Light-struck wines?
A version of this article first appeared in Harpers Wine and Spirit, March 2006, updated July 2009.
Light is well known to promote chemical changes in foodstuffs.
Brewers have know about the flavour impact of light on beer for many years. It’s widely reported in the beer industry that beer goes ‘skunky’ when it’s exposed to light, even just in the time it takes you to nurse your beer on a lazy summer’s afternoon.
Beer off flavour in the presence of light was noted as long ago as 1875, which is why, until the marketeers got involved, pretty much all beer was bottled in dark glass, amber having been found to be the best at filtering out most of the harming wavelengths.
For wines, it is much more of an issue for whites and rosés. In reds the high tannin content offers a protective veil, binding riboflavin (more below). And the issue can pretty much be avoided by using amber bottles, but with the marketing of any product, many considerations are important, not least the packaging, and we are seeing an increasing number of packaging options including clear glass, in both the beer and wine aisles.
A photochemical reaction occurs in both beer and wine. The responsible light wavelengths are between 350-500 nm, but activity is noted to peak at 370 and 440 nm, so in the near-ultraviolet and part of the visible wavelengths. Amber bottles filter out 97-98% of these wavelengths, which is why they’re the best. Dark green glass filters out just 63% of these critical wavelengths, while white flint (clear) glass filters a meagre 10%, so it’s only when we put beer and wine into clear glass (either bottles for the shop or pints/wine glasses for the garden) that the light degradation issue arises.
Light penetrating the glass excites riboflavin (vitamin B2, present in beer in concentrations around 1mg/litre, and in wine at concentrations at 0.4mg/litre), which acts as a catalyst, cleaving the compounds responsible for the classic bitter flavours in beer.
In wine, riboflavin, found naturally in must, reacts with cysteine and methionine (amino acids). Wine naturally contains 1-4g/l of amino acids. This reaction forms hydrogen sulphide and mercaptans, notably 3-methyl-2-butene-1-thiol (MBT). MBT is the skunky thiol (mercaptan), smelling pungently of leek, onion, cooked cabbage, wet wool, soy etc. It is highly flavour and aroma active, and humans can smell it at concentrations similar to TCA, namely about 4 ng/l. In pale beer concentrations as low as 1ng/l can make the beer unpalatable.
Paul Hegarty, head of communications for Coors brewers, has 15 years technical experience for brewers where ‘sunstruck’ was high on the agenda. He was prosaic about the issue for beer, saying “light struck has a skunky character, and rubbery – the flavour of beer when you’re sitting in the garden on a sunny afternoon. Experts might define it as a fault, but consumers might quite like it. It reminds them of sunny holidays.”
Dr Ellen Norman, head of analytical development and Brewing Research International said: “Research has suggested that consumers do not agree on whether they view lightstruck as a negative flavour in beer – some actually prefer it. We have also seen studies where consumers are unable to pick up differences between a beer before and after the beer has sat in the light.”
However, hop cones can be extracted using supercritical CO2 (of late renowned in the wine industry for reducing TCA levels in cork granules). The MBT-forming compounds undergo another process to make them light-stable, enabling the resulting beer to be stored safely in clear glass. It is thought the number of brands produced using these processed hop products is relatively low.
Wine’s propensity to reductive flavours
But wine never got that ‘skunky’ aroma. Peter Godden, group manager of industry development and support, at AWRI, said: “the formation of lightstruck flavour is exacerbated by low oxygen in both beer and wine, where the wine already has a propensity to develop reductive characters, but the reasons for this are different for beer and wine, because the compounds thought to be responsible in beer do not exist in wine. However, lees-aged sparkling wine may be an exception, containing more methionine, and other compounds, than other wines, and so may therefore be susceptible to formation of lightstruck flavour through a mechanism that is similar to that in beer.”
We know some grape varieties and wine styles are more susceptible to reductive characters than others. Not every wine reacts the same way. It also depends on how the wine is made, and how many amino acids it contains. Given the right (or wrong) light conditions, mercaptans develop. Gerd Stepp, who was winemaker for Marks and Spencer at the time (now consulting to them), explained: “mercapans are relevant in winemaking. A lot of the precursor in wine is hydrogen disulphide (H2S). H2S has a high threshold, but it can continue to mercaptans, and these have a low threshold [4 ng/l]. So tiny amounts of H2S, almost undetectable, [can go into the bottle] but once the bottle is closed, the wine can develop hints of mercaptan.”
In the case of lightstruck cysteine and methionine amino acids come from yeast autolysis, the time spent on lees. Thus it is not so common to find champagne in clear glass. For wine with low amino acid content, so without lees contact, the chance for wine to get lightstruck is much smaller. But, Stepp added: “You taste it much more easily in aromatic or neutral white wine, onion flavours come through. With red wines there is more air, oak, open rackings, so H2S is lower.”
Paolo Bisol of Ruggeri said: “Our rosé has been in clear glass since 1950, but pinot bianco and pinot nero are less sensitive. It might be better in dark glass, but you can see the colour. However, it is important for prosecco to be in dark glass because prosecco is sensitive to light. It is light, fresh, young and fruity.”
