Poling procedures and piezoelectric response of (Ba0.85Ca0.15Zr0.1T0.9)O3 ceramics

ABSTRACT The goal of the current report was to find the optimal poling conditions: poling field, poling time and poling temperature for lead-free (Ba0.85Ca0.15Zr0.1T0.9)O3 ceramic. It has been noticed that the low poling field (< 2 kV/mm) was insufficient for obtaining ideal poling state (θ = 90°) due to the incomplete switching of domains and thus low piezoelectric properties (d33 ≤ 438 pC/N and kp ≤ 49%). However, relevance of higher poling voltage (> 2 kV/mm) may induce electric breakdown and physical defects such as cracks in the sample, which also leads degradation of piezoelectric properties. The optimal poling field, poling time and poling temperature was found to be 2 kV/mm, 10 min and 24°C, i.e., near room temperature, respectively to achieve nearly ideal poling state (θ = 86°) obtained by impedance spectra, and thus enhanced piezoelectric constant d33 = 505 pC/N and kp = 56% at room temperature. The O-T phase transition point was identified at 32.6°C, for which a peak value of d33(E=0V) = 622 pm/V was realized.


Introduction
In recent years, lead-free piezoelectric materials are receiving a considerable attention for the alteration of lead zirconate titanate (PZT)-based materials. In 2009, Liu and Ren [1] discovered (Ba 0.85 Ca 0.15 Zr 0.1 T 0.9 )O 3 (BCZT) system, which show similar piezoelectric properties than that of PZT. BCZT ceramic exhibits a temperature-dependent morhotropic phase boundary (MPB) and show very high piezoelectric constant (d 33 > 600 pC/N). However, such high piezoelectric constant is not easy to achieve and can be influenced by different processing techniques [2], sintering conditions [3,4] and poling conditions [5][6][7][8]. For polycrystalline ceramics, the domain orientations are randomly distributed, which gives a zero net polarization. Therefore, electrical poling is required by applying a DC field for certain amount of time and temperature that can align polar axis toward the direction of an applied field resulting net piezoelectric response. However, the poling conditions for BCZT ceramic are scattered in the literature [5,6,9,10] and is worth to compile a detailed report particularly focused on it. Therefore, the prime focus of the current research is to optimize the poling conditions such as poling field, poling time, and poling temperature to achieve enhanced piezo-response. Poling studies carried out by Su et al. [5]and Wu et al. [9] have reported high piezoelectric properties for BCZT ceramics when poled at very high poling field of 4 kV/mm and poling temperature of 40°C. Li et al. [10] discovered optimal piezoelectric properties when BCZT ceramic was poled at 120° C, which is higher than the Curie temperature (T C ). On the other hand, Praveen et al. [6] showed that a poling field of 3 times coercive field (E C ), which is approximately as low as 0.4 kV/mm to obtain d 33 > 600 pC/N. Therefore, in this article, we compiled a detailed report on the influence of wide range of poling field, poling time, and poling temperatures on the piezoelectric properties of BCZT ceramic using a standard DC poling in silicon oil. The effect of lower and excessive poling field on piezoelectric and ferroelectric properties is discussed and the optimized poling conditions for BCZT ceramic has been identified.

Phase characterization
The room temperature XRD patterns (2θ = 20°-70°) were obtained for BCZT ceramics between 0.4 and 3.6 kV/mm poling fields for 10 min as illustrated in Figure 1 (a). BCZT ceramics formed a pure perovskite phase and no trace of impurity or secondary phase was detected, which directs that the equilibrium of the ferroelectric phase was reached in BCZT ceramics. The peak splitting of (022)/(200) between 2θ = 44.4-45.6° (Figure 1(b)) and the formation of triplet (004)/(040)/ (222) between 65.2-66.4 (Figure 1(c)) suggests the presence of the orthorhombic (O) phase. Therefore, the BCZT ceramic shows O-phase at room temperature, regardless of applied poling field, which clearly indicates that the poling field do not affect phase structure of BCZT ceramic. Similar pure O-phase are previously reported by Fu et al. [11] for (Ba 0.89 Ca 0.11 Zr 0.1 T 0.9 )O 3 , Bijalwan et al. [12] and Keeble et al. [13] for (Ba 0.85 Ca 0.15 Zr 0.1 T 0.9 )O 3 ceramics at room temperature.

