In this work, dimethyl isosorbide (DMI) and 1-butylpyrrolidin-2-one (NBP), as biobased and greener organic solvents, were used for the first time as entrainers in extractive distillation to separate a close-boiling mixture of methylcyclohexane and toluene. Vapor–liquid equilibrium (VLE) data were collected for pseudoternary mixtures consisting of methylcyclohexane and toluene in the presence of DMI and NBP at various entrainer-to-feed ratios (E/F) and pressures. The VLE measurements were conducted by using a Fischer Labodest VLE602 ebulliometer, and the thermodynamic consistency of the data was verified by using the Van Ness test. Both DMI and NBP were found to increase the relative volatility of methylcyclohexane to toluene, successfully eliminating close-boiling behavior. Compared to benchmark entrainers, both outperformed 1-methylpyrrolidin-2-one (NMP) and sulfolane under certain conditions. In comparison with other green entrainers, DMI and NBP showed similar performance to gamma-valerolactone (GVL) and Cyrene under specific conditions. The VLE data were accurately correlated by using the nonrandom two-liquid (NRTL) model.

Green solvents have emerged as promising green entrainers to substitute conventional entrainers in extractive distillation to separate azeotropic mixtures. However, the limited availability of thermodynamic data for green-solvent-containing mixtures continues to hinder their practical implementation in this process. This study is the first to report experimental vapor–liquid equilibrium (VLE) data for the n-hexane + ethanol azeotropic system containing the greener entrainer 1-butylpyrrolidin-2-one (NBP) alongside the benchmark entrainer 1-methylpyrrolidin-2-one (NMP). Using a Fischer Labodest VLE602 ebulliometer, VLE measurements were performed at pressures of 50.0 and 100.0 kPa and various entrainer-to-feed ratios (E/F). The reliability of the reported VLE data was tested and confirmed using the Van Ness thermodynamic consistency test. The results show that NBP enhances relative volatility and effectively eliminates the azeotrope, performing comparably to the benchmark entrainer NMP. The nonrandom-two-liquid (NRTL) model was utilized to regress the investigated VLE data and determine the optimum binary interaction parameters (BIPs). As a result, the NRTL model demonstrates good agreement with the experimental data. This thermodynamic modeling confirms the data’s reliability and suitability for process design, highlighting NBP’s potential as an environmentally friendly alternative entrainer in extractive distillation.

The isobaric vapor–liquid equilibrium (VLE) data of the binary mixture of methylcyclohexane (1) + toluene (2) at 101.3 kPa; the pseudoternary mixture of methylcyclohexane (1) + toluene (2) + gamma-valerolactone (GVL) (3) with the entrainer-to-feed ratio (E/F) = 1 (mass basis) at 50, 80, and 100 kPa, and E/F = 2 and 3 at 100 kPa; and the pseudoternary mixture of methylcyclohexane (1) + toluene (2) + 1-methylpyrrolidin-2-one (NMP) (3) with E/F = 1 at 100 kPa were measured using a Fischer Labodest VLE602 ebulliometer. The reliability of the experimental VLE data was tested and confirmed by Van Ness and Fredenslund thermodynamic consistency tests. The experimental results indicate that the presence of GVL and NMP increases the relative volatility of methylcyclohexane to toluene; therefore, both entrainers remove a close-boiling behavior in the mixture. Non-random two-liquid (NRTL) and universal quasi chemical (UNIQUAC) thermodynamic models were applied in the experimental data correlation to obtain the optimum binary interaction parameters. For the mixture involving GVL, the experimental VLE data were accurately correlated by NRTL and UNIQUAC. However, NRTL has more accurate results compared with UNIQUAC. For the mixture containing NMP, both the UNIQUAC and NRTL models show favorable regression results.

Green organic entrainers and natural deep eutectic solvents (NADESs) possessing high boiling points and decomposition temperatures, exhibit considerable potential as advanced materials for green entrainers in extractive distillation. However, there is a wide range of solvents to choose from and their properties are only rarely available for use in process design and simulation. The present study aims to assess the selection parameters and examine the performance of and select green organic entrainers and NADESs for the separation of the close boiling mixture methylcyclohexane-toluene. The evaluation carried out was based on selectivity and relative volatility values obtained by using predictive models. COSMO-RS (a unimolecular quantum chemical calculation) was employed to predict the selectivity at infinite dilution, while group contribution methods such as UNIFAC and modified UNIFAC (Dortmund) were used to predict the relative volatility. According to the calculated results, the selectivity appeared a more important selection parameter than the performance index. The relative volatility prediction using the UNIFAC and UNIFAC Dortmund methods exhibits comparable trends to the selectivity results derived from COSMO-RS. However, the use of UNIFAC and modified UNIFAC (Dortmund) in predicting the relative volatility of NADES containing mixtures is limited due to the absence of functional group parameters. This CAPE study reveals that, based on the calculated selectivity (using COSMO-RS) and relative volatility (using UNIFAC, and modified UNIFAC (Dortmund)), most of the proposed green organic entrainers and NADESs exhibit a higher or comparable selectivity and relative volatility as the benchmark entrainers. This confirms the potential of the evaluated green entrainers with higher selectivity and relative volatility to enhance the design of extractive distillation. Therefore, the cost, energy and water consumption, as well as CO2 emissions in the methylcyclohexane and toluene separation can be reduced.