Subscriber access provided by Queen Mary, University of London

Article
Reversible reaction-assisted intensification process for
separating the azeotropic mixture of ethanediol and 1,2-
butanediol: Vapor-liquid equilibrium and Economic evaluation
Hong Li, Zhenyu Zhao, Jie Qin, Rui Wang, Xingang Li, and Xin Gao
Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.7b04921 • Publication Date (Web): 21 Mar 2018
Downloaded from http://pubs.acs.org on March 22, 2018

Just Accepted

“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical

Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 Reversible reaction-assisted intensification process for
5
6 2 separating the azeotropic mixture of ethanediol and
7
8
9 3 1,2-butanediol: Vapor-liquid equilibrium and Economic
10
11 4 evaluation
12
13
14 5 Hong Li, Zhenyu Zhao, Jie Qin, Rui Wang, Xingang Li, Xin Gao∗
15
16
17 6 School of Chemical Engineering and Technology, National Engineering Research Center of Distillation
18
19
7 Technology, Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin University,
20
21
22 8 Tianjin 300072, China
23
24 9 Abstract: Vapor-liquid equilibrium data of six binary systems composed of ethylene
25
26
27 10 glycol (EG), 1,2-propylene glycol (1,2-PG), 1,2-butanediol (1,2-BD) and
28
29 11 2,3-butanediol (2,3-BD) were measured and regressed by NRTL model. With the
30
31
32 12 binary interaction parameters regressed by experimental data, four separation
33
34 13 processes to obtain polymer grade EG from the synthesis products were simulated,
35
36
37 14 including conventional distillation route, azeotropic distillation route, distillation
38
39 15 coupled with liquid-liquid extraction route and reaction-assisted distillation route. The
40
41
16 total annual costs and profits of the four routes were estimated and compared, which
42
43
44 17 indicates that the annual profit was mainly determined by the yield of EG. The results
45
46 18 also demonstrate the reaction-assisted distillation route was proved to be the best
47
48
49 19 strategy currently among those separation methods.
50
51 20 Keywords: Ethylene glycol, Vapor-liquid equilibrium, Process simulation, Economic
52
53 21 evaluation, Reactive separation
54
55 ∗
Corresponding author: Xin Gao
56 E-mail: [email protected]. Tel: +86-022-27404701.
57
1
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 2 of 32

1
2
3
4 1 1. Introduction
5
6 2 Ethylene glycol (EG) is widely used in petrochemical industry and other fields, such
7
8
3 as polymer ester manufacturing and anti-freezer. It is mainly produced by ethylene
9
10
11 4 epoxidation and ethylene epoxide hydrolysis. However, the worldwide oil shortage
12
13 5 and the increasing demand of EG call for alternative non-oil routes to produce
14
15
16 6 glycols.1 Among those routes, C1 chemical process based on syngas mainly from coal
17
18 7 is proved to be promising.2 The liquid product of indirect syngas-to-EG reaction
19
20
21 8 contains 48.0wt%-51.0wt% methanol, 39.0wt%-43.0wt% EG, 1.5wt-2.5wt% water,
22
23 9 2.6wt-3.5wt% ethanol (EtOH), 2.0wt%-2.5wt% glycols including 1,2-butanediol
24
25
26 10 (1,2-BD), 2,3-butanediol (2,3-BD), 1,2-propylene glycol (1,2-PG), 1.5wt%-2.1wt%
27
28 11 unreacted methyl nitrite (MN), dimethyl oxatate (DMO), methyl glycolate (MG),
29
30
12 dimethyl carbonate (DMC) and 0.3wt-0.5wt% heavy glycols, including diethylene
31
32
33 13 glycol (DEG), triethylene glycol (TEG) and others.3
34
35 14 Currently, the traditional mode of separating the mixture consumes excessive amount
36
37
38 15 of energy because of the application of distillation column with large reflux ratio and
39
40 16 tray number to remove other glycols from EG, resulting in poor competitive
41
42
43 17 advantage of coal-based glycol production. Therefore, several methods have been
44
45 18 proposed to enhance the separation process, for example, azeotropic distillation.
46
47
48 19 Much work4, 5
has been devoted to selecting proper entrainer for the system but
49
50 20 reports on industrial application have rarely been seen. Besides, extraction is expected
51
52
21 to be used in separation process but the recovery and consumption of extracting
53
54
55 22 solvent should be taken into account.6 Moreover, reaction-assisted separation for the
56
57
2
58
59
60 ACS Paragon Plus Environment
Page 3 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 glycols developed in recent years was proved to be of great potential.7 In addition,
5
6 2 other novel separation ideas have been reported, such as extractive distillation2, 8 and
7
8
3 adsorption9, but they didn’t quite come off due to high toxicity or large wasting and
9
10
11 4 were dismissed in this paper.
12
13 5 Those ideas are all great but it is still not clear which route is more applicable to the
14
15
16 6 glycol system. Therefore, in this paper, four separation processes were designed and
17
18 7 evaluated based on process simulation by Aspen Plus. To ensure the accuracy of
19
20
21 8 process simulation, the vapor-liquid equilibrium data of glycol systems should be
22
23 9 firstly measured to retrieve the parameters of thermodynamic model.
24
25
26 10 2. Vapor-liquid equilibrium data
27
28 11 2.1 Measurement of thermodynamic data
29
30
12 The isobaric vapor-liquid equilibrium (VLE) data of the glycols has already been
31
32
33 13 measured by many researchers but it is still not systematic, especially the data of
34
35 14 binary systems composed of 1,2-BD, 2,3-BD, 1,2-PG and EG.10-15 Therefore, the
36
37
38 15 double circulating vapor-liquid equilibrium still was used to take measurements. The
39
40 16 temperature was measured with a precise digital thermometer (1552A-12-DL model,
41
42
43 17 provided by Fluke Calibration) with the accuracy ±0.01K. The pressure was regulated
44
45 18 by a pressure control system, including one vacuum pump, one digital manometer
46
47
48 19 (AOB-20 model) with the accuracy ±0.1 kPa and an aneroid barometer (DYM3 model)
49
50 20 with the accuracy ±0.2 kPa.
51
52
21 The VLE measurements for the systems (EG +1,2-BD), (EG +1,2-PG), (EG +2,3-BD),
53
54
55 22 (1,2-BD +1,2-PG), (1,2-BD +2,3-BD), (1.2-PG +2,3-BD) at pressure of 10.0kPa,
56
57
3
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 4 of 32

