La mémoire est considérée comme étant gloutonne en consommation d'énergie, un problème sensible, particulièrement dans les systèmes embarqués. L'optimisation globale de fonctions multimodales est également un problème délicat à résoudre du fait de la grande quantité d'optima locaux de ces fonctions. Dans ce mémoire, je présente différents nouveaux algorithmes hybrides et distribués afin de résoudre ces deux problèmes d'optimisation. Ces algorithmes sont comparés avec les méthodes classiques utilisées dans la littérature et les résultats obtenus sont encourageants. En effet, ces résultats montrent une réduction de la consommation d'énergie en mémoire d'environ 76% jusqu'à plus de 98% sur nos programmes tests, d'une part. D'autre part, dans le cas de l'optimisation globale de fonctions multimodales, nos algorithmes hybrides convergent plus souvent vers la solution optimale globale. Des versions distribuées et coopératives de ces nouveaux algorithmes hybrides sont également proposées. Elles sont, par ailleurs, plus rapides que leurs versions séquentielles respectives. / Résumé en anglais : Memory is considered to be greedy in energy consumption, a sensitive issue, especially in embedded systems. The global optimization of multimodal functions is also a difficult problem because of the large number of local optima of these functions. In this thesis report, I present various new hybrid and distributed algorithms to solve these two optimization problems. These algorithms are compared with conventional methods used in the literature and the results obtained are encouraging. Indeed, these results show a reduction in memory energy consumption by about 76% to more than 98% on our benchmarks on one hand. On the other hand, in the case of global optimization of multimodal functions, our hybrid algorithms converge more often to the global optimum solution. Distributed and cooperative versions of these new hybrid algorithms are also proposed. They are more faster than their respective sequential versions.

L'implantation d'hydrogène à forte dose est utilisée dans le procédé Smart Cut(tm) afin de transférer des couches de silicium assez épaisses (>200 nm) sur un autre substrat. En utilisant l'implantation à très basse énergie, la co-implantation d'H et d'He pour des doses totales bien plus faibles que celles requises lorsque l'hydrogène est implanté seul ouvre la voie à un transfert de couches beaucoup plus minces (< 50 nm). Cependant, les phénomènes mis en jeu ainsi que les mécanismes responsables de l'interaction, près de la surface libre du wafer, entre l'H et l'He, et les interstitiels et les lacunes qu'ils génèrent, restent à ce jour largement incompris. Dans ce travail, nous avons tout d'abord déterminé l'effet de la réduction des énergies d'implantation d'H et d'He sur la formation et le développement, lors d'un recuit, des cloques qui se forment à partir de micro-fissures en l'absence d'un raidisseur collé à la plaque implantée. Une approche basée sur la comparaison entre les caractéristiques dimensionnelles des cloques obtenues expérimentalement et la simulation par éléments finis, nous a permis de déterminer la pression et la quantité d'He et d'H2 hébergées dans ces cloques. En comparant ces résultats avec les doses d'ions implantées, nous avons pu mettre en évidence l'absence d'exo-diffusion d'He et d'H lors d'un recuit quelle que soit la distance entre la surface et les profils d'ions implantés, qui montre une forte efficacité des cloques à préserver les molécules. Nous avons pu identifier, puis expliquer, la différence en efficacité de coalescence des cloques en fonction de leurs positions en profondeur en la reliant à la variation de l'augmentation d'énergie élastique des cloques par rapport à leur surface. Nous avons ensuite étudié le rôle du dommage ionique, c'est-à-dire