Effects of cerium doping on dielectric properties and defect mechanism of barium strontium titanate glass-ceramics

Effects of cerium doping on dielectric properties and defect mechanism of barium strontium titanate glass-ceramics

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx Contents lists available at ScienceDirect Journal of the European Ceramic Society journal...

2MB Sizes 0 Downloads 17 Views

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc

Original Article

Effects of cerium doping on dielectric properties and defect mechanism of barium strontium titanate glass-ceramics Zhangyuan Zhao, Xuewei Liang, Tianyuan Zhang, Kangjia Hu, Shenhou Li, Yong Zhang



Beijing Key Laboratory of Fine Ceramics, State Key Laboratory of New Ceramics and Fine Processing, Institute of Nuclear and New Energy Technology, Tsinghua University, Beijing, 100084 PR China

A R T I C LE I N FO

A B S T R A C T

Keywords: Barium strontium titanate Cerium substitution Defect mechanism Dielectric properties Oxygen vacancy

The effect of cerium content on phase evolution, dielectric properties and defect mechanism has been investigated in (Ba,Sr)TiO3 glass-ceramics. Cerium mainly acts as an isovalent dopant in the B-site of ABO3 perovskite structure at low content (1 mol%) and then cerium substitution gradually occurs in the A-site with increasing cerium content. A compensation mechanism related to variation in oxygen vacancy concentration has been identified. When cerium content increased to 2 mol%, the maximum values of dielectric constant and energy storage density were simultaneously achieved. The impedance spectra revealed the highest conductivity. It is due to the increase in the concentration of charge carriers accompanied by the decrease in the activation energy of oxygen vacancy migration. With a further addition of cerium to 3 mol%, the opposite trend was observed. The result is related to the presence of more cation vacancies, which, in turn, limits the diffusion rate of oxygen vacancy.

1. Introduction In recent years, barium strontium titanate (Ba,Sr)TiO3 (BST) ferroelectric glass-ceramics have shown great promise for application as the energy storage capacitors in pulsed power devices, high power microwaves, and distributed power systems because of their excellent properties [1–3]. BST ferroelectric glass-ceramics with both high dielectric constant and large dielectric breakdown strength can be achieved based on the synergistic effects of admirable dielectric properties from the ceramic and the defect-free nature of the glass [4,5]. It is well known that various dopants have great effects on the dielectric properties of the ferroelectrics under external fields. Thus, the desired performance of these ferroelectric materials can be tailored through careful control of the doping species as well as doping concentration at A-site or B-site in ABO3 perovskite structure [6,7]. Up to now, a variety of perovskite structure solid solutions with rare-earth ions doping have achieved significance. In this context, numerous solid solutions were doped with rare-earth ions as donors. These elements include La [8,9], Nd [10], Sm [11] and Eu [12]. While smaller rareearth ions, such as Yb, prefer to occupy B-site acting as acceptors [13,14]. Some intermediate rare-earth ions, such as Ce, Ho and Er, show an amphoteric site-occupation preference for both A/B-site by the self-compensation mode, but the lower solid solubility of Ho and Er limits their applications as compared with Ce [13,15,16]. ⁎

Cerium is an attractive element with unique properties benefiting from its ability to shift easily between different valence states (Ce3+ and Ce4+). Considering the ionic radii, Ce3+ should exchange with dodencahedrally coordinated A-site and Ce4+ could enter octahedrally coordinated B-site [17]. In general, Ce4+, which itself is more stable than Ce3+ under high oxygen partial pressure, can easily be reduced to Ce3+ under low oxygen partial pressure or reducing atmosphere. But some studies have shown that CeO2 could react with other oxides to form stable crystalline phases containing Ce3+ even under an oxidizing atmosphere [17,18]. When CeO2 is added to barium titanate BaTiO3 (BT) with an excess amount of TiO2, it is easier for Ce to enter the Ba2+ sites in the form of Ce3+ ion during the sintering process [19]. These reports have revealed that whether Ce enters A-site or B-site are affected by the ratio of Ba/Ti in the samples and sintering atmosphere. It is demonstrated that cerium substitution can take place at both sites (A and B) in the ABO3 perovskite unit cell. And the corresponding site occupancy preference will result in different properties. Therefore, it is of prime importance to investigate the cerium occupancy and defect mechanism in the ABO3 perovskite lattice. Hwang et al reported that the substitution of Ce3+ for A-site can reduce the electrical insulation of BT ceramics [20,21]. These ceramics exhibited a typical donor-doped behavior. In addition, this substitution also decreases the spontaneous polarization of BT and then leads to a significant weakening in ferroelectricity [20,22]. By contrast, a

Corresponding author. E-mail address: [email protected] (Y. Zhang).

https://doi.org/10.1016/j.jeurceramsoc.2019.10.023 Received 21 June 2019; Received in revised form 8 October 2019; Accepted 12 October 2019 0955-2219/ © 2019 Elsevier Ltd. All rights reserved.

