DETERMINATION OF DIMETHYL SULPHIDE IN BREWERY SAMPLES BY HEADSPACE GAS CHROMATOGRAPHY MASS SPECTROMETRY (HS-GC/MS)

Headnote

ABSTRACT

Dimethyl sulfide (DMS) is an important beer flavour compound and when there is more than 100 µg/L as free DMS, it is considered an off-flavour because it may give an undesirable cabbage flavour and smell to the beer. The aim of this work was to set up a gas chromatography-mass spectrometry (GC-MS) method, coupled with headspace, to determine the DMS present in the beer chain. This could then be useful for setting up the quality control of parameters in malting and brewing. A case study for application of the proposed method is reported for barley variety screening.

- Key words: barley, beer, dimethyl sulfide, gas chromatography-mass spectrometry, malt -

INTRODUCTION

A wide variety of sulphur compounds has been reported to occur in beer (NYKÄNEN and SUOMALINEM, 1983). Even if the malt, hops and water used in brewing may all be possible sources of sulphur compounds, the principal volatile and semi-volatile compounds do not come directly from the raw materials, but they mainly originate during malting and fermentation (HILL and SMITH, 2000; WALKER and SIMPSON, 1992;). The main volatile sulphur compound in beer is dimethyl sulfide (DMS), a thioether which plays an important role in beer flavour (BAMFORTH, 1998; HILL and SMITH, 2000; KUNZE, 2004). The contribution of DMS in beer is well known (YANG and SCHAWARZ, 1998a). Its threshold level in beer is about 3050 µg/L (BAMFORTH, 1998; ZHAO et al, 2006). In the range of 50-100 µg/L, DMS contributes to the taste of lager-type beers, whereas above 100 µg/L it may give an undesirable vegetablelike or cabbage flavour and smell to the beer. The control of DMS formation, or its removal, is therefore desirable.

Two DMS precursors (DMS-p) or two routes of DMS formation during malting and brewing have been proposed and different techniques have been used to identify the precursors: Smethyl-L-methionine (SMM) and dimethyl sulfoxide (DMSO). Both of these precursors originate from malt but they form DMS in different ways (YANG and SCHAWARZ, 1998a).

SMM, which is synthesized in the seed germ and rootlets during barley germination, is converted into DMS under high temperatures during malt kilning and wort boiling. DMSO, instead, is reduced to DMS by yeast during wort fermentation (ANNESS and BAMFORTH, 1982).

It has widely been accepted that the final DMS level in beer will depend mainly on the SMM content of the malt. However, it is also believed that the level of DMSO in wort plays a major role in determining the DMS level in the final beer (KUNZE, 2004; ZHAO et al, 2006; YANG and SCHAWARZ, 1998a).

During steeping and germination some attempts have been made to transform SMM as little as possible into active free DMS (KUNZE, 2004). The residual part of SMM is thermally degraded into DMS mainly during malt kilning, and a small amount is lost to the atmosphere with the evacuation of gases.

During the kilning process, part of the SMM is degraded into DMS, and, at the same time, some DMS is oxidized to DMSO. During wort fermentation, DMSO can be transformed into DMS and water (DMSO [arrow right] DMS + H2O) through metabolic reduction by brewer's yeast, catalyzed by DMSO-reductase in a NADPH-dependent manner (BAMFORTH, 1980). The amount of DMS produced by yeasts changes from strain to strain (WALKER and SIMPSON, 1992) and this kind of DMS production is one of the main causes of beer off-flavours. The reduction of DMSO is probably affected by the formation of methionine sulfoxide (MetSO). MetSO competitively inhibits DMSO-reductase because it is the preferred substrate of the enzyme. The affinity of DMSO reductase for MetSO is higher than that of DMSO (ANNESS and BAMFORTH, 1982; HANSEN et al, 2002).

Intensive wort boiling causes the DMS concentration in the final beer to be reduced (HERTEL and SOMMER, 2008; KUNZE, 2004; VAN DEN EYNDE, 1991).

The DMS produced from DMSO during fermentation is removed together with the other volatile compounds present by mean of stripping with CO2. (VAN DEN EYNDE, 1991).

Given the importance of DMS in beer flavour, and taking into account the influence of technical parameters of malting and brewing on the DMS content of beer, a sensitive instrumental method of analysis for the routine determination of this compound is required (COGHE et al, 2004).

