key: cord-0853254-ladvfjma
authors: de Jonge, Niels
title: Chapter One CISCEM 2014 Proceedings of the Second Conference on In situ and Correlative Electron Microscopy, Saarbrücken, Germany, October 14–15, 2014
date: 2015-12-31
journal: Advances in Imaging and Electron Physics
DOI: 10.1016/bs.aiep.2015.02.004
sha: 0460400d16183e5173d6fe46f5ac9e726b74c9eb
doc_id: 853254
cord_uid: ladvfjma

Abstract This chapter contains the extended abstracts of the second conference on in situ and correlative electron microscopy (CISCEM 2014), held October 14–15, 2014, in Saarbrücken, Germany. The conference was housed at the INM-Leibniz Institute for New Materials. The aim of the conference was to bring together an interdisciplinary group of scientists from the fields of biology, materials science, chemistry, and physics to discuss future directions of electron microscopy research. The topics of the different sessions were correlative and in situ electron microscopy in biology, in situ observations of biomineralization processes, designing in situ experiments, high-temperature and other experiments, and in situ transmission electron microscopy of catalytic nanoparticles. A corporate session was also held.

One of the key challenges at the forefront of today's electron microscopy research is to observe processes at the nanoscale under relevant conditions. For samples from the materials science, this is accomplished by in situ electron microscopy. Movies-even at atomic resolution-are recorded of processes at high temperatures, in gaseous environments, or in liquids, while carefully taking into account the effect of the electron beam. For most biological samples, the electron beam impact prevents acquiring the time-lapse date, and research is mostly directed toward correlative light-and electron microscopy often using proteins labels. Single electron microscopic images are preferentially recorded in amorphous ice, or liquid. A conference discussing these topics was held for the second time at the INM-Leibniz Institute for New Materials on October 14-15, 2014, in Saarbr€ ucken, Germany. The conference on in situ and correlative electron microscopy CISCEM 2014 aimed to bring together an interdisciplinary group of scientists from the fields of biology, materials science, chemistry, and physics to discuss future directions of electron microscopy research. The venue was the Aula at Saarland University.

The conference opened with a session on correlative and in situ electron microscopy in biology. Keynote speaker Wolfgang Baumeister gave a broad overview of in situ transmission electron microscopy (TEM) of proteins and cells embedded in amorphous ice. Recent advances in correlative light and electron microscopy were discussed by the invited speakers, including Ben Giepmans 1 presented the topic of electron microscopy of cells and viruses in liquid. The second session involved in situ observations of biomineralization processes. A highlight on this topic was a presentation by James de Yoreo, showing movies of such processes with atomic resolution. Bio-degradation processes were studied in situ by Damien Alloyeau.

The first day of the conference also accommodated a session with (nonscientific) corporate presentations (not reflected in this chapter). This day ended with a poster session including a total of 29 posters on the following topics, movement of nanoparticles, designing in situ experiments, hightemperature and other in situ experiments, experiments in biology, and experiments in metals.

The second day started with a dedicated session discussing various aspects of the design of in situ experiments. The two invited speakers at this session were Eva Olsson and Patrica Abellan. The most important topic was experiments in a liquid environment. High-temperature and other experiments were presented in the fourth session, including also a presentation on in situ characterization of battery materials. The conference concluded with a session on in situ TEM of catalytic nanoparticles in gaseous environments. Invited speakers were Renu Sharma and Jakob Wagner.

This chapter contains a selection of the extended abstracts submitted for the conference. (1) electron crystallography, (2) single-particle analysis and (3) electron tomography. Ideally, all three imaging modalities are applied to frozen-hydrated samples, ensuring that they are studied in the most lifelike state that is physically possible to achieve. Vitrified aqueous samples are very radiation sensitive and consequently, cryo-EM images must be recorded at minimal electron beam exposures, limiting their signal-to-noise ratio. Therefore, the high-resolution information of images of unstained and vitrified samples must be retrieved by averaging-based noise reduction, which requires the presence of repetitive structure. Averaging can obviously not be applied to pleomorphic structures such as organelles and cells (Fitting Kourkoutis et al., 2012; Leis et al., 2009) . Electron crystallography requires the existence of two-dimensional crystals, natural or synthetic, and averaging is straightforward given the periodic arrangement of the molecules under scrutiny. In principle, electron crystallography is a high-resolution technique as demonstrated successfully with a number of structures, particularly membrane proteins. Often, however, the same structures can be studied by X-ray crystallography, which tends to be faster and can attain atomic resolution more easily.

In contrast, EM single particle analysis (arguably a misnomer since it involves the averaging over large numbers of identical particles) has become one of the pillars of modern structural biology. The amount of material needed is minute, and some degree of heterogeneity, compositional or conformational, is tolerable since image classification can be used for further purification in silico. It is particularly successful in structural studies of very large macromolecular complexes where the traditional methods often fail. In principle, single-particle analysis can attain nearatomic resolution, but in practice, this often remains an elusive goal. This is changing, however, with the advent of new technology, particularly detectors with improved performance. But even intermediate resolution (subnanometer) structures of very large complexes can provide an excellent basis for hybrid or integrative approaches in which high-resolution structures of components and orthogonal data, such as distance restraints, are used to generate atomic models.

Electron cryotomography can be used to study the three-dimensional organization of nonrepetitive objects (Lucic et al., 2005) . Most cellular structures fall into this category. In order to obtain three-dimensional reconstructions of objects with unique topologies, it is necessary to acquire data sets with different angular orientations of the sample by physical tilting. The challenge is to obtain large numbers of projections covering as wide a tilt range as possible and, at the same time, to minimize the cumulative electron dose. This is achieved by means of elaborate automated acquisition procedures. Electron cryotomography can provide medium-resolution, three-dimensional images of a wide range of biological structures from isolated supramolecular assemblies to organelles and whole cells. It allows the visualization of molecular machines in their unperturbed functional environments (in situ structural biology) and ultimately the mapping of entire molecular landscapes (visual proteomics; Robinson et al., 2007) .

Until recently, the use of electron cryotomography was restricted to relatively thin samples, such as prokaryotic cells or the margins of eukaryotic cells. This has changed with the advent of focused-ion beam (FIB) micromachining and developments that allowed the application of this technology to samples embedded in vitreous ice. This allows the cutting of "windows" providing views of the interior of thicker samples such as eukaryotic cells. By combining the FIB with correlative fluorescence microscopy, it is now possible to navigate large cellular landscapes and to select and target specific areas of interest (Villa et al., 2013) .

Given the full signal-to-noise ratio of the tomograms, it can be challenging to interpret them and take advantage of their rich information content. Image denoising can improve the signal-to-noise ratio by reducing the noise while preserving the features of interest. Segmentation separates the structures of interest from the background and allows their three-dimensional visualization and quantitative analysis. Larger molecular structures can be identified in tomograms by pattern recognition methods using a template structure, and once their location and orientation is determined, identical structures can be extracted computationally and averaged. Therefore, electron tomography has unique potential to bridge the divide between molecular and cellular structural studies, perhaps the most exciting frontier in structural biology (Villa et al., 2013) .

Ben N.G. Giepmans* Department of Cell Biology, University of Groningen, University Medical Center Groningen, The Netherlands *Corresponding author: e-mail address: b.n.g.giepmans@umcg.nl; www.cellbiology.nl Today, I will first focus on recent developed labeling strategies for probes that allow Correlated light and electron microscopy (CLEM) (Giepmans, 2008; Sjollema et al., 2012) . These include particles (gold, quantum dots) to highlight endogenous proteins, but also genetically encoded probes, as well as traditionally used stains for light microscopy (LM) that aid in electron microscopy (EM)-analysis of samples. Probes that can be detected only in a single modality and require image overlay, as well as combinatorial probes that can be visualized both at LM and EM will be discussed. In addition, published (Ravelli et al., 2013) and new approaches for large-scale EM to visualize macromolecules and organelles in the context of organized cell systems and tissues will be covered (www.nanotomy.nl). Matching the areas of acquisition in CLEM and EM will not only increase understanding of the molecules in the context, but also is a straightforward manner to combine the LM and EM image data. Covering a wide variety of probes and approaches for image overlay will help to enable (new) users to implement CLEM to better understand how molecules (mal)function in biology. Figure 1 snapshots taken from www.nanotomy.nl, an open-source database, zooming into the boxed areas (clockwise). Note that the islet of Langerhans is identifiable (left), but also cells and organelles, as well as macromolecular complexes (ribosomes, nuclear pores, etc). Data from Ravelli et al. (2013) .

See Ravelli et al. (2013) Correlative light electron microscopy combines the strengths of light and electron microscopy in one experiment, and the sum total of such an experiment should provide more data/insight than each technique alone (hence 1 +1¼ 3). There are many ways to perform a CLEM experiment, and a variety of microscopy modalities can be combined. The choice of these instruments should primarily depend on the scientific question to be answered. A CLEM experiment can usually be divided into three parts; probes, processing, and analysis. I will discuss three processing techniques based on light microscopy in conjunction with transmission electron microscopy, each with its advantages and challenges.

The first is based on the use of coverslips with a finder pattern, it allows live cell imaging and captures an event of interest using chemical fixation ( Figure 1 ; Hodgson et al., 2014a) , A second uses the Tokuyasu cryo immuno Figure 1 CLEM allows the identification and subsequent high-resolution analysis of 1 special cell among hundreds. In the upper-left row, a dividing cell (1 in >100) is identified based on its DNA staining (blue). With the aid of embossed coverslips (for full details, see Hodgson et al., 2014a) the dividing cell can be traced back in the EM and studied at higher resolutions. labelling to trace back objects of interest ( Figure 2 ; Hodgson et al., 2014b) , this allows for relatively high immunolabeling efficiencies but is almost impossible in combination with live cell imaging. The third is based on cryofixation to obtain the best possible preservation of ultrastructure (Verkade, 2008; Brown et al., 2012) . This allows us to capture events that would be lost because of chemical fixation (e.g., membrane tubules). It allows for live cell imaging, but immunolabeling options are limited. Understanding the fundamental properties of macromolecules has enhanced the development of emerging technologies used to improve biomedical research. Currently, there remains a critical need for innovative platforms that can illuminate the function of biological objects in a native liquid environment. To address this need, we have developed an in situ TEM approach to visualize the dynamic behavior of biomedically relevant macromolecules at the nanoscale. Newly designed silicon nitride-based devices containing integrated microwells were used to enclose active macromolecular specimens in liquid for TEM imaging purposes ( Figures 3A, B ). With each specimen tested, the integrated microwells could adequately maintain macromolecules in discrete local environments (Dukes et al., 2014) while enabling thin liquid layers to be produced for high-resolution imaging purposes as previously exemplified using gold nanorods (Dukes et al., 2013) . This success permitted us to utilize the integrated microwell-designed microchips to examine actively transcribing rotavirus assemblies having native contrast (Dukes et al., 2014) . In developing biochemical experiments to assess viral attributes, we first needed to manufacture competent viral specimens. To accomplish this objective, we purified simian rotavirus double-layered particles (DLPs) (strain SA11-4 F) from monkey kidney MA104 cells, as previously described (Dukes et al., 2014) . The proteins that comprised the purified DLPs were analyzed using SDS-PAGE and silver staining ( Figure 4A ). We found that DLPs produced in the MA104 cells contained four proteins (VP1, VP2, VP3, and VP6) and that VP4 and VP7 were absent from the formed particles. To verify that our purified DLPs could transcribe viral RNAs, we utilized an in vitro messenger RNA (mRNA) synthesis assay. Each reaction mixture contained DLPs, each NTP, and [ 32 P]-UTP, and they were allowed to incubate for 30 min at 37°C ( Figure 4A , +ATP). Negative control reactions also contained each transcription cocktail component except ATP ( Figure 4A , ÀATP). Radiolabeled mRNA products were detected in the reaction mixtures containing a complete transcription cocktail, and no radiolabeled products were detected in the reaction mixtures lacking ATP. Therefore, this functional analysis confirmed that the purified DLPs used for subsequent imaging analysis were enzymatically active.

