John B. Butt

Internal mass and heat transport effects on catalytic behavior—activity, selectivity, and yield in a complex reaction system

Abstract The effects of internal mass and heat diffusion on catalytic activity and on the selectivity and yield of intermediate product in a consecutive reaction system are investigated. It is shown that, over a wide range of conditions, the selectivity property is annihilated and there is a net consumption rather than production of intermediate. In other regions, selectivity is greatly decreased and the yields of intermediate are small. For consecutive reactions in which the activation energy of the first reaction is greater than that of the second, however, improved selectivity is possible and the yield of intermediate may surpass that realized in the absence of internal gradients. Expressions for estimation of activity over a large range of thermal effects are given, and a convenient numerical technique for solution of the nonlinear equations involved in description of the system is discussed.

An experimental study of the effect of intraparticle temperature gradients on catalytic activity

Abstract A study has been made of temperature profiles incident within porous catalyst particles in the course of a highly exothermic chemical reaction, the hydrogenation of benzene. A newly developed experimental technique which permits catalyst pellets to be formed around fine wire thermocouples has been employed. Using thermocouple wires 0·001 in. in diameter it has been possible to measure temperatures simultaneously at four radial positions in 1 2 -in. cylindrical pellets formed by compressing a nickel on kieselguhr catalyst powder. Pellets were pressed at three density levels, 1·71, 1·85 and 1·95 g/cm 3 and the reaction experiments carried out with a mole fraction of hydrogen in the entering mixture to the reactor of 0·83. Temperature of reactants entering the reactor varied from 40 to 133°C. For runs at the lower temperature levels, the internal temperature profile was found to increase monotonically from the surface to a center line maximum along the radius of the cylinder. Although centerline measurements were somewhat distorted by axial conduction effects at higher temperatures, it was possible to obtain consistent interpretation based on measurements of internal temperature near the surface. It was also observed that the slope of the internal temperature profile passed through a maximum as surface temperature was increased. This observation is consistent with the kinetic behavior of the system: the rate of reaction, as reported by others, passes through a temperature maximum somewhere between 150 and 220°C. The results of this work indicate that the uncertainty surrounding the temperature at which maximum rate occurs may well be due to the existence of diffusional limitations in previously reported kinetic studies. An analysis of the internal temperature profiles was achieved using an approach similar to that outlined by Weisz and Hicks, employing a simple pore structure model based on the bimodal volume-area distribution of the catalyst pellet to evaluate transport parameters. At a given pellet density it was found that the theory accurately predicted temperature profiles over the range of surface temperatures below 150°C. The analysis requires a value of activation energy be assumed for computations and values corresponding to the range reported in the literature, 6500 to 11,700 cal/g-mole, were employed. The best agreement between theory and experiment was obtained with a value of 6500 cal/g-mole, corresponding to reported activation energies from experimentation with nickel film catalysts. It was not possible to apply the results obtained with catalyst at one density level to the prediction of effects at a second density level. It is felt that this is a result of the inadequacies of the pore structure model employed for estimation of transport properties.

Monte carlo simulation of a catalytic surface: Activity and selectivity of γ-alumina for dehydration

Abstract The acidic properties and catalytic activity of oxide surfaces have been shown in a number of cases to be directly influenced by their degree of hydration. In the present work the rates and selectivities of ethanol dehydration on a well characterized gamma alumina have been measured experimentally over the temperature range of 180 to 400 °C with feeds containing 0, 10, 20, and 40 mole % water. The effects of surface hydration and feed composition are compared with the predictions of a Monte Carlo simulation based on a model of the alumina surface and the mechanism of the dehydration reaction. The simulation is in good agreement with experiment, and points clearly to the distinctions which can be made between the separate effects of reaction energetics and surface configurations on rate and selectivity in the reaction.

