Switchable Polarization in an A-site Deficient Perovskite through Vacancy and Cation Engineering

TL;DR

This work shows that deliberate A-site vacancy ordering in the perovskite $Y_{1/3}$TaO$_3$ can stabilize a polar ground state through coupling between vacancy-ordered layers and TaO$_6$ rotations, yielding room-temperature switchable ferroelectricity in a polar $Pb_2_1m$ structure. Diffraction, SHG, polarization measurements, and first-principles phonon analysis reveal a temperature-driven transition to a $(3+2)$-dimensional incommensurate paraelectric phase near $T \\simeq 750$ K, with domain topology reminiscent of hybrid improper ferroelectrics. The study quantifies a measured polarization of about $0.8 \\mu$C/cm$^2$, and DFT suggests that modest compressive strain could elevate the intrinsic polarization to around $21.8 \\mu$C/cm$^2$ by suppressing polar instabilities. Overall, vacancy engineering emerges as a general strategy to amplify polarization in improper ferroelectrics, including magnetoelectric multiferroics, and strain-tuned designs may unlock higher-performance polar states in this materials class.

Abstract

While defects are unavoidable in crystals and often detrimental to material performance, they can be a key ingredient for inducing functionalities when tailored. Here, we demonstrate that an A-site-deficient perovskite Y$_{1/3}$TaO$_3$ exhibits room-temperature ferroelectricity in a $Pb2_1m$ phase, enabled by ordered vacancies coupled with TaO$_6$ octahedral rotations. Defect-ordered perovskites are frequently trapped in centrosymmetric incommensurate states due to competing structural instabilities; we circumvent this by favoring rotational over polar instability through compositional selection. Unlike canonical improper ferroelectrics that are \textit{ferrielectric}, the vanishing dipoles on vacancy layers in Y$_{1/3}$TaO$_3$ allow for a net ferroelectric alignment of local dipoles, resulting in enhanced polarization. Upon heating, Y$_{1/3}$TaO$_3$ transforms to a paraelectric incommensurate phase at $\simeq$750 K, whose atomic arrangement mirrors the domain topology observed in hybrid improper ferroelectrics. Superspace analysis of the modulated phase reveals a route to improve room-temperature polarization, achieved through epitaxial strain, as confirmed by our lattice-dynamics calculations. This defect-ordering strategy should be generalizable to other improper ferroelectrics, including magnetoelectric multiferroics, providing a pathway to amplify otherwise limited macroscopic polarization.

Figures (10)

• Figure 1: Schematic illustration of (a) a simple perovskite, (b) A/A$^{\prime}$ layer-ordered double perovskite, and (c) A-site deficient perovskite with an alternate ordering of vacancies, all of which exhibit octahedral rotations corresponding to the Glazer notation of $a^-a^-c^+$glazer1972ACB. In panels (b) and (c), the asymmetry introduced by the distinct adjacent A layers leads to the absence of inversion centers at the B cations. Red and blue arrows show local electric dipoles arising from A and A$^{\prime}$ cation displacements. Dotted arrows in panel (c) indicate vanishing local polarization due to the complete absence of cations.
• Figure 2: Phonon dispersion curves calculated along a high-symmetry path [$\Gamma$(0 0 0)--X($\frac{1}{2}$ 0 0)--M($\frac{1}{2}$$\frac{1}{2}$ 0)] for the nondistorted $P4/mmm$ structure of Ca$_{1/2}$BO$_3$ (B = Nb and Ta). This model represents Y$_{1/3}$BO$_3$ systems, where half of the A-site is randomly occupied by Y$^{3+}$. Eigendisplacements for key vibrational modes, including $\Gamma_5^-$, X$_1^+$, and M$_5^-$ modes, are depicted in the lower panel.
• Figure 3: Temperature dependence of (a) $\frac{1}{2}$$\frac{1}{2}$3, $\frac{1}{2}$$\frac{3}{2}$1, and $\frac{3}{2}$$\frac{1}{2}$1 superlattice reflections upon heating and (b) 112 reflection upon heating and cooling (SXRD, $\lambda$ = 0.501146 Å). The Miller indices are defined with respect to the $P4/mmm$ unit cell with the possible highest symmetry. (c) Temperature-variable ND patterns ($\lambda$ = 1.594 Å) highlighting the same superlattice reflections and peak splitting at lower temperatures. The green arrows indicate satellite reflections of the paraelectric phase.
• Figure 4: Rietveld refinement results for 300-K (a) SXRD ($\lambda$= 0.42112 Å), (b, c) 2$\theta$-dispersive ND, and (d--f) time-of-flight ND data of Y$_{1/3}$TaO$_3$ with a $Pb2_1m$ structural model. Orange circles, black lines, and blue lines represent the observed, calculated, and difference profiles, respectively. The green ticks indicate the position of Bragg reflections. All six data sets were fitted simultaneously. Broad satellite peaks are excluded from the refinement.
• Figure 5: Atomic structure illustration of the [100]-projected $Pb2_1m$ phase and the layer-resolved local polarization along the $b$ axis calculated for the refined $Pb2_1m$ structure with a point-charge approximation. The blue dashed square represents the unit cell.