Fluid Phase Equilibria 275 (2009) 27–32

Contents lists available at ScienceDirect

Fluid Phase Equilibria

journal homepage: www.elsevier.com/locate/fluid

Vapor–liquid equilibria for water + acetic acid + (N,N-dimethylformamide

or dimethyl sulfoxide) at 13.33 kPa
Yong Peng ∗ , Lijuan Ping, Shengli Lu, Jianwei Mao
Key Laboratory of Agricultural Products Chemical and Biological Processing Technology of Zhejiang Province,
College of Biological and Chemical Engineering, Zhejiang University of Science and Technology, Hangzhou 310023, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Isobaric vapor–liquid equilibrium (VLE) data of the systems acetic acid + N,N-dimethylformamide (DMF),
Received 7 March 2008 acetic acid + dimethyl sulfoxide (DMSO), DMSO + water, water + acetic acid + DMF, and water + acetic
Received in revised form acid + DMSO have been measured at 13.33 kPa by using an improved Rose equilibrium still. The association
18 September 2008
of acetic acid in vapor phase has been considered, and the nonideality of vapor phase was accounted for
Accepted 19 September 2008
using the Hayden–O’Connell (HOC) method. The experimental binary data have been correlated by the
Available online 30 September 2008
NRTL, Wilson and van Laar models. The NRTL model parameters obtained from the binary data have been
used to predict the ternary VLE data. The ternary predicted values obtained in this way agree well with
Keywords:
Acetic acid
the experimental values.
N,N-Dimethylformamide © 2008 Elsevier B.V. All rights reserved.
Dimethyl sulfoxide
Vapor–liquid equilibrium (VLE)

1. Introduction mal VLE data of the systems acetic acid + (DMF or DMSO) are
reported [9,10]. In this article, the VLE data of the systems
Acetic acid is an important compound, and is widely used in acetic acid + DMF, acetic acid + DMSO, DMSO + water, water + acetic
the petroleum and dyestuff industries, and also in fiber synthesis, acid + DMF, and water + acetic acid + DMSO were measured at
acrylic spinning, and polyester leather synthesis. The separation 13.33 kPa. NRTL models parameters of the systems water + acetic
of acetic acid from aqueous solutions is industrially important. acid, DMSO + water, DMF + water, acetic acid + DMF, and acetic
Because of the closeness of their boiling points and the deviation acid + DMSO were obtained, and based to predict the VLE data of
from ideal solution behavior, acetic acid cannot be easily removed the ternary system water + acetic acid + (DMF or DMSO).
from its mixtures with water by distillation. Extractive distilla- The nonideality in the vapor phase is accounted for by using the
tion is an attractive process for the separation of acetic acid from Hayden–O’Connell (HOC) method in combination with the chemi-
water. cal theory [11–13] to account for association between components.
Several solvents [1–4], such as dimethyl sulfoxide (DMSO), N,N- The liquid-phase nonideality is considered by the nonrandom two-
dimethylformamide (DMF), N-methyl pyrrolidone, and N-methyl liquids model (NRTL) [14], the Wilson model [15], and the van Laar
acetamide, have been used as solvents for the separation of acetic equation [16]. The NRTL model was used to predict the VLE data of
acid and water. DMSO and DMF are good solvents for the sepa- the ternary systems.
ration of acetic acid, but they have a high normal boiling point,
and DMSO easily decomposes at high temperature. Therefore,
it is necessary to study the VLE of the systems water + acetic 2. Experimental
acid + (DMF or DMSO) at reduced pressure. Many VLE data of
the systems water + acetic acid [5] or DMF [5] or DMSO [6–8]) 2.1. Materials
have been reported. Unfortunately, little work has been done on
VLE of the systems acetic acid + (DMF or DMSO), only isother- DMSO (99.7 mass%) was purchased from Wuxi Haishuo Bio-
logical Company, DMF (≥99.5 mass%) from Rugao Jinling Reagent
Company, and acetic acid (≥99.5 mass%) from Hangzhou Chem-
ical Reagent Company. All samples were purified by distillation.
∗ Corresponding author. Tel.: +86 13958097318. The water sample was distilled twice. The purity of the samples
E-mail address: [email protected] (Y. Peng). was checked by measuring boiling points and further confirmed

0378-3812/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.fluid.2008.09.010
28 Y. Peng et al. / Fluid Phase Equilibria 275 (2009) 27–32

Table 1
Normal boiling points Tb of pure components and constants of Antoine equation.

