CHEM. RES.

CHINESE UNIVERSITIES 2011, 27(4), 678—682

Vapor-liquid Phase Equilibria for CO2+Tertpentanol

Binary System at Elevated Pressures

WANG Lin1,2, LUO Jian-cheng3, YANG Hao1 and CHEN Kai-xun2*

1. Chemistry and Pharmaceutical Engineering College, Nanyang Normal University, Nanyang 473061, P. R. China;
2. Chemical Engineering College, Northwest University, Xi’an 710069, P. R. China;
3. Chemical Engineering College, Nanyang Institute of Technology, Nanyang 473004, P. R. China

Abstract Vapor-liquid phase equilibrium data of tertpentanol in carbon dioxide were measured at temperatures of
313.4, 323.4, 333.5 and 343.5 K and in the pressure range of 4.56―11.44 MPa. The phase equilibium apparatus used
in the work was a variable-volume high-pressure cell. The experimental data were reasonably correlated with
Peng-Robinson equation of state(PR-EOS) together with van der Waals-2 two-parameter mixing rules. Henry’s Law
constants and partial molar volumes of CO2 at infinite dilution were estimated with Krichevsky-Kasarnovsky equa-
tion, and Henry’s Law constants increase with increasing temperature, however, partial molar volumes of CO2 at in-
finite dilution are negative whose magnitudes decrease with temperature. Partial molar volumes of CO2 and tertpen-
tanol in liquid phase at equilibrium were calculated.
Keywords Carbon dioxide; Tertpentanol; Phase equilibrium; Henry’s Law constant; Partial molar volume
Article ID 1005-9040(2011)-04-678-05

1 Introduction design purposes, the measured vapor-liquid equilibrium data

were correlated with Peng-Robinson equation of state
Supercritical CO2 has been widely used in chemical engi- (PR-EOS)[13]; the vapor-liquid equilibrium data of every iso-
neering separation and chemical reactions for its nontoxicity, S
therm were also used to calculate Henry’s law constant( H 1, 2 P2 )
moderate critical conditions, low cost and small impact on the
earth’s environment since the last decade[1,2]. The thermody- ∞
and partial molar volumes of CO2 at infinite diluton( V1 ) with
namic properties and derivative properties(like partial molar
Krichevsky-Kasarnovsky equation[14]; finally, partial molar
quantities) of the systems containing supercritical CO2 in
volumes of CO2 and tertpentanol in liquid phase at equilibrium
high-pressure phase equilibria are required for the design and
were calculated.
operation of separation equipments in many chemical processes
and separation operations. 2 Experimental
The mixtures of supercritical CO2 and alcohols have re-
ceived much attention, for the mixtures are often encountered 2.1 Materials
in the petroleum, oil and natural gas. In addition, CO2 and al-
Carbon dioxide with a purity of 99.99% was purchased
cohol systems have a high potential as supercritical cosolvent
from Shanxi Xinghua BOC Co., Ltd.(China), and tertpentanol
pairs for separating metabolites[3―5]. Up to now, studies on
with a purity of more than 99.9% by mass was provided by
mixtures containing both CO2 and alcohol have been abundant,
Shanghai Chemical Reagent Plant(China). The purities of the
even involving the four-carbon alcohols[6―10], but fewer phase
materials were checked with GC102M gas chromatography.
equilibrium reports on tertpentanol in CO2. Lee et al.[11] re-
They were used directly without further purification.
ported the high-pressure phase equilibrium data for the carbon
dioxide+tertpentanol system at 313.2 K in 1999. Heo et al.[12]
2.2 Apparatus
studied the single-phase limit of tertpentanol+CO2 system at
temperatures from 313.15 K to 353.15 K and pressures from 3 The experimental apparatus used in the work was a varia-
MPa to 24 MPa in 2001. However, these references could not ble-volume visual cell similar to that in the literature[15]. A
provide the full data of phase equilibrium and thermodynamic schematic diagram of the variable-volume visual cell is shown
properties of the system. Therefore, in the present work, the in Fig.1. The cell has one quartz visual window and its internal
high pressure vapor-liquid phase equilibrium data for the binary volume can be adjusted from 25 mL to 100 mL by moving an
system of tertpentanol+CO2 from 313.4 K to 343.5 K were internal movable piston, which separates the fluid in the
measured in a varible-volume visual cell. Furthermore, to pro- cell from pressure medium(water). The stainless steel piston
vide phase equilibrium information for high-pressure process is sealed with double O-rings against the pressures of the

