Search

3D bioprinting is manufactural biotechnology which was used to create a drug delivery system with advanced functions and personalised tissue regeneration. In this study, 3D scaffolds were bio-printed using different bio-ink formulations including pure alginate, alginate-methylcellulose (MC), polylactic acid (PLA) with polyethylene glycol (PEG), and calcium phosphate cement (CPC). The properties of biomaterials were analysed by viscometer and texture analyser. Confocal microscopy, scanning electron microscopy (SEM) and X-Ray powder diffraction (XRD) were applied for surface evaluation of the samples. In vitro cell studies were used for testing cell viability and anti-cancer function of 3D bio-printed scaffolds. The mathematical models between the concentration of pure alginate (C) and viscosity (V) in different pH have been explored, which were V=0.0019C4.9061 at pH=1.2, V=73.097e0.1788C at pH=6.6 and V=0.4059C2.7997 at pH=7.2 respectively. Among all the 3D bio-printed alginate-MC scaffolds, the scaffold made from alginate-MC with a ratio of 1:1 showed a positive effect on cancer cells and inhibited the growth of cancer cells within 4 days, indicating its effective function in anti-cancer drug delivery system. Besides, 3D bio-printed PLA/PEG scaffolds loaded with growth factor erythropoietin (EPO) were proved to be valid for assisting cell repair. And 5-Fluorouracil (5-FU) coated CPC scaffolds showed effective controlled-release drug delivery and anti-cancer function within 4 days. Those 3D bio-printed novel scaffolds illustrated their outstanding potential for future clinical use in both drug delivery and tissue engineering field. Finally, the change of cell behaviour was observed visually by culturing scaffolds with 1BR, HEK293T-GFP and U2OS-GFP cell lines for 28 days. The quantitative analysis showed that the size and structure of scaffolds were two important factors affecting the behaviour of cells and 3D CPC scaffolds in size S (diameter ~9.4mm) is the most suitable one culturing with 1BR cells. The whole thesis explored 3D bio-printed scaffolds in three perspectives, formulations, scaffolds and in vitro functions, which contributed to understanding bio-inks/scaffolds' physical-chemical properties, clinical functions and cell-scaffold interactions.

Mixed selectivity characterises neurons that simultaneously respond to different input stimuli. Neurons with mixed selectivity have been observed in multiple brain regions, and are hypothesised to play important roles in neural computation. Recently, both experimental and theoretical work demonstrated the importance of mixed selectivity in context-dependent decision tasks. This thesis extends existing theoretical work on mixed selectivity, arguing for a general and statistical role of mixed selectivity in learning complex dependencies of input stimuli. This role can be motivated from unsupervised learning of generative models, and is exhibited in increased mutual information between the stimuli and their neural representation. This argument is supported empirically using simulation of a sequence disambiguation task that incorporated key aspects of related behaviour experiments. Mixed selectivity neurons that resembled hippocampal place cells were emerged from models optimised only for behaviour. To understand these results, as well as to generalise the findings to a wider range of computations, I provided a formal connection between learning robust models and mixed selectivity.

This PhD thesis seeks to explore the importance of affiliate strategic roles in determining affiliate survival in foreign markets, reverse knowledge transfer from the affiliate to the parent company, and lateral knowledge transfer between affiliates. I group affiliates into four different types including upstream, horizontal, downstream and unrelated affiliates based on their activities within multinational enterprises' global value chains. Affiliates that are either upstream or downstream in the value chains are regarded as vertical affiliates. Robust to a number of different specifications and drawing on large MNE parent-affiliate linkage data in the period from 2004 to 2016, I find that the likelihood to divest poorly-performing affiliates and the effect of intra-MNE knowledge transfer are contingent on affiliate strategic roles. More specifically, the likelihood to divest poorly-performing affiliates is higher when the overseas affiliates are vertical rather than horizontal FDI, and when they are downstream, rather than upstream, in the MNE value chains. Besides, I find that the performance benefits from intra-MNE knowledge transfer are greater when the overseas affiliates are horizontal rather than vertical FDI, and when they undertake upstream, rather than downstream, activities. This thesis contributes to the international business literature by emphasizing the importance of affiliate strategic roles.

