A guy visiting my neighbour told me about his cousin's friend who was fiddling around with this technique. I think he wanted to see if pre-dosing limonene extracts with a bit of acetone would reduce precipitation of fumaric acid in the process. Here is an excerpt from his practical report:

"Roughly 5l of 3 or 4 year old MHRB basic slurry which had already been 3x naphtha extracted was[...] extracted with a relatively small amount of limonene, which acquired a strong yellow colour originally having been colourless. The limonene was kept and the basic slurry was discarded. The limonene was then put in a tightly stoppered bottle and stored for a further 2 1/2 years.

To the approximately 225ml limonene was added about 2g anhydrous sodium carbonate. This was swirled, left to settle and decanted from the sediment.

Anhydrous acetone was prepared using anhydrous magnesium sulfate.

'FASA' was prepared by adding 100ml anhydrous acetone to 0.63g fumaric acid and swirling to dissolve with occasional heating on a hot water bath.

The limonene MHRB extract was decanted once more from any further sediment and 25ml anhydrous acetone was added.

A 100ml portion of this 10% acetone in limonene MHRB extract was transfered to a conical flask and part of a 1ml portion of FASA was dripped carefully into the limonene/acetone/MHRB extract. A cloudiness was immediately apparent and the addition of FASA was continued, now with swirling of the L/A/MHRBE.

Further 1ml portions of FASA were added dropwise, with careful observation of the formation of cloudiness and spot checking of the pH of the L/A/MHRBE on moistened, narrow range pH 6-8 indicator paper. When the clouding appeared to be greatly diminished, and the pH was found to be about 7, the addition was stopped.

The remaining approximately 150ml L/A/MHRBE was added to the flask. The addition of FASA was continued as before until the solution was again brought to pH 7.

By this point it had become increasingly difficult to observe formation of any further cloudiness so the solution was allowed to settle for about 90 minutes. One further drop of FASA was added and very little cloudiness was observed to form. The flask was then covered and allowed to settle for 24 hours.

The precipitate was very fine and after this time the solution was still highly turbid. The upper 80-90% of the solution was less turbid than the lower remainder, with a fairly noticeable line demarcating the two zones.

It may be inferred that either two quite distinct particle size ranges or particles of two different densities have formed."

It's a shame no controls were tried, without the added acetone or with a larger amount of acetone. That's as much as I know, although I think he wanted to see if there would be a precipitate rather than simply washing it with FASW or vinegar. The pH checks at least probably show the precipitate wasn't just fumaric acid.

Here is an update from the curious fellow:
"After a further 24hours the solution had cleared appreciably. There was still a light turbidity and the overall yellow coloration of the solution remained. At the bottom of the flask, the white, powdery precipitate had started to reform itself into very small crystalline clusters.

Small samples taken from both the surface and the middle of the solution showed a pH value of about 8 when placed on damp narrow range (6-8 ) pH test strips. Using a long Pasteur pipette a further 3ml of FASA was added slowly and carefully in 1ml portions to the surface and throughout the body of the solution, this time without swirling in order to avoid disturbing the crystals. The FASA became turbid as it was added to the limonene solution. The turbid acetone solution was observed to float forming a more turbid upper layer which over the course of several hours would diffuse downwards while the turbidity decreased.

During the addition of the FASA further pH checks were made as before until pH 7 was reached. At this point the FASA addition was stopped, the flask was tightly covered and left to settle.

A further 36 hours had passed when the pH was again tested and found to have a value of around 7.5. Whatever more turbid layer in the solution there may previously been was no longer distinct and the precipitate at the bottom of the flask had become more crystalline.

1ml FASA was added as before but this time in 1/2 ml portions. The pH was checked as before and FASA addition was again stopped at pH7. The FASA again became turbid and floated to form an upper more turbid layer which diffused downward and dissipated over the course of about 6 hours. The crystalline formations at the bottom of the flask continued to grow, by this time some of them having reached a diameter of about 4mm.

After this above mentioned 6 hours the pH was checked as before and found to be slightly alkaline. A further 1ml FASA was added as before with the same methods and observations. The flask was tightly covered and left to stand for another 18 hours, after which time another pH test was conducted. The pH of the solution was 6.5. The solution was still slightly turbid but much less so than at any point since the first addition of FASA. Some of the crystalline clusters at the bottom of the flask had grown to 7mm diameter."

Here is a photo of his flask:

Seems like its going well, then...

It just takes a few days to crystallize, I wait at least 62hrs for precipitation, sometimes up to a week. After three days or so I usually try to add more FASI/FASA to see if there is more cloudying.

