A method for the congener-specific analysis of multiple classes of persistent organic pollutants (POPs) has been developed. It consists of high-speed blending extraction, automated FMS cleanup/fractionation or GPC-Florisil cleanup, and final determination by gas chromatography and mass spectrometry (GC-MS). Fifty-five ¹³C₁₂- or ²H-labeled compounds were used as surrogates to evaluate the analytical efficiency. Proper polychlorinated biphenyl (PCB) and polybrominated biphenyl (PBB) congeners were used as internal standards to monitor the GC performance. Potential GC-MS chromatographic interferences were studied. The method was demonstrated to be reproducible, effective, and reliable and was applied successfully to determine POP concentrations in northwestern Atlantic harbor seal (Phoca vitulina concolor) blubber samples. PCBs and organochlorine pesticides (OCPs) account for more than 95% of the POP contamination, while non-ortho- and mono-ortho- substituted PCBs are the principal contributors to the total toxic equivalents (TEQs) in seal blubber.

Four non-ortho- and eight mono-ortho-substituted PCBs are potent inducers of hepatic aryl hydrocarbon hydroxylase (AHH). Therefore, evaluation of their environmental behaviors in terms of their physicochemical characteristics, such as Henry's Law constant (HLC), is very important. A modified gas purging technique was used to determine HLCs for these 12 PCBs. Measured HLCs range from 5.6 to 21.8 Pa·m³/mol and they are comparable with literature values. Meta-analysis technique and principal component regression (PCR) was applied to model the relationship between experimentally-determined HLCs for 97 PCBs and the congeners' structures. Results suggest that HLCs for PCB are primarily affected by meta-chlorine substitution and chlorine substitution on the biphenyl ring generally leads to smaller HLCs. The PCR-predicted HLCs are in good agreement with the experimental values and the predicted HLCs by a complicated partial least-square regression (PLSR) model using the molecular connectivity index as predictor.

Finally, photocatalytic degradation of 4,4'-dichlorobiphenyl was conducted in the TiO₂ aqueous suspensions. 4,4'-dichlorobiphenyl degradation follows the Langmuir-Hinshelwood kinetic model. Three hydroxylated byproducts, 4,4'-dichloro-2-biphenyl-ol, 4-chloro-4'-bipenyl-ol, and 4,4'-dichloro-3-biphenyl-ol were identified and they underwent further photocatalytic decomposition. Kinetic modeling indicates that hydroxylation is the major pathway for PCB decomposition by UV/TiO₂. A sketch of the reaction pathways for 4,4'-dichlorobiphenyl based on the byproducts formed has been proposed.