Hacker News new | past | comments | ask | show | jobs | submit login
Dilution of whisky – the molecular perspective (nature.com)
134 points by Tomte on Aug 17, 2017 | hide | past | favorite | 68 comments



It's disappointing they did not reference the seminal work by the father of the Periodic Table on ethanol/water solutions [1]. This work is often cited as justification for why Russian vodkas are 80 proof.

1. Mendeleev, D.I., O soedinenii spirta s vodoi, sochineniya (Total Collection of Transactions: On the Combinations of Alcohol with Water), Leningrad: ONTI-Khimteoret, 1937, vol. 4.

UPDATE on Mendeleev's work with alcohols: https://www.acs.org/content/acs/en/pressroom/newsreleases/20...

Dale Schaefer and colleagues, who conducted the study, note that vodka has a long-standing reputation as a colorless, tasteless solution of 40 percent pure ethyl alcohol, or ethanol, and 60 percent pure water. All such beverages should have the same faint or undetectable taste. But sales of premium vodka brands have surged in recent years. Schaefer’s group at the University of Cincinnati worked with colleagues from Moscow State University in Russia to find an answer.

They knew that famed Russian chemist Dmitri Mendeleev, noted for work on the Periodic Table of the Elements, made a key observation on alcohol solutions in his 1865 doctoral dissertation. Mendeleev believed that a solution of 40 percent ethanol and 60 percent water would develop peculiar clusters of molecules, called hydrates. That solution became the global standard for vodka, which usually is sold as an 80-proof, or 40 percent alcohol, beverage.


Can you give a brief explanation of the justification? I googled for the paper, and had difficulty finding the original or citations.


Not sure who is right, but there is this: https://en.wikipedia.org/wiki/Dmitri_Mendeleev#Vodka_myth


From the wikipedia page you're citing: However, Mendeleev's dissertation was about alcohol concentrations over 70% and he never wrote anything about vodka.

This is clearly not true. Page 81 (Chapter 4) of Mendeleev's dissertation [1] starts as follows:

На осно­ваніи этого, желая съ возможною точностію опредѣлить составъ той смѣси спирта и воды, которой при разныхъ температурахъ соотвѣтствуетъ наибольшее сжатіе, я изслѣдовалъ смѣси, содержащія отъ 40 до 50 процентовъ безводнаго спирта. Это изслѣдованіе состоитъ изъ двухъ главныхъ частей: приготовленія смѣ­сей и опредѣленія ихъ удѣльнаго вѣса. Опишу первое, второе изложено во второй главѣ.

[1] https://upload.wikimedia.org/wikipedia/commons/3/3f/%D0%9C%D...


where does that say whiskey? google translate provides:

    On the basis of this, wishing, with the utmost precision, to 
    determine the composition of the mixture of alcohol and water, which 
    corresponds to the greatest compression at different temperatures, I 
    used a mixture containing from 40 to 50 percent of anhydrous alcohol. 
    This study consists of two main parts: the preparation of the sms and 
    the definition of their extra vias. I'll describe the first, the 
    second is set out in the second chapter.


Here's the original: https://upload.wikimedia.org/wikipedia/commons/3/3f/%D0%9C%D...

He actually determined that the highest compression of the mixture takes place at roughly 46% alcohol. (p. 93).


>All such beverages should have the same faint or undetectable taste. But sales of premium vodka brands have surged in recent years.

Not sure what the threshold is for "premium" vodka, but anything not relatively high in price very objectively tastes like Satan's armpit


This is mostly seen from the angle of guaiacol, which is relevant for peaty whiskies. Does this hold for less peat-forward whiskies? Could water dilution hurt some other compounds?


It's a good question, but much (in a lot of cases maybe most?) of the flavor of a whiskey comes from the cask, not the mash bill, and guaiacol is a component of the wood flavor as well.


Ummm citation please? I'm not even sure of a measurement that you could cite for "flavor" that would be generally agreeable to experts.

