Sciencemadness Discussion Board

Extracting caffeine from coffee - photo essay

peach - 10-7-2010 at 09:05

Someone over at psychonaut was asking about extracting caffeine. He can't easily get hold of dichloromethane, so wanted to do it with acetone. I also have ethyl acetate handy and know this is used as an alternative, but that's even trickier for most people to find so I'm not using that.

I'm trying to do a little experiment each day, for fun. Whether it's making a salt, drying them out or something like this. I started brewing 10l of alcohol yesterday, more on my fiendish plans for that later. For now, here's caffeine. If you have the solvents, give it a go, it's a reasonably fun.

DCM EXTRACTION OF CAFFEINE FROM COFFEE

SAFETY:
  1. This extraction uses DCM, which is now labelled as a toxic chemical. Familiarize yourself with the WIKI at the end before using it. Aside from being toxic, DCM HURTS if it gets in your eyes, itches, melts a lot of plastics and strips all the oil out of skin
  2. Do not boil solvents over a gas hob. It'll have to be electric. And don't smoke standing next to it or flick electrical appliances on and off too much
  3. Heat them in a water bath (pan of water) or place a bowl over the top. Turn on the fan hood and open the windows.
  4. Think carefully about your solvent purity and drying method if you plan to eat the results
  5. Keep in mind that caffeine can kill you, and has killed people consuming large amounts of it


Here's some of the bits I'll be using. I didn't need the hotplate. Neither will you need a toaster or a jar of lime jam.


I have some very nice gear which isn't in this picture, and was trying to show that it could be done with everyday items. You don't even really need the sep funnel or flask, a cup and a pipette will do (or a syringe with a blade - a turkey baster won't :P the DCM will pour out of it as you try to transfer it, and it'll rot the rubber in seconds). One item I don't yet have is a nice set of balances. I'm stuck with these nasty, sweat shop, drug dealer looking, micro things. But to my surprise, I put the first new penny I had to hand on them and got this, which is dead on the minted weight of a penny (3.56g). Not so bad after all for £9.99. Particularly since, as you can see, I haven't taken very good care of them over their years from calibration in the land where people work harder than here


Weighing out a large amount of Nescafe Gold Blend, freeze dried, instant coffee. I'm using this much since it'll just about fill my 250ml sep funnel when put into solution


I'm expecting somewhere around 0.5g of caffeine from that. I wanted to do as much as possible so I could get a decent, visible result for the photos. Incidentally, this is around five times the recommend daily allowance. If you're planning on eating, drinking, smoking or injecting the results, you should be aware that people have actually died as a result of eating sheets of caffeine tablets on nights out, as it causes tachycardia (the syncing clock of the heart ticks incorrectly, creating irregular beating which can cause a heart attack); you've been warned...



Adding some bicarbonate of soda to encourage the caffine out of hiding. Don't add more. I tried with 10g and it turned the whole lot into foam. Trying to extract it produced a persistent emulsion, which salt didn't improve. I binned that and started again with this much


Starbucks ain't got shit on this... the "heart attacker" coffee. Note the stained DCM clearly collecting at the bottom. I've used around 50ml of DCM here


You need to be very gently with this. DCM usually splits from water nicely. But you have some base here and if you swirl slightly too hard, it'll go to an emulsion. I swirled a few times and the DCM puddle gradually decreased. When it suddenly disappeared, I stopped. It soon began to reappear however


Whilst it settles, I do some tidying up to make better use of my time. It takes about five or ten minutes to reach this point.


Good to go. You see there are some bubbles left in there and coffee trapped at the boundaries? You can sometimes remove them by tapping the funnel. Another trick I use, which I did here, is to drain the funnel very slowly. As it drains, the bubbles will 'pop' against each other.


I left some gunk behind, purposefully. I'd rather have clean caffeine than a mg more. If you were doing this with a pipette, use a long test tube or a cup. If it's a cup, tilt it at 45 degrees to make the DCM accumulate in one spot. Then poke the pipette down to the DCM and suck it up. You could also use a syringe.


The crude DCM extract


The DCM has a little water absorbed in it, making it cloudy, and that will remain during evaporation, so I add a few grains of CaCl2. You can use any dry salt, like sodium sulphate (Gluabers salt) or magneiusm sulphate (Epsoms salt, available from the pharmacy isle or counter). Just be sure to bake the salt dry first.


Most of the salt remains behind and is easily separated....


But I don't want salty grit in my ultra pure caffeine, so it goes for a ride through the old kitchen chemist's favourite, coffee paper. Yes, I have genuine, rated papers and sintered glass, but it's not in the spirit of things for this, or necessary. Coffee filters cost pence, where as genuine lab papers can be quite a lot more. The coffee papers are available from ASDA. Since that's Walmart, steal them. Or buy them from Tescos, the UK alternative.


