Screening alternative degreasing solvents using multivariate analysis

Multivariate analysis was used to explore physicochemical properties of org anic chemicals that would characterize and identify degreasing solvents. Th e exploratory techniques used in this study include cluster analysis, discr iminant function analysis, and canonical discriminant analysis. Out of a co mpilation of 16 physicochemical properties evaluated, aqueous solubility, H enry's constant, and surface tension were identified as relevant properties that could effectively screen degreasing solvents from among 30 chemicals of similar chemical classes. The suitability of these three properties and the multivariate techniques used in classifying degreasing solvents were de monstrated on an external testing set of 10 solvent- and nonsolvent-type ch emicals. On the basis of the results of these studies, canonical discrimina nt analysis is recommended as a potential tool for screening purposes. The cluster analysis procedure was informative for explorative purposes; the di scriminant function analysis procedure was not efficient in separating solv ents from others.