Acetonitrile Dehydration via Extractive Distillation ... - ACS Publications

Article pubs.acs.org/jced

Cite This: J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Acetonitrile Dehydration via Extractive Distillation Using Low Transition Temperature Mixtures as Entrainers Bandhana Sharma, Neetu Singh,* Tarun Jain, Jai Prakash Kushwaha, and Parminder Singh Department of Chemical Engineering, Thapar Institute of Engineering and Technology, Patiala-147004, India

Downloaded via KAOHSIUNG MEDICAL UNIV on July 9, 2018 at 14:13:17 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Low transition temperature mixtures (LTTMs) are versatile alternatives to ILs. They share many properties with ILs, so they become a suitable choice for entrainers in extractive distillation processes. In this study, glycolic acid and choline chloride in a 3:1 molar ratio (GC3:1) were synthesized and explored as entrainers for separation of acetonitrile + water azeotropic mixtures. Isobaric vapor−liquid equilibrium data for the pseudobinary mixtures of ACN + GC3:1 and water + GC3:1 were measured at atmospheric pressure (101.32 kPa). For the pseudoternary system ACN + water + GC3:1, also VLE data were measured at different GC3:1 mole fractions of 0.05, 0.1, and 0.15. The thermodynamic modeling of these systems was performed using the nonrandom two-liquid (NRTL) model. Furthermore, a study was conducted for synthesized GC3:1 recoverability. A good agreement were found between experimental data and predicted values for these systems. Results showed that LTTM (GC3:1) eliminated the acetonitrile + water azeotrope by manipulating the relative volatility of the acetonitrile + water mixture. Therefore, LTTM (GC3:1) can be concluded as an efficient entrainer for the separation of an acetonitrile + water azeotropic mixture by extractive distillation.

1. INTRODUCTION Acetonitrile (ACN) is widely applied in various areas such as pharmaceuticals,1 chromatography,2 organic synthesis,3 and light-sensitive materials.4 In chemical industries, there are several processes which produce mixtures of acetonitrile and water. Inhalation of higher concentrations of acetonitrile causes irritation of mucous membranes and can lead to many other health related problems.5 Acetonitrile has been reported to be toxic similar to severe cyanide poisoning. However, somewhat delayed toxic effects of acetonitrile have been observed comparative to inorganic cyanides or various other saturated nitriles. The environment may be contaminated in acetonitrile if the effluents from industries/municipal wastewater treatment plant are discharged to water bodies. Due to such significant demand and high toxicity of this compound, separation of acetonitrile from water is necessary. However, its separation/recovery from waste effluents is often difficult, as acetonitrile + water solution forms a minimum-boiling azeotrope with a composition of 82.2 mass % (67.4 mol %) acetonitrile at 76.5 °C and standard atmospheric pressure.6 Various separation methods such as extractive distillation,7−10 azeotropic distillation,11 pressure-swing distillation,12 liquid−liquid extraction,13,14 pervaporation,15−17 and adsorption18,19 have been reported to eliminate the acetonitrile + water azeotropic mixture to recover acetonitrile from its aqueous solution. Extractive distillation using solvents and salts was widely used for the separation of such mixtures. However, salt distillation required high-quality © XXXX American Chemical Society

equipment and the use of organic solvent as an entrainer demanded high energy.20 It is obvious that an economical and effective design of extractive distillation is based on the selection of a good solvent. Some typical solvents had been used previously for the separation of an acetonitrile + water mixture such as ethylene glycol21,22 and butyl acetate.23 However, the main problems associated with the use of ethylene glycol are that it does not mix evenly with the azeotropic mixture and the solvent dose requirement is also quite high. On the other hand, due to the low boiling point of butyl acetate, it is not easy to recycle it. Zhang et al.7 used dimethyl sulfoxide (DMSO) as an entrainer for the dehydration of an acetonitrile + water mixture. DMSO could successfully eliminate the azeotrope and was found to be an effective solvent for separating the acetonitrile + water azeotropic binary system. From the last two decades, application of ionic liquids (ILs) as an entrainer for azeotropic mixture separation significantly encouraged the researchers due to their distinctive prominent features. Different types of ILs such as 1-butyl-3-methylimidazolium chloride ([Bmim][Cl]), 1-butyl-3-methylimidazolium tetrafluoroborate ([Bmim][BF4]), and 1-butyl-3- methylimidazolium dibutyl phosphate ([Bmim][DBP]) have been explored for acetonitrile + water separation.24,25 The use of ILs for extractive distillation is Received: March 21, 2018 Accepted: June 20, 2018

A

DOI: 10.1021/acs.jced.8b00228 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

Table 1. Chemicals Used name

CAS no.

source

purity (wt %)

purification method

acetonitrile glycolic acid choline chloride Karl Fischer reagent isopropyl alcohol double distilled water

75-05-8 79-14-1 67-48-1 7446-09-5 67-63-0

Merck chemicals Spectrochem TCI chemicals Merck chemicals Sigma-Aldrich our laboratory

