Abstract
Following an experimental design replicating typical winery conditions, a Riesling wine was bottled with different headspace oxygen levels and sealed with either a coextruded closure or a screwcap to investigate the impact of headspace oxygen and closure oxygen transfer rate on wine evolution. Using luminescence technology, dissolved oxygen and headspace oxygen, as well as oxygen ingress through the closure, were monitored during 24 months of bottle storage. Under typical winery conditions, headspace oxygen introduced at bottling was found to be a major component of oxygen in bottled wine. Headspace oxygen at bottling influenced loss of sulfur dioxide during bottle storage, being the main cause of sulfur dioxide decline during the first four months after bottling in 375 mL bottles. The loss of sulfur dioxide was not correlated with the evolution of dissolved oxygen, but with the total amount of oxygen consumed by the wine. After 24 months in the bottle, color differences due to different headspace oxygen and closure oxygen transfer rate were generally minor. Conversely, differences in closure oxygen transfer rate were responsible for significant differences in the final concentration of the off-odor compound hydrogen sulfide, with screwcap generally associated with higher levels of this compound. Even if less significantly, the amount of oxygen present in the headspace at bottling also had an effect on final hydrogen sulfide, with higher concentrations observed in wines bottled with lower headspace oxygen.
Oxygen is a key reactant that can change chemical and sensory properties of wine (Ribereau-Gayon 1933, Wildenradt and Singleton 1974, Du Toit et al. 2006). In bottled wine, oxygen derives from the bottling process or enters the package during storage. During bottling operations, contact of the wine with air (e.g., during bottle filling) can result in increased levels of dissolved oxygen (DO) (Kielhöfer and Würdig 1962, Perscheid and Zürn 1978, Kettern 1985, Schneider 2005). Additionally, gaseous oxygen present in bottle headspace (HS) is another major source of oxygen, which can vary depending on HS volume and HS management technology, such as evacuation or inerting (Müller-Späth 1977, Kettern 1985, Schneider 2005). Subsequently, during storage, oxygen ingress through the closure, as determined by the oxygen transfer rate (OTR), is responsible for additional oxygen uptake (Perscheid and Zürn 1978, Schneider 2005, Müller-Späth 1977, Skouroumounis et al. 2005a, Lopes et al. 2006).
Several studies have investigated the role of oxygen ingress through the closure on wine development. The general conclusion, which is now widely accepted in the wine industry, was that after a certain time of storage, wines bottled with closures with different OTR will exhibit different chemical and sensory characteristics (Skouroumounis et al. 2005a, 2005b, Lopes et al. 2006, Godden et al. 2002, Hart and Kleinig 2005, Lopes et al. 2009, O’Brien et al. 2009). In particular, after three years of storage, a Semillon wine sealed with screwcap retained higher sulfur dioxide (SO2) and showed lower browning compared to the same wine sealed under synthetic closures, natural corks, and technical corks (Godden et al. 2002). In addition, wines that retained higher concentrations of SO2 also had higher scores in citrus and fruity aroma. However, wines also had unwanted reductive character. Similar conclusions were proposed later for Riesling and Chardonnay wines (Skouroumounis et al. 2005a, 2005b) and for red wines (Hart and Kleinig 2005). More recently, it has been suggested that an optimal degree of oxygen exposure should be identified in order to prevent formation of reductive off-flavors such as hydrogen sulfide (H2S) without incurring an excessive loss of fruitiness due to oxidation (Lopes et al. 2009, O’Brien et al. 2009).
Only a limited number of studies have investigated the contribution of different levels of oxygen at bottling, such as DO and particularly HS oxygen. It was suggested that, in addition to OTR, variations in DO at bottling as well as in HS oxygen could be responsible for chemical and sensory differences following bottle storage (Lopes et al. 2009). However, different HS volumes, and therefore levels of HS oxygen, had no influence on the reduced characters of a Semillon wine 24 months postbottling (Godden et al. 2005). Brajkovich et al. (2005) also did not find significant differences in aroma and SO2 concentrations between different HS volumes in Sauvignon blanc wine. In a study on the impact of different HS volumes on a Cabernet Sauvignon wine sealed by screwcap, after 12 months wines bottled with higher HS volume had lower SO2 concentrations and darker color than those with lower HS volume (Kwiatkowski et al. 2007). However, these were extreme situations, with 4 mL and 64 mL HS volume, respectively.
