Abstract Catalytically converting CO2 to methanol by hydrogenation offers a method to effectively reduce the excessive CO2 emission in the atmosphere and produces value-added chemicals simultaneously. Thus, the investigations on catalysts in methanol synthesis reaction has gained attraction in the past few years. Commercial catalysts based on Copper, Zinc, and Zirconium are popular, but increasingly, researchers are looking for other options with superior conversion and selectivity. From prior literature, catalysts based on Nickel and Gallium, specifically a Ni5Ga3 bimetallic catalyst exhibits a similar CO2 conversion and higher methanol yield compared with commercial Copper-based catalysts. However, the purities of Ni5Ga3 catalysts were found to be restricted during the reported synthesis process. Thus, a simpler and reproducible method to prepare highly pure Ni5Ga3 is desirable. In this study, we developed a method to synthesize highly pure Ni5Ga3 catalyst from hydrotalcite-like compounds (HTlc) precursors for CO2 hydrogenation to methanol. A series of Ni-Ga HTlc precursor was synthesized in the temperature range between 90 C and 150 C. The results indicated the HTlc phase in the nickel-gallium precipitant became better crystallized and the structure became more stable as the synthesis reaction temperature increased. Bimetallic alloy Ni5Ga3 was obtained by reducing the as-prepared HTlc precursors in a hydrogen atmosphere. X-ray absorption spectroscopy (XAS) investigation confirmed that a stable and complete HTlc precursor structure assisted in the synthesis of a steady and perfectly structured Ni5Ga3 alloy, where the bond distance of Ni-Ga and cell volume increased with temperature. Ni-Ga HTlc precursor prepared at a hydrothermal temperature of 110 C resulted in the formation of bimetallic alloy, Ni5Ga3, which demonstrated characteristics such as smaller crystal size and stable structure under optimized conditions. The enhanced performance was demonstrated by an endurance test with a constant CO2 conversion and 100% methanol selectivity at 200 C, and the turnover frequency reached 0.27 s-1. Metal oxide promoters are well known to enhance catalytic properties, thus, a modification by incorporating promoters, such as Mg, Zn and Zr, was investigated. A new series of Ni-Ga-X HTlc precursors (X represented Mg, Zn and Zr) were prepared by a similar synthesis procedure, followed by a H2 reduction process. The results revealed that the main Ni-Ga phase transformed from Ni5Ga3 to Ni3Ga when promoters were incorporated in the Ni-Ga catalytic system, due to an unstable HTlc structure as additional elements were incorporated in the parent precursor. Mg and Zr were present as metal oxides, while ZnGa2O4 structure was present in Zn-promoted Ni-Ga catalysts. The BET surface area was measured for all prepared Ni-Ga-X catalysts, and the surface area exhibited a sharp increase after the promoter modifications. Among all samples, the Ni-Ga-Zr revealed the highest BET surface area. TEM-mapping measurements, for Ni-Ga-Zr catalyst, showed Ni-Ga assembly as a core, while Zr surrounded the core, which isolated and separated Ni-Ga catalysts. Thus, the average particle sizes of Ni-Ga-Zr catalysts were considerably decreased compared with other samples, resulting in a relatively large surface area. However, the promotion effect was not obvious in other samples, because Mg could not be completely precipitated in the catalysts and ZnGa2O4 was formed instead. Furthermore, ZrO2 also facilitated the reduction of Ni-Ga-Zr HTlc precursor due to an enhanced electron transfer. Additionally, incorporation of promoters generated additional strong basic sites in the catalytic system, as demonstrated by CO2-TPD measurement. The catalytic properties were evaluated, and a maximum methanol yield (3.8%) was obtained over a Zr-modified Ni3Ga catalyst at 300 C, 30 bar, which exhibited a similar reactivity of commercial Cu-based catalysts. The Ni-Ga-Zr catalysts were subsequently mixed with a commercial high-temperature CO2 adsorbent (MG50). The Ni-Ga-Zr (NGZr) and MG50 were well-mixed, as revealed from SEM images, and the Ni3Ga phase did not change when MG50 was introduced in the Ni-Ga-Zr catalytic system. A series of mixed samples, with different ratios of MG50 and NGZr, was prepared. The corresponding CO2 conversion exhibited a mild decrease as the amount of Ni-Ga-Zr decreased due to loss of active sites, however, the methanol space-time yield was greatly improved as MG50 increased, which suggested that the catalytic property was considerably promoted in the presence of MG50. The highest space-time yield was observed in 25%NGZr/MG50 mixture, with 123.5 gmeth/gcat/h at 300 C. The promotion was ascribed to enhanced CO2 adsorption on MG50 adsorbent, resulting in higher CO2 concentration adjacent to NGZr active sites, contributing to a higher reaction rate and CO2 conversion. Despite the great improvement in methanol space-time yield in NGZr/MG50, the overall CO2 conversion was lower than that of Cu-based catalysts under moderate temperatures, such as 200 C - 250 C. Thus, the NGZr catalysts were subsequently modified by optimizing the Zr amount in the NGZr catalytic system. The TEM-mapping revealed that once the ZrO2 concentration increased above 15%, ZrO2 experienced a severe agglomeration between the Ni-Ga particles instead of surrounding them. Consequently, the interactions between Ni3Ga and ZrO2 was not further increased as Zr content increased from 15% to 25%. The batches of NGZr catalysts were tested for catalytic performance, respectively. The Ni3Ga catalysts with 15% Zr content exhibited a higher CO2 conversion under the entire reaction temperature range when compared to Cu-based catalysts, which indicated that the Ni-Ga-Zr (15%) catalyst area promising candidate for future catalytical CO2 conversion to methanol.

