Microstructure, dielectric properties and impedance spectroscopy of Ni doped CaCu₃Ti₄O₁₂ ceramics

Abstract

Ni doped CaCu₃Ti₄O₁₂ ceramics (CaCu_3−xNi_xTi₄O₁₂, x = 0, 0.05, 0.1, 0.2) were prepared by a sol–gel method and the influence on the microstructure, dielectric properties and impedance spectroscopy has been investigated. CaCu_3−xNi_xTi₄O₁₂ ceramics can cause a remarkable increase in the grain size and exhibit a very high dielectric constant (ε′) over a wide frequency range at room temperature. The ε′ value of the CaCu_2.9Ni_0.1Ti₄O₁₂ ceramic sintered at 1060 °C for 8 h was between 7.1 × 10⁴ and 9.6 × 10⁴ over the frequency range of 20 Hz to 100 kHz. Besides, a very low dielectric loss (tan δ) of 0.025 was observed for the CaCu_2.95Ni_0.05Ti₄O₁₂ ceramic sintered at 1060 °C for 8 h with ε′ ∼ 4.2 × 10⁴ at about 1 kHz. These results indicate that Ni doping at the Cu site may improve the dielectric properties of CaCu₃Ti₄O₁₂. The giant dielectric response behavior can be attributed to the internal barrier layer capacitance (IBLC) effect.

---

1. Introduction

In recently years, dielectric materials with high dielectric constants (ε′) have attracted much attention due to application in the miniaturization of electronic devices, especially those with giant ε′ at a low frequency. CaCu₃Ti₄O₁₂(CCTO) has recently attracted considerable attention due to its giant ε′ in the order of 10⁴ at room temperature (RT) with a good temperature stability over a wide temperature range from 100 K to 600 K.^1–11 However, an increase in ε′ is usually accompanied by a strong increase in the dielectric loss (tan δ).¹² In other words, high tan δ has restricted the practical application of CCTO.

To date, the mechanism for giant dielectric response of CCTO is still ambiguous.⁶ Most studies attribute the cause of the giant ε′ phenomenon to extrinsic origins. An internal barrier layer capacitance (IBLC) effect associated with insulating grain boundaries and semiconducting grains was proposed by Sinclair group.^3–5 They considered that the novel giant dielectric response of CCTO comes from heterogeneities, and the explanation has been the most widely accepted one so far. According to the IBLC model, the change of electric properties of grain and grain boundary will affect the dielectric properties considerably. In other words, the dielectric properties of CCTO are very sensitive to the preparation process as well as ion-doping. For this hypothesis, many groups have tried doping CCTO with various ions, such as La for Ca,¹³ Ni for Cu,^14–16 Nb for Ti¹⁷ etc. Among these reports, Ni substitution of Cu site seems to be a useful way.^9–11 Rai et al. reported that Ni doping (CaCu_2.9Ni_0.1Ti₄O₁₂) could increase the ε′ from 1800 to 3000 at RT and 1 kHz. However, the tan δ also increased from 0.2 to about 0.25;¹⁴ Zhang et al. synthesized CaCu₃Ni_xTi₄O_12+x (x = 0, 0.1, 0.2, 0.3) ceramics by the sol–gel method. With x = 0.2, they found that CaCu₃Ni_0.2Ti₄O_12.2 ceramics exhibits a high ε′ about 4000 and a tan δ about 0.4 at RT and 1 kHz;¹⁵ Li et al. prepared CCTO–xNiO ceramics by the conventional solid state reaction method. They found that, with x = 0.003 and 0.006, the ε′ could increase to 3.2 × 10⁴ at RT at 1 kHz while the ε′ of pure CCTO is only about 1.2 × 10⁴. On the other hand, the tan δ decreases obviously by Ni doping. With x = 0.020, the tan δ was as low as 0.12 while the tan δ of pure CCTO is 0.34 at RT and 1 kHz.¹¹ As described above, one can find that Ni doping can significantly change the dielectric properties of CCTO. However, both the microstructure and dielectric properties of similar Ni-doping samples prepared by different group or method were very different,^14–16 indicating the cause of the giant ε′ phenomenon is extrinsic.

