Liquid–Liquid Equilibria for the Ternary System ... - ACS Publications

May 20, 2014 - The Othmer−Tobias and the Hand correlations were used to check the reliability of the experimental LLE data. The distribution coeffic...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/jced

Liquid−Liquid Equilibria for the Ternary System Water + Benzyl Alcohol + Methylbenzene at (303.2 to 343.2) K Hui Wang,† Qinbo Wang,*,† Zhenhua Xiong,‡ and Chuxiong Chen‡ †

Department of Chemical Engineering, Hunan University, Changsha, 410082 Hunan P. R. China Zhejiang Shuyang Chemical Co. Ltd., Quzhou, 324002 Zhejiang P. R. China



S Supporting Information *

ABSTRACT: Liquid−liquid equilibrium (LLE) data for the ternary system water + benzyl alcohol + methylbenzene were measured at (303.2 to 343.2) K under atmospheric pressure. The Othmer−Tobias and the Hand correlations were used to check the reliability of the experimental LLE data. The distribution coefficient and separation factor were used to investigate the ability of water to extract benzyl alcohol from methylbenzene. The results show that the distribution coefficient of benzyl alcohol increases with the increasing mass fraction of water in the aqueous phase. The experimental data were correlated by both the nonrandom two-liquid (NRTL) and the universal quasichemical (UNIQUAC) activity coefficient models. The relevant model interaction parameters were regressed by data fitting. The model calculated results show good agreement with the measured data. The obtained interaction parameters can be used in the calculation of LLE for the ternary system water + benzyl alcohol + methylbenzene as well as for the design and optimization of the related separation process.

1. INTRODUCTION Benzyl alcohol is an important fine chemical. It is widely used in the field of food, medicine, daily chemicals, and so on. Traditionally, it is produced by the hydrolysis process of benzyl chloride, which would simultaneously produce a large amount of acidic wasting water and cause serious environmental problems. Though there are also some studies about the air oxidation of methylbenzene to produce benzyl alcohol, this method is still less than ideal because of the poor selectivity. In recent years, some researchers have found a green way to oxidize methylbenzene.1−3 In this way, methylbenzene is oxidized by hydrogen peroxide, and this method is cleaner, safer, and more cost-effective compared with the traditional ways. These existing studies provide a new idea to produce benzyl alcohol. Hydrogen peroxide is an excellent oxidant because of its high oxygen content, and water is the sole theoretical coproduct. In this oxidation system, the reaction mixture would split into two liquid phases due to the existence of water. It is designed to separate benzyl alcohol from the liquid mixtures after the oxidation. It is therefore of greatest importance to measure the liquid−liquid equilibrium (LLE) data of the ternary system water + benzyl alcohol + methylbenzene at different temperatures and © 2014 American Chemical Society

different compositions. Until now, liquid−liquid equilibria for binary mixtures of water + methylbenzene4−18 and water + benzyl alcohol19−23 could be found, but except for the work of Skrzecz,24 no publication or reports on the ternary system water + benzyl alcohol + methylbenzene are available. Furthermore, in the work of Skrzecz, the vapor−liquid−liquid equilibrium (VLLE) data for the ternary system water + benzyl alcohol + methylbenzene were only measured at 293.2 K and (362.2 ± 0.5) under atmospheric pressure. The experimental data was difficult to extrapolate to other temperatures. In this work, the LLE data for the ternary system water + benzyl alcohol + methylbenzene were measured at (303.2 to 343.2) K under atmospheric pressure. The reliability of the experimental LLE data was confirmed according to the Othmer− Tobias and the Hand correlations. The experimental data were correlated by the NRTL (nonrandom two-liquid) activity coefficient model25 and UNIQUAC (universal quasichemical activity coefficient) model.26 The relevant interaction parameters Received: February 26, 2014 Accepted: May 10, 2014 Published: May 20, 2014 2045

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053

Journal of Chemical & Engineering Data

Article

Fischer titrator (Mettler-Toledo V20). The analytic procedure is described in detail elsewhere.28 For the aqueous phase, the concentrations of benzyl alcohol and methylbenzene were measured, and the concentration of water was determined by the normalized calculation. For the organic phase, the concentrations of benzyl alcohol and water were measured, and the concentration of methylbenzene was determined by the normalized calculation. To verify the reliability and reproducibility of the GC analysis method, five methylbenzene and benzyl alcohol solutions of known concentrations were analyzed. For each solution, the composition was measured at least three times. The standard uncertainty of benzyl alcohol in mass fraction was determined to be 0.0010, and that for methylbenzene was determined to be 0.0014. To verify the reliability and reproducibility of the water concentration determination method, five benzyl alcohol and water solutions of known concentration were analyzed, too. Similarly, each solution was measured at least three times. The standard uncertainty of water in mass fraction was determined to be 0.0008.

are expected to be regressed with the experimental data. The obtained interaction parameters might be used in the calculation of LLE for the ternary system water + benzyl alcohol + methylbenzene as well as for the design and optimization of the related separation process which is an ongoing part of our work.