≥99.8% >98% >98% >99% >99%

none vacuum drying vacuum drying none none

restricted due to the complex synthesis process, difficulties in purification, high cost, potential toxicity, and poor degradability. Thus, there is a need to seek a more suitable solvent which can replace the drawbacks associated with these traditional solvents and ILs. Newly emerging solvents, low transition temperature mixtures (LTTMs), are versatile alternatives to ILs. LTTMs are eutectic mixtures of two or more solid, hydrogen-bond donor (HBD)

and hydrogen-bond acceptor (HBA) compounds, forming liquids with much lower melting point compared to individual HBDs and HBAs used, and have properties comparable to those of ILs. Unlike ILs, LTTMs are easy to prepare in a pure form. Production cost and simultaneous recyclability or biocompatibility/biodegradability are comparable to ILs. This new family of solvents was presented for the first time by Abbott et al. in 2003 and given the name deep eutectic solvents (DESs).26 Many of such DESs do not demonstrate the eutectic melting point, rather a glass transition temperature, and do not represent/ envelope the entire group of such solvents.27 For the first time, these mixtures were explored as possible entrainers for extractive ́ distillation by Rodriguez et al.28,29 The aim of this work is to investigate the influence of glycolic acid + choline chloride (in 3:1 molar ratio) (GC3:1) as an entrainer for separation of an ACN + water azeotropic binary mixture by extractive distillation. Isobaric vapor−liquid equilibrium (VLE) data for the pseudobinary systems (water + GC3:1 and ACN + GC3:1) and pseudoternary system (ACN + water + GC3:1) were generated at 101.32 kPa for the first time. Isobaric VLE measurements have been performed for different LTTM concentrations, 5, 10, and 15% (mol/mol). The experimental VLE data were successfully correlated using the nonrandom two-liquid (NRTL) model.30 For the NRTL fitting,

Figure 1. Hydrogen bonding mechanism in glycolic acid:choline chloride 3:1 (GC3:1).

Figure 2. Schematic diagram of a modified Othmer type recirculation still (1, boiling still; 2, heating mantle; 3, magnetic stirrer; 4, liquid phase sampling port; 5, 6, thermometers; 7, air-cooled condenser; 8, water-cooled recirculation condenser; 9, condensate sampling port; 10, water-cooled condenser). B



Article

Table 2. Experimental VLE Data and Correlated Results of Pseudobinary Subsystems, Activity Coefficients γi, Deviation in Activity Coefficients Δγi, and Deviation in Equilibrium Temperature ΔT at 101.32 kPaa water (1) + GC (3:1, mol/mol) (2)

Figure 3. Comparison of isobaric T−x−y data for the IPA (1)−water (2) system at atmospheric pressure: ▲, this work; ●, ref 35.

the LTTM was considered as a single compound (HBD + HBA pair). Furthermore, the recoverability of the synthesized LTTM in this study was also investigated.

2. EXPERIMENTAL SECTION 2.1. Chemicals. The chemicals used were acetonitrile, isopropyl alcohol, glycolic acid, choline chloride, and double distilled water. Table 1 summarizes the list of chemicals, including their purity level and supplier. Acetonitrile (HPLC grade) with a high purity of ≥99.8% was supplied by Merck chemicals. Isopropyl alcohol with a purity of >99% was supplied by Sigma-Aldrich, and glycolic acid (>98%) was supplied by Spectrochem. Both acetonitrile and isopropyl alcohol were used as procured without further purification. Choline chloride was provided by TCI Chemicals with a minimum purity of >98%. Due to the hygroscopic nature of glycolic acid and choline chloride, both were first dried in a vacuum oven before processing the LTTM. The trace water content in these chemicals was analyzed using Karl Fischer titration analysis (type Esico 1760) and was observed to be less than 0.3 wt %. 2.2. Synthesis and Characterization of LTTM (GC3:1). The LTTM, GC3:1, was prepared in the same manner as reported earlier.31 A Mettler Toledo electronic balance with a precision of ±0.0001 g was used. Both the glycolic acid and choline chloride were introduced into a closed 50 mL glass flask at a molar ratio of 3:1. The standard uncertainties in moles of glycolic acid and choline chloride were 4.158 × 10−6 and 1.012 × 10−6, respectively, in the mixture. Both were homogeneously mixed, and afterward, the mixture was heated using a thermostatic oil bath and stirred once the melting of the mixture provided enough liquid. The temperature of the oil bath was controlled at 70 °C (±0.1 °C) with a temperature controller (IKA ETS-D5). Once a colorless and transparent liquid formed, the heating was stopped and liquid was allowed to cool gradually to ambient temperature. The water content in the prepared GC3:1 was analyzed using the Karl Fischer titration method (Esico, 1760). Density and viscosity were also measured. The thermal stability of synthesized GC3:1 was checked by TGA analysis (PerkinElmer, STA6000) and is shown in Figure S1. It has been characterized by FT-IR ((PerkinElmer), and the

mole % of GC3:1

x1

γexp 1

Δγ1b

ΔTc (K)

S. No.

T (K)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

373.1 373.3 373.5 373.8 374.2 374.5 374.9 375.6 376.5 377.7 379.3 381.2 383.7 386.0 388.0

S. No.

T (K)

mole % of GC3:1

x1

γexp 1

Δγ1b

ΔTc (K)

1 2 3 4 5 6 7 8 9 10 11 12 13 14

354.7 355.5 356.7 357.4 358.3 359.4 360.7 362.1 363.7 365.2 366.8 369.9 373.2 376.5

0.00 4.11 8.30 12.57 16.92 21.35 25.87 30.48 35.19 39.99 44.88 49.88 54.99 60.20

1.000 0.956 0.916 0.872 0.830 0.784 0.740 0.694 0.647 0.601 0.550 0.501 0.452 0.398

1.000 1.019 1.025 1.054 1.077 1.103 1.124 1.150 1.177 1.212 1.263 1.269 1.281 1.328

0.0000 0.0126 0.0046 0.0117 0.0103 0.0064 −0.0023 −0.0078 −0.0120 −0.0050 0.0167 −0.0030 −0.0132 0.0126

