Determination and modeling for solid–liquid phase equilibrium of ternary caprolactam + cyclohexanone oxime + methyl tert-butyl ether system

Determination and modeling for solid–liquid phase equilibrium of ternary caprolactam + cyclohexanone oxime + methyl tert-butyl ether system

Accepted Manuscript Determination and modeling for solid-liquid phase equilibrium of ternary caprolactam + cyclohexanone oxime + methyl tert-butyl eth...

644KB Sizes 0 Downloads 4 Views

Accepted Manuscript Determination and modeling for solid-liquid phase equilibrium of ternary caprolactam + cyclohexanone oxime + methyl tert-butyl ether system Gan bing Yao, Zhan xiang Xia, Zhi hui Li, Chao Shao PII:

S0378-3812(16)30114-5

DOI:

10.1016/j.fluid.2016.03.002

Reference:

FLUID 11038

To appear in:

Fluid Phase Equilibria

Received Date: 26 December 2015 Revised Date:

20 February 2016

Accepted Date: 3 March 2016

Please cite this article as: G.b. Yao, Z.x. Xia, Z.h. Li, C. Shao, Determination and modeling for solidliquid phase equilibrium of ternary caprolactam + cyclohexanone oxime + methyl tert-butyl ether system, Fluid Phase Equilibria (2016), doi: 10.1016/j.fluid.2016.03.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Cyclohexanone oxime 100 278.15 K 293.15 K 308.15 K

100 w2

90 80 30

10 10

20

80

100 w1

90 100 Caprolactam

AC C

EP

TE D

M AN U

SC

0 0 MTBE

RI PT

20

1

ACCEPTED MANUSCRIPT

RI PT

Determination and modeling for solid-liquid phase equilibrium of ternary caprolactam + cyclohexanone oxime + methyl tert-butyl ether system Gan bing Yao*, Zhan xiang Xia, Zhi hui Li, Chao Shao

SC

College of Chemistry & Chemical Engineering, YangZhou University, YangZhou, Jiangsu 225002,

AUTHOR INFORMATION

Corresponding author

M AN U

People’s Republic of China

Gan bing Yao. Tel: + 86 514 87975568; Fax: + 86 514 87975244. E-mail address:

EP

ABSTRACT

TE D

[email protected] (G.B. Yao).

AC C

The solid-liquid equilibrium (SLE) for ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether were determined at three temperatures of (278.15, 293.15 and 308.15) K under pressure of 101.2 kPa. Three isothermal phase diagrams of the system were constructed based on the measured mutual solubility data. There were two pure solids formed in the ternary phase diagram, including pure caprolactam and pure cyclohexanone oxime, which were identified by the method of Schreinemakers’ wet residue. At each temperature, the phase diagram contained three crystallization regions (which corresponded to caprolactam, cyclohexanone oxime, and a mixture of

1

ACCEPTED MANUSCRIPT caprolactam and cyclohexanone oxime), two crystallization curves, and one co-saturated point. The crystallization field of caprolactam was larger than that of cyclohexanone oxime. The NRTL model was employed to correlate and calculate the ternary phase diagram. The calculated ternary phase diagram with the NRTL model agreed well with the experimental ones. The solid-liquid phase

RI PT

equilibrium and phase diagram for the ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether could provide basis for the purification process of caprolactam.

SC

Keywords: Caprolactam; Cyclohexanone oxime; Methyl tert-butyl ether; Solid-liquid equilibrium;

M AN U

Phase diagram.

1. Introduction

Caprolactam is an important raw material for producting of Nylon 6 fiber and Nylon 6 resin

TE D

[1,2]. Up to date, many methods have been developed for the manufacture of caprolactam [3-8]. Most of caprolactam is produced from cyclohexanone, which is first converted to its oxime, and

EP

then treated the oxime with acid to obtain caprolactam [5-6]. The other main production process of caprolactam in industrial scale involves formation of oxime from benzene, toluene and cyclohexane

AC C

through the beckmann rearrangement [3,7,8]. During the production process, no matter which method is used, the impurity, cyclohexanone oxime, is brought to the product of caprolactam through the intermediate by-products [9-11]. The quality index for caprolactam is very rigorous because even a small amount of impurity significantly affects the physical–mechanical properties of Nylon 6 [12-14]. For example, the presence of 0.1 % (mass fraction) cyclohexanone oxime will cause a considerable decrease in the relative viscosity of polycaproamide [14]. At present, the crude product of caprolactam is purified via multistage distillation and extraction in industrial process, 2

ACCEPTED MANUSCRIPT which are highly energy-demanding separation operations to remove these impurities. Recently an adiabatic evaporative cooling crystallization procedure for the system of caprolactam-water was proposed by Diepen et al. [15]. Solvent crystallization is an attractive method for purifying the caprolactam from its crude product in industrial manufacturing process which leads to the product

caprolactam, especially methyl tert-butyl ether [18,19].

