Search

The intracellular Ca2+ handling system in a cardiac myocyte is of crucial importance. It regulates the contraction and relaxation of the myocyte during the excitation-contraction (EC) coupling. A normal intracellular Ca2+ handling system keeps the contraction of the heart orderly, which represents a powerful force to pump blood to the whole body. However, disarrayed or remodelled cellular structure associated with the intracellular Ca2+ handling system at the subcellular level, such as loss of T-tubule network in diseased conditions, may promote abnormal cardiac EC coupling, leading to genesis of cardiac arrhythmias impairing cardiac mechanical functions. Up to date, it is still incompletely understood how the intracellular Ca2+ handling system is altered by changes in subcellular structures of Ca2+ handling systems. In this thesis, biophysically detailed computational models for the intracellular Ca2+ handling system of a cardiac cell were developed, providing a powerful platform to investigate the spatio-temporal complexity associated with the intracellular Ca2+ handling, and its role in generating abnormal cardiac EC coupling. First, a well-validated single cell model was used to investigate how the diastolic and systolic Ca2+ concentration responded to alterations in the model parameters related to the Ca2+ handling system, from which the mechanisms underlying the rate-dependence of EC coupling were analysed. Then, a novel single cell model, with a 2D presentation of the spatial structures of subcellular Ca2+ handling and membrane action potential, of a sheep atrial myocyte was developed for simulating the abnormal intracellular Ca2+ regulation system due to the loss of T-tubules during atrial fibrillation. Variant scenarios of T-tubule loss were considered to investigate the role of the T-tubule in affecting the intracellular Ca2+ regulation. Furthermore, membrane currents' alterations due to the electrical remodelling arising from atrial fibrillation were considered together with the loss of T-tubule. Three typical types of abnormal Ca2+ cycling phenomenon, namely intracellular Ca2+ alternans, spontaneous Ca2+ sparks and intracellular Ca2+ waves were observed in AF conditions. The relationship between T-tubule loss, AF-remodelling and the genesis of delayed afterdepolarizations (DADs) was also investigated. It was shown that the loss of T-tubule in AF condition played an important role in disturbing the Ca2+ regulation system, which increases the risk for a cell to generate impaired contraction.

The purpose of this research is to study, evaluate, and compare the different electro deposition processes for producing nickel-based composite coatings for application in tribology, including: Ni/BMA luminescent coating, Ni-P/MoS2 and Ni-P/WS2 self-lubricating coatings. In the first part, a new luminescent Ni coating containing an embedded, blue emitting rare-earth mixed metal oxide (BaMgAl11O17:Eu2+) BAM was electrodeposited successfully from an aqueous electrolyte. Two types of surfactants were utilised to investigate the effective co-deposition of these phosphors into the nickel matrix. The surfactants of non ionic PEG (polyethylene glycol) and cationic CTAB (cetyl trimethylammonium bromide) were observed to increase the phosphor contents in the deposit from zero to 4.6% and 11.5%, respectively. A mixture of these two surfactants produced the highest particle embedded coverage (15.6%). Systematic study of the hardness, corrosion, tribological and luminescent properties of the as-deposited coating was performed before using it for wear sensing. In the second and third parts of this research, a self-lubricating Ni-P/MoS2 composite coating hasbeen designed and deposited on a mild steel substrate by electrodeposition. The effect of current density, electrolyte additive and MoS2 concentration on the structure of the coating have been investigated. The Ni-P/MoS2 coating exhibited a dramatic reduction in friction compared to an electrodeposited Ni-P coating, the coefficient of friction against a bearing steel ball in sliding wear being as low as 0.05. The worn surfaces and the wear debris were characterised by surface analysis techniques. The composite coating showed only slight wear and oxidation compared to the severe wear and oxidation observed in the pure Ni-P coating or the uncoated mild steel substrate. It has been speculated that MoS2 in the coating underwent shear stress-induced reorganization, fragmenting into fine crystallites and ultimately forming a smooth lubricating film. In the final part, Ni-P/WS2 composite coatings have been electrodeposited from an aqueous bath containing suspension of WS2 nanoparticles (80 nm) and their tribological properties have been evaluated by a reciprocating test with bearing steel balls. Increasing the bath temperature did completely eliminate the cracks in coatings by reducing the internal stress and prolonged the duration of the low friction coefficient. By optimizing the combinations of the WS2 and CTAB concentrations in solution, the amount of the WS2 particles incorporated was greatly boosted, demonstrating a highest WS2 content of 4.8 wt%. Significantly, this Ni P/WS2 coating exhibited a greatly lower friction coefficient of 0.17 compared to single Ni-P (0.6).