Substance Antoine constantsa [5] Tb (K)

Ai Bi Ci This work Literature [17]

Water 8.07131 1730.63 233.426 373.10 373.15

Acetic acid 8.02100 1936.010 258.451 391.00 391.035
DMSO 6.88076 1541.52 191.797 461.92 462.19
DMF 7.1085 1537.780 210.390 425.70 426.2
a
Antoine equation: log (Pi s /0.133) = Ai − Bi /(t + Ci ), Units: p (kPa), t (◦ C).

by gas chromatography. Good agreement between the measured

boiling points and literature values [17] is obtained as shown in
Table 1.

2.2. Apparatus and procedure

The experimental apparatus for measuring VLE data is the mod-

ified Rose–Williams still [18], the favorable performance of which
has been verified in the previous measurements [18–21]. The total
volume of the still was 100 cm3 . The energy was applied to the still
with the help of electrical heating [20]. A given liquid solution was
Fig. 2. VLE data of the system water (1) + acetic acid (2) at 101.33 kPa , This work;
charged into the equilibrium still and heated. The vapor was con- -䊉-, Gmehling and Onken [22,29].
densed in the condenser by the circulation of cooling water and was
delivered to the mixing chamber through the vapor-phase sampling
port. In the mixing chamber vapor and liquid were mixed again was removed from the buffer tank by the vacuum pump until the
and returned to the boiling still for recirculation. The vapor–liquid pressure is 10.00 kPa. If the pressure of the still was higher than
equilibrium was usually reached after about 60 min. Samples of 13.33 kPa, the pressure controller two opens the electromagnetic
the equilibrium phases were taken out from a vapor and liquid valve. When the pressure of the still reached 13.33 kPa, the elec-
sampling port, respectively. tromagnetic valve closes again. When the pressure of buffer tank
The temperature was measured with a calibrated precision mer- was higher than 12.00 kPa, the pressure controller one activates the
cury thermometer. The accuracy of temperature measurements vacuum pump. When the pressure of buffer tank reached 10.00 kPa,
was 0.1 K. Pressure was controlled by using an automatic con- the vacuum pump stops working instantly.
trol system. The accuracy of pressure control was within ±30 Pa.
The pressure control system mainly included one vacuum pump,
2.3. Sample analysis
one buffer tank, one electromagnetic valve, and two pressure
controllers. Fig. 1 shows the schematic diagram of the pressure
Compositions of samples of the equilibrium phases were ana-
control system. The buffer tank was connected with the modified
lyzed using a gas chromatograph SP6800-A with a dual-channel
Rose–Williams still by the electromagnetic valve. At first the air
chromatography workstation N-2000 and a calibrated thermal con-
ductivity detector. Calibration curves were obtained by producing
chromatograms for a set of solutions of known composition. The
peak area of each component was correlated with the correspond-
ing mass. The stainless steel column was 2 m long, 4 mm o.d., and
filled with Porapak Q; the optimum operational conditions were
an follows: oven temperature 463 K; injector temperature 473 K;
detector temperature 473 K; bridge current 100 mA; sample vol-
ume 1 ␮L. Hydrogen was used as carrier gas with a flow rate of
50 mL min−1 .

3. Results and discussion

3.1. Experimental data

To test the equilibria apparatus and experimental procedure,

VLE data of the system water + acetic acid was measured at
101.33 kPa as shown in Fig. 2. It is found that the experimental data
of this work agree well with the literature data [22,29].
Experimental VLE data of the binary systems DMSO + water,
acetic acid + DMSO, and acetic acid + DMF have been measured at
13.33 kPa. The results were reported in Table 2 and Figs. 3–5. The
systems acetic acid + (DMSO or DMF) show a maximum tempera-
ture azeotropic point. Also, experimental VLE data of the ternary
Fig. 1. Experimental setup: (1) vacuum pump; (2) buffer tank; (3) pressure con-
troller 1; (4) pressure controller 2; (5) Dewar with cryohydrate; (6) trap; (7) the systems water + acetic acid + DMF and water + acetic acid + DMSO
modified Rose–Williams still; (8) electromagnetic valve. were measured at 13.33 kPa, and the results are reported in Table 3.
 Y. Peng et al. / Fluid Phase Equilibria 275 (2009) 27–32 29

