Interfaces in Heterogeneous Catalysts: Advancing Mechanistic

Publication Date (Web): February 16, 2017. Copyright © 2017 American ... Her research interest is in the development and application of in situ and f...
101 downloads 13 Views 9MB Size
Download PDF
Article pubs.acs.org/accounts

Interfaces in Heterogeneous Catalysts: Advancing Mechanistic Understanding through Atomic-Scale Measurements Wenpei Gao,†,‡ Zachary D. Hood,†,§ and Miaofang Chi*,† †

Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United States Department of Chemical Engineering and Materials Science, University of California, Irvine, Irvine, California 92697, United States § School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, United States ‡

CONSPECTUS: Developing novel catalysts with high efficiency and selectivity is critical for enabling future clean energy conversion technologies. Interfaces in catalyst systems have long been considered the most critical factor in controlling catalytic reaction mechanisms. Interfaces include not only the catalyst surface but also interfaces within catalyst particles and those formed by constructing heterogeneous catalysts. The atomic and electronic structures of catalytic surfaces govern the kinetics of binding and release of reactant molecules from surface atoms. Interfaces within catalysts are introduced to enhance the intrinsic activity and stability of the catalyst by tuning the surface atomic and chemical structures. Examples include interfaces between the core and shell, twin or domain boundaries, or phase boundaries within single catalyst particles. In supported catalyst nanoparticles (NPs), the interface between the metallic NP and support serves as a critical tuning factor for enhancing catalytic activity. Surface electronic structure can be indirectly tuned and catalytically active sites can be increased through the use of supporting oxides. Tuning interfaces in catalyst systems has been identified as an important strategy in the design of novel catalysts. However, the governing principle of how interfaces contribute to catalyst behavior, especially in terms of interactions with intermediates and their stability during electrochemical operation, are largely unknown. This is mainly due to the evolving nature of such interfaces. Small changes in the structural and chemical configuration of these interfaces may result in altering the catalytic performance. These interfacial arrangements evolve continuously during synthesis, processing, use, and even static operation. A technique that can probe the local atomic and electronic interfacial structures with high precision while monitoring the dynamic interfacial behavior in situ is essential for elucidating the role of interfaces and providing deeper insight for fine-tuning and optimizing catalyst properties. Scanning transmission electron microscopy (STEM) has long been a primary characterization technique used for studying nanomaterials because of its exceptional imaging resolution and simultaneous chemical analysis. Over the past decade, advances in STEM, that is, the commercialization of both aberration correctors and monochromators, have significantly improved the spatial and energy resolution. Imaging atomic structures with subangstrom resolution and identifying chemical species with single-atom sensitivity are now routine for STEM. These advancements have greatly benefitted catalytic research. For example, the roles of lattice strain and surface elemental distribution and their effect on catalytic stability and reactivity have been well documented in bimetallic catalysts. In addition, three-dimensional atomic structures revealed by STEM tomography have been integrated in theoretical modeling for predictive catalyst NP design. Recent developments in stable electronic and mechanical devices have opened opportunities to monitor the evolution of catalysts in operando under synthesis and reaction conditions; high-speed direct electron detectors have achieved sub-millisecond time resolutions and allow for rapid structural and chemical changes to be captured. Investigations of catalysts using these latest microscopy techniques have provided new insights into atomic-level catalytic mechanisms. Further integration of new microscopy methods is expected to provide multidimensional descriptions of interfaces under relevant synthesis and reaction conditions. continued... Received: November 25, 2016 Published: February 16, 2017 © 2017 American Chemical Society

787

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795

Article

Accounts of Chemical Research

In this Account, we discuss recent insights on understanding catalyst activity, selectivity, and stability using advanced STEM techniques, with an emphasis on how critical interfaces dictate the performance of precious metal-based heterogeneous catalysts. The role of extended interfacial structures, including those between core and shell, between separate phases and twinned grains, between the catalyst surface and gas, and between metal and support are discussed. We also provide an outlook on how emerging electron microscopy techniques, such as vibrational spectroscopy and electron ptychography, will impact future catalysis research. between the support and the NPs.10 Many interesting phenomena, which otherwise do not exist in pure-phase catalysts, were reported on oxide supported NPs. For example, Haruta reported enhanced activity of CO oxidation on supported Au NPs by hightemperature calcination11,12 where the particle sizes are beyond the theoretically most active size of 2 nm;13 the activity enhancement was attributed to a strong metal−support effect and possible migration of atoms in the support to the Au NPs. In Cu NPs supported on ZnO, thin ZnO encapsulation layers were observed on Cu NPs, which were also catalytically active toward CO2 reduction.14 The encapsulation of oxide supported Pd NPs with a secondary oxide was also shown to enhance the selectivity for H2O2 synthesis.15 In gas reactions, the dynamic interaction between the surface of the catalyst and gas molecules, their derivatives, intermediates, and adsorbed products forms a solid−gas interface, which is directly related to the catalysis pathway. The solid−gas interaction is expected to modify the catalyst structures in two ways: shape change, that is, the modification of enclosing of surface facets,16 and surface reconstruction.17 These dynamic structural changes modulate the catalytic activity of NPs. Probing the dynamic evolution of surface structure and chemistry at the atomic scale under simulated reaction conditions is crucial to the identification of intermediate species and to the fundamental understanding of reaction mechanisms. To experimentally probe the catalyst interfaces, techniques that offer atomic-scale structural and chemical analysis are required. Studying the kinetics of catalysis in situ necessitates a characterization technique, which provides not only an adequate temporal resolution that can resolve intermediate reactions and products but also operation under reaction conditions. In the remainder of this Account, we briefly introduce the capabilities and limitations of current characterization techniques with a focus on recent progress in electron microscopy that could impact the study of heterogeneous catalysis. We then discuss recent experimental results of catalyst interfaces using advanced characterization tools coupled with theoretical calculations. Lastly, we provide a future perspective on new possibilities and opportunities for catalyst research using microscopy.

