Symposium CH02—Advances in Cryogenic Transmission Electron Microscopy and Spectroscopy for Quantum and Energy Materials

Solution based synthesis is a common method for production of energy-related materials and a pervasive process in environmental settings tied to energy technology through the impact on carbon emissions and climate. The realization that crystallization in aqueous systems occurs along diverse structural pathways has motivated extensive TEM-based investigations to understand the nature of those pathways, the associated dynamics, and the underlying thermodynamic and kinetic factors. Ex situ TEM studies of such systems suffer from the unknown impact of drying, as well as an inability to directly relate structures seen at one time point with those present at another, rendering an understanding of pathways nearly impossible. For systems amenable to in situ liquid phase TEM, these two problems can be readily circumvented, but the effect of the electron beam through the creation of radiolysis products that alter the solution chemistry and, in the case of organic compounds, the breakdown of the solutes themselves often cast doubt on the applicability of the results to synthetic processes outside the microscope. CryoTEM offers a middle ground that gives a true picture of crystallization at any given time point and allows for the construction of likely pathways per the ergodic principle by analyzing structures present over large areas as a substitute for following individual structures over time. Here we present two examples of crystallization processes for which cryoTEM provides critical information concerning the structural state. The first involves the process of crystal growth through repeated oriented attachment (OA) events, i.e., the growth of single crystals by repeated fusion of nanocrystals that adopt crystallographic coalignment prior to attachment. Using data from minerals aluminum and iron oxide growing by OA, we show that cryogenic high-resolution TEM (HRTEM) can provide definitive information on the temporal evolution of the growing crystals and the structural relationship of the primary particles. The second example involves the crystallization of the water itself. Visualizing the crystalline network of ice with atomic resolution in real space is challenging due to the weak directional hydrogen bonds between water molecules and the resulting instability of ice under an electron beam in vacuo. However, we find that crystallizing water encapsulated between amorphous carbon membranes and freezing it in liquid nitrogen to form hexagonal ice (Ih ice) enables successful HRTEM of the resulting ice sections. Even though the sections show overall single-crystallinity by diffraction standards, the crystal may exhibit structural variation and contain sub-domains ~ 10 nm in size. Quantitative strain and tilt mapping reveals the sub-surface flexibility of ice and is used to demonstrate the correlation between defect structures and strain accumulation. Both sharp low-angle grain boundaries and gradual lattice bending are associated with sub-domains with different crystal orientations. Taken together, this work shows promise in elucidating the molecular and defect structures of ice and may shed light on the physics of ice surfaces and their interfaces with substrates that promote ice nucleation.

Air pollution is becoming one of the most pressing issues because of the increasing levels of pollution around the world. Airborne particulate matter (PM) can trigger cellular oxidative stress and inflammation, which are the basis for respiratory diseases associated with pollution. Understanding the effects of PM pollution on respiratory system cells is key to develop strategies to reduce personal exposure, especially in congested areas. Currently, there is no correlated cell-level information about the effects of PM on primary human epithelial cells metabolism, the structural damage of cellular organelles and the size and chemistry fractions of PM that cause this cellular damage. In a collaboration with the National Heart and Lung Institute, we have been focusing on the impacts of pollution PM collected from a London Underground platform on the health of cell organelles and cellular metabolism. We have shown that the PM collected from the London Underground is composed of trace quantities of ultrafine particles composed of redox-active transition metals, including Cu, Cr, Fe and Mo that could damage respiratory cells by generating damaging free radicals¹. We have exposed human nasal epithelial cells from healthy and asthmatic volunteers to this pollution PM. We have chemically fixed and stained the cells, embedded them in a resin and cut ultrathin sections for transmission electron microscopy and analysed the chemistry of intracellular nanoparticles of pollution using high resolution electron microscopy. In parallel, cryo soft X-ray tomography (cryo SXT) with structured illumination microscopy (cryo-SIM) (B24 Beamline, DLS, UK) was employed to generate whole-cell structural and metabolically relevant information about the cells (cell lines and human cells) exposed to PM in their near-native state. I will discuss the changes observed to ultrastructure and metabolism of cell organelles in cells exposed to London Underground pollution particles and the challenges of preparing and imaging human cells for correlative cryoSXT, cryo-SIM and cryo electron microscopy. I will also discuss how optimisation of cryo-focussed ion beam milling scanning electron microscopy (FIB-SEM) techniques could be used subsequently in specific areas for higher resolution imaging and complementary compositional analysis, as well as for lamella preparation for high-resolution cryo-TEM, where oxidation states and composition of damaging PM can be attained. In future, the development of this multidimensional workflow using human derived cells will bring together imaging of whole-cell structure and metabolism with chemical analysis of the PM to identify which chemistries are most damaging to intracellular cell organelles and their mechanisms of toxicity. This novel approach will offer new insights into how to mitigate the health effects of pollution PM.

^1. Kumar P, Chung KF, Porter AEc et al, 2022, Characteristics of fine and ultrafine aerosols in the London underground., Sci Total Environ, Vol: 858

In recent years, all-polymer solar cell performance has accelerated, in part due to significant improvements in controlling microstructure by optimizing processing parameters. However, quantitative relationships between processing, microstructure, and performance remain weak. While charge generation in all-polymer blends is highly dependent on local chain orientation and local stacking structure, common microstructural characterization techniques only provide bulk average characteristics of film morphology. On the other hand, cryogenic High Resolution Transmission Electron Microscopy (cryo-HRTEM) has emerged as a promising method to image and quantify local polymer structure at sub-nanoscale resolutions.

In this study, we use cryo-HRTEM to analyze the structural organization of the fluorinated all-polymer blend PBDB-T-2F:F-N2200, and we show that donor microstructure can be selectively tuned with processing additives. By applying Fourier analysis and statistical methods, we map the local orientation of lamellar stacking, characterize donor/acceptor domains and interfaces, and quantify inter- and intra-species chain alignment. The results demonstrate how advances in quantitative analysis of cryo-HRTEM images can reveal new structural characteristics in polymer blends.

Organic mixed ionic–electronic conductors (OMIECs) are soft organic materials that are advantageous for varied bioelectronic devices, ranging from organic electrochemical transistors (OECTs) and biosensors to batteries and supercapacitors.¹ Their ability to operate in aqueous environments is an important virtue that qualifies them for these applications. One widely used successful strategy for obtaining OMIECs with enhanced water and ion uptake and hence increased capacitance, is placing ethylene glycol side chains on the polymer backbone.² However, the structural understanding of these glycolated OMIECs is still limited, particularly in operational states such as being hydrated in water, or used in aqueous electrolytes, or when electrochemically doped. One significant challenge in filling this gap is the challenge in characterizing hydrated or solvent swollen OMIECs, that is incompatible with many standard relevant techniques that involve high vacuum conditions.³ Here we describe experiments where we used cryogenic vitrified solid-state samples and advanced scanning nanobeam diffraction (4D-STEM). We have previously applied 4D-STEM to study these materials in the dry solid-state, mining structural information from the molecular to the mesoscale level, providing a baseline and analysis pipeline for the more complex states. Applying this hybrid approach, we demonstrate the structural changes induced in glycolated OMIECs by hydration. The most prominent change is the lamellar expansion. Noteworthy, this is not an obviously predicted structural change, and theoretical simulation even predicted a lamellar contraction.⁴ We believe this methodology can bring us one step closer to a deeper understanding of the structures of these materials in their operational modes. Our results show how cryogenic 4D-STEM can provide true structure-function relationships that can help improve our rational design of OMIECs thin films for more efficient devices.

____________________________________________________________
1. Nat. Mater. 2020,19, 13–26.
2. (a) Chem. Mater. 2018, 30, 9, 2945–2953. (b) Energy Environ. Sci., 2019,12, 1349-1357. (c) PNAS. 2016, 113, 12017.
3. Chem. Rev. 2022, 122, 4, 4493–4551.
4. Adv.Mater.2022, 34, 2204258.

The development of cryogenic electron microscopy started in the 1970s in order to study the structure of proteins and other biological material. However, only recent advances in detector technology and software algorithms have allowed the determination of biomolecular structure at atomic resolution [1]. The application of cryogenic transmission electron microscopy (cryo-TEM) methods to materials science and energy-related fields has been mainly limited to the study of beam sensitive materials such as lithium-ion batteries, organic semiconductors, perovskite-based solar cells, polymers and metal organic frameworks (MOFs). Most low-temperature in situ TEM experiments are carried out using cryogenic holders based on liquid nitrogen cooling and using 3- mm-sized TEM support grids. Despite their widespread use, grid-based cryogenic systems suffer from poor stability, making it challenging and time-consuming to obtain adequate atomic-resolution imaging [2]. Additionally, these systems often have the inability to set intermediate temperatures, which limits the experiments to be performed at either liquid nitrogen temperature or room temperature. Furthermore, such systems are limited in applications when understanding the structure, electronic and transport properties of materials under an applied electrical stimuli at low temperatures is of interest. For such applications, a system with low sample drift that combines liquid nitrogen cooling with an external voltage bias is needed. Such a development can enable a vast range of applications in the field of quantum materials [2], magnetic materials and nanostructures, ferroelectrics [3], topological insulators, metal-to-insulator transitions, superconductors.
In this talk, we will share our latest developments with respect to a combined in situ cooling and biasing system. The system includes a novel double-tilt cryogenic holder which is based on micro-electro-mechanical-systems (MEMS) technology, and uses liquid nitrogen for cooling. It has multiple electrical contacts and is compatible with the DENSsolutions heating and biasing Lightning Nano-Chips, enabling the setting of any intermediate temperature between LN2 and room temperature. Without the presence of the cooling agent, it is also possible to perform in situ heating experiments up to 1000oC. By exploiting the high stability of this system and its double tilt capability, we will demonstrate how atomic resolution imaging can be achieved while applying a bias to the sample and present a number of application examples, including ferroelectric focused ion beam (FIB) lamellas.

