An intuitive algorithm developed at the Center for Nanophase Materials Sciences of Oak Ridge National Laboratory is being used to teach microscopes to drive discoveries. This could guide breakthroughs in new materials for sensing, energy technologies, and computing.
In material science, there are a large number of potential substances that cannot be studied with conventional tools, and require more systematic and efficient approaches to be designed and synthesized.
The exploring new materials can be undertaken using smart automation, and a shareable, reproducible path can be created for discoveries that have not been previously possible.
The method published in Nature Machine Intelligence combines machine learning and physics to automate microscopy experiments designed to study functional properties of materials at the nanoscale.
In fact, functional materials are responsive to external stimulus such as electricity or heat, and are engineered to support everyday as well as emerging technologies that range from solar cells and computers to artificial muscles and shape-memory materials.
The unique properties of functional materials are associated with atomic structures and microstructures that can be examined with advanced microscopy. However, developing efficient methods to locate regions where these properties emerge and can be examined has been challenging.
Importantly, scanning probe microscopy is an essential instrument for understanding structure-property relationship in functional materials. The instrument that is used scans the surface of materials with an atomically sharp probe to chart the structure at the nanometer scale.
In addition, scanning probe microscopy can also discover response to a range of stimuli, and provide insights on fundamental mechanisms of polarization switching, quantum phenomena, electrochemical reactivity, or plastic deformation.
Microscopes that are available today can perform a point-to-point scan of a nanometer square grid, but can be painstakingly slow, wherein measurements can be collected over days for one single material.
