>      I've been working on releasing SiO2 cantilever beams
> from Si with KOH wet etching. SiO2 beams was attached to Si
> layer at one side. The temperature is 80 degree-C and
> stirring is 220RPM. The etching time is about 15mins. By
> caculation, only 15um Si layer needs to be etched away to
> release cantilever beam. I didn't get satisfactory etching
> selectivity on Si/SiO2 etching rate. SiO2 beams began to be
> etched even before Si was cleaned. The resulted cantilever
> beam wasn't integrated. Do you have some suggestions about
> how to increase the etching selectivity or any ideas about
> releasing SiO2 cantilever beams from Si. Thanks.
>

There are many parameters to control to get a good etch using SiO2 as an
etch mask. A good thermal oxide is preferred over a plasma oxide. Some
people like to "condition" their KOH solutions by etching an entire scrap
wafer before they etch a device wafer. I think I've even heard speculation
about the role that iron may play in attacking oxides (if you're using a
stainless steel etch tank). Some surfactants can increase the etch rate of
SiO2, so if you're using a soapy mixture to clean the beaker before the
etch, you need to rinse and rinse and rinse to make sure you get rid of all
surfactants.

Kendall and Shoultz wrote a very nice chapter on KOH etching in one of the
SPIE microlithography books (D.L. Kendall and Robert A. Shoultz, "Wet
Chemical Etching of Silicon and SiO2, and Ten Challenges for
Micromachiners", in Handbook of Microlithography, Micromachining, and
Microfabrication, Volume 2: Micromachining and Microfabrication, ed. P.
Rai-Choudhury, SPIE Press, 1997.) To maximize the selectivity of KOH, you
should use a concentration of 14 wt% KOH. Lower etch temperatures also
increase the selectivity.

An empirical formula from Kendall and Shoultz for etch rates is

R_si = 2.6E6*(W^2.5)exp[-(W/300 + 0.48)/(k)(T+273)

R_sio2 = 2.2E9*W^(1.5E-4*W^2.15 + 1)exp[-(0.795+6E-6*W^2.5)/k(T+273)]

Etch rates, R, in µm/hr. W is wt% KOH. T is temperature in Celsius. k is
Boltzmann's constant 8.617E-5 eV/K.

One last note. To get an accurate measure of wt%, it's best to use density
or specific gravity measurements of your solution. KOH pellets will absorb
water very quickly (especially in Louisiana!) which will screw up your
measurements if you're trying to weigh the pellets and mix them into
solution.

Bill Eaton, Ph.D.
Materials & Analysis Manager
NP Photonics
5706 Corsa Avenue, Suite 100
Westlake Village, CA  91362
Voice:  (818) 991-7044 x211
eFax:   (503) 214-5559
mailto://[email protected]
www.parvenutech.com