Study on the dielectric properties of Mg-doped NaBiTi6O14 ceramics
Chen Yong1, Xue Simin1, Luo Qian1, Su Huyin1, Chen Qi1, Huang Zhen1, Xu Linfang1, Cao Wanqiang1, 2, †, Huang Zhaoxiang3, ‡
School of Physics and Electronic Science, Key Laboratory of Ferro & Piezoelectric Materials and Devices of Hubei Province, Hubei Collaborative Innovation Center for Advanced Organic Chemical Materials, Hubei University, Wuhan 430062, China
School of Materials Science and Engineering, Hubei University, Wuhan 430062, China
Department of Information Science and Technology, WenHua College, Wuhan 430074, China

 

† Corresponding author. E-mail: caowanq@163.com huangzhaox@163.com

Abstract

With the interest in using lead-free materials to replace lead-containing materials increasing, the use of Na Bi TiO (NBT) has come into our sight. We studied the composition of NBT and found that NaBiTi O ceramics can be compositionally tuned by Mg-doping on the Ti-site to optimize the dielectric properties. In this study, Mg-doped NaBiTi O (NaBi(Ti Mg O ) ceramics were prepared by a conventional mixed oxide route at different sintering temperatures, and their dielectric properties have been studied at a wide temperature range. X-ray diffraction (XRD) patterns of the NBT-based ceramics indicate that all samples have a pure phase without any secondary impurity phase. The experimental data show that after Mg-doping, the relative permittivity and dielectric loss become lower at 1040, 1060, and 1080 °C except 1020 °C and at different frequencies from 10 kHz, 100 kHz to 1 MHz. Take 1060 °C for example, when the sintering temperature is 1060 °C at 1 MHz, the minimum relative permittivity of NaBiTi O is 32.9 and the minimum dielectric loss is 0.01417, the relative permittivity of NaBi(Ti Mg O under the same condition is 25.8 and the dielectric loss is 0.000104. We explored the mechanism of Mg-doping and surprisingly found that the dielectric property of NaBi(Ti Mg O becomes better owing to Mg-doping. Thus, NaBi(Ti Mg O can be used in microwave ceramics and applied to new energy materials.

1. Introduction

Ferroelectric materials are particularly significant in the development of modern technology due to their promising dielectric and piezoelectric properties. Piezoelectric materials have been widely used in electronic devices such as sensors, actuators, and transducers.[13] In addition, ferroelectric materials are also widely used as dielectric materials in modern electronics applied to bypassing, coupling, filtering, smoothing, and power conditioning.[4] We must point out that we mainly use Pb(Zr Ti )O -based piezoelectric materials nowadays. As the use of lead and other toxic elements has been restricted, the interest in substituting lead-containing materials for lead-free materials has surged recently in order to meet environmental standards such as Restriction of Hazardous Substances (RoHS) and Waste Electrical and Electronic Equipment Directive (WEEE).[58] Compared with BaTiO , PZT and other materials in the perovskite system, NBT has excellent dielectric and piezoelectric properties with higher relative permittivity ( ) and lower dielectric loss. NBT exhibits the maximum relative permittivity at ( and possesses a distorted perovskite structure with extensive chemical, cation-displacement, and octahedral tilt disorder.[911] We must emphasize that NBT has a pretty wide operational temperature, so it can be used in many devices and adapt to various temperature conditions. However, NBT can very easily cause Bi leakage. A new Mg-doped NaBiTi O (NaBi(Ti Mg O ceramic was prepared through the chemical reaction of Na CO , Bi O , MgO, and TiO . We studied the composition of NaBiTi O and surprisingly found that NaBiTi O . can be compositionally tuned by Mg-doping on the Ti-site to optimize its piezoelectric and dielectric properties. The samples were prepared by the following chemical equation:

This NaBiTi O material with Mg-doping has high relative permittivity and low dielectric loss with a wide operational temperature range. NaBiTi O can be comprehensively used in ceramic capacitors, especially high-temperature capacitors and applied to electronic control systems for hybrid or electrically powered vehicles and microwave ceramics. NaBiTi O also develops a new direction for ceramic capacitors and new energy materials.[1215]