RI PT

with high purity [16,17]. Some organic solvents have shown potential uses in purification of

SC

In the design and optimization of solvent crystallization process, the solubility data and reliable phase diagrams of binary and multicomponent mixtures are of great importance. In the literatures,

M AN U

although the solubility of caprolactam and cyclohexanone oxime in several pure organic solvents including methyl tert-butyl ether, isopropyl ether, 1-propanol, 2-propanol, and 1-butanol are reported [18,19], no further study is made on solid-liquid equilibrium and phase diagrams for the ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether at different

TE D

temperatures. In order to provide guidance and basic data for designing the crystallization process which is employed to develop the optimal crystallization process for caprolactam purification, the

EP

main object of this investigation is to study and generate the ternary phase diagrams for the ternary caprolactam + cyclohexanone oxime + methyl tert-butyl ether system at (278.15, 293.15 and 308.15)

AC C

K by using the method of Schreinemakers’ wet residue [20-22] and indicate the relation between temperature and the ternary phase diagrams.

2. Experimental section 2.1. Materials

Caprolactam and cyclohexanone oxime (0.98 and 0.975 in mass fraction, respectively) were supplied by Aladdin Chemical Reagent Co., Ltd., China. The crude materials were repeatedly purified by the method of recrystallization in ethanol. The mass fraction purity of caprolactam used 3

ACCEPTED MANUSCRIPT in phase equilibrium determination is higher than 0.996, and of cyclohexanone oxime, 0.994. The compositions of the two compounds were confirmed by gas chromatography (GC) [18] with a type of Agilent 7890A Infinity GC provided by Agilent Technologies. The solvent methyl tert-butyl ether was purchased from Sinopharm Chemical Reagent Co., Ltd., China with the mass fraction

RI PT

purity of 0.995. No further purification was made before use. The detailed information of the materials is presented in Table 1.

SC

2.2. Apparatus and procedure

The ternary solid-liquid phase equilibrium was studied by employing the method of isothermal

M AN U

solution saturation [21,22]. The pure equilibrium solid phases were identified via the method of Schreinemakers’ wet residue [20-22]. A jacketed dissolution vessel with a volume of 50 mL was used to dissolve the excess caprolactam and cyclohexanone oxime in methyl tert-butyl ether during the experiment. The original composition of system was approximately back-calculated based on

TE D

the “lever rule” and “linking rule”. The components were taken in such ratios that the composition of the resulting saturated mixture fell in the desired portion of the solubility curve. A condenser was

EP

attached with the vessel to prevent the solvent from escaping. An isothermal water bath (Blon DCW 4600, Shanghai) was employed to maintain the experimental temperature constant. The real

AC C

temperature of studied system was displayed by using a precision mercury thermometer inserted into solution of the vessel, and the standard uncertainty of which was 0.01 K. The solution procedure of the mixture was facilitated by stirring at a constant temperature. The system equilibrium was verified by testing repeatedly the composition of caprolactam and cyclohexanone oxime. The equilibration was considered to reach when the composition of solution didn’t change. It showed that it would take about 22 h to reach equilibrium. After the ternary system arrived at equilibrium, the stirring was stopped. Thirty minutes later, the saturated liquid and equilibrium solid 4

ACCEPTED MANUSCRIPT adhering some saturated liquid were taken out respectively and analyzed by GC [18]. In order to acquire different compositions of the equilibrium solid phase and equilibrium liquid phase, the components of initial systems were taken by varying the ratio of caprolactam to cyclohexanone oxime. Generally, the selected temperature range should be as wide as possible. However, the

studied at three temperatures of (278.15, 293.15 and 308.15) K.

SC

2.3. Analytical method

RI PT

boiling point of methyl tert-butyl ether is 328.4 K, the ternary solid-liquid phase equilibrium was

The samples of wet solid phase and saturated liquid phase containing caprolactam and

M AN U

cyclohexanone oxime were dissolved in methanol, respectively. The mass fractions of caprolactam and cyclohexanone oxime in solution were quantitatively tested by GC [18]. The chromatographic column was a 122-1032G DB-1 capillary column (30m×0.320 mm; 0.25µm film thickness). The temperature of injector was 520 K; and FID, 530 K. Pure nitrogen was used as carrier gas with a

TE D

flow rate of 2.2 mL·min-1. Temperature program was from 323 K (10 min) to 530 K (20min) at 12 K⋅min-1. The injection mode was split with a split ration of 50:1. An analytical balance (Mettler

EP

Toledo ME) having a standard uncertainty of 0.0001 g was employed in the experiment to determine the mass of wet solid phase and saturated liquid phase. Each test was performed three

AC C

times to ensure repeatability of the solubility determination, and the average value of corresponding experimental measurements was considered to be the solubility data. The relative stand uncertainty of the test was evaluated to be 0.020 in mass fraction.

