A formação de espinélio de aluminato de magnésio (\'MG\'\'AL IND.2\'\'O IND.4\'; espinélio), através da combinação de óxido de alumínio (\'AL IND.2\'\'O IND.3\'; alumina) com geradores de poros à base de óxido de Magnésio (\'MG\'\'O\'; magnésia) ou hidróxido de magnésio (\'MG\'(\'OH)IND.2\'; HM), resulta em estruturas de elevada porosidade (acima de 50%) e com baixa tendência a densificar mesmo em temperaturas elevadas (1500-1650°C). Devido a isso, esse sistema tem grande potencial de aplicação tecnológica para isolamento térmico e filtração em altas temperaturas. Considerando sua utilização em larga escala como isolante térmico verifica-se a necessidade de desenvolvimento de novas rotas de síntese mais eficazes. A reação de hidroxilação do \'MG\'\'O\' pode ser controlada, utilizando como matéria-prima tanto a magnésia sínter como com a magnésia caustica, desta forma, a expansão de peças de cerâmicas porosas feitas a base de \'MG\'\'O\' também podem ser controladas. Deste modo, neste trabalho pretende-se investigar uma nova rota para a obtenção de espinélio utilizando o \'MG\'\'O\' como ligante hidráulico e incorporar poros, in situ, utilizando-se da decomposição do HM em uma matriz de alumina. Os resultados mostraram que a área superficial específica (ASE) da magnésia influencia fortemente nas propriedades mecânicas do material moldado, comprovando que para as ASEs maiores a magnésia é viável como ligante hidráulico, sendo 0,6 MPa para o modulo de ruptura por compressão diametral, o melhor resultado deste trabalho, enquanto o sistema à base de cimento de aluminato de cálcio (CAC) teve 0,5 MPa de módulo de ruptura, e o sistema à base de \'alfa\'-Bond teve 0,3 MPa de módulo, ambos sistemas de ligantes hidráulicos conhecidos na literatura. A grande diferença no módulo de ruptura por compressão diametral mostrou que o tempo de cura e a ASE da magnésia influenciaram-no fortemente.Enquanto, as amostras com alta ASE de magnésia, em torno de ~60 \'M POT.2\'/g, tiveram 0,6 MPa de módulo de ruptura, as amostras com baixa ASE, ~1 \'M POT.2\'/g, não tiveram alteração no módulo de ruptura. Verificou-se também um aumento gradual no módulo de ruptura segundo a ASE, para um valor intermediário de ASE, ~30 \'M POT.2\'/g, o módulo foi de 0,4 MPa, valor ainda comparável a outros sistemas que se utilizam ligantes hidráulicos. Observou-se que a expansão das amostras durante a cura foi influenciada pela sua ASE, quanto maior a ASE, maior a expansão. Os resultados de PTG e do módulo de ruptura combinados com as imagens de MEV e os difratogramas, das amostras calcinadas, mostraram a influência da temperatura e da ASE na formação da cerâmica porosa por essa rota, sendo a temperatura uma variável de controle já conhecida, observada no diagrama de equilíbrio de fases. Entretanto, a influência da ASE do \'MG\'\'O\' para o controle da formação do espinélio é desconhecido da literatura / Magnesium aluminate spinel (\'MG\'\'AL IND.2\'\'O IND.4\'; spinel) formation by the combination of aluminum oxide (\'AL IND.2\'\'O IND.3\'; alumina) with magnesium oxide (\'MG\'\'O\', magnesium) or magnesium hydroxide (\'MG\'(\'OH) IND.2\'; HM), results in structures of high porosity (above 50%) and with low tendency to densify even at elevated temperatures (1500-1650°C). Due to this, this system has great potential of technological application for thermal insulation and filtration in high temperatures. Considering its large scale use as thermal insulation, it is necessary to develop new and more efficient routes of synthesis. The hydroxylation reaction of \'MG\'\'O\' can be controlled using both sinter magnesia and caustic magnesia as a feedstock, so the expansion of porous ceramic pieces made with \'MG\'\'O\'can also be controlled. Thus, in this work we intend to investigate a new route to obtain spinel using \'MG\'\'O\' as a hydraulic binder and to incorporate pores, in situ, using the decomposition of HM in an alumina matrix. The results showed that the specific surface area (ASE) of the magnesia strongly influences the mechanical properties of the molded material, proving that for the larger ASEs the magnesia is viable as a hydraulic binder, being 0.6 MPa for the diametral compression rupture modulus, the best result of this work, while the calcium aluminate cement (CAC) system had 0.5 MPa of modulus of rupture, and the \'alfa\'-Bond based system had 0.3 MPa of modulus, both systems of hydraulic binders known in the literature. The large difference in the diametral compression rupture modulus showed that the curing time and the ASE of the magnesia strongly influenced it. While high ASE magnesia samples, around ~60 \'M POT.2\'/g, had 0.6 MPa of modulus of rupture, samples with low ASE, ~1 \'M POT.2\'/g, had no change in modulus of rupture. There