In this work, the piezoelectric and dielectric properties of [001]-oriented 0.7Pb(Mg1/3Nb2/3)O3-0.3PbTiO3 single crystals under alternating current poling (ACP) at different temperatures were studied. The piezoelectric coefficients (d33 ∼ 1930 pC/N, d31 ∼ −850 pC/N) and free dielectric permittivity (εT33/ε0 ∼ 7570) reached their highest values when the poling temperature (Tpoling) was 70 °C. Compared with traditional direct current electric field poling at 20 °C, 70 °C-ACP samples showed an enhancement of 40%, 35%, and 49% for d33, d31, and εT330, respectively. Meanwhile, d33 and εT33/ε0 were enhanced by about 9% when Tpoling increased from 20 °C to 70 °C under ACP, while d31 remained the same value and the dielectric loss was lowered from 0.29% to 0.22%. Moreover, ACP samples with different Tpoling have similar electromechanical coupling factors (k31 ∼ 0.44, kt ∼ 0.60). A discussion of the mechanism for the ACP enhancement was based on the domain observation using piezoresponse force microscopy, and the results showed that the domain densities of ACP samples with different Tpoling were positively correlated with their piezoelectric properties. This work demonstrated the enormous potential of ACP optimization for relaxor-PT single crystal applications.

The complex perovskite structure (1–x)Pb(Mg1/3Nb2/3)O3xPbTiO3 (PMN-xPT) single crystals constitute a milestone for the development of piezoelectric materials because their properties outperform those of lead zirconate titanate (PZT) ceramics.1,2 Today, PMN-PT single crystals are still attractive to researchers, due to both their growing commercialization, and their great potential for property enhancement and reliability improvement.3 After decades of systematic studies on the structure–property relationship of relaxor-PT crystals,4–6 further material property enhancement is focused on two main strategies: one is intrinsic or composition-changing method,7–9 and the other is extrinsic and represented by domain engineering10 (or domain wall engineering11,12) methods, including nanoscale electrode engineering13–17 and tuning the poling conditions.18–20 

Domain engineering is first well-known in crystallographic anisotropic characteristic studies and used to describe the polarization rotation induced by the poling process.21,22 The relationship between piezoelectric properties and domain morphology was then studied in BaTiO3 by Wada et al.,23 where it was found that domain engineered crystals with high domain wall densities have high piezoelectric properties. Although the IEEE standard24 recommends PMN-PT poling under a typical field of up to 5 kV/cm for less than 1 min, many studies show that utilizing specific poling methods and the corresponding domain engineering will further enhance the piezoelectric property.18,25,26 Recently, Yamamoto and Yamashita et al.19,20 introduced an alternating current poling (ACP) method in their patents showing a low-cost and time-saving domain engineering method which has attracted much research focus recently. Our previous study18 showed the effectiveness, potential, and stability of this ACP method for PMN-0.3PT single crystals. The property enhancement was attributed to the engineered heterogeneous domain structure with unprecedented domain wall density. At the same time, Xu et al.'s study27 on PMN-0.25PT single crystals reported an even higher enhancement of the longitudinal piezoelectric coefficient (d33) and a higher depoling temperature which makes PMN-0.25PT more favorable for ultrasonic transducer applications. More studies include the thickness dependence of dielectric and piezoelectric properties for ACP relaxor-PT crystals from Qiu et al.28 and temperature stability of ACP PMN-0.3PT single crystals from Zhang et al.29 An increasing number of ACP studies suggest the promise of this domain engineering method on PMN-PT and other relaxor-PT crystals.

The previous ACP studies optimized ACP by varying poling voltage or electric field, cycle number, and frequency18,27 However, the poling temperature (Tpoling) has not been discussed yet in ACP studies while there are many reports that discuss the high Tpoling in the case of the traditional direct current electric field poling (DCP) method, meaning that temperature matters in the poling process. For instance, Feng et al.25 recommended that the PMN-0.3PT single crystal should be poled in silicon oil for 15 min under an applied field of 2–5 kV/cm at 120 °C, followed by the field cooling. The higher Tpoling (120 °C) in this case brought about 20% enhancement on d33 compared to the room temperature poling. Shabbir et al. found that the high temperature poled state of the [001]-oriented PMN-0.33PT single crystal remained stable even after removal of the field and dielectric permittivity was a complex function of temperature and poling field along [001].26 The domain engineering under the high Tpoling was also discussed in the studies from Shen et al.30 and Hu et al.31 In both reports, the domain densities were higher in the high-temperature-poled samples than those room-temperature-poled ones, and the higher domain densities contributed to the enhanced piezoelectric and dielectric properties.