[1] Cheng Y et al, A primer to single-particle cryo-electron microscopy. Cell 2015 161 (3) (438).
[2] E. Tyukalova et al, Challenges and Applications to Operando and In situ TEMimaging and spectroscopic capabilities in a cryogenic temperature range. Accounts of Chemical Research 2021 54(16) (3125).
[3] JL Hart and JJ Cha, Seeing Quantum Materials with Cryogentic Transmission Electron Microscopy. Nano Lett 2021 21 (5449).
[4] J. Mun et al, In situ Cryogenic HAADF-STEM observation of spontaneous tranistion of ferroelectric polarization domain structure at low temperatures. Nano Lett. 2021 13 (8679).

Quantum materials experience novel phenomena at low temperatures, either developing new structural or ferroelectric phases or new properties such as superconductivity and spin liquids. Because of its ability to simultaneously derive complex information about a material’s electric field, charge density, and magnetic structures, four-dimensional scanning transmission electron microscopy (4D-STEM) has gained favor in cryogenic experiments to investigate a material's temperature dependent properties and phase transitions. However, 4D-STEM detectors usually require a per-pixel acquisition time in the region of a millisecond, resulting in a relatively lengthy acquisition time and potential sample damage due to excessive electron dose. One possible strategy is to only capture the probe positions required to recover the desired information. If a proper sampling method is taken and the data at redundant sites can be reconstructed using an algorithm, the total acquisition time and electron dose could be substantially reduced.
In this work, we developed a machine learning (ML) based compressive sensing algorithm to reconstruct 4D-STEM datasets from a limited subset of the available data. Our ML model consists of an autoencoder that extracts a low-dimensional feature space from the limited subset of the diffraction patterns, and a fully connected neural network to map from the integrated images, e.g., bright field, center of mass, to the extracted feature space. Once the neural networks are trained, for each missing pixel, i.e., the pixel that is not scanned during the acquisition, we can use compressive sensing to interpolate the pixel values in the integrated images, map the interpolated values to the feature space via the fully connected network, and then push the feature sample through the decoder to construct the missing diffraction pattern. [1]
Our 4D-STEM compressive sensing algorithm has been successfully implemented on a pair of datasets including different features in real space, including periodical atomic structural imaging as well as samples including interfaces and defects. Using this algorithm, we demonstrate significantly reduced sampling points (down to 25%) is sufficient to reconstruct STEM images retaining atomic-scale structural information.

Acknowledgments
[1] This research is supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research, the Scientific Machine Learning Program, and performed at Oak Ridge National Laboratory (ORNL). ORNL is operated by UT-Battelle, LLC, for the U.S. Department of Energy under Contract DE-AC05-00OR22725.
[2] The microscopy work was supported by an Early Career project supported by DOE Office of Science FWP #ERKCZ55–KC040304. All microscopy technique development was performed and supported by Oak Ridge National Laboratory’s (ORNL) Center for Nanophase Materials Sciences (CNMS), which is a DOE Office of Science User Facility.

Transition metal trihalides are two-dimensional van der Waals interlayer-bonded quantum materials with complex phase transitions and intrinsic magnetic orderings at lower temperatures. For instance, RuCl₃ stirred up considerable interest due to its potential as a spin-1/2 honeycomb-lattice magnet, which is predicted to have characteristics of a spin liquid [1]. For these 2D materials, shifts in interlayer stacking can cause distinct transitions in magnetic properties [2]. However, atomic scale understanding of variations in layer-by-layer stacking arrangements of metal trihalides at relevant temperatures have yet to be empirically explored. Scanning transmission electron microscopy (STEM) is an exemplary technique to visualize changes in stacking structures. We select CrCl₃ as a model material as it is a relatively stable metal trihalide with an observed bulk phase transition at 240 K and ferromagnetic characteristics between 14 K and 17 K [3]. Cryogenic-STEM techniques are employed to examine low-dimensional CrCl₃ samples at varying thicknesses and temperatures from room temperature to 103 K. Unexpected stacking shifts deviating from expected phase transitions in the bulk are directly observed through high annular dark field (HAADF) imaging. Implications for these structural changes will be further discussed. [4]

References:
[1] Y Kubota, H Tanaka, T Ono, Y Narumi, and K Kindo. Phys. Rev. B 91 (2015), p. 094422. DOI: 10.1103/PhysRevB.91.094422.
[2] Y Xu, A Ray, Y Shao, S Jiang, K Lee, D Weber, J E Goldberger, K Watanabe, T Taniguchi, D A Muller, K F Mak, and J Shan. Nat. Nanotechnology 17 (2022), p. 143–147. DOI: 10.1038/s41565-021-01014-y
[3] M A McGuire, G Clark, Santosh K C, W M Chance, G E Jellison, Jr., V R Cooper, X Xu, and B C Sales. Phys. Rev. Materials 1 (2017), p. 014001 DOI: 10.1103/PhysRevMaterials.1.014001
[4] Microscopy research was performed at the Center for Nanophase Materials Sciences at Oak Ridge National Laboratory, which is a US Department of Energy (DOE), Office of Science User Facility.

An interfacial understanding is necessary for developing strategies to commercialize high-energy density rechargeable lithium metal anode batteries, as currently, the lithium anode/electrolyte interface is unstable with prolonged cycling. We have used several strategies to improve the cycling performance of lithium metal anodes, including reducing the parasitic reactions between lithium metal and the electrolyte, and improving the electrodeposited lithium metal morphology. These strategies have generated unconclusive electrochemical data, that has required the need for nanoscale interfacial characterization of these solid-liquid interfaces. Our team has used the cryogenic transfer workflow developed by Leica in collaboration with cryo-SEM/FIB tools by Thermo Fisher Scientific to cross-section lithium metal anodes and intact coin cell batteries to observe the interfacial structures, lithium morphology, and failure mechanisms relative to changes in electrode contract pressure, electrode coatings, and electrolyte chemistry. Cross-sectional SEM images and EDS maps of the lithium metal anodes have provided a better understanding of the electrodeposited lithium morphology, quantity of ‘dead’ lithium metal, and quantity of solid electrolyte interphase material that has formed alongside the lithium metal. In understanding lithium metal battery failure at the system level, we used a cryogenic stage in a laser plasma FIB to cross-section through the coin cell’s cap for imaging/mapping the entire battery stack under cryogenic conditions. The tools, methods, and results of these studies will be detailed in this presentation.

The great majority of detectors currently used in electron microscopy are frame-based. Each time data is read from the sensor, a "frame" is generated by reading out all of the pixels on the sensor, or in a region of the sensor. In applications involving sparse signals, such as electron counting for imaging in cryogenic transmission electron microscopy, frame-based detectors are inefficient, as sparsity means that the majority of each frame does not contain signal, and therefore reading out frames of pixels involves large volumes of redundant data. In contrast, event-based detectors are an alternative technolgy in which pixels are read out asynchronously when they register a significant change in intensity. When event-based readout is employed on a direct detector with a very high single-electron signal to noise ratio, the detector can be designed such that pixels are only read out when they are excited by an incident electron. This strategy avoids the redundancy of reading out large numbers of pixels that contain no signal, and as a result, event-based direct detectors promise higher electron counting performance at a lower cost than frame-based detectors. In this presentation, we will review the development of event-based direct detectors, and their current and potential applications in transmission electron microscopy.

An imaging process can exert an inadvertent effect on the object being observed. Consequently, what we observe does not necessarily represent what had been present before the observation. Normally, this effect can be ignored if the consequence of such a change is believed not to be significant. Electron beam based imaging and spectroscopic techniques appear to be one of the indispensable approaches for probing the characteristics of rechargeable batteries. Modification of battery materials by the electron beam during their imaging has been widely noticed, which, for some cases, appears to be not amendable for capturing any useful information. This is especially true for the case of dynamically formed entity during battery cycling, such as solid electrolyte interphase (SEI) layer, a thin layer formed on the electrode surface that is essential for electrochemical reactions in batteries and critically governs the battery stability. Recently, it has been demonstrated that, cooling of the battery sample, as termed as cryo-microscopy for materials science, can effectively mitigate the imaging electron induced modification effect, therefore to allow the capturing of structural features of beam sensitive battery materials at multi-scale down to atomic resolution. In this presentation, we will check into the latest advances of cryo-TEM and its application for probing into the structural and chemical evolution of rechargeable batteries, typically in the system of Si and Li anode and their SEI layers. In general, the pros and cons of cryo-TEM for material science will also be discussed.

