Gas-ETEM study of catalysis

Heterogeneous nanoparticle catalysts that catalyze reduction or oxidation reactions at the solid–gas interface are of paramount importance in a wide range of chemical and energy applications. Heterogeneous catalysis is a complex process that involves a number of critical steps, including diffusion, reactant adsorption, surface reaction, and product desorption. The interaction of the active catalyst with reactant gases varies with temperature, gas pressure, and surface structure of the catalyst. Transformation and restructuring of a catalytic particle often occur under reaction conditions.^{Reference Tao and Salmeron18,Reference Tao, Grass, Zhang, Butcher, Renzas, Liu, Chung, Mun, Salmeron and Somorjai19} Understanding and elucidating the mechanism of catalysis requires knowledge of the structural and chemical evolution of the nanoparticle catalysts during the reaction. Because such information is difficult to access through ex situ characterization in a vacuum, in situ characterization that enables the probing of dynamic catalytic processes under real reaction conditions is necessary.

In situ ETEM studies of gas–solid reactions under controlled reaction conditions can provide great opportunities for the characterization of heterogeneous catalysts. A local gaseous environment can be created by a number of approaches, including the use of a closed-gas TEM cell and gas injection into an open sample area with differential pumping in the TEM column.^{Reference Xin, Niu, Alsem and Zheng20–Reference Lee, Dewald, Eades, Robertson and Birnbaum22} There have been many studies on the use of ETEM to visualize nanoparticle catalysts with atomic resolution during gas reactions. Local structural and chemical information about a catalyst particle can also be obtained through a combination of imaging, diffraction, and spectroscopy.^{Reference Gai23,Reference Sharma24} Key insights have been acquired on catalytic active sites, defect structural evolution, the nature of bonding in redox reactions, and the correlation between microstructure and catalytic performance.^{Reference Yoshida, Kuwauchi, Jinschek, Sun, Tanaka, Kohyama, Shimada, Haruta and Takeda6,Reference Hansen, Wagner, Helveg, Rostrup-Nielsen, Clausen and Topsøe25–Reference Simonsen, Chorkendorff, Dahl, Skoglundh, Sehested and Helveg28}

Some examples of the ETEM study of nanoparticle catalysts are as follows: Formation of a subsurface oxide on a metal catalyst was identified during the catalytic oxidation of carbon monoxide due to the incorporation of oxygen into the metal at elevated temperatures.^{Reference Somorjai and Li29} Copper nanoparticle catalysts were found to exhibit remarkable restructuring in various gaseous environments, including reversible surface faceting, which was a result of preferential adsorption of reactant molecules on different crystalline facets.^{Reference Hansen, Wagner, Helveg, Rostrup-Nielsen, Clausen and Topsøe25} Another recent study of an Au/CeO₂ catalyst by aberration-corrected ETEM visualized the restructuring of {100} facets of gold nanoparticles during CO oxidation at room temperature.^{Reference Yoshida, Kuwauchi, Jinschek, Sun, Tanaka, Kohyama, Shimada, Haruta and Takeda6} The CO molecules adsorbed onto the on-top sites of gold atoms in the undulating hexagonal lattice, and the restructured {100} facets could sustain CO adsorption at higher surface coverages (Figure 1).

Figure 1. In situ environmental transmission electron microscope study of catalytic Au nanoparticles (NPs) supported on CeO₂. Au NP on CeO₂ (a) in ultrahigh vacuum (UHV) and (b) in a gas environment that contained 1 vol% CO in air at 45 Pa at room temperature. Higher magnification images of these regions are shown at the bottom of the corresponding panels. The difference between (a) and (b) shows that the interlayer distance of Au NPs in 1 Torr (∼133 Pa ∼1.3 mbar) CO increases in contrast to that in UHV. (c) High-magnification image of Au NPs with adsorbed CO. (d) Simulated image based on an energetically favorable model, which corresponds to the selected region in (c).^{Reference Yoshida, Kuwauchi, Jinschek, Sun, Tanaka, Kohyama, Shimada, Haruta and Takeda6} Adapted from Reference 6. © 2012 American Association for the Advancement of Science.

