The Chemistry of Whiskey

"Understanding the delicate interplay between chemistry and aroma could be a huge help to distillers looking to tweak their whiskey to encapsulate that perfect blend of smoky and spicy….It’s research that means that the perfect whiskey–smoky, spicy, or however you want it–might not be so elusive after all."(2)

Inside the Jack Daniels Distillery

This Thanksgiving, as per usual, I'll be making the long overland journey to Nashville to see my brother. Although the visits are, of course, enjoyable the question always arises among my  family: what on earth are we going to do for four days while we're not stuffing our faces? This year, it looks like we're taking a tour of the Jack Daniel's Distillery in Lynchburg, TN.

My immediate reaction to this was, I bet they have a lab. Then, I wonder if they're hiring? Well, like most companies which produce alcoholic beverages, they do indeed have a lab and a fairly extensive one at that.

According to an article in the New York Times the use of high-tech analytical chemistry equipment and techniques wasn't fully implemented at the company until 2002, 136 years after it's founding. Sensory scientists were concerned that not all the key aroma compounds of their whiskey had been identified and that therefore there could be room to improve its taste (if not by changing the formulation then through ensuring the balance of compounds is perfect for each and every batch). This new research into their own product went as far as analyzing the effects of the type of oak used for aging barrels (1).

Whiskey Lactone: β-methyl-γ-octalactone

The basic process for making whiskey starts with fermentation of a blend of corn, rye, malt and barely. That product runs through several feet of sugar maple wood charcoal, is distilled, and allowed to age in an oak barrel (usually for years). Interestingly - well, there's two really interesting things here - the charcoal removes the esters from the whiskey making it less fruity and pungent, and the distillate leaves behind a material with upwards of 500 unknown compounds, some of which might be aroma or flavor chemicals. When this chemical mass was sifted through for potential whiskey additives, scientist's at Jack Daniels "identified 28 critical compounds that seemed to give the most flavor to the bourbon, including beta-damascenone, which give the taste of cooked apples; lactones, which provide coconut flavors and eugenol from the oak barrels, which give clove like flavors" and "when the 28 compounds were added to the alcohol and other compounds, the result was a pretty good bourbon" (1).   

Similarly, a Smithsonian Magazine article divulges the chemical complexity of whiskies in general. A University of California, Davis researcher (Tom Collins) and his team analyzed 60 American whiskies and found over 4,000 different non-volatile compounds in total, many of which overlapped among the brands and types. Because whiskey is made from a variety of grains and because yeast is involved there is an overwhelming number of potential chemicals that can be produced during the process (2). 

The purpose of the research was to answer on a fundamental level the question: why are all bourbons whiskey but not the other way around? The team found a more technical answer than production regulations can supply, "there are 50 to 100 chemical compounds such as fatty acids and tannins that can be used to distinguish a Tennessee whiskey from a bourbon" (2). 

An article from the Royal Society of Chemistry gets into the fine details and complexities of Scotch Whiskey production. The Scotch Whiskey Research Institute in Edinburgh, Scotland delves into the chemistry behind Scotch whiskey as its own class of liquor with its unique flavor and aroma compounds. 

Like most liquors scotch whiskey contains many of the major classes of flavor chemicals: fatty acids, aldehydes, esters, and more. Unlike Tennessee whiskey, the only raw material involved is barely (becomes malt upon fermentation, hence "malt whiskey") so the flavor combinations from starting materials is relatively limited. However, traditional scotch production has some interesting conventions: smoking the aging barrels over a peat-fire, the use of copper stills, and aging in oak barrels for 12 to 16 years. 

The complexities of peat-smoke is practically a field on its own: one study found that peat in different locations around Scotland produced a different chemical profile (4). Like dripping the undistilled brew through sugar maple wood charcoal in Tennessee whiskey production, the firing of the casks produces a layer of active carbon which removes undesirable chemicals. Additionally, lignin from the wood breaks down during ethanolysis which produced aromatic aldehydes and whiskey lactones which are often fruity and even coconut-like (3). 

Peat Site in the Shetland Islands

Like the charcoal, the use of a copper still reduces the amount of sulfur compounds that end up in the final product. Although these chemicals are important to produce meaty flavor, as noted in last week's post, no whiskey purist wants their drink to taste like turkey (3). 

The type of barrel used for aging also has effects on the product. Bourbon producers and some Tennessee whiskey producers like Jack Daniels only use their oak casks once before selling them to other companies (which seems incredibly wasteful to me…). Casks from different alcohol production processes effects the Scotch whiskey. For example, if rum or sherry casks are re-used, the whiskey will be darker and sweeter. More important than the type of cask used, however, are the reactions that occur when the liquor in is contact with the wood. These include the absorption of dimethyl sulfide into the wood, reducing the pungency of the final product and the oxidation of alcohols and aldehydes to form characteristic esters (3). 

