Home Search Collections Journals About Contact us My IOPscience

Ferroelectric and Piezoelectric Properties of (Bi1/2K1/2)TiO3 Ceramics

This content has been downloaded from IOPscience. Please scroll down to see the full text.

2005 Jpn. J. Appl. Phys. 44 5040

(http://iopscience.iop.org/1347-4065/44/7R/5040)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 129.130.252.222
This content was downloaded on 15/07/2014 at 14:39

Please note that terms and conditions apply.

Japanese Journal of Applied Physics
Vol. 44, No. 7A, 2005, pp. 5040–5044
#2005 The Japan Society of Applied Physics

Ferroelectric and Piezoelectric Properties of (Bi1=2 K1=2 )TiO3 Ceramics

Yuji H IRUMA, Rintaro A OYAGI, Hajime N AGATA and Tadashi TAKENAKA
Faculty of Science and Technology, Tokyo University of Science, 2641 Yamazaki, Noda, Chiba 278-8510, Japan
(Received February 4, 2005; accepted March 23, 2005; published July 8, 2005)

The dielectric, ferroelectric and piezoelectric properties of bismuth potassium titanate, (Bi1=2 K1=2 )TiO3 (BKT), ceramics
were studied. Single-phase BKT ceramics with a high relative density of 97% were obtained by the hot pressing (HP) method.
The resistivities of BKT ceramics hot-pressed at 1060 and 1080
C (hereafter abbreviated as BKT-HP1060
C and
BKT-HP1080
C) were fairly high being of the order of 1013
cm at room temperature (RT). The Curie temperature Tc of
BKT-HP1060
C was 437
C, which is relatively higher than those of other lead-free piezoelectric materials. In this study, the
ferroelectric properties of BKT ceramics were successfully obtained with fully saturated hysteresis loops. The remanent
polarization Pr and coercive field Ec of BKT-HP1080
C were 22.2 mC/cm2 and 52.5 kV/cm, respectively. D–E hysteresis
loops for these ceramics were observed even at 260
C. The electromechanical coupling factor k33 and piezoelectric constant
d33 of BKT-HP1080
C were 0.28 and 69.8 pC/N, respectively. The second-phase transition temperature T2 of 340
C was
determined from the temperature dependence of piezoelectric and dielectric measurements.
[DOI: 10.1143/JJAP.44.5040]
KEYWORDS: lead-free piezoelectric ceramics, (Bi1=2 K1=2 )TiO3 , dielectric properties, ferroelectric properties, D–E hysteresis
loop, Curie temperature

