PREDICTION OF VAPOR-LIQUID EQUILIBRIUM FROM HEAT OF MIXING DATA FOR BINARY HYDROCARBON-ALCOHOL MIXTURES

The method of Hanks, Gupta, and Christensen is applied to the Continuous Linear Association Model (CLAM) of Renon and Prausnitz to predict vapory-liquid equilibrium (VLE) data from heat of mixing (hE) data for 35 binary mixtures of aliphatic or aromatic hydrocarbons with alcohols. The CLAM equation appears to account properly for the pronounced deviations from ideal solution behavior (large hE) which arise from alcohol association. VLE data for all 35 hydrocarbon-alcohol mixtures are predicted to within 10% standard deviation or less, averaging 5.5% for all systems. This is to be compared with an average 24.6% deviation from Raoult's law with some deviations therefrom being in excess of 71%. VLE data were predicted and are presented for eight hydrocarbon-alcohol mixtures, pentanol-heptane, pentanol-octane, pentanol-nonane, pentanol-decane, ethanol-nonane, octane-hexanol, nonane-octanol, and decane-octanol, for which only hE data existed in the literature. A method for predicting VLE data for murticomponent mixtures of aliphatic and aromatic hydrocarbons with one alcohol is suggested. Possible methods for handling multiple alcohols are also suggested. © 1979, American Chemical Society. All rights reserved.