Fluid Phase Equilibria 219 (2004) 25–32

Binary and ternary solubilities of disperse dyes and their

blend in supercritical carbon dioxide
Kazuhiro Tamura∗ , Takashi Shinoda
a Department of Chemistry and Chemical Engineering, Division of Applied Sciences, 40-20 Kodatsuno 2-chome,
Kanazawa University, Kanazawa, Ishikawa 920-8667, Japan

Abstract

Binary and ternary solubilities of C.I. Disperse Blue 134 (1,4-bis(isopropylamino)anthraquinone) C.I. Disperse Yellow 16 (3-methyl-1-
phenyl-5-pyrazolone) and their dye mixture in supercritical carbon dioxide (SC-CO2 ) were measured by a flow-type apparatus. The solubility
measurements were carried out at the pressure ranges from 10.0 to 25.0 MPa for the binary systems at the temperatures from 323.15 to
383.15 K and for the ternary system at 383.15 K. An empirical equation was used to correlate the experimental binary solubilities of the dyes
in terms of the density of carbon dioxide. To represent accurately the binary solubility of the dyes in terms of temperature and pressure, we
used a modified Peng–Robinson–Stryjek–Vera equation of state (PRSV EOS). The ternary solubilities of the dye blend could be predicted
successfully from binary parameters with the modified PRSV EOS.
© 2004 Elsevier B.V. All rights reserved.

Keywords: Solubility; Disperse dye; Supercritical carbon dioxide

1. Introduction SC-CO2 and their accurate representation using a thermo-

dynamic model.
Supercritical fluid dyeing (SFD) processes have attracted Dye solubilities in supercritical fluids have been reported
considerable attention as a replacement for the conventional [3–11], but those of mixed dye compounds in SC-CO2 have
wet dyeing from a viewpoint of environmental protection not yet been measured so far. On the other hand, good pre-
and sustainable development [1,2]. In the ordinary wet dye- dictive and correlation methods for the dye solubilities in
ing process, a large amount of water and antibiodegradable SC-CO2 are desirable. Whenever representing the dye solu-
additives such as surfactants and dispersion agents in addi- bilities in SC-CO2 within the thermodynamic framework of
tion to dyestuffs have been discharged and then waste water solid–fluid equilibria, we need to evaluate the fugacity coef-
including the toxic chemicals is treated to achieve acceptable ficient in supercritical phase, vapor pressure and molar vol-
emission standards. To reduce water pollution and energy ume of dye. However, experimental values are lacking and
consumption related to the wet dyeing process, we adopted reliable estimation for these properties of solid dye by the
SFD using carbon dioxide as an alternative to the conven- well-known group contribution methods. Above all, the va-
tional dyeing. Carbon dioxide is a benign, nonflammable and por pressures of the dye cannot be calculated reliably owing
inexpensive solvent, and acts as a useful processing fluid for to several auxiliary functional groups unavailable in their
SFD at supercritical conditions. Supercritical carbon dioxide chemically complex structure of dye. Even when the group
(SC-CO2 ) can be separated easily from the dyeing medium parameters are available, the models often fail to represent
to recycle perfectly by expansion. Moreover the SFD pro- the solid-like properties accurately.
cess can save and reuse the additives and dyes more than In this work, we measured the binary solubilities of C.I.
the wet dyeing process. To develop and assist the proper Disperse Blue 134 (1,4-bis(isopropylamino)anthraquinone)
design of the SFD process, we need experimental solubili- and C.I. Disperse Yellow 16 (3-methyl-1-phenyl-5-pyrazo-
ties of various kinds of dyes and the blends of their dyes in lone) in SC-CO2 at temperatures of 323.15, 353.15 and
383.15 K and at pressures from 10.0 to 25.0 MPa and the
∗ Corresponding author. Tel.: +81-76-234-4804; ternaries of their disperse dye mixture at 383.15 K over
fax: +81-76-234-4829. pressures from 10.0 to 25.0 MPa. The experimental results
E-mail address: [email protected] (K. Tamura). of the dyes were compared with those calculated using

0378-3812/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.fluid.2004.01.009
26 K. Tamura, T. Shinoda / Fluid Phase Equilibria 219 (2004) 25–32

the equations of state (PR EOS and PRSV EOS) proposed

by Peng–Robinson [12] and Peng–Robinson–Stryjek–Vera
[13], and a modified PRSV EOS published previously [11].
For accurate representation of the binary solubilities of the
dyes in SC-CO2 , we evaluated simultaneously the vapor
pressure and molar volume of the dyes as well as the fugac-
ity coefficients of the dyes in the supercritical phase using
the PR, PRSV EOS and modified form of PRSV EOS. The
ternary experimental results were compared with those pre-
dicted using the modified PRSV EOS with the vapor pressure
and molar volume of the pure dyes, and binary parameters
alone obtained from the constituent binary correlation.
Fig. 2. Schematic diagram of experimental apparatus.

