Supramolecular assembly provides a versatile pathway for engineering bespoke materials, such as metal–organic hybrid materials. Metal–phenolic networks (MPNs), constructed from the coordination-driven assembly of phenolic ligands and metal ions, are an emerging class of hybrid materials with a rich choice of building blocks. Due to their strong adhesion to different substrates (particles, planar surfaces, microorganisms), high degree of modularity, and tuneable degradability, MPNs have garnered considerable attention in fields such as drug delivery, bioimaging, antimicrobials, separation, and catalysis. However, fundamental research in the material aspects of MPNs and how these influence biomedical applications are essential yet overlooked. This thesis explores the fundamental principles of MPNs and uses this insight to examine MPN materials in a range of biomedical applications. First, MPN microcapsules comprised of various building blocks are engineered, and their programmable gating mechanisms are explored in terms of intermolecular dynamics. This fundamental study not only provides insight into the dynamic nature of MPNs but also offers a route to engineer smart delivery systems and selective gating materials. Next, MPN coatings are used as a versatile and cytocompatible platform to trigger the endosomal escape of nanoparticles, which has been regarded as a key bottleneck for the intracellular delivery of therapeutics. The escape mechanism is systematically investigated and determined to be the “proton-sponge effect”, arising from the buffering capacity of MPNs. Notably, this buffering-enabled escape capability is preserved after the post-functionalization of MPN coatings with polymers, showing the generalizable nature of the platform. Therefore, a subsequent in-depth exploration of the buffering effects of MPNs sheds mechanistic insight into metal–organic systems and their emergent buffering capacity based on coordination dynamics and building block choice. Finally, the advantages of different polyphenol-enabled supramolecular networks are integrated to expand the MPN platform from thin films to self-assembled nanoparticles. Bioactive metal–phenolic nanoparticles are developed via robust and template-free assembly. Hydrophobic interactions and coordination play dominant roles in the assembly and stabilization of the nanoparticles. Furthermore, the incorporation of diverse biomacromolecules (e.g., functional proteins and genes) during assembly enables the potential use of these metal–phenolic nanoparticles in various biomedical applications, anticancer treatments, cascade reactions, and gene knockdown.

Traditional solvent extraction processes typically rely on volatile organic compound (VOC) solvents from petrochemicals to extract desired compounds. Due to their unsustainable nature, as well as other regulatory, economic, environmental and health-related concerns, pressure to replace VOCs with green alternatives is increasing, and, in addition, a number of new sustainable bio-based solvents have become available. This thesis examines solvent selection techniques and their usefulness in identifying new solvents for a given process. This includes consideration of the ‘green-ness’ of a solvent and the process it is used in. Bio-based molecular solvents offer a promising alternative to VOCs, and key solvents of interest are discussed. One such industry investigating the replacement of VOCs with bio-based solvents is the natural products industry. In this work, the extraction of natural pharmaceutical products (e.g., alkaloids) from poppy straw using petrochemical based solvents at Sun Pharmaceutical Industries, Ltd. was investigated as an application for bio-based alternative solvent systems. The present research aimed to evaluate a range of alternative solvent systems for extraction of model compounds and alkaloids, including bio-based solvents and extractants. This required investigation of solvent selection tools including green solvent guides, Hansen solubility parameters (HSP), and COnductor-like Screening Models (COSMO) which can be used for a range of extraction applications as well as new and evolving solvent alternatives. The strengths and challenges of these selection tools were provided in the context of phenol extraction, with results evaluated for the ability to predict distribution in the two-phase system. The solubility predictions indicated that in considering quantitative accuracy of distribution results, COSMO-Segment Activity Coefficient (SAC) with directional hydrogen bond approach was the best option, as the distribution coefficients were closer to those determined experimentally. For qualitative predictions, the 2010 COSMO-SAC model implemented in MATLAB was superior for the present application, as it was the only theoretical solubility tool which correctly predicted the experimental solvent ranking. However as a qualitative tool, HSP predictions had minimal deviations from the experimental trends, with only one or two deviations to the solvent order depending on methodology of calculation. Physical properties and equilibrium distribution data encouraged the use of Cyanex 923 as an extractant, and this was studied in detail using bio-based solvent d-limonene and a traditional solvent, xylene, as diluents. Finally, to test the performance of the extractant molecule at pilot scale, Karr reciprocating column studies were completed. A range of operating conditions were studied to observe the effect of flowrate and agitation on hydrodynamics and mass transfer performance for these new solvents. Dispersed phase holdup, drop size and mass transfer coefficients were calculated, and correlations to predict these were developed to enable the performance of the new solvents in existing equipment to be predicted. The results support the use of Cyanex 923 as an extractant in the application of Karr column extraction of phenol and morphine from natural products.