2. Experimental procedure

Reagent grade Na CO , Bi O , MgO, and TiO were used as the starting materials of NaBi(Ti Mg O (99.9%Na CO , 99.9%Bi O , 99.9%TiO , 99.9%MgO). These high-purity powders were weighed stoichiometrically and ball-milled with C H OH for 24 h, then the powders washed and dried in a drying oven were calcined at 950 °C for 2 h. The phase identification was done using a powder x-ray diffractometer (XRD; Bruker D8, Karlsruhe, Germany, Cu Ka A) with Cu Ka radiation; scan speed of 1 °C/min. After being re-milled, the powder was pressed into disks of 10 mm in diameter and 2 mm in thickness. The PVA (polyvinyl alcohol) was heated for 10 h at 650 °C in the air. The compacted disks were sintered at 1020 °C–1100 °C for 9 h in stove (Energy-save-type Quick temp raising Electric furnace), then cooled down to room temperature. The sintered disks were polished and examined by x-ray diffraction on a Philips X'Pert diffractometer again. Finally, the dishes were covered with silver paste and the electrode of the dishes was baked. After the preparation of NaBiTi O high-temperature IS measurements were performed using an impedance analyzer (Agilent 4194 A, Agilent) in a non-inductively wound tube furnace. The relative permittivity and dielectric loss were tested by an impedance analyzer (HP Agilent, 4192 A Hewlett Packed, Santa Clara, CA) between 100 °C and 700 °C. The microstructure of NaBiTi O was observed by scanning electron microscope (SEM JEOL Ltd., Tokyo, Japan).

3. Results and discussion
3.1. XRD results

The XRD patterns of sintered pellets of NaBiTi O and NaBi(Ti Mg O are displayed in Fig. 1. The x-ray diffraction patterns of the NBT-based ceramics indicate that all samples have a pure phase. Figures 1(a) and 1(b) show the XRD patterns of well-sintered NaBiTi O and NaBi(Ti Mg O ceramic disks respectively. The XRD results show a major phase of NaBiTi O solid solution. It can be seen that all samples maintain the tetragonal perovskite structure. According to the XRD spectrum analysis, the peak shifted to the right after Mg-doping, which indicated that interplanar spacing became larger which was mainly attributed to the substitution of relatively larger Mg into Ti sites. Based on Bragg’s Law, , under the condition that the wavelength of incident light remains unchanged. What has been mentioned above elaborates the reason why the peak appears later in the XRD spectrum. This could be understood by the oxygen vacancy compensation mechanism:

When the samples were sintered at a temperature from 1020 °C to 1080 °C, the substitution of Mg into Ti sites which follows oxygen vacancy compensation mechanism explains the formation of TiO . It has been reported that substituting Ti ions for larger ions in the oxygen octahedral could modify the polarization of NBT-based materials effectively. In this case, when larger Mg occupied Ti sites, the Mg was centrally located in the oxygen octahedral and got “stuck”. Since the polarization of the system mainly originated from the Ti deviating from the central region of the octahedra, relatively larger Mg prevented the deviation and thus produced changes in the polarization, besides, the larger Mg inflated the octahedra which probably altered the adjacent octahedra and thus inhibited the formation of ferroelectric domains. As a result, the tetragonality decreased and the formation of the ferroelectric domains was suppressed. What should be mentioned here is why there is no TiO phase. TiO content is too low so that it cannot be detected by XRD. In addition, TiO may be involved in the reaction in Eq. (1). Therefore, the existence of TiO will not affect XRD results.

