Interlayers of montmorillonite were pillared with TiO2, SiO2 -TiO2 and SiO2-Fe2O3 sols. The pillared structures were thermally stable at least up to 500°C and retained unusually large basal spacings in the range of 24–45Å and surface areas as high as 300–500 m2 /g. The TiO2 pillared clay showed Type IV adsorption isotherm for nitrogen. Although SiO2 -TiO2 and SiO2-Fe2O3 pillared clays had basal spacings much larger than that of TiO2 pillared clay, these mixed oxide pillared clays had small pores and exhibited Type I isotherm. The acidic strength distributions were determined by a titration method using n-butylamine and Hammett indicators. All of the pillared clays had large acidities, but the acidic strength decreased in the following order: TiO2 - SiO 2 » SiO2-Fe2O3 pillared clays. Temperature-programmed desorption (TPD) spectra of ammonia were measured and interpreted in relation to the acidity distribution. Infrared spectra of pyridine adsorbed on TiO2 pillared clay indicated that the acidity predominantly arises from Lewis acid sites.

Highly dispersed metal oxide/molecular sieve catalysts can be prepared through modification of metallosilicate molecular sieves. The resulting materials have been shown to exhibit enhanced catalytic activities for selected reactions. The metallosilicates examined include the gallosilicates, iron (ferri-) silicates and cobalt silicates. The size of the metal oxide particles as well as their location (in the pores or on the surface) can be controlled through post-synthesis methods. Thermal or mild hydrothermal treatment of the metallosilicate produces a highly dispersed metal oxide phase, while higher temperatures or longer treatment causes the metal oxide to migrate and form larger agglomerates of the metal oxide phase. The size of these agglomerates are strongly dependent on the conditions of hydrothermal treatment. Dispersion, location and agglomeration of the metal oxide phase have been characterized for the iron silicates using both physical and catalytic techniques. The magnetic properties of the iron silicates are described.

Bismuth(III) oxide, which is pale yellow, functions as an amphoteric semiconductor, although it is relatively unstable upon illumination with UV light. The energy level positioning, corrosion limits and photoactivity of samples of n-type bismuth oxide have been determined by electrochemical techniques. It was found that this material is able to oxidise water to molecular oxygen upon illumination with UV light.

Bismuth oxide forms a wide range of well-defined solid solutions with many other oxides and, in particular, niobium oxide can be used to form several distinct structures, each based upon defective fluorite. Each of these structures will function as an n-type semiconductor, retaining the basic properties of undoped bismuth oxide. However, the exact composition of the solid solution determines the band-gap energy of the semiconductor. Low concentrations of niobium in the lattice induce a substantial lowering in the band-gap energy and give rise to orange coloured materials that can collect a higher fraction of the solar spectrum. Photoelectrochemical studies showed that the presence of niobium caused the valence band to move to lower potentials but had little effect upon the energy of the conduction band. The photochemical properties of these solid solutions were studied using the photooxidation of alcohols as a test system. The results are discussed in terms of the structure of the oxide.

If the ionic radii are compatible, alkali metal ions will substitute for the divalent metal ions in magnesium oxide, calcium oxide and zinc oxide. At high temperatures in the presence of molecular oxygen, centers of the type [M+O−] are formed, where MM+– Group IA ion. The O− ions in equilibrium with these centers are effective in H” abstraction from CH4, which is the first step in the oxidative dimerization reaction. At reaction temperatures >720°C alkali metal oxides formed on the catalytic surface may themselves become active centers. Alkali metal carbonates also may inhibit the activity of the host oxide and sinter the oxide, thereby eliminating corner sites which result in complete oxidation.