0.00 −0.41 −0.15 −0.37 −0.32 −0.19 0.07 0.23 0.35 0.14 −0.46 0.08 0.37 −0.35

0.00 1.000 1.000 0.0000 1.44 0.985 1.009 0.0093 2.99 0.969 1.018 0.0117 4.67 0.954 1.023 0.0101 6.49 0.936 1.027 0.0042 8.47 0.913 1.042 0.0109 10.63 0.893 1.050 0.0053 13.00 0.870 1.051 0.0021 15.61 0.844 1.050 0.0014 18.50 0.814 1.044 0.0058 21.72 0.784 1.025 0.0038 25.33 0.747 1.008 0.0091 29.39 0.706 0.980 −0.0164 34.01 0.660 0.969 −0.0260 39.30 0.608 0.984 −0.0087 ACN (1) + GC (3:1, mol/mol) (2)

0.00 −0.26 −0.32 −0.28 −0.12 −0.30 −0.14 −0.06 −0.04 −0.16 −0.11 −0.27 0.49 0.80 0.27

a Standard uncertainty u(x1) = 0.001, u(mole % of GC3:1) = 0.1, cal c exp − Tcal. u(T) = 0.1 K, u(P) = 0.05 kPa. bΔγ1 = γexp 1 − γ1 . ΔT = T The superscripts “exp” and “cal” represent calculated and experimental values, respectively.

spectra are shown in Figure 8a. In the FT-IR spectra, at lower wavenumber, three bands with peaks at 1734, 1193, and 1083 cm−1 were observed. This indicates the carbonyl group of glycolic acid and hydrogen bonding between the OH group of glycolic acid and Cl− of choline chloride, respectively. Near 3302 cm−1, a broad band was also observed exhibiting a OH stretching region with the existence of extensive hydrogen bonds between OH of glycolic acid/choline chloride and Cl− of choline chloride.32,33 On the basis of this, the possible hydrogen bonding between glycolic acid and choline chloride is shown in Figure 1, and the formula unit of prepared LTTM is (C5H14NO)+Cl−·3(CH2OH)−COOH. Hydrogen bond formation between the HBD and HBA decreases the freezing point of the mixture, which ultimately results in the formation of LTTM. 2.3. Apparatus and Procedure. Improved Othmer type recirculation still was used to measure isobaric VLE data at 101.3 kPa.34 Figure 2 represents the schematic of modified Othmer recirculation still currently used in the present work. C



Article

The setup consists of a boiling still (1) of 150 mL capacity in which a nichrome wire heater (2) (glass-sealed) is provided to boil the mixture. A magnetic stirrer (3) (REMI Instruments2MLH model, Bombay, India) was used to stir the still contents. The still has two side-angled openings: the left one (4) to draw the liquid phase sample and to accommodate a thermometer (5) and the right one to receive condensate recycle. Vapors generated in the still pass through an air-cooled condenser (7) in which a thermometer (6) is incorporated. The uncondensed vapors and condensate move to the water-cooled condenser (8), and condensate flows down from this condenser to the condensate sample bulb (9) which is connected with a condensate sampling port. Almost all of the vapor condenses in the watercooled recirculation condenser; if some uncondensed vapor is remaining, that undergoes condensation in another water-cooled condenser (10). An autotransformer was applied to supply the electricity to the nichrome wire heater (2). To measure the temperature at equilibrium, a precision and calibrated thermometer was used of 0.1 K uncertainty count. 2.4. Composition Analysis of Samples. The composition of the condensed vapor phase (ACN + water) at equilibrium was analyzed using a Bruker Gas chromatograph (SCION 456-GC) facilitated with FID (flame ionization detector) and TG-Bond Q column (30 m × 0.53 mm I.D., 20.0 μm film thickness) of Thermo Scientific. The detector temperature was set to 250 °C. Injections (1 μL) were performed in the split mode at a split ratio of 100:1 with an injector temperature of 200 °C. As carrier gas in GC, helium was used at a flow rate of 5.0 mL min−1. The oven temperature was fixed at 180 °C. No presence of LTTM was observed in the vapor phase. The acetonitrile mole fractions in the liquid phase were determined using the same GC method, while Karl Fischer moisture analysis (Esico model 1760) was used for determination of water content. While analyzing the liquid samples in GC, a glass liner (Agilent Technologies) was used before the column. The glass liner is generally used to prevent any viscous liquid from entering the column; otherwise, it can damage the GC column. LTTM was prevented from entering into the GC column by applying a clean glass liner after each gas chromatogram, and the same was found effective. In liquid samples, the mole fraction of LTTM was analyzed by applying the mass balance. A calibration, prepared from gravimetric standard solutions, was used to determine the equilibrium compositions of the samples. The uncertainty in the component mole fraction of the vapor and liquid phases was 0.001.

Figure 4. Effect of GC3:1 on the normal boiling point of ACN and water at 101.32 kPa. Symbols: experimental data for ACN (▲) and water (●). Solid lines: calculations based on the NRTL model.

