You are on page 1of 12

Ferroelectrics, 319:4556, 2005

Copyright Taylor & Francis Inc.


ISSN: 0015-0193 print / 1563-5112 online
DOI: 10.1080/00150190590965370
Electric Field Applicability Limits for PZT Ceramics
T. MALYSH

AND J. ERHART
Department of Physics and International Center for Piezoelectric Research,
Technical University of Liberec, H alkova 6, CZ-461 17 Liberec 1,
Czech Republic
Electric eld applicability limits were measured by the resonant methods for D.C. and
pulse electric eld depoled PZT ceramics. It is found that the limiting elds for soft PZT
are in the range of 500700 V/mm for D.C. depoling and in the range of 200500 V/mm
for pulses. Hard PZT ceramics is resistant to the depoling elds up to 1000 V/mm either
for D.C., or pulse depoling elds. Relative changes in the magnitude of the electrome-
chanical properties are 320% for the electric elds up to 1000 V/mm. Some electrome-
chanical tensor components appear more pronounced sensitivity to the external elds
than others.
Keywords PZT ceramics; poling
PACS: 77.84.Dy; 77.22.Ej
1. Introduction
Nowadays, poled PZT ceramics serve as a basis for the production of many components in
electromechanical devices. In recent years many authors were engaged in investigations of
conditions under which changes of electromechanical properties of PZT ceramics would be
possible. PZTceramics is commercially poled by the strong D.C. voltage at elevated temper-
atures, but the poling conditions are not usually provided to the customers. In applications,
alignment of the dipole moments in ferroelectric ceramics is inuenced by the electric eld
in the direction opposite to the poling eld. If it is strong enough, such electric eld could
completely destroy the polarization orientation and therefore also the piezoelectric proper-
ties. While the poling (as well as the following application of electric eld, i.e. so called
depoling) process is basically the dynamic process, different poling (and depoling) results
are supposed from the D.C. or pulse-poling voltage. Poling-dependent material properties
were recently studied in soft and hard lead zirconate titanate (PZT) [1, 2], lead titanate
(PbTiO
3
) [3] and barium titanate (BaTiO
3
) [4] ceramics for the electric elds exceeding the
coercive eld.
Main aim of this contribution is the measurement of the electromechanical properties
for several types of PZTceramics (hard as well as soft) as a function of the depoling voltage.
Electric eld limits for the application of PZT ceramics without the substantial change of
the material properties will be specied.
Received September 6, 2004; In revised form February 28, 2005.

Corresponding author. E-mail: tamal@centrum.cz


[271]/45
46/[272] T. Malysh and J. Erhart
Table 1
Material Properties of Poled PZT Ceramics Before Depoling
d
33
d
31
k
31
k
p
k
t

T
33
s
E
11
c
E
33

E
PZT type [pC/N] [pC/N] (%) (%) (%) [
0
] [10
12
Pa
1
] [10
11
Pa] [1]
APC 850
APC 856
480
609
186
196
35
32
60
62
48
47
2150
2870
14.8
14.4
1.16
0.99
0.32
0.45
APC 840
APC 841
APC 880
274
301
299
97
114
92
29
32
29
55
59
54
46
46
46
1140
1284
1044
11.2
11.0
11.1
1.25
1.31
1.24
0.44
0.40
0.44
Figure 1. Piezoelectric strain constant d
33
after depoling: (a) by D.C. and (b) after voltage impulses
applied for bar samples.
Electric Field Applicability Limits [273]/47
2. Experiment
Samples of PZTceramics have been supplied by the American Piezo Ceramics International
Ltd., Mackeyville, PA, USA. Samples include soft (APC 850 and 856) as well as hard
PZT (APC 840, 841 and 880) ceramics. Due to the resonant method used for the material
properties measurement [5], sample shapes were bar and disc for each PZTtype. Dimensions
of bar-shaped samples werelength 15 mm, width 4 mm and thickness 1 mm. Discs were
15 mm in diameter and 1 mm in thickness. Main faces of the discs and bars were electroded
by the red silver paste.
Figure 2. Piezoelectric strain constant d
31
after depoling: (a) by D.C. and (b) after voltage impulses
applied for bar samples.
48/[274] T. Malysh and J. Erhart
Virgin PZT samples were poled at elevated temperature in strong D.C. electric eld.
According to our previous tests, the poling temperatures have been selected high enough
to ensure the saturated material properties with moderate aging after the poling. Soft PZT
ceramics has been poled at 90

