Thermodynamic investigation of 1,3,5-trioxane, methyl acrylate, methyl

Subscriber access provided by Nottingham Trent University

Thermodynamics, Transport, and Fluid Mechanics

Thermodynamic investigation of 1,3,5-trioxane, methyl acrylate, methyl acetate and water mixtures in terms of NRTL and UNIQUAC models Hui Zhao, Jie Li, Lei Wang, Chunshan Li, and Ping Li Ind. Eng. Chem. Res., Just Accepted Manuscript • Publication Date (Web): 04 Sep 2019 Downloaded from pubs.acs.org on September 4, 2019

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.

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 20 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 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Thermodynamic investigation of 1,3,5-trioxane, methyl acrylate, methyl acetate and water mixtures in terms of NRTL and UNIQUAC models Hui Zhao a,b, Jie Li b, Lei Wang b,c, Chunshan Lib,c,* and Ping Lid a University b State

of Chinese Academy of Sciences, Beijing 100049, China

Key Laboratory of Multiphase Complex Systems, Beijing Key Laboratory of Ionic Liquids Clean Process,

Zhongke Langfang Institute of Process Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China c d

Innovation Academy for Green Manufacture, Chinese Academy of Sciences, Beijing 100190, China

State Key Laboratory of High-efficiency Utilization of Coal and Green Chemical Engineering, School of Chemistry

and Chemical Engineering, Ningxia University, Yinchuan, 750021, PR China

Abstract The thermodynamics of 1,3,5-trioxane, methyl acrylate, methyl acetate and water were systematically investigated in terms of NRTL and UNIQUAC models. Solubility data of 1,3,5-trioxane in methyl acetate and methyl acrylate was reported. VLE dataof 1,3,5-trioxane + methyl acetate and 1,3,5-trioxane + methyl acrylate binary systems were measured. The LLE of 1,3,5-trioxane + methyl acrylate + water and 1,3,5-trioxane + methyl acrylate + methyl acetate + water mixtures were investigated. The thermodynamic consistency of VLE and LLE experimental data were tested by using Van Ness point method and Othmer-Tobias method respectively. Binary interaction parameters for VLE and LLE systems were regressed and compared according to AAD and AAD% values. Proper model and parameters were applied in the designed separation process of the quaternary mixture. Keywords: Methyl acrylate; Methyl acetate; 1,3,5-Trioxane; Phase equilibria 1. Introduction Methyl acrylate (MA), which is widely used in the production of paintings, adhesives and leather modification,1,2 is synthesized by the two-step oxidation of propylene.3,4 Recently, due to the rising price of raw materials and the necessity of environmental friendly, our lab has been devoted to develop

ACS Paragon Plus Environment

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 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 20

a new alternative route, so the aldol condensation of methyl acetate (MAc) with 1,3,5-trioxane is proposed.5,6 To separate MA from other products and recycle raw materials, reliable phase equilibrium data is of great importance to be obtained to design advisable separation process. The thermodynamics of multiphase equilibria comes down to the interaction between components, lots of studies have been published in the field. In 1995, Martin and Mato measured the vapor-liquid equilibrium (VLE) for MAc, methanol (Me) and water ternary system by an ebulliometric method at 101.3 kPa, and fitted the results with activity coefficient models.7 Tu et al. reported the VLE for the Me, MAc and MA system at atmospheric pressure, and the experimental data of three binaries fulfilled the thermodynamic requirements adequately via Herington’s test.8 Many other works have also been done to investigate the phase equilibrium for the mixtures of MAc, Me, water and MA, which is not only physical equilibrium, but also chemical equilibrium.9-11 However, due to the partial miscibility in mixtures, two liquid phases may coexist, the data for VLLE and liquid-liquid equilibrium (LLE) are indispensable.12,13 Zuo et al. studied the VLE and LLE for MAc, Me, MA and water systems, introduced the thermodynamic behavior of MA and water mixture in detail, finally established a separation process for quaternary acrylic systems.14 While studies on the thermodynamics of system that containing 1,3,5-trioxane were almost in aqueous solution. Brandani et al. reported the isothermal VLE data for Water + 1,3,5-Trioxane system and found a positive azeotrope under pressure of 93.3 kPa.15 In the same year, they studied the isothermal VLE and solubility of the system methanol and 1,3,5-trioxane.16 Albert et al. added new experimental results for VLE of 1,3,5-Trioxane + Water binary system and Formaldehyde + 1,,3,5-Trioxane + Water ternary system.17 From the above, the phase equilibrium data for the mixture of MAc, MA, 1,3,5-trioxane and water is required. In the present work, solubility data of 1,3,5-trioxane in MAc and MA was reported. Isobaric VLE data for binary systems (MAc + 1,3,5-Trioxane, MA + 1,3,5-Trioxane) were measured and the thermodynamic consistency was tested using Van Ness point method. The LLE data for ternary and quaternary systems were determined at isothermal conditions (313.15 K and 323.15 K) and correlated with Othmer-Tobias equation to test quality. Interaction parameters for NRTL and UNIQUAC equations were regressed in Aspen Plus. Proper activity coefficient model and interaction parameters were obtained by comparing the average absolute deviations between experimental and estimated data.


