FPA-based Micro-FTIR Analysis of Pre-extracted Particles in Sediment Samples
Between 29 and 64 particles per sample were analyzed by FTIR microscopy resulting in a total number of 404 particles analyzed from the ten samples provided by the NLWKN. Only 16 particles (12 fragments and 4 fibres), i.e. 4 % of the total number of investigated particles, differed in their shape from the abundant granular material.
With 388 counts the majority (96 %) of the particles could clearly be assigned to the category of granular particles. Most of the granular particles were transparent to whitish-opaque and between 100 and 300 µm in size (Fig. 8.2). Of these, 320 particles showed a spectrum that strongly differed from common polymer IR spectra. By comparison with the IR spectrum of laboratory quartz (p.a. grade, Merck) we were able to show, that all these particles were in fact quartz sand and not microplastics (Fig. 8.3). 68 further granular particles that were very similar in shape and appearance to the quartz particles also displayed non-polymer spectra. These spectra were similar to the quartz spectra (Fig. 8.4) and were thus certainly also of mineral origin. However, because of the lack of reference spectra a definite identification was not possible in those cases.
Fig. 8.3 Comparison of the spectrum of laboratory quartz (p.a. grade Merck) (blue) and a spectrum obtained from the measurement of a typical granular particle (red) by FPA-based microFTIR spectroscopy
Fig. 8.4 Four examples of IR spectra of granular particles that were very similar in shape and appearance to quartz particles but displayed different non-polymer spectra when analyzed by FPA-based micro-FTIR spectroscopy
18.104.22.168 Fragments and Fibres
Within the 16 particles that differed from the granular material, we found four black malleable fragments with a size between 260 and 390 µm, which exhibited high absorbance and showed non-interpretable transmission IR spectra (Fig. 8.5). According to their bitumen-like, malleable properties and coloration, those particles are most likely oil residues from ship spillages or road wear. In total, six particles were of biological origin. Four particles of fibrous appearance (length: 290–900 µm, width 16–100 µm) showed organic IR spectra similar to cellulose and were most likely of plant origin. Two particles (200–800 µm) exhibited chitinlike IR spectra and were thus assigned to particles of animal origin.
Within the differentially shaped particles, only six fragments (i.e. 1.4 %) of the analyzed particles, displayed a plastic polymer spectrum. Three fragment-shaped particles in the size between 440–1200 µm were polystyrene and another three fragments between 180–450 µm were polyethylene (Fig. 8.6). With the help of chemical imaging the localization of characteristic spectra within the measured area could be visualized accurately and the microplastic particles could be identified (Fig. 8.7).
As a consequence of the applied sample extraction and purification techniques, it was assumed that the provided samples should have comprised mainly microplastic fragments and other organic material of low density. However, plastic fragments were exceptionally rare as they made up only 1.4 % of all analyzed particles. Since we manually screened the bright-field microscopic images for areas with heterogeneous particle appearance and measured at these locations, it can be assumed, that in total, the abundance of microplastics in the samples was
Fig. 8.5 IR spectra of bitumen-like black particles that showed high absorbance when measured by FPA-based micro-FTIR spectroscopy
Fig. 8.6 Examples of the IR spectra of the six microplastic particles found in the sediment samples, which were measured by FPA-based micro-FTIR spectroscopy: upper panel polystyrene, lower panel polyethylene (FPA-measured spectra in red, ATR reference spectra in blue)
even lower. Surprisingly, the majority of particles could be clearly identified as quartz particles, although these should have been excluded via the density separation step. While it remains to be clarified why the samples contained such large amounts of sand grains, the case study proves the absolute necessity of spectroscopic (e.g. FTIR) measurements for microplastics analysis. In this context FPAbased micro-FTIR spectroscopy proved to be a very promising technique for verifying polymer origin of microplastic particles and thus should be mandatorily included in future monitoring programs on microplastics after SOPs for the purification and measurement approaches have been established.
Fig. 8.7 Chemical imaging of an integration of typical spectral bands of a quartz (1931– 1832 cm−1), b polyethylene (2981–2780 cm−1) in the sample Wu_Ku_Eu 4 analyzed by FPAbased micro-FTIR spectroscopy. The color scale represents the intensity of the chosen spectral bands according to peak area (bluish color scale intensity of the peak 1,931–1,832 cm−1; reddish color scale intensity of the peak 2,981–2,780 cm−1) and clearly shows quartz particles in light blue (a) and a polyethylene particle in orange (b). Scalebars: 500 µm