Crystallization kinetics, breakdown strength, and energy-storage properties in niobate-based glass-ceramics

Accepted Manuscript Crystallization kinetics, breakdown strength, and energy-storage properties in niobate-based glass-ceramics Jinhua Liu, Haitao Wang, Bo Shen, Jiwei Zhai, Zhongbin Pan, Ke Yang, Jinran Liu PII:

S0925-8388(17)32053-4

DOI:

10.1016/j.jallcom.2017.06.066

Reference:

JALCOM 42134

To appear in:

Journal of Alloys and Compounds

Received Date: 22 March 2017 Revised Date:

17 May 2017

Accepted Date: 5 June 2017

Please cite this article as: J. Liu, H. Wang, B. Shen, J. Zhai, Z. Pan, K. Yang, J. Liu, Crystallization kinetics, breakdown strength, and energy-storage properties in niobate-based glass-ceramics, Journal of Alloys and Compounds (2017), doi: 10.1016/j.jallcom.2017.06.066. 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.

ACCEPTED MANUSCRIPT

Crystallization kinetics, breakdown strength, and energy-storage properties in niobate-based glass-ceramics JinhuaLiu1,2, HaitaoWang1, Bo Shen1, Jiwei Zhai1*, Zhongbin Pan1 , Ke Yang1, Jinran Liu1 1

Key Laboratory of Advanced Civil Engineering Materials of Ministry of Education, Functional Materials

RI PT

Research Laboratory, School of Materials Science & Engineering, Tongji University, 4800 Caoan Road, Shanghai 201804, China

School of Physics and Electronic Science, Guizhou Normal University, Guiyang 550001, China

SC

2

Abstract ：

M AN U

In this work, (100-x)(0.12 SrO-0.3 Na2O-0.1 Nb2O5)-xSiO2 (x=30, 35, 40, 45, 50 mol%) glass-ceramics were prepared by using the meth-quenching-controlled crystallization method. Phase evolution, Crystallization mechanism, dielectric properties, dielectric breakdown strength (DBS), and energy-storage performances were comprehensively studied by varying SiO2 content. DSC studies revealed

TE D

simultaneous occurrence of surface and internal crystallization mechanism in the glass-ceramics. XRD results showed three tungsten bronze structure SrNb2O6, Sr6Nb10O30, NaSr2Nb5O15 phases and perovskite structure NaNbO3 phase, which was

EP

quantified by the Reietveld refinement. It was found that dielectric constant and theoretical energy-storage density increased firstly and then decreased with the increase of the SiO2 contents. For x=35 mol%, the theoretical energy-storage density

AC C

reaches the maximal value of 15.3 J/cm3 due to the highest dielectric constant of 124 and DBS of 1669 kV/cm. And the highest DBS is related to the uniform and dense microstructure for x=35 mol%. For practical applications in pulsed RLC circuit, the discharged efficiency increases from 74.2% to 91.5% with the increase of the SiO2 contents.

Keywords: Glass-ceramics; Crystallization mechanism; Energy storage density; Discharged energy efficiency * Author to whom correspondence should be addressed. E-mail: [email protected] Tel: 86-21-69584759, Fax: 86-21-69584759.

ACCEPTED MANUSCRIPT 1. Introduction Dielectric capacitors with high energy-storage density are a crucial component to discharge large amounts of electric energy in a shorter time, which results in very high

RI PT

power density for applications of power electronics and pulsed power system [1, 2]. To meet ever-increasing needs for dielectric capacitors with high energy-storage capacity, ferroelectric glass-ceramics were developed and attracted considerable

SC

attentions since this composite material has potential advantages of high DBS due to

M AN U

the glass matrix with strong bond energy of Si-O bond (~ 424 kJ/mol) [3] and high dielectric constant due to ferroelectric phase. Then, ferroelectric glass-ceramics were expected to have high energy-storage density and high power density. A clear understanding of the crystallization kinetic behavior and crystallization

TE D

mechanism is important to manufacture glass-ceramics, tailor dielectric properties, and improve the microstructure of glass-ceramics. Generally, the crystallization kinetic behavior and crystallization mechanism could be revealed by the

EP

crystallization activation energy and the Avrami parameter n [4-9]. For the activation