C and hard PZT at 130

C, both types by the D.C. electric


eld 2 kV/mm for 5 min in silicon oil bath. Due to the aging in the material coefcients,
the measurements were done one day after the poling. Then the samples were aged for one
decade (i.e. for the next ten days) and material properties were measured again. At the same
Figure 3. Electromechanical coupling factor k
31
after depoling: (a) by D.C. and (b) after voltage
impulses applied for bar samples. Data missing in curves for APC850 and APC856 are due to the
non-resonant response of samples in certain depoling eld range.
Electric Field Applicability Limits [275]/49
time, the electric eld depoling of the samples was performed. PZTaging and changes in the
material properties due to the depoling process would be separated in such way. Depoling
has been done by the electric elds of 250 V/mm, 500 V/mm, 750 V/mm and 1000 V/mm
of the opposite direction as poling eld at room temperature. Depoling eld was applied
either as D.C. eld, or in the form of unipolar rectangular pulses (pulse shape: zero eld for
1 s, applied D.C. eld for 1 s etc.) for the same total time as poling eld (i.e. for 5 minutes,
which means approximately 100 pulses and PZT samples were subjected to the non-zero
electric eld for 2.5 minutes).
Figure 4. Electromechanical coupling factor k
p
after depoling: (a) by D.C. and (b) after voltage
impulses applied for disc samples. Data missing in curves for APC850 and APC856 are due to the
non-resonant response of samples in certain depoling eld range.
50/[276] T. Malysh and J. Erhart
Material property measurement has been done by the resonant methods according to
the IRE Standard 1961 [5]. Length-extensional vibrations of the bar (measured resonant f
r
,
antiresonant f
a
frequency and sample capacitance C
0
at 1 kHz frequency) resulted in the
values of material properties s
E
11
, k
31
, d
31
,
T
33
. Radial disc vibrations (measured resonant
f
r
frequencies for fundamental and the third overtone mode, antiresonant f
a
frequency of
the fundamental mode and sample capacitance C
0
at 1 kHz frequency) give us the values
of s
E
12
, k
p
,
T
33
and Poissons ratio
E
. Thickness-extensional disc vibrations could be used
for c
E
33
, k
t
measurements (measured resonant frequencies of the fundamental f
1
and third
overtone mode f
3
). Measurements of the impedance spectra have been performed on the
Figure 5. Electromechanical coupling factor k
t
after depoling: (a) by D.C. and (b) after voltage
impulses applied for disc samples.
Electric Field Applicability Limits [277]/51
impedance analyzer HP4192A. Moreover, d
33
has been directly measured on d
33
-meter
(ZJ-3C type). Material properties of the poled PZT ceramics before depoling are listed in
Table 1.
3. Results and Discussion
D.C. Voltage Depoling
Signicant changes in the material coefcients occur in soft PZT ceramics for the values of
s
E
12
,
E
, k
31
, k
p
, d
31
, d
33
,
T
33
. At certain electric eld value, the piezoelectric coefcient d
33
Figure 6. Elastic compliance s
E
11
after depoling: (a) by D.C. and (b) after voltage impulses applied
for bar samples.
52/[278] T. Malysh and J. Erhart
changes its sign (from + to ) for soft PZT. In such case, the sign of d
33
could be directly
measured by the d
33
-meter. The change in sign means that reverse poling occurred. Similar
change in the sign of d
31
coefcient cannot be found by the resonant methods, but it is also
supposed when d
33
changes its sign. Depoling by the electric eld of this magnitude would
depolarize soft PZTceramics more or less; however no virgin state will be reproduced again.
Generally, in case of homogeneous perovskite material, the antiparallel domain states
(such as ) have all even-rank tensors equal in all components. Odd-rank tensors, on the
Figure 7. Poissons ratio after depoling: (a) by D.C. and (b) after voltage impulses applied for disc
samples. Data missing in curves for APC850 and APC856 are due to the non-resonant response of
samples in certain depoling eld range.
Electric Field Applicability Limits [279]/53
contrary, should have all components of the opposite sign for the antiparallel domains. Due
to the ceramics non-homogeneity and non-uniform domain structure such relationship is
not exactly fullled for the macroscopic material properties.
Depoling behavior for the measured quantities is shown in Figs. 19 for the piezoelec-
tric, elastic and dielectric properties. Even-rank materials tensors (e.g. the permittivity
T
33
and elastic compliance s
E
11
) do not change the sign by reverting the sign of polarization ori-
entation, but just show the substantial property changes in the range of electric eld where
d
33
changes sign. Odd-rank tensors (e.g. piezoelectric coefcient d
33
) change their signs as
Figure 8. Elastic modulus c
E
33
after depoling: (a) by D.C. and (b) after voltage impulses applied for
disc samples.
54/[280] T. Malysh and J. Erhart
Table 2
Electric Field Applicability Limits for PZT Ceramics
D.C. depoling Pulse depoling
APC 850 700 V/mm 500 V/mm
APC 856 200500 V/mm 200 V/mm
Figure 9. Permittivity
T
33
after depoling: (a) by D.C. and (b) after voltage impulses applying for disc
samples.
Electric Field Applicability Limits [281]/55
Table 3
Relative Changes (in %) of Material Coefcients for Hard PZT Ceramics
Relative changes of material coefcients (in %)
D.C. depoling Pulse depoling
Bar Disc Bar Disc
PZT type d

