Weekly Article
Sydney Time
Copyright © Ric Einstein 2009
More on the Fizzics of
Corks - (Article with Questions and Answers)
(12 May)
(John Casey is the former Chief
Chemist for McWilliams Wines Pty Ltd. He completed his Associate
Diploma in Applied Chemistry at The Melbourne Technical College, (now RMIT) and worked for a number of research and industrial
organisations, including 32 years at McWilliams Wines. After
leaving the wine industry, he worked for four years in the ‘residue
chemistry’ of foodstuffs, soil, water and air. He has written about
thirty technical articles about wine packaging, closures, the chemistry
of wine, and the physical chemistry of sparkling wines.)
Introduction
(This article started off the same way as the
first one which John did, with email feedback that lead to a series of
questions from by me and answers provided by him.)
Things may not always be the way they appear, especially
when it comes to science and what goes on inside a bottle of wine. As
much as science knows about the development of wine, there is a gigantic
amount that is either not known or not fully understood. In some cases,
what may appear to be simple logic is incorrect and this further
confuses the situation. Take this for example. In July 2007 in the Drops
'n Dregs section, the following appeared...
These Guys Couldn't Lie Straight in Bed
Thanks to David More for sending me this one. The Australian cork
spin witchdoctors are at it again. Have a look at
And if you open up a bottle that has oxidised, ask yourself how could it
be so? It must be your imagination because the cork people tell us corks
are impenetrable to air. But hang on a second, one of the main arguments
used by the cork lobby for using their product is that natural corks
breath; now how do they do that if they are impenetrable to air? But
then they do say corks are "resistant to humidity and oxidation." "
John Casey responded to these comments and his answers
are both fascinating and enlightening.
1. Yes, cork is impenetrable to liquids, but it is
permeable to liquid vapour by virtue of the minute channels (plasmodesmata)
through the cell walls. With diameters of about 0.6 microns, the
passage of liquids as such is not possible. When exposed to water
vapour, the surface of the cell walls eventually become saturated,
and net permeation stops at about 12 -16% moisture. However when
there is excessive hydraulic pressure on the cork in a bottle
of wine, there is a far greater entry of liquid vapour, (1 – 3+g),
which condenses as a free liquid in the cellular structure. This can
be seen under a microscope as a clear liquid, usually in the lower
end of the cork. The fact that red wine pigments are never seen in
the cell structure of the cork shows that there is no penetration by
liquid as such. In brief, cork is permeable to gases and vapours
but not to liquids.
2. A cork is a pneumatic seal. When a cork is compressed
in the neck of a bottle, its volume is reduced by some 40%, and the
air pressure in the cells is double atmospheric pressure. Under
these circumstances, there can be no net penetration of
atmospheric oxygen into or through the cork, although the cork
remains permeable to gases and vapours. This is ‘the cork paradox’.
Despite its permeability to gases, a cork in a bottle is an
effective barrier to atmospheric oxygen, and the only significant
permeation of oxygen is from the cork to the atmosphere. Any
ingress of atmospheric oxygen is between the cork and the glass.
This is why 25 mm diameter corks give better exclusion than 24 mm
and are preferred for long term storage.
“Breathing corks” was consigned to the scientific rubbish bin in
the 1930s, but winemakers are a sentimental lot, and they have a
irresistible urge to exploit obscurantism as a marketing tool. Cork
suppliers are just as bad in pandering to them. Even more alarming
is the huge amount of ‘research?’ resources used over the last
decade or so in trying unsuccessfully to prove that corks have
some involvement in the oxidation of bottled wines; in NSW
police terminology, it has been a gigantic WOFTAM. (- for our non
Australian readers, that Waste Of ****ing Time And Money.)
Wow - when I first read this
stuff I was wondering what John had been smoking; some of it sounded
pretty far out to me, wacky in fact but given that John has a
lifetime of experience in this subject and is a trained scientist, I
respected that and decided to ask some questions to see if there was
any real sense and scientific rational behind John's comments. (For
the record, John is a scientist; he is not a member of the cork
lobby or a pro screwcap advocate, so we are just getting an unbiased
scientific opinion from a man with no axe to grind. But before we
get to that, in the same email from John, there were more comments
in relation to this from the Drops 'n Dregs Section.
What the Heck is Going On?
Recently a winery was kind enough to send me samples of three of
their wines to review. The email exchange between myself and the
winery tells the story.
"Hi Xyz, Thank you for sending the samples of your 2003 wines.
Unfortunately the news is not good and I need to get your opinion
and possible further information from you before I do anything with
my tasting notes.
Three bottles and three dodgy corks! I was wondering what sort of
response I would get; shock, surprise or denial? Here is what it
said.
