Oxynitride and oxycarbide of niobium present metallic and acidic catalytic functions as evidenced by molecular probe reactions: isomerization of cyclohexane and hydrodenitrogenation of 1–4 tetrahydroquinoline. The proximity of the two functions leads to the concept of “dual site” by selective inhibition of the metallic and acid sites. Substitution of nitrogen of the oxynitride by carbon to form the oxycarbide produces a large enhancement of the metallic character of the surface. The oxynitride was shown to present a more acidic and a less hydrogenating activity as compared to the activity of the oxycarbide. Both compounds of niobium have been compared to molybdenum oxynitride.

Several methods of synthesizing molybdenum nitride or oxynitride fine powders are presented.

We have prepared a γ-Mo2N type oxynitride phase by reacting ammonia with MoO3. The surface area and morphology of the oxynitride powders depend on the synthesis conditions. Characterization of the solids by elemental analysis, X-ray and neutron diffraction, and thermogravimetric analysis shows dramatic modification of the stoichiometry of conventional Mo2N nitride. Aging at room temperature under air results in decreasing the material surface area. The initial surface area can be recovered be fine tuning of experimental conditions. MoCl5 and Ca3N2 are reacted in a molten CaCl2 medium leading to a new Mo2N structure type.

The reaction between molybdenum sulfide and NH3 produces two different phases depending on the reaction conditions. They are structurally related to δ-MoN.

Molybdenum nitrides have been shown to be active hydrotreating catalysts. We explored the hydrotreating activities of supported Mo nitrides. It was observed that the hydrodenitrogenation and hydrodesulfurization activities were inverse functions of the molybdenum loading. By using oxygen chemisorption and carbon monoxide temperature programmed desorption, we were able to characterize the particle morphology of the nitrides.

Nanoscale and well dispersed catalyst particles offer a large number of advantages: no diffusion resistance, easy accessibility to reactants, and a large number of active sites. In coal liquefaction, highly dispersed catalysts are especially needed because the catalyst particles are only able to influence reactions within their immediate vicinity.

The thermolytic decomposition of Mo(CO)6 with hydrogen sulfide or ammonia vapor (in a He carrier stream) at temperatures ranging from 300 to 1100 °C produces high surface area molybdenum sulfides (MoS2 or Mo2S3) or molybdenum carbides (hexagonal Mo2C) and carbonitrides, (hexagonal MoN(C) or cubic Mo2N(C)), respectively. The MoS2 surface areas range from 16.7 to 82.0 m2/g, while the surface areas of molybdenum carbides and carbonitrides vary from 14.9 to 21.1 m2/g. The maximum surface area for MoS2 is achieved at 500 °C and decreases with increasing or decreasing temperature. The surface area of the carbonitrides formed from 300 to 800 °C increases with increasing temperature up to 950 °C, where lower surface area Mo2C is formed. Crystallographically pure hexagonal MoN is prepared by decomposing Mo(CO) 6 in pure ammonia. Fe(CO) 5 decompositions in ammonia produce FexZ (where 5.8≥x≥1.6 and Z=C and N), and in some cases elemental Fe. Hexagonal Fe3 N(C) forms when Fe(CO) 5 is thermolyzed in ammonia from 300 to 600 °C, with surface areas ranging from 9.5 to 13.7 m2/g, whereas orthorhombic Fe3C and cubic Fe are produced at 700, 800, 900 and 1000 °C with surface areas of 6.7, 7.6, 2.2 and 2.0 m2/g, respectively. Within the same phase, the surface areas of the carbonitrides increase with increasing reaction temperature. These iron and molybdenum carbonitrides catalyze the conversion of CO/H2 to alkanes and methanol. Based on preliminary catalytic studies, the highest rate of methane (2850 g/kg/hr at 374 °C) and methanol (440 g/kg/hr at 284 °C) formation was accomplished with an FeMo carbonitride prepared by decomposing Mo(CO)6 and Fe(CO)5 in ammonia at 800 °C.

By vapor-phase pyrolysis of Fe(CO) 5 and NH3 induced by CW CO2 laser, γ'-Fe4N and ε- Fe3N nanoparticles(≤35nm) were prepared in Ar and N2 atmospheres. In Ar, γ'-Fe4N accompanied by a little α-Fe was formed at low temperatures below 650°C, while higher temperatures favored the formation of γ-Fe. In N2 atmospheres, a mixture of Fe4N and Fe3N nanoparticles were obtained at high temperatures. In addition, the influences of synthesis parameters on the structure, morphology and magnetic properties of the nanoparticles were discussed.

