Fluid Phase Equilibria 259 (2007) 45–50

New isothermal vapor–liquid equilibria for the

CO2 + n-nonane, and CO2 + n-undecane systems
Luis E. Camacho-Camacho, Luis A. Galicia-Luna ∗ ,
Octavio Elizalde-Solis, Zenaido Martı́nez-Ramı́rez
Instituto Politécnico Nacional, ESIQIE, Laboratorio de Termodinámica, UPALM, Edif. Z, Secc. 6,
1ER piso, Lindavista, C.P. 07738, México, D.F., Mexico
Received 23 December 2006; received in revised form 10 April 2007; accepted 12 April 2007
Available online 24 April 2007

Abstract
Experimental vapor–liquid equilibria (VLE) for the CO2 + n-nonane and CO2 + n-undecane systems were obtained by using a 100-cm3 high-
pressure titanium cell up to 20 MPa at four temperatures (315, 344, 373, and 418 K). The apparatus is based on the static-analytic method; which
allows fast determination of the coexistence curve. For the CO2 + n-nonane system, good agreement was found between the experimental data and
those reported in literature. No literature data were available for the CO2 + n-undecane system at high temperature and pressure. Experimental data
were correlated with the Peng–Robinson equation of state using the classical and the Wong–Sandler mixing rules.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Vapor–liquid equilibria; Equation of state; Static-analytic method

1. Introduction
Carbon dioxide has received more attention in chemical
areas, especially in the supercritical fluid technology due to
its well-known advantages [1]. Systems involving mixtures of
CO2 + alkanes, or +alkanols are also important because of its
applications in separation processes. Accurate information about
phase equilibrium behavior, volumetric properties and critical
properties is essential for the development of these processes.
The enhanced oil recovery is the main area, which requires cor-
rect thermodynamic data for the CO2 + hydrocarbon systems in
a wide range of temperature and pressure. This information can
be obtained by using thermodynamic models. These are appro-
priate tools to predict phase behavior and reduce the number
of experimental measurements, however interaction parameters
for the binary mixtures are needed.
Although high-pressure phase equilibria for the carbon diox-
ide and n-alkane systems has been widely investigated by several
authors [2], scarce data concerning the vapor–liquid equilibria
∗ Corresponding author. Tel.: +52 55 5729 6000x55133;
fax: +52 55 5586 2728.
E-mail address: lgalicial@ipn.mx (L.A. Galicia-Luna).
for the CO2 + n-nonane and CO2 + n-undecane systems is found
in literature.
The vapor–liquid equilibria for the CO2 + n-nonane system
are only reported by Jennings and Schucker [3] at 343.25 K. On
the other hand, the high-pressure phase behavior for the CO2 + n-
undecane system was studied at low temperatures by Schneider
et al. [4].
In this work, measurements of vapor–liquid equilibria for
the CO2 + n-nonane and CO2 + n-undecane systems were per-
formed from 314 to 418 K. The experimental data for the first
binary mixture was compared with the previously reported at
∼344 K [3], good agreement between both sets of data can be
observed. The isothermal data for the CO2 + n-undecane sys-
tem provide new phase behavior for literature. The degree of
smoothness in the equilibrium ratios in both binary mixtures
indicates an excellent internal consistency of the experimental
measurements.
The representation of the vapor–liquid equilibria was car-
ried out with the Peng–Robinson equation of state [5] (PR EoS)
using classical and Wong–Sandler [6] mixing rules. For this last
mixing rule, the non-random two liquid (NRTL) model [7] was
used to predict the excess Helmholtz free energy. Good agree-
ment was found between the experimental and calculated data
sets.

