Premox in White Burgundy
Premox – maybe we should be more forgiving to SO₂
Last week we opened a vertical of five Meursault Blagny 1er Cru from Domaine Antoine Jobard at our annual Non-Primeur. The wines were showing exceptionally well, and it was fun comparing vintages as well as seeing how white ages.
Any serious white Burgundy tasting seems to open up the discussion of Premox. Thankfully, we had no premoxed bottles, but the issue was naturally a topic of conversation. As someone passionate about Burgundy and prone to opening too many bottles for work and personal consumption, the issue of premature oxidation (or premox) is very real. It is as frustrating as opening a cherished bottle to find it is affected by cork taint. Like cork taint, premox seems random, affecting a few bottles in a case whilst others in the same case shine. Unlike corked bottles, the real source of premox remains undetermined despite extensive research. The topic is intrinsically linked to white Burgundy, though is not exclusive to the region. It seems that the issue first arose with the 1995/1996 vintages. Though some producers were more noticeably affected, it turned out that no one was immune. The worst affected vintages are 1996-1998[i] and a decreasing number of incidences have been reported from recent vintages. Yet, premox continues to cast a shadow over white Burgundy and this is most likely due to the fact that despite more than a decade of intensive research and writing, the origin of premox remains a mystery. There are many potential theories, many accepted factors, but the whole picture remains a mystery. As I explored the issues, it soon became apparent that this blog was becoming longer and also far more technical in places than one would expect from a wine-merchant’s blog. Huge apologies in advance.
Premox refers to a process of unexpected fast ageing. In a shorter time than one would normally expect, the wines become darker in colour with a ‘flat’ taste or start to display aromatic notes of baked apples, lees and in extreme cases a ‘madeirised’ note. As complexity advances too soon, it can result in a loss of balance. It is important to stress that I am only referring to cases where the development of the wine is unusually fast. Most white Burgundy should be drunk within 2-5 years of vintage and (with the exception of the very best producers and vintages) I would usually drink my 1er Crus within 3-7 years of vintage.
Some technical background. The process of oxidation is linked to two important chemical reactions:
Both types of oxidation usually happen simultaneously, but it is possible for a wine that looks youthful to taste old and vica versa. In addition, the oxidation of flavonoids in certain instances can be reversible, whereas my understanding is that the oxidation of ethanol is not reversible.
Arne Ronold MW[ii] argues that a trend for less sulphur dioxide (SO₂) is the main cause of premox. SO₂ is a powerful anti-oxidant and has been used in winemaking since Roman times to protect the wine from oxidation. At higher levels, however, it has an organoleptic effect, either masking the fruit flavours or in higher doses giving a burning sensation. Increased tastings of young wines by critics and private individuals during trade tastings has increased the need for the wines to show fruit and purity in youth and therefore winemakers have reduced the amount of SO₂ at bottling. Normally, there is a sharp reduction of SO₂ during the first 18 months after bottling due to it binding to oxygen, highlighting the importance of adding sufficient SO₂. In the 1980s it was common to have an excess of SO₂ (easily 50parts per million (ppm)). During the mid-1990s demand for less SO₂ combined with the ultra-clean nature of the 1996 vintage, started a trend towards less SO₂ in wines. Allen Meadows also observed that the introduction of sorting tables further supported this trend (Burghound issue #16, Oct 4 2004). SO₂ at bottling was reduced by some producers to below 20ppm, which in many cases was insufficient to protect the bottled wine. In response to research by the Vins de Bourgogne (BIVB), levels of SO₂ have gone back up to 35-40ppm in recent years for most producers. Though it seems likely that SO₂ levels may be linked to the cause of premox, it does not explain why only a few bottles in a case are affected and not the whole batch.
The randomness of premox may be related to corks, especially as there is an intrinsic link between corks and the oxygen transmission rate (OTR). The elasticity of corks is important to OTR and it has been suggested that as the world woke up to wine during the 1980-1990s and the international demand for cork increased beyond what was expected this had a negative effect on the quality of the corks. Clive Coates MW [iii] argues that international demand for corks has resulted in trees being stripped too soon or too low on the tree and cork harvested from trees that were planted in the wrong places reducing the elasticity of the corks and increasing OTR. Though there certainly was a period in which spongy corks became more common, this is not a problem unique to Burgundy. The density of the individual cork and its length is important, and this also explains why magnums age slower and suffer less from premox as they have a decreased surface area exposed to oxygen (and why half bottles age more rapidly). Using longer corks could be a possible way of reducing the OTR, but there is a problem with traditional shaped Burgundy bottles. Burgundy bottles taper outwards resulting in a looser environment near the bottom of the cork, increasing the OTR. Producers are increasingly aware of this and are using different shaped bottles with longer necks, giving a uniform diameter along the length of the cork as well as using longer corks. Some have suggested that a move towards using silicone coating (rather than paraffin) for corks may be linked to premox as silicone has oxidative properties and in addition is more slippery allowing more oxygen ingress. Julia Harding MW (2008) reported that experiments (on 2004 vintage) showed no difference in oxidation between paraffin or silicon treatments. Most producers are aware of the issues related to corks (both in terms of premox and cork taint) and huge investments are being made in buying high quality natural corks as well as introducing (and experimenting with) alternatives such as Diam/Nomacorc, screwcaps or Vinoloks. Diams especially seems to have acquired a strong following in recent years with leading producers such as Lafon (since 2013) and Domaine Leflaive (since 2014) using Diam corks while Sauzet and Niellon are using longer natural corks[iv].
