Accepted Manuscript Title: The effects of MnO2 addition on the structure and dielectric properties of the strontium barium niobate glass-ceramics Authors: Shaomei Xiu, Bo Shen, Jiwei Zhai PII: DOI: Reference:
S0025-5408(17)30262-3 http://dx.doi.org/doi:10.1016/j.materresbull.2017.08.008 MRB 9486
To appear in:
Received date: Revised date: Accepted date:
17-1-2017 25-6-2017 2-8-2017
Please cite this article as: Shaomei Xiu, Bo Shen, Jiwei Zhai, The effects of MnO2 addition on the structure and dielectric properties of the strontium barium niobate glass-ceramics, Materials Research Bulletinhttp://dx.doi.org/10.1016/j.materresbull.2017.08.008 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
The effects of MnO2 addition on the structure and dielectric properties of the strontium barium niobate glass-ceramics
Shaomei Xiu, Bo Shen*, Jiwei Zhai
Key Laboratory of Advanced Civil Engineering Materials of Ministry of Education, Functional Materials Research Laboratory, School of Materials Science & Engineering, Tongji University, 4800 Caoan Road, Shanghai 201804, China
*Corresponding author. Tel.: +86 21 65980544; fax: +86 21 65985179. E-mail address: [email protected]
(1). A small amount of MnO2 doping the SBN-glass-ceramics makes BDS reach to 1470.6 kV/cm
(2). The theoretical energy storage density of the SBN-glass-ceramics with MnO2 addition gets 9.2 J/cm3.
(3). A small amount MnO2 doping the BSN-glass-ceramics makes the leakage current densities decrease to 10-6 A/cm2. And a small amount of MnO2 doping the SBN-glass-ceramics can effectively decrease the dielectric loss of the materials.
Abstract The effects of MnO2 content on the structure and dielectric properties of the SBN-glass-ceramics were studied. The results show that a small amount of MnO2 doping of the SBN-glass-ceramics can make their microstructure become denser and more uniform and MnO2 doping of the SBN-glass-ceramics, as a grain growth inhibitor, has an evident effect on the reduction of grain sizes. The Mn ions exist in the form of Mn3+ and Mn4+ ions in the SBN-glass-ceramics, as confirmed by XPS measurements and Mn3+ and Mn4+ ions easily form charge defect complexes, which causes the leakage current densities of the SBN-glass-ceramics to obviously decrease. And a small amount of MnO2 doping of the SBN-glass-ceramics can effectively decrease the dielectric loss of the materials. When 0.05 mol% MnO2 is added to the SBN-glass-ceramics, the BDS is up to 1470.6 kV/cm and the theoretical energy storage density reaches the maximum value of 9.2 J/cm3.
Keywords: Strontium barium niobate, Glass-ceramics, MnO2 addition, Dielectric loss, Energy storage property
Introduction For energy storage capacitors used in pulsed power applications, high energy storage densities and low dielectric loss are extremely important properties [1-4]. As is well known, the high energy storage densities of dielectric materials are related to the breakdown strength (BDS) and dielectric constant . Therefore, if the energy storage densities of dielectric materials are improved, these materials must have both a higher BDS and a higher dielectric constant at the same time. At present, ferroelectric glass-ceramics materials, as the most promising candidates for application in high energy storage capacitors, have attracted much attention in pulsed power applications [6, 7]. This is because the synergistic effect of the good dielectric properties from the ferroelectric phase and the defect-free nature of the glass matrix results in ferroelectric glass-ceramics that can have a high dielectric constant and a high BDS . Many investigations on the high energy storage densities of ferroelectric glass-ceramics materials have been reported [9, 10]. Zheng et al.  reported that the BDS increased from 608 to 1127 kV/cm in niobate glass-ceramics with the addition of La2O3 and the energy storage density reached 7.1 J/cm3. Chen et al.  reported that a certain amount of P2O5 can modify the microstructure and dielectric properties of niobate-based-glass-ceramics, and caused the energy storage density increase to 9.1 J/cm3. Zhang et al.  studied influence of the different methods of heat-treatment on the phase evolution and microstructure of glass-ceramics and found that microwave treatment could restrain the formation of the dendritic microstructure and improve the dielectric BDS. However, for practical energy storage capacitors materials, it is desirable not only to have high energy storage densities, but also to maintain lower dielectric loss. This is because the high energy loss in the capacitor easily leads to heating and detrimental effects on the performance and reliability of the capacitor , which requires energy storage capacitors materials with higher energy storage densities and the lower dielectric loss. Nevertheless, only a few researchers have paid attention to the reduction of the dielectric loss for ferroelectric glass-ceramics [11, 14]. MnO2 addition can effectively improve the microstructure of the materials
niobate-based-glass-ceramics with MnO2 addition. In this work, we have investigated the influence of different MnO2 addition on the phase composition, microstructure and dielectric properties of the SBN-glass-ceramics in order to obtain the material with a high energy storage density and a lower dielectric loss. Experimental procedure The SBN-glass was prepared from well mixed powders of BaCO3, SrCO3, Nb2O5, Al2O3, B2O3, SiO2 and MnO2. The powders with the composition of 20BaCO3, 20SrCO3, 20Nb2O5, 5Al2O3, 1.5B2O3, 33.5SiO2 and xMnO2 (x = 0, 0.05, 0.1, 0.2 and 0.5) (mol%) were weighted, mixed by ball milling for 20 h, and then melted in an alumina crucible at 1550 oC for 2 h. The as-quenched SBN-glass was annealed at 650 oC for 5 h to remove residual stresses. Finally these samples with different MnO2 content were heat-treated in air at 1100 oC for 3 h to convert the SBN-glasses into SBN-glass-ceramics. X-ray diffraction (XRD) (D/max 2550V B3+/PC, Rigaku, Japan) with Cu Kα radiation was used to investigate the phase structure of the SBN-glass-ceramics. Field emission scanning electron microscopy (FE-SEM) (HITACHI S-4700) was used to investigate the microstructure of the samples. The samples for the SEM measurements were prepared by polishing and etching with the solution (15 wt% HNO3+10 wt% HF) to remove the glass matrix on the surface so as to observe the crystals distribution clearly. The chemical state of MnO2 in the SBN-glass-ceramics was analyzed by X-ray photo electron spectroscopy (XPS, Thermo ESCALAB 250Xi). The current as a function of voltage (I-V) was measured by using a Keithley 2400 source meter unit interfaced with a computer to perform the measurement and record data. The temperature dependence of the dielectric constant and dielectric loss was measured using an LCR meter (Agilent E4980A) at the frequency of 10 kHz and in the temperature range from -80 oC to 120 oC. Before the dielectric properties of the crystallized samples were measured, both sides of the samples that were polished to achieve parallel and smooth faces were covered with silver paint and then fired at 600 oC for 30 min in order to form Ag electrodes. Polarization-electric (P-E) field loops were measured by a Premier II ferroelectric test system in silicone oil to avoid electrical discharges. The direct current dielectric BDS measurement was carried out with a voltage-withstand testing device (ET2671B, Entai, Nanjing, China) at room temperature. The samples for voltage-withstand measurement were ground flush to about 0.07 mm thickness. During the testing process, the samples were immersed in silicone oil to prevent arcing. Complex impedance spectrums were measured by an LCR meter (E4980A, Agilent, USA) over frequencies from
0.15 kHz to 2,000 kHz in the temperature range of 560-610 oC with an AC electric field of 4 V/mm. Results and discussion XRD patterns for the SBN-glass-ceramics with MnO2 addition are shown in Fig.1. Besides the tetragonal tungsten bronze phase Ba0.27Sr0.75Nb2O5.78, the celsian phase BaAl2Si2O8 was indexed as the second phase in all samples. The XRD patterns of the SBN-glass-ceramics with MnO2 addition are similar to each other, which indicates that variation of the MnO2 content has little effect on the phase structure of the SBN-glass-ceramics. Average grain sizes were calculated by the Scherrer formula for all samples and are shown in the inset of Fig.1. With increasing MnO2 content, the average grain sizes of the SBN-glass-ceramics have a downward trend, which indicates that MnO2 addition can restrain the growth of the grains. Fig.2 presents the SEM images of the SBN-glass-ceramics with MnO2 addition. Comparing the average grain sizes shown in the inset of Fig.1 with the particles shown in Fig.2, it can be deduced that reunited grains occur in the SBN-glass-ceramics. The serious phenomenon of reunited grains is observed in Fig.2 (a) and (e). When a small amount of MnO2 (from 0.05 to 0.2 mol%) is added, the microstructure of the SBN-glass-ceramics become denser and more uniform, as shown in Fig.2 (b), (c) and (d). This may be because a small amount of MnO2 doping of the SBN-glass-ceramics can weaken reunited grains and excess MnO2 content (0.5 mol%) intensifies the reunited grains again . Also, with increasing MnO2 content, the average grain sizes of the SBN-glass-ceramics decrease slightly. This result reveals that a certain amount of MnO2 doping the SBN-glass-ceramics, as a grain growth inhibitor and a particle dispersant, has an evident effect on grain sizes reduction and scatters the reunited particles . Fig.3 shows the curves of the leakage current densities (J) for the SBN-glass-ceramics with MnO2 addition in the field region from 0 to 90 kV/cm. In Fig.3, the leakage current densities of all samples in the higher field region are lower than 10-6 A/cm2, indicating that the materials in capacitor applications have a good reliability. The leakage current densities are lower for the SBN-glass-ceramics with MnO2 content from 0.05 to 0.2 mol% than for the undoped SBN-glass-ceramics and SBN-glass-ceramics with 0.5 mol% MnO2, which indicates that a small amount of MnO2 doping of the SBN-glass-ceramics can effectively decrease the leakage current densities. And the leakage current densities for the SBN-glass-ceramics with MnO2 addition also indirectly explain the trend of the dielectric loss shown in Fig.4 (b).
The influence of MnO2 addition on the temperature dependent dielectric properties of the SBN-glass-ceramics in the temperature range from -80 oC to 120 oC at 10 kHz is illustrated in Fig.4 (a). The dielectric constant and dielectric loss maintain a good stability within a wide temperature range. The Curie temperatures of all samples are lower than -80 oC, thus all of the samples present a paraelectric behavior at room temperature. In addition, Fig.4 (b) shows the MnO2 content dependence of the dielectric constant and dielectric loss of the SBN-glass-ceramics. The values of the dielectric constant and dielectric loss are measured at room temperature. In Fig.4 (b), the dielectric constant presents a higher value (about 80~110), which is due to the tetragonal tungsten bronze structured phase Ba0.27Sr0.75Nb2O5.78 in the SBN-glass-ceramics , and is consistent with the XRD results. With the increase of MnO2 content from 0 to 0.5 mol%, the dielectric constant is slightly decreased. This may be because the average grain sizes shown in the inset of Fig.1 gradually decrease with increasing MnO2 content in accordance with the findings of other studies [18, 19]. The dielectric loss obviously decreases from 0.012 to 0.004 as the MnO2 addition increase from 0 to 0.2 mol%. And after 0.5 mol% MnO2 content is added to the SBN-glass-ceramics, the dielectric loss rises to 0.009 again, as shown in Fig.4 (b). The dielectric loss is caused by several factors and there are also many papers that have reported that MnO2 addition reduced the dielectric loss [19, 20]. It has been suggested that the homogeneous microstructure  and leakage current [21, 22] of the materials may be the main factors that lead to the obvious variation of the dielectric loss. It is believed that the tetragonal tungsten bronze structure possesses A1/A2-sites, B1/B2-sites and C-sites . In general, it is easy for low-valence ions (such as Ba2+, Sr2+, Na1+) to enter the A1/A2-sites, and for high-valence ions (such as Nb5+, Ti4+, Zr4+, Mn4+) to get into B1/B2-sites, while C-sites are not often occupied by ions [23, 24]. Therefore, when MnO2 content is added to the SBN-glass-ceramics, Mn ions easily enter the B1/B2-site of the tetragonal tungsten bronze structure. Mn ions exist in the form of Mn3+ and Mn4+ ions in the SBN-glass-ceramics, as confirmed the XPS measurements, and Mn3+ and Mn4+ ions easily form charge defect complexes (Mn′′Nb▬V••O) and (Mn′′Nb▬V••O▬Mn'Nb) . They both lead to a lower defect concentration. By increasing the MnO2 contents from 0 to 0.2 mol%, more of (Mn′′Nb▬V••O) and (Mn′Nb▬V••O▬Mn′Nb) could be formed and the mobility of V ••O becomes more difficult, with the result that the dielectric loss originated from the leakage current is decreased. The above conclusion that the charge defect complexes
SBN-glass-ceramics decrease is consistent with the analysis of the leakage current densities of the
SBN-glass-ceramics in Fig.3. Thus, the dielectric loss obviously decreases from 0.012 to 0.004 as the MnO2 addition increases from 0 to 0.2. In addition, in the SBN-glass-ceramics with a small amount of MnO2 addition (0.05~0.2 mol%), the microstructure becomes more uniform and denser compared with the undoped SBN-glass-ceramics, which results in the dielectric loss obviously decreasing . For the SBN-glass-ceramics with 0.5 mol% MnO2 content, the dielectric loss rises to 0.009 again. The main reason may be that excess MnO2 addition makes the microstructure of the SBN-glass-ceramics become uneven. This induces the dielectric loss to rise again. Fig.5 shows the polarization-electric (P-E) field hysteresis loops of the SBN-glass-ceramics measured at room temperature. It is shown that the remanent polarization of all samples is extremely low and the varying trend of the remanent polarization is the same as that of the dielectric loss. And in Fig.4 (a) the samples present a paraelectric behavior in the temperature range from -80 to 120 oC. Therefore, in combination with the P-E hysteresis loops of the SBN-glass-ceramics, the SBN-glass-ceramics can be regarded as a linear dielectric material at room temperature. Impedance analysis  is introduced to study the effects of MnO2 addition on the interfacial polarization and the breakdown mechanism for the SBN-glass-ceramics. Series of Cole-Cole figures were obtained when the samples were measured in the temperatures range from 570 oC to 610 oC at every 10 oC interval as shown in the inset of Fig.6 (a). The measurement temperature T and lnτ obey the Arrhenius relationship ： ln
1 Ea ln 0 , where Ea is the activation energy for the kBT
relaxation process, kB is the Boltzmann constant and T is the absolute temperature. Fig.6 (a) shows the Ea of the SBN-glass-ceramics with MnO2 addition. The values of Ea can be calculated from the slope of the function between the relaxation time and measuring temperature. And these values of Ea are 1.16, 0.98, 1.00, 1.09, 1.19 eV, respectively. In our measurement condition, the activation energy Ea corresponds to the relaxation level of the space charge . The values of the activation energy Ea are lower in the SBN-glass-ceramics with MnO2 addition from 0.05 to 0.2 than in the undoped SBN-glass-ceramics and SBN-glass-ceramics with 0.5 mol% MnO2 content, which indicates that the more uniform and denser microstructure can reduce the accumulation of charge in the interfaces between the ferroelectric phase crystals and glass matrix [25, 27] in accordance with Fig.2. The accumulation of charge in the interfaces affects the BDS of materials. More charges accumulated in the interfaces can result in the BDS decreasing. Therefore, the correlation between the BDS and the Ea is
