C Pascal1, N Champeau1, E Charpentier1, E Brenon1, JB Diéval1, S Vidal , M Moutounet2
1 Vinventions, team enologico, Francia
2Consulente freelance, Montpellier, Francia
Abstract
The longevity of wines depends on their oxidative evolution, which is more or less rapid under certain conditions, and on the acceptability of the presence of signs of oxidation for a given organoleptic profile. Determining the susceptibility of wines to oxidation, i.e. the risk of oxidative notes appearing on contact with oxygen, is of practical interest in order to better control vinification, tank/wood ageing and bottle ageing, and to consciously choose the operations carried out on wines.
The objective of this work is to develop a method for assessing the sensitivity of wines to oxidation. To this end, the air resistance test method was implemented with electrochemical monitoring of a series of wines whose sensitivity to oxidation had previously been classified by experienced oenologists. The variation in the electrochemical signal of the wine between contact with air and after 2 hours of contact was used to classify samples sensitive or, on the contrary, resistant to oxidation. To understand the origin of these signal variations, electrochemical measurements were made on solutions of catechin, chosen here as a model phenolic compound, in the presence of Fe(II) or Fe(III) ions under anoxic conditions. The greater or lesser ability of semiquinones, the oxidation products of phenolic compounds, to form complexes with Fe(II) ions could prevent the latter from reacting according to the Fenton reaction and thus prevent the formation of oxidation markers in wines.
One of the winemaker’s main challenges is to assess a wine’s ability to resist the appearance of signs of oxidation, so as to adapt the technical route to its target profile and lifespan.
In practice, the signs of oxidation in wines are described as a change in colour with the appearance of orange hues, an aromatic change described as a loss of varietal notes (boxwood, passion fruit) and/or the appearance of so-called oxidative notes (honey, wax, chocolate, stone, dried fruit, dried grass, nuts, etc.), and sometimes a change in the taste balance with an increase in astringency and/or a loss of fat. These different signs of oxidation may appear successively or simultaneously, without the appearance of one being indicative of the possible appearance of another.
The point at which a wine is considered oxidised depends on the acceptability of these characteristics in relation to the type of product and the winemaker’s intent. For example, the presence of dried fruit notes in naturally sweet wines (e.g. raisin wines) is acceptable, if not desirable, but not in dry Sauvignon Blanc wines of certain appellations, which must retain citrus or exotic fruit notes.
Assessing a wine’s ability to resist oxidation is equivalent to determining through a predictive test its capacity for ageing and its longevity, provided it is stored under the right conditions.
In wine, certain molecules are known for their antioxidant capacities, e.g. phenolic compounds or antioxidants added during the process, such as sulphites. Consequently, the assessment of the concentration of antioxidants has been considered as being able to predict the longevity of a wine. The evaluation can be carried out using different methods, which generally consist of assessing the wine’s ability to capture radicals, to react with different types of oxidants (DPPH test, ABTS, FRAP, Folin Ciocalteu analysis) or to oxidise on electrodes of a certain type (voltammetry). Finally, the antioxidant content of a wine has been shown to be strongly correlated with the concentration of phenolic compounds (Büyüktuncel et al. 2014; Romanet et al. 2019; de Beer et al. 2004, 2006; Kilmartin 2001).
Several studies have shown that the concentration of antioxidants decreases as wines oxidise (Rodrigues et al. 2007, Ugliano 2013 and 2019). In general, older wines contain fewer antioxidants than younger wines (Rodrigues et al. 2007), but the variability of vintages and winemaking processes (especially the extraction steps of phenolic compounds) means that the age of a wine cannot be determined by its antioxidant concentration (Romanet et al. 2019). Finally, the low antioxidant content of a wine has been correlated with the presence in the sensory profiles of notes associated with the oxidative evolution of wines (Romanet et al. 2019). However, these studies do not show that antioxidant content is predictive of a wine’s ability to resist oxidation.
To make a connection with oenological practices, if the intrinsic content of phenolic compounds (the main contributors to antioxidant content) were an indicator of a wine’s ability to resist the appearance of oxidative notes, oenologists would use the Folin index or even the total polyphenol index for this purpose. In contrast, oenologists currently assess this capacity empirically by tasting before and after contact with air (air resistance test). Changes in the sensory profile of the wine are assessed, in particular the appearance of notes considered oxidative.
