Plantas do gênero Hypericum são bem conhecidas por sua utilização na medicina tradicional, sendo Hypericum perforatum a espécie mais estudada. No Sul do Brasil, esse gênero é representado por cerca de 20 espécies, entre elas, destaca-se H. polyanthemum, o qual apresenta em extrato lipofílico a presença de uliginosina B, um derivado de floroglucinol, e três benzopiranos: HP1, HP2 e HP3. Partes aéreas de H. polyanthemum foram submetidas à extração com fluido supercrítico (SFE), sendo realizadas extrações sucessivas com dióxido de carbono nas pressões de 90, 120, 150 e 200 bar, em diferentes temperaturas (40, 50 e 60 °C). Posteriormente, as frações obtidas foram comparadas com extrato n-hexano obtido por técnica convencional (ultra-som). SFE apresentou maior seletividade que a extração com nhexano, sendo a 50 °C a melhor condição para extrair os metabólitos bioativos. Para extração do composto com maior peso molecular, uliginosina B, pressões mais elevadas foram necessárias. Além disso, avaliou-se a atividade anti- Trichomonas vaginalis do extrato obtido por SFE a 50 °C/150 bar e dos compostos isolados. Os resultados indicam que todas as amostras testadas possuem atividade anti-T. vaginalis, porém, HP1 demonstrou melhor seletividade frente ao protozoário (isolado metronidazol-resistente e -suscetível), sem apresentar citotoxicidade contra células de mamíferos. Ainda, o HP1 apresentou satisfatória atividade contra isolado metronidazol-resistente (52% de trofozoítos viáveis), apresentando efeito sinérgico quando testado com baixa concentração de metronidazol (23% de trofozoítos viáveis). Todos os compostos isolados provocaram danos na membrana do parasito (liberação de LDH > 90%) e nenhuma das amostras apresentou efeito hemolítico importante, sendo que HP2 e uliginosina B apresentaram citotoxicidade frente a células de mamíferos. Portanto, benzopiranos e derivados de floroglucinol são moléculas promissoras como protótipos para novos fármacos tricomonicida, principalmente HP1, visto que corrobora com a ação do metronidazol em isolados de T. vaginalis resistentes. No entanto, mais estudos são necessários a fim de melhorar a atividade antiprotozoária das moléculas testadas, sem que as mesmas apresentem toxicidade para células de mamíferos. / Plants of the genus Hypericum are well known for their use in traditional medicine, being H. perforatum the most important species. In South Brazil, this genus is represented by approximately 20 species and among them, H. polyanthemum showed in the lipophilic extract the presence of uliginosin B, a phloroglucinol derivative, and three benzopyrans: HP1, HP2 and HP3. Aerial parts of H. polyanthemum were submitted a supercritical fluid extraction (SFE), being successively extracted with supercritical carbon dioxide under pressures of 90, 120, 150 and 200 bar at different temperatures (40, 50 and 60 ºC), and compared with the n-hexane extract obtained by ultrasound-assisted extraction. SFE presented higher selectivity than the n-hexane extration. The best condition to extract the target metabolites has been determined to be at 50 °C, being that for the high molecularweight compound, uliginosin B, higher pressures was required. Moreover, the anti- Trichomonas vaginalis activity of the extract obtained by SFE at 50 °C/150 bar and the isolated compounds was evaluated. The results showed that all samples have anti-T. vaginalis activity; however, HP1 demonstrated the best selectivity against this protozoa (metronidazole-resistant and -susceptible isolates), with no cytotoxicity against mammalian cells. In addition, HP1 has good activity against metronidazoleresistant isolate (52% of viable trophozoites), presenting synergic effect when tested with low concentration of metronidazole (23% of viable trophozoites). All isolated compounds caused damage in the parasites membrane, revealed by more than 90% of LDH release. None of the samples presented important hemolytic effect, and HP2 and uliginosin B presented cytotoxicity against mammalian cells. Therefore, benzopyrans and phloroglucinol derivative are promising molecules as prototypes for new antiprotozoal drugs, especially HP1, which improve metronidazole action against a resistant T. vaginalis isolate. Nevertheless, more studies are necessary in order to enhance their in vitro antiprotozoal activity without toxicity for mammalian cells.

