Influence of processing parameters on the structure and properties of barium strontium titanate ceramics

Influence of processing parameters on the structure and properties of barium strontium titanate ceramics

Materials Research Bulletin 43 (2008) 1989–1995 www.elsevier.com/locate/matresbu Influence of processing parameters on the structure and properties o...

906KB Sizes 0 Downloads 24 Views

Materials Research Bulletin 43 (2008) 1989–1995 www.elsevier.com/locate/matresbu

Influence of processing parameters on the structure and properties of barium strontium titanate ceramics Sining Yun a,b,*, Xiaoli Wang b, Delong Xu a a

School of Material Science and Engineering, Xi’an University of Architecture & Technology, Xi’an 710055, China b Department of Materials Physics, School of Science, Xi’an Jiaotong University, Xi’an 710049, China Received 3 May 2007; received in revised form 7 August 2007; accepted 4 October 2007 Available online 10 October 2007

Abstract Barium strontium titanate (BST) with the molar formula (Ba0.8Sr0.2TiO3) has been prepared by two different processing methods: mixed-oxide (BST-MO) and reaction-sintering (BST-RS). X-ray powder diffraction study shows differences in grain size and crystal symmetry for both these ceramics. The former shows a tetragonal symmetry while the latter presents a cubic symmetry. The occurrence of polar micro-regions associated with the higher chemical non-homogeneous distribution of ion defects from the influence of the processing parameters is the main reason for the higher peak dielectric constant (Km), the higher remanent polarization (Pr), the higher coercive field (Ec), the higher peak current density (Jm), and the lower temperature of peak dielectric constant (Tm) in BST-MO ceramics. # 2007 Elsevier Ltd. All rights reserved. Keywords: A. Ceramics; C. X-ray diffraction; D. Dielectric properties; D. Ferroelectricity

1. Introduction Ferroelectric materials of BaTiO3-based solid solutions with perovskite structure are extensively used in the electronic industry [1–3]. Barium strontium titanate (Ba1 xSrxTiO3, BST) [4–17], one of these ferroelectric materials with good properties, has attracted much attention. In recent years, it has been focused on the processing methods, such as precipitation [8,9], co-precipitation [10], sol–gel [5,6,11,12], hydrothermal [13,14], polymeric precursor [4], continuous supercritical [15], mechanochemical [17] and conventional methods. Compared with the conventional method, the synthesis of BST powder through chemical methods offers several advantages, such as high-purity, homogeneity and precise composition. However, most of these chemical methods require high-purity inorganic or organometallic reactants that are more expensive than the widely available oxides and carbonates. Moreover, these processing methods for synthesizing BST are more complicated than the conventional one. In this sense, this conventional method seems to be the most available. Although there are many methods reported to prepare BST powders, little research has been published regarding the effect of different processing methods on the microstructure and electrical properties of BST ceramics. The purpose of the present paper is to obtain explicit results for the influence of two ceramics processing methods (mixedoxide and reaction-sintering) on the microstructure, the dielectric properties and ferroelectric properties of BST * Corresponding author. Tel.: +86 29 82205280; fax: +86 29 85535724. E-mail address: [email protected] (S. Yun). 0025-5408/$ – see front matter # 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.materresbull.2007.10.003

1990

S. Yun et al. / Materials Research Bulletin 43 (2008) 1989–1995

ceramics. The corresponding mechanisms of the different processing methods on the microstructure and electrical properties of BST ceramics are also discussed. 2. Experimental BST ceramics with the nominal composition (Ba0.8Sr0.2)TiO3 samples were prepared by two different processing methods. In the first method (mixed-oxide), the TiO2 (98%) was milled with the corresponding amounts of BaCO3 (99%) and SrCO3 (99%). The dried mixture was calcined at 1150 8C for 5 h and then milled to obtain the mixed-oxide prepared BST ceramics precursors, designated hereafter BST-MO. In the second method (reaction-sintering), the first stage was the preparation of the precursor BaTiO3 and SrTiO3. The appropriate amounts of reagent grade BaCO3 (99%), SrCO3 (99%) and TiO2 (98%) powders were ball-milled in agate balls media with alcohol for 4 h and, after drying, the mixture BaCO3 + TiO2 and SrCO3 + TiO2 were aircalcined in an alumina crucible at 1150 8C for 5 h to obtain BaTiO3 and SrTiO3 pre-reacted powders. The corresponding amounts of pre-reacted powders BaTiO3 and SrTiO3 were ball-milled in alcohol for 6 h using agate balls in a planetary mill, the slurry was dried, and then calcined at 1250 8C for 4–5 h. The calcined powder were ballmilled and dried again to obtain the homogeneous reaction-sintering prepared BST ceramics precursors, and the resulting materials after sintering are designated hereafter as BST-RS. The two different methods prepared powders were pressed into pellets of 10 mm in diameter and 1 mm thick using 5% PVA binder. The binder was burnt out by slow heating up to 500 8C for 3–4 h. The samples were sintered at 1300–1400 8C in air for 3 h, with heating rates of 200 8C/h. The samples were cooled with the furnace. X-ray powder diffraction (XRD) patterns at room temperature were obtained on an automated Rigaku D/max 2400 X-ray diffractometer with rotating anode using Cu Ka radiation. The microstructures were examined by a scanning electron microscopy (SEM). After polishing, the dimensions were measured before silver electrodes were deposited on the pellets, then the specimens were fired at 810 8C for 10 min. Dielectric properties without bias field at frequencies ranging from 100 Hz to 1 MHz were measured with an Agilent 4284 A LCR meter, as samples were heated at a rate of 2 8C/min from 20 to 200 8C. Hysteresis loops were measured using a computer controlled, modified Sawyer–Tower circuit at frequency of 1 Hz. Current-field relation was measured on an automatic ferroelectric test system of aixACT TF-ANALY2ER2000. Applied electric field signal was triangular, and the period time was 1 s. 3. Results and discussion SEM observation of the pre-reacted mixture (Fig. 1(a) BaCO3 + TiO2 and (b) SrCO3 + TiO2) shows that they are highly agglomerated. These agglomerates in both powders are made up of similar particles in size. XRD patterns