Optimization of poling conditions: piezoelectric and ferroelectric properties
Poling is necessary step for any polycrystalline piezoelectric ceramics to get their piezo-response. For this, electrical poling is required by which an external electric filed is supplied and randomly oriented ferroelectric domain aligned or becomes polarized with a non-zero polarization developed along applied field direction. Poling conditions such as poling voltage, poling temperature and sufficient poling time is required to get optimal piezoelectric response of BCZT ceramics [6]. Figure 2(a) depicts impedance spectra of BCZT ceramics at different poling voltage (ranges from 0.4 to 3.6 kV/mm) at room temperature for 10 minutes. Upon poling, the phase angle (θ) first increases from 79.2° (poling field, 0.4 kV/mm) to 86° (poling field, 2 kV/mm) and then decreases to 83.4° with further increase in the poling field. Beyond 2.4 kV/mm, θ values stabilized between 82.1° and 82.7°. The highest θ = 86° is realized for BCZT sample poled at 2 kV/mm for 10 min, which indicates that the sample is approaching toward ideal poling state (θ = 90°). The low d 33 values at poling field (< 2 kV/mm) can be ascribed to the incomplete switching of domains. Relevance of higher poling voltage (> 2 kV/mm) seemed also inefficient to reach the ideal poling state; however, it may induce electric breakdown and physical defects such as cracks in the sample leading to the reduction in the electrical properties [14]. The comparison of various poling conditions and corresponding piezoelectric response is shown in Table 1.
In response of approaching ideal poling state (θ = 86°), the piezoelectric constant d 33 and planar coupling coefficient k p reached the peak value of 505 pC/N and 56%, respectively, for the sample poled at 2 kV/mm for 10 min as shown in Figure 2(b). There was a clear correlation found between degree of poling (phase angle values) and piezoelectric properties. The poling voltage for BCZT ceramic in this study was found to be lower compared to the PZT-5 H (poling  [7]. Tao et al. [7] found optimal poling voltage for BCZT ceramic as 1.6 kV/mm, while d 33 was found as low as 381 pC/N. Electric field dependent polarization curves (P-E loops) were measured at 1 Hz as shown in Figure  3(a), which represents ferroelectric behavior of BCZT ceramics. The remnant polarization (P r ) values first increases from 10.74 µC/cm 2 to a maximum of 12.5 µC/cm 2 for the sample poled at 2.4 kV/mm for 10 min (24°C). However, similar P r value, 12.43 µC/cm 2 was also attained for the sample poled at 2 kV/mm as depicted in Figure 3(b). Further increase in poling voltage led to the saturation in the P r values (12.18-12.38 µC/cm 2 ). The coercive field (E C ) was found below 0.2 kV/mm for BCZT ceramics at different poling fields, which indicate the soft behavior of BCZT ceramics.
The optimized poling voltage for BCZT samples is further optimized with holding time during poling. For this, BCZT sample (poling voltage, 2kV/mm) poling time was prolonged i.e. 20 min and 30 min as represented in Figure 4(a). When the poling time is increased from 10 to 30 min, the phase angle value reduced from 86° (poling time, 10 min) to 83.4° (poling time, 30 min). As a result, the d 33 value decreased from 505 pC/N to 480 pC/N and similarly, k p values reduced from 56% to 50%. The decrease in k p values can also be confirmed by decrease in the anti-resonance frequency (f r ) as depicted in Figure 4(b). Therefore, large poling time was not found to be efficient so as to enhance piezoelectric properties of BCZT ceramics. Figure 5    maximum d 33 and k p values were kept for the samples poled at room temperature i.e. 24°C as represented in Figure 5(b). When poling temperature further increased, the d 33 values first decreases and then stabilized in the range of 426-460 pC/N. It is evident from the above results that the highest d 33 and k p values were obtained for the BCZT sample poled at 24°C (poling field, 2 kV/mm and poling time, 10 min). The comparison of different poling conditions with piezoelectric properties are summarized in Table 2.

