Hi Sonia,

I don't know if the ultrasounds are enough, but I would try stirring the bath
(with a magnetic stirrer, for example) in order to help the H2 bubbles formed in
the etched surface to go away.

I would also recommend an immediate washing with DI water of the wafer after
pulling it out from the bath to avoid "clouds" formed in the surface. I think it
could also help using a lower concentration (20% wt KOH) etchant.

I don't know if IPA helps, I have never used it for KOH dilution. I use SiO2 for
the mask.

H2 bubles travel upwards to the surface of the bath, and if you have to etch big
surfaces maybe you have "shadows" over the structures, in the direction of the
flow of the H2 bubbles, especially if you have high rate of aspect structures.

Best regards,

Pablo.

Pablo de la Fuente Prado
pdlfuente@ceit.es

-----Original Message-----
From: mems-talk-bounces@memsnet.org
[mailto:mems-talk-bounces@memsnet.org]On Behalf Of GARCIA BLANCO Sonia
Sent: 11 November 2005 15:47
To: 'General MEMS discussion'
Subject: [mems-talk] KOH etch fo silicon with very low surface roughness

Hi,

I am starting to setup a KOH silicon etch bath. I would like to get 54.7deg
and 45deg grooves with very good surface roughness (less than 30nm would be
ideal).

Could anybody give me some suggestions about an starting point for my
experiments and an idea of what are the most important parameters and in
which way they will influence my etch?

(I did a first trial with 35%wt KOH, 20% vol IPA, 80C, ultrasounds, very
strong agitation and orientation of the sample vertical with the patterns
facing the current of the flow and the surface roughness was really bad. I
etched 80microns in 1.5h, so I think the etch rate is as reported in the
literature. I used a PECVD silicon nitride mask).