Harmala extraction always includes the step of precipitating Harmala*HCl as a purification stepnamed after Alkaloid-Chemist Richard Manske. During my first Harmala extraction I totally failed at it, as the resulting crystals were way too fine to filter so it was not doing anything. I even tried a budget camping pump in order to vacuum pump it with my car-cigarette-lighter and nothing worked

From that point I simply made
- cook seeds in pH 3 water
- filter
- basify to pH 10+
- wash until neutral
- decant

And at least from the practical side I did never encounter any unwanted effects or similar that could be created by contaminants.

Now years ago there was an NMR analysis on some freebase powder seen here It shows no Vasicin and also no other signals anywhere.

But these might still be hiding below the measured signals and furthermore NMR is not capable of catching up trace components. Still any other substances hiding here would only be trace compounds, so from a perspective of consumption that material would be indeed totally safe and at least 95-100 % pure Harmalas and the Manske-step probably rather cosmetic. As far as I know this step is anyways designed to get rid of vasicin, which would be present here at > 1 %, which would make 2 mg eaten when you ingest a full dose, so even at maximum near 0. VDS protocols are even saying kitchen chemistry harmala extraction would be producing dangerous products, that was the original intent for harm reduction. I am not really aware of what methods were used which could keep dangerous by-products, but at least the analysis above cant prove anything which would account for that. But well maybe they account for kitchen-methods for THH conversion and not specify Harmala extraction methods as dangerous.


But it made me still wondering because of 2 reasons:

But this stuff will not dissolve 100 % even in strong HCl and always some brown stuff remains ...

When measuring NMR there is always some brown stuff remaining, too, even if you use DMSO.

In both cases it is quite a considerable amount, like 20 % of the total alkaloids. But as those extracts are still 95+ % pure Harmalas this seems rather strange to me.



So as I cant measure MS / HPLC / GC since a long time I did not bother, but now sent a sample to a friend for analysis and to see if there is indeed still something else hiding, which would ultimately justify this additional Manske step.

First just an NMR from that recent sample:



First analysis above was roughly 1:1 Harmin:Harmalin and this is now 1:1,3 Harmin:Harmalin. Besides still shows no real sign of contaminants, but as last time something could be still hiding here and also I already told that not everything dissolves, so maybe that remainder is a true contamination.

Now here is the HPLC Chromatogram. This method has the benefit of using a much lower concentration than NMR. Therefore I was told all the material dissolved 100 % in ultra-sonic bath and therefore we will be sure that we analyze 100 % of the material. Seems it was overloaded, but in this case that might even be desired as that increases sensitivity for trace compounds even more:



Only one very big peak = Harmin + Harmalin ( ~ 8,7 min)

Those are chemically nearly identical so they are not separated here due to high concentration. But any other Alkaloid that would be realistically found in those seeds will run at a noticable different time. Still everything that can be seen is a minor peak earlier (~ 8,2 min) and even with this heavy overload they only make up 1,1 %, so in reality probably less than 0,1 %. Therefore I would conclude this is basically a pure compound.

Now also just for complecity the ESI MS included to show that the (nearly) pure chemical is just the 2 Harmala Alkaloids:



So 2 masses found that exactly correspond to Harmine / Harmaline + 1H+ = 212+1 u / 215+1 u




So to conclude the NMR results never showed any signs of contaminations if you simply skip the Manske step. But there still remains some uncertainty which cannot ultimately prove that Manske is not needed for purification. Now with the HPLC / MS it shows that even when 100 % material is analyzed it shows far more than 99 % of pure Harmin/Harmalin mix. So I would really say that this step is unnecessary. Sadly I cannot read in the VDS protocols what exactly is avoided when going this route.



Literature quotes given here as far as I see mostly speak about intoxication of simply too high levels of Harmalas? No mentioning of dangerous by-products of extraction.

So now with a shorter isolation pathway the product seems just perfectly fine for me. Cannot conclude anything now that would justify the concerns or the Manske step - just pure Harmine / Harmaline.


