Search

Partial hydrodeoxygenation (HDO) of bio-oil derived from the pyrolysis of lignin can yield higher value added products, such as phenol derivatives. The HDO of 2-methoxyphenol (guaiacol) was studied over nickel phosphide (Ni2P) supported on various acidic substrates. The substrates were the microporous zeolite ZSM-5, a fluidized catalytic cracking (FCC) support consisting of USY zeolite embedded in a silica–alumina matrix, and an amorphous silica–alumina (ASA) material. Guaiacol was used as reactant because it is a good model compound for species derived from the lignin portion of biomass. The order of activity was Ni2P/ASA > Ni2P/FCC > Ni2P/ZSM-5. Contact time measurements indicated that the main pathway on Ni2P/ASA, the most active catalyst, was conversion of guaiacol to the primary intermediate catechol, which was then dehydroxylated to phenol.

A series of nickel phosphide catalysts supported on ultrastable Y zeolites USY (Si/Al = 40) with microporous/mesoporous structure were prepared by impregnation and temperature-programmed reduction and were studied for the hydrodeoxygenation (HDO) of 2-methyltetrahydrofuran (2-MTHF). The loading of the active phase was varied from 0.58, 1.16, and 1.74 to 2.3 mmol (g support)−1 and the corresponding samples were denoted as 0.5, 1.0, 1.5, and 2.0 Ni2P/USY. The two lowest loading supports did not show X-ray diffraction (XRD) lines, but X-ray absorption fine-structure spectroscopy (XAFS) indicated the formation of a Ni2P phase, with low Ni–Ni coordination, consistent with high dispersion. The two highest loading supports showed XRD patterns typical of Ni2P, and XAFS indicated similar bond distances to bulk Ni2P and high Ni–Ni coordination. Furthermore, the XAFS data indicated that the low-loading samples had shorter bond distances and more Ni in square-pyramidal coordination compared to the high loading samples and the reference Ni2P material, suggesting that there were differences in structural properties in the samples. This likely was due to preferred termination of the small crystallites with pyramidal Ni sites. The HDO of 2-MTHF was studied at 0.5 MPa and 513–593 K and the main products were n-pentane and n-butane for all catalysts. The low-loading samples showed higher turnover frequency (based on sites titrated by CO chemisorption), and this was attributed to the higher intrinsic activity of the pyramidal Ni sites. In addition, the low-loading samples showed higher selectivity to n-pentane, and this was attributed to lower C–C hydrogenolysis type reactions, which are favored by metallic ensembles as found in the high-loading samples.

Hybrid membranes are promising materials for the purification of natural gas from carbon dioxide. The present paper investigates the effect of the incorporation of primary and secondary amine functional groups on the performance of an organic–inorganic hybrid silica membrane for CO 2 /CH 4 separation. Hybrid membranes were synthesized by chemical vapor deposition using 3-aminopropyltrimethoxysilane and (3-methylaminopropyl)trimethoxysilane as primary and secondary alkylamine–silica precursors, respectively. The amino functionalized membranes were compared to an amine-free membrane prepared using propyltrimethoxysilane as precusor. The amine-free membrane had a pore size of 0.37 nm, and at 393 K a CO 2 permeance of 2.1×10 −8 mol m −2 s −1 Pa −1 and a CO 2 /CH 4 selectivity of 4. The primary amine membrane had a pore size of 0.36 nm, and at 393 K displayed a CO 2 permeance of 2.1×10 −8 mol m −2 s −1 Pa −1 and a CO 2 /CH 4 selectivity of 70. The secondary amine hmembrane had a pore size of 0.43 nm and achieved a CO 2 permeance of 1.3×10 −7 mol m −2 s −1 Pa −1 and a CO 2 /CH 4 selectivity of 140. The pore sizes were estimated by Tsuru׳s method. The transport mechanism of CO 2 throughout the amino-silica hybrid membranes was surface diffusion. The secondary amino-silica hybrid membrane was stable for 60 h under a relative humidity of 20%.

A one-pot redox deposition method is described to prepare silica-supported bimetallic NiPt and CuPt catalysts using differences in standard reduction potentials. The synthesis involved the initial reduction of non-noble metal ions (Ni2+ and Cu2+) by NaBH4, followed by reduction of Pt ions by these reduced metals. Alloy formation and phase purity were established by X-ray diffraction (XRD), bulk composition by X-ray fluorescence (XRF), and surface composition by X-ray photoelectron spectroscopy (XPS). Alloy formation was enhanced by increase in the amount of base metals available to reduce Pt, or through the use of excess NaBH4 after redox deposition. The catalysts were evaluated for methane oxidation at 300-400 °C and atmospheric pressure. Reference samples prepared using the incipient wetness impregnation (IW) method had inferior performance and characteristics. The redox deposition method provides a simple and versatile protocol for the preparation of bimetallic alloys between noble metals and base metals.

The hydrodeoxygenation of 2-methyltetrahydrofuran (2-MTHF) at a medium pressure of 0.5 MPa is studied over a Ni 2 P/SiO 2 catalyst. Contact time studies show the formation of n -pentanal, 1-pentanol, 2-pentanone, 2-pentanol, and 2-pentene, as reaction intermediates and the production of pentane and butane as major products. The results are consistent with adsorption of 2-MTHF followed by rate-determining ring-opening to form either 1-pentoxide or 2-pentoxide alkoxide intermediates. Subsequent hydrogen-transfer steps produce the various intermediates, a decarbonylation step of the pentanal forms n -butane and CO, and further hydrodeoxygenation steps result in n -pentane. Fitting of the results using a rake mechanism that considers adsorbed intermediates gives excellent agreement with experimental data, and agrees with a simulation with a simpler first-order model. The more detailed rake analysis indicates that the surface species from the 1-pentoxide intermediate are ten-fold more plentiful than those produced from the 2-pentoxide intermediate. In situ infrared measurements support this reaction mechanism.

Selective CO methanation over Ru/TiO2, Ru-Ni/TiO2, Ru-Co/TiO2, Ru-Fe/TiO2, Ru-La/TiO2, Ru-K/TiO2, and Ru-Ni-La/TiO2 was investigated as a CO removal method from reformate gas for polymer electrolyte fuel cell applications, and the selectivity and activity were examined. The addition of Co and La to Ru/TiO2 improved both CO and CO2 methanation, while the addition of Ni raised only the activity for CO methanation. The addition of La increased the electron density in the Ru species, which likely enhanced the dissociation of the C O bond of CO on Ru probably due to back donation of electrons from Ru to CO. This led to high CO methanation activity over Ru-La/TiO2. The trimetallic catalyst Ru-Ni-La/TiO2 showed the highest CO methanation activity among the prepared catalysts, resulting in a wider temperature range for selective CO methanation at low temperatures compared to Ru/TiO2.

