Abstract
Background and goals Every year, some white wine is lost during distribution because it has unexpectedly started to show oxidized character. One reason some wines taste “reduced” is that the winemaker has added excessive SO2 to avoid over-oxidation. While SO2 and ascorbate (where it is used) are the key preservatives in wine, some assays (FRAP, DPPH, and Folin-Ciocalteu method) used to predict a wine’s protection from oxygen include a response to components such as phenolics. However, phenolics are not antioxidants; they function more like catalysts that are continually recycled by SO2, at least in white wine before over-oxidation. This research aimed to develop a rapid shelf-life test for white wine by predicting SO2 exhaustion during aging.
Methods and key findings Commercial white wines (n = 8) experienced accelerated aging (air saturated and heated at 45°C) and the total SO2 to O2 reaction ratio was calculated. Additionally, the SO2 addition method was used to determine the sulfite buffering factor of each sample. From those values, we derived a new formula, “potential wine shelf-life”, that factors in O2 uptake through wine closures and initial and minimum SO2 levels to predict when a particular wine would exhaust its free SO2.
Conclusions and significance “Potential wine shelf-life” can forecast when free SO2 levels drop below a necessary minimum, and the wine will start to exhibit oxidized character. The simple formula should provide winemakers with a much more precise estimate of the shelf-life of their white wines.
Introduction
Every year, wine loss due to poorly predicted shelf-life is significant during wine distribution and in personal wine cellars. Both winemakers and wine consumers are interested in the question ‘how long will this wine last?’. Currently, such estimates are usually based on expert experience and tasting notes (Langlois et al. 2010), but not on chemical analysis.
Existing methodologies reported to predict a wine’s ability to resist oxidation measure the antioxidant capacity of wine using various methods, including the ferric reducing antioxidant power (FRAP), 2,2-diphenyl-1-picrylhydrazyl (DPPH), and Folin-Ciocalteu assays, which respond primarily to polyphenolics and sulfur dioxide (SO2) (Danilewicz 2015). The reactions between the phenolics in the sample and the added oxidants (e.g., Fe3+, DPPH radicals, or a phosphotungstate-molybdate complex) leads to the formation of quinones, and the SO2 in the sample recycles quinones back into phenolic compounds (Figure 1). However, since these methods apply excessive oxidants, the reaction hits its endpoint when SO2 is exhausted and the quinones cannot be recycled. Consequently, the FRAP and DPPH assays correlate strongly with total flavanol levels in wine, while the Folin-Ciocalteu assay is a standard method to determine total phenolics (Arnous et al. 2001). Alternatively, a UV-visible spectroscopy method was developed to monitor catechin, Fe3+ ions, quinones, and browning to determine the oxidative stability of model wines (containing Fe2+ and methyl catechol) during accelerated aging with oxidation using hydrogen peroxide (Lopresti et al. 2024). This method focused on oxidative products of wine aging instead of wine antioxidants. However, antioxidants in wine play a key role in the early stage of wine aging, as discussed later.
Mechanism of quinone reduction to polyphenolics by sulfur dioxide (Nikolantonaki and Waterhouse 2012).
The oxidation mechanism that is widely accepted as occurring during the aging of white wine is shown (Figure 2). To simplify the process, Fe2+ is oxidized by O2 to Fe3+, which then oxidizes phenolics to quinones (Danilewicz 2016). Bisulfite, the most abundant form of SO2 under wine pH, then recycles quinones back to phenolics, coupled with the reduction of Fe3+ to Fe2+ (Nguyen and Waterhouse 2021). During regular bottle aging, wine exposure to O2 happens in a controlled and slow manner by O2 transmission through the closure. In one study, the median O2 transmission rates for natural corks, synthetic stoppers, microagglomerated stoppers, and screwcaps were 2.05, 5.16, 0.27, and 0.93 mg/year, respectively (Crouvisier-Urion et al. 2018). Compared to the antioxidant capacity tests mentioned above, the oxidants involved in the wine aging process are present in lower amounts and react more slowly than during accelerated aging, so most of the quinones produced are recycled back to phenolics by SO2, thiols, and ascorbic acid (Waterhouse and Nikolantonaki 2015). Consequently, phenolics should be viewed as catalysts of wine oxidation. On the other hand, white wine aging capacity originates from SO2, thiols, and ascorbic acid, but not from phenolics. As a result, the FRAP, DPPH, and Folin-Ciocalteu assays, which respond strongly to phenolics, can overestimate wine aging capacity.
Proposed mechanism for wine oxidation with the presence of sulfite (Danilewicz 2016).
