75 years of science and technology
Integrating electron microscopy and atomic imaging with big data technologies is a monumental task, but the end result is a deeper, more powerful understanding and control over materials functionality at the atomic level.
That understanding is what attracted Sergei Kalinin, a researcher at the Center for Nanophase Materials Sciences.
The steady progress in microscopy in the last decade has “opened the floodgates of high-veracity structural information and has offered insights into physical and chemical functionalities on an atomic level,” Kalinin said.
Working at ORNL, Kalinin saw the promise of high-performance computing to drive machine learning and artificial intelligence in materials research, making sense of the enormous amount of data captured by electron and probe microscopes. These methods not only generate high-resolution images but can also help researchers understand the underlying physics, which in turn will help computational researchers create higher-fidelity simulations.
The simulations go beyond imaging and can help researchers control and direct matter atom by atom with an atomically focused electron beam, creating atomic configurations optimized for specific physical and chemical properties.
The knowledge gained through this merger can be used to improve material functionalities, improve synthesis, and design and predict novel materials. The electron-beam-based, atom-by-atom fabrication techniques developed by Kalinin and
others at ORNL can even open pathways for applications in quantum computing, single-spin magnetoelectronics and atomic robotics.
“These two directions—materials fabrication and understanding—are closely intertwined,” Kalinin said. “Only if we understand material on the atomic level will we know what we need to fabricate and how to do it.”