In the above-mentioned studies HS oxygen was not measured, so a comprehensive rationalization of the different results is not possible. In addition, HS oxygen was generally adjusted by changing wine volume in the bottle, while in the winery HS oxygen is often adjusted by applying a flush of inert gas to remove oxygen. In this sense, the impact of HS oxygen on wine evolution is unknown and the question remains open as to whether HS volume and HS composition are equivalent in terms of wine development, given that changes in the concentration of oxygen in the headspace can affect oxygen diffusion in the wine. In common industry practice, screwcaps are bottled with a larger headspace and cylindrical closures are bottled with a smaller headspace. In addition, no study has been carried out to date to investigate whether oxygen at bottling is equivalent to oxygen delivered during storage through the closure and whether different types of closures require different types of HS management.
The aim of this study was to investigate the influence of HS volume and composition on wine development during bottle storage and to study the effects of combining different HS oxygen at bottling with closures with different OTR. An experimental design replicating typical winery conditions was adopted. Oxygen evolution inside the bottles was accurately measured using a luminescence-based technology (Dieval et al. 2009, Nygaard et al. 2009) in a first attempt to describe the correlation existing between oxygen consumption and wine chemical changes during bottle storage.
Materials and Methods
Wine.
Approximately 1000 L of Riesling wine (Rheingau region, vintage 2007) was vinificated at Geisenheim Research Center. The fermentation took place in stainless-steel tanks between 18 and 22°C and bentonite fining was performed four months later. The wine was stored in a tank with no ullage until bottling. Analytical parameters of the wine at bottling were as follows: 12.7% alcohol, 9.7 g/L sugar, pH 3.33, 7.1 g/L acidity, 54 mg/L free SO2, 135 mg/L total SO2, and 0.3 mg/L DO.
Bottles and closures.
A coextruded (Co) synthetic closure (Nomacorc Classic, 43 x 22 mm; Nomacorc SA, Thimister Clermont, Belgium) and a screwcap (Sc) closure (CSI 5SE, 28 x 15 mm, PVC-free; Alcoa, Kelkheim, Germany) were used. The bottles were colorless, Saint Gobain 0.375 L glass (375 mL at 52 mm for the cylindrical closure finish and at 28 mm for the screwcap finish).
Oxygen measurements.
Dissolved oxygen of the wine prior to bottling was measured in the tank with a Fibox 3-Trace fiber-optic oxygen meter coupled to an oxygen dipping probe (PreSens GmbH, Regensburg, Germany). Dissolved oxygen and HS oxygen were measured with the Fibox 3-Trace fiber-optic oxygen meter coupled to Pst3 oxygen sensors (linearity range from the manufacturer: 0 to 50% oxygen). Five bottles fitted with two oxygen sensors each were used for each modality. The sensors were glued inside the bottle with silicone (RS Components, Mörfelden-Walldorf, Germany) at a height of 8 cm from the bottom of the bottle for DO measurement and in the headspace for HS oxygen measurement. These bottles were equally distributed in the bottling line during bottling (approximately one bottle with sensors every 30 bottles). Manufacturer calibration was used for all the sensors, with no further calibration. All oxygen measurements in the bottle were taken right after bottling, weekly during the first two months of storage, and at 4, 8, 10, 14, and 24 months after bottling (without shaking of the bottles).
Bottling.