———————————
*Corresponding author. E-mail: [email protected]
Received November 17, 2010; accepted March 4, 2011.
Supported by the Youth Key Teacher Foundation of Henan Educational Committee, China(No.2010GGJS-162).
No.4 WANG Lin et al. 679
equilibrium cell. The inner diameter of the cell is 22.12 mm. volume by capillary and needle-type valve. After the desorption
The maximum operating temperature and pressure of the appa- tank was evacuated, carbon dioxide was desorbed to the de-
ratus are 423 K with an accuracy of 0.1 K and 20 MPa with an sorption tank very slowly until the pressure of the tank didn’t
accuracy of 0.01 MPa, respectively. The pressure in the cell is change. The amount of CO2 in each phase was determined by
generated via the pressurization of the piston driven by a measuring the corresponding pressure variation in the desorp-
hand-pump. The equilibrium pressure in the cell is measured by tion tank. The mass of tertpentanol was equal to the sample
a CYB-20S differential pressure gauge which is calibrated with mass minus CO2 mass. The content of water in tertpentanol was
a standard pressure gauge; cell temperature is controlled by a detected with gas chromatography, and the O-rings were re-
ZCY-15B digit constant temperature encircling liquid bath placed when the content of water was over 0.5%. The molar
controller and measured by a calibrated thermocouple. The fractions in two phases were calculated.
motion distance of the piston is measured with a NS-YB05C- The volume of the sample was calculated by the product
A-I-0-0-1 displacement meter with an accuracy of 0.01 mm, of the known inner diameter of the cell and the distance
which is confirmed with a standard micrometer. A magnetic between the positions of the piston before and after taking the
stirrer which is driven by a pneumatic drive pump is used for sample. The densities of the vapor and liquid phases were
mixing the fluid in the cell. The reliability of measurement was obtained via dividing the appropriate mass by the volume of
tested by comparation with the published data of an ethanol+ each phase. Finally, the molar volumes of the vapor and liquid
CO2 system[16]. phases were obtained from sampling volume and the composi-
tions.
The reproductivity of the composition results was verified
by several measurenents under the same pressure and tempera-
ture. Vapor and liquid phase compositions were found repro-
ducible with a molar fraction of ±0.0008 and ±0.0010.

3 Results and Discussion

3.1 Experimental Data
To test the reliability of the apparatus, the vapor-liquid
Fig.1 Schematic diagram of the variable-volume
equilibrium data of CO2+ethanol was measured at 333.40 K.
visual cell
The data were compared with the highly accurate data at
1. Handpump; 2. digit constant temperature liquid bath; 3. magnet;
4. liquid phase sampling valve; 5. intake valve; 6. variable- 333.40 K by Joung et al.[16]. The comparison is drawn in Fig.2,
volume cell; 7. high pressure pump; 8. refrigerator; 9. CO2 in which xCO2 and yCO2 are the molar fractions of CO2 in liquid
cylinder; 10. displacement indicator; 11. temperature indicator; phase and vapor phase, respectively. The results show that both
12. pressure indicator; 13 quartz visual window; 14. thermocouple;
the measured and literature data of CO2+ethanol agree well.
15. vapor phase sampling valve; 16. pressue gauge; 17. safety
valve; 18. magneton; 19. displacement meter; 20. water cup. Based on the comparison we conclude that the experimental
apparatus used in the present study is reliable and can generate
2.3 Experimental Procedure accurate vapor-liquid data.

