Solvent Extraction Research and Development, Japan, Vol. 22, No 1, 1 – 15 (2015) Liquid-Liquid Equilibria of Three Ternary Systems: Glycerol + Acetone + Water, Glycerol + 1, 4-Dioxane + Water, and Glycerol + Acetonitrile + Water Hirotake KATAYAMA* and Tomoya SATOH Department of Applied Chemistry, Hosei University, 3-7-2 Kajinocho, Koganei, Tokyo 184-8584, Japan (Received August 31, 2014, Accepted October 13, 2014) Liquid–liquid equilibria of three ternary systems {glycerol + acetone + water}, {glycerol + 1, 4-dioxane + water}, and {glycerol + acetonitrile + water} were measured at temperatures of 288.15, 298.15, and 308.15 K. The results of the first two systems were satisfactorily correlated with the UNIQUAC equation with average deviations of 0.74 and 0.81%, respectively. Those of the last system were well correlated with a deviation of 2.49%. Glycerol was verified as a suitable solvent to remove water from acetone + water and acetonitrile + water solutions, through investigation of the distribution coefficients and selectivity for water, combining vapor-liquid equilibria of the systems {acetone + water} and {acetonitrile + water} at atmospheric pressure. 1. Introduction Solvent extraction is an important separation technique for organic compounds. The operation requires reliable knowledge of the liquid-liquid equilibrium (LLE) between the mixture to be separated and the solvents selected. Because of strong polarity differences between the molecular species of the mixture and those of the solvents, LLE predictions are much more difficult than those for vapor-liquid equilibrium (VLE). Since the VLE of two binary mixtures of {1, 4-dioxane and water} and {acetonitrile and water} give azeotropic mixtures at atmospheric pressure, the mixtures cannot be separated by conventional distillation. The mixture of {acetone and water} has no azeotrope; however the relative volatility of the mixture shows a near unity value in the high purity region for acetone. To obtain high-purity acetone from an aqueous mixture, a distillation column with a large number of theoretical plates is required. Glycerol is partially miscible with acetone, 1, 4-dioxane, and acetonitrile, but arbitrarily miscible with water. In addition, it has a high normal boiling point (563.7 K) and a high density (1261.3 kg/m 3 at 293 K) [1]. After extraction of acetone, 1, 4-dioxane, and acetonitrile with glycerol, the separation of the mixtures and glycerol by distillation is easy. In this work, in order to determine the feasibility of glycerol as a solvent to eliminate water from aqueous solutions containing acetone, 1, 4-dioxane, and acetonitrile, the LLE of these three systems were measured at temperatures of 288.15, 298.15, and 308.15 K, and their phase diagrams were determined. As far as we know, the LLE of the glycerol + acetone + water system has been measured at 293.15 and 298.15 K by Hampe and Schermuly [2], and Krishna et al. [3], respectively, but the LLE of the last two systems have not been studied. - 1 -
15
Embed
Solvent Extraction Research and Development, Japan Vol. 22 ...1)/22_1.pdf · Liquid – liquid equilibria of three ternary systems {glycerol + acetone + water}, {glycerol + 1, 4-dioxane
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Solvent Extraction Research and Development, Japan, Vol. 22, No 1, 1 – 15 (2015)
Liquid-Liquid Equilibria of Three Ternary Systems: Glycerol + Acetone + Water, Glycerol +
1, 4-Dioxane + Water, and Glycerol + Acetonitrile + Water
Hirotake KATAYAMA* and Tomoya SATOH
Department of Applied Chemistry, Hosei University, 3-7-2 Kajinocho, Koganei, Tokyo 184-8584, Japan
(Received August 31, 2014, Accepted October 13, 2014)
Liquid–liquid equilibria of three ternary systems {glycerol + acetone + water}, {glycerol + 1, 4-dioxane +
water}, and {glycerol + acetonitrile + water} were measured at temperatures of 288.15, 298.15, and 308.15
K. The results of the first two systems were satisfactorily correlated with the UNIQUAC equation with
average deviations of 0.74 and 0.81%, respectively. Those of the last system were well correlated with a
deviation of 2.49%. Glycerol was verified as a suitable solvent to remove water from acetone + water and
acetonitrile + water solutions, through investigation of the distribution coefficients and selectivity for water,
combining vapor-liquid equilibria of the systems {acetone + water} and {acetonitrile + water} at
atmospheric pressure.
