Ferroelectrics and piezoelectrics for MEMS

Luca Lutterotti: Luca.Lutterotti@ing.unitn.it

2006 Meeting Trondheim January 5
MEMS materials tree
Piezoelectric
Pyroelectric
Electrooptic
SAW/BAW
(NORSPACE)
RF-MEMS Ferroelectric
memories
Magnetostrictive
Magnetic
Academic
Soon commercial
Commercial
M
a
te
ria
ls
s
c
ie
n
c
e
M
ic
ro
ele
ctro
n
ic
s
Chemistry
Physics
Ultrasonic imaging
Positioning systems
Flow sensors
Accelerometers
Microactuators
Gas sensors
Energy harvesting
IR devices/imaging
Optical switches
Modulators
Microstructuring Life science
Gyros/accelerometers
Capacitive actuation/sensing
Piezoresistive sensing
Thermal devices
Pressure sensors (PRESENCE)
Airbag sensors (SENSONOR)
Optical emission/detection/reflection
(SIMRAD OPTRONICS)
Microfluidics
Bio-MEMS
Microreactors
Optical MEMS
CD-technology
Lab-on-a-chip (NORCHIP)
Spectrophotometer (TOMRA)
Mechanical actuation and sensing in MEMS
2006 Meeting Trondheim January 5
Mechanical actuation and sensing in MEMS
Piezo-
electric
Electro-
magnetic
Capaci-
tive
Piezo-
resistive
Thermal Mechan-
ism
MEMS with functional materials Si-only MEMS
+
+
Difficult to scale
down/make
inductor
Few materials
Piezoelectricity
!! " !E
Large amplitude
Linear
High energy
conversion
Capacitive
force on
charged
membrane
F " 1/d
Non-linear
Piezoresistivity
!R " !!
Sensing only
Small amplitude
Thermal
expansion
!! " !T
Slow, small
amplitude
Bad energy
conversion
Comment
+++ + - +
Actuation
+++ +++ ++ -
Sensing
Comparison different material classes
Applicazioni come attuatori (PZT)
TABLE II Comparison of characteristics of shape-memory alloys, piezoelectric ceramics and magnetostrictive materials as actuation
materials
Properties Shape-memory alloy Piezoelectric Magnetostrictive
(TiNi) (PZT) (Terfenol-D)
Compressive stress (MPa) :800 60 700
Tensile strength (MPa) 8001000 3055 2835
Youngs modulus (GPa) 5090 (P) 6090 (Y#) 2535 (Y&)
1035 (M) :110 (Y") 5055 (Y)
Maximum strain :0.1 :0.001 :0.01
Frequency (Hz) 0100 120 000 110 000
Coupling coecient : 0.75 0.75
Eciency (%) 35 50 80
Energy density (kJ m\) 300600 :1.0 1425
modulus for constant electric eld
modulus for constant electric displacement
modulus for constant magnetizing eld
modulus for constant induction eld
conventional piezoelectric or electrostrictive ceramics
have a superior dynamic response but their displace-
ments are quite small and most of them are very
brittle. Combining SMAs with piezoelectric or mag-
netostrictive materials, eld-activated smart com-
posites can be designed, which may generate a larger
displacement than conventional piezoelectric ceramics
or magnetostrictive materials and have an improved
dynamic response as compared to monolithic SMAs.
More recently, some pioneers have explored the tech-
nical feasibility of smart thin-lm heterostructures by
depositing the SMA thin lms on piezoelectric or
magnetostrictive substrates. However, the complexity
of the fabrication processing and the interface bonding
and dynamic coupling of dissimilar components re-
main tough issues for the composites [213].
The alternative is to improve the monolithic shape-
memory materials by employing new processing tech-
niques or to design a new generation of shape-memory
materials. The development of deposited thin-lm
shape-memory alloys, as we described above, is one of
the eorts directed to this objective. Also worth men-
tion are the recently developed porous shape-memory
alloys [214, 215]. Bulk TiNi alloys with dierent por-
osity, exhibiting superelasticity and shape-memory
eect, have been successfully manufactured via the
powder metallurgical route. The porous SMAs are
very desirable for some biomedical applications be-
cause the alloys have good biocompatibility and their
porous structure favours in-growth of living tissues
and rm xation. Naturally, it reminds us of bone
a typical biomimetic model. Bone is also porous;
moreover, it exhibits pyroelectricity and piezoelectric-
ity, and maintains the skeletal homeostasis and min-
eral homeostasis for the body [1]. After the model,
biomimetic articial bone materials based on the
porous SMAs and other advanced materials may be
developed. For instance, microballoons or microtubes
coated by some functional material layers can be
constructed in the porous SMAs which may provide
a suitable substrate or skeleton to grow heterostruc-
tures with certain intelligence.
In principle, the deformation of the polydomains in
the ferromagnetic and ferroelectric materials by ap-
plying external elds can be controlled just the same
way as the stress-induced deformation of the marten-
sites in ferroelastic SMAs. The next challenging
objective, therefore, is to explore new potentially com-
mercial materials wherein the martensitic-like trans-
formations and the reorientation of the domains can
be induced by magnetic elds or electric elds at
ambient temperatures. The design concepts and strat-
egies for nding new ferromagnetic and ferroelectric
shape-memory materials have been proposed [142,
205211]. In this aspect, the remarked common fea-
tures shared by several smart material systems, and
the successful development story of the giant mag-
netostrictive materials Terfenol-D [176, 216218]
may oer some clues or inspirations.
Acknowledgements
The authors thank Prof. J. Van Humbeeck of K. U.
Leuven and Mr J. Cederstrom, Scandinavian Memory
Metals, AB, for critical reading of the manuscript,
Professor R. D. James, University of Minnesota, and
