Sources of clean and sustainable energy are important vectors for economic growth and development. The current global energy supply depends heavily on fossil fuels, which in the future may have added costs of carbon dioxide emissions. This makes technology such as direct water splitting from harvesting solar energy in photo-electrochemical (PEC) systems potentially attractive. The principle of this technology utilises semiconductors to absorb photons of energy greater than their band gap energy, generating an electron-hole (absence of electron) pair. The hole could oxidise water to produce oxygen at the anode, while the electrons could reduce water to form hydrogen at the cathode. This project aims to design, model, characterise the performance and optimise a photo-electrochemical reactor that could efficiently harvest and store solar energy by splitting water to produce hydrogen and oxygen. α-Fe2O3 ,which is a cheap and abundant material, has shown promise as a photo-anode material, so was chosen as the photo-anode in the development of the PEC reactor. α-Fe2O3 thin films were produced by spray pyrolysis of alcoholic FeIII solutions onto fluorine-doped tin oxide film on glass. Effects of deposition precursor, post deposition heat treatment and SnIV-doping were studied. Results showed that both SnIV-doping and heat treatment were required to produce the best results (photocurrent of ca. 1-2 Am-2 at applied potential of 0.5 V vs. HgO |Hg). A charge carrier transport model was developed to understand and predict the behaviour of the Fe2O3. The model suggested that the magnitude of photocurrent was dependent on the photo-electrochemical reaction rate at the electrolyte | electrode interface, and would be limited by the intensity of illuminated photon flux. Operating the reactor at higher temperatures favoured the electrolysis process in the absence of light, but experimental results showed it was unfavourable for the net photo-generation of charge. A reactor system with a 0.1 m x 0.1 m photo-anode, Ti/Pt cathode and cation-permeable membrane was built to investigate the effects of operating parameters and operational issues of process scale up. COMSOL Multiphysics™ software was used to model the reactor and to study the reactor performance effects of fluid flow, light intensity and electrical potential drop in the thin conducting layer on glass. Results showed that at electrode area of 10-2 m2 scale, a significant electrical potential drop occurred across the photo-anode, due to its sheet resistance, resulting in non-uniform distribution of current density / rate of H2 (and O2) production (solar to hydrogen conversion efficiency of 0.16%), much of the photo-anode area being inactive. A reactor model was developed to provide a better understanding of larger-scale PEC reactor performance and was used to re-design and optimise the next reactor prototype.

The prevalent business model in the pharmaceutical industry requires rapid and robust process development and flexible manufacturing processes. This work reports the attempts to develop structured procedures for membrane process development to meet these requirements. The Donnan Steric Pore Model, in conjunction with a computational molecular dynamics programme, was evaluated as tool for membrane performance predictions to circumvent the need for tedious membrane screening experiments. However, the computational effort required was too onerous, making experimentation more efficient than computational method at this stage. Process chemistry manipulation enabled the use of otherwise incompatible membranes for separation and reduced the time needed for membrane scoping. Firstly, through pH manipulation to selectively increase electrostatic sieving, the permeation selectivity of a membrane to 2 different solutes was changed. Secondly, a structured procedure for polyalkylation of an ‘anchor’ molecule to increase the steric hindrance of an organocatalyst was used to enable the total retention of the catalyst so that a single stage membrane recycling strategy for the catalyst could be enacted. Published membrane processes were analysed and found to lack robustness and to be too sensitive to slight deviations in membrane performance. Hence new membrane processes were devised to address these challenges. Firstly, a membrane cascade process was used to enhance the rejection of an active pharmaceutical ingredient (API) over the single pass membrane rejection. This cascade process was then used for concurrent API concentration and solvent recovery. Secondly, a permeable stripping cascade configuration was used for the removal of an excess reagent from an API to enable the excess loading of the reagent to increase the yield of the API. The membrane cascades benefited from enhanced reliability, increased productivity and improved robustness.

