Colossal dielectric performance of pure barium titanate ceramics consolidated by spark plasma sintering

Abstract

A facile sol–gel-hydrothermal method was developed to prepare well-dispersed BaTiO₃ (BT) nanocrystals with a size of about 20 nm. By using such nanopowder, dense BT nanoceramics were densified via spark plasma sintering (SPS). Dielectric measurements reveal that pure BT nanoceramics have an extremely high permittivity up to 6 × 10⁴ and a low dielectric loss. On the basis of microstructure characterization and dielectric measurement, it was proposed that the polaron dipoles configured by oxygen vacancies and Ti³⁺ cations within the grains could contribute to the colossal dielectric permittivity of the ceramics. By using the thermally activated polaron hopping model, a large activation energy E_A has been determined (0.135 eV) below the Curie temperature, possibly due to the coupling of polaron dipoles and intrinsic ferroelectric dipoles. This simple method offers the possibility to produce pure BT nanoceramics with a colossal permittivity.

---

Introduction

Owing to a high dielectric constant, ferroelectric barium titanate (BaTiO₃, BT) has found many applications in various electronic devices, highly attractive for the usage in multilayer ceramic capacitors (MLCC).¹ As the size of device is reduced, there is a growing demand for BT nanoceramics with a high dielectric constant. For example, the thickness of MLCC layers has been reduced to around 1 μm recently, and it is highly desirable to achieve several hundred nanometers.² In recent years, there has been an increasing interest in developing BT ceramics with an average grain size well below 100 nm. However, it remains a great challenge to prepare BT ceramics with enhanced performance for MLCCs, and the difficulty arises from the fact that the ferroelectric and dielectric properties suffer degradation when the size of particles is reduced into nanoscale.

To obtain high-performance BT nanoceramics, the key step is to prepare BT powders (the raw material) with controlled stoichiometry, uniformity and fine particle size. Currently, there are two prevailing methods to synthesize BT powders, i.e., solid phase^3–6 and liquid phase methods.^7–12 Traditional solid phase method requires high sintering temperatures that may cause serious agglomeration and coarse grains. The liquid phase routes include co-precipitation,^7,8 microemulsion⁹ and hydrothermal methods.^10–12 Among them, the first two methods still require a high sintering temperature, which will also result in large particle size. In contrast, the hydrothermal process combined with sol–gel method has been demonstrated to be an effective way to synthesize high purity, uniform, ultrafine, well-dispersed BT powder at low temperature, providing a prerequisite for preparing BT nanoceramics.^13,14

Also, sintering is a critical processing step for obtaining dense ceramics with high performance. BT ceramics consolidated by conventional sintering approaches always have a low relative permittivity, with a range from 4000 to 6000.^15,16 In comparison, spark plasma sintering (SPS) can produce well densified ceramics with fine microstructure due to a high heating rate, lower sintering temperature and shorter dwell time needed.^17–21 The relative permittivity of SPSed BT ceramics is up to 3000–8000.^22,23 To improve the permittivity of BT ceramics, several methods were developed, including the substitution of a small amount of lanthanum in the perovskite lattice,^24,25 and fabricating core–shell architecture or a metallic layers on BT grain.^26–29 However, the main drawback of them is the complexity and uncontrollability of the process that always leads to a large fluctuation in the overall performance of the material. Until now, pure-phase BT ceramics with colossal permittivity have been rarely reported.

In this work, well-dispersed BT nanoparticles with amorphous layer and a size of 10–40 nm were synthesized by a sol–gel-hydrothermal method. By using such nanopowders, dense BT nanoceramics were consolidated through SPS. The obtained BT ceramics showed extremely high permittivity up to 6 × 10⁴, with relatively low dielectric below 0.058. Temperature dependence of dielectric properties and loss tangent at different frequencies were evaluated. A mechanism, based on the thermally activated hopping polaron dipoles and their coupling with intrinsic ferroelectric dipoles was proposed to explain the colossal dielectric permittivity.

