Barrel alternatives supply oxygen to wines during elevage.
Two physico-chemical phenomena contribute to achieving this change in wine: extraction of volatile and non-volatile wood compounds with simultaneous dissolving of atmospheric oxygen in the wine. Oak is a porous material, permeable to gas, allowing exchanges with ambient air throughout the elevage period. (15)
Although the traditional cooperage market is stable, barrel alternatives such as staves, chips, powders, sticks and inserts, used in fermentation or during elevage, continue to become integrated into the everyday practice of modern enology. The choice of the rate at which these alternatives are used constitutes a decisive factor in the final quality of the wine.
It is undeniable that the olfactory and taste contribution from oak aging, and in particular from contact with an optimal quantity of staves, contributes to the production of quality wine. For example, many users agree that using staves during the aging of white and red wine can limit the formation of reductive odors, thereby improving the aromatic clarity of their wines. Their use brings oak flavors and complexity.
However, when an enologist decides to add these alternatives in excessive quantity, the resulting wine can show symptoms of over-oaking: loss of fruit and original "style" accompanied, sometimes, by an irreversible change in the wine's aroma (oxidation faults).
Mastering the oaky character of oakaged wines requires knowledge of the compounds associated with these nuances and flavors. Oaky flavors in oak-aged wines come from a large number of aromatic molecules. They include whiskey lactone, vanillin, eugenol and isoeugenol.
These compounds, which come directly from the wood, constitute part of the impact compounds: They contribute significantly to the oaky character of the wines. (12, 14)
The change in the taste of wines aged in oak has been the subject of much research, whether focusing on changes in its astringent, bitter character or, on the contrary, in its sweetness. (5,6,7)
Research undertaken by the Institut des Sciences de la Vigne et du Vin for many years on premature oxidative development of aromas in red wines has resulted in identification of a new molecule: 3-methyl-2,4-nonanedione (MND). This very aromatic diketone (sensory threshold: 16 ng/L) is reminiscent of the odor of prune. Formed through oxidative-type mechanisms, it is found in large quantity (more than 200 ng/L) in wines stored in the presence of oxygen. (11)
On the sidelines of this characterization research, there is an important area of study that has generated very little research: the study of oxygen supply during elevage. This process occurs in two stages. In the first weeks after filling, the wine impregnates the pores of new oak barrels, replacing the air present in the wood. This phenomenon is accompanied by a rapid and extensive uptake of oxygen. An equilibrium is then established between the wine and the wood, during which slow interaction occurs between the outside air and the wine inside the hermetically sealed barrel.
Given the complexity of the phenomena involved, estimating the oxygen concentration supplied to the wine during elevage in contact with wood has always been a subject of interest.
The first to pursue this field of investigation was Pascal Ribereau-Gayon, who published the first results of his research on oxygen intake during elevage in 1931. Using a rudimentary system of constant topping up with sulfur dioxide, he established that the transfer of oxygen through the wood was limited, amounting to between 2 and 5 ml of oxygen per liter per year.
Almost 60 years later, N. Vivas in Bordeaux, (15) F. Feuillat in Dijon (3) and M. Moutounet in Montpellier (9) took up this research while emphasizing other aspects--particularly measurement of wood porosity--an indirect way to estimate the extent of gas exchange across the mass of wood.
The use of barrel alternatives for aging in stainless steel tanks makes any exchange between the wine and the atmosphere impossible. But as with barrels, the high porosity of oak wood means oxygen trapped in oak pores can progressively dissolve in the wine.
The first studies carried out by F. Feuillet, (2) and more recently by others, agree on the measurement of the total porosity of oak wood: 63% on average. (8) It is therefore possible to estimate the maximum quantity of oxygen contained in 1 g of dry wood to be 0.4 mg. With reasonable, traditional use of staves (a rate of 8 g/L), the wine can therefore receive a maximum theoretical supply of 3.2 mg/L of oxygen.
This paper presents the first results relating to the study of oxygen supplied by staves: from development of a protocol for measurement of extractable oxygen from wood to the study of the impact of using staves on oxidation markers (SO, and MND).