The super licker...


Rather than put it on the hotplate and risk bursting this bit of kitchenware, ruining my efforts, and realizing a lot of beginners don't have a hotplate, I simply place the ceramic dish inside a plastic bowl and add hot water


Not long after...


Shortly after that...


Then it's minutes until I have caffeine...


This is the weight with the boat


And without the caffiene it's 0.33g, indicating a yield of 330mg crude caffiene. You can stop here if you're after the crude yield, but I think we can do a little better than that


Preparing for recrystalization [edit: reading this back, that looks a lot like Farther Jack's earwax candles]


Adding some ethyl acetate. The guide I'm using suggests 1 drop per mg, which means 8ml for me; as one drop is around 0.05ml. The solvent is then heated in a pan of boiling water. I found that wouldn't dissolve all my solid, so I ended up adding around 30ml of acetate, and it still wasn't dissolved. So there are insoluble impurities in there. Instead, I hot filtered the 30ml.


Here's the approximate 30ml of ethyl acetate after hot filtering. I then boiled off the acetate until my solid returned, which looked to require around 8ml of solvent; as calculated. Cutting it short, it should be about 8ml of solvent to 330mg of caffeine, the persistent solid needs hot filtering. Recrystalizing from this solvent excess would net you not a lot



In the MIT digital lab videos (which I highly recommend beginners check out, online, for free), they place a recrystalization in a beaker lined with tissue to slow the rate of cooling. This produces bigger, often purer, crystals, as they have longer to form. But NurdRage from youtube suggests this far more impressive trick in one of his videos, which is to place the vessel inside a thermos, which slows the rate of cooling right down, netting you the best chances of impressive, clean crystals. I can only improve on his suggestion by saying you could warm the inside of the flask up with boiling water first


Back at room temperature, around an hour or so later. Very pretty. Making crystals has to be one of the most enjoyable aspects of chemistry


I'm now dealing with a very tiny sample, so I am switching to using a POR1 sintered funnel under vacuum and I'll be lining the frit with some genuine paper. Whilst I'm setting this up, the sample has gone in the freezer to crash as much out as possible, along with 5ml of ethyl acetate


Here's the sintered disc if you've never seen one before. This is POR1, which means it has pores around 100-160um in size. They can go up to POR4 (10-16um) for polishing filtrates. You shouldn't clean sintered wash heads or filters with Na/KOH, as these rot the surface of the glass off. Not usually a problem, but it'll cause frits to fall apart over time. Use concentrated acids, aqua regis or piranha mix instead; all three of these are genuinely dangerous and not advisable for beginners (the aqua regis is abnormally corrosive, as is piranha mix, but the latter also explodes on contact with organic solvents and such). Filter paper will do the POR4 region. Below that, you need membrane filters for bacteriological / spore based sterilization nm numbers


I cut a little disc of paper to line the frit


Jam it in the bitch and get her wet...


Here's the solution after a five minute visit to the ice world


I scrape the coffee smelling slushy out onto the filter, then triple rinse the vial and empty the washings over the cake, then leave it to suck dry for a few minutes


CANDY FLOSS! Caffeine flavor


Lost a fair amount due to laziness, but I now have my reasonably clean crystals. See the improvements section for why my yield is likely so low (the filtrate probably needs re-recrystalizing to get the rest)


The extracted crystals can live in this little amber bottle, with a PTFE wrapped thread, and join my ever expanding collection of curiosities...








ACETONE EXTRACTION OF CAFFEINE FROM COFFEE

I reran this with a similar mass of coffee and simply poured 250ml of acetone over it, then shook the ass out of it. The acetone remained somewhere colorless. I filtered, then boiled off the acetone to yield a scum on the dish. When I scraped this up, the following was the total result


INTERESTING FACTS

Caffeine is a psychoactive stimulant

It scores level 2 on the NFPA 704 system of health warnings, with level 4 being the highest. I would recommend all beginners get used to the NFPA warning system, as it's effectively the metric of warnings and far more informative than a skull and cross bones, fire or cross, yet also very simple to read.

An LD50 for someone my weight using the rat model is around 11g, which make it around 15 times more dangerous than table salt and 40 times more dangerous than alcohol. Nicotine, however, is around 200 times more deadly than caffeine; not counting the cancer.

Caffeine seems to interact with the stress related endocrine hormone system and may increase physical fatigue of the body by stimulating metabolism and other stress / fear related systems into an excessively active state

Caffeine can be sublimated to clean it up. I may give this a go.

The molecule of caffeine looks like this



WHERE CAN I FIND THE THINGS TO TRY THIS MYSELF?

