Prediction of Boiling and Melting Points for Some Substituted Alkanes and their Isomers using Group Contribution Techniques

Abstract

The prediction of thermodynamic properties such as boiling (T_b) and melting (T_m) points is fundamental in chemical process design. This study evaluates the performance of three Group Contribution Techniques (GCTs): Constantinou – Gani (C & G), Marrero – Gani (M & G), and Joback – Reid (J & R) — in predicting two thermodynamic properties for selected substituted alkanes (SAs) and their isomers. The results indicate significant variations in predictions, with the J & R method generally estimating higher T_b, while the M & G method provides lower T_m values for most compounds. Discrepancies among the methods highlight the influence of molecular structure and functional groups on phase transition properties. The mean absolute error (MAE) across the methods was found to be 35.38 K for T_b and 39.06 K for T_m, highlighting variability in predictive accuracy. Standard deviation analysis further revealed significant fluctuations, with an average of 50.17 K for T_b and 51.21 K for T_m. Analysis of variance indicated no statistically significant difference between the three methods for T_b and T_m, suggesting that all three models perform comparably despite individual discrepancies. The findings underscore the need for further refinement of predictive models to enhance their reliability for industrial and research applications.