The image reversed pillars resist thermal flow like silylated resist does.
First expose the pillar regions.  Then image reverse using ammonia technique,
dozens of papers on this technique.  Then flood expose remaining resist, this
is the exposure level that determines the pillar wall angle.  Any angle 22
degrees +/- from vertical is easily achievable.  Then develop away the newly
exposed resist leaving the original exposed and reversed pillars.  This resist
is more resistant to thermal flow.  Flow TEOS around the resist.  Then CMP or
back lap off the hillocks over the pillars.  This exposes the tops of the
pillars and they can be removed with simple Oxygen plasma giving vertical side
walls down to the contacts.  Contact me by phone and I will walk through any
variations and details of reversal down to 800 angstrom lines and spaces.  I
think in later production they used spin on glass.  This process is one we
tried to do in the 80's called the lost contact process but we etched and the
contacts got bigger so we had no benefit.  With the reversal approach this
gave a 40% reduction in chip size and a 60% increase in speed. I think.  My
phone number if you want to hear more details is 408 954 8353.  Bill

-----Original Message-----
From: mmf [mailto:[email protected]]
Sent: Friday, May 17, 2002 1:11 PM
To: [email protected]
Subject: Re: [mems-talk] PDMS pillar shape


Bill,

"Then flowing TEOS around the reversed image photo resist he defined
contacts with vertical side walls..."

How did he do it?

mateen
_______________________________________________
[email protected] 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.mems-exchange.org/