Rate of coffee staling

[RapidCoffee]

For every 10°C change, the speed of a chemical reaction changes by a factor of 2.

Ignoring boffins with spectrographs, coffee staling and freezing is simple chemistry :

- Changing the temperature from +20°C to -20°C slows staling by a factor of 16. In laymans terms, two weeks in your typical refrigerator freezer equates to one day at room temperature.
- Changing the temperature from +20°C to -30°C slows staling by a factor of 32. In laymans terms, one month in your typical chest deep freeze equates to one day at room temperature.

Dave's [and Ken's] freezer @ -15°F is midway between, so let's say 3 weeks = 1 day.
Ken's coffee, frozen immediately after roast for 4 to 8 weeks, accumulates ~1.5 to 3 days.

Ian, thanks for your feedback. But I doubt that the rate of coffee staling has any simple relationship (exponential or otherwise) to temperature... Is there a substantive basis for this statement?

[cafeIKE]

But I doubt that the rate of coffee staling has any simple relationship (exponential or otherwise) to temperature... Is there a substantive basis for this statement?

"Staling of coffee occurs gradually as the result of numerous chemical processes affecting the coffee at different rates. The actual amount of time in which these processes take place will depend upon the state of the coffee (whole bean or ground) and conditions of storage (amount of oxygen contact, heat, moisture, and light)." from Coffee Analysts -> Packaging.

If temperature did not affect coffee staling, it would be a chemical anomaly.

[RapidCoffee]

If temperature did not affect coffee staling, it would be a chemical anomaly.

Sure. I didn't question whether higher temperatures accelerate staling (of course they do). I just wanted to know if there was a substantive basis for your claim that every 10C accelerates staling by a factor of two.

I assume from your response that there isn't.

[cafeIKE]

Sure. I didn't question whether higher temperatures accelerate staling (of course they do). I just wanted to know if there was a substantive basis for your claim that every 10C accelerates staling by a factor of two.

I assume from your response that there isn't.

It's not my claim. It's basic chemistry.

Turning your argument on it's head, are you saying there are special coffee properties that cause the chemical processes of staling to ignore the laws of nature?

Ken's short term and Dave's long term freezer tests are close enough to what one would expect from the science to validate the assertion.

[RapidCoffee]

Turning your argument on it's head, are you saying there are special coffee properties that cause the chemical processes of staling to ignore the laws of nature?

Ken's short term and Dave's long term freezer tests are close enough to what one would expect from the science to validate the assertion.

Kindly refrain from misquoting me. I'm simply asking for substantiation of what seems to be an extraordinary claim: coffee stales at twice the rate for every 10C rise in temperature. Neither Ken nor Dave's results come anywhere close to proving this assertion. I don't buy it, but I'm willing to be convinced...

BTW, reaction rate doubling for every 10C is not a "law of nature", it's merely a rule of thumb in chemical kinetics.

[AndyS]

"Staling of coffee occurs gradually as the result of numerous chemical processes affecting the coffee at different rates. The actual amount of time in which these processes take place will depend upon the state of the coffee (whole bean or ground) and conditions of storage (amount of oxygen contact, heat, moisture, and light)." from Coffee Analysts -> Packaging.

If temperature did not affect coffee staling, it would be a chemical anomaly.

Of course temperature affects coffee staling. I believe the point RapidCoffee is trying to make is that it's probably an oversimplification to say that coffee neatly follows the 10C rule of thumb.

Here's one reason why your Coffee Analysts quote is an oversimplification: coffee staling is not only a chemical process, it's a physical process as well. As the coffee ages, CO2 and aromatic compounds gradually work their way out of the coffee and are lost into the surrounding space.

It's reasonable to assume that this diffusion rate is temperature dependent, but does it follow the same 10C rule as the chemical processes? I don't know, do you?

[luca]

Hi Andy,

Thanks for succinctly stating the question!

