Reactivity of 3-Mercaptohexanol in Red Wine: Impact of Oxygen, Phenolic Fractions, and Sulfur Dioxide

Reactivity of 3-Mercaptohexanol in Red Wine 115 Reactivity of 3-Mercaptohexanol in Red Wine: Impact of Oxygen, Phenolic Fractions, and Sulfur Dioxide Louis Blanchard, 1 Philippe Darriet, 2 * and Denis Dubourdieu 2 Abstract: 3-Mercaptohexanol formed during alcoholic fermentation from an odorless precursor present in grapes can contribute to the fruity aromas of Cabernet Sauvignon, Cabernet franc, and Merlot wines. The concentrations of this compound decreased during aging as a result of oxygen addition each time the wine was handled. 3- Mercaptohexanol decrease did not result from a direct oxidation by oxygen, as the kinetics of 3-mercaptohexanol disappearance in a red wine supplemented with oxygen was delayed compared to the kinetics of oxygen consumption. The implication of wine phenolic compounds in these mechanisms was studied in a model medium. When catechin was dissolved in the presence of oxygen, the 3-mercaptohexanol content decreased more rapidly than in a solution containing only 3-mercaptohexanol and oxygen. Anthocyanins did limit the decrease in 3-mercaptohexanol. The vital role of sulfur dioxide in protecting 3-mercaptohexanol in the model medium and in wines was shown. Key words: oxidation, 3-mercaptohexanol, phenolic compounds, sulfur dioxide Several hundred volatile substances may be involved in the aroma of a red wine. Most are present in all wines, while some are specific or more abundant in wines of particular grape varieties and may contribute to their varietal aroma. For example, the contribution of 2-methoxy-3-isobutylpyrazine to the herbaceous, green pepper aroma at times detected in Cabernet Sauvignon wines has been clearly demonstrated (Allen et al. 1994). Cabernet Sauvignon and Merlot wines, particularly in the Bordeaux area, may develop complex fruity, minty, meaty, or roasted nuances (Peynaud 1980). Several volatile compounds with a thiol function have been identified in red Cabernet Sauvignon and Merlot wines from Bordeaux: 2-mercaptoethyl acetate (Lavigne-Cruège et al. 1997), 3- mercaptohexanol (3MH), 3-mercapto-2-methylpropanol (Bouchilloux et al. 1998a), 2-methyl-3-furanthiol (Bouchilloux et al. 1998b, Kotseridis et al. 2000), furfurylthiol (Tominaga et al. 2000), and ethyl-2-mercaptopropionate (Blanchard 2000). After a specific method for assaying volatile thiols in white wines was developed (Tominaga et al. 1998), the contribution of 3MH and its acetate to the aroma of Sauvignon blanc wines was clearly established. This assay method was later adapted to red wines (Tominaga et al. 1 Seguin Moreau South-America, Seguin-Moreau South America, Las Quilas 4232, Vitacura, Santiago, Chile; 2 Faculté d Oenologie, Université Victor Segalen Bordeaux 2, 351 Cours de la Libération, 33405 Talence, France. *Corresponding author [Fax : 33(0)54 00 06468; email: philippe.darriet@œnologie.u-bordeaux2.fr] Acknowledgments: The Conseil Interprofessionnel du Vin de Bordeaux (CIVB) and the Conseil Régional d Aquitaine are thanked for funding the research. Manuscript submitted July 2003; revised December 2003 Copyright 2004 by the American Society for Enology and Viticulture. All rights reserved. 2000), making it possible to determine the olfactory impact of these compounds in various Bordeaux wines. It was established that 3MH and furfurylthiol can contribute to the aroma of red Bordeaux wines (Blanchard et al. 1999). More recently, various authors (Murat et al. 2001, Ferreira et al. 2002) have shown that 3MH can play an important role in the fruity aroma of rosé wines obtained from Cabernet Sauvignon, Merlot, and Grenache varieties. In red wines, phenolic compounds, essential for the structure and color of red wines, are particularly reactive with oxygen. These reaction mechanisms play a decisive role in wine development, particularly in stabilizing coloring matter through the formation of cross-linking of tannins and anthocyanins in the presence of ethanal as well as the polymerization mechanisms that