Ferroelectricity in CsPb₂Nb₃O₁₀ and exfoliated 2D nanosheets

Abstract

Pb-based perovskites play pivotal roles in ferroelectric research. In the search for new Pb-based ferroelectrics, we investigated the ferroelectric properties of Dion–Jacobson type CsPb₂Nb₃O₁₀ and exfoliated 2D nanosheets. Ferroelectricity in CsPb₂Nb₃O₁₀ was demonstrated for the first time. CsPb₂Nb₃O₁₀ adopted a polar tetragonal structure with a modest T_C = 260 °C and polarization P_S = 7.93 μC cm⁻²; the polarization mainly arose from the out-of-plane displacements of Nb⁵⁺ ions and nearby oxygens. CsPb₂Nb₃O₁₀ layered perovskite offers additional advantages for tailoring ferroelectric nanomaterials, as exfoliated 2D nanosheets provide novel platforms for investigating ferroelectric properties down to the 2D limit. Piezoresponse force microscopy confirmed stable ferroelectricity even in exfoliated 2D Pb₂Nb₃O₁₀ nanosheets.

---

Introduction

Lead (Pb) is a magic element for ferroelectric materials. Pb-based perovskites have always played pivotal roles in the history of ferroelectric science and technology. One of the most important materials is PbTiO₃, where ferroelectric polarization is linked to the off-centering of Ti⁴⁺ ions in TiO₆ octahedra due to so-called second-order Jahn–Teller effects.^1,2 In PbTiO₃, Pb²⁺ ions on the A site also contribute to this off-centering, influenced by the unique characteristics of 6s² lone-pair electrons.^3,4 These lone-pair electrons on Pb²⁺ ions play a significant role in the structural and physical properties of Pb-based perovskites; the strong covalent nature of Pb–O bonds and steric hindrances of Pb²⁺ ions contribute to a large polarization with a high Curie temperature (T_C) of 490 °C.^5–7

Similar lone-pair effects are expected in layered perovskites containing Bi³⁺ and Pb²⁺ ions with lone-pair electrons. Significant efforts have been devoted to developing new ferroelectric materials, focusing on Bi-based layered perovskites with Aurivillius (AU) structures (e.g., Bi₄Ti₃O₁₂)^8,9 and Dion–Jacobson (DJ) structures (e.g., RbBiNb₂O₇, CsBi₂Ti₂NbO₁₀)^10,11 as primary targets for Pb-free high-T_C ferroelectrics due to environmental concerns. While Pb-based compounds remain important targets, Pb-based layered perovskites have been rarely synthesized to date. Experimental confirmation of ferroelectricity has been achieved in AU-type Bi-layered perovskites (PbBi₂Ta₂O₉, PbBi₄Ti₄O₁₅, and Pb₂Bi₄Ti₅O₁₈).^12–14

Our primary concern is that new solutions can be found in the DJ-type layered perovskite CsPb₂Nb₃O₁₀, which has been first reported in 1988 (ref. 15). Although pioneering studies on dielectric properties appeared in 1999 (ref. 16), this compound has neither been characterized nor documented in the ferroelectric database. Furthermore, DJ-type layered perovskites offer additional advantages for ferroelectric research, as exfoliated 2D nanosheets provide novel platforms for investigating ferroelectric properties down to the 2D limit. While the exfoliation of Rb-analogue (RbPb₂Nb₃O₁₀) has been reported by several groups,^17–19 previous research on Pb₂Nb₃O₁₀ nanosheets has primarily focused on photo-functionalities such as solar cells and photocatalysts. Exfoliated Pb₂Nb₃O₁₀ nanosheets have not yet been explored as 2D ferroelectrics. If CsPb₂Nb₃O₁₀ and its exfoliated 2D nanosheets exhibit ferroelectricity, how does the reduced dimensionality affect their ferroelectric properties? These questions are particularly intriguing from both fundamental and practical perspectives, as CsPb₂Nb₃O₁₀ could potentially illuminate new properties of Pb-based perovskites.

