Improvement of electrical and energy harvesting properties of new lead-free BST modified 0.995BNKT–0.005LN ceramics

ABSTRACT

New lead-free (1-y)[0.995Bi_0.5(Na_0.80K_0.20)_0.5TiO₃-0.005LiNbO₃]-y[(Ba_0.7Sr_0.3)TiO₃] or (1-y)[0.995BNKT-0.005LN]-y[BST] (where y = 0, 0.01, 0.02, 0.03, 0.04 and 0.05 mol fraction) ceramics were fabricated by a conventional sintering method with two calcination steps. All ceramics sintered at 1125°C for 2 hours showed an optimum relative density of ~98% and a linear shrinkage of ~18%. X-ray diffraction patterns indicated that all ceramics possessed a pure perovskite phase with coexisting rhombohedral and tetragonal structures. Electrical and energy harvesting properties of 0.995BNKT–0.005LN ceramics were improved with 0.03 mol fraction BST addition (d₃₃ = 287 pC/N, d*₃₃  = 440 pm/V, g₃₃ = 13.40 × 10⁻³ V/mN and FoM = 3.85 pm²/N). From the observed results, the 0.97[0.995BNKT-0.005LN]-0.03[BST] ceramic is a promising candidate for lead-free piezoelelectric sensor-actuator and energy harvester.

KEYWORDS:

• Bismuth sodium potassium titanate
• strain behavior
• piezoelectric properties
• lead-free piezoelectric
• energy harvesting

1. Introduction

Piezoelectric materials are well-known smart materials that can convert mechanical energy to electrical energy and vice versa. With these characteristics, the piezoelectric materials have been used in a variety of technological applications such as sensor, actuator, transducer, filter, resonator, micro-electromechanical system (MEMS), electromechanical device, high energy storage capacitor, and so on [1–7]. Pb(Zr,Ti)O₃ or PZT is a commercial piezoelectric material that has been selected as a material for piezoelectric applications due to its high piezoelectric performance. However, the toxicity of lead compounds released to the environment during a production process has been a concern [8–11]. This problem has led to the development of new lead-free piezoelectric materials with useable piezoelectric properties.

Perovskite-type lead-free piezoelectric materials such as BaTiO₃ (BT), (Bi_0.5Na_0.5)TiO₃ (BNT), K_0.5Na_0.5NbO₃ (KNN)-based compounds have been observed to possess high piezoelectric coefficients. However, these kinds of piezoelectric materials have always suffered from some drawbacks, i.e. a low Curie temperature (T_c) and a difficulty in poling treatment [12]. Therefore, many efforts have been paid to solve these problems [13]. Among lead-free piezoelectric materials, a bismuth-based Bi_0.5Na_0.5TiO₃ (BNT) material is known to be a potential candidate for applications due to its high T_c (~ 320°C) and large remnant polarization (P_r ~ 32 µC/cm²). However, it has limited use due to high electrical conductivity, a large coercive field (E_c), and a lower normalized strain coefficient [14–16]. Previous research revealed that dielectric, ferroelectric, and piezoelectric properties of BNT-based materials were enhanced by the addition of Bi_0.5K_0.5TiO₃ (BKT). This was due to a formation of a morphotropic phase boundary (MPB) at BKT content of 16 to 20% [13,16–18]. Moreover, Zhou et al. have synthesized a complex (1-x-y)Bi_0.5Na_0.5TiO₃-xBi_0.5K_0.5TiO₃-yLiNbO₃ ternary system to improve electrical properties of BNKT-based materials. They found that the optimum piezoelectric constant (d₃₃) and electromechanical coupling factor (k_p) of 195 pC/N and 0.336, respectively, were observed at x = 0.18 and y = 0.01 [19]. It was also found in our previous study that electrical properties of BNKT ceramics were significantly modified by the addition of a small amount of LN (0.005 mol fraction [20].

(Ba_0.7Sr_0.3)TiO₃ (BST) is an insulating material with a large relative dielectric permittivity and small dielectric loss near ambient temperature, which is acceptable for a wide range of electrical applications [21–23]. Previous research has shown improved electrical properties of BNKT-based materials by BST modification. For example, room temperature dielectric constant, ε_r, and d₃₃, bipolar field-induced strain, dynamic piezoelectric coefficient, d*₃₃, of BNKT ceramics were improved by BST addition [24,25].

