Symposium CP03—Advances in in situ Techniques for Diagnostics and Synthetic Design of Energy Materials

In situ transmission electron microscopy has become a powerful technique to study functional nanomaterial dynamics in response to electrical, chemical, electrochemical, thermal, and/or electromagnetic fields prevalent under operating conditions. The time resolution, however, is typically limited by the maximum frame rate of the detector, which is at best in the kHz regime. Ultrafast transmission electron microscopy combines the high time resolution of laser spectroscopy with the excellent spatial resolution of electron microscopy techniques. The structural and electronic changes in the material are initiated by short (fs, ps, ns) laser pulses, which are followed by similarly short photoelectron pulses for probing the dynamics by means of imaging, diffraction, or energy-loss spectroscopy (EELS) within the electron microscope. In this talk I will present our dynamic environment transmission electron microscope (DETEM) at the University of Illinois. This versatile instrument enables time-resolved in-situ studies of fast (fs-ms) material processes in gaseous and liquid environments that are triggered by light, heat, chemical flux, electrical bias or strain. I will give examples of studies using light excitation as the trigger, including nanosecond-resolved single-nanoparticle electron diffraction and real-space imaging of spin crossover phase transitions and ultrafast (fs,ns) EELS of light-induced dynamics in graphite thin films.

Nanoscale and atomic scale interactions between light and matter in water govern a number of important processes. One example is photocorrosion, which negatively impacts many areas including energy conversion technologies and structural materials. The process takes place in aqueous environments and, for semiconductors, is driven by electron-hole pair generation from photon illumination^1-2. For example, cadmium sulphide (CdS) based photocatalysts, undergo severe photocorrosion during photocatalytic water splitting³. Understanding the nano- and atomic scale changes induced by photocorrosion may offer new insights into fundamental questions regarding the initiation and evolution of photocorrosion and the impact of factors such as the presence of solute ions.

To probe photocorrosion and other light/matter interactions in solution, we have developed an open cell method for condensing liquid water around nanoparticles and illuminating them inside an environmental transmission electron microscope (ETEM).
To facilitate liquid water formation, nanoparticles are dispersed on a TEM sample grid alongside ball milled sodium chloride (NaCl), which are strongly hydrophillic⁴. Samples are loaded into the ETEM in a cryogenic holder held at ~5^oC, and are then exposed to 4-10 Torr of water vapor in the ETEM column, causing water droplets to form and dissolve the NaCl, and enveloping neighbouring nanoparticle clusters in NaCl solution. To illuminate particles immersed in solution, we have modified an earlier in situ light system⁵ to use a 405 nm laser connected to the TEM with an optical fibre inserted through the objective aperture port of the microscope.

Aberration corrected TEM imaging before and after exposure to light and water will allow us to characterize changes that occur in clusters of particles, We apply this technique to CdS photocatalyst nanoparticles in a variety of light and solute conditions.


References:
1. Gerischer, H.J. Electroanal. Chem. & Interfacial Electrochem. 82, 133-143 (1977).
2. Chen, S.Y. & Wang, L.W. Chem. Mater. 24, 3659-3666 (2012).
3. Meissner, D. et al. Chem. Phys. Lett. 127, 419-423 (1986).
4. Wise, M.E. et. al. Aerosol Science and Technology, 42, 281–294 (2008).
5. Liu, Q. et. al. Microsc. Microanal. 22(S3), 730-731 (2016).
We gratefully acknowledge support from the U.S. DOE (DE-SC0004954) and ASU’s John M. Cowley Center for High Resolution Electron Microscopy.

Energy dispersive X-ray spectroscopy (EDS) experiments at elevated temperatures up to the range of 1000°C, will be used to discuss challenges of in situ spectroscopy and element mapping. Heat radiation at elevated temperatures changes the low energy part of EDS spectra. Still, low energy element ID and chemical phase analysis are available in STEM and SEM within certain limits. The agglomeration and evaporation of material in a Pd-Au-nanoparticle test structure during heating in a conventional STEM were evaluated using statistical methods and phase analysis for. The results demonstrate the capabilities of in-situ EDS. The heating holder and its stability at changing temperatures plays a crucial role for successful experiments [1].

Further options of using EDS for in situ analysis are explored. One certain request is speed. Optimum photon detection geometry is crucial for efficient EDS in the electron microscope. Important to achieve that are a large solid and high take-off angle for photon collection, good collimation and suitable sample holders. The holders should not shadow the detector(s). Multi-detector arrangements enable large solid angles at high take-off angle particularly easily. Geometric optimization in combination with high brightness electron sources and aberration correction allows the identification and tracking of single atoms at low accelerating voltages in STEM within a few seconds [2]. Specimen holder materials and window materials need to be well known to ensure correct quantitative analysis of spectroscopic results, which is of particular interest in case of reaction cells holding liquid samples. Additionally, dealing with constantly changing data streams demands new ways of data acquisition and processing, which are being developed.

[1] T. T. van Omme et al., Ultramicroscopy 129 (2018) 14
[2] R. M. Stroud et al. Appl. Phys. Lett. 108 (2016) 163101

Melting and crystallization are fundamental and practically important first-order phase transitions in condensed-matter physics, material science and climate change, yet a detailed understanding of their relevant kinetic pathways is still evolving [1,2]. To date, many theoretical models have been developed from homogeneous classical nucleation theory (CNT) model, but they rarely address the exact preferential nucleating sites and the potentially relevant role played by defects, surfaces, dimensionality and their combinations in phase transformations [3]. Recently, Samanta et al. conducted large-scale atomistic calculation of a phase transformation process of a metal from solid to liquid and predicted that the process takes place via multiple competing pathways involving the formation and migration of point defects or dislocations [4]. Although these calculations indeed provide a rare look at real phase transformations, much confusion still exists regarding the atomistic understanding of a dynamic process of a phase transformation due to the lack of direct experimental observations at the atomic scale as well as due to the experimental intricacies in tackling such a challenging topic. Bismuth (Bi) is an ideal inorganic model material suitable for gaining insights into nucleation dynamics due to its low melting point (even down to room temperature due to size effect).
Here, we report in-situ atomic-scale observations of a real dynamic process of melting or crystallization in supported Bi nanoparticles under heating or cooling conditions using a heating holder within an aberration-corrected (scanning) transmission electron microscope. We provide direct evidence that pre-nucleation in either melting or crystallization takes places via multiple intermediate state pathways involving the formation and migration of domain boundaries, dislocations and the ordering of interface and surface at the atomic scale. The pre-melting of the nanoparticles initiates at a grain boundary and expands to interfaces and dislocations and finally undergoes a catastrophic transformation from a solid-liquid structure to a liquid droplet as a whole in a rather short time when their size exceeds a threshold value. When the size is smaller than the threshold value, the melting of nanocrystals takes place via two barrier-crossing pathways, i.e. pre-melting at grain boundary and a catastrophic solid-liquid transformation. Interestingly, pre-crystallization in a droplet occurs first at a solid-liquid interface and subsequently at the liquid surface, and eventually the droplet undergoes a fast complete transformation to a solid nanocrystal when undercooled. Thus the ability to conduct in-situ atomic-scale observation of the evolutional pathways in phase transformations of supported nanoparticles represents a significant step forward in understanding microscopic mechanisms of phase transitions at the atomic scale. The findings in the present study demonstrate that the melting/crystallization processes cannot be viewed as a simple single barrier-crossing event but as a complex multiple intermediate state phenomenon, which enhances our general understanding of nucleation and growth, melting/crystallization phenomena, phase transformations and helps to clarify atomic origins of temperature dependent behaviours in other nanomaterials and thin films [5].
References
[1] AM Alsayed et al, Science 309 (2005) p. 1207.
[2] H Wang et al, Nat. Commun. 6 (2015) p. 6412.
[3] A Tabazadeh, YS Djikaev and H Reiss, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) p. 15873.
[4] A Samanta, ME Tuckerman, TQ Yu, E Weinan, Science 346 (2014) p. 729.
[5] J Li, Z Wang and FL Deepak, J. Phys. Chem. Lett., 9 (2018) p. 961.

In the understanding of durability and longevity of emerging photovoltaic materials, an essential component is the study of structural stability in response to photoexcitation. The study of photoinduced structural dynamics in inorganic semiconductors, such as silicon and gallium arsenide, has been established over recent years [1,2], thus enabling insights into thermal- and carrier-transport properties, especially pertaining to the effects of morphology and defects. Ultrafast electron microscopy (UEM), a pump-probe technique that utilizes a pulsed photoelectron beam generated by a femtosecond (fs) pulsed laser to extend the conventional capabilities of transmission electron microscopy (TEM) into the ultrafast (fs to picosecond) time domain [3], is one technique by which such studies can be conducted. Indeed, accessing the picosecond-nanometer spatiotemporal parameter space in the study of radiation-sensitive organic materials would doubtless allow for the discovery and the elucidation of light-induced excitation and decay mechanisms, especially with respect to the impact of defects on carrier-scattering dynamics and macroscopic transport [4]. However, to properly interpret the results of such measurements, it is vital to identify and quantify the effects of the pulsed electron beam on the intrinsic material structure and morphology, as the segmented nature of the dose has the potential to alter the dominant damage mechanisms relative to those occurring during conventional TEM. Importantly, the pulsed beam can be precisely controlled in terms of the number of electron pulses per second (EP/s) via the instrument repetition rate and the number of electrons per pulse (E/P) via the laser fluence, thus enabling meticulous management of the delivery of electron dose. Indeed, aside from uncovering any anomalous effects of the pulsed beam, such an approach has the potential to empower high-resolution studies of nanoscale architectures and defects in materials that are currently challenging to study with prolonged exposures, in addition to the aforementioned facilitation of illumination of ultrafast behaviors.
To address this, we are exploring the effects of pulsed electron-beam properties on the beam-induced damage of organic crystalline lattices. Here, we have used the waxy paraffin hexatriacontane (C₃₆H₇₄) as a test case owing to its well-defined crystal structure and the large amount of previous work that has been done on determining the intrinsic damage thresholds and the various behaviors occurring under electron-beam illumination. Following convention, we quantify the degree of damage by monitoring the attenuation of Bragg-reflection intensities arising from the beam-induced loss of crystallinity. Because of the potential for large variability in such measurements, experiments are routinely repeated as many times as is practical (i.e., a minimum of n = 5 for most experiments), damage is quantified via rigorous and robust peak-fitting methods, and statistical analyses (e.g., one-way analysis of variance) are employed to check for significance of the results. Two types of experiments are undertaken: (1) pulsed beams are compared to stochastic thermionic beams, and (2) pulsed-beam parameters (EP/s and E/P) are systematically varied to uncover existing trends. We discuss both the fundamental mechanisms of damage pertinent to certain regimes within the parameter space – for example, a relaxation of crystal bending that only occurs at high dose rates regardless of the total dose – and the practical conclusions regarding best methods to mitigate damage in high-resolution study of these types of materials.

[1] J. Cho, et al., Proc. Natl. Acad. Sci. U.S.A. 111 (2014) 2094.
[2] E. Najafi, et al., Science 347 (2015) 164.
[3] D. A. Plemmons, et al., Chem. Mater. 27 (2015) 3178.
[4] X. Wu, et al., Sci. Adv. 3 (2017) e1602388.

Aqueous interfaces are ubiquitous in technological applications and in the environment and govern numerous important phenomena, including corrosion, photoelectrochemical splitting of water, weathering of rocks, cloud droplet nucleation, the concentration of trace gases in polar regions, and thundercloud electrification. [1,2] A detailed understanding of these processes requires the investigation of aqueous interfaces with chemical sensitivity and interface specificity under ambient conditions, i.e., temperatures above 200 K and water vapour pressures in the milliTorr to tens of Torr pressure range. Ambient pressure X-ray photoelectron spectroscopy [3,4] is an excellent method to probe the heterogeneous chemistry of these interfaces and has been applied to a number of different aqueous interfaces, including ultrathin water films on metals and oxides, bulk liquid/solid interfaces, the surface of aqueous solutions in the presence of the equilibrium vapour pressure as well as trace gases, and the surface of ice, which at temperatures close to the melting point is covered by a thin liquid-like layer (see Figure, which is adapted from Ref. [1]). The challenge for XPS experiments on aqueous interfaces under ambient conditions is not only the elevated pressure at which these experiments have to be carried out, but equally so the preparation of clean, well controlled aqueous samples with high cleanliness and reproducibility. In this talk I will review the progress that has been made in the area of APXPS experiments on aqueous systems and provide an outlook on opportunities for future research in this field.