Membrane gas separation is now a mature technology for a number of applications such as hydrogen (H2) recovery from ammonia purge gas streams, natural gas sweetening for carbon dioxide (CO2) and hydrogen sulphide (H2S) removal, and air separation for nitrogen (N2) enrichment. There are numerous other applications for which membrane gas separation technology holds promise. Mathematical models of membrane gas separation enable the viability, separation performance and membrane system characteristics to be determined for emerging gas separation applications, hence these models are very important in engineering membrane-based processes. A significant number of membrane gas separation models have been developed in the academic literature, but most cannot be used for a broad range of gas separation applications, as these models are unable to consider the influences of some of the most important challenges of membrane gas separation. These challenges either originate from the intrinsic properties of membranes or are imposed by gas separation operating conditions and adversely affect the separation performance. Consequently, any models neglecting such challenges are prone to generate erroneous results. Chapters 4 and 5 of this PhD thesis investigate the influences of the Joule-Thomson effect, real gas behaviour, pressure loss, concentration polarisation, plasticization, and competitive sorption using a rigorous membrane gas separation model. For this purpose, a mathematical model of gas separation in polymeric hollow-fibre membrane modules is developed under steady-state conditions. The rigorous membrane model is compared with a simplistic one, similar to conventional membrane gas separation models, to highlight the need to account for these challenges. For a pre-combustion CO2 capture case study (CO2/H2 separation), negligible deviations between the rigorous and simplistic models are reported as the difference in the activation energies of permeation of CO2 and H2 increases, meaning the deviations are not noticeable as H2/CO2 selectivity decreases. As this selectivity increases, 2-5% and 0.5-8% differences are calculated in CO2 concentrations on the feed and permeate sides, respectively. H2-selective and CO2-selective membranes are chosen for further analysis. With 60% CO2 in the feed gas at 100C and 60 bar, deviations as large as 20% are calculated between the rigorous and simplistic models in terms of permeate flow rate and purity. As the H2 fugacity coefficient is greater than unity, real gas behaviour results in a greater driving force for permeation for the H2-selective membrane. On the contrary, fugacity coefficients of CO2 are less than unity, meaning CO2 permeation is reduced due to a negative departure from ideal gas behaviour. A further analysis shows concentration polarisation effects are negligible even at extremely large stage-cuts within the feed flow rate range studied. Verified by H2 and CO2 Joule-Thomson coefficients, retentate heating and cooling are observed for the H2-selective and CO2-selective membranes, respectively. However, due to the greater Joule-Thomson coefficient of CO2, the extent of cooling is larger than that of heating, meaning that the Joule-Thomson effect is more pronounced for the CO2-selective membrane. It is concluded that deviations between the rigorous and simplistic models generally grow at high CO2 concentrations, low feed temperatures, high feed pressures and high stage-cuts, as real gas behaviour and Joule-Thomson effects are intensified. For a biogas upgrading case study, (CO2/CH4/water vapour (H2O) separation), the rigorous membrane model is modified to account for competitive sorption, plasticization, and component blocking effects in glassy polymers. A fugacity-dependent permeability model is developed that considers the effects of plasticization by CO2 and component blocking by H2O. The membrane model performance using this fugacity-dependent permeability is compared with two simpler permeability scenarios: pure gas and constant permeabilities. Results show a 30% difference in retentate composition between the pure gas and fugacity-dependent permeability scenarios. Using the fugacity-dependent permeability, higher CH4 recoveries are predicted relative to the pure gas permeability mode, but this comes at the expense of less CO2 removal. At the process conditions studied, one analysis shows the influence of concentration polarisation on CH4 recovery and CO2 removal is the smallest of all effects, even in the presence of highly permeable water vapour. Real gas behaviour impacts CO2 removal but has no effect on CH4 recovery. A comparison shows competitive sorption, plasticization and component blocking effects have greater impact on CH4 recovery and CO2 removal than real gas behaviour. In the penultimate chapter, the rigorous membrane model is exported to Aspen Hysys and Aspen Plus to simulate two- and three-stage membrane processes for Xenon (Xe) and Krypton (Kr) separation at sub-ambient temperatures provided by a simple propane refrigeration loop. Membrane materials with suitable separation properties are identified to yield 99 mol% or more of both components. For the two-stage membrane process, Xe/Kr selectivity must be greater than 56 and 44 at pressure ratios 10 and 15, respectively, to fulfill this separation target. The three-stage membrane process shows higher potential than the two-stage process, as it needs moderate Xe/Kr selectivity equal to or greater than 24 at a pressure ratio 5 and 16 at pressure ratios 10 and 15 to ensure at least 99 mol% purity of both components. Energy demand of the three-stage membrane process is calculated and compared with that of cryogenic distillation-a commercial technology to separate Xe from Kr. Results show the membrane technology saves at least 20% energy. The present thesis’s outcomes indicate the importance of the inclusion of non-ideal phenomena in rigorously evaluating membrane separation performance. The thesis presents a model considering non-ideal effects in membrane gas separation and shows idealistic membrane models may result in significant errors in the simulation of membrane systems at industrially relevant conditions. Other than accounting for the non-ideal effects, the membrane model developed