Validation of near infrared spectroscopy as an age-prediction method for plastics

This work has two aims. Firstly, to validate the ability of experimental models derived through near infrared spectroscopy for acrylonitrile butadiene styrene (ABS), low-density polyethylene (LDPE), polyethylene terephthalate (PET), and polypropylene (PP) in predicting polymers' aging; focusing on the degree of oxidation. Secondly, to assess the reliability of non-invasive age-predictive models on waste plastic samples and on mechanically recycled samples. Aging time, temperature and number of extrusion cycles were selected as independent variables to build the aging-prediction models, where they were calibrated on samples subjected to controlled conditions. The accuracy of the prediction models was assessed on external samples (aged under known conditions) through the cross correlation technique and the Root Mean Square Error (RMSE). The models exhibited good collinearity for the aging temperature and the number of extrusion cycles factors for all tested polymers, but not for the aging time factor. The RSME value of the aging time factor was far from zero for all polymers. Plastic waste samples provided analogous results; the aging time estimation was mostly negative in value. The estimations of aging time and number of extrusion cycles were always positive in values, where the most reasonable aging factor estimation was the number of extrusion cycles.