Temperature-Dependent Properties of a 1–3 Connectivity Piezoelectric Ceramic–Polymer Composite

Introduction

1–3 connectivity piezoelectric ceramic–polymer composites have been widely used for piezoelectric transducers in many fields, such as hydrophones for low-frequency (Zhang et al. 1993), ultrasonic transducers (Chen et al. 2013), non-destructive testing for high frequency (Kirk and Schmarje 2013) and so on. Comparing with monolithic 0.90Pb(Zr_0.52Ti_0.48)O₃ (PZT) ceramics, the 1–3 composites show advantages of higher piezoelectric constants g_h, higher value of k_t, flexible processing in manufacturing, and relatively low impendence Z which leads to better material performance in matching with human tissue and water (Dyaz and Castillero 2001; Wang et al. 2008). At present, methods developed for processing fine-scale 1–3 piezoelectric composites include dice-and-fill (Xu et al. 2012), injection molding (Leslie, Gentilman, and Pham 1993), viscous-polymer processing (VPP) (Abrar et al. 2004; Bernassau et al. 2012), among which dice-and-fill is the uppermost process because of its mature technique and easy setup preparation. On the other hand, in the VPP, the size and shape of fibers can be changed flexibly as well as the distribution in the composites, therefore the prepared 1–3 composites have advantages in optimizing the material properties to meet various requirements for different applications (Bowen and Stevens 2006). Now there are many studies (Zhen and Li 2008; Zhou et al. 2011; Shen, Xu, and Li 2009; Chaipanich et al. 2011; Potong et al. 2012; Ramesh et al. 2011) focusing on the 1–3 composites, mostly concentrating on the preparation, the dielectric and ferroelectric properties at room temperature. The material properties of 1–3 composites are decisively determined by the piezoelectric phase and the polymer matrix. However, comparing with the piezoelectric phase, the epoxy matrix is much more sensitive to temperature for the creeping behavior of the molecular chain and the segment of the epoxy at high temperature, which makes the epoxy phase temperature sensitive with poor heating stability (Majda and Skrodzewicz 2009). Transducers or sensors containing the 1–3 composites are always served at different or even extreme environment with varying temperatures, so it is quite important and necessary to study the temperature dependent properties of the 1–3 composites. However, the performance of the 1–3 composites concerning the temperature changing has not been researched sufficiently. In this study, temperature dependences of the vibration mode, dielectric properties, piezoelectric properties, ferroelectric hysteresis behavior of the 1–3 connectivity PZT composites are investigated.

Experiment

The piezoelectric fibers of 0.90Pb(Zr_0.52Ti_0.48)O₃–0.05Pb(Mn_1/3Sb_2/3)O₃–0.05Pb(Zn_1/3Nb_2/3)O₃ (abbreviated as PZT–PMS–PZN) were prepared by VPP, including the preparation of PZT powder, and a sintering process at 1,100°C for 3 h. The diameter of a single sintered piezoelectric fiber is about 300 μm. The piezoelectric fibers were aligned vertically in a sample holder, followed by casting low viscosity epoxy to form a fiber/epoxy 1–3 composite, where the epoxy was firstly degassed for 30 min and then cured at 80°C for 4 h. The composite was then cut into thin disk and painted with low-temperature silver paste on both sides. The thickness of the 1–3 composite sample is 1 mm and the volume fraction of the piezoelectric fibers (φ) is about 50%. The composite samples were then poled at 80°C for 30 min under an applied field of 40 kV/cm in silicone oil.

The density of the 1–3 composites was measured by Archimedes method with an average value of 4.38 g/cm³. The crystal structure of the piezoelectric fibers was examined using an X-ray diffractometer (XRD, Bruker D8 Advanced, CuKα radiation, λ=0.15418 nm). The microstructures of the piezoelectric fibers were observed by a scanning electrical microscopy (SEM, Leo 1550, Zeiss Company). The piezoelectric coefficient (d₃₃) was measured using a quasi-static piezo-d₃₃ meter (ZJ-3A, Institute of Acoustics Academia Sinica) after aging the samples for 24 h. All the temperature-dependent properties such as dielectric permittivityεr, the planar coupling coefficient (k_p), the thickness coupling coefficient (k_t), acoustic velocity (υL) and the acoustic impedance (Z) were measured by using an impedance analyzer (Agilent 4294A) in a temperature control system (Linkam), where the values of k_p, k_t and the mechanical quality factor (Q_m) were determined from the resonance and antiresonance method based on the IEEE standards. The dielectric permittivity ε33T determined by measuring the capacitance C at 1 kHz was then calculated from the equation:

where S is the electrode area and t is the thickness of the sample, and ε0 is the permittivity of the free space. The sound velocity υL and acoustic impedance Z were calculated from the following formulas:

where t, f_a and ρ are the thickness, antiresonant frequency and density of the sample, respectively. Ferroelectric hysteresis loops were measured under 35 kV/cm electric field at room temperature by using a ferro-analyzer (TF2000, aixACCT GmbH).