result in tannin formation (Haslam 1966, Timberlake et al. 1976, Glories 1978, Escribano-Bailon et al. 1996, Es Safi 1999). Sulfur compounds with thiol function are also highly reactive compounds, easily oxidized to disulfides with metals, particularly iron and copper, at trace concentrations (Jocelyn 1972). Moreover, their nucleophilic properties produce numerous addition reactions, and, in enology, reactions involving nonvolatile or volatile thiols in grape juice with oxidized phenolic compounds have been reported (Singleton et al. 1984, Cheynier et al. 1986, Ribéreau-Gayon et al. 1999a,b). Recently, Murat et al. (2003) have demonstrated the stabilization of a volatile thiol, 3MH, in the presence of anthocyanins in model medium. This article presents the incidence of some technological aspects during red wine aging on the evolution of 3MH concentration and focuses on the impact of oxygen and sulfur dioxide on the 3MH content in the presence of various phenolic fractions isolated from wine. 115

116 Blanchard et al. Materials and Methods Reagents and solutions. Catechin (hydrate form, 98% pure) was provided by Fluka (Sigma-Aldrich-Fluka, Saint Quentin Fallavier, France). The anthocyanin fraction was extracted from 5 L of wine sampled at the end of alcoholic fermentation, using the method described by Glories (1978). Hydrogen peroxide (35% w/v) was obtained from Acros (Noisy le Grand, France). The sulfur dioxide was a 10% (w/ v) potassium bisulfite solution (Laffort, Bordeaux, France). 3-Mercaptohexanol (99% pure) was from Interchim (Montluçon, France). 3-Methylthiohexan-1-ol (99.5% pure) was from Oxford Chemicals (Hartlepool, UK). Wine samples. The wines were from several appellations in the Bordeaux area and were made from Cabernet franc, Cabernet Sauvignon, and Merlot grapes. They were vinified in stainless steel vats for alcoholic fermentation and malolactic fermentation, then poured into barrels after malolactic fermentation, and racked every three months by pumping from one barrel to another. Samples were taken at different stages in the winemaking process (after alcoholic fermentation and malolactic fermentation in stainless steel vats and during aging in oak barrels). Samples were taken from vats as soon as the wine had been homogenized (postfermentation) or from a series of three barrels (during aging); contact with air was avoided. The 3MH content was assayed one week after sampling. During aging, the free sulfur dioxide content in the wine was controlled at 30 mg/l (± 5 mg/l), using the Ripper method (OIV 1990). Samples were stored in corked bottles. 3MH assay in wine and model media. The 3MH content of the red wines was assayed using the method described by Tominaga et al. (2000). Reproducibility of the 3MH assay was better than 5%. After incubations in the model media, the internal standard used in Tominaga et al. (2000) method (1-methoxy-3- mercapto-3-methylbutane:1,3,3-mmb) was not used in the assay due to its possible degradation by peroxides or reaction with oxidized phenolic compounds. An alternative assay using an external standard, 3-methylthio-1-hexanol (0.1 mmol added), was developed according to the protocol proposed by Murat et al. (2003). Repeatability of the 3MH assay, determined by calculation from Murat et al. (2003) results, was better than 7%. Free anthocyanin assay. The free anthocyanins were assayed using the method described by Ribéreau-Gayon and Stonestreet (1965). Racking protocols in the cellar. Three racking modalities were used: racking from one barrel to another (protocol 1); racking as in protocol 1 but with nitrogen in the reception barrel (protocol 2); and racking from one barrel to a small vat and again to the barrel (protocol 3). The control sample was not racked. The oxygen content was measured with a Clark electrode (OxyGuard Handy MKII, Birkeroed, Denmark). Wine sampling and oxygen measurements during kinetic studies. A 5-L sample of red wine, taken at the beginning of aging in vat, was supplemented with oxygen (99.9% purity; Air Liquide, Floirac, France) up to a