Stress release-induced piezoresponse in lead-free piezoceramics with high-symmetry morphotropic phase boundary

1 Stress release induced piezoresponse in lead-free piezoceramics with high-symmetry morphotropic phase boundary Xavier Vendrell,1,a) Fernando Rubio-Marcos,2 Adolfo del Campo,2 Lourdes Mestres,1 and Jose E. Garcia3,a) 1 Department of Inorganic and Organic Chemistry, Universitat de Barcelona, 08028 Barcelona, Spain. 2 Departmen t of Electroceramics, Instituto de Cerámica y Vidrio (CSIC), 28049 Madrid, Spain 3 Department of Physics, Universitat Politècnica de Catalunya - BarcelonaTech, 08034 Barcelona, Spain The (1-x)(Bi0.5Na0.5)TiO3-xBaTiO3 (BNT-BT) system has attracted a great deal of interest because it presents a morphotropic phase boundary (MPB) between the rhombohedral and tetragonal phases for 0.05<x<0.08. Finding the MPB in BNT-BT system often leads up to obtain a material that shows a high symmetry (pseudo)cubic X-ray diffraction pattern. However, this singular composition exhibits ferroelectricity, which has been explained as a consequence of a field-induce phase transformation. Here, we demonstrate that the stress release after poling, from the virgin state of the sample, is a crucial phenomenon to obtain piezoelectric response in MPB BNT-BT. The mechanism behind the unusual poling-depoling process in piezoceramics exhibiting high-symmetry MPB is elucidated by combining X-ray diffraction measurements and advanced Raman spectroscopy. This underscores the importance of post-poling stress release from the virgin state as a critical factor in attaining piezoelectric response in lead-free piezoceramics with high symmetry MPB configurations. Keywords: Raman spectroscopy; piezoelectric ceramics; ferroelectricity; stress relaxation; morphotropic phase boundary a) Authors to whom correspondence should be addressed: [email protected]; [email protected] This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678 2 Solid solutions showing morphotropic phase boundary (MPB) between two ferroelectric polymorphs have attracted a lot of attention because the unique properties these compositions may exhibit.1-3 For instance, exceptional properties were found in lead zirconate titanate (PZT) solid solution for compositions in which the MPB forms between the tetragonal and rhombohedral polar phases,4 being these materials the undisputed market leaders of piezoelectric oxides. However, environmental concerns related to lead toxicity have led to the development of lead-free piezoelectric materials, which has focused the attention of many researchers in the last twenty years.5-7 Since the better piezoelectric properties of lead-based piezoelectric materials was achieved in MPB compositions, the search for lead-free piezoceramics with potential high piezoelectricity has therefore been focused on solid solutions exhibiting MPB.8-10 The (1-x)(Bi0.5Na0.5)TiO3-xBaTiO3 (BNTBT) system has attracted a great deal of interest because it presents (like the PZT) a MPB between the rhombohedral and tetragonal phases for 0.05<x<0.08.11-16 Although BNT-BT has demonstrated to own moderate piezoelectric response,17-19 some intriguing features of this system still deserve attention because understand them may serve to the design of improved lead-free piezoelectric compositions. A delicate balance of polymorphic phases is usually reached at MPB, which has demonstrated to be strongly dependent of the material processing. In the case of BNT-BT, not only a mixed phase of tetragonal and rhombohedral polymorphs is obtained but also a high symmetry, pseudo-cubic phase may be achieved for the same nominal composition (x=0.06).13,20-22 The functional properties depend on the resulting phases balance; therefore, variable performance is attained depending on the processing condition. Here, we report an in-depth analysis of the local scale structural stress of BNTBT ceramics exhibiting high symmetry MPB. Results demonstrate that, in our specific case, the bulk sample undergoes increased stress compared to their powdered counterpart. While this may suggest a stress-induced phase phenomenon, it is not the sole origin of the ferro-piezo response, given the well-recognized potential contribution of field-induced effects