Objective of the Study
The objective of this study was to assess the performance of some common preservation methods and “systems” used by wine drinkers and wine/restaurant tasting bars to preserve wine in opened, unfinished bottles.
The preservation methods and systems included: replacing the cork without flushing the air out from the headspace; replacing the cork by first flushing the headspace with an inert gas (Private Reserve™); resealing with a rubber stopper and pumping the air out using a hand vacuum pump tool; resealing with a glass stopper fitted with an O-ring (similar to a Vinolok glass closure) with no flushing of headspace; resealing with a Hayley’s Corker® with no flushing of headspace; and resealing with a decorative metal stopper (fitted with rubber rings) with no flushing of headspace. The Coravin Wine System was tested under a separate study.
The major method of analysis and assessment was to measure changes in headspace oxygen (HSO) and dissolved oxygen (DO), the sum of which is referred to as total package oxygen (TPO). Wines were also assessed using a 20-point scale for appearance, aroma and bouquet, taste, aftertaste, and overall impression, and compared to a control wine stored in the same type of bottle and a natural-cork stopper.
Summary
All preservation methods and systems tested in this study had an impact on wine quality after 2 to 5 days of storage at room temperature in half-full, 750-mL bottles.
The control wine exhibited aroma and flavor complexities with subtle, pleasant aromas of toasted oak. Only the wines preserved under the vacuum pump sealer system and glass stopper still exhibited, though somewhat less intense, those toasted oak aromas; all other methods and systems had completely or almost completely lost those aromas.
The wine preserved using a vacuum pump sealer system was the only wine with a residual amount of FSO2 comparable to the control wine, which confirms that this system offers the best protection against oxidative effects. Interestingly, however, this wine had the largest increase in volatile acidity (VA); VA increased in all wines (except the control) but could not be detected as the measured concentrations were still below the threshold level.
The wine preserved under a blanket of inert gas was rated third best although it had the biggest drop in FSO2 and dropped a lot of sediment; sedimenting was not observed in any other wine.
The wines resealed under the original cork with no flushing and under a Hayley’s Corker rated fourth best; they both had lost the oak character.
The wine under the metal stopper scored the lowest as the wine had lost most of its aromatic complexities and the taste was becoming more aggressive.
There were no perceptible (to the eye) changes in color or appearance in any of the wines. The control wine was assessed the same score as the bulk wine prior to bottling.
Introduction
Wine drinkers, winery tasting bars, and restaurants serving wine by the glass protect opened, unfinished bottles from adverse effects and spoilage using various preservation methods and gadgets (“systems”). These range from simply replacing the cork in the bottle to flushing air out of the headspace and resealing the bottle with a special closure to placing bottles in the refrigerator.
There is ample anecdotal evidence favoring one method or system over another. But such comparisons are not valid unless preservation methods and systems are tested on the same wine as wine chemistry greatly influences how it reacts to atmospheric oxygen once the bottle is opened. The main wine chemistry factors include the level of free sulfur dioxide (SO2), or FSO2, the amount of dissolved oxygen (DO), the concentration of headspace oxygen (HSO) over the wine once the bottle is opened, total package oxygen (TPO) – which is the sum of DO and HSO and which can remain constant if DO and HSO change by the same amount and in opposite directions – and storage temperature. By examining the changes in these parameters and monitoring pH and volatile acidity (VA), color intensity (IC), hue (H), and total phenol concentration (TPC), these can provide valuable insights on how wine in an opened, unfinished bottle evolves and whether there are any adverse effects.
In opened, resealed bottles, the increased amount of HSO would cause an increase in DO, which would be catalytically reduced into hydrogen peroxide (H2O2). H2O2 would then oxidize the ionic form of SO2 into its sulfate form and therefore causing a drop in FSO2. This series of autoxidation reactions is fairly quick, and therefore, a drop in FSO2 is a sure sign of oxidation (and possibly loss to the atmosphere around the closure).
Concurrently, albeit at a much lower rate, polyphenols are oxidized into their o-quinone forms, which are then regenerated back into their polyphenol form via sulfite ion interactions; this too causes a small drop in FSO2. Sulfite ions can also attach to o-quinones and form bisulfite addition products that can alter the color of wine. O-quinones can also polymerize with other polyphenols, including anthocyanins, the red pigment molecules in red wine, to form pigmented polymers that can cause further alterations in color.