Professor D. S. Grummon, Michigan State University,
for providing their papers prior to publication. One of
the authors (Z G W) wishes to acknowledge the Royal
Institute of Technology (KTH) for oer of a fellowship
to work at KTH.
References
1. M. V. GANDHI and B. S. THOMPSON, Smart Materials
and Structures (Chapman & Hall, New York, 1992).
2. T. TAKAGIL, J. Intelligent Mater. System Struct. 1 (1990)
149.
3. Idem, in Proceedings of the 3rd International Conference on
Intelligent Materials, edited by P. F. Gobin and J. Tatibouet
(Technomic, Lancaster, 1996) p. 2.
4. C. A. JAEGER and C. A. ROGERS, in Proceedings of the
ARO Workshop on Smart Materials, Structures and Math-
ematical Issues, edited by C. A. Rogers (ARO, Washington,
1988) p. 14.
5. S. NOZAKI and K. TAKAHASHI , in Proceedings of the
2nd International Conference on Intelligent Materials,
edited by C. A. Rogers and G. G. Wallace (Technomic,
Lancaster, 1994) p. 1230.
6. G. BECK and P. F. GOBIN, in Proceedings of the 1st
International Conference on Intelligent Materials, edited by
T. Takagi, K. Takahashi, M. Aizawa and S. Miyata (Tech-
nomic, Lancaster, 1992) p. 9.
3758
Different actuator types
Confronto tipi di attuatori
IBM SYSTEMS JOURNAL, VOL 35, NOS 3&4, 1996 FLETCHER 633
been discovered but are still in the developmental
stage.
18
Piezoresistive polymers and polymer composites.
Another promising class of materials for force trans-
duction consists of electronic polymer materials. The
eld of polymer science continues to advance tremen-
dously, and there now exist various types of polymers
that behave to some degree like the more common
metal- or ceramic-based smart materials.
19
However,
since most smart polymers are still in early develop-
ment, an alternate approach has been to use polymers
for sensing or measuring force indirectly. These mate-
rials are grouped separately from smart materials
because the mechanism employed is not reciprocal
(e.g., applying an electrical stimulus will not produce
a force output). Rather than performing force trans-
duction through direct conversion of mechanical
energy to electrical energy, as in PVDF, certain types of
polymers can be used to modulate an electric current
via a force-dependent electrical resistance. Polymer
sensors can be used in compression as well as exten-
sion and can function over a greater range of strains
than conventional metal strain gauges.
20
The two most common physical mechanisms em-
ployed in resistive polymer force sensors are piezo-
resistance and geometric deformation. The deforma-
tion type of force sensor is comprised of a conducting
polymer strip or foam (such as polypyrole) and relies
simply on the deformation of the polymer to vary the
conducting cross section, thus changing the resis-
tance. Piezoresistive sensors, in contrast, generally
consist of a semi-insulating polymer matrix contain-
ing some type of conducting particulates, such as
graphite. Such compounds are macroscopically
piezoresistive but also exhibit noise and hysteresis
due to the percolative nature of the conduction mecha-
nism. Although some mildly piezoresistive polymers
have been synthesized, practical force sensors com-
prised of a homogeneous piezoresistive polymer are
not yet commercially available.
Rare-earth permanent magnets. The discovery of
new rare-earth magnetic materials, such as NdFeB,
has led to the development of compact electromag-
netic actuators, which represent perhaps the most
attractive technology for force actuators requiring
large displacements, as a possible substitute for
hydraulic actuators.
21
Using a geometry similar to that
of solenoid actuators, linear electromagnetic actuators
made from permanent magnets can produce forces as
high as 2000 foot-pounds over a 24-inch stroke.
Cleaner and less noisy than hydraulic actuators, com-
pact powerful permanent magnets driven by magnetic
elds may also be employed as large-displacement
dynamic elements to transmit sound or vibration.
Other materials. Some of the materials mentioned in
this paper will most likely be superseded by new
classes of force transduction materials that are cur-
rently being synthesized or are yet to be discovered.
In some cases, new process technologies will enable
new materials to be processed, such as the powder
metallurgy and rapid solidication process that
enabled the fabrication of amorphous metals. In other
cases, the advent of powerful computers and the
increasing ability to engineer materials on the molec-
Table 1 Comparison of active materials
PZT5H
Max strain (! 10
6
)
Max stress (MPa)
Density (kg/m
3
)
Modulus (GPa)
Efciency (%)
Bandwidth (Hz)
300
19
7500
62
56
1000
Energy density (KJ/m
3
) 2.9
600
72
7500
120
75
1000
22
3000
180
~7500
~60
?
?
?
200
~1
1780
3
2
1000
~1
1800
90
9250
40
40
100
19
70000
190
6450
78
>3
3
>10
20000
180
~1500
5
>30
>1
>1
400000
0.3
~1300
<0.1
30
.1
0.4
400000
0.3
1037
.06
>35
4
0.8
PMN PLZT PVDF Ternol-D SMA
(NiTi)
Conducting
Polymer
Polymer
Hydro-gel
Human
Muscle
PZT-PLZT phase diagram
2006 Meeting Trondheim January 5
Phase diagram of the PZT and PLZT solid-solution systems
Haertling(1999)
Strong
! piezoelectric
! pyroelectric
! electrooptic
! relaxor
effects are
observed in this
system
FE = ferroelectric
AFE = antiferroelectric
PE = paraelectric
Pyroelectric
application
area
Piezoelectric
application
area
Relaxor
phases
Morphotropic
phase boundary
area
Production PZT-MEMS
Production PZT-MEMS (cantilever-membrane)