level of 5 mg/l dissolved oxygen (at 20 C), and placed in a hermetically sealed flask under a nitrogen blanket. Samples were taken using nitrogen counter-pressure. Dissolved oxygen was measured using a Clark electrode, submerged in the wine at all times. In this case, the 3MH assay was carried out immediately after sampling. Constitution of model wines with different phenolic fractions. A model wine with a similar composition to wine (ultrapure water [MilliQ, Millipore]/ethanol [88:12 v/v]; 5 g/ L tartaric acid, ph 3.5) was kept under a nitrogen stream until the dissolved oxygen content was as low as possible. Iron and copper concentrations in the medium were less than 1 µg/l. 3MH at 1000 ng/l was then added to the solution. Partially purified anthocyanins (200 mg/l) or catechin (500 mg/l) were added in the presence (4.7 mg/l dissolved oxygen) or absence of oxygen, with or without sulfur dioxide (30 mg/l). After homogenization, the model wines were placed, under nitrogen atmosphere, into bottles (500 ml), hermetically sealed (first-quality corks), and kept in the dark at 20 C. After one week, each sample was assayed twice for 3MH content and oxygen consumed. Results and Discussion Determination of 3MH content during vinification and barrel aging. The 3MH content in young wines can be considerably higher than the olfactory perception threshold (60 ng/l in model wine with a similar composition to wine) (Tominaga et al. 1998), just after alcoholic and malolactic fermentation (Table 1). The 3MH content was also usually higher than the olfactory perception threshold in older vintages analyzed, indicating that this compound can contrib- Table 1 3-Mercaptohexanol (3MH) content in nine Bordeaux red wines. Controlled appellation a Vintage Grape variety b 3MH (ng/l) Aromatic index c Montagne St Emilion (end AF) 1998 100% Cf 4560 76 Graves (end AF) 1998 100% CS 1730 29 St Estèphe (end MLF) 1998 100% M 850 14 Côtes de Bordeaux (end MLF) 1998 100% CS 1250 20 Graves 1996 80% CS 20% M 660 11 Saint Estèphe 1995 50% M 50% CS 700 11 Pauillac 1993 80% CS 20% M 336 5.6 Saint Julien 1990 80% CS 20% M 400 6.9 Pomerol 1989 100% M <60 <1 a AF: alcoholic fermentation; MLF: malolactic fermentation. b Cf: Cabernet franc; CS: Cabernet Sauvignon; M: Merlot. c Aromatic index was defined as the ratio between the concentration and the olfactory perception threshold of 3MH in model solution.

Reactivity of 3-Mercaptohexanol in Red Wine 117 Table 2 Evolution of 3MH content in red wines during barrel aging (Cabernet Sauvignon, Pessac Leognan, 1996 and 1998) after alcoholic fermentation (AF). Days 3MH after AF (ng/l) C. Sauvignon, 1996 0 6800 15 5780 150 3572 240 2700 330 2100 390 1508 C. Sauvignon, 1998 0 1760 20 1420 60 1183 150 741 240 620 360 478 Table 3 Incidence of oxygen level incorporation on 3MH content in red wines (Cabernet Sauvignon, Pessac Leognan, 1996 and 1998). 1996 1998 3MH Dissolved O 2 3MH Dissolved O 2 Assay (ng/l) (mg/l) (ng/l) (mg/l) Control sample 2900 0.2 986 0.1 Protocol 1 2200 2.2 745 2.2 Protocol 2 2400 0.5 896 0.5 Protocol 3 1100 5.5 366 5.7 ute to the fruity aromas of red Merlot, Cabernet franc, and Cabernet Sauvignon wines. The 3MH content of red Bordeaux wines from the 1996 and 1998 vintages was monitored during barrel aging (Table 2). Concentrations were high at the end of alcoholic fermentation and decreased gradually during the winemaking and aging process. Concentrations at the end of aging were only 22% and 27% of those measured just after alcoholic fermentation. The incidence of oxygen incorporation on 3MH evolution was assessed, particularly during transfer operations (racking). Red wines aged in oak barrels (1996 and 1998 vintages) were subjected to three levels of added oxygen, corresponding to different racking techniques (Table 3). After one week, wines were sampled and 3MH concentrations in wines from different modalities were determined. In oxygenated wines, 3MH content was systematically lower than that of the non-oxygenated wine (Table 3). Protocol 1 (2.2 mg/l dissolved oxygen) induced a decrease between 24 and 25% in the 3MH content; protocol 2 (0.5 mg/l dissolved oxygen) induced a decrease between 9 to 17%. 