We attempted to visualize transcribing rotavirus DLPs using reaction mixtures that were prepared as described previously. An aliquot of each transcription cocktail was added onto Ni-NTA-coated SiN microchips that were previously decorated with His-tagged protein A and IgG polyclonal antibodies against the VP6 capsid protein (Degen et al., 2012; Gilmore et al., 2013) (Figure 4A , schematic). The fluidic microchamber that contained the antibody-bound transcribing DLPs was assembled into the Poseidon specimen holder. Active DLPs were examined using a FEI Spirit Bio-Twin TEM equipped with a LaB 6 filament and operating at 120 kV. Images of transcribing DLPs were recorded using an Eagle 2 k HS charge-coupled device (CCD) camera under low-dose conditions (approximately 0.5 electrons/Å 2 ) to minimize beam damage to the viral specimens. The resulting images revealed dynamic attributes of RV pathogens in liquid at 3-nm resolution ( Figure 4B ). We could also distinguish discrete strands emerging from numerous DLPs. These strands had characteristic shapes and dimensions consistent with being single-stranded viral mRNA transcripts ( Figure 4B , 1-4, right panels). No strands were identified in images of our negative control transcription reactions that lacked ATP. We could subsequently use the RELION software package to compute 3D reconstructions of the active DLPs from a single image ( Figure 4C ). The interiors of the DLP cores revealed movements indicative of protein rearrangements during mRNA synthesis. The spatial distribution of internalized nanoparticles (NPs), and of membrane proteins tagged with NP labels were studied by imaging whole and hydrated cells with an environmental scanning microscope (ESEM), equipped with a scanning transmission electron microscope (STEM) detector ( Figure 5 ). COS7 fibroblast, A549 lung cancer, and SKBR3 breast cancer cells were grown on silicon microchips with silicon nitride (SiN) membrane windows. One ESEM-STEM study followed the fate of gold nanoparticles (AuNPs) within cells, an important topic in view of the high potential of AuNPs for medical applications. Here, A549 cells took up serum protein coated AuNPs of 10-, 15-, or 30-nm diameters. One or two days after the AuNP uptake cells were fixed and investigated with ESEM-STEM Figure 5 A schematic of ESEM-STEM of whole cells in a wet state. A focused electron beam (30 keV) is scanned over the fixed and hydrated cells. Contrast is obtained on QDs labeling EGFR or HER2 receptors or on internalized AuNPs. A gaseous secondary electron detector, located above the sample, and a STEM detector, located beneath the sample, simultaneously collect the signals.

( Figure 6A ), AuNPs were found in a distinct lining pattern within intracellularly scattered lysosomes. The dimensions of 1,106 AuNP-storing lysosomes were determined from 145 whole cells, within a total time (including imaging and analysis) of only 80 h. This study revealed a statistically relevant enlargement effect on the size of the lysosomes of the 30-nm AuNP compared to the smaller NPs . Studied membrane proteins included the epidermal growth factor receptor (EGFR), and the related receptor tyrosine kinase HER2. The mapping of monomers and dimers of these two receptors is important for basic research, as well as for the study of the molecular mechanisms involved in certain anti-cancer drugs. Membrane-bound EGFR or HER2 on live cells were labeled with probes consisting of small protein ligands and AuNPs or fluorescent quantum dots (QDs). After fixation, EGF-AuNP labeled cells were examined directly with ESEM-STEM ( Figure 6B ), whereas cells labeled with QD probes were studied with correlative fluorescence microscopy and ESEM-STEM ( Figure 2C and D). In all cell lines, significant fractions of the labeled receptors appeared as dimers and in small clusters. Hundreds of ESEM-STEM images were recorded of several tens of cells providing nanometer-scaled data from thousands of labels, which were analyzed and quantified by automated image software algorithm. In addition, correlative microscopy confirmed the heterogeneity of nonisogenic cancer cells, manifesting in large variations of EGFR and HER2 expression and distinct spatial distributions on the cell membrane.

In conclusion, ESEM-STEM is an exciting EM methodology for analytic studies of whole cells in their hydrated state with nanometer resolution and in very short timeframes. to combine the two techniques in one instrument, we are now routinely able to register and-using fluorescence probes-accurately localize regions of interest anywhere on a standard EM grid while using most types of TEM sample preparations. We have also shown that we can use intrinsic fluorescence to provide complementary information that is not accessible by EM alone. However, light and electron microscopy remain vastly different methods, with an accordingly large gap in the relevant length scales. In order to bridge this gap, we have recently begun to combine super-resolution light microscopy with TEM. All positive-strand RNA viruses replicate in the cytoplasm in distinct membranous compartments serving as replication factories. Membranes building up these factories are recruited from different sources and serve as platforms for the assembly of multi-subunit protein complexes (the replicase) that catalyze the amplification of the viral RNA genome. In this study, we found that hepatitis C virus (HCV), a major causative agent of chronic liver disease, induces profound remodeling of primarily endoplasmic reticulum (ER)derived membranes. By using correlative light and electron microscopy (CLEM), we observed that HCV triggers the formation of double membrane vesicles (DMVs), surrounding lipid droplets and residing in close proximity of the ER ( Figure 9A ; Romero-Brey et al., 2012) . Furthermore, by means of electron tomography (ET), we showed that these DMVs emerge as protrusions from ER tubules ( Figure 9B ; Romero-Brey et al., 2012) . CLEM allowed us to confirm the important contribution of one of the HCV nonstructural proteins (NS5A) to the formation of DMVs ( Figure 10A ; unpublished data). Importantly, inhibitors that are currently tested in clinical trials and targeting NS5A disrupt biogenesis of these HCV-induced mini-organelles and completely block virus replication ( Figure 10B ; Berger et al., 2014) .

These results unravel the mode of action of highly potent HCV inhibitors and disclose unexpected similarities between membranous replication factories induced by HCV and the very distantly related picornaviruses and coronaviruses. 

Solution-based growth of single crystals through assembly of nanoparticle precursors is a pervasive mechanism in many materials and mineral systems. Morever, the dominance of particle-mediated growth processes increases the importance of understanding the mechanisms and controls on nucleation of the precursor particles. Yet many longstanding questions surrounding nucleation remain unanswered and the postnucleation assembly process is poorly understood, due in part to a lack of experimental tools that can probe the dynamics of synthetic processes in liquids with adequate spatial and temporal resolution. Here, we report the results of using fluid cell TEM to investigate nucleation of calcium carbonate (Nielsen et al., 2014a) and particle assembly in both the iron oxyhydroxide (Li et al., 2012) and calcium carbonate systems (Nielsen et al., 2014b) .

To examine nucleation of calcium carbonate, we used a custom-built fluid holder that enabled us to mix two reagents near the entrance to the cell and thus explore a wide range of solution conditions 1 . We observed the formation of amorphous calcium carbonate (ACC) over the entire range of conditions. In addition, we found that all common crystalline phases of calcium carbonate, including calcite, vaterite, and aragonite could form directly. Multiple phases often formed within a single experiment and the direct formation of the crystalline phases occurred under conditions in which ACC also readily formed. These observations demonstrate that multiple phases of calcium carbonate can form directly from solution without the intermediate stage of ACC. For all phases measured, we found radial/edge growth rates after nucleation were linear with respect to time, showing that growth was reaction limited. Beyond these direct formation pathways, we observed transformation from ACC to aragonite and vaterite, but, significantly, not to calcite ( Figure 11 ). In these observations, ACC transformed directly to the crystalline phases rather than undergoing a process of dissolution and reprecipitation. Nucleation of the second phase began on or just below the surface of the ACC particle and was preceded by a brief period of particle shrinkage, perhaps associated with expulsion of water. These formation pathways were confirmed by collecting diffraction information of the various phases of calcium carbonate.

To understand how the introduction of an organic matrix, which is common in biomineral systems, affects the nucleation of calcium carbonate, we performed a similar set of experiments in solutions containing the polyelectrolyte polystyrene sulfonate (PSS). Here, the cell was initially filled with CaCl 2 solution through one inlet and carbonate ions were introduced by diffusion from an ammonium carbonate source through the second inlet. In the absence of PSS, vaterite formed randomly throughout the fluid cell. Figure 11 In situ liquid phase TEM enables the observation of phase evolution during nucleation (Nielsen et al., 2014a) . (A-F) Time series showing CaCO 3 nucleation via a two-step process. The first phase to form is amorphous CaCO 3 (ACC) (A,B). This is followed by surface nucleation of aragonite (C, D) and consumption of the ACC (E, F). Just before the moment of aragonite nucleation, the ACC partcle shrinks in size, indicating expulsion of its structural water. Scale bar: 500 nm.

When PSS was introduced, it complexed more than half of the Ca 2+ ions and formed a globular phase. As carbonate diffused into the cell, the first solid phase to appear was ACC, which nucleated only within the globules. These results demonstrate that ion binding can play a significant role in directing nucleation, independent of any control over the free energy barrier to nucleation, which is usually inferred to be the primary factor leading to matrixcontrolled nucleation (Habraken et al., 2013) .