Here, we present the characterization of the structural and ferroelectric properties of CsPb₂Nb₃O₁₀. Our findings demonstrate that CsPb₂Nb₃O₁₀ exhibits ferroelectric behaviour with a moderate Curie temperature (T_C) of 260 °C and a spontaneous polarization (P_S) of 7.93 μC cm⁻². Piezoresponse force microscopy (PFM) reveals that the ferroelectric properties remain stable even in exfoliated 2D nanosheets.

Results and discussion

Synthesis and structural characterizations of CsPb₂Nb₃O₁₀

CsPb₂Nb₃O₁₀ was synthesized via solid-state reaction at 1000 °C for 24 h in air (see Experimental in ESI†). Powder X-ray diffraction (XRD) confirmed that CsPb₂Nb₃O₁₀ crystallized in the tetragonal system (Fig. 1a and b); detailed structural features will be discussed later. The formation of CsPb₂Nb₃O₁₀ was further verified using field emission scanning electron microscopy (FE-SEM) equipped with energy-dispersed X-ray spectroscopy (EDS) (Fig. S1, ESI†). FE-SEM observations revealed a plate morphology characteristic of layered perovskites. EDS analysis confirmed the composition of the product as Cs_1.03Pb_2.20Nb_3.00, consistent with CsPb₂Nb₃O₁₀.

Fig. 1 Structural features of CsPb₂Nb₃O₁₀. (a) Rietveld refinement profiles of the powder sample. The parameters are used for characterizing the quality of refinement including the weighted profile residual (R_wp), profile residual (R_p), goodness of fit (S), Bragg residual (R_B), and the structure parameter (R_F). (b) Crystal structure defined by the single-crystal XRD data and Rietveld refinement. (c) Contributions to the net polarization of each ion. The calculated polarization P_S was 7.93 μC cm⁻² along the c axis.

To investigate the detailed structure, single-crystal XRD measurements were performed on CsPb₂Nb₃O₁₀. A single crystal with dimension of 0.02 × 0.16 × 0.29 mm³ was selected for the XRD analysis. The crystal structure was characterized by the direct method with the SHELXT software. The XRD pattern was successfully refined with P-lattice symmetry, and possible space groups included P4/mmm, P4mm, P4̄m2, P422 or P4̄2m. Second harmonic generation (SHG) measurements (Fig. S2, ESI†) indicated that CsPb₂Nb₃O₁₀ exhibited SHG activity, suggesting a polar non-centrosymmetric structure (Fig. 1b). Considering the polar nature, CsPb₂Nb₃O₁₀ adopted a tetragonal system with the polar space group P4mm (no. 99).

We calculated the polarization (P_S) value based on the structural data (Fig. 1b) using the formula:^11,20

where m_i is the site multiplicity, Δz_i is the atomic displacement along the c-axis from the corresponding position in the ideal tetragonal structure, Q_ie is the ionic charge for the i^th constituent ion, and V is the volume of the unit cell. For CsPb₂Nb₃O₁₀, the calculated polarization P_S was 7.93 μC cm⁻². This value was modest compared to DJ-type CsBi₂Ti₂NbO₁₀ (39–49 μC cm⁻²) with triple perovskite layers.¹¹ However, in CsPb₂Nb₃O₁₀, the predominant polarization direction was out-of-plane, unlike the in-plane polarization observed in other layered perovskites containing Bi³⁺ and Pb²⁺ ions. We also analyzed the contribution of each ion to the net polarization (Fig. 1c). The main polarization stemmed from the outer Nb⁵⁺ ions and adjacent oxygens (O(4,5) and O(6,7) sites), while the contribution from Pb²⁺ ions was minimal. This differed from observations in Bi₄Ti₃O₁₂ and CsBi₂Ti₂NbO₁₀, where contributions from Bi³⁺ ions were dominant.