1. Introduction
Solvent extraction is an important separation technique for organic compounds. The operation requires
reliable knowledge of the liquid-liquid equilibrium (LLE) between the mixture to be separated and the
solvents selected. Because of strong polarity differences between the molecular species of the mixture and
those of the solvents, LLE predictions are much more difficult than those for vapor-liquid equilibrium
(VLE).
Since the VLE of two binary mixtures of {1, 4-dioxane and water} and {acetonitrile and water} give
azeotropic mixtures at atmospheric pressure, the mixtures cannot be separated by conventional distillation.
The mixture of {acetone and water} has no azeotrope; however the relative volatility of the mixture shows
a near unity value in the high purity region for acetone. To obtain high-purity acetone from an aqueous
mixture, a distillation column with a large number of theoretical plates is required.
Glycerol is partially miscible with acetone, 1, 4-dioxane, and acetonitrile, but arbitrarily miscible with
water. In addition, it has a high normal boiling point (563.7 K) and a high density (1261.3 kg/m3 at 293 K)
[1]. After extraction of acetone, 1, 4-dioxane, and acetonitrile with glycerol, the separation of the mixtures
and glycerol by distillation is easy.
In this work, in order to determine the feasibility of glycerol as a solvent to eliminate water from
aqueous solutions containing acetone, 1, 4-dioxane, and acetonitrile, the LLE of these three systems were
measured at temperatures of 288.15, 298.15, and 308.15 K, and their phase diagrams were determined. As
far as we know, the LLE of the glycerol + acetone + water system has been measured at 293.15 and 298.15
K by Hampe and Schermuly [2], and Krishna et al. [3], respectively, but the LLE of the last two systems
have not been studied.
- 1 -
The experimental results were correlated with the UNIQUAC equation. The correlations are discussed.
Distribution coefficients for water and degree of selectivity were also investigated, combining the VLEs of
the related binary mixtures at atmospheric pressure.
2. Experimental
2.1 Reagents
The organic reagents used (glycerol, acetone, 1, 4-dioxane, acetonitrile, ethylene glycol, and
1-propanol) were purchased from Wako Pure Chemical Industries, Ltd., all of which were stated to be
minimum assays of 99.5 mass %. After dehydrated with molecular sieves 4A, the reagents were used
without further purification. Tap water was distilled before use.
2.2 Apparatus and Procedure
The experimental method and apparatus are similar to those already described [4,5]. Temperature
was measured using an F25 platinum resistance thermometer (Automatic System Laboratories, Ltd.) with a
stated accuracy of 0.03 K and a stated resolution of 0.001 K. The temperature fluctuation of the water bath
was within 0.08 K. Each 50-mL mixture of glycerol, acetone (or 1, 4-dioxane or acetonitrile), and water
was poured into six 50-cm3 flasks sealed with glass stopcocks, and the flasks were immersed in the bath.
The mixtures were agitated for 4 h, and then allowed to settle for more than 12 h. Samples (1-mL) were
withdrawn from both phases with long needle syringes, and then 1-mL of ethylene glycol for the systems
{glycerol (1) + acetone (2) + water (3)} and {glycerol (1) + acetonitrile (2) + water(3)}, or 1-mL
1-propanol for the system {glycerol (1) + 1, 4-dioxane (2) + water (3)},was added to the sample as an
internal standard.
The samples were analyzed by gas chromatography on a Shimadzu GC-8A Model unit with a thermal
conductivity detector, in which a 2-m column of Porapak-Q was installed. Helium gas was used as the
carrier gas.