Equipment and methods
The samples came from 15 staves of French oak cut either into small staves of 3 cm x 5 cm x 2 cm or into chips using an industrial grinder. The samples were taken from stocks at the Seguin-Moreau cooperage. Half of the samples were toasted (medium toast) according to the cooper's protocol. Before use, the samples were weighed using a precision balance.
The oak samples were macerated (about 16 g/L) in an alcohol-water solution previously desorbed using nitrogen ([Colima! less than 50 g/L). The dissolved oxygen was measured using a PreSens sensor placed on the wall of the flask. The level of dissolved oxygen was measured at regular intervals in the solution that had been prepared in a hermetically sealed flask and placed in a thermostatically controlled chamber.
The analysis of the 3-methyl-2,4-nonanedione (MND) was performed by liquid-liquid extraction and GC-MS using a previously validated method.
Development of method to monitor oxygen supplied by oak wood
The quantity of oxygen supplied by oakwood staves was estimated according to several different parameters. An example of the kinetics of oxygen liberation by a stave in a model solution is presented in Figure 1. In our experimental conditions, this includes three phases:
At the first contact between the wood and the solution, the oxygen concentration increases very significantly (Phase 1). More than 80% of the maximum oxygen concentration measured (maximum Cg during the experiment is achieved within 48 hours. The oxygen content reaches a plateau lasting two to three days, corresponding to what is referred to as the maximum concentration measured (maximum Cg.
An equilibrium is therefore established between the oxygen still present in the crevices in the wood and oxygen present in solution (Phase 2).
The last phase corresponds to a lowering of the oxygen concentration through its consumption by oak wood extractives, which outweigh the extraction kinetics (Phase 3). Given the physico-chemical phenomena involved, it is clear that direct measurement of the quantity of oxygen supplied by the oak wood to a model solution is scarcely feasible.
The experiment was repeated many times with stave fragments from the same piece of wood, and also with other wood samples. The example shown in Figure 1 corresponds to two staves from the same source.
In this experiment, the stave was removed from the solution at the stage corresponding to maximum 02 concentration. This was in order to evaluate only the speed of consumption of oxygen by the alcohol-water solution containing the oak wood extractives. In this way, we were able to make an initial approximation of the average balance between the residual oxygen yielded by the wood and its consumption by phenolic compounds.
We have shown that, under experimental conditions, the rate of oxygen consumption by the solution containing wood extractives (Phase 3) is between 20 and 30 pg/L per day. These values are low compared to the average rate of oxygen dissolution observed during the first days of maceration (Phase 1) that approaches 300 pg/L per day for the same staves.
In the example of the use of staves, and of barrel alternatives in general, the extraction is rapid because it takes place simultaneously in three directions across the oak vessels: radial, tangential and longitudinal.
Given the preliminary results, it became evident that the rate of oxygen consumption by the phenolic compounds yielded by the wood could interfere with accurate assessment of the quantity of oxygen supplied by the wood. For this reason we optimized the maceration conditions so as to minimize the consumption of oxygen by these compounds. This step allowed us to obtain a linear, replicable method, i.e. the quantity of oxygen measured (maximum 02) is proportional to the number of staves macerated.
Influence of anatomical structure of the wood and toast
"Grain" is a term used by coopers to describe the average width of growth rings, and in consequence the growth rate of a tree. They traditionally make reference to two "standards" to define grain type: Limousin oak with quick growth (4 mm or more), large structure for "coarse grain;" and Allier oak with slow growth (1 mm), tight structure for "fine grain." (4) This systematic relationship between growth rate, structure and porosity is commonly accepted by cooperage professionals.
We have assessed the impact of certain parameters on the maximum O, values measured in the samples: They range from 0.48 mg/L to more than 3.3 mg/L depending on the sample. The detailed analysis of the results (Figure 2A) shows that the selection of wood by grain width is not a criterion that enables them to be classified according to oxygen content.