Experimental procedures

The chemical reagents used in the work were barium acetate [Ba(CH₃COO)₂] (Shanghai Silian Chemical Co., Ltd.), tetrabutyl titanate [Ti(OC₄H₉)₄] (Sinopharm Chemical Reagent Co., Ltd.), nitric acid (HNO₃) (Zhejiang Zhongxing Chemical Reagent Co., Ltd.) and potassium hydroxide (KOH) (Sinopharm Chemical Reagent Co., Ltd.). All the chemicals have analytical grade purity and were used in their original forms without further purification. To synthesize BT nanoparticles, tetrabutyl titanate (0.005 mol), nitric acid (0.004 mol), hydrolysis inhibitor (0.002 mol) and deionized water (25 ml) were mixed together. The mixture was then incubated in a water bath (60 °C) with constantly vigorous stirring for 10 h and a homogeneous sol was formed. The obtained gel and barium acetate were used as the precursors to introduce the 0.015 mol Ba and 0.005 mol Ti cations into the solution. Potassium hydroxide was added as a mineralizer. The concentration of the KOH solution was 8 M. The hydrothermal treatment was performed in an autoclave at 200 °C for 6 h. The products were filtered and washed several times with distilled water and absolute ethanol and finally oven-dried in air at 80 °C for 12 h to obtain white powders for further characterization.

Consolidation of the BaTiO₃ nanopowders was carried out in a SPS apparatus (Dr Sinter 2050, Sumitomo Coal Mining Co., Tokyo, Japan). In a typical cycle, BaTiO₃ nanopowder with a mass of 1.1 g was loaded in a cylindrical graphite die with an inner diameter of 12 mm. Four sintering regimes were performed, referred as BT-1, BT-2, BT-3 and BT-4, respectively. The temperature was automatically raised to 650 °C (BT-1) or 600 °C (BT-2, BT-3, and BT-4) over a period of 6 min (BT-1) or 3 min (BT-2, BT-3, and BT-4). Then the samples were heated to different sintering temperatures (950–1100 °C) with a heating rate of 50 °C min⁻¹ and dwell time of 3 min and naturally cooled down to room temperature. A radiation pyrometer, focused on the bottom of a deep hole drilled on the wall of the graphite die that is only 2 mm away from the center of the powder compacts, was employed for monitoring the temperature. For all the experiments, a uniaxial pressure of 40–100 MPa was applied at room temperature and held constant from the beginning till the end of the sintering cycles. Finally, all the samples were annealed in air at 550 °C for 5 h in order to remove the residual carbon on the surfaces. The densitometer was employed to measure the relative density of BT ceramics (GH-120D, Xiamen qunlong Instrument Co. Ltd., China). The heating rate, sintering temperature mechanical pressure values and relative density are listed in Table 1.

Table 1 The sintering parameters and sintered relative densities of BaTiO₃ ceramics

Sample name	BT-1	BT-2	BT-3	BT-4
Heating rate (°C min^−1)	50	50	50	50
Sintering T (°C)	950	1100	1050	1050
Pressure (MPa)	40	50	75	100
Relative density	95.4%	94.3%	96.8%	96.4%

---

Crystalline phase composition was determined by an X-ray diffractometer (XRD, D/max-RA, Rigaku Corporation, Japan) using CuKα radiation (λ = 1.5418 Å) over a 2θ range of 10–90° at room temperature. Microstructural investigations were conducted by using a scanning electron microscope (SEM, MODEL S-480, Hitachi, Japan) and a transmission electron microscope (TEM, F20, FEI, USA) with an accelerating voltage of 200 kV. Raman spectra were obtained by using a micro-Raman spectrometer (inVia, Renishaw, UK) with a Nd:YAG laser (532 nm). X-ray photoelectron spectroscopy (XPS) was carried out by Thermo ESCALAB 250Xi using a monochromatized Al Kα (hν = 1486.6 eV, 150 W) radiation. Scanning transmission electron microscopy (STEM) and electron energy loss spectroscopy (EELS) were obtained by using spherical aberration correction transmission electron microscope (FEI Titan G² 80-200 ChemiSTEM, FEI, USA). The BT-1 sample was polished by precision ion polishing system (PIPS, II 695, GATAN, USA).

The frequency spectra of complex permittivity of the materials were measured by a precision impedance analyzer (Agilent 4294 A) over the frequency range from 40 Hz to 110 MHz. Prior to it, the samples were well-polished and Ag electrodes were formed on both surfaces from Ag paint by heat-treated at 550 °C for 10 min.

Results and discussions

The XRD pattern of the as-synthesized powder is shown in Fig. 1a. All the diffraction peaks can be indexed to cubic BT (JCPDS 31-0174). No peaks from other phases are detected, implying a high purity of the single-phase BT powder. Fig. 1b shows a XRD pattern selected from literature³⁰ as a comparison with that of BT powder in this work. On the basis of result in Fig. 1, the phase reported here agrees well with that of BaTiO₃ powder in the literature.³⁰ Fig. 1c shows SEM image of the BT powder. It can be observed that the powder consists of well-dispersed BT nanoparticles with a spherical morphology and a size of 10–40 nm. Fig. 1d and e are the HRTEM images of the BT nanoparticle. An amorphous layer can be observed clearly on the surface of the nanoparticle. In contrast, within the nanoparticle, the regular interplanar spacing of the observed lattice is 0.276 nm, as shown in Fig. 1e which is an enlarged image of the rectangular area in Fig. 1d. Index of this spacing confirmed the (110) plane of the cubic BT.