This thesis describes work which was focused on stratifying annular flow in horizontal tubes. Stratifying-annular flows in horizontal tubes are characterized by the tube being wetted with liquid over its whole periphery but with a tendency for more of the liquid to be present in the layer at the bottom of the tube. For such a condition to exist, there have to be some mechanisms by which the liquid is transported at the top of the pipe in opposition to the influence of gravity. Such flows are typically experienced in hydrocarbon recovery and in many other applications as for instance in gas-condensate lines. Unless the liquid phase (to which a corrosion inhibitor is often added) can adequately wet the top of the pipe, corrosion and ultimate pipe failure may occur in this region; this is a crucial problem for the petroleum industry. Three physical mechanisms have been identified for the transport of the liquid phase to the top of the tube, namely: droplet transport, wave spreading and mixing, and secondary flow. The secondary flow mechanism seems unlikely to contribute significantly and the wave spreading mechanism is only significant in smaller diameter tubes (typically 25 mm). For larger pipes, it is the droplet transport mechanism which is likely to occur; in this mechanism, droplets are entrained from the liquid layer at the bottom of the pipe and transported in the gas core at the top where they deposit to form a liquid film. Two mechanisms of droplet transport seem to be significant, namely ballistic transport in which larger droplets move in a direction governed by their initial release velocity and diffusional transport in which droplets move randomly under the influence of the gas flow turbulence. For pipes of medium diameter (for example in the 78 mm diameter pipe used in the experiments described here), ballistic droplet transport is likely to be the dominant mechanism but diffusional transport is expected to be dominant for large diameter pipes. The work described in the thesis comprised both experimental and computational studies. In the experimental work, a special visualization method (namely axial view photography) was employed to study the droplet entrainment and transport mechanisms. The distribution of liquid film flow rate around the pipe was determined using a film extractor device. The axial viewing system was used to investigate droplet entrainment in stratifying-annular air-water flows in a 0.078 m diameter horizontal pipe. The process of droplet entrainment was captured using a high speed cine camera. Both ligament and bag breakup mechanisms leading to droplet entrainment were identified. The creation of a ballistic droplet from a ligament was clearly observed. The film flow rate measurements were used in conjunction with previous measurements of droplet mass flux in the core in comparisons with a computational model which aimed to predict the transport of droplets using Reynolds Averaged Navier Stokes (RANS) modelling embodied in a commercial CFD code (CFX). In this model, the turbulent gas core was modelled and the motion of droplets emitted from the liquid layer at the bottom of the pipe was tracked in this turbulent field. Ultimately, the droplets are deposited on the tube wall and the film flow rate may be calculated using the predicted deposition rates as input data. The thesis closes with a description of a numerical experiment aimed at investigating the influence of the turbulence model on droplet transport. Specifically, comparisons were made between RANS and Large Eddy Simulation (LES) models. [For supplementary files please contact author].

Milling has been the key unit operation in controlling particle size of pharmaceutical powders at scale. The work carried out in this thesis is a comprehensive study of the stability of pharmaceutical solids post-milling and upon storage, from molecular level up to bulk handling scale. It is an attempt to fill key gaps in knowledge with regard to the anomalous behaviour and physical instability of milled powder through the development of advanced novel techniques. The physical instability of milled or amorphous pharmaceutical powders often manifest in changes in derived powder properties. Moisture induced dimensional changes of amorphous lactose compacts were monitored by in-situ environmental controlled optical profilometry. The complex volumetric behaviour involves glassy-rubbery phase transition followed by amorphous-crystalline transformation under the influence of water. These associated changes were not observed in physical aging of amorphous lactose compacts by measuring specific surface area. At the molecular level these physical changes are governed by relaxation processes. By operating within the linear viscoelastic region, low strain uni-axial indentation of small molecule organic glasses at a range of temperature generated master curves using WLF analysis. Viscoelastic behaviour of these materials were determined to be controlled by local β-relaxation around the glass transition rather than globally for polymers. At the bulk level, due to the non-equilibrium nature of milled and amorphous powders, their surface energies tends to be significantly higher than the equivalent crystalline forms. This can be detrimental as highly cohesive and poor flowing powders are difficult to process. The unconfined compression test was adapted to measure cohesion of small weak pharmaceutical powder compacts. More significantly, a positive relationship was confirmed between surface energetics and cohesion of modified D-mannitol. At the particle level, the mechanism(s) by which milling or micronisation creates low levels of amorphicity remains unclear. MOUDI fractionation of bulk micronised α-lactose monohydrate and characterisation of fine fractions has clearly demonstrated that micronisation as well as mechanical particle size reduction also generates low levels of highly amorphous ultrafine particles within bulk crystalline powder which will have a significant effect on powder physical stability post-milling and upon storage. In conclusion, using the novel techniques developed here, significant progress has been towards understanding the physical behaviour of milled and amorphous pharmaceutical solids.