Table 2
Experimental VLE data and correlated results of binary subsystems at 13.33 kPa.
exp exp
No. T (K) x1 y1 1 2 T y1

Water (1) + DMSO (2)

1 331.10 0.8146 0.9975 0.9017 0.3155 −0.31 0.0001
2 335.65 0.7329 0.9913 0.8070 0.5780 −0.04 0.0024
3 338.15 0.6743 0.9832 0.7771 0.7898 −0.96 0.0053
4 344.05 0.5848 0.9756 0.6864 0.6430 −0.76 −0.0010
5 349.95 0.5200 0.9420 0.5812 0.9599 0.69 0.0156
6 361.15 0.3763 0.8797 0.4798 0.8682 1.09 0.0057
7 362.65 0.3443 0.8535 0.4804 0.9354 0.00 0.0061
8 368.35 0.2754 0.7847 0.4461 0.9518 −0.11 0.0020
9 370.55 0.2439 0.7612 0.4509 0.9151 −0.70 −0.0178
10 373.45 0.2158 0.7119 0.4295 0.9349 −0.35 −0.0136
11 377.65 0.1708 0.6282 0.4131 0.9501 −0.39 −0.0176
12 378.95 0.1612 0.5977 0.3981 0.9613 −0.02 −0.0086
13 380.45 0.1508 0.5786 0.3913 0.9334 0.46 −0.0141
14 383.15 0.1235 0.4820 0.3633 0.9933 0.43 0.0112
15 386.27 0.0939 0.4080 0.3645 0.9661 0.49 −0.0048
16 388.15 0.0880 0.3818 0.3422 0.9301 1.75 0.0017

Acetic acid (1) + DMSO (2)

1 351.75 0.8025 0.9976 0.8259 0.1391 −0.96 0.0000
2 358.45 0.7464 0.9952 0.7486 0.1482 −1.07 0.0015
3 362.30 0.7146 0.9885 0.7046 0.2545 −1.19 0.0003
4 368.48 0.6650 0.9825 0.6427 0.2370 −1.19 0.0058
5 373.05 0.6304 0.9633 0.5909 0.3527 −0.84 0.0035
6 379.65 0.5724 0.9202 0.5237 0.4692 −0.90 0.0041
7 383.25 0.5402 0.8801 0.4834 0.5437 −0.68 0.0012
8 386.43 0.5126 0.8265 0.4410 0.6285 −0.17 −0.0093
9 389.35 0.4828 0.7671 0.4037 0.6845 0.14 −0.0162
10 391.65 0.4608 0.7257 0.3775 0.6916 0.71 −0.0159
11 394.52 0.4189 0.6346 0.3396 0.7398 0.78 −0.0168
12 396.55 0.3742 0.5328 0.3068 0.7904 0.55 −0.0142
13 397.45 0.3400 0.4548 0.2855 0.8312 0.20 −0.0109
14 398.65 0.2809 0.3106 0.2367 0.9001 0.14 −0.0169
15 398.65 0.2041 0.1774 0.1951 0.9557 −0.16 −0.0091

Acetic acid (1) + DMF (2)

1 363.65 0 0 – – – –
2 364.57 0.0609 0.0161 0.2344 1.0168 −0.26 −0.0006
3 366.25 0.1385 0.0520 0.2847 1.0044 −0.34 −0.0021
4 367.93 0.2149 0.1069 0.3229 0.9818 −0.39 −0.0094
5 369.85 0.3177 0.2449 0.4061 0.9152 −0.28 −0.0187
6 370.65 0.3832 0.3841 0.4718 0.8337 0.04 −0.0159
7 370.30 0.4277 0.4898 0.5194 0.7825 0.11 −0.0129
8 367.55 0.5133 0.6897 0.6186 0.6801 0.01 −0.0050
9 363.30 0.5851 0.8114 0.6979 0.6185 −0.23 −0.0070
10 358.35 0.6449 0.8940 0.7844 0.5317 −0.67 −0.0005
11 354.00 0.6960 0.9458 0.8551 0.4016 −0.91 0.0061
12 350.25 0.7411 0.9669 0.9009 0.3490 −0.99 0.0037
13 347.47 0.7760 0.9804 0.9347 0.2758 −1.00 0.0040
14 344.65 0.8161 0.9867 0.9593 0.2629 −0.86 0.0013
15 336.55 1 1 – – – –
exp
Note: T = Texp − Tcal ; y1 = y1 − y1cal .

