Abstract: Electron energy loss spectroscopy (EELS) is a well established technique in electron microscopy that yields information on the elemental content of a sample in a very direct manner. One of the persisting limitations of EELS is the requirement for manual identification of core-loss edges and their corresponding elements. This can be especially bothersome in spectrum imaging, where a large amount of spectra are recorded when spatially scanning over a sample area. This paper introduces a synthetic dataset with 736,000 labeled EELS spectra, computed from available generalized oscillator strength tables, that represents 107 K, L, M or N core-loss edges and 80 chemical elements. Generic lifetime broadened peaks are used to mimic the fine structure due to band structure effects present in experimental core-loss edges. The proposed dataset is used to train and evaluate a series of neural network architectures, being a multilayer perceptron, a convolutional neural network, a U-Net, a residual neural network, a vision transformer and a compact convolutional transformer. An ensemble of neural networks is used to further increase performance. The ensemble network is used to demonstrate fully automated elemental mapping in a spectrum image, both by directly mapping the predicted elemental content and by using the predicted content as input for a physical model-based mapping.

Abstract: Electron donors are widely exploited in visible-light photocatalytic hydrogen production. As a typical electron donor pair and often the first choice for hydrogen production, the sodium sulfide-sodium sulfite pair has been extensively used. However, the resultant thiosulfate ions consume the photogenerated electrons to form an undesirable pseudocyclic electron transfer pathway during the photocatalytic process, strongly limiting the solar energy conversion efficiency. Here, we report novel and bioinspired electron donor pairs offering a noncyclic electron transfer pathway that provides more electrons without the consumption of the photogenerated electrons. Compared to the state-of-the-art electron donor pair Na2S-Na2SO3, these novel Na2S-NaH2PO2 and Na2S-NaNO2 electron donor pairs enable an unprecedented enhancement of up to 370% and 140% for average photocatalytic H-2 production over commercial CdS nanoparticles, and they are versatile for a large series of photocatalysts for visible-light water splitting. The discovery of these novel electron donor pairs can lead to a revolution in photocatalysis and is of great significance for industrial visible-light-driven H-2 production. [GRAPHICS] .

Abstract: Cryogenic electronmicroscopy can provide high-resolution reconstructionsof macromolecules embedded in a thin layer of ice from which atomicmodels can be built de novo. However, the interactionbetween the ionizing electron beam and the sample results in beam-inducedmotion and image distortion, which limit the attainable resolutions.Sample charging is one contributing factor of beam-induced motionsand image distortions, which is normally alleviated by including partof the supporting conducting film within the beam-exposed region.However, routine data collection schemes avoid strategies wherebythe beam is not in contact with the supporting film, whose rationaleis not fully understood. Here we characterize electrostatic chargingof vitreous samples, both in imaging and in diffraction mode. We mitigatesample charging by depositing a single layer of conductive grapheneon top of regular EM grids. We obtained high-resolution single-particleanalysis (SPA) reconstructions at 2 & ANGS; when the electron beamonly irradiates the middle of the hole on graphene-coated grids, usingdata collection schemes that previously failed to produce sub 3 & ANGS;reconstructions without the graphene layer. We also observe that theSPA data obtained with the graphene-coated grids exhibit a higher b factor and reduced particle movement compared to dataobtained without the graphene layer. This mitigation of charging couldhave broad implications for various EM techniques, including SPA andcryotomography, and for the study of radiation damage and the developmentof future sample carriers. Furthermore, it may facilitate the explorationof more dose-efficient, scanning transmission EM based SPA techniques.

Abstract: We explore the flatness of conduction and valence bands of interlayer excitons in MoS2/WSe2 van der Waals heterobilayers, tuned by interlayer twist angle, pressure, and external electric field. We employ an efficient continuum model where the moire pattern from lattice mismatch and/or twisting is represented by an equivalent mesoscopic periodic potential. We demonstrate that the mismatch moire potential is too weak to produce significant flattening. Moreover, we draw attention to the fact that the quasi-particle effective masses around the Gamma-point and the band flattening are reduced with twisting. As an alternative approach, we show (i) that reducing the interlayer distance by uniform vertical pressure can significantly increase the effective mass of the moire hole, and (ii) that the moire depth and its band flattening effects are strongly enhanced by accessible electric gating fields perpendicular to the heterobilayer, with resulting electron and hole effective masses increased by more than an order of magnitude – leading to record-flat bands. These findings impose boundaries on the commonly generalized benefits of moire twistronics, while also revealing alternative feasible routes to achieve truly flat electron and hole bands to carry us to strongly correlated excitonic phenomena on demand.

