J.T. Richardson

Origin of superparamagnetism in nickel oxide

It has been known for some time that particles of nickel oxide of less than about 100 nm in size show superparamagnetism that increases as the particle size decreases. The origin of the particle magnetic moment responsible for this behavior has never been fully explained. This research indicated that the size of the particle rather than the presence of nonstoichiometry or impurities of reduced nickel determines the moment. The critical experiment was the measurement of magnetization versus magnetic field for a sample of nickel oxide prepared under conditions that preclude metallic nickel. Almost identical results were found for the original sample, which was black in color and thus nonstoichiometric, and after mild reduction in hydrogen at 400 K, which produced stoichiometry and changed the color to green. The magnetic susceptibility was inversely proportional to the particle size for a given method of preparation. This is consistent with a simple model of incomplete edges on the bounding planes of the cr...

Reduction of model steam reforming catalysts : NiO/α-Al2O3

Abstract The effect of NiO loading (5–21 wt.-%) on the reduction of NiO/α-A 2 O 3 catalysts, prepared by multiple impregnation of nickel nitrate solution followed by calcination at 650°C has been characterized usingtemperature-programmed reduction, isothermal hydrogen consumption, magnetization, X-raydiffraction, and electron microscopy. X-ray diffraction analysis of fresh catalysts indicated normal NiO crystallites about 30 nm in size. Studies from 270°C to 350°C show hydrogen consumption at lower temperatures is faster than the subsequent growth of nucleated clusters of nickel atoms into crystallites, with the rates of the two processes approaching each other at higher temperatures. As NiO loading increases, chemical reduction becomes more difficult but nickel crystallite growth is not affected. This decreased reducibility is believed due to Al 3+ ion incorporation into NiO surface layers during impregnation. Isothermal hydrogen consumption from 270°C to 450°C follows a shrinking core model with NiO crystallites decreasing progressively in size. Magnetic measurements show crystallite growth is dependent on diffusion-controlled nucleation. Decreasing hydrogen flow has a profound effect on chemical reduction and nucleation but not on growth. Similar results are found with added water vapor. X-ray diffraction and transmission electron microscope measurements reveal 23 nm nickel crystallites with some evidence for Al-Ni alloy formation. A mechanism is proposed in whichadsorbed water inhibits chemical reduction and nucleation, and foreign ions such as A1 3+ increase this effect.

Reduction of model steam reforming catalysts: effect of oxide additives

This paper is part of a series on the reduction of NiO catalysts and reports results for 4–15 wt.-% NiO, supported on α-Al2O3 and containing small amounts of additive oxides, such as SrO, La2O3, CaO, and MgO. Rupture of NiO bonds by H2 was measured by H2 consumption, both isothermally and with temperature programmed reduction, and growth of Ni atoms into metallic crystallites with isothermal magnetization techniques. Chemical reduction follows a shrinking core model, whereas crystallite growth is best described with a nucleation model in which diffusion of Ni0 atoms from reduction sites to nucleation centers controls the rate. The rates of both these processes are impeded by adsorbed H2O. Hydrophilic additives retain surface H2O, even when low levels of these additives are present, thereby slowing reduction rates so that reduction must be carried out at significantly higher temperatures. A model is presented to explain these effects and to rationalize the order for reduction temperature: none

Preparation variables in nickel catalysts

Abstract Nickel-on-silica catalysts were prepared by the homogeneous precipitation of slowly decomposing urea. After reduction the crystallite size distribution (CSD) of the nickel was determined from magnetic measurements. This technique was found to give very narrow CSDs and is much more reproducible than conventional impregnation methods. The nickel loading may be controlled by the solution concentrations and time of precipitation but higher values lead to broader distributions. The amount of reduced nickel is a function of the hydrogen flow rate and time, but these factors do not influence the CSD. Increasing reduction temperatures produce broader CSDs. Calcination has no effect on the resulting CSDs, nor does passivation in air after reduction. The uniform and stable crystallites resulting from this method are a consequence of the precipitation of nickel hydrosilicate homogeneously over the silica surface. This compound reduces slowly so that the metal does not sinter during reduction and small, uniform crystallites are produced.

Reduction of impregnated NiO/α-A12O3 association of A13+ ions with NiO

Contrary to common belief that α-Al2O3 is an inert support for the impregnation of base metal species, evidence is presented for the incorporation of Al3+ ions into NiO impregnated on α-Al2O3, thereby causing the observed increase in reduction temperature. In agreement with this proposed model of support dissolution/incorporation into the impregnated material, temperature programmed reduction measurements on samples of NiO prepared by coprecipitation with small amounts of Al3+ and by multiple impregnation of α-Al2O3 gave similar results. As predicted, treating α-Al2O3, in the form of rings or powder, with HNO3 extracted small amounts of Al3+ at 30°C and larger amounts at 90°C. During the calcination step following impregnation with Ni(NO3)2,6H2O, NiO is formed as the Ni(NO3)2 crystals decompose and Al3+ ions are extracted from the surface. These Al3+ ions are then retained either in or in the vicinity of evolving NiO crystallites. Successive impregnation steps result in further Al3+ extraction from the support and additional increases in reduction temperature.