Originating from Silicon Valley in the early 1990s, high-technology-oriented entrepreneurs-turned-philanthropists have applied venture capital principles to philanthropy in order to address intractable social problems, coining the term venture philanthropy (VP). Evolving from an emergent to a pervasive model in Europe in the last two decades, the VP approach has been considered as an innovative alternative to the traditional philanthropy (TP) type of benevolence and cheque-writing (Anheier and Leat 2006). With increasing expectations, in the context of governmental hollowing-out of social services, debate seems to have become polarised. VP is criticised for not being a solution to changes in the social landscape and for its business approaches failing to address fundamental social issues (Sievers 2001; Anheier and Leat 2006; Shiller 2012) and so remaining simply a myth. This research explores the nature of VP based on the organisational perspectives of four funders in Scotland, with a focus on the engagement process. The new empirical data regarding the funding distribution process are gathered with the aim of answering the core questions: 'why give', 'what to give' and 'how to give'. A new operational framework for analysing funders is developed and is used to analyse processual trajectories mapping the convergence and divergence amongst the four funders, citing new evidence from Scotland. Case studies from the years 2011 to 2014 present four grant-giving modes respectively: 1) pure grant-giving but emerging to a business approach applied to funding distribution; 2) grant-giving but applying venture capital approaches (VP); 3) mixed grants and repayable business loans; and 4) repayable business loans. To map the feature of emergent trajectory, a new operational framework is proposed and utilised for analysis. Research findings suggest that a pattern of resource heterogeneity is emerging in the four funding models in response to isomorphic forces. While dealing with inward (governance) and outward (market and political) legitimacy forces, hero-entrepreneurs are shown in the four cases as the key driver to identify the need for change and drive change forward. Meanwhile, hero-entrepreneurship behaviour is associated with the setting of goals, shaping the rationale of the funding scheme, marshalling resources and aligning with partners to demonstrate value adding through the engagement process. The contribution of this research to the philanthropic field is threefold. Firstly, with regard to its theoretical contribution, the findings support conventional isomorphic change theory by arguing: a) that the agent-conduit-roles of funders are not determined by structure, but rather individual agents (hero-entrepreneurs) play a cementing role in the change process of initiating, leading, diffusing influence and levering power for social change; and b) that in their agent-conduit-roles funders act as an active but reflective intermediary, change taking place in the process of legitimacy and resource distribution through the cycle of change-model shaping; convening and conducting; reflecting, dismantling and reshaping. This contribution enhances and complements the discovery by Mair and Hehenberger (2014), which suggests TP and VP create shared space for negotiation, shared objectives and a reflective isomorphic process (Nicholls 2010a). Arguably, funders should strategically consider complex and plural elements of funding and integrating a competitive market and a cooperative rationale with emotional motives into a decision-making. Realisation of social objectives will ultimately be achieved through reflective isomorphic processes, adjusting the funding structure to fit social contexts with convergent resources alignment. Secondly, with regard to its empirical contribution, this research proposes a new typology of funders. Different from the typology proposed by Ostrower (2006), the new typology proposal is based on what the funding is for. The elements of the new typology are synthesised from why, how and what in action, i.e. grant-giving mode, engagement approach and level of risks. Thirdly, practical contributions emerging from the implications of the proposed framework, which are discussed in the concluding chapter, may improve the quality of decision-making in funding behaviour and may also help to shape modes of governance for social problem-solving.

Robot learning from demonstrations (RLD), also known as imitation learning, plays an important role in human-robot interactions (HRI) because it creates a userfriendly environment for non-expert end-users to teach a robot new skills. Recent advances in sophisticated humanoid robots provide an opportunity for these robots to manipulate everyday tools. This thesis focuses on some key aspects in enabling humanoid robots to learn manipulation of tools using an RLD approach as a threefold investigation. Firstly, this work investigates the outstanding issues in existing RLD frameworks for applications in a real-world HRI environment and presents an RLD model using template-based approach integrated with an online learning algorithm to address these issues. Further investigations probe into biological systems for inspirations on improving the reusability of learned skills which results in an integrated framework using the proposed RLD model as the base model. Attention is then turned to an important aspect tightly coupled with RLD learning for tool manipulation - recognition of tools. A biologically-inspired framework based on RLD-learned tool affordance is proposed to address the issues of applying to tools the traditional static feature matching approach for object recognition. In a series of experiments, we show that our RLD model is capable of learning skills without demanding the users for repeated demonstrations. The learned skills can be generalised to similar tasks with different environmental constraints and updated as more demonstrations of the skill are presented. We also demonstrate the reusability of the complete framework in comparison to the base model by a realworld HRI application in playing the tic-tac-toe game. Through the experiments on tool recognition using a set of human demonstrations, the benchmarked performance indicates that our proposed tool representation framework is a suitable supplement to the current object recognition models on tools.