We investigated the postnucleation growth of iron oxyhydroxide (Li et al., 2012) and calcium carbonate (Nielsen et al., 2014b) though particle assembly processes using a custom-built static fluid cell that enabled subnanometer resolution. We found that primary particles of ferrihydrite interacted with one another through translational and rotational diffusion until a near-perfect lattice match was obtained either with true crystallographic alignment or across a twin plane ( Figure 12 ). Oriented attachment (OA) then occurred through a sudden jump-to-contact, demonstrating the existence of an attractive potential driving the OA process. Following OA, the resulting interfaces expanded through ion-by-ion attachment at a curvaturedependent rate. However, when a significant mismatch existed between the sizes of the two particles and attachment failed to occur over extended periods of particle interaction, the larger crystals still grew in size through Ostwald ripening, resulting in the disappearance of the smaller ones. In contrast to the clear role played by OA in the case of ferrihydrite, analysis of the assembly of akaganeite nanorods to form single-crystal hematite spindles showed that attachment did not result in coalignment; rather, the initial mesocrystal was disordered and recrystallized over time to become a well-ordered single crystal. Calcium carbonate exhibited still a different style of particle-mediated growth. In this system, we also observed that nanoparticles interacted and underwent aggregation events; however, the smallest particles often appeared to be amorphous, with crystallinity presumably arising as a result of attachment to the larger crystalline particle ( Figure 13 ).

The results presented here highlight the wide array of pathways that are accessible during the nucleation process, as well as the diversity of mechanisms possible in particle mediated growth of single crystals. In both cases, the availability of liquid-phase TEM opens up new opportunities to decipher these underlying pathways and mechanisms. Using liquid-cell TEM holder in an aberration-corrected TEM is a major technological rupture for understanding the complex phenomena arising at the liquid/solid interface. Recent microelectromechanical system (MEMS)based technology allows imaging the dynamics of nano-objects in an encapsulated liquid solution within an electron-transparent microfabricated cell. The environmental conditions are finely controlled with a micro-fluidic system which enables to mix different reaction solutions at the observation window. Here, we performed the direct in situ study of two crucial phenomena in materials science: (1) the growth mechanisms of gold nanoparticles (NPs). (2) The degradation mechanisms of iron oxide NPs in a solution mimicking cellular environment.

We studied the growth of gold NPs via the reduction of metal salt. These growth mechanisms observed with a resolution below 0.2 nm, is indirectly induced by the electron beam. Indeed, 200-kV incident electrons radiolyze the water, creating free radicals and aqueous electrons that reduce metallic precursors. We have shown that the growth mode of gold NPs depends highly on the electron dose. High electron doses result in a diffusion-limited growth mode, leading to large dendritic structures, while low electron dose allows the formation of faceted NPs due to reaction-limited growth. These Figure 14 In situ growth of facetted gold NPs observed by low-dose STEM-HAADF. We observe a shape transition between two nano-polyhedra.

latter conditions enable the fascinating study of the 2D growth mechanisms of nanoplates. (Figure 14 ).

If the understanding of the formation mechanisms of inorganic NPs is very important for controlling upstream their shape related-properties, studying their reactivity and transformation mechanisms in cellular environment is essential for evaluating their long-term efficiency as diagnostic or therapeutic agents. Here, we demonstrated for the first time that liquid-cell TEM is a relevant method to follow the (bio)degradation of iron oxide NPs in a solution mimicking the intracellular environment to which they are exposed during their life cycle in the organism (Figure 15 ). Biominerals possess shapes, structures, and properties not found in synthetic minerals. These defining characteristics arise from the interplay of the mineral with a macromolecular matrix, which directs crystal nucleation and growth (Lowenstam & Weiner, 1989; Mann, 2001) . Within this threedimensional biomolecular assembly, the developing mineral interacts with acidic macromolecules, either dissolved in the crystallization medium or associated with insoluble framework polymers such as chitin or collagen (Palmer et al., 2008) . Although acidic macromolecules are known to affect growth habits and phase selection, or even to completely inhibit precipitation in solution (Sommerdijk & With, 2008; Meldrum & C€ olfen, 2008; Gower, 2008) , little is known about the role of matrix-immobilized acidic macromolecules in directing mineralization. This lack of understanding is, in part, due to the difficulty of studying biomimetic mineralization systems with sufficient spatial and temporal resolution (Dey et al., 2010) . However, liquid phase transmission electron microscopy (LP-TEM) can visualize events in situ in thin liquid volumes (about 500 nm in height) confined within two electron transparent silicon nitride (SiN) membranes. Here, we use LP-TEM to visualize the nucleation and growth of CaCO 3 in a biomimetic matrix of polystyrene sulfonate (PSS). In particular, we utilized a dual inlet flow stage where we started out with a CaCl 2 solution in the confined cell, after which carbonate was introduced through in-diffusion of vapor released from the decomposition of solid (NH 4 ) 2 CO 3 (mainly CO 2 (g) and NH 3 (g) ) via the second inlet port. Within minutes, we directly observed the nucleation of randomly distributed vaterite nanoparticles over the surface, which we attribute to heterogeneous nucleation on the SiN.

When we introduced a PSS solution together with our CaCl 2 solution, in situ imaging of this model system showed that the PSS by binding of Ca 2+ ions is able to form globules of about 10-100 nm that can adsorb onto the SiN surface, as shown by our previous work (Smeets et al., 2013) . We find that calcium is able to bind to the sulfonate groups of the polymer (as determined by FTIR measurements); 56 AE 6 mole % of the total added calcium is bound as determined by performing Ca 2+ ion selective electrode measurements.

The immobilized Ca-PSS globules inside the TEM cell were subsequently exposed to vapor from the decomposition of solid (NH 4 ) 2 CO 3 . After a delay time of about 20 min the nucleation of CaCO 3 nanoparticles was observed, growing to sizes of 10-20 nm within seconds ( Figure 16A ,C). In situ electron diffraction showed that the as-formed particles were amorphous ( Figure 16C , inset). Growth rate profiles of ACC (with PSS) and vaterite (without PSS) were extracted from the time-lapse series, where the initial rates were higher for ACC particles (16-23 nm/s) compared to vaterite crystallites (3-6 nm/s) ( Figure 16B ). To validate that the observed nucleation of nanoparticles was not the result of the continuous exposure to Figure 16 Nucleation of ACC from Ca-PSS globules imaged in LP-TEM. (A) Image sequence after 45 min of (NH 4 ) 2 CO 3 diffusion showing initial nucleation and growth of a CaCO 3 particle inside (i) a primary Ca-PSS globule within 4 s (viii) (ACC 1, scale bar: 20 nm). (B) Extrapolated growth rates versus average radius for two ACC particles (ACC 1 & 2 with PSS) compared to those of three vaterite particles (without PSS). A logarithmic function is used to fit each set of data points and to extrapolate to zero radius. (C) Lower magnification (scale bar: 50 nm) image exhibiting many nuclei in Ca-PSS globules and an electron diffraction pattern (inset) showing they are amorphous. Differences in the radial distribution of black and white contrast in the nucleated ACC particles is due to differences in the location relative to the focal plane: Black signifies near-focus condition (near the bottom of the SiN membrane) and white reflects overfocus (near the top of the membrane) (Smeets et al., 2015) . the electron beam, the reaction solution was imaged at different time points between 20 and 65 min at unexposed areas using a single exposure with an electron dose of about 50-300 e/Å 2 , comparable to what is used in low dose cryo-TEM imaging (Friedrich et al., 2010) . In these images, comparable amorphous nanoparticles were identified, again forming only at the sites of Ca-PSS globules.

Thus, we verify that the binding of Ca 2+ in the globules is a key step in the controlled formation of metastable amorphous calcium carbonate, an important precursor phase in many biomineralization systems (Addadi et al., 2003) . These findings now reveal the significant role that ion binding can play in directing nucleation independent of any controls over the free energy barriers.

Biological systems and many industrial applications depend on a precise control of crystallisation processes. In living systems, a variety of proteins regulate the progress of the mineralisation leading to materials with enhanced properties (Arias & Fernández, 2008) . In order to improve biomimetic materials engineering approaches, it is necessary to gain a better understanding of how organics interact with inorganics in the process of particle formation.

In this work, we focus on the impact of the protein AP7 on the in vitro crystallisation of calcium carbonate. AP7 is an intracrystalline protein found in the nacre layer of the pacific red abalone (Michenfelder et al., 2003) . It forms intrinsically instable oligomers that assemble into larger aggregates when calcium ions are present (Amos et al., 2011) . Our potentiometric experiments show that this protein influences the progress of the crystallization already in the prenucleation stage by associating with calcium carbonate prenucleation species, thereby prolonging the time till the onset of nucleation (Perovic et al., 2014) . Protein phase-calcium carbonate nanoparticle interactions in the postnucleation stage were observed by a conventional TEM time series study showing entrapped calcium carbonate particles within the protein phase. However, conventional TEM requires a high vacuum that commonly leads to a number of undesired drying artefacts.

Fluid cell scanning transmission electron microscopy allowed us to overcome this shortcoming and provided additional time-resolved information of the particle formation process. For these experiments, we utilized a commercially available three-port liquid cell holder where the liquid is situated between two silicon chips with electron transparent silicon nitride windows in the centre. Solutions of calcium ions and bicarbonate/carbonate ions, respectively, were mixed within the TEM holder tip in order to see the precipitation reaction from the onset. Using the electron microscope in scanning mode enabled us to image through liquid thicknesses of a few microns and moreover restrict the area exposed to the electron beam only to the investigated location. In this study, we found that AP7 assembles and forms calcium rich-nanocrystal-protein networks if both calcium and carbonate ions were present. These protein networks are incorporated into the growing calcium carbonate crystals and even remain unchanged when the surrounding crystal is dissolved by a variation of the experimental conditions ( Figure 17 ). 

Eva Olsson* Department of Applied Physics, Chalmers University of Technology, 412 96 Gothenburg, Sweden *Corresponding author: e-mail address: eva.olsson@chalmers.se As material synthesis and nanofabrication methods are refined and the control of material structure reaches beyond the nanoscale, the role of individual interfaces, defects, and atoms is pronounced and can dominate the properties. The strong influence of local atomic structure offers the possibility of designing new components with tailored and unique properties. Electron microscopes offer the possibility of correlating the structure to transport properties with a spatial resolution that reaches the atomic scale. A knowledge platform of how to design new materials by combining experiments and theory can thus be established. In addition, inserting a scanning tunnelling or an atomic force microscope in the electron microscope enables studies of dynamic events. This talk will cover aspects of in situ studies using manipulators with high spatial precision in electron microscopes, including experiments in low-vacuum conditions and give examples from different material systems. Using fluid stages in the TEM, in situ observations of dynamic processes in liquids can be made with nanometer and even atomic spatial resolutions. The flexibility of this approach for studying liquid phase reactions relies upon both the possibility of obtaining information by using well established as well as novel capabilities in the TEM platform and the possibility of fine-tuning the design of the experiment. Fluid stages are designed to fit in any TEM, allowing for local chemical information in microscopes equipped with electron energy loss spectrometers, subangstrom spatial resolution using Cs-correction, or high temporal resolution on the 10 ns to 1 μs range, if using a dynamic TEM (DTEM). Additional flexibility comes from the design of the fluid stage, where the liquid chamber is created and separated from the high vacuum in the microscope column using two microfabricated chips. Thus, the experiment layout can be adapted for each sample by using chips designed for a specific application including different configurations of chips for continuous flow applications, mixing of two chemicals or with electrical contacts. As the technique is increasingly refined, the challenge now becomes obtaining quantifiable and reproducible data that is free from beam-induced artifacts. Once the effect of the electron beam on the liquid sample is calibrated, reliable information on the fundamental mechanisms behind the colloidal synthesis of nanostructured materials, self-assembly or the structural processes during battery operation, and even biological conformational changes can be uncovered, just to name a few.