has two distinct advantages over most conventional membrane gas separation models. First, the model uses an equation-oriented programming language requiring much less effort to develop a computer code compared with an algorithm-based programming approach. Specifically, this becomes a significant advantage when gas separation with membranes is not limited to mass transport phenomenon only and complex energy and momentum transport phenomena affect the separation properties of the membranes. At such conditions, mass, energy, and momentum balance equations should be solved simultaneously and computer coding of these equations using the algorithm-based approach is not as easy as that of the equation-oriented method. Second, unlike most conventional membrane gas separation models, the model developed in the present thesis is easily interfaced with commercial process simulators, enabling the simulation of a broad range of single- or multi-stage membrane processes.

Climate change resulting from the emission of greenhouse gases, in particular CO2 into the atmosphere has become a global challenge in recent years. Post-combustion CO2 capture is an effective way to mitigate CO2 emissions from power plants. Gas-solvent hollow fibre membrane contactors have been introduced as a promising technology for post-combustion CO2 capture, as contactors combine the benefits of membrane technology with chemical absorption. Although the membrane acts as a permeable barrier adding further mass transfer resistance approximately by 40% to the contactor system, the membrane module provides 10-fold reduction in equipment size and 30 times higher interfacial area than that of gas absorbers. Therefore, there are three main resistances to mass transfer in gas-solvent membrane contactor systems; namely the membrane (porous and non-porous layers), solvent phase boundary layer and gas phase boundary layer. To reduce the membrane mass transfer resistance, ultra-thin film hollow fibre membranes have been utilised. However, the solvent phase boundary layer and gas phase boundary layer resistances remain considerable, and therefore techniques are required that will reduce these resistances and enhance membrane contactor performance. In this thesis, solutions to overcome the mass transfer challenges of gas-solvent membrane contactor systems for CO2 capture from simulated flue gas have been investigated. Inducing additional mixing to the solvent and gas phase boundary layers is one of the most effective methods for improving mass transfer. To generate mixing and changing the hydrodynamics of the solvent and gas flow, oscillating and vibrating conditions have been designed and integrated into gas-solvent membrane contactor systems. Oscillating the flow of either the solvent or gas phases induces additional mixing within the respective boundary layers, while applying vibration to the membrane module induces mixing across all stages of mass transfer. The effect of these new techniques on the solvent and gas phase mass transfer coefficients have been investigated and reported here. The solvent side mass transfer coefficient increased by up to 50% and 13% under oscillating and vibrating conditions, respectively. The reason for these enhancements was due to disruption to the solvent phase’s boundary layer, resulting in additional mixing that produced a higher CO2 flux through the solvent phase. Moreover, the gas phase mass transfer coefficient increased by 28% on average under oscillating flow due to an increase in gas phase mixing within the lumen side of the membrane fibres. Vibrational conditions had a stronger impact, enhancing the gas phase mass transfer coefficient by a factor of two. The frequency and amplitude have been presented as the key adjustable parameters of the oscillating operating modes. For the vibrating unit, oscillating and displacement amplitude have been contributed to the definition of operating conditions. The mass transfer performance on the gas and solvent phases of membrane contactor systems has been investigated for variable oscillating/vibrating frequencies and amplitudes. Oscillating solvent flow frequency has been investigated between 0.83 and 2.03Hz, while the vibrating unit worked within the range of 0.5 to 2.3Hz. The solvent boundary layer thickness has been found to reduce with increasing oscillating/vibrating solvent flow frequency. The enhancement in the gas phase was strongly linked with oscillating/displacement amplitude, associated with the magnitude of pressure wave propagating through the gas phase. The present thesis has introduced novel experimental apparatuses and mass transfer analysis methods that add to our understanding of mass transfer phenomena in gas-solvent membrane contactors. Empirical correlation models, in terms of the dimensionless Sherwood number, have been proposed for the first time to predict the mass transfer performance of gas-solvent membrane contactor systems under both oscillating and vibrating conditions. To indicate the effect of oscillating parameters on the Sherwood number correlations, the frequency and amplitude have been incorporated into the definition of root mean squared velocity of oscillating solvent/gas flow resulting in oscillating Reynolds number. The effect of vibrational parameters has been presented in the form of dimensionless vibration number incorporated within the new vibrating Sherwood number correlation of the gas-solvent membrane contactor system. The novel Sherwood number correlations have enabled the influence of oscillation and vibration on the solvent and gas boundary layers (e.g. the boundary layer thickness) to be quantified. Overall, the aim of this thesis was to improve the mass transfer performance of gas-solvent membrane contactors for CO2 capture. This thesis demonstrated that the application of oscillating solvent flow, oscillating gas flow and vibrating solvent/gas flow are three novel approaches that achieve higher CO2 flux and mass transfer within membrane contactors than possible under conventional flow regimes. Furthermore, the application of oscillating and vibrational conditions to larger scale membrane modules is readily achievable, therefore the methods developed here show great potential for being used on industrial scales.

