Analysis of Whiskeys by GC-FID
Daniel Iversen, R&D Chemist, Lucidity
Introduction
While spending time with my family over the holiday season, I was given a bottle of whiskey by my father-in-law, who boasted that it was a very fine bottle of whiskey and he paid top dollar for it. This got me wondering if there was a way to tell analytically if there was a difference between top shelf whiskey and lower quality whiskey.
Experimental
Between all of us at Lucidity we had 5 different whiskeys to test. A small sample of each of the whiskeys was brought into the lab. The procedure to test the whiskeys isn’t too difficult as the main parts are ethanol and water. Due to the high water content and not wanting to do anything to change the profile of the whiskey, I decided to run the whiskey neat. In order to run the whiskey including the water, I used an MXT-Wax column.
I just wanted to see if there was a chemical difference so no standard was run. I was, however, able to find several studies that showed the composition of the chromatograms and my chromatogram matched almost exactly in retention times. I decided to use these chromatograms as a reference for my study.
| Lucidity MiniGC Conditions | ||
| Carrier Gas | Helium | |
| Control | Pressure | |
| Flow | 1 mL/min | |
| Split | 10:1 | |
| Column | MXT-Wax | 30 m x 0.25 mm, 0.25 μm |
| Injector | 250 °C | |
| FID | 300 °C | |
| Oven Program | ||
| Rate | Temperature | Time |
| 150 °C | 0 min | |
| 20 °C/min | 250 °C | 0 min |
| 2 °C/min | 280 °C | 2 min |
Results
There were 8 main peaks for each of the whiskeys. They were as follows:
- Ethyl acetate
- Methanol
- Ethanol
- 1-Propanol
- Isobutyl alcohol
- 3-Methyl-1-butanol
- Acetic acid
- Phenyl ethyl alcohol
Each of these compounds is responsible for some of the unique taste and smells of these whiskeys. Each of these whiskeys had a different concentration of these compounds. The lower priced whiskeys had low concentrations of the ethyl acetate, isobutyl alcohol and phenyl ethyl alcohol and higher concentrations of acetic acid, shown in Figure 2.
The acetic acid will give a vinegar flavor and smell to the whiskey. The ethyl acetate will give a very sweet smell. The isobutyl alcohol will give a sweet, musty smell and the Phenyl ethyl alcohol will give a floral scent.
Figure 1. Overlay of all 5 whiskeys showing the consistency of the chromatograms.
Figure 2. Chromatogram of the lowest priced whiskey showing a lack of ethyl acetate and phenyl ethyl alcohol.
Figure 3 shows a more expensive whiskey with higher concentrations in ethyl acetate, isobutyl alcohol and Phenol ethyl alcohol. Presumably these three compounds are important to the scent and taste and overall quality of the whiskey.
Figure 3. Chromatogram of a high-end whiskey
| Whiskey | Ethyl Acetate | Methanol | Ethanol | 1-Propanol | Isobutyl Alcohol | 3-Methyl-1-butanol | Acetic Acid | Phenyl Ethyl Alcohol |
| Sweet smell | Faintly sweet, pungent odor | Alcohol content | Alcohol-like, sweet, and pleasant odor | Sweet, musty smell | Apple brandy aroma, spicy flavor | Vinegar | Floral scent | |
| #1 | 1 | 6 | 37201 | 15 | 10 | 268 | 17 | 11 |
| #2 | 7 | 9 | 70003 | 18 | 72 | 541 | 18 | 18 |
| #3 | 6 | 6 | 51909 | 31 | 171 | 556 | 32 | 6 |
| #4 | 3 | 15 | 44547 | 12 | 16 | 238 | 16 | 4 |
| #5 | 9 | 11 | 50283 | 37 | 57 | 433 | 14 | 7 |
Figure 4. Peak areas of constituent compounds for each whiskey
| Whisk(e)y | Ethanol Peak Area | Labelled Alcohol Content (%) | Peak Area / Content Ratio |
| #1 | 37201 | 40 | 930 |
| #2 | 70003 | 64 | 1094 |
| #3 | 51909 | 50 | 1038 |
| #4 | 44547 | 44 | 1012 |
| #5 | 50283 | 47 | 1070 |
Figure 5. Peak areas of ethanol compared to labelled alcohol content for each whiskey
Conclusion
Next, we conducted a single blind taste test around the office. The participants were only allowed to see the chromatograms of each of the whiskeys labeled 1-5 and the actual whiskeys labeled A-E. They were asked to sample the whiskey by sipping and match the letter with the chromatogram. They were also informed of the compounds each of the peaks represented and what each of these peaks contributed to the whiskey.
Being novice whiskey drinkers and not used to drinking whiskey for different flavors, most participants were not able to tell the difference between all of the different whiskeys. One was able to differentiate between the two most expensive on the scale but was not able to tell the lower two apart.
Everyone in the experiment was able to tell the difference between a blended whiskey and a small batch whiskey as the blended whiskey was very even tasting across the board, which is the idea behind blending the whiskey.
The labelled alcohol content of the whiskeys correlates fairly nicely with the area of the Ethanol peak for each whiskey. This was a quick study so we didn’t use internal standards or inject different dilutions of each whiskey. We simply injected each whiskey into the miniGC and reported the peak areas of each of the peaks of interest. But, from a cursory study the peak areas of the Ethanol seem to do a reasonable job of representing the alcohol content of each whiskey as can be seen in Figure 5.
Although it was difficult for our novice group to discern between the different whiskeys, the group was able to agree that one whiskey in particular was much harsher and 2 of the group tasted much sweeter. The harshest was Whiskey #2, the one with the highest alcohol content.
The two whiskeys that tasted sweeter were #3 and #5. These two show higher amounts of 1-Propanol (Peak 4), Isobutyl Alcohol (Peak 5), and 3-Methyl-1-butanol (Peak 6), all known for sweetness. The 1-Propanol peak was uniquely high in these two whiskeys. Peaks 5 and 6 were also high in Whiskey #2, which was noticed to be uniquely harsh, so possibly the high Ethanol content of this whiskey overwhelmed some of the sweeter compounds.
Another point worth mentioning is that it did seem that we could see a correlation between the price of the whiskey and the number of unidentified peaks eluting later in the run around Peaks 7 and 8. In the lower priced whiskeys we noticed quite a few different compounds in varying quantities, but in the more expensive whiskeys, we didn’t see as many or as much of these compounds, as they produced “cleaner” chromatograms. (See Figures 2 and 3).
All in all, it was a fun experiment, and we confirmed that none of us know much about whiskey, but that by using GC-FID you can profile a whiskey and learn some interesting things about its makeup.
More results from the miniGC can be found here: https://luciditysystems.com/products/minigc/minigc-results/
More information on the miniGC can be found here: https://luciditysystems.com/products/minigc/
To see the miniGC in operation go here: https://luciditysystems.com/products/minigc/minigc-interface/