Phenolic Chemistry And Winemaking
I’m sorry, but it has to be done. Yes, I will now attempt to render palatable the topic of phenolic chemistry, at least to open a door to its baffling collection of hexagonal pictograms guaranteed to drive the bravest enologist into fits of terror and ennui.
After a year of beating around the bush, it is time for my loyal readers to join with me and take on red wine’s defining process, the peculiar and counterintuitive reaction with which red wine builds structure, in so doing elevating its soulful resonance and graceful longevity. To comprehend why, we must descend Frodo-like into the depths of enology’s darkest recesses and attempt to drag its most daunting secrets into the light, hex signs and all.
Why have I suddenly turned so cruel? Because an appreciation of phenolic structure is key to the postmodern view of wine. This journey will provide basic tools that will allow us to move beyond the modern approach we all learned in school. Red wine is not a solution; moreover, the extent to which it deviates from solution behavior is a pretty good working definition of quality.1 Phenolics are the principal inhabitants of the non-solution world, which is where soulfulness lives. A grasp of their behavior will empower us to steer the course of wine development.
Put another way, red winemaking is a type of cooking related to sauce-making. Good béarnaise doesn’t taste like its ingredients (tarragon, shallots, vinegar, chervil); it tastes like, well, béarnaise. The structure, not the composition, determines the flavor. Curdle the sauce and it loses its integrative properties and tastes terrible. Likewise, the shape and size of suspended particles determines the wine’s sensory characteristics and, more importantly, its aesthetic impact.
(Hint: When I say “aesthetic impact,” I’m talking about cash flow.)
A clear picture of how wine behaves benefits winemakers of every stripe. I’m not pushing here for micro-ox, lees stirring or any other winemaking technique. No recipes will be forthcoming. But to make wine is to choose a path. Increasingly, great winemakers elect, when they can, to do nothing. But even if the practitioner in no way intervenes, the choice of a wine’s structure is still central to the art.
The phenols in young wine resemble those of cocoa and may remain dry and harsh, or through pathways similar to chocolate-making may transform into silky, visceral, flavor-carrying textures. The engine of this transformation is that very reductive strength that modern enology calls a defect. This force may be harnessed by the winemaker who is familiar with wine’s nature.
Are you with me? Then hang on tight and let’s get started.
What’s a “phenolic?”
The basic unit of all phenolics is phenol itself, which is simply a ring of six carbons with alternating double bonds (called a benzene ring) attached to oxygen and hydrogen (-OH) atoms. This very stable structure lies in a flat plane, allowing many phenols to be stacked compactly like ping pong paddles. Plants find phenols very handy, and grapes manufacture hundreds of different kinds for various purposes by attaching additional chemical groups to the ring.
There’s no need to worry about understanding all the details of phenolics in wine. Nobody does. Take the dizzying array of thousands of different phenolic compounds, then start to tack these monomeric units together like Legos, and shortly you end up with many millions of combinations as unique as snowflakes.
These polymers go on to form complicated structures in wine sometimes as big as a bacterial cell, no two alike. If your eyes are beginning to cross, you’re getting the picture. Wine structure does not lend itself to analysis. A full understanding of its diversity is neither possible nor useful.
What does prove handy is an overview of the parts of the process we can measure and control. I will present a model, however true or false it may eventually turn out to be, that has served postmodern winemakers well as a predictive tool.
The -OH group of phenol is mildly acidic, and can be persuaded in basic solution (about pH 9) to ionize by giving up its H+, resulting in a negatively charged -O- still attached to the benzene ring. This ion is called phenolate, and it has the ability to react with oxygen to form a double bond=O called a quinone (see graphic below).
Since wines are generally below 4.0 on the logarithmic pH scale, there is not much phenolate around—only about one molecule in 100,000. This means the reaction proceeds very slowly. Quinones absorb oxygen and bind SO2, but that’s about it. This isn’t a very interesting reaction, but a special case of it is far more consequential, the subject of this article.
Phenols repel water and love to ring-stack. When placed in water, they try to get away, either by evaporating (there are many phenolic aromatics common to wine—vanilla, clove, menthol and so on) or by aggregating, driven by water molecules into tiny beads called colloids, herded like cattle into tiny holding pens.
