Advances in Energy Harvesting Using Low Profile Piezoelectric Transducers
Advances in Energy Harvesting Using Low Profile Piezoelectric Transducers
Advances in Energy Harvesting Using Low Profile Piezoelectric Transducers
DOI 10.1007/s10832-007-9043-4
Received: 31 May 2006 / Accepted: 8 September 2006 / Published online: 14 March 2007
# Springer Science + Business Media, LLC 2007
Table 1 Sources of energy available in the surrounding which are/can be tapped for generating electricity.
Breathing, blood pressure, Aircraft, UAV, Bridges, roads, tunnels, Motors, compressors, Wind, solar, temperature
exhalation, body heat helicopter, farm house structures chillers, pumps, fans gradient, daily
automobiles, trains temperature
Walking, arm motion, finger Tires, tracks, peddles, Control-switch, HVAC Conveyors, cutting and Ocean currents, acoustic
motion, jogging, swimming, brakes, shock systems, ducts, cleaners, dicing, vibrating waves, EM waves, RF
eating, talking absorbers, turbines etc. mach. signal
power are possible to be continuously harvested from practical method is to develop self-powered sensors by
human and industrial activity. harvesting energy from the wasted energy. Practically, on-
A large network with several sensor nodes and data site small scale successful harvesting of electrical energy
acquisition components requires a centralized energy source from the available energy source (as listed in Table 1) can
that has to be charged or replaced over time. In remote solve various challenges and enable new applications.
applications such as structural health monitoring of aircrafts Figure 1 compares the various alternatives for power
or ships, recharging, battery replacement or wiring can be sources. The comparison is made using the experimental
very tedious and expensive task. In many other cases, these values available in literature and the data measured in our
operations may be prohibited by the infrastructure. Further, laboratory [1–14]. The outside solar energy has the
in order for the sensor nodes to be conveniently placed and capability of providing power density of 15,000 μW/cm3
used they should be as small as possible which puts an which is about two orders of magnitudes higher than other
upper limit on their life time. If an electronic device with a sources. Definitely, solar energy is a very attractive source
1 cm3 of non-rechargeable lithium battery (at max energy for powering sensor networks and the solar technology has
density of 2,880 J/cm3 or 800 Wh/l) were to consume matured over the years. One of the major challenges in the
100 μW of power on average, the device would last implementation of solar technology on “energy on demand”
333 days. A lifetime of approximately 1 year is not platforms has been the requirement of bulky electronics.
practical. Even though the nodes in the wireless network Further, the variation in light intensity (cloudy vs. sunny
will be much smaller than the 1 cm3 area, the power day) can drop the efficiency significantly. Also, the solar
requirements will force them to use a battery of much larger power drops down significantly inside the building. The
size enhancing the system volume [2–8]. other most attractive source is kinetic energy comprising of
There are four possible ways to realize a distributed mechanical vibrations, air flow and human power. The
sensor network with adequate performance as following (1) kinetic energy can be converted into electric energy using
enhance the energy density of the storage systems (2) reduce piezoelectric, electromagnetic or electrostatic mechanism. It
the power consumption of the sensor, (3) develop self- can be easily shown that piezoelectric transducers are more
powered sensors by generating or harvesting energy or (4) suitable as kinetic to electrical energy converters. In
transmit the power from a centralized source to the sensor. addition to the advantage of being smaller and lighter the
Out of these various possible solutions the most efficient and piezoelectrics have three times higher energy density as
Table 2 The daily average power consumption for a wearable device with three functions.
Profile (24 h)
Based on-Shutdown 100 μW; MP3 110 mW; Organizer 50 mW; Communication 300 mW; Standby 10 mW.
J Electroceram (2007) 19:165–182 167
Table 3 Average power consumption of common household devices. generated is proportional to frequency and strain and higher
Product Average power consumption energy can be obtained by operating at the resonance.
