The removal of trace amounts of acetylene in ethylene streams is a high-volume industrial process that must possess high selectivity of alkyne hydrogenation over hydrogenation of alkenes. Current technology uses metallic nanoparticles, typically palladium or platinum, for acetylene removal. However, problems arise due to the deactivation of the catalysts at high temperatures as well as low selectivities at high conversions. Pore expanded MCM-41 is synthesized via a two-step strategy in which MCM-41 was prepared via cetyltrimethylammonium bromide (CTMABr) followed by the hydrothermal treatment with N,N-dimethyldecylamine (DMDA). This material was washed with ethanol to remove DMDA, or calcined to remove both surfactants. PE-MCM-41 based materials were impregnated with palladium, gold, and palladium-gold nanoparticles. The removal of DMDA had an effect on both the conversion and selectivity, in which they were found to drop significantly. However, by using the bimetallic PdAu catalysts, higher selectivity could be achieved due to increased electron density.

The green production of formic acid and utilization of methanol over heterogeneous catalysis system were investigated in this study. The heterogeneous catalysts are widely used in the chemical industry. They offer high stability and reusability which can enhance the production ability and lower the production cost, it can be considered as the sustainable energy solution for the future. In this work, we demonstrated several different heterogeneous catalysts for sustainable formic acid production and methanol utilization, including the heteropoly acid supported mesoporous silica catalysts and multi-function mixed metal oxide catalysts. Detailed characterizations of the final products were carried out by N2 adsorption and desorption, XRD, HR-TEM, SEM, ICP-OES, XANES, NH3- TPD, Raman spectroscopy, and FTIR to identify the chemical properties and physical properties of the catalysts. We obtained 60 % glycerol conversion and 30 % formic acid selectivity with at least 3 rounds of usages in batch system over PV1Mo/SBA-15-p-DS catalyst. Moreover, the continuous methyl formate production with significantly high formation rate (16.7) has been achieved via our CuMgO-based catalysts, and the best Cu5MgO5 catalyst gives more than 80 % methanol conversion with constant selectivity to methyl formate even after 4 catalytic test (more than 200 h), revealing their potential for industrialization. For the methanol utilization reaction, the methanol homocoupling to form dimethoxymethane (DMM) has been investigated. The redox and acidic properties of catalysts both play a critical role in this reaction and the related to different product. The supported V2O5 catalyst achieves the best catalytic performance (62.1 % conversion and 85.6 % DMM selectivity) with a Ce/Al specific ratio of 1. This research not only provided the efficient catalysts for numerous application with high activity, but also discovered the relation between the catalytic performance and the nature of the materials. These findings might further help the researcher to solve the global environmental and energy issues in the near future.

Artificial Neural Networks (ANNs) have been established as one of the most important algorithmic tools in the Machine Learning (ML) toolbox over the past few decades. ANNs' recent rise to widespread acceptance can be attributed to two developments: (1) the availability of large-scale training and testing datasets; and (2) the availability of new computer architectures for which ANN implementations are orders of magnitude more efficient. In this thesis, I present research on two aspects of the second development. First, I present a portable, open source implementation of ANNs in OpenCL and MPI. Second, I present performance and scaling models for ANN algorithms on state-of-the-art Graphics Processing Unit (GPU) based parallel compute clusters. / Master of Science / Artificial Neural Networks (ANNs) have been established as one of the most important algorithmic tools in the Machine Learning (ML) toolbox over the past few decades. ANNs’ recent rise to widespread acceptance can be attributed to two developments: (1) the availability of large-scale training and testing datasets; and (2) the availability of new computer architectures for which ANN implementations are orders of magnitude more efficient. In this thesis, I present research on two aspects of the second development. First, I present a portable, open source implementation of ANNs in OpenCL and MPI. Second, I present performance and scaling models for ANN algorithms on state-of-the-art Graphics Processing Unit (GPU) based parallel compute clusters.

Multicore processors often increase the performance of applications. However, with their deeper pipelining, they have proven increasingly difficult to improve. In an attempt to deliver enhanced performance at lower power requirements, semiconductor microprocessor manufacturers have progressively utilised chip-multicore processors. Existing research has utilised a very common technique known as thread-level speculation. This technique attempts to compute results before the actual result is known. However, thread-level speculation impacts operation latency, circuit timing, confounds data cache behaviour and code generation in the compiler. We describe an software framework codenamed Lyuba that handles low-level data hazards and automatically recovers the application from data hazards without programmer and speculation intervention for an asymmetric chip-multicore processor. The problem of determining correct execution of multiple threads when data hazards occur on conventional symmetrical chip-multicore processors is a significant and on-going challenge. However, there has been very little focus on the use of asymmetrical (heterogeneous) processors with applications that have complex data dependencies. The purpose of this thesis is to: (i) define the development of a software framework for an asymmetric (heterogeneous) chip-multicore processor; (ii) present an optimal software control of hardware for distributed processing and recovery from violations;(iii) provides performance results of five applications using three datasets. Applications with a small dataset showed an improvement of 17% and a larger dataset showed an improvement of 16% giving overall 11% improvement in performance.

