We have been observing lot of adhesion related issues while processing MEMS devices using UV-LIGA process. Typical problems such as poor adhesion between thick SU-8 mould and Cr-Au seed layer are observed... there is poor adhesion between titanuim seed layer and electroplated nickel film..
How to quantify the adhesion between SU8 and seed layer/ substrate/ Nickel etc.?
How to reduce internal stress so that the delamination can be avoided?
Stress in SU-8 films can be minimized by adjusting the pre-bake and post-bake temperatures and timings. Merely following datasheets will not help in achieving stress free films. There are several research papers available, that discusses how to minimize the stress in thick and thin SU-8 films.
SU-8 gives best adhesion to Titanium. If you want to increase the adhesion of SU-8 on any substrate, clean the substrate in acetone, IPA and heat the substrate at 120 degree for 15 mins, give few mins cooling time and spin coat the film immediately.
I highly recommend this paper:
"Processing of thin SU-8 films", Stephan Keller et al 2008 J. Micromech. Microeng. 18
it talks about effect of soft-bake conditions, exposure dose and post-exposure-bake parameters on residual film stress. I took some ideas from there which helped me a lot in working with thin film SU-8.
As crosslink and networking density of SU-8 is extremely high, best you can do as do a "slow bake" schedule, to make sure film is baking on a thermodynamic and not a kinetic schedule. What you really need is a lower tensile strength material.
The other consideration here might also be lower modulus, as SU-8 can't absorb the energy associated with interfacial stress. As to quantitative measure of adhesion of Cu/seed layer, as well as other layers, MicroChem uses sheer stress adhesion, such as measured on a Dage 4000 tool, as our primary guide.
- Badges
- Science topic