Ferroelectric properties of CsPb₂Nb₃O₁₀

To investigate the dielectric and ferroelectric properties, we performed electrical characterization using a ceramic pellet of CsPb₂Nb₃O₁₀ (Fig. 2). Polarization–electric field (P–E) measurements confirmed a well-saturated ferroelectric hysteresis loop at room temperature (Fig. 2a, inset). The remanent polarization (P_r) value was 4 μC cm⁻², consistent with the P_S value (7.93 μC cm⁻²) calculated from the structural data. Clear polarization switching behavior was also observed in piezoelectric response measurements, with a bipolar strain curve displaying a characteristic butterfly strain loop (Fig. S3, ESI†). The piezoelectric coefficient was estimated from the formula d₃₃ = S_max/E_max, where E_max and S_max are the maximum applied electric field and corresponding unipolar strain output the piezoelectric coefficient, respectively. The piezoelectric coefficient (d₃₃) was determined to be 14.3 pm V⁻¹, a value relatively modest compared to simple perovskites like PbTiO₃. The temperature dependence of the dielectric constant (ε_r) in CsPb₂Nb₃O₁₀ was examined to elucidate its phase transition behavior (Fig. 2a). The dielectric constant exhibited a transition with a broad maximum occurring at 260 °C, where the T_C value was estimated. The ε_r–T curves showed no clear frequency dependence, similar to behaviors observed in conventional ferroelectrics.

Fig. 2 Ferroelectric properties of CsPb₂Nb₃O₁₀. (a) ε_r–T curves measured at 1 and 10 kHz. (Inset) P–E loop at 25 °C. (b) Temperature dependence of low-frequency Raman spectra. (c) Relationship between the polarization and the T_C value for CsPb₂Nb₃O₁₀ and previously reported Bi and Pb-based layered perovskites.^{9–14,25–29} The symbols represent AU (squares) and DJ-type perovskites (circles), respectively.

To gain a deeper insight into the structural origin of the observed ferroelectric responses, we measured the temperature-dependent Raman spectra of CsPb₂Nb₃O₁₀ (Fig. 2b). At room temperature, CsPb₂Nb₃O₁₀ exhibited intense phonon modes (at 40–80, 180–260, ∼580, ∼700, and ∼960 cm⁻¹) and many weak features (Fig. S4, ESI†). These modes and general spectral features were similar to those observed in other layered perovskites.^21–24 In CsPb₂Nb₃O₁₀, the Raman modes of CsPb₂Nb₃O₁₀ could be classified as lattice vibrations involving the cations (<200 cm⁻¹) and internal modes of the NbO₆ octahedra (>200 cm⁻¹). In this study, we discuss low-frequency modes related to polar displacements. Upon heating, some distinctive changes were observed in the low-frequency region, particularly for the low-frequency mode (at 65 cm⁻¹). The 65 cm⁻¹ mode exhibited gradual softening upon heating and disappeared at ∼260 °C. These changes resembled the so-called “soft-mode” behaviour observed in a typical displacive-type ferroelectric transition (proper ferroelectricity). Since the polarization mainly originated from the out-of-plain displacements of Nb⁵⁺ ions and nearby oxygens (O(4,5) and O(6,7) sites), the soft mode was strongly coupled with polar displacements of Nb⁵⁺ ions.

In Fig. 2c, we compare the T_C and P_S values among CsPb₂Nb₃O₁₀ and previously reported Bi and Pb-based layered perovskites.^{9–14,25–29} As documented in previous literature, PbTiO₃ exhibits both a high P_S value (75 μC cm⁻²) and a high T_C (490 °C) due to the strong covalent nature of Pb–O bonds and steric hindrances of Pb²⁺ ions. Triple-layered perovskites such as Bi₄Ti₃O₁₂ and CsBi₂Ti₂NbO₁₀ also show similar properties with high P_S and T_C value. In contrast, CsPb₂Nb₃O₁₀ was quite unique; it exhibited a modest T_C of 260 °C and polarization P_S of 7.93 μC cm⁻² possibly due to the negligible contribution of Pb²⁺ ions to the polarization. Regarding CsPb₂Nb₃O₁₀, a distinctive feature was its insensitivity of the lowest-frequency mode at 40 cm⁻¹ (Fig. 2b). In typical ferroelectrics like Bi₄Ti₃O₁₂, the lowest-frequency mode strongly correlates with polar displacements of the heaviest Bi³⁺ ions.^10,30 From the mass consideration of phonon modes, the lowest-frequency mode at 40 cm⁻¹ would normally be expected to couple with the heaviest Pb²⁺ ions. The insensitivity of the Pb mode is indicative of weak lone-pair effects, consistent with the modestly high P_S and T_C values observed.

