Symposium CM06 : Frontiers in Functional Imaging in Aberration-Corrected Electron Microscopy

2D crystals can be layered together to create new van der Waals crystals with bespoke properties. However, the performance of such materials is strongly dependent on the quality of the crystals and the interfaces at the atomic scale. Transmission electron microscopy (TEM) is the only technique able to characterize the nature of buried interfaces in these engineered van der Waals crystals and hence to provide insights into their optical, electronic and mechanical properties. I will report the use of scanning TEM imaging and analysis to aid the development of 2D heterostructures.

For example, it has been shown that confinement between two closely space graphene sheets can have a dramatic effect on the confined material [1,2]. By studying van der Waals structures where encapsulated graphene layers contain channels it is possible to measure the transport behavior of water through such channels. We observe significant enhancement in fluid flow rates for few atomic layer channel heights compared to larger channels [1]. We also observe that confinement in such channels can drive chemical transformations in aqueous salts [2].

Encapsulation with inert 2D crystals (e.g. graphene or hBN) also provides environmental protection from air or vacuum. We have performed mechanical exfoliation of air sensitive 2D materials in an inert argon atmosphere and used hBN or graphene encapsulation to allow the novel electrical properties of air sensitive 2D crystals to be realized [3]. However we find that even when fabricated in an inert atmosphere, NbSe₂ monolayers contain point defects that can be observed by high resolution TEM [4]. Furthermore, cross sectional scanning TEM imaging reveals that even comparatively stable materials like MoSe₂ and WSe₂ have different interlayer separation when exfoliated in a glove box compared to fabrication in air [5]. Finally, we will demonstrate how 2D heterostructures can be used to fabricate TEM liquid cells with precisely controlled liquid volumes and advanced imaging capabilities.

[1] Boya et al. Nature, 538, 222–225 (2016)
[2] Vasu et al. Nature Communications, 7, 12168 (2016)
[3] Cao et al., Nano Letters, 15, 4914-4921 (2015)
[4] Nguyen et al., ACS Nano, 11, 2894-2904 (2017)
[5] Rooney et al., Nano Letters, 2017, 17, 5222–5228 (2017)

Intriguing phenomena arising at the atomic scale have been found in functional complex oxide materials in recent years due to significant progress in technical and methodological development in scanning transmission electron microscopy (STEM). New possibilities for direct visualization of light elements have become feasible with introduction of the annular bright-field (ABF)-STEM technique [1]. Simultaneous acquisition of high-angle annular dark-field (HAADF)- and ABF-STEM images has enabled combined imaging of materials systems consisting of light and heavy elements.
In this work, high quality NdBa₂Cu₃O₇ (NBCO) thin films have been deposited on SrTiO₃ (STO) substrate by sputtering. The stacking sequence of NBCO layers in direction of the crystallographic c-axis is as follows: CuO-BaO-CuO₂-Nd-CuO₂-BaO [2]. The perovskite-type structure layers of NBCO are separated by CuO₂ planes with Nd atoms present in-between the copper-oxygen planes. Chains of CuO run parallel to the copper-oxygen planes with barium atoms placed between the planes and chains. Variation of the oxygen content in NBCO results in major changes of its physical properties [2]. In the case of non-stoichiometric NdBa₂Cu₃O_7-x, x denotes the amount of O vacancies present in the CuO chain. A previous report shows high dependency of T_c on the charge balance between the copper-oxygen chains and copper-oxygen planes [3]. The chain serves as a charge reservoir from which electrons are transferred to the copper-oxygen planes due to a decrease in oxygen content.
We have employed atomically resolved quantitative STEM imaging to investigate the NBCO film lattice with special emphasis on Cu-O distortion in NBCO by using an aberration-corrected JEOL JEM-ARM200F microscope equipped with a DCOR probe corrector operated at 200 kV. To improve the signal-to-noise ratio and to effectively reduce the image distortion, HAADF- and ABF-STEM images were acquired as series of 10 frames using a short acquisition time (1µs/pixel). Afterwards the images were aligned and merged. The simultaneously acquired HAADF- and ABF-STEM images enabled us to quantitatively analyze the local lattice and copper-apical-oxygen distances.
In addition, by employing extensive STEM image simulations, the lowest detectable oxygen concentrations present in the CuO chains, as visible in ABF images, were related to the sample thickness.

Reference:
[1] S.D. Findlay et al., Appl Phys Lett 95, 191913-1 (2009)
[2] H. Shaked et al., Phys Rev B 41, 4173 (1990)
[3] R. J. Cava et al., Physica C 165, 419 (1990)

MXenes are two-dimensional transition metal carbides or nitrides that have recently gained interest with applications geared towards energy storage, catalysis, and electronic devices. However, until now, bottom-up synthesis of MXenes have not been reported. In this work, using in situ aberration-corrected scanning transmission electron microscopy (STEM), we observed homoepitaxial growth of an additional hexagonal TiC (h-TiC) layer on both surfaces of 2D MXene (Ti₃C₂) monolayer flakes, forming new 2D MXenes Ti₄C₃ and Ti₅C₄, at temperature above 500 °C. The growth of single-layer h-TiC is controlled by a small diffusion barrier and a large step-edge barrier as revealed by density functional theory (DFT) and ReaxFF molecular dynamics simulations. The in situ heating experiments also reveal the edge structure of the MXene and the h-TiC add-layer, and the unique properties of the edges are understood using DFT. These findings thus provide insights on MXenes growth and pave the way to fabricate MXenes with controlled morphology for tailored functionality.

Research supported as part of the Fluid Interface Reactions, Structures and Transport (FIRST) Center, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences. Aberration-corrected STEM imaging was conducted as part of a user proposal at Oak Ridge National Laboratory’s Center for Nanophase Materials Sciences (CNMS), a U.S. Department of Energy Office of Science User Facility.

Electron energy-loss spectroscopy (EELS) and energy-dispersive X-ray spectroscopy are two of the most common means for chemical analysis in scanning transmission electron microscopy (STEM). The marked progress of the instrumentation hardware has made chemical analysis at atomic resolution readily possible nowadays. However, the acquisition and interpretation of atomically resolved spectra can still be problematic due to image distortions and poor signal-to-noise ratio (SNR) of the spectra, especially for investigation of energy-loss near-edge fine structures.

By combining multi-frame spectrum imaging and automatic energy-offset correction, we developed a spectrum imaging (SI) technique [1] implemented into customized DigitalMicrograph scripts for suppressing the image distortion and improving the signal-to-noise ratio of the EELS spectra. As widely used in STEM imaging, the former technique can efficiently remove scan distortions in the ADF and ABF images. The latter technique helps significantly reducing correlated noise [2], as for successive spectra different camera pixels are exposed which precludes amplification of small gain normalization errors. We implemented these techniques into STEM spectrum imaging for atomically resolved EELS elemental and fine structure mapping. Available methods for realizing the energy-offset on modern Gatan GIF quantum energy filters, and their reliability and influences on the final EELS energy resolution are tested and discussed. Using practical examples, we show that multi-frame SI and post-alignment can efficiently suppress image distortions and improve the final elemental map quality. We demonstrate that the energy-offset correction method reduces the correlated noise and helps resolving weak features of the near-edge fine structures. The final SI with improved SNR enables extracting individual component maps of the Ti-L_2,3 near-edge fine structure and of a Ti-O-Ti bonding direction map at atomic resolution, which has been theoretically predicted but extremely difficult to detect experimentally due to the poor SNR of the spectrum [3]. Combining the multi-frame SI and auto energy-offset-correction we demonstrate that these techniques will open new opportunities for atomically-resolved EELS fine structure mapping.

References:
[1] Y. Wang, M. R. S. Huang, U. Salzberger, K. Hahn, W. Sigle, P. A. van Aken, Ultramicroscopy, DOI: 10.1016/j.ultramic.2017.10.014.
[2] M. Bosman, V.J. Keast, Ultramicroscopy, 108 (2008) 837–846.
[3] M.J. Neish, N.R. Lugg, S.D. Findlay, M. Haruta, K. Kimoto, L.J. Allen, Phys. Rev. B, 88 (2013) 115120.

The discovery of Q-carbon has drawn a lot of attention in the past two years due to its interesting physical properties. Q-carbon is synthesized by rapid quenching (~10¹⁰ K/s) of highly undercooled carbon melt and is constituted of ~80% sp³ and ~20% sp² hybridized carbon. In the present study, we present a correlation of electronic structure of Q-carbon with the ferromagnetism and superconductivity properties. In contrast to the other diamagnetic derivatives of carbon, such as graphite, it is shown that Q-carbon nanostructures exhibit room temperature ferromagnetism with finite coercivity. Using electron energy-loss spectroscopy (EELS), we demonstrate that the C K-edge of Q-carbon consists of a sharp π* peak and a broad σ* peak. On comparing the C K-edge of amorphous Q-carbon with various diamond-like-carbon (DLC) films having a different sp³-sp² ratio, it is found that π* peak intensity is exceptionally high in spite of having just ~20% sp² content. This increase in the intensity corresponds to the increased unpaired spin electron density in Q-carbon due to the highly non-equilibrium synthesis route and gives rise to the room temperature ferromagnetism. Q-carbon, due to this dramatic increase in unpaired spin electron density, also exhibits the extraordinary Hall Effect characteristics.
Using EELS, we also demonstrate the correlation between superconductivity and the role of B doping in Q-carbon. We show that the nanosecond laser melting and rapid quenching of C results in strongly bonded unique superconducting phase of B-doped Q-carbon. This results into a type II superconductivity in B-doped Q-carbon with a transition temperature of 36.0±0.5 K. The EELS results show that we can achieve a homogeneously distributed B doping in Q-carbon as high as 17.0±1.0 at% with the employed synthesis process.^[1] An essential conduction for superconductivity in B-doped C is that B stays in sp³ hybridized state with carbon. We quantify that ~60% B atoms bond with sp³ hybridized C and contribute in the superconducting state of B-doped Q-carbon. With monochromated low-loss EELS and Raman spectroscopy, we demonstrate a higher electronic density of states near the Fermi energy level, which leads us to achieve remarkably high superconductivity transition temperature in B-doped Q-carbon. With this study, we present an insight on the role of electronic structure in achieving high-temperature superconductivity.
[1] Bhaumik, A; Sachan, R; Narayan, J. High-Temperature Superconductivity in Boron-Doped Q Carbon. ACS Nano 2017, DOI: 10.1021/acsnano.7b01294.

Recent advances in instrumentation, such as the introduction of advanced, high-resolution monochromators have allowed for new exciting experiments in the electron microscope. Spectroscopic signatures of optical and acoustical phonons, excitons and defect gap states are now accessible with an atom size probe and in tandem with high precision imaging. Here, we present results on the structure and electronic structure of thermoelectric (TE) materials for heat recovery applications, using advanced electron microscopy. Accurate information on the crystal structure and the resulting electronic properties is of paramount importance for understanding and predicting TE materials properties, as macroscopic quantities governing the material performance like the Seebeck coefficient and electronic conductivity are directly related to the electronic states in the vicinity of the Fermi level. High energy resolution scanning transmission electron microscopy (STEM) and electron energy loss spectroscopy (EELS) were used to inform theoretical predictions from density functional theory. In particular, we investigate the misfit-layered layered cobalt oxide bismuth strontium cobaltate (BSCO). We shed further light on the structure of this material, whose high efficiency makes it one of the most exciting TE oxides, highlighting the widespread occurrence of stacking faults, and finding that changes in the relative arrangements of the CoO2 and BiSrO layers can easily arise as their formation requires a very low energetic cost. We further demonstrate that Bi deficiency has a paramount importance on BSCO’s electronic, magnetic and transport properties. Atomic-layer-resolved electron energy loss spectroscopy shows how Bi vacancies lead to the hole-doping of the CoO2 layer which our theoretical modelling then demonstrates is in turn responsible for the high positive Seebeck coefficient of the material measured experimentally. Furthermore, we reveal the effect of the A-site occupancy in the structure and electronic structure in an A-site deficient perovskite system based on the Nd2/3TiO3 double perovskite. This system, another promising candidate for thermoelectric applications, has attracted significant attention due to the presence of a peculiar superstructure originating in part in cation vacancy ordering of the A-site. Using high precision atomically resolved monochromated core loss EELS measurements, acquired with an energy resolution better than 90 meV with a Nion UltraSTEM 100MC, it is possible to map individual components of the Ti L2,3 and O K near edge fine structures (ELNES). First-principles multiplet calculations are used to explain subtle changes in the ELNES, and associate them predominantly with Coulombic interactions from the A-sites. Annular Bright Field Imaging can then correlate the presence of tilting domains in the TiO6 sub lattice with these electronic structure changes observed by EELS.