But certainly mash bill vs cask is the biggest debate in whiskey right?

Like I'm a mash bill guy but I'll buy that I'm wrong cause the science is unclear.


Woodford Reserve does a rye thing where they sell a pair of identical mash bill bottles, aged for the same amount of time, one in new oak and one in "used" (ex-rye). The difference is huge; they're not remotely the same whiskeys.

Meanwhile, Four Roses Single Barrel does annual releases with different "recipes", varying in mash bill and yeast strain. If you actually go to Four Roses, you can buy a bottle of every recipe. My brother did that and we did a tasting, and it is very hard to tell them apart.

I can't give you a scientific cite, but from experience: cask is a huge flavor influence. Chuck Cowdery suggests it's why indie distillers have such a hard time turning out good product (they use smaller casks, which age whiskey poorly).

Note though that expecting exactly this kind of response, I said "in a lot of cases". And, of course, I'm not saying you couldn't immediately tell the difference between a non-peated bourbon-aged barley Scotch and a corn whiskey, or the difference between a wheated bourbon and a high-rye bourbon!

My only point is: wood flavors are important to non-peated whiskey. :)


Caveat: I am not a scientist, just a humble lover of a dram. I don't generally go for peaty whiskies, much preferring the Speyside flavour profile and some of them benefit from a couple of drops of water. It differs from whisky to whisky, and age to age. I'll generally try one without, one with, and decide what my favourite is.


Takeaway/tl;dr: "Overall, there is a fine balance between diluting the whisky to taste and diluting the whisky to waste"


I tend to drink it cask-strength, neat. I like the burn.


Now these are the articles I wished I read in Academia!


Whilst still in academia, I surely wasn't interested in diluting my whiskey.


The whisky is diluted as part of the manufacturing process to the optimum(ish) level: 80 proof / 40% ABV. You are correct to not dilute it further.

Ref: Whisky is distilled to around 70% alcohol by volume (vol-%) then diluted to about 40 vol-%, and often drunk after further slight dilution to enhance its taste.


It depends very much on the whiskey! Most whiskeys are somewhere in the 40-48 ABV band (they're definitely not all diluted to exactly 40%), but "cask strength" whiskeys go all the way up to the mid-60s (the idea of cask strength whiskey is that you're supposed to dilute it yourself; it's sold undiluted so that more of your money is going towards ethanol than water).


The 2016 release of George T Stagg came out of the cask at 144 proof, or ~72%. You might want to recalibrate your high-water mark. ;-)


Wow, that's very high. I'm curious about whether the angels' share noticeably alters the ABV, from the time the distillate is put into the cask, to when it's bottled (as I'm assuming different evaporation rates for water & ethanol).

Edit: https://www.bruichladdich.com/article/the-angels-share Is quite interesting:

"Casks stored in humid conditions lose a greater proportion of alcohol than those stored at lower humidity levels. Conditions of low humidity draws more water through the staves - leaving a higher proportion of alcohol in the cask. "


So, in the case of the aforementioned Stagg, the distillate went into the barrels at 125 proof, and came out at 144. Clearly, the angel's share preferred to take water over distillate.

That said, the yield was horrible -- 10-15 gallons of spirit per barrel out of an initial barrel capacity of 55 gallons.

That said, I question whether humidity has as much to do with it as atmospheric pressure does. The benefits of barreling something like bourbon in a place like Kentucky is that they have 4 full seasons, and the variance in atmospheric pressure from cold winter to hot summer creates something of a capillary effect in the distillate -- as the weather gets hotter (and maybe more humid) the distillate pushes into the barrels, but pulls out during the cooler months.

Just putting the barrels in a hot place for its entire aging process might accelerate the aging, or it might mean that because it loses the breathing effect, it might extract more of the tannins and less desirable elements from the barrels.


Cheers those figures are very interesting.

Wow that's a massive loss though, I hadn't realised it could be that much lost.

That's also interesting regarding a potential for accelerated ageing.

