Piezoelectricity and excellent temperature stability in nonferroelectric Bi₁₂TiO₂₀–CaTiO₃ polar composite ceramics

Abstract

Nonferroelectric Bi₁₂TiO₂₀–CaTiO₃ (BT–CT) composite ceramics were prepared through an interfacial reaction between presynthesized crystalline CaTiO₃ and crystalline Bi₁₂TiO₂₀ phases at various sintering temperatures. After sintering at temperatures above the melting point of Bi₁₂TiO₂₀, both direct and converse piezoelectric effects were observed in these composites for the first time. Because neither CaTiO₃ nor Bi₁₂TiO₂₀ is ferroelectric and because no obvious crystallographic orientation was found in these sintered composites, the temperature gradient-driven plastic flexoelectricity of the grain boundary amorphous phases might be the main poling mechanism. In this work, the highest d₃₃ value of 8 pC N⁻¹ was obtained in the samples sintered at 860 °C, which were found to contain a large amount of amorphous Bi₁₂TiO₂₀ and to possess the lowest density. For these BT–CT polar composite ceramics, the piezoelectric activity, the dielectric loss (tg δ ≈ 0.1%), the mechanical quality factor (Q_m ≈ 2300), the depoling temperature (T_d ≈ 880 °C) and the temperature stability of the resonance frequency are all comparable to those of the well-known bismuth layer-structured ferroelectrics, which indicates that these new polar composite ceramics are promising candidates for high-temperature piezoelectric applications.

---

Introduction

Traditional piezoelectric ceramics, such as Pb(Zr, Ti)O₃ and BaTiO₃ ceramics, are ferroelectric materials. In general, these ferroelectric piezoelectric materials all undergo an electrical poling process, only after which can they possess macroscopic polarization and exhibit piezoelectricity due to the orientation of ferroelectric domains.¹

In fact, nonferroelectric ceramics that contain no ferroelectric components can also be used as piezoelectric materials because of the presence of flexoelectricity.^2–6 As known, flexoelectricity, which describes the coupling between electric polarization and strain gradients, is a universal effect that is allowed by symmetry in all insulating and semi-conducting materials. For bulk homogeneous inorganic solids, the macroscopic flexoelectric polarization is generally very small.⁷ Thus, previous studies of the flexoelectric effect in solid materials primarily focused on low-dimensional systems, in which a large strain gradient can be easily generated and even sustained. These materials include thin films,^8,9 superlattices,¹⁰ nanocomposites^3,5,6 and other special nanostructures,^11–13 and their excellent ferroelectric or piezoelectric properties indicate the potential for utilizing flexoelectricity in the design of new artificial polar materials.¹⁴

We recently reported a special class of bulk nonferroelectric polar composite ceramics, including Bi₁₂TiO₂₀–SrTiO₃ and Bi₁₂TiO₂₀–BaSnO₃, which can exhibit both direct and inverse piezoelectric effects.^15–19 These sintered nonferroelectric composites do not contain obviously oriented grains, and their piezoelectricity also does not depend on ferroelectricity; thus, they are very likely to possess a frozen-in macroscopic flexoelectric polarization that is generated and then sustained during the sintering process. Based on the previous experimental results and inspired by the random network of local bonding units (RN-LBU) theory of quasi-amorphous films, we hypothesize that this poling effect might also be classified as plastic flexoelectricity and that macroscopic polarization should mainly be influenced by the plastic strain gradient-induced partial alignment of grain boundary amorphous phases. Due to the lack of a ferroelectric phase transition, these nonferroelectric systems generally exhibit more temperature-stable dielectric and piezoelectric behavior. However, the piezoelectricity in these materials is relatively weak. Improving the piezoelectric properties in nonferroelectric ceramics is the key issue that should be addressed for practical applications. According to our previous study, composition is of importance to the piezoelectric properties. Furthermore, the sintering temperature might also have an effect on the interfacial interaction, which would further influence the piezoelectric properties. Therefore, it is necessary to investigate the effect of the sintering temperature and composition on the structures and properties of this type of ceramic.

In this work, the dielectric and piezoelectric properties of novel BT–CT composite ceramics sintered at different temperatures were investigated, and a relationship between the piezoelectric properties and density was present. In addition, the thermal-depoling behavior and temperature stabilities of the ceramics were also evaluated.

