Ind. Eng. Chem. Res.

2005, 44, 5757-5760 5757

GENERAL RESEARCH

Solubility of Thiophene + Pentane and Thiophene + Octane Binary

Mixtures in Supercritical Carbon Dioxide at Temperatures from 333
to 383 K
Octavio Elizalde-Solis and Luis A. Galicia-Luna*
Instituto Politécnico Nacional, ESIQIE, Laboratorio de Termodinámica, Edif. Z, Secc. 6, 1ER piso, UPALM,
C. P. 07738, Lindavista, México D. F., México

The solubility of thiophene + hydrocarbon mixtures in supercritical carbon dioxide was studied
in this work. Phase behavior for the ternary system thiophene + pentane + carbon dioxide was
measured at 333.66, 362.13, and 382.71 K. Additionally, behavior for the ternary system
thiophene + octane + carbon dioxide was determined at 334.25, 363.73, and 383.28 K.
Measurements were carried out to analyze the influence of pentane and octane on the solubility
of thiophene in carbon dioxide. The results show that, for the system thiophene + pentane +
carbon dioxide, the solubility of thiophene in carbon dioxide was higher than that obtained for
the system thiophene + octane + carbon dioxide. The Peng-Robinson equation of state (PReos)
using classical mixing rules was used to predict the experimental vapor-liquid equilibium (VLE)
data of both ternary systems. The calculated values agree within an average absolute deviation
<10%.

Introduction As a continuation of this study, the solubility of

The hydrotreating process has been successfully ap- thiophene + pentane and thiophene + octane binary
plied to reduce the sulfur content in fuels in the mixtures in carbon dioxide from 333 to 383 K are
petroleum industry. However, severe operational condi- reported in this work. The PReos5 using classical mixing
tions, i.e., high temperature and high hydrogen con- rules was used to calculate the phase behavior.
sumption, are required to obtain lower sulfur content
in fuels, making the operative costs high.1 New alterna- Experimental Section
tive processes are needed to complement or substitute
Materials. Carbon dioxide (CO2) with a minimum
the hydrotreating process. The extraction of sulfur
purity of 99.995% and helium with a minimum purity
compounds instead of reaction could help to reduce
of 99.998% were supplied by Air Products-Infra. Thio-
energy consumption. An attempt to develop an extrac-
phene (99+%), pentane (99+%, HPLC grade), and
tion process of sulfur from synthetic fuels has been done
octane (99+%) were obtained from Aldrich Chemical
recently by Huang et al.2 and Bösmann et al.3 using
Co., Inc. These compounds were not purified further.
ionic liquid technology. We are focused on the super-
The thiophene and the alkanes were degassed under
critical fluid extraction (SCFE) of sulfur compounds
vacuum while stirring.
from oil fractions. The supercritical fluid technology
Apparatus. The experimental apparatus is based on
strongly depends on accurate phase equilibrium data.
the static-analytic method. A schematic flow diagram
Besides, this information is required for the develop-
of the apparatus is illustrated in Figure 1. It mainly
ment and validation of thermodynamic models. Our
consists of an equilibrium cell (ECT), pressure and
attention is focused on the search for an appropriate
temperature measurement systems, and a movable
cosolvent4 and the study of the influence of fuel com-
capillary sampler (MCS; Rolsi6). The apparatus prin-
pounds on the solubility of sulfur in CO2. Selectivity and
ciple was described in detail elsewhere.7
capacity of the supercritical solvent are some variables
The ECT has an internal volume of ∼100 cm3 and can
that can give a clue about the feasibility of the SCFE of
be operated at up to 60 MPa and 673.15 K. It is
sulfur from oil fractions. Because of the scarce informa-
equipped with a magnetic rod (MR) activated with a
tion about CO2 + sulfur compounds, we are carrying
magnetic stirring device (MSD). The temperature of the
out a systematic study to obtain phase equilibria and
ECT was kept constant by an air bath (AB). Two
transport properties in order to know the behavior of
platinum probes (PTPi; Pt100, Specitec), immersed in
such systems. We have reported the solubility of
thermometric wells at the top and bottom of the cell
thiophene in carbon dioxide + 1-propanol mixtures as
body and connected to a digital indicator (F250, Auto-
part of a systematic study.4
matic Systems Laboratories), were used for temperature
* To whom correspondence should be addressed. Tel.: measurements. Pressure was monitored with a thermo-
(52) 55 5729-6000 Ext. 55133. Fax: (52) 55 5586-2728. regulated pressure transducer (PT; PDCR 910-1756,
E-mail: [email protected]. Druck) connected to an indicator (DPI 150, Druck).
10.1021/ie048780y CCC: $30.25 © 2005 American Chemical Society
Published on Web 06/14/2005
5758 Ind. Eng. Chem. Res., Vol. 44, No. 15, 2005

