Dielectric and ferroelectric characteristics of barium zirconate titanate ceramics prepared from mixed oxide method

Transcription

1 Journal of Alloys and Compounds 462 (2008) Dielectric and ferroelectric characteristics of barium zirconate titanate ceramics prepared from mixed oxide method F. Moura a,1, A.Z. Simões a,, B.D. Stojanovic a,b,1, M.A. Zaghete a,1, E. Longo a,1, J.A. Varela a,1 a Chemistry Institute, Department of Chemistry-Physics, UNESP C.P. 355, CEP , Araraquara, SP, Brazil b Center for Multidisciplinary Studies University of Belgrade, Department of Materials Science, Kneza Viseslava 1a, Belgrade, Serbia Received 27 December 2006; received in revised form 17 July 2007; accepted 20 July 2007 Available online 26 July 2007 Abstract Barium zirconium titanate (BZT) ceramics were prepared by mixed oxide method. X-ray diffraction showed the presence of a single phase while Raman scattering confirmed structural transitions as a function of different Zr/Ti ratio. The addition of Zr strongly influenced the crystal structure and electrical properties of the ceramics. A typical hysteresis loops were observed for all investigated compositions. BZT ceramics with 15 mol% Zr have shown a ferroelectric to paraelectric transition at around 77 C Published by Elsevier B.V. Keywords: Ferroelectrics; X-ray diffraction; Dielectric response 1. Introduction Barium zirconium titanate ceramics are attractive candidates for dynamic random access memories and tunable microwave devices. Such, lead-free, environmental friendly materials are known to exhibit relaxor behaviour in bulk materials with increasing the Zr content. The interest in high strain piezoelectric materials is increasing for electromechanical transducers and various related applications [1]. Though the large family of lead-based perovskites and relaxors has shown a great potential, lead-free compositions in these families will be of interest due to obvious environmental concern in the future [2 6]. BaTiO 3 is known to have a large electromechanical coupling factor. Substitution of Ti 4+ (atomic weight of 47.9, ionic radius of nm) with Zr 4+ (atomic weight of 91.2, ionic radius of nm) exhibits several interesting features in the dielectric behavior of Corresponding author. Tel.: ; fax: address: alezipo@yahoo.com (A.Z. Simões). 1 Tel.: ; fax: BaTiO 3 ceramics. When the Zr content is less than 10 mol%, the BZT ceramics show normal ferroelectric behavior and dielectric anomalies corresponding to cubic to tetragonal (Tc), tetragonal to orthorhombic (T 2 ), and orthorhombic to rhombohedral (T 3 ) phase transitions. At around 27 mol%, Zr-doped BZT ceramics exhibit typical diffuse paraelectric to ferroelectric phase transition behavior, whereas Zr-richer compositions exhibit typical relaxor-like behavior in which Tc shifts to higher temperature with increase of frequency [7]. The modified Ba(Zr,Ti)O 3 has shown the systematic changes in the dielectric, piezoelectric, and phase transition characteristics in the bulk ceramic and single crystal forms [8]. In the paraelectric state, just above Tc, BZT ceramics are attractive candidates for dynamic random access memories and tunable dielectric devices. It is well known that Zr, like Sr, is an effective substituent in BaTiO 3 to decrease and shift the Curie temperature below room temperature [9]. Moreover, Zr 4+ ion is chemically more stable than Ti 4+ and has a larger ionic size to expand the perovskite lattice. Therefore, the substitution of Ti by Zr would depress the conduction by electronic hopping between Ti 4+ and Ti 3+ and it would also decrease the leakage current of the BaTiO 3 system. It is reported that an increase in the Zr content induces a reduction /$ see front matter 2007 Published by Elsevier B.V. doi: /j.jallcom