1
2
3
4 1 30.0kPa and 101.3 kPa were performed in the still with 110 cm3 capacity. Liquid
5
6 2 phase and condensed vapor phase were continuously circulated to ensure rapid
7
8
3 establishment of vapor-liquid equilibrium. The system was thoroughly cleaned by the
9
10
11 4 standard sample before experiment to eliminate contaminants. During the experiments,
12
13 5 the equilibrium was assumed to be reached when the pressure and temperature
14
15
16 6 maintained for 45 minutes.
17
18 7 The composition of the sample was analyzed by gas chromatograph (GC7890Ⅱ,
19
20
21 8 Techcomp (China) Ltd. ) equipped with a hydrogen flame ionization detector (FID)
22
23 9 and a DB-WAX capillary column (30m × 0.25 mm × 0.25 µm, Agilent). The
24
25
26 10 temperature of injector, oven and detector were kept at 523.15K, 393.15K and
27
28 11 543.15K, respectively. The carrier gas was high purity nitrogen (mass fraction 0.9999)
29
30
12 and the flow rate was 1 mL•min−1. Each VLE data was measured at least three times
31
32
33 13 to ensure accuracy. The external standard was chosen to obtain quantitative analysis
34
35 14 results by using the external standard curve to convert the peak areas into mole
36
37
38 15 fractions of each sample. Taken the measurement error into consideration, the
39
40 16 uncertainties of the vapor and liquid mole fractions were both calculated to be within
41
42
43 17 0.001. The isobaric VLE data (T, xi, yi) for the six binary mixtures were measured and
44
45 18 shown in Table S1-S6 in Supporting Information. The activity coefficients γi of the
46
47
48 19 liquid phase could be calculated from the experiment data by Eq.(1), as the vapor
49
50 20 phase was assumed as ideal gas under low experimental pressures.
51
52
(1)
53
54
55 21 where was the saturated vapor pressure of pure substance i at the system
56
57
4
58
59
60 ACS Paragon Plus Environment
Page 5 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 temperature T, which was calculated by Extended Antoine equation based on Aspen
5
6 2 Plus database. yi is the vapor phase mole fraction, xi is the liquid phase mole fraction,
7
8
3 and P is the total pressure of the system. The VLE data of EG +1,2-PG and 1,2-PG
9
10
11 4 +1,2-BD at 10.0kPa and EG +1,2-BD at 101.3kPa has already been measured by other
12
13 5 researchers10,13 and was not measured repeatedly.
14
15
16 6 2.2 Consistency test and data regression
17
18 7 To verify the measured quality of VLE values, all experimental results were checked
19
20
21 8 with Herington test and van Ness test proposed by Kang et.al16. As Herington17
22
23 9 suggested, if , the pertinent experimental isobaric VLE data can be
24
25
26 10 considered thermodynamically consistent. In the process, D and J were calculated as
27
28 11 Eq.(2) and Eq.(3).
29
30
31
32 (2)
33
34
35 (3)
36
37
38
39 12 where Tmax and Tmin were the highest and lowest temperature in the system,
40
41
13 respectively. The γ values of the systems (2,3-BD +1,2-PG, 2,3-BD +1,2-BD, EG
42
43
44 14 +1,2-PG, 1,2-PG +1,2-BD) were all between 0.95 and 1.10 at 10.0kPa, 30.0kPa and
45
46 15 101.3kPa, hence the area test was not employed. The values of D -Jfor the system
47
48
49 16 {EG +1,2-BD} were calculated to be 1.48, 0.09, 1.05 at 10.0kPa, 30.0kPa, 101.3kPa,
50
51 17 respectively, while the values of the binary system {2,3-BD +EG} were 3.71, 2.27
52
53
54 18 and 1.88 at 10.0kPa, 30.0kPa and 101.3kPa, which meant the experimental data
55
56 19 passed the Herington test.
57
5
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 6 of 32

1
2
3
4 1 Van Ness point test described by Fredenslund18 was applied to confirm the accuracy
5
6 2 of the experimental data. The index average absolute deviation (AAD) was defined by
7
8
3 following equation:
9
10
11 (4)
12
13 4 where Ci is the independent variable (bubble point temperature, vapor phase mole
14
15
16 5 fraction) of the pure component i, N is the number of experimental data points, the
17
18 6 superscript exp denotes data from experiments and cal denotes data calculated with
19
20
21 7 thermodynamic models.
22
23 8 The experimental data were correlated by minimizing object function Q, which was
24
25
26 9 defined by the sum of squared differences between calculated and experimental data
27
28 10 as Eq.(5):
29
30 N Te ,i - Tm ,i Pe,i - Pm ,i xe ,i - xm ,i y e ,i - y m ,i
31 Q = ∑[( )2 + ( )2 + ( )2 + ( )2 ] (5)
32 i =1 σ T ,i σ P ,i σ x ,i σ y ,i
33
34 11 where y, x represent mole fraction of pure component in the vapor phase and in the
35
36
37 12 liquid phase. T and P are temperature and pressure. i represents each pure component
38
39 13 in the binary mixtures, N is the number of experimental data points, and σ represents
40
41
14 the estimated standard deviation of measured experimental values. e, m represent
42
43
44 15 estimated value and measured value, respectively.
45
46 16 Correlated interaction parameters, together with the average absolute deviation (AAD)
47
48
49 17 of equilibrium temperature, vapor phase mole fraction for the binary systems between
50
51 18 experimental values and calculated values (obtained from Wilson, NRTL and
52
53
54 19 UNIQUAC models) were shown in Table 1.
55
56 20 Table 1 Correlated Interaction Parameters and the Average Absolute Deviations (AAD) between
57
6
58
59
60 ACS Paragon Plus Environment
Page 7 of 32 Industrial & Engineering Chemistry Research

1
2
3
1 Experimental and Calculated Values
4
5
AAD AAD (P)/
6 p/kPa model B12/K B21/K α AAD (y)
7 (T)/K kPa
8 1,2-PG(1) +EG(2)
9 Wilson 11.2114 -890.5198 0.0059 0.0285 6.7907e-06
10
10.0 NRTL 1025.1681 -164.0382 0.3 0.0060 0.0402 9.3288e-06
11
12 UNIQUAC -481.7063 102.8876 0.0061 0.0404 9.3784e-06
13 Wilson -303.8596 -435.0841 0.0086 0.0290 2.2166e-05
14 30.0 NRTL 379.5311 407.0284 0.3 0.0087 0.0367 2.9864e-05
15 UNIQUAC -149.2076 -160.0364 0.0087 0.0363 2.9381e-05
16
17 Wilson -303.8596 -502.8533 0.0068 0.0898 2.9896e-04
18 101.3 NRTL 306.2324 598.3744 0.3 0.0081 0.0942 3.0985-04
19 UNIQUAC -59.9569 -301.4157 0.0080 0.0932 3.0700e-04
20 1,2-BD(1) +EG(2)
21
22 Wilson -365.3776 177.9303 0.0074 0.0055 1.3376e-06
23 10.0 NRTL -274.0664 471.9168 0.3 0.0074 0.0094 2.2593e-06
24 UNIQUAC 76.2143 -137.4402 0.0074 0.0057 1.3666e-06
25 Wilson -553.1914 218.5276 0.0039 0.0137 1.1115e-05
26
30.0 NRTL -344.9806 688.6855 0.3 0.0046 0.0225 1.8333e-05
27
28 UNIQUAC 137.4240 -253.4244 0.0040 0.0132 1.0805e-05
29 Wilson -357.4847 129.9490 0.0030 0.0099 3.3322e-05
30 101.3 NRTL -234.7664 476.8356 0.3 0.0032 0.0110 3.6732e-05
31
UNIQUAC 12.3188 -87.5923 0.0030 0.0098 3.2597e-05
32
33 2,3-BD(1) +EG(2)
34 Wilson -466.6218 251.7104 0.0085 0.0468 1.0411e-05
35 10.0 NRTL -343.8058 550.6957 0.3 0.0086 0.0313 6.8939e-06
36 UNIQUAC 154.1807 -233.8679 0.0084 0.0450 1.0131e-05
37
38 Wilson -345.4576 182.9973 0.0039 0.0152 1.2047e-05
39 30.0 NRTL -293.5719 477.0140 0.3 0.0038 0.0206 1.6138e-05
40 UNIQUAC 72.6219 -126.3332 0.0039 0.0149 1.2037e-05
41 Wilson -174.4196 44.6094 0.0082 0.0118 4.0027e-05
42
101.3 NRTL -94.9450 227.5011 0.3 0.0082 0.0094 3.3662e-05
43
44 UNIQUAC -91.3732 31.7564 0.0082 0.0128 4.2203e-05
45 1,2-PG(1) +1,2-BD(2)
46 Wilson 214.5182 -309.1564 0.0039 0.0494 1.1516e-05
47
10.0 NRTL 385.0055 -290.5894 0.3 0.0040 0.0510 1.1953e-05
48
49 UNIQUAC -192.5574 150.5187 0.0039 0.0492 1.1472e-05
50 Wilson -104.6842 82.7958 0.0066 0.0197 1.6738e-05
51 30.0 NRTL -139.0940 167.2831 0.3 0.0066 0.0193 1.6403e-05
52 UNIQUAC 98.9830 -119.0317 0.0066 0.0186 1.5832e-05
53
54 Wilson 242.3635 -369.2433 0.0047 0.0122 3.8772e-05
55 101.3 NRTL 408.5374 -300.8118 0.3 0.0050 0.0130 4.0436e-05
56 UNIQUAC -225.1837 171.5125 0.0047 0.0122 3.8509e-05
57
7
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 8 of 32