It is possible to classify the solid state catalyst systems into two categories; structure static catalyst and structure dynamic catalyst. The former is the catalyst in which the original structural circumstances of surface active center, such as coordinatively unsaturated state of active element, length and order of metal-ligand bond or of a bond between structural component, and their bond angle, are maintained unchanged during catalysis. The example is molybdenum sulfid which is a good catalyst for the isomerization and hydrogenation of olefins. On the contrary, the catalysts, in which the above circumstances of surface active center are changing dynamically during catalysis, are classified into the latter category. Molybdate catalysts for olefin oxidation by molecular oxygen at high temperature belong to this category. Main difference between two categories is whether the originally coordinated ligand on metal cation, such as sulfur anion of molybdenum sulfid and oxygen anion of metal molybdate, can participate in the catalysis directly or not. When the coordinated ligand is involved in the reaction, resulting abnormal coordinatively unsaturated state of surface active center must be sustained by the dynamic structure changes of both surface and bulk for the steady state catalysis. The bulk structure, therefore, affects strongly the properties of structure dynamic catalysts, whereas the properties of structure static catalysts depend on the surface structure like coner and edge mainly. The relationship between catalytic properties and structural stability of molybdate catalysts in the selective oxidation of propylene was discussed on the basis of 18O tracer study.

Recently a lot of work has attempted to correlate the activity of Co and Mo sulfide catalysts with observations of high resolution electron microscopy. Hydrodesulfuration (HDS) bulk sulfide catalysts are specially suited for this type of study. We prepared unsupported molybdenum sulfide by three methods: the homogeneous sulfide precipitation (HSP) (1), co-maceration (CM) (2), and impregnated thiosalt decomposition (ITD) (3). Pure cobalt sulfide was prepared only by HSP, and two Co-Mo sulfides in the ratios of 0.3 and 0.5 at. % were prepared by HSP and ITD. The samples were treated with a mixture of 20% H2S/H2 for 4 hrs. at 400 C. In all Co/Co+Mo mixtures the presence of Co9S8 and MoS2 was identified in both, ESP and ITD treatments. Characterization was done by high resolutibn electron microscopy. The ITD method can be considered a better method, as opposed to ESP for the preparation of mixed catalysts. This method produces poorly crystalline specimens that are usually associated to higher chemical activity.

Discrete, organometallic clusters containing S, Mo, and Fe or Co have been deposited intact on metal oxide supports. These supported clusters are subjected to temperature programmed decomposition (TPDE) which causes loss of the organic ligands and produces catalysts active for CO hydrogenation and thiophene hydrodesulfurization (HDS). In-situ Mossbauer spectra reveal that the surface is remarkably uniform and that high spin ferrous iron is the predominate species (ca. 90%) under catalytic reaction conditions. EXAFS and XANES spectra show that the clusters are irreversibly oxidized by the surface at ca. 400 K under H2.

Adjustable (hydrogenolysis/hydrogenation) ratios (i.e., selectivity) of thiophene over sulfided ruthenium catalysts have been observed to depend on the method of presulfidization. Sulfidization in 10% H2S/H2 at 300–800K yielded a selectivity greater than 100, whereas a ratio of ca. 1 was noted after presulfidization in 100% H2S. The transition from one selectivity regime to another is reversible and can be changed via subsequent annealing or resulfidization procedures. X-Ray photoelectron spectroscopy shows similar results where sulfidization in 100% H2S yields: (i) a higher sulfur uptake with a corresponding 0.5 eV shift of the S 2P1/2–3/2 peaks to higher binding energy; (ii) multilayer sulfur incorporation at lower temperatures; and (iii) similar reversible behavior following subsequent annealing or resulfidization. Secondary ion mass spectroscopy and temperature programmed desorption studies also reveal that sulfur coverages affect thiophene adsorption, orientation and cracking on ruthenium surfaces.

A new synthesis technique produces a family of iron and iron alloy catalysts with a micro-morphology uniquely suited for ammonia synthesis catalysts.

Generalized Valence Bond (GVB) and Local Density Functional (LDF) calculations were carried out to explore the influence of electronic structure on the metal-support interaction in Pt/γAl2O3. On the basis of GVB calculations it was found that ethylene can form either di-σ- or π complexes on Pt depending on the oxidation state of the metal. The LDF calculations showed that chemisorbed Pt transfers charge to γ-Al2O3. These results and experimental observations led us to propose a model of the metal-support interaction in which the Pt exists in two oxidation states for reduced Pt/ Al2O3 catalysts.