The study and determination of DMS is difficult principally due to its low concentration in beer. Various techniques have been reviewed: amperometry, titrimetry (GARZA-ULLOA, 1980), colorimetry (INSTITUTE OF BREWING, 1982), fluorimetry (BLOCKMANS et al, 1981) and potentiometry (NARZISS et al, 1985) but none of them provide the level of sensitivity and selectivity usually required.

Actually, the simplest analytical methods involve headspace sampling of the beer followed by direct injection into a gas cromatograph (GC) (WALKER, 1991), but this type of analysis is not useful when the concentration of the sulphur compounds is low or when they are partially volatile. Generally, in these cases methods, which involve "purge and trap", whereby the volatile components in the beer headspace are concentrated onto a suitable adsorbent material, have been described (LEPPANEN, 1979). Furthermore, it is possible to set up an analytical method using columns and detectors specific for separating and identifying sulphur compounds (MUNDY, 1991; HILL and SMITH, 2000). Several researchers have indicated the use of headspace gas chromatography or sometimes Gas -Liquid Chromatography (GLC) equipped with a Flame Photometric Detector (FPD) or a Sulphur Chemiluminescent Detector (SCD) or a Pulsed Flame Photometric Detector (PFPD) or a Flame Ionization Detector (FID) as possible analytical devices, with or without preliminary Solid-Phase Micro Extraction (SPME) (BAKER, 1988; DROZD et al, 1990; DUPHIRE, 1999; HILL and SMITH, 2000; VAS, 1997; WALKER, 1992; YANG and SCHWARZ, 1998b). SPME has already been applied in many areas of the food and beverage industries with important results (HILL and SMITH, 2000).

Little information is available, however, on the use of gas chromatography coupled with Mass Spectrometry (MS) for DMS analysis in barley, malt, wort and beer.

The aim of the present work was to demonstrate the possibility of setting up a HeadspaceGas Chromatography-Mass Spectrometry (HSGC-MS) method able to identify and quantify DMS as free DMS, total-DMS and DMS precursors (DMS-p) (total DMS - free DMS) in barley, malt, wort and beer, in order to control the malting and brewing process.

With GC, DMS in beer can be sampled and at the same time, coupling GC with MS, it is possible to use the high selectivity, precision and sensitivity of MS to identify the substance of relevance from barley and malt as well as from elaborate matrices such as wort and beer.

Different varieties of barley and their respective malts, worts and beers were analysed as a case study for the application of the proposed method.

MATERIALS AND METHODS

Reagents

DMS and ethyl methyl sulfide (EMS) standards were purchased from Sigma-Aldrich (Milwaukee, WI). Ethanol 99% (v/v) was purchased from VWR International S.R.L. (Milano, IT). Sodium hydroxide was purchased from J.J. Baker (Deventer, NL). Distilled water was of high purity.

Barley samples

Six two-row varieties of barley were used: Henley, Chert Scarlett, Barke, Regina and Puffin. The first four were spring and the last two were autumnal. All samples were grown in the experimental fields of the University of Perugia.

Malting process

For the evaluation of barley malts, a standard malting program was followed. The steeping, germination and drying programs were suitable for the production of "pilsner" type malt. The barley samples were malted in a micromalting pilot-plant (Custom Laboratory Product, Keith, Scotland, UK) consisting of 4 independent steeping/germination tanks, each having a 5 kg capacity.

The steeping phase lasted 20 hours and was divided into 3 steps:

- Step 1: 5 h of soaking in water at 18°C followed by 7 h of resting at 17°C.

- Step 2: 2 h of soaking in water at 16°C followed by 5 h of resting at 16°C.

- Step 3: 1 h of soaking in water at 16°C before starting of germination.

The germination phase lasted 4 days, during which time the temperature was regularly decreasing: 16°C during the first day, 15°C during the second day and 14°C for the last two days. The green malts were dried using a drying plant (Custom Laboratory Product, Keith, Scotland, UK) consisting of 4 units and applying the following drying temperatures:

- 55°Cfor 15 h;

- 72°C for 4 h and 30 min;

- 82°C for 3 h and 30 min;

- Cooling (to room temperature).