Before each experiment, the entire internal loop of the

equilibrium cell was rinsed several times with carbon dioxide
to remove traces of air. Then, the equilibrium cell was eva-
cuated with a vacuum pump. A known amount of tertpentanol
was sucked in the equilibrium cell, and CO2 was pressurized
into the cell from CO2 inlet valve. The cell was slightly pressu-
rized to the experimental pressure by the hand-pump and was
heated to the experimental temperature with isothermal liquid.
Once the pressure and temperature were adjusted to designated
values, the mixture was stirred for a few hours to equilibrium
state. Then, the stirrer was switched off. About 1.5 h later, the
coexisting phases were separated completely. Phase interface Fig.2 Phase behavior of CO2+ethanol at 333.40 K
could be clearly observed from the visual window. Samples □ Data obtained in the work; ■ data reported in ref. [16].
of the vapor and liquid phases were collected with a small Vapor-liquid equilibrium data for CO2+tertpentanol binary
steel vessel from the upper and lower sampling valves by system at 313.4, 323.4, 333.5 and 343.5 K were measured with
pushing hand pump slowly, respectively. The samples were the apparatus. The results are listed in Table 1, where x1 and y1
weighed up with a FA1104 electronic balance with an accuracy are the molar fractions of CO2 in liquid phase and vapor phase,
of 0.0001 g. respectively, and Vm,l and Vm,v are the molar volumes of liquid
The small steel vessel after sampling was placed in an ice- phase and vapor phase, respectively，and ρl and ρv are the
water bath and connected with a desorption tank of calibrated densities of liquid phase and vapor phase, respectively.
680 CHEM. RES. CHINESE UNIVERSITIES Vol.27
Table 1 Vapor-liquid phase equilibrium data for CO2(1)+tertpentanol(2) system
T/K p/MPa x1,expt. y1,expt. ρl/(g·mL–1) ρv /(g·mL–1) Vm,l /(mL·mol–1) Vm,v /(mL·mol–1)
313.4 4.56 0.3587 0.9939 0.8158 0.1031 88.6411 429.3201
5.02 0.4384 0.9937 0.8162 0.1184 84.2865 374.0127
5.88 0.5583 0.9930 0.8153 0.1521 77.8924 291.3101
6.54 0.6462 0.9929 0.8116 0.1845 73.4675 240.2024
7.06 0.7569 0.9916 0.7951 0.2196 68.8501 202.2116
7.46 0.8213 0.9907 0.774 0.2538 67.0512 175.4613
7.76 0.8941 0.9862 0.7221 0.2932 67.4224 151.7104
7.99 0.9236 0.9768 0.6856 0.4738 69.1143 95.2800
323.4 4.81 0.3206 0.9977 0.7999 0.1024 92.5101 430.8103
5.78 0.4109 0.9964 0.7986 0.1332 87.6718 331.9315
6.56 0.5177 0.9956 0.7947 0.1618 82.1694 273.2424
7.55 0.6397 0.9920 0.7851 0.2104 76.3201 210.8201
8.61 0.7961 0.9858 0.7455 0.3066 71.1104 146.2112
9.04 0.8541 0.9793 0.7109 0.3639 70.9724 123.4100
9.44 0.9182 0.9583 0.6349 0.5191 74.9900 88.3016
333.5 5.21 0.3099 0.9904 0.7884 0.1082 94.4612 410.8121
6.16 0.3777 0.9935 0.7867 0.1349 90.8634 328.3201
7.24 0.4921 0.9948 0.7807 0.1706 85.0901 259.3102
8.28 0.5858 0.9892 0.7727 0.2237 80.6215 199.9212
9.17 0.7089 0.9826 0.748 0.2739 76.0217 163.5123
9.26 0.7240 0.9810 0.7431 0.2824 75.6207 158.8404
10.03 0.8092 0.9713 0.7043 0.3614 74.4541 125.3228
10.15 0.8235 0.9658 0.6946 0.4071 74.5804 112.3811
10.37 0.8377 0.9597 0.6852 0.4284 74.6768 106.8891
10.62 0.9159 0.9253 0.5767 0.5544 82.7398 85.3201
343.5 5.33 0.2409 0.9961 0.7658 0.1033 101.2112 427.9103
6.33 0.3206 0.9935 0.7589 0.1311 97.5201 340.6213
7.3 0.4075 0.9911 0.7505 0.1597 93.5202 278.1402
8.33 0.5161 0.9901 0.7373 0.195 88.7445 227.9201
9.51 0.6389 0.9794 0.7153 0.2668 83.8078 170.3716
10.27 0.7056 0.9697 0.6975 0.3065 81.7303 147.9027
11.44 0.8908 0.9416 0.5534 0.4407 88.2325 105.7205

The corresponding isothermal pressure-composition dia- liquid phase decreases as the pressure increases. In the course
grams are drawn in Fig.3. Meanwhile, the reported values in of approaching to critical pressure, subtle change of pressure
the literature[11] are also drawn in Fig.3. As can be seen from can cause the noticeable variation of densities of vapor and
Fig.3, the experimental data in the work at 313.4 K are in good liquid phases.
agreement with those reported by Lee et al.[11] at 313.2 K. The
experimental results also show that the solubility of CO2 in
tertpentanol increase with pressure.

