Nie, Y.; Li, C.; Sun, A.; Meng, H.; Wang, Z. Extractive Desulfurization
of Gasoline Using Imidazolium-Based Phosphoric Ionic Liquids.
Energy Fuels 2006, 20, 2083–2087.
(6) Alonso, L.; Arce, A.; Francisco, M.; Rodr´ıguez, O.; Soto, A. Gasoline
Desulfurization using Extraction with [C8min][BF4] Ionic Liquid.
AIChE J. 2007, 53, 3108–3115.
Figure 7. Solute distribution ratio as a function of the mole fraction of
solute in the hydrocarbon-rich phase (x′2) for the [C8mim][NTf2] (1) +
C4H4S (2) + C7H8 (3) ternary system. b, exptl; -, NRTL; - - -, UNIQUAC
model.
(7) Holbrey, J. D.; Lo´pez-Martin, I.; Rothenberg, G.; Seddon, K. R.;
Silvero, G.; Zheng, X. Desulfurisation of oils using ionic liquids:
selection of cationic and anionic components to enhance extraction
efficiency. Green Chem. 2008, 10, 87–92.
(8) Alonso, L.; Arce, A.; Francisco, M.; Rodr´ıguez, O.; Soto, A.
Measurement and correlation of Liquid-Liquid Equilibria of two
imidazolium ionic liquids with thiophene and methylcyclohexane.
J. Chem. Eng. Data 2007, 52, 2409–2412.
(9) Alonso, L.; Arce, A.; Francisco, M.; Soto, A. (Liquid-Liquid) Equilibria
of C8mim][NTf2] ionic liquid with a sulfur-component and hydrocar-
bons. J. Chem. Thermodyn. 2008, 40, 265–270.
(10) Alonso, L.; Arce, A.; Francisco, M.; Soto, A. Phase behaviour of
1-methyl-3-octylimidazolium bis[trifluoromethylsulfonyl]imide with
thiophene and aliphatic hydrocarbons: The influence of n-alkane
chain length. Fluid Phase Equilib. 2008, 263, 176–181.
(11) Pa´dua, A. A. H.; Fareleira, J. M. N. A.; Calado, J. C. G.; Wakeham,
W. A. Density and Viscosity Measurements of 2,2,4-Trimethylpentane
(Isooctane) from 198 to 348 K and up to 100 MPa. J. Chem. Eng.
Data 1996, 41, 1488–1494.
(12) Riddick, J. A.; Bunger, W. B.; Sakano, T. K. Organic SolVents.
Physical Properties and Methods of Purification, 4th ed.; John Wiley
& Sons: New York, 1986.
(13) Papaiconomou, N.; Yakelis, N.; Salminen, J.; Bergman, R.; Prausnitz,
J. M. Synthesis and Properties of Seven Ionic Liquids Containing
1-Methyl-3-octylimidazolium or 1-Butyl-4-methylpyridinium Cations.
J. Chem. Eng. Data 2006, 51, 1389–1393.
Figure 8. Selectivity as a function of the mole fraction of solute in the
hydrocarbon-rich phase (x′2) for the system [C8mim][NTf2] (1) + C4H4S
(2) + C7H8 (3). b, exptl; -, NRTL; - - -, UNIQUAC model.
(14) Tokuda, H.; Tsuzuki, S.; Susan, M. A. B. H.; Hayamizu, K.; Watanabe,
M. How Ionic Are Room-Temperature Ionic Liquids? An Indicator
of the Physicochemical Properties. J. Phys. Chem. B 2006, 110, 19593–
19600.
(15) Renon, H.; Prausnitz, J. M. Local compositions in thermodynamic
excess functions for liquid mixtures. AIChE J. 1968, 14, 135–144.
Conclusions
The liquid-liquid-phase equilibria for ternary systems 1-meth-
yl-3-octylimidazolium bis[trifluoromethylsulfonyl]imide +
thiophene + 2,2,4-trimethylpentane or 1-methyl-3-octylimida-

Journal of Chemical & Engineering Data, Vol. 53, No. 8, 2008 1755
(16) Abrams, D. S.; Prausnitz, J. M. Statistical thermodynamics of liquid
mixtures: A new expression for the excess Gibbs energy of partly
or completely miscible systems. AIChE J. 1975, 21, 116–128.
(17) Daubert, T. E.; Danner, R. P. Physical and Thermodynamic
Properties of Pure Chemicals Data Compilation; Hemisphere: New
York, 1989.
(18) Prausnitz, J. M.; Anderson, T. F.; Grens, E. A.; Eckert, C. A.; Hsieh,
R.; O’Connell, J. P. Computer Calculations for Multicomponent
VLE and LLE; Prentice-Hall: Englewood Cliffs, New York, 1980.
(19) Simoni, L. D.; Lin, Y.; Brennecke, J. F.; Stadtherr, M. A. Modeling
Liquid-Liquid Equilibrium of Ionic Liquid Systems with NRTL,
Electrolyte-NRTL, and UNIQUAC. Ind. Eng. Chem. Res. 2008, 47,
256–272.
(20) Sorensen, J. M.; Arlt, W. Liquid-Liquid Equilibrium Data Collec-
tion; DECHEMA Chemistry Data Series: Frankfurt, 1980.
Received for review January 30, 2008. Accepted April 10, 2008.
The authors are thankful for the financial support under project
CTQ2006-07687/PPQ from the Ministerio of Ciencia y Tecnolog´ıa
(Spain). L. Alonso is grateful to the Ministerio for financial support via
“Juan de la Cierva” Programme and M. Francisco for FPI grant
(BES-2007-16693).
JE800071Q