In this thesis testing of solid oxide cells in fuel cell and electrolysis operation have been performed. Attempts to fabricate fuel cells are described, equipment for testing solid oxide electrolysis cells has been constructed and the development process for this described. Cells of a number of different types have been tested in which initial work was performed using microtubular cells. Work on the fabrication of planar solid oxide cells is described, anode supports were prepared by pellet pressing however the application of a suitably dense electrolyte was unsuccessful which resulted in a poor cell OCV. The initial degradation of commercial solid oxide cells has been investigated. During cyclic testing at low current density the cells were found to degrade at twice the rate in electrolysis operation compared to fuel cell operation. This leads to the conclusion that the degradation observed in electrolysis is reversible and that there is a disconnect between the electrolysis and fuel cell degradation processes. During testing at different current densities the cells were found to undergo severe degradation when operated with very low water content supplied to the cells. The degradation was 512 mV kh\(^{−1}\) at 2.5 vol% H2O and reduced to 45mV kh\(^{−1}\) at 50 vol% H2O. Over the timescales investigated in this work and due to the reversible nature of the electrolysis degradation identifying a degradation mechanism was very difficult.

Catalytic deactivation caused by coking was studied in ZSM5 and zeolite Y catalysts during the isomerisation of 1-hexene under sub and supercritical conditions. The effects of varying temperature and pressure, from 220–250 °C and 10-70 bar respectively, on conversion and coke deposition were studied in both zeolites. TGA, DRIFTS, nitrogen sorption isotherms for fresh and coked catalysts and catalyst acidity measurements were compared. In ZSM5 the catalyst was stable for 96 hours. TGA and DRIFTS results show coke deposits were mainly polyolefinic and the amount decreases considerably from 18.8 wt% in the subcritical region to 10 wt% in the supercritical region. In zeolite Y, decay in conversion was observed with the rate of deactivation being slower at supercritical conditions at 235 °C and 40 bar. Naphthalene hydrogenation on NiMo/γ-Al2O3 catalyst was also studied. The effect of temperature, pressure, varying naphthalene feed concentration and operating in sub and supercritical conditions were studied. Coke deposit decreased by 38 wt% in the supercritical region. SC CO2 (Tc 31.04 °C, Pc 73.8 bar) was also used to re-activate the coked catalysts. This resulted in recovering 93% of the catalytic activity and 37% of the coke was effectively extracted by SC CO2.

This thesis pursues an enhanced understanding of flow dynamics and mixing within Sulzer SMX mixers. A number of techniques were used, with the main focus on Positron Emission Particle Tracking (PEPT), as well as Particle Image Velocimetry (PIV) and high speed image capture. PEPT tracer location data was processed to derive properties such as local velocity fields, mixing efficiencies, occupancies and changeover efficiencies, for a number of model Newtonian and non-Newtonian fluids, under industrially relevant flow conditions. It was demonstrated that velocity fields within SMX mixers are not adversely affected by the fluid rheology. Comparable velocity maps were also obtained using PIV, where transparent 3D printed mixer elements were successfully used with the technique for the first time. The assessment of the mixing patterns illustrated that the concentric feed orientation offers the fastest reduction in variance across the mixer cross-section, when compared to side-by-side feed patterns. Analysis of occupancies demonstrated a sharp breakthrough front, reminiscent of plug flow, while the assessment of the changeover patterns further emphasised the resemblance to plug flow within the mixer. A model was derived predicting the time required to achieve a desired level of changeover within a system with known rheological properties and geometry.

This thesis focuses on the atomic structure of Platinum-based nanoclusters, and covers two main areas: Binary platinum-titanium clusters, and pure platinum clusters. Both were produced with a magnetron sputtering, gas-aggregation cluster beam source, and imaged with a Scanning Transmission Electron Microscope (STEM) with detailed image analysis. For the study of Pt-Ti clusters, identifying their overall morphology is key. For oxidised clusters, a Pt core with a TiOx shell is found for smaller clusters, whereas for larger clusters multiple Pt cores are seen within the TiOx shell. The Pt-Ti clusters have been transferred under nitrogen, to reduce oxidation. Here, the morphology of the clusters is more amorphous, with the Pt and Ti atoms forming an alloy core within the cluster. Experimentally, clusters were produced containing 10- 600 Pt atoms. The structural motif of these clusters shows that the large clusters (> 250 Pt atoms) typically present with a cubic structure which matches that of bulk FCC Pt. This experimental work has been complemented by theoretical modeling, to identify dominant motifs within a large size range from 55-10,000 Pt atoms, using empirical potentials. The results show that there is a persistent switching between the decahedral and octahedral motifs.

