Department of Physical Chemistry
Nanostructured Model Catalysts
Department of Physical Chemistry
Nanostructured Model Catalysts
Methanol steam reforming, that is, the generation of hydrogen and carbon dioxide from methanol and water is to date one of the most promising heterogeneously catalyzed chemical reactions for on-board hydrogen storage and use. Its particular advantage is mainly based on the high hydrogen-to-carbon ratio (3:1), if the reaction is led highly CO2-selective. To achieve this goal, at the moment a range of also technologically used and already well-characterized catalyst systems are exploited, including copper-zinc oxide catalysts or intermetallic compounds on palladium basis. However, the used catalysts suffer from the severe drawbacks of either being prone to serious sintering upon catalyst activation or even during reaction, or, in the case of palladium-based intermetallic compounds, being structurally highly dynamic systems. This typically hampers the determination of the catalytically active centers leading to high CO2 selectivity. The copper-zirconium oxide catalysts discussed in the project “The Quest to Reliable Structure-Property Relationships in Methanol Steam Reforming” exhibit not only a high CO2 selectivity with at the same time multiple activities compared to used materials, but as a particular advantage also offer also an increased structural and chemical stability. This is mainly due to the inherent chemical inactivity of zirconium oxide, which simplifies the determination of the associated catalytically active and selective sites. To reach the set goal, new grounds of catalyst preparation and synthesis need to be broken as well as dedicated structural and chemical characterization techniques capable of resolving structure, morphology and chemistry during reaction applied to optimize the chemistry of the two components copper and zirconium oxide. Only by this knowledge-based approach, a new catalyst material with enhanced activity and selectivity with at the same time structurally stable catalytic center finally results, enabling setting-up reliable structure-property relationships for methanol steam reforming.
Goal of this approved program proposal is the development, set-up, application and successful commissioning of experimental method/tool for in-situ X-ray powder diffraction at elevated temperatures (up to 1100°C) and controllable gas atmospheres. Generally, such a tool offers the possibility for studies in a wide range of research fields where exact and time-resolved knowledge of either crystalline-structural transformations or the confirmation of phase stability at certain temperatures and gas environments is of scientific interest. Potential studies can address questions in materials processing (heat treatment, calcination, even ammonolysis, … ), materials performance (corrosion (protection) in harsh environments, stress, catalysis, ) and operando studies of e.g. catalytic active materials. Our focus is (i) technical-wise the development of this new experimental capability (ii) scientifically-wise the investigation of catalytic functional materials (binary and complex (bi-)metallic-oxide systems) in terms of clarification of temperature/gas-environment-dependent stability ranges as well as in operando studies under realistic operating conditions. The tool-development takes place in close collaboration with the beamline staff and stands in continuation with our past successful cooperation that resulted in the setup of a high temperature heating unit. It is our aim to make this experimental method available to the general user population as soon as it is commissioned.
Viele moderne Technologien beruhen auf den spezifischen Eigenschaften von Metalloxiden. Die Chemie und Physik von Oxidoberflächen und -grenzflächen steuern z.B. Prozesse in der Katalyse, Energieerzeugung (Brennstoffzellen), Mikroelektronik, Sensorik, Oberflächenvergütung, etc. Die atomare und elektronische Struktur von Oxidoberflächen kann jedoch stark von den bekannten Volumenstrukturen abweichen, d.h. Oxidoberflächen sind oft nicht nur einfache Schnitte durch den "bulk". Für Oxiddicken <10 nm sind die erreichten physikalischen, chemischen, elektrochemischen, elektronischen, magnetischen etc. Eigenschaften meist sehr verschieden von jenen der Volumenoxide. Modifizierte Oxide (Mischoxide, Dotierung, Metall-Oxid Kontakte) spielen in der Technologie eine noch wichtigere Rolle, allerdings fehlt z.Zt. weitgehend ein grundlegendes Verständnis des Zusammenhangs zw. Struktur und Eigenschaften. Wenn Oxide in der Technik Flüssigkeiten oder Gasen ausgesetzt werden, also als "funktionelle Oxide" wirken, werden die Struktur und Eigenschaften nochmals verändert. Die Weiterentwicklung oxidbasierter Technologien setzt ein grundlegendes Verständnis der Oberflächenchemie und -physik von Oxiden voraus und hängt v.a. von der Identifizierung von Struktur-Funktions-Beziehungen ab. Dies verlangt ein breites Spektrum wissenschaftlicher Herangehensweise und Methoden, was nur durch konzertierte Anstrengungen spezialisierter Forschergruppen erreicht werden kann. Dieser SFB vereint 11 experimentelle und theoretische Gruppen, die sich funktionellen Oxiden verschrieben haben, mit Schwerpunkten auf der Chemie/Physik von Oberflächen/Grenzflächen, auf Brennstoffzellen und auf Mikroelektronik. Der SFB schließt Forschergruppen zusammen, die zu verschiedenen "Scientific Communities" gehören, was zu einer neuen Qualität von Synergie und Interaktion führt. Erst durch den "Pool" an experimenteller und theoretischer Expertise wird eine umfassende Untersuchung von Prozessen an komplexen Oxiden möglich. Thematische