Академический Документы
Профессиональный Документы
Культура Документы
Abstract
Isobaric vaporliquid and vaporliquidliquid equilibria were measured for the waterethanolhexane system at normal atmospheric pressure.
The apparatus used for the determination of vaporliquidliquid equilibrium data was an all-glass dynamic recirculating still with an ultrasonic
homogenizer coupled to the boiling flask.
The experimental data demonstrated the existence of a ternary heterogeneous azeotrope at 329.2 K with a composition of 0.105, 0.236 and
0.658 mol fraction of water, ethanol and hexane, respectively. The experimental vaporliquid and vaporliquidliquid equilibria data were found to
be thermodynamically consistent. The correlation of these data was carried out using NRTL and UNIQUAC models and predicted with UNIFAC.
The equilibrium data of the system were compared with other hydrocarbon systems previously analyzed and commonly used as entrainers in
the dehydration of the ethanol by azeotropic distillation such as benzene and cyclohexane.
2007 Elsevier B.V. All rights reserved.
Keywords: Vaporliquid equilibria; Vaporliquidliquid equilibria; Ethanol; Hexane
1. Introduction
Heterogeneous azeotropic distillation is a widely used
technique to separate binary azeotropic mixtures into their components. For many years, the traditional process to obtain pure
ethanol from a waterethanol mixture has been heterogeneous
azeotropic distillation using benzene as entrainer [1]. In the last
few years, because of the carcinogenic effect of benzene, this
entrainer has been substituted by other compounds [2,3].
The absolute ethanol produced with this process can be used
as a carburant mixed with gasoline. Since hexane is a common compound found in gasoline and is suitable to use it as an
entrainer in the azeotropic distillation of ethanol, any remaining
amount of hexane in the anhydrous ethanol will not be a problem
for its subsequent use as fuel. Many studies have been carried
out in this field, either by experimentation or by simulation.
However, it is not possible to find a complete experimental
study in literature of the isobaric vaporliquidliquid equilibrium (VLLE) and vaporliquid equilibrium (VLE) equilibrium
Corresponding author. Tel.: +34 965 90 38 67; fax: +34 965 90 38 26.
E-mail address: vgomis@ua.es (V. Gomis).
0378-3812/$ see front matter 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.fluid.2007.04.011
67
(a) The sampling of the gaseous phase was carried out using a
UW Type, 6-port valve from Valco Instruments Co., which
injected the samples automatically into the GC.
(b) For the sampling of the liquid phase in the heterogeneous
region, a small amount of the liquid coming from the separation chamber of the instrument was diverted to a tube
using a solenoid valve. In this tube, the dispersed liquid
phases enter and separate into two layers at their bubble
point, since the tube is placed in a thermostatic bath at the
boiling point temperature of the mixture. A sample of each
layer was taken and placed in a vial with a small amount of
an internal standard.
(c) Liquid samples in the homogeneous region were withdrawn
with a syringe of the liquid coming from the separator chamber and put into a vial.
To obtain quantitative results in the analysis of the liquid
phases, an internal standard method was used. For this reason,
2-propanol (which is completely miscible with water, ethanol
and hexane) was also added to the sample vials. Furthermore, the
addition of the standard prevents phase-separation effects when
changing the temperature after the separation of the phases.
A further description of the sampling procedure is given in a
previous paper [4].
Table 1
Vaporliquidliquid equilibrium data (mole fraction) for the water (1)ethanol (2)hexane (3) ternary system at 101.3 kPa
Organic phase
BIN
