L’objectif de la thèse a été de comprendre la dynamique protonique et l'adsorption d'eau dans de nouveaux matériaux poreux identifiés expérimentalement comme des candidats prometteurs pour des applications dans le domaine de la conduction protonique et du transfert de chaleur par adsorption. Dans ce contexte, des simulations à l’échelle électronique (Théorie de la fonctionnelle de la Densité) et atomique (Monte Carlo et Dynamique Moléculaire classique) ont permis (i) d’élucider les mécanismes de conduction protonique assistées par l’eau de deux matériaux hybrides de type MOFs, MIL-163(Zr) et KAUST-7', et d'un phosphate de titane, TiIVTiIV(HPO4)4 à l’origine de leurs performances exceptionnelles et (ii) d’interpréter les comportements d’adsorption de l’eau d’une série de matériaux hybrides CUK-1(Me), MOF-801(Zr) and MIL-100(Fe) qui peuvent être modulées par la nature de leur centre métallique, la création de défauts et l’incorporation de sites de coordination insaturés. Cette connaissance fondamentale devrait permettre de voir émerger de façon plus efficace des matériaux pour les deux applications visées. / The objective of this PhD thesis was to gain insight into the proton dynamics and water adsorption mechanisms in novel porous materials that have been identified experimentally as promising candidates for low temperature proton conduction and adsorption-based heat reallocation-related applications. This was achieved by combining advanced computational tools at the electronic (Density Functional Theory) and atomic (force field_based Monte Carlo and Molecular Dynamics) levels to (i) reveal the water-assisted proton migration pathway through the pores of the hybrid metal organic frameworks MIL-163(Zr) and KAUST-7’and the inorganic phosphonate TiIVTiIV(HPO4)4 materials at the origin of their outstanding proton conduction performances and (ii) explain the water adsorption behaviors of a series of metal organic frameworks CUK-1(Me), MOF-801(Zr) and MIL-100(Fe) that can be tuned by changing the nature of the metal center, creating defects and incorporating coordinatively unsaturated sites. Such a fundamental understanding is expected to pave the way towards a more efficient development of materials for the two explored applications.

The androgen receptor (AR) fulfils important roles for both sexes. By mediating the biological function of androgens, the AR has remained the target for endocrine therapies treating prostate cancer. The AR also determines the effectiveness of medroxyprogesterone acetate (MPA) in treating AR positive breast cancer. Every man will be affected by prostate cancer if he lives long enough. Prostate cancer continues to be a leading cause of death for males despite research into this cancer covering more than 60 years since Huggins' seminal 1941 study showing that androgens play a key role in this cancer. Unfortunately, significant advances have not been forthcoming and the effect of treatment has remained largely the same over past decades, whereby initial treatment provides temporary remission but eventually advanced cases become refractory to further intervention and the disease recurs in a more aggressive form. A plethora of factors are exquisitely sensitive to minute changes in the AR's structural profile, which can be altered by a single mutation, resulting in aberrant activity. A principal feature of these variant ARs associated with prostate cancer, is enhanced capacity to bind a number of molecules other than its cognate ligand, dihydrotestosterone (DHT). The promiscuous activity of this receptor leads to continued AR signalling and stimulus for the cancer despite low androgen levels induced by treatment regimes. A key question is whether mutations occurring within the AR occur as a result of cancer or contribute to the propagation of the cancer. Recent research has demonstrated that treatments incorporating anti-androgens such as flutamide, which are designed to impede prostate cancer progression by inhibiting AR activity, may actually provide selective pressure favouring somatic mutation of the receptor to take place. The specific changes to the AR which are responsible for gains of function have not been resolved as their crystal structures, which are used to provide conformational analysis of proteins, are tremendously problematic to produce with little success found in literature. Generating representative crystals of the AR protein involves producing soluble recombinant protein. Unfortunately the AR is prone to aggregation and is highly unstable, especially in the presence