Bi nonstoichiometry and composition engineering in (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ ceramics

Abstract

Bi is easily evaporated during the preparation of BiFeO₃, and moreover its phase structure and electrical properties can also be modified by doping BaTiO₃. In this work, the effects of Bi nonstoichiometry as well as BaTiO₃ on the phase purity, microstructure, and electrical properties of (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ (0.25 ≤ x ≤ 0.35 and −0.1 ≤ y ≤ 0.25) ceramics were studied. A rhombohedral–pseudocubic (R–C) phase boundary was formed in the ceramics with a wide composition range of x = 0.30 and −0.05 ≤ y ≤ 0.15, leading to an enhanced piezoelectricity (d₃₃ ∼ 180 pC/N). In addition, one can find that dense microstructure as well as the improvement of Curie temperature (T_C ∼ 506 °C) and thermal stability of piezoelectricity (T_a = 400 °C) can be achieved in the ceramics with excess Bi (x = 0.30, y = 0.05) with respect to the ones with Bi deficiency. We expect that the investigation on Bi nonstoichiometry and the addition of BaTiO₃ will point out an effective method to fabricate high-performance BFO-based ceramics.

---

1. Introduction

Lead-free BiFeO₃–BaTiO₃ (BFO–BTO) solid solutions have become one of the most promising candidates for high-temperature piezoelectric applications due to their environment-friendly characteristics and relatively high piezoelectricity.^1–7 The key properties of BFO–BTO ceramics, such as piezoelectricity and ferroelectricity, have been extensively investigated in the past several years.^2–6 Particularly, the electrical properties of BFO–BTO ceramics can be improved by the construction of phase boundaries.^3,7 Among these studies, the main methods to improve their electrical properties can be shown here: (1) improvement of preparation process.^8–14 It was previously reported that the modified conventional solid-state method with a quenched process is effective for the enhancement of electrical properties due to the freezing of disordered defect states.^8,9,14 (2) Site engineering.^15–21 Ion substitutions for Bi (e.g., La, Nd, Eu) or Fe (e.g., Al, Sc, Co, Cr) in BFO–BTO ceramics were carried out, and improved electrical properties were also achieved.^15–21 (3) Construction of ternary systems.^22–26 A series of third members [e.g., Bi_0.5Na_0.5TiO₃, Bi_0.5K_0.5TiO₃, BiM_0.5Ti_0.5O₃ (M: Zn, Ni, Mg)] were utilized to modify BFO–BTO ceramics with the aim of improving their resistivity and electrical properties.^22–26 Recently, a large piezoelectricity (d₃₃ = 402 pC/N) together with a high Curie temperature (T_C = 454 °C) have been reported in BFO–BTO–BiGaO₃ ceramics, indicating that the BFO–BTO material systems may become the most promising candidate.²⁷ However, little attention was paid to the phase boundary and electrical properties of pure BFO–BTO ceramics with Bi nonstoichiometry.

For BFO-based ceramics, one of the tough issues is low electrical resistivity, thus leading to the difficulty in poling process. Previously, it was reported that the Bi evaporation during high-temperature process can induce the formation of some defects (i.e., free electrons or charged vacancies), thus resulting in a high conductivity.²⁸ It was found that the Bi content strongly affected electrical properties of Bi-based piezoceramics. For example, the difficulty in the poling process for (Bi_0.5+xNa)TiO₃ ceramics may be ascribed to the pinning of domain walls because the Bi evaporation can bring about oxygen vacancies.²⁹ An enhancement of piezoelectricity and ferroelectricity can also be achieved in Bi₂O₃-doped (K_0.5Na_0.5)NbO₃ ceramics.³⁰ Therefore, it is essential to investigate the effects of Bi nonstoichiometry on phase structure and electrical properties of BFO–BTO ceramics. Although the effects of Bi excess on the structure and piezoelectricity of BFO–BTO ceramics were carried out, there are few reports on the effects of Bi deficiency on their electrical properties, and the corresponding ferroelectricty as well as thermal stability were also few mentioned.³¹ Therefore, it is necessary to conduct the systematic researches about the effects of Bi nonstoichiometry as well as BaTiO₃ on the phase structure and electrical properties of BFO ceramics.

In this study, the effects of Bi nonstoichiometry and BaTiO₃ on the phase structure, microstructure, and electrical properties of (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ (x = 0.25–0.35 and y = −0.1–0.25) ceramics were studied systematically. Rhombohedral–pseudocubic (R–C) phase boundary was formed in the ceramics through the addition of BaTiO₃. A large piezoelectric coefficient (d₃₃ ∼ 180 pC/N) can be achieved in a wide composition range of x = 0.30 and −0.05 ≤ y ≤ 0.15. In addition, appropriate excess Bi can promote the ferroelectricity, the Curie temperature, and the thermal stability of BFO–BTO ceramics. The related physical mechanisms were also illuminated.