Schwerpunkte sind: i) Struktur von funktionelle Oxidoberflächen/-grenzflächen; ii) Gas-Oberflächen Wechselwirkungen; iii) funktionelle Oxide unter Arbeitsbedingungen. Die gewählten Oxide (ZrO2 (auch Y-stabilisiert), CeO2, Perowskite, Oxid-Oxid und Oxid-Metall Kontakte) sind Schlüsselmaterialien der Katalyse, von Brennstoffzellen und in der Mikroelektronik, und werden mittels physikalischer (PVD, MBE, ALD, PLD, Lithographie) und chemischer (Sol-Gel) Methoden hergestellt. Die Charakterisierung der Oberflächen-, Grenzflächen- und Volumenstruktur erfolgt durch eine Vielzahl mikroskopischer/spektroskopischer Methoden (STM, AFM, HRTEM/SEM, PEEM/FIM, SXRD, XPS, AES, LEIS, PM-IRAS, SFG, FTIR, TPD, TOF-SIMS, UV-Vis, PL, Impedanz, C-V/I-V, DFT etc.), einige davon sind besonders geeignet, Oxide unter funktionellen Bedingungen zu charakterisieren. Die gemeinsamen Anstrengungen im SFB erlauben die physikalisch-chemischen Eigenschaften komplexer Oxidoberflächen und -grenzflächen auf molekularer Ebene zu verstehen, was u.a. Einblicke in die katalytischen Eigenschaften, in die Kinetik der Sauerstoffaktivierung an Perowskiten, und in die elektronischen Eigenschaften von Oxiden mikroelektronischer Bauteile erlaubt, aber auch verwandte Gebiete wie Sensorik, Korrosion, anorganische Strukturchemie u.a. beeinflusst.
The diversity of the application of ZrO2-based materials is considerable and directly reflects the manifold specific properties of this oxide. Different ZrO2 polymorphs have been intensively studied over decades, with applications in ceramic engineering, sensor technology, and in catalysis. In catalysis, ZrO2 is used both as catalyst itself and as support, e.g. for conversion of hydrocarbons in biomass-related reactions or doped with Y2O3 as an anode support and electrolyte material in fuel cells. It encompasses use over a wide temperature range between few hundred K up to 1273 K.The requirements on the intrinsic properties of the used ZrO2 material are thereby strongly depending on the specific application. Many empirical and fundamental research attempts have therefore been made on ZrO2-based materials to study their active sites and their selective generation and exact control with suitable stability. Parameters that have a strong influence are preparation routine, pre-treatment, phase (purity), termination, surface area, particle size, doping level or stoichiometry. However, open questions especially about phase stability, surface termination, surface activity and gas-phase-dependent stability limits still remain. The Zr-O phase diagram exhibits five stable ZrO2 polymorphs. The three most commonly used and best characterized phases exhibit the monoclinic (m), tetragonal (t) and cubic (c) structure. As m-ZrO2 is the by far most stable polymorph under realistic experimental conditions, it is most widely studied among the un-doped ZrO2 materials. The crystalline tetragonal phase and its stability (in terms of thermally or gas-phase-induced transformation into monoclinic ZrO2) is reported to be crucial for the performance of ZrO2-containing catalysts. Note that many preparation and pre-treatment attempts lead to different stabilities, but also activity and surface reactivity. To access the other un-doped polymorphs, specific annealing treatments are necessary. The temperature-induced phase transitions from m-ZrO2 to t-ZrO2 occurs at ~ 1400 K, and to c-ZrO2 above ~ 2600 K. At lower temperatures the c and t polymorphs are usually stabilized by doping with different metal cations (e.g. Y3+, Si4+, Zn2+)- or by either grain size or interfacial effects control (below a critical diameter d of d ≤ 33 nm). Also anions such as OH-, SO42- are potential stabilizers for low-temperature stable tetragonal ZrO2. In our recently submitted study we characterized the surface chemistry and gas-phase dependent transformation behavior of t-ZrO2 under oxidative, reductive and inert gas atmospheres of dry and moist gases He, O2, CO, CO2 and H2. The sample powder is prepared via a sol-gel route starting from a solution of zirconium(IV)isopropoxide in isopropanol without any further additives. Calcination of the resulting hydroxide aerogel at 673 K leads to structurally pure, un-doped and nano-crystalline t-ZrO2 with a defective and strongly hydroxylated structure. This t-ZrO2 phase is most likely stabilized via OHgroups, lattice anionic defects and crystallite size effects.16 Detailed information about its general kinetic stability in air has already been compiled.16 Transformation to m-ZrO2 is pronounced only upon annealing to 873 K and prolonged isothermal treatments. Less conversion takes place at isothermal treatments at higher temperatures up to 1273 K, but the main part of the sample is transformed to m-ZrO2 upon recooling to room temperature, once the critical temperature of 873 K is overcome. The transformation into the monoclinic phase is suppressed up to temperatures of ~723 K, independent of the gas phase composition15. At higher temperatures, the dependence of the extent of phase transformation on the gas phase environment is significant. In inert He atmospheres, the structural defectivity remains, leading to a high stability of t-ZrO2, even after a heating-cooling cycle up to 1273 K. In contrast, treatment in CO2 and H2 triggers the formation of a significant amount of m-ZrO2 (> 85 wt.