The interaction of NH3 and CH3NH2 with the surfaces of f β-Mo16N7, γ-Mo2N, and δ-MoN films was investigated using thermal desorption spectroscopy. Ammonia temperature programmed desorption (TPD) spectra for the films were similar. Ammonia TPD (adsorption at ∼280 K) produced an NH3 peak at ∼360 K. Some of the NH3 decomposed into H2 and N2. Two H2 desorption peaks were produced: a low temperature peak due to recombination of surface hydrogen and a high temperature peak due to hydrogen that emerged from the nitride subsurface. The N2 desorption spectrum consisted of a peak at ∼340 K and several peaks in the range 500–900 K. The desorption kinetics depended on the structure and composition of the film. 15NH3 TPD experiments indicated that the low temperature N2 desorption peak was due to NH3 decomposition, while the origin of the high temperature peaks was the nitride itself. We believe that nitrogen desorption from the nitride was induced by the presence of hydrogen which altered the Mo-N bonding. Three different CH3NH2 decomposition processes were observed: complete decomposition of CH3NH2 into H2, N2 and C, partial decomposition into HCN, and simple C-N bond hydrogenolysis into CH4 and NH3. The decomposition pathways depended on the structural and compositional properties of the films. All three processes were observed for the β- Mo16N7 (400), β-Mo16N7 (203) and δ-MoN(002) films, while only trace amounts of HCN were detected for the γ-Mo2N(200) film, suggesting that the partial decomposition into HCN was not favored on this surface. The β-Mo16N7 (400) and δ-MoN(002) films had the high selectivities for simple C-N bond cleavage.

An unsupported sample of Mo2N has been subjected to a first-stage hydrocracking test. The evolution of the HDS and HDN performance indicated a transformation of Mo2N into MoS2. This was substantiated by XPS and TEM, the latter technique showing that the transformation is limited to only a few surface layers.

A very active and selective molybdenum based catalyst for C6+ linear alkanes isomerization as been found. Physical characterization of this new active phase, such as XPS, XRD, MASN-MR and HRTEM is described.

Using both XRD and HRTEM lattice imaging, we have shown that CO2 laser pyrolysis (LP) produces nanoscale transition metal carbide and nitride catalysts, including cubic Mo2C, Mo2N, and W2N, which possess highly crystalline structures in their as-synthesized form In contrast, LP-produced W2C in its hexagonal phase is disordered. Clear lattice expansion, induced by the small crystallite size of the nanoparticles has been observed for LP-produced Mo2C particles, which have a typical crystallite size of 2 nm. No carbon coating was observed in HRTEM for LP-produced Mo2C particles. Furthermore, Mo=N and Mo=C bonding in Mo2N and Mo2C, respectively, were identified by an XPS measurement, which also reveals the presence of a thin oxide layer formed on the particle surface during the passivation process. Finally, the average crystallite sizes determined from HRTEM and XRD are in good agreement, indicating that the line broadening observed in XRD is due to the small crystallite size of the nanoparticles.

High surface area silicon carbide (SiC) of 31 m2/g has been synthesized by the catalytic conversion of activated carbon. The thermal stability in non-oxidizing environments is shown to be excellent; no significant sintering has been observed after ageing in nitrogen for 4 hours at 1273 K. The presence of 2v% steam or the use of air results in SiC oxidation into SiO2 and considerable sintering at 1273 K. Air oxidation of SiC is shown to cause substantial SiC conversion, viz. 60 % after 10 hours at 1273 K. Complete conversion is achieved at 1080 K in about 100 days. This rate of oxidation agrees with reports on the oxidation of non-porous Acheson SiC and Chemical Vapour Deposited SiC coatings. The use of SiC based catalysts is, therefore, limited to (1) high temperature gas phase reactions operating in the absence of oxidizing constituents (O2 or H2O) and (2) liquid phase processes at demanding pH. Syntheses of highly dispersed and highly loaded Ni/SiC catalysts are feasible by applying an ion-exchange technique, resulting in supported nickel particles of 4 nm.

The presence of PdCx, a metastable palladium carbide phase, was detected in Pd dispersed on high surface area carbon blacks using a Pd acetylacetonate precursor. TEM and XRD measurements indicated well dispersed Pd after reduction at 573 K; however, hydride formation and chemisorption of gases was suppressed on these catalysts and this was accompanied by lattice expansion of Pd as seen from XRD. Cleaning pretreatments with O2 slightly contracted the Pd lattice and enhanced hydride formation and chemisorption, while a 673 K reduction gave normal hydride ratios and Pd lattice parameters although chemisorption was still suppressed. Pd lattice expansion and suppression of hydride formation and chemisorption are attributed to carbon atoms present both in the bulk and on the Pd surface. The performance of these catalysts in probe reactions was consistent with this model of C contamination of the Pd surface and bulk.

Especially with the production of fine chemicals catalytic reactions are being performed in which solid catalysts suspended in liquids are used. Carbon is stable in both acid and alkaline liquids and is not preferentially wetted by water. Carbon is therefore an attractive support for catalysts to be used in liquid-phase reactions. The relatively low diffusion coefficients in liquids call for catalyst bodies having wide pores and dimensions between 1 and 10 μm. Since particles smaller than 1 μm cannot be readily separated from liquids, attrition of carbon support bodies must be avoided. However, it is difficult to produce active carbon bodies of 1 to 10 μm of a high mechanical strength. Also with gas-phase reactions, the low mechanical strength of active carbon bodies limits the applicability of active carbon.

