Ed:

What you just wrote should be read by everyone in the MEMS community. It is
spot on in every respect. Having been an etch engineer in the 70's I too
notice that a lot of the questions from this group revolves around
technologies that were developed and used in main line processing in the
70's and early 80'. The advent of single wafer plasma processing tended to
make a lot of these techniques obsolete and in some cases they remain only
in old processing minds. Many seem to be relavent for MEMS today as an
example the ability to reflow photoresist and use an oxygen rich process to
make a positive slope on a variety of metal films to make subsequest
depositions like oxide have good step coverage. The same if you need to
increase your step coverage in a dielectric film for conformal coverage with
a subsequent metal for good coverage. Wet etching also has a variety of
tricks like you just outlined to prevent or reduce the amount of lifting of
the photoresist masking material. HMDS vapor prime has to be done properly
in a vacuum oven but I think it works well on many different films for good
photo results. Many of the MEMS specific processes like proper release have
gone thru some interesting challanges like critical CO2 initially to some
other methods currently in use that are more production (cost) worthy. There
are many new challanges ahead for specific processes but in essence many of
the older tried and true techniques work and just need to be made available
to the MEMS community. I think this forum is a start.
Bob Henderson
----- Original Message -----
From: "Edward Sebesta" 
To: "General MEMS discussion" 
Sent: Friday, November 16, 2007 7:01 AM
Subject: RE: [mems-talk] Best resist for HF etching


For SiO2 surfaces, the positive resist won't stick in a flouride
solution unless it is buffered to around pH 7. You need to use BOE,
Buffered Oxide Etch. This is an old standard in wet oxide etching in the
semiconductor industry. It is a solution of ammonium flouride and HF.
Almost, any semiconductor chemical vendor makes them. They come with
surfactants and are called Superwet, etc. This is important for wetting
the wafer and getting good oxide etch, especially for contacts and other
small features and high aspect ratio features.

There is a classic paper on BOE etching back in the 70s on all this
where they did SiO2 etching with positive photoresist under a series of
pHs. It is one of the breakthroughs that enabled semiconductor
processing.

The problem has nothing to do with the type of resist. Any positive
resist at an adequate thickness to cover your topology and not have
pinholes should do, if the solution is buffered. None will stick, unless
there is some radically different positive resist chemistry out there,
with out pH buffering.

Also, you need to make sure your surface is properly prepared. If you
have monolayers of hydrocarbons on it your resist may not stick even
during develop.

The best method is sulfuric peroxide or an acid peroxide solution, good
rinsing and drying, and then Vapor Priming. Use HMDS, Hexamethyldisilane
(?) and it abstracts the OH groups off the surface, the other popular
theory is that it coats a monolayer of bonded organic groups to the
wafer surface, either way it works.

However, if your wafer has metal or something, just a good clean in
peroxide or just plain water will do. Vapor Prime is thought to be
restricted to SiO2 or Silicon, however, I think it works on metal
surfaces by being a form of chemical drying. I used it on metal layers
in photomasking for years. Though it may have been just an extra set of
suspenders.

What concerns me most about this question, which I have nothing against
in itself, is that we are building on what is already known. How is the
MEMS profession using what is already known and building knowledge in
general.

The Semiconductor industry progressed since almost everyone was running
hugh volumes of similar technologies and drove the technology, Moores
law and all that. MEMS has diverse technologies. Are we to be merely
derivative or largely derivative of semiconductor technology? Can we
progress technologically like semiconductor processing/

Ed


-----Original Message-----
From: mems-talk-bounces@memsnet.org
[mailto:mems-talk-bounces@memsnet.org] On Behalf Of Michael A Gingras
Sent: Thursday, November 15, 2007 11:13 AM
To: mems-talk@memsnet.org
Subject: [mems-talk] Best resist for HF etching

Hello,

Me and my colleagues are trying to etch a 4 micron Pyrex layer with HF
using photoresist as a mask. We've seen delamination of the mask as seen
by others. We are currently using a Shipley 7 micron resist and 4:1
HF:H2O etch.

Any recommendations on the best resist and HF concentration to help us
get the whole film etched before delamination? Current the Pyrex is
etching about 300A/s, so we'd need something that would last 2-3
minutes.

Thanks
-Mike