Im not sure why the extra complication in your description regarding pH testing and changing of containers or what not.. Why not just keep adding FASA or FASI to the main container with your limonene? Also measuring pH won't work well considering there isn't water involved, if your acetone is dry and so is your limonene...

Yes, all seems to be going fine - and as endlessness said, pH measurements mean nothing as pH manifests only in aqueous solutions,

That's why he used pH paper moistened with water. So I'm told. I would reckon you can get at least something of an indication of the degree of neutralisation that way.

He also mentioned that the solution is clearing nicely and the crystals are very clearly visible now. The plan is to let it all settle, decant off the limonene and wash the crystals with some more anhydrous acetone. He'll then put the acetone wash into the decanted limonene, distil the acetone back out of the limonene and do a water wash on the limonene to see how much fumaric acid, if any, remains in there.

It sounds as though a FASW wash would have been much quicker and simpler, but there's nothing wrong with trying something in the name of science, I suppose! So unless his crystals turn out to be of extraordinary purity or something, I can think of no advantage in following this technique besides it being a great exercise in patience.

Seeing as the whole process was started some 6 or 7 years ago I don't think the guy is that bothered about how long things take! But it's probably safe to say this technique can be applied to fresh, brand new limonene extracts with a reasonable chance of success.

A quick recap would be:

1. Make limonene extract from basified source material.

2. Dry limonene with anhydrous sodium carbonate, settle and decant.

3. Add sufficient anhydrous acetone to make a 10% concentration in the limonene. This is done by adding 1 part acetone to 9 parts limonene, ie divide volume of limonene by 9 and add that amount of acetone. You may find adding more acetone works better, or you may find adding less would be sufficient. Experiment!

4. Add ANHYDROUS FASA slowly and carefully to the limonene/acetone solution. A long glass pipette is apparently very useful here. Take your time - 96 hours seems reasonable. Perhaps.

5. Consider spot checking with moistened pH paper if you have it. Alternatively, each time the solution has been allowed to clear between rounds of FASA addition, very carefully add one drop of FASA and observe whether there is any significant clouding. It's not clear how reliable this is - and pH paper is always a useful thing to have around anyhow.

*6. When the FASA addition is complete and the solution seems to be clear (check for turbidity by shining a thin beam of light through the solution. If it is visible side-on, there are still particles in suspension), you should have some beautiful crystals at the bottom of your flask. Very carefully decant the supernatant liquid, using a filter if there are any free-floating crystals.

*7. Wash the crystals 3 times with pure anhydrous acetone. Drain the washings into the decanted limonene (which already has acetone in anyhow), checking carefully for any further formation of precipitate. If there are still crystals in the bottom of the flask - congratulations! Allow the crystals to dry. Gentle heat will help to evaporate the acetone, but keep it away from flames and incandescent materials!

*7a. If there are no crystals, the experiment has been a complete failure. Sit down and consider how lucky you are not to have thrown anything away yet.

*8. (notwithstanding 7a, above) Perhaps it's time to consider an anhydrous IPA recrystallisation as that's about as easy as trying to scrape the crystals out from the bottom of the flask with a spoon. Or perhaps just use water to get them out and rebasify to recover alkaloidal freebase.

Steps marked * have not been tried yet and are therefore theoretical. But they'll work!

PS Sorry for the long post!

PPS Thanks for the replies, guys!

The changing of containers was initially done to test the process on a small portion of the solution just in case anything weird happened. He was going to precipitate the remaining 150ml in its original flask but then had to decant away from solid that settled out after decanting off from the main portion of the sodium carbonate, and he was running out of clean flasks at that point! Anyhow, the first 100ml was working fine so it seemed reasonable just to carry on and precipitate the lot in one go.

The sodium carbonate should probably have been left to settle for longer in the first place and the second decant was indeed an unnecessary complication to the process. I think he only mentioned it because it was an account of what had happened. Leaving your limonene pull in a stoppered bottle for 2 1/2 years is also unnecessary!

But, to reiterate, this is someone who is most clearly not concerned by taking his time over things!

The pH testing was partly out of experimental interest, to see if it would be any use to the process. If the limonene was causing precipitation of the fumaric acid it may have been unclear as to where the endpoint of the neutralisation actually was. He didn't want to assume his experimental technique would actually work, having devised it as a possible solution to the reported acid precipitation problem.

Also the pH test in part showed that, possibly, some fumaric acid was precipitating and either settling out or - hopefully - redissolving as it reacted with the remaining base in the solution. (Remember, the turbid solution floated on top at first...)