Page 3 of 20 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 51 52 53 54 55 56 57 58 59 60


The optimum parameters were applied in a designed separation process for the quaternary mixture and showed a very well products purity. 2. Experiment 2.1 Materials The chemicals in this work were purchased and used as received. The specification of chemical samples is shown in Table 1 and the purity is as stated by the supplier. MA was inhibited with 10-4 g/g 4-methoxyphenol in order to avoid polymerization. The water was from the deionization setup in our lab. Table 1. Specification and boiling point (Tb) of chemicals. Tb/K

Chemical

CAS

Supplier

Mass content

methyl acetate

79-20-9

Damao Chemical Reagent Factory

≥ 0.980

330.20

methyl acrylate

96-33-3

Sinopharm Chemical Reagent Co.,Ltd

≥ 0.985

353.64

1,3,5-trioxane

110-88-3


≥ 0.980

387.57

1,4-dioxane

123-91-1

Xilong Chemical Co., Ltd

≥ 0.995

4-methoxyphenol

150-76-5


≥ 0.990

water

7732-18-5

deionization setup

This work

Literature 330.0221 330.2033 353.3522 352.3614 387.6518 388.0634

2.2 Apparatus and procedure SLE Measurement. The solubility experiment of 1,3,5-trioxane in MAc and MA was conducted in a stoppered flask thermostated by water bath. The mixtures of MAc/MA and excess 1,3,5-trioxane were stirred and then equilibrated at several temperature points. After 2 h, the clear solutions were sampled and analyzed. VLE Measurement. As shown in Figure 1a, an improved Rose-Williams still (Zhejiang University) with a provision for both vapor and liquid recirculation was used for the VLE experiments.



Page 4 of 20

The reliability of the equipment has been proved in the early work.14 The cyclic process of liquid and vapor phase followed the red and blue arrows in the diagram respectively. Before heating, a certain proportion of binary (MAc + 1,3,5-Trioxane, MA + 1,3,5-Trioxane) mixtures were introduced into the VLE still. The equilibrium state was reached in about 1-2 h when the boiling temperature remained unchanged. The temperature was measured using a mercury thermometer with an accuracy of ± 0.1 K. 0.5 h after reaching the equilibrium, the liquid and condensed vapor phases were sampled, recording the equilibrium temperature at the same time. The vapor phase was condensed in two steps due to the solubility of 1,3,5-trioxane in liquid has the maximum. Most 1,3,5-trioxane and a small part of MAc was condensed in the first step by circulating hot water, , then the remaining MAc was totally condensed by tap water in the second step. The temperature of hot water was controlled between 318.15 K and 328.15 K. LLE Measurement. The liquid-liquid equilibrium experiments were carried out in the LLE still (Zhejiang University) shown in Figure 1b. The temperature was kept by circulation thermostat (DC2006, Ningbo Scientz Biotechnology Co.,LTD.). All the experiments were carried out at 101.3 ± 0.1 kPa. The mixture to be measured was fully stirred in the LLE still and then stood at a constant temperature for layered. 1 h later, the upper and under layers were sampled for analysis.

Figure 1. Schematic diagrams of (a) VLE and (b) LLE still.


Page 5 of 20 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 51 52 53 54 55 56 57 58 59 60


Sample Analysis. All the vapor and liquid samples were analyzed by using gas chromatography (GC-2010 Plus, Shimadzu) with 1,4-dioxane as the internal standard. Special standard curves were established for each phase equilibrium system with the accuracy of > 0.99. The column type was RTXWax (30 m × 250 mm) with helium as the carrier gas. The temperature of gasification and BID detector was 393.15 K and 453.15 K respectively. The experimental data was repeated two times. The experimental data of composition in liquid and vapor phase were denoted as x and y respectively, mole fraction. The composition in oil and water phase for LLE system were written as o and w respectively, mole fraction. 2.3 Thermodynamic modeling The NRTL and UNIQUAC models were used to correlate the experimental data for the systems containing 1,3,5-trioxane, MA, MAc and water. The NRTL model was first proposed by Renon and Prausnitz in 1968, the equation is based on Scott’s two-liquid model and an assumption of nonrandomness.19 The vapor phase is considered as ideal gas under ambient pressure and the excess Gibbs energy gE is used to express the nonideality of liquid mixture.20 𝑔𝐸 = 𝑅𝑇

𝑚

𝑚

∑𝑗 = 1𝜏𝑗𝑖𝐺𝑗𝑖𝑥𝑗

∑𝑥 ∑ 𝑖

𝑖=1

𝑚 𝐺 𝑥 𝑘 = 1 𝑘𝑖 𝑘

The activity coefficient for component i is: N

ln  i 

 j 1 N

ji

G ji x j

G k 1

x

ki k

N    lj Glj xl    N xG    N j ij  ij  l N1   j 1 Gkj xk  Gkj xk    k 1 k 1  

where 𝜏𝑗𝑖 = 𝑎𝑖𝑗 + 𝑏𝑖𝑗 𝑅𝑇, 𝜏𝑖𝑖 = 0, 𝐺𝑗𝑖 = exp ( ― 𝛼𝑗𝑖𝜏𝑗𝑖), 𝛼𝑖𝑗 = 𝛼𝑗𝑖. 𝑎𝑖𝑗 and 𝑏𝑖𝑗 are binary interaction parameters for NRTL model. The UNIQUAC model is applied for highly non-ideal chemical systems, and used for VLE and LLE regression. The excess Gibbs energy gE is determined by molecule size and structure, and intermolecular forces. The equation for the UNIQUAC model is:



ln  i  ln

i

  z  qi ln i  li  i i xi 2 xi

x l

j j

j

Page 6 of 20

    qi  qi ln  j ji   qi  j ij j   k kj  j  k

where 𝜑𝑖 =

𝑧 𝑥𝑖𝑟𝑖 ∑ ( 𝑥𝑞 𝑥 𝑟 ), 𝜃𝑖 = 𝑖 𝑖 ∑𝑗(𝑥𝑗𝑞𝑗), 𝑙𝑖 = 2(𝑟𝑖 ― 𝑞𝑖) +1 ― 𝑟𝑖, 𝜏𝑖𝑗 = 𝑒𝑥𝑝(𝑎𝑖𝑗 + 𝑏𝑖𝑗 𝑇). r, q 𝑗 𝑗 𝑗

are volume and surface parameter respectively. The values of r and q of the four components were listed in Table 2. 𝑎𝑖𝑗 and 𝑏𝑖𝑗 are binary interaction parameters determined by experimental data. The parameters were estimated using the Maximum likelihood objective function. The function minimized by data regression can be stated as follows. 𝑁

Q=

[

∑(

𝑇𝑒𝑥𝑝 ― 𝑇𝑒𝑠𝑡 𝑖 𝑖 𝜎𝑇

𝑖=1

2

) ( +

𝑃𝑒𝑥𝑝 ― 𝑃𝑒𝑠𝑡 𝑖 𝑖

2

) (

𝜎𝑃

+

𝑒𝑠𝑡 𝑥𝑒𝑥𝑝 1,𝑖 ― 𝑥1,𝑖

2

) (

𝜎𝑥

+

𝜎𝑦

2

)]

𝑒𝑠𝑡 𝑦𝑒𝑥𝑝 1,𝑖 ― 𝑦1,𝑖

where σ is the standard deviation of the indicated data. Table 2. Values of parameters r and q of componentsa.

a

Parameters

1,3,5-Trioxane

MA

MAc

Water

r

2.75

3.25

2.80

0.92

q

2.34

2.90

2.58

1.40

Taken from Aspen property databank, Aspen Plus V8.4.

Deviation. The deviation between experimental and estimated data is presented as Average Absolute Deviation (AAD) and AAD in percentage (AAD%). The deviation value is calculated according to the following equations, c is the mole fraction of component. 𝑁

AAD = (1 𝑁)

∑ |𝑐

𝑒𝑥𝑝 𝑖

― 𝑐𝑒𝑠𝑡 𝑖 |

𝑖=1 𝑁

AAD% = (100/𝑁)

∑|

𝑖=1

|

𝑐𝑒𝑥𝑝 ― 𝑐𝑒𝑠𝑡 𝑖 𝑖 𝑐𝑒𝑥𝑝 𝑖

3. Results and discussion 3.1 Solid-liquid equilibrium 1,3,5-Trioxane, as a heterocyclic solid compound, has the maximum solubility in MAc and MA.


Page 7 of 20

The mixture used for experiments should be prepared within the allowable range of solubility under ambient conditions. The solubility data as a function of temperature was shown in Table S1 (Supporting Information) and plotted in Figure 2. The solubility of 1,3,5-trioxane in MAc was more than in MA at the same temperature. With the increase of temperature, the solubility of 1,3,5-trioxane in MA increased faster than in MAc. The SLE curves in Figure 2 tended to intersect. This is because when the mole fraction of 1,3,5—trioxane was 1.0, the temperature would be the melting point of 1,3,5-trioxane (337.15 K)18, at which the substance changed its state from solid to liquid.

330 325 320

Temperature / K

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 51 52 53 54 55 56 57 58 59 60


315 310 305

1,3,5-Trioxane in MA 1,3,5-Trioxane in MAc

300 295 0.5

0.6

0.7

0.8

0.9

Mole fraction of 1,3,5-Trioxane Figure 2. Solubility diagram of 1,3,5-trioxane in MAc/MA.

3.2 Vapor-liquid equilibrium The mole fraction of 1,3,5-trioxane in MAc is less than 0.5 at ambient temperature (297.15 K), and the value in MA is less than 0.46. In consequence, the VLE of binary systems were investigated among the mixture in which the mole fraction of 1,3,5-trioxane was less than 0.5. The experimental data of 1,3,5-Trioxane + MAc and 1,3,5-Trioxane + MA systems were listed in Table S2 and S3 respectively. To obtain the binary interaction parameters, the experimental data were regressed with NRTL and UNIQUAC models. The estimated data were shown besides the experimental data. The regressed interaction parameters were used to predict the binary T-x-y diagrams in the whole



concentration, which were shown in Figure 3 (solid line). From the binary diagrams, the experimental data points fit well with the blue curves in both systems, indicating the NRTL model performed better in data regression. The AAD and AAD% in Table 2 were used to present the accuracy of thermodynamic models quantitatively. The regressed data for liquid phase were more accurate than that for vapor phase.