Fig. 1. (color online) X-ray diffraction patterns of calcined and sintered NaBiTi O (a) and NaBi(Ti Mg O powders (b).
3.2. Relative permittivity

Impedance spectroscopy measures the dielectric properties of ceramic samples as a function of frequency. It is based on the interaction of an external field with the electric dipole moment of the sample, usually expressed by permittivity. The frequency dependence of relative permittivity ( ) of NaBiTi O and NaBi(Ti Mg O ceramics sintered at 1020, 1040, 1060, and 1080 °C are shown in Fig. 2. By processing the experimental data, we can find that the relative permittivity of NaBi(Ti Mg O is a little lower than that of NaBiTi O . Figure 3 shows the temperature dependence of relative permittivity of NaBiTi O and NaBi(Ti Mg O ceramics at 1 MHz. Despite 1020 °C, the relative permittivity of NaBi(Ti Mg O is lower than that of NaBiTi O . The relative permittivity drops pretty fast at first, and then keeps stable, and finally increases quickly. We can describe the general trend of permittivity as a ‘U’ shape. At different frequencies, the general trend remains the same. Relative permittivity decreases with the increase of frequency. At different sintering temperatures, the general trend still remains a ‘U’ shape. Relative permittivity decreases with the increase of sintering temperature. When the sintering temperature is 1040 °C at 1 MHz, the relative permittivity is 27.2. With respect to ion radius, the radius of Mg is 0.072 nm, the radius of Ti in the B site is 0.064 nm. The radius of Mg is similar to that of Ti , so it is possible to occupy Ti ’s position. When Mg occupies the B site of Ti , oxygen vacancies will appear. The deformation of the perovskite structure makes it difficult for the domain to move. When oxygen vacancies diffuse to the domain wall, they will nail the domain easily and block the movement of the domain which makes it difficult for the domain to change direction. Therefore, piezoelectric and dielectric properties of ceramics will decrease.[16]

Fig. 2. (color online) Relative permittivity of NaBiTi O and NaBi(Ti Mg O ceramics sintered at (a) 1020 °C, (b) 1040 °C, (c) 1060 °C, and (d) 1080 °C.
Fig. 3. (color online) The relative permittivity of NaBiTi O and NaBi(Ti Mg O ceramics at 1 MHz at different temperatures.\vglue3pt
3.3. Dielectric loss

The dielectric loss of the samples sintered at 1020, 1040, 1060, and 1080 °C is shown in Fig. 4. Figure 4 shows the frequency dependence of the dielectric loss tangent of NaBiTi O and NaBi(Ti Mg O at different temperatures. In all sintering temperatures, the general trend is that of NaBi(Ti Mg O is lower than that of NaBiTi O . For example, when the sintering temperature is 1040 °C at 1 MHz, the minimum dielectric loss of NaBi(Ti Mg O is 0.00248. The minimum dielectric loss of NaBiTi O under the same condition is 0.100. It is very obvious that after Mg-doping, the dielectric loss becomes lower. Figure 4 also shows that the dielectric loss remains a very low level from around 100 °C to 600 °C, which implies that NaBi(Ti Mg O can be comprehensively used in a ceramic capacitor, microwave ceramics with a pretty wide operational temperature. Similar to the change trend of the relative permittivity, the dielectric loss drops very fast at first, then keeps stable, and finally increases quickly. We can describe the general trend of dielectric loss as a ‘U’ shape. At different frequencies, the general trend remains the same. Dielectric loss decreases with the increase of frequency. At different sintering temperatures, the general trend still remains a ‘U’ shape. Dielectric loss decreases with the increase of sintering temperature. Figure 5 shows the temperature dependence of the dielectric loss tangent (tan of the NaBiTi O and NaBi(Ti Mg O at 1 MHz. As the temperature increases, the dielectric loss keeps on decreasing until the temperature reaches 1060 °C, and then increases a little between 1060 °C and 1080 °C. There is a theory that may explain the reason why the dielectric loss increases between 1060 °C and 1080 °C. The interface polarization between the grain and grain boundary is weak. With the increase of temperature, the resistivity of the ceramics decreases and the conductivity of the grains becomes obvious, namely that the interaction between ceramics and electrodes begins to influence ceramic properties.[17] Consequently, at high temperatures, the dielectric loss increases with the increasing temperature. Generally, the dielectric loss of NaBi(Ti Mg O is about one order of magnitude lower than NaBiTi O at the same sintering temperature, which indicates that the dielectric property of NaBi(Ti Mg O is better than that of NaBiTi O . The reason why the dielectric loss becomes lower after Mg-doping in NaBiTi O might be attributed to the decrease of the grain size. Large grain size causes the decrease of the ratio of the grain boundary. Generally, the defects in the grain boundary are less than that in grain, leading to the decrease of the dielectric loss.[1820]