Results and Discussions

The typical piezoelectric properties for the 1–3 composite and piezoelectric ceramic are listed in Table 1. The data comparison displays that PZT ceramic has a much larger value of ε33T, Q_m, d_33,Z, k_p, and a relatively smaller value of tanδ, k_t. The 1–3 composites usually work in thickness mode, so the dynamic properties such as electromechanical coupling coefficient k_t, mechanical quality factor Q_m and the ratio of k_t to k_p were significantly determined by the resonant f_r and antiresonant frequencies f_a at thickness resonance mode. Usually, for piezoelectric material, higher thickness coupling coefficient k_t means higher power transduction capability. For 1–3 composites, a larger k_t/k_p value is necessary for better resolution in anisotropy. From Table 1, the k_t of the 1–3 composite is 0.59 and the k_t/k_p value is 2.1, showing a clear thickness mode. Moreover, another typical dynamic property for the 1–3 composite is the mechanical quality factor Q_m (about 10), which is beneficial to application in the imaging transducers for the characteristic of obtaining broad operating bandwidth (Lee et al. 2012). Besides, the acoustic impendence Z of the 1–3 composites is also much lower than PZT ceramics because of the lower acoustic impendence of polymer, therefore the 1–3 composites are more suitable for applications requiring good matching with human tissue and water. In all, the special characteristics of the 1–3 composites in high quality ensure the wide applications in many fields.

Figure 1(a) depicts the XRD patterns of PZT–PMS–PZN fibers in 2θ range between 20° and 60°. It can be seen that the specimen exhibits a pure perovskite structure; no discernible peaks of second phases such as pyrochlore is observed, indicating that the adding organics do not affect the forming of the perovskite structure. Figure 1(b) shows the SEM micrograph for the cross section of the sintered piezoelectric fibers with a high-density morphology and the diameter of a single piezoelectric fiber is about 300 μm. Moreover, Figure 1(c) shows the SEM micrograph of inner grains of piezoelectric fiber with the average grain size around 2–4 μm.

Figure 1:

(a) XRD patterns of PZT–PMS–PZN fibers; (b) SEM micrograph for sectional piezoelectric fiber; and (c) SEM micrograph revealing inner grains of piezoelectric fiber.

Figure 2(a) shows the contrast in the impedance characteristic of the 1–3 composite and the monolithic piezoelectric ceramic. As shown, the piezoelectric ceramic has two major resonance modes: planar mode and thickness mode, in which the planar mode is the typical working mode for monolithic piezoelectric ceramic. On the contrary, for the 1–3 composites, the low-frequency resonance at planar mode has been remarkably depressed and only the thickness resonance mode with resonance frequency over 1 MHz is seen. Figure 2(b) shows the trend about the impedance Z and phase angle of the 1–3 composite under various temperatures from 20°C to 100°C. From Figure 2(b), we can see that the antiresonant frequency fa and resonant frequency fr at the thickness mode and planar mode both decreased as temperature increases. Because of the temperature sensitivity of the epoxy phase, its elastic compliance increases distinctly when temperature increases. The 1–3 composite contains epoxy phase; thus, the overall elastic compliance shows increasing trend, resulting in decreased resonant frequency and antiresonant frequency correspondingly. However, the resonance peak at thickness mode is still clear even at the temperature of 100°C. In comparison, as to the planar mode, the resonance peak has been further depressed by rising temperature, and the major resonance peak almost disappeared at 100°C. Based on temperature-dependent vibrations at planar and thickness mode, this kind of 1–3 composites could work in the environment up to 100°C for the stable continuous vibration at thickness mode.

Figure 2:

(a) Vibration mode comparison between the 1–3 composite and piezoelectric ceramic; (b) the temperature-dependent planar mode and thickness mode of the 1–3 composite with volume fraction ϕ = 50%.