Chemical solution deposition (SINTEF)

RF-sputtering (EPFL)

Pulsed laser deposition

2006 Meeting Trondheim January 5
Processing steps MEMS-pie
TiO2 barrier/adhesion
layer (20 nm)
Pt electrode and seed layer
(100 nm)
Piezoelectric thin film layer (2 m)
Top electrode deposition
and pattering (lift off patterning)
PZT patterning
Back side etch (wet etch
with protected front side)
Patterning and release
etch of cantilevers
Starting point: SOI wafer (buried oxide layer)
Si
SiO2
Si
Types of piezoelectric MEMS
Type of MEMS piezoelectric
Recent studies have shown that for PZT thin films, d
33
measured using interferometric
techniques is roughly independent of sample thickness [7]. However, on application
of switching voltages a linear behaviour between the material properties in the
fatigued state and the sample thickness was found.
3 Choice of Materials
At present, there are basically three main types of structures that have been
considered. These are discussed in this section, with uses for these basic structures
elaborated on further in a later section. In most of the structures examined so far,
silicon has been used as a substrate, onto which the active part of the device is
deposited.
3.1 Whole Wafer
In this structure, shown schematically in figure 3.1, the piezoelectric region is
deposited across the whole of the wafer. The top electrode is most likely to consist of
an array of dots, rather than a continuous electrode. This allows regions to be
activated with different voltages, and can help avoid any problems with pin-hopes or
conducting pathways for very thin piezoelectric films.

Figure 3.1: Schematic of the MEMS structure based on a whole
wafer.
The silicon substrates are approximately 500 m thick, standard for 100 mm diameter
wafers, but could be around half this value for special wafers. Larger diameter wafers
are found to be thicker. The substrates are generally single sided (polished on one side
only), although for special cases, double-sided wafers can be found.
The support layer is usually silicon dioxide, SiO
2
, which can be thermally grown with
ease to produce a thin stoichiometric layer. Silicon nitride or a combination of this
and SiO
2
are both possible. In each case, the thickness of this layer is in the region of
1 200 nm.
The bottom electrode is found to be Pt/Ti or Pt/TiO
2
, both being used as standard,
with TiO
2
showing better adhesion properties [8]. The Pt/Ti electrode structure has a
thickness of approximately 100 nm / 5 nm. As well as Pt/Ti, oxide electrodes can be
considered. Oxide electrodes can offer the advantage of improving the fatigue
characteristics of the piezoelectric, but are usually thicker, being up to about 200 nm.
For the upper electrode, Cr/Au is normally used, but the same material as the lower
electrode may be used. This electrode, being exposed to the atmosphere, is usually
Top electrode
Piezoelectric
Bottom electrode
Support layer
Substrate
Whole wafer
structure
MEMS
membrane
thicker and can be patterned to different sizes during testing. The electrodes can
typically be 1 m or so in thickness.
The piezoelectric layer generally consists of PZT. High strain relaxor materials will
be of interest but little work has been performed in this area thus far. The PZT film
thickness is usually in the range 0.5 5 m.
3.2 Membrane
In considering membrane structures, such as that shown in figure 3.2, a similar
structure is used as for patterning a whole wafer. To date, membranes are either
square or circular, but other shapes could be fabricated. The dimension, d, is typically
in the range 0.5 5 mm. As part of the silicon structure has been removed in this
configuration, the thickness of the support layer, along with its stiffness, become more
important, and may well be thicker than encountered previously. A variation of the
membrane structure has been used in biomedical applications for micropumps.