(a)

Temperature / K

380

V

370

360

V-L

L

350

L Phase V Phase NRTL UNIQUAC

340

330 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.9

1.0

Mole fraction of 1,3,5-Trioxane

(b)

385 380

V

Temperature / K

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 51 52 53 54 55 56 57 58 59 60

Page 8 of 20

375

L

V-L

370 365

L Phase V Phase NRTL UNIQUAC

360 355 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Mole fraction of Trioxane

Figure 3. Binary VLE diagrams of (a) 1,3,5-Trioxane + MAc and (b) 1,3,5-Trioxane + MA systems.


Page 9 of 20 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 51 52 53 54 55 56 57 58 59 60


Table 2. AAD and AAD% values for experimental data with estimated data of binary VLE systems. Model

AAD

AAD%

T/K

y1

T

y1

VLE for 1,3,5-Trioxane(1) + MAc(2) NRTL

0.12

0.0024

0.04

10.08

UNIQUAC

0.99

0.0039

0.29

0.41

VLE for 1,3,5-Trioxane(1) + MA(2) NRTL

0.22

0.0037

0.26

2.42

UNIQUAC

0.82

0.0207

0.23

16.23

There are two methods for the thermodynamic consistency test of VLE experimental data, Herington area test and Van Ness point test. The Herington area test is a semiempirical method, requiring the experimental data cover the whole concentration range from 0 to 1.23 So the area test is not suitable in this work. The Van Ness point test was put forward by van Ness and modified by Fredenslund et al.24-25 It was used to verify the reliability of experimental data.26-28 The equation is shown as the following: 1 ∆y = 𝑁 1 ∆P = 𝑁

𝑁

∑100|𝑦

𝑒𝑥𝑝 𝑖

― 𝑦𝑒𝑠𝑡 𝑖 |

𝑖=1 𝑁

∑

100

𝑖=1

|

|

𝑃𝑒𝑥𝑝 ― 𝑃𝑒𝑠𝑡 𝑖 𝑖 𝑃𝑒𝑥𝑝 𝑖

where N is the number of experiment points, P is the pressure. The experimental data pass the thermodynamic consistency test when ∆y and ∆P < 1. The results of Van Ness point test for the two VLE binary systems were provided in Table 3. The experimental data of 1,3,5-Trioxane + MAc system passed the test both in NRTL and UNIQUAC model. However, the experimental data of 1,3,5Trioxane+MA system did not pass the Van Ness test in UNIQUAC model. According to the Van Ness equation, the estimated data of thermodynamic models were considered. The deviation of regression will influence the results of point test method. From AAD and AAD%, the UNIQUAC model did not correlate the experimental data well. In conclusion, the experimental data of the two VLE binary systems passed the thermodynamic consistency test according to the results performed by NRTL model.



Page 10 of 20

Table 3. The results of Van Ness test for VLE binary systems. Systems 1,3,5-Trioxane(1)+MA(2) ∆𝑦1

NRTL

Results

UNIQUAC

Results

0.3653 0.0565

Passed Passed

2.0719 0.2840

Failed Passed

0.2407 0.1251

Passed Passed

0.3864 0.3180

Passed Passed

∆P 1,3,5-Trioxane(1)+MAc(2) ∆𝑦1 ∆P

Due to the NRTL model correlated the experimental data points well in both binary systems, the interaction parameters of NRTL model were collected in Table 4. The estimated boiling points of 1,3,5-trioxane, MA and MAc by regressed NRTL parameters were summarized in Table 1. The estimated boiling points are close to the literature value, indicating the regressed parameters are suitable for further application.

Table 4. NRTL parameters for 1,3,5-Trioxane(1), MA(2) and MAc(3) mixtures. i-j

aij

aji

bij/K

bji/K

αij

1-2 1-3

2.51 -3.37

31.02 58.70

-1488.11 555.58

-10000.00 -18662.01

0.3 0.3

3.3 Liquid-liquid equilibrium In the MA production process, equal mole of water was produced with MA. The mixture of 1,3,5trioxane, MA, MAc and water will separate into two liquid layers. Organic components of lower density concentrate in the upper layer called oil phase, the under layer is water with a small quantity of other compounds called water phase. The mixtures of 1,3,5-trioxane, MA and water ternary system were prepared to investigate the liquid-liquid equilibrium. The composition of oil and water phases measured at 313.15 K and 323.15 K were shown in Table S4. The experimental data quality was determined by Othmer-Tobias equation.29 The equation formula is shown as the following: ln

(

1 ― 𝑚𝑜𝑀𝐴 𝑚𝑜𝑀𝐴

)

(

= A + Bln

1 ― 𝑚𝑤 𝐻2𝑂 𝑚𝑤 𝐻2𝑂


)

Page 11 of 20 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 51 52 53 54 55 56 57 58 59 60


where 𝑚𝑜𝑀𝐴 is mass fraction of MA in oil phase, 𝑚𝑤 𝐻2𝑂 is mass fraction of water in water phase. In many ternary liquid systems, a straight line obtained from the equation indicates the reliability of experimental data.30-32 Parameters A and B are correlated with experimental data and square correlation coefficient R2 is used to estimate the quality of data points. The value of A, B and R2 at different temperatures were shown in Table 5. The values of R2 were greater than 0.95, indicating the good linearity of equation and the reliability of experimental data.