[1] O. Björneholm, M.H. Hansen, A. Hodgson, L.-M. Liu, D.T. Limmer, A. Michaelides, Ph. Pedevilla, J. Rossmeisl, H. Shen, G. Tocci, E. Tyrode, M.-M. Walz, J. Werner, H. Bluhm, Chem. Rev. (2016).
[2] A. Verdaguer, G.M. Sacha, H. Bluhm, M. Salmeron, Chem. Rev. 106, 1478 (2006).
[3] H. Siegbahn, J. Phys. Chem. 89, 897 (1985).
[4] D.E. Starr, Z. Liu, M. Hävecker, A. Knop-Gericke, H. Bluhm, Chem. Soc. Rev. 42, 5833 (2013).

Fast XPS is a popular method for observing reactions and transitions at surfaces. The combination of chemical analysis and surface sensitivity with fast and time resolved data acquisition offers insight into the process itself where “static” measurements merely show its outcome.
Applications can be found in many fields of research where the nature of the process is studied, e.g. on-surface synthesis of polymers and 2d materials, adsorption mechanisms during solid-gas interactions or thermally activated reactions at surfaces.
This method is mostly applied at synchrotron beamlines where short acquisition times for time resolution are compensated by high photon flux for signal strength.
In this talk enhancement of this measurement technique due to recent advances in instrumental development is presented and demonstrated with applications. Not only does it increase the spectroscopic information to multiple, wide range peak areas, but opens up the possibility for off-line and laboratory based experiments with conventional X-ray sources on time scales down to the millisecond range. This way, parameters for controlled and reproducible processes or interactions can be found and optimised, and the dynamics of these processes systematically observed.
Furthermore, specific properties important e.g. in device fabrication can be engineered with fast and dynamic XPS using its time resolved aspect.

Recent developments in X-ray Photoelectron Spectroscopy have enabled studies, in which the surface chemistry of catalytically active materials is probed under quasi in situ conditions. However, the insight into the mechanisms of catalytic reactions gained from such studies is entirely dependent on our ability to interpret the recorded spectra. Often this can be extremely challenging, for example when reference data that is typically used to guide "peak-assignment" is either unavailable or conflicting.

First principles calculations of core electron binding energies can guide the interpretation of experimental results. We have developed a computational framework for the calculation of core electron binding energies of surface species that combines the use of slab and cluster models of the surface. In the first step, the geometry of the adsorbate is optimized using a periodic slab model of the surface. Then, a cluster is cut from the slab, and the core electron binding energy is calculated for that cluster using the all-electron delta-SCF method. We have carefully examined the convergence of the results with respect to the size of the cluster used to model the surface, and found that finite size effects can be effectively reduced to the point where they are smaller than the functional error.

Using this method, we have calculated C1s nad O1s core electron binding energy shifts for a series of adsorbates on the Cu(111) surface. The results are in good overall agreement with experimental binding energies that have been determined from studies of copper based catalysts. Where there is ambiguity about experimental peak assignments, the theoretical binding energies can help to resolve the issue.

We have also tested the accuracy of the all-electron delta-SCF method by comparing the results to gas phase reference data, and found that the meta-GGA exchange correlation functional SCAN is particularly well suited for core electron binding energy calculations. For the 1s core levels of C-F, and the 2p core levels of Si-Cl, the absolute values of core electron binding energies can be calculated with mean unsigned errors of ~0.1 eV and ~0.25 eV respectively, without the use of any empirical corrections.

The energy barrier for most of the chemical reactions is overcome by heating the reactants in the presence of a suitable catalyst, where the thermal energy provided by heat helps in surpassing the barrier that has been reduced by the adsorption of the reactant molecules on the catalyst nanoparticle surfaces. In recent years, localized surface plasmon (LSP) resonance energy has been used to replace the thermal energy to initiate reaction at low temperatures, even at room temperature.^1,2 As most of the plasmonic materials, Au, Ag or Al, are not generally catalytically active, they have been successfully used by combining them with active catalysts such as Pt, Ni, Ru, etc. Most of the researchers so far have used optical methods to generate and measure the LSP resonance energies. However, the combination of low spatial resolution, of the order of 100 nm, achievable by optical methods, does not provide sub-particle level distribution of coupling efficiency of different modes. On the other hand, high energy electrons not only excite all LSP modes simultaneously, but also provide high spatial resolution in the nanometer range to resolve their energy loss probability that indicates the efficiency of coupling the location-specific electron excitation of the LSP modes at certain resonance energies on the plasmonic nanoparticle. Therefore, we use a monochromated electron source in an environmental scanning-transmission electron microscope (ESTEM), combined with metallic nanoparticle boundary element method (MNPBEM) and density functional theory (DFT) simulations, to characterize various LSP modes and their coupling efficiency distribution within the shape and size-controlled nanoparticles of Au and Al. We demonstrate that the knowledge gained from low-loss and core-loss EELS measurements can be used to design catalyst-plasmonic particle combination for selected chemical reactions.


1 Mukherjee, S. et al. Hot-Electron-Induced Dissociation of H-2 on Gold Nanoparticles Supported on SiO2. Journal of the American Chemical Society 136, 64-67, doi:10.1021/ja411017b (2014).
2 Linic, S., Aslam, U., Boerigter, C. and Morabito, M. . Photochemical transformations on plasmonic metal nanoparticles. Nature materials 14, 567-576 (2015).

Tailoring the chemical reactivity of nanomaterials at the atomic level is one of the most important challenges in catalysis research. In order to achieve this elusive goal, we must first obtain a fundamental understanding of the structural and chemical properties of these complex systems. In addition, the dynamic nature of the nanostructured films and nanoparticle (NP) catalysts and their response to the environment must be taken into consideration, since their working state might not be the state in which the catalyst was prepared, but rather a structural and/or chemical isomer that adapted to the particular reaction conditions.
This talk provides examples of recent advances in the preparation, plasma-functionalization and characterization of Cu- and Ag-based nanostructured films and NP catalysts (CuZn, CuNi, CuCo) with well-defined morphology, size, and shape. I will discuss how to resolve the structure and composition of nm-sized metal catalysts via a combination of in situ and operando microscopy (EC-AFM, STM, TEM) and spectroscopy (XAFS, NAP-XPS) methods, and how to follow its evolution under different gaseous or liquid chemical environments in the course of a catalytic reaction. Examples of catalytic processes which will be discussed include the hydrogenation of CO₂ at high pressure as well as its electrochemical reduction. Emphasis will be given to elucidating the role of the NP size, shape, support, composition and chemical state of the catalysts in the activity and selectivity of the former reactions.

The study of numerous materials systems would benefit from nanoscale observation as the chemical processes are progressing, but attempting to see these dynamics on the necessary time and length scales can be quite challenging. Liquid cell Transmission Electron Microscopy (TEM) provides a window into these processes and can reveal mechanistic insights to aid future materials development. To improve the spatial resolution, we use graphene to encapsulate small pockets of liquid, and this high spatial resolution allows the size, shape, and facets of nanocrystals to be followed during their dynamic processes. Once we were able to understand and control the chemistry of graphene liquid cell TEM pockets, the dynamics of nonequilibrium etching and ligand effects on etching were able to be elucidated.

The electron beam interacts with the water solution in our liquid cell, generating reactive radiolysis products that can be used to promote the dynamics that we are trying to study. For our study of oxidative etching, the liquid cell is preloaded with the nanoparticles of interest and a small concentration of FeCl₃ (25-45 mM). At this concentration of FeCl₃, the rods are not etched until exposure by the electron beam. Highly oxidative beam-induced species such as the hydroxide radical and hydrogen peroxide are generated. The electron beam dose rate controls the amount of oxidative species in solution, and the etching rate varies linearly with the electron beam dose rate. The FeCl₃ concentration determines the relative amount of the oxidative species in solution which controls the oxidative potential of the etching environment. With this control of the liquid cell pockets, oxidative etching reactions can be systematically controlled and studied.

Previous work has shown that nonequilibrium etching of cubes with {100} facets and rhombic dodecahedra (RDD) with {110} facets leads to intermediate tetrahexahedra (THH) shapes with {hk0} facets. Nonequilibrium shapes such as the THH are difficult to synthesize, so understanding the mechanisms of nonequilibrium etching could lead to unprecedented control of high index surface facets. With our developed ability to tune the chemical potential of the liquid cell by modulating the FeCl₃ concentration, we systematically varied the oxidation potential for the etching process and followed the etching trajectory while comparing the results with Monte Carlo simulations. When etching from the {100} faceted cubes, the chemical potential determined the steady {hk0} facet. However, the {110} faceted RDD etched to {hk0} THH with h/k = 2.5 regardless of the chemical potential. Modelling this process on the atomic level, etching 6-coordinated edge atoms on the nanocrystals reveals 7-coordinated inner atoms for cubes and 6-coordinated inner atoms for RDD. The difference in coordination of newly exposed atoms causes the differing chemical potential dependence. This mechanistic insight into nonequilibrium etching gives important directions for synthetic control of high index facets.

Etching nanocrystals in the graphene liquid cell can also provide useful information about the organic ligands on the surface of the nanocrystal. We see that changing the ligands on the surface of the nanocrystal varies the etching rate of the nanocrystal. Using ligands that selectively bind to specific sites or facets on the nanocrystal leads to unique etching trajectories. This technique shows where the ligands are binding on the nanocrystal as well as potential pathways to unique nanocrystal shapes.

This work provides mechanistic insight into the shape transformations of nanocrystals in their native liquid environment. In addition, the development of chemical control in the liquid cell will open up future opportunities for materials researchers who seek to further understand a variety of nanoscale systems.

Atomically-thin two-dimensional (2D) layered materials, especially the semiconducting 2D transition metal dichalcogenides (TMDs) and their heterostructures, exhibit remarkable quantum properties that areenvisioned for energy-efficient photovoltaics, flexible optoelectronics, and quantum information science. However, significant synthesis and processing challenges currently limit their development, including wafer-scale, bottom-up synthesis of uniform layers of crystalline 2D materials that are comparable in quality to exfoliated flakes of bulk materials. As-synthesized crystals of 2D TMDs display remarkable heterogeneity on both the atomistic level (e.g., vacancies, dopants, and edge terminations) and on the mesoscopic length scale (e.g., misoriented grains, layer orientations, and interactions with substrates and adsorbates) that can strongly influence the structure and electronic properties in 2D materials. While this heterogeneity offers a serious challenge for synthesis and processing, it also offers a tremendous opportunity to tailor functionality. In situ spectroscopic diagnostics during the synthesis and processing of 2D materials offers a rare chance to ‘close the loop’ between synthesis and properties to tailor optoelectronic functionality with real-time feedback.

Here we present principally non-equilibrium, laser-based synthesis and processing approaches for the synthesis and processing of 2D TMD materials that permit correlated in situ spectroscopy and imaging measurements as real-time diagnostics of heterogeneity. First, advances in pulsed laser deposition (PLD) of 2D TMDs (e.g., MoSe₂, WSe₂) at elevated temperatures are described using in situ diagnostics of the growth environment (e.g., ICCD imaging and ion probe characterization of the plasma plume) and in situ diagnostics of the growing film (e.g., in situ laser reflectivity) to understand and optimize the growth process. Structure, composition, and heterogeneity of the films are characterized principally with AFM, atomic resolution HA-ADF STEM microscopy, and EELS.