如今世界貿易的型態大多以中間財貿易為主，這些中間財貿易，來自於國與國之間的委外代工，中間財或者零組件透過貿易的方式運送回母國組裝。然而Melitz(2003)指出生產力差異是新貿易理論重要的議題，也就是說，研究今日的委外代工行為，必須考慮廠商之間生產力的差異，而本文就是以此點當作立論的依據，進行模型分析。 　　因為異質生產力的因素，使得每一家廠商的邊際生產力皆不相同，於是中間財委外代的程度亦不相同。並且，考慮委外代工產生的固定成本，生產力對於委外代工程度的影響關係曲線呈現凹性，這使得工資水準和委外代工程度的關係呈現不確定的現象。 　　我們將產業的類型區分為高中間財依存度廠商，例如製造產業就是屬於這類型；以及高最終財依存度廠商，例如高科技產業屬於這類型。在高中間財依存度的產業中，開放海外生產，會使得產業的平均工資上漲，壓縮良好的工作，擴張劣等工作，均衡失業率上升。當開放海外生產，高最終財部門依存度產業中，國內的平均工資與失業率變化是不確定的。 / Most of the types of world trade today are dominated by intermediate trade, which are derived from outsourcing or offshoring. Melitz (2003) pointed out that the difference in productivity is an important issue in the new trade theory, that is, the study of today's outsourcing foundry behavior, we must consider the difference between the productivity of manufacturers, and this article is based on this point as a basis, Carry out model analysis. Because the heterogeneity of productivity factors, making each company's marginal productivity are not the same, so the level of the middle finance committee is not the same. Moreover, considering the fixed costs generated by the outsourcing of foundry, the relationship between the productivity and the degree of subordinate foundry is concave, which makes the relationship between the wage level and the degree of outsourcing found an uncertain phenomenon. We classify the types of industries into Manufacturing and Headquarter industry. In Manufacturing industries, offshoring will make the industry's average wage rise, a good job compression, expansion of poor work, increases unemployment rate. However, under Headquarter industry, the domestic average wage and unemployment rate changes will be uncertain, if they decide offshoring.

