It's possible to get 0.5 um lines with this recipe. The values are optimized for sapphire substrates, and the last bake and the development time may need to be adjusted depending on the substrate.

1. Spin LOR-3A in the class 2 room for 60 s at 3000 RPM with 1000 RPM/s acceleration.
2. Bake 5 min at 180 C.
3. Spin S-1813 for 60 s at 3000 RPM with 1000 RPM/s acceleration.
4. Bake 1:15 at 115 C for sapphire substrates.
5. Expose 0.26 seconds in the 5x stepper.
6. Develop for 1:30 in 726-MIF developer using minimal agitation.
7. Rinse in DI water.

• For liftoff processes, soaking overnight in room temperature 1165 followed by ~5 sec of sonication is normally sufficient.

• For etch processes, soaking 1-2 hours in 180 C 1165 is recommended.

• Is your piece larger than the center dot but smaller than the smallest largest vacuum ring?
  • Good! Use the ring-shaped wafer to block the other vacuum rings on the 3“ stage.

• Is your piece largest than the smallest vacuum ring in some dimensions?
  • That's ok! Take cleanroom safe paper, preferably the blue printer kind. Carefully cut out the shape of a 3” wafer without wrinkling the paper. Next, carefully cut out a central region which will fit entirely inside your smallest sample. Use this underneath your sample. This type of paper should be thin enough that you should not need to use a stage of a different height.

Mahmoud Elzouka, July 27, 2015

• LOR thickness should be at least 1 to 1.25 times the target thickness of metal to have the liftoff process complete satisfactorily within 12-24 hours. More LOR may speed up the liftoff process.
• For LOR or DS-K101 liftoff resists the bake temperature and bake time highly affect the amount of undercut, with bake temperature being the dominant factor by far. Read the plots in the following link around p. 3 for exact undercut rates vs bake time and temperature http://microchem.com/pdf/PMGI-Resists-data-sheetV-rhcedit-102206.pdf
• The LOR undercut is visible in the optical microscope, the edge of the LOR and the edge of the top photoresist is offset bu around 500 nm to a couple um based on process parameters.