Unveiling microplastic spectral signatures under weathering and digestive environments through shortwave infrared hyperspectral sensing.

Microplastic (MP) pollution presents a novel challenge for marine environmental protection, necessitating comprehensive and long-term monitoring and assessment approaches. Environmental MPs can undergo weathering and microorganism-related digestive processes, altering their original surface properties and chemical structure, thus complicating their quantification and identification. This study aims to establish a comprehensive hyperspectral database for weathered and digestion-degraded MPs, using a wide variety of polymer types collected as either virgin particles or commercial products (within a size range of approximately 3 mm), and to investigate the impact of these processes on their spectral characteristics. Polypropylene (PP) and polyethylene (PE) MPs exhibited significant responses to weathering treatment, as indicated by the formation of new characteristic peaks or slight peak shifts around 1679-1705 nm, which can be attributed to the formation of carbonyl and vinyl functional groups through Norrish reactions. Similarly, polyethylene terephthalate (PET), acrylonitrile butadiene styrene (ABS), and polystyrene (PS) MPs demonstrated notable degradation following digestive treatment, as evidenced by the emergence of new absorption peaks at approximately 1135-1165 nm, possibly associated with alterations involving carbonyl and vinyl functional groups. The results were further validated based on their comparable spectral characteristics of the resultant MPs to reference polymers and possible additives, considering a reasonably accurate match of approximately 80% for the studied MP samples. This study showcases the significant advantage of using shortwave infrared hyperspectral sensing for rapid identification of virgin and exposed MPs with a relatively large scan area after a simple sample preparation. This approach, combined with other complementary characterization techniques, shall provide highly throughput results for MPs identification. This research provides valuable insights into the features extracted from environmental MPs and establishes a foundation for improving their classification efficiency for environmental applications.