3. Thermodynamic background At a given temperature and pressure, if a solid-liquid system is in equilibrium, the fugacity of liquid equals to that of corresponding solid. For a component i, the equilibrium condition can be expressed as: 5

x γ f =xγ f L L L i i i

ACCEPTED MANUSCRIPT

s s s i i i

(1)

Here fi is the fugacity of pure component i, and γ i , the activity coefficient. L and S represent the liquid state and solid state, respectively. The solid solubility in a liquid can be described in a very general form by equation (2), which is

ln ( xi ⋅ γ i ) =

∆H p  1 1  ∆Cp  − − R  Tp T  R

 1 1  ∆Cp  Tp  ln    −  − T T R T  p  

(2)

RI PT

deduced in accordance with the classical theory of solid-liquid phase equilibrium [23].

SC

Where ∆Hp is the molar fusing enthalpy at triple-point temperature Tp. ∆Cp is the difference of

M AN U

solute heat capacity between in the solid state and the liquid state. In general, the triple-point temperature is approximately equal to the normal melting point temperature; and the fusion enthalpies at the two temperatures are almost same. The triple-point temperature and corresponding fusing enthalpy can be replaced by those at normal conditions. By these replacement and canceling

TE D

the last two terms relating to ∆Cp in equation. (2), a simple equation describing the solid-liquid phase equilibrium is obtained and expressed as equation (3) [24].

∆H m  1 1   −  R  Tm T 

(3)

EP

ln ( xi ⋅ γ i ) =

Where Tm is normal melting temperature and ∆Hm is the fusion enthalpy of a solid at normal

AC C

conditions.

Thermodynamic relations of the activity coefficient should be known in modeling solid-liquid phase equilibrium. In the present work, the NRTL model developed by Renon and Prausnitz [25] is used to compute the activity coefficient. It is established on the basis of local composition concept with the assumption of a non-random molecular distribution. The activity coefficient for a constituent i based on the NRTL mode can be expressed as

6

ACCEPTED MANUSCRIPT

N   xiτ ijGij  ∑ N  xG  ln γ i = j=1N + ∑ N j ij τ ij − i=1N  j=1 Gij xi Gij xi Gij xi  ∑ ∑ ∑    i =1 i=1 i =1

(4)

G ji = exp(−α jiτ ji )

(5)

αij = α ji

(6)

N

τ ij =

ji Gji x j

g ij − g ii RT

=

∆g ij

(7)

RT

RI PT

∑τ

SC

∆gij ( = gij − g jj ) and ∆g ji ( = g ji − gii ) are temperature independent parameters (J⋅mol−1) relating to energy interactions between solute and solvent molecules. α is an adjustable parameter

M AN U

which is in relation to the nonrandomness of solution. Its value is frequently in the range from (0.2 to 0.47). In this work, it is taken as 0.3. The values of ∆gij can be obtained by regression from experimental solubility data.

4. Results and discussion

TE D

4.1. Solid-liquid phase equilibrium data

The equilibrium mutual solubility for the ternary system of caprolactam + cyclohexanone oxime +

EP

methyl tert-butyl ether at temperatures of 278.15 K, 293.15 K and 308.15 K are tabulated in Tables 2, 3 and 4, respectively. The corresponding ternary phase diagrams are constructed and shown

AC C

graphically in Figs. 1, 2 and 3.

Points MTBE, caprolactam and cyclohexanone oxime in Figs. 1-3 denote methyl tert-butyl ether, pure solid caprolactam and pure solid cyclohexanone oxime, respectively. Points S1, S2 and S3 represent the equilibrium solubility of caprolactam in methyl tert-butyl ether at 278.15 K, 293.15 K, 308.15 K respectively, and S1* , S 2* , and S 3* represent the solubility of cyclohexanone oxime at three corresponding temperatures. Points C1, C2 and C3 are invariant points, which demonstrate that the two pure solids caprolactam and cyclohexanone oxime are saturated with equilibrium liquid 7

ACCEPTED MANUSCRIPT phase in the ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether at three studied temperatures. In Figs. 1-3, S1C1, S2C2, and S3C3 are solubility curves of pure solid caprolactam at 278.15 K, 293.15 K, 308.15 K respectively, which indicate that the compositions of saturated liquid are in

RI PT

equilibrium with pure solid caprolactam. Along the solubility curves S1C1, S2C2, and S3C3, linking the composition points of saturated liquid and corresponding moist residue and extended, the joint

SC

point of the tie-lines is nearly the component of pure solid caprolactam. Similarly, S1*C1 , S 2*C2 and

S3*C3 are solubility curves of pure solid cyclohexanone oxime, which show that the compositions

M AN U

of saturated liquid are in equilibrium with the pure solid cyclohexanone oxime. Along the solubility curves S1*C1 , S 2*C2 and S3*C3 , linking the composition points of saturated liquid and corresponding moist residue and extended, the joint point of the tie-lines is nearly the component of pure solid cyclohexanone oxime.

TE D

Figs. 1, 2 and 3 further demonstrate that solid solution is not formed in the ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether at studied temperatures. The ternary

EP

phase diagram at each temperature is divided into four regions by two solubility curves. I is unsaturated region; II is the crystallization region of caprolactam; III is the crystallization region of

AC C

cyclohexanone oxime, and IV is the crystallization region of mixture solids of caprolactam and cyclohexanone oxime. At a certain temperature, along the solubility curve of cyclohexanone oxime, the solubility of cyclohexanone oxime decrease with increasing in concentration of caprolactam. Similarly, along the solubility curve of caprolactam, the solubility of caprolactam decrease with increasing in concentration of cyclohexanone oxime. Figs. 1-3 further reveal the equilibrium phase diagram dependence on temperature for the ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether. The solubility of 8

ACCEPTED MANUSCRIPT caprolactam and cyclohexanone oxime increase with increasing in temperature from T = (278.15 to 308.15) K, and the co-saturated points move upward with the increase in temperature. The crystalline fields of the two pure solids increase as the temperature decreases. The crystallization field of caprolactam is larger than that of cyclohexanone oxime at a given temperature.

RI PT

The schematic crystallization route is displayed in Fig. 4. The point O stands for the original composition of caprolactam and cyclohexanone oxime mixture. The crystallization path was plotted

SC

in red color. The point O is diluted to point A at 308.15 K by adding MTBE. The solution crystallizes out caprolactam, and the concentration of the mixture moves into invariant point C3. By

M AN U

decreasing the temperature from 308.15 K to 278.15 K, the solution cools down and crystallizes out the mixture of caprolactam and cyclohexanone oxime along the curve C3C1. The paths used for recycling is from C1P by adding the raw material O, then the solution can be processed along the C1C3P.

TE D

4.2. Solid-liquid phase equilibrium correlation and calculation In the present work, the NRTL model is employed to correlate and calculate the solid-liquid

EP

equilibrium for the ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether at different temperatures. The solubility of caprolactam in methyl tert-butyl ether and

AC C

cyclohexanone oxime in methyl tert-butyl ether are taken from refs. [18,19], and are correlated by NRTL model. The object function of regression process, F, is the difference between the measured solubility and calculated values, which is described in equation. (8). F = ∑ ( ln γ ie − ln γ ical )

2

(8)

i=1

Here ln γ ie is logarithm of the activity coefficient achieved by solid-liquid phase equilibrium equation (3), and ln γ ical is logarithm of calculated activity coefficient obtained with NRTL model. 9

ACCEPTED MANUSCRIPT Furthermore, the root-mean-square deviation (Rmsd) and the relative average deviation (RAD) are employed to estimate the NRTL model, which are expressed as equations (9) and (10).

1 N

N

∑ i =1

1 Rmsd =  N

wi − wical wi

∑ (w N

j=1

j

−w

(9)

)

cal 2 j

  

12

(10)

RI PT

RAD =

Where N is the number of solubility data points; wical is the calculated mass fraction

of binary parameters, and wi , experimental ones.

SC

solubility of caprolactam or cyclohexanone oxime in methyl tert-butyl ether in terms of the values

M AN U

During regression process, the melting temperature Tm and melting enthalpy ∆Hm for caprolactam are cited in reference [26], and for cyclohexanone oxime, in reference [27]. The obtained values of binary interaction parameters ∆gij for the two binary systems of caprolactam + methyl tert-butyl ether and cyclohexanone oxime + methyl tert-butyl ether are regressed by nonlinear regression

TE D

method using Mathcad software on the basis of the solubility data [18,19]. The obtained interaction parameters for the binary systems and the root-mean-square deviation (Rmsd) are presented in

EP

Table 5. It shows that for the binary systems of caprolactam + methyl tert-butyl ether and cyclohexanone oxime + methyl tert-butyl ether, the values of Rmsd are less than 5.36×10-3. The

AC C

NRTL model gives better results for the binary systems. Based on the acquired binary interaction parameters for the systems of caprolactam + methyl tert-butyl ether and cyclohexanone oxime + methyl tert-butyl ether, the interaction parameters between a pair of caprolactam-cyclohexanone oxime molecules are acquired by using the method of nonlinear regression on the basis of the mutual solubility data for the ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether at three temperatures. The obtained parameters’ values together with the values of Rmsd are tabulated in Table 5. Furthermore, the calculated mass 10

ACCEPTED MANUSCRIPT fraction of caprolactam and cyclohexanone oxime in equilibrium liquid phase via NRTL at 278.15 K, 293.15 K and 308.15 K are presented in Tables 2, 3, and 4, respectively, and the calculated phase diagrams are plotted in Fig. 4. Table 5 shows that the Rmsd value for the ternary system of caprolactam + cyclohexanone oxime + methyl tert-butyl ether is 0.48, and the maximum RAD value

RI PT

obtained with NRTL model is 0.03%. As a result the calculated values by NRTL model have good agreement with the experimental data.

