Oxidation and reduction reactions play a key role in wine maturation, throughout fermentation, in the tank or barrel, and during more lengthy bottle aging. A measure of the tendency of wine components to be oxidized or reduced would be a welcome aid to the winemaker in making decisions about how a wine is to be handled. The redox potential recorded at a platinum electrode has been presented as such a measure of the state of oxidation or reduction of a wine (Ribéreau-Gayon et al. 2000, Vivas et al. 1996). While claims have been made about a link between wine quality and a low redox potential, or that a particular value should be sought for different wine types (Dikanovic-Lucan et al. 1995), others have failed to find a link with quality and have questioned the significance of the measure for practical winemaking (Rankine 1998). Nevertheless a design which lessens past problems in measurement time and electrode stability is now available (Vivas et al. 1996) and was adopted by the OIV General Assembly in 2000 (Resolution Oeno 3), while the measure continues to be applied in studies of the state of oxidation or reduction of wines (Del Alamo et al. 2006).
In this paper, consideration is given to the main oxidation and reduction processes which occur at a platinum electrode in wine, particularly ethanol oxidation coupled to oxygen or proton reduction, which indicates that the measure has limited value in wine science. This point will be made through an exposition of the standard theory behind redox potential measurements in solutions of metals ions and using results for electrochemical studies of wines at a range of electrode materials (Tomlinson and Kilmartin 1997, Kilmartin and Zou 2001).
Redox titrations of metal ions
The use of platinum electrodes to follow the course of redox titrations is commonly outlined in Analytical Chemistry textbooks (e.g. Skoog et al. 2004). In the case of the oxidation of Fe2+ by Ce4+:
the potential (E) leading up to the end-point is given by the Nernst equation for the iron species in relation to the electrode potential (Eo) on the standard hydrogen scale (she):
allowing the titration curve in Figure 1 to be constructed. Conversely, at any stage leading up to the end-point, the measured redox potential of the solution can be used to determine the ratio of Fe2+ to Fe3+, and in this sense the state of oxidation or reduction of the iron species – on the other hand it is the quantity of Fe2+ remaining which determines how much Ce4+ oxidant can be consumed.
Figure 1. Change in potential at a platinum electrode during the redox titration of a solution containing Fe2+ ions using Ce4+ as an oxidant.
When the redox potential is taken in a solution containing both Fe2+ and Fe3+ ions there is no net current flowing at the platinum electrode, but the situation is far from static. The redox potential can be viewed as the point at which the currents due to the oxidation of Fe2+ and the reduction of Fe3+ are equal and opposite, as illustrated in Figure 2. In this figure the magnitude of the current is seen to increase exponentially for a slow increase (for the oxidation) or a decrease (for the reduction) in the potential. The absolute value of the current for the potential at which the currents are equal and opposite is called the exchange current density, although no net change in Fe2+ or Fe3+ concentration occurs (Bard and Faulkner 2001). The position of each curve for current against potential depends upon the amount of Fe2+ and Fe3+ present, and the lower potential obtained in Figure 2b versus Figure 2a illustrates the effect of an increased concentration of Fe2+.
|Figure 2. Schematic illustration of current versus potential curves for Fe2+ oxidation and Fe3+ reduction, and the determination of the redox potential for equal and opposite currents in the case of (a) a lower concentration of Fe2+, and (b) a higher concentration of Fe2+.
In a solution containing a mixture of metal ions and other active species such as redox indicator dyes (e.g. ferroin), which rapidly and reversibly exchange electrons with other metals, the concentrations of the oxidized and reduced forms of each metal ion adjust so that the potentials given by their respective Nernst equations is the same. In this ideal situation (equilibria involving metal precipitates and complex ions complicate the situation further), the redox potential of a solution of metal ions, such as a soil water, should provide useful information about the state of oxidation or reduction of the solution, but this may be limited to anaerobic solutions (Eshel and Banin 2002) and those dominated by iron (Grundl and Macalady 1989).
When the oxidation and reduction processes which occur at an electrode involve different species, net chemical changes can occur. This is seen when metals such as iron or steels are inserted into an aqueous solution - now the electrode potential provides a commonly used measure of the rate of corrosion (Crow 1994). The “mixed” potential which develops in this case, also called the open circuit or corrosion potential, occurs at the point at which the currents due to metal oxidation and the reduction of protons or oxygen is equal and opposite, as illustrated in Figure 3. Larger positive currents are seen when oxidation of a more active metal (such as iron versus a stainless steel) couples with a higher rate of reduction processes at a lower potential. In a given solution, a lower corrosion potential thus points to a higher rate of corrosion.