Este projeto de pesquisa, com o apoio dos pequenos produtores de café do Estado de São Paulo/Brasil, visou estudar e caracterizar o óleo de grãos de café verde obtido pela extração supercrítica e extração com fluido pressurizado. Foi feita a caracterização deste produto, considerando os componentes minoritários com atividade biológica demonstrada na literatura, como os componentes fenólicos e os diterpenos cafestol e kahweol. Com o objetivo de obter o óleo de café verde enriquecido, principalmente nestes dois principais diterpenos, componentes relevantes para as indústrias cosmética e farmacêutica, aplicaram-se como inovação tecnológica a extração supercrítica (SFE) e extração com fluido pressurizado (PFE). Os ensaios de extração foram realizados no Laboratório a Alta Pressão e Produtos Naturais (FZEA/USP/Brasil) para as SFE e no Laboratoire de Procédés (IUT/Université d\'Orléans/França) para PFE. No desenvolvimento da metodologia de análise dos diterpenos foram realizados estudos do método de saponificação do óleo de café verde Arabica (Coffea arabica) cv. Catuaí amarelo, obtido por prensagem (OP). E, concomitantemente a este estudo analítico, foram feitas as extrações do óleo dos mesmos grãos de café verde via SFE e PFE. Os diferentes óleos de grãos de café verde, obtidos pelos três métodos de extração foram analisados e comparados entre si considerando principalmente a concentração dos diterpenos. A metodologia analítica desenvolvida na primeira fase do trabalho constatou que para uma melhor análise dos diterpenos a reação de saponificação deve ser feita utilizando NaOH e a 70 °C. A fração insaponificável foi analisada para os diferentes extratos. Para os extratos obtidos via SFE, a cafeína teve que ser separada do óleo. A análise de quantificação dos diterpenos via cromatografia gasosa, utilizando o próprio cafestol e kahweol como padrão externo, indicou que o óleo obtido por SFE apresentou 50.180,41 mg cafestol/kg óleo e 63.764,32 mg kahweol/kg óleo na condição otimizada. Para a PFE as concentrações máximas obtidas foram de 65.876,29 mg cafestol/kg óleo e 53.328,47 mg kahweol/kg óleo e, no óleo de café verde obtido pelo método convencional a concentração dos diterpenos mostrou-se inferior com 7.508 mg cafestol/kg óleo e 12.822 mg kahweol/kg óleo. As propriedades físicas medidas visando a caracterização do óleo dos grãos de café verde obtido por SFE e prensagem dos grãos indicaram que tanto a densidade quanto a viscosidade diminuem com o aumento da temperatura. Para ambos os óleos, os comportamento da densidade foi semelhante, para aquele obtido pelo método convencional, os valores variaram de 0,94177 g/cm3 (25 °C) a 0,92143 g/cm3 (55 °C) e para o óleo obtido por SFE a variação foi de 0,93645 g/cm3 (25 °C) a 0,9157 g/cm3 (55 °C). A viscosidade dinâmica para o óleo obtido por prensagem variou de 127,8798 mPa×s (25 °C) a 35,0510 mPa×s (55 °C) e, para o óleo obtido por SFE, estes valores foram menores, entre 84,0411 mPa×s (25 °C) e 24,2555 mPa×s (55 °C). O índice de refração para ambos os óleos foi semelhante. Na extração com fluido pressurizado, (PFE), um valor ótimo de rendimento do extrato foi obtido quando se empregou metanol como solvente para uma temperatura de 100 °C e tempo de equilíbrio estático de 20 min. Este valor foi 183% superior a media dos resultados obtidos nos outros ensaios. Na concentração de diterpenos nestes extratos, nota-se também que o etanol foi menos eficiente que o metanol. Na comparação dos dois processos inovadores (SFE e PFE), a concentração de kahweol apresenta-se como o principal componente em quase todos os ensaios, com valores da relação cafestol/kahweol entre 0,73 e 0,96. No entanto, ao ser analisado o processo individualmente, percebe-se que a extração com fluido supercrítico indica que a maioria dos extratos obtidos teve uma concentração de kahweol superior ao cafestol. / This research project, with the support of small coffee producers in the State of São Paulo/Brazil, aimed to study and characterize the green coffee oil obtained by the supercritical fluid extraction and pressurized fluid extraction. Was made to characterize this authentic product, considering the minor components with biological activity demonstrated in the literature, such as phenolic compounds and diterpenes cafestol and kahweol. In order to get the green coffee oil enriched, mainly in these two major diterpenos, components relevant to cosmetic and pharmaceutical industries, applied technological innovation as the supercritical fluid extraction (SFE) and pressurized fluid extraction (PFE). The extraction tests were performed at the High Pressure Laboratory and Natural Products (FZEA/USP/Brazil) for SFE and the Laboratoire de Procédés (IUT/Université d\' Orléans/France) for PFE. In the development of