Ferroelectric properties of exfoliated 2D nanosheets

The exfoliation reaction of DJ compounds has research significance because exfoliated perovskite nanosheets offer novel platforms for investigating ferroelectric properties down to 2D limit.^31,32 Having established the ferroelectric CsPb₂Nb₃O₁₀, we synthesized Pb₂Nb₃O₁₀ nanosheets via an intercalation-based exfoliation method involving acid treatment and bulky guest intercalation^33,34 (Fig. S5, ESI†). Through acid treatment, the Cs⁺ ions in CsPb₂Nb₃O₁₀ were replaced by H⁺ ions, resulting in the formation of a protonic phase (HPb₂Nb₃O₁₀) while maintaining the layer structure. From the EDS analysis, the composition of the product was Pb_2.06Nb_3.00, indicating the success of cation exchange (Fig. S6, ESI†). Intercalation of tetrabutylammonium (TBA⁺) ions to the protonated compound resulted in total delamination into Pb₂Nb₃O₁₀ nanosheets (Fig. 3a).

Fig. 3 Characterization of Pb₂Nb₃O₁₀ nanosheet. (a) Structural model of Pb₂Nb₃O₁₀ nanosheet. (b) AFM image taken from an individual nanosheet. (c) TEM image and (inset) SAED from an individual nanosheet. (d) EDS spectrum taken from the selected point in (c).

The morphological and structural features of Pb₂Nb₃O₁₀ nanosheets were investigated using atomic force microscopy (AFM) and transmission electron microscopy (TEM). AFM images revealed a sheet-like morphology with a thickness of 1.8 nm and lateral dimensions of ∼1 μm (Fig. 3b). This thickness corresponded to the crystallographic thickness of the Pb₂Nb₃O₁₀ nanosheet, with slight differences due to adsorbed TBA⁺/H⁺ ions and water molecules. TEM analysis confirmed the high crystallinity of the 2D nanosheets (Fig. 3c). The selected area electron diffraction (SAED) pattern (Fig. 3c, inset) displayed a single-crystal-like pattern with distinct sharp spots, which can be indexed to the square cell structure characteristic of the 2D host layer in the parent CsPb₂Nb₃O₁₀. The successful formation of Pb₂Nb₃O₁₀ nanosheets was further validated by EDS (Fig. 3d), demonstrating that the composition was retained throughout the exfoliation process. These findings confirm the successful synthesis of Pb₂Nb₃O₁₀ nanosheets.

For the ferroelectric characterization, we conducted piezoresponse force microscopy (PFM) measurements on individual Pb₂Nb₃O₁₀ nanosheets (Fig. 4).³⁵ The nanosheets displayed a typical butterfly loop in the amplitude–voltage curve (Fig. 4b). The piezoelectric coefficient for Pb₂Nb₃O₁₀ nanosheets was 5 pm V⁻¹, which was smaller than that of 3D bulk CsPb₂Nb₃O₁₀ (13 pm V⁻¹) (Fig. S7, ESI†). In the phase–voltage curve, we observed 120° phase flipping at coercive voltages of ±2 V (Fig. 4c), indicating the robust ferroelectric switching behaviour of the Pb₂Nb₃O₁₀ nanosheets. Furthermore, both the amplitude and phase curves exhibited identical coercive voltages (±2 V), confirming the ferroelectric nature of the Pb₂Nb₃O₁₀ nanosheets.

Fig. 4 PFM measurements of Pb₂Nb₃O₁₀ nanosheet. (a) AFM image of a Pb₂Nb₃O₁₀ nanosheet on a Pt substrate. (b) PFM amplitudes and (c) phase signals from the nanosheet and the Pt substrate. (d) PFM lithography taken after writing a box-in-box pattern with reversed DC bias.