0378-3812/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.fluid.2007.04.022
46 L.E. Camacho-Camacho et al. / Fluid Phase Equilibria 259 (2007) 45–50
Table 1
Temperature conditions for the gas chromatography analyses
System Temperature (K)
Injector Oven
CO2 + n-nonane 483 493
CO2 + n-undecane 483 503
2. Experimental
2.1. Materials
Anhydrous n-nonane (nC9 H20 ) and n-undecane (nC11 H24 )
with a certified purity of 99+% were obtained from
Sigma–Aldrich, USA. Carbon dioxide (CO2 , mole frac-
tion purity > 0.99995) and helium (He, mole fraction
purity > 0.99998) were provided by Air Products-Infra,
Mexico. Alkanes were used without any further purification,
these were only degassed before being used, and gases were
used as received.
2.2. Static-analytic apparatus
The experimental apparatus has been previously used for
vapor–liquid equilibria measurements [8], it is based in the
static-analytic method. Its main component includes a 100-cm3
equilibrium cell made of titanium alloy that can be operated
up to 60 MPa and 673 K. Two windows at the front and the
backside of the cell allow the visual observation of the num-
ber of phases involved into the studied mixture. The cell is
connected on-line to a gas chromatograph (GC). A detailed
description of the apparatus has been given by Elizalde-Solis et
al. [9].
For the cell temperature measurements, the cell has two
horizontal wells (at the top and at the bottom) in which each plat-
inum probe Pt100 (Specitec, France) was inserted. The probes
were connected to a digital indicator (F250, Automatic Systems
Laboratories, USA), they were calibrated by using a 25- ref-
erence probe (Rosemount, model 162CE ± 0.005 K) connected
to a calibration system F300S (Automatic Systems Laborato-
ries, USA). Temperature uncertainties were calculated within
±0.03 K. Pressure inside the cell was measured by means of a
thermoregulated pressure transducer (PDCR 910-1756, Druck,
UK) connected to a digital indicator (DPI 145, Druck, UK). The
pressure transducer was calibrated against a dead weight bal-
ance (Desgranges & Huot, Model 5304 Class S2, ±0.005%).
Pressure uncertainty was estimated within ±0.04%.
The temperature conditions for the GC are listed in Table 1.
The packed column used in the analytical equipment was a
stainless steel porapak Q 80/100 (Alltech) diam.: 1/8 in. The
column length was 3 and 2 ft for the composition analyses in the
CO2 + n-nonane and CO2 + n-undecane systems, respectively.
The thermal conductivity detector (TCD) of the GC was cali-
brated according to the procedure described by Elizalde-Solis
and Galicia-Luna [8]. Uncertainties in TCD calibration using
the n-alkanes and CO2 were less than 1%.
2.3. Procedure
The high-pressure view cell was initially loaded with about
TCD
50 cm3 of the n-alkane, and placed into the air bath. The pres-
sure transducer and the loading circuit were connected to the
503 cell. Then the n-alkane was degassed under vigorous stirring
513
for 30 min. After that by carefully feeding CO2 , the pressure
of the cell was raised up to 1.5 MPa by means of an Isco
syringe pump (100 DM, USA). Next, the temperature of air
bath was set to the desired value. When temperature was kept
constant, pressure was set to a desired value by adding CO2
or removing CO2 + n-alkane mixture from the cell. The stir-
ring device allows faster pressure stabilization into the cell. The
phase equilibrium measurements were started when tempera-
ture and pressure remained constant. Finally, mole fractions
were determined by taking samples from the cell using a
ROLSITM sampler-injector [10] and sent to the GC through
a thermoregulated line to ensure the complete vaporization
of the liquid samples and to avoid condensation of vapor
samples due to the difference between temperatures of the
room and the equilibrium cell. At least five consecutive sam-
ples of each phase are taken at a stated pressure. The phase
composition uncertainties were estimated within ±1%. The
isothermal phase envelopes were measured by adding CO2 to the
system.
3. Results and modeling
Fig. 1 shows the agreement between the vapor–liquid equilib-
rium data for the CO2 + n-nonane system measured at 344.53 K
in this work and the data reported at 343.25 K by Jennings
and Schucker [3]. The four isothermal data at 315.12, 344.53,
373.28, and 418.82 K are presented in Table 2. The new mea-
sured vapor–liquid equilibria for the CO2 + n-undecane system
reported at 314.98, 344.46, 373.13 and 418.30 K, are presented
in Table 3 and illustrated in Fig. 2. In both cases the solubil-
Fig. 1. Experimental and correlated vapor–liquid equilibria for the CO2 + n-
nonane system. (䊉) 315.12 K; () 344.53 K; () 373.28 K; (♦) 418.82 K; ()
343.25 K, Jennings and Schucker [3]; (—) PR EoS using Wong–Sandler mixing
rule.
 L.E. Camacho-Camacho et al. / Fluid Phase Equilibria 259 (2007) 45–50 47