2. Experimental
preheating coil and equilibrium cell, monitored by a digi-
2.1. Materials tal thermometer (1502A, Hart Scientific) with a PRT probe,
was controlled within ±0.1 K in the oven. The pressure of
C.I. Disperse Blue 134 and C.I. Disperse Yellow 16 whose the cell was maintained within ±0.2 MPa with the back-
purities are more than 97% were kindly supplied by Arimoto pressure regulator. The pressure transducer in the regulator
Chemical Co. Ltd. Any additives such as dispersing agents was calibrated using a pressure indicator (DPI145, Druck).
and surfactants were not detected in the disperse dyes, dis- The line between the exit of the oven and the cold trap is
solved in ethanol, by scanning the UV-Vis wavelength from wapped with a flexible heater to prevent from clogging with
190 to 1100 nm. The chemical structures of the dyestuffs dry ice or precipitation of dye in the flow lines. After the
are shown in Fig. 1. Carbon dioxide (purity >99.9%) was temperature and pressure in the system reached to the super-
obtained from Uno Sanso Co. These compounds were used critical conditions at a steady state, the six-way valve was
without further purification. turned to flow the carbon dioxide into the extraction cell and
mix it with the dye. The supercritical fluid mixture including
2.2. Apparatus and procedure the dye was released at the back-pressure regulator. The dye
was trapped in a two-step cold trap filled with ethanol. The
A detailed description of the equipment and operating dye precipitated in the system during the depressurization
procedure, and the reliability of the solubility measurement was additionally recovered by washing out with ethanol. A
is given elsewhere [11]. A flow-type apparatus (SCF-201, UV-Vis spectrophotometer (BioSpec-1600, Shimadzu) was
JASCO) was used to measure the solubility of disperse dye in utilized to analyze the concentration of the dye dissolved in
SC-CO2 . The schematic diagram of the apparatus is shown ethanol solution. A wet gas meter (W-NK-1B, Shinagawa)
in Fig. 2. The dye packed with glass beads was loaded into was placed down-stream of the two-step trap to measure the
the extraction cell. Carbon dioxide was delivered from the flow rate of carbon dioxide. The solubility of the dye can be
gas cylinder and flowed through the extraction cell at a rate calculated from the dye concentration and volume of car-
of 2.0 cm3 min−1 on the basis of liquid carbon dioxide at bon dioxide. After every experimental run, the whole line
the circulation pump. The flow rate used in this work was of the apparatus was rinsed with acetone and ethanol, re-
determined from preliminary experiments carried out at dif- spectively. The solvent remaining in the line was completely
ferent flow rates of 1.0, 1.5, 2.0, 2.5 and 3.0 cm3 min−1 removed by flowing carbon dioxide through the line heated
and the solubilities measured were independent of the flow at 373.15 K with the oven and flexible heater.
rate between 1.0 and 3.0 cm3 min−1 . The temperature of the

3. Results and discussion

O NHCH(CH3)2
HOC N
3.1. Experimental results
N=N C N
C At pressure ranges from 10.0 to 25.0 MPa, the experi-
O NHCH(CH3)2 mental solubilities of C.I. Disperse Blue 134 and C.I. Dis-
CH 3
perse Yellow 16 in SC-CO2 at the temperatures of 323.15,
(a) (b) 353.15 and 383.15 K are given in Table 1. In the ternary
Fig. 1. Chemical structures of (a) C.I. Disperse Blue 134 (1,4-bis(isopro- measurement, the solid dye blends prepared by equal molar
pylamino)anthraquinone); and (b) C.I. Disperse Yellow 16 (3-methyl- quantity of C.I. Disperse Blue 134 and C.I. Disperse Yellow
1-phenyl-5-pyrazolone). 16, mixed vigorously in a mortar with a pestle, were loaded
 K. Tamura, T. Shinoda / Fluid Phase Equilibria 219 (2004) 25–32 27