The isobaric VLE data for the pseudoternary system of ACN (1) + water (2) + GC3:1 (3) at different entrainer contents (5, 10, and 15 mol %) were experimentally measured at 101.32 kPa. Experimental data are reported in Table 3, where x1 and x1′ are the mole fractions of ACN in the liquid phase and the mole fraction of ACN in the liquid phase based on a GC3:1-free basis, respectively; y1 is the mole fraction of ACN in the vapor phase; and T is the equilibrium temperature. The T−x, y diagrams are plotted in Figure 5 to explore the salting-out effect triggered by GC3:1 at 5, 10, and 15 mol %, respectively. The x−y diagrams for the above experimental conditions are also represented by Figure 6. Nonvolatile entrainer addition (LTTM) increases the system equilibrium temperature. The more LTTM, GC3:1, addition in the system, the higher the equilibrium temperature is established. At the azeotropic point, the relative volatility for the ACN + water system is unity, and to destroy the azeotrope, the relative volatility should be greater than 1. We can see from Table 3 that GC3:1 is fairly effective to increase the relative volatility of the ACN + water system to greater than 1 in the entire

3. RESULTS AND DISCUSSION 3.1. Verification of the Apparatus. In order to verify the reliability of the experimental setup and method used in the present work, the VLE data of the IPA and water binary system were measured at atmospheric pressure using the present setup, and the same was compared with literature data.35 Table S1 and Figure 3 represent the experimental T−x, y data obtained in this work. It can be observed that the measured VLE data of the IPA and water system in this work are in good agreement with the literature data, indicating that the VLE apparatus used in the work is reliable. 3.2. Experimental VLE Data. The isobaric VLE data for the pseudobinary systems of water (1) + GC3:1 (2) and ACN (1) + GC3:1 (2) were measured at 101.32 kPa and are presented in Table 2. Figure 4 represents the enhancement in the normal boiling point Tb of water and ACN by addition of the nonvolatile component GC3:1. The boiling points of both were higher after addition of GC3:1. D



Article

Table 3. Experimental Isobaric VLE Data for the ACN (1) + Water (2) + GC3:1 (3) System, Experimental Activity Coefficients a,b γexp i , and Relative Volatilities α12 at 101.32 kPa S. No.

T (K)

x1′

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

356.0 356.5 357.1 358.2 359.0 359.2 361.0 362.1 363.5 364.3 366.4 367.0 368.6 369.1 371.5 372.0 372.5 373.0 373.9 374.2 374.8

1.000 0.893 0.873 0.813 0.723 0.662 0.533 0.433 0.412 0.329 0.222 0.211 0.187 0.162 0.132 0.122 0.119 0.102 0.100 0.0876 0.000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

357.0 357.6 357.6 358.4 359.1 359.3 361.7 362.5 364.0 364.9 366.3 366.9 367.8 371.0 372.0 372.8 373.5 374.0 374.1 374.2 374.8 375.5

1.000 0.912 0.870 0.821 0.765 0.693 0.569 0.496 0.433 0.390 0.354 0.327 0.294 0.214 0.200 0.182 0.165 0.158 0.138 0.119 0.088 0.000

1 2 3 4 5 6 7 8 9 10 11 12 13 14

358.3 358.3 358.5 358.7 359.2 359.4 364.3 368.3 370.4 370.9 371.1 372.1 373.2 373.8

1.000 0.932 0.867 0.824 0.765 0.702 0.456 0.312 0.271 0.267 0.234 0.221 0.199 0.187

x1

x2

y1

Glycolic Acid/ChCl 3:1 = 5 mol % 1.000 0.849 0.102 0.919 0.829 0.121 0.896 0.774 0.178 0.865 0.687 0.263 0.815 0.629 0.321 0.800 0.506 0.444 0.728 0.412 0.538 0.657 0.391 0.559 0.639 0.313 0.638 0.563 0.211 0.741 0.490 0.200 0.750 0.466 0.178 0.773 0.436 0.154 0.797 0.381 0.125 0.825 0.339 0.116 0.835 0.320 0.113 0.837 0.312 0.096 0.854 0.301 0.095 0.854 0.251 0.083 0.867 0.233 0.000 Glycolic Acid/ChCl 3:1 = 10 mol % 1.000 0.822 0.079 0.965 0.784 0.117 0.947 0.739 0.161 0.926 0.689 0.212 0.901 0.624 0.276 0.883 0.512 0.388 0.812 0.446 0.454 0.784 0.390 0.510 0.734 0.351 0.550 0.713 0.319 0.581 0.680 0.294 0.605 0.638 0.264 0.635 0.609 0.192 0.708 0.512 0.180 0.720 0.508 0.164 0.735 0.452 0.149 0.751 0.442 0.142 0.758 0.428 0.124 0.776 0.385 0.107 0.792 0.341 0.079 0.822 0.261 0.000 Glycolic Acid/ChCl 3:1 = 15 mol % 1.000 0.792 0.058 0.985 0.737 0.113 0.967 0.700 0.150 0.949 0.650 0.200 0.926 0.597 0.253 0.912 0.388 0.463 0.812 0.265 0.584 0.703 0.230 0.619 0.643 0.227 0.624 0.641 0.199 0.652 0.605 0.188 0.663 0.581 0.169 0.681 0.552 0.159 0.691 0.533 E

γexp 1

γexp 2

α12

1.023 1.003 1.003 1.039 1.108 1.187 1.276 1.251 1.346 1.636 1.609 1.621 1.612 1.646 1.656 1.634 1.823 1.507 1.581

1.492 1.577 1.336 1.193 1.049 0.963 0.959 0.923 0.949 0.882 0.892 0.861 0.899 0.851 0.848 0.841 0.823 0.854 0.852

1.36 1.25 1.46 1.69 2.04 2.35 2.51 2.53 2.62 3.37 3.26 3.36 3.18 3.37 3.39 3.36 3.81 3.02 3.16

1.074 1.104 1.119 1.143 1.230 1.283 1.387 1.423 1.494 1.508 1.507 1.558 1.645 1.663 1.620 1.644 1.632 1.753 1.904 2.098

0.795 0.820 0.801 0.789 0.711 0.742 0.706 0.730 0.706 0.707 0.752 0.748 0.745 0.727 0.754 0.756 0.758 0.767 0.777 0.805

