Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / O estudo da estabilidade de emulsões de biocombustíveis destaca-se, dentre outros aspectos, pela necessidade de identificação das características destas. Esta necessidade reforça-se pela dificuldade de detecção de biodiesel e suas misturas quando estão fora de especificação, além de se aplicar aos processos de separação e purificação do biodiesel após a transesterificação. O entendimento deste problema poderá ser alcançado a partir de estudos de estabilidade associados aos modelos termodinâmicos de equilíbrio líquido-líquido (modelo NRTL) em sistemas ternários, que serão a base da modelagem proposta neste trabalho. O levantamento de dados envolve técnicas de observação direta em titulação, e ensaios de estabilidade e tensiometria, com algumas alterações das técnicas clássicas. Desta forma, o objetivo principal do trabalho é propor uma maneira de identificar e caracterizar as interfaces das emulsões formadas e estudar a estabilidade das fases envolvidas. Além disso, estuda-se a introdução de alcoóis como agentes tensoativos para viabilizar a utilização de misturas de combustíveis, mesmo na região de mais de uma fase. Para o desenvolvimento do trabalho, emulsões envolvendo os constituintes básicos serão combinadas entre si em diferentes proporções. A etapa inicial constitui-se de um mapeamento das misturas e proporções que formam as emulsões, baseadas nas curvas de equilíbrio. Em seguida, faz-se o estudo de caracterização e estabilidade destes sistemas. Com estas informações é possível mapear o sentido de incremento de tensão na interface e instabilidade da separação de fases das emulsões em relação aos diagramas ternários correspondentes. Por fim, propõe-se uma estratégia para estimar os parâmetros termodinâmicos do NRTL para as espécies presentes nos sistemas estudados e prever o comportamento dos mesmos e de sistemas similares. Este modelo matemático proposto tem como objetivo complementar a técnica experimental adotada para construção das curvas de equilíbrio, sendo bem sucedido / The study of the stability of biofuels emulsions calls attention, among other things, to the need to identify their characteristics. This need is reinforced by the difficulty of detecting biodiesel and its blends when they are out of specification, and apply the processes of biodiesel separation and purification after transesterification. The understanding of this issue could be reach from stability studies associated with thermodynamic models of liquid-liquid equilibrium (NRTL) in ternary systems, which will be the basis for modeling proposed in this paper. The research involves direct observation techniques for titration, and stability tests and tensiometry, with some changes of classical techniques. Thus, the main objective is to propose a way to identify and characterize the interfaces of emulsions formed and study the stability of the phases involved. Furthermore, it studies the introduction of alcohols as surfactants to enable the use of mixtures of fuel, even in the region of more than one phase. For development work, emulsions involving basic constituents will be combined in different proportions. The first step consists of mapping the mixtures and proportions which form emulsions, based on the equilibrium curves. Then, it is done the study of characterization and stability of these systems. With this information it is possible to map the direction of an increase in the interface tension and phase separation instability of emulsions in relation to the corresponding ternary diagrams. Finally, it proposes a strategy to estimate the thermodynamic parameters of the NRTL for the species present in the systems studied and predict the behavior of these and similar systems. This mathematical model aims to complement the experimental technique adopted for construction of the equilibrium curves, being successful

Uma das dificuldades relacionadas com a exploração de petróleo é o transporte de óleo pesado, que devido a sua alta viscosidade, acarreta em uma elevada perda de carga no sistema. Para proporcionar economia de energia aplica-se o método do Core Annular Flow (CAF) onde é utilizado um escoamento anular bifásico em que a água escoa na periferia da tubulação para redução do gasto energético. O presente trabalho visou compreender e reproduzir este fenômeno, desenvolvendo simulações em CFD através do pacote comercial ANSYS FLUENT considerando o escoamento 3D, turbulento, isotérmico e incompressível para casos estacionários e transientes. A interface entre a água e o óleo foi adequadamente reproduzida em diferentes geometrias como tubo reto e com curva. O método LES para simulação de grandes escalas provou ser o melhor método de turbulência dentre os testados, como k-epsilon e modelo de tensores de Reynolds, de forma com que a interface fosse representada corretamente. O modelo para sistema multifásico adotado foi o Volume de Fluido (VOF), comparado com o comportamento experimental e com dados da literatura. Os fenômenos de swirl observados experimentalmente também foram reproduzidos de forma satisfatória. / One of the difficulties related to oil exploration is the transportation of heavy oil, which due to its high viscosity, causes a high pressure drop in the system. In order to provide energy savings, the Core Annular Flow (CAF) method applies where a two-phase annular flow occurs in which water flows at the periphery adjacent to the pipe to reduce energy expenditure. The present work aimed to understand and reproduce this phenomenon, developing CFD simulations through the commercial package ANSYS FLUENT considering flow as 3D, turbulent, isothermal and incompressible for stationary and transient cases. The interface between water and oil has been properly reproduced in different geometries such as straight pipe and pipe with a curve. The LES method for large scale simulation proved to be the best turbulence method among the tested, such as k-epsilon and Reynolds stress model, so that the interface was correctly represented. The model for the multiphase system adopted was the Volume of Fluid (VOF), compared to the experimental behavior and with data from the literature. The experimentally observed swirl phenomena were also reproduced satisfactorily.