opposite. The distribution of the average BDS value can be estimated by the Weibull distribution :
X i ln Ei , Yi ln( ln(1 i /( n 1))) , where Ei is the specific breakdown voltage of each specimen in the experiments, n is the sum of specimens, i is the serial number of specimens and in this work, n=8. According to the Weibull distribution, the average BDS value can be extracted from points where the fitting lines intersect with the horizontal line through Yi=0. Fig.6 (b) illustrates the Weibull plot of the BDS for the SBN-glass-ceramics with MnO2 addition. The β values indicate that the two parameters Xi and Yi have a good linear relationship . The correlation between the BDS and the theoretical energy storage densities is shown in Fig.7. In Fig.7, the BDS of the SBN-glass-ceramics with 0.05 mol% MnO2 addition dramatically increases to 1470.6 kV/cm. And with increasing MnO2 addition from 0.05 to 0.2 mol%, the BDS of the SBN-glass-ceramics slightly decreases from 1470.6 to 1331.9 kV/cm. But after the addition of 0.5 mol% MnO2 to the SBN-glass-ceramics, the BDS of the SBN-glass-ceramics obviously decreases to 1085.3 kV/cm. The BDS of dielectric materials is determined by several factors, such as microstructure, grain size, and the dielectric constant [25, 29, 30]. Generally, the relationship between values of the BDS and grain sizes conforms to the relations: EBDS G
, where EBDS is the breakdown strength, G is the grain size, and α is a
constant. Apparently, in this study, the relationship between values of the BDS and grain sizes is not consistent with the formula (as the grain size decreases, the BDS has an increased trend.). To the authors’ knowledge, the microstructure of the SBN-glass-ceramics has a greater influence on the BDS [12, 32]. A small amount of MnO2 addition leads to enhanced densification and homogenization of the SBN-glass-ceramics, which results in the increase of the BDS of the SBN-glass-ceramics . Excess MnO2 addition makes the microstructure no uniform and a mass of reunited grains occur in the SBN-glass-ceramics, which is the main reason that the BDS obviously is decreased. As stated above, the activation energy Ea in Fig.4 (a) also indirectly proves the variation trend of the BDS for SBN-glass-ceramics with MnO2 addition. The SBN-glass-ceramics can be regarded as linear dielectric materials at room temperature. The theoretical energy storage densities of the SBN-glass-ceramics can be evaluated by the formula :
1 W 01 E 2 . The theoretical energy storage densities of the SBN-glass-ceramics at room 2 temperature are shown in Fig.7. It is obvious that the trend between the BDS and theoretical energy
storage densities of the SBN-glass-ceramics is coincident. Although the dielectric constant of the SBN-glass-ceramics with different MnO2 content is slightly decreased, the BDS is obviously varied. And the BDS of the SBN-glass-ceramics has the main influence on the theoretical energy storage density. The theoretical energy storage density of the SBN-glass-ceramics with 0.05 mol% MnO2 is higher than that of the other sample and reaches 9.2 J/cm3. Conclusions Strontium barium niobate (SBN)-based glass-ceramics with different MnO2 content were synthesized through melt casting followed by controlled crystallization. The MnO2 content mostly impacts on the microstructure, dielectric properties and energy storage properties of the SBN-glass-ceramics. A small amount of MnO2 can enhance the densification and homogenization of the SBN-glass-ceramics, while the dielectric loss of the SBN-glass-ceramics obviously decreases from 0.012 to 0.004 and the BDS of the SBN-glass-ceramics increases from 1152.0 to 1470.6 kV/cm. The theoretical energy-storage density of the SBN-glass-ceramics is found to be 9.2 J/cm3. Dielectric materials with a higher BDS and lower dielectric loss are technologically very important. Acknowledgments The authors would like to acknowledge the support from National Key Fundamental Research Program (2015CB654601).