Through this understanding and tuning of the PLD process, energetic atoms/molecules, to clusters and amorphous nanoparticles can be utilized for synthesis. Two effects will be presented. First, the effect of high kinetic energy atomic and molecular species to directly convert existing 2D crystals into alloys by chalcogen replacement is explored using photoluminescence and Raman spectroscopy as diagnostics of band gap shift and structure. Selective conversion of bilayer crystals is possible with this technique to form vertical heterojunctions. Second, amorphous nanoparticle ‘building block’ precursors are deposited as precursors for subsequent laser annealing studies to understand the timescales and mechanisms for their crystallization and evolution into 2D layers. This conversion is studied at two levels. First, in a custom vacuum microchamber under an optical microscope tailored for in situ Raman spectroscopy, absorption spectroscopy, photoluminescence, and SHG mapping. And second, in a prototype setup that allows laser illumination to be coupled into a (scanning) transmission electron microscope (TEM) for real-time observations of non-equilibrium synthesis directly on TEM grids.

These laser-based synthesis and processing approaches, coupled with in situ spectroscopic and imaging diagnostics, allow the stepwise evolution of synthesis to structure and optoelectronic function to be correlated through associated theory and modeling, forming a novel strategy to ‘close the loop’ for directed synthesis.

Research sponsored by the U.S. Dept. of Energy, Office of Science, Basic Energy Sciences, Materials Science and Engineering Div. (synthesis science) and Scientific User Facilities Div. (characterization science). Throughout the presentation, facilities available for collaboration at the Center for Nanophase Materials Sciences (CNMS) user facility will be presented.

The complexity of the catalytic process makes it difficult to capture the structure of the catalytically active sites. The catalytic process occurs at multiple time- and length-scales and the structure of a heterogeneous catalyst is non-uniform and deciding which of the observed species is relevant for catalysis is far from obvious. Simplification of the system by using model structures and conditions helps to determine structure – performance relationships. However, such model systems and conditions are not necessarily relevant to the actual catalytic process. Determining the structure of a catalyst while is performs its catalytic function is the domain of operando characterization and is seemingly the best way of capturing structure – performance relations. Performing an operando characterization does not necessarily yield the catalytically relevant structures.
Catalysis occurs over reactive surfaces, which may change their structure upon interaction with reactants, intermediates and products. During catalytic conversion reactants are converted into intermediates and products, thus changing the gas composition. With such change, the catalyst surface and its catalytic reactivity may change in accord. Thus, multiple structures as function of position within the reactor may exist. Also, the observation of structure in an operando experiment is no guarantee that that structure is relevant to the catalytic reaction. When taking into account relatively simple concepts from reaction kinetics, a distinction between spectator and active species can be made.

Recent advances in liquid cell transmission electron microscopy (TEM) have enabled the study of reactions in liquid-solid interface at high spatial resolution in real-time environment. The capability to apply stimuli such as electrical, heating and electrochemical measurements has already started to provide new insights on the dynamics and structural changes during nanoparticles synthesis¹, lithium charge and discharge², crystal growth³ and metal corrosion⁴. The further ability to introduce light within the confined liquid environment of TEM strengthens the capability by allowing the study of light interaction in materials during reaction at solid-liquid interface, thus enabling photoelectrochemistry in TEM. The study of hydrogen generation via photoelectrochemical method⁵ is one relevant application that can benefit from this technique and can provide fuel for energy storage solution in a clean and environmental friendly manner. However, the atomic scale mechanisms of the photocatalysts that facilitate the water splitting for efficient hydrogen generation are currently poorly understood. Further understanding of the role of various photocatalysts and the physics governing the active hydrogen evolution sites will allow for better and efficient design of photoelectrochemical devices. Here, we present the development of a unique in-situ liquid TEM holder with photo-capable light source to study photocatalytic reactions in real time at nanometer length scales. Using some of the model materials such as Au nanoprisms and MoS₂ flakes, we will present the correlation of I-V characteristics with water splitting and simultaneous structural changes at the catalytically active sites. For accurate quantitative information and reliability of materials interface in liquid, we will also present strategies to deposit materials of interests in the desired electrodes of liquid cell chips by using an inkjet materials printer. The availability and utilization of photo stimuli in liquid cell TEM can provide important fundamental insights into the understanding of several other photoelectrochemical systems.
References:
¹ Tan et al., J. Am. Chem. Soc., 2018, 140 (37), 11680-11685
² Lim et al., Science, 2016, 353 (6299), 566-571
³ Nielsen et al., Science, 2014, 345 (6201), 1158-1162
⁴ Chee et al., Micros. Microanal., 2014, 20 (2), 462-468
⁵ Maeda, K and Domen, K., J. Phys. Chem. Lett., 2010, 1(18), 2655-2661

Metal silicide, as a contact material, has played a vital role in advancing the performance of electronic devices due to its excellent electrical properties and thermal stability. As the device dimension is constantly reduced, precise understanding of the scaling behaviour of the silicide materials as well as silicide-Si interfaces is very crucial for both process integration and the device performance. For example, the typical phase evolution of nickel silicide (Ni-Si) with increasing temperature is, in general, from the metal-rich phase Ni₂Si to mono-silicide NiSi and finally the silicon-rich phase NiSi₂. The intermediate phase NiSi is often desired because it has the lowest resistivity and is stable up to 600 - 700^oC [1]. However, it is reported that for a film thickness lower than 6 nm NiSi₂ can directly grow epitaxially, bypassing the formation of the two former phases [2]. The morphologically stability of NiSi phase has been improve by the incorporation of platinum 5 - 10% (Pt) [3]. With the current fast-paced development of electronic devices, there is a substantial demand for the precise control of the silicide layer thickness at the scale of less than 5 nm, leading to the requirement of advanced characterization techniques.
Rutherford backscattering spectrometry (RBS) has been a versatile method for material research for more than five decades [4]. It can measure the composition, the thickness and the crystallinity of the materials in a non-destructive way. However, conventional RBS at the energy of 2 MeV is only effective for depth profiling films with thickness larger than 20 nm. The depth resolution can be improved by using low and medium energy ion scattering (LEIS and MEIS). The achievable depth resolution of this instrument is below 2 nm [5]. Moreover, an advanced detection method such as a position-sensitive time-of-flight (TOF) detector is able to acquire the blocking pattern of the scattered beam and reveal the crystal structure of the materials, particularly at the near surface.
In this presentation, an in-situ characterization of the silicide formation will be presented for film thicknesses of 5 - 20 nm. For this aim the samples are subjected to thermal annealing from 100 - 600^oC in the high vacuum scattering chamber of the ToF-MEIS system. We study in details the phase evolution of the silicide films as indicated by the film compositional profile. In parallel, we attempt to analyse the blocking patterns for the epitaxial layers and the role of Pt in stabilizing the NiSi phase. The real-time studies will provide insights into the silicidation process, relevant for fine-tuning the manufacturing pathways.

Reference
[1] C. Lavoie, F. M. d’Heurle, C. Detavernier, and C. Cabral, Microelectronic Engineering 70, 144 (2003).
[2] R. T. Tung, J. M. Gibson, and J. M. Poate, Physical Review Letters 50, 429 (1983).
[3] Z. Zhang et al., Applied Physics Letters 97, 252108 (2010).
[4] W.-K. Chu, Backscattering spectrometry (Elsevier, 2012).
[5] M. K. Linnarsson, A. Hallén, J. Åström, D. Primetzhofer, S. Legendre, and G. Possnert, Review of Scientific Instruments 83, 095107 (2012).

Viscosity is an essential physical property when studying the rheological behavior. However, rheological studies remain a challenging task for many laboratories, as traditional techniques are time-consuming and faced to volatility and drying issues. In a new approach to confined, high shear viscosity measurements, a novel microfluidic rheometer is equipped with an enclosed system suitable for volatile and fast drying products, providing viscosity measurements as a function of shear rate and temperature in a single experiment set-up with an excellent precision and accuracy even at very low viscosity values. Visual analysis of a laminar flow interface generated from a reference and an unknown sample allows for viscosity to be measured by taking a picture of the main channel inside the microfluidic chip, using dedicated algorithms to automatically extract viscosity data over a controlled range of shear rates and temperatures. The technology allows flow viscosity measurements of liquid products from water-like to thick formulations using a small sample volume under a wide range of shear rates from 10 to 10⁵ s^-1. In this work, we have analyzed a model blend of two solvents aimed for Li-ion electrolyte batteries at 35 °C consisting of dimethyl carbonate, ethyl methyl carbonate and dissolved lithium salts.

GaN is a material of interest for power electronics due to its high critical breakdown electric field of 3.3-3.75 MV/cm, compared to 0.3 MV/cm for silicon.A typical GaN-on-GaN power device has regions that are doped differently. For example, vertical-junction field-effect transistors (VJFET) have p-GaN region as the gate, intrinsic region as the drift layer, and the n-GaN region as the source and drain. Identifying these regions with a good spatial resolution is critical for the understanding of the device characteristics and failure analysis after malfunction. Dopant sensitive techniques like SIMS, electron holography are destructive and time consuming. Also, SIMS does not have the capability to resolve lateral variations in the layer design.
Low-voltage scanning electron microscopy has been used to study dopant contrast in GaN p-i-n epitaxial structures for power electronics applications. Secondary electron emission intensity is highest from p-layer and lowest from the unintentionally-doped (UID)-layer. Quantification of dopant contrast requires identification of all the parameters effecting the secondary electron emission intensity. In this work, we discuss about the factors effecting the secondary electron emission intensities from p-, i-, and n-type GaN layers. Also, we examine the effect of surface preparation on the dopant contrast. This technique is reliable for dopant profiling in GaN power electronics.

Gallium nitride (GaN) based power electronics has attracted much attention in recent years due to its advantages over traditional Si-based power devices in terms of energy conversion efficiency, switching speed, operation temperature, and system size. In the fabrication of advanced vertical power devices, such as junction barrier Schottky diodes, superjunction diodes, and normally-off vertical junction field effect transistors (VJFETs), it is important to produce laterally patterned p-n junctions, i.e., alternating n-type and p-type regions arranged in lateral directions. However, the mechanism of selective area doing of GaN is poorly understood. The characterization of the lateral dopant distribution in selectively doped GaN regions is difficult, due to the spatial resolution limitation of secondary ion mass spectroscopy.
In this work, selective area p-type doping of GaN on GaN bulk substrates was achieved by etching and epitaxial regrowth. The optical properties of p-GaN epilayers grown on an unintentionally-doped-GaN mesa structure have been studied using spot mode cathodoluminescence (CL) and monochromatic CL mapping. A strong variation in the luminescence characteristics has been observed, with the sidewall of the mesa structure being different to the flat upper and lower mesa regions. A comparison with luminescence characteristics of p-GaN films grown with different Mg concentrations indicate that the Mg concentration at the sidewall is significantly lower. The lower Mg concentration at sidewall is attributed to inefficient Mg incorporation of semi-polar growth. This work demonstrates that CL spectroscopy can be used as a nondestructive, contact-free method to probe local p-GaN doping concentration with submicron resolution. The determination of Mg distribution in p-GaN on a mesa structure is important for GaN-based power devices.

AAlthough there are many techniques that can detect bandgap states associated with point defects in the lattice, it is not routinely possible to determine the type of defect at submicron spatial resolution. Here we show that high-resolution electron energy loss spectroscopy (EELS) in a scanning transmission electron microscope can locate and identify point defects with a sub-nanometer probe in a wide-bandgap BAlN semiconductor.
Two BAlN thin films of 50 nm thickness were grown by metalorganic chemical vapor deposition on top of AlN templates, with B/(B+Al) gas-flow ratios of 0.12 and 0.18, respectively. In the atomic-resolution high-angle annular dark-field images, the two films exhibit considerable variation in the B concentration. The high-resolution monochromated spectra from the high B concentration BAlN film taken at 60kV show energy losses in the bandgap region. The film with low B concentration has similar energy thresholds, but with lower intensities. These energy-loss thresholds are still detectable in the AlN substrate, although with diminishing intensity as the beam is moved further from the interface, due to the long-range nature of the electromagnetic interaction. Since there is an analytical theory for the spatial variation of the signal as a function of distance between the probe and the dipole, it would be in principle possible to fit this functional form to the observed signal to localize the dipole with nanometer resolution.
To identify the origin of these features, calculations of densities of states with the VASP Density Functional Theory (DFT) code were used to explore whether postulated point defects give rise to states in the band gap, and whether transitions involving these states match experimental observations. VASP DFT calculations were performed using projection augmented wave, local-density approximation potentials for single Al, B and N interstitial atoms, single Al and N vacancies in a supercell constructed from 3x3x2 AlN unit cells. Although the bandgap energy is underestimated as 4.6 eV, it could be argued that the DFT energies for the point defect states are reliable.
In our calculations, there are no bandgap states when B is substituted for Al, and the density of states is very similar to that of AlN. In all the other cases the calculations showed states in the band gap, consistent with previous reports on point defects in AlN. The EELS thresholds at 0.39 eV and 0.79 eV respectively, are attributed to B interstitials. It is expected that they would be present in both the high and low concentration B films, albeit with lower intensity in the lower concentration film, given that the solubility limit of B in AlN is reportedly 2.8%. It would seem that other thresholds at 0.53 eV and 0.6 eV, arise from N or Al vacancies, and possibly Al interstitials. Not surprisingly these thresholds attributed to displaced Al and N atoms are more likely in the film with higher B concentration.

