Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 168 (2016) 844 – 847

EUROSENSORS 2016

Vibration Modes of Piezoelectric Diaphragms for Ultrasonic

Microsensors and Influence of Top Electrodes
Kaoru Yamashita*, Tomoki Nishioka, Taiki Nishiumi, Minoru Noda
Kyoto Institute of Technology, Matsugasaki, Kyoto 606-8585, Japan

Abstract
Vibration modes of MEMS piezoelectric diaphragms responding to an ultrasound pulse were investigated and influence of top
electrodes to the vibration modes was discussed. The top electrode has a centrosymmetric shape divided into inner part for
ultrasonic sensing and outer part for resonant frequency tuning. Piezoelectric output signal is generated in the inner electrode by
centrosymmetric vibrations. The piezoelectric diaphragms were fabricated in flat or buckled shape and the top electrodes of various
thickness were formed on them. Pulse-induced vibration modes were evaluated by using scanning laser Doppler vibrometry. The
buckled diaphragms showed resonant peaks in a narrower frequency range compared to the flat diaphragm. The top electrode
showed an influence on the vibration modes so that vibrations with a large amplitude spontaneously concentrated in the sensing
electrode area, even in the case of a thin electrode film around 10 nm.
©c 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
 2016 The Authors. Published by Elsevier Ltd.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of the 30th Eurosensors Conference.
Peer-review under responsibility of the organizing committee of the 30th Eurosensors Conference
Keywords: resonant mode, vibration mode, diaphragm, piezoelectric, ultrasonic sensor

1. Introduction

Piezoelectric micromachined ultrasonic sensors have been developed [1–3] to utilize for three-dimensional mea-
surement in applications to under water or medical imaging, or gesture recognition using airborne ultrasound. The
authors have been developing two-dimensional array sensors on silicon-micromachined diaphragm structures with
lead-zirconate-titanate (PZT) films for three-dimensional airborne ultrasonic measurement [4,5]. Figure 1 illustrates
a structure of a single element of the array sensor. A PZT capacitor is formed on a thermally oxidized silicon (SiO2 )
diaphragm. The top electrode of the capacitor is divided into two parts of the inner electrode for ultrasonic sensing
and the outer electrode for resonant frequency tuning [6]. The array sensor is used as a phased array in the three-
dimensional measurement to detect the incident angles of the reflecting ultrasound, that is, the output signals from the

∗ Corresponding author. Tel.: +81-75-724-7446 ; fax: +81-75-724-7400.

E-mail address: yamashita.kaoru@kit.ac.jp

1877-7058 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of the 30th Eurosensors Conference
doi:10.1016/j.proeng.2016.11.287
 Kaoru Yamashita et al. / Procedia Engineering 168 (2016) 844 – 847 845

(a) (b)
Sensing (inner) top-electrode:
for signal read-out

7 μm
7 μm
Tuning (outer) top-electrode:
Ultrasound for resonant frequency tuning

m
pulse

0μ

0μ
Contact hole:

80

80
to bottom electrode 800 800
μm μm
Au: 50 nm
PZT: 1 μm (c) (d) (e) (f)
Pt/Ti: 160/20 nm
240 μm SiO2: 1 μm
Si: 300 μm
800 μm
300 μm 300 μm 300 μm 300 μm
Fig. 1. A schematic illustration of the sensor structure with
the complete double top-electrode configuration. Fig. 2. Three-dimensional profiles of (a) flat and (b) buckled diaphragms, and top
view photographs of the buckled diaphragms having top electrode with thickness
of (c) 0 nm, (d) 12.5 nm, (e) 25 nm and (f) 50 nm.

sensor elements of the array are processed with delay-and-summation technique. The summed-up signal is strongly
affected by frequency scattering of each signal [5], and the outer tuning electrode of the sensor shown in Fig. 1 re-
solves this issue through resonant frequency tuning on each sensor element. On the other hand, the third dimension
of radial distance is determined by time-of-flight of the ultrasound, and a short pulse waveform is used from the
viewpoint of distance resolution. However, the short ultrasound pulse has a broad frequency spectrum beyond the
designed fundamental resonant frequency of the sensor, and might cause higher order vibration modes on the sensor
diaphragm. Previously the authors reported that vibration modes in impulse response of piezoelectric diaphragms
without top electrode [7]. In this work, we investigated the influence of the top electrode to the vibration modes of the
piezoelectric diaphragms.