The bottling set up is shown (Figure 1). Two HS volumes, 6 mL (HS6) and 18 mL (HS18), were applied for each closure. For each HS volume, oxygen content was adjusted by CO2 flushing, with the highest HS oxygen obtained with no flushing, and therefore the headspace consisted of ambient air. Medium (Med) and High levels of HS oxygen were obtained for each closure. For the coextruded closure, an extra set of wines was prepared with lower HS oxygen (Low) to study the effect of high inerting on coextruded closures. The HS oxygen was measured in hPa, converted into mg in headspace, and then to potential mg/L in wine, taking into account the wine volume (376 mL for the cylindrical closure HS6, 365 mL for the cylindrical closure HS18, and 379 and 367 mL for the screwcap, respectively). The bottling line included a plate filter (Seitz Enzinger Noll, Worms, Germany), a 100 L filler tank, and a three-head manual filler (KTM-Troxler, Ettenheim-Münchweier, Germany), a semiautomatic corking machine with vacuum (GAI 4040, Prospero Equipment Corp., Pleasantville, NY), an industrial corker with CO2 flushing (Seitz Enzinger Noll), and a semiautomatic screwcap machine (VAW Aluminium AG, Grevenbroich Germany). The bottles were rinsed with SO2 and sterile water directly before bottling (Stroh GmbH, Schloßböckelheim, Germany).
The filling level was adjusted manually using a silicone stopper fitted with two adjustable tubes. One tube was adjusted to the desired filling height, the other was set shorter, to be in the headspace. After each bottle was filled, the silicon stopper was held for some seconds on the top of the bottle neck closing the headspace, so that the first tube was immersed in the wine as deep as the desired filling height. By applying a gas pressure through the second tube, excess wine was pushed upward and removed from the bottle neck via the second tube, which was connected to a vacuum pump for increased efficiency. The gas pressure was obtained by air for the High HS oxygen, whereas CO2 (purity 100%; Air Liquid, Düsseldorf, Germany) was used for the Med oxygen level to obtain increased removal of HS oxygen. For the Low oxygen level (only cylindrical closures), filling height was adjusted by CO2 pressure and residual oxygen was removed by flushing the headspace directly with CO2. Bottles were then sealed with one of the two closure types. In no case was vacuum applied at closure insertion. All bottles were stored upright in the storage room of the cellar in Geisenheim Research Institute at 14 to 16°C and 55% humidity.
Following bottling, DO was 1.08 ± 0.15 mg/L, confirming consistency across the different wines. Given that DO before bottling was 0.3 mg/L, DO increase due to bottling was ~0.8 mg/L, which is consistent with other studies (Kielhöfer and Würdig 1962, Perscheid and Zürn 1978, Lopes et al. 2009) and indicates a well-controlled process.
Determination of total consumed oxygen.
Total consumed oxygen (TCO) at each time point was calculated as the sum of oxygen present at bottling (HS plus DO), plus the oxygen entering the bottle through the closure during storage, minus HS and DO oxygen measured at each time point, the latter accounting for any residual oxygen not consumed by the wine. The result obtained was the total amount of oxygen consumed by the wine. In order to quantify oxygen ingress through the closure during time, 10 bottles for cylindrical closure and 10 for screwcap were fitted with one Pst6 oxygen sensor each (linearity range 0 to 4.2% oxygen) at 8 cm from the bottom. These empty bottles were filled with pure nitrogen (purity 99.8%; Air Liquid) until complete removal of the oxygen present and then were directly sealed with one of the closures used. Measurements of oxygen were taken at the time points previously indicated to quantify oxygen ingress through the closure in the course of time.
Sulfur dioxide measurements.
Duplicate measurement of free and total SO2 was carried out on three bottles from each treatment right after bottling, weekly during the first month of storage, and after 4, 10, 14, and 24 months of storage. The measurements were carried out by flow injection analysis with a FIAstar 5000 Analyzer (Foss, Rellingen, Germany). Calibration of the analyzer was done before each set of measurements using a set of 10 known solutions (0, 10, 20, 30, 50, 100, 120, 150, 180, and 200 mg/L SO2).
Color measurements.
The CIELab parameters L*, a*, and b* were calculated using the ASTM software (Standards on Color and Appearance Measurement, West Conshohocken, PA) after scanning samples of three bottles per treatment (same bottles used for SO2 analysis) in a CADAS 200 spectrophotometer (1 cm cuvette; Hach Lange GmbH, Düsseldorf, Germany).
Hydrogen sulfide.