The unique capability of piezoelectric materials to convert between mechanical and electri-cal energy holds tremendous potential in enabling a range of emerging applications. Pol-ymers, as soft and biocompatible materials, are excellent candidates for the use in power-ing wearable and implantable electronics, as well as for the primary sensing mechanism in soft robotic interfaces. However, piezoelectric polymers are sparsely utilised due to their chemical and structural complexity, and the tremendous energetic cost to maximise their energy conversion efficiency. Fluoropolymers have piezoelectric figures of merit rivalling those of the widely used ceramics and are therefore promising to investigate. The common processing techniques for fluoropolymers revolve around solution casting from toxic, hazardous, and/or high boiling point solvents, which require lengthy solvent evaporation times and arduous post-processing by electrical poling, applying high electric fields to align the dipoles. Recent advances in three-dimensional (3D) printing show promise in order to process fluoropolymers into piezoelectric devices, inducing shear forces on the polymer chains during extrusion toward greater alignment and tailored architectures. In this work, pathways to improving the piezoelectric output of a fluoropolymer, poly(vinylidene fluoride-co-trifluoroethylene) (PVDF-TrFE) were thoroughly investigat-ed. The solvent evaporation-assisted (SEA) 3D printing technique was adapted to printing fluoropolymers, investigating the effects of layer-by-layer deposition on the optical, pol-ymorphic and electromechanical properties. In combination with 3D printing, two classes of nanoscale additives were further investigated, single-walled carbon nanotubes (SWCNTs) and transition metal carbides (MXenes), to elucidate their role in the evolution and alignment of the piezoelectric polarisation. The first part of the thesis focused on the development and optimisation of 3D printing capabilities for fluoropolymers. A binary solvent mixture was optimised by Hansen sol-ubility parameters and the rheological properties were thoroughly probed to optimise the polymer concentration. The effects of printing parameters were further investigated in or-der to minimise spreading of the resultant ink post-printing. The polymers, 3D printed up to 19 layers were transparent and exhibited piezoelectricity, with minimal changes in the electroactive phase fraction and without electrical poling. These results confirmed that shear stresses impart partial polarisation on the extruded materials, and provided a strong foundation for the further studies investigated in this thesis. The second study of this thesis critically investigated the effects of the incorporation of SWCNTs, as a nanoscale additive, into the PVDF-TrFE coupled with the developed 3D printing process. The composites were printed as single-layer films, found to be transpar-ent at carbon nanotube loadings up to 0.05 wt%, with low haze. The piezoelectric proper-ties were investigated through two techniques, piezoresponse force microscopy (PFM) and bulk electromechanical characterisation, finding the greatest enhancement in piezoelec-tric properties at a 0.02 wt% loading of the SWCNTs from both techniques. Molecular dynamics (MD) simulations of the carbon nanotube interface with the polymer confirmed a polarisation enhancement effect, providing the first report of polarisation in the absence of electrical poling. Furthermore, the composites were found to be recyclable in pure ace-tone, a green and low boiling point solvent, allowing the printed piezoelectric polymers to be reprinted, with minimal changes in the chemical, physical and electroactive properties. The final part of this thesis utilised two-dimensional (2D) MXene nanosheet additives as a model non-piezoelectric system to deduce and provide the first report on the mechanism of physical polarisation locking in the PVDF-TrFE, building on the knowledge of the first two studies. The composites were printed directly from acetone as a physical gel, allowing for a faster solvent evaporation rate and therefore improved crystallisation kinetics. MD simulations found a suppressed electroactive phase fraction of the polymer directly adja-cent the surface of the additive, confirmed experimentally by Raman microscopy and dif-ferential scanning calorimetry. Furthermore, the MD simulations found the polarisation vector direction was locked perpendicular to the basal plane of the MXene, which was governed by electrostatic interactions. PFM results confirmed the dipole locking phenom-enon, whereby the polarisation magnitude increased logarithmically with an increase in the MXene loading, while demonstrating the transition metal carbide had no discernible out of plane polarisation. Direct piezoelectric effect measurements via macroscale electromechan-ical testing showed that the composite material exhibited a larger piezoelectric coefficient relative electrically poled polymer, implying that physical poling and polarisation locking from a nanomaterial template can impart greater polarisation than standard electrical poling techniques. In summary, this thesis has developed and applied a fundamental understanding of the origins of piezoelectricity in fluoropolymers and how this phenomenon can be controlled at the nanoscale. The implications of this research are far-reaching, enabling commercial viability of piezoelectric materials in a multitude of emerging applications.