2.65 2.64 2.72 2.80 3.34 3.27 3.69 3.61 3.89 3.88 3.63 3.74 3.86 3.97 3.70 3.73 3.68 3.90 4.18 4.42

1.113 1.167 1.199 1.241 1.324 1.565 1.768 1.755 1.746 1.871 1.849 1.900 1.909

0.453 0.505 0.584 0.623 0.581 0.564 0.606 0.637 0.624 0.653 0.657 0.656 0.661

4.79 4.50 3.99 3.85 4.40 5.14 5.22 4.84 4.90 5.01 4.88 4.98 4.94



Article

Table 3. continued S. No. 15 16 17 18 19 20

T (K) 374.0 374.5 374.8 375.6 376.1 376.6

x1′ 0.176 0.154 0.143 0.101 0.068 0.000

x1

x2

y1

Glycolic Acid/ChCl 3:1 = 15 mol % 0.150 0.700 0.526 0.132 0.718 0.471 0.122 0.728 0.433 0.086 0.764 0.310 0.057 0.794 0.231 0.000

γexp 1

γexp 2

α12

1.990 1.997 1.976 1.958 2.155

0.657 0.702 0.734 0.827 0.873

5.17 4.84 4.57 3.99 4.15

a Standard uncertainty u(x1) = u(x1′) = u(y) = 0.001, u(T) = 0.1 K, u(mole % of GC3:1) = 0.1, u(α) = 0.025. bThe presence of GC3:1 was not observed in the vapor phase.

Figure 5. (a) Temperature−composition diagram for the acetonitrile (1) + water (2) + GC3:1 (3) system at 101.3 kPa with GC3:1 = 5 mol %. x′1 vs T (●) and y1 vs T (○), --- GC = 0%. (b) Temperature−composition diagram for the acetonitrile (1) + water (2) + GC3:1 (3) system at 101.3 kPa with GC3:1 = 10 mol %. x′1 vs T (▲) and y1 vs T (Δ), --- GC3:1 = 0%. (c) Temperature−composition diagram for the acetonitrile (1) + water (2) + GC3:1 (3) system at 101.3 kPa with GC3:1 = 15 mol %. x1′ vs T (■)and y1 vs T (□) , --- GC3:1 = 0%.

also studied and presented in Table 4. The plot of relative volatility versus GC3:1 concentration (Figure 7) shows that, at 15 mol % GC3:1 concentration, the relative volatility α12 was 4.51. This value is 4.5 times higher than the GC3:1-free system.

concentration range studied. The azeotrope of the ACN + water binary mixture disappears at the lowest GC3:1 content (5 mol %). The variation in the relative volatility α12 of the ACN + water mixture (close-to-azeotrope composition, x1′ = 0.674) by varying the GC3:1 concentration in the range 0−15 mol % was F



Article

By using the γ−φ approach, the equilibrium, eq 2, is obtained yi φ̂ V p = xi γi pis φisPEi

(2)

where fV̂ is the fugacity of component i in the vapor phase, fL̂ is the fugacity of component i in the liquid phase, y1 is the mole fraction of component i in the vapor phase, φVi is the fugacity coefficient of component i in the vapor phase, P is the total system pressure, xi is the mole fraction of component i in the liquid phase, yi is the activity coefficient of component i in the liquid phase, psi is the saturation vapor pressure of component i at system temperature T, and φsi is the fugacity coefficient of saturated pure i at the system temperature T. The Antoine constant parameters36 are listed in Table 5. Table 5. Antoine Constants for Pure Substances36 a Antoine constants

Figure 6. Experimental and calculated VLE data for the ACN (1) + water (2) + GC3:1 (3) pseudoternary system at 101.32 kPa. For GC = 5 mol % (●), for GC = 10 mol % (▲), for GC = 10 mol % (■), and solid lines, calculations based on the NRTL model.

relative volatility (α12)

0.0 2.0 3.3 5.0 6.8 10.4 15.0

1.06 1.20 1.25 1.32 1.78 2.69 4.51

Ai

Bi

Ci

acetonitrile water

7.5305 8.07131

1609.86 1730.63

264.7 233.426

log pisat = Ai −

a

Table 4. Effectiveness of GC3:1 Concentration on the Relative Volatility α12 (x′1 ≈ 0.674) of the ACN (1) + Water (2) System, Relative Volatilities α12 at Atmospheric Pressure (101.32 kPa): GC3:1 = 0−15.0 mol %a mol % of GC3:1

component

Bi , T − 273 + Ci

where psat i is in mm Hg and T is in K.

The virial equation of state reduced at the second coefficient with Tsonopoulo’s correlations was applied to calculate the fugacity coefficients.37,38 Further, for Poynting effect (PE) estimation, DIPPR equations were used to calculate saturated liquid molar volumes and vapor pressures.39 ÄÅ L ÉÑ ÅÅÅ vi (p − pis ) ÑÑÑ ÑÑ PEi = expÅÅÅÅ ÑÑ ÅÅ ÑÑ RT (3) ÅÇ ÑÖ Therefore, the activity coefficients can be given as eq 4

a

Standard uncertainty u(mole % of GC3:1) = 0.1, u(α) = 0.025.