Strongly correlated electrons and spin-orbit interaction have been the two major research directions of condensed matter physics in recent years. The discovery of high temperature superconductors in 1986 not only brought excitement into the field but also challenged our theory on quantum materials. After almost three decades of extensive study, the underlying mechanism of high temperature superconductivity is still not fully understood, the reason for which is mainly a poor understanding of strongly correlated systems. The phase diagram of cuprate superconductors has become more complicated throughout the years as multiple novel electronic phases have been discovered, while few of them are fully understood. Topological insulators are a newly discovered family of materials bearing topological non-trivial quantum states as a result of spin-orbit coupling. The theoretically predicted topological Kondo insulators as strongly correlated systems with strong spin-orbital coupling make an ideal playground to test our theory of quantum materials. Scanning tunneling microscopy (STM) is a powerful technique to explore new phenomena in materials with exotic electronic states due to its high spacial resolution and high sensitivity to low energy electronic structures. Moreover, as a surface-sensitive technique, STM is an ideal tool to investigate the electronic properties of topological and non-topological surface states. In this thesis, I will describe experiments we performed on high temperature superconductors and topological Kondo insulators using STM. First, I will describe our experiments on a Bi-based high temperature superconductor $\mathrm{Bi_2Sr_2CuO_{6+\delta}}$. The quasiparticle interference technique uncovers a Fermi surface reconstruction. We also discovered the coexistence of Bogoliubov quasiparticle and pseudogap state at the antinodes. Afterwards, I will discuss our discovery of $d$-form factor density wave in the same material, showing the omnipresence of $d$ form factor density wave above and below the Fermi surface reconstruction. The relation between the $d$-form factor density wave and the pseudogap state is discussed. Second, I will describe our experiments on topological Kondo insulator $\mathrm{SmB_6}$ where spin-orbit coupling plays an important role in the strongly correlated electron system. I will present the spectroscopic evidence of Kondo hybridization based on a spectral decomposition technique. I will introduce a dimension reduction method in the fitting procedure to reduce the computation time by two order of magnitude. I will also discuss the possible quasiparticle interference patterns we discovered in $\mathrm{SmB_6}$., Physics

The aim of this study was to utilize a combination of quantum chemical modeling calculation and state of the art experimental techniques to develop inexpensive non-noble metal catalysts for contemporary reactions. Specific focus was placed on understanding how to systematically tune the surface chemical reactivity as a function of the compositional constitution within the solids, e.g. transition metal + post-transition metal and transition metal + p-block element materials. The efforts upon developing fundamental understandings on surface chemistry drive the discovery of catalysts that exhibit appropriate reactivity to enable selectively activating C–C, C–O, and C–H bonds within linear and aromatic unsaturated hydrocarbons.The work presented here provides two unique platforms to show how to rationally design in-expensive non-noble metal catalysts. In the first example, a computational surface science study was performed upon understanding several fundamental bonding and surface chemistry factors of TM ceramics and their connections to reaction mechanisms of biomass transformation towards benzene production. Our studies show that the general surface reactivity towards carbon, oxygen, and hydrogen can be systematically manipulated as a function of composition change within TM ceramics and can dictate catalytic activity and selectivity. For selective biomass deoxygenation catalysts, an enhanced surface reactivity towards oxygen and a balanced reactivity towards carbon and hydrogen are needed. Our calculations demonstrated that several 1st row transition metal (TM) nitrides and phosphides that display relatively appropriate surface reactivity exhibited more reasonable energetics in the deoxygenation of biomass-model compound, guaiacol.In the second example, we utilized a combination of DFT calculations and surface composition controlled synthesis approach to design highly selective (up 95%) supported Ni+Ga intermetallic compounds for propane dehydrogenation. Our studies demonstrated that the surface composition of supported Ni+Ga IMCs can be manipulated as a function of support selection, actual loading of Ni and Ga, chemical potential of reduction gas, and annealing condition. These aspects allowed us to investigate how the surface composition change affects catalytic activity and selectivity experimentally as well as to model the surface in the theoretical study to understand the effect of surface composition change on local electronic structure and on the general surface chemistry towards carbon that can affect the energetics of reaction mechanisms.

