Science says, dab sugar water on a baby’s lips and she’ll smile. Swap in salt water and you’ll elicit a more neutral reaction, while sour or bitter flavors will make her fuss and frown. In other words, more Negronis for mommy.
Some scientists point to the experiment above as proof that we are hardwired to enjoy sweet things. To the developing human body, sugar—a carbohydrate—represents an easy fuel source. But too much sugar brings cavities and bad skin; more still can invite diabetes and obesity. The key to good health is a balanced diet—one needs some broccoli to offset all those carbs. Similarly, the key to a good wine is balance.
Baby wine drinkers also often crave the sweet stuff. How many of us began our fine wine journey with a taste of Boone’s Farm, Liebfraumilch, or a wine cooler? Sweetness is a reliable characteristic of bottom-shelf brands, yet many of the world’s greatest wines contain some degree of residual sugar. Even so, there has been a dry wine revolution in the last few decades. Consider the rise of the VDP and Grosses Gewächs in Germany, the swelling category of non-dosé Champagne, and the increasingly sluggish sales of dessert wines.
In a recent conversation with importer Terry Theise, he remarked that there was nothing in wine as poorly understood as the role of sugar. Indeed, sugar seems to have become an almost political topic in the wine trade, with some buyers refusing to even taste Champagne above certain dosage levels. The controversy invites closer consideration of the topic—where sugar comes from, how its presence or absence affects certain wines, and some of its more mystical attributes.
Sugars are the main source of perceived sweetness in wine, and they come in many forms. Inside of the grapevine, indeed any plant, the primary energy substrate is sucrose (in animals, it is glucose). Sucrose is a disaccharide, the combination of a single glucose and a single fructose molecule. This breaks down during the ripening of the grape, so a fermenting must always starts with equal parts glucose and fructose.
Saccharomyces (whose very name translates to “sugar fungus”) preferentially consumes glucose; as a result, residual sugar is typically composed of 60 to 70% fructose, though this varies according to grape variety and yeast strain. Because fructose molecules interact with the sweetness receptors on our taste buds at a very high rate of efficiency, fructose tastes twice as sweet as glucose. This is part of the reason the packaged food industry relies so heavily on high-fructose corn syrup—that, and it is extremely cheap to produce.
Other sugars are present in wine in small, often negligible amounts. These include cellobiose, galactose, and pentoses, which are five-carbon sugars such as arabinose, rhamnose, and xylose. As Saccharomyces won’t consume them, they are often referred to as “unfermentable sugars,” and as they don’t taste very sweet to humans, they are rarely discussed. That said, some of them can be enjoyed by Brettanomyces or certain strains of lactic acid bacteria, which will mark a wine, if not leave it flawed.
Oak, especially in the form of toasty new barrels, has long been thought to contribute sweetness to wine, but the mechanism wasn’t completely understood until recently. In 2017, scientists in Bordeaux discovered a set of molecules called quercotriterpenosides, which are released from oak during fermentation and élevage. These molecules are small but mighty, influencing the taste of wine at even low doses due to their extreme sweetness. And yet, they aren’t sugars. According to Dr. Andrew Waterhouse of UC Davis, “Oak sugars could almost be categorized as artificial sweeteners, like Stevia. Their sweetness is a trick of their physical composition. They are simply able to interact with the receptors on our tongues at a much higher rate than actual sugar.”
Oak aromas can also add to the perception of sweetness. Guaiacol, eugenol, vanillin, and oak lactones evoke vanilla, coconut, and baking spices—scents that often herald dessert. These odors can trick our brain into perceiving a higher level of sweetness than is chemically measurable in the wine. Fruity aromas can do the same thing. “Our brains don’t work like adding machines,” Dr. Waterhouse explains, elaborating that what we think of as “taste” is created in the brain, informed by a variety of inputs that includes scent, touch, sight, temperature, and even memory. “Things interact and can emphasize or mask each other. Sweet and fruity are tied together in our brains, as ripe fruit is generally sweet.” This association would have been especially important to early humans, who undoubtedly used their sense of smell (along with touch) to select the ripest and therefore most energy-giving fruit for consumption.
Alcohol also adds to the sensation of sweetness and viscosity of a wine. Glycerin, also known as glycerol (a sugar alcohol), is a byproduct of yeast metabolism and the second biggest component in wine after water and ethanol. Because it is only half as sweet as sucrose, its low concentrations move the sweetness needle only slightly. And while it is often said to increase viscosity in wine, Dr. Waterhouse’s book Understanding Wine Chemistry shows that its concentrations are rarely above threshold, even in dessert wines.
