Estimating binary liquid composition from density and refractive index measurements : a comprehensive review of mixing rules

Abstract

A comprehensive review of refractive index mixing rules of binary solutions is presented. All published mixing rules, with the exception of the Wiener mixing rule, can be reformulated into expressions with a linear dependence in either mole- or volume-fractions. Comparison of the predictive performance of both types of linear mixture models show that mixture models which are linear in volume fractions fail to represent real data trends. Mixture refractive index data reveal that both positive and negative deviations from linear behaviour in this property with volume fraction can occur, explaining the inability of linear models in volume fraction to be generally applicable. Improved data fits are possible with the modified Eykman and the power mean mixing rules, provided that in both cases the adjustable constant can be tuned for each binary system of interest. Only the Lorentz-Lorenz molar refraction mixing rule that is linear in mole fraction provides excellent data fits for all the systems investigated, allowing for prediction of composition of solutions from combined density and refractive index measurements. The aforementioned conclusions are derived from extensive experimental data of binary systems consisting of n-alkanes (C12 to C32) and different organic solvents, ultimately providing a tool to determine liquid composition from refractive index measurements.