Vapor−liquid equilibrium measurements of the binary mixtures CO2 + acetone

and CO2 + pentanones

Chieh-Ming Hsieh*,a and Jadran Vrabecb

a
Department of Chemical and Materials Engineering, National Central University, Jhongli,

Taiwan
b
Thermodynamics and Energy Technology, University of Paderborn, Paderborn, Germany

Abstract

Vapor−liquid equilibrium (VLE) data of the binary mixtures carbon dioxide (CO2)

+ acetone and CO2 + pentanones (i.e. all C5-ketones, which are 3-methyl-2-butanone,

2-pentanone, and 3-pentanone) are measured at 313.15 K, 333.15 K, and 353.15 K up

to a pressure of 11.8 MPa using a high pressure view cell technique based on the

synthetic method. The VLE measurement results of CO2 + acetone are compared to

the available experimental data from the literature in order to validate the employed

experimental apparatus. In case of binary CO2 + pentanone mixtures, only for CO2 +

3-methyl-2-butanone and CO2 + 2-pentanone at 313.15 K experimental data are

available in the literature. The present measurements thus complete the literature

database for binary VLE of CO2 + pentanones. All present data are compared to the

results from the Peng−Robinson equation of state (PR EOS) with the van der Waals

one-fluid mixing rule and from the combination of the PR EOS with the UNIQUAC

model through the Huron−Vidal mixing rule.

Keywords: Peng-Robinson equation of state, phase equilibrium, experiment, carbon

dioxide, ketones

*Corresponding author. Email [email protected]. Tel: +886-3-4227151#34220

1
1. Introduction

Carbon dioxide (CO2) is a popular solvent in industrial applications because it is

nontoxic, nonflammable, and inexpensive and has a moderate critical temperature and

critical pressure. Supercritical carbon dioxide (ScCO2) is widely used in industry, such

as for extraction of natural flavors, decaffeination of coffee beans, or as a solvent in

polymer synthesis processes [1-8]. Sometimes cosolvents are necessary either to

increase the solubility of the solute in ScCO2 or to modify the properties of ScCO2.

Thermodynamic properties, in particular vapor-liquid equilibrium (VLE) data, for

mixtures of ScCO2 and cosolvents are important for designing and optimizing

processes containing these mixtures [9, 10]. E.g., operating conditions during

crystallization processes that employ supercritical anti-solvents strongly affect the size

distribution and the mean size of particles [7, 11, 12].

The most reliable way to obtain thermodynamic properties and fluid phase

equilibrium data is by experiments. These data are necessary to evaluate the accuracy

of thermodynamic models. Several publications have focused on VLE of ScCO2 + low

molecular weight cosolvents because they are good candidates for the extraction of

drugs and food ingredients [5, 13-15]. The most popular cosolvents for ScCO2 are low

molecular weight alcohols because they can greatly increase the solubility of polar

solutes in ScCO2 [1, 6, 13, 14, 16]. The phase behavior of ScCO2 + alcohols was

systematically investigated by several research groups [17-22]. Compared with

alcohols, relatively few data were measured for other types of solvents, such as

ketones or ethers, which are potential cosolvents for ScCO2 in some applications [5].

E.g., the solubility of pharmaceutical compounds in ScCO2 + acetone was extensively

studied by experimental measurements [12, 13], thermodynamic models [23, 24], and

2
molecular simulation [25]. It was found that acetone may be a good cosolvent

candidate in processes for recovering nondegraded triglycerides from used frying oil

and in the extraction of cocoa butter [26, 27]. Acetone is the most popular ketone

cosolvent for ScCO2 applications. The main reason may be that acetone is a very

popular solvent in general. Another possible reason may be that insufficient

information (such as experimental data) is available for other ketones.

In this study, the isothermal VLE of CO2 + ketones at 313.15 K, 333.15 K, and

353.15 K were systematically investigated by experiments using a high pressure view

cell technique based on the synthetic method. There are numerous measured data

available for CO2 + acetone (the only C3-ketone) and CO2 + 2-butanone (the only

C4-ketone) [28], but there are very few experimental data in the literature for binary

mixtures CO2 + pentanones (C5-ketones). Only for CO2 + 3-methyl-2-butanone and

CO2 + 2-pentanone at 313.15 K data were reported. The first step in this work was to

measure VLE for the binary mixture CO2 + acetone at three temperatures to validate

the measurements of the employed experimental apparatus on the basis of data from

the literature. Next, VLE of binary mixtures CO2 + pentanones (3-methyl-2-pentanone,

2-pentanone, and 3-pentanone) were measured at 313.15 K, 333.15 K, and 353.15 K

and also compared with the available experimental data from the literature. The

present measurements thus complete the literature database of binary VLE for CO2 +

ketones up to pentanones. Furthermore, all present data were compared to the results

from the Peng−Robinson equation of state (PR EOS) [29] with the van der Waals

one-fluid mixing rule and from the combination of the PR EOS with the UNIQUAC

model [30] through the Huron−Vidal mixing rule [31].

3
2. Materials and methods

2.1. Materials

Acetone (purity ≥ 99.9%, CAS No.: 67-64-1), 3-methyl-2-butanone (purity 99%,

CAS No.: 563-80-4), 2-pentanone (purity 99.5%, CAS No.: 107-87-9), and

3-pentanone (purity 99%, CAS No.: 96-22-0) were purchased from Sigma-Aldrich,

Germany. Carbon dioxide (volume fraction 99.995%) was supplied by Air Liquide,

Germany. All chemicals were used without further purification.

