Hydrogen as Carrier Gas for the Analysis of Dibenzo-p-dioxins and Dibenzofurans (PCDD/Fs) in Food and Feed Samples Using an Agilent 7010 Series Triple Quadrupole GC/MS

Importance of the topic

Polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) are persistent, bioaccumulative toxicants regulated at ultra-trace levels in food and feed. Gas chromatography coupled to triple quadrupole mass spectrometry (GC/TQ MS/MS) is an accepted confirmatory approach for these analytes. Switching GC carrier gas from helium to hydrogen offers faster chromatography and lower operating cost but poses analytical challenges (reduced signal-to-noise, altered spectra, possible chemical reactions and safety requirements). This study demonstrates a validated conversion of a helium-based PCDD/F GC/MS/MS method to hydrogen using an Agilent 7010 series GC/TQ with a high-efficiency EI source (HES), preserving compliance-level performance while substantially reducing run time.

Objectives and study overview

The study aimed to:

• Evaluate sensitivity and qualitative performance differences between Agilent 7000C and 7010 series triple quadrupole instruments under helium as a baseline.
• Translate and optimize the PCDD/F method from helium to hydrogen carrier gas and shorten chromatographic runtime without compromising quantitation at ultra-trace levels.
• Validate method performance across multiple food and feed matrices, assess matrix effects and limits of quantification (LOQs), and run a multi-month robustness study under routine high-throughput conditions.

Methodology

Key experimental design elements:

• Standards and calibration: 13-point isotope-dilution calibration including 17 native congeners and 13C12-labeled analogs; test-mix (TMS) and window-defining mix (WDM) prepared in toluene.
• Sample set: 167 diverse food and feed extracts (eggs, milk, infant formula, fats and oils, spices, fish, meat meals, mineral feeds) with native congener amounts ranging roughly 10 fg to 100 pg per sample.
• Analysis mode: GC/TQ MS/MS in Multiple Reaction Monitoring (MRM) with two precursor→product transitions per congener.
• Validation endpoints: linearity, residuals, relative response factor (RRF) stability, precision, repeatability, LOQs (process-blank based and S/N based), accuracy (SRMs), and long-term robustness (~2,000–3,000 injections over months).

Used instrumentation

The main instrumentation and critical settings:

• GC: Agilent 7890B gas chromatograph with Gerstel MPS autosampler and cold injection system (CIS) using PTV solvent vent injection (5 µL).
• MS: Agilent 7010 series triple quadrupole GC/MS (HES source) used for hydrogen experiments; 7000C used for baseline helium comparison (the 7000C was upgraded during the study).
• Column (final hydrogen method): VF-Xms, 40 m × 0.18 mm × 0.18 µm (Agilent CP9049); constant hydrogen flow 0.9 mL/min (average linear velocity ≈ 45.9 cm/s).
• GC oven and inlet: CIS PTV program with high temperature ramping; detailed oven program optimized to compress run to ~27.5 min.
• MS parameters (representative): transfer line ≈300 °C, ion source temperatures adjusted (conditioning up to 300 °C reported), electron energy 70 eV, collision gas used for CID, quench gas turned off when running hydrogen.
• Safety & system modifications for H2: hydrogen sensor, venting of split/septum purge lines, removal of loose covers, conditioning protocol (increased source/MS temps, EMV adjustments, overnight filament operation) and attention to pumping/vacuum load.

Main results and discussion

Performance comparison (7010 vs 7000 under helium baseline):

• 7010 with HES produced substantially higher ion yield (reported ~20× higher peak areas), yielding better linearity and lower LOQs (improvements of ~4–5× for select congeners) and improved calibration stability (RSD of calibration procedure 1.1–3.1% for 7010 vs 2.4–5.9% for 7000C).
• At the lowest calibration points, the 7010 produced clear signals (e.g., S/N = 13 at 1.73 fg/µL for 2,3,7,8-TetraCDD) where the 7000C was not reliably detectable.

Hydrogen conversion and method optimization:

• Initial conversion to H2 produced expected effects: narrower chromatographic peaks and faster elution, but lower absolute peak areas (~2× lower initially vs helium), increased baseline noise, and elevated early background. A conditioning period (about one week) reduced the background and stabilized performance.
• Final hydrogen method used a shorter, narrower 40 m column (0.18 mm i.d.) and 0.9 mL/min H2 flow, achieving ~27.5 min cycle time—a ~30–40% reduction in runtime compared with the 60 m helium configuration.
• Separation efficiency was comparable or improved for many congeners (hexa congeners showed up to ~10% better separation under H2), although a ~10% deterioration was observed for one tetraCDF separation that could not be resolved by method changes.

