*2.1. Sample Preparation*

Highly purified metal oxides of SrCO<sup>3</sup> (99%), Bi2O<sup>3</sup> (99%), TiO<sup>2</sup> (99%), CeO<sup>2</sup> (99.99%) and Gd2O<sup>3</sup> (99.99%) were weighed according to the stoichiometric formula of three designed compositions: SrBi4Ti4O<sup>15</sup> (SBT), Sr0.92Gd0.053Bi4Ti4O<sup>15</sup> (SGBT) and Sr0.96Ce0.04Bi4Ti4O<sup>15</sup> (SCBT), respectively. These powders as starting materials were milled for 10 h under the condition of ethanol as a dissolvent and zirconium ball as a milling medium. The dried mixtures were calcined at 850 ◦C for 4 h, and then the calcined powders of SGBT and SCBT were divided into four equal parts according to their quality, respectively. Secondly, 0.2 wt% of CeO<sup>2</sup> (99.99%), MnO<sup>2</sup> (99%) and Cr2O<sup>3</sup> (99%) were severally added into three parts of SGBT powders, while 0.2 wt% of Gd2O<sup>3</sup> (99.99%), MnO<sup>2</sup> (99%) and Cr2O<sup>3</sup> (99%) were severally added into three parts of SCBT powders. Then, the mixtures of SBT, SGBT, SGBT-Ce, SGBT-Mn, SGBT-Cr, SCBT, SCBT-Gd, SCBT-Mn, SCBT-Cr were milled again in the same requirement. After drying, polyvinyl alcohol (PVA) as binder was added to the uniform mixture to form granules. The granules were pressed into pellets of 10 mm in diameter and 1 mm in thickness. After burning out PVA at 600 ◦C for 2 h, these pellets were sintered between 1100 ◦C and 1200 ◦C for 2 h in a sealed alumina crucible to obtain the ceramics with the maximum density.

### *2.2. Sample Characterization* Meanwhile, the relative density of all sintered samples was calculated as the ratio of the

*2.2. Sample Characterization* 

The crystallographic structure of all sintered samples was determined by an X-ray diffractometer (DX2700, Dandong, China) employing Cu-Kα radiation (*λ* = 1.5418 Å). Meanwhile, the relative density of all sintered samples was calculated as the ratio of the apparent density measured by the Archimedes method to the theoretical density obtained from crystallographic structures (XRD). In order to measure the electrical properties, the samples were polished and fired with silver paste as the electrodes at 700 ◦C for 10 min. The dielectric constant (*ε*r) and loss tangent (*tan δ*) as a function of temperature were recorded using an LCR analyzer (TH2829A, Tonghui, China) attached to a programmable furnace. Samples were poled under a DC field of 6–10 kV/mm for 15 min in a silicone oil bath at 150 ◦C. The electrical impedance (|*Z*|) and phase angle (*θ*) as a function of frequency was measured using an impedance analyzer (TH2829A, Tonghui, China). The planar electromechanical coupling factor (*k*p), mechanical quality factor (*Q*m) and planar frequency constant (*N*p) were calculated by the IEEE standard. Thermal depoling behavior was investigated by annealing the polarized samples at different temperatures for 4 h, and then the piezoelectric charge coefficient (*d*33) was remeasured using a quasi-static *d*<sup>33</sup> m (ZJ-6AN, IACAS, Beijing, China) when the samples were cooled to room temperature. apparent density measured by the Archimedes method to the theoretical density obtained from crystallographic structures (XRD). In order to measure the electrical properties, the samples were polished and fired with silver paste as the electrodes at 700 °C for 10 min. The dielectric constant (*ε*r) and loss tangent (*tan δ*) as a function of temperature were recorded using an LCR analyzer (TH2829A, Tonghui, China) attached to a programmable furnace. Samples were poled under a DC field of 6–10 kV/mm for 15 min in a silicone oil bath at 150 °C. The electrical impedance (|*Z*|) and phase angle (*θ*) as a function of frequency was measured using an impedance analyzer (TH2829A, Tonghui, China). The planar electromechanical coupling factor (*k*p), mechanical quality factor (*Q*m) and planar frequency constant (*N*p) were calculated by the IEEE standard. Thermal depoling behavior was investigated by annealing the polarized samples at different temperatures for 4 h, and then the piezoelectric charge coefficient (*d*33) was remeasured using a quasi-static *d*33 m (ZJ-6AN, IACAS, Beijing, China) when the samples were cooled to room temperature. **3. Results and Discussion** 

The crystallographic structure of all sintered samples was determined by an X-ray diffractometer (DX2700, Dandong, China) employing Cu-Kα radiation (*λ* = 1.5418 Å).

*Materials* **2021**, *14*, x FOR PEER REVIEW 3 of 11

### **3. Results and Discussion** *3.1. Phase Structure of Ceramics*

### *3.1. Phase Structure of Ceramics* The appearances of the oxide-modified SBT piezoceramics are presented in Figure 1.

The appearances of the oxide-modified SBT piezoceramics are presented in Figure 1. As can be seen from these figures, the pure SBT ceramic seems to be taupe; after it was doped with different oxides, different colors were presented. However, all these oxide-modified SBT piezoceramics were sintered in a uniform color and free of cracks, blotches, striations and holes, at least seen from their surfaces. The change of color also proves that the oxides as additives have dissolved into SBT, leading different color emerging mechanisms to the ceramics. As can be seen from these figures, the pure SBT ceramic seems to be taupe; after it was doped with different oxides, different colors were presented. However, all these oxidemodified SBT piezoceramics were sintered in a uniform color and free of cracks, blotches, striations and holes, at least seen from their surfaces. The change of color also proves that the oxides as additives have dissolved into SBT, leading different color emerging mechanisms to the ceramics.

