In a decade-long quest, scientists at Berkeley Lab, the University of Hawaii, and Florida International University uncover new clues to the origins of the universe – and land new chemistry for cleaner combustion engines
Connected moments math shortcut shaves time and cost of quantum calculations while maintaining accuracy
Molecular dynamics is central to many questions in modern chemistry. However, computer models of molecular dynamics must balance computational cost and accuracy. Scientists have now used a machine learning technique called transfer learning to create a novel model of molecular motion that is as accurate as calculations that use quantum-mechanical physics but much faster.
Chemical reactions transform reactants to products through intermediate states. These intermediates are often short-lived, making them hard to study. But by bringing a molecule to a temperature barely above absolute zero, scientists can “trap” the reaction in the intermediate stage for a much longer time. In this study, scientists used photoionization to directly observe a reaction’s reactants and products.
As a staff member in the Theory and Computation Group at Brookhaven Lab’s Center for Functional Nanomaterials, Qu applies various approaches in artificial intelligence to analyze experimental and computational nanoscience data.