I'm also wondering how oxygen uptake is effected by differing levels of spirit in the cask, I'm assuming that might effect the taste somewhat.

Edit: The thesis 'Aging of whiskey spirits in barrels of non-traditional volume' has lots of interesting information.

"As evaporation occurs and the fill level decreases, oak not in contact with liquid spirit dries, contracts, and becomes more porous to the entry of outside air (28). Oxygen present in the barrel fuels oxidation reactions that are essential in the production of mature spirits"


Yes, the ABV raises or lowers depending on temperature and humidity. The Stagg almost certainly went in at 125 proof (barrel entry proof), so it got a big boost during the considerable aging process.

I've heard descriptions (from the producers) of barrels that gained / lost proof depending on what floor of the warehouse they were stored on. Pretty neat.


*Whisky ;-)


Only if he's talking about scotch. Pretty sure it applies to bourbon and other non-scotch whiskeys, too.


This is truly the tabs-versus-spaces pointless debate of drinking culture.


The Martini is the tabs-versus-spaces debate of drinking culture.

(But there is only one right way for that as well)


I thought the Martini was the toilet-paper-roll-orientation debate of drinking culture.


People still use tabs? I thought tabs became obsolete the moment IDEs and text editors supported space-for-tabs with automatic column alignment.


Tabs are semantic. Using spaces for indentation is like using tables for layout.


Yup. I wouldn't want to ruin a good thing. I don't even want ice in it.

On the other hand, I've tried stuff like Everclear. That needs dilution. It's absolutely horrible, by itself.

Good whiskey needs no adulteration.


> Good whiskey needs no adulteration.

Most whiskey is not sold at cask strength so evidently most distillers disagree. On the other hand, the higher quality whiskies are usually the ones that have lower cask strength and higher bottle strength, i.e. less diluted.


Correct. Good whiskey is perfect when you get it. The folks who made it know how to make good whiskey.

I never buy cask strength.


That's a shame. There are some Lagavulin and Laphroaig bottlings that are amazing at cask strength, without any additional dilution.


I haven't found one I really like. I've been enjoying some 18 year old Buchanan. I refuse to spell it without the 'e.' Not that I have a war to win, but spellcheck prefers it and I let spellcheck make many choices for me. It's for the best, really.


Well, except with coke or orange juice.


These types of articles are the reason I come to HN!


Huh, they used two different versions of the Amber software: v10 for molecular dynamics, v15 to calculate spatial distribution functions. Amber: http://ambermd.org/ I wonder why.


I use Amber to do similar calculations. It's distributed as a free set of analysis and preparatory programs called AmberTools, which is the V15 software cited, and a licensed simulation program, which is v10. Interestingly, that is a super old version (usually it's released every other year or so with the number corresponding to the year)


> Spatial distribution functions (SDFs) were calculated using the Grid command implemented in the AmberTools suite of programs (v. 15, USCF, San Francisco, USA)

Likely has to do with improvements to this Grid feature?


I was told about this by some kind strangers at a pub while traveling in Scotland and never looked back. It specially improves the flavoring in very tarry/peaty whisky.


So guaiacol boils at 205°C. How does it end up in the whisky, after distillation? I mean, the water should come over before the guaiacol does.


Guaiacol is present in wood smoke. It's maybe being leached out of the charred barrels into the whiskey during the aging process.


It does mention "It is likely that the concentration of guaiacol in Isley whiskies is even higher.", since Islay whisky often uses peated malt, won't at least some of it be coming from the malt.

As far as I understand, they don't treat the casks differently to other distilleries.


I believe it is present in both the original distillate and also is leached from barrels. Some info here on the leaching from barrels:

AGING OF WHISKEY SPIRITS IN BARRELS OF NON-TRADITIONAL VOLUME

http://etd.lib.msu.edu/islandora/object/etd:1059/datastream/...


Right - it's the barrels that infuse a lot of the flavor.