2. Experimental

2.1. Fabrication of piezoelectric diaphragms

A 2-inch silicon wafer with 1 μm-thick SiO2 was used as the starting substrate. Bottom electrode of platinum and ti-
tanium films were deposited on the substrate by using rf magnetron sputtering. Piezoelectric PZT film was then formed
through sol-gel method up to 950 nm-thick (Mitsubishi Materials, type E1, 15wt% in 1-butanol, Pb/Zr/Ti=115/52/48).
Next, the top electrode of gold film was deposited by using rf magnetron sputtering and patterned to form the double
top electrode shape by etching with I2 +KI solution. Finally the SiO2 on the backside was etched with buffered hy-
drofluoric acid to form windows, and silicon substrate was vertically etched through the windows by using ICP-RIE
(SAMCO, RIE-400iPB) with conventional Bosch process to complete the diaphragm structures.
Buckled diaphragms were fabricated for a comparison to flat diaphragms because upward bucked diaphragms
yield higher sensitivity than flat or downward ones [8]. The buckling behavior was controlled by residual stress of
PZT through sol-gel preparation condition [9]. Thickness of the top electrode was controlled simply by the sputtering
time. A normal sensor has the top electrode with 50 nm-thick gold film deposited for 2 minutes at 40 W. We also
prepared top electrodes with the deposition time of 1 minute and 30 s. These thickness are referred to as 25 nm
and 12.5 nm for convenience, although the thickness of these thin films were not precisely confirmed. Figure 2
shows three-dimensional profiles of the flat and the buckled diaphragms, and top view photographs of the buckled
diaphragms having the top electrodes with the various thickness.

2.2. Evaluation of pulse induced vibration modes

An ultrasound pulse was generated by using a spark discharge sound source [7]. A pair of needles is set to face
each other and parallelly connected to a capacitor which is being charged from a battery. When air between the
846 Kaoru Yamashita et al. / Procedia Engineering 168 (2016) 844 – 847

needles electrically breaks down with a spark discharge, a short ultrasound pulse is generated. The pulse has typically
a duration around 25 μs and a broad spectrum centered at 80 kHz in the range from audible frequencies to 150 kHz,
which were confirmed with a calibrated microphone (Brüel & Kjær, Type 4138). Vibration modes of the diaphragms
responding to the pulse were measured by using a scanning laser Doppler vibrometer (Polytec, PVS-400) [7].

(i) (ii) (iii)

(a)
contour map velocity map

Displacement [rel. unit] Displacement [rel. unit]

56.7 kHz 142 kHz 204 kHz
Theoretical Measured

Flat diaphragm
no electrode

Intensity [rel. unit]

1

56.7 kHz

142 kHz

204 kHz
0

1 ms
0 –1
(b) 1
velocity map contour map velocity map

65.3 kHz 92.2 kHz 103 kHz 92.2 kHz Buckled diaphragm
Theoretical Measured

no electrode
Intensity [rel. unit]

1
65.3 kHz

103 kHz

1 ms
0 –1
(c) 63.1 kHz 133 kHz 222 kHz 1

Displacement [rel. unit] Displacement [rel. unit] Displacement [rel. unit]

Measured

Buckled diaphragm
electrode 12.5 nm
Intensity [rel. unit]

1
133 kHz

0
63.1 kHz

222 kHz

(No corresponding theoretical modes)

1 ms
0 –1
(d) 64.4 kHz 115 kHz 203 kHz 1
velocity map

Buckled diaphragm
Measured

electrode 25 nm
Intensity [rel. unit]

1
115 kHz

0
203 kHz
64.4 kHz

(No corresponding theoretical modes)

1 ms
0 –1
(e) 1
68.3 kHz 108 kHz 196 kHz
velocity map

Buckled diaphragm
Measured

electrode 50 nm
Intensity [rel. unit]

1
108 kHz

0
196 kHz
68.3 kHz

(No corresponding theoretical modes)

1 ms
0 –1
0 100 200 300
Frequency [kHz]

Fig. 3. Evaluation results of the pulse response of the diaphragms. (a) flat diaphragm without top electrode, and buckled diaphragms having
(b) no top electrode and (c) 12.5 nm-thick, (d) 25 nm-thick and (e) 50 nm-thick top electrode. (i) major vibration modes of measured results and
corresponding theoretical modes derived from a flat, no-stressed plate without an electrode (if available), (ii) frequency spectra at the center of the
diaphragm, and (iii) time course vibration waveforms at the center of the diaphragm.
 Kaoru Yamashita et al. / Procedia Engineering 168 (2016) 844 – 847 847

