Beyond the known: Integrated non-target screening and in silico toxicity prediction of substances migrating from plastic and paper-based food contact materials

This study presents a comprehensive non-targeted screening workflow for the identification and prioritization of chemicals migrating from food contact materials (FCMs). In sixteen commercially sourced materials - representing paper, plastic, and multilayer paper/plastic materials - migration tests were conducted under standardized conditions (EU 10/2011) with simulants A, B C and D1 and analyzed by a UHPLC-Orbitrap high-resolution mass spectrometer (HRMS) system. The resulting data were processed through a multi-layered, customized Python-based pipeline that combined a curated database, and advanced spectral matching using in silico MS/MS predictions. Confidence levels were refined through retention time prediction, isotopic pattern verification, and cosine similarity scoring. A total of 114 migration compounds were identified, 4 were confirmed using reference standards, while 21 reached Level 2a based on high-confidence spectral annotation. Plasticizers and non-intentionally added substances (NIAS) represented the most frequently detected categories, underscoring both their widespread use and relevance to safety evaluation. Dual toxicity assessment was employed using Cramer classification and LC₅₀ prediction, capturing both structural alerts and de novo toxicity insights. Migration profiles across material types highlighted class-specific markers, with 17 compounds uniquely present in paper-based FCMs and 6 in plastics. Three compounds-phytosphingosine, hexadecenylsuccinic acid, and (9S)-9-hydroxyoctadecadienoic acid-were consistently found in all paper-based materials, while N,N-dimethyloctadeca-9,12-dienamide and 4-methylbenzophenone were frequent in plastics. Stabilizer tris(2,4-di-tert-butylphenyl) phosphate, was detected in 100% of the samples. Overall, this work demonstrates the utility of integrating advanced analytical, cheminformatics, and predictive toxicology tools for profiling FCMs and contributes a scalable framework for future safety assessments and regulatory prioritization.