On shape normalization for non-uniformly active catalyst pellets—II

A shape normalization is developed for non-uniformly active catalyst pellets with an isothermal first-order reaction. A variety of activity profiles, including step-distribution and those which increase or decrease towards the pellet surface, are considered. Mass transfer in the outer inert zone is found to be characterized by a single dimensionless parameter, called the internal Biot number. The effect of finite external mass transfer is also taken into account. Analytic solutions and numerical comparisons in terms of the normalized significant parameters are reported for various activity profiles and pellet shapes. The shape normalized curves, obtained for different geometries, activity profiles and external mass transfer, match perfectly for small or large values of the Thiele modulus. However, as in part I of this paper (and as for positive-order reactions in a uniformly active slab catalyst where a normalization for the reaction order is sought), the normalized curves can differ significantly for intermediate values of the Thiele modulus. The particular case of activity distributions which drop continuously to zero within the pellet is excluded from the shape normalization, when mass transfer resistance—both in the bulk phase external to the pellet and in the outer inert region of the pellet—is negligible.     

On the inevitability of non-uniform foulant deposition within a catalyst pellet

The series pore model has been used to investigate the interaction of pore structure and foulant deposit geometries during the deposition of a foulant     

Shape normalization of catalyst pellet

A shape normalization, which is applicable in the entire range of Thiele modulus φ, is developed. A shape normalization established here for small φ and the shape normalization already established here for large φ are used in developing the normalization for all φ. This normalization brings the η - φ curves for all pellet shapes to a single curve corresponding to infinite slab geometry for all φ. The effectiveness factor for any shape of catalyst is simply the effectiveness factor for an infinite slab when the Thiele modulus for the slab is properly defined in terms of the characteristic pellet length and the reaction kinetics. The shape normalization is shown to give negligible error for any pellet configuration and first order reaction, and is postulated to hold for general kinetics and any pellet configuration, by proper definition of the Thiele modulus.     

A basic study of catalyst aging in the H-coal process

Samples of CoMo/Al₂O₃ catalysts used in an H-coal process demonstration run were studied to determine causes of catalyst deactivation. Physical and surface properties of the aged and regenerated catalysts were examined. Model compounds were used to assess four catalyst activity functions, viz., hydrodesulfurization (HDS), hydrogenation, cracking and hydrodeoxygenation (HDO). Other tests were performed to study the effects of coke and metals separately on the four catalyst activity functions.Catalyst coke content and metals deposits first increased rapidly, then more gradually with exposure time in the process run. Surface area and pure volume markedly decreased with exposure time. Catalyst activities of aged catalysts showed a rapid decline with exposure time. One-day exposure to coal resulted in significant losses in HDS and hydrogenation activities and nearly complete loss in cracking and HDO activities. Although metal deposits caused some permanent catalyst deactivation, coke had a much greater effect. Regenerated catalysts showed less recovery of catalytic activity as processing time increased. These results agreed well with product inspections from the process run.Oxygen chemisorption on aged—regenerated catalysts decreased with catalyst exposure time, indicating a significant loss of active sites. However, ESCA results showed no evidence of extensive sintering of the active MoS₂ phase. Permanent deactivation of the longer-time exposed catalysts can be ascribed, at least partly, to lateral growth of the active molybdenum sulfide phase. In addition, some loss in cobalt promotion occurred early in the process, which may account for the rapid loss in HDS and HDO activity in regenerated catalysts.     

The effect of catalyst temperature zoning on hydrotreating a coal-derived liquid

The effect of catalyst temperature zoning on hydrotreating reactions, coke formation, and changes in catalyst surface area and pore volume was examined. A coal liquid was hydrotreated with the top zone operated in the range of 400–500° C (752–932° F) while the lower zone was maintained at 400° C (752° F). Increasing the top zone from 400° C (752° F) to 450° C (842° F) had beneficial effects on hydrogenation, HDN, and HDM at no cost of increased catalyst coke content. A further temperature increase was detrimental because of excessive coking and possible thermodynamic equilibrium limitations.     