des défauts résultants de la co-implantation d'He et d'H, sur la formation et l'évolution thermique de la microstructure du silicium implanté. Cette étude a été menée soit en fonction de l'ordre d'implantation, soit en fonction de la position nominale en profondeur du profil d'He par rapport au profil d'H, soit en fonction du ratio entre les doses d'implantation d'He et d'H. Nous avons montré que la distribution en profondeur de l'H n'est jamais affectée par la co-implantation d'He. L'He est toujours piégé dans la zone où le dommage est maximal. Lorsque le dommage est maximal dans la zone du profil d'H, l'He y diffuse et y est piégé dans des nano-bulles et/ou des microfissures. Mais si le dommage généré dans la zone où est distribué l'He est supérieur à celui généré autour du profil d'H, l'He reste piégé en dehors du profil d'H dans des nano-bulles. L'He contenue dans des nano-bulles, quelle que soit leur distribution en profondeur, ne contribue pas à la pressurisation des cloques ce qui ralenti la coalescence des cloques. Finalement, nous avons pu proposer différents scénarii permettant de rendre compte des similarités et des différences mises en évidence tant avant recuit qu'après recuit, à basse ou plus haute température selon le type d'implantation réalisé. / The high dose hydrogen ion implantation is used in the Smart Cut (tm) process to transfer relatively thick (i.e. >200 nm) Si layers from a donor substrate onto a host material. Hydrogen and helium co-implantation at low energies for a much lower total fluence opens the way for transferring extremely thinner (i.e. <50 nm) layers. However, the phenomena and the mechanisms responsible for the interaction, close to a wafer surface, between H, He, silicon interstitials and vacancies they generate remain poorly understood. First, we studied the effect of reducing the ion energies during both H and He implantations onto the formation and the development of blisters during annealing. Blisters were formed from the micro-cracks since a stiffener was not bonded to the implanted wafer. An approach, based on the comparison between experimentally obtained size characteristics of blisters with the finite element method simulations, allowed us to deduce the pressure inside blister cavities and the fraction of the implanted fluences used to pressurize them. We showed that even when implanted at very low energy, H and He atoms do not exo-diffuse out of the implanted region during annealing. We were able to identify, and then relate the efficiency of blister coalescence to a variation in the elastic energy of blisters as a function of their depth position. In a second part, we studied the role of the damage, produced by He and H coimplantation, on the formation and the thermal evolution of the microstructure of the implanted silicon. These investigations were realized as a function of either the order of co-implantation, or the nominal position of the He profile with respect to the H one, or the ratio between He and H fluences. We showed that the H depth distribution was never affected by He co-implantation. Helium was always trapped at the depth where the damage was maximum. When the damage was highest within the H profile, He diffused and was trapped there in the nano-bubbles and /or the blister cavities. However, when the damage was higher within the He profile than within the H one, He remained trapped in the nano-bubbles outside the H profile. Helium contained in the nano-bubbles, whatever their depth distribution, did not contribute to a pressurization of blister cavities that slowed down their coalescence. Finally, we have proposed various scenarios accounting for the similarities and the differences evidenced both before and after annealing at low or higher temperatures depending on the type of realized implantation.