Analytical figures under hydrogen (selected):

• LOQs from S/N for TMS: approximately 10 fg (2,3,7,8-TetraCDD) and 18 fg (OctaCDD) absolute (n ≈ 27). Process-blank-based LOQs for routine samples ranged 0.05–1.15 pg/sample for the 17 native congeners, consistent with helium-based LOQs.
• Cross-method agreement: comparison of 167 samples measured in duplicate under He and H2 produced regression slopes near unity (0.96–1.04) and generally good correlation (most congeners R2 ≥ 0.86). Average native-congener ratio H2/He ≈ 91.6% (n = 16 congeners) and average labeled recovery ≈ 101.4% (n = 17 labeled congeners).
• Precision and accuracy: calibration RSDs 1–4% under H2; relative response factors stable with RSDs comparable to helium; SRM accuracy within ±20% and sample recoveries typically within 50–130% for both gases.

Robustness and long-term behavior:

• Over 5–10 months and ~2,000–3,000 injections, method parameters (linearity, RRFs, precision) remained stable. Some gradual performance decline attributed to normal system aging: LOQs by S/N increased modestly (e.g., 2,3,7,8-TetraCDD S/N LOQ increased from ~10 to ~20 fg absolute) and sensitivity decreased roughly 3× after extended use.
• Switching carrier gases repeatedly caused transient sensitivity losses and required long stabilization periods; periodic alternation of He and H2 is not recommended.

Potential chemical artifacts:

• Concerns about hydrogen-induced dechlorination (reported historically) were investigated. This study found no evidence of significant formation of lower-chlorinated artifact peaks that would compromise quantitation or TEQ assessment for the matrices tested, although minor dechlorination could not be entirely excluded and should be actively checked in other applications.

Benefits and practical application of the method

Practical advantages demonstrated:

• Substantial runtime reduction (≈30–40%) while retaining regulatory-level quantitation and confirmatory capability for PCDD/Fs in complex food/feed matrices when using a highly sensitive detector (Agilent 7010 with HES).
• Lower operating gas cost with hydrogen and reduced dependence on helium supply.
• Comparable or improved chromatographic separation for many congeners using appropriately chosen shorter/narrower columns and optimized flow programs.

Future trends and potential applications

Outlook and recommendations:

• Hydrogen is a viable alternative carrier gas for ultra-trace GC/TQ analyses of stable hydrophobic organic compounds when instrument sensitivity is sufficient (detection capability at or below ~10 fg on-column). The approach is especially attractive where helium supply or cost is problematic.
• Careful method transfer is required: evaluate potential chemical reactivity (dechlorination, hydrogenation), re-optimize column selection (phase and geometry), and implement safety and vacuum/pumping accommodations for hydrogen.
• Other compound classes may present different reactions with hydrogen; each analyte group should be explicitly evaluated for spectral changes, fragment stability, and artifacts before routine adoption.
• Future improvements may include wider adoption of high-efficiency ionization sources, column chemistries tailored for H2, and automated conditioning protocols to shorten stabilization after gas changes.

Conclusions

The study demonstrates that PCDD/F confirmatory analysis by GC/TQ MS/MS can be transferred from helium to hydrogen carrier gas without compromising regulatory performance, provided a highly sensitive instrument (7010 series with HES) is used, method parameters are re-optimized (column geometry, flow, oven program), and appropriate safety and conditioning measures are implemented. Hydrogen permits notable runtime reduction and operating-cost savings while maintaining LOQs and accuracy for most congeners. Routine high-throughput use over months showed good robustness; gradual sensitivity loss observed over very long operation is consistent with system aging rather than being intrinsic to hydrogen operation.

References

1. Alvarado JS, Silzer J, Lemley F, Erickson MD. Separation of Polychlorinated Biphenyls by Fast Gas Chromatography. Analytical Communications. 1985;34:381–383.
2. Cramers CA, Janssen H‑G, Van Deursen MM, Leclercq PA. High-Speed Gas Chromatography: An Overview of Various Concepts. Journal of Chromatography A. 1999;856:315–329.
3. Muñoz-Guerra JA, Prado P, Vargas García-Tenorio S. Use of Hydrogen as a Carrier Gas for the Analysis of Steroids with Anabolic Activity by Gas Chromatography–Mass Spectrometry. Journal of Chromatography A. 2011;1218:7365–7370.
4. Agilent Technologies. EI GC/MS Instrument Helium to Hydrogen Carrier Gas Conversion. User Guide; publication 5994-2312EN, 2022.
5. L’Homme B, Scholl G, Eppe G, Focant J‑F. Validation of a Gas Chromatography–Triple Quadrupole Mass Spectrometry Method for Confirmatory Analysis of Dioxins and Dioxin-Like Polychlorobiphenyls in Feed Following New EU Regulation 709/2014. Journal of Chromatography A. 2015;1376:149–158.
6. Quimby BD, Prest FH. Conversion of Agilent EI GC/MSD Systems from Helium to Hydrogen Carrier Gas. Agilent Technologies presentation, 2012.
7. Lau BP‑Y, Sun W‑F, Ryan JJ. Complication of Using Hydrogen as the GC Carrier Gas in Chlorinated Dibenzofuran and Dibenzodioxin GC/MS Analysis. Chemosphere. 1985;14(6/7):799–802.