3MH concentrations were noticeably higher than those in the wine supplemented with the highest quantities of dissolved oxygen in protocol 3, which had a 62% decrease. Impact of oxygen and on reactivity of 3MH with phenolic compounds. To identify the mechanisms responsible for the decrease in 3MH in the presence of oxygen, the evolution kinetics of 3MH and dissolved oxygen in a red wine were monitored in the laboratory (Figure 1). The dissolved oxygen content in the wine decreased rapidly, stabilizing at 0.5 mg/l after 48 hr (Figure 1), while the 3MH content started to decrease only after 48 hr, with a 30% decrease by the end of the experiment (168 hr, 7 days). The kinetics of the decrease in oxygen and 3MH were not the same, as the 3MH content started to decrease only when the oxygen content had stabilized. Therefore, the decrease in 3MH was not directly linked to dissolved oxygen and was probably due to other molecules present in red wines Figure 1 Monitoring the evolution of 3MH and oxygen contents in red wine (Côtes de Bordeaux, 1998) (oxygen initial content, 5 mg/l). that react more readily with oxygen than thiols. Their reactivity in oxidized form might have been responsible for the decrease in 3MH. A model wine containing 1000 ng/l 3MH, was supplemented with different grape phenolic compounds at concentrations close to those found in the wines (Table 4). After seven days, the 3MH content in the control solution without added oxygen had dropped by 33%. The addition of oxygen (4.7 mg/l) caused a further 25% decrease in the 3MH content (a total decrease of 49%), although little oxygen was consumed (0.3 mg/l). In contrast, the addition of (30 mg/l) significantly attenuated the decrease in 3MH, producing an additional decrease of only 8% (a total of 38%), with 0.9 mg/l oxygen consumed in seven days. The standard model wine consistently contained low quantities of dissolved oxygen (approximately 0.2 mg/l), and after seven days 0.05 to 0.1 mg/l oxygen remained in the solution. Volatile thiols have been reported not to oxidize easily in the presence of oxygen, but oxidation may be catalyzed by trace amounts of metal ions (iron, copper) (Jocelyn 1972). The presence of metals at concentrations less than 1 µg/l in the model solution cannot be ruled out. It is therefore probable that the decrease in 3MH in the control was due to the aforementioned reaction mechanisms. These same mechanisms might also explain the decrease in 3MH in the presence of larger quantities of oxygen (4.7 mg/l). In the presence of 10 mg/l hydrogen peroxide, corresponding to the concentration of molecular oxygen initially incorporated in the medium (4.7 mg/l), the 3MH content decreased by over 60% (a total decrease of 73%), compared

118 Blanchard et al. to the solution kept without oxygen for seven days. This result confirms that thiols are easily oxidized by peroxides (Jocelyn 1972). There was a considerably greater decrease in 3MH in the presence of catechin and oxygen (65% decrease; total decrease 76%) than in the absence of oxygen (only a 31% decrease; total decrease 54%). Another assay, performed in duplicate, with a 10-fold 3MH concentration has shown a greater decrease of this compound in the presence of oxygen and catechin than in the presence of oxygen (a 46% decrease with catechin and oxygen versus a 33% decrease with oxygen). Catechin is likely to oxidize to quinone when exposed to air (Haslam 1966, Singleton 1987). o-quinones are powerful oxidizing agents and strongly electrophilic (Singleton 1987) and they can easily react with thiols by a Michael addition reaction (Cheynier et al. 1986). By a series of coupled reactions, o-quinones can also generate peroxides (Timberlake et al. 1976, Wildenradt and Singleton 1974) known for their oxidative properties against thiols. The disappearance of 3MH in the presence of catechin and oxygen could be due to one or the other mechanism or both. Adding in the presence of catechin significantly