Novel ultra-thin Pd and Pd-alloy (Pd–Ag, Pd–Cu) membranes, with a phosphorus component, were prepared on a porous hollow fiber α-alumina support. The aim was to study the effect of the presence of a P additive on the behavior of the membranes towards sulfur poisoning. To incorporate P in the structure, the membranes were exposed to PH 3 as a gaseous phosphorous source. The performance and sulfur resistance of the membranes with and without P was evaluated by measuring the hydrogen permeance and the H 2 /N 2 selectivity at 673 K before and during exposure to a gas mixture of 100 ppm H 2 S in H 2 , and also after regeneration in H 2 . The membranes were characterized by X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and scanning electron microscopy (SEM), both before exposure to H 2 S and after exposure/regeneration. XRD shows no detectable amounts of sulfur or phosphorus compounds in any of the exposed or P-treated membranes respectively. XPS indicates that S was located on the surface of the H 2 S/H 2 treated Pd and Pd–Ag membranes. No S was detected on the surface of the Pd–Cu membranes indicating that no irreversible S compounds were formed in the latter. P was also detected on the surface of the P-treated membranes. The presence of phosphorus lowered the hydrogen permeance and H 2 /N 2 selectivity; however it drastically improved the Pd-alloy membranes' capacity of recovery in H 2 after H 2 S poisoning. The Pd–Cu membrane recovered only part of its H 2 permeance after regeneration, going from 11.2×10 −7 mol m −2 s −1 Pa −1 (as-prepared) to 2.3×10 −7 mol m −2 s −1 Pa −1 , while in the presence of P the H 2 permeance varied from 5.2×10 −7 mol m −2 s −1 Pa −1 (as-prepared) to 6.5×10 −7 mol m −2 s −1 Pa −1 and its H 2 /N 2 selectivity increased from 30,000 to 40,000. Similarly, the Pd–Ag membrane recovered better in the presence of P going from 17×10 −7 mol m −2 s −1 Pa −1 (as-prepared) to 6.3×10 −7 mol m −2 s −1 Pa −1 in contrast to 52×10 −7 mol m −2 s −1 Pa −1 (as-prepared) to 5.9×10 −7 mol m −2 s −1 Pa −1 in the absence of P. SEM images show that in the absence of P, more cracks and pinholes are visible on the surface of the H 2 S/H 2 exposed membranes. The presence of P produces more homogeneous surfaces and less segregation, conferring a structural integrity to the membrane by suppressing the formation of cracks and pinholes.

Bio-oil obtained by the pyrolysis of woody biomass contains many oxygenated organic compounds which degrade the product quality and make necessary upgrading for its use as a liquid fuel. Hydrodeoxygenation (HDO) is a catalytic hydrotreating process for the removal of the problematic oxygen functionalities and is promising for bio-oil upgrading. In this work, 2-methyltetrahydrofuran (2-MTHF) was chosen as a model oxygenated compound, and its HDO reaction mechanism was studied on a silica-supported nickel phosphide catalyst (Ni 2 P/SiO 2 ) at a medium pressure of 0.5 MPa. The temperature dependency of the catalyst activity was determined and it was found that at 350 °C Ni 2 P/SiO 2 showed 100% conversion and 85% selectivity to n -pentane, with higher oxygen removal activity and less C C bond cracking activity than commercial noble metal Ru/C and Pd/Al 2 O 3 catalysts based on the same amount of active sites. A contact time study allowed the determination of a reaction sequence for 2-MTHF HDO on Ni 2 P/SiO 2 and it was found that C O bond cleavage of the furanic ring to generate either 2-pentanone or 1-pentanal was the rate-determining step. This was followed by hydrogen transfer steps to produce oxygen free compounds, n -pentane or n -butane. A partial pressure analysis of 2-MTHF and H 2 was consistent with a rate equation derived using a Langmuir–Hinshelwood (L–H) mechanism. This suggested that adsorption of 2-MTHF and hydrogen occurred competitively and that these species reacted on the Ni 2 P/SiO 2 surface. Although high partial pressure of H 2 was favorable for hydrogenation, too much H 2 competed with 2-MTHF adsorption, which caused lower conversion.

Photocatalytic water splitting is the cleanest method to produce H2, but often requires the removal of O2, which is simultaneously generated. To achieve this separation, use was made of supported liquid membranes formed from perfluorotributylamine (PFTBA) imbued in porous alumina supports. The PFTBA membranes were subject to single and mixed H2/O2 gas permeance tests at room temperature and showed a sieving transport behavior. The membranes had an average H2/O2 selectivity of 140 with a H2 permeability of 4500 barrers. The performance of the membrane showed only a slight dependence with increases of temperature and gas flow rate, but these increases reduced the lifetime of the membranes because of enhanced volatilization of the perfluoro compound. Studies showed that at small support pore sizes, the membrane displayed longer lifetime because of reduced vapor pressure of the liquid inside the porous support. A partial pressure analysis revealed that the selectivity improved when the partial pressure of the gases was reduced. Additionally, it was shown that separation at humid conditions reduced the stability of the membrane, but did not affect its permeance. The diffusivity and solubility of the gases across the membrane were obtained by the time-lag technique. It was found that H2 has higher solubility and diffusivity than O2 giving rise to a higher overall permeance for the smaller sized gas.

The removal of CO from reformate streams by selective CO methanation was investigated over TiO 2 supported Ru Ni bimetallic and monometallic catalysts. The combination of Ru and Ni enhanced CO methanation at low temperatures. The introduction of Ni into Ru/TiO 2 decreased the CO 2 conversion rate at 260 °C from 10 to 7.3 μmol min −1 g cat −1 . The use of Ru and Ni, thus, expands the temperature range of selective CO methanation. Transmission electron microscopy and temperature programmed reduction by H 2 confirmed that Ru species were in close proximity to Ni species on Ru Ni/TiO 2 , indicating a decrease in direct contact between Ru and TiO 2 . Fourier transform infrared spectroscopy techniques revealed that the decomposition of the formate species, formed during CO 2 methanation, is slow over Ru Ni/TiO 2 , in contrast to Ru/TiO 2 .

