Wine. It’s a liquid made of chemicals. And some of these chemicals have smells and tastes. That, in a nutshell, is wine flavour chemistry—a dauntingly complex but utterly engrossing topic. It’s my goal with this article to try to introduce the subject, outline some of the emerging concepts, and make sure that you stay reading to the end.
There are two ways of approaching wine flavour chemistry. One is to begin with the chemical composition of wine itself. The other is to look at how we as humans perceive tastes and smells. Both approaches are vital, because it’s possible to argue that the taste of wine isn’t a property of the wine itself, but of the interaction between the wine and the taster. This might seem to be an overly philosophical approach, but it’s important to hold this concept in mind as we explore the whole area of the taste and smell of wine.
Currently, the academic field of wine flavour chemistry is undergoing a quiet revolution. Until recently, most sensory scientists took a reductionist approach to studying wine. That is, they broke it down into its component parts, and studied each in isolation. This is a powerful method of enquiry, but has its limitations, chief of which is that wine aroma and flavour isn’t simply the result of an additive combination of the different flavour compounds present. Instead, there are complex interactions among the different flavour molecules, and also with other chemicals present in the wine, such that the final result isn’t predictable from simply knowing the composition of the wine.
An example of this would be where one aroma compound, present below the concentration at which we are able to smell it, actually affects the way another aroma compound smells. As a result, scientists are now turning to more sophisticated holistic ways of looking at wine aroma. We’ll come to some of these approaches in a while. For now, suffice to say that these sorts of experiments are a lot more tricky than simply looking at chemicals in wine in isolation.
Where wine flavor comes from
Take grape juice, from freshly pressed wine grapes. It’s sweet and not particularly complex, and nothing at all like wine. This tells us that it’s the action of microbes during the process of fermentation, plus the process the French call élevage—the bringing up of the wine—that is responsible for most of the flavour chemistry that we can sniff and slurp in the finished wine. Much of wine’s flavour comes from the action of yeasts and (where malolactic fermentation takes place) bacteria. As these organisms grow in the wine, they take up a range of nutrients from the grape must, and use these to build other chemical components that they need for growth, with sugar as the basic energy source to achieve this growth. The complicated metabolic pathways end up producing a wide array of chemicals that are then excreted into the must. In addition, yeasts die and release chemicals into the developing wine, as well as creating the lees—dead yeast cells at the bottom of the fermenting vessel—that are able to interact with the wine and alter its composition. During élevage, small amounts of oxygen contribute to the chemical reactions taking place, and if wooden barrels are used, these can also directly contribute flavour compounds themselves. And we mustn’t forget that the way that the wine is finished—for example, whether it is filtered or not, and any fining processes employed, plus the bottling conditions and closure choice—also have an impact on wine flavour chemistry.
Studying the flavour of wine
So we have this complicated chemical soup that is wine. How do scientists begin to unravel this complexity and work out the contribution of specific chemicals? It’s important to know which chemicals are important and why, because this then makes the next step of working out how they got to be in the wine at this particular level possible. And from here it’s then feasible to begin to understand how the winemaking process and vineyard environment influences the presence of this chemical (or, if it is produced by yeasts, its precursor compound, i.e. the one that is made from). This sort of scientific understanding is a tool to enable winemakers and viticulturists to have more control. Some may argue whether or not this is a good thing. I like to think that tools like this can be powerful when they are in the right hands. For example, if you desire to make an authentic, profound, complex wine that expresses a sense of place, then this sort of knowledge is your friend.
The long-standing approach to wine flavour chemistry is to take what amounts to a chemical fishing expedition. Individual chemicals are identified one by one from the wine, separated out, and then examined to see whether they smell of anything. This has been a useful approach, especially for flavour chemicals that are present at higher concentrations. Where it has struggled is with those chemicals that are present at very low concentrations and thus are hard to isolate, but which have a significant aromatic influence.
More recently, a technique called gas chromatography/olfactometry (GC/O) in tandem with mass spectrometry (MS) has been employed. Gas chromatography is a powerful way of analysing the volatile composition of wine. It works by separating out the aromatic portion of wine in much the way that a piece of blotting paper separates out the different pigments present in ink. Mass spectrometry can identify each of the chemicals that have been separated out in time, as they leave the chromatography column one by one, and olfactometry is the process where someone sits and smells each of the chemicals in turn as they are released. Now pet owners will be aware that the human nose isn’t as powerful as a dog’s, but it’s still able to pick out some smelly chemicals at incredibly low concentrations. These can be matched to their respective chemical identities, and a list of the volatile aroma compounds in wine can be built up.
The holistic view
One of the leaders in the field of wine flavour and aroma is a researcher called Vicente Ferriera, who is based in Zaragoza, Spain. His group recently published a really interesting paper, which is likely to have a big impact on the way that wine flavour chemists work. In this experiment he took red and white wines, and stripped them of their aromatic components to produce what is referred to as a non-volatile wine matrix. The aromatic components were held in reserve, and then recombined with the non-volatile matrix, to recreate wine. But the clever thing he did was to switch the two components around, so that the white wine volatiles were mixed with the red wine matrix, and vice-versa. You’d expect that the aromatic component would be the bit that determines whether the wine tastes red or white. But what he found was that in fact it was the non-volatile wine matrix that influenced how the wine was perceived. The red wine volatiles, when mixed with a white wine matrix, produced a wine that smelled like a white wine. And the opposite occurred when it was the red wine matrix that had the white wine volatiles added.
This is an important observation, because it shows that it’s not just the volatiles that are present that matter for wine aroma. The non-smelly bit of wine influences how the smelly bits of wine smell. That’s really unexpected and has important implications for how we study wine flavour and aroma. Any approach needs to be holistic.