A presente pesquisa teve como objetivo desenvolver e validar um método analítico utilizando microextração líquido-líquido (LLME) e cromatografia líquida acoplada ao detector por arranjo de diodos (HPLC-DAD) para análise do metil, etil, propil e butilparabeno em amostras de esgoto sanitário provenientes da Estação de Tratamento de Esgoto de São Carlos (ETE Monjolinho) antes e após o tratamento, bem como em amostras de águas superficiais dos córregos Santa Maria Madalena, Tijuco Preto, Gregório e Monjolinho. Por meio de planejamentos experimentais pode-se desenvolver o método analítico, e inicialmente, otimizou-se as condições cromatográficas, selecionando-se a coluna Zorbax SB-C8 (250 x 4,6 mm, 5 µm), 20 µL de volume de injeção, comprimento de onda de 257 nm, vazão da fase móvel de 1 mLmin-1 e temperatura da coluna cromatográfica de 30°C. A fase móvel com modo de eluição isocrática foi composta por metanol e solução de ácido acético 1% (v/v). No método de extração LLME, o procedimento otimizado utilizou 5 mL de amostra de esgoto, 10 mL de acetato de etila como solvente extrator e 0,4 g de cloreto de sódio (NaCl). Os ensaios de validação demonstraram boa adequação do método desenvolvido, com limites de quantificação de 10 µgL-1 e coeficientes de correlação superiores a 0,99. A recuperação deu-se entre 71 e 99% e a precisão entre 0,4 e 7,4 %, valores estes que estão dentro da faixa de aceite estabelecida pela Agência Nacional de Vigilância Sanitária (ANVISA). Foram realizadas duas coletas em estações diferentes do ano e em ambas foi possível encontrar e quantificar os parabenos. Na primeira coleta, feita em período de estiagem, a concentração encontrada dos parabenos foi maior do que aquela encontrada na segunda coleta. As maiores concentrações encontradas na primeira coleta foram: 0,98 µgL-1 de metilparabeno no esgoto bruto, 9,7 µgL-1 de etilparabeno na amostra de água do rio Monjolinho, 7,9 µgL-1 de propilparabeno na amostra do córrego do Gregório e 11 µgL-1 de butilparabeno no esgoto bruto. Já na segunda coleta, o metilparabeno pode ser quantificado apenas na amostra de esgoto bruto: 0,03 µgL-1 e a maior concentração encontrada de etilparabeno foi de 2,1 µgL-1 no rio do Monjolinho. O butilparabeno foi detectado em seis dos sete pontos amostrados, sendo a maior concentração encontrada no rio do Monjolinho (6,99 µgL-1). O propilparabeno não foi encontrado em nenhuma das amostras da segunda coleta. / The present research had the objective of developing and validating an analytical method using liquid-liquid microextraction (LLME) and liquid chromatography coupled to the detector by diode array (HPLC-DAD) for the analysis of methyl, ethyl, propyl, and butylparaben in wastewater samples from the São Carlos wastewater treatment plant (ETE Monjolinho) before and after treatment; it was also accomplished in surface water samples from Santa Maria Madalena, Tijuco Preto, Gregório, and Monjolinho brooks. By means of experimental planning, the analytical method could be developed, and, initially, the chromatographic conditions were optimized by selecting the Zorbax SB-C8 column (250 x 4,6 mm, 5 μm), injection volume of 20 μL, 257 nm wavelength, mobile phase flow rate of 1 mLmin-1, and chromatographic column temperature at 30ºC. The mobile phase with isocratic elution mode was composed of methanol and 1% (v/v) acetic acid solution. In the LLME extraction method, the optimized procedure used 5 mL of wastewater, 10 mL of ethyl acetate as the extracting solvente, and 0,4 g of sodium chloride (NaCl). The validation tests showed a good fit of the developed method, with quantification limits of 10 μgL-1 and correlation coefficients higher than 0,99. The recovery was between 71% and 99%. Furthemore, the precision was between 0,4% and 7,4%. Those values are within the acceptable range established by the National Agency of Sanitary Administration (ANVISA). Two samplings were carried out in different seasons of the year and, in both, it was possible to find and to quantify the parabens. In the first sampling, during the dry season, the concentration