Brewing process

Mashing

Each sample of barley malt was processed in a pilot-scale plant (Braumaister, Feltre, I) programmed in order to produce 25 litres of wort suitable for "pilsner" type beer. Five kilograms of malt were milled in a two-roller mill (120 kg/h capacity, ENGL maschinen Gmbh, Schwebheim, Germany,) with a 0.5 mm gap between the crushing rollers. The grist was then mixed with 23 L of water. The properties of the brewing liquor were: 0.5 Millival, 15 mg/L of carbonate (CO3-2), 30.5 mg/L of hydrogen-carbonate (HCO3-2), 1 1 mg/L of dissolved CO2 and pH 6.5. The milled malt was mashed in the brewing liquor at 52°C for 30 min to allow protein breakdown. The temperature was increased to 65°C in 15 min and maintained for 45 min for development of the beta-amylase activity. The temperature was increased to 72°C in 5 min and maintained at this temperature for 20 min to allow the alpha- amylase activity to develop. The efficiency of saccharification was tested by the iodine solution test.

The last temperature step was a short rest at 76°C to inactivate the enzymes. Afterwards, the mash was transferred to a 30 L lauter-tun vessel The first wort was collected from the lauter-tun, then the spent grains were sparged with 12 L of water (78°C) to wash out the second wort. The total amount of wort was collected in a heated 30 L kettle for boiling; the temperature was raised to 100°C and maintained for 75 min. 100 grams of hops pellets (Saaz variety, 2. 1% of a- acids) were added: 70% of hops was added at the beginning of boiling, and 30% was added 15 min before the end of boiling.

Fermentation

The wort was cooled to 18°C and transferred to a 30 L fermentation vessel (Spadoni, Orvieto, I). 11.5 grams of ale yeasts (Safale S-04, Fermentis, Marcq-en-Baroeul Cedex - France) were activated with 100 mL of sterile wort. The fermentation temperature was maintained at 18°C for 5 days. When the gravity of the wort reached the limit of attenuation (85%), the beer was bottled to continue fermentation at 18°C. At the end of maturation, the bottles were kept at 4°C for two weeks and then analysed.

Apparatus

The DMS concentration was determined in static headspace with an Agilent Technologies 6850 gas Chromatograph equipped with an Agilent Technologies Mass Spectrometer 5975C (Santa Clara, CA) coupled with Maestro Autosamples Gerstel Multi Purpose Sampler (MPS) (Baltimore, MD).

GC-MS conditions

The gas chromatograph-mass spectrometer was fitted with a DB-5MS capillary column, 60 m X 0.32 mm 1µm (J&W Scientific Inc, Folsom, CA) consisting of cross-linked 5% phenyl methyl siloxane. Helium was the carrier gas at a flow rate of 1 .2 mL/min. The injection port temperature was 250°C, and the detector was set at 280°C. The source and the quadrupole temperature were 230° and 150°C, respectively. The ionization energy was 7OeV. Detection and data acquisition was performed in Selected Ion Monitoring (SIM) mode.

The target molecular ions for EMS were selected to be (m/z) 48, 61 and 76 (Fig. 1) and those for DMS were selected to be 47 and 62 (Fig. 2).

The oven temperature program used was 40°C for 2 min, followed by an increase of 3°C/min to 100°C and 10°C/min to 150°C. The final temperature was held for 3 min. The headspace conditions were 30 min of incubation at 50°C on unrest (250 rpm). The shaker temperature was 60°C, the syringe capacity was 2.5 mL, and the temperature was 70°C. The volume injected was 1 mL.

Standard solutions and calibration

DMS stock solution was prepared at 2,553 mg/L in ethanol whereas the DMS working solution was prepared at 25.53 mg/L in ethanol. EMS was used as internal standard (IS). The EMS stock solution was prepared at a concentration of 8,422 mg/L in ethanol and the working standard solution at 5 mg/L in ethanol.

DMS calibration curve (Fig. 3) was determined for five different concentrations in a range from 1.66 µg/L to 150.00 µg/L and with EMS as IS (0.05 mg/L). The calibration solutions were prepared from the working standard solution. Each concentration was analyzed three times.

Sample preparation

The sample preparation was carried out following European Brewery Convention (EBC) and Mebak methods with some modifications. The free DMS concentration of barley and malt samples was determined on a cold water extract of 10 +/-0.01 g of ground barley and malt, respectively, mixed with 100 mL of water under magnetic agitation at 150 rpm for 35 min. at room temperature (20°C). Then, 5 mL of the extract were mixed with 5mL of water and 100 µL of IS, and the solution was directly analysed by HSGC-MS. The total DMS concentration was determined by adding 1 mL of IN NaOH, 6.5 mL of water and 100 µL of IS to 2.5 mL of the barley and malt extract. Then the solutions obtained (10.1 mL), with an IS concentration of 0.05 mg/L, were boiled and analyzed by HS-GC-MS.