Vaccines are an effective tool for preventing and controlling various diseases by inducing adaptive immunity. Nanomaterials play an important role in vaccine development. Micro- and nanocarriers can be engineered to improve the therapeutic efficacy of vaccines by (i) preventing the degradation and systemic clearance of vaccine antigens and (ii) facilitating the uptake of vaccines in antigen-presenting cells (iii) co-delivering adjuvants and antigens at desired intracellular compartments for optimal immunotherapy. However, it is important to engineer a carrier that is both effective and safe. Micro- and nanoparticles based on DNA have shown great potential for biological applications, owing to the programmable sequences, predictable interactions, versatile modification sites, and high biocompatibility of DNA strands. This thesis aims to develop facile strategies to synthesize DNA particles for vaccine delivery by self-assembly approaches. First, a simple strategy to synthesize DNA microcapsules is reported. The cytosine-phosphate-guanosine oligodeoxynucleotides (CpG) motif is an efficient vaccine adjuvant that can effectively stimulate the immune system to secrete cytokines. By loading and crosslinking Y-shaped DNA building blocks (containing CpG motifs) into sacrificial calcium carbonate templates, monodisperse and spherical DNA capsules were obtained. These DNA microcapsules were internalized into cells efficiently, accumulated in endosomes, and induced immune cells to secrete high-level of cytokines. Next, we developed a template-assisted and versatile approach for synthesizing a new set of multifunctional particles through the supramolecular assembly of tannic acid (TA) and DNA molecules. Uniform and stable DNA-TA particles with different morphologies could be easily synthesized by using different types of DNA strands. Intriguingly, different DNA sequences can be encoded into this DNA-TA particle for applications in immunotherapy or gene delivery. The incorporation of CpG motifs and ovalbumin into the particles allows the intracellular antigen/adjuvant co-delivery to amplify cytokines production in macrophages, through synergistic effects. In addition, green fluorescent protein (GFP)-expressing plasmid DNA could be transfected by using the DNA-TA particles in HEK293T cells. Finally, nanometer-sized particles were engineered by exploiting the one-pot supramolecular assembly of TA, DNA, and PEG for intracellular delivery of CpG motifs. TA-DNA-PEG nanoparticles with different sizes could be fabricated by adding different molecular weight PEG chains. TA-DNA nanoparticles with tunable size were also synthesized by varying the molar ratio of TA and DNA. The obtained nanoparticles can enhance the cellular uptake of CpG oligonucleotides and consequently the production of cytokines in macrophages. Overall, the engineered DNA-based particles have potential for co-delivering nucleic acids and protein antigens in immune cells to enhance the immunological response against infectious diseases and cancer.