2.2. Experimental apparatus and procedure

Figure 1 shows the apparatus for experimentation in this study. This setup was

used for VLE measurements in prior work [17, 32]. The only modification made for

this study was that the surrounding climate chamber was removed because all ketones

included in this study are liquid at ambient temperature. All pressure transducers (P1

and P2, model Super THE, Honeywell test & measurement with measuring ranges 6.8

MPa and 20 MPa and an uncertainty of ± 0.05 % of their full measuring ranges),

platinum resistance thermometers (Pt1 to Pt4, with a basic resistance of 100 Ω and an

uncertainty of ± 0.04 K), and spindle presses A and B were calibrated. Briefly

summarized, the measurement procedure was as following:

1. Liquid CO2 from the gas cylinder was loaded into the calibrated spindle press A

which was cooled down by ice water to about 275 K.

2. Liquid ketone was filled into the calibrated spindle press B from the reservoir.

3. The high pressure view cell was filled with the desired quantity of ketone from

spindle press B and heated up to the measuring temperature.

4. Liquid CO2 from spindle press A was slowly loaded into the view cell until it was

4
 observed that the CO2 + ketone mixture was in a homogenous liquid state. The

mixing process was accelerated by a magnetic stirrer in the view cell.

5. The mixture in the view cell was maintained in its homogeneous liquid state for at

least 20 minutes.

6. Spindle press A was operated to decrease the pressure in the view cell in very

small steps. This expansion was terminated at the saturated liquid state, i.e. when

the first tiny vapor bubbles appeared in the view cell.

7. Temperature and vapor pressure of the mixture were measured with the

thermometers Pt1 and Pt2 and the pressure transducer P2. The mole fraction of

CO2 (xCO2) was obtained straightforwardly by the procedure described in previous

work [17, 32]. The error analysis in terms of xCO2 was taken from Ref. [33] and is

briefly summarized in the Supplementary Material.

3. Thermodynamic models

Two approaches, both based on the PR EOS [29], were used to correlate the

experimental data of the four binary mixtures studied in this work. The PR EOS

describes the relation between temperature T, pressure P, and molar volume v of a

fluid

RT a
P= − , (1)
v − b v (v + b ) + v (v − b )

where R is the gas constant. For a pure fluid, the substance-specific parameters, i.e.

the energy parameter a and covolume parameter b, were determined from the critical

temperature Tc, the critical pressure Pc, and the acentric factor ω by

5
 2
R 2Tc2   T 
a (T ) = 0.457235 1 + κ 1 −  , (2)
Pc   Tc 

RTc
b = 0.077796 , (3)
Pc

where κ = 0.37464 + 1.54226ω − 0.26992ω 2 . The required pure substance properties

for all relevant substances are summarized in Table 1.

In the case of mixtures, the energy parameter a and the covolume parameter b are

not only a function of temperature and pure substance properties, but also a function

of mole fraction. Therefore, a mixing rule must be used to consider the concentration

effects on these two parameters. In this study, two mixing rules were assessed: the van

der Waals one-fluid mixing rule and the Huron-Vidal mixing rule [31].

3.1. Van der Waals one-fluid mixing rule

The van der Waals one-fluid mixing rule is the simplest mixing rule. It assumes

that a mixture is a pseudo one-fluid composed of all substances in the mixture and its

energy and covolume parameters are given by

n n
a(T , x) = ∑∑ xi x j a i (T )a j (T ) (1 − kij ) , (4)
i =1 j =1

n
b( x) = ∑ xibi , (5)
i =1

where kij is an adjustable binary interaction parameter between species i and j, n the

number of components in the mixture, and xi the mole fraction of component i. kij is

set to be zero in prediction mode or obtained from correlation to experimental data in

correlation mode. Values for the binary interaction parameter kij in correlation mode

were obtained from the regression of experimental VLE data generated in this study

and are summarized in Table 2. These two models are denoted as PR+VDW and
6
PR+VDWkij in the following. The results are discussed in section 4.

3.2. Huron-Vidal mixing rule

The Huron-Vidal mixing rule is based on the excess Gibbs free energy Gex as

supplied by additional models. It also specifies the covolume parameter b with Eq. (5).

The expression for the energy parameter a is

a (T , x) n a (T ) G ex
= ∑ xi i + , (6)
b( x ) i =1 bi C HV

12+ 2 
where CHV is an equation of state dependent constant, i.e. C HV = ln  in
2 2  2 - 2 

case of the PR EOS. Gex is the excess Gibbs free energy determined from an activity

coefficient model, G = RT ∑i xi ln γ i . In this study, the UNIQUAC model [30] was

ex

used for this task

ln γ i = ln γ icomb + ln γ ires . (7)

