QUANTITATIVE STRUCTURE- PROPERTY RELATIONSHIP STUDY FOR PREDICTION OF BOILING POINT OF ALIPHATIC ALKANES

摘要

Objective: QSPR (quantitative structure property relationship) models represent well-established tools for the molecular design of new compounds with desired properties. These are statistically based and aimed at extracting the maximum information from experimental data on compounds of known structure. The main objective of the study is to calculate the topological descriptors and to generate the multiple linear regression models for the prediction of boiling point of aliphatic alkanes. Methods: Quantitative structure-property relationship (QSPR) studies had been performed on 30 compounds of a series of aliphatic alkanes. Multivariable models were developed for calculating a number of topological descriptors to predict the boiling points of 30 alkanes having 1-10 carbon atoms. The QSPR models were generated to determine multiple correlation coefficient and standard error by using SPSS software. The experimental boiling point values have been correlated with calculated boiling point by using multiple linear regression (MLR ) analysis. Results: This study produced good predictive models which gave statistically significant correlations with good multiple correlation coefficient (R = 0.997) and minimum standard error (SE = 4.753) using topological descriptors as QSPR parameters. Conclusion: The models were used to predict the boiling points of alkanes for a set of test data from 1-10 carbon atoms for which no experimental boiling point data existed