of antagonistic molecules or absence of a stabilising ligand, preventing the protein from being maintained in the soluble state required for crystallization. In order to produce sufficient quantities of soluble material for crystallization, the androgen receptor's ligand binding domain (LBD) was produced as a recombinant protein in Escherichia coli bacteria strain BL21 (DE3) in the presence of DHT, flutamide, as well as in the absence of ligand. Since soluble unbound AR-LBD has not been produced until now, the bacterial culture containing no ligand was further processed to the stage of cleaving the purification tag from the recombinant protein and represents considerable progress into producing soluble material for crystallizing the troublesome yet considerably important AR in the absence of ligand. Although distinct from prostate cancer in males, AR activity in breast tissue is also a factor determining the action of drugs, such as MPA, included in therapies aimed at breast cancer. The use of MPA has declined primarily due to its adverse effects including unsuccessful generation of a biological response, as well as the advent of other drugs administered for hormonal therapies treating breast cancer. Alternative drugs are needed when breast cancer therapies fail as tumours develop resistance to primary drugs. Although there are a number of drugs on the market, success would be maximised if the determined therapy is matched with the patient, based for example, on their genetic makeup. There is a conundrum whereby some patients with an AR do not respond to MPA, a drug normally recognised by the receptor. In clinical trials it was discovered that an AR with threonine instead of methionine at residue 780 (M780T) fails to activate in response to MPA, but the exact mechanism has remained elusive and needs to be answered at the molecular level. The X-ray crystallographic studies that generate 3D images of macromolecules and wet chemistry, which have traditionally been used to provide insight into science in these dimensions, are incorporated with computer based molecular simulation. This is both complementary and distinct to traditional scientific methodologies, enabling further elucidation of protein-protein interactions, and the influence applied to such inter-relations by natural and drug ligands. This approach has been used, and is continually developed, to understand the binding mechanisms of current drugs as well as designing new drugs. In order to produce a receptor representing the M780T variant, the crystal structure representing the AR-LBD was mutated in silico, into which MPA was then docked. It was found that MPA binds into the M780T AR-LBD with considerably more spatial displacement compared to the position of DHT in the crystal structure, and is predicted to be the primary reason for the inability of MPA to activate this variant AR. The clarification of MPA binding and failure to elicit a response from the variant AR is significant for a cohort of breast cancer patients, as not only does the presence of an AR in the tumour determine the effectiveness of MPA, but correct composition of the AR, specifically, the absence of a M780T mutation. In the absence of this AR mutation, MPA could effectively be used either as an alternative to primary drugs, or in secondary therapies when primary therapies fail. Aberrant activity of variant ARs in response to MPA should also be taken into consideration when analysing drug studies about the effectiveness of MPA. The findings on the loss of response to MPA by the M780T variant AR have been included in the journal article &quotDecreased Androgen Receptor Levels and Receptor Function in Breast Cancer Contribute to the Failure of Response to Medroxyprogesterone Acetate" appearing in the September 2005 issue of Cancer Research journal.

Submitted by ALEXANDRE BORGES null (alex.brgs@hotmail.com) on 2016-06-28T19:11:21Z No. of bitstreams: 1 ESTUDOS DE MODELAGEM MOLECULAR DE LIGNANAS EM COMPLEXOS COM CICLOOXIGENASES-1 E 2.pdf: 4325586 bytes, checksum: 2f6ab56677aea7746bd28dad4b24ea23 (MD5) / Approved for entry into archive by Ana Paula Grisoto (grisotoana@reitoria.unesp.br) on 2016-06-29T17:41:12Z (GMT) No. of bitstreams: 1 borges_a_dr_ilha.pdf: 4325586 bytes, checksum: 2f6ab56677aea7746bd28dad4b24ea23 (MD5) / Made available in DSpace on 2016-06-29T17:41:12Z (GMT). No. of bitstreams: 1 borges_a_dr_ilha.pdf: 4325586 bytes, checksum: 2f6ab56677aea7746bd28dad4b24ea23 (MD5) Previous issue date: 2016-05-11 / Os inibidores seletivos da ciclooxigenase-2 (COX-2), como o