Figure 3.2: Schematic of the MEMS membrane structure
3.3 Cantilever
A schematic of the cantilever MEMS structure is shown in figure 3.3, and can be
found in micro-relays for example. As for the membrane, due to part of the substrate
being removed, the thickness and strength of the support layer will be determined by
the applications of the device as well as the size and shape of the cantilever. As shown
in figure 3.3, if the cantilever is to act as a switch, additional contacts will need to be
included in the package. The beam length is typically around 0.1 1 mm, with the
width approximately 20 200 m.

d
Top electrode
Piezoelectric
Bottom electrode
Support layer
Substrate
Figure 3.3: Schematic of the MEMS cantilever structure,
showing typical dimensions of the beam.
4 Literary Review of Piezoelectric MEMS
Many different types of structures have been fabricated by using piezoelectric
materials in MEMS. 3-D structures such as nozzles, relays, filters, inductors and
power dividers have been developed. As well as traditional piezoelectric materials
such as PZT and BaTiO
3
, work has been performed using polymeric piezoelectrics.
This section details some of the published work to date. Examples of some of the
devices produced in different industrial sectors are described along with some of the
aspects of modelling of these devices.
4.1 PZT incorporated into MEMS Devices
The majority of work into piezoelectric MEMS devices has focussed on PZT, to
provide small-scale actuation or sensing. PZT has already been extensively used in
areas such as ultrasound and in nanoscience (for example actuators for atomic force
microscopy). An overview of the issues of integrating PZT into MEMS devices has
been reported, which gives figures of merit for various applications [9].
Several MEMS devices have been developed for a wide range of applications. MEMS
accelerometers based on cantilever beams and PZT have been demonstrated [10].
There has also been some discussion of the use of ferroelectric thin films as part of
MEMS sensors in infrared detectors and MEMS actuators for use in optical and
acoustic devices [11].
Piezoelectric MEMS actuators may provide high force output at low driving voltages,
but applications may be limited due to small strains involved. This has led to the idea
of a strain amplifier using large-scale micro-fabrication techniques [12]. Using PZT as
a compliant mechanism gives the required amplification at the expense of some
piezoelectric force, but which is still desirable.
Another area of interest in actuation is the production of nanometre steps over a range
of hundreds of microns, which may be used for precision positioning of optical
components or microsurgical instruments [13].
length
width
Actuation
electrodes
Switching
contacts
MEMS
cantilever
MEMS fabrication (texture and stresses)

PZT on Pt/Ti/SiO
2
/Si or Pt/Ti/TiO
2
/SiO
2

Texture and residual stresses are important

Texture determines the properties along different

directions

Residual stresses affect also the properties and life of the

components.

Residual stresses are due to different thermal expansion

properties and crystallographic matching

Applying a stress opposed to the observed one can

improve properties.

Domain wall motion is little observed in thin lms

Applications: MEMS
Applications: MEMS
CIRP Seminar on Micro and Nano Technology
Copenhagen, Denmark, November 13-14, 2003

photo steps. The cantilever structure is released by an
isotropic XeF
2
vapor etch step, which avoids stiction.
2. DESIGN AND FABRICATION
2.1 Design
Figure 1 shows the device schematic of the d
33
mode
piezoelectric device. The basic design of the multilayer
structure is as follows: Layer 1: Membrane layer (SiO
2
) for
controlling stress and bow of the cantilever structure;
Layer 2: Diffusion barrier/buffer layer (ZrO
2
) for preventing
electrical charge diffusion from the piezoelectric layer
above it; Layer 3: PZT piezoelectric layer; Layer 4: Top
interdigitated electrode (Pt/Ti). Layer 5: Optional SU-8
Proof Mass.
The mechanical resonance frequency of the cantilever
can be tuned by altering the cantilever dimensions (i.e.
layer thicknesses, length of the beam and the inclusion of
a proof mass, if necessary). This affects the k stiffness of
the composite beam. The resonance frequency can be
approximated by calculating the composite beam stiffnes,
K, the value of the lumped mass, M (A proof mass was
included in our experimental device.) and using the
standard formula for the resonance frequency, !
n
:
M
K
n
! "
(3)
The vibrational excitation of the PMPG micro cantilever
can be modeled as a seismically excited, spring-mass
damper:
i i o o o
x B Kx Kx x B x M ! ! ! ! # ! # #
(4)
where x
i
is the input, seismic base displacement, x
o
is the
output, cantilever tip displacement, M is the lumped
mass, B is the damping coefficient and K is the stiffness
of the composite cantilever beam.
The ratio of the cantilever tip displacement to base
displacement at resonance is a well known result as
shown in equation (5):
Q
w s w s
s w w
s x
s x
n n
jw s
n n
n n
j s
i
o
!
# $ #
$ #
!
! !
2
2
2
2
2
) (
) (
%
%
"

