My previous attemps to separate harmine from DHH using sodium carbonate were unsuccessful, so I tried again with slightly different approach.

I prepared 300 ml of hot 0.25% HCl solution. While stirring, I added 5 grams of mixed harmala freebase that dissolved immediately. Some tiny amount of freebase stayed undissolved and I filtered that out. I assume all the HCl was consumed and the solution is just water and harmala hydrochlorides. The starting pH was around 2.5 so maybe there was some excess acid anyway.

I calibrated a precise pH meter using 2-point calibration (4.01 and 7.01 buffers) and put it on continous measurement.

Then I added 10% ammonia solution drop wise from a syringe until observing precipitation. The solution stayed cloudy even before the solution was neutralized!



I stopped adding base at a pH around 8.0:



The precipitate was filtered:



I returned the filtrate on a stirrer and added base until pH 8.75. Some clouding occurred but not much:



Filtered again:



I repeated the process, now adding excess ammonia until pH well over 10.0. No precipitation happened:



I have gone to overkill and added excess 50% NaOH solution. This caused the liquid to clear up and pushed the rest of the alkaloid from solution:



There was not much left on a filter. Here are the fractions side by side:



So another trial and I am still unable to separate harmaline from harmine. One possibility is that the seeds contain only harmine but every syrian rue analysis I've seen have roughly 1:1 harmine:harmaline content. And I have used even the same vendor of the seeds as the study successfully separating the two compound.

Do you have any ideas how to get the harmaline? No matter how careful I am, everything simply crashes out before pH 7.0.

Stop at pH 7.

PS: careful with sensitive pH probes, those don't like solutions with very high pH (over 12).

Is there any theoretical basis for that? One should get 92% harmine even at pH 8.75 yet everything crashes out before 8.00.

This means that even if I remove precipitate at pH 7.0, the rest will still precipitate between 7.0-8.0 and all that must still be harmine, right?

I found removing the alkaloids makes the solution more basic which is even weirder. It seems something happens and some equilibrium is pushed by mere act of removing the precipitate. Mayube that has positive effect on the further separation of harmaline?

I will read the whole VDS DHH separation thread (it is very long so I haven't had time yet). I hope there will be some answers.

BTW the Wiki needs to be updated. The information there is alarmingly outdated yet it is impossible to change that People need easier way to access high quality information than browsing dozens of threads...

In theory, theory and practice agree.

In practice they don't.

I think that there might be a good theoretical explanation for the observed behavior, namely precipitation influencing the equilibrium. The problem with the hitherto applied pKa theory is that it doesn't take solubilities into account, while these are very relevant to the effective behavior of the system. The additional consideration of solubility equilibria explains why precipitation starts at a very early point in the pKa curve and why the pH stays put during most of the precipitation window of each component.

Everything is logical when you use the right theory. The problem is that only practice will show what theories actually apply.

I know I know. The teks on wiki shows it works in practice too. The problem is that my practice produces different results than others' practice following the same protocol. The author of the wiki article even showed photos of separated harmine and harmaline so it really worked for him based on the simple pKa separation model which we now know is innacurate to the point of not working at all in many cases.

The 8.75 separation point appeared on several different places only solidifying my conviction it works. Different people seems to have success separating the two compounds simply throwing in excess sodium bicarbonate, filtering and then throwing in excess sodium carbonate. No careful pH measurements, no titration... yet it works for them and not for me which deepens the mystery.

The only explanation might be hidden parameters the author was not aware of. This only shows the importance of peer-review as this can reveal problems of the said tek.

Unfortunately, it's not the first time I found completely outdated and proven-wrong information on the Wiki. If only it would be possible to add notes to that...

Maybe and empiric experiment could be 2 or more people do the same experiement, with same chemicals, recipe, & source material and compare results ?
It's could be hard to do exctly the same recipe for 2 or more remote people, but is used the same ingredients & recipe, could it be better to compare than same material using different chemicals & recipe ?

That was my 2 cents,

Below my 1 cent,

As we're doing energetic manipulations ( as playing around ions, polarity, so energies) could the energy of a subject can add an another parameter to the recipe ?