rofecoxibe (2) e o celecoxibe (1), formam uma importante classe de medicamentos anti-inflamatórios desenvolvidos a partir da descoberta das duas isoformas das ciclooxigenases (COX-1 e COX-2) na década de 1979. A isoforma 1 esta relacionada com a citoproteção gástrica, agregação plaquetária e função renal e a isoforma 2 relacionada a processos inflamatórios. Estes inibidores seletivos apesar de não apresentarem os efeitos colaterais (ulceras e gastrites) dos anti-inflamatórios não esteroidais (AINEs) clássicos por inibirem apenas a COX-2, apresentam grave risco cardiovascular, o que motivou à retirada do rofecoxibe do mercado. Porém, por ser um eficiente inibidor seletivo da COX-2 a estrutura do rofecoxibe tornou-se referência no estudo de novas substâncias capazes de inibir seletivamente a COX-2. Dentre as ferramentas utilizadas na busca destas novas estruturas está a modelagem molecular através de programas como o GOLD 5.1, que foi utilizado neste trabalho. O uso do GOLD 5.1 possibilitou o estudo do comportamento das estruturas avaliadas em ligação com as ciclooxigenases. O objetivo foi obtenção de estruturas com comportamento semelhante ao rofecoxibe (em relação às COXs) como potenciais candidatos ao desenvolvimento de novos inibidores seletivos para a COX-2. O estudo foi realizado com 480 estruturas modeladas a partir de lignanas naturais como a hinoquinina, cubebina, deoxipodofilotoxina e podofilotoxina, que apresentam atividade anti-inflamatória in vivo ou in vitro, além de semelhanças estruturais com o rofecoxibe. A deoxipodofilotoxina por apresentar seletividade para a COX-2 em ensaio in vitro também foi utilizada como estrutura de referência além do rofecoxibe. Os resultados observados a partir da simulação molecular permitiram concluir que embora tanto o rofecoxibe como a deoxipodofilotoxina (3) inibam seletivamente a COX-2 in vitro, o fazem de modo diferente. Em relação a COX-2 as duas estruturas ocupam a mesma região do sítio ativo, mas o rofecoxibe apresenta interações mais fortes com o bolso hidrofílico desta isoforma (condição necessária para a inibição seletiva para os coxibes). Já para a COX-1 enquanto o rofecoxibe ocupa a porção superior do canal hidrofóbico (sítio ativo) como os demais AINEs, a deoxipodofilotoxina ocupa uma região vizinha. Pelos resultados obtidos é possível sugerir que tanto a maior flexibilidade das estruturas como a presença do anel lactônico, são importantes para um comportamento análogo ao rofecoxibe ou à deoxipodofilotoxina. Com relação à interação com o bolso hidrofílico da COX-2, os resultados sugerem que a presença de grupos aceptores de prótons menos volumosos nas posições C3 e C4, C3’ e C4’ ou C4 levam a resultados melhores que grupos aceptores de maior volume. A presença de grupos doadores de prótons apesar de permitirem forte interação com o bolso hidrofílico da COX-2 leva a resultados globais insatisfatórios, pois formam interações fortes com o resíduo Arg120 do sítio ativo da COX-1, interação considerada importante para a inibição não seletiva. Resultado semelhante à deoxipodofilotoxina foi observado apenas para a estrutura 17. As estruturas 37, 188, 266, 267, 348 e a hinoquinina (4) apresentam resultados semelhantes ao rofecoxibe, para as duas isoformas. Deste modo permite-se sugerir a partir dos resultados obtidos neste estudo que a hinoquinina (4) e as estruturas 17, 37, 188, 266, 267 e 348 apresentam-se como possíveis protótipos de fármacos que atuem como inibidores seletivos para a COX-2. / The selective inhibitors of the cyclooxygenase-2 (COX-2) as rofecoxib (2) and celecoxib (1), form an important class of anti-inflammatory drugs developed from the discovery of two isoforms of cyclooxygenases (COX-1 and COX-2) in the late 1979. Isoform 1 is related to the gastric cytoprotection, platelet and renal function and isoform 2 related to inflammatory processes. These selective inhibitors although they did not side effects (ulcers and gastritis) of the classic NSAIDs to inhibit only COX-2, have severe cardiovascular risk, which led to the withdrawal of rofecoxib from the market. However, to be an effective selective COX-2 to rofecoxib structure has a reference in the study of new substances capable of selectively inhibiting COX-2. Among the tools used in the search of these new structures is by molecular modeling program such as GOLD 5.1, which was used in this work. Using GOLD 5.1 made it possible to study the behavior of structures evaluated in binding with the cyclooxygenases. With the objective of obtaining structures with similar behavior to rofecoxib (regarding behavior with COX) as potential candidates for the development of new selective inhibitors for COX-2. The study was conducted with 480 structures modeled from natural lignans as hinokinin, cubebin, deoxypodophyllotoxin and podophyllotoxin, which have anti-inflammatory activity in vivo or in vitro as well as structural similarities with rofecoxib. The deoxypodophyllotoxin for presenting selectivity for COX-2 in the in vitro assay was also used as a reference structure beyond rofecoxib. The results observed from the molecular simulation showed that although both rofecoxib (2) as deoxypodophyllotoxin (3) selectively inhibit COX-2 in vitro, they do differently. In relation to COX-2 the two structures occupy the same region of the active site, but rofecoxib has stronger interactions with the hydrophilic pocket of this isoform (a necessary condition for the selective inhibition for coxibs). As for the COX-1 while rofecoxib occupies the upper portion of the hydrophobic channel (active site) like other NSAIDs, the deoxypodophyllotoxin occupies a neighboring region. From the results it is possible to suggest that the greater flexibility of the structures such as the presence of the lactone ring, are important for a similar behavior to rofecoxib or deoxipodofilotoxina. With respect to the interaction with the hydrophilic pocket COX-2, the results suggest that the presence of acceptors groups less bulky protons in posítions C3 and C4, C3 ' and C4' and C4 lead to better results than acceptors groups of larger volume. The presence of proton donors groups despite allowing strong interaction with the hydrophilic pocket COX-2 lead to poor overall results, since they form strong interactions with Arg120 residue of COX-1 active site, considered important interaction for inhibiting non-selective. Results similar to deoxipodofilotoxina was only observed for structure 17. Structures 37, 188, 266, 267, 348 and hinokinin (4) show results similar to rofecoxib for the two isoforSA. Thus it allows suggest from the results obtained in this study hinokinin (4) and structures 17, 37, 188, 266, 267 and 348 are shown as possible prototype drugs that act as selective inhibitors for COX-2.

Orientador: Rosangela da Silva de Laurentiz / Resumo: Os inibidores seletivos da ciclooxigenase-2 (COX-2), como o rofecoxibe (2) e o celecoxibe (1), formam uma importante classe de medicamentos anti-inflamatórios desenvolvidos a partir da descoberta das duas isoformas das ciclooxigenases (COX-1 e COX-2) na década de 1979. A isoforma 1 esta relacionada com a citoproteção gástrica, agregação plaquetária e função renal e a isoforma 2 relacionada a processos inflamatórios. Estes inibidores seletivos apesar de não apresentarem os efeitos colaterais (ulceras e gastrites) dos anti-inflamatórios não esteroidais (AINEs) clássicos por inibirem apenas a COX-2, apresentam grave risco cardiovascular, o que motivou à retirada do rofecoxibe do mercado. Porém, por ser um eficiente inibidor seletivo da COX-2 a estrutura do rofecoxibe tornou-se referência no estudo de novas substâncias capazes de inibir seletivamente a COX-2. Dentre as ferramentas utilizadas na busca destas novas estruturas está a modelagem molecular através de programas como o GOLD 5.1, que foi utilizado neste trabalho. O uso do GOLD 5.1 possibilitou o estudo do comportamento das estruturas avaliadas em ligação com as ciclooxigenases. O objetivo foi obtenção de estruturas com comportamento semelhante ao rofecoxibe (em relação às COXs) como potenciais candidatos ao desenvolvimento de novos inibidores seletivos para a COX-2. O estudo foi realizado com 480 estruturas modeladas a partir de lignanas naturais como a hinoquinina, cubebina, deoxipodofilotoxina e podofilotoxina, que apre... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The selective inhibitors of the cyclooxygenase-2 (COX-2) as rofecoxib (2) and celecoxib (1), form an important class of anti-inflammatory drugs developed from the discovery of two isoforms of cyclooxygenases (COX-1 and COX-2) in the late 1979. Isoform 1 is related to the gastric cytoprotection, platelet and renal function and isoform 2 related to inflammatory processes. These selective inhibitors although they did not side effects (ulcers and gastritis) of the classic NSAIDs to inhibit only COX-2, have severe cardiovascular risk, which led to the withdrawal of rofecoxib from the market. However, to be an effective selective COX-2 to rofecoxib structure has a reference in the study of new substances capable of selectively inhibiting COX-2. Among the tools