The present thesis explores the cognitive operations underlying word recognition and production of late bilingual adults in their second language (L2). Experimental psychology and electrophysiology have made a case for the activation ofthe first language (Ll) when bilingual individuals process words in L2. Evidence for cross-language activation has shaped current models of bilingual lexical processing and influenced our conception ofthe bilingual lexicon. However, previous studies have made extensive use of interlinguallexical stimuli (e.g., cognates, interlingual homographs) and/or translation equivalents to compare L1 and L2 processing in bilingual individuals. Experiments mixing stimuli from two languages create an artificial context which may differ significantly from real-life situations and bias behavioural performances toward a language-nonselective processing pattern. In the present thesis we tested bilingual participants reading, listening to, and producing words exclusively in their L2. In the first experiment series, Chinese-English bilinguals read and listened to pairs of English words, half of which shared a character repetition in their Chinese translations. Evidence of eve"nt-related potentials (ERPs) showed that Chinese translations were accessed automatically and unconsciously. In the second experiment series, the same paradigm was used except that phonological and orthographic repetitions in, Chinese translations were independently tested. Significant priming was found for phonological but not orthographic repetitions, independently of the input modality (visual or aud~ory), demonstrating that cross-language activation is mediated by phonology. In the third experiment series, speech production was studied using a covert picture naming paradigm involving rhyming decisions. Here again, L1 access was detected but it was delayed in comparison to L2 access as well as more conscious in comparison to reading and IV listening in L2. Moreover, cross-language activation in picture naming was asymmetric, featuring strong influences ofL! on L2, hut no effect ofL2 on Ll. Findings ofthe thesis shed new light on the dynamic nature ofhilinguallanguage processing, as well as constraints affecting cross-language activation. Implications for current models ofhilinguallexical access are discussed.

The Internet has been viewed as a revolutionary means for including individuals in public deliberation on an equal basis. However, there is insufficient empirical work on applications of computer mediated communication to public deliberation under non-democratic social circumstances. This thesis explores the potential of computer-mediated communication in constructing the public sphere in post-Communist China and focuses on the current affairs discussion on Qiangguo Luntan (QGLT), one of the most popular Internet bulletin boards in China. The production-content-reception study shows that: 1) Internet bulletin boards in China are not 'online dissident avenues' without administrative surveillance. In addition, self-censorship works in a long-term and covert way in restraining the democratic potential of online discussion; 2) Internet bulletin boards enable ordinary Chinese to have their identities as politically activated citizens constructed in cyberspace. Their enthusiasm in voicing differentiated political standpoints proves that online public opinion could enrich the political discourse in China, and has the potential of leading to tensions between the public and the state; 3) QGLT SARS postings display the growth of a public critique that ventured to touch upon a taboo issue that was originally banned from public discussion. Non-localized and dialogic forms of communication among the users have created an alternative form of publicity outside the mainstream media's agenda and could potentially check the government's policy-making. The Internet may work to undermine public deliberation by increasing inequalities of access, fragmenting public discourse, and accommodating non-progressive rhetoric. Nevertheless by enlarging the scale of civic participation, advancing alternative and oppositional public discourses, and tackling problems at a global level, Internet bulletin boards could make a significant contribution to the construction of an alternative public sphere representing divergent political views. In summary, under the articulated forces of commercialisation, democratisation, and globalisation, the virtual public sphere in China today has been put in tension between democratisation and degeneration.

Engineering conceptual design is a knowledge-intensive process that generates solutions to a product specification. It is a process that can benefit from past experience of similar designs. In reality however, designers often have limited time to build up the necessary experience and are, in any event, unlikely to become experts in all relevant fields. Hence there is a need to capture, store and reuse valuable knowledge. Currently available conventional CAD systems offer limited possibilities for the re-use of existing designs. Techniques from the field of Artificial Intelligence (Al) may be applied to aid the conceptual design phase, which is known as the area of intelligent computer-aided design. The aim of this work is to identify and externalise design knowledge using a fractal-like model, to understand the role of design knowledge in conceptual design and to use design knowledge as a guide for every stage of concept development. This research provides a framework for supporting conceptual design, which uses the techniques of Case-Based Reasoning (CBR) and fractal theory, for reasoning about the design and development of computer-based design aids. The framework is comprised of three parts. The first is case representation. This research proposes a new representation technique, Fractal-like Design Modelling (FDM), which integrates design knowledge in a graph-based form and has fractal-specific characteristics. The second is case retrieval. Based on FDM, the similarity between a new design and the existing designs is assessed by concurrently applying a feature-based similarity measure and a structure-based similarity measure. The third is case adaptation. With the help of fractal characteristics, an approach of adaptive design is developed by performance revision and by goal-oriented substitution. These three parts work together to achieve an automated, case-based, conceptual design method: Fractal-Based Re-design.