AC C

energy, its value close to the disassociation energy of ion bonds denotes that crystallization mechanism is attributed to the breakage of ion bonds, while its value needed for the diffusivity of ions indicates that crystallization mechanism is attributed to the diffusion of ions. Avami parameter n is a characteristic parameter that characterizes the crystallization mechanism of the glass-ceramics: surface or/and internal crystallization mechanisms [4, 6]. Recently, numerous studies focused on titanate-based and niboate-based

ACCEPTED MANUSCRIPT ferroelectric glass-ceramics to improve dielectric properties and energy-storage density. In these studies, most researchers mainly changed the compositions of ceramic phases in the glass-ceramics, such as Ba/Sr [10, 11, 12], Ba/Na [13], Ba/K

RI PT

[14], Sr/Na[15], and Ba/Ti [16]. In addition, the rare-earth doping amount was adjusted to improve the microstructure, dielectric properties, and DBS of the ferroelectric glass-ceramics, such as CeO2 [17], Gd2O3 [18], AlF3 [19], BaF2 [20],

SC

MnO2 [21], and so on. No researcher studies effect of the glass former SiO2 content

M AN U

on crystallization kinetics, dielectric properties, and energy-storage performances in the SrO-Na2O-Nb2O5-SiO2 (SNN-Si) glass-ceramics.

In this work, the crystallization kinetics, phase structure, dielectric properties, microstructure, and energy-storage performances of SNN-Si glass-ceramics will be

TE D

studied with the variation in the glass former (SiO2). 2. Experimental procedure

The SNN-Si glass-ceramics were prepared from reagent grade SrCO3 (99.9%,

EP

American Alfa) Na2CO3(99.9%, American Alfa) , Nb2O5(99.99%, Zhuzhou, China)

AC C

and SiO2 (99.99% Chinese Medicine) powders. The nominal composition of the SNN-Si was (100-x)(0.48 SrO-0.12 Na2O-0.4 Nb2O5)-xSiO2 (mol%) with x=30 mol%, 35 mol%, 40 mol%, 45 mol%, and 50 mol%. Correspondingly, the five compositions are referred to as Si30, Si35, Si40, Si45, and Si50, respectively. The weighted powders were well mixed by ball milling in the ethyl alcohol with zirconia for 20 h. The dried powers were melted at 1520oC in an alumina crucible for 2 h. After the copper mold was heated to 600oC for 10~20 min, the glass lava was quickly poured

ACCEPTED MANUSCRIPT into it. Then the copper mold with the glass lava was quickly put into an annealing oven at 600oC and the annealing time is 6 h to remove the residual stress. The transparent glass was cut into the piece-shaped sheets with about 1.2 mm thickness.

RI PT

Thereafter, all glass samples were crystallized at 850oC for 3 h in air with the heating rate of 2oC/min. The chosen crystallization temperature 850oC to obtain simultaneously high dielectric constant and high breakdown strength according to

SC

DSC curve in Fig. 1.

M AN U

The crystallization behavior of the glass powders was analyzed by using differential scanning calorimetry (DSC, model STA449C, Netzsch). The phase structure of the glass-ceramics was detected by X-ray diffraction (XRD, D/max 2550V B3+/PC, Rigaku, Japan) with Cu Kα radiation in the 2θ range from 20oC to

TE D

80oC. Quantitative phase analysis was carried out by using Rietveld analysis program GSAS. Crystallinity of samples was calculated using peak separation technique in Jade software. A LCR meter (Agilent E4980A, Palo Alto CA USA) was used to

EP

measure dielectric constant and dielectric loss of the glass-ceramics at the frequency

AC C

of 100 kHz and in the temperature range of -100oC to 100oC. The microstructure was observed by a field scanning electron microcopy (SEM) (HITACHI S-4700, Japan). A voltage-withstand testing device (ET2671B, Entai, Nanjing, China) was used to measure DC breakdown strength of the glass-ceramics at room temperature. A DC voltage ramp of about 0.2 kV/s was applied to the specimens until the DBS occurred. The glass-ceramic samples were ground into about 0.06~0.08 mm thickness and immersed in silicone oil to prevent surface flashover.

ACCEPTED MANUSCRIPT 3. Results and discussions 3.1 Thermal Properties and Crystallization Mechanism To analyze crystallization process depending on the glass former SiO2 content, the

RI PT

differential scanning calorimetry (DSC) was employed. Fig. 1 shows the DSC patterns of the glass with different SiO2 content at the heating rate of 15oC/min. From the DSC patterns, glass transition temperature (Tg) of 692oC ~ 697oC, dilatometric

SC

softening temperatures (Td), crystallization peak temperature (Tp1, Tp2, Tp3), and

M AN U

crystallization onset temperature (Tx) of 790oC~808oC were observed. It should be noted that crystallization peak temperatures (Tp3) were only found in DSC patterns for high SiO2 contents x=45 mol% and 50 mol%. All data are presented in Table 1. Here, the glass stability could be measured by the stabilization temperature

TE D

(∆T=Tx-Tg), as shown in Table 1. The result indicates the glass for x=35~50 mol% has the similar stability (∆T=108~112).

As is known, the evolution of crystallization exotherms depends on the heating rate

EP

Φ of DSC [6, 9]. In fact, the crystallization temperature (Tp) shifts to higher

AC C

temperature with the heating rate increasing. To analyze the crystallization kinetics, the Kissinger equation [22] is given by ln  = − Φ

 



+C

(1),

where Φ denotes the heating rate, Tp is crystallization peak temperature, Ea is the activation energy of crystallization, R is the gas constant (R=8.314 J·mol-1·K-1), and C is constant. According to Eq. (1), the activation energy Ea of crystallization could be calculated from the slope (-Ea/R) of a fitting straight line ln(Φ/Tp2) versus 1/Tp. Fig. 2

ACCEPTED MANUSCRIPT presents the fitting straight lines with the heating rates Φ = 10oC/min, 15oC/min, 20oC/min, and 25oC/min for (a) Tp1, (b) Tp2, and (c) Tp3, respectively. The calculated activation energies of crystallization are summarized in Table 2. During the

RI PT

crystallization process of the glass-ceramics, the activation energy of crystallization is connected to the breakage of bonds and the diffusion of ions. Since the activation energies of crystallization (446 kJ/mold of Si30 for Tp1, 393 kJ/mol of Si35 for Tp1,

SC

444 kJ/mol of Si45 for Tp2, 395 kJ/mol of Si45 for Tp3) are close to the dissociation

M AN U

energy of Si-O bonds (424 kJ/mol) [3], the crystallization mechanism is attributed to the breakage of Si-O bonds in the glass. The possibility of the diffusion of the ions in the glass, such as Si4+ (555 kJ/mol) [8], Sr2+ (165~255 kJ/mol) [23], is rule out because their activation energies are not consistent with or not close to the diffusion

TE D

energy of the ions. Furthermore, to determine the crystal growth mechanism of the glass-ceramics, i.e., surface and/or internal crystallization, the Avrami parameter n can be estimated by the following equation [5]:

 

(2),

EP

n = 2.5   (∆)  



AC C

where (∆T)p is the full width at half maximum of the crystallization peak. The n value can be used to determine the crystallization mechanism including (i) surface crystallization for n ~ 1, (ii) simultaneous occurrence of both surface and internal crystallization for 1 < n < 3, and (iii) internal crystallization for n ≥ 3. In Table 2, it is observed that the average values of n for different heating rates Φ are in the range from 1 to 3, which implies a combining effect of surface and internal crystallizations. As is known, a broad crystallization peak is associated with surface crystallization,

ACCEPTED MANUSCRIPT whereas a sharp peak is associated with internal crystallization [22]. For example, the n value approaching to 1, such as n = 1.13 at Tp2 for Si35, denotes that the surface crystallization mechanism is dominant due to the broader crystallization than others,

RI PT

while the n value approaching to 3, such as n = 2.52 at Tp2 for x=45 mol%, indicates that the internal crystallization mechanism dominates due to the sharp crystallization. 3.2 Phase structures

SC

In the glass-ceramics, phase segregation must occur in crystallization process. The

M AN U

occurrence of phase segregation is mainly attributed to the competition between Si4+ and other cations (Sr2+, Nb5+, Na+) to capture the oxygen ion (O2+). Thus the crystal phases were formed. To study effect of the glass former SiO2 content on phase structures of the SNN-Si glass-ceramics, XRD patterns for Si30-Si50 samples are