Fig. 1 Comparison of various potential power sources for the sensor networks [7, 14, 18, 26]
168 J Electroceram (2007) 19:165–182
proper selection of the materials and design. The loss factor response of the electromechanical system it is easier to
(roughly equal to 2ς) for some commonly used structural perform the analysis using the equivalent circuit model.
materials are quite small, e.g., aluminum ∼0.007 and steel Table 5 lists the analogy between the mechanical and
∼0.05. In terms of the acceleration the above expression can electrical parameters that can be used to derive the
be written as: equivalent circuit for a given mechanical system [18]. The
dimension of mechanical impedance is same as that of
m! 2
Pmax ¼ ð4Þ mechanical resistance and is expressed in the same unit,
45n B N·s/m, often defined as mechanical ohms. Figure 6 shows
which implies that output power is proportional to the square the electrical equivalent circuit of the vibration source and
of the acceleration. Another important conclusion that can be elastic load. These equivalence relations can be used in
drawn from Eq. 4 is that output power is directly analyzing the performance of the system.
proportional to the proof mass of the system and thus In the off-resonance condition the modeling of the
reducing the size of the converter reduces the conversion generator can be significantly simplified as illustrated in
efficiency. Table 4 provides a list of acceleration and peak Fig. 7. Figure 7a shows the schematic of the piezoelectric
frequency values for the common vibration sources [7]. It bimorph transducer of height H, length L and width W
has been found that accelerations of 2.5 m/s2 at 120 Hz are subjected to the AC force F. The load dependence of the
typical of many low level vibrations. Figure 5 shows the bimorph in the frequency range far from the resonance can
example of a vibration spectrum from a car driven at be computed by using the equivalent circuit representation
65 mph and washing machine. In this figure the spectral shown in Fig. 7b. In this circuit the voltage source is taken
peaks at 1.2, 14 and 35 Hz correspond to fundamental to be the open circuit voltage across the bimorph. The
frequency, structural imbalance, and engine speed (http:// voltage across the load can than be expressed as [19]:
www.vibrationdata.com/).
The power generated by the converter is measured across RLoad
VLoad ¼ VOC ð5Þ
the electrical load. In order to analyze the frequency RLoad þ jwC
1
þ Rs
ðoff resonanceÞ
ð10Þ
2
2k31 tc c
m 2
p
RCb2
2
e Am k2
P¼ pffiffiffiffiffiffi ð11Þ
2k 4 B 2 þ k31
4 ðRC Þ2 k þ 4 B k 2 RC
b 31 b km þ 4 B m
2
2
Vibration Source Peak acceleration Frequency of
8 lw d31 (m/s2) peak (Hz)
CA ¼ 2 ð9Þ
π b sE11
Base of a 5 HP 3-axis 10 70
where l is the length and w is the width of the rectangular machine tool
plate, s is the elastic compliance and d is the piezoelectric Notebook computer while 0.6 75
constant. A rigorous analysis for the piezoelectric bimorph CD is being read
loaded in cantilever form with the tip mass m has been Clothes dryer 3.5 120
Second story floor of a wood 0.2 100
conducted by Roundy et al.. They used the transformer
frame office building
based equivalent circuit for the analysis and arrived at the
170 J Electroceram (2007) 19:165–182
Fig. 5 Power spectral density of the car (2003 Ford Taurus) driven on
highway at 65 mph. The data was recorded using Nicolet Vision
System. (Source: Website—http://www.vibrationdata.com. Permission
granted by T. Irvine to reproduce the figure). The spectral peak at
1.2 Hz is the fundamental frequency of the vehicle mass and
suspension system and is typical for passenger cars. Other peaks can
be associated to the tire imbalance, engine vibrations, and rotational
vibrations from drive train system, alternator, and air conditioning unit
higher charge output. For example—a rectangular bimorph Fig. 6 Electrical equivalence of the mechanical system for vibration
of the dimension 60×35×0.6 mm3 mounted in cantilever source and elastic load representing the relationship shown in Table 5
(adopted from: Permission granted by H. Kim to reproduce the
fashion and operating in bending mode has the resonance
figure). The unit of mechanical impedance is similar to resistance and
frequency of 65 Hz. By adding four additional layers to this is given as N s/m
bimorph the resonance frequency can be lowered to 45 Hz
[21]. An multimorph (more than six layers) of size 24×6× analysis revealed that large part of the impact energy was
0.3 mm3 can sense an acceleration frequency of as low as returned to the steel ball in the form of kinetic energy
0.5 Hz [22]. Results on accelerometer devices have shown causing it to bounce off the plate. In case the steel ball did
that by increasing the number of layers from six to 12 the not bounce off rather vibrated with the piezoelectric
charge output is increased by 30% [22]. diaphragm the efficiency of the system was computed to
a F
3 Energy conversion efficiency L
δl δz
The efficiency of the vibration to electric energy conversion
process is dependent upon the method of applying the
oscillating stress onto the piezoelectric and material
H
parameters. Umeda et al. have analyzed the electric energy
l
generation from the piezoelectric diaphragm structure
which is vibrated by dropping a steel ball (d31 mode) [23, b
24]. The impact of the steel ball induced vibration in the
piezoelectric diaphragm resulting in AC voltage. Their
UTA
condition
Morgan
15000
DongIl
Energy harvesting in the on-resonance condition is limited
m /N)
Ferroperm
due to limitation in terms of vibration amplitude at high
2
APC
12000
d33.g33 (10
Fuji
harvesting device operating under high frequency reso-
APC
EDO
DongIl
piezoelectric transformer. A piezoelectric transformer con-
Ferroperm
EDO
9000 verts the electrical energy into mechanical energy at the input
Morgan
terminal and mechanical to electrical energy at the output
Channel
QuickPack Bimorphs
174 J Electroceram (2007) 19:165–182
Smart Materials Corp. have recently revealed a micro- QuickPack (rigid) and PowerAct (flexible) transducer
energy harvester TI MSP430 μController based on MFC’s are packaged bimorphs in a protective skin with pre-
which is integrated with EnOcean STM100 and has the attached electrical leads, producing a highly reliable
ability to measure the local temperature, pressure and other component with no soldered wires (http://www.mide.com).