of parabens was higher than that found in the second sampling. The highest concentrations found in the first sampling were: 0,98 μgL-1 of methylparaben in the raw wastewater, 9,7 μgL-1 of ethylparaben in the water sample of Monjolinho river, 7,9 μgL-1 of propylparaben in the water sample of Gregório brook, and 11 μgL-1 in the raw wastewater. In the second sampling, methylparaben could be quantified only in the raw wastewater: 0,03 µgL-1. The highest concentrations of ethylparaben was 2,1 µgL-1 in the Monjolinho river. Butylparaben could be detected in six of seven points, and the highest concentration was in the water sample of Monjolinho river. Propylparaben could not be found in any of the samples.

O presente estudo teve como objetivo desenvolver e validar um método analítico para a investigação de parabenos em tecido de peixe. Para o preparo de amostra empregou-se como método a microextração líquido-líquido (LLME) seguido de análise cromatográfica líquida de alta eficiência acoplada a um detector de arranjo de diodos. Empregou-se coluna Nucleodur C8 (4,6 x 250 mm, 5 µm), com injeção de 20µL, fluxo de 1mL min-1, temperatura de 25 °C, comprimento de onda 258nm e fase móvel composta por água acidificada com 1% de ácido acético e metanol em modo gradiente na proporção de 60% de metanol até 8 minutos e de 65% até 20 minutos. A recuperação do método para o metilparabeno nas concentrações 0,05; 0,1; 0,25; 0,5; 0,8 e 1,0 µgg-1 foi de 94; 95; 89; 83; 100 e 74% respectivamente e o desvio padrão relativo foi de 7,0; 0,5; 3,2; 10,0; 5,2 e 2,9 %para as seis concentrações. A recuperação do etilparabeno foi de 107; 103; 80; 89; 89 e 88% para as seguintes concentrações 0,05; 0,1; 0,25; 0,5; 0,8 e 1,0 µgg-1 e o desvio padrão foi de 7,4; 4,1; 12,2; 3,2; 7,7 e 3,6 % respectivamente. A recuperação do propilparabeno foi de 106; 96; 94; 97; 100 e 76 % para as seguintes concentrações 0,05; 0,1; 0,25; 0,5; 0,8 e 1,0 µgg-1 e o desvio padrão foi de 14,2; 8,6; 14,9; 4,0; 7,9 e 2,6 % respectivamente. A recuperação do butilparabeno foi de 102; 74; 96; 79; 112 e 78% para as seguintes concentrações 0,05; 0,1; 0,25; 0,5; 0,8 e 1,0 µgg-1 e o desvio padrão foi de 7,7; 4,0; 9,1; 5,5; 2,0 e 4,2 % respectivamente. Os limites de detecção variaram de 0,0007 a 0,0221 µgL-1 e os de quantificação de 0,0021 a 0,0738 µgL-1. Os coeficientes de correlação ficaram entre 0,9837 a 0,9906 dentro do limite estabelecido pela Agência Nacional de Vigilância Sanitária para matrizes biológicas. O método foi aplicado em três amostra de peixes adquiridos no comércio de São Carlos - SP. Em 50% das amostras analisadas encontrou-se picos cromatográficos nos mesmos tempos de retenção dos padrões dos parabenos. A concentração do etilparabeno foi de 0,0438 µgg-1, para o propilparabeno a concentração foi de 0,0365 a 0,0564 µgg-1 e para o butilparabeno a concentração foi de 0,0322 a 0,0678 µgg-1. O metilparabeno foi o único que não foi encontrado nas amostras de tecido dos peixes analisadas. / his research aimed to develop and validate a new chromatographic method combined with liquid-liquid microextraction (LLME) for the determination and quantification of parabens in fish tissue. A high performance liquid chromatography coupled with diode array detection was employed with a Nucleodur C8 ec column (250 x 4.6 mm, 5 µm), injection volume of 20?L, wavelength of 258nm, flow rate of 1 mL min-1, temperature of 25 ° C and gradient mode, using 1% of acetic acid diluted in water and methanol, in the proportion of 60% of methanol until 8 minutes and 65% within 20 minutes. The liquid-liquid microextraction (LLME) provided adequate recoveries of 94; 95; 89; 83; 100 and 74% for the methylparaben at concentration of 0,05; 0,1; 0,25; 0,5; 0,8; 1,0 µgg-1 respectively, with accuracy of 7,0; 0,5; 3,2; 10,0; 5,2 and 2,9 %. The recoveries for the ethylparaben was 107; 103; 80; 89; 89 and 88% at concentration of 0,05; 0,1; 0,25; 0,5; 0,8 and 1,0 µgg-1 respectively,with accuracy of 7,4; 4,1; 12,2; 3,2; 7,7 and 3,6 % The recoveries for the propylparaben was 106; 96; 94; 97; 100 and 76 % at concentration of 0,05; 