B
M
Q
Q
n
"
% ! ! and
2
1
where
(5)
The Q of the system can be found by taking the ratio of
the tip displacement at resonance over the tip
displacement far outside the resonance. For the
actuation mode the Q turns out to be approximately 47. t
is expected to be higher for the sensor mode.
The rectification bridge circuit consists of four
STMicroelectronics& 1N5711 small signal Schottky
diodes and a 10 nF mylar storage capacitor. These
diodes were chosen specifically because, compared to
most discreet components, they have the smallest
forward voltage drop (approximately 0.2V). This allows
for the largest possible DC voltage to develop across the
cap/load. The 10 nF mylar cap was chosen because it
does not leak currently easily. The resistors are also
discreet components.
2.2 Fabrication
We fabricated a d
33
mode, thin film PZT cantilever device
with DT electrode using just three photo masks. First, the
membrane layer (PECVD oxide) is deposited on the Si
wafer. Each PECVD oxide layer is annealed at 750C for
30 minutes. The 50 nm thick ZrO
2
layer is deposited via a
sol-gel spin-on process. t acts as a buffer layer on top of
the various membrane layers and is dried at 350C for 1
minute, then annealed at 700C for 15 minutes. Our
device uses 0.05'm ZrO
2
/0.4'm PECVD oxide.
The PZT solution is spun on the substrate at 500 rpm for
3 seconds and 1500 rpm for 30 seconds. The precursor
gel film is pyrolyzed at 350C for 5 minutes on a hot plate
in several repeated cycles to create a PZT layer with 0.48
um thickness. The PZT film is annealed at 700C for 15
minutes in a box furnace. The PZT layer and membrane
layers are patterned via RE with the first mask. The
interdigitated top electrode requires the second mask and
is deposited using an e-beam evaporation and lift-off
procedure with 20nm Ti and 200nm Pt. The SU-8 is spin
coated on top of the existing layers and patterned with
the third mask to create the proof mass. The cantilever
membrane is then released with the XeF
2
vapor etcher.
No mask is required because XeF
2
has a high etch
selectivity between Si and the other layers. An SEM
image of the fabricated PMPG device is shown in Figure
3.

Figure 1: Device schematic of d
33
mode
piezoelectric device.

Figure 2: Higher d
33
Open Circuit Voltage vs. d
31


Figure 3. SEM image of the fabricated device
2.3 Packaging, Poling and Assembly
The MEMS chip with multiple PMPG devices was super
glued to a multiple pin, ceramic package. Each MEMS
device has two bond pads: One for the positive potential
and one for the negative. The chip was packaged using
(a) (b)
+ - + - + - +
PZT
ZrO
2
membrane
Inter-digitated
Electrode
d
33
mode d
31
mode
V
31
= (
xx
) t
pzt
) g
31
V
33
= (
xx
) L

