Development and Evaluation of Parsimonious Solvation Models for the Estimation of Partitioning Properties of Passive Samplers

Abstract

In this thesis, a (2p-LFERs) two parameter Linear Free Energy Relationship have been designed and applied for varied sets of aquatic and air passive samplers. LFER models for polyethylene passive sampler to water, polyoxymethylene passive sampler to water, polyacrylic passive sampler to water, polydimethylsiloxane passive sampler to air and polyurethane foam passive sampler to air were designed. For water system the new 2p-LFER models excellently explained partitioning variability in the datasets with R2 and root mean square error (RMSE) ranging from 0.81 to 0.94 and from 0.45 to 0.53 log units respectively. For air system my 2p-LFER models exhibited R2 and root mean square error in the range of 0.94 to 0.93 and from 0.38 to 0.54 log units. All models designed in this study were found statistically robust after testing by using four independent cross validations tests. The models developed provide us how the traits of hydrophobicity, volatility – octanol-water partition coefficient (Kow), air-water partition coefficient (Kaw) and organic carbon partition coefficient (Koc), manage the transfer of contaminants from one phase to another and indicates the dominance of octanol-water partition coefficient (Kow), hydrophobicity, in both aquatic and air phases. The significance of my research models is that they have a benefit over the predefined estimation approaches. The previous models either require super-fast computers, or parameter demanding, or the required parameters are not experimentally accessible or they are remarkably expensive. Whereas, the suggested LFER methods in this study are not only simple and accessible but also productive than different multi-parameter estimation methods. My 2p-LFER model is not only statistically potent but also theoretically-accurate and compliments the principle of parsimony.