Experimental Liquid Equilibrium Data Research Articles

Energy-saving extractive distillation system using o-xylene as an entrainer for the high-purity separation of dimethyl carbonate/methanol azeotrope

The pressure-swing distillation traditionally used in the industry for separating dimethyl carbonate/methanol (DMC/MeOH) commonly suffers from high energy consumption and challenges in achieving high-purity separation due to the existence of pinch zone. Hence, energy-saving extractive distillation system is explored to achieve the high-purity separation of DMC/MeOH azeotrope. With the analyses of isovolatility and residue curve maps based on Wilson model calibrated by vapor–liquid equilibrium experimental data, an extractive distillation process using O-Xylene (OX) as the most suitable entrainer were modeled and experimentally validated. Furthermore, design, optimization, and comparison of an extractive distillation process and a pressure-swing distillation process for producing 10000 t DMC/a are investigated. It is demonstrated that the extractive distillation process presents significant economic and emission advantage, specifically with 34.66 % and 26.95 % reductions in operating cost (OC) and in total annual cost (TAC) respectively, as well as 42.53 % reduction in CO2 emissions. This work should provide a feasible and promising approach to energy saving, low-carbon, and high-purity separation of DMC/MeOH azeotrope.

Surrogates for Liquid-Liquid Extraction.

Fuel surrogates are mixtures that mimic the properties of real fuels with only a small number of components, simplifying the calculation and simulation of fuel-related processes. This work extends a previously published surrogate optimization algorithm toward the generation of fuel surrogates with a focus on liquid-liquid extraction characteristics. For this purpose, experimental liquid-liquid equilibrium data from batch extraction experiments are incorporated into the calculation procedure as an additional constraint. The use of the method is demonstrated by optimizing a surrogate for the catalytic reformate. Application of the surrogate to an extraction process and comparison with experimental data demonstrate that the resulting surrogate accurately depicts the properties of the real mixture with regard to liquid-liquid extraction performance. This demonstrates that the use of such surrogates is of particular interest for mixtures used as extracting agents for biofuels.

Liquid–Liquid Equilibrium Measurement for the Separation of n-Propanol + n-Propyl Acetate Using Imidazolium-Based Ionic Liquids with Different Anions at T = 303.15 K

Separating the mixture of the target product n-propyl acetate (NPAC) and the unreacted reactant n-propanol (NPA) is necessary during the production of NPAC. NPA and NPAC cannot be separated by simple distillation because the two could form an azeotrope under atmospheric pressure. Liquid–liquid extraction is an energy-saving and environment-friendly method for separating azeotropic mixtures. Ionic liquids (ILs) are promising alternatives to conventional extractants for separating azeotropes due to the advantages of designability and negligible vapor pressure. In this study, the liquid–liquid equilibrium (LLE) data for the ternary systems of NPAC + NPA + 1-butyl-3-methylimidazolium dihydrogen phosphate ([BMIM][H2PO4]) or 1-butyl-3-methylimidazolium thiocyanate ([BMIM][SCN]) or 1-butyl-3-methylimidazolium tetrafluoroborate ([BMIM][BF4]) were determined at 303.15 K and 101.3 kPa. The accuracy of the experimental LLE data was checked by the Bachman, Hand, and Othmer–Tobias equations. The selectivity and partition coefficient were calculated to evaluate the extraction capability of the selected extractants. The experimental LLE data were correlated with the nonrandom two-liquid model, and the binary interaction parameters were regressed. Finally, the separation mechanism of the three ILs was explored through σ-profile analysis.

