Chemical Additive Fingerprinting of Microplastics for Source Tracking: The Effect of Weathering and Sample Processing

The use of chemical fingerprinting of microplastics (MPs) was evaluated to support source identification by (1) identifying treatment-driven changes in MP chemical fingerprints caused by artificial weathering and sample processing, and (2) testing the identification of MPs within MP mixtures to mimic environmental samples. A nontarget chemical fingerprint library was constructed from 15 pristine MPs using High-Performance Liquid Chromatography coupled with Quantitative Time-of-Flight Mass Spectrometry, and a Random Forest classifier was trained to classify treated MP counterparts. Results indicated that chemical fingerprints comprised both polymerization-based and functional additive-derived chemicals. The classifier correctly assigned all pristine MPs and retained 90% accuracy after artificial weathering and sample processing, despite treatment-driven changes in features that shifted fingerprint composition. Suspect screening analysis and Random Forest classifications showed that oxidative digestion had the greatest impact on MP fingerprints and identification, decreasing prediction probabilities by an average of 30% and altering MP fingerprint composition. In mixed-MP tests, MPs with unique fingerprints remained detectable, such as tire rubber, supporting targeted applications of the method. Overall, this work serves as a proof-of-concept of a novel fingerprinting method based on MP identification using plastic additive fingerprints, which could be further developed to identify MP sources, such as local polluters.