Metal–phenolic networks (MPNs), which are made using metal ions and phenolic ligands, have attracted widespread interest owing to their hybrid physicochemical properties and high affinity to diverse substrates. The combination of MPNs with functional materials can lead to MPN composites capable of outperforming the individual components in a wide range of emerging applications. However, the small pore size of MPNs limits the possibilities of loading MPN-based materials with other functional components. In addition, the controlled assembly of MPN composites for functional thin films with tailored structures has been largely unexplored. This thesis focuses on the engineering of the composition and structure of MPN composites for various applications. Firstly, a supramolecular fluorescent labeling strategy was developed using luminescent MPN composites that consist of a MPN and commercially available dyes. To demonstrate the versatility of this strategy, 16 types of particle substrates that are formed from different materials, and have different sizes and surface chemistries, were successfully labeled. This strategy obviated the need to covalently conjugate the dyes or to modify the surface chemistry of the substrates. In addition, customized luminescence regimes (e.g., red, blue, multichromatic, and white light) were readily achieved using common fluorophores. The fluorescent coating is stable in many biological environments, such as in serum and the cytosol, which demonstrates its potential to study the cell association and internalization of particles in real-time. Secondly, a cubosome templating strategy was developed to prepare ordered mesoporous MPN particles with uniformly large pores (around 40 nm). The large mesopores allow various cargos (e.g., biomacromolecules) to diffuse into the particles while the phenolic groups stabilize the cargos. This led to considerably higher loading amounts than those typically achieved when using commercially available SiO2 with 50 nm pores. In addition, meso-MPN particles that are loaded with enzymes acted as highly efficient bioreactors, displaying catalytic activities that exceed those prepared from porous silica particles. Thirdly, cubosome templates were also engineered in the form of monoliths using diffusion-induced self-assembly, and the formation mechanism and precise molecular organization of the monoliths were investigated by both experiments and all-atom molecular dynamics simulations. The large pores of the polymer monoliths were then used to synthesize ordered MPN-based monoliths. These results show the significant potential of using MPN composites in various fields, including chemistry, biology, and materials science.

Metal-phenolic chemistry, owing to its facile and versatile functions to manipulate dynamic interactions (e.g., interfacial adhesion and coordination crosslinking) at the molecular level, has emerged as a powerful tool for the rational design and engineering of hybrid metal-organic materials. In particular, leveraging a myriad of advantages, including a rich choice of building blocks, the combined properties of phenolics and metals, and dynamic coordination bonds, metal-phenolic network (MPN) particles and films have been explored in diverse fields such as drug delivery, functional elastomers, and water treatment. Despite significant progress, the mechanisms related to the kinetics (e.g., assembly process) and thermodynamics of MPNs are poorly understood, resulting in less control over the physicochemical properties of MPN particles and films. This thesis (1) reviews the fundamental insights about metal-phenolic interactions underpinning the dynamic nature of coordination bonds and universal adhesion to surfaces; (2) develops various assembly strategies, including oxidation-mediated assembly, enzyme-mediated assembly, and spray assembly, to endow MPN particles and films with controllable properties (e.g., thickness, pore size, roughness, and wettability) and functionalities (e.g., fluorescence, catalysis, radical scavenging, and UV-shielding), thereby (3) expanding the applications of MPN particles and films, specifically in gas separations, biomineralization, biomolecule conjugation, and oil-water separations; and (4) discusses the underlying mechanism governing the assembly process and resultant coordination states of MPNs by various characterization methods. Some unsolved challenges and perspectives related to MPN chemistry are also highlighted. This thesis provides insightful perspectives into the chemistry of MPN assembly and other metal-organic coordination complexes.

The clinical success of monoclonal antibody therapy has inspired research in understanding the fundamental molecular basis of antibody-antigen interactions and the engineering of antibodies with enhanced or novel properties. With the emergence of nanomedicine, antibodies have been widely applied as targeting ligands decorated on the surface of therapeutic nanostructured modalities – including liposomes, protein nanoparticles, and polymeric assemblies – for drug delivery and imaging applications. However, little is known about how antibodies assembled in a cluster or particulate form interact with antigens in a biological system, largely due to the challenge in preparing ‘pure’ antibody assemblies with controlled physicochemical properties. In this thesis, a mesoporous silica template-mediated assembly platform was applied to fabricate well-defined nano-assemblies of therapeutic antibodies, including conventional monoclonal antibodies and antibody-drug conjugates. The antibody nano-assemblies (AbNAs), crosslinked with poly(ethylene glycol)-N-hydroxysuccinimide (PEG-NHS), preserved the selectivity of the monoclonal antibody and induced receptor-mediated internalization of antibodies to achieve enhanced intracellular response, such as growth inhibition. This strategy presents opportunities for intracellular delivery of monoclonal antibodies, as well as a versatile platform for fundamental studies on the interactions between antibody assemblies and cells. Facile engineering of AbNAs can be achieved by leveraging the intrinsic property of the PEG crosslinker, such as chain architecture (PEG arm numbers and arm length), to regulate bio-nano interactions. As a widely recognized stealth material, PEG can prolong blood circulation time to allow the accumulation of nanoparticles in target tissues, however, it could also result in decreased targeting efficacy by blocking the antigen-binding sites. This thesis investigates the influence of PEG crosslinking, specifically the effect of using PEG crosslinkers with different chain architecture on the formation of AbNAs and their bio-interaction with respect to specific binding and uptake by phagocytic cells. PEG crosslinkers with less arms but longer arm length were found to be more beneficial for AbNAs to achieve both minimal phagocytic capture and optimal targeting. Furthermore, the targeting efficacy of AbNAs could be enhanced by substituting conventional monoclonal antibodies with engineered antibody fragments. Nanobodies, also known as single-domain antibodies (sdAb), are the smallest antigen-binding unit (12-15 kDa) that solely bind to the target antigen. The unique structure of nanobodies offers several desirable features, including small size, high stability, strong antigen-binding affinity and low immunogenicity, which makes nanobodies superior for antibody nano-assembly engineering. The nanobody nano-assemblies (NanoNAs), prepared via the template-mediated assembly platform, exhibited significantly enhanced selective association to target cells and reduced phagocytic association in comparison with full-sized AbNAs, owing to the unique structure of nanobodies that allowed a large amount of active binding sites to be presented on the particle surfaces and eliminated crystallisable fragment (Fc) receptor-mediated capture by phagocytic cells. Overall, the versatile antibody nano-assembly systems expanded our understanding of antibody-antigen interactions, and provides a facile platform to engineer antibody assemblies with novel or enhanced properties for biomedical applications.