Fig.4 p-ρ diagrams for CO2+tertpentanol system

at different temperatures
■,□ 313.4 K; ●,○ 323.4 K; ▲,△ 333.5 K; ▼,▽ 343.5 K. Solid points
Fig.3 p-x(y) diagrams for CO2(1)+tertpentanol(2) represent the liquid phase, and hollow points represent vapor phase.
system at different temperatures Fig.5 gives Vm-x(y) diagram for CO2+tertpentanol at dif-
■,□ 313.4 K; ●,○ 323.4 K; ▲,△ 333.5 K; ▼,▽ 343.5 K; — calculated
ferent temperatures, where the thin lines are isopiestic conju-
value; ref.[11]: ×┈313.2 K. Solid points represent the liquid phase,
and hollow points represent vapor phase. gating lines. At a constant temperature, the molar volume of the
Fig.4 gives p-ρ diagram for CO2+tertpentanol binary sys- liquid phase decreases slowly with increasing pressure, but the
tem at four different temperatures. At a constant temperature, molar volume of the vapor phase decreases sharply with
the density of the vapor phase increases and the density of the increasing pressure.
No.4 WANG Lin et al. 681

Fig.5 Vm-x(y) diagrams of CO2+tertpentanol system at different temperatures

● Liquid phase; ○ vapor phase. Temperature/K: (A) 313.4; (B) 323.4; (C) 333.5; (D) 343.5.