Este trabalho apresenta o primeiro estudo dos estados aniônicos de baixa energia da para-Benzoquinona (p-BQ) por meio de cálculos de espalhamento elástico de elétrons pela molécula neutra. Foi utilizado o método multicanal de Schwinger implementado com pseudopotenciais de Bachelet, Hamann e Schlüter, na aproximação de núcleos fixos, e nas aproximações estático-troca e estático-troca mais polarização. As seções de choque integrais foram calculadas em duas geometrias do alvo, otimizadas para o estado fundamental da molécula neutra e do ânion. Os estados aniônicos encontrados foram aracterizados e comparados com resultados experimentais, realizados com a técnica de espectroscopia de transferência eletrônica (ETS). Três estados * foram observados, sendo um ligado e dois metaestáveis, em acordo com os experimentos. Uma ressonância de Feshbach foi encontrada nos cálculos, tal qual é observada experimentalmente e sugerida em diversos estudos teóricos de estado ligado. Acredita-se que a aproximação utilizada para descrever estados excitados do alvo poderá ser reconsiderada, em que é sugerida uma proximação mais sofisticada, a ser utilizada em cálculos futuros. Com este estudo, é possível uma melhor compreensão dos resultados experimentais ETS, assim como é permitida uma interpretação de estudos teóricos sobre os estados aniônicos da p-BQ. Ademais, este trabalho possibilita próximos estudos de espalhamento envolvendo quinonas mais complexas. / This work present the first study of low-energy anionic states in para-benzoquinone (p-BQ) by electron scattering calculations. The elastic scattering calculations were done in the fixed-nuclei approximation, with the Schwinger multichannel method implemented with pseudopotentials of Bachelet, Hamann and Schlüter, in the static-exchange and in the static-exchange plus polarization approximations. The integral cross sections were performed in two target geometries optimized for the neutral and anion ground states. The anionic states found were characterized and compared with experimental results obtained by electron transfer spectroscopy (ETS). Three * states were observed, a bound state and two metastable, in good agreement with experimental results. A Feshbach resonance was found in calculations, also it is observed experimentally and it was suggested in several theoretical studies of bound state. It is believed that the approach used to describe the target excited states can be reconsidered, as it is suggested a more sophisticated approach to be used in future calculations. With this study, a better understanding of experimental results ETS is possible, as it allowed an interpretation of theoretical studies of the anionic states of p-BQ. Moreover, this work enables further scattering studies involving more complex quinones.

Une approche thérapeutique innovante utilisant l'adjonction d'éléments de numéro atomique élevé à une radiothérapie de basse énergie semble offrir une voie prometteuse pour le traitement des tumeurs cérébrales résistantes. Une telle technique est notamment développée sur la ligne médicale de l'ESRF (European Synchrotron Radiation Facility) utilisant un rayonnement monochromatique allant de 25 à 90 keV [Adam 2003, Adam 2006]. Des résultats encourageants ont été obtenus en traitant des souris cancéreuses après injection de nanoparticules d'or (AuNP) [Hainfeld 2004]. Cependant, les processus physiques et l'impact biologique issus de la photoactivation de nanoparticules sont encore aujourd'hui mal compris et ne peuvent être expliqués par des calculs de doses macroscopiques [Cho 2005, Zhang 2009]. Le but de ce travail est d'évaluer par simulation Monte Carlo l'augmentation locale de dose en présence de nanoparticules ainsi que les caractéristiques des électrons secondaires produits. Dans un premier temps, des simulations ont été réalisées en utilisant une géométrie cellulaire, de manière à comparer les données simulées aux expérimentations menées à l'ESRF. Des tests de clonogénicité ont été réalisés pour mesurer le taux de radiosensibilité des cellules pour une irradiation de 4 Gy (SER4Gy) en présence de gadolinium, pour différentes énergies d'irradiation (25 keV à 1250 keV). Ces études, expérimentales et numériques, ont permis l'évaluation de l'influence de la localisation du gadolinium au sein de la cellule et la forme de ce dernier (nanoparticules ou agent de contraste). D'autre part, une étude comparative a été menée pour caractériser le comportement d'une nanoparticule sous irradiation à une échelle nanométrique, en fonction de l'énergie de faisceau, du rayon de la nanoparticule et de l'élément lourd (or et gadolinium). / An innovative approach using