Combustion synthesis and characterization of Sr and Ga doped LaFeO3

Abstract A homogeneous perovskite oxide La 0.5 Sr 0.5 Ga 0.2 Fe 0.8 O 3− δ (LSGFO) has been synthesized by self-propagating high-temperature synthesis (SHS). The homogeneity and the particle size of the combustion product may be increased by decreasing the cooling rate of the sample, either by increasing the sample diameter or by controlling the post-combustion temperature. The particle morphology depends on the gaseous and molten compounds formed as the combustion front passes through the sample. The perovskite oxide maintained its cubic structure at all temperatures in air. However, a decomposition of LSGFO occurred at 860°C under a simulated syngas environment (22% CH 4 +57% H 2 +21% CO 2 , oxygen partial pressure of about 10 −17 atm). The maximum electrical conductivity of a disc prepared from the SHS powder was 142 S/cm at 580°C under oxygen pressure of 1 atm. The LSGFO may be suitable for use as a membrane in syngas production since its thermal expansion in air and reducing environment are rather close at high temperature.

Alkali promotion of nickel catalysts for carbon monoxide methanation

Differential reactor studies of carbon monoxide hydrogenation have been carried out on a series of 8% NiSiO2Al2O3 catalysts in which the supports have been impregnated with increasing amounts of sodium. Magnetic measurements show that the extent of reduction and the nickel crystallite size are not influenced by the sodium on the support. Hydrogen adsorption measurements reveal that the exposed surfaces are in agreement with the crystallite sizes and are not significantly poisoned by the alkali. The turnover number increases with sodium content and passes through a maximum at 0.3% Na with an enhancement by a factor of 6. Kinetic data best fit the surface reaction mechanism for which rate = kKCOPCOPH12H(1 + KCOPCO)2 For KCO, the pre-exponential factor and heat of adsorption are unchanged by Na, but the parameters of k are not. The pre-exponential factor passes through a maximum and the activation energy decreases as the Na level increases. These results are best explained by an initial decrease in acidity, which keeps the nickel surface clean, followed by a poisoning via direct metal interaction or support modification by the alkali.

Crystallite size distributions of sintered nickel catalysts

Abstract Sintering of nickel on silica was followed by measuring the crystallite size distribution with a magnetic granulometry method. The initial distribution of a sample prepared by slow, homogeneous precipitation-deposition was very narrow but could be broadened by thermal treatment. The effects of temperature, time, initial distribution, and nickel concentration were studied. The dispersion was very stable at 673 K but showed some broadening for sintering times up to 100 hr. Above 723 K, however, sintering occurred with the disappearance of small crystallites, reaching a limiting log-normal shape, independent of initial distribution or nickel concentration below 17 wt%. These results were interpreted as particle migration. The decline of surface area showed an order of 10 and an activation energy of 200 kJ/mole. At temperatures approaching 873 K, the resulting distributions were bimodal, suggesting that the influence of pore-size distributions should be considered in sintering models.

Effects of promoter oxides on the reduction of nickel oxide

Abstract Bulk nickel oxide samples promoted with 2–3 wt.% CuO, Ag 2 O, Al 2 O 3 and ZrO 2 were tested for reducibility in hydrogen using three independent techniques: magnetization measurements, thermal gravimetric reduction and temperature-programmed reduction. All three showed the same trend. Reducibility, as measured by the extent of reduction at a given temperature, follows the order CuO ≈ Ag 2 O ≫ none ≫ Al 2 O 3 ≫ ZrO 2 . X-ray diffraction indicated the absence of any promoter crystallites larger than 2.5 nm. This and the variation of derived lattice strain parameters are consistent with their successful incorporation into the NiO lattice. Lattice parameters and crystallite sizes were essentially the same. A model is proposed in which the initial stage of nickel oxide reduction is nucleation of nickel clusters, aided or hindered by the presence of more or less easily reduced cation-oxygen bonds in the surface. The conflict about the role of silver in promoting reduction of nickel oxide has been resolved. A steam reforming-type catalyst was tested for activity during self-reduction with and without promotion by Ag 2 O. Moderate improvement in reducibility occurred but not when the catalyst was precalcined at 400 °C, showing that decomposition to silver metal destroys the effect.

Crystallite size effects in nickel catalysts: Cyclohexane dehydrogenation and hydrogenolysis

Abstract Crystallite size dependence of cyclohexane dehydrogenation and hydrogenolysis has been studied over silica-supported nickel catalysts. The samples were freshly prepared and sintered catalysts, produced by homogeneous precipitation-deposition and ranging from 25 to 40 wt% Ni on SiO 2 . Crystallite size distributions were determined from magnetization data. Other experimental measurements, such as hydrogen chemisorption and catalytic kinetics, were made in the same cell in order to provide in situ precision. Average crystallite sizes varied from 2 to 4 nm and critical differences were found in this region. Areal rates for dehydrogenation more than doubled, whereas cyclohexane hydrogenolysis decreased by a factor of 5. These findings were shown to be consistent with a model wherein dehydrogenation occurs on crystal face sites and hydrogenolysis on edge positions. Benzene hydrogenolysis followed the same pattern as dehydrogenation, suggesting that common intermediates are involved.