Hydrogen sulfide (H2S) was analyzed after 24 months of storage by gas chromatography (GC) coupled with a pulsed flame photometric detector (PFPD), using static headspace sampling. A 6890 gas chromatograph (Agilent Technologies, Santa Clara, CA) was used, equipped with a MPS 2 headspace sampler (Gerstel, Mülheim an der Ruhr, Germany), a CIS-4 cooled injection system (Gerstel), and a 5380 PFPD (OI Analytical, College Station, TX). Chromatographic separations were performed on a SPB-1 sulfur column (30 m x 0.32 mm i.d., 4 μm film thickness; Supelco, Sigma-Aldrich, Munich, Germany). Analytical conditions were as follows: injector temperature program, −100°C, 12°C/sec until 40°C for 1 min, then 12°C/sec until 180°C for 8 min; oven temperature program, 29°C for 7 min, 10°C/min until 180°C for 10.5 min. Helium was used as carrier gas. Detector temperature was 250°C. Analyses of the samples were carried out in duplicate. More details on sample preparation and analytical parameters can be found elsewhere (Rauhut et al. 1998, 2005, Irmler et al. 2008).
Statistical analyses.
Analysis of variance and Tukey tests (0.05 significance) were carried out using SPSS 15.0 (IBM SPSS, Chicago, IL). Regression analysis and Pearson’s correlation test (0.05 significance) were conducted with XLSTAT 2010 software (Addinsoft Deutschland, Andernach, Germany).
Results and Discussion
Headspace oxygen.
Headspace oxygen data at bottling (Figure 1) showed that different levels of HS oxygen were created by CO2 flushing of the headspace. Differences in HS oxygen between the modalities with same closure and HS volume were statistically significant (p < 0.05). Although similar values of HS oxygen were obtained by different combinations of HS volumes and degree of HS inerting (e.g., Co/HS6/High and Co/HS18/Med), HS volume has to be consistently considered in conjunction with expansion of the wine inside the bottle. Indeed, too small of HS volumes could increase the risk of wine leakage. Therefore inerting a larger headspace could be a safer solution for HS oxygen management than reducing HS volume. In addition, high inerting a small headspace (Co/HS6/Low) was overall the most effective approach to reduce HS oxygen levels (Figure 1).
The HS volumes of 6 mL and 18 mL represent typical industry settings for cylindrical and screwcap closures, respectively (Paine 1991, Schneider 2005). However, as 375 mL bottles were used here, the levels of oxygen contained in these headspaces (expressed in mg/L wine) would be 50% lower if 750 mL bottles were used. When the values are calculated for 750 mL bottles, final values of 0.2, 1.4, and 3 mg/L wine in bottles with cylindrical closure (HS6), and 2.5 and 7.3 mg/L wine in screwcap bottles (HS18) were obtained for the three inerting levels. It can be concluded that the range of HS oxygen concentrations in this study is similar to that found in other studies (Vidal et al. 2004, O’Brien 2009). Therefore, although obtained in an experimental setup, the observations of this study provide meaningful indications regarding bottling management in large-scale winemaking.
Headspace oxygen decreased during storage in all treatments because of its dissolution into the wine and consequent consumption through different chemical reactions (Figure 2A, B). The time required for HS oxygen to become undetectable varied between two weeks and less than eight months depending on initial concentration and closure type; the two extremes were Co/HS6/Low and Sc/HS18/High with 0.4 mg/L wine and 14.5 mg/L wine initial HS oxygen, respectively. However, in the majority of the cases, HS oxygen was completely consumed within one to three months.
Dissolved oxygen.
Starting from a DO after bottling of 1.08 ± 0.15 mg/L across all modalities, a decrease in DO was observed over the entire experiment (Figure 2C, D), with complete consumption occurring within one to eight months, depending on initial HS oxygen. In no case did DO increase again once it reached zero (data not shown). Changes in DO reflect the net balance between dissolution of oxygen from the headspace into the wine and consumption by the wine itself. Therefore, in general, the latter was higher than the former during the timeframe of the study. In samples with lower HS oxygen, DO started decreasing immediately after bottling, whereas in samples with higher HS oxygen, an early transient increase in DO was observed. This suggests that, during the early stages of bottles storage, in these samples the rate of oxygen dissolution into the wine was higher than consumption. In general, the magnitude of this DO increase was proportional to the concentration of oxygen in the headspace. Among the concentrations of HS oxygen tested, a critical concentration of 3.9 mg/L wine was identified as the threshold value of HS oxygen causing this transient DO increase (Sc/HS6/High). Below this value of initial HS oxygen, no DO increase was observed with a HS oxygen of 2.9 mg/L wine (Co/HS6/Med), suggesting that for this HS oxygen the rate of consumption was higher than that of dissolution. Therefore, it is likely that, in terms of DO increase, the actual critical HS oxygen was between 2.9 m and 3.9 mg/L wine for the particular wine tested and the size of the bottles used in this study.