1. Introduction
Environmental destruction has been a serious problem BKT
worldwide. Such a problem includes the release of harmful
materials (e.g., Pb) from electrical industries, and some
regulations to prevent the release of harmful waste materials
such as WEEE, RoHS and ELV have already been
implemented. Piezoelectric ceramics are widely used in
various applications such as ceramic filters, actuators and
sensors. However, PbZrO3 –PbTiO3 (PZT)-based three-com-
ponent systems that include with a large amount of Pb have
been mainly used in piezoelectric applications because of
their excellent properties.1,2) Therefore, lead-free piezoelec-
tric materials have recently been extensively studied. In
lead-free piezoelectric materials, single crystals of the
langasite series and ferroelectric ceramics with a tungsten
Fig. 1. Temperature dependence of lattice constants a and c of BKT
bronze structure and a bismuth layer structure are piezo-
ceramics reported by Ivanova et al.
electrics whose frequency constant is stably temperature-
dependent. These materials have a good thermal stability,
but small electromechanical coupling factors. For this piezoelectrics with a high Tc ; nevertheless, the same reason
reason, various lead-free piezoelectric solid solutions with for BNT, it is necessary to clarify the electrical behavior at
a perovskite structure, such as BaTiO3 , (Bi1=2 Na1=2 )TiO3 T2 . Although the electrical properties of some BKT-based
and KNbO3 , have been actively studied.3–7) ceramics have been reported,10–12) few studies on BKT
Bismuth potassium titanate, (Bi1=2 K1=2 )TiO3 (BKT), is a ceramics and their electrical properties have been reported
typical ferroelectric material with a perovskite structure of because it is difficult to prepare a dense ceramic body by a
tetragonal symmetry at room temperature and a relatively conventional ceramic fabrication process. Recently, to
high Curie temperature, Tc , of 380
C.8) Ivanova et al. improve the sinterability of BKT-based ceramics, some
investigated the crystal structure of BKT ceramics by X-ray chemical processes have been studied;13,14) however, there is
diffraction analysis, and they reported that the lattice no data on ferroelectric properties. For this reason, pure
constants of BKT ceramics are a ¼ 0:3913 nm, c ¼ 0:3993 BKT ceramics were prepared using the hot pressing (HP)
nm (Fig. 1).9) They also reported that the second-phase method in this study, and their dielectric, ferroelectric and
transition temperature T2 , that is, from tetragonal to piezoelectric properties were investigated.
pseudocubic, existed at about 270
C. However, the electrical
behavior at T2 has not been investigated. The Tc of 2. Experimental Procedure
(Bi1=2 Na1=2 )TiO3 , which belongs to the same series as BKT powder was prepared using a conventional ceramic
bismuth alkali titanate, is 320
C; however, BNT has a T2 of fabrication process. Reagent-grade metal oxide or carbonate
about 200
C, the same as that of BKT. That is, the phase powders of Bi2 O3 (99.99%), TiO2 (99.99%) and K2 CO3
boundary between the rhombohedral ferroelectric and anti- (99.9%) were used as starting materials. These oxide or
ferroelectric phase fields, and piezoelectricity disappear at carbonate powders were mixed for 10 h in ethanol with 2-
T2 . BKT ceramics seem to be a candidate for lead-free mm-diameter and 10-mm-diameter zirconia balls by ball
5040
Jpn. J. Appl. Phys., Vol. 44, No. 7A (2005)
milling, followed by calcination at 800
C for 4 h and 950
C
for 2 h in an alumina crucible. After calcination, the ground
powder was ball-milled for 20 h and then pressed into discs
of 20 mm diameter and 5 mm thickness. The bulks formed
were sintered using an ordinary firing (OF) method and the
HP method.
The HP method was carried out at 1060–1080
C for 3 h. A
pressure of 150 kg/cm2 was applied for 2 h at the maximum
sintering temperature. In the HP method, it was difficult to
confirm actual sample temperatures during sintering; how-
ever, it was thought to be 20 to 30
C lower than the observed
temperature. During hot pressing, bulk samples were tightly
covered with alumina powders under poor-oxygen condi-
tions. After hot pressing, the resulting samples were
annealed at 900
C in air for reoxidization and reduction in
inner stress. Compared with hot pressing, the OF method
was carried out at 1000–1070
C for 2 h in an alumina
crucible in air.
The observed densities of the obtained ceramics were
measured using the Archimedes method. These ceramics
were cut and polished for various physical and electrical
measurements. The ceramics were polished to the same