γi =

yi φ̂ V p xi pis φisPEi

(4)

The Poynting correction is close to unity when the difference pi − psat i is small, and this equation can thus be simplified to yi φî V p = xi γi pis φis

Since GC3:1 is a nonvolatile component and found to be absent in the vapor phase, these systems can be simplified as pseudobinary and pseudoternary considering the LTTM as a single component.28,29 However, while calculating the activity coefficient of the ACN + water mixture in the liquid phase, the GC3:1 concentration was considered. Generally, to calculate the activity coefficient, γcal, of systems comprising LTTMs, the NRTL model is employed.28,29,40 Experimental activity coefficients for pseudobinary systems of ACN + GC3:1 and water + GC3:1 were calculated using eq 5 and were fitted to the NRTL model. For the pseudoternary system ACN + water + GC3:1, correlation was done on the solvent-free basis.38,40 Regression analysis was performed using optimization techniques in the MatLab program for the estimation of parameters. A relative least-squares error objective function (OF) was used to fit the experimental data41 ÄÅ exp ÉÑ2 n Å ÅÅ γi − γicalc ÑÑÑ Å ÑÑ OF = ∑ ÅÅ exp ÅÅ ÑÑÑ γ ÑÑÖ i (6) i=1 Å ÅÇ

Figure 7. Effect of GC3:1 concentration on the relative volatility α12 (x1 ≈ 0.674) of the ACN (1) + water (2) system at 101.32 kPa.

3.3. Correlation of VLE Data. Fugacity equality (eq 1) is applied to calculate the VLE in both vapor and liquid phases. f̂

V

L = f̂

(5)

(1) G



Article

Table 6. NRTL Parameters and Their Correlation Statistics at 101.32 kPaa system

Δg12

Δg21

α

DT (K)

dt (K)

Dy1

ACN (1) + GC3:1 (2) water (1) + GC3:1 (2) ACN (1) + water (2) + GC3:1 (3)

7382.1 −930.9 3587

−3247.6 835.4 −542.3

0.3 0.3 0.3

0.25 0.24 0.42

0.46 0.79 0.98

0.0093

n (1/n)∑k = 1 |y1exp

y1cal |k ;

max(|yexp 1

Δg12 = (g12 − g21); DT =

a

n (1/n)∑k = 1 |T exp

cal

− T |k ; dT = max(|T

exp

− T |); Dy1 = cal

−

dy1 =

dy1

0.016

−

ycal 1 ).

Figure 8. Typical FTIR spectrum of GC3:1 before (a) and after (b) use.

where i represents the component and n the number of experimental data. The superscripts exp and cal represent calculated and experimental values, respectively. The OF was minimized in order to estimate the pseudobinary and pseudoternary parameters for components in the ACN + water + GC3:1 system. The relative volatility for ACN (1) + water (2) can be determined using eq 7 α12 =

Pseudoternary VLE data for the ACN + water + GC3:1 system were also fitted to the NRTL model. Due to LTTMs possessing a high viscosity, the NRTL model parameters calculated from pseudobinary systems cannot be applied for accurate prediction of the VLE behavior of the pseudoternary (ACN + water + GC3:1) system.28,29 Therefore, the NRTL model was used directly for correlation of ternary system VLE data, and the estimated results of the average difference in equilibrium temperature and ACN mole fraction in the vapor phase are presented in Table 6 and shown in Figure 6. The estimated values of the average difference in equilibrium temperature and ACN mole fraction in the vapor phase were found to be 0.009 and 0.42 K, respectively. This indicates that the NRTL model can be used for prediction of VLE data for a LTTM (GC3:1) containing system. 3.4. Recoverability Test. In order to evaluate the recoverability of the used GC3:1 as an entrainer, an aqueous solution of about 5% (mol/mol) was prepared for GC3:1, and in a subsequent process, it underwent boiling for 2 h in a rota evaporator. After each run, the entrainer was recovered by evaporating the water in a vessel and was reused for the subsequent run. Recovered GC3:1 was vacuum-dried, and the FT-IR spectrum was then obtained. The FT-IR spectrum of recovered GC3:1 is given in

y1 /x1′ y2 /x 2′

(7)

where x′1 and x′2 represent the mole fractions of ACN and water on a LTTM-free basis in the liquid phase, respectively, and y1 and y2 represent the mole fractions of ACN and water in the vapor phase, respectively. For pseudobinary systems (ACN + GC3:1, water + GC3:1), the correlated results for the average deviation in equilibrium temperature and ACN vapor mole fractions are presented in Tables 2 and 6, respectively, and the same are plotted in Figure 4. The results represent a good correlation between experimental VLE data and predicted values by the NRTL model. H



Article

VL

Figure 8b. While comparing the spectra of freshly prepared and recovered GC3:1, no significant change is observed. Hence, recovered GC3:1 can be reused for distillation, as its chemical properties are stable.

Acronyms

LTTMs VLE FTIR IPA

4. CONCLUSIONS GC3:1 has been identified as a potential entrainer for the separation of acetonitrile + water mixtures. Isobaric VLE data for the pseudobinary mixtures of ACN + GC3:1 and water + GC3:1 were measured at atmospheric pressure. The VLE data for the pseudoternary system ACN + water + GC3:1 were also measured at different LTTM mole fractions of 0.05, 0.1, and 0.15. The relative volatility of ACN + water mixtures was found to increase with increasing GC3:1 concentration. The experimental results of the pseudoternary system ACN + water + GC3:1 showed that the azeotropic behavior can be eliminated by using a GC3:1 entrainer. The experimental VLE data for the pseudobinary and pseudoternary systems were fitted/validated to the NRTL model. The estimated values of the average difference in equilibrium temperature and ACN mole fraction in the vapor phase were found to be 0.42 K and 0.009, respectively. This demonstrates that the NRTL model is capable of predicting the VLE data for a LTTM (GC3:1) containing system.

