NCAFM2023 Programme Booklet

Wednesday 1640 - 1700

IMAGE INTERPRETATION METHODS FOR HIGH RESOLUTION SPM

Lauri Kurki 1 , Niko Oinonen 1 , Ondřej Krejčí 1 , Shigeki Kawai 2,3 , Adam S. Foster 1,4

1 Department of Applied Physics, Aalto University, Espoo, Finland 2 Research Center for Advanced Measurement and Characterization, National Institute for Materials Science, Tsukuba, Japan

3 Graduate School of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Japan 4 WPI Nano Life Science Institute (WPI-NanoLSI), Kanazawa University, Kakuma-machi, Japan

Email: lauri.1.kurki@aalto.fi Scanning tunnelling microscopy (STM) and atomic force microscopy (AFM) functionalized with a CO molecule on the probe apex are methods capable of capturing sub-molecular level detail of the electronic and physical structures of a sample [1]. While high-resolution STM is a widely adopted method in materials science, the produced images are often difficult to interpret due to the convoluted nature of the signal. We propose image interpretation tools to extract physical information directly from STM images using machine learning. In recent years, there has been rapid development in image analysis methods using machine learning, with particular impact in medical imaging. These concepts have been proven effective also in SPM in general and in particular for extracting sample properties from atomic force microscopy (AFM) images [2,3,4]. We build upon these models and show that we can extract atomic positions directly from STM images. We also further explore how the accuracy of these predictions varies with the use of a simultaneous AFM signal and finally establish the limits of the approach in an experimental context by predicting atomic structures from STM images of silico-organic compounds [5].

Fig. 1 An example prediction from a simulated STM image of a C8O3H6NF molecule. We predict the (a) atomic disks, (b) vdW spheres and (c) height map descriptors [2]. References [1] Cai, S., Kurki, L., Xu, C., Foster, A. S., Liljeroth, P. Water Dimer-Driven DNA Base Superstructure with Mismatched Hydrogen Bonding. J. Am. Chem. Soc. 2022, 144, 44, 20227–20231 [2] Alldritt, B., Hapala, P., Oinonen, N., Urtev, F., Krejci, O., Canova, F. F., Kannala, J., Schulz, F., Liljeroth, P., Foster, A. S. Automated structure discovery in atomic force microscopy. Sci. Adv. 2020; 6 : eaay6913 [3] Carracedo-Cosme, J., Romero-Muñiz, C., Pérez, R. A Deep Learning Approach for Molecular Classification Based on AFM Images. Nanomaterials 2021, 11, 1658. [4] Oinonen, N., Kurki, L., Ilin, A., Foster, A. S. Molecule graph reconstruction from atomic force microscope images with machine learning. MRS Bulletin 2022, 47, 895-905 [5] Sun, K., Silveira, O. J., Ma, Y., Hasegawa, Y., Matsumoto, M., Kera, S., Krejčí, O., Foster, A. S., Kawai, S. On-surface synthesis of disilabenzene-bridged covalent organic frameworks. Nature Chemistry 2022. 15, 136-142

65