Influence of sintering conditions on microstructure and electrical properties of CaCu3Ti4O12 (CCTO) ceramics

Influence of sintering conditions on microstructure and electrical properties of CaCu3Ti4O12 (CCTO) ceramics

Journal of Alloys and Compounds 650 (2015) 59e64 Contents lists available at ScienceDirect Journal of Alloys and Compounds journal homepage: http://...

2MB Sizes 1 Downloads 40 Views

Journal of Alloys and Compounds 650 (2015) 59e64

Contents lists available at ScienceDirect

Journal of Alloys and Compounds journal homepage:

Influence of sintering conditions on microstructure and electrical properties of CaCu3Ti4O12 (CCTO) ceramics Pei Liu a, b, Yuanming Lai a, Yiming Zeng a, *, Shuang Wu a, Zihan Huang a, Jiao Han a a

State Key Laboratory of Advanced Technologies for Comprehensive Utilization of Platinum Metals, Kunming Institute of Precious Metals, Kunming, 650106, PR China b Aerospace Science and Industry Wuhan Magnetism-electron Co., Ltd, Wuhan, 430074, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 February 2015 Received in revised form 20 July 2015 Accepted 27 July 2015 Available online 29 July 2015

CaCu3Ti4O12 (CCTO) ceramic was prepared by the solid-state method. The microstructure, dielectric properties, complex impedance and nonlinear IeV characteristic were studied. The results show that increasing the sintering temperature, most grains grow up firstly, and then some diminish, resulting in depravation of the comprehensive properties in varying degrees. While prolonging sintering time promotes grain growth, microstructural densification, and improves the dielectric and nonlinear IeV properties. It should be noted that the CCTO ceramics sintered at 1050  C for 12 h exhibit giant dielectric constant of 105 and low dielectric loss <0.1 with weak frequency dependence below 1 MHz, as well as a nonlinear IeV coefficient of 5.27. © 2015 Elsevier B.V. All rights reserved.

Keywords: CCTO Microstructure Dielectric properties IeV characteristic

1. Introduction In recent years, CaCu3Ti4O12 (CCTO) has attracted much interest due to the unusual perovskite structure and remarkable electric 0 00 properties. CCTO is a 1: 3 A-site ordered perovskite (A A 3B4O12) compound with space group Im-3, containing square-planar Cu atoms on A sites that cause tilting of TiO6 octahedra [1]. CCTO presents an extraordinarily high dielectric constant of ~105, which is almost frequency-independent up to 106 Hz and shows good temperature stability from 100 K to 600 K. Furthermore, CCTO does not also undergo any structural phase transitions from 20 Ke600 K [2]. In addition, CCTO also exhibits unusual nonlinear currentevoltage (IeV) characteristic [3,4]. Therefore, CCTO has been considered as a promising material for capacitor-based applications, microwave communication devices, switching, gas sensing devices and energy storage devices [4e6]. However, the fact that giant permittivity of CCTO accompanied by high dielectric loss and sensitivity of dielectric performance to the preparation technology [7,8], seriously hinders the application of CCTO. So, lowering dielectric loss and developing stable processing technology has been becoming an urgent issue. Till now,

* Corresponding author. E-mail address: [email protected] (Y. Zeng). 0925-8388/© 2015 Elsevier B.V. All rights reserved.

some theoretical models have been proposed to explain the origin of giant permittivity and high loss, and finally help to develop applicable CCTO materials, such as internal domain [9], electrode polarization effect [10], bimodal grain size model [11], internal barrier layer capacitance (IBLC) [7] and nanoscale barrier layer capacitance model (NBLC) [12]. Where, the IBLC model has been widely accepted as the most approbatory explanation for the abnormal dielectric response in CCTO ceramics [8,13,14], which proposes that semiconducting grains are separated by insulating grain boundaries, producing many small capacitances and resulting in high apparent permittivity values. Afterward, the NBLC model, based on the existence of stacking faults [12], was proposed to reconcile the opposing views of intrinsic versus extrinsic debate about the origin of high dielectric constant. At the same time, the dielectric loss and nonlinear IeV properties were also discussed based on different models and most results indicated that they were closely related to the microstructure in CCTO [15e18], which can be markedly influenced by element doping [19e21] and sintering conditions [7,22]. In this work, CCTO ceramics were prepared by the conventional solid-state route at different sintering temperature and for different sintering time, in order to study the evolution of microstructure and properties with the sintering conditions. The microstructure, dielectric and impedance performance, currentevoltage behaviors of obtained CCTO ceramics were characterized, in order to discuss