methodology for analyzing diterpenes were conducted studies on the method of saponification of the green oil Arabica coffee (Coffea arabica) cv. yellow Catuaí, obtained by pressing (OP). And, concomitantly this analytical study were extracted from the oil of the same green coffee beans by SFE and PFE. Different green coffee beans oils, obtained by the three extraction methods were analyzed and compared with each other considering mainly the concentration of diterpenes. The analytical methodology developed in the first phase of the work found that for a better analysis of the diterpenes saponification reaction must be made using NaOH and 70 °C. The unsaponifiable fraction was analyzed for the different extracts. Specifically for extracts obtained by SFE, caffeine had to be separated from the oil. The analysis and quantification of diterpenes made by gas chromatography, using himself as cafestol and kahweol external standard, indicated that the oil obtained by SFE showed cafestol 50,180.41 mg/kg oil and kahweol 63,764.32 mg/kg oil in optimal condition. For the PFE maximum concentrations were obtained cafestol 65,876.29 mg/kg oil and kahweol 53,328.47 mg/kg oil and the green coffee oil obtained by the conventional method the concentration of diterpenes was lower than with 7,508 mg cafestol/kg oil kahweol and 12,822 mg/kg oil. The physical measures for the characterization of oil from green coffee beans obtained by SFE and pressing the grains indicated that both the density and the viscosity decrease with increasing temperature. For both oils, the density behavior was similar, to that obtained by the conventional method, the values ranged from 0.94177 g/cm3 (25 °C) to 0.92143 g/cm3 (55 °C) and the oil obtained by the SFE variation was 0.93645 g/cm3 (25 °C) to 0.9157 g/cm3 (55 °C). The dynamic viscosity for the oil obtained by pressing ranged from 127.8798 mPa×s (25 °C) to 35.0510 mPa×s (55 °C) and the oil obtained by SFE, these values were lower, between 84.0411 mPa×s (25 °C) and 24.2555 mPa×s (55 °C). The index of refraction for both oils was similar. In the extraction pressure, (PFE), an optimal value of the extract yield was obtained when methanol was used as solvent at a temperature of 100 °C and static equilibrium time of 20 min. This value was 183% above the average of the results obtained in other tests. At the concentration of these diterpenes extracts, note also that the ethanol was less effective than methanol. In comparing the two innovative processes (SFE and PFE), the concentration of kahweol presents itself as the main component in almost all tests, compared with values of cafestol/kahweol between 0.73 and 0.96. However, the process to be analyzed individually, it is clear that the supercritical fluid extraction indicates that most of the extracts had a concentration higher of kahweol than cafestol.

Orientador : Martin Aznar / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica / Made available in DSpace on 2018-08-01T10:07:01Z (GMT). No. of bitstreams: 1 Arce-Castillo_PedroFelipe_M.pdf: 13948249 bytes, checksum: 89b291cb7fd018bc32c50812717efa25 (MD5) Previous issue date: 2002 / Resumo: A modelagem e simulação de processos de extração supercrítica requer uma boa predição das condições do equilíbrio entre fases, condições que determinam a maior concentração do líquido ou sólido (soluto) a ser extraído pelo fluido supercrítico (solvente). O problema principal na modelagem dos sistemas que se encontram em processos de extração supercrítica é a grande diferença de tamanho (peso molecular) entre as substâncias envolvidas (soluto e fluido supercrítico) e a baixa concentração do soluto no fluido supercrítico. Atualmente, as regras de mistura em uso não consideram de forma adequada o problema da grande assimetria; portanto, não permitem uma boa predição do equilíbrio entre fases. É proposta uma regra de mistura (Regra Não Quadrática Generalizada) baseada nas regras de mistura clássicas de van der Waals, com modificações nos parâmetros de interação, tanto na constante de energia quanto na de volume. Um parâmetro de interação binário, dependente da concentração, é usado para estimar a constante de energia; outro parâmetro de interação binário, independente da concentração e com um efeito diferenciado sobre o componente mais pesado quando comparado ao fluido supercrítico, é usado para estimar a constante de volume. Este novo arranjo considera como casos particulares os modelos conhecidos de Panagiotopoulos-Reid e Adachi-Sugie, entre outros. A forma não quadrática generalizada proposta tem todas as boas características desses modelos, permitindo maior flexibilidade na correlação dos dados experimentais. Usou-se outra regra de mistura, Regra de Kurihara-Tochigi-Kojima, a qual está baseada na parte residual