1. INTRODUCTION: INTERFACES IN PRECIOUS METAL-BASED HETEROGENEOUS CATALYSTS Chemical reactions take place on the surface and interfaces of heterogeneous catalyst systems.1 Depending on the phase of the reactant, the reactive interfaces can include those between solid and gas, those between solid and liquid, and the triple interfaces of solid−gas−liquid. At these interfaces, the catalyst provides active sites where the reactants are absorbed, activated, and converted to new chemical species that are eventually released from the catalyst surface. The ability of catalysts to promote these reactions is determined by the surface bonding energy, which can be notably modified through atomic structuring by tuning interfacial atomic arrangements or forming new interfaces. For example, postsynthesis acid leaching modifies the atomic arrangements and chemical distribution at catalyst surfaces. Introducing core−shell structures has proven to be a useful strategy to fine-tune surface strain and therefore the surface electronic structure. Structural twining and internal phase boundaries within single catalyst particles are reported to modify local surface atomic configurations and therefore can be used to tune catalytic activities. Understanding the role of these interfaces during catalytic reactions, in particular how the atomic and chemical configurations of these interfaces modify the reaction mechanisms and how they evolve during dynamic electrochemical processes, is important for designing novel catalysts with optimal performance. In this Account, we discuss the latest atomic-scale understanding of critical interfaces in precious metal-based heterogeneous catalysts, by emphasizing recent insights gained through the utilization of new scanning transmission electron microscopy (STEM) techniques. 1.1. Current Understanding of Interfaces in Heterogeneous Catalysts

Metallic catalysts for different molecular reactions can be engineered in the forms of core−shell and alloyed nanoparticles (NPs).2 It has been reported that adopting a transition metal−Pt core−shell structure can significantly enhance the specific and mass activity of Pt in comparison with commercial Pt/C catalyst.3−5 Such enhanced activity is attributed to the modifications of surface electronic structures, that is, a narrower d-band state and a higher d-band energy, resulting in a stronger interaction between the catalyst surface and gas molecules, which has been demonstrated both experimentally and theoretically.6,7 On the other hand, the performance of catalysts was improved by alloying different elements. Somorjai et al. demonstrated that alloying Pt with Sn results in NPs with a SnOx component, which improves the overall resistance of the catalyst against CO poisoning.8 This approach, relying on the formation of interfaces between noble metal and reducible metal oxide, increases the stability and durability of the catalyst. In oxide-supported NPs, the formation of metal−oxide interfaces can critically influence the resulting catalytic activities. The interaction between the oxide surface and the metal atoms can exhibit various metal−support effects, including the change in the morphology of NPs,9 the anchoring or encapsulation of NPs, the formation of interphases, and charge-transfer effects

1.2. Microscopy Techniques to Probe Interfacial Atomic Structures

The unprecedented spatial resolution makes TEM a major characterization technique for revealing atomic structure of catalysts and their interfaces. In TEM, two imaging modes can be operated: TEM imaging, which uses a parallel electron beam, and scanning TEM (STEM), which uses a convergent electron probe to raster across the specimen.18 High-resolution TEM images provide coherent phase contrast, while STEM can offer incoherent image contrast when a high angle annular dark field (HAADF) detector is used to integrate the scattered electrons. HAADFSTEM contrast is highly sensitive to the mean atomic number and thus can be used to differentiate elements within catalysts, which is especially valuable in determining elemental distributions in alloyed or core−shell catalysts or identifying the potential elemental interdiffusion between metal NPs and their supports. The advancements of aberration correctors have significantly 788

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795

Article

Accounts of Chemical Research

shell and the dimmer contrast from Pd illustrate a clear interface between the two metals. The images collected at higher magnifications in Figure 1C,D showcase the atomic registry of

improved the spatial resolution to subangstrom in both TEM and STEM imaging. In particular, STEM has been more broadly implemented in catalyst research because of its relatively simple interpretation of image contrast, the ease of elemental identification, and the capability of incorporating spatially resolved spectroscopic techniques, including energy-dispersive X-ray spectroscopy (EDS) and electron energy-loss spectroscopy (EELS). Due to the significantly enhanced beam intensity enabled by spherical aberration correctors,19 chemical analysis of single atom or single atomic columns is also feasible.11,20 Another important development in microscopy critical to catalyst research is imaging and chemical analysis under different in situ conditions, such as gas/liquid environments or thermal annealing. These conditions allow for real-time observation of structural and chemical evolutions of working catalysts. Although it is now feasible to achieve high-resolution STEM imaging in a gas environment or during thermal annealing, it is very challenging to achieve atomic resolution in STEM in a liquid environment. This difficulty is mainly due to significant scattering from the cell windows and the relatively thick liquid layer. The gas environment can be achieved by two different configurations: an environmental TEM, which uses differential pumping, or a MEMS-based in situ holder.21 Each configuration has its own unique advantages. The first configuration provides opportunities to integrate various in situ holders with gas environments, such as heating, cooling, or biasing, while the latter one can be operated at higher gas pressures (up to 4 atm), therefore mimicking high pressure conditions.22 Thermal instability has long been the major hurdle limiting high resolution imaging during annealing. MEMS-based chips have allowed researchers to overcome this limitation, where atomic-scale imaging upon in situ annealing is now feasible for heterogeneous catalysts. The recent implementation of these developments for in situ STEM have sparked new insights regarding the fundamental understanding of both synthesis and working mechanisms of catalysts.23 Here, we must point out that strong electron−matter interaction in STEM and TEM can introduce beam damage to the specimens, inducing undesirable chemical, structural, and bonding changes. According to the nature of the specimen, different damage mechanisms, that is, radiolysis, knock-on displacement, and heating or electrostatic charging, can happen. Rational experimental conditions, such as electron accelerating voltage, beam current, cooled stage, and pulsed electron exposures, need to be carefully selected to minimize beam damage. In addition, it is also important to consider the possible electron beam damage during data analysis and interpretations. More detailed discussion on possible radiation damage in STEM and TEM can be found in the literature.24,25 Below, we focus on recently reported investigations of catalyst interfaces using in situ and ex situ atomicresolution STEM experiments.

