Nanoscale light element identification using machine learning aided STEM-EDS

Abstract

Light element identification is necessary in materials research to obtain detailed insight into various material properties. However, reported techniques, such as scanning transmission electron microscopy (STEM)-energy dispersive X-ray spectroscopy (EDS) have inadequate detection limits, which impairs identification. In this study, we achieved light element identification with nanoscale spatial resolution in a multi-component metal alloy through unsupervised machine learning algorithms of singular value decomposition (SVD) and independent component analysis (ICA). Improvement of the signal-to-noise ratio (SNR) in the STEM-EDS spectrum images was achieved by combining SVD and ICA, leading to the identification of a nanoscale N-depleted region that was not observed in as-measured STEM-EDS. Additionally, the formation of the nanoscale N-depleted region was validated using STEM–electron energy loss spectroscopy and multicomponent diffusional transformation simulation. The enhancement of SNR in STEM-EDS spectrum images by machine learning algorithms can provide an efficient, economical chemical analysis method to identify light elements at the nanoscale.

Mapping the nanoscale effects of charge traps on electrical transport in grain structures of indium tin oxide thin films

“…Previous reports say that it has been very challenging to perform defect analyses through elemental mapping or electron microscopies at a nanoscale resolution. 23 Recently, we reported a scanning noise microscopy method that allows one to map the activities of electrical noise sources in lateral and vertical electrical channels at a nanoscale resolution. 20,24,25 In the conducting film, lateral transport is the dominant conduction mechanism between the electrodes.…”

STEM Tools for Semiconductor Characterization: Beyond High-Resolution Imaging

“…Machine learning algorithms also find applications facing spectral data, which is useful, for example, to distinguish the signal sub-space from the noise by dimensionality reduction of EDX data, which in turn improves the identification of light elements (i.e., N) [ 95 ], thus offering a promising alernative for the study of semiconductor materials at the nanoscale, such as dilute nitrides.…”

Exploring the Surface Segregation of Rh Dopants in PtNi Nanoparticles through Atom Probe Tomography Analysis

“…24 APT also exhibits an equal sensitivity to both light and heavy elements, whereas STEM-EDS is relatively less sensitive to light elements. 25 Here, we analyzed Rh-doped PtNi nanoparticles that have been reported to exhibit excellent oxygen reduction activity and superior stability. impressive long-term ORR durability, experiencing only a 9.2% reduction in mass activity after 10,000 cycles.…”

“…For characterization of multicomponent nanocatalysts, scanning transmission electron microscopy-energy-dispersive X-ray spectroscopy (STEM-EDS) is commonly used for elemental mapping of nanocatalysts and is increasingly complemented by atom probe tomography (APT). − APT has the ability to measure elemental distribution in three dimensions with high chemical sensitivity and subnanometer spatial resolution and as such can address issues associated with the two-dimensional projected image obtained by STEM-EDS that makes analysis of in-depth compositional distribution in three dimension (3D) challenging . APT also exhibits an equal sensitivity to both light and heavy elements, whereas STEM-EDS is relatively less sensitive to light elements …”