Miscounting microplastics as a result of matrix molecules: optimizing identification and quantification of microplastics in natural freshwater systems

Microplastics have become a major focus of research in environmental chemistry; over the past decade, microplastic sampling has shifted from marine and large lake settings to areas such as wastewater treatment plants, inland lakes, and rivers, which present various environmental sampling challenges due to more enriched and complex sample matrices. This study highlights the various methods that are used in microplastic sample processing, tests a new methodological approach to address the matrices found in a series of inland lakes in Minnesota, United States, and investigates the FTIR spectra and potential library misidentification of natural matrix materials, such as plant waxes and bird feathers that sometimes survive oxidative treatments. Specifically, the new methodological approach incorporates the use of ethanol to address lipid-rich organic materials and an additional filtering step following Fenton oxidation to separate the retained plastics from smaller clay matrix materials and avoids the use of enzyme digestion. This method was validated through a series of particle recovery tests and method blanks, which highlighted minimal to no particle loss and small amounts of fragmentation/discoloration. Further testing confirmed that the method exhibits low to negligible contamination levels. We determined that plant wax materials are not removed in sample processing and are often misidentified as plastic polymers within library searches. However, we identified distinguishable, albeit small, spectral differences from plastics, which may require adjusting libraries and adding to training sets for AI identification of plastics. Overall, this study has resulted in an adaptable sample processing workflow to ensure robust and accurate assessment of microplastic contamination in a variety of environmental compartments.