Characterization and high piezoelectric performance of Pb(Fe1/2Nb1/2)O3–Pb(In1/2Nb1/2)O3–PbTiO3 ternary ceramics

Guanyu Zhu, Hui Liu, Shengdong Sun, Botao Gao and Jun Chen*
Beijing Advanced Innovation Center for Materials Genome Engineering, Department of Physical Chemistry, University of Science and Technology Beijing, Beijing 100083, China. E-mail: junchen@ustb.edu.cn

Received 13th August 2019 , Accepted 21st September 2019

First published on 24th September 2019


A (0.93 − x)Pb(Fe1/2Nb1/2)O3–0.07Pb(In1/2Nb1/2)O3xPbTiO3 (0.06 ≤ x ≤ 0.11) ternary system was synthesized using a B-site precursor solid-state reaction method. The maximum piezoelectric coefficient d33 = 673 pC N−1 was achieved at the morphotropic phase boundary (MPB) composition x = 0.09 and its Curie temperature was 141 °C. In situ high energy synchrotron X-ray diffraction (SXRD) results demonstrate that tetragonal and monoclinic phases coexist at the MPB composition. It is interesting to observe that the stable single monoclinic phase induced by the electric field exhibits the characteristics of flexible lattice parameters in response to the electric field. The large piezoelectric response of PFN–PIN–PT is found to be mainly ascribed to the intrinsic lattice strain.


Introduction

Owing to the excellent piezoelectric properties, the prototype Pb(Zr1−xTix)O3 (PZT)-based piezoelectric materials have been widely utilized as sensors, actuators, and so on.1,2 In order to satisfy the requirements of various applications, a variety of piezoelectric systems have been investigated based on morphotropic phase boundary (MPB), in which tetragonal, orthorhombic, rhombohedral, or monoclinic phases can coexist. Piezoelectric materials with high piezoelectric coefficients are of great interest. Recently, ultrahigh piezoelectricity with a piezoelectric coefficient d33 of up to 1500 pC N−1 has been achieved in Sm-doped Pb(Mg1/3Nb2/3)O3–PbTiO3 (PMN–PT) ceramics.3 Besides, the corresponding single crystals present an extremely high d33 (3400 to 4100 pC N−1).4

As a typical relaxor ferroelectric material, the binary system Pb(Fe1/2Nb1/2)O3–PbTiO3 (PFN–PT) with a low sintering temperature has attracted considerable attention.5,6 Interestingly, a low symmetric monoclinic phase has been observed near the MPB composition of PFN–PT (8 mol% PT),6,7 which is believed to play an important role in its piezoelectric performance.8,9 However, the undesired oxygen vacancies and Fe2+ formed during the sintering process of PFN–PT ceramics result in high dielectric loss and low resistivity,10,11 which leads to inferior piezoelectric performance. Various strategies have been employed to overcome this limitation, such as addition of different end numbers like PbZrO3 (PZ),12–14 to form a ternary system to improve the electrical behaviors. Besides, a small amount of additive such as Li2CO3 was used to achieve an enhanced piezoelectric coefficient of 590 pC N−1.15

Pb(In1/2Nb1/2)O3 (PIN) is also a classical relaxor ferroelectric with a TC of about 90 °C.16 It can easily form a full range of binary solid solutions with PbTiO3, in which good piezoelectric properties (d33 = 400 pC N−1) and a relatively high Curie temperature (TC = 320 °C) can be found at the MPB composition.17–19 It would be interesting to add PIN into the PFN–PT system to form a ternary system, which is expected to exhibit an enhanced piezoelectric performance. Herein, a small amount of the PIN (7 mol%) component was added and a novel ternary system of (0.93 − x)PFN–0.07PIN–xPT was formed. The structure and piezoelectric, ferroelectric and dielectric properties were investigated in detail. In particular, in situ high energy synchrotron X-ray diffraction (SXRD) was performed to understand the origin of the high piezoelectric performance of the PIN–PFN–PT system.