The analytical limit for each component was 0.0002 mole fraction. However, because of the
reproducibility of the numerical values, the concentrations were reported to three decimal places.
3. Results and Discussion
3.1 Experimental data
The results for the LLE of the ternary systems {glycerol (1) + acetone (2) + water (3)}, {glycerol (1)
+ 1, 4-dioxane (2) + water (3)}, and {glycerol (1) + acetonitrile (2) + water (3)} are given in Tables 1-3 and
in Figures 2-4 (shown as filled circles), respectively. The compositions are given in terms of mole fractions.
Figure 1 shows the experimental results at 298.15 K for the system {glycerol (1) + acetone (2) + water (3)},
which are in good agreement with those of Krishna et al.
Figures 2-4 show that the magnitudes of the two-phase areas for the three ternary systems are in the
order of {glycerol + acetonitrile + water} > {glycerol + acetone + water} > {glycerol + 1, 4-dioxane +
water}, and those of their tie-line slopes are also in the same order. From these data, the polarity of the
substances used is considered to be in the order of glycerol > water >1, 4-dioxane > acetone > acetonitrile.
- 2 -
Figure 1. Comparison of the experimental results with those of the literature for the LLE of the system
{glycerol (1) + acetone (2) + water (3)} at 298.15 K. Solid and dotted lines show the tie lines from this
work, and those of Krishna et al., respectively.
3.2 Correlation with the UNIQUAC equation
The experimental results were correlated by the UNIQUAC equation [6]. The parameters A12 and A21
for the binary systems{glycerol (1) + acetone (2)}, {glycerol (1) + 1, 4-dioxane (2)}, and {glycerol (1) +
acetonitrile (2)} were obtained by means of the least-squares method, using the respective binary data of
Tables 1-3. The A12 and A21 values obtained are shown in the first and second columns of Table 4. The four
remaining Aij parameters (i.e., A23, A32, A31, and A13) for the systems {glycerol (1) + acetone (2) +water (3)},
{glycerol (1) + 1, 4-dioxane (2) + water (3)}, and {glycerol (1) + acetonitrile (2) + water (3)} were
determined by minimizing the following objective function, Fobj, using the modified Marquardt method [7]:
2 32
, , , , , ,
1 1 1
( , )( )N
obj k j n exptl. k j n cal.
n k j
F w k j x x
(1)
where k is the number of phases (1 and 2), j is the number of components (1, 2, and 3), n is the number of
data (1 to N = 8-16), w (k, j) is a weighting factor of phase k and component j, and x is the mole fraction.
Subscripts (exptl. and cal.) represent experimental and calculated values, respectively. Parameter Aij is in
Kelvin.
- 3 -
Table 1. Liquid-liquid equilibria for the system {glycerol (1) + acetone (2) + water (3)}
Temp. acetone-rich phase (I) glycerol-rich phase (II) β 3 S
[K] x 1 x 3 x 2 x 3
288.15 0.0274 0.0000 0.144 0.0000
0.0238 0.0024 0.142 0.0151 6.40 43.9