2. EXPERIMENTAL SECTION 2.1. Materials. The materials used include methylbenzene, benzyl alcohol, and 1,4-dimethylbenzene, which were obtained from Xilong Chemical Co. and had a declared purity of 0.990 in mass fraction. Gas chromatography was used to check the purity of these chemicals. Purified water manufactured by Hangzhou Wahaha Group Co. was obtained from the supermarket (596 mL each bottle). All of the chemicals used in the experiments were used without further purified. The suppliers and the mass fraction of the used chemical reagents are shown in Table 1. Table 1. Suppliers and Mass Fractions of the Chemical Reagent component methylbenzene benzyl alcohol 1,4-dimethylbenzene water a

suppliers Chemical Reagent Co. Xilong Chemical Co. Xilong Chemical Co. Hangzhou Wahaha Group Co.

mass fraction

analysis method

>0.990 >0.990 >0.996 >0.999

GCa GCa GCa

3. RESULTS AND DISCUSSION 3.1. Data Reliability Checking. The reliability and consistency of the experimental LLE data can be tested by the Othmer−Tobias29 and the Hand correlations.30The equations are shown as eq 1 and eq 2:

Gas chromatograph.

2.2. Apparatus and Procedures. The experimental apparatus and sampling methods used in this work was described in detail by Wang et al.27 and Shen.28 Briefly, in each experiment, a given amount of water, methylbenzene, and benzyl alcohol were added into a 100 mL round-bottom glass bottle and then heated the bottle to the desired temperature within 0.1 K by putting it in a thermostatic water bath. To prevent the evaporation of solvent, the bottle was sealed by a rubber stopper. After reaching the desired temperature, the mixture was stirred about 3 h to accelerate the partition of benzyl alcohol between methylbenzene and water. Then the mixture was left undisturbed for 24 h for phase equilibrium. To verify the attainment of LLE, the upper and lower liquid phases were sampled once an hour, and then the phase composition was determined. Experimental results show that 3 h was enough for benzyl alcohol + methylbenzene + water to reach LLE, because repetitive measurements during the following several hours indicated the results were reproducible with 3 %. A series of LLE measurements were carried out by changing the temperatures and the compositions of the mixture. The uncertainty of experimental temperature was 0.1 K, and the uncertainty of balance used in this work is 0.0001 g. 2.3. Analysis. In each measurement, about 1.0 g of the solution was sampled for the aqueous phase, and about 0.2 g of the solution was sampled for the organic phase. The concentrations of benzyl alcohol and methylbenzene were analyzed by the gas chromatograph (GC-1690) equipped with a flame ionization detector (FID) produced by KeXiao Chemical Instrument Co. Ltd. The GC column was a capillary column (30m × 0.32 mm × 0.50 μm; stationary phase, AT SE-54). The internal standard method was used, and 1,4-dimethylbenzene was chosen as the internal standard substance. The concentration of water was determined by using a compact volumetric Karl

⎛ 1 − W33 ⎞ ⎛ 1 − W11 ⎞ ln⎜ ⎟ ⎟ = A + B ln⎜ ⎝ W11 ⎠ ⎝ W33 ⎠

(1)

⎛W ⎞ ⎛W ⎞ ln⎜ 21 ⎟ = M + N ln⎜ 23 ⎟ ⎝ W11 ⎠ ⎝ W33 ⎠

(2)

In eq 1, A and B are constants of the Othmer−Tobias correlation, respectively. W11 is the mass fraction of water in the water-rich phase, and W33 is the mass fraction of benzyl alcohol in the benzyl alcohol-rich phase. In eq 2, M and N are constants of the Hand correlation, respectively. W11 and W21 are the mass fractions of water in the aqueous phase and the organic phase, and W23 is and W33 are the mass fractions of benzyl alcohol in the aqueous phase and the organic phase, respectively. The parameters in Othmer−Tobias and the Hand equations are linear-regressed by fitting the experimental data with eq 1 and eq 2. The results are listed in Tables 2 and 3. The Othmer− Tobias plot (Figure 1) and Hand plot (Figure 2) were also used to evaluate the thermodynamic consistency of the experimental data. The results indicated that the linear correlation coefficients, Table 2. Constants of the Othmer−Tobias Equation for the Water (1) + Benzyl Alcohol (2) + Methylbenzene (3) System at (303.2 to 343.2) Ka Othmer−Tobias T/K

A

B

R2b

303.2 313.2 323.2 333.2 343.2

−3.501 ± 0.009 −3.434 ± 0.007 −3.407 ± 0.009 −3.333 ± 0.010 −3.352 ± 0.010

−0.250 ± 0.010 −0.259 ± 0.007 −0.258 ± 0.010 −0.283 ± 0.010 −0.245 ± 0.011

0.987 ± 0.007 0.993 ± 0.003 0.989 ± 0.006 0.989 ± 0.008 0.994 ± 0.008

a

Standard uncertainty u is u(T) = 0.1 K. bR2 is the linear correlation coefficient. 2046