The Staverman−Guggenheim combinatorial term [34, 35] was taken to consider

molecular size and shape effects

ln γ icomb = 1 −
ri
+ ln
ri z  r
− qi 1 − i
∑xq
j j j
+ ln
ri ∑xq j j j

 . (8)
∑ j x j r j ∑ j x j r j 2  qi ∑xr
j j j qi ∑xr j j j 

The expression for the residual term is

 ∑ j x j q jτ ji − x j q jτ ij 
ln γ ires = qi 1 − ln ∑ j x q τ  , (9)
 ∑j x jq j ∑k k k kj 
where z = 10 is the coordination number. ri and qi are relative van der Waals volume

and surface area of substance i, respectively. In this study, the values of these two

parameters were either taken from the literature [30] or determined from the UNIFAC

group contribution method [36]. τij = exp(-∆uij / RT) is an empirical parameter and ∆uij
7
is the binary interaction parameter between components i and j. The values for the

binary interaction parameter ∆uij were obtained from a regression to experimental

VLE data generated in this study and are summarized in Table 2. This approach is

denoted as PR+HV+UNIQUAC in the following.

4. Results and discussion

4.1 VLE measurements of CO2 + acetone

The aim of measuring VLE of the binary mixture CO2 + acetone was to validate

the employed experimental apparatus and operation procedure. This is the simplest

system of binary mixtures CO2 + ketones and there are several sets of experimental

data available in the literature at 313.15 K [37-46], 333.15 K [42, 47-49], and 353.15

K [47-49]. VLE data of this system were measured here at these three temperatures in

a pressure range from 4 MPa to 10 MPa with respect to the saturated liquid line. The

measurement results are summarized in Table 3 and are compared with the available

experimental data from the literature in Fig. 2. As illustrated there, the relationship

between the mole fraction of CO2 (xCO2) and the vapor pressure is almost linear at all

temperatures, except for the critical region. The solubility of CO2 in acetone is

decreased with rising temperature. At 313.15 K, the present measurement results are

consistent with those of Katayama et al. [41], Adrian and Maurer [42], Chiu et al. [43],

and Stievano and Elvassore [43], but are slightly lower than the data of Aida et al. [46]

and are slightly higher than those of Day et al. [37, 40]. Nevertheless, the

experimental data from eight different sources, including our measurements, do not

deviate too much from each other (the maximum deviation is about 0.4 MPa at xCO2 ≈

0.52 mol/mol). At 333.15 K and 353.15 K, the present measurement results are also in

8
good agreement with those of Sato et al. [47], Aida et al. [46], and Wu et al. [49].

However, the experimental data reported by Han et al. [48] are obviously lower than

those measured by other investigators.

4.2 VLE measurements of CO2 + pentanones

The VLE measurement results of the three binary mixtures CO2 + pentanones at

313.15 K, 333.15 K, and 353.15 K are summarized in Table 3. They are compared

with other experimental data, if available in the literature, for CO2 +

3-methyl-2-butanone in Fig. 3, for CO2 + 2-pentanone in Fig. 4, and for CO2 +

3-pentanone in Fig. 5. The temperature dependence of CO2 solubility in pentanones

and the relationship between xCO2 and the vapor pressure for these three binary

mixtures are qualitatively the same as in case of CO2 + acetone.

As shown in Figs. 3 to 5, only the binary mixtures CO2 + 3-methyl-2-butanone

and CO2 + 2-pentanone at 313.15 K were investigated by Aida et al. [46]. No other

experimental data are available in the literature for these two binary mixtures at

333.15 K and 353.15 K nor for the binary mixture CO2 + 3-pentanone at any

temperature. Therefore, our measurements enrich the literature database for binary

VLE of CO2 + pentanones significantly. The present measurement results of the

binary mixture CO2 + 3-methyl-2-butanone at 313.15 K are in good agreement with

those of Aida et al. (Fig. 3). However, the present vapor pressure data of the binary

mixture CO2 + 2-pentanone are slightly higher than those of Aida et al. [46] (Fig. 4),

especially around xCO2 = 0.3 ~ 0.7 mol/mol. A possible explaination is that the purity

of 2-pentanone in the measurements of Aida et al. [46] was insufficient (in Ref. [46],

> 95% was reported). Such impurities (up to 5% in 2-pentanone [46]), which are

usually heavy components, will typically lead to a lower vapor pressure. Therefore,

9
we believe that the present measurements are more reliable because 2-pentanone with

a high purity of 99.5% was used here. Furthermore, several data points were randomly

chosen and measured again to check for reproducibility.

4.3 VLE data correlation with thermodynamic models

The VLE data were correlated with two different approaches, PR+VDWkij (only

one binary interaction parameter kij per binary mixture in the van der Waals one-fluid

mixing rule) and PR+HV+UNIQUAC (two binary interaction parameters ∆uij and ∆uji

per binary mixture are necessary). Except for experimental data reported by Han et al.

[48] for CO2 + acetone and by Aida et al. [46] for CO2 + 2-pentanone, all available

experimental VLE data, including the present measuring results, were considered in

the correlations. The correlated values of all binary interaction parameters are

summarized in Table 2.

As shown in Figs. 2 to 5, the correlation results of PR+HV+UNIQUAC (solid

lines) for the four binary mixtures considered in this study are in very good agreement

with the experiments, including the temperature dependence. The predictions by

PR+VDW (dashed lines) always underestimate the vapor pressure at a given xCO2 for

these four binary mixtures. These deviations of PR+VDW from experiments were

eliminated by introducing one binary interaction parameter kij per binary mixture,

leading to the PR+VDWkij model. The results of PR+VDW are illustrated in Figs. 2

to 5, but not those of PR+VDWkij, because they are very similar to those of

PR+HV+UNIQUAC.

