Enhancement of noisy EDX HRSTEM spectrum-images by combination of filtering and PCA

Leveraging Machine Learning for Advanced Nanoscale X-ray Analysis: Unmixing Multicomponent Signals and Enhancing Chemical Quantification

Energy dispersive X-ray (EDX) spectroscopy in the transmission electron microscope is a key tool for nanomaterials analysis, providing a direct link between spatial and chemical information. However, using it for precisely determining chemical compositions presents challenges of noisy data from low X-ray yields and mixed signals from phases that overlap along the electron beam trajectory. Here, we introduce a novel method, non-negative matrix factorization based pan-sharpening (PSNMF), to address these limitations. Leveraging the Poisson nature of EDX spectral noise and binning operations, PSNMF retrieves high-quality phase spectral and spatial signatures via consecutive factorizations. After validating PSNMF with synthetic data sets of different noise levels, we illustrate its effectiveness on two distinct experimental cases: a nanomineralogical lamella, and supported catalytic nanoparticles. Not only does PSNMF obtain accurate phase signatures, but data sets reconstructed from the outputs have demonstrably lower noise and better fidelity than from the benchmark denoising method of principle component analysis.

Enhancing spatial resolution in Fourier transform infrared spectral image via machine learning algorithms

Owing to the intrinsic signal noise in the characterization of chemical structures through Fourier transform infrared (FT-IR) spectroscopy, the determination of the signal-to-noise ratio (SNR) depends on the level of the concentration of the chemical structures. In situations characterized by limited concentrations of chemical structures, the traditional approach involves mitigating the resulting low SNR by superimposing repetitive measurements. In this study, we achieved comparable high-quality results to data scanned 64 times and superimposed by employing machine learning algorithms such as the principal component analysis and non-negative matrix factorization, which perform the dimensionality reduction, on FT-IR spectral image data that was only scanned once. Furthermore, the spatial resolution of the mapping images correlated to each chemical structure was enhanced by applying both the machine learning algorithms and the Gaussian fitting simultaneously. Significantly, our investigation demonstrated that the spatial resolution of the mapping images acquired through relative intensity is further improved by employing dimensionality reduction techniques. Collectively, our findings imply that by optimizing research data through noise reduction enhancing spatial resolution using the machine learning algorithms, research processes can be more efficient, for instance by reducing redundant physical measurements.

Room Temperature Spin-to-Charge Conversion in Amorphous Topological Insulating Gd-Alloyed BixSe1-x/CoFeB Bilayers

Disordered topological insulator (TI) films have gained intense interest by benefiting from both the TI's exotic transport properties and the advantage of mass production by sputtering. Here, we report on the clear evidence of spin-charge conversion (SCC) in amorphous Gd-alloyed BixSe1-x (BSG)/CoFeB bilayers fabricated by sputtering, which could be related to the amorphous TI surface states. Two methods have been employed to study SCC in BSG (tBSG = 6-16 nm)/CoFeB(5 nm) bilayers with different BSG thicknesses. First, spin pumping is used to generate a spin current in CoFeB and detect SCC by the inverse Edelstein effect (IEE). The maximum SCC efficiency (SCE) is measured to be as large as 0.035 nm (IEE length λIEE) in a 6 nm thick BSG sample, which shows a strong decay when tBSG increases due to the increase of BSG surface roughness. The second method is THz time-domain spectroscopy, which reveals a small tBSG dependence of SCE, validating the occurrence of a pure interface state-related SCC. Furthermore, our angle-resolved photoemission spectroscopy data show dispersive two-dimensional surface states that cross the bulk gap until the Fermi level, strengthening the possibility of SCC due to the amorphous TI states. Our studies provide a new experimental direction toward the search for topological systems in amorphous solids.

High Power Factor Nb-Doped TiO2 Thermoelectric Thick Films: Toward Atomic Scale Defect Engineering of Crystallographic Shear Structures

Donor-doped TiO₂-based materials are promising thermoelectrics (TEs) due to their low cost and high stability at elevated temperatures. Herein, high-performance Nb-doped TiO₂ thick films are fabricated by facile and scalable screen-printing techniques. Enhanced TE performance has been achieved by forming high-density crystallographic shear (CS) structures. All films exhibit the same matrix rutile structure but contain different nano-sized defect structures. Typically, in films with low Nb content, high concentrations of oxygen-deficient {121} CS planes are formed, while in films with high Nb content, a high density of twin boundaries are found. Through the use of strongly reducing atmospheres, a novel Al-segregated {210} CS structure is formed in films with higher Nb content. By advanced aberration-corrected scanning transmission electron microscopy techniques, we reveal the nature of the {210} CS structure at the nano-scale. These CS structures contain abundant oxygen vacancies and are believed to enable energy-filtering effects, leading to simultaneous enhancement of both the electrical conductivity and Seebeck coefficients. The optimized films exhibit a maximum power factor of 4.3 × 10^-4 W m^-1 K^-2 at 673 K, the highest value for TiO₂-based TE films at elevated temperatures. Our modulation strategy based on microstructure modification provides a novel route for atomic-level defect engineering which should guide the development of other TE materials.