From supported metal particles carbon filaments of a diameter between 10 and 300 nm can be grown rapidly. The diameter is generally of the same order of magnitude as the size of the metal particles. The graphite layers within the fibrils, that have a circular cross-section, are oriented either parallel to the fibril axis or at an angle, which leads to cone-shaped graphite layers. The mechanism of growth of carbon fibrils from supported metal particles will be dealt with. The parallel orientation of the graphite layers leads to a very high mechanical strength, while the porous structure is extremely open. A representative value for the surface area is 225 m2/g and for the pore volume 1.6 ml/g, which leads to an average pore dimension of 28 nm.

To apply active precursors, such as, palladium salts, the surface of the carbon fibrils must be oxidized, e.g., by treatment with nitric acid. A very high dispersion of palladium can thus be achieved. The dispersion of components active in hydrogenation can be assessed by the ability to react the support with hydrogen to methane. The activity of catalyst on carbon fibrils in the hydrogenation of nitrobenzene as well as the filterability of carbon fibril catalysts will be compared with that of commercial catalysts.

Catalytically grown carbon nanofibers have been prepared by the thermal decomposition of carbon containing gases over copper-nickel and iron surfaces. This material is found to be highly graphitic in nature when prepared from certain catalysts and gaseous reactants. In the as-grown state, carbon nanofibers have surface areas in the range 200 to 300 m2/g, and by following careful activation procedures this value can readily be increased to ˜700 m2/g. Electrical measurements indicate that the material has a conductivity approaching that of single crystal graphite. This material combines the attributes of active carbon and graphite and in addition, the physical form of carbon nanofibers offers some interesting opportunities for the design of unique catalyst systems.

The most widespread methoid for buckytube synthesis is by arc discharge in an inert gas ambient. At arc discharge the major kind of charge carriers are the electrons generated from the cathode by thermal emission. For this reason up until now the deposit was always obtained on the cathode. By supplementary electrode heating and cooling we studied the effects of the anode-cathode temperature difference and found that it plays a major role in defining the quality and quantity of the deposit obtained. Given appropriate heat-up or cool-down of the electrodes, a state can be reached such that the temperature at the cathode is higher than the temperature at the anode. A deposit was obtained on the anode, observed by a TEM. In our view employing this method yields the most favorable conditions for buckytube growth. Our results show that except buckytubes with a high ratio of their length with respect to their width, a multitude of small size polyhedral carbon systems closed in by spatial carbon nets and thin flat interfering carbon plates were also obtained. When using electrodes of different cross-section, with the deposition of an ever thicker deposit on the anode, its temperature regime changes. The lower initial temperature at the anode begins to grow steadily with time tending towards the temperature at the cathode. Thus, buckytube growth proceeds at the lowest possible rate, at a small temperature difference between the vapor source and growing structure. In our view employing this method yields the most favorable conditions for buckytube growth.

We describe synthetic strategies for designing cobalt Fischer-Tropsch catalysts in order to maximize synthesis rates and C5+ selectivity. The support cobalt-precursor interaction must be balanced to maximize dispersion yet allow facile low temperature reduction. The cobalt time yield is independent of support and varies directly as the cobalt dispersion for dispersions in the 0.02 to 0.12 range [1].

Catalyst drying procedures are often given little attention in the experimental section of papers on supported metal catalysts. In general, drying appears to be regarded as a method to remove water and other volatile components prior to calcining or reduction, but not as a method to affect the surface properties of the catalyst. This study uses x-ray photoelectron spectroscopy (XPS) to examine the surface properties of silica supported cobalt catalysts, prepared using incipient wetness impregnation of cobalt nitrate hexahydrate, and finds a wide range of cobalt distributions, extent of nitrate decomposition, and reducibility for various drying procedures. After UHV annealing and subsequent reduction, the final cobalt surface properties are found to depend on the length of heating and the environment during the drying process. Maximum cobalt metal surface area is obtained for samples exposed to limited amounts of air and dried under conditions where gas phase species generated during the precursor decomposition are rapidly removed from the surface of the sample.

Chemical reactions of metals and strongly interacting alloys such as Cu-Mn, Ni-Ti, Ni-Hf and Ni-Zr with oxygen and hydrogen play important roles in the preparation, activation, and regeneration of many important heterogeneous catalytic systems involving supported and unsupported metals and alloys. Recent advances in the understanding of metal and alloy oxidation is bringing new insight into the reactive design and activation of bi- and multi-metallic catalysts. By surface studies of oxidation, thermal annealing and reduction of selected alloys and their thin films and reaction layers and products we have been able to delineate the factors which are most important to the oxide formation processes and the oxide reduction processes. Reaction models developed from these results are permitting the design of new catalyst systems and providing long sought understanding to explain specific aspects of well established metallic catalysts.

A novel supported metal catalyst has been prepared, that consists of Ag clusters a few nanometers in diameter which are encapsulated in an amorphous silica coating only a few Ångstroms in thickness and are deposited on an oxide support. The synthesis of this catalyst and its physical characterization are described. Its potential use as a selective oxidation catalyst is discussed.

The characterization of nanoparticles is of prime importance in catalysis. High Resolution Electron Microscopy coupled with image processing has produce a lot of new information on the detailed structure of the particles. In this paper we discuss the possibility of applying these techniques to the study of surface roughness.