Interview with Roger B. Boulton

“Both white wines and red wines can consume oxygen at similar rates”

Tanto los vinos blancos como los tintos pueden consumir oxígeno a tasas similares

María del Álamo e Ignacio Nevares

Which is the relevance of oxygen in wine, and how does it act?

The oxygen consuming reactions in wine have puzzled researchers and winemakers for more than a century. While it is widely known that the phenolic components form the main substrates, and that acetaldehyde is the main characteristic sensory product, it does not seem to be widely understood that the oxygen consumption is the rate-limiting reaction. The rates of this reaction however do not depend on the concentration of any phenol, or that of all phenols, are first order in oxygen concentration and catalyzed by ferrous ion, with different rates being observed in different wines at the same oxygen concentration. Studies of the time course of oxygen consumption as a function of temperature, of white wines, of model catechin solutions and of model wine with free SO2, have been shown to follow either pseudo-first order or first order kinetics with respect to oxygen concentration. This would occur only if some of the ferrous ion is quickly returned to the reduced state, so that its concentration is essentially constant, and not declining due to consumption. Under these conditions the rate constant would be related to the ferrous ion concentration, but the rate law would depend only on the oxygen concentration.


How does oxygen interact with the wine components?

The initial reaction step was proposed to be ferrous limited Professor Jean Ribereau-Gayon in 1933 and recent experiments have confirmed that ferrous ion not ferric ion is essential and that the nature of the phenol (catechin versus caffeic acid) is of minor importance. This has led to a proposed reaction sequence that forms the most plausible one to date.  It involves the reaction between ferrous ion with oxygen to form ferric ion and hydrogen peroxide, and that ferric ion reacts with the phenol to regenerate the ferrous ion and produce a quinone. The ferrous ion then reacts with hydrogen peroxide to produce the hydroxyl radical in the Fenton sequence and also with any remaining oxygen in the first sequence, both resulting in ferric ion formation. In the absence of oxygen, the hydroxyl radical reaction with ethanol yields the hydroxyl ethyl radical which reduces ferric ions back to ferrous ions and creates acetaldehyde. In the presence of oxygen the hydroxyl ethyl radical can react with oxygen to form the peroxyl radical, which on decomposition yields acetaldehyde and the hydroperoxyl radical. This radical decomposes into hydrogen peroxide and oxygen, repopulating these pools for further reaction. If acetaldehyde is formed when oxygen is present, the yield of aldehyde per peroxide can be 50% higher than when it is absent, so the yield would seem to be changing during the progression of these reactions.


Does oxygen impact to the same extent on red and white wines?

The hydroxyl radical is considered to be one of the most reactive entities and indiscriminate in its reaction chemistry and is considered to react with quenching species based on their abundance in solution. This is why ethanol is its main reactant. However, the quenching of hydroxyl radicals by phenolics, in particular flavonoids, cinnamates, flavonols and anthocyanins, was demonstrated by the Bors group in the early 90s. These reactions lead to lower the acetaldehyde formation from both oxygen and peroxide and hence less perceived oxidation. Recently it has been shown that caffeic acid is especially reactive in this way, giving further evidence that the influence of the phenolic composition is more related to the peroxide and radical reactions than to those involving oxygen. It has been expected that tannins from skin and seed are significant radical quenching agents, and maybe oak tannin fractions as well, even though they show little role in the oxygen and ferrous reactions. This is why both white wines and red wines can consume oxygen at similar rates yet the aldehyde production of white wines is much more obvious than that of red wines.


What does determine the oxygen consumption degree in wines? 

Since the ferrous ion reaction with oxygen is controlling the speed of reaction, the reactions involved in the regeneration of ferric to ferrous are of key importance. The picture in which only these two species are in a redox couple at wine pH is probably oversimplified, since in the presence of oxygen, at pH of 3.0 to 4.0, the predominant forms appear to be ferric hydroxide complexes (Fe(OH)++ and Fe(OH)2+) and ferrous. Other complexes involving iron and copper ions are expected to be pectin fractions, quercetin and some tannin fractions. The reactions in which ferric is reduced back to ferrous are not known and yet they seem to determine the extent to which a saturation of oxygen can be consumed in the first few hours.


Which is the role of SO2 in these oxygen-consuming processes?

The speed at which the quinone is returned to the phenol seems to dominate that of other reactions and Danilewicz has shown very clearly that sulfur dioxide greatly increases this return reaction, forming sulfate ion at the same time. A second role of free SO2 is also to react with hydrogen peroxide, again forming sulfate ion but competing for it with ferrous in the Fenton sequence.  Thus in the presence of free SO2, the rate of oxygen consumption is accelerated and yet the yield of acetaldehyde might be lowered, and any aldehyde formed will be bound. It is easy to see why for so long the addition of sulfur dioxide to wine was generally viewed as “stopping” oxidation since the oxygen disappeared quickly, and acetaldehyde formation was not detected sensorially. The return of the quinone to the phenol is why so little browning is observed even when large amounts of oxygen are delivered into a wine, compared with a saturation of two in a juice, where the enzyme reaction has far more extensive quinone condensation. The form of free SO2 involved in these reactions is in question with some suggestion of sulfite radicals and others of moleceular SO2 and needs to be studied further.


Which do you think that next steps in the oxigen - wine relationship research should be?

The rates of radical forming and quenching reactions need to be studied for all major phenolic components under wine conditions, not only the hydroxyl radical, but the hydroxyl ethyl, the peroxyl and the hydroperoxyl radicals. The quenching reactions in particular need to be considered with other wine components ranging from the organic acids, glycerol and glutathione,  characteristic sensory impact components related to varietal identity such as terpenes, pyrazines and thiols and both the major anthocyanins and their copigmentation cofactors.