Figure 3. Schematic illustration of current versus potential curves for metal oxidation and coupled O2 or proton reduction, and the determination of the “mixed” or corrosion potential for equal and opposite currents in the case of (a) a less active metal such as a stainless steel, and (b) a more active metal such as iron.
Electrochemistry of wine
Electrochemical studies of platinum, gold and carbon electrodes in wine have led us to conclude that the redox potential in wine should be viewed as the point at which particular oxidation and reduction processes at the electrode are coupled, rather than as a static (equilibrium) point for all of the species present.
As part of a research project to develop a more rapid and reproducible redox potential measure we considered the use of high surface area platinized (black) platinum as an alternative to the polished (bright) platinum electrodes which have been generally used in wine (Tomlinson and Kilmartin 1997). We can also note that platinized platinum is used as the standard electrode (in a 1 M acid solution with hydrogen gas bubbling over at 1 atmosphere pressure) to provide the universal scale of electrode potentials, due to the rapidity with which the potential is established and its stability (Skoog et al. 2004). Rather than produce the same reading, platinized electrodes gave redox potential values well below those recorded at bright platinum electrodes, with a consistent decrease seen in potential as more platinum black was deposited on the electrode surface (Tomlinson and Kilmartin 1997). Once an initial value was established a drift in the potential to higher values was seen to occur over the course of several hours as the electrode was poisoned by wine components. The redox potential at platinized platinum of 21 wines (with no exposure to the air) was also strongly correlated (r2 = 0.87) with the pH of the wine. Further trends often noted in previous studies on the redox potential of wine included a marked increase in potential with aeration of the wine, while wine stirring led to an increase in the potential by a few mV.
In a further study redox potentials at bright platinum electrodes were compared with values at two other “inert” electrodes, namely gold and glassy carbon (Kilmartin and Zou 2001). Once again, rather than produce the same reading, the potentials at gold were typically 40 mV higher than those at platinum, and a further 40 mV higher again at glassy carbon. Little difference was seen between red and white wines, indicating that the polyphenols present in red wines (which are reducing agents as part of their antioxidant activity) are not major determinants of the redox potential.
Further insight into the processes occurring at the electrodes is given by scanning the potential away from the redox potential and recording the current response. Typical voltammograms obtained in a Sauvignon blanc white wine are shown in Figure 4 for different electrode materials. The cathodic currents for negative potential sweeps (likely to be due to proton and oxygen reduction) were larger on platinum than gold, and smaller again at the carbon electrode; upon purging with air the currents at the platinum electrode increased markedly, with more moderate increases at the other electrodes. Scanning the potential in the positive direction also led to higher currents at the platinum electrode, until a potential of over 600 mV (she) was reached and the carbon electrode gave the highest current. At this point the oxidation of catechol-containing polyphenols comes into play which occurs readily at the carbon electrode (Kilmartin et al. 2002). An oxidation current beyond 250 mV was also seen at platinum in a model wine solution containing 12% (v/v) ethanol and 0.033 M tartaric acid with added NaOH to give pH 3.6 (but with much higher current values than in the white wine where further wine species adsorb onto the platinum blocking active electrodes sites), indicating that the oxidation of ethanol will occur at platinum for redox potential values typically seen in wines (Kilmartin and Zou 2001). The current at platinum in the model wine solution was greatly lowered in the presence of SO2, where the adsorption of sulfur-containing species onto the electrode surface affects the efficiency of ethanol oxidation – the influence of adsorbates adds a further level of complexity to redox potential values. On the other hand the current at a carbon electrode in a model wine solution was negligible until much higher potentials (c. 1000 mV) were applied.
Figure 4. Experimental voltammetry curves taken at 100 mV s-1 of a deaerated Sauvignon blanc wine taken at platinum, gold and glassy carbon electrodes.
The redox potential of wine as a mixed potential
The different redox potential readings given at the various electrode materials indicate that there is no unique value for a given wine – instead different electrodes are catalytically active for certain species in wine to varying extents. Ethanol is oxidized very readily at platinum making it a very effective electrode material in fuel cell applications involving small alcohols (Zhou et al. 2003). Likewise protons and oxygen are reduced very efficiently at platinum. Further, the standard potential for the oxidation of ethanol to ethanal (pH = 0 and equal concentrations of the two species) has been determined to be c. 220 mV (she) (Clark 1960), which at pH 3.5 converts to +15 mV, or with an ethanol concentration 1,000 times that of ethanal, a value of –70 mV is obtained. Given that redox potential values in bottled wine are higher than this, typically from +100 to 300 mV at bright platinum (Dikanovic-Lucan et al. 1995), and down to 0 mV at pH 3.5 for platinized platinum (Tomlinson and Kilmartin 1997), the oxidation of ethanol is to be expected at platinum in wine at these potential values. On the other hand, these potentials are too low for the oxidation of wine polyphenols (Kilmartin et al. 2002). While other species in wine, including metal ions, may also be oxidized and reduced at the redox potential, the high levels of ethanol and protons, and at times oxygen, and their high electroactivity, mean that the redox potential in wine can be explained as a mixed potential involving these species. The concept of mixed potentials in potential readings at platinum electrodes has been further outlined by Durliat and Comtat (2005), with a consideration of the influence of dissolved oxygen.
