Search

Given the relatively weak strength of indium salts as Lewis acids, trimethylsilyl halides or other silane derivatives are often employed together with In(III) salts as robust combined Lewis acid catalysts. This thesis focuses on the development of novel organic reactions using In(III)-TMSBr catalyst system. A novel example of In(tfacac)3-TMSBr-catalyzed [2+2] cycloaddition reaction of aryl alkynes with acrylates that afford cyclobutenes with aryl and ester substituents in moderate to good yields is developed. Moreover, selective cleavage of C-C/C=C bond of cyclobutene products is also attained to give synthetically useful 1,4- and 1,5-dicarbonyl compounds. Furthermore, a novel In(tfacac)3-TMSBr-catalyzed cascade dearomatization reaction of aryl alkyne with acrylates is also studied for efficient assembly of densely functionalized polycyclic ene skeletons with high stereocontrol. Mechanistic studies conducted aid in elucidation of a plausible mechanism involved for this serendipitously discovered reaction. Using In(III)-TMSBr catalyst system, the synthesis of 1,5-di-bromo-1,4-pentadienes, 1,3,5-triaryl benzenes as well as [4+2] cycloaddition of ɑ-aryl vinyl bromide with methyl acrylate are also demonstrated. ​Doctor of Philosophy (SPMS)

Regulation of eukaryotic gene expression by microRNAs, 19-25 nt long non-coding RNAs, plays a major part in organismal development and function. Individual miRNAs often exist as populations of length variants but the mechanisms underlying this heterogeneity are poorly understood. Here we identified a large set of miRNAs that are significantly shortened during mouse nervous system development. This phenomenon, which we refer to as miRNA trimming, occurs at the 3’ end of mature miRNAs following the Dicer processing step. Of the four miRNA-interacting Argonaute paralogs encoded in mammalian genomes, Ago2-bound miRNAs are trimmed with the highest efficiency. This is consistent with the high expression of Ago1 in embryonic brain and an increase in the relative abundance of Ago2 at the later developmental stages. We further show that the trimming efficiency is defined by the structure of the Argonaute PAZ domain interacting with the miRNA 3’ end as well as the identity of the 3’-terminal nucleotide. These results uncover the previously unknown Argonaute role in developmentally regulated changes in the composition of miRNA isoforms. Doctor of Philosophy

In this work, we study the extended Shastry-Sutherland model which is a Quantum spin system with geometrical frustration. It is originated from the compound SrCu2(BO3)2 which exhibits nontrivial magnetization plateaus in the external magnetic field. In the first part, we use a generalized spin wave theory (spin wave theory in plaquette representation) to investigate the intermediate phase (extremely frustrated range) of the Shastry-Sutherland model and the associated quasiparticle dispersions. We confirm the existence of this plaquette singlet state by explictly calculating its energy with quantum corrections from second order perturbation theory. We also propose a more general plaquette valence-bond-solid (PVBS) phase when the anisotropy is turned on. The quasi- particle dispersion changes qualitatively when it passes a critical line within the PVBS phase. The gap splits from k = (0,0) to four degenerate points which may imply a crossover to the resonanting valence bond state. In the second part, we study the magnetization plateaus and supersolid phases in the extended Shastry-Sutherland model which is expected to be the effective model of the rare earth tetraborides family. Analytical (spiral plaquette representation) and numerical (Stochastic Series Expansion QMC) methods are applied to the Ising and Ising-like XXZ models, respectively. Results from the two methods are qualitatively consistent. Expected plateaus 1/2 and 1/3 have been observed in the phase diagram and a few new plateaus including 5/9 and 2/9 have been observed in a narrow regime of the phase diagram. Besides, we also study the Spin-1 Heisenberg model on a bipartite lattice to test the consistency of the QMC and spin wave method. The excellent quantitative and qualitative agreement guarantees the efficiency of the lowest order spin wave theory. DOCTOR OF PHILOSOPHY (SPMS)

Quantitative understanding of complex biological systems has been a major challenge. The synthetic biology toolkit helps in engineering genetic circuits to design such complex systems. To develop effective precisely-controlled circuits, rapidly regulate-able gene expression systems are necessary. Opto-genetic modules offer unprecedented ways to control cellular functions in precise spatial and temporal resolution. However, current opto-genetic systems have been limited to low dynamic expression and switchable control. Henceforth, the design of synthetic circuits for coordinated cell-cell communication has been difficult. A novel blue light-regulated genetic system is characterised in this work that demonstrates precise and stable switchable gene expression control with light intensity and pulse-width percentages. The system provides effective means of coordinated functions and thus, is applied to optically synchronize cells to perform different control functions. With such precise control, the system serves as a basis for underlying biological patterns and new developments in Systems biology and Biotechnology. MASTER OF ENGINEERING (EEE)

Complex bis(u-chloro)-bis{(R)-1-(dimethylamino)ethyl] naphtyl-C2,N} palladium (II) was found to successfully undergo insertion reactions at the palladium-carbon with the diphenylarsinoprop-1-yne ligand at room temperature which leads to the formation of a new 7-membered metallacyce in the inserted product. This reaction was also carried out on the platinum (II) metal template containing the naphythylamine moiety. The monoarsine precursor complexes has shown to undergo a C-H bond activation followed by a C-C bond coupling at room temperature. This reaction can only be initiated with the presence of a platinum (II) metal. The chiral organopalladium template was used to promote asymmetric hydroarsination reaction between diphenylarsine and various functionalised alkenyl phosphines with hydroxy, methoxy, dimethylamino, ester and ketone as functional groups. The asymmetric Diels-Alder reaction promoted by chiral platinum (II) metal template between DMPP and phenyldi[(Z)]prop-1-enyl]phosphine oxide showed poor stereoselectivity with two isomers (1:1) being formed. The cycloaddition reaction between DMPP and phenyldi[(Z)]prop-1-enyl]phosphine sulphide similarly showed poor selectivity. DOCTOR OF PHILOSOPHY (SPMS)

This study addressed the tissue specific function of pleotropic heterochromatin component HP1a in the Drosophila female germline, particularly to determine whether there was crosstalk with the piRNA pathway. Abolishing HP1a in the female ovarian germline led to complete sterility. A significant de-repression of the telomeric retrotransposon- HET-A, TART and TAHRE (which regulate telomere length in Drosophila), was also observed. piRNAs against the telomeric retrotransposons were severely reduced suggesting that HP1a regulates telomeres in piRNA dependent mechanism in the ovarian germline. Furthermore, we determined that HP1a occupies telomeric transposon loci and probably functions in the primary piRNA cluster transcription for the telomeric retroelements. DOCTOR OF PHILOSOPHY (SBS)

A simple and portable colorimetric assay for creatinine detection is fabricated based on the synergistic coordination of creatinine and uric acid with Hg(2+) on the surface of gold nanoparticles, which exhibits good selectivity and sensitivity. Point-of-care clinical creatinine monitoring can be supported for monitoring renal function and diagnosing corresponding renal diseases at home.

Gossips, rumors and casual conversations are a part of the informal, routine exchange of information that humans engage in. During events of crisis, rumors feature prominently due to the extreme uncertainty and lack of information surrounding these situations. Despite recent research that explores rumoring on social media platforms, limited work has focused explicitly on how crowds self-correct rumors through the expression of feelings of anxiety, confusion and skepticism. This dissertation examines how users of social media express uncertainty and subsequently correct online rumors during a period of collective uncertainty. Focusing on tweets from Twitter that were collected and analyzed during an incident surrounding the death hoax of Singapore’s first Prime Minister, different ways of expressing doubts and information-correcting behaviors were identified. Over 8,100 tweets over a 4-day period in 2015 were harvested from Twitter. They were then coded into two dimensions. The first dimension is to identify crowd corrections and to determine whether the rumor is supported or refuted. This is achieved by coding the tweets along 5 mutually exclusive categories. Subsequently, the contents of the tweets were then coded according to the second dimension categories, which captures expressed uncertainty in the rumor corpus following three Shielding behavior sub-categories and five Milling sub-categories. The approach extends the understanding of how expressed uncertainties in online rumoring by users of social media helped in shaping the collective sensemaking process. The results have implications for government agencies, businesses and companies during crisis management. As the findings suggest, expressed uncertainty may be an earlier indicator of rumors than denials or corrections, which could benefit organizations, especially government agencies, in improving the speed of detection of online rumoring, identify them at the initial stage and prevent the spread of false rumors in the online space. If left uncorrected, the misinformation could have a detrimental effect on the stability of the country’s interests and tarnish its reputation. This dissertation also adds to the growing knowledge of rumor studies, rumor correction and the expression of uncertainties by users of social media during crisis events. Master of Science (Information Studies)

