A.J. Burggraaf

Importance of the surface exchange kinetics as rate limiting step in oxygen permeation through mixed-conducting oxides

Attention is drawn to the possible involvement of the surface exchange kinetics in limiting the rate of oxygen permeation through mixed-conducting oxide ceramics. A theoretical approach is provided with which it is possible to distinguish between surface exchange- and bulk diffusion controlled kinetics of oxygen permeation. New results on the oxygen permeability of perovskites La0.8Sr0.2CoO3−σ and SrCo0.8Fe0.2O3−σ are presented. The importance of the exchange reaction re to diffusion in limiting overall oxygen transport through (La,Sr)(Co,Fe)O3−σ perovskite-type oxides is emphasized.

Effect of impurities on sintering and conductivity of yttria-stabilized zirconia

The effect of low concentrations of Fe2O3, Al2O3 and Bi2O3 on the sintering behaviour of (ZrO2)0.83 (YO1.5)0.17, made by alkoxide synthesis, has been investigated. The best results are achieved with Bi2O3 as a sinter agent and a relative density of 95% is obtained at 1200 K. The effects of these impurities on the electrical conductivity of the bulk and the grain boundaries has been investigated using frequency dispersion analysis (101-106 Hz). All investigated impurities have a negative influence on both the bulk and grain-boundary conductivity. For Fe2O3 and Al2O3 grain-boundary segregation factors of about two are calculated.

Analysis and theory of gas transport in microporous sol-gel derived ceramic membranes

Sol-gel modification of mesoporous alumina membranes is a very successful technique to improve gas separation performance. Due to the formed microporous top layer, the membranes show activated transport and molecular sieve-like separation factors. This paper concentrates on the mechanism of activated transport (also often referred to as micropore diffusion or molecular sieving). Based on a theoretical analysis, results from permeation and separation experiments with H2, CO2, O2, N2, CH4 and iso-C4H10 on microporous sol-gel modified supported ceramic membranes are integrated with sorption data. Gas permeation through these membranes is activated, and for defect-free membranes the activation energies are in the order of 13?15 kJ.mol?1 and 5?6 kJ.mol?1 for H2 and CO2 respectively. Representative permeation values are in the order of 6×10?7 mol.m?2.s?1.Pa?1 and 20×10?7 mol.m?2.s?1.Pa?1 for H2 at 25°C and 200°C, respectively. Separation factors for H2/CH4 and H2/iso-butane are in the order of 30 and 200 at 200°C, respectively, for high quality membranes. Processes which strongly determine gas transport through microporous materials are sorption and micropore diffusion. Consequently, the activation energy for permeation is an apparent one, consisting of a contribution from the isosteric heat of adsorption and the activation energy for micropore diffusion. An extensive model is given to analyse these contributions. For the experimental conditions studied, the analysis of the gas transport mechanism shows that interface processes are not rate determining. The calculated activation energies for micropore diffusion are 21 kJ.mol?1 and 32 kJ.mol?1 for H2 and CO2, respectively. Comparison with zeolite diffusion data shows that these activation energies are higher than for zeolite 4A (dpore=4A), indicating that the average pore size of the sol-gel derived membranes is probably smaller.

Formation and characterization of supported microporous ceramic membranes prepared by sol-gel modification techniques

The formation is described of supported microporous membranes (by IUPAC definition rpore < 1 nm), prepared by modification of mesoporous γ-alumina membranes with polymeric sols. The mesoporous γ-alumina membranes, with a top-layer thickness in the order of 7–10 μm, and with pore radii of 2–2.5 nm, have a very high surface finish (mean roughness 40 nm). The amorphous microporous top-layer thickness is in the order of 60–100 nm. Gas transport properties are effectively improved as is shown by activated permeation and molecular sieve-like separation factors in the order of 50–200 for H2/CH4. These microporous top-layers can be prepared from a relatively wide range of sol structures; from inorganic oligomers which are too small to result in significant scattering with SAXS, to polymeric structures with fractal dimensions in the range: 1

Transport properties of alkanes through ceramic thin zeolite MFI membranes

Polycrystalline randomly oriented defect free zeolite layers on porous -Al2O3 supports are prepared with a thickness of less than 5 μm by in situ crystallisation of silicalite-1. The flux of alkanes is a function of the sorption and intracrystalline diffusion. In mixtures of strongly and weakly adsorbing gases and a high loadings of the strongly adsorbing molecule in the zeolite poze, the flux of the weakly adsorbing molecule is suppressed by the sorption and the mobility of the strongly adsorbing molecule resulting in pore-blocking effects. The separation of these mixtures is mainly based on the sorption and completely different from the permselectivity. At low loadings of the strongly adsorbing molecules the separation is based on the sorption and the diffusion and is the same as the permselectivity. Separation factors for the isomers of butane (n-butane/isobutane) and hexane (hexane/2,2-dimethylbutane) are respectively high (10) and very high (> 2000) at 200°C. These high separation factors are a strong evidence that the membrane shows selectivity by size-exclusion and that transport in pores larger than the zeolite MFI pores (possible defects, etc) can be neglected.