Table 1
Binary solubility data of dyes (2) in SC-CO2 (1)
T = 323.15 K T = 353.15 K T = 383.15 K

P (MPa) 105 y 2 P (MPa) 105 y 2 P (MPa) 105 y2

C.I. Disperse Blue 134

10.0 0.1171 10.0 00.03000 10.0 00.05578
12.0 1.111 12.0 00.1445 12.0 00.1741
15.0 2.956 15.0 00.9805 15.0 00.7372
20.0 5.652 20.0 6.771 20.0 4.567
25.0 8.744 25.0 13.08 25.0 15.36
C.I. Disperse Yellow 16
10.0 00.2246 10.0 0.09245 10.0 0.2089
12.0 02.111 12.0 00.3317 12.0 00.5423 Fig. 3. Solubility of C.I. Disperse Blue 134 in SC-CO2 as a function of
15.0 5.453 15.0 02.090 15.0 01.775 pressure. Experimental: (䊊) 323.15 K; () 353.15 K; (䊐) 383.15 K; (—)
20.0 0009.613 20.0 9.988 20.0 9.616 calculated by modified PRSV EOS.
25.0 13.65 25.0 22.03 25.0 26.94

in an extraction column. Table 2 gives the solubilities of

the equimolar dye compounds in SC-CO2 at 10.0–25.0 MPa
and 383.15 K. The results were obtained from an arithmetic
mean value of two or three measurements. The reproducibil-
ity of the solubilities of the dyes was within ±4% in mole
fraction. The solubilities of the dyes in SC-CO2 were deter-
mined from a calibration curve established between known
amounts of the dyes diluted in ethanol and UV-peak heights
whose absorbance was observed at 640 nm in wavelength
for C.I. Disperse Blue 134 and at 393 nm for C.I. Disperse
Yellow 16. The melting points of these dyes are 428 K [14]
Fig. 4. of C.I. Disperse Yellow 16 in SC-CO2 as a function of pressure.
for the C.I. Disperse Blue 134 and (452–453) K [15] for
Experimental: (䊊) 323.15 K; () 353.15 K; (䊐) 383.15 K; (—) calculated
C.I. Disperse Yellow 16, which are higher than 383.15 K of by modified PRSV EOS.
the experimental conditions. The solid dyes were not melted
during the experiments and also we did not find the dyes
melted in the remains after the solubilities measurements. the temperature. In contrast, at lower pressures, the former
The experimental solubility data of C.I. Disperse Blue 134 effect is controlled by the latter because of the densities
and C.I. Disperse Yellow 16 in SC-CO2 are shown in Figs. 3 in this range greatly depend on the temperature. Figs. 5
and 4 as a function of pressure. The solubility curves of and 6 depict the solubilities of C.I. Disperse Blue 134 and
the dyes illustrated in Figs. 3 and 4 intersect at each other C.I. Disperse Yellow 16 in SC-CO2 as a function of the
around 20–25 MPa. This can be explained by the two fol- density of SC-CO2 . At a constant density the solubility of
lowing effects. First, at constant pressure, the solubilities of dye increases with the temperature. Comparing the binary
dye decreases as the density of carbon dioxide decreases and ternary experimental results at 383.15 K, the solubility
with increasing temperature. Second, at constant pressure,
the solubility of dye increases with increasing temperature
because of a rising diffusion power of the solute. At higher
pressures, the former effect is dominant over the latter one
because of the densities in this range slightly depend on

Table 2
Ternary solubility data of C.I. Disperse Blue 134 (2) and C.I. Disperse
Yellow 16 (3) in SC-CO2 (1) at T = 383.15 K
P (MPa) 105 y2 105 y3
10.0 00.05669 00.1832
12.0 00.1764 00.4834
15.0 00.6618 01.495
Fig. 5. Solubility of C.I. Disperse Blue 134 in terms of density of carbon
20.0 05.091 08.652
dioxide. Experimental: (䊊) 323.15 K; () 353.15 K; (䊐) 383.15 K; (—)
25.0 16.46 26.49
calculated by Eq. (1).
28 K. Tamura, T. Shinoda / Fluid Phase Equilibria 219 (2004) 25–32