1
2
3 2,3-BD(1) +1,2-PG(2)
4 Wilson -304.6439 215.6571 0.0043 0.0210 4.692e-06
5
10.0 NRTL -296.5335 389.0352 0.3 0.0042 0.0246 5.4823e-06
6
7 UNIQUAC 149.0168 -188.3749 0.0043 0.0217 4.8438e-06
8 Wilson -346.6179 240.3541 0.0058 0.0032 3.1805e-06
9 30.0 NRTL -316.4019 416.5423 0.3 0.0056 0.0053 4.3514e-06
10 UNIQUAC 165.1422 -210.8901 0.0058 0.0032 3.1468e-06
11
12 Wilson 219.9860 -299.5595 0.0093 0.0030 1.1982e-05
13 101.3 NRTL 333.0377 -268.0018 0.3 0.0092 0.0042 1.716e-05
14 UNIQUAC -29.0645 24.5504 0.0087 0.0226 7.0655e-05
15 2,3-BD(1) +1,2-BD(2)
16
Wilson 277.1605 -454.6703 0.0054 0.0931 2.0524e-05
17
18 10.0 NRTL 418.4868 -310.2117 0.3 0.0061 0.0792 1.7419e-05
19 UNIQUAC -259.2443 192.6849 0.0056 0.0864 1.9031e-05
20 Wilson 361.2147 -817.1036 0.0053 0.0081 6.3589e-06
21
30.0 NRTL 799.0103 -464.4162 0.3 0.0056 0.0427 3.3472e-05
22
23 UNIQUAC -457.6764 282.4736 0.0052 0.0251 1.9698e-05
24 Wilson -412.5052 290.0799 0.0041 0.0547 1.7443e-04
25 101.3 NRTL -366.2347 469.8231 0.3 0.0042 0.0548 1.74244-04
26
UNIQUAC 198.3000 -249.2973 0.0041 0.0561 1.7866e-04
27
28 1 The AAD of bubble point temperature, pressure and vapor phase mole fraction are
29
30
31 2 less than 0.1K, 1Pa and 0.01 respectively in all of cases, which indicate that the three
32
33 3 models agree well with experimental data and can be used in industrial application.
34
35
36 4 Among those models, non-random two liquid (NRTL) model, as a common tool for
37
38 5 fitting in many literatures19-21, has been reported for VLE and LLE in multicomponent
39
40
41 6 systems consisting alcohols, water, aromatic hydrocarbons, e.g. toluene and glycols
42
43 7 that are all involved in the simulation while the other two models have rarely been
44
45
8 reported. Therefore, NRTL thermodynamic model was applied to simulate the
46
47
48 9 separation process. The activity coefficient calculated in Aspen Plus is shown as
49
50 10 Eq.(6).
51
52
53
(6)
54
55
56
57
8
58
59
60 ACS Paragon Plus Environment
Page 9 of 32 Industrial & Engineering Chemistry Research

1
2
3
4
1 where
5
6 2
7
8 3
9
10
11 4 In order to retireve the binary interaction parameters correspongding the models, the
12
13 5 data regression module of Aspen Plus was employed based on the least-squares
14
15
16 6 method subject to the maximum likehood principle following Britt-Luecke
17
18 7 algorithm22.The model parameters regressed by experimental data were shown in
19
20
21 8 Table 2.
22
23 9 Table 2 Model parameters fitted by experimental data
24
25 component i 1,2-PG 1,2-BD 2,3-BD 1,2-PG 2,3-BD 2,3-BD
26 component j EG EG EG 1,2-BD 1,2-PG 1,2-BD
27
0 0 0 0 0 0
28
29 0 0 0 0 0 0
30 (K) 306.2324 -234.7664 -94.9450 408.5374 333.0377 -366.2347
31 (K) 598.3744 476.8356 227.5011 -300.8118 -268.0018 469.8231
32
0.3 0.3 0.3 0.3 0.3 0.3
33
34 10 The NRTL parameters of the binary systems composed of methanol, ethanol, glycols,
35
36
37 11 DMC, DMO and MG were adjusted based on the reported literatures1, 23-25.
38
39 12 According to VLE results correlated by NRTL model, the key to obtain EG with high
40
41
42 13 quality is removal of other glycols, including 1,2-PG, 2,3-BD as these components
43
44 14 form close boiling point mixtures at the entire range of their compositions, especially
45
46
47
15 1,2-BD which forms azeotrope with EG shown in Figure 1. The difficulty to separate
48
49 16 the mixture of glycols only by conventional distillation requires application of other
50
51
17 special distillation technology.
52
53
54 18 3. Processes descriptions and simulations
55
56 19 To obtain polymer grade EG, three separation processes were proposed and simulated
57
9
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 10 of 32

