We are using Shipley 1813 as a mask for our process. The mill is an AJA argon ion mill running at 500 V, 56 mA. The resist seems to hold up well for a 10 minute etch, but by 40 minutes it is badly pitted, resulting in a nonuniform "haze" across the substrate.

We have also tried argon-ion etching using e-beam resist as a mask. We have followed the technique of Cai, et al. (AIP Advances 5, 117216 (2015)). A very high e-beam dose is used to cross-link PMMA, making it effectively a negative resist. We have made wires in the PMMA on order of 100 nm this way. However, when we try etching in the AJA, the PMMA is almost immediately etched away. Cai, et al report no such issues in their paper.

The AJA instrument is quite new. I can think of two possible issues. First, perhaps the sample is not well-cooled. It is greased to a copper platen that sits on a water-cooled copper block, but the platen is not greased to the block, so maybe this is a weak point in the thermal transfer. Second, the system is plumbed to the argon bottle with plastic tubing, so perhaps there's some oxygen getting in? We did try pumping out the tube all the way to the tank valve, but it made no difference.

Anyone have experience in this sort of fabrication who could make some suggestions?

Thanks!