Deep Learning in Atomically Resolved Imaging: From Mechanisms of Solid-State Reactions to E-Beam Atomic Fabrications
Abstract: Understanding the elementary mechanisms behind solid-state phase transformations and reactions is the key to optimizing the desired functional properties of many technologically relevant materials. Recent advances in scanning transmission electron microscopy (STEM) allow the real-time visualization of solid-state transformations in materials, including those induced by an electron beam and temperature, with atomic resolution. However, despite the ever-expanding capabilities for high-resolution data acquisition, the inferred information about kinetics and thermodynamics of the process and single-defect dynamics and interactions is minimal, because of the inherent limitations of manual ex situ analysis of the collected volumes of data and lack of feedback to theory.
In this presentation, I discuss research coordinated by the Institute for Functional Imaging of Materials aimed at bridging imaging and theory via big data technologies to enable design of new materials with tailored functionalities. I will illustrate several examples of using deep learning networks and inverse modelling to extract materials specific physics from imaging in STM and STEM.
For atomistic systems, we developed a deep learning framework for dynamic STEM imaging that is trained to find the structures (defects) that break a crystal lattice's periodicity and apply it to map solid-state reactions and transformations in layered WS2 doped with molybdenum. This framework allows extracting thousands of lattice defects from raw STEM data (single images and movies) in a matter of seconds, which are then classified into different categories by using unsupervised clustering methods.
We further expanded our framework to extract parameters of diffusion for the sulfur vacancies and analyzed transition probabilities associated with switching between different configurations of defect complexes consisting of molybdenum dopant and sulfur vacancy, providing insight into point defect dynamics and reactions. This approach is universal, and its application to beam-induced reactions allows mapping chemical transformation pathways in solids at the atomic level. Several additional examples of deep-learning analytics of molecular systems, theory-assisted image recognition, and extraction of mesoscopic and atomistic physical parameters will be illustrated.
Finally, incorporation of the real time feedback in electron microscopy opens the pathway toward the use of atomically focused STEM beam to control and direct matter on atomic scales. I will introduce several examples of beam-induced fabrication on the atomic level and demonstrate how beam control, rapid image analytics, and image- and ptychography-based feedback allows for controlling matter on the atomic level.