Changes in winemaking techniques in general have been linked to premox. Julia Harding MW reported experiments done by BIVB that showed that the timing of malolactic fermentation (MLF) may be relevant. MLF is a natural process where the harsh malic acids are converted to softer lactic acids and in Burgundy this usually takes place after alcoholic fermentation is complete. Traditionally, an early start to MLF reduced the risk of oxidation between alcoholic fermentation and the start of MLF allowing a more gradual integration of SO₂ and higher levels of sulphur containing compounds (thiols) were found that protected the must[v]. During the mid-1990s MLF started to take place later (as a result of climate change and winemaking) and this may be linked to the issue of premox. It is certainly supported by the fact that premox seems more obvious in high acid vintages such as 1996, 1999, 2002 where acidity delayed MLF, whereas there are surprisingly few incidences in 2003 (yet, it could also be linked to the fact that lower levels of sulphur dioxide (SO₂) were added to high acid vintages as the acidity was deemed to have protective qualities). Etienne de Montille at a debate on Burgundy organized by the Institute of MWs in 2013 argued that it is important to increase the phenolic content of the wine to protect it against oxidation and argued for a higher quantity of lees during ageing for their protective anti-oxidative influence. Similarly, it has been suggested that premox may have been linked to the introduction of new oak, which started in the 1980s, but reached its height in the 1990s (in terms of both the number of growers using new oak and the percentage of new oak used), but has been decelerating since. Nomacork previously pointed towards bottling lines as a potential cause for premox[vi], suggesting that the free oxygen available at the beginning and end of a bottling run is higher. This combined with the small production of Burgundy, especially for the top wines, would explain a comparatively high incidence of premox on white Burgundy with randomly affected bottles reflecting the start or finish of the bottling run. The research, however, does not explain why this ‘suddenly’ became an issue in 1995/1996.
The use of presses has also hit the headlines with the use of pneumatic presses becoming more common in the mid 1990s. A modern pneumatic press keeps fruit whole, whereas the hydraulic basket press breaks up the skins liberating more phenols to the juice. As a result, pneumatic presses allow for a much gentler extraction in a more reductive environment. Bernard Hervet Faiveley, at the same debate on Burgundy mentioned earlier, discussed how they had returned to basket presses as the increased amount of phenolic particles have a protective influence against oxidation. In addition, it seems that the reductive environment of the pneumatic press may have a negative effect in terms of oxidation. This is a slightly inverse argument, as the old fashioned basket press allows for a certain amount of exposure to oxygen during the pressing, yet this seems to make the finished wine less prone to oxidation. Earlier I noted that there are two types of oxidation and strangely it seems that if the flavonoids (extracted from skins, pips and stems and more likely to be extracted in an old-fashioned basket press) oxidise before fermentation this can be corrected with sulphur dioxide and adds stability to the final wine. This is a process referred to as hyperoxidation (or browning of the must) and though it is widely practiced, it is still unclear exactly how it provides this extra layer of protection.
A number of other sources have been quoted as being linked to premox. Some have suggested that rot (noble and grey rot) at harvest can be the cause of premox as it produces laccase, which accelerates oxidation of wine, but this does not explain why ultra clean vintages such as 1996 have been affected most. Global warming has been pointed to as a cause of premox as it produces lower acidity, increased alcohol and riper fruit, which means less phenolic content to protect the wine and less acidity, which offers natural protection against oxidation. In fact, much of the early research linked premox to acidity levels. The lower the pH the higher the free SO₂ needed to protect the wine. However, as mentioned before this does explain why high acid vintages such as 1996 were affected, while 2003/2009 seem (so far) to have been less affected. A number of articles have focused on viticultural changes as the source of premox, with covercrops being highlighted as a possible culprit. Covercrops allegedly decrease the amount of glutathione, a natural sulphur containing anti-oxidant. However, this does not explain why those that adhere to the strictest forms of biodynamic viticulture and lute raisonée seem to have had fewer incidences of premox.
It is increasingly evident that the randomness of premox is at the heart of the complexity of understanding the problem and frustratingly, it still remains impossible to determine the origin of premox. As we have seen there are many hypotheses given as to the origin of premox, but none are fully satisfying and most likely it has been a combination of issues. Jasper Morris MW points towards the ‘zeitgeist of what I call eco-elegance’ as the possible source. In the aftermath of Claude Bourguignon’s soil research and the rise of biodynamics, it is clear that Burgundian producers were looking to produce more ecologically sound wines. This combined with a desire to produce more elegant, less phenolic wines that were approachable earlier with no interference of SO₂ for critics to score and consumer en primeur samples has resulted in a number of changes in the vineyard and winery. If the issue of premox is related to oxidation, then it does seem likely that the issue has been a combination of more fragile juice in search of elegance (expressed by using pneumatic presses and changing the habits of lees handling) and a reduction of free SO₂ at bottling to produce wines that show clean in youth and a problem related to corks (and the OTR of individual corks playing an important part in the randomness of the problem). The impact of viticultural practices such as covercrops, the increased use of new oak and the timings of MLF are more difficult to establish.
