wa2.gif (4241 bytes)

abut9.gif (3095 bytes)

abut12.gif (3207 bytes)
abut10.gif (3636 bytes)

abut11.gif (4039 bytes)

Fighting cork taint: are screwcaps and plastic corks the answer?

Jamie Goode analyses the results from an significant independent study on the effectiveness of wine bottle closures. This scientific paper, published on July 12 2001, has thrown up some surprising results, and the ongoing trial it describes promises to answer the key question of whether alternatives to cork are suitable for long-term ageing of wine. 

What's the single most controversial issue in the wine world today? Without doubt I'd argue that it's the good old cork taint debate. Plenty has been written on this subject, but what have been lacking are decent independent data. In fact the whole area has been so sullied by commercial interests (cork manufacturer Amorim has been busy offering hospitality to journalists, and I'd be surprised if synthetic cork manufacturers haven't been doing the same), that it is only solid results that will settle this issue once and for all. Yet more anecdotal observations and strongly held opinions won't really do. This is why I was delighted to receive through the post the first results of the Australian Wine Research Institute (AWRI) Wine Bottle Closure Trial. Published in the Australian Journal of Grape and Wine Research, this highly significant study looks like it will help answer the question of what really is the best way of sealing wine bottles. To those unfamiliar with the scientific literature, this 41 page paper would be pretty heavy going, so in this feature I'll attempt to unpack what the data really show, and assess their real significance.

The key issues
The widespread existence of cork taint, caused by chloroanisole compounds, is now universally acknowledged, although there is still debate over the prevalence. Whether it is 2%, or 5% - or as some claim, as high as 7% - it's a huge problem, and so wine producers have been looking for alternative closures for some time. But are these alternatives - whether they are screwtops, plastic corks or agglomerate corks - suitable for long-term storage of wine? It's one thing bottling an inexpensive quaffer with a plastic cork, but would a manufacturer be wise in using the same closure on a wine intended for 10 years in the cellar? And even if the plastic cork would maintain a good seal for this length of time, would the wine age differently than if it were sealed with a real cork? These are the questions this study is designed to address.

Experimental design and statistical treatment
The key to any study is good experimental design. You have to plan the experiment so that with the appropriate use of statistics, you can be sure you are able to answer the questions you are trying to assess (the results have to be shown to be 'statistically significant'). In this sense, the study here seems to be quite carefully thought out, and the statistics look pretty solid. The essential details of the experimental design are as follows:

  • 14 different closures were analysed (see different closure types used in the study): one ROTE (stands for 'roll-on tamper evident'; a posh name for a screwcap, or 'Stelvin'), two 'technical corks' (part cork, part synthetic substance), two different grades of standard cork and nine different plastic corks
  • The wine used was a 1999 Clare Valley Semillon that had been fermented and stored in stainless steel tanks
  • Care was taken to ensure consistency in bottling with the different closures
  • Two batches each of 300 bottles were sealed with each closure, with the exception of the screwcap (which had a single bottling run of 800)
  • After bottling the wines were stored in a temperature controlled facility, where storage was randomised, with each bottle given a separate number

A range of tests were then performed at three monthly intervals, with the variables studied including:

  • Extraction force required
  • Ease of reinsertion into the neck of the bottle
  • Incidence of leakage
  • Concentration of free and total sulphur dioxide (SO2)
  • Concentration of ascorbic acid
  • Browning measures
  • Wine sensory analysis, using a panel of experienced tasters working under strict guidelines
  • Interrelationships between sensory data and wine composition analysis

The most important results and conclusions
First, it's important to bear in mind that these are the first results to be published from this ongoing study. There is enough wine bottled for the study to continue for a decade, and as such these results are best viewed as preliminary. Having said that, even at this early stage some conclusions can tentatively be drawn.

Importantly, this study showed that the wines that retained the highest concentration of SO2 and ascorbic acid showed the lowest degree of browning, and in the sensory analysis tasted the freshest. It seems that SO2 concentration, which is relatively easy to measure, can act as a good predictor of future browning and a useful proxy for oxidation. It therefore offers a convenient way of measuring the effectiveness of a stopper.