used in the search of these new structures is by molecular modeling program such as GOLD 5.1, which was used in this work. Using GOLD 5.1 made it possible to study the behavior of structures evaluated in binding with the cyclooxygenases. With the objective of obtaining structures with similar behavior to rofecoxib (regarding behavior with COX) as potential candidates for the development of new selective inhibitors for COX-2. The study was conducted with 480 structures modeled from natural lignans as hinokinin, cubebin, deoxypodophyllotoxin and podophyllotoxin, which have anti-inflammatory activity in vivo or in vitro as well as structural similarities with rofecoxib. The deoxypodophyllotoxin for presenting selectivity for COX... (Complete abstract click electronic access below) / Doutor

Homogeneous crystal nucleation in binary hard sphere mixtures is an active area of research for last two decades. Although Classical nucleation theory (CNT) gives a qualitative picture, it fails at high super saturations because of the following reasons. CNT assumes that the cluster formed is spherical in shape, its properties can be modeled using the bulk properties of the stable solid phase and the interfacial free energy γ between the nucleus and the surrounding metastable fluid is equal to the planar surface tension between two phases at coexistence. These assumptions get increasingly tenuous at higher degrees of super saturations where the critical nucleus formed is microscopic in size leading to breakdown in the predictions of CNT. In addition direct experimental observation of critical nucleus is very difficult because, 1. Critical nucleus is microscopic in size, consisting of few hundreds of particles. 2. Formation of critical cluster is very rare (typically of the order of 101– 106nuclei/cm3/s) 3. Its life time is very short (it either rapidly grows to form a solid phase or melts back to fluid) In these circumstances molecular simulations are an attractive tool to study the crystal nucleation, because in these simulations microscopic size critical nucleus properties can be calculated. However, brute force molecular dynamic (MD) simulation techniques to study the homogeneous crystal nucleation is currently not feasible due to long times involved. Hence, an indirect approach is needed. In this work, Monte Carlo Abstract v (MC) molecular simulation techniques are used to calculate free energy barrier height during the crystal nucleation. Phase behavior of Binary hard sphere mixtures with varying ratios of smaller diameter to larger diameter (α) is very similar to that of binary organic liquids. By studying the crystal nucleation in hard sphere system, the physics behind the nucleation for binary organic liquids can be understood. This is the key motivation to study the homogeneous crystal nucleation in binary hard sphere mixtures using MC simulations. Simulations were done using umbrella sampling in combination with local bond order analysis for the identification of crystal nuclei and to compute shape and height of nucleation barrier. Parallel tempering scheme of Geyer and Thomson was utilized to sample phase space more efficiently. Parallel tempering technique was implemented using Message passing interface (MPI) libraries. By using all the above Monte-Carlo simulation techniques, nucleation barrier was calculated during crystallization of binary hard sphere mixtures under the moderate degrees of super cooling in Isothermal-Isobaric semi grand ensembles. Crystal nucleation in binary hard sphere mixtures has been studied for size ratios α = 0.85, 0.42 and 0.43. For α=0.85, phase diagram contains eutectic point. In this system, the effect of eutectic composition on the nucleation barrier height was observed, by calculating nucleation barriers at various fluid mixture compositions keeping Laplace pressure constant. It is observed that as the fluid mixture composition move towards the eutectic point, free energy barrier height, surface tension and critical cluster sizes are increased and the nucleation rate is drastically decreased by a factor of 10-31. Thus the difficulty of homogenous crystal nucleation increases near the eutectic point. For α=0.42 and 0.43 in the hard sphere system, compound solids such as AB and AB2 solids are stable respectively. In these systems crystal nucleation study was done to observe the compound solid formation. It is observed that in these systems crystallization kinetics are very slow and more advanced simulation techniques need to be developed in order to study crystal nucleation.

