Round robin on analysis of clusters in atom probe tomography data
Current and past team members
On the round robin experiment, our collaborators in the round robin include Auriane Etienne & Bertrand Radiguet (Rouen), Alex Frolov & Svetlana Fedotova (Kurchatov Institute), Katsuhiko Fujii & Koji Fukuya (INSS), Anabelle Lopez (CEA), Andrew London, Sergio Lozano-Perez (Oxford), Yasuyoshi Nagai &Takeshi Toyama (Tohuku), Kenji Nishida &Naoki Soneda (CRIEPI), Daniel Schreiber (PNNL), Mattias Thuvander (Chalmers).
On the modeling of field evaporation, our collaborators include Wolfgang Windl and Christian Oberdorfer (OSU).
Understanding artefacts & defining error & uncertainty: modeling APT data
The objective of this project is the development of modeling based methods designed to improve the reliability of atom probe tomography data. The effort is split into three main tasks:
– The development of a modular simulation platform or virtual APT simulator that will include physics of evaporation, electrostatic field simulator, trajectory calculations, and a reconstruction module;
– An experimental component to generate and analyze specific aspects of the data generation procedure: mass spectrum ranging, atom position assignment, and interface reconstruction;
– An optimization routine that compares the experimental data (APT and other) to the simulated data and determines the best-possible reconstruction in an iterative process.
Previously atoms were selected based on location of the highest applied field ignoring any effect of surface structure. This simplistic assumption would not be appropriate when simulating defects or atoms of different chemistry. We introduced a more physical approach with the selection of evaporation events based on density functional theory calculations. Applying it to pure Al, we showed that local surface structure strongly influences the strength of the evaporation field required to overcome the energy barrier for field evaporation. The model reproduces a number of evaporation characteristics observed experimentally that were not reproduced by earlier simplistic models. These include spatial resolution degradation, density patterns formed by evaporated ion on the detector.
Combining Oberdorfer new combined TAPSIM+LAMMPS simulation package with experimental measurements, we focused on a dilute Al-Cu alloy to understand the non-uniform reconstruction of Cu atoms from random solid solution specimens. A traditionally used explanation relies on the diffusion of species driven by the non-uniform field distribution on the surface or the uncontrolled evaporation of species that are therefore not detected. Both mechanisms would result in locally high density of solutes with crystallographic dependence. Instead, we found a third possible mechanism that does not rely on field distribution or detector deficiencies. We showed that binding and atomic surface morphology can drive Cu atom mobility to specific terraces.
Solute clustering in nuclear and structural materials is commonly analyzed using atom probe tomography. The simple maximum separation method remains the most common data analysis procedure used, even though not all microstructures are amenable to its application. Moreover, the open literature reveals significant reporting gaps, raising concerns from data interpretation, validity, quality, to reproducibility. We address these concerns through two parallel efforts. A round robin experiment, conducted among ten international research groups and led by Prof Marquis, brings awareness about algorithm limitations, community testing and vetting, and the ethical question of data reproducibility and openness in a field with significant financial and social consequences. The outcomes will be recommendations for best practices and a data reporting template that will assist future researchers in need of re-interpretation of published data. In parallel, we explore alternative analysis methods, including hierarchical density-based cluster analysis methods, well developed in computer science, as well as simulation tools for APT data.
- Quantitative atom probe tomography characterization of microstructures in a proton irradiated 304 steel, Y Chen, P Chou, EA Marquis, J. Nuclear Materials (2014) 451 130-136 link
- Effects of the local structure dependence of evaporation fields on field evaporation behavior, L Yao, T Withrow, O Restrepo, W Windl, EA Marquis, Applied Physics Letters 107 (2015) 241602
- Three dimensional imaging of shear bands in bulk metallic glass composites, AH Hunter, M Gibbons, V Araullo-Peters, OD Restrepo, SR Niezgoda, W Windl, KM Flores, DC Hofmann, EA Marquis, Journal of Microscopy 264(3):304-310 (2016) link
- On the use of density-based algorithms for the analysis of solute clustering in atom probe tomography data. EA Marquis, V Araullo-Peters, Y Dong, A Etienne, S Fedotova, K Fujii, K Fukuya, E Kuleshova, A Lopez, A London, S Lozano-Perez, Y Nagai, K Nishida, B Radiguet, D Schreiber, N Soneda, M Thuvander, T Toyama, F Sefta, P Chou,In: Jackson J., Paraventi D., Wright M. (eds) Proceedings of the 18th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors. EDM 2017. The Minerals, Metals & Materials Series. Springer, Cham Link
- Influence of surface relaxation on APT analysis of solute atoms, C Oberdorfer, T Withrow, L-J Yu, K Fisher, EA Marquis, and W Windl, Materials Characterization (2017) Invited article for special issue on Atom Probe Tomography – link
- Did you say clusters? Workshop on atomic scale structures of interfaces, Bernkastel, Germany, May 2016
(Invited) From atom probe tomography imaging to microstructural quantification I – Modeling field evaporation. International Congress on 3D Materials Science, June 2016
(Invited) Characterization of Nuclear Materials Using Atom Probe Tomography, M&M, St Louis, August 2017
(Invited)Progress in modeling atom probe tomography, Frontiers of Electron Tomography, NCEM, October 2017
(Invited) Atom Probe Tomography Round Robin on the analysis of solute clustering in Irradiated Stainless Steels, APT&M, Gaithersburg, MD, June 2018