Hello, list.  I am building an array of pillars (spikes, rods, poles,
etc.) that are 20um in diameter, 200um tall, and 430um apart.  There is
nothing around an 8x8 group of pillars for about 8mm and there is a
large amount of active etch area (about 60-70%) on the wafer (mostly the
area to be cleared surrounding the pillars).  I am having difficulty
preventing tapering (and eventual collapse) toward the bottom of the
pillars during the etch using a bosch process on an STS DRIE machine.
Grass is also a problem, as the longer the passivation step, the more
grass is created it seems.  Uniformity is a problem as well.  I have
tried 12.5 sec etch cycle time and 7 sec passivation cycle time.
  I have tried decreasing the etched surface area by isolating the
groups of pillars and etching them, and then covering them and etching
the bulk of the wafer with a second mask, but there have been problems
making a clean boundary between these steps that does not leave silicon
behind (no net etch) or attack the 6kA of oxide (etch stop) beneath
(double net etch).
  Is there a solution to the undercutting and grass problem when a large
area of silicon needs to be clered?  Maybe a suggestion as to how to do
a more precise double etch process?  thanks for your help
 Andrew