Description
Microplastic (MP) contamination is a critical environmental challenge with potential conse-quences on the ecosystems and for human health. For this reason, there is an urgent need for developing reliable methods enabling to monitor the presence of these particles in different environmental compartments and biota, especially in aquatic samples, because the marine en-vironment is demonstrated to be the final sink for MP contamination.
In light of this consideration, the present work proposes near infrared hyperspectral imaging (NIR-HSI) as an analytical method for the identification and characterisation of small MP, down to 50 μm, in aquatic samples.
For testing the proposed strategy, 54 cellulose filters were analysed: 27 samples were pre-pared by the filtration of the digested soft tissue of individual mussels, while 27 samples were prepared by filtering sea water samples (1 L each). Before filtration, all the 54 samples were artificially contaminated with different MP (polypropylene, polystyrene and polyamide) at 3 levels of addition – high (≅ 1.5 mg), medium (≅ 1 mg) and low (≅ 0.5 mg).
Moreover, 6 samples of sea water and digested soft tissue of mussels without any artificial contaminations were analysed, independently, for validating the analytical strategy.
NIR-HSI images were acquired by a SWIR-3 camera equipped with a 40x20 cm Lab Scanner (Specim Ltd, Finland), recording a short-wave infrared (SWIR) spectrum, in the region be-tween 1000 nm up to 2500 nm, for each pixel of the image.
On the acquired data-cubes, chemical maps of the filters were obtained by calculating the normalised difference image (NDI), a data processing strategy able to enhance spectral differ-ences between the cellulose background of the filter and the polymer of the MP. The combi-nation of NDI with the evaluation of the spectral signature allows the analyst to differentiate between the 3 MP polymers investigated; the outcomes of the present study demonstrated the reliability of NIR-HSI coupled with the NDI, as an automatable and cost and time- effective method, for the identification of MP down to 50 μm on heterogenous sample matrices.
Keywords: Near infrared hyperspectral imaging, microplastics, aquatic samples.