-%) and the associated formation of a Zr-surface-(oxy-)carbide and dissolved hydrogen – which is not the case for the undefective m-ZrO2 reference sample. The factors that effectively steer the properties and stability of t-ZrO2 are (i) defect chemistry, (ii) hydroxylation degree and (iii) crystallite size. In particular moist conditions promote the t-to-m phase transformation, although at significantly higher temperatures as previously reported for doped t-ZrO2 samples. Our time resolved in-situ heating X-ray diffraction synchrotron studies of heating-cooling cycles in – up to date – CO2 and H2 atmosphere support our aforementioned findings concerning the t-ZrO2 stability and t-to-m transformation. The Rietveld-refinement reveals interesting details about the evolution of the t-ZrO2 crystallite sizes during cycling, that has not been reported so far and requires a further, detailed HR-TEM and EELS characterization: The initial t-ZrO2 samples with crystallite sizes of 21-26 nm (=below the critical 33 nm t-ZrO2 stabilization size) remain stable in size and structure up to ~1200 K. Upon further heating, the crystallite sizes increase to 38-42 nm at 1273 K while the t-ZrO2 structure still remains. At the beginning of the subsequent cooling cycle, the size of t-ZrO2 crystallites remains constant in both atmospheres. Upon further cooling below the t-to-m transformation temperatures (625 K in H2; 715 K in CO2) the t-to-m transformation is accompanied by a sharp decrease in crystallite size of the remaining t-ZrO2 down to 6-11 nm, while the newly formed m-ZrO2 has sizes of 22-27 nm (comparable to the initial t-ZrO2 size). The higher transformation temperature in CO2 in comparison to H2 atmosphere may be due to the bigger crystallite size of t-ZrO2 in CO2 than in H2 atmosphere. In essence, this is consistent with the suggestion that as the crystallite size of t-ZrO2 increases, the monoclinic-totetragonal phase transformation temperature increases.The interaction of different gas atmosphere with site defects on ZrO2 could serve as explanation for the lower the transformation temperature in H2 atmosphere. However, a crystallite size of the remaining t-ZrO2 phase being below the initial t-ZrO2 crystallite size has – to the best of our knowledge – so far never been reported. The relevant studies known to us report identical starting and finishing t-ZrO2 crystallite sizes.
The main aim of this proposal is to determine the crystal structure of a series of oxygendeficient lanthanum strontium ferrite La0.6Sr0.4FeO3-d (LSF) specimen in order to understand the role of oxygen and Fe deficiencies in their structural and physicochemical properties. As a mixed ionic and electronic conductor (MIEC), LSF is frequently investigated as a new material for solid oxide fuel cells (SOFCs) to lower the operating temperatures. Due to the duality of their conductivities, MIECs are suited for use as electrode as well as electrolyte materials. Our work showed that iron exsolution occurs in reductive environments at higher temperatures, leading to the formation of Fe rods on the surface. Utilizing an array of complementary in situ and ex situ techniques such as XPS, in situ XRD, scanning EXAFS, in situ TGA/DSC, and in situ TEM, we currently characterize this phenomenon regarding its thermodynamic and kinetic properties. According to in situ XRD, there are two distinct processes taking place in reductive (H2 and N2 atmospheres were tested) conditions: In H2, subsequently to the Fe exsolution, the perovskite exhibits a nonlinear shift of the diffraction peaks to smaller 2 theta at 300-350°C. At temperatures above 450°C Fe starts to crystallize (not shown here). The nonlinear reflex-shift can be assigned to a rhombohedral (R-3c) to cubic (Pm-3m) structural transition that may go along with a relocation of Fe atoms, possibly forming diffusion channels before segregating to the surface and crystallize as metallic iron agglomerates. Our time-resolved synchrotron in-situ XRD experiments, performed at Berkeley Labs Advanced Light Source (ALS), show that La0.6Sr0.4FeO3-δ heated and cooled in air undergoes reversible phase transition from rhombohedral (R-3c) to cubic (Pm-3m) perovskite at 300-350 °C. In contrast, this phase transition, taking place at ~350 °C for LSF heated in reducing H2 atmosphere was found irreversible upon cooling. Furthermore, small amounts (1-2 wt%) of metallic iron were segregated out from the cubic LSF phase above 450 °C, indicating Fe and oxygen deficiencies in the cubic LSF structure. Interestingly, in N2 atmosphere the cubic phase can be stabilized as well, the non-linear reflex shifting appears in a less pronounced manner as compared to H2 but no iron segregation is observed at elevated temperatures. TG analysis reveals that the LSF specimen heated in H2 atmosphere exhibits higher weight loss (≈2.5 wt%) in comparison to the sample heated in air atmosphere (≈1.3 wt%), suggesting the generation of oxygen vacancies in the sample heated in H2 atmosphere, the weight loss in N2 is about 0.6 wt% higher compared to air. These results suggest that the cubic (Pm-3m) perovskite recovered to room temperature upon cooling in H2 atmosphere is most probably stabilized by either oxygen vacancies or Fe deficiencies generated during heat treatment. However, the in-situ XRPD experiments do not allow us to refine the site occupancies of oxygen and Fe atoms in the cubic phase in order to confirm this hypothesis, therefore we propose to perform high-resolution neutron diffraction measurements on a series of La0.6Sr0.4FeO3-d samples treated in H2 and N2 and at different temperatures in order to have different O and Fe deficiency stages and to follow the transformation and stabilization process. Crystalline structure and especially O and Fe site occupancies can be refined with this data. The expected results will further expand our understanding of the relationship between structural and physicochemical properties in LSF.