I moved to wine country nine years ago. Back then, when speaking with winemakers, we would discuss the ripeness of grapes in terms of Brix. At some point, I noticed that winemakers were eschewing talk of Brix for something called “gluc/fruc.” Gluc/fruc is exactly what it sounds like: the total amount of glucose and fructose in a solution. How is this different than Brix? Don’t they both measure sugar and therefore potential alcohol?
Dr. Rich DeScenzo, of ETS Laboratories in St. Helena, breaks it down. “Brix, which is often checked in the field with a refractometer, is effectively a measurement of density, not of sugar. All sorts of things affect Brix—sugars, acids, tannins—basically any soluble solid.” As freshly harvested grapes are composed of around 25% sugar, sugar is the dominant component of the density. Because of this, there is often a strong correlation between Brix and final alcohol, but it is not infallible. Depending on the conditions of the vintage, there may be more or less solids in a given year, which will affect how closely Brix correlates to the resulting alcoholic strength of the wine.
Brix also fluctuates in accuracy depending upon the stage of fermentation. “When you go over 18 Brix, there is more sugar than Brix. Under that, there is less sugar than Brix,” Dr. DeScenzo explains. What’s more, as the yeast convert sugar to ethanol, the density of the solution changes. Pure water has a Brix reading of zero. Ethanol is less dense than water, which is why finished fermentations show a negative Brix reading.
Data courtesy of ETS
By contrast, gluc/fruc is the precise measurement of the sugar that the yeast will ferment. According to Dr. DeScenzo, “16.83 grams of sugar gives you 1% of alcohol. There’s some wiggle room there (16.7–17.5) based on the temperature of fermentation, the architecture of the tank, and yeast strains, but this is basically the number.” Following each harvest, he makes a point to calculate the ratio of Brix to gluc/fruc. “Last year, we saw lower alcohol conversions than expected. Ethanols were down three-tenths per 25 Brix. The reason is a bunch of geeky science stuff, but the point is that Brix is variable and therefore not the best way to predict your alcohol. Also, there is more variability at higher Brix. Back when people were harvesting at 22 or 23 Brix, this wasn’t much of an issue.”
How we measure residual sugar is also important, and Gordon Burns is part of an organization that is working to standardize it. FIVS (Fédération Internationale des Vins et Spiritueux) is an international group of producers and scientists, essentially the trade counterpart to OIV (International Organisation of Vine and Wine), which is an intergovernmental association. “[It’s] basically the UN of wine,” Burns jokes. “So, the question has been, what do we mean when we say residual sugar? Are we including pentoses and other unfermentable sugars? And how do we measure it?”
The OIV has decreed that residual sugar will refer exclusively to the presence of glucose, fructose, and sucrose in a wine. The inclusion of sucrose allows for wines that have been chaptalized, though that sucrose eventually breaks down into equal parts glucose and fructose in the bottle. But while this ruling sounds refreshingly simple, the conundrum for FIVS is that not all wineries use the same methods to measure residual sugar. This is reflected in the fact that, historically, RS stood for reducing sugars, a nomenclature to which some countries and wineries still cling.
Measuring exclusively for the presence of glucose and fructose requires a specialized piece of equipment, a spectrophotometer, which runs around $2,000. Cash poor or less modern wineries and laboratories that don’t possess this equipment test for sugar by measuring a wine’s ability to reduce copper ions. “The old way was very satisfying and recalls the science done in a Muppet lab. It’s all stir plates and swirling flasks that change color and emit vapors,” Burns reminisces, “but it isn’t very specific. Pentoses and tannins and aldehydes and ketones also reduce copper.”
FIVS recently concluded a study that looked at the difference in reporting between labs running gluc/fruc panels and those checking for reducing sugars. “It confirmed what we tell clients. Given a wine with close to zero gluc/fruc, the reducing sugar analysis reports RS of one gram per liter for white wines and two grams per liter for red wines. That’s a huge bias.” Why does this matter? Money. “There are certain countries, like China, that base the tax classification of their wines on sugar content, not alcohol. And a bias of two grams per liter will bump certain wines into a different tax status.” Burns continues, “Ultimately, it’s not an issue of sensory perception. It’s about dollars and cents.”
As noted, sweetness is not an empirical quality wedded solely to the physical amount of sugar in a wine; it is relative to both the other components in that wine as well as the sensitivity of the taster. That said, certain generalizations can be made.