Table 2. Experimental VLE for the System Thiophene (1)

+ Octane (2) + CO2 (3)
P/MPa x1 x2 y1 y2
T ) 334.25 K
4.163 0.1175 0.5265 0.0056 0.0061
5.063 0.1052 0.4757 0.0052 0.0061
5.981 0.0922 0.4185 0.0050 0.0066
7.141 0.0749 0.3433 0.0050 0.0077
8.025 0.0610 0.2813 0.0053 0.0093
9.138 0.0415 0.1915 0.0059 0.0136
9.968 0.0244 0.1090 0.0083 0.0263
10.223 0.0178 0.0774 0.0101 0.0362
T ) 363.73 K
4.102 0.1289 0.5924 0.0117 0.0154
5.002 0.1211 0.5503 0.0110 0.0153
6.012 0.1125 0.5056 0.0098 0.0146
7.031 0.1032 0.4593 0.0099 0.0160
8.046 0.0937 0.4119 0.0101 0.0172
9.021 0.0846 0.3684 0.0104 0.0193
10.002 0.0747 0.3205 0.0111 0.0231
11.077 0.0627 0.2614 0.0125 0.0289
12.128 0.0507 0.2048 0.0156 0.0415
12.944 0.0337 0.1242 0.0238 0.0786
T ) 383.28 K
3.017 0.0258 0.0296
4.545 0.1517 0.5730 0.0200 0.0252
Figure 1. Experimental apparatus: AB, air bath; CT, titanium 5.124 0.1452 0.5503 0.0185 0.0243
cap; CNi, connecting nuts; ECT, titanium equilibrium cell; MCS, 6.109 0.1345 0.5122 0.0170 0.0239
movable capillary sampler; MR, magnetic rod; MSD, magnetic 6.979 0.1248 0.4806 0.0164 0.0243
stirring device; OR, O-ring; PT, pressure transducer; PTPi, plati- 8.104 0.1136 0.4406 0.0161 0.0260
num temperature probe; TRi, thermal regulator i, and Vi, shutoff 9.107 0.1027 0.4012 0.0160 0.0281
valve i. 10.155 0.0920 0.3635 0.0160 0.0310
11.031 0.0827 0.3287 0.0166 0.0352
Table 1. Experimental VLE for the System Thiophene (1) 12.067 0.0722 0.2868 0.0183 0.0433
+ Pentane (2) + CO2 (3) 13.114 0.0599 0.2383 0.0203 0.0553
P/MPa x1 x2 y1 y2 14.031 0.0451 0.1759 0.0267 0.0879