So what flavour and aroma chemists are now doing is creating what is called a model wine. The first stage in this process is to use GC/O to find out which aroma compounds are present in the wine under study at concentrations where they are above the human olfactory threshold. These are likely to be the most important ones, although, as we discussed earlier, sometimes volatile compounds present below the level at which they can be smelled may be having an effect.
Stage two is to take the wine in question, and strip it of volatiles, so that you are left with the wine matrix. This should be done as cleanly as possible with the minimum effect on the other components present in the wine. Then, the most important volatile compounds—perhaps 20 or 30—are added back in their original concentrations. The result is tested by sensory analysis to see whether it smells and tastes like the original. If it does, then it’s game on. Using this approach it is then possible to create wines where one or more of the volatile compounds are excluded one by one, or in groups, to see what the impact is on the wine. This holistic approach has incredible power, and can yield quite unexpected results.
What’s in a wine?
So what makes wine taste and smell the way it does? It’s possible to categorize flavour compounds in a number of categories.
First, the base aroma. Ferreira’s group has identified 20 aromatic chemicals that are present in all wines, and which form a global odour that has been dubbed ‘wine odour’. Of these 20 aromas, just one is present in grapes (β-damascenone); the rest are produced by the metabolism on yeasts, in many cases working on precursors present in the grape juice.
• Higher alcohols (e.g. butyric, isoamylic, hexylic, phenylethylic)
• Acids (acetic, butyric, hexanoic, octanoic, isovaerianic)
• Ethyl esters from fatty acids
• Acetates and compounds such as diacetyl
The influence of alcohol (ethanol) is quite strong. Ethanol has been shown to modify the solubility of many of the aroma compounds, and makes them more reticent to leave the solution, thus making the wine less aromatic.
Second, we have a group of contributory compounds. There are 16 compounds which are present in most wines, but at relatively low levels. They are usually below the level at which they can be smelled on their own, but they have odour activity that is synergistic, contributing to characteristic scents despite being at lower concentrations that would normally lead to them being smelled.
• Volatile phenols (guiaicol, eugenol, isoeugenol, 2,6-dimethoxyphenol, allyl-2,6-dimethoxyphenol)
• Ethyl esters
• Fatty acids
• Acetates of higher alcohols
• Ethyl esters of branched fatty acids
• Aliphatic aldehydes with 8, 9 or 10 carbon atoms
• Branched aldehydes such as 2-methylpropanol, 2-methylbutanol, 3-methylbutanol, ketones, aliphatic γ-lactones
• Vanillin and its derivatives
Finally, we have impact compounds: the chemicals that are responsible for giving characteristic aromas to certain wines, even when they are present at extremely low concentrations. These are of great interest because they often contribute to distinctive varietal aromas. Examples of impact aromas include:
The perception of wine
With all this emphasis on what is actually in a wine, it’s possible to forget that the flavour of a wine isn’t the property of the wine: it is what happens when we encounter a wine. And we bring quite a bit to this wine tasting experience. We are far more than measuring devices. So it’s important that we discuss, at least briefly, the nature of sensory perception.
Our experience of wine is based on our perception of chemicals in the wine, but a lot happens to the signals that come from our tongue, mouth and nose. The inputs from the senses smell, taste, touch and vision all affect each other, and are recombined and processed a fair bit by the brain before we are consciously aware of them. Flavour is a multimodal perception: that is, it’s the result of the combination of more than one sensory modality. Added to this, we also bring our previous knowledge and experience of wine into play, and this can affect how we perceive the wine. And there’s also the issue of biological differences that needs to be considered. Some individuals are much more sensitive to certain tastes than others, and it is well known that about a third of people are quite bad at detecting ‘bitter’ taste. We all differ in our ability to smell, too. The best example of this is the compound rotundone, which is responsible for pepperiness in Syrah: about a fifth of people can’t smell this at all.
This sort of complexity is mightily inconvenient, but we need to bear it in mind when we approach wine tasting and evaluation. Currently, the way wine education is taught assumes that everyone is experiencing the same thing when they taste wine, which we now know to be a false assumption. But the fact that what we know about a wine has the potential to shape our perception of that wine reinforces the fact that the culture, history and context of wine is really important, and reinforces the role of the sommelier and wine service in restaurants as vital in helping people to get the most out of their wine.
So there we have it: a brief introduction to a topic where our knowledge is currently growing at quite a rate. In some ways, this sort of science talk grates a bit with the wonderfully artistic side of wine. Most of the world’s great wines were made before we knew very little about wine chemistry at all, by people who observed, experimented and had a language for wine totally at odds with scientific thinking and terminology. But I’d argue that in the right hands – of those who really understand what makes wine special, and who can tell the difference between authentic and industrial wine – then the sort of understanding that science brings to wine can help more people produce interesting, compelling wine at prices that normal people can afford. And that has to be a good thing.
Jamie Goode is a London-based wine writer with a scientificbackground. After completing a PhD in plant biology he worked as ascience editor. Smitten by the wine bug, he began consumer-focusedwebsite www.wineanorak.com in 2000. The success of this site led tomore work, and in 2005 he landed a national newspaper column and abook deal. Now he devotes all his energies to wine, and his secondbook, Authentic Wine, was published by University of California Pressin September 2011, authored in conjunction with consultant winemakerSam Harrop.
As both a wine lover and a science geek, I love the piece.
I was wondering if you could answer, from a chemical point of view, why it is that some spirits (i.e. wine, beer) can improve and then deteriorate in the bottle overtime and others (i.e. a 12 year old bottle of scotch) won't improve and/or deteriorate in the bottle over time. Does it come down to alcohol level? Or is that just a piece of the puzzle.
Always been curious. If you don't have the time to answer, but you could point me in the direction of a source for the answer, I'd love to hear it. Thanks.