Intermetallic Pd-Zr precatalysts show strongly improved methane dry reforming activites relative to pure Pd. Alterations in reactant activation and binding, e.g. strongly improved CO2 activation and pronounced redox and carburization behaviour of segregated surface Zr species are expected. Simultaneous oxidation of segregated surface Zr species by CO2 and carburization by C from CH4 adsorption is most likely inducing an “oxycarbidic” reaction mechanism. Reactive carbon species will be identified by studying the transient kinetics of carburization and carbon clean-off. Kinetic and spectroscopic effects related to (re)carburization and CO/hydrogen selectivity and activity will be investigated.
Tailored functional materials such as nanoparticles dispersed on a variety of different supports play an ever-increasing role in a wide range of research fields, such as heterogeneous catalysis, photocatalysis or energy conversion. Recently, an elegant pathway of creating highly regular well-dispersed metal nanoparticles on different perovskite materials has been reported, following efficient control of the anion non-stoichiometry of the material. In due course, exsolution of metal particles from a variety of such materials, including lanthanum strontium ferrites La0.6Sr0.4FeO3-δ or La0.3Sr0.7Fe0.7Cr0.3O3-δ has been documented and shown to have a significant impact on the physico-chemical or catalytic properties. In that respect, treatment of LSF under mild reductive conditions (using controlled cathodic polarization in a water electrolysis cell) in a H2/H2O mixture already led to iron exsolution and to a significant enhancement in the water-splitting kinetics [6]. Effective control of the experimental parameters, including reaction temperature, oxygen partial pressure and gas atmosphere is a key criterion for successful preparation in situ. The controlled exsolution does not only allow to steer the chemical properties, but also the morphology of the exsolved material. Besides irregular iron particles, reduction of LSF in dry flowing H2 at 600°C yields well-defined iron whiskers of about 20 nm thickness and several hundreds nanometers length. If small Ni particles are in addition dispersed on LSF, growth of strontium oxide nanorods of similar morphology and sizes and lengths has also been observed. This, too, raises the question about the metal-support interaction on metal-on-complex oxide systems. In that respect, the reduction induces segregation of Fe from the perovskite lattice and, depending on the perovskite systems, complete or partial alloying with Ni to form Ni-Fe solid solutions. In consequence, this alloying process has a significant influence on the catalytic behavior of the systems in the inverse water-gas shift reaction and methane reforming. Formation of the Ni-Fe solid solution appears to significantly suppress the overall activity of the catalytic systems.
Despite the progress made so far, the picture of exsolution remains on a rather phenomenological level and no systematic, knowledge-based approach has been followed yet. Reduction has been restricted to very high temperatures (Tmax=600°C) and also the parameter space (effective control of the reduction potential, reversible antisegregation) has not been fully explored up to date. This, however, is the prerequisite of a thorough understanding of the chemistry of the materials under question, allowing a knowledge-based transfer of general ideas and mechanisms also to similar systems and subsequently tailoring the physico-chemical and catalytic properties accordingly.
Within this proposal we focus on a systematic approach of resolving the mechanism of exsolution on two pure perovskite materials, namely La0.6Sr0.4FeO3-δ (LSF, lanthanum strontium ferrite) and SrTi0.7Fe0.3O3-δ, strontium titanium ferrite) and the respective Ni-decorated (10 mol%) LSF samples. The choice of the two model perovskites is fuelled by their distinctively different reduction behavior: LSF is much more easy to reduce than STF, thereby facilitating the exsolution of metal and/or oxidic particles from the perovskite lattice. To directly follow the exsolution process of iron and strontium (oxide) and to narrow down the parameter space, we anticipate to perform
(1) temperature-controlled scanning nano-XAFS or full-field XAFS experiments between room temperature and 500 °C in reductive atmospheres of different reduction strength (diluted pure hydrogen and H2/H2O mixtures of variable composition to adjust the reduction potential) on the pure LSF and STF materials,
(2) kinetic, time-resolved isothermal experiments at different reduction temperatures at and below 500 °C,
(3) temperature-controlled and time-resolved scanning nano-XAFS or full-field XAFS experiments between room temperature and 500 °C in reductive atmospheres of different reduction strength (diluted pure hydrogen and H2/H2O mixtures of variable composition to adjust the reduction potential) on the Ni-LSF material to follow the mechanism and kinetics of strontium oxide exsolution.
(4) antisegregation experiments using oxygen to controllably re-oxidize the respective samples to eventually close the full reduction-oxidation cycle and to determine if and to what extent the exsolution is a structurally reversible phenomenon.
Using this approach the obtained 2D images will thus reveal the distribution of iron and strontium as a function of time and temperature within the sample and in combination with XANES and EXAFS analysis also their chemical states during the exsolution process. For analysis the Fe K (at 7113 eV), Ni K (8383 eV), La K (38938 eV) and L3 (5489) eV as well as the Sr K (16108 eV) will be used for analysis. In the case of Ni-LSF, as the Ni K-edge appears after the Fe K-edge, EXAFS at the Fe K-edge and both XANES and EXAFS at Ni K-edge can therefore not be determined at the same time, because the Ni K-edge is superimposed on the Fe K-edge). Thus, the Fe K-edge XANES will be used for analysis.