If there are enough phenolic colloids around, the aromatic compounds will enter them and ring-stack also, diminishing their aromatic impact. This phenomenon is called “aromatic integration,” and it explains why wines with good structure can contain large amounts of oak extractives, Brettanomyces-induced 4-ethyl phenol or vegetal aromatics like methoxypyrazines without deleteriously affecting the wine’s aroma.
Aromatic integration is a source of soulfulness in food, be it a béarnaise, a bisque or a Bordeaux. When aromas are integrated into a single voice, the food speaks to the soul the way a symphony does when the entire orchestra plays as one.
The core reaction
One type of phenolic plants find very useful in fruits is the vicinyl (or ortho-) diphenol. This is simply a phenol with two -OH groups attached to adjacent positions on the ring. This molecule has the magical ability, in reaction with oxygen, to attach itself to another phenol and recreate its original reactive diphenol structure. This odd trait allows it to react over and over, daisy-chaining to create long polymers. The browning of apples and bananas is the polymerizing of diphenols, which seal the fruit in case of injury .
It was a bright feather in UC Davis’s cap when Dr. Vernon Singleton in 1987 elucidated the mechanism of oxidative polymerization in all its bizarre aspects.2 He explained quite clearly why, contrary to all common sense, a young wine challenged with oxygen behaves homeopathically, increasing its reductive strength. He also showed why oxygen uptake is so strongly temperature-dependent, a critical morsel of cellar knowledge.
Grape skins are rich in a special type of phenols called flavonoids. Some of these are red-colored anthocyanins, but most are tannin building blocks that contain adjacent diphenols. When ortho-diphenols see oxygen, magic happens.
Without going into the details of Singleton’s explanation, the phenolic ring basically goes Rambo, hungrily searching out a nearby molecule (“R”) to attach to the ring in any unoccupied position. Choices include another diphenol, an anthocyanins pigment, a sulfide or a protein fragment—usually from lees. Once this bond is formed, the diphenol is recreated and can react again.
There’s one important difference. Because the R-substitution stabilizes its ionization, the phenolate is more stable, thus more favored, and its equilibrium pH shifts from 9.0 to 8.5. As a result, there is now one phenolate molecule in 50,000 instead of 100,000, so the reaction occurs twice as fast. In effect, the introduction of oxygen has made the wine hungrier for oxygen. Instead of oxidation, we get increased reductive strength.3
For those who claim there is no scientific support for homeopathy, it’s time to pay attention.
Since the reactive diphenol is recreated and now reacts faster, the cascade repeats again and again, resulting in a complex of random linkages. The shape of the resulting polymer is thus dependent on the concentration of the various reactants and their relative affinity for the reaction.
In effect, oxygen acts like a wire whisk, and the tannins, like egg whites, firm up into a rich, light structure similar to a meringue.
The ability for wines to consume oxygen varies unbelievably. Musts consume 10ppm per hour. Young reds consume 1ppm in a day. The same wine, post ML will consume 0.1ppm per day. After two years in barrel, it may consume 0.01ppm per day.
Young, unpolymerized anthocyanins are the key to textural finesse and aromatic integration. They are vulnerable to destruction until incorporated into oxidative polymers. Reductive winemaking results in clumsy red wines that display excessive oak, Brett and vegetal aromatics and coarse tannins.
Low cellar temperatures promote spoilage by diminishing red wine’s ability to protect itself from oxygen. Between 50ºF and 59ºF, oxygen uptake decreases five-fold. A 1ºF decrease in temperature alters O2 protective reactivity by one-third.
Early lees exposure destroys color and inhibits structuration, leading to dry tannin. Late lees introduction has the opposite effect, adding fatness and roundness to the palate.
Sulfite addition completely alters phenolic reactivity. Oxygen uptake rate drops ten-fold. Aldehyde is no longer available to stabilize anthocyanins. The major opportunity to enhance structure is after alcoholic fermentation and pre-ML.
The key role of color
The anthocyanins responsible for red color readily combine with tannins, which are similar in composition, with one important difference. Anthocyanins are not ortho-diphenols and cannot daisy chain, so they act like bookends, terminating the polymer. The more anthocyanins, the shorter the resulting polymers and the finer the tannins. Heavily colored wines like Syrah tend to handle their tannins better than light wines like Pinot Noir, which have a tendency towards dryness.
The more anthocyanins we start with, the shorter the polymers we end up with. Smaller polymers lead to smaller colloids. Not only do these feel softer and finer, they have higher surface area. This gives aromatic compounds more interface, easier access to ring stacking in the structure as opposed to appearing in the nose. That’s why the finer the texture of a sauce or a wine, the dreamier it tastes.