Overall, the correlation of VLE data for CO2 + acetone and CO2 + pentanones

with PR+VDWkij and PR+HV+UNIQUAC was investigated in this study. Both

approaches have a similar accuracy in correlating experimental VLE data for the

10
considered mixtures. The correlated binary interaction parameters as given in Table 2

can be used in further applications.

4.4 Comparing VLE data of four binary mixtures at 333.15 K

The present results from experiment and correlation of all four binary mixtures at

333.15 K are compared in Fig. 6. In addition to the P-xCO2 phase diagram (top), a

deviation plot (bottom) is shown. The correlation results of the CO2 + acetone mixture

were used as the baseline to better distinguish these four systems, where

∆P ( x CO 2 ) = P( x CO 2 ) − PCO2
correlation
+ acetone ( x CO 2 ) at the same mole fraction xCO2. The vapor

pressure deviations between measurements and correlation for the CO2 + acetone

binary mixture are always positive (∆P ≈ 0.15 MPa) because the correlation was done

using all available experimental VLE data. At xCO2 = 0 mol/mol, the vapor pressure of

pure acetone is higher than that of all pentanones. In the xCO2 < 0.4 mol/mol range, the

correlation line shows that the vapor pressure of the three CO2 + pentanone mixtures

is higher than that of CO2 + acetone above a certain mole fraction xCO2. The tendency

of vapor pressure of these four binary mixtures at xCO2 = 0.4 mol/mol is P(CO2 +

3-pentanone) ≈ P(CO2 + 2-pentanone) > P(CO2 + 3-methyl-2-butanone) ≈ P(CO2 +

acetone). In the 0.4 < xCO2 / mol/mol < 0.85 range, it can be seen from the present

measurements that the tendency of the vapor pressure of these four binary mixtures at

the same xCO2 is P(CO2 + 3-pentanone) > P(CO2 + 2-pentanone) > P(CO2 +

3-methyl-2-butanone) ≈ P(CO2 + acetone). It can be seen from both measurements

and correlations that the vapor pressure of the three CO2 + pentanone mixtures is very

similar in the critical region (xCO2 > 0.85 mol/mol). According to the correlation

results, the ∆P of the three CO2 + pentanone mixtures goes up again in the critical

region because the CO2 + acetone mixture reaches the critical point slightly earlier

11
with increasing mole fraction xCO2. Overall, the VLE phase boundaries of these four

binary mixtures are slightly different and the maximum difference between the vapor

pressure at the same mole fraction xCO2 is only about 0.4 MPa.

5. Conclusions

Experimental VLE data for the binary mixtures CO2 + acetone and CO2 +

pentanones were generated with a high pressure view cell technique based on the

synthetic method at 313.15 K, 333.15 K, and 353.15 K up to a pressure of 11.8 MPa.

First, the measurements of CO2 + acetone were used as a case study for the validation

of the employed experimental setup and operation procedure. A comparison of the

present measurement results with the experimental data in the literature shows that the

present data are in very good agreement with most data from the literature, but not all

experimental data sets are consistent with each other. Second, experimental VLE data

for three binary mixtures CO2 + pentanones (3-methyl-2-butanone, 2-pentanone, and

3-pentanone) were measured. Among these three binary mixtures, CO2 +

3-methyl-2-butanone and CO2 + 2-pentanone were investigated only at 313.15 K in

the past. The mixture CO2 + 3-pentanone was not measured before. The present

measurements thus broaden the database on VLE data of CO2 + pentanones. Finally,

all data were compared to the results from the Peng−Robinson equation of state (PR

EOS) with the van der Waals one-fluid mixing rule and from the combination of the

PR EOS with the UNIQUAC model through the Huron−Vidal mixing rule. Both

approaches may correlate these experimental data well. The present experimental

measurements and thermodynamic models can be used for the design of processes

using CO2 + acetone and CO2 + pentanones.

12
Acknowledgements

The authors would like to thank for the financial support from the Ministry of

Science and Technology of Taiwan (MOST 103-2218-E-008-003-MY2) and the

Alexander von Humboldt Stiftung. We wish to thank Elmar Baumhögger for his

support during the experimental investigations.

Appendix A. Supplementary Material

The Supplementary Material provides a briefly summary of error analysis in terms of

xCO2 and can be found online.

13
References

[1] G. Brunner, Supercritical fluids: technology and application to food processing,

