LiquidâLiquid Extraction of Lower Alcohols Using ... - ACS Publications

Article pubs.acs.org/IECR

Cite This: Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Liquid−Liquid Extraction of Lower Alcohols Using Menthol-Based Hydrophobic Deep Eutectic Solvent: Experiments and COSMO-SAC Predictions Rupesh Verma and Tamal Banerjee* Department of Chemical Engineering, Indian Institute of Technology Guwahati, Guwahati, Assam 781039, India S Supporting Information *

ABSTRACT: The lower alcohols tend to form azeotropic solution with water which makes the separation challenging. The current work primarily focuses on the synthesis and application of a menthol-based hydrophobic deep eutectic solvent (DES) for the removal of lower alcohols from its aqueous solutions. The DES is synthesized by the addition of DL-menthol and lauric acid (dodecanoic acid) with a molar ratio of 2:1. Liquid−liquid equilibria (LLE) experiments are then performed to evaluate the performance of the synthesized DES for the extraction of lower alcohols such as ethanol, 1propanol, and 1-butanol. LLE corresponding to the pseudo ternary systems of lower alcohols (1-butanol, ethanol, and 1propanol) + hydrophobic DES + water are measured at T = 303.15 K and p = 1 atm. The composition of the tie lines were evaluated using 1H NMR analysis for both extract and raffinate phases. Thereafter, the extraction efficiency of the DES is analyzed and compared by determining the solute distribution coefficients and the selectivity values. Finally, the experimental LLE data for the systems were regressed using the excess Gibbs free energy model, namely the Non Random Two Liquid (NRTL). Further the predictions of the tie lines were also confirmed through the COnductor like Screening MOdel Segment Activity Coefficients (COSMO-SAC) model. The average root-meansquare deviations (RMSD) obtained were 0.01 and 0.07 for NRTL and COSMO-SAC model, respectively.

■

INTRODUCTION The demand for energy is increasing proportionally with the current population. In such a scenario, energy generation is the key to sustain such a fast pace development. Currently fossil fuel replenishes nearly 80% of the energy demand globally. Hence, there is a dire need to explore alternate energy sources in order to lessen the dependency on the nonrenewable fossil fuel which are limited.1−4 Lower alcohols are considered a potential replacement for conventional fuels. Lower alcohols such as ethanol, 1-propanol, and 1-butanol are vital and are potential renewable energy sources. 1-Butanol has higher calorific value, higher hydrophobicity, and lesser flammability than other alcohol fuels.5 Qureshi et al.4,6 described that 1butanol with a lower vapor pressure and higher flash point is less corrosive. Lower alcohols have also shown properties similar to gasoline. Lower alcohols hence can be used as a renewable biofuel with little or no modification to the engine. One of the sources of lower alcohols is the acetone-butanolethanol (ABE) fermentation where alcohols exist as an azeotrope having a water-rich phase. Hence its extraction from aqueous phase is essential. In the chemical process, industrial separation of azeotropic mixtures is of great importance. Methods such as extraction, adsorption, pervaporation, gas stripping, and membrane separation have been used conventionally for separation of © XXXX American Chemical Society

lower alcohols from the fermentation broth. Membrane separation and pervaporation are expensive due to low mass transfer rates and requirement of low pressure.1,7−10 Typically, removal of lower alcohols from fermentation broth by adsorption from the liquid phase can only be used in laboratory scale due to the small-capacity of adsorbents. The other option for removal of lower alcohols can be used, such as membrane reactors where the immobilization of microorganisms occurs in the membrane. On industrial scale, cell immobilized technique gives more disadvantages due to poor mechanical strength and an increase in mass transfer resistance.2 One normally switches to liquid−liquid extraction when component separation (from a mixture of many components) cannot be achieved economically by other mass transfer operations such as distillation, evaporation, and crystallization. Azeotropic distillation, extractive distillation, and liquid−liquid extraction, which are three of the most important industrial separation techniques for azeotrope breaking, involve the use of an extracting agent. In such a scenario, solvent extraction is a suitable operation particularly when the solvents possess Received: Revised: Accepted: Published: A

December 21, 2017 February 6, 2018 February 12, 2018 February 12, 2018 DOI: 10.1021/acs.iecr.7b05270 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research Table 1. Compound Name, Solubility, Purities, and Source of the Chemicals Used in the Work

a

Sl. no.

compound name

solubility in water (g/lit)

B.P. (°C)

densitya (g/cm3)

purity

sources

1 2 3 4 5 6

DL-menthol dodecanoic acid ethanol 1-propanol 1-butanol DES

0.42 0.059 infinite infinite 75 NMa

214.6 298.9 78.24 98 117.7 NMb

0.890 1.007 0.789 0.803 0.810 0.894

≥95% ≥99% ≥99.9% ≥99% ≥99% NMb

Sigma-Aldrich, Germany Merck, Germany Merck, Germany Merck, India Merck, India this work

298.15 K. bNo presence of DES was observed from 1H NMR spectra (Figure S3).

water. In the concluding section, we validate the experimental tie lines with COSMO-SAC predictions.23−28

simultaneously high affinity for alcohol and low solubility with water. An important aspect with these hydrophobic solvents is its comparable density difference when compared to water, making separation easier. Hydrobhopic ionic liquids (IL’s) have been known to perform better for the extraction of lower alcohols. In the literature, two approaches to separate alcohol− water azeotrope mixture namely (a) hydrophilic ILs and (b) hydrophobic ILs are available. Hydrophilic ILs work in instances where there is a concentrated feed mixture of 1butanol and water.10−14 However, in many situations, separation from dilute feed stream becomes challenging. It is in these context that the synthesis of low cost hydrobhobic solvent is desired. However, due to its high cost and nature of toxicity, cheaper and sustainable solvents are now being explored. This has also prompted researchers to use solvents such as DES for the extraction of lower alcohols and generate the thermodynamic phase equilibria data. Keeping these advantages in mind, the extraction of lower alcohols by alternative hydrophobic solvents or DES has been proposed in the present study. Deep eutectic solvents (DESs) were introduced as analogues and alternative green solvents to the conventional ILs, with the advantage of easy preparation with high purities and low cost. By definition, DESs result from the establishment of specific interactions, mainly hydrogen bonds, between two compounds namely a hydrogen bond donor (HBD) and a hydrogen bond acceptor (HBA). Originally, low temperature transition mixtures (LTTM) were called deep eutectic solvents (DESs), but this name does not cover the complete class of solvents because many mixtures do not show an (eutectic) melting point but a glass transition instead. It is also important to mention that the scientific community19,20 does not yet agree on how large the melting point decrease of the mixture should be in order to be called a DES. Most researchers do agree that DESs are not only formed at the eutectic composition but a mixture with a melting point lower than a certain threshold (e.g., room temperature). For example, mixtures of ammonium salts with ethylene glycol and glycerol have also been called DESs by other researchers. Overall DES are new chemical entities with melting points lower than that of the initial compounds. Most of the DESs proposed so far in the open literature have a hydrophilic character and thus are unstable in water, leading to the separation of both components. In such a scenario, the choice of HBA and HBD is crucial.15−18 With respect to the hydrophobicity of the DES from previous literature, in the current work, DL-menthol and lauric acid were chosen as HBA and HBD, respectively.19,20 It should be noted that their toxicity are yet to be too documented; hence at this point, we shall merely refer to them as a potential substitute for IL’s.17,19,21,22 The present study discusses the synthesis of DES and then uses it to measure the LLE with lower alcohols and

