Comparison of experimentally and theoretically determined oxidation and photochemical transformation rates of some organohalogens to promote prediction of persistence

Sammanfattning: The diversity of choices we have to make everyday influence our environment and ourselves in more ways than most of us realise. Anthropogenic substances, such as flame retardants, date back as early as 450 BC when the Egyptians used alum to reduce flammability. The increasing demand for new articles has led to an increased production of chemical substances, for which many are commercially produced without complete knowledge on properties such as persistence, bioaccumulation and toxicology (PBT). Commercial compounds may be properly tested and denominated as “safe” regarding PBT properties, but their degradation products and/or metabolites may cause environmental impact.The availability of uniform and accurate data for prediction of persistence is of key importance for the understanding of chemical fate. A method to determine the susceptibility of chemicals to undergo oxidation in water has been developed and applied on several organohalogens, including PBDEs and OH-PBDEs. The method was used to determine reaction rates and the group of OH-PBDEs were subsequently subjected to photolysis by use of UV-light. Hence, susceptibility to undergo both oxidation and photolysis for the OH-PBDEs were investigated and compared to previously reported degradation rates on PBDEs.As a final step in promoting the prediction of persistence, Quantitative structure-property relationship (QSPR) models were performed on a set of compounds which had undergone photolytic degradation under similar conditions. The QSPRs were used as a preliminary step in predicting photolysis half-lives for chemical substances and to determine which physicochemical descriptors are of greatest importance thereof.This thesis presents the possibility of performing and assessing oxidation transformations on compounds of low and high water solubility, photolysis transformations in various media and using obtained data to predict behaviour via QSPR models, to promote predictions of persistence.