Fig. 1 Characterizations of the as-prepared BaTiO₃ nanoparticles by a sol–gel-hydrothermal method, (a) XRD pattern, (c) SEM image, (d) and (e) HRTEM images of the BaTiO₃ nanoparticles, (b) XRD pattern of the as-prepared BaTiO₃ sample synthesized by the hydrothermal process at 200 °C for 6 h with 0.1 M KOH.³⁰

The XRD patterns of our samples and standard PDF card of BT are together shown in Fig. 2. All peaks of the patterns can be indexed into tetragonal (JCPDF 05-0626) and cubic (JCPDF 31-0174) perovskite BT without any impurity. As shown in Fig. 2b, the enlarged peaks at 45–46° suggest a pseudo-cubic perovskite structure of the ceramics without the characteristic splitting of the tetragonal BT. This is due to the existence of a small spontaneous lattice strain in the ceramics in addition to the broadening effect related to the small grain size.³¹

Fig. 2 XRD patterns of (a) the as-sintered BaTiO₃ ceramics and (the tetragonal (JCPDF 05-0626) and cubic (JCPDF 05-0626) PDF card of BaTiO₃), (b) enlarged diffraction peak pattern at around 45.2°.

In order to verify the existence of tetragonal phase in BT ceramics, Raman spectroscopy was used to check the presence of Raman-active modes. As shown in Fig. 3, the Raman peaks at around 715 cm⁻¹ and 298 cm⁻¹ are identified as the characteristic peaks of tetragonal BT.^32,33 The typical sharp band at around 298 cm⁻¹ corresponds to the asymmetrical vibration mode of TiO₆ octahedra.³⁴ The peak at about 715 cm⁻¹ is related to the highest frequency longitudinal optical mode (LO) of A1 symmetry of TiO₆ octahedron, and is also an implication for the eventual Ba²⁺ defects in the BT lattice.³⁵ These results reveal the existence of tetragonal distortion in the BT ceramics, which are in good agreement with the XRD patterns.

Fig. 3 Raman spectra of the as-sintered BaTiO₃ ceramics.

Table 1 lists the relative density values of the sintered samples, and the SEM images of their fracture surfaces are shown in Fig. 4. It can be seen that BT-3 and BT-4 possessed the highest sintered densities (96.8% & 96.4%), owing to the high sintering temperature and pressure applied. The high pressure also suppressed the grain growth as their grain size are estimated to be 10–50 nm for BT-3 and 10–20 nm for BT-4, the grain size in BT-3 and BT-4 is much smaller than that the average grain size was determined to be ∼300 nm in the literature.²⁵ Sample BT-2 is the least dense one with a relative density of 94.3%. The highest sintering temperature (1100 °C) led to significantly coarser grains (80–500 nm) compared with others. The relative density of sample BT-1 is 95.4% with a grain size of 20 nm to 1 μm because of the lowest sintering temperature (950 °C). All the results show that the relative density and microstructure of the samples can be tuned by varying sintering parameters (temperature, pressure, etc.).

Fig. 4 SEM images of the fracture surfaces of the as-sintered BT ceramics. (a) BT-1, (b) BT-2, (c) BT-3, and (d) BT-4.

The dielectric properties of the BT ceramics were measured and discussed. Fig. 5 shows the frequency dependence of dielectric permittivity (ε_r) for different samples at room temperature. All samples have a good stability of the permittivity in the wide frequency range from 40 Hz to 10 kHz, and the dielectric constant at low frequency is up to 10⁴. For BT-1, BT-2, BT-3 and BT-4, the maximum values of the dielectric constant are 65 700, 38 000, 58 400 and 35 300, respectively. In comparison, the dielectric constants of these BT samples at 1 kHz are nearly 5–11 times higher than the previously reported values (3500, <1000, 5000, 3800 and 4000).^36–40 To the best of our knowledge, the present values could be the highest among the reported for the pure-phase BT ceramics.

Fig. 5 Frequency dependence of permittivity (solid line) and loss tangent (dash line) for (a) BT-1, (b) BT-2, (c) BT-3 and (d) BT-4 at room temperature.