Gallium nitride offers several advantages over current silicon based devices, such as high-temperature applications, high-power electronics, and space-based or other high radiation exposure applications [1]. Measurements of the minority carrier diffusion length in GaN can be studied using cathodoluminescence (CL) or electron beam induced current (EBIC) in a scanning electron microscope (SEM). However, as recently reported in Yakimov et. al., the EBIC planar geometry in SEM often leads to an over estimation of the minority carrier diffusion length in n-type GaN [2]. This over estimation is attributed to the interaction volume in SEM, which is on the order of hundreds of nanometers to a few microns and within the same order of magnitude as reported minority carrier diffusion lengths of n-GaN, which are reported to be between a few tens of nanometers to a few microns. Instead of SEM EBIC, we use bulk scanning transmission electron microscopy (STEM) EBIC to measure the minority carrier diffusion length in n-GaN. Our sample consists of a high purity single crystal n-type GaN substrate (275 µm thick) grown by hydride vapor phase epitaxy (HVPE) with a threading dislocation density of ~10⁶/cm² and patterned with a nickel Schottky contact as well as an indium ohmic contact. Using a 100 kV or 200 kV accelerating voltage causes the interaction volume in our bulk specimen to be large enough to separate from the minority carrier diffusion length. We then fit the STEM EBIC line profiles in order to determine the minority carrier diffusion length of n-GaN as well as the interaction volume. In addition, we have imaged the growth of the depletion edge during an applied reverse bias, and indirectly measure the depletion width at zero applied bias.

This research was performed at the University of Maryland NanoCenter Advanced Imaging and Microscopy Laboratory. The data was taken on a JEOL 2100F TEM operating at 100kV and 200 kV in STEM mode using a Nanofactory STM-TEM holder.

[1] S.J. Pearton, et al., ECS J. Solid State Sci. Technol., 5, Q35, (2016)
[2] E.B Yakimov, J. Alloys Compd., 627, p 344-351, (2015)

The authors acknowledge funding from Nanostructures for Electrical Energy Storage (NEES), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, the National Science Foundation Graduate Research Fellowship Program under Grant No. DGE 1322106, and support from NIST Grant No. 70NANB15H218. Sandia National Laboratories is a multi-mission laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC0494AL85000.

Transmission electron microscopy (TEM) sample holders with biasing capabilities in which a mechanical mobile probe acts like one of the electrodes allow to investigate, simultaneously, the electrical properties of materials and their structural and chemical states, by using the full battery of characterization techniques available in a TEM (high-resolution imaging, electron-based spectroscopy, electron diffraction, etc.). This in-situ biasing microscopy technique has been used to investigate the conductance through rows of metal and carbon atoms [1,2], the switching of resistive memories [3], anode materials for batteries [4,5], etc. Here, we present several examples in where a recently developed in-situ TEM sample holder with a mobile probe was used to bias samples locally and discuss the practical considerations of the technique, with the objective of facilitating the experiments—which can be challenging during the mobile probe mechanical approach to the sample.

Specifically, we discuss how to approach the sample with coarse and piezoelectrically-driven fine movements and how to prepare the mobile probe, inside the TEM column, to obtain a good electrical contact between the sample and the probe prior to the experiment; by “good” we understand contacts without contamination or oxide layers that can disrupt the contact’s intrinsic electrical properties. The sample holder used during these experiments allows to approach the sample with the mobile probe in manual coarse mode with smooth and uncoupled steps of μm (or fractions of μm) among the three spatial axes. Likewise, it allows to do the final contact approach with a piezoelectric controlled mechanism that allows uncoupled nanometer-sized (or less) steps among the three spatial axes. We will present three examples: 1) observation of the Li-alloying process in yolk-shell p-Si nanorods@void@nitrogen-doped carbon nanostructures relevant for anode materials in Li-ion batteries [4]; 2) the investigation of the high capacity retention of Sb2Te3/C composites relevant for sodium-ion batteries [5]; and 3) the reaction of a solid-sate electrolyte particles under load. These examples illustrate the convenience of this in-situ biasing technique when creating a contact prior to the experiment is difficult.

[1] Hideaki Ohnishi, Yukihito Kondo, Kunio Takayanagi. Nature 395(1998), p. 780.
[2] Ovidiu Cretu, et al., Nano Letters 13(2013), p. 3487.
[3] David Cooper, et al., Advanced Materials 29(2017), p. 1700212.
[4] Fei-Hu, et al., ACS Nano11(2017), p. 8628.
[5] Ze Yang, et al., Energy Storage Materials 9(2017), p.214

Nanoparticles immersed in liquids (i.e. water or electrolyte solutions) change their solid-gas interface into a solid-liquid interface which substantially modifies the surface chemistry of the particles. Consequently, existing excess surface charges lead to an electrochemical double layer formation around the dispersed nanoparticles. Furthermore, the structure of the electrochemical double layer has a profound influence on reaction kinetics [1]. Many properties and applications of nanoparticles such as ion exchange, adsorption or catalytic reactions rely on the structure of the solid-liquid interface. The solid-liquid interface plays a major role in photocatalytic water splitting, electrocatalysis, solid-electrolyte batteries and fuel cells. SnO₂ is a material that finds application as photocatalyst or in lithium ion batteries as anode material.
Here, chemical vapor synthesized (CVS) SnO₂ nanocrystals are studied with X-ray absorption spectroscopy (XAS) in liquid colloids. The local structural environment of tin dioxide – water interface is analyzed over a range pH values (using HBr and RbOH) and the ionic strength (RbBr) to reveal changes in the interfacial region due to the modified the particle-ion interaction at the particle surface. XAS spectra at the Sn K-edge, Br K-edge and Rb K-edge are measured and compared to dry SnO₂ nanocrystals.

[1] P. Mulvaney, V. Swayambunathan, F. Grieser, D. Meisel, Effect of zeta potential on electron transfer to colloidal iron oxides, Langmuir, 6 (1990) 555-559.

Acknowledgement:
Advanced Photon Source (APS) at Argonne National Laboratory for providing beamtime

Heterogeneous catalysis of photocatalytically active perovskite nanoparticles like LaFeO₃ occurs by the interaction of molecules with active sites of the particle surface. While the crystalline core is resolved using conventional X-ray diffraction (XRD), structural changes of the boundary layer during catalytic activity are accessible by In situ X-ray spectroscopy (XAS) [1].

Since nanoparticles have a high ratio of surface to volume atoms, a large fraction of the measured signal originates from the surface. By Reverse Monte Carlo (RMC) analysis, pair distribution functions can be retrieved from XAS measurements which contain the structural evolution and information of the local structure of the particle and its surface.

To constrain the RMC model, it is important that the initial single particle model closely represents measured particles. Since in situ XAS is an ensemble measurement over thousands of individual nanoparticles, the measured nanoparticles must be phase pure, monodisperse, very small, highly crystalline and dispersible. The requirements of colloidal perovskite photocatalytic nanoparticles for in situ probing of the structure-activity-relationships are discussed and first ex situ measurements using RMC analysis will be presented.

[1] Sandmann, A., Kompch, A., Mackert, V., Liebscher, C. H. & Winterer, M. Interaction of L-Cysteine with ZnO: Structure, Surface Chemistry, and Optical Properties. Langmuir : the ACS journal of surfaces and colloids 31, 5701–5711 (2015).

Acknowledgements:
This work is supported by the German Research Foundation (DFG) through the Collaborative Research Center 247, Projekt B01. The authors gratefully acknowledge the provision of XAS beamtime at Diamond and the Advanced Photon Source and support by Fred Mosselmans and Sungsik Lee at beamlines I20 and 12-BM-B.

Recently, the research on carbon/carbon electrochemical capacitors (ECs) is essentially dedicated to increase their extractable energy which is directly related to the maximum operating voltage of the electrode/electrolyte system [1, 2]. Long-time operation at maximum voltage (so-called potentiostatic floating) accelerates ECs ageing, revealed by loss of capacitance, increase of resistance and internal pressure. The worsening of these three parameters is related to carbon surface oxidation and/or pore blockage by electrolyte decomposition products, together with ionic starvation in the electrode due to decrease of electrolyte concentration and gas evolution which entails weakening of the adhesion between the active mass and the current collectors [3, 4].
In aqueous medium, the most pronounced symptom of the degradation phenomena is H₂ and CO evolution appearing at the negative and positive electrodes, respectively [5, 6]. Besides, the surface of the aged positive carbon electrode is modified by new oxygenated groups, while the aged negative electrode is also functionalized, although one could expect that it should not occur at this electrode under reductive conditions. Indeed, H₂O₂ which is first produced at the positive electrode as by-product of water electrolysis, migrates through the separator towards the negative compartment being then further involved in oxidation of the carbon surface [7].
In this context, it will be shown that it is possible to obtain information about all the realistic mechanisms of ECs ageing as well as to determine the relative impact of each process on the system degradation by using a combination of operando analyses consisting of electrochemical measurements, mass spectrometry and cell pressure records, together with post-mortem surface functionality analysis of individual carbon electrodes by temperature programmed desorption (TPD). On the example of an electrochemical carbon/carbon capacitor in aqueous Li₂SO₄, the approach is based on assigning the charge passing through the cell under potentiostatic floating (known as leakage current) to the charge: i) spent at each electrode for production of gases (taking into account operando EMS analysis and pressure records); ii) utilized to oxidize the electrodes surface (from post-mortem surface functionality analyses realized by TPD); iii) being redistributed throughout the electrochemical cell (taking into account the steady-state leakage current measurements).

Metal-Organic Frameworks with surface areas exceeding 5000 m²/g are promising materials for gas storage and sensing applications. Ultrahigh porosity gives rise to new counterintuitive phenomena such as flexing, gating, breathing and negative gas adsorption (NGA). The switching mechanism of structural transformation in flexible crystalline frameworks involves huge cell volume changes exceeding 200 %. Parallelized diffraction-adsorption instrumentation (in situ analysis) at temperatures ranging from 60 – 300 K connected to a high resolution adsorption instrument (0.01 Pa – 100 kPa) is an enabling technique to shine light on the structural transformations via Rietveld refinement of intermediary phases using synchrotron X-ray powder diffraction. Molecular simulations complement the findings and provide means for rationalizing the energetics along the adsorption pathway. The interplay of customized syntheses, in situ X-ray and NMR characterization of dynamics, and simulation gives detailed insights into switching mechanisms and their kinetic barriers. Switchable MOFs show great potential as threshold sensors, gas storage materials, or highly selective adsorbents. An in depth understanding of switching phenomena may enable the highly selective recognition of small molecules in biomimetic devices in future.

