Whoa! Too much coffee dude. Maybe just limit yourself to the aroma... (the chemistry of coffee aroma)

So I see Ben has put up a post on the various reviews out there on Dawkins' book, and I noted one comment saying:

This is utterly ridiculous. The distilled argument here is that the discussion of religion and god(s) should include only believers. Perhaps you should think about that for 3 or 4 seconds.

I won't comment further on the post itself, since I happen to think Ben does a good job in raising a valid point, but I will say to the commenter to maybe lay off the coffee for a while (although I will note that he did weigh in again in a much calmer fashion).

Anyway, I bring this up more as a segway segue into a recent magazine request I got, that asked if I could help in a graphic related to the chemistry of coffee roasting - in so far as what actually happens to the green (and otherwise sensory null) coffee bean during roasting, which in turn gives it its brown colouring and flavor/aroma rich attributes.

Initially, this started off as a question on what "caramelization" would look like in a chemical graphic sort of way. But, this is actually a tough one to broach, since the chemical act of caramelization is a loosy goosy sort of affair, as well as being only one of a variety of processes that occur during the roasting process.

Look at it this way: The browning effect is pretty complicated (and to a large extent even uncharacterized), in that it's akin to taking something composed of a kazillion different things (i.e. the umpteen molecules in a piece of tissue - in this case a bean), heat it to the point that the likelihood of all manner of enzymatic and chemical reactions is significantly enabled, and then stand back and see what happens when these kazillion or so things get to hang out and interact.

In other words, there's a lot going on, and I'm not sure how pertinent a graphic on one potential chemical change would be.

Still, one train of thought I had would be to look up whether there were any terpenoids that have been characterized with respect to coffee. I have a colleague down the hall who works on the biochemisrty of these molecules, so know quite a bit about them myself.

Just to bring you up to speed, terpenoid molecules are a major class of volatile phenolic molecules (have a big honking ring structure) that all plants produce (via enzymes known as terpene synthases). There's like tens of thousands of different ones so far characterized, but basically they're responsible for most of the scents and flavours you get from the plant kingdom. i.e. smell the pine forest? That's a terpenoid. Garlic breath? That's a terpenoid. Whatever it is that attracts bees for pollination? That's a terpenoid. Stinky stuff like citronella to repel bugs - also a terpenoid. So I thought that maybe there's one (or a few) that are a big part of the flavour of coffee (on top of the caramelization and the meat browing stuff - whose name I forget - oh wait it's called the Maillard reaction). I should also note, while I'm of mind, that terpenoids also do a lot of other things (particularily in the plant defense arena).

Anyway, from this train of thought, I found this really cool paper (Laser Mass Spectrometry as On-Line Sensor for Industrial Process Analysis: Process Control of Coffee Roasting), that basically looks at the compounds responsible for coffee aroma, using pretty high tech means to do so.

Check this figure out:

Although, it's a little busy, it's really quite extraordinary. Just remember that the placement of the peak along the mass [m/z] axis is the identifier for a particular compound (i.e. caffeine, etc should have a mass to charge ratio at just so), the height being the relative amount of that compound, and the final (in front of you) axis is the roasting time. If you look closely, you can even see that the caffeine seems to come about at very define and discrete times, which turns out to be indicative of a single bean popping (they use 4 to 5 beans per experiment) from the heat. This popping essentially releases a whole whack of caffeine at once.

Anyway, this paper does a pretty extensive look at a variety of compounds released during roasting, as best summarized in the following table:

None of these are actually terpenoids, but in hindsight that makes sense, since production of terpenoids is a biochemical process, dictated by enzymes. These compounds (which are still largely phenolics), on the other hand, are made by the chemistry induced through high heat of roasting. So it's more of a free for all of production.

Pretty neat in any respect. And in the end, I suggested using a couple of these compounds for the graphic - I guess we'll just wait and see what they'll use in the print article (in an upcoming issue of MAKE magazine).