This article presents the development of a method to determine the sensitivity of wines to oxidation. Wine samples were empirically selected by oenologists and subjected to an air resistance test monitored by line-scan voltammetry. A different evolution of the voltammetric signal was observed for wines empirically categorised as oxidation-resistant and oxidation-sensitive. Lastly, the voltammetric signal of catechin in hydroalcoholic solution in the presence and absence of Fe(II) and Fe(III) ions in an inert atmosphere was recorded, allowing hypotheses to be advanced on the involvement of phenolic compounds-Fe(II) or Fe(III) ions in the recordings obtained and on the mechanisms underlying this test.
Materials and methods:
134 wine samples (87 reds, 32 whites and 15 rosés from different French and Spanish regions) were selected at the end of alcoholic or malolactic fermentation by experienced oenologists (cellar managers or consultant oenologists with several years of experience in the production area in question and who generally include micro-oxygenation in some of their technical protocols). For each sample, the oenologist empirically assessed the wine’s risk of premature oxidation through tasting, classifying it as resistant or sensitive to oxidation. Wine samples were taken from 750 ml bottles previously inerted, filled to overflowing and sealed with a screw cap. These wines were subjected to an air resistance test by taking 50 ml of sample with a volumetric pipette and placing it in a 125 ml bottle. Linear scanning voltammetry measurements (0-600 mV, 100 mV/s, 10 mV steps) were carried out when the bottle was opened and then after 2 hours, using a potentiostat (WQS Polyscan, Vinventions), on moulded electrodes (carbon working electrode, Vinventions). The potentials are expressed in relation to an Ag/AgCl reference electrode.
The solutions of catechin (1150.0 mg/L and 1165.0 mg/L), iron sulphate (10.3 mg/L) and iron chloride (10.7 mg/L) were prepared in 13% ethanol containing 4.0 g/L tartaric acid and adjusted to pH 3.3 with 1N HCl then 0.1N. These stock solutions were deaerated by bubbling nitrogen and placed in an inert fume hood by flushing with nitrogen. The % O2 in the hood was monitored by a Nomasense O2 P300 (Vinventions) and maintained at less than 2% throughout the test. The intensity-potential curves of the freshly prepared stock solutions and the 50/50 v/v catechin/Fe ion mixture were recorded in triplicate as previously described using line-scan voltammetry.
Results
Empirically, the air resistance test is only carried out by winemakers on samples that have not been in contact with oxygen in the preceding days. 2To approximate these conditions, the air resistance tests were only carried out here on samples stored for 3 weeks prior to measurement in the absence of O . 58% of the samples selected by the winemakers were empirically categorised as resistant to oxidation.
Comparison of the intensity/potential curves obtained on a wine immediately after opening the sample bottle with those obtained after 2 hours of contact with air revealed several changes in the signals.
For 47 of the 134 wines, a significant increase in intensity (greater than 5%, the method’s measurement uncertainty) in the range of 100-600 mV was observed on the measurement after 2 hours of contact with air (Figure 1 (a)). The 47 samples in this group had all been characterised as oxidation-resistant by the oenologists. None of the ‘oxidation-sensitive’ samples showed this behaviour.
Figure 1: Measured intensity/potential curves of a wine when the sample is placed in contact with air (T0) and then after 2 hours of contact. (a) samples described as oxidation-resistant by oenologists (b) samples described as oxidation-sensitive by oenologists. (a1) e (b1): red wines. (a2) e (b2): vini rosati. (a3) e (b3): white wines.
For 38 of the 147 samples, however, a significant decrease in intensity (greater than 5%, the method’s measurement uncertainty) in the range of 100-600 mV was observed on the measurement taken after 2 hours of contact with air (Figure 1 (b)). These 38 wines had all been characterised as sensitive to oxidation by the oenologists. None of the samples characterised as ‘oxidation-resistant’ showed this behaviour.
For the remaining 49 wines, the variation in intensity between the initial measurement and that recorded after 2 hours of contact with air was not significant. These wines could potentially have less oxidative behaviour than the 2 categories described above, making the test less sensitive to their discrimination. Moreover, no intensity scale for resistance or sensitivity to oxidation was required in the categorisation of the wines, which means that it was not possible to distinguish this category of wines, subsequently defined as ‘intermediate’.
