Hi Mikael,

Regarding using SU-8: if you form your 50µm-thick from patterned SU-8 that
hasn't been fully cross-linked yet (i.e. spin-coat, soft bake, expose,
lightly post-expose bake, and develop), you might be able to bond it
directly to the other glass substrate by simultaneously pressing the two
substrates together and applying heat.  This should cause the SU-8 to reflow
a bit before it fully crosslinks.  I had some success using this technique
to bond silicon to glass when I needed an air gap of a specific size in
between the patterned SU-8 features, and it might be that when optimized
this technique could produce a bond suitable for microfluidics.

Another technique you might try is to spin your 50µm of SU-8, pattern and
develop (again lightly post-exposure baking), and then basically repeat the
process on the other glass substrate but using a very thin film of SU-8 and
patterning it such that your electrodes are exposed.  You'll want to use a
low-viscosity SU-8 formulation for the thin layer.  Then, when you align and
bond, you'll have two SU-8 surfaces mated together which would probably give
you a better bond than SU-8 reflowing onto glass.  I have not tried this
method personally, but it is a small extension of the work presented here:

Li, Sheng; Carl B. Freidhoff; Robert M. Young; Reza Ghodssi. Fabrication
of Micronozzles Using Low-Temperature Wafer-Level Bonding with SU-8.
*Journal of
Micromechanics and Microengineering*, *2003*, 13, 732-738.

Good luck,

Brian C. Stahl
Graduate Student Researcher
UCSB Materials Research Laboratory
brian.stahl@gmail.com / bstahl@mrl.ucsb.edu
Cell: (805) 748-5839
Office: MRL 3117A


On Fri, Dec 4, 2009 at 3:56 PM, Mikael Evander  wrote:

> Hi Bob.
>
> Doesn't SU-8 require an additional spun-on layer of SU-8 in order to bond
> after developing? I was looking at SU-8 from the beginning but couldn't
> find a decent way of bonding the structured/developed SU-8 to the second
> glass wafer? I'm looking at a glass - SU-8 - glass structure.
>
> Thanks!
>
> /Mikael