Mo3Ni2N Nanoparticle Generation by Spark Discharge

Spark ablation is an advantageous method for the generation of metallic nanoparticles with defined particle sizes and compositions. The reaction of the metal particles with the carrier gas during the synthesis and, therefore, the incorporation of those light elements into structural voids or even compound formation was confirmed for hydrides and oxides but has only been suspected to occur for nitrides. In this study, dispersed nanoparticles of Mo₃Ni₂N and Mo with Janus morphology, and defined particle sizes were obtained by spark discharge generation as a result of carrier gas ionization and characterized using transmission electron microscopy and powder X-ray diffraction. Metal nitrides possess beneficial catalytic and thermoelectric properties, as well as high hardness and wear resistance. Therefore, this method offers the possibility of controlled synthesis of materials which are interesting for numerous applications.

Machine learning in electron microscopy for advanced nanocharacterization: current developments, available tools and future outlook

In the last few years, electron microscopy has experienced a new methodological paradigm aimed to fix the bottlenecks and overcome the challenges of its analytical workflow. Machine learning and artificial intelligence are answering this call providing powerful resources towards automation, exploration, and development. In this review, we evaluate the state-of-the-art of machine learning applied to electron microscopy (and obliquely, to materials and nano-sciences). We start from the traditional imaging techniques to reach the newest higher-dimensionality ones, also covering the recent advances in spectroscopy and tomography. Additionally, the present review provides a practical guide for microscopists, and in general for material scientists, but not necessarily advanced machine learning practitioners, to straightforwardly apply the offered set of tools to their own research. To conclude, we explore the state-of-the-art of other disciplines with a broader experience in applying artificial intelligence methods to their research (e.g., high-energy physics, astronomy, Earth sciences, and even robotics, videogames, or marketing and finances), in order to narrow down the incoming future of electron microscopy, its challenges and outlook.

Mapping short-range order at the nanoscale in metal-organic framework and inorganic glass composites

Characterization of nanoscale changes in the atomic structure of amorphous materials is a profound challenge. Established X-ray and neutron total scattering methods typically provide sufficient signal quality only over macroscopic volumes. Pair distribution function analysis using electron scattering (ePDF) in the scanning transmission electron microscope (STEM) has emerged as a method of probing nanovolumes of these materials, but inorganic glasses as well as metal-organic frameworks (MOFs) and many other materials containing organic components are characteristically prone to irreversible changes after limited electron beam exposures. This beam sensitivity requires 'low-dose' data acquisition to probe inorganic glasses, amorphous and glassy MOFs, and MOF composites. Here, we use STEM-ePDF applied at low electron fluences (10 e^- Å^-2) combined with unsupervised machine learning methods to map changes in the short-range order with ca. 5 nm spatial resolution in a composite material consisting of a zeolitic imidazolate framework glass a_gZIF-62 and a 0.67([Na₂O]_0.9[P₂O₅])-0.33([AlO_3/2][AlF₃]_1.5) inorganic glass. STEM-ePDF enables separation of MOF and inorganic glass domains from atomic structure differences alone, showing abrupt changes in atomic structure at interfaces with interatomic correlation distances seen in X-ray PDF preserved at the nanoscale. These findings underline that the average bulk amorphous structure is retained at the nanoscale in the growing family of MOF glasses and composites, a previously untested assumption in PDF analyses crucial for future non-crystalline nanostructure engineering.

Towards Automatic Detection of Precipitates in Inconel 625 Superalloy Additively Manufactured by the L-PBF Method

In our study, the comparison of the automatically detected precipitates in L-PBF Inconel 625, with experimentally detected phases and with the results of the thermodynamic modeling was used to test their compliance. The combination of the complementary electron microscopy techniques with the microanalysis of chemical composition allowed us to examine the structure and chemical composition of related features. The possibility of automatic detection and identification of precipitated phases based on the STEM-EDS data was presented and discussed. The automatic segmentation of images and identifying of distinguishing regions are based on the processing of STEM-EDS data as multispectral images. Image processing methods and statistical tools are applied to maximize an information gain from data with low signal-to-noise ratio, keeping human interactions on a minimal level. The proposed algorithm allowed for automatic detection of precipitates and identification of interesting regions in the Inconel 625, while significantly reducing the processing time with acceptable quality of results.

Deep learning-based denoising for improved dose efficiency in EDX tomography of nanoparticles

The combination of energy-dispersive X-ray spectroscopy (EDX) and electron tomography is a powerful approach to retrieve the 3D elemental distribution in nanomaterials, providing an unprecedented level of information for complex, multi-component systems, such as semiconductor devices, as well as catalytic and plasmonic nanoparticles. Unfortunately, the applicability of EDX tomography is severely limited because of extremely long acquisition times and high electron irradiation doses required to obtain 3D EDX reconstructions with an adequate signal-to-noise ratio. One possibility to address this limitation is intelligent denoising of experimental data using prior expectations about the objects of interest. Herein, this approach is followed using the deep learning methodology, which currently demonstrates state-of-the-art performance for an increasing number of data processing problems. Design choices for the denoising approach and training data are discussed with a focus on nanoparticle-like objects and extremely noisy signals typical for EDX experiments. Quantitative analysis of the proposed method demonstrates its significantly enhanced performance in comparison to classical denoising approaches. This allows for improving the tradeoff between the reconstruction quality, acquisition time and radiation dose for EDX tomography. The proposed method is therefore especially beneficial for the 3D EDX investigation of electron beam-sensitive materials and studies of nanoparticle transformations.

Anisotropic Angstrom-Wide Conductive Channels in Black Phosphorus by Top-down Cu Intercalation

Intercalation in black phosphorus (BP) can induce and modulate a variety of the properties including superconductivity like other two-dimensional (2D) materials. In this perspective, spatially controlled intercalation has the possibility to incorporate different properties into a single crystal of BP. We demonstrate anisotropic angstrom-wide (∼4.3 Å) Cu intercalation in BP, where Cu atoms are intercalated along a zigzag direction of BP because of its inherent anisotropy. With atomic structure, its microstructural effects, arising from the angstrom-wide Cu intercalation, were investigated and extended to relation with macrostructure. As the intercalation mechanism, it was revealed by in situ transmission electron microscopy and theoretical calculation that Cu atoms are intercalated through top-down direction of BP. The Cu intercalation anisotropically induces transition of angstrom-wide electronic channels from semiconductor to semimetal in BP. Our findings throw light on the fundamental relationship between microstructure changes and properties in intercalated BP, and tailoring anisotropic 2D materials at angstrom scale.

Dissecting complex nanoparticle heterostructures via multimodal data fusion with aberration-corrected STEM spectroscopy

With nanostructured materials such as catalytic heterostructures projected to play a critical role in applications ranging from water splitting to energy harvesting, tailoring their properties to specific tasks requires an increasingly comprehensive characterization of their local chemical and electronic landscape. Although aberration-corrected electron spectroscopy currently provides sufficient spatial resolution to study this space, an approach to concurrently dissect both the electronic structure and full composition of buried metal/oxide interfaces remains a considerable challenge. In this manuscript, we outline a statistical methodology to jointly analyze simultaneously-acquired STEM EELS and EDX datasets by fusing them along their shared spatial factors. We show how this procedure can be used to derive a rich descriptive model for estimating both transition metal valency and full chemical composition from encapsulated morphologies such as core-shell nanoparticles. We demonstrate this on a heterogeneous Co-P thin film catalyst, concluding that this system is best described as a multi-shell phosphide structure with a P-doped metallic Co core.

Evidence of a strong perpendicular magnetic anisotropy in Au/Co/MgO/GaN heterostructures

We report a strong perpendicular magnetic anisotropy (PMA) in Au/Co/MgO/GaN heterostructures from both experiments and first-principles calculations. The Au/Co/MgO heterostructures have been grown by molecular beam epitaxy (MBE) on GaN/sapphire substrates. By carefully optimizing the growth conditions, we obtained a fully epitaxial structure with a crystalline orientation relationship Au(111)[1̄10]//Co(0001)[112̄0]//MgO(111)[101̄]//GaN(0002)[112̄0]. More interestingly, we demonstrate that a 4.6 nm thick Co film grown on MgO/GaN still exhibits a large perpendicular magnetic anisotropy. First-principles calculations performed on the Co (4ML)/MgO(111) structure showed that the MgO(111) surface can strongly enhance the magnetic anisotropy energy by 40% compared to a reference 4ML thick Co hcp film. Our layer-resolved and orbital-hybridization resolved anisotropy analyses helped to clarify that the origin of the PMA enhancement is due to the interfacial hybridization of O 2p and Co 3d orbitals at the Co/MgO interface. The perpendicularly magnetized Au/Co/MgO/GaN heterostructures are promising for efficient spin injection and detection in GaN based opto-electronics without any external magnetic field.

Optimal principal component analysis of STEM XEDS spectrum images

STEM XEDS spectrum images can be drastically denoised by application of the principal component analysis (PCA). This paper looks inside the PCA workflow step by step on an example of a complex semiconductor structure consisting of a number of different phases. Typical problems distorting the principal components decomposition are highlighted and solutions for the successful PCA are described. Particular attention is paid to the optimal truncation of principal components in the course of reconstructing denoised data. A novel accurate and robust method, which overperforms the existing truncation methods is suggested for the first time and described in details.