Christina Choi | Searcharxiv

arXiv2021

Machine Learning Pipeline for Segmentation and Defect Identification from High Resolution Transmission Electron Microscopy Data

In the field of transmission electron microscopy, data interpretation often lags behind acquisition methods, as image processing methods often have to be manually tailored to individual datasets. Machine learning offers a promising approach for fast, accurate analysis of electron microscopy data. Here, we demonstrate a flexible two step pipeline for analysis of high resolution transmission electron microscopy data, which uses a U-Net for segmentation followed by a random forest for detection of stacking faults. Our trained U-Net is able to segment nanoparticle regions from amorphous background with a Dice coefficient of 0.8 and significantly outperforms traditional image segmentation methods. Using these segmented regions, we are then able to classify whether nanoparticles contain a visible stacking fault with 86% accuracy. We provide this adaptable pipeline as an open source tool for the community. The combined output of the segmentation network and classifier offer a way to determine statistical distributions of features of interest, such as size, shape and defect presence, enabling detection of correlations between these features.

eess.IV↗

arXiv2021

Comparative analysis for meaningful interpretation of rare-earth oxide M$_{4,5}$ energy loss edges

The magnetic, electronic, and optical properties of rare earth-oxides are directly influenced by the valency of the metallic cation. With the development of next generation electron energy-loss spectrometers, high-energy lanthanide fine structure can be studied with improved signal-to-noise for quantitative analysis. Unfortunately, the behavior of rare-earth $4f$ orbital electrons is not well understood. To establish best practices for analysis of energy-loss spectra from lanthanide oxides, we have performed a comparative study of the four traditional white line analysis methods extended to lanthanide $M_{4,5}$ edges resulting from $3d \rightarrow 4f$ orbital transitions using data from Gatan's EELS Atlas. The ${M_4}/{M_5}$ spectral feature ratios were examined as a function of $4f$ occupancy. The ${M_4}/{M_5}$ spectral feature ratio decreases exponentially as $4f$ occupancy increases, except for a plateau between S$\text{m}^{3+}$ and D$\text{y}^{3+}$. The full-width at half the maximum intensity of the $M_4$ edges shows increased broadening for S$\text{m}^{3+}$ through D$\text{y}^{3+}$. We suggest that the plateau results from $4f$ orbital half-filling and is explained through the relationship between electron transition probability and transition lifetime as expressed through Fermi's Golden Rule. Of the four spectral analysis methods described, only the integrated area method can be ascribed a quantitative physical interpretation.

cond-mat.mtrl-sci↗