4. Calculated results

Assuming that no supercritical fluid dissolves in solid

phase, the solid is incompressible and its vapor pressure is
very low, the solubilities of the solid in supercritical fluid
can be expressed as
Pi sat exp[Vi S (P − Pi sat )/RT]
yi = (2)
φi SCF P
where ViS is the molar volume of dye, Pisat the saturated
vapor pressure of dye and φiSCF the fugacity coefficient of
Fig. 6. Solubility of C.I. Disperse Yellow 16 in terms of density of carbon dye in supercritical phase calculated from the thermody-
dioxide. Experimental: (䊊) 323.15 K; () 353.15 K; (䊐) 383.15 K; (—) namic relation [17] using an equation of state (EOS). From
calculated by Eq. (1). a practical point of view, we used the Peng–Robinson [12],
Peng–Robinson–Stryjek–Vera [13], and a modified form of
PRSV EOS [11] for the representation of the binary and
of C.I. Disperse Blue 134 in the ternary system is close ternary solubilities of disperse dyes in supercritical condi-
to the solubility in the binary system. However, the sol- tions. The modified form of PRSV EOS proposed in our pre-
ubility of C.I. Disperse Yellow 16 in the ternary system vious work takes into account the temperature dependence
was lower (average 10.4%) than the solubility in the binary of a pure characteristic parameter κ1 in the PRSV EOS. The
system. expressions for these EOSs are described fully in the ap-
pendix. The critical constants and acentric factor ω of pure
components used in the EOSs were estimated by following
3.2. Empirical correlation of solubility data group contribution methods. The critical properties were es-
timated by the Miller method [18], and the acentric factor
The solubilities of disperse dyes in SC-CO2 were cor- and vapor pressure of dye were evaluated using the Lee-
related using an empirical equation proposed by Sung and Kesler method [17]. The estimated values for C.I. Disperse
Shim [6]. Blue 134 and C.I. Disperse Yellow 16 are summarized in

 Table 4. The conventional van der Waals one-fluid mixing
B D rule used for the cross-interaction parameters aij and bij of
ln y2 = A + + C+ ln ρ (1)
T T the attraction and repulsion term in the binary and ternary
mixtures was expressed as follows.
y2 is the mole fraction of dye and ρ the density of SC-CO2 aij = (aii ajj )0.5 (1 − kij ) (3)
calculated from an equation of state which can describe the
P–V–T behavior accurately over the temperature and pres- bij = 21 (bii + bjj ) (4)
sure ranges [16]. The values of A, B, C and D were obtained
by fitting the Eq. (1) to the experimental results. Table 3 where aii and bii are the parameters for pure component i and
lists the constants in Eq. (1) together with the absolute arith- kij is the binary interaction parameter between components
 exp i and j.
metic mean deviations of AAD (%) = 1/ND ND i=1 |{y 2,i −
exp
cal }/y
y2,i 2,i | × 100 between the experimental and calculated 4.1. Binary correlation
solubilities of C.I. Disperse Blue 134 and C.I. Disperse Yel-
low 16 in SC-CO2 . Figs. 5 and 6 compare the experimen- The binary solubilities of dye in SC-CO2 were repro-
tal solubilities of the dyes with the values calculated from duced by two kinds of calculation procedures. One is that
Eq. (1). the solubilities of dye are correlated using Eq. (2), wherein

Table 3
Parameters in Eq. (1)
System Parameter AAD (%)a

A B (K−1 ) C D (K−1 )

CO2 + C.I. Disperse Blue 134 −5.6359 −13276 1.9381 1166.3 5.9
CO2 + C.I. Disperse Yellow 16 12.518 −18542 −0.85373 2005.0 4.4

  
ND  exp
 
cal 
a AAD (%) = 1  {y2,i − y2,i } 
  × 100 .
ND  exp
y2,i 
i=1
 K. Tamura, T. Shinoda / Fluid Phase Equilibria 219 (2004) 25–32 29

Table 4
Characteristic parameters for pure dyes
C.I. Disperse C.I. Disperse
Blue 134 Yellow 16
Tc (K)a 940.6 917.1
Pc (Pa)a 1.862 2.356
ωa 1.2658 0.9046
Ps (Pa)
323.15 Kb 8.526 × 10−8 1.094 × 10−4
353.15 K 1.026 × 10−6 4.403 × 10−3
383.15 K 5.162 × 10−4 9.089 × 10−2
vsc (×106 m3 mol−1 ) 276.9 218.9
a Estimated by Miller method [18].
b Estimated by Lee-Kesler method [17].
c Estimated by Yamamoto et al. method [19].