To further investigate the switching behaviour, we performed PFM lithography of the Pb₂Nb₃O₁₀ nanosheets was taken after writing a box-in-box pattern with reversed DC bias (Fig. 4d). The PFM image showed a clear reversal of phase contrast, indicating the polarization switching. Ferroelectric properties can often be affected by leakage current. However, the Pb₂Nb₃O₁₀ exhibited highly insulating behaviour, indicating that its ferroelectric response was not a consequence of leakage current. Notably, the Pb₂Nb₃O₁₀ nanosheet showed strong dielectric endurance with withstand voltages of 4.0 V and −4.5 V (Fig. S8, ESI†), attributed to its 2D size effects. This highly insulating nature is favorable for maintaining stable ferroelectricity even in Pb₂Nb₃O₁₀ nanosheets.

For understanding the impact of dimensional reduction on ferroelectric properties, we performed additional PFM measurements of samples with different thicknesses (Fig. S9, ESI†). By comparing the PFM amplitude of samples with different thicknesses, the PFM amplitude reduced with the layer thickness, consistent with size effect commonly found in ferroelectric thin films.^36,37 The critical thickness for ferroelectricity in perovskite materials has been thoroughly explored using different experimental methods, such PFM, TEM, Raman spectroscopy, and synchrotron XRD. Research indicates that ferroelectric polarization is notably diminished in thin films with fewer than 10 unit cells (∼4 nm). However, ferroelectric properties have been observed even in films as thin as 4–7.5 unit cells (1.6–3 nm) for BaTiO₃,^38–40 3–10 unit cells (1.2–4 nm) for PbTiO₃,⁴¹ 1.5 unit cells (around 0.6 nm) for PbZr_0.2T_0.8O₃,⁴² and 7.5 unit cells (about 3 nm) for BiFeO₃.⁴³ Recent TEM studies have suggested that ultrathin PbZr_0.2T_0.8O₃ films might not exhibit a critical thickness or size-dependent effects, with observed polarization values of approximately 22 μC cm⁻² for films with 2.5 unit cells and 16 μC cm⁻² for films with 1.5 unit cells.⁴² Despite these findings, determining the minimum thickness at which perovskite layers can retain ferroelectricity remains challenging due to the difficulties in synthesizing nearly perfect ultrathin perovskite films. Our research presents new opportunities to explore perovskite ferroelectrics with minimal critical thickness, as evidenced by the successful experimental confirmation of ferroelectricity in a 1.8 nm thick (3 unit cells) Pb₂Nb₃O₁₀ nanosheet, marking it as the thinnest free-standing perovskite with confirmed ferroelectric properties.

Conclusions

We investigated the ferroelectric properties of DJ-type CsPb₂Nb₃O₁₀ and its exfoliated 2D nanosheets. Ferroelectricity of CsPb₂Nb₃O₁₀ was demonstrated for the first time. CsPb₂Nb₃O₁₀ adopted a polar tetragonal structure with a moderate T_C of 260 °C and polarization P_S of 7.93 μC cm⁻², primarily arising from the out-of-plane displacements of Nb⁵⁺ ions and nearby oxygens. Importantly, the contribution from Pb²⁺ ions was nearly negligible, contrasting with observations in Bi₄Ti₃O₁₂ and CsBi₂Ti₂NbO₁₀. Having established the ferroelectric CsPb₂Nb₃O₁₀, we synthesized Pb₂Nb₃O₁₀ nanosheets via an intercalation-based exfoliation method. Through PFM measurements, we found that ferroelectricity remained stable even in the exfoliated 2D Pb₂Nb₃O₁₀ nanosheets. Our study opens new avenues for investigating 2D perovskite ferroelectrics at critical thicknesses.

Compared with conventional ferroelectrics in bulk and thin-film forms, 2D ferroelectric nanosheets have various advantages such as high capacitance density, low voltage operation, flexibility and bandgap tunability, which are suitable for technological applications, such as non-volatile memories, field-effect transistors, sensors and energy harvesting devices. Furthermore, the self-assembly of 2D nanosheets can provide a potentially useful route to ferroelectric nanodevices with precisely controlled thickness and properties. These studies are currently underway.

