Bulk-Metal-Based MEMS Fabricated by Micro-Electro

Bulk-Metal-Based MEMS Fabricated by Micro-Electro
BULK-METAL-BASED MEMS FABRICATED BY MICRO-ELECTRO-DISCHARGE MACHINING
Kenichi Takahata
Yogesh B. Gianchandani
Department of Electrical and Computer Engineering
University of British Columbia, Vancouver, Canada
Email: [email protected]
Department of Electrical Engineering and Computer Science
University of Michigan, Ann Arbor, USA
Abstract—This paper presents the recent development of microelectro-mechanical systems (MEMS) and devices realized by planar
micromachining of metal foils. New types of cardiac stents
including sensor-integrated antenna stents, a micromachined
Kelvin probe with an integrated actuator, an intraluminal flow
sensor cuff, and a mechanically/chemically robust capacitive
pressure sensor are reported. Micro-electro-discharge machining
(µEDM) and the modified processes were used for the fabrication of
the devices with mechanical or electromechanical functionality.
The cost-effective manufacturing is potentially available with the
use of the batch mode µEDM technology that employs
lithographically fabricated microelectrode arrays.
Keywords: MEMS; Bulk metals; Micro-electro-discharge
machining; Stent; Wireless.
I. INTRODUCTION
The advancement of micromachining techniques has led
to the evolution of micro-electro-mechanical systems (MEMS).
These techniques are typically based on semiconductor
manufacturing processes, which offer significant advantages
such as batch manufacturing of end products and monolithic
integration with microelectronics. Surface micromachining
has been used to construct fairly complex microstructures, but
their structural geometries are two-dimensional, which often
limits their mechanical abilities. This constraint has been
addressed by use of bulk micromachining techniques that
involve both etching and deposition processes. For etching,
anisotropic wet etching and deep reactive ion etching have
been widely used to create three-dimensional (3D) geometries
in MEMS. However, these processes are mainly applicable
only to silicon. As for deposition, electroplating has been
used to form 3D metallic microstructures, but practical
materials are also limited to a few metals and their alloys.
Diversifying bulk materials is a key to achieve new
functionalities and higher performance in MEMS, extending
their application fields.
Micro-electro-discharge machining (µEDM) is a bulk
micromachining technique that is applicable to virtually any
electrically conductive material. The technique is capable of
creating complex 3D mirostructures with features as small as 5
µm and aspect ratios up to 20-30. It involves the sequential
discharge of electrical pulses between a microscopic electrode
and the workpiece while both are immersed in dielectric oil
[1]. Although it has been commercially used for applications
such as ink-jet nozzles and magnetic heads for digital VCRs
[2], traditional µEDM is limited in throughput because it is a
serial process based on the use of single machining tools,
which are typically cylindrical tungsten elements with 5-300
µm diameter. This also limits precision because the electrodes
themselves are individually shaped by using a µEDM
technique, wire electro-discharge grinding [3], which may
cause variation in the electrode shape. It was previously
demonstrated by the authors that these constraints could be
Fig. 1:
Batch µEDM
with highaspect-ratio
microelectrode
arrays
fabricated by a
LIGA process.
addressed by using lithographically formed arrays of planar
microelectrodes (Fig. 1) [4]. The parallelism, or, batch mode
process in µEDM offers not only opportunities to manufacture
micro devices and components from a variety of bulk metals
with high throughput and precision but also compatibility with
other planar microfabrication techniques based on lithography
processes, which are the mainstream of MEMS manufacturing.
The effort was extended to address the constraint in
MEMS, i.e., lack of diversity of bulk materials by applying the
µEDM technology. The approach was to use planar metal foil
as starting material for the fabrication, which permits the
benefit of the parallelism to be exploited, offering high
throughput and repeatability. The following section presents
the devices developed through the approach.