The free DMS concentration in wort and beer was quantified directly in 10 mL of sample, adding 100 µL of IS solution. The total DMS concentration was determined by adding 2.5 mL of IN NaOH, 1 mL of water and 100 µL of IS solution to 6.5 mL of wort sample and boiling them for 60 min. The sample solutions were prepared in 20-mL vials in duplicate and then analyzed by HS-GC-MS. The beer samples were not degassed to remove the carbon dioxide in order to avoid removing of DMS, because of its high volatility.

The total and free DMS concentration in barley, malt, wort and beer was obtained by mean of the calibration curve (Fig. 3). The DMS-p was calculated by subtracting free DMS from tot-DMS.

RESULTS AND DISCUSSION

The first step was the setting up of an innovative analytical method by GC-MS to identify and quantify DMS in different kinds of samples after elimination of interferring substances. Barley and malt samples for the analysis of the DMS content were mainly prepared following the official methods, but with some modifications (ANALYTICA-EBC 3.39; HILL and SMITH, 2000; MEBAK ROHSTOFFE 3. 1.4. 17; WALKER, 1991; YANG and SCHWARZ, 1998b). Filtration of the extract was excluded because it causes DMS losses. Fig. 4 shows that a satisfactory separation of DMS was achieved in a short time.

The performance of the GC-MS method was evaluated taking into account the calibration curve, the linearity, the repeatability, the limit of detection (LOD) and the limit of quantification (LOQ) (Table 1). The confidence interval of the calibration curve included the origin of the axes, and the coefficient of determination was r^sup 2^=0.996. The LOD was calculated as sb+3s, where sb was the average signal of ten blank injections and s the standard deviation. The LOQ was calculated as s,+10s, where s, was the average signal of ten blank injections and s the standard deviation (EURACHEM GUIDE, 1998). The blank sample was obtained by boiling the beer for 30 min to evaporate all free DMS. The LOD was 0.09 µg/L and the LOQ was 0.23 µg/L.

The repeatability of the method was measured by performing the determination of the same beer sample with a free DMS concentration of 25 µg/L 10 times on the same day and estimating their RSD (%) (EURACHEM GUIDE, 1998). The RSD calculated on each concentration used to work out the calibration curve was 4.96%.

The accuracy of the method for DMS in malt was verified by participating in the International Proficiency MAPS test (Malt Analytes Proficiency Testing Scheme) (LGC Standards Proficiency Testing, Lancashire, UK) every three months. The results of the participation in the MAPS rounds are reported in Table 2. The accuracy of the method for DMS in beer was verified by participating in the International Proficiency BAPS test (Brewing Analytes Proficiency Scheme). The results of the participation in the BAPS rounds are reported in Table 3.

The efficiency of the proposed method was evaluated by performing the recovery test on beer samples with the addition of three different DMS concentrations: 25, 50 and 75 µg/L. The recovery test results are reported in Table 4. The recovery ranged from 69 to 101% with a mean value of 80%. The data showed good results at the lowest spiked level which corresponds to the acceptable range for free DMS in beer. Therefore, the recovery test and the MAPS results confirmed the efficiency of the method.

The analytical screening of different samples allowed the development of DMS to be studied during malting and brewing. A comparative study of the results enables the following remarks to be made.

The majority of total DMS in barley was represented by DMS-p which was more than 80% in each case. The Regina, Puffin and Barke varieties showed a similar DMS-p concentration, with values of 0.23, 0.14 and 0.20 mg/kg, respectively, confirming that a low amount of DMS-p is present in barley (Table 5) (YANG and SCHAWARZ, 1998a). The differences between the six barley varieties are probably due to their different genetic characteristics.

Table 6 reports the free DMS, total DMS and DMS-p contents of the malts analysed. All malts contained DMS and DMS-p. It is well known that during germination the level of SMM increases and then kilning which leads to an increase in the temperature until 82°C, causing an enhancement of free DMS because of the thermodegradation of SMM to DMS.

The Cheri variety malt showed the highest amount of DMS-p, while the other five varieties showed similar concentrations (p<0.05). Among the six varieties, Scarlett had a significant concentration of free DMS with 8.47 mg/kg. This can be explained by the hypothesis that a high level of SMM in Scarlett green malt was produced during germination and then it was broken down during kilning to produce DMS. Scarlett is probably the variety with the highest amount of enzymes involved in SMM synthesis (ANNESS and BAMFORTH, 1982). Therefore, considering that all the barley samples were subjected to the same malting process, the different DMS concentration should be related to the variety and its genetic factors. It is interesting to focus on the reaction of the six different barley types to the heat treatment.

