The interface between metal catalyst and support plays a critical role in heterogeneous catalysis. An epitaxial interface is generally considered to be rigid, and tuning its intrinsic microstructure with atomic precision during catalytic reactions is challenging. Using aberration-corrected environmental transmission electron microscopy, we studied the interface between gold (Au) and a titanium dioxide (TiO2) support. Direct atomic-scale observations showed an unexpected dependence of the atomic structure of the Au-TiO2 interface with the epitaxial rotation of gold nanoparticles on a TiO2 surface during carbon monoxide (CO) oxidation. Taking advantage of the reversible and controllable rotation, we achieved in situ manipulation of the active Au-TiO2 interface by changing gas and temperature. This result suggests that real-time design of the catalytic interface in operating conditions may be possible.

Imaging a reaction taking place at the molecular level could provide direct information for understanding the catalytic reaction mechanism. We used in situ environmental transmission electron microscopy and a nanocrystalline anatase titanium dioxide (001) surface with (1 × 4) reconstruction as a catalyst, which provided highly ordered four-coordinated titanium “active rows” to realize real-time monitoring of water molecules dissociating and reacting on the catalyst surface. The twin-protrusion configuration of adsorbed water was observed. During the water–gas shift reaction, dynamic changes in these structures were visualized on these active rows at the molecular level.

Pd/CeO 2 has attracted great attention owing to its unique activity for methane catalytic oxidation; however, the active sites for CH 4 catalytic oxidation still remain elusive, which affects the comprehensive understanding of the catalytic mechanism. In this work, the structures of PdO x nanoparticles (NPs) loaded on octahedrons, cubes, and rods of nanocrystal CeO 2 supports were systematically studied by Cs-corrected HRTEM/STEM, XPS, and Raman spectroscopy. Our results indicate that the Pd species on CeO 2 supports are morphology-dependent: PdO NPs (Pd 2+ ) on octahedrons, PdO x (x = 1−2) clusters (1−2 nm) on cubes, and dispersed Pd 4+ ions on the CeO 2 rods. Additionally, the chemical states of Pd can be tuned in oxidizing/reducing atmospheres via interactions between Pd and CeO 2 . Detailed studies reveal that the Pd 2+ species are the active centers for the catalytic oxidation of methane. The activity of Pd 0 could be ascribed to Pd 2+ produced through the gradual oxidation of Pd 0 during the CH 4 oxidation. Further, Pd 4+ in the CeO 2 lattice is inactive for CH 4 oxidation. In situ Fourier transform infrared spectroscopy results suggest that the mechanism of CH 4 oxidation reaction on PdO x /CeO 2 follows the Mars−van Krevelen mechanism, and adsorbed CO can be produced in CH 4 decomposition over Pd 2+ in the absence of gas-phase oxygen. As revealed by density functional theory calculations, the incomplete coordination of Pd 2+ ions and adjacent oxygen atoms has excellent activity in cracking the C−H bond of CH 4 , which leads to high methane oxidation ability.

Preventing sintering of supported nanocatalysts is an important issue in nanocatalysis. A feasible way is to choose a suitable support. However, whether the metal–support interactions promote or prevent the sintering has not been fully identified. Now, completely different sintering behaviors of Au nanoparticles on distinct anatase TiO2 surfaces have been determined by in situ TEM. The full in situ sintering processes of Au nanoparticles were visualized on TiO2 (101) surface, which coupled the Ostwald ripening and particle migration coalescence. In contrast, no sintering of Au on TiO2 anatase (001) surface was observed under the same conditions. This facet‐dependent sintering mechanism is fully explained by the density function theory calculations. This work not only offers direct evidence of the important role of supports in the sintering process, but also provides insightful information for the design of sintering‐resistant nanocatalysts.