Per Dr. Waterhouse, there is no such thing as a completely dry wine. “It is physically impossible for yeast to consume 100% of the sugars in a wine, so when we talk about a ‘dry’ wine, we are generally talking about anywhere between 0.5 and 2 grams per liter of residual sugar.” Though very experienced tasters might be able to detect those 2 grams, he believes the typical threshold is much higher: 4 grams per liter for expert tasters and potentially as high as 10 grams per liter for beginners. “In a standard wine that has around 5 to 6 grams per liter of acid, you generally need about 1% sugar by volume (10 g/L) before most tasters will register the wine as sweet. For comparison, dessert wines range from 5 to 15% sugar (50–150 g/L), your standard Ports are around 10% sugar (100 g/L), and there are really sweet wines that hit 20% sugar (200 g/L)—but those are crazy sweet.”
Individual thresholds aside, there are elements in wine that influence how sweet it will taste. The perception of tannin and acid notoriously diminishes—and is diminished by—residual sugar, while high alcohol increases the sensation of sweetness. According to Dr. Gordon M. Shepherd’s Neurogastronomy, the reverse is also true: sweetness increases perceived viscosity. But of these ballasts, the most important is acidity. And it is the relationship between sugar and acid vis-à-vis perceived sweetness that the International Riesling Foundation (IRF) admirably set out to quantify.
Riesling is widely considered one of the world’s most noble grapes, and yet even its most fervent advocates admit that selling the wines can be difficult. While cumbersome German nomenclature is occasionally the culprit, more often the issue lies with consumer insecurity. People simply have a hard time predicting how sweet a Riesling will be just by looking at the label.
Riesling Rendezvous, a Seattle-based international gathering of Riesling producers, held its first symposium in 2008. According to current IRF President Janie Brooks Heuck, that year, the debate between panelists regarding perceived sweetness got so heated that the organization decided to intervene. Clearly the confusion regarding what constituted a dry wine existed on the producer as well as the consumer level. The result was the IRF scale, a useful if imperfect tool for communicating sweetness that has since graced the back label of over 26 million bottles worldwide.
Producers determine their wine’s position on the scale via a three-step process. First, they divide the grams of residual sugar by the grams of titratable acidity (TA). If the resulting number is less than one, the wine is considered dry. This means that a wine with 5 grams per liter of acid and 1 gram per liter of sugar would be dry, but so would a wine with 11 grams per liter of sugar so long as the acidity weighed in at 11 grams per liter or more. Ratios between 1.0 to 2.0 are considered medium dry, 2.1 to 4.0 are medium sweet, and over 4.1 is decidedly sweet. This is similar to Germany’s rules for trocken Riesling, which state that residual sugar can be up to nine grams per liter so long as total acidity falls within two grams per liter of the sugar.
The second step uses the wine’s pH to further refine its setting. A lower pH might inch the wine toward dryness, while a higher pH may push it to sweet (see the thresholds here). The final step is a taste test, in case the sensation of the wine differs from its mathematical projection. And while this last step is not overseen by the IRF, Heuck insists that an official committee regularly stages blind tastings, and that they agree with the wineries’ positioning 99% of the time. Fallible or not, at the very least, the IRF scale takes a considerable amount of guesswork off the shoulders of the consumer.
Very sweet wines require elevated acidity to make them palatable, but the reverse can also be true. It naturally follows that dry wines from some of the highest acid varieties (Riesling, Chenin Blanc) and the most marginal climates (Champagne, the Mosel, the Finger Lakes, parts of Canada) regularly feature some degree of residual sugar. This is an especially hot topic in Champagne, where drier styles are trending.
“In Champagne, the grapes are harvested early, so it is not uncommon for wines to have acidity at or in excess of 10 grams per liter. Most people would consider a wine with that level of acidity undrinkable without residual sugar,” remarks Dr. Waterhouse. “But, of course, tastes change. One hundred years ago, Champagne used to be a lot sweeter. Bottles found at the bottom of the Baltic Sea contained 15% sugar!”
Frédéric Panaiotis, winemaker for Ruinart, believes that dosage is important, especially for the bigger brands. “When I joined in 2007, the dosage for our non-vintage Champagne was 12 grams per liter. I moved it down to seven to eight grams per liter, perhaps too fast, and we received letters of complaint from some of our older clientele that the wines were getting too green!” He laughs and continues, “I find a little touch of dosage is necessary for the wines to show their best. Without it, the wines are too extreme. I can understand that if you are a small grower speaking to a smaller, more geeky population, but as a house we are speaking to a broader audience and we need it."
Among Champagne producers, there seems to be a dual credit for the move toward dryness: shifting consumer tastes and global warming. The logic of the latter is that higher temperatures lead to naturally riper grapes with lower acidities that therefore require less dosage. Panaiotis finds this explanation simplistic. “I hate this journalistic shortcut of global warming,” he exclaims. “Yes, there is global warming, but we have adjusted our farming to it. The vineyard management, in association with the change in weather, has led to grapes that are balanced differently, not just riper. Plus, the way we’ve made wine has changed across the region. We are better winemakers now, so the final balance needs less sugar.”