in MPB systems. Polycrystalline samples comprising 0.94(Bi0.5Na0.5TiO3)-0.06BaTiO3 (BNT-6BT) were synthesized using conventional solid-state reaction techniques. Detailed procedures for the preparation of both the powder and ceramics are provided elsewhere.11 After sintering, a section of the sintered sample was crushed and annealed, giving a powder with particle size between 1 and 5 μm. Crystalline phases were analyzed using X-ray diffraction (XRD) with a PANalytical X'pert Pro instrument. Patterns were captured within an angular range of 20 to 80 degrees (2θ) with a step size of 0.0334 degrees and a duration of 100 seconds per step. CuKα radiation was employed with a working voltage of 40 kV and a current of 40 mA. The ferroelectric phase distribution and stress degree of each sample were determined using a Witec alpha-300RA Confocal Raman Microscope. Raman images of the phase This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678 3 were obtained using a 100x objective lens with a numerical aperture of 0.95 on powder and ceramics regions of 120 x 120 µm and 70 x 70 µm on the surface, respectively. The Raman images, revealing the phase distribution and stress degree induced by the polarization and depolarization processes, were constructed from 60 x 60 and 70 x 70 spectra (i.e., 3600 and 4900 spectra per Raman image, respectively) with an acquisition time of 0.5 seconds per spectrum and a laser power of 20 mW. Gold electrodes were sputtered on parallel polished faces of the bulk sample for poling, which were carried out in a silicone bath at 25 C under a dc electric field of 5 kV/mm for 30 min. The polarized sample was aged for a week before measurements were performed in order to prevent the influence of aging processes. Subsequently, the sample was annealed by heating at a rate of 5 C/min until reaching 250 C, followed by a slow cooling. Before another week, the sample was polarized again. The ferroelectric hysteresis loops of the unpoled sample were recorded using a modified Sawyer-Tower circuit at 1 Hz and at room temperature. XRD is a well-established technique to evaluate the structural characteristics of MPB ferroelectric materials. Fig. 1 displays the XRD patterns of BNT-6BT powder (a) and ceramics (b-d). Notably, the fabrication of bulk ceramics inherently introduces internal stress, primarily due to grain confinement during the sintering process, resulting in boundary effects. Detailed XRD analysis of the powdered sample reveals the presence of a cubic or pseudocubic phase (Fig. 1a). This is evidenced by examining the 2θ regions between 39.5°<2θ<40.5° and 46°<2θ<47°, where a single peak is observed in both cases, corresponding to the (111) and (200) reflections, respectively. In the former 2θ region, two peaks would be anticipated in a rhombohedral structure, representing the (003) and (021) planes, while a tetragonal phase would also manifest a splitting of the (200) reflection. The XRD diffraction pattern of the bulk sample displays an initial separation of the cubic reflections on its pristine state. This indicates that the BNT-6BT undergoes a mixture of rhombohedral and tetragonal phases upon ceramics is formed after sintering. Ferroelectric response at this state (Fig. 1e) exhibit a typical P-E loop with expected values of remnant polarization and coercive field.23 The emergence of the rhombohedral and tetragonal phases is enlarged after poling the ceramic, where the peak-splitting in both regions becomes more pronounced (Fig. 1c). A similar behavior has already been examined by Daniels et al.,24 where a ferroelectric phase (non-cubic) transformation can be induced by the application of an electric field. When annealing the polarized sample, it becomes apparent that the structure tends towards increased symmetry compared to the nonpolarized case, as indicated by the reduced visibility of the diffraction maxima splitting, although it does not fully revert to the initial (pseudo)cubic state observed in the powdered sample. More importantly, a second poling process does not significantly change the XRD pattern of the unpoled sample after annealing (Fig. 1d), unlike what happened during the first polarization. This suggests a stress-triggered phase phenomenon, although