The Redlich-Kister expression is primarily an empirical fit to data. The use of either the mole fraction or any other property depends on the type of interaction one would expect in the mixture of interest. If the mixture contains polar compounds and you expect the electro-static forces to be dominant, you would use something like an equivalent surface area of the molecule. If the mixture contains molecules of vastly different sizes, you can use a cell-structure (e.g. Flory-Huggins model) and then the number of cells occupied by a single molecule depends on the molar volume of the compounds.

The utilization in the Redlich-Kister equation of a particular variable of concentration depends on the definition of the excess or deviation property. For example, if you are correlating excess volumes the choice will be molar fractions; on the other hand if you are correlating excess refractive indices (J. C. R. Reis, et al., Refractive index of liquid mixtures: theory and experiment, Chemphyschem 11, 2010, 3722-3733 ) the choice will be volume fractions.

G=a+bT=H-TS juste the tow first parameters

Thank you for your interesting answers

Generally, the use of the molar fraction is the more suitable. It's a good thermodynamic variable.

Thank you for your answer 

It's a request:

Please, help me how to purchase an ancient article:

Industrial & Engineering Chemistry

Algebraic Representation of Thermodynamic Properties and the Classification of Solutions

Otto Redlich, A. T. Kister

Ind. Eng. Chem., 1948, 40 (2), pp 345–348

DOI: 10.1021/ie50458a036

The Publisher permits only to their members!!

@Noureddine Ouerfelli. You can find the file attached