Terry Theise also takes issue with blaming global warming. “That implies that sugar is a corrective device. I prefer to think of it as a flavor enhancer.” While many of his producers work in a drier style and more than one puts out a non-dosé, Theise is a known advocate for the balancing effects of dosage. “There are degrees of sweetness, yes, but we also need to acknowledge that there are degrees of dryness. And often, the perfectly dry wine is not perfectly dry.”
Theise has sat in on several of his producers’ dosage trials. “One of the strange Zen things I have learned is [that] as you adjust sweetness, you reach a point where the sugar sticks out like a sore thumb. The beginner would think that is because there’s too much sugar, but I learned it is because there’s often too little. And when you hit the exact right amount, the sugar melts back into the wine and becomes harmonious.” Theise is especially fond of René Geoffroy and Chartogne-Taillet’s approach to dosage. “They are all over the place in terms of sugar. I love that about them, but it drives the market crazy. A Champagne might be 10 grams per liter one year and 3 the next. What this tells me is the producer is tasting for balance and ignoring the lab results.”
The shift in consumer tastes is harder to explain, though theories abound. On an anecdotal level, my personal experience in the trade has led me to believe that dry wines are often regarded as more sophisticated than off-dry or sweet wines, especially in Champagne. This view is reinforced by the marketplace. Most producers’ entry level non-vintage Champagne is measurably sweeter than their vintage or tête de cuvée bottlings. Also, it is not uncommon for larger houses to send their sweeter wines to the US and/or Asia and reserve their drier efforts for the European market. Whatever the specific motivations behind these decisions, it is hard to avoid the subconscious messaging: dry Champagne is better Champagne, and those in the know drink dry.
Though the primary purpose of dosage in Champagne is to balance acidity, it also serves to counter the lifting effect of CO2 (which forms a weak carbonic acid in solution) as well as the phenolics of the grapes. The soothing of phenolic bite, specifically the bitterness expressed by Viognier, Gewurztraminer, and Pinot Gris, is a common application for residual sugar in dry white wines.
Alex Davis from Porter Creek Winery in Sonoma thinks that fear of bitterness is one of the reasons so many producers let their Viognier get too ripe. Davis, who studied under Guigal in the Northern Rhône, believes bitterness is what gives the variety its nerve and isn’t afraid to harvest earlier than the California average. Though he aspires to get his Viognier under four grams per liter, other producers will sometimes leave in a little residual sugar to ease its phenolic edge. Zoltán Kovács of The Royal Tokaji Wine Company in Hungary does the same thing with Furmint, especially in leaner years. “Sometimes in a higher acid vintage, we will leave two to three grams of residual sugar to make a better balance. We do this by cooling down the fermentation, racking, and sulfuring.”
Alsace is notorious for walking the line that separates dry from sweet, often to the consternation of consumers. But to diehard fans, this confusion only speaks to the multifaceted character and complexity of Alsace. And while the regional style skewed a bit sweeter in the 1980s and ‘90s, Trimbach has always been firmly within the dry camp. But as previously established, dry is a relative concept.
“It has to do with acidity, of course, but residual sugar also has to work with the terroir, the salinity, the dry extract, the mouthfeel, the texture of a wine,” Jean Trimbach explains with great enthusiasm. “You can have a wine with five or eight grams of residual, which will still taste dry because of the power of the vintage.” Tasting dry is essential to Trimbach, who makes dry wines and sweet wines and nothing in between. “We don’t want anything to be off-dry. It’s a kind of mystery taste that we don’t like at all.” To achieve this, they ensure that the residual sugar of their top Pinot Gris and Gewurztraminers range from 5 to 12 grams per liter, but their Rieslings always fall beneath 6 grams per liter. “Sometimes the fermentations stop themselves, especially the Rieslings. Or sometimes, like if we have a Gewurztraminer that was harvested at a potential alcohol of 14.5%, we cool it and stop at 13.5% because that last degree of alcohol would make the wine burn.”
For Mathieu Deiss, who works with his father at Domaine Marcel Deiss, the decision whether to allow residual sugar is dictated solely by site—specifically, whether a vineyard regularly sees botrytis. “For me, botrytis is the only way to produce a wine with residual sugar that is integrated into the body. If you just push maturity and then stop the fermentation to leave in RS, that is a stylistic goal, but it is the opposite of agriculture.” The magic of botrytis, Deiss believes, is its ability to concentrate sugar and acid simultaneously. “When you have noble rot, sometimes you have a wine with 10 or 15 grams of sugar and you don’t feel it at all.”