TE D

illustrated in Fig. 3 (a). It is shown that the tungsten structure SrNb2O6, Sr6Nb10O30 phases and cubic perovskite Na0.5Sr0.25NbO3, NaNbO3 were formed. Also, the crystallinity of the glass-ceramics firstly increased from 75% for Si30 to 87% for Si35,

EP

and then decreased to 80% for Si50, which indicates the SiO2 contents in

AC C

glass-ceramics firstly decrease and then increase. Specially for x>35 mol%, phase segregation becomes weaker and weaker due to the reduction of met cat ion contents with the glass former SiO2 contents increasing, which leads to the decrease of the crystallinity. According to the crystal phases from XRD patterns, Rietveld refinements were performed by utilizing the pseudo-Voigt function, the contents of crystal phases are determined with the variation of glass former SiO2 contents, as shown in Table 3. For example, the Reietveld fitting for XRD of glass-ceramic for Si35 with x=35 mol %

ACCEPTED MANUSCRIPT was shown in Fig. 3 (b). The result displays a fine fit between the measured and calculated data. In addition, Table 1 shows percentage of each phase wt% and residual error Rp through Rietveld fit.

RI PT

3.3 Dielectric properties Fig. 4(a) illustrates the temperature dependence of dielectric constant and dielectric loss with the SiO2 contents at the frequency of 100 kHz. With the glass former SiO2

SC

content increasing, the dielectric constant slightly changes, which indicates its

M AN U

excellent stability in the temperature range from -70oC to 100oC. At room temperature, dielectric constant and dielectric loss with the glass former SiO2 contents are shown in Fig. 4 (b). It is shown that dielectric constant firstly increases from 75±4 for x=30 mol% to 124±5 for x=35 mol%, and then decreases down to 31±4 for x=50 mol%. This is

TE D

related to the change of crystallinity in inset of Fig. 3(a). In Fig. 3(a), the crystallinity behavior with the glass former SiO2 contents indicates that the amount of total phases (wt%) in a sample increases firstly and decreases. As shown in Fig. 4(b), the dielectric

EP

loss varies between 0.0135 and 0.0267 at room temperature with the increase of the

AC C

SiO2 content. Dielectric loss increases firstly, and then decreases with the increase of SiO2 content, which is also related to the crystallinity. The change behavior of crystallinity implies that the weight of amorphous phase SiO2 with lower dielectric loss in the SNN-Si glass-ceramics decreases firstly and then increases, which results in the change behavior of dielectric loss. 3.4 Breakdown strength and Microstructure The DBS could be analyzed by utilizing a two-parameter Weibull distribution

ACCEPTED MANUSCRIPT function (E ) = 1 − exp!−(E /E# )$ % [24], where P(Ei) is the cumulative probability of electric failure, Ei is DBS of the i-th specimen in the experiments, Eb is the characteristic DBS, and β is the shape parameter related to linear-regression fit of

RI PT

the data distribution. Based on this function, the Weibull distribution plots of DBS of the glass-ceramics are shown in Fig. 5(a). Here, the shape parameters (β) corresponding to the slope of the fitting lines are bigger than 11 in Fig. 5(b), which

SC

indicates a good linearity. Further, it implies that the DBS values obey the Weibull

M AN U

distribution. In Fig. 5(b), it is observed that the maximum DBS of 1669 kV/cm is achieved for x=35 mol% in SNN-Si glass-ceramics, which is attributed to more uniform and dense microstructure as shown in Fig 6 (b), compared to other cases for x=30, 40, 45, and 50 mol%. In Fig. 6, there are more pores in microstructures for (a)

TE D

x=30 mol%, (c) x=40 mol%, (d) x=45 mol%, for x=50 mol% than that for x=35 mol%. In (a), (d), and (e), the cluster of crystal particles occur. Pores and the cluster can lead to the non-uniform distribution of crystal particles in microstructures.