The test seems to be able to classify wines:
those that show an increase in signal between the initial measurement and that taken after 2 hours of contact with air appear to be systematically resistant to oxidation,
those that show a decrease in signal between these two measurements appear to be systematically sensitive to oxidation.
Impact of complexation of Fe(II) and Fe(III) ions by phenolic compounds on voltammetric signal
In an attempt to explain the chemical mechanism behind the test, further work was carried out in a model solution, in an inert environment. For information, it has been shown (Kilmartin et al. 2002, Ugliano et al. 2019) that carbon electrode voltammetry can be used to analyse phenolic compounds.
It was also considered that the oxidation mechanism in wines (Figure 2) underlies the redox cycle of the Fe(II)/Fe(III) pair (Ribereau Gayon 1931, Danilewicz 2018, du Toit et al. 2006, Waterhouse and Laurie 2006).
Furthermore, these same Fe(II) and Fe(III) ions are known to form complexes with phenolic compounds, particularly structures with a di- or tri-hydroxylated B ring (Amorim Porfírio et al 2014). 2+2+Recently, several studies (Le Nest et al. 2004, Amorim Porfirio et al 2014) have shown that complexes between phenolic compounds and some metal ions (Zn , Fe ) can exhibit intensity/potential curves different from those of the phenolic compound in the absence of metal ions. The signal of the complex is often higher. In these studies, several model phenolic compounds (morin, quercetin, fisetin, catechin, chrysin, taxifolin, etc.) were analyzed, showing different impacts of complexation on their respective voltammetric signals.
However, these studies were conducted at neutral pH to reproduce physiological conditions and do not show the signal of the phenolic-Fe(III) compound complex. Because wine has an acidic pH and organic acids, particularly tartaric acid, are known to complex Fe(II) and Fe(III) ions, the voltammetric signal of catechin, chosen as the model phenolic compound for this study, was recorded in a model wine solution (13% hydroalcoholic solution, 4 g/L tartaric acid, pH 3. 5) before and after the independent addition of Fe(II) and Fe(III) ions in an inert atmosphere to prevent the establishment of the oxidation mechanism (performed in an inert fume hood with nitrogen flow, previously deaerated solutions). The independent addition of Fe(II) and Fe(III) ions to the catechin solution produced changes in the potential intensity curve recorded for catechin alone (Figure 3), suggesting that catechin-Fe(II) and Fe(III) ion complexation did indeed take place under the model wine conditions and impacted the voltammetric signal. It was observed that the voltammetric signal of the catechin-Fe(II) complex was close to that of catechin (Figure 3(b)), as observed by Porfirio et al (2014), or even slightly lower between 450 and 500 mV. In contrast, the signal of the catechin-Fe(III) complex was significantly higher than that of catechin alone between 450 and 600 mV (Figure 3 (a)).
Figure 3: Potential intensity curves for : (a) catechin (1150mg/L, 13% hydroalcoholic solution, pH 3.3, 4g/L tartaric acid) in the absence or presence of Fe(III) ions (5 mg/L) (b) catechin (1165mg/L, 13% hydroalcoholic solution, pH 3.3, 4g/L tartaric acid) in the absence or presence of Fe(II) ions (5 mg/L).
The complexation of Fe(II) ions makes them less available to participate in the Fenton reaction.
Accordingly, the following hypotheses were put forward to explain the evolution of the intensity-potential curves of a given wine after 2 hours of air contact.
The first measurement is made immediately after opening the sample, which has been stored for a minimum of 3 weeks under anoxic conditions after being taken under inert conditions. Under these conditions, iron ions are a priori mainly in the Fe(II) form (Danilewicz, 2016 and 2018, Nguyen & Waterhouse 2019). The voltammetric signal therefore corresponds a priori to that of phenolic compounds or phenolic compound-Fe(II) complexes, depending on the ability of the phenolic compounds present to complex Fe ions.