1
2
3
4 1 by Aspen Plus to calculate equipment sizes and energy cost, including azeotropic
5
6 2 distillation route, distillation coupled with liquid-liquid extraction and distillation
7
8
3 coupled with acetalization reaction. Here separation process currently applied in
9
10
11 4 conventional industry was also simulated as an investment for comparison. The
12
13 5 distillaton columns were firstly designed by DSTWU modules to determine tray
14
15
16 6 number, reflux ratio and distillate rate. The columns were then rated by RadFrac
17
18 7 modules. The operational conditions of the equipments were optimized by sensitive
19
20
21 8 analysis with the object to decrease energy cost, e.g. reduce reflux ratio and reboiler
22
23 9 duty under the prerequisite of product quality.
24
25
26 10 3.1 Conventional distillation route (CD route)
27
28 11 The four-column distillation process widely used in coal-based ethylene glycol
29
30
12 industry was shown in Figure 2. The operational conditions of columns and main
31
32
33 13 information of streams were also demonstrated in the figure while some details were
34
35 14 omitted due to space reasons.
36
37
38 15 Note that methanol could be served as raw materials to produce MN and DMO in
39
40 16 previous CO coupling reaction.3 Therefore, the majority of methanol was recovered
41
42
43 17 by Col-11 with 57 stages, where the reflux ratio was set at 1.9 to ensure the methanol
44
45 18 purity according to simulation results. The recovery of methanol was determined
46
47
48 19 according to sensitive analysis of Col-11. From Figure 3, when the recovery of
49
50 20 methanol is too high a sharp increase of reboiler duty occurs due to increasing
51
52
21 difficulty of separation according to the tendency of minimum reflux ratio Rmin. Thus
53
54
55 22 the methanol recovery was reasonably set at 95.0%. B11 from the bottom stream of
56
57
10
58
59
60 ACS Paragon Plus Environment
Page 11 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 Col-11 was separated by Col-12 to remove the remaining methanol, by-product
5
6 2 ethanol and other light components. D12 was assumed to be used as fuel and was
7
8
3 unnecessary to be purified. Then unreacted MG and DMO was recovered by Col-13.
9
10
11 4 For separation of glycols, it is widely used as expediency in industry to distillate a
12
13 5 part of EG carrying almost all other glycols to the top of column, of which the
14
15
16 6 concentration of EG should be up to 95wt% so that it could be still sold as qualified
17
18 7 product at relative lower price. Then EG (> 99.9wt%) could be gained from the
19
20
21 8 sidestream while other heavy glycols (DEG, TEG) were removed from the system at
22
23 9 the bottom of Col-14. Obviously, it is necessary to apply large tray number and large
24
25
26 10 reflux ratio to Col-14 to ensure the purity of EG at the sidestream. The concentration
27
28 11 profile of EG in Col-14 (at RR=24) was shown in Figure 4. As it will keep up to
29
30
12 99.9wt% below the 15th plate in vapor phase, the drew-position was accordingly set at
31
32
33 13 the 16th plate considering tower performance. Col-14 was operated at 0.05bar to
34
35 14 decrease reboiler temperature so that cheaper low pressure steam could be used.
36
37
38 15 3.2 Azeotropic distillation route(AD route)
39
40 16 The form of close boiling point mixture and even azeotropic mixture discounts EG
41
42
43 17 quality and yield in CD route and hence causes large diseconomy in manufacture.
44
45 18 Therefore, many researches have been devoted to development of azeotropic
46
47
48 19 distillation technology for glycols system.
49
50 20 The application of azeotropic distillation to separating glycols system was firstly
51
52
21 proposed by Berg26 in 1990. He listed about 30 potential entrainers to enhance the
53
54
55 22 separation via heterogeneous azeotropic distillation. The past two decades has
56
57
11
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 12 of 32

1
2
3
4 1 witnessed developing performance of azeotropic distillation technology in separation
5
6 2 of EG and 1,2-BD. Song5 has used homotopy-Newton united method to predict
7
8
3 azeotropic component and proved ethylene benzene was the best entrainer by fuzz
9
10
11 4 mathematics. Wang27 simulated separation process for EG and 1,2-BD through
12
13 5 azeotropic distillation and compared it with conventional distillation. The result
14
15
16 6 showed that azeotropic distillation is superior to conventional distillation on the
17
18 7 operating cost and investment. Li28 conducted several experiments to separate EG and
19
20
21 8 1,2-BD and the purity of EG can be up to 99.7wt% and recovery reached 80.32%.
22
23 9 Furthermore, the feasibility of application of azeotropic distillation to separating
24
25
26 10 glycols has been studied by experiments.4 Based on those researches, a feasible
27
28 11 azeotropic distillation process with the optimized results was shown in Figure 5.
29
30
12 Similar to CD route, methanol and DMO should be recovered firstly and meanwhile
31
32
33 13 light impurities, including ethanol and water were removed from the system by
34
35 14 Col-21, Col-22 and Col-23 which were similar with Col-11, Col-12 and Col-13 and
36
37
38 15 not shown repeatedly in Figure 5. Then the mixture of glycols was pumped into
39
40 16 Col-24, where the mixture of EG, 1,2-BD, 1,2-PG, 2,3-BD was distillated out. The
41
42
43 17 majority of EG mixed with heavy glycols from the bottom was separated by Col-25.
44
45 18 1,2-BD in the mixture from the top of Col-24 was expected be removed by azeotropic
46
47
48 19 distillation using ethylbenzene (EB) as entrainer. The model parameters of binary
49
50 20 mixture of EG and ethylbenzene were adjusted according to Yang’s work29 to ensure
51
52
21 the accuracy of simulation. According to the result of property analysis in Aspen Plus,
53
54
55 22 1,2-PG is bad for the process of azeotropic distillation. Therefore, the mixture should
56
57
12
58
59
60 ACS Paragon Plus Environment
Page 13 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 be pretreated to remove a part of 1,2-PG from the system by Col-26 with 69 trays to
5
6 2 ensure quality of the final product. The bottom product of Col-26 was pumped into
7
8
3 Col-27, together with EB, where 1,2-BD was carried by EB to the top and thus EG
9
10
11 4 (>99.9wt%) could be obtained at the bottom. Then ethylbenzene-rich phase from the
12
13 5 decanter flowed back to Col-27 while glycol-rich phase was treated by Col-28 to
14
15
16 6 recover EB.
17
18 7 The mass ratio of entrainer to feed was 0.5-0.8 according to the result of sensive
19
20
21 8 analysis shown in Figure 6. Under the goal that the recovery of 1,2-BD in Col-27 was
22
23 9 set at 0.995, the addition of entrainer makes it easier to separate the two glycols,
24
25
26 10 therefore the minimum tray number and input energy decrease with the increase of EB.
27
28 11 However, too much entrainer will cause the increase of column capacity and thus the
29
30
12 increase of energy cost.
31
32
33 13 3.3 Distillation coupled with liquid-liquid extraction route(ED route)
34
35 14 Apart from azeotropic distillation, there are some researches on application of
36
37
38 15 liquid-liquid extraction on separating EG and 1,2-BD. It has been proved that toluene
39
40 16 was the best agent among aromatics, ethers, alkanes as well as ketones.6, 9, 28 Sun23
41
42
43 17 simulated several extraction processes for separating EG and 1,2-BD using toluene as
44
45 18 extracting solvent. Based on those researches, a detailed conceptual design and
46
47
48 19 operation of separation process was shown in Figure 7. Col-31-Col-33 were similar
49
50 20 with Col-11-Col-13 in Figure 2 and not shown.
51
52
21 After the recovery of methanol, MG and DMO, the stream B33 was pumped to the
53
54
55 22 extraction section. The extraction column was operated at 313K and 1 bar based on
56
57
13
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 14 of 32