Therefore the procedure I always do seems just fine for me to create Harmalas, I never got any problems and analysis is proving too:

1. Boil seeds at pH 3 for 30 min at 80 °C
2. Strain and repeat
3. Combine and settle down particles
4. Decant
5. Basify to pH 10+

"Workup":
6. Pour Harmala-water into BIG container and fill with water
7. Wait 10 min for harmalas to settle
8. Decant and repeat

Do this as many times until you reach pH < 8. Interestingly that nearly always converged with having a totally clear water surfactant, no colour anymore in the water - in case you have no pH meter. You can easily throw a 500 g extraction into a 10 L bucket and decant from this, which will make it much faster. Harmala solubility is so bad, that it goes to nearly 0 in water and you will not loose any even with that huge water volume.

10. Put last Harmala droplets on a plate and dry in oven at 50 °C with ventilation on.

Super easy and not even 1 single filtering step included

Ok but why the color change? Crude extract is quite brown and the first few Manske steps yield long red needles. The subsequent Manske steps make the crystals lighter and also change the crystal geometry. I never get long needles in the latter steps but very fine needles arranged in a round flower like structure. Their texture is very delicate as compared to the red needles which are fairly sturdy and keep their form when the mother liquor is poured out. There's also a significant difference in weight.

Now, if indeed the crude extract is mostly just pure harmalas, it may be that only the trace amounts of other stuff is responsible for the significant differences in morphological properties. One explanation for the weight loss is that some of the harmalas don't precipitate via Manske or via basification. I've noticed that after filtering the freebase harmalas, if you leave the basified water, some more stuff crashes out after 12-24 hours. It has a different texture than the freebase harmalas. Also, the mother liquor after Manske steps usually does not precipitate anything on basification (if you do the Manske right that is), but here again, if you leave the basified liquid standing for several hours, some more stuff crashes out. Now the weird thing is, after multiple Manske steps, the mother liquor after Manske precipitation starts to yield a significant precipitation upon basification. This precipitate again has a different morphology than the precipitates of initial steps. The odd thing is, that you can filter out the crystals after Manske and add more salt into the mother liquor but that usually doesn't crash anything out (that is, if you did the Manske right!). But only on basification more stuff seems to crash out. Aside from that there's the yellow color and the florescence which doesn't seem to completely vanish no matter what you do.

So... All in all, I think the matter is a bit complicated. I can think of dozens of different hypotheses... One good starting point would be to test the mother liquors and the different precipitates of different steps.

But BW those Manske crystals are so beautiful!

Seriously though, I am not in your league in terms of chemistry, but I have also wondered about a few things in that process and have had some strange results when working with harmalas up to the zinc THH reduction like one time the liquid and subsequent freebase powder became pink instead of the expected pale white.

When the liquor(?) is in an acid state there seems no end to the solids that end up in a filter. No matter how golden or clear the liquid. Always some black stuff. Every time. That has always fascinated me.

Maybe VDS just needed to justify his chosen research and 'harm reduction' did the trick? He certainly nailed the pH reversal separation tek. Eternally grateful for that.

I also ignored the 'abortion' warning (for obvious reasons it didnt seem a risk to a male) and have never had any issue with taking the 'crude' harmine/harmaline combo. As you say it might be even more overrated as a potential harm than I even thought. Certainly had no unusual effects from it. Maybe its like all the oft-repeated warnings that are attached to discussions about taking MAOIs and that ignore the reversible qualities of the harmala MAOIs.

I do enjoy the 'Manske' from a satisfyingly aesthetic point of view, especially the 'golden banana slice' crystal clumps - but that is all really, plus of course the pursuit for (apparent) 'purity' in terms of the powder and liquid colours along the way. The separation of the harmaline in order to get THH was also another reason I went for the clearest golden liquids and whitest tan powders. But that was all assumptions really as I said.

As Dasein says there are many other strange things that seem to precipitate or affect the process. Some weird sharp clusters of crystals I got once after forgetting a flask with Manske left over liquor. I posted some photos of them somewhere here. And of course my 'pink' THH.