■

EXPERIMENTAL DETAILS Table 1 shows the compound name, solubility, purity, and source of the chemicals used in the present study. Purities of ethanol, 1-propanol, 1-butanol, lauric acid, and DL-menthol were confirmed by 1H NMR spectroscopy. The analysis of the peaks indicated negligible impurities. The NMR solvent, Dimethyl sulfoxide-d6 (DMSO-d6 ≥ 99.8%) supplied by Merck, Germany was used as received. All chemicals were of analytical grade and were used without further purification. Hydrophobic DES Preparation. As per previous authors, the hydrogen bond acceptor (HBA:DL-menthol) is combined with hydrogen bond donor (HBD:lauric acid) with a molar ratio of 2:1.19,20 They were added in a flat-bottom flask which was fitted with a reflux condenser for 1 h at 50 °C with magnetic stirring until a clear liquid was formed. For reducing the water content and volatile compounds of DES to negligible values, a vacuum at T = 60 °C for at least 48 h was applied to the DES samples prior to the LLE measurements. Thereafter the synthesized DES was evaluated for its purity using 1H NMR spectra. Figure S1 reports the 1H NMR spectra of the synthesized DES (DL-menthol:lauric acid) along with the peak assignment. It should be noted that the density and viscosity are also essential physical properties for efficient piping design. Keeping this in mind, the density and viscosity were measured in the temperature range of 20−85 °C (Table 2). The density of the all the solvents and DES was measured by a DMA 4500 M densitometer (Anton Paar Make) with a relative expanded uncertainty of 0.003. The viscosity of DES was measured by an interfacial rheometer (model: Physica Table 2. Experimental Density and Viscosity Data of Pure DES at Atmospheric Pressure (p = 1 atm) and Different Temperaturesa density (g cm−3)b

viscosity (mPa)b

temperature (K)

present work

lit19,20

present work

lit19,20

293.15 303.15 313.15 323.15 333.15 343.15 353.15 AAD

0.8971 0.8898 0.8826 0.8753 0.8678 0.8603 0.8526 0.003

0.9002 0.8930 0.8857 0.8780 0.8703 0.8631 0.8549

21.810 12.500 7.657 5.112 3.623 2.670 2.088 2.84

29.689 16.957 10.527 7.057 4.872 3.599 2.650

a

The standard uncertainty u are u(T) = 0.1 K, u(p) = 1 kPa. bThe relative expanded uncertainty U are Ur(ρ) = 0.003, and Ur(η) = 0.033; n

AAD = ∑i = n B

|wical − wiexp| n DOI: 10.1021/acs.iecr.7b05270 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

with the results of both Florindo et al.19 and Rebeiro et al.,20 where the corresponding value of 27 and 34 °C were obtained. Thus, overall the melting point of the DES is now lower than either HBA (menthol) or HBD (lauric acid) and hence qualifies the definition for the scientific community. In order to confirm the effect of the ratio of HBD to HBA, a preliminary attempt has been devised to predict the same through the COSMO-SAC model. The mutual solubility of two DES, namely (a) choline chloride and 1-ethyl-3-methylimidazolium chloride and (b) lauric acid and DL-menthol are predicted by the COSMO-SAC model.29−32 The pure component parameters (melting point and heat of fusion) are obtained as per the literature given in Table 3. While the first

MCR301, Anton-Paar Make) with a relative expanded uncertainty of 0.033. In order to confirm the eutectic nature of the mixture, the melting point of the synthesized DES was performed by recording the DSC measurements (model: TGA/DSC1 Star System, make: Mettler Toledo, Switzerland) (Figure 1). The

Table 3. Melting Properties for COSMO-SAC Calculations name of the cmpd

Tm (K)

ΔHf (kJ/mol)

reference

[Ch]Cl [Emim]Cl menthol lauric acid

597.15 350.42 308.8 317.48

4300 8588 11000 37830

Fernandez et al.33 Fernandez et al.33 Corvis et al.46 Fernandez et al.33

one is benchmarked and validated with the experimental data obtained by Fernandez et al.,33 the latter have been predicted and validated with the current work (Table 4, Figure 2). This is essentially a solid liquid equilibrium problem where the simplified form is given by eq 1 as below: ln(xsLγsL) = −

Ts,fus ⎞ ΔfusHs ⎛ ⎟ ⎜1 − RTs,fus ⎝ T ⎠

(1)

Table 4. Coordinates of the Eutectic Points (Menthol:Lauric Acid) Estimated by Experimental Study and Validated with COSMO-SAC Model experimental

a

COSMO-SAC

system

xlauric acid

T (K)

xlauric acid

T (K)

[Ch]Cl + [Emim]Cl menthol:lauric acid

0.430a 0.333

295.0 288.19

0.433 0.311

291.40 283.15

Fernandez et al.33

Where ΔfusHs and Ts,fus are the heat of fusion and melting temperature of the pure solute, respectively. T is the equilibrium temperature in K, R is the ideal gas constant, xLs is the mole fraction of lauric acid in liquid phase, and γLs is the activity coefficient of the solute (lauric acid) in the liquid phase (i.e., menthol). For the choline chloride based DES, the eutectic point at 0.43 closely matches with the predicted mole fraction of 0.433 as per COSMO-SAC predictions. In a similar manner, the eutectic temperatures are in close vicinity of each other, thereby indicating the robustness of the COSMO-SAC method. For our DES, it is evident that a liquid solution or eutectic point is predicted at x = 0.311, where x corresponds to the mole fraction of lauric acid. This essentially implies a molar ratio of 0.689/0.311 ∼2, which is the same as obtained in the experimental evaluation. Further as can also be seen from Figure 2, the eutectic point is obtained at T = 283.15 K by COSMO-SAC predictions. Therefore, it implies an agreement between the experimental and COSMO-SAC prediction. We have also attempted to synthesize molar ratios of 0.5:1, 1:1, 1.5:1, and 2.5:1 for the synthesis of DL-menthol:lauric acid as

Figure 1. Differential Scanning Calorimetry (DSC) for (a) DES and (b) menthol.

solid−liquid phase transition was measured from −30 to 50 °C with a heating and cooling rate of 5 °C/min. The condensation in the furnace was avoided with the use of dry nitrogen as purge gas. The flow rate of the purge gas was kept at 60 mL/min. Indium with a melting point of 156.6 °C was used as a standard for calibration. Samples of DES and DL-menthol ranging from 5 to 10 mg were then transferred to an aluminum DSC pan and hermetically sealed so as to prevent its vaporization. The uncertainty in the melting point temperature obtained by calculation of the standard deviation of three consecutive measurements for the same sample was found to be better than ±1 K. The melting point of the DES was obtained as 15.04 °C (Figure 1a). This is lower than the melting point of DLmenthol (Figure 1b). It is to be noted that DL-menthol presents two melting points, namely 28.31 and 35.3 °C. This is primarily due to its polymorphs α and β. This closely agrees C

DOI: 10.1021/acs.iecr.7b05270 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article


Figure 2. Eutectic composition of lauric acid for DES composed of (a) choline chloride and [C2mim]Cl33 and (b) lauric acid and DL-menthol as predicted by the COSMO-SAC model.