Among the preservative antioxidants in wine, SO2 is typically present at relatively high levels, with a mean concentration of free SO2 in white wine of 26 mg/L and total SO2 of 130 mg/L (Godden et al. 2015). The antioxidant capacity of SO2 is involved in two mechanisms. First, recycling of phenolics from quinones as described above, which prevents accumulation of quinones and browning, and second, reacting with peroxide to prevent the Fenton reaction. SO2 also prevents emergence of oxidation aroma by binding with the oxidation marker, aldehydes. Aldehyde in wine comes from the fermentation process, where a small fraction of the aldehyde produced by glycolysis is trapped by SO2 or is not reduced by NADH. Some aldehydes can also be produced from amino acids via Strecker degradation. Aldehydes bind with SO2 (Figure 3), so some SO2 is present in the bound form. The bonds between different carbonyls and SO2 are of different strength. For instance, the acetaldehyde-bisulfite bond has a very low dissociation constant, with the acetaldehyde strongly bound to SO2, while other carbonyls, including pyruvic acid, α-ketoglutaric acid, galacturonic acid, and glucose, have higher dissociation constants (i.e., they bind weakly to SO2) (Waterhouse et al. 2024).
The sulfite and acetaldehyde binding mechanism.
During wine oxidation, as the free form of SO2 is being depleted, the weakly bound SO2 slowly dissociates and replenishes the free SO2 pool. The ratio of the change in total SO2 to the change in free SO2 during oxidation, such as aerial exposure of wine, is >1 (Sacks et al. 2020). This ratio represents the sulfite buffering capacity of a wine system, and a higher ratio indicates a greater sulfite buffering capacity. However, there is a limit to sulfite buffering and it is reported that there is no buffering when free SO2 is <5 mg/L (Sacks et al. 2020). It is important to note that when free SO2 is very low, free aldehyde levels rise enough to be detected; one report states that at <10 mg/L free SO2, oxidized aromas are notable (Godden et al. 2001).
Another important factor describing SO2 oxidation reactions is the SO2:O2 molar reaction ratio, which refers to the molar ratio of total SO2 consumption versus total O2 consumption. Theoretically, the SO2:O2 molar reaction ratio should be 2, as 1 mole of O2 reacts with 2 moles of SO2 (while a direct reaction is unlikely, SO2 does react with oxidation products to yield the same stoichiometric result). However, the reported SO2:O2 molar reaction ratio of commercial white wines varies between 1.29 and 2.03 (Danilewicz 2016). This indicates that in some wines, other antioxidants such as ascorbic acid, thiols, or phenolics are reacting with O2 or other oxidants instead of SO2.
As SO2 is the key preservative in wine, a method projecting the free SO2 level decline during aging could predict when levels drop to a minimum safe level that could be used to predict the wine’s shelf-life. The aim of this study was to develop a rapid shelf-life test by predicting the time it will take to reach a minimum level of free SO2 (chosen by the winemaker) during bottle aging.
Materials and Methods
Materials and chemical reagents
Sodium thiosulfate (0.05 mol/L) and sodium hydroxide solution (2 mol/L) were purchased from Sigma-Aldrich. Starch (powder for iodometry) and potassium metabisulfite were purchased from Thermo-Fisher Scientific Inc. Iodine (0.01 mol/L) was purchased from Exaxol Chemical Corporation and sulfuric acid (25% v/v) was purchased from Ricca Chemical.
Commercial wines
Eight commercial white wine samples were used, including seven Chardonnay and one Sauvignon blanc, sourced from California, New Zealand, and Argentina, as described (Table 1).
Basic analyses of eight commercial white wines.
Accelerated wine oxidation procedure
750 mL of wine was transferred into a 1-L flask with a magnetic stir bar and stirred for 40 min to dissolve O2 to 7 mg/L. 2 × 750 mL of air-saturated wine were mixed (1500 mL total) and used to fill 16 75-mL biological oxygen demand (BOD) bottles, leaving no headspace. Each bottle was sealed using silicone grease and a glass stopper to ensure an airtight environment. The BOD bottles had O2 sensor spots (SP-Pts3-NAU; Presens) attached inside for dissolved O2 analysis. All sealed bottles were heated at 45°C for 5 days in an oven (Precision Scientific Group) in the dark. Dissolved O2 levels of wine samples were measured in unopened bottles using a Fibox 3 LCD trace O2 meter (Presens). On days 0, 1, 3, and 5, two bottles were opened to determine free and total SO2 by the Ripper method, chosen for its wide availability in the wine industry. While alternatives such as aeration-oxidation and headspace methods could be applicable for our proposed test, further study would be required to ensure they would work in a comparable manner. The remaining eight bottles of wine were poured out and underwent a second round of air saturation. The accelerated oxidation and testing procedures were repeated.