Experimental procedure

The ceramic samples of (0.93 − x)PFN–0.07PIN–xPT (x = 0.06–0.11, abbreviated as PFN–PIN–xPT) were fabricated using the modified two-step B-site precursor method.20–22 Oxide powders PbO, In2O3, Nb2O5, Fe2O3 and TiO2 with analytical purity of not less than 99.0% were used as the starting materials. First, to form FeNbO4, a mixture of stoichiometric amounts of Fe2O3 and Nb2O5 was wet-milled in ethanol for 24 h in zirconia milling media and calcined at 1050 °C for 6 h. Similarly, to form InNbO4, In2O3 and Nb2O5 were mixed in stoichiometric amounts and calcined at 1100 °C for 6 h. Then stoichiometric amounts of PbO (excess 1 mol%), FeNbO4, InNbO4, and TiO2 were ball-milled in ethanol for 24 h, and then calcined at 850 °C for 2 h to form the perovskite structure. The reacted powders were subsequently ball-milled again in ethanol for 12 h. After the second ball-milling, the powders were mixed with polyvinyl alcohol (PVA) and pressed into disks, 8.5 mm in diameter and 1.5 mm in thickness, under a uniaxial pressure of 20 MPa. Finally, after burning out the PVA binder at 550 °C for 2 h, the sintering process was performed at 1150 °C for 2 h in sealed crucibles. It should be noted that the sintering temperature of PFN–PIN–PT is at least 50 °C lower than that of PIN–PT.18,19

The structural analysis of the obtained samples was performed using an X-ray powder diffractometer with Cu-Kα radiation (XRD, X'Pert PRO, PANalytical, The Netherlands). The surface morphologies of the samples were obtained by using a scanning electron microscope (SEM, LEO1530, Germany).

For electrical performance characterization, the gold electrode was sputtered on both sides of the polished ceramic samples using a small ion sputtering apparatus (ETD-3000, China). Electric poling was carried out under an electric field of 20 kV cm−1 for 20 min at room temperature. A ferroelectric analyzer (TF Analyzer 1000, aixACCT, Germany) was used for measuring the ferroelectric PE hysteresis loop and the strain versus electric field (SE) curve of the samples at 1 Hz. The temperature dependent dielectric constant and dielectric loss were measured for the poled samples using an LCR meter (E4980, Agilent, Santa Clara, CA) operated under 0.5 V at 1 kHz. A quasi-static piezoelectric d33 meter (Institute of Acoustics, Chinese Academy of Sciences, ZJ-4A) was used to measure the piezoelectric coefficient (d33).

At the 11-ID-C beamline of the Advanced Photon Source (APS), Argonne National Laboratory, we performed the in situ high-energy synchrotron X-ray diffraction (SXRD) measurements. The wavelength of the X-ray was 0.1173 Å. The details of the experimental procedure can be obtained from our previous studies.9,23,24

Results and discussion

It is well known that perovskite-type solid solutions in the MPB region exhibit the optimum piezoelectric properties, thus most of the studies are focused on this region. The XRD patterns of the crushed powders of PFN–PIN–xPT (x = 0.06–0.11) are shown in Fig. 1a. One can see that all specimens exhibit a pure perovskite phase without any secondary pyrochlore phase. The (200)PC profile gradually changes from a singlet to two distinct peaks with increasing PT content, indicating a phase transition from the monoclinic or rhombohedral phase to the tetragonal one. Such evolution is analogous to the previous reports on PFN–PT25 and Li-doped PFN–PT.15 The MPB region of the present PFN–PIN–xPT system is determined at x = 0.09, which shifts to a higher PT content than the binary system of PFN–PT.7
image file: c9qi01022j-f1.tif
Fig. 1 (a) XRD patterns of PFN–PIN–xPT (x = 0.06–0.11). (b) Microstructure of the surface of the PFN–PIN–0.09PT ceramic sintered at 1150 °C.

The relative density of the PFN–PIN–xPT (x = 0.06–0.11) ceramics is more than 95% of theoretical density. Fig. 1b shows the representative SEM image of the MPB composition. It can be seen that the ceramic is dense and well sintered without apparent pores. The average grain size is about 9.3 μm evaluated using a freeware SmileView.

The ferroelectric PE loops and SE curves measured at 1 Hz at room temperature are shown in Fig. 2. Saturated ferroelectric PE loops without apparent leakage currents are observed for all compositions. The maximum polarization (Pmax), remnant polarization (Pr), and coercive field (EC) of the samples are plotted in Fig. 2b. It can be seen that the maximum values of Pmax and Pr (32 μC cm−2 and 24 μC cm−2, respectively) are obtained at the MPB composition x = 0.09, both of which are larger than those of the MPB composition of Li doped PFN–PT.26 The EC decreases with decreasing PT content, which agrees with the reduced structure distortion observed in the XRD patterns. It should be noted that the EC significantly decreases from 6 kV cm−1 to 3 kV cm−1 with the PT content decreasing from 0.11 to 0.09, which also confirms that MPB has the composition x = 0.09. The EC value of PFN–PIN–0.09PT is similar to that of the 0.675PMN–0.325PT ceramic.7


image file: c9qi01022j-f2.tif
Fig. 2 (a) PE loops, (b) EC, Pr, and Pmax, (c) bipolar SE curves, and (d) unipolar SE curves of PFN–PIN–xPT (x = 0.06–0.11).

