Unfortunately, I have no way to be sure of how these powders were prepared. Wish I did.

I still have one more powder to try, just looking at the paste, it's definitely a different cactus than the others--using the same ingredient proportions as prior runs, this one mixed up to a tough, rubbery consistency, and stuck to the spatula aggressively as I mixed. Really made my arm hurt. Not sure if more water will fix it, but at the risk of introducing more variables than I can deal with, I'm going to carry on with the next steps, hoping it'll loosen up when the EA is introduced.

For comparison to any others who may report back on results with whole cactus powder and drying, I've been using Na2CO3. I may not have allowed it enough time to completely dry the solvent, so I can try leaving it for longer, or switching to K2CO3 if Na2CO3 isn't a good substitute for it. I also have molecular sieves now, so I could try that out as well.

I do see the benefit of drying, even if my procedure resulted in incomplete drying. With fumaric acid, the extra step gave bigger crystals than when I left it out.

It does seem that what I'm observing with citric acid is consistent with the formation of a biphasic system and ultimately a mixture of acid and salt like Nemoo described. With the addition of CA, my solvent partitions into layers, and the oil forms in the lower layer, almost exclusively near to where crystals of solid CA were present. Adding a quantity of CA calculated to be slightly in excess of the yields I got from my successful run with fumaric acid also produced oil, in that case, the oil formed exclusively around where the pile of citric acid had been before it dissolved.

I'll put extra effort into improving my drying process and give dissolved citric acid a try as well. Thanks Nemoo for the suggestions

I updated the goo FAQ section of the TEK. Let me know if I got anything wrong or someone disagrees. Thanks.

I think I have some strong support for the citric acid/citrate mixture hypothesis here:

I had a jar of citrate oil from a week or two ago, I discarded the extraction solvent into my waste jar and added a small amount of clean solvent, capped it and forgot about it.

I just rediscovered it, and found crystals beginning to form out of the oil! Even more convincing, the solvent is acidic. Seems that once the oil is formed, it takes quite a while to diffuse back out into solution, so a quick wash won't cut it, but it eventually does move out of the precipitate and crystals can form.

Just a note: the table I'm working on is very beat up, what looks like particles in the solvent are just chips in the laminate below.

Extracted the third powder yesterday evening. Went better than expected! I was worried on account of the paste turning out so rubbery, I swear it would've bounced back if dropped. The first two pulls required some extra attention to break apart the paste, but it did become the expected consistency after that. The resultant extract was much, much lighter in color than with the other two were. A bright lime green vs an almost black forest green for the other pastes.

I proceeded with the drying step as I had before in order not to introduce any additional variables, and reserved a small portion for salting with citric acid.

It crystallized this time! The solution still partitioned into two layers as it had in earlier runs, so I expected oil again, but this morning I was pleasantly surprised to find that crystals formed in that lower layer instead of oil. The crystals were all attached to the wall of the jar in a thin layer. It looked a lot like a lichtenberg figure, very pretty. They easily detached from the wall, so I don't expect any issues in collecting them.

The same solution salted with fumaric acid produced crystals within 10 minutes of introducing the acid and giving it a gentle stir. The stirring seems to have resulted in fluffier crystals, so I can't really eyeball the yield, but it appears to be a lot more than the other powders produced. Both solutions are completely clear at this point, will filter and weigh it tonight.

Thanks for the updates orchidist.

A possible interpretation is that some kind of paste sources are more difficult to crystalize with citric. Not sure why, perhaps the presence of different alkaloids that are more sensitive or less robust to the xtalization details.

Fumaric seems more robust so far and is giving you easier xtalization in general.

Perhaps we should consider reccomending fumaric over citric for the general case?

For anyone getting goo have you checked there's not a addative like an anti-caking agent added to the citric acid you're using?

Probably unlikely the issue but just a thought.

I'm very happy with the results I'm getting from fumaric acid, the only thing I've run into with it is that it seems to be far less soluble in EA than the citric acid is, at least when using solvent fresh out of the bottle. Using the 30mg/mL figure posted earlier as a guide, I attempted to dissolve 500 mg of it in 20mL of EA. Even with heating and stirring, most didn't dissolve. I've had it sitting on a stir plate (no heat) for almost 48 hours and incrementally increased the volume to 100mL over that time, and still not all of the acid has dissolved. Due to that, I'd recommend more care in measuring the quantity of FA to avoid adding much more than is needed to salt the expected yield.

