New questions have been a little light lately.  I want to share a correspondence I had due to Ask the Alchemist #211.  Because I have something planned Ask the Alchemist #212 (is it obvious to everyone what the subject will be?) I’m going to call this one 211.5.

I really enjoy these kinds of interactions and conversations.  It try really hard to walk the line between too simple and too complex with a mind to not mansplainng.  My hope is that if you are coming here you are doing so for me explaining things and realize it is to a broad audience and what is over the head to one person is missing too many details for another and yet to another it is downright condescending. 

In this particular case I got the fact wrong that semi-suitable solvents were not readily available.  I’ll own that.  They are.  I will also remind everyone that I was passing on what I did in the past and in hindsight there are things that could have been better and would do differently now.


What you don't have a huge supercritical fluid decaffeinating machine in your garage?....
BTW eBay is great resource for slighly used lab equipment, like teflon stirbars and hotplates at almost reasonable prices.
Here in Colorado, the street solvent of choice to extract THC from herb is plain 'ol butane. Chilled in the freezer, it is a liquid. Big advantage is ease of evaporation, BUT big downside is extreme volatility and FLAMMABILITY. Not easy to avoid filling your enclosed space with an explosive mixture of air-fuel if no chemical exhaust hoods to work in.
I was paid to develop some methods to do just this for various seed oils that some local cosmetic companies wanted to extract themselves from raw and toasted oilseeds. They needed non lab grade solvents to do it and they are available at low cost from your auto parts store and Home Depot hardware stores.
Here is the cross to hardware grade solvents. 
Everclear 200  aka Ethyl Alcohol absolute USP, cross to denatured alcohol with 5% methanol as denaturant. Works great less cost no paperwork to obtain, 7 bucks a Qt.
Petroleum ether is the petroleum fraction consisting of aliphatic hydrocarbons and boiling in the range 35‒60 °C; commonly used as a laboratory solvent. Petroleum Ether (Ligroin), ACS, 35-60°C, 1L (32oz) not sold at hardware stores but see white gas and Coleman fuel below.
Hexanes, HPLC grade, cross to paint thinner,AKA light Naphtha, hard to find in US but White Gas or Coleman Fuel is a pretty good substitute b.p. 80-120 C Coleman fuel is a mixture of cyclohexane, nonane, octane, heptane, and pentane.  10 bucks a Gal or less.
Paint Thinner deodorized aka Heavy Naphtha b.p. 150-200C   Klean-Strip 1 qt. Odorless Mineral Spirits 7 bucks a Qt

Real diethyl ether aka cold engine start fluid is 90% diethyl ether 5 bucks a spray can at auto parts stores nationwaide. 
MEK methyl ethyl ketone a low boiling ketone available at Home Depot in Qt cans for 12 bucks
My method used white gas to extract the oil fraction from ground toasted oil seeds, Weigh oil seeds, 100g aliquot x 3 and extract with hot (hot water bath hot only) paint thinner x 3 x 250 mL. Combine fractions evaporate on electric stirring hotplate with teflon stirbar OUTSIDE or in chemical exhaust hood to apparent dryness, weigh oil fraction. Evaporate oilseed dry mass to constant dry weight and add both numbers to comopare losses.
Percent oil = oilseed oil weight (grams) / 100( grams)
Error in method due to extraction of gums, phospholipids and other oil soluble non fatty acid triglycerides.
These methods work, just source the right solvents!


You are absolutely right about those solvents and the hardware store.  I have no clue where my brain was this morning.  I'd still hold that they are not going to be really safe to use and not something I'd recommend.

Thanks for keeping me honest and accurate.