I think that it's relevant to point out that espresso roasts run the whole gamut of resting periods; from 2 or 3 days to 2 or 3 weeks, and perhaps even longer if nitrogen flushed. So I'd be a bit surprised if there was some rule of thumb that could be applied, although I'm sure that the rest times are repeatable for any given pro roast.

Cheers,

Luca

[cafeIKE]

Here's one reason why your Coffee Analysts quote is an oversimplification: coffee staling is not only a chemical process, it's a physical process as well. As the coffee ages, CO2 and aromatic compounds gradually work their way out of the coffee and are lost into the surrounding space.

Let's try a famous AndyS thought experiment: Roast some coffee. Split the roast in to equal parts. Grind one part as for drip. Seal both halves for a couple of weeks. Would you expect a markedly different gas and aromatic volume? Yes, if there was a lot of CO2 and aromatics physically trapped in the bean released by grinding. : No, it's likely that CO2 and aromatic compounds are by-products of the staling process.

I don't know for certain, do you? But my money is on NO.

[AndyS]

Let's try a famous AndyS thought experiment: Roast some coffee. Split the roast in to equal parts. Grind one part as for drip. Seal both halves for a couple of weeks. Would you expect a markedly different gas and aromatic volume? Yes, if there was a lot of CO2 and aromatics physically trapped in the bean released by grinding. : No, it's likely that CO2 and aromatic compounds are by-products of the staling process.

I don't know for certain, do you? But my money is on NO.

Hmmm...further evidence of the futility of thought experiments?

I don't understand the point you're trying to make, but yes, there is a LOT of CO2 and a lot of aromatics physically trapped in freshly roasted beans. They eventually make their way out, whether the beans are ground or not.

Carbon dioxide formed during roasting is trapped in the cellular structure of the bean and is only released over a period of weeks following roasting, resulting in a 1.5-1.7% weight loss....The driving force at the basis of carbon dioxide and volatile release from roasted coffee is given by a diffusion flow due both to concentration and pressure gradients....The few data available about volatile partition kinetics from roasted coffee packed in air indicate that the rates of CO2 and volatiles release are of the same magnitude.

It is possible that some of the CO2 released in staling is newly formed as the result of chemical reactions, but the vast majority has already been produced in the roasting process. As far as the aromatic compounds are concerned, there are hundreds of desirable aromatic compounds formed during roasting (according to Illy). In the aging process, many of these are gradually oxidized to form characteristic stale coffee aromas.

Interestingly, in looking through Illy's book to find the section quoted above, I found a paragraph that speaks directly to the effect of temperature on CO2 and volatiles release:

in the experimental conditions adopted, for any 10C increase, the rate of volatile release increased 1.5-fold

If it's true that the rate of chemical staling reactions doubles with every 10C, and the rate of CO2/volatiles release increases 1.5x with every 10C, then your original assertion is, again, an oversimplification:

coffee staling and freezing is simple chemistry :

- Changing the temperature from +20°C to -20°C slows staling by a factor of 16. In laymans terms, two weeks in your typical refrigerator freezer equates to one day at room temperature.
- Changing the temperature from +20°C to -30°C slows staling by a factor of 32. In laymans terms, one month in your typical chest deep freeze equates to one day at room temperature.

[cafeIKE]

Ok, I should have said ...slows the chemical processes of staling by about a factor of...

mea culpa

[another_jim]

It's true that chemical change and diffusion rates are always a simple exponential function of temperature; but each reaction or diffusion process has a different rate constant, and most of the many reactions that occur in staling are unknown. The rub here is that there is no closed form solution for the inverse of a sum of exponential terms. This means that there is no method for generalizing about staling rates, the numerical rates you observe in your experiment does not lead to knowledge of what it would be in other circumstances.

Trust me, I know all about this gotcha.

[AndyS]

there is no closed form solution for the inverse of a sum of exponential terms.

Thanks, Jim, that's what Ian and I were trying to say all along!

But seriously, part of what makes staling (or perhaps, "aging") so interesting is related to what you say: "most of the many reactions that occur are unknown." It's particularly intriguing to me that at least in espresso, a coffee may be peaking in high note flavors only a few days out of the roaster. And then, after a decline for week or even two, it may hit another sweet spot of more mid-range and low notes.