X-ray interactions with heavy elements seems to open a promising way of treatment for resistant cancers, such as high-grade gliomas. Such a technique is developed at the medical beam line of ESRF using monochromatic X-rays in the 25-90 keV range for the treatment of brain tumors [Adam 2003, Adam 2006]. The use of gold nanoparticles (AuNP) to treat mice bearing subcutaneous tumors led to encouraging results [Hainfeld 2004]. However, the physical processes and biological impact of the photon activation of nanoparticles are not yet well understood. The experimental results cannot be explained by macroscopic dose calculations [Cho 2005, Zhang 2009]. The aim of this work was to evaluate, at the sub-cellular level, the dose enhancement in presence of nanoparticles and the properties of the secondary electrons production using Monte Carlo simulations. In a first step, simulations were performed using cell geometry, in order to compare the simulated data to the experiments realized on the ID17 beamline of ESRF. Clonogenic assays have been performed on F98 cells to measure the “Sensitizer Enhancement Ratio” for an irradiation of 4 Gy (SER4Gy) in the presence of gadolinium, for several beam energies (25 to 80 keV). These experimental and numerical studies were done to evaluate the influence of the gadolinium location within the cell and its shape (nanoparticles or contrast agent). On the other hand, a comparative study has been performed to evaluate the behavior of a nanoparticle under irradiation at a nanometer scale. Electron spectra have been studied for two heavy elements - gold and gadolinium - and several beam energies from 25 keV to 2 MeV. Experiments have shown that gadolinium nanoparticles (GdNP) incubated during 5 h with the cells were strongly effective compared to non-incubated nanoparticles and contrast agent, for the same concentration of gadolinium. A part of radiosensitivity could possibly be explained by a biological action of GdNP on the cell cycle. Another part could be attributed to the important dose enhancement factor (DEF) calculated in the vicinity of GdNP, highlighted from two-dimension DEF maps. The DEF can reach two orders of magnitude within a few nanometers of the GdNP surface and is mainly due to high-linear energy transfer electrons (< 5 keV). By modeling the case of nanoparticles randomly distributed on the cell membrane (closest to the experimental case), we showed that a good correlation exists between the SER4Gy and the membrane DEF. On the other hand, the comparison of the two elements showed that GdNP could produce more electrons (of lower energy) than AuNP (with same mass), but that the local DEF due to AuNP was more important. Interesting results were obtained by comparing the local DEF with experimental results on plasmid DNA. However, it seems important to carry on these studies by taking into account the post-irradiation chemical processes in modeling.

Os procedimentos braquiterápicos atuais seguem as diretrizes apresentadas no protocolo da AAPM TG - 43, introduzido em 1995, que define a metodologia de cálculo de dose ao redor de fontes encapsuladas. Este protocolo, dentre outras considerações, adota a água como meio dosimétrico padrão e desconsidera a composição dos tecidos, densidades e dimensões do paciente nas estimativas de dose absorvida. Com o objetivo de realizar os cálculos de dose em condições mais próximas à realidade, a AAPM publicou em 2012 o TG-186, que introduz os algoritmos de cálculos de dose baseados em modelos na braquiterapia (MBDCA- model-based dose calculation algorithm). Estes algoritmos são capazes de considerar as complexidades descritas acima, além de permitirem estimativas de dose diretamente nos tecidos biológicos. Apesar deste advento, ainda há controvérsia sobre a melhor forma de se reportar a dose absorvida, com pontos favoráveis tanto para estimativas de dose na água, quanto nos tecidos. Esta tese se insere dentro deste contexto, buscando a correlação entre todo o conhecimento previamente adquirido baseado em água com os modernos algoritmos de cálculo de dose. A relação entre a dose na água e no meio é realizada através da teoria da cavidade, que assume que a fluência dos fótons na água e no meio é idêntica. Parte dos objetivos do presente trabalho foi avaliar a fluência energética de fótons em diferentes meios irradiados com fontes braquiterápicas de energias baixas (<50,0 keV), propondo uma forma eficiente de correlacionar a dose na água e no meio em situações nas quais a fluência dos fótons é relevante. Avaliou-se a dose absorvida na água e em diferentes tecidos humanos, para quantificar fatores de conversão entre as estimativas de dose, simulados por Monte Carlo. Para validar tais fatores foi proposta uma metodologia experimental com o uso de um objeto simulador, desenvolvido especificamente para a realização deste trabalho e com capacidade de medir os efeitos da heterogeneidade do meio utilizando doses absorvidas em dosímetros termoluminescentes. As correções baseadas na fluência energética obtidas neste estudo, quando necessárias, são capazes de correlacionar a dose absorvida no tecido e na água com uma precisão melhor do que 0,5 % nos casos mais críticos (ex. osso). Os fatores de conversão calculados mostraram que a dose absorvida na água subestima a dose absorvida no osso em até 80 %, mas superestima a dose no tecido adiposo em aproximadamente 75 %, ressaltando a necessidade de se considerar a composição e a densidade do meio nas estimativas de dose. Os resultados experimentais permitiram validar os fatores de conversão de dose simulados com diferenças máximas de 8,5 %, entre os valores experimentais e simulados. Todos os resultados obtidos comprovaram que a estimativa da dose absorvida em procedimentos braquiterápicos com baixas energias diferem significativamente quando realizadas na água e nos tecidos biológicos, evidenciando a necessidade do uso de algoritmos que considerem a heterogeneidade do meio. Tais resultados também enfatizaram a necessidade de se considerar com precisão a composição do corpo, uma vez que variações nas composições médias dos tecidos podem afetar as estimativas dosimétricas e aumentar as incertezas dos resultados. / Brachytherapy treatments are commonly performed using the American Association of Physicists in Medicine (AAPM) Task Group report TG-43, introduced in 1995, which defines the formalism for the calculation of absorbed dose to water, and neglects human tissue densities, material compositions, body interfaces, body shape and dose perturbations from applicators. In order to perform dose calculations in conditions closer to reality, the AAPM published in 2012 the TG-186, which introduces the model-based dose calculation algorithm (MBDCA) in brachytherapy. These algorithms are able to consider the complexities described above, in addition, it allows dose estimates directly into biological tissues. Despite the advent, the best way to report the absorbed dose is still a matter of debate, with favourable points for both water and tissue absorbed dose estimates. The present thesis is inserted within this context, searching for the correlation between all previously acquired knowledge based in absorbed dose to water with the absorbed dose calculated using modern MBDCA. The correlation between the doses in water and the doses in tissue is performed through the cavity theory, which generally assumes that the fluence of the photons in water and in tissue are identical. Part of the purpose of the present work was to evaluate the energy-fluence of photons in different medium irradiated with low energy brachytherapy sources (<50,0 keV), proposing an efficient way to correlate absorbed dose to water and absorbed dose to tissue in brachytherapy in situations in which the fluence of the photons is relevant. In addition to this objective, the dose absorbed in water and in different human tissues was evaluated to quantify conversion factors between these dose estimates, simulated by Monte Carlo. To validate such factors, an experimental methodology was proposed in a phantom with the capacity to quantify the effects of the heterogeneity of the medium measuring absorbed doses in thermoluminescent dosimeters. The energy-fluence based corrections given in this work, when necessary, are able to correlate the absorbed dose to tissue and absorbed dose to water with an accuracy better than 0.5 % in the most critical cases (e.g.: bone tissue). The calculated conversion factors showed that the absorbed dose to water underestimates the absorbed dose to bone by up to 80 % but overestimates the dose in adipose tissue by approximately 75 %, emphasizing the need to consider the composition and the density of tissue in the dose estimates. The experimental results allowed validating the simulated conversion factors with maximum differences of 8.5 % between the experimental and simulated dose values. All the results obtained showed that the estimation of the dose absorbed in low energy brachytherapy procedures differ significantly when performed in water and in biological tissues, evidencing the necessity of using MBDCA. These results also emphasized the need to accurately consider tissue composition, since the smallest variations in tissue compositions may affect dosimetric estimates and increase uncertainties of the results.