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053

Journal of Chemical & Engineering Data

Article

were also compared with the literature data of the binary mixtures which were mentioned before. The results are shown in Figures 5 and 6. We found that the obtained experimental data have no important deviations with the literature data. The experimental results given in Table 4 indicate that water could be regarded as insoluble with methylbenzene. The solubility of benzyl alcohol in the aqueous phase is greater than methylbenzene in the aqueous phase. The result given in Figures 3 and 4 shows that the mutual solubility of water and methylbenzene increases slightly with the increase of benzyl alcohol’s mass fraction at settled temperature. Within the temperature range of the measurements, the experimental data shows that the mutual solubility of the mixtures slightly changed but not obviously. The ability of the solvent water to extract benzyl alcohol from methylbenzene could be demonstrated by the separation factor S and distribution coefficients k1 and k2. These parameters are defined as eq 3 to 5:

Table 3. Constants of the Hand Equation for the Water (1) + Benzyl Alcohol (2) + Methylbenzene (3) System at (303.2 to 343.2) Ka Hand T/K

M

N

R2b

303.2 313.2 323.2 333.2 343.2

−10.02 ± 0.374 −10.75 ± 0.295 −8.886 ± 0.382 −10.30 ± 0.428 −8.464 ± 0.372

2.297 ± 0.136 2.647 ± 0.109 2.021 ± 0.143 2.630 ± 0.165 1.969 ± 0.143

0.972 ± 0.164 0.986 ± 0.075 0.961 ± 0.168 0.969 ± 0.205 0.992 ± 0.167

a

Standard uncertainty u is u(T) = 0.1 K. bR2 is the linear correlation coefficient.

Figure 1. Othmer−Tobias plots of the ternary water (1) + benzyl alcohol (2) + methylbenzene (3) system at different temperatures: ■, 303.2 K; ▲, 313.2 K; □, 323.2 K; ○, 333.2 K; ◇, 343.2 K. W11 is the mass fraction of water in the aqueous phase, and W33 is the mass fraction of benzyl alcohol in the organic phase.

S=

k2 k1

(3)

k1 =

w31 w33

(4)

k2 =

w21 w23

(5)

where k 1 and k 2 are the distribution coefficients of methylbenzene and benzyl alcohol between organic and aqueous phases, respectively; w31 and w33 are the mass fractions of methylbenzene in aqueous and organic phases, respectively;w21 and w23 are the mass fractions of benzyl alcohol in aqueous and organic phases, respectively. The separation factors and distribution coefficients at different temperatures are summarized in Table 5. The plots for the solubility and distribution coefficient of benzyl alcohol are showed in Figures 7 and 8. It can be found from the plots that there is no obviously influence of temperature on the distribution coefficient at lower amounts of water. However, at higher amounts of water, a higher distribution coefficient is achieved. 3.3. Correlation of Experimental Data. The activities of each component i in the two coexistent liquid phases of a system at equilibrium might be regarded as equal, and the mole fractions xIi and xIIi of LLE phases can be determined using eqs 6 to 9: xi =

Figure 2. Hand plots of the ternary water (1) + benzyl alcohol (2) + methylbenzene (3) system at different temperatures: ■, 303.2 K; ▲,313.2 K; □, 323.2 K; ○, 333.2 K; ◇, 343.2 K. W11 and W21 are the mass fractions of water in the aqueous phase and in the organic phase, respectively. W23 and W33 are the mass fractions of benzyl alcohol in the aqueous phase and in the organic phase, respectively.

R2, were all close to 1, which shows that the experimental data was highly consistent. It verified that the experimental LLE data were reliable. 3.2. Tie-Lines, Distribution Coefficients, and Separation Factor. The measured LLE data of the ternary system of water (1) + benzyl alcohol (2) + methylbenzene (3) at (303.2 to 343.2) K are summarized in Table 4, and the experimental deviations in the measurement are listed in Table A, which is given as Supporting Information. The diagrams for the ternary system at (303.2 to 343.2) K are supplied in Figures 3 and 4. All concentrations are given in mass fraction. The experimental data

wi /Mi 3 ∑i = 1 wi /Mi

(6)

γi IxiI = γi IIxiII

(7)

∑ xiI = 1

(8)

∑ xiII = 1

(9)

where, γIi and γIIi are the activity coefficients of component i in the aqueous phase and organic phase. The key to solve the equations is to calculate the activity coefficients. As UNIFAC parameters are available for each group constituting selected molecules, maybe the UNIFAC activity coefficient model is a good choice to predict the LLE of water (1) + benzyl alcohol (2) + methylbenzene (3). Thus, the UNIFAC model was used as a try to predict the studied LLE system. The predicted LLE results at 303.2 K by the UNIFAC model are 2047