Journal of Food Engineering 67 (2005) 21-33.
[2] A.I. Cooper, Polymer synthesis and processing using supercritical carbon dioxide,
Journal of Materials Chemistry 10 (2000) 207-234.
[3] P. Girotra, S.K. Singh, K. Nagpal, Supercritical fluid technology: a promising
approach in pharmaceutical research, Pharmaceutical Development and
Technology 18 (2013) 22-38.
[4] T. Merker, C.M. Hsieh, S.T. Lin, H. Hasse, J. Vrabec, Fluid-phase coexistence for
the oxidation of CO2 expanded cyclohexane: Experiment, molecular simulation,
and COSMO-SAC, AIChE Journal 59 (2013) 2236-2250.
[5] P.G. Jessop, B. Subramaniam, Gas-expanded liquids, Chemical Reviews 107 (2007)
2666-2694.
[6] K.A. Araus, R.I. Canales, J.M. del Valle, J.C. de la Fuente, Solubility of
beta-carotene in ethanol- and triolein-modified CO2, Journal of Chemical
Thermodynamics 43 (2011) 1991-2001.
[7] S.C. Chang, T.H. Hsu, Y.H. Chu, H.M. Lin, M.J. Lee, Micronization of aztreonam
with supercritical anti-solvent process, Journal of the Taiwan Institute of
Chemical Engineers 43 (2012) 790-797.
[8] S.E. Bozbag, C. Erkey, Supercritical fluids in fuel cell research and development,
Journal of Supercritical Fluids 62 (2012) 1-31.
[9] S.I. Sandler, Chemical and Engineering Thermodynamics, 3rd ed., John Wiley &
Sons, New York, 1999.
[10] B.E. Poling, J.M. Prausnitz, J.P. O'Connell, The Properties of Gases and Liquids
5th ed., McGraw-Hill, New York, 2001.
[11] C. Roy, D. Vrel, A. Vega-Gonzalez, P. Jestin, S. Laugier, P. Subra-Paternault,
Effect of CO2-antisolvent techniques on size distribution and crystal lattice of
theophylline, Journal of Supercritical Fluids 57 (2011) 267-277.
[12] Z.M. Liu, J.P. Wang, L.P. Song, G.Y. Yang, B.X. Han, Study on the phase
behavior of cholesterol-acetone-CO2 system and recrystallization of cholesterol
by antisolvent CO2, Journal of Supercritical Fluids 24 (2002) 1-6.
[13] S.S.T. Ting, S.J. Macnaughton, D.L. Tomasko, N.R. Foster, Solubility of
naproxen in supercritical carbon dioxide with and without cosolvents, Industrial
& Engineering Chemistry Research 32 (1993) 1471-1481.
[14] F. Temelli, A. Cordoba, E. Elizondo, M. Cano-Sarabia, J. Veciana, N. Ventosa,
Phase behavior of phytosterols and cholesterol in carbon dioxide-expanded
ethanol, Journal of Supercritical Fluids 63 (2012) 59-68.
[15] N. De Zordi, I. Kikic, M. Moneghini, D. Solinas, Solubility of pharmaceutical
compounds in supercritical carbon dioxide, Journal of Supercritical Fluids 66
(2012) 16-22.
[16] S. Camy, J.S. Condoret, Modelling and experimental study of separators for
co-solvent recovery in a supercritical extraction process, Journal of Supercritical
Fluids 38 (2006) 51-61.
[17] C.M. Hsieh, T. Windmann, J. Vrabec, Vapor-liquid equilibria of CO2+C1-C5
alcohols from the experiment and the COSMO-SAC model, Journal of Chemical
and Engineering Data 58 (2013) 3420-3429.
[18] C. Secuianu, J.W. Qian, R. Privat, J.N. Jaubert, Fluid Phase Equilibria
Correlation for Carbon Dioxide+1-Heptanol System with Cubic Equations of
State, Industrial & Engineering Chemistry Research 51 (2012) 11284-11293.
14
[19] C. Secuianu, V. Feroiu, D. Geana, Measurements and modeling of high-pressure
phase behavior of the carbon dioxide + pentan-1-ol binary system, Journal of
Chemical and Engineering Data 56 (2011) 5000-5007.
[20] C. Secuianu, V. Feroiu, D. Geana, Phase behavior for carbon dioxide plus ethanol
system: Experimental measurements and modeling with a cubic equation of state,
Journal of Supercritical Fluids 47 (2008) 109-116.
[21] J.O. Valderrama, J. Zavaleta, Generalized binary interaction parameters in the
Wong-Sandler mixing rules for mixtures containing n-alkanols and carbon
dioxide, Fluid Phase Equilibria 234 (2005) 136-143.
[22] I. Polishuk, J. Wisniak, H. Segura, Simultaneous prediction of the critical and
sub-critical phase behavior in mixtures using equation of state I. Carbon
dioxide-alkanols, Chemical Engineering Science 56 (2001) 6485-6510.
[23] C.S. Su, Prediction of solubilities of solid solutes in carbon dioxide-expanded
organic solvents using the predictive Soave-Redlich-Kwong (PSRK) equation of
state, Chemical Engineering Research & Design 91 (2013) 1163-1169.
[24] S.N. Reddy, G. Madras, A new semi-empirical model for correlating the
solubilities of solids in supercritical carbon dioxide with cosolvents, Fluid Phase
Equilibria 310 (2011) 207-212.
[25] A.I. Frolov, M.G. Kiselev, Prediction of Cosolvent Effect on Solvation Free
Energies and Solubilities of Organic Compounds in Supercritical Carbon
Dioxide Based on Fully Atomistic Molecular Simulations, Journal of Physical
Chemistry B 118 (2014) 11769-11780.
[26] E.K. Asep, S. Jinap, M.H.A. Jahurul, I.S.M. Zaidul, H. Singh, Effects of polar
cosolvents on cocoa butter extraction using supercritical carbon dioxide,
Innovative Food Science & Emerging Technologies 20 (2013) 152-160.
[27] J. Rincon, R. Camarillo, L. Rodriguez, V. Ancillo, Solubility of Used Frying Oil