Thin film synthesis methods developed over the past decades have unlocked emergent interface properties ranging from conductivity to ferroelectricity. However, our attempts to exercise control over interfaces are constrained by a limited understanding of growth pathways and kinetics, as well as by imprecise probes of local electronic structure. Here we examine the rearrangement of atomic planes at a polar / non-polar junction of LaFeO₃ (LFO) / n-SrTiO₃ (STO) using aberration-corrected scanning transmission electron microscopy (STEM) and electron energy-loss spectroscopy (STEM-EELS). While surface characterization confirms that substrates with two different (TiO₂ and SrO) terminations were prepared prior to LFO deposition, STEM-EELS measurements of the final heterojunctions reveal a predominantly LaO / TiO₂ interface configuration in both cases. Furthermore, we observe a reduction in the Fe L₂₃ edge that points toward possible interfacial conductivity. From ab initio simulations, we discuss several possible routes for the apparent disappearance of the FeO₂ / SrO interface. Our results illustrate how advanced characterization coupled with judicious control of deposition parameters may be used to map growth pathways, opening new avenues to control the structure and properties of functional interfacial systems.

CdTe is one of the most promising photovoltaic materials due to its direct band-gap and a high absorption coefficient. However, the practical efficiencies of poly-crystalline CdTe photovoltaic cells are still significantly below the theoretical limit. Reduction of non-radiative recombination at grain boundaries is believed to be the key to improve the efficiency of polycrystalline CdTe-based solar cells. Atomic-scale characterization of grain boundaries is crucial in developing a fundamental understanding how grain boundaries effect the efficiency. However, due to the small grain sizes, separating bulk and grain boundary contribution to the solar cell efficiency is nearly impossible.
In this work, atomic-resolution scanning transmission electron microscopy (STEM) is combined with in-situ heating experiments of poly-crystalline CdTe solar cells and CdTe bi-crystals to simulate PV cell aging on the atomic and electronic structures of the grain boundaries. More specifically, we use aberration-corrected electron microscope to identify gain boundaries in poly-crystalline solar cell. CdTe bi-crystals are synthesized to create a specific grain boundary model system of poly-crystalline solar cells. Using in-situ heating experiments, we measure how atomic structures of grain boundaries vary during the PV cell aging. Based on the atomic-resolution characterization, structural models are built to use for first- principles density functional theory (DFT) calculations to understand how grain boundaries and PV cell aging effect efficiencies of CdTe photovoltaic cells.

Multiferroic materials that exhibit simultaneous -and strongly coupled- magnetic and ferroelectric order above room temperature offer exciting potential for room-temperature device integration. Thus, magnetoelectric multiferroic films are ideal candidates for applications in next-generation memory devices which utilize low consuming electric fields to control magnetic order. However, due to competing requirements for displacive ferroelectricity and magnetism, only a hand-full of single-phase materials displaying multiferroic properties above room temperature are known. Most multiferroics are not suitable for practical applications either because they exhibit antiferromagnetic or weak ferromagnetic alignments, small spontaneous polarization, week coupling between the order parameters, or because their properties only emerge at extremely low temperatures. Therefore, much effort is devoted to search new single-phase multiferroic materials that exhibit high ordering temperatures.

Additionally, the multiferroic domain structures in these materials are considered to be an important factor to improve the efficiency and performance of future multiferroic devices. Therefore, it is crucial to investigate the domain structures in multiferroic oxides. Recent advances in aberration-corrected (scanning) transmission electron microscopy (S/TEM) and in microelectromechanical (MEMS) technology for miniaturized TEM specimen holders have opened up a wide range of new opportunities for in-situ studies. Thus, probing ferroelectric domain dynamics at atomic resolution by means of in-situ heating/electrical biasing TEM is now feasible thanks to the better spatial and temporal resolution of in-situ TEM.

In this contribution, we will show recent advances on the characterization of ferroelectric domain structures in multiferroic oxides. In particular, we will address three different multiferroic oxide systems. Firstly, we will show that substitutional Ca dopants in BiFeO₃ lead to the spontaneous formation of a layered structure of substitutional dopants which creates a complex ferroelectric structure constituted by the alternation of polar and non-polar domains. Secondly, the effect of doping and epitaxial strain on the structure and the ferroelectric properties of the potential multiferroic Aurivillius compound Bi₅FeTi₃O₁₅ will be discussed. Finally, the improper ferroelectric phase transition in the hexagonal YMnO₃ perovskite will be addressed.

Layered carbon nitrides have recently emerged as metal-free, visible-light absorbing semiconductors with growing interest as photocatalysts for hydrogen production¹. While transmission electron microscopy (TEM) and electron energy-loss spectroscopy (EELS) are powerful tools to characterize nanoscale features in (photo)catalysts, their application to layered carbon nitrides is difficult due to radiation damage. Graphitic carbon nitride (g-CN_xH_y) compounds, based on layers of amine-bridged heptazine (C₆N₇) chains², differentiated by their residual H-content, represent a wide-ranging class of material with ill-defined variation in structure. Molten salt synthesis routes have yielded crystalline, layered carbon nitrides based on triazine (C₃N₃) motifs with correspondingly less H-content³. These are referred to as poly(triazine imide) with intercalated halide ions (such as Li and Cl), or PTI/LiCl. Amine (e.g. N-H_x) content in both g-CN_xH_y and PTI/LiCl are correlated to changes in photocatalytic hydrogen evolution under visible-light, suggesting these defects can regulate optical absorption or charge transfer kinetics depending on how the host structure is modified. While infrared (IR) spectroscopy enables comparison of H-content, it lacks the spatial resolution needed for unambiguously correlating defects with catalytically relevant sites.

Here, we utilize vibrational EELS to locally probe bonding in g-CN_xH_y and PTI/LiCl, which are expected to contain a high and low degree of chemical heterogeneity, respectively. A radiolytic damage mechanism involving loss of amine groups is proposed based on variable dose-rate TEM imaging and core-loss EELS observations. To mitigate radiolysis, an ‘aloof-beam’ configuration is employed, meaning the electron beam is placed several nanometers outside the specimen by a distance called the ‘impact parameter’. By comparing to photon-based vibrational spectroscopies (e.g. IR absorption and Raman), the aloof-beam EELS resembles a broadened IR absorption spectrum with two major peaks attributed to C-N ring and N-H_x vibrations. Two types of heterogeneity in the vibrational EELS is observed: (i) an extra, weak peak at 265-meV associated with cyano (C≡N) defects and (ii) variation in the amine content defined as the ‘N-H’ to ‘C-N’ peak intensity ratio. By considering the consistency in probed-volume and lack of correlation with impact parameter, it is concluded that amine and cyano defects spatially-fluctuate throughout the g-CN_xH_y structure. On the other hand, PTI/LiCl does not exhibit spatially-varied cyano defects and the increased type-(ii) heterogeneity was attributed to radiolysis.

[1] X. Wang et al. Nat. Mater. 2009, 8, 76-80. [2] B. Lotsch et al. Chem. Eur. J. 2007, 13, 4969-80. [3] E. Wirnhier et al. Chem. Eur. J. 2011, 13, 3213-21. [4] We gratefully acknowledge the support of DOE grant DE-SC0004954, ASU’s John M. Cowley Center for High Resolution Electron Microscopy and ASU’s Center for Solid State Science.

Unusual transport or magnetic phenomena induced by octahedral distortions at the interfaces between substrates and thin films have been studied intensely for the last few years. In transition metal perovskite oxides, the rotation or tilt of the oxygen octahedra in the substrates can couple into the thin films, resulting in atomic or electronic structure configuration not seen in bulk materials. Ferromagnetic La_0.5Sr_0.5CoO_3-δ thin films grown on SrTiO₃ have been shown to favor periodic oxygen vacancy ordering as the result of interfacial strain from the substrate. Moreover, SrTiO₃ undergoes an antiferrodistortive phase transition from cubic to tetragonal structure at 105 K, driven by the TiO₆ octahedral rotation. This breaks the structural symmetry and induces orbital reconstructions in the interfacial Ti-O-Co bonds, which will distort both the CoO₆ octahedral and CoO₄ tetrahedral structure in the thin film. In this work, we use aberration-corrected scanning transmission electron microscopy (STEM) combined with in-situ cooling to investigate the atomic and electronic structure of La_0.5Sr_0.5CoO_3-δ thin films grown on SrTiO₃ with respect to the SrTiO₃ transition at 105K. Atomic-resolution imaging and electron energy-loss spectroscopy (EELS) are used to examine changes in the local density of states and magnetic moments as a function of sample temperature.

Acknowledgement: This work is supported by a grant from NSD (DMR-1408427). Support from the UIC Research Resources Center (RRC), in particular A.W. Nicholls and F. Shi is acknowledged. Work at UMN is supported by the DOE.

The ultimate performance of semiconductor materials and devices relies on the ability to characterize and optimize their structural, chemical, and electronic properties at the nanoscale. While high-resolution structural characterization via aberration-corrected scanning transmission electron microscopy (STEM) is approaching routine, the same cannot be said yet of electronic structure/properties analysis. However, with drastic improvements in energy resolution via recent advances in monochromation, low-loss electron energy-loss spectroscopy (EELS) has the potential to fill this much-needed role. Here, we present work on the application of low-loss EELS, using a monochromated FEI Titan3 G2 STEM, toward two important electronic characterization thrusts in CuIn1-xGaxSe2 (CIGS) solar cells: spatially-resolved bandgap profiling using a newly-developed analysis approach, and detection of sub-gap defect states with high spatial and energy space accuracy.

Compositional nonuniformity can play a significant role in device performance due to resultant changes within a material’s bandgap. However, accurate characterization of such issues, especially in a complex, phase-rich material system like CIGS, is far from straightforward. A new, simplified bandgap extraction method, based on straightforward Gaussian fit model, was developed to enable more rapid and robust bandgap profiling. The applicability of this technique was demonstrated within the cross-section of a CIGS solar cell containing intentional Ga/(Ga+In) composition (and thus bandgap) gradients. Comparison of the EELS-based bandgap profile to the nominal profile calculated using STEM-based energy dispersive X-ray spectroscopic composition data shows excellent spatially-resolved agreement. While this approach sacrifices a small degree of absolute (systematic) accuracy, excellent internal precision is maintained, and the effectively intervention-free methodology improves analytical speed and robustness.

Electrically active defects, a well-known problem in CIGS (and many other semiconductor materials), limit achievement of maximum device performance. However, as before, correlating these defect levels with specific defect structures is exceptionally difficult. To this end, we present work on the energy- and spatially-resolved detection of sub-gap defect levels within two different CIGS samples with two different trap energies (EV + 0.43 eV and EV + 0.56 eV). Low-loss EELS is shown to not only enable spatially-resolved detection of these states, but is also found to provide identical energies to those obtained using conventional deep level transient spectroscopy (DLTS). Furthermore, correlation between a new scanned probe DLTS method and low-loss EELS show accurate correlation in both spatial localization and sub-gap energy position. Taken together, these results indicate the potential of high-resolution low-loss EELS for the accurate nanoscale characterization of important electronic structure details.