Please cite this article as: Zhangyuan Zhao, et al., Journal of the European Ceramic Society, https://doi.org/10.1016/j.jeurceramsoc.2019.10.023

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx

Z. Zhao, et al.

Table 1 Original compositions of the cerium doped BST glass-ceramics. Sample C0 C1 C2 C3

CeO2

BaO

SrO

TiO2

Al2O3

SiO2

0 1 2 3

29.6 28.8 28 27.2

7.4 7.2 7 6.8

29 29 29 29

12 12 12 12

22 22 22 22

Unit: mol%.

decreased transition temperature and a pronounced dielectric relaxation behavior have been observed for the addition of Ce4+ at B-site [23–25]. It is also reported that the maximum energy storage density can be obtained for the Ce-doped BT ceramics [25,26]. However, there are few studies on the energy storage properties and defect mechanism in the Ce-doped BST glass-ceramics. In this work, the BST glass-ceramics with cerium content of 0, 1, 2 and 3 mol% have been successfully prepared by a melt-quenching technique. The four kinds of glass ceramic samples are denoted as C0, C1, C2, and C3 respectively. The focus of the present paper is to evaluate the influence of cerium content on the phase evolution, defect mechanism, dielectric properties, energy storage properties and impedance behavior of the BST glass-ceramics.

Fig. 1. Circuit diagram of the setup for discharge curve measurement.

R4. Here, R4 = 3 MΩ was chosen to keep the discharge time appropriate. In order to minimize the influence of resistance on the voltage across the sample, the resistance value of R1 was selected to be much higher than that of R4, namely, R1 > > R4. In the meantime, R2 + R3 > > R1. For the impedance characterization, the sample was examined with an LCR meter (HP 4284A, Palo Alto, CA) over frequencies from 20 Hz to 1 MHz at a temperature range from 300 °C to 550 °C. 3. Results and discussion 3.1. Phase evolution

2. Experimental procedure

Fig. 2(a) shows the X-ray diffraction patterns of the BST glass-

Glass samples in this study were prepared from well-mixed powders of BaCO3, SrCO3, TiO2, Al2O3, SiO2, and CeO2 to achieve various compositions as listed in Table 1. The powders were melted in a platinum crucible at 1550 °C for 2–3 h and then poured rapidly into a container filled with distilled water to obtain glass frits. The glass frits were subsequently ball-milled for 4 h to produce glass powders using ethyl alcohol and zirconia balls as grinding media. Then the powders were mixed with polyvinyl alcohol (PVA) and uniaxially pressed into small pellets with 8 mm in diameter and 0.30 mm in thickness under the pressure of 5 MPa. A two-step sintering process was used for nucleation and subsequent crystal growth. The green pellets were first heated at 700 °C for 4 h and then sintered at 950 °C for 2 h. All samples were painted with silver paste and fired at 600 °C for 20 min to prepare electrodes for dielectric properties, energy storage properties and impedance measurements. X-ray diffraction (XRD, Model-D8 Advance, Bruker AXS, Germany) was used to investigate the phase evolution. The BST glass-ceramics were ground into powders for XRD in order to avoid the adverse influence of solid internal stress. Bulk density of the sintered specimen was measured by Archimedes method using an electronic balance with a minimum scale division of 0.0001 g. The surface chemical information was obtained from X-ray photoelectron spectroscopy (XPS, 250XI, Thermo Scientific, UK) using a physical electronics system (Model 1361), equipped with an Al Kα X-ray source (1486.6 eV). The binding energy (BE) was obtained from high-resolution scans (pass energy 30 eV and step size 0.05 eV). All peaks were calibrated to a standard C 1s peak at 284.8 eV. The measurements of dielectric constant and dielectric loss for the glass-ceramics were performed using an LCR meter (E4980A, Agilent, USA) in the temperature range of −100 to 300 °C at 1 kHz. The released energy storage density of the samples was measured with a specially designed measurement system based on a resistance load circuit, as shown in Fig. 1 [27]. Firstly, the sample was charged by a high voltage power supply (Trek 609-B, NY, USA) through resistance R1, and then discharged through resistance R4 by a high voltage relay. The relay was controlled by a waveform generator with a rise time of 100 ms and a width of 1 s. An oscilloscope (MDO3014, Tektronix, USA) was used to record the voltage U3 across the resistance R3 and then the discharge energy storage density of the sample was calculated based on R2, R3 and

Fig. 2. (a) X-ray diffraction patterns for the BST glass-ceramics with various cerium concentrations; (b) depicts the evolution of the (110) perovskite reflection; (c) cerium content dependence of the lattice parameters for the BST glass-ceramics. The inset shows the cerium content dependence of c/a ratio. 2

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx

Z. Zhao, et al.

Fig. 3. Lattice distortion caused by cerium occupying the (a) Ti4+ sites and (b) Ba2+/Sr2+ sites in the BST lattice. The red bar is the TieO bond.