Normally, the dielectric constants at low frequency (<1 kHz) increased with increasing relative density. However, the dielectric constant of sample BT-4 at low frequency is slightly lower than sample BT-2, which could be attributed to the smaller grain size of sample BT-4 compared with that of sample BT-2 (“size effect”).⁴¹ The results indicate that fine grain and high relative density should be beneficial for improving the dielectric constant. The relaxation frequency is about 100 kHz for BT-2 and BT-4, 300 kHz for BT-3, and 1 MHz for BT-1. The corresponding mechanism for sample BT-1 should be a typical ferroelectric dipole relaxation, whereas for other samples an additional relaxation mechanism such as defects is possible. At the same time, the corresponding dielectric losses (loss tangent, tan δ) for different samples as a function of frequency at room temperature are also shown in Fig. 5. The dielectric loss peak frequency is consistent to the relaxation frequency. The minimum losses of BT-1, BT-2, BT-3, and BT-4 are 0.0379, 0.044, 0.04 and 0.058, respectively. It should be noted that the dielectric loss of all samples in the range of 50–1 kHz gradually decreased and thus grew at high frequency downward. Such phenomenon could be toughly related to the space charge relaxation due to oxygen defects and ferroelectric polarization relaxation.

Different mechanisms have been proposed to discuss the colossal dielectric performance of BT ceramics. Generally, grain boundaries (GB), domain walls and inhomogeneity are critical factors to influence the dielectric constants of the ceramics. Oyama et al.⁴² showed that the grain boundaries will limit the diffusion of oxygen vacancies, particularly in fine-grained ceramics. Oxygen vacancies associated with Ti³⁺ cations were present in the powders/ceramics and their localizations at grain boundaries have been suggested to be the key factor to control the dielectric properties.⁴³ It is important to note that device fabrication with high-κ dielectrics favors amorphous layers.⁴⁴ Recently, Jiang et al.⁴⁵ found an increase in the dielectric constant as a function of decreasing the particle size of PbTiO₃. The reason was the presence of an amorphous surface layer and surface energy associated with domain walls.⁴⁶ In addition, the presence of an amorphous layer of SrTiO₃ prepared hydrothermally resulted in the anomalous dielectric behavior of the SrTiO₃ nanocrystals.⁴⁷ Accordingly, we assumed that the amorphous layers of the as-sintered BT nanoparticles could facilitate the rapid formation of GB with fine grains. This offers the field for the numerous oxygen vacancies to be concentrated. These oxygen vacancies can act as blocking barriers against any further oxygen diffusion. Additionally, intrinsic ferroelectric dipole and defect dipole existing on the GB may also contribute to the colossal dielectric permittivity of the BT ceramics.

According to the results shown in Fig. 5, the dielectric property of BT-1 is the best among all the four samples, and thus was chosen for further characterization. Fig. 6a and b show HAADF-STEM images of BT-1, where grain boundaries are clearly observed and the grains grew up during the sintering process rather than aggregation by small nanoparticles. Electron energy loss spectra (EELS) of Ti across the grain boundary were collected along the area marked with green line (see Fig. 6b). Experimental EELS spectra of position located at 1, 2 and 3 were fitted by standard EELS spectra of Ti³⁺ (green curve) and Ti⁴⁺ (pink curve) (see Fig. 6c). The percentage of Ti⁴⁺, Ti³⁺ and Ti³⁺/Ti⁴⁺ ratio at 1, 2 and 3 are listed in Table 2. In addition, by the same method, the percentage of Ti³⁺, Ti⁴⁺ and Ti³⁺/Ti⁴⁺ ratio along the green line were depicted in Fig. 6d. According to the results in Table 2 and Fig. 6d, Ti³⁺ ions indeed exist within grain and at grain boundary. Particularly, large-scale Ti³⁺ ions are located within grain considering its size compared to that of grain boundary.

Fig. 6 (a) HAADF-STEM image of partial BT-1. (b) Corresponding HAADF-STEM image of the red rectangular area in (a). (c) The fitted curves of EELS of Ti at 1, 2 and 3 point in (b), standard spectra for Ti⁴⁺ (green) and Ti³⁺ (pink) are shown at the bottom. (d) The fitted percentage of Ti³⁺, Ti⁴⁺ and Ti³⁺/Ti⁴⁺ ratio along the green line of (b).