II. µEDM MEMS BASED ON BULK METALS
The application of µEDM in the effort involved both
purely mechanical and electromechanical devices. The µEDM
technique can be directly applied to the fabrication of
mechanical devices or components by patterning metallic
materials. The stent described in Section A is included in this
group. For the development of electromechanical devices,
dielectric portions need to be incorporated in µEDM structures
in order to form electrical circuitry in the structures. This
challenge was addressed through the development of the
devices presented in Sections B-E, i.e., antenna stent, micro
Kelvin probe, electromagnetic flow sensor, and capacitive
pressure sensor.
A. MICROMECHANICAL STENT
Stents are mechanical devices that are chronically
implanted into arteries in order to physically expand and
scaffold blood vessels that have been narrowed by plaque
accumulation. The vast majority of stents are made by laser
machining of stainless steel tubes [5], creating mesh-like walls
that allow the tube to be expanded radially upon the inflation
of an angioplasty balloon. The use of µEDM is another option
for cutting metal microstructures. The batch machining
approach mentioned above potentially enables the technology
to be a promising method for the stent manufacturing.
The fabrication approach in this effort was to µEDM 50-
1
0840-7789/07/$25.00 ©2007 IEEE
electrical characteristics of a stent during balloon angioplasty,
allowing the stent to be a helical-shaped antenna (stentenna)
[7]. The planar design of the stent permits the combinational
use of lithography-based micromachining techniques for direct
fabrication of sensors on the stent as well as the integration of
separate micromachined sensors fabricated by the techniques
with the stent. This effort took advantage of the latter benefit.
The planar microstructure for the stentenna had a series
of cross bands that had involute contours, similar to those in
the mechanical stent in Section A, with a bridge to a
longitudinal beam at the center of the device. The involute
bands formed dual inductors, whereas the beam was a
common electrical node. Two micromachined capacitive
pressure sensors [8] were connected across the common line
and the inductors, implementing dual inductor-capacitor (LC)
tank configuration (Fig. 5a). The resonant frequency of the
tank, which depends on local pressure or flow rate, was
wirelessly interrogated through an external antenna that was
magnetically coupled to the stentenna. In contrast to the
mechanical stent, the stentenna needed to play an electrical
role. To warrant this functionality, the fabricated device was
coated with Parylene-CTM for electrical protection while
granting biocompatibility to the device. The stentennas were
then deployed inside a silicone-based mock artery using a
standard angioplasty balloon, resulting in a helical shape with
inductance of approximately 110 nH (Fig. 5b).
µm-thick stainless steel foil into a planar structure that could
be slipped over an angioplasty balloon, plastically deforming
it into a cylinder shape when deployed [6]. The planar pattern
had two longitudinal side-beams, which were connected
transversely by expandable cross bands, each of which
contained identical involute loops (Fig. 2a). In the manner
identical to commercial stents, the deployment of the stent was
emulated by inflating an angioplasty balloon that was threaded
through the planar structure such that the transverse bands
alternated above and below it. Figure 2b shows an SEM
image of the expanded stent with the balloon removed. The
final diameter was 2.65 mm in this case.
Figure 3 shows the measured result of radial stiffness test
for the developed stent with type 304 stainless steel. A
commercial stent with type 316L stainless steel of thickness
varying over 90-130 µm (Guidant Co, IN, USA, Multilink
TetraTM) was also tested for comparison. The test indicated
that the developed stent had almost the same radial strength
even though its walls were only 50-µm thick. (Note that the
mechanical properties of types 304 and 316L of stainless steel
are almost identical.) The radial stiffness was similar when
the loading was applied at two extreme orientations, i.e.,
perpendicular to the original plane of the pre-expansion planar
microstructure and parallel to the plane [6].
Fig. 2: (a: left) A 7-mm-long planar stent sample as cut by µEDM
from stainless steel foil; (b: right) an expanded state of the planar
structure.
Fig. 4: A wireless cardiac monitoring system and its link to a
database permitting physicians to review the sensed parameters.
Fig. 3:
Measurement of
the radial stiffness
of a developed
stent and
comparison to a
commercial stent
with similar
diameter and
twice the
thickness.
