Revisiting microplastic pollution: A novel method for detecting small-sized microplastics in natural waters

Abstract

Abstract Microplastics are widespread, yet methods to measure polymer type simultaneously and particles size smaller than 20 µm are limited, hindering toxicity assessment and policy formulations. We introduce a machine learning-assisted spectral flow cytometry approach that identifies six major polymers relevant to the environment and human health within the 5–100 µm range and allows projection in the 1–100 µm range. High-throughput, sensitive detection, straightforward sampling, and strong quality control make this technique suitable for routine monitoring of industrial and natural waters. Clear differences in microplastics levels and polymer composition were found between Lake Geneva, nearby rivers, and surface versus deep waters. With up to 97% of microplastics between 1 and 20 µm, data indicates a substantial underestimation of pollution. Consistent with laser-infrared measurements, concentrations of small-sized microplastics exceed previous reports in lake waters up to 656-fold. Environmental risk assessment of microplastics suggests that risk might be expected at some sampling sites.