More recently, ETEM studies of nanoporous catalysts were performed to understand the dependence of catalytic activity on pore size and residual elements.^{Reference Fujita, Guan, McKenna, Lang, Hirata, Zhang, Tokunaga, Arai, Yamamoto and Tanaka26,Reference Xin, Pach, Diaz, Stach, Salmeron and Zheng30} In situ ETEM observations of dealloyed nanoporous gold provided compelling evidence that surface defects are active sites for the catalytic oxidation of CO, and residual silver stabilizes the atomic steps by suppressing {111} faceting kinetics.^{Reference Fujita, Guan, McKenna, Lang, Hirata, Zhang, Tokunaga, Arai, Yamamoto and Tanaka26} Another study on a nanoporous cobalt catalyst was performed in the presence of hydrogen at various temperatures using atomic imaging and EELS. The results revealed that during H₂ reduction, the valence state of CoO_x nanoporous particles changed from cobalt oxide to metallic cobalt.^{Reference Xin, Pach, Diaz, Stach, Salmeron and Zheng30} In their article in this issue, Crozier and Hansen provide additional examples on in situ and operando TEM of catalytic materials.

Dynamic processes in liquids

Wang et al. describe liquid-cell TEM, a new experimental platform that allows imaging through liquids with subnanometer resolution. The development of liquid-cell TEM is largely due to technical advances in nanofabrication and membrane technology. Recently, graphene liquid cells have opened new opportunities to study liquid reactions in situ in the transmission electron microscope with improved resolution (Figure 2).^{Reference Yuk, Park, Ercius, Kim, Hellebusch, Crommie, Lee, Zettl and Alivisatos7}

Figure 2. (a) Graphene liquid cell and in situ high-resolution transmission electron microscope imaging of Pt nanoparticle growth.^{Reference Yuk, Park, Ercius, Kim, Hellebusch, Crommie, Lee, Zettl and Alivisatos7} (b) The arrow indicates a small cluster attaching to the existing nanoparticle. A twinned Pt nanoparticle is achieved with a (111) mirror plane, as shown in the fast Fourier transform pattern at the bottom right. Scale bar in the image sequence is 2 nm. Note: Z.A., zone axis. Adapted from Reference 7. © 2012 American Association for the Advancement of Science.

Liquid-cell TEM has been applied to the study of nanoparticle growth mechanisms, and important insights have been achieved by direct observations of nanoparticle growth trajectories.^{Reference Liao, Cui, Whitelam and Zheng1,Reference Liao and Zheng9,Reference Park, Kodambaka, Ross, Grogan and Bau31–Reference Jungjohann, Bliznakov, Sutter, Stach and Sutter38} These studies illustrate that the growth of a nanowire involves attachment of nanoparticles grown from the solution, and recrystallization and rearrangement of nanoparticles are essential for nanoparticle coalescence.^{Reference Liao, Cui, Whitelam and Zheng1,Reference Zheng, Smith, Jun, Kisielowski, Dahmen and Alivisatos36} The direct observation of Pt nanocube facet development revealed that growth by following the conventional surface-energy-minimization law breaks down at the nanoscale.^{Reference Liao, Zherebetskyy, Xin, Czarnik, Ercius, Elmlund, Pan, Wang and Zheng2} The application of liquid-cell TEM has become increasingly important for studying a wide range of other materials transformations in materials science, as well as in chemistry and biology.

The dynamics of nanobubbles^{Reference White, Mecklenburg, Singer, Aloni and Regan39} and water nanodroplets^{Reference Mirsaidov, Zheng, Bhattacharya, Casana and Matsudaira40} were revealed using liquid-cell TEM. The properties of water at interfaces (see the December 2014 issue of MRS Bulletin on “Water at Functional Interfaces”) play a crucial role in functional biological membranes, flow of liquids through pores and over surfaces, hydration of biomolecules, and chemical reactions in aqueous solutions. The ability to directly study water in contact with the substrate interface could provide insights into the dynamic properties of water. Liquid-cell TEM has also made it possible to directly image biomaterials in their native environment at ambient temperature.^{Reference Mirsaidov, Zheng, Casana and Matsudaira12,Reference Mirsaidov, Zheng, Bhattacharya, Casana and Matsudaira40} Whole cells in liquids have also been captured at nanometer resolution with the assistance of Au nanoparticle labels.^{Reference De Jonge, Peckys, Kremers and Piston8,Reference Liu, Wu, Huang, Peng, Chang, Chang, Hsu and Yew41} These studies provide the basis for future nanometer-scale dynamic imaging of biological materials in liquids using TEM with proper management of the electron dosage.

