Sunday, December 11, 2011

Science of Wood Maturation

This has been a long time coming, I'll admit. And it's not that I've been putting it off, per se. It's just that it's a huge topic to broach. The science behind wood maturation is the last bastion of scientific advancement, the last frontier, of distilling science research. Even to this point, there are men and women far my superior that still don't know how reactions take place within the barrel. Me, I'm just going by what I've learned from school, good old fashioned hands-on work, and some linking of things. But, since it has been so heavily requested, I present to you:

The In With Bacchus Guide On: The Science of Wood Maturation

Wood maturation has been done, on purpose, for hundreds of years to improve taste and flavor of spirit. Realistically, probably at about the mid to late 1600s. Aging spirit really began as a happy coincidence. In the early days, most places and people preferred their spirit straight off the still. The Glenlivet that King George IV drank in Edinburgh (at that time an illicit liquor made in the hills and dales of Scotland) probably wasn't aged. At that point, distilling wasn't entirely about flavor and, hell, not even really about getting loaded. It was about not wasting things. It was easier to transport porcelain/clay jugs of new make than the equivalent acreage of barley to market. The container didn't really mean much, it was just a means of transportation. With the inception of long distance shipping (read: wooden ships) and the need for more sturdy (read: wooden) containers, barrels came into usage. Even then it was a matter of shipping and storage rather than flavor. But once barrels that had sat, slowly rocking below the decks of a ship as it sailed about Europe, people began to realize: "Wow, this tastes better." At that point, putting spirit into casks was a matter of not just necessity but of choice. Not only did it hold up under duress and travel much better but you got a tastier product out of it to boot. Everybody wins! But even though we've been intentionally putting spirit into barrels since, as said, about the dawn of the 18th century...we still didn't know WHY it was doing what it was doing. It wasn't really until the last hundred years and the advancement of analytical methodology have we discovered what happens in a cask. And we still only have an overview; a lot of the chemistry is a mystery.

What we HAVE discovered is that aging spirits can be broken down into three categories, all ending with "-tion": extraction, reaction, and interaction. Let's get to it, shall we?

Extraction: This has to deal with the wood in a two-fold manner. To start, let's go back to the major basics and start with cellular level chemical make-ups. For us humans, our cells have what's known as a cell membrane, shown below:

Courtesy of Wikipedia
 Ignore the stuff on the top, what we want is the thing in the center. Technically the photo isn't an ACTUAL cell membrane (it's missing some parts) but it's the important part of the cell membrane: the phospholipid bilayer. It's made up of phospholipids, or long chain fats with a phosphorous atom on the end. The phosphorous is very hydrophilic, it LOOOVES water because of it and water are polar ions. The fats are hydrophobic; they HAAAAATE water because they're non-polar and water's polar. So they arrange so that the phosphorous is on the outside, near the water and the fats are on the inside. As you can see, the non-polar fats stick together on the inside (since there's water on the inside and outside of the cell) so it forms this dual layered membrane. Keep this in mind. It's
 important later.

Plant cells are different. Plant cells have a cell wall, shown below:

Courtesy of the University of Georgia.
It's made up of different stuff. No fats for these babies. What gives plants rigidity and strength is their cell wall, made up of a composition of long-chain sugars. As you can see in the above photo, the cell wall is made up of three things: cellulose, hemicellulose, and lignin. These are where we want to focus our attention.

Trees are plants so the rigidity of a barrel comes from its cell walls. Cell walls are comprised of about 45% cellulose, 30% lignin, 15% hemicellulose, and 10% other extractables. Toasting and charring the wood break down that cellulose and hemicellulose by severing bonds in them and breaking them down into smaller chain sugars that will dissolve into the spirit with time. The most quickly taken up are arabinose and glucose with fructose and xylose trailing in terms of uptake time in 55% spirit. Some of these will also breakdown into furan products (like furfural, which gives whiskey a nutty flavor). What primarily adds flavor, however, is the lignin. Lignin will break down (both by heating and by ethanolysis) into aromatic aldehydes (syringaldehyde, sinapaldehyde, coniferaldehyde, vanillin) and their counterpart acids (syringic acid, synapic acid, ferulic acid, vanillic acid). Depending on the level of toasting/charring, phenolics will be derived from the lignin as well and taken up by the ethanol. These are your guaiacyl and syringyl phenols that will give toasty, smokey flavors.