Measuring the monomeric anthocyanin concentration of a young red wine to gauge this potential is very useful but slightly tricky. It’s bad news for the amateur wine enthusiast with a cellar that red wine’s darkness itself is not useful for determining this potential.
Color density doesn’t account for the degree to which field oxidation may have polymerized pigment, particularly in cases of long hang time. Wines that look very dark may actually be quite fragile. It helps to look for an overly amber edge, a bad sign. You can’t make a soufflé from eggs that are already scrambled.
For technical winemakers, there is more bad news. The standard Somers-Boulton spectrophotometric methods for color aren’t much better. They look at monomeric anthocyanins at pH 3.6, where only 15%-30% of anthocyanins are in the colored form, and this amount varies widely according to the precise ratio of the five anthocyanins types as well as alcohol content and other factors. Fuhgeddaboudit.
The best available molar estimator for anthocyanins is an offshoot of the Adams astringency assay4 called BP (bleachable pigment). Easy enough to run, its LPP (long-chain polymeric pigment) is also a good analytical index of dryness. Problem solved. You’re welcome.
Young pigments are very vulnerable and easily lost to adsorption by lees and attack by enzymes. Once incorporated i nto polymers, they are protected. Early stabilization of color through cascading polymerization benefits wine’s appearance, texture and aromatic properties. The diphenol cascade promotes color stabilization directly through polymerization, but also indirectly by aldehyde bridging.
Each turn of the diphenol cascade produces a molecule of hydrogen peroxide (H2O2), a highly reactive oxidant with a strong affinity for SO2. If there is any present, the peroxide will combine with it to form sulfuric acid (H2SO4), a strong acid that lowers pH and lends a firm authority to the finish. Prior to malolactic, thus before SO2 has been added, peroxide will instead oxidize ethanol to acetaldehyde, which is believed to form bridges between tannins and anthocyanins which further stabilize color..
Timing is everything
Aldehyde bridging is believed to be initiated by a low pH form of acetaldehyde called a carbo-cation. For all these reasons, red wines benefit most from oxygen introduction pre-ML, pre-SO2. This is when oxygen reactivity is highest, anthocyanins concentration is highest and most vulnerable and aldehyde is readily formed and consumed.
In the absence of oxygen, non-oxidative polymerization will take place, which does not as readily protect anthocyanins, resulting in long polymers that cooperatively bind more readily to salivary protein, resulting in drier, harsher mouthfeel. Early cellar choices determine whether oxidative or non-oxidative polymerization will predominate.
The presence of lees in young reds is deleterious for two reasons: Lees react readily with oxygen, dampening phenolic activity and suppressing the cascade. Lees also adsorb anthocyanins and contain glycosidase enzymes that attack them. Just as egg yolks prevent the formation of a meringue structure, lees must be separated through clarification in young red wine. As with a soufflé, their incorporation after the structure is formed can be beneficial to add fatness and richness.
The vicinal diphenol cascade reaction is extremely temperature dependent. Because it involves two reactions (one with each -OH group), we label its kinetics of the second order, a fancy way of saying that its temperature dependence is squared. Patrick Ducournau’s5 empirical determination was that red wines take up oxygen five times as fast at 59ºF as at 50ºF. This means that a 1ºF drop in cellar temperature can reduce oxygen uptake by one-third.
In 2000, research at the University of Bordeaux6 reported that acetic acid bacteria in red wine are not inhibited by the pigment-bound SO2. Because it is in rapid equilibrium, our standard analyses report this as free SO2, but this is false. There is essentially no truly free SO2 unbound to pigments in young red wine. It is O2 uptake by the vicinyl diphenol cascade that protects the wine, and cold cellars suppress its action. This explains why the Pinots of Burgundy’s cold cellars are prone to VA.
Overripe fruit, as a general rule, also has very low comparative phenolic vigor—often losing 90% of its reactivity in three weeks on the vine. It is possible to measure the phenolic O2 uptake capacity of a wine. A report about efforts to re-introduce an 80-year-old method for measuring this critical attribute will be the subject of an upcoming column.
Clark Smith is winemaker for WineSmith, founder of the wine technology firm Vinovation. He lectures widely on an ancient yet innovative view of American winemaking.