I'm still going to continue small scale trials with citric acid. On Nemoo's recommendation, I ordered some K2CO3 after a quick search indicated that unlike Na2CO3, it is deliquescent, so it'll dry the solvent much more effectively.



A good thought! The citric acid I've been using (Milliard brand) is the one recommended by the TEK. It is labeled as pure and anhydrous, citric acid as the only listed ingredient. Under the suspicion that my original one might have absorbed moisture from the air while sitting dubiously sealed in a cabinet for months, I ordered a new package of it after the first few failed attempts. Unfortunately it didn't fix the issue. As an extra check for insoluble impurities, I dissolved the acid in water and EA but found both had dissolved cleanly.

Yes, fumaric acid is know to be less soluble than citric giving a smaller process window. That is a reason why we went with citric (+ we never saw an issue that was unique to citric).

However, the fumaric acid process window should be plenty big.

30mg/ml is definitely overkill for fumaric. I think 10mg/ml was ok if I recall correctly. Once saturated, you can dry a volume of EA and see how much fumaric you recover to get a solubility estimate.

Loveall, I'll work on solubility data soon. I was very conservative with my fumaric on this run, only using 500mg for this jar. Theoretically, only ~200mg of that should have been consumed by the neutralization of the freebase.

I filtered the portion of FA salted solvent and calculated yield. Because more crystals formed in the citric portion after I scratched the crystals that formed yesterday, I'm going to leave that jar for another day in case more grow.

Cactus: 60g
Volume of collected solvent: ~550mL
Portion reserved for citric acid: 50mL

I calculated these yield percents based on the entire cactus mass, ignoring whatever is left in the citrate jar, so these represent a low estimate for yield from this material:

Minimum mass (%) as fumarate: 560 mg (0.93%)
Minimum mass (%) as freebase: 361 mg (0.60%)
Minimum mass (%) as citrate: 690mg (1.15%)

Using a scaling factor of 550mL/500mL, I calculated the following extrapolations to predict my yield had I not set aside any extract to test the citrate:

Extrapolated mass (%) as fumarate: 616 mg (1.02%)
Extrapolated mass (%) as freebase: 397 mg (0.66%)
Extrapolated mass (%) as citrate: 759 mg (1.26%)

I guess I know which material I'm working with going forward! Definitely breaking out the stand mixer for it though. This stuff is TOUGH.

So to recap, one of your cactus sources xtalized with both fumaric and citric. Other source needed fumaric, but citric started working after decanting to keep goo and adding fresh EA to that.

Also, drying the EA made no difference.

So.... Looks like some cacti are more difficult to xtalize, but fumaric is more robust than citric in your research sample.

Is that correct?

I tried three different cactus products (San Pedro, grown in Peru, Peruvian Torch, location unknown, and San Pedro grown in Ecuador), only the last gave the expected result with citric acid, all crystallized quickly with fumaric acid.

For the others, drying with sodium carbonate did not prevent it forming an oil. I plan to try again with potassium carbonate, which should be more effective at removing water. Drying with sodium carbonate seems to be beneficial for getting larger fumaric acid crystals, and may improve yield.

And yes, the oil formed crystals after an extended period under fresh EA. When I checked the pH of that solution, it was acidic, suggesting that excess citric acid is a major component of the oil, as Nemoo suggested.

Thanks for the other updates/confirmation.

Regarding the pH and excess acid, how do you know the pH strip reading is not from excess citric in residual EA above the oil? I thought you didn't decant it completely off?

Sorry, I wasn't completely clear--I did decant it completely. The remaining oil was washed with a small portion of solvent, which was also decanted and discarded, then I added more solvent and let it sit untouched for a week or two. There is a picture of it with the red pH strip a few posts back

That's great info. Like we have been saying, your results suggest that some cacti could have issues crystalizing with citric but not fumaric.

I added the info to the TEK. If someone repeats your result (fumaric gives xtals where citric gives goo) I think we should update the TEK to suggest fumaric instead.

Glad I could contribute, I'm taking a look at the diffs on the wiki now.

On suggestion 2 of the "goo, what do I do?" answer, is Tarak execution a typo?

For suggestion 3, I think the amount of time it took sitting under fresh EA for crystals to form is worth mentioning. I used approximately ~50ml and then added another ~50 when I discovered the crystals. It will probably take a few days, at least, to diffuse enough citric acid out to start seeing crystal formation, then a few more to go to completion, and it may take more than one change of solvent to completely convert the oil to crystals. At the time I reported it, there was still oil present alongside the crystals.

