I usually did 500 g rue seed extractions with low yeld (little over 10 g extract) until this time I've used more vinegar for cooking and WHOA - there were several times more precipitated alkaloids than usual.

This causes many struggles along the way. For example, I dissolved all the alkaloids in 1 500 ml of hot 3% vinegar, then added 150 g of salt (the prescribed 10% of volume). To my surprise, the harmala HCl crystals crashed too quickly, making yellow sediment of the bottom.

I filtered the liquid and added base to see how much alkaloids were left behind. It was still a lot, like half the previous amount.

I therefore ran another Manske on these alkalkoids, dissolved in 1 250 ml of vinegar, added 125 g salt. Unfortunately, nothing crystalized this time, even after agitation and scraping with a stirring rod.

My theory is that maybe all the harmalas crystallized and these are companion non-harmala alkaloids so they don't crystallize from saline solution. On the other hand, the alkaloids are of same color as the harmalas.

I am tempted to run an experiment, testing many harmala/salt concentrations in different conditions (vinegar, water, impure harmalas, pure harmalas). This would take 100 test tubes and lots of time, but maybe we will finally have empirical data for estimating how much salt to add to get the best crystals.

From my fiddling with the reaction I have found 20g salt per 100ml of solution to be functional, dissolved whilst still hot and then left to cool at room temp before being moved to the fridge overnight. More salt than this gives a chance of salt contamination and less than this I found all the alkaloids did not precipitate. Ambient temperature here is >25c.

In one experiment 25g of salt per 100ml solution dissolved, there was salt contamination but in another identical experiment there was not. I just redissolved the contaminated sample in water in and it was cleaned during the basing stage anyway. Both yielded the same. So there is something there too.

My water here is pretty soft, low in minerals and does not contain fluoride. I've pondered about the ongoing effect of this through the extraction process and whether or not this would effect the mankse at all, but have not sourced other water to find out.

In two scenarios, one using 250g of seeds and another with 500g of seeds, extracted alkaloids dissolved in about 1200ml of vinegar, I found that the the precipitated alkaloid yield was almost identical by percentage, ~4.5%. The downside to cramming moar alkaloids in this amount of solution is they did not have enough room to crystallize properly and formed a hard solid lump in the precipitation vessel. I had to punch a hole in the crust with a spoon to pour out the remaining solution after the reaction. The 250g batch however precipitated beautiful formations and I would call it a total success, in aesthetic terms.

I ran a tiny experiment, dissolving harmala.HCl in 10% saline (distilled water and pure NaCl). 5 mg/ml yielded the nicest crystals, long needles. 15 and 25 mg/ml seem too cramped (the latter caused harmalas to crash too quickly as a yellow powder), so I will try something like 10 mg/ml. I hope it will work similarly for crude harmalas too.

I will compare efficiency of the crystallizations by filtering and weighing the remaining freebased alkaloids.

Maybe crystallizing at 10%, then increasing adding salt to 20% or even 25% will work to push more harmalas out.

My earlier extractions yielded about 12g of harmalas and 1 200 ml of water was okay - this translates to the seemingly optimal 10 mg/ml.

I will try another test run with different concentrations, then maybe try different salt concentrations too.

I am also pondering converting freebase harmalas to HCl directly (i.e. mixing with small amount of dilute hydrochloric acid). I calculated about 50 ml of 1% HCl per gram of freebase should be sufficient (works for harmine and harmaline but THH needs much more concentrated acid for some reason). This will save me from making litres of vinegar in the workup stage.

Okay I filtered, dried and weighed the samples and here are the yields for the respective harmala.HCl concentrations:

5 mg/ml: 310 mg from 500 mg = 62%
15 mg/ml: 1 380 mg from 1 500 mg = 92%
25 mg/ml: 2 400 mg from 2 500 mg = 96%

It seems higher concentraction of harmalas is more beneficial to yield, but the extra weight can as well be a salt contamination.

A better experiment would be to estimate salt contamination by measuring salt content in the remaining liquid.

I ran a second experiment, comparing freebase to HCl crystallization from 3% acetic and 1% hydrochloric acid (see pic). The acetic acid gave nicer, longer crystals but HCl crystallized much faster (process started almost instantly).

I recovered 93.2% of the material from acetic acid solution and 100.9% material from HCl solution. Obviously, the latter must have been contaminated or I measured the source freebase badly.

Anyway, it seems crystallizing directly from HCl is superior to acetic acid as it gives better yields and we will maybe save some salt. The only concern now is contamination and overall purity.

I will try smaller harmala or salt concentration and take more samples and measure the salt contaminations. This will give more conclusive method or harmala (re)crystallization from saline.

