- Research Note -
Post-bottling Hydrogen Sulfide in Wines:
2,4,6-Trimethyl-1,3,5-Trithiane As a Source
by
C.J. Muller and K.C. Fugelsang
CATI Publication #960303
© Copyright March 1996, all rights reserved
INTRODUCTION
There are few events as potentially disastrous for a winery as the return of bottled wine because of the presence of unacceptable levels of hydrogen sulfide. Its characteristic odor, reminiscent of rotten eggs, can be detected at levels as low as 20 ppb (20 µg/l) in white wines (6), and can be nauseatingly overwhelming at levels ca. 100 ppb. Yet, on occasion, and increasingly as of late, some wines present themselves with excessive hydrogen sulfide. This problem is not new; it has been with the wine industry for a long time. It is indeed perplexing that even with improved winemaking practices the problem has defied solution. Detectable levels of hydrogen sulfide are certain to elicit immediate rejection of affected wines.
The prevailing thought regarding post-bottling hydrogen sulfide is that it develops primarily from elemental sulfur or from sulfur dioxide that is reduced to hydrogen sulfide by the highly reducing conditions present during and after fermentation (2,3). Both of these potential sources have been effectively dealt with by ascertaining a prudent time interval between vineyard application of sulfur and harvest, and, by minimizing use of sulfur dioxide during winemaking. Also, by using yeast strains with less powerful reducing ability (2). Another source is thought to be sulfur-containing amino acids, primarily cysteine, whereby upon the action of pyridoxal on the amino acid, hydrogen sulfide is released (3). However, there is little if any active pyridoxal (or the enzymes that use it as a coenzyme) in finished wines.
Most wines presenting sensorially detectable hydrogen sulfide are whites, although the defect is occasionally found in some reds. Whites, particularly those made in a "pristine" manner, that is from grapes with minimal nitrogen fertilization, partly in an effort to prevent high levels of ethyl carbamate formation, that result in free amino-nitrogen (FAN) levels well below 100 mg/l, and from highly clarified juice with total solids of considerable less than 1% are the most commonly encountered. Usually, these wines have been fermented cold, at 55° F (12.8° C), or less. Typically, this is the wine that, more often than not, is also prone to a protracted or stuck fermentation.
In addition, most of these wines are made with little sulfur dioxide at crush, thus promoting the oxidation of those grape constituents (phenolics) that are easily oxidized. Although some of these oxidized constituents (i.e., quinones) often precipitate and are removed at racking, some will be present in the must throughout fermentation (6). Conditions as described above are ideally suited to provide for the formation of acetaldehyde from a secondary oxidation of ethanol by the quinones during and after fermentation.
Hydrogen sulfide is naturally formed when nitrogen is limiting; where the yeast, in order to survive, must utilize cysteine as a source of nitrogen with the concomitant liberation of hydrogen sulfide (3). Under these conditions, excess acetaldehyde can also be formed by the pyridoxal-catalyzed cleavage of threonine.
Physical laws pertaining to solubility predict that both acetaldehyde and hydrogen sulfide will be more soluble in a cold wine. Thus, the practice of fermenting and storing at low temperatures allows most acetaldehyde and hydrogen sulfide to remain in solution.
But the wine is also acidic; this provides ideal catalytic conditions for the interaction of acetaldehyde and hydrogen sulfide for the formation of 2,4,6-trimethyl-1,3,5-trithiane (hereupon referred to as trithiane), a cyclic thioacetal that has been postulated by us as a source (perhaps the primary source) of post-bottling hydrogen sulfide (4). This is perhaps the reason why hydrogen sulfide is not sensorially apparent in these wines at bottling, inasmuch as the trithiane has an odor reminiscent of "dusty," "earthy," and "nutty," (5) not unlike that used to describe and associated with "corkiness."
We have postulated that post-bottling hydrogen sulfide arises after bottling by the acid-catalyzed hydrolysis of these trithianes (4). We surmise that, in wines, a delicate equilibrium exists among acetaldehyde and hydrogen sulfide (the more volatile constituents) on one side, and trithiane. If bottled wines are allowed to warm, then the trithiane will hydrolyze, liberating hydrogen sulfide (and acetaldehyde, in addition to other as yet unidentified compounds possessing notes reminscent of onion and garlic). It is interesting to note that there is a sensorially significant synergistic effect by concomitant exposure to both hydrogen sulfide and acetaldehyde (Muller, Fugelsang, unpublished data).
In order to test our hypothesis that trithiane(s) might be responsible (at least in part) for post-bottling hydrogen sulfide production, we synthesized both isomers and studied the behavior of each in a model wine solution, and in a white wine.
MATERIALS AND METHODS
SYNTHESIS OF TRITHIANE
There are two isomers of trithiane, (an α-, and a β-form). The following describes our method for the synthesis of each.
Alpha-form: To 100 ml of aqueous 2N HCl cooled to 0° C in an ice bath, 13.2 g (0.03 moles) of freshly distilled acetaldehyde (Aldrich) was added with stirring. To this solution, a slight excess of H2S (Matheson) gas (11-12 g, 0.32 moles) was bubbled while keeping the solution at 0° C. The solution was then cooled overnight to -20° C. A crop of faintly yellow crystals was collected and recrystallized from 2N HCl. The crystals were then washed with cold water to pH 7, and then with cold methanol. The crystals were then dried under a gentle stream of dry nitrogen. Yield: approx. 15.3 g (65 %) of white crystals [m.p. 100-101° C, lit.: 101° C (1)].