In our recent work, a very low tan δ of 0.028 was observed in pure CCTO ceramic (prepared by the sol–gel method) sintered at 1100 °C for 16 h with ε′ ∼ 1.8 × 10⁴ at RT and 1 kHz.¹⁸ In this paper, stoichiometric CaCu_3−xNi_xTi₄O₁₂ (CCNTO, x = 0, 0.05, 0.1, 0.2) ceramics were successfully prepared by the sol–gel method. The influences of different Ni doping and sintering temperature/duration on the microstructures, dielectric properties and impedance spectroscopy were analyzed and discussed. Value of ε′ was extremely increased while value of tan δ considerably decreased by Ni doping. A ε′ of ∼8.7 × 10⁴ and tan δ of ∼0.047 have been achieved for CaCu_2.9Ni_0.1Ti₄O₁₂ ceramics sintered at 1060 °C for 8 h in the atmosphere measured at RT and 1 kHz. And a very low tan δ ∼ 0.025 was found in the CaCu_2.95Ni_0.05Ti₄O₁₂ ceramics sintered at 1060 °C for 8 h with ε′ ∼ 4.2 × 10⁴ at RT and 1 kHz. In comparison with the previous work of Ni doped CCTO,^14–16 the CCNTO ceramics in our work can show higher ε′ and lower tan δ, which may be related to the preparation and sintering process.

2. Experimental

CaCu_3−xNi_xTi₄O₁₂ (x = 0, 0.05, 0.1, 0.2) powders were prepared by the sol–gel method. Appropriate amount of calcium nitrate (Ca(NO₃)₂·4H₂O, 99%), nickel nitrate (NiNO₃·6H₂O, 99%), copper nitrate (Cu(NO₃)₂·3H₂O, 99%), tetrabutyl titanate ([CH₃(CH₂)₃O]₄Ti, 99%) and citric acid (C₆H₈O₇, 98.5%), all from aladdin, were weighted in precision and dissolved in ethanol to be mixed uniformly (the pH value needs to be adjusted to 2–3 with nitric acid). After magnetic stirring, homogeneous sol was obtained. The gel was obtained by stirring the sol at 80 °C for about 3 h and dried at 100 °C for about 16 h. The CCNTO precursor powders were obtained by sintering the gel at 650 °C for 2 h in air. Then, the precursor was fully grounded and pressed into small pellets (12 mm in diameter, 2 mm in thickness) with a pressure of 350 MPa. Finally, the pellets were sintered at 1000–1060 °C in a muffle furnace for 8 h to get the CCNTO ceramic samples.

X-ray diffraction with Cu Kα radiation (XRD) (MSAL-XD2) and scanning electronic microscope (SEM) (HITACHI S-520) were used to characterize the phase composition and fractured cross-sectional microstructures. The step size is 0.05° and the scan speed is 6° min⁻¹. Both surfaces of the CCNTO ceramics for dielectric measurements were coated with silver paste as the electrode. The dielectric properties were measured by an LCR meter (Agilent E4980A) with an oscillation voltage of 0.5 V over the frequency range from 20 Hz to 2 MHz and temperature range from 20 °C to 360 °C.

3. Results and discussion

Fig. 1 shows the XRD patterns for different Ni-doped CCNTO ceramic samples sintered at 1060 °C for 8 h. Ni has successfully substituted Cu atom in CCTO due to no second phase was observed. The lattice parameters are calculated to be 7.379 Å, 7.374 Å, 7.385 Å and 7.379 Å for the CaCu_3−xNi_xTi₄O₁₂ ceramic samples with x = 0, 0.05, 0.1 and 0.2 sintered at 1060 °C for 8 h, respectively. Lattice parameters nearly remain unchanged due to the ionic radius of Ni²⁺ (0.72 Å) is almost the same as Cu²⁺ (0.71 Å). Fig. 2(a)–(d) shows the SEM images of the fractured surfaces of CaCu_2.95Ni_0.05Ti₄O₁₂ ceramics sintered at 1000–1060 °C for 8 h, respectively. Though the lattice parameter does not changed a lot, the grain size is influenced by Ni doping in a significant manner. The average grain size (estimated by a line intercept technique) and relative densities (by the Archimedes' method) of CCNTO ceramic samples discussed in this work are listed in Table 1.