2. Experimental procedure

Lead-free (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ (0.25 ≤ x ≤ 0.35 and −0.1 ≤ y ≤ 0.25) ceramics were synthesized by the conventional solid-state method with a quenching method. Here, we need to point out that the y range of −0.1 to 0.25 is the nominal value, and then the actual y composition range was identified to be −0.02–0.22 according to the EDS data. In addition, the error of atom percent of each sample is not more than 10%, indicating that chemical composition is relatively stable in all samples. Therefore, in order to simplify the reader's understanding, we will still discuss related phenomenon using the nominal composition range. And we will also point out the corresponding actual Bi content in the manuscript. Raw materials including Bi₂O₃ (99%), Fe₂O₃ (99%), BaCO₃ (99%), and TiO₂ (98%) were weighted and ball milled for 24 h with alcohol. Those mixing slurries were dried and calcined at 700 °C for 2 h. After that, the pellets with 10 mm diameter and 0.6 mm thickness were pressed under a pressure of 10 MPa using 6–8 wt% polyvinyl alcohol (PVA) as a binder. After burning off PVA at 500 °C for 3 h, the pellets were sintered at 960–1000 °C for 3 h in air and then quenched in water, and the heating and cooling rates are 10 °C min⁻¹. For electrical measurement, both sides of the sintered samples were pasted on silver slurry and then fired at 600 °C for 10 min. The samples were poled at 120 °C in a silicone oil bath under a dc field of 5 kV mm⁻¹.

X-ray diffraction (XRD) (Bruker D8 Advanced XRD, Bruker AXS Inc., Madison, WI, CuKα) was used to measure the phase structure. The temperature dependence of capacitance and dielectric loss were tested using an LCR analyzer in the temperature range of 20–550 °C. The field emission-scanning electron microscopy (FE-SEM) (JSM-7500, Japan) was used to characterize the surface microstructure of the sintered samples. Their polarization–electric (P–E) hysteresis loops were measured by Radiant Precision Workstation (USA) at f = 10 Hz and room temperature. The piezoelectric constant (d₃₃) was measured by piezo-d₃₃ meter (ZJ-3 A, China) and approximately 24 h after poling.

3. Results and discussion

Due to the close relationships between phase structure and piezoelectricity, it is essential to identify the phase structure of BFO–BTO ceramics. Fig. 1(a) and (b) show the room-temperature XRD patterns of the ceramics as a function of x (y = 0.05) and y (x = 0.30), measured in the 2θ range of 20–70°. The addition of BaTiO₃ into BiFeO₃ matrix can form a stable solid solution, resulting in a pure perovskite structure. The phase transition of the ceramics is confirmed by the expanded XRD patterns at 2θ = 31–32° and 38–40°. It is clearly to find that there is a composition dependence of their phase structure. For example, the ceramics can endure the phase transition from rhombohedral to pseudocubic phases with the increase of x, leading to the formation of R–C phase boundary for 0.30 ≤ x ≤ 0.33. In addition, a pseudocubic phase can be detected in the ceramics with y = −0.1 (actual y content: y = −0.02) [Fig. 1(b)], which can also be identified by its ε_r–T curve in Fig. 5(c). With the further increase of y contents, the multiphase coexistence containing rhombohedral and pseudocubic phases was formed. Therefore, composition design has induced the formation of a wide R–C phase boundary in the ceramics. In addition, it can be observed from Fig. 1(b) that the characteristic peak shifted to low angle with the increase of y contents (0 ≤ y ≤ 0.15), and then shifted to high angle when y = 0.25 (actual y content: y = 0.22).

Fig. 1 XRD patterns of (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ ceramics with different (a) x (y = 0.05) and (b) y (x = 0.30) contents.

As we all know, the ion occupying can result in the formation of point defects, which has a close relationship with microstructure and electrical properties of perovskite piezoceramics.³² Here, we systematically explored the doping mechanisms of the ceramics with different y contents. Obviously, the severe loss of Bi₂O₃ induced the formation of Bi vacancies (V_Bi³⁺). The volatilization of Bi₂O₃ also caused the formation of both Bi (V_Bi³⁺) and O vacancies (V_O²⁺). However, the excess Bi contents compensated the Bi volatilization and resulted in different ion occupying mechanisms. The ionic radii of Bi³⁺, Ba²⁺, and Ti⁴⁺ are 0.103 nm, 0.135 nm, and 0.0605 nm respectively. Therefore, excess Bi³⁺ occupied the Ba or Ti site. Combined with the characteristic peaks shift, the variations of crystal cell volume and the ionic radius, we can deduce that slight excess of Bi³⁺ (0.05 ≤ y ≤ 0.15) can substitute Ti site, leading to the expansion of crystal cell volume as well as the peak shift to low angle. However, the excessive Bi contents of y = 0.25 (actual y content: y = 0.22) can substitute Ba site, resulting in the shrinkage of crystal cell volume and the peak shift to high angle. The defect reactions were listed below:

2Bi³⁺ + 3O²⁻ → Bi₂O₃↑ + 2V_Bi³⁻ + 3V_O²⁺ (1)

Bi₂O₃ → 2Bi_Ti⁻ + 3O_O + V_O²⁺ (2)

Bi₂O₃ → 2Bi_Ba⁺ + 3O_O + V_Ba²⁻ (3)

As a result, the point defects contain a large number of Bi vacancies (V_Bi³⁻) and some O vacancies (V_O²⁺) for the ceramics with Bi deficiency due to both the severe loss of Bi₂O₃ and the volatilization of Bi₂O₃. The eqn (1) exhibits the Bi volatilization as well as the formation of both Bi V_Bi³⁻ and V_O²⁺ for the ceramics with y = 0. Eqn (2) illustrates that the Bi occupied Ti site, leading to the formation of a small number of V_O²⁺ in the ceramics with 0.05 ≤ y ≤ 0.15. Finally, the formation of V_Ba²⁻ for the ceramics with y = 0.25 (actual y content: y = 0.22) was due to the occupying of Bi for Ba site.

In order to analyze the valence state and oxygen vacancies of BFO–BTO ceramics with Bi nonstoichiometry, we carried out the element composition by XPS spectra, as shown in Fig. 2(a)–(c). In addition, the fitting results were included in Table 1. Fig. 2(a) shows the fitting results of Bi 4f for the ceramics with y = −0.1 (actual y content: y = −0.02), 0.05, and 0.25 (actual y content: y = 0.22). The main peaks corresponding to Bi 4f_5/2 and Bi 4f_7/2 were observed, and the separation between two peaks is about 5.3 eV, confirming the existence of Bi³⁺.^33–35 One can also see a small amount of atomic or metallic Bi exists located at ∼159 eV, indicating the existence of Bi⁺ (ref. 34) In addition, the percentage of Bi⁺ decreased firstly and then increased with the increase of Bi contents, reaching the lowest value of 5% for y = 0.05 [Table 1]. Fig. 2(b) shows the asymmetrical O 1s photoelectron peaks divided into three sub-peaks. The lowest peaks in binding energy located at ∼529 eV correspond to the cation–oxygen bonds, while two further higher energy peaks located at ∼530 eV and at ∼532 eV were assigned to adsorbed surface H₂O as well as the presence of oxygen vacancies.^35–37 The percentage of oxygen vacancies of the ceramics with the increase of y content was 64.0%, 63.1%, and 59.6%, respectively. This result indicated that oxygen vacancies decreased gradually as the Bi contents increased. The experimental results just coincide with the previous theoretical analysis. At last, the high resolution XPS scans of Fe 2p region was exhibited in Fig. 2(c), and the energy peak corresponding to Fe 2p_3/2 can be used to calculate the ratio of Fe³⁺ and Fe²⁺. Fe³⁺ locates at ∼711 eV, while Fe²⁺ locates at a lower energy peak of 709 eV. According to the fitting results, the concentration ratios of Fe³⁺ to Fe²⁺ were respectively 50.5 : 49.5, 53.7 : 46.3, and 59.6 : 40.4 with increasing y contents (see Table 1), demonstrating that the concentrations of Fe²⁺ decreased gradually with the increase of Bi. Generally speaking, Fe²⁺ and oxygen vacancies can emerge simultaneously in BFO for charge compensation, and thus less Fe²⁺ means less oxygen vacancies.^35,38 In this work, the concentration of Fe²⁺ decreased gradually with the increase of Bi contents, which is just coincident with the results of the concentration variation of oxygen vacancies. This result indicates that excess Bi contents are indeed beneficial to the decrease of Fe²⁺ and oxygen vacancies.

Fig. 2 Valence states of the elements in (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ ceramics with y = −0.1, 0.05, and 0.25. (a) Bi 4f, (b) O 1s, and (c) Fe 2p.

Table 1 Fitted peaks of Bi 4f, O 1s, and Fe 2p

Bi 4f	Peaks	Atomic%	O 1s	Peaks	Atomic%	Fe 2p3/2	Peaks	Atomic%
y = −0.1	163.8	42.3	y = −0.1	531.6	64.0	y = −0.1	711.1	50.5
	159.3	6.1		531.0	3.1
	158.5	51.6		529.1	32.9		709.6	49.5
y = 0.05	164.1	43.6	y = 0.05	531.9	63.1	y = 0.05	711.1	53.7
	159.5	5.0		530.4	10.5
	158.8	51.4		529.3	26.4		709.5	46.3
y = 0.25	163.9	43.5	y = 0.25	531.7	59.6	y = 0.25	711.0	59.6
	159.3	5.9		530.4	12.4
	158.5	50.5		529.1	28.0		709.5	40.4

---

Fig. 3 displays the surface microstructure of the ceramics with different x (y = 0.05) and y (x = 0.30). As shown in Fig. 3(a)–(d), dense microstructure, conspicuous grain boundaries as well as similar grain sizes can be found in those ceramics with different x, demonstrating that the microstructure cannot be greatly affected by the variations of BT content. Nevertheless, there is an obvious change in microstructure for the ceramics with different y, as shown in Fig. 3(e)–(h). For example, the seriously refined grain sizes can be formed in the ceramics with y = −0.1 (actual y content: y = −0.02), a much denser microstructure can be obtained when appropriate excess of Bi₂O₃ were introduced, while too many Bi₂O₃ results in a loose microstructure of the ceramics with y = 0.25 (actual y content: y = 0.22). Thus a severe deficiency of Bi₂O₃ prohibited the grain growths and caused the refinement of grain sizes due to the formation of a large number of V_Bi³⁻ and V_O²⁺ [Fig. 3(e)]. However, too excessive Bi₂O₃ led to a loose microstructure [Fig. 3(h)]. As a result, appropriate compensation of Bi₂O₃ can contribute to the densification of the microstructure in BFO–BTO ceramics due to the easy volatilization of Bi during their high-temperature sintering. In order to clearly exhibit the variations of grain sizes, their grain size distribution can be calculated and presented in Fig. 4. One can see that the average grain sizes increased firstly and then decreased as both x and y increased, reaching a maximum value of about 7.07 μm for x = 0.30 and y = 0.05. The reduced grain sizes for the ceramics with y = −0.1(actual y content: y = −0.02) and y = 0.25 (actual y content: y = 0.22) may be due to the prohibited grain growth caused by massive point defects, such as V_Ba²⁻, V_Bi³⁻ and V_O²⁺.

Fig. 3 The surface microstructure of (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ ceramics with different x (y = 0.05) and y (x = 0.30) contents.

Fig. 4 Grain size distributions of the (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ ceramics with different x (y = 0.05) and y (x = 0.30) content.

As discussed above, Bi nonstoichiometry seriously affected the microstructure of the ceramics. Therefore, we measured the EDS data of different areas of the ceramics and then calculated the average values of the actual atom percent, as shown in Table 2. First of all, it can be seen that the measured Bi atom percent increased with the increase of y contents. The average atom percent of Bi element is 13.76%, 14.02%, and 16.62% for the ceramics with y = −0.1, 0, and 0.25, respectively. In addition, the error of atom percent of each sample is not more than 10%, indicating that chemical compositions are relatively stable in all samples. At last, according the EDS data, the actual composition range of y can be considered to be −0.02 to 0.22. Therefore, we have given out the nominal and actual composition range of y in this manuscript at the same time for the easier understanding.

Table 2 EDS data of Bi element in the ceramics with y = −0.1, 0, and 0.25

Composition	Area 1	Area 2	Average value	Nominal value	Error (%)
y = −0.1	13.67	13.86	13.76	12.78	7
y = 0	14.06	13.99	14.02	14	0.1
y = 0.25	16.92	16.32	16.62	16.91	2

---

Fig. 5(a) and (c) exhibit the temperature-dependent dielectric constant (ε_r–T) of the ceramics as a function of x (y = 0.05) and y (x = 0.30), measured at room temperature∼550 °C and at f = 100 kHz. First, it is obvious to find that the ε_r–T curve of the ceramics with y = −0.1 (actual content: y = −0.02) was seriously suppressed. The severe deficiency of Bi caused the formation of massive V_Bi³⁻, then leading to the increase of disorder degree in perovskite structure and the serious suppression of ε_r–T curves. Besides, the composition dependence of Curie temperature (T_C) of the ceramics was plotted in the inset of Fig. 5(a) and (c). We can know that their T_C declined rapidly with increasing x contents. However, their T_C values increased firstly and then decreased slightly with the increase of y, reaching a maximum value of ∼506 °C for x = 0.30 and y = 0.05. Consequently, appropriate compensation for Bi₂O₃ can greatly promote the T_C values of BFO–BTO ceramics. Fig. 5(b) and (d) exhibit the temperature dependence of tan δ of the ceramics with x (y = 0.05) and y (x = 0.30). It can be seen that their tan δ reduced gradually when the measurement temperatures increased from room temperature to 250 °C, and their tan δ values ascended higher and higher with the further increase of temperatures due to the increased conductivity.^18,22,25

Fig. 5 The temperature dependence of: (a, c) ε_r, and (b, d) tan δ of the (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ ceramics as a function of x (y = 0.05) and y (x = 0.30) content. T_C against x and y was plotted in the inset of (a) and (c), respectively.

Fig. 6(a) and (b) show the composition dependence of polarization–electric field (P–E) loops of the ceramics, measured at room temperature and f = 10 Hz. Compared with ions-modified BFO ceramics, it is easier to attain the saturated P–E loops in BaTiO₃-modified ceramics. Their P–E loops almost remained unchanged with the variations of BT contents [Fig. 6(a)], while the ceramics with too excess Bi₂O₃ contents (actual y content: y = 0.22) exhibited a slim P–E curve [Fig. 6(b)]. In order to distinctly reveal the composition dependence of ferroelectricity, we plotted the remnant polarization (P_r) and coercive field (E_C) of the ceramics against x and y, as shown in Fig. 6(c) and (d). According to recent literature, P_r almost has no changes in the composition range of 0.30 ≤ x ≤ 0.40 for (1 − x)BFO − xBTO ceramics, and a much poorer P_r value was observed in the ceramics with x = 0.20 and 0.50.²⁷ Therefore, in this work, we conjectured that the week composition dependence of ferroelectricity was ascribed to the quenched process. The quenched process can partly compensate the ferroelectricity for the compositions deviated from phase boundaries region. Appropriate excessive Bi contents (y = 0.05) can enhance the ferroelectricity of BFO–BTO ceramics due to the dense microstructure and improved electrical insulation. The weakened ferroelectricity was observed in the ceramics with severe loss and excess Bi content (actual y content: y = −0.02 and y = 0.22). It can be interpreted that the weakened ferroelectricity is due to the difficult mobility of domain walls – induced by the increased amount of defects.

Fig. 6 (a) and (b): P–E loops of (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ ceramics with different x (y = 0.05) and y (x = 0.30) contents. (c) and (d): P_r and E_C values as a function of x (y = 0.03) and y (x = 0.30) contents.

The composition dependence of d₃₃, ε_r, and tan δ of the ceramics was presented in Fig. 7(a) and (b). It can be seen that their ε_r has weak composition dependence at room temperature. Here, the relatively high tan δ (∼0.1) may be attributed to the involvement of dielectric loss peak around room temperature. Some researchers have reported a strong composition dependence of dielectric properties at room temperature.^39,40 However, there is a week composition dependence of dielectric properties for BFO–BTO ceramics with a quenched process reported by Song et al.²⁷ Therefore, the week composition dependence of dielectric properties may be attributed to the quenched technique. In the past, Yu et al. studied the dielectric behavior of BFO–BTO–BZT ceramics with MnO₂ addition and pointed out that high tan δ can be observed in the undoped ceramics due to the space charges, while the addition of MnO₂ can greatly decrease its tan δ.²⁴ In addition, the piezoelectricity of BFO–BTO ceramics can be greatly affected by the variations of x and y contents. As shown in Fig. 7(a), d₃₃ increased from 84 pC/N to 180 pC/N with the increase of x (0.25–0.30), and then decreased to 130 pC/N as x further increased to 0.35. In addition, the observed d₃₃ increased rapidly from 40 pC/N to 165 pC/N as y increased, and then slightly increased and fluctuated in the vicinity of 180 pC/N for y = −0.05–0.15 [Fig. 7(b)]. This variation of d₃₃ with different y contents are similar to those of BNT or KNN-based ceramics with the addition of Bi₂O₃.^28–30 As discussed above, the excessive deficiency or excess of Bi (actual y content: y = −0.02 or 0.22) can lead to the formation of a large number of V_Bi³⁻ or V_Ba²⁻. It was thought that the pinning of domain walls resulted in the degraded piezoelectricity of a material.⁴¹ As a result, the pinning of domain walls caused by defects, the pseudocubic phase, and loose microstructure are totally responsible for the serious reduction of d₃₃ in the ceramics with y = −0.1 (actual y content: y = −0.02) and 0.25 (actual y content: y=0.22). In addition, the R–C phase boundary and dense microstructure can totally contribute to the enhancement of piezoelectricity in BFO–BTO ceramics with −0.05 ≤ y ≤ 0.15 (x = 0.30).

Fig. 7 d₃₃, ε_r and tan δ as a function of (a) x (y = 0.05) and (b) y (x = 0.30) contents.

Fig. 8(a) and (b) exhibit the variations of d₃₃ with annealing temperatures (T_a) of the ceramics. The conventional thermal depolarization theory illustrated that the switching back of the domains to the initial state leads to the loss of piezoelectricity with the increase of T_a.⁴² For all ceramics except for y = −0.1 (actual y content: y = −0.02), d₃₃ has a slight increase and can maintain a relatively high value with the increase of T_a and then drops greatly with the further increase of T_a. For example, the ceramics with x = 0.30 have a much better thermal stability for T_a = 400 °C, while the ceramics with x = 0.25, 0.33, and 0.35 respectively show a poor thermal stability of d₃₃ for a lower T_a of 300 °C, 350 °C, and 300 °C [Fig. 8(a)]. In addition, all ceramics except y = −0.1 (actual y content: y = −0.02) have a similar changed trend with the increase of T_a, whose T_a is 400 °C. Especially, the ceramics with y = 0.05 have a much larger d₃₃ value (∼150 pC/N) as compared with other samples even if T_a increases to 450 °C, indicating that doping with optimum Bi can enhance the thermal stability of d₃₃ for BFO–BTO ceramics. An enhanced thermal stability was also reported in lead free high temperature BF–BT ceramics with excess Bi content.⁴³ As a result, large piezoelectricity (d₃₃ ∼ 180 pC/N) and its good thermal stability can be simultaneously achieved in the ceramics with appropriate excess Bi content.

Fig. 8 Thermal stability of (1 − x)Bi_1+yFeO_3+3y/2 − xBaTiO₃ ceramics with different (a) x (y = 0.05) and (b) y (x = 0.30) contents.

4. Conclusion

The phase structure, microstructure, and electrical properties of BFO were sensitive to Bi nonstoichiometry and the addition of BaTiO₃. An obvious refinement of grain sizes, serious diffused ε_r–T curves, and degraded electrical properties appeared in the ceramics with severe Bi deficiency. However, the improvement of piezoelectricity (d₃₃ ∼ 180 pC/N), ferroelectricity (P_r ∼ 25 μC cm⁻²), Curie temperature (T_C ≥ 500 °C), and thermal stability can be observed in the ceramics with excess Bi because of the existence of phase boundary and dense microstructure as well as the increase of electrical resistivity. As a result, it is necessary to improve the electrical properties of BFO ceramics by introducing both Bi nonstoichiometry and BaTiO₃.

Acknowledgements

Authors gratefully acknowledge the supports of the National Science Foundation of China (NSFC No. 51102173 and 51472169). We also thank Ms. Hui Wang for measuring the SEM patterns.

References

1. M. Mahesh Kumar, A. Srinivas and S. V. Suryanarayana, Structure property relations in BiFeO₃/BaTiO₃ solid solutions, J. Appl. Phys., 2000, 87, 855.
2. S. O. Leontsevw and R. E. Eitel, Dielectric and Piezoelectric Properties in Mn-Modified (1-x)BiFeO₃-xBaTiO₃ Ceramics, J. Am. Ceram. Soc., 2009, 92(12), 2957–2961.
3. H. B. Yang, C. R. Zhou, X. Y. Liu, Q. Zhou, G. H. Chen, W. Z. Li and H. Wang, Piezoelectric properties and temperature stabilities of Mn- and Cu-modified BiFeO₃-BaTiO₃ high temperature ceramics, J. Eur. Ceram. Soc., 2013, 33, 1177–1183.
4. Y. Wei, X. T. Wang, J. T. Zhu, X. L. Wang and J. J. Jia, Dielectric, Ferroelectric, and Piezoelectric Properties of BiFeO₃-BaTiO₃ Ceramics, J. Am. Ceram. Soc., 2013, 96(10), 3163–3168.
5. Z. Y. Cen, C. R. Zhou, H. B. Yang, Q. Zhou, W. Z. Li and C. L. Yuan, Structural, ferroelectric and piezoelectric properties of Mn-modified BiFeO3-BaTiO3 high-temperature ceramics, J. Mater. Sci.: Mater. Electron., 2013, 24, 3952–3957.
6. Y. X. Wei, X. T. Wang, J. J. Jia and X. L. Wang, Multiferroic and piezoelectric properties of 0.65BiFeO₃-0.35BaTiO₃ ceramic with pseudo-cubic symmetry, Ceram. Int., 2012, 38, 3499–3502.
7. H. B. Yang, C. R. Zhou, X. Y. Liu, Q. Zhou, G. H. Chen, H. Wang and W. Z. Li, Structural, microstructural and electrical properties of BiFeO₃-BaTiO₃ ceramics with high thermal stability, Mater. Res. Bull., 2012, 47, 4233–4239.
8. T. Rojac, M. Kosec, B. Budic, N. Setter and D. Damjanovic, Strong ferroelectric domain-wall pinning in BiFeO₃ ceramics, J. Appl. Phys., 2010, 108, 074107.
9. S. T. Zhang, M. H. Lu, D. Wu, Y. F. Chen and N. B. Ming, Larger polarization and weak ferromagnetism in quenched BiFeO₃ ceramics with a distorted rhombohedral crystal structure, Appl. Phys. Lett., 2005, 87, 262907.
10. J. Walker, P. Bryant, V. Kurusingal, C. Sorrell, D. Kuscer, G. Drazic, A. Bencan, V. Nagarajan and T. Rojac, Synthesis-phase-composition relationship and high electric-field-induced electromechanical behavior of samarium-modified BiFeO₃ ceramics, Acta Mater., 2015, 83, 149–159.
11. S. X. Zhang, L. Wang, Y. Chen, D. L. Wang, Y. B. Yao and Y. W. Ma, Observation of room temperature saturated ferroelectric polarization in Dy substituted BiFeO₃ ceramics, J. Appl. Phys., 2012, 111, 074105.
12. X. X. Shi, X. Q. Liu and X. M. Chen, Structure evolution and piezoelectric properties across the morphotropic phase boundary of Sm-substituted BiFeO₃ ceramics, J. Appl. Phys., 2016, 119, 064104.
13. G. L. Yuan, K. Z. Baba-Kishi, J. M. Liu, S. W. Or, Y. P. Wang and Z. G. Liu, Multiferroic Properties of Single-Phase Bi_0.85La_0.15FeO₃ Lead-Free Ceramics, J. Am. Ceram. Soc., 2006, 89(10), 3136–3139.
14. T. Zheng and J. Wu, Quenched bismuch ferrite-barium titanate lead -free piezoelectric ceramics, J. Alloys Compd., 2016, 676, 505–512.
15. C. R. Zhou, H. B. Yang, Q. Zhou, Z. Y. Cen, W. Z. Li, C. L. Yuan and H. Wang, Dielectric, ferroelectric and piezoelectric properties of La-substituted BiFeO₃-BaTiO₃ ceramics, Ceram. Int., 2013, 39, 4307–4311.
16. Q. J. Zheng, L. L. Luo, K. H. Lam, N. Jiang, Y. Q. Guo and D. M. Lin, Enhanced ferroelectricity, piezoelectricity, and ferromagnetism in Nd-modified BiFeO₃-BaTiO₃ lead-free ceramics, J. Appl. Phys., 2014, 116, 184101.
17. M. J. Tian, L. Zhou, X. Zou, Q. J. Zheng, L. L. Luo, N. Jiang and D. M. Lin, Improved ferroelectricity and ferromagnetism of Eu-modified BiFeO3-BaTiO3 lead-free multiferroic ceramics, J. Mater. Sci.: Mater. Electron., 2015, 26, 8840–8847.
18. Z. Y. Cen, C. R. Zhou, H. B. Yang, Q. Zhou, W. Z. Li, C. L. Yan, L. Cao, J. Song and L. Peng, Remarkably High-Temperature Stability of Bi(Fe_1-xAl_x)O₃-BaTiO₃ Solid Solution with Near-Zero Temperature Coefficient of Piezoelectric Properties, J. Am. Ceram. Soc., 2013, 96(7), 2252–2256.
19. Y. Wan, Y. Li, Q. Li, W. Zhou, Q. J. Zheng, X. C. Wu, C. G. Xu, B. P. Zhu and D. M. Lin, Microstructure, Ferroelectric, Piezoelectric, and Ferromagnetic Properties of Sc-Modified BiFeO₃-BaTiO₃ Multiferroic Ceramics with MnO₂ Addition, J. Am. Ceram. Soc., 2014, 97(6), 1809–1818.
20. C. R. Zhou, Z. Y. Cen, H. B. Yang, Q. Zhou, W. Z. Li, C. L. Yuan and H. Wang, Structure, electrical properties of Bi(Fe,Co)O₃-BaTiO₃ piezoelectric ceramics with improved Curie temperature, Phys. B, 2013, 410, 13–16.
21. X. H. Liu, Z. Xu, X. Y. Wei and X. Yao, Ferroelectric and Ferromagnetic Properties of 0.7BiFe_1-xCr_xO₃-0.3BaTiO₃ Solid Solutions, J. Am. Ceram. Soc., 2008, 91(11), 3731–3734.
22. Y. Li, N. Jiang, K. H. Lam, Y. Q. Guo, Q. J. Zheng, Q. Li, W. Zhou, Y. Wan and D. M. Lin, Structure, Ferroelectric, Piezoelectric, and Ferromagnetic Properties of BiFeO₃-BaTiO₃-Bi_0.5Na_0.5TiO₃ Lead-Free Multiferroic Ceramics, J. Am. Ceram. Soc., 2014, 97(11), 3602–3608.
23. D. M. Lin, Q. J. Zheng, Y. Li, Y. Wan, Q. Li and W. Zhou, Microstructure, ferroelectric and piezoelectric properties of Bi_0.5K_0.5TiO₃-modified BiFeO₃-BaTiO₃ lead-free ceramics with high Curie temperature, J. Eur. Ceram. Soc., 2013, 33, 3023–3036.
24. Y. Lin, L. L. Zhang and J. Yu, Stable piezoelectric property of modified BiFeO3-BaTiO3 lead-free piezoceramics, J. Mater. Sci.: Mater. Electron., 2015, 26, 8432–8441.
25. Q. Zhou, C. R. Zhou, H. B. Yang, G. H. Chen, W. Z. Li and H. Wang, Dielectric, Ferroelectric, and Piezoelectric Properties of Bi(Ni_1/2Ti_1/2)O₃-Modified BiFeO₃-BaTiO₃ Ceramics with High Curie Temperature, J. Am. Ceram. Soc., 2012, 95(12), 3889–3893.
26. L. F. Zhu, B. P. Zhang, S. Li, L. Zhao, N. Wang and X. C. Shi, Enhanced piezoelectric properties of Bi(Mg_1/2Ti_1/2)O₃ modified BiFeO₃-BaTiO₃ ceramics near the morphotropic phase boundary, J. Alloys Compd., 2016, 664, 602–608.
27. M. H. Lee, D. J. Kim, J. S. Park, S. W. Kim, T. K. Song, M. H. Kim, W. J. Kim, D. Do and I. K. Jeong, High-performance lead-free piezoceramics with high curie temperatures, Adv. Mater., 2015, 27, 6976–6982.
28. R. Z. Zuo, S. Su, Y. Wu, J. Fu, M. Wang and L. T. Li, Influence of A-site nonstoichiometry on sintering, microstructure and electrical properties of (Bi_0.5Na_0.5)TiO₃ ceramics, Mater. Chem. Phys., 2008, 110, 311–315.
29. Y. S. Sung, J. M. Kim, J. H. Cho, T. K. Song, M. H. Kim and T. G. Park, Effects of Bi nonstoichiometry in (Bi_0.5+xNa)TiO₃ ceramics, Appl. Phys. Lett., 2011, 98, 012902.
30. H. L. Du, D. J. Liu, F. S. Tang, D. M. Zhu, W. C. Zhou and S. B. Qu, Microstructure, Piezoelectric, and Ferroelectric Properties of Bi₂O₃-Added (K_0.5Na_0.5)NbO₃ Lead-Free Ceramics, J. Am. Ceram. Soc., 2007, 90(9), 2824–2829.
31. C. R. Zhou, H. B. Yang, Q. Zhou, G. H. Chen, W. Z. Li and H. Wang, Effects of Bi excess on the structure and electrical properties of high-temperature BiFeO3-BaTiO3 piezoelectric ceramics, J. Mater. Sci.: Mater. Electron., 2013, 24, 1685–1689.
32. W. J. Wu, J. Li, D. Q. Xiao, M. Chen, Y. C. Ding and C. Q. Liu, Defect dipoles-driven ferroelectric behavior in potassium sodium niobate ceramics, Ceram. Int., 2014, 40, 13205–13210.
33. Z. Ai, W. Ho, S. Lee and L. Zhang, Efficient photocatalytic removal of NO in indoor air with hierarchical bismuth oxybromide nanoplate microspheres under visible light, Environ. Sci. Technol., 2009, 43(11), 4143–4150.
34. G. Tan, W. Liu, X. Xue and H. Hao, Relative thickness effects on multiferroic Bi_0.9Gd_0.1Fe_0.96Mn_0.04O₃/NiFe₂O₄ bilayered thin films, J. Alloys Compd., 2014, 617, 265–270.
35. Y. Guo, P. Xiao, R. Wen, Y. Wan, Q. Zheng, D. Shi, K. Lam, M. Liu and D. Lin, Critical roles of Mn-ions in enhancing the insulation, piezoelectricity and multiferroicity of BiFeO₃-based lead-free high temperature ceramics, J. Mater. Chem. C, 2015, 3(22), 5811–5824.
36. W. Hu, Y. Liu, R. L. Withers, T. J. Frankcombe, L. Norén, A. Snashall, M. Kitchin, P. Smith, B. Gong, H. Chen, J. Schiemer, F. Brink and J. W. Leung, Electron-pinned defect-dipoles for high-performance colossal permittivity materials, Nat. Mater., 2013, 12(9), 821–826.
37. J. Lv, X. J. Lou and J. Wu, Defect Dipoles-induced Poling Characteristics and Ferroelectricity of Quenched Bismuth Ferrite-based Ceramics, J. Mater. Chem. C, 2016, 4.
38. Y. Li, M. Cao, D. Wang and J. Yuan, High-efficiency and dynamic stable electromagnetic wave attenuation for La doped bismuth ferrite at elevated temperature and gigahertz frequency, RSC Adv., 2015, 5(94), 77184–77191.
39. S. C. Yang, A. Kumar, V. Petkov and S. Priya, Room-temperature magnetoelectric coupling in single-phase BaTiO₃-BiFeO₃ system, J. Appl. Phys., 2013, 113, 144101.
40. H. H. Zhang, W. Jo, K. Wang and K. G. Webber, Compositional dependence of dielectric and ferroelectric properties in BiFeO₃-BaTiO₃ solid solutions, Ceram. Int., 2014, 40, 4759–4765.
41. C. A. Randall, N. Kim, J. P. Kucera, W. W. Cao and T. R. Shrout, Intrinsic and Extrinsic Size Effects in Fine-Grained Morphotropic-Phase-Boundary Lead Zirconate Titanate Ceramics, J. Am. Ceram. Soc., 1998, 81(3), 677–788.
42. Y. Hiruma, H. Nagata and T. Takenaka, Thermal depoling process and piezoelectric properties of bismuth sodium titanate ceramics, J. Appl. Phys., 2009, 105, 084112.
43. J. G. Chen and J. R. Cheng, Enhanced thermal stability of lead-free high temperature 0.75BiFeO₃-0.25BaTiO₃ ceramics with excess Bi content, J. Alloys Compd., 2014, 589, 115–119.

---