Sm-doped PZT thin film with high piezoelectric properties by sol-gel method

Lead zirconate titanate (PZT) is widely used in the fields of microelectronics, sensors, actuators, transducers, harvesting devices, and semiconductors due to its outstanding piezoelectric and ferroelectric properties.^1–5 Rare-earth-doped modified PZT materials have attracted considerable research interest in recent years.^6–11 Rare-earth ions, with their unique ionic radius and different chemical valence states, can replace the original ions and create vacancies, leading to lattice distortion and, consequently, affecting the performance of PZT films.¹² In particular, the substitution of Sm (samarium) ions in lead-based ferroelectric materials has exhibited unexpectedly high piezoelectric responses, attracting significant attention in recent studies.^13–17 

Currently, the research on Sm-doped PZT materials primarily focuses on piezoelectric ceramics, with relatively little investigation on Sm-doped PZT films. PZT films, as more suitable piezoelectric and ferroelectric materials for miniaturization and integration, offer advantages such as high piezoelectric coefficients, outstanding ferroelectric properties, and enhanced sensing capabilities. Consequently, they have received widespread attention in applications for non-volatile memory devices and MEMS piezoelectric devices.^18–21 Therefore, investigating the preparation and properties of Sm-doped PZT films is of significant importance for developing high-performance piezoelectric and ferroelectric devices. However, the piezoelectric, dielectric, and ferroelectric properties and crystal growth mechanism of Sm-doped PZT films have not been systematically studied so far, and the performance of the films requires further improvement.

In this paper, the sol-gel method was chosen as the method to prepare Sm-doped PZT films. As a simple, low-cost, and efficient coating technique, the sol-gel method can produce uniform and high-quality films on various substrates.²² Moreover, this method enables precise control over the film's composition and facilitates the fabrication of films with multiple and complex components. Therefore, it stands as an excellent choice for producing doped PZT thin films.²³ This study employed the sol-gel method to prepare Sm-doped PZT films with varying doping concentrations of 0, 0.5, 1, 1.5, 2, and 3 mol. %. These films were characterized by assessing their surface morphology, density, crystal structure, crystal orientation, piezoelectric performance, dielectric properties, and ferroelectric properties. The objective of this study is to investigate the impact of Sm doping on the properties of PZT films and to significantly enhance the performance of Sm-doped PZT thin films.

The preparation and thermal treatment processes of Sm-PZT films are shown in Fig. 1. Initially, solution A was prepared by combining lead acetate trihydrate (Pb(CH₃COO)₂⋅3H₂O) and samarium acetate hydrate (C₆H₉O₆Sm⋅xH₂O) in 2-methoxyethanol (C₃H₈O₂). More than 10 mol. % of lead acetate was added during the preparation process to mitigate losses due to lead evaporation during sintering. Following this, zirconium n-propoxide (C₁₂H₂₈O₄Zr) and titanium isopropoxide (C₁₂H₂₈O₄Ti) were added to 2-methoxyethanol (C₃H₈O₂), and formamide (CH₃NO) was included to prevent film cracks, thereby forming solution B. Subsequently, solution B was combined with solution A. After the solution cooled to room temperature, it was filtered to obtain a transparent, stable, and precipitate-free Sm-PZT solution. Finally, the solution was allowed to stand for 24 h to further enhance its stability.

A 200 nm thick SiO₂ layer was grown on a Si (100) substrate by wet oxidation. Next, 50 nm Ti and 200 nm Pt layers were deposited by magnetron sputtering to form the substrate for PZT film growth. The Sm-PZT sol was spin-coated on the Pt (200 nm)/Ti (50 nm)/SiO₂ (200 nm)/Si (100) substrate at 3000 rpm for 40 s. Then, the film was preheated on a hot plate at 250 °C for 10 min to remove the organic solvent. Afterward, the film was heated at 550 °C for 15 min. After 15 cycles of coating and heat treatment, the film was annealed at 600 °C for 30 min to fully crystallize the film. PZT films doped with Sm at concentrations of 0, 0.5, 1, 1.5, 2, and 3 mol. % were prepared by repeating this procedure.

Piezoresponse Force Microscopy (PFM) is an excellent choice for testing the piezoelectric coefficient of thin films.²⁴ After the prepared Sm-PZT films were cut into small pieces, a portion of it underwent treatment using the photolithography-etching method to selectively remove sections of the Sm-PZT films, thereby exposing the bottom electrode for subsequent Scanning Probe Microscopy (SPM) testing. As for another portion of the completed cut Sm-PZT films, to perform dielectric performance testing and the characterization of polarization–electric field (P–E) hysteresis loops, the processing steps are illustrated in Fig. 2. The photoresist coated on the PZT films was patterned using standard photolithography development techniques, followed by ion beam sputtering deposition of Au (100 nm)/Cr (50 nm) films. Then, square top electrodes with a side length of 1.2 mm were fabricated by lift-off process. Subsequently, the photolithography-etching method was employed to remove sections of the PZT layer, exposing the Pt bottom electrode. Electrical connections were made using silicon-aluminum wires through ultrasonic bonding techniques.

The crystal structure and orientation of Sm-PZT films were analyzed by X-ray Diffraction (XRD) (λ = 1.5406 Å, 10° ≤ 2θ ≤ 60°). Surface morphology and cross-sectional images of the films were characterized by Scanning Electron Microscope (SEM) (Nova NanoSEM 650, FEI), and the thickness of the films was obtained. The Atomic Force Microscopy (AFM) module in the Scanning Probe Microscope (SPM, Dimension FastScan, Bruker) further characterized the surface morphology and root mean square roughness (R_q) of the films, and the Piezoresponse Force Microscopy (PFM) module was used to measure the piezoelectric properties of the films. The dielectric properties of the films within the frequency range of 40 Hz–100 kHz were analyzed using an impedance analyzer (4294A, Agilent) at room temperature. A standard ferroelectric test system (Radiant Technologies) was used to evaluate the polarization–electric field (P–E) hysteresis loops.

The XRD results of (Pb_1−xSm_x)(Zr_0.54Ti_0.46)O₃ (x = 0, 0.5, 1, 1.5, 2, 3 mol. %) films are shown in Fig. 3, where no impurity phases were observed in any samples, indicating good crystallinity. In both undoped (0 mol. %) and lightly doped (0.5 mol. %) PZT films, a distinct preference for the (110) plane was observed. As the doping concentration increased, at Sm doping concentrations of 1 and 1.5 mol. %, the (100) peak gradually increased, while the (200) peak showed asymmetry or slight splitting, indicating that the Sm-PZT films were in the coexistence of rhombohedral and tetragonal phases and were in the morphotropic phase boundary (MPB) region.^25,26 Further increasing the doping concentration (2, 3 mol. %) led to the diminished observability of the (100) peak in the films.

Figures 5(a)–5(f) show the surface morphology and cross-sectional profiles of Sm-PZT films with different doping concentrations of 0, 0.5, 1, 1.5, 2, and 3 mol. %. The surface morphologies of all the Sm-PZT films exhibit clear grain structures, with smooth surfaces and no apparent cracks. Table I provides the average grain size and film thickness of the Sm-PZT films. The surface morphologies of all the Sm-PZT films exhibit clear grain structures, with smooth surfaces and no apparent cracks. Nanostructures appear on the surface of all Sm-PZT films. Previous studies have indicated that in sol-gel synthesis, the surface of PZT films with an excess of 10% lead exhibits nanostructures. Moreover, this surface phase is not pyrochlore but in fact consists of zirconium oxide nanostructures.³² Our XRD results also confirm this, showing no pyrochlore phase in any of the Sm-PZT films.

Figures 6(a)–6(f) display the grain size distribution of Sm-PZT films with different doping concentrations of 0, 0.5, 1, 1.5, 2, and 3 mol. %. The average grain size of Sm-PZT films is shown in Fig. 7. A previously reported study on Sm-doped PZT ceramics showed that the introduction of Sm inhibited the grain growth of PZT ceramics.¹³ However, Sm-doped PZT films were different. The results indicated that all Sm-PZT films exhibit a significant increase in the grain size compared to undoped PZT films. A possible reason for this difference is the effect of the film interface and growth substrate on the growth of PZT crystals. Figure 7 shows that the grain size of Sm-PZT films increases first and then decreases with the increase in the Sm doping concentration. This suggests that moderate Sm doping promotes the growth of PZT film grains. Excessive doping leads to the gradual accumulation of Sm at grain boundaries, restraining the growth of PZT film grains. Nevertheless, the grain size remains larger than that of undoped films.

The cross-sectional SEM results show that the undoped and low-doped (0.5 mol. %) PZT films exhibit more pores, while highly doped Sm-PZT films present fewer pores. This further confirms that in the state of films, Sm can promote crystal growth, enhance crystal density, and reduce porosity. The layered structure of Sm-PZT films is clearly visible and also exhibits distinct features. For undoped PZT films or those with lower doping concentrations (0, 0.5, and 1 mol. %), distinct layering is evident and it becomes more pronounced with increasing doping concentration. However, as the doping concentration further increases to 1.5, 2, and 3 mol. %, the layered structure of the PZT films gradually weakens, eventually no longer exhibits distinct stratified characteristics.

Figures 8(a)–8(f) show the 3D AFM images of Sm-PZT films with doping concentrations of 0, 0.5, 1, 1.5, 2, and 3 mol. % in the scanning area of 10 × 10 μm², respectively. The surface roughness Rq (root mean square roughness) of the films is presented in Table I. Figure 9 illustrates the surface roughness of Sm-PZT films with respect to doping concentrations, indicating an initial increase followed by a decrease with rising doping concentration, consistent with the trend observed in Sm-PZT's average grain size. For PZT films with doping concentration below 1.5 mol. %, the addition of Sm promoted crystal growth, resulting in a gradual increase in the grain size and subsequently increasing the surface roughness. However, as the doping concentration further increased, Sm accumulated at the grain boundaries, inhibiting further grain growth and, consequently, leading to a reduction in the surface roughness.

The film was polarized in a 10 × 10 μm² region using a + 15 V direct current bias through the PFM tip. Figures 10(a)–10(f) illustrate the PFM test results for Sm-PZT films with doping concentrations of 0, 0.5, 1, 1.5, 2, and 3 mol. %. The piezoelectric coefficients d₃₃ for the respective films were obtained by fitting the slope of the piezoresponse amplitude and the voltage. Figure 11 shows the piezoelectric coefficients d₃₃ for various doping concentrations of Sm-PZT films. The research findings demonstrate that Sm doping significantly enhances the piezoelectric coefficient d₃₃ of PZT films, and there exists an optimum doping concentration that maximizes the enhancement of the film's piezoelectric properties. In particular, the film with a doping concentration of 1.5 mol. % exhibits a high piezoelectric coefficient d₃₃, reaching 279.87 pm/V. This significantly exceeds the piezoelectric coefficient d₃₃ reported in previous studies for PZT films.^33–42 Compared to the undoped PZT with 61.45 pm/V, the piezoelectric coefficient d₃₃ is 4.55 times of it. Moreover, the film with 1.5 mol. % Sm-doped was in the morphotropic phase boundary (MPB) range, with the largest average grain size and the optimal (100) orientation based on the previous analysis.

Sm doping acts as a donor dopant where samarium ions replace Pb (lead) ions, creating lead vacancies. The XRD results show that the crystal orientation of Sm-PZT changes significantly under different doping concentrations. Specifically, the (100) peak is more prominent at doping concentrations of 1 and 1.5 mol. %, resulting in higher piezoelectric performance. This suggests that within a specific doping range, the Pb vacancies and lattice distortions caused by doping induce the crystal growth along the (100) crystal orientation, consequently enhancing the piezoelectric properties. For Sm-PZT films with similar crystal orientations (0, 0.5, 2, 3 mol. % doping), the grain size is the main factor influencing their piezoelectric performance, and the piezoelectric performance increases with the grain size.

Figure 12 illustrates the variation in the dielectric constant (ɛ) and loss tangent (tan δ) of Sm-PZT films at room temperature within the frequency range of 40 Hz–100 kHz. With an increase in the frequency, the dielectric loss of all films increases, while the dielectric constant (ɛ) decreases. Figure 13 shows the dielectric constant (ɛ) and loss tangent (tan δ) of Sm-PZT films with different doping concentrations at a frequency of 1 kHz. It can be seen that samarium doping causes an increase in the dielectric loss of the PZT films. The undoped and 0.5 mol. % doped films have comparatively lower dielectric losses. With a further increase in the doping concentration (1, 1.5, 2, 3 mol. %), the dielectric losses increase notably. SEM cross-sectional images indicate that the samples with doping concentrations of 2 and 3 mol. % exhibit almost no pores and relatively fewer defects. Therefore, compared to Sm-PZT films with doping concentrations of 1 and 1.5 mol. %, these two samples exhibit lower dielectric losses. Compared to other films with higher doping concentrations (1.5, 2, 3 mol. %), the film with a doping concentration of 1 mol. % has smaller grain sizes and the most distinct stratified structure. Smaller grain sizes lead to increased grain boundary density, and both grain boundary density and stratified structure serve as leakage paths, consequently causing an increase in dielectric losses.

Sm is a type of donor dopant, and research findings indicate that donor doping can enhance the dielectric constant of PZT films, which is in agreement with our experimental results.^7,13,43 The Sm-PZT film shows the optimal dielectric constant at a doping concentration of 1 mol. %. The reason for the lower dielectric constant in lightly doped (0.5 mol. %) and undoped (0 mol. %) Sm-PZT films is that, compared to highly doping concentrations of Sm-PZT films, they exhibit lower density and more porosity, as shown in the SEM cross-sectional images.

Figure 14 illustrates the polarization–electric field (P–E) hysteresis loops of Sm-PZT films. The remnant polarization (P_r) and coercive field (E_c) of Sm-PZT films are shown in Fig. 15. At an electric field of 550 kV/cm, all Sm-PZT films exhibit a significant increase in remnant polarization compared to the undoped PZT film. In particular, the Sm-PZT sample with a doping concentration of 1.5 mol. % exhibits a remarkable P_r of up to 25.7 μC/cm². This exceeds the P_r reported in previous studies for rare-earth-doped PZT films.^44–47 The trend of P_r variation in different Sm-PZT films aligns with the trend of piezoelectric coefficients d₃₃. Previous studies on PZT materials have shown a significant correlation between remnant polarization and piezoelectric performance,⁴⁸ and our research findings are fully consistent with this observation. Furthermore, with an increase in the doping concentration, there is a certain degree of decrease observed in the coercive field. In general, the coercive field of PZT films with 1–3 mol. % doping concentration is lower than that of undoped samples. This suggests that Sm doping can somewhat reduce the resistance to domain reversal, which lowers the value of the coercive field.

Sm-PZT films with various doping concentrations (0, 0.5, 1, 1.5, 2, 3 mol. %) were prepared by the sol-gel method. The effects of different Sm doping concentrations on the surface morphology, density, crystal structure, piezoelectric, dielectric, and ferroelectric properties of PZT films were systematically studied. The results showed that unlike Sm-doped ceramics, in the case of films, Sm doping promoted the growth of PZT grains, which might be attributed to the film interface and growth substrate on the crystal growth of PZT. Sm doping affected the surface roughness R_q and density of the films, as well as the crystal orientation of the PZT films. Moreover, Sm doping significantly enhanced the piezoelectric coefficient d₃₃ and remnant polarization of PZT films while also decreased the coercive field. In addition, Sm doping raised the dielectric constant of the PZT films, although it caused an increase in dielectric loss. In particular, the Sm-PZT sample with a doping concentration of 1.5 mol. % exhibited the optimal (100) preferred orientation among all the Sm-PZT films, with a maximum average grain size of 163.76 nm. It has a dielectric constant of 1343.55, with a remarkable remnant polarization of 25.7 μC/cm² under an electric field of 550 kV/cm, and a high piezoelectric coefficient d₃₃ of 279.87 pm/V, which is 4.55 times that of the non-doped PZT film. These findings reveal the promising prospects of Sm-doped PZT films for various applications related to piezoelectric and ferroelectric devices.

This work was supported by the National Natural Science Foundation of China (Grant Nos. 12174420 and 11874388); the National Key Research and Development Program of China (Grant Nos. 2023YFF0716500, and 2023CSJZN0200); the Goal-oriented project independently Deployed by Institute of Acoustics, Chinese Academy of Sciences (Grant No. MBDX202112); and the Frontier Exploration Project Independently Deployed by Institute of Acoustics, Chinese Academy of Sciences (Grant No. QYTS202002).

The authors have no conflicts to disclose.

Jinming Ti: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Software (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Junhong Li: Conceptualization (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Supervision (equal); Writing – review & editing (equal). Qingqing Fan: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Visualization (equal). Qing Yu: Investigation (equal); Validation (equal). Yuhan Ren: Investigation (equal); Visualization (equal). Chenghao Wang: Methodology (equal); Resources (equal); Supervision (equal).

The data that support the findings of this study are available from the corresponding author.