Development of new spectroscopic and multivariate chemometric methods for the characterization of microplastics in the marine environment

Microplastics are ubiquitous and gain increasing attention in research as well as media and politics worldwide. In many cases, the presence of microplastics in different environmental compartments was reported and the risk potential was discussed. However, there is a lack of standardization starting with a clear definition of microplastics up to suitable analytical methods for their detection and characterization. Consequently, this work reviewed the literature to determine the status quo of microplastics analysis, and developed new suitable analytical and chemometric methods for microplastics identification and characterization. Based on the studies reviewed, Fourier Transform Infrared Spectroscopy (FTIR) was confirmed as the most used method to identify microplastics. However, it could be obtained that the documentation of the microplastics identification process itself was mostly underrepresented, which indicated a lack of awareness of this important aspect. Therefore, a practical guide was developed that addressed data evaluation for spectroscopic data of microplastics. Moreover, a critical examination of the principles of infrared spectroscopy with special regard to the analysis of microplastics was presented. In its core, this work focused on a new chemometric concept to identify microplastics based on FTIR spectroscopy, which is called µIDENT. The developed algorithm automatically extracts accurate vibrational band lists of individual spectra using a very robust and fast multi component curve fitting approach. In a second step, every list is transformed into a peak intensity ratio pattern. This is highly robust and characteristic for each polymer, and is the basis for reference pattern library search to identify microplastics. Furthermore, a rapid and intelligent method for chemical imaging using µFTIR (µMAP) was developed to challenge the problem that those kinds of measurements are very time consuming. Therefore, the µFTIR scans a defined area step-by-step, and only stops to start a detailed infrared measurement, if microplastics are at the current spot. This significantly reduces the number of measurements, which saves up to 92 % of the total measurement time. The method development is concluded by presenting an innovative concept for separation of microplastics from sediment samples for analytical purposes (µSEP). In this context, microplastics adhere to fine air bubbles and are transported onto a filter, while matrix components remain in the closed loop like approach. In contrast to other separation concepts, the presented prototype requires only water and air, which reduces costs and lowers the risk of losing microplastics due to aggressive substances. To all method developments made, there were similar contributions by other research groups. However, this work especially addressed highly practical, automatable and robust concepts. The focus was always on the application of the methods, which was underlined, for example, by publishing all source codes. In conclusion, the tools provided could improve current microplastics analysis, and could play an important role in future challenges in this area.