SO2:O2 reaction ratio
The SO2:O2 molar reaction ratio was determined by adapting the procedure described by Danilewicz (2016). Measurements of dissolved O2 and total SO2 were collected for the wine samples on days 0, 1, 3, and 5, across two rounds of the accelerated wine oxidation procedure. The relationship between O2 consumption (x-axis) and total SO2 level (y-axis) was analyzed. The slope of the calculated linear regression of these variables (total SO2/total O2 consumption) was used to determine the SO2:O2 molar reaction ratio for each wine sample.
Sulfite buffering capacity
Sulfite buffering capacity describes the ability of a wine to release SO2 from its bound forms during wine oxidation and equals the ratio of change of total SO2 (Δ total SO2) to the change of free SO2 (Δ free SO2). The methodology for measuring sulfite buffering capacity was adapted from described procedure (Sacks et al. 2020), in which the ratio of total SO2 to free SO2 was measured during the oxidation process (the loss of SO2). The adapted analysis involved sequential SO2 addition, allowing the reaction to be completed in 2 days. A 10 g/L SO2 stock solution was prepared by dissolving 868 mg of potassium metabisulfite in 50 mL of Milli-Q water. SO2 concentrations of 25, 50, 75, and 100 mg/L were added to the wine samples. The samples were then allowed a two-day period for equilibration. Following this, measurements of free and total SO2 were performed at room temperature. The data were plotted with total SO2 on the y-axis and free SO2 on the x-axis. A linear regression analysis was conducted and the slope of this regression line was Δ total SO2/Δ free SO2, the sulfite buffering capacity of the wine.
Free SO2 depletion factor
Free SO2 depletion factor is the ratio of free SO2 depletion to consumed O2. It is the conceptual framework of this research to illustrate the evolution of free SO2 during aging. Free SO2 and dissolved O2 were analyzed at days 0, 1, 3, and 5 on wine samples undergoing accelerated oxidation. The relationship between O2 levels (x-axis) and free SO2 concentrations (y-axis) was analyzed, and the slope of the linear regression of the variables equals the free SO2 depletion factor.
Results and Discussion
This study was performed to predict wine shelf-life by forecasting SO2 depletion during the aging process. Previous research identified key factors such as the SO2:O2 molar reaction ratio and the free SO2 to total SO2 ratio. The objective of this research, therefore, was to further investigate a linkage between these factors, determine whether they remain constant throughout the aging process, and subsequently, develop a simpler and more effective prediction factor. The insight could be valuable to predict SO2 depletion and thus improve wine shelf-life predictions, leading to more precise SO2 additions to attain a desired wine shelf-life, or time until free SO2 is depleted to a desired minimum.
SO2:O2 molar reaction ratio
Total SO2 levels versus total O2 consumption from accelerated oxidation is illustrated by the example of wine #4, which underwent two rounds of accelerated oxidation (Figure 4). These data were used to construct a linear regression between SO2 and O2 consumption. Total SO2 showed a good linear correlation to total O2 consumption, and the slope of the linear regression line defines the SO2:O2 molar reaction ratio. The linear correlation lines of seven out of eight white wines showed correlation coefficients >0.91 (Table 2). The correlation coefficient of wine 1 was 0.8103, which was lower than the others due to the very low free SO2 of this wine, which started at 8 mg/L, before dropping to 3 mg/L after accelerated oxidation. When SO2 approaches very low levels, oxidation reactions must necessarily involve other components, and thus the ratio of reaction with O2 must change. However, until that point, the rate differential between oxidants and either SO2 or other antioxidants appeared to be fairly constant for a particular wine. The slope of the linear regression line defines the reaction ratio of SO2:O2 and the SO2:O2 molar reaction ratio. The range of SO2:O2 molar reaction ratios of seven wines with more normal SO2 levels was 1.0972 to 1.6621, similar to a previous finding that the SO2:O2 molar reaction ratio of white wines was in the range of 1.3 to 2 (Danilewicz 2016). A ratio <2 indicates oxidation of other antioxidants such as phenols, thiols, and ascorbic acid, in the presence of SO2. Under those conditions, the wine appears to be more susceptible to oxidation.