environmental variables [40]. The picture of the prototype These transducers have been demonstrated in energy
is shown in Fig. 16a. MicroStrain has also developed a harvesting applications by exciting them with a mechanical
wireless sensor node using the MFC’s as shown in Fig. 16b shaker at resonance frequency and measuring the response.
which has been demonstrated to operate with one of the QuickPack model QP20W having dimensions of 5.08×
highest efficiencies [41]. 3.81×0.051 cm3 with a nominal capacitance of 200 nF
NASA LaRC’s Radial Field Diaphragm (RFD) is an
exciting development in the area of low profile piezoelec-
tric transducers. The RFD comprises of a thin circular
piezoelectric ceramic disk sandwiched between two poly-
imide dielectric films with copper-etched spiral electrodes
as shown in Fig. 11d [42]. On each spiraled electrode
pattern, positive and negative electrodes spiral inward to the
center of the disk in a serpentine manner. This novel
electrode pattern induces an electric field into the piezo-
electric ceramic that extends out radially from the center of
the wafer. The dielectric film serves as the electrode carrier
and insulator. At low frequencies of DC—10 Hz, RFD’s
have been shown to provide 2–3 times higher displacement
Fig. 14 Variation of the output voltage as a function of the fiber
as compared to conventional unimorphs [42]. Experimental
composite thickness for various fiber diameters (Source: Advanced
demonstration of the RFD’s in the energy harvesting Cerametrics Inc. website, http://www.advancedcerametrics.com/. Per-
application has not yet been reported. mission granted by ACI to reproduce the figures)
176 J Electroceram (2007) 19:165–182
driven at 50 Hz has been shown to provide 30.66 mW surface of the earth heats and cools unevenly, creating
power with the open circuit voltage of 68 V [15, 43]. atmospheric pressure zones that make air flow from high-
In general, piezoelectric bimorph transducer is the to low-pressure areas. Approximately 2% of the sun’s
simplest and well known low frequency resonance struc- energy reaching the earth is converted into wind energy.
ture. Bimorphs can be easily mounted into several config- For a flow of air with velocity ν and density ρ through unit
urations as shown in Fig. 17 providing high degree of area A perpendicular to the wind direction, the kinetic
adaptability to the available vibrations. Recently, we have energy per unit time is given by P (http://www.Windworkers.
demonstrated wind energy harvesting using the bimorph com; http://www.windturbinewarehouse.com; http://www.
transducer commercially available from APC International, windpowerstore.com) [44]:
Mackeyville, PA (http://www.americanpiezo.com). Bimorph
transducer has following advantages in terms of suitability P ¼ 1=2mv2 ¼ 1=2ðAvρÞv2 ¼ 1=2Aρv3 ð18Þ
for windmill:
The air density varies with altitude, atmospheric pressure
1. They have enough mechanical strength in the vibration and temperature. The power density function p(ν) of a
frequency range of 1–10 Hz. In this range the applied windmill may be expressed as (http://www.Windworkers.
load on the bimorph can be of the order of few com; http://www.windturbinewarehouse.com; http://www.