The experimental VLE data of the system acetic acid + DMF, theory [11,12]. The liquid-phase activity coefficient is calculated by
acetic acid + DMSO and DMSO + water were found to be thermo- the van Laar, Wilson and NRTL models.
dynamically consistent by the point-to-point method of Van Ness The general equilibrium relationship for VLE can be expressed
et al. [23], as modified by Fredenslund et al. [13]. The average as
absolute deviation using NRTL mode between yexp and ycal were  
ViL (p − psi )
0.0067, 0.0084 and 0.0088, respectively, which is less than 0.01. ˆ V yi p = xi i s ps exp
 (1)
i i i RT
So, all binary VLE data of this work were thermodynamically
consistent.
The Poynting correction, the exponential term in Eq. (1), is close
to unity at low or moderate pressure, and Eq. (1) can be simplified
3.2. Vapor–liquid equilibria model to
ˆ V yi p = xi i s ps
 (2)
i i i
In systems containing carboxylic acids, because two acid
molecules tend to form a pair of stable hydrogen bonds, large nega- where xi is the mole fraction of component i in liquid phase; yi is
tive deviations from vapor ideality occur even at very low pressures. the apparent mole fraction of component i in vapor phase; p is the
In the present work, the fugacity coefficients of water and solvents total pressure in the system; ␥i the activity coefficient of component
are calculated by using the virial equation. The vapor phase fugac- i in the liquid phase; psi is the staturation vapor pressure of pure
ity coefficients of acetic acid are computed by using the chemical liquid i at system temperature T, which can be obtained from the
30 Y. Peng et al. / Fluid Phase Equilibria 275 (2009) 27–32

Table 3
Experimental VLE data and correlated results of the ternary systems at 13.33 kPa.

No. T (K) x1 x2 y1 y2 T y1 y2

Water (1) + acetic acid (2) + DMSO (3)

1 357.62 0.2788 0.3164 0.8069 0.1431 −1.22 0.0000 −0.0089
2 363.70 0.2376 0.3193 0.7683 0.1523 0.21 0.0022 −0.0198
3 362.15 0.2238 0.3215 0.7492 0.1712 −2.90 0.0005 −0.0083
4 365.10 0.2136 0.3237 0.7426 0.1810 −1.10 0.0083 −0.0072
5 368.55 0.2038 0.3213 0.7269 0.1966 1.01 0.0050 0.0057
6 369.40 0.1968 0.3237 0.7120 0.2085 1.09 0.0018 0.0102
7 372.35 0.1756 0.3241 0.6796 0.2185 1.32 0.0048 0.0066
8 371.80 0.1702 0.3248 0.6512 0.2272 0.10 −0.0133 0.0104
9 373.21 0.1454 0.3282 0.5886 0.2298 −1.69 −0.0231 −0.0121
10 373.80 0.1492 0.3248 0.5990 0.2430 −0.72 −0.0227 0.0102
11 373.90 0.1660 0.3237 0.6330 0.2088 1.59 −0.0240 −0.0103
12 383.32 0.1000 0.2641 0.4744 0.1813 0.00 −0.0240 −0.0090
13 385.08 0.0880 0.2612 0.4361 0.1869 0.00 −0.0202 −0.0097
14 386.52 0.0926 0.2606 0.4574 0.1824 2.08 −0.0155 −0.0095
15 387.84 0.0690 0.2616 0.3612 0.2199 0.01 −0.0204 0.0053
16 388.20 0.0556 0.2627 0.3123 0.2184 −1.62 −0.0100 −0.0096
17 391.92 0.0432 0.2624 0.26187 0.2090 0.20 0.0001 −0.0314
Ave. abs. – – – – – 0.99 0.0115 0.0108
Max. abs. – – – – – 2.90 0.0240 0.0314

Water (1) + acetic acid (2) + DMF (3)

