Ashwini,

You may check pressure differential first as a possible cause. If this is not
possible then, it sounds like you are right, a compressive stress in the
remaining silicon is causing the bow. This is typical of ebeam evaporated Al, in
my experience, as well as some amorphous silicon films.

In order to measure the stress, you may try to build a beam structure, fixed on
both ends, about 20-50 microns wide into the device. This may give a more
accurate read of the stress (one dimensional). Check Roark's for the analysis.

Rob MacDonald
[email protected]


Dear Friends,

Can anybody suggest how to calculate stress developed in Anodically
bonded thinned Silicon wafer with slelectively unbonded region.

We are observing the following:

step 1: pockets are etched  in silicon, and bonded to glass from
etched side (anodic bonding)
pocket dimension
depth: 5 micron
length and width 500 micron

step 2: wafer is thinned from backside in DRIE.
It is observed that the pocket region(unbonded region) pops up, when
wafer is thinned to 30 micron

Is it only due to stress related to Anodic bonding, or it is also due
to the fact that silicon in thinned.

In any case is there any way to analytically find out the stress.

Ashwini.