Experimental

BT–CT composite ceramics, which have a nominal composition of Bi₁₂TiO₂₀–12CaTiO₃, were fabricated via an interfacial reaction between presynthesized Bi₁₂TiO₂₀ and CaTiO₃ phases. High-purity Bi₂O₃, TiO₂ and CaCO₃ powders were used as the starting materials. First, Bi₁₂TiO₂₀ powder and CaTiO₃ powder were presynthesized via a conventional solid-state reaction; the calcining temperatures of Bi₁₂TiO₂₀ and CaTiO₃ were 750 °C and 1150 °C, respectively. Similar to the composite ceramics that were reported in our previous work, the calcined Bi₁₂TiO₂₀ and the calcined CaTiO₃ were both completely reacted and well crystallized, and no impurity phases could be detected.¹⁹ Subsequently, the Bi₁₂TiO₂₀ and CaTiO₃ powders were mixed, ball-milled, and then pressed into disks with a diameter of 15 mm and thickness of 1 mm. Finally, the pressed BT–CT composites were placed on an Al₂O₃ ceramic plate, sintered in air atmosphere for 2 h and then naturally cooled to room temperature. The sintering temperature was in the range of 845 to 870 °C.

The constituent crystalline phases and the microstructures of the sintered composites were studied using an X-ray diffractometer (XRD) (Bruker AXS D8 Advance) and a high-resolution scanning electron microscope (SEM) (FEI Nova NanoSEM 450). The local structures and chemical bond environments were analyzed using Raman spectroscopy (Horiba Jobin-Yvon, LabRAM HR-800). Differential scanning calorimetry (DSC) measurements were performed using a TA SDT Q600 instrument. The piezoelectric constant d₃₃ was measured using a YE2730A d₃₃ meter. The dielectric and piezoelectric resonance properties were measured using an Agilent 4294A impedance analyzer, and the frequency constants were determined using the resonance method according to the IEEE standard on piezoelectricity.

Results and discussion

Fig. 1(a) presents the XRD patterns of the inner layer of the BT–CT ceramics sintered at 845 °C (abbreviated as BT–CT845) and at 860 °C (abbreviated as BT–CT860). Note that the unsintered BT–CT composites were well crystallized and that no impurity phase could be detected. After sintering at 845 °C and 860 °C, respectively, and then naturally cooling to room temperature, the Bi₁₂TiO₂₀ and CaTiO₃ crystalline phases were still present, and no new crystalline phases were found.

Fig. 1 XRD patterns of BT–CT845 and BT–CT860 composite ceramics.

To study the detailed contrast of the XRD patterns, Fig. 1(b) presents an enlargement of the typical baseline of the XRD pattern. Compared with the well-crystallized BT–CT845 ceramic, the amorphous phase of the BT–CT860 sample was confirmed by the broad diffusion halo.²⁰ Comparing the areas under the well-defined diffraction peaks and their amorphous holes with the total peak area, the weight fraction of amorphous content in the BT–CT860 sample is estimated to be approximately 20%.²¹ In addition, more detailed contrast of the XRD peaks is observed in the insets, which show an expanded view in the vicinity of the sillenite (321) peak and the perovskite (040)/(202) peaks as a representative. Both Bi₁₂TiO₂₀ peaks and CaTiO₃ peaks of the BT–CT860 sample slightly shifted toward higher diffraction angles compared with the BT–CT845 sample, which implies a little decrease in the lattice constant.²² Moreover, the XRD peaks of the Bi₁₂TiO₂₀ crystalline phase for BT–CT860 were observed to be broadened and weakened compared to those for BT–CT845, and no obvious change was found for the CaTiO₃ phase. This result suggests that the amorphous phase in BT–CT860 may correspond to the Bi₁₂TiO₂₀ phase.

Although the BT–CT ceramics contain no ferroelectric phases and no oriented grains, they can still exhibit detectable piezoelectricity. The piezoelectric properties were studied using the same physical coordinate system described in our previous work.¹⁷ For the BT–CT system, the sintering temperature was changed within a narrow range from 845 °C to 870 °C. Fig. 2 compares the changes in the obtained piezoelectric constants and density of the BT–CT composite ceramics as a function of the sintering temperature. The largest density was found in the BT–CT845 sample, in which no piezoelectricity was detected. In contrast, the highest piezoelectric constant value of 8 pC N⁻¹ was obtained in the BT–CT860 sample, which has the lowest density. We attempted to study the sample sintered at higher temperatures; however, we failed to obtain dense samples due to excessive volatilization of the Bi₁₂TiO₂₀ during the sintering process.²³

Fig. 2 (a) Density change and (b) piezoelectric constant d₃₃ of Bi₁₂TiO₂₀–CaTiO₃ samples as a function of the sintering temperature.

