Flourides might work, because they set HF free at a low rate. I always wanted to try SiO2 etching with ArmourEtch (glass etching paste, i.e. flourides) but got

Flourides might work, because they set HF free at a low rate. I always wanted to try SiO2 etching with ArmourEtch (glass etching paste, i.e. flourides) but got sidetracked. There might be other problems with that specific paste, for example I don't know if it contains abrasives, but if you have access to a chemical lab then flourides should be easy to obtain. AFAIK the chip industry uses flourides too (buffered oxide etch) because unbuffered HF has a too unpredictable etch rate.