All-inorganic perovskite nanocrystals (CsPbX₃, X=Cl,Br,I; IP-NCs) enjoy currently great research interest due to their stability and outstanding optical properties and low production costs, serving as promising candidates for optoelectronic and photovoltaic applications. The IP-NCs combine the advantages of perovskites (bandgap engineering through composition, low temperature, fast and low-cost synthesis) with the benefits of quantum confinement. As the NC diameter decreases and approaches the Bohr radius, the quantum confinement (QC) sets in, modifying the wave function of the free electron and hole. Hence, the energy band structure is affected and their bandgap energy increases. Here we synthesize CsPbBr₃ NCs, and investigate the bandgap energies of individual IP-NCs making use of electron energy loss spectroscopy (EELS) in a state-of-the-art low-voltage monochromatic scanning transmission electron microscope (STEM).[1] This is made possible due to recent developments in the low-energy monochromatic transmission electron microscope with advanced aberration correctors have enabled the combination of electron spectroscopy with ultrahigh spatial (below 1.6 Å) and energy resolutions (FWHM of the zero loss peak ~50 meV).[2,3] In that way, the absorption spectrum is directly correlated with the structural parameters of a single NC. A direct relation between the NC size and its bandgap is obtained on a single object level. We demonstrate that the bandgap is governed by the smallest dimension of the cuboidal perovskite NC. Further, we explicitly show an effective coupling between proximal NCs in an ensemble, leading to their band structure modification. In addition, our methods allowed us to discover the presence of non-perovskite configurations, which we identified as nanocrystals of Cs₄PbBr₆ and CsPbBr₃/Cs₄PbBr₆ nanohybrids.[4] Cs₄PbBr₆ introduces additional emission and absorption bands, which affect the emission quantum yield of the ensemble in the UV. These results highlight the incredible developments in the fields of microscopy techniques and materials science, and are of a general interest to the scientific community working on perovskites.

[1] L. Gomez, J. Lin & C. de Weerd et al., Nano Lett. 2016, 16.
[2] L.H. Tizei et al. Phys. Rev. Lett. 2015, 114.
[3] T. Sasaki, et al., Ultramicroscopy 2014, 145.
[4] de Weerd, J. Lin & L. Gomez et al., J. Phys.Chem. C. 2017, 121.

Intermediate temperature solid oxide fuel cells (IT-SOFC) are an attractive energy source that can be integrated into the current power generation infrastructure. IT-SOFC are capable of converting a wide range of carbonaceous fuels into energy. CeO₂ (ceria) functionalized with catalytic nanoparticles has demonstrated potential as an anode which may be capable of fuel reforming for these systems. Oxygen exchange is a critical reaction that controls the functionality of many oxide-based electrochemical processes. Understanding the key factors which impact the oxygen exchange reaction at the surface is essential in order to integrate and adopt the next generation of fuel flexible, renewable energy technologies. Here we explore how structural motifs evolve on CeO₂ surfaces and how these features manifest themselves in high-resolution transmission electron microscope (TEM) images.
To gain fundamental insight into the factors which regulate the oxygen exchange mechanism at various surfaces of ceria nanoparticles, molecular dynamics (MD) simulations will be performed on a number of vacancy configurations. Different vacancy configurations and concentrations will be considered on (111), (110), and (001) surfaces, as well as step edges, to obtain the equilibrated surface structures. The surface energy of each structure will be obtained to evaluate the stability trends for the various proposed structures, as well as the potential energy of edge and corner site atoms. The high resolution TEM images associated with these structural motifs will be simulated and compared with experiment. The structures and simulated images may provide insights into vacancy configurations and will provide a theoretical framework for interpreting experimental images.
[1] We gratefully acknowledge ASU’s HPC staff for support and assistance with computing resources, NSF grant DMR-1308085, and ASU’s John M. Cowley Center for High Resolution Electron Microscopy.

Due to the rapid progresses in segmented/pixelated detector developments combined with the state-of-the-art aberration correction technologies, atomic-resolution differential phase contrast (DPC) imaging in scanning transmission electron microscopy (STEM) is now becoming feasible. It has been shown that atomic resolution DPC STEM can visualize atomic electric field, the field between positively charged atomic nucleus and negatively charged surrounding electron clouds [1-3]. In this study, we use high-speed segmented type detector combined with a center of mass detection method [4] to quantitatively image atomic electric field of atomic columns in crystalline materials and even inside individual single atoms. We show that the (projected) atomic electric field points outward from the center of the atomic columns very clearly, reflecting the presence of positive atomic nucleus and negative surrounding electrons. In this talk, current status and prospects for aberration-corrected DPC STEM with high-speed segmented detector will be discussed. This work was supported by SENTAN, JST, and the JSPS KAKENHI Grant number JP17H01316.


[1] N. Shibata et al., Nature Phys., 8, 611-615 (2012).
[2] K. Müller et al., Nature. Comm. 5, 5653 (2014).
[3] N. Shibata et al., Nature Comm. 8, 15631 (2017).
[4] R. Close et al., Ultramicroscopy 159, 124–137 (2015).

The development of novel materials and heterostructures relies heavily on the ability to correlate structural and functional properties across length scales, especially down to the nanometer or atomic scale. Advances in aberration corrected electron microscopy has allowed for real space imaging, diffraction, and spectroscopy at the atomic scale; moreover, it has improved the capabilities of in situ experiments in the electron microscope. One such example is imaging magnetic domain structures as a function of applied magnetic field and temperature using aberration corrected Lorentz transmission electron microscopy (LTEM) and, more recently, differential phase contrast scanning TEM (DPC-STEM). Both techniques operate with the specimen in variable applied magnetic field by shutting off or reducing the strength of the objective lens, which surrounds the specimen in most microscopes. With the specimen in field-free or low field conditions, a wide range of important and interesting magnetic structures can be studied in a large swath of field/temperature space.
Over the last few years, the magnetic skyrmion phase has piqued scientific interest for potential applications in next-generation electronic devices such as racetrack memory due to their robustness to defects and extremely low pinning energy, which can be orders of magnitude lower than that of domain wall motion. Skyrmions are chiral magnetic spin textures that can form either in isolation or in extended lattices, with sizes ranging from the micron to single nanometer scale. The most well-studied skyrmion form is the Bloch skyrmion which is circular like a vortex; however, in thin and ultra-thin film heterostructures, a different kind of skyrmion can form with radial symmetry called a Néel skyrmion.
In our work, perovskite bilayers of SrIO₃/SrRuO₃ are grown on (001) SrTiO₃ via off-axis magnetron sputtering. In this system, skyrmions have been reported to be sub-10nm in size and Néel-type, making them difficult to image using LTEM without large specimen tilts; however, using DPC-STEM, overall spatial resolution in the image, as well as sensitivity to the gradient of in-plane magnetism should be improved. Imaging skyrmions in these materials as a function of applied magnetic field and temperature, as well as investigating the effect of varying SrRuO₃ and SrIrO₃ layer thickness gives direct insight into the stability and properties of skyrmions in this type of system and is only possible through such advanced imaging techniques.

Intermediate temperature solid oxide fuel cells (IT-SOFC) are an attractive energy source that can be integrated into the current power generation infrastructure. IT-SOFC are capable of converting a wide range of carbonaceous fuels into energy. Doped CeO₂ (ceria) has demonstrated potential as an electrolyte for these systems. However, resistive grain boundaries (GB) in these polycrystalline materials lead to a decrease in the total ionic conductivity. Doping is one avenue to tailor material properties and increase GB ionic conductivity. Experimental evidence has shown that doping at different nominal concentrations could result in increased GB ionic conductivity relative to the undoped samples [1]. This effect has been attributed to increases in local dopant/solute concentrations at the GB. Probing local structural motifs associated with low migration energies for oxygen ion transport across grain boundaries is a very challenging problem. Low migration energy is often associated with more open regions at the grain boundary core. Electron energy-loss spectroscopy (EELS) performed with a focused electron probe in a scanning transmission electron microscopy is sensitive to local changes in bonding and coordination. Such an approach may offer the potential to detect sites in grain boundaries with more open structures leading to higher oxygen transport.
To investigate this question, computational modeling is employed using molecular dynamics to identify high and low migration energies sites. The energy-loss spectrum associated with the sites are then calculated to look for changes in the near-edge fine structure of the oxygen K-edge. First, preliminary electron energy-loss spectroscopy (EELS) simulations will be performed to study the effects of tensile and compressive strain on the O K-edge from pure CeO₂. Substitutional Ca atoms will also be simulated to gauge how dopant atoms effect the near-edge fine structure. The changes in spectra at low and high migration energy sites at grain boundaries will be explored to determine if there is an experimentally detectable difference. Select spectral simulation will be compared with experimental data from Ca doped grain boundaries [1]. These studies may provide insights into the utility of EELS to locate oxygen migration pathways at grain boundary cores.

[1] Bowman, W. J. et al. (2017) Nanoscale, DOI: 10.1039/C7NR06941C
[2] We gratefully acknowledge ASU’s HPC staff for support and assistance with computing resources, NSF grant DMR-1308085, and ASU’s John M. Cowley Center for High Resolution Electron Microscopy.

There are a variety of techniques for electromagnetic field mapping in STEM. Of particular interest are differential phase constrast (DPC) and ptychography, due to the increasing availability of segmented detector systems and fast-readout electron cameras. In differential phase contrast we relate the mean transverse deflection of the beam to the electromagnetic field at each probe position in the STEM scan. Ptychographic techniques use diffraction pattern measurements for each scan position of the probe to solve for the scattering potential and hence the electromagnetic field structure of the sample. Application of these techniques to smaller length-scales leads to some subtleties in the interpretation of experimental results.

There is a long history of STEM techniques being used to study micron-sized magnetic field domains. At these resolutions the magnetic field causes a rigid transverse diffraction shift of the electron probe. When these techniques are applied to objects less than a few tens of nm across, a more complicated diffraction plane redistribution occurs. It is possible to account for these redistributions using an analytic scattering model. We use experimental data from imaging a 17 nm wide p-n junction in [001] GaAs as a case study.

There is interest in applying STEM electromagnetic field mapping techniques at atomic resolution to directly measure charge transfer due to bonding. I will show quantitative reconstruction of the electrostatic potential of a MoS₂ monolayer as a example of the atomic resolution capabilities of this technique. However application of these techniques to thicker samples is difficult: I show that quantitative reconstruction of a 7.8 nm thick sample of SrTiO₃ sample fails. This is because it is necessary to assume a very thin object to apply DPC or ptychography (the projection approximation) and this assumption fails for even moderately thick samples (> 3 nm) at atomic resolution. Possible ways of circumventing this limitation are discussed.

Abstract not available.