Table 3 Binary interaction parameters(k12 and n12) and

3.2 Data Correlation with EOS-based Model average relative deviation for CO2+tertpentanol
In this work, the Peng-Robinson equation of state(PR- system from PR-EOS together with van der
EOS)[13] together with Panagiotopoulous-Reid[17], Stryjek-Vera Waals-2 two-parameter mixing rule
mixing rules[18], van der Waals-1[19] or van der Waals-2[20] T/K p/MPa k12 n12 AARDxa(%) AARDyb(%)
mixing rules was chosen for description of the phase behavior 313.4 4.56―7.99 0.0818 –0.0556 1.2046 1.1967
323.4 4.81―9.44 0.0820 –0.0658 0.8211 0.8118
of the binary system, respectively. Having been tried, PR-EOS
333.5 5.21―10.62 0.0876 –0.0613 0.6214 0.6177
coupled with van der Waals-2 mixing rule was found to be the
343.5 5.33―11.44 0.0611 –0.1145 0.8851 0.8745
most accurate for the description of the phase behavior of the
1 M
binary system. a. AARDx(%) = ∑ ( x1, exp t. − x1, calcd. ) / x1, exp t. × 100%;
N i =1 i
The objective function(F) applied in the work is
1 M
⎡⎛ b. AARDy(%) = ∑ ( y1, expt. − y1, calcd.) / y1, expt. ×100%.
x i , calcd. ⎞ ⎤
2
1 M N
y i , calcd. ⎞ ⎛ (1) N i =1
∑∑ ⎢ ⎜ ⎟ ⎜ ⎟ ⎥
i
F = 1 − + 1 −
M i = 1 j = 1 ⎢ ⎜⎝ y i , expt. ⎟⎠ ⎜⎝ x i , expt. ⎟⎠ ⎥
⎣ ⎦ 3.3 Data Correlation with Krichevsky-Kasar-
where M and N are the number of acquired phase equilibrium
novsky Equation-based Model
and the number of components in the system, respectively.
Subscripts calcd. and expt. represent calculated and experi- Krichevsky-Kasarnovsky equation[14] was used to
mental values, respectively. represent the dependence of the solubility of CO2 in liquid
The critical properties and acentric factors for the pure phase on pressure and given by equation (2):
component used in the work came from ref. [21] that are sum- ⎛ fl⎞
∞
Ps V1 (2)
marized in Table 2. ln ⎜⎜ 1 ⎟⎟ = ln H 1, 2 2 + ( p − p 2s )
x
⎝ 1⎠ RT
Table 2 Critical values and acentric factors for
where subscripts 1 and 2 represent CO2 and tertpentanol, re-
the pure component
spectively; f is fugacity.
Component pc /MPa Tc /K w
By linear fitting ln(f1l/x1) as a function of (p–p2s)/(RT),
CO2 7.376 304.20 0.225
H1,2 and V1 can be obtained. The application of equation
P2
s ∞
Tertpentanol 3.790 545.00 0.478
The optimized binary interaction parameters kij, nij and de- (2) requires the knowledge of saturated vapor pressures of tert-
rivation analyses with PR-EOS and van der Waals-2 mixing pentanol at experimental temperatures, which comes from the
rule are summarized in Table 3. The results calculated from literature[22]. The optimized parameters and deviation analys-
these binary interaction parameters at four different experi- es based on the Krichevsky-Kasarnovsky equation[14] are
mental temperatures were obtained that are also drawn in Fig.3 summarized in Table 4. The results in Table 4 show that Hen-
in order to direct comparison with experimental data. Results in ry’s Law constants for CO2+tertpentanol binary system and
Fig.3 also show the calculated results agree well with the expe- partial molar volumes of CO2 at infinite dilution only depend
rimental data. Therefore, PR-EOS coupled with van der on the temperature, and Henry’s law constant increased with
Waals-2 two-parameter mixing rule can be applied to the better increasing temperature and partial molar volume of CO2 at
description of the phase behavior for CO2 +tertpentanol binary infinite dilution are negative in value whose magnitude de-
system at high pressures and near-critical areas. crease with temperature.
682 CHEM. RES. CHINESE UNIVERSITIES Vol.27
[13]
Table 4 Thermodynamic parameters of solubility of MPa to 11.44 MPa. Peng-Robinson equation of state toge-
CO2 in tertpentanol based on Krichevsky- ther with van der Waals-2 two-parameter mixing rule[22] could
Kasarnovsky model describe well the phase behavior for CO2 +tertpentanol binary
T/K Ps
H 1, 2 2 /MPa ∞ –1
V1 /(mL·mol ) ra SDb AARDc(%) system at elevated pressures and near-critical areas. Henry’s
313.4 21.2063 –439.5539 –0.9982 d
0.0142 0.0459 Law constant and partial molar volume of CO2 at infinite dilu-
323.4 23.9611 –380.1028 –0.9992d 0.0108 0.0353 tion were estimated with Krichevsky-Kasarnovsky equation.
333.5 25.1769 –371.0424 –0.9974d 0.0166 0.0725 Partial molar volumes of CO2 and tertpentanol at equilibrium
343.5 35.7568 –323.3361 –0.9984d 0.0187 0.0366 were calculated. The phase equilibrium data, predicted models
a. Corrective coefficient; b. standard deviation; c. average relative obtained and partial molar volume properties are expected to be
deviation; d. highly significant.
a useful tool for the process design and recovery of tertpentanol
with supercritical CO2 technique.
3.4 Calculation of Partial Molar Volumes at
Equilibrium
References
Partial molar volumes of CO2 and tertpentanol in liquid [1] Liu Z. M., Han B. X., Adv. Mater., 2009, 21, 825
phase at vapor-liquid equilibrium are determined by partial [2] Yuan H. B., Shang L. N., Wei C.Y., Ren B. Z., Chem. Res. Chinese
molar properties[23] and PR-EOS. The expressions are as Universities, 2010, 26(6), 888
follows: [3] Fourie F. C. V. N., Schwarz C. E., Knoetze J. H., J. Supercrit.
RT ∂Z Fluids, 2008, 47, 161
V1 = [ Z + x2 ( )] , V2 = RT [ Z − x1 ( ∂Z )] (3)
P ∂x1 P ∂x1 [4] Schwarz C. E., Fourie F. C. V. N., Knoetze J. H., J. Supercrit.
in which Fluids, 2009, 51, 128
∂B ∂A [5] López J. A., Trejos V. M., Cardona C. A., Fluid Phase Equil., 2009,
[2 Z (1 + 3 B ) − Z + A − B (2 + 3 B )]( ) + ( B − Z )( )
2

