Chemical Profiling of Commercially Available Cola Beverages Using GC-TOFMS and ChromaTOF Sync

Significance of the topic

GC‑MS based volatile profiling provides a direct window into the composition and stability of complex flavored beverages such as cola. Volatile flavorants and their degradation products determine aroma, shelf life and perceived product quality; profiling these compounds supports authenticity testing, quality control, sensory correlation and formulation development. High‑resolution, non‑targeted approaches reveal both expected ingredients and unanticipated markers (off‑odors, degradants), enabling manufacturers and analysts to detect subtle differences between commercially similar products.

Objectives and overview

This study aimed to produce a comprehensive non‑targeted volatile chemical profile of 13 commercially available cola beverages (from eight brands) using headspace solid‑phase microextraction (HS‑SPME) combined with GC‑TOFMS. Data from triplicate analyses were processed collectively with ChromaTOF Sync to automate peak finding, deconvolution and alignment across the sample set. The goals were to identify characteristic flavorants and degradation markers, quantify relative differences (e.g., caffeine), and uncover sample clustering or brand similarities via multivariate analysis (PCA).

Methodology

Sample set and design:

• Thirteen commercial cola beverages representing eight brands; all analyzed in triplicate.

Sample preparation and acquisition:

• Volatile extraction by HS‑SPME followed by analysis on GC‑TOFMS (Pegasus BT platform).
• Acquisition provided full mass spectral (m/z) data suitable for library matching and deconvolution.

Data processing and identification:

• ChromaTOF Sync was used to process the complete set of chromatograms together: automated peak finding, spectral deconvolution, alignment, and compilation of a peak table across samples.
• Automated matching to NIST library spectra and retention index (RI) comparison; minimum spectral similarity and processing thresholds reported (e.g., S/N ≥ 10, library similarity scores used for confidence).
• Multivariate analysis: Principal Component Analysis (PCA) and heatmap visualization to detect compositional trends and groupings.

Instrumentation used

• GC‑TOFMS: Pegasus BT (LECO) for high‑speed, high‑quality time‑of‑flight mass spectral acquisition.
• Software: ChromaTOF Sync for alignment, automated peak finding, deconvolution, heatmaps and cross‑sample visualizations.
• Library: NIST mass spectral library and retention index matching for putative identifications.
• Sample prep: Headspace solid‑phase microextraction (HS‑SPME).

Main results and discussion

• Approximately 150 deconvoluted peaks were detected and matched to the NIST library with similarity scores ≥ 700, yielding a broad volatile chemical map across the 13 colas.
• Representative identifications included flavorants and degradation products: p‑cymene (degradation product; similarity 916; RIobs 1028 vs RIlib 1025), D‑limonene (flavorant; similarity 863; RIobs 1033 vs RIlib 1031), and eucalyptol (flavorant; similarity 816; RIobs 1037 vs RIlib 1032). These assignments combined spectral match and RI agreement for confidence.
• Caffeine was observed in all samples (Rt ~648.8 s) with a high library match (similarity 938) and RI agreement (RIobs 1857 vs RIlib 1837). Relative abundance trends showed the expected energy drink had elevated caffeine, but an unsweetened product (Cola 5) exhibited the highest measured caffeine signal among the set.
• Terpenes, terpenoids, sesquiterpenes and esters varied across the products; unsweetened colas (Cola 5 and 6) and the energy drink (Cola 7) showed many elevated terpenoid/ester features. These compound classes are important for aroma contribution and may reflect formulation or added essential oils.
• PCA grouping indicated notable chemical similarity between two globally dominant cola products (Cola 3 and Cola 4), despite different branding and marketing claims, suggesting closely aligned core formulations that underlie comparable sensory profiles.
• ChromaTOF Sync features (heatmaps, side‑by‑side chromatogram overlays, bar charts) streamlined discovery of trends and facilitated navigation of complex peak tables, reducing manual review time.

Benefits and practical applications

• Quality control: detection of degradants (e.g., p‑cymene) that indicate acid‑catalyzed breakdown of flavor constituents and potential off‑odors.
• Authenticity and competitive analysis: chemical fingerprints enable differentiation or confirmation of product equivalence between brands and batches.
• Formulation and R&D: identification of terpene/ester profiles guides flavor optimization and reformulation efforts.
• Shelf‑life monitoring: tracking volatile changes over time to predict aroma loss or the appearance of off‑note compounds.

Future trends and applications

• Integration with quantitative targeted methods to translate relative signal trends into absolute concentrations for regulatory or nutritional labeling purposes.
• Expanded reference libraries and retention index databases tailored to beverage matrices to increase identification confidence and reduce false positives.
• Machine learning and chemometric pipelines that combine high‑dimensional GC‑TOFMS data with sensory or consumer data to predict perceived flavor and acceptance.
• High‑throughput screening workflows for routine QC that automate anomaly detection and batch comparison across large sample repositories.
• Cross‑platform data sharing to build commercial beverage fingerprints for authenticity databases and anti‑counterfeiting efforts.

Conclusion

Non‑targeted HS‑SPME GC‑TOFMS combined with ChromaTOF Sync alignment and deconvolution successfully revealed detailed volatile profiles across 13 commercial cola beverages. The approach identified key flavorants, degradants and variable compound classes (terpenes, esters) and highlighted unexpected trends such as elevated caffeine in an unsweetened product and strong compositional similarity between two major cola brands. These results illustrate the value of high‑resolution, automated workflows for beverage quality assurance, product differentiation and formulation insight.

References

Lorjaroenphon Y., Cadwallader K.R., Journal of Agricultural and Food Chemistry, 2015, 63, 776–789. DOI: 10.1021/jf504954p