Figure 1. (A) HAADF-STEM image of Pd@Pt core−shell icosahedra. (B) Atomic resolution image of a single particle along a 2-fold symmetry axis, showing the sharp interface between three Pt atomic layers in the shell and the Pd core. (C,D) HAADF-STEM images on the two edges in boxes in panel B: green dots, Pd atoms; red dots, Pt atoms. Reproduced with permission from ref 3. Copyright 2015 Nature Publishing Group.

the Pd−Pt interface in the multiply twinned NP for the first time. The authors discovered that the Pt atoms in the shell do not stack perfectly on top of Pd. Instead, the expended surface plane on the icosahedron allows for extra Pt atoms to be packed into each Pt overlayer relative to the Pd layers, which creates compressive strain on the Pt shell.5 The twin boundaries further limit the relaxation of Pt atoms only along the surface normal direction, leading to a corrugated shell. This experimental result provides a model structure for density functional theory (DFT) calculations, which suggest that the new surface has weakened binding with hydroxyl radicals. The destabilization of hydroxyl radicals enhances the oxygen reduction reaction (ORR) activity, explaining the improved specific and platinum mass activities of the Pd@Pt icosahedra, relative to a commercial Pt/C catalyst. Tuning strain in binary alloyed NPs can also be achieved by electrochemical leaching (or so-called “dealloying”) that is based on the different dissolution rates of elements in particular solutions.26,27 Although slightly different procedures have been reported among different research groups, the same mechanism is shared: leaching away a trace amount elements that are more active toward acid leaching (such as transition metals) and leaving the surface layer(s) rich in elements less active toward acid leaching (often Pt). As the atomic structure often remains epitaxial between the core and shell, the different chemical distribution can lead to lattice strain and therefore influence catalytic activity of the NPs. Due to chemical sensitivities with atomic resolution, Z-contrast imaging is often used to evaluate the success of such methods.26,27 Figure 2 shows an example of Pt−Ni NPs before and after acid leaching, where the intensity line profiles clearly reveal the Pt-rich surface in the leached NP.

2. INTERFACIAL STRUCTURES IN CORRELATION WITH CATALYST FUNCTION 2.1. Atomic Structure of Interfaces in Core−Shell and Alloyed NPs

An exemplary study on the atomic interface in Pd@PtnL core−shell icosahedral NPs using STEM is shown in Figure 1.3 When projected along one of the 2-fold symmetry axes in an icosahedron, the STEM image reveals the deposition of Pt shells on the {111} facets of a Pd icosahedral seed. Due to the large difference in atomic number, the brighter contrast from the Pt 789

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795

Article

Accounts of Chemical Research After thermal annealing at 400 °C, the surface atomic structure is smoothed and the Pt-rich surface is about 2−3 atomic layers thick. This multilayered Pt-skin surface was found to exhibit an enhanced ORR catalytic activity, ∼6 times greater than Pt/C particles with similar size (∼5 nm). Defects at the interface are also critical in engineering the activity of NPs. Hsieh et al.28 demonstrated that desirable nanostructures can be achieved when the lattice defect level is minimized in Ru@Pt core−shell NPs. When Ru NPs were annealed before depositing a Pt shell, the Ru core became highly ordered with reduced lattice defects. A sharp boundary forms between the fcc Pt shell on the hcp Ru core as confirmed by the HAADF-EELS mapping using STEM. DFT calculations predicted that two stacking sequences of Ru(ABAB)−Pt(AC) and Ru(ABAB)−Pt(CA) are preferred in energy, which was observed in the HAADF images (Figure 3A,B). The morphology

Figure 3. (A,B) Z-contrast images with superimposed DFT-optimized structural models for close-packed Pt bilayer on a Ru(0001) surface. Scale bar, 0.5 nm. The shifts from the hcp to the fcc lattice sites at the top layers are indicated by yellow rectangles, blue circles, and blue arrows. (D) Intensity profile of the indicated region in panel C (black circles) with calculated curves using Ru@Pt core−shell (red line) and Ru (blue line) models. (E) Calculated STEM image based on a Pt bilayer model. Reproduced with permission from ref 28. Copyright 2013 Nature Publishing Group.

