Enomoto, Kei; (2003) Application of linear solvation energy relationships to the prediction of important physico-chemical properties of agrochemicals. Doctoral thesis (Ph.D), UCL (University College London).

Text Application_of_linear_solvatio.pdf Download (33MB)

Abstract

Linear Free Energy Relationships (LFERs) were used to determine solvation descriptors for agrochemicals and predict physico-chemical and biological properties important in determining their biological efficacy and environmental fate. In the course of this work, an overview on important agrochemical properties was given, followed by a description of the LFER approach. The various descriptor estimation methods were illustrated and compared using a carefully selected representative dataset based on the Pesticide Manual 12th ed. The experimental determination of LFER descriptors showed the importance in the selection of reliable literature data and allowed the introduction of a new water-solvent partition coefficient measurement approach, the microshakeflask method. The results of this agrochemical study was then used to estimate a large number of physico-chemical properties including the Chromatography Hydrophobicity Index (CHO, aqueous solubility (log Sw), water-solvent partition coefficients (log Ps), air-solvent partition coefficients (log Ls) and other properties important in the study of agrochemistry. In addition, a comparative study was included of the chemistry of agrochemicals and pharmaceuticals as well as an LFER profile for compounds of environmental interest. New LFERs were established for the prediction of soil sorption (log Koc), vapour pressure (log VP) and melting point (log MPt), illustrating: the importance of the choice of the compounds in the training set (log VP) ; the importance of defining the property under study carefully (log MPt) ; the introduction of new descriptors (number of rotatable bonds for log MPt and aqueous solubility). Studies showed that, when reliable descriptors are available, the coefficients of the LFERs obtained using an agrochemical dataset are in agreement with those already established using a different training set. As a conclusion, this work showed that: LFER can be applied to a wide range of chemical classes; LFER can be reliable in predicting a wide range of physico-chemical properties; LFER can be easily applied, with the introduction of new user-friendly software such as Desalt and Absolv.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item