Inverse QSPR estimation of molar mass and unsaturation of FAME, BD, and blends from refractive index and speed of sound measurements

Abstract

Quantitative Structure-Properties Relationships (QSPR) are widely used for the prediction of physical and chemical properties in many pure substances. They have also been successfully applied to liquid mixtures, such as fatty acid methyl esters and biodiesel of known molecular composition. For these substances, we propose that molar averages of the number of carbon atoms and double bonds in the methyl ester molecules are relevant parameters for QSPR estimations. In this work, we define a system of Inverse QSPR (IQSPR) to estimate the molaraverages of carbon chain length, number of double bonds, and molar mass of fatty acid methyl esters, biodiesel, and blends from refractive index and speed of sound measurements. The comparison with chromatography data for saturated and unsaturated fatty acid methyl esters, with a chain length between 8 and 18 carbon atoms, and biodiesel samples from different feedstock, shows that the relative difference in the estimation of the molar mass is lower than 2.1 %. The difference (absolute value) between the estimated and experimental molar averages of chain length and number of double bonds per ester molecule are less than 0.34 and 0.14, respectively.