Oxygen-diluted Combustion (OdC) technology has evolved from the concept of Excess Enthalpy Combustion and is characterized by reactants of low oxygen concentration and high temperature. Recent advances in this technology have demonstrated significant energy savings, high and uniform thermal field, low pollution, and the possibility for downsizing the equipment for a range of furnace applications. Moreover, the technology has shown promise for wider applications in various processes and power industries. The objectives of this thesis are to analyze the thermodynamic aspects of this novel combustion technology and to quantify the enhancement in efficiency and heat transfer inside a furnace in order to explore the potentials for reduced thermodynamic irreversibility of a combustion process and reduced energy consumption in an industrial furnace. Therefore, theoretical and experimental investigations were carried out. The 2nd law of thermodynamics analyses of OdC systems have been carried out for cases in which the oxidizer is either oxygen (Flameless-oxy-fuel) or air (High Temperature Air Combustion, HiTAC). The analyses demonstrate the possibilities of reducing thermodynamic irreversibility of combustion by considering an oxygen-diluted combustion process that utilizes both gas- and/or heat-recirculation. Furthermore, the results showed that an oxygen-diluted combustion system that utilizes oxygen as an oxidizer, in place of air, results in higher 1st and 2nd law efficiencies. Mathematical models for heat regenerators were developed to be designing tools for maximized heat recovery. These models were verified by heat performance experiments carried out on various heat regenerators. Furthermore, experiments were performed in a semi-industrial test furnace. It was equipped with various regenerative burning systems to establish combustion and heat transfer conditions prevailing in an industrial furnace operating based on HiTAC. The tests were carried out at seven firing configurations, two conventional and five HiTAC configurations, for direct and indirect heating systems. Measurements of energy balance were performed on the test furnace at various configurations in order to obtain the 1st law efficiency. Moreover, local measurements of temperature, gas composition, and heat fluxes in the semi-industrial test furnace were performed to find out the main characteristics of HiTAC flame and the effects of these characteristics on the heating potential, i.e., useful heating in the furnace. In the case of HiTAC, these measurements showed uniformities of chemistry, temperature, temperature fluctuation, and heat fluxes profiles. The values of fluctuations in temperature were small. The high speed jets of the fuel and air penetrated deep into the furnace. The fuel gradually disappeared while intermediate species gradually appeared in relatively high concentrations and at broader regions inside the furnace. These findings indicate: a large reaction zone, low specific combustion intensity in the flame, low specific fuel energy release, and high heat release from this large flame. In addition to the thermodynamic limitations to the maximum temperature of the Oxygen-diluted Combustion, the low specific energy release of the fuel and the high heat release from the flame to its surroundings cause this uniform and relatively moderate temperature profile in a HiTAC flame, consequently suppressing thermal-NO formation. Heat flux and energy balance measurements showed that heating potential is significantly increased in the case of HiTAC compared to that in the conventional case, implying much more energy savings than the apparent heat recovery from the heat regenerators, and consequently much less pollutants emissions. Therefore, it is certain that this large HiTAC flame emits more thermal radiation to its surroundings than the conventional flame does, in spite of the moderate-uniform temperature profile of the flame. This intense heat flux was more uniform in all HiTAC configurations, including the indirect heating configuration, than that of the conventional-air combustion configuration. / QC 20101011

Se presenta un modelo matemático para el diseño y simulación de un reactor de lecho fijo con enzimas inmovilizadas en partículas esféricas porosas. La ecuación de diseño del reactor se ha resuelto para el caso de un sistema enzimático, con limitaciones difusionales internas, que obedece a una cinética de Michaelis-Menten reversible.La validez del modelo se ha comprobado con el sistema enzimático naringina/naringinasa, aplicable al proceso de desamargado de zumos cítricos. / A mathematical model for design and simulation of a fixed bed reactor with immobilized enzymes in spherical particles is presented. The reactor design equation is solved for an enzymatic system taking into account internal diffusional limitations. Moreover, the enzyme obeys a reversible Michaelis-Menten kinetic. The validity of the model is checked by using the enzymatic system naringine/naringinase, which is used for fruit juice debbitering processes.

