| dc.description.abstract | The application of molecularly imprinted polymers (MIPs) as sorbents in solid phase extraction (SPE) leading to molecularly imprinted SPE (MISPE) is among the most advanced application scenarios for molecular imprints, and is nowadays considered a viable alternative to conventional procedures of sample pre-treatment. Furthermore, the selectivity and stability that MIPs provide evidently facilitates thorough sample pre-treatment, which is important for ultra-trace level detection. Alternative promising sorbents are multi-walled carbon nanotubes (MWCNTs), which have been the subject of extensive research mainly due to their outstanding chemical and mechanical stability, electrical properties, and high surface area. In addition, owing to the π-electron system of the graphene sheets, MWCNTs strongly interact with aromatic compounds, and were therefore together with MIPs investigated during the present thesis research as potential sorbents for the pre-concentration of polychlorinated biphenyls (PCBs) from various environmental matrices. Therefore, following the dummy template approach - whereby the template used is not necessarily the target analyte in the real-life application - MIPs for recognition of PCBs were synthesized. PCB 15 was used as the dummy template to model the compounds targeted for analysis in the present study, i.e., PCB 28, 52, 101, 138, 153, and 180. Since the synthetic protocol was aqueous suspension polymerization, the resulting particles were microspheres in the size range of 6 to 6 μm, as determined by scanning electron microscopy (SEM). The observed rough surfaces were in agreement with the recorded high specific surface areas ranging from 134.45 to 285.56 m2 g-1. Batch rebinding studies resulted in an imprinting factor of 1.09, whereby the low molecular recognition was attributed to the weak forces of interaction (i.e., π-π stacking) during the synthesis. In depth characterization of the polymers using Langmuir and Freundlich adsorption isotherms showed that the MIP had slightly higher maximum binding capacities than the non-imprinted polymer (NIP) (60.2 vs. 54.9 μg g-1). The MIP was also more heterogeneous than the NIP, which is characteristic of imprinted polymers because of the heterogeneous distribution of the binding site affinities. In addition, a blank prepared in the absence of both the template and the functional monomer gave the lowest binding capacity (23.5 μg g-1) and revealed a homogenous distribution of binding site affinities.
xiv Closer to the intended application, optimization of the MISPE for application in cases where the analytes are in organic medium resulted in recoveries of 94.9-99.0% compared to conventional C18 materials, which gave recoveries between 69.2 and 78.9%, with significant losses in the loading step. C18 therefore demonstrated poor interactions with non-polar PCBs, as compared to MIPs. On the other hand, commercially available sulfoxide-modified silica showed enhanced recoveries (83.8-90.2%).Further optimization for application in aqueous environment resulted in decreased recoveries for both C18 and sulfoxide-modified silica (<70%), while the MISPE maintained recoveries >70%. Consequently, the two commercial sorbents were ruled out as possible sorbents for pre-concentration of PCBs in aqueous environment; therefore, MISPE was applied in pre-concentration of spiked real-world water samples giving recoveries >80%. The realized detection limits ranging from 0.002 to 0.01 ng mL-1 were lower than the set maximum contaminant level of 0.5 ng mL-1 for ∑PCBs in drinking water; implying that the developed protocol may be applied for monitoring of PCBs in environmental contamination scenarios. Towards the determination of PCBs in complex matrices, the MISPE was then modified and optimized for this purpose. Soil and sediment were the matrices of interest, as they are considered geological sinks for organics, thus serving as important matrices for monitoring of PCBs. Due to the complexity of these two matrices, a specific column for clean-up of extracts from the two matrices was designed using MIP particles and acidified silica gel. The cartridge was optimized and applied in validation of a full method from extraction to quantification using spiked soil and sediment samples. Recoveries >70% for all the analytes were realized. The method was then applied for the determination and quantification of PCBs in real-world soil and sediment samples, which had been collected in Kenya. Besides being able to quantify PCBs in both soil and sediment, the developed method offered clean extracts showing little or no interferences during the subsequent chromatographic analysis, and achieved low detection limits. Furthermore, the volume of extraction solvent used was reduced significantly, as compared to conventional clean-up methods. Since MIPs have in the past proved to be stable and robust materials, in the present study these polymers have confirmed their unequalled stability by maintaining high recoveries (>80%) even after 30 cycles of use and regeneration. | en_US |