For suggestion 4, I found no change to the oil from exposing the oil to fumaric acid. I would change this to suggest that if oil forms repeatedly, try substituting the citric acid with fumaric acid in a new run through the process.

Do you mind if I make those changes to suggestions 3 and 4 on the wiki?

Yes, please go ahead and edit the Wiki. Thanks!

First, I think it would be wise for all interested in making crystals to do some basic googling to understand the crystallization process.

Second, after getting a grasp on how crystallization works, do some further googling on Liquid-Liquid Phase Separation (LLPS), aka “oiling out”, aka goo formation.

The problems that orchidist has encountered are NOT likely to be source material related. Without naming names, I believe he and I have sourced our “incense” from the same vendor. I have extracted all 3 product offerings from that vendor, from multiple batches, and have never developed goo at the salting step, and believe me, I have tried.

Rather, I believe that there is something happening in his extraction process that is interfering with the nucleation step of crystallization and causing the creation of a second liquid phase… in short, something in his crystallization process - not his material - is causing the oiling out.

I suggest you focus on dealing with this rather than changing the ingredients. This is a common problem in crystallization, particularly pharmaceutical crystallization, and I am greatly surprised that the chemical experts on this site have not dealt with the prevention and/or mitigation of this issue.

Good luck in your research and progress on dealing with oiling out. This is a great and solid tek, don’t screw it up!

Sorry for the slow reply.
Im not new here or to these processes, just wanted to give my encouragement and advice to the thread. I won't be participating in any experiments.

I actually did read 99% of the thread, so I know some things have been tried. It was from seeing everyone working hard to develop a method that got me to reply.

Ive not used Na2CO3 as a drying agent and it’s not listed in my reference guides as a drying agent. I’ve seen (from reading the thread) that many people don’t need to dry the solvent to get crystals, but I also saw they got continued precipitation of the alkaloid salts for days after adding the acid, and some people got goo. Drying solvents is just a good habit, and no need to worry about water for crystallisation, there will likely be plenty left for that in there. I think, if you try a drying agent that is recommended for the solvent and alkaloid, that you will find a much greater difference than trying to use CaCl2. Try K2CO3 and make sure you give it time to work- up to 30mins. The precipitation should result in clouding (crystals precipitating), the fine crystals then growing in size as they fall, and yes it can take time but shouldn't take days.

I did see that fumaric acid gave crystals, I actually think its probably a better acid choice than citric which is appreciably more polar and more soluble in your ethyl acetate (and therefore would likely act more strongly on salting out not just alkaloids but water from the ethyl acetate, hence goo). You don't need much to precipitate the alkaloids in the batches you're running, dissolve it in ethyl acetate and add that so you don't add too much. Also the beauty of gassing with HCl or using insolubility to crash out the alkaloid salts is that you don't overshoot the end point of the titration, this has major benefits for purity that are hard to achieve with a aqueous acid.

Re goo: Decanting the solvent from the goo, then adding fresh ethyl acetate and grinding/stirring the goo in the fresh solvent would be called triturating, the water and excess acid are soluble in the ethyl acetate, the alkaloid salt is not, the grinding action and fresh solvent pulls the impurities and allows the alkaloid salts to crystallise. Acetone would probably work better.

Re the reuse of the solvent:
Vogel’s mentions different grades of ethyl acetate, the anhydrous is 99% and has a bp of 76-77C. Lower grades are contaminated with water, ethanol and acetic acid. To purify the lower grades of solvent he recommends adding 100ml of acetic anhydride and 10 drops of conc sulphuric acid to 1L of ethyl acetate and reflux for 4hrs then fractionally distill. The distillate is shaken with anhydrous K2CO3, filtered and redistilled from calcium hydride.

So that is what you should aim for, but I also realise that is unobtainable. As a minimum, washing the used ethyl acetate with d.H2O several times will remove some of the acetic acid, ethanol, citric acid, and other water soluble impurities. The ethyl acetate should then be dried with anhydrous K2CO3 to remove water, acetic acid, excess citric acid and then filtered- this drying may need to be repeated. If one was able to distill the solvent, then distillation and drying with K2CO3 would probably be sufficient and better than the first idea. A still can just be a metal can with copper pipe on an electric hotplate in a well ventilated area.

On the issue of method validation, if you wanted to show its efficiency you would do the following.
1. Take dried cactus material that did not contain mescaline and add x grams of mescaline to it as a solution, then dry powder again.
2. Use said material in your procedure and report the amount reclaimed.