The goals of this thesis were to explore the impact of architectural design on the potency of antimicrobial polypeptide polymers. To this aim, we firstly investigated the synthesis and kinetics of flexible polymeric macroinitiators obtained via photo reversible addition-fragmentation chain-transfer polymerization (RAFT) for the ring opening polymerization of N-carboxyanhydrides. This study revealed the bottlebrush polymers benefit from a cooperative folding of alpha helices in chlorinated solvent, resulting in fast kinetic rates. Furthermore, free primary amine and trimethylsilylated macroinitiators were shown to offer similar initiation efficiency, kinetics, and polypeptide control, which allowed the selection of appropriate initiators and polymerization conditions for future applications. Based on this synthetic knowledge, amphipathic bottlebrush polypeptide polymers, termed brush Structurally Nanoengineered Antimicrobial Peptide Polymers (brush SNAPPs) were synthesized. Their direct antibacterial activity was evaluated against Gram-positive and Gram-negative pathogens by means of in vitro assays. The outcome of the structure activity investigation revealed the bottlebrush morphology is indeed antibacterial, with short bottlebrushes of backbone degree of polymerization of 16 behaving similarly to star shaped SNAPPs and displaying preference for Gram-positive bacteria. In contrast, a brush SNAPP with longer backbone length of 190 displayed greater efficacy when challenged with Gram-negative Acinetobacter baumannii. These results, combined with the reduced cytotoxicity of the brush SNAPP architecture provide guidance for the treatment of Gram-negative infections. Aiming to further reduce the observed mammalian toxicity of polypeptides, we next focused on PEGylation as a means to improve SNAPP biocompatibility. Three different avenues for the polymerization of brush poly(ethylene glycol) methyl ether acrylate (PEGA) with high chain end fidelity were investigated. Blue LED-activated RAFT polymerization yielded remarkable alpha- and omega- group retention compared to thermal and UV activated RAFT methods. This technique was applied toward the synthesis of discrete polypeptide nanogels comprised of a comb-brush PEGA and bioactive polypeptide corona, aiming to selectively target Gram-positive pathogens. Antibacterial evaluation of the SNAPP nanogel hybrid against Escherichia coli and Staphylococcus aureus demonstrated this architectural design furnishes bacteriostatic and bactericidal properties, with selective targeting of S. aureus whilst also achieving an improved therapeutic profile and antifouling protection. Together, these results have generated important considerations for the future design of antimicrobial polypeptide therapeutic agents of varied architecture.