Reaction:  The reaction phase deals with evaporation and chemical reactions but this is where things kinda get weak. We know that there's a lot of chemical processes going on in the barrel but we don't really know all of it (or at least I don't). But what I do know is this. For evaporation, there is both evaporation of positive and negative attributed chemicals. For the negative aspect, the release of the polymethyl sulfides (dimethylsulfide, trimethylsulfide) that is commonly attributed to a “sulfur” smell and taste in new make evaporates. However, here is evaporation of “good” chemicals as well such as acetaldehyde, ethyl hexanoate, and acetic acid (however, the acetaladehyde / acetic acid levels are generally in equilibrium throughout maturation, meaning that the overall level upon disgorging the cask is similar to that at the very beginning). Within the reaction phase is also the aforementioned oxidation of components. Two of the key oxidation / acetal formation reactions within the cask is the transformation of acetaldehyde and acetic acid from the ethanol within the spirit. Also, the formation of dimethyl sulfoxide from dimethylsulfide (which, once again, limits the sulfur content of the final spirit). There is also esterification reactions within the barrel, such as the formation of ethyl acetate from the previously mentioned acetic acid (it can also be extracted from the wood itself as opposed to reaction with ethanol). There is also the reaction of ethanol with the aromatic aldehydes. The presence of many hydroxyl (OH) groups, afforded by the ethanol, will cause breakdowns in the aldehydes to their constituent acids and even further down the reaction chain. An example would be coniferaldehyde. In the presence of ethanol, it changes to vanillin, then vanillic acid and then to ethyl vanillate. Or sinapaldehyde, perhaps. It will change to syringaldehyde, to syringic acid, to ethyl syringate. Thus, the longer you keep it in wood, the more of these "deeper", or further progressed down the chemical reaction chain, products you get.

Interaction: The last stage is interaction. Interaction comes in two forms: pH based interactions and ethanol / water interactions. While the reactions stated above seem to be numerous and consequential, the reality is that the concentrations of the volatiles don't really change too much during maturation. Yep, dead serious. However, fluctuations in pH cause changes in the ionization states of weak bases within the spirit which affects their volatility. By changing the pKa of the solution within the cask (either by addition of acidic / basic components from the wood or the evaporation / concentration of the solution itself), the evaporative losses of some volatiles may be greatly increased. Then there's the ethanol / water interaction thing. This is where me explaining the cell membrane comes in handy. Truth be told, if you pick up a bottle of vodka, you see that it's perfectly clear and you'd probably say that the ethanol is evenly distributed within it. That if you were to pour a glass it would have as much ethanol in it as the next glass and the next glass.

You'd be wrong.

Ethanol and water are a funny pairing. They're both "polar" so they should dissolve evenly within each other. But the structure of ethanol keeps that from happening. Let's take a peek, shall we? Here's ethanol:

Courtesy of Wikipedia
Take a peek at that. You have your polar hydroxyl group on the right (the O-H, or OH group). To your left, you have your non-polar ethane group. Sounds like your phospholipids, huh? Well, you'd be right. Ethanol in concentrations above 20%, is heterogeneously distributed through the solution (unevenly distributed). What they will tend to do is form ethanol clusters. A bunch of ethanol molecules will bundle up such that their OH groups are sticking outward (the polar OH group preferring the polar water) and the ethane group sticking inward. So, above 20% ABV, this happens. And not all at once, either. This is why older spirits can taste "smoother". As time passes, these ethanol clusters become more and more compact, making the solution more and more heterogeneous. New spirits will have more of a "burning" taste because the ethanol hasn't had the time to cluster as effectively. This is also why watering down a spirit (with a mixer or whatever) makes it easier to drink. Below 20% ABV, the ethanol evenly distributes into a homogeneous mixture so you won't get random clumps of pure ethanol. When you taste it, it's like drinking a shot of 200 proof alcohol and pure water at the same time. When drinking a heterogeneous mixture, it's like taking shots of 200 proof alcohol, then a shot of water and repeating this at infinitesimally small time periods. Significantly more burning on the latter.

So there you have it. The In With Bacchus guide to the science of wood maturation. Bear in mind that I could be wrong. Distinctly possible, in all likelihood. And that this is just an overview. As said before, maturation is the final frontier of distilling science so there's constantly papers being put out about it. But only nerdy people like me constantly seek them out.


16 comments:

  1. Nice. Thanks for writing this up. Maybe the next installment could focus on applying this knowledge practically.

    ReplyDelete
  2. I wouldn't call this the final frontier. I'd call it the next frontier. There's so much more that hasn't been touched, like artificial atmosphere...

    ReplyDelete
  3. I'd honestly say that artificial atmosphere is just an extension of this. Investigations into artificial atmosphere are investigations into the effect of higher or lower temperature/humidity/elevation/what have you on wood maturation. We've covered almost all the science of yeast fermentation and byproducts, malt biology and enzyme interaction, and atmospheric effects of whisk(e)y post-bottling. We've kinda run out of things to investigate but, luckily enough, what we have left can go deeper than the Mariana Trench.

    ReplyDelete
  4. A few comments from the biochemist here:

    1) cell membranes are made up of more than phospholipids. Often the lipids have a hydrophilic but uncharged end. Some are charged. If all the membrane's lipids were charged, this would cause massive electrostatic issues such as in acquiring food.
    2) plants have both a cell wall and a cell membrane. So they do have the fatty stuff. Same with yeast which have a cell wall and a cell membrane.