Hydrogen separation membranes are important in the gas separation field and among these, silica-based membranes have emerged as promising materials at high temperatures due to their high permeation rates, high selectivity, hydrothermal stability, resistance to poisons, and mechanical strength. A critical review of the preparation of silica membranes by chemical vapor deposition (CVD) is given, with special attention placed on the two major supports used, Vycor glass and porous alumina. The coverage includes the different gas transport mechanisms that occur through silica membranes, which are convective flow, Knudsen diffusion, molecular sieving, activated diffusion, and solid-state diffusion. A description is made of the two main CVD geometries, which are the “one-sided” feed and “opposing reactants” feed configurations. The results of numerous studies in which CVD precursors such as silicon alkoxides and organosilicon species, as well as non-silicon element compounds are used to control pore size and surface properties are presented. Also the effect of varying reaction conditions such as temperature, pressure, and reaction time are compared. Finally, the permeation properties of the silica membranes are compared to those of palladium membranes, both types of membranes are employed for hydrogen separation in membrane reactors, and the advantages of silica membranes in these reactor systems are discussed.

A general mixed mechanism model formulated within a grid framework is proposed. In particular, the framework is employed to account for the variability in the diffusion of carbon dioxide and methane in a certain class of inorganic–organic membranes. This variability is a basis for separation of gas mixtures into the various components. Various sizes of the cells within the grid are developed for a series of hybrid inorganic–organic membranes. Parameters in each model are estimated from the permeance results of both carbon dioxide and methane. Acceptable grid models are determined by comparing the estimated parameters of a surface diffusion mechanism with the values reported in the open literature. For each membrane, the model with the minimum number of grid cells is studied. The diffusion in the studied membranes is dominated by activated Knudsen at lower temperatures and limited by surface diffusion when the temperature is above a critical value. Activated Knudsen diffusion is the major mechanism that af...

A series of silica-supported metal phosphides was prepared by two methods involving the reduction of phosphite (I) or phosphate (A) precursors and was studied for the hydrodeoxygenation (HDO) of 2-methyltetrahydrofuran (2-MTHF). The I method required lower temperature than the A method and resulted in catalysts with higher surface area. The activity was evaluated in a packed-bed reactor on the basis of equal CO chemisorption sites (30 μmol) loaded in the reactor with comparison made to a commercial Pd/Al 2 O 3 catalyst. At 300 °C and 1 atm, the order of activity was Ni 2 P > WP > MoP > CoP > FeP > Pd/Al 2 O 3 . The principal HDO products for the iron group phosphides (Ni 2 P and CoP) were pentane and butane, whereas for the group 6 metal phosphides (MoP and WP), the products were mostly pentenes and pentadienes. For the Pd/Al 2 O 3 and the low-activity FeP/SiO 2 catalyst the products were mostly pentenes and C4 mixtures. There were no significant differences in the turnover frequency between materials prepared by the two methods, except possibly for the case of WP. There were likewise no great changes in selectivity toward HDO products at 5% total conversion, except for WP. The differences in the case of WP were attributed to the surface P/W ratio that X-ray photoelectron spectroscopy (XPS) showed to be twice as large for the I method than the A method. Contact-time studies were used to develop reaction networks for the most active catalysts, Ni 2 P/SiO 2 and WP/SiO 2 . For Ni 2 P/SiO 2 by both methods, the selectivity profiles were similar and could be explained by a rake mechanism with pentenes as primary products, 2-pentanone as a secondary product, and pentane as a final product. In contrast, for WP/SiO 2 , the selectivity depended greatly on the preparation method, but produced a preponderance of unsaturated compounds. The results could be explained from the surface composition.

A series of supported nickel–iron phosphide catalysts were used to study the hydrodeoxygenation (HDO) of 2-methyltetrahydrofuran (2-MTHF), a model compound for biomass-derived pyrolysis oil. The catalysts were prepared by incipient wetness impregnation of the active components onto potassium ion-exchanged USY zeolites followed by temperature-programmed reduction. The samples were denoted as Ni2P/KUSY, NiFeP(3:1)/KUSY, NiFeP(1:1)/KUSY, NiFeP(1:3)/KUSY, and FeP/KUSY, where the numbers in parenthesis are the Ni:Fe molar ratios. The results of the studies can be understood from ensemble and ligand effects. X-ray diffraction analysis indicated the presence of alloys in the mixed composition samples, and this was supported by CO chemisorption and infrared measurements. Uptakes of CO decreased as the iron content increased, suggesting that iron present on the surface of the catalysts blocked chemisorption sites. Fourier transform infrared (FTIR) measurements showed a single peak for linearly adsorbed CO whose intensity decreased with Fe content, in agreement with the uptake results. Moreover, the FTIR peak shifted monotonically consistent with a ligand effect in an alloy formed in the Ni2P and FeP particles. The reactivity in 2-MTHF HDO was studied at 250–325 °C and 0.5 MPa and it was found that the conversion was highest for the Ni2P/KUSY sample and decreased steadily with Fe content. However, the turnover frequency did not change significantly, indicating that the rate-determing step was activation of 2-MTHF on single Ni sites. The selectivity at low conversion (3 %) changed from mostly n-pentane and n-butane for Ni2P/KUSY to mostly 1-pentanol for the iron-containing samples, suggesting that in subsequent steps an ensemble effect was operational with the iron influencing the reaction chemistry. At high conversion (70 %) all the samples produced mostly n-pentane and n-butane.

The ethanol steam reforming (EtOH SR) reaction was carried out over a Co–Na/ZnO catalyst both in a packed bed reactor (PBR) and in membrane reactors (MRs) equipped with ultrathin Pd or Pd–Cu membranes to evaluate the benefits of employing membranes. For all conditions, ethanol conversion and hydrogen production were significantly higher in the MRs than in the PBR. Average ethanol conversion enhancement and hydrogen production enhancement were measured to be 12% and 11% in the Pd MR and 22% and 19% in the Pd–Cu MR, respectively. These enhancements of the conversion and product yield can be attributed to the shift in equilibrium by continuous hydrogen removal by the Pd based membranes. A significant contamination of the Pd layer by CO or carbon compounds during the reaction can be the reason for the comparatively low enhancement in the Pd MR compared to the Pd–Cu MR in spite of the high H 2 permeability of the original Pd membrane. A one-dimensional modeling study of the MRs and the PBR was conducted and their predicted results were compared to those obtained from the experimental study. The model was developed using a simplified power law expression but the predicted values fit the experimental data with only minor deviations. Enhancements of ethanol conversion and hydrogen yield with space velocity (SV) could be explained by the increased H 2 flux through the membranes with SV in the MRs.

