The advancements in aberration correctors, pixelated direct electron detectors, and monochromators represent significant milestones in the development of transmission electron microscopy (TEM). These innovations have enabled the imaging of materials' structure, chemistry, and functional properties at the atomic scale. In this talk, I will introduce a novel four-dimensional scanning transmission electron microscopy (4D STEM) method. This method facilitates the imaging of local polarization, electric fields, and charge density in multiferroic oxide nanostructures with sub-angstrom resolution. I will demonstrate how polarization at ferroelectric/insulator interfaces influences functional properties, such as strain, bonding, electric field and charge distributions. Through the combination of 4D STEM and scanning probe microscopy, we have observed unique skyrmion-like polar nanodomains in freestanding PbTiO3/SrTiO3 bilayers transferred onto silicon. These nanodomains can be toggled between states by an applied electric field, leading to substantial alterations in their resistive behaviors. I will present some new phenomena of ferroelectric interfaces revealed by novel space- and angle-resolved electron energy-loss spectroscopy (EELS) methods. I will show the phonon dynamics at interfaces between ferroelectric and insulating oxides. These novel methodologies hold valuable potential for studying real nanodevices and enhancing our comprehension of charge distribution and heat dissipation in nanostructures and interfaces.

Extraordinary physical properties arise at ferroelectric domain walls, where electronic reconstruction phenomena can be driven by bound charges. Great progress has been achieved in the characterization of such domain walls and first tomography techniques have been established, providing unprecedent insight into their 3D structure. Here, we present a non-destructive strategy for tomographic nanoscale imaging of ferroelectric domain walls using secondary electrons. Utilizing conventional scanning electron microscopy (SEM), we show that it is possible to reconstruct the position, orientation, and charge state of hidden domain walls located at distances up to several hundreds of nanometers beneath the surface. A mathematical model links the SEM intensity variations at the surface to the local domain wall properties, enabling non-destructive tomography with good noise tolerance on the timescale of seconds. Our SEM-based approach facilitates high throughput screening of materials with functional domain walls and domain-wall-based devices. This is crucial for real-time monitoring during the production of device architectures and quality control.

Science traditionally uses data for decision-making. Previously, data was manageable for human analysis. However, the emergence of advanced sensing technologies has led to a flood of large, fast-moving data from varied sources, overwhelming traditional analysis methods. Despite advancements in AI and models like ChatGPT, their limitations are evident. AI excels in interpolation but struggles with extrapolation, often producing unrealistic outcomes when beyond their training data's scope. The rush towards AI adoption has created a misleading perception of its potential in ferroics, overshadowing significant technical innovations and transformative insights.
We delve into the convergence of massive data growth and artificial intelligence, underscoring their combined strengths and weaknesses in augmenting decision-making processes, especially in the realm of data-driven infrastructure. We explore the co-design of experimental systems, encompassing algorithms, comprehensive software solutions, and hardware, all aimed at actualizing scientific machine learning.
We then tackle the challenges inherent in applying machine learning within the context of ferroelectrics, where concepts like order, symmetry, and periodicity form essential semantic relationships. Our discussion extends to the development of high-availability computational frameworks designed for deploying resilient, self-repairing services in science, thus facilitating materials science at the exascale. Additionally, we spotlight innovations in parsimonious neural networks, adept at learning geometric transformations in reciprocal space. We emphasize the role of stochastic averaging in enhancing noise robustness, surpassing traditional algorithmic approaches. We apply these techniques to a variety of common synthesis and characterization techniques in ferroelectrics.
Finally, we examine the progress in AI co-design, wherein algorithms are optimized for programmable logic (e.g., FPGAs) for real-time <1 ms inference. This optimization enables rapid, intelligent analysis, decision-making, and control on ultra-low-cost, low-power devices at unparalleled speeds. We illustrate how this approach is instrumental in real-time data analysis, data reduction, and dose-controlled imaging across various scientific platforms, including electron microscopy.