The developments in polymer synthesis and their applications over the last 100 years have had a profound effect on the world as we know it. While polymers abound in our everyday lives, significant advances continue to be made in the field of polymer chemistry. The first reversible deactivation radical polymerization (RDRP) method was discovered in the 1980s, enabling the synthesis of polymers with tailored molecular weights and advanced architectures. Reversible addition-fragmentation chain transfer (RAFT) polymerization is one type of RDRP that has shown exceptional promise for the synthesis of advanced materials, with recent developments in photo-mediated RAFT polymerization leading to unprecedented control of polymer size and architecture. Star polymers are one type of advanced architecture readily synthesized by RDRP and other techniques for a range of applications. Many of the recent advances in RAFT polymerization have yet to be applied in advanced architectures such as stars. Thus, the objective of this thesis is to investigate the synthesis of star polymers and their applications, with focus on stars synthesized by photo-mediated RAFT polymerization and N-carboxyanhydride ring-opening polymerization (NCA ROP). We first investigated the synthesis of star polymers via photoiniferter RAFT polymerization using a core-first approach. Multifunctional cores were synthesized for star synthesis via RAFT polymerization and early results demonstrated the highly “living” nature of the star polymers. This highly living nature allowed us to synthesize ultra-high molecular weight (UHMW) star polymers with molecular weights in excess of 20 MDa, the largest reported for any RDRP techniques to date. Although photoiniferter RAFT polymerization resulted in highly living stars, it does suffer from two significant limitations. The polymerizations are sensitive to oxygen and can only be mediated by a small portion of the electromagnetic spectrum. Photoinduced energy/electron transfer (PET)-RAFT polymerization can overcome these limitations by employing a photoredox catalyst or photosensitizer. We synthesized a novel self-assembled photocatalyst with broadband and near-infrared (NIR) absorbance, facilitating polymerization across the UV, visible light, and NIR regions. Using this photocatalyst we reported the first NIR-mediated RDRP in aqueous conditions and NIR-mediated star polymerization. Finally, the application of star polymers with degradable polypeptide segments was investigated for lysosomal escape of a model drug in vitro. Endo-lysosomal escape of therapeutics to reach their site of action remains a challenge in the field of nanoparticle drug delivery. Star polymers have exceptional potential in drug delivery stemming from their nanometer size and compact, unimolecular structure that allows for functionalization of distinct core, arm, and peripheral regions. Dye-loaded polypeptide star polymers with PEG-brush coronas were biocompatible, demonstrated successful internalization by cells, and allowed the dye to escape the lysosome due to enzymatic degradation of the polypeptide arms while a nondegradable control star accumulated in the lysosome, demonstrating the advantage of polypeptide stars for drug delivery. Together, these results mark exciting advances in the synthesis of star polymers and their applications. The versatility and exquisite control of photo-mediated RAFT polymerizations are demonstrated, while the advances reported in this thesis show great potential for numerous biomedical applications including biomaterials, cell surface modification, and drug delivery.

Replacing coal with natural gas is one of the feasible approaches to mitigate CO2 emission at present. To satisfy the increasing global demand for natural gas, more research and development efforts are being put in converting the highly sour natural gas into applicable energy sources. Adsorption processes offer a promising method to remove the bulk sour component, CO2, from high pressure sour natural gas. Adsorptive properties of six zeolites and one silica gel (CBV-760, CBV-780, HSZ-320HOA, HSZ-350HUA, HSZ-385HUA, HSZ-390HUA, and Sorbead WS) were characterised by various methods. Their adsorption isotherms of CO2 and CH4 in the low-pressure regime (0 – 10 bar) were used to determine the isosteric heat of adsorption, while the high-pressure adsorption isotherms (0 – 50 bar for CO2 and 0 – 100 bar for CH4) were fitted to the Toth and the Langmuir adsorption isotherm models. The model fitting parameters were further used in a rigorous simulation model to evaluate the performance of the selected adsorbents in separating a gaseous mixture with 30% CO2 and 70% CH4. The simulated CH4 recovery were used as a major adsorbent performance metric. Besides, another traditional adsorbent performance indicator, IAST (Ideal Adsorbed Solution Theory) selectivity, was employed for the screening of the adsorbents. Both evaluation metrics gave different results about the performance ranking for the same CO2/CH4 separation application. Although HSZ-320HOA was suggested to be the best adsorbent among selected materials by the IAST selectivity, the process simulation preferred Sorbead WS which achieved the highest CH4 recovery (71%) among selected adsorbents. To push the limitations of the traditional adsorbent performance indicators, a new indicator was proposed. The new indicator, adiabatic column working selectivity, contained all features that other indicators have, and also involved a new factor, the components remaining in the void gas phase. The adsorbent evaluation results of the adiabatic column working selectivity were compared with the results of other indicators and the rigorous simulation model. It was found that only the new indicator could match the results of simulations well, while other indicators deviated more or less from the simulation results. The new indicator was proved effective for screening the adsorbents for the application of CO2 removal from natural gas, but to apply this indicator, two process variables, the temperature and bed composition at the end of the desorption step must be determined beforehand. To overcome this difficulty, two simplified simulation models were developed to facilitate the calculation of all critical process variables. The newly developed simplified models could determine the comparable desorption conditions as well as the rigorous simulations, but within much less computation time. A pressure swing adsorption (PSA) plant was operated in the CO2CRC Otway CO2 Capture Research Facility applying Sorbead WS as the adsorbent to separate CO2 from gas mixtures with various compositions (28 mol% to 78 mol% CO2 balanced with CH4) under different conditions from May 2017 to May 2019. Using the recorded process conditions, the process performance predicted by the simplified models could agree well with the actual process performances. This outcome verified the validation of the simplified models. Overall, we found a novel adsorbent performance indicator, adiabatic column working selectivity to evaluate the performance of adsorbents for the application of CO2 removal from natural gas at high-pressure. At the same time, with the help of a newly developed simplified verified by a real PSA plant, the calculation results of the adiabatic column working selectivity can more precisely predict and screen adsorbents for the targeted application.