Hybrid organic-inorganic H2-selective membranes consisting of single-layer or dual-layers of silica incorporating aromatic groups are deposited on a porous alumina support by chemical vapor deposition (CVD) in an inert atmosphere at high temperature. The single-layer silica membranes, which are made by the simultaneous decomposition of phenyltriethoxysilane (PTES) and tetraethylorthosilicate (TEOS), have good hydrothermal stability at high temperature and a high permeance for hydrogen in the order of 10−7 mol m−2 s−1 Pa−1 at 873 K, while preventing the passage of other larger molecular gases such as CH4 and CO2. The dual-layer silica membranes, which are obtained from the sequential decomposition of PTES and TEOS, exhibit an extremely high permeance for hydrogen of 3.6 × 10−6 mol m−2 s−1 Pa−1 at 873 K with a permselectivity of hydrogen over methane of 30. A normalized Knudsen based permeance method is applied to measure the pore size of PTES-derived silica membrane on the dual-layer silica membrane before...

The methane dry-reforming and steam reforming reactions were studied as a function of pressure (1–20 atm) at 973 K in conventional packed-bed reactors and a membrane reactors. For the dry-reforming reaction in a conventional reactor the production yield of hydrogen rose and then decreased with increasing pressure as a result of the reverse water-gas shift reaction in which the hydrogen reacted with the reactant CO2 to produce water. For the steam reforming reaction the production yield of hydrogen kept increasing with pressure because the forward water-gas shift reaction produced additional hydrogen by the reaction of CO with water. In the membrane reactors the methane conversion and the hydrogen production yields were higher for both the dry-reforming and steam reforming reactions, but for the dry reforming at high pressure half of the hydrogen was transformed into water. Thus, the dry-reforming reaction is not practical for producing hydrogen.

The effects of hydrogen permeance and selectivity in the performance of membrane reactors for the ethanol steam reforming (EtOHSR) were studied at 1–10 atm and 623 K. The studies were performed with Pd–Cu and SiO 2 –Al 2 O 3 composite membranes prepared by depositing the permselective components onto porous alumina supports with intermediate layers. The hydrogen permeances of the membranes varied from 5.2 × 10 −8 to 3.9 × 10 −6 mol m −2 s −1 Pa −1 and the selectivity ranged from 200 to 1000 (H 2 /CO 2 ) at 623 K, which allowed a broad set of conditions to be probed. Comparison studies with packed-bed reactor (PBR) and membrane reactor (MR) operation showed that higher ethanol conversions and hydrogen molar flows were obtained in the MRs for all pressures studied. It was determined that both hydrogen permeance and selectivity had a favorable effect on the EtOHSR reaction, with the highest ethanol conversion enhancement of 44% and hydrogen molar flow enhancement of 69% obtained in a MR fitted with a membrane with the highest hydrogen permeance. A criterion for the required permeance for industrial applications is developed, which for 1 cm diameter membrane tubes is 2.5 × 10 −7 mol m −2 s −1 Pa −1 .

Structural changes in Ni2P/MCM-41 were followed by quick extended X-ray absorption fine structure (QEXAFS) and were directly related to changes in X-ray absorption near-edge structure (XANES) which had been used earlier for the study of the active catalyst phase. An equation is proposed to correct the transient QEXAFS spectra up to second-order in time to remove spectral distortions induced by structural changes occurring during measurements. A good correlation between the corrected QEXAFS and the XANES spectral changes was found, giving support to the conclusions derived from the XANES in the previous work, namely that the formation of a Ni—S bond in a surface NiPS phase is involved in the active site for the hydrodesulfurization reaction.

Supported Ni2P catalysts are active for hydrodesulfurization (HDS) of petroleum feedstocks and have attracted considerable attention. In order to find out the structure of the active catalyst and to obtain insight into the reaction mechanism of thiophene HDS over Ni2P/MCM-41, we conducted a simultaneous time-resolved (STR) study with in situ X-ray absorption fine structure (XAFS) spectroscopy and Fourier transform infrared (FTIR) spectroscopy together with time-resolved product analysis. At 513 K, adsorbed tetrahydrothiophene (THT) was observed by FTIR, following the formation of a nickel phosphosulfide phase (NiPS). Detection of the HDS products coincided with the THT formation. All these experimental data indicated that NiPS worked as the active phase and identified THT as a reaction intermediate for the HDS reaction. The STR study proved to be a powerful tool for the precise study of reaction intermediates and the catalytic reaction path.

The permeation properties of hybrid inorganic–organic membrane with a silica crystalline framework are studied using quasi-non-equilibrium molecular dynamics. In particular the authors consider the single gas permeation of carbon dioxide and methane using the silica framework with phenyl groups inserted. The authors compare two types of membrane models, namely (a) single-pore with phenyl groups inserted (PSPM), and (b) single-pore without phenyl groups inserted (SPM) are developed. The pore diameter in the SPM model affects the performance of gas permeation. For the PSPM models the diameter significantly affects the selectivity; the optimum diameter is found to be 0.4 nm. This leads to a carbon dioxide permeance of 2.87 × 10−4 mol m−2 s−1 Pa−1, methane permeance of 1.66 × 10−5 mol m−2 s−1 Pa−1 and a selectivity of 17.3. Finally based on the pore size two permeation mechanisms are analyzed from the atomic level and the effect of the number of phenyl groups (phenyl group pairs) in one type of membrane model is discussed.

The hydrodesulfurization (HDS) of the refractory compound 4,6-dimethyldibenzothiophene (DMDBT) normally proceeds through a hydrogenation pathway that removes the planarity of the ring system and makes the hindered sulfur atom more accessible to the desulfurization centers. In this study, a highly active dispersed bimetallic NiFeP catalyst is found to have high selectivity for a direct desulfurization pathway, which does not require prior hydrogenation. The compound has equal numbers of Ni and Fe atoms which extended X-ray absorption fine structure analysis indicates are distributed randomly in the hexagonal Fe2P structure, with just a slight enrichment of Fe on the surface. This is supported by density functional theory calculations. The remarkable properties of the catalyst are ascribed to a ligand effect of Fe on the active Ni atoms.

Nickel phosphide (Ni 2 P) is an emerging catalyst for hydrodesulfurization and other important environment- and energy-related catalytic reactions. To understand its high performance, the surface structure of a Ni 2 P (0 0 0 1) single crystal surface was investigated using dynamical LEED analysis. The obtained surface structure for Ni 2 P (0 0 0 1)-1 × 1 is a P-covered Ni 3 P 2 structure (Ni 3 P_P structure) as opposed to the expected bulk terminated surface structures. This Letter discusses the driving force for the formation of the Ni 3 P_P surface, which involves the minimization of dangling bonds.