Optical microscopy is often used to characterize the surface microstructure of polycrystalline materials because of its simplicity and low-cost. However, the type of information acquired is limited by the diffraction barrier of optical lenses. For this reason, optical microscopy cannot be employed to resolve materials’ atomic structure and its use is typically limited to quantifying the geometry and distribution of the microstructural features that can be resolved, such as crystal grains, phases, and defect. In this project we will use modified optical techniques and develop novel numerical analysis methods to overcome the diffraction barrier of optical microscopy and assess grain orientation information. Our techniques rely on the quantification of reflected light intensity as a function of incident light direction or light polarization. Our results lay the foundation for the development of new optical-based techniques for cost-effective and high-throughput microstructure characterization of polycrystalline materials. Master of Science (Aerospace Engineering)

Light-responsive hydrogel particles with multi-compartmental structure are useful for applications in microreactors, drug delivery and tissue engineering because of their remotely-triggerable releasing ability and combinational functionalities. The current methods of synthesizing multi-compartmental hydrogel particles typically involve multi-step interrupted gelation of polysaccharides or complicated microfluidic procedures with limited throughput. In this study, a two-step sequential gelation process is developed to produce agarose/alginate double network multi-compartmental hydrogel particles using droplets assemblies induced by superhydrophobic surface as templates. The agarose/alginate double network multi-compartmental hydrogel particles can be formed with diverse hierarchical structures showing combinational functionalities. The synthesized hydrogel particles, when loaded with polypyrrole (PPy) nanoparticles that act as photothermal nanotransducers, are demonstrated to function as near-infrared (NIR) light triggerable and deformation-free hydrogel materials. Periodic NIR laser switching is applied to stimulate these hydrogel particles, and pulsatile release profiles are collected. Compared with massive reagents released from single-compartmental hydrogel particles, more regulated release profiles of the multi-compartmental hydrogel particles are observed.

Ecosystems may undergo nonlinear responses to stresses or perturbations. Hence there can be more than one stable state or regime. It is not known whether alternate regimes also occur in coupled social-ecologica-l systems, in which there is the potential for intricate feedbacks between natural and social processes. To find out, we investigated the management of rice paddies by Balinese farmers, where ecological processes impose constraints on the timing and spatial scale of collective action. We investigated responses to environmental and social conditions by eight traditional community irrigation systems (subaks) along a river in Bali to test the intuition that older and more demographically stable subaks function differently than those with less stable populations. Results confirm the existence of two attractors, with sharply contrasting patterns of social and ecological interactions. The transition pathway between the two basins of attraction is dominated by differences in the efficacy of sanctions and the ability of subaks to mobilize agricultural labor. Published version

Most of the surface-enhanced Raman scattering (SERS) substrates are 2D planar systems, which limits the SERS active area to a single Cartesian plane. Here, we fabricate 3D SERS substrates with the aim to break the traditional 2D SERS active area limitation, and to extend the SERS hotspots into the third dimension along the z-axis. Our 3D SERS substrates are tailored with increased SERS hotspots in the z-direction from tens of nanometers to tens of micrometers, increasing the hotspots in the z-direction by at least an order of magnitude larger than the confocal volume (∼1 μm) of most Raman spectrometers. Various hierarchical 3D SERS-active microstructures are fabricated by combining 3D laser photolithography with Langmuir-Blodgett nanoparticle assembly. 3D laser photolithography creates microstructured platforms required to extend the SERS-active area into 3D, and the self-assembly of Ag nanoparticles ensures homogeneous coating of SERS-active Ag nanoparticles over the entire microstructured platforms. Large-area 3D Raman imaging demonstrates that homogeneous signals can be collected throughout the entire 3D SERS substrates. We vary the morphology, height, and inclination angles of the 3D microstructures, where the inclination angle is found to exhibit strong influence on the SERS signals. We also demonstrate a potential application of this hierarchical 3D SERS substrate in information tagging, storage and encryption as SERS micro-barcodes, where multiple micro-barcodes can be created within a single set of microstructures.

Systemic lupus erythematosus (SLE) is characterized by the production of autoantibodies (ANAs), these form self-reactive immune complexes with self-antigens which results in tissue inflammation. The SLAM family (SLAMF) receptors are regulators of the innate and adaptive immune response. In mice, SLAMF are located within the sle1 lupussusceptibility locus. When present on a B6 background, sle1 drives B and T cell hyperreactivity and ANAs production. ​Doctor of Philosophy (SBS)

We discuss three different topics in combinatorial number theory. In the first topic, we form several truncated series from the quintuple product identity and its specialised versions, then prove that the coeffcients of these series exhibit uniformity in sign. In other words, all the coefficients are either positive or negative. In the second topic, we are interested in N(m,n), which denotes the number of partitions of n with rank m. Our numerical computations suggest that for m ≥ 0, n ≥ 39 and n ≠ m+ 2, we have N(m,n) ≥ N(m+ 1, n). We analyse this conjectural monotonicity and presents several useful results which could help solving it. Finally, in the last topic, we study many special eta quotients such that, when written as series, their coeffcients are interlinked in a specific manner which could be generalised. By observing that these series are associated with the quadratic form x^2 + ky^2, we construct a system of transformation formulas to prove this phenomenon of interlinked coefficients. ​Doctor of Philosophy (SPMS)

Graphene has been touted as the next "wonder material" for various applications, owing to its excellent properties. The development of novel graphene materials enables hitherto unavailable characteristics for electrochemical applications, while in-depth and systematic study of graphene materials is necessary to understand its electrochemistry. Hence, new synthesis methods were explored using carbon nanotubes as a precursor to produce graphene materials, with analysis on the types of products formed and the effect of residual metallic impurities. A novel method for the reduction of graphene oxide nanoribbons was demonstrated to afford graphene nanoribbons with superior properties than those using conventional chemical reduction. The synthesis of graphene materials using rapid microwave treatment was also studied to provide insight on the influence of the precursor graphite oxide, with the procedure then being modified to successfully synthesize sulfur-doped materials that exhibited electrocatalytic activity. On the other hand, comprehensive investigations on current synthetic procedures were also conducted, revealing the impact of different methods on the levels of metallic impurity contamination of the graphene materials. The choice of graphite source was also shown to have varying, compounding effects on the properties of the resulting graphite oxides and graphene materials. Different techniques for the transfer of CVD graphene were compared in order to determine the most suitable method to obtain graphene films for electrochemical applications. These studies thus serve the dual purpose of expanding the synthetic toolbox available for graphene production as well as provide insight on the suitability of graphene materials for various electrochemical applications. ​Doctor of Philosophy (SPMS)

This thesis focuses on techniques to improve the performance of voice conversion. Voice conversion modifies the recorded speech of a source speaker towards a given target speaker. The resultant speech is to sound like the target speaker with the language content unchanged. This technology has been applied to create personalized voice in text-to-speech or virtual avatar, speech-to-singing synthesis or spoofing attacks in speaker verification systems. To perform voice conversion, the usual approach is to create a conversion functions which is applied on the source speaker’s speech features such as timbre and prosodic features, to generate the corresponding target features. In this past decade, most of voice conversion researches had focused on spectral mapping, i.e. conversion of the features representing the timbre characteristics in a frame by frame manner. In chapter 3, we investigate a comprehensive approach to train the conversion function using DNN which considers both timbre and prosodic features simultaneously. For better modelling, we have used high-dimension spectral features. However, this further worsen the ability to robustly train a DNN which typically requires large amount of training data. To overcome the issue of limited training data, we propose a new pretraining process using autoencoder. The experimental results show the proposed comprehensive framework with pretraining performs better than conventional voice conversion systems including the state-of-the-art GMM-based system. The technique introduced in chapter 3 only learns a DNN system to convert between a pair of speaker. To reduce the need for parallel training data of new speaker pair, in chapter 4 we examine a novel DNN adaptation technology for voice conversion by including two bias vector representing both source and target speaker. By this configuration, new speaker pair conversion are archived. Our preliminary results show that conversion to new target speakers’ voices could be achieved. Master of Engineering (SCE)

With increasing importance placed on energy efficiency and indoor-air-quality, the advanced energy efficient liquid desiccant assisted air-conditioning (LDAC) systems have attracted attention in the recent years. The thesis aims to develop a vacuum graded LDAC system with multi-stage-effect dehumidification and regeneration systems employing membrane technology namely VMEMD for improving the efficiency of the conventional LDAC system. The proposed multi-effect regeneration test-bed maximizes not only the heat transfer between the heating fluids and salt solution but also mass transfer of water vapour across the hydrophobic membrane. Subsequent findings are obtained through the experimental investigations of multi-effect membrane based LDAC system employing various control parameters of the test facility. We found that the conventional VMEMD is not suitable for the regeneration of LiCl solution for dehumidification purposes. Therefore, the VMEMD system has been modified with one steam three solution (1S-3L) passes system through modelling and simulation. Finally, an experimental investigation is conducted to prove the test-of-concept. The use of 1S-3L is the main findings of this work. The thermal efficiency of the proposed 1S-3L VMEMD system is calculated in terms of performance ratio (PR) and change in feed concentration between the inlet and outlet of the regenerator for various solution flow rates and regeneration temperatures. The proposed system is driven at the lowest temperature of 55°C. The PR value is found ~2 for VMEMD system, which can be considered a new finding. In this thesis, a comprehensive testing is conducted for the thermal performance of the conventional VMEMD regenerator of the LDAC test bed under various configurations. The desiccant heating load is handled by both the solar thermal and auxiliary heat pump systems. It is found that the PR is mainly influenced by the initial feed concentration and flow rate delivered from the absorber module. The thermodynamic frameworks of the proposed 1S-3L VMEMD system is developed employing the heat and mass balances for all the stages of VMEMD system, which simulates the working principles of the regeneration process for various driving heat source temperatures. The simulation result proves the validity of changing VMEMD configuration for handling higher inlet feed concentration. The present study shows a test-of-concept to utilize the modified VMEMD for dehumidification purposes in tropical climate. Doctor of Philosophy (IGS)

