Chemie und Prozesstechnik
Filtern
Erscheinungsjahr
- 2019 (2) (entfernen)
Dokumenttyp
- Vortrag (1)
- Posterpräsentation (1)
Sprache
- Englisch (2)
Referierte Publikation
- nein (2)
Schlagworte
- Principal component analysis (PCA) (2) (entfernen)
Organisationseinheit der BAM
Eingeladener Vortrag
- nein (1)
This talk was given within the scope of the SIMS-22 conference in October 2019 in Kyoto (Japan). It deals with the surface analytical investigation of nanoparticles by PCS-assisted ToF-SIMS. This technique is applicable to core-shell nanoparticles, in order to distinguish a complete encapsulation from an incomplete encapsulation of the core by the shell material. Furthermore, the depletion process of organic nanoparticle coatings caused by UV-weathering is investigated. Finally, the significance of grouping studies for nanomaterials research and risk assessment is demonstrated.
Working towards a comprehensive understanding of introduction pathways, number, and fate of micro¬plastics in the environment, suitable analytical methods are a precondition. Micro-spectroscopic methods are probably the most widely used techniques. Besides their ability to measure single spectra of a particle or fiber, most modern FTIR- and Raman microscopes are also capable of two-dimensional imaging. This is very appealing to microplastics research because it allows to simultaneously characterize the analytes chemically as well as their size (distribution) and shape.
Two-dimensional imaging on extensive sample areas with FTIR-micros¬copes is facilitated by focal plane array (FPA) detectors resulting in large data sets comprised of up to several million spectra. With numbers too large for manual inspection of each individual spectrum, automated data evaluation is inevitable. Identifying different polymers based on the comparison with known reference spectra (library search) has proven to be a suitable approach. For that purpose, FTIR-spectra of common plastics can be collected to create an individual reference library.
To Supplement this ‘targeted analysis’, looking for known substances via library search, an exploratory approach was tested. Principal component analysis (PCA) proved to be a helpful tool to drastically reduce the size of the data set while maintaining the significant information. Subsequently, cluster analysis was used to find groups of similar spectra. Spectra found in different clusters could be assigned to different polymer types. The variation observed within clusters gives a hint on chemical variability of microplastics of the same polymer found in the sample. Spectra labeled according to the respective cluster/polymer type were used to build a classification model which allowed to quickly predict the polymer type based on the FTIR spectrum. Classification was tested on a second, independent data set and results were compared to the spectral library search procedure.