1
2
3
4 1 the experimental results in the literature28, in which the feed is counter-currently
5
6 2 brought into contact with toluene i.e. extracting solvent. The extract stream leaving
7
8
3 the extraction column contains 1,2-BD, coextracted EG and extracting solution.
9
10
11 4 The process designs included extraction and distillation, which were simulated by
12
13 5 using the Extract and RadFrac modules of Aspen Plus, respectively while the
14
15
16 6 equilibrium stage model was used, of which the distribution coefficient of toluene was
17
18 7 estimated by UNIFAC group-contribution method with the help of LLE data20. To
19
20
21 8 ensure the quality of EG product, the mass fraction of 1,2-BD in raffinate phase
22
23 9 should be below 0.001, then the minimum solvent to feed mole ratio was calculated to
24
25
26 10 be 0.1 using Eq. (7).
27
28 (7)
29
30
11 where S is flow rate of the solvent, F is flow rate of the feed, Xin is mass fraction of
31
32
33 12 the solute i.e. 1,2-BD in the feed, Xout is mass fraction of the solute in the raffinate
34
35 13 phase, Yin is the mass fraction of the solute in the entering solvent, and Yout is the mass
36
37
38 14 fraction of the solute in the extractant phase, which is in equilibrium with Xin. As
39
40 15 suggested by King30, an appropriate estimation of the optimal solvent to feed ratio
41
42
43 16 should be 1.15 ~2 times of (S/F)min to ensure product quality and in this work it was
44
45 17 set to be 1.5. Therefore, the flow rate of solvent was calcultated to be about 5000kg/h.
46
47
48 18 The extraction phase (D34) including the majority of toluene and 1,2-BD flowing out
49
50 19 from the top of extraction tower was separated by Col-36 to recover the solvent while
51
52
20 the raffinate phase (B34) from the bottom was separated by Col-35, where the mixture
53
54
55 21 of 1,2-PG and toluene was distillated from the system and treated as fuel. The bottom
56
57
14
58
59
60 ACS Paragon Plus Environment
Page 15 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 liquid of Col-35 was separated by Col-38, where EG (>99.9wt%) was obtained at the
5
6 2 top. Col-36-Col-38 were operated under lower pressure to reduce operational
7
8
3 temperatures to keep products/solvent from spoilling and save energy.
9
10
11 4 3.4 Reaction-assisted distillation route (RD route)
12
13 5 The novel idea has been developed in recent years that chemical reactions could be
14
15
16 6 used to enhance separation process. Dhale’s work31 has shown excellent performance
17
18 7 in the recovery of EG, 1,2-PG and 1,2-BD from aqueous solution through
19
20
21 8 acetalization reaction. The acetals formed by acetaldehyde (ALK) and glycols were
22
23 9 easier to be separated and then additional reactive distillation columns were needed
24
25
26 10 for hydrolysis of acetals back to EG and 1,2-PG, which showed great advantage for
27
28 11 EG with low concentration. As for manufacturing of EG, Huang et al.32 studied
29
30
12 reaction kinetics and developed a reactive distillation process for separation of EG
31
32
33 13 and 1,2-BD. Based on those previous researches, a conceptual design was established
34
35 14 shown in Figure 8.
36
37
38 15 Similar to AD route in Figure 5, light components were removed by Col-41-Col-43
39
40 16 and the majority of EG was obtained after distillation by Col-44 and Col-45. Here
41
42
43 17 they were omitted as they were similar with AD route. The brief introduction of
44
45 18 Reaction-assisted distillation (RD) route begins with the stream D44 from Col-44.
46
47
48 19 Unlike AD route, the distillates from Col-44 (similar with Col-24) was pumped,
49
50 20 together with acetaldehyde into the reactor R41, which was operated under 1.5 bar,
51
52
21 333.15 K. A 2:1 mole ratio of acetaldehyde to glycols was set to ensure the conversion
53
54
55 22 of glycols based on the previous study32. Besides, the conversions of EG and 1,2-BD
56
57
15
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 16 of 32

1
2
3
4 1 were fixed to 0.61 and 0.70, respectively, based on the previously published
5
6 2 experimental results7. The conversions of 1,2-PG and 2,3-BD have not been reported
7
8
3 and they were assumed to be 0.70, due to their chemical properties were similar with
9
10
11 4 1,2-BD and their concentrations were too little to have effect on energy cost. The
12
13 5 acetalization reactions of glycols following Eq(8)-(11) occurred in the reactor.
14
15 OH
16 H O
17 + -H2O
18 (8)
O
19 OH O
(EG) (ALK) (2-MD)
20
21 OH
H O
22 -H2O
+
23 (9)
24 OH
O
O
25 (1,2-PG) (ALK) (2,4-DMD)
26
27 OH
28 H O

29 + -H2O (10)
30 OH
O
O
31 (1,2-BD) (ALK) (4-EMD)
32 OH
H O
33 -H2O
+
34 (11)
35 OH
O
O
36 (2,3-BD) (ALK) (2,4,5-TMD)
37
38 6 The distillation tower Col-46 was to recover the unreacted acetaldehyde and glylcols.
39
40 7 The distillation tower Col-47 and Col-48 were aimed to separate the acetals.Then EG
41
42
43 8 could be obtained by acetals hydrolysis in the continuous reactive distillation column
44
45 9 R42, where the conversions of hydrolysis reaction were set at 100% according to the
46
47
48 10 experimental results in a pilot scale column as mentioned by Atul D. Dhale33. The
49
50 11 hydrolysis of 2,4-DMD, 4-EMD and 2,4,5-TMD occurred in R43, where acetaldehyde
51
52
12 was recovered. Then water in S42 and S43 was removed by Col-48 and Col-49,
53
54
55 13 respectively.
56
57
16
58
59
60 ACS Paragon Plus Environment
Page 17 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 The process included reactor and distillation tower, which were simulated using the
5
6 2 Rstoic and RadFrac modules of Aspen Plus, respectively, while the reactive
7
8
3 distillation columns R42-R44 were simulated by using RadFrac modules. The binary
9
10
11 4 interaction parameters of acetals and water were obtained from regression of the VLE
12
13 5 data measured by Chopade et al.34, 35
. The VLE data as well as liquid-liquid
14
15
16 6 equilibrium (LLE) data of the binary system 2MD +4EMD were measured in the
17
18 7 preliminary work7.
19
20
21 8 4. Economic estimation
22
23 9 After the initial steady state design of each section, optimization was performed to
24
25
26 10 find a set of optimal design and operating variables with minimizing total annual cost
27
28 11 (TAC), including adjusting tray number and reflux ratio, choosing the optimal
29
30
12 position to feed and draw-out36. According to Luyben’s work37, the four processes
31
32
33 13 were evaluated by annual cost and profit. Total annual cost (TAC) contains capital
34
35 14 investigation and operational cost.
36
37
38 (12)
39
40 15 Payback period was usually set at 3 years in industry. The capital investigation
41
42
43 16 included columns and heat exchangers and the purchase costs of pumps and other
44
45 17 equipments were too small to be taken in account38.
46
47
48 18 Stainless steel was specified as the material in coolers, reboilers, tanks, and internals
49
50 19 to avoid iron contamination in the final product. Therefore, the cost of columns was
51
52
20 calculated by Eq.(13) based on the height of column L and the diameter of column D
53
54
55 21 esitimated by Tray Sizing module in Aspen Plus following Eq.(14).
56
57
17
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 18 of 32

