Biomedical Engineering Reference
In-Depth Information
during heating, indicating the microfingers slightly open owing to bimorph thermal effect.
A hysteresis is observed during the thermal cycle due to differences in forward and mar-
tensitic transformations upon heating/cooling (i.e., different contents of martensite and
austenite at a certain temperature during thermal cycling).
A rough estimation according to Equation 9.13 indicates that if the thickness ratio of TiNi/
DLC is 1, there will be a maximum bending effect. However, the thickness of TiNi should be
larger than a few hundred nanometers, below which the SME will be too weak for an efficient
actuation (Fu et al., 2006). When the film is too thin, surface oxide and film/substrate interfacial
diffusion layers exert dominant constraining effect that renders high residual stress and low
recovery capabilities (Ishida et al., 2003). The surface oxide and interdiffusion layer restricts
the phase transformation, alters the chemical stoichiometry of the remaining TiNi film, which
effectively reduces the volume of the material available for phase transformation. On the other
hand, there is also a limit to the DLC thickness. When the thickness of DLC is above 100 nm,
the DLC layer may peel off from the Si substrate because of intrinsic stress. This has severely
restricted the usage of DLC of a few hundred nanometers thick.
Microcages of five, six, and seven fingers were designed. The width of the fingers and the
gap between the beams were 4
μ
m. The fingers were connected to each other with bond pads.
The central part of the microcage was large enough so that it remained attached to the sub-
strate after the fingers were released from the substrate. A DLC film of 100 nm was deposited
on Si substrate using a filtered cathodic vacuum arc (FCVA) method with a graphite source.
The compressive stress of the film was 5 GPa as determined via curvature measurement. A
TiNi film of 800 nm thick was deposited on top of the DLC layer by magnetron sputtering in
an argon gas environment at a pressure of 0.8 mTorr from a Ti
50
Ni
50
target (using a 400-W r.f.
power) and a 99.99% pure Ti target (using a 70-W dc power). Post-annealing of the TiNi/DLC
bimorph layer was performed at 480°C for 30 min in a high vacuum condition for crystalliza-
tion. The DLC/TiNi microcage was fabricated by photolithographically patterning of 4.8
μ
m
thick AZ4562 photoresist on top of the TiNi films. An HF/HNO
3
/H
2
O (1:1:20) solution was
employed to etch the TiNi films to form the microcage patterns. The exposed DLC underlayer
was etched off in oxygen plasma at a flow rate of 80 sccm and a power of 100 W. A deep reac-
tive ion etching machine was used to isotropically etch the silicon substrate with SF
6
plasma at
a flow rate of 70 sccm and pressure of 72 mTorr. The coil and platen powers were set at 50 and
100 W, respectively. A time-controlled etching was performed to release the fingers leaving
the middle parts of the microcage remained attached to the Si substrate. The actuation perfor-
mance of the released microgrippers was evaluated using a peltier device (with a temperature
range from 5°C to 100°C). The displacements of the TiNi pattern were measured using a video
camera from the top view.
An SEM morphology of the fabricated TiNi/DLC microcages on a 4-inch silicon wafer is
shown in Figure 9.69a (Fu et al., 2008). The microcages have different finger numbers and
beam lengths. After released from the Si substrate, the microcages show significant curl-
ing up of the microfingers, depending on the beam lengths. The fabricated seven-inger
microcages with different beam lengths. As designed, with increase in beam length, the
microfinger patterns of the microcages change from fully open, under-closed to over-
closed (Fu et al., 2008).
Figure 9.70 shows the top-view optical images of the deformation behavior of a five-finger
microcage upon heating (Fu et al., 2008). Actuation of the microcage is mainly determined
by SME. With temperature increased above martensitic start transformation temperature
(about 50 to 60°C), martensite (loose structure) changes to austenite (a dense structure),
causing the closing of microcages, and capturing an object. Further increase in tempera-
ture above 100°C causes the slight opening of the microfingers due to thermal effects.
