Advanced High-Resolution LC/MS-Based Characterization of Extractables from Food Contact Materials Enabled by a Newly Developed Chemical Database

Importance of the topic

The safety of food is directly affected by chemicals migrating from polymer-based food contact materials (FCMs) into food during production, storage, processing and use. Comprehensive, reliable identification of extractable and leachable (E&L) chemicals across a wide chemical space is therefore critical for risk assessment, regulatory compliance and formulation development. High-resolution LC/MS complemented by application-specific spectral libraries and dedicated data workflows improves confidence in identifying semi-volatile, polar, and high-molecular-weight FCM-related compounds that are not amenable to GC/EI approaches.

Objectives and overview of the study

This study demonstrates an end-to-end high-resolution LC/MS workflow to characterize ethanol extracts from commercial vacuum-seal food storage bags. Key aims were to: (1) develop an FCM-focused MS/MS database and library to improve compound identification, (2) optimize LC/Q-TOF acquisition and chromatographic separation for structural isomers and higher-molecular-weight oligomers, and (3) show practical identification of known additives, degradation products and cyclic oligomers across multiple suppliers and extraction conditions.

Methodology

Samples: Ethanol extracts were prepared from three commercially sourced vacuum-seal bags (samples A, B and C). Pieces of each bag were heated at 70 °C for 10, 20 and 60 minutes; extracts were analyzed undiluted and after 10- and 100-fold dilution. An ethanol blank was included to account for mobile-phase and system impurities.

Acquisition strategy: Iterative data-dependent acquisition (DDA) was used, with each extract injected three times. Iterative DDA built exclusion lists to prioritize MS/MS acquisition for lower-abundance species. MS/MS data were collected at three collision energies (10, 20 and 40 eV) to capture complementary fragmentation patterns useful for structural elucidation.

Data processing: Non-targeted data processing was performed using Agilent MassHunter Explorer 2.0 to extract RT-aligned high-resolution MS and MS/MS spectra. Extracted spectra were searched against a custom FCM library and, when needed, exported to NIST MS Search and SIRIUS for additional interpretive support.

Used instrumentation

The main analytical platform and parameters (summarized):

• LC: Agilent 1290 Infinity II with a Poroshell Aq-C18 analytical column (2.1 x 150 mm, 2.7 µm) and Poroshell EC-C18 delay column (4.6 x 50 mm, 2.7 µm); column temperature 40 °C; injection volume 5 µL; flow 0.35 mL/min; reverse-phase gradient ramping from 2% to 100% organic and returning to 2% after 26.1 min.
• MS: Agilent Revident LC/Q-TOF with Dual AJS ESI operating in positive polarity; MS range 40–1700 m/z; acquisition rates ~4 spectra/s (MS) and ~6 spectra/s (MS/MS); collision energies 10/20/40 eV; source gases and voltages optimized to minimize in-source fragmentation (drying gas ~250 °C, sheath gas ~300 °C, nebulizer ~35 psi, capillary ~4000 V).
• Software: MassHunter Explorer 2.0 for non-targeted workflows; Agilent ChemVista for library/database management; SIRIUS (Bright Giant) and NIST MS Search used for structure proposal and additional spectral matching.

FCM database and library construction

An application-specific FCM database was assembled by merging curated lists from Food Packaging Forum resources (FCCdb, FCCprio, FCCmigex totaling >18,000 entries combined) and >50 commercially available polymer additives plus two AChemTek standards kits. Public MS/MS spectra from MassBank (MONA and MassBank EU) and in-house reference standards were incorporated. The result was a downloadable custom FCM library of more than 3,000 compounds with associated MS/MS spectra, exportable in formats compatible with NIST and SIRIUS workflows.

Main results and discussion

- Identification scope: The workflow successfully identified a range of additive-related species, degradation products and previously unannotated cyclic oligomers across the three bag samples.

- Sample-specific composition: Using volcano plots and a Unique Features analysis, the study found 104 features unique to sample C, 17 unique to A, and 11 unique to B, demonstrating supplier-dependent chemical profiles.

- Key chemical classes: Sample A was enriched in cyclic nylons and amides; Sample B showed breakdown products associated with the antioxidant Irganox 1076; Sample C contained amides and a homologous series of cyclic esters/oligomers (formulas ranging approximately C19H30O8 to C24H24O10) eluting between ~12.3–14.5 min with highly similar MS/MS spectra.

- Structural inference: High-resolution MS and MS/MS (multi-energy) in combination with SIRIUS and the custom FCM library permitted formula confirmation and plausible structure proposals for cyclic oligomers and related modifications (e.g., observed species consistent with -CH2 or -COH2 modifications of a C20H32O9 core). The iterative DDA approach increased MS/MS coverage of low-abundance species useful for distinguishing isomeric series.

- Analytical considerations: The LC gradient and stationary phase selection improved separation of structural isomers and retained hydrophobic oligomers, while optimized source conditions reduced in-source fragmentation—both essential for reliable spectral matching and structure elucidation in LC/MS-based E&L work.

Benefits and practical applications

• Improved confidence in non-targeted LC/MS E&L identification through an application-specific MS/MS library focused on FCM-relevant chemistry.
• Enhanced detection of low-abundance and high-molecular-weight migration products via iterative DDA and optimized chromatographic separation.
• Interoperability with established tools (NIST MS Search, SIRIUS) allows complementary spectral matching and in silico structure prediction for knowns and unknowns.
• Practical relevance for manufacturers, QA/QC labs and regulatory assessment teams evaluating migration profiles from polymeric food contact articles.

Future trends and potential uses

• Expansion and community curation of FCM-specific MS/MS libraries to cover more formulation additives, processing aids and oligomer families, improving identification rates.
• Integration of machine-learning spectral prediction and retention-time prediction to prioritize candidate structures for unknowns when spectral library matches are absent.
• Hybrid analytical strategies combining GC/EI libraries for volatile/semi-volatile compounds with LC/HRMS workflows for polar and high-mass migrants to achieve comprehensive E&L screening.
• Automation and higher-throughput extraction and data processing pipelines to support routine screening in compliance testing and supplier surveillance.
• Tighter linkage of annotated databases to regulatory lists and toxicological metadata to streamline risk-assessment workflows.

Conclusion

This work demonstrates that high-resolution LC/Q-TOF with iterative DDA, when combined with a purpose-built FCM MS/MS library and integrated data-processing tools, significantly improves the ability to detect and identify diverse extractables from polymer food-contact materials. The approach enables discrimination of supplier-specific chemistries, characterization of cyclic oligomers and identification of additive breakdown products—capabilities that strengthen analytical support for safety assessment and material selection.

Reference

1. Food Packaging Forum Foundation. FCCmigex database (detailed information on >4,000 FCMs and >24,000 database entries). Cited in study materials.
2. Food Packaging Forum Foundation. FCCdb — a database of intentionally used food contact chemicals; FCCprio — Food Contact Priority List (high-risk compounds). Cited in study materials.