Open Access

Cordula Witzig

• Plastics have many beneficial properties, making them an integral part of everyday life. However, the continuously increasing global production of plastics, as well as their durability and extensive use, entail a serious consequence: The accumulation of plastic fragments of various sizes in the environment. Small plastic fragments, so-called microplastics (MPs), are particularly mobile and have been detected not only in all environmental compartments but also in the most remote parts of the world. Assessing the extent of environmental pollution by MPs is crucial to understanding their impact on ecosystems and, ultimately, on humans, as well as for implementing effective countermeasures. Indeed, the occurrence of MPs in the environment has been intensively researched since the early 2000s, but MP concentrations reported by different studies often deviate by several orders of magnitude, even for the same environmental compartment. Possible reasons for this discrepancy include (i) a highPlastics have many beneficial properties, making them an integral part of everyday life. However, the continuously increasing global production of plastics, as well as their durability and extensive use, entail a serious consequence: The accumulation of plastic fragments of various sizes in the environment. Small plastic fragments, so-called microplastics (MPs), are particularly mobile and have been detected not only in all environmental compartments but also in the most remote parts of the world. Assessing the extent of environmental pollution by MPs is crucial to understanding their impact on ecosystems and, ultimately, on humans, as well as for implementing effective countermeasures. Indeed, the occurrence of MPs in the environment has been intensively researched since the early 2000s, but MP concentrations reported by different studies often deviate by several orders of magnitude, even for the same environmental compartment. Possible reasons for this discrepancy include (i) a high spatial and temporal variability of MP concentrations in the different environmental compartments and (ii) the use of non-uniform methods for sampling, sample processing, and analysis by different studies, as well as insufficient quality assurance and quality control (QA/QC). This thesis focuses on investigating the temporal variability and representativeness of MP concentration and composition in streams and wastewater treatment plant (WWTP) effluents. The temporal variability of the MP concentration and composition in streams is of great importance in many respects, as streams are the habitat of freshwater biota, a source of drinking water for humans and wildlife, and a transport pathway for MPs into the oceans. In turn, WWTP effluents are a continuous, localized input pathway for MPs into surface waters, and the temporal variability of the MP concentration and composition in WWTP effluents is likely to influence that of receiving water bodies. To obtain not only temporally representative but also reliable MP data, QA/QC, including the avoidance of false positive MP identification, are a fundamental part of this research. To gain reliable results, MP losses, cross-contamination, and false positive MP findings must be minimized and quantified as far as possible. Two hitherto mostly unknown sources of false positive MP findings are disposable gloves and sodium dodecyl sulfate (SDS). Disposable gloves are commonly worn during sample processing to avoid cross-contamination or for personal protection. SDS is often applied during sample processing to aid the decomposition of animal and plant residues. It was found that SDS and the stearates released from disposable gloves can be falsely identified as polyethylene (PE) when analyzed using Raman microspectroscopy (µ-Raman), Fourier transform infrared microspectroscopy (µ-FTIR), and pyrolysis-gas chromatography/mass spectrometry (pyr-GC/MS). The extent of the potential PE overestimation due to the use of disposable gloves was experimentally determined for all three analytical methods and their suitability to distinguish between SDS, PE, and stearates was investigated and assessed. Measures to avoid false positives were derived and applied to all subsequent laboratory and field experiments carried out as part of this thesis. The temporal variability of MP concentration and composition in aquatic matrices has only been researched to a limited extent to date. In particular, MP studies that cover long investigation periods and collect temporally representative mixed samples are rare. This is partly due to a lack of sampling devices specifically designed for long-term sampling of MPs. Therefore, a custom-built, automatic sampling unit for MPs (SAM) was developed within this thesis, which is capable of fully automated, repetitive collection of volume-reduced mixed samples from freshwater matrices over long periods of time. A recovery experiment (positive control) using five different polymer types revealed an overall recovery of 77 ± 29% for sampling with SAM, subsequent sample processing, and µ-FTIR analysis for particles ≥ 11 µm. A low level of cross-contamination was verified based on process blank values, which recorded cross-contamination over the entire workflow from sampling preparation to sample analysis. These process blank values resulted in low, polymer-specific detection limits, demonstrating a high sensitivity of the overall method. The good suitability of SAM for long-term sampling was proven by two field applications. SAM was applied to collect 24-hour mixed samples over a period of eight days in a stream and over a period of 31 days in a WWTP effluent. By analyzing the corresponding mixed samples with µ-FTIR, the concentration and composition of MPs ≥ 11 µm was determined for each sampling day. In both the stream and the WWTP effluent, high day-to-day variability of MP concentration was observed. In the stream, no clear relationship could be identified between the time series of MP concentration and the daily precipitation in the catchment area, the discharge of the stream, the turbidity, or the total particle count in the stream. Similarly, for the WWTP effluent, no correlation could be established between the MP concentration and the discharge from the WWTP effluent, or the daily precipitation sum. However, a high positive correlation was found between the time series of the concentration of many relevant polymer types in the WWTP effluent. Furthermore, a connection was observed between the temporal proximity of sampling days and the similarity in MP concentration and composition of the corresponding samples. The temporal variability of the MP composition by shape, size, and polymer type was much more consistent over the study periods in both the stream and the WWTP effluent than the variability of the MP concentration. Thus, it was concluded that repetitive or continuous collection of 24-hour mixed samples is required to obtain MP concentration data that is representative over long periods of time, while individual mixed samples may be sufficient for making a rudimentary estimate of the MP composition in streams and WWTP effluents. This thesis has revealed two hitherto mostly unknown causes for overestimation of MP concentrations and has derived effective countermeasures. In addition, an automated device for the repetitive collection of mixed samples over long time periods was developed and validated. Application of this device for sampling from a stream and a WWTP effluent yielded reliable data on the temporal variability of MP concentration and composition and enabled a first assessment of potential variability influencing factors. This work therefore represents a first step towards the collection of reliable, temporally representative MP data from streams and WWTP effluents, and serves as a basis for further research on the temporal variability of MP concentration and composition in water matrices.…