Fluid Phase Equilibria 584 (2024) 114140

Contents lists available at ScienceDirect

Fluid Phase Equilibria

journal homepage: www.journals.elsevier.com/fluid-phase-equilibria

Density measurements of homogeneous phase fluid mixtures comprising

CO2/methanol and CO2/ethanol binary systems and correlation with
equations of state
Hiroaki Matsukawa a, Masamune Yomori b, Tomoya Tsuji c, Katsuto Otake a, *
a
Department of Industrial Chemistry, Faculty of Engineering, Tokyo University of Science, Niijuku 6-3-1, Katsushika-ku, Tokyo 125-8585, Japan
b
Department of Industrial Chemistry, Graduate School of Engineering, Tokyo University of Science, Niijuku 6-3-1, Katsushika-ku, Tokyo 125-8585, Japan
c
Universiti Technologi Malaysia, Malaysia Japan International Institute of Technology, Off Jalan Sultan Yahya Petra, Kuala Lumpur 54100, Malaysia

A R T I C L E I N F O A B S T R A C T

Keywords: Equations of state (EoS) are powerful tools for estimating a wide variety of physical properties. However, the
Density applicability of parameter sets derived from specific physical properties for the correlation and estimation of
Carbon dioxide various other physical properties has received limited attention. Additionally, the estimation of physical prop­
Methanol
erties using EoS is anticipated to be affected by the association of molecules. The densities of homogeneous phase
Ethanol
Sanchez-Lacombe equation of state
fluid mixtures comprising carbon dioxide (CO2)/methanol (MeOH) and CO2/ethanol (EtOH) binary systems were
PC-SAFT equation of state measured in the current study using a high-pressure vibration-type density meter equipped with a circulation
pump and a variable-volume viewing cell. Homogeneity was ensured by observing the fluid through the viewing
window of the variable volume cell. The measurements were carried out at a temperature range of 313–353 K,
the CO2 mole-fraction range of 0–80 mol%, and at pressures up to 20 MPa. Subsequently, the as-obtained
experimental data were correlated with two EoSs, viz. Sanchez-Lacombe (SL) EoS and Perturbed Chain statis­
tical associating fluid theory (PC-SAFT) EoS. The density correlations between SL and PC-SAFT EoS were almost
identical in accuracy. Additionally, the association between CO2 and alcohols in PC-SAFT EoS had no discernible
effect on the reliability of the density correlations. The vapor liquid equilibria (VLE) of the CO2/MeOH and CO2/
EtOH mixtures were further estimated using parameter sets determined from the density measurements. Both the
EoSs demonstrated comparable estimation accuracy; however, the pressure was estimated primarily near the
critical region of the mixture, which yielded a lower estimation accuracy. Additionally, the densities of the binary
systems were determined using characteristic parameters derived from the VLE correlations. Of the EoSs, the PC-
SAFT EoS yielded a good correlation of the VLE, including the region near the mixture’s critical region, while
taking the association between CO2 and alcohols into consideration. Although few of the correlations were
observed to be inferior, the density of the homogeneous fluid mixture was accurately estimated using the two
EoSs, with the parameters obtained from the VLE correlations. The findings of the study thus suggest that in order
to estimate the density and VLE using EoS-shared parameters, the parameter sets must first be determined using a
VLE that exhibits a wide range of conditions affected by the system’s associations.

1. Introduction Moreover, Kim et al. [2] investigated the use of methanol as an inhibitor
during the formation of CO2 clathrate hydrate. Additionally, MeOH has
Owing to the widespread use of carbon dioxide (CO2)/alcohol binary also been used in the Rectisol® process as a physical absorbent for CO2
systems in the industry, their phase equilibria have been measured and capture [3]. Owing to its biocompatibility, EtOH is frequently used as a
modeled. For instance, Rudyk et al. [1] reported on the extraction of co-solvent in the extraction of natural products such as caffeine [4], and
crude oil using supercritical (sc) CO2 and co-solvents viz. methanol α-pinene [5] using scCO2. Furthermore, EtOH is also used to prepare
(MeOH) and ethanol (EtOH) . When compared to pure CO2, the total microparticles of poly(vinylpyrrolidone) [6], and poly(vinyl alcohol)
recovery yield increased significantly by an average of about 20 %. [7] via supercritical anti-solvent approaches.

* Corresponding author.
E-mail address: [email protected] (K. Otake).

https://doi.org/10.1016/j.fluid.2024.114140
Received 11 January 2024; Received in revised form 15 May 2024; Accepted 17 May 2024
Available online 18 May 2024
0378-3812/© 2024 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Table 1
Reported densities for the CO2/MeOH and EtOH binary systems.
System Measurement object T/K P / MPa Authors Refs.

CO2/MeOH homogeneous phase fluid 313.2 to 323.2 10 to 20 Smith et al. [41]

323.20 to 398.15 2.5 to 39.50 Galicia-Luna et al. [42]
saturated liquid 298.15 to 473.15 0.99 to 14.75 Brunner et al. [43]
303.2 to 323.2 1 to 6 Kariznovi et al. [44]
291.15 to 313.14 0.5 to 7.72 Chang et al. [45]
298 to 313 1.27 to 7.67 Sih et al. [46]
CO2/EtOH homogeneous phase fluid 313.15 1.0 to 10.0 Kato et al. [47]
308 to 323 5 to 30 Cunico et al. [48]
313 to 363 8 to 25 Zuniga-Moreno et al. [49]
308.15 to 338.15 15 to 45 Zhu et al. [50]
saturated liquid 303.2 to 323.2 1 to 6 Kariznovi et al. [44]
291.15 to 313.14 1.16 to 7.92 Chang et al. [45]
313.2 to 328.2 0.62 to 8.91 Tsivintzelis et al. [51]
291.15 to 323.15 3.30 to 8.11 Stievano et al. [52]
291.15 to 313.14 0.87 to 7.92 Day et al. [53]

The phase equilibria of CO2/alcohol binary systems, particularly selection of an EoS for the accurate estimation of phase equilibria [16,
those of CO2/MeOH, have been reported repeatedly to verify the reli­ 17]. Focusing on CO2/alcohol binary systems, Joung et al. [8] correlated
ability of high-pressure phase equilibria measurement systems [8,15]. VLE of CO2/alcohol (MeOH, EtOH, 2-methoxyethanol and 2-ethoxye­
However, phase equilibrium is not the only physical property required thanol) systems using the Peng-Robinson (PR) EoS [18], as well as the
for the design of chemical industrial processes. In this context, numerous multi-fluid nonrandom lattice fluid hydrogen-bonding (MF-NLF-HB)
other properties, such as density, viscosity and interfacial tension, must EoS [19,20]. Except for the critical regions, the MF-NLF-HB EoS was
be considered. However, the number of reports on these properties is observed to accurately correlate with the liquid phase composition.
significantly less than the number of reports on phase equilibria. Addi­ Thus, they concluded that the PR EoS provided a more accurate fit for
tionally, the measurement of physical properties under high pressure, the data. Additionally, Khalifa et al. [21] correlated the phase behavior
including phase equilibria, requires specialized equipment and long of CO2/MeOH and CO2/EtOH binary systems using different types of
acquisition periods. This adds to the already low number of reports for EoS, viz. Perturbed Chain statistical associating fluid theory (PC-SAFT)
physical properties other than phase equilibrium. Thus, the estimation EoS [22,23], Perturbed Chain Polar (PCP)-SAFT EoS [24-26],
of physical properties through calculations is a critical complement to Cubic-Plus-Association (CPA) EoS [27], and General cubic (G) EoS
experimental measurements. [28-32]. As indicated by the findings, no significant difference in EoS in
In high-pressure systems, such as mixtures containing scCO2, the the low-pressure regions was discernible, although, similar to Joung
equation of state (EoS) is a powerful and extremely useful tool for et al., there was a difference in EoS in the critical region, with PC-SAFT
estimating physical properties. EoS is often used to calculate vapor EoS exhibiting slightly superior performance.
liquid equilibrium (VLE), as well as the solubility of a system, because it On the other hand, few reports exist that verify the validity of
not only describes the basic pressure-volume-temperature relationships parameter sets obtained from specific physical properties for the cor­
but also enables the calculation of thermodynamic parameters such as relation and estimation of various other physical properties. Although
Gibbs free energy and enthalpy. Furthermore, EoS is also capable of only PR EoS was used, Cardoso et al. [33] discussed the vapor-liquid
calculating a variety of physical properties, including interfacial tension. equilibrium and volume expansion of organic solvents by carbon diox­
Because chemical process design requires a wide variety of physical ide using several mixing rules. Consequently, the findings demonstrated
properties, it is inefficient to use only specialized equations to estimate that none of the mixing rules was able to satisfactorily predict the
specific properties. Thus, the ability of EoS to estimate a wide range of mixture density or volume expansion. Additionally, Moine et al. [16]
physical properties is extremely appealing. However, to utilize EoS for compared the Soave-Redlich-Kwong (SRK) EoS [34] and the PC-SAFT
obtaining these physical properties, the ability to correlate and estimate EoS for estimating the physical properties of pure components. The
as many properties as possible using appropriate parameter sets deter­ findings indicated that the SRK EoS failed in accurately reproducing the
mined from a single set of experimental data is highly desired. liquid density, while the PC-SAFT EoS overestimated the critical pres­
Although numerous EoSs exist, they can be broadly classified into sure due to the difference in the parameterization approach for the pure
three primary categories, viz. those based on van der Waals model, those component parameters. Furthermore, the findings also suggested that by
based on the lattice fluid model, and those based on the perturbation unifying the parameterization approach for pure component parameters
theory. Numerous discussions have taken place regarding the proper involving the combined analysis of saturated vapor pressure and liquid

Table 2 Table 3
Specifications of the pure components. Pure component parameters for the two EoSs.
Component CAS Mw / g Supplier Mass Moisture Sanchez-Lacombe equation of state
number mol− 1 fraction content
3
purity mole P* / MPa T* / K ρ* / kg m − Refs.
(Supplier) fraction CO2 574.5 305 1510 [37]
Carbon 124–38–9 44.01 Fujii 0.9999 MeOH 1202 468 922 [37]
dioxide, Bussan EtOH 1069 413 963 [37]
CO2 Co.
Methanol, 67–56–1 32.04 Kanto 0.995 0.00013 PC-SAFT equation of state
MeOH Kagaku 1
m σ/Å εk− /K κAi Bi / - εAi Bi k− 1
/K Refs.
Co.
Ethanol, 64–17–5 46.07 Kanto 0.995 0.00019 CO2 2.0729 2.7852 169.21 [22]
EtOH Kagaku MeOH 1.5255 3.2300 188.90 0.0352 2899.5 [23]
Co. EtOH 2.3830 3.1770 198.24 0.0323 2653.4 [23]