**Figure 1.** Appearances of the oxide-modified SBT ceramics (the corresponding chemical compositions (marked with (**a**–**i**) respectively) are located above the samples). **Figure 1.** Appearances of the oxide-modified SBT ceramics (the corresponding chemical compositions (marked with (**a**–**i**) respectively) are located above the samples).

The XRD patterns of the oxide-modified SBT piezoceramics are shown in Figure 2. It can be seen that these samples display a single SrBi4Ti4O15 phase crystallized in the orthorhombic structure with *Bb*21*m* (36) space group (JCPDS No: 43-0973). There is no impurity detected from XRD patterns, which indicates that these oxide additives have been incor-The XRD patterns of the oxide-modified SBT piezoceramics are shown in Figure 2. It can be seen that these samples display a single SrBi4Ti4O<sup>15</sup> phase crystallized in the orthorhombic structure with *Bb*21*m* (36) space group (JCPDS No: 43-0973). There is no impurity detected from XRD patterns, which indicates that these oxide additives have been

porated into the crystal lattice of SrBi4Ti4O15. The strongest diffraction of all these samples

incorporated into the crystal lattice of SrBi4Ti4O15. The strongest diffraction of all these samples appears at the (1 1 9) peak, stating the fact that SrBi4Ti4O<sup>15</sup> belongs to the BLSF with the structure of four layer (*m* = 4) [30]. Some variations observed from the details of XRD patterns can be related with the lattice distortions of SBT caused by doping. By contrasting with the pure SBT, three diffraction peaks of the doped SBT: (0 0 10), (0 0 16) and (0 0 20) are weakened, which indicates that their grains orientating along the *c*-axis becomes fewer [31]. Inversely, the diffraction peaks of (2 0 0)/(0 2 0) are enhanced, which states that the number of grains oriented along the *a*-*b* plane increased. appears at the (1 1 9) peak, stating the fact that SrBi4Ti4O15 belongs to the BLSF with the structure of four layer (*m* = 4) [30]. Some variations observed from the details of XRD patterns can be related with the lattice distortions of SBT caused by doping. By contrasting with the pure SBT, three diffraction peaks of the doped SBT: (0 0 10), (0 0 16) and (0 0 20) are weakened, which indicates that their grains orientating along the *c*-axis becomes fewer [31]. Inversely, the diffraction peaks of (2 0 0)/(0 2 0) are enhanced, which states that the number of grains oriented along the *a*-*b* plane increased.

**Figure 2.** XRD patterns of the oxide-modified SBT ceramics: (**a**) SGBT-series; (**b**) SCBT-series. **Figure 2.** XRD patterns of the oxide-modified SBT ceramics: (**a**) SGBT-series; (**b**) SCBT-series.

The lattice parameters of the oxide-modified SBT ceramics are given in Table 1. The lattice parameters (*a*, *c* and *v*) of the oxide-modified SBT decrease, whereas the values of orthorhombic distortion (*b*/*a*) increase, which may be attributed to the ion-substitution effect caused by the addition of different oxides. The bismuth oxide layer is very strong and bismuth ion in the bismuth oxide layer are difficult to be substituted by other ions [32], meanwhile ions with similar ionic radius and same coordination number are more likely to be mutually substituted [31], consequently Sr2+ located at the A-site in the perovskite layers are substituted by Gd3+/Ce4+ with smaller ion radius. Ti4+ located at the B-site in the perovskite layers would be substituted by Mn3+ and Cr3+. The lattice distortion caused by ion substitution can result in the change of electrical properties for ferroelectric compounds [33]; the larger *b*/*a* value is, more distorted the lattice is. Among these compo-The lattice parameters of the oxide-modified SBT ceramics are given in Table 1. The lattice parameters (*a*, *c* and *v*) of the oxide-modified SBT decrease, whereas the values of orthorhombic distortion (*b*/*a*) increase, which may be attributed to the ion-substitution effect caused by the addition of different oxides. The bismuth oxide layer is very strong and bismuth ion in the bismuth oxide layer are difficult to be substituted by other ions [32], meanwhile ions with similar ionic radius and same coordination number are more likely to be mutually substituted [31], consequently Sr2+ located at the A-site in the perovskite layers are substituted by Gd3+/Ce4+ with smaller ion radius. Ti4+ located at the B-site in the perovskite layers would be substituted by Mn3+ and Cr3+. The lattice distortion caused by ion substitution can result in the change of electrical properties for ferroelectric compounds [33]; the larger *b*/*a* value is, more distorted the lattice is. Among these compositions, the unit cell of SGBT-Cr has the largest orthorhombic distortion with a *b*/*a* value of 1.0024.



*V* (Å3) 1209.54 1207.65 1208.08 1206.68 1206.46 1208.83 1209.12 1208.38 1207.85 *b/a* 1.0010 1.0014 1.0016 1.00155 1.0024 1.0019 1.0018 1.0020 1.0022 Table 2 lists the density of the oxide-modified SBT ceramics. The relative density of SBT is measured as 94.8%, which has been changed after the addition of different oxides. Table 2 lists the density of the oxide-modified SBT ceramics. The relative density of SBT is measured as 94.8%, which has been changed after the addition of different oxides. According to the results given by Table 2, the addition of CeO<sup>2</sup> and MnO<sup>2</sup> played a positive effect on the densifying process of the SBT ceramic during sintering.

According to the results given by Table 2, the addition of CeO2 and MnO2 played a posi-


**Table 2.** Density data of the oxide-modified SBT ceramics.