Studying liquid samples in the TEM represents specific challenges as compared to the study of crystalline/amorphous solid specimens. Solutions decompose upon radiolysis and chemical species are generated in the liquid phase and interact with the sample, most times beyond the observation area since they are mobile by nature. Factors such as the electron dose applied to the sample (Zheng et al., 2009; Woehl et al., 2012; , accelerating voltage (Abellan et al., 2014a) and imaging mode (e.g., TEM, STEM, SEM; Abellan et al., 2014a; , liquid thickness (Jungjohann et al., 2013) , and solution composition strongly determine the amount of radiation damage produced, and their effect must be calibrated for. In many systems, as for the case of an aqueous Ag precursor solution, radiolysis of deionized water in the TEM leads to the nucleation and growth of particles (Woehl et al., 2012; Abellan et al., 2014a) . As an example, Figure 18 illustrates (Abellan et al., 2014a) . For the specific case of growth of nanomaterials, using solvents with higher chemical complexity, such as organic solvents, leads to the production of a larger variety of radicals (as compared to water) increasing the number of chemical reactions involved. Therefore, understanding how these solutions are affected by the electron beam becomes even more crucial. Besides the study of particle growth, the damaging effect of the electron beam can be used for other applications, such as an effective tool for studying the degradation mechanisms of electrolyte solutions used for Li-ion battery technology (Abellan et al., 2014b) . This approach provides the possibility of rapidly screening electrolytes candidates for an improved stability upon reduction. For all of this, we believe that as the interest on understanding the interaction of the electron beam with a higher number of liquid systems increases so too do the possibilities of application and the range of dynamic processes that can be explored. Techniques allowing real-time, high-resolution imaging of nanoscale processes by in situ TEM are currently under rapid development based on microchip systems that enable real-time imaging of controlled processes to be combined with, for instance, correlated electrical measurements (Huang et al., 2010; Kallesøe et al., 2012) . In addition to gas-phase ETEM studies, microchip-based systems also allow the imaging of processes in liquids (Williamson et al., 2003) and recently with good resolution (Li et al., 2012; Liao et al., 2012) . Here, we report on our recent advances in joule heated microcantilevers (Molhave et al., 2008) used for gas-phase in situ TEM studies of nanowire growth (Kallesøe et al., 2010) and devices, and on our efforts in creating a suspended microchannel system for TEM imaging of liquid samples and processes. We created and characterized nanowire devices by growing silicon nanowires in situ TEM using the VLS mechanism (Ross, 2010) , from one joule heated microcantilever to an adjacent cantilever, thereby forming a bridge. Building on our previous experiments, we studied the details of contact formation when a hot nanowire grows into contact with a similarly heated adjacent cantilever. A silicon-to-silicon junction forms as the catalytic gold diffuses away. The suspended microcantilevers system opens up novel ways to create electrically contacted nanowire devices (Kallesøe et al., 2012) in addition to allowing in situ TEM studies of the many simultaneous processes taking place upon contact formation and in situ electrical measurements carried out after connected nanowires were formed.

To study liquids encapsulated between membrane windows on microchips, the reported microchip devices today all involve two bonded chips with thin membranes of silicon dioxide or nitride . To allow better control of the encapsulated volume and geometry, we are exploring devices using monolithic chips ( Figure 19 ) with suspended microfluidic channels made from silicon nitride (Jensen et al., 2014) . On the channels, thinned window regions allow higher resolution than in the supporting part of the channel. The system's novel geometry is expected to improve the imaging resolution by providing precise control over the channel height and the possibility of achieving small regions with ultrathin membranes for the best possible resolution. We have also developed an electrochemical setup suitable for use in SEM (Fig 2) ( Jensen, Køblera, Jensen, & Mølhave, 2013) . . Electron microscopy of specimens in liquid. Nature Nanotechnology, 6(11), 695-704. Kallesøe, C., et al. (2012) . In situ TEM Creation and electrical characterization of nanowire devices. Nano Letters, 12 (6) Figure 20 ). Nanoscale resolution is achievable for nanomaterials of a high atomic number (Z) in a low-Z liquid resulting from the Z contrast of STEM. It is also possible to study thin liquid samples using the environmental scanning electron microscope (ESEM) with STEM detector. Time-resolved Liquid STEM is feasible and can be used to study the movements of nanoparticles in liquid. We have found that the movement of nanoparticles in close proximity to the SiN membrane is up to three orders of magnitude slower than of Brownian motion in a bulk liquid . Liquid STEM is useful to explore growth processes of nanomaterials, such as gold dendrites (Kraus & de Jonge, 2013) .

The capability to achieve nanoscale resolution in liquid is of great interest for the study of membrane proteins in cells, which is challenging because the involved nanoscale dimensions require a spatial resolution beyond that of state-of-the-art fluorescence microscopy, while cells have to be prepared into thin solid samples for conventional electron microscopy preventing the study of whole cells. An example involves the dimerization of the EGFR, a transmembrane receptor playing a critical role in the pathogenesis and progression of many different types of cancer. An important question is under which conditions and in which cellular regions dimerization occurs. Liquid STEM provided a spatial resolution of a few nanometers to identify the protein complex subunits in the images while the cell remains intact and 

Andrei Kolmakov* Center for Nanoscale Science and Technology, NIST Gaithersburg, MD 20899-6204 *Corresponding author: e-mail address: andrei.kolmakov@nist.gov Electron microscopy and spectroscopy of complex liquid-gas, solid-liquid, and solid-gas interfaces in situ under realistic conditions is largely demanded in bio-medical, energy, catalysis research, microelectronics, and environmental studies. The performance of the traditional environmental scanning electron microscopy (ESEM), as well as novel membrane-based environmental cells for atmospheric pressure SEM, depends on electron mean free path in the ambient media and/or membrane itself, which usually is very small in solids, liquids, and dense gases. Novel 2D materials such as graphene and its derivatives have only a single atomic layer or a few thin layers and possess the unique combination of the mechanical strength (Lee et al., 2008) , molecular impermeability (Bunch et al., 2008) , optical and electron transparency (Stoll et al., 2012) in a wide electron energy range (see Figure 22 ). The latter makes these materials as prospective electron windows for ambient pressure SEM and electron spectroscopy, as well as for correlative imaging. In this report, we describe few tested designs of the graphene based environmental cells (GE-cell) along with protocols for their high yield fabrication (Krueger et al., 2011) . We compare these graphene windows with standard SIN membranes and demonstrate the capability to perform low-energy scanning electron microscopy (Stoll et al., 2012) , Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS) (Kolmakov et al., 2011) at atmospheric pressure using GE cells, which are hard to accomplish using standard approaches. The process of water radiolysis and bubble formation at the water-graphene interface was studied in vivo using SEM and XPS to demonstrate the capabilities of the technique (Figure 23 ; Kraus et al., 2014) . Sintering could be defined as the transformation of a powdered compact to a cohesive material under heating at high temperatures. It appears as a key-step in the preparation of ceramic materials such as nuclear fuels UOx and MOx (mixed oxide U and Pu). The sintering is usually described as having three different stages. The initial stage involves the elaboration of necks between the grains and leading to the cohesion of material, the intermediate and final stages are dedicated to the elimination of porosity between the grains by the means of grain growth mechanisms (Bernache-Assolant, 1993). The first stage of sintering is generally described through numerical simulation using simplified systems constituted by two or three spherical grains in contact (Wakai, 2006) . Presently, only few experimental works are devoted to the kinetics of necks elaboration (corresponding to the first stage of sintering) using ex situ TEM (Qin et al., 2010) and SEM (Slamovich & Lange, 1990) observations. These studies allowed experimental investigations of this stage of sintering on metallic materials with spherical grains.

In this study, we report the first experimental observations of the initial stage of sintering from grains of lanthanide oxides with controlled shapes by using environmental scanning electron microscopy (ESEM) at high temperatures (HT-ESEM). Actually, the use of HT-ESEM allowed the in situ observation of the samples during long term heat treatments up to 1400°C under various atmospheres (Podor et al., 2012) . In a first step, lanthanide and actinide oxide grains with spherical shapes were synthesized by means of soft chemistry (Nkou Bouala et al., 2014) , in order to investigate chemical systems with shapes close to those used in models. Then the HT-ESEM was used to investigate the first stage of sintering on three different systems (a single grain and two and three grains in contact) in the temperature range of 900-1200°C:

• Monitoring of a single grain led to the evolution of the number of crystallites included in the sphere. From the micrographs series, the time necessary to reach a spherical single crystal was determined, as well as the activation energy necessary for the growth of crystallites. • The observation of the morphological modifications of two-and threegrain arrangements led to assess the evolution of several parameters of interest, such as neck size, dihedral angles between the spheres, and distance between the grain centers. From the micrographs series, it was possible to identify experimentally for the first time the mechanisms of neck growth between the grains (Figures 24-25 ). The use of HT-ESEM observations appears to be of great interest for the study of sintering phenomena. The exploitation of micrograph series allows for determining original and fundamental experimental data and characteristics of the processes occurring during the initial stage of sintering. Figure 24 In situ HT-ESEM micrographs series of two CeO 2 nanospheres showing the evolution of neck size and dihedral angles between the grains at T ¼ 1100°C.

In situ HT-ESEM micrographs series of three CeO 2 nanospheres showing the evolution of neck size, dihedral angles and porosity between the grains at T ¼ 1100°C.

Podor, R., Clavier, N., Ravaux, J., Claparède, L., Dacheux, N., & Bernache-Assollant, D. (2012) . Dynamic aspects of cerium dioxide sintering: HT-ESEM study of grain growth and pore elimination. Journal of the European Ceramic Society, 32, 353-362. Qin, J., Yang, R., Liu, G., Li, M., & Shi, Y. (2010) . Grain growth and microstructural evolution of yttrium aluminium garnet nanocrystallites during calcination process. Materials Research Bulletin, 45, 1426 -1432 . Slamovich, E.B., & Lange, F.F. (1990 . Densification behavior of single-crystal and polycrystalline spherical particles of zirconia. Journal of the American Ceramic Society, 73 (11), 3369-3375. Wakai, F. (2006) . Modeling and simulation of elementary process in ideal sintering, Journal of the American Ceramic Society, 89 (5), 1471-1484.