was also a gradual increase in the ASE burst modulus, for an ASE intermediate value ~30 \'M POT.2\'/g, the modulus was 0.4 MPa,a value still comparable to other systems using hydraulic binders. It was observed that the expansion of the samples during curing was influenced by their ASE, the higher the ASE, the greater the expansion. The results of PTG and the rupture modulus combined with SEM images and the diffractograms of the calcined samples showed the influence of temperature and ASE on the formation of the porous ceramic by this route, the temperature being a known control variable, observed in the phase equilibrium diagram. However, the influence of \'Mg\'\'O\' ASE on the control of spinel formation is unknown in the literature

Le dragage des ports français génère chaque année entre 25 et 40 millions de tonnes de matériaux non contaminés qui sont clapés en mer. Il existe, de plus, un stock important de sédiments contaminés (10 millions de tonnes environ) qui n’ont pas été dragués, du fait de la réglementation interdisant un rejet en mer lorsque les niveaux de contamination dépassent les seuils définis par l’arrêté du 14 juin 2000. En outre, ce stock de sédiment, qui devra impérativement être dragué dans les dix ans à venir, est quasiment orphelin de filières de traitement et de valorisation adaptées. L’urgence de mettre en place des solutions (filières), respectueuses des fondements du développement durable, a engendré de nombreux programmes de recherche. Le programme SEDiGEST (Gestion des Sédiments de dragages des ports) envisage notamment un scénario de remblaiement des cavités terrestres par des sédiments traités. Cette Thèse de Doctorat, menée dans le cadre de ce programme, contribue à l’amélioration de la compréhension du comportement géo-physico-chimique d’un sédiment marin stabilisé aux liants hydrauliques (chaux + ciment). Pour répondre à cet objectif la démarche expérimentale a été conduite en trois étapes : tout d’abord la caractérisation du solide, puis l’évaluation de la mobilité potentielle des polluants inorganiques et enfin, la modélisation du comportement à la lixiviation. La synthèse des résultats a montré que les polluants inorganiques cibles de la matrice d’étude (cuivre, plomb et zinc) étaient majoritairement associés aux carbonates, aux sulfures/sulfates, aux (oxy)hydroxydes, et/ou à la matière organique. L’étude comparative de la matrice étudiée à trois stades « d’évolution » (avant et après traitement et vieilli artificiellement) a permis de mettre en évidence que le procédé de stabilisation aux liants hydrauliques n’était pas une solution pérenne. En effet, le sédiment traité présente un risque de pollution à plus ou moins long terme, notamment par relargage de plomb et de cuivre, en cas de carbonatation de la matrice sédimentaire. En particulier, la réversibilité du procédé de stabilisation/solidification induit par la réactivité des phases cimentaires avec le CO2 atmosphérique, peut également conduire à une mobilisation des polluants à long terme. / The dredging of French ports generates each year between 25 to 40 million tons of uncontaminated materials which are piled onto sea. Moreover, an important stock of contaminated sediments (about 10 million tons), which has to been dredged, exists. Indeed, the decree of June 14th, 2000 prohibits disposal to sea when levels of contamination exceed regulations. This important stock of sediment, which must be imperatively dredged within 10 years, is almost orphan of adapted treatment and valorization. In this context many research program appear to propose and to develop solutions. The SEDiGEST program (Management of the dredging sediments of ports) intends a scenario of filling terrestrial cavities with treated sediments. This Ph.D. thesis, carried out within the framework of this program, contributes to the improvement in understanding the geo-physico-chemical behavior of marine sediment stabilized with hydraulic binders (cement + lime). In this objective, the experimental approach was conducted in three steps. Firstly, a characterization of the solid was realized. An evaluation of the potential mobility of inorganic pollutants have was then determined. Finally, a modeling of the behavior towards leaching was operated. Data shown that target's inorganic pollutants of the matrix of study (copper, lead and zinc) were mainly associated with carbonates, sulfides/sulfates, (oxy)hydroxides, and/or with organic matter. Comparative study of the matrix at three levels “of evolution” (before and after treatment and artificially aged) highlights the reversibility of the stabilization by hydraulic binders. Indeed, treated sediment presents a pollution risk to less or more long term, leaching of copper and lead, during carbonation.