Atomic resolution imaging of soft materials using electron microscopy is challenging because of the radiation damage under the electron beam. By leveraging low-dose cryogenic transmission electron microscopy (cryo-TEM) and the sophisticated image processing approaches developed by the structural biology community, atomic-scale structural information can be obtained from the nanostructures formed by synthetic soft materials. Our experiments were conducted on crystalline nanosheets and nanofibrils formed by self-assembly of amphiphilic polypeptoid molecules. Low-dose cryo-TEM micrographs were obtained from vitrified nanosheets and nanofibrils. A combination of electron crystallographic and single particle analysis was used to obtain two-dimensional (2D) and three-dimensional (3D) high resolution images of crystal motifs and single molecules. 2D atomic-scale imaging reveals the importance of halogen bonds in the design of crystal motifs in self-assembled nanostructures. Moreover, the hidden lattice symmetry in halogenated nanosheets is clearly illustrated in 3D density map. In addition, 3D reconstruction of nanofibrils shows the direct view of chain conformation, the unique internal structures comprising two opposite layers and the effect of capping group interactions on the design of nanostructures. These discoveries thus allow us to explore the effect on inter/intra interactions on the self-assembled nanostructures at the atomic level.

Lithium-ion batteries (LIBs) commercially dominate portable energy storage and have been extended to hybrid/electric vehicles by utilizing electrode materials with enhanced energy density. The energy density and cycling life of LIBs must extend beyond the current reach of commercial electrodes to meet the performance requirements for transportation applications. Recently, researchers proposed to use Li metal as the anode to achieve higher energy density. However, the dendritic growth of Li metal during cycling will result in low coulombic efficiency as well as safety issue which is detrimental for practical applications. It is proposed solid state electrolytes (SSE) are compatible with lithium metal because of lithium penetration resistance, which can be the ultimate choice to achieve the energy density of 500 Wh/kg. The performance of existing all-solid-state batteries is still not optimal, which is mainly attributed to the poor interfacial stability. Further improvements and developments of these energy storage materials rely on a fundamental understanding of their electrochemical cycling mechanisms at atomic level. Characterization of Li anodes and SSE/Li-metal interfaces remains difficult due to their sensitivity to beam damage, the nano-scale of interfacial decomposition products, and their buried nature. Our group has demonstrated the importance of cryogenic techniques for preparation (cryogenic focused ion beam, cryo-FIB) and characterization (cryogenic transmission electron microscopy, cryo-TEM), to preserve the morphology and structure of lithium metal. These breakthroughs enabled unprecedented characterization of plated Li metal and are currently being extended to all-solid-state batteries.

The first demonstration of the functionality of the cryo-TEM will focus on characterizing the morphology and crystallinity of electrochemically cycled lithium metal with different electrolytes. This work will also cover recent advances on cryo-EM development for Li/LiPON/LNMO interfaces. The combination of cryo-FIB and cryo-EM is necessary for quantitative structural and chemical analysis due to extreme susceptibility of both lithium metal and LiPON to air and beam damage. The obtained results provide new insights of decomposition, diffusion of chemical species, and chemical evolution along the interface, which leads to a better understanding of the cycling stability for Li metal batteries.

Cryo-EM received the 2017 Nobel Prize in Chemistry for its ability to elucidate the nanostructure of biomolecules in their native state, revolutionizing the field of structural biology. By leveraging this powerful technique to study battery materials, we revealed the atomic structure of reactive battery materials and interfaces for the first time (Science 358, 506–510, 2017), demonstrating its potential impact for battery research. Indeed, one of the most important yet unanswered questions in battery research still remains: what are the structures and chemistries present across liquid-solid battery interfaces and how do they evolve with time? To bridge this gap in understanding, we adopt and innovate cryo-EM techniques that can resolve these sensitive liquid-solid interfaces and correlate them with battery performance. Here, we further innovate cryo-EM techniques towards battery research by leveraging a thin film vitrification method to preserve the sensitive battery interfaces in their native liquid electrolyte environments for high resolution imaging and spectroscopy. Our preliminary data reveals a significant swelling ratio between dry and wet states of the solid electrolyte interphase (SEI), a ubiquitous yet poorly understood corrosion film formed on the surface of battery anodes (Science 375, 66, 2022). This surprising discovery highlights the promising potential of cryo-EM to enable new fundamental understanding and insights for materials research. <div id="accelSnackbar" style="left: 40px; top: 40px;"> </div>

The formation of a cathode electrolyte interphase (CEI) during initial cycles plays a crucial role in stabilizing the performance of Li-ion cathode materials. However, understanding the oxidation mechanism and surface evolution of the CEI remains elusive due to the lack of reliable in situ surface analysis tools and the much smaller dimension of the CEI than that of the anode counterpart. Capturing the formation of metastable and/or intermediate species formed during the initial charge will provide a critical understanding of the oxidation pathway of electrolyte components. Cryogenic X-ray Photoelectron Spectroscopy (cryo-XPS) is a powerful and precise surface technique to analyze the interfacial reaction at a solid-liquid interface and the low temperatures also prevent decomposition of unstable species under high temperature and ultrahigh vacuum conditions. In this study, we used cryo-XPS to monitor the CEI evolution of a lithium iron phosphate (LFP) cathode by monitoring the surface chemistry of LFP at various states of charge/discharge during the first cycle. The result showed that the decomposition of lithium hexafluorophosphate (LiPF₆) occurred as early as 3.4 V vs. Li/Li⁺ as evidenced by the presence of its decomposition products. Interestingly, the concentration of decomposition products changed with increased cell voltage, which is consistent with the unusual strain evolution of the LFP electrode cycled in LiPF₆ electrolyte obtained via in situ strain measurement of the LFP electrode. Through this study, we aim to provide a comprehensive picture of the evolution of the CEI and its correlation to the strain of electrodes, which is important for controlling the CEI formation during the formation process in manufacturing automotive lithium-ion cells and better design of future electrolyte components.

Accessing the atomic scale structural and electronic properties is crucial to understand fundamental correlated properties in quantum materials. In recent times, analytical electron microscopy techniques such as scanning transmission electron microscopy (STEM)-electron energy loss spectroscopy (EELS) under cryogenic conditions have been applied to image local bosonic states such as excitons and phonons. My talk will be focused on probing these bosonic states at low temperatures. I would start my talk with moiré two-dimensional transition metal dichalcogenides which have correlated electronic states resulting from atomically varying strains associated with the stacking configuration. I will talk about the recent progress that we have made in cryogenic imaging excitons in moiré two-dimensional transition metal dichalcogenide heterostructures at the atomic scale from barely detecting the moiré exciton signal in EELS [1] to imaging these excitonic states at the atomic scale [2] to recognizing the fine structure of the moiré excitons [3]. I would also elaborate on the future cryogenic experiments that could be performed on these systems. In the second part, I will take the example of a thin film PbTiO₃ layer grown on DyScO₃ substrate [4] and explain how the phonon properties vary near the domain walls. I hope to end my talk by provinding insights about future for cryogenic electron spectroscopy to be successful in accessing fundamental structural and electronic states in quantum materials with some of the pathways that our group is adopting to solve this problem.

[1] S. Susarla, L. M. Sassi, A. Zobelli, S. Y. Woo, L. H. G. Tizei, O. Stéphan, and P. M. Ajayan, Nano Lett. 21, 4071 (2021).
[2] S. Susarla, M. H. Naik, D. D. Blach, J. Zipfel, T. Taniguchi, K. Watanabe, L. Huang, R. Ramesh, F. H. da Jornada, and S. G. Louie, ArXiv Prepr. ArXiv2207.13823 (2022).
[3] M. Dandu, J. Hachtel, M. H. Naik, S. Sankar, T. Taniguchi, K. Watanabe, S. G. Louie, F. Jornada, P. Ercius, A. Raja, and S. Susarla, (In Prep.).
[4] S. Susarla, (In Prep).

Metal–organic frameworks (MOFs) are crystalline coordination networks composed of metal ions connected by organic linkers. One of the most prominent advantages of MOFs originates from interactions of the framework with small molecules; these are referred to as host-guest interactions. Visualizing interactions between host MOFs and guest molecules with cryoTEM enables us to explore their structure under different conditions. For example, it is possible to compare partial loading and visualize the nanostructural morphology of individual crystals, and ultimately optimize loading of MOFs.¹ Currently, host-guest interactions in MOFs are either understood via theoretical simulations or bulk characterizations using x-ray/neutron diffraction, in situ diffuse reflectance infrared spectroscopy, or nuclear magnetic resonance.² These experimental techniques average out information such as heterogeneity among adsorption sites.