2
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Fig. 1. Comparison of the density of pure MeOH measured in this work against previously reported results: (a) density-pressure correlations of the experimental
results, (b) variations in the deviation of density with pressure.

density, the estimation of various physical properties between the two evident from the spectra for the combination bands for CO2, CO2 and
EoSs is rendered equivalent in terms of accuracy. Mallepally et al. [17] MeOH were observed to interact with each other. Thus, it was estab­
further compared the SRK EoS, PC-SAFT EoS, and a modified lished that the interactions between CO2 and MeOH were Lewis
Sanchez-Lacombe (MSL) EoS [35] for estimating the phase behavior and acid-base interactions rather than hydrogen bonds. Furthermore, the
densities of propylene/toluene and ethylene/toluene systems. The MSL association of molecules was observed to affect physical properties.
EoS is a modification of Neau’s version of the SL EoS [36,38], which Therefore, several EoSs with modifications for these associations were
includes a Péneloux-type volume translation as demonstrated by the also reported [19,20,27].
findings, both SRK and PC-SAFT EoSs accurately estimated the phase Table 1 summarizes the reported densities for the CO2/MeOH and
behavior and densities of a variety of systems, with only minor over­ EtOH binary systems. The density of the homogeneous phase fluid is
prediction of densities. In a previous study [39], we evaluated the cor­ distinct from the density of the saturated liquid (the density at the VLE
relation and estimation accuracy of homogeneous mixed-fluid density line in the phase diagram) and represents the density in the homoge­
and phase equilibrium for the CO2/Toluene binary system using three neous phase region. As the table shows, while there are numerous re­
EoS, PR, SL, and PC-SAFT. The results indicate that the accuracies of the ports on the saturated liquid density, reports on homogeneous phase
calculations are significantly affected by the pure component parame­ fluid density are highly limited due to the measurement conditions.
ters of the individual EoS. Smith et al. [41] reported the densities of the CO2/MeOH mixtures at
It is well established that alcohol molecules associate with each other temperatures from 313.2 to 323.2 K and at pressures ranging from 10 to
via hydrogen bonds. Additionally, it is well known that CO2 and alcohols 20 MPa. Additionally, Kato et al. [47] and Cunico et al. [48] measured
associate under high pressures. Using FT-IR spectroscopy, Reilly et al. the densities of the CO2/EtOH mixtures at temperatures of 313.15 K, as
[40] investigated the solvent properties of liquid CO2 and MeOH. As well as 308 and 323 K, respectively, and pressures ranging from 10 MPa

3
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Fig. 2. Comparison of the density of pure EtOH measured in this work against previously reported results: (a) density-pressure correlations of the experimental
results, (b) variations in the deviation of density with pressure.

to 30 MPa. Furthermore, Galicia-Luna et al. [42,49] measured the several rigid hard spheres in the EoS. Additionally, for systems in which
densities of both the CO2/MeOH and CO2/EtOH mixtures using a the association of molecules occurs, the contribution by association is
vibrating tube densimeter. added [23,54–56] to the main hard chain reference state and dispersion
In previous studies on CO2/Toluene binary systems [39], the asso­ terms. On the other hand, the SL EoS is derived from the Flory–Huggins
ciation terms were not considered because no interactions between the lattice-fluid model and is primarily used to study systems containing
components occurred. However, the current study focuses on the asso­ macromolecules, owing to the absence of critical constants such as pure
ciation of molecules. The densities of homogeneous phase fluid mixtures parameters. Using the parameter sets obtained from the density data, the
comprising CO2/MeOH and CO2/EtOH binary systems were measured VLE of the CO2/MeOH and CO2/EtOH systems were subsequently esti­
using a high-pressure vibration-type density meter at temperatures mated. Furthermore, the validity of the parameter sets obtained by
ranging from 313.2 to 353.2 K and at pressures up to 20 MPa. The CO2 correlating the density to the estimation of the VLE and vice versa was
composition was varied between 0 and 80 mol% at 20 mol% intervals. discussed, and the difference in the estimation accuracy for the densities
The experimental data were correlated with the SL EoS and the PC-SAFT and phase behavior was determined using the basic calculation method
EoS [PR EoS was found to be unsuitable for density calculations in of each EoS.
previous studies [39]. Previous studies demonstrated that density cal­
culations using the van der Waals-type EoS, including the PR EoS, are 2. Experimental
complicated by differences in the pure-component parameters and the
manner by which the repulsion term is expressed. Thus, results of PR EoS 2.1. Materials
were summarized in Supporting information 1]. The PC-SAFT EoS is
based on perturbation theory, comprising a methodology for obtaining Carbon dioxide (CO2, CAS number [124–38–9], mass fraction purity
residual Helmholtz free energy by adding perturbation terms to the main >99.99 %) was purchased from Fujii Bussan Co. Methanol (CH3OH, CAS
hard chain reference state. Thus, each molecule is regarded as a chain of number [67–56–1], mass fraction purity 99.5 %) and Ethanol (C2H5OH,

4
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

pump (Nihon Seimitsu Kagaku Co., NP-AP-20(J)). All components

except for the circulation pump were installed in a constant temperature
air bath (Yamato Scientific Co., Ltd., DKN602). To maintain a constant
temperature, the piston head of the circulation pump was equipped with
a heater and a temperature controller. The high-pressure vibration type
density meter measured the oscillation period of the U-tube, which
contained the sample. The following equation [57] was used to deter­
mine the densities from the measured period of oscillation.

ρ = Aτ2 − B
{
A = a + bP + cP2 (1)
with
B = d + eP + fP2

where, ρ, P, and τ are the density, pressure, and oscillation periods,

respectively, and a, b, c, d, e, and f are the apparatus constants [Appa­
ratus constants are summarized in Supporting Information 2]. To
determine these constants, two known density samples with densities
greater and less than the unknown sample must be measured at the
experimental pressure and temperature. In this work, MeOH and eth­
ylbenzene, and EtOH and ethylbenzene were used as the reference
samples of known density for the CO2/MeOH and CO2/EtOH binary
systems, respectively. The density of MeOH and EtOH were obtained
using the high-accuracy Helmholtz energy NIST equations of state in the
reference fluid properties (REFPROP) software (ver. 10), while that of
ethylbenzene was obtained from the literature [58,59]. Additionally, the
temperature dependence of the apparatus constants was not considered
due to the lack of measurement data. As a result, the apparatus constants
were determined for each experimental temperature condition. The
details of the experimental apparatus and procedures employed have
been described elsewhere [60]. The homogeneity of the phase was
confirmed from the VLE phase diagram, as well as visual observation
through the viewing window of the variable-volume cell.
The temperature and pressure standard uncertainties for all mea­
surements were 0.153 K and 0.033 MPa, respectively. For each data
point, the combined standard uncertainties in the composition and
experimental density values were calculated. Details of the calculation
methods have been described previously [60]. The standard uncertainty
of the density was determined for the periodic measurement, and the
combined expanded uncertainty of the density, Uc(ρ), with confidence
level of 0.95 (k = 2) was obtained by considering the standard un­
certainties of the pressure and temperature. The method of calculating
the density uncertainty is summarized in Supporting Information 3. The
Fig. 3. Density behavior of (a) the CO2(1)/MeOH(2) and (b) the CO2(1)/EtOH
Uc(ρ) at each experimental point is shown in Tables 4 and 5.
(2) binary systems. The data points in the figure were obtained by the inter­
polation of the density data from Tables 4 and 5.
2.3. Experimental data correlation and estimation of density
CAS number [64–17–5], mass fraction purity 99.5 %) were obtained
2.3.1. SL EoS
from Kanto Kagaku Co. Alcohols were stored using a molecular sieve to
Because it does not require a critical value as a pure parameter, SL
prevent them from absorbing water from the air. The water content of
EoS [36; 37] is primarily used to study systems containing large mole­
alcohols was measured using a Karl Fischer titrator (HIRANUMA Co.,
cules, such as macromolecules. In this study, the experimental data were
AQV-7) immediately prior to performing density measurements. The
correlated to SL EoS and is shown below:
alcohol had a moisture content of less than 0.019 mol%, which was
assumed to be insignificant in terms of density uncertainty. During the ρ2 + P
̃ ̃ + T{ln(1
̃ ρ) + (1 − 1/r)̃ρ} = 0
− ̃
density measurement, the system was sealed to ensure no contact with ̃ = T/T∗ P
T ̃ = P/P∗ , ̃
ρ = ρ/ρ∗ = V ∗ /V (2)
moisture in the air. The specifications of pure CO2, MeOH, and EtOH are r = P V /RT
∗ ∗ ∗

listed in Table 2. It should be noted that in the table, the unit of the
moisture content given in mass fractions by the measuring apparatus where, T, P, and ρ are the temperature, pressure, and density, respec­
was converted to mole fractions. tively. T,
̃ P,
̃ and ̃ ρ represent the reduced parameters; T ∗ , P∗ , and ρ∗ are
characteristic parameters of the pure component; R is the gas constant;
2.2. Apparatus and procedure and r is the number of segments. To calculate the mixture using the EoS,
it is necessary to apply the mixing rule. The mixing rule for SL EoS used
The density of the homogeneous phase of the CO2/MeOH and CO2/ in this study is described by the following set of equations [61].
EtOH binary mixtures was determined in this study with the help of an 1 ∑ϕ
i
experimental apparatus comprising a variable-volume viewing cell = (3)
V∗ i
Vi∗
(Tamaseiki Ind. Co., H-050,151–1), high-pressure vibration-type density
meter (Anton Paar Co., DMA 512), and double plunger type circulation

5
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Fig. 4. Comparison of the density of the homogeneous mixtures measured in this work against previously reported results. (a) CO2(1)/MeOH(2) system at 15 MPa,
(b) CO2(1)/EtOH(2) system at 10 MPa, (c) CO2(1)/EtOH(2) system at 15 MPa. The data points in the figure were obtained by the interpolation of the density data
from Tables 4 and 5.