∂Z ∂x1 ∂x1 275, 1

=
∂x1 3Z 2 − 2(1 − B ) Z + ( A − 2 B − 3 B 2 ) [6] Wang L., Luo J. C., Li R. L., Chen K. X., Chem. J. Chinese
∂A P Universities, 2010, 31(7), 1416
= [2a1 x1 − 2a2 x2 + 2(1 − 2 x1 )(1 − k12 ) a1a2 ] (4)
∂x1 R 2T 2 [7] Wang L., Hao X. S., Zheng L., Chen K. X., Chinese J. Chem. Eng.,

∂B P 2009, 17, 642

= [2b1 x1 − 2b2 x2 + (1 − 2 x1 )(1 − n12 )(b1 + b2 )] [8] Wang L., Li R. L., Sun L. T., Chen K. X., Chem. J. Chinese
∂x1 RT
Universities, 2009, 30(8), 1631
The results are presented in Fig.6. As can be seen from
l
[9] Jennings D. W., Lee R. J., Teja A. S., J. Chem. Eng. Data, 1991,
Fig.6, at equilibrium, V1 is positive and increases with in- 36, 303
l
creasing pressure, and V2 is positive at low pressure and [10] Chen H. I., Chen P. H., Chang H. Y., J. Chem. Eng. Data, 2003, 48,
l
decreases with increasing pressure. Finally, V2 becomes ne- 1407
gative at subcritical area. The above results show that the addi- [11] Lee H. S., Mun S. Y., Lee H., Fluid Phase Equil., 1999, 157, 81
tion of CO2 and tertpentanol in liquid phase will cause dilato- [12] Heo J. H., Shin H. Y., Park J. U., Joung S. N., Kim S. Y., Yoo K. P.,
metric effect of mixing volume at equilibrium. Furthermore, the J. Chem. Eng. Data, 2001, 46, 355
dilatometric effect of CO2 increases and the dilatometric effect [13] Peng D. Y., Robinson D. B., Ind. Eng. Chem. Fundam., 1976, 15,
of tertpentanol decreases with the ascent of pressure. 59
[14] Krichevsky I. R., Kasarnovsky J. S., J. Am. Chem. Soc., 1935, 57,
2168
[15] Han F., Xue Y., Tian Y. L., Zhao X. F., Chen L., J. Chem. Eng.
Data, 2005, 50(1), 36
[16] Joung S. N., Yoo C. W., Shin H. Y., Kim S. Y., Yoo K. P., Lee C. S.,
Huh W. S., Fluid Phase Equil., 2001, 185, 219
[17] Panagiotopoulos A. Z., Reid A. C., Watanasiri S., Equation of
State, Theories and Applications, ACS Symp. Sere. 300, American
Chemical Society, Washington DC, 1986, 156
[18] Stryjek R., Vera J. H., Can. J. Chem. Eng., 1986, 64, 323
Fig.6 Partial molar volume properties for CO2+ [19] Soave G., Chem. Eng. Sci., 1972, 27, 1197
tertpentanol system at equilibrium [20] Soave G., Chem. Eng. Sci., 1984, 39, 357
■□ 313.4 K; ●○ 323.4 K; ▲△ 333.5 K; ▼▽ 343.5 K. [21] Bruce E. P., John M. P., John P. O.; Trans. Zhao H. L., Wang F.
Solid points represent V1l , and hollow points represent V2 l . K., Chen S. K., The Properties of Gases and Liquids, Chemical
Industry Press, Beijing, 2006, 260
4 Conclusions [22] Liu G. Q., Ma L. X., Liu J., Chemical Property Data Handbook,
Chemical Industry Press, Beijing, 2002, 602
The vapor-liquid equilibrium data for CO2+tertpentanol [23] Stanley M. W., Phase Equilibria in Chemical Engineering, But-
binary system were measured in a variable-volume view cell at ter-worth, New York, 1985, 137
temperatures from 313.4 K to 343.5 K and pressures from 4.88