Deiss feels there are many types of vineyards in Alsace, with two extremes. On one hand, there are sites where ripening happens quickly and “you can harvest a beautiful, ripe fruit with mature skin and no bitterness, plus the freshness from the early picking,” and on the other, ripening is slow and “botrytis hits late, at the end, and you can make a very elegant wine with some sweetness because the acidity wasn't ripened away by the weather.” In his experience, botrytis tends to favor vineyards on the valley floor, where it is cooler and the soils are heavier.
Winemaking is extremely hands-off at Deiss. Fermentations happen via native yeast and are never prematurely arrested, so all residual sugar levels are arrived at naturally. “The whole point is to get mature skin. We harvest the botrytis wines at a potential alcohol of around 15% and they stop naturally at 13 or 13.5%. The dry sites we harvest at 13 to 14%, depending on the weather.” Because they don’t sulfur until late, most of the wines undergo malolactic fermentation, which is unusual in wines with residual sugar. Interestingly, they hold the botrytis wines back from the market for two to three years before release to allow the sugar to integrate. “For me, Alsace is maybe one of the few places where you can make an elegant wine with sugar and also elegant dry wines. This makes it complex to understand from the outside, especially if people think about the region only in terms of variety. For me, what dictates the sugar is not the variety; it is always the terroir.”
This article has so far discussed residual sugar primarily in terms of its role as a balancing agent in fine wine, but sugar has more devious applications as well. Residual sugar can be used to mask flaws, smooth rough edges, and—because all of us started life as sugar-craving infants—increase appeal.
David Ramey is a self-described classicist. As such, he believes that all the classic table wines of the world are dry (excluding Germany). “It’s not conceivable to me to have a Red Bordeaux or White Burgundy with residual sugar, even in a botrytis year.” And yet, he is also a winemaker. And as a winemaker, he is fascinated by both the art and the science of his profession. For his own brand, Ramey, he employs classical thinking, but he has been known to bend his own rules at the behest of clients. His experience in this began early. “When I started as assistant winemaker at Simi in 1980, we finished the rosé and Chenin Blanc with 10 grams per liter of sugar,” he tells me. “One of my jobs was to dial down the temp in the tank and hit it exactly at 1% sugar. It was just like landing a plane.”
Nearly 40 years later, Ramey is a well-established winemaker who has worked with vineyards and people across Northern California. He is especially fascinated by wine chemistry and has built an impressive laboratory at his winery in Sonoma. He loves blinding groups of tradespeople on wines, having them select their favorites, and then presenting them with the numerical data. The results are often quite revealing. “What can I say? It’s fun for me,” he admits with a laugh. “Running these numbers, it’s like being a superhero. It gives me x-ray vision, and I can see what all these other winemakers are up to.”
Lab analysis from David Ramey of select sweeter dry wines (Click to zoom in)
“For me, the sweetest I’d consider a Napa Cabernet to be dry is at two grams per liter, but I prefer under one gram per liter,” Ramey allows. “But if you harvest at a potential of 16% alcohol and the yeast struggle, you’ll probably end up with 2.5 grams per liter. Add concentrate, and you’re hitting 3 to 3.8 grams per liter. And that—that is sweet.” Grape concentrates such as Mega Purple have effectively two applications. Their primary purpose is to be added during fermentation to increase the sugar and pigment of a must, but some producers introduce them following fermentation to imbue their wines with added richness. This second approach is not unlike adding süssreserve to a fully fermented German Riesling. “You add the concentrate right before bottling, then add Velcorin, then sterile filter. You do it right at the bottling tank. I’ve helped clients do it.” Ramey has also added bulk Port to California Cabernet at the behest of clients.
Why would a producer who is paying for a consultant and therefore likely making a high-end wine add concentrate or Port right before bottling? Most likely: points. While a touch of sugar in red wines is commonly employed to entice low price-point drinkers, many believe sweetness appeals to critics as well. While some are known to love rich and inky (and potentially slightly sweet) reds, other critics may simply be experiencing palate fatigue. As anyone who’s tasted too many wines in one sitting can attest, sugar is a salve to the weary tongue. It’s not impossible to imagine that low levels of residual sugar give a wine an advantage during a marathon tasting. It’s also possible that the producers in question—along with their clientele—simply prefer their wine off-dry. After all, America remains a nation of soda drinkers, and an analysis of the sugar content of popular vermouths suggests a preference for sweetness in cocktails as well.