Table 5. Othmer-Tobias equation parameters. System Ternary Quaternary

T/K

R2

A

B

313.15

0.98

1.87

1.69

323.15

0.99

1.57

1.53

313.15

0.95

2.21

2.13

323.15

0.98

0.78

1.17

The experimental data were regressed using NRTL and UNIQUAC models (Table S5). The estimated data were plotted in Figure 4 with scatter diagram of experimental data. The increase of temperature did not cause much change in solubility. In addition, the AAD and AAD% value in Table 6 show that NRTL and UNIQUAC model both performed well in the LLE correlation. 0.00

1,3,5-Trioxane

O Phase W Phase UNIQUAC NRTL

1.00

(a) 0.25

0.75

0.50

0.50

0.75

0.25

1.00

Water

0.00

0.25

0.50

0.75

0.00 1.00 MA



0.00

(b)

1,3,5-Trioxane

Page 12 of 20

O Phase W Phase UNIQUAC NRTL

1.00

0.25

0.75

0.50

0.50

0.75

0.25

1.00

Water

0.00

0.25

0.50

0.75

0.00 1.00 MA

Figure 4. LLE diagrams of 1,3,5-Trioxane+MA+Water system at 313.15 K (a) and 323.15 K (b).

Table 6. AAD and AAD% for experimental data with estimated data of 1,3,5-Trioxane(1) + MA(2) + Water(3) LLE system at 313.15 K and 323.15 K. Oil Phase Model

NRTL 313.15 K UNIQUAC

NRTL 323.15 K UNIQUAC

Water Phase

Deviation o1

o2

o3

w1

w2

w3

AAD

0.0046

0.0077

0.0095

0.0005

0.0013

0.0016

AAD%

1.59

1.30

5.28

1.25

9.81

0.18

AAD

0.0037

0.0153

0.0177

0.0004

0.0011

0.0012

AAD%

1.23

2.97

9.52

1.19

8.33

0.13

AAD

0.0145

0.0057

0.0181

0.0008

0.0012

0.0013

AAD%

6.75

1.19

8.60

3.92

7.96

0.15

AAD

0.0138

0.0128

0.0254

0.0010

0.0011

0.0007

AAD%

6.19

2.48

12.75

3.49

7.17

0.08

The liquid-liquid equilibrium of 1,3,5-trioxane, MA, MAc and water were also investigated at 313.15 K and 323.15 K. The quality of experimental data (Table 7) was measured by Othmer-Tobias equation. The value of 𝑚𝑜𝑀𝐴 in the equation was replaced with 𝑚𝑜𝑀𝐴 + 𝑀𝐴𝑐, the mass fraction of MA and MAc in oil phase. The correlated equation parameters were listed in Table 6, line “Quaternary”. The experimental data of LLE quaternary system was reliable according to the value of R2. The estimated data were supplied in Table S6 (NRTL) and S7 (UNIQUAC). The AAD and AAD% value between experimental and regressed data were shown in Table 7, indicating the NRTL model correlates the experimental data well in quaternary liquid-liquid equilibrium system. The regressed


Page 13 of 20 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 51 52 53 54 55 56 57 58 59 60


NRTL parameters were concluded in Table 8 for further application. Table 7. Experimental LLE data, AAD and AAD% for 1,3,5-Trioxane(1) + MA(2) + MAc(3) + Water(4) quaternary system at 101.3 kPaa. Oil phase/mol∙mol-1