The purpose of this work was to study the effect of Tpoling on PMN-0.3PT single crystals under ACP. The piezoelectric and dielectric properties of ACP samples poled at different temperatures were measured and compared with those of DCP samples. Moreover, the influence of Tpoling on domain structures was investigated, and the correlation between properties and the domain density was analyzed.

[001]-oriented PMN-0.3PT single crystals (from TRS Technologies Inc., State College, Pennsylvania, USA) were diced into 15 × 4 × 0.5 mm3 plates with Ti/Au (10/100 nm) electrodes on top and bottom surfaces (15 mm × 4 mm). Six samples were tested and repeated three times each. Each sample was heated up to 250 °C for 30 min with shorted top and bottom electrodes to fully depole before the poling process,18 and the property at each Tpoling was the average value of at least three samples' measurement results. According to the crystal vendor, the phase transition temperature of rhombohedral to tetragonal is 99 °C and the Curie temperature is 136 °C.

The poling process at different temperatures was conducted in a temperature-controlled chamber. Before the poling, the sample and the chamber were heated up to the target temperature at a ramp of 10 °C/min and remained stable for 5 min. For the ACP samples, a bipolar voltage was generated by a function generator (Agilent 33250A, Santa Clara, California, USA), and then amplified by a high-voltage amplifier (Trek Model 2219, Lockport, New York, USA) before applied to the samples. Referring to our previous study,18,28 the d33 and free dielectric permittivity (εT33/ε0) of ACP samples became saturated with 7 or more cycles by an alternating current electric field with a 10 kV/cm peak-to-peak voltage and 1 Hz frequency. Thus, in this work, all the ACP conditions remained the same except for poling temperature. For DCP, the direct current field was applied for 60 s according to the IEEE standard.24 The overpoling issue which degrades the properties of the DCP sample was also considered, but it was found that there are no obvious different piezoelectric and dielectric properties between DCP samples poled under 5 kV/cm (IEEE standard) and 10 kV/cm (in this work) at different temperatures.

After poling, samples were air cooled to room temperature. Then, d33 was measured using a quasi-static piezo d33 meter (Model ZJ-4B, Chinese Academy of Science). εT33/ε0 was calculated from the electrical capacitance measured by a multi-frequency LCR meter (Agilent 4294A, Santa Clara, California, USA) at 1 kHz. The transversely poled length-extensional piezoelectric coefficient (d31), thickness mode coupling factor (k31), and the thickness mode coupling factor (kt) were calculated per the IEEE standard.32 The corresponding domain structures were observed by piezoresponse force microscopy (PFM) (Dimension Icon, Bruker, Santa Barbara, California, USA). PFM images were scanned on the freshly cracked cross-section surface since the domain structures of PMN-PT single crystal are strongly affected by the dicing and lapping process conditions.33 The final domain width statistics was analyzed based on six scanned areas for each sample.

Figure 1 summarizes the property characterization results for both ACP and DCP samples with different Tpoling. As shown in Figs. 1(a) and 1(b), the Tpoling corresponding d33 and εT33/ε0 curves from ACP and DCP samples have similar profiles while the maximum values of d33 and εT33/ε0 were not obtained at the same Tpoling. Specifically, the peak Tpoling is 70 °C for ACP samples, with d33 and εT33/ε0 of 1930 pC/N and 7570, respectively. The d33 and εT33/ε0 from 70 °C-ACP samples are both enhanced about 9% compared to room temperature ACP (20 °C-ACP) samples, and when compared with those from room temperature DCP (20 °C-DCP), the enhancement is 40% and 49%, respectively. In the other case of DCP, the Tpoling at which peak d33 and εT33/ε0 occurred is slightly higher (80 °C) than that from ACP, where the 80 °C-DCP sample has the highest d33 and εT33/ε0 of 1650 pC/N and 5940, respectively. The difference between ACP and DCP samples here can be explained from the field-induced phase transition and thermodynamic theory. From our previous study,18 it was reported that 20 °C-DCP samples remained in the rhombohedral (R) phase and 20 °C-ACP samples experienced R to monoclinic A (MA) field-induced phase transition after poling. The different phase transitions can be explained according to the free-energy density profiles (or Landau energy density profiles34) of the DCP and ACP samples. The ACP sample has a flatter free-energy density profile between R and MA phases compared to DCP one, which makes the phase transition from R to MA phase easier and leads to higher piezoelectric properties34 in the ACP sample. In this work, when the Tpoling is changed, the flatter free energy also helps the ACP sample to reach the property peak at a slightly lower Tpoling than that of DCP samples.18 