2 130 F. Moura et al. / Journal of Alloys and Compounds 462 (2008) in the average grain size, decreases the dielectric permittivity (ε r ), maintaining a low and stable leakage current [10 12]. This is possible because Zr 4+ ion has larger ionic size (0.087 nm) than Ti 4+ (0.068 nm). Thus, BZT ceramics have good properties and can be applied as a storage capacitor for the next DRAM generation and a dielectric material for MLCC. Zr/Ti ratio is important in BZT system and a 0.2/0.8 ratio is known to have excellent bulk properties [13]. In this work, the effect of Zr/Ti ratio on the structural and electrical properties of BZT ceramics was investigated. 2. Experimental procedure Fig. 1. X-ray diffraction pattern of the BZT powders calcinated at 1200 C for 2 h in different concentrations: (a) Ba(Ti 0.95 Zr 0.05 )O 3 ; (b) Ba(Ti 0.9 Zr 0.10 )O 3 and (c) Ba(Ti 0.85 Zr 0.15 )O 3. The powders were homogenized in a ball mill using isopropyl alcohol. All oxides were analytical grade: BaCO 3 (Vetec), ZrO 2 (Inlab), TiO 2 (Vetec). The investigated systems were Ba(Ti 0.95 Zr 0.05 )O 3, Ba(Ti 0.9 Zr 0.10 )O 3 and Ba(Ti 0.85 Zr 0.15 )O 3. After drying, the powders were compacted in the form of pellets with 1 mm of thickness and 12 mm of diameter, isostatically pressed at 210 MPa. The pellets were sintered at 1550 C for 4 h in static air and cooled to room temperature (5 C/min). X-ray diffraction data were collected with a Rigaku Rint 2000 diffractometer under the following experimental condition: 50 kv, 150 ma, 20 2θ 80, 2θ = 0.02, λcu k monocromatized by a graphite crystal, divergence slit = 2mm, reception slit = 0.6 mm, step time = 10 s. Dilatometric analyses were performed in a Nezstch dilatometer 402E, up to 1550 C, at a heating rate of 5 C/min in static atmosphere and the results of Fig. 2. Linear shrinkage as a function of temperature for BZT powders calcinated at 1200 C for 2 h in different concentrations: (a) Ba(Ti 0.95 Zr 0.05 )O 3 ; (b) Ba(Ti 0.9 Zr 0.10 )O 3 and (c) Ba(Ti 0.85 Zr 0.15 )O 3.

3 F. Moura et al. / Journal of Alloys and Compounds 462 (2008) Fig. 3. Linear shrinkage rate as a function of temperature for BZT powders calcinated at 1200 C for 2 h in different concentrations: (a) Ba(Ti 0.95 Zr 0.05 )O 3 ; (b) Ba(Ti 0.9 Zr 0.10 )O 3 and (c) Ba(Ti 0.85 Zr 0.15 )O 3. linear shrinkage rate (d( l/l 0 )/dt) and linear shrinkage l/l 0 were determined. Raman measurements were performed using an ISAT triple monochromator. An optical microscope was employed to focus the nm radiation from a Coherent Innova 99 Ar + laser on the sample and to collect the back-scattered radiation. The scattered light dispersed by the spectrometer was detected by a charge-coupled device (CCD) detection system. Specific surface area was determined based on nitrogen adsorption isotherms by using the BET method (Sorptomatic 1990). Gold electrodes for electrical measurements were applied by evaporation through a sputtering system on a polished surfaces of sintered discs. Ferroelectric properties were measured on a Radiant Technology RT6600A tester system equipped with a micrometer probe station in a virtual ground mode. The dielectric characterization was accomplished with HP 4192 impedance analyser, and measurements of the capacitance as a function of temperature at a frequency of 10 khz were performed. From the capacitance dependence temperature curves, the Curie temperature was determined. 3. Results and discussion Fig. 1 shows the room temperature XRD plots of Ba(Ti 0.95 Zr 0.05 )O 3, Ba(Ti 0.9 Zr 0.10 )O 3 and Ba(Ti 0.85 Zr 0.15 )O 3 powders. All the ceramics composition were crystallized into a single-phase perovskite structure. This is a clear indication that the addition of Zr is forming a stable solid solution with the BaTiO 3 lattice. As reported in other studies [2 6], depending upon the Zr content, BZT may have orthorhombic, rhombohedral and tetragonal structures at room temperature. The linear shrinkage l/l 0 and the linear shrinkage rate (d( l/l 0 )/dt) as a function of temperature for different Zr/Ti ratio are shown in Figs. 2 and 3. The sintering process is strongly influenced by Zr/Ti ratio leading to a reduction in the densification temperature. The maximum shrinkage rate occurred around 1410 C for the composition Ba(Ti 0.95 Zr 0.05 )O 3. Meanwhile, a reduction in the maximum shrinkage rate was observed for the Ba(Ti 0.85 Zr 0.15 )O 3 composition, probably due to the segregation of ZrO 2 at the grain boundaries. The peak close to 980 C indicates powder agglomeration during the sintering process (intraand inter-agglomerates) as a consequence of different surface area of the raw powders [ZrO 2 (5.67), BaCO 3 (1.35) and TiO 2 (53.71)]. Room temperature Raman spectra are displayed in Fig. 4. Because of the random grain orientations in the powders, the directions of the phonon wave vectors are randomly distributed from one grain to another with respect to the crystallographic axes. The evolution of Raman spectra with Zr substituting on Ti sites shows some interesting changes. As observed, with increasing Zr content the Raman line at 123 cm 1 has not been observed. Taking the mass ratio Zr/Ti = 1.9 into consideration this mode frequency is expected to be about