3. Results and Discussion

Typical examples of major vibration modes on the whole diaphragm, and frequency spectra and vibration wave-
forms at the center of diaphragms are shown in Fig. 3. Diaphragms without top electrode showed three major cen-
trosymmetric vibration modes corresponding to theoretical ones, but resonant peaks of the buckled diaphragm dis-
tributed in a narrower frequency range than that of the flat one. In both cases, all these modes have a large amplitude
and the vibration waveforms showed distorted damping oscillation due to interference of the vibrations with different
frequencies. On the other hand, the diaphragms with top electrode showed no corresponding theoretical vibration
modes, and one of the centrosymmetric mode has a dominant amplitude. Since the side peaks have much smaller
amplitude, the damping oscillation of the vibration waveform showed almost no distortion.
Vibration on the diaphragms with the top electrode was concentrated into the inner sensing electrode region in
all electrode thickness of 50, 25 and 12.5 nm, and each vibration mode showed much more similar patterns among
them compared to the difference of those between the diaphragms with and without the electrode. The gold film
with the thickness 12.5 nm of the top electrode has a mass component of only 2.5% to the whole diaphragm, which
might have no enough mechanical influence to the vibrations. On the other hand, the electrode can have a large
electrical influence by homogenizing the electric field in the piezoelectric layer under the electrode, even with a very
thin gold film. Under the homogenized electric field, piezoelectric polarization can not distribute according to only
the mechanical vibration, and converse piezoelectric stress would be generated from the difference between the free
mechanical polarization distribution and the restricted distribution under the homogeneous field. This might cause the
concentration of the vibration into the central electrode region.

4. Conclusions

Vibration modes of micromachined piezoelectric diaphragms responding to an ultrasound pulse were investigated.
The piezoelectric diaphragms were fabricated in flat or buckled shape and the top electrodes of various thickness were
formed on them. The buckled diaphragms showed resonant peaks in a concentrated narrow frequency range compared
to the flat diaphragm. The top electrode showed influence on the centrosymmetric vibration modes to spontaneously
concentrate on the sensing electrode, which might be caused through homogenized electric field by the electrode.

Acknowledgements

This work was supported in part by Grant-in-Aid for Scientific Research (C) 15k06016 from Japan Society of the
Promotion of Science and by Research Grant from Nippon Sheet Glass for Materials Science and Engineering.

References
[1] J. T. Bernstein, S. L. Finberg, K. Houston, L. C. Niles, H. D. Chen, L. E. Cross, K. K. Lee, K. Udayakumar, Micromachined high frequency
ferroelectric sonar transducers, IEEE Trans. Utrrason. Ferro. Freq. Cont., 44 (1997) 960–969.
[2] D. Akai, T. Yogi, I. Kamja, Y. Numata, K. Ozaki, K. Sawada, N. Okada, K. Higuchi, M. Ishida, Miniature ultrasound acoustic imaging devices
using 2-D pMUTs array on epitaxial PZT/SrRuO3 /Pt/γ-Al2 O3 /Si structure, Transducers 2011, Beijing, China, June 5–9 (2011) 910–913.
[3] R. J. Przybyla, H.-Y. Tang, S. E. Shelton, D. A. Horsley, B. E. Boser, 3D ultrasonic gesture recognition, IEEE ISSCC 2014, San Francisco,
USA, Feb. 9–13 (2014) 210–211.
[4] K. Yamashita, H. Katata, M. Okuyama, H. Miyoshi, G. Kato, S. Aoyagi, Y. Suzuki, Arrayed ultrasonic microsensors with high directivity for
in-air use using PZT thin film on silicon diaphragms, Sens. Actuat. A, 97-98 (2002) 302–307.
[5] K. Yamashita, L. Chansomphou, H. Murakami, M. Okuyama, Ultrasonic micro array sensors using piezoelectric thin films and resonant
frequency tuning, Sens. Actuat. A, 114 (2004) 147–153.
[6] K. Yamashita, K. Tomiyama, K. Yoshikawa, M. Noda, M. Okuyama, Resonant frequency tuning of piezoelectric ultrasonic microsensors by
bias voltage application to extra top-electrodes on PZT diaphragms, Ferroelectrics, 408 (2010) 48–54.
[7] T. Nishiumi, K. Yamashita, H. Tanaka, M. Noda, Vibration modes in impulse response of piezoelectric diaphragms for ultrasonic microsensors,
IEEE-NEMS 2016, Matsushima, Japan, Apr. 17–20 (2016) B2L-A-2.
[8] K. Yamashita, H. Nishimoto, M. Okuyama, Diaphragm deflection control of piezoelectric ultrasonic microsensors for sensitivity improvement,
Sens. Actuat. A, 139 (2007) 118-123.
[9] K. Yamashita, K. Arai, H. Tanaka, T. Nishiumi, M. Noda, Buckling behavior of piezoelectric diaphragms for highly sensitive structures of
ultrasonic microsensors controlled through intrinsic stress of PZT films, ISAF 2015, Singapore, May 24-27 (2015) 44–47.