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Aqueous phase
Vapor phase
Tb (K)
x1
x2
x3
x1
x2
x3
y1
y2
y3
0.002
0.001
0.002
0.002
0.003
0.003
0.004
0.005
0.006
0.007
0.008
0.010
0.012
0.015
0.016
0.022
0.029
0.037
0.047
0.055
0.061
0.000
0.004
0.015
0.021
0.028
0.039
0.049
0.060
0.065
0.080
0.087
0.101
0.113
0.133
0.142
0.172
0.203
0.235
0.272
0.297
0.316
0.998
0.995
0.983
0.976
0.969
0.958
0.947
0.935
0.930
0.913
0.905
0.889
0.875
0.852
0.842
0.806
0.768
0.728
0.680
0.648
0.622
1.000
0.965
0.893
0.869
0.791
0.754
0.732
0.704
0.674
0.616
0.555
0.524
0.489
0.431
0.379
0.353
0.324
0.296
0.248
0.234
0.219
0.000
0.035
0.106
0.131
0.208
0.244
0.266
0.293
0.322
0.376
0.433
0.461
0.492
0.541
0.575
0.593
0.613
0.624
0.620
0.609
0.604
0.000
0.000
0.001
0.001
0.001
0.002
0.002
0.003
0.004
0.008
0.012
0.014
0.019
0.029
0.046
0.053
0.063
0.079
0.132
0.157
0.178
0.214
0.159
0.139
0.137
0.128
0.124
0.123
0.122
0.120
0.117
0.114
0.112
0.110
0.107
0.104
0.103
0.101
0.100
0.096
0.096
0.095
0.000
0.076
0.150
0.162
0.185
0.195
0.197
0.202
0.207
0.214
0.221
0.225
0.229
0.235
0.239
0.242
0.246
0.248
0.253
0.254
0.256
0.786
0.765
0.711
0.702
0.688
0.681
0.680
0.676
0.673
0.670
0.665
0.663
0.661
0.658
0.657
0.656
0.653
0.653
0.652
0.650
0.649
334.44
332.57
330.80
330.54
330.05
329.92
329.82
329.77
329.71
329.56
329.51
329.48
329.45
329.42
329.26
329.28
329.48
329.50
329.53
329.55
329.56
68
Table 2
Vaporliquid equilibrium data (mole fraction) for the water (1)ethanol
(2)hexane (3) ternary system at 101.3 kPa
Liquid phase
Vapor phase
Tb (K)
x1
x2
x3
y1
y2
y3
0.317
0.139
0.234
0.135
0.408
0.253
0.138
0.390
0.246
0.263
0.460
0.428
0.344
0.156
0.154
0.123
0.263
0.267
0.291
0.035
0.030
0.029
0.035
0.073
0.272
0.103
0.190
0.019
0.020
0.016
0.013
0.013
0.013
0.014
0.012
0.009
0.013
0.013
0.018
0.094
0.110
0.094
0.072
0.055
0.054
0.049
0.048
0.145
0.122
0.679
0.844
0.756
0.831
0.579
0.724
0.803
0.594
0.693
0.670
0.528
0.557
0.609
0.729
0.559
0.624
0.718
0.713
0.708
0.964
0.936
0.880
0.848
0.676
0.628
0.574
0.653
0.697
0.799
0.828
0.665
0.592
0.524
0.456
0.371
0.309
0.247
0.750
0.815
0.658
0.678
0.635
0.700
0.783
0.774
0.676
0.734
0.666
0.727
0.004
0.016
0.010
0.034
0.014
0.023
0.059
0.016
0.061
0.066
0.013
0.015
0.046
0.115
0.287
0.253
0.019
0.020
0.001
0.001
0.033
0.090
0.117
0.251
0.100
0.323
0.157
0.284
0.181
0.156
0.323
0.395
0.463
0.530
0.616
0.682
0.740
0.237
0.167
0.248
0.212
0.272
0.228
0.162
0.172
0.275
0.218
0.189
0.151
0.215
0.105
0.159
0.089
0.209
0.135
0.078
0.189
0.102
0.100
0.216
0.175
0.110
0.077
0.103
0.098
0.080
0.154
0.220
0.036
0.025
0.020
0.025
0.054
0.102
0.079
0.090
0.017
0.012
0.012
0.012
0.013
0.014
0.015
0.015
0.017
0.017
0.008
0.012
0.062
0.070
0.068
0.051
0.037
0.037
0.039
0.033
0.080
0.068
0.683
0.669
0.636
0.563
0.529
0.485
0.438
0.494
0.340
0.297
0.504
0.435
0.282
0.313
0.253
0.257
0.279
0.536
0.752
0.957
0.686
0.484
0.437
0.313
0.250
0.277
0.270
0.355
0.403
0.408
0.351
0.338
0.333
0.327
0.324
0.319
0.313
0.375
0.419
0.301
0.300
0.298
0.324
0.377
0.371
0.332
0.351
0.289
0.316
0.103
0.227
0.205
0.347
0.261
0.380
0.484
0.317
0.558
0.604
0.280
0.389
0.608
0.610
0.644
0.645
0.641
0.310
0.029
0.008
0.289
0.495
0.539
0.634
0.648
0.644
0.640
0.628
0.586
0.580
0.637
0.649
0.653
0.657
0.661
0.665
0.670
0.617
0.570
0.637
0.631
0.633
0.625
0.586
0.593
0.629
0.616
0.632
0.616
349.66
346.18
346.66
342.64
345.91
341.71
338.04
344.36
335.28
332.55
345.88
342.20
332.71
332.66
329.51
329.71
330.34
343.86
351.10
351.20
329.73
329.73
329.73
329.71
330.23
330.25
330.63
332.36
334.33
334.64
332.08
331.72
331.50
331.39
331.29
331.25
331.26
333.16
334.59
331.09
331.21
330.75
331.84
333.91
333.95
331.79
331.90
332.03
332.22
Table 3
Antoine equation parametersa of the pure substances
Fig. 1. VLE (mol%) diagram for the water (1)ethanol (2)hexane (3) ternary
system at 101.3 kPa: () liquid phase and (+) vapor phase.
