The assement of chemical persistence is an important part of legislative protection of the environment and human health. Of the vast number of chemicals on the market today few have been properly assessed. The coordination between testing guidelines from different frameworks is limited and especially the methods for determination of biodegradation show poor reproducibility because of their highly complex nature.

In order to circumvent the multifactorial assessment methods that involve the use of e.g. soils and sediments an attempt to create a new approach to chemicals assessment was postulated by Green and Bergman in 2005. This approach puts the focus on testing the chemical reactivity of the compound in environmentally relevant transformation/degradation reactions, i.e. reduction, oxidation, hydrolysis-substitution-elimination (hse), radical reactions, and photolysis. These tests are to be performed in controlled abiotic laboratory experiments ensuring that the results reflect the transformation rate of the intended type of reaction for the investigated substance. To achieve an assessment of the presistence of the compound, the test results are then combined with data on physicochemical properties of the compound and a mathematic matrix describing the reactive power of the different types of reactions in each environmental compartment (air, water, soil, and sediment).

Thus far methods for testing of oxidation, photolysis, and hydrolysis-substitution-elimination reactions have been developed. Within this thesis a method for determining reduction was developed and further utilised to determine transformation products from reductive debromination of the three nonabrominated diphenyl ethers. The previously established method for hse was evaluated and further developed in a study of selected chlorobenzenes. Some novel brominated flame retardants were investigated using the previously developed photolysis method, and transformation products and quantum yields were determined. All of the papers presented within this thesis intend to build on the project of a new persistency assessment model. The results presented also contributes important information on the properties and transformation of some common organohalogen pollutants.

A method was developed to study reductive transformation of highly brominated diphenyl ethers (BDEs). The method development is a part of a broader project where it will be used to determine the susceptibility of environmental pollutants to reductive conditions, in an attempt to create a scheme for determination of chemical’s persistence. This paper focuses on identification of octabrominated diphenyl ether transformation products from reductive debromination of the three nonabrominated diphenyl congeners (nonaBDE), BDE-206, -207 and -208. Sodium borohydride was used to explore the reductive debromination of the nonaBDEs. The transformation products were collected at two time-points and identified products were quantified by GC–MS. The reduction of the nonaBDEs lead primarily to debrominated products, mainly octaBDEs. The three nonabrominated DEs gave isomer-related transformation product patterns. BDE-207 and BDE-208 showed a propensity for ortho-debromination in the initial reaction step, while no discrimination between initial debromination positions was seen for BDE-206. All three nonabrominated DEs displayed a preferred initial debromination on the fully brominated DE ring.

An existing substitution reaction method was improved as part of a system to measure the persistency of selected chemicals in the environment by evaluating their chemical reactivity. Hexachlorobenzene (HCB), pentachlorobenzene, 1,2,4,5-tetrachlorobenzene and 1,2,4-trichlorobenzene were selected as model compounds. Sodium methoxide in methanol was used as a nucleophile and found to be stable for at least 35 days after preparation. The substitution reaction system was modified so that nitrogen protection was not necessary to avoid oxidation and hydrolysis effects, which led to improved results. The reactivity of polychlorinated benzenes (PCBz), which substituted chlorine for methoxide according to a second-order rate constant, increased as the number of chlorine atoms on the benzene ring increased. HCB was selected as a standard for the k₂ calculations of the substitution reactions. A normalized k₂ (k_N) was calculated as k_N = k_PCBz/k_HCB. GC–MS analysis confirmed that the reactions were pure nucleophilic aromatic substitutions without side reactions.