da energia livre excedente à pressão infinita, atuando diretamente sobre a constante de energia da equação de estado. Os resultados obtidos com as duas regras de misturas citadas anteriormente são comparados com os resultados obtidos da regra modificada de W ong-Sandler e da regra convencional de van der Waals. As regras de mistura descritas anteriormente, junto com as equações cúbicas de estado de Soave-Redlich-Kwong, Peng-Robinson e a equação cúbica de estado generalizada de Patel-Teja-Valderrama, são usadas para descrever o equilíbrio líquidovapor e sólido-vapor em misturas assimétricas binárias (CO2 supercritico + um componente pesado ou produto natural). Neste trabalho foram estimados os parâmetros de interação binários tanto para a modelagem do equilíbrio líquido-vapor dos sistemas binários: CO2 +limoneno, linaloo_ ac. láurico, ac. palmítico, ac. oléico, ac. linoleico, 2-metil-pentanol, 1octanol, l-decanol e a.-pineno, quanto para a modelagem do equilíbrio sólido-vapor dos sistemas binários: CO2 + naftaleno, 2,3-dimetilnaftaleno, 2,6-dimetilnaftaleno, fenantreno,antraceno, p-colesterol, cafeína, p-caroteno e capsaicina. A predição dos parâmetros de interação binários realiza-se usando um método modificado de Marquardt com uma função objetivo que contém a pressão de saturação e a concentração na fase gás para o equilíbrio líquido-vapor e apenas a concentração na fase gás para o equilíbrio sólido-vapor. As sub-rotinas computacionais foram escritas considerando a possibilidade de múltiplas soluções e conseqüentemente, a busca dos parâmetros ótimos realiza-se sob um amplio intervalo de soluções possíveis. Os resultados (desvios na pressão e na fração molar na fase vapor para o ELV e apenas os desvios na fração molar na fase vapor para o ESV) indicam que as regras de mistura NQG e WS modificada permitem predizer melhor o comportamento dos sistemas binários estudados. A influencia das EDEs para uma mesma regra de mistura não é apreciável / Abstract: Modeling and simulation of supercritical extraction processes request a good prediction of conditions of phase equilibrium, conditions that determine the higher concentration of liquid or solid (solute) to be extracted by supercritical fluid (solvent). The main problem in modeling of systems found in supercritica1 extraction processes is the great size difference (molecular weights) between involved substances (solute and supercritical fluid) and low concentration of solute in supercritical fluid. Now, mixing roles in use don't consider of appropriate way the problem of great asymmetry, therefore don't alIow a good prediction of phase equilibrium. A mixing role is proposed (Generalized Non Quadratic Rule), which is based on c1assic mixing rules of van der Waals, with modifications in the interaction parameters, as much in energy constant as in volume constant. A binary interaction parameter, dependent of concentration, is used to estimate energy constant; another binary interaction parameter, independent of concentration and with a differentiated effect on heavier component when compared with supercritical fluid, is used to estimate volume constant. This new arrangement considers as private cases well-known models of Panagiotopoulos-Reid and Adachi-Sugie among others. This generalized non quadratic form has all good characteristics of those models, allowing larger flexibility in correlating experimental data. Other mixing rule was used, Kurihara-Tochigi-Kojima Rule (based on the residual part of the excess free energy to infinite pressure), acting direct1y on energy constant of the state equation. Results obtained with the two mixing roles previously mentioned are compared with obtained results of modified role of Wong-Sandler and conventional role of van der Waals. The mixing roles described previously, joined with the Soave-Redlieh-Kwong and Peng-Robinson's cubic equations of state and the Patel-Teja-Valderrama generalized cubic equation of state, are used to describe liquid-vapor and solid-vapor equilibrium in asymmetrie binary mixtures (supercritical CO2 + heavy component or natural product). In this work were estimated the binary interaction parameters as much for the modeling of liquid-vapor equilibrium of binary systems: COz + limonene, linalool, lauric acid, palmitic acid, oleic acid, linoleic acid, 2-methyl-pentanol, l-octanol, l-decanol and a-pinene, as for modeling of solid-vapor equilibrium of the binary systems: CO2 + naphtalene, 2,3dimethylnaphtalene, 2,6-dimethylnaphtalene, phenanthrene, anthracene, B-eholesterol, caffeine, B-carotene and capsaicin. A Marquardt modified method with a objective function (saturation pressure and vapor phase concentration for liquid-vapor equilibrium and on1y vapor phase concentration for solid-vapor equilibrium) was used to predict binary interaction parameters. The computational subroutines were written considering the possibility of multiple solutions and consequently, the search of the optimum parameters was done on a large interval of possible solutions. Results (deviations in pressure and molar fraction in vapor phase for ELV and on1y deviations in fraction molar in phase vapor for ESV) indicate that mixture roles: NQG and modified WS allow to predict better the behavior of studied binary systems. The influence of EDEs for a same mixture role is not appreciable / Mestrado / Desenvolvimento de Processos Químicos / Mestre em Engenharia Química