In Figure 5 the potential at point (a) represents the coupling of ethanol oxidation and proton reduction at a bright platinum electrode – this represents the case in a wine bottle with very low levels of dissolved oxygen. When a platinum black electrode is used instead, the currents for ethanol oxidation are much higher at each potential value, and the mixed redox potential occurs at a lower value (b) than for bright platinum (the curve for proton reduction will also have shifted to more negative current values, but we can assume that the effect on ethanol oxidation is greater). The effect of an increase in pH upon both ethanol oxidation and proton reduction at any particular potential is to make the ethanol oxidation current more positive and the proton reduction less negative, which has the effect of moving the mixed potential to lower values. The strong linear correlation between the redox potential with pH seen at platinum black electrodes (and a weaker correlation at bright platinum) is consistent with this expectation and suggests that the reading was dominated by the ethanol and acid content (Tomlinson and Kilmartin 1997). When oxygen is introduced into a wine, a large current due to the reduction of oxygen can couple with a larger current for ethanol oxidation (Figure 5 point (c)) to give a higher redox potential reading. On the other hand stirring may increase the reduction currents preferentially, leading to the small increase in potential observed experimentally.
Figure 5. Schematic illustration of current versus potential curves for ethanol oxidation and coupled proton (a,b) or O2 (c), and the determination of the redox potential for equal and opposite currents in the case of bright platinum (a,c) and platinum black (b) electrodes.
In the absence of a platinum catalyst, the oxidation of ethanol in wine does not proceed rapidly, but appears to require the coupled oxidation of polyphenols and the presence of iron and copper as catalytic ions (Wildenradt and Singleton 1974, Danilewicz 2007). Given that polyphenols are the major substrates of interest for wine oxidation, but are not oxidized at platinum for redox potential values commonly found in wine, the significance of redox potential measurements for wine aging must be questioned. At best it can provide a broad measure of the oxygen content of a wine or a must undergoing fermentation (Berovic et al. 2003). A more accurate measurement of dissolved oxygen can always be made using specific probes such as the Clark electrode (Vidal et al. 2003). Many statements in enology texts in which the redox potential is mentioned will remain valid if they are replaced with statements about oxygen levels. For example, the claim that a low redox potential leads to the production of reduced sulfur off-odours in some wines (Jackson 2000), can be restated as (very) low levels of dissolved oxygen lead to the production of reduced sulfur off-odours. The change in redox potential values during fermentation can likewise be restated as changes in oxygen levels.
A more promising application of redox potentials with beverages has been the recent application of potential measurements at a platinum electrode to the ‘fermentation’ of teas, in which the transformation of green tea polyphenols into black tea theaflavins and related compounds was successfully monitored (Chen et al. 2007). In the absence of ethanol oxidation at the platinum electrode, the ratio of polyphenol catechol and galloyl groups to the oxidized quinone forms may well be responsible for the shift in redox potential values observed during tea fermentation.
The redox potential in biological systems has also been criticized because of the lack of reversible equilibrium systems (Kohen and Nyska 2002). These authors instead promote a measure of the “reducing power” given by the concentration of reducing equivalents present and determined using voltammetry at an inert electrode. Returning to the example of the oxidation of Fe2+by Ce4+, this is equivalent to the amount of oxidant required to complete the titration, as opposed to the ratio of Fe2+ to Fe4+ at any particular point. A measure of this sort is obtained for wine in the Folin-Ciocalteu assay for total polyphenols and by the intensity of peaks given by voltammetry (Kilmartin et al. 2002, Piljac et al. 2004, Rodrigues et al. 2007, Petrovic 2009, Makhotkina and Kilmartin 2009). The oxidisability of the wine in these cases is related mainly to the polyphenol content, and the key role polyphenols play in wine maturation.
Bard A.J., Faulkner L.R. Electrochemical methods. Fundamentals and applications. John Wiley & Sons, New York, 2001.
Berovic M., Mavri J., Wondra M., Kosmerl T., Bavcar D. «Influence of temperature and carbon dioxide on fermentation of Cabernet Sauvignon must». Food Technology and Biotechnology 2003; 41: 33-359.