T ) 333.66 K
3.013 0.1463 0.5671 0.0075 0.0747 The GC (HP-5890 series II) used for the composition
4.121 0.1242 0.4846 0.0066 0.0621 analyses is equipped with a thermal conductivity detec-
5.102 0.1041 0.4107 0.0067 0.0590 tor (TCD). A packed column (Porapak Q, Alltech) with
6.108 0.0876 0.3466 0.0072 0.0577 a length of 1.2 m and an external diameter of 0.32 cm
7.054 0.0691 0.2753 0.0079 0.0576 was used with helium as the carrier gas at a flow rate
7.941 0.0523 0.2157 0.0098 0.0620
8.350 0.0439 0.1768 0.0108 0.0641
of 30 mL/min.
Calibration. The temperature probes were calibrated
T ) 362.13 K
3.338 0.1871 0.5769 0.0190 0.1288
against a calibration system (F300S, Automatic Systems
4.291 0.1689 0.5289 0.0170 0.1104 Laboratories) fitted with a 25-Ω reference probe (model
5.273 0.1510 0.4780 0.0169 0.1038 162CE with (0.005 K certified accuracy, Rosemount)
5.989 0.1396 0.4480 0.0173 0.1008 and a triple point cell. The pressure transducer was
7.164 0.1162 0.3904 0.0181 0.1020 calibrated against a dead weight gauge (model 5304
8.072 0.1043 0.3523 0.0192 0.1036 Class S2, (0.005% full scale precision up to 138 MPa,
9.041 0.0870 0.2976 0.0224 0.1121
10.167 0.0648 0.2299 0.0330 0.1451
Desgranges & Huot). Estimated uncertainties were
within (0.03 K and (0.04% for temperature and
T ) 382.71 K pressure, respectively.
3.844 0.2111 0.5947 0.0318 0.1750
5.133 0.1886 0.5278 0.0289 0.1497 The calibration of the TCD detector was performed
5.877 0.1808 0.4976 0.0294 0.1453 by relating the known amounts of external standards
6.959 0.1568 0.4271 0.0306 0.1416 with the chromatographic areas obtained. These stan-
8.089 0.1489 0.3872 0.0329 0.1410 dards (carbon dioxide, thiophene, pentane, and octane)
9.168 0.1315 0.3552 0.0368 0.1493 were injected into the GC using gas chromatographic
9.997 0.1100 0.3207 0.0453 0.1709
10.455 0.0966 0.2859 0.0519 0.1811
syringes.
Experimental Procedure. Experimental phase be-
About 1 µL of liquid or vapor phase can be sampled havior data were obtained by using the following
by the MCS. The amount of sample depends on the procedure:
temperature and pressure conditions inside the cell, and (a) Calibrating the temperature probes, pressure
these are sent to the gas chromatograph (GC), through transducer, and the TCD detector.
a thermoregulated transferring circuit. Resistance heat- (b) Loading the cell: the binary mixture of liquid
ers were used to heat the expansion chamber of the compounds was introduced into the equilibrium cell at
MCS and the transferring circuit. The temperature of a fixed volume ratio (thiophene/pentane ) 1.0:4.51 and
the heaters was set above the boiling point of the thiophene/octane ) 1.0:6.28).
heaviest compound. This avoids the condensation of the (c) Setting up the experimental conditions: the liquid
samples, making it easy for them to be carried by helium mixture was degassed under vacuum and vigorously
flow. stirred with the MR. After degassing for 20 min, the
 Ind. Eng. Chem. Res., Vol. 44, No. 15, 2005 5759

Figure 2. Solubility of thiophene in CO2 with the presence of Figure 3. Solubility of thiophene in CO2 at ∼383 K for the
pentane at: b, 333.66 K; 1, 362.13 K; and 9, 382.71 K; and with systems: b, thiophene + pentane + CO2; 3, thiophene + octane
the presence of octane at: O, 334.25 K; 3, 363.73 K; and 0, + CO2; and 9, thiophene + CO2.4
383.28 K.