Cette thèse s’inscrit dans un projet Européen TOPCOMBI de 22 partenaires. Plus spécifiquement, ce travail estle résultat d’une collaboration entre ENI (Italie), ITQ (Espagne), Repsol (Espagne) et IRCELYON (France).Ce travail est composé de 2 thématiques différentes dont les améliorations peuvent s’obtenir en trouvant denouveaux matériaux adaptés aux besoins.Les demandes énergétiques mondiales sont et seront en constante hausse ces prochaines années. Dans l’optiquede pallier à ce besoin, de nouvelles ressources doivent être trouvées et d’autres optimisées. Les énergies fossiles fontparties des ressources les plus utilisées dans le monde. Parmi c’est 3, le gaz naturel semble être le plus prometteur dupoint de vu du rendement énergétiques ou de l’impact écologique. Cependant, de nombreux champs de gaz ne peuventpas être traités car trop petit ou trop contaminés pour être économiquement viable. L’un des moyens pour les rendreattractifs est d’abaisser le coût de purification en utilisant de nouvelles techniques de séparation comme le système PSA.Cependant, il n’existe pas à l’heure actuelle d’adsorbant efficace pour permettre une purification économique viable.De nos jours, les réactions d’alkylation représentent un intérêt économique très important. Les procédésindustriels sont généralement effectués par des réactions acides homogènes ou non. A la vue des nouvelles restrictionsécologiques, certains procédés de catalyse homogène doivent être remplacés par des réactions catalytiques hétérogènespossédant les mêmes rendements. Pour ce faire, de nouveaux matériaux à caractères acide ont été utilisés : les zéolithes.En revanche la faible taille de pores de ces composés empêche les réactions d’alkylation sélective de molécules tropgrosse comme les composés poly-aromatiques.Depuis une vingtaine d’année, de nouveaux composés cristallins microporeux ont vu le jour. Les MOFs, MetalOrganic Frameworks. Ces composés ont la particularité d’être obtenus avec différents cations métalliques et ligandsorganiques. Ces combinaisons donnent une très grande diversité de ces composés au niveau des réactivités, du volumeporeux et de la taille des pores. Dans ce travail, nous avons essayé d’obtenir différents matériaux avec descaractéristiques précises afin de pouvoir les utiliser dans des procédés de purification du méthane ou dans la catalyseacide. Pour ce faire, nous avons mis en place un protocole de synthèses de MOFs à haut débit ainsi que d’un protocolede caractérisation adapté à nos besoins. Les différents composés remplissant les différents critères étant testés dans lesréactions adéquates. / This thesis is a European project TOPCOMBI of 22 partners. More specifically, this work is the result ofcollaboration between ENI (Italy), ITQ (Spain), Repsol (Spain) and IRCELYON (France).This work consists of 2 different themes which improvements can be obtained by finding new materials tailoredto the needs.The global energy demands are and will be constantly rising in the coming years. In order to meet this need,new resources must be found and further optimized. Fossil fuels are among the most used resources in the world. Amongthis 3, natural gas appears to be the most promising point of view of energy efficiency and ecological impact. However,many gas fields cannot be treated because there are too small or too contaminated to be economically viable. One way tomake them attractive is to lower the cost of purification using novel separation techniques such as the PSA system.However, there is no currently effective adsorbent to allow a viable economic cleansing.Today, the alkylation reactions represent a very important economic interest. Industrial processes are typicallycarried out by homogeneous acid reactions or not. Seeing the new environmental restrictions, some homogeneouscatalytic processes must be replaced by heterogeneous catalytic reactions with the same yields. To do this, new materialsto acid characters were used: the zeolites. However the small size of pores of these compounds prevents selectivealkylation reactions of molecules too large compounds such as poly-aromatic.For twenty years, new microporous crystalline compounds have emerged. The MOFs, Metal OrganicFrameworks. These compounds have the characteristic to be obtained with different metal cations and organic ligands.These combinations give a wide variety of these compounds at the level of reactivity, pore volume and pore size. In thiswork, we tried to get different materials with specific characteristics in order to use them in methods for purification ofmethane in acid catalysis. To do this, we have established a protocol for the synthesis of MOFs with high speed and acharacterization protocol suited to our needs. The different compounds fulfilling the various criteria being tested in theappropriate responses.

Thesis advisor: Chia-Kuang Tsung / Thesis advisor: Eranthie Weerapana / Metal-Organic Frameworks (MOFs) are crystalline porous materials constructed of metal ions and organic linkers, and have been widely utilized in gas storage, sensing, and chromatographic separation. The combination of MOF nanoparticles with other materials will broaden the utilization of MOF materials to a great extent. Several approaches for creating composites with the MOF, Zeolitic Imidazolate Framework-8 (ZIF-8), have been developed: dye and model drug molecules were encapsulated in ZIF-8 pores for potential drug delivery; mesoporous silica monolayer was epitaxially grown on the ZIF-8 surface for structural stability enhancement and hollow structure formation; UiO-66, another MOF subclass, was hierarchically encased inside ZIF-8 for double-phase gas separation and heterogeneous catalysis. By exploring the versatile ZIF-8 platform, these nanocomposites could have great applications in fields such as heterogeneous catalysis and drug delivery. / Thesis (MS) — Boston College, 2015. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Chemistry.