Modern deep learning has enabled amazing developments of computer vision in recent years (Hinton and Salakhutdinov, 2006; Krizhevsky et al., 2012). As a fundamental task, semantic segmentation aims to predict class labels for each pixel of images, which empowers machines perception of the visual world. In spite of recent successes of fully convolutional networks (Long etal., 2015), several challenges remain to be addressed. In this thesis, we focus on this topic, under different kinds of input formats and various types of scenes. Specifically, our study contains two aspects: (1) Data-driven neural modules for improved performance. (2) Leverage of datasets w.r.t.training systems with higher performances and better data privacy guarantees. In the first part of this thesis, we improve semantic segmentation by designing new modules which are compatible with existing architectures. First, we develop a spatio-temporal data-driven pooling, which brings additional information of data (i.e. superpixels) into neural networks, benefiting the training of neural networks as well as the inference on novel data. We investigate our approach in RGB-D videos for segmenting indoor scenes, where depth provides complementary cues to colors and our model performs particularly well. Second, we design learnable dilated convolutions, which are the extension of standard dilated convolutions, whose dilation factors (Yu and Koltun, 2016) need to be carefully determined by hand to obtain decent performance. We present a method to learn dilation factors together with filter weights of convolutions to avoid a complicated search of dilation factors. We explore extensive studies on challenging street scenes, across various baselines with different complexity as well as several datasets at varying image resolutions. In the second part, we investigate how to utilize expensive training data. First, we start from the generative modelling and study the network architectures and the learning pipeline for generating multiple examples. We aim to improve the diversity of generated examples but also to preserve the comparable quality of the examples. Second, we develop a generative model for synthesizing features of a network. With a mixture of real images and synthetic features, we are able to train a segmentation model with better generalization capability. Our approach is evaluated on different scene parsing tasks to demonstrate the effectiveness of the proposed method. Finally, we study membership inference on the semantic segmentation task. We propose the first membership inference attack system against black-box semantic segmentation models, that tries to infer if a data pair is used as training data or not. From our observations, information on training data is indeed leaking. To mitigate the leakage, we leverage our synthetic features to perform prediction obfuscations, reducing the posterior distribution gaps between a training and a testing set. Consequently, our study provides not only an approach for detecting illegal use of data, but also the foundations for a safer use of semantic segmentation models.

Twinning plays critical role in the deformation of hexagonal close-packed (HCP) metals. Developing HCP alloys with improved properties requires a fundamental knowledge on the twinning mechanisms which, however, remain largely theoretical contentions. Here, by using state-of-the-art in situ transmission electron microscopy, the atomic mechanisms of twinning in HCP metals were unprecedentedly revealed. Classically, the most prevalent {1 0 -1 2} twinning proceeds by shuffle-gliding of twinning dislocations (TDs). However, this mechanism has been challenged by extensive discrepancies in experiments. Here, it was found that the twinning nucleation and growth were primarily mediated by direct transformation of matrix prismatic planes into twin basal planes, which established the same lattice correspondence as an ideal {1 0 -1 2} twin. In detwinning, the twin boundaries retreated from incoherent basal│prismatic and prismatic│basal interfaces to coherent twin boundaries; thereafter, the classical TD mechanism dominated in the detwinning. The findings demonstrate that the twinning formation is not necessarily a shear-dominated process on the twinning plane. Moreover, a new mechanisms of HCP→FCC transformations was discovered, which proceeded by {1 -1 0 0}hcp→{1 1 1}fcc transformation. The FCC phase assisted the {1 0 -1 2} twinning of the HCP matrix as it transformed back to HCP structure through {1 1 1}fcc→(0 0 0 1)hcp transformation. It is generally believed that {1 1 -2 1} twinning is mediated by TDs. Here, direct evidences for the TD mechanism and theoretical twinning elements were unambiguously revealed. In stark contrast to the general consensus, atomic shuffles happened along 1/6 on alterative basal planes. Classically, {1 1 -2 2} twinning is a contraction twinning mode. Here, a {1 1 -2 2} tension twinning mode was discovered, wherein basal plane was K2. The finding was further corroborated by the large twinning shear and dominant elementary TDs. This dissertation resolves long-standing controversies in HCP twinning, reveals prolific new transformation mechanisms pertinent to twinning, and provides crucial information for the twinning-based design and processing of HCP alloys with desired mechanical properties.