The first measurement is made immediately after opening the sample, which has been stored for a minimum of 3 weeks under anoxic conditions after being taken under inert conditions. Under these conditions, iron ions are a priori mainly in the Fe(II) form (Danilewicz, 2016 and 2018, Nguyen & Waterhouse 2019). The voltammetric signal therefore corresponds a priori to that of phenolic compounds or phenolic compound-Fe(II) complexes, depending on the ability of the phenolic compounds present to complex Fe ions. This would mean that wines classified as oxidation-resistant by this test have a pool of phenolic compounds capable of complexing Fe ions under the environmental conditions of these wines. The effect of exogenous metal chelators to stabilize wines against oxidation was considered by Kreitman et al. (2013) following a study in model solutions. In addition, in a physiological context, Lopes et al. (1999) suggested that tannic acid limits the oxidation of 2-desoxyribose, a component of DNA, due to its ability to chelate Fe ions, making Fe(II) ions unavailable to participate in the Fenton reaction and thus limiting the formation of HO° hydroxyl radicals. In the case of resistant wines, it is possible that the Fe(II) ions (Figure 4) remain complexed to the semichinone (Perron et al. 2009 and 2010), the oxidized form of the phenolic compound that has reduced an Fe(III) ion to an Fe(II) ion. There would be no continuation of the mechanism to a quinone form that reduced 2 Fe(III) ions, and the Fe(II) ions would remain complexed to the phenolic compounds. Perron et al. (2010) also suggested that the stability of the phenolic compound-Fe(III) complex would promote the autoxidation of complexed Fe(II). Fe(II) ions might therefore be less willing to react with hydrogen peroxide and give rise to the Fenton reaction, which would explain the lower occurrence of notes described as oxidative when these wines are in contact with oxygen.
On the other hand, when the signal of the wine decreased after 2 hours of contact with air, it is likely that the oxidation mechanism led to the formation of quinones, oxidized forms of phenolic compounds that cannot contribute to the voltammetric signal in the potential zone under consideration. The pool of phenolic compounds in the environmental conditions of these wines would not be able to maintain complexed Fe ions (Figure 4). The Fe(II) ions could then react with hydrogen peroxide to initiate the Fenton reaction (Singleton 1987) and give rise to very high-energy radicals, the HO° hydroxyl radicals that oxidize ethanol to ethanal, for example. Without being exhaustive, it has also been hypothesized that hydroxyl radicals give rise to the appearance of brown/yellow/orange pigments linked to condensation reactions of flavan-3-ols (tannins) in the presence of ethanal or pyruvic acid (Oszmianski 1996, Fulcrand 1996, Guyot 1996) or attack compounds with alcoholic functions leading to the formation of so-called oxidation aldehydes (methional, phenylacetaldehyde) (Nikolantonaki and Waterhouse 2012).
Conclusions:
In conclusion, the determination of a wine’s susceptibility to oxidation is of great interest to winemakers having the purpose of best matching the winemaking method and storage conditions to the desired profile of each product. The air resistance test is an empirical method to simply assess this resistance. When this test is monitored using linear voltammetry, the change in signal between the time of air contact and 2 hours of contact can be used to identify oxidation-resistant wines, which show an increase in voltammetric signal, and oxidation-sensitive wines, which show a decrease. These changes in signals could be related to the greater or lesser ability of wine phenolic compounds to complex Fe ions, with complexation leading to a higher voltammetric signal for the phenolic compound-Fe(III) complex and making Fe(II) ions less available to generate the Fenton reaction. In addition to the link between the results of voltammetric measurements and the empirical classification of wines by winemakers, white and red wines were subjected to this test, then bottled in a controlled manner and stored to verify the predictive character of the test. These studies, which confirm the possibility of assessing the sensitivity of wines to oxidation, will be published in a forthcoming article.
Amorim Porfírioa D., de Queiroz Ferreiraa R., Renata Malagutti A., Maíra Agostini Valle E., (2014) Electrochemical study of the increased antioxidant capacity of flavonoids through complexation with iron(II) ions, Electrochimica Acta, 141, 33-38.
Büyüktuncel E., E Porgalı, C Çolak, Comparison of Total Phenolic Content and Total Antioxidant Activity in Local Red Wines Determined by Spectrophotometric Methods Food and Nutrition Sciences, 5 (17), 1660-1667
Danilewicz J.C. (2007) Interaction of Sulfur Dioxide, Polyphenols, and Oxygen in a Wine-Model System: Central Role of Iron and Copper. Am. J. Enol. Vitic. 58:1, 53-60.