Climate change due to anthropogenic carbon dioxide emissions (e.g., combustion of fossil fuels) represents one of the most profound environmental disasters of this century. According to the Special Report completed by Intergovernmental Panel on Climate Change (IPCC) in October 2018, maintaining global warming at 1.5 Celsius degree requires a reduction in CO2 emissions of 49% by 2030. To meet this urgent target, power plants have to equip with carbon capture and storage (CCS) technology. For postcombustion CO2 capture from flue gas, there are four main challenges: (1) the low pressure (ca. 1 bar) of the flue gas produced by power plants; (2) the lower CO2 concentration (15-16%) of flue gas; (3) a high CO2 removal requirement (50-90%); and (4) a low energy consumption of the applied technology. Membrane-based CO2 separation is an attractive technology that meets many of the requirements for economic postcombustion CO2 capture. Within this field, thin-film composite (TFC) membranes are particular attractive due to their high gas permeance. TFC membranes are usually composed of three layers: (1) a bottom porous support layer to provide mechanical strength; (2) a top thin (<1 micrometer) species-selective layer to provide selective function; and (3) a highly permeable intermediate gutter layer to improve the compatibility between the support layer and selective layer. Currently, a key challenge in the development of high performance TFC membranes has been to simultaneously maximize the transmembrane gas permeance (by minimizing the gas permeation resistance of each layer) while maintaining high gas pair selectivities. Two-dimensional (2D) nanosheets are recognized as promising candidates for preparing highly permeable and selective membranes by virtue of their nanosized thickness along with their regular in-plane or interlayer pore arrays, providing minimum transport resistance and maximum selectivity based on the molecular size or solubility. In this thesis, based on 2D nanosheet materials, novel gutter layers and selective layers were developed for the fabrication of flexible, ultrapermeable, and highly processable TFC membranes for economic postcombustion CO2 capture. This includes: (1) the synthesis of ultrathin metal-organic framework (MOF) nanosheets using a surfactant or solvent modulation method; (2) the development of gutter layers employing pristine 2D MOF nanosheets or blending 2D MOF nanosheet into poly(dimethylsiloxane) (PDMS) materials; (3) deciphering the physical aging behavior of polymers of intrinsic microporosity (PIMs) materials; and (4) using PIM@MOF composite as selective layers. The fabricated TFC membranes in this thesis showed much enhanced CO2 permeance (> 1,000 GPU) along with good CO2/N2 selectivity (> 25), meeting the requirements for economic postcombustion CO2 capture. Thus, the research method included in this thesis provides new strategies for the preparation of high performance TFC membranes, which may also be used for other gas or liquid separation applications.

Cheddar cheese is formed by enzyme-induced coagulation of milk proteins, known as ‘rennet gelation’. During rennet gelation of milk, only the casein proteins are coagulated while the whey proteins are expulsed from the coagulum. This results in approximately 20% of the protein being lost during cheese production. Although whey proteins are nutritionally valuable and can be converted to commercial products such as powder, they are often disposed of as a waste. Incorporating them back into cheese is an attractive in-situ method of utilizing the whey protein released during cheese making. It also offers the opportunity to produce cheese with elevated protein and nutritional properties. Concentrating native milk proteins using membrane filtration and denaturing whey proteins using heat are conventionally used to increase cheese yields and reduce protein loss. However, the impact of formulating cheese milk with altered whey protein contents and functionality on the rennet gelation stage is not yet fully understood. This knowledge is vital to producing cheese curds with acceptable coagulation times and curd properties. Therefore, this thesis aimed to deepen our understanding of the influence of whey proteins on the rennet gelation process. Interactions between whey proteins (in native and denatured forms), the rennet enzyme, and coagulating para-casein micelle particles were investigated experimentally. Previous studies report that denaturing whey proteins in the presence of caseins impairs the coagulation of para-casein micelle particles during rennet gelation. However, less is known about the impact of whey proteins in their native form, which is relevant to the use of cheese milk with a protein composition altered using membrane filtration. Therefore, the first phase of this study aimed to investigate the effect of native whey proteins on rennet gelation kinetics. Cheese milks with a wide range of whey protein:casein ratios (with standardised casein concentrations) were formulated using native protein concentrate powders produced by membrane filtration. Oscillatory rheometry and casein macropeptide release measurements during rennet gelation demonstrated that native whey protein impaired enzymatic hydrolysis and significantly delayed the subsequent aggregation of para-casein micelle particles. These observations were independent of changes in the ionic balance or the viscosity of the different milk systems. Binding between whey protein and casein micelles or whey protein and rennet was not observed by dynamic light scattering particle size measurements or native poly-acrylamide gel electrophoresis. While there was no evidence of binding between native whey proteins (approximately 5nm in size) and casein micelles (approximately 200 nm in size) that could be responsible for the impaired rennet gelation, it was instead proposed that whey proteins passively occupy the gaps in the ‘kappa-casein hairy layer’ on the casein micelle surface, which arise as a result of enzyme