The SE curves of PFN–PIN–xPT are shown in Fig. 2c and d. It can be seen that the curves for all compositions exhibit a well-defined butterfly shape in accordance with the typical ferroelectric PE loops. Although the largest positive strain at 2 kV mm−1 is obtained at x = 0.10, the total peak-to-peak strain is the largest at the MPB composition x = 0.09 (Fig. 2c).

The small-signal d33 can be directly obtained using a quasi-static piezoelectric d33 meter. The large-signal image file: c9qi01022j-t1.tif can be estimated as image file: c9qi01022j-t2.tif according to the unipolar SE curves shown in Fig. 2d. The evolution of small and large-signal image file: c9qi01022j-t3.tif as a function of PT content is shown in Fig. 3. Both small-signal d33 and large-signal image file: c9qi01022j-t4.tif reach the maximum value (673 pC N−1 and 797 pm V−1, respectively) at the MPB composition x = 0.09. Interestingly, it should be noted that the piezoelectric response is much enhanced compared with the PIN–PT binary system (400 pC N−1).17 As a comparison, the small-signal d33 of x = 0.09 is larger than those of the systems of BiFeO3–PFN–PT (351 pC N−1),27,28 PMN–PT–PFN (545 pC N−1),25 PMN–PIN–PT (550 pC N−1),17 Cu modified PIN–PMN–PT (584 pC N−1),29 Li-doped PFN–PT (590 pC N−1),15 and PMN–PFN–PZT (626 pC N−1),30 and even comparable to the classical high performance system of PMN–PT (670 pC N−1).31


image file: c9qi01022j-f3.tif
Fig. 3 Small and large-signal image file: c9qi01022j-t9.tif of PFN–PIN–xPT as a function of PT content.

In situ high-energy SXRD was employed at the MPB composition of PFN–PIN–0.09PT to explore the structural origin of high piezoelectric response in PFN–PIN–xPT. It has been demonstrated that the influence of the texture induced by the electric field can be neglected at the 45° sector.9,23,24,32 Therefore, the diffraction patterns at the 45° sector are adopted to determine the structure using the full-profile Rietveld refinement method. As shown in Fig. 4a, the electric field induces an irreversible phase transition. For the unpoled state, the phase structure is determined as the phase coexistence of P4mm + Cm, whereas after poling, it exhibits a single monoclinic phase (Cm). The representative structural refinement results are shown in Fig. 4b and c. One can see that the observed experimental profiles can be fitted well by the models. The agreement Rwp factor is as low as 4.5%, indicating highly reliable results. In the unpoled state, the phase content is 72% for the monoclinic phase and 28% for the tetragonal phase. It should be noted that once the coexisting phases are completely transformed into the monoclinic phase at E = 0.4 kV mm−1, the monoclinic phase does not come back to the original unpoled state of phase coexistence in the subsequent electric field loading process. This scenario has also been observed in PZT and PMN–PT.23,32


image file: c9qi01022j-f4.tif
Fig. 4 (a) Diffraction peak profiles and contour plots of (200)PC with respect to the electric field at the 45° sector of PFN–PIN–0.09PT. The blue arrows indicate the direction of increasing electric field while the pink arrows indicate the reverse process. Full-profile Rietveld refinement of PFN–PIN–0.09PT at the 45° sector, (b) at 0 kV mm−1 (unpoled state), and (c) at E = 0.4 kV mm−1. The insets show the enlarged profiles of (111)PC and (200)PC. (d) The evolution of lattice parameters of the monoclinic phase with respect to the bipolar electric field.