FIG. 1.

The piezoelectric and dielectric properties of ACP and DCP samples with different poling temperatures: (a) the piezoelectric coefficient d33; (b) the free dielectric permittivity and loss.

FIG. 1.

The piezoelectric and dielectric properties of ACP and DCP samples with different poling temperatures: (a) the piezoelectric coefficient d33; (b) the free dielectric permittivity and loss.

Close modal

Further increasing Tpoling above the peak Tpoling brought a decrease in both d33 and εT33/ε0 in two steps. The first step is from the peak Tpoling to 110 °C, where the d33 decreased to 1320 pC/N, and εT33/ε0 dropped to 5200 in ACP samples, and these values are close to those from DCP samples at the same Tpoling. The second dropping step appeared between 110 °C and 120 °C, where both d33 and εT33/ε0 dropped to the unpoled level, showing that the poled status cannot be maintained during the cooling for both ACP and DCP samples when Tpoling is higher than 110 °C. In addition to εT33/ε0, the dielectric loss data are also illustrated in Fig. 1(b). The dielectric loss of 20 °C ACP samples is 0.29%, which is close to that of DCP samples. The loss gradually decreases to 0.22% and 0.26% for ACP and DCP samples, respectively, when the Tpoling is increased to 70 °C. When Tpoling is higher than 110 °C, the rapidly increased loss values and the decreased d33 and εT33/ε0 suggest that the samples should not be poled at such high temperatures.

Electromechanical coupling factors of length-extensional mode (k31) and thickness-extensional mode (kt) were also measured through the electrical impedance and phase spectra of ACP and DCP samples. k31 and kt were calculated based on the corresponding resonance (fr) and antiresonance (fa) frequencies, which are listed in Table I. Figure 2(a) shows the first frequency response corresponding to the fundamental mode of k31 of the 70 °C ACP sample. The value of k31 was estimated to be 0.44 (fr = 47.92 kHz, fa = 52.08 kHz). Figure 2(b) further shows that k31 and kt remain stable in a certain range with the change of Tpoling.

TABLE I.

Coupling factors of ACP and DCP samples with different Tpoling.

Poling MethodTpoling (°C)k31 modeakt modeb
fr (kHz)fa (kHz)k31d31 (pC/N)fr (MHz)fa (MHz)kt
DCP 20 52.84 57.04 0.42 -630 3.84 4.565 0.58 
ACP 20 47.48 51.50 0.43 -840 3.83 4.59 0.59 
ACP 70 47.92 52.08 0.44 -850 3.80 4.60 0.60 
Poling MethodTpoling (°C)k31 modeakt modeb
fr (kHz)fa (kHz)k31d31 (pC/N)fr (MHz)fa (MHz)kt
DCP 20 52.84 57.04 0.42 -630 3.84 4.565 0.58 
ACP 20 47.48 51.50 0.43 -840 3.83 4.59 0.59 
ACP 70 47.92 52.08 0.44 -850 3.80 4.60 0.60 
a

k31 mode: (small signal transversely poled) length-extensional mode.24 

b

kt mode: thickness-extensional mode.

FIG. 2.

Electromechanical coupling properties: (a) frequency response of k31 mode (b) coupling factors of ACP/DCP samples with different Tpoling.

FIG. 2.

Electromechanical coupling properties: (a) frequency response of k31 mode (b) coupling factors of ACP/DCP samples with different Tpoling.

Close modal

d31 at different Tpoling were also calculated according to the IEEE standard.32Table I shows that although ACP samples have a 35% improvement in d31 compared to DCP samples (from −630 to −850 pC/N), the Tpoling hardly affects the d31 of ACP samples. Thus, raising Tpoling to 70 °C is proved to enhance the properties of the ACP sample effectively, and supplementary Fig. 3s further shows the excellent consistency of 70 °C-ACP compared to the DCP among six samples. The error bars here are from the different measuring points of the same sample.

