Hello Robert,

if you have the possibility to choose for bonding either in vacuum or in
atmospheric pressure then prefer the bonding in vacuum. The main reason for
bonding in near vacuum is to avoid the generation of voids.
Voids are appearing due to particles at the bonding interface silicon-glass
which inhibit the bond process locally or caused by air bubbles. Moreover
the bonding front is propagated at several points randomly (depends on
surface properties of your substrates), since you have no control of the
bonding points and the bonding fronts would meet and might crack the glass
(multiple bond fronts colliding).
However if you want to bond at atmospheric pressure Suss MicroTec star
shaped electrode allows you perfect anodic bonds without voids. Several
independent controllable electrical circuits are integrated in this bond
tool, which allows you to start the bond front in the center of the
substrate and control the bond front until it reached the wafer edge and the
bond is finished. Controlling the bonding front in near vacuum is important
too. Therefor our SB6e/8e is well prepared for perfect bonds.

Standard Bonding recipe for HV anodic bonding:

1) load the wafers in your bonder
2) heat up to 400C
3) evacuate your chamber down to 10e-5 torr
4) pull the spacers to bring the substrates in contact
5) apply voltage -800V on glass
6) stop bond when 10 percent of current limit is reached (termination
criteria for bonding)
7) purge the chamber and cool down


Best Regards,
Volkan Cetin

Message: 9
From: Robert Dean 
Subject: [mems-talk] anodic bonding in a near vaccum

I need to anodically bond some die at near vacuum to fabricate a MEMS
device.  I have heard that anodic bonding in a near vacuum (~10e-6 Torr) is
somewhat different that anodic bonding at atmospheric conditions, but I do
not have any data on the procedural differences.  Does anyone have
experience with this and could describe the differences?