Optimization of the post-acquisition data processing for the non-targeted screening of trace leachable residues from reusable plastic bottles by high performance liquid chromatography coupled to hybrid quadrupole time of flight mass spectrometry

Lei Tian; Lan Lin; Stéphane Bayen

doi:10.1016/j.talanta.2018.09.070

Optimization of the post-acquisition data processing for the non-targeted screening of trace leachable residues from reusable plastic bottles by high performance liquid chromatography coupled to hybrid quadrupole time of flight mass spectrometry

Talanta. 2019 Feb 1:193:70-76. doi: 10.1016/j.talanta.2018.09.070. Epub 2018 Sep 20.

Authors

Lei Tian¹, Lan Lin¹, Stéphane Bayen²

Affiliations

¹ Department of Food Science and Agricultural Chemistry, McGill University, Canada.
² Department of Food Science and Agricultural Chemistry, McGill University, Canada. Electronic address: stephane.bayen@mcgill.ca.

PMID: 30368300
DOI: 10.1016/j.talanta.2018.09.070

Abstract

Food safety regulations for food contact materials (FCM) usually rely on the assessment of chemical migration in order to reduce human exposure to chemical residues that could leach from the FCM into the food. In this field, there is a need for non-targeted analytical tools which can identify unknown or unexpected leachable residues, and therefore avoid unwanted human exposure. In this study, a method based on high performance liquid chromatography coupled to hybrid quadrupole time of flight mass spectrometry (HPLC-QTOF-MS) was developed and optimized to investigate leachable residues from 30 reusable plastic bottles. Firstly, a method was validated for the targeted analysis of the 11 bisphenol analogues. None of the bisphenols were detected in the food simulants (ethanol/water; 50:50 v/v), indicating that all tested bottles are free of BPA, and that bisphenol analogues were not applied as BPA replacement in bottle manufacture. The effect of post-acquisition data processing parameters on the feature extraction in non-targeted analysis was then systematically investigated. Several parameters significantly reduced the number of correct identifications of some target trace residues, which confirms that data post-processing has to be carefully optimized to decrease the risk of false negatives. The optimized method was effectively applied to the 30 bottle samples, and monomethyl terephthalate was identified at trace level in food simulants in contact with Tritan™ bottles (migration rate of 0.054 ± 0.002-0.53 ± 0.021 µg cm^-2 per 10 days at 40 °C). This method can therefore be applied to study the leachable residues from other FCMs and offer some novel information for human risk assessments.

Keywords: Data treatment; Food contact material; Non-targeted analysis; Omics; Plasticizer.