S. Krause, V. Bon, I. Senkovska, D. M. Többens, D. Wallacher, R. S. Pillai, G. Maurin, S. Kaskel, Nat. commun. 2018, 9 (1), 1573; P. Freund, I. Senkovska, S. Kaskel, ACS Appl. Mater. Interfaces 2017, 9 (50), 43782-43789; S. Krause, V. Bon, U. Stoeck, I. Senkovska, D.M. Többens, D. Wallacher and S. Kaskel, Angew. Chem. Int. Ed., 2017, 56, 10676-10680; S. Krause, V. Bon, I. Senkovska, U. Stoeck, D. Wallacher, D. M. Tobbens, S. Zander, R. S. Pillai, G. Maurin, F. X. Coudert and S. Kaskel, Nature, 2016, 532, 348-352.

The energy flux to a surface during plasma exposure and the associated surface heating are of long standing interest as they contribute to the physico-chemical changes that occur during plasma-based materials synthesis and processing. Indeed, the energy delivered to the surface, via a flux of particles and photons, in concert with a flux of reactive species serves to chemically modify, etch, and/or deposit materials, with an efficacy that depends on the plasma processing environment. A unique feature of plasma synthesis and processing is that most of the delivered energy is absorbed at or very near the surface over short (picosecond) time scales. The dissipation of thermal energy proceeds through electron-electron and/or electron-phonon interactions as they propagate through the material, with relaxation time scales that can be orders of magnitude slower. Typically then, the surface is not in thermal equilibrium with the bulk material. Fast, surface-sensitive techniques are thus required to fully appreciate the dynamics of the plasma-surface interaction. In this work, we employ pump-probe Time-Domain Thermoreflectance, a surface sensitive technique typically used to measure thermal properties of thin films, to determine electron heating of thin metal films during exposure to an atmospheric pressure plasma jet. The results, in conjunction with current measurements, are used to develop a first order understanding of plasma jet-surface interactions. The results show that the energy delivered by the plasma jet causes a localized increase in electron energy within the thin film over an area commensurate with the plasma jet radius. We also describe measurements in which we monitor the time evolution of the surface temperature durign and after plasma excitation, demonstrating both heating and cooling due to chemical changes to the surface.

Semiconductor nanowires are promising building-blocks for the next generation of solar cells [1]. With diameters of a few hundred nanometers, vertical nanowires can act as waveguides for the incoming sunlight and thereby enhance the light absorption. The nanoscale geometry further enables the growth of defect free heterostructure interfaces including also defect free growth on lattice mismatched substrates, not possible for bulk structures due to the magnitude of interfacial strain. Because of these and other unique properties, nanowire-based solar cells have the potential to provide higher efficiencies and lower material costs than their thin-film counterparts. One of the main challenges for realizing these high-performance nanowire solar cells lies in the optimization of the electrical and photovoltaic properties at the individual nanowire level. Here characterization techniques play a vital role. Standard solar cell characterization methods do not reveal the contribution of individual nanowires. Techniques involving single nanowire lift-off alter the working conditions, do not include the effect of the nanowire/substrate interface and generally demand a great deal of sample preparation. Hence, techniques where single nanowires can be probed directly on the growth substrate are advantageous.
In this work, we have developed a technique where a scanning tunneling microscope (STM)-probe is operated inside a scanning electron microscope (SEM) for characterization of the photovoltaic and electrical properties of single as-grown semiconductor nanowires. The SEM provides live imaging with nanoscale resolution while the piezo-controlled STM-probe contacts individual nanowires forming a closed electrical circuit. Reproducible ohmic contacts between the STM-probe and the nanowires were established through electron beam induced Pt deposition and Joule heating. Photocurrent measurements were enabled by mounting a light emitting diode (LED) in the SEM chamber. Current – voltage (I-V) measurements in both dark and illuminated conditions were performed on single GaAs nanowires with built-in radial p-i-n junctions. The best performing nanowire had an apparent power conversion efficiency of around 11 % under white LED illumination. Electron beam induced current (EBIC) maps were recorded to show the local charge separation ability within a nanowire with a spatial resolution of around 250 nm. Furthermore, the STM-probe was used to bend single nanowires, which affected the I-V characteristics. In conclusion, the in situ STM-SEM technique offers the possibility to investigate and compare the functional properties of individual as-grown nanowires that are part of a nanowire solar cell. The properties can be directly linked to the structure of the individual nanowires. This may play a key role in the further improvement of nanowire solar cells.

References.
[1] G. Otnes and M. T. Borgström, “Towards high efficiency nanowire solar cells,” Nano Today, vol. 12, pp. 31–45, 2017.
[2] J. Holmér, L. Zeng, T. Kanne, P. Krogstrup, J. Nygård, L. de Knoop, E. Olsson, "An STM – SEM setup for characterizing photon and electron induced effects in single photovoltaic nanowires", Nano Energy, vol. 53, pp. 175-181, 2018.

In situ diagnostics can yield fundamental insights into the materials formation process and aid in the design and optimization of materials for energy technologies, but in situ measurement has not been widely adopted for many time-resolved spectroscopic measurements. In particular, transient absorption (TA) spectroscopy can reveal the kinetic pathways of photoexcited species that dictate materials function, but TA is rarely performed in situ during materials formation. In TA a pump pulse photogenerates electronic excitations in the material, and the differential absorption caused by these excited species is measured by a probe pulse after a controlled delay time. The measurements at each delay time are typically collected in series, resulting in data acquisition times from many minutes to multiple hours. Many materials undergo formation processes on a shorter timescale than the data acquisition time for TA, precluding the accurate measurement of excited state dynamics in these systems. In this work, we introduce a novel implementation of transient absorption spectroscopy that can measure transients with up to a 60 ps pump-probe time delay in one shot. Data with an excellent signal-to-noise ratio can be acquired in just a few seconds. The measurement is validated by comparison to typical implementations of TA in static systems, and we report the results of in situ measurements during the formation of materials with applications in photovoltaics and light emitting devices. The information gained using this technique will be used to modify environmental parameters during the materials formation process to steer the material towards kinetically trapped states with excited state dynamics tailored for particular types of devices.

Synchrotron-based soft x-ray spectroscopies, namely soft x-ray absorption (XAS) and emission (XES) spectroscopy, as well as their combination resonant inelastic soft x-ray scattering (RIXS) are powerful techniques to investigate the electronic structure of a variety of systems. We have intensively used these tools for the investigation of applied materials including thin-film solar cells, photoelectrochemical devices, batteries, … (e.g., [1,2,3]). For these systems, in situ and operando studies are often required to answer the pertinent materials questions necessary for a further optimization of the respective device. Due to the photon-in-photon-out nature of XAS, XES, and RIXS with soft x-rays, it is, in fact, possible to investigate gaseous and liquid samples as well, making such in situ and operando investigations possible. However, since these spectroscopies and the corresponding synchrotron beamline environment usually require ultra-high vacuum (UHV) conditions, this requires the development of sophisticated experimental setups.
In this presentation, we will discuss our recent progress in developing suitable environmental cells for in situ and operando soft x-ray studies. Showing selected examples, it will be presented how these techniques give insight into the electronic and chemical properties in the gas and liquid phase, as well as at liquid/solid interfaces. In many cases, the investigated materials are very sensitive to x-ray radiation, which requires careful experimenting with optimized approaches that will be discussed as well.

[1] L. Weinhardt, D. Hauschild, and C. Heske, Advanced Materials, in print (2019).
[2] T.R. Hellstern, D.W. Palm, T.F. Jaramillo, J. Carter, L. Weinhardt, M. Blum, C. Heske, A.D. DeAngelis, K. Horsley, N. Gaillard, and W. Yang, ACS Appl. Energy Mater., in print (2019).
[3] A. Léon, A. Fiedler, M. Blum, A. Benkert, F. Meyer, W. Yang, M. Bär, F. Scheiba, H. Ehrenberg, L. Weinhardt, and C. Heske, J. Phys. Chem. C 121, 5460 (2017).

Synthesizing new electrode materials with the desired phases and properties has often found difficult and mostly relies on trial and error since there are a variety of parameters that can have a strong influence on the structural properties (crystal structure, morphology, particle size), and eventually on electrochemical performance (e.g., capacity, rate capability, and durability) of the synthesized materials. Materials synthesis is such a complex process and often undertaken under non-equilibrium conditions, therefore posing a great challenge to theoretical predictions. In-situ, real-time tracking of the structural chemistry evolution during the synthesis reactions can provide details on reaction process, elucidating intermediate phases and how temperature, pressure, time and the precursor concentrations affect the kinetic reaction pathways. Synchrotron X-ray powder diffraction (SXPD), with its high beam intensity and angular resolution, allows in situ structural characterization in short time scales capable of catching subtle structural changes in the materials under synthesis. The presentation will introduce the in situ SXRD techniques developed at NSLS-II, as applied in the study of synthesis processes of cathode materials for Li-ion batteries. Specific examples on different materials and synthesizing methods will be given to show how the insights gained from in-situ studies may provide guidance to rational design and synthesis of new battery materials.

Lithium-rich layered oxides (LrMOs; Li1+xMOy, M=Ni, Co, Mn) are among the most promising high energy cathode candidates for next-generation Li-ion batteries, with excess capacity enabled by both cationic and anionic redox. However, activating anionic redox (AR) may cause voltage hysteresis and voltage decay and so rational design of LrMOs of controlled AR activity becomes crucial to practical application. In this presentation, we will present recent results from combined in situ Synchrotron X-ray scattering and absorption studies on the AR activity in Li1+xMOy under operating conditions. Findings from this study, particularly about correlation between AR activity and local structural ordering, along with its implication to rational design of LR layered oxide cathodes will be discussed.

Materials storing energy via alloying reaction are promising anode candidate in rechargeable lithium ion batteries due to their much higher energy density than current graphite anode. Among these materials, bismuth provides a high theoretical volumetric capacity of 3800mAh/cm⁻³ when it alloys with Li⁺ to form Li₃Bi. Yet, its further development has been hurdled by the unclear alloying mechanisms with Li⁺. In this work, we carry out real-time study of the (de)alloying kinetics of Bi nanomaterials working as a LIB anode combining both in situ synchrotron XRD and in situ TEM. In situ synchrotron XRD reflects the phase transition information of the ensemble of many Bi nanoparticles, while in situ TEM demonstrates the structural evolution of single Bi nanoparticle at both high spatial resolution and high temporal resolution. Both two methods reveal that the Li-BI alloying is a two-step process featuring two two-phase reactions with formation of Li₁Bi as the intermediate phase. More importantly, our work reveals an energetically favorable lithiation pathway along Bi-(012) planes during the Bi-Li₁Bi phase transition process, which is further theoretically understood by DFT calculations. The fundamentals discovered in this work are expected to further guide the design of Bi-based electrode materials for battery performance enhancement in the future.

High-Ni transition metal layered oxides are among the most promising cathodes for Li-ion batteries. Their electrochemical activity is largely determined by the cationic ordering in octahedra (e.g., the basic building units, BBUs, of the systems). Therefore, mechanistic understanding of how the BBUs are inherited from precursors and subsequently evolve into the desired ordering during synthesis, may provide guidance to preparation of high-performance high-Ni layered oxides. Herein, a multimodal in situ x-ray characterization approach is employed to investigate structural/chemical evolution at scales varying from long range to the local individual octahedral units during solid-state synthesis of LiNi_0.77Mn_0.13Co_0.10O₂. Real-time observation corroborated by first-principles calculations reveals a topotactic transformation throughout the entire process, during which the breaking and reconstruction of NiO₆ octahedra occurs, leading to Li/Ni mixing and ordering. Findings from this study shed light on designing high-performance layered oxide cathodes and, broadly, various functional materials through synthetic control of the constituent BBUs.

Neutron and synchrotron x-ray facilities are indispensable scientific resources and provide powerful instruments and experimental techniques for both fundamental and applied researches. Neutrons and X-rays interact with matter in different ways, thus are often used as complementary tools for studying materials at the electronic, atomic and molecular levels. We have been utilizing neutron and synchrotron techniques to investigate a variety of materials for energy storage applications. In this talk, we will discuss the difference and similarity of neutrons and synchrotron x-rays, and their complementary use in battery materials research. We will present our recent results on in-situ/operando characterization of advanced materials for rechargeable batteries, ranging from material synthesis to safety. We will also discuss some challenging issues in battery materials characterization.

