Replace using BBr3 or Bcl3 or use Pcl3. This may provide an alternative demethylation procedure for catechol derivative.

If you got the catechol, aqueous NaOH would seem a good extractive, but first you should be sure you don't have your product as the boradioxol derivative (i.e an ester of boric acid and the catechol). I know amine-boron complexes can be destroyed with HCl, NaOH or alcaline peroxide, but I don't know if any of that is compatible with your reaction product. Good luck!

Instead of BBr3, try BF3-SMe2, or methionine in MeSO3H for demethylation. BBr3 in my experience gives gross workups. Also, stoichiometric BCl3 and catalytic TBAI is reported to work well.

If HBr reflux doesn't work (shown above), I found that when using BBr3, I had to first quench the reaction, then filter through celite to remove the bulk boron bromide junk. Then extraction from the aqueous mixture using chloroform/isopropanol 9:1 was successful usually giving above 80% yield.

Article Conversion of Thebaine to Oripavine and Other Useful Interme...

I disagree with Juan, under basic conditions catechols become very prone to autooxidation. And unless you are doing your extraction under completely inert atmosphere with degassed solvents/water etc... the catechols will oxidize and polymerize pretty quick.

A 2 step procedure that works okay is demethylation with catalytic B(C6F5)3 and stochiometric Et3SiH.  The methyls are removed at the same time that the catechols are protected with -TES groups, which can be removed with acidic fluoride.

I had to make a double catechol recently and it was prone to oxidize very quickly so be careful and get the conditions acidic quickly to slow the rate of oxidation when you work it up. 

I don't know if this will help you but for my molecule with double catechols, they protected with benzyl groups in the first step, and in the final step they were deprotected by hydrogenolysis with Pd/C and H2 gas.  Because hydrogenolysis is usually a clean, quantitative reaction, and because the double catechol is so highly polar, after you purge the reaction with argon, you just filter off the Pd/C with A PTFE .45um membrane filter from acrodisc or similar as opposed to using celite or silica which will bind your double catechol.