..but I did find these at the patent office:
Patent GB482414
:
Example 1: 4-methoxy-N-methylamphetamine base is refluxed in an excess of 48% HBr for 1 hour. Yield of 4-hydroxy-N-methylamphetamine: 80-90%
Example 2: 4-methoxy-N-methylamphetamine base is dissolved in 5x the amount of conc. HCl and heated at 130°C in a sealed reactor for 1 hour. Yield: 90%
Example 3: If one substitutes the acids in the preceeding examples by HI/P, the yield is 90%
Patent US2015579
(by Gordon Alles):
While several procedures are possible, involving different acids and time of heating, this demethylation can be carried out with constant boiling hydrobromic acid (48%) in water in good yield in the following manner:
One mol of 1-(p-methoxyphenyl)-2-methylaminopropane, as the free base or as a salt such as the hydrochloride or hydrobromide, is dissolved in one liter of constant boiling aqueous hydrobromic acid solution (48%), and the mixture heated to boiling for four hours using a reflux condensor under atmospheric pressure. The water solution of the product, obtained by evaporation of the excess aqueous acid present, addition of water and decolorization with charcoal, is made alkaline with a concentrated aqueous sodium carbonate solution. The desired 1-p-(hydroxyphenyl)-2-methylaminopropane is precipitated as a gummy material which soon solidifies. this solid can be crystallized from ethanol, alone or with the addition of ether or benzene, or may directly be converted into an addition salt by combining it with an acid.
With the JCS article, and the patent Ganesha posted in mind, it can be assumed that 3,4-dihydroxyamphetamine or its N-methyl analog will give similar results.
This route may prove to be less problematic as a way to MDA than the eugenol demethylation, as mr. Alles has just demonstrated the scalability of it.