MOFs are inherently electron beam-sensitive and typical electron irradiation causes collapse of the structure within seconds.¹ Since gas adsorption in a MOF is usually optimized to be reversible (so that the MOF can be cycled and reused), the host-guest interaction is ideally somewhat weak. At atmospheric pressure, small gas molecules and any other volatile solvent easily desorbs from the material. The high vacuum in the TEM further promotes desorption.²

For this project, a novel, air-free workflow to activate and gas dose metal-organic frameworks on a TEM grid has been devised. This is followed by cryogenic transfers that ensure that the gas molecules remain adsorbed in the pores of the material for imaging in the TEM.

This workflow was used on Co(bdp) (bdp²⁻ = 1,4-benzenedipyrazolate), a MOF in which adsorption induces a structural phase transition. Scanning nanodiffraction or “4D-STEM”³ was employed to map the change in lattice parameters upon desorption. First, 4D-STEM at cryogenic temperature was used to image several crystals of CO2-dosed Co(bdp). To induce a phase transition, the sample was then heated to room temperature, and 4D-STEM was carried out on the same crystals. A change in lattice spacing was observed after warming, indicating the release of CO2 from the framework. Measured lattice spacings around 13 Angstroms suggest the framework itself, rather than dry ice, is being imaged. In the future, larger ensembles of crystals will be scanned to investigate the heterogeneity of adsorption, and the methodology will be used on an air-stable, industrially relevant framework that is isostructural to Co(bdp).

Sources
1. Gong, X.; Gnanasekaran, K.; Chen, Z.; Robison, L.; Wasson, M. C.; Bentz, K. C.; Cohen, S. M.; Farha, O. K.; Gianneschi, N. C. J. Am. Chem. Soc. 2020, 142, 17228.
2. Li, Y.; Wang, K.; Zhou, W.; Li, Y.; Vila, R.; Huang, W.; Wang, H.; Chen, G.; Wu, G.-H.; 2. Tsao, Y.; Wang, H.; Sinclair, R.; Chiu, W.; Cui, Y. Matter 2019, 1, 428.
3. Ophus, C. Microsc. Microanal. 2019, 25, 563.

Polymers that dissolve lithium ions are promising electrolyte materials for the next generation of rechargeable batteries. Both high mechanical stability and high ionic conductivity can be achieved through microphase separated block copolymer electrolytes by using one phase to conduct ions while the other provides rigidity[1]. The ion transport properties of such composite materials are not easily predicted solely from ion transport properties of the conducting domain alone [2], as the microphase and morphology play an important role in electrochemical performance [3, 4]. In previous research, heavy element staining bright field imaging and Energy-filtered transmission electron microscopy spectrum-imaging (EFTEM-SI) have been used to study PS-b-PEO [1, 2, 5]. Here, we utilize cryo-4D-STEM (four-dimensional scanning transmission electron microscopy) to investigate both the polymer morphology as well as the crystallinity as a function of Li concentration. 4D-STEM is a technique by which a nearly parallel electron nanoprobe is rastered across a sample (two dimensions in real space), recording the diffraction pattern (two dimensions in reciprocal space) at each scan point.[6] The diffraction patterns acquired form a 4D dataset provide real space information about the phase distribution and orientation of the polymer materials. It is a powerful tool to unveil the microphase of PS-b-PEO and provide further information about how the structure effects electrochemical performance.[2, 3]

Adding lithium salt into PS-b-PEO is a crucial step for real applications. Here we used mixtures of poly (styrene-block-ethylene oxide) copolymers (SEO) and bis(trifluoromethane)sulfonimide lithium salt (LiTFSI). By utilizing an air-free glove box and a cryo-transfer holder, we have established a workflow that avoids air exposure to these air-sensitive samples. 4D-STEM was later applied to show how the phase and DPs change when lithium concentrations change from r = 0 to 0.1. We calculated and carefully controlled the beam damage and successfully produced 4D-STEM results in cryo-condition. Compared with traditional RuO₄ staining 4D-STEM bright field imaging, 4D-STEM is shown to produce similar phase information without using toxic chemicals. Furthermore, 4D-STEM can provide additional phase distribution and orientation information inside the crystalline phase. The relation between the electrochemical performance and phase distribution, in relation to the crystallinity with and without the Li salt will also discussed.

1. Enrique D. Gomez, Ashoutosh Panday, Edward H. Feng, Vincent Chen, Gregory M. Stone, Andrew M. Minor, Christian Kisielowski, Kenneth H. Downing, Oleg Borodin, Grant D. Smith, and Nitash P. Balsara*, Effect of Ion Distribution on Conductivity of Block Copolymer Electrolytes. Nano letters, 2009.
2. Maslyn, J.A., et al., Limiting Current in Nanostructured Block Copolymer Electrolytes. Macromolecules, 2021. 54(9): p. 4010-4022.
3. Galluzzo, M.D., et al., Measurement of Three Transport Coefficients and the Thermodynamic Factor in Block Copolymer Electrolytes with Different Morphologies. J Phys Chem B, 2020. 124(5): p. 921-935.
4. Loo, W.S., et al., Phase Behavior of Mixtures of Block Copolymers and a Lithium Salt. J Phys Chem B, 2018. 122(33): p. 8065-8074.
5. Allen, F.I., et al., Deciphering the three-dimensional morphology of free-standing block copolymer thin films by transmission electron microscopy. Micron, 2013. 44: p. 442-50.
6. Bustillo, K.C., et al., 4D-STEM of Beam-Sensitive Materials. Acc Chem Res, 2021. 54(11): p. 2543-2551.

Membrane-based gas separation requires highly delicate engineering of molecular structures for the desired separation performance. For instance, natural gas processing including CO₂/CH₄ and N₂/CH₄ separations or olefin/paraffin separation involves a minimal molecular size difference of less than 0.5 Å. However, the extent of improvement in separation performance via engineering the chemical structure of polymeric membranes is very limited due to the trade-off relationship between permeability and selectivity. Moreover, most polymeric membranes are susceptible to condensable gas-induced plasticization at high pressures, suffering from significant selectivity loss. Carbon molecular sieve (CMS) membranes are promising molecular sieving candidates to overcome the intrinsic drawbacks of polymeric membranes including the limited separation performance and weak plasticization resistance. Recently, our group proposed a new approach to engineering the selective micropore structure of CMS membranes for various gas separations, including CO₂/CH₄, N₂/CH₄, or C₃H₆/C₃H₈ separation by pyrolyzing homogeneous blends of polyimide and ladder-structured polysilsesquioxanes. In particular, the SiOx phases in a carbonaceous matrix contributed to enhancing the molecular sieving ability of the resultant carbon-silica hybrid microporous membranes for various gas separations. In addition, our newly proposed approach enabled the formation of asymmetric CMS hollow fiber membranes with an excellent separation performance owing to the enhanced thermal stress resistance.

Observing materials at atomic resolution gives us insight into how these systems operate and their failure mechanisms. This understanding is required to develop clean energy technology for a sustainable future on earth. In most cases, the key materials of the system under study (e.g., lithium batteries, perovskite solar cells, and MOFs/COFs) are sensitive to moisture, air, and electron beam irradiation, thus making it quite challenging, and in some cases, even impossible to succeed if conventional S/TEM methods are applied.
Using results from various materials, we will review some of these challenges. We will highlight how recent developments at Gatan in specimen holder (cryo-transfer vacuum holder), detection technology (high-speed electron counting direct detectors for imaging, diffraction, and EELS), and software (in-situ and multi-modal data acquisition and processing) enable the observation of such beam-sensitive materials at low dose, low temperature, and low accelerating voltage.

Cryogenic transmission electron microscopy allows for the observation of samples in near native conditions and for higher electron beam tolerances than room temperature observations. Crucially, it provides the opportunity of obtaining quantified information from solid-liquid interfaces. In practice, accessing the solid-liquid interface at high-resolution represents specific challenges that are sample dependent and that must be addressed. Typically, sample preparation for cryo-(S)TEM is central and as a matter of fact, a significant bottleneck in many experiments. For the case of hard-soft interfaces in hydrated systems there are additional challenges which arise by the mismatch between the bio(chemical), mechanical and physical properties of both sides of the interface. Imaging conditions that are suitable for the different electron scattering produced by the different materials must be found and there can be limitations by the fact that a specific site within the sample (with a specific orientation) must be accessed. In this presentation, I will show our latest experimental results and discuss the specific challenges met on the application of 3D Cryo-FIB/SEM to the study of hard-soft tissue interfaces, to probe biomolecules fouled in polymeric filtration membranes and for the analysis of biomaterials interfaces. One specific challenge that is common to all experiments in hydrated samples is to understand the effect of the electron beam in our observations to disregard possible artefacts. Despite the more limited diffusion of secondary species at cryogenic temperatures, the high-energy electron beam used to probe the sample will unavoidably induces chemical processes due to radiolysis. Methods to directly probe the effects of radiation chemistry inside the EM have not been developed yet. In this talk, I will also present our investigations combining scanning transmission electron microscopy (STEM) and electron energy loss spectroscopy (EELS) on ice and discuss the effects of high-dose rate electron irradiation on water and on water-solid interfaces in the EM.