Everyone knows acetone is nail polish remover. Provided you don't plan to consume your crystals, you could probably get away with actual nail polish remover for a crude extraction. The cheaper it is, and so the fewer special additives it has in it, the better the results will be. You may even be able to find it pure.

DCM makes up around 50% of paint stripper - the kind that stinks, feels cool, then hot and itches if it touches your skin. If you can distill things, you can obtain DCM by atmospheric distillation of paint stripper (after washing the polymer based thickener out first). Again, the cheaper and shittier the container looks, the more DCM it's likely to have in it and it'll probably have fewer other additives.

Ethyl acetate is used in some acetone free nail polish removers. I've only ever seen it for sale once as pure acetate however.

To dry your extract, you can use use a number of salts, as mentioned. But they must be entirely baked dry first.

DON'T use colored, scented nail polish remover to produce things you're going to eat. And be very careful eating the results, if that's what you plan to do. Remember that you need to preferably vacuum boil off the DCM, because it comes in a box like this;



POSSIBLE IMPROVEMENTS

I don't think you need to use very much DCM at all. When boiling mine down, I didn't see the majority of the product until the last had gone, indicating it is perfectly soluble in the solvent.

You could try adding a little more base, but I'd try 4-5g, not 10g.

Use a stronger, cheaper blend of coffee, and a lot of it. I used about £1's worth, Nescafe is fairly classy and not that strong.

Try changing the recrystalization solvent.

Extract it from Pro-Plus pills if you want a mountain of it cheaply. There's 50mg of caffeine per pill, and a 48 pack costs about £4. With 48 yielding around 2.4g, that's about £1.70 a gram. If I find any lying around the house, I'll run it against them and might sublime the result. Depending on how kinky I feel, I might even try using the toaster for that.

I probably lost a fair bit in the ethyl acetate I threw out after the first recrystalization. Reheat your collection flask, pour the filtrate back into the small vial and repeat the cooling process to see if there's anymore in solution.

RESOURCES

MIT DIGITAL LAB MANUALS - VIDEOS <---- If you watch one film this weekend, make it theirs

The PDF guide I used

The WIKI for dichloromethane

The WIKI for caffeine

THE END, FOR TODAY...

I hope some of you liked that and have a go yourself. I will likely be posting a number of photo essays over the coming weeks and I intend to gear as many of them as I can to simplicity if possible. I will also be doing some more complex work. There will be much quickfit glassware porn, inert, dry atmospheres and other interesting bits and pieces.

[Edited on 10-7-2010 by peach]

mnick12 - 10-7-2010 at 10:28

Good job,
I dont know if you have one, but I have heard of caffeine being purified by subliming it on to a cold finger. I have tried caffeine extractions before from the hole bean which kinda works, but you get a super concentrated coffe tar. This sorta reminds me of the time I tried to extract theobromine from cocoa powder. I think I was using toluene and DCM as the solvent, but I neglected to take into account how much natural oil is in the cocoa. So I ended up with a bunch of waxy solids and thought I must have done an amazing job. But after tasting it I realized it was mostly cocoa butter.

Oh also in the US you can find "maximum strength" caffeine pills which are 200mg ea, and if I remember correctly is about 5.00$ for 200 of them.

Good job again I really like your posts.

entropy51 - 10-7-2010 at 10:58

Now use your filter flask (fitted with a cold finger) and your pump to sublime the stuff and it will be a satisfying white color. You've done a nice job, so go ahead and finish it up!

peach - 10-7-2010 at 12:20

I'm going to get a cold finger, but I've just spent a fortune buying some more gear, so it's on the list of things to come for now; I bought two vacuum pumps last week, then spent £60 on shot blasting and paint, then another £60 on filtration glass and a pig receiver, then another £80+ on flasks, £50 on another condenser and flasks... And there are more bills on the way. I need to get my equalizing funnel repaired, which will be another £45. There's more.

If I find a pack of Pro-Plus, I may try the MIT sublimation method using a petri dish. Or balance a round bottom flask a top a beaker and fill it with ice.

I suppose I could probably use my coil condenser and then rinse it back off.

I can also use test tubes to effect a cold finger in a flask.

Need the pack of Pro-Plus as well. I think there's one somewhere around the house. But it's too expensive to run it just for fun, so I need the free pack first.

I'd like to see someone try this with an entire jar of own brand coffee and redo the recrystalization a few times perhaps.

Remarkably, after all that solvent pounding, there was still a coffee smell right at the end. Maybe some of the aromatics are coming over with the caffeine.

I pulled some lithium out of cells last night and the alcohol is destined for some nifty manipulation if things work out. I don't want to hint at it though because, if it works well, it'll be funnier to see the posts.

I'd do nicotine but for the mondo tax on tobacco and low yield.