it is not the primary source of the ferro-piezo This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678 4 response, taking into account the potential contribution induced by the electric field. Note that the ferroelectric response after annealing (Fig. 1f) reveals a harder P-E loop, indicated by a noticeable increase in the coercive field along with a decrease in remnant polarization. Furthermore, the polarization does not seem to saturate at the same maximum applied electric field, all of which indicates that the polarization orientation of this material in this state becomes more difficult. It is worth noting that although laboratory XRD patterns primarily provide information from the surface of the sample,25 the comparison of diffraction patterns under different sample conditions (Fig. 1b-d) should still be considered relevant. Once we have confirmed that stress-induced changes in crystal structure can be detected using XRD measurements, our focus is to demonstrate how stress-induced phase phenomena can alter domain configuration, thus influencing the macroscopic polarization of the crystal. To assess the effect of stress levels on the ceramic microstructure, experiments were conducted using confocal Raman microscopy (CRM).26 Fig. 2 shows a comparison between powder and ceramic samples using CRM. In both instances, a sizable work area was chosen to assess the homogeneity of the systems. Fig. 2a-b displays images obtained through optical microscopy of the analyzed areas, while Fig. 2c-d illustrate the Raman images of the powder and ceramic, respectively. Notably, the primary compositional difference between the powder and ceramic samples lies in the presence of a secondary phase in the ceramic (highlighted as blue regions), although we can assume this is a residual phase on the surface since it is not detected by XRD. Fig. 2(e–f) depicts the average Raman spectra derived from the Raman images shown in Fig. 2(c-d), corresponding to the BNT-BT perovskite phase, represented by the regions in red. The deconvolved spectra is detailed in Table I. As evidenced, the primary Raman mode (labeled as 2) is centered at approximately in 270 cm⁻¹, and it is employed to determine the relative homogeneity and stress of the samples. A more intriguing explanation emerges from the statistical analysis of the primary Raman mode for both powder and ceramic samples. The powder sample (Fig. 2g) shows a narrow distribution of mode 2, indicating lower compositional variation compared to the broader distribution in the ceramic sample (Fig. 2h). Another noteworthy observation is the increase in the average Raman shift of mode 2 in the ceramic sample (271 cm⁻¹) compared to the powder material (267 cm⁻¹). This shift should be attributed to the distortion of the crystal lattice caused by the emergence of ferroelectric domains, which distribute stresses within the grains. This is corroborated by XRD analysis (Fig. 1), indicating the appearance of ferroelectrically active phases. Consequently, the ceramic in its pristine state exhibits a higher degree of stress. This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678 5 TABLE I. Raman modes and their mode symmetry assignments in rhombohedral BNT ceramics. According to the nuclear site group analysis, Raman active phonons of the rhombohedral symmetry with space group R3c (C63v) are represented by ΓRaman = 7A1+6E. Raman Shift (cm-1) Symmetry Labeled Description ~ 150 E (TO) 1 Network modifiers or distorted octahedral [BiO6] and [NaO6] clusters ~ 275 A1 (TO2) 2 Polar vibration of A1 character highly sensitive to the local polar order 380 Mode originated from A1 (TO2) mode of Ti-based perovskites. ~ 520 A1 (TO3) 3 Ascribed to the (O–Ti–O) stretching symmetric vibrations of the octahedral [TiO6] clusters. ~ 590 4 ~ 760 E (LO4) 5 Modes due to the presence of the sites within the rhombohedral lattice pre containing octahedral distorted [TiO6] clusters. ~ 830 A1 (LO3) 6 Similar to the observations in XRD, when the ceramic sample is polarized and depolarized, a relaxation-stress phenomenon occurs. Fig. 3 illustrates a polarization-depolarization cycle monitored by CRM. Fig. 3a-b shows the average Raman spectra and statistical analysis of the primary Raman