Carbon nanotubes (CNTs) are molecular-scale tubes of graphitic carbon with outstanding electronic, mechanical and thermal properties. The numerous industrial and commercial applications of CNTs are driving the growth of production of the nanotubes. As such, release of CNTs into the environment inevitably occurs. Therefore, quantification of CNTs in environmental matrices becomes critical to answer research questions regarding their environmental fate/transport and potential risks to environment and human health. However, most current analytical methods have some limitations on the environmental determination of CNTs. This dissertation aims to systematically explore the potential of an innovative microwave induced heating system for quantification of different types of CNTs in various environmental media. First, a microwave induced heating apparatus was established in our lab, connected with a custom-built temperature acquisition and control system. At a given microwave condition, a linear relationship between microwave induced temperature rises and mass of well-characterized CNTs (single-walled CNTs, multi-walled CNTs or carboxylated multi-walled CNTs) was successfully generated in quartz sand, agricultural soil and wastewater treatment plant sludge, respectively. Moreover, our results indicated that the presence of inorganic carbon, organic carbon and other carbon-based nanomaterials (fullerene, activated carbon or graphene oxide) in environmental samples did not affect the microwave measurements of CNTs when the microwave energy was high enough. The detection limits of the microwave method were determined for all three types of CNTs in different environmental matrices.Secondly, the impacts of heterogeneity of multi-walled CNTs (length, diameter and aggregation) on the microwave responses of multi-walled CNTs were investigated. We found that the microwave performances of multi-walled CNTs were independent of the length and diameter of the nanotubes, whereas the aggregated multi-walled CNTs were not able to convert the microwave energy to heat. To apply the microwave method to the aggregated multi-walled CNTs, a two-step pre-treatment process, involving elevated temperature exposure and acetone-based surfactant assisted dispersion, was proposed to suspend the agglomerated multi-walled CNTs that have been embedded in the environment before microwave quantification. With the suggested pre-treatment process, the microwave method could effectively quantify the aggregated multi-walled CNTs in environmental matrices. Thirdly, the simultaneous quantification of individual type of CNTs in environmental samples containing several types of CNTs was studied using Chemometrics based multivariate calibration methods. The environmental samples with two or three types of CNTs were exposed to varied microwave conditions to generate microwave induced temperature rises spectra. The relationship between the temperature rise spectrum and individual mass of CNTs was explored by such statistical tools as partial least square regression, least square-support vector machine and artificial neutral network. The performance and prediction ability of fitted models were evaluated in terms of R2, root mean square error of cross validation and root mean square error of prediction. Among three multivariate calibration approaches, the developed least square-support vector machine model showed best fitness and prediction results.Overall, this dissertation provides demonstration and insights for the development of a viable and low-cost quantification method for CNTs in environmental matrices, which could be beneficial for researchers and scientists in all CNTs-related studies.