Table 2
Experimental vapor–liquid equilibria for the CO2 + n-nonane system
p (MPa) xCO2 yCO2 KCO2 KnC9

T = 315.12 K
2.033 0.2207 0.9987 4.5254 0.0016
3.098 0.3254 0.9986 3.0691 0.0021
4.309 0.4601 0.9986 2.1705 0.0025
5.452 0.5698 0.9984 1.7522 0.0038
6.668 0.7085 0.9978 1.4082 0.0075
7.204 0.7975 0.9971 1.2504 0.0142
7.800 0.8993 0.9960 1.1075 0.0393
8.026 0.9340 0.9955 1.0658 0.0681
T = 344.53 K
2.111 0.1779 0.9951 5.5950 0.0060
3.748 0.2987 0.9958 3.3338 0.0060
5.464 0.4149 0.9955 2.3996 0.0077
6.942 0.5127 0.9948 1.9404 0.0106
8.651 0.6297 0.9932 1.5774 0.0183 Fig. 2. Experimental and correlated vapor–liquid equilibria for the CO2 + n-
10.059 0.7177 0.9875 1.3758 0.0443 undecane system. (䊉) 314.98 K; () 344.46 K; () 373.13 K; (♦) 418.30 K;
11.236 0.8346 0.9778 1.1715 0.1344 (—) calculated with PR EoS using Wong–Sandler mixing rule.
11.989 0.9049 0.9635 1.0648 0.3835
T = 373.28 K
2.064 0.1472 0.9873 6.7050 0.0149
3.508 0.2379 0.9891 4.1583 0.0143
4.979 0.3213 0.9894 3.0798 0.0156
6.481 0.4059 0.9898 2.4382 0.0172
7.941 0.4746 0.9887 2.0832 0.0215
9.594 0.5768 0.9861 1.7096 0.0329
11.031 0.6465 0.9832 1.5207 0.0475
12.578 0.7306 0.9758 1.3357 0.0897
13.550 0.7878 0.9680 1.2287 0.1509
14.839 0.8678 0.9335 1.0757 0.5030
14.970 0.8965 0.9188 1.0249 0.7848
T = 418.82 K
2.548 0.1460 0.9506 6.5126 0.0578
3.729 0.2008 0.9618 4.7889 0.0478
5.058 0.2599 0.9674 3.7225 0.0441
6.620 0.3381 0.9691 2.8663 0.0466
7.990 0.3978 0.9697 2.4375 0.0503
9.476 0.4657 0.9679 2.0785 0.0601
11.122 0.5371 0.9660 1.7985 0.0735 Fig. 3. Vapor–liquid equilibrium ratios for the CO2 + n-nonane system. (䊉)
12.509 0.5889 0.9613 1.6325 0.0940 315.12 K; () 344.53 K; () 373.28 K; (♦) 418.82 K; () 343.25 K, Jennings
14.154 0.6733 0.9479 1.4079 0.1593 and Schucker [3]; (—) PR EoS using Wong–Sandler mixing rule.
15.627 0.7321 0.9297 1.2699 0.2625
16.773 0.8110 0.8891 1.0963 0.5868

ity of CO2 in n-alkane decreases when temperature increases at

constant pressure.
Behavior of the equilibrium ratios (Ki = yi /xi ) for the
CO2 + n-nonane and CO2 + n-undecane systems are illustrated
in Figs. 3 and 4, respectively. Fig. 3 also shows a good agree-
ment between the experimental data and those from literature
[3]. The logarithmic scale for the equilibrium ratios permits a
better observation for the four isotherms behavior in function of
pressure. K-values for the carbon dioxide (light component) are
greater than one, while the n-alkane have K-values lower than
one. The isothermal equilibrium ratios converge to the unity at
the critical point, however the critical point is not presented. Fig. 4. Vapor–liquid equilibrium ratios for the CO2 + n-undecane system. (䊉)
According to these figures both systems exhibit a good internal 314.98 K; () 344.46 K; () 373.13 K; (♦) 418.30 K; (—) PR EoS using
consistency due to the smoothness of the isothermal curves, and Wong–Sandler mixing rule.
48 L.E. Camacho-Camacho et al. / Fluid Phase Equilibria 259 (2007) 45–50