Gas transport and separation with ceramic membranes. Part II. Synthesis and separation properties of microporous membranes

Non-supported microporous silica (amorphous) and titania thin films were made by the polymeric gel route. The titania system consisted of particles smaller than 5 nm. Reproducible modification of supported γ-alumina films with silica demands a strict control of every modification step. Silica films of 30–60 nm thickness on top of and presumably partly inside the γ-alumina film were realised. The permeabilities of helium and hydrogen through this film are activated, while the propylene permeability was below the detection limit. Separation factors of a H2---C3H6 mixture are larger than 200 at 200 °C with a flux of the preferentially hydrogen of 1.6 × 10−6 mol/m2-sec-Pa. The pores must be of molecular dimensions to realise this (< 1 nm diameter). Preliminary research shows that changes in the synthesis parameters result in higher activation energies and improved separation properties. The relation between synthesis, resulting microstructure and gas separation properties, however, is not yet fully understood.

High oxygen ion conduction in sintered oxides of the

The phase diagram of the Bi2O3-Er2O3 system was investigated. A monophasic f c c structure was stabilized for samples containing 17.5–45.5 mol% Er2O3. Above and below this concentration range polyphasic regions appear. The f c c phase showed high oxygen ion conduction. The ionic transference number is equal to one for specimens containing 30 mol% Er2O3 or less, while an electronic component is introduced at low temperatures for specimens containing 40–60 mol% Er2O3. Between 673 K and 873 K a maximum in the conductivity was found at 20 mol% Er2O3. (Bi2O3)0.8.(Er2O3)0.20 is found to be the best oxygen ion conductor so far known. The conductivity at 773 K and 973 K is 2.3 Ω−1m−1 and 37 Ω−1 m−1 respectively. These values are 2–3 times higher than the best oxygen ion conductor reported for substituted Bi2O3 systems and 50–100 times higher than those of stabilized zirconia (ZrO2)0.915(Y2O3)0.085 at corresponding temperatures.

Bi_2O_3-Er_2O_3

The transport of pure gases and of binary gas mixtures through a microporous composite membrane is discussed. The membrane consists of an alumina support with a mean pore diameter of 160 nm and an alumina top (separation) layer with pores of 2-4 nm. The theory of Knudsen diffusion, laminar flow and surface diffusion is used to describe the transport mechanisms. It appears for the composite membrane that Knudsen diffusion occurs in the toplayer and combined Knudsen diffusion/laminar flow in the support at pressure levels lower than 200 kPa. For the inert gas mixture H2/N2 separation factors near 3 could be achieved which is 80% of the theoretical Knudsen separation factor. This value is shown to be the product of the separation factor of the support (1.9) and of the top layer (1.5). The value for the top layer is rather low due to the relatively small pressure drop across this layer. This situation can be improved by using composite membranes consisting of three or more layers resulting in a larger pressure drop across the separation layer. CO2 surface diffusion occurs on the internal surface of the investigated alumina membranes. At 250-300 K and a pressure of 100 kPa the contribution of surface diffusion flow measured by counterdiffusion is of the same order of magnitude as that resulting from gas diffusion. The adsorption energy amounts —25 kJ/mol and the surface coverage is 20% of a monolayer at 293 K and 100 kPa. The calculated surface diffusion coefficient is estimated to be 2-5 x 10-9 m2/sec. Modification of the internal pore surface with MgO increases the amount of adsorbed CO2 by 50-100%. Modifications with finely dispersed silver are performed to achieve O2 surface diffusion.

system

A single-phase monoclinic zirconia (the thermodynamically stable modification up to a temperature of 1170°C), having a specific surface area of 67 m2g?1 and a well-developed mesoporous texture, has been prepared by gel-precipitation followed by calcination at 450°C. A commercially available high-surface area monoclinic zirconia powder (SBET=71 m2g?1) has also been studied. It was found that the specific surface area and pore volume of monoclinic zirconia both decreased markedly on increasing the calcination temperature; despite the fact that the crystal structure was that of the stable modification, this did not seem to impart any substantial resistance to thermal sintering. The thermal stability of monoclinic zirconia could however be improved significantly by addition (by an impregnation technique) of various oxides: CaO, Y2O3, La2O3 all led to an improvement in the thermal stability up to 900°C while MgO exhibited stabilizing properties only up to 700°C; the best results were obtained with La2O3. All the additives investigated other than MgO were found to bring about a partial transition of the monoclinic to a fluorite-like phase of zirconia upon heat treatment; this phase has been shown in the case of the CaO-doped sample to be cubic zirconia and in the cases of the Y2O3- and La2O3-doped samples to be tetragonal zirconia. As little as 20?50% of a theoretical monolayer quantity of La2O3 was sufficient to give satisfactory thermal stability. The results can be explained by a model involving mass transport by a surface diffusion mechanism.

Gas separation mechanisms in microporous modified γ-al2o3 membranes

Theoretical expressions for single gas permeation are analysed and evaluated with selected literature and some new experimental data on Silicalite/ZSM-5 membranes. The phenomenological sorption–diffusion (PSD) description (and its equivalent Maxwell–Stefan equation) covers both the microscopic models based on configurational diffusion (CD) with δ=1.0 and on surface diffusion (SD) with δ≥1.24. The ratio δ of the pore diameter over the (kinetic) molecular diameter is important. For 1.24≤δ Ar>Kr>Xe at given values of δ close to unity is predicted by the SEMP model.