Toward an automated tool for dislocation density characterization in a scanning electron microscope

• Published in: Materials characterization Vol. 158; p. 109954
• Main Authors: Cazottes, S., Bechis, A., Lafond, C., L’Hôte, G., Roth, C., Dreyfus, T., Steyer, P., Douillard, T., Langlois, C.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.12.2019; Elsevier
• Subjects: Data clustering; Dislocations; Electron channeling contrast imaging; Engineering Sciences; Scanning electron microscopy; Scanning electron microscopy; Electron channeling contrast imaging; Data clustering; Dislocations
• ISSN: 1044-5803; 1873-4189; 1873-4189
• DOI: 10.1016/j.matchar.2019.109954

We propose a methodology for quantitative dislocation characterization of a bulk sample in a scanning electron microscope without requiring pre-orientation of the sample before analysis. In this method, a series of backscattered electron images are acquired while rotating the sample, and an intensity profile as a function of the rotation angle is obtained for each pixel of the observed area. These intensity profiles are used to determine the orientation condition of the analyzed grain. The nature of the pixel is defined as what dominates the pixel intensity (matrix, defect or noise). As the intensity profiles are also characteristic of the pixel nature, a data clustering algorithm is applied to the intensity profiles to classify the pixel nature. As a result, the defect density, such as the dislocation density, can be automatically measured. The proposed method is fast and efficient compared with transmission electron microscopy analysis and could enable the future characterization of multiple grains in a deformed sample within a reasonable amount of time. [Display omitted] •Dislocation were observed by means of Electron Channeling Contrast Imaging without any prior orientation of the crystal.•The nature (dislocation, matrix or noise) of each pixel of the ROI was determined automatically using the Intensity Profiles.•The dislocation density was directly determined and is similar to the one determined from single image analysis.