Looking for advice on wafer bonding

Kam Arnold karnold at berkeley.edu
Tue Jul 1 08:48:37 PDT 2008

Direct silicon-silicon fusion bonding is quite strong (can be stronger
than the silicon itself, if wafers are clean, etc.) It is a high
timperature process, though, and highly degraded by particles. The
standard Berkeley fusion bonding process is described here:

A colleague, Erik Shirokoff, has done a lot of wafer bonding research.
He is currently using spun-on BCB, I believe.


On Mon, Jun 30, 2008 at 11:40 PM, ben.jian
<ben.jian at arrayedfiberoptics.com> wrote:
> Dear labmembers,
> I have been doing silicon-silicon wafer bonding for years using gold-silicon
> eutectic bonding.  While this is a reasonable wafer bonding method,  it
> leaves much to be desired.  Briefly,  the bond is not very strong -
> frequently even ultrasonic cleaning can break the bonded chip stack.
> I am looking for a better silicon-silicon wafer bonding process.  Factors to
> consider include process simplicity, bonding yield/strength, tolerance to
> dust particle, etc.  If every step can be performed at SNF,  it would be
> always better.  A lower temperature process is desirable.
> Your kind advice is greatly appreciated.
> Ben Jian

Kam Arnold
Graduate Student Researcher
UC Berkeley Physics Department
351 LeConte Hall
Berkeley, CA 94720-7300
Office: 1 510 643 8161
Fax: 1 510 643 5204

More information about the labmembers mailing list