6
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Table 4
Densities of the CO2(1)/MeOH(2) system (pressure P, density ρ, CO2 mole fraction x1, and the combined expanded uncertainties in experimental density Uc(ρ) and
composition Uc(x))a,b.
313.2 K 333.2 K 353.2 K
3 3 3 3 3 3
P / MPa ρ / kg m− Uc(ρ) / kg m− P / MPa ρ / kg m− Uc(ρ) / kg m− P / MPa ρ / kg m− Uc(ρ) / kg m−

x1 = 0
19.99 792.0 1.1 20.13 774.7 1.6 20.09 756.3 1.9
19.01 791.0 1.0 18.92 773.3 1.4 19.11 755.4 1.5
18.18 790.2 0.9 17.98 772.7 1.3 17.92 754.1 1.1
16.91 789.0 0.7 16.93 771.3 1.2 17.03 753.2 0.8
15.98 788.2 0.6 15.99 770.6 1.1 16.13 752.3 0.6
14.99 787.2 0.5 15.12 769.5 1.0 15.07 751.2 0.4
14.04 786.4 0.5 14.17 768.4 0.9 13.99 749.9 0.4
13.00 785.4 0.4 13.14 767.6 0.8 13.01 748.8 0.4
11.92 784.3 0.3 11.94 766.4 0.7 12.07 747.7 0.4
11.10 783.4 0.3 10.96 765.0 0.7 10.96 746.4 0.5
9.99 782.4 0.3 10.08 764.2 0.7 10.10 745.4 0.5
9.15 781.6 0.3 9.08 763.2 0.6 9.11 744.2 0.5
8.15 780.4 0.4 7.98 762.1 0.6 8.12 742.9 0.5
6.85 779.2 0.4 6.85 760.7 0.6 7.10 741.6 0.5
5.95 778.3 0.4 6.11 759.9 0.6 6.00 740.2 0.6
4.96 777.3 0.5 5.00 758.7 0.6
3.92 776.3 0.5 3.96 757.3 0.6
x1 = 0.200 (Uc(x) = 0.003)
20.00 829.3 0.7 19.94 797.9 1.1 20.06 771.6 1.3
18.97 828.1 0.6 18.88 796.5 1.0 18.98 770.2 1.0
17.95 826.8 0.6 18.04 795.6 0.9 17.94 768.9 0.8
16.97 825.6 0.5 16.92 794.0 0.9 17.01 767.7 0.7
16.04 824.4 0.5 16.11 793.1 0.8 15.90 766.2 0.5
15.02 823.1 0.5 15.06 791.8 0.7 14.96 765.0 0.5
14.06 821.8 0.4 14.06 790.4 0.7 14.01 763.8 0.5
13.09 820.6 0.4 13.02 789.1 0.7 12.97 762.3 0.5
12.07 819.3 0.4 12.06 787.8 0.6 11.90 760.6 0.5
11.07 818.1 0.4 11.13 786.5 0.6 10.99 759.3 0.5
9.99 816.5 0.4 9.93 784.8 0.6 10.10 758.0 0.6
9.07 815.3 0.4 8.96 783.4 0.6 8.85 756.0 0.6
7.87 813.8 0.5 8.02 782.1 0.6 8.07 754.8 0.6
6.99 812.5 0.5 7.06 780.6 0.6 6.86 752.8 0.6
6.02 811.2 0.5 5.94 778.9 0.6 5.91 751.3 0.6
4.95 809.7 0.5 5.02 777.5 0.6 5.08 749.8 0.6
3.94 808.1 0.5 4.03 776.1 0.6 3.99 747.9 0.7
3.07 806.2 0.5
x1 = 0.400 (Uc(x) = 0.003)
20.01 859.3 0.6 19.90 821.1 0.8 20.06 789.7 1.0
18.96 857.4 0.6 19.07 819.8 0.8 19.06 787.6 0.9
18.06 855.9 0.6 18.05 817.8 0.8 18.02 785.3 0.8
16.91 853.8 0.5 17.03 815.7 0.8 17.01 783.0 0.7
15.94 851.9 0.5 16.00 813.7 0.7 16.01 780.6 0.7
15.06 850.2 0.6 14.90 811.6 0.7 14.92 778.0 0.7
13.99 848.1 0.6 13.91 809.5 0.7 14.10 775.9 0.7
12.99 846.1 0.6 12.94 807.3 0.7 12.93 772.8 0.7
12.03 844.2 0.6 12.06 805.5 0.7 11.94 770.2 0.8
11.14 842.3 0.6 11.06 803.3 0.7 10.92 767.3 0.8
9.99 839.8 0.6 10.08 801.2 0.7 10.09 764.9 0.8
9.07 838.0 0.6 9.08 798.7 0.7 8.95 761.2 0.8
7.92 835.2 0.7 8.05 796.2 0.7
7.02 833.1 0.7 7.02 793.6 0.8
6.10 831.0 0.7
x1 = 0.600 (Uc(x) = 0.003)
19.97 872.6 0.8 20.06 829.2 1.0 19.97 790.0 1.2
18.96 870.1 0.8 18.96 826.0 1.0 18.94 787.2 1.1
18.09 867.7 0.8 17.95 823.0 1.0 17.94 784.3 1.1
17.02 864.8 0.8 16.99 820.1 1.0 17.01 781.3 1.1
16.07 862.1 0.8 16.09 817.3 1.0 15.94 777.8 1.1
15.05 859.1 0.8 15.05 814.0 1.0 15.02 774.7 1.2
14.06 856.1 0.8 14.05 810.6 1.0 14.07 771.5 1.2
13.07 853.0 0.9 13.03 806.9 1.1 13.07 767.7 1.3
12.05 849.5 0.9 12.08 803.4 1.1 12.05 763.8 1.3
11.01 846.1 0.9 10.96 799.1 1.1 11.09 759.9 1.4
10.03 842.6 0.9 10.02 795.2 1.2
8.99 838.6 1.0
7.95 834.5 1.0
x1 = 0.801 (Uc(x) = 0.002)
19.89 865.9 1.2 20.06 788.8 1.4 20.21 702.6 1.8
19.07 861.3 1.2 18.95 779.4 1.5 19.13 688.5 1.9
18.18 856.1 1.2 18.01 770.6 1.6 17.97 670.5 1.9
17.12 849.6 1.2 16.97 759.9 1.6 16.90 651.5 2.1
(continued on next page)

7
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Table 4 (continued )
313.2 K 333.2 K 353.2 K
3 3 3 3 3 3
P / MPa ρ / kg m− Uc(ρ) / kg m− P / MPa ρ / kg m− Uc(ρ) / kg m− P / MPa ρ / kg m− Uc(ρ) / kg m−

16.13 843.1 1.3 16.00 748.7 1.7 15.95 630.7 2.2

15.12 836.0 1.3 15.00 735.5 1.8 14.89 621.5 2.5
14.17 828.8 1.4 14.00 719.8 1.9 13.97 599.1 2.8
13.07 819.7 1.4 12.99 700.3 2.1
11.95 809.2 1.5 11.89 670.6 2.3
11.03 799.5 1.5
10.15 788.8 1.6
9.11 773.4 1.7
7.98 749.4 1.9
a
Standard uncertainties u are u(T) = 0.15 K, u(P) = 0.033 MPa.
b
Under all conditions, the mixture was in a compressed uniform liquid phase state.

/
wi ρ∗ Eq. (10):
ϕi = ∑ /i (4) ( )
wj ρ∗j a
∂̃res
j Z=1+η (10)
∂η T,xi
∑ ∑∑
P∗ = ϕi P∗i − RT ϕi ϕj χ ij (5)
i j i<j
where, η is the packing fraction. Details of the calculation methods for
hc disp
a and ̃
̃ a have been described previously [39]. In this study, differ
( )( )1/2 from the previous work, association will occur between alcohols as well
P∗i + P∗j − 2 1 − δij P∗i P∗j
χ ij = (6) as between CO2 and alcohols. The contribution by association to the
RT Helmholtz free energy is shown as follows [54,56]:
[ ]
P ∗ v0 ∑ ∑( XAi
)
1
T∗ = (7) a
assoc
Xi lnXAi − + Mi (11)
R ̃ =
2 2
i Ai
( ∗ )
1 ∑ Pi
= ϕi (8) where, XAi is the mole fraction of ith component not bonded at site A,
v0 i
RT ∗
i
and Mi is the number of association sites on ith component. XAi is given
by

⎡ ⎤− 1
∑∑
X = ⎣1 + NAv
Ai
ρj X Δ
Bj Ai Bj ⎦

j Bj
⎛ ⎞ (12)
∑ ∑
⎝ over all sites on jth component, Aj , Bj , Cj , ⋯; over all components⎠
Bj j

where, v0 is the volume of a lattice site. The characteristic temperature

T ∗ of the mixture is given by where ϕi and wi are the close-packed vol­ where, NAv and ρj are the Avogadro’s number and molar density of jth
ume fraction and mass fraction of the ith component, respectively. χ ij is component, respectively, and ΔAi Bj is the strength of interaction between
the i-j interaction term, and δij is the i-j interaction parameter. The site A on ith component and site B on jth component. ΔAi Bj is shown
interaction parameters were used as the fitting parameters during cor­ below:
relation. Methods underlying the calculations involved have been ( )seg [ ( / ) ]
detailed in a previous study [39]. The pure component parameters used ΔAi Bj = dij gij dij κAi Bj exp εAi Bj kT − 1 (13)
in this study are summarized in Table 3 [37]. ( )/
dij = di + dj 2 (14)
2.3.2. PC-SAFT EoS
PC-SAFT EoS [22,23], which is representative of the SAFT-type EoS, where, k is the Boltzmann constant. di is the temperature-dependent
is derived from the perturbation theory. Perturbation theory is a meth­ segment diameter of ith component and is given by
odology for obtaining solutions by adding perturbation terms to the [ ( εi )]
main solution. In this study, the experimental data were correlated with di = σ i 1 − 0.12exp − 3 (15)
kT
the PC-SAFT EoS and is shown below:
where σi and εi are the segment diameter and depth of the pair potential
Ares res
a =̃
=̃
hc
a +̃
disp
a +̃ a
assoc
(9) of ith component, respectively. gij(dij)seg is the temperature-dependent
NkT
segment diameter of ith component and the segment radial distribu­
hc disp assoc tion function. gij(dij)seg can be approximated to the radial distribution
where, Ares is the residual Helmholtz free energy, and ̃
a ,̃a , and ̃
a
function of the hard-sphere fluid gij(dij)hs as follows:
are contributions to the Helmholtz free energy by a chain of hard
spheres, by dispersion, and association, respectively. The compression
factor Z is expressed in terms of the Helmholtz free energy as shown in

