Streamlined static headspace GC–MS approach for detection of impurities in recycled PET

Importance of the topic

Headspace gas chromatography–mass spectrometry (HS‑GC–MS) is a rapid and minimally invasive approach to detect volatile and semi‑volatile contaminants in recycled polyethylene terephthalate (rPET). Evaluating NIAS (non‑intentionally added substances) such as acetaldehyde, benzene and d‑limonene in rPET is essential for assessing food contact safety, organoleptic quality of bottled beverages, and compliance with regulatory limits. Robust analytical workflows that combine targeted quantitation with untargeted screening support process control in recycling streams and risk assessment for recycled‑content packaging.

Goals and overview of the application note

This application note demonstrates a streamlined static headspace GC–MS workflow for qualitative and quantitative determination of acetaldehyde, benzene and d‑limonene in rPET. The work illustrates analytical performance (linearity, limits of detection, repeatability), the complementarity of simultaneous SIM and full‑scan acquisition for targeted quantitation plus untargeted identification, and typical concentration ranges observed in commercial rPET bottles used for bottled water.

Methodology

Samples: rPET bottled water containers were cut into chips (~5×5 to 8×8 mm). Aliquots of ~1.0 g were placed into 20 mL headspace vials without solvent extraction to avoid loss of volatile analytes.
Headspace sampling: Static HS incubation at elevated temperature and pressure was used to transfer volatiles to the headspace prior to injection. Key headspace operational parameters included vial incubation at 120 °C for 30 min, vial pressurization, and a dedicated loop/sample path maintained at elevated temperature to prevent analyte losses. Split injection (approx. 15:1) with an injector set to 100 °C was employed.
GC–MS separations and detection: Separation used a 60 m × 0.25 mm × 1.4 µm TraceGOLD TG‑624 SILMS capillary column (low–mid polarity, high thermal stability). Analysis cycle was ≈14 min. The mass spectrometer was operated in electron ionization (EI) mode with SMART tune and combined acquisition: full scan (approx. 10–250 m/z) together with SIM for target ions to maximize sensitivity and enable screening.
Calibration and QA: External calibration used a broad multi‑level calibration series (eleven levels) prepared in acetonitrile and spiked into vials in total evaporation mode; calibration ranges covered sub‑ppb to high ng/µL levels. Repeatability was evaluated by replicate injections of low‑level standards (n=9) and recovery/linearity checks across the dynamic range.

Used instrumentation

• TriPlus 500 HS autosampler (direct connection to GC column, solventless headspace extraction)
• Thermo Scientific TRACE 1610 Gas Chromatograph with iConnect split/splitless pneumatic control
• ISQ 7610 single quadrupole mass spectrometer operated in EI mode (SMART tune); combined FS and SIM acquisition
• TraceGOLD TG‑624 SILMS capillary column (60 m × 0.25 mm × 1.4 µm)
• Chromeleon Chromatography Data System (instrument control, data processing, reporting compliant with regulatory requirements)

Main results and discussion

• Linearity and dynamic range: Calibration showed linear response across four to five orders of magnitude. Quantitation ranges reported were 0.25–500 ng/µL for acetaldehyde and 0.01–500 ng/µL for benzene and d‑limonene, with coefficients of determination R2 ≥ 0.997 for all targets.
• Sensitivity: Calculated limits of detection (LODs) were low: benzene ≈ 0.01 ng/µL, acetaldehyde ≈ 0.02 ng/µL, and d‑limonene ≈ 0.03 ng/µL, demonstrating sub‑ppb capability in vial concentrations.
• Repeatability: Injection repeatability for low‑level standards (n=9) produced absolute peak area RSDs below 5% (reported values ~3–4.6%), evidencing robust pneumatic control and a thermally inert sample path.
• Targeted quantitative results in rPET samples: Measured concentrations (expressed as µg/kg) in six commercial rPET bottled water samples fell into typical literature ranges: acetaldehyde (≈1,100–2,300 µg/kg), benzene (≈1.0–16.4 µg/kg), and d‑limonene (≈22–54 µg/kg). These values underscore sample variability and the importance of routine monitoring.
• Untargeted screening: Full‑scan data enabled detection and putative identification of additional NIAS such as 2‑methyl‑1,3‑dioxolane and ethylene glycol derivatives by extracted ion chromatograms and spectral matching against the NIST 2023 library, illustrating the value of combined FS+SIM acquisition for both confirmatory and exploratory analysis.

Benefits and practical applications of the method

• Solventless, minimal sample preparation: Static headspace avoids laborious extraction steps and reduces risk of artefact formation or loss of volatile analytes.
• Fast throughput: Short GC run (~14 min) with automated HS sampling supports higher sample throughput for QC laboratories.
• Combined targeted + untargeted capability: SIM increases sensitivity for regulatory targets, while full scan enables discovery of unexpected contaminants and spectral confirmation.
• Direct autosampler‑to‑column connection: Minimizes transfer line length and potential active surface interactions, improving sensitivity and reproducibility.
• Data integrity and automation: Integrated CDS control enables streamlined workflows, reporting and compliance with electronic record regulations.
• Applicability: Suitable for routine monitoring of rPET used in food contact applications, process control in recyclate production, and screening in migration studies.

Future trends and potential applications

• Integration with high‑resolution mass spectrometry (HRMS) for improved identification confidence of unknown NIAS and isotopic profiling.
• Development of standardized HS protocols and reference materials for rPET to harmonize interlaboratory comparisons and regulatory compliance testing.
• Coupling HS‑GC with complementary sample enrichment (e.g., SPME, sorptive extraction) for even lower detection limits when required.
• Automated data mining and machine learning‑assisted spectral interpretation to accelerate untargeted screening and prioritize hazardous hits.
• Greater adoption of online monitoring within recycling plants to control thermal history and contamination sources that drive benzene formation.

Conclusions

This application demonstrates that static headspace GC–MS, implemented with a Thermo Scientific TriPlus 500 HS autosampler coupled to TRACE 1610 GC and ISQ 7610 MS, is an effective approach for sensitive, repeatable quantitation of acetaldehyde, benzene and d‑limonene in rPET. The workflow combines solventless sample handling, broad dynamic range, low LODs and simultaneous targeted/untargeted acquisition to support routine QC and screening for NIAS in recycled plastic packaging.

References

1. How does PET plastic recycling work? Recycle the One.
2. U.S. Food and Drug Administration. Use of Recycled Plastics in Food Packaging (Chemistry Considerations): Guidance for Industry.
3. Welle F. Maximum concentrations of limonene in mineral water bottles containing postconsumer PET recyclates without organoleptic deterioration. Fraunhofer Institute for Process Engineering and Packaging (IVV).
4. Shen X., Hed Y., Annfinsen S., Singh N., Anwar H., Mylvaganam B.T., Emmer A. (2025) Investigating polyethylene terephthalate beverage packaging: impact of recycled content on acetaldehyde, benzene, and other contaminants. Journal of Polymers and the Environment 33:2362–2370.
5. Commission Regulation (EU) No 10/2011 of 14 January 2011 on plastic materials and articles intended to come into contact with food.