| About this Abstract |
| Meeting |
2020 TMS Annual Meeting & Exhibition
|
| Symposium
|
Atom Probe Tomography for Advanced Characterization of Metals, Minerals and Materials III
|
| Presentation Title |
Quantifying Compositional Uncertainty Arising from Peak Overlaps |
| Author(s) |
Andrew James London |
| On-Site Speaker (Planned) |
Andrew James London |
| Abstract Scope |
There are many sources of error atom probe composition quantification, including when mass peaks overlap. A degree of difficulty and hence error in solving overlaps stems from their relative intensities. We show how uncertainty can be minimized using a maximum likelihood methodology, which furthermore yields a quantitative description of "overlap difficulty". Misidentification of peaks, often yields large systematic errors, which can be additionally complicated by overlapping peaks. However, an analyst must select only the set of ions that are reasonable to consider before solving the overlap. By using an objective criterion from information theory, it is shown how such systematic bias can be avoided when choosing which ions to consider when analyzing the mass spectrum. We show how combining these techniques improves compositional certainty and how this approach can improve experimental planning. |
| Proceedings Inclusion? |
Planned: Supplemental Proceedings volume |