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053

Journal of Chemical & Engineering Data

Article

Table 4. Experimental Liquid−Liquid Equilibrium Data for theb System Water (1) + Benzyl Alcohol (2) + Methylbenzene (3) at (303.2 to 343.2) K and Pressure p = 101.3 kPaa aqueous phase

aqueous phase

organic phase

w1

w2

0.9969 0.9833 0.9781 0.9769 0.9741 0.9712 0.9701 0.9683 0.9666 0.9656 0.9649 0.9552

0.0000 0.0152 0.0194 0.0212 0.0242 0.0266 0.0282 0.0301 0.0321 0.0332 0.0343 0.0448

0.9958 0.9815 0.9778 0.9751 0.9732 0.9708 0.9683 0.9653 0.9647 0.9636 0.9622 0.9529

0.0000 0.0154 0.0196 0.0226 0.0250 0.0275 0.0301 0.0325 0.0341 0.0353 0.0368 0.0471

0.9942 0.9814 0.9778 0.9736 0.9728 0.9704 0.9671 0.9660

0.0000 0.0157 0.0196 0.0241 0.0252 0.0276 0.0306 0.0322

w3 T = 303.2 0.0031 0.0015 0.0025 0.0019 0.0017 0.0022 0.0017 0.0016 0.0013 0.0012 0.0008 0.0000 T = 313.2 0.0042 0.0031 0.0026 0.0023 0.0018 0.0017 0.0016 0.0022 0.0012 0.0011 0.0010 0.0000 T = 323.2 0.0058 0.0029 0.0026 0.0023 0.0020 0.0020 0.0023 0.0018

w1 K 0.0013 0.0044 0.0084 0.0153 0.0204 0.0264 0.0329 0.0388 0.0430 0.0474 0.0520 0.1027 K 0.0014 0.0048 0.0105 0.0171 0.0224 0.0285 0.0338 0.0406 0.0447 0.0500 0.0564 0.1056 K 0.0022 0.0066 0.0142 0.0180 0.0262 0.0311 0.0357 0.0434

w3

w1

w2

0.0000 0.1042 0.2027 0.2963 0.3738 0.4508 0.5167 0.5742 0.6216 0.6732 0.7054 0.8973

0.9987 0.8914 0.7889 0.6884 0.6058 0.5228 0.4504 0.3870 0.3354 0.2794 0.2426 0.0000

0.9631 0.9622 0.9613 0.9521

0.0352 0.0363 0.0374 0.0479

0.0000 0.1224 0.1946 0.2863 0.3697 0.4199 0.5157 0.5712 0.6184 0.6612 0.7019 0.8944

0.9986 0.8728 0.7949 0.6966 0.6079 0.5516 0.4505 0.3882 0.3369 0.2888 0.2417 0.0000

0.9935 0.9802 0.9777 0.9734 0.9711 0.9681 0.9649 0.9618 0.9594 0.9583 0.9577 0.9515

0.0000 0.0169 0.0198 0.0244 0.0269 0.0300 0.0340 0.0367 0.0388 0.0399 0.0406 0.0485

0.0000 0.1005 0.2021 0.2858 0.3739 0.4432 0.5103 0.5538

0.9978 0.8929 0.7837 0.6962 0.5999 0.5257 0.4540 0.4028

0.9914 0.9799 0.9760 0.9715 0.9697 0.9674 0.9660 0.9639 0.9615 0.9598 0.9554 0.9499

0.0000 0.0172 0.0214 0.0259 0.0280 0.0304 0.0329 0.0341 0.0366 0.0384 0.0429 0.0501

ln γi =

xjGij 3 j = 1 ∑k = 1 Gkjxk



3 ⎛ ∑ xτ G ⎞ ⎜τij − k = 1 k kj kj ⎟ 3 ⎜ ∑k = 1 Gkjxk ⎟⎠ ⎝

T = 323.2 0.0017 0.0015 0.0013 0.0000 T = 333.2 0.0065 0.0029 0.0025 0.0022 0.0020 0.0019 0.0011 0.0015 0.0018 0.0018 0.0017 0.0000 T = 343.2 0.0086 0.0029 0.0026 0.0026 0.0023 0.0022 0.0011 0.0020 0.0019 0.0018 0.0017 0.0000

w1 K 0.0472 0.0529 0.0596 0.1128 K 0.0025 0.0044 0.0115 0.0176 0.0282 0.0321 0.0392 0.0437 0.0534 0.0610 0.0675 0.1250 K 0.0050 0.0076 0.0178 0.0247 0.0300 0.0368 0.0461 0.0482 0.0552 0.0601 0.0686 0.1411

w2

w3

0.6067 0.6559 0.7011 0.8872

0.3461 0.2912 0.2393 0.0000

0.0000 0.1019 0.1928 0.2846 0.3727 0.4393 0.5172 0.5683 0.6238 0.6572 0.6992 0.8750

0.9975 0.8937 0.7957 0.6978 0.5991 0.5286 0.4436 0.3880 0.3228 0.2818 0.2333 0.0000

0.0000 0.1044 0.2066 0.2976 0.3788 0.4508 0.5305 0.5727 0.6336 0.6766 0.6940 0.8589

0.9950 0.8880 0.7756 0.6777 0.5912 0.5124 0.4234 0.3791 0.3112 0.2633 0.2374 0.0000

Standard uncertainties u are u(T) = 0.1 K, ur(p) = 0.05, u(x1) = 0.0008, u(x2) = 0.0010, and u(x3) = 0.0014.