References  E.P. Gorzkowski, M.J. Pan, B.A. Bender, C.C.M. Wu, J. Am. Ceram. Soc., 91 (2008) 1065-1069.  E.P. Gorzkowski, M.J. Pan, B.A. Bender, C.C.M. Wu, J Electroceram. 18 (2007) 269-276.  K.M. Slenes, P. Winsor, T. Scholz, M. Hudis, IEEE Transactions on Magnetics, 37 (2001) 324-327.  J. Luo, J. Du, Q. Tang, C. Mao, IEEE Transactions on Electron Devices, 55 (2009) 3549-3554.  R. Gordon, J. Am. Ceram. Soc. 73 (2005) 323-328.  A. Herczog, J. Am. Ceram. Soc. 47 (1964) 107–115.  C.T. Cheng, M. Lanagan, B. Jones, J.T. Lin, M.J. Pan, J. Am. Ceram. Soc. 88 (2005) 3037–3042.  A.L. Young, G.E. Hilmas, S.C. Zhang, R.W. Schwartz, J. Mater. Sci. 42 (2007) 5613-5619.  N.J. Smith, B. Rangarajan, M.T. Lanagan, C.G. Pantano, Mater. Lett. 63 (2009) 1245-1248.  J. Zheng, G.H. Chen, C.L. Yuan, C.R. Zhou, X. Chen, Q. Feng, M. Li, Ceram. Int.42 (2016) 1827-1832.  G.H. Chen, J. Zheng, C.L. Yuan, C.R. Zhou, X.L. Kang, J.W. Xu, Y. Yang, Mater. Lett. (2016) 46-48.  W. Zhang, S. Xue, S. Liu, J. Wang, B. Shen, J. Zhai, J. Alloys. Compd. 617 (2014) 740-745.  H. Tang, H.A. Sodano, Nano Lett. 13 (2013) 1373-1379.  S. Xue, S. Liu, W. Zhang, J. Wang, L. Tang, B. Shen, J. Zhai, J. Alloys. Compd. 617 (2014) 418-422.  X.J. Li, Q. Wang, Q.L. Li, 20 (2008) 89-94.  J. Hao, Z. Xu, R. Chu, Y. Zhang, G. Li, Q. Yin, Mater. Chem. Phys. 118 (2009) 229-233.  S.B. Deshpande, H.S. Potdar, P.D. Godbole, S.K. Date, J. Am. Ceram. Soc. 75 (1992) 2581-2585.  S. Xiu, S. Xiao, S. Xue, B. Shen, J. Zhai, J. Electron. Mater. 45 (2015) 1-6.  J. Chen, Z. Yong, C. Deng, X. Dai, J. Am. Ceram. Soc. 92 (2009) 1863-1866.  X.R. Wang, Y. Zhang, T. Ma, C.S. Deng, X.M. Dai, Ceram Int. 38 (2012) S57-S60.  L. Wang, W. Ren, W. Ma, M. Liu, P. Shi, X. Wu, Aip Advances, 5 (2015) 4925-4951.  Y. Kizaki, Y. Noguchi, M. Miyayama, Appl. Phys. Lett.89 (2006) 142910-142913.  L.A. Bursill, P.J. Lin, Acta Crystallogr Sect B. 43 (1987) 49-56.  S. Podlozhenov, H.A. Graetsch, M. Ulex, J. Schneider, M. Wöhlecke, K. Betzler, Acta Crystallogr Sect B. 62 (2007) 960-965.  J. Huang, Y. Zhang, T. Ma, H. Li, H. Zhang, Appl. Phys. Lett. 96 (2010) 2902-042902.  X. Wang, Y. Zhang, X. Song, Z. Yuan, T. Ma, Q. Zhang, C. Deng, T. Liang, J Eur Ceram Soc. 32 (2012) 559-567.  J. Wang, L. Tang, B. Shen, J. Zhai, J. Mater. Res. 29 (2014) 288-293.  A. Kishimoto, K. Koumoto, H. Yanagida, J. Mater. Sci.24 (1989) 698-702.  J. Song, G.H. Chen, C.L. Yuan, Y. Yang, Mater. Lett. 117 (2014) 7-9.  L. Tang, W. Wang, B. Shen, J. Zhai, L.B. Kong, J. Electron. Mater.44 (2015) 227-234.  T. Tunkasiri, G. Rujijanagul, J. Mater. Sci. Lett. 15 (1996) 1767-1769.  S. Xiu, S. Xiao, W. Zhang, S. Xue, B. Shen, J. Zhai, J. Alloys. Compd. 670 (2016) 217-221.
Fig.1 XRD patterns of the SBN-glass-ceramics with MnO2 addition.
Fig.2 SEM images of the SBN-glass-ceramics with MnO2 addition.
Fig.3 The curves of the leakage current densities (J) for the SBN-glass-ceramics with MnO2 addition
Fig.4 (a) Temperature dependence of the dielectric constant and dielectric loss of the SBN-glass-ceramics with MnO2 addition. Fig.4 (b) MnO2 addition dependence of dielectric constant and dielectric loss of the SBN-glass-ceramics with MnO2 addition at room temperature.
Fig.5 Polarization-electric field (P-E) hysteresis loops of the SBN-glass-ceramics with MnO2 addition
Fig.6 (a) Relaxation time as a function of 1000/T for the BSN-glass-ceramics with MnO2 addition and the inset of Fig.6 (a) shows the complex impedance spectra measured at various temperatures for sample with 0.05 mol% MnO2 addition. Fig.6 (b) Weibull plots of the BDS values of the BSN-glass-ceramics with MnO2 addition
Fig.7 MnO2 addition dependence of the BDS and theoretical energy-storage density of the BSN-glass-ceramics with MnO2 addition