Equilibrium was confirmed when two successive sample concentrations gave the same molar concentration at the desired temperature. For the LLE experiments, desired components namely water, alcohol, and DES were added in a 30 mL stoppered bottle so that they formed a heterogeneous mixture. The sample bottles were kept inside the incubator shaker (Daihan Lab Tech, China), with the set temperature of 303.15 K at 200 rpm for 6 h. After shaking, the mixture was kept for 24 h so as to ensure equilibrium. Thereafter using 1H NMR spectroscopy, the compositional analysis were performed on two clear transparent phases, namely the DES-rich phase (upper layer) and the water-rich phase (lower layer). A sample of 0.1 mL were taken from both the phases and added with 0.5 mL DMSO− D6 in a NMR tube (thrift grade, Sigma-Aldrich). Para film has been used on the cap of the NMR tube to avoid any evaporation losses. A 600 MHz NMR spectrometer (Make: Bruker) was used for the 1H NMR of all samples so as to measure the peak areas of the distinct types of hydrogen nuclei. On the basis of the theory of NMR, the area under the curve is proportional to the number of hydrogen for the referred component. The NMR spectra for pure DES, the DES-rich phase, and the water-rich phase have been provided in Figures S1−S3. The reference peak for the NMR solvent namely DMSO−D6 has been recorded at 2.5 ppm. The peaks referenced in the composition analysis for DES in the extract phase was −CH− grouping from DL-menthol at 3.15 ppm (Figure S2a). In the extract phase, the spectra of water gets merged with the −CH2− attachment of 1-propanol, which is at ∼3.4 ppm (Figure S2a). In such a scenario, suitable equations were formed to evaluate the water and alcohol composition separately by using peak information from the −CH2−OH peak of alcohols at ∼1.4 ppm (Figure S2a). Figure S2a has been expanded in the region for which the contribution of water and alcohol needs to be elucidated. As per Figure S2a, the peaks of both 1-propanol and water merge at 3.35 ppm. This is true since pure 1-propanol 1H NMR depicts the −CH2−OH peak at 3.34 ppm (Figure S2b).This is the precise reason that the water peak (3.35 ppm in Figure S2a) gets overlapped with 1-propanol in the extract phase. Thus, the area corresponding to −CH2− of 1-propanol (Figure S2b) at 1.4 ppm is subtracted from the combined area of water and 1-propanol at 3.35 ppm (Table S1).

shown in the visual observation (Figure 3). It is clearly seen that the ratio’s below 1:1 shall have crystal formation. This

Figure 3. Formation of DES with different molar ratio of menthol to lauric acid.

diminishes as the mole ratio increases or we go toward the left of the x axis of Figure 2. Moreover this also agrees with the reported literature, where the composition (2:1) have been used for synthesizing the similar DES.19,20 Further, the same authors have also confirmed the aqueous stability of the DES.

■

MEASUREMENT OF LIQUID−LIQUID EQUILIBRIA The 1H NMR spectra is a widely used technique to obtain phase composition in multicomponent mixtures.30 The reliability of the NMR method was checked using known compositions of binary (alcohol + DES) and pseudo ternary mixtures which gave an uncertainty range of ±0.001. For experimental uncertainty, we have also triplicated a few experiments and have found that the average uncertainty of the mole fraction does not exceed 0.001. We have compared our estimated uncertainty with similar experimental work such as those of LLE measurements with ionic liquid34 and other DES35 and have found them to be reliable. For the equilibrium state in our study, samples were drawn at different time intervals with concentrations determined by 1H NMR. The experiments were performed with the help of Hamilton Syringes of 0.1 mL. These were taken from both the phases and added with 0.5 mL DMSO−D6 in a NMR tube (thrift grade, Sigma-Aldrich). Due to short time response, the syringes after crossing the upper phase still did not show any contamination. This has been further checked with known binary mixtures by subsequently recording their 1H NMR. D


Article


Table 5. Experimental LLE Data for the Pseudo Ternary System, DES (1) + Ethanol (2) + Water (3) at T = 303.15 K and p = 1 atma extract phase

a

raffinate phase

xDES [DL-menthol + lauric acid] (2:1)

xethanol

xwater

xDES

xethanol

xwater

βethanol

selectivity (S)

0.702 0.639 0.478 0.388 0.336 0.236

0.080 0.130 0.275 0.302 0.334 0.406

0.218 0.231 0.247 0.310 0.330 0.358

0.000 0.002 0.003 0.002 0.001 0.001

0.044 0.043 0.082 0.101 0.117 0.159

0.956 0.955 0.915 0.897 0.882 0.840

1.818 3.023 3.354 2.990 2.855 2.553

7.973 12.499 12.423 8.652 7.630 5.991

Standard uncertainties are u(T) = 0.01 K, u(x) = 0.001.

Table 6. Experimental LLE Data for the Pseudo Ternary System, DES (1) + 1-Propanol (2) + Water (3) at T = 303.15 K and p = 1 atma extract phase

a

raffinate phase


xpropanol

xwater

xDES

xpropanol

xwater

βpropanol

selectivity (S)

0.671 0.581 0.491 0.439 0.267 0.330

0.128 0.186 0.276 0.313 0.430 0.381

0.201 0.233 0.233 0.248 0.303 0.289

0.002 0.003 0.001 0.000 0.004 0.002

0.006 0.014 0.025 0.024 0.038 0.034

0.992 0.983 0.974 0.976 0.958 0.964

21.333 13.286 11.040 13.042 11.316 11.206

105.287 56.051 46.150 51.325 35.777 37.379

Standard uncertainties are u(T) = 0.01 K, u(x) = 0.001

Table 7. Experimental LLE Data for the Pseudo Ternary System, DES (1) + 1-Butanol (2) + Water (3) at T = 303.15 K and p = 1 atma extract phase

a

raffinate phase


xbutanol

xw

xDES

xbutanol

xw

β1‑butanol

selectivity (S)