T/K

o1

o2

o3

Water phase/mol∙mol-1 o4

w1

w2

w3

w4

0.1566 0.2972 0.4612 0.0850 0.0160 0.0066 0.0228 0.9546 0.1984 0.2439 0.3773 0.1804 0.0212 0.0065 0.0219 0.9504 0.1834 0.2772 0.3557 0.1837 0.0224 0.0070 0.0204 0.9502 313.15 0.2597 0.1628 0.2498 0.3277 0.0382 0.0069 0.0223 0.9326 0.2753 0.1883 0.2599 0.2765 0.0375 0.0073 0.0216 0.9336 0.2521 0.2001 0.2781 0.2697 0.0321 0.0070 0.0204 0.9405 AAD-NRTL 0.0061 0.0017 0.0059 0.0122 0.0005 0.0001 0.0004 0.0008 AAD%-NRTL 2.99 0.80 1.74 7.78 2.02 1.72 1.76 0.09 AAD-UNIQUAC 0.0066 0.0040 0.0079 0.0155 0.0006 0.0001 0.0004 0.0010 AAD%-UNIQUAC 3.15 1.71 2.25 11.08 2.18 1.53 2.08 0.11 a The standard uncertainties are u(P)=0.1 kPa, u(T)=0.1 K, u(o )=0.0057, u(o )=0.0062, u(o )=0.0093, 1 2 3 u(o4)=0.0211, u(w1)= u(w2)=0.0001, u(w3)=0.0002, u(w4)=0.0004. 0.2389 0.1837 0.2635 0.3139 0.0339 0.0082 0.0225 0.9354 0.1708 0.2272 0.3076 0.2944 0.0220 0.0081 0.0215 0.9484 0.2478 0.1534 0.2348 0.3640 0.0412 0.0091 0.0254 0.9243 323.15 0.2557 0.1705 0.2477 0.3261 0.0392 0.0088 0.0237 0.9283 0.2462 0.1933 0.2687 0.2918 0.0335 0.0085 0.0221 0.9359 0.2555 0.1346 0.2220 0.3879 0.0465 0.0088 0.0261 0.9186 AAD-NRTL 0.0020 0.0018 0.0015 0.0016 0.0003 0.0002 0.0003 0.0006 AAD%-NRTL 1.02 0.91 0.60 0.52 0.89 1.82 1.50 0.07 AAD-UNIQUAC 0.0088 0.0037 0.0066 0.0190 0.0004 0.0003 0.0007 0.0010 AAD%-UNIQUAC 3.91 1.77 2.33 6.24 1.06 3.20 3.15 0.11 a The standard uncertainties are u(P)=0.1 kPa, u(T)=0.1 K, u(o )=0.0033, u(o )=0.0031, u(o )=0.0067, 1 2 3 u(o4)=0.0133, u(w1)=0.0003, u(w2)=0.0001, u(w3)=0.0003, u(w4)=0.0006. Table 8. NRTL parameters for 1,3,5-Trioxane(1) + MA(2) + MAc(3) + Water(4) LLE system. Model aij

i-j 1-2

1-3

1-4

2-3

2-4

3-4

-21.53

-17.69

24.33

-29.78

-0.62

38.95



aji bij/K bji/K

33.14 6541.66 -10000.00

-23.36 5164.93 10000.00

-9.19 -7527.10 3577.18

-9.41 10000.00 2643.50

Page 14 of 20

-7.34 1096.04 3509.04

7.06 -10000.00 -1167.41

3.4 Separation process simulation In order to refine MA and recycle MAc from the mixture, a separation process was designed. The regressed VLE and LLE parameters were used for distillation and decantation processes respectively. The interaction parameters for 1,3,5-trioxane + water, MAc + water, MA + water and MA + MAc systems were also needed in the process simulation, they were referenced to the reported literatures and Aspen database (Table 9). The designed separation process was shown in Figure 5.

Table 9. Referenced NRTL parameters for binary systems. Binary system

aij

aij

bij/K

bij/K

αij

Source

1,3,5-Trioxane+Water MAc+Water MA+Water MA+MAc

0 -2.91 -24.82 -0.25

0 3.53 18.92 -0.94

120.44 1240.36 9440.76 -20.12

830.50 -308.60 10866 423.09

0.3 0.3 0.3 0.3

Aspen Database Aspen Database Literature 14 Literature 8

Figure 5. Separation flow scheme for 1,3,5-Trioxane+MA+MAc+Water quaternary system.

The first column C101 is distillation for MAc recovery. The exchange of material on each stage improved the concentration of light component and made the steam that reached the top of the tower contain light component of high purity. The mass concentration of MAc distillated from the top was 99.4%. The bottom liquid was introduced into decanter V104 to remove water by layering. Due to the decrease of mutual solubility under low temperature, the liquid-liquid separation was conducted at 303.15 K and 1,3,5-trioxane enriched in organic phase. The second column C102 is distillation for


Page 15 of 20 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 51 52 53 54 55 56 57 58 59 60


second dehydration. The azeotrope of MA-Water mixture was distillated from the top and the condensed liquid separated into two phases in V105, the organic phase went back to the column and enriched with 1,3,5-trioxane together at the bottom of the tower. The last column C103 is distillation for MA refining and 1,3,5-trioxane recovery. The purity of MA at the top and 1,3,5-trioxane at the bottom were both 99.5%, reaching the level of commodity. The total flow rate of quaternary mixture was 1000 kg/h. The operation condition of each column, decanter and the information of all streams were supplied in Table S8 and S9. The results of separation process indicate the regressed parameters are suitable for the simulation and can express the interaction between components well. 3. Conclusions

The solubility of 1,3,5-trioxane in MA/MAc at normal condition is less than 0.46 and 0.5 mol·mol1

respectively. The isobaric VLE experimental data of 1,3,5-Trioxane + MA and 1,3,5-Trioxane +

MAc binary systems were measured in the range of 1,3,5-trioxane solubility. The LLE experimental data were obtained for 1,3,5-Trioxane + MA + Water ternary system and 1,3,5-Trioxane + MA + MAc + Water quaternary system at 313.15 K and 323.15 K. Van Ness point method and Othmer-Tobias method were applied to verify the reliability of VLE and LLE experimental data respectively. The data were in good quality and reliable according to the testing results. The experimental data were regressed by NRTL and UNIQUAC models and compared with estimated data. The NRTL model performed well in the correlation and the parameters for VLE and LLE systems were obtained. The VLE parameters were applied to predict the binary phase diagram in the whole range and the estimated boiling point fitted well with the reported data. The designed process was carried out using regressed VLE and LLE parameters for distillation and decantation respectively. The purity of refined MA and 1,3,5-trioxane was 99.5%, the distillated MAc was 99.4% meeting the requirement of recycling. Associated Content Supporting Information Table S1. Solubility of 1,3,5-trioxane (solid) in MAc/MA at 101.3 kPa; Table S2. Experimental and estimated VLE data for 1,3,5-Trioxane(1)+MAc(2) binary system at 101.3 kPa; Table S3. Experimental and estimated VLE data for 1,3,5-Trioxane(1)+MA(2) binary system at 101.3 kPa; Table