Metal–organic materials (MOMs), a type of porous crystalline structure composed of organic ligands jointed with metal ions, have captured the interest of scientists as potentially useful in gas sorption applications. Some of the most crucial avenues of investigation are in H2 storage (for use as a clean burning fuel source) and CO2 capture and sequestration (to remove the greenhouse gas from the environment). A major advantage of MOMs for such applications is their high variability in terms of physical dimensions and chemical moieties, based on composition and synthesis conditions, making them potentially customizable for specific application if necessary structural characteristics are known. Computational experimentation is an important avenue for determining such specifications as it allows examination of gas/MOM interaction at the molecular level. In this dissertation a number of MOM structure are computationally studied in order to elucidate gas sorption mechanisms. These systems were probed by classical simulation using grand canonical Monte Carlo with a carefully chosen set of intermolecular interaction parameters. While the focus of this work is specifically H2 and CO2 sorptive behavior, the insights gained from simulation extend beyond these specific applications. Addressed first are a series of MOMs with rht topology, which possesses asymmetric copper paddle-wheels and easily functionalized linkers. Beginning with a prototypical structure and then branching out into more chemically interesting variants revealed surprising gas sorption behavior about the metal paddle-wheels (with a definite preference for one copper over its counterpart). A synthetic strategy for controlling the preferred open-metal sorption site through the inclusion of electron rich functionality in the linker bodies, was also revealed. An additional MOM with similar composition components, exhibiting zyg topology, also showed this metal preference effect on the asymmetric paddle-wheels. A second class of MOMs, composed of square-pillared grids and known as the SIFSIX series (due to the inclusion of SiF62− as pillaring units) was also examined. These structures have been shown excellent results for CO2 sorption making the elucidation of the sorptive mechanisms of great interest. Six different structures were examined, probing the effects of linker length, metal selection, and interpenitration of unbonded scaffolds. The nature of the CO2-MOM sorption interactions were revealed through simulation and provided insights regarding the synergistic effect of pore dimensions and SiF62− functionality for specifying specific behavior (i.e. high selectivity vs. high uptake). A final MOM, composed of Y3+ ions and chemically complex linkers, was also examined. Disorder in the crystallographic data (e.g. single atoms with multiple positions) indicated the coexistance of notably different unit cells in the same system. Nevertheless, simulations revealed favored sorption sites in conjunction with results from physical experimentation.

Ayant une structure particulière avec une tête hydrophile et une queue hydrophobe, des molécules amphiphile ont de nombreuses applications importantes, comme par exemple, la fabrication des détergents, la protection et la fonctionnalisation de surfaces, etc. Des agrégats de diverses formes, micelles, véhicules, membranes etc., peuvent se former à partir des amphiphiles. La complexité de ces agrégats moléculaires rend l’étude théorique de ce type de systèmes extrêmement difficile. Jusqu’à présent, notre connaissance sur l’interaction entre des nanoparticules et des agrégats des amphiphiles reste encore incomplète. A l’aide de certaines méthodes de simulations moléculaire et une approche théorique, nous avons entrepris une série d’études pour mieux comprendre les questions fondamentales suivantes :1. Comment la présence de nanoparticules, notamment la courbure de ses surfaces, affecte l’agrégation de molécules amphiphile ?2. Comment une bicouche de lipide, une forme d’agrégat particulier des amphiphile, peut induire l’assemblage auto-organisé de nanoparticules hydrophobes ?3. Est-ce que la présence des nanoparticules peut provoquer des transitions morphologiques d’un nanotube membranaire ? / Amphiphile molecules, endowed with a particular structure containing a hydrophilic head and a hydrophobic tail, have many important applications, e.g., fabrication of detergents, surface coating or surface functionalization, etc. Molecular aggregates of various forms, micelles, vehicle, membranes, etc. can be formed from amphiphile molecules. The complexity of these molecular aggregates involving a large number of atoms make the theoretical study of these system very challenging. Up to now, our understanding of the interaction between nanoparticles and aggregates of amphiphiles remains quite incomplete. Using a variety of molecular simulation methods and some theoretical approaches (Helfrich theory and perturbation theory), we have studied the following issues in the present thesis: 1. How the presence of nanoparticles, especially due to their highly curved surfaces, affects the aggregation of the amphiphiles? 2. How a lipid bilayer, a particular amphiphile aggregate, induces the self-assembly of hydrophobic nanoparticles.3. How the morphology transition of a membrane nanotube can be induced by nanoparticles?

No description available.