1 330.00 0.5275 0.1807 0.9311 0.0443 −2.22 0.0312 0.0052
2 331.23 0.4956 0.1948 0.9212 0.0501 −1.76 0.0311 0.0063
3 333.08 0.4659 0.2075 0.9057 0.0556 −0.68 0.0256 0.0073
4 333.69 0.4463 0.2151 0.8933 0.0592 −0.61 0.0201 0.0082
5 334.80 0.4132 0.2283 0.8749 0.0647 −0.47 0.0145 0.0083
6 335.34 0.3910 0.2358 0.8639 0.0692 −0.63 0.0128 0.0097
7 335.74 0.3702 0.2455 0.8497 0.0731 −0.91 0.0083 0.0088
8 336.16 0.3612 0.2516 0.8419 0.0759 −0.79 0.0052 0.0081
9 337.26 0.3355 0.2575 0.8280 0.0808 −0.65 0.0041 0.0102
10 338.31 0.3125 0.2598 0.8145 0.0884 −0.54 0.0031 0.0168
11 339.21 0.2933 0.2678 0.8021 0.1095 −0.45 0.0029 0.0327
12 339.61 0.2784 0.2742 0.7903 0.0923 −0.71 0.0014 0.0110
13 342.41 0.2678 0.2744 0.7762 0.0894 1.57 −0.0054 0.0078
14 342.48 0.2479 0.2821 0.7555 0.0823 0.65 −0.0103 −0.0055
15 344.28 0.2335 0.2907 0.7611 0.1015 1.71 0.0082 0.0065
Ave. abs. – – – – – 0.96 0.0123 0.0102
Max. abs. – – – – – 2.22 0.0312 0.0327
exp exp
Note: T = Texp − Tcal ; y1 = y1 − y1cal ; y2 = y2 − y2cal .

Antoine equation. The Antoine constants taken from the literature 3.3. Calculation of binary vapor–liquid equilibria
are listed in Table 1. is is the fugacity coefficient of saturated pure
i at system temperature T.  ˆ V , the apparent fugacity coefficient in To correlate the VLE data of the system water + acetic
i
vapor phase, can be calculated by Eq. (3) [11]: acid + solvents it is necessary to determine the binary adjustable
model parameters. For each binary system, the three correspond-
ˆ V yi = zi #
 (3) ing parameters were estimated by the minimization of the objective
i i
function F [26]:
In Eq. (3), i# , the true fugacity coefficient of component i monomer,
is expressed by Eq. (4) [11]: 
N
exp 2 exp 2
F= [(1 − 1cal ) + (2 − 2cal ) ]i (7)
BiF p i=1
ln i# = (4)
RT

where BiF , the “free” contribution to the second virial coefficient of where  is the activity coefficient and N denotes the experimental
component i, can be obtained from the HOC equation [24]. point number.
In Eq. (3), zi , the true vapor-phase mole fraction of component i The results for the system water + DMSO at 13.33 kPa are shown
monomer, is expressed by Eq. (5) [11,25]: in Fig. 3. The experimental data of this work agree well with
the literature data [27]. For the systems acetic acid + (DMSO or
 DMF), the average deviation in temperature and vapor mole frac-
1 + 4Kp exp(BiF /RT )yi (2 − yi ) − 1
Zi = (5) tion are listed in Tables 2 and 4 and plotted in Figs. 4 and 5.
2Kp exp(BiF /RT )(2 − yi ) It can be found from these tables and figures that the NRTL
model correlated values agree well with the experimental VLE
K, the associated equilibrium constant of acetic acid, is found from data.
the relation [13]: The VLE data of the system water + DMF was cited in Ref. [27],
 101.33 × K  the system acetic acid + water at 13.33 kPa cited in Refs. [22,28], and
3166 NRTL model parameters of these systems were correlated using the
−lg = 10.4205 − (6)
760 T literature data.
 Y. Peng et al. / Fluid Phase Equilibria 275 (2009) 27–32 31

Table 4
The parameters and correlation deviations of NRTL Wilson or van Laar models.