In the past decade, the rapid growth of electric vehicles and consumer electronics market leads Li-ion batteries to attract significant attention. In order to further advance their performance for higher energy and better safety, fundamental understanding of battery materials structures and chemistry is essential. Nowadays, in order to pursue higher battery performance, more and more materials used in battery are beam sensitive and inherently air sensitive. This brings challenges to preserve their native structure and property via routine electron microscope practice.
Low-dose EELS STEM analysis carried out at cryogenic temperature on a Krios (S)TEM under cryogenic conditions has proven to be a very successful approach to study the solid electrolyte interface (SEI) in the Li metal anode region. The main challenges that until recently has limited the accurate investigation of SEI samples are the extreme sensitivity to air and probing sources such as the electron beam. However, by combining the cryo transfer approaches developed for life science tools and the extra stability of the dedicated cryo-stage in the Krios, it was possible to successfully carry out, morphological, chemical and tomography studies of the solid electrolyte interface region in Li-anodes [1].
Motivated by cryogenic electron microscope (Cryo-EM) technology in structural biology, the application and challenges of cryo-EM in battery research is discussed here.

References:
[1] B. Han, X. Li, S. Bai, Y. Zhou, B. Lu, M. Zhang, Z. Ma, Z. Chang, Y. S. Meng and M. Gu, Matter 4, 1-12, November 3, 2021

Limited knowledge of the chemistry and time dependent formation of Ca or Mg solid electrolyte interphases exists as current tools do not combine surface sensitivity, high spatial resolution, chemical specificity and preservation of the electrolyte environment. In response to this challenge, we have developed and demonstrated cryogenic transmission electron microscopy (cryo-TEM) combining the above attributes to determine the identity and properties of anode interphases that enable long cycle lifetimes in emergent Mg and Ca metal batteries. We find that a nanometric, heterogeneous oxide, not a hydride as previously thought, is the interphase responsible for reversible Ca cycling in a calcium borohydride - tetrahydrofuran electrolyte. Ca²⁺ conductivity in such a ubiquitous material, specifically along oxide, borate, and carbonate phase boundaries, raise the possibility of controlling SEI formation to support working cation conductivity in a broader class of chemistries. We demonstrate a strong dependence of SEI composition and properties on Ca and Mg cation solvation structure, highlighting the possibility of directing reactivity by controlling speciation within the electrolyte. In this presentation, we explore this concept of directing reactivity and correlate outcomes with SEI performance.

Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

The Imperial Centre for Cryo Microscopy of Materials (I(CM)²) is a new facility at Imperial College London for the development and application of cryo microscopy to environmentally sensitive materials. The facility consists of the three state of the art microscopes (Cameca Local Electrode Atom Probe 5000 XR, FEI-Thermofisher Scientific Helios Hydra DualBeam with cryo-stage and FEI-Thermofisher Scientific Spectra 300 with cryo-holders) connected via an inert gas glovebox with cryo-vacuum transfer capability. By combining the complementary capabilities of these instruments, nanoscale compositional and structural analysis on environmentally sensitive materials of interest to a wide range of fields can be achieved.
Cryogenic focused ion beam (cryo-FIB) sample preparation for site specific liftout has significant challenges compared to the standard room temperature process flow. A number of applications that require cryo-FIB have successfully bypassed the requirement for liftout such as on-grid thinning for cryo electron microscopy (cryo-EM) [1] or “satellite dish” samples for Atom Probe Tomography (APT) [2]. However, there are many number of situations in life sciences and materials science where the cryo-FIB liftout stage is essential and so reliable and reproducible processes to enable this to become routine are required. Complementary Scanning Transmission Electron Microscopy (STEM) and APT analysis on the same sample [3] is increasingly commonplace and the application of this to materials requiring cryo-FIB is an emerging field of interest. This requires samples to be fabricated and mounted in a manner that is appropriate for the specific requirements for high yield and data quality from both techniques.
Room temperature FIB liftout generally uses site specific decomposition of an organometallic precursor gas supplied by a Gas Injection System (GIS) to provide protective layers and to attach liftout lamella to micromanipulators and support structures such as grids. If this gas is released into a FIB chamber with a cryogenically cooled stage then the gas immediately condenses. The thickness of this condensate is hard to control, it requires “curing” via an electron or ion beam to become conductive and using it significantly increases the difficulty of the liftout process [4]. GIS-free cryo-FIB liftout of TEM lamelle using redeposition methods has been shown to be viable [5] [6] and a cryo-FIB approach using a combined controlled cryogenic GIS deposition and redeposition has been demonstrated for APT samples [7] but has not been optimised for the extreme conditions experienced during APT analysis.
In this presentation, we demonstrate a reproducible, GIS-free, process flow for cryo-FIB liftout and mounting of APT samples using selective resputtering [8] of an in-situ tungsten micro-manipulator. APT samples made using cryo-FIB liftout from single crystal silicon using this method have been successfully analysed and the data collected was found to be comparable in quality to that collected from commercial pre-sharpened silicon calibration specimens (Cameca) [9]. Progress of this approach and applications of this process to facilitate cryo-FIB sample preparation for complementary cryo-STEM and APT on the same sample will be discussed along with the numerous challenges of cryo and vacuum transfer between instruments involved in this process.
References:
[1] F. R. Wagner et al, Nature Protocols, 15 2041-2070 (2020).
[2] A. A. El-Zoka et al., Science Advances, 6 49 (2020) .
[3] M. Herbig, P. Choi, D. Raabe, Ultramicroscopy, 153:32-9 (2015).
[4] C. D. Parmenter, Z. A. Nizamudeen, Journal of Microscopy, 281 (2021),p.157-174.
[5] S. Klumpe et al., Microscopy Today, 30(1) (2022) p.42-47.
[6] J. Kuba et al., Journal of Microscopy, 281 (2) (2021),p112-124
[7] D.K. Schreiber et al., Ultramicroscopy, 194 (2018) p.89-99.
[8] S. Kölling, W Vandervorst, Ultramicroscopy, 109(5) (2009) p.486-491.
[9] J. Douglas et al, Microscopy and Microanalysis, In Review.

Solid-liquid interfaces are important to electrochemical energy technologies from electrical energy generation/storage to chemicals/materials synthesis. These electrochemical interfaces are complex and experimentally difficult to study partly due to a lack of effective tools to simultaneously access both solid and liquid phases at the nanoscale. This gap in understanding has contributed to insufficient experimental control over interfacial structure and reactivity. In addition, these interfaces are often closely related to the development of nanoscale intraparticle electrochemical inhomogeneities in phases and compositions. Such heterogeneities decrease the reversibility of the electrochemical reactions. For example, the solid-electrolyte interphase (SEI), an interfacial layer formed at the electrode-electrolyte interface due to the electrochemical/chemical decomposition of electrolytes. It is a key component responsible for the reversible operation of Li-ion and Li metal batteries and thus efforts have been made to engineer the SEI to enable battery chemistries with higher energy density and longer cycle. However, fundamental understanding of the interfacial phenomena in these battery chemistries is still limited. Elucidating the nanoscale structures and chemistries at the electrode-electrolyte interface is thus critical for developing high-energy-density batteries.
We adopted and developed the thin film vitrification method to preserve the SEI in native liquid organic electrolyte environments to enable high resolution cryo-TEM and spectroscopy. We quantified the electron dose limits for several conventional and state-of-the-art organic liquid electrolyte systems in vitrified states for Li-ion and Li metal batteries to carry out low dose cryo-TEM analysis. In combination of mechanical nanoindentation results with atomic force microscopy, we report substantial swelling of the SEI on Li metal anodes in electrolyte environments. The existence of electrolyte in SEI is rarely discussed or considered in the past and sheds light on the SEI growth mechanism. More importantly, we found that the swelling behavior is dependent on electrolyte chemistry and highly correlated to battery performance. Higher degrees of SEI swelling tend to exhibit poor electrochemical cycling for Li metal anodes. Based on these understandings, we further proposed a suspension electrolyte design to improve lithium metal anode performance. We tuned the local Li⁺ solvation structure with inorganic nanoparticle additives to form favorable anion-rich and less swelling SEI. Improved Coulombic efficiency (up to ~99.7%), reduced Li nucleation overpotential, stabilized interfaces and prolonged cycle life of anode-free cells (~70 cycles at 80% of initial capacity) were achieved with the suspension electrolyte design.
Furthermore, we leveraged cryogenic electron tomography (cryo-ET) to resolve in three dimensional the growth of extended SEI which is likely to be result of reaction between electrolyte in SEI and active Li metal. Combining cryo-ET reconstruction and atomic resolution cryo-TEM, we were able to map out the spatial distribution of extended SEI on the electrode. Moreover, we identified the growth LiH phase inside Li metal dendrites. The existence of an atomically sharp interface between Li and LiH phases suggested the growth proceeded with a phase transition mechanism. We hypothesized that the growth of LiH is related to the release of active H in solvent molecules or water. Using convolutional neural network, we were able to reconstruct the LiH phase inside Li metal dendrites. We discovered that the growth of LiH and extended SEI seemed to be spatially mutual exclusive events and therefore should be considered as different battery failure modes. These findings will serve as critical guides to the electrolyte design towards practical high energy density batteries.

