Search

Gas-liquid processing in microreactors remains mostly restricted to the laboratory scale dueto the complexity and expenditure needed for an adequate numbering-up with a uniform flowdistribution. The barrier-based distributor is a multiphase flow distributor which assures flowuniformity and prevents channeling between the two phases.Experimentally, design criteria for the barrier-based distributor are obtained in a setupmade of capillary and tube fittings. The design criteria are illustrated for the case of anitrogen-water Taylor flow in four parallel microchannels. Gas-liquid channeling is preventedat equal pressures in the gas and liquid manifolds. An optimal operational window is realizedwhen the gas to liquid flow ratio is kept constant and the ratio of the maximum over the minimumflow rates remain less than 15. The effect of variations in the inner diameters due to thefabrication tolerance of the barrier channels and the microchannels on the flow distribution isdemonstrated.Numerically, the flow distribution is studied using a method based on hydraulic resistivenetworks (RN). The single phase hydraulic RN model is extended to account for two phasesgas-liquid Taylor flow. For ReGL

In this thesis the performance and applicability of rotating solid foam stirrers is investigated. The stirrer consists, thereby of a solid, highly porous structure, which is used as stirrer and catalyst support simultaneously. The solid foam block occupies a large part of the reactor volume. Catalyst loadings of 1-2 wt% can be achieved. The rotation of the foam block drives the liquid outwards through the solid foam pores. Baffles mix gas and liquid and direct the fluid flow back into the reactor center. The thesis follows chronologically the steps of mass transfer in a three-phase reaction. First the external mass transfer processes, the gas-liquid mass transfer and the liquid-solid mass transfer, are studied. Then the gas-holdup in the reactor is studied using g -ray tomography for various liquids and stirrer conformations. The effects on the external mass transfer is discussed. Then the catalyst coating is described before a three-phase reaction is used to test the reactor performance. For comparison a slurry reactor equipped with a Rushton stirrer is used. The mass transfer and reaction rates are always related to to the power input, as these are the main criteria for the choice of a reactor for a given process. Chapter 1 gives a brief introduction on mass and heat transfer processes in three-phase reactors. Options for determination of crucial steps limiting the reaction rate are discussed. Subsequently, some widely used three-phase reactor types are discussed, emphasizing fixed bed reactors. This includes random and structured packings, trickle beds and packed bubble columns and rotating packed bed reactors. The chapter ends with an outline of the thesis. In Chapter 2 the gas-liquid mass transfer is discussed. This mass transfer step is often limiting three-phase reaction rates due to the low solubility of gaseous compounds in liquids. A fluorescence probe is used to study the absorption/desorption rates of the oxygen-water system. Gas-liquid mass transfer data are presented for different solid foam stirrer configurations, mainly in a blade and a block design, and compared to a six-blade Rushton stirrer. Both foam reactor designs work at stirring rates below 600 rpm. Using the foam blade design, gas bubbles are mainly created by the turbulence at the gas-liquid interface. Large bubbles are broken up by collisions with the foam blades. Using a foam block design, rotation leads to the structuring of the reactor volume into sections strongly differing in gas holdup, flow behavior and bubble size distribution. Gas-liquid mass transfer coefficients have been achieved, which are 50 % higher than measured for a Rushton stirrer used as reference. The foam block structure occupies a larger reactor volume than the foam blade stirrer. In the following chapters we focus on this stirrer type. Chapter 3 covers the liquid-solid mass transfer. The rotating foam block stirrer is compared to a slurry reactor equipped with a Rushton stirrer using the dissolution of copper as model reaction. For the stirred reactor consisting of a Rushton stirrer and a slurry with small particles, the particle density has a large influence on the liquid-solid mass transfer rate. Heavy particles show high slip velocities as their path is less influenced by the liquid turbulence. Furthermore, the mass transfer is enhanced by neighboring particles passing each other. In case of solid copper particles, a kLS-value of 3·10-3 ms-1 was observed. For particles having a density comparable to industrially catalyst supports, the kLS-value is one order of magnitude lower. For the rotating foam block stirrer also high kLS-values of 2.5 ·10-3 ms-1 have been achieved due to the high liquid velocity along the foam struts. The influence of several parameters on the hydrodynamics and the liquid-solid mass transfer has been studied. Among these the foam block height, the foam pore size, and the space above the foam block stirrer have a large effect on the liquid circulation, the bubble formation and therefore on the liquid-solid mass transfer. The reduction of the foam pore size has two counterbalancing effects resulting only in a slight increase of the rate of mass transfer: i) the liquid-solid interfacial area is increased; ii) the liquid circulation is, however, reduced due to the higher frictional pressure drop. Gas bubbles passing the foam struts affect the liquid-solid mass transfer by inducing liquid velocity fluctuations. For the foam block stirrer a maximum kLSaLS-value of 0.6 s-1 was obtained. An additional advantage of the foam stirrer reactor is that the solid phase is fixed and the catalyst does not need to be separated downstream the reactor. In Chapter 4 the gas holdup of rotating foam block reactors is studied using g-tomography. The influence of liquid properties, such as viscosity and surface tension, and the rotational speed on the gas/liquid distribution in the different reactor sections is investigated. Especially the viscosity has a strong effect on the entrapment of gas bubbles in the foam block structure, while the surface tension is the dominant parameter in the outer reactor section. The influence of these parameters on the inset of foaming and the collapse of the gas/liquid dispersion is investigated. For non-foaming Newtonian liquids the two phase flow through the foam block stirrer is mainly influenced by the solid foam pore size and the liquid viscosity. For low viscosity, the optimal foam block pore size was identified in the range between 10 and 20 ppi. With smaller pore size, the gas-holdup inside the foam block strongly increases due to bubble entrapment. For higher viscosity, pore sizes larger than 10 ppi have to be used in order to achieve a sufficient liquid flow rate through the foam block to avoid a strong gradient over the reactor height. The effect of the hydrodynamics on the gas-liquid and liquid-solid mass transfer and the reactor performance are discussed and recommendations for further optimizations of the reactor design and the operational conditions depending on the liquid properties have been developed. Chapter 5 discusses the coating of solid foam structures. Highly stable alumina layers have been prepared on aluminum foams by a combination of anodization and wash coating. The structure of the coating layers has been investigated using electron microscopy and X-ray diffraction. Nitrogen adsorption and mercury-porosimetry have been applied to determine the pore size distribution and specific surface area. Mechanical stability was tested using ultrasonic vibration. Anodizing conditions used in this work, produce a highly stable oxide coating with a specific surface area up to 2.5 m2g-1 foam. These layers can be used directly as low surface area catalyst support, but also as sub layers in order to increase the mechanical stability of an applied wash coat. For the wash coat, alumina particles in micrometer size range have been used. By adjusting the alumina content in the wash coat slurry between 20 and 40 wt%, homogeneous alumina coatings have been produced with a thickness between 5 and 25 µm, a porosity of around 50 % and a specific surface area up to 24 m2g-1 foam. The interparticle pores in the micrometer size range provide the space for a high mass transfer of gas and/or liquid reactants to the catalytic surface. The addition of colloidal alumina particles with a particle size of less than 200 nm increases the coating stability. Additional use of an anodization layer has resulted in a mass loss of less than 2 wt%. In Chapter 6 the catalytically active solid foam blocks are applied in a three-phase model reaction, the oxidation of glucose on platinum promoted by bismuth. Using catalyst nanoparticles and measuring the oxygen concentration in the bulk liquid, the kinetic parameters and mass transfer characteristics have been determined at a temperature of 333 K. The overall reaction rate has been studied experimentally using three different support types: slurry catalysts, pellets and a solid foam stirrer. The glucose conversion rate and the deactivation rate of the catalyst depend strongly on the ratio between mass transfer and reaction rate. At low catalyst concentrations the glucose oxidation process is liquid-solid mass transfer limited. The block stirrer shows a superior performance over the slurry catalysts, due to the high liquid-solid volumetric mass transfer coefficient. The bimodal pore size distribution of the catalyst layer further increases the conversion rate. Using slurry catalysts, a loading of 1 wt% in combination with pure oxygen feed is required to achieve acceptable conversion rates. Under these conditions the gas-liquid mass transfer and partially the liquid-solid mass transfer are the rate limiting steps. The foam block stirrer shows good gas-liquid mass transfer rates and high liquid-solid mass transfer rates, which still increase at high power input. Working under external mass transfer control and using this stirrer type, the catalyst loading can be strongly reduced to loadings less than 0.4 wt%, while the conversion rate remains comparable to slurry particles with loadings of 1 wt%. As the catalyst is fixed, attrition and agglomeration in high viscosity liquids are circumvented. There is further no need for filtration and the catalyst can simply be re-used. Chapter 7 summarizes the main conclusions concerning the gas-liquid and liquid-solid mass transfer processes. The influence of various parameters on the individual mass transfer processes and the overall process is discussed. Finally recommendations concerning further optimizations are given. These include the reactor design, but also the development of new porous materials for rotating packed bed reactors.