Data courtesy of Jason Heller
“These guys that want that stuff added, they are better businessmen than I am,” Ramey admits. “On the one hand, I’m glad their customers are drinking wine and not beer. On the other, I’m hoping that one day, they hone their tastes to appreciate dry wines that work with food.”
The reason David Ramey would follow a dose of grape concentrate with Velcorin (an anti-microbial enzyme) and sterile filtration is because the presence of residual sugar typically makes for an unstable wine. Sugar, lest we forget, is the preferred snack of microbes. And yet, of the world’s greatest wines, dessert wines are often the longest lived. How could something so sweet survive for months, never mind decades? And once they are open, why do bottles of dessert wine stay fresh for so long? And why does the impression of sweetness in dessert wines seem to diminish with age? The more I thought about these things, the more impossible it all seemed.
It was time to phone some friends.
“It’s certainly true that the longest-lived wines I have come across have been Sauternes, particularly Yquem,” Jancis Robinson, MW, attests, “but I assume that it was the botrytis plus sugar rather than just sugar, because some simple sweet wines don’t seem to last at all well.” She continues, “In Vouvray, the sweet wines seem to age longer than the dry ones—though here, presumably acidity is coming into play.” Legendary sommelier Rajat Parr does not equivocate. “One hundred percent, yes. Sweet wines age better than dry wines.” He goes on to describe the remarkable freshness and vitality of the greatest wine he’s ever had, an 1847 Château d’Yquem, as well as the immortality of 1947 Cheval Blanc, a wine that is regularly described as “Port-like” and is rumored to have been bottled with some degree of residual sugar. “And Champagne with heavy dosage will age longer than Brut Nature, too. That’s a fact.”
“I believe that sweet wines age better than dry wines,” admits Kovács. “But it’s not just about the sugar. It’s also how the wine was treated during vinification. If everything is made perfectly, the wines with the highest sugar will age better.” As a maker of Essencia, he is uniquely experienced to comment. “Essencia is the sweetest wine made in the world—some have as much as 600 grams of residual sugar, and you can still enjoy this wine for up to a century. Fifteen years ago, I tasted the 1913, and it was still wonderful. A lot of tertiary aromas, understandably, but it was in a nice place.”
Botrytized vines in Tokaj (Photo credit: Bryce Wiatrak)
Not everyone agrees that sweet wines have the advantage. “I don’t necessarily believe that sweet age better than dry wines,” Egon Müller says. “The secret to aging is balance. I’m very good friends with Jean Trimbach, and we regularly taste our old wines. I’ve had his dry wines back to the 1950s, and they are as much alive as the sweet wines that we make. In my opinion, it is not a question of sweet or dry; it’s a question of do we have the right grapes, and did the winemaker do his job of turning them into a great wine.”
Regardless of whether sweet wines exhibit superior aging, it is still remarkable, from a chemical perspective, that they survive. Fortified wines are easier to understand; high alcohol is a known preservative. But what about unfortified styles? Dr. DeScenzo claims that osmotic pressure plays a critical role. For a microbe to live, it must maintain a comparable concentration of water inside its cell as exists in the surrounding solution. High concentrations of sugar force the water within a microbe to rush outward, and its cell walls collapse. This is the mechanism by which fruit can be preserved in syrup.
That said, certain strains of yeast and bacteria can thrive at high concentrations of sugar (Zygosaccharomyces rouxii, the bane of Mega Purple, can survive in solutions of 85 to 90%), so osmotic pressure by itself is insufficient. “Most dessert wines have around 100 to 150 grams per liter residual sugar,” DeScenzo states, “but a fresh grape must that was harvested at 28 Brix has around 300 grams per liter of sugar, and yeast happily ferment that!” The secret to stability, he believes, is osmotic pressure plus alcoholic pressure (plus sterile filtration, widely practiced by producers of dessert wine).
Acidity, as Robinson surmised, and SO2 are also essential. And they all—osmotic and alcoholic pressure, acidity and sulfur—interact with and reinforce each other. Kovács describes the ideal Tokaji. “If we have sufficiently high sugar, 140 to 150 grams per liter, and a nice alcohol like 11%, with a high acidity, then our free sulfur will be around 30 milligrams, with a total sulfur around 150 to 200 milligrams. We wouldn’t need to have the maximum allowed 400 milligrams because the high sugar, low pH, and nice alcohol make for a good stability.” When asked to provide his ideal Sauternes chemistry, Luc Planty, co-owner of Château Guiraud replied, “We look for wines between 13 and 13.5 degrees of alcohol, between 110 and 130 grams per liter of residual sugar, a pH of around 3.5 or 3.6, and a free sulfur as low as possible. [But] there is no exact formula, and the equilibrium can change according to each lot and vintage.” Once again, it seems that a balance of multiple forces is more powerful than any one factor alone.
