Giant piezoresponse in nanoporous (Ba,Ca)(Ti,Zr)O3 thin film

Lattice strain effects on the piezoelectric properties of crystalline ferroelectrics have been extensively studied for decades; however, the strain dependence of the piezoelectric properties at nano-level has yet to be investigated. Herein, a new overview of the super-strain of nanoporous polycrystalline ferroelectrics is reported for the first time using a nanoengineered barium calcium zirconium titanate composition (Ba0.85Ca0.15)(Ti0.9Zr0.1)O3 (BCZT). Atomic-level investigations show that the controlled pore wall thickness contributes to highly strained lattice structures that also retain the crystal size at the optimal value (<30 nm), which is the primary contributor to high piezoelectricity. The strain field derived from geometric phase analysis at the atomic level and aberration-corrected high-resolution scanning transmission electron microscopy (STEM) yields of over 30% clearly show theoretical agreement with high piezoelectric properties. The uniqueness of this work is the simplicity of the synthesis; moreover the piezoresponse d33 becomes giant, at around 7500 pm V−1. This response is an order of magnitude greater than that of lead zirconate titanate (PZT), which is known to be the most successful ferroelectric over the past 50 years. This concept utilizing nanoporous BCZT will be highly useful for a promising high-density electrolyte-free dielectric capacitor and generator for energy harvesting in the future.

nm.The ellipsometry characterization confirmed that a film thickness of 170 nm was obtained.The substrate was then annealed at 120 °C for 5 min, calcined at designated temperatures with a ramping rate of 1 °C min -1 , and held for 10 min at the highest temperature.

Materials characterizations.
The surface structures of the composites were observed with fieldemission scanning electron microscopy (FESEM) (Hitachi SU7001 SEM).Images were collected with a low-voltage Hitachi SU7001 SEM at an accelerating voltage of 5 kV and a current of 10 mA.The crystallinity of the composites was evaluated with wide-angle X-ray diffraction (XRD) (Rigaku RINT 2500X).The XRD pattern was collected with monochromated Cu Kα radiation (40 kV and 40 mA) at a scanning rate of 0.5 °C min -1 .Scanning transmission electron microscopy (STEM) for nanoporous BTO and BCZT films was performed using equipped with a Cs corrector in the irradiation and imaging system (JEOL JEM-ARM200F).TEM was equipped with two EDS detectors and a total solid angle was 0.98 Sr at an acceleration voltage of 200 kV.The drift correction function was used in the highresolution EDS measurements.TEM samples were prepared by peeling the films from substrates using a cutter knife and dispersing them in ethanol.Following dispersion, the suspension was dropped onto a Cu mesh with a carbon support film.The doping levels in the nanoporous BCZT film were determined using X-ray photoelectron spectroscopy (XPS) (Kratos AXIS-ULTRA DLD).

S1
Electronic Supplementary Material (ESI) for Chemical Science.This journal is © The Royal Society of Chemistry 2024 PFM amplitude distributions.The amplitude distributions were observed using high-voltage piezoresponse force microscopy (HV-PFM) (Cypher DART, Asylum Research).A heavily doped conductive silicon cantilever with a spring constant of 2.7 N was used.Its medium stiffness with high contact resonance and fairly tight adhesion to the sample surface (essential for compensating local capacitive forces) were found to be the best suited of all the studied materials.The sample deformation becomes at 1 V [1], and the amplitude image of PFM becomes a direct representation of the vertical piezoresponse ( ), as plotted in Figure 5. 33 Strain study.The strain for bulk BCZT was calculated with high-resolution scanning transmission electron microscopy (HR-STEM) (Hitachi HF5000 aberration-corrected STEM).We have taken the geometric phase approach for this purpose, which is based on combining real-space and Fourier-space information.By studying the amplitude of an electron micrograph image, the local contrast of the fringes and their phase represent their positions.Consequently, the strain can be measured by calculating the local Fourier components of the lattice fringes in an image, following Hÿtch et al. [2].
Measurement of dielectric properties.The dielectric constant ( ) was measured using a semiconductor parameter analyzer (Keysight, B1500A Semiconductor Device Parameter Analyzer).
A 0.5 wt.% Nb-doped STO substrate was used as the bottom electrode, and platinum (Pt) was sputtered as the top electrode ( 100 ) on the synthesized film.A DC voltage of 3 mV was applied to the ∅≅  synthesized nanoporous and bulk films with a thickness of 170 nm. was measured at 298 K under the above conditions, and the dielectric loss ( ) was calculated using Equation (S1):