These results are consistent with a 2011 study by M. Mirabel, which emphasizes, among other things, the absence of correlation between oak density and grain width. (8) However they contradict the results obtained by Riad Bakour. (1) After analysis of many anatomical parameters of oak (on nearly 290 samples), small but significant correlations were found between grain width, wood porosity and its density.
However, our results tend to show that the maximum 02 value is fairly well correlated to the density of wood used in this protocol (Figure 3). In other words, low-density wood is likely to supply more oxygen to the solution than high-density wood.
We thought the toast might change the amount of oxygen contained in the wood. It is not impossible that toasting of staves could lead to changes in the internal structure, that would likely change the porosity, or at least cause cracking of the vessel walls, allowing the liquid to extract air from the cavities more easily.
In order to validate this hypothesis, we used the previously described protocol to compare the maximum amount of oxygen yielded by the oak, whether toasted or not. The analysis of the results (Figure 2B) shows that the toast has no effect on the level of maximum 02 measured. This corresponds to the results of the recent research by M. Mirabel, also showing that the toast has no effect on porosity. (8)
Influence of type of barrel alternative
Oak chips are used as alternatives to oak barrels. The previously described protocol for staves was used to calculate the maximum 02 for chips. The maximum 02 levels for staves and chips from the same source are shown in Figure 4. Oak chips, like staves, contain a significant amount of oxygen. However, the level of maximum 02 for chips is, on average, one-half that for staves.
Change in S[O.sub.2] level according to different parameters
We have shown that the use of oak can supply significant amounts of oxygen. We illustrate this phenomenon by showing the change in the level of sulfur dioxide in an alcohol-water solution in which increasing concentrations of untoasted oak chips were macerated.
During this experiment, regular samplings were performed for 32 days (Figure 5). The loss of free S[O.sub.2] is quicker and bigger when the quantity of chips is increased. For the model wine with 15 g/L, one-half of the initial free S[O.sub.2] had been consumed after 32 days of maceration. This significant decrease demonstrates the presence of pronounced oxidation phenomena.
This result can be explained by two simultaneous phenomena. The first corresponds to what we just described, namely the greater supply of oxygen according to the amount of wood used. The second is a result of the dissolution of oak wood extractives that may catalyze oxidation reactions. (13) This effect can also be observed at a rate of 5 g/L, close to that used in practice, although to a lesser extent.
It is well-known that the level of ethanol in the medium plays an important role in the extractability of volatile and non-volatile compounds from oak. We have assessed the effect of alcohol content of the model medium containing 10 g/L of oak chips on the change in free SO,. In the range of ethanol concentrations in wine, and under our experimental conditions, its presence does not increase oxidation phenomena and therefore has a negligible effect on the change in the free SO, level.
Validation of these observations in winery: Red wine elevage
The trials were performed on a 2012 Merlot wine from the Bergerac appellation. After malolactic fermentation, the wine was racked, sulfured (initial free SO, = 29 mg/L) and then divided into four 50H L stainless-steel tanks.
The control wine (without staves) was monitored and three tanks treated with increasing concentrations of French oak staves (Oenostave, Seguin Moreau; 950 cm x 5 cm x 18 mm; medium toast) at the following rates: T2 = 5 g/L, T3 = 8 g/L and T4 = 10 g/L of oak). Under experimental conditions, the level of free S[O.sub.2] was adjusted each month so that the concentration was the same for all samples (target free S[O.sub.2] = 30 mg/L) during five months of elevage.
At the end of this experiment we have shown that the more staves are used, more S[O.sub.2] must be added in order to maintain the free S[O.sub.2] at its initial level (Figure 7).
We analyzed an olfactory marker for oxidation of red wines: 3-methyl-2,4-non-anedione (MND), whose odor is reminiscent of prune. The analysis in Figure 7 shows that its concentration increases significantly with reduction in free S[O.sub.2] in the trial at 10 g/L. However, sensory analysis of these wines does not show the appearance of oxidation faults. This result is not surprising given that the concentrations found in our trials remain below the perception threshold in wine, which is 62 ng/L.
