Prediction of Liquid-Liquid Equilibria of Binary Systems Containing Alcohols Using Eos-Ge Models

In this work, the equation of state-excess gibbs energy (eos-ge) model has been employed to predict liquid-liquid equilibria (lle) of some binary mixtures at 0.1 mpa. three binary systems containing decane+methanol, cyclohexane+methanol and cyclohexane+2,2,2-trifluoroethanol (tfe) have been tested using peng–robinson eos modified by stryjek and vera (prsv), along with the excess gibbs mixing rules mhv1 and mhv2 and two excess gibbs models: wilson equation modified by tsuboka and katayama (t–k–wilson) and uniquac equation. the interaction parameters of t-kwilson and uniquac ge models for three binary systems have been determined from lle data points at 0.1 mpa. the prediction ability of the models has been evaluated by comparison of the results with experimental data. average absolute error (aae) of 0.047 for mhv1 and t–k–wilson model, 0.0117 for mhv2 and t–k–wilson model, 0.0317 for mhv1 and uniquac model and 0.0109 for mhv2 and uniquac model have been obtained. as it is clear, the combination of mhv2 excess gibbs mixing rule with uniquac equation shows a satisfactory agreement with experimental data.