0,1; 0,25; 0,5; 0,8; 1,0 µgg-1 respectively, with accuracy of 14,2; 8,6; 14,9; 4,0; 7,9 and 2,6 % . The recoveries for the butylparaben was 102; 74; 96; 79; 112 and 78%at concentration of 0,05; 0,1; 0,25; 0,5; 0,8; 1,0 µgg-1 respectively,with accuracy of 7,7; 4,0; 9,1; 5,5; 2,0 and 4,2 %. The method detection limits (MDLs) ranged from 0,0007 a 0,0221 µgL-1 and the method quatification limists (MQLs) ranged from 0,0021 a 0,0738 µgL-1. The correlation coefficients ranged from 0.9837 to 0.9906 not further than the limits established by the Agência Nacional de Vigilância Sanitária for complex samples. The method LLME/HPLC-DAD was applied to the analysis of three samples from the market of São Carlos-SP. In 50% of the samples analysed, chromatographic peaks were found at the same retention times of parabens. The ethylparaben concentrations founded was 0,0438 µgg-1, for propylparaben concentration ranged from 0,0365 to 0,0564 µgg-1 and for butylparaben concentration ranged from 0,0322 to 0,0678 µgg-1. The methylparaben was not found in the fish tissue sample analyzed.

Ce travail de thèse vise à apporter des connaissances structurales et fonctionnelles sur les protéines du gluten. Pour cela, nous utilisons les concepts et méthodes de la physique des polymères et de la matière molle. Plus précisément, nous optimisons un protocole d’extraction basé sur la séparation de phases liquide-liquide. Ce dernier permet d’obtenir des isolats de protéines à différents rapports massiques gluténines/gliadines que nous étudions ensuite dans un solvant eau/éthanol 50/50 (v/v). Les résultats, montrent que les protéines se comportent comme des chaînes de polymères en solvant θ, en régime dilué et semi-dilué avec des tailles caractéristiques définis par diffusion de rayons X et de neutrons aux petits angles. De plus, 2 tailles d’objets sont distinguées en régime dilué par diffusion dynamique de la lumière: d’une part des protéines monomériques de l’ordre d’une dizaine de nanomètres associées aux  et -gliadines et à des polymères de gluténines de faibles masses molaires et d’autre part des assemblages polymériques de l’ordre de 100 nm, principalement composés de ω-gliadines et polymères de gluténines de haute masse molaire. Ces assemblages sont mis en avant par une combinaison de mesures réalisées par chromatographie d’exclusion de taille et par fractionnement par flux de forces asymétrique et permettent de rationaliser les diagrammes de phases de ces mélanges protéiques, en fonction de la température. L’étude de la dynamique de séparation de phases de ces mélanges protéiques, par diffusion de rayons X aux petits angles, montre que celle-ci est pilotée par un phénomène de décomposition spinodale. Cette décomposition peut être arrêtée lors de trempes en température profondes mais également observée à toutes les températures de trempe, pour les échantillons les plus riches en gluténines, formant un gel dès le régime monophasique, d’après leur étude par rhéologie / The aim of this thesis is to provide structural and functional knowledge on wheat gluten proteins. For that, we use the physical methods and the concept of soft matter. We optimize an extraction protocol based on a liquid-liquid phase separation. With this protocol, we obtain protein batches with different glutenin/gliadin mass ratios, which we then study in a 50/50 water/ethanol solvent (v/v). We show that proteins behave like polymer chains in θ solvent in dilute and semi-dilute regime, whose characteristic size are extracted by small angle X-ray and neutron scattering. Moreover, two sizes of objects are evidenced in dilute regime by dynamic light scattering: monomeric proteins with a size around 10 nm which can be associated to α/β, and γ-gliadins and polymeric glutenins with low molecular weight and polymeric assemblies with a size around 100 nm composed of ω-gliadins and glutenins polymers with high molecular weight. These assemblies are revealed by a combination of size exclusion chromatography and asymmetric flow field flow fractionation and allow one to rationalize the phase diagrams of the protein mixtures with temperature. The study of the dynamics of the phase separation of these protein mixtures by small angle X-ray scattering shows that the phase separation proceeds through a spinodal decomposition phenomenon. An arrested phase separation is observed for deep quenches but also at all temperature quenches for the most glutenin rich samples, which are gels in the monophasic regime, as confirmed by rheology.