H2O2 from oxygen reduction can also oxidize ethanol into acetaldehyde in the absence of FSO2. Changes in acetaldehyde would also be very valuable, however, the equipment and methods for this analysis were not available.
As only tiny amounts of ethanol (and unmeasurable with the available equipment) would be oxidized, changes in ethanol concentration were not monitored. The amount of acetic acid produced from acetaldehyde oxidation was however measured using a Cash Still and titrating of VA. High levels of VA could potentially mask aromas in wine.
As changes could be very subtle, particularly over only one or two days of storage, the study was carried out over a 5-day period.
As advanced techniques and equipment (e.g., gas chromatography) to measure changes in aroma compounds were not available, the wines were assessed by standard, albeit subjective, tasting techniques and scored using a 20-point scale.
Results obtained in this study and discussed here only apply to the wine tested. Results will vary greatly depending on wine chemistry. The wine used in this study was a homemade Merlot vinified from grapes from the 2011 vintage. The wine is characterized by oak aromas and flavors resulting from aging in an oak barrel and which are mainly due to the compound known as whiskey lactone, known chemically as 5-butyl-4-methyl-4,5-dihydro-2(3H)-furanone.
Test Method
A homemade unfined, unfiltered Merlot vinified from Lodi grapes from the 2011 vintage was chosen for this study. The wine was analyzed for DO, pH, VA, FSO2, TSO2, IC, H and TPC just prior to bottling. FSO2 was adjusted to approximately 25 mg/L. A drop in FSO2 immediately post bottling is expected due to sulfite binders, mainly carbonyls and polyphenols, binding to bisulfite ions. The wine was assessed prior to bottling using a 20-point scale: 3 points for appearance, 6 points for aroma and bouquet, 6 points for taste, 3 points for aftertaste, and 2 points for overall impression.
The wine was bottled in seven (7) bottles equipped with oxygen sensor spots that would allow measuring HSO in the ullage and DO in the wine. All bottles were sealed with identical natural-cork closures. One bottle was used for control and one bottle was used for each of the wine preservation methods. The bottles were left to stand for approximately 2 hours to allow the wine to reach equilibrium in each bottle. Then, test bottles were uncorked and exactly 375 mL of wine was carefully poured out. The control bottle was left corked and standing upright for the duration of the study. Test bottles were resealed as follows:
- One bottle was resealed with the original cork closure re-inserted approximately halfway down with no flushing of headspace.
- One bottle was first flushed of air in the headspace using a can of Private Reserve™ inert gas (a mix of argon, carbon dioxide and nitrogen), and then resealed with the original cork closure re-inserted approximately halfway down.
- One bottle was resealed with a rubber stopper and the air pumped out using a vacuum pump sealer tool.
- One bottle was resealed with a glass stopper fitted with an O-ring with no flushing of headspace.
- One bottle was resealed using a Hayley’s Corker® with no flushing of headspace.
- One bottle was resealed with a decorative metal stopper (fitted with rubber rings) with no flushing of headspace.
All bottles were held at approximately 20°C (68°F) for the duration of the study. Testing preservation methods and systems by placing bottles in cold storage was not performed as the intent was to not alter the speed of oxidation and polymerization reactions.
HSO, DO and TPO were measured in all bottles every 6 hours over a 5-day period. pH, VA, FSO2, TSO2, IC, H and TPC were then measured in all bottles at the conclusion of the study, i.e. at t=120 hours. At the end of the study, all wines, including the control, were assessed using a 20-point scale. The 6-hour interval would allow the wine to reach equilibrium in terms of oxygen dynamics and the monitoring of oxygen changes. This would allow detecting changes in HSO and DO as HSO would become dissolved and become oxidized and start interacting with other wine components.
Since the study examined changes in oxygen levels in the headspace and in the wine to assess changes in wine chemistry, bottles were not reopened nor was the wine retasted until the end of the study so as not to alter the oxygen (and FSO2) equilibrium.
A NomaSense P300 Oxygen Analyzer was used to measure HSO, DO and TPO (this is simply the sum of HSO and DO). A Hanna HI 902 Potentiometric Titration System with pH and ORP probes was used to measure pH, VA, FSO2 and TSO2. A Cash Still was used to separate the volatile acid fraction in wine. A Hanna HI 83742 Photometer for the Determination of Color and Total Phenols in Wine was used to measure IC, H and TPC.