Page 16 of 20

S4. Experimental LLE data for 1,3,5-Trioxane(1)+MA(2)+Water(3) ternary system at 101.3 kPa; Table S5. Estimated LLE data for 1,3,5-Trioxane(1)+MA(2)+Water(3) ternary system; Table S6. Estimated data for 1,3,5-Trioxane(1)+MA(2)+MAc(3)+Water(4) quaternary system by NRTL model; Table S7. Estimated data for 1,3,5-Trioxane(1)+MA(2)+MAc(3)+Water(4) quaternary system by UNIQUAC model; Table S8. Operation condition of distillation columns and decanters; Table S9. Streams information of separation process; Figure S1. Separation flow scheme for 1,3,5Trioxane+MA+MAc+Water quaternary system with stream number. Author Information Corresponding Author *Email: [email protected] Acknowledgments The work was supported by the National Natural Science Funds (No. 21878293, No. 21576261), Key Research Program of Frontier Sciences, CAS (No. QYZDB-SSW-SLH022) and K.C.Wong Education Foundation (No. GJTD-2018-04).

References (1) Xu, X. B.; Lin, J. P.; Cen, P. L. Advances in the research and development of acrylic acid production from biomass. Chinese J. Chem. Eng. 2006, 14, 419-427. (2) Straathof, A. J. J.; Sie, S.; Franco, T. T.; van der Wielen, L. A. M.

Feasibility of acrylic acid

production by fermentation. Appl. Microbiol. Biotechnol. 2005, 67, 727-734. (3) Vitry, D.; Morikawaa, Y.; Duboisb, J.L.; Ueda, W. Mo-V-Te-(Nb)-O mixed metal oxides prepared by hydrothermal synthesis for catalytic selective oxidations of propane and propene to acrylic acid. Appl. Catal. A: Gen. 2003, 251, 411-424. (4) Kum, S. S.; Jo, B. Y.; Moon, S. H. Performance of Pd-promoted Mo-V-Te-Nb-O catalysts in the partial oxidation of propane to acrylic acid. Appl. Catal. A: Gen. 2009, 365, 79-87. (5) Yang, D.; Li, D.; Zhang, G. L.; Jiao, T. T.; Li, Z. X.; Li, C. S.; Zhang, S. J. Reaction of formalin with acetic acid over vanadium-phosphorus oxide bifunctional catalyst. Ind. Eng. Chem. Res. 2015, 54, 6865-6873.


Page 17 of 20 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 51 52 53 54 55 56 57 58 59 60


(6) Zhang, G. L.; Zhang, H. H.; Yang, D.; Li, C. S.; Peng, Z. P.; Zhang, S. J. Catalysts, kinetics and process optimization for the synthesis of methyl acrylate over Cs-P/γ-Al2O3. Catal. Sci. Tech. 2016, 6, 6417-6430. (7) Maria, C.; Martin, M. C.; Mato, R. B. Isobaric vapor-liquid equilibrium for methyl acetate + methanol + water at 101.3 kPa. J. Chem. Eng. Data 1995, 40, 326-327. (8) Tu, C. H.; Wu, Y. S.; Liu T. L. Isobaric vapor-liquid equilibria of the methanol, methyl acetate and methyl acrylate system at atmospheric pressure. Fluid Phase Equilibr. 1997, 135, 97-108. (9) Bernatová, S.; Aim, K.; Wichterle, I. Isothermal vapour-liquid equilibrium with chemical reaction in the quaternary water + methanol + acetic acid + methyl acetate system, and in five binary subsystems. Fluid Phase Equilibr. 2006, 247, 96-101. (10) Calvar, N.; Domĭnguez, A.; Tojo, J. Vapor-liquid equilibria for the quaternary reactive system ethyl acetate + ethanol + water + acetic acid and some of the constituent binary systems at 101.3 kPa. Fluid Phase Equilibr. 2005, 235, 215-222. (11) Dirk-Faitakis, C. B.; An, W. Z.; Lin, T. B.; Chuang, K. T. Catalytic distillation for simultaneous hydrolysis of methyl acetate and etherification of methanol. Chem. Eng. Process 2009, 48, 1080-1087. (12) Li, C. S.; Wozny, G.; Suzuki, K. Design and synthesis of separation process based on a hybrid method. Asia-Pac J. Chem. Eng. 2009, 4, 905-915. (13) Jiao T. T.; Zhuang, X. L.; He, H. Y.; Zhao, L. H.; Li, C. S.; Chen, H. N.; Zhang, S. J. An ionic liquid extraction process for the separation of indole from wash oil. Green Chem. 2015, 17, 37833790. (14) Zuo, C. C.; Li, Y. P.; Li, C. S.; Cao, S. S.; Yao, H. Y.; Zhang, S. J. Thermodynamics and separation process for quaternary acrylic systems. AIChE J. 2016, 62, 228-240. (15) Brandani, S.; Brandani, V.; Flammini, D. Isothermal vapor-liquid equilibria for the water-1,3,5trioxane system. J. Chem. Eng. Data 1994, 39, 184-185. (16) Brandani, S.; Brandani, V.; Isothermal vapor-liquid equilibria and solubility in the system methanol + 1,3,5-trioxane. J. Chem. Eng. Data 1994, 39, 203-204. (17) Albert, M.; Hasse, H.; Kuhnert, C.; Maurer, G. New experimental results for the vapor-liquid equilibrium of the binary system (trioxane + water) and the ternary system (formaldehyde + trioxane + water). J. Chem. Eng. Data 2005, 50, 1218-1223. (18) Zheng, C. Q. Concise Materials Dictionary. China, 2002, pp. 320.