0.651 0.517 0.434 0.330 0.260 0.184

0.150 0.269 0.310 0.381 0.480 0.508

0.199 0.214 0.256 0.289 0.260 0.308

0.022 0.001 0.000 0.003 0.001 0.009

0.007 0.012 0.006 0.012 0.014 0.039

0.971 0.987 0.994 0.985 0.985 0.952

21.43 22.41 51.66 31.75 34.28 13.02

104.56 103.39 200.61 108.21 129.89 40.26

Standard uncertainties are u(T) = 0.01 K, u(x) = 0.001

of DL-menthol (1.0) and lauric acid (0.5) are in the molar ratio of 2:1. This is exactly the same as obtained from the COSMOSAC predictions (Table 2 and Figure 3). Stability of DES in Water. For analyzing the hydrophobic nature of DES, two parts by volume water and one part of DES were added in a 15 mL bottle. Shaking was performed with a magnetic stirrer for 30 min. Thereafter, the mixture was kept for 12 h for equilibrium. The samples were drawn from both the water-rich (lower) and the DES-rich (upper) phases for 1H NMR. The NMR spectra clearly show an absence of water in the upper phase (solvent-rich phase) (Figure S4). Figure S5 shows the 1H NMR of the washed sample repeatedly after four times. The NMR spectra clearly reflects an absence of DES in the bottom phase (aqueous-rich phase) even after repeated washing. It implies that the DES synthesized are hydrophobic in nature.

In order to ensure correct water content in the extract phase, the compositions were confirmed through a Karl Fisher Titrator (MetroOhm 787 KF Titrino). The uncertainty of the method for water determination was 0.01 wt %. Considering triplicate runs performed at each point, it was determined that the uncertainty in the water content was always lower than 0.03 wt %. The water exhibited a dominant peak at 4 ppm (Figure S3) in the raffinate-rich phase. The concentration of other components in each phase was then calculated by eq 2 as below27

xi =

Hi n ∑i = 1 Hi

(2)

Here Hi denotes the peak area of single hydrogen of component “i” and xi the mole fraction of component “i”. As observed from 1H NMR spectra of synthesized pure DES (Figure S1), the molar ratio of −OH group in DL-menthol (peak number 15) is twice than that of the corresponding −OH group (peak number 14) of lauric acid. In another conformation, −CH− group of DES (peak number 9) has an area of unity, while peak 12 resembling 16 H atoms of lauric acid has an area of 8. So an effective contribution of a single hydrogen atom of lauric acid is 0.5 (i.e., 8/16). Hence the ratios

■

RESULTS AND DISCUSSION The LLE for the pseudo ternary systems, namely, DES (1)ethanol (2)-water (3), DES (1)-1-propanol (2)-water (3), and DES (1)-1-butanol (2)-water (3) were measured at T = 303.15 K and atmospheric pressure. The experimental tie line data are reported in Tables 5, 6, and 7 for ethanol, 1-propanol, and 1E


Article

Industrial & Engineering Chemistry Research butanol, respectively, in terms of mole fraction for the extract and raffinate phase. The compositions have been obtained after the 1H NMR analysis as discussed in the previous section. Tables 5−7 also report the distribution ratio, which is the ratio of the concentration of alcohols in the DES phase to the concentration in the aqueous phase. Another important descriptor, namely the selectivity, is simply the ratio of the distribution coefficient of alcohol and water. The distribution coefficient (β) and selectivity (S) are then calculated using eqs 3 and 4 as below:24 DES xalc β = water xalc

s=

DES xalc water xalc

Table 8. Comparison of Distribution Coefficients and Selectivities for Ethanol Extraction in Aqueous Media Using Ionic Liquids and DES system DES [DL-menthol/lauric acid with molar ratio (2:1)] DES [glycerol/choline chloride with molar ratios (4:1)] DES [glycerol/choline chloride with molar ratios (2:1)] DES [glycerol/tetramethylammonium chloride with molar ratios (4:1)] DES [glycerol/tetramethylammonium chloride with molar ratios (2:1)] [TDTHP][Phosph] [TDTHP][Deca] [TDTHP]Cl [TDTHP][CH3SO3] [TDTHP]Br [TDTHP][N(CN)2] [TDTHP][Tf2N] [BMIM][Tf2N]

(3)

x wDES x wwater

(4)

Here xalc and xw are the mole fractions of alcohol and water, respectively. The superscripts DES and water represents the extract (E) and raffinate (R) phase, respectively. Among the alcohols, high values of selectivity were reported for 1-butanol (Table 5). This indicates that the DES has a preferential ability to extract 1-butanol when compared to ethanol and 1-propanol. In general, the separation factor decreases as the concentration of alcohol in the feed increases. This is primarily due to the reduction of the two-phase region with increase in concentration of alcohol. This indicates the fact that the separation capacity of the solvent is reduced. A similar pattern is also seen for the distribution coefficient which refers to the amount of solvent required for the desired separation. As the DES is hydrophobic in nature, the mole fraction is negligible in the raffinate phase. The extraction efficiencies, EE, were calculated from the alcohol concentration in the water-rich phase prior and after extraction. This is given as below19 %EE =

R − molefr alc,1 R molefr alc,0

distribution coefficient

selectivity

0.505

12.5

present study

0.811

21.9

Rodriguez et al.17

0.618

15.2

Rodriguez et al.17

0.643

13.3

Rodriguez et al.17

0.725

14.7

Rodriguez et al.17

0.83 0.82 0.88 0.82 0.70 0.51 0.31 0.15

5.1 4.9 6.6 4.6 8.4 6.8 2.0 7.5

references

Neves et al.11 Neves et al.11 Neves et al.11 Neves et al.11 Neves et al.11 Neves et al.11 Neves et al.11 Cháfer et al.47

Table 9. Comparison of Distribution Coefficients and Selectivities for Propanol Extraction in Aqueous Media Using Ionic Liquids and DES system


selectivity

references

DES [BMP][Tf2N] [TDTHP][Phosph]

3.2 0.37 1.37

105.28 19.8 88.5

present study Cháfer et al.48 Bharti et al.30

It should also be noted that the chloride ion is responsible for the corrosion of reactor vessels when heated at high temperatures. This is due to the fact that at high temperature it can underdo hydrolysis thereby corroding the vessel surface. Thus, ionic liquids or DES consisting of halogen or halogencontaining anions such as [AlCl4], [PF6], [BF4], [CF3SO3], and [(CF3SO2)2N] thus limit their “greenness”. Overall the presence of halogen atoms is known to cause serious concerns if the hydrolysis stability of the anion is poor or if a thermal processing and recycle of spent ionic liquids is desired.37 In both cases, additional operations are required to avoid the liberation of toxic and corrosive HF or HCl into the environment. Thus, from an industrial point of view, halogencontinuing DES or ionic liquids are generally not recommended, even though they have higher selectivity as well as distribution coefficient. However, for the current DES, the selectivity and distribution values are higher with respect to ethanol (Table 8), 1-propanol (Table 9), and 1-butanol (Table 10), even when compared with a halogen-containing IL, namely [TDTHP]Br (Table 8). A comparison with ionic liquids also points to a similar fact for ethanol extraction as measured by Neves et al.11 with trihexyl tetradecyl based phosphonium cations. On the contrary both the selectivities and distribution coefficient are larger for DES when compared to [BMIM][Tf2N] for the case of 1-propanol (Table 9) or 1-butanol (Table 10). In 1-butanol both the selectivities and distribution coefficient are larger than imidazolium-,7 phosphonium-,7 or morpholium-based ionic liquids.38 Looking at the distribution coefficients, it implies that the current DES will be beneficial as