attenuated the decrease in thiol content (Table 4), which is consistent with both hypotheses. This observation agrees with having a synergistic antioxidant effect on catechin, as indicated by Saucier and Waterhouse (1999), and with antioxidant properties of against peroxides. This decrease in thiol content may also be explained by the fact that reduces quinone forms of catechin, avoiding reaction of the quinone with thiols. The decrease in 3MH content observed in the catechin solution without oxygen might be due to the presence of peroxides or quinone forms of catechin in the commercial extract. As noticed previously, this decrease did not occur when was present (Table 4). The 3MH content in the solution containing 200 mg/l anthocyanin without added oxygen decreased by only 10%, a much lower decrease than in the control solution (Table 4). This finding was also observed by Murat et al. (2003). Adding oxygen induced a total 25 % decrease of 3MH concentration after seven days (Table 4), and in the presence of and oxygen, the 3MH content after seven days was only slightly lower than it was initially (10% total decrease). The addition of anthocyanins attenuated the decrease in volatile thiols, which is paradoxical considering the oxygen consumed in the solution containing anthocyanins (between 2.2 and 2.3 mg/l) but consistent with the antioxidant properties usually associated with anthocyanins (Ghiselli et al. 1998). Adding in the presence of anthocyanins and oxygen attenuated the decrease in the 3MH content that occurred when oxygen was added. It could be due to the antioxidant effect of sulfur dioxide against peroxides potentially produced by oxidation of anthocyanins. Another hypothesis is that reduced the quinonic and semiquinonic forms of cyanidin-3-glucoside and delphinidin-3-glucoside, which might react with the thiols. These results reveal a synergy between and anthocyanins in stabilizing 3MH in the model wine. To complement previous studies, the incidence of free levels on the evolution of 3MH in red wines supplemented with oxygen (5 mg/l) was evaluated. Free content varied between 13 and 40 mg/l (Table 5). After seven days, there was only a limited decrease in the 3MH concentration in wines with a high free content. The decrease was much greater when free content was low (Table 5). In another experiment, ph value was modified in the same red wines by adding potassium hydroxide and 3MH content was measured 7 or 15 days after oxygen incorporation (Table 6). 3MH content consistently decreased in wines at low free content (Table 6). However, no significant decrease in 3MH content may be established, in relation to the ph value of the wines. Table 4 Evolution of 3MH content and dissolved oxygen in model wines with added phenolic fractions, and O 2 after seven days conservation at 20 C. Sample 3MH Initial dissolved O 2 O 2 consumed (ng/l) (mg/l) (mg/l) Immediate Control 1000 0.2 nd a After seven days Control 670 0.2 0.1 Sample + O 2 507 4.7 0.3 Sample + O 2 + 617 4.7 0.9 Sample + H 2 O 2 269 nd nd Sample catechin 459 0.2 0.1 Sample catechin + 668 0.2 0.1 Sample catechin + O 2 238 4.7 1.4 Sample catechin + O 2 + 647 4.7 1.2 Sample anthocyanin 887 0.2 0.1 Sample anthocyanin + O 2 750 4.7 2.2 Sample anthocyanin + O 2 + 883 4.7 2.3 a nd: not determined. Table 5 Effect of free levels on 3MH concentration in red wines in the presence of dissolved oxygen after seven days. Storage time Free 3MH in Initial 3MH (days) (mg/l) wine (ng/l) content (%) Wine 1 0 15 800 100 Wine 1 7 40 750 94 Wine 1 7 15 520 47 Wine 2 0 13 1050 100 Wine 2 7 40 1000 95 Wine 2 7 13 550 52 Wine 3 0 22 650 100 Wine 3 7 40 490 75 Wine 3 7 22 340 52