8
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Table 5
Densities of the CO2(1)/EtOH(2) system (pressure P, density ρ, CO2 mole fraction x1, and the combined expanded uncertainties in experimental density Uc(ρ) and
composition Uc(x))a, b.
313.2 K 333.2 K 353.2 K
3 3 3 3 3 3
P / MPa ρ / kg m− Uc(ρ) / kg m− P / MPa ρ / kg m− Uc(ρ) / kg m− P / MPa ρ / kg m− Uc(ρ) / kg m−

x1 = 0
19.96 790.0 0.3 19.92 774.0 1.2 20.15 757.2 0.7
19.00 789.0 0.3 18.94 773.1 1.0 19.09 756.1 0.5
17.87 788.1 0.3 17.96 772.2 0.9 18.16 755.3 0.4
16.75 787.2 0.3 16.92 771.2 0.8 17.15 754.4 0.3
15.72 786.3 0.3 15.83 770.1 0.7 16.12 753.3 0.3
14.90 785.9 0.3 14.97 769.3 0.6 15.11 752.4 0.4
14.02 785.1 0.3 14.04 768.5 0.6 14.12 751.2 0.4
12.94 784.2 0.4 13.01 767.7 0.6 13.14 750.1 0.4
11.98 783.2 0.4 12.15 766.5 0.6 12.05 749.1 0.5
10.93 782.3 0.4 11.16 765.7 0.6 11.06 748.0 0.5
9.97 781.2 0.4 10.02 764.8 0.6 10.17 747.0 0.4
8.99 780.4 0.4 9.11 763.7 0.6 9.04 745.7 0.4
8.02 779.8 0.4 7.96 762.6 0.7 8.01 744.5 0.4
6.93 778.7 0.4 7.06 761.7 0.8 7.02 743.4 0.4
5.92 777.8 0.4 6.06 760.6 0.9 6.01 742.2 0.5
4.95 776.8 0.4 4.95 759.6 1.0 4.97 740.8 0.6
3.97 775.9 0.4 4.03 758.6 1.1 4.00 739.6 0.7
x1 = 0.202 (Uc(x) = 0.004)
19.94 819.5 0.4 19.90 796.5 0.8 19.988 774.2 0.5
18.99 818.3 0.4 19.06 795.5 0.8 18.94 772.7 0.5
17.94 817.0 0.4 17.97 794.1 0.7 17.84 771.2 0.4
16.87 815.7 0.4 16.94 792.8 0.6 17.01 770.0 0.4
15.97 814.6 0.4 15.95 791.4 0.6 15.976 768.5 0.4
14.96 813.3 0.4 14.90 790.1 0.5 14.93 766.8 0.4
13.98 812.1 0.4 13.99 788.8 0.5 14.05 765.4 0.4
13.03 810.9 0.4 12.98 787.5 0.5 12.954 763.5 0.5
12.05 809.6 0.4 11.96 786.0 0.5 11.98 762.0 0.5
11.06 808.3 0.4 11.01 784.6 0.5 11.08 760.4 0.5
10.04 806.9 0.4 10.08 783.3 0.5 10.08 758.9 0.5
9.02 805.7 0.4 9.04 781.8 0.5 9.06 757.2 0.5
8.07 804.4 0.4 8.09 780.3 0.6 8.05 755.4 0.4
7.04 803.0 0.4 7.04 778.8 0.6 7.06 753.5 0.4
6.12 801.8 0.4 6.03 777.2 0.7 6.04 751.6 0.5
4.91 800.1 0.4 5.04 775.6 0.8
4.15 799.1 0.4
x1 = 0.401 (Uc(x) = 0.003)
20.01 845.7 0.6 20.15 812.8 0.7 20.05 780.8 0.6
19.06 844.0 0.6 19.09 810.7 0.6 19.07 778.6 0.5
18.09 842.2 0.6 18.12 808.6 0.6 18.19 776.4 0.5
16.99 840.1 0.6 17.07 806.3 0.6 17.24 773.9 0.5
16.01 838.2 0.6 16.01 804.0 0.6 15.99 770.9 0.5
15.02 836.2 0.6 15.08 801.9 0.6 15.15 768.3 0.5
13.97 834.1 0.6 14.11 799.7 0.6 14.08 765.1 0.6
13.01 832.1 0.6 13.15 797.3 0.6 13.09 762.0 0.6
12.03 830.1 0.6 12.10 794.7 0.6 12.15 759.0 0.6
11.15 828.2 0.6 11.12 792.2 0.6 11.08 755.4 0.6
10.20 826.1 0.6 10.18 789.7 0.6
9.03 823.5 0.6 9.09 786.6 0.6
7.98 821.0 0.6
6.92 818.5 0.6
x1 = 0.603 (Uc(x) = 0.003)
19.97 856.1 0.8 20.09 814.8 0.8 20.11 774.8 0.8
18.99 853.5 0.8 19.08 811.8 0.8 19.12 771.2 0.8
18.02 851.0 0.8 18.04 808.7 0.8 18.19 767.7 0.8
16.95 848.0 0.8 17.11 805.8 0.8 17.20 763.8 0.8
16.01 845.4 0.8 16.16 802.8 0.8 16.04 758.9 0.8
15.00 842.4 0.8 15.12 799.2 0.8 15.12 755.6 0.9
14.02 839.5 0.8 14.16 795.9 0.8 14.12 749.8 0.9
13.11 836.7 0.7 13.11 792.0 0.8 13.10 744.4 0.9
12.08 833.4 0.7 12.03 787.8 0.8 12.07 738.3 0.9
11.04 829.9 0.7 11.08 783.8 0.8
10.04 826.5 0.8 9.96 778.6 0.9
9.03 822.8 0.8
7.98 818.7 0.8
x1 = 0.800 (Uc(x) = 0.002)
19.99 863.5 1.1 20.07 793.8 1.2 20.12 717.5 1.4
19.00 858.4 1.1 18.98 785.9 1.2 18.99 705.0 1.4
18.01 853.2 1.1 18.01 778.5 1.3 17.98 692.2 1.5
16.95 847.3 1.1 17.05 770.4 1.3 16.93 677.2 1.6
16.00 841.8 1.1 16.07 761.0 1.3 15.99 661.2 1.7
15.03 835.8 1.1 15.04 750.0 1.4 14.94 639.6 1.9
(continued on next page)

9
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Table 5 (continued )
313.2 K 333.2 K 353.2 K
3 3 3 3 3 3
P / MPa ρ / kg m− Uc(ρ) / kg m− P / MPa ρ / kg m− Uc(ρ) / kg m− P / MPa ρ / kg m− Uc(ρ) / kg m−

14.01 829.0 1.1 14.01 737.1 1.5 13.94 611.7 2.0

13.05 821.9 1.1 12.94 721.3 1.5 13.50 596.1 2.1
12.06 814.2 1.2 11.91 701.3 1.6
11.04 805.2 1.2 10.90 684.4 1.8
9.98 794.5 1.3
9.09 783.9 1.3
8.09 768.8 1.4
a
Standard uncertainties u are u(T) = 0.15 K, u(P) = 0.033 MPa.
b
Under all conditions, the mixture was in a compressed uniform liquid phase state.

pure component parameters independent on the contribution by asso­

ciation. When the contribution by association is introduced, the asso­
ciation parameters κAi Bi and εAi Bi are added as pure component
parameters. In the current study, it was assumed that alcohols have two
association sites, one for the oxygen atom (O) of the hydroxy group and
the other for the hydrogen atom (H), and that association occurs be­
tween the O and H atoms. This is classified as type 2B [56]. All pure
component parameters of the PC-SAFT EoS including the association
parameters were obtained by fitting the saturated liquid density and
saturated vapor pressure of the pure components. The pure component
parameters of the PC-SAFT EoS used in this study are summarized in
Table 3 [22,23].
The following mixing rule for the PC-SAFT EoS was used in this
study:
1( )
σ ij = σi + σj (18)
2
( )√̅̅̅̅̅̅̅
εij = 1 − θij εi εj (19)

where θij is the interaction parameter and is a fitting parameter in the

Fig. 5. Comparison of the excess molar volume of the CO2(1)/MeOH(2), correlation. For association, instead of applying the mixing rule to
CO2(1)/EtOH(2), and CO2(1)/Ethane(2) binary systems at 20 MPa and 313.2 K. obtain the mixing parameters, it is needed to consider the way of asso­
ciation between different components and the association site. In this
( )seg ( )hs study, the 1st component, CO2, was not expected to cause intermolec­
gij dij ≈ gij dij
( ) ( )2 ular association, while the 2nd component, alcohol, was expected to
1 di d j 3ζ2 di dj 2ζ22 cause intermolecular association (OROH and HROH). Reilly et al. [40]
= + + (16)
1 − ζ3 di + dj (1 − ζ3 )2
di + d j (1 − ζ3 )3 reported that the association occurs between CO2 and alcohol, where the
association occurs between the oxygen atom (OCO2 ) of CO2 and the
where, ζn is defined as hydrogen atom (HROH) of the hydroxy group of the alcohol. We intro­
π ∑ duced the contribution by the association between CO2 and alcohol into
ζn = ρm xi mi dni n ∈ {0, 1, 2, 3} (17)
6 i
the PC-SAFT EoS in a similar way to Fu et al. [62] The mole fraction not
bonded at each site is given by:
where, xi and mi are the molar fraction and number of segments per [ ]− 1
chain of ith component, respectively. ρm is the total number density of XOCO2 = 1 + NAv ρ2 XHROH ΔOCO2 HROH (20)
molecules. Finally, κAi Bj and εAi Bj are the volume of interaction and as­ [ ]− 1
sociation energy of interaction between site A on ith component and site XOROH = 1 + NAv ρ2 XHROH ΔOROH HROH (21)
B on jth component, respectively. In the PC-SAFT EoS, mi, σ i, and εi are

Table 6
Results from the density correlations for the CO2(1)/MeOH(2) and CO2(1)/EtOH(2) binary systems using the EoSs (temperature T, interaction parameters in SL EoS
and PC-SAFT EoS δij and θij, and volume of interaction and association energy of interaction between an oxygen atom on CO2 and a hydrogen atom on a hydroxy group
of alcohol κOCO2 HROH and εOCO2 HROH k− 1 ).
SL EoS PC-SAFT EoS CACA-PC-SAFT EoS

T (K) δij AARD (%) θij AARD (%) θij κOCO2 HROH / - εOCO2 HROH k− 1 / K AARD (%)

CO2/MeOH system
313.2 0.120 0.59 0.039 0.42 0.050 0.0018 2317 0.45
333.2 0.102 0.81 0.043 1.05 0.043 0.0000 2389 1.05
353.2 0.072 1.25 0.032 1.69 0.032 0.0000 2426 1.69
CO2/EtOH system
313.2 0.078 0.91 0.082 0.62 0.116 0.0179 4632 0.48
333.2 0.055 1.63 0.081 0.52 0.132 0.0769 5386 0.47
353.2 0.025 3.06 0.071 0.72 0.073 0.0000 4037 0.73

10
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Table 7
Vapor–liquid equilibrium data set used in current study.
T/K Number of Standard Authors Refs.
data uncertainties
points

