Development of a fluorescence-based optical sensor for detection and identification of microplastics in water samples

Microplastics (MPs) with a size of less than 5 mm are emerging contaminants that are ubiquitous in our daily lives and present still not clear, yet potentially high risks for the environment and human health. Despite a great research effort, the methods for quantitative identification still mainly rely on long and subjective visual counting procedures carried out on optical microscopes by skilled human operators. This study presents a new automatic, portable, low-cost, and fast method for the quantitative detection of MPs in water. The proposed method automatically processes and counts the fluorescence pulses emitted by dye-stained MPs in flowing liquids after excitation with a low-power laser beam. Nile Red (NR) dye was used for the quantification of MPs in commercial bottled water and tap water samples after staining parameter optimization for higher fluorescence signal-to-noise level, the result is in good agreement with microscope observations. In addition, for further studies on the identification and classification of MPs with polymerselective fluorescent dyes, solvatochromic dye DANS was investigated and optimized with fluorescence microscopy and spectroscopy. Image analysis with machine learning of fluorescent channel images proved the ability of DANS to selectively classify polymer types. The present investigation demonstrated a methodology for quick automated counting and classification of MPs with fluorescent dyes in water through a cost-effective and efficient approach.