Software vulnerabilities allow an attacker to reduce a system's Confidentiality, Availability, and Integrity by exposing information, executing malicious code, and undermine system functionalities that contribute to the overall system purpose and need. With new vulnerabilities discovered everyday in a variety of applications and user environments, a systematic study of their characteristics is a subject of immediate need for the following reasons: (1) The high rate in which information about past and new vulnerabilities are accumulated makes it difficult to absorb and comprehend. (2) Rather than learning from past mistakes, similar types of vulnerabilities are observed repeatedly. (3) As the scale and complexity of current software grows, better mental models will be required for developers to sense the possibility for the occurrence of vulnerabilities. While the software development community has put a significant effort to capture the artifacts related to a discovered vulnerability in organized repositories, much of this information is not amenable to meaningful analysis and requires a deep and manual inspection. In the software assurance community a body of knowledge that provides an enumeration of common weaknesses has been developed, but it is complicated and not readily usable for the study of vulnerabilities in specific projects and user environments. Also the discovered vulnerabilities from different projects are collected in various databases with general metadata such as dates, person, and natural language descriptions but without the links to other relevant knowledge, they are hard to be utilized for the purpose of understanding vulnerabilities. This research combines the information sources from these communities in a way that facilitates the study of vulnerabilities recorded in large software repositories. We introduce the notion of "Semantic Template" to integrate the scattered information relevant to understand and discover vulnerabilities. We evaluate the use of semantic templates by applying it to analyze and annotate vulnerabilities recorded in software repositories from the Apache Web Server project. We refer to software repositories in a general sense that includes source code, version control data, bug reports, developer mailing lists and project development websites. We derive semantic templates from community standards such as the Common Weaknesses Enumeration (CWE) and Common Vulnerabilities and Exposures (CVE). We rely on standards in order to facilitate the adoption, sharing and interoperability of semantic templates. This research contributes a novel theory and corresponding mechanisms for the study of vulnerabilities in large software projects. To support these claims, we discuss our experiences and present our findings from the Apache Web Server project. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Cellular structures are networks of interconnected struts or walls with porosities and are widely found in many natural load-bearing structures such as plants and bones. Cellular structures offer unique functional characteristics such as high stiffness to weight ratio, tailorable heat transfer coefficient, and enhanced mechanical energy absorption, which makes them highly attractive in various engineering disciplines such as biomedical implants, electrodes, heat exchangers, and lightweight structures. There exists an abundance of literatures that have investigated various mechanical properties of various cellular structures such as Poisson’s ratio, elastic modulus, ultimate strength, yield strength, and failure characteristics. Based on the classic cellular structure model, these mechanical properties are highly dependent on both the relative density and the topologies of the unit cell designs. Cellular structures with higher relative densities generally exhibit higher overall mechanical properties. In addition, there also exist multiple general design rules for cellular unit cell topology designs, such as nodal connectivity-based deformation mechanism and re-entrant auxetic mechanism. However, currently most theoretical knowledge for cellular structures are established based on infinite pattern sizes, i.e. infinite numbers of unit cells along all principal symmetry directions. On the other hand, for the cellular structures with finite sizes that are commonly designed in real-world applications, in addition to relative density and cell topology, the cellular pattern size effects, which are introduced by the non-ideal boundary conditions, also plays important roles in determining the overall mechanical characteristics of the cellular structure. As a result, many equations and conclusions from the classic Ashby and Gibson models cannot be directly applied to these finite-size cellular structures, which significantly limits the designability of cellular structures for various dimension-limited applications. Besides, due to the complex geometry of cellular structures, additive manufacturing (AM) processes have been considered as the only practically viable option for their fabrication, which introduce various manufacturing-related design variables with material properties such as material anisotropy and material imperfection. In order to adequately design for cellular structures realized by AM processes, a modeling approach that enables comprehensive analysis of all these factors are desirable. In this work, an analytical model framework was established for the analysis of mechanical characteristics of the finite-size cellular structure with imperfect local material properties. The model was verified by the experimental results for both mechanical properties and cellular fracture failure propagation patterns with samples fabricated via powder bed fusion (PBF) process. The results showed that the models could not only provide good predictions to both average mechanical properties and their variabilities, but also adequately capture the effects of the finite pattern size effects and local material heterogeneity effects. Based on the established model, the topology-material-mechanical properties of the finite-size AM cellular structures were investigated in detail. More specifically, the effects of pattern size-topology, material anisotropy and the material imperfections were studied systematically. Various new insights were obtained, including: 1. Based on the modeling analysis, the effects of size and topology on the tensile failure behavior of multiple representative cellular structures (2D auxetic, 2D diamond, 2D triangular1 and 2D triangular2) under various geometry design conditions (including cell topology, cell size and number of unit cells) were systematically investigated. It was found that the 2D bending-dominated structures with lower nodal connectivity (number of struts that meet in joints) (2D auxetic and 2D diamond) exhibited a relatively progressive crack propagation pattern, while the 2D stretching-dominated structures with higher nodal connectivity (2D triangular1 and 2D triangular2) appear to exhibit rather catastrophic brittle fracture failure. During the failure fracture propagation, the energy absorption of the 2D stretching-dominated structures were significantly higher than that of the 2D bending-dominated structures. Moreover, for all cellular designs, the tensile failure behaviors tend to converge to more consistent patterns when the cellular structure pattern sizes increase beyond certain thresholds that are dependent on the cellular topology designs. 2. The material anisotropy effects, which are characteristic to AM processes, were explored through both analytical modeling analysis and experiments on three representative 3D cellular structures (auxetic, BCC and octahedral). The established models were verified via experimentation with samples fabricated by electron beam PBF (EB-PBF) process using Ti6Al4V as material, using the material anisotropy information established experimentally using single struts with different build orientations (0°, 15°, 30°, 45°, 60°, 75° and 90°). The predicted mechanical properties of the Ti6Al4V cellular structures showed good agreement with experimental results. It was shown that both the strength and elastic modulus anisotropy of the materials affect the strength of the cellular structures, which must be determined based on the topology design. In addition, the material anisotropy-topology effects on cellular structures of varying cellular pattern sizes were also investigated in order to quantify the pattern size effects. It was also found that the pattern size effects and the material anisotropy effects can be decoupled during the design of the mechanical properties of these cellular structures. 3. The local material property fluctuation caused by the material imperfection is another important factor to consider for adequate design of AM cellular structures. The local material and feature imperfections affect the overall structural properties of cellular structures and are typically unavoidable with the current AM process technologies. Three representative 2D cellular designs including auxetic, diamond and triangular structures were modeled and analyzed based on the established model, which allows for the implementation of heterogeneous material imperfection at full-scale cellular structure level. The material property imperfection was represented by 3 levels of variabilities (2%, 5% and 10%) for both elastic modulus and strength, defined at local cellular element level. Experimental verification using Ti6Al4V cellular structures fabricated via laser PBF (L-PBF) process demonstrated the potential of the established model in providing accurate predictions to the mechanical property variability of the cellular structures. In addition, the results also revealed new insights into the topology-material imperfection coupling relationships for the cellular structures.