1
2
3
4 (13)
5
6 (14)
7
8
1 where NT is tray number of each column.
9
10
11 2 The estimation of reactor cost followed Eq. (13), either. Under the assumption that the
12
13 3 reactor could be considered as CSTR(continuous stirred tank reactor) and the ratio of
14
15
16 4 D and L was generally set at 3. D and L were estimated by reactor volume, which was
17
18 5 calculated by Eq. (15).
19
20
21 (15)
22
23 6 where Q is the flow rate of input stream, t is space time depending on optimal time of
24
25
26 7 reaction at different operating conditions. α is the ratio of effective volume to actual
27
28 8 volume, which is always set at 1.5~2 in industrial design to take the cost of additional
29
30
9 volumes and internals into account. The cost of catalyst is too low to be considered.
31
32
33 10 The cost of heat exchangers was calculated following Eq.16 based on the heat transfer
34
35 11 area of each heat exchanger. The latter was estimated by Aspen Exchanger Design and
36
37
38 12 Rating following Eq.(16).
39
40 (16)
41
42
43 13 Energy cost was calcuated based on the heat duty of heat exchangers. For reboilers
44
45 14 and heaters, the price of high pressure (41barg, 251℃) steam was 9.87$/GJ while the
46
47
48 15 prices of middle pressure (10barg, 184℃) and low pressure (5barg, 160℃) steam
49
50 16 were 8.22$/GJ and 7.78GJ$, respectively. In the four separation processes, the
51
52
17 condensers were all cooled by cooling water, whose price was 0.354$/GJ.
53
54
55 18 Apart from input energy, those separations need additional materials for supplement,
56
57
18
58
59
60 ACS Paragon Plus Environment
Page 19 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 including ethylbenzene in AD route, toluene in ED route as well as acetaldehyde in
5
6 2 RD route. The output value was determined by production of EG. The prices of
7
8
3 chemicals were inquired in Molbase database. The ecomomic estimation results of the
9
10
11 4 four separation processses were shown in Table 3.
12
13 5 Table 3 Economic estimation of separation processes
14 Price CD AD ED RD
15 Column&Reactor/10 $ 6
9.72 8.99 7.88 9.19
16
heat exchanger/106$ 4.36 6.25 5.63 6.08
17
18 energy cost/106$/a 44.00 17.94 25.10 23.76
19 additional material//106$ 0.00 0.40 3.74 0.01
20 TAC/106$ 48.69 23.42 33.34 28.86
21 output/(106$/a) 296.28 317.75 318.42 336.76
22
23 profit/(106$/a) 247.59 294.33 285.08 307.89
24
6 Table 3 reveals that TAC is mainly determined by energy cost and the annual profit
25
26
27 7 depends mainly on the yield of EG (output). In CD route, the existence of 1,2-PG,
28
29 8 1,2-BD and 2,3-BD will result in decrease of EG output and increase of imported
30
31
32 9 energy and hence cause reduction of annual profit. Compared with the CD route, the
33
34 10 application of azeotropic distillation technology improves the yield of EG and
35
36
37 11 reduced the energy cost of separation, which demonstrated great potential. However,
38
39 12 ethylbenzene, as is proved to be the best entrainer though, has danger of serious
40
41
42 13 damage to health by prolonged exposure through inhalation and if swallowed.
43
44 14 Therefore, the application of azeotropic distillation technology on separation of
45
46
15 glycols system in industry requires development of non-toxic entrainer, e.g.
47
48
49 16 ionic-liquid. The application of liquid-liquid extraction on separating EG and 1,2-BD
50
51 17 is able to increase the productivity of EG but the process to remove 1,2-PG and
52
53
54 18 recover extracting agent still consumes large quantity of energy. In addition, owing to
55
56 19 great quantity and high toxicity of extracting solvent, ED route shows no advantage
57
19
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 20 of 32

1
2
3
4 1 whether from economic angle or from the perspective of the environment. By
5
6 2 comparison, reaction-assisted separation with acetaldehyde is an effectively way to
7
8
3 solve the puzzle of separating glycols mixture due to low energy consumption and
9
10
11 4 large EG output while additional products such as 1,2-PG and 1,2-BD could be
12
13 5 obtained. Thus, the RD route was calculated to be the best strategy for separation of
14
15
16 6 liquid product of ethylene glycol synthesis from syngas. Furthermore, as an idea
17
18 7 developed in recent years, reverse reacion-assisted separation needs more researches,
19
20
21 8 such as, developing new suitable reverse reactions e.g. reaction with butanone and
22
23 9 propionaldehyde39, application of reactive distillation to promote reaction conversion
24
25
26 10 and save energy, etc.
27
28 11 5. Conclusion
29
30
12 In this work, the rigorous steady state design and optimization work of four separation
31
32
33 13 processes for liquid products of EG synthesis from syngas were investigated. The
34
35 14 thermodynamic data of glycol system was measured to ensure the accuracy of process
36
37
38 15 simulation by Aspen Plus.
39
40 16 The result of economic analysis indicated that azeotropic distillation owned great
41
42
43 17 potential due to lower energy cost compared to CD route but it still needed promotion,
44
45 18 for example, development of entrainer. Among those separation routes, the distillation
46
47
48 19 process coupling with acetalization reaction was currently proved to be the best
49
50 20 strategy for the glycols system in industry due to high yield of EG and low energy
51
52
21 cost.
53
54
55 22 Supporting Information
56
57
20
58
59
60 ACS Paragon Plus Environment
Page 21 of 32 Industrial & Engineering Chemistry Research