the fugacity coefficient was calculated by the PR and PRSV

EOSs, and the vapor pressure and molar volume of dye,
given in Table 4, were estimated by the group additive meth-
ods [18,19]. The binary interaction parameter kij in the PR
and PRSV EOSs was correlated from the experimental re-
sults by minimizing the following objective function
 exp cal 2
1  yi,j − yi,j
ND
Q= exp (5) Fig. 8. Calculation procedure for vapor pressure and molar volume of
ND
j=1
y i,j dye obtained simultaneously in binary solubility correlation.

where ND is the number of experimental data and yi,j is the between dye and SC-CO2 phases during the minimization
binary solubility of dye in SC-CO2 . Fig. 7 shows a com- process of the deviations in solubility between the experi-
parison of the binary solubilities of C.I. Disperse Blue 134 mental and calculated ones. Fig. 8 illustrates the calculation
in SC-CO2 at 383.15 K. Whenever the solubilities of the procedure for the vapor pressure, molar volume and fugac-
disperse dye were correlated according to above-mentioned ity coefficient of dye determined simultaneously from the
procedure, the PR and PRSV EOSs could not represent the solubilities of dye using the EOSs. Using the PR EOS in
solubilities of the dye satisfactorily. the binary correlation of the solubilities of dye by means of
In contrast, another method for calculation is to corre- Eq. (5), we determined only the binary interaction param-
late the binary solubilities of dye using the fugacity coeffi- eter kij as temperature-dependent fitting the model to the
cient, vapor pressure and molar volume of dye determined experimental solubilities. In the PRSV EOS, the binary in-
by solving the equivalent in chemical potentials of the dye teraction parameter kij as well as the adjustable parameter
κ1 of the PRSV EOS were fitted to the experimental solu-
bilities. Furthermore the modified PRSV EOS assumes the
temperature-dependent parameter κ1 for accurate represen-
tation of the binary solubilities. The temperature-dependent
parameter κ1 in the modified PRSV EOS was obtained fit-
ting to the experimental solubilities of dye in SC-CO2 at
every temperature. Fig. 7 compares the experimental and
calculated results for the solubilities of C.I. Disperse Blue
134 in SC-CO2 at 383.15 K. From the correlated results,
the modified PRSV EOS can represent the solubilities of
the dye in SC-CO2 more accurately than the PR and PRSV
EOSs. We obtained a considerable improvement on the bi-
nary solubility correlation in calculating simultaneously the
Fig. 7. Solubility of C.I. Disperse Blue 134 in SC-CO2 at 383.15 K: vapor pressure and molar volume of dye in addition to the
(—) m-PRSV EOS, using P2sat , V2S and φ2SCF calculated by modified fugacity coefficient of dye in SC-CO2 . However, we could
PRSV EOS simultaneously; (- - -) PRSV EOS, using P2sat , V2S and φ2SCF
not compare the vapor pressure of the dye calculated with
calculated by PRSV EOS simultaneously; (- · - · -) PR EOS, using P2sat , V2S
and φ2SCF calculated by PR EOS simultaneously; (– – –) PRSV GC, using the experimental results exactly because no experimental va-
P2sat and V2S estimated by group contribution method; (- · · - · · -) PR GC, por pressures were available for the dyes used in this work.
using P2sat and V2S estimated by group contribution method. Table 5 summarizes the binary correlation results obtained
30 K. Tamura, T. Shinoda / Fluid Phase Equilibria 219 (2004) 25–32

Table 5
Correlation of solubility of dyes (2) in SC-CO2 (1) with PR EOS, PRSV EOS and modified PRSV EOS
System PR EOS PRSV EOS Modified PRSV EOS

T (K) AAD (%)a k12 κ1 AAD (%)a k12 κ1 b AAD (%)a

CO2 + C.I. Disperse Blue 134 323.15 48.5 0.1687 0.8270 10.4 0.1784 0.9114 6.0
353.15 60.9 0.1576 0.8270 7.4 0.1576 0.8268 7.4
383.15 48.1 0.1506 0.8270 12.5 0.1313 0.7423 5.4
CO2 + C.I. Disperse Yellow 16 323.15 27.6 0.1639 −0.1991 17.7 0.1786 −0.3145 7.7
353.15 15.9 0.1546 −0.1991 5.2 0.1564 −0.1989 5.2
383.15 5.7 0.1519 −0.1991 14.8 0.1226 −0.0834 5.2


  
ND  exp
 
cal 
1  {y2,i − y2,i } 
a AAD (%) =   × 100 .
ND  exp
y2,i 
i=1
b κ1 was obtained using Eq. (6) with the constants given in Table 6.

temperature.
kij = A0 + A1 T (7)

where A0 and A1 are the constants given in Table 6.