Abstract: Hybrid pixel direct electron detectors are gaining popularity in electron microscopy due to their excellent properties. Some commercial cameras based on this technology are relatively affordable which makes them attractive tools for experimentation especially in combination with an SEM setup. To support this, a detector characterization (Modulation Transfer Function, Detective Quantum Efficiency) of an Advacam Minipix and Advacam Advapix detector in the 15–30 keV range was made. In the current work we present images of Point Spread Function, plots of MTF/DQE curves and values of DQE(0) for these detectors. At low beam currents, the silicon detector layer behaviour should be dominant, which could make these findings transferable to any other available detector based on either Medipix2, Timepix or Timepix3 provided the same detector layer is used.

Abstract: Layered Li-rich oxides, demonstrating both cationic and anionic redox chemistry being used as positive electrodes for Li-ion batteries,have raised interest due to their high specific discharge capacities exceeding 250 mAh/g. However, irreversible structural transformations triggered by anionic redox chemistry result in pronounced voltagefade (i.e., lowering the specific energy by a gradual decay of discharge potential) upon extended galvanostatic cycling. Activating or suppressing oxygen anionic redox through structural stabilization induced by redox-inactivecation substitution is a well-known strategy. However, less emphasishas been put on the correlation between substitution degree and theactivation/suppression of the anionic redox. In this work, Ti4+-substituted Li2MnO3 was synthesizedvia a facile solution-gel method. Ti4+ is selected as adopant as it contains no partially filled d-orbitals. Our study revealedthat the layered “honeycomb-ordered” C2/m structure is preserved when increasing the Ticontent to x = 0.2 in the Li2Mn1-x Ti (x) O-3 solidsolution, as shown by electron diffraction and aberration-correctedscanning transmission electron microscopy. Galvanostatic cycling hintsat a delayed oxygen release, due to an improved reversibility of theanionic redox, during the first 10 charge-discharge cyclesfor the x = 0.2 composition compared to the parentmaterial (x = 0), followed by pronounced oxygen redoxactivity afterward. The latter originates from a low activation energybarrier toward O-O dimer formation and Mn migration in Li2Mn0.8Ti0.2O3, as deducedfrom first-principles molecular dynamics (MD) simulations for the“charged” state. Upon lowering the Ti substitution to x = 0.05, the structural stability was drastically improvedbased on our MD analysis, stressing the importance of carefully optimizingthe substitution degree to achieve the best electrochemical performance.

Abstract: e investigate atmospheric pressure microwave (MW) plasma (2.45 GHz) conversion in CO2 and CH4 mixtures (i.e., dry reforming of methane, DRM) focusing on reaction performance and carbon formation. Promising energy costs of ~2.8–3.0 eV/molecule or ~11.1–11.9 kJ/L are amongst the best performance to date considering the current state-of-the-art for plasma-based DRM for all types of plasma. The conversion is in the range of ~46–49% and ~55–67% for CO2 and CH4, respectively, producing primarily syngas (i.e., H2 and CO) with H2/CO ratios of ~0.6–1 at CH4 fractions ranging from 30% to 45%. Water is the largest byproduct with levels ranging ~7–14% in the exhaust. Carbon particles visibly impact the plasma at higher CH4 fractions (> 30%), where they become heated and incandescent. Particle luminosity increases with increasing CH4 fractions, with the plasma becoming unstable near a 1:1 mixture (i.e., > 45% CH4). Electron microscopy of the carbon material reveals an agglomerated morphology of pure carbon nanoparticles. The mean particle size is determined as ~20 nm, free of any metal contamination, consistent with the electrode-less MW design.

Abstract: From ripples to defects, edges and grain boundaries, the 3D atomic structure of 2D materials is critical to their properties. However the damage inflicted by conventional 3D analysis precludes its use with fragile 2D materials, particularly for the analysis of local defects. Here we dramatically increase the potential for precise local 3D atomic structure analysis of 2D materials, with both greatly improved dose efficiency and sensitivity to light elements. We demonstrate light atoms can now be located in complex 2D materials with picometer precision at doses 30 times lower than previously possible. Moreover we demonstrate this using WS2, in which the light atoms are practically invisible to conventional methods at low doses. The key advance is combining the concept of few tilt tomography with highly dose efficient ptychography in scanning transmission electron microscopy. We further demonstrate the method experimentally with the even more challenging and newly discovered 2D CuI, leveraging a new extremely high temporal resolution camera.

Abstract: Sr2FeO3+x F1-x (x = 0.08, 0.2), an n = 1 Ruddlesden-Popperphase, was synthesized from the oxidationof Sr2FeO3F in air at high temperature followinga fluorine for oxygen substitution and Fe3+ to Fe4+ oxidation. A structural investigation of both compounds was performedusing complementary and high-resolution techniques (Synchrotron X-rayand electron diffraction, Mo''ssbauer spectroscopy, HR-STEM)coupled to DFT calculation. This study reveals that oxidation leadsto a high degree of apical anion disorder coupled to antiphase boundaries. Sr2FeO3F, an oxyfluoride compoundwith an n = 1 Ruddlesden-Popper structure,was identifiedas a potential interesting mixed ionic and electronic conductor (MIEC).The phase can be synthesized under a range of different pO(2) atmospheres, leading to various degrees of fluorinefor oxygen substitution and Fe4+ content. A structuralinvestigation and thorough comparison of both argon- and air-synthesizedcompounds were performed by combining high-resolution X-ray and electrondiffraction, high-resolution scanning transmission electron microscopy,Mo''ssbauer spectroscopy, and DFT calculations. While the argon-synthesizedphase shows a well-behaved O/F ordered structure, this study revealedthat oxidation leads to averaged large-scale anionic disorder on theapical site. In the more oxidized Sr2FeO3.2F0.8 oxyfluoride, containing 20% of Fe4+, two differentFe positions can be identified with a 32%/68% occupancy (P4/nmm space group). This originates due to the presenceof antiphase boundaries between ordered domains within the grains.Relations between site distortion and valence states as well as stabilityof apical anionic sites (O vs F) are discussed. This study paves theway for further studies on both ionic and electronic transport propertiesof Sr2FeO3.2F0.8 and its use in MIEC-baseddevices, such as solid oxide fuel cells.

Abstract: Twin boundaries have been exploited to stabilize ultrafine grains and improve mechanical properties of nanomaterials. The production of the twin boundaries and nanotwins is however prohibitively challenging in carbide ceramics. Using a scanning transmission electron microscope as a unique platform for atomic-scale structure engineering, we demonstrate that twin platelets could be produced in carbides by engineering antisite defects. The antisite defects at metal sites in various layered ternary carbides are collectively and controllably generated, and the metal elements are homogenized by electron irradiation, which transforms a twin-like lamellae into nanotwin platelets. Accompanying chemical homogenization, alpha-Ti3AlC2 transforms to unconventional beta-Ti3AlC2. The chemical homogeneity and the width of the twin platelets can be tuned by dose and energy of bombarding electrons. Chemically homogenized nanotwins can boost hardness by similar to 45%. Our results provide a new way to produce ultrathin (< 5 nm) nanotwin platelets in scientifically and technologically important carbide materials and showcase feasibility of defect engineering by an angstrom-sized electron probe.

Abstract: Interface engineering in complex oxide superlattices is a growing field, enabling manipulation of the exceptional properties of these materials, and also providing access to new phases and emergent physical phenomena. Here we demonstrate how interfacial interactions can induce a complex charge and spin structure in a bulk paramagnetic material. We investigate a superlattice (SLs) consisting of paramagnetic LaNiO3 (LNO) and highly spin-polarized ferromagnetic La2/3Ca1/3MnO3 (LCMO), grown on SrTiO3 (001) substrate. We observed emerging magnetism in LNO through an exchange bias mechanism at the interfaces in X-ray resonant magnetic reflectivity. We find non-symmetric interface induced magnetization profiles in LNO and LCMO which we relate to a periodic complex charge and spin superstructure. High resolution scanning transmission electron microscopy images reveal that the upper and lower interfaces exhibit no significant structural variations. The different long range magnetic order emerging in LNO layers demonstrates the enormous potential of interfacial reconstruction as a tool for tailored electronic properties.

