> I am doing anisotropic etching of (110) Si Wafers.
> I use KOH solution with IPA.
> The problem is that at the bottom of (110) surface form line like hills.
> the short axis of this lines points to [100] direction.
> the concentration and dimentions of this fetures become smaller by adding
> IPA but don't disapear.
> I think it has something to do with, may be local, etch rate changes of
> {110}/{100} planes but I dont know how to suppress it.
> Any suggestion will be welcome.

Fairly recently a japanese group published a paper called "Groove depth
uniformization in (110) Si anisotropic etching by ultrasonic wave...". I think
it was a Hilton Head Conference. Authors are K. Ohwada, Y. Negoro,. They used
an ultrasonic cleaner capabel of of "Dynashock" waves. We could not find
such a device in the US. We bougt a pulsed wave cleaner. It does not to
work as nicely as theirs.

Using 55 wt.% KOH gave me almost as good smoothness as on a (100) wafer. Thry
that, but your oxide mask has to be thicker. For these high concentrations
(>35 wt.% or so), planes start to form were the (110) bottom intersects with
the (111) vertical side walls. D. Kendall et al "Orientations of the third
kind: The coming age of (110) Silocon" claims these are (113) type planes.
We could only find it is close to that plane and that it etches slower that
the bottom (110) plane. (That reduces your aspect ratio if you try to etch
narrow slots).
In yet another article "Vertical etching of silicon at very high aspect
ratios" he mentions certain additives that make the bottom essentially
flat, as a reference.
Good luck

Alexander Hoelke
University of Cincinnati
MEMS group