I've just re-read the "A harm-reduction approach to the isolation of harmine and its hydrogenated derivatives..." paper and they show pH variations as well as correct pH-metry for the separation of the two compounds. Indeed it is way below 8.75, the harmine fraction is taken at pH 7.5, then mixed harmalas at 7.5-7.7 and finally DHH fraction over pH 8.5. The pH of the mother liquor fluctuates (drops when certain pH levels are reached) which needs to be taken into account.

This really should be in the Wiki instead of the outdated and misleading information. The current article is not chemically correct and even potentially causing harm

Just done this separation again, sticking closely to VDS protocol and it worked fine. First time I used 10% ammonia and added it too fast, missing the first pH depression for harmine. Using 1% ammonia, adding dropwise and stirring continuously allowed the first depression to be observed. This occurred at just over pH 7, with precipitate forming. Eventually pH started to rise steadily again and I filtered at 7.5.

Again, 1% ammonia was added dropwise until the second depression occurred at about 8.5. Kept adding base till pH rose steadily again till about 9.75 at which point the harmaline was filtered off.

So try going slow and steady with 1% ammonia and hopefully you should have more success.

Thanks. I wanted to use 5% ammonia but I will therefore prepare 1% ... it's always good to be on the safe side.

I will stick with the harmala/acid concentration used in the mentioned paper and then maybe experiment with higher concentrations and take notes of the stoichiometry involved to speed up the process next time.

Hopefully this will get distilled into a tek one day but it won't make it to the Wiki as the wiki seems dead

BTW till now I thought the "VDS protocol" is some special HPLC machine file format or something, but now I found the VDS is probably an abbreviation of "Van Der Sypt", the author of the paper.

This is why I never read the thread, I always thought it's some highly advanced/technical chromatography-related stuff.

Again, concentration-wise, I stuck to Van Der Sypt's ratios, which I cant remember off the top of my head. His harmaline-thh conversion works well too btw. His pH values for the depressions were slightly different to what I got, but this could be due to calibration errors etc., overall pattern was the same.

Your consternation that the pKa system (8.75 sep-point) flawed was also the final conclusion of VDS that this approach is fault. You came to witness the same thing as VDS did and he felt thus called for to publish a correcting article, even warning that sites like nexus and others were wrong.

We have now actually 2 ways of separating harmine from DHH:

1) the circumstances are ideal for VDS mechanics to work, like in your case;
2) when the circumstances are not allowing for the VDS mechanics to work, we end up in the good ole pKa based 8.75 separation point.

What are those circumstances that determine for the VDS mechanics to work:
First: concentration should be within a certain window, the VDS papers work around 15 to 20. You had a concentration of around 16 mg per ml at the start, that is just ideal for the VDS mechanics to work;
Second: do not use a strong base like lye. Use a weak base like ammonia or carbonates but carbonates s*ck and gives rapidly a mess. So here you fit in too.

So you complied to VDS mechanics and therefore you failed on the pKa front, as you noticed.

What are the VDS mechanics, what does it do?
the crystallization undergoes a different process, one that particularly happens more rapidly, as you noticed, circa all fell out at pH 8.
When the first element precipitates (harmine) a fallback of pH is to be seen, usually it starts at 7 and the fall back goes back to near 6 even when dripping in drops of ammonia, this is the so called pH depression. When pH eventually starts rising and hit 7 again, there is your harmine, filter it out.
Now you can base further to get the DHH, the second depression for DHH is much shallower, sometimes just a pH plateau.

***

I would say your findings are perfectly in line with true expectations, you only expected to see a pKa based theory while you actually were performing a VDS situation (if I might call it so). If you just switch to lye or KOH you would see a complete different style of precipitation more in line with pKa based approach.