The ability to perform high-resolution mapping of adsorbate molecules or layers and correlate this with atomic structure would provide a transformative new tool for investigating the surface chemistry taking place on nanoparticles. Such a tool can be used to detect dissociative and non-dissociative sites on adsorbent surfaces and study different bonding arrangements between the adsorbate and the adsorbent. Recent work using high-resolution electron energy-loss spectroscopy (EELS) in the scanning transmission electron microscope (STEM) has shown how surface and interfaces modify the spatial extent of vibrational excitations [1]. For a comprehensive understanding of the detection of such vibrational signals, experiments need to be performed on simple model systems. The spectra need to be compared with results from conventional vibrational spectroscopies like Raman and Fourier-transform infrared (FTIR) spectroscopy to understand how electron excitation differs from photon excitation. Radiation damage may be substantial when an electron probe is placed directly on the adsorbate layer of interest. We adopt the aloof beam EELS technique that makes use of the long-range Coulomb interaction between electrons and adsorbate molecules to minimize damage when the probe is placed just outside the adsorbate layer [3]. We explore three classes of adsorbate/substrate systems because of their overall scientific importance and their suitability for developing the aloof beam vibrational EELS technique, viz. PVP ligand shell on Au nanoparticles, CO on Pt nanoparticles supported on CeO₂ nanocubes, and CO₂ on MgO nanocubes.
All our results were obtained using a NION UltraSTEM 100 aberration-corrected microscope operated at 60 kV and equipped with a monochromator. The aloof beam energy-loss spectrum from the PVP ligand shell/Au nanoparticle system suffers from a low signal-to-noise ratio and consists of two weak and broad vibrational signals centered at 160 meV and 205 meV corresponding to peaks associated with the C-N stretch, CH₂ scissor and wag, and C=O stretch modes, convolved with the experimental energy-resolution. Comparison shows that the FTIR spectrum convolved with the EELS energy-resolution is a better match to the vibrational energy-loss spectrum than the similarly convolved Raman spectrum [4]. For CO chemisorbed on a Pt nanoparticle/CeO₂ nanocube system, diffuse reflectance infrared Fourier-transform spectroscopy (DRIFTS) was performed to provide a reference data set for the aloof beam EELS experiment. Further vibrational EELS results from the CO on Pt/CeO₂ and CO₂ on MgO systems will be presented, along with comparisons with theory, FTIR and Raman spectra.
References:
[1] K. Venkatraman et al., Microscopy (under review).
[2] Y. Borodko et al., J. Phys. Chem. B, 110 (2006), p. 23052.
[3] The support from National Science Foundation CHE-1508667 and the use of (S)TEM at John M. Cowley Center at Arizona State University is gratefully acknowledged.

Ferroelectric materials have become a prototypical example of functional oxides, attracting considerable interest both in fundamental research and device engineering. They have been utilized in a broad range of electronic, optical, and electromechanical applications and hold promise for the design of future high-density nonvolatile memories and multifunctional nanodevices. The functionality of the ferroelectric-based devices depends strongly on the structures of ferroelectric domains with different polarization orientations. Advanced imaging techniques based on aberration-corrected transmission electron microscopy (TEM) or scanning transmission electron microscopy (STEM) have been a powerful method to study the domain structures in ferroelectric thin films, allowing nanoscale polarization states to be resolved unambiguously with sub-Angstrom resolution. It has been shown that nanoscale ferroelectric structures under special boundary conditions may exhibit unusual polarization patterns that challenge the conventional understanding of ferroelectric polar states, and possess functional properties that can be useful for new applications¹. In recent years, several new techniques based on aberration-corrected TEM have been under developing, including the STEM scanning diffraction method that allows the electric field distribution in thin foils to be measured at atomic scale², the atomic electron tomography (AET) that allows the 3D coordinates of all the atoms in a nanosystem to be reconstructed accurately³, and the low-energy electron energy-loss spectroscopy (or vibrational spectroscopy, with energy resolution < 10 meV) that allows the local vibrational spectrum of TEM specimens to be probed with high spatial resolution⁴. These techniques provide new opportunities for the study of fundamental physics in ferroelectric nanostructures. Here, we demonstrate the capability of these techniques for the characterization of emergent phenomena in ferroelectric heterostructures. The obtained results provide insights into 1) the effects of polarization bound charge, interface built-in field, free charge compensation, and magnetoelectric coupling on polarization structures across domain walls and different types of interfaces; 2) the 3D polarization structures of domain walls, domain states induced by surface reconstruction, or exotic polarization states such as vortices; and 3) the interaction between the polarization and the lattice vibration (phonon modes) of ferroelectric materials.

[1] A. K. Yadav, et al., Nature 530, 198 (2016).
[2] N. Shibata, et al., Nat. Phys. 8, 611 (2012).
[3] Y. S. Yang, et al., Nature 542, 75 (2017).
[4] O. L. Krivanek, et al., Nature 514, 209 (2014).

Widely used vibrational spectroscopies like Raman and Fourier-transform infrared (FTIR) spectroscopy provide chemical fingerprints of bonding arrangements for materials characterization. The spatial resolution using photons as probes is limited by the photon wavelength. Advances in monochromation in the modern scanning transmission electron microscope (STEM) has made it possible to detect vibrational excitations using electron energy-loss spectroscopy (EELS) [1]. This new capability can be used to investigate the delocalized behavior of bulk as well as the localized behavior of surfaces and interfaces. For a comprehensive understanding of the technique, experiments need to be performed on simple model systems. We explore the spatial variation in the vibrational stretch signals from the bulk, surface, and interface when an electron beam is scanned across a SiO₂/Si interface. The results are interpreted it in terms of the classical dielectric theory. This investigation provides baseline data which can be used to further explore the influence of more complex geometries on vibrational modes in materials.
A 3 μm layer of SiO₂ on a Si wafer was prepared for STEM-EELS analysis by lifting out a focused ion beam (FIB) lamella using a Nova 200 NanoLab (FEI) FIB. A NION UltraSTEM 100 aberration-corrected electron microscope operated at 60 kV and equipped with a monochromator was used to perform EELS linescans across the SiO₂/Si interface. The energy-loss spectrum from SiO₂ shows the Si-O bond-stretch mode at 144 meV and bond-bend mode at 100 meV. An initial drop in the bond-stretch signal is observed when the electron probe is 200 nm from the Si due to the long-range nature of the electrostatic interaction. However, the distance from the interface at which this signal intensity drops to half its maximum value is 5 nm. Calculations show that surface effects influence the energy position of a vibrational signal, and the surface contribution dominates the energy-loss spectrum over bulk for typical TEM sample thicknesses (≤ 100 nm). We show that, in practice, nanometer resolution is possible when selecting a part of the SiO₂/Si interface signal that is at a different energy position than the bulk signal. Calculations also show that, at 60 kV, the signal in the SiO₂ can be treated non-relativistically (no retardation) while the signal in the Si, not surprisingly, is dominated by relativistic effects. The surface effect allows local thickness in thin SiO₂ films to be determined based on the peak energy. Further geometrical effects on the vibrational modes of the system from edges and corners will also be presented.
References:
[1] O.L. Krivanek et al., Nature, 514 (2014), p. 209.
[2] The support from National Science Foundation CHE-1508667 and the use of (S)TEM at John M. Cowley Center at Arizona State University is gratefully acknowledged.

Controlling the nanolocalization of light through engineering the optical responses of dielectric structures has drawn ever increasing interest, which relies on a fundamental understandings of how a material’s composition, size and morphology affect its optical behaviors. STEM EELS provides a unique way of detecting local optical responses, as the probing electron or the so-called fast electron in a STEM can be viewed as an evanescent source of supercontinuum light, which is often referred to as the virtual photons. Under certain circumstances, a target material may exhibit resonance excitations when interacting with this supercontinuum light source, causing energy losses of the fast electrons which can then be detected using EELS. Because this work focuses on studying resonant modes which are nanoparticle geometry specific, they are noted as “cavity modes”. Two oxides (TiO₂ nanoparticles and CeO₂ nanocubes) with similar refractive indices but different geometries are employed here as the target materials, and complicated spectral peaks within the bandgap regions of these two oxides are detected using monochromated EELS in the aloof beam mode. In this mode, the electron probe is positioned only a few nanometers away from the particle surfaces and signals are generated from delocalized electron-solid interactions [1]. Since energy losses smaller than the bandgap energies are detected, these signals are not the result of electronic excitations (significant bandgap states are not expected to present in these materials) but rather from the excitations of optical-frequency geometric modes in these oxides. Simulations based on classical electrodynamics are performed to interpret the complex spectral features. Combination of experiments and simulations reveal that many factors influence the energy and strength of the cavity modes, including materials’ refractive indices, particle aggregation and fast electron velocity. Analytical Mie analysis also reveals that these geometric cavity modes are encoded in the scattering properties of the oxide particles when they are exposed to light or electron irradiation.
[1] P.A. Crozier, Ultramicroscopy 180 (2017), 104.
[2] Funding from DOE (DE-SC0004954) and NSF CHE-1253775 and the use of NION microscope at John M. Cowley Center for Microscopy at Arizona State University are greatly acknowledged.

Novel methods in time-resolved electron microscopy, diffraction and spectroscopy promise unprecedented insight into the dynamics of structural, electronic and magnetic processes on the nanoscale. A key to the realization of such technologies is the generation of high-quality beams of ultrashort electron pulses. This talk will present our recent development of imaging and spectroscopy capabilities using laser-triggered field-emitter sources in electron microscopy [1].
Based on custom modifications of Schottky-field-emission transmission electron microscope, an ultrafast electron probe beam with temporal resolution down to 200 fs, a spectral width of 0.6 eV and sub-nanometer beam focusability is demonstrated.
First applications of this instrument will be discussed, which include studies of compressive and shear strain wave dynamics in thin films probed by ultrafast convergent beam electron diffraction [2], magnetization dynamics in nanostructures observed by ultrafast Lorentz microscopy[3], and imaging of optical near-fields using inelastic electron-light scattering. Moreover, new approaches to free-electron quantum optics [4,5] and the generation of attosecond pulse trains [6] will be introduced.

References:
[1] A. Feist et al., “Ultrafast transmission electron microscopy using a laser-driven field emitter: Femtosecond resolution with a high-coherence electron beam”, Ultramicroscopy 176, 63 (2017).
[2] A. Feist et al., “Nanoscale diffractive probing of strain dynamics in ultrafast transmission electron microscopy”, arXiv:1709.02805 (2017).
[3] N. Rubiano da Silva et al., “Nanoscale mapping of ultrafast magnetization dynamics with femtosecond Lorentz Microscopy”, arXiv:1710.03307 (2017).
[4] A. Feist et al., “Quantum coherent optical phase modulation in an ultrafast transmission electron microscope”, Nature
521, 200 (2015).
[5] K. E. Echternkamp et al., “Ramsey-type phase control of free electron beams”, Nature Phys. 12, 1000 (2016).
[6] K. E. Priebe et al., “Attosecond Electron Pulse Trains and Quantum State Reconstruction in Ultrafast Transmission Electron Microscopy”, arXiv:1706.03680 (2017).

Many processes in materials science, chemistry and biology take place in a liquid environment. In many of these cases, the final desired outcome of the process is a result of a series of complicated transients (occurring on timescales of milliseconds to nanoseconds), where a change in the order, magnitude or location in each of the steps in the process can lead to a radically different result. Understanding and subsequently controlling the final outcome therefore requires the ability to directly control and observe the kinetics of these transients as they happen. Additionally, if we would like to acquire a time sequence of images from the same transient event, the effect of the beam must be taken into account – the longer the sequence of events the more the beam has an effect. In this case, the aim is to more efficiently use the dose that is supplied to the sample and to extract the most (spatial and temporal) information from each image. For the STEM mode of operation, optimizing the dose/data content by the use of sub-sampling and inpainting can increase imaging speed, reduce electron dose by 1-2 orders of magnitude while at the same time compressing the data by the same amount. Here, we discuss the use of inpainting to generate high quality, interpretable images from sub-sampled datasets obtained from crystalline materials – the highly ordered structures allow the physical principles behind inpainting to be identified. The differences between crystalline approaches and the application of the same methods to 3-D, 4-D, spectroscopic and lower resolution in-situ images will be highlighted. New results showing the use of in-situ liquid stages to study nucleation and growth using inpainting will be presented and the potential insights that can be gained by increasing the image acquisition speed and/or decreasing the electron dose will be described. Importantly for all in-situ observations, the kinetic control of the nucleation and growth process using sub-sampling highlights the role of the interfaces in the cell in controlling the process. Sub-sampling and inpainting is not limited to STEM as similar methods can be used in TEM mode to increase the speed of any camera. The potential to apply these methods together to extract quantitative image information from a wide range of methods will also be discussed.

Colloidal quantum dots (CQDs) are attractive materials for lasers, displays and light-emitting applications due to their narrow and brighter spectral emission bandwidth, size-tunable bandgap and high-photoluminescence quantum yield (PLQY). However, these CQDs undergo inevitable degradation of their unique optical properties overt time due to their sensitive surface chemistry. To overcome these limitations, several approaches have recently been used such as over coating with an inorganic semiconductor shell of a wider band gap (core-shell hetrostructures), surface functionalization with new ligands or polymer coating and composites etc. In particular, core-shell heterostructured quantum dots with thick shell (so called “giant” quantum dots (g-QDs)) has shown higher PLQYs and improved photochemical stability than traditional thin-shelled or core only CQDs. The outstanding properties of g-QDs essentially depend on both the structure (defects, surface chemistry etc.) and the properties of interfacial layer (sharp or smooth core/shell interface). These structural and interfaces properties of g-QDs are strongly influenced and can be tailored by the synthetic parameters. Here, we will present our detailed characterization of the interface of the g-QDs nanocrystal both structurally (for defects) and as a function of elemental composition (for alloying) using aberration-corrected electron microscopy. These results will provide the deeper understanding needed to improve or understand better the optoelectronics properties of these core-shell nanocrystals.

Combinations of iron oxide nanoparticles with metallic nanoparticles have been extensively studied due to their large variety of applications in biomedical and technological fields [Souza, 2017; Moraes Silva, 2016]. Particularly, dual heterostructures composed of gold and iron-oxide have been made by following the so-called seeded-growth reaction sequence, involving the synthesis of gold nanoparticles seeds, which subsequently serve as nucleation and growth sites for the iron oxide phase [Fantechi, 2017]. In this context, to determine the composition and the iron-oxidation state of the iron oxide nanoparticles is of paramount importance to understand their functional properties.

This report focuses on Electron Energy Loss Spectroscopy (EELS) analysis of three sets of Au/Fe_yOx heterostructures to identify variations in the iron-oxidation state at the core and surface of the iron oxide nanoparticles, as well as at their interface with the gold counterpart. For this purpose, hybrid nanoparticles of Au/Fe_yO_x were produced employing three sets of Au seeds with 3 nm and 7 nm. The use of these seeds leads to hybrid systems with three morphologies; dimers, trimers, and flowers. High angular annular dark field (HAADF) STEM images were acquired on a double corrected FEI Titan Themis operated at 300 keV, while dual EELS was performed at 80 keV.

The results show a predominant magnetite (Fe₃O₄) phase at the iron oxide core, which is different from the surface layer that exhibits lower electron energy loss. Only for specific nanoparticle morphologies, an additional phase is shown, as a result of the reduction of magnetite to form wustite, which is predominant at the surface. These results may justify some peculiar phenomenologies often observed in this type of hybrid nanoparticles, as for example exchange bias effects or spin polarization transfer [Pineider, 2013; Feygenson, 2015].

[Fantechi, 2017]. E. Fantechi et. al., Chem. Mater. 2017, 29, 4022−4035
[Souza, 2017]. L. Souza da Costa. et. al. Catal. Today. 2017, 282 (2), 151-158
[Moraes Silva, 2016]. S. Moraes Silva. et. al. Chem. Commun., 2016, 52, 7528-7540.
[Pineider, 2013]. F. Pineider. et. al. ASCNano. 2013, 7 (1), 857-866.
[Feygenson, 2015]. M. Feygenson. et. al. Phys. Rev. B. 2015, 92, 054416

Data science approaches have been successfully applied to a variety of electron microscopy measurements, resulting in highly detailed analysis of composition, electronic structure, crystallography and atomic structure of complex systems. The goal in all of these approaches is to take a large set of measurements from a sample and then to utilise the redundancy in the data to recover a model, or set of significant components that accurate describe the real system.
Statistical decompositions are one widely used method for this, broadly using matrix factorisation methods to isolate those particular signals that describe the majority of the structured part of the data as efficiently as possible. This has proven a vaulable approach in the analysis of scanning diffraction data, especially in the fairly common situation where individual diffraction patterns can contain information about one or more overlapping phases in the microstructure being studied. This approach will be validated both with simulated as well as experimental data.
One major issue with any decomposition lies in determining how many significant components are needed to produce the most robust model of the measured data. Too few and real structural differences can be merged in with similar signals leading to the possibility of missing a significant change. Too many and the efficiency of the algorithm is lost and every piece of noise becomes a significant factor in the decomposition. In an attempt to make a more automatic guided decomposition the idea of data-clustering has been developed alongside decomposition. This uses the structure of the data (at least a suitably projected version of the data) to guide the correct number of components needed in the model. Under some circumstances the analysis of structure in the reduced dimensions needed can even provide physical insight into the structure and crystallography of the sample.

Transition metal dichalcogenides (TMDs), a major group of 2-dim materials beyond graphene, have shown extraordinary properties as candidates for future electronics. When the dimensions of materials reduce from bulk to 2-dim, the thermal expansion coefficient (TEC) dramatically increases due to the lack of the constraint from inter-layer interactions. The subsequent larger TEC mismatch becomes a significant problem in the design of the electronic nano-device. The knowledge of the thermal expansion of TMDs is one of the most important aspects of modern device design with such materials. Particularly, an effective method to control the thermal expansion coefficient of TMDs and an understanding of the related mechanism are needed.

In this contribution, we will introduce alloy engineering to tune the thermal expansion coefficients of monolayer Mo_1-xW_xS₂ and study the interplay between thermal expansion and local defects using a combination of scanning transmission electron microscope (STEM), electron energy loss spectroscopy (EELS) and first-principles calculations. More specifically, we will measure thermal expansion coefficient based on the plasmon energy shift in a range of temperatures and first-principles modeling of the low-loss EELS signals. With this combination, we have previously measured the thermal expansion coefficients of monolayer MoS₂ and WS₂, which shows a considerable mismatch with each other. Free-standing Mo_1-xW_xS₂ flakes are then prepared and we measure the local thermal expansion of flakes in different concentrations over a range of temperatures. It is important to note that the TEC can be tuned through controlling the doping concentration. Atomic-resolution Z-contrast images and the corresponding nanometer-scale thermal expansion maps will be used to explore the influence of local strains and defects on thermal expansion.


Acknowledgement: This work was supported by the National Science Foundation EFRI 2-DARE Grant 1542864. The acquisition of UIC JEOL JEM ARM200CF is supported by an MRI-R2 grant from the National Science Foundation (Grant No. DMR-0959470). This work made use of instruments in the Electron Microscopy Service and the High Performance Computing Clusters at Research Resources Center, UIC.

Tuning the Metal-Insulator transition (MIT) of nikelate perovskites, RENiO₃, where RE refers to a rare-earth, is possible by changing the Rare-earth cation of the structure [1,2]. Alternatively, this transition can also be shifted by other mechanisms, for instance, inducing structural distortions via epitaxial strain [3]. However, it still remains unclear how the ReNiO₃ films accommodate when they are epitaxially grown onto different substrates. Here, we use the aberration corrected Scanning transmission Electron Microscope (STEM) to precisely characterize the atomic structure of tensile and compressive strained LaNiO₃ (LNO) and NdNiO₃ (NNO) thin films grown by chemical solution deposition (CSD) processes. Firstly, we discuss the influence of the selected substrate on the final defect landscape, which is different in each case. Secondly, we elucidate a fundamental link between strain and the commonest defect observed in nikelate films, the Ruddlesden-Popper fault (RPF), which will ultimately impinge on the electrical properties of the films. Finally, we identify the exact position of each atomic column by applying a center of mass refinement routine, enabling us to locally quantify, with subatomic resolution, any atomic-atomic spacing or displacement, with which we unveiled unforeseen polar-like distortions appearing on either side of the RPF’s rock salt fault.

[1] L. Medarde, J. Phys. Condens. Matter 9, 1679 (1997).
[2] G. Catalan, Phase Transitions 81, 729 (2008).
[3] S. Catalano, M. Gibert, V. Bisogni, F. He, R. Sutarto, M. Viret, P. Zubko, R. Scherwitzl, G. A. Sawatzky, and J. Triscone, 62506, (2015).

Authors acknowledge the MICIN (NANOSELECT, CSD2007-00041 and MAT2014-51778-C2-1-R), Severo Ochoa SEV-2015-0496 grant, the RyC-2012–11709 contract of J.G. and the Generalitat de Catalunya (2014SGR 753 and Xarmae) project. The STEM microscopy work was conducted in the ICTS-CNME at UCM as well as at the Laboratorio de Microscopias Avanzadas (LMA) at Instituto de Nanociencia de Aragon (INA) at the University of Zaragoza. Authors acknowledge the ICTS-CNME for offering access to their instruments and expertise.

To understand material properties and functionality at the fundamental level, one must know the 3D positions of atoms with high precision. For crystalline materials, crystallography has provided this information since the pioneering work of Max von Laue, William Henry Bragg, and William Lawrence Bragg over a century ago. However, perfect crystals are rare in nature. Real materials often contain crystal defects, surface reconstructions, nanoscale heterogeneities, and disorders, which strongly influence material properties and performance. Here, I present atomic electron tomography (AET) for 3D structure determination of crystal defects and disordered materials at the single-atom level. Using a Fourier based iterative algorithm, we first demonstrated electron tomography at 2.4-Å resolution without assuming crystallinity in 2012. We then applied AET to image the 3D structure of grain boundaries and stacking faults and the 3D core structure of edge and screw dislocations at atomic resolution. Furthermore, in combination of AET and atom tracing algorithms, we localized the coordinates of individual atoms and point defects in materials with a 3D precision of ~19 pm, allowing direct measurements of 3D atomic displacements and the full strain tensor. More recently, we determined the 3D coordinates of 6,569 Fe and 16,627 Pt atoms in an FePt nanoparticle, and correlated chemical order/disorder and crystal defects with material properties at the individual atomic level. We identified rich structural variety with unprecedented 3D detail including atomic composition, grain boundaries, anti-phase boundaries, anti-site point defects and swap defects. We showed that the experimentally measured coordinates and chemical species with 22 pm precision can be used as direct input for density functional theory calculations of material properties such as atomic spin and orbital magnetic moments and local magnetocrystalline anisotropy. Looking forward, AET will not only advance our ability in 3D atomic structure determination of crystal defects and disordered materials, but also transform our understanding of materials properties and functionality at the individual atomic level.

J. Miao, P. Ercius and S. J. L. Billinge, Science 353, aaf2157 (2016).
J. Miao, F. Förster and O. Levi, Phys. Rev. B. 72, 052103 (2005).
M. C. Scott et al., Nature 483, 444–447 (2012).
C. C. Chen et al., Nature 496, 74–77 (2013).
R. Xu et al., Nature Mater. 14, 1099-1103 (2015).
Y. Yang et al., Nature 542, 75-79 (2017).
A. Pryor Jr. et al., Sci. Rep. 7, 10409 (2017).

Complex crystals, formed by solid solutions or crystal growth, are characterized by locally varying composition, lattice, symmetry and chemical bonding. Examples include semiconductor heterostructures, where modulation of composition and lattice leads to quantum confinement, relaxor-ferrielectrics with locally varying polarization and doped high-T_c superconductors. Characterization of local lattice and bonding in these crystals is an outstanding challenge. Here, we report on a scanning CBED (convergent beam electron diffraction) approach to meet this challenge. Scanning CBED is performed by deflecting the electron probe and recording CBED patterns at each probe position. When it is combined with energy filtering, this technique provides all information carried by electron scattering, plus the advantage of scanning and localization using a nm-sized electron probe. In this talk, we will present methodologies for extracting local lattice, symmetry and bonding information and discuss their limitations. Applications to the determination of local lattice [1], symmetry and symmetry fluctuations in ferroelectric perovskites [2] and determination of bonding in oxides [3] will be demonstrated. Future applications of scanning CBED using fast detectors promises a full determination of crystal potential, electric and magnetic fields, however, realization of such promises require better understanding of electron diffraction through non-idealized potential and magnetic fields that requires new and improved inversion schemes. The research was supported by DOE BES.

[1] R Yuan, Y Meng, J Zhang, JM Zuo, Microscopy and Microanalysis 23 (S1), 180-181, 2017
[2] Y-T Shao, J-M Zuo, Acta Cryst. B73, 708-714, 2017
[3] JM Zuo, JCH Spence, Advanced Transmission Electron Microscopy, Springer Science+ Business Media New York, 2017

In scanning transmission electron microscopy (STEM), the highly convergent Å-size electron probe forms convergent beam electron diffraction (CBED) in the back focal plane of the objective lens. In four-dimensional (4D) STEM, a high-speed direct electron camera records a series of these CBED patterns as the beam is scanned. The CBED patterns are extremely rich in information, encoding information about the local electric field and magnetic field, local thickness and orientation, three-dimensional defect crystallography, phonon spectra and more. They are also a fairly large dataset, ranging from 10s of GBs to TBs for one scan of the sample.

Convolution neural networks (CNNs) may offer a computationally efficient, flexible, and accurate approach to identifying features in 4D STEM data sets and extracting meaningful information about the sample^[1]. Modern CNNs can classify features in natural images (photographs) when trained against a large data set of images tagged with the features present in each image. In 4D STEM, we can generate tagged training data using multislice simulations from known structures, use it train a CNN, then use the CNN to extract information about the sample from experimental CBED patterns.

As a first step, we propose an approach to measuring sample thickness from 4D STEM data using a CNN. We developed the CNN in Keras (Python) by applying the transfer learning method on a pre-trained VGG-16 model^[2]. The training dataset was prepared by performing frozen phonon simulations on a 100 nm SrTiO₃ model along [100] axis, and the simulated CBEDs are categorized in 2 nm steps. Random, Poisson-distributed noise is applied to the simulated CBEDs to imitate the noise in the experiment, and the training dataset was further expanded through image augmentation. Preliminary results suggest that the CNN achieves a >98% accuracy on predicting the sample thickness on the test dataset, similar to the performance achieved by Lebeau using a CNN to analyze position-averaged convergent beam electron diffraction data^[3]. An analogous CNN method to determine sample tilt will also be discussed.

A remarkable demonstration of the importance of the atomic-scale properties of interfaces is found in monolayer FeSe on SrTiO₃ that exhibits a superconducting gap consistent with a critical temperature an order of magnitude higher than that seen for the bulk material. Here a combination of aberration-corrected scanning transmission electron microscopy and quantum mechanical calculations is used to determine the atomic structure of the film and its interface with the substrate and to identify what drives the increase in T_c. A comparison of experimental images of the interfacial region with multislice STEM simulations based on the calculated structure shows essentially perfect agreement in all respects (structural and compositional) and leads us to conclude we have accurately determined the FeSe/ SrTiO₃ interface. Our calculations show that the interlayer in the FeSe/SrTiO₃ system is not merely a passive glue holding substrate and film together. We find that electrons in the interfacial layer (IL) form electronic distributions (at 0 K) that stabilize a small distortion of the FeSe lattice that has been shown to enhance T_c in a strained, freestanding monolayer of FeSe. The shear-like distortions in the FeSe monolayer that develop naturally from the interfacial layer are not seen in calculations of freestanding FeSe nor in FeSe on the usual TiO₂-terminated SrTiO₃ surface. The continued study of this system is expected to lead to insights into 2D superconductivity and other novel phenomena.

Strain is widely employed in semiconductor technology due to its ability to enhance carrier mobility, which provides an alternative to scaling for higher performance. For the length scale of latest generation devices like FinFET, electron nanodiffraction is the method of choice for strain characterization. A nano-sized probe is scanned across desired sample area and a series of diffraction patterns are acquired over a region of interest. With this technique, strain measurement can be done with high spatial resolution as well as a large field of view. The spatial resolution can be further improved by increasing the convergence angle of the electron beam. Thus, the technique eventually becomes scanning CBED (convergent beam electron diffraction), where diffraction spots are no longer sharp peaks but are disks. Identification of the disk position is crucial for strain analysis. Here we report on a systematic study of scanning CBED for high-accuracy strain measurement. We propose a new method to find the center of diffraction disk using circular Hough transform, and calculate the strain by fitting multiple disks. The method was applied to do strain mapping on a SiGe MOSFET. The limiting factors of measurement precision and accuracy are discussed to optimize the experimental conditions. Multislice simulations were done to examine the effects of 3D strain and curved lattices. The results are used to pave the way for strain characterization in complex crystals and nanostructures with high spatial resolution and accuracy.

Heat dissipation in integrated nanoscale devices is a major issue that requires the development of nanoscale temperature probes. Here, we report the implementation of a method that combines monochromated electron energy gain and loss spectroscopy to provide a direct measurement of the local temperature in the nano-environment. Loss and gain peaks corresponding to an optical-phonon mode in boron nitride were measured from room temperature to ~ 1300 °C. We find that both peaks present a red shift (towards lower energies) as the sample is heated up, with a linear behavior over the temperature range studied here. First-principles calculations reveal that the red shift is due to a combination of lattice thermal expansion and anharmonic phonon scattering, with the latter being the dominant factor to reduce the energy of the optical phonon as the temperature of the sample increases. The gain peak exhibits a clear increase of intensity as a function of temperature, in accordance with the occupation probability of the phonon energy state. The spectroscopy presented here shows that by detecting both gain and loss peaks, the local temperature of a material can be obtained directly by statistical principles; and in conjunction with theory, open the doors to the study of anharmonic effects in materials by directly probing phonons in the electron microscope [1].
[1] This research was supported by the Center for Nanophase Materials Sciences, which is a DOE Office of Science User Facility (JCI & RRU), and by the Materials Sciences and Engineering Division Office of Basic Energy Sciences, U.S. Department of Energy (ARL). This research used resources of the National Energy Research Scientific Computing Center, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. This work used the Extreme Science and Engineering Discovery Environment (XSEDE). Theoretical work at Vanderbilt University was supported by DOE grant DE-FG02-09ER46554 and by the McMinn Endowment (TLF, STP).

Determination of the electric field at the atomic scale in materials is critical in understanding the mechanisms of functional materials such as ferroelectrics, light emitting diodes and heterogeneous catalyst. To map the electric field in such small scale, differential phase contrast imaging (DPC) in scanning transmission electron microscopy (STEM) has been developed by using segmented electron detectors to collect scattered electrons. The variation in electron count among the detectors is related to the shift of the electron beam by the local electro-magnetic field in materials; when coupled with the high spatial resolution in aberration corrected STEM, DPC can reveal the electric field at atomic resolution. Recently, the development of high speed direct electron detectors makes it possible to acquire the entire diffraction pattern during the scanning of the electron probe at the speed of ~1000 fps, which could improve the accuracy in the quantification of electron momentum transfer over regular DPC. However, from previous experiments and theoretical calculations, quantitative correlation between electron momentum transfer and the electric field remains a challenge.

We carried out a study on the calculation of electric field based on the measured average momentum transfer in both STEM experiment and simulations. SrTiO3 is employed as the model system. The momentum transfer map was calculated using the shift of center of weight (CoW) of each diffraction pattern. In the maps from simulation, the momentum transfer is not seen to increase uniformly around all atomic columns as the thickness of the modeled sample increases. For structures with one unit cell thickness, the momentum transfer map always points radially inwards from all atomic columns, as would be expected based on the projection of the coulomb potential surrounding each positive nucleus and the negative charge of the electrons in the beam. However, for larger structures, while the measured field near oxygen atoms only increased in magnitude, near Sr and Ti sites, the field also switches direction, pointing radially outwards in some regions.

To understand the behavior described above, we have examined the 3D trajectory of the electron probe in the model using a multislice simulation. From our preliminary results, we found that the electron probe does not propagate following a straight pathway and that the deviation depends on the atomic number of the closest atomic column. The electron momentum transfer measured using the diffraction pattern is therefore the average electric field along the path, not the localized average electric field where the electron probe enters the sample surface.

Ferroelectric single crystals had been widely applied for medical imaging, actuation, and sensor technologies due to their piezoelectric effect. However, fundamental questions regarding the microscopic origin of ferroelectric phase transitions are still under debate despite many past efforts. Here, we demonstrate the energy-filtered scanning convergent beam electron diffraction (EF-SCBED) technique that can be simultaneously used for (1) the identification of polarization domains, (2) a determination of the local crystal symmetry within a single domain, and (3) identification of structural distortions.
Results from SCBED show that the crystal symmetry is not homogeneous; regions of few tens nm retaining almost perfect symmetry are interspersed in regions of lower symmetry. The highly symmetric regions confirm the acentric tetragonal, orthorhombic, and rhombohedral symmetry for the ferroelectric phases of BaTiO₃ at different temperatures, which is consistent with the displacive model of ferroelectric phase transitions in BaTiO₃. On the other hand, the observed nanoscale symmetry fluctuations are consistent with the predictions of order-disorder phase transition mechanism. For future work, the SCBED-based techniques, quantitative SCBED (SQCBED) and rocking SCBED (SRCBED), will be further developed for simultaneous identification of polarization, domain wall structures, crystal potential, and chemical bonding.

Scanning transmission electron microscopy (STEM) has traditionally involved recording a single, or very few, integrated signals as an electron probe is scanned across a sample. Advances in detector technology now allow the entire scattering distribution to be recorded, resulting in a four-dimensional (4D) dataset that includes two real-space dimensions and two reciprocal-space dimensions for every probe position. New fast cameras recently installed on low-voltage and monochromated aberration-corrected STEMs have greatly simplified the acquisition of these datasets. Each convergent beam diffraction pattern (CBED or Ronchigram) can contain as many pixels as a traditional TEM image, making these datasets extremely large. Thus the problem of how to extract the useful information remains a challenge.

The simplest approach is to mask regions of the data, simulating traditional STEM detectors. More advanced ptychographic methods allow details of the probe aberrations to be extracted, and potentially corrected, to improve the contrast or detectability of light atoms. Machine learning or deep data approaches will allow the local symmetry or changes in structure to be automatically identified and perhaps tracked in real-time.

Free-standing 2D materials, such as graphene, form an ideal test suite for new imaging modalities because every atom in the structure can be directly imaged. Moreover, they have some advantages over 3D samples, because unknown factors, such as background contributions from contamination or sample preparation and surface reconstructions, can be either imaged directly or ruled out. Two dimensional samples are also particularly amenable to simulations, since they contain a well-defined number of atoms. These patterns are scientifically interesting because they are recorded in diffraction-space, with a near atomic-sized probe, effectively generating diffraction from a single atom. Applications to the imaging and control of single dopant atoms will be discussed.

Work supported in part by US Department of Energy, Office of Science, Basic Energy Sciences, Materials Sciences and Engineering Division and in part by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U. S. Department of Energy.

The family of two-dimensional (2D) materials and their applications continue to expand as novel layered materials are synthesized and used in various devices. One common application is in field effect transistors (FETS), where the 2D material serves as a channel between the metallic source and drain contacts. The interface between the metal and the 2D material is critical to the device performance, but its local atomic and electronic structure remains poorly understood. Theoretical studies using atomistic simulations have shown that the embedded 2D material interacts with the metal contacts and alters its atomic structure. However, few, if any, experimental studies have supported or expanded on these theoretical claims, likely because the device-embedded 2D material is difficult to access and investigate experimentally at the necessary atomic scale. At this scale, transmission electron microscopy (TEM) may be the only viable option, but TEM investigations of devices embedded with 2D materials still remains scarce due to the susceptibility of 2D materials to electron irradiation damage.