"Hi Ric, Over twelve months ago we made a decision to abandon
natural cork and the closure we chose was the 47ml Sabaté Diam. This
is a TCA treated composite cork, it is totally in round, it has no
seams or capillaries and it satisfies the premium end of the
restaurant market here. Further to this we have spent a considerable
amount on a new corking machine so we will never have a cork related
issue again. I must say that I am very pleased with the result of
this move.
John had some really
enlightening comments in relation to this situation and they really
made me start to think that perhaps he did really know what he was
talking about. The more I thought about what he has said, and the
further the email exchange went, the more sense it made. As the
exchange unfolded, the more the revelations and claims made by those
in the industry started to look, well .... like rubbish! Here are
John's comments on ":wine travel" and "leakage."
Headspace pressure is
the motive force for all leakage/venting in bottled wines.
Water does not run uphill, and wine does not come out of a
sealed bottle unless the internal pressure is greater than the
sealing pressure of the closure/cork. Strictly speaking, corks
do not “leak” at all, but they do vent excess pressure when gas
or liquid is forced past the cork. Moreover, it has been shown
that excessive hydraulic pressure causes a greater and faster
uptake of liquid vapour by the cork from the wine, and that this
leads to a premature softening of the cork and a significant
reduction in its sealing pressure; the absorption of the
additional water and ethanol vapour also leads to a lowering of
the wine level by several millimetres.
The major cause of
headspace pressure is the compression of headspace gases
by the piston-like action of the cork as it moves into the neck
of the bottle. Headspace pressure is not a problem in
well-managed bottling operations, although constant monitoring
of the corking machine is imperative. Further contributions to
the headspace pressure are made by the vapour pressures of
gases dissolved in the wine, mostly air and carbon dioxide,
and by expansion of the wine with increases in temperature.
Common sense tells us that
if we open several bottles of wine and in only one of them, the
cork is a soggy mess, the wine level is down and the wine is
showing signs of oxidation, it just has to be a ‘bad’ cork.
However, ‘common sense’, or ‘naïve realism’ also tells us that
the earth is flat and that the heavens revolve around my house
in North Ryde. Thus the bleeding obvious explanation of
“leaky” [sic] bottles is hopelessly wrong. As Julius Caesar
said, “The fault dear Brutus lies not in our corks but in our
bottling and corking operations.” Naïve realism is one of
the ‘great comforters’, (cf. Manning Clark), but unfortunately,
one of its by-products is ‘invincible ignorance’. The wine
industry prefers scape-goats to scientific facts.
The gauntlet has certainly been
thrown down with those comments (and some of them are more like an
old fashioned slap in the kisser!) Here are the questions I posed to
John and his answers.
Question:
If I read what
you say and take it literally, it seems that all oxidised wine
would be the fault of bottling issues rather than cork failure.
Answer:
Most post-bottling oxidation is caused by air contact in the
days before and during bottling operations. Malfunction at the
corking machine can cause leakage/venting, and it also adds that
little bit more air to the bottle.
Question: What
about flaws in corks where they leak down one side in a groove
in the cork? Surely that is a cork problem?
Answer:
‘Creased corks are caused by failure to lubricate the ‘jaws’ on
the corking machine, or if the jaws are worn or not aligned
correctly. Depending on the size of the crease, less pressure is
needed to expel liquid past the cork.
Question: There
has been a huge problem with random oxidation of some vintages
of White Burgundy; if it is a pressure issue, how come the Red
Burgundy wines have not been affected?
Answer:
White wines are much more susceptible to oxidation than red
wines because of differences in composition and differences in
their relationship to sulphur dioxide. Many winemakers know very
little about the chemistry of sulphur dioxide in wine.
Question:
Finally,
( as per my first question) surely
you must admit that “some” corks are “defective” and allow air
into the wine, and if that is the case, what do you estimate the
percentage of failure of corks are due to pressure problems vs
defective corks?
Answer:
From
testing and casual observation about ten years ago, I estimated
that about 15% of bottled wines in Australia had excessive
headspace pressure. About five years later, it seemed to be
about 10%. I am guessing that
excessive headspace pressure
is now less than 10%. Over the last thirty years, I have
only seen maybe 5, 10 at most corks with an inherent defect, in
a bottle of wine. I once saw a 38mm inserted sideways into a
bottle, it must have fallen over in
the corking chamber. It looked like a lolly, but there was no
leakage.
I wasted some of the best years of my working life in the futile
pursuit of ‘good’ and ‘bad’ corks only to realize after a couple
decades that the major problem was sporadic, excessive headspace
pressure. Among other things, I kept a record of the performance of
corks of different sources, grades and prices over five years. With
one exception, performance was pretty much the same for all batches.
