Insufficient development does sound likely. 3 - 4 mins may not be
enough especially for your structure.

A well enough exposed/crosslinked structure should withstand the
developer for much longer periods than the suggested datasheet times.
I also find putting in fresh developer after the IPA rinse helps
remove partially exposed resist.


2008/10/27 Liang Zhao :
> Hi All,
> I am currently working on SU-8 microfabrication. My propose is fabricating a
> PDMS seive for immobilizing cells by using SU-8 mold. We designed a
> microchannel with a seive pattern that consist of 3 short arc, gap between
> them is 8 μm, middle arc is 20μm in width and 80μm in length and the two
> short arcs located in each sides of the middle arc with much smaller size in
> length. we do microfabrication with SU-8 2050, and we hope that we could get
> a PDMS channel with a hight of 50μm. Unfortunately, during our fabrication
> process we find that it's hard to get a sieve pattern that thoughly from top
> to bottom. By using this SU-8 mold we cannot immoblize the cells, because
> the PDMS sieve doesnot touched the bottom that all cells just flows away.
> We basically attribrite this phenomena to an unefficient development, but
> things does not work when we prolong the develop time. Could any one give me
> some advice? thanks.
>
> PS: our procedurs are as follows:
>
> ・         Spin PR (SU8 2050): 500rpm 15s and 3000 rpm 40s (expected PR film
> thickness 50μm)
> ・         Soft bake: 65C / 5min, 95C / 15 min
> ・         Expose: 15 seconds or 10 seconds
> ・         Post Expose bake (PEB): 65C / 5min, 95C / 15 min
> ・         Development: 3~4 min with agitation in SU-8 developer
>  .       Hard bake (Cure): 200C in 20 mins
>
> lowing the temprature slowly after each baking step.
>
> --
> Liang Zhao
> School of Chemistry and Chemical Engineering, Nanjing University