But how did the different stoppers perform? The curves plotting free SO2 concentration against time show a pronounced dip with all the stopper types over the first year, that then begins to flatten out. The best performance is from the ROTE (screwcap), closely followed up by the Altec and then the Twintop. The two standard corks perform pretty well, as do the Auscork, NuKork, Nomacorc, Aegis, Supremecorq and Integra: these all perform surprisingly similarly. Slightly less effective are the Tage and the ECORC, and the only significantly poorer performer seems to be the poor old Betacorque (which showed dismally). With regard to browning, it's even harder to separate the various stoppers, with only the Betacorque performing less well than the others. The ROTE seems to perform the best again, but not significantly so.

The sensory analysis threw up some surprising results. The ROTE closure produced a rubber-like flavour/aroma in the wine after 18 months. This is unexpected and alarming, considering that the most of the Clare Valley Riesling producers decided to switch to screwcaps for the 2000 vintage, in a well publicised move. Considering that there is a long track record of bottling Riesling using screwcaps, and many library reference samples are available, it's an odd result. The authors of this paper suggest that it could be a consequence of the lack of oxygen, and that leaving a slightly higher headspace may have alleviated this anomalous result. Apart from this, no plastic-type taint was associated with any of the synthetic corks. [Note added later: Peter Godden, lead author of this trial, communicated the following by e-mail: "We are very confident that the 'rubber-like' character is not a taint, but is an unwelcome modification due to chemically reduced sulfur, as a result of lack of oxygen. However, it is certainly an important character in ROTE-closed wine, and we have highlighted its existence to avoid mass-bottling of wine under extremely anaerobic conditions which might then develop a similar character somewhere in the future. However, you can see from the other sensory data that it has not detracted from the fruit characters and intensity of aroma of the wine to a great extent so far, although the intensity seems to have increased at 21 and 24 month post-bottling testing."]

As many professionals have long suspected, the Altec cork tainted all the bottles with TCA from an early stage, despite providing a very good seal. Surely the results of this study should spell the death knell for the use of this stopper by producers who care at all about the quality of their wines. Two of the 14 one-plus-one closures were affected, and for both of the normal corks, four of the 14 bottles were tainted with TCA. These results were confirmed with GC/MS (gas chromatography/mass spectrometry; an analytical technique).

I was reassured by the fact that the performance of the synthetic closures was very similar to that of the normal corks, but without the taint. What this study doesn't show is that plastic stoppers are unsuited to long-term ageing of wines. We'll have to wait for a few more years' data yet, but I suspect that the best performing synthetic corks will be just as good as real corks, but without the variability in performance. This would be an exciting result for all those who've gone to the cellar only to find that their irreplaceable aged bottle has been ruined by cork taint.

Of the synthetics, the best performers in the sensory analysis seem to be the Auscork, Aegis, NuKork, Integra, Supremecorq and Nomacorc. The poorer performers are the ECORC and Tage, and of course the poor old Betacorque. Thus both the extruded and moulded plastic corks show a mixed performance.

An important question to consider: do we really want a perfect seal? Assuming that people generally like the way that wines change with age when they are sealed with corks, my assumption would be that all we are asking from the perfect stopper is that it should seal as well as a cork and no better, without the risk of taint from TCA. The fact that screwcaps provide a better seal than corks may actually work against them. Indeed, in the discussion of this paper the possibility that the rubber-like aromas were the result of them providing an almost hermetic seal: the report suggests an extended period of anaerobic storage may not be desirable, and this 'reduced' aroma may be avoided by allowing a slightly greater headspace or by treating to eliminate sulfide compounds before bottling.

Rather teasingly, this paper mentions that a separate study using measurements of oxygen permeability will be reported elsewhere. These will be of great interest with regard to the role the stopper plays in the normal ageing of wine. I'm also looking forward to seeing the next set of results from this well planned and executed ongoing study.

See also: different closure types used in the study

Reference for the study discussed here: Godden P, Francis L, Field J et al 2001 Wine bottle closures: physical characteristics and effect on composition and sensory properties of a Semillon wine. 1. Performance up to 20 months post-bottling. Australian Journal of Grape and Wine Research 7:64-105

Conflict of interests: None. I have never received hospitality or remuneration from any cork or synthetic stopper manufacturer, and I have no axe to grind or hidden agenda on this topic. Acknowledgements: Thanks to Val Rechner of the AWRI for sending me a copy of this report.

Back to top
Back to the cork, plastic and screwcaps index page
Back to wine anorak index page