Sol-Gel Synthesis and Piezoelectric and Structural Properties of Zr –rich PZT Nanoparticles

INTRODUCTION

Pb-based perovskite solid solution of lead zirconate titanates (PZT; Pb(Zr_xTi_1-x) O₃) have been considered due to its high spontaneous polarization abilities, piezoelectric coefficient, dielectric permittivity and pyroelectricity [1,2]. The highest values for piezoelectric coefficient (d₃₃) and the permittivity (ɛ_r) have been observed for chemical composition of x=0.52, the so-called morphotropic phase boundary (MPB), at which the tetragonal (Ti-rich) and rhombohedral (Zr-rich) phases coexist [3,4]. The excellent piezoelectric properties make it a promising material for sensors, optoelectronic and electromechanical transducer application [5]. Nano-sized particles of ferroelectric materials exhibited significant properties which recently motivated many in-depth researches on future applications [6-8]. For example, PZT nanoparticles (PZT-NPs) can also be suspended in PZT sol to serve as seeds to lower sintering temperature and reduction of Pb loss [9].

Currently, there are two approaches to make PZT-NPs. The first way is a top-down approach often used in the industry [10]. The second method is a bottom-up approach that is still underactive investigation by various researchers. Size and quality of PZT-NPs are controlled by various process parameters, such as temperature, time, and mineralizer concentration [11-15]. A wide variety of preparation routes have been employed to produce PZT-NPs such as hydrothermal [16,17], electrodynamics atomization [18,19], ultrasonic spray combustion synthesis (USCs) [20] and sol-gel [21-23]. Among these techniques, sol-gel process provides the advantage of homogenous chemistry in multi-component system, ease of composition variation, and low processing temperature [24]. Acetic acid based sol–gel processing of PZT, introduced by Yi et al. [25], and later developed by Assink and Schwartz [26], offers a simpler preparation route compared to the 2-methoxyethanol based sol–gel. The method was based on the reduced reactivity of transition metal alkoxides by acetic-acid modification.

In the present study, PZT precursor solution was prepared from different metal-organic compounds and different experimental processes. Then, PZT nanopowders were synthesized by acid acetic based sol-gel method. The properties of PZT nanopowders were investigated to evaluate the effect of calcination temperature on the crystal structure of PZT nanopowders by using an X-ray diffractometer (XRD). The electromechanical and piezoelectric behavior of PZT was determined by the resonance/anti-resonance method.

MAERIALS AND METHODS

The chemicals used to prepare the PZT sol were; lead acetate trihydrate (Pb(C₂H₃O₂)₂. 3H₂O, denoted as Pb(OAc)₂. 3H₂O, Merck.), tetra-iso-propylitanate (aka titanium-isopropoxide, Ti[OCH(CH₃)₂]₄, denoted as Ti(Oⁱpr)₄, Sigma-Aldrich Co.), zirconium n-propoxide solution with 70% ^w/w in n-propanol (Zr[O(CH₂)₂CH₃]₄, denoted as Zr(Opr)₄, Sigma-Aldrich Co.), glacial acetic acid (HOAc, Sigma-Aldrich Co., 99.7% pure) and acetylacetone (C₅H₈O₂, denoted as AcAc, Merck Co.).

Synthesis

First, 13.9 gr Pb(OAc)₂.3H₂O was dissolved in acetic acid on a 1:3 molar ratio while stirred and refluxed at 110 ^°C during 3 h for dehydration and homogeneity purposes (solution A). Subsequently, 0.48 gr Ti (Oⁱpr)₄ was mixed with (2.52 gr) acetylacetone in order to avoid fast hydrolysis of reactants and continuously stirred at room temperature for 4h. Then 14.8 gr Zr(Opr)₄ was t added into the mixture, which will be referred hereafter as solution B. Pb and Zr/Ti precursor solutions were then mixed and refluxed for 3 h at 85 ^°C. For hydrolysis, 1.8 gr of distilled water was added to it and pH value was adjusted using Hydrochloride acid to allow the nanoparticle precursor to slow grow. The solution was stirred overnight at room temperature to get slightly yellowish clear transparent sol and then heated at 70 ^°C to obtain the gelation of Zr -rich PZT sol. The gel was dried at 120 °C and heat treated in two stages. The first heat treatment was carried out at 400 °C/ 2h with a heating rate of 5°C/min. The obtained powder agglomerates were manually milled and then calcined at 700°C for 2h (5°C/min) to promote the formation of the perovskite structure of PZT. After this second heat treatment, the carbon compounds were eliminated leading to nano-sized PZT powders with high purity. The columbic synthetic route employ is shown schematically in Fig. 1. Nano-sized PZT powders obtained were plasticized by 4 wt.% polyvinyl alcohol (PVA) addition, granulated and cold-pressed to disc-pellets at 120 MPa. Green ceramic discs (diameter (D): 10mm & thickness (t): 4.5 mm) were sintered (5°C/min) in PbZrO₃ rich atmosphere, at various temperatures.