mode of powdered (Fig. 3a) and ceramic (Fig. 3b) BNT-6BT. Upon polarization (Fig. 3c), the sample returns to its initial state, with the distribution of the primary Raman mode appearing at 267 cm⁻¹, indicating a relaxation phenomenon.26 However, upon depolarization (Fig. 3d), the Raman shift increases to 270 cm⁻¹, signifying an increase in stress. The observed Raman shift can be attributed to changes in the strength of the constant force, caused by the variation of the distance between Ti4+ ions and the coordinated oxygen, which alter the ferroelectric order. However, while intrinsic lattice strain increases due to domain alignment during poling, stress release is associated not only with the intrinsic properties of the material but also with extrinsic factors. Localized macroscopic (extrinsic) stress redistribution, particularly at grain boundaries, may contribute to localized relaxation effects.27 Upon repolarization of the ceramic (Fig. 3e), the sample does not revert to its relaxed (i.e., first polarized) state, suggesting a tendency towards increased symmetry compared to the non-polarized state. Clearly, it lacks piezoelectric characteristics. Attempts to repolarize the ceramic are futile, suggesting that the sample fails to relax again, thus remaining devoid of electromechanical properties. Essentially, it is not the stabilization of a pseudocubic phase but rather the relaxation of stress during the initial polarization that enables the sample to exhibit favorable piezoelectric This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678 6 properties. This suggests that the initial polarization process involves stress relaxation, explaining why complete polarization is achieved almost instantly and at room temperature.23 To comprehensively assess the bulk sample, given its small grain size, we conducted Raman imaging on smaller areas. This approach enabled detailed observation of phase orientation and system tension (i.e., stress) within each grain, contingent upon the polarization angle of the light. Fig. 4a displays an optical image of the sample surface, while Fig. 4b-c depicts the Raman image and stress distribution derived from Raman mode 2 (Table I) of the region delineated by the red rectangle in Fig. 4a. To narrow the focus of the study, we isolated two grains labeled as G1 and G2 in Fig. 4b. The average spectra of both zones are depicted in Fig. 4d, revealing that the G1 grain exhibits an elevated average Raman shift of mode 2 (272.6 cm⁻¹) compared to the G2 grain (265.6 cm⁻¹), a distinction also evident in Fig. 4c. This indicates that a stress gradient exists, causing not all grains to develop the same tension. Thus, grain G1 presents a higher degree of stress compared to grain G2, which can be explained by the emergence of different ferroelectric phases. The mixed phase region induces a high degree of polarization directions close to room temperature, associated with polarization directed along the [001]p and [111]p primitive cell edges for the tetragonal (T) and the rhombohedral (R) phases, respectively.28 Therefore, the R-T phase coexistence at room temperature produces a lattice deformation associated with the high degree of polarization directions, contributing to the generation of different domain configurations as a consequence of the existing stress gradient in the system. Since the polarization vector correlates directly with Raman displacement,29 this suggests varying degrees of polarization across these grains. Moreover, to ascertain the orientation of the polarization vector (or crystallographic orientation), we analyzed the impact of light polarization angle on the Raman modes of the grains. Fig. 4e illustrates the fluctuations in the average spectrum of each selected grain for polarized light angles (θ) ranging from 0° to 90°. Notably, the Raman modes of the spectrum corresponding to the G1 grain show no dependence on θ variations (neither intensity nor Raman shift undergo modification), indicating that the polarization vector in grain G1 lies out of plane (as schematically represented in panel 4e). Conversely, the G2 grain exhibits a clear dependence on θ, with both the intensity and shift of the Raman modes evolving with the laser polarization