A Crystallography-Mediated Reconstruction (CMR) Approach for Atom Probe Tomography: Solution for a Singleton Pole

Abstract

Spatially accurate atom probe tomography reconstructions are vitally important when quantitative spatial measurements such as distances, volumes and morphologies etc. of nanostructural features are required information for the researcher. It is well known that the crystallographic information contained within the atom probe data of crystalline materials can be used to calibrate the tomographic reconstruction. Specifically, the crystallographic information projected into the field evaporation images is used. This offers a powerful and accurate enhancement of the atom probe technique. However, this is often difficult to do in practice. In previously reported approaches, it was necessary to index at least two poles to compute the image compression factor ‘ξ’ and observe crystallographic planes in at least one of the pole regions to obtain a measure of the field factor ‘kf’ while also manually accounting for a change in reconstruction parameters throughout the dataset. Not only is this error-prone and time consuming, it does not work for materials that exhibit limited crystallographic information in their field evaporation image. Here, we extend the applicability of the crystallographic calibration of atom probe data by proposing a reconstruction methodology where only one pole with observable lattice planes is required in the projected detector image. Our proposal also accounts for dynamic variations in the reconstruction parameters throughout the 3D dataset. The method is simpler and significantly faster to implement and is applicable to more atom probe situations than previously approaches. Our single-pole crystallography mediated reconstruction (SP-CMR) utilizes the Hawkes-Kasper projection model (equivalent to the equidistant-azimuthal projection model) and the direct fourier (DF) fit algorithm to determine the precise reconstruction parameters required to produce flat atomic planes. It is applied to experimental Al and highly Sb-doped Si data. The discrepancies between the spatial dimensions of the SP-CMR reconstructions compared to uncalibrated reconstructions are visually apparent. Consistent plane spacings and angles between crystallographic directions matching the theoretically known values for each crystal structure are demonstrated.