Validity of the strain values.
Here, the strain values of BCZT were calculated to ensure the validity of the measured piezoelectric properties in this study by comparing the strain values estimated from TEM analysis in this study.The strain values were calculated using Equation (S1), where , , E and are the electric field-induced strain, piezoelectric constant, electric field and proportionality factor of strain for the applied voltage, respectively [3].
The value of was determined from the calibration plot (Figure S6) of the effective piezoelectric () sensitivity obtained from an optical beam deflection sensor equipped with PFM and was strongly S2 frequency-dependent and can amplify the contribution to the PFM signal at a frequency near 300 kHz [3].It was estimated that an amplification factor from 10 to 100 might arise subject to the   -1 optimization of the electrostatic contributions, local capacitive forces, and tip bias, as broadly explored for the contact resonance between the sample and tip from 200 to 300 kHz.Therefore, the value of was set to 10.The values of electric field-induced strain were calculated to be 0.003 ( 0.3%) () ≅ ≅ and 0.245 ( 25%) for the bulk and nanoporous films, respectively, using 10 and the ≅ ≅ () = calculated values from the piezoelectric hysteresis loops in Figure 4.They are in good agreement with the values of 0.3% and over 30% for the bulk and nanoporous films estimated from TEM  ≥ ãnalysis, respectively.Therefore, it is considered that the evaluation of piezoelectric properties in this study was reasonable.
Comment on leakage current.The leakage current was measured using a semiconductor parameter analyzer (Keysight, B1500A Semiconductor Device Parameter Analyzer).A Si/5 wt.% Nb-doped STO substrate was used, and titanium (Ti)/Pt was sputtered as the bottom electrode.A voltage of 1 V was applied to the synthesized films with a thickness of 170 nm.The measurement temperature was 298 K. Comment for Figure S5: It is clear that the wide-range XPS spectra demonstrate the existence of predictable Ba, Ca, Zr, Ti, and O peaks (Figure S5a).Ba 3d  S5b).In the current analysis, the C 1s peak centered at 284.85 eV is the reference for calibrating the binding energies of the elements present in the BCZT composition.The asymmetrical high-resolution XPS peak of the O 1s orbital in each spectrum is deconvoluted into three different component peaks (Figure S5c).The peaks are centered at 529.15, 530.26, 531.25, and 532.58 eV for nanopores [4].The peak centered at 529.15 eV corresponds to lattice oxygen ions (O 2-) in the perovskite structure, whereas the peak centered at 530.26 eV can be attributed to the O 2-/O -ions caused by oxygen vacancies [5].The peak centered at 531.25 eV is due to chemisorbed oxygen on the thin film surface, as reported earlier by Tu et al. [6] and Usman et al. [7] for ZnO nanorods and ZnO thin films, respectively.The satellite peak at 532.58 eV may be due to the overlap of orbitals.However, a slight shift in the binding energy for the O 1s spectral components in the BCZT thin films occurred due to different chemical environments and may also be due to charging during XPS experiments [8].The high-resolution XPS spectra of the Ba 3d orbital exhibit two peaks at 709.12 eV and 693.6 eV, attributed to Ba 3d 5/2 and Ba 3d 3/2 , respectively (Figure S5d), whereas the Ti 2p spectra exhibit two peaks at 457.75 and 463.5 eV, attributed to Ti 4+ 2p 3/2 and Ti 4+ 2p 1/2 , respectively (Figure S5e) [9].Moreover, Zr 3d exhibits two peaks at 181.25 and 183.75 eV, attributed to Zr 3d 5/2 and Zr 3d 3/2, respectively (Figure 5f).Ca 2p exhibits two peaks at 346.52 and 350.22 eV, attributed to Ca 2p 3/2 and Ca 2p 1/2 , respectively (Figure S5g).with increasing frequency, although that for the nanoporous BCZT film shows an abrupt decrease above approximately 0.3 MHz.This is because, as frequency increases, the wavelength size approaches the size of the 'grains' the medium can no longer be represented as homogenous.Each transition of the wave between particles becomes a boundary condition, leading to dispersion and loss in the waves due to slight variations in the dielectric parameters of each particle.The values for the   nanoporous film are 1.5 times larger than those for the bulk film up to 0.3 MHz.This higher value   is attributed to the strain induced by nanopores [11,12].On the other hand, the values for both tan  nanoporous and bulk films are almost the same at approximately 0.03 to 0.04 in the frequency range from 0.001 to 0.01 MHz, whereas they show an increase above 0.01 MHz.In addition, the value tan  for the nanoporous BCZT film shows a more gradual increase than that for the bulk BCZT film.It is approximately 0.12 at 1 MHz, which is approximately half that of the bulk film.Therefore, higher   and lower for the nanoporous film than for the bulk film are observed over this frequency range.tan  nanoporous BCZT films, respectively (Figure S8a).It is confirmed that the current density increases linearly with the applied voltage for both in the voltage range up to 0.05 V, although the current density for the nanoporous BCZT film is larger than that for the bulk film (Figure S8b).In our previous work, the current density of an STO/BTO hybrid single layer [12] was found to be A while 10 -5  -2 maintaining a high dielectric constant.In a later work [13] on an STO/BTO multilayer stack, the current density was significantly reduced to A , but no significant ferroelectric properties 10 -11  -2 were found.In the same context, a recent study by Mohamed et al. reported the synthesis of mesostructured HfO 2 /Al 2 O 3 composite thin films with a reduced leakage current of A at 10 -9  -2 1 V [14].It is thought that the leakage current improvement may be caused by the effective stress at the interfaces between the mesoporous HfO 2 and Al 2 O 3 domains, consequently affecting the stability of the mesostructured composite thin film.On the other hand, our nanoporous BCZT film reported here has an adequately low leakage current for use as a dielectric material.From the results of the dielectric properties and the leakage current measurements, a large dielectric constant is confirmed, and an adequately low leakage current is fit for practical use.Therefore, it is considered that nanoporous BCZT has more potential as a dielectric capacitor than bulk BCZT.

Figure S8 .
Figure S8.Leakage current.(a) J-V curves for bulk and nanoporous BCZT prepared on an Si/5 wt.% Nb-doped STO substrate at 650 .(b) Magnified graph at a very low voltage marked in red in (a).℃ .63, 2.48, 9.14, 0.72, and 50.97 mol% of the synthesized composite are observed, which confirms successful doping.The above calculation of the stoichiometric composition from the XPS result of nanoporous Ba 0.84 Ca 0.16 (Ti 0.93 Zr 0.07 )O 3 clearly fits with the nominal composition of nanoporous Ba 0.85 Ca 0.15 (Ti 0.9 Zr 0.1 )O 3 synthesized in this study.The XPS spectrum of C 1s is generally considered to indicate surface contamination of the synthesized film (Figure , Ca 2p, Ti 2p, Zr 3p, and O 1s peaks with S9 compositions of 11