) g
33
V
33
= 2L / t
pzt
) V
31
* 20 ) V
31
* 3 V
L
( if g
33
> 2g
31
, L * 10 t
PZT
)
(
xx
E
t
PZT
(
xx
(
xx
(
xx
t
PZT
E
2006 Meeting Trondheim January 5
Integration Integration of Functional of Functional
Ceramics in MEMS Ceramics in MEMS
Henrik Rder
SINTEF Materials and Chemistry
Functional Ceramics Group
Thanks to colleagues at SINTEF M&C:
Frode Tyholdt, Nicolas Lecerf, Hallgeir Klette, Camilla Haavik, Arne Ryset, Steinar Neegaard, Ragnar
Fagerberg, John Walmsley, Martin Fleissner and Rune Bredesen
and SINTEF ICT:
Wilfred Booij, Niels Peter stb, Erik Poppe, Dag Wang, Andreas Werner, Ralph Bernstein and Anders
Hanneborg
and UiO and NTNU
2006 Meeting Trondheim January 5
Different materials and deposition methods
! Common flexible process
route for both membranes
and cantilever structures
! PZT
! Chemical solution
deposition (SINTEF)
! RF-sputtering (EPFL)
! Pulsed laser deposition
(University of Twente,
oxide electrodes)
! AlN
! RF sputtering (EPFL)
! AlN technology to be
established at SINTEF in
2006
4 wafer design
Processed wafer
!"#$%&"'()*+&$,+$)-&.&(&/, !"#$%&"'()*+&$,+$)-&.&(&/,
Ferroelectric/piezoelectric-diamond hybrid heterostructures
for high performance MEMS/NEMS devices
S. Srinivasan, O. Auciello, G. Bai,
J. Hiller, and J.A. Carlisle
Schematic diagram of a molecular
recognition biosensor based on
PZT-UNCD heterostructured cantilever
Figure 1. Cross section SEM picture of PZT thin
film grown by MOCVD on a Pt (100 nm) /TiAl
(10 nm) /UNCD (1 !m) layered heterostructure
Figure 2. Characterization of first
Pt/PZT/Pt/TiAl/UNCD capacitor showing
excellent polarization hysteresis
Figure 3. SEM image
of UNCD cantilevers
1 !m
1 !m
Figure 4. SEM image of UNCD
paddle resonator structures
Future directions:
! Optimization of processing for integration of functional PZT
thin films on UNCD cantilevers for MEMS/NEMS applications
! Characterization of PZT/UNCD
cantilevers performance
! Extend the integration to
other functional materials
(e.g. multiferroics) on UNCD
Piezoelectric microcantilever working principle
2006 Meeting Trondheim January 5
Piezoelectric microcantilever working principle:
Sensor
(Grabiec, 2005)
Actuator and
sound emitter
Energy harvester
Mass
Battery
!
V
E
T
PZT
Si
Transverse piezoelectric effect is utilised, d
31
+ e
31
Piezoelectric microsensor
Piezoelectric microsensor
!"#$!$%$&'()'*+$&,!+-#
,&#&+,./'+"0'&"1!"&&,!"1
!"#$%"&'()'(*$+,$#-#./&'.$.+01+(#(/%$2(3$.'&.4'/%$4%'(*
/+3256%$%#0'.+(34./+&$'(/#*&2/#3$.'&.4'/$789:$/#."(+-+*5$"2%
%'*(','.2(/-5$."2(*#3$/"#$;+&-3<$='('2/4&'%2/'+(>$;"'."$2-%+
&#34.#%$.+%/>$"2%$?##($&#.+*('%#3$2%$+(#$+,$/"#$02'(
/#."('.2-$/&#(3%$?5$?+/"$2.23#0'.%$2(3$'(34%/&5<$@,,+&/%$'(
0'('2/4&'%'(*$#-#./&+0#."2('.2-$3#A'.#%$2&#$-#23'(*$/+$2
&21'3$#B12(%'+($+,$&#%#2&."$2(3$3#A#-+10#(/$'(
0'.&+#-#./&+0#."2('.2-$%5%/#0%$7=@=C:<
D/$8=E@>$;#$"2A#$3#A#-+1#3$2$(40?#&$+,$23A2(.#3
1'#F+#-#./&'.$02/#&'2-%$;'/"$1+/#(/'2-$,+&$=@=C$211-'.2/'+(%>
'(.-43'(*$"'*"$1#&,+&02(.#$1'#F+#-#./&'.$/"'($,'-0%>$/"'.)
,'-0%>$2(3$04-/'-25#&$.#&20'.%<$G#$"2A#$3#0+(%/&2/#3
%#A#&2-$/'(5$1'#F+#-#./&'.$2./42/+&%$2(3$4-/&2%+('.$0+/+&%>
2(3$"2A#$023#$%4?%/2(/'2-$1&+*&#%%$'($,4&/"#&$%.2-'(*$3+;(
+4&$1'#F+#-#./&'.$3#A'.#%<$8($+4&$.+--2?+&2/'+($;'/"$C+(5
@-#./&+('.%$7C'(*21+&#:>$;#$"2A#$.+0?'(#3$+4&$1'#F+#-#./&'.
&#%#2&."$;'/"$%'-'.+($0'.&+H,2?&'.2/'+($1&+.#%%#%$/+$1&+34.#
2$0'.&+H02."'(#3$1'#F+#-#./&'.$2./42/+&$1&+/+/51#<$!"#
3#A#-+10#(/$;+&)$(+/$+(-5$'(A+-A#%$2./42/+&$%/&4./4&#
3#%'*($2(3$04-/'1-#$02/#&'2-%$'(/#*&2/'+(>$?4/$2-%+$04-/'1-#H
!"#"$%&'"()*+ ,")-.)/)0%'"0+ 1)2"0)(
!"#$%#&'(#)*+
Toon nonbos. Yoo Ko| (|-,oo_|no.o-s|o.ooo.sq;, Soo||oooo Shooo|qoh| (soo||-s_|no.o-s|o.ooo.sq;, Tooq \|oosooq (xs-|ooq_|no.o-s|o.ooo.sq;,
|o \o|ooq (x-ho_|no.o-s|o.ooo.sq;, Coo 8oo Kooo (b|-qoo_|no.o-s|o.ooo.sq; ooo Choo /o|no (nn-choo_|no.o-s|o.ooo.sq;
Co||oboo|o. S|oqooo osooch Loboo|o, - Soo, E|oc|oo|cs (S|oqooo; f|o L|o