Total sulfur dioxide consumption versus total oxygen consumption during accelerated oxidation (n = 8) for wine #4. The slope indicates the SO2:O2 molar reaction ratio.
The SO2:O2 molar reaction ratio, determined by linear regression analysis of eight wine samples (n = 8).
Sulfite buffering capacity
When SO2 is added to a wine, both the free and total SO2 levels increase, but not necessarily at the same rate, because there are weak binders in wine that bind only to a fraction of the added SO2. Comparing those values can be very informative. For example, in wine #1 the total SO2 versus free SO2 using the SO2 addition method was plotted and a linear correlation line was created (Figure 5). The slope of the linear correlation line equals the change in total SO2 versus the change in free SO2. This ratio represents the sulfite buffering capacity. The correlation coefficients for the linear regression equations of the eight white wines consistently exceeded 0.92 (n = 15; Table 3), which suggests that the ratio of the change in total SO2 to free SO2, i.e., the sulfite buffering capacity of wine, is a constant for specific wines with typical SO2 levels. The sulfite buffering capacities of the eight wines were in the range of 1.25 to 1.98, a larger range than a limited prior report on the sulfite buffering capacity of four Chardonnay wines, which spanned 1.38 to 1.45 (Sacks et al. 2020). The greater the sulfite buffering capacity, the higher the percentage of weakly bound SO2 in wine. In such wines, more bound SO2 can be released as free during wine oxidation, i.e., more free SO2 is released by that bound SO2. Thus, by knowing the buffering ratio for a particular wine, one can then estimate how much free SO2 will be consumed with a particular quantity of oxidation, at least until the weakly bound SO2 is exhausted. That leads to Equation 1, the free SO2 depletion factor.
Total SO2 versus free SO2 by the SO2 addition method (n = 15) for wine #1. The slope indicates sulfite buffering capacity.
The sulfite buffering capacity, determined by linear regression analysis of eight wine samples (n = 15).
Free SO2 depletion factor
The SO2:O2 molar reaction ratio is the ratio of total SO2 change to O2 consumption, while sulfite buffering capacity equals total SO2 change divided by free SO2 change. By cancelling Δ [Total SO2], the ratio of SO2:O2 molar reaction ratio to sulfite buffering capacity equals the ratio of free SO2 change to O2 consumption. This is referred to as the free SO2 depletion (to O2 consumption) factor. Since the SO2:O2 molar reaction ratio and the sulfite buffering capacity are constants (within the normal range of wine values), the quotient of the two constants is also constant. Thus, the free SO2 depletion factor is constant for each wine.
The free SO2 depletion amounts versus dissolved O2 level during accelerated oxidation of wine #7 were plotted and a linear regression line was created (Figure 6). The slope of the linear correlation equation is the free SO2 depletion factor. The correlation coefficient of the linear correlation line of free SO2 versus dissolved O2 of the eight wines was >0.86 (Table 4). These data support our contention that the free SO2 depletion factor is a constant, over a typical range of SO2 values, and can be measured in white wines in this manner.
The dissolved oxygen level versus the free sulfur dioxide in wine #7 (n = 12). The slope indicates the free SO2 depletion factor.
The free SO2 depletion factor, determined by linear regression analysis of eight wine samples (n = 12).
The free SO2 depletion factor of the eight wine samples was in the range of 0.74 to 1.89. The lower the free SO2 depletion factor, the less the free SO2 decreased with the same level of O2 consumption; in other words, such wines are not losing much SO2 during oxidation. However, the free SO2 depletion factor doesn’t directly correlate with the aging capacity of wine, as the relatively low free SO2 depletion factor could be related to two possible scenarios. First, the free SO2 in the wine may be too low to react with O2, giving the wine a very low aging potential, such as wine #1. Alternatively, the wine has high levels of weakly bound SO2 available to replenish the free SO2 pool (high sulfite buffering capacity), so the free SO2 level does not change much with oxidation and the wine is well protected by SO2, such as wine #6. As a result, the free SO2 depletion factor can be low in both very well and very poorly protected wines, so by itself it cannot reflect the aging potential of wine or provide useful information for a wine’s shelf-life.
