Quantification of small (1-10 µm) microplastic particles in soil matrices using automated scanning electron...

Scanning electron microscopy (SEM) provides superb resolution power and allows distinguishing microplastic particles (MPs) from other particles based on elemental ratios derived from energy dispersive x-ray (EDX) analysis. Polycarbonate (PC) membranes are typically used to isolate MPs from suspensions for SEM investigations. This, however, results in poor (atomic weight) contrast between MPs and the underlying PC membrane, challenging automated particle detection. To introduce an elemental contrast between MPs and the PC membrane, a gold (Au) coating was applied to the PC membrane before use. Monte Carlo simulations conducted to optimize operational conditions of the SEM suggest an acceleration voltage of 3 kV in combination with a 40 nm Au coating of PC membranes as most appropriate. Stock suspensions made from polyethylene (PE) fragments, polyvinyl chloride (PVC) fragments, and standard soils, all sieved to 0.45-10 µm, were mixed at different ratios. Selected mixtures were density separated to isolate the MPs from the mixtures. Whereas up to 7% of PE was detected in PVC stocks, likely representing false positives, only negligible amounts of PVC were detected in PE stocks. In soils, neither PE nor PVC was detected in significant amounts. Polyethylene and PVC particles diluted in soil suspensions at particle number ratios of 1:1:2 and 1:1:2000 were density separated, and largely concentration independent recoveries of around 70% and 50% were obtained for PE and PVC, respectively. Good agreement between SEM-EDX and Raman measurements was obtained for MPs, underlining the strength of our approach to address the smallest size fraction of MPs in complex matrices.