All results obtained during this study clearly show that oak used in the form of staves is likely to supply oxygen to wines quickly and in large quantity. This result runs counter to the common idea that elevage in the presence of barrel alternatives is reductive-type maturation compared to oxidative-type maturation occurring in traditional barrels.
The benefits associated with a moderate supply of oxygen during elevage of wine are understood: limitation of reductive aromas, stabilization of color in red wines and a refined wine taste.
Meanwhile, given the large amount of oxygen supplied by staves, the use of alternatives requires careful monitoring of the free S[O.sub.2] level by the winemaker, at the risk of development of oxidative phenomena and, in certain specific cases, the development of spoilage microflora (Brettanomyces D.; Acetobacter sp.).
In practical terms, a winemaker should anticipate the loss of S[O.sub.2] caused by addition of barrel alternatives to a wine. This loss is dependent on quantity of oak used and its form (chips, powder or staves), the maximum theoretical loss of free S[O.sub.2] being possible to reach 0.75 mg/L per gram of oak added.
The winemaker should also consider the quantity of oxygen introduced by barrel alternatives in total oxygen balance of the process. That could be important, for example, for adjusting micro-oxygenation rates used for wine production.
The authors would like to thank Richard Nadal (winemaker from Domaine de Marsalet) and Cyril Fourcade (consultant enologist) for setting up the trials.
Caption: OENOSTICK barrel insert in 225L barrel.
Caption: Figure 1: Example of the kinetics of development in oxygen supplied by a stave of oak wood placed in a model solution with composition similar to wine, according to whether contact continues through the experiment (stave ?) or whether the stave is removed once the maximum 02 value is reached (stave -).
Caption: Figure 2A and 2B. Box-plot representation of the maximum 02 level measured in alcohol-water solution. The vertical line corresponds to the median of the measured values. Figure 2A: Comparison of coarse and fine grain. Figure 2B: Comparison of toasted and non-toasted staves.
Caption: Figure 3: Maximum quantity of oxygen supplied by an un-toasted oak stave placed in a model solution, according to the density (n = 20).
Caption: Figure 4: Comparison of the level of maximum 02 obtained from maceration of the same quantity of oak staves and chips (17 g/L) in an alcohol-water solution (n = 6).
Caption: OENOBLOCKs in infusion bag.
Caption: OENOSTAVE fan of twenty 0.28 mm thick French oak staves that are 37.4 inches long and 2 inches wide.
Caption: Figure 5: Impact of the addition of increasing quantities of oak chips on the change in the level of free S[O.sub.2] in a model solution similar to wine (n = 3).
Caption: Figure 6: Impact of the increasing alcohol concentration (% volume) on the change in the level of free S[O.sub.2] in a model wine solution supplemented with oak chips (10 g/L) (n = 3).
Caption: Figure 7: Total quantity of S[O.sub.2] added to the wine after five months of aging in tank (50 HL) in order to maintain the free S[O.sub.2] at 30 mg/L, and the concentration of MND in the wine according to the quantity of staves.
(1.) Bakour, R. 2003 "Influence de I'espece et de la provenance des deux principaux chenes francais (Quercus robur L, Quercus petraea Liebl.) sur la structure anatomique et les proprietes physiques du bois de merrain." ENGREF, Nabcy. 250 pp.
(2.) Feuillat, F. 1993 "Variabilite de quelques proprietes du bois de chene (Quercus robur L. et Quercus petraea Liebl.) susceptible d'influencer la qualite des vins. In: Elevage des vins en futs de chene," Ed. Y. Glories (Edition CEPS: Melun) pp. 40-49.
(3.) Feuillat, F. 1997 "Contribution a l'etude des phenomenes d'echanges bois/vin/atmosphere a I'aide d'un 'fut' modele: relations avec 1'anatomie du bois de chene." ENGREF, Nancy. 396 pp.
(4.) Feuillat, F. and R. Keller. 1994 "Structure anatomique et porosite du bois des chenes rouvre et pedoncule--Consequence sur 1'utilisation en tonnellerie." Rev. Oenol. 71, 13-18.