Increasing energy demand in the world and the environmental concerns due to air pollution and the greenhouse effect of fossil fuel resources stimulate the research into renewable alternative energy sources. Among these alternatives, hydrogen has a life changing opportunity for the human being for the future. The chemical energy stored in hydrogen can be converted into electricity via various types of fuel cells. The proton exchange membrane fuel cell is the most promising type providing high power densities at operating temperatures typically less than 100 °C. The necessity of increasing the platinum utilization in PEMFCs, due to its cost and availability, requires an optimization of the catalyst support and the gas diffusion layers. In many long term fuel cell tests, a decrease in water removal capacity leads to a gradual decrease of the PEMFC performance. The use of more graphitic and highly oriented carbon as an alternative material for use in the gas diffusion media combines a high electronic conductivity with an acceptable surface area. One approach, introduced in this thesis, is the direct growth of carbon nanofibers (CNFs) on a gas diffusion layer providing a strong network with a high surface area. Two methods are developed for the direct growth of CNFs on a carbon based gas diffusion layer: CNF growth via nickel complex particles (chapter 2) and via homogeneous deposition precipitation of nickel (chapter 3). The production of CNF grown carbon paper via the nickel complex particles was rather fast compared to the homogenous deposition precipitation of nickel, since deposition of nickel complex particles on a carbon paper took less than one hour whereas it took more than two days for the homogeneous deposition precipitation of nickel. The time of the deposition of nickel hydroxide in the deposition precipitation depends upon the decomposition rate of urea at the deposition temperature. Since the deposition was carried out near the boiling point of water, the decomposition rate of urea was nearly constant. It was attempted to increase the total amount of nickel by working with highly concentrated nickel and urea solutions. However, this caused a non-uniform deposition of nickel hydroxide which resulted in the detachment of CNFs during ultrasonic treatments. The loading of the nickel catalyst was controlled precisely during homogeneous deposition precipitation of nickel on carbon paper in this slow deposition process. In addition, the growth of nickel hydroxide layers covered the entire surface of the carbon paper and thereby closed the pores bigger than 1 µm. In our experiments, it was observed that open pores on the CNF grown carbon paper surface caused instabilities in operation due to water accumulation at these locations of the cathode compartment of the PEMFCs. Considering the two main factors described here, homogeneous deposition precipitation of nickel was utilized to obtain a highly controlled nickel deposition and thereby a highly controlled CNF loading on the carbon papers. The decoration of carbon paper with CNFs was investigated as a water management layer (chapters 4 and 5) and as a direct catalyst layer for platinum (chapter 6) in order to give more accessibility for the gas transport. The CNF grown carbon paper provided effective gas diffusion to the catalyst layer when they were used as a water management layer, especially under high gas flows. It suffered from performance losses at the mid-current density region (400-800 mA.cm-2), which can be improved by loading of a hydrophobic polymer. The CNF grown carbon paper can be a good alternative as a catalyst support, however, further optimization of the Nafion-platinum contact is required to get the benefits of using CNF grown carbon paper as a direct catalyst layer.