Emulsions are a vital part of many products in everyday use, such as foods, cosmetics, and even construction materials. Membrane emulsification is a technique which has been used to produce emulsions in a manner contrary to the traditional methods where droplets are broken and re-broken to make smaller and smaller droplets, and instead each droplet is individually formed at a pore on the surface of the membrane. This research compared two of the most favoured membrane emulsification techniques; cross-flow and rotating membrane emulsification. Two systems were built for producing emulsions using tubular microporous membranes, made from shirasu porous glass, polymer, ceramic and stainless steel. One device employed a cross-flow system providing shear to detach the nascent droplets from the membrane pores whilst the other system employed a rotated membrane to produce both shear and potentially centripetal force at the membrane surface. Both systems were used to create emulsions, and the effects of various settings of the systems were investigated. A direct comparison between cross flow membrane emulsification and rotating membrane emulsification were achieved for the first time, as the same membranes were available for both systems. The modular interchangeable nature of the membranes in the systems also allowed direct comparison between the various different membrane types tested. The distinct differences in the structure and materials of the membranes tested was compared, and its effects elucidated, as the different membrane types each showed different advantages and disadvantages when producing droplets. It was shown that the membrane pore size is a major factor on the size of the droplets produced, and the membrane pore size distribution span affects the size distribution span of the droplets. Increasing the emulsifier concentration decreases droplet size, as does increasing the shear force applied to the forming droplets, either by increasing the cross-flow velocity or the rotation rate. Increasing the pressure applied to force the dispersed phase through the membrane increases flux rate, but also increases droplet size slightly. The relative viscosity of the two phases being emulsified has an effect on the droplet size; increasing the continuous phase viscosity decreases droplet size, and increasing dispersed phase viscosity increases droplet size. The systems performed equally well making water in oil, as oil in water emulsions. Although the rotating membrane system produces lower shear rates than the cross-flow system, similar droplet diameters were produced, implying that detachment is enhanced by the rotation, showing a clear advantage to rotating membrane emulsification. The systems were also used to produce various more complex particles, including double emulsions and gelled beads, and the level of control over the phases afforded by membrane emulsification was shown to be an advantage in the production of such microstructures.

Twin screw extruder (TSE) has been studied extensively as a granulator because it allows continuous processing. Initial work was carried out by comparing the TSE with conventional granulator shows that the mechanism of TSE granulation is different from conventional granulation with the absence of the consolidation stage. PEPT was also utilised and it reveals that the flow stream of the material is not only due to the conveying capacity but also the granulator fill, in particularly for the 90o mixing zone which is believed to be a dispersion type of mechanism driven by the granulator fill gradient. Residence time distribution was measured and simulated by fitting the experimental data using a continuous stirred tank reactors model. The model describes the experimental curves reasonably well when a plug flow fraction was considered. Generally the mean residence time (MRT) of the system is proportional to the mixing zone angle and is inversely proportional to the screw speeds and flowrate. A study using the variance reduction ratio demonstrates that the TSE granulator used in the present study is able to remove the feed instability given that the ratio of the frequency of the input stream fluctuation to the MRT is high.

Knowledge of internal porous structure is essential in improving the development of a catalytic system and consequently leading to an optimized performance. XRCT has been used to show differences in density and pore distribution and that differences can be related to the process route by which the specimen was prepared. Alumina samples with defects prepared using different conditions have been investigated in order to acquire information on the introduction and development of cracks. Results indicated that cracks became larger when the water content, the ram speed of the extruder and the drying temperature were higher. The calcination process increased the number and the 3-D size of the cracks. Both concentration of the bulk solution and impregnation time appeared to have a significant impact on the metal distribution profile during the impregnation process. Drying had a strong impact on the metal profile, as a redistribution of copper was noticed. The redistribution was observed exclusively in the case of weak adsorption of copper with alumina support. Experiments on the newly established I13 beamline synchrotron in Diamond was carried out in order to gain an understanding of the cracks induction/propagation and of the diffusion of metallic solution inside the catalyst support.