Small particles moving near a wall experience particle-scale inertial lift forces in a direction normal to the wall and hence, migrate away from the wall. In addition, particles in suspensions experience hydrodynamic collision forces and migrate away from or towards the wall. These forces are critical in the biological context as they contribute to the separation between platelets and red blood cells that ensure the repair and integrity of blood vessels. These migration mechanisms have also been utilised to design and optimise micro-scale cell-sorting microfluidics for (e.g.) novel health detection systems and `smarter' industrial shear enhanced membrane filtration devices. Despite these important applications, a comprehensive model that can predict the lift and drag forces acting on a small particle moving near a wall is not available. Models that have previously been developed are limited to specific wall separation distances, fluid shear rates, and particle slip velocities, which do not cover practically relevant parameter ranges. Further, particle-scale lift models have not been previously employed in the context of suspension modelling to predict the averaged motion of many particles, as opposed to the motion of a single particle. Such suspension modelling is necessary to predict the performance of real biological and industrial multiphase flows. Hence, this thesis aims to develop a comprehensive model of particle lift applicable to the parameter ranges found in typical biological and industrial flows and apply this model to predict the behaviour of particle suspensions in these flows. The work is conducted in two parts. Firstly, the hydrodynamic forces acting on a small spherical particle moving with a finite particle Reynolds numbers in single wall-bounded flows are investigated via direct numerical simulation. Based on these results, new lift and drag models are proposed for rigid spherical particles moving in quiescent and simple shear flows, valid for any wall separation distance, and shear and slip particle Reynolds numbers of order 0.1 or less. The models are used to examine the behaviour of single buoyant and neutrally-buoyant particles moving near walls, and the results are validated against existing experimental and numerical data. Secondly, a two-fluid model, which includes the developed wall-bounded forces, is implemented to predict particle migration in mono-disperse, dilute suspensions at low particle Reynolds numbers. Different implementation methods related to solid phase velocity boundary conditions, the force application phase, and secondary wall effects are discussed. Using this model, the transient solid concentration profiles in Taylor-Couette flows are examined and compared against available experimental results.

Process Safety is a key element of engineering that ensures the rapid human advancements can occur safely. It is important for the people who are already working in the process industries, and for those who will join them in the future. Process Safety education is a mandatory component and should be in the chemical engineering syllabus in every university around the world. A concept map is a graphical representation of information that shows the relationships between concepts. Since its development, it has been used to represent ideas and concepts in a simple and holistic manner. This study investigates whether concept maps may be used to assess learning by individual students and by cohorts of students of the Process Safety domain. To achieve that objective, a method of assessing concept maps through a categorical scoring system is proposed by developing appropriate categorisations for the concepts, links and propositions. The concept categories for the Process Safety domain were developed and validated via a novel process to remove human subjectivity. The process involved defining ten categories into which each of the concepts could be assigned. Several sets of concept maps were analysed independently by three assessors, requiring the assessors to assign every concept into one of the proposed categories. Analysis of the assessors’ responses was aided by presenting their responses in a three-way table. The use of the novel table allowed the assessors’ responses to be compared effectively. This comparison tool enabled identification of problematic categories for further refining. The analysis of the distribution of the concepts, with the help of the proposed Link Quality Index, revealed more information about students’ understanding of the topic. This study analysed different types of connections between concepts to obtain more context and understanding of the concept maps, which represents students’ grasp of the Process Safety knowledge. The students generally appreciated the non-physical preventative measures, such as procedures and maintenance, but they did not recognise the importance of the education, training and values, such as responsibility towards Process Safety. This information is useful in helping to redesign the curriculum. This study also proposes a method that may be applied beyond Process Safety domain; the classification of propositions into one of three proposed attributes, Professional Practice, Values, or Technical Knowledge. The application of this method to engineering students’ concept maps revealed that students show high awareness towards Technical Knowledge. Students also demonstrate the ability to use higher order thinking in explaining the relationship of the behaviour of the Professional Practices and Values. Upon applying this method to nursing students’ concept maps, on the topic of Oxygenation, it is found that nursing students had similar patterns in their awareness of technical attribute. However, nursing students generally were more oriented towards using lower order thinking in explaining relationship between concepts. Overall, this study found that concept maps are a useful method to be applied in Process Safety domain; however, students need to be aware of the importance of having complete and clear propositions which are essential in indicating their understanding.