Fig. 1 The X-ray diffraction patterns of CaCu_3−xNi_xTi₄O₁₂ (x = 0, 0.05, 0.1 and 0.2) ceramics sintered at 1060 °C for 8 h.

Fig. 2 The FE-SEM images of the fractured surfaces of CaCu_2.95Ni_0.05Ti₄O₁₂ ceramics sintered at (a) 1000 °C, (b) 1020 °C, (c) 1040 °C, (d) 1060 °C for 8 h.

Table 1 The average grain size and relative densities for CaCu_3−xNi_xTi₄O₁₂ ceramics

Sintering temperature/duration	Ni dopant concentration	Average grain size (μm)	Relative density (%)
1000 °C/8 h	x = 0.05	2.65	80.2%
1020 °C/8 h		2.74	81.4%
1040 °C/8 h		5.19	82.8%
1060 °C/8 h		6.53	84.0%
1060 °C/8 h	x = 0	1.68	89.5%
	x = 0.05	2.74	84.0%
	x = 0.1	31.5	89.5%
	x = 0.2	1.77	88.1%

---

As Sinclair et al. reported in their work,^3–5 the one-step IBLC model is widely accepted as an effective method to explain the dielectric properties of CCTO. And the ε′ of CCTO samples can be approximated by the equation as follows.^5,19,20

where ε_r and ε_gb represent the ε′ of the samples and grain boundary, respectively. A and t represent the average grain size of semiconducting grains and the average thickness of grain boundaries, respectively.

Eqn (1) provides the general guidance in interpreting the dielectric responses of systems consisting of semiconducting grains and insulating boundaries.^19,20 According to eqn (1), if one can achieve a high ratio by controlling the grain size and the thickness of grain boundaries, a material with an extremely high ε′ can be obtained. Fig. 3(a) and (b) illustrates the frequency dependence of ε′ and tan δ for CaCu_2.95Ni_0.05Ti₄O₁₂ ceramics sintered at 1000–1060 °C for 8 h measured at RT, respectively. Combining the Fig. 2, 3(a) and the data in Table 1, one can find that the trend of average grain size was nearly the same as the ε′. In addition, an equivalent circuit model has been proposed to clarify the dielectric properties of CCTO. The equivalent circuit contains three RC elements (R_gC_g, R_gbC_gb, and R_xC_x, respectively) and a frequency dependent term Z_UDR, which represents the effect of hopping conduction of localized charge carriers.²¹ In order to explain the behavior of tan δ for different CCNTO samples, the typical Cole–Cole plots of impedance are shown in Fig. 4. To make it clearer, the curves are amplified in the inset of Fig. 4. According to the IBLC model, the tan δ at low frequency mainly rests with the equation as follow¹⁶

tan δ ≈1/ωR_gbC_p (2)

where R_gb represents the resistance of insulating grain boundary, which can be obtained by the extrapolated intercept of the arc on the Z′ axis at the low-frequency, while C_p is the capacitance of the sample, which is proportional to the value of ε′ which depends on eqn (1). It can be seen from Fig. 4 that R_gb increases with the increasing sintering temperature. Since another factor C_p (∝ ε′) also increases with the increasing sintering temperature (Fig. 3(a)), the lower value of tan δ for the sample sintered at 1060 °C for 8 h at low frequency should be caused by the increase of both factors. At the range of high frequency, the tan δ is approximately to be^16,22

tan δ ≈ ωR_gC_p (3)

where R_g represents the resistance of the semiconducting grain, which can be estimated by another nonzero intercept on the Z′ axis shown in the inset of Fig. 4. It can be seen that the R_g of all samples changes within a relatively small range from 0 to 5 Ω cm. Therefore, C_p with remarkable change shown in Fig. 3(a) is considered as a predominant factor at high frequencies in eqn (3). Reviewing the changes of ε′ at high frequency shown in Fig. 3(a), the values increased with the increasing sintering temperature. These can explain why the tan δ value of the ceramics sintering at 1040 °C and 1060 °C is larger than those ceramics sintered at lower temperature, as shown in Fig. 3(b). The results confirm a good correlation between tan δ and ε′ at high frequencies.

Fig. 3 The frequency dependence of ε′ (a) and tan δ (b) for CaCu_2.95Ni_0.05Ti₄O₁₂ ceramics sintered at 1000–1060 °C for 8 h.

Fig. 4 The complex impedance spectroscopies for CaCu_2.95Ni_0.05Ti₄O₁₂ ceramics sintered at 1000–1060 °C for 8 h.

Fig. 5(a)–(d) shows the SEM images of the fractured surfaces of CaCu_3−xNi_xTi₄O₁₂ (x = 0, 0.05, 0.1, 0.2) ceramics sintered at 1060 °C for 8 h, respectively. One can clearly find that Ni doping was benefit for the grain growth of CCTO (see Table 1), and the ceramic with x = 0.1 owned the largest grain size. However, the relative density did not increase significantly as the grain size which may be caused by the increasing porosity resulting from Ni doping (see Fig. 5). Moreover, the sintering temperature which may be adjusted by the amount of Ni doping can be another reason. In future work, we will try treating CCNTO ceramics for longer time in order to reduce the porosity. Fig. 6(a) and (b) illustrates the frequency dependence of ε′ and tan δ for CaCu_3−xNi_xTi₄O₁₂ (x = 0, 0.05, 0.1, 0.2) ceramics sintered at 1060 °C for 8 h measured at RT, respectively. At 1 kHz, the ε′ value is about 1.3 × 10⁴, 4.2 × 10⁴, 8.8 × 10⁴ and 5.3 × 10⁴ for CaCu_3−xNi_xTi₄O₁₂ (x = 0, 0.05, 0.1, 0.2) ceramics, respectively, indicating Ni doping not only increases the grain size but also extremely increase the ε′. The trend of grain size was found to be nearly the same as the ε′ (see Fig. 5 and 6(a)), which can be explained by eqn (1).

Fig. 5 The FE-SEM images of the fractured surfaces of CaCu_3−xNi_xTi₄O₁₂ ceramics with (a) x = 0, (b) x = 0.05, (c) x = 0.1 and (d) x = 0.2 sintered at 1060 °C for 8 h.

Fig. 6 The frequency dependence of ε′ (a) and tan δ (b) for CaCu_3−xNi_xTi₄O₁₂ (x = 0, 0.05, 0.1 and 0.2) ceramics sintered at 1060 °C for 8 h.

Until now, the greatest challenge for CCTO ceramics is to lower the dielectric loss and keep a high dielectric constant.²³ The effect of sintering temperature on tan δ was shown in Fig. 3(b), a very low tan δ value ∼ 0.025 was found at the CaCu_2.95Ni_0.05Ti₄O₁₂ ceramic sintered at 1060 °C for 8 h with ε′ ∼ 4.2 × 10⁴ at about 1 kHz. For the ceramic sintered at 1040 °C, the two values are 0.026 and 4.0 × 10⁴, not very different from the ceramic sintered at 1060 °C. The effect of different Ni doping ratio on the tan δ can be observed in Fig. 6(b). One can find that, sintering at the same condition, at low frequency (f < 2 kHz), the tan δ of CaCu_2.95Ni_0.05Ti₄O₁₂ ceramic can be lower than pure CCTO; whereas the other two Ni doping ceramics increased the tan δ at almost the whole frequency range. The temperature dependence of ε′ and tan δ measured at some typical frequencies for the CaCu_2.95Ni_0.05Ti₄O₁₂ ceramic sintered at 1060 °C for 8 h are plotted in Fig. 7. There is an obvious dielectric relaxation at low temperature in both the ε′–T and tan δ–T plots, and a high temperature dielectric relaxation probably existed which is out of the temperature range of measurement (ε′–T plots). The characteristic temperature increases with the increasing measurement frequency. As reported in the previous work, the high temperature relaxation is closely related to the high temperature conductivity according to the almost same activation energy.²⁴

Fig. 7 The temperature dependence of ε′ (a) and tan δ (b) for the CaCu_2.95Ni_0.05Ti₄O₁₂ ceramic sintered at 1060 °C for 8 h measured at some typical frequencies.

4. Conclusions

CaCu_3−xNi_xTi₄O₁₂ ceramics were prepared by the sol–gel method, and then sintered at various temperatures and durations. The dielectric properties as well as microstructure of CCTO were successfully improved by Ni doping, and Ni doping lead to increased porosity which may be due to the fact that the sintering time is not long enough. CaCu_2.9Ni_0.1Ti₄O₁₂ ceramic sintered at 1060 °C for 8 h exhibited very high ε′ (7.1 × 10⁴ to 9.6 × 10⁴) over the frequency range of 20 Hz to 100 kHz. The CaCu_2.95Ni_0.05Ti₄O₁₂ ceramic sintered at 1060 °C for 8 h show both a very low tan δ ∼ 0.025 and a very high ε′ ∼ 4.2 × 10⁴ at RT and 1 kHz. Correspondingly, dielectric relaxations can be observed in ε′/tan δ–T curves, and ε′ tends to be slightly dependent on temperature at high frequency.

Acknowledgements

This work was supported by National Natural Science Foundation of China (NSFC) (No. 11404236), Special Funds of the National Natural Science Foundation of China (No. 11547176) and Natural Science Foundation of Shanxi (No. 2015021026).

References

1. M. A. Subramanian, D. Li, N. Duan, B. A. Reisner and A. W. Sleight, High Dielectric Constant in ACu₃Ti₄O₁₂ and ACu₃Ti₃FeO₁₂ Phases, J. Solid State Chem., 2000, 151, 323–325.
2. A. P. Ramirez, M. A. Subramanian, M. Gardel, G. Blumberg, D. Li, T. Vogt and S. M. Shapiro, Giant dielectric constant response in a copper–titanate, Solid State Commun., 2000, 115, 217–220.
3. T. B. Adams, D. C. Sinclair and A. R. West, Characterization of grain boundary impedances in fine- and coarse-grained CaCu₃Ti₄O₁₂ ceramics, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 73, 094124.
4. D. C. Sinclair, T. B. Adams, F. D. Morrison and A. R. West, CaCu₃Ti₄O₁₂: one-step internal barrier layer capacitor, Appl. Phys. Lett., 2002, 80, 2153.
5. T. B. Adams, D. C. Sinclair and A. R. West, Giant Barrier Layer Capacitance Effects in CaCu₃Ti₄O₁₂ Ceramics, Adv. Mater., 2002, 14, 1321–1323.
6. C. C. Homes, T. Vogt, S. M. Shapiro, S. Wakimoto and A. P. Ramirez, Optical response of high-dielectric-constant perovskite-related oxide, Science, 2001, 293, 673–676.
7. Z. M. Dang, T. Zhou, S. H. Yao, J. K. Yuan, J. W. Zha, H. T. Song, J. Y. Li, Q. Chen, W. T. Yang and J. Bai, Advanced Calcium Copper Titanate/Polyimide Functional Hybrid Films with High Dielectric Permittivity, Adv. Mater., 2009, 21, 2077–2082.
8. L. Zhang, X. Shan, P. Wu, J. Song and Z. Y. Cheng, Microstructure and Dielectric Properties of CCTO-P(VDF-TrFE) Nanocomposites, Ferroelectrics, 2010, 405, 92–97.
9. W. Yang, S. Yu, R. Sun and R. Du, Nano- and microsize effect of CCTO fillers on the dielectric behavior of CCTO/PVDF composites, Acta Mater., 2011, 59, 5593–5602.
10. L. Zhang, X. Shan, P. Wu and Z. Y. Cheng, Dielectric characteristics of CaCu₃Ti₄O₁₂/P(VDF-TrFE) nanocomposites, Appl. Phys. A, 2012, 107, 597–602.