Table 2 The percentage of Ti⁴⁺, Ti³⁺ and Ti³⁺/Ti⁴⁺ ratio at different position corresponding to Fig. 6

Location	Ti^3+ (%)	Ti^4+ (%)	Ti^3+/Ti^4+
1	33.7	44.5	0.758
2	29.9	44.9	0.666
3	33.9	49.9	0.679

---

Moreover, X-ray photoelectron spectroscopy (XPS) was used to examine the valence state of the elements in BT ceramics. Fig. 7 shows the XPS narrow scanning spectra of Ba 3d and Ti 2p peaks of BT-1. All of the binding energies at various peaks were calibrated using that of C 1s (284.04 eV), and each component peak in the spectra was fitted with Lorentzian curves. As shown in Fig. 7a, the 5/2 and 3/2 spin–orbit doublet components of the Ba 3d photoelectron are located at about 780.97 eV and 796.31 eV, respectively. In Fig. 7b, the two peaks locating at 459.08 eV and 465.18 eV can be ascribed to Ti 2p peaks. The deconvolution of the slightly broad peak centered at 459.04 eV indicates the existence of an additional peak at around 457.86 eV. Similarly, the peak located at 463.64 eV is associated with the one at 465.18 eV. Both of them proved the co-existence of Ti³⁺ and Ti⁴⁺, which is in good agreement with the EELS results. Ti³⁺ species was also present in BT powder in the literature,⁴⁸ two additional peaks located at 456.8 eV and 462.0 eV, which are ascribed to Ti³⁺ 2p_3/2 and Ti³⁺ 2p_1/2 peaks, are identified, suggesting the presence of Ti³⁺ species. In addition, the two peaks located at 457.9 eV and 463.6 eV can be assigned to the core levels of Ti⁴⁺ 2p_3/2 and Ti⁴⁺ 2p_1/2, respectively. The difference of binding energy for Ti 2p in our samples and the reported one is possibly due to the sample preparation and processing. The presence of Ti³⁺ probably give rise to the observed colossal permittivity, such as small polarons of electronic hopping of Ti³⁺/Ti⁴⁺ within the grain. It has also been found in Ba_0.95La_0.05TiO_3−x ceramics.²⁵

Fig. 7 XPS narrow scanning spectra of BT-1, (a) Ba 3d and (b) Ti 2p.

Temperature dependence of dielectric properties and loss tangent of the BT-1 at different frequencies are shown in Fig. 8. At the same temperature, the dielectric constant and loss tangent decreased with increasing frequency, and the corresponding dielectric peaks are gradually broadened with increasing frequency, indicating the occurrence of a diffuse phase transition.⁴⁹ Generally, the degree of diffuse phase transition is enhanced with decreasing of grain size,^50,51 which could be the reason for the enhanced diffusion of phase transition in our experiment. Generally, the temperature corresponding to the dielectric peak is considered to be the phase transition temperature. The phase transition temperature of sample BT-1 at different frequency is listed in Table 3. It shows that the phase transition temperature of BT-1 decreased with increasing frequency. At the frequency of 1 kHz, the Curie temperature of BT-1 is 144 °C, higher than the empirical value (120 °C). The increasing in Curie temperature may be caused by the smaller grain size.⁵⁰ In the paraelectric region at 1 kHz, the permittivity of ceramics closely follows the Curie–Weiss (C–W) law at 143 °C. Two temperatures are very close, indicating that diffuse phase transition took place in the BT ceramics.

Fig. 8 Temperature dependence of (a) permittivity and (b) loss tangent measured at different frequencies for sample BT-1.

Table 3 Phase transition temperature of BaTiO₃ ceramics at different frequency

Frequency	1 kHz	5 kHz	10 kHz	50 kHz	100 kHz
Temperature (°C)	144	144	144	136	128

---

Furthermore, Fig. 9a shows the Cole–Cole diagram of BT-1. It provides the information of the dielectric relaxation in the BT ceramics.⁵² It is well known that the ideal Debye model of monodispersive relaxation corresponds to a standard semicircle, i.e., the circle center locates exactly on the ε′, whereas for a polydispersive process the circle center locates below the ε′ axis. In our case, there is a parabola above the ε′ axis in the sample, implying that the relaxation behavior exists and its character is a polydispersive relaxation rather than an ideal monodispersive Debye relaxation.⁵³ The Debye relaxation time τ (the characteristic time that the dipoles response to an applied electric field) is calculated to be 1.0662 ×o10⁻⁷ s from the peak frequency f_m of the Cole–Cole diagrams by equation τ = 1/(2πf_m).