The direct observation of atomic processes during structural phase transitions or grain boundary motion is of fundamental importance for understanding the properties of phase change materials. In situ transmission electron microscopy (TEM) enables excellent spatial and analytical resolution allowing the acquisition of data about the atomic structure in combination with heating experiments. The system under study is the ternary Ge-Sb-Te alloy, which is characterized by an amorphous state at room temperature and phase transitions to cubic (about 150-165°C) (Seo et al., 2000; Friedrich et al., 2000) and hexagonal (about 340-360°C ) (Friedrich et al., 2000; Kooi et al., 2004) crystal structures at higher temperatures.

The Ge-Sb-Te (GST) thin films were deposited on a cleaned Si-(111) substrate in an ultra-high vacuum chamber by using effusion cells for the three elements. One sample was prepared at room temperature and subsequently heated to form a polycrystalline film, the other was epitaxially grown on Si (Rodenbach et al., 2012) . Cross-sectional TEM samples were prepared using the standard procedure of mechanical grinding, dimpling and ion beam milling. In situ TEM is performed with a JEOL 3010 operated at 200 kV using a JEOL double-tilt specimen heating holder and controller unit. The temperatures stated here referred to the temperature given by the controller unit. The temporal resolution of the recorded videos are 0.04 s.

A thin polycrystalline Ge 1 Sb 2 Te 4 film (composition confirmed by energy electron loss spectroscopy) is annealed to 400°C for in situ observation of the motion of grain boundaries. Dark-field imaging conditions are used for achieving an accurate determination of the grain boundary position. Figure 26 shows a sequence of dark-field snapshots taken from a video, where the bright grain is characterized by a preferential orientation to the substrate: hexagonal basal planes are parallel to Si-(111) planes. During the grain boundary motion, the preferential grain grows as measured by the distance d 0 between the actual boundary position and an obstacle at the interface versus time. The corresponding graph in Figure 27 reveals a steplike character instead of a continuously increasing progression. This behavior could be caused by the cubic-to-hexagonal phase transition accompanying the grain boundary motion with the formation of the complex layered structure (Kooi et al., 2002) of hexagonal Ge 1 Sb 2 Te 4 .

Another important aspect of phase transitions addresses the crystallization dynamics and the evolution of the growth front morphology. An epitaxially aligned cubic Ge 2 Sb 2 Te 5 film (Rodenbach et al., 2012) is annealed at about 400°C. At the interface region, unexpected amorphization and recrystallization is observed under e À -beam irradiation. The crystallization proceeds along the highly stepped surface similar to the step-flow growth mode during epitaxy on vicinal surfaces ( Figure 27 ). However, within our temporal resolution, the multiple steps grow both block by block and in a two-layer Figure 26 Dark-field TEM images taken from a video sequence showing the recrystallization of GST thin film at 400°C: The grain with the preferred orientation of basal planes parallel to Si (111) planes grows laterally. The graph shows the result of the growth analysis: a stepwise and fast growth process. growth, rather than by simple atom migration. This behavior could again be caused by the complex structure. In order to get the right composition, the material needs at least two layers to replicate the surrounding matrix. 

In recent years, the environmental transmission scanning electron microscope (ESTEM) has been successfully employed to elucidate the structural and chemical changes occurring in the catalyst nanoparticles under reactive environments. While atomic-resolution images and the combination of high spatial and energy resolution is ideally suited to distinguish between active and inactive catalyst particles and identify active surfaces for gas adsorption, unambiguous data can only be obtained from the area under observation. This lack of global information available from TEM measurements is generally compensated for by using other, ensemble measurement techniques such as X-ray or neutron diffraction, X-ray photoelectron spectroscopy, infrared spectroscopy, and Raman spectroscopy. However, it is almost impossible to create identical experimental conditions in two separate instruments to make measurements that can be directly compared. Moreover, ambiguities in ESTEM studies may arise from the unknown effects of the incident electron beam and uncertainty of the sample temperature. We have designed and built a unique platform that allows us to concurrently measure atomic-scale and micro-scale changes occurring in samples subjected to identical reactive environmental conditions by incorporating a Raman spectrometer on the ESTEM. We have used this correlative microscopy platform (1) to measure the temperature from a 60-μm 2 area using Raman shifts, (2) to investigate light-matter interactions, and (3) as a heating source for concurrent optical and electron spectroscopy, such as cathodoluminescence, electron energy loss spectroscopy (EELS), and Raman. Details of the design, function, and capabilities will be illustrated with results obtained from in situ combinatorial measurements. ETEM is a unique tool combining imaging and spectroscopy capabilities with a spatial resolution approaching 1 Å at elevated temperatures and controlled gaseous environments. As with any other characterization technique, ETEM has drawbacks. Limitations in sample geometry and field of view call for use of complementary in situ characterization techniques. Such techniques could be in situ X-ray based techniques such as X-ray diffractometer (XRD), X-ray absorption spectroscopy (XAS), and extended X-ray absorption fine structure (EXAFS). The combination of techniques will paint a more complete picture of the sample under a reactive environment.

In order to take full advantage of the complementary in situ techniques, transfer under reaction conditions is essential. Here, we introduce the in situ transfer concept by use of a dedicated TEM transfer holder capable of enclosing the sample in a gaseous environment at temperatures up to approx. 900°C. The holder is compatible with other in situ technique set-ups (Damsgaard et al., 2014) . Figure 28 shows the concept of the transfer holder. In this experiment, cuprous oxide has been initially characterized in the ETEM at room temperature in a vacuum, then transferred to an in situ XRD setup, reduced in hydrogen, and finally transferred back to the microscope in 10 5 Pa of H 2 at 220°C.

In general, the catalytic behavior of nanoparticle systems can be greatly influenced by the synthesis procedure used. The different stages of the **Present address: Haldor Topsøe A/S, 2800 Kgs. Lyngby, Denmark formation process, such as drying, calcination, and reduction, can all be optimized by tweaking parameters such as temperature, time of treatment, and gaseous environment. Following the process of identified volume subsets of the catalyst material by means of electron beam-based characterization at the different stages of synthesis gives insight in the formation mechanisms of individual nanoparticles, from precursor to active catalyst. Converting solar energy into chemical bonds, and thereby increasing the storage capability of light energy harvesting, require the use of photocatalysts. In order to characterize and study such catalysts under relevant conditions, we have developed a TEM holder capable of exposing a sample to visible light inside the microscope. In this way, photo induced reactions and phenomena such as Cu 2 O degradation ( Figure 29 ) are studied in situ under different gaseous environments (Cavalca et al., 2013) .

We have used two experimental approaches to study the motion of gold nanoparticles in liquid: scanning transmission electron microscopy in liquid (liquid STEM), and environmental scanning electron microscopy (ESEM) equipped with a wet-STEM detector . In liquid STEM, a microfluidic chamber assembled from two Si microchips with electron-transparent SiN windows is used to enclose the nanoparticle solution, and separate it from the vacuum in the microscope chamber ( Figure 30A ). The images were recorded at 200 keV in STEM mode using the annular-dark-field (ADF) detector providing Z-contrast. With this method, the particles show a high contrast compared to the background signal originating from the liquid layer. A spatial resolution of 1.5 nm was achieved for gold nanoparticles . In comparison to liquid STEM, there is no need of a microfluidic chamber in ESEM STEM ( Figure 31B ). The nanoparticles were imaged in a droplet of water on an SiN window that was cooled to 3°C. To maintain a stable liquid layer thickness, the water vapor pressure in the microscope chamber is adjusted to 750 Pa. To image the particles, we used 30 keV, resulting in a decreased but still nanoscale resolution compared to STEM at 200 keV.

For the liquid STEM experiments, 30-nm-diameter gold nanoparticles were investigated in a glycerol solution with 20% water. This solution has a higher viscosity compared to pure water and therefore slows the nanoparticle movement. Nanoparticles stabilized by a coating of positively charged thiolated chitosan (TCHIT) were used. The electron irradiation induced the detachment of the particles from the SiN membrane, and we observed a subsequent movement and agglomeration of the particles (Figure 31) . A bubble was formed during imaging by the interaction of the liquid with the electron beam. As a result, the liquid layer thickness in the bubble region was decreased, and we observed that the velocity of the nanoparticles depended on the thickness of the liquid layer. The thinner the liquid layer, the slower the particles move. The calculated theoretical displacement due to Brownian motion was more than two orders of magnitude higher than our experimental data. In the ESEM STEM experiments, we imaged 30-nm, citrate-stabilized nanoparticles in pure water. Two different regimes of velocities were observed, very fast ones and much slower ones. Interestingly, even the velocities of the fastest nanoparticles were much smaller than expected on the basis of Brownian motion. 

Brownian motion slowing down the motion of nanoparticles in liquid. Possible explanations could involve van der Waals forces or electrostatic interactions between the gold nanoparticles and the SiN window or the surface of the droplet . The reduced velocities compared to Brownian motion are beneficial for electron microscopical investigations of particles in liquid since nanoscale resolution is not much degraded by the movement of the nanoparticles within the image acquisition.