Multidimensional hybrid systems play a major role in most applications of modern materials science, including relevant research areas of energy conversion and storage (solar cells, batteries), chemical and biochemical processes (photocatalysis, drug delivery, sensors), nanoelectronics (LEOs, FETs), and nanophotonics (light sources and amplifiers, plasmonic waveguides and modulators). By taking advantage of two or more components with distinctly different physical and electronic properties, for instance organic and inorganic constituents, hybridization at the atomic level and exploitation of dimensionality effects allow designing nanostructured material with properties "on demand". From the theoretical standpoint, the in-depth understanding of structure-property relationships of such inhomogeneous systems poses new challenges: methods that proved successful for describing homogeneous components may not be applicable to the hybrid systems, and are likely to fail to capture the physics of abrupt heterointerfaces. In this thesis, we consolidated a theoretical framework to study hybrid nanostructured materials by a combination of computational strategies and first-principle approaches, including density functional theory (OFT), time-dependent OFT and many-body perturbation theory. Within this framework, we studied the structural, vibrational, electronic and photophysical properties of model material systems with different dimensionality and functions, including polymer/fullerene blends, polymer/Ill-V heterointerfaces, organic-inorganic perovskites for photovoltaic applications, and metallic/dielectric topological insulators for active plasmonics. Our main findings are the following: i) a complete description of the vibrational fingerprints of polaronic and excitonic states m conjugated polymers; ii) the quantification of dimensionality and surface polarity effects on charge transfer and separation in hybrid polymerllll-V (film and quantum dots) photovoltaic systems; iii) the demonstration of dimensionality and self-trapping effects on charge transport and photoluminescence of multidimensional organic-inorganic perovskites for photovoltaic and light-emission applications; iv) the determination of composition-dependent dielectric and metallic optical properties of topological insulator crystals for low-loss plasmonic devices operating in the visible part of the spectrum. Our predictions compare favorably with structural and spectroscopic data of the actual hybrid material systems, demonstrating the flexibility and potential of first principle calculations for the computational discovery of novel material systems, and their optimization for real-life applications. ​Doctor of Philosophy (SPMS)

Electrohydrodynamic Atomization (EHDA), also referred to as the electrospray process, has been recently developed to fabricate polymeric micro- or nano-particles as well as fibers for drug encapsulation. EHDA in the cone-jet mode can produce highly monodisperse aerosols. As a result of strong electric forces, liquid droplets can be created by the electrospray process and converted into solid particles. In this study EHDA was carried out in a glass chamber in which solid pharmaceutical particles were fabricated after evaporating the solvent from sprayed droplets. A high voltage nozzle with a concentric ring and a grounded collection plate were enclosed in an EHDA glass chamber. A modified design for the glass chamber was used in this study in which the particles are collected inside the chamber. The effects of different parameters on the particle collection efficiency were experimentally investigated. Important factors affecting the particle collection efficiency were the flow rate of the liquid solution used to create the particles, the nitrogen flow rate and the nozzle voltage. Systematic scaling analysis was conducted on the governing equations in the current EHDA system. The final results confirmed that the particle acceleration, gravitational and Saffman lift forces are not important in comparison to the electrical field and flow field forces for the EHDA apparatus and conditions used in this study. Furthermore, scaling analysis of the describing equations for the electrospray process permitted collapsing all the collection efficiency data onto a universal plot as a function of one dimensionless group that incorporates all the process parameters. This plot can be used to determine the effect on the collection efficiency of changing the liquid solution flow rate, nitrogen flow rate and nozzle voltage as well as geometrical parameters such as length and diameter of the chamber, and location of the collection plate.

The interfacial properties of Ni clusters grown on the stoichiometric and reduced rutile TiO2(0 0 1) surfaces were investigated by means of X-ray photoelectron spectroscopy (XPS). The binding energies (BE's) of elements from both overlayers and substrates were found to be affected by the formation of interfacial dipole. Regardless of the TiO2 surface stoichiometry, the Ni 2p3/2 BE's move monotonically toward lower value with the increase of Ni thickness due to the cluster size effect. However, the Ni 2p3/2 BE shift is much smaller on reduced TiO2(0 0 1) surfaces compared to that on the annealed stoichiometric surface. For stoichiometric and lightly reduced TiO2 surfaces, O 1s BE's exhibit an unexpected upward shift with increasing Ni thickness below 2 A, and then downward shift to lower BE's when the Ni thickness increases further. This opposite tendency is attributed to the insulator-to-metal transition. On heavily reduced surface, only monotonically downward shift of the O 1s BE's was observed with the increase of Ni thickness. The different behaviors are well elucidated by collective contributions of interfacial charge transfer and image charge effect.

In coding theory, quasi-twisted (QT) codes form an important class of codes which has been extensively studied. We decompose a QT code to a direct sum of component codes – linear codes over rings. Furthermore, given the decomposition of a QT code, we can describe the decomposition of its dual code. We also use the generalized discrete Fourier transform to give the inverse formula for both the nonrepeated-root and repeated-root cases. Then we produce a formula which can be used to construct a QT code given the component codes.

A model for the dry-casting process for polymeric membrane formation is generalized to include a broader spectrum of boundary conditions. This model is used to conduct a sensitivity analysis to assess which properties and process parameters are most influential and hence need to be determined most accurately. The model predictions are relatively insensitive to changes in the nonsolvent/polymer and nonsolvent/solvent friction coefficients but moderately sensitive to changes in the solvent/polymer friction coefficient. However, the predictions are very sensitive to changes in the gas-phase mass-transfer coefficient, which needs to be well-characterized and carefully controlled during the dry-casting process. Experimental studies for the water/acetone/cellulose acetate system demonstrate that changes in the boundary conditions can result in dramatically different membrane morphologies. The results of the modeling, sensitivity analysis and experimental studies are used to advance a simple phenomenological model for the dry-casting process that explains all the observed trends. This is used to develop a set of design heuristics for dry-casting based on the effect of changes in the properties and parameters on the casting solution and gas-phase mass-transfer resistances, the solvent concentration at the interface and the mass-transfer rate.

Graphitic carbon nitride (g-C3N4), an emerging graphene-like two-dimensional (2D) material, has received wide attention by its fascinating photocatalytic performances as a photocatalyst in a variety of photocatalytic reactions such as photocatalytic water splitting and photocatalytic degradation. Owing to a bandgap of 2.7 eV, g-C3N4 is able to make use of the renewable and sustainable solar energy and conduct photocatalytic reaction under visible-light. Moreover, g-C3N4 is inexpensive, abundant and eco-friendly. It also has high thermal and chemical stability. Therefore, researchers believe that g-C3N4 is the most promising photocatalyst to serve in environmental protection and energy conservation. Studies on the chemical, optical and photocatalytic properties of g-C3N4 is very important and of great interests to the public. In the thesis, we report the synthesis, characterizations, and photocatalytic application of g-C3N4 products. High-yield synthesis of g-C3N4 powder via sealed heating method and half-sealed heating method are demonstrated. Based on the g-C3N4 powder, the g-C3N4 films are further produced by vacuum heating deposition and thermal evaporator. The chemical composition and crystallization of the as-prepared g-C3N4 products are demonstrated by multiple characterizations such as thermal gravimetric analysis, X-ray diffraction, X-ray photoelectron spectroscopy, transmission electron microscopy, scanning transmission electron microscopy and electron energy loss spectroscopy. On the other hand, a systematic optical study of g-C3N4 is carried out with several approaches. Using infrared and Raman spectroscopy, the chemical band vibrational modes of g-C3N4 products are revealed, and the absorption and emission property of g-C3N4 are investigated by UV-visible spectroscopy and photoluminescence spectroscopy. Moreover, the photocatalytic property of g-C3N4 is explored by the photodegradation experiment. We also build g-C3N4\hematite heterojunctions and study its photocatalytic water splitting performance. ​Doctor of Philosophy (SPMS)

