Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 35: Poster Solid-Liquid Interfaces: Structure
O 35.3: Poster
Dienstag, 18. März 2025, 13:30–15:30, P3
Informed Automated Structure Discovery of Atomic Force Microscopy Images — •Azin Alesafar1, Joakim Jestilä1, and Adam Foster1,2 — 1Department of Applied Physics, Aalto University, Espoo, Finland — 2Nano Life Science Institute (WPI-NanoLSI), Kanazawa University, Kanazawa, Japan
Atomic Force Microscopy (AFM) enables direct imaging of atomic-level features however, the interpretation of non-planar molecules is challenging due to the fact that only the top layers of these systems interact with the microscope tip. This leads to images deviating from structures familiar to us. Recent Advances in machine learning-based image recognition tools have provided a framework suited to tackle this challenge. However, these methods rely heavily on training data and may produce inaccurate results when faced with unfamiliar structures. An alternative approach is to develop an iterative algorithm that generates realistic 3D structures by comparing simulated and experimental AFM images in a fully automated manner. The final workflow enables the generation of candidate structures using different techniques, such as molecular dynamics, minima hopping, or machine learning models. A deeper understanding of the simulated structural information is achieved through feature detection algorithms and image registration. Furthermore, the simulated structures, and consequently their corresponding AFM images, are automatically evaluated for similarity to reference AFM images using image quality metrics. These approaches are tested on water clusters modeled on gold and copper surfaces using the Neural equivariant interatomic potential (NequIP).
Keywords: AFM; Image Quality Metrics; Feature Detection; Image Registration; Minima Hopping