Piezoelectric Characteristics of 0.55Pb(Ni1/3Nb2/3)O3-0.45Pb(Zr,Ti)O3 Ceramics with Different MnO2 Concentrations for Ultrasound Transducer Applications

In this study, we investigate the piezoelectric characteristics of 0.55Pb(Ni1/3Nb2/3)O3-0.45Pb(Zr,Ti)O3 (PNN-PZT) with MnO2 additive (0, 0.25, 0.5, 1, 2, and 3 mol%). We focus on the fabrication of a piezoelectric ceramic for use as both actuator and sensor for ultrasound transducers. The actuator and sensor properties of a piezoelectric ceramic depend on the piezoelectric strain coefficient d and piezoelectric voltage coefficient g, related as g = d/εT. To increase g, the dielectric constant εT must be decreased. PNN-PZT with MnO2 doping is synthesized using the conventional solid-state reaction method. The electrical properties are determined based on the resonant frequencies and vibration modes measured by using an impedance analyzer. The MnO2 addition initially improves the tetragonality of the PNN-PZT ceramic, which then saturates at a MnO2 content of 1 mol%. Therefore, the dielectric constant and piezoelectric coefficient d33 steadily decrease, while the mechanical properties (Qm, Young’s modulus), tanδ, electromechanical coupling coefficient k, and piezoelectric voltage coefficient g were improved at 0.5–1 mol% MnO2 content.


Introduction
Piezoelectric materials have attracted considerable interest for various applications such as multi-layer ceramic actuator, transducer, sensor and actuator applications, and for analyses on fundamental science. Lead-based piezoelectric ceramics such as Pb(Zr,Ti)O 3 (PZT) have been extensively used in electrical devices because of their excellent piezoelectric properties [1]. Recently, the policies suggesting lead elimination have triggered studies on alternative compounds such as (K 0.44 Na 0.52 Li 0.04 )-(Nb 0.86 Ta 0.10 Sb 0.04 )O 3 (KNL-NTS) [2]. However, the lead-based compositions have exhibited higher piezoelectric performances than those of the lead-free compositions such as KNL-NTS.
The aim of this study was to better understand the effects of doping on PZT-based complex ceramics and improve the sensing performances and mechanical properties by using MnO 2 as an additive in the 0.55Pb(Ni 1/3 Nb 2/3 )O 3 -0.135PbZrO 3 -0.315PbTiO 3 ternary ceramic. The effects of the MnO 2 content on the piezoelectric, dielectric, and mechanical properties of the ceramic were investigated.

Synthesis of the MnO 2 -Doped PNN-PZT
The ceramics of PNN-PZT + MnO 2 (0, 0.25, 0.5, 1, 2, and 3 mol%) were synthesized by using the conventional solid-state method. Raw material powders (PbO, NiO, Nb 2 O 5 , ZrO 2 , and TiO 2 ) (Sigma Aldrich, 99.99%, Gillingham, UK) and MnO 2 additive (Sigma Aldrich, 99.99%, Gillingham, UK) were weighted in chemically stoichiometric proportions and ball-milled with distilled water for 24 h. After the ball milling, the slurry was dried at 80 • C, and then calcined at 900 • C (heating rate: 100 • C/h) for 2 h. The calcined powders were ball-milled again for 24 h with a 5 wt% polyvinyl alcohol (PVA) (Sigma Aldrich, 99+%, Gillingham, UK, M w = 89000-98000) solution as a binder for ceramic formation. The mixture was dried and crushed by a high-energy ball mill. The powder was sieved to control the particle size below 5 µm. The sieved powder was pressed into a mold (Φ = 30 mm) under a pressure of 100 MPa. The PVA in the ceramic disk was burnt out at 600 • C (heating rate: 100 • C/h) for 2 h. Subsequently, the ceramic disk was sintered at 1200 • C (heating rate: 100 • C/h) for 2 h with a spacer powder in an alumina crucible [14]. The sintered PNN-PZT ceramic disk was polished and coated with silver paste to obtain electrodes on both surfaces. The samples were polarized by applying a direct-current (DC) electric field of 2.5 kV/mm for 30 min at 50 • C in silicon oil.
The crystal structures of the sintered samples were characterized by X-ray diffraction (XRD) (X'pert Pro Powder, PANalytical, Netherlands). The surface microstructures of the as-sintered ceramics were observed by using field-emission scanning electron microscopy (SEM) (JSM-5900, JEOL, Akishima City, Japan). Three PNN-PZT specimens were fabricated for each composition. Their properties were measured, and then the average values were calculated. The piezoelectric coefficients (d 33 ) of the piezoelectric ceramics were measured by using a quasi-static piezoelectric d 33 meter (HY2730, Yangzhou, China). The planar electromechanical coupling coefficients (k p , k 31 ), piezoelectric coefficient (d 31 ), mechanical factor (Q m ), dielectric constant (ε r = ε T 33 /ε 0 , ε 0 = 8.854 × 10 −12 F/m), piezoelectric voltage coefficients (g 33 , g 31 ), and Young's modulus (Y E 11 ) were determined according to the method of resonance and antiresonance frequencies by using an impedance analyzer (HP 4194A, Agilent, Santa Clara, CA, USA) based on the Institute of Electrical and Electronics Engineers (IEEE) standards. Length-mode specimens (22 by 4 by 0.8 mm 3 ) were used to calculate Y E 11 , k 31 , d 31 , and g 31 . properties were measured, and then the average values were calculated. The piezoelectric coefficients (d33) of the piezoelectric ceramics were measured by using a quasi-static piezoelectric d33 meter (HY2730, Yangzhou, China). The planar electromechanical coupling coefficients (kp, k31), piezoelectric coefficient (d31), mechanical factor (Qm), dielectric constant (εr = ε T 33/ε0, ε0 = 8.854 × 10 −12 F/m), piezoelectric voltage coefficients (g33, g31), and Young's modulus (Y E 11) were determined according to the method of resonance and antiresonance frequencies by using an impedance analyzer (HP 4194A, Agilent, Santa Clara, CA, USA) based on the Institute of Electrical and Electronics Engineers (IEEE) standards. Length-mode specimens (22 by 4 by 0.8 mm 3 ) were used to calculate Y E 11, k31, d31, and g31. Figure 2 shows XRD patterns of the 0.55PNN-0.45PZT ceramics doped with different MnO2 contents (0, 0.25, 0.5, 1, 2, and 3 mol%). All samples exhibited typical ABO3 perovskite structures without the pyrochlore phase. Rhombohedral and tetragonal phases were found to coexist in the PNN-PZT ceramics. As shown in Figure 3, the XRD patterns were fitted by Gaussian functions in the 2θ range of 44.5-45.5°. The (200) peak consisted of three peaks, tetragonal (200) and (002) diffractions peaks presented in green and blue, respectively, and rhombohedral (200) diffraction peak presented in magenta. The red peak represents the overlap intensities of T(200), T(002), and R(200). Figure 2 shows the apparent changes in diffraction peaks, which indicate a gradual rhombohedral-to-tetragonal phase transition. Tetragonality is a crucial structural parameter of the perovskite lattice because it may affect the material properties. The tetragonality was calculated as IT (200) Figure 3. As shown in Figure  4, with the increase in MnO2 content, the tetragonality of the PNN-PZT ceramic initially largely increases, and then saturates at MnO2 contents higher than 1 mol% [13].   Figure 4, with the increase in MnO 2 content, the tetragonality of the PNN-PZT ceramic initially largely increases, and then saturates at MnO 2 contents higher than 1 mol% [13]. Materials 2019, 12, x FOR PEER REVIEW 5 of 10                Figure 6 shows the dielectric constants ε r and dielectric losses tanδ (%) of the PNN-PZT ceramics with different MnO 2 concentrations measured at 1 kHz at room temperature. The dielectric loss largely decreased with the MnO 2 content of 1 mol%, and then steadily increased in the MnO 2 range of 1 to 3 mol%. The dielectric loss reached the maximum (5.6%) for the undoped PNN-PZT ceramic and minimum (1.6%) for the 1 mol% MnO 2 sample. Therefore, the dielectric loss decreased by almost five times upon slight MnO 2 doping. This is consistent with the change rate of the tetragonality. The dielectric loss rapidly decreased with the increase in tetragonality from 0 to 1 mol%. The dielectric loss was improved at MnO 2 content above 1 mol%, because the tetragonality was saturated at 1 mol%. Figure 6 shows the negative and positive effects on the dielectric constant ε r and dielectric loss tanδ (%). A negative effect, decrease in dielectric constant, was observed with the increase in tetragonality with the MnO 2 content. In terms of the dielectric loss, the minimum value could be explained by the competition between the positive effect (increase in tetragonality) and negative effect of the Mn ions on the motion of the wall domains.