Aquaporins are a family of small, integral membrane proteins which facilitate water transport across the plasma membrane of cells in response to osmotic gradients. Since the discovery of these water channels more than 20 years ago, there has been an encouraging advancement in the understanding of their structural and functional roles in human physiology, in particular in water balance disorders, and in cell migration and angiogenesis. However, the progress in aquaporin-based drug discovery is challenged by lack of suitable assays for the identification of modulators, as well as the overall poor druggability of aquaporins. In this work, we have developed a novel high throughput functionality assay which has the potential for biotechnological applications such as drug screening as well as the identification of aquaporin functional mutants. Several biophysical techniques have been employed to assess the binding and inhibitory efficiency of the hits to the human aquaporin 1. Overall, we hope that this research effort can play a part for future treatment of the devastating human diseases related to aquaporins. ​Doctor of Philosophy (SBS)

NT5C2 relapse-associated mutations, K359Q and R367Q, are correlated with large gains in their enzymatic profiles. Crystal structures obtained in this study provide insights to the allosteric regulation mechanism of the enzyme by the disorder-to-order transition of a regulatory helix, which forms part of the active site. In addition, the mutant enzymes possess an increased basal level of activity and sensitivity to activation by effectors. In this study, the structural aspects of how each mutation potentially gains an advantage over the wild-type protein is revealed, with supporting evidence from biochemical and biophysical assays. ​Master of Science

The aim of this study was to identify and understand the factors that contribute to the effectiveness and efficiency of knowledge transfer during employee turnover, under time constraint. This was done by analysing the case of International Enterprise Singapore (hereafter referred to as IE Singapore), a government agency that spearheads Singapore's external economy. Master of Science (Knowledge Management)

Typical 2-Cys peroxiredoxins (Prxs) are mainly involved in antioxidant defense and the signal transduction pathway of H2O2 by undergoing inactivation once intracellular H2O2 concentrations exceed the threshold. This inactivation mechanism is only present in eukaryotes. In order to glean molecular insights into the sensitive balance behind this mechanism, Escherichia coli mutant and chimeric proteins such as EcAhpC1-186-YFSKHN were generated. A combination of crystallographic and stopped-flow spectroscopic techniques were employed so as to gain greater insight into the essential role of the C- terminal tail in the robustness of prokaryotic Prx. Based on enzymatic and structural assays, the transition steps between the main conformational states of Prx were determined, and a physical linkage between the C-teminal tail and the oligomer interface was established. Furthermore, the C-terminal tail of EcAhpC was identified as a molecular switch between the conformational states and is proposed to be the determining factor of enzymatic inactivation. Masters of Science

The aim of this dissertation was to investigate non-catalytic and catalytic synthetic routes for crucial ROS-resistant deuterium labelled essential polyunsaturated fatty acids (PUFAs). In order to perform selective catalytic H/D exchange of the bis-allylic position in methyl linoleate, a new complete series of cyclopentadienyl ruthenium complexes [(η5-C5MenH5-n)Ru(κ2-P,N-tBu2PIm)(CH3CN-κN)]PF6 has been successfully developed. In accordance to our initial proposal, structural modifications in these complexes did restrict the most deleterious isomerization pathway that previously resulted in conjugation of double bonds in methyl linoleate. Additionally, we conducted a series of kinetic and mechanistic studies of isomerization and deuteration reactions catalyzed by these complexes using simple terminal olefin as a substrate. These studies contrasted the influence of steric hindrance degree in their structure on their activity towards alkene isomerization. Finally, these complexes have been characterized by means of 1H, 13C and 31P NMR spectroscopy, mass spectroscopy and single crystal X-Ray analysis that provided evidence towards structural changes arisen from increased steric hindrance of cyclopentadienyl moiety. ​Doctor of Philosophy (SPMS)

Light emitting diodes (LED), is used prevalently in many current lighting applications. Creating high powered and efficient LEDs decreases energy wastage, bringing forth environmental and financial benefits. We propose an original method of increasing the efficiency of LEDs by fabricating epoxy-based micro-hemisphere of different spacing on the surface of the LED using a micro-plotter, through a simple and inexpensive fabrication process. This new methodology will improve the power output, external quantum efficiency and tune the emission color of Gallium Nitride (GaN) LED, with a cheaper and less complex procedure as compared to current methods used in the industry. From the results, a 10 mA injection current and 50!!m separation of the micro- hemisphere is able to achieve 23 % increase in the power output, with an enhancement factor of 1.23 in the external quantum efficiency as compared to one without the micro- hemisphere arrays. Internal heating within the LED is also reduced significantly as observations show a 0.79 nm decrease in red shift from the photoluminescence spectrum of the LED with fabricated micro-hemisphere array. Reduced internal heating will improve the lifetime of the LED devices. The color of the LED is tuned by doping the fabricated micro-hemisphere with suitable dye material. It is shown that by doping with Rhodamine 6G (R6G) and Coumarin 540A (C540A) dye, GaN LED is able to emit red and green light through the area fabricated with the micro-hemisphere array. When color emission can be properly controlled via doping, white color LED light with more practical uses can be achieved. Master of Science

Natural photosynthesis processes have as their first step light harvesting (LH) in which excitation energy transfer occurs from units with higher energy gaps to those with lower energy gaps, and as their second step coupled electron transfer (oxidation/reduction) reactions to produce organic compounds. The exquisite organization of multiple chromophores in LH complexes plays a crucial role in enhancing the efficiency of photosynthesis, and investigations of artificial light-harvesting systems have looked at both covalent and non-covalent assembly techniques for obtained such ordered arrangements of chromophores, while their photophysical investigation has been extended from linear one-photon (1P) excitation to multi-photon (MP) excitations. The features of Zn2+ and terpyridine complexation including facile preparation, and a highly directional, kinetically liable assembly of units are of interest as one could construct potential light harvesting prototypes by the insertion of luminescent units within such an assembly. Two carbazole bisterpyridyl ligands: 1-H (3,6-bis(4-([2,2':6',2''-terpyridin]-4'-yl)phenyl)-9H-carbazole) bearing a hydrogen atom, and 1-C12 (3,6-bis(4-([2,2':6',2''-terpyridin]-4'-yl)phenyl)-9-dodecyl-9H-carbazole) with a dodecyl chain at the 9-position, were synthesized for comparison of the effect of H-bonds on their assembly behavior with Zn2+. Ligand 1-H formed a single pentametric Zn2+ metallocycle, as confirmed by NMR, ESI-MS and X-ray photoelectron spectroscopy (XPS) (Zn-1-H), whereas 1-C12 produced a mixture of two metallocycles (Zn-1-C12). Molecular packing and assembly of Zn-1-H studied by microscopy showed it has a high propensity to form ordered layers. Spectroscopic studies revealed 1-H and its Zn2+ coordination complex (Zn-1-H) display rare excitation wavelength dependent dual blue or green emission either under both 1P and MP excitation circumstances, and multi-stimuli-response behaviours. The emission of the complex Zn-1-H could be facilely manipulated to obtain either pure blue or green emission by respectively changing the solution basicity through addition of triethylamine or tetraethylammonium hydroxide, or by changing the ion strength by adding anionic polyelectrolyte-polystyrene sulfonate (PSS), whereas the emission from the free ligand 1-H did not exhibit such complete transformations. The metal ion selectivity towards Fe2+, Cu2+, and Co2+ resulted in a superior binding through secondary amine of 1-H for Cu2+, and substitutional binding of Fe2+ through terpyridine units in Zn-1-H. A second bisterpyridine based on a red emitting dithienyl benzothiadiazole unit as a photonic acceptor was combined with two different donor bisterpyridine units (D, bisterpyridine linked respectively to carbazole at 3,6- (1-H), and 2,7- positions (2) (blue-greenish emission) to form statistical Zn2+-tpy copolymers P1 and P2. Spectroscopic studies revealed an approximately 4 fold enhancement of A emission for both P1 and P2. Transient fluorescence studies revealed decreased lifetime and quantum yield of D unit in P1 compared to 1-H, indicating a FRET mechanism in P1. By contrast, the fluorescence decay curve was unchanged and re-absorption was suggested as the energy transfer mechanism in P2. Moreover, the large multiphoton absorption cross-section in the donor part also resulted in obvious enhanced A emission under two- and three- photon circumstances. The discovery of TICT state facilitated one- to multi- photon energy transfer can be used to help design novel and efficient light harvesting systems. Doctor of Philosophy (MSE)

Human Coronavirus 229E (HCoV-229E or 229E) is one of the six human coronaviruses associated with multiple respiratory illnesses. Despite its global prevalence, the viral-host interactions of 229E have not been well studied. Here, we demonstrate that 229E replicated efficiently and induced apoptosis in lung adenocarcinoma H1299 cells. The c-Jun N-terminal kinase (JNK) pathway was activated in 229E-infected H1299 cells. Of the two known upstream protein kinases, MKK7, but not MKK4, is responsible for JNK activation during 229E infection. Although JNK did not significantly affect the overall 229E-induced apoptosis, it served an anti-apoptotic function by modulating B-cell lymphoma 2 (Bcl2) and Bcl2 homologous antagonist killer (Bak) in a mechanism not mediated by c-Jun. Moreover, JNK also mediated the induction of interferon β (IFNβ) and interleukin-8 (IL8) at transcriptional levels during 229E infection. Collectively, our findings show that JNK activation regulates expression of Bcl2 family proteins and innate immunity during 229E infection. Master of Science

Coronary Artery Disease (CAD), one of the most common heart disease, is the leading cause of death in the world. One of the main causes for CAD is atherosclerosis. There have been many studies done to identify genes responsible for atherosclerotic plaques formation. Recently, Genome-wide association studies (GWAS) done by Ghosh et al. has shown a strong association between semaphorin pathway with CAD. Additionally, the protease Furin was found to be a member in 6 of the 32 replicated, CAD associated pathways. To follow up for these novel findings, present studies examined the regulation of Semaphorin pathway gene expression and the effects of Furin knock-down in RAW264.7 macrophages. CAD characterized cells were induced using human high OxLDL and Furin knock-down was carried out using siRNA. Preliminary results suggest that Semaphorin pathway could be regulated by OxLDL with decrease in NRP1 and PLXNA2 expression. Presence of OxLDL also reduced the expression of Sma and Icam1 by more than 50%. Knock-down of Furin reduced expression of IL-1B, CD86 and CD205 by more than 50%. Future studies will include measurement of protein levels for candidate gene and examine of OxLDL and Furin knock-down effect in endothelial and vascular smooth muscle cells. Doctor of Philosophy (SBS)

Terahertz time-domain spectroscopy (THz-TDS) is a powerful tool which has been used to probe the low-energy excitations in metals and semiconductors. The THz transmission signal is sensitive to the presence of free carriers, phonons, excitons, and the interplay among these degrees of freedom. The most remarkable advantage of THz-TDS is the simultaneous acquisition of the amplitude and phase of the THz pulses, which allows us to obtain the complex permittivity (or equivalently, complex optical conductivity) without using the Kramers-Kronig relationship. The ability to obtain the real and imaginary parts of the optical conductivity severely restricts the physical models that can explain the data. In this thesis, after introducing the experimental setup and data analysis procedure, the main experimental results are presented. Topological insulators (TIs) are fascinating materials with an insulating bulk and conducting surface. The surface states of topological insulators display Dirac-cone dispersion in which the electrons obey a linear energy-momentum ‘relativistic’ relationship. Due to the very small photon energy of THz radiation, the photons can be absorbed, causing intra band transitions in the Dirac cone. By performing THz measurement on a composite of Bi2Te3 TI nanoparticles deposited on single-layer graphene (SLG), we observed an enhanced THz modulation compared to pure SLG, which suggest that there is electron injection from the TI nanoparticles into the graphene. Our results also reveal the presence of multiple phonon modes in Bi2Te3 thin film in our frequency window of 0.3 – 3 THz. Doped ZnO is a type of transparent conductive oxide which can be used in organic photovoltaics. The properties of transparency and conductivity are both desirable, but are often mutually exclusive. To improve the conductivity while retaining the transparency, a good way is to improve the carrier mobility. Our THz-TDS results show that the Ag nanoparticles can improve the electron mobility of Al-doped ZnO. The recently-discovered organometallic halide perovskite CH3NH3PbI3 has been shown to achieve a photovoltaic efficiency of up to 19%. By measuring the THz con-ductivity via THz-TDS, we observed that two phononmodes in the high-temperature tetragonal phase split into four phonon modes in the low-temperature orthorhombic phase. The presence of new phonon modes in the low-temperature orthorhombic phase might have implications on the photovoltaic efficiency at low temperatures. ​Doctor of Philosophy (SPMS)

The aim of this project is to further develop the application of BSA and its analogues in asymmetric phase-transfer catalyzed alkylation reactions. In conclusion, we have developed the first highly enantioselective bis-guanidinium salt catalyzed phase-transfer alkylation reactions of cyclic ketones and silyl enol ethers by the using of BAS and its analogues. And we found that, for the alkylation reactions of cyclic ketones, the enantioselectivity can be improved a lot by tuning the structure of non-chiral BSA analogues, which opened a new way to achieve high enantioselectivity. In this catalysis system, the catalyst, fluorine source, BSA analogues, all the three parts are tunable to improve the enantionselectivity. Furthermore, as for the alkylation reactions of silyl enol ethers, we found that by the adding of catalytic amount of BSA, the yield was greatly improved with even better enantioselectivity, which provide a new strategy to solve the low yield problem of Mukaiyama type reactions. These results surely indicate the great potential of application of BAS and it’s analogues in asymmetric phase-transfer catalysis, which provides a new powerful strategy to active high pKa substrates and solve the low yield problem of Mukaiyama type reactions in asymmetric version. Firstly, a series of BSA analogues were synthesized. Then, we applied BAS and it’s analogues in bis-guanidinium salts catalyzed alkylation reactions. As for the alkylation reactions of cyclic ketones up to 99% ee was obtained, and for most of the substrates, the yield was excellent. As for the alkylation reactions of silyl enol ethers , up to 91% ee was obtained. At last, suggestions to further improvement were proposed and detailed experimental procedures were provided. ​Master of Science

After a review of Hamiltonicity of graphs and related concepts, we discuss several generalizations of Hamilton cycles: k-walks, k-trees, Hamilton-prisms and edge-dominating cycles, and investigate the relationship between them. In particular, we focus on the Jackson-Wormald conjecture and show that it holds for a graph with an edge-dominating cycle. The latter gives us our central result: an efficient algorithmic proof of Jackson-Wormald conjecture for 2K_2-free graphs. Another main result is that each (1+\epsilon) -tough 2K_2-free graph is prism-Hamiltonian. Generally, being prism-Hamiltonian is a stronger property than admitting k-walks for all k\ge2, but weaker than being traceable. Finally, we present several results on the existence of 2-walks under the 1-toughness assumption for some other graphs, and pose conjectures for further research. ​Doctor of Philosophy (SPMS)

In this thesis we concentrate on the usage of lattices to construct explicit sphere packings with high density, packing families with good asymptotic properties, and cryptographic schemes with special functionalities, including (delegatable) policybased signature and revocable identity-based encryption. First we introduce new methods using codes and lattices from number fields to construct dense packings in Euclidean spaces. Let ω = (−1 + √−3)/2. For any lattice P ⊆ Z^n, P = P + ωP is a subgroup of On K, where OK = Z[ω] ⊆ C. As C is naturally isomorphic to R^2, P can be regarded as a lattice in R 2n. Let P be a multiplicative lattice (principal lattice or congruence lattice) introduced by Rosenbloom and Tsfasman. We concatenate a family of special codes with t`P·(P + ωP), where tP is a generator of a non-zero prime ideal P of OK. Applying this concatenation to a family of principal lattices, we obtain a new family with asymptotic density exponent λ > −1.26532182283, which is better than −1.87 given by Rosenbloom and Tsfasman by considering only principal lattice families. For a new family based on congruence lattices, the result is λ > −1.26532181404, which is better than −1.39 obtained by considering only congruence lattice families. More generally, via the canonical Q-embedding of arbitrary number field K into R^[K:Q], both the prime ideal p and its residue field κ can be embedded as discrete subsets in R^[K:Q]. Thus we can concatenate the embedding image of the Cartesian product of n copies of p together with the image of a length n code over κ. This concatenation leads to a packing in the Euclidean space R^n[K:Q]. Moreover, we extend the single concatenation to multiple concatenations to obtain dense packings and asymptotically good packing families. For instance, with the help of Magma, we construct a 256-dimensional packing denser than the Barnes-Wall lattice BW256. Next we develop the recent techniques in lattice-based cryptography to construct a (delegatable) policy-based signature scheme and a revocable identity-based encryption scheme from lattice assumptions. Policy-based signature, introduced by Bellare and Fuchsbauer at PKC 2014, is a new type of digital signature in which a signer is only allowed to sign messages satisfying certain policy specified by the authority, but the signatures should not reveal the underlying policy. We adapt Langlois et al.’s zero-knowledge argument system (PKC 2014) for the Bonsai tree signature scheme (Eurocrypt 2010) to enable the prover to convince the verifier that its secret witness satisfies an additional condition. Making the protocol non-interactive via the Fiat-Shamir transformation, we obtain a policy-based signature scheme supporting polynomially many policies, which satisfies the two security requirements (simulatability and extractability) in the random oracle model. Furthermore, our construction can be efficiently extended to a delegatable policy-based signature, thanks to the hierarchical structure of the Bonsai tree. In view of the expiration or revelation of the user’s private credential (or private key) in a realistic scenario, identity-based encryption (IBE) schemes with an efficient key revocation mechanism, or for short, revocable identity-based encryption (RIBE) schemes, become prominently significant. We present an RIBE scheme from lattices by combining two Agrawal et al.’s IBE schemes (Eurocrypt 2010) with the subset difference method. In particular, our scheme may serve as a solution to a question posed by Chen et al. (ACISP 2012). DOCTOR OF PHILOSOPHY (SPMS)