Results and Discussion
Control: The control wine behaved as expected. Sulfite binders (mainly carbonyls and polyphenols) and DO at bottling quickly reacted to reduce FSO2. DO remained relatively constant throughout the study, increasing only 3%, while HSO declined steadily and became dissolved in the wine, further consuming FSO2. HSO dropped by about 58% while TPO dropped by 53%. FSO2 dropped to about half the level at bottling by the end of the study; test wines resealed for preservation would therefore be expected to drop by at least the same amount, and ideally be at the same level to perform well.
There were no significant changes (beyond instrumentation error) in all other parameters.
At the end of the study, the wine tasted exactly the same as the bulk wine just prior to bottling; it was awarded a score of 16.5.
Cork/No flushing: Wine preserved in a resealed bottle using the original cork with no flushing of headspace had lost its subtle oak aromas and tasted a touch more aggressive; there was some loss of taste complexity. The wine was awarded a score of 15.0.
Once the bottle was opened and wine poured, there was a spike in HSO, as expected, which caused a further drop in FSO2 compared to the control wine. HSO declined steadily and became dissolved in the wine, further consuming FSO2, while DO increased by about 238%. TPO decreased steadily and dropped by approximately 20%. This was the only test wine where HSO declined (21%) over the course of the study. This suggests 1) that the cork, when pushed in halfway, seals the bottle fairly well and does not let air in, and/or 2) that HSO becomes dissolved at a faster rate than air enters the bottle around the cork. FSO2 dropped by approximately 60%, and therefore more than the control wine did, causing the wine to suffer a little more oxidation.
The only other significant parameter change beyond instrumentation error is the 11% increase in VA.
Cork/Inert gas flushing: Wine preserved in a resealed bottle using the original cork with inert gas flushing of headspace had lost its subtle oak aromas but still tasted as good as the control wine. The wine was awarded a score of 15.5.
The flushing operation was not as effective as expected; there was still about 0.4% oxygen in the headspace even after more than 10 seconds of flushing. This wine was the only one to have HSO increase during the study. HSO increased 5-fold in the first 24 hours with DO remaining relatively constant. This suggests that there was a loss of inert gas and/or a poor seal from reinsertion of the original cork. HSO continued to increase to about double the HSO measured after 24 hours while DO almost tripled. These factors all caused a further reduction in FSO2; this wine recorded the largest drop (75%) of all test wines. At the end of the study, TPO and HSO had increased by 583% and 301%, respectively, while DO dropped by 78%.
The only other significant parameter change beyond instrumentation error is the 14% increase in VA.
Quite unexpectedly, significant sediment was observed at the bottom of the bottle when the wine was poured out at the end of the study. No significant changes in TPC was recorded although there was a slight increase in color intensity with no change in hue, suggesting proportional absorbance increases at both 420 nm (yellow) and 520 nm (red) wavelengths, but no changes could be detected visually.
Vacuum pump sealer: Wine preserved in a bottle resealed with a rubber stopper and the air pumped out using a vacuum pump sealer tool was comparable to the control wine with only a slight loss of oak aromas. The wine was awarded a score of 16.0.
The flushing operation was not nearly as effective as expected; there was still about 4.2% oxygen in the headspace even after many more pulls than normally performed, even after there was considerable resistance in the pull/push operation. After reaching equilibrium within 24 hours, HSO remained fairly constant for the duration of the study. DO too remain remained realtively constant with a small increase observed at the end of the study, suggesting that the Vacu-Vin system starts losing effectiveness after 5 days. At the end of the study, TPO and HSO had decreased by 15% and 14%, respectively, while DO rose by 64%. FSO2 was almost identical to the control wine, confirming the efficacy of this type of preservation method.
Interestingly, this wine recorded the highest jump (15%) in VA.
There were no significant changes (beyond instrumentation error) in all other parameters, including pH.
Glass closure: Wine preserved in a bottle resealed with a glass stopper was comparable to the control wine with only a slight loss of oak aromas. The wine was awarded a score of 16.0.
