I am etching SiO2 to use as hardmaks to a diamond device. I've noticed that walls in diamond were not good due to the roughtness of the hardmas and now I have to improove the SiO2 etching process. I'm using CSAR (AR-P 6200.09) which I spin coat at 4000 rpm through 60 s to form a ~ 200 nm thickness resist layer over a PECVD deposited 100 nm thick SiO2 layer (I'm also using HMDS to improove CSAR adhesion on SiO2). Then I do soft bake for 60 s at 150ºC in a hotplate before e-beam pattering.
I'm using raith e-beam with: doses 80uC/cm²; stepsize of 4nm; 10um apperture; 30kV EHT.
I'm doing postbaking at 130ºC for 60 s in the hotplate before etching.
To develop CSAR i'm using AR 600-546 (60 s) and AR 600-60 as stopper, and drying with dry N2.
I've tried to etch SiO2 using C4F8/SF6 and CHF3/O2 recipes by ICP-RIE and both recipes resulted in bad walls (the angle is good meaning I have a good proportion between chemical and physical etching, but the roughness is not goot and this seems to be a problem with the resist).
I'm attaching files with SEM figures of the process before and after the dry etching. After etching, I left the sample in Dioxolane for 60 s to chemically remove the resist but is was clearly no fully removed. What can I do to reduce walls roughness? Increasing postbaking time and temperature should make resist walls more smooth? Dos the HMDS affect my results? Or should I phocus in the etching recipes?
Hi, the structure shown in the SEM is the oxide mask after etching using the resist mask?
Yes. The first one is just the mask before etching, the second one is the oxide mask after etching and the texture over it is probably residual of the resist that I tried to remove by dipping the sample 60 s in the resist remover. I will try sonication next time to see if I can remove it properly.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Powders