The interaction of supported vanadium oxide with Al2O3 and TiO2 substrates is examined with Raman spectroscopy. The Raman spectra of the supported vanadium oxide reveal that the strong interaction of the vanadium oxide with the Al2O3 and TiO2 supports results in the formation of an atomically dispersed surface vanadium oxide phase as well as supported crystalline V2O5. The relative concentrations of the atomically dispersed surface vanadium oxide and crystalline V2O5 depend on the vanadium oxide loading and the surface area of the oxide support.

A series of iron oxide-chromia on alumina and iron oxide on magnesia materials were made by the aerogel process. These materials had high surface areas (300 to 700 m2/g) and demonstrated good activity for the selective reduction of nitric oxide by ammonia.

A combination of 4 parts Fe to 1 Part Cr gave the best catalyst for the reaction. Fe-Cr to alumina ratios were varied from 0.1 to 0.4. Activity per unit weight increased with increasing Fe-Cr/A12O3 ratios, however, activity per unit Fe-Cr favored the lower Fe-Cr to alumina ratios.

The fresh material from the autoclave was found reduced (Fe3O4) and initially had poor activity for the selective reduction of nitric oxide by ammonia (525 ppm NO & 525 ppm NH3 in air feed conditions). However, after several testing cycles, in which the catalyst was in continuous contact with oxygen (as air), activity improved dramatically. Upon removal from the reactor the material had a distinct color change (from brown to reddish brown) and magnetic properties were lost. This suggests that Fe2O3 is part of the active site for the selective reduction of nitric oxide by ammonia. At higher temperatures activity decreased because of a competing reaction, ammonia oxidation to nitric oxide.

Styrene-butadiene triblock copolymer PS-PB-PS was hydrogenated in the bulk using the Crabtree catalyst [Ir(COD)(py)(tcyp)]PF6 (COD = 1,5-cyclooctadiene, py = pyridene, tcyp = tricyclohexylphosphine). Since this polymer/catalyst mixture reacts rapidly with hydrogen at ambient temperature and low hydrogen pressures, it should act as an effective hydrogen getter.

FeTi is considered an excellent candidate for the reversible storage of hydrogen and has been studied extensively in an attempt to understand the bulk activation needed for this material before use. Segregation of TiO2 to the surface has been noted to occur during activation explaining a slight loss of efficiency per hydride cycle, however, characterization by a host of bulk and surface sensitive techniques has not revealed the cause of this decomposition process.

10 nm FeTi samples were prepared in a UHV evaporator both with and without palladium coatings. Post treatment characterization was performed with backscatter conversion electron Mossbauer spectroscopy (CEMS), XPS and SIMS. CEMS is a powerful tool for providing stoichiometric, electronic, magnetic, chemical, and particle size information of iron at depths down to 100 nm. XPS and SIMS are useful to gain quantitative and chemical state information from the topmost 2 nm and the topmost monolayer, respectively. Activation treatments consisted of annealing at 573K and 623K followed by reduction at 573K. Results indicate that ppm levels of H2O in H2 are sufficient to decompose the FeTi alloy and produce TiO2 and Fe metal domains at the surface. Also, at 573K in vacuum, a solid-state reaction was found to occur between Fe oxides and FeTi to produce Fe metal and TiO2. The Pd-FeTi interface was probed with CEMS and the results demonstrate hydrogen dissociation and migration in the absence of alloy decomposition. Our approach uses nondestructive-depth profiling of non-Pd coated FeTi samples along with interfacial information from Pd-FeTi specimens to obtain unique insight into the decomposition process.

Amorphous Cu70Zr30 ribbons heat treated at 473 K in a hydrogen atmosphere yield a very active catalyst for the hydrogenation of ethene and buta-diene. The catalytic activity is due to finely dispersed Cu particles formed on a surface containing numerous cracks. Auger depth profiles and scanning electron micrographs reveal details of the surface chemistry and morphology. The observations are related to the free-energy curve of bulk amorphous Cu-Zr.