Nanomaterials have been promised a great future and are already being used in various applications. The recently updated definition of the term nanomaterial, to be used in all European Union (EU) legislation, is "A nanomaterial means a natural, incidental, or manufactured material containing particles, in an unbound state or as an aggregate or as an agglomerate and where, for 50% or more of the particles in the number size distribution, one or more external dimensions is in the size range 1 nm-100 nm." In this definition, a nanoparticle has all three dimensions in the nanoscale. Because their physical and chemical properties depend very much on their size and shape, their characterization is of utmost importance. An in situ observation technique such as dynamic light scattering (DLS) is commonly used to probe the dispersion state of colloids dispersed in a fluid. This technique is very precise in determining mean particle size and its distribution for monomodal particle samples. But when multimodal distributions evolve (for example, induced by aggregation), the quantification becomes challenging due to the assumptions made in the applied models. Unfortunately, the median of the number-weighted size distribution, which has become an instrumental parameter in nanomaterial's legislation, cannot be retrieved by scattering methods. Transmission electron microscopic (TEM) analysis can provide model-free data and the number size distribution. However, in order to obtain TEM samples of nanoparticles in a fluid, the solvent (mostly water) must be removed. Drying effects will take place during sample preparation, which leads to the accumulation and aggregation and hence introduces bias or artifacts in the analysis (Michen et al., 2014) . It is impossible to differentiate nanoparticle aggregates caused by drying effects from aggregates that were formed prior to drying (e.g., due to colloidal instability of the system). Thus, TEM has long been thought to be unsuitable for quantitative investigations of nanoparticles and colloidal aggregates of nanoparticles in liquids. Here, we present a straightforward, cost-effective protocol that circumvents the drying issue ( Figure 32 ) by stabilizing the colloidal system using a macromolecular agent bovine serum albumin (BSA). TEM samples with various ratios of proteins to nanoparticles have been prepared and resulted in different spatial distributions of nanoparticles on the TEM grid. When a sufficient high-protein concentration was used, the resulting TEM images allow for automated data collection on a large number of nanoparticles which is in very good agreement with data obtained from dynamic light scattering and thus the actual dispersion state. With the introduction of commercially available transmission electron microscopy (TEM) holders for liquid specimens in recent years, the interest in liquid-cell TEM has greatly risen. Nowadays, this method is applied in scientific fields as different as battery research, cell biology, and materials science in order to gain information about structures and processes on the nanoscale in a liquid or the natural environment (de Jonge & Ross, 2011). However, the presence of liquids and, in most cases, also solutes dramatically increases the number of chemical reactions conceivable during an experiment in the electron microscope compared to TEM in a vacuum. In addition to interactions with solid materials, as known from conventional electron microscopy experiments, the electron beam influences the chemistry of the liquid. Electrons penetrating through the liquid excite electronic states in molecules or atoms or remove electrons from the electron shell, resulting in the generation of highly reactive, transient species (Henglein et al., 1969) . Some of these compounds are strong-reducing (e.g., solvated electrons) others are powerful oxidizing agents (e.g., hydroxyl radicals), that are capable of dissolving metals as noble as gold. These by-products are often not desired, as they may have severe consequences on the outcome of liquid cell transmission electron microscopy experiments, contingently leading to entirely unexpected results. Hence, a better understanding of these electron beam-induced processes, and avoiding them is needed in order to design experiments comparable to those without electron beam exposure.

In this study, we focused on the impact of the electron beam on gold nanoparticles in an aqueous environment. We utilized TEM experiments in scanning mode (e.g., STEM) in order to obtain a good handle on the dose introduced into the observed sample area. By varying the properties of the solution (e.g., the salt concentration or pH) and the imaging conditions, we altered the composition of the transient species generated by the electron beam . It became apparent that the addition of high concentrations of chloride ions, destabilized gold nanoparticles. Moreover, the size of the nanoparticles influenced the susceptibility toward dissolution, with smaller particles dissolving faster than bigger ones (Figure 33 ). Generally, in order to make the water and carbon energy cycles economically competitive in terms of today's fossil and nuclear energy, energy conversion materials have to be further optimized and therefore studied in more detail. More specifically, the fundamental details of platinum-based proton exchange membrane fuel cell (PEM-FC) catalyst degradation are still not clear (Mayrhofer et al., 2008; Meier et al., 2012) . Missing insights can only be gained with an introduction of new advance characterization techniques like identical location transmission electron microscopy (IL-TEM) and in situ electrochemical TEM. A generally accepted technique, besides electrochemistry, used to study the structure of nanoparticles is a vacuum TEM equipped with numerous detectors enabling many characterization methods, such as electron diffraction, high-resolution TEM (HR-TEM), scanning TEM (STEM), and energy dispersive spectroscopy (EDS), which provide a complete description of studied nanomaterial characteristics. This information is commonly used to interpret nanoparticle activity and stability. However, the concerns still remain whether the surface structure and morphology change of random spots is representative for the whole sample and if the structure and morphology integrity get disrupted when nanoparticles come in contact with the atmosphere or get immersed in the liquid electrolyte. It can reconstruct or deconstruct the metal surface depending on the interaction energy of thermodynamic driving forces. It can also change unevenly across the sample. In order to access the information of real surface structure at relevant conditions, identical locations and, more important, in situ observations are mandatory. Before IL-TEM was the only technique showing degradation on the same nanoparticle (Hodnik et al., 2014; Mayrhofer et al., 2008; Meier et al., 2012) . Combination of this approach with conventional electrochemistry provides a complete description of the catalytic activity and stability in relation to the surface structural characteristics.

In this study, we present dealloying of polydisperse PtCu 3 nanoparticles. IL-TEM images ( Figures 34A and B) show that particles larger than 20 nm form a porous structure upon a potential hold experiment (1.2 V vs RHE for 2 h in 0.1 M HClO 4 ) (Hodnik et al., 2014) . We provide the first preliminary results done with electrochemical in situ TEM (Poseidon 500, Protochips) confirming the pore formation upon electrochemical biasing ( Figure 34C and D). The implementation of in situ liquid-phase electron microscopy has increased significantly in the last decade . Several chemical reactions, with some induced by the electron beam, have been observed in situ (Radisic et al., 2006; Williamson et al., 2003; Zheng et al., 2009) , but the effects of the electron beam on these experiments is still not fully understood .

The electron beam will generate redox active species, as well as local changes in pH, and in turn affect the local chemical environment being imaged. In this study, we tested the influence of irradiation on the fluorescence signal from a liquid sample in the TEM (Figure 35 ). The combined TEM/fluorescence microscope imaging modes offered by the FEI Technai iCorr was used to image a vacuum compatible ionic liquid (1-butyl-3-methylimidazolium tetrafluoroborate) mixed with two fluorescent dyes: fluorescein and Nile blue. The two dyes were mixed into the ionic liquid such that the fluorescent signal could be imaged with the fluorescence microscope.

The investigation showed varied results. In the fluorescein sample ( Figure 35A ), the fluorescent signal was increased after electron exposure. The ionic liquid became polymerized, as evidenced by the irradiated areas not gradually becoming diffuse after electron exposure. For the Nile blue sample, no such evidence was found. Rather, the fluorescent signal deteriorated over time, which indicates that the reaction stabilized or that the fluorescent dye was replenished by diffusion.

When performing experiments at different dose rates, a nearly linear correlation between intensity loss and dose rate was observed for the fluorescein sample, but the Nile blue showed no correlation. Plunge freezing is commonly used as a cryofixation technique for the stabilization of biological materials prior to electron microscopy (EM) (Dobro et al., 2010; Milne et al., 2013) . However, even if plunge freezing is the easiest way to retain the original structure of the sample, it is very difficult to achieve the high cooling rates (Ryan et al., 1992) necessary for obtaining amorphous ice and thus avoiding freezing artefacts such as ice crystal distortions. We present the development of a low-mass, microchip-based system for the measurement of ultra-fast freezing rates during cryo-immobilization of small hydrated volumes such as living cells and thin liquid films. Our system enables the change in temperature to be tracked accurately over the short timescales involved, by measuring the resistance of metallic thin-film resistors lithographically defined on the membrane. The chip system was used to determine cooling rates in slush nitrogen, propane and a mixture of propane and ethane with and without liquid samples. We measured cooling rates of over 100.000 K/s (see Figure 36 ) and heat transfer coefficients over 10.000 W/m 2 K on the dry chip, with lower values when liquid layers were present, and occasionally even higher values in slush nitrogen. Electroplating of copper is a widely used technique for aesthetic coatings, creating electrodes, and precoating prior to additional electroplating. Recently it has become possible to directly see the growth on a nanometer scale. Theory on the nucleation density laid down by Hyde and Compton (2003) was augmented by Radisic et al. (2006a Radisic et al. ( , 2006b after they found by means of in situ TEM that the nucleation density was three orders of magnitude more than predicted. More recently, studied the effect of depletion on growth morphology.

While Radisic et al. (2006a, b) . used a TEM chip with a very limited volume and looking through 1-2 μm of electrolyte, we are using an in situ SEM setup that allows us to have an electrolyte volume on the order of 1 mL . A larger total volume and more open liquid cell geometry ensures a better diffusive supply from the bulk, which means that the electrolyte close to the electrode remains more representative of the bulk for a longer duration of electron radiation and electrochemical depletion.

In addition to studying the morphology in different deposition current regimes, we investigated how the electron beam affects the experiment by blanking the beam during deposition and dissolution and relating its effect to changes in deposition rate and coating topography. Figure 37A shows the schematic layout of our SEM setup during electroplating. An Au thin-film electrode on an electron transparent silicon nitride membrane serves as the working electrode. Copper was electroplated from an acidic copper sulfate solution and imaged in situ ( Figure 37B ). This resolution approaches that of despite the lowresolution tungsten filament SEM used. Higher resolutions will be achieved with FEG SEM in future experiments. After the deposition, the morphology of the flushed and dried Cu films was inspected in SEM ( Figures 37C and D) .

Experiments performed outside the SEM obtained a very high correlation between deposit thickness and electrical charge, as indicated by the results in Figure 38 . From integration of the current, we calculated that the layer of deposited copper had a thickness of 5.1 μm ( Figure 38A ). During dissolution ( Figure 38B ), we measured the same charge before all the deposited material was consumed and the electrode became passivated.

In summary, we have developed an electrochemical cell for in situ SEM to serve as a flexible alternative to TEM when large liquid volumes are advantageous. We were able to correlate the charge to the measured film thickness and study the copper growth on the Au electrode. Figure 38 Electroplating was demonstrated by optical imaging on a gold electrode (A). Layer thickness was calculated based on the total charge during the deposition. During dissolution (B), the current dropped to zero as the copper became completely dissolved. The discrepancy in copper dissolved versus copper plated occurred with the second decimal, which indicates that potentially unwanted side reactions are not a concern. (C) and (D) show the in situ optical microscopy growth of copper on an electrode. that are first identified and localized with WF, with limited EM resolution loss (Liv, 2014) . However, SEM only investigates the upper area of the holder, whereas WF collects fluorescence from different sample depths in the imaging area. Hence, a direct correlation between WF and liquid EM image is relatively inaccurate due to differences in z-resolution.

An improved correlation in iCLEM requires the optical sectioning capabilities of confocal microscopy (CM). For this reason, we are currently working on an integrated setup that combines SEM with CM by means of a commercial Nikon C2 module ( Figure 39C ). We first plan to demonstrate this setup by studying membrane proteins on the cell surface. The z-resolution of 0.8 μm achievable with CM will allow localization and functional studies of membrane receptors in this iCLEM system. Further development of integrated microscopes like this one will allow CLEM to become a powerful tool for industrial applications, fundamental biological research, and medical diagnostics.

This work is in collaboration with Delmic BV. We want to thank Ruud van Tol and Carel Heerkens for their assistance, and STW for financial support. (Liv et al., 2014) .

