Regensburg 2022 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
MM: Fachverband Metall- und Materialphysik
MM 22: Data Driven Materials Science: Experimental Data Treatment and Machine Learning
MM 22.3: Vortrag
Mittwoch, 7. September 2022, 11:00–11:15, H46
Correcting density artifacts in Atom Probe reconstructions: A tip shape-corrected volume reconstruction approach — •Patrick Stender1, Daniel Beinke1, Felicitas Bürger2, and Guido Schmitz1 — 1Institute for Materials Science, University of Stuttgart — 2Fakultät für Mathematik, Universität Regensburg, D-93040 Regensburg, Germany
Atom Probe Tomography enables the chemical investigation of nanometric volumes with single atomic sensitivity in 3D. The tip shape sample is evaporated atom by atom. From the obtained data sequence, the respective volume is reconstructed.
Conventionally, this reconstruction is performed with the assumption of a hemispherical tip apex. This practice can lead to serious volume distortions (local-magnification effect). Instead of using in-situ correlative microscopy to discover the evolution of the tip shape during the measurement, we extract the emitter shape numerically from the event statistics on the 2D detector plane.
The method is based on the fundamental postulate that the detected density of events is linked to the local Gaussian curvature of the tip apex. Knowing the variation of this curvature, the surface profile is determined by a finite difference scheme. Except for convexity, no further restriction is imposed on the possible tip shapes.
Different simulated and experimental data sets of complex tip shapes will be discussed and compared. The method largely suppresses the local magnification effects appearing at interfaces between materials of contrasting evaporation thresholds.