Fig. 1. SEM micrographs of (a) pre-reacted BaCO3 + TiO2 mixture and (b) SrCO3 + TiO2 powder calcined at 1150 8C for 5 h.

S. Yun et al. / Materials Research Bulletin 43 (2008) 1989–1995

1991

Fig. 2. XRD patterns of BaTiO3 and SrTiO3 powder calcined at 1150 8C for 5 h.

(Fig. 2) show that the perovskite phase is the principal one in BaCO3 + TiO2 and SrCO3 + TiO2 powder calcined at 1150 8C for 5 h. For the former, the pre-calcination leads to almost complete reaction, forming a large amount of the perovskite phase. But for the latter, the temperature is not sufficient to complete the reaction, and the second phase can be observed in Fig. 2. SEM observation of the pre-reacted powder (Fig. 3(A) BST-RS and (B) BST-MO) shows that they are also highly agglomerated. These agglomerates in BST-RS and BST-MO powder are not made up of similar particles in size and are more remarkable in the BST-MO powder. Fig. 4 shows XRD patterns of BST-MO and BST-RS ceramics samples sintered at 1360 8C for 3 h. The crystal structure exhibits tetragonal symmetry for BST-MO ceramics while cubic symmetry for BST-RS ceramics at room temperature. In all of them a single phase, perovskite solid solutions, was observed. No reflection associated with the SrTiO3 was found. These results confirm that the tetragonality is related to the processing method, which is not in agreement with that the crystal symmetries should depend on the Ba/Sr ratio of BST solid solutions [18,19]. Fig. 5 shows typical SEM micrographs of BST ceramics sintered at 1360 8C for 3 h. A homogeneous microstructure was observed in BST-MO ceramics. It can be seen that a more complicated mixture of grain size occurred with the rise of pre-reacting and sintering time. In the samples sintered for BST-RS ceramics abnormal large grains were formed in a matrix of smaller grains. Similar results were obtained in the literature [20,21]. It can be assumed that the increase in diffraction maxima for BST-RS ceramics (Fig. 4) might be caused by a different content of

Fig. 3. SEM micrographs of (A) pre-reacted BST-RS powder calcined at 1250 8C for 5 h and (B) BST-MO powder calcined at 1150 8C for 5 h.

1992

S. Yun et al. / Materials Research Bulletin 43 (2008) 1989–1995

Fig. 4. XRD patterns of BST ceramics samples sintered at 1360 8C for 3 h.

the crystal phase in the ceramics. It is obvious that a small-grained sample contains a phase with smaller particle size within the notably more developed total grain boundary (Fig. 5), indicating the small-grained ceramics containing a considerably less crystal phase in its volume. It appears that the different microstructure evolution of BST ceramics materials would be affected by the processing methods. The microstructure of BST-RS ceramics materials indicates that the reaction-sintering method may be not very preferable for obtaining dense and homogeneous BST ceramics. Both types of sintered ceramics show the characteristic dispersion of the dielectric constant. Electrical properties of BST ceramics sintered at 1360 8C for 3 h were listed in Table 1. Fig. 6 shows the temperature dependence of the dielectric constant of BST ceramics sintered at 1360 8C for 3 h for different frequencies of the measuring field. The peak dielectric constant, Km, is strongly dependent on the frequency of the electric field applied to both these ceramics, whereas the temperature of peak dielectric constant, Tm, is weakly dependent on the frequency (Fig. 6). Comparing both types of sintered ceramics with a considerable high dielectric constant, Km value of BST-MO ceramics samples is about 36% higher than that of BST-RS. Though the compositions of these two samples are same (Ba0.80Sr0.20TiO3), their Tm value is somewhat different: one is 53 8C for BST-MO and the other is 54 8C for BST-RS. The dielectric properties of ferroelectric ceramics materials result from the sum of the intrinsic single-crystal response of the system and extrinsic effects, similar to those associated with the dynamic of the ferroelectric domains and defects. When the behavior of the two BST samples is compared, it can be observed that grain size diminution

Fig. 5. Typical SEM micrographs of (A) BST-MO and (B) BST-RS ceramics sintered at 1360 8C for 3 h.

S. Yun et al. / Materials Research Bulletin 43 (2008) 1989–1995

1993

Table 1 Electrical properties of BST-MO and BST-RS ceramics sintered at 1360 8C for 3 h Ceramics

Km

Tm (8C)

Pr (mC/cm2)

Ec (kV/cm)

Jm (A/cm2)

BST-MO BST-RS

17,557 12,945

53 54

5.95 4.91

1.2 0.9

4.79  10 2.61  10

4 4

Fig. 6. The temperature dependence of the dielectric constant of BST ceramics sintered at 1360 8C for 3 h for the different frequencies (0.1, 1, 10, 100, and 1000 kHz, from top to bottom).

causes a decrease of the temperature of the cubic to tetragonal transition (see Figs. 4–6). This transition is accompanied by a spontaneous strain and a small volume increase. For ceramics, the grain is constrained, the volume change causes internal stresses to develop, and which are minimized by the formation of 908 twins. Domain twinning in ferroelectric ceramics is dependent on grain size, and a banded lamellar structure relieves stress in three dimensions for coarse grain ceramics while a simple lamellar structure relieves stress in two dimensions for fine grain ceramics (<10 mm) [22,23]. For even finer grained samples, 908 domains cannot form. If there is not a 908 domain, the switching occurs only by 1808 domain wall nucleation and growth [17]. The smaller grain (10 mm) can be observed in BST samples in Fig. 5. The 908 domains can form in BST ceramics prepared in present work because a threshold of the disappearance of 908 domain is not reached. The experimental results indicate that this threshold is around 0.4 mm [23] and ferroelectric switching has been observed down to 50 nm [24]. The decrease of grain size for BST-MO ceramics causes the increase of Km in the ferroelectric phase, the reduction of Tm of the ferroelectric-paraelectric phase

Fig. 7. Field dependence of polarization at room temperature for BST ceramics samples sintered at 1360 8C for 3 h.

1994

S. Yun et al. / Materials Research Bulletin 43 (2008) 1989–1995

Fig. 8. Field dependence of current density at room temperature for BST ceramics samples sintered at 1360 8C for 3 h.

transition and the broadening of the anomaly (see Figs. 5 and 6 and Table 1). This result is similar to the result in the reports [17,23,25]. Figs. 7 and 8 show that field dependence of polarization (P) and current density (J) at room temperature for BST ceramics samples at 1 Hz, respectively. A well-behaved hysteresis loop can be observed. BST-RS ceramics presents four unusual J peaks (Fig. 8), indicating that the J peak might be related to either induced polar-regions or the extension of polar-regions [26]. The value of J between the two peaks varies larger with E, and P–E relationship is nonlinear. Comparing both ceramics, the small difference of Tm was found between two synthesizing methods. This phenomenon can be explained by the fact that the transition temperature of this compound is not only determined by the chemical composition of the solid solution. It should be noted that remanent polarization (Pr), coercive field (Ec) and peak current density (Jm) values increase in the samples (BST-MO) prepared by the first method (Figs. 7 and 8 and Table 1). This means that the properties of the same-composition compounds depend on various factors. It is necessary to discuss the factor related to domains. 908 domain and 1808 domain can form in the samples with grain size (10 mm). 908 walls are more difficult to move than 1808 walls are, and thus Pr, Ec and Jm are also higher [17]. For our samples prepared, Pr, Ec and Jm of BST-MO samples are higher than those of BST-RS samples. This is consistent with a transition from a 908 lamellar domain structure to one without 908 domains, for which polarization switching occurs by 1808 domain wall nucleation and growth. On the other hand, the occurrence of the ion and electron space charges associated with ion defects, in particular, the oxygen vacancies, should be taken into consideration when interpreting behavior of the dielectric characteristics. The possible non-homogeneous distribution of the space charge assembling on the various interfaces participate in the screening process of the spontaneous polarization, in part of the domains and grains, particularly in those situated in the surface layers of the ceramics and grains. The pinning effect excludes a number of domains from the reversal process and causes a decrease of the Pr, Ec and Jm determined from hysteresis loop measurements in BST-RS ceramics (Figs. 7 and 8 and Table 1). Obviously, the abnormal behavior observed in BST samples is attributed to the occurrence of polar micro-regions, which is associated with the non-homogeneous distribution of ion defects from the influence of the processing parameters. 4. Conclusions The processing method used for preparing BST ceramics has a significant influence on the final properties of the ceramics. When the mixed-oxide method (BST-MO) is used, the chemical homogeneity of the ceramics increases leading to 36% increase of the peak dielectric constant at Tm. In the reaction-sintering method (BST-RS), abnormal grain size forms. The microstructure evolution of ceramics materials confirms the importance of the processing method. The values of peak dielectric constant (Km), remanent polarization (Pr), coercive field (Ec) and peak current density (Jm) of BST-MO ceramics were higher than those of BST-RS ceramics. Four remarkable J peaks were observed in BST-RS, which confirms again the influence of the processing method on the final properties of the