Fig. 9 (a) Cole–Cole diagrams of sample BT-1 at room temperature; ε′ is the real part of the relative permittivity (the dielectric constant); ε′′ is the imaginary part; (b) activation energy of BT-1 for polaron hopping polarization at 291–401 K; (c and d) schematic illustrations for defects dipole (P_D) configuration corresponding to spontaneous polarization (P_s) before and after annealing, (c) proposed state before annealing, in which P_D was randomly distributed, with respect to P_s; (d) proposed state after annealing in the single domain, in which defect symmetry followed spontaneous polarization in single domain, (e) proposed state after the ferroelectric phase transition from tetragonal to cubic, in which P_D was randomly distribution again and P_s disappeared.

On the basis of the results in Fig. 6 and 7, the thermally activated polaron hopping model is used to get insight into the physical characteristics of the processes driving the relaxations. The maximum of the imaginary part of the relative permittivity (ε′′), is related to the number of hopping polarons (N) by⁵⁴

where μ represents the hopping dipole moment. N can be written as⁵⁴where N₀ is the pre-exponential factor and E_A is the activation energy.⁵⁵ Thus, substituting eqn (1) into (2) results in⁵⁴

The ε′′_max obtained for each temperature of BT-1, were then plotted in the Arrhenius form (ln(ε′′_maxT) vs. 1/T) to calculate the activation energy for hopping polarization (Fig. 9b). The activation energy of the sample is calculated from the fitted line for high and low temperature region. The ε′′_max increases as temperature increases, which confirms that thermally activated polarization, associated with polaronic dipoles, is a contribution polarization mechanism to the colossal permittivity of BT ceramics. The activation energy (E_A) of 0.135 eV (291–389 K) and 0.065 eV (393–401 K) is obtained in the different temperature regions. The value of activation energy in the low temperature region is about 4 and 2 times larger than the values reported in the literature.^54,56 The transition temperature for the activation energy is around 389 K, which could correspond to a ferroelectric phase transition temperature, as discussed in Fig. 8. One should note that the activation energy below Curie temperature range is much higher than that above the Curie temperature. We attempted to explain why the activation energy in the low temperature region is so high and a transition temperature exists by a simple model described in Fig. 9c–e. As revealed in defect BaTiO₃ single crystal after aging, the defect symmetry in each domain followed the polar tetragonal crystal symmetry, and exhibited a defect dipole moment following the polarization direction of the residing domain.⁵⁷ Encouraged by this work and considering the annealed treatment of the BT nanoceramics before our dielectric measurement, we proposed that defect dipoles tend to be aligned along the spontaneous polarization direction after annealing as shown in Fig. 9d. In contrast, before annealing, defect dipoles were randomly distributed within BaTiO₃ single domain as shown in Fig. 9c. Below ferroelectric Curie temperature, the aligned defect dipoles coupled with P_s could be pinned, leading to the higher E_A (291–389 K) for polaron hopping. Above the Curie temperature, P_D was random distribution again, as shown in Fig. 9e, and the defect dipoles could be easily activated, thus the E_A is lower (393–401 K).

Conclusions

In conclusion, BT nanoparticles with a size of ∼20 nm was synthesized via a sol–gel-hydrothermal method and subsequently densified by SPS. The average grain size of the sintered ceramics was less than 300 nm. The BT nanoceramics exhibited excellent behavior at room temperature as they combined both colossal permittivity (>35 000) at low-frequency and very stable permittivity within a wide frequency range from 40 Hz to 10 kHz, while relaxation frequencies of all samples are larger than 100 kHz. At 1 kHz, the Curie temperature of BT-1 could be around 144 °C. On the basis of microstructure characterizations and dielectric measurement, it was proposed that the polaron dipoles configured by oxygen vacancies and Ti³⁺ cations within the grains could be attributed to the colossal dielectric permittivity of the ceramics. By using thermally activated polaron hopping model, large activation energy E_A has been determined to be 0.135 eV below Curie temperature, probably due to a coupling of polaron dipoles and intrinsic ferroelectric dipoles. This simple method offers a possibility for the large-scale production of BT nanoceramics with a colossal permittivity.

Acknowledgements

This work was financially supported by National Natural Science Foundation of China (51232006, 51472218, and 51102212), China Postdoctoral Science Foundation (2015M571866), and Fundamental Research Funds for the Central Universities (2016FZA4005).