Corks are graded for quality by the size and distribution of
lenticels, (the small, dark channels of plant debris at right angles
to the length of the cork). The ’higher’ grades of cork have fewer
and smaller lenticels. Aesthetics is the main consideration, as
there is not much difference in sealing pressures in the short term.
However, it seems likely that the higher grades would be more
enduring in the very long term. The other part of the problem was
the enormous expansion in the production of table wines beginning in
the mid 1960s. Wine-bottling changed from leisurely manual or
semi-automatic operations to mass production with multi-head
automatic corking machines. Not only did the speed-up multiply
problems, it allowed less time for close scrutiny of the operations.
That’s my excuse for being a bit slow out of the blocks; mind you, I
am a stayer.
Another historical note.
Post-bottling oxidation, or so-called “random oxidation”, only
became a problem in the late 80s and 90s, following the widespread
mistaken belief that ascorbic acid in wine is an anti-oxidant, (not
so, according to the text books at the time), and the very bad
advice that sulphur dioxide levels in bottled wine should be
minimized. This trend is in the process of being reversed, but it
may take time, as the original proponents are unlikely to recant
publicly. It was only in the early to mid 1990s that “random
oxidation” was invented. I think some research organisations may have given up trying
(unsuccessfully) to prove that corks are a cause of oxidation, but the
malady lingers on. I really think that wine journalists were taken
for a ride on this one.
Two of John's "papers"
certainly opened up my eyes and provided a huge amount of clarifying
information, and are they certainly worth reading. John has been
kind enough to allow reproduction of them here.
and
The Cabernet Sauvignon wasn't at all bad; in fact it was a good
result for the vintage. However what concerned me about this wine
when I opened it was the condition of the cork. The cork
looks like it had not sprung back properly and there was staining on
about 80 percent of the cork, right up to the top and it looked like
some drops had actually leaked through, staining the top of the
cork.
I then opened the Shiraz and it was also suffering from slight
leakage, but unfortunately the bottle was absolutely rank with
cork taint; the sort of level where even inexperienced wine drinkers
would think there was a problem with the wine.
I then opened up the Merlot. This cork also showed signs of
leakage; the bottom half of the cork was almost completely wet
and it had leaked through to the top on one side. It also looked
like it had not sprung back as far as it should have, but it was not
as bad as the Cabernet cork. Unfortunately this wine was not sound
either. The aroma was rank and despite vicious swirling it had not
blown off after a couple of hours. I left the bottle for 24 hours by
which time the stench had dissipated and whilst there was a pretty
good level of fruit flavour, I doubt the bottle was sound. I'm not
sure if it was TCA or some other fault, but it was not a good look.
I look forward to hearing your comments."
Prior to the 2004 vintage we had been paying top dollar for
Reference 1 corks and clearly not getting them. The quality has been
more likely Ref 2 or worse. Of course this was happening at a time
when there was enormous demand for cork and the cork industry was
satisfying the demand with inferior product which it pushed out to
smaller wineries like ourselves. Not surprisingly these wineries got
pissed off and made choices on other closures. There is no doubt
that if I compare the level of cork taint we are getting today with
the cork taint we were getting in say vintage 2001 or 2002, cork
taint today is more frequent.
However there are two issues here and it is my belief that
neither are related. The first is cork taint. The second
is wine seeping down the sides of the cork. It is our
experience and I suppose we are opening about 2 dozen bottles per
week, that most of the corks which have seeped have not seeped the
full way, have not leaked externally and the bottles have not
displayed ullage of any kind. Whilst I agree this is not a great
look we have found the wine to be completely unspoiled and sound.
The bottles which have taint would have had taint anyway. We have
not found a bottle which has leaked externally or where the wine has
become oxidised. It would appear to us that in the bad cases
where wine has seeped a lot, the particular corks in question appear
softer and more spongy than the better corks or have had wine
running down a fault line in the cork. It is also my belief that not
every cork is like this."
Feel free to submit your comments!
From: Rory Shannon
05/18/2007 22:25:37 HHmmm,I agree with the premise that oxidisation can occur in the bottling line, and as far back as the process of turning grape juice into wine.
But two things I amm not convinced about:
1, If there is a problem in the process of sealing the bottles with cork, in my experience, it is usually gone through at least a number of dozen, more likely 10 to 20 dozen, before it is noticed. Yes, it the process SHOULD be kept an eye on, but I have seen it happen too many times that as humans, we make mistakes. God knows the bottling operations are usually under enormous pressure to finish one contract to move on to another. So with this premise, "random" oxidisation would cover more than just one bottle in your dozen.