τij = aij +

bij (11)

T

Gij = exp( −αijτij)

αij = αji ,

(12)

τij ≠ τji ,

τii = 0

(13)

where T is the absolute temperature, K. aij and bij are the NRTL binary interaction parameters. The activity coefficient equations for the UNIQUAC model used in this work are ln γi = ln

3

+

w3

a

plotted in Figure 9. For comparison, the experimental LLE data at 303.2 K are also shown in Figure 9. All concentrations are given in mole fraction. It can be seen that apparent deviations appears, which indicate the existing UNIFAC model parameters might be unsuitable for this ternary system. The NRTL and the UNIQUAC models could also be used to correlate the experimental data and could be extrapolated to a wider temperature range. Therefore, the obtained experimental data were correlated by both the NRTL and the UNIQUAC models in this work. The experimental data can be used to determine the interaction parameters between water, benzyl alcohol, and methylbenzene by data fitting. The NRTL activity coefficient equations used in this work are 3 ∑ j = 1 τjiGjixj 3 ∑k = 1 Gkixk

organic phase

w2

ψi xi

+

3 ψ 3 θ z qi ln i + li − i ∑ xjl j − qi ln(∑ θτ j ji) xi j = 1 2 ψi j=1 3

θτ j kj 3 j = 1 ∑k = 1 θkτkj

+ qi − qi∑

·

lj =

(10) 2048

⎛Z⎞ ⎜ ⎟(r − q ) − (r − 1) j j ⎝2⎠ j

(14)

(15)

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053

Journal of Chemical & Engineering Data

Article

Figure 3. LLE of the ternary water (1) + benzyl alcohol (2) + methylbenzene (3) system at different temperatures: ●---●, experimental data tie line; □- -□, correlated data tie line using the NRTL model. wi is the mass fraction of component i in water (1) + benzyl alcohol (2) + methylbenzene (3) solvent mixtures.

ψi =

xiri 3 ∑i = 1 xiri

(16)

θi =

xiqi 3 ∑i = 1 xiqi

(17)

⎛ bij ⎞ τij = exp⎜aij + ⎟ T⎠ ⎝

f minsearch in the optimization toolbox of Matlab (Matwork, MA) uses the Nelder−Mead Simplex approach and can be employed for the minimization of the objective function, which is the corresponding relative-mean-standard deviation (RMSD) between the experimental and the correlated mole fractions defined as 1 RMSD = N

(18)

Here ψi and θi represent the volume fraction and the area fraction. aij and bij are UNIQUAC model parameters. The pure component structural parameters (r and q) are listed in Table 6. Using NRTL model and UNIQUAC model, the experimental data were correlated, and the model parameters were optimized. The optimum algorithm applied in the parameter estimation program was the Nelder−Mead Simplex approach.31 Function

⎛ x exp − x cal ⎞2 ijkt ijkt ∑ ∑ ∑ ∑ ⎜⎜ exp ·100⎟⎟ xijkt ⎠ t k j i ⎝

(19)

cal where N is the total number of tie lines, xexp ijkt and xijkt are the experimental and calculated mole fractions, and subscripts i, j, k, and t denote the component, phase, tie-line, and temperature, respectively. The regressed NRTL binary interaction parameters aij and bij, along with the RMSD are shown in Table 7. The

2049

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053

Journal of Chemical & Engineering Data

Article

Figure 4. LLE of the ternary water (1) + benzyl alcohol (2) + methylbenzene (3) system at different temperatures: ●---●, experimental data tie line; □- -□, correlated data tie line using the UNIQUAC model. wi is the mass fraction of component i in water (1) + benzyl alcohol (2) + methylbenzene (3) solvent mixtures.

In the work of Skrzecz et al.,24 most of the VLLE data points for the ternary system water + benzyl alcohol + methylbenzene were measured at 293.2 K and (362.2 ± 0.5) K under atmospheric pressure. To verify the reliability of the regressed NRTL and UNIQUAC model parameters and evaluate the precision of model extrapolation, we mathematically predicted the LLE data at 362.2 K by using the obtained NRTL model parameters as an illustration. The predicted results were scattered in Figure 10. It can be seen that the predicted results have no apparent deviations with the experimental liquid−liquid data reported by Skrzecz et al.24 It indicates the obtained NRTL parameters were predictable and could be extrapolated to a wider temperature range. As the UNIQUAC method is better than the NRTL method for this system, it could be inferred that the UNIQUAC parameters are predictable, too.

