Dear reader:
Thank you very much for looking into this question!
I am testing a dry etching process that transfers the photoresist profile into silicon. The target is using O2/SF6 reactive ion etching (RIE) to achieve a 1:1 selectivity (Si etching rate: photoresist etching rate 1:1). But when I am close to the 1:1 selectivity by: changing O2/SF6 ratio; changing the power; changing the pressure, I always see a dark, rough surface.
I notice that this phenomenon might be due to the SiO2 mask layer forming on the wafer surface, so I am adding CHF3 for Si etching, but similar things happen when selectivity is close to 1:1. So would there be any other reasons for this phenomenon? How can I improve the roughness? I can only tune the power, pressure, and gas mixture, without control for temperature and DC bias.
Thank you very much if you can give me any clue!
This might be the "micromasking effect". I also experienced this previously using O2/SF6 to do the deep RIE on the Si wafer. In my opinion, this phenomenon may happen because the plasma hit the PR, and the sputtering of the PR remained on the rest area of your wafer and formed the micron-masks. When you proceed RIE with these micron masks, you will receive a grass-like, rough, black surface of your wafer.
There were two ways that I did to improve this RIE process to avoid the micro masking effect:
1. Instead of using PR as an etch mask, I use Cr.
2. Do 2 or 3 steps of RIE (e.g., if you plan to do 10 mins RIE, split it into one 5 min and another 5 min steps.), and between steps, clean your wafer with O2 plasma, make sure there are no contaminations on the wafer. Also, do the O2 plasma cleaning to the chamber to purge the remaining gases.
Hope this answer will help.
Yudong Wang :
Thank you very much for the answer and the solution! I will check whether an inter-cleaning step is working in my case. Although an O2 plasma for cleaning might change the etching results (selectivity, surface smoothness), at least I can check whether it is due to PR micro masking.
Yuchen Song 02 Plasma may remove your PR, and I used Cr as my etch mask. The process is listed below:
1. Deposite 100 nm Cr on the Si wafer
2. Spin coating PR on the Cr-coated Si wafer
3. Lithography on the PR and making the pattern
4. Cr etching with PR etch mask
5. O2 plasma cleaning or Acetone cleaning to remove the PR (dependent on your PR material)
and you will have a Cr (or other metal) etch mask with your pattern.
Yudong Wang
No, I do not have the metal mask, only the PR pattern, like PR microlens. The target is 1:1 selectivity etching, transferring the PR pattern to silicon. Here is an example: Article Fabrication of refractive silicon microlens array with a lar...
The tool is limited in settings for the plasma process (temperature, DC bias), so I am not surprised with the results. Just to check whether I can improve, thanks again for the answer.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials Chemistry
- Polymer Chemistry