CO2/MeOH system
313.15 12 u(T) = 0.05 K, Joung et al. (2001) [8]
330.00 12 u(P) = 0.01 MPa
342.80 11
313.15 8 No contributions Kodama et al. (1996) [9]
313.2 7 u(T) = 0.1 K, Laursen et al. (2002) [10]
u(P) = 0.01 MPa
313 9 u(P) = 0.001 MPa Ohgaki et al. (1976) [11]
313.15 8 No contributions Naidoo et al. (2008) [12]
313 5 u(T) = 0.1 K, Matsukawa et al. (2021) [13]
330 5 u(P) = 0.01 MPa
343 5
330 13 u(T) = 0.1 K, Reighard et al. (1996) [14]
u(P) = 0.01 MPa
343 9 No contributions Hong et al. (1988) [15]
CO2/EtOH system
313.4 11 u(T) = 0.05 K, Joung et al. (2001) [8]
333.4 11 u(P) = 0.01 MPa
313.15 8 u(T) = 0.1 K, Secuianu et al. (2008) [32]
333.15 9 u(P) = 0.05 MPa
353.15 12
312.82 5 u(T) = 0.03 K, Galicia-Luna et al. [74]
313.15 8 u(P) = 0.02 MPa (2000)
333.82 11
333.75 4

for the density of the two systems (CO2/MeOH, CO2/EtOH). Figs. 1 and
2 illustrate a comparison of the density of pure MeOH and EtOH to
published data [9,63-70]. The solid lines in the figure represent values
calculated using the NIST EoS by the REFPROP. The NIST EoS is a
high-accuracy Helmholtz-energy EoS, and REFPROP uses the most ac­
curate NIST EoS and models available currently. Figs. 1(b) and 2(b)
depict the absolute deviations (AD) from the calculated values:
AD = ρexp − ρcal (24)

where the subscripts exp and ref denote the experimental and calculated
values, respectively. As shown in the figures, the experimental data and
calculated values agree well. Thus, it can be concluded that the exper­
imental apparatus is sufficiently reliable for this measurement.
The density-composition relationships for the CO2/MeOH and CO2/
Fig. 6. Correlation of the densities at 313.2 K with the SL, PC-SAFT, and CACA- EtOH systems are depicted in Figs. 3(a) and (b), respectively. Addi­
PC-SAFT EoS: (a) CO2(1)/MeOH(2) system and (b) CO2(1)/EtOH(2) system. tionally, Figs. 3, 4, 6, and 9 show an interpolation of the density data
The data points in the figure were obtained by the interpolation of the density from Tables 4 and 5. As shown in Fig. 3, at a low CO2 mole fraction, the
data from Tables 4 and 5. density of the CO2/MeOH and CO2/EtOH system increases with
increasing CO2 mole fraction, whereas it decreases (approaches the
[ ( )]− 1 density of pure CO2) at higher CO2 mole fractions. This tendency is
XHROH = 1 + NAv ρ1 XOCO2 ΔOCO2 HROH + ρ2 XOROH ΔOROH HROH (22)
explained by the fact that at low CO2 mole fractions, CO2 penetration
between the organic molecules causes the mixture to become dense. In
where, ΔOCO2 HROH is shown as:
contrast, at high CO2 mole fractions, the organic molecules are sur­
[ ( / ) ]
ΔOCO2 HROH = d12 g12 (d12 )seg κOCO2 HROH exp εOCO2 HROH kT − 1 (23) rounded by CO2, causing the CO2 properties to take precedence. In
general, the density decreases with increasing temperature, with a
In this study, the experimental data were correlated in two different particularly high rate of decrease at high CO2 mole fractions. This
ways, with and without the association between CO2 and alcohol. When observation could be explained by the mixture’s thermal expansion
the association is not introduced, the interaction parameter θij is the characteristics. The thermal expansion coefficient αp of CO2 is 0.010 K −
fitting parameter, and when introduced, the three parameters θij, 1
(at 12 MPa, 300 K) [71], while that of MeOH and EtOH are much lower
κOCO2 HROH , and εOCO2 HROH are the fitting parameters. When the association at 10.68×10− 4 K− 1 (at 20.0 MPa, 303.15 K) [72] and 9.70×10− 4 K− 1 (at
between CO2 and alcohol was considered, it was denoted as CACA 20.0 MPa, 303.15 K) [73], respectively. Thus, the densities of mixtures
(Considering the Association between CO2 and Alcohol)-PC-SAFT EoS. with high CO2 concentrations are significantly different from those with
low CO2 mole fractions, as the properties of CO2 become more dominant
3. Results and discussion as the mole fraction increases.
Figs. 4(a), (b), and (c) compare the experimental results obtained in
3.1. Experimental results this work for the CO2/MeOH and CO2/EtOH systems obtained with
published data [41,42,47-50]. Comparing the results is difficult due to
Figs. 1–4 and Tables 4,5 illustrate the experimental measurements the scarcity of results under the same conditions. However, at 313.2 K,

11
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Fig. 7. Estimation of the vapor-liquid phase diagram at 313.2 K with the SL, PC-SAFT, and CACA-PC-SAFT EoS: (a) CO2(1)/MeOH(2) system and (b) CO2(1)/EtOH(2)
system (closed symbol: liquid phase, open symbol: vapor phase).

capillary viscometer. Their measurement methods and conditions

Table 8 differed from those of other researchers, including us. A concern
Estimation of the vapor-liquid equilibrium lines for the CO2(1)/MeOH(2) and regarding this was that partial condensation might have occurred in the
CO2(1)/EtOH(2) binary systems using the EoSs with interaction parameters
capillary and thus resulted in slightly higher densities. Thus, the trend of
obtained by the density correlations.
density change observed as a function of composition and temperature
T/K AARD /% Number of data points in this work was found to be consistent with the data published in the
SL EoS PC-SAFT CACA-PC-SAFT literature.
EoS EoS In Fig. 5, the excess molar volume of the CO2/MeOH and CO2/EtOH
CO2/MeOH system system at 20 MPa and 313.2 K was compared to the CO2/ethane system
313 92.20 14.43(0) 15.66(0) 49 [8, 9, 10, 11, 12, 13] quoted from REFPROP. The excess molar volume was calculated using
(0) the values from Table 2–3 and the REFPROP values. As evident from the
330 87.11 15.06(0) 31.81(0) 30 [8, 13, 14]
(0)
graph, The CO2/ethane system had a positive excess molar volume,
343 81.66 15.13(0) 28.40(0) 25 [8, 13, 15] whereas the CO2/MeOH and CO2/EtOH systems had a negative excess
(0) molar volume. Additionally, the volume of the CO2/ethane system
CO2/EtOH system expanded due to the repulsion between the non-polar molecules.
313 61.08 36.87(0) Uncalculated 32 [8, 32, 74]
Conversely, it was expected that the volume would contract in the CO2/
(0)
333 33.60 36.00(0) Uncalculated 32 [8, 32, 74] MeOH and CO2/EtOH systems due to attractive interactions.
(2)
353 9.13(0) 24.58(0) 25.76(0) 12 [74]
3.2. Correlation of density with EoS
*The values in parentheses are the number of data points that could not be
estimated and were excluded.
The absolute average relative deviation (AARD) for density given
below was optimized by changing the fitting parameters of the two EoS.
the results of this work were in good agreement with those of Smith et al. ⃒ ⃒
in the CO2/MeOH system, and with those of Kato et al. in the CO2/EtOH 100 ∑ ⃒⃒ρcal − ρexp ⃒⃒
AARD = (24a)
system. In Fig. 4(c), the densities reported by Zhu et al. were slightly Ndata ⃒ ρexp ⃒
higher than the other densities including that obtained in this study.
They performed the measurements under CO2-rich conditions using a where Ndata denotes the number of data points. The parameters and
AARD obtained by correlating the experimental density values with the

12
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Table 9
Results from the vapor-liquid equilibrium line correlations for the CO2(1)/MeOH(2) and CO2(1)/EtOH(2) binary systems using the EoSs.
SL EoS PC-SAFT EoS CACA-PC-SAFT EoS Number of data points

T (K) δij AARD /% θij AARD /% θij κ OCO2 HROH

/- ε OCO2 HROH − 1
k /K AARD /%

CO2/MeOH system
313 − 0.009 18.18(0) 0.016 11.32(0) 0.048 0.0033 2333 5.72(0) 49 [8, 9, 10, 11, 12, 13]
330 − 0.011 17.81(7) 0.017 11.87(0) 0.055 0.0010 2892 3.50(0) 30 [8, 13, 14]
343 − 0.003 25.36(8) 0.014 14.11(0) 0.064 0.0010 3002 7.14(0) 25 [8, 13, 15]
average − 0.008 0.016 0.055 0.0018 2742
CO2/EtOH system
313 0.026 9.46(0) 0.047 8.49(0) 0.089 0.0033 2493 3.52(2) 32 [8, 32, 74]
333 0.030 10.65(2) 0.034 13.58(0) 0.122 0.0098 2401 10.00(0) 32 [8, 32, 74]
353 0.032 4.93(0) 0.039 5.76(0) 0.146 0.0102 2602 3.14(0) 12 [32]

* The values in parentheses are the number of data points that could not be correlated and were excluded.