Beta-form: As above, 13.2 g (0.3 moles) of freshly distilled acetaldehyde were dissolved in 100 ml of 6N aqueous HCl at 20° C. Hydrogen sulfide was bubbled to the solution and the reaction mixture allowed to stand overnight at room temperature; white crystals of the b-isomer were then recrystallized from 6N HCl. Crystals were washed and dried as above. Yield: 20.1 g (85%) [m.p. 125-126° C, lit.: 125-126° C (1)].
MODEL WINE SOLUTION
A model wine system was prepared by dissolving neutral spirits in distilled water to 12% (v/v) and adjusting to pH 3.4 with tartaric acid. Trithianes, (ca. 0.002 g/l) were added individually and in combination to aliquots of this solution. A portion of the solution was stored at 20° C, and another portion was subjected to accelerated storage by warming to 70° C for 48 hrs in a closed container, followed by cooling to 20° C. Production of H2S from these solutions was determined sensorially by a trained panel, immediately after cooling of the accelerated storage sample, and at monthly intervals for the sample held at 20° C.
WHITE WINE
A wine was made from California State University, Fresno-grown French Colombard grapes following standard industry practice. The wine had no sulfur dioxide added. To this wine, trithianes were added as per the model wine system and production of H2S tested in analogous manner.
TRITHIANE FORMATION
In order to ascertain if trithianes can form under conditions present in the model system, and in the white wine, a 1-liter sample of each was treated with acetaldehyde (0.132 g, 0.003 moles) and an equivalent amount of H2S bubbled through the samples at 12.8° C (ca. 55° F). After allowing the samples to attain equilibrium conditions (ca. 2 weeks), the samples were sensorially examined for any detectable H2S. If no H2S could be detected, the samples were then subjected to accelerated storage at 70° C as described previously, followed by cooling to room temperature. The samples were then submitted to a trained panel to ascertain if H2S had formed.
ANALYTICAL
Routine analyses on model solutions and wine were performed according to the methods described by Zoecklein, et a (6).
RESULTS AND DISCUSSION
I. MODEL SYSTEM
Hydrogen sulfide became apparent (by sniff test) in the model wine solution to which trithiane had been added, but not in the control (without trithiane) after six to eight weeks of storage at 20° C. H2S was also very much apparent immediately after holding the samples under accelerated storage conditions at 70° C. In addition, in both instances of samples containing added trithiane, the presence of odors reminscent of green onion and garlic were also detected.
II. WHITE WINE
Hydrogen sulfide was similarly detected in white wine samples spiked with trithiane that had been held at 20° C, but only after 8 to 12 weeks. The odor associated with "garlicky" notes was present; however, much subdued as compared to the model system samples. These notes were not detected in control wine (without trithiane added). In those samples containing trithiane, there were also other notes described by some panelists as "burnt match"; these notes are often associated with sulfur dioxide; yet the white wine used in this study was made without SO2. The chemical identity of none of these notes has been ascertained by us at this time.
III. FORMATION OF TRITHIANE IN MODEL SYSTEM AND WINE
All panelists were able to recognize the characteristic odor of trithiane in the model system samples in which H2S had been bubbled after adding acetaldehyde. However, not all of the panelists were able to detect the odor of trithiane formed from a similar addition of acetaldehyde and H2S to a white wine.
IV. FORMATION OF H2S FROM MODEL SYSTEM AND WINE
Hydrogen sulfide was unmistakenly detected in both model and white wine treated as described above with both acetaldehyde and H2S when samples of these were subjected to accelerated storage at 70° C. This is not surprising in view of the fact that H2S has a threshold of approximately 20 µg/l and also that any excess dissolved H2S from the reaction would leave the liquid phase and become apparent upon heating.
CONCLUSION
From the result described herein, it is apparent that indeed 2,4,6-trimethyl-1,3,5-trithiane might be a precursor of hydrogen sulfide and thus a potential source of post-bottling hydrogen sulfide in wines. However, as of this writing, we have not attempted to isolate 2,4,6-trimethyl-1,3,5-trithiane from either model systems or from wine to achieve unequivocal chemical/instrumental characterization.
AUTHORS' NOTE
This publication contains preliminary results and has not undergone peer review.
ACKNOWLEDGEMENTS
We thank the California Agricultural Technology Institute (CATI) at California State University, Fresno for its support.
REFERENCES
1. Campaigne, E. Thiones and Thials. Chem. Rev. 39:1-77 (1946).
2. Eschenbruch, R. Sulfite and sulfide formation during winemaking. A review. Am. J. Enol. Vitic. 25:157-161 (1974).
3. Gruenwedel, D.W. and R.K. Patnaik. Release of hydrogen sulfide and methyl mercaptan from sulfur-containing amino acids. J. Agr. Food Chem. 19:775-779 (1971).
4. Muller, C.J. and K.C. Fugelsang. Post-bottling hydrogen sulfide and 'corkiness' - Any relationship? Pract. Winery Vnyrd. March/ April:35-36 (1994).
5. Werkoff, B., W. Bretschneider, R. Emberger, M. Guntert, R. Hopp, and M. Kopsel. Recent developments on the sulfur flavour chemistry of yeast extracts. Chem. Mikrobiol. Technol. Lebensm. 13:30-57 (1991).
6. Zoecklein, B.W., K.C. Fugelsang, B.H. Gump, and F.S. Nury. Wine Analysis and Production. Chapman & Hall, N.Y. 1994.
{ page top }
{ CATI , VERC , VERC - Research Publication }
Copyright © 1996. All rights reserved.
CALIFORNIA AGRICULTURAL TECHNOLOGY INSTITUTE - CATI
College of Agricultural Sciences and Technology
California State University, Fresno