Although an increase in DO was observed for both closures, there was a difference between coextruded closures and screwcaps. Coextruded closures with high and medium initial HS oxygen showed an increase of DO during the first two days after bottling (Figure 2C), after which DO started to decrease because of oxygen consumption. Conversely, in screwcap samples, DO initially decreased and only after ~10 days postbottling was a decrease observed. This difference could be partly due to extra pressure generated in the headspace by cylindrical closures, which could have accelerated the dissolution of oxygen into wine (Kielhöfer and Würdig 1962). In addition, the differences in wine–headspace contact surfaces have to be considered. In the case of cylindrical closures, HS18 resulted in a filling height below the bottle shoulder, and therefore in a larger wine–headspace contact surface, which probably favored rapid dissolution of oxygen in the wine. In the case of coextruded closures, high HS treatments were only introduced for comparison (as they are not commercially realistic), whereas both HS volumes are in theory possible for screwcap closures.
Another interesting difference between the two types of closures is that, despite similar values of initial HS oxygen (for example Co/HS6/Med and Sc/HS6/Med), complete consumption of initial DO was longer with cylindrical closures. Elsewhere, a faster DO drop in white wine was measured under screwcap than under natural cork (Perscheid and Zürn 1978). These findings reflect the fact that, due to their porous nature, cylindrical closures contain air, and therefore oxygen (Jung and Zürn 2000, Lopes et al. 2007, Ugliano et al. 2011), which is in part released gradually into the headspace following closure insertion in the bottle. In addition, ingress of oxygen through the closure, although limited at this stage, was higher for the coextruded closures used here.
Free and total SO2.
The evolution of free SO2 in the different treatments during storage was determined (Figure 3). Similar profiles were obtained for total SO2 (data not shown). Consistent with other studies (Godden et al. 2005, Brajkovich et al. 2005, Kwiatkowski et al. 2007, Lopes et al. 2009), a rapid decrease in free SO2 was observed early in the study. Differences between treatments became significant in the first week (p < 0.05), and wines with higher HS oxygen showed a more rapid decline of free SO2. Differences reached their maximum at about four months, and remained mostly unchanged for the following 20 months of bottle storage. At 10 months, the evolution of SO2 was not significantly affected by initial HS oxygen (p < 0.05), while closure OTR started to be the main source of differences between the modalities (OTR data presented later). In general, ~55% of the global free SO2 decrease under coextruded bottles and 80% under screwcap bottles occurred during the first four months, confirming that the greatest loss of free SO2 during bottle storage is associated with the oxygen present at bottling, which is largely HS oxygen. Therefore, management of HS oxygen at bottling allows great control of SO2 decline during bottle storage and, consequently, during wine shelf life.
Following the first four months of storage, wines sealed with screwcaps showed virtually no further loss of free SO2, consistent with the very low OTR of this type of closure. Conversely, a further decline was observed for cylindrical closures, given their permeability to oxygen (Figure 4). During the first four months, when free SO2 evolution was mainly dependent on initial headspace, losses of free SO2 were consistently smaller when lower HS oxygen (and smaller HS volume) was applied. In the following 20 months, initial HS oxygen had a minor impact on the decline of free SO2, while closure OTR was the main modulator of SO2 concentration. From these results it is clear that management of both HS oxygen at bottling and OTR offer the potential to control, to a certain extent, the decline of free SO2 during bottle storage. However, when the HS contained higher oxygen, sealing with extremely low OTR closures such as screwcap did not prevent significant loss of SO2 early in wine life (Sc/HS18/High; Figure 3). On the contrary, removal of oxygen by inert gas flushing significantly reduced initial SO2 loss, even for closures with higher OTR (Co/HS6/Low). In this study, CO2 inerting of the headspace reduced the loss of free SO2 up to 17 mg/L in the first four months of storage (Co/HS18/Low vs. Co/HS18/High). At this point it should be emphasized that the wine volume in the bottles was half the volume typically obtained in practice (375 instead of 750 mL), whereas the HS volume, the HS oxygen, and the oxygen ingress through the closure were realistic for 750 mL bottles. To that effect, the time of storage in these smaller bottles corresponds to twice as much as in larger bottles.