The study of liquid-cell TEM is rapidly expanding. The development of electrochemical liquid cells allows monitoring of electrochemical processes in liquid electrolytes in real time.^{Reference Zeng, Liang, Liao, Xin, Chu and Zheng10} There have been many recent studies in this area, and this work is of great interest to electron microscopists as well as battery researchers.

Solid-state batteries

A solid electrolyte might be the ultimate solution for safe batteries. Brazier et al. reported the first cross-sectional ex situ TEM observations of an all solid-state lithium ion “nanobattery,”^{Reference Brazier, Dupont, Dantras-Laffont, Kuwata, Kawamura and Tarascon42} where they used a focused ion beam to make a nanobattery from a pulsed-laser-deposited solid-state battery. For probing electrodes and electrolyte materials and their interfaces, ex situ experiments often provide only limited insights because of the sample preparation and transfer needed, which prevents the determination of the time constants of electrochemically induced phase transformations. More importantly, electrochemical systems often operate at states far from equilibrium, where a system tends to relax to its equilibrium state with time.

To probe the kinetics in an electrochemical system, progress has been made with in situ techniques that are able to probe the structural, morphological, and chemical changes that take place during electrochemical processes at the nanoscale, particularly at the interfaces between an electrode and electrolyte. Yamamoto et al. reported on the in situ dynamic visualization of the electric potential in an all solid-state battery with electron holography and EELS.^{Reference Yamamoto, Iriyama, Asaka, Hirayama, Fujita, Fisher, Nonaka, Sugita and Ogumi43} Interestingly, Ruzmetov et al. found that a substantial reduction in the electrolyte thickness, into the nanometer regime, can lead to rapid self-discharge of the battery even when the electrolyte layer is conformal and pinhole-free.^{Reference Ruzmetov, Oleshko, Haney, Lezec, Karki, Baloch, Agrawal, Davydov, Krylyuk and Liu44} A procedure for fabricating electrochemically active, electron-transparent solid-state batteries using a focused ion beam were developed, and compositional and structural changes have been observed at the interfaces between the LiCoO₂ cathode and LiPON electrolyte as well as at the interfaces between the Si anode and Cu current collector, as shown in Figure 3.^{Reference Santhanagopalan, Qian, McGilvray, Wang, Wang, Camino, Graetz, Dudney and Meng45}In situ TEM with EELS is becoming one of the most important tools for studying solid–solid interfaces in a variety of solid-state devices, including batteries, solar cells, and solid-oxide fuel cells.

Figure 3. Electron energy-loss spectroscopy (EELS) mapping and scanning transmission electron microscopy (STEM) images of a solid-state Si/LiPON/LiCoO₂ rechargeable battery. Reprinted with permission from Reference 45. © 2013 American Chemical Society.

Ferroelectric and ferromagnetic switching and behavior

The need to continuously decrease the size of magnetic-bit elements in magnetic-storage media, such as magnetoresistive random access memory, is perhaps the strongest driving force underlying studies of magnetic nanostructures. A similar trend occurred for ferroelectric nanostructures. Intriguing structures and properties emerge as the size of the structure contracts. Switchable spontaneous polarization, domain engineering, and strain control of ferroelectrics were recently targeted for energy-efficient nonvolatile memories and ferroelectric field-effect transistors. Nevertheless, these efforts were hindered by a lack of experimental methods, especially ones to characterize domain-defect interactions and their dynamics, as well as to directly link local atomic displacement to polarization.^{Reference Yu, Onose, Kanazawa, Park, Han, Matsui, Nagaosa and Tokura5,Reference Zweck, Dehm, Howe and Zweck13,Reference Huang, Schofield and Zhu46–Reference Yu, Kanazawa, Zhang, Nagai, Hara, Kimoto, Matsui, Onose and Tokura49}

Lorentz microscopy has been widely used to study skyrmions (a hypothetical particle related originally to baryons; skyrmions as topological objects are important in solid-state physics), including magnetic switching and domain configuration evolution. The dynamic process of generation and annihilation of magnetic biskyrmions was observed by in situ TEM.^{Reference Yu, Tokunaga, Kaneko, Zhang, Kimoto, Matsui, Taguchi and Tokura48} As an alternative approach, a dedicated magnetizing stage can be used to generate a magnetic field at the specimen area. For example, a magnetizing stage can be built by adding Helmholtz coils on each side of the specimen,^{Reference Budruk, Phatak, Petford-Long and De Graef50,Reference Kohn, Kovács, Fan, McIntyre, Ward and Goff51} which can be used to study the field-induced motion of magnetic domain walls. Another possible design is to bring a piezo-driven sharp needle made of a permanent magnet close to the specimen.^{Reference Shindo, Park, Gao and Park52,Reference Park, Park, Gao, Shindo and Inoue53} Furthermore, to obtain the relationship between magnetic structure and temperature, a thermal element also needs to integrate with the holder.