Characterization

Density (ρ) of sintered discs was measured by the liquid displacement technique based on the Archimedes’ principle. Silver paste was applied on both sides of ceramic discs as electrodes to provide a better ohmic contact for electrical measurements. Polarization was done at 120 °C/30 min in a silicon oil bath under a DC electric field of 15 kV/cm. The piezoelectric strain coefficient (d₃₃) of each disc was measured 24 h after poling, by a quasi-static d₃₃ -meter (RM3500, KCF Tech). Coupling factor (k_p) and mechanical quality factor (Q_m) were determined by the resonance/anti-resonance method using an Agilent 4294A impedance analyzer (hp Hewlett Packard). The parameters k_p and Q_m were related to the resonance frequencies by means of the following expressions [14,29,30]:

where f_r is the resonant frequency and f_a is the anti-resonant frequency in the fundamental vibration mode. Z_m and C represent the resonant impedance and the capacitance on the mechanical branch of the equivalent circuit, respectively.

Apparatus

Thermo-gravimetric analysis (TGA) and differential thermal analysis (DTA) with a heating rate of 5 °C/min was done to follow the phase evolution during heat treatment in air. Additionally, the phase purity and crystalline size were determined by an X-ray diffraction (XRD, D8-Advance Brucker diffractometer) with Cu kα radiation (λ=1.5418 °A), scanned 2θ from 20° to 85°). Energy dispersive X-ray spectroscopy (EDS) attaching to SEM was used to analyze the elemental composition. Particle size and morphological characteristics of PZT nanoparticles were monitored via scanning electron microscope (SEM; VEGATESCAN/LMU) and transmission electron microscope (TEM; Philips CM 30) at an accelerating voltage of 100 kV. The fine powder was dispersed onto a carbon coated copper grid. The electro-mechanical properties_ including; voltage constant (g₃₃), mechanical quality (Q_m) and coupling (k_p) factors, dielectric constant (ε), etc._ were examined using an Agilent 4294A impedance analyzer (hp Hewlett Packard) and the piezoelectric constant d₃₃ was measured with a quasi-static d₃₃ –meter (RM3500, KCF Tech.).

RESULT AND DISCUSSION

DTA-TG

Thermal decomposition behavior of PZT gel was examined by thermo gravimetry (TG) and differential thermal analysis (DTA), shown in Fig. 2. At temperature of up to 400 ^°C, endothermic (Ed) and exothermic (Ex) reactions may be attributed to the volatilization and burnout of residual solvent in the sol [5]. The Ed₁ peak at around 70 ^°C is due to the evaporation of water absorbed in the PZT solution (rate of weight loss: 4.173%). Also, the vaporization of light molecular organics occurs between 210-310 ^°C which corresponds to Ed₂, accompanied by a 27.492% weight loss. As a result, the exothermic peaks at 390, 480 and 530 ^°C were considered to be related to progressive decomposition of an organic group from Pb-Zr-Ti alkoxide and due to crystallization of metal oxide phase [3,5]. Therefore, crystallization of PZT powders takes place at 480 ^°C, and perovskite structure is established above 530 ^°C.

XRD

X-ray diffraction patterns of PZT nano-powders calcined at various temperatures are shown in Fig. 3. A series of characteristic peaks: (012), (110), (202), (024) and (018) crystal planes at 2θ values are observed of nanopowders, which can be indexed based on rhombohedral phase (ICDD No. 1-070-0740). The perovskite -type structure emerged at a higher temperature of 400 °C which is due to large grain size and high extent of crystallinity [2-5]. The relative intensity of pyrochlore peak (29.4°) at calcination temperature of 400 °C with respect to PZT peak (30.62°) is much stronger. At temperature of 500 and 600 °C, the pyrochlore and perovskite phases coexisted, which is consistent with the results reported elsewhere [2,3,12]. However, the pyrochlore phase at 29.4° of 2θ was transformed into perovskite phases at temperatures of 700 °C. By comparison, it can be found from Fig. 3; (i) At 400 °C the powders calcined were already crystalline, and composed of phases of cerussite (PbCO₃: Ref. code 5-417), hydrocerussite (Pb₃(CO₃)₂(OH)₂: Ref. code 13-131), and laurionite (PbClOH: Ref. code 31-680), monoclinic zirconia (ZrO₂: Ref. code 5-543), and anatase titanium oxide (TiO₂: Ref. code 2-387). (ii) Following calcination at 500- 600 °C, the PZT precursors were composed of phases of litharge (PbO: Ref. code 1-796), PbZrO₃ (Ref. code 1-75-1607) and PbTiO₃ (Ref. code 1-75-1605). Lead hydroxyl carbonate phases (i.e., cerussite and hydrocerussite) were mostly converted into their oxides (i.e., litharge). (iii) Upon calcination at 700°C, the phases of litharge and pyrochlore completely disappeared, and the initial formation of rhombohedral PZT (Ref. code. ICDD 1-70-740) was observed. The average particle size (S_p) was calculated by the Debye– Scherrer formula [17]; S_p= 0.89λ/βCosθ, where λ is the X-ray wavelength (1.54 ˚A), θ the diffraction angle and β is the peak width of half maximum (the S_p of PZT nano-powders calcined at various temperatures are shown in Fig. 3.).