Here, we present an experimental study that shows direct observation of bonding interactions between the metallic contact and the 2D material channel within a device using scanning TEM (STEM) in conjunction with electron energy loss spectroscopy (EELS). Using FETs with Ti contacts and MoS₂ channels serving as the representative system, atomic-resolution STEM images of the Ti-MoS₂ interface show distinct differences in the crystal structure between interacting and non-interacting areas within the FET. Furthermore, EELS measurements collected layer-by-layer shows a clear and systematic change in the local density of electronic states (DOS) at each layer of MoS₂ as a function of distance from the Ti contact. These experimental results together with theoretical predictions show quantitatively not only how the metallic contact affects the innate structure of 2D materials but also how many layers the 2D material must be to preserve its expected properties, both of which has significant implications for device performance.

The surface structures of metal oxides play important roles in a wide variety of phenomena such as catalytic reaction and epitaxial film growth, and have been extensively investigated in the past decades. Among the wide variety of metal oxide materials, SrTiO₃ (STO) has an ABO₃ type perovskite structure that is considered as one of the model in oxide materials. The free surface of STO is well known to form an atomically flat, step-terrace structure when annealed at high temperatures, and therefore is widely used as a substrate for thin film growths. Since the properties of the grown films strongly depend on the quality of the substrates, the control of surface atomic structures of the substrates is of great importance to obtain high quality films. To date, the surface atomic structures of annealed STO have been mainly observed by STM (Scanning Tunneling Microscopy). While STM enables imaging of atomic surface structures, its application is restricted to conducting materials, and moreover it is still challenging to identify atom species. On the other hand, ADF STEM (annular dark-field scanning transmission electron microscopy) is a versatile technique that provides atomic scale, element identifiable image of the bulk structure. In plane-view observations, STEM has not been employed to investigate surface structures due to the transmitting nature. However, it is expected that surface structures may be determined by taking a series of defocused ADF STEM images. In this study, we have investigated the (001) surface step structure of STO annealed in ambient air, using a defocus series of ADF STEM images.
The annealed STO single crystal was observed by both TEM and STEM. We confirmed that a step-terrace structure was formed by annealing at 1323 K, with steps faceted in either by {100} or {110}. In order to investigate the three-dimensional atomic structure of the faceted surface, we acquired a defocus series of ADF STEM images at exactly the same region. For each atomic column in the image, the standard deviation of the intensity was plotted as a function of focus depth. Then, column-by-column surface position was determined by the focus depth value with the largest standard deviation of the intensity, which was obtained by fitting the plotted standard deviation with a bell-shaped function. The annealed surface was consisted of three major atomic steps with the heights of 5 to 10 nm, and all of them were faceted in atomically flat planes oriented in either {100} or {110}.
Acknowledgement: This research was partly supported by Research and Development Initiative for Scientific Innovation of New Generation Batteries (RISING2) project of the New Energy and Industrial Technology Development Organization, Japan.

Graphitic carbon nitrides are layered, polymeric compounds that are widely studied as a semiconducting support for visible-light driven photocatalytic water splitting¹. However, structural disorder in these materials is poorly understood due to the difficulty of performing high spatial resolution characterization through transmission electron microscopy (TEM) on these beam-sensitive materials. Fully-condensed g-C₃N₄ contains hexagonally periodic heptazine (C₆N₇) or triazine (C₃N₃) motifs, bridged by 3-fold coordinated N-atoms, creating a graphite-like structure with regularly-spaced voids. Synthesis involves calcination of N-rich precursors to yield yellow powders of varying paleness. In this way, g-CN_xH_y compounds are formed with a sub-stoichiometric C/N ratio at the expense of residual hydrogen preventing complete structural condensation. In 2007, nuclear magnetic resonance and electron diffraction were combined to solve the average in-plane structure of a g-CN_xH_y compound which consisted of amine (N-H_x) -bridged heptazine molecules arranged in zig-zag chains³. Based on x-ray and neutron diffraction, the bulk stacking order was determined giving a 3D structural description⁴. While these works set the foundation for understanding the structure of g-CN_xH_y’s, low-dose TEM may be applied to directly image variations of the in-plane structure and obtain a more precise description of structural disorder.

Using an aberration-corrected TEM operating at 300-kV under low dose rate conditions (e.g. ~20 e^-/Ų/s) coupled with a direct electron detector, we have acquired images from three g-CN_xH_y specimens with different degrees of presumed structural disorder. So far, several techniques are being considered to elucidate (in-plane) disorder based on TEM images taken with the basal plane perpendicular to the beam. Firstly, rotationally-averaged Fourier transforms (FT’s) of large areas (e.g. (50 nm)²) show the improvement in the in-plane scattering information when compared to powder XRD and maintain a similar trend in diffraction broadening. Scanned FT’s could be used to determine local variations in layer buckling, chain separation, and azimuthal orientation of the heptazine building blocks. Autocorrelations have been applied to the TEM images to generate pseudo Patterson functions and radial distribution functions, which may be used to compare the real-space extent of structural correlations. Correlographs, azimuthal auto-correlations performed on an FT, are another tool being used to evaluate symmetry axes and are particularly useful in FT’s containing speckled rings.

[1] X. Wang et al. Nat. Mater. 2009, 8, 76-80. [2] B.V. Lotsch et al. Chem. Eur. J. 2007, 13, 4969-80. [3] F. Fina et al. Chem. Mater. 2015, 27, 2612-18. [4] We gratefully acknowledge the support of DOE grant DE-SC0004954, ASU’s John M. Cowley Center for High Resolution Electron Microscopy, ASU's Center for Solid State Science and Gatan Inc. for the use of a K2-IS direct electron detector.

Li[Mn₂]O₄ (LMO) is a well-known cathode material for Li-ion batteries, but it is plagued with cyclability problems associated with the surface disproportionation of Mn (2Mn³⁺ → Mn²⁺ + Mn⁴⁺) and consequent loss of Mn²⁺ to the organic liquid electrolyte during electrochemical cycling.

In this paper, we use a combination of high-angle annular dark-field (HAADF) aberration-corrected scanning transmission electron microscopy (STEM) and electron energy loss spectroscopy (EELS) to identify the atomic surface structure and composition of LMO. We confirm the underlying spinel structure and for the first time we find, in as-processed LMO, a surface structure composed of Mn₃O₄ and a lithium-rich Li_1+xMn₂O₄ subsurface layer which occurs as a result of the surface reconstruction.

In addition, we have applied an aqueous acid treatment, a non-aqueous chemical delithiation, and an oxygen plasma treatment to LMO in order to understand how this surface reconstruction is affected by chemical treatments. We find that Mn₃O₄ is a robust surface phase in the Li_1-x[Mn₂]O₄ system regardless of the chemical treatment and level of lithiation. The surface Mn₃O₄ phase is cubic whereas bulk Mn₃O₄ undergoes a cooperative Jahn-Teller distortion to tetragonal symmetry. Thicker Mn₃O₄ surface layers are tetragonal.

Mesoporous silica nanoparticles have been intensively studied due to their potential use in catalysis and biomedicine, including cancer treatment and drug delivery applications. However, when mesoporous particles are produced at the nanoscale, the arrangement of pores is modified, hindering the characterization of the porous structure. In order to determine their morphology and structure, advanced microscopy techniques are usually required. However, the complexity of the pore structure makes the characterization very challenging, in particular, an accurate representation in 3D.

This work combines molecular dynamics techniques and electron microscopy computer simulations with experimental results to provide an insight into the structure of amorphous SiO₂ NPs. The amorphous silica model is prepared using a simple melt-quench molecular dynamics (MD) method, while the reconstruction of the mesoporous nanoparticle is carried out using an isotropic unit cell to avoid false symmetry in the final model. For the high-resolution STEM simulations, the QSTEM software package is used based on a multislice technique. Finally, for comparison with the simulated images high angular annular dark field (HAADF) STEM images were taken using an aberration-corrected FEI Titan ChemiSTEM microscope, operated at 200 kV.

The amorphous models are analyzed using the radial distribution function (RDF) and mass density, demonstrating a good agreement with the experimental results. Depending on the quenching model, the local density can be modified obtaining isotropic values between 2.2 and 4 g cm^-3, with radial distribution function similar to the bulk values reported experimentally. The highest probability of finding Si-O pairs is at 1.58 Å, O-O is at 2.62 Å and Si-Si 3.08 Å. The multislice STEM images demonstrate that the density of the models does not have a significant impact on the STEM images for isolated SiO₂ phases. However, a detailed analysis reveals that the intensities of the systems show that denser SiO₂ structures result in a more intense signal, due to the increase in the scattering power.
When constructing the nanoporous particles, an isotropic unit cell of the previously mentioned amorphous SiO2 model was utilized. In order to avoid false symmetry in the STEM images, the unit cell was randomly rotated finding good agreement with the experimental images obtained. The results show the possibility of accurately modeling amorphous SiO₂ amorphous phases, and opens the possibility of simulating nanoparticles when functionalized for catalysis and biomedical applications.

Proton Exchange Membrane Fuel Cells (PEMFCs) have gained increasing attention due to their high energy/power density, high efficiency. The sluggish oxygen reduction reaction (ORR), occurred in the catalyst layer in the membrane electrode assembly (MEA), is the major bottleneck for the FEMFC performance. The ORR is strongly depends on the nanostructure of the catalyst layer, in particular, the ionomer/catalysts/carbon interface in the catalyst layer. Thus, it is imperative to determine the nanostructure of the interface in terms of the ionomer coverage, thickness of the ionomer film over the catalyst nanoparticles.
In this context, Aberration Corrected-Scanning Transmission Electron Microscope (ac-STEM) coupled with Electron Energy Loss Spectroscopy (EELS) are used to study the nanostructure of the interface. Most of the previous research was focused on observing the presence of fluorine, which is a distinct element in the perfluorosulfonated ionomer. However, the use of fluorine as the marker may not be as accurate as one expected as there is a significant F loss under electron beam [1].
In this work, EELS analysis of Vulcan XC72 (carbon) and nafion (ionomer) is employed to distinguish between these two kinds of amorphous carbon.. A single spectrum of both Vulcan XC72 and Nafion, with an energy resolution of 0.3 eV, is obtained in ac-STEM coupled with a monochromator. Decomposition of the carbon K-edge in Nafion and Vulcan XC72 showed that the ratio of intensities between σ edge (at 287ev) and π edge (295ev) is significantly different, as well as the energy loss near edge structure (ELNES) of the σ edge. Moreover, an extra peak at 292ev is shown in the spectrum of Nafion, which is primarily attributed to the carbonyl group in Nafion .
To prepare electron transparent MEAs, a partial embedded method on ultramicrotome is applied to avoid resin penetration. Spectrum images on the catalyst layers are obtained under the same condition for obtaining single spectrums for Nafion and Vulcan XC72. Principal component analysis (PCA) and multiple linear least square (MLLS) fitting is then applied to electrode spectrum images. The Nafion and carbon particle distribution are clearly observed and distinguished. Meanwhile, the spectrum images of the catalyst layer from the fluorine K-edge is also obtained, which shows good agreement with the results from the carbon K-edge but with a lower signal-to-noise ratio.
In summary, STEM-EELS characterization of the carbon K-edge provides a novel method to determine the interplay between the ionomer/carbon/catalysts distribution on the catalyst layers of PEMFCs. As the cross section for carbon K-edge is normally higher than F-edge in EELS, this new method also provides insight into detecting low perfluorosulfonated ionomer contents in the PEMFCs.