The research carried out within this thesis studied the hydrodynamics, reaction applications and scale-up of liquid-liquid microreactors. The liquid-liquid flow patterns in microchannels were evaluated in terms of stability, surface-to-volume ratio, achieved throughput and extraction efficiency. The research focused on the optimal flow patterns, the slug and bubbly flow. A liquid-liquid slug flow pressure drop model was developed and validated on examples of water-toluene and water/ethylene glycol-toluene slug flow. The fluidic control over the interfaces in a slug flow microreactor was employed to study a complex system of liquid-liquid phase transfer catalyzed alkylation of phenylacetonitrile. Last, a novel microstructured redispersion reactor was developed and tested on the example of phase transfer catalyzed esterification, allowing for significant increases in yield, process safety and waste reduction compared to the conventional process.

Propene oxide is a very important intermediate for the synthesis of commercial products, including adhesives, paints, and cosmetics. The gas-phase epoxidation of propene over Au/Ti-based catalysts is an intriguing scientific topic, not only because of the industrial importance of the production of propene oxide, but also because this reaction offers fundamental insights into the working principles of Au-based catalysis. The current methods used for propene oxide synthesis combine several reaction steps and have economic or environmental problems. Au/Ti-based catalysts are able to produce propene oxide with high selectivity out of propene, hydrogen and oxygen in a simple process since a single reactor is needed. However, a fundamental understanding of the reaction mechanism is necessary for developing more efficient Au/Ti-based propene epoxidation catalysts. The aim of this PhD thesis was to shed further insight into the mechanism of Au/Ti-based catalysts for the epoxidation of propene with O2 and H2. The first effort of the research has focused on the synthesis of well-defined Au/Ti-SBA-15 catalysts. Different methods have been used for the synthesis of the Ti-SBA-15 supports. Au/Ti-SBA-15 prepared by grafting Ti in the SBA-15 structure contains a higher absolute amount of tetrahedral Ti than the ones prepared by a hydrothermal method of Ti-SBA-15 support. The catalytic activity of Au/Ti-SBA-15 catalysts was found to be dependent on the amount and dispersion of Ti as well on the Au particle size. Further more, the epoxidation activity of Au/Ti-SBA-15 catalysts was improved by performing a post-synthesis NH4NO3 treatment. Formation of a large number of surface hydroxyl groups together with ammonium species on the surface support have a beneficial effect on the homogeneous deposition of Au. The superior activity of treated Au/Ti-SBA-15 materials is related to the increased number of Au adsorption sites. Several spectroscopic techniques were used for characterization of the resulted catalysts and for understanding of the reaction mechanism. In-situ XAFS spectroscopy provided evidence on the adsorption of propene on Au nanoparticles. The results indicated that the adsorption of propene is a key step in the epoxidation mechanism over Au/Ti-based catalysts. Furthermore, steady state isotopic kinetic analysis performed on Au/TiO2 and Au/Ti-SBA-15 was used to investigate if the support oxygen is playing a role in the production of propene oxide. The results obtained from our studies were discussed together with the literature reports in order to understand the related propene epoxidation mechanism.