To study the mechanism behind the property enhancement from the high-temperature ACP, PFM was used to observe the corresponding domain structures. Figure 3 shows the domain patterns at the cross sections of samples poled at different temperatures. The domain patterns show typical ACP ‘4R’ domain configurations containing broad and nearly [100]-orientated 109° domain walls, and fine [101]-orientated 71° domain walls. The most significant difference in the domain morphology from samples with different Tpoling is the distribution of 109° domain walls. The 109° domain walls of 70 °C ACP samples have a more uniform width in each domain and are parallel to each other to form a longer range of ordering. Such ordering is much weaker in the 110 °C ACP and 20 °C ACP samples, where the domain wall orientation is poor. From the comparison of magnified areas in each panel, the 71° domain wall density can also be calculated from the half of the width between the two contrast peaks. Although Fig. 3 shows only one scanned area for each Tpoling, the domain width statistics, as listed in Table II, is collected from six scanned areas or PFM images for each sample.

FIG. 3.

PFM images: (a) 20 °C ACP sample, (b) 70 °C ACP sample, and (c) 110 °C ACP sample. The contrast is from the out-of-plane polarization phases, and the output amplitude is normalized.

FIG. 3.

PFM images: (a) 20 °C ACP sample, (b) 70 °C ACP sample, and (c) 110 °C ACP sample. The contrast is from the out-of-plane polarization phases, and the output amplitude is normalized.

Close modal
TABLE II.

Domain width of ACP samples poled at different temperatures.

Poling MethodTpoling (°C)109° Domain width (nm)71° Domain width (nm)d33 (pC/N)Reference
DCP 20 2000 ± 800 920 ± 80 1650 ± 20 18  
ACP 20 700 ± 10 160 2000 ± 10 18  
ACP 20 710 ± 90 102 ± 11 1770 ± 30 This research 
ACP 70 640 ± 160 96 ± 6 1930 ± 20 This research 
ACP 110 850 ± 40 119 ± 18 1320 ± 80 This research 
Poling MethodTpoling (°C)109° Domain width (nm)71° Domain width (nm)d33 (pC/N)Reference
DCP 20 2000 ± 800 920 ± 80 1650 ± 20 18  
ACP 20 700 ± 10 160 2000 ± 10 18  
ACP 20 710 ± 90 102 ± 11 1770 ± 30 This research 
ACP 70 640 ± 160 96 ± 6 1930 ± 20 This research 
ACP 110 850 ± 40 119 ± 18 1320 ± 80 This research 

As shown in Table II, a positive correlation between the domain wall density and the Tpoling can be established. ACP samples at the peak Tpoling (70 °C) have the highest domain wall densities of both 109° and 71° domains. Thus, the high piezoelectric properties are associated with the uniformly distributed 109° domain walls and the high domain wall densities of both 109° and 71° domains.

This observation further explains how Tpoling of the ACP method dynamically influences the domain engineering. The uniform domain distribution corresponds to the low coercive field at elevated temperatures.35 According to Landau theory, when Tpoling is increased, the polarization rotation becomes easier, and the domain wall movement will be faster. On the other hand, the higher domain density is closely related to the higher internal energy of the system.3 The higher Tpoling raises the maximum level of the internal energy, and such enhancement is found relaxed when Tpoling is higher than the peak Tpoling.

To conclude, this paper studies how the poling temperature affects the piezoelectric and dielectric properties of PMN-0.3PT single crystal using the ACP method. Experiments showed that d33, d31, and εT33/ε0 reached their maximum values when ACP poling was conducted at 70 °C. At 70 °C, the enhancement was more than 40% and 9% compared with 20 °C-DCP/ACP samples. Meanwhile, the dielectric losses on ACP samples were lower than those of DCP samples. In addition, ACP samples with different Tpoling have similar electromechanical coupling factors - k31 and kt. All the improvement can be explained by the domain observation which shows that 70 °C-ACP samples have the highest domain wall density from both 109° and 71° domains, while domain densities at different Tpoling are positively correlated with their piezoelectric properties. We demonstrated an excellent cost-effective domain engineering method of ACP optimization for relaxor-PT single crystals that have large potential for future single crystal applications.