P. Liu et al. / Journal of Alloys and Compounds 650 (2015) 59e64

the influence of sintering time and temperature on CCTO ceramics systematically. The results offer some useful information for study and developing available high-performance CCTO materials in future. 2. Experimental section Ceramic samples of CaCu3Ti4O12 were prepared by solid state reaction and conventional sintering route, using analytical-grade (99.5%) powders of CaCO3, CuO and TiO2 as starting materials. The reagents were weighted according to the stoichiometric rations 1: 3: 4 and mixed thoroughly in a planetary ball mill for 4 h with zirconia balls, using deionized water as dispersant. The mixed powder was calcined at 950  C in air for 4 h to synthesize CCTO powder. The calcined powder was ball-milled for a further 4 h to obtain fine powder (D50 z 1.5 mm), and then pressed into disks with 12.7 mm in diameter and 2 mm in thickness under pressure of 120 MPa by using polyvinyl alcohol as organic binders. The pellets were sintered in air at 1000  C, 1050  C, 1100  C for 12 h and at 1050  C for 4 h, 8 h and 12 h, with the heating rate of 5  C/min, then furnace-cooled to room temperature. The samples were characterized by different analytic techniques. Powder X-Ray diffraction (XRD) patterns were performed on a diffractometer Panalytica X'Pert PRO with Cu Ka radiation (l ¼ 1.5406 Å) to check the phase. The 2q angles were scanned from 10 to 80 with step of 0.03 . Microstructure and the grain size distribution in different samples were characterized by field emission scanning electron micrograph (FESEM, TM-1000, Hitachi, Japan), the SEM micrographs were performed at an accelerating voltage of 15 kV. Pellet densities of samples were measured by the Archimedes method, r ¼ m1/(m1m2), where r is density, m1 is the dry mass and m2 is the mass in deionized water. In order to characterize electrical properties, the sintered pellet were polished and printed with silver paste on both sides and then heated at 850  C for 15 min to form electrodes. Dielectric and impedance properties were measured from 40 Hz to 100 MHz using the impedance analyzer (Agilent 4294A, Agilent, USA) with an oscillation voltage of 0.5 V at room temperature. The IeV behaviors of the obtained CCTO ceramics were measured using Digital Source Meters (Model 2400 and 6487, Ketthley, USA) at room temperature. 3. Results and discussion The XRD patterns for CCTO ceramics sintered under different conditions are shown in Fig. 1. All the diffraction peaks in the patterns can be perfectly indexed to the perovskite CCTO phase (JCPDS No.075-2188), and no impure peaks can be observed from possible secondary phases such as CuO, Cu2O, TiO2, CaTiO3, especially the CuO, Cu2O, which appeared by sintering at 1115  C for different time [23]. The results indicate that the sintering conditions in this work have not caused observable decompositions of CCTO structure, and could be used as references for production of CCTO materials in future. The morphology and grain size of the as-prepared ceramics are shown in Fig. 2 (different sintering temperature) and Fig. 3 (different sintering time). Fig. 2 shows SEM micrographs and relative densities of the CCTO ceramics sintered for 12 h at 1000  C, 1050  C and 1100  C. As can be seen, when sintered at 1000  C, most of the particles in ceramics are very small, about 0.3e1 mm with narrow size distribution (Fig. 2a). When sintered at 1050  C, many particles can grow up to about 20e30 mm in size, and the number of small particles is fewer than sintered at 1000  C (Fig. 2b). Interestingly, when sintered at 1100  C, the grain size distributes in disorder and many particles present regular shape (Fig. 2c). The densities of ceramics are 4.71 g/cm3 (standard deviations s ¼ 0.014)

Fig. 1. XRD patterns of CCTO ceramics sintered at 1050  C for (a) 4 h, (b) 8 h, (c) 12 h and for 12 h at (d) 1000  C, (e) 1100  C.

at 1000  C, 4.87 g/cm3 (s ¼ 0.016) at 1050  C and 4.76 g/cm3 (s ¼ 0.009) at 1100  C, respectively, corresponding to 93.2%, 96.4% and 94.2% of theoretical density of 5.05 g/cm3. Fig. 2d presents the change of relative density with the different sintering temperature. Obviously, the relative density of CCTO ceramics increases firstly and then decreases with increasing sintering temperature. It reveals that, the CCTO particles can grow up at suitable temperature and result in high density, low temperature will result in insufficient sintering, overheating will promote crystallization but reduce the amount of large grains and then lower densification. Fig. 3 shows SEM micrographs of the surface morphologies and relative densities of the CCTO ceramics sintered at 1050  C for 4 h, 8 h and 12 h. Remarkable changes in microstructure with sintering time are clearly observed. When sintering for 4 h, almost all particles are still small, about 0.5e1.5 mm with narrow size distribution (Fig. 3a) in ceramics. With increasing to 8 h, some particles grow rapidly to large grains (about 10e20 mm) (Fig. 3b), and the microstructure shows a bimodal distribution of particles with small ones (about 1e3 mm) surrounding the large ones. When increasing to 12 h (Fig. 3c), the small particles are further swallowed by large ones to grow. As shown in micrography, the grain size and number of large grains increase with prolonging sintering time, which significantly promotes the grain growth. Fig. 3d shows the relative density changes with the sintering time. The densities of CCTO ceramics are 4.71 g/cm3 (s ¼ 0.025) for 4 h, 4.83 g/cm3 (s ¼ 0.036) for 8 h and 4.87 g/cm3 for 12 h, respectively, corresponding to 93.2%, 95.7% and 96.4% of theoretical density. The microstructural densification of CCTO ceramics increases with sintering time, revealing that sufficient time allow more matter transport from ultra-fine particles for grain growth under the driving force from surface energy difference, and finally forming the bimodal distribution of particles, which will reduce the porosity effectively and increase the density. The frequency dependence of the dielectric constant of the CCTO ceramics at room temperature is shown Fig. 4. It can be observed that the permittivity of ceramics sintered at 1050  C for 8 h and 12 h maintain almost constant from 102 Hz to 106 Hz, presenting good frequency-stability, while the others decrease with frequency increasing in different degrees. When the frequency exceeds 106 Hz, the behaviors of permittivity vs. frequency for all samples are linearly declined in semi-logarithmic coordinates, corresponding to the well-known Debye-like relaxation

P. Liu et al. / Journal of Alloys and Compounds 650 (2015) 59e64


Fig. 2. SEM micrographs and relative densities of CCTO ceramics sintered for 12 h at (a) 1000  C, (b) 1050  C, (c) 1100  C.

Fig. 3. SEM micrographs and relative densities of CCTO ceramics sintered at 1050  C for (a) 4 h, (b) 8 h, (c) 12 h.

[11,22e25]. By comparison, it can be found that the permittivity increases quickly with the sintering time increased from 4 h to 8 h, but no virtually changes from 8 h to 12 h, which is in accord with

the changes of microstructure in ceramics, as shown in Fig. 3. On the other hand, when the temperature increases from 1000  C to 1050  C (12 h), the corresponding permittivity changes from about


P. Liu et al. / Journal of Alloys and Compounds 650 (2015) 59e64

Fig. 4. Frequency dependence of dielectric constant for CCTO ceramics.

104 to 105 between 102 Hz and 106 Hz. The permittivity at 1100  C is higher than at 1000  C and 1050  C, but it has obvious changes between 102 Hz and 106 Hz. In conclusion, the frequency-stability of permittivity increases firstly and then deteriorates seriously, it may be due to the changes of microstructure as discussed above and more oxygen vacancies at high sintering temperature in the CCTO ceramics [8]. The frequency dependence of dielectric loss (tand) of the CCTO ceramics measured at room temperature is shown in Fig. 5. As can be seen, the dielectric loss tand of all CCTO ceramics slowly declines from 100 Hz to 105 Hz and then sharply increases in high-frequency range (105 ~ 107 Hz), which could be attributed to the Debye-like relaxation process [11,26,27]. By comparison, it can be found that the dielectric loss decreases with extending sintering time, but decreases firstly and then rises rapidly with increasing sintering temperature, which is consistent with the microstructure analysis of Figs. 2 and 3. The results reveal that the dielectric loss of CCTO ceramics maybe mainly come from the structural defect (such as pores, metal valence change and oxygen vacancies) [12] and then be sensitive to the densification. Longer sintering time will promote densifying, reduce defects and lower loss. But too high sintering

temperature (1100  C) will make the microstructure of CCTO ceramics loose, and the associated defect chemistry in CCTO materials are metastable defects, which lead to high dielectric loss combined with huge dielectric properties. It is worthy to note that, CCTO ceramics sintered at 1050  C for 12 h exhibits low tand (~0.1) over wide frequency range of 102 ~ 105 Hz and the minimum is 0.07 near 20 kHz, which endows CCTO materials more competitiveness for practical applications. From above analysis, it can be concluded that more large grains is helpful for enhancing the permittivity but bad for lowering the dielectric loss of CCTO ceramics. From the complex impedance spectroscopy of CCTO ceramics (Fig. 6), we can see that all samples exhibit two parts in the measuring frequency range: an approximate straight line with a large slope and a single semicircular arc, indicating that there are two different contributions for impedance in CCTO. According to the IBLC theory [7,14,28], the straight line corresponds to the contribution of grains at high frequency and the semicircular arc corresponds to the contribution of grain boundaries at low frequency respectively, which can be described using an equivalent electrical circuit model with two parallel RC elements in series [8,29] (Fig. 7), Rg represents the high frequency grain impedance and Rgb represents the low frequency grain boundary impedance. The resistance is larger at low frequency, implying an extrinsic GBtype relaxation, which is dominated by the insulting grain boundary. Thus, the carriers are blocked near grain boundary and forming double layer, then the low frequency spur in the impedance spectra behaves like the double layer in electroceramics, which is correlated to the electrodes [30]. The values of Rg and Rgb can be estimated from the intercept of Z0 axis, and are summarized in Table 1. The low Rg and high Rgb indicate that the grains are semiconducting and the grain boundaries are insulating in CCTO ceramics. As can be seen, prolonging sintering time from 4 h to 12 h, the Rg increase stably from 400 U to 680 U, but the Rgb increase sharply from 7.5  103 U to 8.5  104 U at first and then gently to 9.6  104 U. While raising the sintering temperature from 1000  C to 1100  C, the Rg increase at first and then drop remarkably, and the Rgb monotonically from 3.5  103 U to 1.2  105 U. These results indicate that the insulativity of boundaries will increase with sintering degree deepening, no matter prolonging time or raising temperature, but the semiconductivity of grains will reduce with prolonging time and fluctuate violently with raising temperature. It may be affected by the interfacial polarization process known as MaxwelleWagnereSillar, which is produced by the traveling of charge carriers [31]. Base on above analyses, the sintering conditions have important impact on the intrinsic properties of grains and boundaries, and the boundaries may play a greater role than grains for the giant dielectric constant of CCTO ceramics. Except the giant dielectric constant, CCTO also presents good nonlinear currentevoltage (IeV) properties, which is empirically expressed by the equation as follow [32]:

I ¼ KV a

Fig. 5. Frequency dependence of dielectric loss for CCTO ceramics.