Examples of dessert wine chemistry
As for botrytis, first, to cross the line from aging to aging well, complexity is surely essential, and botrytis lends a hand there. It is also known to concentrate acidity, which is stabilizing. But as Müller muses, its presence is not always a gift. “Botrytis changes the grape; that is a given. It will add its own layer of flavors and complexity, alter the mouthfeel, and make a wine seem more substantial. I believe botrytis contributes to ageability, but it can also go the other way.” He goes on to make the point that sometimes rot is just rot. “[The year] 2000 was the birth year of my son, so I made an extra effort, but it was not ideal noble rot and the wines are aging very quickly. But in great noble rot vintages like ‘99, ‘89, ‘05, the wines have unbelievable aging potential, greater than the same wines without botrytis. I feel it is mostly a question of weather. If botrytis attacks and the weather is fine, the rot will be noble. But if it rains and everything becomes diluted, then it is not a good thing.”
At one point during the research for this article, I thought I had cracked the case wide open. A 2015 article in Plant Physiology detailing the study of botrytis-infected Sémillon berries had come across my desk. In the paper, scientists reported that, in addition to concentrating malic acid, botrytis led to the genesis of sugar alcohols, pentoses (formed from the breakdown of pectin), and phenolic compounds along with anthocyanins. Anthocyanins—that is, tannins and pigment—are known antioxidants and are traditionally produced by red wine grapes. Botrytis had effectively reprogrammed Sémillon to behave like a red grape! Could these anthocyanins be the secret to the long life of botrytized dessert wines?
Dr. Waterhouse thinks this is unlikely. “The amount of anthocyanin is quite small, and at that level, unlikely to be a substantial pool of antioxidants. Plus, botrytis infections induce laccases that cause oxidative damage during winemaking.” Laccase is an enzyme known for eating anthocyanins, which many producers fight by adding powdered tannin (for red wine infections) or, more commonly, loads of sulfur. Indeed, per Egon Müller, “When I have botrytis, I add far more sulfur. Maybe you can bottle non-botrytis wines at 70 or 100 milligrams total sulfur and still have 30 free. When you have a botrytis wine, depending on the concentration, you very quickly add up to 200 milligrams total in order to have the same amount of free SO2.” In short, if botrytis is contributing to ageability, it is either through increasing complexity and concentration or an as-yet-unknown pathway.
What about dessert wine’s uncanny ability to resist oxidation once opened? Eric Hervé, of ETS labs, was uncertain, but suggested it might have to do with the fact that many late harvest wines are made in an oxidative fashion. “The juice can get quite brown. So basically, anything that can oxidize in the wine has already been oxidized by the time it gets into the bottle.” Still others believe that it is the sugar itself that protects the wine. Not from oxidation, per se, but from showing signs of oxidation. To understand this more completely, we must return to Champagne.
Tom Stevenson has regularly contended that low levels of dosage can adversely affect Champagne, making for wines that oxidize more easily and display a diminished capacity to age. To test his theory, he conscripted a group of producers to engage in dosage aging trials. Participating houses bottled a handful of magnums at zero, three, six, and nine grams per liter of dosage, to be opened at three and six years of age. The results of the first tasting were published in a 2016 World of Fine Wine article called “Sweet Truth.” In the piece, Stevenson references a conversation with Bertrand Robillard, a Champagne-based chemist. Robillard allows that part of the problem might be the tendency of producers who bottle at a low dosage to also employ reduced levels of sulfur. But the article also describes how sugar prolongs a wine’s freshness via two distinct mechanisms.
The first involves sulfur. In wine chemistry, two sulfur numbers are typically cited: total SO2 and free SO2. Total SO2 represents both free and bound SO2, and this number is mostly fixed. By contrast, free SO2 is the sulfur that is unbound in a wine and therefore able to actively protect it against microbes and oxidation. This number diminishes with time, eventually disappearing altogether, at which point a wine is considered relatively naked and unprotected. Residual sugar binds SO2 but only loosely, and it is believed that, as free SO2 is depleted, the binds between sugar and sulfur break, which thereby releases bonus free SO2 into the wine. To quote Stevenson, “Thus the dosage acts as a reservoir of potential free SO2, and is an additional line of defense against oxidation.” The same rule applies to residual sugar in still wines.