This thesis reports the successful development and use of surface-initiated polymer films via ring-opening metathesis polymerization (ROMP) and solid-state continuous assembly of polymers reactions as capable bottom-up processes for nanofabrication of organic and inorganic materials. Chapter 1 presents an overview of existing bottom-up nanofabrication methods. Chapter 1 also highlights the drawbacks of current top-down fabrication methods and the opportunities to expand bottom-up methods to nanofabrication, of which are the focus of this thesis. In Chapter 2, the development of a protocol for solid-state continuous assembly of polymers via ring-opening metathesis polymerization (ssCAPROMP) from silanized substrates is presented. This method expands upon previously reported protocols by utilizing an olefinic silane to form highly crosslinked and tailorable polymer thin films which are covalently attached to surfaces. Polymer films constructed using the new anchoring layer showed similar growth characteristics and film thicknesses when compared with previous studies on CAP film growth, but overcome the high degree of steric hindrance and typical minimal macro-cross linker attachment encountered when previously working with silanized silicon surfaces and CAP processes. Films were also successfully reinitiated multiple times, a key property of the CAP process. Optimization of the protocol was carried out via thorough testing of reaction conditions, offering insight into the reactivity of the SI-ROMP process at play. The protocol was further developed to work on the organic material SU-8, a negative photoresist commonly used in microfluidic devices and bio-microelectromechanical systems (bioMEMS) and could be patterned using two methods: masking of the silanization step and subsequent lift-off prior to initiation and film growth, or through selective surface functionalization via the use of polymeric stamps and micro-contact printing procedures. The study presented in Chapter 2 translates the significant advantages of the solid-state CAPROMP protocol when forming polymer films into a method that can be integrated into the fabrication of robust sensors, bioMEMS or microfluidic devices which require tailorable yet stable polymer thin films. In Chapter 3, the ssCAPROMP protocol developed in Chapter 2 was used as a reactive ink system for a micro/nano 3D printing platform to access spatially defined, crosslinked polymer features. The printing platform used was a modified atomic force microscope whereby a hollow cantilever and an aperture at the tip was used to deliver material from a reservoir through a nanofluidics channel to the substrate. Material delivery was controlled by applying pressure to the reservoir. Utilizing the control over polymer properties offered by CAP, a reactive polymer crosslinker was created that encompassed a number of critical parameters in order to be printed successfully; the ink must undergo rapid crosslinking, it must have a low glass-transition temperature to be printed, and the crosslinking must be of a living nature to enable the printing of multiple layers continuously. The reactive ink could be used without solvent and was delivered through the aperture of the tip onto an initiated surface where it was found to crosslink almost immediately, even in ambient conditions and without requiring an inert atmosphere. This rapid crosslinking enabled the delivery of several layers of which each crosslinked, allowing the build-up of various line heights. A comparison of line heights after washing and drying found that when printing with both the 4 micrometer aperture and the 300 nm aperture, repeated depositions in the same location resulted in an almost quantitative addition of material based on the number of depositions. Overlapping lines were also printed and showed that the height at the overlapped location was the sum of the heights of both lines, highlighting the ability to print on top of existing lines. Experiments showed that by waiting a short period of time between overlapping lines, line heights could be increased as the feature could “cure” before the next deposition. Print directionality was shown to affect line widths based on the contact between the deflected tip and the direction it moves. Structures with three distinct layers were also created, showing that this method of delivering reactive ink and crosslinking in-situ could create three-dimensional patterns. By combining this printing platform with the versatility in ink and polymer chemistry offered by CAP, a robust platform for creating three dimensional structures with a layer resolution of down to 2 nm and a minimum line width down to 450 nm was demonstrated. In Chapter 4, the surface-initiated ring-opening metathesis polymerization of norbornene was used to direct the bottom-up construction of TiO2 and ZnO via atomic layer deposition (ALD). Norbornene monomers were used in the vapor phase to avoid excessive polymerization and cross-metathesis side reactions that occur when using solution-based SI-ROMP. This process afforded surface-bound polymer films in an extremely rapid fashion, with 100 nm films achievable in less than one minute of vapor exposure. The polymer films were exposed to TiO2 and ZnO thermal ALD processes and then analyzed via XPS to observe any inorganic material growth. It was found that 100 nm of surface-initiated polynorbornene could resist ALD of at least 1200 cycles, challenging a current paradigm of using small molecules to prevent ALD deposition. Several norbornene-based monomers were synthesized as surface-binding initiators that were selective for copper and copper oxide over silicon oxide. These initiators were used in conjunction with coplanar and topographical copper features on silicon oxide wafers in order to selectively attach initiator to the patterned features. Once coated, the functionalized features were used to grow polymer in an area-selective fashion and these substrates were subjected to ALD to test the ability of the polymers to perform area selective deposition (ASD). ASD of ZnO was achieved using a hydroxamic acid-based initiator, enabling the deposition equivalent of 38 nm of ZnO before any nucleation was observed on the polymer surface. This was also tested on large areas, where a large Sierpinski’s triangle 300 micrometers across was created on a substrate using e-beam lithography to demonstrate that large scale ASD could be performed. Polymers grown on the copper surfaces showed inhibition of ALD for up to 675 cycles of ZnO. These results show that SI-ROMP is not only an excellent tool for the bottom-up construction of organic materials as is shown in Chapters 2, 3 and 4, but that long macromolecules can be used to drive the patterned bottom-up construction of inorganic films relevant to semiconductor and device fabrication to great effect. Finally, in Chapter 5 we propose several pathways where expansion upon the work contained in this thesis may lead to further advances in the bottom-up construction of organic and inorganic materials for nanofabrication.