Portable Pyrolysis-Point Discharge Optical Spectrometer for In Situ Plastic Polymer Identification by Coupling with Machine Learning

Rapid and in situ identification of specific polymers is a challenging and crucial step in plastic recycling. However, conventional techniques continue to exhibit significant limitations in the rapid and field classification of plastic products, especially with the wide range of commercially available color polymers because of their large size, high energy consumption, and slow and complicated analysis procedures. In this work, a simple analytical system integrating a miniature and low power consumption (22.3 W) pyrolyzer (Pyr) and a low temperature, atmospheric pressure point discharge optical emission spectrometer (mu PD-OES) was fabricated for rapidly identifying polymer types. Plastic debris is decomposed in the portable pyrolyzer to yield volatile products, which are then swept into the mu PD-OES instrument for monitoring the optical emission patterns of the thermal pyrolysis products. With machine learning, five extensively used raw polymers and their consumer plastics were classified with an accuracy of >= 97.8%. Furthermore, the proposed method was applied to the identification of the aged polymers and plastic samples collected from a garbage recycling station, indicating its great potential for identification of environmentally weathered plastics. This portable Pyr-mu PD-OES system provides a cost-effective tool for rapid and field identification of polymer types of recycled plastic for proper management and resource recycling.