The pharmaceutical industry is going through a rather turbulent period. Many blockbuster drugs have fallen off patent over the past two years and many more are expected to do so in the near future. In response, pharmaceutical companies have continued searching for products that will replace those that have lost patent protection. However, drug development and approval is extremely time-consuming and costly. So that this critical issue is addressed, industry experts and regulatory agencies have jointly proposed the implementation of Quality by Design (QbD) principles in the development and manufacture of all new drugs. Adoption of QbD is expected to reduce drug development cost and approval time. It is also expected to encourage innovation by developing drugs, and the processes used to manufacture them, around the mechanisms that relate process inputs with end product quality. Within this context, monoclonal antibodies (mAbs) are currently the highest-selling products of the biopharmaceutical industry and are projected to account for nearly half of the world’s top-selling drugs by 2018. All currently commercialized mAbs contain N-linked glycans (complex carbohydrates) bound to their protein backbone. These carbohydrates, in turn, have been widely reported to impact the safety and efficacy of mAbs. Furthermore, it has widely been reported that bioprocess conditions heavily impact the composition and distribution of these glycans. For these reasons, mAb glycosylation is considered a critical quality attribute (CQA) of these therapeutic proteins under the QbD scope. Based on QbD principles, the objective of this thesis was to generate a mathematical model that mechanistically relates the effect of nutrient availability throughout cell culture with the glycan profile of a mAb. The model was constructed from three individual ones. The first model describes the N-linked glycosylation process which occurs in the Golgi apparatus. The second model is unstructured and describes cell culture dynamics. The third and final model describes the biosynthetic pathway for nucleotide sugars. All three models were developed independently, but were adapted with features so that they could be interconnected. The glycosylation model approximates the Golgi apparatus to a single plug flow reactor where resident proteins (glycosylation enzymes and transport proteins) are recycled from distal portions of the Golgi space to proximal ones. Optimisation-based methods were developed to estimate unknown parameters of the model. The cell culture dynamics model was developed to represent cell growth, nutrient consumption and mAb synthesis. It was originally based on Monod kinetics, but was adapted to include experimentally-encountered complexity. The model for nucleotide metabolism was heuristically reduced from 35 constituting reactions to 7. Additional mechanistic features were adapted or included to ensure model fidelity. Experimentally, batch cultures were performed with hybridoma (CRL-1606 from ATCC). Data for viable cell density, glucose, glutamine, lactate, ammonia and mAb titre were collected. Intracellular samples were produced by perchloric acid extraction. These samples were then analysed for nucleotide sugar content using a high performance anion exchange chromatographic method which was optimized to quantify eight nucleotide sugars and four nucleotides in 30min. mAb bound glycans were analysed by MALDI mass spectrometry. The experimental data was used to estimate the unknown parameters of the models. The models – along with their associated parameters – were then combined to produce a coupled model that mechanistically relates nutrient availability with mAb glycosylation-associated quality. With further validation, such a model could be used for bioprocess design, control and optimization.

One key barrier to the timely and efficient production of biopharmaceuticals is that they are prone to both chemical and physical instability including aggregation. Protein-protein molecular interactions are known to be a factor in protein solution aggregation behaviour; however their practical and elementary importance has not been fully established. The osmotic second virial coefficient (B22) is a fundamental physiochemical property that describes the molecular interactions between proteins in solution which could result in aggregation. This experimental study reports on the B22 of five different proteins (lysozyme, lactoferrin, catalase, concanavalin A and anti-TNFα dAb) using Self-Interaction Chromatography (SIC) over a wide range of solution conditions including the effects of pH, salt concentration, salt type as well as excipient stabilisers. It was established that current practise for SIC peak analysis was inadequate, and an improved method of SIC peak analysis was deployed which provided improved B22 data quality and robustness. Protein aggregation performance was evaluated under identical solution conditions to those used for the B22 determinations using a simple Dynamic Light Scattering based measurement of initial solution aggregate size. This data shows a direct and strong correlation between B22 values and protein aggregation performance for all systems studied here. Specifically, proteins solution systems with B22 values of ~ 1 x 10-4 mL mol g-2 or less, regardless of solution composition, all exhibited aggregation behaviour. This thesis establishes the decisive importance of B22 as a predictor of aggregation performance. Preliminary validation data is presented for a method which allows for B22 estimation from osmotic cross virial coefficients (B23), removing the need for immobilisation of the protein of interest and therefore allowing rapid B22 screening.

No description available.

No description available.

No description available.

No description available.