Analytical testing strategy for identification and in silico toxicity assessment of non-intentionally added substances in repeatedly recycled flexible mono-plastic food contact material

Abstract

Polyethylene terephthalate (PET) is a polymer approved for recycling for food grade applications in the EU. Recycling PET for food grade applications is desirable from sustainability perspective; however, its safety assessment is a challenge due to the potential presence of different types of non-intentionally added substances (NIAS). Identification of NIAS is important for risk assessment and devising appropriate exposure limits. An analytical testing strategy based on high-resolution LC-MS has been presented and applied to food contact materials of virgin PET (vPET) and repeatedly recycled PET (rPET) produced using Starlinger Decon technology. We devised an analytical strategy using data-dependent acquisition technique and used InChI strings to identify the peaks with a high level of confidence. We identified 23 peaks in the samples by using the developed analytical testing strategy that were largely oligomers of different series and lengths, acids, and cyclic monomer. We semi-quantified the peaks and did not notice any statistically significant difference between vPET and repeatedly recycled PET samples. The identified substances were screened for potential Carcinogenicity, Mutagenicity and Reproductive toxicity (CMR); and for endocrine potential models of action represented as Estrogen, Androgen, Thyroid, Steroidogenic – and Retinoic acid (EATS-R) related modalities using the Danish quantitative structure activity relationship (QSAR) Database. The QSAR screening showed predicted flags for CMR and EATS-R for some of the substances. As the experimental information on the toxicity of PET oligomers is very limited, there is a need for in vitro testing, followed up with further testing depending on the findings, to perform hazard characterization.