I don't agree thatThis looks VERY interesting for synthing ephedrine derivatives
There are too many steps for this to be of use in synthesising ephedrine derivatives. You would have to:
1. Make the nitrostyrene from the benzaldehyde
2. Reduce the double bond to give the nitroalkane
3. Oxidise the nitroalkane to the 1-oximino-1-phenylacetone
4. Hydrolyse this to 1-phenyl-1,2-propanedione
5. Perform a selective reductive amination to give the methcathinone derivative (STAB as reducing agent may work; this is expensive but has been recommended in the past)
6. Reduce the benzylic carbonyl of the methcathinone derivative to the alcohol to finally give the ephedrine derivative.
If you are starting from a benzaldehyde and nitroethane, you are far better off making the nitroalcohol via a Henry condensation, followed by reduction to the aminoalcohol and N-methylation.
But if you have a better idea in mind, please post it.
You could also proceed from any propiophenone to give an ephedrine derivative in around 70% overall yield. The synthesis is incredibly easy, but then I am a little biased with regards that route.