Fig. 4. (color online) The dielectric loss of NaBiTi O and NaBi(Ti Mg O ceramics sintered at (a) 1020 °C, (b) 1040 °C, (c) 1060 °C, and (d) 1080 °C.
Fig. 5. (color online) The dielectric loss of NaBiTi O . and NaBi(Ti Mg O ceramics at 1 MHz at different temperatures.
3.4. Dielectric properties

The bulk densities of disks were measured as a function of the sintering temperature and the results are listed in Table 1. It is obviously seen in Table 1 that the bulk density of undoped samples (NaBiTi O ) increases with the sintering temperature increasing. However, after Mg-doping, the bulk density of samples is always maintained at about 3.6 g/cm which is not very different from the undoped samples. As a result, it is concluded that densification of the system can be improved by higher sintering temperatures. We can also see from Table 1 that insulation resistance of all samples is roughly maintained at six party orders of magnitude. High insulation resistance demonstrates all the samples are dielectrics. Table 2 shows the dielectric properties (the minimum relative and permittivity dielectric loss) of the samples sintered at 1020, 1040, 1060, and 1080 °C at 1 MHz. In general, with the increasing sintering temperature, the relative permittivity of all samples is between 22 and 36, and there is not much difference. The relative permittivity and the dielectric loss of NaBiTi O increase gradually with the sintering temperature increasing. The relative permittivity of NaBi(Ti Mg O does not vary much with the change of sintering temperature. It is demonstrated once again that the relative permittivity and the dielectric loss of NaBi(Ti Mg O are lower than that of NaBiTi O at the same sintering temperature except 1020 °C, indicating that NaBi(Ti Mg O ceramics have excellent dielectric property. These intriguing properties make this material qualify for practical applications.

The scanning electron microscope (SEM) micrographs of NaBiTi O and NaBi(Ti Mg O sintered at different temperatures are shown in Fig. 6 and Fig. 7, respectively. To a certain degree, the dielectric constant can reflect the function of materials. Comparing the two pictures, we can observe that in general, the grain size of Mg-doped samples is slightly smaller than that of NaBiTi O . ceramics. More specifically, the grain shown in Fig. 6 is more homogeneous. The domain reorientation of the ceramics becomes more difficult and restricts the movement of the domain wall due to the strong coupling between the grain boundaries and the domain wall. Then the value of remnant polarization decreases, and the relative permittivity also increases accordingly. In the end, we can draw a conclusion that the relative permittivity and the dielectric loss of NaBi(Ti Mg O is smaller than that of NaBiTi O at the same condition, which is consistent with Table 2.

Table 1.

Dielectric properties of NaBiTi O and NaBi(Ti Mg O samples sintered at different temperatures.

.
Table 2.

Dielectric properties of NaBiTi O and NaBi(Ti Mg O samples sintered at different temperatures.

.
Fig. 6. SEM of NaBiTi O ceramics sintered at (a) 1020 °C, (b) 1040 °C, (c) 1060 °C, and (d) 1080 °C.
Fig. 7. SEM of NaBi(Ti Mg O ceramics sintered at (a) 1020 °C, (b) 1040 °C, (c) 1060 °C, and (d) 1080 °C.
4. Conclusion

In this article, we studied the dielectric property by comparing NaBiTi O and NaBi(Ti Mg O , we mainly focused on the relative permittivity and the dielectric loss of the samples and found that the dielectric property of NaBi(Ti Mg O is better than that of NaBiTi O .