Dielectric Properties
With the increase in MnO 2 content, the dielectric constant decreased because the hardener Mn ions affected the domain movement. The decrease in dielectric constant was caused by the oxygen vacancies generated by the substitutions of the high-valence Ti 4+ and Zr 4+ in the perovskite lattice by the low-valence Mn 2+ and/or Mn 3+ . The dielectric constant rapidly decreased up to the MnO 2 content of 1 mol%. However, the decrease rate of the dielectric constant was smaller at MnO 2 contents of 1 to 3 mol%.  Figure 6 shows the dielectric constants εr and dielectric losses tanδ (%) of the PNN-PZT ceramics with different MnO2 concentrations measured at 1 kHz at room temperature. The dielectric loss largely decreased with the MnO2 content of 1 mol%, and then steadily increased in the MnO2 range of 1 to 3 mol%. The dielectric loss reached the maximum (5.6%) for the undoped PNN-PZT ceramic and minimum (1.6%) for the 1 mol% MnO2 sample. Therefore, the dielectric loss decreased by almost five times upon slight MnO2 doping. This is consistent with the change rate of the tetragonality. The dielectric loss rapidly decreased with the increase in tetragonality from 0 to 1 mol%. The dielectric loss was improved at MnO2 content above 1 mol%, because the tetragonality was saturated at 1 mol%. Figure 6 shows the negative and positive effects on the dielectric constant εr and dielectric loss tanδ (%). A negative effect, decrease in dielectric constant, was observed with the increase in tetragonality with the MnO2 content. In terms of the dielectric loss, the minimum value could be explained by the competition between the positive effect (increase in tetragonality) and negative effect of the Mn ions on the motion of the wall domains.

Dielectric Properties
With the increase in MnO2 content, the dielectric constant decreased because the hardener Mn ions affected the domain movement. The decrease in dielectric constant was caused by the oxygen vacancies generated by the substitutions of the high-valence Ti 4+ and Zr 4+ in the perovskite lattice by the low-valence Mn 2+ and/or Mn 3+ . The dielectric constant rapidly decreased up to the MnO2 content of 1 mol%. However, the decrease rate of the dielectric constant was smaller at MnO2 contents of 1 to 3 mol%.  Figure 7 shows the mechanical properties of the PNN-PZT ceramics with varying MnO2 content (0-3 mol%). The mechanical quality factor Qm reflects the steepness of the resonance of the mechanical vibration around the resonance frequency. Therefore, Qm has been considered the main parameter of an ultrasonic actuator. Qm was improved by approximately five times (from 42.70 to 202.26) with the increase in MnO2 content. In addition, the Young's modulus Y E 11 was improved from 7.14 to 10.56 with the increase in MnO2 content (a similar curve shape was observed). The changes in mechanical properties with the MnO2 content can be attributed to the oxygen vacancies generated by the accepter doping effect. Therefore, the densifications of the ceramics resulted from the introduction of MnO2, which improved the piezoelectric properties [13]. The density was improved from 7771.09 to 7938.41 kg/m 3 with the increase in MnO2 content ( Table 2). The oxygen vacancies improved the mechanical properties.  Figure 7 shows the mechanical properties of the PNN-PZT ceramics with varying MnO 2 content (0-3 mol%). The mechanical quality factor Q m reflects the steepness of the resonance of the mechanical vibration around the resonance frequency. Therefore, Q m has been considered the main parameter of an ultrasonic actuator. Q m was improved by approximately five times (from 42.70 to 202.26) with the increase in MnO 2 content. In addition, the Young's modulus Y E 11 was improved from 7.14 to 10.56 with the increase in MnO 2 content (a similar curve shape was observed). The changes in mechanical properties with the MnO 2 content can be attributed to the oxygen vacancies generated by the accepter doping effect. Therefore, the densifications of the ceramics resulted from the introduction of MnO 2 , which improved the piezoelectric properties [13]. The density was improved from 7771.09 to 7938.41 kg/m 3 with the increase in MnO 2 content ( Table 2). The oxygen vacancies improved the mechanical properties.

Piezoelectric Properties
The electromechanical coupling coefficient (kp, −k31) is a constant representing the piezoelectric efficiency of a piezoelectric ceramic, i.e., it represents the efficiency of conversion of electrical energy into mechanical energy. As shown in Figure 8, kp decreased with the increase in MnO2 content to 0.25 mol%. However, it was largely increased at MnO2 contents of 0.25 and 1 mol%, where it reached the maximum of 0.41. Above this concentration, it steadily decreased in the MnO2 content range of 1 to 3 mol%. The curve shape of −k31 was different. It had the maximum value for the undoped PNN-PZT ceramic, and then steadily decreased with the increase in MnO2 content. The −k31 values at 0.25-1 mol% were similar. At MnO2 concentrations less than 1 mol%, the changes may be attributed to the relatively large increase in tetragonality. Above this concentration, kp and −k31 decreased owing to the hardener doping effect (acceptor substitution inducing point defects) [12,13]. The piezoelectric coefficients (−d31, d33) reflect the distortion originating from the application of an electric field having a uniform strength without stress. Therefore, the piezoelectric coefficients (−d31, d33) have been considered the primary parameters of actuators. The piezoelectric coefficient was calculated by using −k31, Y E 11, and ε T 33:

Piezoelectric Properties
The electromechanical coupling coefficient (k p , −k 31 ) is a constant representing the piezoelectric efficiency of a piezoelectric ceramic, i.e., it represents the efficiency of conversion of electrical energy into mechanical energy. As shown in Figure 8, k p decreased with the increase in MnO 2 content to 0.25 mol%. However, it was largely increased at MnO 2 contents of 0.25 and 1 mol%, where it reached the maximum of 0.41. Above this concentration, it steadily decreased in the MnO 2 content range of 1 to 3 mol%.

Piezoelectric Properties
The electromechanical coupling coefficient (kp, −k31) is a constant representing the piezoelectric efficiency of a piezoelectric ceramic, i.e., it represents the efficiency of conversion of electrical energy into mechanical energy. As shown in Figure 8, kp decreased with the increase in MnO2 content to 0.25 mol%. However, it was largely increased at MnO2 contents of 0.25 and 1 mol%, where it reached the maximum of 0.41. Above this concentration, it steadily decreased in the MnO2 content range of 1 to 3 mol%. The curve shape of −k31 was different. It had the maximum value for the undoped PNN-PZT ceramic, and then steadily decreased with the increase in MnO2 content. The −k31 values at 0.25-1 mol% were similar. At MnO2 concentrations less than 1 mol%, the changes may be attributed to the relatively large increase in tetragonality. Above this concentration, kp and −k31 decreased owing to the hardener doping effect (acceptor substitution inducing point defects) [12,13]. The piezoelectric coefficients (−d31, d33) reflect the distortion originating from the application of an electric field having a uniform strength without stress. Therefore, the piezoelectric coefficients (−d31, d33) have been considered the primary parameters of actuators. The piezoelectric coefficient was calculated by using −k31, Y E 11, and ε T 33:

=
(1) The curve shape of −k 31 was different. It had the maximum value for the undoped PNN-PZT ceramic, and then steadily decreased with the increase in MnO 2 content. The −k 31 values at 0.25-1 mol% were similar. At MnO 2 concentrations less than 1 mol%, the changes may be attributed to the relatively large increase in tetragonality. Above this concentration, k p and −k 31 decreased owing to the hardener doping effect (acceptor substitution inducing point defects) [12,13]. The piezoelectric coefficients (−d 31 , d 33 ) reflect the distortion originating from the application of an electric field having a uniform strength without stress. Therefore, the piezoelectric coefficients (−d 31 , d 33 ) have been considered the primary parameters of actuators. The piezoelectric coefficient was calculated by using −k 31 , Y E 11 , and ε T 33 : Materials 2019, 12, 4115 8 of 10 Figure 9 shows the piezoelectric coefficients (−d 31 , d 33 ) of the PNN-PZT ceramics with varying MnO 2 content (0-3 mol%). −d 31 and d 33 exhibited similar curve shapes; they decreased with the increase in MnO 2 content because the decrease rate of ε T 33 was considerably higher than the increase rates of k 31 and Y E 11 . The piezoelectric coefficient exhibited the minimum decrease rate between 0.5 and 1 mol% MnO 2 , because of the relatively significant increase in tetragonality at 0.5-1 mol%.
Materials 2019, 12, x FOR PEER REVIEW 8 of 10 rates of k31 and Y E 11. The piezoelectric coefficient exhibited the minimum decrease rate between 0.5 and 1 mol% MnO2, because of the relatively significant increase in tetragonality at 0.5-1 mol%.  The piezoelectric voltage coefficients (−g31, g33) reflect the field strength, originating from a uniform applied stress without electrical displacement, and thus they represent the sensor properties. The values of −g31 and g33 were calculated by using the relationship between the piezoelectric coefficient d and dielectric constant, g = d/ε T . −g31 exhibited a similar tendency to that of k31 ( Figure  10). g33 steadily increased to a high value of 20.31 × 10 −3 Vm/N at 1 mol%. However, it was decreased at 2 mol%, and then largely increased to the maximum of 21.13 × 10 −3 Vm/N at 3 mol%, as the decrease rate of d33 was smaller than the rate of decrease in ε T 33. Table 2 shows the densities and dielectric, mechanical, and piezoelectric properties of the PNN-PZT samples with different MnO2 contents (0, 0.25, 0.5, 1, 2, and 3 mol%).
The MnO2 doping can affect the PNN-PZT ceramic properties owing to the increase in number of oxygen vacancies generated by the substitutions of the high-valence Ti 4+ and Zr 4+ in the perovskite lattice by the low-valence Mn 2+ and/or Mn 3+ , as mentioned above [4]. The oxygen vacancies improved the mechanical properties (Young's modulus, Qm). The Young's modulus and Qm exhibited similar curve shapes. Initially, they were rapidly improved, but their increase rates were reduced at MnO2 contents above 1 mol%. The phase of PNN-PZT transformed from rhombohedral to tetragonal with the increase in MnO2 content. The phase transition is attributed to the enhancements in electrical The piezoelectric voltage coefficients (−g 31 , g 33 ) reflect the field strength, originating from a uniform applied stress without electrical displacement, and thus they represent the sensor properties. The values of −g 31 and g 33 were calculated by using the relationship between the piezoelectric coefficient d and dielectric constant, g = d/ε T . −g 31 exhibited a similar tendency to that of k 31 ( Figure 10). g 33 steadily increased to a high value of 20.31 × 10 −3 Vm/N at 1 mol%. However, it was decreased at 2 mol%, and then largely increased to the maximum of 21.13 × 10 −3 Vm/N at 3 mol%, as the decrease rate of d 33 was smaller than the rate of decrease in ε T 33 . Table 2 shows the densities and dielectric, mechanical, and piezoelectric properties of the PNN-PZT samples with different MnO 2 contents (0, 0.25, 0.5, 1, 2, and 3 mol%). rates of k31 and Y E 11. The piezoelectric coefficient exhibited the minimum decrease rate between 0.5 and 1 mol% MnO2, because of the relatively significant increase in tetragonality at 0.5-1 mol%.  The piezoelectric voltage coefficients (−g31, g33) reflect the field strength, originating from a uniform applied stress without electrical displacement, and thus they represent the sensor properties. The values of −g31 and g33 were calculated by using the relationship between the piezoelectric coefficient d and dielectric constant, g = d/ε T . −g31 exhibited a similar tendency to that of k31 ( Figure  10). g33 steadily increased to a high value of 20.31 × 10 −3 Vm/N at 1 mol%. However, it was decreased at 2 mol%, and then largely increased to the maximum of 21.13 × 10 −3 Vm/N at 3 mol%, as the decrease rate of d33 was smaller than the rate of decrease in ε T 33. Table 2 shows the densities and dielectric, mechanical, and piezoelectric properties of the PNN-PZT samples with different MnO2 contents (0, The MnO 2 doping can affect the PNN-PZT ceramic properties owing to the increase in number of oxygen vacancies generated by the substitutions of the high-valence Ti 4+ and Zr 4+ in the perovskite lattice by the low-valence Mn 2+ and/or Mn 3+ , as mentioned above [4]. The oxygen vacancies improved the mechanical properties (Young's modulus, Q m ). The Young's modulus and Q m exhibited similar curve shapes. Initially, they were rapidly improved, but their increase rates were reduced at MnO 2 contents above 1 mol%. The phase of PNN-PZT transformed from rhombohedral to tetragonal with the increase in MnO 2 content. The phase transition is attributed to the enhancements in electrical properties, which led to the piezoelectric property changes. Therefore, the dielectric constant, Young's modulus, electromechanical coupling coefficients (k p ,−k 31 ), and piezoelectric properties (−d 31 , d 33 , −g 31 , and g 33 ) of PNN-PZT changed with the density of MnO 2 [4].
Q m and Young's modulus (Y E 11 ) also increased with the MnO 2 content owing to the oxygen vacancies generated by the MnO 2 doping. The dielectric properties and electromechanical coupling coefficient k were optimal at MnO 2 contents of 0.5-1 mol%, where tanδ and k p had the maximum values. The changes in electrical properties were attributed to the increased tetragonality. PNN-PZT is a soft piezoelectric material more suitable for actuator applications. The addition of MnO 2 to PNN-PZT showed its potentials for use in sensory actuators. In following studies, we aim to investigate vibration control and nondestructive testing applications based on the enhanced PNN-PZT ceramics.

Conflicts of Interest:
The authors declare no conflict of interest.