Wireless sensor networks are becoming important in several monitoring and sensing applications. Ultra low power consumption in the sensor nodes is important for extending the battery life of the nodes. In this dissertation, two low power BFSK receiver architectures are proposed and verified with prototype implementations in silicion. A 2.4 GHz 1 Mb/s polyphase filter (PPF) BFSK receiver demonstrates ±180 ppm frequency offset tolerance (FOT) and 40 dB adjacent channel rejection (ACR) at a modulation index (MI) of 2, with a power consumption of 1.9 mW. High FOT at low MI is achieved by a frequency-to-energy conversion architecture using PPFs without any frequency correction. The proposed hybrid topology of the PPF provides an improved ACR at reduced power. To further improve the energy efficiency, a low energy 900 MHz mixer-less BFSK receiver is designed. High gain frequency-to-amplitude conversion and better sensitivity is achieved by a linear amplifier with Q-enhanced LC tank, eliminating the need for local oscillators and mixers. With a power consumption of 500 μW, the receiver achieves sensitivities of -90 dBm and -76 dBm for data rates of 0.5 Mb/s and 6 Mb/s, respectively. The energy efficiency is 80 pJ/b when operating at 6 Mb/s. / Graduation date: 2013 / Access restricted to the OSU Community at author's request from March 25, 2013 - March 25, 2014

The primary objective of this research was to investigate and develop an electrostatic energy-harvesting voltage-constrained CMOS/BiCMOS integrated circuit (IC) that harnesses ambient kinetic energy from vibrations with a vibration-sensitive variable capacitor and channels the extracted energy to charge an energy-storage device (e.g., battery). The proposed harvester charges and holds the voltage across the vibration-sensitive variable capacitor so that vibrations can induce it to generate current into the battery when capacitance decreases (as its plates separate). To that end, the research developed an energy-harvesting system that synchronizes to variable capacitor's state as it cycles between maximum and minimum capacitance by controlling each functional phase of the harvester and adjusting to different voltages of the on-board battery. One of the major challenges of the system was performing all of these duties without dissipating the energy harnessed and gained from the environment. Consequently, the system reduces losses by time-managing and biasing its circuits to operate only when needed and with just enough energy while charging the capacitor through an efficient inductor-based precharger. As result, the proposed energy harvester stores a net energy gain in the battery during each vibration cycle.

Hydrogen is an efficient, carbon-free and safe energy carrier. However, its compact and weight-efficient storage is an ongoing subject for research and development. Among the intensively investigated hydrogen storage materials, lithium aluminum hydride (LiAlH4) is an attractive candidate because of its high theoretical hydrogen density (volumetric: 96.7g H2/l material; gravimetric: 10.6 wt.%-H2) in combination with rather low decomposition temperatures (onset temperature <100°C after doping). Although the reversible dehydrogenation of LiAlH4 must be carried out with the help of organic solvent, LiAlH4 can serve as single-use hydrogen storage material for various special applications, for example, hydrogen fuel cell systems. This thesis deals with transition metal (TM)-doped LiAlH4 aiming at tailored dehydrogenation properties. The crystal structure and morphology of TM-doped LiAlH4 is characterized by XRD and SEM respectively. The positive effects of four dopants (NiCl2, TiCl3, ZrCl4 and TiCl4) on promoting the dehydrogenation kinetics of LiAlH4 are systematically studied by thermal analysis. Based on the state of each TM chloride (solid or liquid), three low-energy-input doping methods (1. ball-milling at low rotation speed; 2. manual grinding or magnetic