R molefr alc,0

(5)

molefrRalc,0

Here is the mole fraction prior extraction and molefrRalc,1 after equilibrium at 0.2 mole fraction of alcohol in the feed mixture. It can be observed that in general the extraction efficiencies of the lower alcohols with DES follow the order: 1butanol (∼90%) > 1-propanol (∼80%) > ethanol (∼50%). Hence the hydrophobic water stable DES may be recommended for use in the removal of lower alcohols from aqueous solutions. In order to further explore the role of the hydrophobic character of the extractant, a comparison has been conducted with both reported DES and the ILs in Tables 8, 9, and 10, for ethanol, 1-propanol, and 1-butanol, respectively. In Table 8, the distribution coefficient and selectivity is found to be less that of the other DES due to the presence of chloride ion in the hydrogen bond donor. So in the case of choline chloride (HBA)- or tetramethylammonium (HBA)-based DES, the hydrogen bonding with the incoming ethanol is much higher when compared to menthol (HBA)-based DES. This is due to the fact that the chloride ion possessing a higher charge density results in a higher fraction of active site for initiating hydrogen bonding. Further, it is also reported that the anionic part of IL or DES generally plays an important role in the thermal stability. For example, the IL-based organic anions have higher thermal stability than those based on inorganic anions.36 F


Article

Industrial & Engineering Chemistry Research Table 10. Comparison of Distribution Coefficients and Selectivities for 1-Butanol Extraction in Aqueous Media Using Ionic Liquids and DES system


selectivity

references

DES [Im10,1][TCB] [P6,6,6,14][TCB] [Im8,1][FAP] [Im10,1][ Tf2N] [Mo10,1][TCB] [Mo10,1][Tf2N]

8.9 3.2 2.0 0.8 5.7 4.8 2.1

200.6 100 500 420 90 70 99.7

present study Heitmann et al.7 Heitmann et al.7 Heitmann et al.7 Nann et al.38 Nann et al.38 Nann et al.38

it will require a lower solvent to feed ratio for the same performance. In order to propose a technique for recovery of alcohols, the thermal stability of the DES mixtures needs to be ascertained, so that distillation can be proposed to recover the alcohols. TGA of the DES20 shows that it is stable up to 231.49 °C, such that the alcohols with boiling points less than 117 °C can be easily recovered as an overhead product. With respect to the tie lines as evident from Figures 4, 5, and 6, it exhibits a type-I LLE behavior. Further from the NMR

Figure 5. Experimental and COSMO-SAC predicted tie lines for the pseudo ternary system: DES (1) + 1-propanol (2) + water (3) at 303.15 K and 1 atm.

Figure 6. Experimental and COSMO-SAC predicted tie lines for the pseudo ternary system: DES (1) + 1-butanol (2) + water (3) at 303.15 K and 1 atm. Figure 4. Experimental and COSMO-SAC predicted tie lines for the pseudo ternary system: DES (1) + ethanol (2) + water (3) at 303.15 K and 1 atm.

interaction parameters (Aij) are given in Table 11. A detailed methodology is given in our earlier work and is not discussed here.23,24,27,39 m

spectra, the raffinate phase of all the systems have negligible solvent and contains primarily water. It should be noted that due to the baseline correction of the NMR spectra, concentrations lower than 0.001 are very difficult to obtain. This usually signifies the uncertainty levels of the NMR spectra. In all the alcohols, the raffinate phase lies on the extreme corner of the binary axis of water-1-butanol and water-1-propanol. A negligible amount of alcohol is present in the raffinate phase after extraction. NRTL24 has been used to obtain interaction parameters which are regressed from experimental data, while the COSMO-SAC model24,25 has been used to determine the activity coefficient using a statistical mechanical framework. A suitable minimization process using GA was carried out with the objective function (eq 6) which was regressed from the experimental pseudo ternary compositions. The binary

II

c

maximize: F with respect to Aij = − ∑ ∑ ∑ wikl(xikl − xik̂ l )2 , ( where i , j = 1,2,3

)

wikl = 1

k=1 l=I i=1

and j ≠ i

(6)

The goodness of the fit was predicted by the root-meansquare deviation (RMSD) which is defined as40 1/2 m c II (xikl − xik̂ l )2 ⎤ ⎛ F ⎞1/2 ⎡ ⎥ ⎟ RMSD = ⎜ − = ⎢∑ ∑ ∑ ⎝ 2mc ⎠ ⎥⎦ 2mc ⎣⎢ k = 1 i = 1 l = I

(7)

where m refers to the number of tie lines and c refers to the number of components (viz. three for the present system). Here xlik and x̂lik are the experimental and predicted values of composition (mole fraction) for component i in the kth tie line for phase l, respectively. It should be noted that the parameters G


Article


basis set and density fitting basis set DGA1 are used. While calculating the Coulomb interaction, the density fitting basis set expands the density in a set of atom-centered functions. Thus for a medium-sized system, a performance gain in terms of accuracy in molecular structure and relative energies are achieved. The COSMO files once generated were then used to generate the sigma profile of the molecules, which were then used in the statistical framework namely segment activity coefficient (SAC) to obtain the activity coefficient. The activity coefficients were then used with the modified Rashford Rice algorithm to generate the tie lines. The COSMO-SAC parameters used were namely effective area (aeff) = 7.5 Å2, cutoff value for hydrogen-bonding interaction (σhb) = 0.0084 e/ Å2, and constant for hydrogen-bonding interaction (chb) = 85580 (kcal/mol)(Å4/e2). The details of the calculation procedure and the COSMO-SAC parameters can be found elsewhere.29 For the sigma profile calculation of DES, a molar ratio of 2:1 was used for DL-menthol (HBA) and lauric acid (HBD) molecule, respectively. The sigma profile of DL-menthol was taken as the average of dextro and levo isomers [pHBA(σ)].The charge distribution for such a DES will then be the algebraic sum of the sigma profiles calculated separately.45 It takes the form:

Table 11. NRTL Binary Interaction Parameters for the Systems at T = 303.45 K and p = 1 atm NRTL model parameters

a

Aji

Fa

RMSDb (in %)

i−j

Aij

1−2 1−3 2−3

5.76 −0.06 12.34

DES (1) + ethanol (2) + water (3) 13.53 −0.10 0.52 6.64 17.97

1−2 1−3 2−3

5.00 4.75 11.80

DES (1) + propanol (2) + water (3) 15.45 −0.15 0.65 8.79 19.21

1−2 1−3 2−3

3.27 −4.66 3.69

DES (1) + 1-butanol (2) + water (3) 19.68 −0.11 0.55 15.58 2.99

COSMO-SAC RMSD (in %) 8.75

7.71

5.83

Calculated by eq 4.

