After Foxy2 posted the CTH reductive amination, we started checking it. We thought about that possibility time ago, but never tried it, but in any case we didn't thought in a water:alcohol system. Our first tests, acording with TLC, where not satisfactoy, but we finally discovered that we were using little of a not very active catalyst. Then we tried it again using the best chemicals, and we found that the reaction works fine for MDA, with the only disadvantage of the use of big amounts of catalyst. I made the research with the help by a friend, it was very appreciated, all was easier.
We started our test using a 10:1 methanol:water solvent because we didn't have at that moment anhydrous chemicals, finally my friend made a test with the 9:1 system and a bigger amount of ketone and it worked similar, probably that difference is not very significant. Accidentally we found as well that ammonium acetate can substitute the ammonium formate as amino group donor, what makes possible use less formate for the reduction.
Here you are what we found in our research (all yields molar ratio):
3 gr ketone, 8.5 gr AA, 3.75 gr AF dissolved in 28 ml 1:10 solution, added 300 mg. of 10 % Pd/C stirred 48 hours (TLC showed completion, but workup was done at 96 hours), 64 %.
3 gr ketone, 9.8 gr AF in 49 ml of 1:10 solution, added 300 mg of 10 % Pd/C, stirred 48 hours, 65 %.
2 gr ketone, 6.9 gr AF in 33 ml of 1:10 solution, added 1.35 gr of 10 % Pd/C (second time used), stirred 16 hours, 81 %. (Pd/C similar to reference)
25 gr of ketone, similar proportions of the reference, full catalyst, 85 %. The higher yield could be explained by less mechanical lost of product and/or the 9:1 solvent system.
Product was checked with mp and bioassayed showing it was the racemic and desired mixture, so the good MDA.
In conclusion, the rxn works as claimed in the reference for MDP2P and probably with other interesting ketones, and yields will depend on the amount and activity of the catalyst. Probably a 20 % weight from ketone of a good catalyst is enough to get the best yields.
As you see yields are very good, and the rxn is very simple, add the ketone and ammonium formate to the solvent, and when it is all dissolved, add the catalyst. With this small amounts we didn't care about to add the catalyst wet, we added it dried at once and we hadn't any accident, but for big ammounts, it is recommendable to take all kind of precautions, but respecting the amount of solvent, we have observed that incresing de solvent volume decrease yields.
Workup was filter and add NaOH solution to release ammonia, then concentrate with the aspirator, add HCl solution, wash, basify, extract, dry and crystallize. May be we could ommit the NaOH addition, but we wanted to be sure that we didn't have ammonia salts contamination in the product.
Temperature was around 20 C, and I would say if room temp is higher, like now in summer, say around 30 C, yields could be not so good. I made a test recently at that temp and the final solution was much darker and TLC was not so good.
Carbonates of ammonia and the amine are formed, and you can see fizzing when adding the HCl solution, or a crystalline mass mixed with the catalyst in the last part of the rxn. I guess that the amine carbonate is more soluble than the ammonia bicarbonate, so may be that crystals doesn't contain amine, but it's not sure. When I observed the crystals, after filtration, I added dilute HCl to the catalyst+crystals in other flask to remove them and filterd the catalyst again adding that solution to the workup.
The chemicals were the best, ketone was 98+ % stored now 4 years in a freezer, what explain the small amounts we used, we don't want to loss that treasure. Ammonium formate, acetate and Pd/C were well branded too. This means that if you use homemade ketone, that can be from 75 to 90 %, the yield will diminish proportionally.
Ammonium formate could be easily homemade, and for this case it seems the best to generate it from aqueous ammonia over solid NaOH (or NaOH and ammonium sulfate) and bubble it into a mixture of methanol, water and formic acid till get a pH of 6-7.
When we tought time ago about a similar rxn to this one, we thought as well in making MDMA. We added a bit of aq methylamine and formic acid to methanol, and then a bit of Pd/C and we observed a clear decomposition of the formate, so we tried the rxn with aq methylamine and formic acid and TLC was very good, then we make a test with 3 gr:
3 gr of ketone, 4.4 ml of 40 % methylamine, 2.4 ml of 98 % formic acid in 30 ml of methanol, 300 mg of Pd/C stirred 48 hours, 83 %.
Yields are better that with ammonium formate using the same catalyst, but the disadvantage is now the use of more methylamine than in the other procedures, TLC tests with less methylamine were not so good. We used and exccess of formic acid to reuse the methylamine forming more reducing agent, but we don't now if it is a good solution or it affects the yields, it's a field to investigate. Unfortunately it is not a very interesting way becuase the problem of the methylamine, but I guess it could be overcome finding a good way to decarboxylate glycine.
I made this research in January, but only communicated it to a few bees, excuse me for delaying so much time these interesting results, but there were reasons to do it. Recently Rhodium sent me a review of the reference in wich they make secondary and tertiary amines using this method.
https://www.thevespiary.org/rhodium/Rhodium/pdf/redamin.cth-pd.pdf (https://www.thevespiary.org/rhodium/Rhodium/pdf/redamin.cth-pd.pdf)
May be me and them were working at the same time in the rxn of not primary amines.
We made recently a homemade catalyst to see if it could work. We thouhgt about an OTC an easy way, first we got activated charcoal from a pet shop, it was powdered with a cofee grinder, and then purified by heating it at 80 C one hour at half with 10 % HCl, filtered and dried, then PdCl2 was diluted in a HCl solution, the purified charcoal was added and then we added grindered aluminium slowly, at 40-60 C, we used about 1 gr for 140 mg of PdCl2 but probably much less is enough, (HCl was added if not rxn). Finally it was filtered and washed extensively with water and dried. The I compared his activity with a new commercial one by adding the same amount of both catalysts in two flasks containing the same amount of diluted formate. My catalyst worked, but not so well like the other (a 50 % ?). I forgot the idea of making a comparative test with both catalysts, but after a week in wich the homemade catalyst was in an open dish in a closed armchair (exposed to air), I thought that the lack of activity could be compensated by doubling the amount of catalyst, I tested the catalysts again and my catalyst had lost a lot of activity (10 % of the other ?) so I forgot the idea.
I've made different homemade catalysts for CTH rdxns and my homemade catalyst always gave poor yields. I used formaldehyde and NaBH4, and now Al, and results seem similar. I believe now that the question is the support that has been always the same, and as you understand is very primitive. I tried pharmaceutical C from capsules, but it so thin that can't be filtered. May be that's the key, find a good activated carbon to get a enough good catalyst.
Have a nice summer.
I posted info on pd/c catalysts here a few weeks ago... it appears you did not do an acid prewash on your homemade catalyst... pore size seems to be a factor as well in Pd/C catalysts
here's the post... Post 334044 (https://www.thevespiary.org/talk/index.php?topic=9014.msg33404400#msg33404400)
(ClearLight: "Effects of carbon activation on Pd", Methods Discourse)
Infinite Radiant Light - THKRA