Data availability

The data supporting this article have been included as part of the ESI† at [https://doi.org/10.1039/x0xx00000x].

• Experimental section.

• FE-SEM image and EDS spectrum of CsPb₂Nb₃O₁₀; SHG spectra; S–E curve; Raman spectra; exfoliation process; FE-SEM image and EDS spectrum of HPb₂Nb₃O₁₀; PFM images of CsPb₂Nb₃O₁₀ single crystal; I–V curve of Pb₂Nb₃O₁₀ nanosheets; PFM signals from 3L film of Pb₂Nb₃O₁₀ nanosheets; thickness dependence.

• Powder X-ray crystallographic data of CsPb₂Nb₃O₁₀ (CIF).

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was funded by the Grant-in-Aid for Scientific Research KAKENHI (21H05015, 22H01907, 23K17956), Japan Society for the Promotion of Science (JSPS), Design & Engineering by Joint Inverse Innovation for Materials Architecture (DEJI²MA), MEXT, and the joint usage/research program of the Institute of Materials and Systems for Sustainability (IMaSS), Nagoya University.

References

1. R. E. Cohen, Nature, 1992, 358, 136.
2. B.-G. Kim, S. M. Cho, T.-Y. Kim and H. M. Jang, Phys. Rev. Lett., 2001, 86, 3404.
3. Y. Kuroiwa, S. Aoyagi, A. Sawada, J. Harada, E. Nishibori, M. Takata and M. Sakata, Phys. Rev. Lett., 2001, 87, 217601.
4. T. Abe, S. Kim, C. Moriyoshi, Y. Kitanaka, Y. Noguchi, H. Tanaka and Y. Kuroiwa, Appl. Phys. Lett., 2020, 117, 252905.
5. G. Shirane, R. Pepinsky and B. C. Frazer, Acta Crystallogr., 1956, 9, 131.
6. S. C. Abrahams, S. K. Kurtz and P. B. Jamieson, Phys. Rev., 1968, 172, 551.
7. V. Gavrilyachenko, R. Spinko, M. Ma and E. Fesenko, Fiz. Tverd. Tela, 1970, 12, 1532.
8. S. E. Cummins and L. E. Cross, J. Appl. Phys., 1968, 39, 2268.
9. Y. Noguchi and M. Miyayama, Appl. Phys. Lett., 2001, 78, 1903.
10. B.-W. Li, M. Osada, T. C. Ozawa and T. Sasaki, Chem. Mater., 2012, 24, 3111.
11. E. E. McCabe, E. Bousquet, C. P. J. Stockdale, C. A. Deacon, T. T. Tran, P. S. Halasyamani, M. C. Stennett and N. C. Hyatt, Chem. Mater., 2015, 27, 8298.
12. E. C. Subbarao, J. Phys. Chem. Solids, 1962, 23, 665.
13. Y.-I. Park, M. Miyayama and T. Kudo, J. Ceram. Soc. Jpn., 1999, 107, 413.
14. S. Dubey and R. Kurchania, Bull. Mater. Sci., 2015, 38, 1881.
15. M. A. Subramanian, J. Gopalakrishnan and A. W. Sleight, Mater. Res. Bull., 1988, 23, 837.
16. M. Fang, C. H. Kim and T. E. Mallouk, Chem. Mater., 1999, 11, 1519.
17. Y. Hu and L. Guo, ChemCatChem, 2015, 7, 584.
18. B. Zheng, L. Mao, J. Shi, Q. Chen, Y. Hu, G. Zhang, J. Yao and Y. Lu, Int. J. Hydrogen Energy, 2021, 46, 34276.
19. Y. Zhang, J. Yao, Z. Zhang, R. Zhang, L. Li, Y. Teng, Z. Shen, L. Kang, L. Wu and X. Fang, J. Mater. Sci. Technol., 2023, 164, 95.