0.0262 0.0119 0.146 0.0515 4.31 28.4
0.0275 0.0296 0.139 0.149 5.05 34.2
0.0290 0.0410 0.147 0.177 4.32 27.2
0.0302 0.0645 0.153 0.289 4.47 26.4
0.0303 0.0874 0.163 0.351 4.01 21.7
0.0356 0.110 0.175 0.391 3.56 17.4
0.0356 0.120 0.179 0.414 3.44 16.2
0.0371 0.139 0.190 0.434 3.13 13.6
0.0420 0.171 0.204 0.465 2.73 10.5
0.0440 0.196 0.221 0.489 2.49 8.56
0.0501 0.217 0.237 0.493 2.27 7.01
298.15 0.0301 0.0000 0.157 0.0000
0.0267 0.0011 0.152 0.0065 6.06 38.8
0.0259 0.0021 0.151 0.0126 6.04 38.8
0.0325 0.0110 0.151 0.0404 3.67 23.3
0.0266 0.0255 0.153 0.118 4.63 28.7
0.0323 0.0355 0.150 0.159 4.49 28.0
0.0318 0.0442 0.158 0.210 4.76 27.8
0.0357 0.0544 0.162 0.236 4.34 24.4
0.0414 0.0816 0.163 0.289 3.54 19.1
0.0378 0.0973 0.180 0.343 3.52 16.9
0.0489 0.131 0.183 0.386 2.95 13.2
0.0511 0.155 0.193 0.419 2.69 11.1
0.0549 0.181 0.213 0.448 2.47 8.85
0.0517 0.189 0.219 0.458 2.43 8.40
0.0595 0.220 0.242 0.475 2.15 6.41
0.0643 0.217 0.235 0.470 2.16 6.62
308.15 0.0395 0.0000 0.169 0.0000
0.0378 0.0023 0.178 0.0068 2.88 15.6
0.0401 0.0116 0.166 0.0403 3.48 19.9
0.0410 0.0359 0.166 0.144 4.00 22.3
0.0446 0.0592 0.168 0.222 3.75 20.0
0.0481 0.0768 0.178 0.280 3.64 18.0
0.0508 0.107 0.188 0.336 3.13 14.1
0.0547 0.132 0.198 0.378 2.87 11.8
0.0565 0.146 0.211 0.401 2.75 10.4
0.0609 0.171 0.218 0.419 2.45 8.62
0.0683 0.205 0.236 0.443 2.16 6.64
0.0788 0.252 0.266 0.459 1.82 4.60
- 4 -
Table 2. Liquid-liquid equilibria for the system {glycerol (1) + 1, 4-dioxane (2) + water (3)}
Temp. 1, 4-dioxane-rich phase (I) glycerol-rich phase (II) β 3 S
[K] x 1 x 3 x 2 x 3
288.15 0.0467 0.0000 0.171 0.0000
0.0440 0.0228 0.172 0.120 5.26 28.5
0.0476 0.0414 0.182 0.216 5.22 26.1
0.0455 0.0600 0.196 0.280 4.67 21.3
0.0479 0.0771 0.204 0.332 4.30 18.4
0.0507 0.0914 0.211 0.362 3.97 16.1
0.0520 0.114 0.230 0.398 3.50 12.7
0.0567 0.154 0.261 0.441 2.86 8.65
298.15 0.0547 0.0000 0.192 0.0000
0.0565 0.0282 0.195 0.119 4.22 19.8
0.0593 0.0495 0.208 0.214 4.33 18.6
0.0606 0.0675 0.216 0.274 4.05 16.3
0.0631 0.0906 0.227 0.321 3.55 13.2
0.0665 0.112 0.240 0.358 3.20 10.9
0.0701 0.138 0.259 0.388 2.81 8.58
0.0813 0.196 0.305 0.421 2.14 5.08
308.15 0.0714 0.0000 0.210 0.0000
0.0727 0.0307 0.226 0.124 4.22 19.8
0.0745 0.0585 0.234 0.212 4.33 18.6
0.0744 0.0846 0.248 0.265 4.05 16.3
0.0789 0.113 0.258 0.310 3.55 13.2
0.0844 0.137 0.282 0.347 3.20 10.9
0.0936 0.171 0.307 0.364 2.81 8.58
0.109 0.234 0.393 0.370 2.14 5.08
- 5 -
Table 3. Liquid-liquid equilibria for the system {glycerol (1) + acetonitrile (2) + water (3)}
Temp. acetonitrile-rich phase (I) glycerol-rich phase (II) β 3 S
[K] x 1 x 3 x 2 x 3
288.15 0.0069 0.0000 0.145 0.0000
0.0083 0.0074 0.144 0.0154 2.08 14.2
0.0067 0.0076 0.138 0.0304 3.99 28.6
0.0076 0.0210 0.135 0.175 8.33 59.8
0.0081 0.0409 0.123 0.348 8.52 65.7
0.0091 0.0527 0.125 0.421 7.98 60.0