HSO remained fairly constant for the duration of the study, decreasing by only 5% while TPO decreased by only 3%. DO remain remained relatively constant for the first three days and then started increasing, reaching a 407% increase from the level measured at the start of the study after the initial pour, suggesting that the glass stopper starts losing effectiveness after 3 days. FSO2 dropped by approximately 59%, very similar to the drop seen in the wine resealed under the original cork with no headspace flushing.
There were no significant changes (beyond instrumentation error) in all other parameters, including pH.
Hayley’s Corker: Wine preserved in a bottle resealed with a Hayley’s Corker was comparable to the wine resealed under the original cork with no headspace flushsing, exhibiting a loss of oak aromas and aroma and taste complexity. The wine was awarded a score of 15.0.
TPO remained fairly constant (2% increase) for the duration of the study with only a small decrease (5%) in HSO and a commensurate increase in DO, though a 433% increase from the level measured at the start of the study after the initial pour. The increase occurred after three days from the start of the study, suggesting that the Hayley’s Corker starts losing effectiveness after 3 days. FSO2 dropped by approximately 66%, a slight increase as measured in the wine resealed under the original cork with no headspace flushing and the wine resealed under a glass stopper.
There were no significant changes (beyond instrumentation error) in all other parameters, including pH.
Metal stopper: Wine preserved in a bottle resealed with a metal stopper was comparable to the wine resealed under a Hayley’s Corker, exhibiting a loss of oak aromas and a more pronounced loss of aroma and taste complexity. This wine was assessed the lowest score (14.0) in this study.
TPO remained fairly constant (2% increase) for the duration of the study with only a small decrease (5%) in HSO and a commensurate increase in DO, though a 458% increase from the level measured at the start of the study after the initial pour. The increase occurred within two days from the start of the study, suggesting that the metal stopper starts losing effectiveness in less than 2 days. FSO2 dropped similarly as the Hayley’s Corker by approximately 65%, a slight increase as measured in the wine resealed under the original cork with no headspace flushing and the wine resealed under a glass stopper.
There were no significant changes (beyond instrumentation error) in all other parameters, including pH.
Oak aromas: Oak aromas are mainly due to the compound whiskey lactone, known chemically as 5-butyl-4-methyl-4,5-dihydro-2(3H)-furanone. No correlation between oak aroma loss and measured parameters could be established. Condensation to varying degrees was noticed in bottles as an equilibrium was established between the liquid and gaseous phases. The low partition coefficient of whiskey lactone causes the compound to volatize and enter the gaseous phase as equilibrium is established. Volatized whiskey lactone is either loss, to some extent, through closures during the course of the study or upon opening bottles at the conclusion of the study to pour and assess the wines. A greater partial pressure of gaseous whiskey lactone would cause greater retention of aromas in the wine; this could not be measured. Observed results, measured parameters and the type of preservation methods and systems do not provide sufficient conclusive data.
Conclusions
Leftover wines in half-full 750-mL bottles resealed with a vacuum pump sealer system or glass stopper performed best in preserving wine over a 5-day period. Although the wine resealed under the original cork but with the headspace flushed with inert gas was assessed second best to the vacuum pump sealer system or glass stopper, a lot of sediment was observed in the bottle; there was also a loss of oak aromas. Wines resealed with the original cork with no flushing and a Hayley’s Corker were assessed third best having lost oak aromas and some aroma and taste complexity. Both the glass stopper and Hayley’s Corker are recommended for preserving wine for no more than 3 days. The metal stopper was assessed the lowest as it became devoid of all oak aromas and some loss of aroma and taste complexity; it is recommended for preserving wine for no more than 2 days, perhaps less.
References
Biondi Bartolini, A., F. Cavini and M. de Basquiat. 2008. Oxygène et vin – Du rôle de l’oxygène à la technique de micro-oxygénation. Parsec Édition, Florence (IT).
Ribéreau-Gayon, P., D. Dubourdieu, B. Donèche and A. Lonvaud. 2012. Traité d’œnologie, Tome 1 – Microbiologie du vin. Vinifications. 6e édition. Dunod, Paris.
Schmidtke, L.M., A.C. Clark and G.R. Scollary. 2011. Micro-oxygenation of red wine: techniques, applications, and outcomes. Crit. Rev. Food Sci. Nutr. 51:115-131.
Vivas, N. 1999. Les oxydations et les réductions dans les moûts et les vins. Éditions Féret, Bordeaux.
