Detecting microplastics faster: OIST researcher automates the analysis of spectra
A particle's spectrum reveals whether it is plastic – but analysing it is laborious. In a video from Japan's OIST, Samantha Phan presents a tool designed to standardise this analysis.
Photo: Martin Wagner et al. / Wikimedia Commons, CC BY 4.0 (cropped and resized)
Video: YouTube · Okinawa Institute of Science and Technology (OIST) · youtube.com
The Okinawa Institute of Science and Technology (OIST) in Japan presents Samantha Phan's project in a video released on 19 October. The researcher is developing a machine-learning tool that identifies microplastics from spectral data. It is based on an extensive identification system built from thousands of collected samples. Phan is supported by a Technology Pioneer Fellowship from the institute's Proof of Concept Program.
In infrared and Raman microspectroscopy, each particle produces a chemical fingerprint that shows whether it is made of plastic. The analysis is still time-consuming and requires experts; common standards are lacking. The project, called Plastitective, aims to automate identification and save time and effort.
According to the project description, the tool is aimed primarily at the seafood industry, which needs faster, standardised contamination tests. The method is also meant to work for other environmental samples. Phan hopes to help build a clearer picture of how strongly different industries are affected by microplastics.