Independently, the La L3-edge and Sr K-edge will also be measured. The requested beamlines are either BL37XU or BL39XU.
The anticipated results are expected to yield combined data on the spatial distribution of chemically different iron species within the sample and at the surface, thus allowing to establish a potential exsolution mechanism especially for pure LSF. For the corresponding Ni-LSF sample, the role of Ni could be clarified, since analytical aberration corrected electron microscopy suggests that strontium oxide exsolution proceeds exclusively on Ni sites, but not on the pure Ni-free perovskite. Up to date, a surface-bound mechanism on Ni (similar to extrusion carbon filament formation) is favored, but the detailed mechanism is unclear up to now
Fuel cells ("Solid Oxide Fuel Cells" SOFC) are promising elements for the production of energy, both in small, decentralized plants as a central component within a largely CO2 neutral energy cycle based on biogas as well as for portable electronic devices such as mobile phones and portable computers. Bimetallic Ni-Cu anode materials have compared to nickel some advantages, such as fewer tendencies to carbon precipitation. Thermodynamic and kinetic processes control the deposition of carbon in Ni-Cu-materials and yttriumstabilized zirconium oxide electrodes and have a negative effect on the energy efficiency, long term stability and reliability of SOFC from such material combinations. Cost, durability, reliability and dimensions of SOFC can be improved, by using micro-SOFC, produced by PVD thin film technologies. In this exploratory project Ni-Cu / YSZ model systems and a model µm-sized SOFC cell are prepared by magnetron sputtering and pulsed laser deposition. The composition, structure and interfaces of the model systems and micro-SOFC and deposition of carbon on the Ni-Cu-surfaces as well as its solution in the total volume before and after catalytic experiments at realistic working conditions are characterized by atomic force microscopy, scanning electron microscopy, Raman and FTIR spectroscopy, X-ray diffraction and high resolution transmission electron microscopy. It should provide new fundamental insights in the preparation and optimization of thin-film technology-based Ni-Cu/YSZ SOFC model fuel cell in the µm-range.
Perovskite-type electrode materials are successfully applied as solid oxide fuel cell (SOFC) cathodes and they are also promising SOFC anodes. For cathodes such as (La,Sr)(Co,Fe)O3-d (LSCF) the role of strontium segregation to the electrode surface is assumed to be responsible for performance changes. On perovskite-type anodes we could recently show that a polarization-induced modification of the surface strongly influences their electrochemical activity. In both cases the chemical surface composition and the near-surface cation valence states play a crucial role for heterogeneous (electro-)chemical reactions. Since most of the above mentioned phenomena occur only under SOFC operation conditions, in-situ investigations are urgently needed. Thus, nearambient pressure XPS investigations of LSCF electrodes in polarized electrochemical cells are proposed at 400 – 700 °C in different atmospheres (O2, H2/H2O/CO/CO2) with simultaneous measurement of the electrochemical properties.
Studien der letzten 20 Jahre haben immer wieder gezeigt, dass nominell wasserfreie Minerale unter Bedingungendes Oberen Erdmantels deutliche Gehalte an OH-Defekten aufweisen, und dass die Kenntnis des Einbaus und derVerteilung von OH von großer Bedeutung für viele Bereiche der Geowissenschaften ist, wie z.B. Petrologie,Geochemie und Mineralphysik. Im Oberen Erdmantel können Pyroxene als Hauptwasserträger angesehen werden.Beschaffenheit und Kozentration von OH-Defekten in Pyroxenenhängt von vielen petrologisch interessantenParametern ab, wie z.B. Druck, Temperatur, SiO2-Aktivität und Sauerstoff-Fugazität und kann daher als Sensor fürdie physiko-chemischen Bedingungen benutzt werden.