thickness of 0.5 mm. Fired-on silver paste was used as the
electrode for the determination of the dielectric and ferro-
electric properties of the ceramics. The crystal structures and
lattice constants of the sintered ceramics were determined by
X-ray diffraction analysis with Cu K
radiation through a Ni
filter using an X-ray diffractometer (Rigaku; RINT2000).
The microstructures of the obtained ceramics were observed
by scanning electron microscopy (SEM: Hitachi S-2400).
For these ceramics, the temperature dependences of the
dielectric constants "s and loss tangent tan  were measured
to determinate Curie temperature Tc at 10 kHz, 100 kHz and
1 MHz using an automated dielectric measurement system
with a multifrequency LCR meter (YHP 4275A). D–E
hysteresis loops at 50 Hz of the prepared specimens were
observed using a ferroelectric testing system (Toyo Corpo-
ration; Model 6252 Rev. C) in the temperature range from
room temperature (RT) to 260
C. The electrical strains at
room temperature were determined using a contact-type
displacement sensor (Millitron; Model 1240) at 0.1 Hz.
Specimens for piezoelectric measurements were poled in a
stirred silicone bath by applying a dc electrical field of 7–
8 kV/mm for 10 min. The piezoelectric properties were
measured by a resonance-antiresonance method based on
EMAS-6002 standards using an impedance analyzer (YHP
4294A). The electromechanical coupling factors of the
longitudinal extensional mode k33 were measured using a
rectangular specimen of 2  2  5 mm3 .
3. Results and Discussion
As a primary study, the OF method has been carried out in
the temperature range of 1000–1070
C to confirm the
sinterability and to identify the crystal structure of BKT
ceramics. The X-ray diffraction patterns of the ordinary fired
BKT ceramics are shown in Fig. 2. The single-phase
perovskite structures of tetragonal symmetry were observed
at a sintering temperature lower than 1030
C. However, it is
confirmed that the number of impurity peaks such as that of
K2 Ti6 O13 gradually increased in the sintering temperature
range from 1040 to 1060
C, and the sample was then melted
Y. HIRUMA et al. 5041
K2Ti6O13
101
111
100
200
211
202
(e)
(d)
(c)
(b)
(a)
10 20 30 40 50 60 70
2θ ( ° )
Fig. 2. X-ray diffraction patterns of BKT ceramics ordinary fired at
1030
C (a), 1050
C (b) and 1060
C (c), and those of BKT-HP1060
C (d)
and BKT-HP1080
C (e).
at 1070
C. In the OF method, a dense body, 93% in
theoretical density, was obtained at a sintering temperature
of 1060
C; however, the resistivity was low being of the
order of 1011
cm. It is thought that this low resistivity was
caused by impurity phases. From these results, it is fairly
difficult to fabricate dense BKT ceramics with single-phase
perovskite structures by the OF method without using a
chemical process,13,14) the HP method and the hot isostatic
pressing (HIP) method, because impurity peaks appear at
temperatures lower than the appropriate sintering temper-
ature, at which dense ceramic bodies have been obtained.
For fabricating BKT ceramics, the HP method is superior
to the OF method, because it is easy to fabricate dense
bodies at low sintering temperatures. The X-ray diffraction
patterns of the BKT ceramics sintered by the HP method are
shown in Fig. 2. The samples were hot-pressed at 1060
C
and 1080
C (hereafter abbreviated as BKT-HP1060
C and
BKT-HP1080
C). The theoretical densities of the obtained
ceramics were all higher than 97%. From Fig. 2, it can be
noted that impurity peaks were suppressed by the HP
method, and single-phase perovskite structures were ob-
tained at a sintering temperature of 1060
C. The lattice
parameters of BKT-HP1060
C were a ¼ 0:3925 nm and
c ¼ 0:3993 nm, which agree well with (a ¼ 0:3913 nm,
c ¼ 0:3993 nm) reported values.9) The resistivities of both
samples were fairly high of the order of 1013
cm at RT.
The temperature dependences of dielectric constants, "s ,
and dielectric loss tangents, tan , for BKT-HP1060
C and
BKT-HP1080
C are shown in Fig. 3. The dielectric proper-
ties were measured in the temperature range from room
temperature to 600
C, and at frequencies of 10 kHz, 100 kHz
and 1 MHz. The Tc ’s at 10 kHz, 100 kHz, and 1 MHz were
415, 425 and 437
C for BKT-HP1060
C, and 387, 397 and
410
C for BKT-HP1080
C, respectively. This frequency
dispersion of Tc indicates that BKT ceramics are typical
relaxor-structure ferroelectrics. The Tc of BKT-HP1080
C
was lower than that of BKT-HP1060
C. It is considered that
BKT-HP1080
C is slightly different from stoichiometric
BKT. The "s ’s and tan ’s at RT of BKT-HP1060
C and
5042 Jpn. J. Appl. Phys., Vol. 44, No. 7A (2005) Y. HIRUMA et al.