Newton’s. Laboratory scale measurements have shown windpowerstore.com):
that a bimorph vibrating under a force of 2 N at low
frequencies of 10 Hz do not suffer from any mechanical pðν Þ ¼ Cp 1=2Ar ρν 3 f ðν Þ and
ð19Þ
degradation. k1 ðν= Þ
k
Table 6 Comparison of the sensor and actuator characteristics for the d33 and d31 mode MFCs (Permission granted by T. Daue, Smart Materials
Corp to reproduce the figure).
Device Operation voltage Capacity Sensor Characteristic Actuator characteristic Generator characteristic
þ eff eff
Vop ½V Vop ½V Cpol [nF/cm2] d31 ½pC=N d31 ½pC=N Strain/Volt. [μstrain/V] Charge/Strain[pC/ppm]
3–3 MFC 1,500 −500 0,42 460 – 0,7..0,9[0 ... 1,500 V] 1,670 [>100 V]
3–1 MFC 360 −60 4,5 – −370 −2 [0 ... 360 V] 3,250 [<100 V]
J Electroceram (2007) 19:165–182 177
7 Energy harvesting using piezoelectric polymers temperature operations [50–54]. Our investigations have
shown that the piezoelectric properties maintain high
Researchers from Ocean Power Technologies have pro- magnitude until 70 °C. Deposition of PVDF films is now
posed energy harvesting Eel that uses piezoelectric polymer a well established technique [51, 54]. In the sol-gel process,
to convert the mechanical flow energy available in oceans PVDF powder is dissolved in a solvent composed of 90%
and rivers to electrical power [49]. In a nonturbulent flow, acetone and 10% dimethylformamide (DMF) and sonicated
the bluff body regularly sheds alternating vortices on either in an ultrasonic water bath. The solution is applied on the
side of the bluff body with a frequency and size substrate through spin coating at room temperature, fol-
corresponding to the flow speed and width of the bluff lowed by drying at room temperature or in an oven.
body. The resulting pressure differential caused by the Sohn et al. have investigated the feasibility of using
vortices forces the Eel to move in an oscillating motion. PVDF films of thickness 9–110 μm for powering the Bio-
The resulting strain on the piezoelectric polymer generates MEMS using the FE analysis package MSC/NASTRAN
a low frequency ac voltage signal. The electrical power, P,
produced by an Eel undulating in water flow is given by [49]:
h1 h2 h3 ArV 3
P¼ ð20Þ
2
Voltage [VDC]
Power [mW]
30
PVDF has a relatively low value of d31 (3.3×10−11 pC/N)
90
resulting in lower power. This prompted research on 20
electrostrictive polymers where it is possible to induce 60
large piezoelectric effect by applying high DC bias field
10 30
due to Maxwell stress. Electrostrictive polymers recently
have been discovered that generate large strain (above 5%)
0
under moderate electric field intensity (400–800 V on a 0
0 200 400 600 800 1000
20 μm film) [58]. Poyl(vinylidene fluoride-trifluoroethylene),
or P[VDF:TrFe] copolymers have been shown to exhibit d31 Resistive Load [kΩ]
of 8.0×10−11 pC/N, and polyurethane has d31 of 1.7× a Without DC-DC converter
10−10 pC/N. A theoretical investigation conducted by Liu et
al. has shown that the energy densities of the order of 40 40
10 10
8 Energy storage circuit
0 0
In the bulk harvester cases, a DC-to-DC converter circuit is
commonly used to harvest the generated electric energy and
0 10 20 30 40 50
transfer it across the load as schematically shown in Fig. 20
Resistive Load [kΩ]
[15, 43, 47, 48]. The impedance for applications such as
charging a battery and lighting a bulb is in the order of few b Output of DC-DC converter
Fig. 21 Output voltage and output power of the DC–DC converter as
hundred ohms while the piezoelectric is a high impedance
a function of resistive load condition with 1 kHz switching frequency
load. This converter circuit is called a buck converter and it and 2% duty cycle using single layered high-g PZT Cymbal
operates in the discontinuous current mode (DCM). The transducer at 8 N
180 J Electroceram (2007) 19:165–182
Several other techniques are used in conjunction with the integrated on the chip, and (2) LDR can only perform
circuit for reducing the matching impedance to lower levels step-down operation. Since the input voltage of the dc–dc
such as multilayering the piezoelectric transducer structure converter can be as low as 0.1 V, LDR is not suitable [62].
and increasing the area aspect ratio (area/thickness). Amirtharajah and Chandrakasan have reported a test chip
Recently, Guyomar et al. have presented a nonlinear fabricated using CMOS process that integrates an ultra-low
processing technique “Synchronized Switch Harvesting on power controller to regulate the generator voltage and a low
Inductor” (SSHI) for harvesting energy, which consists of a power subband filter DSP load circuit [63]. The entire
switching device in parallel with piezoelectric element [60]. system was found to consume a total of 18 μW power.