SC

5. Conclusion

The solid-liquid phase equilibrium and the mutual solubility for the ternary system of

M AN U

caprolactam + cyclohexanone oxime + methyl tert-butyl ether at (278.15, 293.15 and 308.15) K were measured and the corresponding ternary phase diagrams were constructed. Two pure solid phases were formed in the ternary system at each temperature, which were confirmed by the method of Schreinemakers’ wet residue and correspond to caprolactam and cyclohexanone oxime. There

TE D

are three crystallization regions (caprolactam, cyclohexanone oxime, and a mixture of caprolactam and cyclohexanone oxime), two equilibrium solubility curves, and one co-saturated point in each

EP

ternary phase diagram. The crystallization field of cyclohexanone oxime is smaller than that of caprolactam at the same temperature. The ternary phase diagram for solid-liquid phase equilibrium

AC C

data were correlated and calculated by using the NRTL model, and the corresponding binary interaction parameters are acquired. The NRTL model gives acceptable results for the investigated system.

Acknowledgment

11

ACCEPTED MANUSCRIPT This work was financially supported by the Priority Academic Program Development of Jiangsu Higher Education Institutions. The authors would like to express their gratitude for the Yangzhou City Science and Technology Bureau, China (Project number: 2012038-3 and YZ2011139) for their

RI PT

support.

References [1] X. Wang E.N. Kalali

D.Y. Wang, An in situ polymerization approach for functionalized

SC

MoS2/nylon-6 nanocomposites with enhanced mechanical properties and thermal stability, J. Mater. Chem. A 3 (2015) 24112-24120.

M AN U

[2] H.C.H.J. Turk, W.P. Kloosterman, D.K. Ninaber, K.P.A.M. Kolen, K. J. Knutova, E. Suir, M. Schürmann, P.C. Raemakers-Franken, M. Müller, S.M.A. Wildeman, L.M. Raamsdonk, R. Pol, L. Wu, M.F. Temudo, R.A.M. Hoeven, M. Akeroydm R,E, Stoel, H.J. Noorman, R.A.L.

TE D

Bovenberg, A.C. Trefzer, Metabolic engineering toward sustainable production of Nylon-6, ACS Synth. Biol. 5 (2016) 65-73.

[3] K.Y. You, Study on one-step synthesis of ε-caprolactam from nitrosation of cyclohexane,

EP

Dissertation for Ph.D. Degree, Xiangtan Univ. 2008.

AC C

[4] G. Dahlhoff, J.P.M. Niederer, W.F. Hoelderich, ε-Caprolactam: new by-product free synthesis routes, Catal. Rev. 43 (2001) 381-441. [5] T. Curtin, J.B. Mcmonagle, M. Ruwet, B.K. Hodnett, Deactivation and regeneration of alumina catalysts for the rearrangement of cyclohexanone oxime into caprolactam, J. Catal. 142 (1993) 172-181.

12

ACCEPTED MANUSCRIPT [6] M. Ishida T. Suzuki

H. Ichihashi

A. Shiga, Theoretical study on vapour phase Beckmann

rearrangement of cyclohexanone oxime over a high silica MFI zeolite, Catal. Today, 87 (2003) 187-194. [7] K.Y. You, B.H. Wu, L.Q. Mao, D.L. Yin, P.L. Liu, H.A. Luo, A novel route to one-step

RI PT

formation of ɛ-caprolactam from cyclohexane and nitrosyl sulfuric acid catalyzed by VPO composites, Catal. Lett. 118 (2007) 129-133.

SC

[8] K.Y. You, Q. Mao, L. Chen, D.L. Yin, P.L. Liu, H.A. Luo, One-step synthesis of ε-caprolactam from cyclohexane and nitrosyl sulfuric acid catalyzed by VPO supported transition metal

M AN U

composites, Catal. Commun. 9 (2008) 2136-2139.

