Some compounds supporting coffee’s aroma can evoke animal-like odors when smelled alone. Skatole is one example. A Japanese study published in 1998 investigated why a coffee drink’s aroma changes during heat sterilization, focusing on decreases in five trace compounds, including skatole.
More important than the detection of skatole itself was the finding that changes in abundant compounds did not fully explain the aroma change. Combining instrumental analysis with the human nose revealed compounds that can matter to aroma even in very small amounts.
- Heating coffee with sugar and milk at 123°C for 20 minutes
- The largest peaks did not explain the aroma difference
- The five declining compounds had coffee-like, potato-like, and animalic notes
- Testing heat responses in aqueous solutions of individual compounds
- Not just disappearing, but changing into other compounds
- Summary
Heating coffee with sugar and milk at 123°C for 20 minutes
The researchers paper-filtered an infusion made with 58 grams of roasted Arabica beans and hot water, then added 58 grams of sugar, 100 grams of milk, an emulsifier, sodium bicarbonate, and other ingredients to make one liter of beverage.
They sealed 190-milliliter portions in cans, heated some at 123°C for 20 minutes, and immediately cooled them. These were compared with unheated beverage. The conditions represented sterilization of canned coffee containing sugar and milk, not reheating a cup of coffee at home.
The heated drink and its aroma concentrate were reported to have a less roasted, coffee-bean-like character. However, the paper does not specify the number of assessors or statistical tests for this sensory comparison.
The largest peaks did not explain the aroma difference
The researchers first compared samples using gas chromatography, which separates volatile compounds and allows changes in their amounts to be examined. Some compounds, including furfuryl acetate, changed, but many major peaks did not change substantially, and this did not adequately identify the compounds strongly involved in the aroma change.
They therefore used GC-O, in which people smell the separated compounds, progressively diluting the aroma concentrate to see which odors remained detectable. There were 23 peaks with a flavor dilution factor of at least 100 before heating and 20 afterward.
The FD factor is not a direct measure of a compound’s concentration or of the overall aroma intensity of the drink. Under the same analytical conditions, it provides a clue for identifying compounds likely to make important contributions to aroma.
The five declining compounds had coffee-like, potato-like, and animalic notes
For peaks whose FD factors fell after heating, the researchers further concentrated the samples and conducted mass spectrometry, comparing retention times, odors, and mass spectra with authentic standards. They identified the following five compounds.
- 2-Furfurylthiol: roasted coffee-like
- Methional: cooked potato-like
- 3-Mercapto-3-methylbutyl formate: roasted coffee bean-like
- β-Damascenone: sweet, honey-like
- Skatole: animalic
These compounds were present in amounts too small to appear as separate peaks in the conventional analytical chromatogram, yet they can be perceived at low concentrations. The authors inferred that decreases in these trace compounds played an important role in the aroma difference after heating.
Skatole is structurally related to indole but is a different molecule. An animalic note of an individual compound does not translate directly into the same character in coffee as a whole. Its detection here neither indicates contamination of the coffee nor demonstrates that adding skatole makes coffee taste better.
Testing heat responses in aqueous solutions of individual compounds
The researchers next prepared separate aqueous solutions containing 5 ppm of each of the five compounds, adjusted the pH to 6.8, and applied the same heating conditions of 123°C for 20 minutes. This simplified system was intended to investigate changes that might occur in the beverage.
The proportions remaining relative to before heating were 3.6% for 2-furfurylthiol, 26.0% for methional, and 0.3% for 3-mercapto-3-methylbutyl formate. These three sulfur-containing compounds showed particularly large decreases under the conditions tested.
By contrast, 75.7% of β-damascenone and 79.0% of skatole remained. The extent of change thus differed between compounds despite identical heating. These percentages were obtained in single-compound aqueous solutions; they do not mean that the same proportions were lost from the actual coffee drink.
Not just disappearing, but changing into other compounds
For the sulfur compounds showing large decreases, the researchers also analyzed volatile products formed after heating. Products from 2-furfurylthiol included difurfuryl disulfide, which the authors interpreted as evidence of oxidation.
Dimethyl disulfide was detected in the methional model. The authors discussed the possibility that methanethiol formed through thermal degradation was subsequently oxidized, although methanethiol itself was not detected in this experiment.
Heating 3-mercapto-3-methylbutyl formate also produced substantial amounts of the corresponding alcohol. This is consistent with hydrolysis, in which water cleaves the ester group. Heat-related aroma changes may therefore involve molecules changing into other forms, not only escaping into the air.
However, the study focused on volatile compounds and did not examine nonvolatile products. Some highly volatile compounds may also have been lost during sample concentration. The authors note that further reactions may be involved in real beverages containing milk and sugar.
Summary
Heating a coffee drink containing sugar and milk at 123°C for 20 minutes reduced an indicator of aroma contribution for five trace compounds, including coffee-like sulfur compounds and skatole. In single-compound aqueous models, the three sulfur compounds decreased particularly strongly, and products associated with oxidation, degradation, and hydrolysis were identified.
Changes in coffee aroma can be missed when attention is limited to abundant compounds. This study highlights the importance of combinations of compounds perceptible at trace levels, including unexpected animalic and potato-like notes, for understanding roasted coffee character.
All four authors were affiliated with Ogawa & Company at the time of the research. No separate research funding or conflict-of-interest statement was identified in the text.
Reference
Kumazawa, K., Masuda, H., Nishimura, O., & Hiraishi, S. Change in Flavor of Coffee Drink during Heating. Nippon Shokuhin Kagaku Kogaku Kaishi. 1998;45(2):108–113. doi:10.3136/nskkk.45.108 (Original full text)
