On Thu, Feb 22, 1996 10:48:42 AM at by way of [email protected] wrote:
>Here at the university of CIncinnati we are trying to etch (110) wafers
and more
>thaan 700 um deep and the trenches are 100 um wide.  We are have problems
of
>the side walls not being straight.  we observe a wavyness on the surface.
>These are thermally grown oxide.  Nitride seems to work better but does
not
>completly remove this problem.  The trenches are 1 inch long.  We observe
>an increase in the underetch of the trenches.  Wonder if you experienced
>such problems.
>
It sounds very much like a problem we've had once. After 4 to 6 different
experiments to track down the cause (e.g., TMAH vs KOH, etc.), we finally
concluded that it was some kind of defect or impurity in the p-type wafers
(relatively lightly doped). The problem was most severe after a substantial
thermal cycle like growing an oxide. Try using a different batch of wafers
(different manufacturer) and see if that produces different results - this was
the only solution we found.

Good luck!

Perry