20. Y. Shimakawa, Y. Kubo, Y. Nakagawa, S. Goto, T. Kamiyama, H. Asano and F. Izumi, Phys. Rev. B, 2000, 61, 6559.
21. J. M. Jehng and I. E. Wachs, Chem. Mater., 1991, 3, 100.
22. R. Nozaki, J. N. Kondo, C. Hirose, K. Domen, A. Wada and Y. Morioka, J. Phys. Chem. B, 2001, 105, 7950.
23. B.-W. Li, M. Osada, Y. Ebina, T. C. Ozawa, R. Ma and T. Sasaki, Appl. Phys. Lett., 2010, 96, 182903.
24. S. Pal, S. Das, R. Nagarajan and S. Uma, J. Mater. Chem. C, 2022, 10, 14675.
25. Q. Hou, B. Yang, C. Ma, Z. Zhou, R. Liang, H. Li and X. Dong, Ceram. Int., 2022, 48, 16677.
26. T. Zeng, X. Yu, S. Hui, Z. Zhou and X. Dong, Mater. Res. Bull., 2015, 68, 271.
27. Y. Noguchi, M. Miyayama, K. Oikawa, T. Kamiyama, M. Osada and M. Kakihana, Jpn. J. Appl. Phys., 2002, 41, 7062.
28. H. Irie, M. Miyayama and T. Kudo, J. Am. Ceram. Soc., 2000, 83, 2699.
29. I.-S. Yi and M. Miyayama, Mater. Res. Bull., 1997, 32, 1349.
30. M. Osada, M. Tada, M. Kakihana, T. Watanabe and H. Funakubo, Jpn. J. Appl. Phys., Part 1, 2001, 40, 5572.
31. M. Osada and T. Sasaki, Adv. Mater., 2012, 24, 210.
32. M. Osada and T. Sasaki, APL Mater., 2019, 7, 120902.
33. Y. Ebina, T. Sasaki and M. Watanabe, Solid State Ionics, 2002, 151, 177.
34. Y. Ebina, K. Akatsuka, K. Fukuda and T. Sasaki, Chem. Mater., 2012, 24, 4201.
35. B. J. Rodriguez, C. Callahan, S. V. Kalinin and R. Proksch, Nanotechnology, 2007, 18, 475504.
36. J. Junquera and P. Ghosez, Nature, 2003, 422, 506.
37. M. Stengel and N. A. Spaldin, Nature, 2006, 443, 679.
38. K. Hagiwara, K. N. Byun, S. Morita, E. Yamamoto, M. Kobayashi, X. Liu and M. Osada, Adv. Electron. Mater., 2023, 9, 2201239.
39. D. A. Tenne, P. Turner, J. D. Schmidt, M. Biegalski, Y. L. Li, L. Q. Chen, A. Soukiassian, S. Trolier-McKinstry, D. G. Schlom, X. X. Xi, D. D. Fong, P. H. Fuoss, J. A. Eastman, G. B. Stephenson, C. Thompson and S. K. Streiffer, Phys. Rev. Lett., 2009, 103, 177601.
40. Z. Wen, C. Li, D. Wu, A. Li and N. Ming, Nat. Mater., 2013, 12, 617.
41. D. D. Fong, G. B. Stephenson, S. K. Streiffer, J. A. Eastman, O. Auciello, P. H. Fuoss and C. Thompson, Science, 2004, 304, 1650.
42. P. Gao, Z. Zhang, M. Li, R. Ishikawa, B. Feng, H.-J. Liu, Y.-L. Huang, N. Shibata, X. Ma, S. Chen, J. Zhang, K. Liu, E.-G. Wang, D. Yu, L. Liao, Y.-H. Chu and Y. Ikuhara, Nat. Commun., 2017, 8, 15549.
43. Y. H. Chu, T. Zhao, M. P. Cruz, Q. Zhan, P. L. Yang, L. W. Martin, M. Huijben, C. H. Yang, F. Zavaliche, H. Zheng and R. Ramesh, Appl. Phys. Lett., 2007, 90, 252906.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4dt02343a

‡ These two authors contributed equally to this work.

---