,+-*%.#/012#34.#567#84(9(#(:#'#8+.;(.<.=9%+=#>+=%(0'=9*'9(%#?'>8<.#@.A.<(8.@
:%(>#(*%#B'9=4#>+=%(0:'B%+='9+(C#8%(=.??2#34+?#'=9*'9(%#=(>8%+?.?#/#8+.;(.<.=9%+=
94+C#:+<>#.<.>.C9?D#E+94#9E(#.<.>.C9?#(C#'#B.'>2
,+-*%.#/0F2#G#>+=%(0>'=4+C.@#8+.;(.<.=9%+=#>+=%(0'=9*'9(%#?'>8<.#B.+C-#9.?9.@2
%/#1$0'.&+H12//#&('(*$2(3$/"&##H3'0#(%'+(2-$0'.&+H
,2?&'.2/'+(<$8((+A2/'A#$02/#&'2-%$'(/#*&2/'+($2(3$0'.&+H
,2?&'.2/'+($0#/"+3%$"2A#$?##($4%#3$/+$%+-A#$02/#&'2-%$2(3
1&+.#%%'(*$.+012/'?'-'/5$'%%4#%<$G2,#&H?2%#3$?2/."
1&+.#%%'(*$/"2/$2--+;%$-2&*#H%.2-#$1&+34./'+($+,$0'.&+H
2./42/+&%$"2%$2-%+$?##($%4..#%%,4--5$'(A#(/#3<$!"'%
%'*(','.2(/-5$&#34.#%$1&+34./'+($.+%/%<
I'*4&#$JHK$%"+;%$/"#$C@=$1"+/+$+,$2$1'#F+#-#./&'.
0'.&+H2./42/+&$%201-#$3#A#-+1#3$4%'(*$+4&$?2/."$0'.&+H
,2?&'.2/'+($1&+.#%%<$!"#$2./42/+&$.+01&'%#%$J$1'#F+#-#./&'.
/"'($,'-0$0'.&+#-#0#(/%>$;'/"$/;+$#-#0#(/%$+($2$?#20<
I'*4&#$JHL$%"+;%$2$%201-#$?#'(*$/#%/#3$;'/"$+4&$-2%#&
,+-*%.#/0H2#34.#A+B%'9+(C#A.<(=+9"#(:#'#>+=%(0'=9*'9(%#?'>8<.#*C@.%#.<.=9%+='<
@%+A+C-#+?#9.?9.@#E+94#'#<'?.%#A+B%(>.9.%2
!"#$!$%$&'()'*+$&,!+-#
,&#&+,./'+"0'&"1!"&&,!"1
!"#$%"&'()'(*$+,$#-#./&'.$.+01+(#(/%$2(3$.'&.4'/%$4%'(*
/+3256%$%#0'.+(34./+&$'(/#*&2/#3$.'&.4'/$789:$/#."(+-+*5$"2%
%'*(','.2(/-5$."2(*#3$/"#$;+&-3<$='('2/4&'%2/'+(>$;"'."$2-%+
&#34.#%$.+%/>$"2%$?##($&#.+*('%#3$2%$+(#$+,$/"#$02'(
/#."('.2-$/&#(3%$?5$?+/"$2.23#0'.%$2(3$'(34%/&5<$@,,+&/%$'(
0'('2/4&'%'(*$#-#./&+0#."2('.2-$3#A'.#%$2&#$-#23'(*$/+$2
&21'3$#B12(%'+($+,$&#%#2&."$2(3$3#A#-+10#(/$'(
0'.&+#-#./&+0#."2('.2-$%5%/#0%$7=@=C:<
D/$8=E@>$;#$"2A#$3#A#-+1#3$2$(40?#&$+,$23A2(.#3
1'#F+#-#./&'.$02/#&'2-%$;'/"$1+/#(/'2-$,+&$=@=C$211-'.2/'+(%>
'(.-43'(*$"'*"$1#&,+&02(.#$1'#F+#-#./&'.$/"'($,'-0%>$/"'.)
,'-0%>$2(3$04-/'-25#&$.#&20'.%<$G#$"2A#$3#0+(%/&2/#3
%#A#&2-$/'(5$1'#F+#-#./&'.$2./42/+&%$2(3$4-/&2%+('.$0+/+&%>
2(3$"2A#$023#$%4?%/2(/'2-$1&+*&#%%$'($,4&/"#&$%.2-'(*$3+;(
+4&$1'#F+#-#./&'.$3#A'.#%<$8($+4&$.+--2?+&2/'+($;'/"$C+(5
@-#./&+('.%$7C'(*21+&#:>$;#$"2A#$.+0?'(#3$+4&$1'#F+#-#./&'.
&#%#2&."$;'/"$%'-'.+($0'.&+H,2?&'.2/'+($1&+.#%%#%$/+$1&+34.#
2$0'.&+H02."'(#3$1'#F+#-#./&'.$2./42/+&$1&+/+/51#<$!"#
3#A#-+10#(/$;+&)$(+/$+(-5$'(A+-A#%$2./42/+&$%/&4./4&#
3#%'*($2(3$04-/'1-#$02/#&'2-%$'(/#*&2/'+(>$?4/$2-%+$04-/'1-#H
!"#"$%&'"()*+ ,")-.)/)0%'"0+ 1)2"0)(
!"#$%#&'(#)*+
Toon nonbos. Yoo Ko| (|-,oo_|no.o-s|o.ooo.sq;, Soo||oooo Shooo|qoh| (soo||-s_|no.o-s|o.ooo.sq;, Tooq \|oosooq (xs-|ooq_|no.o-s|o.ooo.sq;,
|o \o|ooq (x-ho_|no.o-s|o.ooo.sq;, Coo 8oo Kooo (b|-qoo_|no.o-s|o.ooo.sq; ooo Choo /o|no (nn-choo_|no.o-s|o.ooo.sq;
Co||oboo|o. S|oqooo osooch Loboo|o, - Soo, E|oc|oo|cs (S|oqooo; f|o L|o
,+-*%.#/012#34.#567#84(9(#(:#'#8+.;(.<.=9%+=#>+=%(0'=9*'9(%#?'>8<.#@.A.<(8.@
:%(>#(*%#B'9=4#>+=%(0:'B%+='9+(C#8%(=.??2#34+?#'=9*'9(%#=(>8%+?.?#/#8+.;(.<.=9%+=
94+C#:+<>#.<.>.C9?D#E+94#9E(#.<.>.C9?#(C#'#B.'>2
,+-*%.#/0F2#G#>+=%(0>'=4+C.@#8+.;(.<.=9%+=#>+=%(0'=9*'9(%#?'>8<.#B.+C-#9.?9.@2
%/#1$0'.&+H12//#&('(*$2(3$/"&##H3'0#(%'+(2-$0'.&+H
,2?&'.2/'+(<$8((+A2/'A#$02/#&'2-%$'(/#*&2/'+($2(3$0'.&+H
,2?&'.2/'+($0#/"+3%$"2A#$?##($4%#3$/+$%+-A#$02/#&'2-%$2(3
1&+.#%%'(*$.+012/'?'-'/5$'%%4#%<$G2,#&H?2%#3$?2/."
1&+.#%%'(*$/"2/$2--+;%$-2&*#H%.2-#$1&+34./'+($+,$0'.&+H
2./42/+&%$"2%$2-%+$?##($%4..#%%,4--5$'(A#(/#3<$!"'%
%'*(','.2(/-5$&#34.#%$1&+34./'+($.+%/%<
I'*4&#$JHK$%"+;%$/"#$C@=$1"+/+$+,$2$1'#F+#-#./&'.
0'.&+H2./42/+&$%201-#$3#A#-+1#3$4%'(*$+4&$?2/."$0'.&+H
,2?&'.2/'+($1&+.#%%<$!"#$2./42/+&$.+01&'%#%$J$1'#F+#-#./&'.
/"'($,'-0$0'.&+#-#0#(/%>$;'/"$/;+$#-#0#(/%$+($2$?#20<
I'*4&#$JHL$%"+;%$2$%201-#$?#'(*$/#%/#3$;'/"$+4&$-2%#&
,+-*%.#/0H2#34.#A+B%'9+(C#A.<(=+9"#(:#'#>+=%(0'=9*'9(%#?'>8<.#*C@.%#.<.=9%+='<
@%+A+C-#+?#9.?9.@#E+94#'#<'?.%#A+B%(>.9.%2
Double cantilever: 4
piezoelectric layers
Controlled vibration
electrically induced
Accelerometers (MEMS inertial sensors)
Accelerometers (MEMS inertial sensors)
Automotive:
- Airbag: accelerometers
- ESP: gyroscopes
The smallest gyroscope
is a piezoelectric one by
Seiko-Epson: 21 mm
3