evolution and phase transformation of a single Pt3Co NP during thermal annealing from RT to 800 °C. Upon heating to 350 °C, the exterior monolayer shows brighter intensity than adjacent inner atomic layers, implying the segregation of Pt toward the surface (Figure 4B). Multislice simulation and EDS mapping were employed to confirm the formation of a Pt-rich layer. Further annealing at 550 °C leads to an ordered Pt3Co phase with disappearance of the surface Pt layer and evolution of the surface facets. These findings brought the first atomic-scale insights of the roles of elemental diffusion, surface faceting, phase transformation, and their interplay during the process of catalyst NP reformation by taking advantage of the high resolution and elemental probes provided by in situ STEM. 2.2. Heterogeneous Interfaces of NP Catalysts on Oxide Supports Figure 2. (A) Z-contrast images of a representative as-synthesized Pt−Ni particle (left) and the particle after acid treatment (right). (B) Background subtracted, normalized intensity line profiles extracted for the regions marked in panel A. (C) Atomic structure models of as-prepared and acid treated PtNi catalysts. Reproduced with permission from ref 26. Copyright 2011 American Chemical Society.

The metal−oxide interface is inherently complex. As such, a variety of advanced techniques in STEM and TEM have been applied to study the metal−oxide interface. The Wulff construction is the result of surface energy minimization on particles. By investigation of the modification of Wulff construction at the interface of Pt−SrTiO330 and Au−TiO2,9,31 the results suggest that tuning the surface of the support could manipulate the exposed surface area of the NPs, which would impact the catalyst activity, selectivity, and stability. The measured interfacial energy also has a range of variation, possibly from defects or reconstruction of the oxide surface, existence of oxygen vacancies, and presence of interfacial metal atoms. While much research assumes heterogeneous metal−oxide systems with perfect support surfaces,11,31,32 evidence suggests the activity of catalysts toward CO oxidation is increased by the formation of oxygen vacancies near the interface on reducible oxides.33 Oxygen at the interface is difficult to detect due to its small scattering cross section for electrons. The recent development of annular bright field (ABF) imaging has made imaging light elements in STEM feasible.34 Using ABF imaging,

of the NP indicates that the Pt bilayer shell shifts downward on the left following the curvature of the Ru core in an energetically favorable manner (Figure 3A). The finely tuned Ru@Pt core− shell structure shows enhanced CO tolerance and improved stability in accelerated stress tests, a progress toward better catalytic performance in proton exchange membrane fuel cells. Postsynthesis annealing is a widely used method to tune atomic structure and chemical distributions in bimetallic nanocatalysts. In transition metal−Pt NPs, ex situ macroscopic measurements revealed that the catalytic activity varies with annealing temperature. However, the exact atomic configurations responsible for the specific activity change were not clear.4 In recent work, Pt3Co NPs were studied by in situ Z-contrast imaging upon annealing.29 Figure 4 shows the structural 790

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795

Article

Accounts of Chemical Research

Figure 4. HAADF-STEM images of a Pt3Co nanoparticle at RT, 350 °C, and 550 °C, respectively. Dynamic elemental segregations and phase transformation was revealed. Reproduced with permission from ref 29. Copyright 2015 Nature Publishing Group. Figure 6. (A) Schematic showing the principle of depth sectioning in STEM. (B) Z-contrast images taken at a series of focuses at the Au−TiO2 interface. (C) Depth-sectioning images constructed using the intensity profiles from the Au, TiO, and interfacial layers, respectively (yellow box, Au-column; blue box, Ti-column; red box, interfacialcolumn). Reproduced with permission from ref 38. Copyright 2015 American Chemical Society.

the interface of Au−Fe2O3 was imaged with directly interpretable oxygen contrast.35 As shown in Figure 5, the Fe and O lattice fits

probe position at different depths. The focal-series images provide depth information as shown in Figure 6B.38 The intensity profile along each atomic column at the interface indicates a restructured interface with embedded Au atoms, distinctive from stoichiometric TiO2. Fitting the three-dimensional intensity profiles with multislice simulation estimates the numbers of embedded Au atoms. The interface was formed by annealing in air at 650 °C, a highly oxidative environment, similar to the activation conditions reported by Haruta,12 while the interface is highly reduced according to DFT calculations36 and the reduced interfacial spacing.38 Therefore, the detection of interfacial Au atoms suggests a strong metal−support effect at the Au−TiO2 interface during annealing processes in air. Supported NP catalysts were also investigated using in situ STEM and TEM. In several works, the dynamics of Au NPs supported on CeO2 are captured by both environmental TEM and HAADF-STEM. In environmental TEM, the Au NP was found to have a stepwise shift and rotation on oxygen terminated CeO2 surface upon heating in CO/O2.39 The stepwise shift by 0.09 nm was directly observed from the cross-sectional image of the Au−CeO2 interface, as shown in Figure 7. The rotation of a particle around an axis is difficult to observe because the lattice of a small Au NP overlaps with that from the CeO2 support. The authors used the fast Fourier transform of the images to differentiate the Au lattice from CeO2, determine the orientation of Au, and quantify the rotation. The shift and rotation of Au atoms at the interface demonstrated that under a reactive environment, the Au−CeO2 interface is not rigid. In STEM, continuous irradiation by the electron beam can lead to the flattening of NPs on oxide supports, and the particle could reform after blocking the electron beam.40 The evolution is closely related to the oxidation state of the CeO2 surface suggested by EELS, where the flattened Au layer occurs on reduced CeO2 surfaces and Au clusters form on oxidized surfaces.41 Hence, advanced STEM and TEM imaging techniques have brought fruitful insights