Reference:
[1] D A Cullen, R Koestner, R S Kukreja, Z Y Liu, S Minko, O Trotsenko, A Tokarev, L Guetaz, H M Meyer III, C M Parish, K L More, J. Electrochem Soc 161, F1111 (2014)

The optimal functionalities of materials often appear at phase transitions involving simultaneous changes in the electronic structure and the symmetry of the underlying lattice. It is experimentally challenging to disentangle which of the two effects––electronic or structural––is the driving force for the phase transition and to use the mechanism to control material properties. In this talk, we present a concurrent pumping and probing of Cu₂S nanoplates using an electron beam to directly manipulate the transition between two phases with distinctly different crystal symmetries and charge-carrier concentrations, and show that the transition is the result of charge generation for one phase and charge depletion for the other. We demonstrate that this manipulation is fully reversible and nonthermal in nature. Our observations reveal a phase-transition pathway in materials, where electron-induced changes in the electronic structure can lead to a macroscopic reconstruction of the crystal structure. This control method is in contrast to conventional chemical doping, which is irreversible and often introduces unwanted lattice distortions [1].

On the other hand, Cu₂S material recently has attracted considerable attention for its ionic and thermoelectric properties. Unveiling the structural origin of this material's functionality is of great interest. Taking advantage of in situ electron diffraction and aberration-corrected imaging techniques, we further reveal the structural characteristics of Cu₂S, including the Cu arrangement in the S framework through different thermal processes and the strain mapping at the phase boundaries during phase transitions.

[1] J. Tao et al., PNAS 114, 9832 (2017)

Aberration-corrected scanning transmission electron microscopy (STEM) has proved indispensable for understanding the atomic structures of nanomaterials and heterogeneous catalysts [1]. Nanoscale architectures of individual nanocomponents can provide synergistic and unique physicochemical properties that would not be possible with the individual components. We have previously developed a robust process to produce well faceted ZnO nanowires (NWs) as support materials for nanostructured catalysts or solar cell applications [2]. The ultraclean ZnO NWs primarily consist of ZnO {10-10} and {11-20} nanoscale facets. We have reported that Bi₂O₃ selectively deposit on the {11-20} nanofacets of the ZnO NWs with an epitaxial relationship [2]. We recently discovered individual Ir atoms self-assemble into single-atom size chains onto the {10-10} nanoscale facets of the ZnO NWs. Such self-assembly of Ir atom-size chains occurs under calcination treatment at selected temperatures. Aberration-corrected STEM-HAADF (high-angle annular dark-field) imaging technique was extensively used to help optimizing the synthesis processes, determine the atomic structure of the Ir/ZnO system, and to investigate the stability of the Ir atom chains under various gas treatment. By tiling the ZnO NWs both profile and plan view STEM-HAADF images were obtained which help determine the 3D structural relationship of the Ir atom chains with respect to the surface structure of the ZnO NWs. The catalytic properties of the newly discovered Ir/ZnO system were also evaluated for selected catalytic reactions [3].


[1] J Liu, ChemCatChem 3 (2011), p. 934.
[2] J Xu and J Liu, Chemistry of Materials 28 (2016), p. 8141.
[3]This work was funded by NSF under CHE-1465057. We acknowledge the use of facilities in the John M. Cowley Center for High Resolution Electron Microscopy at Arizona State University.

Transmission electron microscopes primarily employ indirect cameras (IDC) for electron detection in imaging, diffraction and EELS modes. Such cameras convert incident electrons to photons which, through a fiber optic network or lens, are coupled to a light sensitive camera. This indirect detection method typically has a negative impact on the point spread function (PSF) and detective quantum efficiency (DQE) of the camera. Over the last decade, radiation tolerant CMOS active pixel sensors, which directly detect high-energy incident electrons and have the speed to count individual electrons events, have been developed. These detectors result in greatly improved PSF and DQE in comparison to conventional IDCs. Such direct detection cameras (DDCs) have revolutionized the cryo-TEM field as well as have strong advantages for in-situ TEM in both imaging and diffraction applications. EELS applications can benefit from the improved PSF and the ability to count electrons. The improved PSF allows spectra to be acquired over larger energy ranges while maintaining sharp features and greatly reduced spectral tails. The ability to count electrons nearly eliminates the noise associated with detector readout and greatly reduces the proportional noise associated with detector gain variations. This effectively leaves the shot noise as the limiting noise source present. The implication for EELS acquisition is that fine structure analysis becomes more straightforward for typical conditions and even possible for the case of low signal levels.
Very high-energy edges above 3000eV have always been very hard or almost impossible to acquire using EELS due to the very limited amount of signal. With the introduction of DD detectors the amount of noise has been enormously reduced and as result low intensity signals can now be observed and detected. EELS spectra of Cu K and Ni K-edges at about 9keV and 8.3keV can now be collected and easily observed and the quality is such that high contrast elemental maps can be generated. Until now, such high energy edges have been collected using synchrotron based techniques such as XAS with very limited spatial resolution. Now, by acquiring EELS data in counting mode using DD detectors, high energy edges can be collected and their signal mapped out with high spatial resolution. A new world is about to open up.
In this presentation, we will review the current state of electrons counting detectors for electron microscopy with an emphasis on system for EELS measurements.

Properties of low-dimensional materials are largely influenced by its structural imperfections, such as defects, impurities, edges or boundaries. Hence, analytical technique at the single-atom level is becoming crucial to fully understand their physical/chemical performance. In my presentation, single atom spectroscopy by means of electron energy-loss spectroscopy (EELS) will be shown to discriminate individual atoms in low-dimensional materials at their interrupted periodicities. It is emphasized here that information of the bonding/electronic states has become accessible for single atoms through the EELS fine-structure analysis [1] as well as the spin state [2, 3]. Large variations of local electronic properties of 1D and 2D materials with different atomic coordinates will be investigated [4, 5].
I will also show some examples for the optical-range electron-spectroscopy of low-dimensional materials [6, 7, 8]. Differences between EELS and optical absorption will be discussed for surface materials in diluted systems [9].


References
[1] YC. Lin et al., Nano Lett., 15 7408-7413 (2015).
[2] YC. Lin et al., Phys. Rev. Lett., 115 206803 (2015).
[3] G. Liu et al., Nature Chem. 9 810-816 (2017).
[4] YC. Lin et al., Nano Lett., 17 494-500 (2017).
[5] HP. Komsa et al., Nano Lett., 17 3694-3700 (2017).
[6] R. Senga et al., Nano Lett., 16 3661-3667 (2016).
[7] J. Lin et al., Nano Lett., 16 7198-7202 (2016).
[8] L. Tizei et al., Phys. Rev. Lett., 114 107601 (2015).
[9] R. Senga et al., (unpublished)
[9] JSPS KAKENHI is acknowledged for financial support.

The atomic structure of materials can be characterized by transmission electron microscopy (TEM) and scanning TEM (STEM). However, determining the positions of atoms in three dimensions from two-dimensional images is non-trivial. In this work, we use atomistic modeling with first principles density functional theory (DFT) or empirical potentials, in conjunction with machine learning, to tackle the S/TEM image inversion problem. We discuss the use of single and multi-objective evolutionary and basin-hopping approaches for S/TEM-guided atomistic structure determination, incorporating comparison of simulated and experimental S/TEM images using computer vision approaches. We show that the combined use of energetic and experimental information is effective in arriving at physical solutions.

Convergent beam electron diffraction (CBED) patterns acquired with scanning transmission electron microscopy (STEM) contain abundant specimen information that is inaccessible in conventional STEM with annular detectors. However, use of this “4D-STEM” technique has been hindered by two primary challenges: (1) Pixel array detectors are much slower than conventional annular detectors, which limits the specimen field-of-view of 4D-STEM experiments, and (2) Data sets from 4D-STEM experiments are large and unwieldy, which makes interpretation of results difficult. To address the first challenge of detector speed, we have customized a high-speed direct detection camera for synchronized 4D-STEM acquisition at several kilohertz, which is faster than other currently-available pixel area detectors. This increased speed enables collection of larger specimen areas without unacceptable specimen drift. Then, to address the second challenge, we have developed an open-source Python program for robust analysis of large 4D-STEM datasets. As a proof-of-concept, this combination of hardware and software have been used to analyze the structure of several materials specimens in an aberration-corrected electron microscope.

The graphene growth and its interaction with the metal support are interesting both from a fundamental and technological point of view. In particular, graphene on ferromagnets has the potential to be used for the realization of spintronic devices, acting as a conductive channel, into which spin-polarized electrons are injected from the ferromagnet support. For this reason, the study of the growth and the electronic structure of the graphene-metal interface is the prerequisite for any practical application.
Here, we follow the temperature-dependent growth of graphene on Co/W(110) by means of low energy electron and photoemission electron microscopy (LEEM/PEEM). LEEM studies show that a highly ordered graphene layer with single azimuthal orientation can be only grown within a certain temperature range (>500°C). At lower temperatures, graphene grows predominantly with misoriented domains. This was confirmed using different techniques: LEED shows a (1x1) structure with an additional blurred ring indicating graphene domains rotated relative to the Co lattice. In addition, the C 1s XPS spectra indicate the presence of defects in the graphene network along with carbide components at lower binding energy.
At higher temperatures, the low quality graphene transforms into a (1x1) epitaxial graphene with better structural quality, as monitored by LEEM and μ-LEED. In this case, the diffraction ring due to azimuthally misoriented domains disappears and the LEED pattern is composed only by a (1x1) structure in registry with the Co substrate, indicating the presence of an epitaxial graphene overlayer. Moreover, the C 1s XPS spectrum presents a sharp single-component feature centered at 284.8 eV binding energy, characteristic in graphene-cobalt complexes.

Determination of bandgaps and optical properties using electron energy loss spectroscopy (EELS) has attracted interest since monochromated transmission electron microscopes (TEM) with excellent spatial resolution and an energy resolution that is in the tens of meV range have become available. [1].
However, for small bandgap materials (like Si, GaAs) Čerenkov losses changes the valence EEL spectrum because of the large refractive index [1]. The phase velocity of light in the sample material scales with the refractive index, and beam electrons which move faster than this phase velocity of light will suffer. The occurrence of Čerenkov losses causes unwanted spectral artifacts which affect the interpretation of low-loss EELS and complicates precise measurements of bandgap values. In addition to Čerenkov losses, contributions from retardation and surface effects, and the excitation of guided light modes can complicate the analysis of valence EELS data. For determination of optical properties, one has to ensure that no relativistic effects impact the low-loss signal [2].

In our previous work [3], we presented an experimental set-up that allows bandgaps of high refractive index materials to be determined. In this method, semi-convergence and -collection angles in the micro-radian range were combined with off-axis or dark field EELS to avoid relativistic losses and guided light modes in the low loss range to contribute to the acquired EEL spectra. Off-axis EELS further suppressed the zero loss peak. The bandgap of several GaAS-based materials were successfully determined by simple regression using this method.

In this work we have continued band gap measurements of GaAs-based materials to improve our methodology and to transfer it to slightly different compositions. Materials used in this work are all potential candidate structures for use in intermediate band solar cell applications. On the top of a (001) GaAs substrate, one of the samples has a periodic stack of thin In_x Ga_x-1As quantum well layers grown by molecular beam epitaxy. The spacer layers in between consist of 50 nm GaAs. In the other sample, on top of the substrate a stack of Al_xGa_1-xAs layers with 30 nm thick layers are grown. A double corrected Titan TEM with monochromator was used, operating at 60 kV in all experiments.

References

[1] M Horak and M.S Pollach, Ultramicroscopy, 157(2015) 73-78
[2] R Erni and N D. Browning, Ultramicroscopy, 108 (2008) 84-99
[3] M. Vatanparast, R. Egoavil, T.W. Reenaas, J. Verbeek, R. Holmestad and P.E. Vullum, Ultramicroscopy, 182 (2017) 92-98

Acknowledgments
The Norwegian Research Council is acknowledged for funding the HighQ-IB project under contract no. 10415201