Activity coefficients for binary and multicomponent liquid mixtures from unary property model for Wilson parameters

Abstract

A model has been developed for prediction of Wilson parameters from single component properties. Activity coefficients and VLE data for miscible binary or multicomponent systems can be calculated from the standard form of the Wilson equation using these properties. The major advantage of this model is that experimental determinations of binary data are not required. Relevant properties for a fairly large number of chemical species have been provided. This method has been compared with UNIFAC, which is generally used for prediction of activity coefficients and has been found to give comparable results with considerably less computational effort. (Tassios, 1971;Ladurelli et al., 1975;Hiranuma and Honma, 1975;Krumins et al., 1980) or to predict the binary parameters by using empirical correlations (Ghosh and Chopra, 1975;Gothard et al., 1976), but none has been of general utility. An approximate theoretical model, based on the assumption that the solution is locally regular, has been developed that successfully expresses Wilson parameters in terms of pure compound properties. The model is quite general in nature and uses the solubility parameter, "effective molar volume," and two deviation parameters for pure substances to predict the Wilson parameters. These properties for a number of polar and nonpolar compounds have been obtained. The results obtained by this method for a large number of binary and multicomponent systems have been compared with those predicted by the UNIFAC method. S. N. ASH CONCLUSIONS AND SIGNIFICANCEThe correlation obtained from the proposed model for prediction of Wilson binary parameters uses only three types of single component properties; the solubility parameter, effective molar volume, and the deviation parameter. This has been tested for a large number of binary and multicomponent systems, including hydrocarbons (aliphatic, cyclic, and aromatic), alcohols (primary and secondary), carbonyl compounds, esters, halogenated hydrocarbons, nitriles, and water. The computed average deviation in bubble point pressure and vapor phase The Wilson equation (Wilson, 1964) has been quite satisfactory for correlation of binary vapor-liquid equilibrium (VLE) data and for the prediction of multicomponent VLE behavior from relevant binary data. In his model Wilson used the energies of interaction At{ and hj or A, , between like and unlike pairs of molecules, but was not successful in attributing precise physical significance to or in obtaining expressions for these quantities. Fortunately, however, the model was so constructed that only the difference terms, (Xi, -hi) and (A, -A j j ) , appeared in the final expressions.Hence, Wilson could use these as adjustable parameters to be obtained from fitting of experimental binary equilibrium data. composition is small, normally within 2%. The results have also been compared with and found to be as good as those predicted by the UNIFAC method.The proposed correlations and expressions are remarkably simple and require only pure compound data. The method will con...

Experimental Binary VLE Data of Morpholine with 1-Butanol and 3-Methyl-1-butanol Systems

“…The Wilson single parameter for the data at each condition has been evaluated using the Ash version of the Wilson one-parameter model [Ash et al], and the results are given in Table . Barring the data of 1-butanol–morpholine, the Wilson single parameter, in general, correlated the data of the binary mixtures of this study with a little loss of accuracy compared to the two-parameter Wilson model.…”

“…Wilson single parameter models are quite useful in the online control of distillation columns. The approach of Ash et al is followed to evaluate the Wilson single parameter for a binary mixture at each condition. The binary interaction parameter K 12 is introduced in the expression for the Wilson model parameter.…”