Hi,

If i understand correctly, you are experiencing an adhesion issue at this time, i.e. the SU-8 pattern is lifted off from the wafer during development.

If this is the case, i would suggest you to try the following:

1. Instead of HMDS, I would use Omnicoat to increase adhesion to the silicon wafer. It is applied by spin-coating followed by baking at 190 C for 30 s-1 min. If Omnicoat is not available, try simply by cleaning the wafer with O2 plasma and proceed with SU-8 coating.

2. 1 min development for 1 um thick SU-8 sounds too long, based on my experience. Keep it to 15-20 s to start and see if more is needed.

Good luck!

Hi Gianluca, thanks for your suggestions. I will try these to see how it is improved. By the way, I wonder how O2 plasma can help. As far as I know, SU-8 is hydrophobic while the plasma-treated surface is hydrophillic. Is there another coating step in between? Thank you.

I think a minute or two of high power O2 plasma right before you spin should activate the surface enough to sort you out, if not the HF dip might also be a good option.

Hi,

about the role of O2 plasma before coating: indeed the surface of the silicon will become more hydrophilic, but the main purpose in this case is to have as clean as possible a silicon surface. As pointed out by others, a brief HF dip could help but i find it inconvenient for safety issues. When i deal with adhesion issues, e.g. very thick SU-8 or high aspect ratio structures (>5), i often pre-coat the wafer with a thin SU-8 layer and flood-expose it, so that after post-baking it provides for a SU-8 on SU-8 contact surface. I wouldn't expect this to be necessary for your case, though.

Hi,

Other concerns, 1 min development time might be too long for the 1 um of SU-8 thickness. You may start from 10-15 s.

Thanks Ed, Gianluca, and Iiro for all the answers. I have tried the plasma cleaner (air plasma, Harrick Plasma, Medium RF level, at 1100 microTorr, for 20 sec) on a preclean, dried wafer. It works very good now. However, I have to develop it in a long process:

- Post bake at 95 C for 2 min, then 150 C for 1 min, and 180 for 20 sec before developing.

- Develop in SU-8 developer diluted with IPA (2:8 ratio) for 10 sec or until seeing a little white stuff coming out. Rinse with IPA throughly. Heat at 180 C for 3 sec.

- Repeat the same thing with diluted developer (4:6 ratio)

- Repeat the same thing with diluted developer (1:1)

- Finally, rinse with 100% developer and IPA.

I know this process is miserable but I can't do better without it. :(

FYI, all are done in regular wet lab, not inside the clean room. Not sure if this is the reason I have so much trouble with fabricating small features.

SU-8 ia negative photoresist which means that the uv exposed area must get resistant against developer and remains unremoved by the developer. When your all photo resist is removed by the developer it means that you may not enough subjected your photresist to uv light. Which means that light dose is not sufficient to strengthen the cross links in the resist. No one raised this point previously.

The second point is that the surface of your wafer must be free of moisture and completely dry before you deposit the photoresist.

Best wishes

Hi Long Pham

Can you please tell me what rpm did you use to get a 1 micron thick layer of SU8 2000.2?