Thin and defect-free Pd–Cu composite membranes with high hydrogen permeances and selectivities were prepared by electroless plating of palladium and copper on porous alumina supports with pore sizes of 5 and 100 nm coated with intermediate layers. The intermediate layers on the 100 nm supports were prepared by the deposition of boehmite sols of different particle sizes, and provided a graded, uniform substrate for the formation of defect-free, ultra-thin palladium composite layers. The dependence of hydrogen flux on pressure difference was studied to understand the dominant mechanism of hydrogen transport through a Pd–Cu composite membrane plated on an alumina support with a pore size of 5 nm. The order in hydrogen pressure was 0.98, and indicated that bulk diffusion through the Pd–Cu layer was fast and the overall process was limited by external mass-transfer or a surface process. Scanning electron microscopy (SEM) images of the Pd–Cu composite membrane showed a uniform substrate created after depositing one intermediate layer on top of the alumina support and a dense Pd–Cu composite layer with no visible defects. Cross-sectional views of the membrane showed that the Pd–Cu composite layer had a top layer thickness of 160 nm (0.16 μm), which is much thinner than previously reported.

This review describes palladium and palladium alloy membranes for hydrogen separation prepared by different fabrication methods and using different membrane supports. Several correlations of structure and function for those membranes are provided based on mechanistic considerations of permeance along with structural properties and membrane morphologies. Particular attraction is placed in analysis of the hydrogen permeance and selectivity of membranes reported in recent papers. Composite palladium membranes prepared by the electroless plating technique deposited on alumina substrates are found to be the most promising for practical applications. It is concluded that the prospects for the use of palladium membranes in industrial applications are improving due to extensive research addressing current problems such as durability, hydrogen embrittlement, fouling by hydrocarbons or hydrosulfide compounds, and the high cost of palladium.

Gold nanoparticles on titanium oxide and titania-silica supports are active for the formation of propylene oxide by the oxidation of propylene with hydrogen and oxygen mixtures. Cyanide treatment of gold supported on titanosilicate zeolite supports (Au/TS-1) yielded unexpected results. Catalysts treated with weak solutions of sodium cyanide resulted in preferential removal of small gold particles, while catalysts treated with strong solutions resulted in dissolution of the gold and re-precipitation as gold (+1) cyanide. X-ray absorption spectroscopy demonstrated that catalysts that produce propylene oxide in the presence of hydrogen and oxygen mixtures had supported gold (+3) oxide nanoparticles of 3 nm size after synthesis, which were reduced to gold metal at reaction conditions. Samples treated with strong solutions of sodium cyanide resulted in supported gold (+1) cyanide particles of large size, 9–11 nm. These particles did not produce propylene oxide but, surprisingly, showed high selectivity toward propylene hydrogenation. Increasing gold (+1) cyanide particle size resulted in a decrease in hydrogenation activity.

Ni2P/K-USY (K-promoted ultrastable Y-type zeolite) is an efficient catalyst for hydrotreatment of petroleum feedstocks. The structural changes of the passivated catalyst during activation were foll...

Ultra-thin Pd composite membranes with a thickness of 1 μm were prepared by a novel electric-field assisted activation technique followed by electroless deposition of Pd on a hollow-fiber α-alumina support. The novel activation method places Pd precursors and a reducing agent on opposite sides of a porous substrate and uses an electric field to cause migration of Pd ions to the outer surface where they are reduced to form seeds in high density in a narrow spatial region. The resulting membranes showed a high hydrogen permeance in the range of 4.0–5.0 × 10{sup −6} mol m{sup −2} s{sup −1} Pa{sup −1} and stable H{sub 2}/N{sub 2} selectivity of 3000–9000 during stability tests for 150 h at 733 K with H{sub 2} flow. The formation of the thin, defect-less and robust Pd layer can be ascribed to the evenly distributed Pd seeds on the support layer and the enhanced bonding between the Pd layer and the support layer resulting from the strong anchoring of the Pd seeds onto the support in the new activation step.

The major mechanisms of gas permeation through solid membranes are described and the applicable equations describing the permeance are presented. The mechanisms depend on the relative size of the permeating molecules and the diameter of the pores. As pore size decreases the operable mechanisms are Hagen-Pouiselle flow, Knudsen diffusion, surface diffusion, gas-translation, and finally solid-state diffusion. The Hagen-Pouiselle mechanism involves flow through large pores, while the Knudsen mechanism involves collision of molecules with the walls of pores of intermediate size. Surface diffusion deals with movement of molecules trapped in the potential field of the walls of pores of relatively small size, while gas-translation involves molecules that can escape the field, but are constrained by the small pores. Finally, solid-state transport comprises dissolution and transport by diffusion within the solid. These mechanisms are illustrated for hydrogen permeance with the use of two membranes, an alumina membrane with intermediate sized pores and a silica on alumina membrane of dense structure.

The catalytic properties of gold catalysts leached with sodium cyanide solutions were investigated for the propylene epoxidation reaction with H 2 /O 2 mixtures. Unexpectedly, the selectivity of the Au/TS-1 catalysts changed from propylene oxide to propane as a result of the cyanide treatment. Measurements by in situ X-ray absorption spectroscopy at in situ reaction conditions demonstrated that the active catalyst before leaching consisted of gold in the zero-valent state, while after the cyanide treatment had gold in the Au + state. Examination of the catalysts by high resolution transmission electron microscopy revealed the formation of precipitates of a material with lattice spacings corresponding to Na[Au(CN) 2 ]. It is found that leaching with cyanide does not result in positively-charged gold ions substituting in cation vacancies, as previously thought, but instead produces a gold(I) cyanide precipitate with catalytic hydrogenation properties typical of ptatinum. These results explain previously reported findings on highly dispersed gold nanoparticles, in which the peripheral atoms are positively charged.

The surface structure of Ni2P (1010), a model for highly active hydrodesulfurization catalysts, was studied using scanning tunneling microscopy (STM) and low energy electron diffraction (LEED) in order to understand the reconstruction of the surface layers. Annealing at 573 K revealed a (1 × 1) LEED pattern which changed to a c(2 × 4) arrangement by further heating to 723 K. Atomic scale STM images were obtained for both the (1 × 1) and c(2 × 4) structures. Bright spots observed in the STM images were interpreted to be due to surface phosphorus atoms and this was supported by a density functional theory (DFT) simulation. Several possible models for the c(2 × 4) reconstructed structures were proposed including a P-dimer defect model, a missing-row model and a missing-row + added-row model. The last model gave the best explanation for the c(2 × 4)structure. The mechanism for the c(2 × 4) reconstruction on the Ni2P (1010) surface is discussed.