4.2. Ternary prediction

The ternary solubilities of C.I. Disperse Blue 134 and

C.I. Disperse Yellow 16 in SC-CO2 were predicted using
the modified PRSV EOS with the binary interaction param-
eter kij obtained from the constituent binary systems. The
Fig. 9. PRSV parameters of dyes in modified PRSV EOS plotted against parameter k23 between C.I. Disperse Blue 134 and C.I. Dis-
reduced temperature. Experimental: (䊊) C.I. Disperse Blue 134; () C.I. perse Yellow 16 was set to be zero. The absolute arithmetic
Disperse Yellow 16; (—) calculated by Eq. (6). mean deviations in mole fraction between the experimental
and calculated results were AAD = 5.7% for C.I. Disperse
Blue 134 and AAD = 9.9% for C.I. Disperse Yellow 16.
by using the vapor pressure, molar volume and fugacity co- From the calculated results, the molecular interaction be-
efficient of C.I. Disperse Blue 134 and C.I. Disperse Yellow tween two dyes in SC-CO2 could be expressed by combining
16. Figs. 3 and 4 show that good agreement between the ex- only pure component parameters of the dyes with the binary
perimental and correlated solubilities of the dyes in SC-CO2 interaction parameter obtained in the modified PRSV EOS.
was obtained by the modified PRSV EOS. As shown in Fig. 11, the modified PRSV EOS reproduced
Fig. 9 shows the temperature dependence of the parameter the experimental solubilities of the ternary CO2 –C.I. Dis-
κ1 in the modified PRSV EOS obtained from Table 5. The perse Blue 134–C.I. Disperse Yellow 16 system with good
temperature-dependent parameter κ1 was regressed by the accuracy.
reduced temperature, Tr = T/TC
κ1 = κ2 + κ3 Tr (6)

The constants κ2 and κ3 for the parameter κ1 in the mod-

ified PRSV EOS are given in Table 6. Moreover, the bi-
nary interaction parameter kij in the modified PRSV EOS,
shown in Fig. 10, can be expressed by a linear function of

Table 6
Coefficients in Eqs. (6) and (7) for modified PRSV EOS
Compound κ2 κ3 (×103 K−1 ) A0 A1 (×104 K−1 )

C.I. Disperse −1.822 2.819 0.4338 −7.879

Blue 134 Fig. 10. Binary interaction parameters of dyes plotted against temperature.
C.I. Disperse −1.559 3.851 0.4813 −9.327 Experimental: (䊊) C.I. Disperse Blue 134; () C.I. Disperse Yellow 16;
Yellow 16 (—) calculated by Eq. (7).
 K. Tamura, T. Shinoda / Fluid Phase Equilibria 219 (2004) 25–32 31

Subscripts
C critical
i, j components i and j
2, 3 solid solute of second and third components

Superscripts
cal calculated
exp experimental
sat saturated
S solid
SCF supercritical fluid
Fig. 11. Solubility of C.I. Disperse Blue 134 (2) and C.I. Disperse Yellow
16 (3) in SC-CO2 (1) at 383.15 K. Experimental: (䊉) C.I. Disperse Blue
134; (䉱) C.I. Disperse Yellow 16; (—) predicted by modified PRSV EOS Acknowledgements
with parameters determined by Eqs. (6) and (7) (k23 = 0.0).
We thank Arimoto Chemical Co. Ltd. for providing the
5. Conclusions C.I. Disperse Blue 134 and C.I. Disperse Yellow 16.