τij from eq 6 are merely regressed parameters for the NRTL model. It has no physical interpretation. It was observed from our earlier work on aqueous systems (Singh et al.41) and ionic liquids (Sahoo et al.42) that with a population size and generation greater than 100 and 200, respectively, the objective function remains asymptotically flat. In our earlier work, a local minimization was also carried out in the nearby optimized solution so as to confirm the global optimized parameters. Hence different runs of GA with population size 100 and number of generations 200 has been adopted in this current work. GA was required as the objective function is highly convex in nature with the result that traditional optimization algorithms fail.

pDES (σ ) = pHBA (σ ) + pHBD (σ ) = fHBA pHBA (σ ) + fHBD pHBD (σ )

(8)

pHBA (σ ) = pMTBP (σ ) + pBr (σ )

(9)

Here pHBA(σ) and pHBD(σ) are the sigma profile of the components of DES, namely the HBD and HBA. f HBA = 2 and f HBD = 1 are the mole ratios that have been adopted in the experimental work. The sigma profile of DES is then normalized so that it would appear as the sigma profile of a single molecule. The pseudo ternary tie lines are then predicted and compared with experimental tie lines for ethanol (Figure 4), 1-propanol (Figure 5), and 1-butanol (Figure 6). Using both the NRTL and COSMO SAC models, the average root mean square deviation (RMSD) values obtained were 0.01 and 0.07, respectively. The NRTL model is expected to agree with a higher accuracy when compared to the COSMO-SAC model, owing to its regression from the experimental data. The COSMO-SAC model is predictive in nature where the only input required is the optimized molecular structure and the temperature. However, in order to further elucidate the nature of interaction, a key indicator within the COSMO scheme,31,32 namely sigma potential, has been computed and discussed in the next section. Sigma Potential. The sigma potential is a useful indicator for evaluating the hydrogen bond acceptor and donor segments. The expression for the sigma potential of the segments based is given by

■

COSMO-SAC MODELING The pseudo ternary composition as discussed above contains molecules which do not have any repeatable long-range structure. Hence all possible interactions should be incorporated so as to correctly model the system. This prediction is quite difficult with excess Gibbs free energy models such as NRTL and UNIQUAC. The latter is a group contribution model (GCM), which depends on the parameters for different functional groups. Quantum-chemical-based solvation models such as COSMO-SAC provide an alternative means of predicting activity coefficients and other thermodynamic properties using a statistical mechanical framework. The model is well-presented in the excellent work of Lin et al.31 and Klamt et al.,32 hence this is not discussed here. The first step for a COSMO-based scheme is to obtain an accurate geometry optimized structure of the molecules (water, alcohol, DL-menthol, and lauric acid) in gas phase. The geometry optimization was carried out with Gaussian 09,43 from at least three initial geometries of each complex systems. The geometries were drawn by Gauss View 5.044 visualization package. For DL-menthol which has equal contribution from both dextro and levo stereoisomer, the geometries for each stereoisomer were optimized in Gaussian09. Frequency analysis was carried out on each complex at the same level of theory. Absence of imaginary vibrational frequencies determine the true energy minimum structures. For the COSMO-SAC calculation, BVP86 functional with SVP

⎧ ⎡ −Epair(σ , σ ′) + μ (σ ′) ⎤⎫ ⎪ ⎪ S ⎥⎬ + kT ln pS (σ ) μS (σ ) = −kT ln⎨ ∑ exp⎢⎢ ⎪ ⎪ ⎥ kT ⎣ ⎦ ⎩ σ′ ⎭

(10)

Here μS(σ) is the chemical potential of a surface segment with charge density σ in a solution S. Epair(σ,σ′) is the total interaction energy as provided in COSMO formalism of two adjacent screening charges σ,σ′ in the solution S. pS(σ) is the sigma profile of the solution.31,32 This distribution has been termed as σ-potential, and it describes the affinity of the solvent for a molecular surface of polarity σ. In this regard, it should be H


Article


entrapped in the DES core primarily due to the difference in shape and sizes of the molecules. The influence of this ratio can be inferred from the molecular dynamic simulation of DES with 1-butanol so as to obtain the coordination numbers. This will also provide us more insight into the extraction regime. In summary, the ratio of DES and lauric acid or HBA to HBD needs to be judiciously selected based on both water and thermal stability for future aqueous phase extraction.

noted that an extra amount of free energy is expanded from the newly formed hydrogen bonds, especially with the addition of a hydrogen-bond donor or acceptor segments. Hence an opposing effect is seen between the free energy which is used for removing the screening charges (positive in magnitude) and that of forming hydrogen bonds which is negative in magnitude. Thus if a strong hydrogen bond is formed, the net restoring free energy, which is a sum of the misfit and the hydrogen bonding contribution, would be negative. Water is both a hydrogen-bond donor and acceptor, therefore its σ potential extends outside both H-bond cutoff values (±0.0084 e/Å2) which is reflected in its negative values of sigma potential on both the side of axes (Figure 7). Sigma

■

CONCLUSIONS The application of a hydrophobic deep eutectic and lowtransition temperature mixture for the lower alcohol extraction has been investigated for the first time. Deep eutectic solvent has been synthesized with a hydrogen bond acceptor, namely DL-menthol, along with a hydrogen bond donor, lauric acid. Thereafter with the synthesized DES, extraction of ethanol 5− 40 mol %, 1-propanol 5−30 mol %, and 1-butanol 10−50 mol % from the aqueous solution was measured. The distribution coefficient and selectivity were found to be much higher for 1butanol as compared to ethanol and 1-propanol. The 1H NMR analysis depicted an absence of DES in the raffinate phase. This indicates that the DES does not contaminate the water phase, thereby enabling the ease of solvent recycling. The NRTL and COSMO-SAC model gave RMSD value of 0.01 and 0.07, respectively, implying that NRTL model gave an excellent match with the experimental data. Both NRTL and COSMOSAC model gave RMSD value of 0.01 and 0.07, respectively, thereby giving good prediction with the experimental data.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.iecr.7b05270. The 1H NMR spectra of pure DES, saturated DES, extract phase, and raffinate phase is provided (PDF)

Figure 7. Sigma potentials of 1-butanol, water, and DES.

profile of DES, ethanol, 1-propanol, and 1-butanol is shown in Figure 8. Both 1-butanol and DES exhibits a negative value of sigma potential beyond the H-bond cutoff values (> −0.0084 e/Å2) primarily due to its respective oxygen atom(s). However, for DES, the segment potential is positive and parabolic in nature within the donor region (> +0.0084 e/Å2). This indicates that the contribution is only due to misfit interactions. Hence it can be said that the molecules of 1-butanol are

■

AUTHOR INFORMATION

Corresponding Author

*Tel.: +91-361-2582266. Fax: +91-361-2582291. E-mail: [email protected]. ORCID

Tamal Banerjee: 0000-0001-8624-6586 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors are thankful to Mr. Mood Mohan for the useful discussion regarding the computation of solid liquid equilibria for DES solvents.

■

REFERENCES

(1) Marszałek, J.; Kamiński, W. Concentration of butanol-ethanolacetone-water using pervaporation. Proc. ECOpole 2012, 6 (1), 31−36. (2) Kaminski, W.; Tomczak, E.; Gorak, A. Biobutanol-production and purification methods. Ecol. Chem. Eng. S 2011, 18 (1), 31−37. (3) Maddox, I. S. The acetone-butanol-ethanol fermentation: recent progress in technology. Biotechnol. Genet. Eng. Rev. 1989, 7 (1), 189− 220. (4) Qureshi, N.; Liu, S.; Ezeji, T. Cellulosic Butanol Production from Agricultural Biomass and Residues: Recent Advances in Technology. In Advanced Biofuels and Bioproducts; Springer, 2013.

Figure 8. Sigma profile of DES, ethanol, 1-propanol, and 1-butanol. I


Article


Extraction Systems Using COSMO-RS. Ind. Eng. Chem. Res. 2007, 46, 1292−1304. (25) Banerjee, T.; Singh, M. K.; Sahoo, R. K.; Khanna, A. Volume, surface and UNIQUAC interaction parameters for imidazolium based ionic liquids via Polarizable Continuum Model. Fluid Phase Equilib. 2005, 234 (1−2), 64−76. (26) Mohan, M.; Balaji, C.; Goud, V. V.; Banerjee, T. Thermodynamic Insights in the Separation of Cellulose/Hemicellulose Components from Lignocellulosic Biomass Using Ionic Liquids. J. Solution Chem. 2015, 44, 538−557. (27) Mohan, M.; Goud, V. V.; Banerjee, T. Solubility of glucose, xylose, fructose and galactose in ionic liquids: Experimental and theoretical studies using a continuum solvation model. Fluid Phase Equilib. 2015, 395, 33−43. (28) Gouveia, A. S.; Oliveira, F. S.; Kurnia, K. A.; Marrucho, I. M. Deep eutectic solvents as azeotrope breakers: liquid−liquid extraction and COSMO-RS prediction. ACS Sustainable Chem. Eng. 2016, 4 (10), 5640−5650. (29) Kundu, D.; Banerjee, T. Multicomponent vapor−liquid−liquid equilibrium prediction using an a priori segment based model. Ind. Eng. Chem. Res. 2011, 50 (24), 14090−14096. (30) Bharti, A.; Kundu, D.; Rabari, D.; Banerjee, T. Phase Equilibria in Ionic Liquid Facilitated Liquid−Liquid Extractions; CRC Press, 2017. (31) Lin, S. T.; Sandler, S. I. A priori phase equilibrium prediction from a segment contribution solvation model. Ind. Eng. Chem. Res. 2002, 41 (5), 899−913. (32) Klamt, A. Conductor-like Screening Model for Real Solvents: A New Approach to the Quantitative Calculation of Solvation Phenomena. J. Phys. Chem. 1995, 99 (7), 2224−2235. (33) Fernandez, L.; Silva, L. P.; Martins, M. A.; Ferreira, O.; Ortega, J.; Pinho, S. P.; Coutinho, J. A. Indirect assessment of the fusion properties of choline chloride from solid-liquid equilibria data. Fluid Phase Equilib. 2017, 448, 9−14. (34) García, S.; Larriba, M.; García, J.; Torrecilla, J. S.; Rodríguez, F. Liquid−liquid extraction of toluene from n-heptane using binary mixtures of N-butylpyridinium tetrafluoroborate and N-butylpyridinium bis (trifluoromethylsulfonyl) imide ionic liquids. Chem. Eng. J. 2012, 180, 210−215. (35) Gonzalez, A. S. B.; Francisco, M.; Jimeno, G.; de Dios, S. L. G.; Kroon, M. C. Liquid−liquid equilibrium data for the systems {LTTM+ benzene+ hexane} and {LTTM+ ethyl acetate+ hexane} at different temperatures and atmospheric pressure. Fluid Phase Equilib. 2013, 360, 54−62. (36) Tang, B.; Bi, W.; Tian, M.; Row, K. H. Application of ionic liquid for extraction and separation of bioactive compounds from plants. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2012, 904, 1− 21. (37) Wasserscheid, P.; van Hal, R.; Bösmann, A. 1-n-Butyl-3methylimidazolium ([bmim]) octylsulfatean even ‘greener’ionic liquid. Green Chem. 2002, 4 (4), 400−404. (38) Nann, A.; Held, C.; Sadowski, G. Liquid−liquid equilibria of 1butanol/water/IL systems. Ind. Eng. Chem. Res. 2013, 52 (51), 18472− 18481. (39) Mohan, M.; Banerjee, T.; Goud, V. V. Solid Liquid Equilibrium of Cellobiose, Sucrose, and Maltose Monohydrate in Ionic Liquids: Experimental and Quantum Chemical Insights. J. Chem. Eng. Data 2016, 61 (9), 2923−2932. (40) Harvianto, G. R.; Kim, S. E.; Jin, I. B.; Kang, K. J.; Lee, M. Liquid−Liquid Equilibria Data for the Quaternary System of Acetic Acid, Water, p-Xylene, and Ethyl Acetate at 313.15 K and 101.325 kPa. J. Chem. Eng. Data 2016, 61 (2), 780−787. (41) Singh, M. K.; Banerjee, T.; Khanna, A. Genetic algorithm to estimate interaction parameters of multicomponent systems for liquid−liquid equilibria. Comput. Chem. Eng. 2005, 29 (8), 1712− 1719. (42) Sahoo, R. K.; Banerjee, T.; Ahmad, S. A.; Khanna, A. Improved binary parameters using GA for multi-component aromatic extraction: NRTL model without and with closure equations. Fluid Phase Equilib. 2006, 239 (1), 107−119.