1
2
3
4 1 Tables of vapor-liquid equilibrium data measured for the binary mixture composed of
5
6 2 EG, 1,2-PG, 1,2-BD and 2,3-BD are available in the Supporting Information.
7
8
9 3 Acknowledgements
10
11 4 The authors are grateful for the financial support from National Natural Science
12
13 5 Foundation of China (Nos. 21690084, 21336007).
14
15
16 6 Corresponding Author
17
18 7 *E-mail: [email protected] (Xin Gao) Tel: +86-022-27404701
19
20
21 8 Notes
22
23 9 The authors declare no competing financial interest.
24
25
26 10 References
27
11 (1) Sun, L.; Fu, J.; Li, W.; Zhan, X. L.; Xiao, W. D. Binary vapor-liquid equilibrium of methyl
28
29 12 glycolate and ethylene glycol. Fluid Phase Equilib. 2006, 250, 33-36.
30 13 (2) Chen, Y.; Hung, S.; Lee, H.; Chien, I. Energy-Saving Designs for Separation of a
31 14 Close-Boiling 1,2-Propanediol and Ethylene Glycol Mixture. Ind. Eng. Chem. Res. 2015, 54,
32 15 3828-3843.
33
16 (3) Yu, B.; Chien, I. Chemical Engineering Research and Design Design and optimization of
34
35 17 dimethyl oxalate ( DMO ) hydrogenation process to produce ethylene glycol ( EG ). Chem. Eng.
36 18 Res. Des. 2017, 121, 173-190.
37 19 (4) Niu, Y.; Liu, Z.; Qiao, K.; Chen, M. Study on Separation of Glycol and 1,2-Propanediol with
38
20 Heterogeneous Azeotropic Distillation. Contemp. Chem. Ind. (China) 2011, 40, 560-561.
39
40 21 (5) Song, G. Study on the Separation of Glycol-Propanediol-Butanediol System Using
41 22 Azeotropic Distillation. Thesis, Tianjin University, Tianjin, 2006.
42 23 (6) Dai, C.; Pan, G.; Liu, Y.; Li, Q. A novel method to separate ethylene and 1,2-butanediol with
43 24 azeotropic distillation coupled with extraction technology. China Patent CN 103396290-A, 2013.
44
25 (7) Li, H.; Huang, W.; Li, X.; Gao, X. Application of the Aldolization Reaction in Separating the
45
46 26 Mixture of Ethylene Glycol and 1,2-Butanediol: Thermodynamics and New Separation Process.
47 27 Ind. Eng. Chem. Res. 2016, 55, 9994-10003.
48 28 (8) Figueirêdo, M. F.; Brito, K. D.; Ramos, W. B.; Vasconcelos, L. G. S.; Brito, R. P.; Brito, K.
49
29 D.; Ramos, W. B.; Vasconcelos, L. G. S.; Brito, R. P., Optimization of the Design and Operation of
50
51 30 Extractive Distillation Processes Optimization of the Design and Operation of Extractive
52 31 Distillation Processes. Sep. Sci. Technol. 2015, 50, 2238-2247.
53 32 (9) Xiao, J.; Zhong, L.; Guo, Y. Separation method for glycol, propylene glycol and butylene
54 33 glycol. China Patent CN 102372600 A, 2012.
55
34 (10) Zhu, L. Yan, J.; Xiao, W. Determination and correlation of VLE data for ethylene glycol and
56
57
21
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 22 of 32

1
2
3 1 1,2-butanedioi system. Chem. Eng. (China) 2012, 40, 34-37.
4 2 (11) Chiavone-filho, O.; Proustj, P., Vapor-Liquid Equilibria for Glycol Ether + Water Systems. J.
5
3 Chem. Eng. Data 1993, 38, 128-131.
6
7 4 (12) Yue, H.; Zhao, Y.; Ma, X.; Gong, J. Ethylene glycol: properties, synthesis, and applications.
8 5 Chem. Soc. Rev. 2012, 41, 4218.
9 6 (13) Yang, Q.; Jin, F.; Feng, X.; Zhi, J.; Yang, C., Isobaric Vapor–Liquid Equilibrium for Two
10 7 Binary Systems ( n -Butanol + 1,4-Butanediol and γ-Butyrolactone + 1,4-Butanediol) at p = (30.0,
11
8 50.0, and 70.0) kPa. J. Chem. Eng. Data 2016, 61, 3034-3040.
12
13 9 (14) Zhong, Y.; Wu, Y.; Zhu, J.; Chen, K.; Wu, B.; Ji, L. Thermodynamics in separation for the
14 10 ternary system 1,2-ethanediol + 1,2-propanediol + 2,3-butanediol. Ind. Eng. Chem. Res. 2014, 53,
15 11 12143-12148.
16
12 (15) Yang, C.; Feng, X.; Sun, Y.; Yang, Q.; Zhi, J. Isobaric Vapor–Liquid Equilibrium for Two
17
18 13 Binary Systems{Propane-1,2-diol + Ethane-1,2-diol and Propane-1,2-diol + Butane-1,2-diol} at p
19 14 = (10.0, 20.0, and 40.0) kPa. J. Chem. Eng. Data 2015, 60, 1126-1133.
20 15 (16) J.W., Kang; V., Diky; R.D. Chirico; J.W. Magee; C.D. Muzny; I. Abdulagatov; A.F. Kazakov;
21 16 Michael. F. Quality assessment algorithm for vapour-liquid equilibrium data. J. Chem. Eng. Data
22
17 2010, 55, 3631-3640.
23
24 18 (17) Herington, E. F. G. Tests for the consistency of experimental isobaric vapour-liquid
25 19 equilibrium data. J. Pet. Ins. 1951, 37, 457-470.
26 20 (18) Hack C. W. Winkle M. V. Vapor-Liquid Equilibria of the Diacetone Alcohol–Water System at
27
21 Subatmospheric Pressures. Ind. Eng. Chem. Res., 1954, 46, 2392-2395.
28
29 22 (19) Esteban J.; Ladero M.; García-Ochoa F. Liquid–liquid equilibria for the systems ethylene
30 23 carbonate+ethylene glycol+glycerol; ethylene carbonate+glycerol carbonate+glycerol and
31 24 ethylene carbonate+ethylene glycol+glycerol carbonate+glycerol at catalytic reacting temperatures.
32 25 Chem. Eng.g Res. Des., 2014, 94, 440-448.
33
26 (20) Mohsen-Nia M.; Doulabi F. S. M.; Manousiouthakis V. I. (Liquid + liquid) equilibria for
34
35 27 ternary mixtures of (ethylene glycol + toluene + n-octane). J. Chem. Thermodyn. 2008, 40,
36 28 1269-1273.
37 29 (21) Cháfer, A.; Burguet, M. C.; Monton, J. B.; Lladosa, E., Liquid–liquid equilibria of the
38
30 systems dipropyl ether+n-propanol+water and dipropyl ether+n-propanol+ethylene glycol at
39
40 31 different temperatures. Fluid Phase Equilib. 2007, 262, 76-81.
41 32 (22) Britt, H. I.; Luecke, R. H., The Estimation of Parameters in Nonlinear, Implicit Models.
42 33 Technometrics 1973, 15, 233-247.
43 34 (23) Sun, L. Binary Vapor-liquid Equilibrium of Methyl Glycolate and Ethylene glycol and
44
35 Separation Process for Hydrogenation of Dimethyl Oxalate to Ethylene Glycol. Thesis, East China
45
46 36 University of Science and Technology, Shanghai, 2007.
47 37 (24) Matsuda, H.; Takahara, H.; Fujino, S.; Constantinescu, D.; Kurihara, K.; Tochigi, K.; Ochi,
48 38 K.; Gmehling, J., Fluid Phase Equilibria Selection of entrainers for the separation of the binary
49
39 azeotropic system methanol + dimethyl carbonate by extractive distillation. Fluid Phase Equilib.
50
51 40 2011, 310, 166-181.
52 41 (25) Ma, X.; Liu, X.; Li, Z.; Xu, G. Vapor–liquid equilibria for the ternary system methanol +
53 42 dimethyl carbonate + dimethyl oxalate and constituent binary systems at different temperatures.
54 43 Fluid Phase Equilib. 2004, 221, 51-56.
55
44 (26) Berg, L. Recovery of ethylene glycol from butanediol isomers by azeotropic distillation. U.S.
56
57
22
58
59
60 ACS Paragon Plus Environment
Page 23 of 32 Industrial & Engineering Chemistry Research