regressed UNIQUAC binary interaction parameters aij and bij, along with the RMSD are shown in Table 8. By using the regressed model parameters, the LLE data under the experimental conditions could be simulated, and the correlated results were also given in Figures 3 and 4 for comparison. As can be seen, the experimental data agree well with the calculated data. Further, from the RMSD results, the predicted results by the UNIQUAC method is slightly better in agreement with the calculated data than that of the NRTL method. 3.4. Model Verification. In this work, LLE data for the ternary system water + benzyl alcohol + methylbenzene have been measured at (303.2 to 343.2) K under atmospheric pressure. The NRTL and UNIQUAC model parameters listed in Tables 7 and 8 were regressed at (303.2 to 343.2) K. 2050

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053

Journal of Chemical & Engineering Data

Article

Figure 7. Mass fraction of benzyl alcohol in the borganic phase versus its mass fraction in the aqueous phase at different temperatures: ■, 303.2 K; ▲, 313.2 K; □, 323.2 K; ○, 333.2 K; ◇, 343.2 K. W21 and W23 are the mass fractions of benzyl alcohol in the aqueous phase and in the organic phase, respectively.

Figure 5. Comparisons between experimental solubility of methylbenzene in the organic phase with that reported in literature. ●, experimental data; □, literature data from Stephenson;6 ○, literature data from Chen and Wanger;8 △, literature data from Jou and Mather;12 ▽, literature data from Marche et al.;14 ◇, literature data from Neely et al.;16 ⬡, literature data from Brown et al.;10 ⬠, literature data from Miller and Hawthorne.11

Figure 8. Distribution coefficients of benzyl alcohol versus the mass fraction of water in the aqueous phase at different temperatures: ■, 303.2 K; ▲, 313.2 K; □, 323.2 K; ○, 333.2 K; ◇, 343.2 K. W11 is the mass fraction of water in aqueous phase, and k2 is the distribution coefficient of benzyl alcohol. Figure 6. Comparisons between experimental solubility of benzyl alcohol in the aqueous phase with that reported in literature. ●, experimental data; □, literature data from Stephenson and Stuart;19 ○, literature data from Solimo and Gramajo de Doz;21 △, literature data from Banerjee.23

+ methylbenzene was regressed.17 Even in the order of magnitude, a great difference was found in the model parameters determined in our work and that in Saien’s work. However, from the calculated Gij listed in Table B in the Supporting Information, they gave the same order of magnitude, which indicated that the two sets of NRTL model parameters were suitable for the binary water + methylbenzene system. It might be due to the nonuniqueness of the parameters during the data-fitting process.

Saien et al. measured the LLE for ternary system water + acetic acid + methylbenzene and correlated the experimental results by the NRTL model. The NRTL model parameters between water

Table 5. Distribution Coefficients k and Separation Factors S of Benzyl Alcohol at Different Temperaturesa T = 303.2 K

a

T = 313.2 K

T = 323.2 K

T = 333.2 K

T = 343.2 K

k

S

k

S

k

S

k

S

k

S

0.146 0.096 0.072 0.065 0.059 0.055 0.052 0.052 0.049 0.049

86.69 29.04 25.92 23.07 13.41 14.46 12.68 13.32 11.48 14.75

0.126 0.101 0.079 0.068 0.065 0.058 0.057 0.055 0.053 0.052

34.32 30.79 23.91 22.84 21.25 15.46 10.04 15.48 14.02 11.52

0.156 0.097 0.084 0.067 0.062 0.060 0.058 0.058 0.055 0.053

48.10 29.23 25.52 20.22 17.23 11.84 12.33 11.81 10.74 9.82

0.166 0.102 0.086 0.072 0.068 0.066 0.065 0.062 0.061 0.058

52.94 32.52 27.18 21.62 19.00 26.51 16.70 11.15 9.50 7.97

0.165 0.104 0.087 0.074 0.067 0.062 0.060 0.058 0.057 0.062

50.45 30.90 22.68 19.00 15.71 23.87 11.29 9.46 8.30 8.63

Standard uncertainty u is u(T) = 0.1 K. 2051

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053

Journal of Chemical & Engineering Data

Article

Figure 9. Comparisons between experimental data and calculated data by the UNIFAC model for the ternary system of water (1) + benzyl alcohol (2) + methylbenzene (3) at 303.2 K. ○- - -○, experimental data tie line; □- - -□, calculated data tie line. xi is the mole fraction of component i in the water (1) + benzyl alcohol (2) + methylbenzene (3) solvent mixtures.

Figure 10. Correlated LLE data and literature data for the ternary water (1) + benzyl alcohol (2) + methylbenzene (3) system at 101.3 kPa: ■---■, literature data tie line; □- - -□, correlated data tie line using the NRTL model. wi is the mass fraction of component i in water (1) + benzyl alcohol (2) + methylbenzene (3) solvent mixtures.