Table 3 where p and T are the pressure and temperature, and v is the
Experimental vapor–liquid equilibria for the CO2 + n-undecane system molar volume. a(T) and b are the parameters obtained from the
p (MPa) xCO2 yCO2 KCO2 KnC11 critical temperature Tc and pressure pc , and the acentric factor
T = 314.98 K
ω of each pure component:
2.373 0.2433 0.9985 4.1036 0.0020
R2 Tc2
3.162 0.3087 0.9993 3.2372 0.0010 a(T ) = 0.45724 α(Tr ) (2)
4.010 0.3801 0.9995 2.6296 0.0009 pc
4.920 0.4544 0.9994 2.1991 0.0012
RTc
5.698 0.5258 0.9993 1.9003 0.0016 b = 0.07780 (3)
6.287 0.5825 0.9991 1.7154 0.0021 pc
6.879 0.6366 0.9989 1.5692 0.0029
7.638 0.7370 0.9982 1.3543 0.0069 2
8.294 0.9134 0.9926 1.0867 0.0859 α(Tr )=[1 + (0.37464 + 1.54226ω−0.26992ω2 )(1 − Tr0.5 )]
T = 344.46 K (4)
2.558 0.1851 0.9988 5.3961 0.0015
4.057 0.2805 0.9985 3.5600 0.0020 the subscript r denotes the reduced property for each compo-
5.526 0.3680 0.9985 2.7134 0.0024 nent, respectively. R is the gas constant, α is a temperature
7.063 0.4727 0.9981 2.1114 0.0036
8.586 0.5569 0.9971 1.7905 0.0064
dependent factor. The critical properties and the acentric fac-
10.052 0.6425 0.9958 1.5500 0.0117 tor for each substance taken from literature [11] are presented
11.518 0.7189 0.9890 1.3757 0.0392 in Table 4. Appendix A describes the equations for classical and
12.946 0.8516 0.9699 1.1389 0.2030 Wong–Sandler mixing rules.
13.388 0.9161 0.9518 1.0390 0.5740 The k12 , A12 , and A21 parameters from Appendix A were
T = 373.13 K optimized by using the Levenberg–Marquardt method [12] and
3.468 0.2253 0.9962 4.4221 0.0049 minimizing the following objective function, F:
5.517 0.3282 0.9965 3.0361 0.0053 ⎡  2  cal 2 ⎤
7.013 0.3990 0.9958 2.4957 0.0070 
NP  Nc
yijcal − yijex pj − pex
j
8.527 0.4635 0.9950 2.1465 0.0093 F= ⎣ + ⎦ (5)
9.955 0.5185 0.9934 1.9159 0.0138 yijex pex
j
j=1 i=1
11.613 0.5842 0.9918 1.6976 0.0197
13.148 0.6412 0.9894 1.5432 0.0294 where NP is the number of data points, Nc the number of com-
14.840 0.7146 0.9805 1.3722 0.0682
16.001 0.7624 0.9595 1.2584 0.1706
ponents, y the vapor mole fraction, and the superscripts cal and
17.108 0.8705 0.9321 1.0708 0.5240 ex denote the calculated and experimental data, respectively.
The relative percentage deviation for the bubble pressure was
T = 418.30 K
2.052 0.1125 0.9709 8.6274 0.0328 calculated by
3.570 0.1736 0.9849 5.6722 0.0183
NP
5.017 0.2335 0.9859 4.2213 0.0184 100  |pex i − pi |
cal
p = (6)
6.433 0.2930 0.9856 3.3640 0.0204 NP pi
ex
8.175 0.3547 0.9854 2.7778 0.0227 i=1
9.538 0.3990 0.9843 2.4673 0.0261 the absolute mean deviation in the mole fraction vapor phase
11.209 0.4605 0.9818 2.1321 0.0337
12.535 0.5225 0.9788 1.8732 0.0444
was established as

