How much caffeine in Pu-er?

Ok. At the risk of appearing difficult, may I ask why you want to know?

The question of how much caffeine is in this-or-that comes up again, and again, and the answers never seem to be satisfactory or conclusive. For instance, despite there being good experimental data, it's not obvious that the results apply far beyond the sample used in the experiment, not to mention the variation in brewing parameters.

If you say why you want to know, you might get a much better answer.

Cheers,

- Joel

All tea leaf have the same (appox) amount of caffeine. However it is easier to extract it in one steep from broken leaf oxidized teas (like average black) and therefore the cup of black tea has more caffeine than others. Black tea is rarely re-steeped and if it is, the second steep has much less caffeine. Whole leaf teas and large leaf teas have the same amount of caffeine, but give it up slower (and not only caffeine, but all its ingredients) and therefore are steeped several times. This way each cup may have less caffeine. Steeping time and temperature also play their roles.

Puerh has the same amount of caffeine (by weight) and depending on technology can give it up quicker and slower. Same while leaf vs broken, temperature, steeping time etc. But nothing is better than direct measurment. If you have a liquid chromatograph handy.

Sasha.

Different puers do indeed contain different amounts of caffeine and it does appear that green puer is lower than most. See

for more info. They used reversed phase liquid chromatography with UV detection for their study.

Caffeine is extremely water soluble. Most traditional methods of brewing puer involve "washing" the tea, literally rinsing it one or two times with hot, often boiling, water. I suspect that this is tantamount to significantly decaffeinating the tea. While I have never seen a study of available caffeine "after" washing, deductive reasoning certainly supports this theory.

Mike Petro snipped-for-privacy@pu-erh.net

remove the "filter" in my email address to reply

Two things, Mike:

1. "Different puers do indeed contain different amounts of caffeine"? Common, Mike - 1.7 vs. 3.4% is statistically not a difference - although it appear to be "twice", but it is only with a 1.7 total percent!. I am sure they would get the same difference if they would measure not a one, but several samples of the same tea with different temperatures!

1. The extraction from green teas is slower and the obly way to really measure teh caffeine content is to do multiple steeps until there is no caffeine left in teas and calculate the sum.

2. Caffeine is highly soluble, you are absolutely right. But you also have to take into consideration that to enter solution it has first be extracted from the dead cells of the tealeaf, and that takes time. And mechanical properties of the tea play critical role in that delay. I very much doubt that any significant portion of caffeine is extracted during short washing period, because the leaves must first be saturated with water to expand and this take time. I actually think that with puerhs and green teas the second and the third steeps (especially the second) is the most caffeinated.

Sasha.

50% more is indeed a difference in my book. Twice as much of a little bit is still twice as much, just ask anyone who is susceptible to caffeine jitters.

I agree that since the extraction method is not listed here, beyond the fact they used powdered leaf, that there is room for question. My assumption is that anyone with access to this type of equipment would know how to do a proper sampling, but that is only an assumption.

Possibly, I would love to see a proper study on it though. I cant help but feel that hot water will dissolve a significant portion of the caffeine immediately, but yes the second steep may indeed still contain a considerable amount. However if you do 2 rinses, as is popular with black puer, then where do we stand?

It is an old trick with black (aka red) tea that if you want to decaffeinate it substantially all you need to do is steep it for 30 seconds and throw away the first infusion. The "washing tea" may not be long enough to saturate the leaf totally but it is bound to have an effect.

Mike Petro snipped-for-privacy@pu-erh.net

remove the "filter" in my email address to reply

The 30 second trick is indeed famous. It's been mentioned countless times, but I don't recall ever seeing good numbers on what it achieves.

/Lew

Sorry, Mike, mathematics does not work this way. The 'twice" thing only makes sense if the you first established that your standard variation is significantly smaller. That can only be done on a number of measurements. If they would have said that they measured several green puerhs of different types and the caffeine content cluster around a certain number (plus/minus standard deviation) and that cluster is significantly smaller or larger than the cluster of black puerhs, then the numbers will make sense. Otherwise they are senseless.

Mike, you know how I respect you, but thi sis very naive of you. You can have access to LH at about $50/hour here in teh US or send the sample to the lab from here to India for about $10/sample. The way they described their experiment without multiple measurments, standard deviations and margins of error (even political pollsters do that!) are self-evident.

OK, let me put it this way - for a "wash" to work, the water has to penetrate the insides of the dried cells of the leaf. The best indicator of that is that the bulk of the leaf is not floating on the sirface anymore. So put your puerh in hot water and measure the time when the bulk of the leaf will settle on the bottom. This is the shortest possible time for caffeine to be dissolved. But it also has to get out into the external part of the solution. If you help it by say, centrifuge, I am quite sure that you will get rid of 90% of caffeine. However as you know for non-boken or large leaf tea to settle on teh bottom takes a LONG time. No 30-40 sek or even 1-3 min wash would even come close. Sorry, pal.

I would recommend a different experimental approach for the goal of decafing a tea (based on extraction kinetics) - put it in ice-cold water until the tea leaves sink (caffeine has a very high solubility - much higher than other tea components) and ice-cold (or just cold) water will extract it with no problem. Then, drain and brew (taking into consideration that it is soaked in cold water, so you may need hotter water or even pre-heat the soaked tea)

Exactly. See the above explanation and technique based on the "floating time"

Sasha

Joel snipped-for-privacy@cindy.panacea.null.org10/29/04

21: snipped-for-privacy@panacea.null.org

Joel,

I didn't ask the question, but I want to know for medical reasons.

Michael

That's still not very specific, but it's a reasonable one to throw at me. I suspect it's a hypothetical and you're testing me. :)

If the caffeine intake needs to be known for medical reasons, then it's something you don't want to take any chances with. So in such cases you want to leave margin for error.

If your doctor says something like "little or no caffeine", then you're stuck with *properly* decaffeinated tea if any at all. On the other hand if you're told "no more than x mg per day" then you have to brew such that you're sure you stay under that limit. That means that caffeine on a per tea-type basis is unlikely to be something you want to mess with. Find the largest reputable result for caffeine content, and then treat your tea consumption as if each cup will yield that amount, regardless of the type of tea. To be honest I don't know what medical conditions would require that kind of limit, but if it's medical than better safe than sorry.

A useful analogy might be blood alcohol limits. It's unreasonable to ask which varieties of scotch have less alcohol, because the differences are likely to be so small that trying to use that information to regulate your intake would only make a difference if you're already drinking close to the limit. And you shouldn't. So such specific information should not be needed for this reason.

My answer would be very different if the reason for asking the question was something like "I don't want the jitters" or "I want to sleep". Leaving a margin for error is not so important there.

Cheers,

- Joel