George R. Gavalas

Statistical and continuum models of fluid-solid reactions in porous media

In this review, we discuss past theoretical works on fluid-solid reactions in a porous medium. Such reactions are often accompanied by a continuous alteration of the pore structure of the medium, and at high conversions they exhibit percolation-type behavior, i.e. the solid matrix of the medium and/or the fluid phase lose their macroscopic connectivity. These phenomena are, therefore, characterized by a percolation threshold which is the volume or area fraction of a phase (solid or fluid) below which that phase exists only in isolated clusters or islands. Important classes of such processes are acid dissolution of a porous medium and gas—solid reactions with pore volume growth, e.g. coal gasification, and with pore closure, e.g. lime sulfation, and catalyst deactivation. These processes are characterized by continuous changes in the pore space as a result of a chemical reaction. We also consider here other processes such as the flow of fines, stable emulsions and solid particles in a porous medium which also alter the structure of the pore space, but by physical interaction of the particles and the solid surface of the pores. In this review we compare two different modelling approaches to reactions accompanied by structural changes. First we review the continuum approach, which is based on the classical equations of transport and reaction supplemented with constitutive equations describing the effect of structural changes on reaction and transport parameters. We then outline the relevant concepts, ideas and techniques of percolation theory and the statistical physics of disordered media, and review their application to the phenomena mentioned above. In particular, we emphasize the fundamental role of connectivity of the porous medium in such phenomena. Since in both approaches one needs to estimate the effective transport properties of the porous medium that is undergoing continuous change, we also review continuum and statistical methods of estimating the effective transport properties of disordered porous media.

Deposition of H2-permselective SiO2 films

Abstract Films of amorphous SiO 2 were deposited within the walls of porous Vycor tubes by SiH 4 oxidation in an opposing-reactants geometry: SiH 4 was passed inside the tube while O 2 was passed outside the tube. The two reactants diffused opposite to each other and reacted within a narrow front inside the tube wall to form a thin SiO 2 film. Once the pores were plugged the reactants could not reach each other and the reaction stopped. At 450°C and 0.1 and 0.33 atm of SiH 4 and O 2 , the reaction was complete within 15 min. The thickness of the SiO 2 film was estimated to be about 0.1 μ. Measurements of H 2 and N 2 permeation rates showed that the SiO 2 film was highly selective to H 2 permeation. The H 2 : N 2 flux at 450°C varied between 2000 and 3000. Thermal annealing at 600°C reduced somewhat that selectivity. Thermal annealing in the presence of H 2 O vapor decreased further the flux of H 2 and increased the flux of H 2 . Permeation of H 2 is believed to occur through an activated diffusion mechanism. Applications of such H 2 -permeable films to membrance reactors for equilibrium-limited reactions are discussed.

A New Algorithm for Automatic History Matching

History-matching problems, in which reservoir parameters arc to be estimated from well pressure data, are formulated as optimal control problems. The necessary conditions for optimality lead naturally to gradient optimization methods for determining the optimal parameter estimates. The key feature of the approach is that reservoir properties are considered as continuous functions of position rather than as uniform in a certain number of zones. The optimal control approach is illustrated on a hypothetical reservoir and on an actual Saudi Arabian reservoir, both characterized by single-phase flow. A significant saving in computing time over conventional constant-zone gradient optimization methods is demonstrated.

Structure and aging characteristics of H2-permselective SiO2-Vycor membranes

Hydrogen-permselective membranes were prepared by atmospheric pressure chemical vapor deposition (APCVD) of SiO_2 layers in porous Vycor tubes. The deposition was carried out by passing the reactants SiCl_4 and H_2O through the bore of the support tubes at temperatures ranging from 600 to 800°C. The deposit layers were examined by TEM, SEM, and EPMA. When the deposit was confined inside the pores of the Vycor substrate, the membranes were mechanically stable but when it extended substantially outside of the porous matrix the stresses induced by thermal cycling led to crack formation and propagation. Electron microscopy revealed that the SiO_2 deposit density is maximum in a region ≈0.5 μm thick adjacent to the bore surface and gradually declines to zero within a depth of ≈10 μm from the surface. The thin region of maximum deposit density is responsible for permselectivity, for it essentially blocks the permeation of nitrogen and larger molecules while allowing substantial permeation of hydrogen. This region contains ≈10% by volume trapped voids and as a result has relatively high permeability as suggested by the percolation theory. Annealing at high temperatures causes densification of the deposited material as evidenced by increased activation energy for H_2 permeation and correspondingly reduced permeance. The presence of H_2O vapor accelerates the densification process. The densified membranes had a H_2 permeance as high as 0.1 cm^3(STP)/min-atm-cm^2 at 500°C and a H_2/N_2 permeance ratio above 500.

Preparation of highly selective zeolite ZSM-5 membranes by a post-synthetic coking treatment

Zeolite ZSM-5 membranes with high n-butane:isobutane selectivities, e.g., 322 at 185°C, are obtained by a selective deposition of coke into non-zeolitic pores. The zeolite membranes are prepared by in situ crystallization on either bare porous α-Al_2O_3 support disks or disks that are pretreated to include a diffusion barrier. The post-synthetic coking treatment is accomplished by impregnating these membranes with liquid 1,3,5-triisopropylbenzene (TIPB) for 24 h at room temperature and then calcining them in air at 500°C for 2 h. Calcination at 500°C for up to 30 h does not destroy the high n-butane:isobutane selectivity. Thermogravimetric analysis (TGA) experiments on two model pore systems ZSM-5 (5.5 A) and Vycor glass (40–50 A) suggest that micro-defects are selectively eliminated by the TIPB coking treatment while the intracrystalline pore space of the ZSM-5 is not affected. The elimination of non-zeolitic pores results in a large increase of n-butane:isobutane pure gas flux ratio (45 vs. 320 at 185°C) accompanied by a fourfold reduction of the n-butane flux. The permeation experiments reveal that the n-butane flux increases nonlinearly with the partial pressure in the feed while the n-butane:isobutane pure gas flux ratio remains relatively unchanged.