Interesting blue.magic.

The Cat approves. Thinks about making some further experiments.

Yes I will definitely have to find the best harmala/salt concentrations.

Last time I dissolved 10.02 grams of harmala freebase in a minimum amount of 0.5% HCl (just to dissolve everything), topped up the water to exactly 1 liter, heated up in microwave and added 100 grams of salt.

So the solution was 10 g/l harmala and 100 g/l salt. It precipitated too fast leaving yellow crystalline "sand".

I filtered the sand for next Manske, this time adding 70 g/l salt. Almost nothing precipitated so I increased salt concentraction to 100 g/l. And everything crashed out too fast again...

It seems there is a sweet spot for harmala and salt concentration to get the best crystals (largest and least contaminated).

I think I will have to run the experiment and test many different combinations.

So far it seems the sweet spot lays somewhere around 12 g/l harmala and 80 g/l NaCl.

And of course the sweet spot may be different for crude freebase harmalas and for the presence of acetic acid.

The problem is we don't know how much salt there is already in the saline liquid you're adding more salt to. Some of the initial salt reacted with the alks to form harmala HCl, some got trapped in the crystals.

I like to do that too, but I also reduce the liquid, so not only the NaCl concentration increases, but also the harmala concentration. And I often end up with too much salt.

In my experience, the harmala HCl crystals start forming between 90C and 70C. Ensuring that the cooling down between those temperatures happens as slowly as possible allows for bigger crystals. Cover the crystallization vessel to limit evaporative cooling and wrap in a blanket. Set the vessel aside and down disturb it until it has reached room temperature. Big crystals need time.

Another relevant detail is to make sure that the salted harmala solution starts at a temperature close to its boiling point.

That is exactly what would happen if you did everything correctly. The alkaloids that were not pushed out by salt, but were freebased out, and then again could not be pushed out by salt are the accessory alkaloids. Vascicine, etc. A problem would have been if some alkaloids did crash out on the second salting.
If your calculating hydrochloride yield from starting freebase mass did you remember to calculate the product as harmala hydrochloride dihydrates?
Manske crystals have two waters of hydration.
Harmaline hydrochloride dihydrate mw = 286.762
Subtract 1 for an equal mix of harmine and harmaline

Amount of plant gunk and how close you are to a single pure alkaloid also effects it.
In my experience, lots of plant gunk makes crystallization much slower compared to fairly pure harmaline/harmine and once you separate the alkaloids they precipitate very fast. On P. harmala seed decoction I can just dissolve rock salt into the solution, let it cool slowly, and it'll crystallize out in the refrigerator overnight. On pure separated alkaloids my only option is to salt out with saturated salt solution because the hydrochlorides will be crashing out before the salt is even dissolved, otherwise.

Thanks pitubo, Elrik.

I usually heat up the solution only to 70 °C, so this might be still too low. I will therefore try near-boiling temperature before adding salt.

As for the dihydrate, I powdered and dried the HCl crystals in a food dehydrator at maximum temperature so I assumed it was anhydrous. But it might not be the case. I don't know at which temperature it loses all the water.

Yes the crystallization is done in a 2l beaker sitting in a glass dish that fits it perfectly and provides additional insulation. But I will try other means such as alu foil and some textile wrap.

I was trying to find good conditions for crystallization since I keep getting half the yield that I should. Getting little over 10 grams of freebase haramals from 500 g of rue seeds is little too low. People report getting over twice more with just throwing in some vinegar and salt, not very precisely. So I need to find where is the problem in my technique and provide some good numbers to others. Unfortunately, the current teks give mixed results even on the same batch of seeds.

Every time I do a Manske step I try to make sure I catch all the harmalas, so I reduce to push out more until nothing comes out. The yields are around 5% alks of unknown purity, after typically 2-3 Manske cycles.

Adding too much salt - even so much that the hot solution becomes saturated and some salt doesn't dissolve - doesn't seem to do much harm. The resulting crystals may be smaller, but not so small that they would pass through coffee filters. The most obvious difference is that instead of settling at the bottom as a solid 'cake', they float suspended in solution, so decanting doesn't work well and you have to filter them.

Yes last time I got tiny crystals, but no caking, easy to filter. The resulting harmala HCl is yellow-golden colour, much lighter in colour than my earlier extractions wich are brownish (even after 4 Manskes). I was afraid the lighter colour is due to salt contamination but currently have no way to determine that.

I usually just filter the Manske solution with Buchner funnel, flush the filtering flask and pour fresh water in the Buchner - this will redissolve harmalas and the solution goes through the filter while impurities stay on the filter.