I use a collimated Hg lamp with a similar spec to yours.
50 microns is certainly no problem - I don't generally do smaller than
this but would expect 10 microns to be feasible as long as you can
ensure good contact between your mask and the substrate (make sure the
printed emulsion side is in contact with the substrate otherwise there
will be a gap). However I am not sure if 25,400 dpi is good enough for
10 microns - the service I use can do 64,000 dpi but for this they
quote a minimum feature size of 8 microns.
The UV lamp I use has IR filters which prevents heating of the SU-8
(otherwise you may need to put gaps in your exposure time to allow it
to cool).  It also has a long pass filter to cut down "T-topping"
which is especially important for thicker layers and to allow good
release of moulded structures.


On 03/01/2008, Dhananjaya Dendukuri  wrote:
> Hello everyone,
>
> I am looking to make masters for some single-layer microfluidic devices
> using SU-8 2050 photoresist on 3" Silicon wafers.  I am trying to see if
> I can make do with a UV(Hg) lamp based flood exposure system that
> provides a collimated 4" X 4" beam with <5% intensity variation and a
> beam divergence angle of 2.6 degrees.  The idea is to perform contact
> mode photolithography but without the help of a full-fledged mask
> aligner.  The feature sizes that I will be printing are down to 10
> microns but typically 50 microns and greater. I will be using
> transparency masks that have been printed at greater than 25,400 dpi.  I
> want to know if any of you have performed i-line photolithography with
> SU-8 using just a UV lamp (not an aligner)?   If so what sort of
> resolution have you achieved with the device features?
>