Untargeted approaches based on GC-Orbitrap-HRMS and two-dimensional GC–MS for the identification of intentionally and non-intentionally added substances from bio-based food contact materials

The increasing adoption of bioplastics is rapidly expanding in food contact materials (FCMs) to minimise environmental impact and improve consumers’ acceptance. However, several aspects must be considered throughout their lifecycle, from bioplastic processing to final product application. This study presents an untargeted analytical workflow to characterise volatile compounds of bioplastic FCMs based on GC–Orbitrap HRMS coupled with chemometrics and complemented by GC × GC-TOF–MS. The bio-based FCMs included polylactic acid (PLA), its heat-resistant variant crystallised PLA (cPLA) and no-PLA materials. The approach based on GC-Orbitrap HRMS coupled to supervised PLS-DA and variable importance in projection (VIP) method allowed the identification of differential compounds able to discriminate samples according to material composition and maximum end-use temperature. Chemometric analysis applied to samples classified according to maximum end-use temperature identified cyclic oligomers and hydrocarbons as markers for PLA samples suitable for use up to 40°C. Considering classification by the type of material resistant to moderate temperature (cPLA vs. no-PLA FCM), analysis of the dataset showed almost completely different VIP selection, with some PLA-related molecules (lactic acid, lactate esters, and linear oligomers) characterising cPLA, some hydrocarbons as key variables of the non-PLA class, and two succinate esters present in both the materials suggesting the presence of PLA/polybutylene succinate blends. In addition to polymer-related molecules, plasticisers emerged as chemical additives incorporated into the investigated materials during manufacturing. In addition, GC × GC-TOF–MS successfully identified distinct series of PLA oligomers, citric acid esters, and mineral oil saturated hydrocarbons, with regular chromatographic patterns, supporting structural hypotheses and integrating GC–Orbitrap HRMS findings.