Page 18 of 20

(19) Orye, R. V.; Prausnitz, J. M. Multicomponent equilibria with the Wilson equation. Ind. Eng. Chem. 1965, 57, 18-26. (20) Scatchard, G. General discussion. Trans. Faraday Soc. 1937, 33, 156-159. (21) Zhang, X. M.; Liu, H. P.; Liu, Y. X.; Jian, C. G.; Wang, W. Experimental isobaric vapor-liquid equilibrium for the binary and ternary systems with methanol, methyl acetate and dimethyl sulfoxide at 101.3 kPa. Fluid Phase Equilibr. 2016, 408, 52-57. (22) Stull, D. R. Vapor pressure of pure substances organic compounds. Ind. Eng. Chem. 1947, 39, 517-540. (23) Wisniak, J. The Herington test for thermodynamic consistency. Ind. Eng. Chem. Res. 1994, 33, 177-180. (24) van Ness, H. C.; Byer, S. M.; Gibbs, R. E. Vapour-liquid equilibrium: Part I An appraisal of data reduction methods. AIChE J. 1973, 19, 238-244. (25) Fredenslund, A.; Gmehling, J.; Rasmussen, P. Vapor-liquid equilibria using UNIFAC: A group contribution method, Elsevier, Amsterdam, 1977. (26) Wu, J. Y.; Xu, D. M.; Shi, P. Y.; Gao, J.; Zhang, L. Z.; Ma, Y. X.; Wang. Y. L. Separation of azeotrope (allyl alcohol + water): Isobaric vapour-liquid phase equilibrium measurements and extractive distillation. J. Chem. Thermodynamics 2018, 118, 139-146. (27) Zhang, Y.; Liu, K.; Wang, Z. J.; Gao, J.; Zhang, L. Z.; Xu, D. M.; Wang, Y. L. Vapour-liquid equilibrium and extractive distillation for separation of azeotrope isopropyl alcohol and diisopropyl ether. J. Chem. Thermodynamics 2019, 131, 294-302. (28) Zhang, Y.; Diao, B. T.; Xu, D. M.; Jiang, H. H.; Zhang. L. Z.; Gao, J.; Wang, Y. L. Separation of the mixture (isopropyl alcohol + diisopropyl ether + n-propanol): Entrainer selection, interaction exploration and vapour-liquid equilibrium measurements. J. Chem. Thermodynamics 2019, 135, 2734. (29) Othmer, D. F.; Tobias, P. E. Liquid-liquid extraction data - Tie line correlation. Ind. Eng. Chem. 1942, 34, 693-696. (30) Shi, M. D.; Wang, L. P.; He, G. Y.; Yu, X. M.; Li, Q. S. Measurement and correlation of liquidliquid equilibria data for low concentrated methylal aqueous solution and extractants. Natural Gas Chemical Industry 2018, 43, 34-39. (31) Liu, X. K.; Zhang, X. L.; Liu, Q.; Li, X. Determination and correlation of liquid equilibrium of


Page 19 of 20 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 51 52 53 54 55 56 57 58 59 60


the n-Dodecane-Toluene-Phenol ternary system. Journal of Chemical Engineering of Chinese Universities 2018, 32, 275-279. (32) Cheng, K.; Feng, W. X.; Wang, L. H.; Chen, Y. Phase equilibria for quaternary systems of water + methanol or ethanol + ethyl methyl carbonate + octane at T = 298.15 K. J. Solution Chem. 2017, 46, 1337-1348. (33) Jiang, S. L.; Gao, J.; Li, R.; Xu, D. M.; Zhang, L. Z.; Wang, Y. L. Isobaric vapor-liquid equilibrium measurements for separation of azeotrope (methanol + methyl acetate). J. Chem. Eng. Data 2019, 64, 296-302. (34) Zhang, Q. H.; Qiu, Z. M.; Luo, Z. C.; Hu, Y. Vapor-liquid equilibria of trioxane-water and trioxane-1,2-dichloroethane system. Journal of East China University of Science and Technology 1996, 22, 7-12.



For Table of Contents (TOC) graphic only


Page 20 of 20

Thermodynamic investigation of 1,3,5-trioxane, methyl acrylate, methyl

Recommend Documents