hydrolysis of the kappa-casein. The whey proteins thereby provide a steric hindrance to rennet reaching the casein micelle surface and a barrier to intimate contact between destabilised casein micelles leading to slower gelation. Incorporating denatured whey protein aggregates into cheese gels has been previously proposed as a means to improve the overall cheese yield. However, the potential of modifying whey protein aggregate properties to mitigate the impaired rennet gelation caused by native whey proteins, has not yet been properly studied. Therefore, in the second phase of the study, protein aggregates with a wide range of sizes were produced by heat and power ultrasound. The effects of size and hydrophobicity differences in the whey protein aggregates produced by heat and heat coupled with ultrasound were investigated in relation to the kinetics of rennet gelation and protein retention in model non-fat cheddar cheeses. Rheological measurements showed that sufficiently large, denatured whey protein aggregates could avoid impairment of rennet gelation caused by native whey proteins, irrespective of changes in the soluble calcium concentration or the surface hydrophobicity of the aggregates. Whey protein aggregates formed by the combined heat and ultrasound treatment were more hydrophobic than the larger heat-treated aggregates and were better retained in the cheese. However, inclusion of sufficiently large aggregates in cheeses milks conferred an openness to the cheese microstructure, and showed promise in terms of improving the otherwise rigid non-fat cheese microstructure. In the third phase of the study, the potential of power ultrasound to generate protein-stabilised water-in-oil-in-water double emulsions containing encapsulated whey protein was investigated as a means of incorporating whey proteins into cheese. Ultrasound was successfully applied to form whey protein-enriched water-in-oil-in-water double emulsions using minimal amounts of food-grade emulsifiers. These emulsions had a markedly higher rate of protein encapsulation than previously reported studies. The size distributions and protein encapsulation of the double emulsions could be tailored by manipulating the emulsion formulation and ultrasonic emulsification parameters. Whey protein-rich double emulsions were successfully incorporated into cooked curds formed by rennet gelation, without impairing gelation kinetics, to increase the retention of whey proteins that are otherwise lost during syneresis. In this body of work, fundamental understanding of the role of native and denatured whey proteins during rennet gelation was developed. The specific interactions among whey proteins, casein micelles and the rennet enzyme during the different stages of rennet gelation and cheese making were systematically studied. Further, power ultrasound was successfully used to formulate whey enriched double emulsions that can markedly improve the whey protein retention in cheeses. This understanding can be useful to implementing strategies to better incorporate whey protein into cheese matrices.

Pulmonary delivery has proven to be a promising delivery route for either local lung targeting or systemic delivery. A variety of particle systems such as polymeric particles and lipid-based particle systems have been developed as therapeutic delivery carriers for pulmonary delivery. However, the majority of current inhaled particles have limited retention time and low bioavailability in the target lung region, leading to suboptimal efficacy of therapeutic delivery and needing increased drug dosage or dose frequency, which could cause severe side effects. This is mainly due to the clearance and metabolic degradation mechanisms in the lungs. The mucociliary clearance and the alveolar macrophage clearance defend the airway and deep lung region, respectively, and are responsible for the elimination of inhaled particles. Therefore, it is important to understand the interactions of particles with the complex lung physiological environment in order to design more effective drug delivery carriers that can overcome various biological barriers or exploit the defence mechanisms to achieve improved biological outcomes. This PhD thesis focuses on engineering particle systems for pulmonary drug delivery, with specific aims of studying the interactions between inhalable particles and complex biological systems in the lungs including particle–mucus interactions and the role of a pulmonary corona in the uptake or clearance of particles by alveolar macrophages. Poly (ethylene glycol) (PEG) as a low-fouling material commonly used for ‘stealth’ modification of particles to reduce immune clearance was first investigated. The use of PEG building blocks with various architectures resulted in PEG-based particles with different structures and mechanical properties, which further affected the interactions of particles with proteins and immune cells in a complex biological environment (e.g., human blood). The particle–mucus interactions were then studied in the second part of this PhD research by comparing different polymer particles with potentially mucoadhesive and mucus-penetrating properties, obtaining a basic understanding of mucociliary clearance of particles in the lungs. The role of the pulmonary protein corona in alveolar macrophage clearance of polymer particles was then studied. The presence of a protein corona on particles resulted in increased or reduced macrophage uptake depending on the particle properties. When particles were transferred from one biological environment into another (e.g., blood to lungs), the interplay of protein coronas formed in each environment determined the composition of the eventual mixed protein corona and the subsequent particle–cell interactions. Finally, drug (structurally nanoengineered antimicrobial peptide polymers) loading and intracellular delivery using promising polyphenol-based carriers were investigated as potential antimicrobial therapies against lung infections (e.g., tuberculosis).