[RapidCoffee]

It's true that chemical change and diffusion rates are always a simple exponential function of temperature; but each reaction or diffusion process has a different rate constant...

This is a big part of what concerns me with the 10C rate doubling. There are many compounds in coffee that contribute to flavor, each of which undoubtedly stales at its own rate. For example, Illy's Complexity of Coffee article (Scientific American, 2002) states:

The aroma of green coffee contains some 250 different volatile molecular species, whereas roasted coffee gives rise to more than 800.

Another concern is phase change. In particular, some lipids will transition from liquid oils to solids over the range of temperatures involved (water too, of course). I'm sure there is dramatically different staling behavior in different states of matter.

Ok, I should have said ...slows the chemical processes of staling by about a factor of...

This is not a triviality. For a 40C temperature difference, assuming 2X per 10C yields a factor of 16X. Assuming 1.5X per 10C yields a factor of only 5X. Regardless, for the reasons mentioned above, I still don't buy into such a simple model.

[cafeIKE]

Regardless, for the reasons mentioned above, I still don't buy into such a simple model.

Doesn't Ignoring boffins with spectrographs, say the same thing?

[AndyS]

This is not a triviality. For a 40C temperature difference, assuming 2X per 10C yields a factor of 16X. Assuming 1.5X per 10C yields a factor of only 5X.

It leads to some interesting possibilities. Normally, coffees are aged at room temperature (say, 21C). If one aged them at 1C, theoretically the degassing would proceed at ~44% of the room temperature rate while the chemical staling reactions would slow down to ~25% of their room temp rate (again, the scenario is oversimplified, ignoring the possible phase changes you mentioned and a host of other factors). *

But still, 7-10 days aging at 1C might be about right. If one properly protected the coffee from moisture, we might get a different (and improved) flavor profile: degassed coffee with less chemical staling.

Or maybe not.

* [edit] To clarify, one could try aging the coffee at 1C, for 2.25x longer than one's typical room temperature aging period. At that point, it would theoretically be degassed about the normal amount, but would have experienced only about 56% of the usual chemical staling reactions.

[cannonfodder]

I am just a computer guy, not a PhD so I have a much simpler set of guidelines for aging. Dave's rules of coffee aging...

1. all coffee ages
2. freezing slows aging
3. different coffees have different prime flavor windows
4. to prevent coffee from aging beyond its prime flavor window, drink it faster
5. coffee past its espresso prime will often still make a good drip/press pot
6. don't order more than you can drink before rule #4 is enforced
7. if your coffee is all stale and the next batch is not ready, drink tea or beer and plan better next time.

[RapidCoffee]

I suppose I am a "boffin" (had to look that one up ), but I agree with Dave's guidelines. I might add:

8. Any coffee that you will not consume within roughly two weeks should go into the freezer.
(Personally, I believe that freezing does impact the taste of coffee... but it's much better than the alternative of letting it go stale.)

Thanks to Ian for bringing up this interesting topic. It's certainly worthy of further exploration.

[cannonfodder]

I keep a small air tight jelly jar of beans in the freezer. That is my 'plan better next time' stash. I let a coffee age to within two days of its flavor window, put some in the jar, date it and put it in the chill chest. If I run short, I have two days worth in storage ready to go the next day. I cycle that stash every month so it does not sit around getting stale.

[welone]

Here's a long answer where that rule of thumb comes from and for what it can be used. And after that an alternative calculation method, which is imo more useful when dealing with gases escaping the bean.