stirring; 3. magnetic stirring in ethyl ether) are compared in order to prepare LiAlH4 with the maximum amount of hydrogen release in combination with fast dehydrogenation kinetics. The dehydrogenation properties of the TM-doped LiAlH4 powders are measured under isothermal conditions at 80°C at a H2 pressure of 1 bar, which is within the operating temperature range of proton exchange membrane (PEM) fuel cells, aiming at applications where the exhaust heat of the fuel cell is used to trigger the dehydrogenation of the hydrogen storage material. Furthermore, the mid-term dehydrogenation behavior of TM-doped LiAlH4 was monitored up to a few months in order to test its mid-term storability. In addition, the pelletization of TM-doped LiAlH4 is investigated aiming at a higher volumetric hydrogen storage capacity. The effects of compaction pressure, temperature and the H2 back-pressure on the dehydrogenation properties of TM-doped LiAlH4 pellets are systematically studied. Moreover, the volume change through dehydrogenation and the short-term storage of the TM-doped LiAlH4 pellets are discussed in view of practical applications for PEM fuel cell systems. / Wasserstoff ist ein effizienter, kohlenstofffreier und sicherer Energieträger. Jedoch die kompakte und gewichtseffiziente Speicherung ist ein permanentes Forschungs- und Entwicklungsthema. Unter den intensiv untersuchten Materialien für die Wasserstoffspeicherung ist aufgrund der hohen theoretischen Speicherdichte (volumetrisch: 96,7 g H2/L, gravimetrisch: 10.6 Gew.%-H2) in Kombination mit sehr niedrigen Zersetzungstemperaturen (Anfangstemperatur < 100°C nach Dotierung) Lithium Aluminiumhydrid (LiAlH4) ein vielversprechender Kandidat. Obwohl die reversible Dehydrierung von LiAlH4 mit Hilfe von organischen Lösungsmitteln durchgeführt werden muss, kann LiAlH4-Pulver als Einweg-Speichermaterial für verschiedene Anwendungen dienen, beispielsweise für Wasserstoff/Brennstoffzellensysteme. Diese Doktorarbeit beschäftigt sich mit LiAlH4 dotiert mit Übergangsmetall, mit dem Ziel maßgeschneiderte Dehydrierungseigenschaften zu erreichen. Die Kristallstruktur und die Morphologie der mit Übergangsmetallen dotierten LiAlH4-Pulver wurden mit Röntgenbeugung (XRD) und Rasterelektronenmikroskopie (REM) charakterisiert. Weiterhin wurde der positive Effekt der Dotanden auf die reaktionsfördernde Dehydrierung von LiAlH4 systematisch mit Hilfe thermoanalytischer Methoden untersucht. Für jedes Übergangsmetall, welches in Form von Übergangsmetallchloriden vorlag, wurden drei Dotierungsmethoden mit niedrigem Energieeintrag (Kugelmahlen mit geringer Rotations-geschwindigkeit, manuelles Schleifen/Magnetrühren, Magnetrühren mit Ethylether) verglichen, um LiAlH4-Pulver mit einer maximalen Wasserstofffreisetzungsmenge in Kombination mit einer schnellen Dehydrierungskinetik zu erzielen. Die Dehydrierung des dotierten LiAlH4-Pulvers wurde unter isothermen Bedingungen bei 80°C und einem H2-Druck von 1 bar gemessen, was im Bereich der Betriebstemperatur von PEM-Brennstoffzellen (Proton Exchange Membran) liegt. Dadurch sollen Anwendungen anvisiert werden, bei denen die entstehende Abwärme der Brennstoffzelle genutzt wird, um die Dehydrierung des Wasserstoffspeichermaterials auszulösen. Zudem wurde das Dehydrierungsverhalten des dotierten LiAlH4 bis zu einigen Monaten kontrolliert, um die mittelfristige Haltbarkeit zu testen. Weiterhin wurde die Pelletierung des mit Übergangsmetallen dotierten LiAlH4 mit dem Ziel untersucht, eine hohe volumetrische Speicherkapazität zu erreichen. Der Einfluss des Pressdrucks, der Dehydrierungstemperatur und des H2-Gegendrucks auf die Dehydrierungseigenschaften der mit Übergangsmetallen dotierten LiAlH4-Presslinge wurde systematisch analysiert. Außerdem wird die Volumenveränderung durch die Dehydrierung und die Kurzzeitspeicherung der mit Übergangsmetallen dotierten LiAlH4-Presslinge im Hinblick auf praktische Anwendungen unter Nutzung der Brennstoffzelle diskutiert.