After full-profile Rietveld refinement is carried out, lattice parameters of the monoclinic phase under an applied electric field can be obtained. As shown in Fig. 4d, the lattice parameters of the monoclinic phase show a well-defined butterfly curve as a function of bipolar electric field, suggesting that the macroscopic piezoelectric properties are closely related to the intrinsic structural evolution. Such evolution of lattice parameters of the monoclinic phase has also been observed in the prototype system of PMN–PT.9 The lattice parameters are am = 5.6892 Å, bm = 5.6769 Å, cm = 4.0244 Å and βm = 90.096° at zero electric field (poled state, Table 1), which coincide with the previously reported values of PFN–PT.6,7

Table 1 Structural refinement results for the monoclinic phase of the poled 0.84PFN–0.07PIN–0.09PT ceramics at zero electric field
Phase Cm
a (Å) 5.6892(2)
b (Å) 5.6769(1)
c (Å) 4.0244(2)
β (°) 90.096(3)
x-Nb/Fe/In/Ti 0.482(1)
y-Nb/Fe/In/Ti 0
z-Nb/Fe/In/Ti 0.536(1)
x-O(I) 0.458(3)
y-O(I) 0
z-O(I) 0.121(1)
x-O(II) 0.226(3)
y-O(II) 0.245(2)
z-O(II) 0.580(2)
Biso(Pb)2) 3.01(1)
Rwp (%) 4.14
Rp (%) 6.05
χ2 1.67


The spontaneous polarization (PS) of the monoclinic phase can be calculated by considering a pure ionic crystal based on the structural parameters obtained from the structure refinements.9,32 The calculated value of PS is about 43 μC cm−2 at all electric field amplitudes, which is larger than the measured value from the PE hysteresis loops (Pmax = 32 μC cm−2). The monoclinic phase in the present study belongs to MA, in which the polarization direction varies between 17.3° and 26.0° off the [100]PC direction.9,32 It should be noted that the electric-field-induced monoclinic phase of MA has also been observed in other high performance piezoelectric systems.9,23,32

According to the previous studies,9,23,24 diffraction patterns parallel to the electric field direction, namely at the 0° sector, can be used to evaluate lattice strain. The lattice strain (ε) can be calculated by the equation ε = di,200PC/d0,200PC − 1, where di and d0 are the d spacings calculated from the (200)PC peak at the applied electric field i and at 0 kV mm−1 of the poled state, respectively. The comparison between the calculated lattice strain of the (200)PC profile and the measured macro strain is shown in Fig. 5. Interestingly, the lattice strain is in good agreement with the measured macro strain. This is in accordance with our previous observation in PZT and PMN–PT.9,23 It is well known that intrinsic lattice strain and extrinsic domain switching are the main contributors to the piezoelectric response.33 The monoclinic phase exhibits the characteristic of negligible domain switching, namely the domain switching makes a limited contribution to the piezoelectric response in the monoclinic phase.23 Hence, the high piezoelectric response of PFN–PIN–PT mainly stems from the intrinsic lattice strain. The calculated lattice stain at 2 kV mm−1 is 0.12%, corresponding to a piezoelectric parameter of 600 pm V−1. As a comparison, the calculated piezoelectric parameter in the present system is higher than that in the monoclinic phase of PZT (520 pm V−1).23


image file: c9qi01022j-f5.tif
Fig. 5 A comparison between the calculated lattice strain of the (200)PC profile and the measured macro strain in PFN–PIN–0.09PT.

It is well known that thermal stability is an important property of piezoelectric materials. The temperature dependent ferroelectric and piezoelectric properties of PFN–PIN–0.09PT are displayed in Fig. 6. One can see that EC, Pr, Pmax and negative strain decrease with increasing temperature (Fig. 6a and b). The reduced Pr and Pmax should be due to the weakened spontaneous polarization in the unit cell, since polarization always decreases with increasing temperature.34 The reduced EC and negative strain would be attributed to the reduction of structure distortion with increasing temperature.35 As shown in Fig. 6c and d, the strain decreases with increasing temperature corresponding to the reduced large-signal image file: c9qi01022j-t5.tif. The large-signal image file: c9qi01022j-t6.tif is 797 pm V−1 at room temperature, and it decreases to 208 pm V−1 at 150 °C. This indicates that the large-signal image file: c9qi01022j-t7.tif of the present system is sensitive to temperature. The reduced image file: c9qi01022j-t8.tif could be ascribed to the domain texture degradation, reduced structural distortion of the monoclinic phase, and the emergence of a paraelectric cubic phase.36


image file: c9qi01022j-f6.tif
Fig. 6 Temperature dependence of (a) PE loops, and (b) bipolar SE curves, (c) unipolar SE curves, and (d) large-signal image file: c9qi01022j-t10.tif of PFN–PIN–0.09PT.

