--=====================_1648843281==_.ALT
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Mems Clearinghouse members:
I want to thank all who helped me solve the etch problem that I posted last
month. I received many helpful suggestions and comments. I particularly=
want
to thank Uwe Breng from Freiburg, Germany for the solution and Paul=
Danielson,
from Corning for confirming it. =20
Sorry for the length, but several people asked for me to share what I=
learned,
and I thought more might benefit. I, for one, would like to see the=
solutions
people have for their problems that they post. When I search through the
postings I'm often looking for someone who has run into similar challenges,
but
I usually only find questions. Maybe someone would disagree. What do you
think? Thank you for your help!
My previous posting:
I am etching a microfluidics channel into Corning 2947 soda lime glass using=
a
1:1 mixture of (BOE 6:1) : (BOE 10:1) from Baker. The etch profile is
unpredictable, often with strange crystal-like bumps within the etched
channel. These bumps are definitely glass, not salt. No amount of rinsing=
or
cleaning removes it. Can anyone explain the strange etch phenomenon in this
picture?
http://neuron.chem.psu.edu/particulates_or_what.jpg
This channel was etched (unagitated) for 24 minutes in the above etch
mixture.=20
The depth of etch is ~10=B5m, and the crystal-looking features are raised=
bumps,
measuring up to 6-8=B5m in height. I use a gold/chrome sacrificial layer to
mask
the etch, and I am certain that all the resist/masking was sufficiently
removed
where the bumps are before etching.
Can anyone help? I have asked everyone in our cleanroom, but nobody knows.=
=20
They don't use soda lime.
Thank you for helping!
Drew Manica
The solution:
The bumps are called "dendritic crystals" formed by the precipitation of
insoluble fluoride salts. These would be salts of Ca, Mg, Li, which
precipitate onto the glass, effectively masking the etch. The crystals=
cannot
be removed by physical means nor by HF. I believe it's important to note=
that
while the impurities in the glass are the cause of the bumps, it's not just=
a
random problem of hitting hard spots in the glass; instead it's something=
that
can be avoided by adding HCl to dilute buffered HF. =20
Stjernstrom and Roeraade showed this phenomenon using EDAX, however, my
picture shows a much more obvious crystallization problem extending into the
channel. (See Journal of Micromechanics and Microengineering, 1998, 8,
33-38.) =20
The free chloride forms soluble salts with Ca, Mg, and Li and other earth
metals, preventing the fluoride salts from forming. Stjernstrom reports
masking with VERY hard baked photoresist, but I observed severe undercutting
and slow etch rates strictly using their protocol, which uses a 1:2:17
ratio of
[(BOE 6:1) : (HCl) : (H2O)]. (My channel etched 2 microns deep and 185
microns
total laterally.) By masking with Cr/Au sacrificial layers and changing the
composition to 1:2:4 and etching the plate upside down while stirring the
solution @ 250 rpms for 7 minutes, the channel etched 6 microns deep and 15
microns total laterally. The channel had no crystals at all, and the
sidewalls
were clean. =20
Thanks again,
Drew Manica
**********************************
=B7=B7=B7=B7=B7=B7=B7=B7
"We ought to be no less persuaded that the propitious smiles of Heaven=20
can never be expected on a nation that disregards the eternal rules of=20
order and right which Heaven itself has ordained."=20
--George Washington
--=====================_1648843281==_.ALT
Content-Type: text/html; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Mems Clearinghouse members:
I want to thank all who helped me solve the etch problem that I posted
last month. I received many helpful suggestions and comments.
I particularly want to thank Uwe Breng from Freiburg, Germany for the
solution and Paul Danielson, from Corning for confirming it.
Sorry for the length, but several people asked for me to share what I
learned, and I thought more might benefit. I, for one, would like
to see the solutions people have for their problems that they post.
When I search through the postings I'm often looking for someone who has
run into similar challenges, but I usually only find questions.
Maybe someone would disagree. What do you think? Thank you
for your help!
My previous posting:
I am etching a microfluidics
channel into Corning 2947 soda lime glass using a 1:1 mixture of (BOE
6:1) : (BOE 10:1) from Baker. The etch profile is unpredictable,
often with strange crystal-like bumps within the etched channel.
These bumps are definitely glass, not salt. No amount of rinsing or
cleaning removes it. Can anyone explain the strange etch phenomenon in
this picture?
http://neuron.chem.psu.edu/particulates_or_what.jpg
This channel was etched (unagitated)
for 24 minutes in the above etch mixture. The depth of etch is
~10=B5m, and the crystal-looking features are raised bumps, measuring up to
6-8=B5m in height. I use a gold/chrome sacrificial layer to mask the
etch, and I am certain that all the resist/masking was sufficiently
removed where the bumps are before etching.
Can anyone help? I have asked everyone in our cleanroom, but nobody
knows. They don't use soda lime.
Thank you for helping!
Drew Manica
The solution:
The bumps are called "dendritic crystals" formed by
the precipitation of insoluble fluoride salts. These would be salts
of Ca, Mg, Li, which precipitate onto the glass, effectively masking the
etch. The crystals cannot be removed by physical means nor by
HF. I believe it's important to note that while the impurities in
the glass are the cause of the bumps, it's not just a random problem of
hitting hard spots in the glass; instead it's something that can be
avoided by adding HCl to dilute buffered HF.
Stjernstrom and Roeraade showed
this phenomenon using EDAX, however, my picture shows a much more obvious
crystallization problem extending into the channel. (See Journal of
Micromechanics and Microengineering, 1998, 8, 33-38.)
The free chloride forms soluble salts with Ca, Mg, and Li and
other earth metals, preventing the fluoride salts from forming.
Stjernstrom reports masking with VERY hard baked photoresist, but I
observed severe undercutting and slow etch rates strictly using their
protocol, which uses a 1:2:17 ratio of [(BOE 6:1) : (HCl) : (H2O)].
(My channel etched 2 microns deep and 185 microns total laterally.)
By masking with Cr/Au sacrificial layers and changing the composition to
1:2:4 and etching the plate upside down while stirring the solution @ 250
rpms for 7 minutes, the channel etched 6 microns deep and 15 microns
total laterally. The channel had no crystals at all, and the
sidewalls were clean.
Thanks again,
Drew Manica
**********************************
=B7=B7=B7=B7=B7=B7=B7=B7
"We ought to be no less persuaded that the propitious smiles of
Heaven
can never be expected on a nation that disregards the eternal rules of
order and right which Heaven itself has ordained."
--George Washington
--=====================_1648843281==_.ALT--