 |
Center for Nonlinear Studies |
This talk highlights the advancement of two computational methodologies developed to analyze the mechanisms of optoelectronic 2D nanoscale devices, with a particular focus on devices based on monolayer graphene. The first methodology employs a ML potential neural network called TorchANI to investigate atomic-level interactions, advancing our knowledge of molecular non-covalent adsorption phenomena. This is achieved by simulating over 1.6 million initial configurations to delineate metastable ground state energy structures. The second methodology integrates density functional theory with machine learning to optimize the design of 2D sensors for detecting ammonia, for environmental monitoring. Both approaches have been validated through experimental surface characterizations. The purpose of these models is to enhance the selection of polymers for polymer-assisted transfer of CVD-grown graphene and improving the screening of sensing devices.
Host: Ben Nebgen (T-1)