in High Pressure CO2-Cosolvent Mixtures, Industrial & Engineering Chemistry
Research 50 (2011) 9314-9323.
[28] Dortmund Data Bank, http://www.ddbst.com/ddb.html.
[29] D.Y. Peng, D.B. Robinson, New two-constant equation of state, Industrial &
Engineering Chemistry Fundamentals 15 (1976) 59-64.
[30] D.S. Abrams, J.M. Prausnitz, Statistical thermodynamics of liquid-mixtures -
New expression for excess Gibbs energy of partly or completely miscible
systems, AIChE Journal 21 (1975) 116-128.
[31] M.-J. Huron, J. Vidal, New mixing rules in simple equations of state for
representing vapor-liquid-equilibria of strongly non-ideal mixtures, Fluid Phase
Equilibria 3 (1979) 255-271.
[32] T. Windmann, A. Koster, J. Vrabec, Vapor-liquid equilibrium measurements of
the binary mixtures nitrogen + acetone and oxygen + acetone, Journal of
Chemical and Engineering Data 57 (2012) 1672-1677.
[33] T. Windmann, Vapor-liquid equilibrium properties from molecular simulation and
experiment, in: Fakultät für Maschinenbau, Universität Paderborn, Paderborn,
2014.
[34] A.J. Staverman, The entropy of high polymer solutions - Generalization of
formulae, Recueil Des Travaux Chimiques Des Pays-Bas-Journal of the Royal
Netherlands Chemical Society 69 (1950) 163-174.
[35] E.A. Guggenheim, Mixtures, Oxford University Press, Oxford, 1952.
[36] J. Gmehling, J.D. Li, M. Schiller, A modified UNIFAC model .2. Present
parameter matrix and results for different thermodynamic properties, Industrial
& Engineering Chemistry Research 32 (1993) 178-193.
15
[37] C.Y. Day, C.J. Chang, C.Y. Chen, Phase equilibrium of ethanol plus CO2 and
acetone plus CO2 at elevated pressures, Journal of Chemical and Engineering
Data 41 (1996) 839-843.
[38] C.J. Chang, C.Y. Day, C.M. Ko, K.L. Chiu, Densities and P-x-y diagrams for
carbon dioxide dissolution in methanol, ethanol, and acetone mixtures, Fluid
Phase Equilibria 131 (1997) 243-258.
[39] C.M.J. Chang, K.L. Chiu, C.Y. Day, A new apparatus for the determination of
P-x-y diagrams and Henry's constants in high pressure alcohols with critical
carbon dioxide, Journal of Supercritical Fluids 12 (1998) 223-237.
[40] C.Y. Day, C.J. Chang, C.Y. Chen, Phase equilibrium of ethanol plus CO2 and
acetone plus CO2 at elevated pressures. (vol 41, pg 839, 1996), Journal of
Chemical and Engineering Data 44 (1999) 365-365.
[41] T. Katayama, K. Ohgaki, M. G., M. Goto, T. Nagano, Isothermal vapor-liquid
equilibria of acetone-carbon dioxide and methanol-carbon dioxide systems at
high pressures, Journal of Chemical Engineering of Japan 8 (1975) 89-92.
[42] T. Adrian, G. Maurer, Solubility of carbon dioxide in acetone and propionic acid
at temperatures between 298 K and 333 K, Journal of Chemical and Engineering
Data 42 (1997) 668-672.
[43] H.Y. Chiu, M.J. Lee, H.M. Lin, Vapor-liquid phase boundaries of binary mixtures
of carbon dioxide with ethanol and acetone, Journal of Chemical and
Engineering Data 53 (2008) 2393-2402.
[44] P. Traub, K. Stephan, High-pressure phase equilibria of the system
CO2-water-acetone measured with a new apparatus, Chemical Engineering
Science 45 (1990) 751-758.
[45] M. Stievano, N. Elvassore, High-pressure density and vapor-liquid equilibrium
for the binary systems carbon dioxide-ethanol, carbon dioxide-acetone and
carbon dioxide-dichloromethane, Journal of Supercritical Fluids 33 (2005) 7-14.
[46] T. Aida, T. Aizawa, M. Kanakubo, H. Nanjo, Dependence of volume expansion
on alkyl chain length and the existence of branched methyl group of
CO2-expanded ketone systems at 40 degrees C, Journal of Supercritical Fluids
55 (2010) 71-76.
[47] Y. Sato, N. Hosaka, K. Yamamoto, H. Inomata, Compact apparatus for rapid
measurement of high-pressure phase equilibria of carbon dioxide expanded
liquids, Fluid Phase Equilibria 296 (2010) 25-29.
[48] F. Han, Y. Xue, Y.L. Tian, X.F. Zhao, L. Chen, Vapor-liquid equilibria of the
carbon dioxide plus acetone system at pressures from (2.36 to 11.77) MPa and
temperatures from (333.15 to 393.15) K, Journal of Chemical and Engineering
Data 50 (2005) 36-39.
[49] W.Z. Wu, J. Ke, M. Poliakoff, Phase boundaries of CO2 + toluene, CO2 +
acetone, and CO2 + ethanol at high temperatures and high pressures, Journal of
Chemical and Engineering Data 51 (2006) 1398-1403.
[50] J.F. Mathews, Critical constants of inorganic substances, Chemical Reviews 72
(1972) 71-100.
[51] H. Orbey, S.I. Sandler, Modeling Vapor-Liquid Equilibria: Cubic Equations of
State and Their Mixing Rules., Cambridge University Press, New York, 1998.
[52] D. Ambrose, B.E. Broderick, R. Townsend, The critical temperatures and
pressures of thirty organic compounds, Journal of Applied Chemistry and
Biotechnology 24 (1974) 359-372.
[53] K.A. Kobe, H.R. Crawford, R.W. Stephenson, Industrial Design Data—Critical
Properties and Vapor Presesures of Some Ketones, Industrial & Engineering
16
 Chemistry 47 (1955) 1767-1772.
[54] DIPPR database, http://www.aiche.org/dippr.