Liquid-liquid equilibria of the ternary systems water + C1 – C3 alcohols + dimethyl adipate at 298.15 K and atmospheric pressure: Experimental data and modeling

The aim of this study was to examine the possibility of using a green organic solvent, dimethyl adipate (DMA), as a separation agent in the extraction of different alcohols from aqueous solutions. For that purpose, the liquid–liquid equilibria (LLE) of four ternary systems were analytically determined at 298.15 K and atmospheric pressure using stirred and thermo-regulated cells: {water + methanol + dimethyl adipate (DMA)}, {water + ethanol + DMA}, {water + 1-propanol + DMA}, {water + 2-propanol + DMA}. Phase diagrams for these four ternary systems were obtained by determining a binodal curve and tie-lines data using a visual cloud-point method, but also gas chromatography. Distribution coefficients and separation factors were calculated for the immiscibility region. Hand and Othmer-Tobias correlations were used to examine the reliability of the tie-line data. In addition, the experimental ternary LLE data were correlated with the UNIversal QUAsiChemical (UNIQUAC) activity coefficient models. The obtained results showed that DMA is a suitable candidate for the extraction of methanol, ethanol, 1-propanol, and 2-propanol from water. With some alcohols, this potential is more noticeable than with others. The potential of DMA as a separation agent is increasing in the following sequence: methanol < ethanol < 2-propanol < 1-propanol, and that also could be noticed from the selectivity values that also increase in this sequence. The UNIQUAC model proved as successful in correlating the LLE data.

Liquid-liquid phase equilibrium measurement and thermodynamic modeling for the separation of n-hexane and methanol with four solvents

To solve the problem of generating n-hexane in the industrial production of methanol, four common and inexpensive solvents {dimethyl sulfoxide, 1,2-propanediol, furfuryl alcohol, and furfural} were chosen as extractants to separate n-hexane and methanol. The liquid–liquid equilibrium (LLE) data of n-hexane + methanol +{dimethyl sulfoxide, 1,2-propanediol, furfuryl alcohol, or furfural} were measured at 303.2 K and 101.3 kPa. The extraction performance of four extractants was evaluated by the distribution coefficient (D) and separation factor (S). At the same time, NRTL and UNIQUAC models were used to fit the experimental LLE data to obtain regression parameters. The reliability of the binary interaction parameters was verified by the mixed surface analysis based on Gibbs energy topology. The root mean square deviation (RMSD) of the NRTL and UNIQUAC models are less than 0.0040 and 0.0057, respectively, indicating that the NRTL and UNIQUAC models are suitable for the study of this system. Furthermore, the interaction energy between extractant and methanol molecules was analyzed by the DMol3 module and the effect difference of liquid–liquid extraction was explained from the perspective of interaction energy using σ-Profile.

Mechanism analysis of liquid-liquid extraction separation for dimethyl carbonate and methanol by choline-based DESs as extractants

Dimethyl carbonate (DMC) is a widely used chemical raw material and chemical product with excellent performance. In this manuscript, based on electrostatic potential (ESP), reduced density gradient (RDG) function, non-bond interaction energy, radial distribution function (RDF), spatial distribution function (SDF) and self-diffusion coefficient, the separation of DMC / methanol (MeOH) by four deep eutectic solvents (DESs) was studied. The results of ESP analysis showed that the hydrogen bond receptor played a leading role in the separation process, and the urea in the four hydrogen bond donors was more likely to form a stable structure with MeOH. The weak interaction region between different DESs and azeotropes was studied by RDG function. Based on the non-bond interaction energy, it can be seen that Cl− can form a very strong interaction with MeOH, and Electrostatic energy (Eelec) plays a leading role in the separation process. The charge density around the central atom can be observed by RDF and SDF, so as to determine the interaction between azeotrope and DESs. The diffusion rate of azeotrope in different DESs was studied by self-diffusion coefficient. The results showed that MeOH moved the slowest in the DES composed of urea and choline chloride (CHCl). The effect of DES structure on the liquid level of methanol dimethyl carbonate system was clarified by analyzing the experimental liquid-liquid equilibrium data. In this study, based on quantum chemistry (QC) calculation and molecular dynamics (MD) simulation, this study analyzed the liquid-liquid extraction and separation mechanism of DMC and MeOH. The extraction effect of DES composed of CHCl+urea is the best, which has certain theoretical guiding significance for liquid-liquid extraction to realize the green and efficient separation of DMC and MeOH.