(i) According to the above discussion, we find that despite 1020 °C, the relative permittivity ( ) of NaBi(Ti Mg O is lower than that of NaBiTi O at the same sintering temperature. In addition, the dielectric loss remains at a very low level around 100 °C to 600 °C, indicating that the dielectric property of NaBi(Ti Mg O is better than that of NaBiTi O . The dielectric property of NaBi(Ti Mg O implies that NaBi(Ti Mg O can be comprehensively used in microwave ceramics with a pretty wide operating temperature.

(ii) The change of the permittivity after Mg-doping is mainly contributed to the oxygen vacancies. When Mg occupies the B site of Ti , oxygen vacancies will appear. The deformation of the perovskite structure octahedron makes domain moving difficult. When oxygen vacancies diffuse to the domain wall, they will nail the domain easily and block the movement of the domain which makes it difficult for the domain to change direction in poling. Therefore, piezoelectric and dielectric properties of ceramics will decrease.

(iii) The change of the dielectric loss after Mg-doping is mainly attributed to the interface polarization and the grain size. The interface polarization between the grain and the grain boundary is weak. With temperature rising, the resistivity of the ceramics decreases and the conductivity of grains becomes obvious, namely that the interaction between ceramics and electrodes begins to influence ceramic properties. Consequently, at high temperatures, the dielectric loss increases with the increasing temperature. As for grain size, large grain size causes the decrease of the ratio of the grain boundary. Generally, the defects in the grain boundary are less than that in the grain, leading to the decrease of the dielectric loss.

Reference
[1] Cao W Q Liu P Z Chen Y et al. 2016 Acta Phys. Sin. 65 137701 in Chinese
[2] Chen Y Gong S Zhang B et al. 2013 Microelectron. Eng. 103 106
[3] Huang H H Chen Y Li Z et al. 2015 J. Wuhan University of Tech. -Mater. Sci. Edn. 30 674
[4] Cao W Q Shang X Z 2015 Ferroelectr. Lett. 42 132
[5] Chen Y Zhang S Li Z et al. 2016 J. Wuhan University of Tech. -Mater. Sci. Ed. 31 87
[6] Chen Y Gong S P Fu Q Y et al. 2010 Chin. Phys. 19 117301
[7] Hansen P Hennings D Schreinemacher H 1998 J. Am. Ceram. Soc. 81 1369
[8] Kishi H Kohzu N Iguchi Y Sugino J Kato M 2001 J. Eur. Ceram. Soc. 21 1643
[9] Nam S P Lee S G Bae S G Lee Y H 2007 J. Electrical Eng. Technol. 2 98
[10] Zang J Li M Sinclair D C J Wook R Jürgen 2014 J. Am. Ceram. Soc. 97 1523
[11] Li M Pietrowski M J De Souza R A Zhang H Reaney I M Cook S N Kilner J A Sinclair D C 2014 Nat. Mater. 13 31
[12] Huang X Gao C Wei M 2011 Rare Met. 30 72
[13] Guo Y Xiao P Luo L et al. 2014 J. Mater. Sci.: Mater. Electron. 25 3753
[14] Zang J Li M Sinclair D.C. Jo W. Rödel J. 2014 J. Am. Ceram. Soc. 97 1523
[15] Li M Feteira A Mirsaneh M Lee S Lanagan M T Randall C A Sinclair D C 2010 J. Am. Ceram. Soc. 93 4087
[16] Li Z F Chan W Li Y X Kwok K W Choy S H 2010 J. Alloys Compd. 41 70
[17] Feng C Yang C H Geng F J Lv P P Yao Q 2015 J. Eur. Ceram. Soc. 36
[18] Lei W Yan Y Y Wang X H Lu W Yang Z B Lu W Z 2015 Ceram. Int. 41 521
[19] Li M Feteira A Mirsaneh M Lee S Lanagan M T Randall C A Sinclair D C 2010 J. Am. Ceram. Soc. 93 4087
[20] Chen Y Bian M Y Zhang C C Chen Q Luo Q Huang Z Cao W Q Pan R K Qi Y J 2017 Sci. China Ser. E 47 32 in Chinese