Berlin 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
O: Fachverband Oberflächenphysik
O 4: New Methods: Experiments
O 4.6: Talk
Monday, March 18, 2024, 11:45–12:00, MA 005
Autonomous nanoARPES Experiments — •Steinn Ýmir Ágústsson1, Alfred J. H. Jones1, Davide Curcio1, Søren Ulstrup1, Jill Miwa1, Davide Mottin2, Panagiotis Karras2, and Philip Hofmann1 — 1Dpt. of Physics and Astronomy, Aarhus University — 2Dpt. of Computer Science, Aarhus University
Angle-resolved photoemission spectroscopy (ARPES) is a technique used to map the occupied electronic structure of solids. Recent progress in X-ray focusing optics has lead to the development of ARPES into a microscopic tool, permitting the electronic structure to be mapped across the surface of a sample. This comes at the expense of a time-consuming scanning process to cover the whole surface.
We implemented a protocol which leverages Gaussian Processes to autonomously search the surface area in order to find positions of particular interest, based exclusively on the observed spectra. The protocol promises significant efficiency gains by avoiding redundant measurements and maximizing information gain from the data already measured. Furthermore, it can easily be expanded to explore a larger parameter space, including temperature or external perturbations.
The autonomous experimental control is implemented on the SGM4 micro-focus beamline of the synchrotron radiation source ASTRID2, where pilot experiments were used to quickly identify regions of interest to study in further detail. The successful implementation of the protocol shows the value of machine learning in the context of controlling complex experiments, and the potential for further development of autonomous experiments.
Keywords: ARPES; Machine Learning; Artificial Intelligence; Autonomous Experiments; Gaussian Processes