Both harmine and harmaline hydrochloride seem to form hydrates.

This might not make much difference pharmacologically (the dihydrate is 'only' 17% heavier compared to anhydrous HCl salt), but it's a huge difference from other standpoints (analytical, economical) and the presence of water might be an issue when recrystallizing harmalas from other solvents than water (?).

Another issue is that the water of crystallization skews melting point measurement when assessing purity.

I assume the 'Manske' procedure results predominantly in the dihydrate salt, as harmalas crystallize from aqueous solution, picking up water in the process.

I am still pondering the idea of recrystallizing harmalas from anhydrous solvents, avoiding multiple 'Manskes' and lots of salt (which is a water contaminant).

Another advantages of the anhydrous salt are higer potency and less bulk (per gram). It has also more specific melting point.

I found some indirect suggestions that harmine and harmaline crystallize differently in various solvents, giving an opportunity of further purification of the pH-separated compounds. Finally, it seems having a solid acetate crystals is possible, which is also attractive:



To have a better idea about the weight differences between freebase and various HCl hydrates, here are the weights corresponding to 1 mole of harmaline, both absolute and relative (to freebase):

freebase: 214.263 g (1.00)
hydrochloride: 250.723 g (1.17)
hydrochloride hydrate: 268.738 g (1.25)
hydrochloride dihydrate: 286.75354 g (1.34)

As it happens I was thinking it might be worth trying recrystallisation of harmala salts via boiling aqueous ethanol followed by refrigeration. So, for anhydrous salts we'd need to use absolute ethanol perhaps - but does this form the fabled harmala red on evaporation?

Don't know about the harmala red, but I had ethanolic harmala tinctures of very high concentration (dissolved as acetate) and it seemed stable.

I've got only about 50 g freebase from 100 g harmala.HCl. This suggests either heavy salt contamination (unlikely) or presence of hydrates (or both).

I will try to vacuum-dry the HCl salt to see if there is significant weight loss ( => hydrates).

To remove salt contamination, perhaps just quick washing of the crystals with ice cold dilute hydrochloric acid will help.

I will also try to make HCl salt directly by neutralizing the freebase and evaporating the solvent. Once the harmalas are pure enough, the conversion can be done without Manske.

Recrystallizing freebase from alcohol is also very appealing.