Another interesting observation was that, in the first four months, the greatest improvement in free SO2 loss was observed when inerting was applied to screwcap closures with larger headspace, which is the typical industry setting for this type of closure. This highlights the importance of management of HS oxygen for screwcap closures and, in general, in situations where larger HS volumes are used, as they involve higher levels of oxygen.
Correlation between SO2 evolution and oxygen.
A Pearson’s correlation test was carried out to explore the correlation between SO2 loss and decline of DO (Table 1). A correlation coefficient higher than 0.7 could only be obtained for the first two weeks of bottle storage, while at four months the correlation was already lower. Although our data is consistent with the empirical observation that, at least in the presence of high concentrations of available oxygen (i.e., HS oxygen or DO), a rapid decline of SO2 is observed (Brajkovich et al. 2005, Kwiatkowski et al. 2007, Lopes et al. 2009), the direct reaction between oxygen and SO2 is extremely slow under wine conditions (Waterhouse et al. 2006). Conversely, SO2 loss is linked to oxygen through reaction of SO2 with the products of wine oxidation, in particular hydrogen peroxide (Danilewicz et al. 2008). Additionally, the DO measurement only reflects the oxygen that is present in the wine at any given time. In a highly reactive environment such as wine, it reflects the occurrence of any excess oxygen that the wine has not yet consumed at that time point. Therefore, once DO reaches a value of zero, consumption of oxygen is still taking place, but cannot be quantified, explaining why a good correlation between SO2 loss and DO was observed only at the early time points.
In order to overcome this limitation, total consumed oxygen (TCO) was calculated as the sum of the oxygen present at bottling (HS oxygen plus DO), plus the oxygen entering the bottle through the closure during storage, minus DO and HS oxygen measured at each time point. Oxygen ingress through the closure was measured by quantifying oxygen ingress in bottles filled with nitrogen, sealed with the closures used in the study, and stored under the same conditions as the experimental wines. In total, during 24 months, 13 ± 1 mg/L oxygen was delivered by the synthetic closure and 1 ± 0.1 mg/L by the screwcap closure. The amount of oxygen entering the bottles sealed with the closure with higher OTR (13 mg/L) is similar to the initial HS oxygen in a screwcap bottle with large headspace (Sc/HS18/high, 14.8 mg/L wine), a setting that is commonly used in the industry for a closure with lower OTR. Clearly, bottling can be a key component of the total oxygen in the bottle.
The TCO obtained 24 months after bottling ranged from 5 to 25 mg/L (Figure 5). Headspace oxygen right after bottling ranged from 0.4 to 14.5 mg/L wine, indicating the contribution of the HS oxygen reservoir to total consumed oxygen. Correlation analysis confirmed that TCO was well correlated with SO2 decline throughout bottle storage (Table 1). Calculation of TCO can therefore provide a valuable tool to predict SO2 loss during bottle storage and, therefore, to estimate the shelf life of wine.
Color measurements.