Im Rahmen dieses Projektes beabsichtigen wir, die räumliche Verteilung von OH-Defekten in Pyroxenen ausHochdruck-Synthesen zu untersuchen. Die Defekt-Verteilung soll dabei mit Hilfe einer neuartigen Technik auf derμm-Skala analysiert werden. Dabei werden die Proben mit einem FT-IR Mikroskop untersucht, dass mit einem sogenannten Flächendetektor (engl.: focal plane array detector, FPA) ausgestattet ist. Der FPA Detektor besteht aus 64 x 64 MCT Detektoren, wobei jeder Messpunkt durch die Position des jeweiligen Detektor-Elementes bestimmt wird und keine optische Apertur zur Maskierung des Messpunktes mehr notwendig ist. Auf diese Weise ist es möglich, auf einer Fläche von 170 x 170 μm IR-Spektren mit einer Ortsauflösung von 2,65 μm (also genauso gut oder besser als die physikalische Auflösung bedingt durch Wellenlänge der Strahlung) in weniger als einer Minute zu erfassen. Räumliche Variationen von OH-Defekten können so auch in sehr kleinen Kristallen entdeckt und als
Monitor für die vorherrschenden Kristallisations-Bedingungen während des Wachstums verwendet werden. Die Proben sollen auch mittels Transmissions-Elektronen-Mikroskopie (TEM) untersucht werden, um zu überprüfen, inwieweit die Bildung planarer Defekte (eines sehr effizienten Weges, um OH in den Kristall einzubauen) berücksichtigt werden muss. Außerdem sollen die Proben mittels Mikrosonde, Ionensonde, Raman- und Mößbauer-Spektroskopie charakterisiert werden. Die neuen Einblicke, die aus den Experimenten gewonnen werden, sollen auf natürliche Proben aus dem Erdmantel angewendet werden. Die Ergebnisse dieses Projektes lassen tiefere Einblicke bezüglich des Einflusses der SiO2-Aktivität und der Sauerstoff-Fugazität auf die Defekt-Chemie von Pyroxenen und der Entwicklung von Kristallisations-Bedingungen in Hochdruck-Experimenten im Allgemeinen erwarten. Neben dem Beitrag auf dem Gebiet der experimentellen Petrologie könnte dieses Projekt zum Verständnis des globalen Wasserkreislaufes beitragen, welcher seinerseits die Plattentektonik und das Ausmaß vulkanischer Aktivität beeinflusst und damit schlussendlich Einfluss auf das Klima.hat
Methanol steam reforming (MSR) is one of the most promising reactions to clean hydrogen. Upon usage of appropriate catalysts with a high CO2-selectivity, the CO content in the reformate gas can be sufficiently suppressed. The latter is e.g. a prerequisite for optimal solid oxide fuel cell (SOFC) operation. Different catalysts are in use, ranging from Cu/ZnO to oxide-supported Pd intermetallic systems [1]. Especially the latter have attracted recent interest due to the possibility of overcoming the inherent drawbacks of the industrially used Cu/ZnO catalysts: they are much more sinterstable, while still exhibiting comparable CO2 selectivities. Despite the frequent use, the nature of the active and selective state of these catalysts is still under discussion: while the necessity of the presence of the intermetallic compounds ZnPd or Ga2Pd is a widely accepted fact, the role of the oxide and/or the intermetallic interface is largely unknown. However, it is clear from recent experiments on ZnPd/ZnO and both surface science and catalytic studies of Pd-Ga intermetallic phases, that the mere presence of the unsupported isolated intermetallic compounds alone is insufficient to obtain high CO2 selectivities [2,3]. For ZnPd/ZnO, rather, a dynamic state of the catalyst with segregated and subsequently surface oxidized Zn (forming ZnO) on top of the ZnPd particles is adopted during the catalytic MSR reaction [4]. This has been deduced from operando X-ray absorption measurements [5], catalytic experiments and ex-situ electron microscopy [4]. However, using ex-situ methods, the possibility of structural changes introduced during transfer to the electron microscope cannot be avoided and are in turn highly likely for intermetallic compounds encompassing elements like Ga being even more prone to surface oxidation.
In this respect, the proposed operando high-resolution TEM measurements, the first of its kind to elucidate the nature of the active and selective catalytic site, are imperative. In the proposal, the focus of the experiments lies entirely on the corresponding Ga2Pd/Ga2O3 system, which for chemical reasons is much more complex than its already ex-situ studied ZnPd/ZnO system: spectroscopic measurements and ex-situ TEM work also suggest segregation and gas-intermetallic surface interaction to yield a dynamic catalyst state, steered by altered segregation behavior of Ga and a generally easier oxidation to different Ga-oxide species [6,7].
The proposed experiments will be performed on the most CO2-selective Ga2Pd/Ga2O3 powder catalyst as determined in catalytic tests prior to the actual TEM work. In due course, this catalyst will be first heated in hydrogen to induce the formation of the Ga2Pd phase on Ga2O3 at temperatures between 250 and 500°C, after which the sample will be exposed to a MSR mixture (methanol and water) with a defined, fixed gas phase composition. Control and stability of this composition is crucial to obtain a highly CO2-selective catalyst state. This will allow for the first time operando assessment of the working state of the catalytic active site in methanol steam reforming, which is indispensable for a thorough understanding of any catalyst system. Subsequently, the plan is also to test the stability of the intermetallic compounds by operando oxidative decomposition, thereby closing the full reduction-reaction-reoxidation cyle. As an ultimate goal, we aim at visualizing the working state of the catalyst as shown for the ZnPd/ZnO case under operando conditions. In due course, the nanoreactor experiments will be backed up by additional spectroscopic and tomography measurements, which will directly reveal the electronic structure and changes of especially the Ga2O3 support in the working state of the catalyst and the threedimensional distribution of Ga inside the intermetallic particle and at the metal-oxide interface.
In addition, we also aim at visualizing the transformation to the selective state operando with time: Exposure of a low temperature-reduced catalyst to methanol - water mixture induces a more gradually and slowly formation of the selective state compared to pure reduction in H2).
[1] D. R. Palo, R. A. Dagle and J. D. Holladay, Chem. Rev. 107 (2007) 3992-4021
[2] L. Mayr et al., J. Catal. 309 (2014) 231-240
[3] C. Rameshan et al. Angew. Chemie Intl. Ed. 49 (2010) 1-5
[4] M. Friedrich, S. Penner, M. Heggen, M. Armbrüster, Angew. Chemie Intl. Ed. 52 (2013) 4389-92
[5] A. Haghofer, K. Föttinger, M. Nachtegaal, M. Armbrüster, G. Rupprechter, J. Phys. Chem. C 116
(2012) 21816.