This device is composed of a switch and an inductor Several other methodologies have been proposed in
connected in series. The switch is in open state except when literature power conditioning of piezoelectric micro devices
the maximum displacement occurs in the transducer. At that [64–69]. These results clearly indicate the possibility of
instant the switch is closed and the capacitance of the realizing an efficient power harvesting system based on the
piezoelectric element and inductance together constitute an architecture depicted in Fig. 22. One other challenge which
oscillator. The switch is kept closed until the voltage on the needs to be tackled in order to implement this scheme for
piezoelectric element has been reversed. In case of powering sensors is design and fabrication of an ASIC
nonlinear AC device, a resistive load is directly related on since outputs from individual layers have widely different
the piezoelectric element in parallel with the switching impedance.
device. This nonlinear technique has been shown to In this review, a brief description is presented on the
significantly enhance the performance of the energy middle layer consisting of piezoelectric cantilevers. Lu et al.
harvesting circuit and will be well suited for the resonating have analyzed the case of ‘31’ mode piezoelectric micro-
structures [60]. cantilever structure and developed a model for power and
coupling efficiency (kinetic to electric energy conversion) as
a function of frequency [70]. It was found that power
9 Energy harvesting using piezoelectric micro devices increases with frequency while the coupling efficiency
decreases. The modeling results also indicated that at low
Self-powering at chip scale can be achieved by developing frequencies the output power requires smaller dielectric
a smart architecture which utilizes all the environmental constant (e33) while at higher frequencies it’s the ratio of
resources available for generating electrical power as shown piezoelectric constant (e31) and dielectric constant that
in Fig. 22 for vibrations, light, and temperature gradients. makes significant contribution. The power generated from
This architecture has been proposed and analyzed by Priya the cantilever beam depends on several factors including
et al. for the self powered ultra high efficiency sensor resonant frequency, end mass, and mechanical damping.
network with static and mobile nodes [61]. The power The resonant frequency of the free-end beam is approxi-
harvested can be stored using switched-capacitor dc–dc mately given by:
converter (regulated charge pump) through on-chip chem- rffiffiffiffiffiffiffiffiffirffiffiffiffiffiffiffiffiffiffiffi
ical storage media. Regulated charge pump is preferred k Eh3 w
fo ¼ ¼ ð22Þ
over the switched-mode power converter (SMPC) and low- m 4L3 m
dropout regulator (LDR) because (1) SMPC requires where k is the mechanical spring constant, m is the mass, E
inductor for dc–dc conversion; however the inductor to is the Young’s modulus, h is the height of the beam, w is
sustain milliampere current is complicated for being the width of the beam, and L is the length of the beam. The
viscous damping constant for squeeze-film damping of a
plate can be approximated by [7]:
KA2 η
B ¼ ð23Þ
x3
where ζ is the damping constant, K is a constant of the
order of 0.4–0.5, A is the area of beam, η is the viscosity,
and x is the gap between the beam and substrate. From
Eqs. 22 to 23 it can be seen that a large mass and gap (x) is
desirable for obtaining high power.
Fang et al. have reported results on the PZT cantilever
Fig. 22 Schematic representation of the power module harvesting
beam of the size 2,000 μm (length)×500 μm (width)×
energy from environmental sources light, vibrations and temperature 12 μm (thickness) with Ni proof mass of size 500×
gradients 500 μm2 fabricated on to Si/SiO2/Ti/Pt substrate [71]. A
J Electroceram (2007) 19:165–182 181
power level of 1.15 μW with 432 mV AC voltage was the characterization methods so that a comparative study
measured across a 20.4 kΩ load under resonant vibration of between the systems could be realized.
about 609 Hz with 1 g acceleration. This is an interesting
result since the power magnitude can be enhanced by Acknowledgement The author is grateful for the support from
arraying the cantilevers in parallel with in-phase operation. Texas ARP grant 003656-0010-2006.