Here we present the full upgrade of neutron time-of-flight spectrometer NEAT at Helmholtz Zentrum Berlin and show first experimental results related to the in-situ investigation of the gas storage in porous materials [1], [2]. NEAT spectrometer has a long history of successful application to the study dynamics and function and is best suited to probe dynamic phenomena directly in space and time in the large time domain 10^-14 – 10^-10 s and on the length scale ranging from 0.05 to up to about 5 nm [3]. The upgrade started in 2010 after a rigorous internal and external selection process and resulted in 70 to 300 fold gain in neutron count rate compared to predecessor. The advanced features of the new instrument include novel integrated guide-chopper system, which is optimized for both hard and soft matter studies and is several times more efficient in the use of the source power than world leader neutron spectrometer IN5 at ILL/ France. The neutron guide delivers neutrons with flexible beam properties: either highly homogeneous beam with low divergence for single crystals studies or “hot-spot” neutrons suitable for small samples. Substantial increase of the detector angle coverage was achieved by using 416 ³He linear position sensitive detectors equipped by advanced technology electronics.
The increase in sensitivity of the instrument coming from high count rate and low background allowed us to reduce substantially the amount of sample measured on NEAT. This is particularly important in the investigations of novel materials where only limited sample amounts are available. Further, the advanced design of NEAT resulted in new instrumental capabilities and enable new research applications. Higher flux allows NEAT to have an access to the wavelength as short as 1.3 Å consequently doubling the range at neutron energy loss studies to up to 40 meV and substantially improving experimental conditions for low energy studies. The non-magnetic environment of the sample and detector chamber is ideal for studies with high-magnetic field or polarization analysis. One of the most important benefits is the event recording data collection which allows us to follow changes of sample conditions in real time on the scale of minutes or seconds. In combination with implemented sample environment for the controlled gas atmospheres and high pressures it opens up new opportunities to study the kinetic processes in materials.
References
[1] M. Russina, G. Günther, V. Grzimek, L. Drescher, M.C. Schlegel, R. Gainov, T. Kaulich, W. Graf, A, Daske, K. Grotjahn, R.Hellhammer, G. Buchert, H. Kutz, L. Rossa, O. P. Sauer, M. Fromme, “Upgrade Project of TOF spectrometer NEAT at Helmholtz Zentrum Berlin”, Physica B Condenssed Matter, 2018, doi: 10.1016/j.physb.2017.12.026
[2] M.-C. Schlegel, D. Többens, R. Svetogorov, M. Krüger, N. Stock, H. Reinsch, D. Wallacher, R. Stewart and M. Russina, “Conformation-controlled hydrogen storage in the CAU-1 metal–organic framework”, Phys. Chem. Chem. Phys. 18 (2016) 29258, DOI: 10.1039/C6CP05310F
[3] M. Russina, E. Kemner, F. Mezei, ”Intra-cage dynamics of molecular hydrogen confined in cages of two different dimensions of clathrate hydrates”, Scientific Reports. 6, 27417 (2016)

Although the lots of research have improved the performance of lithium-ion batteries, their cycle stability, and energy density at high rate condition are still insufficient. Recently, Silicon-based material, having much higher theoretical capacity than conventional graphite anode, has emerged as one of the most promising new anode material to follow those requests. However, the general problems of silicon active materials, large volume expansion and unstable formation of a solid-electrolyte interface layer, have hindered the real application in industry field. In this work, silicide-nanowires anchored on the inner surface of 3D graphene micro-tubular electrode has been synthesized using the two-step chemical-vapor-deposition process. The 3D Graphene gives not only conductivity to electrode but also the effect of controlling electric potential for inhibition of solid-electrolyte interphase formation. With these properties, a solid-electrolyte interphase forms on the outer surface of the graphene, while solid-electrolyte interphase-free nanowires enable fast electronic and ionic transport, as well as strain relaxation over 2000 cycles, with 84% capacity retention even at ultrafast cycling (>20C). Moreover, these anodes exhibit high rate capabilities with capacity retention higher than 88% at 80C (vs. the capacity at 1C).

A major limitation to the advanced lithium ion batteries with silicon based anode is the huge volume expansion during lithiation, which leads the breakdown the electrode structure, a low initial coulombic efficiency and capacity fading upon cycling. A number of approaches, such as downsizing the silicon structures, modifying the structures with nanowires or porous matrix or making Si/C composites have been performed to reduce the volume expansion phenomena. However, several critical issues such as un-satisfactory cyclability and the capacity decay still could not be avoided despite the prolonged cyclic life of Si anode. In present work, a synchrotron radiation based x-ray spectroscopic techniques with in-situ/operando capabilities were conducted to elucidate the structural evolutions of silicon anode upon electrochemical reactions in real time. SiO/C anodes were employed as a target material since they have revealed the excellent capacity retention. The results demonstrate that in-situ/operando x-ray characterization techniques provide the unique information for understanding the real reaction mechanism.

As the in-situ X-ray powder Diffraction has become a routine method to characterize the crystal structure of battery particles under operando condition, it reflects the assemble average of all fundamental particles in the X-ray illumination volume. However, the local behavior remains a mystery until Ulvestad et al, [1] first applied Coherent X-ray Diffractive Imaging on a single LiNi_0.5Mn_1.5O₄ nanocrystal under operando condition. The application of the Coherent Hard X-ray Diffraction in Bragg geometry enable us to visualize the structural, morphological and strain field evolution of crystals in operando battery materials. Coherent multi-crystal diffraction provides collective information of tens of crystals that bridges the statistical and individual behaviors of battery particles. In this presentation, Coherent Hard X-ray Diffraction experiment is demonstrated with a layered cathode material under operando condition [2]. I will show in details the fundamentals of Coherent X-ray Diffraction, the sample screening, instrumentation, data acquisition and interpretation.

[1] Ulvestad, A., Singer, A., Cho, H. M., Clark, J. N., Harder, R., Maser, J., Meng, Y. S. & Shpyrko, O. G. (2014). Nano Lett. 14, 5123–5127.
[2] Li, L., Xie, Y., Maxey, E., & Harder R. (2019) J. Synchrotron Rad. 26, 220-229.

There is significant interest worldwide to develop new battery systems for transportation and grid-storage applications. Development of new transformational technologies will require tailoring advanced materials and incorporating additional chemical or electrochemical processes. Using X-ray absorption, X-ray emission, and X-ray Raman spectroscopy, we seek to capture the fine details of the redox behavior (electronic structure) and the complex local structure of battery materials under device operating conditions. Such studies provide unique insight to develop structure-property correlations that are important to predict long-term battery performance and also help tailor new novel materials with high energy and power densities and stable crystal structures. In this talk, we will discuss operando studies of battery systems performed at the Advanced Photon Source.

In the past decades, renewable energy storage and energy conversion systems, such as fuel cells, water electrolysis, and metal-air batteries, have attracted great attention. Oxygen evolution reaction (OER) is a key half reaction of water splitting to produce clean fuels. However, the sluggish kinetics of OER has significantly limited the performance and commercialization of such energy conversion devices. Up to now, the most efficient OER catalysts are still noble metal and metal oxides of Ruthenium (Ru) and Iridium (Ir), which are not cost-effective catalysts and unstable under high potentials. Recently, single atom catalysts have been used to improve the surface-to-volume ratio to increase OER catalytic activity. In our work, we used layered double hydroxides (LDH) known as anionic or hydrotalcite-like clays, which are inexpensive but are good catalysts or support for single atom catalyst to exhibit much improved OER activity. In particular, we utilized in situ XAS to identify the catalytic active centers. Our in situ works also provide insights into the mechanism of these catalysts during OER.

The success of material discovery and engineering is partly based upon the development and application of highly advanced characterisation methods. Hard X-ray photoelectron spectroscopy (HAXPES) has traditionally found its application in the core topics of condensed matter physics, but the slowly growing number of beamlines worldwide has widened its appeal to other interest groups. HAXPES uses X-rays in the 2-10 keV range to excite photoelectrons, which are used to non-destructively study the chemical environment and electronic structure of materials. As its soft X-ray counterpart (SXPS or simply XPS) it delivers qualitative and quantitative information on elemental distributions, chemical states, and electronic structure.

In contrast to the very surface-sensitive XPS, HAXPES is much more bulk sensitive. This makes it applicable to bulk materials and structured samples, e.g. layered samples and heterostructures. In addition, its bulk sensitivity means that realistic samples can be investigated without the need of prior surface preparation. Furthermore, results can easily be correlated to a whole range of other bulk-sensitive characterisation results, and therefore to structural information and material properties. However, the number of existing HAXPES systems is very small and they are predominantly located at synchrotrons (approx. 20 beamlines worldwide) due to low photoionization cross sections necessitating high X-ray intensities, limiting their availability to users and applications.

This work presents a new laboratory-based instrument capable of delivering monochromated hard X-rays with an energy of 9.25 keV and a focused 30x45 μm² X-ray spot, giving excellent energy resolution of <0.5 eV. Systematic reference measurements are presented outlining the systems capability, including energy resolution, X-ray and signal intensity, and stability. Results from a rutile TiO₂ bulk sample show the wide range of possible signals that can be collected, such as shallow and deep core levels, Auger lines, and valence bands. The valence band measurements are benchmarked against theoretical results from density functional theory. Beyond pure bulk measurements a range of metal oxide multilayer structures, as are used in transistor and diode applications, is investigated to showcase the system’s ability to detect signals from buried layers and to analyse them for their local chemical states.

Ultimately, this spectrometer presents an alternative to synchrotron-based endstations and will help to expand the number and range of HAXPES experiments performed in the future. HAXPES is a cutting edge characterisation method and the advancement of this technique will tremendously increase the potential to study an ever increasing range of inorganic materials and beyond.

Determining the structure of active species and active sites, and their evolution in reaction conditions is a required step towards understanding catalytic mechanisms. In heterogeneous catalysts, multiple species of the same element: single atoms, larger clusters and nanoparticles may appear, disappear and coexist during the reaction. There is therefore a strong need to monitor the charge and geometry of metal centers, as well as their dynamic changes, in reaction conditions. Even with that capability achieved, the reaction mechanism may not be understood until it is known which species are catalytically active, and which ones – spectators. One possible way to address that is to combine multiple in situ probes that can resolve between the heterogeneously distributed species with the online product analysis. This combination will correlate the changes in the spectra (and thus – in the structure) with the changes in activity and thus help identify the active species during the reaction.

To illustrate these ideas, and highlight an exemplary application of this multimodal approach I will show how we treated the complexity of a newly synthesized and highly active selective oxidation catalyst, a model single atom catalyst formed by the dilute substitutional doping of Ceria by isolated Pt atoms, (Pt)CeO₂. Among the techniques chosen for its characterization, STEM probes the distribution of single atoms and nanoparticles in the sample, in situ XANES and EXAFS probe the local structure around metal atoms/ions, XPS and RIXS – their charge states, DRIFTS – surface composition, and online MS – product formation. We used this multimodal approach to reveal the correlation of the number of Pt²⁺ ions in Pt/CeO₂ catalysts with the observed change in the water-gas shift reaction activity—data definitively establishing that Pt nanoparticles formed during the reaction, and not the Pt single atoms, are the active species. The Pt atoms on the nanoparticle perimeter with proximate Ceria bonding interactions comprise the catalytically active sites.

This work was supported by U. S. Department of Energy, Office of Basic Energy Science under Grant No. DE-FG02-03ER15476

Optically-excited localized surface plasmon (LSP) resonances have been used to induce chemical reactions, such as hydrogen dissociation, water splitting, and ethylene epoxidation¹. Energy harnessed by plasmonic nanostructures and transferred to adsorbed reactants on nanostructure surfaces initiates chemical processes that compensates for the thermal energy required otherwise. However, sub-particle level information, regarding the relevant LSP resonance modes and their response to reactants, has not been identified by variants of optical spectroscopy.