angle. Hence, it can be inferred that the polarization vector of grain G2 lies in the plane, rendering it susceptible to variations with θ. As shown in Fig. 4c, stressed grains with out-of-plane polarization are surrounded by distended grains exhibiting in-plane polarization, being this configuration unfavorable to being reoriented as verified by the P-E loop in Fig. 1f. In summary, we have demonstrated that the MPB BNT-BT can undergo a conformation-induced phase transformation from pseudocubic to a non-cubic, tetragonal plus rhombohedral, polymorphic phase. The transformation can be explained in terms of stress-relaxation cycles where samples This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678 7 undergo depolarization and maintain the ferroelectric phase, yet Raman analysis indicates a reemergence of stress, and therefore, no piezoelectric response is observed. Attempts to repolarize the sample are unsuccessful, suggesting that the sample fails to relax again, thus remaining devoid of electromechanical properties. Essentially, it is not the stabilization of a pseudocubic phase but rather the relaxation of stress during the initial polarization that enables the sample to exhibit favorable piezoelectric properties. This implies that stress relaxation plays a crucial role in the initial polarization, explaining the rapid achievement of complete polarization at room temperature. This holds particular significance in the context of MPB systems, which exhibit high symmetry in powder form. Hence, it is reasonable to suggest inducing a degree of phase "instability" during processing to yield a ferroelectric powder phase. This elucidates the notable discrepancies in properties observed among samples apparently possessing identical compositions but varying levels of symmetry in their diffraction patterns. Therefore, the release of stress post-poling, starting from the pristine state of the sample, is a crucial phenomenon for attaining a piezoelectric response in lead-free piezoceramics characterized by high symmetry MPB configurations. Acknowledgement This work was supported by the Spanish Ministry of Science and Innovation, grant numbers PID2020114192RB-C41 and TED2021-130957B-C51 and PID2023-153398OB-I00, funded by MICIU/AEI/10.13039/501100011033 and the European Social Fund Plus (FSE+). F. R-M. is indebted to Comunidad de Madrid for the financial support through the Doctorados Industriales project, grant number IND2020/IND-17375, which is co-financed by the European Social Fund. X. V. is a Serra Húnter Fellow and is grateful to the Generalitat de Catalunya project 2023 CLIMA 00009 AGAUR. AUTHOR DECLARATIONS Conflict of Interest The authors have no conflicts to disclose. Author Contributions Xavier Vendrell: Formal analysis; Investigation; Resources; Visualization; Writing – original draft. Fernando Rubio-Marcos: Formal analysis; Investigation; Validation; Visualization; Funding acquisition; Writing – review & editing. Adolfo del Campo: Formal analysis; Investigation. Lourdes Mestres: Methodology; Resources; Validation. Jose E. Garcia: Conceptualization; Methodology; Visualization; Writing – review & editing; Supervision; Project administration. DATA AVAILABILITY The data that support the findings of this study are available from the corresponding author upon reasonable request. This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678 8 REFERENCES 1 B. Noheda, Curr. Opin. Solid. State Mater. Sci. 6, 27–34 (2002). 2 S. Trolier-Mckinstry, S. Zhang, A. J. Bell, and X. Tan, Annu. Rev. Mater. Res. 48, 191-217 (2018). 3 S. Zhang and F. Li, J. Appl. Phys. 111, 031301 (2012). 4 R. Guo, L. E. Cross, S-E. Park, B. Noheda, D. E. Cox, and G. Shirane, Phys. Rev. Lett. 84, 5423-5426 (2000). 5 J. Rödel, W. Jo, K. T. P. Seifert, E. Anton, T. Granzow, and D. Damjanovic, J. Am. Ceram. Soc. 92, 1153–1177 (2009). 6 J. Rödel, K. G. Webber, R. Dittmer, W. Jo, M. Kimura, and D. Damjanovic, J. Eur. Ceram. Soc. 35, 1659–1681 (2015). 7 T. Zheng, J. Wu, D. Xiao, and J. Zhu, Prog. Mater. Sci. 98, 552–624 (2018). 8 R. Wang, K. Wang, F. Yao, J. Li, F.H. Schader, K.G. Webber, W. Jo, and J. Rödel, J. Am. Ceram. Soc. 98, 2177– 2182 (2015). 9 J. Wu, D. Xiao, and J. Zhu, Chem. Rev. 115, 2559–2595 (2015). 