OK, here's a couple more pictures of the progress:

Update:

"After another 30 hours, the solution was observed to have substantially cleared, with but a faint turbidity remaining. A further test drop of FASA was added which caused formation of more turbidity, so a further 0.5 ml FASA was added slowly and carefully throughout the body of the solution using a long Pasteur pipette. This again floated to form a turbid upper layer.

12 hours later, the turbid layer had dissipated angd the solution was nearly clear as before, so a drop of FASA was added. Turbidity formed, so a 0.8 ml portion was added in the usual way, with the usual observations.

48 hours later, while inspecting the experiment, the flask got knocked over and all but 50 ml of the solution was spilled. Fortunately it appeared that most of the crystalline formations had stuck to the bottom of the flask and the remaining solvent had saved a significant part of any loose crystals.

An additional 30-40 ml of solvent was recovered from a fortuitously placed paper napkin which was then drained into a separate flask and tightly covered.

After a further 12 hours, the solution in the flask with the crystals had settled out and cleared completely so the liquid was carefully decanted off into another flask. After draining off any residual solvent, the crystals were washed with 3x 2 ml anhydrous acetone, the washings being added to the decanted limonene /acetone mixture.

There remained in the bottom of the flask a quantity of white, possibly hygroscopic crystals, some of which were coloured yellow from trapped solvent. Next these crystals will be recrystallised from boiling anhydrous isopropanol..."

More news:

"Dry [actually, it seems, not that dry at all] isopropanol was prepared by adding anhydrous magnesium sulphate to pharmaceutical grade 70% isopropanol, decanting into a boiling flask with anhydrous sodium carbonate and distilling.

The crystals were washed with a total of 5 ml room temperature freshly distilled isopropanol in 3 portions. The crystals almost completely dissolved as a result of this, so the flasked was thoroughly washed with a further 5 ml of freshly distilled isopropanol.

The 10 ml of washings had been transferred into a clean beaker and was a very faintly cloudy, yellow solution. The beaker was tightly covered with cling film and placed in the refrigerator for 12 hours. The solution became only slightly more cloudy. The beaker was then transferred to a freezer, at about -22 to -24 degC. After 48 hours there were no significant changes to the solution so it was removed from the freezer and allowed to return to room temperature."

It seems either the isopropanol was too damp, or the crystals were just fumaric acid, or possibly both. Further tests are required.

"After a couple of days of nothing happening, so the container of liquid was placed underneath an axial fan with the airflow going upwards, away from the container. After a few hours the liquid had evaporated down to a viscous, yellow oil with a strong smell of oranges due to the residual limonene. On standing for three more days a few very fine crystals became visible in a part of the container where most of the liquid had drained away."

So do you think he should either: add more water to dissolve any fumarate and allow the limonene to float off?

or: dry the goo in a chamber with anhydrous MgSO4 and wash off the limonene with anhydrous acetone?

Id probably mix it with sodium carbonate and a bit of water, let it dry and pull with acetone and add FASA to precipitate fumarates again. Or just redissolve in water and precipitate the freebase by adding a base, if you want the freebase instead of fumarates.

When washing fumarates with acetone or IPA, it has to be fully dry, and ideally have the solvents cold (put solvent in the freezer for a little while before). This way it still cleans but there's very little yield loss. Also do the wash quickly and filter.

"Two days later, the goo had started to set up as crystals."

I finally heard from this chap again after several weeks:

"The goo was dissolved in about 10 mL distilled water, transfered to a small volumetric flask and washed 3 times with 2 mL light petroleum ether in order to remove most of the limonene contamination.

"The aqueous solution was then returned to an evaporating dish and placed underneath an axial fan as before albeit this time with very gentle heating from a terrarium mat. The resulting goo was redissolved in 91% isopropanol and evaporated again in a similar manner, to remove most traces of petroleum ether.

"When the solution had again evaporated down to a nearly solid goo it was left to crystallise over the course of a week or so. Any larger solid masses that formed were carefully broken up so as to assist the drying process. Feathery crystal formations developed where the layer of goo was thinner.

"The crystals were scraped up and weighed. The yield was 590 mg. Presumptive tests carried out with Marquis reagent and Mandelin reagent gave colour changes corresponding to those seen for DMT in the literature - a dirty yellow with the former and a bright intense yellow with the latter - although this is by no means definitive, especially considering the crystals still had a faint orangey odour and pale yellow tinge. Limonene reacts with both of the above reagents to give a bright orange-yellow, which at lower concentrations could give a similar result to that described above."

I'm told the crystals actually appear much less yellow than in this photo.

So it seems like he may have been successful... what other presumptive tests are there that might distinguish DMT from, say, 2-Me-THBC?