NP
14.000 0.5777 0.9755 1.6886 0.0581
1  ex
15.556 0.6329 0.9716 1.5353 0.0772 y = |yi − yical | (7)
17.082 0.6842 0.9659 1.4117 0.1079 NP
i=1
18.390 0.7361 0.9539 1.2960 0.1745
19.184 0.7626 0.9342 1.2250 0.2772 The calculated VLE data are in good agreement with the exper-
20.045 0.8419 0.9063 1.0765 0.5929 imental values for the CO2 + n-nonane and CO2 + n-undecane
systems using temperature dependent parameters. The obtained
deviations in p and y, as well as the optimized parameters are
listed in Table 5, for both systems. As were expected, small
deviations in pressure and vapor mole fraction were obtained
the data points do not exhibit severe fluctuations. Uncertain-
from the correlation with the PR EoS using the Wong–Sandler
ties for isothermal equilibrium ratios for carbon dioxide were
mixing rule. These calculations are plotted in Fig. 1 for the
estimated to be within ±5.0% of the reported value.
The phase equilibrium data for these systems were correlated
Table 4
with the well-known Peng–Robinson equation of state (PR EoS) Critical properties and acentric factor of the pure substances [11]
[5] using classical and Wong–Sandler [6] mixing rules. The PR
EoS is expressed as Tc (K) pc (MPa) ω

CO2 304.12 7.374 0.225

RT a(T ) n-Nonane 594.60 2.290 0.445
p= − (1) n-Undecane 639.00 1.980 0.537
v − b v(v + b) + b(v − b)
 L.E. Camacho-Camacho et al. / Fluid Phase Equilibria 259 (2007) 45–50 49

Table 5
Interaction parameters, deviations in pressure and vapor composition
T (K) PR EoS using (mixing rules)

Classical Wong–Sandler

k12 p (%) y k12 NRTL parameters p (%) y

A12 (kJ mol−1 ) A12 (kJ mol−1 )

CO2 (1) + n-nonane (2)

315.12 0.1009 2.2 0.001 0.6771 9.5872 −1.2644 0.9 0.0003
344.53 0.0989 1.9 0.006 0.6831 10.0994 −1.8890 0.6 0.001
373.28 0.0915 1.7 0.006 0.6923 9.2364 −2.3899 0.5 0.003
418.82 0.0959 1.7 0.011 0.7133 8.9580 −2.9527 1.1 0.006
T.I. 0.0968 2.1 0.010 0.6966 8.1874 −2.0026 4.6 0.009
CO2 (1) + n-undecane (2)
314.98 0.1138 3.1 0.001 0.7305 13.5297 −0.6452 0.7 0.001
344.46 0.1034 6.9 0.006 0.7509 11.1918 −1.2447 1.2 0.006
373.13 0.1099 3.3 0.003 0.7483 13.2764 −2.3272 0.6 0.002
418.30 0.1211 4.9 0.003 0.7740 11.7147 −2.6455 2.0 0.004
T.I. 0.1104 7.1 0.008 0.7486 11.2059 −1.6915 6.3 0.013

T.I.: temperature independent parameters.

CO2 + n-nonane system, and in Fig. 2 for the CO2 + n-undecane Appendix A. The mixing rules
system.
Higher deviations in p and y were found using a temperature The classical mixing rule is defined for the a and b constants
independent (T.I.) parameter to calculate the vapor–liquid equi- as follows:
librium data at the four temperatures in each CO2 + n-alkane 
system compared with those VLE values calculated with a tem- am = xi xj (ai aj )1/2 (1 − kij ) (A1)
perature dependent parameter. These results are also presented i j

in Table 5. 
The obtained parameters at 344.53 K for the CO2 + n-nonane bm = xi bi (A2)
i
system was used to predict those data reported by Jennings
and Schucker [3] at 343.25 K with the Peng–Robinson equa- kij is the interaction parameter for the binary mixture.
tion of state coupled to the Wong–Sandler mixing rule. The The Wong–Sandler mixing rule is represented as
predicted values for the literature data at 343.25 K gave good 
agreement according to the following deviations: p = 0.80% i j xi xj (b − (a/RT ))ij
bm =  (A3)
and y = 0.001. 1 − i xi (ai /(bi RT )) − AE∞ /CRT
with
4. Conclusions
 a (bi − (ai /RT )) + (bj − (aj /RT ))
b− = (1 − kij )(A4)
New high-pressure vapor–liquid equilibria for the CO2 + n- RT ij 2
nonane and CO2 + n-undecane systems were isothermally
measured via a static-analytical set up from 315 to 418 K and
 
in the pressure range of 2–20 MPa. The experimental data  ai AE∞
for the CO2 + n-nonane system at ∼344 K have a good am = bm xi + (A5)
bi C
trend with those reported in literature. The experimental data i
were well correlated by using the PR EoS coupled to the √ √
For the PR EoS C = ln( 2 − 1)/ 2. The excess Helmholtz
Wong–Sandler mixing rule for these CO2 + n-alkane systems.
free energy (AE∞ ) at infinite pressure can be expressed with the
Besides, the predicted VLE data for the CO2 + n-nonane
excess Gibbs free energy GE at low pressure. In this work, we
system using the temperature dependent parameters showed
use the NRTL model [7] given by
good agreement with those published at 343.25 K in Ref.

[3]. AE∞  j xj τji Gji
= xi  (A6)
RT k xk Gki
Acknowledgments i

where
The authors thank to Conacyt and IPN for the financial sup-
port. Gji = exp(−αji τji ) (A7)
50 L.E. Camacho-Camacho et al. / Fluid Phase Equilibria 259 (2007) 45–50

and [5] D.Y. Peng, D.B. Robinson, A new two constant equation of state, Ind. Eng.
Chem. Fundam. 15 (1976) 59–64.
Aji
τji = (A8) [6] D.S.H. Wong, S.I. Sandler, A theoretically correct mixing rule for cubic
RT equation of state, AIChE J. 38 (1992) 671–680.
For both systems the nonrandomness parameter αji was fixed [7] H. Renon, J.M. Prausnitz, Local composition in thermodynamic
excess functions for liquid mixtures, AIChE J. 14 (1968) 135–
to 0.3. Gji is the local composition factor, τ ji are two binary 144.
interaction parameters, and Aji = (gji − gii ), gji is the interaction [8] O. Elizalde-Solis, L.A. Galicia-Luna, High pressure vapor–liquid equilibria
energy between the components i and j. for the systems thiophene + nonane + CO2 , and thiophene + decane + CO2 ,
Fluid Phase Equilib. 236 (2005) 178–183.
References [9] O. Elizalde-Solis, L.A. Galicia-Luna, S.I. Sandler, J.G. Sampayo-
Hernandez, Vapor–liquid equilibria and critical points of the CO2 + 1-
hexanol and CO2 + 1-heptanol systems, Fluid Phase Equilib. 210 (2003)
[1] M.A. McHugh, V.J. Krukonis, Supercritical Fluid Extraction Principles and
215–227.
Practice, 2nd ed., Butterworth-Heinemann, USA, 1994.
[10] P. Guilbot, A. Valtz, H. Legendre, D. Richon, Rapid on-line sampler-
[2] H. Knapp, R. Döring, L. Plöcker, J.M. Prausnitz, Vapor–Liquid Equilibria
injector: a reliable tool for HT-HP sampling and on-line GC analysis,
for Mixtures of Low Boiling Substances, Chemistry Data Series, vol. 6,
Analusis 28 (2000) 426–431.
Dechema, Berlin, 1981.
[11] B.E. Poling, J.M. Prausnitz, J.P. O’Connell, The Properties of Gases
[3] D.W. Jennings, R.C. Schucker, Comparison of high-pressure vapor–liquid
and Liquids, 5th ed., McGraw-Hill, New York, 2001, pp. A6, A15,
equilibria of mixtures of CO2 or propane with nonane and C9 alkylben-
A16.
zenes, J. Chem. Eng. Data 41 (1996) 831–838.
[12] W.H. Press, S.A. Teukolsky, W.T. Vetterling, B.P. Flannery, Numerical
[4] G. Schneider, Z. Alwani, W. Heim, E. Horvath, E.U. Franck, Phasengle-
Recipes in Fortran, vol. 77, 2nd ed., Cambridge University Press, USA,
ichgewichte und kritische Erscheinungen in binären Mischsystemen bis
1999, pp. 678–683.
1500 bar, Chemie-Ing.-Techn. 39 (1967) 649–656 (in German).