Investigating Sulphur Components in Beer Using Gas Chromatography and Sulphur Chemiluminescence Detection

David Prahl/stock.adobe.com

Organosulphur compounds are important substances in the food industry because they can contribute to the flavour impression of a product. The human olfactory system’s sensitivity to sulphur leads to low flavour threshold values, making the analysis of these substances a challenging task. Gas chromatography (GC) with sulphur chemiluminescence detection (SCD) is a highly sensitive and selective technique for the analysis of sulphur compounds in various matrices. Using a range of different beers as an example, an approach is presented to reliably qualify and quantify sulphur components in beverages using headspace sampling and GC–SCD.

Sulphur has been known since prehistoric times as it exists in elementary form. However, many volatile sulphur compounds are known for their unpleasant odour, such as hydrogen sulphide (H₂S), which smells of rotten eggs (1). Consequently, sulphur compounds like sulphides or thiols are important substances in the food industry as they can contribute to the flavour impression of a product. The human olfactory system is sensitive to sulphur, leading to low flavour threshold values, making the analysis of these components an important yet challenging task (2).

Sulphur Chemiluminescence Detection

Selective sulphur detection is particularly essential for high matrix load samples because the hydrocarbon background cannot be separated chromatographically from the sulphur compound, which is typically present in low concentration. Sulphur chemiluminescence detection (SCD) is both a sensitive and selective technique and is therefore common in petrochemistry to study the sulphur content of diesel or gasoline. This detection technique can also be an alternative for less severe matrix cases
(3).

In food chemistry flame photometric detection (FPD) is widely used, although it suffers from hydrocarbon quenching and non-linearity of calibration curves. SCD offers increased selectivity and linearity when compared to FPD (3). SCD was therefore chosen for the analysis of beer samples regarding their sulphur content, as presented in this article.

Materials and Methods

Samples and Reagents: A matrix of 5% ethanol in water was used to check instrument performance and sample preparation quality.

Eighteen different beers were chosen randomly to test their sulphur content. To increase sample diversity, 10 came from Germany, while the remainder were of Dutch origin. Different types of beer were chosen from each country.

The overall setup included Pilsner beer, wheat, alt, bock, lager, and dark beer, as well as the so-called Kellerbier, an unfiltered and unpasteurized type from Germany (4). To gain an impression of sulphur content change in the process of converting a regular beer into its non-alcoholic version, three wheat and one Pilsner beers from the setup were compared qualitatively against their alcohol-free counterparts.

Ethyl methyl sulphide was used as internal standard (ISTD) for all measurements. Quantitation was performed on three sulphur compounds: dimethyl sulphide (DMS), S-methyl thioacetate, and dimethyl disulphide (DMDS). All sulphur standards were purchased as pure substances and weighed to create the respective standard solutions.

Sample Preparation: Sulphur analysis is challenging, and not just in terms of analytical setup. Sample preparation steps need to be planned carefully as many sulphur compounds show high volatility, making efficient sample preparation necessary. Slow working during the sample preparation sequence decreases reproducibility when measuring a series of control samples. Furthermore, negative sample preparation effects contribute directly to deviations of the calibration curve, leading to a decrease in calibration quality.

To avoid such negative effects, all standards and samples in this work were kept cold, and the vials were also precooled before filling to ensure the least possible evaporation during sampling. A 10 g measure of beer was weighed into the cold headspace vial and spiked with the standard solutions at the respective concentrations. The internal standard was added as the last compound to ensure minimum deviation. To minimize losses of the already spiked standards, the vials were covered by caps between the spiking series of standards and ISTD.

Experimental Conditions: The beer samples were analyzed using a GC‑2030 system equipped with SCD‑2030 chemiluminescence detector and HS‑20 headspace sampler (Shimadzu). The beer samples were incubated at 80 °C for 15 min before being transferred to the GC system with a split of 1:5. Compound separation was achieved using a 30 m × 0.32 mm, 4.0-µm SH‑Rxi-1MS column (Shimadzu), within a chromatographic run time of 15 min. Quantitation was based on a three‑level standard addition of the standard compounds. The calibration range was chosen as 8 to 80 ppb for DMS, and 1 to 10 ppb for S-methyl thioacetate and DMDS, as the latter two were suspected to be present in lower amounts than DMS.

Results

Reliability of the workflow was observed by checking reproducibility of the internal standard in a 5% ethanol in water matrix. This excluded possible effects from real beer samples and therefore gave a measure of the overall suitability of the approach. The area reproducibility of the ISTD was found to be below 2% relative standard deviation (RSD), offering a reliable measurement of sulphur compounds when following the sample preparation workflow and instrument setup reported above. For real beer samples poured from freshly opened bottles, the reproducibility of the internal standard was still found to be below 5% RSD. As each beer sample was weighed out to 10 mg, the increased RSD value indicates the impact of CO₂ on sample preparation.