The effect of inclusion of explicit polarization is investigated through several theoret- ical studies of crystalline porous materials herein. In addition to the use of Monte Carlo simulation for such studies, a robust molecular dynamics software is presented which is suitable for analyzing time dependent properties of gases or other molecules in porous materials and other condensed phase systems. Metal-organic frameworks (MOFs) are the main focus of the work included here, a relatively young class of materials originally in- troduced in the early 1990s. These are usually three dimensional crystalline nanoporous materials that exhibit unique properties such as gas separation, storage and catalysis. They are synthesized by the combination of a metal ion e.g. Cu2+ with an organic linker e.g. benzene dicarboxylate. They are a very popular topic of scientific research due to the diversity in possible structures and manifold utility – finding applications in electron transfer, sensing, drug release etc. Industrially, MOFs like HKUST-1 and others are on the global market for use in gas storage and separation in fuel cell and raw materials processing. These materials are often ideal candidates for computer simulation owing to their crystalline nature – a very large atomic system (that is, moles of particles) can be under- stood by only evaluating one or a few unit cells of the MOF, usually less than 5,000 atoms, and macroscopic properties such as gas sorption capacity and diffusion coefficients can be calculated through extrapolation of atomistic interactions in a mathematically infinite lattice. The software developed by the space group as of 2005, Massively Parallel Monte Carlo (MPMC), allows for sophisticated calculation of repulsion dispersion, electrostatic and polarization energies. In this work, Monte Carlo Molecular Dynamics (MCMD) is in- troduced, which can hybridize both methods to explore the phase space of a system with ease and better efficiency, as well as explore the effects of MOF flexibility and dynamic properties which to-date are rarely studied. Studies involving primarily CO2, H2 and CH4 will be presented, but other gases investigated include C2 H2 , C2 H4 , C2 H6 , N2 , H2 O and others. Metal-organic materials with a wide variety of composition and structure will also be presented. Finally, features of the software MCMD will be presented for use by future studies.

Les réservoirs de type "tight oil" et "shale gaz" ont une partie importante de leur volume poreux occupée par des micropores (< 2nm) et des mesopores (entre 2 et 50 nm). Ce type d'environnement crée de fortes forces d’interaction dans le fluide confiné avec les parois du pores et entre ses propres molécules, ce qui change fortement la thermodynamique du fluide. Un travail important doit donc être effectué sur le développement de méthodes de mise à l'échelle de la distribution de pore pour effectuer des simulations réservoir à grande échelle. Premièrement, des simulations moléculaires sont effectuées sur des fluides confinés afin d'obtenir des propriétés thermodynamiques de référence à l'équilibre liquide/vapeur pour différentes tailles de pore. Ensuite, une comparaison des données de simulation moléculaire avec les résultats issus des équation d'état utilisées dans la littérature a permis de mettre en valeur la méthode de flash avec pression capillaire et changement du point critique comme la meilleure méthode existante pour décrire la physique du fluide confiné. Des simulations fines d'écoulement matrice/fracture ont donc été effectuées pour différentes tailles de pore. Des modèles de mise à l'échelle en maillage grossier ont été ensuite construits à partir du même cas synthétique et les résultats ont été comparés avec ceux des simulations de référence en maillage fin. Un nouveau modèle de triple porosité considérant fracture, petit pores et grand pores avec une approche MINC a donné des résultats très proches du maillage fin. Finalement un réservoir stimulé hydrauliquement à grande échelle a été simulé pour différentes distributions de pores avec le modèle développé. / Tight oil and shale gas reservoirs have a significant part of their pore volume occupied by micro (below 2nm) and mesopores (between 2 and 50nm). This kind of environment creates strong interaction forces in the confined fluid with pore walls as well as between its own molecules and then changes dramatically the fluid phase behavior. An important work has therefore to be done on developing upscaling methodology of the pore size distribution for large scale reservoir simulations. Firstly, molecular simulations are performed on different confined fluids in order to get reference thermodynamic properties at liquid/vapor equilibrium for different pore sizes. Then, the comparison with commonly used modified equation of state (EOS) in the literature highlighted the model of flash with capillary pressure and critical temperature and pressure shift as the best one to match reference molecular simulation results. Afterwards fine grid matrix/fracture simulations have been built and performed for different pore size distributions. Then, coarse grid upscaling models have then been performed on the same synthetic case and compared to the reference fine grid results. A new triple porosity model considering fracture, small pores and large pores with MINC (Multiple Interacting Continua) approach, has shown very good match with the reference fine grid results. Finally a large scale stimulated reservoir volume with different pore size distribution inside the matrix has been built using the upscaling method developed here.