Regensburg 2025 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
O: Fachverband Oberflächenphysik
O 17: Poster Focus Session Molecular Nanostructures on Surfaces: On-Surface Synthesis and Single-Molecule Manipulation
O 17.4: Poster
Monday, March 17, 2025, 18:00–20:00, P2
Autonomous chemical reactions in scanning tunneling microscopy — •Nian Wu1, Peter Liljeroth1, and Adam S. Foster1, 2 — 1Department of Applied Physics, Aalto University, Espoo, Finland — 2WPI Nano Life Science Institute, Kanazawa University, Kanazawa, Japan
Scanning tunneling microscopy (STM) has shown great promise in manipulating atoms or molecules in on-surface molecular synthesis. However, the selection of proper parameters for various manipulations requires extensive explorations and strongly depends on domain knowledge. In particular, the coupling of fragments, as a critical step for polymerizations, necessitates more precise control in rotations and distances. To address this problem, we designed a deep reinforcement learning approach to automate the C-C coupling from Zn(II)-5,15-bis(4-bromo-2,6-dimethylpheny)porphyrin (ZnMe4DPP2.) through learning manipulation parameters in STM, thus enabling the assembly of large polymers based on the building blocks.
Keywords: automation; nanofabrication; scanning tunneling microscope; on-surface synthesis; deep reinforcement learning