Fig. 5: (a: upper) A stentenna integrated with two sensor chips
mounted on a deflated angioplasty balloon; (b: lower) a deployed
device with the balloon removed.
B. ANTENNA STENT WITH MICROSENSORS
After stent implantations, re-narrowing (restenosis) often
occurs. To determine the status, patients are required to take
x-ray angiography periodically, which is an invasive
procedure that inserts a catheter to inject contrast dye and
cannot be taken frequently. The failure is still a concern even
with the recent availability of drug-eluting stents. Wireless
monitoring of cardiac parameters such as blood pressure and
flow can provide advance notices of such failures (Fig. 4).
Toward this end, the planar fabrication approach was utilized
to develop a method that automatically transformed the
Fig. 6:
Measured
amplitude of input
impedance in an
external coil shifted
near 239 MHz due
to flow change.
2
Wireless tests with varying flow rate in a fluidic set-up
demonstrated that an impedance peak was shifted down by
increasing flow rate as shown in Fig. 6. The measured
frequency shift indicated a reduction of 9-31 KHz per mL/min.
increase in the flow range over 370 mL/min. [7]
of MetGlas 2826MB, an alloy of primarily Ni and Fe (Fig. 8a).
First, a traditional step was performed to define the shape of
the epoxy plug in the workpiece (Fig. 8b). A two-part epoxy
was applied onto the machined workpiece to fill the plug and
then cured (Fig. 8c). A lapping step was performed for both
sides of the workpiece to remove the excess cured epoxy (Fig.
8d). The rest of the features were defined by µEDM so that
they were aligned to the epoxy plug, which was released by
cutting along its edges (Fig. 8e). Finally, the finished part was
attached to a glass substrate for testing (Fig. 8f). The
fabricated device shown in Fig. 7b was used for non-contact
sensing of pH of liquid inside microfluidic channels [9].
C. MICRO KELVIN PROBE WITH INTEGRATED ACTUATOR
Kelvin probes are used to measure the contact potential
difference (CPD) between materials, which cannot be
measured directly using a voltmeter. One of the major
applications is the characterization of solid-state devices. A
probe is placed above the surface of a sample in close
proximity, and an AC current is generated by dithering the gap
where a CPD-induced charge is built up. The bias voltage
which nulls the current indirectly determines the CPD.
The micromachined probe developed by µEDM included
an actuator that provided the axial dither motion and a lead
transfer beam for the probe (Fig. 7a) [9]. An electrothermal
bent-beam actuator offered the dither motion with amplitude
in the 10-µm range with drive voltages of a few volts [10].
The low voltage was important to minimize the coupling of
the drive signal to the sense probe, while the large
displacement permitted the dithering frequency and amplitude
to be varied to suit the needs of the measurement.
The device needed an isolation plug that mechanically
coupled the probe to the actuator while electrically and
thermally decoupling them from each other. A large width of
isolation was desired to minimize the capacitive feedthrough
of the drive signal as well as the thermal noise from the
actuator. The incorporation of dielectric components was
achieved by a modified µEDM process shown in Fig. 8. The
starting material was commercially available 30-µm-thick foil
D. INTRALUMINAL CUFF FOR ELECTROMAGNETIC FLOW SENSING
The planar-to-cylindrical reshaping technique used for
the stent fabrication was applied to the development of an
intraluminal cuff for electromagnetic (EM) sensing of flow [7,
11]. The EM detection [12, 13] offers several attractive
features such as direct and linear relationship between the
output and flow, less dependence on cross-sectional flow
profile, and mechanical robustness due to no moving parts
used. EM flow sensors typically have two electrodes located
on inner walls of the fluid channel. In the presence of a
magnetic field, a voltage proportional to the flow velocity is
developed between the electrodes.