The world has woken up to premox and the BIVB hired 2 full time oenologists as early as 2006 to study the problem[vii]. Producers are raising the amount of free SO₂ at bottling and changing corks looking for thicker corks with higher density that reduce OTR over the long term. Cork quality has become a lot better and an increasing number of producers are looking at alternative closures that allow for a consistent regulated OTR. In addition, it is now possible to measure accurately (and cheaply) the amount of dissolved oxygen in the wine (that binds to the free SO₂ ) meaning that it is possible to guess more accurately the amount of SO₂ required. In addition, producers are looking at increasing the phenolic content of the must and increasing oxidation of the must by re-introducing basket presses or by deliberately micro-oxygenating.
As a result, the good news is that premox is far less prevalent than a decade ago. It is great to see increased use of alternative corks and I think we need to accept that great white Burgundy will have higher levels of SO₂ in youth, which may make some of the great wines destined for long term ageing not simply tight but potentially unappealing in youth. Writing this, I fear a huge number of responses from customers and critics, but I truly believe that as consumers we should be more forgiving of SO₂. It is important to note that I only think raised levels of SO₂ are acceptable in wines that require long term ageing. Wines that should be consumed within 1-5 years do not require and should not be spoiled by high levels of SO₂. I would expect that Grand Crus from a great vintage should last a minimum of 10 years, yet one cannot expect this from a lesser vintage or a lesser wine. Personally, I would expect most white Burgundies at 10 years to have passed their prime drinking window. Only the truly great wines from outstanding producers in exceptional vintages could be expected to last much longer. Those that have read my article on drinking windows know that I seek out old bottles, but I know that one takes a huge risk opening a bottle of 1989 Le Montrachet Lafon (two bottles were still divine in 2012) and a 1929 Corton-Charlemagne in 2013 was interesting rather than a pleasure to drink (it would have been so much better 60 years ago).
When it comes to old bottles there are only great individual bottles. White Burgundy that is over 15 years old, even from the best producers, will show variation depending on storage as well as the wine itself. I agree with Burgundy expert Allen Meadows that ‘there is only one almost foolproof way around the premox problem and that is to drink the wine young’. However, although drinking wine young is avoiding risk, it also avoids the excitement that comes with great wine. Personally I continue to choose the excitement of ageing my white Burgundy and the good news for my burgeoning cellar is that premox seems less prevalent now than it was a decade ago.
There are two final points worth mentioning. Firstly, the fact that the issue of premox is intrinsically linked to white Burgundy may have been the result of it having been the most common white wine to be cellared rather than the only region affected. There is some evidence that white Bordeaux, white Rhone and Alsacian wines have also been affected and even that the problem is not solely related to white wine, but also to red wines[viii]. Secondly, Jasper Morris MW has raised the idea of what I call ‘resurrection wines’ when it comes to premox. Comparing it to white Rhone, he comments on wines that appeared to have all signs of premature oxidation but after a few more years of cellaring these re-emerge from this state like a phoenix[ix]. He is not alone is mentioning this ‘apparent reversal of oxidation, yet personally, I have too little experience with this phenomenon, though a number of times customers have reported that a wine that seemed to suffer from premox lost its golden colour and produced fragrant fresh fruit after a few hours in a decanter on ice.
For those interested in reading more about the subject I would look at
- Dedicated website to discussion on premox: http://www.gdeschamps.net/wiki/doku.php
- Tong Issue 11 Chardonnay, The Problem of Premox by Arne Ronold MW, http://www.tongmagazine.com/
- Clive Coates MW: http://www.clive-coates.com/observations/premature-oxidation
- Julia Harding MW on purple pages: http://www.jancisrobinson.com/articles/premature-oxidation-results-from-beaune
- Jasper Morris MW, The Annual Report 2014, Premox: The Premature Oxidation of White Wine
- Jasper Morris MW, White Burgundy out of the Woods?, The World of Fine Wine, November 2015
- Allen Meadows http://www.burghound.com/
[i] Clive Coates MW, premature Oxidation of White Burgundy, An extract from My Favourite Burgundies
[ii] Arne Ronold MW, Tong Issue 11 Chardonnay, The Problem of Premox
[iii] Clive Coates MW, premature Oxidation of White Burgundy, An extract from My Favourite Burgundies
[v] Julia Harding MW, Premature Oxidation: Results from Beaune, JancisRobinson.com 18 dec 2008
[vii] Allen Meadows, Burghound Issue 23, July 2006
[ix] Jasper Morris, The Annual Report 2014, Premox: The Premature Oxidation of White Wine