The free DMS concentration in wort ranged from 9.91 µg/L for Scarlett to 83.59 µg/L for Barke (Table 7). The autumn varieties Regina and Puffin had a DMS-p concentration similar to Henley; among the spring varieties only Scarlett and Barke were similar. The concentration of DMS-p in Cheri was different from the other five varieties. It is important to take into account the value of DMS-p in wort, as DMSO, because the concentration of free DMS in the beer may depend on it (KUNZE, 2004; VAN DEN EYNDE, 1991). In fact, the level of SMM in the wort does not influence the production of DMS in the beer because the yeast is not able to produce DMS from SMM (ANNESS and BAMFORTH, 1982).

The beer samples were Pilsner type, and traditionally they should be characterized by a concentration of free DMS around 50-60 µg/L, but not higher than 100 µg/L. Table 8 reports the free DMS concentration of the beer analyzed, which ranged from 38.34 µg/L to 148.61 µg/L. Brewers should pay attention to the level of free DMS as the off-flavour of beer is caused by the free DMS (KUNZE, 2004). This phenomenon may be the consequence of yeast activity during fermentation which produces free DMS from the DMSO.

The beer samples had acceptable concentrations of free DMS: out of the six beers, Chert Scarlett , Regina and Puffin were within limits. Cheri and Regina barley varieties produced beers with low concentrations of free DMS despite the high concentrations of total DMS in the barleys. This means that the brewing process influences the free DMS concentration in the final product. Henley and Barke samples showed a significant concentration of free DMS (138.63 µg/L and 148.61 µg/L, respectively). However, these data are acceptable because they are in the same range of concentration as the threshold. The results about free DMS concentration in beer showed that the six samples were statistically different (p< 0.05), Taking into account the apparent DMS production during fermentation, as the difference between the DMS concentration of the final beer and the wort, just 2 µg/L of DMS were produced by yeast during fermentation of the Cheri wort, while 106 µg/L were produced for the Henley wort. Hence, it was hypothesized that the DMS-p in Cheri wort were mainly represented by SMM, while in Henley wort they were mainly by DMSO. It also seemed that DMS-p in the wort obtained from the autumnal variety malts were mainly constituted by SMM.

All the barley samples showed low levels of total DMS and precursors because DMS is produced from SMM by heating. Therefore, DMS is not present in barley, but it increases during germination. In the majority of malt samples, DMS was constituted by a significant amount of DMS-p. The wort samples in each case had a small concentration of free DMS. At the end of the brewing process, in the final beer, the DMS concentration was mainly represented by free DMS as reported by YANG and SCHWARZ (1998a) and VAN DEN EYNDE (1991).

CONCLUSIONS

The Head-Space Gas Chromatography-Mass Spectrometry (HS-GC-MS) method set up in this study provides an important tool for the study of DMS development during malting and brewing processes. The precision, accuracy and repeatability of the analytical method allows the quality of raw materials to be controlled, as well as the choice of materials in order to produce a better quality beer. It requires routine and versatile equipment, and therefore it can be easily adopted by laboratories worldwide.

The proposed method enables the presence and concentration of DMS to be checked during the brewing process. The screening of different barley varieties demonstrated the applicability of the proposed HS-GC-MS method. Considering DMS concentrations, the results showed that the six varieties studied produced different beers (p<0.05), indicating that the variety has to be considered due to its influence on beer DMS concentration. This confirms data and knowledge already available (YANG and SCHWARZ, 1998a) . The best results in term of free DMS concentration were obtained from the autumn barley samples, Regina and Puffin, probably due to their lower content of nitrogen compounds suggesting that they are suitable for beer production; among the spring varieties Cheri is suggested.

Some of the data reported were presented at the 1st Mass Spectrometry Food Day, 2-3 December 2009, Parma, Italy.

ACKNOWLEDGMENTS

The authors gratefully acknowledge the financial support of the Italian Ministry of University and Research, the Italian Ministry of Agriculture, and the Italian Ministry of Industry.

References

REFERENCES

Analytica-EBC /European Brewery Convention 3.39 "Dimethyl Sulfide and other Lower Boiling Point Volatile Compounds in Beer by Gas Chromatography", EBC Committee: Fachverlag Hans Carl 2000 p. 219. Nurnbergh, Germany.

Anness B.J. and Bamforth C.W. 1982. Dimethyl sulphide-A review. J. Inst. Brew. 88: 244.