The solid-electrolyte interphase (SEI) controls the transport of ions and electrons during battery operation. Thus, understanding its structure and chemistry is necessary to understand battery function. However, there are significant challenges in characterizing the SEI: it is nanoscale in dimension, heterogeneous in composition, and comprised of organic and inorganics species. The first two points suggest that electron and ion microscopy methods are needed, as they have the appropriate spatial resolution. However, organic materials present significant challenges to these approaches, as these materials are usually susceptible to irradiation damage.

Cryogenic transmission electron microscopy (cryo-TEM) has recently been used to observe lithium metals and their SEIs without electron beam damage. Most cryo-TEM research has probed lithium by electrochemically growing lithium on copper grids or extracting it from current collectors. These approaches necessitate removing liquid electrolytes from active materials by drying and washing before being characterized. However, it is unclear whether the structure and chemistry of the SEIs remain intact during such processes.

Vitrification is one way to observe the SEI: this is done by rapidly freezing liquid electrolyte media. However, as the vitreous liquid electrolytes that cover the active materials are often too thick, it is difficult to probe nanoscale assemblies of SEIs at high resolution through subsequent cryo-(S)TEM imaging. It is, therefore, crucial to develop a strategy to control the thickness of vitrified samples to obtain the desired electron transparency.

Cryogenic focused ion beam microscopy methods (cryo-FIB) allow thickness control through the cryogenic lift-out process; samples can be lifted out from specific sample areas and subsequently attached to TEM support grids for further thinning. Using the cryogenic lift-out process to characterize SEIs in energy storage materials has two advantages: (i) the thickness of the sample of interest can be carefully controlled through FIB milling and (ii) the configuration of battery structures can be preserved and selectively extracted without losing spatial information surrounding the region of interest.

In this work, we have used lithium and its SEI as a model material to identify how ion and electron beam damage can occur and how cryo-FIB can mitigate them. We also demonstrate the potential of cryo-TEM analysis for beam-sensitive battery materials using cryo-FIB.
In this talk, we will answer the following three critical questions:
1. We will answer how lithium gets damaged by ion beam milling at cryogenic temperatures.
2. We will discuss ion beam damage to SEIs.
3. We will finally show the structural alteration of SEIs before and after liquid electrolytes are removed, demonstrating the possibility of high-resolution imaging of nanoscale features in SEIs with vitrified electrolytes.

Ammonia is one of the highest value industrial chemicals, with 50% of global agriculture using industrially produced ammonia in fertiliser. However, the industrial method to make ammonia, the Haber-Bosch process, is extremely environmentally damaging. The requirement for extreme operating conditions and reliance on methane derived hydrogen results in an enormous energy and CO₂ penalty¹. A better method of ammonia production would be distributed, powered by renewable energy, and carbon free. Electrochemical ammonia synthesis via the Nitrogen Reduction Reaction (NRR) could provide a solution. The only rigorously verified method of electrochemical ammonia synthesis is the lithium-mediated nitrogen reduction system^1,2. This system utilises a lithium salt and organic electrolyte to catalyse the reaction between dinitrogen gas and a non-aqueous proton donor, which is often ostensibly sacrificial ethanol. While great strides have been made in terms of Faradaic efficiency, stability and activity^1,3, there are still numerous unanswered questions about exactly why this system is the sole example of efficient electrochemical ammonia synthesis.

One aspect of the lithium-mediated paradigm that could make it unique is the formation of the Solid Electrolyte Interphase (SEI) on the surface of the working electrode¹. This electronically resistive but ionically conductive SEI is formed from the decomposition of the organic electrolyte and lithium salt on the application of the highly reductive lithium plating potential, analogous to that formed in lithium-ion batteries¹. Recent work⁴ reveals that the chemical makeup of the SEI is critical for system stability and controlling the access of reactants (Li⁺, N₂ and protons) to the electrode surface. The role of ethanol as a proton donor has been recently disputed and it is proposed that it may in fact play a greater role in NRR-SEI formation^5,6. It would be beneficial to be able to study NRR-SEI morphology changes with ethanol concentration via electron microscopy. However, the NRR-SEI is highly beam and air sensitive, making both analysis in and transport to the microscope complex. Therefore, a specialised air-free transfer system and cryogenic analysis is required.

While previous cryo-microscopy investigations suggest changes in NRR-SEI makeup with and without ethanol⁶, we present a systematic study of the effect of changing ethanol concentration on NRR-SEI morphology via cryo-Scanning Electron Microscopy and cryo-Focussed Ion Beam milling, and propose how NRR-SEI morphology may affect Faradaic selectivity. Such work also paves the way for the future investigation of the SEI via cryo-Transmission Electron Microscopy. In addition, we present a novel method of cryogenic specimen preparation for Atom Probe Tomography (APT)⁷. APT will allow us to examine low atomic number elements, such as Li⁸, as well as the effect of ethanol concentration on reactant transport through the SEI.

References
1. Westhead, O., Jervis, R. & Stephens, I. E. L. Science (1979) 372, 1149–1150 (2021).
2. Tsuneto, A., Kudo, A. & Sakata, T. Journal of Electroanalytical Chemistry 367, 183–188 (1994).
3. Du, H.-L. et al. Nature 609, 722–727 (2022).
4. Westhead, O. et al. J Mater Chem A Mater (2023) doi:10.1039/D2TA07686A.
5. Westhead, O. et al. Faraday Discuss (2022) doi:10.1039/D2FD00156J.
6. Steinberg, K. et al. Nat Energy (2022) doi:10.1038/s41560-022-01177-5.
7. Douglas, J. O., Conroy, M., Giuliani, F. & Gault, B. ArXiv (2022) doi:https://doi.org/10.48550/arXiv.2211.06877.
8. Kim, S.-H. et al. J Mater Chem A Mater 10, 4926–4935 (2022).

Low-carbon efforts to realize a sustainable society are being vigorously pursued around the world. Rechargeable batteries, which are recognized as one of the most important nextgeneration energy technologies, are urgently needed not only for mobile applications such as cell phones and notebook PCs, but also for applications in transportation technology such as electric vehicles to promote low-carbon society.

Among these, lithium-ion battery materials, which are attracting attention because of their high energy density, react with water in the atmosphere when exposed to air, resulting in deterioration. In most cases, however, the sample is exposed to the atmosphere when it is transported to the electron microscope after preparation, making the observation itself difficult. Even if the sample can be transported to the electron microscope without being exposed to the atmosphere, it must be observed under cooled conditions because the crystal structure of the material tends to change due to electron irradiation during observation.

In order to conduct research on such samples, it is essential to develop a multifunctional system that can perform the entire transport process from sample preparation to electron microscopic observation under non-atmospheric conditions, as well as observation under cooled conditions. Against this background, Melville has been working on the development of non-atmospheric exposure and sample cooling techniques to meet the technical requirements of electron microscopy applications.

In this presentation, we will introduce our new TEM holder with non-exposure cooling and Peltier-cooled stage for FIB-SEM, which were developed based on the technologies we have developed so far.

Scanning transmission electron microscopy (STEM) is highly sensitive to lattice degrees of freedom and enables tracking of lattice displacements with sub-Angstrom resolution and picometer precision. This has ushered novel visualizations and discoveries of functional atomic displacements in ferroelectric materials, magnetoelectric oxides, and 2D materials. Due to stringent stability requirements, the overwhelming majority of high-resolution STEM measurements are limited to room temperature or above. In many classes of correlated electronic materials, however, subtle and exotic ground states emerge exclusively below room temperature. Achieving cryogenic capabilities in STEM is therefore paramount to accessing and probing low temperature phases including high-temperature superconductivity, charge order, metal-insulator transitions, and more.

I will illustrate recent, successful applications of high-resolution cryogenic STEM for discovering and understanding novel correlated phases at low temperature (~90 K) [1,2,3]. I will focus on charge order, a state in which the electrons and the lattice form superstructures that break the translational symmetry of the atomic lattice. In an overdoped manganite, we directly visualize intrinsic topological defects and show how these local fluctuations alter the periodicity and long-range order of incommensurate charge order stripes [1]. In a half-doped manganite, we address the long-standing question of whether charge order resides on the sites or bonds, discover an exotic intermediate state which breaks inversion symmetry, and find that non-linear couplings between distinct lattice distortion modes locally determine phase competition [3]. These cryogenic STEM results pave a clear path to imaging low temperature electronic phases with high resolution and precision.