Figure 5. (A) HAADF- and (B) ABF-STEM images of Au−γFe2O3(111). (C) Enlarged area of the interface in panel B. (D) Schematic of the interface structure. Reproduced with permission from ref 35. Copyright 2015 Elsevier Inc.

the Fe2O3 atomic model projected along the [110] zone axis. These images reveal that the Fe2O3 support is terminated with Fe atoms at the interface, differing from those of Au−CeO2 and Au−NiO2.11 These results therefore prove that the nature of supporting oxides can significantly alter the interfacial atomic and electronic configurations in heterogeneous interfaces, leading to distinctive catalytic mechanisms. In this case, Fe-termination is believed to offer a better oxygen adsorption, compared to O-terminations in Au−CeO2 and Au−TiO2.33 At the Au−TiO2 interface, possible reconstruction mechanisms induced by TiO2 surface reduction suggest the existence of interfacial Au atoms.36 The detection of such interfacial atoms is usually challenging. Aberration-corrected STEM allows for larger probe-forming apertures with higher convergence angles, providing a reduced depth of focus of a few nanometers and depth sectioning along the direction perpendicular to the image plane.37 Figure 6A displays a schematic for changing the electron 791

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795

Article

Accounts of Chemical Research

and cons. For example, a maximum gas pressure of 2000 Pa can be achieved in an ETEM without any obstruction of the electron beam. An enclosed cell can contain a gas pressure of 4 atm. However, the interaction of the electron beam with the two enclosing windows often compromises the imaging and spectroscopy performance. As for any other STEM or TEM methods, a careful selection of in situ techniques prior to the experiments is crucial and must address well-defined scientific questions.

3. FUTURE PERSPECTIVES State-of-the-art STEM and TEM have provided significant insights in understanding the role of interfaces in heterogeneous catalysts. The exact atomic arrangement and bonding configurations have been revealed in many catalyst systems, which are valuable for theoretical calculations to establish reliable models toward the design of catalysts at the single atom level. In core− shell and alloyed NPs, subtle changes in lattice strain and chemical diffusion were found to greatly tune catalytic activity through the modification of d-band structure, therefore providing effective tuning factors toward enhanced catalytic activity.3,27−29 The metal−support interfaces in several heterogeneous catalysts were investigated by combining atomic resolution STEM imaging and EELS analysis.35,38−41 The interfacial atomic structure and oxidation states in such studies provide direct experimental proof of previously predicted strong metal−support interactions. Still, many fundamental questions relative to interfaces in precious metal-based catalysts remain unanswered, particularly regarding their dynamic processes at the atomic scale. For example, what are the local structures and dynamics of surface adsorbates on catalyst surfaces? How do interfacial atomic structures evolve under the different gas conditions? What are the intermediate gas product(s) during the reaction? None of these questions can be straightforwardly answered by existing characterization techniques. In many cases, STEM or TEM combined with various in situ liquid/gas or annealing capabilities provides the most promising route. However, STEM and TEM still have limitations, including the instability induced by gas introduction or the surface reaction itself, the limited detectability of gas molecules, the small sampling size at large magnification, the representativeness of in situ reactive conditions to operation conditions used in catalysis laboratories, and the potential electron beam modification on structural and chemical evolutions. Each of these limitations must be quantitatively understood and compared with the results from complementary characterization techniques. On the other hand, the temporal resolution for in situ ETEM studies is witnessing significant improvement by high-speed direct electron detectors, which can now achieve >1000 frames per second. Such a fast imaging speed, which may still be inadequate to record the detailed instantons of chemical reactions, is expected to make possible significant insights into the fundamental understanding of many catalyst synthesis and reaction mechanisms. It has to be pointed out that a high imaging speed can also mean a low signal-to-noise ratio (S/N) in images due to limited image-forming electrons and intrinsic noise in detectors. In this aspect, a trade-off between time resolution and S/N must be considered. We would like to emphasize that several emerging electron microscopy techniques may bring significant impact on future catalyst research, such as vibrational spectroscopy (VS) and ptychography. VS, taking advantage of high energy resolution provided by recently developed monochromators in STEM, is expected to reveal site-specific absorptions of gas molecules on catalyst surface. Using a monochromated STEM with an energy

Figure 7. (A) Au NP supported on CeO2 at 100 Pa of 1 vol % CO/air at room temperature. The interfacial area is enlarged in panel B, showing the lateral, rigid-body displacement. The micrographs, corresponding simulated images, and models are shown in the top, middle, and bottom rows, respectively. Gold, gray, and red circles represent gold, cerium, and oxygen atoms, respectively. Blue and red arrows represent gold and cerium atomic planes, respectively. Reproduced with permission from ref 39. Copyright 2015 American Chemical Society.

on the structure and dynamics of oxide-supported metal NP catalysts. 2.3. Solid−Gas Interfaces

Many catalytic processes occur at solid−gas interfaces. How a surface responds upon gas adsorption under reaction conditions is of prominent interest. Detecting and locating the absorbed gas molecules is central to study the catalytic reaction mechanisms. In an environmental TEM (ETEM), Yoshida et al. demonstrated the first image of adsorbed CO molecule on the surface of a Au NP under CO/air (Figure 8B).17 The authors showed that CO absorption induces a reconstruction of a {100} surface facet to a {100}-hex facet on Au NPs. By combining these observations with ab initio calculations, they confirmed that the CO molecules only bind with reconstructed hexagonal Au top layers on the (100) surface. Such selective absorption implies dissimilar reaction rates on different surface facets. In their experiment, 1 vol % CO in air (0.34 Torr) at RT was used for the in situ CO environmental experiments. It will be crucial to study the influence of gas pressure on the absorption behavior since practical applications use different or higher gas pressures. It needs to be pointed out that two different in situ environmental STEM and TEM technologies have recently emerged based on the different methods utilized to confine gas environment to the specimen. One is the ETEM, which utilizes additional small pumping apertures, and the other one is a holder-based closed cells that encloses both the gas and specimen between two electron transparent windows.23 These two methods have their own pros 792