As I say, I am running more complex work, I've just been trying to keep myself busy extracting or making something each day, regardless of how complex or simple it is. Previously, I'd go through bursts of energy and overwhelm myself with things to prepare or try. Doing a little each day seems to work exponentially better.

[Edited on 10-7-2010 by peach]

Lambda-Eyde - 10-7-2010 at 13:23

Nice job. This is a very good experiment for beginners, since it involves several different lab techniques.

Though, if you just want pure caffeine powder, eBay is the way to go.



[Edited on 10-7-2010 by Lambda-Eyde]

bbartlog - 10-7-2010 at 15:26

We did something similar in my AP Chemistry class back in 12th grade. I remember getting seriously crappy yields from the black tea we used, something like 15mg from a quart of tea, but IIRC about half the groups got nothing at all. High schoolers: even the good students aren't going to have great technique. Might also have been issues with the available solvents.

Anyway, nice writeup. Makes me want to try this (maybe using chloroform instead of DCM) once I get moved. My lab has been packed up and won't be back in action for a month at least.

Satan - 11-7-2010 at 10:10

Extracting few times with small amount of solvent instead of one time with big amount might improve your yield, read something on extraction theory.


peach - 12-7-2010 at 03:04

I know, I was just doing this quickly

I wish the edit feature stayed on the posts, I didn't add that to the improvements

[Edited on 12-7-2010 by peach]

Hexavalent - 2-3-2012 at 13:58

Satan - You get more product by doing several mini-extractions as opposed to one big extraction due to solubility equilibrium.

Bot0nist - 2-3-2012 at 16:47

Quote: Originally posted by Hexavalent  
Satan - You get more product by doing several mini-extractions as opposed to one big extraction due to solubility equilibrium.


That's what Satan said, I believe. See:

Quote: Originally posted by Satan  
Extracting few times with small amount of solvent instead of one time with big amount might improve your yield, read something on extraction theory.


bahamuth - 3-3-2012 at 03:23

Quote: Originally posted by entropy51  
Now use your filter flask (fitted with a cold finger) and your pump to sublime the stuff and it will be a satisfying white color. You've done a nice job, so go ahead and finish it up!

Quote: Originally posted by peach  
I'm going to get a cold finger,


Several years ago I did a similar extraction of caffeine but from regular black tea (bags), but first boiling the tea "powder" in water with sodium carbonate to basify the caffeine. DCM was the solvent of choice but we spun (centrifuged) the extraction mix to induce a faster and cleaner phase separation (though messy compared to a sep funnel), this was the first time I got intoxicated by a halogenated solvent... The DCM was evaporated in open air and the crude caffeine was sublimated in just such an apparatus that entropy51 describes and I got a clean white powder.

Edit: Black tea contains a suprising amount of caffeine, got 180mg from four bags of black tea.

Edit: Presenting the sublimation apparatus.

The apparatus parts:

Composed of, set of rubber funnel-to-flask sleeves/adapters, filter flask, and a standard ~1 inch test tube pre-taped with several rounds of regular office tape to create a "bulge" to avoid shooting the tube into the flask when vacuum is applied.


The assemble apparatus:

Transfer the crude sublimable compound to the filter flask, put the coldest "stuff" you get your hands on in the test tube (shown in blue), apply highest attainable vacuum and set the flask in a sand/oil bath and watch as the sublimate forms on the test tube.


This simple setup works but care must be taken as not to allow the tube to be sucked into the flask, more tape the better. A test "run" might be a smart thing to do beforehand.

Edit:
The sublimated caffeine from all those years ago:


[Edited on 3-3-2012 by bahamuth]

Bot0nist - 3-3-2012 at 06:31

Nice!

"180mg from four bags of black tea."

That's a pretty small scale extraction. I'm impressed by the yield. Proper apparatus and solid technique, it would seem, are very important to isolating compounds from small amounts of substrate.

bahamuth - 3-3-2012 at 07:46

I would "blame" the good yield on boiling the powdered tea in the Na2CO3/water for several minutes.

The tea was Lipton Yellow Label IIRC.

Not that I need the 180mg from that batch as I came over 1.5kg after cleaning out a apothecary school lab at the same school I did the extraction. Got lots of nice old stuff from there and actually regret not taking more of the different chemicals now that I know alot more of compounds and that one never can get a too varied stock. .

Another remarkable thing about caffeine is the bitterness of its citrate salt, alot worse than any bitterant I've ever tasted, and as a plus one can actually feel the caffeine onset in seconds after sampling a small amount. Read somewhere that they "revive" babies that are half-dead upon delivery with caffeine citrate.

sargent1015 - 7-11-2012 at 20:20

I've always wanted to run this experiment with espresso, ever since my analytical lab when we analyzed it via HPLC. I am currently drying it down now, I'll let you know the progress when its done