Orientador: Maria Angela de Almeida Meireles / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia de Alimentos / Made available in DSpace on 2018-08-04T02:50:47Z (GMT). No. of bitstreams: 1 Leal_PatriciaFranco_M.pdf: 2204024 bytes, checksum: 0b4a27a2c6db03ca4023ca90d6a49f65 (MD5) Previous issue date: 2005 / Mestrado / Engenharia de Alimentos / Mestre em Engenharia de Alimentos

In Canada, about 60% of contaminated sites involve petroleum hydrocarbon (PHC) contamination and most of these sites have been abandoned due to contamination. Among current technologies used for soil remediation, supercritical fluid extraction (SFE) is a relatively recent and potentially viable method. The main aim of this research was to investigate the application of SFE for removal of PHCs from contaminated soils. In the first phase, the effects of SFE operational parameters including fluid pressure, fluid temperature, time duration and mode of extraction on the removal efficiency of PHCs from a spiked sandy soil (with diesel fuel with a ratio of 5 wt%) were investigated. SFE experiments were performed at different pressures (15, 33 and 50 MPa) and temperatures (30, 75 and 120 °C). The combination of 10 min static mode followed by 10 min dynamic mode, repeated for 3 cycles (60 min in total) led to the highest PHC removal percentage. According to response surface methodology (RSM), the optimum pressure and temperature were found to be 50 MPa and 69.3 °C, respectively. According to experimental results, the optimum combination of pressure and temperature determined to be 33 MPa and 75 °C; which resulted in the extraction percentages of 99.2%, 91.7% and 86.1% for PHC F2, F3 and F4 fractions, respectively. In the second phase, the influence of several parameters including soil water content, soil pH and addition of modifier on PHCs removals from a field-contaminated sandy soil using SFE were experimentally investigated. SFE experiments were performed at 33 MPa pressure and temperatures of 45 and 75 °C. Three water content levels of 8%, 14% and 20% at two levels of pH 6.5 and 7.5 were investigated. The extraction of total petroleum hydrocarbon fractions (TPHF), the sum of F2, F3, and F4 fractions, decreased due to the increase in the water content from 8% to 20% at both pH 6.5 and 7.5. The difference of extractions of all PHC fractions at pH values of 6.5 and 7.5 were not statistically significant (at p < 0.05 confidence level) at all three water content levels and pH did not have a significant influence on the PHC removal efficiency. Addition of acetone as a modifier (33.7% TPHF removal) was more effective than hexanes (24.3% TPHF removal) to decrease the concentrations of PHCs for the field contaminated soil. In the third phase, the influence of soil texture and grain size on the extraction of PHC fractions was investigated. SFE experiments were performed at 33 MPa pressure and 75 °C temperature. Three types of soils (soil A, B and C) were spiked with diesel fuel with a ratio of 5 wt%. Soil A, B and C had different particle sizes and were categorized as sand, silt loam and clay, respectively. Soil A (sand) which had the largest particle size resulted in the highest TPHF removal percentage while soil C (clay) with the smallest particle size led to the lowest TPHF removal percentage. A higher clay content in soil C resulted in a lower extraction of PHCs. In the fourth phase, the effects of pressure and temperature on the extraction of PHC fractions from a clay soil spiked with diesel fuel with a ratio of 5 wt% were investigated. SFE experiments were performed at three pressures (15, 33 and 50 MPa) and temperatures (30, 75 and 120 °C). According to the statistical analysis including factorial design and RSM, the optimized combination of pressure and temperature was selected at 42.8 MPa and 120 °C; which resulted in the removal percentages of 74.9% and 65.6% for PHC F2 and F3 fractions, respectively. The optimum combination of pressure and temperature based on the experimental results was selected at 33 MPa and 120 °C that led to 70.3%, 58.4% and 32.6% removal of PHC F2, F3 and F4 fractions, respectively.

The thesis focuses on ‘green’ sample preparation techniques that utilise minimal or no organic solvents thereby producing negligible volumes of organic waste, to ensure safety both to user and the environment. Volatile compounds were extracted and pre-concentrated from Kgalagadi desert truffles (kalaharituber) by headspace solid phase microextraction (HSSPME) and supercritical fluid extraction (SFE). PHWE was employed for the extraction of amino acids and fatty acids. Subsequent analysis of volatile compounds was carried out by gas chromatography coupled with mass spectrometry. Four types of HS-SPME fibers (PDMS 100 μm, PDMS 7 μm, Polyacrylate 85 μm, CAR/DVB/PDMS 50/30 μm) were evaluated. A total of 24 volatile compounds with a molecular weight range from 110 to 354, the most prominent peak being 2-t-Butyl-2,3-dimethyl-3-buten-1-ol (C₁₀H₂₀O, MW 156) were detected after sampling with a PDMS 100 μm fiber. Less volatile compounds were detected after SFE with CO2. A total of 16 amino acids were identified while 17 fatty acids (MW from 132.12 to 367.49) were also identified. The characteristic profile of the Kgalagadi desert truffle was found to contain mainly fatty acid methyl esters and unsaturated aliphatic hydrocarbons. The most prominent compound peaks identified were; 2-t-butyl-2,3-dimethyl- 3-buten-1-ol, disulfide, ethyl benzoic acid 2-4-dhydroxy-3,6-dimethyl-methyl ester, 8,11- octadecanoic acid methyl ester, benzoic acid, 2,4-dihydroxy-3,6-dimethyl-methyl ester, isoquinoline, 1 butyl-3,4-dihydro and 3-heptanone, 6 methyl. Optimization results indicated that fresh slices from the heart of truffles were the best to use for HS-SPME-GCMS volatile compound analysis as they showed a higher sensitivity

Cette thèse est une étude théorique et expérimentale sur la formation de liposomes par des procédés utilisant le dioxyde de carbone (CO2) supercritique. Les liposomes sont des vésicules sphériques nano- ou micrométriques dont la paroi est composée d’une ou plusieurs bicouches concentriques de phospholipides séparant un milieu aqueux d’un autre. L’efficacité des procédés supercritiques, comme alternative aux méthodes conventionnelles pour former des liposomes, a déjà été démontrée. Une synthèse critique des résultats de la littérature a été réalisée au cours de ce travail. Parallèlement, une étude fondamentale sur le comportement des systèmes ternaires CO2/eau/surfactant sous pression a été menée et il a été démontré qu’en fonction du type d’émulsions formées au cours du procédé sous pression, les caractéristiques des liposomes produits lors de la dépressurisation sont différentes. Le procédé Supercritical Anti-Solvent a été utilisé pour microniser des phospholipides. Des liposomes sphériques, multilamellaires et de tailles comprises entre 0,1 et 1µm ont ensuite été formés de manière reproductible par hydratation des phospholipides micronisés. Dans les mêmes conditions (même solvant), la méthode conventionnelle de Bangham n’a pas donné des résultats reproductibles et les liposomes formés n’étaient pas tous sphériques. Le résultat majeur de ce travail est la mise au point d’un procédé supercritique innovant. Innovant car continu et en une seule étape, le procédé Continuous Anti-Solvent permet de former de manière reproductible des liposomes sphériques, multilamellaires et de tailles comprises entre 10 et 100µm. / The present thesis is dedicated to liposome formation using supercritical carbon dioxide (CO2). Liposomes are spherical vesicles composed of one or more concentric phospholipid bilayers surrounding an aqueous core. Dense gas processes offer reliable alternatives to conventional methods in liposome formation. We present a review of the literature and we summarized all of the results in a discussion section, with particular attention to emulsion formation under pressure. As it has been shown, the phase behaviour of the ternary CO2/water/surfactant system under pressure greatly influences liposome formation during depressurization. The Supercritical Anti-Solvent process has been used to micronized phospholipids. Then, the micronized particles were hydrated to form spherical and multilamellar liposomes with diameters between 0.1 and 1µm in a reproducible way. Using the same conditions (the same solvent), the conventional Bangham method did not provide reproducible assay results and formed liposomes were not all spherical. The main result of this work is the design of a new supercritical process dedicated to liposome formation. Unlike the current dense gas technologies, the Continuous Anti-Solvent method breaks new ground because it is a single step and continuous process. Liposomes prepared with the Continuous Anti-Solvent method are spherical and multilamellar with diameters between 10 and 100µm.