***

That VDS thread has grown to a monster because we were checking stuff out and actually all that mass of posts was a living matter of study. There was no concluding VDS tek because the papers end with good directives that are hard to improve. Only mainly 2 things we at nexus concluded differently than VDS:
- to not use carbonates, we all messed up with that, only rigorous following VDS led to an acceptable approach, but you still get that nasty CO2 that skews the whole experiment unnecessary in our humble opinion;
- VDS said that his approach completely overwrites the pKa approach without nuances, while we were able to find boundaries wherein his approach works, and not outside those boundaries.
(- something of lesser importance about a cleaning up step that was doubtful)

***

Just one observation about your starting vessel looks: it was indeed a bit cloudy and should not be like that. I recon heating the solution would make it perfectly clear and also dissolve the chunks. Then let it cool to room temp, I had to do that sometimes. It's better to start clear if only for the eyes

Happy trials,
something gotta give

That's not the starting vessel. This was after adding quite a bit of ammonia (about 2 ml or so of 10% solution).

The starting vessel was at pH 2.5 and perfectly clear, color of a red wine. This is before the first pH drop, much clearer although it turned slightly brownish:





Okay I just took someone's advice "if lye does not work for you, use dillute ammonia" ... it should have been added "...but then use a different pH points as the mother liquor will behave completely differently" I really should have studied the paper and the thread deeply before experimentation. But you know, that's much more fun than reading...

Thanks for summing up the VDS thread. I will therefore stick to the VDS protocol except avoiding carbonates and strong bases. And maybe HCl instead of acetic acid (hmm... would be interesting to compare).

BTW I don't know about the Wiki... it's sometimes misleading to borderline harm-inducing, yet admins have more interest in making April fool's jokes than finally fixing the wiki access The very fact there is a harm-reduction article warning against information on Nexus should be alarming enough to do something about it... (just my little rant)

That red wine starting color looked nice

***

Only when that warning is sound.

I see no harm issue in the maoi difference between harmine and harmaline. I think VDS took it a little too far there. There is a slight advantage for harmaline only, the document he mentions does not really support his harm alarm imho. I agree on body load and duration being soundly different, but maoi-level-wise I really have a hard time believing in harm. I think a failed separation is an academic hell much more than a real harm issue and I think he is an academic person mostly, hence.
Practically:
suppose you think having 250 mg harmine but actually (due failed separation) you ingest 125mg harmine + 125 mg harmaline. Is that a harm warning worth concerning maoi level? I don't think so.

Yet I am glad he took that harm 'excuse' to make the paper, now we know why sometimes harmaline falls out under 8, and sometimes (other conditions) start to fall out above 8.75.

***

Note that ammonia alone is not enough to get VDS mechanics to work, the concentrations must fit also. If you had 100ml per gram of extract, you would not have seen your harmaline preciping that fast, ammonia or whatever base used.

Well harmaline is said to be twice as potent as harmine (Nexus, PsychonautWiki, TiHKAL). I have handed to a friend what I believed was pure harmine (given the precipitation pH) and he could actually take dose of harmine while it is actually harmine:harmaline (1:1) - this could easily end up in overdose, especially if he believes in short half life of harmine and decide to redose.



I will take that into account, thanks.

I am ready to believe harmaline is twice as potent maoi when I see some documents of study.
The only one I got so far is also the one VDS was referring at:
https://www.dmt-nexus.me...&m=764294#post764294

I have been asking for proof documents a few times here at nexus but never got an answer to that. Is it possible for you (or anyone else for that matter) to find the other documents (not just mentioning but proof please) that state harmaline's maoi potency is serious different from harmine please? I would be very thankful.

I have one research paper studying harmine/harmaline half life in blood serum I think.

Other than that, there are anecdotal reports that harmaline has higher body load. But given that most people don't even have access to pure DHH and can have a placebo effect, it's a weak evidence of course.

I will carefully experiment with harmine and DHH on myself. I was using the dosage ranges given by Shulgin in TiHKAL since he proposed 150-300 mg dose for DHH and >300 mg for harmine, respectively, which also looks like DHH is more potent (though maybe not twice).

Okay here is my second separation attempt according to VDS. I used 20 mg/ml concentration of harmalas and 1% acetic acid (though VDS used slighly lower concentrations). Basing with 1% ammonia and final basing with NaOH. Some separation occured given the look:



Now I will do a second run on the harmine and DHH fractions to see how pure it was from the first run.