SEM-TEM

Typical TEM micrographs of represent PZT nano-powders calcined at 700 °C are shown in Fig. 4. From Fig. 4, nearly spherical shape particles, highly crystalline nature and size distribution less than 20nm can be observed which is a relatively agglomeration state due to high surface energy of particles in nano-scale.

The surface morphology of PZT nano-powders nanocrystals was analyzed by scanning electron microscopy. Fig. 5 present the SEM photographs of PZT nano-powders calcined at 700 °C. Though the SEM images did not show well dispersed PZT nano-powders, but the PZT aggregated particles appear to be composed of much smaller crystallites. Also, to better identification of elemental composition of the synthesized material, Energy dispersive X-ray measurements were carried out by the EDX spectrometer attached to SEM. The EDS pattern of PZT is shown in Fig. 5f, which confirms the existence of Pb, Zr and Ti elements.

Piezoelectric study

To determine piezoelectric property of ceramics, resonant vibration spectrum of samples was measured at room temperature. Fig. 6 shows the impedance-phase spectra on frequency for piezo-ceramics sintered at 1150 °C. From these results, the resonant (f_r) and anti-resonant (f_a) frequencies, the minimum impedance Z_min, electromechanical coefficients Ƙ_p, mechanical quality factor Q_m and dielectric loss tanδ are defined and the value are listed in Table 1.

The variation of ceramic density as a function of sintering temperatures (T_sint: 1000-1250 °C) is shown in Fig. 7. A relative density of 7.53 g/cm³ was achieved at sintering temperature of 1150 °C, which is close to the theoretical density (~95%). At higher T_sint, the density of ceramic samples strongly decreases which can be attributed to the evaporation of excess PbO.

The electromechanical properties of PZT ceramics as a function of sintering temperatures are shown in Fig. 8. Both d₃₃ and Q_m show a similar variation trend with increasing temperatures. The optimized values of d₃₃ (64 pC/N) and Q_m (34.3 pC/N) were obtained 1150 °C. The Ƙ_p value of piezo-ceramics initially increases with increasing T_sint. until it reaches a maximum value of 0.41 at 1150 °C, and then decreases for high temperature, which may be due to the evaporations of lead oxide. In addition, with decreasing T_sint., dielectric constant (ε_r) increased continuously. This phenomenon is common for polycrystalline ferroelectric ceramics. In the past decades, effect of grain size on the relative permittivity has been extensively reported and different theoretical models have been introduced to describe the physical origin of this effect [27]. For example; in BaTiO₃, it is now widely accepted that ε_r increases with decreasing grain size, reaching a value of 5000, or higher, as the grain size approaches 1µm. Below 1µm, however, ε_r of BaTiO₃ decreases markedly with further decreasing grain size [28]. In our work, the maximum value (384) of ε_r was obtained for specimen with mean grain size of 1.84 µm sintered at 1000 °C. Further lowering the sintering temperature to obtain smaller grain size is impossible, because the sample is difficult to be densified.

Fig. 9a demonstrates the variations of the dielectric properties, i.e., dielectric constant (ε_r) and loss tangent (tanδ) as a function of frequency. It can be seen for piezo-ceramic sintered at 1150 °C that both ε_r and tanδ are nearly independent of frequency and measured about 388 and 1.5% at 1 kHz, respectively. The dielectric constants decrease slowly at low frequency regime (Fig. 9b; ε_r < 390), which is may be attributed to the dielectric relaxation of PZT nanoparticles. Besides, the dielectric constants climb markedly with increasing frequency beyond 550 kHz (Fig. 9c). Dielectric relaxation of the ferroelectric ceramics could originate from the alternations of the elastic and electric behavior and the movement of the domain walls at high frequency regions [29].

CONCLUSION

The PZT nanopowders were synthesized by the sol-gel method using acetic acid as a complexing reagent. The PZT nanopowders with spherical-shaped morphology, perovskite structure and an average size of 20 nm are synthesized. The single PZT phase starts to form at a calcination temperature of 500 °C. The optimum piezoelectric parameters are obtained at sintering temperature 1150 °C: d₃₃=64 pC/N, Ƙ_p=0.41 and Q_m=34.3.

CONFLICT OF INTEREST

The authors declare that there are no conflicts of interest regarding the publication of this manuscript.