Stevenson also briefly suggests that sugar provides a kind of aromatic screening service via increased vapor pressure. Dr. Waterhouse also believes that sugar affects a wine's aromatics, but in a different way. “With regard to chemical reactions, since the sugars are present at such high concentrations, the sugars tend to be the substance to get oxidized. [Oxidized sugar compounds] generally have negligible flavor and only slight or somewhat desirable aromas. So, the change from aging is subtle or slightly favorable.” This revelation explains not only the seemingly enhanced ageability of sweet or off-dry wines but also why dessert wines can be served for days or weeks before any sign of fatigue appears.
But what about the fact that many dessert wines taste less sweet as they age? Many producers interviewed here have run chemistries on their older vintages and report that the amount of residual sugar remains static. So what happens? Tannins in older wines feel smoother because they bind together and lengthen; color diminishes with age because the anthocyanins fall out of solution and form sediment. Whither wither the sweetness?
For now, my last question remains unanswered. To comfort myself, I will just have to learn to embrace the mystery and fall back on romantic notions. Egon Müller’s thoughts on the subject are especially captivating. “When you taste a young sweet wine, what makes it interesting is the fight between acid and sugar. When you taste an old sweet wine, there is no more fighting. Everything is integrated and the individual sensations are less pronounced. What you are tasting is harmony.”
Wet my lips with harmony, then. I promise to smile.
Blanco-Ulate, Barbara, Katherine C.H. Amrine, Thomas S. Collins, Rosa M. Rivero, Ariel R. Vicente, Abraham Morales-Cruz, Carolyn L. Doyle, et al. “Developmental and Metabolic Plasticity of White-Skinned Grape Berries in Response to Botrytis cinerea during Noble Rot.” American Society of Plant Biologists. Accessed Dec. 20, 2018. http://www.plantphysiol.org/content/169/4/2422.
Boiling, Chris. “How noble rot influences wine flavour – study.” Decanter. Accessed Dec. 20, 2018. https://www.decanter.com/wine-news/noble-rot-wine-flavour-study-370557/.
Gilbert, Avery. What the Nose Knows, the Science of Scent in Everyday Life. Synthetics, Inc., 2014.
Goode, Jamie. I Taste Red, The Science of Tasting Wine. Berkeley: University of California Press, 2016.
Goode, Jamie. The Science of Wine, from Vine to Glass. Berkeley, CA: University of California Press, 2005.
Kringelbach, Morten L. and Kent C. Berridge, eds. Pleasures of the Brain. Oxford: Oxford University Press, 2010.
Marchal, Axel, Pierre Waffo-Teguo, Andrei Prida, and Denis Dubourdieu. “Sweet Triterpenoids in Oak for Cooperage: Uncovering the difference between sessile and pedunculate oak.” Wines and Vines. Accessed Dec. 20, 2018. https://www.winesandvines.com/features/article/182252/Sweet-Triterpenoids-in-Oak-for-Cooperage.
Peynaud, Émile. The Taste of Wine, the Art and Science of Wine Appreciation. New York: John Wiley & Sons, 1996.
Science Direct. “Botrytis Cinerea.” Accessed Dec. 20, 2018. https://www.sciencedirect.com/topics/neuroscience/botrytis-cinerea.
Shepherd, Gordon M. Neurogastronomy, How the Brain Creates Flavor and Why it Matters. New York: Columbia University Press, 2012.
Steiner, J.E. “Innate, discriminative human facial expressions to taste and smell stimulation.” Annals of the New York Academy of Sciences 237 (1974): 229-233.
Turin, Luca. The Secret of Scent, Adventures in Perfume and the Science of Smell. London: Faber & Faber Limited, 2006.
Stevenson, Tom. "Champagne Dosage Trial (Part I): Sweet Truth?" The World of Fine Wine 52 (2016): 174-183.
Waterhouse, Andrew L., Gavin L. Sacks, and David W. Jeffrey. Understanding Wine Chemistry. West Sussex, United Kingdom: John Wiley & Sons, 2016.
Special thanks to Dan Petroski, Erik Johnson, and the French Laundry for assistance sourcing dessert samples, and extra special thanks to Christopher Tanghe for inspiring this article with his forum post!
I really appreciate this well-thought-out article. Thank you!
One question: In the statement "And while it is often said to increase viscosity in wine, Dr. Waterhouse’s book Understanding Wine Chemistry shows that its concentrations are rarely above threshold, even in dessert wines.", you were discussing glycerol. Since I have not read Dr. Waterhouse's book, I ask for just a bit of clarification, if you could. The threshold you speak of is that at which we as humans can perceive a change in viscosity due to glycerol, is that correct?
Yes, the threshold is referencing human perception! Glad you enjoyed the article!