    ReplyDelete
  5. "This is why older spirits can taste "smoother". As time passes, these ethanol clusters become more and more compact, making the solution more and more heterogeneous."

    Do you meant the solution get more "homogeneous" as time passes (smaller cluster = less chunk of ethanol = smoother taste)? If so, I assume these cluster tend to "un-bunch" with time. Why is that?

    ReplyDelete
  6. Fred: Yeah, I'm not the greatest biologist. But I can design the hell out of a still, so that's all that matters! Fair points, all, though.

    Anonymous: It is actually heterogeneous. If it was homogeneous, every volume of liquid you poured (down to the smallest of metric units) would have exactly the same amount of ethanol as all the others. When the ethanol clusters, this doesn't happen. Some volumes will have more ethanol, some less. These tighter ethanol groupings have less of an effect on the taste buds (I think) as they are currently interacting with themselves instead of all reacting chemically on the tongue.

    ReplyDelete
  7. Hi Scott,

    I am the "Ananymous" above. Thanks for the reply. I am still a little bit confused, so please help correct my understandings:

    1. Spirits >20% ABV is heterogeneous; <20% ABV is homogeneous.

    2. Homogeneous spirits tastes smoother than the heterogeneous one because your taste buds are not bombarded by clusters of highly concentrated ethanol.

    3. When >20% ABV, ethanol form clusters. These clusters became tighter as it ages. Does that make the cluster smaller? Why does "tighter" cluster making the solution even more heterogeneous? Doesn't it make the solution a bit "more" homogeneous (more uniform size clusters?) and thus, taste smoother?

    Thanks and really appreciate the article!

    ReplyDelete
  8. 1. Correct.

    2. Correct.

    3. As these clusters tighten, it becomes harder and harder to separate them. The intermolecular forces between the hydrophobic ethane group become greater and greater as time goes on. It's like the difference between splitting apart an orange vs. a bowling ball. So it makes it more heterogeneous because if you were to take a liter of this aged spirit and start drawing off picoliter samples (.000000000001 liter), you wouldn't get an even distribution of those ethanol clusters. Why? Because you can't break them apart. If you had a combined amount of 3 ethanol clusters in two samples, it wouldn't be 1.5 clusters per sample, it would be more like 2 clusters in one sample and 1 cluster in another because the chemical bonds in the cluster prevent it from being split between the two samples. This is why it becomes more heterogeneous: the clusters aren't free to cleave and homogeneously distribute.

    ReplyDelete
  9. Technical post; I'm in admiration of your scholarly efforts! Extremely interesting. I agree with first commentor that this could be the start of a great series.

    ReplyDelete
  10. Awesome write-up, Scott! Thank you for the insight!!!

    ReplyDelete
  11. Just came across this fun post. I'm curious about the clustering of the ethanol molecules. You say that this takes time - is the wood necessary for the clustering process or does it continue to happen even after it's been bottled?

    ReplyDelete
    Replies
    1. I'm not entirely sure but I've got a rough idea. Here's what I think:

      Wood is not necessary...but it speeds things up quite a bit. The addition of the other chemicals from wood will react with the ethanol and water and also catalyze the ethanol-water clustering by providing an even more polar solution.

      Also, there will be interactions between the ethanol (ethanolysis) of the chemicals that will utilize the ethanol. Oxygen passing the barrel barrier will also catalyze reactions that will increase the likelihood of clustering.

      Yet again, I'm not entirely sure. The idea of ethanol clustering is still being tested in laboratories and papers are coming out. It's kind of new-ish (the last 10 years is pretty new in the science world).

      Delete
    2. Interesting, exciting stuff. Thanks for the reply.

      Delete
  12. Very nicely done, Scott. You mention how the addition of water makes it "easier" to drink the booze. Could you offer any insight into why the addition of a little water to a glass of whiskey—most especially whiskey that has not been chill-filtered—not only makes it "easier" to drink but *also* releases additional aromas and flavors from the whiskey? Could I guess that the additional water is causing the clustered ethanol molecules to partially break apart and form new bonds? My name is Jay, thanks—

    ReplyDelete
    Replies
    1. Jay,

      Bingo! As said, the closer to 20% ABV it gets, the weaker the ethanol clusters get. They're not reacting in any way with each other, they're just kinda...huddling together for support. The more water you add, the weaker they huddle together until about 20% ABV in which it becomes a fully homogenous mixture.

      As for the aromas and flavors, I BELIEVE they cluster as well. fatty acids and esters are polar just like ethanol and will behave as such. When you add water, you break them up too. At least that's what I THINK. I haven't seen any papers on what exactly they do but I'll keep looking.

      Delete
    2. Thanks for the feedback, Scott, much appreciated. Perhaps the fatty acids and esters cluster together, but less firmly than the ethanol and are more "easily" encouraged by water additions to pull apart. Certainly it does not take anything like 20% ABV to "loosen" up the aroma; this happens with even a tiny splash of water.

      Delete