Die Arbeit zeigt wichtige Nutzungswege von fester Biomasse in Form von Holzhackschnitzel (HHS). Einleitend wird das Potenzial und der derzeitige Stand dargestellt. Es werden die physikalischen und chemischen Eigenschaften mit dem Schwerpunkt Wassergehalt in bezug auf die energetische Nutzung der HHS behandelt. Kerne der Nutzungswege bilden dabei die Vergasung im Gleichstromvergaser und die Co-Verbrennung in der Zirkulierenden Wirbelschicht. Mit Hilfe eines Versuchsvergasers werden die Auswirkungen von HHS-Eigenschaften auf den Betrieb untersucht. Der Modellvergaser IGEL bietet durch seine Konstruktion die Möglichkeit, innere Vorgänge zu beleuchten und Messungen in verschiedenen Vergaserebenen durchzuführen. Die Auswirkungen von verschiedenen Brennstoffchargen mit unterschiedlichem Wassergehalt führten zu Änderungen in der Gaszusammensetzung. Eigene Untersuchungen ermittelten einen Grenzwassergehalt, mit dem der Vergaser noch betrieben werden kann. Die Experimente an der Pilotanlage mit zirkulierender atmosphärischer Wirbelschicht befass-ten sich mit der wichtigen Frage, ob und welches NOx-Minderungspotenzial beim Einsatz von Biomasse vorliegt. Die mathematische Modellierung verdeutlicht die Nutzbarkeit von Simulationsprogrammen bei der Untersuchung von Einflüssen der Co-Verbrennung auf die NOx-Bildung. Hier wurden die Gesichtspunkte der Luftzahl, der Luftstufung, des Wassergehaltes, das Mischungsverhältnis und die Brennstoffstufung betrachtet. Eine Wirtschaftlichkeitsbetrachtung führt zu dem Ergebnis, dass Anlagen zur reinen Stromerzeugung mit Biomasse nur nahe der gesetzlichen Höchstleistung von 20 MWel zur Einspeisevergütung von wirtschaftlich betrieben werden können. Die ökologisch und ökonomisch günstigste Variante stellt die Co-Verbrennung in vorhandenen Anlagen dar. Die Kalkulationen zu den in Deutschland benötigten 20-MWel-Anlagen verdeutlichen, dass bei den gegenwärtig geplanten Heizkraftwerken das Potenzial an HHS schnell aufgebraucht ist.

<p>Oxygen-diluted Combustion (OdC) technology has evolved from the concept of Excess Enthalpy Combustion and is characterized by reactants of low oxygen concentration and high temperature. Recent advances in this technology have demonstrated significant energy savings, high and uniform thermal field, low pollution, and the possibility for downsizing the equipment for a range of furnace applications. Moreover, the technology has shown promise for wider applications in various processes and power industries.</p><p>The objectives of this thesis are to analyze the thermodynamic aspects of this novel combustion technology and to quantify the enhancement in efficiency and heat transfer inside a furnace in order to explore the potentials for reduced thermodynamic irreversibility of a combustion process and reduced energy consumption in an industrial furnace. Therefore, theoretical and experimental investigations were carried out.</p><p>The 2nd law of thermodynamics analyses of OdC systems have been carried out for cases in which the oxidizer is either oxygen (Flameless-oxy-fuel) or air (High Temperature Air Combustion, HiTAC). The analyses demonstrate the possibilities of reducing thermodynamic irreversibility of combustion by considering an oxygen-diluted combustion process that utilizes both gas- and/or heat-recirculation. Furthermore, the results showed that an oxygen-diluted combustion system that utilizes oxygen as an oxidizer, in place of air, results in higher 1st and 2nd law efficiencies.</p><p>Mathematical models for heat regenerators were developed to be designing tools for maximized heat recovery. These models were verified by heat performance experiments carried out on various heat regenerators.</p><p>Furthermore, experiments were performed in a semi-industrial test furnace. It was equipped with various regenerative burning systems to establish combustion and heat transfer conditions prevailing in an industrial furnace operating based on HiTAC. The tests were carried out at seven firing configurations, two conventional and five HiTAC configurations, for direct and indirect heating systems.