Nick:

What type of system are you using to make the trenches. Is it ion milling or
rie plasma system? How deep are you etching in microns. Have you considered
using a photoresist mask over the aluminum to protect it during the oxide
etch. I have done some work with MicroChem corp with a special formula of
their SU-8 resist and have found it to stand up to plasma real well. The
secret is getting it off after use. Seems that the formulation they made for
us was around 5 micron thick after processing. The selectivity for a process
we do etching silicon was 40:1. The process to etch sio2 would be much less
probably around 5:1 depending on the bias used but if you only have to
protect 1 micron of aluminum it might work. Setting up a process to convert
Al into Al203 using plasma probably won't yield the results you are looking
for. I also agree that wet anodizing would be a problem to control. Drop me a
line with a little more detail and I will try to help.

Bob Henderson
480-558-1156