Based on the articles I read it seems that the chemistry and complexity of whiskey is something that has only recently been deeply explored. It's left me wondering how much research has been done on other liquors and the various processes that go along with them. Wine, of course, has been extensively studied for some time but the in-depth analysis of flavor and aroma compounds of other alcohols are potentially just as intricate yet not fully understood or even appreciated. 

If you want to know everything ever about Scotch Whiskey production, here's a documentary narrated with an excellent accent. 

And for kicks, here's your holiday guide for whiskey tasting. 

Ashley 

1. Chang, Kenneth. "What's in that Bottle of Jack Daniel's? A Chemistry Mystery." The New York Times (Archives), Nov. 5, 2002. Web. Accessed Nov. 23, 2014.

2. Geiling, Natasha. "How Chemistry Can Explain the Difference Between Bourbon and a Tennessee Whiskey." Smithsonian Magazine Online. Sept. 9, 2013. Web. Accessed Nov. 23, 2014.

3. Gills, Victoria. "A Whiskey Tour." The Royal Society of Chemistry "Chemistry World." December 2008. Web. Accessed Nov. 23, 2014.

4. B Harrison et al, J. Inst. Brew.,  2006, 112, 333  

Thanksgiving: A Food Chemist's Holiday

With Thanksgiving rapidly approaching you're surely interested in knowing the science behind roasting the tastiest turkey. The Maillard reaction is one of a few fundamental chemical processes that relates to cooking. While I've long been aware of this I've never delved deeper than casually noting that it causes the browning of toast and the dark streaks on grilled food. Besides changes in appearance the reaction involves a number of well-documented chemical steps that result not only in the darkening of food but also the formation and release of aroma compounds. Just think about this: does raw meat have the same flavor as grilled? Right. So, that's a very clear example of just how much the Maillard reaction affects what we eat.

Raw fish does NOT taste like cooked fish

Just over a hundred years ago Louis Camille Maillard wondered what could make cooked food turn brown and release carbon dioxide. Although he was unable to discover the specific mechanisms involved he did realize that the reaction produced a class of previously unknown compounds which he called melanoidins, i.e. compounds that are produced through a reaction of sugar(s), amino acid(s), and heat.

Louis Camille Maillard

The reaction itself has now been illustrated in a variety of ways. In The Chemistry and Technology of Flavors and Fragrances Liam O'Hare and John Grigor chose to represent the process as the simplified  scheme proposed by John Hodge in 1953. His schematic goes something like this:

A carbonyl group, such as glucose, and an amino compound, such as lysine or glycine, react when heated together. Condensation occurs to produce a Schiff Base. A glycosylamine forms and either an aldose or ketose results. If an aldose forms the Amadori rearrangement creates an unstable intermediate while when a ketose forms the Heyns rearrangement leads to an unstable intermediate. The next step in either case is a dehydration followed by a deamination leading to a variety of flavor and aroma products like furans and thiazoles.

Hodge Schematic from The Chemistry and Technology of Flavors and Fragrances

In this system, water acts as a mobility-enhacing solvent. Additionally, analogous reactions can occur with thiols. Sulfur-containing compounds, while acrid or foul-smelling in high concentrations, are surprisingly critical to flavor of cooked meat.

I found this video helpful in understanding the reaction: 

And that brings us back to Thanksgiving. Why does only the surface of a roasted turkey turn that gleaming golden-brown color? That's due a different - but also important - reaction: caramelization. When sugars are heated above 150 degrees Celsius, anhydrides form. All meat contains ribose and therefore sugar is available for a reaction to occur. The high temperature and lack of water on the surface of say, a turkey in the oven create ideal conditions for caramelization. The anhydrides break down into furfural or 5-hydroxymethylfurfural and with continued application of heat become furans, aldehydes, ketones, aliphatic and aromatic hydrocarbons. Some furans present in cooked meat are critical to their flavor. For example, bis-2-methyl-3-furanyl is a key "beefy" component and has one of the lowest odor thresholds known being detectable at 0.00002ppb.

White on the inside, browned on the outside: caramelization 

"Flavor moderators" also impact the taste of cooked meat. The degradation of phospholipids leads to fatty aldehyde - key components of "meaty"flavor - formation. One experiment to prove their importance extracted all the lipids from a piece of meat and upon cooking the lipid-less tissue a biscuit-like smell was produced.  A high proportion of fatty aldehydes limits the production of heterocyclic aroma compounds during cooking (i.e. the Maillard Reaction).

There's a balancing act going on at the chemical level: too many high-impact aroma compounds can cause strong and unpleasant sulfurous odors while too few aroma compounds will leave your turkey tasting like the biscuits it's served with.

Ashley



Sources

"100 Years of the Maillard Reaction: Why Our Food Turns Brown," J. Agric. Food Chem. 2013. 61. p.10197. http://pubs.acs.org/doi/pdf/10.1021/jf403107k 

Hodge, John. "Chemistry of browning reactions in model system." J.Agric. Food Chem. 1953, 1, 928−943.