Systems A12 a (J mol−1 ) A21 (J mol−1 ) ˛ Dy1 b dy1 c DTd (K) dTe (K)

Acetic acid(1) + DMSO (2)

NRTL −890.6 −6105.2 0.3 0.0084 0.0169 0.65 1.19
Wilson 3782.8 2943.9 0.0117 0.0241 0.67 1.45
van Laar −2.1897 −3.7342 0.0090 0.0218 0.60 1.28

Acetic acid(1) + DMF (2)

NRTL −298.6 −4442.4 0.2 0.0067 0.0187 0.47 1.00
Wilson 1992.7 2475.2 0.0115 0.0327 1.06 2.28
van Laar 1.0020 −0.2212 0.0076 0.0208 0.98 1.84

Water (1) + acetic acid (2)

NRTL 4014.5 2872.1 1.3 0.0038 0.0128 1.58 2.49
NRTLf [22,28] 4105.2 4457.6 1.1566 0.0039 0.0119 1.61 2.53
Wilsonf [22,28] 2069.3 2844.8 0.0191 0.0818 1.55 2.53
van Laarg [22,28] 0.4419 1.7238 0.0221 0.0888 1.53 2.53

Water (1) + DMF(2)

NRTL 872.3 425.9 0.3 0.0086 0.0200 0.44 1.11

Water (1) + DMSO(2)

NRTL 1466.9 −4881.0 0.3 0.0088 0.0217 0.51 1.75
a
Aij = gij − gjj .

N
exp
b
Dy1 = (1/N) |y1 − y1cal |i .
i=1
exp
c
dy1 = max(|y1 − y1cal |).

N

d
DT = (1/N) |T exp − T cal |i .
i=1
e
dT = max(|T − T cal |).
exp

f
Literature parameters transferred to SI units.
g
The van Laar parameters are dimensionless.

3.4. Predication of ternary vapor–liquid equilibria and 0.0108, and the average absolute deviation of the equilibrium
temperature was 0.99 K. For the system water + acetic acid + DMF,
In this work, the NRTL model was used in combination with the the average absolute deviation of the vapor-phase mole fractions of
chemical theory for predicting the VLE data of the ternary systems. water and acetic acid were 0.0123 and 0.0102, and the average abso-
The VLE data of the ternary system and the deviation in tempera- lute deviation of the equilibrium temperature was 0.96 K. These
tures and vapor mole fraction are listed in Table 3. For the system values show that the NRTL model gives a good representation of
water + acetic acid + DMSO, the average absolute deviation of the the experimental data.
vapor-phase mole fractions of water and acetic acid were 0.0115

Fig. 4. Experimental and calculated x–y diagram of the system acetic acid
Fig. 3. Experimental VLE data and regression results using the NRTL model of the (1) + DMSO (2) at 13.33 kPa , experimental data; —, calculation based on the NRTL
system water (1) + DMSO (2) at 13.33 kPa , experimental data; —, calculation based model; - - - -, calculation based on the Wilson model; , calculation based on the
on the NRTL model; , literature value [27]. van Laar model.
32 Y. Peng et al. / Fluid Phase Equilibria 275 (2009) 27–32

Greek symbols
˛ adjustable parameter in NRTL model
i activity coefficient of component i in the liquid phase
ˆ #, 
 ˆ V, 
ˆ s true fugacity coefficient of component i; apparent
i i i
fugacity coefficient of component i; fugacity coefficient
of saturated pure i

Superscripts
cal calculated results
exp experimental results
L, V in the liquid phase; in the vapor phase
s saturated

Subscripts
i component

Acknowledgment

Thanks are due to Yangfang Li for her helpful assistance in the

computation work. This research was supported by the National
Fig. 5. Experimental and calculated x–y diagram of the system acetic acid (1) + DMF Natural Science Foundation of China (NO. 20804038).
(2) at 13.33 kPa , experimental data; —, calculation based on the NRTL model; - - - -,
calculation based on the Wilson model;  , calculation based on the van Laar model. References