The planar design of the ring cuff had a pair of meander
bands comprised of 50 µm-wide beams, electrode plates, and
two dielectric links that mechanically tied the bands but
electrically insulated them from each other (Fig. 9a). This
pattern was µEDMed in 50-µm-thick stainless steel foil, and
then all the surfaces except front-side planes of the electrodes
were coated with an insulating layer. (Without this, spatial
averaging would reduce the voltage.) The dielectric links (of
epoxy in this case) were created by a process similar to that
for forming the isolation plug in the Kelvin-probe device
shown in Fig. 8. The planar structure was mounted on a
deflated balloon of a standard angioplasty catheter so that one
of the bands was located above the balloon whereas the other
band was below it. Figure 9b shows a device that was
expanded inside a silicone tube with 3-mm i.d.. The device in
a wired set-up showed a response that linearly and
symmetrically increased or decreased depending on the
orientation of a magnetic field applied externally (Fig. 10).
The signal reading for this device was also extended to a
wireless implementation using the stentenna [7, 14].
Tip
30 µm
2000 µm
1600 µm
360 µm
Electro-Thermal
Actuator
1215 µm
360 µm
Electrical
Isolation
Signal Lead
30 µm
6.6 ˚
2000 µm
Fig. 7: (a: left) Schematic of the µEDM Kelvin-probe device; (b:
right) a fabricated device bonded to a glass substrate.
A
A
A'
(a)
30 µm Thick A'
MetGlas
Lapping
(d)
EDM Plug Definition
(b)
EDM Microstructure
(e)
Fig. 8:
The
modified
µEDM
fabrication
sequence.
Fig. 9: (a: left) A layout of the
planar cuff; (b: upper) an
expanded cuff inside the tube.
(The lead wires for testing
were removed.)
Epoxy
Epox Fill and Cure
(c)
Finished Microstructure
(f)
3
particular environments for MEMS fabrication.
This
potentially allows us to circumvent constraints and problems
associated with packaging of the devices, broadening
application opportunities for MEMS.
III. CONCLUSION
This paper presented recent study on the use of bulk
metals and the µEDM technology for the development of
MEMS devices. The effort demonstrated that the use of µEDM
was effective to fabricate micromachined devices with both
mechanical and electrical functionalities. The large material
base of the technology enabled us to select appropriate
engineering materials with particular characteristics such as
plasticity, robustness, chemical inertness, and biocompatibility
as well as cost-effectiveness with the use of commonly
available stock metal foil. The result suggests that with the
availability of the batch mode technique, µEDM can be a
unique tool for the development and manufacturing of new
types of MEMS that cannot be achievable with conventional
MEMS fabrication methods.
Fig. 10: (a: left) The fluidic measurement set-up with magnetic field
of ~0.25 T applied at different orientations; (b: right) the measured
results.
E. CAVITY/DIAPHRAGM-LESS CAPACITIVE PRESSURE SENSOR
Micromachined capacitive pressure sensors typically use
an elastic diaphragm with fixed edges and a sealed cavity in
between the diaphragm and the substrate below. Since this
configuration relies on the deflection of a relatively thin
diaphragm against a sealed cavity, in some applications there
is a concern for the robustness of the diaphragm and leaks in
the cavity seal. To achieve mechanical robustness and
simplify the structural configuration, it was aimed to eliminate
the need for diaphragms and cavities from the sensor structure.
This was approached by the configuration that consisted of
two micromachined metal plates with an intermediate polymer
layer [15]. Use of polymeric material soft enough to deform
in a target pressure range allowed the thickness of the polymer,
or capacitance of the parallel plate capacitor, to be dependent
on hydraulic pressure that surrounded the device. The
operation was demonstrated by the device with
micromachined stainless-steel electrodes defined by µEDM
and a liquid-phase polyurethane that was applied and
solidified between the electrodes (Fig. 11a). The pressure
monitoring was demonstrated by measuring frequency shifts
in the LC tank that was fabricated by winding a copper coil on
the sensor and bonding the terminals to the electrodes (Figs.
11b and 12). The combination of the selected materials
potentially permits the direct use in corrosive or biological
environment. As demonstrated in this effort, the use of µEDM
promotes proper choice of materials that are compatible with
REFERENCES
[1] T. Masaki, K. Kawata, T. Masuzawa, “Micro Electro-Discharge
Machining and its Applications,” Proc. IEEE MEMS, pp. 21-26, 1990
[2] Y. Honma, K. Takahashi, M. Muro, “Micro-Machining of Magnetic
Metal Film Using Electro-Discharge Technique,” Advan. Inform. Stora.
Sys., 10, pp. 383-399, 1999
[3] T. Masuzawa, M. Fujino, K. Kobayashi, T. Suzuki, N. Kinoshita, “Wire
Electro-Discharge Grinding for Micro-Machining,” Ann. CIRP, 34, pp.
431-434, 1985
[4] K. Takahata, Y.B. Gianchandani, “Batch Mode Micro-Electro-Discharge
Machining,” IEEE/ASME J. Microelectromech. Sys., 11(2), pp. 102-110,
2002
[5] Y.P. Kathuria, “Laser Microprocessing of Stent for Medical Therapy”,
Proc. IEEE Micromech. Human Sci., pp. 111-114, 1998
[6] K. Takahata, Y.B. Gianchandani, “A Planar Approach for Manufacturing
Cardiac Stents: Design, Fabrication, and Mechanical Evaluation,”
IEEE/ASME J. Microelectromech. Sys., 13(6), pp. 933-939, 2004
[7] K. Takahata, Y.B. Gianchandani, K.D. Wise, “Micromachined Antenna
Stents and Cuffs for Monitoring Intraluminal Pressure and Flow,”
IEEE/ASME J. Microelectromech. Sys., 15(5), pp. 1289-1298, 2006
[8] A. DeHennis, K.D. Wise, “A Double-Sided Single-Chip Wireless Pressure
Sensor,” Proc. IEEE MEMS, pp. 252-255, 2002
[9] L. L Chu, K. Takahata, P. Selvaganapathy, Y.B. Gianchandani, J.L.
Shohet, “A Micromachined Kelvin Probe with Integrated Actuator for
Microfluidic and Solid-State Applications,” IEEE/ASME J.
Microelectromech. Sys., 14(4), pp. 691-698, 2005
[10] L. Que, J. S. Park, and Y. B. Gianchandani, “Bent-beam electro-thermal
actuators-I: Single beam and cascaded devices,” IEEE/ASME J.
Microelectromech. Sys., 10(2), pp. 247–254, 2001
[11] K. Takahata, Y.B. Gianchandani, “A Micromachined Stainless Steel Cuff
for Electromagnetic Measurement of Flow in Blood Vessels,” Proc.
Solid-State Sensor, Actuator and Microsystems Workshop (Hilton Head),
pp. 290-293, 2004
[12] J.G. Webster, The Measurement, Instrumentation, and Sensors Handbook,
28.6, CRC Press, 1999
[13] H.J. Yoon, S.Y. Kim, S.W. Lee, S.S. Yang, “Fabrication of a Micro
Electromagnetic Flow Sensor for Micro Flow Rate Measurement,” Proc.
SPIE Symp. Smart Struc. Mat., pp. 264-271, 2000
[14] K. Takahata, Y.B. Gianchandani, “Micromachined Intraluminal Devices
for Active and Passive Electromagnetic Measurements of Flow,” Proc.
IEEE MEMS, pp. 863-866, 2005
[15]
K.
Takahata,
Y.B.
Gianchandani,
“A
Micromachined
Polyurethane/Stainless-Steel Capacitive Pressure Sensor Without Cavity
and Diaphragm,” Proc. IEEE Transducers, pp. 483-486, 2005
Fig. 11: (a: left) A polyurethane/stainless-steel capacitive pressure
sensor fabricated by µEDM; (b: right) an LC tank form of the device.
Fig. 12:
Pressure vs. frequency
measured with the LC
tank device.
4
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement