Chemometric analysis of polymer spectra for degradation assessment and enhanced identification of meso and microplastics

Abstract

Marine environmental conditions and the weathering undergone by plastic samples, turn plastic identification into a challenge. Several analytical methodologies, such as Fourier transform infrared spectroscopy (FTIR), Raman spectroscopy and pyrolysis-gas chromatography-mass spectrometry (Py-GC-MS) have been employed for this purpose. However, these techniques remain insufficient when degradation is highly advanced. This study specifically focuses on micro and mesoplastics, analyzing the most abundant polymers in the marine debris: polyethylene (PE) and polypropylene (PP). To improve identification under conditions of advanced degradation, analytical tools for post-processing were developed. To this end, the calculation of the carbonyl index was simplified and automated, given its role as an indicator of weathering. Furthermore, its relationship with other parameters associated with degradation and plastic identification, such as the yellowness index and the coincidence match, was thoroughly evaluated. Classification algorithms, specifically Random Forest and Support Vector Machine, were implemented to enhance the discrimination of plastic types, while spectral analysis focusing on characteristic peak features was conducted to improve the accuracy and reliability of polymer identification.