Baker C.D. 1988. Determination of dimethyl suplhide in beer by gas chromatography - collaborative trial. J. Inst. Brew. 95: 325.

Bamforth C.W. 1980. Dimethyl sulphoxide reductase of Saccharomyces spp. FEMS Microbiol. Lett. 7:55.

Bamforth C.W. 1998. "Beer: Tap into the Art and Science of Brewing." Plenum Press, New York, Usa.

Blockmans C., Meersche J., Masschelein C.A. and Devreux A. 1981. Proceedings of the European Brewery Convention Congress, p. 347. Copenhagen, Denmark.

Coghe S., Martens E., D'Hollander H., Dirinck P.J. and Delvaux F.R. 2004. Sensory and instrumental flavour analysis of wort brewed with dark specialty malts. J. Inst. Brew. 110: 94.

Drodz J., Havlová P. and Novák V. 1990. Quantitative evaluation of headspace determination of dimethyl sulphide in malt. J. Inst. Brew. 96: 69.

Duphire S. 1999. Determination of DMS and other volatile compounds in beer by headspace capillary gas chromatography and flame ionisation detection. J. Inst. Brew. 105: 265.

Eurachem Guide 1998. "The Fitness for Purpose of Analytical Methods", LGC, Teddington. LTD. Middlesex, UK.

Garza-Ulloa H. 1980. Analytical control of sulphur compounds in beer. A review. Brew. Digest. 55: 20.

Hansen J., Bruun S.V., Bech L.M. and Gjermansen C. 2002. The level of MXR1 gene expression in brewing yeast during beer fermentation is a major determinant for the concentration of dimethyl sulfide in beer. FEMS Y. Res. 2: 137.

Hertel M. and Sommer K. 2008. La bollitura del mosto. Birra e malto 97-98: 4.

Hill P.G. and Smith R.M. 2000. Determination of sulphur compounds in beer using headspace solid-phase microextraction and gas chromatographic analysis with pulsed flame photometric detection. J. Chrom. A. 872:203.

Institute of Brewing. 1982. "Recommended Methods of Analysis", p. 40. London, UK.

Kunze W. 2004. Malt production. Ch. 2. In: "Technology Brewing and Malting". Ch. 11. 3rd International Ed., VLB (Versuch-und Lehranstalt für Brauerei in Berlin), p. 97. Berlin, Germany.

Leppanen O., Denslow J. and Ronkainen P. 1979. Estimation of volatile sulphur compounds in beer. J. Inst. Brew. 85:350

Mebak Rohstoffe 3.1.4.17 Gerste-Rohfrucht-Malz-Hopfen und Hopfenprodukte -Einführung und Validierung von Analysemethoden. Freising, Germany.

Mundy A.P. 1991. The determination of dimethyl sulphide in beer by headspace gas chromatography - A collaborative investigation of precision. J. Inst. Brew. 97: 45.

Narziss L., Miedaner H. and Kattein U. 1985. Behavior of sulfur-containing aroma components during the brewing process. Monatsschr. Brauwiss., 38: 439.

Nykänen L. and Suomalinem H. 1983. "Aroma of Beer, Wine and Distilled Alcoholic Beverages". Ch. 16. Akademie Verlag, p. 233. Berlin, Germany.

Van den Eynde E. 1991. The D.M.S. story from malt to beer. Cerevisiae Biotech. 16: 45.

Vas G. 1997. Characterization of beer sample, using SPME/ capillary GC Analysis. Supelco, 16:7.

Walker B.D. 1991. Estimation of volatile sulphur compounds in beer. J. Inst. Brew. 98: 283.

Walker M.D. and Simpson W.J. 1992. Production of volatile sulphur compounds by ale and lager brewing strains of Saccharomyces cerevisiae. Lett. App. Microbiol. 16; 40.

Yang B. and Schwarz P. 1998a. Factors involved in the formation of two precursors of dimethylsulflde during malting. J. Am. Soc. Brew. Chem. 56: 85.

Yang B. and Schwarz P.B. 1998b. Application of nitrogen-purging of malt extracts to measure two dimethylsulflde precursors by headspace gas chromatography. J. Am. Soc. Brew. Chem. 56:81.

Zhao F.J., Fortune S., Barbosa V.L., McGrath S.P., Stobart R., Bilsborrow P.E., Booth E.J., Brown A. and Robson P. 2006. Effects of sulphur on yield and malting quality of barley. J. Cereal Sci. 43, 369.

Paper received December 31, 2009 Revised paper received November 5, 2010 Accepted December 3, 2010