Burnt taste when extraction water is too hot

A haven dedicated to manual espresso machine aficionados.
User avatar

Postby rpavlis » Mar 01, 2012, 11:42 am

When lever machines are delivering water that is too hot the espresso tastes "burnt". Everyone who makes espresso with lever machines discovers this unfortunate fact very early in the game.

The boiler water on lever machines typically runs between 112 to 120C. An oven set to this temperature will not cause anything to taste "burnt" even if we leave it in such an oven all day long. When the coffee was roasted in the first place it was heated to somewhere between 220 to 240C. (If it were heated too much above that it would have likely ignited.) The question is, how can heating it to a mere 100 to 110 or so during extraction produce this bad burned taste?

The answer to this question seems to me to be simple: When we extract the coffee with higher temperature water (much above 95) we extract compounds that were produced in the roasting that produce the burnt taste that otherwise remain in the grounds. These compounds are probably highly polar and absorb tightly onto the coffee substrate so it takes higher temperatures to extract them. With water of 100C or more we extract more and more of this stuff which results in espresso that tastes like it had been cremated.

Moka pots often produce a material that tastes burnt because heat is conducted up to the coffee and heats it so much that these same bad flavours extract. Stainless steel ones tend not to do this so much simply because stainless is, for a metal, an unusually bad heat conductor.

User avatar
Team HB

Postby TomC » Mar 01, 2012, 2:37 pm

Now go and apply this to every single variety of coffee you consume and come back with something useful.

User avatar

Postby bostonbuzz » Mar 01, 2012, 2:46 pm

The question is, how can heating it to a mere 100 to 110 or so during extraction produce this bad burned taste?

Answer: We grind coffee very finely, making it far, far, far, far more susceptible to temperature changes. Water conducts temperature far greater than air as well. A bean in a roaster takes minutes to reach a high temperature, but in a portafilter it takes seconds (or less), AND it's finely ground. Burning happens easily, and how it gets in the cup is no surprise, although interesting to theorize about.
LMWDP #353

User avatar

Postby rpavlis » Mar 01, 2012, 3:09 pm

In laboratories we commonly conduct reactions with materials so finely divided that the particles are micron sized. The temperatures required to bring about pyrolysis of sugars, fats, etc. are at least 170C and this is NOT related to particle size.

Since this coffee has been subjected to temperatures far in in excess of this during roasting, there are LOTS of pyrolysis products already in the beans. If you look at the structures of these compounds they are highly polar and hence will be strongly bound to the beans. Any type of chromatographic system will elute more materials with higher temperature. What we really are doing when we get the water too hot is elute these materials that were produced in the roasting that we most definitely do not want in our espresso. Espresso making is really a type of liquid chromatography. We are separating the good tasting materials from the bad tasting things because the good tasting things are less strongly bound. The hotter the water or the more water we run through the more the nasty tasting things come forth.

Hotter water is removing already produced "burned" material from the bean residue and putting it into our espresso. It is simply a matter of elementary physical organic chemistry. (My profession, by the way.)

Lower extraction temperatures leave these nasty tasting things on the beans.

User avatar

Postby rpavlis » Mar 01, 2012, 3:32 pm

Also the primary compounds that are responsible for the odour of burnt sugar and the like are the enol forms of cyclic 1,2-diones in five membered rings that are produced from sugars by dehydration. In the aqueous environment of espresso production there is no way that such dehydration could occur.

User avatar

Postby bostonbuzz » Mar 01, 2012, 6:45 pm

This is very interesting and difficult to understand since I have no chemistry background. It makes sense that lighter roasted coffee needs a finer grind AND a higher temperature to extract well. What I meant in terms of grind size is that it reacts to temperature FASTER the finer the grind.
LMWDP #353


Postby jackson6 » Mar 01, 2012, 10:55 pm

So, an espresso machine is a type of refinery. Different compounds a extracted at different temperatures. I don't know if there is a chart of all the major compounds and their corresponding extraction temperatures. I'm sure how time plays a large part as well. I don't have enough chemistry in my background. Hopefully someone can tell me if this makes any sense.


Postby entropyembrace » Mar 01, 2012, 11:58 pm

If espresso brewing works like liquid chromatography then your ground coffee contains a bunch of chemicals each of which extract at different rates. Some are faster and some are slower...and the nasty bitter stuff seems to be slower. Whenever you do something to increase the extration rate (such as grinding finer or increasing the temperature) you increase the extraction rate of ALL of the compounds in the coffee. The faster extracting compounds are still the faster extracting compounds...the "order" doesn't change, it's just that everything is extracted more quickly.

This makes sense with how coffee brews...you aim for your 25 second extraction but the temperature is too high and you get bitter coffee. If you made the coffee at the correct temperature but with an extraction time that is too long...say 60 seconds...you get bitter coffee then too.

User avatar

Postby allon » Mar 02, 2012, 12:26 am

This doesnt hold true on a commercial spring lever group - even though the dipper tube brings super hot water to the coffee, even though the extraction may take a long time, the pressure and temperature profiles (declining) yield a thick sweet shot - not bitter.

Max temp is only part of the equation.
LMWDP #331

User avatar

Postby orphanespresso » Mar 02, 2012, 4:42 am

But you have to admit this is a very interesting analysis and explanation.