<p>Cilj ove doktorske disertacije je postavljanje<br />matematičkog modela složenog trofaznog reakcionog<br />sistem za epoksidovanje sojinog ulja in situ formiranom<br />persirćetnom kiselinom iz sirćetne kiseline i vodonik<br />peroksida u prisustvu jonoizmenjivačke smole kao<br />katalizatora. Model uzima u obzir koncentracije<br />reaktanata i produkata u vodenoj i uljnoj fazi. Pored<br />osnovnih reakcija stvaranja persirćetne kiseline i<br />epoksida, model obuhvata i sporednu reakciju otvaranja<br />epoksi prstena sa sirćetnom kiselinom. Za modelovanje<br />reakcije formiranja persirćetne kiseline na povr&scaron;ini<br />katalizatora primenjeni su Langmuir-Hinshelwood-<br />Hougen-Watson i Rideal-Eley postulati. Postavljeni<br />trofazni model predstavlja sistem običnih diferencijalnih<br />jednačina prvog reda koji opisuje promenu broja molova<br />komponenata i funkcionalnih grupa sa vremenom<br />izvođenja procesa epoksidovanja, i sadrži vi&scaron;e parametara<br />razvrstanih na kinetičke, termodinamičke i parametre koji<br />se odnose na prenos mase. Parametri modela zavise od<br />uslova izvođenja epoksidovanja i to svi od temperature, a<br />neki i od sastava i inteziteta me&scaron;anja.<br />Za konstantu hemijske ravnoteže reakcije nastajanja<br />persirćetne kiseline je izvedena semiteorijska zavisnost<br />od temperature koja daje vrednosti istog reda veličine i<br />istog smera promene sa temperaturom kao i većina<br />podataka objavljenih u literaturi.<br />Za izračunavanje koeficijenta raspodele sirćetne kiseline<br />između uljne i vodene faze potrebno je odrediti zavisnost<br />konstante fazne ravnoteže tečno-tečno sirćetne kiseline od<br />sastava i temperature. Utvrđeno je da je UNIFAC model<br />grupnih doprinosa za koeficijente aktivnosti nepogodan<br />za predskazivanje ravnoteže tečno-tečno. Eksperimentalni<br />podaci za ovu konstantu ravnoteže su uspe&scaron;no korelisani<br />UNIQUAC modelom za koeficijente aktivnosti.<br />Parametri reparametrizovane Arrheniusove zavisnosti<br />konstanti brzina reakcija i konstanti sorpcije učesnika<br />reakcije stvaranja persirćetne kiseline od temperature<br />određeni su simultano sa parametrima koji se odnose na<br />prenos mase i sa odnosom koeficijenata raspodele<br />persirćetne i sirćetne kiseline između uljne i vodene faze,<br />fitovanjem eksperimentalnih podataka epoksidovanja<br />sojinog ulja, tj. minimizacijom sume kvadrata odstupanja<br />računskih od eksperimentalno određenih vrednosti jodnog<br />broja i sadržaja epoksi kiseonika tokom epoksidovanja.<br />Fitovanje je uspe&scaron;no izvedeno primenom metode<br />Marquardta, dok su pomenute računske vrednosti<br />dobijene numeričkom integracijom sistema<br />diferencijalnih jednačina modela primenom Runge-Kutta<br />metode IV reda.</p> / <p>The objective of this doctoral thesis was development of<br />mathematical model for complex three-phase reaction<br />system for soybean oil epoxidation with peracetic acid<br />formed in situ from acetic acid and hydrogen peroxide in a<br />presence of an ion exchange resin as catalyst. The local<br />concentrations of components in water and oil phases were<br />introduced into the model. In addition to reactions of the<br />peracetic acid and epoxy compound formation, model<br />considers the side reaction of epoxy ring cleavage with<br />acetic acid. Approximate modeling of peracetic acid<br />formation was based on Langmuir-Hinshelwood-Hougen-<br />Watson and Rideal-Eley postulates. Established threephase<br />model is a system of ordinary first order differential<br />equations which describes change of components and<br />functional groups amounts with reaction time. Besides<br />kinetic parameters, model comprises the thermodynamic<br />ones as well as parameters of mass transfer between the oil<br />and water phase. All model parameters are dependent on<br />temperature and some additionally on composition and<br />intensity of stirring.