EP

3.5 Energy-storage density

AC C

The theoretical energy-storage density could be calculated by using the formula Jd=0.5ε0εrEb2, where ε0 is the vacuum dielectric constant, εr relatively dielectric constant, and Eb is dielectric breakdown strength, respectively. The theoretical energy-storage densities of SNN-Si glass-ceramics are 8.1 J/cm3, 15.3 J/cm3, 6.6 J/cm3, 2.6 J/cm3, and 3.6 J/cm3 for Si30, Si35, Si40, Si45, and Si50, respectively. It is obvious that the optimal energy-storage density reaches 15.3 J/cm3 for Si35 due to the highest DBS of 1669 kV/cm and dielectric constant of 124.

ACCEPTED MANUSCRIPT 3.6 Discharged energy efficiency For high power capacitors in pulsed power application, discharged properties, such as discharged speed and discharged energy efficiency, can be obtained by a

RI PT

high-speed RLC discharged circuit [25]. In this circuit, the time dependence of the discharged current (I~t) is measured by monitoring the current (I(t)). Based on the measured current I(t), discharged energy density Wd through the pulsed RLC circuit is

( )(*) + -* .

(3),

M AN U

&' =

SC

described by

where R (≈304 Ω) is a load resistor in the discharged circuit and V is the volume of the measured glass-ceramic sample. The charged energy density (Wc) can be calculated by the equation 0.5CU2/V, where C is the capacitance of the measured

TE D

glass-ceramic capacitor, U is the applied voltage, and V is the volume of the sample. Then, the discharged energy density is given by η=(Wd/Wc)×100%. Fig. 7 displays the discharged energy density as a function of time (Wd-t) for different SiO2 content

EP

and under electric field of 50 kV/cm. According to the definition of the discharged

AC C

time [26], one can obtain the discharged time between 10 ns and 70 ns, as shown in Fig. 7. In this work, the capacitance (C), the charged energy density (Wc), discharged energy density, and discharged energy density (η) are summarized in Table 4. The result indicates that discharged energy efficiency increases from 74.2% to 91.5% with the increase of the glass former SiO2 content, which is attributed to the interfacial polarization [27, 28]. Due to the great difference in dielectric properties of the ferroelectric phase and the glass matrix, the amounts of space charges accumulate at

ACCEPTED MANUSCRIPT interfaces between them. Therefore, interfacial polarization seriously affects the discharged energy efficiency. 4. Conclusions kinetics,

dielectric

properties,

DBS,

and

energy-storage

RI PT

Crystallization

performances were investigated in SNN-Si glass-ceramics with the variation in the glass former SiO2 contents. According to the crystallization kinetic studies, the

SC

surface and internal crystallization simultaneously occurred, which was revealed by

M AN U

the Avrami parameter n between 1 and 3. With the increase of the glass former SiO2 contents, the dielectric constant firstly increased and then decreased, which was attributed to the change of the crystallinity and crystal phase Sr6Nb10O30. The optimal dielectric constant of 124 and DBS of 1669 kV/cm were obtained for the SiO2

TE D

contents x=35 mol%, which results in the highest theoretical energy-storage density of 15.2 J/cm3. In pulsed RLC circuit, the fast charged-discharged time is from 10 ns to 70 ns for different SiO2 contents. Also, the discharged energy efficiency increased

EP

from 74.2% to 91.5% with SiO2 contents increasing. This study indicates that the

AC C

SNN-Si glass-ceramics with high energy-storage density have potential application for high power capacitor. Acknowledgments

This work was supported by the Ministry of Sciences and Technology of China

through National Basic Research Program of China (973 Program 2015CB654601), the China Postdoctoral Science Foundation (2015M581660) and (2017M611615), and the Doctor Fund of Guizhou Normal University.

ACCEPTED MANUSCRIPT Reference [1] C. Liu, F. Li, L. P. Ma, and H. M. Cheng, Advanced materials for energy storage, Adv. Mater. 22(8) (2010) E28-62.

RI PT

[2] J. H. Liu, H. T Wang, B. Shen, J. W. Zhai, P. Li, Z. B. Pan, Significantly enhanced energy-storage density in the strontium barium niobate-based/titanate -based glass-ceramics, J. Am. Ceram. Soc. 100 (2) (2017) 506-510.

SC

[3] W. D. Kingery, H. K. Bowen, and D. R. Uhlmann, Introduction to Ceramics,

M AN U

Chapter 8, 2nd edition. Wiley, New York City, New York, 1980.

[4] C. S. Ray, Q. Yang, W. H. Huang, and D. E. Day, Surface and internal crystallization in glasses as determined by differential thermal analysis, J. Am. Ceram. Soc. 79(12) (1996) 3155-3160.

TE D

[5] E. P. Gorzkowski, M. J. Pan, B. A. Bender, and C. C. M. Wu, Effect of additives on the crystallization kinetics of Barium Strontium titanate glass-ceramics, J. Am. Ceram. Soc. 91(4) (2008) 1065-1069.

EP

[6] M. J. Wang, Y. Zhang, X. L. Liu, and X. R. Wang, Crystallization mechanism and

AC C

ac conductivity studies on strontium barium niobate glass-ceramics, Ceram. Inter. 39(2) (2013) 2069-2076.

[7] J. Zhu, Y. Zhang, X. Z. Song, Q. Zhang, D. L. Yang, and Y. Z. Chen, Dependence of crystallization behavior on particle size in barium strontium titanate glass-ceramics, J. Am. Ceram. Soc. 99(8) (2015) 2340-2343. [8] C. T. Cheng, M. L. Lanagan J. T. Lin, B. Jones, and M. J. Pan, Crystallization kinetics and dielectric properties of nanocrystalline lead strontium barium

ACCEPTED MANUSCRIPT niobates, J. Mater. Res. 20(2) (2005) 438-446. [9] A. R. Molla, A. Tarafder, and B. Karmakar, Fabrication and properties of Nd3+-doped ferroelectric barium bismuth titanate glass-ceramic nanocomposites, J.

RI PT

Alloy. Compd. 680 (2016) 237-246. [10] J. Song, G. H. Chen, C. L. Yuan, and Y. Yang, Effect of the Sr/Ba ratio on the microstructures

and

dielectric

properties

SrO-BaO-Nb2O5-

SC

B2O3 glass–ceramics, Mater. Lett. 117(7) (2014) 7-9.

of

M AN U

[11] T. Y. Liu, G. H. Chen, J. Song, and C. L. Yuan, Crystallization kinetics and dielectric characterization of CeO2-added BaO–SrO–Nb2O5–B2O3–SiO2 glassceramics, Ceram. Inter. 39(5) (2013) 5553-5559.

[12] H. T. Wang, J. H. Liu J. H. Zhai, Z. B. Pan, B. Pan, Effects of Sr substitution for

TE D

Ba on dielectric and energy-storage density of SrO-BaO-K2O-Nb2O5-SiO2 glass-ceramics, J. Euro. Ceram. Soc.DOI: 10.1016/j.jeurceramsoc. 2017. 04. 049. [13] S. X. Xue, J. W. Wang, S. H. Liu, W. Q. Zhang, L. J. Tang, B. Shen, and J. W.

Na2O)–Nb2O5–SiO2

glass-ceramics,

Ceram.

Inter.

40(5)

(2014)

AC C

(BaO,

EP

Zhai, Effect of the Ba/Na ratio on the microstructure and dielectric properties of

7495–7499.

[14] H. T. Wang, J. H. Liu, J. W. Zhai, B. Shen, Z. B. Pan, K. Yang, and J. R. Liu, Effect of K2O content on breakdown strength and energy-storage density in K2O-BaO-Nb2O5-SiO2 glass-ceramics, Ceram. Inter. 43(5) (2017) 4183-4187. [15] H. T. Wang, J. H. Liu, J. W. Zhai, B. Shen, S. M. Xiu, S. Xiao, and Z. B. Pan, Enhanced energy storage density and discharge efficiency in the strontium

ACCEPTED MANUSCRIPT sodium niobate-based glass-ceramics, J Alloy. Compd. 687 (2016) 280-285. [16] J. Chen, Y. Zhang, C. Deng, X. Dai, and L. Li, Effect of the Ba/Ti ratio on the microstructures and dielectric properties of barium titanate-based glass-ceramics,

RI PT

J. Am. Ceram. Soc. 92(6) (2009) 1350-1353. [17] T. Y. Liu, G. H. Chen, J. Song, and C. L. Yuan, Crystallization kinetics and dielectric

characterization

of

CeO2-added

BaO–SrO–Nb2O5–B2O3–SiO2

SC

glass-ceramics, Ceram. Inter. 39(5) (2013) 5553-5559.

characterization

of

M AN U

[18] Y. Zhou, Q. M. Zhang, J. Luo, Q. Tang, and J. Du, Structural and dielectric Gd2O3-added

BaO–Na2O–Nb2O5–SiO2

glass–ceramic

composites, Scripta Mater. 65(65) (2011) 296-299.

[19] X. Wang, Y. Zhang, T. Ma, Z. Yuan, Q. Zhang, C. Deng, and T. Liang, Influence

TE D

of AlF3 concentration on microstructures and energy storage properties of Barium Strontium titanate glass ceramics, Int. J. Appl. Ceram. Technol. 10(2) (2013) 301-306.

on

crystallization

kinetics

and

electrical

properties

of

AC C

addition

EP

[20] G. H. Chen, T. Y. Liu, C. L. Yuan, Y. Yang, and C. R. Zhou, Effect of BaF2

BaO-SrO-Nb2O5-B2O3-SiO2 glass-ceramics, J. Non-Cryst. Solids 378 (2013)

241-246.

[21] X. R. Wang, Y. Zhang, T. Ma, C. S. Deng, and X. M. Dai, Effects of MnO2 concentration on dielectric properties of barium strontium titanate glass ceramics, Ceram. Inter. 38(1) (2012) S57-S60. [22] J. Massera, S. Fagerlund, L. Hupa, and M. Hupa, Crystallization mechanism of

ACCEPTED MANUSCRIPT the bioactive glasses, 45S5 and S53P4, J. Am. Ceam. Soc. 95(2) (2012) 607-613. [23] H. H. Jean and Y. C. Fang, Devitrification kinetics and mechanism of K2O-CaO-SrO-BaO-B2O3-SiO2 glass-ceramic, J. Am. Ceram. Soc. 84(6) (2001)

RI PT

1354-1360. [24] Z. B. Pan, L. M. Yao, J. W. Zhai, B. Shen, et al., Excellent energy density of polymer nanofibers induced by moderate interfacial area, J. Mater. Chem. A

SC

4(34) (2016) 13259-13263.

M AN U

[25] X. F. Chen, H. L. Zhang, F. Cao, G. S. Wang, X. L. Dong, Y. Gu, H. L. He, and Y. S. Liu, Charge-discharge properties of lead zirconate stannate titanate ceramics, J. Appl. Phys. 106(3) (2009) 034105-034108.

[26] Z. B. Pan, L. M. Yao, J. W. Zhai, D. Z Fu, B. Shen, and H. T. Wang,

TE D

High-energy-density polymer nanocomposites composed of newly structured one-dimensional BaTiO3@Al2O3 nanofibers, ACS Appl. Mater. Interfaces, 9(4) (2017) 4024-4033.

EP

[27] H. T. Wang, J. H. Liu, J. W. Zhai, and B. Shen, Ultra high energy-storage density

AC C

in the barium potassium niobate-based glass-ceramics for energy-storage applications, J. Am. Ceram. Soc. 99(9) (2016) 2909-2912.

[28] S. X. Xue, S. H. Liu, W. Q. Zhang, B. Shen, and J. W. Zhai, Correlation of energy conversion efficiency and interface polarization in niobate glass-ceramic for energy-storage applications, App. Phys. Lett. 106(16) (2015) 162903-162906. Figure captions: Fig.1 DSC curves of SNN-Si glass-ceramics at the heating rate of 15oC/min for

ACCEPTED MANUSCRIPT different glass former SiO2 contents. Fig. 2 Plots of ln(Φ/Tp2) as a function of 1000/Tp for (a) Tp1, (b) Tp2, and (c) Tp3 with different SiO2 contents.

RI PT

Fig. 3(a) XRD patterns of the SNN-Si glass-ceramics with different SiO2 contents at the crystallization temperature of 850oC. In inset, the crystallinities of the

SC

glass-ceramics are plotted. (b) The refinement curve of the glass-ceramic for x=35 mol%.

M AN U

Fig. 4 (a) Dielectric constant and dielectric loss as function of the temperature for the different SiO2 contents. (b) The composition dependence of dielectric constant and dielectric loss at room temperature and frequency of 100 kHz.

