Quantitative Phase Analysis for Carbide Characterization in Steel Using Automated Electron Diffraction

Abstract

Carbides are critical to many different properties in a steel. In this paper, the quantification of a carbide powder extracted from a stainless steel analyzed with profile refinement of X‐ray powder diffraction patterns, is compared to a novel technique for automated electron diffraction pattern collection with a transmission electron microscope. The automated analysis has the advantage of collecting patterns from individual particles, hence avoiding issues such as overlapping reflections in powder diffractograms. Electron diffraction patterns with satisfactory quality are selected through a deep convoluted neural network and matched to a library of pre‐calculated diffraction patterns, corresponding to a set of selected carbide crystal structures. The quantitative results from the automated particle analysis are in good agreement with the phases identified by powder profile refinement.

Automated serial rotation electron diffraction combined with cluster analysis: an efficient multi-crystal workflow for structure determination

“…The SerialED method combines computer-controlled stage translation with beam shift to automatically collect diffraction data on a large number of crystals. We previously explored the use of SerialED data in the context of structure determination (Smeets et al, 2018b) and quantitative phase analysis (Smeets et al, 2019). However, indexing a randomly oriented diffraction pattern in SerialED data has been challenging, because it is difficult to accurately determine the crystal orientation.…”

Exploring Polycrystalline Materials: High-throughput Phase Elucidation Using Serial Rotation Electron Diffraction

“…The metrics differ in that SPXRD profile analysis determines weight percent and cluster analysis counts the number of crystals. Although our sampling size is still relatively small, this indicates that SerialRED is capable of quantitative phase analysis for crystalline materials as suggested previously 33,52 , as long as the number of crystals collected is large enough.…”

High-throughput continuous rotation electron diffraction data acquisitionviasoftware automation

“…After initial calibrations, the method is fully automated and can survey an area of approximately 400 Â 400 mm in an hour. We have shown previously that it is possible to use the electron diffraction data from a large number of crystals for phase analysis (Smeets, Å ngströ m & Olsson, 2018) and that the merged data are suitable for structure determination . In this section, we discuss the application of the SerialED method as a way of screening for good crystals for cRED data collection.…”

“…When a diffraction pattern is passed through the CNN, a prediction score between 0.0 and 1.0 is returned, where any value greater than 0.5 corresponds to a good quality diffraction pattern [see also Smeets, Å ngströ m & Olsson (2018)].…”