Phase Mapping in EBSD Using Convolutional Neural Networks

Abstract

The emergence of commercial electron backscatter diffraction (EBSD) equipment ushered in an era of information rich maps produced by determining the orientation of user-selected crystal structures. Since then, a technological revolution has occurred in the quality, rate detection, and analysis of these diffractions patterns. The next revolution in EBSD is the ability to directly utilize the information rich diffraction patterns in a high-throughput manner. Aided by machine learning techniques, this new methodology is, as demonstrated herein, capable of accurately separating phases in a material by crystal symmetry, chemistry, and even lattice parameters with fewer human decisions. This work is the first demonstration of such capabilities and addresses many of the major challenges faced in modern EBSD. Diffraction patterns are collected from a variety of samples, and a convolutional neural network, a type of machine learning algorithm, is trained to autonomously recognize the subtle differences in the diffraction patterns and output phase maps of the material. This study offers a path to machine learning coupled phase mapping as databases of EBSD patterns encompass an increasing number of the possible space groups, chemistry changes, and lattice parameter variations.

An Acquisition Parameter Study for Machine-Learning-Enabled Electron Backscatter Diffraction

“…The Xception CNN architecture (Chollet, 2017) was selected for fitting the model. The selection of this network was based on Xception or derivatives of Xception being used previously in the EBSD community (Ding et al, 2020;Kaufmann et al, 2020aKaufmann et al, , 2020bKaufmann et al, , 2020c. Refer to Figure 5 in Xception: Deep Learning with Depthwise Separable Convolutions (Chollet, 2017) for a complete description of the Xception architecture.…”

“…The application of these tools to image-based tasks in materials science has proved to be useful for classification (Modarres et al, 2017;Ziletti et al, 2018;Foden et al, 2019a;Kaufmann et al, 2020a), segmentation (DeCost et al, 2019;Stan et al, 2020), and other objectives (Xie & Grossman, 2018;de Haan et al, 2019). Examples of techniques where interest in developing artificial intelligence agents for image-based tasks include optical microscopy (DeCost & Holm, 2015;DeCost et al, 2019), scanning transmission electron microscopy (STEM) (Laanait et al, 2019;Roberts et al, 2019), transmission electron microscopy (TEM) (Spurgeon et al, 2020), and electron backscatter diffraction (EBSD) (Shen et al, 2019;Ding et al, 2020;Kaufmann et al, 2020aKaufmann et al, , 2020bKaufmann et al, , 2020c. These efforts are motivated by accelerating data generation rates and the traditional need for tedious or arduous analysis of the data by well-trained individuals with sufficient knowledge of the material domain.…”

“…Improvements to phase differentiation have been proposed and developed including dictionary indexing (Chen et al, 2015;Ram et al, 2017;Ram & De Graef, 2018;Singh et al, 2018) and spherical indexing (Day, 2008;Lenthe et al, 2019;Zhu et al, 2019), although each still requires a user to pre-select phases and further requires simulating the Kikuchi sphere for each selected phase. Recently, the EBSD community has begun to investigate the use of CNNs for indexing, phase differentiation, and determining components of the crystal structure (Foden et al, 2019a;Ding et al, 2020;Kaufmann et al, 2020aKaufmann et al, , 2020bKaufmann et al, , 2020c. It is a goal of several of these efforts that the onus of phase selection and/or structure determination can be at least partially lessened on the user (Kaufmann et al, 2020a(Kaufmann et al, , 2020b.…”

Autonomous EBSD Pattern Classification Performance with Changing Acquisition Parameters

“…In material analysis, these tools have largely been applied to techniques requiring analysis of data collected in the form of images [1,2]. Electron backscatter diffraction (EBSD) is one such technique benefitting from these recent efforts to improve material analysis by leveraging deep neural networks [3][4][5][6][7]. EBSD is an SEM-based technique involving the capture of 2D diffraction patterns from the surface of a well-polished sample [8].…”

“…chemistry or simulated diffraction patterns) to be available. In contrast, deep neural network-based methods have demonstrated effective phase differentiation [5] and identification of phases to the space group level [4] without the need for further information. The deep learning approach to EBSD diffraction pattern analysis is capable of these more advanced analyses because it uses all information in the image when assessing a diffraction pattern, whereas traditional Hough-based EBSD pattern analysis discards a significant amount of information.…”