The perception is that inborn errors of metabolism (IEM) are rare, but the reality is that more than 600 lEMs are now recognized. The organic aciduria, 3-methylcrotonyl-CoA carboxylase (MCC) deficiency arises when 3-methylcrotonyl-Coenzyme A (CoA) carboxylase that participates in the fourth step of the leucine catabolism is defective. Tandem mass spectrometry (MS/MS) based screening programmes in North America, Europe and Australia, showed that MCC deficiency is the most frequent organic aciduria detected, with an average frequency of 1:50 000. Therefore MCC deficiency is considered an emerging disease in these regions. The incidence of MCC deficiency in the Republic of South Africa (RSA) is not yet known. However, one 48 year old male Caucasian individual (HGS) was diagnosed suffering from mild MCC deficiency, since elevated levels of 3-hydroxyisovaleric acid, 3- hydroxyisovalerylcarnitine, 3-methylcrotonylglycine was present in his urine. Several groups are currently working on various aspects of this emerging disease with the focus on the molecular characterisation of MCC deficiency. In the RSA no molecular based diagnostic method which complements MS/MS screening programmes have yet been implemented. Therefore, the aim of this study was to implement the necessary techniques for the molecular characterisation of MCC deficiency, the determination of the sequence of the open reading frame (ORF) of mccA and mccB subunits to determine which mutation(s) are present in the South African MCC deficient patient. For the implementation of the molecular characterisation, a two-pronged approached was used to characterize MCC of a MCC non-deficient individual (CFC). This approach included the reverse transcriptase polymerase chain reaction (RT-PCR) amplification of the ORFs of the associated genes [mccA (19 exons) and mccB (17 exons] and the PCR amplification of selected (genomic deoxyribonucleic acid (gDNA) regions (exons mccA8, mccA11 , mccB5, mccB6 and mccB5-intron 5-6 exon 6 (mccB5-6) which have been found to have mutations associated with MCC deficiency in Caucasians. The sequence analyses produced surprising results of the amplified ORFs (CFCmccA and CFCmccB) of the MCC non-deficient individual CFC. A non-synonymous single nucleotide polymorphism (SNP) (1391C→A, H464P) associated with MCC deficiency (Gallardo et al., 2001) was identified in the CFCmccA subunit. Another SNP (1368G→A, A456A) recently listed in GenBank was observed in the amplified CFCmccB ORF. No significant novel variations or described mutations were identified in the amplified genomic regions mccA8, mccA11 ,mccB5, mccB6 and mccB5-6. The implemented molecular approach was used to characterise MCC of our MCC deficient patient (HGS). The patient did not have any mutation in the four selected exons mccA8, mccA11, mccB5, mccB6 or the genomic region mccB5-6. The RT-PCR amplification of both ORFs (HGSmccA and HGSmccB) resulted in multiple amplicons. Gel extracted amplicons of the expected size were sequenced. Of the 36 exons, 34 exons were sequenced. This includes all 19 exons of HGSmccA and 15 of 17 exons of HGSmccB (exons 1-6 and exons 9-17). The non-synonymous SNP (1391C→A, H464P) detected in CFCmccA (MCC non-deficient individual), seems to be present in the HGSmccA subunit of the MCC deficient individual, HGS. The HGSmccB amplicons could not be entirely sequenced. However, the region exon 1-6 and 9-17 was sequenced but no described or novel mutations were identified. The lack of sequence data of region exon 7-8 led to an incomplete molecular characterisation of the MCC deficiency in HGS. In conclusion, the basic methods and techniques for the molecular characterisation of MCC deficient patients have been implemented locally. A few additional sequencing primers need to be designed to cover mccB7 and mccB8 as well as the entire coding and non-coding strands of each MCC gene (mccA and mccB). The primers for RT-PCR of both mccA and mccB need to be further refined to ensure better specificity. / Thesis (M.Sc. (Biochemistry))--North-West University, Potchefstroom Campus, 2007.

A survey of relevant literature on the topic of wave propagation and scattering in pipes is given first. This review is followed by a theoretical framework which is pertinent to wave propagation in homogeneous, isotropic, pipes. Emphasis is placed on approximate solutions stemming from a computer based, Semi-Analytical Finite Element (SAFE) formulation. A modal analysis of the dynamic response of homogeneous, isotropic pipes, when subjected to a transient ultrasonic excitation, demonstrates that dominant features, i.e., singularities in an unblemished pipe’s displacement Frequency Response Function (FRF) coincide with its cutoff frequencies. This behaviour is confirmed experimentally. A novel technique is developed to deduce such a pipe’s wall thickness and elastic properties from three cutoff frequencies. The resulting procedure is simulated numerically and verified experimentally. Agreement between the new ultrasonic procedure and traditional destructive tests is within experimental uncertainty. Then a hybrid-SAFE technique is used to simulate waves scattered by various open rectangular notches. The simulations show, for the first time, that singularities distinct from the unblemished pipe’s cutoff frequencies arise in a displacement FRF when an axisymmetric notch is introduced. They also suggest that the new singularities depend on the properties of the parent pipe and the finite element region but effects are local to a notch. It is demonstrated further that the difference between the frequency at which a singularity introduced by a notch occurs and the nearest corresponding unblemished pipe’s cutoff frequency is a function of the notch’s dimensions. By plotting contours of constant frequency differences, it is shown that it is usually possible to characterize the notch’s dimensions by using two modes. However, the frequency difference for a third mode may be also needed occasionally. The more general case of nonaxisymmetric notches is shown to be a straightforward extension of the axisymmetric case.

This thesis explores memory performance for smartphone workloads. We design a Video Conference Workload (VCW) to model typical smartphone usage. We describe a trace-based methodology which uses a software implementation to mimic the behaviour of specialised hardware accelerators. Our methodology stores dataflow information from the original application to maintain the relationships between requests. We first study seven address mapping schemes with our VCW, using a first-ready, first-come-first-served (FR-FCFS) memory scheduler. Our results show the best performing scheme is up to 82% faster than the worst. The VCW is memory intensive, with up to 86.8% bandwidth utilisation using the best performing scheme. We also test a Web Browsing and a set of computer vision workloads. Most are not memory intensive, with utilisation under 15%. Finally, we compare four schedulers and find that the FR-FCFS scheduler using the Write Drain mode [8] performed the best, outperforming the worst scheduler by 6.3%.

Magnetic nanoparticles (MNPs) hold great promise in the fields of biology and medicine. The synthesis of functional MNPs with precisely controlled crystallographic, physicochemical, and magnetic properties on a large scale still remains the challenge today. This thesis reports the exploration of liquid-fed flame spray pyrolysis (FSP) in the synthesis of functional MNPs, their surface modifications, and potential bio-applications. Superparamagnetic and ferromagnetic maghemite (γ-Fe2O3) nanoparticles, and silica-coated maghemite (SiO2/γ-Fe2O3) nanocomposites were synthesised using FSP. The size of γ-Fe2O3 was controllable from 6 to 53 nm, with morphology evolving from a disordered near-spherical shape to fully ordered 2-D hexagonal/octagonal platelet. The saturation magnetisation (Ms) increased from 21 to 74 emu/g with increasing particle size, up to 13 nm when Ms approached the bulk γ-Fe2O3 characteristics. In the case of SiO2/γ-Fe2O3, three distinct morphologies, namely the single segregated γ-Fe2O3 core- SiO2 shell, transitional mixed morphologies, and multi γ-Fe2O3 cores embedded in submicron SiO2 shell, were obtained. The core size, composite size, and morphology of γ- Fe2O3 were tunable by varying %SiO2 loading and the use of a quartz tube enclosure during flame synthesis. The magnetic behaviour correlated well with the crystal microstructure. Following the core particle design, protein adsorption-desorption behaviour on FSP-madeMNPs was studied. Bovine serum albumin (BSA) adsorption was found to follow the Langmuir isotherm, with high binding capacities (150−348 mg BSA/g particle) and fast association constants. Electrostatically governed BSA orientations were proposed for different particle-buffer systems. The adsorbed BSA was effectively recovered by pH-shift using K2HPO4. Subsequently, terminal amine, aldehyde, carboxylic, epoxy, mercapto and maleimide functionality were anchored onto the FSP-made γ-Fe2O3 particles. These versatile functional groups led to conjugation of active trypsin. The immobilised trypsin exhibited superior durability with >60% residual activity after one week, and excellent reusability for >5 cycles. The trypsin-conjugated MNPs are promising carriers in proteomics, demonstrating good substrate specificity with equivalent or better sequence coverage compared to free trypsin in insulin and BSA digestion. In another application, a refined silanisation procedure simultaneously reduced γ-Fe2O3 to Fe3O4, and generated thiol enriched surface for matrix metalloproteinase-2 (MMP-2) conjugation. The highly active MMP-2-conjugated MNPs could potentially enhance the interstitial transport of macromolecule/nanoparticles in drug delivery.

