Try to use mems-talk archives. There are tons of answers and your question is
one of the most answered one.

  Good luck!!

  sokwon

[email protected] wrote:
  Sorry, I didn't see the message until just now. I was trying to get as
high as possible, so su8 2025 was spun at 1000 rpm. The sample was
settled for 5 min and heated on hotplate with temperature ramping up
from room temperature to 95 degree. After 10 min at 95 degree,
temperature is ramped down before 5 min exposure at 4.5 mW/cm^2. The
post exposure baking is the same as pre-bake. Basically, the 10 um gap
is still filled by su8 after development.
Thanks in advance for your suggestion.