Jeon et al. developed a d33-mode piezoelectric power
generating device consisting of flat cantilever structure
with interdigitated electrode pattern and a proof mass at the References
end. [72] In this experiment it was found that a 170 μm×
1. J.L. Gonzalez, A. Rubio, F. Moll, Int. J. Soc. Mater. Eng. Resour.
260 μm PZT beam generates 1 μW of continuous electrical
10(1), 34–40 (2002)
power to a 5.2 MΩ resistive load at 2.4 V DC corresponding 2. J.M. Rabaey, M.J. Ammer, J.L. da Silva, D. Patel, S. Roundy,
to an energy density of 0.74 mWh/cm2. The measurements IEEE Computer 33, 42 (2000)
were conducted at the first resonant frequency of 13.9 kHz 3. B. Gates, The Economist, Special Issue: The World in 2003,
December, 99 (2002)
[72]. Several other micro-machined devices utilizing PZT
4. Hitachi, Hitachi unveils smallest RFID chip, RFID Journal, March
cantilever and diaphragm structure have been reported in 14, (2003)
literature [73–75]. 5. T. Douseki, et al., in ISSCC Digest of Technical Papers, 84 (1996)
Micro-devices are currently the most pursued areas in the 6. T. Ieki, et al., in Proceedings of IEICE, September, 39 (1996)
power harvesting and this subject will assume increasing 7. S. Roundy, P.K. Wright, J.M. Rabaey, Energy Scavenging for
Wireless Sensor Networks (Kluwer, Boston, 2004)
importance as the trend towards lab-on-chip system continues. 8. E.H. Calaway Jr., Wireless Sensor Networks (Auerbach, NY, 2004)
9. S. Roundy, E.S. Leland, J. Baker, E. Carleton, E. Reilly, E. Lai,
B. Otis, J.M. Rabaey, P.K. Wright, V. Sundararajan, Improving
10 Summary power output for vibration-based energy scavengers, Pervasive
Computing, January–March, 28–36 (2005)
10. V. Raghunathan, A. Kansal, J. Hsu, J. Friedman, M.B. Srivastava, in
Piezoelectric converters are prominent choice for mechanical to IEEE International Symposium on Information Processing in Sensor
electric energy conversion because the energy density is three Networks (IPSN), April 2005. (TR-UCLA-NESL-200503-10)
times higher as compared to electrostatic and electromagnetics. 11. M. Rahimi, H. Shah, G.S. Sukhatme, in Proceedings of IEEE
International Conference on Robotics and Automation, Taipei,
The electrical power generated is inversely proportional to the Taiwan, 14–19 Sept. 2003, pp. 19–24
damping ratio which should be minimized through proper 12. J. Paradiso, T. Starner, Energy scavenging for mobile and wireless
selection of the materials and design. The power is propor- electronics, Pervasive Computing, January–March, 18–27 (2005)
tional to the square of the acceleration and reaches maximum 13. S. Roundy, P.K. Wright, J. Rabaey, Comput. Commun. 26(11),
1131–1144 (2003)
at the resonance frequency. The resonance frequency of the 14. T. Starner, J.A. Paradiso, in Low-power Electronics Design, ed. by
piezoelectric transducer is dependent upon the configuration, C. Piguet (CRC, 2004), Chapter 45, pp. 1–35
size, and loading conditions. The efficiency (η) of the 15. G.K. Ottman, H.F. Hofmann, A.C. Bhatt, G.A. Lesieture, IEEE
conversion process in resonance condition is dependent Trans. Power Electron. 17(5), 669–676 (2002)
16. C.B. Williams, R.B. Yates, in The 8th International Conference on
upon the coupling coefficient and mechanical quality factor Solid-state Sensors and Actuators, and Eurosensors IX, Stock-
of the piezoelectric. In the off-resonance condition, the holm, Sweden, 25–29 June (1995) pp. 369–372
product (d.g) is directly related to the energy density. 17. C.B. Williams, C. Sherwood, M.A. Harradine, P.H. Mellor, T.S.
Recently, the condition for obtaining large magnitude of d.g Birch, R.B. Yates, IEE Proc., Circ. Devices Syst. 148(6) 337–342
(2001)
was derived to be as [29, 30]: 18. H.W. Kim, Ph.D. dissertation, The Pennsylvania State University,
jd j ¼ e n University Park, PA, 2006
19. S. Priya, Appl. Phys. Lett. 87 184101 (2005)
where e is permittivity of the material and n is a material 20. K. Uchino, Ferroelectric Devices (Marcel Dekker, New York,
2000), p. 169
constant having lower limit of 0.5. Several low profile
21. APC International, Piezoelectric ceramics: principles and applica-
transducers are available commercially which hold signifi- tions Mackeyville, ISBN 0-9718744-0-9, PA, USA, (2002)
cant promise for the energy harvesting applications. Results 22. K. Sato, K. Okamoto, Y. Fuda, T. Yoshida, Jpn. J. Appl. Phys. 33,
demonstrated on the MFCs, AFCs, and bimorphs have 5378–5380 (1994)
shown the possibility of realizing a self powered sensor node. In 23. M. Umeda, K. Nakamura, S. Ueha, Jpn. J. Appl. Phys. 35(5B)
3267–3273 (1996)
parallel to materials and transducer development, researchers 24. M. Umeda, K. Nakamura, S. Ueha, Jpn. J. Appl. Phys. 36(5B)
have also developed advanced energy storage circuits. Recent 3146–3151 (1997)
results on circuits utilizing nonlinear signal processing indicate 25. H.A. Sodano, D.J. Inman, G. Park, Shock Vibr. Dig. 36(3), 197–
205 (2004)
the efficiencies can reach to the theoretical maximum.
26. H.A. Sodano, G. Park, D.J. Inman, Strain 40, 49–58 (2004)
As the progress in the development of the power 27. M. Goldfarb, L.D. Jones, Trans. ASME, J. Dyn. Syst. Meas.
harvesting continues it will be important to standardize Control 121 566–571 (1999)
182 J Electroceram (2007) 19:165–182
28. T. Funasaka, M. Furuhata, Y. Hashimoto, K. Nakamura, Proc. 54. M. Benz, W.B. Euler, J. Appl. Polym. Sci. 89, 1093 (2003)
IEEE Ultrasonics Symposium 1, 959–962 (1998) 55. J.W. Sohn, S.B. Choi, D.Y. Lee, Part C: J. Mech. Eng. Sci., 429–
29. C.D. Richards, M.J. Anderson, D.F. Bahr, R.F. Richards, J. 436 (2005)
Micromechanics Microengineering 14, 717–721 (2004) 56. J.T. Cain, W.W. Clark, D. Ulinski, M.H. Mickle, Int. J. Parallel
30. F. Yamauchi, M. Takahashi, J. Phys. Soc. Jpn. 28(Suppl.) 313 Distrib. Syst. Netw. 4, 140–149 (2001)
(1970) 57. P. Niu, P. Chapman, R. Riemer, and X. Zhang, in 35th Annual IEE
31. R. Islam, S. Priya, Appl. Phys. Lett. 88, 032903 (2006) Power Electronics Specialists Conference, Aachen, Germany
32. R. Islam, S. Priya, J. Am. Ceram. Soc. 89(10), 3147–3156 (2006) (2004)
33. T.T. Wang, J.M. Herbert, A.M. Glass (ed.), The Applications of 58. Q.M. Zhang, J. Schienbeim, Electric EAP, in Electroactive
Ferroelectric Polymers (Chapman and Hall, NY, 1988) Polymer (EAP) Actuators as Artificial Muscles, 2nd edn., ed. by
34. From the catalog of Microfine Materials Technologies Ltd. Y. Bar-Cohen (SPIE, Bellingham, WA 2004), pp. 95–150
available at http://www.microfine-piezo.com 59. Y. Liu, K.L. Ren, H.F. Hofmann, Q. Zhang, IEEE Trans. Ultrason.
35. I.W. Chen, X.H. Wang, Nature 404, 168 (2002) Ferroelec. Freq. Control. 52(12), 2411–2417 (2005)
36. K. Mossi, Z. Ounaies, B. Ball, Smith, in Proceedings of the SPIE, 60. D. Guyomar, A. Badel, E. Lefeuvre, C. Richard, IEEE Trans.
Smart Structures and Materials, San Diego, CA, vol. 5053, Ultrason. Ferroelec. Freq. Control 52(4), 584–595 (2005)
(2003) pp. 423–435 61. S. Priya, D. Popa, F. Lewis, in 2006 ASME International
37. S. Leschin, R.B. Cass, F. Mohammadi, in UTA Workshop on Mechanical Engineering Congress & Exposition, Chicago, IL,
Energy Harvesting, 27 January 2006 5–10 November 2006
38. Application notes: http://www.advancedcerametrics.com 62. H. Lee, P.K.T. Mok, W.H. Ki, in Proceedings of IEEE International
39. Application notes: http://www.smart-material.com Symposium on Circuits and Systems, Geneva, Switzerland, May
40. T. Daue, in UTA Workshop on Energy Harvesting, 27 January 2000, vol. I, (2000), pp. 256–259
(2006) 63. R. Amirtharajah, A.P. Chandrakasan, IEEE J. Solid-State Circuits
41. S.W. Arms, C.P. Townsend, D.L. Churchill, J.H. Galbreath, S.W. 33, 687–695 (1998)
Mundell, in UTA Workshop on Energy Harvesting, 27 January 64. B.H. Calhoun, D.C. Daly, N. Verma, D.F. Finchelstein, D.D.