where K is a constant related to the electrical resistivity of the material, and a is the nonlinear coefficient. Fig. 8 shows the IeV curves of CCTO samples at room temperature, and the values of a, deduced by curve fitting, are summarized in Table 1. For the samples sintered at 1050  C, a increases from 1.27 to 4.49 as prolonging sintering time from 4 h to 8 h, and then up to 5.27 for 12 h. While fixed the sintering time for 12 h, a increase from 1.53 to 5.27 as increasing sintering temperature from 1000  C to 1050  C, then drops to 2.52 at 1100  C. The variations of a are consistent with changes in microstructure, that, large grain particles corresponding to higher a and small one corresponding to lower a. The results indicate that nonlinear currentevoltage properties of CCTO

P. Liu et al. / Journal of Alloys and Compounds 650 (2015) 59e64


Fig. 6. Impedance complex plane plot of CCTO ceramics.

ceramics are very sensitive to the grain boundary, including its composition and microstructure, which were seriously dependent on the sintering conditions. The high variability of a was explained by forming Schottky-type barrier at grain boundary [15,33].

Fig. 7. Equivalent circuit model to account for impedance spectroscopy data of CCTO ceramics.

4. Conclusions Pure phase CCTO ceramics were prepared by a conventional sintering method. The CCTO particles can grow up at suitable temperature and result in high density, but overheating will break large grains and deteriorate the comprehensive properties. While prolonging sintering time can promote grain growth, microstructural densification, and improves the dielectric and nonlinear IeV properties. The results show that the process window of CCTO ceramics is narrow. In summary, dense structure of CCTO will improve the electrical properties and increase its stability. The CCTO ceramics sintered at 1050  C for 12 h exhibit dielectric constant about 105 and low dielectric loss <0.1 with minimal frequency dependence below 1 MHz, and also have a nonlinear coefficient of 5.27. In this regard, the present results suggest that CCTO ceramics can be used as varistors or switching devices of the high-current applications.


P. Liu et al. / Journal of Alloys and Compounds 650 (2015) 59e64

Table 1 Grain resistance Rg grain boundary resistance Rgb and non-linear coefficient a of CCTO ceramics. Sintered conditions 1050  C

4 8 12 12 12

1000  C 1100  C

h h h h h

Fig. 8. Current-voltage characteristics of CCTO ceramics.

Acknowledgments Financial support from the Technology Development Project of Yunnan Tin Group (Grant No. Q/GYGL14042010) and Science & Technology Program of Yunnan Province (NO. 2014DC019) is gratefully acknowledged. References [1] M.A. Subramanian, A.W. Sleight, Solid State Sci. 4 (2002) 347e351. [2] C.C. Homes, T. Vogt, S.M. Shapiro, S. Wakimoto, A.P. Ramirez, Science 293 (2001) 673e676. [3] T. Prasit, P. Bundit, Y. Teerapon, M. Santi, J. Mater. Sci. Mater. Electron. 23 (2012) 795e801. [4] Z.Y. Lu, X.M. Li, J.Q. Wu, J. Am. Ceram. Soc. 95 (2012) 476e479. [5] L.C. Kretly, A.F.L. Almeida, P.B.A. Fechine, R.S.D. Oliveira, A.S.B. Sombra, J. Mater. Sci. Mater. Electron. 15 (2004) 657e663.