Measurement and correlation of liquid-liquid equilibrium for the ternary system (water + 1,2-dichloroethane + sulfolane) at 288.15, 298.15, and 308.15 K

Sulfolane is an important aprotic polar solvent. Liquid-liquid equilibrium (LLE) data for the ternary systems of water + 1,2-dichloroethane + sulfolane were measured at temperatures of 288.15, 298.15 and 308.15 K under the atmospheric pressure. The distribution coefficient and selectivity were determined from the measured LLE data, which showed that 1,2-dichloroethane is a suitable extractant for the recovery of sulfolane from its aqueous solution. The nonrandom two-liquid (NRTL) model and the universal quasi-chemical (UNIQUAC) model were utilized to correlate the experimental LLE data. The low values of RMSD indicated that the ternary system could be fitted well by the NRTL and UNIQUAC models. The consistency of the binary interaction parameters for the two thermodynamic models obtained was confirmed by the topological information contained in the Gibbs energy of mixing function ( G M / RT ).

Capturing VOCs in the pharmaceutical industry with ionic liquids

1-ethyl-3-methylimidazolium acetate ([EMIM][Ac]) was adopted to absorb two common pharmaceutical volatile organic compounds (VOCs), dichloromethane (DCM) and ethanol. The vapor–liquid equilibrium experimental data of the VOC-ionic liquid (IL) systems were measured. The absorption experiment was carried out at 20 °C and at ambient pressure, with a gas flow rate of 1 L/min and an absorbent flow rate of 20 mL/min. The results showed that the removal ratios of DCM and ethanol reached 91.82% and 97.01%, respectively. Furthermore, the mechanism of IL capturing DCM and ethanol was systematically studied using the COSMO-RS model and a quantum chemical calculation. The results demonstrated that the separation capability of [EMIM][Ac] mainly depended on the strong [Ac]−-VOC interaction. Finally, industrial-scale VOC capture processes were established and optimized. The results showed that IL process achieved a lower total heat duty and a higher increase in green degree compared with triethylene glycol process.

Quaternary liquid-liquid equilibria measurements and NRTL modelling for the (Water + Ethanol + 2-Propanol + 2,2,4-Trimethylpentane) system at 298.2K to 328.2K

In this work, the influence of a second alcohol on the phase behaviour of (water + alcohol + entrainer) systems was investigated by measuring Liquid-Liquid Equilibrium (LLE) data for the quaternary system of (water + ethanol + 2-propanol + isooctane) under isobaric (0.1 MPa) conditions. Data were measured at three ethanol: 2-propanol ratios to characterise the quaternary system behaviour, and for the constituent ternary (water + ethanol + isooctane) and (water + 2-propanol + isooctane) systems at 298.2 K. The measured quaternary data exhibit an evolution in phase behaviour between the constituent ternary systems in both the miscibility and total alcohol distribution between the aqueous and organic phases. The individual alcohol behaviour highlights the differing intermolecular interactions occurring in the systems, resulting in an increased recovery of ethanol to the organic phase and 2-propanol to the aqueous phase in the quaternary system. Further, the influence of temperature on the quaternary system's phase behaviour was investigated by measuring the quaternary system's data at four temperatures (298.2 K, 308.22 K, 318.2 K, and 328.2 K).The ternary (water + alcohol + entrainer) LLE literature and experimental data were correlated well using the Non-Random Two Liquid (NRTL) activity coefficient model with binary interaction parameters (BIPs) incorporating linear temperature dependence. These BIPs acceptably predicted the quaternary system's total alcohol behaviour, but only qualitatively captured the correct individual alcohol behaviour.

Liquid – liquid equilibria in binary and ternary systems of phenol + hydrocarbons (n–dodecane or n–hexadecane) and water + phenol + hydrocarbons (n–dodecane or n–hexadecane) at temperatures between 298K and 353K

This study reports liquid–liquid equilibrium (LLE) and liquid–liquid–liquid equilibrium (LLLE) data for binary (phenol + n-dodecane, or n-hexadecane) and ternary (water + phenol + n-dodecane, or n-hexadecane) systems measured under atmospheric pressure. The compositions of coexisting phases were determined with analytical and cloud point methods at temperatures 298 K – 353 K. The Non–Random Two–Liquid (NRTL) excess Gibbs energy model was employed to correlate the measured systems. The binary interaction parameters were regressed using analytical LLE and cloud point data. In addition, the parameters were also calculated using the binary LLE data combined with the isothermal vapor–liquid data from the literature applying the NRTL–RK (Redlich–Kwong) property method. The average absolute deviations in liquid mole fraction obtained with the NRTL model (using six adjusted parameters) for the LLE and VLE experimental data were 0.006 and 0.014 respectively. The phase equilibria of binary phenol + hydrocarbon (n-dodecane or n-hexadecane) systems were modeled at the temperature range of 313 K – 573 K.

