SKM 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 77: Scanning Probe Techniques: Method Development II
O 77.5: Vortrag
Donnerstag, 30. März 2023, 11:30–11:45, REC C 213
Artificial Intelligence finds the optimal STM manipulation parameters of unknown molecules — •Bernhard Ramsauer1, Grant J. Simpson2, Leonhard Grill2, and Oliver T. Hofmann1 — 1Institute of Solid State Physics, NAWI Graz, Graz University of Technology, Petersgasse 16, 8010 Graz, Austria — 2Department of Physical Chemistry, Institute of Chemistry, NAWI Graz, University Graz, Heinrichstraße 28, 8010 Graz, Austria
Scanning probe microscopy gives us the possibility to precisely control the position and orientation of single molecules and unlocks the possibility of nanofabrication of novel structures with enhanced properties. However, interaction processes at the nanoscale are stochastic processes, and because their motion itself is often unintuitive and hard to predict, inducing controlled movements is not trivial at all.
In this study we present how a reinforcement learning algorithm identifies optimal manipulation parameters (i.e., the bias voltage, height, and lateral position of the STM tip relative to the molecule) for any unknown molecule to allow for precise control of its movement.
Leaving all the manipulation parameters open for investigation requires a method to exclude manipulation parameter that pick-up or destroy the molecule. Furthermore, already learned information is used to infer prior knowledge to similar manipulation parameters (e.g.: to infer knowledge at the same bias voltage to adjacent tip positions).
This allows for autonomous control of initially unknown molecules with high sub-nanometer precision that set the basis to construct molecular nanostructures from the bottom up.