Herein, we use a focused electron beam in an environmental scanning transmission electron microscope (ESTEM) to excite LSP resonances on triangular Au nanoprisms on TiO₂ support to drive CO disproportionation at room temperature (commonly driven by thermal energy at 400 °C to 600 °C), and the carbon product labels the catalytically active site at the sub-particle level. ESTEM, operated at 80 kV and equipped with a monochromated electron source with an energy resolution of 0.08 eV, was used to acquire in situ low- and core- loss electron energy loss spectroscopy (EELS) maps to spatially resolve both LSP mode distribution (low-loss EELS) and gas adsorption sites (core-loss EELS) on a Au nanoprism by comparing the data collected in vacuum and CO environment. The optimal electron beam position to excite specific LSP resonance modes on the Au nanoprism can be identified by applying the non-negative matrix factorization (NMF) to a low-loss scanning transmission electron microscopy (STEM) EELS map². We found a dipole mode that resulted in the maximum LSP electric field can be excited by placing electron beam at an aloof position outside the cantilevered nanoprism corner (near-field excitation). This dipole mode displayed a spectral blue-shift in the presence of CO (P_CO = 110 Pa) and returned to the original resonance energy when CO is evacuated. More importantly, STEM cathodoluminescence (CL) measured simultaneously with EELS allowed us to probe the scattering from the Au nanoprism in either vacuum or CO environment, which can be used to apprehend the absorbed energies by the Au nanoprism during the LSP induced chemical process.

A core-loss STEM-EELS map acquired at P_CO = 110 Pa, showed increased carbon K-edge signal at the selective locations on the nanoprism edges, which can be related to CO adsorption. By either raster-scanning an electron beam over the Au nanoprism or deliberately placing it statically at an optimal near-field excitation position in CO environment, the electric field of excited LSP mode can drive disproportionation of the adsorbed CO molecules in the vicinity and deposit amorphous carbon at the cantilevered corner. The location of carbon deposit suggests that the catalytically active sites on Au nanoprisms are those where preferred CO adsorption sites and locations of maximum LSP electric field magnitude (resonance antinodes) superimpose. Our findings provide, for the first time, detail experimental information on sub-particle localization of plasmonic catalysis with nanometer scale precision and possible guidelines to develop antenna based plasmonic catalysts.

References
1. Zhang, Y. et al. Surface-Plasmon-Driven Hot Electron Photochemistry. Chem. Rev. 118, 2927-2954 (2017).
2. Nicoletti, O. et al. Three-dimensional imaging of localized surface plasmon resonances of metal nanoparticles. Nature 502, 80-84 (2013).

Iron containing nanoparticles exhibit a broad spectrum of applications due to a large variety of chemical (e.g. redox potential) and physical properties. Those are highly depending on the oxidation state and crystal structure. Different phases and morphologies can be achieved by changing the nanoparticle synthesis parameters [1]. Chemical vapor synthesis (CVS) is a technique which enables reproducible synthesis of nanoparticles with reliable control of synthesis parameters. Organometallic compounds are used to introduce metals into CVS reactor. Solid precursors offer easier handling, mostly lower toxicity and are advantageous for adjusting gas redox potential through a higher carbon content.
Iron(III)acetylacetonate (Fe(acac)₃, Fe(C₅H₇O₂)₃) is such solid precursor used for CVD (chemical vapor deposition) and CVS to generate iron containing nanoparticles. However, due to a higher molecular mass and intermolecular interaction, the volatility of Fe(acac)₃ is relatively low, still it needs to be sublimated to deliver it as vapor to CVD and CVS processes. In fact, only little is known about the local and electronic structure of Fe(acac)₃ vapor.
In this study we designed a novel vapor cell for measuring vapor species in the gas phase. With this vapor cell Fe(acac)₃ vapor is successfully provided for XAS (X-ray Absorption Spectroscopy) measurements under fluorescence detection. We investigate Fe(acac)₃ vapor in-situ at the Fe-K edge using the vapor cell at conditions close to those in the reactor. We compare those data to solid Fe(acac)₃ and Fe(acac)₃ in solution (water, isopropanol and acetone). Additional, we investigate nanoparticles made from Fe(acac)₃ under different synthesis conditions by XAS. This enables a better understanding of processes occurring in the CVS system using iron as a probe.

[1] Levish, A. and Winterer, M., Nanoparticles generated by combining hot wall and microwave plasma chemical vapor synthesis, MRS Adv. 1-6, (2018).
Acknowledgment:
DFG (German Research Foundation)
DLS (Diamond Light Source), DESY (Deutsches Elektronen-Synchrotron), APS (Advanced Photon Source) for providing beamtime

Electrocatalysis plays a key role in producing/converting molecular energy carriers and metal to either store electricity from solar or wind energy in chemical forms or produces electrical energy through fuel cells and metal-air batteries. In the search of highly efficient electrocatalysts, it is important to understand how catalysts work, then subsequently a ration design rule can be provided. We use several in-situ synchrotron X-ray techniques, particularly X-ray absorption spectroscopy (XAS), to probe the element-specific changes from heavy metal (e.g., Mn, Co, Fe, etc) to light element (e.g., O, N) during various electrochemical reactions including CO₂ reduction and oxygen evolution/reduction reactions. We not only directly observe how catalysts adjust their local structures and valence state to promote reactions, but also identify different reaction centers that weren’t considered before and couldn’t be found from ex situ experiments. In my talk I will show couple of cases to highlight the importance of our in situ studies as well as new thoughts for catalyst designs.

In this presentation, we explore the pore structure of overcoated materials by in situ synchrotron small-angle (SAXS) and wide-angle X-ray scattering (WAXS). Thin films of aluminum oxide (Al₂O₃) and titanium dioxide (TiO₂) with thicknesses of 4.9 and 2.5 nm, respectively, are prepared by atomic layer deposition (ALD) on nonporous nanoparticles. In situ X-ray measurements reveal that porosity is induced in the ALD films by annealing the samples at high temperatures. Moreover, this pore formation can be attributed to densification resulting from an amorphous to crystalline phase transition of the ALD films as confirmed by high-resolution X-ray diffraction and the pair distribution function. Simultaneous SAXS and WAXS results show not only that the porosity is formed by the phase transition but also that the pore size increases with temperature.

Electrophoretic deposition (EPD) is a colloidal forming process which may be described by two steps: (1) migration of charged particles in an electric field due to electrophoresis and (2) deposition on a substrate working as electrode. Nevertheless, due to a lack of in-depth mechanistic understanding of the film deposition and formation process control of the film microstructure remains still a challenge. Generally accepted is that particle agglomeration plays an important role during deposition. Recent studies investigated ligand-stabilized nanoparticles in organic solvents during EPD with small-angle X-ray scattering [1, 2].
We investigate the EPD process of chemical vapor synthesized (CVS) SnO₂ nanocrystals. Purely electrostatic stabilized colloidal SnO₂ nanocrystals in aqueous media are deposited on silicon anode due to their high negative charge in an X-ray transparent liquid EPD cell. We use synchrotron radiation based small-angle X-ray scattering (SAXS) in situ in the near electrode region. This approach allows to study structural changes like radius of gyration, fractal dimension at the nanoscale during EPD by probing locally with high spatial resolution (nanobeam) time dependent at the local electrode environment (200 – 3200 nm).

[1] A.J. Krejci, K.G. Yager, C. Ruggiero, J.H. Dickerson, X-ray scattering as a liquid and solid phase probe of ordering within sub-monolayers of iron oxide nanoparticles fabricated by electrophoretic deposition, Nanoscale, 6 (2014) 4047-4051.
[2] Y. Yu, D. Yu, B. Sadigh, C.A. Orme, Space-and time-resolved small angle X-ray scattering to probe assembly of silver nanocrystal superlattices, Nat. Commun., 9 (2018) 4211.

Acknowledgement:
Deutsches Elektronen-Synchrotron (DESY) for beamtime allocation

The Ion Physics group and the Tandem laboratory at Uppsala University have recently developed substantial infrastructure for in-situ film deposition, modification and high-resolution characterization using ion beam-based methods, by set-ups located at three separate accelerators: a 5 MV Pelletron Accelerator, a 350 kV ion implanter/medium-energy-ion-scattering (MEIS) system and a 10 kV Time-of-Flight low-energy-ion-scattering (ToF-LEIS) system. For example, the 6^th beam line of the Pelletron Accelerator features an analytical chamber equipped with a triple-electron-beam evaporator, an ion source used for sputtering, a solid state detector, an x-ray detector and a pure germanium detector used for a series of ion beam characterization techniques: Rutherford backscattering spectrometry (RBS) providing depth profiles of targets constituents with down to 20 nm depth resolution, nuclear resonance analysis (NRA) and particle-induced X-ray emission (PIXE). Additionally, different gases, such as hydrogen and oxygen, can be introduced into the chamber during thin-film growth or beam exposure [1]. The sample temperature can be controlled from cryogenic levels up to typical values 600 ^oC during or after the growths. A fully independent load-lock chamber can be detached and transferred to the MEIS system for further analysis without exposing the film to air. For the MEIS system, the operational energies range from 20 kV to 350 kV, enabling high-resolution near surface analysis [2]. Scattering data can be acquired by two independent ToF-position-sensitive detectors. One system features a large solid angle (0.25 sr), suitable for acquisition of blocking pattern permitting to study depth-resolved crystallographic properties in real space. The other detector features smaller solid angle (0.02 sr), but is situated at larger distance from the sample in order to achieve superior depth resolution (typically ~5 Å). The sample holder can be heated up to 600 ^oC for in-situ annealing experiments. Finally, for the LEIS system, the main analytical potential is to grow and investigate ultra-thin films in a regime very close to the sample surface (few monolayers). An ultra-high vacuum sample preparation chamber connected to the LEIS system features a triple-electron-beam evaporator, an annealing stage, a sputter gun mainly used for surface cleaning prior the measurements, low-energy-electron-diffraction and Auger electron spectroscopy.

Reference

[1] K. Kantre, M. V. Moro, D. Moldarev, D. Wessman, Max Wolff and Daniel Primetzhofer, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms (Submitted).
[2] M. K. Linnarsson, A. Hallén, J. Åström, D. Primetzhofer, S. Legendre, and G. Possnert, Review of Scientific Instruments 83, 095107 (2012).

In situ processing and comprehensive characterization is essential for design and development of materials used and processed at high-temperatures. Here, a new cluster tool for processing and depth-resolved compositional, structural and optical characterization of layered materials with thicknesses ranging from sub-nm to 1 µm at temperatures of -100 to 1000 °C is described [1]. The implemented techniques comprise magnetron sputtering, ion irradiation, Rutherford backscattering spectrometry, Raman spectroscopy and spectroscopic ellipsometry. The combination of techniques enables sample processing by scalable, clean, waste-free, and industry-relevant technologies, quantitative depth-profiling for elements with Z ≥ 6, structural and chemical characterization, sensitivity and nm-precise thickness and optical information for single layers, multilayers and composites. In this study, the cluster tool was used for i) metal-induced crystallization with layer-exchange of a-Si/ Ag layer stacks, and ii) for high-temperature characterization of two types of solar-selective coatings for concentrated solar power (CSP), namely Al_yTi_1-y(O_xN_1-x)-based single and multilayers [2, 3] and an n-type doped solar-selective transparent conductive oxide [4]. Starting with an a-Si/ Ag bilayer stack, metal-induced silicon crystallization with partial layer exchange occurs at 540 °C. The final stack is approximately described by the sequence crystalline Si (c-Si)/ Ag/ c-Si. All the layers contain minor fractions of the other element. Moreover, the Si volume fraction comprises approximately 10 % of amorphous Si. For the CSP coatings, no compositional and structural changes were found up to a maximum temperature of 840 °C in vacuum. Both types of solar-selective coatings thus represent promising materials for the next generation of CSP technology.
[1] R. Wenisch et al., Anal. Chem. 90, 7837-7824 (2018)
[2] I. Heras et al., Sol. Energy Mat. Solar Cells, 176, 81-92 (2018)
[3] R. Escobar-Galindo et al., Sol. Energy Mat. Solar Cells, 185, 183-191 (2018)
[4] F. Lungwitz et al., submitted (2018)
Financial support by the EU, grant No. 645725, project FRIENDS², and the HGF via the W3 program (S.G.) is gratefully acknowledged.