Wine absorbance at 420 nm (abs420) is commonly used to assess the degree of wine color oxidation. After 24 months of storage, absorbance values were consistently less than 0.097 (Table 2), indicating that the degree of oxidation in the wines was relatively low, although several statistically significant differences were observed. Data collected at different times during storage showed that HS oxygen only affected color development during the first four months (p < 0.05). From the subsequent time point (10 months), differences in color development were due to OTR (p < 0.05). CIELab analysis coupled with calculation of ΔEab was carried out to investigate the probability of sensorially relevant differences. The results confirmed that, in general, color variations among the different experimental wines were not likely to be detected by human eye, as ΔEab was less than 1.0 (data not shown). These results differ from others (Skouroumonis et al. 2005a), which showed that differences in oxygen exposure during storage resulted in significant color differences. Considering that, after 24 months, several wines already had free SO2 less than 10 mg/L, the lack of major color differences observed here is quite interesting, as it has been suggested that below this level there is a high risk of advanced color oxidation (Godden et al. 2001). Our data indicate that this value needs to be considered carefully, suggesting that generalizations are not possible. Wine content of phenolic compounds is highly variable, depending on grape variety, region, vintage, and winemaking technology. The only case in which, after 24 months, ΔEab greater than 1.0 was observed was for the pair Co/HS18/High and Sc/HS18/Med, which had a TCO difference of ~15 mg/L oxygen.
Correlation between TCO and abs420 was quite low (Pearson correlation 0.568), opposite of that observed for SO2, suggesting that not only TCO at a specific time point but also timing of oxygen exposure can affect wine color development. For example, in the screwcap treatments, oxygen exposure occurred mainly in the early stages.
Hydrogen sulfide.
Hydrogen sulfide (H2S) was measured in the wines after 24 months of storage by gas chromatography coupled with pulsed flame photometric detection (GC-PFPD) (Figure 6). Type of closure significantly affected final H2S concentration and in general was the experimental variable introducing the greatest changes in H2S after 24 months of bottle storage (Table 3). Samples under screwcap closures were consistently characterized by increased H2S, with values up to 200% higher than coextruded closures (Figure 6), consistent with previous findings highlighting an association between low oxygen exposure during storage and higher levels of H2S (Lopes et al. 2009). Once all the oxygen present at bottling was consumed, the wine environment became highly reductive under screwcap, favoring accumulation of H2S. Conversely, coextruded closures provided a moderate supply of oxygen throughout storage, decreasing H2S accumulation over time. Among the other variables, HS composition at bottling also affected H2S, with increasing HS oxygen consistently resulting in lower H2S (Figure 6, Table 3), with Sc/HS6/High as the only exception. Interactions between HS oxygen and type of closure were also significant, although to a lesser extent. Thus, within each type of closure, the pattern of H2S accumulation over time can be affected by initial HS oxygen.
Finally, HS volume was the experimental variable with the least effect on final H2S, and in some case larger headspace resulted in slightly higher H2S concentrations. High transient levels of H2S can trigger side reactions between H2S and other wine components, resulting in a net loss of H2S (Marchand 2002). However, in our study, that could explain the differences observed within each treatment between HS6 and HS18, but not the overall pattern of H2S in the final wines. In any case, the results of this study provide evidence that management of oxygen at bottling has the potential to affect wine aroma development during storage. Hydrogen sulfide has been associated with off-flavors of rotten egg, often referred to as “reduced” aromas. An odor threshold of 1.6 μg/L has been proposed for H2S in white wine (Siebert et al. 2009), indicating that in the wines studied here H2S was present in concentrations higher than its threshold, therefore potentially contributing to the aroma characteristics of the wines.
Conclusion
Management of HS oxygen and closure selection can affect the evolution of different wine components during aging, including SO2 and H2S. The decline of SO2 during storage is directly linked to the amount of oxygen consumed by the wine. Therefore, during the first four months in 375 mL bottles, SO2 loss is closely dependent on oxygen present in the headspace, while after 10 months in the bottle, oxygen entering through the closure becomes the main factor affecting SO2 decline. The combination of different forms of headspace management at bottling, coupled with closure selection, offers potential for tailoring SO2 addition and wine shelf life according to different needs. Data collected here on H2S also indicate that headspace management can affect compositional parameters that are potentially linked to wine sensory quality, although the effect of closure type on final H2S appeared to be stronger. Further studies are needed to fully elucidate the implications of HS management at bottling on wine aroma composition and sensory properties, in particular with regard to the balance between reduced and oxidized aromas.
Acknowledgments
Acknowledgments: This study was financially supported by Nomacorc, Thimister Clermont, Belgium.
- Received January 1, 2011.
- Revision received April 1, 2011.
- Accepted May 1, 2011.
- Published online December 1969
- © 2011 by the American Society for Enology and Viticulture