One dimensional copper hydroxide nanostrands, two dimensional Cu(2)(OH)(3)NO(3) nanoribbons and three dimensional CuO nanowalnuts were synthesized from the same diluted copper nitrate solution with ethanolamine at room temperature and 10 °C, respectively. The Cu(2)(OH)(3)NO(3) nanoribbons were formed by slowly hydrolyzing ethanolamine at low temperature. The CuO nanowalnuts were formed through dehydration of copper hydroxide nanostrands in aqueous solution at room temperature. Although their average size is about 500 nm, the specific surface area of the CuO nanowalnuts can be as large as 61.24 m(2) g(-1), due to their particular morphology with assembling of 8 nm grains. The Cu(2)(OH)(3)NO(3) nanoribbons were converted to CuO porous nanoribbons, keeping the shape. The catalytic performance of the CuO nanowalnuts for CO oxidation is 160 mL h(-1) g(cat)(-1) which is 23 times higher than those of the CuO porous nanoribbons and 40 nm commercial CuO nanoparticles, respectively. The electrochemical properties of the CuO nanowalnuts were also examined in a lithium-ion battery. After 30 cycles, the capacity of the as-prepared CuO nanowalnuts could sustain 67.1% (407 mA h g(-1)) of the second cycle (607 mA h g(-1)) at a rate of 0.1 C.

Carbon nanotubes were obtained by pyrolysis of acetylene or ethylene catalyzed by iron or iron oxide nanoparticles. The morphology, microstructure, and lithium insertion properties of these carbon nanotubes were investigated by transmission electron microscopy, high-resolution transmission electron microscopy, X-ray diffraction, Raman spectroscopy, and electrochemical measurements, respectively. The results showed that the structures of the carbon nanotubes play major roles in both specific capacity and cycle life. Slightly graphitized carbon nanotubes showed a specific capacity of 640 mAh/g during the first charge, whereas well-graphitized carbon nanotubes showed a specific capacity of 282 mAh/g during the first charge. After 20 charge/discharge cycles the charge capacity of the slightly graphitized samples degraded to 65.3% of their original charge capacities, but the wellgraphitized samples maintained 91.5% of their original charge capacities. The effects of charge-discharge rates and cycling temperature on lithium insertion properties of carbon nanotubes with different extents of graphitization are discussed.

The strong metal-support interaction (SMSI) is widely used in supported metal catalysts and extensive studies have been performed to understand it. Although considerable progress has been achieved, the surface structure of the support, as an important influencing factor,isusually ignored. We report af acet-dependent SMSI of Pd-TiO 2 in oxygen by using in situ atmospheric pressure TEM. Pd NPs supported on TiO 2 ( 101) and (100) surfaces showed encapsulation. In contrast, no such cover layer was observed in Pd-TiO 2 (001) catalyst under the same conditions.This facet-dependent SMSI, which originates from the variable surface structure of the support, was demonstrated in ap robe reaction of methane combustion catalyzedb yP d-TiO 2 .O ur discovery of the oxidative facet-dependent SMSI gives direct evidence of the important role of the support surface structure in SMSI and provides anew way to tune the interaction between metal NPs and the support as well as catalytic activity.

Metal catalysts are of great importance in the modern chemical industry. It is well-known that the structures of metal catalysts determine their properties. However, recent studies suggested that the structures of metal catalysts change dynamically under reaction conditions, resulting in the deactivation or activation of metal catalysts. This Review summarizes the latest research progresses in the structural reconstruction of metal catalysts via controlled-atmosphere transmission electron microscopy. The stateof-the-art research technologies and crucial factors affecting the nanosized metal catalyst reconstruction are discussed. Various types of reconstruction phenomena are reviewed, including sintering and dispersion, reshaping, composition evolution, surface reconstruction of metal oxides, and strong metal−support interactions. Moreover, recent studies of the structure−property relationship of metal catalysts are also reviewed. Finally, we highlight current challenges and provide the perspectives for future research of this topic. We hope this Review provides insights for the rational design of highperformance metal catalysts.