</p><p>Measurements of energy balance were performed on the test furnace at various configurations in order to obtain the 1st law efficiency. Moreover, local measurements of temperature, gas composition, and heat fluxes in the semi-industrial test furnace were performed to find out the main characteristics of HiTAC flame and the effects of these characteristics on the heating potential, i.e., useful heating in the furnace. In the case of HiTAC, these measurements showed uniformities of chemistry, temperature, temperature fluctuation, and heat fluxes profiles. The values of fluctuations in temperature were small. The high speed jets of the fuel and air penetrated deep into the furnace. The fuel gradually disappeared while intermediate species gradually appeared in relatively high concentrations and at broader regions inside the furnace. These findings indicate: a large reaction zone, low specific combustion intensity in the flame, low specific fuel energy release, and high heat release from this large flame. In addition to the thermodynamic limitations to the maximum temperature of the Oxygen-diluted Combustion, the low specific energy release of the fuel and the high heat release from the flame to its surroundings cause this uniform and relatively moderate temperature profile in a HiTAC flame, consequently suppressing thermal-NO formation.</p><p>Heat flux and energy balance measurements showed that heating potential is significantly increased in the case of HiTAC compared to that in the conventional case, implying much more energy savings than the apparent heat recovery from the heat regenerators, and consequently much less pollutants emissions. Therefore, it is certain that this large HiTAC flame emits more thermal radiation to its surroundings than the conventional flame does, in spite of the moderate-uniform temperature profile of the flame. This intense heat flux was more uniform in all HiTAC configurations, including the indirect heating configuration, than that of the conventional-air combustion configuration.</p>

Orientador: João Claudio Thoméo / Banca: Roger Darros Barbosa / Banca: Ranulfo Monte Alegre / Resumo: Este trabalho teve como objetivo produzir enzimas celulolíticas por fermentação em estado sólido (FES), em um bioreator de leito fixo, empregando-se o fungo termofílico Myceliophthora sp. e utilizando-se como substratos bagaço de cana-de-açúcar e farelo de trigo. Os testes foram realizados inicialmente em sacos de polipropileno apresentando uma boa produção de CMCase (550U/g) no período de 96 horas de fermentação. As variáveis controladas foram a temperatura de fermentação (40, 45 e 50ºC), umidade inicial (75, 80 e 85% b.u.) e proporção do substrato de bagaço de cana e farelo de trigo (1:1, 3:7 e 1:9 peso). Os estimadores estatísticos não indicaram diferença significativa da produção de celulase com as variações da proporção de bagaço de cana e de farelo de trigo, com a temperatura e com a umidade inicial do substrato, contudo, os melhores resultados foram obtidos para a proporção de 70% de bagaço de cana e 30% de farelo de trigo, temperatura de 50oC e umidade inicial do substrato de 80%. Nos testes em escala de bioreator, foi empregado um leito fixo de 7cm de diâmetro e 50cm de comprimento, encamisado, operado nas condições experimentais de melhor resultados obtidos na escala de sacos, sendo a temperatura da camisa e vazão de ar as variáveis de interesse. O fungo apresentou um bom desenvolvimento no reator, tendo uma produtividade enzimática semelhante à encontrada em escala de saco. Foi observado um gradiente de umidade no interior do leito, sendo este crescente do fundo para o topo do bioreator, comportamento inverso ao da atividade enzimática. As vazões variaram entre 80 e 120 L/h e não afetaram estatisticamente a fermentação, possuindo pouca influência sobre as distribuições longitudinais de umidade e de atividade enzimática. Já os ensaios para as temperaturas de 45 e 50oC mostraram-se distintos, sendo a maior produção de enzimas após 96h... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: This work aimed to produce celullolytic enzymes through solid state fermentation in a fixed bed bioreactor, using the fungus Myceliophtora sp. and wheat bran and sugar cane bagasse as substrate. Initial tests were performed in polypropylene plastic bags, and satisfactory production of CMCase (550U/g.d.s.) was observed at 96h of fermentation. The controlled variables were the temperature (40, 45, and 50o C), the initial moisture content (75, 80, and 85%, w.b.), and wheat bran to sugar cane bagasse proportion (1:1, 3:7, and 1:9, weight). The experimental results did not differ statistically for any experimental condition adopted, even though the best results were obtained using the proportion 3:7 wheat bran/sugar cane bagasse 3:7, 50o C temperature, and 80% initial moisture content. A jacketed fixed bed bioreactor 7.62cm ID and 50cm long was used for scaling-up, and temperature, solid initial moisture content, and air flow rate were used as variables. The fungus adapted well to the bioreactor and to the experimental conditions, and the production of enzymes was similar to the tests performed in the plastic bags. It was observed a longitudinal gradient of moisture, increasing from the bottom to the top of the reactor, opposite effect observed for the enzyme production. The air flow rate was varied from 80 to 120L/h and did not affect neither the moisture nor the enzyme distributions. However, the experiments carried out at 45 and 50o C were statistically distinct, and the enzyme production at 50o C at 96h was larger than at 45o C. At 144h of fermentation, the results at 45o C were similar to those observed at 50o C and 96h. It was observed temperature peaks, the higher one at 18h of fermentation, which was 6o C above the temperature fixed through the jacket and the air at the entrance, for 45o C temperature and 80L/h air flow rate / Mestre

Submitted by Johnny Rodrigues (johnnyrodrigues@ufcg.edu.br) on 2018-10-18T15:04:53Z No. of bitstreams: 1 RODRIGO SÉRGIO FERREIRA DE MOURA -DISSERTAÇÃO PPGEA 2001..pdf: 7960772 bytes, checksum: 6c554557385690a7c05980568ee6742b (MD5) / Made available in DSpace on 2018-10-18T15:04:53Z (GMT). No. of bitstreams: 1 RODRIGO SÉRGIO FERREIRA DE MOURA -DISSERTAÇÃO PPGEA 2001..pdf: 7960772 bytes, checksum: 6c554557385690a7c05980568ee6742b (MD5) Previous issue date: 2001-03-16 / Capes / En este trabajo se ha estudiado las isotermas de desorción dei cajuil (Anacardium occidentale L.) vahedad roja, y los efectos de la temperatura y velocidad dei aire en la cinética dei secado de lo mismo, con vistas a la obtención de un modelo empírico que simbolizase el proceso, mediante un "planeamiento experimental". La humedad de equilíbrio fue obtenida con auxilio dei aparato Termoconstanter Novasina TH 200, que emplea el método estático; lo ajuste de los datos se hizo por regresión no linear, pelos modelos de BET, GAB, Halsey e Oswin, eligiéndose lo mejor mediante coeficiente de determinación (R2) y el modulo de la desviación relativa (P); la cinética dei secado, en camada fina, fue obtenida en uno secador de lecho fijo, para un rango de temperaturas variando de 35 a 65° C, y velocidades dei aire de secado de 0,9 a 1,9 m.s"1, de acuerdo con el planeamiento factorial completo (22 + configuración estrella). Adentren los modelos ajustados para las isotermas estudiadas, GAB fue el mejor, habido escogido para presentar las isotermas de desorción dei cajuil; se observo que los parâmetros Xm e C de esto modelo, disminuyeron con el aumento de la temperatura; se constato por médio de las curvas de secado que el tiempo gastado para secar el cajuil en las temperaturas de 35 y 40° C es el doble dei tiempo requerido a 65° C, y que la velocidad dei aire de secado no tiene influencia en el proceso; en la formulación de los modelos empíricos, el modelo de 2a orden presente un valor mayor de R2 y la relación entre F calculado e F tabulado vario de 6,04 a 19,86, aunque para el modelo linear esta relación fue de 0,27 a 5,40. Así, se pudo concluir que la cinética de secado es fuertemente afectada pela temperatura; el modelo de 2a orden es util para fines predectivos de la cinética de secado de esto producto; y, que mejor rango de operación se dio con la temperatura mas elevada. / Este trabalho teve como objetivos determinar as isotermas de dessorção do caju {Anacardium occidentale L.), "variedade" vermelha, e os efeitos da temperatura e da velocidade do ar na cinética de secagem deste produto, visando a obtenção de um modelo empírico que represente esse processo, mediante o emprego de um planejamento experimental. A umidade de equilíbrio foi obtida com o auxílio do equipamento Termoconstanter Novasina TH 200, utilizando-se o método