在使用者較為密集的場合中，常會碰到無線網路壅塞的問題，例如在一個大型會議廳中，常會在各個IEEE 802.11頻道上部署不同的WiFi AP（Access Point），來分散使用者的連線。但是由於IEEE 802.11的連線機制是屬於使用者主導（client driven），只能透過使用者去選定AP進行連線，對於使用者裝置來說,，連線到不同AP的優先順序，是依照接收到不同AP的信號強度（RSSI）作為排序指標。這種做法會讓在空間上使用者分佈不平均的環境中，造成多數使用者UE只連線到少數AP，而其餘AP資源閒置無用的情形。 本論文提出一個IEEE 802.11的負載平衡解決方案，結合藍牙低功耗（Bluetooth Low Energy，BLE）及IEEE 802.11成為一個智慧型AP架構。我們利用藍牙低功耗通訊協定GATT （Generic Attribute Profile）分派AP給不同使用者進行連線，再結合馬可夫鏈平穩狀態分佈（Markov Chains Stationary Distribution）演算法，依照使用者在AP網路拓樸中的歷史分佈紀錄，將多個AP的分派轉化為Erlang-C模型的排隊系統以計算AP分派規則，藉此達到系統的負載平衡。 / Usually, a user crowded space encounters wireless network congestion problem. For example, a large conference hall often deploys different wireless AP (Access Point) on each IEEE 802.11 channel to separate users’ connections. However, since the connection mechanism of IEEE 802.11 is client driven, the AP connection is selected by the user and the selection is according to the received signal strength (RSSI) from different APs. This conventional approach may result in most of the user devices connect to relatively limited number of APs, and the resource of the rest of the APs left unused. This paper proposes a smart AP architecture which is able to manage load balance for IEEE 802.11 Wireless LAN using Bluetooth Low Energy (BLE) GATT (Generic Attribute Profile) protocol in order to appropriately assign AP to different user devices. The core AP assignment algorithm is based on Markov chain stationary distribution. Simulation results show that the proposed BM-MS (BLE Management with Markov-Chains Stationary Load Balance) method outperforms RSSI based method in terms of system throughput and average user data rate.

This dissertation has investigated the adsorption, thermal behavior, and radiation (both photon and electron) processing of prebiotically-relevant ices and surfaces. A custom ultra-high vacuum (UHV) chamber has been built that is coupled with a Fourier Transform-Infrared (FT IR) spectrometer and a Temperature Programmed Desorption (TPD) system that utilizes Quadrupole Mass Spectrometry (QMS) to study selected organic:surface systems. Formamide (HCONH₂) has been studied in two related but distinct studies relevant to primitive Earth and interstellar chemistry. First, in collaboration with a theory group, formamide’s interaction with kaolinite (Al6Si6O36H30), a clay mineral relevant to early Earth chemistry, has been studied experimentally and theoretically. Experimental infrared results are compared with calculated infrared frequencies obtained by our collaborators. TPD analysis is compared with the calculated values of adsorption energy, and the optimal kaolinite termination site for adsorption is reported. Second, the first thermal and radiation damage study of pure formamide and HCONH₂:H₂O mixed ices on an interstellar icy grain analog (SiO₂) is reported. A discussion of the pure formamide ice phases identified with FT-IR upon warm-up, as well as the TPD binding energies of HCONH₂ on SiO₂, is presented. The observed Lyman-alpha photochemical products and proposed formation mechanisms from pure formamide ice is reported and discussed. In addition, results of Lyman alpha processing of mixed HCONH₂:H₂O ices are provided. Low-energy electron irradiation of pure HCONH₂ and HCONH₂:H₂O mixed ices has also been reported for the first time. A third investigation has studied acetylene (C₂D₂) and acetonitrile (CH₃CN) interactions and radiation stability in mixed low-temperature ices to simulate possible prebiotic reactions that may occur on Saturn’s moon, Titan. This investigation contributes to understanding the possible consumption, trapping, and degradation of these species on the surface of Titan.