Cleaning of surfaces in contact with chocolate is necessary in chocolate manufacturing to ensure high production quality. This research was carried out to study the removal behaviour of solid dark chocolate from three surface materials; stainless steel 316, polycarbonate and polytetrafluoroethylene (PTFE). In order to understand the crystallization behaviour and its relationship with the chocolate rheology, a differential scanning calorimeter (DSC) was used to characterize the polymorphs, while a rheometer was used to study the rheological behaviour of chocolate. The polymorphism was found to be influenced by the tempering and cooling processes, with no dependency on the surface material if cooling is properly controlled. Crystallisation can also be studied from the rheological data. Micromanipulation technique and texture analyser were then used to determine the ease of removal of a solid chocolate layer from the surfaces. Finally, both removal behaviour and time to clean were identified using a flow cell cleaning rig with respect to a number of parameters. It was found that surface roughness of the three materials, crystallization and cleaning conditions affect the removal behaviour of solid chocolate layer from those surfaces. Water alone did not produce a clean surface for all the materials used, thus chemical was added. The results from cleaning work and micromanipulation measurements were found to be comparable.

Algae have been studied as a potential biodiesel feedstock by identifying on a global scale suitable cultivation locations for three specific cases (EU, US and Brazil) based on the area requirements. A direct conversion of oil harvested from wet algae to biodiesel was undertaken using ethanol at supercritical conditions, eliminating the use of catalyst, feedstock drying and the oil extraction steps. Chlorella vulgaris with 7.3% wt. lipid content was characterised (by elemental, chemical and thermal analyses) and used to assess the supercritical ethanol approach. A biodiesel yield of 47.5% wt. was achieved in a flow reactor at 260°C, 75 bar, aqueous algae concentration of 6 mg·mL-1 and 2 mL·min-1 flowrate. This result demonstrates the advantages of the flow reactor over a batch process where the maximum biodiesel yield was 26% wt. after 6 hours. A life cycle analysis of the proposed route showed that biodiesel yield must exceed 60% wt. to make the process competitive when compared to the traditional route of oil extraction and catalyst transesterification adopted to algae biodiesel production. In comparison to the soybean biodiesel, the use of algae as feedstock would not be justified unless improvements to reduce energy consumption are made.

The results of the hydrothermal oxidation of 1,8 diazabicyclo[5.4.0]undec-7-ene (DBU), a toxic nitrogen-containing organic compound by different processes is presented here. Subcritical experiments (wet air oxidation (WAO) and catalytic wet air oxidation(CWAO)), were conducted in a batch reactor and supercritical water oxidation (SCWO) experiments were conducted in a tubular reactor. The key operating parameters investigated were temperature, pressure, initial organic concentration; for catalytic experiments, the effect of catalyst metal loading and weight were also investigated. Commercially available Ru/Al\(_2\)O\(_3\) pellets were selected for catalytic study after various catalysts were screened. Results indicated that temperature was the key operating parameter in all 3 processes that affected DBU removal, TOC removal and ammonia yield. Ru/Al\(_2\)O\(_3\) pellets enhanced the DBU removal, TOC removal and decreased ammonia yield compared to WAO. Complete DBU removal was achieved by catalytic wet air oxidation and supercritical water oxidation. Kinetic data was acquired and a pseudo first order kinetic model was used to quantify the oxidation rate. WAO and CWAO were investigated for the treatment of hazardous industrial effluent with a high initial ammonia concentration and complete ammonia removal was achieved using Ru/Al\(_2\)O\(_3\) pellets. A novel ruthenium coated reticulated foam monolith was investigated as an alternative for heterogeneous catalyst instead of Ru/Al\(_2\)O\(_3\) pellets, the results show that it improved catalyst stability and could be an alternative to pellets for wastewater treatment.