estático-indireto; o ajuste dos dados foi feito através de regressão não linear, pelos modelos de BET, GAB, Halsey e Oswin, escolhendo-se o melhor mediante o coeficiente de determinação (R2) e o módulo do desvio médio relativo (P); a cinética de secagem, em camada fina, foi obtida em um secador de leito fixo, nas temperaturas variando de 35 a 65° C e velocidades do ar de secagem de 0,9 a 1,9 m.s"\ conforme o planejamento fatorial completo (22 + configuração estrela). Dentre os modelos ajustados às isotermas estudadas, GAB apresentou melhor ajuste dos dados experimentais, tendo sido escolhido para representar as isotermas de dessorção do caju; observou-se que os parâmetros Xm e C desse modelo, decresceram com o aumento da temperatura; constatou-se mediante as curvas de secagem que o tempo gasto para secar o caju nas temperaturas de 35 e 40° C é o dobro do tempo requerido a 65° C, e que a velocidade do ar de secagem não exerce influência no processo; na construção dos modelos empíricos, o modelo de 2â ordem apresentou um maior valor de R2 e a relação entre F calculado e F tabelado variou de 6,04 a 19,83, enquanto que no modelo linear esta relação foi de 0,27 a 5,40. Assim, pode-se concluir que a cinética de secagem é fortemente afetada pela temperatura; o modelo de 2a ordem é útil para fins preditivos da cinética de secagem desse produto; e, que a melhor faixa de operação é aquela correspondente à temperatura mais elevada.

Pour des raisons de coûts, les tests de catalyseur mis en forme (billes, extrudés, …) en réacteur est lit fixe sont mis en œuvre dans des réacteurs de plus en plus petits alors que la taille des objets catalytiques ne change pas. L'objet de cette thèse est d'explorer les domaines où la réduction d'échelle conduit à des questionnements nouveaux en termes de physique, répétabilité, représentativité et modélisation. Le document s'articule autour de 5 chapitres : 1) une introduction détaillée du contexte avec un état de l'art sur les réacteurs à lit fixe de petite taille et l'identification des questions en suspens, 2) une réflexion autour de la nature des écoulements gaz-liquide descendant dans des lits fixes quand les forces capillaires ne sont plus négligeables, 3) la présentation d'une méthodologie et d'un critère sur un nombre minimal de grain nécessaire pour limiter les conséquences d'une distribution granulométrique en présence de limitations au transfert interne, 4) l'étude des effets d'empilements aléatoires sur la performance apparente du réacteur, d'abord par un modèle réseau de pores qui décrit assez bien les tendances observées expérimentalement mais manque de capacité de prédiction, puis par simulation numérique directe d'écoulement réactif dans des empilements de 8 cylindres qui montre que les effets d'empilements apparaissent, avec les limitations au transfert externe, quand la diffusion moléculaire transverse n'est pas assez rapide pour compenser les gradients dus à la réaction et à la convection, 5) une conclusion qui résume les critères de conception de réacteurs de tests à petite échelle, et propose des pistes de continuation du travail / In order to lower costs, testing of catalytic pellets (spheres or extrudates) is performed in ever smaller fixed bed reactors whereas catalytic pellet size is unchanged. The object of this thesis is to explore domains where downsizing leads to new questions in terms of physics, repeatability and modeling. The thesis is built in 5 chapters, 1) a detailed introduction of the context and a review of the literature on small fixed bed reactors, 2) a discussion about flow patterns in fixed beds filled with fine powder where capilary effects are not negligible, 3) a presentation of a methodology to assess the impact of sampling small number of pellets out of a non uniform set and results for internally mass transfer limited reactions, 4) a study of the effect of randomness in fixed beds first through a pore network model with a good ability to predict trends but lacking accuracy, then using direct numerical simulation of a reactive flow in fixed beds made of 8 cylinders arranged in several configuration showing that packing effects occur, concurrently with external mass transfer limitations, when cross flow diffusion is not fast enough to level out convection and reaction induced gradients, 5) a conclusion that summarizes design criteria and offers a few perspectives for R&D in downsizing

Cette thèse s'inscrit dans le domaine de la catalyse hétérogène pour des applications en raffinage et enpétrochimie. Comme objet d'étude, nous avons ciblé les catalyseurs d'hydrotraitement qui permettentd'éliminer les impuretés contenues dans le pétrole comme le soufre. Ces catalyseurs d'hydrotraitementsont constitués de sulfures de molybdène supportés sur alumine et généralement promus par du nickelou du cobalt.L'enjeu de la thèse porte sur la compréhension des phénomènes mis en jeu au cours de la sulfuration etde la stabilisation (cokage, passivation, évolution de la phase sulfure) des catalyseurs pour la réactiond'hydrodésulfuration des gazoles visant à réduire sa teneur en soufre. Néanmoins, à l'heure actuelleaucune technique ne permet de caractériser le catalyseur lors de la sulfuration en phase liquide (avecou sans DMDS) et du test dans les conditions de température et de pression industrielles.La présente thèse vise le développement d'une « caractérisation operando » de ce type de réaction ens'appuyant sur la spectroscopie Raman, une des seules technique de laboratoire permettant d'effectuerdes analyses avec les contraintes précédemment citées. Une unité, composée d'un réacteur cylindriquetransparent, a donc été conçue et mise au point pour suivre par spectroscopie Raman les phases solideet liquide au cours de la sulfuration et de la réaction d'HDS. Parallèlement, une méthodologie decaractérisation et de focalisation a été développée. Ce montage et cette méthodologie ont permisd'accéder pour la première fois à des profils spatiaux et temporels sur la phase solide dans lesconditions d'activation et de réaction d'hydrotraitement des gazoles en condition d'hydrodésulfuration(30 bar, 350 °C).Malgré un signal de fluorescence probablement liée à la décomposition radicalaire du précurseur desoufre (DMDS) entre 200 et 260°C, il a été possible de suivre au cours du temps à des positions fixesla disparition de la phase oxyde, l'évolution de la phase sulfure et du coke . Ces résultats ontnotamment permis d'étudier l'impact de la charge sur la cinétique de sulfuration .Le suivi spatial, lelong du réacteur par exemple, s'avère plus délicat et nécessitera de développer des méthodes pourcompenser les variations d'intensité du signal Raman induites par le positionnement aléatoire desgrains ainsi que l'écoulement. Concernant la phase liquide, une approche multivariée utilisant desoutils chimiométriques a été appliquée afin de relier l'émission de fluorescence intrinsèque à denombreux diesels à certaines de leurs propriétés (teneur en soufre et en aromatiques, densité…). Lesmodèles ont été développés à partir de spectres acquis à température ambiante et à pressionatmosphérique, mais leurs performances satisfaisantes encouragent à étendre l'approche auxconditions réactionnelles d'HDS qui reste une perspective de ce travail / This thesis is in the field of heterogeneous catalysis for the applications in refining and petrochemistry. As an aim of this study, we have focused on the hydrotreatment catalysts that are applied to remove some of the impurities from crude oils, like sulfur. Such hydrotreatment catalysts consist of alumina supported molybdenum sulfides, being generally promoted by nickel or cobalt.The aim of this work have consisted on the understanding of the phenomena occurring during the stabilization phase (coking, passivation, evolution of the sulfide phase) during the catalysts sulfidation and under the reaction of hydrodesulfurization, HDS, aiming to reduce the diesels sulfur content. A better understanding of these phenomena would ease the development of new generations of more efficient catalysts. This thesis aims at extending the operando characterization methods to allow a spatial and temporal follow-up of liquid and catalyst during this type of reactions. A unit was designed and built to follow the solid and liquid phases during the catalysts sulfidation and under HDS reaction. The operando follow-up was done using Raman spectroscopy through a cylindrical transparent reactor. In parallel, a methodology has been developed to focus and acquire good quality spectra through the reactor..With these reactor and methodology , we were able to access for the first time to time-space resolved profiles of the solid phase during the sulfidation and the diesel hydrodesulfurization (under 350 °C and 30 bar). Temporal profiles concerns the oxide phase disappearance, sulfide phase growth and coke formation. Spatial profiles are more challenging and will require a methodology more robust to signal changes induced by the random position of pellets and flow pattern. With respect to the liquid phase, a multivariate approach based on chemometrics has gave properties of diesels at room temperature and atmospheric pressure. The good results are encouraging enough to propose to extend the approach to HDS conditions that constitutes one of the perspectives of the work

L’acylation de Friedel-Crafts est parmi les réactions les plus fondamentales et les plus utiles pour former des cétones aromatiques. Bien que la mise en oeuvre de ces réactions soit habituellement pénalisante en termes de sous produits formés, de rendement atomique et de quantité de catalyseur utilisée, cette voie est encore largement utilisée dans l’industrie de la chimie fine. Ces dernières années, la prise de conscience de l’impact des activités industrielles sur l’environnement a mené les chimistes à travailler sur de nouvelles voies, moins dangereuses et plus respectueuses de l’environnement.Nous avons considéré ici l’acylation d’un dérivé du benzofurane par un chlorure d’acyle, correspondant à une étape intermédiaire pour la fabrication d’un principe actif pharmaceutique. Dans cette étude, la première proposition a consisté à remplacer les catalyseurs classiques (FeCl3 ou AlCl3) par des catalyseurs solides réutilisables. Dans ce travail, plusieurs catalyseurs (zéolites, résines échangeuses d’ions...) ont d’abord été testés dans des conditions « conventionnelles » , c’est-à-dire en utilisant un solvant organique, afin de déterminer le plus performant, en termes de réactivité, de durée de vie et de réutilisabilité. Une zéolite Y s’est avéré la plus adaptée. Cependant, l’utilisation d’un solvant organique (ici le 1,2-dichlorobenzène) reste discutable et l’utilisation du CO2 supercritique comme solvant a donc été envisagée. Pour cela, nous avons imaginé et dimensionné un réacteur tubulaire à lit fixe de catalyseur fonctionnant sous pression et en continu. Les résultats présentés montrent la faisabilité d’un tel procédé et pose les jalons pour arriver à une ou des solutions pour une meilleure mise en oeuvre industrielle des réactions de Friedel-Crafts. / Friedel-Crafts acylation is among the most fundamental and useful reactions to yield aromatic ketones, but it is one of the less acceptable in terms of unwanted polluting by-products or atom economy because of overconsumption of catalyst which is used in stoichiometric quantities in the conventional process. This route is nevertheless still widely used in the fine chemicals industry. In recent years, awareness of the impact of industrial activities on the environment has lead chemists to work on new chemical routes, less dangerous and more environmentally friendly. We considered here the acylation of a benzofurane derivative by an acid chloride, as an intermediary step for a pharmaceutical product. In this study, one of the first alternative was to replace conventional catalysts (FeCl3 or AlCl3), by reusable solid catalysts. In this work, different catalysts (zeolites, ion-exchange resins...) were first tested in "conventional" conditions, i.e., using an organic solvent (1,2-dichlorobenzene in our case), to determine the best one, in terms of reactivity, lifetime and reusability. The zeolite Y was found the most appropriate. However, the use of an organic solvent still remains questionable and the use of supercritical carbon dioxide as the solvent was also considered. Its inherent properties include non-flammability, mild critical conditions, tuneable solubility near to the critical point and very low environmental impact. The reaction was operated using a specifically designed continuous high pressure fixed bed and results concerning yield and selectivity are presented. These results demonstrate the feasibility of such an approach, which would ultimately yield to better industrial operation of Friedel-Crafts reactions.