Acid whey and salty whey have presented a major disposal issue for the dairy industry. The processing of acid whey has proven challenging due to the presence of lactic acid and high levels of minerals, while salty whey is underutilized due to the high levels of salt. The treatment of these two types of whey will allow the production of high value products, including whey powders, lactose powder, and concentrated salt solutions for use in cheese salting or the chlor-alkali industry. The use of membrane technology has been studied for the treatment and processing of these whey streams. Electrodialysis has been shown to be effective for the treatment of acid whey since the process can achieve high removal of lactic acid and minerals when compared to pressure driven membrane processes, such as nanofiltration. However, electrodialysis is not used widely due to the high operating costs associated with membrane replacement and electrical consumption as a result of membrane fouling and poor process performance. In this thesis, the fouling of ion-exchange membranes during the electrodialysis of fresh sweet and acid whey was investigated. Although the fouling of ion-exchange membrane has been examined by many researchers, the feed solution was generally made using resolubilised powders. It has been demonstrated by other researchers that using fresh solutions provide results and outcomes closer to industrial applications. Furthermore, process optimization is a key parameter to reduce the cost of the treatment process. As a result, the effects of concentrate pH and applied current density were investigated to determine the optimum operating conditions that would minimize membrane fouling and enhance ion removal. Although membrane fouling occurred in all experiments, the effects on system performance were limited. Reductions in the current during pure sodium chloride circulation fell to a minimum of 80% of the original value after 5 hrs of whey processing. The use of an alkaline concentrate resulted in the strongest increase in system resistance, but the mineral deposits formed appeared to detach readily, thereby reducing these effects. The use of an acidic concentrate gave significantly greater rates of lactic acid removal, thus reducing the total membrane area required. A solution of hydrochloric acid with a pH of 1.0 was effective for in-situ cleaning of the mineral deposits. However, protein deposits were not readily removed when using the recommended base cleaning formula of 3% sodium chloride at a pH of 9.2. The concentrate stream in an electrodialysis process is considered a waste stream thus adding to the total volumes of waste generated by the dairy factory. Therefore, the use of salty whey permeate as the concentrate stream was investigated during the electrodialysis of sweet whey. The use of salty whey permeate is expected to reduce freshwater update and the volumes of wastewater generated from the treatment process. The type of concentrate (0.1M sodium chloride or salty whey permeate) did not affect the rate of sweet whey demineralization or the energy consumed per tonne of sweet whey processed, but less sodium and more divalent cations were removed when salty whey permeate was used. In addition, the use of electrodialysis for the demineralization of salty whey permeate was investigated to generate a lactose rich stream as the diluate stream and a concentrated salt solution as the concentrate stream. It was observed that salty whey permeate could be effectively demineralized using either 0.1M sodium chloride or a second stream of salty whey permeate as the concentrate stream. The concentrate purity could be enhanced by using monovalent selective membranes without increasing the energy consumption of the process (3.2 kWh per kg of sodium chloride removed from the diluate at 15 V across 2 cell pairs). Furthermore, combining different membrane technologies can assist in enhancing the treatment of acid whey to produce high quality whey powder. Three process combinations were examined at pilot scale, namely, (1) ultrafiltration and electrodialysis; (2) ultrafiltration, nanofiltration, and electrodialysis; and (3) ultrafiltration, dia-nanofiltration, and electrodialysis. Although all three combinations were successful in reducing the levels of lactic acid and minerals in acid whey, the lowest ratio between lactic acid and lactose (0.017 g lactic acid/g of lactose) was obtained with the process that utilized dia-nanofiltration. The energy required for the electrodialysis of the ultrafiltration permeate and dia-nanofiltration retentate were comparable (7.5 and 7.8 kWh/tonne of feed, respectively). However, the dia-nanofiltration retentate was at least 3.5 times more concentrated than the ultrafiltration permeate, thus reducing the annual energy consumption and capital investment of the electrodialysis unit. The product of the nanofiltration and electrodialysis process was successfully dried to produce a powder with an ash and moisture content of 4% and 2.5%, respectively. To further add value to the acid whey treatment process, the possibility of recovering lactic acid from a salt solution was investigated using either loose reverse osmosis membranes or an electrodialysis process. The recovered lactic acid could be reused in the cheese making process thus reducing fresh acid intake. Partial separation between lactic acid and salts was achieved at low applied pressures and feed pH in the reverse osmosis process, as a greater permeation of salts was observed under these conditions. Furthermore, lactic acid retention was enhanced by operating at room temperature with low feed pH. Partial separation between lactic acid and potassium chloride was also achieved in the electrodialysis process. However, the final concentration of potassium at 70% demineralization of the diluate stream was relatively high resulting in a low purity of lactic acid. Furthermore, the observed losses in lactic acid increased with the addition of sodium chloride to the feed solution. This indicates that the separation becomes more challenging as the complexity of the feed solution increases. Although electrodialysis has been widely studied for the treatment of sweet and acid whey, other electrically driven membrane processes, such as membrane capacitive deionization, have never been investigated. Three different pre-treated acid whey solutions were processed through a lab scale membrane capacitive deionization unit, namely, ultrafiltration permeate, nanofiltration retentate, and dia-nanofiltration retentate. Although a lowest demineralization rate was calculated for the nanofiltration retentate, a higher removal of lactic acid and cations was achieved when compared to the ultrafiltration permeate. Furthermore, similar molar concentration of ions were removed from the ultrafiltration permeate and dia-nanofiltration retentate (41 mEq/L and 43 mEq/L, respectively), however, the total energy consumption was lower for the dia-nanofiltration retentate (0.0122 Wh/mEq of cations removed Vs 0.0243 Wh/mEq of cations removed from the ultrafiltration permeate). Finally, it was found that the energy consumed for the treatment of acid whey ultrafiltered permeate using membrane capacitive deionization was comparable to the energy reported for the electrodialysis process. Overall, the results presented in this thesis have demonstrated that both acid whey and salty whey can be treated and transferred into valuable products for the dairy industry. Future work on this topic could include: (1) investigating the feasibility of using electrodialysis reversal for acid whey treatment; (2) performing a pilot scale assessment of the membrane capacitive deionization process; (3) undertaking an economic evaluation to justify using pressure driven membrane process prior to electrodialysis and membrane capacitive deionization processes; and (4) reassessing the possibility of recovering lactic acid from a salt solution by using other available technologies/ processes such as electrodeionization and selective crystallization of lactic acid.