Understanding human development is critical to understanding human evolution, treating developmental disorders, and creating regenerative therapeutics. Single-cell methods have enabled the high-resolution profiling of developmental trajectories and novel organoid models have facilitated a greater understanding of difficult to access developmental periods, especially through the use of perturbation experiments. Here, we used a novel multi-lineage developmental model along with profiling of the developing human prefrontal cortex to better understand human development and evolution. First, we developed a novel visualization method, Similarity Weighted Nonnegative Embedding (SWNE), which both preserves the structure of single-cell datasets and enables key marker genes and relevant genesets to be embedded alongside the cells. We then leveraged a novel multi-lineage developmental model, the teratoma, to study the role of key developmental genes. We conducted a pooled CRISPR knockout screen of those regulators in the teratoma with a single cell RNA-seq readout, enabling us to better understand the function of these genes across all major human lineages. Finally, we used both single-cell RNA-seq and single-cell chromatin accessibility profiling to study the role of human accelerated regions (HARs), genomic regions thought to influence human-specific evolution, in human corticogenesis. The chromatin accessibility enabled us to assess the activity of HARs in specific developmental cell types and link those HARs to genes using co-accessibility of the HARs and gene promoters, while the RNA-seq enabled us to validate the expression of those HAR-linked genes.

Despite significantly improved technologies in food science and public health and tremendous efforts being put by governments to ensure food safety, foodborne outbreaks are still abundant worldwide. Produce products have been frequently implicated in foodborne illness outbreaks in recent years due to changes in consumer demands, consumption habits and production practices. A better understanding on epidemiology changes of produce outbreaks is needed to evaluate current risks associated with produce supply chain and to understand safety regulations regarding produce safety. In addition, it is evident that water used in produce production plays an important role in potentially introducing microbial contaminations. Therefore, its risk management is crucial for safety assurance of the produce supply chain. The goal of this thesis research is to analyze the epidemiological trends of produce outbreaks and to improve the risk management of microbial quality of irrigation water. It summarizes the changing epidemiology of produce outbreaks in the United States from 1998-2007, establishes the baseline to further evaluate the potential impact from the recently implemented Food Safety Modernization Act (FSMA). The study also describes the development, optimization, and evaluation of a novel selective medium for sensitive enrichment and screening of Shiga-toxin producing E. coli and Salmonella in irrigation water. The developed enrichment-indicator system meets the increasing demand of method for multi-pathogen enrichment and detection in a single assay format allowing cost effective detection of STEC and Salmonella within 24 hours.