ceramics with different cerium contents. The enlarged XRD patterns in the 2θ range from 31.25° to 32.25° are illustrated in Fig. 2(b). For C0, C1 and C2, both the major phase (Ba,Sr)TiO3 (PDF No. 44-0093) and the minor phase BaAl2Si2O8 feldspar (PDF No. 26-0137) were observed. However, a small amount of CeO2 crystal phase (PDF No. 34-0394) appears in C3, indicating that the substitution of cerium into lattices has reached a critical value. The lattice parameters (a and c) of the BST glass-ceramics are calculated based on the XRD results using MDI Jade software. The variation in lattice parameters with cerium content is given in Fig. 2(c). As shown in Fig. 2(b), when the cerium content is increased from 0 to 1 mol%, the (110) diffraction peak shifts toward lower angles, indicating the substitution of larger Ce4+ (r = 0.87 Å) for smaller Ti4+ (r = 0.605 Å) at the B-site of the ABO3 perovskite structure [17]. Thus, both the lattice parameters a and c of C1 are larger than those of C0. Fig. 3(a) shows the schematic illustration of the substitution of Ce4+ for Ti4+ in the BST lattice. Moreover, further addition of cerium (larger than 2 mol%) results in a shift of (110) diffraction peak to higher degrees and a decrease of both a and c values. It demonstrates that cerium is doped into the A-site as donor in the BST. The XRD patterns are in consistency with the results reported by Hennings et al [19] in the Ce-doped BST polycrystalline ceramics, which revealed that higher cerium content and lower (Ba + Sr)/Ti ratio will make cerium easier to occupy the A-site by Ce3+. Fig. 3(b) is the explanatory chart of Ce3+ entering the Ba2+/Sr2+ sites. It is well-known that the ionic radius of Ce3+ (r = 1.34 Å) is smaller than that of Ba2+ (r Ba2 + = 1.61 Å ) and Sr2+ (r Sr 2 + = 1.44 Å ). Thus, the (110) diffraction peak shifts toward larger angles when Ce3+ occupies A-site. When the cerium content is 2 mol% (C2), the Δ2θ associated with the (110) peak is almost equal to that of the undoped sample (C0), indicating that the cerium substitution for both A- and Bsites may exist simultaneously. The inset of Fig. 2(c) illustrates the compositional dependence of the tetragonality factor (c/a). The c/a ratio reduces from 1.0059 for C0 to 1.0022 for C1, and then increases to 1.0038 and 1.0048 for C2 and C3 respectively. The initial decreased c/a is inherently associated with the tolerance factor t. It could be calculated by the following formula, t = (rA+rO)/[1.414 (rB+rO)]

× CeO2 + AO→ Ce×Ti + A×A + 3OO

(2)

× 2CeO2 + 2TiO2 → 2Ce·A + 2Ti×Ti + 6OO + O2 + 2e′

(3)

× 6CeO2 + 6TiO2 → 6Ce·A + 6Ti×Ti + V A′′ + VTi′′′′ + 21OO +

3 O2 2

(4)

Where A represents Ba or Sr, Eq. (2) is the compensation mechanism of Ce4+ at B-site. Eqs. (3) and (4) show the mechanism of Ce3+ substituted ′′ at A-site and Ce·Ba / Ce·Sr compensated by electrons cation vacancies V Ba , ′′ ′′′′ VSr and VTi , respectively. It has not been fully established which mechanism is responsible for and both mechanisms may occur. There are no sharp variations of grain size as a function of cerium content in these glass-ceramics. Sintered densities of cerium doped BST glass-ceramics are 4.143, 4.148, 4.157 and 4.172 g/cm3 for C0, C1, C2 and C3, respectively. It demonstrates that the density slightly increased with the increase of cerium content. 3.2. XPS Fig. 4(a) shows the high-resolution XPS spectra of the O 1s signal and their corresponding fitting curve acquired from C0, C1, C2 and C3 samples. The spectrum of the O 1s signal is resolved using two peaks for their fits: peak 1 is assigned to the presence of adsorbed hydroxyl (−OH) groups on the surface and is valuable in characterizing the oxygen vacancy content [28,29]; the BE of peak 2 is attributed to oxygen in BST lattice, including Ba-O, Ti-O, Sr-O bonding [30]. Furthermore, the ratios between the areas below the peak 1 and 2, A1/A2 [29], calculated for all the spectra are 1.31, 1.34, 1.39 and 1.31 for C0, C1, C2 and C3, respectively. It demonstrated that the maximum value of oxygen vacancy concentration is obtained from the BST glassceramics with 2 mol% cerium content. An increase of A1/A2 values of C1 and C2 compared with C0 intuitively suggests a charge compensation mechanism that may result from an effective acceptor doping. Notably, there was no intentional acceptor dopant present in these glass-ceramics. This is a potential mechanism that the substitution of Ce4+ for Ti4+ makes oxygen octahedron to be expanded as illustrated in Fig. 3(a). Thus, oxygen vacancies appear around oxygen octahedron to alleviate this distortion [31]. As shown in Fig. 3(b), the substitution of Ce3+ for Ba2+/Sr2+ leads to a shrinkage of oxygen octahedron, resulting in the disappearance of local oxygen vacancies around oxygen octahedron. And the A1/A2 value of C3 is smaller than that of C2. It is noted that the BEs of peak 2 are 529.3 eV for C0 and 529.6 eV, 529.7 eV, 529.5 eV for C1, C2 and C3, respectively. A shift toward higher BE from C0 to C2 and then a decrease was observed for C3. Fig. 4(b) shows the high-resolution XPS spectra of the Ti 2p signal acquired from the four samples. The Ti 2p spectrum was separated into two components (Ti4+, Ti3+) by peak fitting. Peak I, III and peak II, IV are attributed to Ti4+ and Ti3+, respectively. Lower valence states than

(1)

Where rA, rB and rO are the radii of the A- and B-site ions and the oxygen ion, respectively. The tolerance factor of (Ba2+,Sr2+)Ce4+O3 (0.88 < t < 0.94) is significantly lower than that of (Ba,Sr)TiO3 (1.00 < t < 1.06). And the later increased value of c/a is due to the higher t value of Ce3+Ti3+O3 (t = 0.94) than that of (Ba2+,Sr2+)Ce4+O3. The compensation mechanisms for cerium substitution can be given by [21]:

3

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx

Z. Zhao, et al.

Fig. 5. Variation of dielectric properties as a function of temperature for the BST glass ceramics with various cerium contents.

%, the value of dielectric constant over the whole temperature range decreases without further shift in TC. The undoped sample has the largest value of tanδ in the temperature range of 0 °C and 175 °C. These results indicate that a moderate cerium-doping level is advantageous to improve the dielectric properties. The increase of dielectric losses at high temperatures has also been observed. It is most probably caused by the relaxation of mobile charge carriers. 3.4. Discharged energy storage density The released energy storage density Jd* could be calculated by the following formula [32]:

Jd* =

tf

∫ U32 (t ) dt ts

(6)

Where U3 (t) is the discharged voltage of R3, t is the discharging time, ts is the start time to discharge, tf is the finish time to discharge and V is the volume of capacitor. The time dependence of the released energy storage densities Jd* at 10 kV/mm of the BST glass-ceramics with various cerium contents is illustrated in Fig. 6. The Jd* values at 0.8 ms for C0, C1, C2 and C3 are 0.107, 0.125, 0.133 and 0.088 J/cm3, respectively. At cerium content of

Fig. 4. (a) High-resolution XPS spectra of the O 1s signal with various cerium contents; (b) high-resolution XPS spectra of the Ti 2p signal with various cerium contents.

3+ were not observed. Moreover, the ratios between the areas below the peak of Ti4+ and Ti3+, AI+III/AII+IV, are 1.95, 1.96, 1.76 and 1.45 for C0, C1, C2 and C3, respectively. The AI+III/AII+IV value of C1 is almost equal to that of C0, but it shows a decrease in C2 and C3, indicating that the amount of Ti3+ gradually increases when cerium content exceeds 2 mol%. Based on the above results, it can be explained that the electrons are generated as shown in Eq. (3) and the following equilibrium develops, i.e.,

Ti4 + + e′ → Ti3 +

R22 R32 R 4 V

(5)

A shift toward higher BEs of the Ti 2p levels is detected among these samples, as indicated by dotted lines in the spectra of Fig. 4(b). This shift trend of BEs is similar to that in O 1s. It indicates that the cerium substitution leads to the change of the surrounding environment of oxygen and titanium in the crystal lattice. 3.3. Dielectric properties The cerium content dependence of the dielectric properties of these BST glass-ceramic samples was also investigated. Fig. 5 is the temperature dependence of dielectric constant (ε) and dielectric loss (tanδ) of the BST glass-ceramics over the temperature range of -100 °C to 300 °C at 1 kHz. The value of ε increases and the Curie temperature (TC) shifts to lower temperatures, from εmax = 120 at TC = 39 °C for C0 to εmax = 151 at TC = 1 °C for C2. When the cerium content is above 2 mol

Fig. 6. Time dependence of the calculated released energy storage densities of the BST glass-ceramics with various cerium contents. 4

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx

Z. Zhao, et al.

Fig. 8. (a) Nyquist plots (Z′ vs. Z′′) for C0, C1, C2 and C3 at 500 °C; (b) Nyquist plots (Z′ vs. Z′′) of experimental data and fitting line at 500 °C for C2. The inset of (b) shows the equivalent circuit of the electrical response model, where the circuit elements R1-C1 and R2-C2 represent the contributions from the crystal and crystal-glass interface, respectively.

Fig. 7. (a) Frequency dependence of Z′ for C0, C1, C2 and C3 at 500 °C; (b) frequency dependence of ac conductivity for C0, C1, C2 and C3 at 500 °C.

2 mol%, the released energy storage density reaches the maximum value. A important feature in Fig. 6 is the siginificant decrease of the released energy storage density for the BST glass-ceramics with 3 mol% cerium content. Similar results and trends have been oberved for the dielectric constant as shown in Fig. 5. This variation could be attributed to the saturation of cerium doping amount as well as accompanying crystallization of CeO2 in these samples. It may also be highly correlated to the transition of substitution site in the BST glass-ceramics.

ac conductivity is the largest. Impedance spectroscopy is used to investigate the electrical properties of different components. Fig. 8(a) shows the Nyquist plots (Z′ vs. Z′′) in a wide frequency range (from 20 Hz to 1 MHz) at 500 °C. It can be seen that the impedance plots display only one strongly suppressed semicircle which can be associated with a strong overlap of two semicircles due to similar relaxation time constant τ of two relaxation processes [33]. And the semicircular arcs become smaller first and then get larger with the increase of cerium content. Fig. 8(b) gives the Nyquist plots (Z′ vs. Z′′) of experimental data and fitting line at 500 °C for C2. The modelling of impedance data (using ZSimpwin, Version 3.10) is usually accomplished by an equivalent circuit, as shown in the inset of Fig. 8(b). R and C denote the resistance and capacitance, respectively. The original data of Fig. 8(b) matched well with the fitting results, indicating that the equivalent circuit is suitable for these samples. According to the impedance analysis, the semicircle on the left is the response at high frequency and ascribed to the crystalline phase, while the other one on the right corresponds to the crystal-glass interface at low frequency [9]. The values of R1, C1, R2 and C2 can be obtained after fitting. The relaxation time τ is defined as:

3.5. Impedance characterization Frequency dependence of the real part of impedance (Z′) at 300–550 °C for all samples is accomplished and the data at 500 °C of these four samples are extracted and plotted to obtain Fig. 7(a). The cerium content dependence at 500 °C is evident, that is, the values of Z′ in the low frequency region decrease when the cerium content increases up to 2 mol% and then start to increase with further increase of cerium content. In order to understand the effect of cerium content on conductivity, ac conductivity σac at 500 °C is plotted as a function of frequency, which is shown in Fig. 7(b). When the cerium content is 2 mol%, the value of 5

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx

Z. Zhao, et al.

Fig. 9. Arrhenius plots of (a) τ1, (b) τ2, and (c) σdc; (d) variation of activation energy with cerium content.

τ1 = R1C1

(7)

τ2 = R2C2

(8)

Fig. 9(c). The activation energy values Eτ1 at high frequency are in the range of 0.74–1.11 eV, while Eτ2 values at low frequency are in the range of 0.70-0.86 eV. The migration of oxygen vacancies with the energy level around 1.0 eV has been widely confirmed to occur in perovskite ceramics and glass-ceramics [9,34,36,37]. Eτ1 could be ascribed to in-crystal oxygen vacancy migration and Eτ2 is typically attributed to oxygen vacancy migration across the crystal-glass interface. The Edc values are in the range of 1.23–1.40 eV, thus the dc conduction mechanism may be long-range diff ;usion of oxygen vacancies [38]. All the above-mentioned defect mechanism, in conjunction with the calculation of activation energy, provide a consistent explanation for the cerium content dependence of impedance behavior observed in the BST glass-ceramics. The highest conductivities occur in C2 at 500 °C as illustrated in Fig. 7(b), indicating that the cerium doping could enhance conductivity and decrease the activation energy (Fig. 9(d)). Obviously, according to Eq. (3), the electron concentration increases with the increase of the substitution of Ce3+ for A-site. But the electron contribution to the conductivity could reduce due to electron scattering in the crystal lattice at 500 °C. While oxygen vacancies are the carriers that dominate conductivity at high temperatures and their concentration may increase with increasing cerium content up to 2 mol%. In addition, the moderate cerium doping level could result in the decrease in the activation energy of oxygen vacancies. Then the conductivities tend to decrease with increasing cerium content up to solid solution limit (3 mol%), while the activation energy increases as shown in Fig. 9(d). This increase could be attributed to the presence of more cation vacancies (shown in Eq. (4)), which is detrimental to the diffusion rate of oxygen vacancy. A similar phenomenon was observed in other perovskite oxides such as SrTiO3 [39] and Na0.5Bi0.5TiO3 [40].

Then, the values of τ1 and τ2 at various temperatures can be measured using Eqs. (7) and (8). The plots of lnτ versus 1/T are linear which obey the Arrhenius equation [34]:

lnτ =

E + lnτ0 kT

(9)

Where E is the activation energy, k is the Boltzmann constant, T is the absolute temperature, τ0 is the pre-exponential factor. The slope of lnτ versus 1/T plot is related to the value of E. The calculated activation energy value can be used to infer the associated relaxation mechanism. Fig. 9(a) and (b) are the variation of relaxation time τ1 (at high frequency) and τ2 (at low frequency) as a function of temperature, respectively. Generally, ac conductivity data can be described by Jonscher’s universal law [35],

σ (ω) = σdc + Aωn

(10)

where σ (ω) is the real part of total conductivity, σdc is dc conductivity, coefficient A and exponent n (0 < n < 1) are dependent on temperature and material properties. σdc obtained from Eq. (10) obeys an Arrhenius law, which is suitable for thermal activated systems,

E σdc = B exp ⎛− dc ⎞ ⎝ kT ⎠

(11)

where B is the pre-exponential factor and Edc is the activation energy of dc conductivity. Arrhenius plots of the dc conductivity are shown in 6

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx

Z. Zhao, et al.

The strong coupling between the cation vacancies and the oxygen vacancies makes the carrier migration more difficult and severely constrains their diffusion rate.

https://doi.org/10.1063/1.371698. [9] Y. Zhang, T. Ma, X. Wang, Z. Yuan, Q. Zhang, Two dielectric relaxation mechanisms observed in lanthanum doped barium strontium titanate glass ceramics, J. Appl. Phys. 109 (2011) 084115, https://doi.org/10.1063/1.3581065. [10] Q. Sun, Q. Gu, K. Zhu, R. Jin, J. Liu, J. Wang, J. Qiu, Crystalline structure, defect chemistry and room temperature colossal permittivity of Nd-doped barium titanate, Sci. Rep. 7 (2017) 42274, https://doi.org/10.1038/srep42274. [11] Khushbu, P. Kumar, V. Kumar, Effect of co-substitution of Sm3+ and Fe3+ ions on structural and dielectric properties of BaTiO3 ceramics, J. Alloys Compd. 731 (2018) 760–765, https://doi.org/10.1016/j.jallcom.2017.10.059. [12] C.B. Samantaray, M.L.N. Goswami, D. Bhattacharya, S.K. Ray, H.N. Acharya, Photoluminescence properties of Eu3+-doped barium strontium titanate (Ba, Sr)TiO3 ceramics, Mater. Lett. 58 (2004) 2299–2301, https://doi.org/10.1016/j. matlet.2004.03.001. [13] A. Garrido-Hernández, A. García-Murillo, Fd.J. Carrillo-Romo, L.A. Cruz-Santiago, G. Chadeyron, Ad.J. Morales-Ramírez, S. Velumani, Structural studies of BaTiO3:Er3+ and BaTiO3:Yb3+ powders synthesized by hydrothermal method, J. Rare Earths 32 (2014) 1016–1021, https://doi.org/10.1016/S1002-0721(14) 60176-9. [14] V.V. Mitic, Z.S. Nikolic, V.B. Pavlovic, V. Paunovic, M. Miljkovic, B. Jordovic, L. Zivkovic, Influence of rare-earth dopants on barium titanate ceramics microstructure and corresponding electrical properties, J. Am. Ceram. Soc. 93 (2010) 132–137, https://doi.org/10.1111/j.1551-2916.2009.03309.x. [15] D.Y. Lu, X.L. Gao, S. Wang, Abnormal Curie-temperature shift in Ho-doped BaTiO3 ceramics with the self-compensation mode, Results Phys. 12 (2019) 585–591, https://doi.org/10.1016/j.rinp.2018.11.094. [16] D.Y. Lu, M. Toda, M. Sugano, High-permittivity double rare-earth-doped barium titanate ceramics with diffuse phase transition, J. Am. Ceram. Soc. 89 (2006) 3112–3123, https://doi.org/10.1111/j.1551-2916.2006.00893.x. [17] D. Makovec, Z. Samardzija, D. KOLAR, Solid solubility of cerium in BaTiO3, J. Solid State Chem. 123 (1996) 30–38, https://doi.org/10.1006/jssc.1996.0148. [18] R.S. Roth, T. Negas, H.S. Parker, D.B. Minor, C. Jones, Crystal chemistry of cerium titanates, tantalates and niobates, Mater. Res. Bull. 12 (1977) 1173–1182, https:// doi.org/10.1016/0025-5408(77)90171-4. [19] D.F.K. Hennings, B. Schreinemacher, H. Schreinemacher, High-permittivity dielectric ceramics with high endurance, J. Eur. Ceram. Soc. 13 (1994) 81–88, https://doi.org/10.1016/0955-2219(94)90062-0. [20] J.H. Hwang, Y.H. Han, Dielectric properties of (Ba1-xCex)TiO3, Jpn. J. Appl. Phys. 39 (2000) 2701–2704, https://doi.org/10.1143/JJAP.39.2701. [21] J.H. Hwang, Y.H. Han, Electrical properties of cerium-doped BaTiO3, J. Am. Ceram. Soc. 84 (2001) 1750–1754, https://doi.org/10.1111/j.1151-2916.2001.tb00910.x. [22] J. Zhi, Y. Zhi, A. Chen, Crystalline structure and dielectric behavior of (Ce,Ba)TiO3 ceramics, J. Mater. Res. 17 (2011) 2787–2793, https://doi.org/10.1557/jmr.2002. 0405. [23] A. Chen, Y. Zhi, J. Zhi, P.M. Vilarinho, J.L. Baptista, Synthesis and characterization of Ba(Ti1-xCex)O3 ceramics, J. Eur. Ceram. Soc. 17 (1997) 1217–1221, https://doi. org/10.1016/S0955-2219(96)00220-8. [24] L.P. Curecheriu, C.E. Ciomaga, V. Musteata, G. Canu, V. Buscaglia, L. Mitoseriu, Diffuse phase transition and high electric field properties of BaCeyTi1−yO3 relaxor ferroelectric ceramics, Ceram. Int. 42 (2016) 11085–11092, https://doi.org/10. 1016/j.ceramint.2016.04.008. [25] S. Liu, Q. Xie, L. Zhang, Y. Zhao, X. Wang, P. Mao, J. Wang, X. Lou, Tunable electrocaloric and energy storage behavior in the Ce, Mn hybrid doped BaTiO3 ceramics, J. Eur. Ceram. Soc. 38 (2018) 4664–4669, https://doi.org/10.1016/j. jeurceramsoc.2018.06.020. [26] K.S. Srikanth, R. Vaish, Enhanced electrocaloric, pyroelectric and energy storage performance of BaCexTi1−xO3 ceramics, J. Eur. Ceram. Soc. 37 (2017) 3927–3933, https://doi.org/10.1016/j.jeurceramsoc.2017.04.058. [27] Q. Zhang, X. Liu, Y. Zhang, X. Song, J. Zhu, I. Baturin, J. Chen, Effect of barium content on dielectric and energy storage properties of (Pb,La,Ba)(Zr,Sn,Ti)O3 ceramics, Ceram. Int. 41 (2015) 3030–3035, https://doi.org/10.1016/j.ceramint. 2014.10.139. [28] W. Hu, Y. Liu, R.L. Withers, R.L. Withers, T.J. Frankcombe, L. Norén, A. Snashall, M. Kitchin, P. Smith, B. Gong, H. Chen, J. Schiemer, F. Brink, J. Wong-Leung, Electron-pinned defect-dipoles for high-performance colossal permittivity materials, Nat. Mater. 12 (2013) 821, https://doi.org/10.1038/nmat3691. [29] E. Ramos-Moore, P. Ferrari, D.E. Diaz-Droguett, D. Lederman, J.T. Evans, Raman and x-ray photoelectron spectroscopy study of ferroelectric switching in Pb (Nb,Zr,Ti)O3 thin films, J. Appl. Phys. 111 (2012) 014108, https://doi.org/10. 1063/1.3675479. [30] M. Wegmann, L. Watson, A. Hendry, XPS Analysis of submicrometer barium titanate powder, J. Am. Ceram. Soc. 87 (2004) 371–377, https://doi.org/10.1111/j. 1551-2916.2004.00371.x. [31] N. Kumar, E.A. Patterson, T. Frömling, E.P. Gorzkowski, P. Eschbach, I. Love, M.P. Müller, R.A. De Souza, J. Tucker, S.R. Reese, D.P. Cann, Defect mechanisms in BaTiO3‐BiMO3 ceramics, J. Am. Ceram. Soc. 101 (2018) 2376–2390, https://doi. org/10.1111/jace.15403. [32] X. Hao, A review on the dielectric materials for high energy-storage application, J. Adv. Dielectr. 03 (2013) 1330001, https://doi.org/10.1142/s2010135x13300016. [33] X. Wang, Y. Zhang, X. Song, Z. Yuan, T. Ma, Q. Zhang, C. Deng, T. Liang, Glass additive in barium titanate ceramics and its influence on electrical breakdown strength in relation with energy storage properties, J. Eur. Ceram. Soc. 32 (2012) 559–567, https://doi.org/10.1016/j.jeurceramsoc.2011.09.024. [34] C. Ang, Z. Yu, L.E. Cross, Oxygen-vacancy-related low-frequency dielectric relaxation and electrical conduction in Bi:SrTiO3, Phys. Rev. B 62 (2000) 228–236, https://doi.org/10.1103/PhysRevB.62.228.

