Leon,

Plasma bonding is a little bit of a misnomer - plasma is not typically present
during the bonding process at all.  Instead, plasma is used to activate the
surfaces - increasing the surface energy of the interface.

This technique has been used extensively to enhance silicon direct bonding (also
called fusion bonding).  The primary benefit seen is that it greatly reduces the
annealing temperature required to finalize the bond (typically, direct bonding
of silicon requires an anneal in the range of 1100ºC to make the bond permanent
- plasma activation allows annealing at 200 - 300ºC).

I'm also not sure if plasma can help with the bonding of aluminum; if we are
looking at a direct-bond mechanism, aluminum is typically too rough (for silicon
direct bonding, surface roughness has to be <2nm RMS - 0.5 nm is better).
However, aluminum bonding will typically follow a diffusion scheme, and the
added energy may enhance this diffusion.

Best Regards,

Chad Brubaker

EV Group
invent * innovate * implement
Senior Process Technology Engineer - Direct: +1 (480) 305 2414, Main: +1 (480)
305 2400 Fax: +1 (480) 305 2401
Cell: +1 (602) 321 6071
E-Mail: C.Brubaker@EVGroup.com, Web: www.EVGroup.com

-----Original Message-----
From: mems-talk-bounces@memsnet.org [mailto:mems-talk-bounces@memsnet.org] On
Behalf Of Leon Nathaniel Maurer
Sent: Tuesday, July 07, 2009 7:04 PM
To: mems-talk@memsnet.org
Subject: Re: [mems-talk] Al-Si wafer bonding

Thanks for the advice everyone.

@Joe,

I knew nothing about that, so thanks for the heads up. I'll have to look in to
it. I did a test run with Si wafers without Al, and they bonded nicely, so
that's an encouraging sign (I'm just working with Si at the moment -- sapphire
will come later if needed).

@Bill,

What is plasma bonding? (Since I've started on this project, I hear about new
types of bonding every day!)

@Felix,

The substrates are ion milled before we sputter the Al on -- we haven't had any
problem with the Al not adhering. Right before bonding (after the Al is on) we
don't do much of a clean (the wafer bonding station I have access to has this
fancy cleaning system, but it's only setup with DI water... so that's what I
use) My understanding is that (given a relatively clean wafer) the native oxide
on the Al does more than anything else to make bonding difficult, but we don't
have a good way to get rid of it since it grows back quickly. Thoughts?

Thanks again.
-Leon