This thesis uses an interdisciplinary approach that draws on both the political science and media and communication fields to analyse how Chinese environmental NGOs use the microblogging site, Sina Weibo, in their online activism. The study of NGOs and how they use the internet in China is widespread. However, in many cases, the way that NGOs in China work, both online and offline, has been analysed through the lens of traditional civil society and internet studies literature, which has mostly focused on the ability of NGOs, and the internet, to give rise to significant political change, and even democratisation. Through a mixture of thematic, network, and organisational analysis, this thesis investigates the communicative functions, themes, and use of interactive features in posts on Weibo, including the use of hashtags, retweets and @mentions. At the organisational level, the ways that NGOs engage with different actors, both online and offline, including fellow NGOs, government departments, their followers, and potential donors are interrogated using four case studies. These analyses found that although the political space afforded to environmental NGOs in China is severely constrained, and the operations of the NGOs could not be seen as overtly activist or confrontational in the traditional sense, the NGOs do in fact retain a certain amount of autonomy and are able to carve out some political space for themselves. The findings of this thesis therefore challenge the notion that NGOs in China are co-opted organisations without autonomy from the state and suggests that there is scope for digital activism by NGOs in an authoritarian context, even though the online and offline political space they inhabit may be tightly regulated and controlled.

The goal of this work is to understand the connections between stress, structural and mechanical property development during the drying of particulate coatings containing rigid particles, and the how these conditions are impacted by coating formulation variables and processing conditions. The motivation is to better predict and control the coating performance by optimizing the formulation design and drying condition. Characterizing drying behavior and correlating the stress development with microstructure evolution are critical in this research. In Chapter 3, drying characterization approaches are introduced. To extend the capability of stress measurement, walled substrates with different dimensions and materials were designed. And a shrinkage measurement method allowing convenient correlation of the microstructure change with stress development was developed using laser profilometry technique. Chapters 4-6 focus on investigating the role of formulation variables on the drying and cracking of coatings prepared with micron-sized particles. Chapter 4 focuses on studying the effect of particle size distribution. The study showed that with similar average particle size, coatings prepared from particles with a wide particle size distribution form a more compact microstructure, but are prone to cracking due to high tensile stress development on drying. Chapters 5 and 6 study the effect of particle shape on coatings cracking resistance. Different levels of clay particles were added to particulate coating systems prepared from the irregular-shaped ground calcium carbonate (GCC) particles and the spherical-shaped silica particles. The different geometry constraints of the mixed particles have altered impacts on the drying shrinkage and mechanical strength of the coatings, thus different cracking behaviors were observed. Finally, in Chapter 7 attention is shifted from the particulate coatings with the micron-sized particle to those prepared from nano-sized particles. Formulation variables of particle size and shape were characterized using coatings of nano-silica particles, nano-zinc oxide particles, and fumed-silica.

This thesis addresses the following research question: How can we best understand the effects of internet campaigning on the campaign practices of the People's Democratic Party (PDP) and the All Progressives Congress (APC) in the 2015 Nigerian Presidential election? This research represents an important contribution to the broader study of internet campaigning; widening the theoretical and empirical scope of the literature. From a theoretical perspective, three major approaches are deployed across the thesis: modernisation, Americanisation, and hybridisation. While all three lenses play an important part in understanding the effect of the internet on Nigerian campaign practices, the hybridisation perspective is particularly important - pointing towards a broader need in the literature to integrate this theoretical emphasis. From an empirical perspective, over 50 original, elite interviews in Nigeria were conducted with members of both parties' campaign teams and campaign consultants. In analysing these data, the thesis unpacks three sub-questions: How was internet campaigning adopted and adapted by the campaign teams? What factors help to explain variations in the internet campaigns practices of the presidential candidates of the PDP and APC? How did the internet affect the intra-campaign organisational dynamics of the presidential candidates of the PDP and APC? The analysis across these questions concludes that the importance of the 2015 Nigerian online campaign should not be underestimated - it clearly impacted on campaign practices and organisation. However, the nature of this impact falls far short of a full realisation of the potential impact that the web could have exerted. Understanding this reality requires that close attention be paid to the national and party contexts within which internet campaigning was adopted - meaning that a hybridisation perspective is central to explaining how the internet impacts campaign practices in states such as Nigeria.