Prediction of free SO2 after bottle aging
Total package O2 as outlined in Equation 2 constitutes a pivotal time-dependent variable and encompasses diverse O2 sources, including dissolved O2 at bottling, O2 in the headspace, and O2 permeating through the closure. The dissolved O2 in wine typically falls to <0.1 mg/L at 3 mo after bottling (Tarasov et al. 2021). At this point, the O2 at bottling and any excess O2 in the closure has been consumed. Consequently, O2 consumption after 3 mo is essentially the closure O2 transfer rate (via the closure/seal in a glass bottle) multiplied by the aging time. The free SO2 depletion in bottle for a specific time equals the total consumed O2 during that time multiplied by the free SO2 depletion (to O2 consumption) factor, as in Equation 3. The free SO2 after aging can be calculated by the free SO2 of wine at the initial time point, minus the free SO2 depletion during aging, as in Equation 4. The free SO2 after bottle aging can be calculated based on Equation 4 for wine with a dissolved O2 level <0.1 mg/L, or 3 mo after bottling.
Prediction of white wine shelf-life
This parameter was developed based on the observation that white wines with exhausted free SO2 levels begin to show oxidized character. At this stage, the strong bonds between SO2 and acetaldehyde break, releasing the free form of acetaldehyde. As described in Equation 4, free SO2 after aging is negatively correlated to aging time, and the minimum free SO2 after aging can be set at 10 mg/L, as wines with <10 mg/L free SO2 have significantly greater oxidative aroma (Godden et al. 2001, Sacks et al. 2020). For wines ~3 mo post-bottling (after the O2 of bottling has reacted with the wine), the potential wine shelf-life can be calculated using Equation 5, as the difference between free SO2 in bottle and 10 mg/L (or any value) divided by free SO2 depletion factor and O2 transmission rate of the closure.
Minimum free SO2 level at bottling
The minimum free SO2 at bottling to achieve a specific aging potential (time) for a wine can be calculated as well. This involves a more complex description of O2 consumption during aging, as in Equation 6, because at the time of bottling, the dissolved O2 level of wine and other O2 sources in the bottle are not negligible. Minimum free SO2 level at bottling can be calculated using Equation 7. The “minimum free SO2 at bottling” can be used to minimize SO2 usage during winemaking and ensure the wine is protected from oxidized character during the expected aging period, without using excess SO2.
Conclusion
The consumption of O2, free SO2, and total SO2 of eight white wine samples was recorded during accelerated aging to determine the SO2:O2 molar reaction ratio. Additionally, the sulfite buffering capacity for samples was determined using the SO2 addition method. Both the SO2:O2 molar reaction ratio and the sulfite buffering capacity were constant in each wine with >10 mg/L free SO2; using this finding as a basis, we hypothesized that free SO2 depletion versus O2 consumption factor is constant. The free SO2 depletion factor equals the slope of the linear correlation line that plots free SO2 against dissolved O2. This showed a strong linear correlation, suggesting that the free SO2 depletion factor is in fact a constant. Therefore, free SO2 depletion during a specified bottle aging time equals the O2 consumption multiplied by the free SO2 depletion factor. Using these factors and the O2 transfer rate, and a minimum free SO2 post-aging, a particular white wine’s shelf-life can be calculated. For instance, wine #7, with a free SO2 level in bottle of 28 mg/L and a free SO2 depletion factor of 1.8998, assuming an O2 transmission rate of the bottle closure of 3 mg/yr and a minimum SO2 level of 10 mg/L, would have a potential wine shelf-life of 3.16 yrs, calculated using Equation 5.
The equations illustrate the utility of knowing the free SO2 depletion factor in white wine shelf-life prediction and SO2 management at bottling. Note that the three equations assume that the SO2 depletion to O2 consumption factor determined under accelerated aging is the same as during regular aging. It would be useful to validate this in real life situations; some correction factor may be needed. The SO2 prediction shelf-life test avoids overestimation of wine antioxidant capacity by focusing on measuring and predicting SO2, so it provides a white wine shelf-life estimate that is better than existing methods.
This research provides winemakers with rapid and industry-approachable methods to estimate white wine shelf-life, providing adequate oxidation protection, while at the same time minimizing the use of SO2.
Footnotes
This work was supported by University of California, Davis. The authors declare no conflicts of interest.
Miao Y and Waterhouse AL. 2025. Rapid white wine shelf-life prediction by forecasting free SO2 loss post-bottling. Am J Enol Vitic 76:0760008. DOI: 10.5344/ajev.2025.24057
By downloading and/or receiving this article, you agree to the Disclaimer of Warranties and Liability. If you do not agree to the Disclaimers, do not download and/or accept this article.
The data underlying this study are available on request from the corresponding author.
- Received October 2024.
- Accepted February 2025.
- Published online April 2025
This is an open access article distributed under the CC BY 4.0 license.
References
Vol 76 Issue 1
Jump to section
Related Articles
Cited By...
More from this TOC section
Similar Articles
