Selected traditional structural descriptors and RM values for estimation and prediction of lipophilicity of homologous series of saturated fatty acids
Alina Pyka, Katarzyna Bober · Journal of the American Oil Chemists Society · 2006
Abstract An homologous series of carboxylic FA, from methanoic to triacontanoic, were characterized by selected structural descriptors based on distance matrix (W, A, 0B, 1B), adjacency matrix (M, Mv, 0χ, 1χ, 2χ, 0χv, 1χv, 2χv, F, N2), and information theory (ISA, ) as well as electrotopological states. Saturated FA, from pentanoic to eicosanoic as well as from heptanoic to tricontanoic acids, were separated by reversed‐phase high‐performance TLC (RP‐HPTLC) using mobile phases of methanol/water. The relationships existing between experimental partition coefficients (log Pexp), partition coefficients obtained from an Internet database (IAlogP and ClogPTM), and RM values (obtained by means of RP‐HPTLC, where RMis the hydrophobicity index) or structural descriptors (based on adjacency matrix: M, Mν, 0χ, 1χ, 2χ, 0χv, 1χv, 2χv, F, N2 and information theory: ISA, ) of saturated FA were investigated. Comparison and estimation of the usefulness of RM values and selected structural descriptors for calculating and predicting partition coefficients of saturated FA are the novel components of the work presented. The authors found that both values for RM and structural descriptors can serve for calculating and predicting the partition coefficients of FA investigated; but the selected traditional structural descriptors better describe the partition coefficients than the RM values.