[Paper Review] Atomic scale symmetry and polar nanoclusters in the paraelectric phase of ferroelectric materials

Andreja Benčan, Emad Oveisi+7 more|arXiv (Cornell University)|Oct 21, 2020

Ferroelectric and Piezoelectric Materials61 references104 citations

TL;DR

This study provides direct atomic-scale evidence of static, 2–4 nm polar nanoclusters in the nominally nonpolar paraelectric phases of (Ba,Sr)TiO3 and BaTiO3 using aberration-corrected scanning transmission electron microscopy (STEM) and Raman spectroscopy. The researchers visualize and quantify static off-site displacements of Ti and O ions from ideal cubic positions, revealing local symmetry breaking and confirming the existence of long-range polar order despite the macroscopic cubic symmetry, resolving a long-standing debate on the dynamic versus static nature of polar nanoregions in classical ferroelectrics.

ABSTRACT

The nature of the "forbidden" local- and long-range polar order in nominally nonpolar paraelectric phases of ferroelectric materials has been an open question since the discovery of ferroelectricity in oxide perovskites (ABO3). A currently considered model suggests locally correlated displacements of B-site atoms along a subset of <111> cubic directions. Such offsite displacements have been confirmed experimentally, however, being essentially dynamic in nature they cannot account for the static nature of the symmetry-forbidden polarization implied by the macroscopic experiments. Here, in an atomically resolved study by aberration corrected scanning transmission electron microscopy (STEM) complemented by Raman spectroscopy, we reveal, directly visualize and quantitatively describe static, 2-4 nm large polar nanoclusters in the nominally nonpolar cubic phases of (Ba,Sr)TiO3 and BaTiO3. These results have implications on understanding of the atomic-scale structure of disordered materials, the origin of precursor states in ferroelectrics, and may help answering ambiguities on the dynamic-versus-static nature of nano-sized clusters.

Motivation & Objective

• To resolve the longstanding debate on whether polar nanoregions in paraelectric perovskites are dynamic or static.
• To directly visualize and quantify local atomic displacements in nominally nonpolar cubic phases of ferroelectric oxides.
• To determine the atomic-scale symmetry breaking and polar character of nanoscale regions in (Ba,Sr)TiO3 and BaTiO3.
• To clarify the origin of macroscopic polarization in paraelectric phases that is forbidden by nominal cubic symmetry.
• To investigate the role of polar nanoclusters in dielectric tunability and loss mechanisms in technologically important ferroelectrics.

Proposed method

• Aberration-corrected scanning transmission electron microscopy (STEM) with high-angle annular dark-field (HAADF) and annular bright-field (ABF) imaging to resolve atomic positions and displacements.
• Energy dispersive X-ray spectroscopy (EDS) to confirm elemental composition and validate structural models.
• Raman spectroscopy to probe local lattice dynamics and detect symmetry-breaking modes associated with polar nanoclusters.
• Simulated ABF images from atomic models with tetragonal (P4mm) and cubic (Pm-3m) phases to validate experimental displacement measurements.
• Lattice strain mapping from HAADF images by extracting x/y lattice parameter ratios to identify strain fields at nanocluster interfaces.
• Complex elastic modulus measurements via dynamic mechanical analysis to detect elastically active nanoscale objects.

Experimental results

Research questions

• RQ1 Do static polar nanoclusters exist in the paraelectric phase of classical ferroelectrics like (Ba,Sr)TiO3 and BaTiO3, despite the nominal cubic symmetry?
• RQ2 What is the atomic-scale structure and symmetry of these nanoclusters, and how do they differ from the average cubic structure?
• RQ3 Are the observed polar nanoclusters dynamically fluctuating or statically ordered, as implied by macroscopic polarization measurements?
• RQ4 How do the displacements of B-site cations (Ti) and oxygen anions deviate from ideal cubic positions, and what is their magnitude and spatial correlation?
• RQ5 To what extent do surface effects and beam entry influence STEM-based displacement measurements in thin films?

Key findings

• Direct visualization of 2–4 nm large polar nanoclusters in the nominally cubic paraelectric phase of (Ba0.6Sr0.4)TiO3 using aberration-corrected STEM.
• Measured Ti displacements from ideal cubic positions reach up to ~15 pm, with O anions displaced by ~12.5 pm, indicating strong local symmetry breaking.
• Raman spectroscopy reveals static, disorder-activated modes at ~220 cm⁻¹ and ~550 cm⁻¹, confirming long-range polar order in the paraelectric phase.
• Lattice strain maps show maximal strain at interfaces between polar nanoclusters, indicating elastic coupling and coherent atomic displacements.
• Complex elastic modulus measurements show a pronounced drop in storage modulus (E') near the Curie temperature, indicating the presence of elastically active nanoscale objects.
• Simulations confirm that the top 4–5 nm surface layer dominates image formation and displacement measurements, necessitating caution in interpreting bulk-like behavior from thin-film STEM data.