4 132 F. Moura et al. / Journal of Alloys and Compounds 462 (2008) Fig. 4. Room temperature depolarized Raman spectra for BZT powders calcinated at 1200 C for 2 h in different concentrations: (a) Ba(Ti 0.95 Zr 0.05 )O 3 ; (b) Ba(Ti 0.9 Zr 0.10 )O 3 and (c) Ba(Ti 0.85 Zr 0.15 )O cm 1 for Zr replacing Ti sites, which will be reduced further by an increase in the ionic radius [R(Ti 4+ ) = nm, R(Zr 4+ ) = nm]. The additional mode could therefore be associated with a normal mode involving Zr atoms. Since this mode disappears in the Ba(Ti 0.9 Zr 0.10 )O 3 and Ba(Ti 0.85 Zr 0.15 )O 3 compositions, it may be considered an indication of the orthorhombic to rhombohedral phase transition. Such observations could not be observed in the X-ray studies due to the different coherence length and time scale involved in the process. SEM analysis of BZT powders calcinated at 1200 C for 2 h are shown in Fig. 5. We observe an increase in the agglomerates size for different Zr/Ti ratio leading to a reduction in the specific surface area. The agglomerates are rounded in form which is typical for the mixed oxide method. Values of the specific surface area of BZT powders were measured from BET analyses (Table 1). There is a tendency to decrease the specific surface area with the increase in the zirconium content due to the strong presence of agglomerates and coarse-grained structure, as observed in the SEM analysis. Measurements on dielectric permittivity as a function of temperature reveal anisotropic behavior (Fig. 6). The highest permittivity (ε r = 14,500) is observed for the Ba(Ti 0.95 Zr 0.05 )O 3 composition at 100 khz. Three distinct phase transitions were Table 1 Characteristics of BZT powders and pellets for different Ti/Zr rat Samples Surface area (nm) P r ( C/cm 2 ) E c (kv/cm) Dielectric permittivity at (100 khz) Ba(Ti 0.95 Zr 0.05 )O ,400 Ba(Ti 0.9 Zr 0.10 )O ,660 Ba(Ti 0.85 Zr 0.15 )O

5 F. Moura et al. / Journal of Alloys and Compounds 462 (2008) Fig. 5. SEM analysis for the BZT powders calcinated at 1200 C for 2 h in different concentrations: (a) Ba(Ti 0.95 Zr 0.05 )O 3 ; (b) Ba(Ti 0.9 Zr 0.10 )O 3 and (c) Ba(Ti 0.85 Zr 0.15 )O 3. observed. The Curie temperature reduces with the increase of zirconium content due the changes in crystal structure, as shown in Raman spectra. A strong reduction in the dielectric permittivity for the Ba(Ti 0.85 Zr 0.15 )O 3 system is a consequence of different crystal structure. A sharp phase transition is indicative of a ferroelectric-relaxor behavior, as observed in the literature [14 16]. Well saturated hystereses loops with regular shape, typical of ferroelectric materials were evident for all investigated systems (Fig. 7a c). There is no evidence of imprint phenomena indicating that our ceramics present low defects as oxygen vacancies which pinning the domain walls difficulting the saturation of hystereses loops. Consequently, there is no increase in the coercive voltage in one direction. The reduction in the remnant polarization for the Ba(Ti 0.85 Zr 0.15 )O 3 system can be related to changes in crystal structure which can explain the relaxor behaviour observed in Fig. 6. Fig. 6. Temperature dependence of dielectric permittivity (ε) for BZT pellets sintered at 1550 C for 4 h in different concentrations: (a) Ba(Ti 0.95 Zr 0.05 )O 3 ; (b) Ba(Ti 0.9 Zr 0.10 )O 3 and (c) Ba(Ti 0.85 Zr 0.15 )O 3.

6 134 F. Moura et al. / Journal of Alloys and Compounds 462 (2008) Fig. 7. P E hysteresis loops for BZT pellets sintered at 1550 C for 4 h in different concentrations: (a) Ba(Ti 0.95 Zr 0.05 )O 3 ; (b) Ba(Ti 0.9 Zr 0.10 )O 3 and (c) Ba(Ti 0.85 Zr 0.15 )O Conclusions Single phase BZT powders were obtained from mixed oxide method at 1200 C for 2 h. Raman spectroscopy indicated a change in the crystal structure with the increase of zirconium content. The transition temperature (ferroelectric to paraelectric phase) was found to be systematically reduced for Zr content equal to 15.0 mol%. In this case, a typical relaxor behavior was observed. A decrease in the dielectric permittivity and remnant polarization with the increase of zirconium content is indicative of changes in crystal structure and predominance of relaxor behaviour. Acknowledgments The authors gratefully acknowledge to the financial support of the Brazilian agencies FAPESP, CNPq, and CAPES and Ministry of Science and Environmental Protection of Serbia. [2] Z. Yu, C. Ang, R. Guo, A.S. Bhalla, Appl. Phys. Lett. 81 (2002) [3] A. Dixit, S.B. Majumder, R.S. Katiyar, Appl. Phys. Lett. 82 (2003) [4] P.S. Dobal, R.S. Katiyar, J. Raman, Spectroscopy 33 (2002) 405. [5] A. Dixit, S.B. Majumder, A. Savvinov, R.S. Katiyar, R. Guo, A.S. Bhalla, Mater. Lett. 56 (2002) 933. [6] D.S. Paik, S.E. Park, S. Wada, S.F. Liu, T.R. Shrout, J. Appl. Phys. 85 (1999) [7] D. Hennings, H. Schell, J. Am. Ceram. Soc. 65 (1982) 539. [8] F.W. McQuarrie, Behnke, J. Am. Ceram. Soc. 37 (1954) 539. [9] R.C. Kell, N.J. Hellicar, Acustica 6 (1956) 235. [10] P.W. Rehrig, S.E. Park, S.T. McKinstry, G.L. Messing, B. Jones, T.M. Shrout, J. Appl. Phys. 86 (1999) [11] Y. Zhi, R. Guo, A.S. Bhalla, J. Cryst. Growth 233 (2001) 460. [12] Y. Zhi, R. Guo, A.S. Bhalla, J. Appl. Phys. 88 (2000) 410. [13] Y. Zhi, R. Guo, A.S. Bhalla, Appl. Phys. Lett. 77 (2000) [14] S. Wada, S. Suzuki, T. Noma, M. Kakihana, S.E. Park, L.E. Cross, T.R. Shrout, Jpn. J. Appl. Phys., Part 1 38 (1999) [15] S.E. Park, S. Wada, L.E. Cross, T.R. Shrout, J. Appl. Phys. 86 (1999) [16] J. Ravez, A. Simon, Eur. J. Solid State Inorg. Chem. 34 (1997) References [1] Z. Yu, R. Guo, A.S. Bhalla, J. Appl. Phys. 88 (2000) 410.