Rowe, David J (ed.) (2005) Chemistry and Technology of Flavors and Fragrances. Blackwell Publishing: Poole. Ch.3.

News Review: Pumpkin Spice

"Retail sales of pumpkin offerings have experienced double-digit growth for the past several years, reaching nearly $350 million in 2013, according to market research firm Nielsen. And that figure doesn't count Starbucks pumpkin spice lattes…." 

Carmen Drahl, C&EN 

I admit to hoarding Chemical and Engineering News (C&EN) articles that mention flavors or fragrances. Or food additives. Or food regulations. Pretty much anything that mentions chemicals and things consumers consume. In addition, I recently discovered - to my joy - a section of their website called "What's That Stuff" (which is also the header of certain articles in the magazine).

I'm far from the first and surely far from the last to bring this topic up, but the most recently featured stuff was the (in)famous "pumpkin spice flavor" in the October 27 issue. According to this article the complexities of natural flavorings don't necessarily need to be present for the human brain to understand what a flavor is meant to be. "Pumpkin spice" is considered to be a mixture of spices containing around 340 flavor-relevant compounds. However, given just a fraction of those compounds and concentrating on the inclusion of key aroma and flavor molecules - about 15-30 components of the spice mixture's complete profile - the overall taste is interpreted as true-to-nature by the consumer.

I wrote a blog post on another website on Natural and Artificial Flavors where I basically concluded that both have their pros and cons but that artificial flavors aren't any "scarier" their "natural" counterparts. That said, there's a lot more that goes into making the perfect product that isn't flavor related. Take, for example, that famous Starbucks Pumpkin Spice Latte.

Is Your Coffee Toxic?

Sugar, preservatives, coloring. There's probably even some sensation-generation molecules in the mix. Things to make warm and cozy feelings all over your mouth. The above info-graphic might be exaggerating a bit about the side effects but then again….

But I digress.

I wanted to tell you about the main chemicals involved in pumpkin pie spice according to this article in C&EN. Zingiberene (ginger), (E)-cinnamaldehyde (cinnamon), sabinene (nutmeg), eugenol (clove) and possibly cyclotene (maple or brown sugar flavor) and - everyone's favorite -  vanillin. Other, smaller contributors to the wildly popular pumpkin spice flavor include allspice, anise, and mace. Find their mug shots below….(thanks to Wikimedia Commons for having awesome drawings of chemicals and pictures of foodstuffs). In addition to the images, definitely check out this resource from NBC to learn more about the molecules in these spices.

Zingiberene

(E)-cinnamaldehyde

Sabinene

Eugenol

Cyclotene

Vanillin

Just as the pictures I've paired with the chemical names show very different things than what the chemicals themselves are, key flavor compounds are representations of the full flavors found in nature. Like the images these critical components of a flavor are enough to give the brain context and therefore understanding for something that is as poor a representation of the true thing as still, flat drawings of an-ever-moving molecule. Pretty tricky on the food industry's part, don't you think?

All of this begs the question: what aroma compounds are actually in a pumpkin? As it turns out real pumpkins, like, the kind from a vine in a patch contain a very different set of aroma compounds. If you've ever taken a bite out of a pumpkin at Halloween (anyone?) you'll recognize that they aren't sweet and have a subdued not-all-that-appealing flavor.

According to this article in Scientific American, pumpkins smell like vegetables or squash (not surprisingly) because of "leaf alcohol" or cis-3-hexenol. Other major flavor and fragrance compounds include 2-hexenal, diacetyl, 2-meythlbutanal, furfural and pyridine.

Credit: "Pumpkin, Hold the Spice" Scientific American
Clockwise, from top left: cis-3-hexenol, diacetyl, 2-methylbutanal, furfural, pyridine

This is a good chance to look at some of these totally natural molecules and get some perspective. Things created in labs aren't necessarily more harmful than things created by plants. I mean, take water hemlock, for example (my go-to for putting nature on the spot).

This isn't to say that all chemicals made in labs are fine and dandy. In fact, there are loads of problems with chemical formulation and manufacturing due to waste, toxic materials usage, and more. But both sides of the natural/artificial arguments have their pros and cons. Take diacetyl, one of the chemicals found in pumpkins. It's responsible for the buttery aroma of pumpkins when it occurs naturally and is harmless. However, in its pure form - or in large quantities on microwave popcorn - it poses a risk of severe lung disease.  Pyridine is another chemical mentioned above and it's commonly used as a solvent in the lab which is certainly not something you'd want to ingest.

So, in conclusion, while pumpkin spice flavors might not contain any real pumpkin they do contain some of the key aroma compounds from the spices they're supposed to represent. And while this might be an unsatisfying answer for consumers the chemicals used are simply mimics of things found in nature. In short, I'll continue to eat my pumpkin spice oatmeal, thanks very much.

Ashley