Page 84 -

MEMS Fabrication 3-39

with 64 pixels each. The pixels increase in size from 30 30 µm for the inner circle to 412 412µm for
the outer circle. The circle radius increases exponentially with eccentricity. The center of the chip, called
the fovea, is filled with 104 pixels measuring 30 30 µm placed on an orthogonal pattern. The total chip
area is 11 11 mm. The chip is designed for those applications in which real-time coordination between
the sensory perception and the system control is of prime concern. The main application area is active
vision, and its potential application is expected in robot navigation, surveillance, recognition, and track-
ing. The system can cover a wide field of view with a relatively low number of pixels without sacrificing
the overall resolution and leads to a significant reduction in required image processing hardware and cal-
culation time. The fast, but rather insensitive, large pixels on the rim of the retina chip pick up a sudden
movement in the scenery swiftly (peripheral vision), prompting the robot equipped with this “eye” to
redirect itself in the direction of the movement in order to focus better on the moving object with the
more sensitive fovea pixels.
The area in which MEMS has seen one of its biggest commercial breakthroughs lately is micromachined
mirrors for optical switching in both fiber optic communications and data storage applications. Optical
switches are to optical communications what transistors are to electronic signaling. What makes Si single-
crystal attractive in this case is the optical quality of the Si surface (see example 4 in section 3.12,“Wet Bulk
Micromachining Examples”). The quality of the mirror surface is primordial to obtain very low insertion
loss even after multiple reflections. Wet bulk micromachining has an advantage here over deep reactive ion
etching (DRIE), as the latter leaves lossier Si mirror surfaces due to the inevitable ripples [Hélin et al., 2000].

3.6.3 Biocompatibility of Si

Biocompatibility has both a mechanical and a chemical component. For example, it is well known that
for better biocompatibility a smooth surface and rounded corners are preferable to a rough surface, sharp
corners, and small crevices; that for a given surface finish and shape, carbon is preferable to copper. The
research community remains somewhat divided on the relative contribution of mechanical effects (e.g.,
surface roughness and shape) vs. chemical effects (e.g., Cu vs. C) to the overall bioresponse of an implant.
von Recum’s work seems to demonstrate that micromachining (microtexturing) of surfaces of different
materials may help deconvolute these relative contributions [von Recum and Cooke, 1988].
Some specific examples of substrate microtexturing effects follow. An increase in biosynthetic activity
and mobility of bone cells on polymer films cast from micromachined molds has been identified (this
might have applications for bone regeneration and bioengineering of prosthetics). The regeneration of
severed tendons only on textured surfaces has been demonstrated and might lead to the use of textured
bandages to accelerate healing of tissue. Many examples in the MEMS literature also illustrate directed
growth of nerve cells on micromachined structures. In a typical procedure, a hydrophobic silane pattern
on a glass slide is produced by “lift-off,” where the hydrophobic silane (dimethyl-trichlorosilane) replaces
the more familiar metallization step. Hydrophobic patterning is followed by the deposition of an amino
silane (3-aminoethylaminopropyltrimethoxysilane) in the remaining areas, the amino group forming an
attachment site for proteins [Cooper et al., 1993; Connolly et al., 1992]. Successful chemical patterning of
glass slide surfaces is demonstrated by the growth pattern of cells. Some cells, such as fibroblasts, grow in
hydrophilic areas but have difficulty adhering and spreading in hydrophobic areas. Sharp edge definition
is thus obtained along edges of hydrophilic/hydrophobic interfaces. Using this technique, neuronal growth
can be controlled, and neuronal processes may be induced to follow the straight lines of the original mask
pattern. It has been suggested that this type of approach could be applied to nerve repair techniques. The
method and its range of applications are described in Britland et al. (1990). Surface characteristics of
BIOMEMS materials that should be considered include not only chemistry and surface texture but also
hydrophilicity, charge, polarity and energy, heterogeneous distribution of functional groups, water absorp-
tion, and chain mobility (for a polymer).
The above background on the many aspects determining biocompatibility may explain why literature on
biocompatibility of Si remains somewhat an enigma; it is often hard to understand what is meant when a Si
device is proclaimed to be biocompatible. Maluf, without providing references, claims that preliminary

© 2006 by Taylor & Francis Group, LLC