G-quadruplex is a non-canonical yet crucial secondary structure of nucleic acids, which has proven its importance in cell aging, anti-cancer therapies, gene expression and genome stability. In this PhD thesis, we studied the five established structural models of human telomeric DNA G-quadruplexes on their thermal stabilities, folding dynamics and ligand binding behaviours. Firstly, temperature-replica exchange MD simulations were carried out by incorporating the five folding topologies as conformational isomers. The frequencies of folded topologies occurring at low temperature replicas indicate that the structure of hybrid-2 type is the most stable G-quadruplex folding topology. Next, parallel tempering metadynamics simulations in the well-tempered ensemble were performed to study the folding/unfolding dynamics of the hybrid-1 and -2 type G-quadruplexes. A hairpin and a triplex intermediate were identified from the free energy landscape of the hybrid-1 type, which is consistent with current experimental findings. More importantly, flip-overs of the N-glycosidic conformations of guanines were observed in the transition structures. It has been found out that the flip-over is to re-orient the directions of the Hoogsteen H-bond donor and acceptor of a guanine to participate in the cyclic Hoogsteen H-bonding for a new G-tetrad, which is usually but not necessarily accompanied by a conversion between the syn- and anti-conformations. Furthermore, ligand binding behaviours of G-quadruplexes with Thioflavin T (ThT) were investigated via conventional and well-tempered metadynamics (WT-MetaD) simulations. It is found that the five folding topologies of G-quadruplexes are differentiated in their binding modes as well as binding affinities with ThTs. The major binding modes are identified as the end, sandwich and base stacking via π-π interactions. Specifically, the sandwich stacking is triggered by the reversible conformational change on G-quadruplex and facilitated by the collaboration between ligands. Finally, a case study on the +1 nucleosomes of yeast was carried out. By extracting the +1 nucleosome sequences on the coding strand from the yeast genome and analysing the frequencies of occurrence of dinucleotides, it is revealed that the bending-resistant AA dinucleotide is more abundant in the first half of +1 nucleosomes. This result suggests that the unwrapping of +1 nucleosomes for DNA transcription in yeast might be mechanically regulated by positioning the rigid AA dinucleotide at the 5’ end. In summary, the work presented in this thesis provides original insights into the dynamics and folding of G-quadruplexes, which can be utilized by experimentalists to stabilize the structures or to intervene their formation in genome. Besides, the bioinformatics study on the +1 nucleosomes of yeast helps to realize the roles of DNA sequence on gene regulation, which will be further investigated in future work. ​Doctor of Philosophy (SBS)

Electro-osmotic flow (EOF) is the flow of fluid in a microchannel which is induced by an external elec-tric field. EOF originates from electrical double layer (EDL) formed spontaneously at the channel-fluid interface. The electrostatic potential at the channel wall is called the zeta potential. Nanostructures have been employed in microfluidic channels for electrophoretic separation of biomolecules [1] and catalytic reaction which yields higher reaction efficiency due to the increased surface area [2]. Howev-er, the effect of nanostructured surfaces on EOF has yet to be fully understood. In this preliminary investigation, we demonstrate experimentally that the presence of nanostructured surfaces affects EOF significantly. Accepted version

In recent years, the interplay between condensed matter physics and optics has resulted into several new research branches, such as topological photonics, novel functional optical devices, realization of quantum physics in optical waveguides, etc. In this thesis, we theoretically investigate several novel optical phenomena which are related to condensed matter physics. Specifically, we focus on optical phenomena in the newly discovered two dimensional material graphene, and a tight-binding lattice model with unusual topological and localization properties. In the first part of this thesis, we demonstrate that coherent perfect absorption(CPA) of light at terahertz frequencies is achievable in graphene. We also study the plasmonic properties of novel graphene devices. This includes the non-local dispersion relation of graphene surface plasmons and a proposed device to directionally couple incident light into propagating graphene surface plasmons. In the second part of the thesis, we generalize an experimentally feasible lattice model Aubry-Andre-Harper(AAH) model by introducing a phase difference between the on-site and off-diagonal modulation strengths. It turns out that the generalized model has new localization behaviors and novel topological phenomena. The new physics of the generalized model could be experimentally observed by coupled photonic waveguides or cold atom systems. Remove selected Master of Science

As the carrier of genomic DNA in eukaryotic cells, chromatin is one of the key factors involved in gene regulation. Many studies have revealed the general behavior of chromatin, whose basic feature is the assembly of nucleosome into array fibers. It has been identified that many factors, like cation environment, histone tails, post-translational modifications, linker histones and linker DNA, have great importance for chromatin higher order structure. Chromatin compaction is necessary for packaging of DNA in the nucleus and occurs by both intra-array folding and inter-array fiber-fiber association. However, little is known about the mechanism of inter- and intra-array interaction at the molecular level. A close contact between the surfaces of the nucleosomes, called nucleosome stacking, is a common structural feature of the condensed state of chromatin, been observed in almost all condensed chromatin. As one of the most significant post-translational modifications, acetylation of the N-terminal tail of histone H4 aa lysine 16 (H4K16) was proved to be involved in many cellular processes like transcription and DNA double strand break (DSB) repair. Several in vitro studies with reconstituted nucleosome arrays indicated that this modification disrupted nucleosome stacking leading to partial unfolding. In this thesis, we adopted a novel approach to investigate the importance of the H4 tail mediated nucleosome interaction on array folding and self-association (fiber-fiber interaction). We reconstituted nucleosome arrays using 12-601-177 DNA and various K16 mutated or modified Xenopus histone H4, including H4K16C, H4K16N, H4K16Q, H4K16R, H4K16me3, H4K16Ac and H4K16Q+CH2 (one thioether bond longer than H4K16Q), to investigate the mechanism of the H4K16 mediated chromatin folding and self-association using sedimentation velocity analytical ultracentrifugation (AUC) and precipitation assay measurements. Novel chemical approaches were employed to produce homogeneous and site-specifically modified H4K16 proteins. H4K16Ac was prepared by using a novel semi-synthetic approach we previously developed. H4K16 mutants and modifications were generated to reveal the possible factors that would affect the H4K16 functions. The importance of the H4K16 positive charge on the NCP-NCP interaction was investigated. Effect of side chain was also studied by comparing the modifications with conserved positive charge but bulkier side chain or varied length of side chain on H4K16. Our result showed the significance of positive charge of H4K16 in affecting nucleosome- nucleosome stacking and chromatin compaction. With the loss of positive charge, arrays became less compact compared to wild type array under high salt concentration. The length of side chain also displayed an important effect for maintaining the H4 tail mediated NCP-NCP interaction. Under the maximal folding condition, arrays with shorter side chain at H4K16 were found to be less compact. H4K16N, with a charge-removed and shorter side chain, failed to interact with surrounding residues of adjacent NCP and showed a similar disruptive effect on array folding as H4K16Ac. The detrimental effect on chromatin folding caused by H4K16Ac was proposed to be the consequence of cumulative effect of charge removal and bulky size of side chain. The interaction of the histone H4 N-terminal tail positive charge region (R17-R23) and the H2A acidic patch region is highly important interaction for NCP stacking. We investigated how charge changes of residues in these two regions affected array folding and self-association. Mutation constructs with positive charge removal on the H4 positive charge region include double mutations (R19L and K20L, RK-LL) and triple mutations (R17L, R19L and K20L, RRK-LLL). The mutant construct with charge neutralization of the residues in the H2A acidic patch was H2A-STT (D90S, E91T and E92T). It was found that the disruptive effect on stacking and folding caused by positive charge neutralization on the H4 N-terminal tail positive charge region (RK-LL and RRK-LLL) was similar to the effect caused by negative charge neutralization on the H2A acidic patch (H2A-STT). No cumulative effect on array unfolding was observed in array with charge neutralization on both these two regions (STT+RRK-LLL). In our study of nucleosome array self-association, the efficiency of salt-induced self-association behavior was generally in correlation with the change of charges of arrays. However, with no change of net charge, array reconstituted with H4 mutations (RRK-LLL) and H2A mutations (STT) required more Mg2+ to self-associate than wild type array, which was possibly due to the irrecoverable H2A-H4 interaction. Our result demonstrated the role of H2A-H4 interaction on chromatin self-association. Besides that, we also investigated how nucleosome arrays behaved under mixed-ion circumstances: (1) K+ and Mg2+, (2) Na+ and Mg2+. It was found that the addition of K+ or Na+ promoted the efficiency of Mg2+ induced self-association, while K+ and Na+ played opposite roles in Mg2+ induced chromatin folding. ​Doctor of Philosophy (SBS)