17
 Table 1. Pure substance parameters of the Peng-Robinson EOS and the

UNIQUAC model

Substance Tc/K Pc/MPa ω q r

carbon dixoide 304.2 [50] 7.382 [51] 0.225 [51] 1.40 [30] 1.30 [30]

acetone 508.1 [52] 4.7 [52] 0.3065 a 2.34 [30] 2.57 [30]

2-pentanone 561.08 [52] 3.694 [52] 0.3433 a 4.147 b 3.6023 b

3-pentanone 560.9 [53] 3.74 [53] 0.3448 a 4.3839 b 3.6023 b

3-methyl-2-butanone 553.4 [53] 3.85 [53] 0.3208 a 4.147 b 3.6023 b

a, ω were estimated by using experimental vapor pressure data from the DIPPR

database [54] if they were not available in literature.

b, r and q were determined from the UNIFAC group contribution method [36].

Table 2. Binary parameters of the van der Waals one-fluid mixing rule (kij) and

the UNIQUAC model (∆uij, ∆uji) adjusted in the present work

Mixture kij (-) ∆uij (J/mol) ∆uji (J/mol)

CO2 + acetone 0.00515 -1740.17 2788.26

CO2 + 3-methyl-2-butanone 0.0196 -1826.81 3333.82

CO2 + 2-pentanone 0.0306 -1836.42 2812.16

CO2 + 3-pentanone 0.0357 -2271.16 4128.10

18
Table 3. Experimental vapor−liquid equilibrium data along the saturated liquid

line of CO2 + ketone mixtures generated in this worka

CO2 + acetone

T = 353.15 K T = 333.15 K T = 313.15 K

P xCO2 P xCO2 P xCO2

(MPa) (mol/mol) (MPa) (mol/mol) (MPa) (mol/mol)
10.53 0.794 (4) 8.80 0.823 (5) 6.67 0.843 (5)

9.73 0.729 (4) 7.98 0.756 (4) 6.04 0.778 (5)

8.60 0.648 (4) 7.04 0.675 (4) 5.36 0.699 (4)

7.16 0.552 (3) 5.90 0.582 (3) 4.55 0.612 (4)

6.20 0.485 (3) 5.13 0.516 (3) 4.06 0.551 (4)

CO2 + 3-methyl-2-butanone

T = 353.15 K T = 333.15 K T = 313.15 K

P xCO2 P xCO2 P xCO2

(MPa) (mol/mol) (MPa) (mol/mol) (MPa) (mol/mol)
11.42 0.840 (5) 9.51 0.866 (5) 7.12 0.883 (5)

10.55 0.776 (4) 8.55 0.797 (5) 6.42 0.814 (5)

9.48 0.709 (4) 7.70 0.731 (4) 5.83 0.750 (4)

9.30 0.704 (4) 7.56 0.726 (4) 5.74 0.740 (4)

8.51 0.649 (4) 6.95 0.669 (4) 5.32 0.691 (4)

7.53 0.586 (3) 6.12 0.602 (4) 4.87 0.640 (4)

6.12 0.489 (3) 5.16 0.517 (3) 4.08 0.545 (4)

6.35 0.500 (3) 5.29 0.526 (3) 4.17 0.554 (4)

5.18 0.426 (3) 4.42 0.465 (3) 3.31 0.465 (3)

3.95 0.343 (2) 3.11 0.343 (2) 2.35 0.343 (3)

3.96 0.347 (2) 3.12 0.347 (2) 2.35 0.347 (3)

19
 CO2 + 2-pentanone

T = 353.15 K T = 333.15 K T = 313.15 K

P xCO2 P xCO2 P xCO2

(MPa) (mol/mol) (MPa) (mol/mol) (MPa) (mol/mol)
11.87 0.888 (5) 9.97 0.908 (5) 7.52 0.923 (5)

11.36 0.822 (5) 9.22 0.840 (5) 7.14 0.886 (5)

10.72 0.778 (4) 8.68 0.797 (5) 6.89 0.856 (5)

9.91 0.725 (4) 8.68 0.799 (5) 6.47 0.813 (5)

8.71 0.645 (4) 8.01 0.746 (4) 6.49 0.811 (5)

7.75 0.588 (3) 7.03 0.668 (4) 6.05 0.762 (4)

6.34 0.493 (3) 6.37 0.608 (4) 5.38 0.688 (4)

5.32 0.424 (3) 5.37 0.526 (3) 5.49 0.692 (4)

