I would suggest two things you should look into:

1) A lot of photoresist defects are actually the product of air
turbulence above the sample and one of the clean rooms I work in just
got a special spinner that uses a bowl to cover the sample as it is
spinning.  I don't know the brand name but if this option attracts
you I can look into it and send you the name.  I should note that you
should *not* try just covering your sample on a normal spinner since
that will probably lead to injury.

2) Is it required that you use PMMA?  The best way I know how to coat
the sidewalls of a sample evenly is with paralene.  The way it is
deposited is very similar to metal evaporation (the paralene is
melted and evaporated in a vacuum chamber) but it is not directional
and deposits everywhere and you can get thick depositions fairly
easily.  Another alternative is using a C4F8 plasma like used in DRIE
etches to protect the sidewall.  I do not know the limitation of
thickness on this however so you will need to experiment.  I am not
certain about the paralene, but I do know that the polymer produced
in the DRIE can be removed by an oxygen plasma so you can control the
directionality of removal (high substrate power for removal on
horizontal surfaces, low power for removal all around).

Hope this helps,

Nicolas Duarte
Ph.D. Candidate at Penn State University


On Oct 24, 2007, at 5:24 PM, Sandip Agarwal wrote:

> I am trying to spin-coat a polymer (1% PMMA-butadiene copolymer in
> toluene) on a substrate that has topographical features that are 50x50
> micron posts and 10 microns tall. I observe striations or ripples that
> run from the center of the substrate to the periphery after
> spin-coating. Any solutions? I have tried varying the concentration
> and
> dynamic dispensing without any success.