That makes sense. But the paper seems to claim that the guaiacol comes from the smoked malted barley:

> Many whiskies, especially those that are made on the Scottish island of Isley, have a typical smoky taste that develops when malted barley is smoked on peat fire. Chemically, the smoky flavour is attributed to phenols, and in particular guaiacol, which is much more common in Scottish whiskies than in American or Irish ones.


I agree that this is where the smoky flavours are supposed to come from. So your question remains unanswered. Perhaps guaiacol is generated from intermediate compounds after distillation? From chemical reactions with the barrel?


I wasn't able to find a source in a quick search, but I wonder if (traditionally) the charring of the barrel is also done with peat.


No, it isn't. Scotch whisky can only be aged in barrels that have already been used to age liquor. Conveniently, bourbon must be aged in NEW barrels, so used bourbon barrels are the primary source used for scotch. So, the barrels are sourced for the US and thus not charred with peat.

Other barrels used for scotch are typically fortified wine products, like sherry or port. Occasionally limited releases will be aged in things like ex-rum or ex-brandy casks.


Scottish (and most, although not all, other types of) whisky is produced using a "pot still" rather than a modern column, so you don't get an exact separation by boiling points. I mean, generally you're running the still at under 100C, but you get a substantial fraction of water (~30%) in the output as well.

I suspect that if you plotted the still's throughput as a function of boiling point of the inputs, what you'd get is a fairly broad normal distribution centered on the temperature of the still "head" during the distillation run, with the width of the curve dependent on the design of the still. The upper tail of that curve could easily (must, really) be catching small amounts of substances that boil at 200+C.


OK, thanks. So they use a still with low BP resolution, and so there's appreciable higher BP stuff in the product. But doesn't that also make it harder to exclude methanol and other low BP stuff?


Liquids partially vaporise below their boiling point (eg water vapour in humid air). See http://pubs.acs.org/doi/abs/10.1021/jf60134a022?journalCode=...


There's also the possibility of forming an azeotrope (like with ethanol/water at 96% ethanol). Azeotropes are cool, they're exploited a lot e.g. as working fluids in heat pumps.


Interactions between ethanol, water and guaiacol are a key concept in the article. So there could be an azeotrope, I suppose. There's one for anisole aka methoxybenzene. Its BP is 153.8°C, and the anisole-water azeotrope (40.5% by weight) boils at 95.5°C.[0] Guaiacol is 2-hydroxyanisole, so maybe there's one for it too.

0) Horsley (1947) Table of Azeotropes and Nonazeotropes http://sci-hub.io/10.1021/ac60008a002


Neat, that's very interesting as it does mention it's present in the ferment.


Well, the water DOES come over. Whiskey is distilled to at most 80% AL so you get 20% water with other things coming along for the ride. If you go much higher then 80% you start loosing so much flavor that you might as well be aging vodka.

That said, most of the flavor still comes from the barrel.


Water boils at 100C, so shouldn't the guaiacol come over first? Or am I misunderstanding how distillation works?


Nope, it will come out last.


> on the Scottish island of Isley

The island is actually called Islay. Hopefully the rest of the article is more accurate.


Intriguing. I always assumed it had to do with greater alcohol strengths anesthetizing the taste buds.


I've always spelled it whiskey.


That's the American/Irish spelling. If it's a Scotch it's whisky.


> If it's a Scotch it's whisky.

Also Canadian, Japanese, Australian [0], English [1] and Welsh [2].

[0] https://www.theguardian.com/lifeandstyle/2014/apr/02/tasmani...

[1] https://www.englishwhisky.co.uk/the-whisky

[2] http://penderyn.wales/


Comes from the word uisce, which is the Irish/Gaelic for water. Technically the Irish for whiskey itself though is 'uisce beatha', meaning 'water of life' ;)


For those unable to remember it without a mental aid:

WhiskEy:

IrEland AmErica

Scotland has no E.. :)


This is not 100% true (George Dickel Tennessee Whisky comes to mind first) but it's close enough.


I thought this was going to be an article by Heston Blumenthal...




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: