I tested the effect of extremely long immersion in developer and
sometimes the su8 layer peeled off from the coverglass. So, my guess
is, if small amount of developer is used, the developer is saturated
and it takes longer to do the development and the su8 layer starts to
peel off especially from the corner of intersecting channels. It is
interesting that when this happens, the following samples will be
doomed even after a very short period of development in fresh
developer.

The reason I was trying to be quantitative in every aspect is that the
process is not very reproducible for me. So even if I succeeded today,
I am still in doubt.


Thanks.

Best,
Liu Yu

On 1/2/06, Zeta Tak For YU  wrote:
> HI Liu,
>
> Really appreciate your work and feedback.
>
> As I mentioned in previous mail, adding more developer will help. My
> experience is that SU-8 developer saturates with PR quickly or slow down
> the development rate soon. I guess that is why the SU8 manual suggests
> second bath.
>
> In fact I feel interesting why you are so serious about the amount and
> time the developer added. I used a 4" crystallizing dish container for
> 3" Si wafer with SU8-2025 1500-3000 rpm, of about 1cm deep developer
> solution. I also leave it there for 5 minutes, or even 10 minutes more.
> Does it really matter to limit the developer amount and also count the
> time so seriously? I am just curious : >
>