The steam reforming of ethanol over Na–Co/ZnO catalysts was investigated in a conventional packed-bed reactor (PBR) and in a membrane reactor (MR) fitted with silica–alumina composite membranes. Promotion with a moderate amount of Na (0.2–1.0 wt%) produced catalysts with stable ethanol conversion and product selectivity. At 523 K, the main products were H2 (51%) and CH3CHO (42%) with minor amounts of CH4 (0.7%) while above 573 K the main products were H2 (70%) and CO2 (15%) with a small amount of CH4 (2.4%). Higher cobalt loading, higher W:E ratio, higher reaction temperature and lower space velocity enhanced the conversion of ethanol to H2 and CO2 while reducing the formation of undesirable acetaldehyde. Contact time studies indicated that acetaldehyde was a primary product of ethanol reforming, while CO2 and CH4 were secondary products. The composite membranes were prepared by a chemical vapor deposition (CVD) method and had moderate H2 permeances ((5.2–6.8) × 10−8 mol m−2 s−1 Pa−1 at 623 K). The MR produced positive yield enhancements of the secondary products, including hydrogen, but yield reductions for the primary product, acetaldehyde, and this is explained from the effect of the membrane separation on the reaction network. Two membranes with similar permeances but different H2 selectivity ratios (H2/CH4 = 60 and 350) were compared and it was found that the membrane with a higher H2/CH4 selectivity ratio gave a higher yield enhancement. A survey of membrane reactor studies indicated that the conversion enhancement in reforming reactions could be described by a parameter denoted the operability level coefficient (OLC) which is related to 1/DaPe.

The hydrodesulfurization (HDS) of thiophene was studied over Ni 2 P/MCM-41 by in situ Fourier transform infrared (FTIR) spectroscopy and X-ray absorption near-edge structure (XANES) measurements at atmospheric pressure and 393–723 K (120–450 °C). FTIR showed the presence of adsorbed molecular thiophene principally on the MCM-41 support. In the presence of hydrogen new bands in the aliphatic C–H stretching region (2953, 2918, and 2876 cm −1 ) and the CH 2 scissoring region (1464 cm −1 ) appeared and grew with temperature up to 553 K (280 °C). These were assigned to a tetrahydrothiophene surface intermediate, whose presence is consistent with the high hydrogenation activity of nickel phosphide. XANES measurements between 353 and 573 K (180–300 °C) also showed growth of a signal (8330.9 eV) attributed to Ni–S interactions of either the adsorbed tetrahydrothiophene or a surface phosphosulfide. Transient XANES measurements showed that the sulfur species responded dynamically at the same rate as the steady-state rate, indicating that it was associated with an intermediate in the reaction.

A hydrothermally stable and H2-selective composite membrane was successfully prepared on a mesoporous alumina support by thermal decomposition of titanium isopropoxide (TIP) and tetraethylorthosilicate (TEOS) at high temperature. The membrane with a H2-selective silica–titania top layer of 10–20 nm in thickness had a H2 permeance 3 × 10−7 mol m−2 s−1 Pa−1 and a H2 selectivity over CH4 and CO2 in the range of 40–60. After 130 h exposure in 75 mol% H2O at 923 K, the H2 permeance was reduced by only 30%, compared to a pure silica membrane which suffered a loss of 90%. The use of the opposing reactants technique improved the quality of the composite membrane. Although considerable more efforts are needed to increase surface area, the stability properties suggest applicability of the composition at high temperature in the presence of high concentrations of steam.

The epoxidation of propylene with H 2 –O 2 mixtures was studied over a gold catalyst supported on a Ge-modified TS-1 catalyst at 170 °C. Shifts of X-ray diffraction lines to lower angle showed that the Ge was incorporated into the structure of the TS-1. The presence of Ge more than doubled the conversion from 1.6% to 4% while increasing selectivity from 87% to 91%. Density functional theory calculations indicated that the origin of the enhanced activity was a ligand effect of Ge on Ti, which lowered the activation energy for the critical epoxidation step.

A series of Au/titanium silicalite-1 (TS-1) catalysts with different Si/Ti ratios and promoted with alkali and alkaline earth cations were prepared by deposition–precipitation (DP) and tested for direct propylene epoxidation. It was found that the gold loading and catalytic activity was highly dependent on the pH of the DP synthesis solution and the final composition of the catalyst. Addition of Group 1 metals such as K or Cs had little effect on the gold content, but increased activity, while Group 2 metals such as Mg, Ca, Sr, and Ba increased both the gold content and the catalytic activity. The highest improvement was provided by a Mg promoted catalyst, which at 443 K and 0.1 MPa with a H 2 /O 2 /C 3 H 6 /Ar = 1/1/1/7 feed mixture gave a propylene oxide (PO) formation rate of 88 gPO h −1 kg cat −1 , compared to 57 gPO h −1 kg cat −1 for an unpromoted catalyst, corresponding to a 50% enhancement of activity. Ammonia temperature-programmed desorption (NH 3 -TPD) measurements indicated little change in adsorption amount with promotion indicating that the yield increase was not due to the elimination of acidic sites on the catalyst. Instead, the improved catalytic performance was ascribed to increased Au capture efficiency and dispersion by the catalyst. The effect of Si/Ti ratio, pH of synthesis, and the promoter ions on the gold content could be understood from their effect on the surface charge of the support.

An operability level coefficient ( OLC ), defined as the ratio of product permeation and product formation rates, and related to the inverse combination of the Damkohler number and the Peclet number (1/ DaPe ), is suggested as a useful tool for estimating performances of membrane reactors (MRs) operating as separators in equilibrium-limited reactions. The OLC s for product hydrogen formation in previously reported MRs for methane dry-reforming (MDR), methane steam-reforming (MSR), methanol steam-reforming (MeOHSR), and ethanol steam-reforming (EtOHSR) were correlated with conversion and yield enhancements. For values of OLC s ranging from 0.03 to 0.78, a clear universal trend for increasing conversions and hydrogen yields with increasing OLC was observed for these different types of reforming reactions. The OLC curve calculated from a numerical simulation without adjustable parameters was found to closely approximate experimental data obtained from the MRs, and was shown not to depend on the assumed kinetics. This study confirms that hydrogen selectivity (from the ratio of single-gas permeances) has a substantial influence on conversion and hydrogen yield enhancements in a MR, and demonstrates that a hydrogen selectivity of 100 is sufficient to achieve high performance in a MR.

