American whiskeys differ from their Scottish counterparts in a very distinctive way: they leave an unusual web-like pattern as droplets dry, and those webs are different for different brands, becoming a kind of “fingerprint”. That’s a trait that could not only be used to tell the difference between brands of American whiskeys, but could one day lead to an effective counterfeit identification method, according to a new paper published in ACS Nano.
As we reported last year, Stuart Williams, a mechanical engineering professor at the University of Louisville in Kentucky, noticed one day that if he dilutes a drop of bourbon and vaporizes it under carefully controlled conditions, it produces what he calls a “whiskey.” forms. web”: thin strands forming various grid-like patterns, akin to networks of blood vessels. Intrigued, he decided to investigate further with different whiskeys, plus a bottle of Glenlivet Scotch whiskey for comparison. It was the perfect sabbatical project to explore colloids (suspended particles in a medium such as Jell-O, whipped cream, wine and whiskey) at North Carolina State University.
Essentially, it’s the same underlying mechanism as the ‘coffee ring effect’, when a single liquid evaporates and the solids dissolved in the liquid (such as coffee grounds) form a telltale ring. It happens because the evaporation happens faster at the edge than in the center. Any remaining liquid flows to the brim to fill in the gaps and drag those solids along. Mixing solvents (water or alcohol) reduces the effect, as long as the droplets are very small. Large drops produce more uniform spots.
In those earlier experiments, Williams tested 66 American whiskeys; only one has not made a web. Whiskey webs seem to be related to alcohol content. Williams and his team had to dilute the whiskeys with water to about 40-50 percent proof. Specifically, they found that if the alcohol-by-volume level were above 30 percent, there would only be a uniform film; lower than 10 percent, and you get the coffee ring pattern. It’s only at an average alcohol-by-volume level of between 20 and 25 percent that you get these unique web structures.
The next step was to try and identify the specific chemicals responsible to see if the cartridges could be used as fingerprints for specific brands. For this latest set of experiments, his lab tested 66 American whiskeys (56 bourbons, 10 others), 13 other whiskeys and six unaged whiskeys, all subjected to the same test conditions: 25 percent alcohol by volume and a 1-millimeter drop. “To date, we’ve tested nearly 100 samples and haven’t found any non-American whiskey to form a web under these conditions,” Williams said.
The types of solids present in the whiskeys also contribute to the webby pattern left behind. Whiskeys aged in new charred barrels, for example, tend to contain more of those solids, especially those that dissolve easily in ethanol. “We believe that the increased solids extracted from American whiskeys are responsible for these patterns,” Williams said. “The chemicals originate in the fermentation and distillation, but undergo really dramatic changes during maturation.”
Next, Williams performed an aggregated pattern ID test to confirm that the distinctive web patterns were repeatable across brands. He used three brands of whiskey from the same distillery in the experiment: Old Weller Antique, WL Weller 12 Year and WL Weller Special Reserve. He and his team tested 30 drops of each (90 drops in all, all about 2 millimeters in diameter, or “the size of a very large period”) and took digital photos of each.
They then chose 25 drops of the same whiskey at random and averaged the results to create a representative “library” measurement for that whiskey for future comparison purposes, according to Williams. “We compared the remaining drops to this whiskey library,” he said. “A whiskey was successfully matched with its brand 90 percent of the time.”
For another experiment, Williams added a “control whiskey” containing chemicals such as vanillin, tannin or lignin (all flavorings) and found that there was a marked change in the resulting whiskey web pattern. “This provides evidence that the chemicals associated with taste affect the final structures,” he said. “In other words, each whiskey nods differently because the chemicals that make up their monolayer are different. So the resulting web structure reflects its contents.”
Williams was also able to produce the web pattern for non-American whiskeys by sprinkling the water-ethanol liquid with fatty acids and running the test on a highly water-repellent surface so that the drop pearled very well.
The real question is whether it is possible to tell the difference between old and young whiskey (or good and bad) simply by vaping a distilled drop under the right environmental conditions. In this way, the technique could be used for quality control in distilleries or for detecting counterfeits. The preliminary results were promising, but whiskey chemistry is incredibly complex and contains thousands of chemicals.
“Unfortunately, we cannot determine the composition of a whiskey based on this analysis,” Williams said. “But we can use this technique to determine whether it is an American whiskey or not, that is, whether it forms a web or not.” More statistical analyzes are needed, with the aim of standardizing test procedures, before the technique will be useful in distinguishing between different American whiskeys. And while it won’t replace more advanced chemical techniques, such as chromatography, it can be useful as a simple pre-analysis test to determine whether it’s worth doing a more expensive and time-consuming chemical test.
Since Williams, like almost everyone these days, works largely from home, he’s used that time to work on developing a smartphone-based visualization platform so anyone can create and visualize their own whiskey web at home, using little more than a cheap lens (which costs about $7 on Amazon) or a USB microscope (which costs about $20). He expects to post instructions on his research website (which also includes a gallery of whiskey web images) in time for this weekend, just in case the cabin fever hits you and you’re looking for a fun DIY project.
That said, “The best images are taken with a standard light microscope,” Williams said. “The light has to come in from the side — the illuminated structures in these images reflect the light. You may have noticed that the image is mostly black and the features are white. These features reflect the light toward the camera.”
DOI: ACS Nano, 2020. 10.1021/acsnano.9b08984 (About DOIs).