Fig. 3 shows the dielectric constant and the piezoelectric resonance frequency for the BT–CT860 composite ceramic as functions of temperature, and the inset of Fig. 3(b) presents the details of the piezoelectric resonance frequency at 550 °C. It can be observed that the dielectric behavior of the ceramic is similar to that of the ordinary CaTiO₃-based incipient ferroelectric.²⁴ The dielectric constant changes slowly in the temperature range without discontinuities or abrupt slope changes. This result indicates that the BT–CT860 sample may exhibit good temperature stability, which was confirmed by Fig. 3(b). The resonance frequency decreased very slightly with increasing temperature. The temperature coefficients of the resonance frequency (T_fr) are always less than 100 × 10⁻⁶/°C in the temperature region from room temperature to 300 °C and less than 160 × 10⁻⁶/°C over the entire temperature range from room temperature to 700 °C, which is a very attractive characteristic for these special polar materials. The room temperature properties of the BT–CT composite ceramic are summarized in Table 1. At room temperature, the dielectric permittivity (ε₃₃/ε₀) was determined to be approximately 110, and the dielectric loss tangent, tg δ, is only 0.1%. The piezoelectric coefficients (d₃₃), electromechanical coupling factor (k) and mechanical quality factors (Q_m) were found to be comparable to those of some bismuth layer-structured ferroelectric ceramics.²⁵

Fig. 3 Temperature dependence of the dielectric constant and resonance frequency of the BT–CT860 composite ceramic.

Table 1 Dielectric and electromechanical properties of the BT–CT860 composite ceramic

ε33/ε0 (1 kHz)	tg δ	d33 (pC N^−1)	k	Qm
110	0.1%	8	3%	2300

---

Fig. 4 presents the Raman spectra of the inside layers of the BT–CT845 and BT–CT860 composite ceramics. The Raman peaks of the BT–CT ceramics can be divided into two groups: the Bi₁₂TiO₂₀-related peaks labeled p1–p11 (ref. 26) and the CaTiO₃-related peaks labeled p12–p15.²⁸ The vibrational modes of Bi₁₂TiO₂₀ can be classified as TiO₄ tetrahedra modes appearing above 700 cm⁻¹ and the low-frequency BiO₅ polyhedra modes.^26,27 Compared to the Raman peaks of BT–CT845 ceramics, the positions and full widths at half maximum of the CaTiO₃-related peaks do not obviously change. In contrast, the p1, p2 and p6 peaks in the BT–CT860 sample weaken and broaden, and even shift toward lower wavenumbers, compared with those in the BT–CT845 sample. The weakening and broadening of Bi₁₂TiO₂₀ low wavenumber peaks can further support the amorphization of the main Bi₁₂TiO₂₀ crystalline lattice. Because these peaks are related to Bi–O bond vibrations of BiO₅ polyhedra, the peak shift implies the appearance of asymmetric Bi–O bond elongations.²⁸ Note that the elongations of the Bi₁₂TiO₂₀ phase are exactly opposite to that of XRD patterns, which show contraction of the crystalline phase. As known, Raman spectroscopy is a more sensitive probe of structural distortions, short-range order, and symmetry in solids than XRD, and thus, it is hypothesized that the Raman expansion may be a reflection of the amorphous Bi₁₂TiO₂₀ phase.

Fig. 4 Raman spectra of the BT–CT845 and BT–CT860 composite ceramics.

According to the combined results of XRD and Raman studies, the BT–CT860 composite ceramics should be mainly composed of CaTiO₃ grains, Bi₁₂TiO₂₀ grains and a certain amount of amorphous Bi₁₂TiO₂₀. The SEM image (Fig. 5) and EDS analysis show that CaTiO₃ grains can be easily identified and their sizes are generally less than 1 μm. However, the Bi₁₂TiO₂₀ grains and amorphous Bi₁₂TiO₂₀ are difficult to be distinguished because the Bi₁₂TiO₂₀ grains seem to have irregular morphology and the grain boundaries are also very fuzzy. Fig. 6 shows the DSC traces of the BT–CT860 composite ceramic and the Bi₁₂TiO₂₀ ceramic. The Bi₁₂TiO₂₀ ceramic has a strong endothermic peak at approximately 860 °C, with a calculated enthalpy (ΔH₁) of approximately 282.5 kJ mol⁻¹. In contrast, the BT–CT composite ceramic has a smaller endothermic peak at approximately 865 °C, with a calculated enthalpy (ΔH₂) of approximately 114.1 kJ mol⁻¹ (ΔH₂). The endothermic peaks in the above two samples are both melting peaks of the Bi₁₂TiO₂₀ crystalline phase, and the slight increase in the melting point may be due to the small compressed distortion of the crystalline Bi₁₂TiO₂₀ phase in the BT–CT composite ceramic.²⁹ Note that by using the ratio of ΔH₂/ωΔH₁ (ω is the weight fraction of Bi₁₂TiO₂₀ in the composites and assuming that the Bi₁₂TiO₂₀ ceramic was 100% crystalline), we can roughly estimate that approximately 63% of Bi₁₂TiO₂₀ crystalline phase is formed after the sintering process. This estimation approximately corresponds to a 28% volume fraction of crystalline Bi₁₂TiO₂₀ in the BT–CT composite ceramics.