Reactivity of 3-Mercaptohexanol in Red Wine 119 Table 6 Effect of free level and ph on 3MH concentration in red wines in the presence of dissolved oxygen after 7 days (wine A) or 15 days (wine B). Storage time Free ph 3MH in Initial 3MH (days) (mg/l) wine (ng/l) content (%) Sample control 0 40 3.5 1640 (no oxygen incorporated) Sample A1 7 40 3.5 1060 65 Sample A2 7 25 3.5 650 40 Sample A3 7 40 4.0 950 58 Sample A4 7 25 4.0 590 36 Sample control 0 40 3.5 700 (no oxygen incorporated) Sample B1 15 40 3.5 524 72 Sample B2 15 15 3.5 238 32 Sample B3 15 40 4.0 330 46 Sample B4 15 15 4.0 143 20 These findings suggest the important role played by sulfur dioxide, as it protects the volatile thiols of red wines, and highlight antioxidant properties of described by Pontallier (1981) and Laborde (1987) in relation to the varietal aroma components of a wine. Conclusion Concentrations much higher than the olfactory perception threshold of 3-mercaptohexanol, a thiol of varietal origin, were found frequently in Merlot, Cabernet franc, and Cabernet Sauvignon wines at the end of alcoholic fermentation. These concentrations decreased during malolactic fermentation and barrel aging. By the end of aging, the wines contained only a low percentage of the 3MH formed during alcoholic fermentation. Oxygen dissolved in the wine during various handling operations led to a decrease in the 3MH content of the red wine. However, this decrease of 3MH was delayed relative to that of oxygen consumption in the wine. Preliminary results indicate that phenolic fractions derived from grapes (anthocyanin and catechin) influence the evolution of 3MH content in a model solution in the presence of oxygen. The disappearance of 3MH with catechin in the presence of oxygen was demonstrated as well as the fact that this mechanism can be partially inhibited by adding sulfur dioxide. A synergistic effect of sulfur dioxide and anthocyanins in the stabilization of 3MH was found. In a model wine, the combination of anthocyanins and sulfur dioxide reduced the oxidative decrease of 3MH. Results also confirmed the importance of sulfur dioxide in winemaking, particularly its protective effect in relation to the oxidation phenomena responsible for the decrease in 3MH. A high level of free diminished the decrease in 3MH during handling operations, thus preserving the fruity aromas in red wines provided by this compound. Literature Cited Allen, M.S., M.J. Lacey, and S. Boyd. 1994. Determination of methoxypyrazines in red wines by stable isotope dilution gas chromatography-mass spectrometry. J. Agric. Food Chem. 42:1734-1738. Blanchard, L. 2000. Recherche sur la contribution de certains thiols à l arôme des vins rouges: Etude de leur genèse et de leur stabilité. Thesis, Université Victor Segalen Bordeaux 2. Blanchard, L., P. Bouchilloux, P. Darriet, T. Tominaga, and D. Dubourdieu. 1999. Caractérisation de la fraction volatile de nature soufrée dans les vins de Cabernet et de Merlot. Etude de son évolution au cours de l élevage en barrique. In Sixth Symposium International d Œnologie de Bordeaux. A. Lonvaud (Ed.), pp. 501-505. Tec & Doc, Paris. Bouchilloux, P., P. Darriet, R. Henry, V. Lavigne Cruège, and D. Dubourdieu. 1998a. Identification of volatile and powerful odorous thiols in Bordeaux red wine varieties. J. Agric. Food Chem. 46:3095-3099. Bouchilloux, P., P. Darriet, and D. Dubourdieu. 1998b. Identification du 2-methyl-3-furanthiol dans les vins. Vitis 37:177-180. Cheynier, V., E.K. Trousdale, V.L. Singleton, M. Salgues, and R. Wylde. 1986. Characterization of 2-S-glutathionyl caftaric acid and its hydrolysis in relation to grape wines J. Agric. Food Chem. 34: 217-221. Escribano-Bailon, M.T., O. Dangles, and R. Brouillard. 1996. Coupling reaction between flavylium ion and catechin. Phytochemistry 35:499-505. Es Safi, N.E., H. Fulcrand, V. Cheynier, and M. Moutounet. 1999. Studies on the acetaldehyde-induced condensation of (-)-epicatechin and malvidin-3-o-glucoside in a model solution system. J. Agric. Food Chem. 47:2096-2102. Ferreira, V., N. Ortin, A. Escudero, R. Lopez, and J. Cacho. 2002. Chemical characterization of the aroma of Grenache rosé wines: Aroma extract dilution analysis, quantitative determination, and sensory reconstitution studies. J. Agric. Food Chem. 50:4048-4054. Ghiselli, A., M. Nardini, A. Balbi, and C. Scaccini. 1998. Antioxidant activity of different phenolic fractions separated from an Italian red wine. J. Agric. Food. Chem. 46:361-367.

