Field-induced polarization rotation and phase transitions in $0.70\mathrm{Pb}\left(\mathrm{M}{\mathrm{g}}_{1/3}\mathrm{N}{\mathrm{b}}_{2/3}\right){\mathrm{O}}_3-0.30\mathrm{PbTi}{\mathrm{O}}_3$ piezoceramics observed by in situ high-energy x-ray scattering

Changes to the crystal structure of $0.70\mathrm{Pb}\left(\mathrm{M}{\mathrm{g}}_{1/3}\mathrm{N}{\mathrm{b}}_{2/3}\right){\mathrm{O}}_3-0.30\mathrm{PbTi}{\mathrm{O}}_3$ (PMN-0.30PT) piezoceramic under application of electric fields at the long-range and local scale are revealed by in situ high-energy x-ray diffraction (XRD) and pair-distribution function (PDF) analyses, respectively. The crystal structure of unpoled samples is identified as monoclinic $Cm$ at both the long-range and local scale. In situ XRD results suggest that field-induced polarization rotation and phase transitions occur at specific field strengths. A polarization rotation pathway is proposed based on the Bragg-peak behaviors and the Le Bail fitting results of the in situ XRD patterns. The PDF results show systematic changes to the structures at the local scale, which is in agreement with the changes inferred from the in situ XRD study. More importantly, our results prove that polarization rotation can be detected and determined in a polycrystalline relaxor ferroelectric. This study supports the idea that multiple contributions, specifically ferroelectric-ferroelectric phase transition and polarization rotation, are responsible for the high piezoelectric properties at the morphotropic phase boundary of PMN-$x\mathrm{PT}$ piezoceramics.

Figure 1

(a) FE-SEM images, (b) dielectric measurements, and (c) $P-E$ loops of the PMN-0.30PT piezoceramic. Inset of subplot (b) shows the deviation of Curie-Weiss law; the dashed line shows the linear fit of $(1/\varepsilon \prime )$ versus $T$ at the paraelectric phase.

Figure 2

The overall fit of the unpoled PMN-0.30PT XRD pattern using monoclinic $Cm$ structure for laboratory XRD (a) and SXRD (b). The observed XRD is shown as black crosses, the calculated fit in red, and the difference in blue.

Figure 3

Reduced total scattering function (a) and PDF pattern after Fourier transform (b) of the unpoled PMN-0.30PT piezoceramic. (c) Fits of the unpoled PMN-0.30PT PDF pattern using $Cm$ model over a range of $2–6\AA$ (representing the local scale structure), and over $8–28\AA$ (representing intermediate scale structure). The observed data are shown as blue circles, the calculated fit in magenta, and the difference in green.

Figure 4

(a) In situ electric field XRD in selected representative $Q$ ranges. The process is categorized into three stages. (b) The best fits using specific model for patterns at 1.5, 1.6, 2.2, and 4.0 kV/mm, respectively. Note for 4.0-kV/mm pattern a $P4mm\left(\mathrm{red}\right)+Pm\left(\mathrm{black}\right)$ mixed-phase model was adopted.

Figure 5

Lattice parameters: (a) cell-edge length, (b) $\beta$ angle, and (c) cell volume of PMN-0.30PT under various fields obtained from the Le Bail fitting. The assigned stage numbers and identified structures are noted. (d) Polarization rotation path for PMN-0.30PT piezoceramic proposed by this study. For the patterns fitted with mixed-phase model, only structure information of $Pm$ is plotted.

Figure 6

In situ electric-field PDF patterns from 0 to 4.0 kV/mm at different length scales. The arrows indicate the evolution of PDF peaks with increasing fields. The dashed lines in subplot (a) show the negligible changes of nearest- and second-nearest Pb-Pb pairs with increasing field.

Figure 7

Representative fits of Pb-B local PDF peaks at $r\approx 3.5\AA$ under selected fields. The observed data are shown as blue circles, the sum fit in magenta, and the resolved individual peaks in red, blue, or green. Subplots (a) and (b) show the three-peaks model fits well for 0–1.8-kV/mm patterns. Subplots (c) and (d) show the two-peaks model better represents the high-field local structures. The gray dashed line shows the fits of local Pb-Pb PDF peaks at $r\approx 4.0\AA$.