Dresden came up with pyrrolidine subbed into mph, but I'll take credit for the idea of introducing the ring into amph
Hydrolysis into 3,4-dcmph acid would terminate all CNS effects I'm guessing, but that doesn't mean the molecule can't hit the 2b receptors in the heart which arent guarded by a BBB. Nevertheless the acid would have a short half life before it's quickly excreted I think, so I would bet that even if it did have 2b affinity, it would be short lived agonism. It's pretty interesting how a tfm group in norfenfluramine increases 2b activity several fold over amph which is pretty much negligible, as well as 5ht releasing effects. I think in terms of binding interact, once the nitrogen picks up a proton it will interact with a negative residue, but an electron rich aromatic binds weakly. Once electron density is removed from the ring, better interactions will occur (ie the electron deficient aromatic repels less) and so greater 2b affinity is the result. With that being said, could one of the chlorines in the molecule be substituted with a methoxy? The idea here is to still block the transporters, but add electron density to the ring to remove 2b affinity. It'll make the ring more prone to hydroxylation but that doesn't happen anyway, so I can't comment on the new half life. As for neurotoxicity of the o demethylated metabolite, I don't think it will enter the neuron ( I don't think ritalin does anyway, correct me if I'm wrong, maybe it just doesn't block vmat 2 and agonise taar1). I think all it's gonna do is sit on the transporter and block it, and this is going to stop say dopamine entering the 5ht cell, so I guess it's similar to how fluoxetine is hypothesised to be neuroprotective post SRA administration.