With theoretical specific capacity of 1672 mAh/g and specific energy of 2600 Wh/kg and practical specific energy of 500 Wh/kg demonstrated recently, values that are highest among all solid elemental redox couples, rechargeable Li-S batteries are regarded as the next step in emerging “green” electrical energy storage technology revolution to replace Li-ion batteries by 2020-2025 for applications in transportation (EVs), national security (UAV, UGV), communications, potable electronics, etc. Produced using abundant, environmentally friendly nontoxic materials, light-weight high-energy density Li-S batteries also offer the possibility of significantly reduced cost. The continuous quest for better, high-performance batteries is currently targeting novel Li-S chemistries, shifting R&D efforts towards battery architectures with optimally engineered functional interphases which offer longer cycle life and high rate capabilities. Despite recent advances, more efforts should be applied to achieve the requisite performance, lifetime and reduced cost needed for new battery technologies to enter the market.

For a long time, practical implementation of Li-S technology has been hindered by poor rate capability and short cycle life due to deleterious effects caused by Li polysulfide redox products. Complex interfacial phenomena and phase transformations inherent to the system require a spatially-resolved instrumental methods capable of characterizing - desirably down to the atomic scale - multicomponent microheterogeneous battery electrodes and their evolution under cycling in situ and ex situ. In this work, we report on multiscale structural and analytical diagnostics of sulfur copolymers, elemental sulfur – MoS₂ composites, carbon nanotube - sulfur and graphene – sulfur - MoS₂ heterostructures as new prospective cathode materials for next generation Li-S batteries by a combination of high spatial resolution electron and ion beam techniques, including advanced multimode spectroscopic STEM imaging and recently introduced low energy scanning focused Li ion beam (LiFIB). Coupled with electrical conductivity, CV, impedance and galvanostatic measurements, this integrated approach enables comprehensive characterization of battery materials and devices using multiple operating modes of FESEM, LiFIB, (HR)S/TEM, electron diffraction, EELS, EDXS, tomography and multivariate statistical phase analysis in relation to their performance. Such measurements form a powerful inter-instrumentation platform for advanced battery research, engineering, and metrology.

Pearl Jam’s hit, “The Light Years,” declares “We were but stones, light made us stars.” Bringing light to the transmission electron microscope (TEM) promises to transform our understanding of photonic materials, enabling direct observation of light-mediated processes with atomic scale resolution. To help achieve this goal, my lab is developing capabilities for concurrent optical and electron microscopy within an environmental TEM. This presentation will describe our efforts to visualize energy-relevant photochemical processes in situ with nanometer-to-atomic scale resolution. In particular, we present our observations of photocatalytic phase transitions in nanoparticles, as well as development of a new technique (electron-and-light-induced stimulated Raman scattering, or ELISR) to monitor molecular transformations in the TEM.

First, we study the gas-phase photocatalytic dehydrogenation of bimetallic nanoparticles, including i) Au-Pd dimer antenna-reactor systems, in which one particle acts as a plasmonic light absorber and the other as the catalyst; and ii) alloyed Ag-Pt bimetallic nanoparticles, where the addition of Ag to Pd induces a visible-frequency plasmonic response. Under controlled hydrogen pressures, temperatures, and illumination conditions, we study the study the kinetics of the desorption reaction triggered by the optical excitation of plasmons. Upon illumination, the system is pushed out of equilibrium, resulting in a Pd phase change from the hydrogen-rich β phase to the hydrogen-dilute α phase. We monitor these changes in real time using a combination of single particle electron diffraction, electron spectroscopy, and direct imaging. We find that plasmons modify the rate of distinct reaction steps with unique time constants, increasing the overall reaction rate more than ten-fold. Additionally, plasmons appear to open a new reaction pathway that is not observed without illumination. Third, reaction nucleation occurs more often at electromagnetic hot-spots. Finally, we find that the local composition of alloyed bimetallic nanoparticles significantly affects the plasmon resonance and reaction kinetics. Full-field electromagnetic and thermal simulations combined with recent ab-initio calculations indicate that this reaction is governed by plasmon-induced population of excited-state hybridized Pd-H orbitals, rather than pure thermal effects.

Next, we present a new technique for monitoring molecular transformations in the TEM: electron-and-light-induced stimulated Raman scattering (ELISR). Unlike conventional stimulated Raman measurements, our technique uses a laser source as the pump and the electron beam as the broadband Stokes excitation. We leverage the electron beam as a highly-localized Angstrom-scale source to locally excite molecular vibrational resonances. Plasmonic nanoparticles amplify the process and enable molecular-scale Raman mapping. Finite difference time domain simulations of molecule-decorated gold nanorods show that laser pump intensities less than 3mW/µm² can locally enhance the Raman peak, with sub-2nm resolution of molecular fingerprints. Experimentally, we synthesize colloidal gold NRs functionalized with 4-Nitrothiophenol (4NTP) as Raman reporter. Upon electron beam excitation of the nanorod plasmon modes, we observe a ten-fold increase in the Raman intensity of the1335cm^-1 and 1570cm^-1 lines of the 4NTP molecule. Monochromated electron energy loss spectroscopy (EELS) confirms the plasmon modes we excite during each individual Raman acquisition. Combined, our results promise high-resolution mapping of ionic and molecular transformations, en-route to site-selective and product specific photocatalysis.

Nanotubes have been observed to have superlative mechanical properties, with promise as a possible structural material. The most demanding applications include lightweight components for space vehicles and radiation shielding. In these environments, carbon nanotubes (CNT) have been shown to be affected by ionizing radiation, which diminishes the desirable properties. Boron nitride nanotubes (BNNT) are an analogous structure with similar mechanical properties to CNT, but with more uniform electronic properties. Here, we present results from beam damage experiments with in-situ environmental transmission electron microscopy (ETEM). Oxygen pressures introduced vary from 10^-5 Pa to 10² Pa and an electron beam dose from 10³ e/nm²-s to 10⁵ e/nm²-s. Our observations show BNNT to have more than one order of magnitude higher resistance to oxidation under the electron beam than that of CNT at comparable oxygen pressures. This demonstrates BNNT as a competitive alternative to CNT with enhanced performance in extreme energetic environments.

In-situ diagnosis appears to be one of the essential methods for gaining insights as how an electrode material failure, therefore feeding back for designing and creating new materials with enhanced battery performances. In this presentation, I will highlight recent progress on ex-situ, in-situ and operando S/TEM and in-situ liquid SIMS for probing into the structural and chemical evolution of energy storage materials. Both ex-situ and In-situ high resolution imaging enables direct observation of structural evolution, phase transformation and their correlation with mass, charge and electron transport, providing insights as how active materials failure during the cyclic charging and discharging of a battery. In perspective, challenges and possible direction for further development of the in-situ S/TEM imaging and spectroscopic methods for energy storage materials and other field will also be discussed. Most importantly, integration of different analytical tools appear to be the key for capturing complementary information.

When exposed in oxidizing environment, most metals and alloys will form surface oxide, which causes problems like corrosion and degradation, but could also be used in applications like catalysts and transistors. With the increasing application demands toward the nanoscale, fundamental understanding of the microscopic mechanisms of surface oxidation is essential for the prevention, prediction and controlled growth of metal oxide. The process of surface oxidation is usually divided into three stages, namely oxygen chemisorption, oxide nucleation and growth, and bulk oxide growth. Among these three stages, the initial stage, which spans from the start of oxygen chemisorption to surface reconstruction and the onset of oxide nucleation, is the least understood regime of oxidation, as it is inaccessible to traditional surface science and bulk material science experimental methods. Current computational capabilities and theoretical modeling are still unable to encompass oxide nucleation due to the large computational expense required to simulate nucleation over sufficiently large time and size scales.

In this work, by combining in situ Environmental TEM (ETEM) and multiscale atomistic simulation, we explore the dynamic processes of the initial stages of Cu(100) oxidation. By in situ oxidizing Cu thin films inside the Hitachi H-9500 ETEM (300 kV, LaB₆, with homebuilt gas delivery system with up to 3 lines of gas injection) with flowing O₂ gas, the initial stages of oxidation were investigated. The dynamic process from surface reconstruction to oxide nucleation and growth was observed, delimiting the initial oxidation stage threshold at which ETEM can effectively discern dynamic processes. Previous papers by Zhou et al reported that the first Cu₂O layer during Cu(100) oxidation grows along the (100) surface from a 2-atomic layer wetting layer. However, our work found the ensuing 3D Cu₂O island grows in a layer-by-layer adatom growth mechanism along the Cu₂O(110) direction. To better understand the observed phenomenon, a multi-scale computational approach was applied. The dynamic and energetic changes during Cu₂O growth under O₂ were calculated. A preferential monolayer formation of Cu₂O monolayer along Cu₂O(110) was found to be the reason. This correlated in situ ETEM and multiscale simulation method forms a feedback loop between ETEM results and simulations that allows computational and experimental results to validate and support one other, and helps to fundamentally explain oxide nucleation preference as function of orientation, defects (such as steps), temperature and pressure, and other environmental factors. Our results will enhance the understanding of surface oxidation and will also provide insights into understanding initial oxide growth mechanisms of other metal/alloy systems.


Acknowledgment
The authors acknowledge funding from National Science Foundation (NSF) grants DMR-1410055, NSF DMR-1508417, and DMR-1410335, as well as support from Hitachi-High-Tech and technical assistance from the Nanoscale Fabrication and Characterization Facility (NFCF) in the Petersen Institute of Nano Science and Engineering (PINSE) at the University of Pittsburgh.

The surface atomic arrangements of reducible oxides govern their oxygen exchange properties, and the ability to control and optimize structural parameters is of crucial importance for many applications, in particular in heterogeneous catalysis and photocatalysis [1]. Characterizing the structural evolution of catalyst nanoparticle surfaces with high spatial and temporal resolution may enable differentiation between functionally-relevant active structures and passive spectator structures, both of which may change during a single reaction “turnover” event. Aberration-corrected transmission electron microscopy (AC-TEM) has been used extensively to study nanoparticle surface dynamics using in situ imaging techniques [2]. Recent advances in imaging technology, namely direct detection cameras with fast acquisition modes, have enabled many new materials phenomena to be investigated. Additionally, with the advent of fast image acquisition, advanced processing techniques are essential for extracting useful information from large, often noisy image datasets. In this work, we use time-resolved AC-TEM to characterize oxygen-exchange induced surface reconstructions on CeO₂ nanoparticles, a reducible oxide that is used extensively in catalysis applications due to its oxygen exchange properties [1].

An aberration-corrected FEI Titan ETEM equipped with a Gatan K2 IS direct detection camera (with high detection quantum efficiency) was used to image CeO₂ nanoparticles at 400 frames/second and 10⁴ e^-Å^-2s^-1 with a pressure of <10^-6 Torr at the sample for ~30 seconds. Due to the fast acquisition rate, large datasets were acquired and individual 1/400 second exposure frames have low signal-to-noise. To reduce noise in each individual frame, images were spatially binned and a Gaussian blurring filter was applied. MIPAR, a commercially available software with a recipe-based image processing method, and custom-written MATLAB codes were used to identify Ce atomic column positions and quantify column intensities in large datasets [3]. The integrated intensity quantization of each atomic column and AC-TEM image simulations were used to determine the number of Ce atoms within each column and track atomic migration at the surfaces. We observe dynamic surface reconstructions and lattice expansions/contractions associated with oxygen vacancy creation and annihilation. Further results providing insight into structure-dependent properties of oxygen exchange reactions on reducible oxides will be presented [4].

References:
[1] Trovarelli, A., and Llorca, J., ACS Catalysis 7 (2017), p. 4716-4735.
[2] Jinschek, J.R., Chemical Communications 50 (2014), p. 2696-2706.
[3] Sosa, J.M., et al, Integrating Materials and Manufacturing Innovation 3 (2014), p. 10.
[4] We gratefully acknowledge support of NSF grant DMR-1308085, the use of ASU’s John M. Cowley Center for High Resolution Electron Microscopy and use of the K2 IS camera courtesy of Gatan.