Liquid-liquid equilibrium studies on the removal of naphthalene / 2-methylnaphthalene / dibenzothiophene from model oil using ionic liquids

Liquid-liquid extraction, has been used as an alternative method, for the dearomatization and desulfurization of a model oil using ionic liquids (ILs). The most recent work employs six ILs of high selectivity to aromatic and sulfur-containing compounds in ternary liquid-liquid equilibrium (LLE) systems. The best three ILs from the LLE ternary systems were used in a model multicomponent systems. The ILs were selected as extractive solvents for the extraction of naphthalene (N), 2-methylnaphthalene (2MN) and dibenzothiophene (DBT) from 1,3-diisopropylbenzene (1,3DIPB) in the ternary LLE data. Ternary mixtures of {IL (1) + N , or 2MN, or DBT (2) + 1,3DIPB (3)} and multicomponent mixtures {IL + N + 2MN + DBT + 1,3DIPB} were investigated at temperature T = 313.15 K and pressure 101 kPa. For this reason, the imidazolium-, pyridinium- and morpholinium-based ILs were selected: 1-ethyl-3-methylimidazolium dicyanamide, [EMIM][DCA], 1-allyl-3-methylimidazolium dicyanamide, [AMIM][DCA], 1‑butyl‑3-methylimidazolium tiocyanate, [BMIM][SCN], 1‑butyl‑4-methylpyridinium dicyanamide, [B 4 MPy][DCA], N-(2-hydroxyethyl)pyridinium tricyanomethanide, [Py2OH][TCM] and (3-hydroxypropyl)-1-methylmorpholinium dicyanamide, [ N C 3 OHMMOR][DCA]. The suitability of the ILs as extractive solvents has been evaluated in terms of solute distribution ratio and selectivity. All the ILs showed high selectivity in the extraction of N, 2MN and DBT from 1,3DIPB in ternary and multicomponent systems. All LLE experimental data were correlated with the NRTL model. The highest extraction efficiency of N, 2MN and DBT in LLE data was observed for [B 4 MPy][DCA] (22.3 wt%, 36.8 wt% and 45.5 wt%, respectively). Much higher extraction efficiency was obtained in extraction of N, 2MN and DBT from the model oil stream in a single extraction stage (2IL:1 M.Str., m/m) and in two-stage extraction (2IL:1 M.Str., m/m). The extraction efficiency of N, 2MN and DBT from the model oil stream (2IL:1 M.Str., m/m) was 71.1 wt%, 80.4 wt% and 76.4 wt%, with [BMIM][SCN], or 69.3 wt%, 86.4 wt% and 82.9 wt% with [B 4 MPy][DCA], or 69.2 wt%, 69.4 wt% and 69.0 wt% with [AMIM][DCA], respectively in two-stage extraction. Graphical Abstract .

COSMO-RS prediction, liquid-liquid equilibrium experiment and quantum chemistry calculation for the separation of n-butanol and n-heptane system using ionic liquids

