Overview
Our group uses transmission electron microscopy and electron energy loss spectroscopy to understand nanomaterials and devices at the atomic scale. Using recent advances in aberration-corrected electron microscopy, we aim to image the structure, bonding, electronic, optical properties of materials with unprecedented precision. Together, this research aims to harness the incredible insights of atom-by-atom microscopy to design novel nanomaterials and devices.
Advanced electron microscopy methods: ptychography and more
How do we push electron microscopy to extract structural, chemical, and dynamic behavior of materials with single atom precision? Our group develops new microscopy techniques, such as electron ptychography, 4D Scanning Transmission Electron Microscopy, and electron spectroscopy. For example, we have developed methods that might make electron microscopes cheaper and more accessible (Science 2024).
Machine Learning and Advanced Data Methods for Electron Microscopy
To interpret these data, we often reach across disciplines to borrow techniques from machine learning, biological imaging, and computational imaging to develop new algorithms for processing electron microscopy and spectroscopy data. We have used these algorithms to reconstruct the 3D structure of materials a single layer thick, pick out and measure the statistical distributions of single dopants, and track each atom in a glass as it bends and breaks. Recently, we used generative AI to train machine learning models to identify and classify millions of atoms in electron microscopy data (npj Computational Materials 2023).
2D and Quantum materials
Low-dimensional materials such as atomic layers of graphene, molybdenum disulfide, black phosphorous, and tungsten ditelluride are leading to new designs for high-mobility, high-efficiency nanoelectronic and nanophotonic devices. In these systems, control at the single atom level is critical because defects, dopants, and interfaces have dramatic effects on the properties of materials only few atoms thick. 2D materials can be strongly tuned through defect or interfacial engineering to create and pattern atomically-thin circuitry, engineer low-power phase change memories, and produce high-efficiency absorbers and emitters of light. We use electron microscopy and spectroscopy, paired with advanced methods in computation and data processing, to probe the quantum phenomena that originate from individual atomic-scale defects and interfaces. By correlating atom-by-atom imaging and spectroscopy with in- and ex-situ device measurements, we can trace a direct link between atomic structure and chemistry, microstructure, and device performance. Examples of the impact of our work include: new designs for scalable, high-mobility graphene devices and 2D heterostructures with atomically precise interfaces. For example, we developed a unifying theory of bending in 2D materials and used it to design ultra-soft, bendable 2D electronics (Nature Materials 2020, Advanced Materials 2021). We also study how to design the properties 2D materials, such as through strain by tailoring moiré interfaces (Science Advances 2024).
Materials for quantum computing
Atomic defects can be sources of inefficiencies and losses in qubits, or they can be used as bright emission sources for quantum sensing. In many cases, the exact nature of the defects responsible is unknown. Using the tools of advanced electron microscopy, we seek to understand and control defects on the atomic level in order to create better materials for quantum computing.