400 550 680 260 320

7500 85500 96100 3500 120000

1.27 4.49 5.27 1.53 2.52

[6] R.K. Pandey, W.A. Stapleton, J. Tate, A.K. Bandyopadhyay, I. Sutanto, S. Sprissler, S. Lin, AIP Adv. 3 (2013) 062126. [7] R. Schmidt, M.C. Stennett, N.C. Hyatt, J. Pokorny, J.P. Gonjal, M. Li, D.C. Sinclair, J. Eur. Ceram. Soc. 32 (2012) 3313e3323. [8] R. Schmidt, D.C. Sinclair, Nova Science Publishers Inc., New York, 2013. [9] T.T. Fang, C.P. Liu, Chem. Mater 20 (2005) 5167e5171. [10] P. Lunkenheimer, R. Fichtl, S.G. Ebbinghaus, A. Loidl, Phys. Rev. B 70 (2004) 172102. [11] M.J. Pan, B.A. Bender, J. Am. Ceram. Soc. 88 (2005) 2611e2614. [12] P.R. Bueno, R. Tararan, R. Parra, E. Joanni, M.A. Ramirez, W.C. Ribeiro, E. Longo, J. Varela, J. Phys. D. Appl. Phys. 42 (2009) 055404. [13] P. Fiorenza, R.L. Nigro, C. Bongiorno, V. Raineri, M.C. Ferarrelli, D.C. Sinclair, A.R. West, Appl. Phys. Lett. 92 (2008) 182907. [14] L. Singh, U.S. Rai, K.D. Mandal, J. Alloys Compd. 555 (2013) 176e183. [15] M.A. Ramirez, P.R. Bueno, J.A. Varela, E. Longo, Appl. Phys. Lett. 89 (2006) 212102. [16] M.A. Ramirez, P.R. Bueno, E. Longo, J.A. Varela, J. Phys. D. Appl. Phys. 41 (2008) 152004. [17] M.A. Ramirez, P.R. Bueno, R. Tararam, A.A. Cavalheiro, E. Longo, J.A. Varela, J. Phys. D. Appl. Phys. 42 (2009) 1855503. [18] T. Li, J. Chen, D. Liu, Z.X. Zhang, Z.P. Chen, Z.X. Li, X.Z. Cao, B.Y. Wang, Ceram. Int. 40 (2014) 9061e9067. [19] A.K. Rai, J. Gim, E.C. Shin, H.H. Seo, V. Mathew, K.D. Mandal, O. Parkash, J.S. Lee, J. Kim, Ceram. Int. 40 (2014) 181e189. [20] D. Xu, K. He, R.H. Yu, L. Jiao, H.M. Yuan, X.J. Sun, G.P. Zhao, H.X. Xu, X.N. Cheng, J. Alloys Comp. 592 (2014) 220e225. [21] Q. Zheng, H.Q. Fan, C.B. Long, J. Alloys Comp. 511 (2012) 90e94. [22] J.L. Zhang, P. Zheng, C.L. Wang, M.L. Zhao, J.C. Li, J.F. Wang, Appl. Phys. Lett. 87 (2005) 142901. [23] S. Kwon, D.P. Cann, J. Mater. Sci. 44 (2009) 4117. [24] S.F. Shao, J.L. Zhang, P. Zheng, W.L. Zhong, C.L. Wang, J. Appl. Phys. 99 (2006) 084106. [25] L. Ni, X.M. Chen, X.Q. Liu, R.Z. Hou, Solid State Commun. 139 (2006) 45e50. [26] S.F. Shao, J.L. Zhang, P. Zheng, C.L. Wang, J.C. Li, M.L. Zhao, Appl. Phys. Lett. 91 (2007) 042905. [27] P. Thongbai, J. Jumpatam, T. Yamwong, S. Maensiri, J. Eur. Ceram. Soc. 32 (2012) 2423e2430. [28] S.I.R. Costa, M. Li, J.R. Fradeb, D.C. Sinclair, RCS Adv. 3 (2013) 7030e7036. [29] D.C. Sinclair, T.B. Adams, F.D. Morrison, A.R. West, Appl. Phys. Lett. 80 (2002) 2153e2155. [30] E. Barsoukov, J.R. Macdonald, Impedance Spectroscopy Theory, Experiment and Applications, second ed., Wiley-Interscience, 2005. [31] L.A. Ramajo, M.A. Ramirez, P.R. Bueno, M.M. Reboredo, M.S. Castro, Mater. Res. 1 (2008) 85e88. [32] D.R. Clarke, J. Am. Ceram. Soc. 82 (1999) 485e502. [33] D.L. Sun, A.Y. Wu, S.T. Yin, J. Am. Ceram. Soc. 91 (2008) 169.