I have some small metallic patterns already made with Ar milling starting from a metallic film (Py/Cu bilayer) in a SiO2 substrate. After that i want to make some connection pads with ZEP520A resist. So i follow this recipe for coating:
1) Pre-baking 5 min 120°C
2) HMDS spin coating
3) Baking of HMDS 90 sec 120°C
4) ZEP coating 3000 RPM (roughly 400 nm thick)
5) Baking of resist 90 sec 180°C
Then the exposure is made and after that I used ZED-N50 as developer during 90 sec and ZMD-B to rinse during 30 sec.
I have followed the exact same recipe previously and normally it does not crack after development (specially in new SiO2 substrates it never cracks), but with certain samples the cracking phenomena appears and then I can try many times different recipes and I can't manage to avoid the cracking. For example I tried without HMDS primer, and also I tried very slow cooling down but nothing works.
The only thing I remember it worked in the past was when I performed a low voltage Ar milling during 1 min or something to the whole sample before coating, and sometimes the cracking issue was solved, but it was more or less random. This milling was supposed to enhance the stickness of the resist to the substrate by cleaning the sample and also making it more rough, but I don't really know what is happening at the nano-scale.
I attach pictures of the cracks. In the first one the cracks occur because the device is there, I think the crack propagation starts in a sudden change of height. In the second one no cracks appeared because there was no device there. But you can see that the pattern is quite distorted, also the channel is wider so it's clear there is a lot of stress in the resist.
So, any suggestions to eliminate these cracking issues?
Update:
Finally the only solution I found was to change to a different resist. I used PMMA instead of ZEP520A and the cracks did not appear anymore. It seems that the ZEP520A resist is designed to produce an undercut in the profile of the developed pattern, and for that reason some stress exist in the resist. For some reason the state of the surface of the substrate seems to be very important for dissipating or not this stress. Or at least that's what I think it's happening...
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling