UTILIZATION OF DIFFERENT CHARGE POLARIZATIONS ON PZT
DIAPHRAGM OF ACOUSTIC ENERGY HARVESTERS
K. Tsujimoto, S. Tomioka, S. Kimura, S. IIzumi, K. Tomii, T. Matsuda, and Y. Nishioka*
Department of Precision Machinery Engineering, College of Science and Technology, Nihon University,
7-24-1 Narashinodai, Funabashi, Chiba 274-8501, Japan
*Presenting Author: nishioka@eme.cst.nihon-u.ac.jp
Abstract: This paper presents improved power generation performances of a PZT MEMS acoustic energy
harvester with dual top electrodes to utilize the different polarization of charges on the surface of a vibrating PZT
diaphragm. Two independent energy harvesters were defined between the peripheral top electrode, the central top
electrode and the lower common electrode. The power density of the acoustic energy harvesters was increased by
connecting the two energy harvesters in series, which can be simply realized by using the top electrodes as the
harvester terminals. The PZT acoustic energy harvester had a diaphragm with a diameter of 2 mm, and it exhibited
the generated power of 9.5x10-11 W and the energy density of 23.7 µW/m2 under the sound pressure level of 100
dB (0.01 W/m2) at 4.92 kHz. The device generated nearly two times larger power than other devices with similar
diaphragm areas.
Keywords: Acoustic energy harvester, PZT, diaphragm
INTRODUCTION
Energy harvesters have attracted much attention for
possible application to independently operating
intelligence systems [1,2]. Pb(Zr,Ti)O3 (PZT) is an
excellent candidate for harvesting power from ambient
vibration sources because it can efficiently convert
mechanical strain to electrical charge [3,4]. Several
bulk PZT piezoelectric energy harvesters have been
reported [5-9]. Also, energy harvesters using thin PZT
films have recently been widely studied since they can
be integrated into silicon integrated circuits (ICs) [1015].
Horowitz et al. have recently reported on a MEMS
acoustic energy harvester that contains a PZT
piezoelectric film as a diaphragm [16]. They claimed
that their energy harvester with the PZT (0.27 µm
thick) diaphragm with the diameter of 2.4 mm
exhibited the power density of 0.34 µW/cm2 at 149 dB
sound pressure level (SPL) under the first-resonance
vibration mode. The reason for measuring the energy
harvester at such a high sound pressure level is the
need for health monitoring systems in aircrafts where
noises from engines are extremely high and wireless
intelligent sensor systems are needed.
Shinoda et al. [17] fabricated PZT energy harvesters
with similar sizes to those of Horowitz et al.,
compared the performances of their energy harvesters
with those of Horowitz et al. at 100 dB SPL, and
reported that higher generated power could be realized
by vibrating PZT diaphragms at the third-resonance
frequency. Kimura et al. reported that PZT acoustic
harvesters with non-wet electrode fabrication
processes exhibited further improved power density at
the first resonance [18]. However, these previous
reports on the PZT acoustic energy harvesters did not
consider to utilize the different polarity charges in the
central part and the peripheral part of the diaphragms.
In this paper, we report on further improved
performances of PZT acoustic energy harvesters that
utilize different charge polarizations at the peripheral
part and the central part of a PZT diaphragm at the
first-resonance vibration.
DEVICE STRUCTURE
Figure 1(a) shows the top view of the PZT energy
harvester investigated here. The two top Al electrodes
cover the peripheral area and the central area of the
PZT diaphragm, respectively. The schematic crosssection of the energy harvester is shown in Fig. 1(b). A
piezoelectric capacitor with the diaphragm structure of
Al (0.1 µm)/PZT (1 µm)/Pt (0.1 µm)/Ti (0.1 µm ) was
formed on 1.5-µm-thick SiO2. The cavity of the
diameter of 2.0 mm was fabricated on a 300-µm-thick
silicon wafer. When sound pressure is applied from above
the device, the diaphragm vibrates, resulting in the
deflection of the diaphragm. The deflected PZT
capacitor generates a voltage difference between the
two electrodes of Al and Pt/Ti owing to the
polarization of PZT. In case of exciting the first
resonance, the generated charge polarizations were
different between the peripheral part and the central
part. This way, the dual top electrode dimensions were
determined to utilize the different charge polarizations.
Therefore, the independent top electrodes that covered
the peripheral area and the central area to form the
independent two energy harvesters were fabricated as
illustrated in Fig. 1(c). The two energy harvesters were
defined between the peripheral electrode and the
common electrode, and the central electrode and the
common electrode. A connected energy harvester was
also defined between the two top electrodes. The
detailed fabrication conditions have been already
reported [17,18].
Figure 2 shows the strain distribution of the PZT
capacitor/SiO2 diaphragm under a static pressure. The
sign of the strain changes at a radius of 730 µm. The
different signs of the strain correspond to different
Positive strain 270 µm
Negative strain 730 µm
Fig. 2. Calculated strain distribution on the 1-µm-thick
PZT film deposited on 1.5-µm-thick SiO2 under a static
pressure. This analysis was carried out using the finite
element analysis method.
60000
Intensity (counts)
signs of charges. Therefore, the dimensions of the
peripheral and the central electrodes were determined
to cover the areas where the strains were relatively
large. The area of the diaphragm that was not covered
with the electrodes was where the strain was relatively
small to avoid the unnecessary increase in energy
harvester capacitance.
The X-ray diffraction measurements of the
crystalline orientation of (111) are shown in Fig. 2.
The sol-gel PZT film investigated here showed smaller
peaks for the PZT (001) and (002) perovskite
structures. However, the large peak of PZT (111)
indicated that the PZT crystal is mostly oriented in the
(111) direction. From the measured hysteresis
characteristics of the PZT capacitor, the resultant PZT
film exhibited ferroelectric behaviors with a coercive
field of 168 kV/cm and a remanent polarization of 19.2
µC/cm2 after poling at 20 V for 15 min. It also showed
a relative dielectric constant of 620 and a dielectric
loss of 0.05.
50000
Pt (111)
40000
30000
PZT (111)
20000
PZT (001)
10000
PZT (002)
0
20
30
40
50
60
2θ (deg)
(a)
Top electrodes
PZT
Common
electrode
Cavity
(b)
Fig. 3. XRD patterns for the 1-µm-thick PZT thin film
deposited using sol-gel process.
The resonance frequencies under a sound pressure
of 100 dB SPL (0.01 W/m2) were measured using the
characterization system described in previous reports
[17,18]. In order to measure the resonance frequency
of the energy harvester, an amplifier with a gain of
10,000 was connected to amplify the output voltage
signals from the energy harvester. In the case of
measuring the generated power, a resistance was
connected to the energy harvester, and the voltage
between the two terminals of the load resister was
monitored using an oscilloscope.
RESULTS AND DISCUSSION
(c)
Fig. 1 Structure of PZT energy harvester; (a) top view,
(b) Schematic cross section, (c) expected charge
distribution, and (d) concept of connected energy
harvester.
Figure 4 compares the frequency dependencies of the
output voltages for the peripheral device and the
central device under the sound pressure level of 100
dB. The output voltage, E was determined by
measuring the open-circuit voltage between the two
electrodes of the energy harvesters. A sharp resonance
for the first resonance for the peripheral device and the
central device were observed at 4.29 kHz, which was
also suggested by the vibration analysis using the finite
element method. The frequency dependence of a
device with the same diaphragm diameter but with the
circular-shaped top electrode covering whole surface
of the diaphragm was also compared as a reference
[18].
The resonance frequencies were measured using
the same characterization system described previously
[17,18]. AC sinusoidal voltage signals with different
frequencies were generated by the function generator
and amplified by the power amplifier. The amplifier
was connected to the speaker. The gain of the power
amplifier was modified to generate 100 dB SPL (0.01
W/m2) at the PZT energy harvester using a standard
microphone (Buel & Kjaer Free Field 1/4 Microphone
Type 4039). In order to measure the resonance
frequency of the energy harvester, an amplifier with a
gain of 10,000 was connected to the harvester to
amplify the output voltage signals from the energy
harvester. In the case of measuring the generated
power, a resistance was connected to the energy
harvester, and the voltage between the two terminals of
the load resistance was monitored using an
oscilloscope.
resistance, R, was calculated from the measured
voltage amplitude V using the relationship P = V2/2R.
The maximum power delivered to the load was derived
when the load resistance was approximately 25-30 Ω.
The connected device generated the largest power.
Note that the powers generated by the peripheral
device and the central devices were larger than that
generated by the Kimura’s device with the same
dimension possibly owing to utilizing the polarizations
of different polarities and the reduced harvester
capacitance.
1st resonance
Fig. 5. Output voltage wave forms measured between
the load resistance terminals R connected to (a)
peripheral energy harvester, (b) central energy
harvester, (c) connected energy harvester, and (d) sum
of (a) and (b).
Fig. 4. Frequency dependences of generated voltage, E,
of PZT energy harvester with different diaphragm
diameters of 2.0 mm; (a) Peripheral energy harvester,
(b) central energy harvester, (c) energy harvester
fabricated by Kimura et al. [18]; the whole surface of
the diaphragm was covered by the Al electrode. The
frequency at each arrow corresponds to first
resonance.
Figure 5 shows comparison of the output voltage
wave forms measured between the load resistance
terminals R connected to (a) peripheral energy
harvester, (b) central energy harvester, (c) connected
energy harvester, and (d) sum of (a) and (b).
Figure 6 shows comparison of the load resistance
dependences of the PZT energy harvesters of the
peripheral device, the central device, the connected
device, and the circular device of Kimura et al. [18]
that covered the whole surface of the PZT diaphragm.
The sound pressure of approximately 100 dB at the
first-resonance frequency, 4.92 kHz was irradiated.
The effective electric power delivered to the load
Table I shows the comparison of the performances
of the PZT power generators prepared by Horowitz et
al. [16], Shinoda et al. [17], Kimura et al. [18] and in
this work. The energy harvesters of this research
utilized the both of the charges at the peripheral and
the central parts of the diaphragm. Thus, nearly double
power was generated at the same energy harvester area.
In conventional cases, the resistance at the
maximum power and the capacitive impedance of PZT
energy harvesters matches well as reported by Harigai
et al. [19]. The load resistances of the energy
harvesters of Kimura et al. [18] and this work at the
maximum power were significantly lower than those
of other researchers possibly owing to the larger
capacitance of the PZT capacitors. Harigai et al. [7]
have also demonstrated that their vibration energy
harvesters with (001)-oriented PZT films generated a
much larger power than conventional energy
harvesters with (111)-oriented PZT films. Thus,
further improvement of acoustic energy harvesters
fabricated using (001)-oriented PZT films can be
expected.
corresponds to the power density of 20.7 µW/m2. The
device exhibited nearly two times larger power than
the other devices with similar diaphragm areas.
REFERENCES
Fig. 6. Load resistance R dependence of generated
power for (a) peripheral energy harvester, (b) central
energy harvester, (c) connected energy harvester, and
(d) energy harvester fabricated by Kimura et al. [18]
whose whole surface of the diaphragm was covered by
Al electrode.
CONCLUSION
This paper presented improved power generation
performances of a PZT microelectromechanical
systems (MEMS) acoustic energy harvester having
dual top electrodes to utilize the different polarizations
of charges on the surface of a vibrating PZT
diaphragm. The PZT acoustic energy harvester had a
diaphragm with a diameter of 2 mm consisting of
Al(0.1 µm)/PZT(1 µm)/Pt(0.1 µm)/Ti(0.1 µm)/
SiO2(1.5 µm). The top Al electrodes independently
cover the peripheral surface and the central surface of
the PZT diaphragm. The peripheral energy harvester
generated a power of 5.28x10-11 W, and the central
energy harvester generated a power of 4.25x10-11 W
under the sound pressure level of 100 dB (0.01 W/m2)
at 4.92 kHz. A connected energy harvester was also
defined between the peripheral electrode and the
central electrode. The total power generated by the
connected harvesters was 8.28x10-11 W that
[1] Y. C. Shu and I. C. Lien: Smart Mater. Struct. 15 (2006)
1499.
[2] S. P. Beeby, M. J. Tudor, and N. M. White: Meas. Sci.
Technol. 17 (2007) R175.
[3] S. Priya: J. Electroceram. 19 (2007) 165.
[4] H. A. Sodano, G. Park, and D. J. Inman: Strain 40 (2004)
49.
[5] M. Ericka, D. Vasic, F. Costa, G. Poulin, and S. Tliba: J.
Phys. IV 128 (2005) 187.
[6] D. Shen, H. C. Wikle III, S.-Y. Choe, and D. J. Kim:
Appl. Phys. Express 1 (2008) 098002.
[7] T. Harigai, H. Adachi, and E. Fujii: J. Appl. Phys. 107
(2010) 096101.
[8] K. Morimoto, I. Kanno, K. Wada, and H. Kotera: Sens.
Actuators A 163 (2010) 428.
[9] P. Rakbamrung, M. Lallart, D. Guyomar, N. Muensit, C.
Thanachayanont, C. Lucat, Be. Guiffard, L. Petit, and P.
Sukwisut: Sens. Actuators A 163 (2010) 493.
[10] D. Shen, J.-H. Park, J. Ajitsaria, S.-Y. Choe, and H. C.
Wikle III: J. Micromech. Microeng. 18 (2008) 055017.
[11] V. Bedekar, J. Oliver, S. Zhang, and S. Priya: Jpn. J.
Appl. Phys. 48 (2009) 091406.
[12] H.-B. Fang, J.-Q. Liu, Z.-Y. Xu, L. Dong, D. Chen, B.C. Cai, and Y. Liu: Chin. Phys. Lett. 23 (2006) 732.
[13] W. J. Choi, Y. Jeon, J.-H. Jeong, R. Sood, and S. G.
Kim: J. Electroceram. 17 (2006) 543.
[14] M. Renaud, K. Karakaya, T. Sterken, P. Fiorini, C. Van
Hoof, and R. Puers: Sens. Actuators A 145-146 (2008)
380.
[15] J.-C. Park, J.-Y. Park, and Y.-P. Lee: J.
Microelectromech. Syst. 19 (2010) 1215.
[16] S. B. Horowitz, M. Sheplak, N. Cattafesta, and T.
Nishida: J. Micromech. Microeng. 16 (2006) S174.
[17] S. Shinoda, T. Tai, H. Itoh, T. Sugou, H. Ichioka, S.
Kimura, and Y. Nishioka: Jpn. J. Appl. Phys. 49
(2010) 04DL21.
[18] S. Kimura, S. Tomioka, S. Iizumi, K. Tsujimoto, T.
Sugou, and Y. Nishioka: Jpn. J. Appl. Phys. 50 (2011)
06GM14.
Table 1 Comparison of performances of various acoustic energy harvesters
Diaphragm
Resonance
Generated
Power density
S. B. Horowitz et
al [1]
S. Shinoda et al.
[2]
Kimura et al. [3]
This work
Diameter (mm)
Frequency (Hz)
Power (W)
(µW/m2)
Resistance at
maximum power
(ohm)
2.4
13568
6.0x10-12
3.6
5232
1.0
1000
7.0x10
-13
0.054
1000
-11
4.9
550
1.5
24020
1.1x10
2.0
18020
5.1x10-12
1.3
550
1.2
16700
1.4x10
-10
98
75
2.0
6400
4.0x10-11
10
70
2.0 Central
4920
4.2x10-11
10.5
30
-11
13.2
25
20.7
50
2.0 Peripheral
4920
5.28x10
2.0 Connected
4920
8.28x10-11