TE D

Fig. 5 (a) Weibull plots of DBS of the SNN-Si glass-ceramics with different former SiO2 contents at the crystallization temperature of 850oC. (b) The values of DBS and

EP

the shape parameters β as a function of the SiO2 contents x. Fig. 6 SEM micrographs of the SNN-Si glass-ceramics at the crystallization

AC C

temperature of 850oC for different SiO2 contents. Fig. 7 (a) Discharge energy density of the SNN-Si glass-ceramic capacitors with different SiO2 contents under the applied electric field of E=50 kV/cm and the load resistor R=304Ω. The discharged time τ0.9 was extracted between 10 ns and 70 ns from the discharged energy density versus time.

ACCEPTED MANUSCRIPT Table captions Table 1 Thermal properties of SNN-Si glass with different glass former SiO2 contents. Table 2 Activation energy of crystallization and average values of Avrami parameters

RI PT

from crystallization exothermal peaks Tp1, Tp2, and Tp3 for different glass former SiO2 contents.

Table 3 Refinement results for XRD data of the SNN-Si glass-ceramics. wt % and Rp

SC

denote the weight percentage of crystallization phases and the refined residual error,

M AN U

respectively.

Table 4 Capacitances, charged energy density, discharged energy density, and discharged efficiency for different glass former SiO2 contents under the applied

AC C

EP

TE D

electric field of 50 kV/cm.

ACCEPTED MANUSCRIPT Tables Table 1 Thermal properties of SNN-Si glass with different glass content. Sample Tg name (±2oC)

Td (±2oC)

Tx (±2oC)

Tp1 (±2oC)

Tp2 (±2oC)

Si30

692

718

790

823.9

906.5

Si35

695

720

807

843.7

910.8

Si40

697

721

808

870.8

924.7

Si45

689

714

798

824.1

Si50

688

711

796

824.6

Tp3 (±2oC)

Tx-Tg (±4oC) 98

---

112

---

111

870.7

908.3

109

875.3

907.1

108

M AN U

SC

RI PT

---

Table 2 Activation energy of crystallization and average values of Avrami parameters

Sample name

Tp1

TE D

from crystallization exothermal peaks Tp1, Tp2, and Tp3 for different glass contents. Tp2

Tp3 Ea(kJ/mol) Average n -----

Si35

393

AC C

EP

Si30

Ea(kJ/mol) Average Ea(kJ/mol) Average n n 446.4 1.68 472.2 1.35 1.65

276

1.13

---

---

Si40

311

1.23

504

1.59

---

---

Si45

340

1.75

444

1.62

395

2.04

Si50

330

1.66

344

2.52

682

1.44

ACCEPTED MANUSCRIPT Table 3 Refinement results for XRD data of the SNN-Si glass-ceramics. wt % and Rp denote the weight percentage of crystallization phases and the refined residual error, respectively. Sr6Nb10O30

(wt%)

(wt%)

(wt%)

Si30

22.50

0.85

52.31

Si35

11.09

8.20

74.11

Si40

10.33

8.02

22.95

Si45

7.86

5.68

13.20

Si50

18.83

SrNb2O6

9.97

Rp

(wt%)

(%)

24.34

22

6.60

20

SC

NaNbO3

RI PT

Na0.5Sr0.25NbO3

58.70

17

73.26

17

55.8

24

M AN U

Samples

15.43

TE D

Table 4 Capacitances, charged energy density, discharged energy density, and discharged efficiency for different glass content under the applied electric field of 50

EP

kV/cm.

Capacitance (pF)

Wc (10 J/cm3)

Wd (10-3 J/cm3)

Efficiency (%)

Si30

97

6.79

5.04

74.2

Si35

183

11.17

8.68

77.7

Si40

72

6.73

5.43

80.7

Si45

26

2.38

2.1

83.8

Si50

15

1.94

1.78

91.5

AC C

Sample

-3

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT (1) The surface and inter crystallization mechanisms simultaneously occurred. (2) The microstructure and breakdown strength were improved.

AC C

EP

TE D

M AN U

SC

RI PT

(3) The highest theoretical energy-storage density reached 15.3 J/cm3. (4) The energy efficiency increased from 74.2% to 91.5%.