Hi dennis_pro, indeed it would be a good idea to clearly establish the identity of the product. I just wanted to point out that the alpha-substitution product Ph-C(=O)-C(Br)(CH3)(NHCH3) you proposed would be highly unstable (bromo- and methylamino-groups on the same carbon). I guess it would eliminate HBr and form the imine Ph-C(=O)-C(CH3)(=NCH3), which again looks pretty unstable and might react to god knows what. I could imagine that A/B extraction does get rid of it, if it was formed. Just some speculation.