QSPR/QSAR Study On Mercaptans, Aliphatic Aldehyde Ketones And Sulfides

QSPR/QSAR (quantitative structure-property relationships, quantitative structure-activity relationships) has been a cutting-edge research in chemistry, environment, biological science and so on. As the invariants derived from the molecular structural graph, the topological indices can describe the size, shapes, branches of molecules efficiently. And the study based on the model constructed has become the importance and cutting-edge of QSPR/QSAR research by calculating and predicting properties/activities of molecules. And mean while, during recent years, in QSPR/QSAR, quantum chemistry has been a promising approach. Quantum chemistry index is a theoretical descriptor obtained by transforming molecular structures into the corresponding molecular graphs on the molecular level. Quantum chemistry has been an important method in QSPR/QSAR research. In this paper, we make the topological method combine with the quantum chemical method, and introduce the quantum parameters calculated by the quantum chemical method into the topological structure of molecules. And then we use the new topological indices to predict the properties or activity of compounds, which together with quantum chemical parameters.This work mainly concentrates on the follows:(1) Geometry optimizations and calculation All geometry optimizations and frequencies analyses were performed with the Gaussian 03 program, the molecules of mercaptans, aliphatic aldehyde ketones and sulfides were calculated by density functional theory (DFT) at the B3LYP level of theory. And calculate these parameters of the molecules : Gibbs free energy (G) , constant volume mole hot melting (CV), Entropy (S) , Partition function(LgQ) , Energy of highest occupied molecular orbital (HOMO) (?H), Energy of lowest unoccupide molecular orbital (LUMO) (?L), Electric charge of carbon and oxygen atom of carbonyl (QC,QO), Dipole Moments (?), Distance of atoms (d) and so on.(2) Index constructed To revise distance matrix of molecules of mercaptans, aliphatic aldehyde ketones and sulfides, change the topological distance of atoms into the space distance of atoms calculated by Gaussian 03 program. Intruduce the equilibrium electronegativity (? ) of the atoms into the main diagonal of distance matrix. Calculate the molecular quantum topological index QT according to Eq. :QT=logN*logλmax, whichλmax is the maximum eigenvalue of revised distance matrix.(3) Model development Build the multiple linear regression (MLR) model of mercaptans, aliphatic aldehyde ketones and sulfides by molecular quantum topological index QT together with quantum chemical parameters. In the models, most of the correlation coefficients of the 9 models of mercaptan were over 0.99; the correlation coefficients of boil point model of 72 aliphatic aldehyde ketones is over 0.99, while the chromatographic retention indices models are over 0.98; the correlation coefficient of the chromatographic retention indices models of 61 sulfides are over 0.93, and three of the four are over 0.95. And the models constructed by this method are reasonable and validity in total.(4) Model validation The leave-one-out cross-validation (LOO CV) method and random sampling prediction (RSP) method were used to testify the stability and predictive ability of the models. The multiple correlation coefficients RCV are close to the R values of regression of the models which demonstrate that the models have statistical significance and good stability.So this mehod should has more application in QSPR/QSAR study.