If the unexposed areas are not being washed off, then it could mean:

(1) your exposure dose is too high, causing the masked areas to become
exposed due to diffraction and/or scattered light; (2) your developer
could be too weak. Microchem does sell an "SU-8 Developer", so you
might want to check if that is the same as EC-11. Also, the age of the
developer may matter

Cracks near the edges are due to thermal stress. I have found most of
the cracks happen around 75C. Moving it from one hotplate to another
is not the best approach.  When you pick up the wafer from one
hotplate, it will rapidly cool down, then when you put it down on the
other hot plate you are rapidly heating it. These are unnecessary
up/down ramps. Leave it on one hot plate, and ramp the temperature
gradually. I use 5C every two minutes, both up and down. Yes, it is a
long ramp, but stress cracks will be greatly reduced.

On Thu, Jan 12, 2012 at 9:52 AM, Yingtao Tian  wrote:
> Hi Andrew,
>
> I was trying to say 'the 4 um (unexposed) features could not be fully
developed and I did not get a good structure'. The resist layer looked ok after
development, I mean, there was no sign the entire SU-8 layer was washed away.
>
> When you say 'I have to make sure the exposure gap is minimized', did you mean
I should expose the wafer after soft bake as soon as possible?
>
> By the way, I did not mention in my previous email that I found lots of micro-
cracks near the edge of the features. I guess these cracks must be related to
the soft and hard bake processes. I was aware of the stress and cracks in SU-8,
so I use two steps to bake and cool down. I put the wafer on the 65 C hotplate
for 1 min, then 95 C hotplate for 3 mins, then put it back to the 65 C hotplate
for half minute, then a piece of tissue. Do you have any comment on this?
>
> Thanks a lot!
>
> Yingtao