Shigella causes Shigellosis or bacillary dysentery which is a major diarrheal disease throughout the world especially in developing countries. The annual incidence of Shigella infections was estimated to be 164.7 million with 1.1 million deaths, with the majority of the cases and deaths involving children under 5 years old. Humans are the only host of Shigella. Among the four species of Shigella, S. flexneri is the most predominant species affecting economically poor populations. S. flexneri is further divided into different serotypes based on the structure of the O-antigen. Up to now, at least 15 serotypes have been recognized. All except serotype F6 belongs to the same evolutionary cluster based on housekeeping genes. Thus understanding the molecular epidemiology of S. flexneri will contribute significantly to the development of new strategies for disease prevention and control. This study aimed to identify highly VNTR loci to develop a MLVA typing scheme for S. flexneri and to study further the molecular evolution of S. flexneri. This study used a new approach to identify VNTRs. Through searching two VNTR databases using a range of parameters this study identified 52 potential VNTRs. Variable VNTRs were further refined from comparison of three available S. flexneri genomes. This reduced the number of VNTRs to 13 for initial screening. However, while this project was ongoing, an MLVA study of S. flexneri was published which tested 36 potential VNTRs. The 13 most variable VNTRs from that study, 10 of which overlapped with our initial screening, were adopted for MLVA typing in this project. The 13 VNTRs were used to type 51 S. flexneri isolates. The average number of isolates that can be typed for a VNTR was 95%. The VNTR SF17 has the lowest typability. Only 67% of the isolates can be typed. Only one VNTR, O157-11, had a typability of 100%. MLVA-13 (all VNTRs) and MLVA-8 (the eight most variable VNTRs) had a discriminatory power of 0.998 while MLVA-4 (the 4 most variable VNTRs) had a slightly lower power of 0.994. MLVA typing schemes developed in this project and by other studies were tested in 55 isolates. The level of differentiation of the 55 isolates into MLVA profiles depends on the number of VNTR loci used. The scheme using all 13 VNTR loci divided the 32 of the 55 isolates that can fully be typed into 31 MLVA profiles. Using the MLVA-8 scheme, 49 of the 55 isolates had a complete VNTR data and can be assigned into 47 MLVA types. Further reducing to the 4 most variable VNTRs (the MLVA-4 scheme), 52 of the 55 isolates had a complete VNTR data and were differentiated into 46 MLVA profiles. Phylogenetic analysis of MLVA data showed the isolates can be divided into 4 clusters with the most prevalent serotype 2a being mostly located in cluster I. This study applied MLST-7 to Australian isolates. The 15 Australian isolates were typed into four different STs, ST245, ST626, ST651 and ST1025 with the latter being a new ST. Together with 36 isolates which has been previously typed, this study showed that the majority of the isolates belonged to ST245. This ST appears to be a globally distributed clone. The isolates of ST245 were further separated by inclusion of two genes from MLST-15 scheme.. MLST-15 ST91 comprised 16 of the 34 ST245 isolates. For the first time this study showed that the ST91 which was first reported in China as a multidrug resistant clone is present in other Asian countries and Australia although the isolates in this study have not been tested for multidrug resistance. This study provided another dataset to measure the variability of these 13 VNTRs and provided data for the general applicability of VNTRs for typing S. flexneri. The data from this project confirmed the proposed two MLVA schemes using four and eight VNTRs for different levels of resolution were applicable to our strains. The combination of MLST-7 and MLST-15 schemes revealed predominant S. flexneri STs circulating globally, demonstrating the usefulness and limitations of MLST for global epidemiology. There is significant consistency between MLVA clustering and MLST sequence types. This study has significantly increased our understanding of epidemiology and evolution of S. flexneri.

Consisting of both theoretical and empirical inquires, this study examines the primary functions of narrative and the relationship between narrative and mental imagery. The study proposes a new framework to interpret semiotic resources. Combining this with the linguistic functional theory of Halliday (1978), a functional method to empirically investigate semiotic representations was also developed. In the empirical inquiry, the study developed a latent construct method to empirically test the effects of narrative in a real learning situation. This study is the first to investigate the functional relationship between narrative and mental imagery, and among the first to suggest a theory and empirically investigate representations of a multimodal nature. The study is also among the first to use latent constructs to investigate the learning experience in a real educational setting. Data were collected from 190 library professionals who enrolled in three sections (two in narrative and one in plain text) of an online course administered through Vista 4.0 and who completed the course and responded to several instruments. Essay data (n = 82 x 2) were analyzed using content analysis based on the narrative analysis framework developed. Quantitative data analysis methods include univariate data analysis, factor analysis, and structural equation modeling that tests the proposed model and verifies the relationships between the latent variables. Overall, the findings support the hypotheses about the functional effects of narrative identified, and narrative is found to provide a favorable and positive learning context which is tested by the proposed model of learning experience measured by several latent constructs (X2 = 31.67, df = 47, p = .9577, RMSEA = .00, SRMR = .047, NNFI = 1.05, CFI = 1.00, and GFI = .94). The results indicate that participants who enrolled in the narrative sections of the course gained higher creative scores and showed better results in performance-based and attribution-based experiences. The model testing results indicate that even though more time spent during learning led to better outcome and performance in both groups, more time spent means more satisfaction for the individuals in the narrative group, but led to less satisfaction for the individuals in the non-narrative group.