shutoff valve (V1) was closed. Then, V1 was opened to

introduce carbon dioxide into the measuring cell using
a syringe pump (100DM, Isco) until the desired pressure
was achieved. Then, AB is set to the measuring tem-
perature.
(d) Performing the equilibrium measurements: the
equilibrium temperature is reached in 120 min for this
apparatus. Once pressure and temperature reached the
set values and were kept constant, measurements were
performed. The MCS was moved into the ECT up to a
certain level where the studied phase was located. At
least five samples of either the liquid or vapor phase
were taken in order to check for the reproducibility
(within the experimental uncertainty) and to perform
mole fraction error analysis for each temperature and
pressure. Composition uncertainties were estimated to
be <1%, according to Figure 4. K1 values for the system thiophene + pentane + CO2
at: b, 333.66 K; 1, 362.13 K; 9, 382.71 K; and s, PReos; and
(xi,yi)max - (xi,yi)min for the system thiophene + octane + CO2 at: O, 334.25 K; 3,
xi,yi(uncertainty) ) × 100 (1) 363.73 K; 0, 383.28 K; and - - -, PReos.
(xi,yi)max
Figure 2. The solubility of thiophene in carbon dioxide
where xi and yi are either the liquid or vapor mole reported by Elizalde-Solis and Galicia-Luna4 was com-
fraction and the superscripts max and min denote the pared with the experimental solubilities of thiophene
maximum and minimum values for xi and yi within the obtained from the ternary systems in this work at
last five consecutive samples. ∼383 K in Figure 3. Solubility values of thiophene
Once the phase compositions had been quantified, the decrease under the presence of the alkanes. This means
next pressure was achieved by adding carbon dioxide
that carbon dioxide dissolves pentane and octane better
into the cell. Isothermal phase envelopes were traced
than thiophene.
out by consecutive increments in pressure.
Modeling. The PReos5 and classical mixing rules
Results and Discussion were used to correlate VLE data for the binary sub-
systems. The critical properties of pure compounds were
Experimental Data. VLE measurements for the taken from the literature.8,9 Binary interaction param-
ternary system thiophene + pentane + carbon dioxide eters (kij) were obtained by correlating VLE data
were carried out at 333.66, 362.13, and 382.71 K. These reported in the literature for the mixtures: carbon
data sets are reported in Table 1. For the ternary system
dioxide + thiophene by Elizalde-Solis and Galicia-Luna,4
thiophene + octane + carbon dioxide, the experimental
results were obtained at 334.25, 363.73, and 383.28 K. carbon dioxide + pentane by Knapp et al.,10 and carbon
These are listed in Table 2. The solubility of thiophene, dioxide + octane by Weng and Lee.11 The obtained kij
pentane, and octane in carbon dioxide had the same values are: 0.0751, 0.1222, and 0.1239, in that order.
behavior with pressure and temperature changes. The Because VLE data for the systems thiophene + pentane
solubilities of these compounds increased with increas- and thiophene + octane have not been reported in the
ing temperature. At supercritical carbon dioxide condi- literature, the kij parameters for both mixtures were
tions, pressure and temperature increases gave a greater considered to be zero. Classical mixing rules were
increment in the solubility of thiophene or alkanes than selected because the lack of data for the thiophene +
that obtained at lower pressures, as is illustrated in alkane systems makes it difficult to use the more
5760 Ind. Eng. Chem. Res., Vol. 44, No. 15, 2005