The diminishing quality of oil feedstocks coupled with increasingly more stringent environmental regulations limiting the content of sulfur in transportation fuels have given rise to a need for improved hydroprocessing technology. This review begins with a summary of the major improvements in hydrodesulfurization (HDS) and hydrodenitrogenation (HDN) catalysts and processes that have been reported in recent years. It then describes a new class of hydroprocessing catalysts, the transition metal phosphides, which have emerged as a promising group of high-activity, stable catalysts. The phosphides have physical properties resembling ceramics, so are strong and hard, yet retain electronic and magnetic properties similar to metals. Their crystal structures are based on trigonal prisms, yet they do not form layered structures like the sulfides. They display excellent performance in HDS and HDN, with the most active phosphide, Ni2P, having activity surpassing that of promoted sulfides on the basis of sites titrated by chemisorption (CO for the phosphides, O2 for the sulfides). In the HDS of difficult heteroaromatics like 4,6-dimethyldibenzothiophene Ni2P operates by the hydrogenation pathway, while in the HDN of substituted nitrogen compounds like 2-methylpiperidine it carries out nucleophilic substitution. The active sites for hydrogenation in Ni2P have a square pyramidal geometry, while those for direct hydrodesulfurization have a tetrahedral geometry. Overall, Ni2P is a promising catalyst for deep HDS in the presence of nitrogen and aromatic compounds.

A kinetic method is described to determine the role of adsorbed intermediates in heterogeneous reaction mechanisms. The method, denoted as analysis of coverage transients (ACT), involves comparing the time response of a spectroscopically observed species in an inert gas and a reactive gas to differentiate between adsorption–desorption processes and reaction. The method is applied in a kinetic study of pyridine hydrodenitrogenation (HDN) on a 12.2 wt% silica-supported nickel phosphide (Ni2P/SiO2) catalyst at 423 K and atmospheric pressure. In situ Fourier transform infrared spectroscopy (FTIR) measurements of pyridine adsorbed on Ni2P revealed the formation of a pyridinium surface intermediate. The concentration of the pyridinium intermediate increased with pyridine partial pressure and decreased in the presence of hydrogen, suggesting that it was a reaction intermediate. However, transient and steady-state kinetic measurements showed that the rate of reaction of the intermediate did not correspond to the overall reaction rate, and it is concluded that the pyridinium intermediate is not directly involved in the HDN reaction of pyridine over Ni2P. The studies demonstrate that mere observation of an adsorbed surface species at reaction conditions is not sufficient to prove that it is a reaction intermediate. The ACT method has potential as it can be used with any type of spectroscopy, as long as the surface coverage can be calibrated.

The surface structures of nickel phosphide (Ni2P) single crystals were studied by scanning tunneling microscopy (STM) and photoemission electron microscopy (PEEM). Atomically resolved 1 x 1 images of the Ni2P(0001) and (1010) surfaces are successfully obtained with STM, whose respective dimensions of (0.59 nm x 0.59 nm) and (0.34 nm x 0.59 nm) match the unit cell lengths. The Ni2P(0001) surface has two possible terminations in which the Ni:P ratios are 3:1 (Ni3P termination) or 3:2 (Ni3P2 termination). In the Ni3P terminated surface the Ni atoms have a square pyramidal structure, and in the Ni3P2 terminated surface the Ni atoms have a tetrahedral structure. Only the P is visible in the STM for Ni2P(0001) and this is explained as being due to the greater extension of the phosphorus p orbitals than the Ni d orbitals at the surface. The surface domain sizes of the Ni3P and Ni3P2 termination structures of Ni2P(0001) are determined to be 500 microm by means of PEEM using a Hg-Xe lamp through a low-pass UV-filter. Evidence is found for dissociative adsorbed hydrogen on the Ni3P termination surface of Ni2P(0001), but not on the Ni3P2 surface, indicating that the square pyramidal sites on the Ni3P surface have high reactivity.

Ni 2 P catalysts supported on SiO 2 and MCM-41 were prepared by temperature-programmed reduction (TPR), and the effect of the dispersion on catalyst structure and hydroprocessing performance was studied. The surface areas of the samples varied from low (Ni 2 P/SiO 2 -L, 88 m 2 g −1 ) to high (Ni 2 P/SiO 2 -H, 240 m 2 g −1 ), to very high (Ni 2 P/MCM-41, 487 m 2 g −1 ), with corresponding Ni 2 P average crystallite sizes decreasing from 10.1 to 6.5 and 3.8 nm. X-ray diffraction (XRD) and extended X-ray absorption fine structure (EXAFS) studies were used to obtain structural parameters for the supported Ni 2 P phase. The catalytic activity in hydrodesulfurization (HDS) was measured at 613 K and 3.1 MPa in a three-phase fixed bed reactor using a model liquid feed containing 4,6-dimethyldibenzothiophene (4,6-DMDBT) and quinoline in a tridecane solvent. At standard conditions using 500 ppm S as 4,6-DMDBT, 6000 ppm S as dimethyldisulfide (DMDS), and 500 ppm N as quinoline, the catalytic activity followed the sequence Ni 2 P/MCM-41 > Ni 2 P/SiO 2 -H > Ni 2 P/SiO 2 -L, based on equal sites (230 μmol) loaded in the reactor. In particular, Ni 2 P/MCM-41 gave an HDS conversion of 90%, which was much higher than that of a commercial Ni–Mo–S/Al 2 O 3 catalyst which gave an HDS conversion of 68%, based on equal number of sites (230 μmol) loaded in the reactor. The sites were counted by CO chemisorption for the phosphide and by low-temperature O 2 chemisorption for the sulfide. EXAFS analysis of the samples confirmed the presence of two types of sites, tetrahedral Ni(1) sites and square pyramidal Ni(2) sites, with the latter growing in number in the same order as the reactivity Ni 2 P/MCM-41 > Ni 2 P/SiO 2 -H > Ni 2 P/SiO 2 -L, as the crystallite size decreased. From the selectivity to the direct desulfurization (DDS) product (dimethylbiphenyl) and the hydrogenation (HYD) products (methylcyclohexyltoluenes and dimethylbicyclohexyls) it is concluded that the Ni(1) sites are responsible for DDS while the Ni(2) are highly active sites for the HYD route.