<br />A semitheoretical temperature dependency of chemical<br />equilibrium constant for peracetic acid formation was<br />established. The order of magnitude and temperature trend<br />of the calculated chemical equilibrium constant are in<br />agreement with the most data given in a literature.<br />For calculation of partition coefficient for acetic acid<br />between oil and water phase, temperature and composition<br />dependency of liquid-liquid equilibrium constant for acetic<br />acid is necessary. It was found that UNIFAC model of<br />group contribution was non-applicable for the prediction<br />of the equilibrium constant. The experimental data for the<br />equilibrium constant were, however, successfully fitted by<br />UNIQUAC model.<br />Temperature dependencies of the reaction rate constants<br />and sorption constants of reactants and products in<br />peracetic acid formation reaction are expressed by<br />reparameterized Arrhenius equation. The parameters of<br />such equation were determined simultaneously with mass<br />transfer parameters and ratio of peracetic acid and acetic<br />acid partition coefficients between oil and water phase by<br />fitting the experimental data i.e. by minimization of least<br />sum of squares of deviation between the calculated and<br />experimentally determined iodine value and epoxy oxygen<br />content . Marquardt method was successfully used to fit<br />the experimental data. A fourth-ordered Runge-Kutta<br />method was applied for integrating the system of<br />differential equations of the model.</p>

Dans le cadre du développement de nouveaux procédés par extraction liquide-liquide pour le recyclage du combustible nucléaire usé, de nouvelles molécules extractantes sont à l’étude. Les molécules à fonction amide (monoamide, malonamide ou diglycolamide) sont particulièrement étudiées. Les objectifs de cette thèse sont d’étudier la spéciation moléculaire et supramoléculaire de solutions organiques représentatives des différents procédés en cours de développement. Cette spéciation a été réalisée en couplant des études expérimentales et théoriques. Après extraction de solutés, la composition des solutions organiques est déterminée expérimentalement. Des boites de simulation par dynamique moléculaire ayant la même composition que les solutions expérimentales sont ensuite construites. Après simulation, les trajectoires de dynamique moléculaire permettent de calculer les masses volumiques et les intensités diffusées aux petits angles théoriques des solutions simulées. Dès lors que les données calculées et les données expérimentales sont en accord, on estime que les simulations sont représentatives des solutions réelles. Ces simulations couplées à des caractérisations expérimentales supplémentaires (spectroscopie infra-rouge – IR – et spectrométrie de masse à ionisation par électrospray – ESI-MS –) permettent de décrire les structures en solution à la fois à l’échelle moléculaire et supramoléculaire. Cette méthodologie a été appliquée à l’extraction d’eau et de nitrate d’uranyle par les monoamides DEHBA et par le malonamide DMDOHEMA ainsi qu’à l’extraction d’eau et de nitrate de néodyme par des solutions à base de TODGA.L’extraction d’eau en phase organique est dépendante de l’organisation de la solution : les solutions de monoamides essentiellement constituées de monomères et dimères solubilisent peu d’eau contrairement aux solutions de DMDOHEMA ou de TODGA majoritairement constituées d’agrégats. L’extraction de nitrate d’uranyle a mis en évidence différents comportements dépendant de la structure de la molécule ou de la concentration d’uranyle. Des complexes UO2(NO3)2L2 sont observés à faible concentration d’uranium après extraction par les monoamides. Lorsque la concentration