The trouble here is that magnesium hydroxide or its carbonate will be as insoluble as freebase harmalas, and will also dissolve in the same acids as harmalas. I'm not sure what other folks did when using magnesium ribbon, but I took advantage of the fact that magnesium ammonium phosphate (MAP) is highly insoluble. By using phosphoric acid along with the magnesium ribbon, it was then possible to precipitate the MAP by careful addition of ammonia. The MAP crashes out as a dense, pure white powder at a pH sufficiently distinct from either harmine or harmaline that these compounds are very easy to tell apart from the MAP. I (or maybe @downwardsfromzero) documented the pH measurements somewhere around here a couple of years or so ago.
I get blobs of something when adding ammonia the first time. It happens at pH levels too low to be harmalas, the same as seen in the video in Q21Q21's post above. I assume this is some kind of magnesium salt? I could probably remove at least the visible stuff by filtering this off before going high with the pH. I have a slight concern that it may have an effect like I described at the end of the first post in this thread regarding excess sodium bicarbonate trapping alkaloids: Rue to separated freebases with less fuss.
After doing a manske on the entire precipitate and then basing again, there was nothing strange about the precipitation and I was able to get different fractions out with a clear pH gap between.
So I'm wondering, do I do the same thing again with this higher pH crashed portion of it, or go straight to manske? I'm feeling like a manske makes more sense, that way the one and only basing should be cleaner as the magnesium salts will already be disposed of.
As for the phosphoric acid approach, I could try that on another batch. It does seem like a manske takes care of the strange globs but one less step would be nice.