Fig. 5 The SEM images for cross-section of the BT–CT860 composite ceramics.

Fig. 6 DSC curves of the BT–CT860 composite ceramic and the Bi₁₂TiO₂₀ ceramic.

Thermal depoling of the BT–CT860 sample is shown in Fig. 7, in which the piezoelectric constants are plotted against the annealing temperature. The thermal depoling experiment was conducted by first holding the sample at 500 °C in atmosphere for 2 hours, cooling to room temperature, measuring the d₃₃ value, and then repeating this procedure at intervals up to 880 °C. It is clear that the piezoelectricity in this type of material is very stable. The piezoelectric constant remains unchanged from their room temperature values up to 860 °C. This result demonstrates that this type of composite ceramic is stable against thermal exposure and can be used in high-temperature piezoelectric applications. As the annealing temperature increased to 880 °C, which is above the melting point of Bi₁₂TiO₂₀, the piezoelectric constant began to decrease. The results reveal that Bi₁₂TiO₂₀ must play a decisive role in determining in the depoling temperature in this type of polar material.

Fig. 7 Effect of the annealing temperature on the piezoelectricity of the BT–CT860 composite.

In the present work, the furnace provided a temperature gradient of approximately 1 °C mm⁻¹ along the thickness direction of the samples during the sintering process. Because the composition contains no ferroelectric phases and no obviously oriented grains, their poling may be due to the temperature gradient-driven plastic flexoelectricity of the grain boundary amorphous phases.³⁰ The larger the temperature gradient, the larger the strain gradient and the more probable the formation of the polar amorphous phase.⁸ Therefore, it might be possible to further increase the piezoelectricity by increasing the temperature gradient during the sintering process. In the BT–CT polar composite ceramics, amorphous CaTiO₃ is very limited; thus, the macroscopic polarization is most likely to be mainly influenced by the amorphous Bi₁₂TiO₂₀. When the BT–CT composites are sintered below the melting temperature of Bi₁₂TiO₂₀, the sintering process is the process of densification which is identical to the traditional ceramics. As the sintering temperature enhanced above the melting point of Bi₁₂TiO₂₀, the main role of the crystalline CaTiO₃ may be to suppress recrystallization and volatilization of the melting Bi₁₂TiO₂₀. The appearance of the amorphous Bi₁₂TiO₂₀ phase generally has lower density than the crystalline Bi₁₂TiO₂₀, thus, the density of the sample is reduced.^31–33 It appears that the sintering temperature can greatly affect the interfacial interaction between CaTiO₃ and Bi₁₂TiO₂₀ and further affect the volume fraction of the amorphous Bi₁₂TiO₂₀ phase. For the BT–CT860 sample, the lowest density indicates the highest content of amorphous Bi₁₂TiO₂₀;^31–33 this may be the main reason that the highest piezoelectric constant can be obtained.

Conclusions

In conclusion, nonferroelectric BT–CT polar composite ceramics that contain a large amount of amorphous Bi₁₂TiO₂₀ and have relatively low densities were fabricated via an interfacial interaction between presynthesized crystalline CaTiO₃ and crystalline Bi₁₂TiO₂₀ phases. Because neither CaTiO₃ nor Bi₁₂TiO₂₀ is ferroelectric and because no obvious crystallographic orientation was found in these ceramics, temperature gradient-driven plastic flexoelectricity of the grain boundary amorphous phases may be the main poling mechanism. The BT–CT polar composite ceramics exhibit relatively strong piezoelectricity (d₃₃ = 8 pC N⁻¹), very low dielectric loss (tg δ ≈ 0.1%), high mechanical quality factor (Q_m ≈ 2300), high depoling temperature (T_d ≈ 880 °C) and excellent temperature stability of the resonance frequency, thus making them promising candidates for high-temperature piezoelectric applications.

Acknowledgements

This work was supported by the National Nature Science Foundation of China (No. 51172129, 11374186 and 51231007), and the Natural Science Foundation of Shandong Province (No. ZR2013EMM018 and ZR2014EMM012).

