Heavy oil and bitumen are known to constitute high-boiling molecules which gives them characteristic high viscosity, high density/low API gravity, low yields of fuel distillates, and high heteroatom content compared to light oil. Upgrading therefore refers to the breaking down of heavy oil into oil with similar characteristics as light crude oil. The toe-to-heel air injection (THAI) and its catalytic add-on CAPRI (CAtalytic upgrading PRocess \(In-situ\)) were developed to achieve this objective down-hole. In this study, the CAPRI process was explored with the objective of controlling catalyst deactivation due to coking while increasing the extent of upgrading. The effects of reaction temperature and weight hourly space velocity on the extent of upgrading were studied in the range of 350-425\(^o\)C and 9.1-28 h\(^-\)\(^1\), respectively. In order to control premature deactivation of the catalysts due to coke and metal deposition, the following were investigated activated carbon guard-bed on top of the catalyst bed, hydrogen-addition, steam environment as a source of hydrogen-donor, and nanoparticulate catalyst. It was found that high reaction temperature of 425\(^o\)C and lower WHSV (9.1 h\(^-\)\(^1\)) improved the cracking as well as increase API gravity (~3-7\(^o\)), viscosity reduction of (81.9 %), demetallisation (9.3-12.3 %), desulphurisation (5.3-6.6 %), and higher yield of fuel distillates, respectively compared to upgrading at 350 and 400\(^o\)C. In spite of the improvement in produced oil at 425 \(^o\)C, the carbon-rejection was high (51-56.6 wt.%) compared to (42-47.8 wt.%) and (48-50.3 wt.%) when reaction was carried out at 350 and 400\(^o\)C for 25 hours operations.

To predict cell breakage in bioprocessing it is essential to have an understanding of the cell wall mechanical properties. This project involved a study of the wall mechanical properties of individual Baker’s yeast cells (Saccharomyces cerevisiae) using compression testing by micromanipulation. An analytical model has been developed to describe the compression of a single yeast cell between flat parallel surfaces. Such cells were considered to be thin walled, liquid filled, spheres. Because yeast cells can be compressed at high deformation rates, time dependent effects such as water loss during compression and visco-elasticity of the cell wall could be and were neglected in the model. As in previously published work, a linear elastic constitutive equation was assumed for the material of the cell walls. However, yeast compression to failure requires large deformations, leading to high wall strains, and new model equations appropriate to such high strains were developed. It was shown that the preferred model, based on work-conjugate Kirchhoff stresses and Hencky strains, fitted Baker’s yeast compression data well up to cell failure. This agreement validated the modelling approach, which might also be useful in characterising the material properties of the walls of other cells and microcapsules. Using the analytical model, the effects of compression speed on the elastic modulus obtained by fitting numerical simulations to experimental compression data was investigated. It was found that above a compression speed of approximately 45 µms\(^{-1}\) the estimated elastic modulus was essentially unchanged. By using a compression speed of 68 µms\(^{-1}\) it could be assumed that water loss during compression was negligible. It was then possible to treat the initial stretch ratio and elastic modulus as adjustable parameters within the numerical simulation. In addition to this, as the numerical simulations fitted experimental data well up to the point of cell rupture, it was possible to extract cell wall failure criteria. This study has given mean cell wall properties for late stationary phase Baker’s yeast of: elastic modulus 185 ± 15 MPa, initial stretch ratio 1.039 ± 0.006, circumferential stress at failure 115 ± 5 MPa, circumferential strain at failure of 0.46 ± 0.03, and strain energy per unit volume at failure of 30 ± 3 MPa. Following this, the effect on the intrinsic material properties of treating Baker’s yeast with dithiothreitol (DTT) was investigated. DTT has the effect on Baker’s yeast cells of breaking the disulphide bonds in the cell wall releasing invertase into the suspending solution. It was found that this did not affect the intrinsic material properties or failure criteria. In addition to this, Baker’s yeast cells were mechanically perturbed by sonication and the resulting intrinsic material properties investigated. The surface modulus was found to decrease with increased sonication time while the surface strain energy at failure remained constant. However, it was not possible to determine the extent of damage to each individual cell, preventing explicit conclusions from being reached.