(5.) Glabasnia, A. and T. Hofmann. 2005 "Sensory-directed identification of taste-active ellagitannins in American white oak (Quercus alba Lj and European oak wood (Quercus robur Lj and dose/activity-Studies in bourbon whiskey and oakmatured red wines." J. Agric. Food Chem. 54, 3380-3390.
(6.) Glabasnia, A. and T. Hofmann. 2007 "Identification and sensory evaluation of dehydroand deoxy-ellagitannins formed upon toasting of oak wood (Quercus alba Agric. Food Chem. 55, 4109-4118.
(7.) Marchai, A., P. Waffo-Teguo, E. Genin, J.M. Merillon, and D. Dubourdieu. 2011 "Identification of new natural sweet compounds in wine using centrifugal partition chromatography-gustatometry and Fourier transform mass spectrometry." Analytical Chem. 83, 9629-9637
(8.) Mirabel, M., R. Teissier du Cros, D. de Beauregard, F. Balu, W. Giraud, and M. Comtat. 2011 "Aging wine in barrels: is there a link between oak wood grain size and porosity?" In: Actes de colloques du 9e symposium international d'oenologie de Bordeaux (Dunod: Bordeaux).
(9.) Moutounet, M., B. Saint Pierre, J.P. Micaleff, and J. Sarris. 1994 "Causes et consequences de micro deformations des barriques au cours de 1'elevage des vins." Rev. Fr. Oenol. 74, 34-39.
(10.) Pons, A., V. Lavigne, P. Darriet, and D. Dubourdieu. 2011 "Determination of 3-methyl2,4-nonanedione in red wines using methanol chemical ionization ion trap mass spectrometry." J. Chromatogr. A 1218, 7023-7030.
(11.) Pons, A., V. Lavigne, P. Darriet, and D. Dubourdieu. 2013 "Role of 3-Methyl-2,4nonanedione in the Flavor of Aged Red Wines." J. of Agric. Food Chem. 61,7373-7380. doi: 10.1021/jf400348h.
(12.) Prida, A. and P. Chatonnet. 2010. "Impact of chemical composition of oak extractives on the olfactory perception of barrel aged wines." Am. J. Enol. Vit. 61,408-413.
(13.) Ribereau-Gayon, P., Y. Glories, A. Maujean, D. Dubourdieu. 1998 "Traite d'Oenologie. Tome 2: Chimie du vin, stabilisation et traitements." (Dunod, Paris).
(14.) Spillman, P., M. Sefton, and R. Gawel. 2004 "The contribution of volatile compounds derived during oak barrel maturation to the aroma of a Chardonnay and Cabernet-Sauvignon wine." Aust. J. Grape Wine Res. 10, 227-235.
(15.) Vivas, N., H. Debeda, F. Menil, N. Vivas De Gaulejac, and M.F. Nonier. 2003 "Demonstration of oxygen transfer through the staves of the barrels by using an original system to measure wood porosity. First results." Sci. Ali. 23, 655-678.
Alexandre Pons, [1,2] Andrei Prida,  Benoit Verdier,  Philippe Darriet  and Denis Dubourdieu 
 Seguin Moreau, Cognac,  Universite de Bordeaux, Unite de recherche Oenologie, Institut des Sciences de la Vigne et du Vin, Villenave d'Ornon, France
Please note: Illustration(s) are not available due to copyright restrictions.
|Printer friendly Cite/link Email Feedback|
|Comment:||Barrel alternatives supply oxygen to wines during elevage.(WINEMAKING)|
|Author:||Pons, Alexandre; Prida, Andrei; Verdier, Benoit; Darriet, Philippe; Dubourdieu, Denis|
|Publication:||Wines & Vines|
|Date:||Feb 1, 2015|
|Previous Article:||Pruning of Sauvignon Blanc vines: influence of cane diameter on gross profit.|
|Next Article:||Non-subtractive approach to potassium tartrate stabilization.|