Definitely a whole lot of interesting going on in there! But nice to know the 'crude' is as good as I suspected it was, because a whole lot of that disappears after the 'Manske'.

One question BW - is the 'crude' (un-Manske'd) freebase easy to separate into harmine / harmaline via the VDS method? I've never done that before without a Manske step. Maybe I will try that sometime.

Great post and good findings! Thank you!

Would it be possible to find out the proportions of harmine, harmaline, and harmalol, especially in roasted seeds? Lightly roasted and dark roasted. There is a suspicion that the harmaline get decomposed by the heat but but so much the harmine.

Has anyone tried adding some activated charcoal to reduce any color contamination? Overdoing it might cause yield loss.

I always thought that risk of ingesting vasicine is exaggerated (but this is understandable as harm reduction works better that way) and that amount of impurities in non Mansked extracts is low.
Still, I always do Manske as good purification step, at least for better, lighter colour of the extract.

Thats what Endlessness told me too He just enjoys the step. Sadly my crystals were really small, maybe also a reason why filtering was so hard.





In general Alkalouds would be present in their salt form anyways, as below pH 9 all Dimethylamines will react as a base. So naturally you would expect anyways that plain water is enough. I even read here that when adding regular water, it will slowly decrease in pH anyways when cooking Harmalas. Still it is sometimes observed in Chemistry that even compounds like XY*HCl will dissolve much faster in for example 10 % H2SO4 than in plain water. And furthermore in natural extractions acid might break up the plant material more and increase extraction efficiency, although base would be better (at depolymerizing the cellulose and maybe even lignin when pressure-boiled).



Yes thats exactly the same process and I think it's really enough





I'm sure colour will go brighter and brighter and that is indeed a sign of increasing purity. But nevertheless purity is basically 100 % in the beginning already. That brown impurity is really likely to not interfere at all, Vasicin would be not coloured.

But still this made me wonder, as I also had the feeling that "true pure" Harmalas are even less coloured than light-tan. When you check here at the end of the post, it shows white Harmala precipitation from DMSO (concentration ~ 100 mg / 1 ml) So I was already thinking that the total brown coloration might come to a high degree only from that. Harmala UV Vis was measured on my other thread and it actually stopps before 400 nm. Even Wikipedia speaks of a brown/tan solid, but maybe that is also because these are simply always extracted from seed material and rarely you would synthesize them from scratch. A trace of that brown stuff always carried over will always cause this brown colour then.



Thats why I want to try exactly this tomorrow.



So I wrote also in that Harmala thread that you can crystallize Freebase Harmalas from DMSO. And this made me again think that pure Harmalas even may come close to white. DMSO is regarded as one of the best non-toxic allround solvents and so it will pretty much make sure that this brown impurity will not precipitate. Harmala solubility is really bad on the other side. That is why it slowly precipitates from DMSO. Now I attach a picture and here you can see how crystal-clear crystals grow and this is not Harmala*HCl, but Harmala Freebase:

(will wait for longer to grow them and then wash with water to see if they can go to completely colourless)

DMSO is not so toxic but it's a great carrier. I would take caution handling it compared to other solvents.

That's true. But if truly only sticking to natural products' extraction worst could happen is getting high

Here is a quick result from the Harmala-activated-charcoal test.

I feel you will always loose quite some here. Harmalas can be viewed just as conformationally-restricted tryptamines (probably reason for their MAO-inhibition) and thus dont have the very flexible side-chain. They can be viewed more flat, which Harmin coming closest to a planar ring system. That stuff is more prone to adsorption itself, as these non-covalent forces just happen with enough "molecular contact area". So a flat molecule will readily adsorb, if it has enough p-Orbitals AKA double bonds. Thats a reason why chlorophyll can be removed so perfectly with activated charcoal, but salvinorin will not stick to it. Maybe Harmalas stick already much better ... but for a quick test yield loss does not matter.