Compound
Temperature
range ( C)
Water
Ethanol
Hexane
8.07131
8.11220
6.91058
1730.630
1592.864
1189.64
233.426
226.184
226.280
+1/+100
+20/+93
30/+170
Fig. 2. VLLE (mol%) diagram for the water (1)ethanol (2)hexane (3) ternary
system at 101.3 kPa: () liquid phase and (+) vapor phase.
Table 4
Ternary azeotrope of the water (1)ethanol (2)hexane (3) system at 101.3 kPa
obtained by different authors
x1
x2
x3
Tb (K)
0.144
0.210
0.105
0.215
0.168
0.236
0.641
0.622
0.658
329.19
329.15
329.21
Kudryavtseva
Fritzweiler
Experimental
Fig. 2 represents some of the VLLE data and includes the nonisothermal binodal curve and the vapor line. This figure shows
the presence of a ternary azeotrope in the heterogeneous region.
The points of the vapor line corresponding to the liquidliquid
equilibrium lines 113 are under the liquidliquid tie line. However, the points corresponding to lines 1420 are above the tie
line. Therefore, there is a point between lines 13 and 14 where
the composition of the vapor coincides with that of a liquid heterogeneous mixture. The composition of the ternary azeotrope
determined by numerical interpolation is x1 = 0.105, x2 = 0.236,
and x3 = 0.658 mole fraction of water (1), ethanol (2), and hexane
(3), respectively, and the temperature is 329.21 K. The composition of the two liquid phases in equilibrium is x1 = 0.401,
x2 = 0.561, and x3 = 0.038 in the aqueous phase, and x1 = 0.016,
x2 = 0.138, and x3 = 0.847 in the organic-rich phase. It is worth
noting that establishing a composition of the ternary azeotrope
by numerical interpolation is difficult. For this reason, the accuracy of these compositions is less than that of the experimental
data. The experimental data obtained previously by Kudryavtseva et al. in 1973 [8] and Fritzweiler and Dietrich in 1933 [9]
(compiled by Gmehling et al. [10]) that can be seen in Table 4
differs significantly from the azeotrope composition obtained in
this work, especially that obtained by Fritzweiler. However, the
bubble temperature is the same.
69
was done. The models used were NRTL and UNIQUAC and
the correlated data were the bibliographic VLE data of the
ethanolhexane [11] and waterethanol [11] systems and the
homogeneous and heterogeneous VLE data of this work. All
the correlations and calculations have been made using the
process simulator CHEMCAD V [12]. However, the parameters obtained do not properly predict the heterogeneous region.
Consequently, an LLE data correlation has been done to try to
improve the calculations of that region. The parameters obtained
and the mean deviations between experimental and calculated
compositions are presented in Table 5.
To properly simulate an azeotropic distillation column to
dehydrate ethanol with hexane, it is important to know the size
and shape of the heterogeneous region of the system and the
slope of the liquidliquid tie lines. These characteristics have a
great influence on the compositions obtained of the two liquids
produced in the decanter after the condensation of the vapor that
leaves the top of the distillation column.
For this reason the non-isothermal binodal curve and some
tie lines calculated with UNIFAC LLE and with the parameters
obtained with NRTL and UNIQUAC are compared in Fig. 3.
The size of the heterogeneous region calculated with the different models is larger than the experimental one, especially in
the right-hand region of the diagram. The liquidliquid region
calculated with the models even includes points that are homogeneous. Furthermore, the tie lines of this region, which are those
with the greatest concentrations of ethanol, are those necessary
for the calculations of the decanter and therefore very important
to the simulation of the column.
In conclusion, in spite of the fact that the parameters obtained
appear to predict small deviations from the experimental data,
these small differences have a great influence on the simulation results. For this reason, an improvement in the calculation
method of the equilibrium compositions is necessary.
On the other hand, in order to compare the equilibrium experimental results obtained with those of other hydrocarbon systems
Fig. 3. Comparison of the non-isothermal binodal curve and some tie lines
calculated with UNIFAC LLE and with the parameters obtained with NRTL and
UNIQUAC.
70
Table 5
Parameters and mean deviations of the LLE correlation
i
Bij
Bji
Uij Ujj
Uji Uii
Water
Water
Ethanol
Ethanol
Hexane
Hexane
765.2826
2130.053
421.547
889.5103
1661.666
275.1408
0.3031
0.2
0.3827
821.2323
1057.531
199.0427
347.0423
1823.642
564.5045
Mean deviation
NRTL
UNIQUAC
D x11
D x21
D x12
D x22
0.003
0.007
0.012
0.013
0.010
0.010
0.010
0.012
Bij binary interaction parameters for the NRTL model (cal/mol). Uij Uii binary interaction parameters for UNIQUAC (cal/mol). D xij (i = component, j = phase)
mean deviations in mole fraction of water (1) and ethanol (2) in organic phase (1) and aqueous phase (2).