Hye Jung Chang

Suppression of Octahedral Tilts and Associated Changes in Electronic Properties at Epitaxial Oxide Heterostructure Interfaces

Epitaxial oxide interfaces with broken translational symmetry have emerged as a central paradigm behind the novel behaviors of oxide superlattices. Here, we use scanning transmission electron microscopy to demonstrate a direct, quantitative unit-cell-by-unit-cell mapping of lattice parameters and oxygen octahedral rotations across the BiFeO3-La0.7 Sr0.3 MnO3 interface to elucidate how the change of crystal symmetry is accommodated. Combined with low-loss electron energy loss spectroscopy imaging, we demonstrate a mesoscopic antiferrodistortive phase transition near the interface in BiFeO3 and elucidate associated changes in electronic properties in a thin layer directly adjacent to the interface.

Mapping Octahedral Tilts and Polarization Across a Domain Wall in BiFeO3 from Z-Contrast Scanning Transmission Electron Microscopy Image Atomic Column Shape Analysis

Oxygen octahedral tilts underpin the functionality of a large number of perovskite-based materials and heterostructures with competing order parameters. We show how a precise analysis of atomic column shapes in Z-contrast scanning transmission electron microscopy images can reveal polarization and octahedral tilt behavior across uncharged and charged domain walls in BiFeO(3). This method is capable of visualizing octahedral tilts to much higher thicknesses than phase contrast imaging. We find that the octahedral tilt transition across a charged domain wall is atomically abrupt, while the associated polarization profile is diffuse (1.5-2 nm). Ginzburg-Landau theory then allows the relative contributions of polarization and the structural order parameters to the wall energy to be determined.

Atomically Resolved Mapping of Polarization and Electric Fields Across Ferroelectric/Oxide Interfaces by Z-contrast Imaging

Direct atomic displacement mapping at ferroelectric interfaces by aberration corrected scanning transmission electron microscopy(STEM) (a-STEM image, b-corresponding displacement profile) is combined with Landau-Ginsburg-Devonshire theory to obtain the complete interface electrostatics in real space, including separate estimates for the polarization and intrinsic interface charge contributions.

Defect-Mediated Polarization Switching in Ferroelectrics and Related Materials: From Mesoscopic Mechanisms to Atomistic Control

The plethora of lattice and electronic behaviors in ferroelectric and multiferroic materials and heterostructures opens vistas into novel physical phenomena including magnetoelectric coupling and ferroelectric tunneling. The development of new classes of electronic, energy-storage, and information-technology devices depends critically on understanding and controlling field-induced polarization switching. Polarization reversal is controlled by defects that determine activation energy, critical switching bias, and the selection between thermodynamically equivalent polarization states in multiaxial ferroelectrics. Understanding and controlling defect functionality in ferroelectric materials is as critical to the future of oxide electronics and solid-state electrochemistry as defects in semiconductors are for semiconductor electronics. Here, recent advances in understanding the defect-mediated switching mechanisms, enabled by recent advances in electron and scanning probe microscopy, are discussed. The synergy between local probes and structural methods offers a pathway to decipher deterministic polarization switching mechanisms on the level of a single atomically defined defect.

Imaging of Light Atoms in the Presence of Heavy Atomic Columns

Aberration correction in the scanning transmission electron microscope (STEM) has led to increased sensitivity and the ability to detect light atoms using annular dark field (ADF) imaging [1]. However, in most cases this requires the light atomic columns to be separated from heavier columns by a significant distance. Here we use Bloch wave simulations [2] to examine the visibility of light atoms situated close to heavier atomic columns using both ADF and electron energy loss spectroscopy (EELS).

Interface Structure-Property Relations Through Aberration-Corrected STEM

The aberration-corrected STEM provides quantitative information on atomic positions, species and chemical bonding with new levels of precision and sensitivity. Multiple signal channels are available, some simultaneously, including the high-angle ADF signal, the phase contrast signal with an axial detector, an annular bright field signal [1] and spectroscopic images. Such data can be directly compared with density functional calculations to reveal the atomistic origins of interfacial properties.

Nano-scale Shell in Phase Separating Gd-Ti-Al-Co Metallic Glass

In the present study, formation of yard and shell has been investigated in as-melt-spun alloy using a variety of transmission electron microscopy techniques. The phase separation during cooling leads to the formation of the microstructure consisting of amorphous droplets with different size scales embedded in the amorphous matrix. Due to the interdiffusion at the interface after the first-step phase separation, ~50 nm-thick yard develops on the surface of the primary droplet particle. Due to the critical wetting phenomenon, ~5 nm thickness shell enveloping the droplet forms. The sell is enriched in Co and Ti, implying that the composition is close to that of the droplet.

Uncovering Interface Structure by Column Shape Analysis in ADF STEM Images

The multifaceted magnetic, electrical, and structural functionalities of perovskite ABO3 materials are underpinned by the subtle distortions of the crystallographic lattice from cubic prototype. These distortions include relative displacements of the cations from the centers of the BO6 oxygen octahedra, deformations of oxygen octahedra, and collective tilts of octahedral network. Further interest in this behavior has emerged from the fabrication of novel heterointerfaces that allow new functionalities that do not exist in the bulk, including interface superconductivity, improper ferroelectricity, and magnetoelectric coupling phenomena. Furthermore, at homophase interfaces such as ferroelastic domain walls, symmetry changes alone often give rise to new properties.

Towards the Thinking Microscope

The capability of electron microscopy and spectroscopy to examine structure and chemical composition from mesoscopic to atomic scales makes it the ultimate characterization tool for condensed matter physics and materials science. In particular, electron energy loss spectroscopy (EELS) in the scanning transmission electron microscope (STEM) allows for spatially resolved studies of composition, electronic and magnetic structure (in the core-loss region), as well as elementary excitations such as plasmons (in the low-loss region). Ideally, structural information obtained via STEM imaging (bright-field and dark field) can be synergistically combined with EELS spectroscopic data, providing a direct correlation between structure and electronic functionality. With aberration-corrected electron optics and improved EELS hardware, such correlations can be examined in unprecedented detail. However, progress in this field is strongly hindered by the critical bottleneck of analysis and interpretation of the 3and higher dimensional datasets (spectrum images) with millions of data points, their registration with the 2D structure images, and interpretation in terms of desired materials functionality. The current method of analyzing such data relies on a large amount of human input on every step of the process. This not only makes the process of analyzing the data extremely time consuming but also tends to introduce error as well as user bias into the data analysis. By automating the process of data analysis it will be possible to increase efficiency and reduce artifacts. In this presentation, I will describe the automated for STEM and EELS data classification and analysis based on the combination of grid search and data mining algorithms. As a first step, I have developed an approach for automatic indexing the atomic lattices. The local maxima are identified using erosion-expansion algorithm, and classified in a operator-defined number of classes (e.g. corresponding to the A and B cationic sublattices in perovskites). The reliably-determined lattice sites are used to determine the most likely lattice vectors, and then adaptive algorithm is used to find the remaining atoms. The example of this algorithm is shown in Figure 1. Thus indexed lattice is analyzed using position refinement using appropriate shape function (e.g. pseudoVoigt) to extract the atomic spacings, from which local strains (e.g. variation in lattice parameters), polarization (relative displacements of A and B sublattice), and octahedral tilts (from bright field – dark field image pair) can be established. The atomic column shape is analyzed using shape PCA analysis to provide information on the octahedral distortions [1]. Currently, we are working on developing the algorithm to match the 2D image data to the 3D EELS data to enable direct recognition of atomic structures and fine details of electronic properties on atomic level, as exemplified by Figure 2 [2,3].