...and Yopo...

...and a little San Pedro Extraction Technique...

The principle of alkaloid purification is to obtain from a plant only that fraction which is basic. This is achieved by a double layer extraction, relying on the principle that amines (as opposed to most of the other compounds in a plant) are soluble in acidic (the salt form) but insoluble in basic (the basic form) aqueous solution. However, the basic form is soluble in non-polar organic solvents whereas the acidic/salt form is not. Thus, by varying the pH, alkaloids can be taken from aqueous solution to organic solution or vice-versa, leaving behind other materials.

Some chemicals and equipment are important for successful extraction of alkaloids from cactus. The chemicals include methanol, dichloromethane or chloroform, sodium hydroxide and hydrochloric acid. The equipment includes a distillation apparatus, a separating funnel, and various beakers and containers, pH tester, and filter. Alternatives can be found for each of these.

Method:

1. Slice and dry the cactus. I haven't worked out the best way to do this; no matter how I do it, I am always afraid that I am destroying the alkaloids. In general, what seems to work is to slice it thinly, and run hot air over it overnight. The more water which can be removed from the cactus at this stage, the easier the process will be. [Another possibility might be to use a drying agent such as Calcium Chloride see the "Drying and Preservation" section of the Magic Mushroom Growers Guide for more info.]
2. Pulverise the dried cactus. I have tried using a blender, and it seems to work moderately well. The cactus is tough so you will have to be patient. The finer the grinding, the better the extraction.
3. Extract dried cactus with methanol. Ideally this is done hot using solvent-extraction apparatus (soxhlet). Various makeshift methods may suffice for a hot extraction, but I have generally merely soaked the stuff for up to a week, cold. Ideally this step should be done three times, and the extracts concentrated. I have done it once for a week, and then washed out the absorbed methanol with fresh methanol once or twice over an hour or two. What you should end up with, after filtering out the bulk of the cactus, is a green methanol extract. Ethanol or acetone could be substituted for methanol, but neither of these is quite as effective. It is generally desirable to use several times the weight of the dried cactus for the methanol extraction, or at least enough to cover it well in a container.
4. Remove the methanol to leave just an extract residue. This is best done using vacuum distillation, but can be done using atmospheric distillation, to recover the solvent. If you don't mind losing several litres of methanol, you can merely boil the stuff into the atmosphere; just avoid starting a fire. Always no matter what use boiling chips (porcelain) to promote even boiling. Methanol superboils easily, as I have found :-(. Once most of the methanol is removed, you will be left with a hundred ml or so of watery, methanoly, green slime. If it weren't for the methanol and the bad taste, this could be consumed at this point. In general, I would say that it may be worth your while going to the next stage if you can manage it.
5. Add dilute hydrochloric acid. Sulfuric acid, etc. could be used instead, but I like to use HCl and NaOH, because the product is NaCl, which is of no consequence if it contaminates anything. HCl is not as oxidising as H₂SO₄. The HCl should be less than 1M, but not weaker than 0.1M (pH 0-1). Add a few times the reduced volume of liquid - e.g. take the stuff to 400 ml from 100 ml, etc. One good idea is to let the bulk cactus residue (post methanol) dry, and then soak it for a few hours in the acid you are going to use to add at this point. This will extract the last of the alkaloids. Unfortunately, cactus being what it is, will swell enormously, and removing the HCl is tricky. I have resorted to large quantities of HCl and a kind of press to squeeze out the acid from the bulk residue. This acid should then be filtered, and added to the methanol extract residue as above.
6. (optional) The stuff at this point will be a bit of a mess. Adding activated charcoal and boiling gently for 10 minutes will help to congeal the chlorophyll etc. which is gumming up the stuff. Do not add too much charcoal - less than a gram should be plenty. Too much will absorb alkaloids. Don't use burnt wood, burnt toast, etc - get the proper stuff from the local pharmacist. Performing this step will make the next stages considerably easier.
7. Filter the HCl extract. This will remove a proportion of the gunge. This will be easier if charcoal was used. The more gunge that can be removed at this stage, the better. Washing the residue with fresh HCl before discarding, and adding this to the rest will ensure no loss of yield.
8. Carefully basify the HCl solution with NaOH. I tend to use around a 5M solution for this, which is OK as long as you stir as it reacts. Take it well above pH 7. You should get white clouds of alkaloids forming in the solution, and the whole will become turbid as some of the acid soluble components precipitate. Ammonia or KOH should work for this purpose as well. I have had some difficulty with ammonia not being quite basic enough in other systems.
9. Add dichloromethane (or chloroform); be generous with the quantities if possible. Ideally, one would like to extract into CH₂Cl₂ 3 times with equal volumes, but the amount of solvents gets huge. Ether is not all that good with mescaline extraction, I believe, even though it is easier to separate from water. CH₂Cl₂ is handy because it has a very low boiling point. It is at this point in the whole operation that the most care and patience is necessary. A separating funnel is really a must - one could plausibly separate the layers with a very tall thin jar and a syringe, but this would be difficult. Ideally, the basic solution and the CH₂Cl₂ will separate into 2 nice layers, the lower one (organic) containing the alkaloids. Unfortunately, while this is not difficult with most plants, it is very difficult with cactus extracts because the cacti contain so much resinous junk and natural surfactents (to retain water). The best way I have found to separate the layers once you have shaken them together is to add plenty of salt (NaCl) to the water/base layer. This is excellent for breaking the emulsions which form. Be prepared to use large quantities of salt.
10. Separate dichloromethane layer from mixture and put aside. Repeat steps 9-10 a few times: once is insufficient, three is good, four is excessive. Combine all the dichloromethane extracts together. This should be a slightly green solution. It will contain a bit of water, most likely.
11. Backwash the dichloromethane once with a solution of salt and NaOH (dilute). This will clean up the last of the junk from the organic solution. Separate the layers as before and discard all aqueous material.
12. Distill off the dichloromethane (or allow to escape to the atmosphere if you are rich and don't like the ozone layer). I have found that once you are down to maybe 20 ml of residue, the best option is to place the remainder in a petri dish (or some flat dish you are going to store it on) and hitting it with a hairdryer to remove any last CH₂Cl₂ and water. You should be left with a small quantity of moderately pure alkaloids. This can be easily consumed by dissolving in vodka, e.g., or should be stable for extended periods if refrigerated, frozen, kept airtight and away from moisture. Do not expect more than a 50% yield the first time you try this: theoretically if everything is done properly, the yield should approach 100%, but this is rarely the case.