Notes and references

1. R. Asiaie, W. D. Zhu, S. A. Akbar and P. K. Dutta, Chem. Mater., 1996, 8, 226–234.
2. M. Randall, D. Skamser, T. Kinard, J. Qazi, A. Tajuddin, S. Trolier-McKinstry, C. Randall, S. W. Ko and T. Dechakupt, Carts-conference, CARTS 2007 Symposium Proceedings, Albuquerque, NM, Components Technology Institute Inc. 2007, 27, p. 403.
3. T. T. Lee, C. Y. Huang, C. Y. Chang, I. K. Cheng, C. L. Hu, C. T. Lee and M. Fujimoto, J. Mater. Res., 2012, 27, 2495–2502.
4. C. Y. Chang, C. Y. Huang, Y. C. Wu, C. Y. Su and C. L. Huang, J. Alloys Compd., 2010, 495, 108–112.
5. M. T. Buscaglia, C. Harnagea, M. Dapiaggi, V. Buscaglia, A. Pignolet and P. Nanni, Chem. Mater., 2009, 21, 5058–5065.
6. M. T. Buscaglia, M. Bassoli, V. Buscaglia and V. Reinhard, J. Am. Ceram. Soc., 2008, 91, 2862–2869.
7. L. Simon-Seveyrat, A. Hajjaji, Y. Emziane, B. Guiffard and D. Guyomar, Ceram. Int., 2007, 33, 35–40.
8. S. K. Lee, T. J. Park, G. J. Choi, K. K. Koo and S. W. Kim, Mater. Chem. Phys., 2003, 82, 742–749.
9. C. Beck, W. Hartl and R. Hempelmann, J. Mater. Res., 1998, 13, 3174–3180.
10. Y. Yang, X. Wang, C. Sun and L. Li, J. Nanotechnol., 2009, 20, 55709.
11. F. Maxim, P. Ferreira, P. M. Vilarinho and I. Reaney, Cryst. Growth Des., 2008, 8, 3309–3315.
12. M. Oledzka, N. E. Brese and R. E. Riman, Chem. Mater., 1999, 11, 1931–1935.
13. W. W. Wang, L. X. Cao, W. Liu, G. Su and W. X. Zhang, Ceram. Int., 2013, 39, 7127–7134.
14. X. Yang, Z. H. Ren, G. Xu, C. Y. Chao, S. Jiang, S. Q. Deng, G. Shen, X. Wei and G. R. Han, Ceram. Int., 2014, 40, 9663–9670.
15. H. Han, D. Ghosh, J. L. Jones and J. C. Nino, J. Am. Ceram. Soc., 2013, 96, 485–490.
16. G. Arlt, D. Hennings and G. de With, J. Appl. Phys., 1985, 58, 1619–1625.
17. Z. A. Munir, U. Anselmi-Tamburini and M. Ohyanagi, J. Mater. Sci., 2006, 41, 763–777.
18. Z. J. Shen, M. Johnsson, Z. Zhao and M. Nygren, J. Am. Ceram. Soc., 2002, 85, 1921–1927.
19. E. A. Olevsky and L. Froyen, J. Am. Ceram. Soc., 2009, 92, S122–S132.
20. D. Li, J. F. Hu, J. Z. Zhang, J. Ma and Z. J. Shen, Adv. Appl. Ceram., 2014, 113, 251–256.
21. D. Li, W. Li, C. Fasel, J. Shen and R. Riedel, J. Alloys Compd., 2014, 586, 567–573.
22. H. Maiwa, Jpn. J. Appl. Phys., 2008, 47, 7646–7649.
23. S. Yoon, J. Dornseiffer, Y. Xiong, D. Gruner, Z. J. Shen, S. Iwaya, C. Pithan and R. Waser, J. Eur. Ceram. Soc., 2011, 31, 1723–1731.
24. Z. Valdez-Nava, C. Tenailleau, S. Guillemet-Fritsch, N. El Horr, T. Lebey, P. Dufour, B. Durand and J. Y. Chane-Ching, J. Phys. Chem. Solids, 2011, 72, 17–23.
25. S. Guillemet-Fritsch, Z. Valdez-Nava, C. Tenailleau, T. Lebey, B. Durand and J. Y. Chane-Ching, Adv. Mater., 2008, 20, 551–555.
26. T. Ishii, M. Endo, K. Masuda and K. Ishida, Appl. Phys. Lett., 2013, 102, 062901.
27. U. C. Chung, C. Elissalde, S. Mornet, M. Maglione and C. Estournes, Appl. Phys. Lett., 2009, 94, 072903.
28. S. Jayanthi and T. R. N. Kutty, J. Mater. Sci.: Mater. Electron., 2005, 16, 269–279.
29. M. Valant, A. Dakskobler, M. Ambrozic and T. Kosmac, J. Eur. Ceram. Soc., 2006, 26, 891–896.
