Estimation of the normal boiling point of organic compounds via a new group contribution method

Abstract

An improved second order group contribution method is developed to determine the normal boiling point of organic compounds. This method is based on the Joback's first order functional groups with some changes and some newly added functional groups developed from experimental data. Each functional group has a group contribution value, and the contribution values have been optimized using 2036 experimental data of organic components ranging from NC = 1–36 and MW = 26–555 (g.mol−1), including heavy and complex polycyclic aromatic compounds. In this method, it is tried to distinguish most of the structural isomers and stereoisomers by second order functional groups to predict their different normal boiling points. First and second order functional groups of the hydrocarbons and hydrocarbon derivatives containing carbon, hydrogen, oxygen, nitrogen, sulfur, fluorine, chlorine, bromine and iodine atoms, are included in this study. The fminsearch mathematical approach from MATLAB software is used in this study to select an optimal collection of functional groups (122 functional groups) and subsequently to develop the model. The results of the new method are compared to several well-known methods. The average absolute deviation of normal boiling point predictions for 2036 organic compounds is found to be 4.35194 K; while the percentage of the average absolute relative deviation is just 1.01075%.