Une étape clé du développement de nouveaux catalyseurs est l'évaluation de leurs performances. Les tests de catalyseurs sont généralement opérés dans des unités pilotes de "petite" taille dont les avantages sont une consommation moindre de réactif et un coût d'opération plus réduit. Les réacteurs d'unité pilote ont des diamètres de l'ordre de 3 à 10 fois la taille des catalyseurs qui sont généralement des sphères ou des cylindres dont la dimension typique est 2 à 3 mm. Des données expérimentales ont mis en évidence des difficultés de répétabilité des résultats sur certains réacteurs très courts. L’objectif de ce travail est de comprendre le lien entre la micro-structure de l’empilement, qui est aléatoire et localement hétérogène, l'hydrodynamique locale et la réactivité dans des réacteurs à lit fixe de petite taille, en vue de définir des critères de conception et des méthodologies de chargement permettant d'obtenir des performances chimiques répétables. La démarche retenue est purement numérique basée sur deux codes développés à IFPEN : Grains3D pour générer numériquement l'empilement et PeliGRIFF pour simuler l'écoulement réactif dans le lit fixe. Une analyse de l'état de l'art sur le sujet donne des informations abondantes sur les lits de particules sphériques, et très peu sur les cylindres, en particulier sur les couplages écoulements et transfert/réaction. En effet, ces travaux ne permettent pas de lier les performances réactives à la structure locale hétérogène de ces empilements. Dans ce travail, la caractérisation globale et locale d'empilements de sphères et de cylindres a permis de confirmer et compléter la littérature sur certains points : porosité au centre décrite de façon approximative par les corrélations, porosité axiale, orientations des cylindres... Des variations sont présentes à l’intérieur des lits de petite taille, variations qui ne s’atténuent pas avec l’augmentation du volume d’étude et qui sont équivalentes d’un chargement à l’autre. L'étude de l'hydrodynamique dans des lits de sphères et de cylindres mono- et polydisperses a également permis de quantifier l'effet de la répétition du chargement sur la perte de pression et les champs de vitesse. Les premiers résultats d'écoulement réactif ont montré que, qualitativement, le champ de concentration dans la particule réactive est sensible à l’écoulement en présence d'une limitation interne au transfert de matière. Ces résultats préliminaires ont permis de définir une méthodologie de travail qui pourra servir pour continuer l'étude. / The evaluation of catalyst performances is needed for their development. Catalytic tests are generally operated in pilot units of "small" size. Their advantages are : the reduction of reactant consumption and the reduction of operating costs. The pilot plant reactors have diameters of about 3 to 10 times the size of the catalyst (mainly spheres and cylinders) and the results obtained should be representative of the same performance of the catalysts in industrial units whose characteristic scale is around 5 meters. Experimental data have shown unacceptable repeatability in some small size reactors. The objective of this work is to understand the link between the local micro-structure of the packed bed, which is random and locally heterogeneous, local hydrodynamics and chemical responses in fixed bed reactors of small size. This study aims to define design criteria, and packing methodologies in order to obtain repeatable results. The approach adopted for this work is essentially based on numerical simulations. Two IFPEN codes are used for this study: Grains3D which is used to generate numerically the packed bed, and PeliGRIFF to simulate the reactive flow within the reactor. Previous works give information on packing, hydrodynamics and reactive flow characterization, mostly on spheres, much less is available on cylinders. However, they do not link local structure to reactive performance of the bed. In our work, geometrical characterization of the randomly packed beds of spheres and cylinders confirm and complete literature on several points: porosity at the centre of the reactor for which the correlations are inaccurate, the effect of bed height on the axial bed porosity, cylinder orientations... Some variations are present within small size beds, variations that are not dampened with an increase of averaging volume, and are similar from one packing to another. Hydrodynamics study carried out in packed beds of spheres and cylinders of different sizes allowed to link the packing local structure to pressure drop and velocity field. Only preliminary simulations have been performed on reactive flow. First qualitative results show an evolution of the concentration field inside pellets with flow in case of internal mass transfer limitations. Continuing these preliminary results will yield to the definition of a methodology that can be used to link local structure to reactive performance in randomly fixed packed bed reactors of small size.