[6] S. Penner et al. Appl. Catal. A 358 (2009) 193-202
[7] A. Haghofer, D. Ferri, K. Föttinger, G. Rupprechter, ACS Catalysis 2 (2012) 2305-2315
[8] H. Lorenz et al. ChemCatChem 5 (2013) 1273-1285; A. Haghofer et al. J. Catal. 286 (2012) 13
[9] R. Leary, F. de la Peña, J.S. Barnard, Y. Luo, M. Armbrüster, J.M. Thomas, P.A. Midgley
ChemCatChem 5 (2013) 2599.
The proposed request is related to a finally better understanding of the nanoscale Kirkendall effect in intermetallic phases, leading to hollow structures. Although widely studied for chalcogenide systems (such as NiO), the understanding of the effect is still in its infant’s shoes regarding intermetallic compounds. This is especially a pity, since these compounds have recently gained a lot of interest in various research areas from ceramics especially also to catalysis – due to their exceptional structural and electronic properties. It has been shown that the nanoscale Kirkendall effect can be a convenient tool to access specially designed nanoparticulate systems with tailor-made physico-chemical properties: studied systems range from bulk samples over hollow particles to more complex entities such as nanotube-like systems.
Driven by diffusivity differences, the nanoscale Kirkendall effect, and the associated formation of hollow nanostructures, is naturally a kinetically dominated phenomenon. Work has so far been mainly focused on preparation issues, less so on the underlying fundamental scientific principles. Input does also arise from associated theoretical investigations, serving as a convenient base to discuss the experimental results. What is missing to obtain kinetic information, and therefore is centered at the core of the proposal, is a detailed operando monitoring of the phase-dependent void formation in different Pd-based intermetallic systems at different time and temperature scales. As the intermetallic compound formation is induced by reduction of an oxide (see below) in hydrogen, access to an operando nanoreactor coupled with high-resolution electron microscopy is highly needed. A typical experiment requires a flow of 1 ml s-1 hydrogen at temperatures between 250 and 600°C with simultaneous microscopy investigation. This would allow to exactly correlate the composition of the particles at the atomic scale with the extent of void formation and to subsequently study the stability of the nanoparticles. By analysis of time-dependent high-resolution microscopy measurements, a very important outcome of the proposal would also be the accessibility to otherwise hard-to-obtain interdiffusion data between Pd and the second metal/the intermetallic compound, to close the theory and experiment gap. Up to now, such values are only available for a very limited amount of systems or not available at all (representing the majority), subsequently hampering the analysis. Additionally, upon oxidation of the hollow intermetallic particles, previous measurements point toward a separation of the constituents of the intermetallic phases, which would suggest another diffusion-dominated effect. Hence, it would also be interesting to investigate this separation by operando oxidation in the nanoreactor (in oxygen).
The samples in question are Pd particles epitaxially grown on NaCl(001) cleavage faces and subsequently embedded in the oxide matrices of Ga2O3, SnO2 and GeO2. The resulting films are floated off and are finally mounted on gold grids for electron microscopy. Film thicknesses are in the range of 15-25 nm and are perfectly electron transparent. As the intermetallic compound particles are supported on the corresponding oxide matrix, a final question on the exact nature of the particle structure remains. This could be answered by associated detailed tomography experiments. So far the question, whether the particle is closed on all sides (therefore exhibiting a real hole), or if it is open on one side, is not solved.
The work outlined in the proposal will add to a better understanding of the structure-activity and selectivity correlation in complex Ni-perovskite systems. Specifically, small Ni particles will be deposited on two representative perovskite systems, namely Strontium-Titanium-Ferrite ((Sr(Fe0.3Ti0.7)O3-δ, STF) and Lanthanum-Strontium-Ferrite (La0.6Sr0.4FeO3-δ, LSF) and subjected to high-temperature reduction in hydrogen (600°C). This reduction induces segregation of Fe from the perovskite lattice and, depending on the perovskite systems, complete or partial alloying with Ni to form Ni-Fe solid solutions. In consequence, this alloying process has a significant influence on the catalytic behavior of the systems in the inverse water-gas shift reaction and methane reforming. Formation of the Ni-Fe solid solution appears to significantly suppress the overall activity of the catalytic systems. The importance of the study is also highlighted by the fact, that perovskite systems in general are highly valuable materials which are increasingly widespread in use in catalysis (see e.g. the overview paper by M. A. Pena and J. L. G. Fierro, Chem. Rev. 101 (2001) 1981-2017), e.g. as cathode material in solid-oxide fuel cells.
The proposal is an extension of a previous short-term scientific mission at ER Centre Jülich, where part of the samples have already been studied in cooperation with Dr. Marc Heggen. Ni particles on STF and LSF have already been characterized after performing the catalytic reaction, clearly indicating the formation of a Ni-Fe alloy phase (cf. Figure 1 below, highlighting EELS maps of individual alloyed Ni particles). In due course, with the present proposal we aim at a systematic approach, characterizing both samples (Ni on STF/LSF) in the fully oxidized state and directly after reduction, to get the most complete picture of the structure at each step of a full catalytic cycle. Combination of high-resolution imaging, EELS mapping and dedicated tomography imaging will directly reveal the three-dimensional Fe distribution inside the Ni particles and the Ni/perovskite interface. Furthermore, information on the chemical state of Fe could additionally be obtained. In this respect, for the first time a direct correlation between the structure of the active state and an associated catalytic profile for a complex perovskite system can be established, which is so far significantly hampered by the inherent structural complexity of these systems.