Fig. 8. Correlation of the vapor-liquid phase diagram at 313 K with the SL, PC-SAFT, and CACA-PC-SAFT EoS: (a) CO2(1)/MeOH(2) system and (b) CO2(1)/EtOH(2)
system (closed symbol: liquid phase, open symbol: vapor phase).

two EoS are tabulated in Table 6 and the fitted curves are shown in absence of an association has no discernible effect on density calculation
Fig. 6. As observed from the table and the figures, the AARD and cor­ by EoS.
relation results for SL and PC-SAFT EoS are identical. Furthermore, in
the PC-SAFT EoS, introducing an association between CO2 and alcohols
3.3. Estimation of VLE with EoS
has no significant effect on the correlation of the density. In particular,
the volume parameter for association,κO1 H2 , was almost zero at 333.2
The VLE line for the CO2/MeOH and CO2/EtOH systems was esti­
and 353.2 K. Furthermore, it was confirmed that the calculated results of
mated using the various EoSs and the interaction parameters determined
density did not change significantly when the parameters related to the
via density correlations. VLE data for the CO2/MeOH [8,15] and
association were changed. These findings suggest that the presence or
CO2/EtOH [8,32,74] systems were obtained for comparison with the

13
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

Table 10 uncertainties and slight discrepancies were indicated in terms of the

Estimation of the density for the CO2(1)/MeOH(2) and CO2(1)/EtOH(2) binary temperature. The VLE dataset summarized in Table 7 was used in the
systems using the EoSs with interaction parameters obtained by the vapor-liquid following sections, with emphasis on the effects of the uncertainties and
equilibrium correlations. slight discrepancies in the temperature on the EoS calculations. The
T/K AARD /% estimated results are shown in Table 8 and Fig. 7 and were compared to
SL EoS PC-SAFT EoS CACA-PC-SAFT EoS published data and the AARD using the following equation.
⃒ ⃒
CO2/MeOH system 100 ∑ ⃒⃒x1,cal − x1,exp ⃒⃒
313.2 4.88 1.17 0.55
AARD = (25)
Ndata ⃒ x1,exp ⃒
333.2 4.87 1.52 1.24
353.2 3.62 1.86 1.85
CO2/EtOH system
where x1 is the CO2 mole fraction. Regrettably, no experimental VLE for
313.2 1.95 1.20 0.72 the CO2/MeOH system operating at the same temperature conditions as
333.2 2.01 2.01 0.65 those used in this work was discovered. As a result, the VLE was esti­
353.2 3.08 1.95 1.30 mated for the experimental temperature conditions. The interaction
* The values in parentheses are the number of data points that could not be parameters for 330 and 343 K used in the estimation are based on the
estimated and were excluded. average values listed in Table 6. The values in parentheses in Table 8
represent the number of data points that could not be estimated and
were therefore excluded. The excluded data points were mainly in terms
of high pressure, high CO2 composition, and were observed in the near-
critical region of the mixture. As shown in Table 8, no composition
satisfied the equilibrium conditions under all conditions when the VLE
of the CO2/EtOH system was deduced using the CACA-PC-SAFT EoS. As
a result, Fig. 7(b) does not include the results of the CACA-PC-SAFT EoS.
The interaction parameters in each EoS represent the differences in
structure and energy between the component molecules and not the
physical properties. Ideally, the VLE line can be estimated using the
parameters determined from the density correlations. As shown in
Table 8 and Fig. 7, PC-SAFT EoS was able to estimate the VLE more
accurately than SL EoS. Additionally, although the PC-SAFT EoS did not
show any effect of considering the association of CO2 and alcohol in the
density correlation, the accuracy of the VLE estimation could be further
improved by considering the association between CO2 and alcohol
(CACA-PC-SAFT EoS). The SL EoS disregards molecular association
because the molecules are randomly arranged on the lattice. From the
results of VLE estimation, it is clear that the consideration of association
affects VLE estimation. However, in the CO2/EtOH system, the SL EoS
had slightly higher estimation accuracy, and several conditions were not
estimable using the CACA-PC-SAFT. As a result, the density correlation
and VLE estimation results alone cannot adequately explain the impor­
tance of considering the association in the EoS calculation.

3.4. Estimation of density with EoS

In Section 3.3, VLE was estimated using the parameter set derived
from the density correlation. As illustrated in Fig. 6, the estimation ac­
curacy was observed to be insufficient, particularly for the CO2/EtOH
system. Thus, the estimation of density was carried out in this section
using the parameter sets determined by the VLE. To optimize Eq. (25),
the interaction parameters were determined from the VLE correlations
using each of the EoS. The correlation results for VLEs are tabulated in
Table 9 and Fig. 8. The number in parentheses represents the number of
data points that could not be correlated and were thus excluded. The
data points excluded were primarily those comprising a high pressure, a
high CO2 composition, and those in the mixture’s near-critical region.
From the table, the AARD of PC-SAFT EoS was lower than that of SL EoS.
Furthermore, the AARD of CACA-PC-SAFT EoS was lower than PC-SAFT,
Fig. 9. Estimation of the densities at 313.2 K with the SL, PC-SAFT, and CACA- and the correlation near the critical region of mixture was relatively
PC-SAFT EoS: (a) CO2(1)/MeOH(2) system and (b) CO2(1)/EtOH(2) system. well. Although the association between CO2 and alcohols was not
The data points in the figure were obtained by the interpolation of the density observed in the correlation of the density of homogeneous mixed-fluids
data from Tables 4 and 5.
though, it is abundantly clear that the incorporation of association term
improves the VLE correlation, particularly near the critical region. The
estimated results. The VLE data obtained are summarized in Table 7. The SL EoS and PC-SAFT EoS have been shown to be ineffective for calcu­
result uncertainties were provided in some previous publications. lations in the critical region. This is because the densities of the vapor
Additionally, some of the temperatures did not match exactly. Fig. 7 and liquid phases are almost identical near the critical region. In fact,
shows the experimental VLE data obtained from the literature. As some researchers attempted to solve this problem by improving the
shown, most of the VLE points were on the same line, although some density calculations near the critical region [75-78]. Based on the results

14
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

of density correlation yielded by the EoS, the CACA-PC-SAFT EoS per­ determined under a wide range of conditions, while considering the
formed the best in calculating density, which resulted in a favorable associations occurring in the system. These results suggest that the
correlation near the critical region. Reilly et al. performed FT-IR mea­ reviewal of the used parameters is required to use EoS for the estimation
surements under CO2-rich conditions, as described in the introduction. of several physical properties. Moreover, in the future, the properties
The CO2-rich condition also comprises the critical region of the mixture. required to determine the parameter set should be a wide range of
In other words, their report established that the association between CO2 experimental conditions, and the ability of each EoS to estimate the
and alcohol occurs at or near the critical region of the mixture. On the density and VLE, as well as other properties, should be examined.
other hand, the association between CO2 and alcohols had a significant
effect on the accuracy of the VLE correlation especially near the critical CRediT authorship contribution statement
point of mixtures. These findings demonstrated the critical nature of the
association near the critical condition. The lack of sensitivity of the Hiroaki Matsukawa: Writing – original draft, Formal analysis, Data
density of the homogeneous mixed-fluid to the incorporation association curation. Masamune Yomori: Data curation. Tomoya Tsuji: Supervi­
term may be explained by the alcohol concentration. Additionally, the sion. Katsuto Otake: Writing – review & editing, Supervision.
mole fraction of 0.8 used in our experiments might be too far from the
critical condition to be affected by the association. Table 10 and Fig. 9 Declaration of competing interest
illustrate the estimated density using the interaction parameters derived
from the VLE correlations. As mentioned previously, no experimental The authors declare that they have no known competing financial
VLE for the CO2/MeOH system operating at the same temperature interests or personal relationships that could have appeared to influence
conditions as those used in this work was discovered. The interaction the work reported in this paper.
parameters for 330 and 343 K used in the estimation are average values
in Table 9. The table and figure demonstrate that the PC-SAFT EoS Data availability
accurately estimates the density of the homogeneous mixed-fluid,
whereas the SL EoS overestimates the change in relation to the CO2 Data will be made available on request.
mole fraction. To summarize, in systems with molecular associations,
various properties can be estimated using EoS by considering all asso­
ciations and a set of parameters determined under a wide range of Acknowledgements
experimental conditions, particularly near the critical region.
Funding sources.
4. Conclusions
Supplementary materials
Using a high-pressure vibration-type density meter, the densities of
homogeneous fluid mixtures comprising CO2/MeOH and CO2/EtOH Supplementary material associated with this article can be found, in
binary systems were determined over a wide range of temperature, the online version, at doi:10.1016/j.fluid.2024.114140.
pressure, and composition. At a pressure of 20 MPa and a temperature of
313.2 K, the excess molar volume of the CO2/MeOH and CO2/EtOH References
systems were compared to the CO2/ethane system. The attractive
interaction was confirmed to exist in the CO2/MeOH and CO2/EtOH [1] S. Rudyk, S. Hussain, P. Spirov, Supercritical extraction of crude oil by methanol-
systems, affecting the mixture’s volume. Additionally, the experimental and ethanol-modified carbon dioxide, J. Supercrit. Fluids 78 (2013) 63–69,
https://doi.org/10.1016/j.supflu.2013.03.006.
data were fitted to three different EoSs, viz. SL EoS, PC-SAFT EoS, and [2] S.H. Kim, S.S. Kang, C. Huh, S.-G. Kang, J.W. Kang, C.S. Lee, Methanol inhibition
CACA-PC-SAFT EoS. Subsequently, the individual parameters sets for effect on hydrate-containing phase equilibria of carbon dioxide and water,
each EoS were determined by fitting the experimental density data. The J. Chem. Eng. Data 57 (2012) 2286–2289, https://doi.org/10.1021/je3002909.
[3] M. Gatti, E. Martelli, F. Marechal, S. Consonni, Review, modeling, heat Integration,
accuracy of the density correlations was nearly identical for all the EoSs. and improved schemes of Rectisol®-based processes for CO2 capture, Appl. Therm.
Additionally, the association between CO2 and alcohols in the PC-SAFT Eng. 70 (2014) 1123–1140, https://doi.org/10.1016/j.
EoS had no discernible effect on the reliability of the density correla­ applthermaleng.2014.05.001.
[4] R.G. Bitencourt, A.M. Palma, J.A.P. Coutinho, F.A. Cabral, A.J.A. Meirelles,
tions. Furthermore, the VLE of the CO2/MeOH and CO2/EtOH binary
Solubility of caffeic acid in CO2 + ethanol: experimental and predicted data using
systems were estimated using the parameter set determined from the Cubic Plus Association Equation of State, J. Supercrit. Fluids 138 (2018) 238–246,
experimental density measurements. Based on the results obtained from https://doi.org/10.1016/j.supflu.2018.04.008.
the examination of the molecular associations, the two PC-SAFT EoSs [5] S. Al-Hamimi, A. Abellan Mayoral, L.P. Cunico, C. Turner, Carbon dioxide
expanded ethanol extraction: solubility and extraction kinetics of alpha-pinene and
performed significantly better than the SL EoS in terms of estimation cis-verbenol, Anal. Chem. 88 (2016) 4336–4345, https://doi.org/10.1021/acs.
accuracy. However, for all EoSs, the estimation accuracy near the mix­ analchem.5b04534.
ture’s critical region was observed to be suboptimal. [6] R. Campardelli, E. Reverchon, I. De Marco, PVP microparticles precipitation from
acetone-ethanol mixtures using SAS process: effect of phase behavior, J. Supercrit.
All EoSs had their parameter sets determined using VLE data from Fluids 143 (2019) 321–329, https://doi.org/10.1016/j.supflu.2018.09.010.
the literature. The CACA-PC-SAFT EoS that considered the association [7] V. Prosapio, E. Reverchon, I. De Marco, Polymers’ ultrafine particles for drug
between CO2 and alcohols demonstrated a strong correlation throughout delivery systems precipitated by supercritical carbon dioxide + organic solvent
mixtures, Powder Technol. 292 (2016) 140–148, https://doi.org/10.1016/j.
the mixture, including the critical region where the association between powtec.2016.01.033.
CO2 and alcohols was observed to be highly significant. Thus, with the [8] S.N. Joung, C.W. Yoo, H.Y. Shin, S.Y. Kim, K.-P. Yoo, C.S. Lee, W.S. Huh,
parameter set determined, the density of CO2/MeOH and CO2/EtOH Measurements and correlation of high-pressure VLE of binary CO2-alcohol systems
(methanol, ethanol, 2-methoxyethanol and 2-ethoxyethanol), Fluid Phase Equilib.
binary systems were further estimated. As observed, the CACA-PC-SAFT 185 (2001) 219–230, https://doi.org/10.1016/S0378-3812(01)00472-1.
Eos was able to accurately estimate the density of the homogeneous fluid [9] D. Kodama, N. Kubota, Y. Yamaki, H. Tanaka, M. Kato, High pressure vapor-liquid
mixture. equilibria and density behaviors for carbon dioxide + methanol system at 313.15
K, Netsu Bussei 10 (1996) 16–20, https://doi.org/10.2963/jjtp.10.16.
The reason that the density estimations using VLE correlations was
[10] T. Laursen, P. Rasmussen, S.I. Andersen, VLE and VLLE measurements of dimethyl
successful but VLE estimation from density correlations was not pre­ ether containing systems, J. Chem. Eng. Data 47 (2002) 198–202, https://doi.org/
sumably attributed to the fact that the experimental condition pertain­ 10.1021/je010154. +.
ing to the density measurements did not include the critical region. [11] K. Ohgaki, T. Katayama, Isothermal vapor-liquid equilibrium data for binary
systems containing carbon dioxide at high-pressures; methanol-carbon dioxide, n-
Additionally, when considering the efficient estimation of a wide variety hexane-carbond dioxide, and benzene-carbon dioxide systems, J. Chem. Eng. Data
of physical properties, it was discovered that the parameter sets must be 21 (1976) 53–55, https://doi.org/10.1021/je60068a015.