Based on previous studies mentioned above, in this study, a new lead-free (1-y)[0.995Bi_0.5(Na_0.80K_0.20)_0.5TiO₃-0.005LiNbO₃]-y[(Ba_0.7Sr_0.3)TiO₃] or (1-y)[0.995BNKT-0.005LN]-y[BST] (where y = 0, 0.01, 0.02, 0.03, 0.04 and 0.05 mol fraction) ternary system ceramics were fabricated by a conventional mixed oxide technique using two calcination steps. The effects of BST modification on phase, microstructure, dielectric, ferroelectric, and piezoelectric properties, and strain behavior of 0.995BNKT–0.005LN ceramics were investigated and discussed in detail.

2. Experimental procedures

Lead-free piezoelectric ceramics with the composition of (1-y)[0.995 Bi_0.5(Na_0.80K_0.20)_0.5TiO₃-0.005LiNbO₃]-y(Ba_0.7Sr_0.3)TiO₃ or (1-y)[0.995BNKT-0.005LN]-y[BST] (where y = 0, 0.01, 0.02, 0.03, 0.04 and 0.05 mol fraction) were fabricated by a conventional mixed oxide technique using two calcination steps. Bi₂O₃ (purity 98%, Fluka), Na₂CO₃ (purity 99.5%, Carlo Erba), K₂CO₃ (purity 99%, Sigma-Aldrich), TiO₂ (98.5%, Sigma-Aldrich), Li₂CO₃ (purity 99.0%, Sigma-Aldrich), and Nb₂O₃ (purity 99.0%, Sigma-Aldrich) for the synthesis of 0.995BNKT–0.005LN. BaCO₃ (purity 98.5%, Fluka), SrCO₃ (purity 99.9%, Sigma-Aldrich) and TiO₂ (purity 98.5%, Sigma-Aldrich) were chosen as the reactant powder materials for (Ba_0.7Sr_0.3)TiO₃. Prior to the reaction process, in order to eliminate any moisture, all carbonate powders, i.e. Na₂CO₃, K₂CO₃, BaCO₃, Li₂CO₃, SrCO₃, were dried in an oven at 120°C for 24 hours. All the reactant powders were weighed according to the stoichiometric formula, wet mixed by ball milling methods for 24 hours in 99% ethanol solution and dried in an oven at 120°C for 24 hours. The dried powder was separately calcined at 900°C for 2 hours with 0.995BNKT–0.005LN and 1100°C for 2 hours for BST with a heating/cooling rate of 5°C/min. The calcined 0.995BNKT–0.005LN and BST powders were mixed following the formula of (1-y)[0.995BNKT-0.005LN]-y[BST] (where y = 0, 0.01, 0.02, 0.03, 0.04 and 0.05 mol fraction) by the ball milling method again for 24 hours and dried in an oven at 120°C for 24 hours. To obtain the (1-y)[0.995BNKT-0.005LN]-yBST ceramics, the polyvinyl alcohol (PVA) binder with a concentration of 4 wt% was added to the mixed powders, which were pressed under 150 MPa pressure into the metal disks having 10 mm in diameter and sintered in an alumina crucible under normal atmosphere at 1125°C for 2 hours at a heating-cooling rate of 5°C/min.

The theoretical densities of all ceramics were calculated based on the theoretical densities of BNKT (5.84 g/cm³) [26] and BT (6.01 g/cm³) [27]. The linear shrinkage percentage of all ceramics was calculated using the ratio of the sample diameter before and after the sintering process. The phase formation was studied by X-ray diffractometer (XRD-Phillip, X-pert). Raman spectra were analyzed using Raman spectroscopy (Raman Jobin Yvon Horiba, T64000). The microstructure of all ceramics was examined using scanning electron microscopy (SEM, JEOL JSM-6335 F). The grain size distribution was measured using an ImageJ program. The average grain size of all ceramics was estimated using a linear intercept method following the ASTM E112-88 procedure. All ceramics were polished to a thickness of 1 mm and the silver paste was coated on both circular surfaces of the sintered sample and heated at 400°C for 15 minutes to formed electrodes before electrical properties measurement. The dielectric constant (ε_r) and dielectric loss (tanδ) as a function of temperature (50–350°C) were measured using a high precision LCR meter (LCR 821, GW INSTEK) under frequencies of 1, 10 and 100 kHz. The ferroelectric properties at room temperature of all ceramics were analyzed using a Radiant Precision ferroelectric tester. The electric field of 30, 40, and 50 kV/cm under a frequency of 1 Hz was applied to all studied ceramics. The remanent polarization (P_r), maximum polarization (P_max), coercive field (E_c), and maximum coercive field (E_max) values were determined from the hysteresis loop curve. The strain-electric field (S-E) behavior was obtained by the optical displacement sensor (Fotonic Sensor model MTI-2100) at room temperatures under frequency of 0.1 Hz. The maximum strain (S_max) and the negative strain (S_neg) values were obtained from the bipolar curve at a maximum electric field (E_max) of 50 kV/cm. The normalized strain coefficient (d*₃₃) was also calculated using the following equation: d*₃₃ = S_max/E_max. Before piezoelectric measurement, all ceramics were poled in a silicone oil bath at 55°C under an applied DC electric field of 4 kV for 15 minutes. The piezoelectric coefficient (d₃₃) of all ceramics was obtained by d₃₃ meter (KCF technologies, S5865). The piezoelectric voltage constant (g₃₃) was calculated using the equation, g₃₃ = d₃₃/ε_r ε₀, where g₃₃ is the piezoelectric voltage constant, ε_r the dielectric constant of the piezoelectric material, ε_o the dielectric constant in vacuum, and d₃₃ the piezoelectric charge constant [28,29]. The off-resonance figure of merit (FoM) values for energy harvesting was calculated from the multiplication between piezoelectric charge constant (d₃₃) and the piezoelectric voltage constant (g₃₃) values according to the equation, FoM = d₃₃g₃₃ [30].

3. Results and discussion

Relative density values of all (1-y)[0.995BNKT-0.005LN]-yBST ceramics were found to be about 98% while those of linear shrinkage were about ~18%. These results illustrated that the BST-added ceramics having near theoretical densities were obtained with the fabrication technique used in this study. It can be seen from the results given in Table 1 that the relative density and linear shrinkage of samples gradually increased with increasing BST content.

Table 1. Physical and microstructure properties of the (1-y)[0.995BNKT-0.005LN]-yBST ceramics.

y	Relative density (%)	Linear shrinkage (%)	c/a	Unit cell volume (Å^3)	Average grain size (µm)
0	98.22 ± 0.03	18.01 ± 0.08	1.0118	60.29	0.63
0.01	98.29 ± 0.05	18.15 ± 0.06	1.0122	60.44	0.48
0.02	98.45 ± 0.07	18.28 ± 0.06	1.0125	60.46	0.47
0.03	98.52 ± 0.04	18.37 ± 0.04	1.0128	60.49	0.43
0.04	98.58 ± 0.05	18.44 ± 0.05	1.0130	60.51	0.42
0.05	98.62 ± 0.08	18.42 ± 0.08	1 .0136	60.57	0.40

The X-ray diffraction patterns of all (1-y)[0.995BNKT-0.005LN]-yBST samples with the 2θ = 20–60 degrees are shown in Figure 1(a). All XRD patterns are similar to the previous reports of the 0.995BNKT–0.005LN-based ceramics [20,31] where a pure perovskite structure with no secondary phases was obtained. This suggested that the BST diffused into the 0.995BNKT–0.005LN to form the final compound. The A-site position of Li⁺ ion had no confirmed reported ionic radius value but the expectation value should be between r_Li⁺ (0.92 Å, CN = 8 [32] and r_Na⁺ (1.39 Å, CN = 12 [32])). Ba²⁺ (r_Ba²⁺ = 1.61 Å [32]) ions likely diffused into Bi³⁺ (r_Bi³⁺ = 1.40 Å [33]), Na⁺ (r_Na⁺ = 1.39 Å [32]), K⁺ (r_K⁺ = 1.64 Å [32]) and Sr²⁺ (r_Sr²⁺ = 1.44 Å [32]) while Nb⁵⁺ (r_Nb⁵⁺ = 0.64 Å [32]) could enter into B-site (CN = 6) of Ti⁴⁺ position (r_Ti⁴⁺ = 0.61 Å [32]). From the narrow angular ranges of 2θ = 39°- 41° (See Figure 1(b)) and 2θ = 45°- 48° (See Figure 1(c)), the y = 0 sample represented mixed rhombohedral and tetragonal structure with the overlapping of (111)_R/(111)_T/($1\bar{1}1$)_R peaks. The tetragonal phase seen near 2θ ~ 46 degree from the splitting of (002)_T/(200)_T peaks was similar to the results previously reported [16,24,25,34]. With increasing BST content, we observed a phase change from mixed rhombohedral and tetragonal to a tetragonal dominant phase. The tetragonality (c/a) value confirmed the appearance of phase transformation (see Table 1). The c/a values increased from 1.0118 (y = 0 sample) to 1.0136 (y = 0.05 sample). Moreover, the increase of BST concentration caused the diffraction peaks to slightly shift toward a lower angle, particularly those associated with the ab-plane of tetragonal phase. This was attributed to the substitutional effects between Bi³⁺, Na⁺, K⁺, Li⁺, Ba²⁺, and Sr²⁺ at position of A-site and Ti⁴⁺ and Nb⁵⁺ at position of B-site. This change induced a structural distortion i.e. enlargement of unit cell size as listed in Table 1. A similar peak shift was observed in previous reports of BNKT-based compounds [24,25,31,34].

Figure 1. X-ray diffraction patterns of the (1-y)[0.995BNKT-0.005LN]-y[BST] ceramics where (a) 2θ = 20–60°, (b) 2θ = 39–41° and (c) 2θ = 45–48°.

Raman spectroscopy is a technique used for investigating the vibrational modes in oxide materials, and probing the stretching and bending modes as well as the local structure in perovskite materials [35] and the large length-scale average structures [36]. Furthermore, this technique is used to understand the phase transformation of the ceramics. Figure 2 shows the Raman spectra measured at room temperature in a wavenumber range of 100–1200 cm⁻¹ for all samples. The observed Raman spectra were similar to the results previously reported in references [16,24,37]. The Raman modes below ~ 150 cm⁻¹ were related to the A-site vibrations. The Raman modes between 200 and 400 cm⁻¹ indicated the Ti-O vibration and those of 500–650 cm⁻¹ were due to the octahedral Ti-O₆ vibrations [38–41]. The high-frequency region above 700 cm⁻¹ specified the A₁ (longitudinal optical) and E (longitudinal optical) overlapping bands [42,43]. In this study, the Raman peak around 100–200 cm⁻¹ wavenumbers could be associated with the vibration of A-site cations such as Bi (Bi-O bond), Na (Na-O bond), and K (K-O bond) in the ABO₃ perovskite structure. The Raman peak centered at ~ 272 cm⁻¹ was attributed to the vibrations of Ti-O. The peak became broader and started to split into two bands with slight shift when the BST concentration increased. This was due to the co-occupancy of different cationic size at the position of B-site, suggesting a unit cell distortion. The two overlapping bands associated with the vibration of the TiO₆ octahedra in the range of 450–650 cm⁻¹ showed a separation into two distinct bands with the increase of BST concentration (y = 0.03–0.05 sample), indicating a particular phonon behavior in the structure evolution [42,43] and the phase change from mixed rhombohedral and tetragonal structures to be more tetragonal phase, in agreement with in the XRD result. The peak at the wavenumber above 700 cm⁻¹ suggested the overlapping of longitudinal optical of A₁ and E bands. This result also agreed with that observed by Hao et al [42] and Malik et al [44].

Figure 2. Raman spectra of the (1-y)[0.995BNKT-0.005LN]-y[BST] ceramics at room temperature.

Figure 3 depicts the SEM images and histograms of the grain size distribution for the as-sintered surfaces of all (1-y)[0.995BNKT-0.005LN]-yBST samples. The SEM data confirmed that all samples were dense with nearly no pores. Moreover, the grain boundaries of all samples indicated the cubic-like morphology. For y = 0 sample, the average grain size was approximately 0.63 µm. The grain size was found to decrease with increasing BST concentration with the minimum value of 0.40 µm observed for the y = 0.05 sample. The average grain size values of all ceramics are given in Table 1. As observed from the density, shrinkage, tetragonality, and unit cell volume results, the properties suddenly changed when 0.01 mol fraction of BST was added. But the slight change in the properties was found with a BST content above 0.01 mol fraction. Due to the difference in diffusibility of solvent and solute ions in this system as well as the possibility of compositional variation near the grain boundaries, the mass transfer was then inhibited during the sintering process. Consequently, the grain growth process of the ceramics was restricted [16,45,46].

Figure 3. SEM micrograph and histograms of the grain size distribution of the (1-y)[0.995BNKT- 0.005LN]-y[BST] ceramics.

Figure 4(a) shows temperature dependence of dielectric constant (ε_r) and dielectric loss (tanδ) measured under frequencies of 1, 10, and 100 kHz at temperature range 25–350°C. For all samples, ε_r values measured at lower frequency were greater as compared to those at higher frequency. This was due to a restriction of large-scale polarization such as space charge polarization which could not follow higher applied frequencies [47,48]. The maximum dielectric constant at high temperature (ε_m) of the unmodified 0.995BNKT-0.005LN ceramic was found to be 5570 with tan δ of 0.0187. This result corresponded to the previous studies reported on the BNKT-based system [16,34]. For the BST-added samples, the ε_m initially increased with BST content and reached a maximum value at y = 0.03 (ε_m = 5764, tan δ = 0.0155) and then dropped after y = 0.04. Furthermore, the dielectric characteristic of all samples exhibited two dielectric anomaly peaks. The first anomaly peak was observed at a lower temperature which is a depolarization temperature (T_d). T_d is determined to be a temperature at which a phase of a ceramic transforms from rhombohedral to tetragonal [24,48,49]. The second anomaly peak observed at higher temperature is a maximum dielectric constant temperature (T_m) where the dielectric constant reaches a maximum value [50,51]. All samples exhibited a slight broadening of both T_d and T_m peaks which could be attributed to compositional fluctuation of A-site and/or B-site lattices due to the occupation of different cations [52]. The 0.995BNKT–0.005LN ceramic studied in this work possessed T_d and T_m of 147°C and 255°C, respectively. The observed values were similar to those reported in previous studies of other BNKT-based materials [16,47]. The T_d of the ceramic decreased to 97°C for y = 0.01 composition. The T_d was nearly unchanged with further increasing y, as shown in Figure 4(b). The decrease in T_d could be resulted from the ferroelectric order destabilization and may be associated with the possible presence of nonpolar phase in BST-doped sample [53,54]. This behavior was in good agreement with the previous work for Zr-doped Bi_0.5(Na_0.78K_0.22)_0.5TiO₃ system [55]. The Curie point or T_m value was found to gradually increase with increasing BST content. This was believed to be due to the decreased grain size which resulted in the increased surface area of space charge layer and subsequently, both the space charge field and the locked-in ferroelectric polarization increased. This phenomenon was also observed in the study of the PLZT system [56].

Figure 4. (a) Temperature dependence of the dielectric constant (ε_r) and dielectric loss (tan δ) under various of frequencies from 1–100 kHz and (b) Plots of T_m and T_d as a function of BST concentration of the (1-y)[0.995BNKT-0.005LN]-y[BST] ceramics (inset show: Plots of ε_m and tan δ as a function of BST content).

The ferroelectric properties and polarization-electric field (P-E) curves at room temperature and at a frequency of 1 Hz under electric fields of 30, 40, and 50 kV/cm for all samples are shown in Figure 5. Plots of remanent polarization (P_r) and coercive field (E_c) values as a function of electric field are also shown. The P_r and E_c values increased with increasing electric field. All P-E curves were well-saturated with typical P-E hysteresis loops at the electric field of 50 kV/cm. The BTS addition significantly influenced the ferroelectric hysteresis loop shape and the polarization response of 0.995BNKT–0.005LN-based material. Hysteresis loop shape became more squared when y > 0.01 which indicated a strong ferroelectric interaction [57]. An unmodified-0.995BNKT-0.005LN ceramic had P_r = 23.10 μC/cm² and E_c = 18.28 kV/cm. The P_r tended to increase with increasing BST content and reached a maximum value of 32.92 μC/cm² at y = 0.03. However, with further increasing BST content (y = 0.04–0.05), the P_r value decreased. Likewise, the E_c value reached a maximum value of 26.85 kV/cm at y = 0.04 and the value slightly dropped to 25.98 kV/cm at y = 0.05 sample. The increasing tendency of P_r could be explained by the increased density of material due to BST addition. In case of the increased E_c value, this suggested that the material became harder, which was consistent with the assumption that oxygen vacancies were formed, and the so-called complex defects were produced by these acceptor dopants [57]. This finding was also in agreement with previous reports [24,54,58].

Figure 5. Polarization-electric field (P-E) hysteresis loop, and plots of P_r and Ec values as a function of electric field of the (1-y)[0.995BNKT- 0.005LN]-y[BST] ceramics, measured at RT under various electric fields of 30, 40, and 50 kV/cm and a frequency of 1 Hz.

In order to study the bipolar electric field-induced strain curve of all (1-y)[0.995BNKT-0.005LN]-yBST samples, strain-electric field (S-E) curves measured at room temperature were evaluated. The S-E curves measured at a frequency of 0.1 Hz with a maximum electric field of 50 kV/cm of the ceramics are shown in Figure 6. The typical S-E curve of this work, which was in agreement with BNKT-based ceramics of the butterfly-like curve [16,24,42,44], indicated the typical piezoelectric response [59]. The maximum strain (S_max), the negative strain (S_neg) and the normalized strain coefficient (d*₃₃ = S_max/E_max) are also summarized in Table 2. The variation of S_max with increasing BST content showed a similar trend to that of d*₃₃ values. The undoped sample had the S_max of 0.15% and calculated d*₃₃ of 278 pm/V, indicating that the ferroelectric domain switching was involved at E_c where the largest S_neg of −0.09% was observed. The S_max and d*₃₃ tended to increase with the higher concentration of BST as shown in Figure 7(a). The maximum value was found in the y = 0.03 sample (S_max = 0.22% and d*₃₃ = 440 pm/V). The S_max and d*₃₃ values were then dropped with further increasing BST content. The decreased S_max and d*₃₃ after y = 0.03 sample might be due to the dominating tetragonal phase and the increasing trend in tetragonality (c/a) which clearly indicated the increase of lattice anisotropy of modified-0.995BNKT-0.005BST samples (i.e. rhombohedral phase decreased while tetragonal phase increased) [60,61]. The increasing trend in lattice anisotropy which showed slight distortion of XRD patterns resulted in the enlargement of lattice constant and lattice energy which induced a phase transformation in order to stabilize the structure [62]. This suggested that ferroelectric order was over a nonpolar region. In addition, the decreasing trend of the nonpolar phase at the higher BST concentration likely suppressed the domain switching ability [63].

Table 2. Electrical properties of the (1-y)[0.995BNKT-0.005LN]-yBST ceramics.

y	Pr (µC/cm^2)	Ec (kV/cm)	Pmax (µC/cm^2)	Emax (kV/cm)	Smax (%)	Sneg (%)	d33* (pm/V)	d33 (pC/N)	g33 (10^−3V/mN)	FoM (pm^2/N)
0	23.10	18.28	32.60	50.40	0.15	−0.09	278	152	8.30	1.26
0.01	21.71	18.59	33.57	50.07	0.16	−0.08	313	189	9.87	1.87
0.02	25.81	23.80	32.60	50.08	0.18	−0.08	359	221	12.07	2.67
0.03	32.92	24.82	38.99	50.07	0.22	−0.07	440	287	13.40	3.85
0.04	28.79	27.81	30.55	50.40	0.10	−0.10	200	221	13.17	2.91
0.05	24.03	25.98	31.58	50.07	0.09	−0.10	180	203	11.40	2.31

Figure 6. Bipolar strain electric-field S-E loop of the (1-y)[0.995BNKT-0.005LN]-yBST ceramics.

Figure 7. (a) Plot of the S_max and d₃₃* as a function of BST content and (b) Plot of the g₃₃ and FOM values as a function of BST content of the (1-y)[0.995BNKT-0.005LN]-y[BST] ceramics.