Table 6. r and q Values of the Pure Compounds for UNIQUAC Model

by NRTL and UNIQUAC methods show good consistency with the measured data. From the result, it was seen that the UNIQUAC method is better than the NRTL method. The obtained interaction parameters might be used in the calculation of LLE for the ternary system water + benzyl alcohol + methylbenzene as well as for the design and optimization of the related separation process.

component

r

q

water methylbenzene benzyl alcohol

0.9200 3.9228 4.0217

1.4000 2.9700 3.3220



Table 7. Optimized Temperature-Independent Binary Interaction Parameters in the NRTL Model for the Ternary Water (1) + Benzyl Alcohol (2) + Methylbenzene (3) System at (303.2 to 343.2) K i−j

αij

aij

bij

aji

bji

RMSD

1−2 1−3 2−3

0.2 0.2 0.2

0.43 5.73 0.81

2931.5 536.51 −1681.97

0.32 7.02 −1.22

2863.3 2444.60 6689.80

0.11 %

ASSOCIATED CONTENT

* Supporting Information S

Tables A and B. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Funding

The project was granted financial support from Key S&T Special Project of Zhejiang Province (2012C13007-2) and the Fundamental Research Funds for the Central Universities.

Table 8. Optimized Temperature-Independent Binary Interaction Parameters for the UNIQUAC Model for the Ternary Water (1) + Benzyl Alcohol (2) + Methylbenzene (3) System at (303.2 to 343.2) K i−j

aij

bij

aji

bji

RMSD

1−2 1−3 2−3

−2.93 −0.64 −1.89

893.12 −45.83 792.36

2.07 1.56 5.03

−715.68 −1291.90 −2018.90

0.09 %

Notes

The authors declare no competing financial interest.



REFERENCES

(1) Wang, X. W.; Wu, J. P.; Zhao, M. W.; Lv, Y. F.; Li, G. Y.; Hu, C. W. Partial Oxidation of Toluene in CH3COOH by H2O2 in the Presence of VO (acac)2 Catalyst. J. Phys. Chem. C 2009, 113, 14270−14278. (2) Wu, J. J. Studies on the Liquid-phase Selective Oxidation of Toluene with Hydrogen Peroxide and The Synthesis of Natural Benzaldehyde. Master Thesis, Hunan Normal University, 2012. (3) Liu, G. X. Liquid-phase Selective Oxidation of Toluene to Benzaldehyde with Hydrogen Peroxide. Master Thesis, Nankai University, 2011. (4) Thomas, E. R.; Newman, B. A.; Long, T. C.; Wood, D. A.; Eckert, C. A. Limiting Activity Coefficients of Nonpolar and Polar Solutes in Both Volatile and Nonvolatile Solvents by Gas Chromatography. J. Chem. Eng. Data 1982, 27, 399−405. (5) Anderson, F. E.; Prausnitz, J. M. Mutual Solubilities and Vapor Pressures for Binary and Ternary Aqueous Systems Containing Benzene, Toluene, m-Xylene, Thiophene and Pyridine in the Region 100−200 °C. Fluid Phase Equilib. 1986, 32, 63. (6) Stephenson, R. M. Mutual Solubilities: Water-Ketones, WaterEthers, and Water-Gasoline-Alcohols. J. Chem. Eng. Data 1992, 37, 80− 95.

However, the comparison results in Figure 10 convinced that the model parameters reported in this work were suitable and predictable for the ternary system water + benzyl alcohol + methylbenzene.

4. CONCLUSIONS Liquid−liquid equilibrium (LLE) data for the ternary water + benzyl alcohol + methylbenzene system have been measured at (303.2 to 343.2) K under atmospheric pressure. The reliability of the experimental LLE data has been checked according to the Othmer−Tobias and the Hand correlations. The distribution coefficient and separation factor were discussed. The experimental data were correlated by the NRTL and UNIQUAC activity coefficient models. The relevant interaction parameters were regressed with the experimental data. The predicted values 2052

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053

Journal of Chemical & Engineering Data

Article

Methyl Ethyl Acetate Ternary Systems. J. Chem. Eng. Data 2007, 52, 2171−2173. (28) Shen, B. W. Solid-liquid and Liquid-liquid Equilibria in the Oxidation of Cyclohexane to Adipic Acid. Master Thesis, Hunan University, 2013. (29) Othmer, D. F.; Tobias, P. E. Tie Line Correlation; Partial Pressures of Ternary Liquid Systems and Prediction of Tie Lines. Ind. Eng. Chem. 1942, 34, 693−700. (30) Hand, D. B. Dineric Distribution. J. Phys. Chem. 1930, 34, 1961− 2000. (31) Nelder, J. A.; Mead, R. A. Simplex Method for Function Minimization. Comput. J. 1965, 7, 308−313.