[1] El Baggari et al., Proceedings of the National Academy of Sciences 115.7 (2018): 1445-1450.
[2] El Baggari et al., Physical Review Letters 125.16 (2020): 165302.
[3] El Baggari, Baek et al., Nature Communications 12, 3747 (2021)

EuAl₄ is one of charge density wave materials where the lowering of the crystal symmetry at a phase transition give rise to incommensurate modulation^1–4. CDW transition occurs below 145.1 K and exhibits four successive magnetic transitions below 16 K². Very recently, a chiral spin skyrmion structure was reported, making it a rare case where one could observe the co-exitance of exotic magnetic order and charge order^5,6. CDW phase transition is accompanied by symmetry-breaking atomic displacement in general. Knowledge of the atomic structure, especially which and how atoms are displaced and how CDW gives rise to these exotic physical properties, is thus of utmost importance for understanding nontrivial quantum properties.

Here, we combine atomic resolution Scanning Transmission Electron Microscopy (STEM) combined with scanning convergent beam electron diffraction (SCBED) at cryogenic temperature, enabled by the development of a new liquid nitrogen holder. We show this combination allows for the visualization and determination of the atomic structure in EuAl₄. SCBED is achieved by acquiring electron diffraction patterns formed by a highly coherent nanometer-size electron probe at different specimen locations. With the high-speed direct electron detector, diffraction patterns from a field-of-view of 200 nm with sub-nanometer sampling can be achieved, this powerful technique allows the mapping of interatomic distances in real space. This talk will show that the atomic displacement changes the intensity of Bragg reflection, and CDW can be visualized by diffraction contrast. Analysis of CBED patterns shows a reduction of inversion symmetry, and dominant displacement modes can be identified. Imperfections of CDW, such as dislocations and phase boundaries, can be revealed by interatomic distance mapping. Further, whether a relatively higher temperature CDW phase could be a precursor of the lower magnetic phases remains an open question.


1. G. Gruner, “The dynamics of charge-density waves.”
2. K. Kaneko, T. Kawasaki, A. Nakamura, K. Munakata, A. Nakao, T. Hanashima, R. Kiyanagi, T. Ohhara, M. Hedo, T. Nakama, Y. Onuki, J Physical Soc Japan. 90 (2021), doi:10.7566/JPSJ.90.064704.
3. S. Shimomura, H. Murao, S. Tsutsui, H. Nakao, A. Nakamura, M. Hedo, T. Nakama, Y. Onuki, J Physical Soc Japan. 88 (2019), doi:10.7566/JPSJ.88.014602.
4. S. Ramakrishnan, S. R. Kotla, T. Rekis, J. K. Bao, C. Eisele, L. Noohinejad, M. Tolkiehn, C. Paulmann, B. Singh, R. Verma, B. Bag, R. Kulkarni, A. Thamizhavel, B. Singh, S. Ramakrishnan, S. van Smaalen, IUCrJ. 9, 378–385 (2022).
5. W. R. Meier, J. R. Torres, R. P. Hermann, J. Zhao, B. Lavina, B. C. Sales, A. F. May, (2022) (available at http://arxiv.org/abs/2204.02319).
6. R. Takagi, N. Matsuyama, V. Ukleev, L. Yu, J. S. White, S. Francoual, J. R. L. Mardegan, S. Hayami, H. Saito, K. Kaneko, K. Ohishi, Y. Onuki, T. hisa Arima, Y. Tokura, T. Nakajima, S. Seki, Nat Commun. 13 (2022), doi:10.1038/s41467-022-29131-9.
7. The microscopy work was supported by an Early Career project supported by DOE Office of Science FWP #ERKCZ55–KC040304. All microscopy technique development was performed and supported by Oak Ridge National Laboratory’s (ORNL) Center for Nanophase Materials Sciences (CNMS), which is a DOE Office of Science User Facility.

Over the last decade, modern monochromated electron energy loss spectroscopy (EELS) in a scanning transmission electron microscope (STEM) has achieved the ability to combine ultrahigh energy and spatial resolution simultaneously. As a result, a new wave of experiments on ultralow energy excitations, such as phonons, phonon-polaritons, molecular vibrations, shallow electronic structure, and infrared optical excitations have achieved exciting results in the STEM.

The emergence of monochromators that can reach the mid-infrared has largely coincided with the development of stable cryogenic TEM holders that can reach ultralow temperatures. However, due to the relative recency of both techniques achieving consistency and accessibility only limited opportunities have been available to combine them on the same materials systems. Here, I will show recent work at ORNL on the combining cryo-STEM and monochromated EELS. Focusing on cryogenic phase changes, emergent behavior at cryo-temperatures, and the effect of low-temperatures on beam-sensitive materials.

This abstract has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).

Fast electrons spectroscopies have had huge success for nano-optics [1]. For phase-locked excitations (e. g. surface plasmons) electron energy loss spectroscopy (EELS) is an optical extinction analogue and cathodoluminescence (CL) that of optical scattering [2]. For "incoherent" excitations, EELS also measures optical extinction for atomically thin materials [3,4,5], while CL measures spectra similar to off-resonance CL [3]. For many technologically and fundamentally interesting materials, including semiconductors, experiments at cyrogenic temperatures is a requirement. For example, many quasi-particles (such as excitons and trions) are only stable at low temperatures, due to their low binding energy. This will be discussed in this contribution, including variable temperature measurements down to about 100 K. This will be used as a motivation for still colder experiments (to be realized).


Despite clear benefits (link to structural and chemical information, atomic-scale spatial resolution and broadband excitation), electron spectroscopies have some penalties which limit applications to nano-optics: lack of resonant excitation and polarization degrees of freedom and still limited spectral resolution (EELS). In this seminar, we will discuss how temporally resolved spectroscopies can mitigate some of these issues.

The lack of excitation energy control can be circumvented by measuring the energy lost by each electron in time coincidence EELS-CL experiments. This has been achieved using a nanosecond-resolved direct electron detector (Timepix3) [6], correlation electronics and a PMT. The information retrieved here is analogous to that of photoluminescence excitation spectroscopy (PLE), hence we name it cathodoluminescence excitation spectroscopy (CLE) [7]. With it, we explored the relative quantum efficiency of different excitation energies and decay pathways towards 4.1 eV defect photon emission in h-BN flakes [7].

The current impressive energy resolution achieved with electron beams (~meV range) is still orders of magnitude away from the necessary one to access the physics of lifetime-limited high quality factor optical modes, such as cavity modes in dielectric spheres [8]. As proposed more than a decade ago [9], electron energy gain spectroscopy (EEGS) should easily overcome this limitation. Here, we will discuss EEGS experiments using fast electron blankers (~ns) in a continuous-gun electron microscope, allowing energy resolution delivered by laser beams compatible with a state-of-the-art electron microscope performance. As an example, we will show measurements of whispering gallery modes in dielectric spheres separated by a few hundreds µeV with energy sampling in the µeV scale [10].

[1] F. J. García de Abajo, Rev. Mod. Phys. 82, 209 (2010).
[2] A. Losquin, et al., Nano Lett. 15, 1229 (2015).
[3] N. Bonnet, et al., Nano Lett. 21, 10178 (2021).
[4] F. Shao, et al., Phys. Rev. Mater. 6, 074005 (2022).
[5] S. Y. Woo, et al., in preparation (2022).
[6] Y. Auad, et al., Ultramicroscopy 239, 113539 (2021).
[7] N. Varkentina, et al., 8, eabq4947 (2022) (2022).
[8] Y. Auad, et al., Nano Lett. 22, 319 (2022).
[9] F. J. García de Abajo and M. Kociak, New J. Phys. 10, 073035 (2008).
[10] Y. Auad et al., in preparation (2022).

Topological polar solitions such as domain walls, polar voritices, skyrmions, etc, in ferroelectrics have attracted much attention owing to their unique functionalities and potential applications in electronic devices. Recent advances in transmission electron microscopy (TEM) and electron energy-loss spectroscopy (EELS) provides powerful tools to study the structure, properties, and dynamic behaviors of nanostructures with atomic resolution. In this talk, I will show how ferroelectric domains nucleate and evolve under applied electric field in TEM. Using quantitative TEM, we can measure and map the electric polarization of nanodomains, vortices and other polar solitons with the atomic resolution. Recently, we have developed a novel four-dimensional STEM (4D STEM) method that can directly map the local electric field and charge density of crystalline materials in real space with sub-angstrom resolution.[1]. I will also show that skyrmion-like polar nanodomains can be created in lead titanate/strontium titanate bilayers transferred onto silicon and can be switched from one type to another by an applied electric field, which substantially modifies their resistive behaviours.[2] The polar-configuration-modulated resistance is ascribed to the distinct band bending and charge carrier distribution in the core of the two types of polar texture. The integration of high-density (more than 200 gigabits per square inch) switchable skyrmion-like polar nanodomains on silicon may enable non-volatile memory applications using topological polar structures in oxides.

