Autori: Oprea, Tudor Ionel; Kurunczi, Ludovic; Timofei, Simona
Editorial: Dyes Pigm., 33(1), p.41-64, 1997.
Twenty-seven disperse azo dyes were analyzed using Quant. Structure-Activity Relationship (QSAR) methods by correlating variations in the chem. structure with –°, the affinity to cellulose fiber. Classical QSAR results, r2 of 0.32 for CLogP (the calcd. octanol-water partition coeff.) and r2 of 0.924 for MTD (min. topol. difference), suggest that steric, but not hydrophobic, effects are important. For Comparative Mol. Field Anal. (CoMFA), a 3-dimensional QSAR (3D-QSAR) method, r2 was 0.925, while q2 (cross-validated r2) was 0.776 for 2 PCs (principal components). CoMFA results imply that the pharmacophore theory of dye-fiber interaction holds true. However, CoMFA was insensitive to the alignment rules. PCA (Principal Component Anal.) shows that PC1 is related to chem. substituents, whereas PC2 is related to mol. length (l). The correlation between –° and l (a 1D descriptor) is similar to the CoMFA results. The validity of the pharmacophore theory of dye-fiber interaction is questioned, illustrated by a case of overfitting in QSAR. Features that could improve disperse azo dye binding to cellulose are proposed.
Cuvinte cheie: Biological Simulation and Modeling, Cellulose fibers, CoMFA, Disperse dyes, Dye-Fiber interaction, MTD, Pharmacophore theory, QSAR (Structure-Activity relationships)