Ionic liquids (ILs) have been extensively used in the separation of alkane + alcohol azeotropes as entrainers by liquid–liquid extraction. n-Butanol is widely added to the fuel because of its high calorific value and low volatility. It is of significance to recovery n-butanol from the azeotropic mixture containing n-butanol. In this work, ionic liquids, formed by combining 17 cations and 12 anions, were screened based on the COSMO-RS model. 1-ethyl-3-methylimidazolium dicyanamide [EMIM][DCA], 1-butyl-3-methylimidazolium dicyanamide [BMIM][DCA] and 1-hexyl-3-methylimidazolium dicyanamide [HMIM][DCA] were selected as entrainers. The liquid–liquid equilibrium data of the ternary system {n-heptane + n-butanol + [EMIM][DCA], n-heptane + n-butanol + [BMIM][DCA], n-heptane + n-butanol + [HMIM][DCA]} were measured at 298.15 K and 101.3 kPa. The quality of experimental data was tested by Othmer-Tobias, Bachman and Hand equations. The distribution coefficient (β) and selectivity (S) were calculated to evaluate the separation effect of the three ILs. The NRTL model was used to correlate liquid–liquid equilibrium experimental data. Finally, quantum mechanics calculations were performed to study the interactions between ILs and n-butanol through the reduced density gradient (RDG) function and quantum theory of atom in the molecule (QTAIM). The results of quantum chemistry calculation showed that the separation of n-butanol and n-heptane mixture by ionic liquids with [DCA] anions were dominated by hydrogen bonding. The research process is devoted to the development of green entrainers.

Liquid-liquid and solid-liquid equilibria of several PEG-based ABS with ionic liquid [C4C1im]Br as adjuvant at 298.15 K

Liquid-liquid equilibrium (LLE) and liquid–liquid-solid equilibrium (LLSE) data were determined, at T = 298.15 K, for aqueous biphasic systems (ABS) composed of PEG (poly(ethylene glycol)) with a molecular weight 400, different inorganic salts (K2CO3, (NH4)2SO4, or MnSO4), and the ionic liquid (IL) [C4C1im]Br as adjuvant. Binodal curves of two-liquid phases, LLE and LLSE data were determined by the addition of 5 wt% [C4C1im]Br to investigate its effect on the phase diagrams. All binodal curves were determined by the cloud-point titration method and described accurately using the Merchuk equation. The results show that the presence of 5 wt% of [C4C1im]Br produce small effects upon the phase separation. Compositions of LLE and LLSE were determined by gravimetric method, UV-spectroscopy and X-ray diffraction analysis made on the solids. For PEG400/K2CO3 and PEG400/(NH4)2SO4 ABS, the obtained LLE data are in good agreement with those obtained by analytical methods quantifying the concentrations of PEG using refractive index measurements, and of salt by flame photometry or the formaldehyde method, respectively, with the maximum absolute average deviation of 1.94%. The tie-line lengths (TLL) were also addressed. Densities of the coexisting phases were also reported at 298.15 K and correlated satisfactorily using the Othmer rule. Moreover, it was found that [C4C1im]Br has a higher affinity for the PEG-rich phase, which increases with increasing the system composition and the water-structuring effect of the anion of the salt used as the salting-out agent to form ABS. For the sulfate-based ABS, whereas, the tendency of [C4C1im]Br to partition into the PEG-rich phase was observed to decrease with increasing the water-structuring effect of the cations, which may be attributed to the complex forming tendency of Mn2+ with the ether oxygens of PEG. Furthermore, the solid phases in equilibrium with the biphasic regions were K2CO3·1.5H2O, anhydrous (NH4)2SO4, and MnSO4·4H2O. Finally, the experimental LLE data were accurately correlated using the NRTL model.

Ethylsulphate based ionic liquids for denitification of liquid fuels

Ternary liquid-liquid equilibrium (LLE) for eight systems containing 1-ethyl-3-methylimidazolium ethylsulphate {[EMIM][EtSO4]}, 1-ethyl-3-methylpyridinium ethylsulphate {[EMPy][EtSO4]} (1) + pyridine/pyrrole/indoline/quinoline (2) + dodecene (3) have been determined at 298.15 K. The reliability of experimental LLE data was tested by Othmer-Tobias and Hand equations. The NRTL and UNIQUAC models were successfully used for experimental LLE data correlation and gave the RMSD less than unity, which indicated a good degree of experimental LLE data fit. Distribution coefficient and the selectivity of [EMIM][EtSO4] and [EMPy][EtSO4] were calculated for all the studied systems. Here, high distribution coefficient and selectivity were obtained for the removal of pyridine/pyrrole/indoline/quinoline from dodecene using EMIM][EtSO4] and [EMPy][EtSO4] ionic liquids at 298.15 K.