The selective oxidation of propane to acetone and 2-propanol with H 2 and O 2 was studied on Au/TS-1 by kinetic and spectroscopic analysis. A kinetic study using a factorial design at conditions where the catalyst was stable and gave propane conversions of 90%, resulted in a power-rate law expression of the form r oxyg. = k oxyg. (H 2 ) 0.74 (O 2 ) 0.36 (C 3 H 8 ) 0.29 . In situ Au L 3 -edge X-ray absorption near-edge spectroscopy (XANES) measurements showed activation of O 2 on Au, whereas in situ ultraviolet–visible (UV–vis) spectroscopy evidenced the presence of Ti-hydroperoxo species. The role of the Ti-hydroperoxo species was probed by a transient technique in which changes in Ti K -edge XANES spectra were used to determine the evolution of coverage with time ( d θ / d t ). It was shown that the rate of reaction by XANES ( 6.7 × 10 −4 s −1 ) was close to the turnover rate measured in a catalytic flow reactor ( 5.6 × 10 −4 s −1 ), indicating that the hydroperoxo species were true intermediates in the reaction. A proposed reaction sequence in which H 2 O 2 forms on Au sites and propane is partially oxidized on Ti centers accounts for the spectroscopic results and the reaction orders obtained experimentally for the power-rate law expression.

The vapor-phase epoxidation of propylene with H2 and O2 over a highly dispersed Au/TS-1 catalyst was carried out in a packed-bed catalytic membrane reactor. The membrane reactor allowed the use of high concentrations of H2 and O2 (40% each) that were well within the explosive regime, and the production rate of propylene oxide was improved by 100–200% compared with that from a conventional packed-bed reactor. At 180 °C, the C3H6 conversion was 10% and the PO selectivity was 80%, corresponding to a space–time yield of 150gPOkgcat−1h−1, whereas at 212 °C, the yield increased to 200gPOkgcat−1h−1. The kinetic dependency, rPO=krPO=k (H2)0.53 (O2)0.26 (C3H6)0.18, obtained free from mass transport and heat transfer limitations at these high concentrations was in good agreement with that reported at low concentrations.

Propane epoxidation was carried out by sequential propane dehydrogenation–propylene epoxidation steps using a two-catalyst bed and H2 and O2 as the oxidant mixture. The propane dehydrogenation step used a Au/TiO2 catalyst that was active at the low temperature (443 K) used for the propylene epoxidation step; the latter used a Au/TS-1 catalyst. In situ Au L3-edge X-ray absorption near-edge structure and ultraviolet–visible measurements on Au/TiO2 under propane dehydrogenation conditions showed activation of oxygen on gold nanoparticles and evidence for the formation of adsorbed oxygen intermediate species responsible for the production of propylene. Propane epoxidation with H2 and O2 at 443 K and 0.1 MPa with the dual Au/TiO2 and Au/TS-1 catalysts resulted in an overall propane conversion of 2%, propylene selectivity of 57%, and propylene oxide selectivity of 8%, corresponding to a propylene oxide space-time yield of 4 g kg cat −1 h −1 . The catalysts showed little deactivation and maintained their conversion and selectivity levels for the 12 h duration of the measurements.

A thin layer (30–40 nm) of a dual-element silica–alumina composition was deposited on a porous alumina support by chemical vapor deposition (CVD) in an inert atmosphere at high temperature. Prior to CVD, an intermediate layer of γ-alumina was coated on the macroporous alumina support. The intermediate layer was prepared by the dip-coating and calcination of boehmite sols of different sizes to give a graded structure that was substantially free of defects. The resulting supported composite membrane had high permeance for hydrogen in the order of 2–3 × 10−7 mol m−2 s−1 Pa−1 at 873 K with selectivities of H2 over CH4, CO and CO2 of 940, 700 and 590, respectively. The membrane operated by a hopping mechanism involving jumps of permeating molecules between solubility sites. The presence of aluminum improved the hydrothermal stability of the membranes for periods in excess of 500 h at 873 K in 16% steam, allowing the permeance to remain above 10−7 mol m−2 s−1 Pa−1, although with decreased selectivities.

In situ ultraviolet−visible (UV−vis) diffuse reflectance spectroscopy was used in combination with in situ Ti K-edge X-ray absorption near-edge structure (XANES) to study the formation of Ti-hydroperoxo species during the gas-phase epoxidation of propylene with H2 and O2 at reaction conditions over a Au−Ba/Ti−SiO2 (Ti−TUD) catalyst. The in situ UV−vis measurements showed growth of a signal due to Ti-hydroperoxo species when the catalyst was put in contact with H2/O2/Ar (1/1/8) and C3H6/H2/O2/Ar (1/1/1/7) gas mixtures at 423 K and 0.1 MPa. Changes in the area of the pre-edge peak centered at 4968.9 eV present in the Ti K-edge XANES spectra of the catalyst were used to estimate the Ti-hydroperoxo species coverages (θ) under operating conditions. Transient Ti K-edge XANES experiments with H2/O2/Ar (1/1/8) and C3H6/H2/O2/Ar (1/1/1/7) gas mixtures allowed the estimation of the net epoxidation rate by a novel method involving the determination of dθ/dt. It is shown that the Ti-hydroperoxo species are true inter...

A novel membrane is described consisting of an ultrathin 20–30 nm permselective layer of silica deposited on an intermediate multilayer γ-alumina substrate with a graded structure. The alumina substrate is formed on top of a macroporous alumina support by sequentially dipping and calcining a series of dilute solutions containing boehmite (AlOOH) sols of different particle sizes. The size of the sol particles is tuned by precisely controlling synthesis parameters including acid type, acid concentration and hydrolysis time. The topmost silica layer is deposited on top of the intermediate alumina layer by chemical vapor deposition of a silica precursor, tetraethylorthosilicate, in an inert atmosphere. Cross-sectional images of the membranes obtained from scanning electron microscopy (SEM) show that the intermediate alumina layers are formed from particles of progressively smaller size so as to form a smooth interface between the rough support and the topmost amorphous silica layer. The resulting silica-on-alumina composite membrane had high permeance of 5.0 × 10 −7 mol m −2 s −1 Pa −1 and good selectivities for hydrogen over CH 4 , CO and CO 2 of over 1500 at 873 K.

The gas-phase epoxidation of propylene was studied over MoO x /SiO2 catalysts in a reaction system with a post-catalytic bed volume. In the reaction of a mixture of propylene and propane with oxygen below 578 K, propylene oxide (PO) was mainly formed from the oxidation of propylene. It was found that the oxidation reaction was very sensitive to the temperature of the post-catalytic space more than the temperature of the catalyst bed, strongly indicating that radical reactions occurring in the post-catalytic bed free space were responsible for the PO formation. The addition of NO increased propylene conversions and PO selectivity at low conversions, confirming that radical reactions were involved in the propylene reactions.

The kinetics of propylene oxidation to propylene oxide (PO) with H 2 /O 2 mixtures on gold supported on the mesoporous titanium silicate, Ti-TUD, was investigated using Langmuir-Hinshelwood (L-H) models and power-rate law (PRL) models. The catalyst gave stable activity and was appropriate for the kinetic studies, giving high selectivity to PO (>95%) at low conversions of propylene (