■

Low transition temperature mixtures Vapor−liquid equilibrium Fourier transform infrared spectroscopy Isopropyl alcohol

Greek letters

α α12 γi ϕVi ϕSi

■

Nonrandomness parameter in the NRTL equation Relative volatility of components 1 and 2 Activity coefficient of component i Fugacity coefficient of the ith component in the vapor phase Saturated fugacity coefficient of the ith component

REFERENCES

(1) Acosta, J.; Arce, A.; Rodil, E.; Soto, A. A thermodynamic study on binary and ternary mixtures of acetonitrile, water and butyl acetate. Fluid Phase Equilib. 2002, 203, 83−98. (2) Rosen, J.; Hellenas, K. E. Analysis of acrylamide in cooked foods by liquid chromatography tandem mass spectrometry. Analyst 2002, 127, 880−882. (3) Xu, K. X. Handbook of fine organic chemical raw materials and intermediates, 2nd ed.; Chemical Industry Press: Beijing, 2002. (4) Srinivasu, M. K.; Raju, A. N.; Reddy, G. O. Determination of lovastatin and simvastatin in pharmaceutical dosage forms by MEKC. J. Pharm. Biomed. Anal. 2002, 29, 715−721. (5) http://www.npi.gov.au/resource/acetonitrile (accessed on March 3, 2018). (6) Acosta-Esquijarosa, J.; Rodrıguez-Donis, I.; Haza, U. J.; NuevasPaz, L.; Pardillo-Fontdevila, E. Recovery of acetonitrile from aqueous waste by a combined process: solvent extraction and batch distillation. Sep. Purif. Technol. 2006, 52, 95−101. (7) Zhang, Z.; Ming, L.; Donghao, H.; Peng, J.; Dezhang, S.; Wenxiu, L. Isobaric Vapor−Liquid Equilibrium for the Extractive Distillation of Acetonitrile + Water Mixtures Using Dimethyl Sulfoxide at 101.3 kPa. J. Chem. Eng. Data 2013, 58, 3364−3369. (8) Fang, J.; Liu, J.; Li, C.; Yanwen, L. Isobaric Vapor−Liquid Equilibrium for the Acetonitrile + Water System Containing Different Ionic Liquids at Atmospheric Pressure. J. Chem. Eng. Data 2013, 58, 1483−1489. (9) Fang, J.; Zhao, R.; Wang, H.; Li, C.; Liu, J. Salting-out effect of ionic liquids on isobaric vapor-liquid equilibrium of acetonitrile-water system. Chin. J. Chem. Eng. 2015, 23, 1369−1373. (10) Sazonova, A. Y.; Raeva, V. M. Recovery of Acetonitrile from Aqueous Solutions by Extractive Distillation−Effect of Entrainer. Int. J. Chem. Mol. Eng. 2015, 9, 288−291. (11) Ruiz, R. A.; Borda, B. N.; Alexander, L. R.; Javier, R.; Guevara, L.; Ivan, D.; Gil, C. Control of an azeotropic distillation process to acetonitrile production, 21st European Symposium on Computer Aided Process Engineering − ESCAPE 21. (12) Repke, J. U.; Forner, F.; Klein, F. A. Separation of homogeneous azeotropic mixtures by pressure-swing distillation-analysis of the operation performance. Chem. Eng. Technol. 2005, 28, 1151−1157. (13) Venkata Siva Rama Rao, C.; Venkateswara Rao, K.; Ravlprasad, A.; Chiranjiv, C. Extraction of Acetonitrile from Aqueous Solutions 1. Ternary Liquid Equilibria. J. Chem. Eng. Data 1978, 23, 23−25. (14) Rao, D. S.; Rao, K. V.; Prasad, A. R.; Chiranjivi, C. Extraction of Acetonitrile from Aqueous Mixtures. 2. Ternary Liquid Equilibria. J. Chem. Eng. Data 1979, 24, 241−244. (15) Mandal, M. K.; Sant, S. B.; Bhattacharya, P. K. Dehydration of aqueous acetonitrile solution by pervaporation using PVA−iron oxide nanocomposite membrane. Colloids Surf., A 2011, 373, 11−21. (16) Khayet, M.; Cojocaru, C.; Zakrzewska-Trznadel, G. Studies on pervaporation separation of acetone, acetonitrile and ethanol from aqueous solutions. Sep. Purif. Technol. 2008, 63, 303−310. (17) Masuda, T.; Takatsuka, M.; Tang, B. Z.; Higashimura, T. Pervaporation of organic liquid-water mixtures through substituted polyacetylene membranes. J. Membr. Sci. 1990, 49, 69−83.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jced.8b00228. Table S-1, experimental isobaric VLE data for the IPA (1) + water (2) system at atmospheric pressure (101.32 kPa); Figure S1, TGA for GC3:1 from room temperature to 500 °C at atmospheric pressure (PDF)

■

Liquid molar volume

AUTHOR INFORMATION

Corresponding Author

*Phone: +91-9876019399. Fax: +91-175-2393005. E-mail: [email protected]. ORCID

Neetu Singh: 0000-0002-1970-7095 Jai Prakash Kushwaha: 0000-0003-4077-8805 Funding

Financial support from Thapar Institute of Engineering & Technology, Patiala, Grant No. TU/SEED/2014/CHE/NS, is gratefully acknowledged. Notes

The authors declare no competing financial interest.