90% reclaimed is the best Ive seen in the literature on mescaline, though that wasn’t crystals-> crystals but crystals-> extract-> HPLC. This would allow fine tuning and development.

Also pH=-log[H+]
Acids won't dissociate the same in an organic solvent as they would in water, so its kind of a tricky metric- though I do think your pH paper test is a valid test for the presence of a base or acid dissolved in the ethyl acetate.

I understand someone giving advice is second to someone posting results, and can be unwanted. but like I said, Im not going to do that so this was the best I have to offer- good luck!

Cheelin, all this time I've been under the impression you were working with personally grown material. That's good to know, and I'm by no means married to the idea that the cactus is the problem, or any other-- I hoped that my habit of never speaking in absolutes would convey that. My most recent runs do suggest that you're right in some respects.

For the vast majority of my trials I've used the basic 1:3 cactus:water ratio, by mass. One of my earlier trials with reducing the ratio for a drier paste led to no extraction at all. So I've been reluctant to deviate from it. The three cacti came to different textures using that recipe:

Cactus A, San Pedro "green dye": spreadable, mashed potato consistency, this is the one where I reduced water and nothing extracted.
Cactus B, Peruvian Torch "incense": slightly rubbery dough, crumbly on mixing. I was more confident in the texture of this one, yet still saw phase separation with citric acid.
Cactus C, Ecuador San Pedro "incense": literal rubber, my first impression was that this was too dry and would not extract at all, but was the only one to produce the desired result.

None of these pastes resulted in an emulsion, so there was no early warning that the water content was out of range until the salting step failed. I think that suggests that a qualitative description of what makes a working paste is not sufficient to give consistent results for all users and all plant material. Considering the cactus powder, and the lack of follow up from some who reported issues with crystallization, it's likely many first time users of the TEK will give up after a single failure. Anything within reason of cost and availability that broadens the process window will prevent newcomers from giving up. That can't be a bad thing if widespread adoption is the goal. With a first time success, even if yields are less than optimal, it would allow for more freedom to optimize for a particular material in future runs.

The advice is much appreciated Nemoo, K2CO3 allowed me to crystallize citrate from cactus A using the normal 1:3 ratio. With fumaric acid, crystals also formed, but it's taking longer than usual. Initially, a suspension of small sparkly crystals appeared, and are settling on the bottom as a powder instead of larger needles. That may not be a disadvantage as long as it's not too fine for a coffee filter to handle (better filter paper is pretty easy to get online though)

1. The 3:1 water ratio is a guideline not a commandment handed down from god, it depends on your powder, as loveall and i have said throughout. Attempts have been made to provide a quantitative characterization of the paste, no luck so far, give it your best shot.

Rather, i have stated numerous times, the water needs to be adjusted so that the resulting paste is cookie-dough to playdough consistency, so that the pulled material looks spongy to applesauce-like.

Your rubber paste indicates that you used too much water, and therefore lost yield that would have boosted your results.

2. None of this has anything to do with the goo, however. That problem is completely related to your salting conditions and process.

One approach that may help all who find a second liquid layer forming during salting is to heat up the solvent (no sparks or open fleme) and stir to dissolve the goo back into solution, then let cool. The problem is that crystal nucleation was disrupted somehow, and further crystallization stalled; so try to restart the process, and add additional steps, as necessary.

Additional things to try, would be stirring to disperse M density into smaller aggregates, seeding to initiate nucleation, and/or fresh solvent substitution to reduce M concentration (too much solvent will be bad as well). Again, read up on the thermodynamics of crystallization and oiling out. The ACS has a great and free paper which visually describes the crystallization process, very helpful in understanding the key steps in crystal formation, and where this goo problem occurs ( https://pubs.acs.org/doi....1021/acscentsci.8b00289 ). Lots of great papers on oiling out, of course the info needs to be kitchen customized, but that is doable.

3. It is not very helpful to have drive-by contributions, by those who evidently don’t understand the difference between trained chemists and kitchen chemists. I would go as far as saying, keep your thoughts to yourself, if you are unwilling to demonstrate that your suggestions offer significant advantages to those who are your audience. For example, it is a complete oversimplification, distraction, and near-malpractice to propose using a pot and copper tube to do fractional distillation of EA in a kitchen lab. Not only is it a safety issue, but the azeotropes have such close boiling points that it would be almost impossible to gain any advantage by non-chemists, from this technique, with the quantities of solvent that we are dealing with. The simple tek that Loveall has proposed, and i worked on, does the job safely, adequately, and with good purity. Buy fresh EA if you want slightly more pure product.