See supplementary material for the schematic experimental setup and poling field profiles (Fig. 1s); mechanism and settings of PFM (Fig. 2s); sample consistency between six samples (Fig. 3s) and the hysteresis loops of ACP and DCP sample poling at room temperature (Fig. 4s).

This work was primarily supported by ONR under Grant Nos. N00014-15-1-2418 and N00014-18-1-2538. This work was performed in part at the Analytical Instrumentation Facility (AIF) at North Carolina State University, which is supported by the State of North Carolina and the National Science Foundation (Award No. ECCS-1542015). The AIF is a member of the North Carolina Research Triangle Nanotechnology Network (RTNN), a site in the National Nanotechnology Coordinated Infrastructure (NNCI).

1.
S.
Zhang
,
F.
Li
,
X.
Jiang
,
J.
Kim
,
J.
Luo
, and
X.
Geng
,
Prog. Mater. Sci.
68
,
1
(
2015
).
2.
S.
Zhang
and
F.
Li
,
J. Appl. Phys.
111
(
3
),
031301
(
2012
).
3.
E.
Sun
and
W.
Cao
,
Prog. Mater. Sci.
65
,
124
(
2014
).
4.
Z. G.
Ye
and
M.
Dong
,
J. Appl. Phys.
87
(
5
),
2312
(
2000
).
5.
T.
Ogawa
,
Ferroelectrics
273
(
1
),
371
(
2002
).
6.
K. K.
Rajan
,
Y. S.
Ng
,
J.
Zhang
, and
L. C.
Lim
,
Appl. Phys. Lett.
85
(
18
),
4136
(
2004
).
7.
S.
Zhang
and
T.
Shrout
,
IEEE Trans. Ultrason. Ferroelectr. Freq. Control
57
(
10
),
2138
(
2010
).
8.
A.
Hussain
,
N.
Sinha
,
A. J.
Joseph
,
S.
Goel
, and
B.
Kumar
,
J. Mater. Sci.-Mater. Electron.
29
(
22
),
19567
(
2018
).
9.
Y.
Li
,
Y.
Tang
,
J.
Chen
,
X.
Zhao
,
L.
Yang
,
F.
Wang
,
Z.
Zeng
, and
H.
Luo
,
Appl. Phys. Lett.
112
(
17
),
172901
(
2018
).
10.
S.
Wada
,
K.
Muraoka
,
H.
Kakemoto
,
T.
Tsurumi
, and
H.
Kumagai
,
Jpn. J. Appl. Phys., Part 1
43
(
9B
),
6692
(
2004
).
11.
S.
Wada
,
K.
Yako
,
T.
Muraishi
,
K.
Yokoh
,
S. M.
Nam
,
H.
Kakemoto
, and
T.
Tsurumi
,
Ferroelectrics
340
(
1
),
17
(
2006
).
12.
S.
Wada
,
T.
Muraishi
,
K.
Yokoh
,
K.
Yako
,
H.
Kamemoto
, and
T.
Tsurumi
,
Ferroelectrics
355
,
37
(
2007
).
13.
D.
Li
and
D. A.
Bonnell
,
Ceram. Int.
34
(
1
),
157
(
2008
).
14.
W.-Y.
Chang
,
C.-C.
Chung
,
Z.
Yuan
,
C.-H.
Chang
,
J.
Tian
,
D.
Viehland
,
J.-F.
Li
,
J. L.
Jones
, and
X.
Jiang
,
Acta Mater.
143
,
166
(
2018
).
15.
W.-Y.
Chang
,
W.
Huang
,
A.
Bagal
,
C.-H.
Chang
,
J.
Tian
,
P.
Han
, and
X.
Jiang
,
J. Appl. Phys.
114
(
11
),
114103
(
2013
).
16.
M.
Gao
,
C.
Luo
,
W.-Y.
Chang
,
C. M.
Leung
,
J.
Tian
,
J.
Li
,
X.
Jiang
, and
D.
Viehland
,
Acta Mater.
169
,
28