■

LIST OF SYMBOLS Ai, Bi, Ci Constants of the Antoine equation for the ith component OF Objective function g NRTL model interaction parameters, J/mol P Total pressure, kPa psat Pure liquid saturation vapor pressure of the ith i component, kPa xi Liquid phase ith component mole fraction Liquid phase ith component mole fraction on LTTMxi′ free basis yi Vapor phase ith component mole fraction Tb Pure substance boiling points at 101.32 kPa, K T boiling temperature at equilibrium, K I



Article

(18) Soukup, J.; Jandera, P. Adsorption of water from aqueous acetonitrile on silica-based stationary phases in aqueous normal-phase liquid chromatography. J. Chromatogr. A 2014, 1374, 102−111. (19) Melnikov, S. M.; Höltzel, A.; Seidel-Morgenstern, A.; Tallarek, U. Adsorption of Water−Acetonitrile Mixtures to Model Silica Surfaces. J. Phys. Chem. C 2013, 117, 6620−6631. (20) Seiler, M.; Jork, C.; Karvanou, A.; Arlt, W.; Hirsch, R. Separation of azeotropic mixtures using hyperbranched polymers or ionic liquids. AIChE J. 2004, 50, 2439−2454. (21) Cui, X. B.; Li, Y.; Feng, T. Y.; Sun, G. X.; Lin, L. Separation of acetonitrile and water by saline extractive distillation. Petrochem. Technol. 2007, 36, 1229−1233. (22) Zhou, J. B.; Cui, X. B.; Dong, B. L.; Wang, Y. F.; Chen, Z. K. Separation of acetonitrile and water mixture by batch extractive distillation. Chem. Ind. Eng. 2009, 26, 482−486. (23) Acosta-Esquijarosa, J.; Rodríguez, I.; Jáuregui, U.; Nuevas, L.; Pardillo, E. Recovery of acetonitrile from aqueous waste by a combined process: Solvent extraction and batch distillation. Sep. Purif. Technol. 2006, 52, 95−101. (24) Fang, J.; Liu, J.; Li, C.; Liu, Y. Isobaric Vapor−Liquid Equilibrium for the Acetonitrile + Water System Containing Different Ionic Liquids at Atmospheric Pressure. J. Chem. Eng. Data 2013, 58, 1483−1489. (25) Fang, J.; Zhao, R.; Wang, H.; Li, C.; Liu, J. Salting-out effect of ionic liquids on isobaric vapor−liquid equilibrium of acetonitrile-water system. Chin. J. Chem. Eng. 2015, 23, 1369−1373. (26) Abbott, A. P.; Capper, G.; Davies, D. L.; Rasheed, R. K.; Tambyrajah, V. Novel solvent properties of choline chloride/urea mixtures. Chem. Commun. 2003, 1, 70−71. (27) Francisco, M.; Van den Bruinhorst, A.; Kroon, M. C. New natural and renewable low transition temperature mixtures (LTTMs): screening as solvents for lignocellulosic biomass processing. Green Chem. 2012, 14, 2153−2157. (28) Rodríguez, N. R.; Gonzalez, A. S. B.; Tijssen, P. M. A.; Kroon, M. C. Low transition temperature mixtures (LTTMs) as novel entrainers in extractive distillation. Fluid Phase Equilib. 2015, 385, 72−78. (29) Rodriguez, N. R.; Kroon, M. C. Isopropanol dehydration via extractive distillation using low transition temperature mixtures as entrainers. J. Chem. Thermodyn. 2015, 85, 216−221. (30) Renon, H.; Prausnitz, J. M. Local compositions in thermodynamic excess functions for liquid mixtures. AIChE J. 1968, 14, 135− 144. (31) Francisco, M.; Bruinhorst, A. V. D.; Zubeir, L. F.; Peters, C. J.; Kroon, M. C. A new low transition temperature mixture (LTTM) formed by choline chloride+ lactic acid: characterization as solvent for CO2 capture. Fluid Phase Equilib. 2013, 340, 77−84. (32) Perkins, L. S.; Painter, P.; Colina, C. M. Experimental and computational studies of choline chloride-based deep eutectic solvents. J. Chem. Eng. Data 2014, 59, 3652−3662. (33) Kaur, P.; Rajani, N.; Kumawat, P.; Singh, N.; Kushwaha, J. P. Performance and mechanism of dye extraction from aqueous solution using synthesized deep eutectic solvents. Colloids Surf., A 2018, 539, 85−91. (34) Othmer, D. F.; Gilmont, R.; Conti, J. J. An Adiabatic Equilibrium Still. Ind. Eng. Chem. 1960, 52, 625−628. (35) Brunjes, A. S.; Bogart, M. J. P. Vapor−Liquid Equilibria for commercially important systems of organic solvents. Ind. Eng. Chem. 1943, 35, 255−260. (36) Perry, R. H. Perrys chemical engineering handbook, 7th ed.; McGraw-Hill: New York, 1934. (37) Tsonopoulos, C. An empirical correlation of second virial coefficient. AIChE J. 1974, 20, 263−272. (38) Gjineci, N.; Boli, E.; Tzani, A.; Detsi, A.; Voutsas, E. Separation of the ethanol/water azeotropic mixture using ionic liquids and deep eutectic solvents. Fluid Phase Equilib. 2016, 424, 1−7. (39) Daubert, T. E.; Danner, R. P. Physical and Thermodynamic Properties of Pure Chemicals. Data Compilation, AIChE, New York, 1992.

(40) Peng, Y.; Lu, X.; Liu, B.; Zhu, J. Separation of azeotropic mixtures (ethanol and water) enhanced by deep eutectic solvents. Fluid Phase Equilib. 2017, 448, 128−134. (41) Kim, I.; Svendsen, H. F.; Børresen, E. Ebulliometric determination of vapor-liquid equilibria for pure water, monoethanolamine, N-Methyldiethanolamine, 3-(methylamino)-propylamine, and their binary and ternary solutions. J. Chem. Eng. Data 2008, 53, 2521− 2531.

J


Acetonitrile Dehydration via Extractive Distillation ... - ACS Publications

Recommend Documents