A typical chromatogram of the beer samples is shown in Figure 1. Up to five sulphur compounds were assigned qualitatively in the beers: methanethiol, DMS, S-methyl thioacetate, DMDS, and dimethyl trisulfide. Of these components, DMS, S-methyl thioacetate, and DMDS were quantified.

Area reproducibility on the target compounds in beer was found to be below 5% RSD for dimethyl sulphide and below 7% RSD for the sulphur components present in trace levels. Detection (DL) and quantification (QL) limits were estimated from the beer sample measurements using signal‑to‑noise (S/N) values of 3.3 (DL) and 10 (QL), respectively. The detection limits obtained were 0.11 µg/L for DMS, 0.49 µg/L for S-methyl thioacetate, and 0.04 µg/L for DMDS. Quantitation limits were assigned as 0.33 µg/L for DMS, 1.50 µg/L for S-methyl thioacetate, and 0.13 µg/L for DMDS.

Sulphur Component Patterns in Alcoholic Beers

For all beer samples analyzed, dimethyl sulphide content was below 0.07 mg/L. In two cases, values above 0.06 mg/L were observed, whereas the other samples contained amounts between 0.012 and 0.043 mg/L. The flavour threshold value for DMS is 0.03 mg/L; above this an undesired sweetish, onion-like contribution to the beer aroma becomes apparent (5).

In eight beer samples, the dimethyl sulphide content was below this threshold, whereas four of the beer samples slightly exceeded this value with amounts between 0.030 and 0.035 mg/L. S-methyl thioacetate content was found to be below 0.02 mg/L in all beers, with only three samples showing more than 0.01 mg/L. This is well below the flavour threshold of 0.05 mg/L, preventing a cooked cabbage flavour contribution
(6).

More than half of the beers tested contained no detectable levels of DMDS. The others were quantified as below 1.9 µg/L, less than the flavour threshold value of 0.0075 mg/L, above which a rotten vegetable character would impact the beer (6). For the German beers, only one alt and one wheat beer showed DMDS levels above the detection limit, with both levels being lower than those found in the Dutch beers.

A principal component analysis (PCA) was performed to investigate variations within the two groups of beer samples. The resulting score and loading plots of German and Dutch beers are shown in Figure 2.

Within the German beer samples, the alt beers differ significantly in their sulphur content pattern from the Pilsner, wheat beer, and Kellerbier, which all cluster. Interestingly, the lager beer also differs from the clustering beers including the Kellerbier, which by definition is a type of lager (4).

The Dutch beers in contrast are generally more diverse and show no pronounced clustering. The four bock beers (B1 to B4) differ significantly from each other and from the other types, with B4 being the closest to the other groups. The two Dutch Pilsner on the other hand were found to be quite similar.

Sulphur Component Patterns in Non‑Alcoholic Beers

In addition to the alcohol containing beers, alcohol-free versions were studied qualitatively. When comparing the sulphur component patterns of the regular beers with their non-alcoholic counterparts, dimethyl sulphide was decreased in all four alcohol-free samples. The biggest decrease was found to be nearly 90% in the Pilsner, whereas the wheat type samples showed a decrease of 50%, 60%, and 80%, respectively. Similarly, S-methyl thioacetate content was decreased, in three cases even below the detection limit. DMDS, on the other hand, was only present in one of the regular wheat beers and showed an increase of around 50% in the alcohol‑free counterpart.

Conclusion

Gas chromatography with sulphur chemiluminescence detection is a highly sensitive and selective approach for the analysis of sulphur compounds in various matrices. This article presents an adapted method for sulphur content analysis in food and beverages using the example of different beers. Up to five sulphur substances were assigned qualitatively, three of them being quantified to investigate a possible contribution to the flavour impression.

References

1. A.F. Holleman, Lehrbuch der Anorganischen Chemie / Holleman-Wiberg (de Gruyter, Berlin, Germany, 1995), pp. 338–339.
2. N. Rettberg, M. Biendl, and L.-A. Garbe, J. Am. Soc. Brew. Chem.76(1), 1–20 (2018).
3. X. Yan, J. Sep. Sci.29, 1931–1945 (2006).
4. H. Dornbusch, in The Oxford Companion to Beer, G. Oliver, Ed. (Oxford University Press, Inc., New York, USA, 2011), p. 512.
5. W. Kunze, in Technology Brewing & Malting, O. Hendel, Ed. (VLB Berlin, Berlin, Germany, 2019), p. 388.
6. I. Hornsey, Brewing (RSC Publishing, Cambridge, UK, 2013), p. 210.

Rebecca Kelting holds a doctorate degree in chemistry from the Karlsruhe Institute of Technology (KIT), in Karlsruhe, Germany. She works as a GC and GC–MS product specialist for Shimadzu Europa in Duisburg, Germany, currently focusing on gas chromatographic application and product support.

E-mail: shimadzu@shimadzu.euWebsite:www.shimadzu.eu