It is derived from what is known as van't Hoff rule (which is also the basis of the Q10 temperature coefficient and known to the german speaking by Reaktionsgeschwindigkeit-Temperatur-Regel). The assumption it makes is that the state change of the compound of interest involves an energy of about 50'000 Joules per mole (=12 kilocalories (kcal)). When making a small adaptation to it (related to the energy involved) it gives a good estimation of the temperature dependence of a whole bunch of biological, chemical and physical processes. The exact relation is described as "temperature dependence of equilibrium constants (K) or rate constants (k) as a function of the corresponding Enthalpy Changes (delta H) or Activation Energies (Ea)" (from Environmental organic chemistry, schwarzenbach et al, 2003). The adaptation you can do requires some knowledge of the kind of reaction your dealing (you have to look-up the 'delta H' or 'Ea' from it) with that you can estimate the average factor for a change in temperature (of 10°C for instance). A few examples: a reaction with a 'delta H' or 'Ea' of 10'000 Joules has an average factor for a change in temperature of 1.2 per 10°C (basis was 0-40°C) one with 30'000 Joules a factor of 1.5, and as already mentioned 50'000 Joules= factor of 2.0).
To estimate the change in volatilization during staling of coffee when cooled, you can calculate the change in vapor pressure of the compound of interest depending on its Enthalpy of Vaporization (for liquids) or Enthalpy of Sublimation (for solids). Generally there's a clear trend that the bigger the molecule the lower its initial vapor pressure and also steeper it diminishes with decreasing temperature. Especially for solids the extra amount of energy needed to sublimate it (equal to the amount to first melt and then vaporize) accounts for a stronger dependence of the vapor pressure from temperature than in the case of liquids (or subcooled liquids).

All this may be used for thermodynamical state changes (like the evaporation of a liquid or the sublimation of a solid), chemical processes (for example hydrolysis or redox reactions) or biochemical processes (enzyme reactions). For biological systems dealing with enzyme reactions it only holds true for temperatures roughly from 10-30°C; and doesn't apply at all for anything in the range of freezing temperatures because the proteins already denatured.

When dealing with gasses it seems more appropriate to use the 'ideal gas law'. Especially in relation to the quote from illy: "The driving force at the basis of carbon dioxide and volatile release from roasted coffee is given by a diffusion flow due both to concentration and pressure gradients...."
You can calculate the pressure change of gases (like CO2) inside the bean by applying the 'ideal gas law' (p*V=n*R*T). It is as easy as forming the quotient of the two temperatures of interest (in Kelvin!) to yield the change in pressure: For example the pressure change from 25°C (=298.15 K) to 0°C (=273.15 K) results in a decrease of pressure of 8 %. Or from 25°C to -20°C a decrease of 15% in pressure.
Calculated with:
pressure change in percent=(1-(273.15 K/298.15 K))*100 ; with the general form of the equation being p2=p1*T2/T1; from p/T=n*R/V=constant
The most critical assumption you make when applying it is, that you can neglect the potential 'loss of substance' and decrease in volume during the initial cool-down phase (which is why p/T is considered to be constant).

What concerns the meaning of the concentration gradients; this causes the beans to lose their CO2 less rapidly when in a closed container surrounded by CO2 instead of air.

greets

marco

[Alan Frew]

Let's try a famous AndyS thought experiment: Roast some coffee. Split the roast in to equal parts. Grind one part as for drip. Seal both halves for a couple of weeks. Would you expect a markedly different gas and aromatic volume? Yes, if there was a lot of CO2 and aromatics physically trapped in the bean released by grinding. : No, it's likely that CO2 and aromatic compounds are by-products of the staling process.

I don't know for certain, do you? But my money is on NO.

Thanks, you can send the $ to my P.O. box address. I've done the experiment, ground vs. whole beans, about an hour out of the roaster, vac packed in non-valved bags. You don't have to wait 2 weeks, the ground bag inflates within 20 minutes, and stays puffed. The whole beans bag did inflate after 24 hours, but never to the extent that the ground bag did.

You haven't been following the literature if you don't already know that a LOT of CO2 is physically trapped within the cellular structure of the bean, and is immediately released by grinding. Again, simple physics will tell you that the increase in exposed surface area caused by grinding will increase the rate of diffusion of any gases trapped within the whole bean.

Chemistry was my first career, and your rule of thumb works as an approximation for dilute solutions. Complex organic solids??? No way.

Alan