(
2019
).
17.
Y.
Yamashita
,
N.
Yamamoto
,
K.
Itsumi
, and
Y.
Hosono
,
Jpn. J. Appl. Phys., Part 1
50
(
9S2
),
09NC05
(
2011
).
18.
W.-Y.
Chang
,
C.-C.
Chung
,
C.
Luo
,
T.
Kim
,
Y.
Yamashita
,
J. L.
Jones
, and
X.
Jiang
,
Mater. Res. Lett.
6
(
10
),
537
(
2018
).
19.
N.
Yamamoto
,
Y.
Yamashita
,
Y.
Hosono
,
K.
Itsumi
, and
K.
Higuchi
, “
Ultrasonic probe, piezoelectric transducer, method of manufacturing ultrasonic probe, and method of manufacturing piezoelectric transducer
,” U.S. patent 2014/0062261 A1 (
2014
).
20.
Y.
Yamashita
,
N.
Yamamoto
,
Y.
Hosono
, and
K.
Itsumi
, “
Piezoelectric transducer, ultrasonic probe, and piezoelectric transducer manufacturing method
,” U.S. patent 2015/0372219 A1 (
2015
).
21.
H.
Fu
and
R. E.
Cohen
,
Nature
403
,
281
(
2000
).
22.
F.
Li
,
S.
Zhang
,
Z.
Xu
,
X.
Wei
, and
T. R.
Shrout
,
Adv. Funct. Mater.
21
(
11
),
2118
(
2011
).
23.
S.
Wada
,
S.
Suzuki
,
T.
Noma
,
T.
Suzuki
,
M.
Osada
,
M.
Kakihana
,
S.-E.
Park
,
L. E.
Cross
, and
T. R.
Shrout
,
Jpn. J. Appl. Phys., Part 1
38
(
9B
),
5505
(
1999
).
24.
IEEE standard for relaxor-based single crystals for transducer and actuator applications
,” ANSI/IEEE Std. 1859 (
2017
).
25.
Z.
Feng
,
X.
Zhao
, and
H.
Luo
,
Mater. Res. Bull.
41
(
6
),
1133
(
2006
).
26.
G.
Shabbir
,
S.
Kojima
, and
C.
Feng
,
J. Appl. Phys.
100
(
6
),
064107
(
2006
).
27.
J.
Xu
,
H.
Deng
,
Z.
Zeng
,
Z.
Zhang
,
K.
Zhao
,
J.
Chen
,
N.
Nakamori
,
F.
Wang
,
J.
Ma
,
X.
Li
, and
H.
Luo
,
Appl. Phys. Lett.
112
(
18
),
182901
(
2018
).
28.
C.
Qiu
,
J.
Liu
,
F.
Li
, and
Z.
Xu
,
J. Appl. Phys.
125
(
1
),
014102
(
2019
).
29.
Z.
Zhang
,
J.
Xu
,
L.
Yang
,
S.
Liu
,
J.
Xiao
,
R.
Zhu
,
X.
Li
,
X. a
Wang
, and
H.
Luo
,
J. Appl. Phys.
125
(
3
),
034104
(
2019
).
30.
Z.-Y.
Shen
,
Y.
Tang
,
S.
Zhang
,
J.
Luo
,
Y.
Li
, and
T. R.
Shrout
,
Scr. Mater.
72–73
,
17
(
2014
).
31.
W.
Hu
,
Z.
Xi
,
P.
Fang
,
W.
Long
,
X.
Li
, and
Q.
Bu
,
J. Mater. Sci.-Mater. Electron.
26
(
5
),
3282
(
2015
).
32.
IEEE standard on piezoelectricity
,” ANSI/IEEE Std. 176 (
1987
).
33.
H.
Takahashi
,
H.
Suzuki
, and
Y.
Namba
,
CIRP Ann. -Manuf. Technol.
65
(
1
),
541
(
2016
).
34.
F.
Li
,
D.
Lin
,
Z.
Chen
,
Z.
Cheng
,
J.
Wang
,
C.
Li
,
Z.
Xu
,
Q.
Huang
,
X.
Liao
,
L.-Q.
Chen
,
T. R.
Shrout
, and
S.
Zhang
,
Nat. Mater.
17
(
4
),
349
(
2018
).
35.
T.-F.
Zhang
,
X.-G.
Tang
,
P.-Z.
Ge
,
Q.
Liu
, and
Y.
Jiang
,
Ceram. Int.
43
(
18
),
16300
(
2017
).

Supplementary Material