complex mixing rules where additional parameters are Nomenclature

required to describe the interaction between the mol-
Ki ) equilibrium ratio of component i
ecules.
kij ) binary interaction parameter between i, j
The kij values obtained from the binary mixtures were P ) pressure, MPa
used to predict the solubility of thiophene in carbon T ) temperature, K
dioxide for the ternary systems reported in this work. x ) liquid mole fraction
The calculated solubilities of thiophene in carbon dioxide y ) vapor mole fraction
for the ternary systems were in agreement with the z ) function for pressure or vapor mole fraction in eq 2
experimental data. Calculated and experimental equi-
librium ratios (K1) for thiophene can be observed in Subscripts
Figure 4. Similar results are obtained for the solubility i, j ) species of the mixture
of hydrocarbons. Deviations in vapor mole fraction and
Superscripts
pressure are presented in the Tables of Supporting
Information. These were calculated by the next equation min ) minimum value
max ) maximum value
exp ) experimental data
(ziexp - zcal
i ) cal ) calculated data
% error z ) | exp
| x 100 (2)
zi
Literature Cited
where z is either the pressure or the vapor mole fraction. (1) Vargas-Villamil, F. D.; Marroquin, J. O.; de la Paz, C.;
Rodriguez, E. A Catalytic Distillation Process for Light Gas Oil
Hydrodesulfurization. Chem. Eng. Process. 2004, 43, 1309.
Conclusions (2) Huang, C.; Chen, B.; Zhang, J.; Liu, Z.; Li, Y. Desulfuriza-
tion of Gasoline by Extraction with New Ionic Liquids. Energy
Experimental vapor-liquid equilibrium data are re- Fuels 2004, 18, 1862.
ported in this work for thiophene in supercritical carbon (3) Bösmann, A.; Datsevich, A. J.; Lauter, A.; Schmitz, C.;
dioxide. These measurements were carried out for two Wasserscheid, P. Deep Desulfurization of Diesel Fuel by Extraction
ternary systems. The solubility of thiophene in super- with Ionic Liquids. Chem. Commun. 2001, 23, 2494.
(4) Elizalde-Solis, O.; Galicia-Luna, L. A. Solubility of Thiophene
critical carbon dioxide for the system thiophene + in Carbon Dioxide and Carbon Dioxide + 1-Propanol Mixtures at
pentane + carbon dioxide was greater than that ob- Temperatures from 313 to 363 K. Fluid Phase Equilib. 2005, 230,
tained for the thiophene + octane + carbon dioxide 51.
system. On the other hand, the solubility values for (5) Peng, D. Y.; Robinson, D. B. New Two-Constant Equation
thiophene in supercritical carbon dioxide obtained in of State. Ind. Eng. Chem. Fundam. 1976, 15, 59.
both ternary systems were lower than those reported (6) Guilbot, P.; Valtz, A.; Legendre, H.; Richon, D. Rapid On-
Line Sampler-Injector: A Reliable Tool for HT-HP Sampling and
in the literature for the binary system thiophene + On-Line GC Analysis. Analusis 2000, 28, 426.
carbon dioxide. It was an expected result, because (7) Elizalde-Solis, O.; Galicia-Luna, L. A.; Sandler, S. I.; Sam-
pentane and octane were added to compose the payo-Hernández, J. G. Vapor-Liquid Equilibria and Critical
ternary systems and the carbon dioxide is more selective Points of the CO2 + 1-Hexanol and CO2 + 1-Heptanol Systems.
for alkanes. The solubility values for the ternary Fluid Phase Equilib. 2003, 210, 215.
systems were predicted with the PReos. Calculated (8) Poling, B. E.; Prausnitz, J. M.; O’Connell, J. P. The Proper-
ties of Gases and Liquids; McGraw-Hill: New York, 2001.
values and experimental data were found to be in (9) Yaws, C. L. Chemical Properties Handbook; McGraw-Hill:
agreement. New York, 1999.
(10) Knapp, H.; Döring, R.; Oellrich, L.; Plöcker, U.; Prausnitz,
J. M. Vapor-Liquid Equilibria for Mixtures of Low Boiling
Acknowledgment
Substances; Chemistry Data Series, Vol. 6; DECHEMA: Berlin,
1981.
The authors are grateful to CONACYT and IPN for
(11) Weng, W. L.; Lee, M. J. Vapor-Liquid Equilibrium of the
their financial support. Octane/Carbon Dioxide, Octane/Ethane, and Octane/Ethylene
Systems. J. Chem. Eng. Data 1992, 37, 213.
Supporting Information Available: Solubility and
pressure deviation data for the systems thiophene (1) Received for review December 17, 2004
+ pentane (2) + CO2 (3) and thiophene (1) + octane (2) Revised manuscript received May 6, 2005
Accepted May 17, 2005
+ CO2 (3). This material is available free of charge via
the Internet at http://pubs.acs.org. IE048780Y