Helminth parasite infections may cause major diseases in humans and animals, so they are prime targets of veterinary and medical research. For morphological studies of parasites, as a part of taxonomy, SEM is commonly used. Sometimes only a small number or only one sample of a rare parasite is available and cannot be used in conventional SEM because the sample has to be fixed, dehydrated, and coated before it can be observed. SEM condemns samples to be destroyed so that further analysis is not possible (e.g., for molecular study or depositing as type material in a museum). Environmental scanning electron microscopy (ESEM) based on new methods (Neděla 2010; Neděla et al., 2015) and instrumentation (Jirák et al., 2010) can be used for the advanced study of parasites in their native state. It was shown by Tihlaříková, Neděla, and Shiojiri (2013) in a study of surviving mites. Only a small number of parasites were investigated using ESEM, having benefits in the unnecessary preparation to conventional SEM, which considerably reduces their post-SEM value as described by Buffington, Burks, & McNeil (2005) in insects.

The aim of this work is to show new results of two groups of helminth parasites (Nematoda, Contracaecum osculatum; and Acanthocephala, Corynosoma pseudohamanni) with minimal shape and volume distortions (keeping them acceptable for taxonomical research) and to emphasize the advantages of dynamical in situ experiments in this field of science.

The most suitable treatment and optimal environmental conditions for observing selected parasites were found. Already-fixed samples (70% ethanol or 4% formaldehyde solution) were washed and rinsed with a drop of distilled water. Observations were made using the experimental ESEM AQUASEM II (Neděla, 2010) equipped with an ionization detector of secondary electrons, a specially designed hydration system and Peltier cooled specimen holder. A unique method for reaching and keeping thermodynamic conditions was used (Tihlaříková, Neděla, & Shiojiri, 2013) . The samples were cooled to 1°C and observed in a high-pressure water vapor environment of 690-670 Pa, probe current of 90 pA, and beam accelerating voltage of 20 kV. Samples were placed on a Peltier cooled silicone holder into a drop of water. Consequently, the water was slowly evaporated from the sample (see Figures 41 and 42 ). Using our ESEM, it is possible to keep and observe susceptible wet samples for a longer time in low-beam-current conditions. Figure 41 Detail of the proboscis of Corynosoma pseudohamanni and its sequential drying documented by ESEM AQUSEM II. Scale bar: 100 μm. Observation parameters were as follows: cooling temperature 1°C; pressure of water vapor 680 Pa; distance between the sample surface and the second pressure-limiting aperture 2.7 mm; accelerating voltage 20 kV; and probe current 90 pA.

Sequential drying of Contracaecum osculatum documented by ESEM AQUSEM II. Scale bar: 100 μm. Observation parameters were as follows: cooling temperature 1°C; pressure of water vapor 690-670 Pa; distance between the sample surface and the second pressure-limiting aperture 2.7 mm; accelerating voltage 20 kV; and probe current 90 pA.

In this study, we show that ESEM allows the examination of specimens in a fully hydrated state and with minimal previous treatment. The surface was imaged in a wet state without the covering of a thick water layer to reveal microstructural specifics of samples. In situ observation of nonconductive parasitological samples was free of charging artefacts, and the visibility of topographical structures was sufficient for taxonomical research with no occurrence of artefacts and shape deformities. Eight time-lapse measurements of four lengths (1-lower end of papilla to start of ridge on body; 2-from one edge of the interlabium to the second; 3-height; and 4-length of head papilla; see Figure 43 ) under stable pressure conditions and low local heating by a low-probe-current electron beam were performed. Negligible changes on nematode parasite tissues were proven by measurement ( Figure 43 ). Objects are not deformed and remain in almost the same state.

This method can be used effectively in morphological and morphometrical studies on parasites where valuable and unique specimens sometimes exist in a small number. A very slow and well controlled decrease in humidity allows the visualization of the surface topography of the sample to be free of distortions and changes in shape or dimension. Samples are well preserved from a taxonomical point of view as well. The optimal state of the sample is shown in Figure 42G . Highly visible morphology is shown in Figure 42H , but to the detriment of morphometrical data, as shown in Figure 42H . Both Figures 42G and h show a lateral view of the well-retained subventral lip equipped with one elliptical single papilla and amphid. Interlabia with undivided tip are also highly visible, as is the cuticular striation behind the lips. The suitability of ESEM for the study of parasites in a wet state and with minimal preparation was proved by our results. We introduce advanced ESEM methods continuing from the previous work by Lopes Torres et al. (2013) , Maia-Brigagão and de Souza (2012), and Mašová et al. (2010) . This method also will be applied with other types of parasites, and future research will be using ESEM with live samples of parasites. Structures with hollow interiors, such as hollow nanospheres, have recently received considerable scientific attention owing to their unique properties, which could facilitate breakthrough applications in various fields of nanoscience. Physical processing routes like pulsed-laser ablation (PLA) in the presence of a background gas have tremendous potential for applications because they offer flexibility in terms of the choice of materials to be ablated and, at the same time, the ability to produce well-defined, hollow nanospheres. It was shown previously that complex hollow nanospheres filled with nitrogen gas can be produced via a single-step procedure by the ablation of a metallic or alloy-based target into ambient nitrogen gas (Šturm et al., 2010, 2013) .

The spatially resolved electron energy loss spectroscopy (EELS) analyses performed in the void and in the shell region of the hollow nanospheres consistently showed the presence of nitrogen only in the voids ( Figure 44A ). Figure 44B shows the resulting N-K edge for Co-Pt and Fe-(SmTa) systems, which is presented together with the reference standard spectrum of N 2 gas obtained from air. The fine structure for all these N-K edges is distinctive for molecular nitrogen and is characterized by a sharply peaked edge at 401 eV, followed by a broad continuum. Following these results, the nanospheres observed in HAADF-STEM images ( Figure 44A ) were defined as hollow spheres filled with nitrogen gas.

To unambiguously determine if measured N signal corresponds to the presence of nitrogen gas trapped inside the hollow spheres, time-resolved EELS measurements combined with the HAADF-STEM imaging were performed. EEL spectra were acquired as a function of time until the perforation of the wall of the sphere, which was achieved by the intense electron probe ( Figure 45A ). The corresponding background subtracted N-K edges measured as a function of the exposure time are shown in Figure 45B . The N-K edge signal remains nearly constant in the first 20 s and drops abruptly to the background noise in the time frame of about 1 s, which clearly demonstrates the release of nitrogen gas from the interior of the sphere through the pinhole created by the intense electron probe. Quantitative analysis of the nitrogen density and pressure in Al, Fe-(SmTa), and Co-Pt systems was additionally performed. The number density of nitrogen atoms (n) in the voids of the nanospheres was determined by the following equation [2]: n ¼ I N /(σ N I ZL d), where I N and I ZL are the intensities measured from the N-K ionization edge and the zero-loss peak, respectively; σ N is the angle-integrated hydrogenic cross section for the nitrogen K-shell ionization, calculated for the experimental collection angles; and d represents the measured diameter of the void. To calculate the pressure in the void accurately, with respect to the given density and the temperature range, a standard correction of an ideal gas law using a virial expansion was applied, as follows: P ¼ nkT(1 + nB/N A + n 2 C/N A 2 + . . .), where k represents the Boltzmann constant; T is the absolute temperature,4; N A stands for the Avogadro constant; and B and C are the second and third virial coefficients, respectively. The calculated pressures ranged between 25 bars for the Co-Pt system up to 450 bars in Fe-(SmTa) and Al systems, respectively. The total error of the calculated pressures amounts to 17%. The obtained results support the idea that gasfilled hollow spheres could be fabricated in various complex metallic systems by applying PLA in the presence of a background gas, taking into consideration that in relation to the background gas high-solubility differences between the melt and corresponding solids are achieved. Blood platelets are anucleate cells that play a central role in the arrest of bleeding after a blood vessel is damaged. Platelets circulate at variable velocities in the bloodstream, and are activated following binding to subendothelial components [von Willebrand factor (vWF), collagen] that are exposed during vascular injury. Stable adhesion to collagen promotes the release of soluble components from platelet secretory granules, leading to the formation of a thrombus. Platelets also play a crucial role in the inflammatory response (Boilard et al., 2010; Semple et al., 2011) , through mechanisms that are less well characterized.

Platelet secretory alpha-granules contain adhesive proteins (vWF, fibrinogen), pro-and anti-inflammatory mediators (PF4, beta-TG), and luminal membranes harboring cytosolic components (IL1β). Timed release of granule content and cell surface expression of integral membrane components are important determinants for platelet contribution in modulating haemostatic and inflammatory responses. Platelet secretory granules are heterogeneous in morphology and content, and this heterogeneity may give rise to differential release patterns. The molecular mechanism that regulates the fine-tuning of alpha granule cargo release is not known.

The relative small thickness of adherent platelets (less than 500 nm) makes them ideally suited for whole cell electron tomography analysis. Here, we describe two CLEM methodologies to study platelet adhesion dynamics and cargo release at the ultrastructural level. Cryo-electron microscopy (cryo-EM) allows for the visualization of cells in a close-tonative state, with nanometer-scale resolution (Faas et al., 2013) . Due to low contrast and low signal-to-noise ratio in images of frozen hydrated samples, it often appears difficult to localize structures of interest within heterogeneous samples. We have used iCorr-integrated correlative microscopy with high-resolution (cryo) electron tomography to study the secretory behavior of human platelet alpha granules. iCorr is designed to automate and accelerate CLEM experiments, resulting in fluorescent labeling information combined with 3D ultrastructural information.

Platelets, isolated from freshly drawn human whole blood were allowed to settle on fibrinogen coated copper grids, and stimulated with collagen related peptide. Flow devices specially developed for simultaneous light microscopy imaging and electron microscopy transfer were used to follow membrane dynamics and granule release patterns under physiological flow. Release of vWF was detected using combined immunolabeling with a fluorophore and protein A gold.

Our results show that we are able to correlate fluorescent signals, arising from released vWF, to the apical cell surface of the adherent platelets (Figures 46-47) . From dual-axis tomography analysis on flow-adherent cells, it appeared that subsets of vWF-containing granules migrate in a downstream fashion, while others remain stationary in the cell center. Downstreamoriented granules release vWF as strings in the direction of the flow ( Figures 46A,B) . Similar results were obtained when adherent platelets were vitrified in liquid ethane, followed by correlative analysis by the EM-integrated iCorr workflow (FEI Company) ( Figure 47 ). Low-dose tomographic analysis revealed subtle substructural modifications at the limiting alpha granule membrane that had not been observed before ( Figure 47D ). Whether these subtle membrane changes, which were visualized only after vitrification, give us a clue how platelets manage to differentially release their granule contents remains to be seen. 