In this work, size-selected molybdenum disulfide (MoS\(_2\)) nanoclusters were produced using a magnetron sputter source and time-of-flight mass filter. Magnetron sputtering is a common industrial method for preparation of MoS\(_2\) thin films. The combination of this technology with accurate size control allows us to produce, in high vacuum, lab-scale quantities of size-selected clusters. The strong spatial confinement effects in MoS\(_2\) suggests that such control will modify the catalytic properties. This method also has potential to enhance MoS\(_2\) performance in areas such as hydrodesulfurisation, intercalation batteries and tribology; as well as elucidating the dynamics of compound formation in the gas-phase. Structural properties of these MoS\(_2\) clusters are studied using aberration-corrected STEM. The optimum catalytic size range of 1-5nm has not previously been studied in detail for gas phase synthesis. This work bridges the gap in the cluster beam literature between small, few atom clusters and the production of large MoS\(_2\) fullerenes and monolayers. It has been found that MoS\(_2\) clusters display a characteristic layered structure down to the smallest studied cluster, 50 units of MoS\(_2\). Growth of clusters is indicative of anisotropic growth from the reactive edge sites, proceeds by subsequent addition of van der Waals bound layers and finally coalescence of smaller units in the case of large clusters. The electrocatalytic properties of these clusters are explored by cyclic voltammetry and show good activity for the Hydrogen Evolution Reaction despite the presence of surface oxides. The reaction current normalised matches to loading matches some of the best catalysts produced to date.

Industrial alkaline electrolysers were electrochemically characterised and analysed on an internal combustion engine. These electrolysers exhibited low efficiencies, low gas flowrates and subsequently zero change in engine emissions due to the poor design and build. An improved alkaline electrolyser was designed, built and tested exhibiting improved efficiency/gas output compared to the industrial electrolysers and an improved reduction in emissions. The increased power consumption of the electrolyser results in a rise in electrode degradation which is responsible for the decrease in electrode lifetime. A method for prolonging the electrode lifetime is proposed through a metallic “oxygen-getter”. Implementation of this has shown to prevent cathodic corrosion of the electrode material and thus reduces oxide layer formation. Electrode lifetime in an alkaline electrolyser increased, but the commercial trend is shifting towards the more attract PEM technology for electrolysis due to higher current densities, ability to handle variable input loads and non-caustic liquid requirement. A commercial on-demand PEM electrolyser was tested and system designed for integration with an existing hydrogen refuelling station at the University of Birmingham. This mimicked the case for a distributed hydrogen system where the hydrogen is produced onsite for fuel cell vehicles resulting in a carbon neutral fuel.

This work had as main objective to recover bioactive compounds, more specifically phenolics, from cider lees using subcritical water, an environmentally benign fluid. Maximum solubilisation was obtained at 150ºC for 30 minutes, and a decrease is observed at higher temperatures, indicating the occurrence of precipitation. Maximum total phenolic content (FCR) was obtained at 220ºC for 10 minutes. Chlorogenic acid was identified as the major phenolic in cider lees, however at 220ºC for 10 minutes, only trace levels were found, therefore the high total phenolic content should be the result of interactions between phenolics and macromolecules present in solution. Sequential subcritical water was employed to reduce complexity of extracts produced. Maximum antioxidant capacity (FCR, ORAC) was obtained at 120ºC and 220ºC, where at 120ºC is due to cider phenolics while at 220ºC, after the removal of most of the compounds, it is only left cell wall carbohydrates and chitin, which should be responsible for the formation of derivatives with high antioxidant capacity. Cyclodextrin was used to improve oxidative stability of the cider lees extracts, however while it improved the oxidative stability for cider lees extracts obtained at low temperatures, to cider lees extracts obtained at higher temperatures, no improvement was registered.

Microfluidic systems attract attention because the benefit they offer such as very high surface area to volume ratio and the precise control of the flow features or droplets formation. However, the study of displacement flows has not been systematically explored. This thesis presents a study of the behaviour of liquid-liquid two-phase flows in microchannel during the displacement of one liquid by another. Liquid displacement using immiscible and miscible fluid pairs were carried out. The kinematic viscosities of the fluids used range from 1 to 100 cSt. Three types of straight channel and a T-junction channel were tested. The flow activities at the fluids interface, with the addition of surfactants, were mainly investigated. Flow regimes were identified and their occurrence was illustrated via flow pattern maps. CFD simulations are applied largely in the study of fluid mechanics aiming to confidently predict flow behaviour from computational methods. The experimental results were used to validate CFD simulations carried out using the Fluent package incorporating the VoF model. Good agreement between simulation and experiment results was achieved.