Liquid-liquid equilibrium data measurement and thermodynamic modelling for ternary mixtures composed of water, diethylene glycol dimethyl ether and different solvents at 298.2 K

Liquid-liquid equilibrium (LLE) data for water + diethylene glycol dimethyl ether + {methyl isobutyl carbinol, methyl isobutyl ketone, methyl isopropyl ketone, methyl tert-butyl ether, n-octanol and 2-ethyl-1-hexanol} was measured at 298.2 K under 101.3 kPa. The effect of separating diethylene glycol dimethyl ether from water is evaluated by the distribution coefficient and the separation coefficient. The Othmer-Tobias and Hand equations were chosen to examine the experimental LLE data. Meanwhile, the data was correlated by the NRTL and UNIQUAC thermodynamic models with AAD% less than 0.55 and RMSD% less than 0.77. The separation performance of six solvents was compared using the σ-profile analysis via the exploration of the intermolecular interaction.

Low viscosity lactam-based ionic liquids with carboxylate anions: Application in the separation of systems toluene/heptane, cyclohexene/cyclohexane, and phenol/water

Low viscosity lactam-based ionic liquids (ILs) were evaluated as extractor solvents in organic components separation. Experimental data for liquid–liquid equilibrium (LLE) were obtained from the toluene + heptane + IL, cyclohexene + cyclohexane + IL and phenol + water + IL systems at 298.15 K and atmospheric pressure. The reliability of experimental data has been verified by the Othmer-Tobias, Bachman and Hand equations. ILs γ-butyrolactam acetate (BTA), γ-butyrolactam butyrate (BTB), and γ-butyrolactam hexanoate (BTH); and ε-caprolactam acetate (CPA) and ε-caprolactam hexanoate (CPH) were used. It has been observed that anion structures influence on all miscibility of ternary mixtures, while cation structures influence on cyclohexene + cyclohexane + IL. Ionic liquids containing lower chain anions (BTA, BTB and CPA) formed two phases with heptane and toluene mixture. However, ILs containing higher chain anions (BTH and CPH) formed heterogeneous systems only with the water + phenol mixture. Only BTA showed partial miscibility in the cyclohexene + cyclohexane system. ILs containing acetate anion presented selectivity higher than 1 in aromatic and aliphatic separation. This also happened on lower concentrations of cyclohexene in olefin and paraffin separation. CPA presented selectivity higher than BTA in liquid–liquid extraction of toluene from heptane. In this system, BTA and CPA can be reused for three times without changes in the chemical structure of ILs and with a small reduction in selectivity (5.4 to 3.8 for BTA and 17.2 to 11.8 for CPA). The thermodynamic model NRTL was used to obtain binary interaction parameters and to correlate LLE experimental data for toluene + heptane + IL system. A comparison between experimental and calculated data indicates a good representation of data. ILs containing hexanoate anion presented degradation in water, the hexanoic acid remained in organic-phase, while γ-butyrolactam was transferred to water-phase. Therefore, it was not possible to determine LLE data.

Liquid-liquid separation of n-hexane/1-hexene and cyclohexane/cyclohexene using deep eutectic solvents

In this study, the feasibility of twelve different deep eutectic solvents (DESs) in separating olefins from paraffins was investigated. The DESs were prepared by combining choline chloride (ChCl) or tetrabutyl ammonium bromide (TBAB) with six different hydrogen bond donors. The preliminary COSMO-RS screening revealed that TBAB/TEG (1:4) DES has the high performance index (PI) for both systems. This performance was confirmed by experimental tests at 20% molar ratio of olefins in the feed mixture with the twelve DESs. When TBAB is used, the best distribution ratios were obtained for cyclohexane/cyclohexene system, however for hexane/1-hexene system high selectivities were gotten. In comparison with the performance of ionic liquids, TBAB-based DESs showed better distribution ratios and comparable selectivity. For the cyclohexane/cyclohexene system, TBAB-based DESs exhibited high distribution ratios because of their hydrophobic character represented by butyl chain. In the case of hexane/1-hexene system, TBAB-based DESs are more selective to 1-hexene due to the affinity of TBAB chain and positive charge bringing by ammonium group and 1-hexene. The liquid–liquid equilibrium (LLE) was explored for two ternary systems of {hexane (1) + 1-hexene (2) + DES (3)} and {cyclohexane (1) + cyclohexene (2) + DES (3)} at 298.15 K and 101 kPa using TBAB/TEG (1:4) DES. Furthermore, the NRTL model was successfully used to correlate the experimental LLE data and good predictions were obtained with COSMO-RS for both selected systems. Finally, the extraction mechanism was elucidated by applying 1H NMR analysis and no DESs were found in the raffinate phase.