Les normes en vigueur de la caractérisation environnementale des déchets s’appuient sur l’analyse de granulats ou réduits en poudre. La modification de propriétés du matériau risque de fausser le relargage d’éléments polluants. L’objectif principal de ce travail de thèse est de mettre au point une méthodologie de lixiviation sur matériau monolithique.Des formulations cimentaires de type mortier ou sous-couche routière incorporant des sédiments traités par le procédé NOVOSOL® (STN) ont été réalisées. Leur caractérisation montre l’augmentation de la porosité et la surface lors de l’ajout de STN. Les caractéristiques mécaniques sont proches de celles des matériaux de type routier.Une longue phase de mise au point a permis de concevoir la cellule d’essai et le protocole permettant de tester des matériaux monolithiques en percolation à débit contrôlé. Un protocole de traçage couplé à une modélisation par un logiciel de transport permet de déterminer les caractéristiques hydrodynamiques des matériaux durant percolation.Les résultats d'essais mettent en évidence la bonne stabilité des métaux lourds contenus dans les STN comparés aux métaux apportés par le liant cimentaire. Ainsi, pour certains métaux, l’introduction de STN tend à diminuer les quantités relarguées, pour d’autres la teneur en polluant des percolats soit stable, soit en augmentation. Même si la complexité des phénomènes mis en jeu nous amène à considérer les données avec précaution, certains résultats tendent à montrer l’existence d’un lien entre surface de contact et relargage. La comparaison des résultats obtenus en laboratoire et des mesures réalisés in situ atteste de la représentativité du test mis au point / The standards of environmental characterisation of wastes recommend the sieving of granular material or the crushing of monolithic material to match a required granulometry. The modification of material characteristics is likely to distort the release of pollutants. The main objective of this thesis is to develop a methodology to test monolithic material by percolation.Various mortar-type and/or under-layer road-type materials incorporating sediment treated by the process NOVOSOL® (STN) designed by the Solvay company have been designed. Their characterization has highlighted the increase of porosity, the decrease of bulk density and the increase of specific area with the addition of STN. The mechanical properties are close to those expected for road type materials.The use of a reference material has allowed designing the test cell and the protocol for testing monolithic materials in percolation with a controlled rate of flow. A breakthrough protocol using an inert salt coupled with a transport modelling software has also been defined, to determine the hydrodynamic characteristics of materials during the percolation.The results of tests showed a good stability of heavy metals contained in the STN compared to metals carried by the cement binder. Thus, for some metals, the introduction of STN tends to reduce the leached quantities, for others, the content of pollutant of leachates either stable or increasing. Despite the complexity of the involved phenomena, some results tend to prove the relation between specific area and heavy metals leaching. Comparison of laboratory results and measurements made in situ demonstrates the representativeness of the test developed