Analysis of nonselective poisoning and its influence on reactor behavior

Abstract Three general cases of nonselective catalyst poisoning are analyzed in terms of their effects on catalyst life and reactor operation. The term ‘nonselective’ is used to indicate a linear variation of catalyst activity with the fraction of active surface (or sites) available. A fourth type of deactivation mechanism considered is a ‘elective,’ sintering process in which catalyst activity is proportional to surface area. It is shown that these types of poisoning are amenable to a generalized treatment in terms of variables defined from properties of the reaction system (rate constants and activation energies) and of the conditions of reactor operation. Catalyst deactivation and reactor behavior are reported as reactor temperature versus time of operation at constant conversion. An analysis of two types of non-isothermal operating policies is also given, and it is shown that control of reactor temperature profiles such that uniform active surface is maintained through the bed can, in some instances, lead to greatly increased catalyst life. The conditions pertaining to such improvement are discussed.

Diffusional mass transport of nonadsorbed gases within porous structures: I. Experimentation and correlation

Abstract Counterdiffusion flux and flux ratio have been measured for the argon-helium system in four different, well-characterized porous structures. Pressures from 1 to 14.6 atm and temperatures from 0 ° to 69 °C were employed in the investigation. Care was taken to avoid the effects of extraneous factors in experimental work, and to ensure the absence of transport by surface diffusion. Variation of flux with pressure, temperature, and porosity is discussed with regard to the equations defining diffusion in transition range capillaries.

The structure and activity of chromium oxide catalysts: I. Structure analysis

Abstract Two forms of activated chromia catalysts are recognized: (1) an amorphous form, catalyst A, prepared by heating under vacuum or in an inert atmosphere, and (2) a microcrystalline form, catalyst C, prepared by heating in a reducing atmosphere. The structural properties of these catalysts have been studied by X-ray line-broadening, radial distribution analysis, electron microscopy, infrared spectroscopy, and low temperature nitrogen adsorption. Catalyst C, prepared by heating in hydrogen, is composed of crystallites of α-chromium oxide supported on amorphous chromia. Detailed analysis of the X-ray diffraction powder pattern indicates little contribution to the linewidth from strain. The shape of the crystallites (characterized by X-ray line-broadening of the (104), (110), and (116) diffraction peaks) is oblong. The dimension along the α-axes is roughly twice that along the c-axis. The electron micrographs are consistent with this picture. Attempts to prepare crystallites with dimensions between 15 and 70 A were unsuccessful. Typically a catalyst activated at 400 °C has an effective crystallite size of 80 × 90 A while a 450 °C activated catalyst has an effective crystallite size of 110 × 200 A. The percentage crystalline phase estimated from infrared absorption in the lattice vibration region in catalysts activated for 12 hr at a temperature between 400–450 °C are in the range 5–30%. Catalyst A, prepared by heating in helium, does not give rise to a discernible X-ray diffraction pattern, but does result in a well defined radial distribution curve. Analysis of the radial distribution curve indicates that there exists local order over approximately a 10 A dimension of the same kind as found in α-Cr2O3. A model for this local order is proposed.

The structure and activity of chromium oxide catalysts: II. Influence of catalyst structure on activity for reactions of cyclopropane

Abstract The initial rates for cyclopropane isomerization and hydrogenolysis on amorphous (catalyst A) and crystalline (catalyst C) chromium oxide are reported. Based on the overall specific rates or turnover numbers, cyclopropane hydrogenolysis is sensitive and isomerization insensitive to the catalyst structure. The activation energies for both reactions are greater on catalyst C than on A, but a strong compensation effect is operative. The changes in activity which accompany catalyst structural changes do not appear to be closely related to changes in pore size, site density or mechanism, but are interpreted in terms of the influence of catalyst structure on site activity.

Multilayer nitrogen adsorption in porous structures

Abstract The results of volume-area distribution calculations are sensitive functions of the correlation of adsorbed multilayer thickness employed in the computation. Disagreement between correlations proposed appears due primarily to the fact that the effects of pore structure on the adsorption are not included. The present work presents a correlation of multilayer nitrogen adsorption in porous structures which is based on reported isotherms for a wide range of pore systems. The consistency of surface area results obtained from computations employing the correlation is demonstrated.

Diffusional mass transport of nonadsorbed gases within porous structures: II. Structural significance of the convergent-divergent pore model

Abstract Experimental results of counterdiffusion studies are correlated with the convergent-divergent pore structure model proposed previously. Relationship of the parameters of the model to porous structure properties is discussed; it is shown that these quantities may be determined from the volume-area distribution of a material and that the model may thus be employed for prediction of porous structure effects on diffusional mass transport.