And 2, excuse my rudimentary knowledge of science, but if a cork is in fact leaking (I have had a few of these in my cellar) is not air getting in? You know, the "for every action, there is an equal and opposite reaction" thing. "Transfer" I think it also known as.
Look forward to some thoughts.
........................................
John Casey Responds:
'Random' events are usually the result of the interplay of a number of factors, and this can make them hard to anticipate or prevent. Just a little bit of air in the delivery line now and again, an erratic filling head on a 24 or >>48 head filling machine, a malfunction of one head on a multi-head corking machine at anything from 50 to 400 bottles per minute. If and when these events coincide, there is going to be a bottle with an unacceptable amount of air/oxygen. So-called 'random oxidation' peaked in the 1990s after wine-bottlers were urged to minimise SO2 levels in bottled wines, thereby reducing the safety margin, particularly for bottled white wines. In 1997-98 a number of AWRI publications stated that 'Reports of oxidation of commercial bottled wines . . . have increased in recent years', that the oxidation is variable and 'may affect 1% to 20% of bottles from a bottling run at any given point in time.', and the 'value of the wine suffering sporadic oxidative damage is between $20-$100 million.'That is a lot of money, even allowing for the need to justify research funding.
A couple of years ago a prominent wine industry person said that he did not want to take pressure off the cork industry about TCA, but post-bottling oxidation was a greater problem than TCA. However, in a survey of 13,000 bottled wines in the UK some four or five years ago, about 1% of bottled wines, (cork and synthetic stoppers), were identified as being oxidised. There was some doubt about the precise figure because of the similarity of slight oxidation to low levels of TCA. The survey seemed to confirm my own impression that there has been a significant reduction in the last five to ten years as wine-bottlers have come to realise what the real cause is/was.
A freshly inserted cork has a sealing pressure of around 200 kPa, and this declines to about 100 kPa over a year or so. Headspace pressures range from near zero up to several or more hundred kPa, and even more at elevated temperatures. After some wine has been expelled, the pressure drops. A subsequent combination of headspace oxygen being absorbed by the wine, and the wine itself contracting with a reduction in temperature is then likely to reduce the headspace pressure to less than atmospheric pressure. However, when the headspace pressure drops below atmospheric pressure, there is usually a release of CO2 from the wine, and measured bottle pressures are rarely, if ever, less than -10 to -20 kPa. Although the cork cannot resist internal pressures of greater than 100 kPa, it still prevents the inflow of air when the pressure difference is less than, say, 50–70 kPa. It seems likely that enlargement of the headspace by expulsion of wine would increase the rate of diffusion of wine between the closure and the glass.
From: Carl Schneider
05/21/2007 08:50:45 Interesting article and one I shall have to digest a while. One initial comment however is the quote "With diameters of about 0.6 microns, the passage of liquids as such is not possible."Doing a little bit of sterile filtration of liquids in hospital pharmacy in a past life and I want to say we regularly used a 0.22 micron filter for this purpose. I don't believe we vaporise the liquid as it goes through the membrane.
I suspect the key is the length of the filter is a lot less than the length of a cork. So, I suggest that a gap of 0.6 micron diameter is more than sufficient to permit the passage of liquid with a viscosity similar to water but the length of a cork reduces this passage. Of course this may reduce to zero passage. I don't know.
........................................
John Casey Responds:
There is a simple explanation. I was wrong and Carl is right. I inadvertently translated 60 nanometers from the textbook as 0.6 microns instead of 0.06. I am not sure whether it was an arithmetical or typographical error. If I had thought 'bubble-point' or 'forward flow 'of 0.45 micron membranes, (membrane-filtration-speak), I would have realised my mistake. Thanks Carl, mea maxima culpa!
From: GraemeG
05/22/2007 22:34:10 I accept that all sorts of things could go wrong at the bottling machine. But why should they be? Are we expected to believe that modern machines are so much worse than their predecesors? Or that maintenance is not being carried out? I would expect rates of oxidation to be falling with modern machines. Multinational companies trying to run their wineries like food factories and dealing with large supermarket chains will be vigilant about machinery maintenance - certainly more so than ma-and-pa operations - or the state of the art 20 or 30 years ago. I wonder if the performance of cork in recent years is not dissimilar to that of persimmon as once used in gold clubs. Vastly increased international demand resulted in more frequent harvest of the raw material-supplying trees, with a gradually commensurate decline in the quality of produce obtained. (I know they only harvest every 75 years or something, but perhaps 40 years ago they were only harvesting every 100...?)I dont give a flying f*** whether wine is seeping around the cork or through it. If the level in the bottle is lower than it was 10 years ago, its damn well gone somewhere. And I dont think its leached through the glass. Dont even start me on TCA. Some corks have maintained some bottles in magnificent conditions. Far more have been ruined by cork. If so much of this ruined wine is due to shoddy bottling lines, then I expect to see similar rates of failure even after the change to alternative seals. And I dont think I will - I havent so far.
