Sensory-Directed Analysis for Beer Aroma Profiling: Stacking Multiple SPME Extractions to Help Connect Olfactory and GC-MS Data

Sensory-Directed Analysis for Beer Aroma Profiling: Stacking Multiple SPME Extractions to Link Olfactory and GC-TOFMS Data

Significance of the topic

Volatile compounds at trace concentrations can disproportionately influence aroma due to widely varying odor thresholds. Reliable detection and identification of these low-level odorants are critical in food, flavor, and fragrance analysis for product quality, troubleshooting off-flavors, and formulation. Combining sensory (GC-Olfactometry) information with highly sensitive instrumental analysis increases the chance of assigning olfactory observations to specific chemical species. Stacking multiple SPME extractions before a single chromatographic injection is a practical sample concentration strategy that can reveal analytes below the standard detection limit of a single SPME extraction.

Objectives and study overview

This application note demonstrates a sensory-directed workflow that couples GC-O data with GC-TOFMS analysis of beer aroma. The specific goals were to:

• Use olfactory notes from prior GC-O analyses to prioritize chromatographic regions for investigation.
• Evaluate whether stacking multiple headspace SPME extractions into a single injection increases detectability of low-level aroma compounds.
• Identify analytes responsible for selected olfactory notes using mass spectral matching and retention index (RI) confirmation.

Used instrumentation

The workflow relied on the following instrumentation and software:

• Autosampler: LECO L-PAL 3 for automated SPME handling and stacked injections.
• SPME fiber: tri-phase DVB/C-WR/PDMS fiber, conditioned 2 min at 250 °C between extractions.
• Cryotrap: GL Science cryofocus used at the head of the GC column to collect sequential SPME desorptions (held at −110 °C, rapid ramp to 250 °C at 60 °C/s for injection).
• Gas chromatograph: Agilent 8890 with Rxi-5ms column (30 m × 0.25 mm i.d. × 0.25 µm).
• Mass spectrometer: LECO Pegasus BT time-of-flight MS (ion source 250 °C, mass range 33–500 m/z, acquisition 10 spectra/s).
• Software: ChromaTOF for acquisition and deconvolution, GL Science Evolution for cryotrap control, prior GC-O data collected with Phaser Pro.

Methodology

Beer headspace SPME conditions: sample incubated at 40 °C for 2 min, extracted 4 min at 40 °C. Single-extraction analyses were compared to stacked-injection analyses where 2, 3, or 4 SPME extractions were sequentially desorbed and trapped on the cryofocus, then released as a single combined injection. Desorption in the GC inlet was 2 min at 250 °C (splitless). GC oven program: hold 5 min at 50 °C, ramp at 10 °C/min to 250 °C, hold 2 min. Helium carrier gas at 1.4 mL/min. An alkane standard run under the same conditions provided retention index (RI) calibration. Mass spectral library matching plus RI agreement were used to support tentative identifications. ChromaTOF deconvolution resolved coeluting components.

Main results and discussion

Linking GC-O notes to GC-TOFMS features allowed focused investigation of chromatographic regions with sensory relevance. Key observations:

• Several olfactory notes corresponded to strong chromatographic peaks that were readily identified. Example: ethyl hexanoate was identified (library similarity ~934, RI match 999) and connected to a fruity aroma note.
• Some olfactory notes mapped to peaks with lower signal-to-noise (S/N); examples include ethyl heptanoate and linalool (similarities ~929 and ~818, RI agreement ~1097–1100), consistent with fruity/citrus descriptors.
• Certain olfactory observations corresponded to chromatographic regions without an obvious peak in single-extraction runs. Deconvolution resolved coeluted components near a ‘‘musty’’ note, yielding tentative IDs (e.g., terpinolene, p-cymenene, 2-nonanone, propyl hexanoate) but their known aroma descriptors did not match the observed musty character.
• Stacking SPME extractions improved detectability of low-abundance analytes. Notable examples discovered when stacking 2–4 extractions:
  • Tetramethyl pyrazine emerged with increasing stacked extractions (confirmed by m/z 136.13 XIC, similarity ≈750 and RI 1089 vs library 1088) and has musty aroma descriptors, providing a plausible chemical source for the previously unexplained musty GC-O note.
  • Isoeugenol became detectable only after stacking, with spectral similarity ≈833 and RI agreement (observed ~1457 vs library 1450). Its spicy/woody aroma descriptors align with the spicy/nutmeg olfactory note.

Stacking effectively increased peak heights and S/N, enabling identification of analytes that were otherwise below detection in single SPME injections. Deconvolution remained a valuable complementary tool for resolving coelutions and isolating mass spectra for library matching.

Benefits and practical applications

Practical advantages demonstrated by this approach include:

• Enhanced sensitivity for trace volatile detection without major changes to chromatographic hardware.
• Sensory-directed prioritization reduces time spent searching complex chromatograms by focusing on olfactory-flagged regions.
• Ability to link low-level chemistry to sensory impacts, useful for product development, quality control, and off-flavor root-cause analysis in beverage and flavor industries.
• Compatibility with automated workflows (autosamplers, cryotrap control) for repeatable stacked injections.

Considerations and limitations

Operational and analytical considerations include:

• Potential for carryover or fiber overloading with multiple sequential desorptions — appropriate conditioning and blanks are required.
• Stacking increases total sampled mass but may also concentrate matrix components, possibly affecting chromatographic behaviour or ion suppression.
• Identification confidence hinges on mass spectral similarity and RI agreement; orthogonal confirmation (standards or high-resolution MS) strengthens assignments.
• Method optimization (number of stacked extractions, fiber type, incubation/extraction times, cryotrap parameters) should be tailored to the matrix and analytes of interest.

Future trends and applications

Potential developments and broader uses of stacked SPME/GC-O/GC-MS workflows include:

• Integration with higher-resolution MS or tandem MS to raise confidence in identifications for trace odorants.
• Wider use of SPME Arrow or larger-volume headspace concentrators combined with cryofocusing to push detection limits further.
• Automated sensory-directed routines where GC-O flags trigger targeted stacked reanalysis or enhanced deconvolution workflows.
• Data fusion and machine-learning approaches to correlate chemical signatures with sensory panel scores across many samples for predictive aroma models.
• Application to other matrices requiring trace volatile profiling, such as environmental, forensic, or clinical samples.

Conclusion

Combining GC-O with GC-TOFMS and employing stacked SPME extractions with cryofocusing is an effective strategy to reveal low-level aroma-active compounds in complex matrices like beer. This sensory-directed workflow improves the probability of detecting and identifying trace odorants (e.g., tetramethyl pyrazine, isoeugenol) that explain previously unassigned olfactory notes. The approach is practical for routine lab use when enhanced sensitivity is needed and can be integrated into automated analytical pipelines.

References

1. Good Scents Company. The Good Scents Company database; aroma descriptors for flavor and fragrance compounds.