I've had a quick scan through that post, and it seems he's been pretty complete in his description. I can add the following information:

• The initial extraction is very important. You only have as much alkaloid to play with as you extract at this point. Similarly, the more gunk you extract at this stage, the more you'll have to eliminate. Methanol is used because it's not too polar or non-polar, and it penetrates cell walls and membranes quite effectively. Acetone would be as effective, but it seems to pull much more gunk out along with it, and for this reason should probably be avoided. Also note that the better your mechanical mulching process is, the more effective the extraction will be. Also, note that a hot solvent extraction under reflux conditions, or using a soxhlet would be more effective, but bear in mind that some alkaloids may decompose with heat - it's probably worthwhile checking the physical characteristics of the alkaloid in question in your handy Merck index or CRC handbook.
• With respect to the degree of acidity and basicity necessary for the extraction, this can only really be found by trial and error. Insufficient difference from neutrality will cause insufficient separation, extremes of pH (particulary too basic) may cause alkaloids to be degraded or other nasty things to happen. Also note that other compounds present may act as buffers - be sure to stir well, and measure the pH using indicator paper or somesuch.
• The final product is a free base. Usually it's unneccessary to go further than this. Forming a salt can be tricky, and would also be wasteful. If you're ingesting the product orally, then it'll salt adequately with the HCl in your stomach. If you're concerned with oxidation by the air, well, use your better judgement (are you really going to wait that long before eating the stuff). Of course, if it's DMT and you're going to smoke it, forming the salt is a waste of time.

Fresh cacti, w/ a water content of 90-95%, are difficult to handle; the juice is not easily separated from the fibers, and the presence of mucilaginous substances causes trouble during extraction. Thus, it is preferable to dry the plants immediately after they are collected, in order to avoid losses and deterioration. The plants should be cleaned, the spines extracted with nippers, and the remaining material cut in thin slices and dried in the sun or better in a low-temperature drier (40-60oC). The dried substance is easy to grind, and the powder can be stored if neccessary.

The powdered material (10kg) was exhaustively extracted in a percolator with hot, acidulated [0.3% acetic acid] 96% ethanol, the extract filtered and concentrated in vacuo to 500 mL. water was added (2.5L) and concentrated again to 2.3L. The solution was filtered to separate oils, resins, chlorophyll, etc and washed twice w/500mL of chloroform. The aqueous solution was neutralized with solid potassium carbonate, and an excess of 20g was added. A precipitate containing only calcium carbonate was eliminated, and the solution extracted four times with the same volume of cholorform (2.3L). The combined [chloroform] extracts, after having been dried [allowed to stand over] with anhydrous sodium sulfate were concentrated to 780mL and filtered. The solution was extracted twice with 150 mL of 3% aqueous HCl and once with 100 mL water. The aqueous solution was washed with ether repeatedly and filtered. This is the purified aqueous solution of total alkaloids. [This would be a good place to slowly evaporate the water and consume the residue, for further purification of the alkaoids, see below...By the time all the water is removed, there should be no excess HCl around.]

This soln. is alkalinized agin with dry potassium carbonate an extracted 5 times with 500 mL of ether. This extract, containing the ether soluable alkaloids, was concentrated to 500 mL. this is A (trichocereine) [this is mostly the more lipid soluable, inactive, dimethyl-mescaline, aka trichocercine]

The aqueous solution [still basic] is extracted three times with 500 mL chloroform, and the extract concentrated to 200 mL; this is soln B (mescaline) [this could be taken to dryness to give mescaline base, probably mostly pure]

Soln B was extracted twice with 40 mL of 5% aq. HCl and once with 20 mL water. The aqueous extract was then washed with 20 mL chloroform and 20 mL ether. After alkalinization with potassium carbonate, the solution was extracted with chloroform and the [chloroform] extract evaporated to dryness. It weighed 4g. After neutralization with alcholic hydrochloric acid, the crude hydrochloide was dissolved in 30 mL. of absolute [ethyl] alcohol and the filtered solution overlayed with 30 mL of ether. The hydrchloride crystallized. re-xtalized from alcohol, the pure salt melted and 181-182oC.

L. Reti Journal of the American Chemical Society, 1951, p 1767

This document from the Journal of the American Chemical Society