L’objectif de la thèse était la préparation et la caractérisation des catalyseurs hétérogènes à base de polyoxométalates. L’étude mécanistique d’oxydation du méthane jusqu’au méthanol a montré que sur des polyoxométalates supportés sur la silice l’activation C-H a lieu déjà à la température ambiante. L’adsorption du méthane-13C sur H4SiMo12O40 supporté, suivie par RMN solide a mis en évidence la création de l’espèce méthoxy [SiMo12O40(CH3)]3-. Le cycle catalytique est complété par l’hydrolyse de cette espèce - méthanol est formé et une molécule de l’eau recrée la structure du départ de polyoxométalate. L’adsorption du méthanol-13C sur des polyoxométalates a montré la création de deux types des espèces methoxy, localisées sur des atomes d’oxygènes terminaux ou pontants est caractérisées par deux signaux RMN distincts – à 58 et 77 ppm, respectivement. En greffant un complexe de platine PtMe2COD sur les sels de césium de polyoxometalates, le dégagement du méthane ou de la mélange du méthane et de l’éthane a été observé et expliqué par la séquence de l’addition oxydative du proton de polyoxometalate au centre métallique, couplage C-H ou C-C et finalement l’élimination réductrice et libération d’une molécule de gaz. Sels d’ammonium de l’acide phosphotungstique H3PW12O40 ont été montrées de catalyser l’isomérisation du n-butane a l’isobutane dans des conditions douces (225°C, 1 atm.). Composé du cuivre Cu(OTf)2 sur la surface des sels inorganiques des polyoxometalates donne des catalyseurs très actifs en insertion des carbènes aux liaisons C-H des éthers cycliques. / The aim of this work was preparation and characterization of catalysts based on polyoxometalates and their use in various catalytic reactions in heterogenous conditions. Methane C-H activation on silica-supported polyoxometalates was shown already at room temperature. Methoxy species [SiMo12O40(CH3)]3- from the 13C-enriched methane adsorption at 200°C on the surface of a silicadispersed silicomolybdic acid was detected by means of 13C SS NMR. Its hydrolysis led to methanol formation, thus completing the catalytic cycle. After 13C-enriched MeOH adsorption presence of two distinct methoxy species on the surface of polyoxometalates was shown, located on terminal (single coordinated) and bridging (double coordinated) oxygen atoms and resulting in the resonances at 58 and 77 ppm in 13C SS NMR. Grafting of PtMe2COD on the surface of various polyoxometalate supports led to methane or combined methane and ethane release, explained by means of oxidative addition/reductive elimination mechanism on metal centers. Ammonium salts of phosphotungstic acid catalyzed efficiently n-butane to isobutane skeletal isomerisation at mild conditions (225 °C, atmospheric pressure). Successful heterogenization of copper catalysts, active in enantioselective C-H carbene insertion reactions, on polyoxometalate supports have been shown

Cette thèse porte sur la production et l’étude de mousses polymères microporeuses, par dissolution de gaz. Plusieurs techniques expérimentales ont été améliorées ou adaptées pour la réalisation des mousses et la caractérisation des systèmes étudiés. De plus il est démontré que les paramètres du procédé, tels que la pression et la température, permettent la différenciation entre les différentes voies de moussage utilisées. Ceux-ci ont une influence significative sur les caractéristiques finales des mousses. / This dissertation focuses on the production and study of microcellular polymer foams by gas dissolution foaming. Due to the novelty of this research field several experimental techniques have been improved or adapted to produce the foams but also to provide valuable information from the systems understudy. It is demonstrated that processing parameters, such as pressure and temperature, allow differentiating between our foaming routes, and present a significant influence on the foaming process and final characteristics of foam.

Propelled by the explosive increases in mobile data traffic volume, existing wireless technologies are stretched to their capacity limits. There is a tremendous need for an expansion in system capacity and an improvement on energy efficiency. In addition, wireless network will support more and more multimedia services and applications, in which user experience has been always an important factor in evaluating the overall network performance. In order to keep pace with this explosion of data traffic and to meet the emerging quality of experience needs, wireless heterogeneous networks have been introduced as a promising network architecture evolution of the traditional cellular network. In this dissertation, we explore video quality-aware spectrum efficiency and energy efficiency in wireless heterogeneous networks|the potentials and the associated technical challenges. In particular, aiming to significantly enhance spectrum efficiency, we need to tackle the interference issue, which is exacerbated in heterogeneous network due to ultra dense node deployment as well as heterogeneity nature of various nodes. Specifically, werst study an optimal intra-cell inter-tier cooperation to mitigate interference between high power nodes and low power nodes. Together with cooperation, optimal mobile association and resource allocation schemes are also intensively investigated in heterogeneous network to achieve system load balancing so that bandwidth at high power and low power nodes can be utilized in the optimal way. The proposed scheme can greatly alleviate inter-tier interference and significantly increase overall system spectrum efficiency in a heterogeneous network. We then further apply advanced algorithms such as precoding, and non-orthogonal multiple access into intra-cell inter-tier cooperation so that the overall system spectrum efficiency and user experience are even more improved. When supporting a video type application in such a heterogeneous network, considering only spectrum efficiency is far from enough as video application is bandwidth consuming, battery consuming, and quality demanding. We develop a video quality-aware spectrum and energy efficiency resource allocation scheme in a wireless heterogeneous network and propose novel performance metrics to establish fundamental relationships among spectrum efficiency, energy efficiency, and quality of experience. Extensive simulations are conducted to evaluate the trade-o performance among three performance metrics.