Lithium ion batteries are widely used to power portable electronic devices such as cell phones and laptop computers due to their high energy density. However, the currently used layered LiCoO2 cathode could deliver only 50 % of its theoretical capacity in practical lithium ion cells (140 mAh/g) due to the chemical and structural instabilities at deep charge with (1-x) < 0.5 in Li1-xCoO2. Also, cobalt is relatively expensive and toxic. These difficulties have generated enormous interest in alternative cathode hosts. In this regard, solid solutions between layered Li[Li1/3Mn2/3]O2 (commonly designated as Li2MnO3) and LiMO2 (M = Mn, Ni, Co)) have become appealing as some of them exhibit much higher capacity (~ 250 mAh/g on charging to 4.8 V) with lower cost and better safety compared to LiCoO2. This dissertation investigates the (1-z) Li[Li1/3Mn2/3]O2 - (z) Li[Mn0.5-yNi0.5-yCo2y]O2 (y = 1/12, 1/6 and 1/3 and 0.25 ≤ z ≤ 0.75) layered oxide cathodes, which belong to a solid solution series between layered Li[Li1/3Mn2/3]O2 and Li[Mn0.5-yNi0.5-yCo2y]O2, with an aim to develop a better understanding of the chargedischarge mechanisms and optimize the electrochemical performance of these materials. To accomplish this, the structural and electrochemical characterization of the (1- z) Li[Li1/3Mn2/3]O2 - (z) Li[Mn0.5-yNi0.5-yCo2y]O2 cathodes is carried out. It is found that the amount of oxygen loss is related to the lithium content in the transition metal layer, and the Co and Mn4+ contents play a role in influencing the electrochemical behavior. In addition, the chemically delithiated samples are found to transform to O1 or P3 structure with a vanishing of the superlattice reflections arising from cationic ordering in the transition metal layer due to the incorporation of protons from the chemical delithiation medium, while the electrochemically charged samples retain the initial O3 structure. These layered solid solution oxides exhibit high irreversible capacity (IRC) loss (difference between first charge and discharge capacity) values (up to 100 mAh/g), which have been reduced significantly by modifying the cathode surface with other materials like Al2O3, AlPO4, and F- . For example, compared to an IRC of 75 mAh/g and a first discharge capacity of 253 mAh/g for the pristine Li[Li0.2Mn0.54Ni0.13Co0.13]O2 (y = 1/6 and z = 0.4), the 3 wt. % Al2O3 modified sample exhibits a lower IRC of 41 mAh/g and a higher first discharge capacity of 285 mAh/g, which is two times higher than that achieved with the LiCoO2 cathode. A careful and systematic analysis of the experimentally observed capacity and IRC values suggest that part of the oxide ion vacancies created during first charge is retained in the layered lattice in contrast to the idealized model (elimination of all oxide ion vacancies) proposed in the literature. The surface modification helps to retain even more number of oxide ion vacancies in the lattice, which leads to a lower IRC and higher discharge capacity values. Additionally, bulk cationic and anionic substitutions of Al3+ and F- in Li[Li0.17Mn0.58Ni0.25]O2 (y = 0 and z = 0.5) are found to sensitively decrease the amount of oxygen loss from the lattice.

The mechanical response of rock blocks around tunnels is affected by factors such as the orientation and shear strength of the rock joints, the redistributed stresses, supportforces and other applied loads, and the shape and size of the tunnel. The safety of rock blocks can be studied by means of block theory. Since block theory ignores surface tractions arising from redistributed stresses, it tends to be over-conservative and may lead to uneconomical designs. When redistributed stresses are considered, however, the system of forces governing block safety becomes statically indeterminate. Indeterminacy in the mechanical response leads to uncertainty in determination of the safety conditions.This thesis presents an optimization method for the safety evaluation of two dimensional and three-dimensional rock blocks, which considers the effects of self weight,redistributed stresses, friction angles, and joint set orientations on the safety of the blocks around tunnels. To investigate safety requirements, a fictitious support force was introduced yielding two limiting values: the minimum support force for which safety can be achieved and the maximum support force for which the unsafe condition can occur. Linear Programming was used to determine these limiting values. Effects of friction angle, jointorientation, distributed stresses, and support pressure were studied. It was found that as the height or apical angle of the block increases, the block safety decreases. Block safety increases as horizontal stress ratio K, function coefficient p of the joints, or the tunnel depth increases.