Page 153 - Microsensors, MEMS and Smart Devices - Gardner Varadhan and Awadelkarim

ETCH-STOP TECHNIQUES 133

Worked Example E5.4: Fabrication of Cantilever Beams
Objective:

The objective is to fabricate a cantilever beam that is a few microns thick on a p-type
silicon that supports wafers using electrochemical etching (Linden et al. 1989).

Process Flow:

4
The starting silicon wafers used are again standard commercial, 280 um thick , (100)-
oriented silicon wafers. The wafers are boron-doped with a resistivity of 7 to 10 £2-cm,
–3
15
which corresponds to a doping concentration approximately 1.50 × 10 cm . Two
different techniques can be used to define the beam.
Technique A: The diffused-pattern technique
In the diffused-pattern technique, an n-type (a diffused n-layer of n-doping concen-
17 –3
tration ~10 cm , as described in Worked Example 5.3) pattern that describes the
beam is diffused into the wafer. The wafer is masked with SiO 2 grown in wet O 2
at 1000°C. Then, by stripping the patterning oxide and by performing the electro-
chemical etch, the beam will be totally defined by the diffusion process. The steps
for the beam fabrication using this method are shown in Figure 5.15. The etching is
performed in an apparatus similar to the one described in Worked Example 5.3 (see
Figure 5.14(b)).
In this technique, the pattern is diffused, thus resulting in an n-type pattern defining
the beam. Because of lateral diffusion during the drive-in, this technique will result in
beams with 'rounded' corners (Figure 5.15 (c)); this means that we will not end up with
the sharp corners that are usually associated with anisotropic etching.

Technique B: The etched-pattern technique

–3
In the etched pattern technique, an n-type layer (dopant concentration ~10 17 cm ) is
diffused over the entire wafer. Oxidation is performed simultaneously with the drive-
in of the dopant. The oxide is patterned to cover the surfaces of the forming beam.
The wafer is then immersed in the KOH etching solution as in Worked Example 5.3
(Figure 5.14(b)) without any bias, until the n-type layer is etched through and the p-type
layer is exposed. Following this nonbias etch, a voltage bias is applied and the beam
will be etched correctly. The processing sequence for the beam fabrication using the
etched-pattern technique is outlined in Figure 5.16(a) through to (d).
This method has all the advantages of the anisotropic KOH etch, that is, giving
perfectly sharp corners defined by the (111) crystal planes. On the other hand, there
will be under-etching alongside the beam under the oxide mask when etching the top
layer without bias (if the beam is parallel to the (100) planes). In addition, the tip of the
beam will be 'polygonal' instead of being square in shape (Figure 5.16(e)) because the
etchant finds faster etching planes at the convex corners. Another possible advantage is
the fact that the beam surfaces are protected by the SiO2 throughout the whole etching

4
Wafer thickness normally increases with diameter, so 280 )im-thick wafers (3" diameter) are now hard to
source; however, larger (e.g. 4") wafers can be thinned down.