4. Conclusion In summary, the phase evolution, dielectric properties, energy storage properties and defect mechanism of the cerium doped (Ba,Sr)TiO3 glass-ceramics were investigated. The XRD patterns show that when the cerium content is low (≤1 mol%), the c/a ratio decreases and cerium mainly acts as an isovalent dopant in the B-site of ABO3 perovskite structure. When the cerium content is more than 1 mol%, the c/a ratio increases and cerium substitution gradually occurs in the A-site. The XPS spectra demonstrated that the maximum value of oxygen vacancy concentration is obtained from the BST glass-ceramics with 2 mol% cerium content. Both the dielectric constant and released energy storage density reach the maximum values at this cerium content. The impedance spectra illustrate that the glass-ceramics with this cerium content also show the lowest value of real part of impedance and the highest ac conductivity. Such an effect could be explained by the increase in the concentration of oxygen vacancies and electrons. Besides, according to the fitting of Nyquist plots and the calculations of activation energy, it is noted that a moderate cerium doping level reduces the activation energy of oxygen vacancy migration. With a further increase of cerium content up to 3 mol%, the changes of properties show an opposite trend. This could be explained by the presence of more cation vacancies which is detrimental to the diffusion rate of oxygen vacancy. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. The authors declare the following financial interests/personal relationships which may be considered as potential competing interests. Acknowledgments This work was supported by National Natural Science Foundation of China (Grant No. 51672157 and 51811530105) and the Ministry of Science and Technology of China through 973-Project (Grant No. 2015CB654604). References [1] J.A. Gaudet, C.J. Barker, J. Dickens, R.P. Joshi, J.F. Kolb, A. Neuber, E. Schamiloglu, J.S. Tyo, Research issues in developing compact pulsed power for high peak power applications on mobile platforms, Proc. IEEE 92 (2004) 1144–1165, https://doi.org/10.1109/JPROC.2004.829006. [2] H.V. Alexandru, C. Berbecaru, A. Ioachim, M.I. Toacsen, M.G. Banciu, L. Nedelcu, D. Ghetu, Oxides ferroelectric (Ba, Sr)TiO3 for microwave devices, Mater. Sci. Eng. B 109 (2004) 152–159, https://doi.org/10.1016/j.mseb.2003.10.034. [3] A.K. Yadav, C.R. Gautamb, P. Singh, Crystallization and dielectric properties of Fe2O3 doped barium strontium titanate borosilicate glass, RSC Adv. 5 (2015) 2819–2826, https://doi.org/10.1039/c4ra11301b. [4] E.P. Gorzkowski, M.J. Pan, B. Bender, C.C.M. Wu, Glass-ceramics of barium strontium titanate for high energy density capacitors, J. Electroceram. 18 (2007) 269–276, https://doi.org/10.1007/s10832-007-9127-1. [5] Z. Yao, Z. Song, H. Hao, Z. Yu, M. Cao, S. Zhang, M.T. Lanagan, H. Liu, Homogeneous/ Inhomogeneous-structured dielectrics and their energy-storage performances, Adv. Mater. 29 (2017) 1601727, https://doi.org/10.1002/adma. 201601727. [6] B. Su, T.W. Button, Microstructure and dielectric properties of Mg-doped barium strontium titanate ceramics, J. Appl. Phys. 95 (2004) 1382–1385, https://doi.org/ 10.1063/1.1636263. [7] A. Jana, T.K. Kundu, Microstructure and dielectric characteristics of Ni ion doped BaTiO3 nanoparticles, Mater. Lett. 61 (2007) 1544–1548, https://doi.org/10.1016/ j.matlet.2006.07.075. [8] F.D. Morrison, D.C. Sinclair, A.R. West, Electrical and structural characteristics of lanthanum-doped barium titanate ceramics, J. Appl. Phys. 86 (1999) 6355–6366,

7

Journal of the European Ceramic Society xxx (xxxx) xxx–xxx

Z. Zhao, et al. [35] A.K. Jonscher, The ‘universal’ dielectric response, Nature 267 (1977) 673–679, https://doi.org/10.1038/267673a0. [36] S.F. Liu, Y.J. Wu, J. Li, X.M. Chen, Effects of oxygen vacancies on dielectric, electrical, and ferroelectric properties of Ba4Nd2Fe2Nb8O30 ceramics, Appl. Phys. Lett. 104 (2014) 082912, https://doi.org/10.1063/1.4867069. [37] R. Meyer, R. Liedtke, R. Waser, Oxygen vacancy migration and time-dependent leakage current behavior of Ba0.3Sr0.7TiO3 thin films, Appl. Phys. Lett. 86 (2005) 112904, https://doi.org/10.1063/1.1874313. [38] B. Rangarajan, S.S.N. Baharadwaja, E. Furman, T. Shrout, M. Lanagan, Impedance

spectroscopy studies of fresnoites in BaO-TiO2-SiO2-system, J. Am. Ceram. Soc. 93 (2010) 522–530, https://doi.org/10.1111/j.1551-2916.2009.03423.x. [39] A. Walsh, C. Catlow, A. Richard, A.G.H. Smith, A.A. Sokol, S.M. Woodley, Strontium migration assisted by oxygen vacancies in SrTiO3 from classical and quantum mechanical simulations, Phys. Rev. B 83 (2011) 220301, https://doi.org/10.1103/ PhysRevB.83.220301K. [40] H. Zhang, A.H.H. Ramadan, R.A. De Souza, Atomistic simulations of ion migration in sodium bismuth titanate (NBT) materials: towards superior oxide-ion conductors, J. Mater. Chem. A. 00 (2018) 1–8, https://doi.org/10.1039/C8TA02545B.

8