Thermodynamics of alcohol-unassociated active component liquid mixtures

An associated solution theory based on the UNIQUAC equation is developed to describe the excess properties of the alcohols in unassociated active components. The theory includes two association constants: the one for the alcohols, the other for alcohol-unassociated active component complexes. The theory reproduces well the activity coefficients of the ethanol—trichloromethane system having a maximum and a minimum and it also fits very skewed excess enthalpy curves for alcohol—benzene, alcohol—tetrachloromethane, etc. Prediction of ternary vapor—liquid and liquid—liquid equilibrium and excess enthalpy data has been carried out with good accuracy from binary parameters without adding any ternary constant.     

Changes in physicochemical properties of active carbon within a grain

Commercial active carbons were subjected to abrasion in a spouted bed and as a result, active carbons, with external layers removed to different degrees, were obtained. For active carbons so obtained, physicochemical properties were determined from densimetric and adsorption investigation.It was found that radial changes in the properties of active carbons resulted from the nature of the activation reaction.It seems that the method of working by abrasion in spouted beds can also be used for investigating other processes taking place in fluid—solid system.     

A simple relationship between particle shape effects and density,flow rate and Hausner Ratio

A study has been made of the relative magnitudes of particle shape effects on three bulk solids properties. The bulk solids used in this study were small plastic pellets of regular shapes and sizes. Experimental results indicate that variations in particle normal and tangential not-roundness have marked effects on the apparent density, flow rate and Hausner Ratio of the bulk solids. The not-roundness is the standard deviation of the radial distribution of the particle profile. It is a particle shape characteristic derived from morphological analysis.A simple relation of the form P(bulk property) = P(ideal) + P(deviation) is proposed and tested, in which P indicates the property under test. P(ideal) indicates the value of the property for perfect spheres; P(deviation accounts for the effect of particle shape.     

Oscillations in crystallite shape during heating in hydrogen of model iron/alumina catalysts

Transmission electron microscopy is used to bring evidence for an alternation in the shape of iron crystallites supported on planar, nonporous, alumina substrates, when the specimens were heated, at 400 °C, in hydrogen contaminated with traces of oxygen (less than 1 ppm) and/or moisture. The crystallites alternated between a torus which encloses a cavity and a shape in which an annular gap separates the torus from a core in the cavity. Electron diffraction indicated that this is accompanied by a corresponding alternation in the chemical state of the catalyst. The torus with an enclosed cavity corresponds to an oxidized state, in which iron oxides, resulting on oxidation by the contaminant oxygen, form solid solutions in alumina, with the tendency to approach the aluminates FeAl₂O₄ and/or Al₂Fe₂O₆. The torus containing a core in the cavity corresponds to a less-oxidized state, in which a part of the previously dissolved oxide diffuses out to be (at least partially) reduced by hydrogen to the metal and/or to a lower oxide (including possibly nonstoichiometric oxides). After prolonged heating, the two alternating shapes become less distinct, probably because of mechanical fatigue which eventually leads to a breakup of the crystallites. Similar behavior is observed at temperatures up to 600 °C, though with diminished sharpness of the two alternating shapes above 500 °C. An attempt is made to explain the shape alternation on the basis of surface phenomena resulting from the physical and chemical interactions between gas, crystallite, and substrate.     

Isothermal diffusion—reaction in a slab catalyst with bimolecular Langmuir-Hinshelwood kinetics: Connections with negative first-order kinetics

The problem of isothermal diffusion-reaction in a slab catalyst with bimolecular Langmuir—Hinshelwood kinetics is studied. Analytic solutions for the concentration profile and the catalytic effectiveness factor are reported for the full range of Thiele modulus values, in the limit of large adsorption inhibition constant. In this limit, connections with negative first-order kinetics are shown, and rather good explicit estimates of the region of multiple steady states are provided. Comparisons with numerical calculations are also made.     

The transient response of a CSTR containing porous catalyst pellets: Asymptotic solutions

Transient response of a CSTR containing porous catalyst pellets is analyzed theoretically using a matched asymptotic expansion technique. This singular perturbation technique leads directly to the conditions under which the minima of reservoir concentration occur. The existence of the minima may be used to estimate some inherent parameters of the catalyst pellet.     

The measurement of catalyst acidity: I. Titration measurements

The n-butylamine titration method has been widely used for the measurement of acid strength distributions on solid catalysts. An implicit assumption with this technique is that both the n-butylamine titrant and the indicators are in adsorption equilibrium with acid sites on the surfaces. This assumption has been tested and found invalid. Serious doubt is therefore cast on the utility of such methods for the assessment of surface acidity. It is shown that instead of determining an acid site distribution, the total acidity of the sample is usually obtained. Having ascertained the problems involved, an alternative method has been devised for characterizing the site energy distribution. A series of bases of varying pK_a dissolved in one of several solvents was employed. The strong irreversible chemisorption of these yielded distributions quite different from those obtained by the n-butylamine technique.     

Kinetics,poisoning and mass transfer effects in liquid-phase hydrogenations of phenolic compounds over a palladium catalyst

The influence of poisoning and mass transport on the performance of a suspended palladium catalyst for the liquid-phase hydrogenation of o-cresol h for the various transport and reaction steps occuring in the three-phase slurry system and these were combined into an overall rate expression. From th analysed either by an experimental approach or by calculation.The hydrogenation in the absence of poison is described by a Langmuir—Hinshelwood adsorption model with adsorption of hydrogen and substrate on diffe present in the feed, the activity loss can be described by an extended model of the same type; experimental results show that the poison is very strong between poison and substrate on the remaining catalyst surface. The selectivity is not affected by increasing poison concentration.     

Studies of a CuY zeolite as a redox catalyst

The stability and behavior of CuY in the redox cycles with $\text{CO}\text{O}{}_2$, $\text{H}{}_2\text{O}{}_2$, and $\text{CO}\text{NO}$ have been studied using a microbalance operating in the flow mode and with a standard (BET) volumetric system. When CO was used as a reducing agent CO₂ was produced, thus removing oxygen from the zeolite lattice, but when H₂ was used only some of the H₂ consumed was evolved as H₂O. The rest was retained as lattice OH groups, but this was minimal when H₂ was used after treating the sample with CO. Oxidation with NO produced only N₂. At 500 °C the sample was stable and could be reversibly oxidized and reduced through many cycles using either $\text{CO}\text{O}{}_2$ or $\text{NO}\text{CO}$. In all cases the ratio $\text{O}\text{Cu}$ was close to 0.5, i.e., $\text{1e}\text{Cu}$. Treatment in CO at higher temperatures did not affect the reversible nature of the oxidation, but now the valence change was substantially larger; it approached $\text{2e}\text{Cu}$. The crystallinity of the exchanged zeolite was studied using X-ray diffraction and by measurement of the pore filling with liquid N₂. No significant changes could be detected after the different treatments, even those performed at 750 °C. Temperature-programmed reduction, temperature-programmed oxidation, and X-ray diffraction studies were made and the different maxima are reported. CuO and Cu ^o appeared in the oxidized and reduced samples, respectively, after treatment at 750 °C in CO but not at lower temperatures. Subsequent redox cycles at 500 °C did not appear to affect the size or amount of Cu ^o crystallites. CuY was active in the oxidation of CO with O₂ or NO. Its activity was lower than that of FeY zeolite when it exhibited an oxygen-carrying capacity of 0.5 $\text{O}\text{Cu}$. Treatment with CO at 750 °C, however, reversed the situation. Kinetic results showed that the fresh CuY catalyst was close to zero order in CO and fractional order in O₂ with an activation energy of 15 kcal/mole. After treatment at 750 °C in CO, the rate law became dependent upon the $\text{CO}\text{O}{}_2$ ratio. It was close to first order in CO and zero order in O₂ under oxidizing conditions ($\text{CO}\text{O}{}_2\text{≤ 2}$), but the orders were reversed under reducing conditions ($\text{CO}\text{O}{}_2\text{> 2}$). The activation energies were 12 and 15 kcal/mole, respectively. The data suggested that the Cu²⁺ with bound oxygen are the species active in the oxidation reaction.     