[9] L.G. Jodra, A. Romero, F. García-Ochoa, J. Aracil, Impurity content and quality definition of commercial ε-caprolactam, Ind. Eng. Chem. Process. Des. Dev. 20 (1981) 562–566. [10] L.G. Jorda, A. Romero, F. García-Ochoa, J Aracil, Analysis of the impurities in industrial

TE D

ε-caprolactam. Hypothesis of formation, J. Appl. Polym. Sci. 26 (1981) 3271-3282. [11] Y.Y. Zhao, Z.Z. Jing, H.P. Li, H.S. Zhang, The determination of impurities in caprolactam by

EP

capillary gas chromatography-mass spectrometry, Microchem. J. 69 (2001) 213-217. [12] L.E. Nikulina, V.R. Koroleva, V.A. Chechko, The effect of impurities in the ε-caprolactam on

AC C

the thermostability of polycaproamide, Fiber Chem. 4 (1972) 78-82. [13] X.K. Meng, X.H. Mu, B.N. Zong, E.Z. Min, Z.H. Zhu, S.B. Fu, Y.B. Luo, Purification of caprolactam in magnetically stabilized bed reactor, Catal. Today 79–80 (2003) 21–27. [14] L.T. Fisyuk, L.A. Lezhnina, V.T. Butkin, The influence of impurities in the caprolactam on the properties of the polycaproamide, Khimicheskie Volokna 1 (1978) 5-7.

13

ACCEPTED MANUSCRIPT [15] P.J. Diepen, O.S.L. Bruinsma, G.M. Rosmalen, Melt crystallization by controlled evaporative cooling. The caprolactam-water syetem in batch operation, Chem. Eng. Sci. 55 (2000) 3575– 3584.

crystallization, Chem. Eng. Res. Des. 75 (1997) 171-175.

RI PT

[16] P.J. Diepen, O.S.L. Bruinsma, G.M. Rosmalen, Crystal size engineering in melt suspension

[17] M.A. Gun, O.S.L. Bruinsma, P.J. Jansens, Purification of polycrystalline ε-caprolactam

SC

particles, Chem. Eng. Sci. 60 (2005) 201-211.

for Ph.D. Degree, Tianjn Univ. 2012.

M AN U

[18] J.L. Li, Solubility of caprolactam and effects of impurities on its crystallization, Dissertation

[19] C.L. Guo, L.J. Li, J.Y. Cheng, J.L. Zhang, W. Li, Solubility of caprolactam in different organic solvents, Fluid Phase Equilib. 319 (2012) 9-15.

(1961) 324-324.

TE D

[20] H.A. Schott, Mathematical extrapolation for the method of wet residues, J. Chem. Eng. Data 6

[21] L. Meng, J. Xu, C.B. Du, S. Han, R.J. Xu, J. Wang, H.K. Zhao, Determination and correlation

EP

of solideliquid phase equilibrium and phase diagram for multicomponent system of mixed dibasic acids. (I) Ternary system of adipic acid + glutaric acid + ethanol system, Fluid Phase

AC C

Equilib. 409 (2016) 256-263.

[22] J. Nývlt, Solid-Liquid Phase Equilibria, Publishing House of the Czechoslovak, Prague, 1977.

[23] J.M. Prausnitz, R.N. Lichtenthaler, E.G. Azevedo, Molecular Thermodynamics of Fluid-phase Equilibria, 3rd ed., Prentice Hall PTR, Upper Saddle River, NJ, 1999.

[24] M.W. Stanley, Phase Equilibria in Chemical Engineering, Butterworth, New York, 1985.

14

ACCEPTED MANUSCRIPT [25] H. Renon, J.M. Prausnitz, Estimation of parameters for the NRTL equation for excess Gibbs energies of strongly nonideal liquid mixtures, Ind. Eng. Chem. Process Des. Dev. 8 (1969) 413-419. [26] G.J. Kabo, A.A. Kozyro, V.S. Krouk, V.M. Sevruk, I.A. Yursha, V.V. Simirsky, V.I.

RI PT

Gogolinsky, Thermodynamic properties of 6-aminohexanoic lactam (e-caprolactam), J. Chem. Thermodyn. 24 (1992) 1-13.

SC

[27] D.H. Zaitsau, Y.U. Paulechka, G.J. Kabo, A.V. Blokhin, V.N. Emel'yanenko, S.P. Verevkin, A

AC C

EP

TE D

M AN U

Heintz, Thermodynamics of cyclohexanone oxime, J. Chem. Eng. Data 53 (2008) 694-703.

15

ACCEPTED MANUSCRIPT

Cyclohexanone oxime 100

RI PT

60

S*1

C1

20

II

I 0

40

60

100 w

TE D

0 S1 20 MTBE

SC

40

M AN U

100 w2

80

1

80 100 Caprolactam

Fig. 1. Ternary phase diagram for system of caprolactam (1) + cyclohexanone oxime (2) + MTBE (3) at 278.15 K. MTBE denotes methyl tert-butyl ether. w1, mass fraction of caprolactam, w2, mass

EP

fraction of cyclohexanone oxime. S1 , solubility of caprolactam in MTBE at 278.15 K; S1* ,

AC C

solubility of cyclohexanone oxime in MTBE; C1 , co-saturated point of caprolactam and cyclohexanone oxime at 278.15 K; I, unsaturated region; II, crystalline region of caprolactam, III, crystalline region of cyclohexanone oxime, IV, crystalline region of caprolactam and cyclohexanone oxime mixture; ■, composition of equilibrium liquid phase; ●, composition of equilibrium wet solid phase.