11. L. J. Romasanta, P. Leret, L. Casaban, M. Hernández, M. A. de la Rubia, J. F. Fernández, J. M. Kenny, M. A. Lopez-Manchado and R. Verdejo, Towards materials with enhanced electro-mechanical response: CaCu₃Ti₄O₁₂–polydimethylsiloxane composites, J. Mater. Chem., 2012, 22, 24705.
12. P. Thongbai, K. Meeporn, T. Yamwong and S. Maensiri, Extreme effects of Na doping on microstructure, giant dielectric response and dielectric relaxation behavior in CaCu₃Ti₄O₁₂ ceramics, Mater. Lett., 2013, 106, 129–132.
13. S. Jin, H. Xia and Y. Zhang, Effect of La-doping on the properties of CaCu₃Ti₄O₁₂ dielectric ceramics, Ceram. Int., 2009, 35, 309–313.
14. A. K. Rai, K. D. Mandal, D. Kumar and O. Parkash, Characterization of nickel doped CCTO: CaCu_2.9Ni_0.1Ti₄O₁₂ and CaCu₃Ti_3.9Ni_0.1O₁₂ synthesized by semi-wet route, J. Alloys Compd., 2010, 491, 507–512.
15. C.-H. Zhang, K. Zhang, H.-X. Xu, Q. Song, Y.-T. Yang, R.-H. Yu, D. Xu and X.-N. Cheng, Microstructure and electrical properties of sol–gel derived Ni-doped CaCu₃Ti₄O₁₂ ceramics, Trans. Nonferrous Met. Soc. China, 2012, 22, s127–s132.
16. T. Li, J. Chen, D. Liu, Z. Zhang, Z. Chen, Z. Li, X. Cao and B. Wang, Effect of NiO-doping on the microstructure and the dielectric properties of CaCu₃Ti₄O₁₂ ceramics, Ceram. Int., 2014, 40, 9061–9067.
17. P. Leret, M. A. de la Rubia, J. J. Romero, J. De Frutos and J. F. Fernández, Phenomenological model of grain boundary behaviour under a bias field in Nb-doped CaCu₃Ti₄O₁₂ ceramics, J. Alloys Compd., 2011, 509, 9719–9723.
18. L. Sun, Z. Wang, Y. Shi, E. S. Cao, Y. Zhang, H. Peng and L. Ju, Sol–Gel synthesized pure CaCu₃Ti₄O₁₂ with very low dielectric loss and high dielectric constant, Ceram. Int., 2015, 41, 13486–13492.
19. L. Marchin, S. Guillemet-Fritsch, B. Durand, A. A. Levchenko, A. Navrotsky and T. Lebey, Grain Growth-Controlled Giant Permittivity in Soft Chemistry CaCu₃Ti₄O₁₂ Ceramics, J. Am. Ceram. Soc., 2008, 91, 485–489.
20. M. H. Cohen, J. B. Neaton, L. He and D. Vanderbilt, Extrinsic models for the dielectric response of CaCu₃Ti₄O₁₂, J. Appl. Phys., 2003, 94, 3299.
21. W. Hao, J. Zhang, Y. Tan, M. Zhao and C. Wang, Giant Dielectric Permittivity Properties and Relevant Mechanism of NaCu₃Ti₃SbO₁₂ Ceramics, J. Am. Ceram. Soc., 2011, 94, 1067–1072.
22. Y. Yan, L. Jin, L. Feng and G. Cao, Decrease of dielectric loss in giant dielectric constant CaCu₃Ti₄O₁₂ ceramics by adding CaTiO₃, Mater. Sci. Eng., B, 2006, 130, 146–150.
23. S. Vangchangyia, E. Swatsitang, P. Thongbai, S. Pinitsoontorn, T. Yamwong, S. Maensiri, V. Amornkitbamrung, P. Chindaprasirt and J. Hu, Very Low Loss Tangent and High Dielectric Permittivity in Pure-CaCu₃Ti₄O₁₂ Ceramics Prepared by a Modified Sol–Gel Process, J. Am. Ceram. Soc., 2012, 95, 1497–1500.
24. L. Ni, X. M. Chen and X. Q. Liu, Structure and modified giant dielectric response in CaCu₃(Ti_1−xSn_x)₄O₁₂ ceramics, Mater. Chem. Phys., 2010, 124, 982–986.

---