15
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

[12] P. Naidoo, D. Ramjugernath, J.D. Raal, A new high-pressure vapour–liquid [40] J.T. Reilly, C.P. Bokis, M.D. Donohue, An experimental investigation of Lewis acid-
equilibrium apparatus, Fluid Phase Equilib. 269 (2008) 104–112, https://doi.org/ base interactions of liquid carbon dioxide using Fourier transform infrared (FT-IR)
10.1016/j.fluid.2008.05.002. spectroscopy, Int. J. Thermophys. 16 (1995) 599–610, https://doi.org/10.1007/
[13] H. Matsukawa, K. Otake, Phase behavior of carbon dioxide/tetrapropyl BF01438845.
orthosilicate and tetrabutyl orthosilicate systems, Fluid Phase Equilib. 548 (2021) [41] J. Richard K. Smith, S.B. Lee, S. Suzuki, C. Saito, H. Inomata, K. Arai, Densities of
113172, https://doi.org/10.1016/j.fluid.2021.113172. carbon dioxide + methanol mixtures at temperatures from 313.2 to 323.2 K and at
[14] T.S. Reighard, S.T. Lee, S.V. Olesik, Determination of methanol/CO2 and pressures from 10 to 20 MPa, J. Chem. Eng. Data 47 (2002) 608–612, https://doi.
acetonitrile/CO2 vapor-liquid phase equilibria using a variable-volume view cell, org/10.1021/je010207n.
Fluid Phase Equilib. 123 (1996) 215–230, https://doi.org/10.1016/S0378-3812 [42] L.A. Galicia-Luna, D. Richon, H. Renon, New loading technique for a vibrating tube
(96)90029-1. densimeter and measurements of liquid densities up to 39.5 MPa for binary and
[15] J.H. Hong, R. Kobayashi, Vapor—Liquid equilibrium studies for the carbon ternary mixtures of the carbon dioxide-methanol-propane system, J. Chem. Eng.
dioxide—methanol system, Fluid Phase Equilib. 41 (1988) 269–276, https://doi. Data 39 (1994) 424–431, https://doi.org/10.1021/je00015a005.
org/10.1016/0378-3812(88)80011-6. [43] E. Brunner, W. Hültenschmidt, G. Schlichthärle, Fluid mixtures at high pressures
[16] E. Moine, A. Piña-Martinez, J.-N. Jaubert, B. Sirjean, R. Privat, I-PC-SAFT: an IV. Isothermal phase equilibria in binary mixtures consisting of (methanol +
industrialized version of the volume-translated PC-SAFT equation of state for pure hydrogen or nitrogen or methane or carbon monoxide or carbon dioxide), J. Chem.
components, resulting from experience acquired all through the years on the Thermodyn. 19 (1987) 273–291, https://doi.org/10.1016/0021-9614(87)90135-
parameterization of SAFT-type and cubic models, Ind. Eng. Chem. Res. 58 (2019) 2.
20815–20827, https://doi.org/10.1021/acs.iecr.9b04660. [44] M. Kariznovi, H. Nourozieh, J. Abedi, Experimental measurements and predictions
[17] R.R. Mallepally, B.A. Bamgbade, N. Cain, M.A. McHugh, Phase behavior and of density, viscosity, and carbon dioxide solubility in methanol, ethanol, and 1-
densities of propylene + toluene and ethylene + toluene systems to 580 K and 70 propanol, J. Chem. Thermodyn. 57 (2013) 408–415, https://doi.org/10.1016/j.
MPa, Fluid Phase Equilib. 449 (2017) 138–147, https://doi.org/10.1016/j. jct.2012.10.002.
fluid.2017.06.022. [45] J. CC, C. Kou-Lung, D. Chang-Yih, A new apparatus for the determination of P-x-y
[18] D.-Y. Peng, D.B. Robinson, A new two-constant equation of state, Ind. Eng. Chem. diagrams and Henry’s constants in high pressure alcohols with critical carbon
Res. 15 (1976) 59–64, https://doi.org/10.1021/i160057a011. dioxide, J. Supercrit Fluids 12 (1998) 223–237, https://doi.org/10.1016/S0896-
[19] K.-P. Yoo, H.Y. Shin, C.S. Lee, Approximate nonrandom two-fluid lattice-hole 8446(98)00076-X.
theory. general derivation and description of pure fluids, Bull. Korean Chem. Soc. [46] R. Sih, F. Dehghani, N.R. Foster, Viscosity measurements on gas expanded liquid
18 (1997) 965–972, https://doi.org/10.5012/bkcs.1997.18.9.965. systems—methanol and carbon dioxide, J. Supercrit. Fluids 41 (2007) 148–157,
[20] K.-P. Yoo, H.Y. Shin, C.S. Lee, Approximate nonrandom two-fluid lattice-hole https://doi.org/10.1016/j.supflu.2006.09.002.
theory. thermodynamic properties of real mixtures, Bull. Korean Chem. Soc. 18 [47] M. Kato, D. Kodama, T. Ono, M. Kokubo, Volumetric properties of carbon dioxide
(1997) 841–850, https://doi.org/10.5012/bkcs.1997.18.8.841. + ethanol at 313.15 K, J. Chem. Eng. Data 54 (2009) 2953–2956, https://doi.org/
[21] M. Khalifa, B. Housam, B. Ahmed, Modeling of the phase behavior of CO2 in water, 10.1021/je900127m.
methanol, ethanol and acetone by different equations of state, Fluid Phase Equilib. [48] L.P. Cunico, C. Turner, Density measurements of CO2-expanded liquids, J. Chem.
469 (2018) 9–25, https://doi.org/10.1016/j.fluid.2018.04.006. Eng. Data 62 (2017) 3525–3533, https://doi.org/10.1021/acs.jced.7b00540.
[22] J. Gross, G. Sadowski, Perturbed-Chain SAFT: an equation of state based on a [49] A. Zúñiga-Moreno, L.A. Galicia-Luna, Compressed liquid densities of carbon
perturbation theory for chain molecules, Ind. Eng. Chem. Res. 40 (2001) dioxide + ethanol mixtures at four compositions via a vibrating tube densimeter up
1244–1260, https://doi.org/10.1021/ie0003887. to 363 K and 25 MPa, J. Chem. Eng. Data 47 (2002) 149–154, https://doi.org/
[23] J. Gross, G. Sadowski, Application of the perturbed-chain SAFT equation of state to 10.1021/je010012f.
associating systems, Ind. Eng. Chem. Res. 41 (2002) 5510–5515, https://doi.org/ [50] T. Zhu, H. Gong, M. Dong, Density and viscosity of CO2 + ethanol binary systems
10.1021/ie0003887. measured by a capillary viscometer from 308.15 to 338.15 K and 15 to 45 MPa,
[24] J. Vrabec, J. Gross, Vapor-Liquid equilibria simulation and an equation of state J. Chem. Eng. Data 65 (2020) 3820–3833, https://doi.org/10.1021/acs.
contribution for dipole-quadrupole interactions, J. Phys. Chem. B 112 (2008) jced.0c00175.
51–60, https://doi.org/10.1021/jp072619u. [51] I. Tsivintzelis, D. Missopolinou, K. Kalogiannis, C. Panayiotou, Phase compositions
[25] J. Gross, J. Vrabec, An equation-of-state contribution for polar components: dipolar and saturated densities for the binary systems of carbon dioxide with ethanol and
molecules, AlChE J. 52 (2006) 1194–1204, https://doi.org/10.1002/aic.10683. dichloromethane, Fluid Phase Equilib. 224 (2004) 89–96, https://doi.org/
[26] M. Kleiner, J. Gross, An equation of state contribution for polar components: 10.1016/j.fluid.2004.06.046.
polarizable dipoles, AlChE J. 52 (2006) 1951–1961, https://doi.org/10.1021/ [52] M. Stievano, N. Elvassore, High-pressure density and vapor–liquid equilibrium for
jp072619u. the binary systems carbon dioxide–ethanol, carbon dioxide–acetone and carbon
[27] G.M. Kontogeorgis, E.C. Voutsas, I.V. Yakoumis, D.P. Tassios, An equation of state dioxide–dichloromethane, J. Supercrit. Fluids 33 (2005) 7–14, https://doi.org/
for associating fluids, Ind. Eng. Chem. Res. 35 (1996) 4310–4318, https://doi.org/ 10.1016/j.supflu.2004.04.003.
10.1021/ie9600203. [53] C.-Y. Day, C.J. Chang, C.-Y. Chen, Phase equilibrium of ethanol + CO2 and acetone
[28] V. Feroiu, D. Geana, Computation of vapour—Liquid equilibrium in N2 + CO2 + + CO2 at elevated pressures, J. Chem. Eng. Data 41 (1996) 839–843, https://doi.
CH4 system by a general cubic equation of state, Fluid Phase Equilib. 55 (1990) org/10.1021/je960049d.
263–270, https://doi.org/10.1016/0378-3812(90)85022-3. [54] W.G. Chapman, K.E. Gubbins, G. Jackson, M. Radosz, New reference equation of
[29] D. Geană, V. Feroiu, Calculation of Joule-Thomson inversion curves from a general state for associating liquids, Ind. Eng. Chem. Res. 29 (1990) 1709–1721, https://
cubic equation of state, Fluid Phase Equilib. 77 (1992) 121–132, https://doi.org/ doi.org/10.1021/ie00104a021.
10.1016/0378-3812(92)85100-M. [55] S.H. Huang, M. Radosz, Equation of state for small, large, polydisperse, and
[30] D. Geană, V. Feroiu, Thermodynamic properties of pure fluids using the GEOS3C associating molecules: extension to fluid mixtures, Ind. Eng. Chem. Res. 30 (1991)
equation of state, Fluid Phase Equilib. 174 (2000) 51–68, https://doi.org/10.1016/ 1994–2005, https://doi.org/10.1021/ie00056a050.
S0378-3812(00)00417-9. [56] S.H. Huang, M. Radosz, Equation of state for small, large, polydisperse, and
[31] V. Feroiu, D. Geană, Volumetric and thermodynamic properties for pure associating molecules, Ind. Eng. Chem. Res. 29 (1990) 2284–2294, https://doi.
refrigerants and refrigerant mixtures from cubic equations of state, Fluid Phase org/10.1021/ie00107a014.
Equilib. 207 (2003) 283–300, https://doi.org/10.1016/S0378-3812(03)00034-7. [57] E.C. Ihmels, J. Gmehling, Densities of toluene, carbon dioxide, carbonyl sulfide,
[32] C. Secuianu, V. Feroiu, D. Geană, Phase behavior for carbon dioxide+ethanol and hydrogen sulfide over a wide temperature and pressure range in the sub- and
system: experimental measurements and modeling with a cubic equation of state, supercritical state, Ind. Eng. Chem. Res. 40 (2001) 4470–4477, https://doi.org/
J. Supercrit. Fluids 47 (2008) 109–116, https://doi.org/10.1016/j. 10.1021/ie001135g.
supflu.2008.08.004. [58] S.K. Garg, T.S. Banipal, J.C. Ahluwalia, Heat capacities and densities of liquid o-
[33] S.G. Cardoso, G.M.N. Costa, S.A.B. Vieira de Melo, Assessment of the liquid mixture xylene, m-xylene, and ethylbenzene, at temperatures from 318.15 K to 373.15 K
density effect on the prediction of supercritical carbon dioxide volume expansion and at pressures up to 10 MPa, J. Chem. Thermodyn. 25 (1993) 57–62, https://doi.
of organic solvents by Peng-Robinson equation of state, Fluid Phase Equilib. 425 org/10.1006/jcht.1993.1007.
(2016) 196–205, https://doi.org/10.1016/j.fluid.2016.06.006. [59] Y.M. Naziev, A.N. Shahverdiyev, V.H. Hasanov, The (p,ρ,T) of (methanol+
[34] G. Soave, Equilibrium constants from a modified Redlich-Kwong equation of state, benzene) and (methanol+ethylbenzene), J. Chem. Thermodyn. 37 (2005)
Chem. Eng. Sci. 27 (1972) 1197–1203, https://doi.org/10.1016/0009-2509(72) 1268–1275, https://doi.org/10.1016/j.jct.2005.03.006.
80096-4. [60] H. Matsukawa, K. Kuwabara, Y. Shimada, T. Tsuji, K. Otake, Densities for carbon
[35] R.A. Krenz, T. Laursen, R.A. Heidemann, The modified Sanchz-Lacombe equation dioxide/cyclohexane, methylcyclohexane, and ethylbenzene systems, J. Chem.
of state applied to polydisperse polyethylene solutions, Ind. Eng. Chem. Res. 48 Thermodyn. 142 (2020) 105902, https://doi.org/10.1016/j.jct.2019.105902.
(2009) 10664–10681, https://doi.org/10.1021/ie801451a. [61] Y. Xiong, E. Kiran, Prediction of high-pressure phase behaviour in polyethylene/n-
[36] R.H. Lacombe, I.C. Sanchez, Statistical thermodynamics of fluid mixtures, J. Phys. pentane/carbon dioxide ternary system with the Sanchez-Lacombe model, Polymer
Chem. 80 (1976) 2568–2580, https://doi.org/10.1021/j100564a009. 35 (1994) 4408–4415, https://doi.org/10.1016/0032-3861(94)90100-7.
[37] I.C. Sanchez, R.H. Lacombe, An elementary molecular theory of classical fluids, [62] D. Fu, X. Hua, Y. Xu, Cross-Association model for the phase equilibria and surface
Pure Fluids, J. Phys. Chem. 80 (1976) 2352–2362, https://doi.org/10.1021/ tensions of CO2− methanol and CO2− ethanol mixtures, J. Phys. Chem. C 115
j100562a008. (2011) 3340–3345, https://doi.org/10.1021/jp108400z.
[38] I.C. Sanchez, R.H. Lacombe, Statistical thermodynamics of polymer solutions, [63] J.R.S. Machado, W.B. Streett, Equation of state and thermodynamic properties of
Macromolecules 11 (1978) 1145–1156, https://doi.org/10.1021/ma60066a017. liquid methanol from 298 to 489 K and pressures to 1040 bars, J. Chem. Eng. Data
[39] H. Matsukawa, T. Tsuji, K. Otake, Measurement of the density of carbon dioxide/ 28 (1983) 218–223, https://doi.org/10.1021/je00032a029.
toluene homogeneous mixtures and correlation with equations of state, J. Chem. [64] F.A.M.M. Gonçalves, A.R. Trindade, C.S.M.F. Costa, J.C.S. Bernardo, I. Johnson, I.
Thermodyn. 164 (2022) 106618, https://doi.org/10.1016/j.jct.2021.106618. M.A. Fonseca, A.G.M. Ferreira, PVT, viscosity, and surface tension of ethanol: new