0.0097 0.0700 0.120 0.492 7.03 53.9
0.0097 0.0792 0.123 0.532 6.71 49.9
0.0093 0.103 0.123 0.594 5.77 41.8
0.0098 0.111 0.125 0.618 5.56 39.1
0.0093 0.128 0.127 0.646 5.06 34.5
0.0093 0.146 0.129 0.677 4.62 30.2
0.0099 0.155 0.130 0.685 4.41 28.3
0.0100 0.171 0.136 0.707 4.14 25.0
0.0111 0.226 0.145 0.738 3.26 17.2
298.15 0.0102 0.0000 0.161 0.0000
0.0114 0.0078 0.156 0.0198 2.54 16.0
0.0098 0.0237 0.148 0.168 7.08 46.2
0.0133 0.0481 0.139 0.344 7.15 48.3
0.0127 0.0642 0.141 0.409 6.37 41.8
0.0113 0.0819 0.138 0.479 5.85 38.5
0.0109 0.0936 0.139 0.517 5.52 35.5
0.0128 0.123 0.142 0.584 4.76 29.0
0.0104 0.135 0.144 0.612 4.52 26.7
0.0124 0.153 0.145 0.632 4.14 23.9
0.0123 0.178 0.153 0.664 3.73 19.8
0.0118 0.184 0.154 0.678 3.69 19.3
0.0126 0.263 0.179 0.716 2.73 11.0
308.15 0.0166 0.0000 0.179 0.0000
0.0186 0.0060 0.175 0.0196 3.28 18.2
0.0169 0.0281 0.166 0.162 5.79 33.2
0.0170 0.0545 0.157 0.326 5.98 35.3
0.0175 0.0727 0.148 0.405 5.57 34.3
0.0172 0.0990 0.159 0.468 4.72 26.2
0.0144 0.119 0.161 0.508 4.26 22.9
0.0184 0.145 0.164 0.563 3.88 19.8
0.0185 0.159 0.167 0.591 3.71 18.2
0.0177 0.183 0.173 0.614 3.37 15.5
0.0184 0.218 0.181 0.637 2.92 12.3
0.0167 0.238 0.185 0.646 2.72 11.0
0.0192 0.331 0.227 0.671 2.03 5.80
- 6 -
Figure 2. LLE for the system {glycerol (1) + acetone (2) + water (3)}. Solid lines: experimental tie lines.
Dotted lines: predicted tie lines.
- 7 -
Figure 3. LLE for the system {glycerol (1) + 1, 4-dioxane (2) + water (3)}. The solid and dotted lines are
defined as in Figure 2. The symbol (□) represents the azeotropic point of the system {1, 4-dioxane + water}
at atmospheric pressure [9]. The broken line passes through the azeotropic point and the point for pure
glycerol.
- 8 -
Figure 4. LLE for the system {glycerol (1) + acetonitrile (2) + water (3)} in the absence of the weighting
factor. The solid and dotted lines are defined as in Figure 2. The symbol (□) represents the azeotropic point
of the system {acetonitrile + water } at atmospheric pressure.
- 9 -
Figure 5. LLE for the system {glycerol (1) + acetonitrile (2) + water (3)} in the presence of the weighting
factor, w(1,1) = 10,000.0. The solid and dotted lines are defined as in Figure 2.
3.2.1 Correlation with the weighting factor w (k, j) = 1.0 (no weighting factor)
As the w (k , j) (k = 1, 2; j = 1, 2, 3) values were set at unity, through minimizing Fobj, the four
- 10 -
parameters of Aij obtained for the three ternary systems are listed along with the root-mean-square
deviations (rmsd) in Table 4. Figures 2-4 show the comparison between the experimental values () and
the correlated ones ().
Table 4. UNIQUAC parameters of A ij for the three ternary systems correlated
without the weighting factor{w (k, j ) = 1.0, k = 1-2 and j = 1-3}
T [K] A 12 [K] A 21 [K] A 23 [K] A 32 [K] A 31 [K] A 13 [K] rmsd* remarks