Dear all,
I am doing a square pattern on a 6 mm x 6 mm GaAs substrate by photolithography. Here are the steps I did:
1. HCl treatment for 1 min. 30 s.
2. Dehydration bake at 120°C for 10 mins.
3. AZ5214 Photoresist spin coating at 4000 rpm, 40 s.
4. Hotplate bake at 90°C for 2 mins.
5. UV exposure with square pattern for 1.2 s (measured lamp power is about 25.2 mW).
6. Positive resist development using NMD for 1 min. in which I swirl the sample only after 30 s of dipping.
I attached a photo showing the result. There are photoresist smears that remain on the cell perimeter, connected to the edge beads. I only need the square pattern to be left on the sample.
Our spin coater is limited to 7500 rpm, so I could not do the fast bead removal. Aside from using an edge bead removing solution, which parameters and/or steps are recommended to be optimized or added while retaining the photoresist thickness and adhering well on the sample?
Thanks in advance for your answers.
In photolithography, forming an edge bead during spin coating can be problematic. The edge bead can be smeared during removal, causing the issue you're observing. Here are a few suggestions to mitigate this problem:
Remember, when making any changes, it's important only to change one parameter at a time so that you can identify what is having an effect. Good luck with your process optimization!
Non-circular samples are not compatible with spin coating. You cannot completely get rid of edge beads in a consistent manner. However, there are some tricks to reduce them:
- use a recessed chuck
- use a carrier wafer to reduce Bernoulli effect
- remove the edge bead with a cleanroom swab or razor before the soft bake
- after the first exposure, cover your pattern with Al foil and perform additional flood exposure with only the edges of the sample exposed. This will enhance the edge bead development rate, but some overdevelopment will still be necessary
However, employing spray/dip coating or working with full wafers will be the best solution.
Thank you very much everyone. After considering your suggestions and a series of optimization experiments, I was able to resolve the edge bead smears by image reversal, in which my spin coating setting was a sequence of 500 rpm for 10 s (slow speed to spread) then 4000 rpm for 40 s (for my desired film thickness). Although the edge bead on the corners still remained after image reversal, careful swab with acetone worked for me. Again many thanks for taking time making suggestions.
Cheers,
Bern
- Badges
- Science topic
- Similar topics
- Agriculture
- Horticulture
- Seeds