Source: WTC
Applications of MEMS accelerometers and gyroscopes in
consumer products

Consumer
product
Function Examples of products MEMS inertial
device(s)
Status of
commercialisation
Cell phones Pedometer,
image rotation,
menu scroll,
gaming, free-fall
detection (HDD
protection),
navigation
NTT DoCoMo pedometer
(2003)
Vodafone image rotation
(2004)
Samsung SGH E760
Nokia 3230 (navigation
and gaming)
2- or 3-axis
accelerometer,
1- or 2-axis
gyroscope
Accelerometer in cell
phone since 2003
Gyroscope expected in
20072008
PDA Navigation IMU,
Web content
navigation
Toshiba PocketPC e740 2- or 3-axis
accelerometer
2-axis
gyroscope
Demonstrator in 2002
at Paris PDA show
Digital still
cameras
(DSC)
Image
stabilisation
All Panasonic DSCs, e.g.
Lumix ($200), Pentax
Optio A10 ($350)
Canon, Sony DSCs
Two 1-axis
gyroscopes or
one 2-axis
gyroscope
Two 2-axis
accelerometers
Gyroscope established
since late 1990s
Accelerometer
emerging
Camcorders Image
stabilisation,
free-fall
detection (HDD
protection)
Upper end: Panasonic
(over $1500)
High end: JVC 30 Gb,
Toshiba 60 Gb
Two 1-axis
gyroscopes or
one 2-axis
gyroscope
Gyroscope established
since late 1990s
Laptops Free fall
detection (HDD
protection), GPS
dead-reckoning
assist (anti-theft)
IBM, Toshiba, Apple
laptops
2- or 3-axis
accelerometer
Free-fall detection
established
Other applications
emerging
MP3 players Free fall (HDD
protection)
iPod with hard disc drive 3-axis
accelerometer
Established
Others: toys,
games,
personal
transporter,
robots
Realistic motion Nintendos Kirby Tilt-n-
Tumble GameBoy
Microsoft gamepad
Sidewinder Freestyle
Pro, Segway, Sony Aibo
robot, Sony PS3
2- or 3-axis
accelerometers,
1- or 2-axis
gyroscopes
Established
Accelerometers for aerospace and military ap.
Accelerometers for aerospace and military applications
Piezoelectric accelerometers
Piezoelectric accelerometers
Accelerations up to 200000g
0.2 grams, 5.21 mm long
From 7 to < 1 grams
Full integration and isolation (aerospace)
Full integration and isolation (for aircraft)
Full isolation from acoustic,
thermal effects to achieve
response at <0.1 Hz
Full integration of
electronic circuits