The validity of predicting properties from molecular structure using a quantitative structure-property approach (QSPR) has been well established A general empirical relationship has been developed for estimatand applied to a variety of classes of compounds and properties, using ing the cloud point of pure nonionic surfactants of the alkyl ethoxtopological, geometrical, electrostatic, and quantum-chemical descriptors ylate class. For a set of 62 structures, composed of linear alkyl, (7). One important aspect of the resulting regressions is that properties can branched alkyl, cyclic alkyl, and alkylphenyl ethoxylates, cloud be predicted directly from the molecular structure, for molecules which points can be estimated to an accuracy of {6.3ะC (3.7ะC median have never been synthesized. The application of topological indices to error) using the logarithm of the number of ethylene oxide residues surfactant property prediction has been established, predicting the critical and three topological descriptors that account for hydrophobic micelle concentration from surfactant structure for the nonionic (8) and domain variation. แญง 1997 Academic Press anionic (9) surfactants.