Electron microscopy of biological materials at the nanometer scale

Visualizing cells at the nanoscale

Structural studies by electron tomography: From cells to molecules

The molecular sociology of the cell

Opening windows into the cell: Focused-ion-beam milling for cryo-electron tomography

Bridging fluorescence microscopy and electron microscopy

Destruction of tissue, cells and organelles in type 1 diabetic rats presented at macromolecular resolution

Capturing endocytic segregation events with HPF-CLEM

Retracing in correlative light electron microscopy: Where is my object of interest?

Development of a quantitative correlative light electron microscopy technique to study GLUT4 trafficking

Moving EM: The rapid transfer system as a new tool for correlative light and electron microscopy and high throughput for high-pressure freezing

Epidermal growth factor receptor subunit locations determined in hydrated cells with environmental scanning electron microscopy

Gold nanoparticle uptake in whole cells in liquid examined by environmental scanning electron microscopy

Integrated fluorescence and transmission electron microscopy: a novel approach to correlative microscopy

Integrated laser and electron microscopy correlates structure of fluid catalytic cracking particles to Brønsted acidity

Optimizing immuno-labeling for correlative fluorescence and electron microscopy on a single specimen

Novel contrasting and labeling procedures for correlative microscopy of thawed cryosections

Microscopy Valley Project

Daclatasvir-like inhibitors of NS5A block early biogenesis of HCV-induced membranous replication factories, independent of RNA replication

Three-dimensional architecture and biogenesis of membrane structures associated with hepatitis C virus replication

Ion-association complexes unite classical and nonclassical theories for the biomimetic nucleation of calcium phosphate

Direction-specific interactions control crystal growth by oriented attachment

situ TEM imaging of CaCO 3 nucleation reveals coexistence of direct and indirect pathways

Investigating processes of nanocrystal formation and transformation via liquid cell TEM. Microsccopy and Microanalysis

Taking advantage of disorder: Amorphous calcium carbonate and its roles in biomineralization

In situ techniques in biomimetic mineralization studies of calcium carbonate

Imaging of self-assembled structures: Interpretation of TEM and cryo-TEM images

Biomimetic model systems for investigating the amorphous precursor pathway and its role in biomineralization

On Biomineralization

Biomineralization, Principles and Concepts in Bioinorganic Materials Chemistry

Controlling mineral morphologies and structures in biological and synthetic systems

Biomimetic systems for hydroxyapatite mineralization inspired by bone and enamel. Special issue on biomineralization

Calcium carbonate nucleation driven by ion binding in a biomimetic matrix revealed by in situ microscopy

Unraveling the CaCO 3 mesocrystal formation mechanism including a polyelectrolyte additive using in situ TEM and in situ AFM

Biomimetic CaCO 3 mineralization using designer molecules and interfaces

A C-RING-like domain participates in protein self-assembly and mineral nucleation

Polysaccharides and proteoglycans in calcium carbonate-based biomineralization

Characterization of two molluscan crystal-modulating biomineralization proteins and identification of putative mineral binding domains

An oligomeric C-RING nacre protein influences prenucleation events and organizes mineral nanoparticles

Factors influencing quantitative liquid (scanning) transmission electron microscopy

Probing the degradation mechanisms in electrolyte solutions for Li-ion batteries by in situ transmission electron microscopy

Controlled growth of nanoparticles from solution with in situ liquid transmission electron microscopy

Bubble and pattern formation in liquid induced by an electron beam

In situ liquid cell electron microscopy of the solution growth of Au-Pd core-shell nanostructures

Electron microscopy of specimens in liquid

Electron-water interactions and implications for liquid cell electron microscopy

In situ observation of the electrochemical lithiation of a single SnO2 nanowire electrode

In situ SEM microchip setup for electrochemical experiments with water-based solutions

Monolithic chip system with a microfluidic channel for in situ electron microscopy of liquids

On the center of the chip is a 50-nm-thick suspended membrane with a microchannel and optional components such as heaters. (B) The ECSEM setup with the lid at the top, the chip holder in the middle

Electron microscopy of whole cells in liquid with nanometer resolution

Electron microscopy of specimens in liquid

Dendritic gold nanowire growth observed in liquid with transmission electron microscopy

Epidermal growth factor receptor subunit locations determined in hydrated cells with environmental scanning electron microscopy

Liquid scanning transmission electron microscopy: Imaging protein complexes in their native environment in whole eukaryotic cells. Microscopy and Microanalysis

Video-frequency scanning transmission electron microscopy of moving gold nanoparticles in liquid

Graphene oxide windows for in situ environmental cell photoelectron spectroscopy

Photoelectron spectroscopy of wet and gaseous samples through graphene membranes

Drop-casted self-assembling graphene oxide membranes for scanning electron microscopy on wet and dense gaseous samples

Measurement of the elastic properties and intrinsic strength of monolayer graphene

Electron transparent graphene windows for environmental scanning electron microscopy in liquids and dense gases

Light-induced reduction of cuprous oxide in an environmental transmission electron microscope

Controlled environment specimen transfer

Electron microscopy of specimens in liquid

Liquid scanning transmission electron microscopy: Imaging protein complexes in their native environment in whole eukaryotic cells. Microscopy and Microanalysis

Video-frequency scanning transmission electron microscopy of moving gold nanoparticles in liquid

Correlating Scattering and Imaging Techniques

Switzerland *Corresponding author: e-mail address: dimitri.vanhecke@unifr.ch REFERENCES de

Einf€ uhrung in die Strahlenchemie mit praktischen Anleitungen

Electron-water interactions and implications for liquid cell electron microscopy

Effect of ordering of PtCu3 nanoparticle structure on the activity and stability for the oxygen reduction reaction

Fuel cell catalyst degradation on the nanoscale

Degradation mechanisms of Pt/C fuel cell catalysts under simulated start-stop conditions

Probing the degradation mechanisms in electrolyte solutions for L-ion batteries by in situ transmission electron microscopy

Controlled growth of nanoparticles from solution with in situ liquid transmission electron microscopy

Bubble and pattern formation in liquid induced by an electron beam

and (C) are before exposure to the electron beam and (B) and (D) are afterward. The TEM grids in (A) and (B) are covered with ionic liquid mixed with 550 μM fluorescein, and the ones in (C) and (D) are covered with ionic liquid mixed with 93-mM of Nile blue. There is a clear fluorescence change from the circular central region's exposure to a dose rate of 12 eÁÅ À2 Ás À1

Quantifying electrochemical nucleation and growth of nanoscale clusters using real-time kinetic data

Electron microscopy of specimens in liquid

Electron-water interactions and implications for liquid cell electron microscopy

Dynamic micoscopy of nanoscale cluster growth at the solid-liquid interface

Nanocrystal diffusion in a liquid thin film observed by in situ transmission electron microscopy

Plunge freezing for electron cryomicroscopy

Cryo-electron microscopy-A primer for the nonmicroscopist

Cryofixation of Tissues for Electron-Microscopy Scanning Microscopy

A review of the analysis of multiple nucleation with diffusion controlled growth

In situ SEM microchip setup for electrochemical experiments with water-based solutions

Quantifying electrochemical nucleation and growth of nanoscale clusters using real-time kinetic data

The morphology and nucleation kinetics of copper islands during electrodeposition

Visualization of active and passive control of morphology during electrodeposition

Although FM allows identification of biomolecules, its resolution is limited by diffraction to about 200 nm, or 20-50 nm for super resolution techniques. Electron microscopy (EM) is capable of improving the resolution to sub-nanometer levels, which is required to study biomolecules and their structural environment. Combining both techniques in an integrated correlative light-and electron microscopy (iCLEM) system (Zonnevylle, 2013), makes fast biomolecule identification within the cell ultrastructure possible at high resolution and sensitivity (Liv, 2013), as shown in Figures 39A and B. However, correlating the live-cell FM image with images of fixed cells in EM is rather complex, and in some cases it may even be impossible. Electron microscopy of specimens in liquid has made great progress in development and applicability over the last decade. Samples under nearnative cell conditions are protected from a vacuum in the EM chamber by means of a liquid cell holder. This special holder contains SiN membranes or glass substrates that are transparent for electrons or photons, respectively. We recently managed to develop a holder that enables simultaneous widefield fluorescence (WF) and liquid scanning EM (SEM) on the same region of interest in an iCLEM system (Figures 40A and B). This allows EM investigation of regions of interest of cells under near-native conditions REFERENCES Liv

Scanning electron microscopy of individual nanoparticle bio-markers in liquid

Integration of a high-NA light microscope in a scanning electron microscope

Advanced techniques for imaging parasitic Hymenoptera (Insecta)

Scintillation SE detector for variable pressure scanning electron microscopes

Comparative analysis of Trichuris muris surface using conventional, low-vacuum, environmental, and field emission scanning electron microscopy

Using environmental scanning electron microscopy (ESEM) as a quantitative method to analyse the attachment of Giardia duodenalis to epithelial cells

Redescription, systematic status, and molecular characterisation of Multicaecum heterotis Petter, Vassiliade`s et Marchand, 1979 (Nematoda: Heterocheilidae), an intestinal parasite of Heterotis niloticus (Osteichthyes: Arapaimidae) in Africa

Controlled dehydration of a biological sample using an alternative form of environmental SEM

The low-temperature method for the study of coniferous tissues in the environmental scanning electron microscope

In situ study of live specimens in an environmental scanning electron microscope

Formation of core-shell and hollow nanospheres through the nanoscale melt-solidification effect in the Sm-Fe(Ta)-N system

Pulsed-laser fabrication of gas-filled hollow CoPt nanospheres

Integrated fluorescence and transmission electron microscopy

Platelets amplify inflammation in arthritis via collagen-dependent microparticle production

Localization of fluorescently labeled structures in frozen-hydrated samples using integrated light electron microscopy

Platelets and the immune continuum

Platelets were double-labeled for vWF with a fluorophore (A) and protein A gold (B,C), and then plunge frozen in liquid ethane. Cryo-TEM images (B) superimposed on the fluorescent image (white box in panel A). Image C zooms in on the specific immunogold labeling for vWF of the area box in B. (D) Slice of a cryo-tomogram revealing substructural details of the limiting alpha granule membrane

Christine Hartmann organized the logistics of the conference and edited the abstract booklet. The INM provided financial support through Eduard Arzt. The following sponsors are greatly acknowledged

North Carolina (high-level sponsor) Photograph of participants CISCEM

These initial experiments ( Figure 36 ) clearly showed a strong change in fluorescence after exposure to low doses of the electron beam. In order to fully quantify these results, it will be necessary to further control the thickness of the ionic liquid.The microfabricated systems point to ways of making recordings of individual sample freezing with sub-millisecond time-resolution processes in liquids with subsequent electron microscopic imaging of the sample on the thin membranes.In conclusion, cryo-CLEM and cryo-tomography are powerful techniques that enable fast immobilization and capturing of membrane dynamics that occur during platelet activation, such as fusion pore formation and localized and timed release phenomena.