The product of reductive amination in this context is always a Lewis base. Resolution is a separate reaction.Actually it's not always, there are asymmetric catalysts, but afaik they aren't used.
Step into the chrial pool
Thankfully enantioselective synthesis of MDA has a few papers on it so people won't have to resolve it with old microwaves anymore. "Surfing the chiral pool" for example.
Homochiral (R)- and (S)-3,4-methylenedioxymethamphetamine (MDMA) were prepared in six steps (each) from the chiral pool precursors D- and L-alanine, respectively.
A divergent two-step process has provided access to optically pure enantiomers of MDMA and MDA, clinically relevant phenylisopropylamine entactogens. Target compounds were synthesized from commercially available alanine-derived aziridines. Critical process parameters were identified, and the reactions were optimized to avoid chromatographic purifications toward gram-scale isolations, providing (R)-(−)-MDMA, (S)-(+)-MDMA, (R)-(−)-MDA, and (S)-(+)-MDA each in greater than 98% purity by UPLC, >99% enantiomeric excess, and net yields between 50 and 60% for the complete process.
Traditional resolution via selective crystallization of diastereomeric salt forms has not proven the most effective route for synthesizing MDMA in high enantiomeric excess.56 Instead, a more successful strategy has relied on the use of removable chiral auxiliaries. The first asymmetric synthesis of MDMA was reported by Nichols and co-workers Reductive amination of ketone 12 with (S)-α-methylbenzylamine (15) produced the (S,S) intermediate 16 following crystallization. The use of Raney nickel at 50 psi appears to be crucial for the selectivity of this reaction. In our hands, Raney nickel catalyzed hydrogenation did not proceed under atmospheric conditions, and the use of hydride reducing agents such as NaBH3CN yielded an inseparable 1:1 mixture of diastereomers (unpublished results). Palladium-catalyzed hydrogenolysis afforded MDA (14), which was converted to MDMA after reduction of the formamide. In 2014, Escubedo and co-workers reported a similar approach using Ellman’s sulfonamide as the chiral auxiliary