Thanks for the reply. I re-read your post, picked up some of my extensive books on lipid chemistry and reviewed some methods. They reported methods that were mostly multi-step (ugh, no fun) with purification (s) thrown in for extra credit. Long story short, I believe your problems are due to using ethanol as a solvent. Hot ethanol is the solvent of choice to extract phospholipids and it does a fantastic job of extracting total phospholipids from any plant source. Phospholipids are notorious for gumming up the works. Think intractable emulsions and worse. Phospholipids will swell up in the presence of water to make those sticky thick gums and film forming emulsions that refuse to be broken even with copious sat. sodium chloride washes. Don't use ethanol as your oil seed solvent. Use straight chain hydrocarbons, Ligroin -heptane and longer should work great and evaporate at a reasonable rate. 
I would try the Coleman fuel aka white gas as my first solvent.  Gasoline is another very flammable and smelly option, as it boils lower than white gas but is full of impurities. Keep the cocoa beans dry before grinding them coarsely. I would use roasted beans and compare to raw cocoa bean results but roasted should grind and extract much more efficiently. Be sure to use hot water bath heated hot solvent aliquots to extract total oils, combine oil extraction aliquots after filtering through coffee filter papers and then try to evaporate to total dryness to recover your oil fraction. It should go pretty quick with heat and stirring. Do not wash the oil fraction with water at any time as phospholipids will also be extracted to some extent. You are doing a crude total oil fraction estimation and it should allow good comparison between cocoa beans from different plantations / geo areas / seasons.  Cocoa butter is the major commodity from cocoa beans and I think it is also available worldwide at spot pricing so the extraction of coco beans must be a highly efficient mega tonnage process performed worldwide. Oil seeds are generally processed by cold pressing for highest quality, then hot pressed, the solvent extracted using mostly hydrocarbons and spray thin film vacuum distillation to recover the hydrocarbon fraction
The major problem with using any higher boiling ligroin aka straight chain hydrocarbon as a solvent for triglyceride extraction is the difficulty in removing it without access to a vacuum and a nice expensive rotary evaporator (aka Buchi or Yamato rotary evaporator). Both cost big bucks new (5K dollars plus is typical) and they are NOT easily found used as they appear to be choice equipment  for the drug extraction crowd. I have spent days sitting in auctions trying to bid on the larger ones for prep labs and never come close as bids go up to 80% of new cost, even when old and trashed. So use a stir bar and electric hotplate stirrer to blow off the ligroin fraction and leave behind the cocoa butter solids. Should work at about 250  F for  overnight. An air pump bubbler stone would really help as would reduced pressure / vacuum aspirator but then the whole setup becomes complicated and unwieldy to use and store away when not in use. The idea is to make the process fast and efficient. Lipid chemistry is usually neither, cough cough.  But since you aren't recovering the cocoa butter fraction for cosmetics or as an edible oil, you can just apply excess heat to drive the hydrocarbons off quickly and maybe suffer a little cocoa butter burning in the process. It will still weigh the same in the end.
BTW should be obvious but NO OPEN FLAME anywhere while this is being done. 
I would guess the major problem with cocoa bean extraction of total lipids is the high lipid fraction and possibly the higher melting point of the lipid fraction. You think they are worse than macadamia nuts or other tropical nuts? I consider all oilseed extractions to be problematic in that phospholipids will always tend to gum things up. Soybeans always gave me fits...

I knew going in the alcohol would not work but had to do it for completeness so I could say I did it first hand.  Intractable emulsions is the perfect term.

I should mention these tests were years ago, not recently.  It is funny how memories go.  I wrote what I remembered and at that time in my experience I didn't think about obtaining MEK, toluene, naptha, etc from the hardware store because I had them in the lab and couldn't imagine they would make them accessible to the public.  I'm remembering now how aghast I was upon discovering I could buy them at The Home Depot.

And yes, the issue with higher bp hydrocarbons was indeed the need for vacuum extraction and at the lab I was in we only had access to low pressure vacuums to pull off vapors and nothing to actively pull a negative pressure.

Coleman fuel is a great idea from an accessibility and potential for success standpoint.  Yes, I think cocoa is worse than many nuts.  Funny you mention Brazil nuts.  I tried them as a semi-control and although a little fussy behaved basically ok.

Thanks for all the conversation.  BTW, I laughed at the supercritical fluid extractor comment.  I kept coming across that technique this week while researching various solubilities of cocoa butter.  It certainly says something about the matrix to need that.

Keep keeping me honest!

John Nanci

Chocolate Alchemy

Founding Alchemist


I still shudder at the thought of someone doing this in their kitchen and likewise evaporating the fuel into the air.  There are just so many safety concerns.  The mention of open flames is a great example.  Since this conversation I’ve had 3 others and one asked about doing this outside (for safety sake) with their propane grill.  Obviously that is a major issue.  Another asked about grinding the cocoa and solvent together in a blender.  Aside from the fact of making it unsuitable for food afterwards, many of the seals in home appliances may dissolve or degrade due to the solvent.  In the lab I used pyrex glassware with teflon seals where needed and an sonicator for assisting in the extraction.

There are some many items that are obvious to someone with lab training but not to everyone else.  Open flames and seals are just a couple.  Basically if you are going to try this, do your homework for both methodology, GLP (good laboratory practices) and safety.

With ebay and the like you have the ability now days to get appropriate equipment.  But you need to research what you need and how to use it properly and safely.  It is way beyond the scope here to try and lay it all out.

And again, I am going to remind you that even with good equipment, the appropriate methods do not have the best reproducibility and the subsequent data isn’t all that helpful.  But it can a fun exercise.

So have fun and be safe.