d’uranium augmente, des espèces polymétalliques sont observées et deviennent majoritaire en solution. Dans le cas du DMDOHEMA, des complexes monométalliques sont majoritaires mais n’ont pas une stœchiométrie unique. L’uranyle peut être coordiné à 1 ou 2 malonamides (monodenté ou bidenté), 2 nitrates et parfois une molécule d’eau. Le néodyme est extrait par le TODGA dans l’heptane sous forme de petits agrégats contenant 2 ou 3 cations liés par des nitrates pontants. Une augmentation de la concentration de néodyme entraine une augmentation de l’agrégation jusqu’à l’apparition d’une démixtion de phase. La présence d’octanol (0,3 mol/L) permet d’extraire une plus forte concentration de Nd sans séparation de phase. Les simulations ont montré que l’octanol se place dans la 1ère sphère de coordination du néodyme à la place des molécules d’eau ou de TODGA améliorant ainsi la solubilité des complexes et agrégats dans la phase organique. L’ajout de 0,5 mol/L de DMDOHEMA à une solution de TODGA permet également d’éviter la démixtion en structurant la solution. En présence de malonamide des agrégats plus petits sont observés.Pour les deux systèmes étudiés (mono et diamides), les phases obtenues après démixtion de la phase organique (phénomène de formation de 3ème phase) ont également été caractérisées.Ces travaux ont permis de déterminer l’organisation moléculaire et supramoléculaire dans des solutions d’extraction par l’utilisation d’une méthode couplant études expérimentales et simulations par dynamique moléculaire. / In the framework of development of new processes for spent nuclear fuel reprocessing, new extractant molecules are studied.The goals of this thesis are to study the molecular and supramolecular speciation of representative organic solutions. The speciation was determined by coupling experimental and theoretical study. After solute extraction, the composition of the organic solutions is experimentally determined. Simulations boxes with the same composition than experimental solutions are build. After simulation, trajectories are used to calculate small angle scattered intensities. The representativeness of the simulations is checked by comparison of experimental and calculated scattered intensities. The use of the simulation, ESI-MS spectrometry and IR spectroscopy provides the description of the structures in organic solution at the molecular and supramolecular scale. This methodology was applied on water and uranyl nitrate extraction by the monoamide DEHBA and MOEHA and by the malonamide DMDOHEMA as well as the extraction of water and neodymium nitrate by TODGA solutions.The extraction of water is dependent of the organization of the solution: monoamide solution made of monomer and dimer solubilize few amount of water in comparison with DMDOHEMA or TODGA solutions.The extraction of uranyl nitrate showed different behavior. UO2(NO3)2L2 complexes are observed at low uranium concentration after extraction by the monoamides. Polymetallic species are observed with the increase of uranium concentration. In the case of DMDOHEMA, monometallic complexes are majority but without a unique stoichiometry. The uranyl can be linked to 1 o 2 malonamides, 2 nitrates and sometimes to a water molecule. Neodymium is extracted by TODGA within small aggregates made of 2 or 3 cations liked by bridging nitrates. The increase of the concentration of neodymium leads to a phase separation. The presence of octanol provides the extraction of higher concentration of neodymium nitrate. Simulations showed the replacement of water and TODGA molecules and the first coordination sphere of the cation with octanol molecule. This increases the solubility of the aggregates. The presence of DMDOHEMA prevents also the phase separation by structuring the solution. With malonamide, smaller aggregates are observed.For both systems, phases after demixing (third phase formation) were characterized.This work provides the study of the molecular and supramolecular organization of organic solutions by combining experimental studies and molecular dynamic simulations.