Abstract: In recent years, quantum dots (QDs) have emerged as bright,color-tunablelight sources for various applications such as light-emitting devices,lasing, and bioimaging. One important next step to advance their applicabilityis to reduce particle-to-particle variations of the emission propertiesas well as fluctuations of a single QD's emission spectrum,also known as spectral diffusion (SD). Characterizing SD is typicallyinefficient as it requires time-consuming measurements at the single-particlelevel. Here, however, we demonstrate multiparticle spectroscopy (MPS)as a high-throughput method to acquire statistically relevant informationabout both fluctuations at the single-particle level and variationsat the level of a synthesis batch. In MPS, we simultaneously measureemission spectra of many (20-100) QDs with a high time resolution.We obtain statistics on single-particle emission line broadening fora batch of traditional CdSe-based core-shell QDs and a batchof the less toxic InP-based core-shell QDs. The CdSe-basedQDs show significantly narrower homogeneous line widths, less SD,and less inhomogeneous broadening than the InP-based QDs. The timescales of SD are longer in the InP-based QDs than in the CdSe-basedQDs. Based on the distributions and correlations in single-particleproperties, we discuss the possible origins of line-width broadeningof the two types of QDs. Our experiments pave the way to large-scale,high-throughput characterization of single-QD emission propertiesand will ultimately contribute to facilitating rational design offuture QD structures.

Abstract: Quantification of annular dark field (ADF) scanning transmission electron microscopy (STEM) images in terms
of composition or thickness often relies on probe-position integrated scattering cross sections (PPISCS). In
order to compare experimental PPISCS with theoretically predicted ones, expensive simulations are needed for
a given specimen, zone axis orientation, and a variety of microscope settings. The computation time of such
simulations can be in the order of hours using a single GPU card. ADF STEM simulations can be efficiently
parallelized using multiple GPUs, as the calculation of each pixel is independent of other pixels. However, most
research groups do not have the necessary hardware, and, in the best-case scenario, the simulation time will
only be reduced proportionally to the number of GPUs used. In this manuscript, we use a learning approach and
present a densely connected neural network that is able to perform real-time ADF STEM PPISCS predictions as
a function of atomic column thickness for most common face-centered cubic (fcc) crystals (i.e., Al, Cu, Pd, Ag,
Pt, Au and Pb) along [100] and [111] zone axis orientations, root-mean-square displacements, and microscope
parameters. The proposed architecture is parameter efficient and yields accurate predictions for the PPISCS
values for a wide range of input parameters that are commonly used for aberration-corrected transmission
electron microscopes.

Abstract: The catalytic performance of heterogeneous catalysts can be tuned by modulation of the size and structure of supported transition metals, which are typically regarded as the active sites. In single-atom metal catalysts, the support itself can strongly affect the catalytic properties. Here, we demonstrate that the size of cerium dioxide (CeO2) support governs the reactivity of atomically dispersed palladium (Pd) in carbon monoxide (CO) oxidation. Catalysts with small CeO2 nanocrystals (~4 nanometers) exhibit unusually high activity in a CO-rich reaction feed, whereas catalysts with medium-size CeO2 (~8 nanometers) are preferred for lean conditions. Detailed spectroscopic investigations reveal support size–dependent redox properties of the Pd-CeO2 interface.

Abstract: Research efforts on the description of the low-temperature magnetic order and electronic properties of U3O8 have been inconclusive so far. Reinterpreting neutron scattering results, we use group representation theory to show that the ground state presents collinear out-of-plane magnetic moments, with antiferromagnetic coupling both in-layer and between layers. Charge order relieves the initial geometric frustration, generating a slightly distorted honeycomb sublattice with Néel-type order. The precise knowledge of the characteristics of this magnetic ground state is then used to explain the fine features of the band gap. In this system, spin-orbit coupling (SOC) is of critical importance, as it strongly affects the electronic structure, narrowing the gap by ∼38%, compared to calculations neglecting SOC. The predicted electronic structure actually explains the salient features of recent optical absorption measurements, further demonstrating the excellent agreement between the calculated ground state properties and experiment.