Analysis Techniques for Atom Probe Tomography

Abstract

One of the uniquely powerful aspects of atom probe tomography is the capacity for combining striking three-dimensional (3D) visualisations of atomic-scale microstructure with rigorous quantitative data analysis. Data-mining algorithms can be applied to identify regions of interest, ranging from isolating a single atom to highlighting discrete phases. In modern instruments, tens of millions and often hundreds of millions of ions are routinely acquired, elementally identified and assigned 3D spatial coordinates. Hence, an APT dataset represents a large quantity of atomic-scale structural and chemical information. There are obvious limitations to how much can be learned from mere visual inspection of the tomograms. Therefore, research into the development and application of rigorous data-mining approaches has grown in recent years and is an important frontier for the technique. A challenge is for the analysis techniques to keep pace with the instrumental developments that are generating vastly larger datasets. These larger datasets themselves open up new computational approaches for data-mining.