Carbon dioxide transport through enzymatically active synthetic membranes

The facilitated transport of CO₂ through thin liquid membranes, Millipore filter membranes, and cross-linked protein membranes has been investigated using a tracer ¹⁴CO₂ technique. Both the uncatalyzed reactions and the enzymatic reactions catalyzed by carbonic anhydrase in homogeneous solution and immobilized in several membrane configurations were studied. The steady-state transport data were reduced in terms of an analytical model for simultaneous reaction and diffusion which permitted the direct determination of the diffusional and enzyme kinetic parameters. The experimental method illustrates a powerful technique for measuring rapid reaction kinetics. In addition, a general multi-layer membrane model is developed which is capable of treating membrane kinetic heterogeneities. The multi-layer model yields a useful definition for the reaction boundary layer and provides a comparison among different membrane kinetic configurations for membrane design purposes.     

Forced cycling of the catalytic oxidation of CO over a V2O5 catalyst—I: concentration cycling

A study was performed on the behaviour of a catalytic reactor in which CO was oxidized over V₂O₅ under transient conditions using square-wave cycling of the reactant ratio. A maximum in the time-average reaction rate was observed at a period of about 20 min, resembling a resonance phenomenon. A minimum in the time-average reaction rate occurred in the range of periods between 40 and 60 min.     

Forced cycling of the catalytic oxidation of CO over a V2O5 catalyst—II Temperature cycling

An experimental study was made of the effects of temperature cycling on carbon monoxide oxidation over a vanadium pentoxide catalyst. Square-wave temperature cycles at periods above 5 hr provided reaction rates which, when time-averaged, were closely approximated by quasi-steady state reaction rates. As the period decreases, reaction rates fall off indicating that time-average rates under transient conditions are lower than those acheived at quasi-steady state. At a period of one hour, the time-average cycling rate falls below the rate at steady state at the mean temperature of the cycle used.     

The oxidative dehydrodimerization of propylene over BiSn oxide catalysts: III. Reduction and temperature-programmed reoxidation of the catalyst

A bismuth-tin oxide catalyst, which is active for the oxidative dehydrodimerization of propylene, was studied with the objective of characterizing its redox properties. The redox properties were investigated by kinetic methods and temperature-programmed reoxidation. The initial rate of reduction of the catalyst exhibited a first-order dependence on propylene partial pressure and an activation energy of 22 kcal/mol. The results of the temperature-programmed reoxidation investigation suggested a low-temperature reoxidation region and a high-temperature reoxidation region. The activation energy for the low-temperature reoxidation was 23 kcal/mol; the activation energy for the high-temperature reoxidation was 45 kcal/mol. Additional information regarding the physicochemical changes which occur in the catalyst in the two temperature regions was obtained from an examination of the catalyst by Auger and ESCA. The results obtained from these investigations suggested that the tin cations are more resistent to reduction than the bismuth cations. In addition, the low-temperature reoxidation appears to be associated with the transformation of Sn⁰ to Sn⁴⁺ and Bi⁰ to possibly an intermediate oxidation state. The high-temperature reoxidation appears to be associated with the full reoxidation of Bi to Bi³⁺. The redox properties and the physicochemical changes also correlated with the mechanism suggested for the oxidative dehydrodimerization of propylene in the earlier studies.