January_2021_AMP_Digital

1 0 A D V A N C E D M A T E R I A L S & P R O C E S S E S | J A N U A R Y 2 0 2 1 IMPROVING DEVICE TOUCHSCREENS Scientists at The University of Tokyo have used electron energy loss fine structure spectroscopy with a scan- ning transmission electron microscope to reveal the local arrangement of atoms within a glass made of 50% alu- minum oxide and 50% silicon dioxide. “We chose to study this system because it is known to phase separate into alu- minum-rich and silicon-rich regions,” says researcher Kun-Yen Liao. When imaging with an electron microscope, some electrons undergo inelastic scat- tering, causing them to lose some of their kinetic energy. The amount of energy dissipated varies on the location and type of atom or cluster of atoms in the glass sample. Electron loss spectroscopy is sensitive enough to tell the difference between aluminum coordinated in tetrahedral as opposed to octahedral clusters. By fitting the profile of the electron energy TESTING | CHARACTERIZATION REIMAGINING ELECTRON MICROSCOPY Scientists are reimagining the tools used for work in atomic research. Published as a commentary in Nature Materials, an international team led by Pacific Northwest National Laboratory (PNNL), Richland, Wash., detailed a vision for electron microscopy infused with the latest advances in data science and artificial intelligence. They pro- posed a highly integrated, autonomous, and data-driven microscopy architec- ture to address challenges in energy storage, quantum information science, and materials design. This approach can provide new insight into materials properties, permitting experiments at a vast scale not possible today. In their article, the research team proposes to inte- grate artificial intel- ligence and machine learning into each step of the micros- copy workflow. They argue strongly for the development of a new microscopy infra- structure—advanced via multi-institute na- tional technology ini- tiatives—that would make data more ac- cessible to research organizations world- wide. In this way, past measurements could be used to select techniques and interpret results in the moment, informing autonomous decision-making algorithms. Large li- braries of past experiments could be used to highlight latent features and offer guidance for the user. A strength of this “crowdsourced” approach is that it is highly scalable, less prone to operator bias, and more repeatable, which will translate into improved results. The team’s vision is the result of the first Next-Generation Transmission Electron Microscopy workshop held at PNNL in 2018. Ultimately, the team hopes to show the immense, untapped potential of data science and its critical role in unlocking the full power of elec- tron microscopy. pnnl.gov. A new research effort at MIT, Cambridge, Mass., aims to advance predictive simulation. The Center for Exascale Simulation of Materials in Extreme Environments (CESMIX) will bring together researchers in materials science, numerical algorithms, quantum chemistry, and computer science. The goal is to connect quantum and molecular simulations of materials with advanced programming languages, compiler technologies, and software performance engineering tools. CESMIX will initially focus on exascale simulation of materials in hypersonic flow environments. news.mit.edu. BRIEF Infusing data science and AI into electron microscopy opens new possibilities in the science of imaging. Courtesy of Timothy Holland/PNNL. Aluminosilicate glass used in smartphone applications. Courtesy of SystemElecktronik.