16
H. Matsukawa et al. Fluid Phase Equilibria 584 (2024) 114140

measurements and literature data evaluation, J. Chem. Thermodyn. 42 (2010) [71] M. Chorążewski, J.-P.E. Grolier, S.L. Randzio, Isobaric thermal expansivities of
1039–1049, https://doi.org/10.1016/j.jct.2010.03.022. toluene measured by scanning transitiometry at temperatures from (243 to 423) K
[65] D. Sommer, R. Kleinrahm, R. Span, W. Wagner, Measurement and correlation of and pressures up to 200 MPa, J. Chem. Eng. Data 55 (2010) 5489–5496, https://
the (p,ρ,T) relation of liquid cyclohexane, toluene, and ethanol in the temperature doi.org/10.1021/je100657n.
range from 233.15K to 473.15K at pressures up to 30MPa for use as density [72] T. Sun, S.N. Biswas, N.J. Trappenlers, C.A.T. Seldam, Acoustic and thermodynamic
reference liquids, J. Chem. Thermodyn. 43 (2011) 117–132, https://doi.org/ properties of methanol from 273 to 333 K and at pressures to 280 MPa, J. Chem.
10.1016/j.jct.2010.08.010. Eng. Data 33 (1988) 395–398, https://doi.org/10.1021/je00054a002.
[66] D.N.F. Muche, G.V. Olivieri, R.B. Torres, Density and derived properties of binary [73] T.F. Sun, C.A. Ten Seldam, P.J. Kortbeek, N.J. Trappeniers, S.N. Biswas, Acoustic
mixtures containing {2-(dimethylamino)ethyl methacrylate + alcohols} at and thermodynamic properties of ethanol from 273.15 to 333.1 5 K and up to 280
temperatures from T = (293.15 to 313.15) K and pressures of up to 35 MPa, MPa, Phys. Chem. Liq. 18 (1988) 107–116, https://doi.org/10.1080/
J. Chem. Eng. Data 64 (2019) 1909–1921, https://doi.org/10.1021/acs. 00319108808078584.
jced.8b00975. [74] L.A. Galicia-Luna, A. Ortega-Rodriguez, New appratus for the fast determination of
[67] A. Goldon, K. Dabrowska, T. Hofman, Densities and excess volumes of the 1,3- high-pressure vapor-liquid equilibria of mixtures and of accurate critical pressures,
dimethylimidazolium methylsulfate + methanol system at temperatures from J. Chem. Eng. Data 45 (2000) 265–271, https://doi.org/10.1021/je990187d.
(313.15 to 333.15) K and pressures from (0.1 to 25) MPa, J. Chem. Eng. Data 52 [75] G. Xu, N.Gerek Ince, Tangent point approach in reliably solving the density roots
(2007) 1830–1837, https://doi.org/10.1021/je700188w. for PC-SAFT equation of state, J. Chem. Eng. Data 65 (2020) 5643–5653, https://
[68] M.J.P. Comunas, E.R. Lopez, P. Pires, J. Garcia, J. Fernandez, pρT measurements of doi.org/10.1021/acs.jced.0c00638.
polyethylene glycol dimethylethers between 278.15 and 328.15 K at pressures to [76] A. Barra, A. Moro, Exact solution of the van der Waals model in the critical region,
12 MPa, Int. J. Thermophys. 21 (2000) 831–851, https://doi.org/10.1023/A: Ann. Phys. 359 (2015) 290–299, https://doi.org/10.1016/j.aop.2015.04.032.
1006606122944. [77] R. Tomaschitz, Phenomenological high-pressure equation of state for nitrogen,
[69] C.K. Zéberg-Mikkelsen, L. Lugo, J. Fernández, Density measurements under methane, methanol, carbon dioxide, and helium, Int. J. Thermophys. (2022) 43,
pressure for the binary system (ethanol+methylcyclohexane), J. Chem. https://doi.org/10.1007/s10765-022-03034-9.
Thermodyn. 37 (2005) 1294–1304, https://doi.org/10.1016/j.jct.2005.03.016. [78] E.M. Apfelbaum, V.S. Vorob’ev, Universal equation of state for critical and
[70] P. Sauermann, K. Holzapfel, J. Oprzynski, F. Kohler, W. Poot, T.W. deLoos, The pρT supercritical regions, High Temp. 60 (2023) S339–S346, https://doi.org/10.1134/
properties of ethanol + hexane, Fluid Phase Equilib. 112 (1995) 249–272, https:// s0018151×21060018.
doi.org/10.1016/0378-3812(95)02798-J.

17