3.75 0.305 (2) 4.55 0.462 (3) 4.98 0.639 (4)

2.96 0.305 (2) 4.30 0.558 (4)

4.23 0.551 (4)

3.42 0.462 (3)

2.24 0.305 (3)

CO2 + 3-pentanone

T = 353.15 K T = 333.15 K T = 313.15 K

P xCO2 P xCO2 P xCO2

(MPa) (mol/mol) (MPa) (mol/mol) (MPa) (mol/mol)
11.79 0.857 (5) 9.65 0.876 (5) 7.22 0.890 (5)

11.52 0.830 (5) 9.36 0.848 (5) 6.98 0.863 (5)

11.53 0.829 (5) 9.33 0.848 (5) 6.97 0.863 (5)

10.65 0.760 (4) 8.56 0.780 (4) 6.42 0.796 (5)

9.97 0.717 (4) 8.07 0.739 (4) 6.09 0.753 (4)

20
 9.98 0.718 (4) 8.08 0.738 (4) 6.09 0.754 (4)

9.11 0.661 (4) 7.42 0.683 (4) 5.64 0.702 (4)

8.10 0.596 (3) 6.66 0.619 (4) 5.14 0.643 (4)

6.49 0.494 (3) 5.45 0.522 (3) 4.32 0.553 (4)

5.51 0.420 (3) 4.93 0.482 (3) 3.64 0.465 (3)

5.43 0.415 (2) 4.76 0.455 (3) 3.54 0.456 (3)

4.13 0.342 (2) 4.68 0.456 (3) 2.51 0.342 (3)

3.29 0.342 (2)

a.
The uncertainties are u(T) = 0.04 K, u(P) = 0.01 MPa, and in case of the mole

fraction u(x) they are given by the number in parentheses, where 0.342 (2) stands for

0.342 ± 0.002 mol/mol.

21
Figure 1. Experimental setup for the present measurements of vapor-liquid equilibria.

V1 indicates a valve, TW-V1 a three-way valve, PtX a thermometer and PX a pressure

transducer.

Figure 2. Comparison of vapor-liquid equilibria of CO2 + acetone at 353.15 K (top),

333.15 K (center), and 353.15 K (bottom) from experiment, correlation by

PR+HV+UNIQUAC (─), and prediction by PR+VDW (---). Experimental data were

taken form the literature (□ [37-40], ○ [41], ◇ [42], + [43], x [44], ☆ [45], ▽

[46], [47], [48], [49]) and from this work ( ).

Figure 3. Comparison of vapor-liquid equilibria of CO2 + 3-methyl-2-butanone at

313.15 K, 333.15 K, and 353.15 K from experiment, correlation by

PR+HV+UNIQUAC (─), and prediction by PR+VDW (---). The experimental data

were taken form the literature (▽ [46]) and from this work ( ).

Figure 4. Comparison of vapor-liquid equilibria of CO2 + 2-pentanone at 313.15 K,

333.15 K, and 353.15 K from experiment, correlation by PR+HV+UNIQUAC (─),

and prediction by PR+VDW (---). The experimental data were taken form the

literature (▽ [46]) and from this work ( ).

Figure 5. Comparison of vapor-liquid equilibria of CO2 + 3-pentanone at 313.15 K,

333.15 K, and 353.15 K from experimental data measured in this work ( ),

correlation by PR+HV+UNIQUAC (─), and prediction by PR+VDW (---).

Figure 6. Comparison of vapor-liquid equilibria of CO2 + acetone (○, dotted line),

22
CO2 + 3-methyl-2-butanone (△, dashed line), CO2 + 3-pentanone (□, long-dashed

line), and CO2 + 2-pentanone (◇, solid line) at 333.15 K between experimental data

measured in this work (symbols) and correlation by PR+HV+UNIQUAC (lines). The

top figure shows the regular P-xCO2 phase diagram and the bottom figure shows a

deviation plot where the correlation results of the CO2 + acetone mixture were used as

the baseline, where ∆P ( x CO 2 ) = P( x CO 2 ) − PCO2

correlation
+ acetone ( x CO 2 ) at the same mole fraction

xCO2.

23
Figure 1.

24
Figure 2.

12 T = 353.15 K

10

8
P (MPa)

0
10 T = 333.15 K

8
P (MPa)

0
8 T = 313.15 K

6
P (MPa)

0
0.0 0.2 0.4 0.6 0.8 1.0
xCO2 (mol/mol)

25
Figure 3.

12 T = 353.15 K
10

8
P (MPa)

4 T = 313.15 K

0
0.0 0.2 0.4 0.6 0.8 1.0
xCO2 (mol/mol)

26
Figure 4.

12
T = 353.15 K
10

8
P (MPa)

4
T = 313.15 K
2

0
0.0 0.2 0.4 0.6 0.8 1.0
xCO2 (mol/mol)

27
Figure 5.

12
T = 353.15 K
10

8
P (MPa)

4
T = 313.15 K
2

0
0.0 0.2 0.4 0.6 0.8 1.0
xCO2 (mol/mol)

28
Figure 6.

10

8
P (MPa)

0.4
P (MPa)

0.2

0.0

-0.2
0.0 0.2 0.4 0.6 0.8
xCO2 (mol/mol)

29