Abstract
Quantum mechanical molecular descriptors and physical properties have been used to estimate log k(ow) values for a range of substituted phenols using multilinear regression. The quantum mechanical molecular descriptors were derived from a commercially available software package. Two models have been proposed. The first model involves molecular mass and E(HOMO) as independent variables whereas the second model combines the following molecular descriptors: molecular mass, melting point, charge on oxygen atom, E(HOMO), E(LUMO), molecular volume, total surface area, refractivity, and polarizability. The equations were derived using a set of 14 reference phenolic compounds with their log k(ow) values being the average of several experimental values reported in the literature. The predictive power of the derived model equations was compared to a set of 18 test compounds, whose log k(ow) values have been estimated by an isocratic liquid chromatography (LC) procedure. log k(ow) values of the test compounds estimated on the basis of the two equations showed good correlation with LC estimated log k(ow) values.
Original language | English |
---|---|
Pages (from-to) | 193-200 |
Journal | Analytica Chimica Acta |
Volume | 315 |
Issue number | 1-2 |
DOIs | |
Publication status | Published - 30 Oct 1995 |
Keywords
- Molecular modelling
- multilinear regression
- octanol-water partition coefficient
- quantitative structure property relationships
- quantum mechanical molecular descriptors