Figure 3 shows the temperature-dependent dielectric permittivity εr of the 1–3 composite at 1 kHz to 1 MHz from 30°C to 100°C. It can be seen that all the dielectric permittivity εr under different frequencies increase as temperature increases, and the dielectric permittivity ε33T value comes to 723 at 100°C, 1 kHz. In addition, the dielectric permittivity εr value decreases when the working frequency rises at a certain temperature, on the performance that ε33S is 676 at 100°C, 1 MHz. The εr value of the 1–3 piezoelectric composite is much lower than the piezoelectric ceramic due to the lower εr of the existing epoxy phase. However, the temperature dependence in dielectric permittivity εr of the 1–3 composite is in accordance with that of the piezoelectric ceramic, and adding of epoxy phase does not affect the typical regularity in dielectric permittivity of the piezoelectric composite materials. Meanwhile, the dielectric loss tanδ shows the same trend with temperature as dielectric permittivity εr does. The dielectric loss of the 1–3 composite is higher than pure piezoelectric ceramic due to the contributions from the epoxy phase with a relatively high dielectric loss property.

Figure 3:

Temperature-dependent dielectric permittivity εr, dielectric loss tanδ of the 1–3 composite at different frequencies.

Figure 4 shows the trend of the temperature-dependent k_p, k_t and k_t/k_p of the 1–3 composite with volume fraction ϕ of 50%. As observed from Figure 4, k_t increases with the temperature and reaches 0.705 at 100°C. On the contrary, the k_p value decreases as temperature increases and drops even faster at higher temperature. The k_p value decreased to almost 0 at 100°C, which is in accordance with Figure 1(b), showing that there is no obvious peak at planar mode. As a result, the k_t/k_p value increases all the time with the temperature. However, in this paper, temperature range is not wide enough considering the role of epoxy in the composite, and this kind of 1–3 composite shows the good characteristic of large k_t/k_p value within temperature of 100°C, which is an important factor to ensure the application of the 1–3 composite for its anisotropy property (Meryer, Lopath, and Yoshikawa 1997).

Figure 4:

Temperature-dependent planar coupling coefficient k_p, thickness coupling coefficient (k_t), the ratio of k_t to k_p (k_t/k_p) of the 1–3 composite.

Figure 5 shows the temperature-dependent acoustic velocity υL and acoustic impedance Z of the 1–3 composite. From formulas about υL and Z (eq. [3]), it is known that the variation about the values of υL and Z shares a similar trend of f_a. From room temperature to 100°C, the average value of υL and Z is 3,018 m/s, 13.22 MRayl, respectively.

Figure 5:

The temperature-dependent acoustic velocity υL and acoustic impedance Z of the 1–3 composite.

Figure 6 shows the P–E hysteresis loop which includes the contributions from PZT fibers and epoxy at room temperature. The extracted P–E loops of the 1–3 composite at different frequencies demonstrate a coercive field E_c of 9.83 kV/cm and a remnant polarization P_r of 4.42 μC/cm² at 5 Hz. At the frequency from 20 to 5 Hz, the P_r value is increased from 3.78 to 4.42 μC/cm², and the coercive field E_c increases from 8.96 to 9.83 kV/cm. Compared with the PZT bulk ceramic (E_c = 8.30 kV/cm, P_r = 7.80 μC/cm², 5 Hz), the coercive filed of the 1–3 composite is larger but the remnant polarization P_r is smaller. It is known that epoxy polymer is without ferroelectric polarization, so the role of epoxy decreases the total P–E effect of the 1–3 composite; therefore, the average value is smaller than the PZT bulk ceramics. Additionally, in the 1–3 composites, the high resistance of the polymer shields the external electric field, so the external electric field could not exert on the PZT fibers totally, which increases the coercive field and decreases the polarization as well (Sun et al. 2010).

Figure 6:

Ferroelectric polarization-electric field (P–E) hysteresis loops of the 1–3 composite at room temperature for different frequencies.

Conclusions

The temperature-dependent dielectric, piezoelectric and ferroelectric properties of the 1–3 composites have been investigated in this study. The observed properties are critical for the application of the 1–3 composites in various environments, and the variation may be due to the temperature effects on the piezoelectric fibers and the epoxy matrix. A clear and uncoupled thickness mode is confirmed even at a high temperature of 100°C, indicating that there is a great potential for the composites to be applied in many circumstances.