I agree with Pierre.  I have used this method in IC production.  We used
boiling concentrated phosphoric in a reflux condenser. We monitored the
temperature and added water as needed to maintain the temperature (and
boiling point) at 150 deg C (I may be remembering the temp wrong.)  If you
allow the boiling point to get too high, the phosphoric will etch the oxide
at a much faster rate.  You need a little bit of thermally oxide underneath
the nitride. Once the phosphoric etches through the oxide underneath, it
will attack the silicon and make it look real ugly.  I can probably find
some notes on the procedure if it is of interest.  It's only been 25 years!

Roger Brennan