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795

Article

Accounts of Chemical Research

Figure 8. Aberration-corrected environmental TEM images in (A) vacuum and (B) 100 Pa 1 vol % CO/air at room temperature taken at 80 keV. Inset, a simulated image based on an energetically favorable model (blue dots, C atoms; red dots, O atoms; dark green dots, Au atoms with absorbed CO within the surface hexagonal lattice; bright green dots, other Au atoms within the surface hexagonal lattice; gold dots, Au atoms within and below the second topmost surface layer). Reproduced with permission from ref 17. Copyright 2012 The American Association for the Advancement of Science. (C) A HAADF image of MgO cubes. (D) Transmission (red) and aloof (green) valence-loss spectra showing a band gap at 7.5 eV and an interband intensity at 5.6 eV. (E) Vibrational spectrum in aloof mode showing an OH− stretch at 430 meV (the vibrational peak after background subtraction is in the inset). Reproduced with permission from ref 42. Copyright 2016 Elsevier B.V.

resolution of ∼9 meV, probing OH− species on the surfaces of MgO nanocubes was demonstrated (Figure 8C−E).42 The major advantage of using VS-STEM to probe chemical species is its relatively high spatial resolution (less than 10 nm) compared to its peer techniques, which may permit the investigation of sitespecific reaction mechanisms on the catalyst surface. With simultaneous atomic-scale imaging and in situ environments with VS, direct correlations between the surface atomic configuration and the activity of specific catalytic reactions can be established. Ptychography, which collects nearly all scattered electrons in STEM, provides opportunities to reconstruct the three-dimensional atomic structure of NPs and their interfaces using one single data set.43,44 Furthermore, all STEM imaging modes (ABF, BF, HAADF, etc.) can theoretically be obtained through postacquisition data analysis of ptychography. This method, therefore, could potentially reveal multidimensional structural, chemical, and bonding information. These new microscopy techniques, with continued studies, are expected to open a new era for characterizing critical interfaces in heterogeneous catalyst systems.

■

Biographies Dr. Wenpei Gao is currently a postdoctoral research fellow at the University of California, Irvine. He received his B.S. degree in Physics in 2010 from Peking University and Ph.D. in Materials Science and Engineering in 2015 at the University of Illinois at Urbana−Champaign, during which he received the Eric Samuel Scholarship and the RachelIntel Fellowship. He was a visiting researcher at ORNL in 2015. His current research interest is on in situ environmental electron microscopy of advanced catalysts. Zachary D. Hood is a Ph.D. candidate at Georgia Tech under the guidance of Prof. Younan Xia and Dr. Miaofang Chi. He graduated from Wake Forest University in 2013 with a B.S. in chemistry and a B.A. in German. He received the NSF graduate research fellowship and ORNLGT fellowship in 2015. His research interest includes the development of materials for biomedical and energy-related applications. Dr. Miaofang Chi is a research scientist at ORNL. She received her Ph.D. in Materials Science and Engineering from the University of California, Davis, in 2008, and was a research fellow at Lawrence Livermore National Laboratory (2006−2008). She was awarded the Burton Medal from the Microscopy Society of America (2016), the ORNL Director’s Award for Outstanding Accomplishment in Science and Technology, the ORNL’s Early Career Research Award (2015), the Distinguished Scholar Award by the Microanalysis Society (2007), and the Lawrence Graduate Research Fellowship (2006). Her research interest is in the development and application of in situ and functional STEM imaging for battery materials and catalysts.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

■

Miaofang Chi: 0000-0003-0764-1567

ACKNOWLEDGMENTS This work was supported by the Center for Nanophase Materials Sciences, which is a DOE Office of Science User Facility. W.P.G.

Notes

The authors declare no competing financial interest. 793

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795

Article

Accounts of Chemical Research

(20) Yang, J. C.; Small, M. W.; Grieshaber, R. V.; Nuzzo, R. G. Recent developments and applications of electron microscopy to heterogeneous catalysis. Chem. Soc. Rev. 2012, 41, 8179−8194. (21) Crozier, P. A.; Hansen, T. W. In situ and operando transmission electron microscopy of catalytic materials. MRS Bull. 2015, 40, 38−45. (22) Yokosawa, T.; Alan, T.; Pandraud, G.; Dam, B.; Zandbergen, H. In-situ TEM on (de) hydrogenation of Pd at 0.5−4.5 bar hydrogen pressure and 20−400 C. Ultramicroscopy 2012, 112, 47−52. (23) Tao, F.; Crozier, P. A. Atomic-Scale Observations of Catalyst Structures under Reaction Conditions and during Catalysis. Chem. Rev. 2016, 116, 3487−3539. (24) Egerton, R. F. Mechanisms of radiation damage in beam-sensitive specimens, for TEM accelerating voltages between 10 and 300 kV. Microsc. Res. Tech. 2012, 75, 1550−1556. (25) Egerton, R. F.; Li, P.; Malac, M. Radiation damage in the TEM and SEM. Micron 2004, 35, 399−409. (26) Wang, C.; Chi, M.; Li, D.; Strmcnik, D.; van der Vliet, D.; Wang, G.; Komanicky, V.; Chang, K.-C.; Paulikas, A. P.; Tripkovic, D.; Pearson, J.; More, K. L.; Markovic, N. M.; Stamenkovic, V. R. Design and Synthesis of Bimetallic Electrocatalyst with Multilayered Pt-Skin Surfaces. J. Am. Chem. Soc. 2011, 133, 14396−14403. (27) Strasser, P.; Koh, S.; Anniyev, T.; Greeley, J.; More, K.; Yu, C.; Liu, Z.; Kaya, S.; Nordlund, D.; Ogasawara, H.; Toney, M. F.; Nilsson, A. Lattice-strain control of the activity in dealloyed core−shell fuel cell catalysts. Nat. Chem. 2010, 2, 454−460. (28) Hsieh, Y.-C.; Zhang, Y.; Su, D.; Volkov, V.; Si, R.; Wu, L.; Zhu, Y.; An, W.; Liu, P.; He, P.; Ye, S.; Adzic, R. R.; Wang, J. X. Ordered bilayer ruthenium−platinum core-shell nanoparticles as carbon monoxidetolerant fuel cell catalysts. Nat. Commun. 2013, 4, 2466. (29) Chi, M.; Wang, C.; Lei, Y.; Wang, G.; Li, D.; More, K. L.; Lupini, A.; Allard, L. F.; Markovic, N. M.; Stamenkovic, V. R. Surface faceting and elemental diffusion behaviour at atomic scale for alloy nanoparticles during in situ annealing. Nat. Commun. 2015, 6, 8925. (30) Enterkin, J. A.; Poeppelmeier, K. R.; Marks, L. D. Oriented Catalytic Platinum Nanoparticles on High Surface Area Strontium Titanate Nanocuboids. Nano Lett. 2011, 11, 993−997. (31) Sivaramakrishnan, S.; Wen, J.; Scarpelli, M. E.; Pierce, B. J.; Zuo, J.-M. Equilibrium shapes and triple line energy of epitaxial gold nanocrystals supported onTiO2(110). Phys. Rev. B: Condens. Matter Mater. Phys. 2010, 82, 195421. (32) Shibata, N.; Goto, A.; Matsunaga, K.; Mizoguchi, T.; Findlay, S. D.; Yamamoto, T.; Ikuhara, Y. Interface Structures of Gold Nanoparticles on TiO2(110). Phys. Rev. Lett. 2009, 102, 136105. (33) Widmann, D.; Behm, R. J. Active Oxygen on a Au/TiO2 Catalyst Formation, Stability, and CO Oxidation Activity. Angew. Chem., Int. Ed. 2011, 50, 10241−10245. (34) Findlay, S. D.; Shibata, N.; Sawada, H.; Okunishi, E.; Kondo, Y.; Ikuhara, Y. Dynamics of annular bright field imaging in scanning transmission electron microscopy. Ultramicroscopy 2010, 110, 903−923. (35) Akita, T.; Maeda, Y.; Kohyama, M. Low-temperature CO oxidation properties and TEM/STEM observation of Au/γ-Fe2O3 catalysts. J. Catal. 2015, 324, 127−132. (36) Yu, M.; Trinkle, D. R. Au/TiO2(110) Interfacial Reconstruction Stability from ab Initio. J. Phys. Chem. C 2011, 115, 17799−17805. (37) Borisevich, A. Y.; Lupini, A. R.; Pennycook, S. J. Depth sectioning with the aberration-corrected scanning transmission electron microscope. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 3044−3048. (38) Gao, W.; Sivaramakrishnan, S.; Wen, J.; Zuo, J.-M. Direct Observation of Interfacial Au Atoms on TiO2 in Three Dimensions. Nano Lett. 2015, 15, 2548−2554. (39) Kuwauchi, Y.; Takeda, S.; Yoshida, H.; Sun, K.; Haruta, M.; Kohno, H. Stepwise Displacement of Catalytically Active Gold Nanoparticles on Cerium Oxide. Nano Lett. 2013, 13, 3073−3077. (40) Akita, T.; Tanaka, S.; Tanaka, K.; Haruta, M.; Kohyama, M. Sequential HAADF-STEM observation of structural changes in Au nanoparticles supported on CeO2. J. Mater. Sci. 2011, 46, 4384−4391. (41) Akita, T.; Okumura, M.; Tanaka, K.; Kohyama, M.; Haruta, M. Analytical TEM observation of Au nano-particles on cerium oxide. Catal. Today 2006, 117, 62−68.

was supported by the National Science Foundation under the Grant Number DMR-1506535. Z.D.H. gratefully acknowledges a Graduate Research Fellowship from the National Science Foundation (Grant DGE-1148903) and the Georgia Tech− ORNL Fellowship.