References

1. G. H. Haertling, J. Am. Ceram. Soc., 1999, 82, 797.
2. A. K. Tagantsev, Phys. Rev. B: Condens. Matter Mater. Phys., 1986, 34, 5883.
3. J. Fousek, L. E. Cross and D. B. Litvin, Mater. Lett., 1999, 39, 287.
4. W. Ma and L. E. Cross, Appl. Phys. Lett., 2002, 81, 3440.
5. L. E. Cross, J. Mater. Sci., 2006, 41, 53.
6. N. D. Sharma, R. Maranganti and P. Sharma, J. Mech. Phys. Solids, 2007, 55, 2328.
7. P. Zubko, G. Catalan, A. Buckley, P. R. L. Welche and J. F. Scott, Phys. Rev. Lett., 2007, 99, 167601.
8. E. Wachtel and I. Lubomirsky, Adv. Mater., 2010, 22, 2485.
9. D. Lee, A. Yoon, S. Y. Jang, J.-G. Yoon, J.-S. Chung, M. Kim, J. F. Scott and T. W. Noh, Phys. Rev. Lett., 2011, 107, 057602.
10. N. D. Sharma, C. M. Landis and P. Sharma, J. Appl. Phys., 2010, 108, 024304.
11. I. Naumov, A. M. Bratkovsky and V. Ranjan, Phys. Rev. Lett., 2009, 102, 217601.
12. M. S. Majdoub, P. Sharma and T. Cagin, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 77, 125424.
13. R. Nath, S. Zhong, S. P. Alpay, B. D. Huey and M. W. Cole, Appl. Phys. Lett., 2008, 92, 012916.
14. R. Mbarki, N. Baccam, K. Dayal and P. Sharma, Appl. Phys. Lett., 2014, 104, 122904.
15. M. L. Zhao, L. H. Wang, C. L. Wang, J. L. Zhang, Z. G. Gai, C. M. Wang and J. C. Li, Appl. Phys. Lett., 2009, 95, 022904.
16. M. L. Zhao, L. H. Wang, C. L. Wang, J. L. Zhang, Z. G. Gai, C. M. Wang and J. C. Li, Scr. Mater., 2010, 63, 207.
17. L. H. Wang, M. L. Zhao, C. L. Wang, J. Wang, W. J. Kuai and X. T. Tao, Appl. Phys. Lett., 2012, 101, 062903.
18. W. J. Kuai, M. L. Zhao, C. L. Wang, L. H. Wang, J. Wang, Z. G. Gai, J. L. Zhang and N. Yin, J. Alloys Compd., 2013, 559, 76.
19. D. Yu, M. Zhao, C. Wang, L. Wang, J. Wang, Z. Gai and J. Zhang, Phys. Status Solidi B, 2015, 252, 2174.
20. M. Tiainen, J. Daavitsainen and R. S. Laitinen, Energy Fuels, 2002, 16, 871.
21. V. R. Ouhadi and R. N. Yong, Appl. Clay Sci., 2003, 23, 141.
22. T. Furuta, S. Kuramoto, J. Hwang, K. Nishino and T. Saito, Mater. Sci., 2005, 46, 3001.
23. J. Wong, J. Appl. Phys., 1980, 51, 4453.
24. V. V. Lemanov, A. V. Sotnikov, E. P. Smirnova, M. Weihnacht and R. Kunze, Solid State Commun., 1999, 110, 611.
25. T. Takenaka and H. Nagata, J. Eur. Ceram. Soc., 2005, 25, 2693.
26. B. Mihailova, G. Bogachev, V. Marinova and L. Konstantinov, J. Phys. Chem. Solids, 1999, 60, 1829.
27. L. Wiehl, A. Friedrich, E. Haussuhl, W. Morgenroth, A. Grzechnik, K. Friese, B. Winkler, K. Refson and V. Milman, J. Phys.: Condens. Matter, 2010, 22, 505401.
28. H. Zheng, G. D. C. Csete, R. Quimby, H. Bagshaw, R. Ubic, I. M. Reaney and J. Yarwood, J. Eur. Ceram. Soc., 2003, 23, 2653.
29. Y. Zhu, C. Lishan and Z. Yanjun, Front. Mater. Sci. China, 2007, 1, 356.
30. P. V. Yudin and A. K. Tagantsev, Nanotechnology, 2013, 24, 432001.
31. R. Zallen, The physics of amorphous solids, Wiley, New York, 1983.
32. W. D. Kingery, H. K. Bowen and D. R. Uhlmann, Introduction to ceramics, Wiley, New York, 1976.
33. I. Ebralidze, V. Lyahovitskaya, I. Zon, E. Wachtel and I. Lubomirsky, J. Mater. Chem., 2005, 15, 4258.

---