I think it would be easy to make it into a Tek. Simply ensure the concentraction by using e.g. 50 ml of 1% vinegar per gram freebase harmalas.

One think I am not sure about is obtaining the mixed fraction step. The VDS paper says to add base until precipitation occurs (pH about 7.7), then filter and discard residue.

1. The pH drops after the said precipitation. Should one continue adding base until pH stays at 7.7 ? Probably not, but just wanting be sure...

2. Why discarding the mixed fraction? Non-harmaloid impurities? Maybe collecting the mixed fraction and doing separation on that will be more viable.

You could blind the study. Put a pill of the same dose of each in opaque bottles labeled on the bottom, get them thoroughly mixed up while intoxicated, wait a week, use one without seeing the label, keep notes, several days later use the other, keep notes, then look at the labels to see which was which. You could even do that with two doses of each for statistical power.
That photo will give me nightmares
That DHH goop looks like it will be hell without vacuum filtration. A possible rescue would be, after filtering the mixed fraction, to add a teaspoon of vinegar and then an equal volume of saturated NaCl and just manske out the DHH. If hydrochlorides were the end goal, anyway, it would just save a few steps. Just a thought.
Another option beyond re-extracting it would be to use it as a mixed fraction.

Yeah it was made late at night so I was a bit sloppy. I actually used vacuum but the liquid got thicker after adding NaOH solution and I forgot to wash it properly and dry it on filter by suction. Better luck next time

I don't care since I have harmalas from previous experiment to process and purify, so this will get cleaned up anyway.



My goal is actually obtaining both HCl and freebase forms of each chemical. This is because I want to make a "caapi" mix for changa and psilohuasca. It currently goes like this:

1. obtain mixed harmalas freebase (the standard process as per wiki, Mansked twice for purity)
2. separate to harmine, mixed and DHH fraction (as per VDS protocol)
3. further purify harmine and DHH, add residues to the mixed fraction -> harmine FB, DHH FB
4. salt harmine and DHH -> harmine.HCl, DHH.HCl
5. reduce DHH to THH -> THH FB
6. salt THH -> THH.HCl

So the result are 6 products to readily mix and experiment with.

The double-blind self study is a good idea for a long term experiment I will start when getting enough of the product - still having 4 kilos of rue to process...

BM you are coming to a point of separation, congrats

A test would be taste, the harmine should not taste, the harmaline should taste like hell

If you reduce harmaline to THH then it should lost the bad taste as well.

Did you notice the pH depression?
After falling out the first signs of harmine, the pH should drop back while you slowly base.
TBH instead of using pH values to determine when to consider a separation, another way is to look at the pH curve, the shape, and let the shape determine when to stop basing.
For example:
first precip start at pH X
Basing further the pH drop goes back to X-1
Basing further the pH start to rise again, and eventually goes back to X
Filter precips out and call that your harmine.
Base further to get the harmaline.

This way one is free of absolute pH points.
Why?
Because deviations from the tek and especially salt content (salt due mixing a base with an acid) can skew the absolute pH points.
Only if you have a very rigorous tek with absolutely no room for deviations then you CAN use the absolute pH points to determine separation. But the pH curve shape is as indicative as well and is free of potential pH offset due whatever.

I did an experiment before with as least as possible acid needed to dissolve my starting material, and used some heat to succeed. I left the VDS numbers so to say. Using my same base the harmine was harvested at pH 6 instead of the usual pH 7.
So the absolute pH points were lowered by a 1 pH, but the shape of the pH curve was unchanged, there was still the pH depression to count on to make separation decision.

I think the real power of VDS tek is to follow the pH depression shape, and only in a second instance the pH value marks.

***

Use lye and see nothing of those pH depressions occur, and your precipitation values shoot up, the first signs of precipitation will be a lot higher.

Use carbonates to base (also a weak base btw) and notice everything goes nuts as the CO2 formed is another beast to deal with. On needs extra steps to deal with the CO2 in both gas and dissolved form. That's why it's not advised if you have ammonia instead.

Keep up the good work.