1
2
3 1 Patent, US4966658. 1990.
4 2 (27) Wang, Z. Separation Process for the Product of Ethylene Glycol Synthesis from Syngas.
5
3 Thesis, East China University of Science and Technology, Shanghai, 2011.
6
7 4 (28) Li, Q. Process Separation of ethylene glycol from the polyols mixture. Thesis, Jilin
8 5 University, Jilin, 2014.
9 6 (29) Yang, Y., Huang, M., Ma, Xiaoxun. Isobaric vapor-liquid equilibrium(VLE) data of the
10 7 binary system 1,2-Ethanediol+Ethylbenzene at 3.55kPa. J. Northwest University(China), 2017.
11
8 (30) King, C. J. Separaton Processes. McGraw-Hill: New York, 1971.
12
13 9 (31) Dhale A. D.; Myrant L K; Chopade S. P. Propylene glycol and ethylene glycol recovery from
14 10 aqueous solution via reactive distillation. Chem. Eng. Sci. 2004, 59, 2881-2890.
15 11 (32) Huang, W.; Li, H.; Wang, R.; Li, X.; Gao, X. Chemical Engineering & Processing : Process
16
12 Intensification Application of the aldolization reaction in separating the mixture of ethylene glycol
17
18 13 and 1,2-butanediol: Kinetics and reactive distillation. Chem. Eng. Process.: Process
19 14 Intensification 2017, 120, 173-183.
20 15 (33) Dhale, A. D.; Myrant, L. K.; Chopade, S. P.; Jackson, J. E.; Miller, D. J. Propylene glycol and
21 16 ethylene glycol recovery from aqueous solution via reactive distillation. Chem. Eng. Sci. 2004, 59,
22
17 2881-2890.
23
24 18 (34) Chopade, S. P.; Dhale, A. D.; Clark, A. M.; Kiesling, C. W.; Myrant, L. K.; Jackson, J. E.;
25 19 Miller, D. J. Vapor-Liquid-Liquid Equilibrium ( VLLE ) and Vapor Pressure Data for the Systems
26 20 2-Methyl-1,3-dioxolane ( 2MD ) + Water and 2,4-Dimethyl-1,3-dioxolane ( 24DMD )+ Water. J.
27
21 Chem. Eng. Data 2003, 46, 44-47.
28
29 22 (35) Shubham P. Chopade; Atul D. Dhale; Angela M. Clark; Chris W. Kiesling; Laurie K. Myrant;
30 23 James E. Jackson.; Dennis J. Miller. Vapor−Liquid−Liquid Equilibrium (VLLE) and Vapor
31 24 Pressure Data for the Systems 2-Methyl-1,3-dioxolane (2MD) + Water and 2,4
32 25 -Dimethyl-1,3-dioxolane (24DMD) + Water. J. Chem. Eng. Data 2003, 1, 44-47.
33
26 (36) Chien, I Lung, B. Y. Yu, and Z. J. Ai. Design of Azeotropic Distillation Systems. Chemical
34
35 27 Engineering Process Simulation. 2017.
36 28 (37) Luyben, W. L., Comparison of extractive distillation and pressure-swing distillation for
37 29 acetone/chloroform separation. Comput. Chem. Eng. 2013, 50, 1-7.
38
30 (38) Dye, R. F. Ethylene glycols technology. Korean J. Chem. Eng. 2001, 18, 571-579.
39
40 31 (39) Li, X.; Wang, R.; Na, J.; Li, H.; Gao, X., Reversible Reaction-Assisted Intensification
41 32 Process for Separating the Azeotropic Mixture of Ethanediol and 1,2-Butanediol: Reactants
42 33 Screening. Ind. Eng. Chem. Res. 2018, 57, 710-717.
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
23
58
59
60 ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 24 of 32

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 470.5
20
liquid phase
21 vapor phase
22 470.0
23
24
25
26 469.5
27
28
29 469.0
30
31
32 468.5
33
34
35
36
468.0
37
38
39 467.5
40
41
42 467.0
43 0.0 0.2 0.4 0.6 0.8 1.0
44
45 Mass fraction of 1,2-BD
46
47
48
49 Figure 1. Binary phase diagram of the binary system {EG +1,2-BD}.
50
51
52
53
54
55
56
57
58
59
60

ACS Paragon Plus Environment

Page 25 of 32 Industrial & Engineering Chemistry Research

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45 Figure 2. Flowsheet and simulation results of CD route.
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

ACS Paragon Plus Environment

 Industrial & Engineering Chemistry Research Page 26 of 32

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 80 1000
20
21 Rmin
22 70
23 Reboiler Duty
24 800
25 60
26
27
28 50
29
600
30
31 40
32
33 400
34 30
35
36
37
20
38 200
39 10
40
41
42 0 0
43 0.90 0.92 0.94 0.96 0.98
44
45 Recovery of Methanol
46
47
48 Figure 3. Minimum reflux ratio and reboiler heat duty of Col-11 at
49
50 different methanol recovery.
51
52
53
54
55
56
57
58
59
60

ACS Paragon Plus Environment

Page 27 of 32 Industrial & Engineering Chemistry Research

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18 1.00
19
20
21
22 0.98
23
24
25 0.96
26
27
28
29 0.94
30
31
32 0.92
33
34
35
36 0.90 Vapor phase
37
38
Liquid phase
39 0.88
40
41
42
43 0.86
44 0 5 10 15 20 76 77 78 79 80 81 82
45 Stage number
46
47
48
49 Figure 4. The concentration profile of EG in vapor and liquid phase of Col-14.
50
51
52
53
54
55
56
57
58
59
60

ACS Paragon Plus Environment

 Industrial & Engineering Chemistry Research Page 28 of 32

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
)LJXUH )ORZVKHHW DQG VLPXODWLRQ UHVXOWV RI $' URXWH
50
51
52
53
54
55
56
57
58
59
60 ACS Paragon Plus Environment
Page 29 of 32 Industrial & Engineering Chemistry Research

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 70 12
20
21
22
NT,min
23 60 Reboiler duty
24 10
25
26
27 50
28 8
29
30
31 40
32 6
33
34
35 30
36
37 4
38
39 20
40
41 2
42
43 10
44 0.2 0.4 0.6 0.8 1.0
45
46 EB/Feed
47
48
49 Figure 6. Minimum stage number and reboiler heat duty of Col-27 at
50
51
different ratio of EB to feed.
52
53
54
55
56
57
58
59
60

ACS Paragon Plus Environment

 Industrial & Engineering Chemistry Research Page 30 of 32

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49 Figure 7. Flowsheet and simulation results of ED route.
50
51
52
53
54
55
56
57
58
59
60

ACS Paragon Plus Environment

Page 31 of 32 Industrial & Engineering Chemistry Research

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46 Figure 8. Flowsheet and simulation results of RD route.
47
48
49
50
51
52
53
54
55
56
57
58
59
60

ACS Paragon Plus Environment

 Industrial & Engineering Chemistry Research Page 32 of 32

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50 Table of Content Graph
51
52
53
54
55
56
57
58
59
60

ACS Paragon Plus Environment