(5) Liu, X.; Wang, H.; Zheng, Z.; Liu, J.; Reitz, R. D.; Yao, M. Development of a combined reduced primary reference fuel-alcohols (methanol/ethanol/propanols/butanols/n-pentanol) mechanism for engine applications. Energy 2016, 114, 542−558. (6) Qureshi, N.; Maddox, I. S.; Friedl, A. Application of continuous substrate feeding to the ABE fermentation: relief of product inhibition using extraction, perstraction, stripping, and pervaporation. Biotechnol. Prog. 1992, 8 (5), 382−390. (7) Heitmann, S.; Krings, J.; Kreis, P.; Lennert, A.; Pitner, W.; Górak, A.; Schulte, M. Recovery of n-butanol using ionic liquid-based pervaporation membranes. Sep. Purif. Technol. 2012, 97, 108−114. (8) Groot, W. J.; Soedjak, H. S.; Donck, P. B.; Lans, R. G. J. M. V. d.; Luyben, K. C. A. M.; Timmer, J. M. K. Butanol recovery from fermentations by liquid-liquid extraction and membrane solvent extraction. Bioprocess Eng. 1990, 5 (5), 203−216. (9) Ha, S. H.; Mai, N. L.; Koo, Y. M. Butanol recovery from aqueous solution into ionic liquids by liquid−liquid extraction. Process Biochem. 2010, 45 (12), 1899−1903. (10) Kraemer, K.; Harwardt, A.; Bronneberg, R.; Marquardt, W. Separation of butanol from acetone−butanol−ethanol fermentation by a hybrid extraction−distillation process. Comput. Chem. Eng. 2011, 35 (5), 949−963. (11) Neves, C. M.; Granjo, J. F.; Freire, M. G.; Robertson, A.; Oliveira, N. M.; Coutinho, J. A. Separation of ethanol−water mixtures by liquid−liquid extraction using phosphonium-based ionic liquids. Green Chem. 2011, 13 (6), 1517−1526. (12) Garcia-Chavez, L. Y. G.; Garsia, C. M.; Schuur, B.; de Haan, A. B. Biobutanol recovery using nonfluorinated task-specific ionic liquids. Ind. Eng. Chem. Res. 2012, 51 (24), 8293−8301. (13) Chapeaux, A.; Simoni, L. D.; Ronan, T. S.; Stadtherr, M. A.; Brennecke, J. F. Extraction of alcohols from water with 1-hexyl-3methylimidazolium bis (trifluoromethylsulfonyl) imide. Green Chem. 2008, 10 (12), 1301−1306. (14) Pereiro, A.; Araújo, J.; Esperança, J.; Marrucho, I.; Rebelo, L. Ionic liquids in separations of azeotropic systems−A review. J. Chem. Thermodyn. 2012, 46, 2−28. (15) Mohsenzadeh, A.; Al-Wahaibi, Y.; Jibril, A.; Al-Hajri, R.; Shuwa, S. The novel use of deep eutectic solvents for enhancing heavy oil recovery. J. Pet. Sci. Eng. 2015, 130, 6−15. (16) Naik, P. K.; Dehury, P.; Paul, S.; Banerjee, T. Evaluation of Deep Eutectic Solvent for the selective extraction of toluene and quinoline at T= 308.15 K and p= 1 bar. Fluid Phase Equilib. 2016, 423, 146−155. (17) Rodriguez, N. R.; Ferre Guell, J.; Kroon, M. C. Glycerol-Based Deep Eutectic Solvents as Extractants for the Separation of MEK and Ethanol via Liquid−Liquid Extraction. J. Chem. Eng. Data 2016, 61 (2), 865−872. (18) Mohan, M.; Naik, P. K.; Banerjee, T.; Goud, V. V.; Paul, S. Solubility of glucose in tetrabutylammonium bromide based deep eutectic solvents: Experimental and molecular dynamic simulations. Fluid Phase Equilib. 2017, 448, 168−177. (19) Florindo, C.; Branco, L.; Marrucho, I. Development of hydrophobic deep eutectic solvents for extraction of pesticides from aqueous environments. Fluid Phase Equilib. 2017, 448, 135−142. (20) Ribeiro, B. D.; Florindo, C.; Iff, L. C.; Coelho, M. A.; Marrucho, I. M. Menthol-based eutectic mixtures: hydrophobic low viscosity solvents. ACS Sustainable Chem. Eng. 2015, 3 (10), 2469−2477. (21) Zhang, Q.; De Oliveira Vigier, K.; Royer, S.; Jérôme, F. Deep eutectic solvents: syntheses, properties and applications. Chem. Soc. Rev. 2012, 41 (21), 7108−7146. (22) Dai, Y.; van Spronsen, J.; Witkamp, G. J.; Verpoorte, R.; Choi, Y. H. Natural deep eutectic solvents as new potential media for green technology. Anal. Chim. Acta 2013, 766, 61−68. (23) Anantharaj, R.; Banerjee, T. COSMO-RS based predictions for the desulphurization of diesel oil using ionic liquids: Effect of cation and anion combination. Fuel Process. Technol. 2011, 92, 39−52. (24) Banerjee, T.; Sahoo, R. K.; Rath, S. S.; Kumar, R.; Khanna, A. Multicomponent Liquid−Liquid Equilibria Prediction for Aromatic J


Article

Industrial & Engineering Chemistry Research (43) Frisch, M.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, M. T.; A, J. J.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Keith, T.; Kobayashi, R.; J., Normand; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision B. 01; Gaussian, Inc: Wallingford, CT, 2010. (44) Dennington, R.; Keith, T.; Millam, J. GaussView, version 5; Semichem Inc.: Shawnee Mission, KS, 2009. (45) Banerjee, T.; Verma, K. K.; Khanna, A. Liquid−liquid equilibrium for ionic liquid systems using COSMO-RS: Effect of cation and anion dissociation. AIChE J. 2008, 54 (7), 1874−1885. (46) Corvis, Y.; Négrier, P.; Massip, S.; Leger, J.-M.; Espeau, P. Insights into the crystal structure, polymorphism and thermal behavior of menthol optical isomers and racemates. CrystEngComm 2012, 14 (20), 7055−7064. (47) Cháfer, A.; de la Torre, J.; Font, A.; Lladosa, E. Liquid−Liquid Equilibria of Water+ Ethanol+ 1-Butyl-3-methylimidazolium Bis (trifluoromethanesulfonyl) imide Ternary System: Measurements and Correlation at Different Temperatures. J. Chem. Eng. Data 2015, 60 (8), 2426−2433. (48) Cháfer, A.; De la Torre, J.; Lladosa, E.; Pla-Franco, J.; Cumplido, M. P. Liquid−Liquid Equilibria of the Water + 1-Propanol + 1-Butyl-1methylpyrrolidinium Bis(trifluoromethylsulfonyl)imide Ternary System: Study of the Ability of Ionic Liquid as a Solvent. J. Chem. Eng. Data 2016, 61 (12), 4006−4012.

K


LiquidâLiquid Extraction of Lower Alcohols Using ... - ACS Publications

Recommend Documents

LiquidâLiquid Extraction of Lower Alcohols Using ... - ACS Publications