Authors: Chandan Raychaudhury Debnath Pal
Publish Date: 2013/07/27
Volume: 51, Issue: 10, Pages: 2718-2730
Abstract
Entropy is a fundamental thermodynamic property that has attracted a wide attention across domains including chemistry Inference of entropy of chemical compounds using various approaches has been a widely studied topic However many aspects of entropy in chemical compounds remain unexplained In the present work we propose two new informationtheoretical molecular descriptors for the prediction of gas phase thermal entropy of organic compounds The descriptors reflect the bulk and size of the compounds as well as the gross topological symmetry in their structures all of which are believed to determine entropy A high correlation hbox r2 = 092 between the entropy values and our informationtheoretical indices have been found and the predicted entropy values obtained from the corresponding statistically significant regression model have been found to be within acceptable approximation We provide additional mathematical result in the form of a theorem and proof that might further help in assessing changes in gas phase thermal entropy values with the changes in molecular structures The proposed informationtheoretical molecular descriptors regression model and the mathematical result are expected to augment predictions of gas phase thermal entropy for a large number of chemical compounds
Keywords: