MDMA with LabTop ratios, even without silica gel in imine formation, produce incredible yields with pure ketone, from 84 to 90 %, and using silica gel our best was at least 91 %, may be a little more (I'm considering the error weighting ketone is the worst case, it has been a recent and small test).
It's the best by far. Clean, easy, avoids mercury salts...
And yes, antibody2, after that test, we made three more trials using LabTop ratios, and we were atonished seeing our poor yields, from 15 (even 10 from a friend) to near 40 %. We thought that scaling didn't work, and may be AA should be in big excess, but finally it is what it is.
By the way, about the pseudonitrosite way, how would you go from the nitroalkene to the amine ? Beaker's 2CH way ? It needs fucked THF and Pd/C (homemade didn't work), and I never got his impressive yields (80>%), but a 65 %. Anyway it could be a good alternative.
Flipper, you don't really need to add so much water, we don't. After reaction (and filter if you have used silica gel and it was not previously removed, what seems now better), we evaporate the solvent with an aspirator using hot/boiling water. First slowly, methylamine produces some foam, then we add water (it prevents bumps) and evaporate the solvent, and then may be add a bit more of water and extract with your prefered solvent. It takes more time, but then you are working with domestic volumes, and you use less solvents what means breathing less toxic things. We love evaporations.