16

ACCEPTED MANUSCRIPT

Cyclohexanone oxime 100

RI PT

60

SC

100 w2

80

S*2

C2

20 I 0

II

40

100 w

TE D

0 S220 MTBE

M AN U

40

60

1

80 100 Caprolactam

Fig. 2. Ternary phase diagram for system of caprolactam (1) + cyclohexanone oxime (2) + MTBE

EP

(3) at 293.15 K. S 2 , solubility of caprolactam in MTBE at 293.15 K; S 2* , solubility of cyclohexanone oxime in MTBE; C2 , co-saturated point of caprolactam and cyclohexanone oxime

AC C

at 293.15 K; I, II, III, IV, ■, ●, MTBE, w1 and w2 have the same meaning described in Fig. 1.

17

ACCEPTED MANUSCRIPT

Cyclohexanone oxime 100

RI PT

60 40

C3

S*3

SC

100 w2

80

II

I 0

0 MTBE

M AN U

20

20S3

40 60 100 w1

80 100 Caprolactam

TE D

Fig. 3. Ternary phase diagram for system of caprolactam (1) + cyclohexanone oxime (2) + MTBE (3) at 308.15 K. S 3 , solubility of caprolactam in MTBE at 308.15 K; S 3* , solubility of

EP

cyclohexanone oxime in MTBE; C3 , co-saturated point of caprolactam and cyclohexanone oxime

AC C

at 308.15 K; I, II, III, IV, ■, ●, MTBE, w1 and w2 have the same meaning described in Fig. 1.

18

ACCEPTED MANUSCRIPT

Cyclohexanone oxime 100

RI PT C3

30 20

P

C1

SC

80

O

M AN U

100 w2

90

A

10 0

10

20

80

100 w1

TE D

0 MTBE

90 100 Caprolactam

Fig. 4. Comparison of the experimental phase diagrams with the calculated ones with NRTL model and Wilson model for ternary system of caprolactam (1) + cyclohexanone oxime (2) + MTBE (3) at

EP

different temperatures. ■, ● and ▲, experimental data at 278.15 K, 293.15 K and 308.15 K,

AC C

respectively; —, calculated curves with NRTL model.

19

ACCEPTED MANUSCRIPT

Materials

Molecular weight

Mass fraction purity

Purification method

113.18

0.996

Recrystallization

Cyclohexanone oxime

113.16

0.994

Recrystallization

Methyl tert-butyl ether

88.15

0.995

---

Gas chromatography.

TE D

Taken from literatures [26] and [27], respectively;

Aladdin Chemical Reagent Co., Ltd. (China)

Sinopharm Chemical Reagent Co.Ltd (China)

EP

c

Caprolactam

AC C

a,b

M AN U

g·mol−1

Sources

SC

Sources, purity and properties of materials used in this work.

RI PT

Table 1

20

Analytical method

Tm

∆Hm

K

kJ·mol−1

342.31a

16.10a

GCc

362.2b

12.45b

GC

---

---

GC

ACCEPTED MANUSCRIPT

Table 2

Mass fraction solubility for the ternary system of caprolactam (1) + cyclohexanone oxime (2) + methyl tert-butyl ether (3) at 278.15 K under pressure p = 101.2 kPaa.

100 w2

100 w1

0

22.09b

---

21.81

---

1.81

21.62

---

21.36

0.65

3.5

21.04

---

20.94

1.07

4.85

20.6

---

20.61

1.81

6.62

20.18

---

20.17

2.62

7.04

19.79

6.54

7.36

17.83

6.73

7.58

13.78

7.14

7.97

10.44

7.49

8.19

7.67

8.58

4.65

8.83

2.81

9.01c

0

Solid phase

100 w2 ---

Capro

71.35

Capro

76.43

Capro

SC

100 w2

70.59

Capro

68.73

Capro

M AN U

100 w1

Moist solid phase cal

100 w1

43.74

45.41

Capro + Cyclo

---

68.59

5.98

Cyclo

---

73.1

3.98

Cyclo

---

75.53

2.68

Cyclo

7.79

---

70.17

2.37

Cyclo

8.14

---

66.27

1.69

Cyclo

8.35

---

73.53

0.79

Cyclo

8.69

---

---

---

Cyclo

TE D

20.07

EP

100 RAD

3.64

w, mass fraction; Capro, caprolactam; Cyclo, cyclohexanone oxime. RAD, relative mean

AC C

a

NRTL cal

RI PT

Liquid phase

deviation, RAD =

1 N

N

wj − wcal j

j=1

wj



. The relative standard uncertainty of the mass fraction solubility

is ur (w) = 0.020, and the standard uncertainties u are u (T) = 0.02 K for the measured temperature, and u (p) = 0.45 kPa for pressure; b,c

Taken from references [18] and [19], respectively.

