Here's an idea:
Make an ester from phenylalanine.
Mannich reaction of the ester: formalin, ammonia, HCl.
Heat to deaminate.
Basify to break the ester and pull into xylene or mineral oil (if it'll be soluble in mineral oil) and heat to 300C to decarboxylate. Alternatively, it could be done with spearmint oil catalyst under much milder conditions, but then you have to worry about the Pictet Spengler reaction. Who knows if those products would be toxic.
Then, reduce with sodium dithionite (rust stain remover)... amphetamine.
Might be workable...
EDIT: Nevermind... obvious oversight... the center is not going to be able to deaminate with the carboxyl still there as it'd result in a pentavalent carbon... and if the carboxyl group is gone, deamination wouldn't be possible.
Make an ester from phenylalanine.
Mannich reaction of the ester: formalin, ammonia, HCl.
Heat to deaminate.
Basify to break the ester and pull into xylene or mineral oil (if it'll be soluble in mineral oil) and heat to 300C to decarboxylate. Alternatively, it could be done with spearmint oil catalyst under much milder conditions, but then you have to worry about the Pictet Spengler reaction. Who knows if those products would be toxic.
Then, reduce with sodium dithionite (rust stain remover)... amphetamine.
Might be workable...
EDIT: Nevermind... obvious oversight... the center is not going to be able to deaminate with the carboxyl still there as it'd result in a pentavalent carbon... and if the carboxyl group is gone, deamination wouldn't be possible.