10 X. Lv, J. Zhu, D. Xiao, X. Zhang, and J. Wu, Chem. Soc. Rev. 49, 671-707 (2020). 11 R. Machado, V. B. dos Santos, D. A. Ochoa, E. Cerdeiras, L. Mestres, and J. E. García, J. Alloys Compd. 690, 568–574 (2017). 12 G. Picht, J. Töpfer, and E. Hennig, J. Eur. Ceram. Soc. 30, 3445–3453 (2010). 13 G. Das Adhikary, G. J. Muleta, and R. Ranjan, Phys. Rev. B 108, 224105 (2023). 14 M. Hinterstein, L. A. Schmitt, M. Hoelzel, W. Jo, J. Rödel, H.-J. Kleebe, and M. Hoffman, Appl. Phys. Lett. 106, 222904 (2015). 15 W. Jo, J. E. Daniels, J. L. Jones, X. Tan, P. A. Thomas, D. Damjanovic, and J. Rödel, J. Appl. Phys. 109, 014110 (2011). 16 P. Singh, A. Sharma, P. Sharma, and J. Kolte, Appl. Phys. A 129, 643 (2023). 15 R. Garg, B. N. Rao, A. Senyshyn, P. S. R. Krishna, and R. Ranjan, Phys. Rev. B 88, 014103 (2013). 16 Y. S. Sung, J. M. Kim, J. H. Cho, T. K. Song, M. H. Kim, and T. G. Park, Appl. Phys. Lett. 96, 202901 (2010). 17 R. R. McQuade and M. R. Dolgos, J. Solid State Chem. 242, 140–147 (2016). 18 L. E. Fuentes-Cobas, M. E. Montero-Cabrera, J. Plaisier, E. E. Villalobos-Portillo, and L. Pardo, J. Solid State Chem. 316, 123585 (2022). 19 E. Mercadelli, A. Sanson, C. Capiani, L. Costa, and C. Galassi, Process. Appl. Ceram. 3, 73–78 (2009). 22 H. Simons, J. E. Daniels, W. Jo, R. Dittmer, A. Studer, M. Avdeev, J. Rödel, and M. Hoffman, Appl. Phys. Lett. 98, 082901 (2011). 20 H. Guo, C. Ma, X. Liu, and X. Tan, Appl. Phys. Lett. 102, 092902 (2013). 21 J. E. Daniels, W. Jo, J. Rödel, and J. L. Jones, Appl. Phys. Lett. 95, 032904 (2009). 22 S. Kong, N. Kumar, S. Checchia, C. Cazorla, and J. Daniels, Adv. Funct. Mater. 29, 1900344 (2019). 23 A. Prado-Espinosa, J. Camargo, A. del Campo, F. Rubio-Marcos, M. Castro, and L. Ramajo, J. Alloys Compd. 739, 799–805 (2018). 24 F. Rubio-Marcos, R. López-Juárez, R. E. Rojas-Hernandez, A. del Campo, N. Razo-Pérez, and J. F. Fernandez, ACS Appl. Mater. Interfaces 7, 23080–23088 (2015). 25 F. Rubio-Marcos, Adolfo Del Campo, R. E. Rojas-Hernandez, M. O. Ramírez, R. Parra, R. U. Ichikawa, L. A. Ramajo, L. E. Bausá, and J. F. Fernández, Nanoscale 10, 705-715 (2018). 27 A. Reyes-Montero, F. Rubio-Marcos, L. Pardo, A. Del Campo, R. López-Juárez, and M.E. Villafuerte-Castrejón, J. Mater. Chem. A 6, 5419 (2018) This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678 9 FIG. 1. X-ray diffraction (XRD) patterns and ferroelectric hysteresis (P-E) loops of BNT-6BT at different state. (a) XRD pattern of the powder showing an apparent cubic phase. (b) The XRD of the pristine state of the ceramics exhibits an incipient appearance of rhombohedral and tetragonal phases that is enlarged when the sample is (c) polarized. (d) XRD pattern after a second poling shows a crystallographic structure more symmetric than the first polarized state. P-E loops at (e) the pristine state of the ceramics and (f) after annealing and before the second poling. FIG. 2. Fast evaluation of the structure and homogeneity of both the powder and the ceramic samples. (a-b) Optical micrographs of the powders and the polished surface of the ceramic, respectively. Scale bar, 20 μm. (cd) Raman images showing chemical (in)homogeneity of the powder and the ceramics. The Raman image is derived by summing the total spectral pixel intensity from 100 cm-1 to 1200 cm-1. Scale bar, 10 μm. Blue areas in panel d show the secondary phase location on the bulk BNT-6BT. The black rectangles marked as A and B in the panels (a-b) show the positions where the XY Raman image are taken. (e-f) Average Raman spectra of the powder and the ceramic samples, respectively. Raman spectra are fitted to the sum of six Lorentzian peaks corresponding to Raman modes of the BNT-BT perovskite phase. The numbers next to the vibrational peaks represent the main atomic motions (for the assignment of the Raman modes see Table I). (g-h) Statistical analysis of the number of pixels versus Raman shift at each pixel corresponding to the Raman mode 2 for the powder and ceramic samples. An enlargement of the Raman shift of of 4 cm-1 is revealed. This is the author’s peer reviewed, accepted manuscript. However, the online version of record will be different from this version once it has been copyedited and typeset. PLEASE CITE THIS ARTICLE AS DOI: 10.1063/5.0238678