21

ACCEPTED MANUSCRIPT

Table 3

Mass fraction solubility for the ternary system of caprolactam (1) + cyclohexanone oxime (2) + methyl tert-butyl ether (3) at 293.15 K under pressure p = 101.2 kPaa. NRTL

100 w1

100 w2

0

26.34b

---

26.42

---

100 w2

100 w1

1.78

25.83

---

25.94

0.56

3.56

25.09

---

25.47

1.02

5.29

24.72

---

25.01

1.52

7.65

24.13

---

24.38

2.74

9.74

23.72

---

23.82

10.82

23.37

10.14

11.26

19.05

10.74

11.58

15.52

11.25

11.92

11.37

12.33

8.38

12.75

4.57

12.97

1.81

13.13c

0

Solid phase

100 w2

Capro

74.85

Capro

78.83

Capro

77.70

Capro

72.59

Capro

2.85

78.46

Capro

23.54

46.94

35.79

Capro + Cyclo

---

69.17

6.83

Cyclo

---

75.96

4.18

Cyclo

11.88

---

71.13

3.70

Cyclo

12.36

---

68.13

3.13

Cyclo

12.99

---

72.28

1.50

Cyclo

13.46

---

77.67

0.42

Cyclo

13.78

---

---

---

Cyclo

EP

TE D

M AN U

---

AC C

100 RAD

a

Moist solid phase cal

SC

100 w1

cal

RI PT

Liquid phase

2.04

w, mass fraction; Capro, caprolactam; Cyclo, cyclohexanone oxime. RAD, relative mean

1 deviation, RAD = N

N

wj − wcal j

j=1

wj



. The relative standard uncertainty of the mass fraction solubility

is ur (w) = 0.020, and the standard uncertainties u are u (T) = 0.02 K for the measured temperature, and u (p) = 0.45 kPa for pressure; b,c

Taken from references [18] and [19], respectively. 22

ACCEPTED MANUSCRIPT

Table 4

Mass fraction solubility for the ternary system of caprolactam (1) + cyclohexanone oxime (2) + methyl tert-butyl ether (3) at 308.15 K under pressure p = 101.2 kPaa. Liquid phase

NRTL

0

31.42b

---

31.59

---

2.77

30.82

---

30.82

0.73

5.81

29.58

---

29.97

1.81

9.82

28.73

---

28.85

3.16

76.35

Capro

12.59

28.06

---

28.07

4.01

78.12

Capro

15.13

27.78

---

27.36

5.11

76.26

Capro

16.71

27.35

16.05

26.91

52.16

27.35

Capro + Cyclo

17.08

25.01

16.47

---

76.26

7.42

Cyclo

18.13

19.06

17.6

---

74.15

6.18

Cyclo

18.94

15.48

18.33

---

77.03

4.54

Cyclo

19.86

10.89

19.32

---

73.89

3.56

Cyclo

20.45

7.14

20.18

---

75.64

2.20

Cyclo

20.78

3.42

21.06

---

72.23

1.33

Cyclo

21.37c

0

21.91

---

---

---

Cyclo

EP

AC C

100 w1

100 w2

RI PT

100 w2

---

Capro

80.67

Capro

78.58

Capro

SC

M AN U

100 w2

Solid phase

100 w1

100 RAD

a

Moist solid phase cal

TE D

100 w1

cal

1.80

w, mass fraction; Capro, caprolactam; Cyclo, cyclohexanone oxime. RAD, relative mean

deviation, RAD =

1 N

N

wj − wcal j

j=1

wj



. The relative standard uncertainty of the mass fraction solubility

is ur (w) = 0.020, and the standard uncertainties u are u (T) = 0.02 K for the measured temperature, and u (p) = 0.45 kPa for pressure; b,c

Taken from references [18] and [19], respectively. 23

ACCEPTED MANUSCRIPT

Table 5

The regressed binary interaction parameters in the NRTL model for the ternary caprolactam (1) + Cyclohexanone oxime (2) + methyl tert-butyl ether (3) system.

△gji

1-2

2546.08

5503.95

1-3

5190.76

829.82

2-3

16256.36

-1096.95

α

M AN U TE D EP AC C 24

1000 Rmsd 3.98

RI PT

△gij

SC

i-j

0.3

5.36

1.78

ACCEPTED MANUSCRIPT

Highlights



Solid-liquid phase equilibrium of caprolactam + cyclohexanone oxime + methyl tert-butyl ether system were determined. The ternary phase diagrams were constructed at three temperatures.



The ternary phase diagrams were correlated and calculated by NRTL model.

AC C

EP

TE D

M AN U

SC

RI PT



1