Solubilities of disperse dyes, C.I. Disperse Blue 134,

C.I. Disperse Yellow 16 and their dye blend in SC-CO2 Appendix A
were measured at the temperatures of 323.15, 353.15
and 383.15 K for the binary systems and at 383.15 K for The PR EOS is expressed by
the ternary system over the pressure ranges from 10.0 to RT a
25.0 MPa. The binary experimental results were correlated P= − (A1)
v − b v(v + b) + b(v − b)
accurately in terms of the density of carbon dioxide us- 
ing an empirical equation of Sung and Shim. The binary 0.457235R2 TC2
solubilities of the disperse dyes in SC-CO2 were corre- a= α(T, ω) (A2)
PC
lated with good accuracy using vapor pressures and molar
volumes of the dyes as well as fugacity coefficient of the 
0.5  2
dyes in SC-CO2 that were calculated simultaneously by T
α(T, ω) = 1 + κ 1 − (A3)
a modified form of PRSV EOS. Furthermore, the ternary TC
solubilities of the solid dye compounds in SC-CO2 were
0.077796RTC
predicted satisfactorily using the modified PRSV EOS with b= (A4)
only binary interaction parameters kij between the dyes and PC
SC-CO2 . Stryjek and Vera proposed the PRSV EOS by assuming that
the coefficient κ in Eq. (A3) can be given by Eq. (A5) in
List of symbols terms of κ0 and κ1 in order to represent the vapor pressures
kij binary interaction parameter more accurately at lower reduced temperatures.
A, B, C, D constants in Eq. (1) 
0.5 
A0 , A1 constants in Eq. (7) T T
κ = κ0 + κ1 1 + 0.7 − (A5)
R gas constant TC TC
V molar volume
y mole fraction in supercritical phase where
ND number of data
Q objective function κ0 = 0.37889 + 1.4897ω − 0.17132ω2 + 0.019655ω3
a, b parameters of EOS (A6)
P pressure
Psat saturated vapor pressure and κ1 is an adjustable parameter for pure component.
T temperature

Greek letters References

ω acentric factor
κ, κ0 , κ1 characteristic constants in Eq. (A5) [1] D. Knittel, W. Saus, E. Schollmeyer, J. Textile Inst. 84 (1993) 534–
552.
κ2 , κ3 characteristic constants in Eq. (6) [2] W. Saus, D. Knittel, E. Schollmeyer, Text. Res. J. 63 (1993) 135–142.
ρ density of pure CO2 [3] A.S. Özcan, A.A. Clifford, K.D. Bartle, D.M. Lewis, J. Chem. Eng.
φ fugacity coefficient in supercritical phase Data 42 (1997) 590–592.
32 K. Tamura, T. Shinoda / Fluid Phase Equilibria 219 (2004) 25–32

[4] S.N. Joung, H.Y. Shin, Y.H. Park, K.-P. Yoo, J. Korean, Chem. Eng. [12] D.-Y. Peng, D.B. Robinson, Ind. Eng. Chem. Fundam. 15 (1976)
15 (1998) 78–84. 59–64.
[5] B. Guzel, A. Akgerman, J. Chem. Eng. Data 44 (1999) 83–85. [13] R. Stryjek, J.H. Vera, J.H. Can, J. Chem. Eng. 64 (1986) 323–333.
[6] H.-D. Sung, J.-J. Shim, J. Chem. Eng. Data 44 (1999) 985– [14] Colour Index, Revised third ed., vol. 7, The Society of Dyers and
989. Colourists, 1975.
[7] D. Tuma, G.M. Schneider, Fluid Phase Equilib. 158–160 (1999) [15] M.S. Simon, J. Am. Chem. Soc. 85 (1963) 1974–1977.
743–757. [16] R. Span, W. Wagner, J. Phys. Chem. Ref. Data 25 (1996) 1509–1596.
[8] J.W. Lee, M.W. Park, H.K. Bae, Fluid Phase Equilib. 173 (2000) [17] R.C. Reid, J.M. Prausnitz, T.K. Sherwood, The Properties of Gases
277–284. & Liquids, fourth ed., McGraw Hill, New York, 1986.
[9] S. Maeda, K. Mishima, K. Matsuyama, M. Baba, T. Hirabaru, H. [18] W.J. Lyman, W.F. Reehl, D.H. Rosenblatt, Handbook of Chemical
Ishikawa, K. Hayashi, J. Chem. Eng. Data 46 (2001) 647–650. Property Estimation Methods, American Chemical Society, Wash-
[10] K. Mishima, K. Matsuyama, H. Ishikawa, K. Hayashi, S. Maeda, ington, DC, 1990.
Fluid Phase Equilib. 194–197 (2002) 895–904. [19] H. Yamamoto, F. Kanegae, K. Mishima, Y. Iwai, Y. Arai, Mem. Fac.
[11] T. Shinoda, K. Tamura, J. Chem. Eng. Data 48 (2003) 869–873. Eng. Kyushu. Univ. 47 (1987) 95–113.