O objetivo do presente estudo foi medir experimentalmente a solubilidade do óleo de café verde em dióxido de carbono supercrítico em condições preestabelecidas de pressão (P) e temperatura (T), e a partir dos dados experimentais estabelecer ajustes em modelos termodinâmicos que empregam a equação de estado (EDE) de Peng-Robinson com a regra de mistura clássica de van der Waals para predição do equilíbrio de fases. Foi obtido óleo a partir de grãos de café verde (Coffea arabica), cultivar (cv.) Catuaí Amarelo moído (dm de 0,8409 mm) usando CO2 supercrítico, nas condições de 40, 50, 60, 70, 80 °C e nas pressões de 300 e 350 bar. A solubilidade foi medida pelos métodos dinâmico e estático nas mesmas condições. Foi determinado o perfil de ácidos graxos por cromatografia gasosa acoplada à espectrometria de massas (CG/EM) e, a partir destes resultados, a composição em triacilgliceróis (TAGs) foi estimada. Foram determinadas as propriedades físico-químicas dos TAGs pelo método de contribuição de grupos. Nas condições de P e T estudadas, os maiores rendimentos do óleo de grãos de café verde, 7,58% e 7,60%, foram obtidos nas condições de 70 °C e 300 bar e 80 °C e 350 bar, respectivamente. Normalmente o rendimento do óleo aumentou com o aumento da temperatura, com exceção das condições de 80 °C e 300 bar e 70 °C e 350 bar. O aumento do rendimento com a temperatura reflete nos dados experimentais da solubilidade (0,0099 a 0,026 g/gCO2) que aumentou com a temperatura a 300 e 350 bar para o método estático e para o método dinâmico a 350 bar (0,0058 a 0,0065 g/gCO2). Um comportamento distinto ocorreu com a medida de solubilidade pelo método dinâmico a 300 bar. Neste caso a solubilidade diminuiu com o aumento da temperatura. Para a maioria das condições de P e T estudadas, o comportamento da solubilidade foi semelhante em função destas variáveis, o mesmo para os métodos estático e dinâmico, no entanto com ordens de grandezas distintas. As condições de extração não influenciaram na composição dos ácidos graxos no óleo nem na predição dos TAGs presentes. Os TAGs apresentaram variações entre número de carbonos (C50 a C58). As propriedades físico-químicas dos seis principais TAGs presentes no óleo de grãos de café verde (PSL, PLL, PLP, PLO, PLA e POP), temperatura do ponto de ebulição normal (Tb), temperatura crítica (Tc), pressão crítica (Pc) e fator acêntrico (&omega;) foram estimadas por métodos de contribuição de grupos. Estas variáveis são dados de entrada para a predição do equilíbrio de fases entre o óleo de grãos de café verde, representado pela mistura dos seis TAGs em maior concentração neste óleo, e o CO2 supercrítico. Os valores de Tb dos TAGs variaram de 821,52 a 844,50 K, os valores de Tc de 909,12 a 930,02 K, os valores de Pc de 5,0115 a 4,7221 bar e o &omega; de 1,5704 a 1,7452. Os dados de propriedades físicas e da solubilidade convertida em fração molar foram usados para predizer os parâmetros de interação binária da regra de mistura clássica de van der Waals entre cada TAG e o CO2 supercrítico, já que estes parâmetros também são dados de entrada para a modelagem termodinâmica que emprega a EDE de Peng-Robinson. No entanto, o programa não rodou, não sendo possível este cálculo, nem tão pouco a predição do equilíbrio de fases. Os dados experimentais da solubilidade do óleo de grãos de café verde em CO2 supercrítico determinados pelos métodos estático e dinâmico indicou que àqueles determinados pelo método estático são mais precisos e confiáveis. / The objective of this study was to experimentally measure the solubility of green coffee oil in supercritical carbon dioxide under predetermined pressure (P) and temperature (T) conditions and, from the experimental data, establish thermodynamic models that use the Peng-Robinson equation of state (EOS) and the classic mixing rule of van der Waals to predict phase equilibrium. The oil was obtained from ground green coffee beans (Coffea arabica L.) cultivar (cv.) Yellow Catuaí (dm of 0.8409 mm) using supercritical CO2 under temperatures of 40, 50, 60, 70 and 80 °C and pressures of 300 and 350 bar. The solubility was measured by dynamic and static methods under the same conditions. The fatty acid profile was determined by gas chromatography-mass spectrometry (GC/MS) and, from these results, the composition of triacylglycerols (TAGs) was estimated. Physicochemical properties of TAGs were determined by the group contribution method. In the studied conditions of P and T, higher yields of green coffee beans oil, 7.58% and 7.60%, were obtained at 70 °C and 300 bar and 80 °C and 350 bar, respectively. Normally, the oil yield increased with increasing temperatures, except for the conditions of 80 °C and 300 bar and 70 °C and 350 bar. The increase of yield with temperature reflects in the experimental solubility data (from 0.0099 to 0.026 g/g