120 Blanchard et al. Glories, Y. 1978. Recherche sur la matière colorante des vins rouges. Thesis, Université Bordeaux 2. Haslam, E. 1966. Chemistry of Vegetable Tannins, pp. 81-90. Academic Press, London. Jocelyn, P.C. 1972. Biochemistry of the SH Group, pp. 94-111. Academic Press, London. Kotseridis, Y., and R. Baumes. 2000. Identification of impact odorants in Bordeaux red grape juice, in the commercial yeast used for its fermentation, and in the produced wine. J. Agric. Food. Chem. 48:400-406. Laborde, J.L. 1987. Contribution à l étude des phénomènes d oxydation dans les vins rouges. Rôle de l anhydride sulfureux. Thesis, Université de Bordeaux 2. Lavigne-Cruège, V., R. Henry, and D. Dubourdieu. 1997. Identification et dosage de composés soufrés intervenant dans l arôme grillé des vins. Sci. Aliments 18:175-191. Murat, M.L., T. Tominaga, and D. Dubourdieu. 2001. Mise en évidence de composés clefs dans l arôme des vins rosés et clairés de Bordeaux. J. Int. Sci. Vigne Vin 35:99-105. Murat, M.L., T. Tominaga, C. Saucier, Y. Glories, and D. Dubourdieu. 2003. Effect of anthocyanins on stability of a key odorous compound, 3-mercaptohexan-1-ol, in Bordeaux rosé wines. Am. J. Enol. Vitic. 54:135-138. Peynaud, E. 1980. Le goût du vin. Dunod, Paris. Pontallier, P. 1981. Recherche sur les conditions d élevage des vins rouges. Rôle des phénomènes oxydatifs. Thesis, Université de Bordeaux 2. Office International de la Vigne et du Vin. 1990. Recueil des méthodes d analyses. OIV, Paris. Ribéreau-Gayon, P., and E. Stonestreet. 1965. Dosage des anthocyanes dans les vins rouges. Bull. Soc. Chim. 2649-2652. Ribéreau-Gayon, P., D. Dubourdieu., B. Donèche, and A. Lonvaud. 1999a In Handbook of Enology. Vol. 1. The Microbiology of Wine and Vinifications, pp. 380-382. Wiley & Sons, London. Ribéreau-Gayon, P., Y.Glories, A. Maujean, and D. Dubourdieu. 1999b. In Handbook of Enology. Vol. 2. The Chemistry of Wine Stabilizations and Treatments, pp. 198-200. Wiley & Sons, London. Saucier, C., and A.L. Waterhouse. 1999. Synergic activity of catechin and other antioxidants. J. Agric. Food Chem. 47:4491-4494. Singleton, V.L. 1987. Oxygen with phenols. Am. J. Enol. Vitic. 38:69-77. Singleton, V.L., J. Zaya, E.K. Trousdale, and M. Salgues. 1984. Caftaric acid in grapes and the conversion to a reaction product during processing. Vitis 23:113-120. Timberlake, C.F., and P. Brindle. 1976. Interaction between anthocyanins, phenolic compounds, and acetaldehyde and their significance in red wines. Am. J. Enol. Vitic. 27:97-105. Tominaga, T., M.L. Murat, and D. Dubourdieu. 1998. Development of a method for analyzing the volatile thiols involved in the characteristic aroma of wines made from Vitis vinifera L. cv. Sauvignon blanc. J. Agric. Food Chem. 46:1044-1048. Tominaga, T., L. Blanchard, P. Darriet, and D. Dubourdieu. 2000. A powerful aromatic volatile thiol, 2-furanmethanethiol, exhibiting roast coffee aroma in wines made from several Vitis vinifera grape varieties. J. Agric. Food Chem. 48:1799-1802. Wildenradt, H.L., and V.L. Singleton. 1974. The production of aldehydes as a result of oxidation of polyphenolic compounds and its relation to wine aging. Am. J. Enol. Vitic. 25:119-126.