The rise of computational techniques in atom probe microscopy

Much effort has been devoted to the development of computational techniques in atom probe microscopy over the past decade. There have been several drivers for this effort. Firstly, there has been effort devoted to addressing the challenges of discerning information from the increasingly large size of the data, and capturing the opportunities that this large data presents. Secondly, there has been significant new effort devoted to the simulation of atom probe data so that pristine datasets that contain microstructural features of increasing complexity can be generated in-silico, and subjected to complex data-mining algorithms. This has enabled the benchmarking of various algorithms, guided the setting of parameters for particular analyses, and exposed the effects of instrumentation parameters such as detector efficiency and aberrations in ionic flight path. The authors are especially interested in the prospects of converging atomic-scale microscopy with atomic-scale materials modelling via first principles approaches. This involves excising parts of the APM data and using these as super-cell inputs to calculations of materials properties via density functional theory. It is our opinion that this represents a major advance for materials science because it enables microscopy to advance microstructure-property relationships to the direct mapping of such relationships based on many-body interactions. As such, this approach has great potential for materials design and development. The final part of this paper focuses on how cloud-based computing represents an exciting frontier of the computational aspects of atom probe microscopy. We discuss the opportunities and the barriers for conducting new materials science through the analysis and visualisation of atom probe data via new generation tools that are cloud-based, and which are managed, curated and governed with significant user-community input and integrated with contemporary electronic laboratory notebook technology. © 2013 Published by Elsevier Ltd. All rights reserved