Chen P.C., Chang F.S., Chen I.Z., Lu F.M., Cheng T.J., Chen R.L.C. «Redox potential of tea infusion as an index for the degree of fermentation». Analytica Chimica Acta 2007; 594: 32-36.
Clark W.M. Oxidation-reduction potentials of organic systems. Williams & Wilkins Co.: Baltimore, 1960.
Crow D.R. Principles and applications of electrochemistry. Blackie Academic & Professional, London, 1994.
Danilewicz J.C. «Interaction of sulfur dioxide, polyphenols and oxygen in a wine-model system: central role of iron and copper». American Journal of Enology and Viticulture 2007; 58: 53-60.
Del Alamo M., Nevares I, Carcel M. «Redox potential evolution during red wine aging in alternative systems». Analytica Chimica Acta 2006; 563: 223-228.
Dikanovic-Lucan Z, Palic A. «Le potential redox dans le processus de vinification : un bilan». Bulletin OIV 1995; 775-776: 762-779.
Durliat H, Comtat M. «Critical evaluation of potentiometric redox titrations in enology». Analytica Chimica Acta 2005; 545: 173-181.
Eshel G, Banin A. «Feasibility study of long-term continuous field measurement of soil redox potential». Communications in Soil Science and Plant Analysis 2002; 33: 695-709.
Grundl T.J, Macalday D.L. «Electrode measurement of redox potential in anaerobic ferric/ferrous chloride systems». Journal of Contaminant Hydrology 1989; 5: 97-117.
Jackson R.S. Wine Science. Principles, practice, perception. Academic Press, San Diego, 2000.
Kilmartin P.A, Zou H. «The effect of electrode material on the measured redox potential of red and white wines». Electroanalysis 2001; 13: 1347-1350.
Kilmartin P.A., Zou H, Waterhouse A.L. «Correlation of wine phenolic composition versus cyclic voltammetry response». American Journal of Enology and Viticulture 2002; 53: 294-302.
Kohen R., Nyska A. «Oxidation of biological systems: oxidative stress phenomena, antioxidants, redox reactions, and methods for their quantification». Toxicologic Pathology 2002; 30: 620-650.
Makhotkina O., Kilmartin P.A. «Uncovering the influence of antioxidants on polyphenol oxidation in wines using an electrochemical method: cyclic voltammetry». Journal of Electroanalytical Chemistry 2009; 633: 165-174.
Petrovic S.C. «Correlation of perceived wine astringency to cyclic voltammetric response». American Journal of Enology and Viticulture 2009; 60: 373-378.
Piljac J., Martínez S., Stipcevic T., Petrovic Z., Metikos-Hukovic M. «Cyclic voltammetry investigation of the phenolic content of Croatian wines». American Journal of Enology and Viticulture 2004; 55: 417-422.
Rankine B. Making good wine. A manual of winemaking practice for Australia and New Zealand. Sun Pan Macmillan, Melbourne, 1998.
Ribéreau-Gayon P., Glories Y., Maujean A., Dubourdieu D. «Aging red wines in vat and barrel: phenomena occurring during aging». A: Handbook of Enology. Vol. 2. Wiley, Chichester, 2000: 353-391.
Rodrigues A., Silva Ferreira A.C., Guedes De Pinho P., Bento F, Geraldo D. «Resistance to oxidation of white wines assessed by voltammetric means». Journal of Agricultural and Food Chemistry 2007; 55: 10557-10562.
Skoog D.A., West D.M., Holler F.J, Crouch S.R. Fundamentals of Analytical Chemistry. Thomson: Australia, 2004.
Tomlinson J.W, Kilmartin P.A. «Measurement of the redox potential of wine». Journal of Applied Electrochemistry 1997; 27: 1125-1134.
Vidal J.-C., Boulet J.-C, Moutounet M. «Oxygenation of wine during processing treatments. Data from cellar operations». Part 2. Revue Française d’Oenologie 2003; 201: 32-38.
Vivas N., Glories Y., Bertrand A., Zamora, F. «Principe et méthode de mesure du potentiel d’oxydoréduction dans les vins». Bulletin OIV 1996; 785-786: 617-633.
Wildenradt H.L, Singleton V.L. «Production of aldehydes as a result of oxidation of polyphenolic compounds and its relation to wine aging». American Journal of Enology and Viticulture 1974; 25: 119-126.
Zhou W., Zhou Z., Song S., Li W., Sun G., Tsiakaras P, Xin Q. «Pt based anode catalysts for direct ethanol fuel cells». Applied Catalysis B : Environmental 2003; 46: 273-285.