4000 20 40 40
10kHz (a) (b)
(a) 100kHz 10kHz 20 20
3000 15

Polarization, P (µC/cm2)
1MHz
-100 100 -100 100

Loss tangent, tanδ ( % )

2000 10 -20
-20
340°C
Dielectric constant, εs

1000 100kHz 5 -40 -40

40 40
1MHz (c) (d)
4000 20 20 20
10kHz 10kHz
3000
(b) 100kHz 15 -100 100 -100 100
1MHz
-20 -20
2000 317°C 10
-40 -40
100kHz 5 Applied field, Ea (kV/cm)
1000
1MHz Fig. 5. D–E hysteresis loops of BKT ceramics hot-pressed at 1080
C and
0 measured at 100
C (a), 200
C (b), 240
C (c) and 260
C (d).
0 100 200 300 400 500 600
Temperature ( °C )
Fig. 3. Temperature dependences of dielectric constant "s and dielectric 0.12
loss tangent tan  for hot-pressed BKT ceramics sintered at 1060
C (a) BKT-HP1080°C
and 1080
C (b). 0.10
Strain (%)

0.08

40 0.06
1080 °C BKT-HP1060°C
Polarization, P ( µC/cm 2)

0.04
1060 °C 20
0.02

-150 -100 -50
-20
-40
Applied field, Ea ( kV/cm )
Fig. 4. D–E hysteresis loops of BKT ceramics hot-pressed at 1060
C (a)
and 1080
C (b).
BKT-HP1080
C are 524 and 6.8%, and 517 and 7.1% at
1 MHz, respectively. In Fig. 3, the tan  curves show two
peaks. High-temperature peaks have a frequency dispersion;
therefore, it is thought that these peaks are related to Tc .
Low-temperature peaks are almost independent of frequen-
cies; therefore, it is considered to indicate another phase
transition, and the temperatures of BKT-HP1060
C and
BKT-HP1080
C are about 340 and 317
C, respectively.
The D–E hysteresis loops for BKT-HP1060
C and BKT-
HP1080
C are shown in Figs. 4(a) and 4(b). Well-saturated
D–E hysteresis loops were obtained for BKT-HP1080
C.
However, the D–E hysteresis loop for BKT-HP1060
C did
not completely saturate at RT. The remanent polarizations
Pr and coercive fields Ec values were 22.2 mC/cm2 and
52.5 kV/cm for BKT-HP1080
C, and 14.2 mC/cm2 and 47.3
0.00
0 20 40 60 80
50 100 150 Applied field, Ea (kV/cm)
Fig. 6. Longitudinal strain behavior under unipolar driving of electrical
field from 0 to 90 kV/cm.
@ 50 Hz
kV/cm for BKT-HP1060
C, respectively. BKT-HP1080
C
includes some impurity peaks; however, it has a higher Pr
than BKT-HP1060
C, which has a single-phase perovskite
structure. The reason for this is thought to be the difference
in grain size between these samples. The temperature
dependence of D–E hysteresis loops for BKT-HP1080
C
was observed in the temperature range from room temper-
ature to 260
C as shown in Fig. 5. Pr and Ec gradually
decreased with increasing sample temperature, and a well-
saturated D–E hysteresis loop was observed even at 260
C.
Figure 6 shows the longitudinal strain behavior under
unipolar driving of the electrical field from 0 to 90 kV/cm.
Unipolar driving is critical for the actuator drive. The
longitudinal strain of BKT-HP1080
C is three times higher
than that (about 0.12%) of BKT-HP1060
C. The d33 of
BKT-HP1060
C and BKT-HP1080
C determined from
these results were 53.7 and 135 pm/V, respectively.
The frequency dependence of the impedance Z of the
longitudinal (33) mode for BKT-HP1060
C and BKT-
HP1080
C was measured. Rectangular specimens of 2 
2  5 mm3 were used for measuring the longitudinal (33)
Jpn. J. Appl. Phys., Vol. 44, No. 7A (2005)
6
10
6 49°
Log impedance, | Z| [Ω]
2 30
5
10
6
4 -30
2 -60
4
10
420 440
Frequency, f [kHz]
Fig. 7. Frequency dependences of impedance jZj and phase  of BKT
ceramics hot-pressed at 1080
C.
mode. Poling treatment was carried out in the temperature
range from room temperature to 200
C and in the electrical
field range from 7 to 8 kV/mm. Both specimens were
difficult to pole completely because of a high coercive field,
and the optimum piezoelectric property of BKT-HP1080
C
was obtaind when the poling temperature was 150
C.
In particular, it was difficult to pole BKT-HP1060
C
completely as you could guess from the result of the D–E
hysteresis loops shown by Fig. 4. The frequency depend-
ences of the impedance jZj and phase  for BKT-HP1080
C
are shown in Fig. 7. The electromechanical coupling factor
k33 and piezoelectric constant d33 of BKT-HP1080
C were
0.28 and 69.8 pC/N, respectively. The maximum phase was
49
; this means that this ceramic was not poled completely.
To improve piezoelectricity, it is necessary to determine the
appropriate poling conditions and increase grain size. The
temperature dependences of k33 and  were measured in the
temperature range from room temperature to 500
