BTW ring-substitution of the diphenidine class results in toxic compounds. If you want a SIMPLE K analogue, replace the piperidine ring with an N-isopropyl. Since the latter has more potent DRI activity, the subjective effects are much better. I suspect that swapping the 2-phenyl for a 2-p-toluyl moiety will increase DRI activity as well (but is likely too costly). There is a great RT route to making the N-isopropyl analogue. Use isopropylamine hydrochloride and add KOH to freebase AND to remove the water formed when the amine reacts with the ketone.
The imine is almost insoluble in methanol and so can easily be decanted. NaBH4 in alkali conditions simply reduces the imine to an amine. The only downside is that one needs 2500ml of methanol to reduce 340g of imine.
I discovered this trick looking at novel routes used to produce sertraline. Of course, in the case of sertraline, methylamine hydrochloride & KOH are used. I like RT reactions and I quite like using methanol as solvent but I love the use of KOH for 2 purposes. It's über cheap, simple & safe (compared with the majority of organic syntheses). It goes without saying that solvent management is also very simple. From what I have seen, 'meth cooks' are terrible at solvent management and not too bright when it comes to managing a lab. Scary.