Liquid–liquid equilibrium calculations of systems containing vegetable oil + fatty acids + ethanol + water using new parameters for the unifac subgroups of ethanol and water

Deacidification removes free fatty acids (FFA) from crude vegetable oils during the refining process. The removal of FFA carried out by liquid–liquid extraction has some advantages over chemical and physical refining processes: it saves energy and decreases neutral oil losses caused by undesirable chemical reactions. The knowledge of phase equilibrium behavior of systems involving vegetable oils, solvents, and FFA is necessary to design and optimize the liquid–liquid extraction processes for deacidification. In this context, the present work aims to predict the liquid–liquid equilibrium (LLE) for systems consisting of vegetable oil + fatty acids + ethanol + water, readjusting new UNIFAC interaction parameters just for the ethanol and water subgroups. Several experimental LLE data for multicomponent fatty systems available in the literature are used to adjust the new UNIFAC interaction parameters for the ethanol (EtOH) and water subgroups. The temperature range studied was 288.15 K to 323.15 K, at atmospheric pressure. The new 18 interaction parameters provide a considerably lower mean square error (0.87%) than the other works available in the literature. The results show that the developed procedures improve the capacity of the UNIFAC method to predict LLE for systems involved in the deacidification process of crude vegetable oil. The methodology employed in this work results in reduced computational calculations and a relatively small split with the complex dataset established by the UNIFAC-LL model.

Application of deep neural network (DNN) for experimental liquid-liquid equilibrium data of water + butyric acid + 5-methyl-2-hexanone ternary systems

LLE data are important for simulation and design of extraction equipment. In this study, deep neural network (DNN) structure was proposed for modelling of the ternary liquid-liquid equilibrium (LLE). LLE data of (water + butyric acid + 5-methyl-2-hexanone) ternaries defined at three different temperatures of 298.2, 308.2, and 318.2 K and P = 101.3 kPa, were obtained experimentally and then correlated with nonrandom two-liquid (NRTL) and universal quasi-chemical (UNIQUAC) models. The performance of the proposed DNN model was compared with that of NRTL and UNIQUAC in terms of the root mean square errors (RMSE). RMSE values were obtained between 0.02-0.06 for NRTL and UNIQUAC, respectively. For DNN, the error values were obtained between 0.00005-0.01 for all temperatures. According to the calculated RMSE values, it was shown that proposed DNN structure can be better choice for the modelling of LLE system. Othmer-Tobias and Hand correlations were also used for the experimental tie-lines. Distribution coefficient and separation factors were calculated from the experimental data.

Liquid-liquid equilibrium for the ternary system of (2-methyl-2-butanol + water + n-octanol) at different temperatures

Liquid-liquid equilibrium (LLE) data for the ternary system of (2-methyl-2-butanol + water + n-octanol) were determined at 303.15 K, 313.15 K, and 323.15 K under atmospheric pressure. The Hand equation, the Othmer-Tobias equation, and the mass balance between the feed and the conjugated phases were used to test the reliability of the experimental LLE data. The results show that the correlation coefficients (R2) of the equations are close to 1, and the relative error of mass balance is less than 0.5%. The selectivity (S) was used to evaluate the capacity of n-octanol to extract 2-methyl-2-butanol from the water, which is much greater than 1. Furthermore, S decreases with the mass fraction of 2-methyl-2-butanol in the aqueous phase and increases with equilibrium temperature. The NRTL and UNIQUAC models were applied to correlate the experimental LLE data, and the binary interaction parameters were obtained. The root-mean square deviations (RMSDs) for the two models obtained are less than 1.1%.