C as
shown in Fig. 8. The maximum k33 and  were observed at
150
C, which is coincident with the poling temperature at
which the optimum piezoelectric property was obtained.
From this results, piezoelectricity disappeared at 315
C,
which is different from the second-phase transition temper-
ature T2 (270
C) reported by Ivanova et al.9) In Fig. 3(b), the
peaks of tan , which are independent of measurement
frequency, are observed at 317
C, which well agrees with
0.35
0.30
Coupling factor, k 33
0.25
150°C 315°C
0.20
0.15
0.10
0.05
0.00
0 100 200 300
Temperature (°C)
Fig. 8. Temperature behavior of electromechanical coupling factor k33
and phase  of BKT-HP1080
C.
Y. HIRUMA et al. 5043
90
(a) (b)
60
Phase, θ [ ° ]
0
-90 (c) (d)
460 480
2 µm
Fig. 9. Microstructures of BKT-HP1060
C (a), BKT-HP1080
C (b),
BKT-HP1060
C annealed at 1040
C for 8 h (c) and BKT-HP1060
C
annealed at 1040
C for 20 h (d).
temperature piezoelectricity disappearance. These results
show that the T2 of BKT-HP1080
C ceramics from
tetragonal to pseudocubic was about 315
C. In this study,
the T2 of BKT ceramics was determined to be about 340
C,
which is decided from the temperature dependence of tan 
in BKT-HP1060
C.
Figure 9 shows the microstructures of the hot-pressed
BKT ceramics determined by SEM. The grain size of BKT-
HP1080
C was about twofold larger than that (about 0.2 mm)
of BKT-HP1060
C. The improvement of electrical proper-
ties such as ferroelectricity and piezoelectric strain in BKT-
HP1080
C was the result of grain growth. BKT-HP1060
C
was annealed at 1040
C for 8 h and 20 h in air for grain
growth. Figures 9(c) and 9(d) show the microstructures of
BKT-HP1060
C annealed at 1040
C for 8 and 20 h,
respectively. The grain size of BKT-HP1060
C annealed
at 1040
C for 8 and 20 h were approximately 0.6 and 1 mm.
These microstructures indicate that the electrical properties
of BKT ceramics can be improved.
60 4. Conclusions
40 Single-phase BKT ceramics with a high theoretical
density ratio of 97% have been fabricated using the HP
Phase, θ ( ° )
20
method. The resistivities of BKT-HP1060
C and BKT-
0
HP1080
C were fairly high of the order of 1013
cm at RT.
-20 The Curie temperature Tc values at 1 MHz were 437
C for
-40 BKT-HP1060
C, and 410
C for BKT-HP1080
C. The
ferroelectric properties of BKT ceramics were successfully
-60
obtained with fully saturated D–E hysterisis loops. The
-80 remanent polarization Pr and coercive field Ec of BKT-
HP1080
C were 22.2 mC/cm2 and 52.5 kV/cm, respectively.
400 500
The electromechanical coupling factor k33 and the piezo-
electric constant d33 of BKT-HP1080
C were 0.28 and
69.8 pC/N, respectively. In this study, the second-phase
transition temperature T2 was determined from the measure-
5044 Jpn. J. Appl. Phys., Vol. 44, No. 7A (2005)
ment of temperature dependences of k33 and , and the peak
of tan  which is independent of measurement frequencies;
the T2 of pure BKT ceramics was about 340
C.
Acknowledgements
The authors would like to thank Toho Titanium Co., Ltd.
for providing titanium oxide powder with high purity. This
work was partially supported by a Grant-in-Aid for Scientific
Reserch (B) (No. 15360352) from the Japan Society for the
Promotion of Science.
1) E. Sawaguchi: J. Phys. Soc. Jpn. 8 (1953) 615.
2) T. Yamamoto: Jpn. J. Appl. Phys. 35 (1996) 5104.
3) H. Nagata, N. Koizumi, N. Kuroda, I. Iigarashi and T. Takenaka:
Ferroelectrics 229 (1999) 237.
4) H. Nagata and T. Takenaka: J. Eur. Ceram. Soc. 21 (2001) 1299.
Y. HIRUMA et al.
5) H. Ishii, H. Nagata and T. Takenaka: Jpn. J. Appl. Phys. 40 (2001)
5660.
6) T. Takenaka, K. Maruyama and K. Sakata: Jpn. J. Appl. Phys. 30
(1991) 2236.
7) S. Tashiro, H. Nagamatsu and K. Nagata: Jpn. J. Appl. Phys. 41 (2002)
7113.
8) C. F. Buhrer: J. Chem. Soc. 36 (1962) 798.
9) V. V. Ivanova, A. G. Kapyshev, Y. N. Venevtsev and G. S. Zhdanov:
Izv. Akad. Nauk SSSR 26 (1962) 354.
10) A. Sasaki, T. Chiba, Y. Mamiya and E. Otsuki: Jpn. J. Appl. Phys. 38
(1999) 5564.
11) Y. Hiruma, H. Nagata and T. Takenaka: J. Ceram. Soc. Jpn. 112
(2004) 1125.
12) H. Nagata, M. Yoshida, Y. Makiuchi and T. Takenaka: Jpn. J. Appl.
Phys. 42 (2003) 7401.
13) T. Wada, A. Fukui and Y. Matsuo: Jpn. J. Appl. Phys. 41 (2002) 7025.
14) Z. F. Li, C. L. Wang, W. L. Zhong, J. C. Li and M. L. Zhao: J. Appl.
Phys. 94 (2003) 2548.