Danilewicz J.C. (2018) [Fe(III)]:[Fe(II)] Ratio and Redox Status of Red Wines: Relation to So-Called “Reduction Potential” Am. J. Enol. Vitic. 69:2, 141-147.
Danilewicz J.C. (2016) [Fe(III)]:[Fe(II)] Ratio and Redox Status of White Wines Am. J. Enol. Vitic. 67:2, 146-152.
De Beer D., J F. Harbertson, P A. Kilmartin, V Roginsky, T Barsukova, D O. Adams, A L. Waterhouse Phenolics: A Comparison of Diverse Analytical Methods Am J Enol Vitic. 2004, 55, 389-400
du Toit W.J., Marais J., Pretorius I.S. and du Toit M. (2006) Oxygen in must and wine: A review. S Afr J Enol Vitic 27, 76-94.
Kilmartin, P.A.; Zou, H.; Waterhouse, A.L. A Cyclic Voltammetry Method Suitable for Characterizing Antioxidant Properties of Wine and Wine Phenolics. J. Agric. Food Chem., 2001, 49, 1957–1965.
Kilmartin P.A., Zou H, Waterhouse A.L. (2002) Correlation of wine phenolic composition versus cyclic voltammetry response, Am. J. Enol. Vitic. 53, 294-302.
Kreitman G.Y., Cantu A., Waterhouse A.L., Elias R.J. (2013) The effect of metal chelators on the oxidative stability of model wine J. Agric. Food Chem. 61(39), 9480-7.
Lopes G.K., Schulman H.M., Hermes-Lima M. (1999) Polyphenol tannic acid inhibits hydroxyl radical formation from Fenton reaction by complexing ferrous ions. Biochim Biophys Acta. 1472(1-2),142-52.
Nguyen T.H., Waterhouse A.L. (2019) A Production-Accessible Method: Spectrophotometric Iron Speciation in Wine Using Ferrozine and Ethylenediaminetetraacetic Acid. J. Agric. Food Chem. 67, 2, 680–687.
Le Nest G., Caille O., Woudstra M., Roche S., Guerlesquin F., Lexa D. (2004) Zn–polyphenol chelation: complexes with quercetin, (+)-catechin, and derivatives: I optical and NMR studies. Inorganica Chimica Acta 357 (3), 775-784
Perron N.R., Brumaghim J.L. (2009) A review of the antioxidant mechanisms of polyphenol compounds related to iron binding. Cell Biochem Biophys. 53(2),75-100.
Perron N.R., Wang, H.C., DeGuire S.N., Jenkins M., Lawson M., Brumaghim J.L. (2010) Kinetics of iron oxidation upon polyphenol binding. Dalton Trans. 39, 9982-9987.
Ribéreau-Gayon J. (1931) Contribution à l’étude des oxydations et réductions dans les vins. Thèse Doctorat ès Sciences Physiques, Bordeaux.
Romanet R., C Coelho, Y Liu, F Bahut, J Ballester, M Nikolantonaki, R D. Gougeon The Antioxidant Potential of White Wines Relies on the Chemistry of Sulfur-Containing Compounds: An Optimized DPPH Assay Molecules 2019, 24(7), 1353
Rodrigues A., Silva Ferreira A.C., Guedes de Pinho P., Bento F., Geraldo D. Resistance to Oxidation of White Wines Assessed by Voltammetric Means J. Agric. Food Chem. 2007, 55, 10557–10562
Ugliano M., Dieval J-B, Tacchini P., Vidal S. Oxidation fingerprints of white wines by linear sweep voltammetry using screen printed carbon electrochemical sensors. 2013, In Vino Analytica Scientia, poster
Ugliano M., Pascal C., Diéval J-B., Vidal S., Wirt J., Bégrand, S. Une nouvelle approche voltamétrique pour l’analyse des polyphénols des raisins blancs et le suivi des opérations pré-fermentaires. Infowine 2019
Waterhouse A.L. and Laurie V.F. (2006) Oxidation of wine phenolics: A critical evaluation and hypotheses. Am J Enol Vitic 57, 306-313.