Emulsion explosives have become the preferred choice as blasting agents for numerous industries including mining, agriculture, and construction. One of the most important components in such an emulsion is an emulsifier, which controls the emulsification properties of the explosive. The present study involves the production of one such emulsifier, which is produced by reacting two immiscible liquids, PIBSA (polyisobutylene succinic anhydride) and MEA (monoethanolamine). The study examines the effect of design variable such as the impeller speed, impeller type and the dispersed phase volume fraction on interfacial area. Experiments were carried out in a 0.15 m diameter fully baffled stirred tank using a 6-bladed Rushton turbine impeller and a marine propeller. Drop size was determined using a microscope with a video camera and image processing system. The transient concentration of PIBSA was determined using FTIR analysis and used to estimate the volume fraction of the dispersed phase (ƒÖ). The effective interfacial area was calculated using the Sauter mean drop diameter, d32 and ƒÖ. Impeller speeds ranging from 150 to 600 rpm and dispersed phase volume fractions, ƒÖ ranging from 0.01 to 0.028 were examined in the experimental study. It was found that that the evolution of Sauter mean drop diameter, d32 has four different trends depending on ƒÖ and impeller speed. At high impeller speeds and high ƒÖ, d32 values decrease initially and reach constant values after a long period of time. This trend is consistent with the findings in previous investigations. Under certain operating conditions, d32 values increase initially with stirring time to reach a maximum value and then decrease to reach a steady state value. The presence of these trends has been attributed to the effect of changing physical properties of the system as a result of chemical reaction. Results indicate that, in general, Sauter mean drop diameter d32 decreases with an increase in agitation intensity. However a decrease in the dispersed phase volume fraction is found to increase d32. These trends are found to be the same for both impeller types studied. Comparing the drop size results produced by the two impellers, it appears that low-power number propeller produces s ignificantly smaller drops than the Rushton turbine. It was found that the concentrations of reactants decrease with time for all impeller speeds thereby leading to a decrease in interfacial area with the progress of the reaction. Interfacial area values obtained at higher impeller speeds are found to be lower in spite of lower d32 values at these speeds. Also, these values decrease with time and become zero in a shorter duration indicating the rapid depletion of MEA. The interfacial area values obtained with the propeller at a given impeller speed are lower as compared to those for Rushton turbine. They also decrease and become zero in a shorter duration as compared to those for Rushton turbine suggesting propeller¡¦s performance is better in enhancing the reaction rate.

Thesis (Ph.D.)--McMaster University, 2005. / Supervisor: Geoffrey Brooks and Kenneth Coley. Includes bibliographical references (p. 152-164).

It was found that there exist three major water environments when water is dissolved in nitrobenzene. 2H NMR has proved that these solvatomers exist irrespective of whether lithium salt is added to the system. 7Li NMR experiments suggested that the first solvatomer is majority nitrobenzene, the second a mixed solvation shell consisting of nitrobenzene and water and the third solvatomer is a large water aggregated at the glass surface. The mixed solvation state is short lived and is promoted by addition of water of by supersaturating the system upon cooling. This is a high energy state and decays either into the homogenous bulk NB state or to the surface of the glass wall, depending on if glass surface is present. In the 7Li NMR experiments, the hydrophobicity of the salt, determined by the anion, affects the relative intensity of the three 7Li resonances. Addition of lithium serves to promote hydrogen bonding in the majority nitrobenzene solvatomer, as confirmed by FTIR and neutron diffraction studies. There is no evidence that it has an effect on the size of the mixed solvatomer or the water aggregate immobilized on the glass surface. A reasonable hypothesis is that lithium exchanges between the water species which are formed independent of lithium involvement. The system is summarized as follows: Below critical water concentration (~200mM) nitrobenzene/water is a homogeneous distribution of water molecules in nitrobenzene. Addition of lithium salt to such a system has two main affects. First, the lithium promotes hydrogen bonding between the dissolved water molecules, as confirmed by FTIR and neutron scattering. Second, the hydrogen bonded water may precipitate causing microheterogeneity of the system, leading to a second resonance observed in both the 2H and 7Li NMR spectra (LiNB/W). In the presence of glass, a third solvation state can nucleate at the glass surface; this solvation state has character even closer to that of bulk water (LiW). These two supplementary solvation states can be artificially induced by either adding aliquots of water or cooling.