The piezoelectric coefficient (d₃₃) values of all (1-y)[0.995BNKT-0.005LN]-yBST samples are summarized in Table 2. The 0.995BNKT-0.005LN sample had a d₃₃ value of 152 pC/N. The d₃₃ value of the modified 0.995BNKT-0.005LN samples increased with increasing BST content, and the composition of y = 0.03 showed the highest value of 287 pC/N. The energy-harvesting utilizing piezoelectric materials has recently attracted extensive attention due to the strong demand of self-powered electronics [64]. The important factor for evaluating piezoelectric energy harvester is the piezoelectric voltage coefficient (g₃₃) value [65]. The result from the calculation of the g₃₃ value are also listed in Table 2. The g₃₃ value in this work was found to be improved by the BST doping. The undoped-0.995BNKT-0.005LN sample had the g₃₃ value of 8.30 × 10⁻³ Vm/N and increased with increasing BST content. The maximum value of 13.40 × 10⁻³ Vm/N was found for the y = 0.03 sample. The d₃₃ value decreased with further increasing of the modifier content greater than y = 0.03. The off-resonance figure of merit (FoM) of all (1-y)[0.995BNKT-0.005LN]-yBST samples are listed in Table 2. The energy harvesting properties were also improved by the BST doping. The undoped-0.995BNKT-0.005LN sample had the FoM value of 1.26 pm²/N. The FoM value increased with the increasing BST content and showed the maximum value of 3.85 pm²/N for the y = 0.03 sample and then dropped with a further increase of the modifier content. It can be seen that the trend of the FoM value was similar to that of the trend of g₃₃ value [66] as shown in Figure 7(b). The comparison of the FoM values with the previously studied lead-free piezoelectric ceramics are shown in Figure 8. Based on the obtained data, the piezoelectric characteristics and energy harvesting behavior improvement was achieved in this system. The addition of BST into the 0.995BNKT-0.005LN system had potential to be one of the promising lead-free compounds for further use in piezoelectric devices.

Figure 8. Comparison of FoM values of (1-y)[0.995BNKT-0.005LN]-yBST ceramics with other lead-free ceramics.

4. Conclusions

The new lead-free (1-y)[0.995BNKT-0.005LN]-yBST system (where y = 0, 0.01, 0.02, 0.03, 0.04 and 0.05 mol fraction) was successfully fabricated by a solid-state mixed oxide method using two calcination steps. X-ray diffraction results showed that all ceramic samples exhibited a single-phase perovskite structure with no impurity phases detected. The decrease of grain size was observed when the BST dopant increased. For y = 0.03 sample, it showed the maximum dielectric properties (ε_m = 5764, tan δ = 0.0155, T_d = 100°C, T_m = 262°C), and ferroelectric parameters (P_r = 32.92 µV/cm², E_c = 24.82 kV/cm). The highest piezoelectric coefficient (d₃₃ = 287 pC/N), piezoelectric voltage constant (g₃₃ = 13.40 × 10⁻³ Vm/N), and the off-resonance figure of merit for energy harvesting (FoM = 3.85 pm²/N) were also obtained for this sample. This material system has the potential to be one of the promising new lead-free piezoelectric candidates for further use in sensor-actuator and energy harvesting applications [67,68]

Acknowledgments

This research was financially supported by Department of Physics and Materials Science, Faculty of Science, Chiang Mai University, Thailand Science Research and Innovation (TSRI), the National Research Council of Thailand (NRCT): NRCT5-RSA63004-15. Partial supports from the Center of Excellence in Materials Science and Technology, Faculty of Science, Chiang Mai University are also acknowledged. MP would like to thank Rubber Product and Innovation Development Research Unit (SCIRU63002).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Department of Physics and Materials Science, Faculty of Science, Chiang Mai University, Thailand Science Research and Innovation (TSRI), the National Research Council of Thailand (NRCT): <#award-id;temp>NRCT5-RSA63004-15. The Center of Excellence in Materials Science a [NRCT5-RSA63004-15]; Rubber Product and Innovation Development Research Unit (SCIRU63002) [SCIRU63002];Department of Physics and Materials Science, Faculty of Science, Chiang Mai University, Thailand Science Research and Innovation (TSRI), the National Research Council of Thailand (NRCT): NRCT5-RSA63004-15. The Center of Excellence in Materials Science a [NRCT5-RSA63004-15]; P. Jaiban would like to thank partial funding by National Science, Research and Innovation Fund (NSRF), and King Mongkut’s University of Technology North Bangkok with Contract no. KMUTNB-FF-65-43.