Influence of Analysis Parameters on the Microstructural Characterization of Nanoscale Precipitates
- PDF / 519,257 Bytes
- 7 Pages / 612 x 792 pts (letter) Page_size
- 26 Downloads / 197 Views
1231-NN07-03
Influence of Analysis Parameters on the Microstructural Characterization of Nanoscale Precipitates Ai Serizawa1 and M. K. Miller1 1 Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6136, U.S.A.
ABSTRACT A series of simulated microstructures containing nanometer-scale precipitates was created with an atom probe data simulator. These data were then analyzed with the proximity histogram by creating isoconcentration surfaces to determine the influence of the analysis method. For simulated 2-nm-radius spherical precipitates, the optimized voxel size and delocalization were found to be 0.5-0.6 nm and 1.0-1.5 nm, respectively. Under optimum analysis parameters, the voxelization/delocalization process only slightly degrades the interface width determined from the proximity histogram to ~0.15±0.04 nm. INTRODUCTION Distributions of ultra-fine precipitates or clusters are important for the properties of many materials. The spatial resolution of atom probe tomography enables the finest scale precipitates to be detected and their parameters quantified. However, for accurate estimation of their sizes, compositions, and number densities, it is important to be able to reliably distinguish them from the random solute atoms in the matrix. In order to establish reliable and consistent estimates of these microstructural parameters, a series of simulated microstructures containing nanometerscale precipitates was created. The generated microstructure will therefore contain a known number and position of precipitates, as well as their size and optimal composition. In order to determine the influence of the analysis parameters These data were then analyzed with Imago Scientific Instruments IVAS implementation of the proximity histogram [1] by creating the isoconcentration surfaces. The definition of the position of the interface between a precipitate and the matrix is important not only for investigating the size and morphology of precipitates, but also for estimating the solute concentrations and the extent and amplitude of solute gradients on both sides of the interface. The accuracy of the position of the interface is particularly important for nanoscale (i.e., < 5 nm diameter) precipitates, as a large proportion of the solute atoms reside in the surface layer. EXPERIMENTAL An atom probe data simulator [2] was used to construct 40 nm × 40 nm × 40 nm cubic datasets containing randomly distributed, non-impinging, nominally spherical (i.e., atomically faceted) 2-nm-radius precipitates with a number density of 6×1023 m-3. The solute concentrations of the precipitates and the matrix, and the detection efficiency of the single atom detector were systematically varied. This volume contained 64 precipitates with some intersecting the edge of the volume and 2 in close proximity.
To create the isoconcentration surface, a 3D grid of bins was constructed with a selected voxel size, V. Then for each atom, a 3D 3 Gaussian centered at the atom position with the width of the selected deloc
Data Loading...