A new design capable of applying gigahertz resonance electric current and pulsed excitations in situ was developed, as shown in Figure 4, to measure the nonadiabatic spin torque effect^{Reference Pollard, Huang, Buchanan, Arena and Zhu16} and to map strongly coupled coaxial vortex motion in the dipolar- and indirect exchange-coupled regimes.^{Reference Pulecio, Warnicke, Pollard, Arena and Zhu54} In this issue, Li et al. present details on in situ measurements of ferroelectric domain switching using in situ biasing, atomic imaging, and displacement mapping. Other methods for probing ferroelectric switching, including off-axis electron holography, are described in References 3 and 14.

Figure 4. Visualizing vortex dynamics in a nanomagnet. Top: The spin precession measurement scheme. Yellow indicates two electrodes allowing GHz electric current (j(t), where t is time) to sweep across the Landau domain structure with a clockwise chirality (in-plane magnetization) and a counter-clockwise vortex-core polarity (out-of-plane magnetization). Bottom: Lorentz micrograph of the permalloy square with the Landau state showing the vortex core orbit under resonance excitation in a transmission electron microscope. The vortex core size is approximately 20 nm.^{Reference Pollard, Huang, Buchanan, Arena and Zhu16}

Mechanical properties

Many fundamental assumptions of classical mechanics and continuum mechanics fail as the system size is reduced to the nanoscale. Thus, mechanical testing in the transmission electron microscope has emerged. The development of a dedicated mechanical TEM holder enables the in situ investigation of mechanical behaviors and properties at the nanoscale and even the atomic scale. Throughout recent decades, the mechanical behavior of sub-10-nm structures has been studied inside the transmission electron microscope. Several novel mechanical phenomena have been reported, such as irradiation-induced high pressure and phase transformation,^{Reference Sun, Krasheninnikov, Ahlgren, Nordlund and Banhart55} nanoextruder effects,^{Reference Sun, Banhart, Krasheninnikov, Rodriguez-Manzo, Terrones and Ajayan56} geometry interlocking effects,^{Reference Tang, Ren, Wang, Wei, Kawamoto, Liu, Bando, Mitome, Fukata and Golberg57} cold welding,^{Reference Lu, Huang, Wang, Sun and Lou58} partial dislocation-induced discrete plastic deformation,^{Reference Zheng, Cao, Weinberger, Huang, Du, Wang, Ma, Xia and Mao59} local kinks in two-dimensional nanomaterials,^{Reference Nikiforov, Tang, Wei, Dumitricaǎ and Golberg60} mechanical annealing,^{Reference Shan, Mishra, Asif, Warren and Minor61,Reference Huang, Li, Shan, Li, Sun and Ma62} stress saturation and deformation mechanism transition,^{Reference Yu, Shan, Li, Huang, Xiao, Sun and Ma63} and liquid-like pseudoelasticity.^{Reference Sun, He, Lo, Xu, Bi, Sun, Zhang, Mao and Li64} In this issue, Minor et al. highlight recent advances in in situ TEM probing mechanical properties at the nanoscale (see Figure 5).

Figure 5. Example of a typical compression-to-failure in situ compression test on an individual nanocrystalline hollow CdS sphere. (a–c) Extracted video frames of the in situ compression test, corresponding to time t = (a) 0 s, (b) 1.8 s, and (c) 3.6 s. The estimated contact diameter is shown in red in (b). (d–e) Dark-field transmission electron microscope images of the hollow nanocrystalline CdS ball resting on the Si substrate (d) before and (e) after the compression test. (f) Load and displacement data from the loading portion of the in situ test versus time. The experiment was run under displacement control. Reproduced with permission from Reference 15. © 2008 Nature Publishing Group.