Over the past decade, there has been unprecedented increase in the number of genetic loci associating with type 2 diabetes (T2D) risk and related glycemic traits, thanks to advances in sequencing technologies and access to large sample sizes. Identification of associated genetic variants across the frequency spectrum can provide valuable insight into disease pathophysiology. However, the translation into biological insights has been slow often due to uncertainties over the underlying effector transcripts. G6PC2/ABCB11 is one locus characterised by common non-coding variants that are strongly associated with fasting plasma glucose (FG) levels in healthy adults. The work presented in this thesis aims to understand how protein-coding variants in glycemic trait loci such as G6PC2 contribute to the variability of glycemic traits and in addition gain further insight into the physiological role of G6PC2. To evaluate the role of coding variants in glycemic trait variation, an exome array genotyping study of non-diabetic European individuals (n=33,407) reported multiple coding variants in G6PC2 that were independently associated with FG. I designed and conducted in vitro assays to functionally assess these variants and showed that they result in loss of function (LOF) due to reduced protein stability. This established G6PC2 as the effector transcript influencing FG and highlighted a critical role for G6PC2 (encoding the islet-specific glucose-6-phosphatase catalytic subunit) in glucose homeostasis. To investigate the role of low frequency (MAF=1-5%) and rare (MAF<1%) coding variants in influencing glycemic traits, recent large-scale exome array meta-analyses and whole exome sequencing were carried out as part of MAGIC (n=144,060) and the T2D-GENES/GoT2D consortia (n=12,940) respectively. G6PC1, a gene homolog of G6PC2 that primarily acts through the liver, was uncovered as a novel glycemic locus. My functional follow-up studies demonstrated that rare coding variants in G6PC1 exhibited LOF to influence both FG and FI levels. As rare variation in G6PC2 not previously identified could also affect G6PC2 function and modulate glycemic traits, I also functionally characterised a suite of rare non-synonymous G6PC2 variants. Most of the variants tested exhibited markedly reduced protein levels and/or loss of glycosylation. Several variants were also found to impact on enzymatic activity through inactivating or activating mechanisms to influence FG levels. Finally, to gain better understanding of the function of G6PC2 I performed gene knockdown studies in the EndoC-βH1 human beta cell model followed by insulin secretion analyses. G6PC2 knockdown resulted in increased insulin secretion at sub-threshold glucose stimulation levels, consistent with studies in knockout mouse models. In addition, expression of LOF G6PC2 variants were found to upregulate ER stress responses. These results warrant further studies of the precise roles that G6PC2, an ER-resident protein, plays in regulating insulin secretory function and ER homeostasis in the beta cell. Overall, my work described multiple rare coding variants in both G6PC1 and G6PC2 that alter protein function to regulate glucose metabolism through diverse mechanisms in different tissues. Improved understanding of these effector transcripts will open up opportunities for the exploration of new therapeutic targets for glucose regulation and T2D.