Analysis of Char Combustion Including the Effect of Pore Enlargement

The combustion of char particles including the effects of external mass transfer, surface reaction, pore diffusion, and pore growth is analyzed theoretically. A random pore model is used to relate pore surface area, pore volume and effective diffusivity to the local burnoff. The equations for oxygen concentration and burnoff as functions of time and position are solved analytically for the limiting case of strong diffusional limitations (high temperature or particle size) and numerically for the general case. The results include the variation of particle size and density with burnoff under conditions pertinent to pulverized and fluidized combustion. The analysis is compared with previous analyses from the literature.

ZSM-5 membrane synthesis with organic-free mixtures

Preparation of supported ZSM-5 membranes using a TPA-free synthesis gel was investigated. After exploring the effect of reaction mixture the composition SiO_2:0.0125Al_2O_3:0.2675Na_2O:46H_2O was selected for membrane synthesis. Membranes were prepared by hydrothermal reaction on asymmetric α-Al_2O_3 tubular supports. The membranes were characterized by scanning electron microscopy, EDS, X-ray diffraction and Ar and N_2 adsorption. Permeation measurements with single gas and mixtures yielded selectivities for H_2 over n-butane above 104 and those for O2 over N_2 were 9–10. Permeation was strongly activated with the activation energies increasing sharply with molecular sizes.

Methane partial oxidation on Pt/CeO2–ZrO2 in the absence of gaseous oxygen

Partial oxidation of methane to synthesis gas over platinum or ruthenium supported on Ce_(1−x)Zr_xO_2 (x = 0, 0.2 and 0.5) was studied at 550–700°C in the absence of gaseous oxygen. The reaction was carried out in a packed-bed reactor under continuous or pulsed flows of methane. Oxidation utilized oxide oxygen and was initially very fast but slowed down as the oxide support became progressively reduced. Addition of ZrO_2 into CeO_2 considerably increased the rate of methane oxidation and enhanced the reducibility of CeO_2 but decreased the selectivity to carbon monoxide and hydrogen. Specifically it was found that significant production of carbon dioxide and water occurred on the freshly oxidized solid until a certain degree of reduction was reached beyond which the selectivity to carbon monoxide and hydrogen rose to over 90%. This critical degree of reduction was 10%, 40% and 65% for the solid compositions x = 0, 0.2 and 0.5, respectively. Additional experiments carried out using carbon monoxide pulses showed that carbon monoxide oxidation declines sharply and becomes negligible beyond this degree of reduction while oxidation of methane continues much further. Comparison of the two metals showed that platinum is more active but the reaction rate did not change much in the range of platinum loadings of 0.25–1 wt.%.

Structure and Activity of NiO/α-Al_2O_3 and NiO/ZrO_2 Calcined at High Temperatures I. Structure

A series of NiO/α-Al_2O_3 and NiO/ZrO_2 catalysts were prepared by impregnation and calcination in air between 750 and 1050 °C. Examination of the catalysts by means of nitrogen adsorption, H_2 chemisorption, X-ray diffraction, scanning electron microscopy, and X-ray photoelectron spectroscopy showed marked variations in physical and chemical properties as a function of calcination temperature. The results indicate that NiO does not interact with ZrO_2 whereas NiO-Al_2O_3 interaction is extensive at the interface between NiO and Al_2O_3 when the calcination temperature exceeds 850 °C. When the calcined samples of NiO/ZrO_2 were reduced in hydrogen, extensive redispersion of the metal was observed.

Preparation of supported carbon membranes from furfuryl alcohol by vapor deposition polymerization

Supported carbon membranes were prepared from furfuryl alcohol (FA) precursor by vapor deposition polymerization (VDP). For this purpose γ-Al_2O_3/α-Al_2O_3 or glass/α-Al_2O_3 support tubes were pretreated with an acid catalyst and exposed to FA vapors at 90°C. The tubes were then heated at 200°C to crosslink the poly(furfuryl alcohol) (PFFA) polymer and carbonized slowly to 600°C. The polymerization and carbonization cycle was repeated once to improve the permeation properties. The membranes were examined by scanning electron microscopy (SEM) and tested in a permeation cell with single gases (H_2, N_2, O_2, CO_2, CH_4) and with the mixture CO_2–CH_4. After the first polymerization/carbonization cycle the membranes had little selectivity for gas separations. After the second polymerization/carbonization cycle the membranes had ideal selectivities 10–13 for O_2:N_2, 80–90 for CO_2:CH_4, and 90–350 for H_2:N_2 at room temperature. The permeance was 0.6–2.5 for H_2, 0.27–0.58 for CO_2 and 0.08 for O_2, all in MPU (1 MPU=10^(−8) mol/m^2 Pa s). The permeances were sharply higher at 150°C but the selectivities were lower, e.g. one of the membranes had H2 permeance 10.6 MPU and H_2:N_2 ideal selectivity 30.