S. Yun et al. / Materials Research Bulletin 43 (2008) 1989–1995

1995

ceramics. The difference of the structure and properties for BST ceramics prepared by different processing methods is attributed to the occurrence of polar micro-regions associated with the higher chemical non-homogeneous distribution of ion defects from the influence of the processing parameters. Acknowledgements This work was supported by the National Nature Science Foundation of China through grant No. 50272052. The authors are grateful to professor Bingfeng Yu of Xi’an Jiaotong University for the critical reading and the English editing of the manuscript. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

C. Pithan, D. Hennings, R. Waser, Int. J. Appl. Ceram. Technol. 2 (2005) 1. R.W. Whatmore, J. Electroceram. 13 (2004) 139. J.F. Scott, Annu. Rev. Mater. Sci. 28 (1998) 79. P.R. Arya, P. Jha, A.K. Guanguli, J. Mater. Chem. 13 (2003) 415. C. Shen, Q.F. Liu, Q. Liu, Mater. Lett. 58 (2004) 2302. C. Mao, X. Dong, T. Zeng, H. Chen, F. Cao, Ceram. Int. 34 (2008) 45. S.S. Gevorgian, E.L. Kollgerg, IEEE Trans. Microwave Theory Technol. 49 (2001) 792. J. Qi, Y. Wang, W. Chen, L. Li, H. Chan, J. Solid State Chem. 178 (2005) 279. I. Selvam, V. Kumar, Mater. Lett. 56 (2002) 1089. M.I. Yanovskaya, N.V. Goluubko, E.A. Nenashova, Inorg. Mater. 32 (1996) 200. P.P. Phule, S.H. Risbud, J. Mater. Sci. 25 (1990) 2571. X. Yang, X. Yao, L. Zhang, Ceram. Int. 30 (2004) 1525. X. Wei, N. Padture, Ceram. Process. Res. 5 (2004) 175. B. Gersten, M. Lencka, R. Riman, J. Am. Ceram. Soc. 87 (2004) 2025. H. Reveron, C. Elissalde, C. Aymonier, C. Bousquet, M. Maglione, F. Cansell, Nanotechnology 17 (2006) 3527. J. Liu, Z. Shen, M. Nygren, J. Am. Ceram. Soc. 89 (2006) 2689. T. Hungria, M. Alguero, A. Hungria, A. Castro, Chem. Mater. 17 (2005) 6205. L. Zhou, P.M. Vilarinho, J.L. Baptista, J. Eur. Ceram. Soc. 21 (2001) 531. H. Guo, J. Cannata, K.K. Shung, J. Mater. Sci. 40 (2005) 1509. L. Szymczak, Z. Ujma, J. Handerek, J. Kapusta, Ceram. Int. 30 (2004) 1003. J.M. Siqueiros, J. Portelles, S. Garcia, M. Xiao, S. Aguilera, Solid State Commun. 112 (1999) 189. G. Arlt, Ferroelectrics 104 (1990) 217. Z. Zhou, V. Buscaglia, M. Viviani, M.T. Buscaglia, L. Mitoseriu, A. Testino, M. Nygren, M. Jonson, P. Nanni, Phys. Rev. B 70 (2004) 024107. M.T. Buscaglia, V. Buscaglia, M. Viviani, J. Petzelt, M. Savinov, L. Mitoseriu, A. Testino, P. Nanni, C. Harnagea, Z. Zhou, M. Nygren, Nanotechnology 15 (2004) 1113. [25] W. Luan, L. Gao, H. Kawaoka, T. Sekino, K. Niihara, Ceram. Int. 30 (2004) 405. [26] X. Wang, W. Cao, Appl. Phys. Lett. 90 (2007) 042913.