This thesis is concerned about the asymptotic behavior of the largest eigenvalues for some random matrices and their applications for high-dimensional data analysis. The first type of random matrix is the F-type matrix. More precisely, let $\bbA_p=\frac{\bbY\bbY^*}{m}$ and $\bbB_p=\frac{\bbX\bbX^*}{n}$ be two independent random matrices, where $\bbX=(X_{ij})_{p \times n}$, $\bbY=(Y_{ij})_{p \times m}$ and all entries are real (or complex) independent random variables with $\mathbf{E}X_{ij}=\mathbf{E}Y_{ij}=0$, $\mathbf{E}|X_{ij}|^2=\mathbf{E}|Y_{ij}|^2=1$. Denote the largest root of the determinantal equation $\det(\lambda \bbA_p-\bbB_p)=0$ by $\lambda_{1}$, which can be considered as a general version of the largest eigenvalue of the F matrix. We establish the Tracy-Widom Law for $\lambda_{1}$ under some mild conditions when both $\frac{p}{m}$ and $\frac{p}{n}$ tend to some positive constants as $p\rightarrow\infty$. More applications are given based on this limiting distribution of $\lambda_1$. The second type of random matrix is the sample covariance-type matrix. This type of random matrix is proposed to deal with the high-dimensional change point detection problem. By modifying the classic sample covariance matrix and then developing the corresponding asymptotic behavior of its largest eigenvalue, without any estimators, this thesis suggests an optimization approach that can figure out both the unknown number of change points and multiple change point positions simultaneously. Moreover, the largest eigenvalue of this matrix type can also be extended to other popular high-dimensional testing problems. Except for the theoretical results established for each random matrix type, various simulation studies are also provided in order to support these conclusions and to show the effectiveness of the proposed statistics. ​Doctor of Philosophy (SPMS)

Cytosine methylation is a critical mechanism for epigenetic regulation on the molecular level, which has profound impacts on heritable gene silencing and plays an essential role in mammalian cell development. Recent discoveries suggest that ten eleven translocation proteins (TETs) can oxidize methylcytosine (mC) to generate 5-hydroxymethylcytosine (hmC), 5-formylcytosine (fC), and 5-carboxylcytosine (caC). The molecular mechanism of DNA demethylation and epigenetic regulation system has been revised. Numerous studies focusing at these epigenetic cytosines have been carried out to explore their biological roles. Although a conclusive answer is not yet known, tremendous progress has been achieved from timely developments of technologies that allow for specifically enriching, detecting and sequencing epigenetic cytosines. Genomic sequencing of hmC in genomic DNA reveals that hmC may be associated with both gene expression activation and inhibition, which is considered as a “dual regulation” function. However, it is inconsistent with previous observation that hmC shows only destabilization effect to DNA duplex. In Chapter 2, we have shown that the destabilization effect of hmC on CpG repeats is highly dependent on the overall methylation level of the flanking DNA regions. The phenomena can be observed in either aqueous buffer or crowding condition when either A-form DNA-RNA hybrids or B-form DNA-DNA duplex were investigated. The results provides better understanding of epigenetic regulation via manipulating helical thermostability and may supply knowledge to alternative mechanisms of RNA transcription and intron-exon recognition during mature mRNA formation. As the oxidation product of mC, hmC and fC are proposed to be a key intermediate of active demethylation pathway. Once mC is oxidized to fC, it is recognized and excised by thymine DNA glycosylase (TDG), subsequently restored as normal C with the help of base excision repair (BER) system. To further understand the function of hmC and fC, chemical methods to genomic mapping of fC and hmC would be highly appreciated. In Chapter 3, we have established an effective method to detect fC and hmC in DNA strands. We have found that piperidine and NaOH treatment is able to selectively react with formylcytidine and cause site-specific DNA strand cleavage. Therefore, the position of fC in DNA strand can be profiled by analyzing strand cleavage site. Combine this method with KRuO 4 induced hmC oxidation, sequencing of hmC can be achieved in single base resolution with good selectivity and high efficiency. ​Doctor of Philosophy (SPMS)

The topic of learning receives diverse perspectives in psychological cognitive and behavioral analysis. In this thesis, an evolutionary computational model is proposed to integrate both the cognitive and behavioral point of view. The constructed artificial intelligence (AI) was made to play Tic-Tac-Toe against experienced human opponents. The observed learning behavior of the AI, exhibited signs that could suggest ‘conscious learning’ behavior. This dialectic of learning is based on the hypothesis that learning is an evolutionary process (behavioral) mediated by certain internal memory organizing mechanism (cognitive). Genetic algorithm and cellular automaton are the two fundamental logic behind this model. Here in the game model, the AI is able to both successfully figure out rule(s) and strategies of the game. Adopting this approach, a simplified version of the AI is constructed to model infant language acquisition. In this infant model, we are able to mimic an infant’s initial phase of language development through the first three levels of consciousness via social interactions. This result could be of paramount importance in identifying infants with difficulty in acquiring language. Master of Science

When the world goes flat, we see through it differently. The key to reveal the secrets of low dimensional world lies within one class of newly emerged material – Two dimensional (2D) nanomaterials. During past few years, two dimensional nanomaterials have drawn great attention for its simple physical model in theoretical research and unique properties in practical application. In this thesis, we concentrate on the optical properties related to the ultrafast dynamics of three typical families of 2D nanomaterials: graphene, transition metal dichalcogenides (MoS2, WSe2). A comprehensive investigation on the ultrafast dynamics studies of these 2D nanomaterials is conducted through a combination of several ultrafast spectroscopy techniques, such as time resolved photoluminescence (TRPL), transient absorption spectroscopy (TAS) as well as supplementary material characterizations including Raman Spectroscopy, x-ray photoemission (XPS), scanning electron microscopy (SEM) and etc. Graphene, as the most classic 2D nanomaterial, possesses unique conical band structure and exceptional optical properties. Presently, high quality large area graphene which is widely used as the flexible transparent electrode, has been successfully prepared through chemical vapor deposition (CVD). Revealing the charge carrier dynamics of this CVD graphene is crucial for its related optoelectronic device. The application of CVD graphene requires the polymer-assisted substrate transfer process, the possible polymer residual macromolecule interaction with the graphene surface address this issue great research significance, the understanding of the properties in this polymer-grafted graphene region is far and few between. There are three processes of the carrier relaxation dynamics in graphene after initial excitation mainly consists of three processes; electron-electron scattering (< 10 fs), electron-optical phonon coupling (500 fs ~ 2 ps), optical phonon and acoustic phonon coupling in sequence. A systematical investigation was performed to clarify the carrier relaxation channel modulation in polymer-grafted region compared to pristine graphene. Meanwhile, the obvious features of Auger recombination and hot phonon bottleneck effect were also observed in the polymer-grafted graphene. Another important family of 2D nanomaterials is transition metal dichalcogenides (TMD). The monolayer MX2 TMD obtains binary hexagonal lattice and intrinsic optical bandgap ranging from the near infrared to visible. Theory predicts that many stacked MX2 heterostructures form type II semiconductor heterojunctions that facilitate efficient electron–hole separation for light detection and harvesting. Regarding all possible combination of 2D crystals, MoS2 and WeSe2 van der Waals heterosturctures is the only type II p-n junctions, spatial temporal dynamics studies was conducted on this heterostructures, and we observe that hole transfer from the MoS2 layer to WSe2 layer occurs with 100 fs after the optical excitation. Such ultrafast charge transfer in van der Waals can enable advanced 2D devices requiring fast processing. CH3NH3PbI3 is one of the most extensively studied materials due to its strong optical absorption and long carrier diffusion length. Integrating the strengths of 2D nanomaterial and CH3NH3PbI3 on device designing is one promising field for the future development of photovoltaics. We achieve to tune the band gap offset of MoS2 flakes through mild oxygen plasma treatment to form a type II interface in 2D MoS2/CH3NH3PbI3 hetero-structure. The ultrafast hole transfer process is observed in the heterostructure interface which suggest that vacancy tuned MoS2 a perfect hole transporting layer for the perovskite-based solar cells. In summary, our spectroscopic studies of these 2D nanomaterials provide valuable information on relaxation dynamics analysis and pave the way for further applications of novel 2D nanomaterials based device for optoelectronic and light harvesting. ​Doctor of Philosophy (SPMS)

