Those comments are really impressive.
I agree with those points especially the oxidation part.

Yi Tao
University of Arkansas

----- Original Message -----
From: "Brubaker Chad" 
To: 
Sent: Friday, July 12, 2002 11:35 AM
Subject: RE: [mems-talk] Al-Si wafer level bonding?


> Dear Yahong,
>
> There are a couple of major reasons for the use of Au-Si eutectic
bonding
> rather than Al-Si eutectic bonding.
>
> The first, as you mentioned, is the temperature.  Especially with
wafer
> level bonding, temperature becomes an extremely critical factor.
>
> First of all, the Al-Si eutectic temperature (577 C) is right on
(and in
> some cases, beyond) the edge of the operational range of most
commercial
> bonding systems.
>
> Additionally, one thing that should always be considered for wafer
level
> bonding is the TCE mismatch of all materials involved.  Because the
> temperature of bond is so much higher for the Al-Si eutectic, as
opposed to
> the Au-Si Eutectic (363 C), the effects of TCE mismatch become more
> pronounced.  Even if you are bonding like material to like material
(two
> silicon wafers, for instance), the intermediate later, Aluminum, has
a
> greatly different TCE from that of silicon.  If the bond is
performed at
> ~600 C, then the amount of stress in the film will be severe when
the
> material is returned to room temperature.
>
> Another major reason for Au over Al has to do with the chemical
properties
> of the two materials.
>
> When it comes to Au-Si eutectic bonding, our experience has been
that oxide
> (even native oxide) on the silicon surface is to be avoided, as it
forms
> something of a diffusion barrier to the migration of Au into the Si
(or vice
> versa).  Our typical recommendation is to perform a brief BOE dip to
strip
> all native oxide prior to bonding.  Then, we advise a 15 minute
window
> between the BOE dip and the removal of the silicon from an oxygen
atmosphere
> (by pumping down in vacuum, and replacing with inert atmosphere, for
> instance).  However, only the silicon surface is of concern, since
gold will
> not form an oxide.
>
> Aluminum will actually form an oxide more readily than will silicon.
This
> makes the process more difficult to work as well.  In cases where
direct
> Al-Al bonding has been performed, the substrates were initially
cured in a
> reducing atmosphere (forming gas) prior to bonding (in this case,
> thermocompression bonding rather than eutectic).
>
> Best Regards,
>
> Chad Brubaker
>
>
>
> EV Group-Technology, Tel: (602) 437 9492 x 119, Fax: (602) 437 9435
> E-mail: C.Brubaker@evgroup.com,  Web: www.EVGroup.com, 07/12/02
>
>  -----Original Message-----
> From: Yahong Yao [mailto:hhelen@hotmail.com]
> Sent: Thursday, July 11, 2002 10:13 PM
> To: mems-talk@memsnet.org
> Subject: [mems-talk] Al-Si wafer level bonding?
>
> Hello Folks,
>
> Could anyone give comparison between Au-Si bonding and Al-Si bonding
on
> wafer level?  I have seen many people do Au-Si bonding but not many
do
> Al-Si bonding.  Besides Al-Si eutectic temperature is higher than
Au-Si,
> what else reason makes it not as popular as Au-Si?  Bonding
strength?
>
> I would think Al is IC compatible and it is much easier to process
in an IC
> fab.  Any comment is highly valued.  If one can tell me some
reference
> papers talking about Al-Si bonding, that will be very much
appreciated.
>
> Regards,
> Yahong
>
> _________________________________________________________________
> Join the worlds largest e-mail service with MSN Hotmail.
> http://www.hotmail.com
> _______________________________________________
> MEMS-talk@memsnet.org mailing list: to unsubscribe or change your
list
> options, visit
http://mail.mems-exchange.org/mailman/listinfo/mems-talk
> Hosted by the MEMS Exchange, providers of MEMS processing services.
> Visit us at http://www.memsnet.org/
> _______________________________________________
> MEMS-talk@memsnet.org mailing list: to unsubscribe or change your
list
> options, visit
http://mail.mems-exchange.org/mailman/listinfo/mems-talk
> Hosted by the MEMS Exchange, providers of MEMS processing services.
> Visit us at http://www.memsnet.org/