In this work, new reconstruction and analysis methods were developed for 3D field ion microscopy (FIM) data, motivated by the goal of atomic scale characterisation of radiation damage for fusion applications. A comparative FIM/ atom probe tomography (APT) study of radiation damage in self-implanted tungsten revealed FIM advantages in atomistic crystallographic characterisation, able to identify dislocations, large vacancy clusters, and single vacancies. While the latter is beyond the detection capabilities of APT, larger damage features were observed indirectly in APT data via trajectory aberrations and solute segregation. An automated 3DFIM reconstruction approach was developed to maintain reliable, atomistic, 3D insights into the atomic arrangements and vacancies distribution in ion-implanted tungsten. The new method was utilized for the automated ‘atom-by-atom' reconstruction of thousands of tungsten atoms yielding highly accurate reconstructions of atomically resolved poles but also applied to larger microstructural features such as carbides and a grain boundary, extending across larger portions of the sample. Additional tools were developed to overcome reconstruction challenges arising from the presence of crystal defects and the intrinsic distortion of FIM data. Those were employed for the automated 3D mapping of vacancies in ion-implanted tungsten, analysing their distribution in a volume extending across 50nm into the depth of the sample. The new FIM reconstruction also opened the door for more advanced analyses on FIM data. It was applied to the preliminary studies of the distortion of the reconstructed planes, found to depend on crystallographic orientation, with an increased variance in atomic positions measured in a radial direction to the centre of the poles. Additional analyses followed the subtle displacements in atomic coordinates on consecutive FIM images, to find them affected by the evaporation of atoms from the same plane. The displacements were found to increase with size as the distance to the evaporated atom decreased, and are likely to be the result of a convolution between image gas effects, surface atoms relaxation, and charge re-distribution. These measurements show potential to probe the dynamic nature of the FIM experiment and possibly resolve contributions from the different processes effecting the final image. Finally, APT characterisation was performed on bulk and pre-sharpened needles to determine the effect of sample's geometry on the resulting implantation profiles, and the extent to which pre-sharpened needles could be employed in radiation damage studies. While the ions depth profiles in needles were not found within a good match to SRIM simulations, the damage profiles exhibited closer agreement. Further, the concentration of implanted ions in bulk samples was found significantly higher than in the respective needle implanted samples, with excessive loss found for the light ion implantation.

Cette thèse s'inscrit dans le contexte de l'intégration tridimensionnelle des dispositifs électroniques. Parmi les différentes techniques permettant d'assembler à la fois mécaniquement et électriquement les puces empilées, le collage direct de surfaces mixtes Cu-SiO2 représente l'option la plus prometteuse à ce jour. En effet, cette méthode permet d'atteindre la densité d'interconnexions de 106/cm² visée par l'industrie, tout en offrant une faible résistivité de contact et une excellente fiabilité. L'objectif de ce travail est de démontrer la compatibilité du procédé de collage direct hybride Cu-SiO2 avec des intégrations et des architectures proches de circuits réels. Dans ce but, des véhicules de tests intégrant des structures de cuivre à deux et quatre niveaux d'interconnexions ont été conçus spécifiquement. De plus, des simulations par éléments finis du procédé collage direct ont été développées au sein du logiciel Abaqus. Dans un premier temps, le procédé de collage direct puce-à-plaque en 200 et 300 mm est validé. Des caractérisations morphologiques et électriques montrent que cette méthode d'assemblage ne dégrade pas l'intégrité et les performances de structures de tests à deux niveaux par rapport à une intégration plaque-à-plaque. Par ailleurs, des tests de cyclage thermique confirment l'excellente robustesse mécanique des empilements. La deuxième partie de cette thèse s'intéresse à la caractérisation de la morphologie, des performances électriques et de la fiabilité de structures de tests à quatre niveaux d'interconnexions. Dans ce cas, l'architecture plaque-à-plaque en 200 mm des véhicules de tests se veut proche d'une intégration industrielle. Les diverses observations par microscopie électronique à balayage et en transmission indiquent une excellente qualité de collage des interfaces Cu/Cu et SiO2/SiO2. Par ailleurs, les mécanismes de formation des cavités nanométriques à l'interface Cu/Cu et le phénomène de diffusion du cuivre dans la silice sont investigués. Les caractérisations électriques révèlent des rendements de fonctionnement supérieurs à 95 % ainsi que des écarts types inférieurs à 3 % après recuit à 200 ou 400 °C. Enfin, les études de fiabilité incluant des tests de stockage en chaleur humide, de cyclage thermique, de stockage en température et d'électromigration attestent de la