Sounds like the sort of bow you can get with a Polyimide layer on top of a
Silicon wafer after baking.  The differences in thermal coefficient of expansion
between Polyimide and Silicon causes a severe bow if the wafer is ramped up too
fast and or ramped down too fast.  A slow ramp up and down fixed this problem.
We used 4 degrees C per minute up and 4 degrees per minute down.  This slow ramp
rate allows the different material to accommodate the different expansion rates
with limited bow.  The temperatures are about the same typical Polyimide bake
temperature is about 400 to 450 degrees C in a vacuum.

    Bill Moffat, CEO
Yield Engineering Systems, Inc.
203-A Lawrence Drive, Livermore, CA  94551-5152
(925) 373-8353

www.yieldengineering.com

-----Original Message-----
From: mems-talk-bounces@memsnet.org [mailto:mems-talk-bounces@memsnet.org] On
Behalf Of Daniela Kögler
Sent: Wednesday, September 24, 2008 6:28 AM
To: 'General MEMS discussion'
Subject: [mems-talk] Anodic bonding

Hi,

I have to anodically bond SOI-Wafers on Borosilkatglas-Wafers. The parameters I
use are 430 °C and 1000 V.

The problem is that the wafer package after the bond process has a very high bow
(220 um +).

Does anybody now how to reduce the bow?

Thanks!
Best regards
Daniela