CO2), which, also increased with T at 300 and 350 bar for the static method and, for the dynamic method, at 350 bar (from 0.0058 to 0.0065 g/g CO2). A different behavior occurred with the solubility measured by dynamic method at 300 bar. In this case, the solubility decreased with increasing temperature. For most of the P and T conditions studied, the solubility behavior was similar in function of these variables, the same for static and dynamic methods, but with different magnitude orders. The extraction conditions did not affect the composition of fatty acids in the oil and nor the TAGs prediction. The TAGs showed variations between carbon numbers (from C50 to C58). The physicochemical properties of the six major TAGs present in green coffee bean oil (PSL, PLL, PLP, PLO, PLA and POP), boiling temperature (Tb), critical temperature (Tc), critical pressure (Pc) and acentric factor (&omega;) were estimated by group contribution methods. These are input variables for the prediction for the phase equilibrium between the green coffee oil, represented by the mixture of the six TAGs in higher concentration in this oil, and supercritical CO2. The Tb values of TAG ranged from 821.52 to 844.50 K, Tc values from 909.12 to 930.02 K, Pc values from 5.0115 to 4.7221 bar and &omega; from 1, 5704 to 1.7452. The data of physical properties and solubility converted to mole fraction were used to predict the binary interaction parameters of the classic mixing rule of van der Waals between each TAG and supercritical CO2, since these are also input parameters for the thermodynamic modeling that use Peng-Robinson EOS. However, the program did not run and this calculation was not possible, nor the prediction for the phase equilibrium. The experimental data on the solubility of the green coffee beans oil in supercritical CO2 determined by static and dynamic methods indicated that those determined by the static method are more accurate and reliable.

A novel chromatographic method was devised that makes use of the superb group separation power of normal phase supercritical fluid chromatography (SFC) combined with a fast second separation by a resistively heated gas chromatograph (GC). The SFC was operated isothermally with stopped flow to provide the time required for the GC analysis. The GC analysis had a typical cycle time of 1 minute. During this time the GC column was independently heated at a rate of 450°C/min to 250°C and actively cooled down again to -50°C before the next GC injection takes place. This was achieved with an in-house designed, resistively heated, temperature programmable gas chromatograph. Various temperature measurement circuits were also evaluated. An interface was developed that allows transfer between the SFC and the GC in such a way that the entire eluent from the first separation is analysed by the second separator. Chromatographic resolution was not lost during the transfer process from the first to the second separation stages. The interface also allows for the exchange of the carrier gas used in the second gas chromatographic separation to provide for the maximum separation speed. In the first separation, a silica gel packed column and the novel application of a silica gel porous layer open tubular capillary column was used for SFC group separation. The SFCxGCftp was applied to petrochemical samples and essential oils and the results were compared to that obtained with a commercially available GCxGC system. / Thesis (PhD)--University of Pretoria, 2003. / Chemistry / unrestricted

(EN) Cellulose tris-(3,5-dimethylphenylcarbamate) chiral stationary phase was used for separation of selected 24 analytes. Enantioseparations were realized using two systems, high performance liquid chromatography and supercritical fluid chromatography. Effect of mobile phase composition was studied. Five different aditives (isopropylamine, diethylamine, triethylamine, trifluoroacetic acid, isopropylamine combined with trifluoroacetic acid) and their influence on enantioseparation were tested. Influence of two different modifiers (methanol, propan-2-ol) combined with all aditives was also tested in supercritical fluid chromatography system. The aim of this work was to find optimized composition of mobile phase which was suitable for separation of the analytes studied and to compare separation potential among all mobile phases and also between used separations systems. The supercritical fluid chromatography was shown to yield better results, i.e. better resolution in shorter analysis time. However examples of analytes better resolved under optimized conditions in high performance liquid chromatography system have also been found. Keywords (EN) Chirality, enantiomers, enantioselective separation, chiral stationary phase, high performance liquid chromatography, supercritical fluid chromatography.