Temperature-dependent dielectric, piezoelectric and ferroelectric curves
The temperature dependence of relative permittivity and d 33(E=0) of unpoled BCZT ceramic is depicted in Figure 6(a,b). Two dielectric peaks near room temperature and around 100°C were observed corresponding to the O -T and T -C phase transitions, respectively ( Figure 6(a)). Similar two distinct peaks were also observed for d 33(E=0) vs. T plot, which agree well with   the dielectric measurements. The piezoelectric coefficient was evaluated at zero bias voltage d 33(E=0) on unpoled samples and therefore this could be the reason in obtaining low d 33(E=0) = 444 pm/V at 24°C as compared to the d 33 value attained by quasi-static method. If the samples are poled, then similar value of d 33 can be obtained by indirect method [15,16]. The d 33(E=0) values rapidly increased when it approached toward O -T phase transition temperature (T O-T = 32.6° C), and attained maximum value of 622 pm/V. This higher values of d 33(E=0) is due to the domain volume fractions of the two phases. Such a phenomenon suggests that the higher piezoelectric response of BCZT ceramic can be obtained at T O-T than in single ferroelectric phase. This behavior of d 33 against temperature is reflected to be a macroscopic illustration of the evolution of temperature-dependent spontaneous polarization [17]. Peaks at phase transition temperatures are strongly ascribed to the lattice softening and structural relaxation, that guide to an enlarged mobility of domains and rare sensitivity to the external stress or electric filed. This is also termed as "constrained negative stiffness effect" in the literature [17], which may contribute to the peaks near phase transitions and thus improved piezoelectricity. Theoretical analysis has shown that the high piezoelectric response can be induced by the constrained negative stiffness because of the coupling between electric and strain field [18]. Negative stiffness can be induced in BaTiO 3 -based ceramics the following way; it is relevant at the macroscopic scale and is anticipated in the context of Landau theory of ferroelastic transformations: when temperature is reduced down from a value above the transition temperature, then spontaneous strain and temperature having single minimum slowly flattens and acquire two or more minima i.e. potential wells. The curvature of the spontaneous strain (energy function of strain) denote an elastic modulus. Therefore, the curve flattening signifies a softening of the modulus close to the critical temperature, while the reversed curvature depicts a negative stiffness at small strain.

Discussion
Optimized poling conditions, (poling field, poling time, and poling temperature) of a piezoelectric material is attributed to the alignment of the ferroelectric as well as ferroelectric domains upon an applied electric field, which is essential to enhance the piezoelectric properties of the BCZT ceramic. It is believed that a suitable poling temperature may also assist accumulation of charge carriers at the grain boundaries that can heighten the piezo-response of the samples [5,10] and it was found to be near RT, i.e. 24°C in the current study. The low poling field makes the switching inadequate as in the case between 0.4 and 1.6 kV/mm poling field. However, excessive poling field and poling time tends to over-pole the sample, which may initiate physical defects and ultimately to the dielectric breakdown of the BCZT sample [14]. This could explain the decrease in the d 33 values when sample poled over poling field > 2 kV/mm (Figure 2(b)) and excessive poling time > 10 min (Figure 4(a,b)) in the current study. Poling dependence study by Su et al. [5] and Praveen et al. [6] revealed that the optimal poling voltage for BCZT ceramics is approximately two or three times of the E C and poling temperature in the range of 30-40°C reaching d 33 values > 600 pC/N and they ascribed it to the multiphase coexistence. Wu et al. [9] reported optimal d 33 = 423 pC/N at poling voltage and poling temperature 4 kV/mm and 40°C, respectively. Our previous work also shows the d 33 value ~ 420 PC/N, when BCZT ceramic poled at 3 kV/mm for 10 min [19]. However, in the current study the BCZT ceramics show d 33 = 505 pC/N at 24°C (RT), when sample poled at 2 kV/mm for 10 min and d 33 = 622 pC/N (assumed from the value d 33(E=0) = 622 pm/V) at T O-T = 32.6°C. Keeping in mind that T O-T is not the poling temperature but phase transition point, which is the main reason for obtaining similar high d 33 values previously reported [1,6]. Change in poling field did not affect the phase structure of BCZT ceramic in the current study as confirmed by XRD spectra.

Conclusion
In summary, high piezoelectric properties (d 33 = 505 pC/N and k p = 56%) were attained at the optimized conditions i.e., 2 kV/mm for 10 min at 24°C i.e., room temperature. Also, piezoelectric constant d 33(E=0) = 622 pm/V was measured at 32.6°C, the point at which orthorhombic-tetragonal phase transition occurs. The crystal structure remains unaffected with the poling field and time. This work notably finds the appropriate poling conditions for BCZT ceramics to reach ideal poling state and thus heighten the piezoelectric response.