I used an excess of 10x Charcoal to 1x harmalas. Dissolved in DCM and ran over a Charcoal column. Result was loss of nearly all colour. Then evaporated gave again Freebase Crystals. Makes me wonder if that is a sign that this tiny brown trace impurities prevent it from crystallizing normally (as freebase) or if Harmalas actually crystallize like DMT, we just never see it happening as precipitation from water with base is a sledgehammer method, will probably never give true crystals for anything.

Anyways also included some microscope pictures. Quite colourless that stuff, so I guess adsorption did something at least? But still loss of colour could be also as I lost much of the initial material? Just a guess, did not weigh though ...
But anyhow it crystallized, which I think never happens when you evaporate Harmalas in DCM. So must be something to this I guess?

Repeating now with IPA and Aceton.

Again a big portion of Harmalas did NOT dissolve. I still threw it on top of the column and let more DCM / Aceton run over it. Nothing dissolved ... so it's truly weird how like always 1/3 of Harmalas never seem to dissolve in anything (HCl or strong NPS). Maybe it's always 1 of the 2 Harmalas that easily dissolves, while the other has a solubility of a rock?

Will check what is that remainder and then would be funny if it would be mostly Harmine or mostly Harmaline

- Harmala in IPA before Charcoal
- Harmala in IPA after Charcoal
- Harmala in DCM after Charcoal evaporation
- same closup
- same closup
- same closup on big rock

So everything analyzed from this very simplilfied Harmala TEK has always been showing up as 100 % Harmala Alkaloids, no other compounds found.

Still originally when doing only NMR analysis, it will not track down traces. But when I first only did NMRs of Harmalas, there was always some insoluble powder remaining and that made me originally believe, that this might be some *chemical brick*, neither dissolving in Water/HCl nor DMSO. This is the stuff that weirdly will not dissolve back, after you put your Freebase Harmalas into vinegar. But now with the data posted from HPLC / MS it also shows that this sample lacks any trace amounts of other Alkaloids, showing that there is nothing which justifies advanced cleanup / Manske.

But now I also wanted to made a check with NMR on that insoluble remainder. To kind of get a "pure" insoluble freaction I just boiled non-refined Harmala powder in DMSO at 180 °C and then removed whatever was left. This should be free of the "normal Harmalas" now and only this insoluble, weird stuff left.

This was left for stirring in DMSO-d6 for 30 h, so I hoped that at least some parts of this stuff will still make it into the solvent and be ready for analysis.

Now here is the picture:



Long story short, again just 100 % Harmalas. Nothing else ... so as this fraction should be basically free of the *well dissolving Harmalas* and this NMR only catches up whatever is the brick that does never dissolve easily, it also simply shows that this is just Harmalas and nothing else. In my opinion a last sign, that the whole Alkaloid fraction extracted from Harmalas is never something else than a mix of Harmine/Harmaline.

But interestingly there is also now a shift in the ratio. This was from the original batch in post #1, which shows 1:1,3 Harmine:Harmaline. Now with the batch "depleted" by hot solvent pulls it shows a ratio much higher to Harmaline. This makes me believe if maybe the solubility of Harmaline is pretty bad in contrast of Harmine, and Harmaline for some reason is accounting strongly for this brown powder, that will always remain when trying to dissolve your "crude" (actually not crude, but pure) Harmalas again in anything (Vinegar, Water/HCl, DMSO, IPA, ...).

Awesome work as always BW.

Thanks!

But I think this could really go into some serious debate. Is really anything except an acidic boil, basifying and washing until neutral needed? I see no sign of this anywhere and TEKs that tell to do more complex workup dont have any analytical data.
Everything that comes out of the seeds is just Harmine / Harmaline. Scientific papers reporting discovery of novel derivates of Alkaloids and similar substances usually just report them as a qualitative finding, yet their quantitative amounts remain minuscule.

For example there are quite some dozens of Cacti Alkaloids, but in reality their quantity seems like less than traces besides Mescaline. I guess the same is true for all the Rue Batches I have tested.

..that's really great info...and i think where things can be simplified is always best..
thanks Brennendes Wasser for your extensive work here at the nexus