Nanostructured hybrid organic-inorganic materials are a unique class of materials showing distinctive properties that have attracted high interest due to their diverse applications in the fields of energy, environment and medicine. In particular, hybrid materials are promising candidates for sensing applications due to the tunable chemical, structural and functional properties of the organic and inorganic components. Hence, the engineering of novel nanostructured catalytic organic/inorganic materials provides opportunities for the fabrication of advanced nanodevices for biosensing. In this thesis, novel hybrid materials have been prepared and their electrocatalytic, catalytic, and optical properties explored. First, nanostructured electrocatalytic microparticles were synthesized in mild conditions and used with an organic binding agent to prepare carbon electrodes applied in the detection of glucose in biologically relevant media. Second, hierarchically structured hybrid particles displaying enzyme-like catalytic activities were synthesized and used to prepare high-throughput micro-reactors for the detection of bioanalytes via a hybrid organic-inorganic cascade reaction. Finally, a natural occurring polysaccharidic nanoparticle, i.e. glycogen, was engineered to impart adhesive functional properties to a hybrid film and used for the coating of various substrates with different chemical composition. These hybrid coatings embedding metal nanoparticles were employed as catalytic and optically active functional interfaces.

Inner ear disease is the leading cause of hearing impairment in developed countries. An estimated 466 million people suffer from hearing loss worldwide and this number is on the rise. Sensorineural hearing loss (SNHL) is the most common form of hearing impairment and is characterised by the degeneration of key structures of the sensory pathway in the cochlea of the inner ear (the cochlea) such as the sensory hair cells, the primary auditory neurons and their synaptic connection to the hair cells. Current research focuses on developing techniques to administer growth proteins such as neurotrophins to repair or regenerate damaged auditory neurons, as well as preventing loss of primary auditory neurons. Drug delivery systems are being developed to treat SNHL, such as cell-based drug delivery systems and gene vectors, however nanoengineered systems show promise to address the specific needs of neurotrophin-based therapies such as safety, high dosing and long-term delivery to the cochlea. Research carried out in this thesis has developed this technology further by the scale-up production of nanoengineered silica-based supraparticles (SPs) (~550 micrometers) with high porosity) and the development of several strategies towards their application as viable drug delivery platforms for achieving sustained drug release in the inner ear, as detailed in Chapters 3-6. In Chapter 3, a gel-mediated electrospray technique was developed to synthesise silica supraparticles (Si-SPs) in high yields. The Si-SPs were assembled from different primary silica particles i.e., particles with no pores, small pores (2-3 nm) and bimodal large pores (2-3 nm and 15-64 nm). A high loading of fluorescently labelled model protein (fluorescein isothiocyanate (FITC)-lysozyme) and neurotrophic factor (a drug for the treatment of inner ear disease) in the Si-SPs was possible and the resulting particle system could achieve sustained drug release for over 150 days. The findings demonstrate that gel-mediated electrospray is a robust and automatable technology to produce Si-SPs, which is a promising platform for clinical translation and commercialisation. In Chapter 4, the pharmacokinetics of the neurotrophin brain-derived neurotrophic factor (BDNF) from Si-SPs was examined as engineering drug delivery systems with well-defined pharmacokinetics is important for clinical translation. BDNF-loaded Si-SPs were surgically implanted either directly into the cochlea, or onto a semi-permeable membrane (the round window membrane; RWM) that is a boundary between the middle and inner ear. Treatment duration was for either 3 or 7 days whereby the fluids from the cochleae were sampled and tested for BDNF levels. The results showed that the BDNF released from the Si-SPs was detected in the cochlear fluids indicating that the approach has potential as a clinically relevant neurotrophin delivery strategy to treat people with hearing impairment. In Chapter 5, a bioengineering coating strategy was developed for retarding the initial burst release of neurotrophins from the Si-SPs. Applying a fibrin coating on the surface of the Si-SPs and embedding the fibrin-coated Si-SPs within an alginate CaCO3 hydrogel both slowed the initial burst release to improve the drug release kinetics. The results demonstrate the suitability of alginate CaCO3 hydrogel systems for surgical handing of the Si-SP system. In Chapter 6, a chitosan and an alginate layer-by-layer coating on the Si-SPs was developed as an alternative strategy for delaying the initial burst release of uncoated Si-SPs. Chitosan and alginate are two biocompatible polysaccharides that interact electrostatically via the carboxyl groups from alginate (negatively charged) and the amine groups on chitosan (positively charged). By varying the layer number and hence, the thickness of the coating, different release profiles were attained. In vitro neurotrophins release profiles showed that chitosan-alginate-coated silica supraparticles ((Chi/Alg)Si-SPs) experienced a delayed initial burst release. Spiral ganglion neurons culture and neurite length analysis indicated that the neurotrophins released from (Chi/Alg)Si-SPs had maintained biological activity. Functional hearing was tested using auditory brainstem responses (ABRs) to determine the safety profile of surgical delivery of coated SPs to the inner ear. Hearing thresholds were maintained within the normal range following RWM, however an increase in thresholds for high frequency sounds were observed following implantation of (Chi/Alg)Si-SPs into the cochlea. Scanning electron microscopy images of (Chi/Alg)Si-SPs collected following in vivo implantation along with a commercial viable fibrin sealant indicated the biodegradability of (Chi/Alg)Si-SPs post-implantation. These results indicate that (Chi/Alg)Si-SPs can potentially be used as a clinically applicable platform for sustained inner ear neurotrophin delivery. In summary, this thesis expands knowledge in the development and engineering of Si-SPs in addressing key neurotrophin delivery issues for the treatment of hearing loss, including high yield, sustained drug release, well-defined pharmacokinetics and biodegradability.

