There has been increasing concern about the environmental release and dispersal of emerging contaminants (ECs) and their potential risks to human and ecosystem health. The in situ passive sampling tool, diffusive gradients in thin films (DGT), has been developed as a promising alternative to traditional grab sampling in environmental research of ECs, such as pharmaceuticals, endocrine disrupting chemicals (EDCs), and some types of flame retardants. This thesis explored the role of DGT in determining ECs and understanding their sources, fate and impact in aquatic environments.
The property range of organic compounds which can be routinely sampled with the present design of DGT device (PTFE membrane filter, agarose gel diffusion layer, and HLB binding layer) was investigated. Sorption experiments and DGT deployment with 9 model chemicals [organophosphate esters (OPEs) with a wide range of log KOW (0.8−9.5), molecular weight (182−435 Da)] and different functional groups showed compounds with high hydrophobicity and aromatic rings are prone to retention on membrane filters, which slows the transport of chemical to the binding resin of the sampler. The limitation of the current DGT device for some trace organics is adsorption in the diffusion layer, mainly in the membrane filter. However, it is possible to extend the DGT technique for a wider range of chemicals, for example, by replacing the current DGT membrane filter with a new type of membrane filter which does not interact with target analytes.
The potential effects of biofouling and post-deployment sample storage on DGT measurements were systematically investigated. Biofilms generated at the surface of DGT devices (8-day and 15-day) in summer and winter from a typical urban wastewater treatment plant were tested with 13 ECs; this study showed no effect on DGT measurements for most compounds. Four storage methods up to 2-month were evaluated; this study showed that intact samplers can be kept for up to 2-months at refrigerated temperature (4 °C) without significant effect on the measured concentration of the compounds, but if no refrigerators were available, keeping binding gels in elution solvent at room temperature would achieve comparable results.
DGT and grab sampling were used together to study sources and environmental fate of ECs in a dynamic river catchment, the River Thames in the United Kingdom. For chemicals that were relatively stable in the rivers, DGT and grab sampling provided equally good representativeness. For chemicals that showed high dynamic variation in water bodies, the DGT provided a better integral of loadings and exposure than grab sampling. It took a similar time to set up and collect the DGT passive sampling system and to collect grab samples. However, for later storage and sample treatment, DGT is much more space-, cost- and time-effective.
DGT, for the first time, was combined with a water quality model (LF2000-WQX) to study sources and environmental fate of ECs, taking trimethoprim as a case study. The model needs the following key input information for the EC: per capita emission, WWTPs and in-river removal rates. DGT measurements in the River Thames network were used to assess the ability of the model to predict reasonable concentrations. This study showed that LF2000-WQX is suitable for predicting point-source ECs; predicted concentrations agreed very well with DGT measurements in winter and the model performance can be improved by improving in-river removal rate, i.e., using different in-river removal rates considering local environmental conditions such as DOC in different river reaches.
The work in this thesis is a step forward to understand the current performance of the DGT sampler and to explore its role in studying organic contaminants. It has shown that DGT is an effective tool for studying environmental issues of trace organic contaminants.