Fig. 7a displays the dielectric properties of the PFN–PIN–xPT samples. The Curie temperature (TC) increases from 128 °C to 144 °C with the PT content increasing from 0.06 to 0.10 (Fig. 7b). The TC values of these compositions are slightly higher than those of PFN–xPT ceramics (121 °C for x = 0.05 and 125 °C for x = 0.06).7,37 It should be noted that the TC of the MPB composition of PFN–PIN–0.09PT is comparable to that of 0.65PMN–0.35PT (150 °C).38 It is evident that a dielectric abnormal peak is observed below 100 °C for x ≤ 0.08, similar to the previously reported PMN–PT–PFN,25 which could be corresponding to the monoclinic-to-tetragonal phase transition. The dielectric constant of PFN–PIN–0.09PT at room temperature (4835) is much higher than those of PIN–xPT (x = 0.37, 2450)17 and PMN–PT–PFN (3094).25 The dielectric losses of all compositions studied in the present work are less than 0.02 at room temperature, which is even lower than those of Li-doped PFN–PT.15


image file: c9qi01022j-f7.tif
Fig. 7 (a) Temperature dependent dielectric constant and loss measured at 1 kHz, and (b) the Curie temperature TC of PFN–PIN–xPT ceramics.

Conclusions

PFN–PIN–xPT ceramics were prepared via the B-site precursor solid reaction method. The MPB composition of PFN–PIN–xPT is determined at x = 0.09, exhibiting optimal ferroelectric and piezoelectric properties, in which a high piezoelectric coefficient d33 of 673 pC N−1 with a Curie temperature of 141 °C is obtained. A single electric-field-induced monoclinic phase has been achieved in the sample PFN–PIN–0.09PT. When a bipolar electric field is applied, the lattice parameters of the monoclinic structure exhibit a butterfly shaped curve. It is found that the large piezoelectric response of PFN–PIN–PT is mainly ascribed to the intrinsic lattice strain.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 21825102, 21731001, and 21590793), and the Fundamental Research Funds for the Central Universities, China (Grant No. FRF-TP-18-001C2). This research used resources of the Advanced Photon Source, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract No. DE-AC02-06CH11357.