33
d

31
k

31
s
E
11
d

33
k

T
33
d

33
d

31
k

31
s
E
11
d

33
k

p

T
33
APC 840
APC 841
APC 880
2.6
7.9
10.4
10.3
2.6
5.4
13.8
15.5
0.7
17.9
3.6
3.6
8.6
11.6
8.8
16.5
3.6
1.1
10.4
5.1
5.5
10.3
15.9
19.3
8.2
4.7
4.0
4.1
2.8
4.5
7.7
2.7
4.5
7.9
4.8
5.8
3.7
3.4
5.4
6.2
4.3
5.4
a result of the reverted polarization orientation as it was observed. Limits for the electric
elds, not changing the properties substantially, are listed in Table 2. For the hard PZT
ceramics, there were no signicant changes of the material coefcients observed up to the
electric elds of 1000 V/mm. Relative changes of the material coefcients over the entire
depoling electric eld range studied are listed in Table 3. Relative changes are expressed as
a ratio of the maximum change of the property vs. the same property value for not depoled
sample (in absolute value).
Pulse Depoling
Similar behavior as for the D.C. voltage depoling occurs also in this case. Soft PZT ce-
ramics is more sensitive to the depoling eld intensity and the limit elds are even smaller
than for D.C. voltage depoling. This clearly illustrates the dynamic nature of the poling
as well as depoling process. Limiting elds for soft PZT ceramics are listed in Table 2.
Relative changes of the material coefcients for hard PZT (listed in Table 3.) have similar
magnitude (i.e. 320%) like for D.C. depoling (i.e. 118%). There is no big difference
between D.C. and pulse depoling for the hard PZTs studied up to the electric elds of
1000 V/mm.
4. Conclusions
D.C. and pulse electric eld applicability limits were measured for the set of hard as well as
soft PZT ceramics (soft APC 850 and 856, hard APC 840, 841 and 880) up to 1000 V/mm
electric eld strength. Soft PZT ceramics has applicability limits 200700 V/mm for D.C.
eld and 200500 V/mm for pulses. Ceramics is more sensitive to the pulse electric eld
than to D.C. signal. Changes in the material properties are realized by the domain wall
movement, which is more complex phenomenon in pulsed electric elds. Hard PZTs have
their electric eld applicability limits higher than 1000 V/mm and therefore out of the
tested electric eld range. Hard PZT ceramics acquired relative changes in the materials
coefcients measured in the range of 3 to 18%. Piezoelectric coefcients are the most
sensitive material properties studied with respect to the depoling eld.
Our measured values agree well with the published data on soft and hard ceramics,
namely with the limits of 100300 V/mm for soft PZT ([6], Philips, grades PXE5 or 52)
under D.C. depoling and of 300500 V/mm for hard PZT ([6], Philips, grades PXE 41,
56/[282] T. Malysh and J. Erhart
42 or 43). Data similar to our ndings were published also for A.C. depoling for Morgan
Matroc PZT ceramics [7], i.e. 700 V/mm for soft PZT-5A type and >1000 V/mm for hard
PZT-4D and PZT-8 types. Manufacturer of studied PZT types (i.e. APC International Ltd.)
published electric eld applicability limits as high as 720 V/mm for D.C. and 360 V/mm
for A.C. eld for APC840, 600 V/mm for D.C. and 320 V/mm for A.C. eld for APC850
and 800 V/mm for D.C. and 400 V/mm for A.C. eld for APC880 [8].
Some tensor components are more microstructure-dependent than others, e.g. the Pois-
sons ratio
E
shows bigger changes than c
E
33
, which corresponds to the polarization ori-
entation for elementary dipole moments in the direction parallel and perpendicular to the
applied eld. Such effects are related to the textures created by the external elds, grain
geometry etc., which are not topic of this paper.
Acknowledgment
This work has been supported by the Ministry of Education, Youth and Sport under the
project MSM 242200002 and by the Grant Agency of the Czech Republic under the project
No. 202/03/0569.
References
1. T. Ogawa, Ceramics International 26, 383390 (2000).
2. T. Ogawa and K. Nakamura, Jpn. J. Appl. Phys. 38, 54655469 (1999).
3. T. Ogawa, Jpn. J. Appl. Phys. 39, 55385541 (2000).
4. T. Ogawa, Jpn. J. Appl. Phys. 40, 56305633 (2001).
5. IRE Standard, Measurements on piezoelectric ceramics (1961).
6. Data handbook, Piezoelectric ceramics, specialty ferrites, Philips Components 1997, p. 16.
7. Data book for designers, Piezoelectric ceramics, Morgan Matroc Ltd., p. 12.
8. Piezoelectric ceramics, Principles and Applications, APC International Ltd., 2002 (see Table 1.7,
p. 29).

You might also like