We propose to study inverse zirconia and alumina modified Cu and Pd model catalysts relevant for CO2-selective methanol reforming and CO oxidation. Cu, Pd will be exposed to variable amounts of Al and Zr by sputtering or ALD/CVD of organic precursors to obtain bimetallic or metal-(hydr)oxide precatalysts. Oxidative segregation of submonolayer (hydr)oxide layers is expected under water-rich or oxidizing reaction conditions. Ultrathin (hydr)oxide layers feature different electronic and geometric structures than the metal-bulk (hydr)oxide contact. Alterations in reactant binding, activation and redox behaviour are expected. The effects on CO and methanol conversion, CO2 selectivity and water activation will be investigated.
In addition to a range of well-studied copper catalysts, a group of highly CO2-selective Pd-based methanol steam reforming catalysts supported on ZnO, Ga2O3, and In2O3 has been identified recently. They all exhibit the common feature of formation of nanoparticulate Pd intermetallic phases, which are in contact with the respective reducible oxide. A combination of structural and electronic influences is likely to lead to high activity and selectivity towards additional hydrogen and CO2, due to the applied reductive activation procedure. The question is whether these “stereoelectronic” effects are limited to a certain bimetallic surface state for all these systems, or if they also affect the contact area between the bimetallic nanoparticles and the partially reduced oxide support, and the support itself. Even for the best-studied system, PdZn/ZnOx, the overall mechanistic role of the PdZn bimetallic surface in the reforming process toward CO2 and the influence of metal-support interactions is unclear. The presence of a bimetallic PdZn surface appears to be crucial for the steering of dehydrogenation steps in between methanol and CO. The most important open questions regard the essential intermediates leading selectively toward CO2, the mechanism and site of water activation, the role of the reduced support and the metal-oxide phase boundary, the stability of the bimetallic (nano)particles under “real” reaction conditions, their surface state and composition depending on size, chemical environment and temperature, i.e. the topic of chemically and thermally induced segregation trends. All these open questions have only in part been addressed for PdZn/ZnOx but not at all for the related Pd-Ga and Pd-In systems. From an argument of chemical analogy we suggest that also Pd/SnOx and Pd/GeOx are promising candidates for enhanced reforming selectivity, and we propose to include the analogous investigation of both the Pd-Sn and Pd-Ge model catalysts.
One of the central goals of this project is therefore to identify (or disprove) a potential “common reasoning” for the high CO2 selectivity of these systems, using a directional model catalyst approach to distinguish the influences of the different involved phases and their boundaries. We will put particular effort on the characterization of the catalytically most relevant surface composition of the models (purely metallic, oxidic, and with defined phase boundary) before and after reaction using a combination of electron spectroscopies and ion scattering techniques, and during reaction under “real” conditions, using in-situ spectroscopy techniques such as high-pressure XPS with tunable photon energy. Synergistic structural and electronic effects are already expected for the unsupported bi- or multi-metallic catalyst surfaces, by alteration of the d-band density of states (e.g. the DOS of Cu can be simulated due to the Zn-Pd “ligand effect”) and formation of bifunctional Pd-Zn surface ensembles acting as a structural promoter. A model catalyst approach is chosen which allows not only to distinguish the catalytic function of the bimetallic ligand- and ensemble-modified surfaces, but also that of the contact area of these with the respective oxide support in a defined state of surface and/or bulk reduction. The final goal is to identify the (combination of) micro- or rather “nanoscopic” catalytic units which enable highly selective low-temperature steam reforming and methanol synthesis in a variety of systems.
We propose to investigate the structure, composition and morphology of a variety of differently prepared bimetallic Pd-In, Pd-Sn and Pd-Ge catalysts relevant for CO2-selective steam reforming. Systems include well-defined Pd particles grown epitaxially on vacuum-cleaved NaCl(001) single crystals and subsequently covered by a layer of crystalline In2O3 or amorphous SnO2 and GeO2 (termed as “thin film model catalysts”). as well as, for comparison, conventional powder catalysts prepared by incipient wetness impregnation. Reduction at elevated temperatures in H2 (473–773 K) probably results in the formation of single Pd-In, Pd-Sn and Pd-Ge alloy phases. whose composition, morphology and electronic structure need to be examined by a combination of high-resolution TEM, SAED and EELS. These results will be the basis of subsequent catalytic measurements with the ultimate aim of establishing unambiguous structure-activity correlations in methanol steam reforming. Comparison of the differently prepared In2O, SnO2 and GeO2 supported Pd catalysts exhibiting different (bi-) metal-support contact area will aslo allow to determine its influence on catalytic activity and selectivity.
On the basis of previous work on In2O3 thin films, the TEM work on the bimetallic Pd particles should be complemented by similar investigations on the pure oxides, especially thin films of SnO2 and GeO2 prepared at different substrate (NaCl(001)) temperatures between 300 and 600 K and subjected to similar reductive treatments usually necessary for inducing alloy formation. This will allow to separate the effects only observed on the pure oxide support and in turn also to gain additional insight into the formation mechanism of alloy phases (e.g. alloy formation accompanied by migration or reduction of the oxide support…)