Bioadhesive mechanisms that employ external triggers for on-demand adhesion were developed to satisfy the following conditions 1) tunable adhesion 2) post-treatment modification on model and 3) application on wet, soft-tissue surfaces. A bioadhesive with these attributes could be activated on-site by an external trigger to overcome the challenges of non-specific protein binding and inadequate adhesion in wet conditions more often faced by clinicians who require better methods of soft tissue and medical implant fixation. Various plasma parameters were investigated for controlling model drug release and soft-tissue adhesion of polyester thin films. When compared to non-treated polyester thin films, plasma intensity could give rise to a ~ 6-fold (for 100 W, 5 min.) increase in shear adhesion strength (157 ± 24 mN/cm2). Shear adhesion was 5-fold higher than L-3,4-dihydroxyphenylalanine-based mussel adhesives tested on similar biomaterials. The results indicated that the adhesion strength was dependent upon plasma treatment time and plasma intensity, but was independent of the ratio of argon/oxygen generated plasmas. This plasma-mediated surface chemistry allows on-demand soft tissue adhesion of polyester thin films. The novel bioadhesive mechanism using diazirines that allows tunable adhesion using external stimuli was demonstrated and can be grafted on biodegradable polyester surfaces. These surface functionalized adhesive thin films works in wet condition by generating carbene radicals upon UV stimulation. The functionalized alkyl and aryl diazirine polyester films when tested for shear adhesion to ex-vivo swine aorta tissues, showed 450±50 mN/cm2 and 476±50 mN/cm2 adhesion strength when activated for 10 min and 2 min respectively using 0.5±0.1 mW/cm2 UV light of 365 nm. Thus, indicating the tuning of adhesion depending on time of UV irradiation and type of diazirines functionalized on polyester films. The covalent crosslinking of hydroxyethyl methacrylate polymer brushes grafted on quartz surface to wetted tissues were studied following long-range UV light activation. The potential applications of PHEMA polymer brushes through use of UV irradiation as a soft tissue adhesive in wet conditions were demonstrated for the first time. The technique demonstrate the dynamic and strong covalent adhesion of polymer brushes, which is reliable and straightforward, producing tissue-substrate bond that could retain their protein bio-functionality. Doctor of Philosophy (MSE)

Poor correlation of mRNA and protein levels suggests a role for translational control of gene expression during the Plasmodium intra-erythrocytic life cycle. Among the components of translational machinery, the dozens of ribonucleoside modifications on transfer RNAs (tRNA) are emerging as critical regulators of cell physiology and stress response. However, little is known about malaria tRNA modifications and their roles in gene expression, making it an unexplored area of Plasmodium translational biology. We have now characterized the repertoire of tRNA modifications of P. falciparum and identified critical functions in tRNA maturation and stage-specific translation. Remarkably, P. falciparum has a standard set of eukaryotic tRNA modifications that resembles the model organism Saccharomyces cerevisiae. We demonstrated that these modifications are tightly regulated in two ways throughout the intra-erythrocytic developmental cycle (IDC) of the parasite. First, we observed a synchronized increase in most of the modifications from ring to trophozoite stage, suggesting tRNA maturation to meet stage-dependent translational needs of the parasite. To better understand the relevance of these changes in modifications, we then performed iTRAQ proteomics to quantify the protein abundances across the IDC. Strikingly, the most abundantly expressed proteins in the late stage of the IDC, but not early stages, have a marked codon bias that can be directly correlated with the corresponding increase in tRNA modifications. Based on these results, we propose a new model for how tRNA modifications govern the late-stage of the malaria IDC by controlling the abundance of stage-specific proteins through facilitating selective translation of codon-biased transcripts. In parallel, we examined the involvement of tRNA modifications in stress-induced translational control by treating the parasites with IC20-90 doses of hydrogen peroxide, artemisinin and chloroquine. Hierarchical clustering and principal component analysis distinguished signature tRNA modifications reprogramming patterns induced by different stresses, implying stress-specific translational control mechanism for survival response. These findings together provide new insights into the translational control mechanisms in the development, pathogenesis, and stress response of Plasmodium parasites. ​Doctor of Philosophy (SBS)

Technology has developed to a point where everyday personal object and ubiquitous “things” can gain increased value when connected to the Internet, and Singapore has begun exploiting such advancement to build a vision of an evolved nation – a Smart Nation. The narratives of this initiative suggest a country connected from home to port, accessible from personal devices such as smartwatches and mobile phones, with analysis done on every bit of data generated by each and every citizen just going about their daily task. As a result, higher efficiency in transportation and public services, amongst others, can be expected. Unfortunately, a Smart Nation cannot function without collecting personal data in movement and being for its analysis and therefore exposes its citizens to the danger of latent and pervasive privacy invasion, making privacy protection improbable, if not impossible. No prior research, however, has been conducted to investigate privacy concerns, specifically from the citizen's perspective, with relation to Smart Nation. This study investigates Singapore citizen's perception towards this potential loss of privacy and what the existing factors that influence the successful adoption of Smart Nation are. A hypothesized model is proposed to investigate the relationships between privacy concerns, perceived privacy risk, perceived benefits, privacy awareness, privacy control, community identity, trust in government and perceived government intrusion, using the calculus perspective of information privacy theory as the basic framework. The model is tested via a survey of 519 Singaporean citizens and analyzed using linear and multiple regression. The results show that, although citizens are largely concerned over their loss of privacy fueled by perceived government intrusion and other perceived risks, their trust of the government's ability to execute this vision and safeguard their data is the key factor for their support in building a Smart Nation. Further, while community identity is positively related to perceived benefit, these benefits do not have an impact in addressing privacy concern of citizens in Smart Nation. Similarly, though privacy awareness has a positive and significant impact on privacy control it did not provide an impact to privacy concern. As one of the first investigation into Smart Nation, this study provides an initial glimpse into the factors affecting successful deploying the platform from a citizen’s perspective. The result also suggests that leveraging on trust in government is effective in counterbalancing the privacy concerns to provide necessary support from the citizens towards building up Smart Nation. ​Master of Science (Information Studies)

This thesis investigates the electronic and catalytic properties of CP and PCP palladium and platinum complexes and their application in the asymmetric syntheses of phosphine ligands and the development of chiral phosphine gold anticancer drugs. Doctor of Philosophy (SPMS)

The construction of viable photoelectrochemical (PEC) devices for solar-driven water splitting can be achieved by first identifying an efficient independent photoanode for water oxidation and a photocathode for hydrogen generation. These two photoelectrodes then shall be assembled with a proton exchange membrane within a complete coupled system. For constructing efficient PEC devices, there are two factors to be considered. One is semiconductor absorbers which determine the upper limit of PEC efficiency and the other is the surface catalysts which help to push the PEC efficiency approaching to the ideal state. For the materials used for catalyzing hydrogen evolution reactions (HER), development of noble-metal-free material is one of the most desired efforts during the past decade. Transition metal sulfide electrocatalysts (e.g. MoS2) are highly attractive and have been proven as very efficient catalysts in both electrolysis and photoelectrolysis cells. However, the catalytic activity of MoS2 could be further enhanced by introducing additional metal ions inside the structure. Unfortunately, these ternary sulfides are not well understood. Therefore, in this thesis, two categories of ternary sulfides were explored. We first investigated the photocatalytic activity of Cu2MoS4 on Cu2O and Si surface. The electrochemical study confirmed the catalytic activity of Cu2MoS4 on both semiconductors. Driven by drop-casted Cu2MoS4 particles on the surface, Cu2O and Si photocathodes could drive HER at the potential of only 0.45V and 0.25V vs. RHE. The result is very competitive compared with the reports in literature. However, by electrochemical impedance measurement, photoactivity of Cu2O and Si was also found to be limited by this drop-casted Cu2MoS4 film. We then explored electrodeposition of novel Co/NiMoSx for hydrogen evolution. For the first time in the scientific society, we were able to synthesize this material precisely by electrodeposition method. The intrinsic electrocatalytic activity of Co/NiMoSx for catalyzing HER was then carefully examined. It was found that CoMoSx required only ~100mV overpotential to start up hydrogen production compared with ~200mV of MoSx prepared by the same approach. In the effort of investigation of its atomic-structure, we successfully evidenced coordinated bonds between different species. Moreover, mechanism study of this material showed the catalytic activity is limited by 1 proton, 1 electron process, which seems to be a general case on molybdenum or tungsten based sulfides. Co/NiMoSx catalyst was also investigated in a photoelectrochemical cell by coupled with p-Si electrodes for solar hydrogen production. By novel photodeposition method, we successfully deposit Co/NiMoSx film on p-type silicon electrodes easily. The obtained Si/CoMoSx turned to perform excellently with the Jsc more than 17mA.cm-2. The result is very competitive among the best-performed photocathode cells reported in the literature. We then inspected its catalytic activity by impedance analysis. Compared with MoSx and NiMoSx, CoMoSx layer on illuminated Si surface acts not just as a very effective catalyst accelerating the H2 generation, but also as a very effective pacifying agent which accelerates photoelectrons’ injection and transfer. Doctor of Philosophy (MSE)

Metasurfaces are promising two-dimensional metamaterials that are engineered to provide unique properties or functionalities absent in naturally occurring homogeneous surfaces. Here, we report a type of metasurface for tailored reconstruction of surface plasmon waves from light. The design is generic in a way that one can selectively generate different surface plasmon waves through simple variation of the wavelength or the polarization state of incident light. The ultra-thin metasurface demonstrated in this paper provides a versatile interface between the conventional free-space optics and a two-dimensional platform such as surface plasmonics., 7 figures, supplementary information at the end of the document