Extraction of pesticides by membrane separation. Transfer prediction by linear discriminant analysis and soft independent modelling of class analogy

Abstract

In this paper we describe a simple and versatile procedure for the extraction of different pesticides from organic matrices into aqueous solutions containing varying percentages of methanol, using an extraction cell equipped with a homogeneous silicone membrane. The pesticides studied show well-differentiated physicochemical properties, which are used to propose a classification: the analytes able to cross the silicone membrane and those unable to do so. The results obtained allowed us to make predictions about the behaviour of these analytes in this type of extraction. Linear Discriminant Analysis (lDA) and Soft Independent Modelling of Class Analogy (sIMCA) were applied to generate the classification and the modelling. In the latter, the variables with the greatest weight in discriminant power among groups were the partition coefficient (n-Octanol-Water) and the solubility of the pesticides in water. The results obtained show that it is possible to predict the ability of the pesticides to cross the silicone membrane on the basis of their physicochemical properties.