Cytochrome P450 enzymes are promising biocatalysts for the pharmaceutical industry due to their ability to catalyse selective oxidations of drug-like substrates that are often difficult to achieve with synthetic organic chemistry. This enzyme family is, however, currently underrepresented in industrial biocatalysis due to issues with enzyme stability and cofactor requirements, as well as the poor aqueous solubility of many cytochrome P450 substrates. In this thesis, cytochrome P450-mediated biotransformation of the opium poppy alkaloid noscapine was used as a model reaction for the development and optimisation of a biotransformation process using bacterial cells. The aims of this thesis were to develop a biocatalytic method for the production of noscapine metabolites, which are of interest for the development of anticancer drugs, and to investigate strategies for the implementation of cytochrome P450s in pharmaceutical manufacturing processes. In order to obtain a suitable P450 enzyme for the biotransformation of noscapine, a library of mutants of the enzyme P450BM3 from Bacillus megaterium was constructed using site-directed mutagenesis. This strategy identified enzyme mutants capable of producing six different noscapine metabolites, several of which may be of interest for drug development. A high selectivity for noscapine N-demethylation was demonstrated by several mutants. This N-demethylation reaction was therefore selected as a focus for further process development in a whole-cell biotransformation system. The permeability of Gram-negative and Gram-positive bacteria to noscapine was assessed in order to minimise mass transfer limitations in a whole-cell biotransformation system. This identified Bacillus megaterium as a suitable expression host, as these cells enabled significantly faster reaction rates compared to the Gram-negative Escherichia coli. As noscapine has poor aqueous solubility, two strategies were investigated to improve substrate loading for the biotransformation. First, a biphasic biotransformation system, which uses a water-immiscible organic solvent phase to supply substrate, was developed with the use of in silico solvent screening based on Hansen solubility parameters. This was effective in increasing product yields but had the disadvantage of slower reaction rates due to substrate partitioning. The second approach used a cyclodextrin/polymer system to enhance aqueous noscapine solubility and proved more effective, enabling faster reaction rates. However, the stability of the mutant P450BM3 enzyme was found to be a major factor limiting product yield. The use of growing cells rather than resting cells was then investigated to address the limited enzyme stability observed during biotransformation. As intracellular cofactor supply can often limit reaction rates in growing cells, a metabolic engineering strategy based on the CRISPR/dCas9 interference (CRISPRi) system was used to redirect carbon flux towards cofactor production in the B. megaterium host. This provided a significant improvement in productivity compared to cells lacking the CRISPRi system, though dynamic control of gene repression was not possible due to ‘leaky’ expression of dCas9 from an inducible promoter. The CRISPRi-enhanced system with growing cells showed comparable productivity to the resting-cell system, with the advantage of sustained P450BM3 regeneration and fewer processing steps. In summary, an enzyme and host cell system for the biocatalytic N-demethylation of noscapine was developed. A number of process limitations were identified and addressed to optimise the yield and productivity of the system. The limitations that were encountered during process development are common to whole-cell biotransformations using P450 enzymes, and the strategies used to address these limitations are likely to be useful for the development of other P450-catalysed biotransformation processes.

This thesis introduces and benchmarks 2 novel AFM based techniques to characterise the interfacial rheology of polymeric fluid-fluid interfaces. We develop and benchmark a time-dependent extension of capsule compression to model oscillatory indentations, determining the interfacial rheology of capsules from thin shells to thick shells. This is first tested and benchmarked using the food-based emulsifier beta-lactoglobulin at the MCT-Oil interface. This system is then crosslinked and the behavioural changes in the capsule oscillatory rheology, bicone shear rheology and pendant drop dilational rheology are tested. This investigation has shown that there is a significant frequency dependent response of beta-lactoglobulin capsules to an oscillatory indentation in both the native and the crosslinked state. We also show that there is a large indentation depth dependence on the time dependent dwell behaviour. When comparing these behaviours to the bicone shear rheology and pendant drop dilational rheology, the magnitude of the moduli at similar frequencies differs strongly between all techniques. Comparatively, the tan(delta) between the bicone and capsule measurements exhibits similar magnitudes for both the native and crosslinked cases. An extension of the capsule technique was made to include thicker interfaces, where the thickness of the interface is more than 5% of the overall capsule radius. This model was tested with the food hydrocolloid chitosan, which in acetate buffer concentrations above 0.3M makes particulates and becomes surface active, forming a thick film at the interface. We determine the properties of this film by assessing the film mechanically and visually. To mechanically characterise the film, we utilize bicone interfacial shear rheology, and both linear and oscillatory capsule compression. The film exhibits characteristic particle like behaviours and rapidly heals when broken in shear. Similar self-healing characteristics are then observed in the linear capsule compression, and the oscillatory capsule compression shows similar moduli and behaviour to that observed by the bicone. The visual characterisation was done with in-situ aqueous imaging and dry Langmuir-Blodgett imaging of a transferred film, where a thick, continuous, particulate film is observed. We continue to push the lengthscale of interfacial rheology by developing a novel method of observing the interfacial rheology at a fluid/fluid interface. This method consists of placing a cantilever tip at the interface between two fluids, laterally translating the tip and observing the torsional force on the cantilever. This force can then be transformed into an interfacial viscosity with the knowledge of the diameter of the tip at the interface. We apply this technique to two types of interfaces, linear polymeric interfaces and a globular protein interface. At the air/water interface, we test the linear polymers PEO and PSS, where we observe that shorter chain length PSS shows a reduction in the interfacial viscosity. beta-lactoglobulin is observed over time at the decane/water interface, where the interfacial viscosity increases over time, and non-linear strain dependent behaviour is observed.