Fouling and cleaning processes impact industrial production, in terms of economics, product quality, product safety, and plant efficiency. Therefore, optimisation of fouling and cleaning processes is a significant issue, and needs a good understanding of fouling and cleaning kinetics. Ideal monitors should determine the right time when a process run should stop and when a plant will be clean in order to improve the process efficiency. This thesis investigated the fouling and cleaning behaviour of dairy fluids in a plate heat exchanger (PHE) and bench scale fouling rig, using whey protein concentrate (WPC) and WPC-/m (with added minerals) as fluid models. Fouling and cleaning monitoring methods were also investigated as new ways to operate and control the processes. Experiments displayed that fouling increased with increasing protein concentration up to a limit of approx. β-Lg 0.3 wt. %. Increasing the flow rate from 100 to 150 l/h decreased the Δ(ΔP) fouling rate for β-Lg concentrations of 0.1, 0.3 and 0.5wt.% by 34, 70 and 72.7%, respectively, due to the increasing of shear stresses at the heat transfer surface. Adding minerals to WPC has lowered the temperature at which β-Lg begins to denature. The differences in fouling behaviour of WPC and WPCm had an effect on cleaning behaviour. Increasing the mineral content in WPC deposits leads to cleaning behaviour which differs completely from that of proteinaceous deposit as no pressure peak is observed.

Chromatography remains the most commonly employed method for achieving high resolution separation of large-sized biomolecules, such as plasmid DNA, typically around 150-250 nm in diameter. Currently, fractionation of such entities is performed using stationary phases designed for protein purification, typically employing pore sizes of about 40 nm. This results into a severe underexploitation of the porous structure of the adsorbent as adsorption of plasmid DNA occurs almost exclusively on the outer surface of the adsorbent. In this study, the effect of two processing parameters, the ionic strength of agarose solution and quenching temperature, on the structure of the resulting particles was investigated. Three characterization methods, Atomic Force and cryo-Scanning Electron microscopy, as well as mechanical testing of single particles where used to quantify the effect of these parameters on the pore size/size distribution and mechanical properties of the adsorbent. In the presence of salt, it was found that agarose fibres tend to aggregate, leading to a gel with large pore size and wide pore size distribution. In fact, for the narrow range of ionic strength used (0-0.1m), a five-fold increase in pore size of the gel was observed. The same type of enlarged agarose structures was observed when slow cooling was applied during the gelation of agarose. The increase in pore size of the gel was also accompanied by an increase in the compression strength and the elastic modulus of the particles, i.e. particles with 200 nm pore size were found to have higher compression strength (1.5-fold difference) than those with 40 nm pore size.

Microencapsulation is an efficient technology to deliver perfume oils from consumer products onto the surface of fabrics. Microcapsules having uniform size/mechanical strength, may provide better release performance. Membrane emulsification in a dispersion cell followed by in-situ polymerization was used to prepare narrow size distribution melamine-formaldehyde (MF) microcapsules containing several types of oil-based fragrances or ingredients. Investigated in this study are the parameters impacting to the size and size distribution of the droplets and final MF microcapsules. A pilot plant-scale cross-flow membrane system was also used to produce MF microcapsules, demonstrating that the membrane emulsification process has potential to be scaled up for industrial applications. In this study, health and environmental friendly poly (methyl methacrylate) (PMMA) microcapsules with narrow size distribution were also prepared for the first time using the dispersion cell membrane emulsification system. Characterization methods previously used for thin-shell microcapsules were expanded to analyse microcapsules with thick shells. The intrinsic mechanical properties of thick shells were determined using a micromanipulation technique and finite element analysis (FEM). The microcapsules structure was also considered in the determination of the permeability and diffusivity of the perfume oils in good solvents.

Build-up is a fouling process in which a small fraction of a powder becomes a film adhered to a surface. Specifically buildup in auger fillers which is the focus of this thesis is a significant problem for industry leading to equipment downtime for cleaning, drives tripping due to high torque and reduced powder quality. The work presented in this thesis shows that for a unblended powder with a particle size similar to that of the auger fillers auger/tube clearance, build-up will be formed when powders have b-1 Kawakita parameter of less than 0.5 MPa and tablet strengths below 0.6 MPa post compression to 58 MPa. It is proposed based upon results the work of Crutchley and Bridgewater [29] and Calvert et al. [30] that powders which form build-up due so due to the stresses they experience as they pass through the auger/tube clearance. Additionally this stress is a function of the ratio of the powder’s particle size to the magnitude of the auger/tube clearance C. Work presented in Chapter 8 supports this conclusion showing that once C/D90 is greater than 3.5, build-up is not formed.