Multivariate statistical analysis of electron energy-loss spectroscopy in anisotropic materials

Comparative of machine learning classification strategies for electron energy loss spectroscopy: Support vector machines and artificial neural networks

Machine Learning (ML) strategies applied to Scanning and conventional Transmission Electron Microscopy have become a valuable tool for analyzing the large volumes of data generated by various S/TEM techniques. In this work, we focus on Electron Energy Loss Spectroscopy (EELS) and study two ML techniques for classifying spectra in detail: Support Vector Machines (SVM) and Artificial Neural Networks (ANN). Firstly, we systematically analyze the optimal configurations and architectures for ANN classifiers using random search and the tree-structured Parzen estimator methods. Secondly, a new kernel strategy is introduced for the soft-margin SVMs, the cosine kernel, which offers a significant advantage over the previously studied kernels and other ML classification strategies. This kernel allows us to bypass the normalization of EEL spectra, achieving accurate classification. This result is highly relevant for the EELS community since we also assess the impact of common normalization techniques on our spectra using Uniform Manifold Approximation and Projection (UMAP), revealing a strong bias introduced in the spectra once normalized. In order to evaluate and study both classification strategies, we focus on determining the oxidation state of transition metals through their EEL spectra, examining which feature is more suitable for oxidation state classification: the oxygen K peak or the transition metal white lines. Subsequently, we compare the resistance to energy loss shifts for both classifiers and present a strategy to improve their resistance. The results of this study suggest the use of soft-margin SVMs for simpler EELS classification tasks with a limited number of spectra, as they provide performance comparable to ANNs while requiring lower computational resources and reduced training times. Conversely, ANNs are better suited for handling complex classification problems with extensive training data.

Dimensionality reduction and unsupervised clustering for EELS-SI

A novel combination of machine learning algorithms is proposed for the differentiation of distinct spectra in a large electron energy loss spectroscopy spectrum image (EELS-SI) dataset. For clustering of the EEL spectra including similar fine structures in an efficient space, linear and nonlinear dimensionality reduction methods are used to project the EEL spectra onto a low-dimensional space. Then, a density-based clustering algorithm is applied to distinguish the meaningful data clusters. By applying this strategy to various experimental EELS-SI datasets, differentiation of several groups of EEL spectra representing specific fine structures was achieved. It is possible to investigate particular fine structures by averaging all of the spectra in each cluster. Also, the spatial distributions of each cluster in the scanning regions can be observed, which enables investigation of the locations of different fine structures in materials. This method does not require any prior knowledge, i.e., it is a data-driven analysis; therefore, it can be applied to any hyperspectral image.

Evaluation of EELS spectrum imaging data by spectral components and factors from multivariate analysis

Multivariate analysis is a powerful tool to process spectrum imaging datasets of electron energy loss spectroscopy. Most spatial variance of the datasets can be explained by a limited numbers of components. We explore such dimension reduction to facilitate quantitative analyses of spectrum imaging data, supervising the spectral components instead of spectra at individual pixels. In this study, we use non-negative matrix factorization to decompose datasets from Fe2O3 thin films with different Sn doping profiles on SnO2 and Si substrates. Case studies are presented to analyse spectral features including background models, signal integrals, peak positions and widths. Matlab codes are written to guide microscopists to perform these data analyses.

Theoretical study of core-loss electron energy-loss spectroscopy at graphene nanoribbon edges

A systematic study of simulated atomic-resolution electronic energy-loss spectroscopy (EELS) for different graphene nanoribbons (GNRs) is presented. The results of ab initio studies of carbon [Formula: see text] core-loss EELS on GNRs with different ribbon edge structures and different hydrogen terminations show that theoretical core-loss EELS can distinguish key structural features at the atomic scale. In addition, the combination of polarized core-loss EELS with symmetry resolved electronic partial density of states calculations can be used to identify the origins of all the primary features in the spectra. For example, the nature of the GNR edge structure (armchair, zigzag, etc) can be identified, along with the degree of hydrogenation. Hence it is possible to use the combination of ab initio calculations with high resolution, high energy transmission core-loss EELS experiments to determine the local atomic arrangement and chemical bonding states (i.e. a structural fingerprint) in GNRs, which is essential for future practical applications of graphene.