Too Much Coffee

Using colorimetric sensor arrays scientists have successfully developed a method of "analytical cupping."
They note that in addition to being able to accurately differentiate between beans or blends they can even detect differences in roasting parameters (time and temperature). They used coffee because it is readily available in different types and qualities, but the method can be used for numerous complex substances.

As a note of clarification: This method does not allow detection of the individual compounds that make up a substance, but rather creates a fingerprint that allows the substance (in this case individual bean variety or blend) to be positively identified and distinguished from other varieties or blends.

(Abstract of the work available here).

BTW, as an aside, I might note that the work was done at the University of Illinois where I (albeit unsuccessfully) began my academic adventure hundreds of years ago.
)

bruceb wrote:Using colorimetric sensor arrays scientists have successfully developed a method of "analytical cupping."
They note that in addition to being able to accurately differentiate between beans or blends they can even detect differences in roasting parameters (time and temperature). They used coffee because it is readily available in different types and qualities, but the method can be used for numerous complex substances.

As a note of clarification: This method does not allow detection of the individual compounds that make up a substance, but rather creates a fingerprint that allows the substance (in this case individual bean variety or blend) to be positively identified and distinguished from other varieties.

(Abstract of the work available here).

Interesting, I tried to read the article to see if they had run tests with greens of different ages and humidities to see if these changed the accuracy of differentiation. Unfortunately the site "politely" advised me that my "credentials do not allow retrieval of the full text". How did they know?

I was also underwhelmed by the statement in the extract that they could distinguish roast time to a resolution better than 5 minutes! That could be impressive if all roasts were to the same end roast colour but it could also be the difference between green and charcoal.

bruceb wrote:BTW, as an aside, I might note that the work was done at the University of Illinois where I (albeit unsuccessfully) began my academic adventure hundreds of years ago.
)

It must have been a difficult time during the civil war.

I certainly agree that the details here are (for TMC members) rather strange looking, but you have to remember that uni boffins usually have no idea about good coffee. I am looking into buying the full article, but it seems they want strange forms of payment (strips of skin from the upper back, etc.) so it may take awhile. I think it is interesting, nonetheless, that they can fingerprint individual varieties. For German readers or those who would struggle with a translation by the Fish I found a review of the whole paper in the Spiegel here.

The civil war wasn't so much the problem, but the tea they served in the student union was repulsive, so much in fact, that I quit my studies for a century or so.

Interesting bit of work and according to the paper, the coffees that they looked at were commercially available mainstream US consumer brands and a Colombian coffee from Huila region which was subjected to 'lab roasting'. The latter process was carried in a lab oven as opposed to a sample roaster or equivalent and the resolution of the array methodology (10 deg C) is perhaps down to claims they can be made around the model matrix (it looked at 10 deg intervals) rather than what it can resolve to in actuality, i.e. more work required.

I think that the major application would be in roasting process quality control particularly for multi-site operations when used in combination with other analytical techniques and sensory analysis.