[1] W.-X. Chang, H. Wan, G.-F. Guan, H.-Q. Yao, Fluid Phase Equilib. 242 (2006)
204–209.
4. Conclusions [2] L. Berg, US Patent 5,167,774, 1992-12-1 (1992).
[3] L. Berg, R.J. Szabados, K.M. Wendt, Chem. Eng. Commun. 89 (1990) 113–131.
[4] L. Berg, US Patent 4,729,818, 1988-3-8 (1988).
The VLE data of the systems acetic acid + DMSO, DMSO + water,
[5] J. Gmehling, U. Onken, Vapor–Liquid Equilibrium Data Collection, DECHEMA
acid + DMF, water + acetic acid + DMSO and water + acetic Chemistry Data Series, Vol. 1. Part 1, DECHEMA, Frankfurt, 1977.
acid + DMF were measured at 13.33 kPa by using an improved [6] Murakami, Yamada, Kagaku Kogaku. 26 (1962) 865.
Rose equilibrium still. The thermodynamic consistency test was [7] Zh. Golubkov, Prikl. Khim. (Leningrad) 50 (1977) 257.
[8] Nishimura, Nakayama, Yano, J. Chem. Eng. Jpn. 5 (1972) 223–226.
passed for the measured binary systems. Parameters of NRTL model [9] Shealy, Sandler, J. Chem. Eng. Data 30 (1985) 455–459.
for the binary systems were obtained. The VLE data of the ternary [10] Schuberth, Quitzsch, Geiseler, Z. Phys. Chem. (Leipzig) 257 (1976) 385–391.
systems water + acetic acid + (DMSO or DMF) at 13.33 kPa were [11] J.M. Prausnitz, E. Anderson, C. Grens, Computer Calculation for Multicomponent
Vapor–Liquid and Liquid–Liquid Equilibrium, Prentice-Hall, New Jersey, 1980.
predicted by using binary NRTL parameters. Ternary predicted [12] A. Tamir, J. Wisniak, Ind. Eng. Chem. Fundam. 15 (1976) 274–280.
values agree well with the experimental data in the experimental [13] A. Fredenslund, J. Gmehling, P. Rasmussen, Vapor–Liquid Equilibra Using UNI-
range of compositions. It can be said that NRTL + HOC model is FAC, Elsevier, Amsterdam, 1977.
[14] H. Renon, J.M. Prausnitz, AIChE J. 14 (1968) 135–144.
suitable for correlating the VLE of binary systems and the NRTL [15] G.M. Wilson, J. Am. Chem. Soc. 86 (1964) 127–130.
interaction parameters are good enough to represent the two [16] J.J. Van Laar, Z. Phys. Chem. 35 (1929) 185.
ternary systems. [17] J.A. Riddick, W.B. Bunger, T.K. Sakano, Organic Solvents, 4th ed., Wiley-Inter
Science, New York, 1986.
[18] Z.-X. Chen, W.-M. Hu, L.-H. Wang, Z.-L. Wu, Fluid Phase Equilibria 70 (1991)
List of symbols 227–238.
A, B, C Antoine equation coefficients [19] Z.-X. Chen, W.-M. Hu, Fluid Phase Equilibria 154 (1999) 89–98.
[20] B. Yang, H. Wang, J. Chem. Eng. Data 47 (2002) 1324–1329.
BiF “free” contribution to the second virial coefficient of com- [21] W.-.M. Hu, L.-M. Shen, L.-J. Zhao, Fluid Phase Equilibria 219 (2004) 265–
ponent i 268.
fis component fugacity [22] J. Gmehling, U. Onken, Vapor–Liquid Equilibrium Data Collection, DECHEMA
Chemistry Data Series, Vol. 1. Part 1, DECHEMA, Frankfurt, 1977, pp. 117–
F objective function 133.
g characteristic energy (J mol−1 ) [23] H.C. Van Ness, S.M. Byer, R.E. Gibbs, AIChE J. 19 (1973) 238–244.
K associated equilibrium constant of acetic acid (kPa−1 ) [24] J.G. Hayden, J.P. O’Connell, Ind. Eng. Chem. Process Des. Dev. 14 (1975) 209–216.
[25] K.-H. Nothnagel, D.S. Abrams, J.M. Prausnitz, Ind. Eng. Chem. Proc. Des. Dev. 12
P system total pressure (kPa)
(1973) 25.
Pis staturation vapor pressure of the pure liquid i (kPa) [26] W.-X. Chang, G.-F. Guan, X.-L. Li, H.-Q. Yao, J. Chem. Eng. Data 50 (2005)
T equilibrium boiling temperature (K) 1129–1133.
[27] Engineering Chemistry Database. The Institute of Process Engineering [DB/OL]
xi mole fraction of component i in the liquid phases
[2008-03-01]. http://mole.ipe.ac.cn/nedb/ (System ID 5937 or ID 10867).
yi apparent mole fraction of component i in vapor phases [28] D.B. Keyes, Ind. Eng. Chem. 25 (1933) 569–575.
Zi true mole fraction of species i in the vapor phase [29] G. Povarnin, A. Markcv, Zh. Russ. Fiz. Khim. Obshch. 55 (1924) 381–387.