(2006) Wentzloff, A. Wang, S-H. Cho, A.P. Chandrakasan, IEEE Trans.
42. R.G. Bryant, R.T. Effinger IV, I. Aranda Jr., B.M. Copeland Jr., Comput. 54, 727–740 (2005)
E.W. Covington III, in Proceedings of SPIE, Smart Structures 65. N. Hama, A. Yajima, Y. Yoshida, F. Utsunomiya, J. Kodate,
and Materials—Active Materials: Behavior and Mechanics, Paper T. Tsukahara, T. Douseki, in 2002 Symposium on VLSI, Circuits
4699-40, San Diego, CA, (2002) Digest of Technical Papers, Paper 20.2, 280–283 (2002)
43. G.A. Lesieture, G.K. Ottman, H.F. Hofmann, J. Sound Vib. 269, 66. P.M. Lin, L.O. Chua, IEEE Trans. Circuits Syst. CAS-24, 517–
991–1001 (2004) 530 (1977)
44. A.G. Davenport, M. Novak, in Shock and Vibration Handbook, 67. J. Siebert, J. Collier, R. Amirtharajah, in ISLPED’05, August 9–10
ed. by C.M. Harris (McGraw-Hill, NY, 1988) (San Diego, California, 2005), pp. 315–318
45. S. Priya, C. Chen, D. Fye, J. Zhand, Jpn. J. Appl. Phys. 44, 104 (2004) 68. S. Roundy, P. Wright, J. Rabaey, Comput. Commun. 26, 1131–
46. C. Chen, R. Islam, S. Priya, IEEE Trans. Ultrason. Ferroelec. 1144 (2003)
Freq. Control 53(3), 656–661 (2006) 69. J. Rabaey, A. Chandrakasan, B. Nikolic, Digital Integrated
47. H. Kim, A. Batra, S. Priya, K. Uchino, R. E. Newnham, D. Circuits: A Design Perspective, 2nd edn. (Prentice-Hall, Upper
Markeley, H.F. Hofmann, Jpn. J. Appl. Phys. 43(9A) 6178 (2004) Saddle River, NJ, 2003)
48. H. Kim, S. Priya, K. Uchino, R.E. Newnham, J. Electroceram. 15, 70. F. Lu, H.P. Lee, S.P. Lim, Smart Mater. Struct. 13, 57–63 (2004)
27–34 (2005) 71. H.-B. Fang, J.-Q. Liu, Z.-Y. Xu, L. Dong, D. Chen, B.-C. Cai,
49. G.W. Taylor, J.R. Burns, S.M. Kammann, W.B. Powers, T.R. Y. Liu, Chin. Phys. Lett. 23, 732–734 (2006)
Welsh, IEEE J. Oceanic Eng. 26, 539–547 (2001) 72. Y.B. Jeon, R. Sood, J.-H. Jeong, S.-G. Kim, Sens. Actuators A
50. Q.X. Chen, D.A. Payne, Meas. Sci. Technol. 6, 249–267 (1995) 122, 16–22 (2005)
51. C.M. Wang, M.C. Kao, Y.C. Chen, Y.H. Lai, Jpn. J. Appl. Phys. 73. M.J. Ramsay, W.W. Clark, Proc. SPIE 4332, 429–438 (2001)
42, 170 (2003) 74. R. Sood, Y.B. Jeon, J.H. Jeong, S.G. Kim, in Tech. Digest 2004
52. H. Kawai, Jpn. J. Appl. Phys. 8, 975–976 (1969) Solid State Sensor Actuator Workshop, Hilton Head, South
53. T.T. Wang, J.M. Herbert, A.M. Glass (ed.), The Applications of Carolina, 2004
Ferroelectric Polymers (Chapman and Hall, NY, 1988) 75. S. Kim, W.W. Clark, Q.-M. Wang, Proc. SPIE 5055, 307–318 (2003)