Sparklers
Wines such as traditional method sparklers, with their long lees ageing and consequently high amino acid content are susceptible to light struck. Jean-Baptiste Lécaillon, director of vineyards and wines, and chef de caves at Champagne Louis Roederer said: “amino-acids such as cysteine are very important in wine as they are responsible for aromatic structure of the wines. In the case of sparkling, a large part of the aromatic profile is due to autolysis of yeast proteins. During the autolysis, some enzymatic reactions separate the proteins into thousands of amino-acids that are aromatic and responsible for the “yeasty”, “nutty” characters.”
He offered a number of solutions. “Our only focus is to protect the wine from UV. You can do it by using green bottles that filter 92% of the UV, which is enough to protect the wine during a reasonable sun exposure. A strong sun exposure would be bad for any compounds of the wine…. so it is to be avoided. For Cristal we protect the clear bottles with coloured cellophane. This cellophane is specially designed to filter 98% of the UV, which means, if the bottles stay wrapped, it is better protected than a green bottle. This is why we add in all our cases of Cristal a leaflet which recommends keeping the cellophane until the last moment.” Of course this doesn’t mean the retailer or consumer will keep the cellophane on. The bottle looks pretty stylish without the foil, even with a warning note.
Marks & Spencer
Marks & Spencer use no clear glass at all for their wine range. Five or six years ago all the wines were moved to green or amber glass. Stepp said: “If wine is in a presentation box then it might be in clear glass e.g. champagne, but the box protects the wine.”
But it’s not always easy to get the most protective colour of glass. Stepp said: “UVAG green [a type of bottle colour] has a higher UV protective quality, they use a different oxide for colouring. But it is limited by availability, so we have to see if we can find it on the local market, and if not we go with the normal green.” He said it costs too much to transport empty UVAG bottles to the country of bottling.
The M&S policy makes it difficult to draw attention to their rosé wines, which many other retailers and brand owners market in clear glass to draw attention to their attractive colour and mood suggestions. The M&S technical team insist on green glass to protect the wine, and this outweighs any marketing considerations. To signify rosé to consumers, the bottles all have overtly pink labels and distinctive pink capsules of the same hue to illustrate the wine’s tint.
Stepp said: “Commercially and presentationally it is quite a challenge. The rosé wine colour is less visible, and the colour of rosé is one of the key features of selling a rosé. Green glass is less attractive.”
Clear glass coating
There has been some experimentation with coatings over clear glass which filter the UV and allow the presentational benefits of clear glass. Godden said: “In 2001 we conducted a trial using a clear plasticised coating for the outer surface of clear glass bottles. Its purpose was to block UV light from damaging wine. A chardonnay wine was packaged in coated bottles, and also in the same bottle without the coating. The bottles were then exposed to UV light (approximately 50 cm from the light source of approximately 360 nm) for 21 days.
“During sensory evaluation the wine in the coated bottles was rated higher for freshness and for fruit characters, and lower for a whole range of reductive descriptors (reduced sulphur, smoky, burnt rubber, bacon, burnt match, gunsmoke etc.). The results were highly statistically significant – the coating appeared to work extremely well.”
Stepp explained why M&S had not pursued this option. “A UV protective coating can be applied once the bottle is finished.. The coating has to be completely transparent on clear glass, otherwise it defeats the purpose. But if the coating is 100% clear, its protection only goes to 400nm. So the 440nm peak was not covered. And to get protection over 400nm, the protective coating becomes coloured.”
Greater UV protective capabilities give a cloudy quality to the film, which may explain a lack of progression with this technology. Stepp suggested this avenue may still be worth exploring if there is some kind of bottle coating that is an improvement on what’s currently avail in the industry.
The in-store environment
In a store environment under some types of fluorescent lighting wine in clear glass can develop noticeable degradation flavours after just a few hours. Stepp said: “Green glass protection is not 100%, so store lighting is also important, as is stock rotation. It doesn’t take long for mercaptans to form. Generally the degradation of quality on shelf in clear glass would be quite high. Rosé is slightly more unstable because colour can turn. We’ve done a lot of work in store lighting, we’ve done tests with filters on certain light. We are using lights now that are better. And this is also relevant for other products, chocolate sensitive to light.
Howard Winn, who was quality manager for beers, wines and spirits at Sainsbury’s, said: “We have a few wines in flint glass, mainly the usual suspects, rosé, sweet Bordeaux. It is an ongoing project to move away from fluorescent lighting, reducing the lighting in the BWS [beers, wines and spirits] aisles and moving to lower energy lights positioned away from the fixtures. Deeper steel shelves have also helped reduce the exposure of bottles to light. Our lighter sherry styles have recently been moved to different glass which offers higher UV protection. Similarly our rosé Champagne moved to green glass a couple of years ago. It’s all about protecting the wine.
Lecaillon was optimistic about the trade’s level of knowledge: “there is the question of bottle shops and fridges, It is obvious that if the wine is under full light for a while, there is an important risk of light struck, but I believe the trade is aware of that problem and take all necessary measures to protect the wine.”
With all the right precautions, light struck need not be a big issue, but the creative tension between brand image and positioning, and technical precision, means the issue is still relevant. Stepp offers yet another option, saying: “There is quite a bit of work from our side on the wine making. If we did progress on clear glass with rosé we would want to make the wines in a way that they are less prone to mercaptans, for example, have less of the amino acids cysteine and methionine; have a slightly faster fermentation, focusing on fresh fruit flavours with less lees contact. Our experiments are ongoing.”