[1] W. P. Gao et al., Nature 575, 490 (2019)
[2] L. Han et al., Nature 603, 63 (2022).

Transmission electron microscopy is a critical asset to basic and applied sciences research as it enables the characterization of, e.g., structural order on the atomic scale. This includes not only crystalline structures and orientation, but also their temporal evolution, behavior at and across interfaces, and disordered, amorphous materials leveraging tomographic imaging. To date, the vast majority of electron microscopy harnesses the superior spatial resolution owing to a de Broglie wavelength of a few picometers. Taking advantage of the wave properties of coherent electrons and their interaction with electromagnetic fields, e.g., matter, allows for enhancing sensitivity and resolution and visualizing the spin degrees of freedom in the form of the in-plane magnetic induction. These experiments are based on the detection of the electron phase using off-axis or in-line holography. The former requires a biprism, the latter works in Fresnel mode/Lorentz mode with post processing involving a phase retrieval algorithm, such as transport-of-intensity or Gerchberg-Saxton. I will discuss the fundamentals, including pros and cons, of phase contrast imaging and review recent applications to visualize non-collinear spin textures in magnetic films in the presence of magnetic bias fields and cryogenic temperatures triggering magnetic phase transitions and spin fluctuations. One focus will be on chiral spin textures in amorphous materials where exit wave reconstruction allows to disentangle electrostatic, magnetization, and magnetic field contributions to the phase shift [1,2].

[1] RS et al., Adv. Mater. DOI: 10.1002/adma.201800199
[2] RS et al., Adv. Mater. DOI: 10.1002/adma.202004830

Supported by the National Science Foundation, Division of Materials Research under Grant No. 2203933.

Magnetic skyrmions are topologically protected magnetic spin textures that can exist both as individuals and in dense quasiparticle lattices. Manipulation of skyrmions then requires a detailed understanding of their stability. While individual skyrmions have been studied extensively, their collective behavior in skyrmion lattices is poorly understood, both on the micron-scale where lattice ordering is observed and the nanoscale where inter-skyrmion interactions occur. In this work, we will present the observation of Néel skyrmion lattice ordering in the van der Waals (vdW) ferromagnet Fe3GeTe2 (FGT) using cryo-Lorentz transmission electron microscopy (LTEM). We will discuss the magnetic domains formed in FGT under various conditions of applied field and as a function of temperature. We will discuss the stability of skyrmions under applied fields and formation of domains with higher topological charge. In order to understand the collective behavior of these Néel skyrmions, when imaged with LTEM, we performed imaging of skyrmion lattices over a large field of view and employed a machine learning algorithm which applies a convolutional neural network (CNN) to identify skyrmion centers. We will discuss the local and global lattice order, and the observed thermal hysteresis in the order. Based on an analytical model, we can elucidate the collective behavior of skyrmion lattices to be dependent on the competition between the Zeeman energy and the domain energy.

Cryogenic transmission electron microscopy offers new prospects for studying quantum materials and electron-beam-sensitive materials. Here, we describe recent progress in the use of phase contrast techniques in the (scanning) transmission electron microscope, in combination with liquid nitrogen and liquid helium cooling, to obtain quantitative information about electromagnetic fields in nanoscale materials and to generate improved contrast at low electron dose from materials that are affected by irradiation damage.

The first examples involve the use of off-axis electron holography and Lorentz imaging at low temperature to study magnetic skyrmions in crystallographically chiral materials such as FeGe and magnetic flux vortices in superconducting materials such as NbN in focused ion beam milled samples that provide geometrical lateral confinement. The influence on the measurements of sample preparation damage, the direction and strength of an applied magnetic field and electrical current, as well as the presence of electron-beam-induced specimen charging, will be discussed.

The second examples involve the use of (integrated) differential phase contrast imaging in the scanning transmission electron microscope at low temperature to record atomic-resolution images of electron-beam-sensitive materials, including metal-organic frameworks, organic materials and minerals. The use of event-driven detectors, automation of experimental workflows and non-standard electron beam scanning protocols will be discussed.

Finally, new concepts for improved instrumentation for such experiments, including liquid-helium-cooled specimen stages, advanced magnetizing units, specimen holders for high frequency experiments, sample transfer cartridges and capabilities for in situ laser irradiation of the sample in the electron microscope, will be presented.

Quantum materials refers to a class of materials with exotic properties that arise from the quantum mechanical nature of their constituent electrons, exhibiting, for example, high-temperature superconductivity, colossal magnetoresistivity, multiferroicity, and topological characteristics. Quantum materials often have incompletely filled d- or f-electron shells with narrow energy bands and the behavior of their electrons is strongly correlated. One distinct characteristic of the materials is that their electronic states are often spatially inhomogeneous, thus is well suited for study using spatially resolved electron beams with its great scattering power and sensitivity to atomic ionicity. Furthermore, most of these exotic properties only manifest at very low temperatures, posing a challenge to modern electron microscopy. It requires extraordinary instrument stabilities at cryogenic temperatures with critical spatial, temporal, and energy resolutions in both static and dynamic manner to probe these materials. On the other hand, the ability to directly visualize the atomic, electronic and spin structures and inhomogeneities of quantum materials and correlate them to their functionalities creates enormous opportunities. At the most elementary levels of condensed matter physics, understanding the competing orders of electron, spin, orbital, and lattice and their degrees of freedom, the impacts of defects and interfaces, and the site-specific quantum phenomena and phase transitions that give rise to the emergent behaviors allows us to discover and control novel materials for quantum information science and technologies [1].

In this presentation, several of our research examples are selected to highlight the use of Cryo-EM to study strongly correlated quantum materials. We focus on the critical roles of heterogeneity, interfaces, and disorder in crystal structure, electronic structure, and spin structure to understand the physical properties of the materials. We show how electron diffraction and diffuse scattering analysis at low temperatures empowers us to reveal the nature of structural disorder and phonons under thermal equilibrium and far-from equilibrium [1]; how atomically resolved imaging and electron energy-loss spectroscopy at 10K can be used to understand the interface-enhanced superconductivity (Fig.1); and how to use Lorentz phase microscopy to explore the intriguing transformations among various topological chiral spin states under applied magnetic field at various cryogenic temperatures (Fig.2). Finally, we review our recent unprecedented development of a closed-cycle cryocooler for Cryo-EM without the need of refilling liquid He. The system is designed to reach a temperature as low as 4k with temperature stability and control better than a few mK.

References
[1] Zhu, Y., “Cryogenic Electron Microscopy on Strongly Correlated Quantum Materials”, Invited review article, special issue on Cryogenic Electron Microscopy. Guest editors: Y. Cui and L. Kourkoutis, Accounts of Chemical Research, 54, 3518−3528 (2021).

Acknowledgement
The research was supported by the US DOE-BES, Materials Sciences and Engineering Division, under Contract No. DE-SC0012704. Collaborations with and assistance from the group members at BNL in the past 30 years are acknowledged.

Modern scanning and transmission electron microscopes (S/TEM) are now almost ubiquitous in materials and biological sciences laboratories. They have radically enhanced our understanding matter at the atomic level, bringing unique information of structure, chemical composition, and electronic properties of materials. Moreover, the recent development of stable cryogenic TEM holders at temperatures ranging from liquid N₂ (100 K) to 300 K with electrical contacts, combined with aberration-corrected and monochromated electron optics has allowed S/TEM to study magnetic, structural and electronic phase transitions with unprecedented spatial and energy resolutions.

In this talk, I will present two examples of cryogenic STEM measurements done in a set of two-dimensional (2D) transition metal dichalcogenides (TMDs) samples. In the first example I will address excitonic dephasing in MoS₂. I will present experimental results that show how monochromated electron-energy-loss spectroscopy (EELS) can reveal the exciton-phonon coupling of a few layers of MoS₂ encapsulated in hexagonal boron nitride (hBN). The measurements are spatial localized beyond the diffraction limit of the associated optical excitation, indicated how STEM can be complementary to conventional optical probes. In the second example I will present measurements that show a spatial dependence of a moiré-exitonic coupling in a hBN/WSe₂/WS₂ heterostructure at LN₂ cryogenic temperatures. I will also present a geometrical phase analysis of the moiré structure that allows to reveal the strain on the lattice at large field of views (>30 x 30 nm²) while still maintaining atomic resolution. Prospects and limitations of future experiment will be discussed in the detail during the talk.

Since 2017 the team of condenZero, a spin-off company from the University of Zurich, Switzerland (founded in 2019), has been developing new design principles for liquid helium cryostats. Initially created for x-ray diffraction experiments in pulsed magnetic fields, condenZero adapts the miniaturizable cryostat design for TEM sample holder applications. The main goal of the design is to provide a stable base temperature (<10K) over long time periods, enabling new experimental possibilities for TEM research. In this talk, we outline a brief history of condenZero, current challenges, related engineering projects and future milestones for the further development of our next-generation cryo-TEM sample holders.