References

  1. L. E. Cross, Ferroelectrics, 1994, 151, 305–320 CrossRef CAS.
  2. T. R. Shrout and S. J. Zhang, J. Electroceram., 2007, 19, 113–126 CrossRef.
  3. 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., 2018, 17, 349–354 CrossRef CAS PubMed.
  4. F. Li, M. J. Cabral, B. Xu, Z. Cheng, E. C. Dickey, J. M. LeBeau, J. Wang, J. Luo, S. Taylor, W. Hackenberger, L. Bellaiche, Z. Xu, L.-Q. Chen, T. R. Shrout and S. Zhang, Science, 2019, 364, 264–268 CrossRef CAS PubMed.
  5. S. P. Singh, A. K. Singh, D. Pandey, H. Sharma and O. Parkash, J. Mater. Res., 2003, 18, 2677–2687 CrossRef CAS.
  6. S. P. Singh, D. Pandey, S. Yoon, S. Baik and N. Shin, Appl. Phys. Lett., 2008, 93, 182910 CrossRef.
  7. S. P. Singh, S. M. Yusuf, S. Yoon, S. Baik, N. Shin and D. Pandey, Acta Mater., 2010, 58, 5381–5392 CrossRef CAS.
  8. J. Fu, R. Zuo and X. Gao, Appl. Phys. Lett., 2013, 103, 182907 CrossRef.
  9. H. Liu, J. Chen, L. L. Fan, Y. Ren, L. Hu, F. M. Guo, J. X. Deng and X. R. Xing, Chem. Mater., 2017, 29, 5767–5771 CrossRef CAS.
  10. S.-B. Lee, K.-H. Lee and H. Kim, Jpn. J. Appl. Phys., 2002, 41, 5266 CrossRef CAS.
  11. Y. Jia, H. Luo, S. W. Or, Y. Wang and H. L. W. Chan, J. Appl. Phys., 2009, 105, 124109 CrossRef.
  12. B. Fang, Y. Shan, K. Tezuka and H. Imoto, Jpn. J. Appl. Phys., 2005, 44, 5035 CrossRef CAS.
  13. B. Fang, Y. Shan, K. Tezuka and H. Imoto, Phys. Status Solidi A, 2005, 202, 481–489 CrossRef CAS.
  14. N. Kumar, A. Ghosh and R. N. P. Choudhary, Mater. Chem. Phys., 2011, 130, 381–386 CrossRef CAS.
  15. E. I. Sitalo, I. P. Raevski, A. G. Lutokhin, A. V. Blazhevich, S. P. Kubrin, S. I. Raevskaya, Y. N. Zakharov, M. A. Malitskaya, V. V. Titov and I. N. Zakharchenko, IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 2011, 58, 1914–1919 Search PubMed.
  16. E. F. Alberta and A. S. Bhalla, Mater. Lett., 1996, 29, 127–129 CrossRef.
  17. Y. Hosono, Y. Yamashita, H. Sakamoto and N. Ichinose, Jpn. J. Appl. Phys., 2003, 42, 535–538 CrossRef CAS.
  18. E. F. Alberta and A. S. Bhalla, J. Korean Phys. Soc., 1998, 32, S1265–S1267 CAS.
  19. C. Augier, M. PhamThi, H. Dammak and P. Gaucher, J. Eur. Ceram. Soc., 2005, 25, 2429–2432 CrossRef CAS.
  20. D. Wang, M. Cao and S. Zhang, J. Eur. Ceram. Soc., 2012, 32, 433–439 CrossRef CAS.
  21. X. Qi, E. Sun, J. Wang, R. Zhang, Y. Bin and W. Cao, Ceram. Int., 2016, 42, 15332–15337 CrossRef CAS.
  22. Y. Hou, M. Zhu, F. Gao, H. Wang, B. Wang, H. Yan and C. Tian, J. Am. Ceram. Soc., 2004, 87, 847–850 CrossRef CAS.
  23. L. Fan, J. Chen, Y. Ren, Z. Pan, L. Zhang and X. Xing, Phys. Rev. Lett., 2016, 116, 027601 CrossRef PubMed.
  24. H. Liu, J. Chen, H. Huang, L. Fan, Y. Ren, Z. Pan, J. Deng, L. Q. Chen and X. Xing, Phys. Rev. Lett., 2018, 120, 055501 CrossRef CAS PubMed.
  25. Z. Xia and Q. Li, Acta Mater., 2007, 55, 6176–6181 CrossRef CAS.
  26. N. Luo, Q. Li and Z. Xia, Mater. Res. Bull., 2011, 46, 1333–1339 CrossRef CAS.
  27. D. Pang, C. He, S. Han, S. Pan, X. Long and H. Tailor, J. Eur. Ceram. Soc., 2015, 35, 2033–2040 CrossRef CAS.
  28. D. Pang, C. He and X. Long, Ceram. Int., 2016, 42, 19433–19436 CrossRef CAS.
  29. J. Wu, Y. Chang, B. Yang, X. Wang, S. Zhang, Y. Sun, X. Qi, J. Wang and W. Cao, Ceram. Int., 2016, 42, 7223–7229 CrossRef CAS.
  30. N. Luo, Q. Li, Z. Xia, X. Chu and S. Zhang, J. Am. Ceram. Soc., 2012, 95, 3246–3253 CrossRef CAS.
  31. S. W. Choi, R. T. R. Shrout, S. J. Jang and A. S. Bhalla, Ferroelectrics, 1989, 100, 29–38 CrossRef CAS.
  32. H. Liu, J. Chen, L. Fan, Y. Ren, Z. Pan, K. V. Lalitha, J. Rodel and X. Xing, Phys. Rev. Lett., 2017, 119, 017601 CrossRef PubMed.
  33. D. Damjanovic, J. Am. Ceram. Soc., 2005, 88, 2663–2676 CrossRef CAS.
  34. J. Chen, K. Nittala, J. S. Forrester, J. L. Jones, J. Deng, R. Yu and X. Xing, J. Am. Chem. Soc., 2011, 133, 11114–11117 CrossRef CAS PubMed.
  35. H. Kungl and M. J. Hoffmann, Acta Mater., 2007, 55, 5780–5791 CrossRef CAS.
  36. T. Leist, J. Chen, W. Jo, E. Aulbach, J. Suffner, J. Rödel and S. Zhang, J. Am. Ceram. Soc., 2012, 95, 711–715 CrossRef CAS.
  37. A. Bokov, L. Shpak and I. Rayevsky, J. Phys. Chem. Solids, 1993, 54, 495–498 CrossRef CAS.
  38. T. R. Shrout, Z. P. Chang, N. Kim and S. Markgraf, Ferroelectr., Lett. Sect., 1990, 12, 63–69 CrossRef CAS.

This journal is © the Partner Organisations 2019