Films of Pt, Pd, Rh, and Re with metal volume fractions of 0.3 to 0.5 have been prepared by mass-transport-limited photoelectrodeposition onto (001) p-InP photocathodes from ∼5 × 10−5 M solutions of the metal ions in 1 M HClO4. These films exhibit their normal catalytic activities (e.g., in hydrogen evolution) and have normal crystal structures, yet are substantially more transparent than equivalent dense films of the same metal loading per unit area. Effective-medium analysis of the spectroellipsometrically measured dielectric functions of these films shows that the anomalous transparency is due to microstructure: depolarization factors and metal packing fractions obtained by best-fit model calculations indicate dendritic (Rh), particulate (Pt, Pd), or platelet (Re) forms that are poorly interconnected in directions parallel to the surface, and whose dimensions are all small compared to the wavelength of light. Transmission electron micrographs confirm these results and reveal that these films consist of primary building blocks of ca. 5 nm crystallites that are organized into relatively loosely packed secondary structures. Potential applications of these films include the formation of efficient metallic-catalyst-coated photoelectrodes on poor-quality semiconductors.

The composition of NiCu surface alloys, prepared by decomposition of nickel carbonyl on a Cu(100) surface, has been studied before as well as after annealing using MeV Rutherford Backscattering Spectroscopy.

NiCu films with a thickness of 2 nm and of 10 nm, respectively, having a sharp interface with the copper substrate, were grown. The nickel fraction is found to be 65% and 93% respectively. The outermost monolayer is slightly enriched in copper. During annealing interdiffusion between the film and the interface occurs and the enrichment of the surface increases. In case of the 10 nm film the observed diffusion rate is much higher than is expected on the basis of diffusion coefficients published in the literature. This is ascribed to the presence of defects and/or grain boundaries in the grown surface alloy.

The copper catalyzed fluorination of silicon is first-order in [F2] and in [Cu]s until the coverage reaches ∼4 monolayers. Above ∼4 monolayers the reaction rate is zero order in copper, suggesting a limited number of catalytically active Cu/Si sites. Surface diffusion of copper leads to decrease in the etch rate as a function of time as well as feature size-dependent etch depths. The copper compounds CuF2, CuO, and copper silicides, Cu5 Si and Cu3 Si all catalyzed the F2-Si reaction which suggests that they are all converted to the same active species. The results can be explained by mechanisms involving copper fluorides or copper silicides as active intermediates.

Recently, a new materials characterization instrument and technique employing simultaneous differential scanning calorimetry/X-ray diffraction/mass spectrometry (DSC/XRD/MS) have been developed at Dow. Use of this technology can be illlustrated by the study of various materials such as polymers, organics/pharmaceuticals, inorganics, and catalysts.

Presented here is the use of the DSC/XRD/MS instrument to study the thermostructural behavior of four model copper-based catalyst systems during activation and regeneration. The instrument allows simultaneous generation of thermal, structural and chemical data in real-time during temperature programmed analysis and provides useful insights into the chemical and physical processes occurring. In addition, the calorimetry data yield qualititative information on the magnitude and rate of heat flow, while the diffraction data provide structural dynamics of reduction, oxidation and crystallite growth.

The results of this study conclusively show cuprous oxide as an intermediate in the reduction of the copper oxide portion of each of the model catalysts. However, such features as the onset temperature and copper surface area varied widely among the four catalysts in response to the same chemical event conducted under similar experimental conditions. On the other hand, oxidation runs on the reduced catalysts were all similar to each other, first producing cuprous oxide from copper metal over a broad range of temperature, followed by the oxidation of cuprous oxide to copper oxide at even higher temperatures.

Diffusion of small molecules in zeolites most often takes place in the configurational diffusion region, where the molecular diameter is approximately the same or slightly greater than the channel diameter. Since two molecules may not pass each other in a pore, the random walk based diffusion equation does not apply under high occupancy conditions in zeolites with one-dimensional pores.

For zeolites with multi-dimensional pores, such as ZSM-5 and A, one sometimes encounters the counter-intuitive result that diffusivity dramatically rises with occupancy. There are two explanations for this behavior: one from irreversible thermodynamics and the Darken equation, which predicts that diffusivity will always rise with occupancy; the other from a Markov model of random activated jumps between low energy positions, such as pore crossings, which predicts that diffusivity will increase with occupancy if the activation energy of diffusion decreases with occupancy– such as due to swelling.