Graeme
........................................
John Casey Responds:
There is an old maxim in the QC business, '99.9% reliability is ten times worse than 99.99%.'
As noted above, I believe that the problem of post-bottling oxidation (PBO) has decreased over the last several years, as has the incidence and extent of so-called 'leakage' and 'wine-travel'. Blaming the corks is like a carpenter complaining about poor and inconsistent quality in nails. You may have had the experience; same hammer, same piece of wood, and yet some of the nails bend, go in crookedly, fall over just before the second blow, take more than the usual number of blows to drive home or cause the hammer to slide off the head of the nail on to your thumb.
Two speakers at the Screwcap Symposium in NZ said that the control of materials and methods for screwcap application was far more demanding than using corks, and one speaker bewailed delays caused by the shortage of skilled maintenance and servicing of capping machines. It seems as though the Ant/Grasshopper ratio in NZ is as low as it is in Australia.
A lowering of the wine level is due to expulsion of wine past the cork by pressure in the bottle and/or by hydraulic pressure forcing the cork to absorb up to several grams of, mostly, water and ethanol vapour. In round terms, a loss of one mL drops the wine level by about 3 mm.
From: Mark Dignam
06/12/2007 19:21:15 Excellent stuff on randox and corks.Now I know why thousands of dollars worth of expensive white Burgundies from numerous producers - 96 to 99 vintages - were prematurely turned into sherry and poured down the sink (I don't like any sherry in the first place).
I hope this article is read by the producers.
From: Chris Taylor
07/30/2007 09:09:35 I'd just like to thank John for sharing his experience and insight on cork. I'd also like to make a few comments on cork myself.Leakage/oxidation issues are as John says due to a breakdown in the seal between cork and glass - not due to the cork per se. Consequently, the quality (smoothness/roundness) of the bore of the bottle is paramount to achieving a good seal. Glass production issues including manufacturers using worn bottle molds and employing poor QA play an important role in the quality of the seal achieved.
One further aspect which has not been mentioned is the recommendation by cork suppliers that bottles remain upright for 15 - 30 minutes post-bottling before laying down. This is recommended in order to allow the cork time to expand within the bottle neck to achieve a good seal. The reality is, this very rarely or never happens as the logistics of bottling operations generally don't allow for bottles to be stood up for this length of time prior to packing and laying down. Failure to allow the cork adequate time to expand can result in wine moving up the side of the cork between the cork and the bottle thereby compromising the seal from the outset. I would suggest that part of the reason we have seen PBO increase is also due to this issue.
I also agree with John regarding his comments on the technical demands of alternative closures, especially screwcaps. The application of this closure places much greater demands on bottling lines and requires much greater technical skill than cork. On multi-head applicators especially, it is not uncommon for variation to be seen in terms of torques and tensions. I have sampled numerous wines sealed under screwcap which have been oxidised due to closure failure!
With respect to cork quality and TCA. Cork is a natural product. Period. You generally get what you pay for. In addition, it is absolutely essential that the winery applies its own stringent QA for cork selection. Winemakers need to understand the statistics and employ appropriate QA screening - something which is time consuming and tiring. Finally, they need to be prepared to send batches of corks back to suppliers if they don't come up to scratch. Yes, cork has issues. But I get a bit annoyed at much of the criticism of cork as I think it is misplaced - complacency, laziness and ignorance has also played a role!
From: James Hartman
08/29/2007 13:24:29 John is quite correct as regards cork sealing abilities but cork defects can play a role in oxidation.Here is an extreme example. I was QA Manager at a custom facility where a client (prior to my arrival)had used colmated corks to bottle his Pinot Noir. For the uninitiated, colmating a cork involves essentially spackling the defects with cork dust and coating with a sealant. Colmated corks are always the worst of the lot containing numerous insect bores, wormholes, pith wood etc.
Three months after bottling 20% of his Pinot Noir was severely oxidized and most of the leakage was through the center of the cork and not the side - primarily do to insect activity in the cork. This is a rare example due to poor judgement from an inexperienced winemaker and should not excuse poor bottling practices. The bottling activity and filled bottles are not static entities.
There are events and forces that can play a part in the end product that is why stringent winemaking, bottling and transportation practices are an absolute must.
Copyright © Ric Einstein 2007