résistance à la corrosion et de la robustesse mécanique de cette intégration. Pour finir, les simulations par éléments finis indiquent que les interactions cohésives à l'interface de collage, combinées à la dilatation thermique du cuivre pendant le recuit, assistent significativement le processus de collage de surfaces de cuivre incurvées par sur-polissage. En outre, la déformation plastique macroscopique du cuivre semble avoir un effet néfaste sur le processus de scellement en freinant la propagation de l'onde de collage. / The context of this work is the three-dimensional integration of electronic devices. Among the various techniques allowing to assemble both mechanically and electrically stacked chips, the direct bonding of Cu-SiO2 mixed surfaces is the most promising option to date. Thanks to this method, the interconnection density of 106/cm² aimed by the industry is achievable, while providing a low contact resistivity and excellent reliability. The objective of this study is to demonstrate the compatibility of the direct hybrid bonding Cu-SiO2 process with integrations and architectures that mimic real circuits. For this purpose, test vehicles incorporating two-layer and four-layer copper test structures have been specifically designed. Furthermore, finite element simulations of the direct bonding process have been developed within the Abaqus software. First, the 200 and 300 mm chip-to-wafer direct bonding process is validated. Morphological and electrical characterizations show that this stacking method does not deteriorate the integrity and performances of two-layer test structures with respect to a wafer-to-wafer integration. Furthermore, thermal cycling tests confirm the excellent mechanical strength of the bonded dies. The second part of this work focuses on morphological, electrical and reliability characterizations of four-layer test structures. In this case, the 200 mm wafer-to-wafer architecture of the test vehicles is close to an industrial integration. The various observations conducted with scanning and transmission electron microscopy indicate an excellent bonding quality of Cu/Cu and SiO2/SiO2 interfaces. Furthermore, the formation mechanisms of cavities at the Cu/Cu interface and the copper diffusion phenomenon in the silica are investigated. Electrical characterizations show functional yields above 95 % and standard deviations below 3 % after annealing at 200 or 400 °C. Finally, reliability studies including unbiased HAST, thermal cycling, temperature storage and électromigration test prove the resistance to corrosion and the mechanical robustness of this integration. Finally, the finite element simulations indicate that the cohesive interactions at the bonding interface, combined with the thermal expansion of the copper during the annealing, significantly assist the bonding process of copper surfaces with a dishing effect. In addition, the macroscopic plastic deformation of the copper appears to have a detrimental effect on the sealing of the interface by slowing the propagation of the bonding wave.

Robotics and autonomous systems (RAS) have great potential to propel the world to future growth. Electroadhesion is a promising and potentially revolutionising material handling technology for manufacturing automation applications. There is, however, a lack of an in-depth understanding of this electrostatic adhesion phenomenon based on a confident electroadhesive pad design, manufacture, and testing platform and procedure. This Ph.D. research endeavours to obtain a more comprehensive understanding of electroadhesion based on an extensive literature review, theoretical modelling, electrostatic simulation, and experimental validation based on a repeatable pad design, manufacture, and testing platform and procedure.

The objectives of this project were to synthesise semi-conducting ceramic nanoparticles including zinc oxide (ZnO) and aluminium doped zinc oxide (AZO) through a wet chemistry route to obtain nanoparticles with a controlled size and morphology. Wet chemistry methods (co-precipitation method and hydrothermal method) were used to synthesise ZnO and AZO particles. In the synthesis, various compounds and morphologies were synthesised. ZnO, Zn(OH)2 and unknown phases were co-precipitated, with only ZnO obtained following hydrothermal treatment. Morphologies ranging from platelets, flower-like, nanorods and microflowers were obtained. Particle sizes as small as 11 nm were characterised. Nanorod and nanosphere AZO particles were also synthesised with the results indicated the average grain size decreasing with increasing Al atomic content. Three orthogonal arrays were carried out to investigate the effects of the reaction parameters on the size and morphology of ZnO particles. The applicability of the orthogonal array was successful, with the optimum parameters of both hydrothermal experiments showing an increase in aspect ratio. The L/D ratio of ZnO nanorods obtained in the confirmation experiment increased to 9.4 which was larger than the ZnO synthesised using other reaction conditions (1.0 8.0). Scanning electron microscopy, transmission electron microscopy and X-ray diffraction were used to characterise the properties of the obtained particles. Morphology, crystallinity and particle size were all characterised.