(7) Keith, C.; Zhou, Z.; Carl, L. Y. Determination of Henry’s Law Constants of Organics in Dilute Aqueous Solutions. J. Chem. Eng. Data 1993, 38, 546−550. (8) Chen, H. P.; Wanger, J. An Efficient and Reliable Gas Chromatographic Method for Measuring Liquid-Liquid Mutual Solubilities in Alkylbenzene + Water Mixtures: Toluene + Water from 303 to 373 K. J. Chem. Eng. Data 1994, 39, 475−479. (9) Ng, H. J.; Chen, C. J. Mutual Solubility in Water-Hydrocarbon Systems Gas Processors Association Mutual Solubility in WaterHydrocarbon Systems, 1995, RR−150, Tulsa, Oklahoma. (10) Brown, J. S.; Hallett, J. P.; Bush, D.; Eckert, C. A. Liquid-Liquid Equilibria for Binary Mixtures of Water + Acetophenone, + 1-Octanol, + Anisole, and + Toluene from 370 to 550 K. J. Chem. Eng. Data 2000, 45, 846−850. (11) Miller, D. J.; Hawthorne, S. B. Solubility of Liquid Organics of Environmental Interest in Subcritical (Hot/Liquid) Water from 298 to 473 K. J. Chem. Eng. Data 2000, 45, 78−81. (12) Jou, F. Y.; Mather, A. E. Liquid-Liquid Equilibria for Binary Mixtures of Water + Benzene, Water + Toluene, and Water + p-Xylene from 273 to 458 K. J. Chem. Eng. Data 2003, 48, 750−752. (13) Atik, Z.; Gruber, D.; Krummen, M.; Gmehling, J. Measurement of Activity Coefficients at Infinite Dilution of Benzene, Toluene, Ethanol, Esters, at Various Temperatures in Water Using the Dilutor Technique. J. Chem. Eng. Data 2004, 49, 1429−1432. (14) Marche, C.; Ferronato, C.; Jose, J. Apparatus for the Determination of Water Solubility in Hydrocarbon: Toluene and Alkylcyclohexanes (C6 to C8) from 30°C to 180°C. J. Chem. Eng. Data 2006, 51, 355−359. (15) Valtz, A.; Coquelet, C.; Richon, D. Solubility Data for Toluene in Various Aqueous Alkanolamine Solutions. J. Chem. Thermodyn. 2007, 39, 426−432. (16) Neely, B. J.; Wagner, J.; Robinson, R.; Gasem, K. A. M. Mutual Solubility Measurements of Hydrocarbon-Water Systems Containing Benzene, Toluene, and 3-Methylpentane. J. Chem. Eng. Data 2008, 53, 165−174. (17) Saien, J.; Mozafarvandi, M.; Daliri, S.; Norouzi, M. (LiquidLiquid) Equilibria for the Ternary (Water + Acetic Acid + Toluene) System at Different Temperatures: Experimental Data and Correlation. J. Chem. Thermodyn. 2013, 57, 76−81. (18) Javad, S.; Maryam, M.; Shabnam, D.; Mahdi, N. (Liquid + Liquid) Equilibria for the Ternary (Water + Acetic + Toluene) System at Different Temperatures: Experimental Data and Correlation. J. Chem. Eng. Thermodyn. 2013, 57, 76−81. (19) Stephenson, R. M.; Stuart, J. Mutual Binary Solubilities: WaterAlcohols and Water-Esters. J. Chem. Eng. Data 1986, 31, 56−70. (20) Solimo, H. N.; Gramajo de Doz, M. B. Liquid-Liquid Equilibrium, Densities, Viscosities, Refractive Indices, and Excess Properties of the Ternary System Water + 4-Hydroxy-4-Methyl-2-Pentanone + Benzyl Alcohol at 298.15 K. J. Chem. Eng. Data 1995, 40, 563−566. (21) Solimo, H. N.; Gramajo de Doz, M. B. Influence of Temperature on the Liquid-to-Liquid Extraction of 4-Hydroxy-4-Methyl-2-Pentanone from Aqueous Solutions with Benzyl Alcohol. Fluid Phase Equilib. 1995, 107, 213−227. (22) Sayar, A. A. Liquid-Liquid Equilibria of Some Water + 2-Propanol + Solvent Ternaries. J. Chem. Eng. Data 1991, 36, 61−65. (23) Banerjee, S. Solubility of Organic Mixtures in Water. Environ. Sci. Technol. 1984, 18, 587−591. (24) Skrzecz, A.; Shaw, D. G.; Maczynski, A. IUPAC-NIST Solubility Data Series 69. Ternary Alcohol-Hydrocarbon-Water Systems. J. Phys. Chem. Ref. Data 1999, 28, 983. (25) Renon, H.; Prausnitz, J. M. Estimation of Parameters for NRTL Equation for Excess Gibbs Energy of Strongly Non-Ideal Liquid Mixtures. Ind. Eng. Chem. Process Des. Dev. 1969, 8, 413−419. (26) Abrams, D. S.; Prausnitz, J. M. Statistical Thermodynamics of Liquid Mixtures: A New Expression for the Excess Gibbs Energy of Partly or Completely Miscrible System. AIChE J. 1975, 21, 116−128. (27) Wang, L. J.; Cheng, Y. W.; Li, X. Liquid-Liquid Equilibria for the Acetic Acid + Water + Amyl Acetate and Acetic Acid + Water + 22053

dx.doi.org/10.1021/je500195p | J. Chem. Eng. Data 2014, 59, 2045−2053