■

REFERENCES

(1) Schlögl, R. Heterogeneous Catalysis. Angew. Chem., Int. Ed. 2015, 54, 3465−3520. (2) Wu, J.; Yang, H. Platinum-Based Oxygen Reduction Electrocatalysts. Acc. Chem. Res. 2013, 46, 1848−1857. (3) Wang, X.; Choi, S.-I.; Roling, L. T.; Luo, M.; Ma, C.; Zhang, L.; Chi, M.; Liu, J.; Xie, Z.; Herron, J. A.; Mavrikakis, M.; Xia, Y. Palladium− platinum core-shell icosahedra with substantially enhanced activity and durability towards oxygen reduction. Nat. Commun. 2015, 6, 7594. (4) Wang, D.; Xin, H. L.; Hovden, R.; Wang, H.; Yu, Y.; Muller, D. A.; DiSalvo, F. J.; Abruña, H. D. Structurally ordered intermetallic platinum−cobalt core−shell nanoparticles with enhanced activity and stability as oxygen reduction electrocatalysts. Nat. Mater. 2013, 12, 81− 87. (5) Wu, J.; Qi, L.; You, H.; Gross, A.; Li, J.; Yang, H. Icosahedral Platinum Alloy Nanocrystals with Enhanced Electrocatalytic Activities. J. Am. Chem. Soc. 2012, 134, 11880−11883. (6) Mavrikakis, M.; Hammer, B.; Nørskov, J. K. Effect of Strain on the Reactivity of Metal Surfaces. Phys. Rev. Lett. 1998, 81, 2819−2822. (7) Kim, D.; Resasco, J.; Yu, Y.; Asiri, A. M.; Yang, P. Synergistic geometric and electronic effects for electrochemical reduction of carbon dioxide using gold−copper bimetallic nanoparticles. Nat. Commun. 2014, 5, 4948. (8) Michalak, W. D.; Krier, J. M.; Alayoglu, S.; Shin, J.-Y.; An, K.; Komvopoulos, K.; Liu, Z.; Somorjai, G. A. CO oxidation on PtSn nanoparticle catalysts occurs at the interface of Pt and Sn oxide domains formed under reaction conditions. J. Catal. 2014, 312, 17−25. (9) Gao, W.; Choi, A. S.; Zuo, J.-M. Interaction of nanometer-sized gold nanocrystals with rutile (110) surface steps revealed at atomic resolution. Surf. Sci. 2014, 625, 16−22. (10) Ta, N.; Liu, J.; Chenna, S.; Crozier, P. A.; Li, Y.; Chen, A.; Shen, W. Stabilized Gold Nanoparticles on Ceria Nanorods by Strong Interfacial Anchoring. J. Am. Chem. Soc. 2012, 134, 20585−20588. (11) Akita, T.; Kohyama, M.; Haruta, M. Electron Microscopy Study of Gold Nanoparticles Deposited on Transition Metal Oxides. Acc. Chem. Res. 2013, 46, 1773−1782. (12) Haruta, M. Catalysis of Gold Nanoparticles Deposited on Metal Oxides. CATTECH 2002, 6, 102−115. (13) Chen, M. S.; Goodman, D. W. The Structure of Catalytically Active Gold on Titania. Science 2004, 306, 252−255. (14) Behrens, M.; Studt, F.; Kasatkin, I.; Kühl, S.; Hävecker, M.; AbildPedersen, F.; Zander, S.; Girgsdies, F.; Kurr, P.; Kniep, B.-L.; Tovar, M.; Fischer, R. W.; Nørskov, J. K.; Schlögl, R. The Active Site of Methanol Synthesis over Cu/ZnO/Al2O3 Industrial Catalysts. Science 2012, 336, 893−897. (15) Freakley, S. J.; He, Q.; Harrhy, J. H.; Lu, L.; Crole, D. A.; Morgan, D. J.; Ntainjua, E. N.; Edwards, J. K.; Carley, A. F.; Borisevich, A. Y.; Kiely, C. J.; Hutchings, G. J. Palladium-tin catalysts for the direct synthesis of H2O2 with high selectivity. Science 2016, 351, 965−968. (16) Popa, C.; Zhu, T.; Tranca, I.; Kaghazchi, P.; Jacob, T.; Hensen, E. J. M. Structure of palladium nanoparticles under oxidative conditions. Phys. Chem. Chem. Phys. 2015, 17, 2268−2273. (17) Yoshida, H.; Kuwauchi, Y.; Jinschek, J. R.; Sun, K.; Tanaka, S.; Kohyama, M.; Shimada, S.; Haruta, M.; Takeda, S. Visualizing Gas Molecules Interacting with Supported Nanoparticulate Catalysts at Reaction Conditions. Science 2012, 335, 317−319. (18) Carter, C. B.; Williams, D. B. Transmission Electron Microscopy; Springer-Verlag: New York, 2009. (19) Pennycook, S. J. A scan through the history of STEM. In Scanning Transmission Electron Microscopy; Springer-Verlag: New York, 2011; pp 1−90. 794

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795

Article

Accounts of Chemical Research (42) Crozier, P. A.; Aoki, T.; Liu, Q. Detection of water and its derivatives on individual nanoparticles using vibrational electron energyloss spectroscopy. Ultramicroscopy 2016, 169, 30−36. (43) Yang, H.; Rutte, R. N.; Jones, L.; Simson, M.; Sagawa, R.; Ryll, H.; Huth, M.; Pennycook, T. J.; Green, M. L. H.; Soltau, H.; Kondo, Y.; Davis, B. G.; Nellist, P. D. Simultaneous atomic-resolution electron ptychography and Z-contrast imaging of light and heavy elements in complex nanostructures. Nat. Commun. 2016, 7, 12532. (44) Lupini, A. R.; Chi, M.; Jesse, S. Rapid aberration measurement with pixelated detectors. J. Microsc. 2016, 263, 43−50.

795

DOI: 10.1021/acs.accounts.6b00596 Acc. Chem. Res. 2017, 50, 787−795