To date, malignancies have been categorized as one of the leading life-threatening diseases worldwide, mainly because of lack of imaging devices for detection of cancers at its early stage of development. In this case, the emergence of novel strategies which can be effectively used in early stage cancer detection and therapy continue to become an inevitable need and indispensable scope of and cancer field researchers, for which implementation of new technologies are required to advance the early detection and management of cancers. The biomedical imaging of the organs in the living system offers a viable approach to label and image the cells. Among the different bioimaging techniques such as X-ray, Commuted Tomography, Emission Tomography and Magnetic Resonance Imaging that have been developed, Quantum-Dots based technique has acquired increasing attention because of its unique optical and electrical properties and good biocompatibility. In this report the synthesis of ZnSe@ZnS core-shell Quantum Dots, which is cadmium-free and non-toxic to the biological cells, is proposed. These core-shell Quantum Dots will be applicable to biomedical imaging and diagnostics operating on living biological cells.

Strongly correlated electrons and spin-orbit interaction have been the two major research directions of condensed matter physics in recent years. The discovery of high temperature superconductors in 1986 not only brought excitement into the field but also challenged our theory on quantum materials. After almost three decades of extensive study, the underlying mechanism of high temperature superconductivity is still not fully understood, the reason for which is mainly a poor understanding of strongly correlated systems. The phase diagram of cuprate superconductors has become more complicated throughout the years as multiple novel electronic phases have been discovered, while few of them are fully understood. Topological insulators are a newly discovered family of materials bearing topological non-trivial quantum states as a result of spin-orbit coupling. The theoretically predicted topological Kondo insulators as strongly correlated systems with strong spin-orbital coupling make an ideal playground to test our theory of quantum materials. Scanning tunneling microscopy (STM) is a powerful technique to explore new phenomena in materials with exotic electronic states due to its high spacial resolution and high sensitivity to low energy electronic structures. Moreover, as a surface-sensitive technique, STM is an ideal tool to investigate the electronic properties of topological and non-topological surface states. In this thesis, I will describe experiments we performed on high temperature superconductors and topological Kondo insulators using STM. First, I will describe our experiments on a Bi-based high temperature superconductor $\mathrm{Bi_2Sr_2CuO_{6+\delta}}$. The quasiparticle interference technique uncovers a Fermi surface reconstruction. We also discovered the coexistence of Bogoliubov quasiparticle and pseudogap state at the antinodes. Afterwards, I will discuss our discovery of $d$-form factor density wave in the same material, showing the omnipresence of $d$ form factor density wave above and below the Fermi surface reconstruction. The relation between the $d$-form factor density wave and the pseudogap state is discussed. Second, I will describe our experiments on topological Kondo insulator $\mathrm{SmB_6}$ where spin-orbit coupling plays an important role in the strongly correlated electron system. I will present the spectroscopic evidence of Kondo hybridization based on a spectral decomposition technique. I will introduce a dimension reduction method in the fitting procedure to reduce the computation time by two order of magnitude. I will also discuss the possible quasiparticle interference patterns we discovered in $\mathrm{SmB_6}$.

The most general AdS$_5 imes Y$ solutions of type IIB string theory that are AdS/CFT dual to superconformal field theories in four dimensions can be fruitfully described in the language of generalized geometry, a powerful hybrid of complex and symplectic geometry. We show that the cone over the compact five-manifold $Y$ is generalized Calabi-Yau and carries a generalized holomorphic Killing vector field $xi$, dual to the R-symmetry. Remarkably, this cone always admits a symplectic structure, which descends to a contact structure on $Y$, with $xi$ as Reeb vector field. Moreover, the contact volumes of $Y$, which can be computed by localization, encode essential properties of the dual CFT, such as the central charge and the conformal dimensions of BPS operators corresponding to wrapped D3-branes. We then define a notion of ``generalized Sasakian geometry'', which can be characterized by a simple differential system of three symplectic forms on a four-dimensional transverse space. The correct Reeb vector field for an AdS$_5$ solution within a given family of generalized Sasakian manifolds can be determined---without the need of the explicit metric---by a variational procedure. The relevant functional to minimize is the type IIB supergravity action restricted to the space of generalized Sasakian manifolds, which turns out to be just the contact volume. We conjecture that this contact volume is equal to the inverse of the trial central charge whose maximization determines the R-symmetry of the dual superconformal field theory. The power of this volume minimization is illustrated by the calculation of the contact volumes for a new infinite family of solutions, in perfect agreement with the results of $a$-maximization in the dual mass-deformed generalized conifold theories.