30. Y. G. Wang, G. Xu, L. L. Yang, Z. H. Ren, X. Wei, W. J. Weng, P. Y. Du, G. Shen and G. R. Han, Mater. Lett., 2009, 63, 239–241.
31. M. T. Buscaglia, M. Viviani, V. Buscaglia, L. Mitoseriu, A. Testino, P. Nanni, Z. Zhao, M. Nygren, C. Harnagea, D. Piazza and C. Galassi, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 73, 0641114.
32. C. H. Perry and D. B. Hall, Phys. Rev. Lett., 1965, 15, 700–702.
33. Y. Ma, E. Vileno, S. L. Suib and P. K. Dutta, Chem. Mater., 1997, 9, 3023–3031.
34. I. J. Clark, T. Takeuchi, N. Ohtori and D. C. Sinclair, J. Mater. Chem., 1999, 9, 83–91.
35. M. F. Lin, V. K. Thakur, E. J. Tan and P. S. Lee, J. Mater. Chem., 2011, 21, 16500–16504.
36. B. R. Li, X. H. Wang, M. M. Cai, L. F. Hao and L. T. Li, Mater. Chem. Phys., 2003, 82, 173–180.
37. W. L. Luan, L. Gao, H. Kawaoka, T. Sekino and K. Niihara, Ceram. Int., 2004, 30, 405–410.
38. B. R. Li, X. H. Wang, L. T. Li, H. Zhou, X. T. Liu, X. Q. Han, Y. C. Zhang, X. W. Qi and X. Y. Deng, Mater. Chem. Phys., 2004, 83, 23–28.
39. T. Takeuchi, C. Capiglia, N. Balakrishnan, Y. Takeda and H. Kageyama, J. Mater. Res., 2002, 17, 575–581.
40. M. T. Buscaglia, V. Buscaglia, M. Viviani, J. Petzelt, M. Savinov, L. Mitoseriu, A. Testino, P. Nanni, C. Harnagea, Z. Zhao and M. Nygren, Nanotechnology, 2004, 15, 1113–1117.
41. S. Yoon, J. Dornseiffer, Y. Xiong, D. Grüner, Z. J. Shen, S. Iwaya, C. Pithan and R. Waser, J. Eur. Ceram. Soc., 2011, 31, 1723–1731.
42. T. Oyama, N. Wada, H. Takagi and M. Yoshiya, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 82, 134107.
43. C. Voisin, S. Guillemet-Fritsch, P. Dufour, C. Tenailleau, H. Han and J. C. Nino, Int. J. Appl. Ceram. Technol., 2013, 10, E122–E133.
44. N. Izyumskaya, Y. Alivov and H. Morkoc, Crit. Rev. Solid State Mater. Sci., 2009, 34, 89–179.
45. B. Jiang, J. L. Peng, L. A. Bursill and W. L. Zhong, J. Appl. Phys., 2000, 87, 3462.
46. J. Varghese, R. W. Whatmore and J. D. Holmes, J. Mater. Chem. C, 2013, 1, 2618–2638.
47. L. Q. Dong, K. Cheng, W. J. Weng and W. Q. Han, CrystEngComm, 2014, 16, 10750–10753.
48. Q. Yu, D. Liu, R. Wang, Z. Feng, Z. Zuo, S. Qin, H. Liu and X. Xu, Mater. Sci. Eng., B, 2012, 177, 639–644.
49. H. C. Yu and Z. G. Ye, J. Appl. Phys., 2008, 103, 034114–034115.
50. X. G. Tang, J. Wang, X. X. Wang and H. L. W. Chan, Solid State Commun., 2004, 131, 163–168.
51. W. H. Tzing, W. H. Tuan and H. L. Lin, Ceram. Int., 1999, 25, 425–430.
52. K. S. Cole and R. H. Cole, J. Chem. Phys., 1941, 9, 341–351.
53. H. B. Yang, H. Wang, L. He and X. Yao, Mater. Chem. Phys., 2012, 134, 777–782.
54. H. Han, C. Voisin, S. Guillemet-Fritsch, P. Dufour, C. Tenailleau, C. Turner and J. C. Nino, J. Appl. Phys., 2013, 113, 024102.
55. S. Komine and E. Iguchi, J. Phys.: Condens. Matter, 2004, 16, 1061–1073.
56. O. Bidault, M. Maglione, M. Actis, M. Kchikech and B. Salce, Phys. Rev. B: Condens. Matter Mater. Phys., 1995, 52, 4191–4197.
57. X. B. Ren, Nat. Mater., 2004, 3, 91–94.

---