• 3-Component Margules equation was used to predict vapor–liquid equilibria. • Infinite dilution activity coefficients are the sole input for solvent pre-selection. • Industrial cases were used to validate the approach. • Molecular solvents, ILs and DESs were incorporated in the screening. We present an easily accessible, open-access approach for fast pre-selection of solvents for extractive distillation at isobaric and isothermal conditions. The method uses the three-component Margules equation, which can predict vapor–liquid equilibria (VLE) in ternary systems with the infinite dilution activity coefficients, γ i ∞ , as sole input parameters. This approach is accessible for anybody regardless of the availability of process simulators or other dedicated software to predict VLE, and can be combined with open access γ i ∞ to perform solvent screening. This approach shows a deviation in VLE of <5% for non-hydrogen bond donating mixtures, while for highly dissimilar (e.g. alkane – alcohol) mixtures the deviation can be >10%. The presented method identifies molecular solvents for case studies where the same or similar solvents have also been reported in literature or are already applied on the industrial level, showing realistic pre-selection outcomes. Furthermore, small, cyclic, polar, molecular solvents are identified to induce preferential interactions and several structurally similar solvents, e.g. ethylene carbonate, dihydrolevoglucosenone and y-valerolactone, are identified to be potential alternative solvents. Among the ionic liquids (ILs) and deep eutectic solvents (DESs), the morpholinium and ammonium structures are identified to have the highest potential for increasing relative volatilities, they also show lower toxicity than other cations. This method thus proves to be able to perform early-stage pre-screening among new classes of solvents which generates information on the minimum required solvent to feed ratio ( S F min ) for energy-efficient distillation from γ i ∞ obtained by measurement or from literature.
Read full abstract