Study of model contaminants desorption from recycled PET in dry air and nitrogen atmosphere by thermogravimetric analysis
A.S.F. Santos, J.A.M. Agnelli, E.S. Medeiros, E. Corradini, S. Manrich
INSTITUTE FOR TECHNOLOGICAL RESEARCH, BR
RECYCLING, SUPER-CLEAN, DESORPTION, TGA, DECONTAMINATION
The polyethylene terephthalate (PET) bottle-to-bottle recycling is a good example of a successful technology that has allowed an increase of recycled plastic market. Nevertheless, the capability of all these processes to decontaminate recycled plastics into a level that offers a negligible risk to public health and does not compromise the organoleptic properties of the packed foods has to be proved in order for recycled plastics to be used in direct contact with food. According to our previous work, which resulted in new bottle-to-bottle process that employs dry air atmosphere under ideal mass and heat transfer conditions to remove contaminants from polymer matrix, dry air is able to increase the productivity of PET super-clean technologies, based on solid-state polymerization processes (SSP). Aiming to evaluate the difference in desorption rate of model contaminants from recycled PET in dry air and nitrogen atmosphere, thermogravimetric analysis were accomplished. Furthermore, it was investigated if sample exposed to synthetic air and nitrogen atmosphere during the thermogravimetric analysis presented any difference in their thermal properties. A mixture of surrogates (10 vol.% of toluene, 1 vol.% of trichloroetane, 1 vol.% benzophenone and 1 vol.% of eicosan) in hexane, as suggested by FDA guidelines, were used to spike PET squares. Then, the contaminated PET squares were isothermally heated at 60ºC in a thermogravimetric analyzer to follow the mass loss for two hours under synthetic air and nitrogen atmosphere. The tests were carried out in duplicate and each sample of PET were afterwards analyzed by differential scanning calorimetry (DSC). According to thermogravimetric results shown in Figure 1, the desorption rate of contaminants in synthetic air atmosphere was slightly higher than in nitrogen. Nevertheless, considering the mean deviation of these results, they seemed similar. Probably, the sensitivity of this analysis for the range of mass loss evaluated was not enough to detect any difference between the rate of contaminants desorption from PET in these two atmospheres. In this way, those results could not confirm our previous study, which was carried out using chromatographic analysis and observed an increase of about 60% in diffusion coeficient of benzophenone in dry air atmosphere when compared with that one in nitrogen atmosphere. Additionally, the DSC results did not show any significant difference in thermal properties of analyzed samples (Table 1). Once this mean that there is no specific interaction between dry air and PET matrix, other factors that could probably collaborate with our previous work are the possible improved co-diffusion mechanism between contaminants and oxygen from dry air atmosphere through the polymer matrix, since nitrogen atmosphere is an inert gas of lower molar mass; the reduced water content of dry air; and others. Finally, the results achieved in this work indicated that thermogravimetric analysis was not sensitive to detect difference in the rate of contaminant desorption from PET in synthetic air and in nitrogen atmosphere, considering the level of contaminant content evaluated. In the same way, no difference in thermal properties, i.e. of interaction between PET matrix and exposed atmosphere, have been observed by DSC analysis.
Back to Program