Orientador: Maria Angela de Almeida Meireles / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia de Alimentos / Made available in DSpace on 2018-08-09T21:18:11Z (GMT). No. of bitstreams: 1 Leal_PatriciaFranco_D.pdf: 2381293 bytes, checksum: 5131ad6a5dbd2eb071d1516e4f5d09fc (MD5) Previous issue date: 2008 / Resumo: Este estudo teve por objetivo comparar o custo de manufatura (COM) de extratos vegetais e óleos voláteis de plantas aromáticas, condimentares e medicinais nativas ou adaptáveis ao cultivo brasileiro obtidos por processos de extração considerados limpos e que não agridem o meio ambiente. Foi realizado o estudo técnico-econômico do processo de extração com fluido supercrítico (SFE) e destilação por arraste a vapor (SD) para obtenção de extratos vegetais e óleos voláteis. A metodologia de estimativa de custo de manufatura utilizada neste estudo resulta em COM classe 4 ou 5, sendo uma estimativa preliminar para a decisão de prosseguimento ou a interrupção do projeto. As estimativas dos custos de manufatura dos extratos foram realizadas através das curvas globais de extração (OEC) dos sistemas determinadas experimentalmente neste trabalho (alecrim + vapor, camomila + vapor, erva-doce + vapor, erva-baleeira + vapor, erva-baleeira + CO2, canelinha + CO2) e também através das OECs da literatura (alecrim + vapor, alecrim + CO2, erva-doce + vapor, erva-doce + CO2,, gengibre + CO2, camomila + CO2, pimenta-do-reino + vapor, pimenta-do-reino + CO2, funcho + CO2,). As curvas globais de extração foram ajustadas a um spline contendo duas ou 3 retas: a primeira reta descreve a taxa de transferência de massa no período de taxa de extração constante (CER), a segunda reta descreve a taxa de transferência de massa no período de taxa de extração decrescente (FER) e a terceira reta descreve a taxa de transferência de massa no período controlado pela difusão (DC). Para o scale-up assumiu-se que o rendimento de extração e o tempo do processo para o equipamento em escala industrial são linearmente correlacionados aos parâmetros da escala laboratorial quando é mantida constante a razão entre a massa de solvente e a massa de alimentação. Os COMs para os extratos obtidos por SD e SFE foram estimados durante todo o processo de extração para tempos pré-selecionados ao longo das OECs. Para os extratos obtidos por SFE, o COM também foi estimado para dois tempos de processo préselecionados: (i) o tempo de ciclo correspondente ao período CER e denotado por tCER e (ii) tempo de ciclo correspondente à intercepção das retas que representam os períodos CER e DC e denotado por tCER2. O parâmetro tCER é amplamente proposto pela literatura e como uma boa aproximação para o tempo de ciclo do processo SFE, tendo em vista que aproximadamente 70% do leito de extração é esgotado nesta etapa. Para as OECs ajustadas a um spline linear contendo três retas o tempo de ciclo foi selecionado como tCER2. Através do estudo de simulação da etapa de separação extrato/solvente realizado para os sistemas funcho + CO2 e canelinha + CO2 foi possível observar que a condição de operação desta etapa tem grande influência na composição do extrato final. A pressão de operação do tanque flash igual a 40 bar foi considerada uma boa alternativa, pois tanto a perda de CO2 na corrente de saída de extrato quanto a perda de extrato na corrente de saída de solvente foram mínimas. O estudo da etapa de separação é uma área promissora e ainda pouco explorada frente à etapa de extração. Foi avaliado o impacto do custo fixo de investimento, custo de matéria-prima, custo de utilidades, custo de mão-de-obra e custo de tratamento de resíduos sobre o COM. Este estudo foi importante, pois quebrou o paradigma de que o custo fixo de investimento seria o fator de maior impacto no COM dos extratos produzidos por SFE. O custo de matéria-prima predominou na composição do COM dos extratos da grande maioria das matrizes vegetais obtidas por SFE e SD. A SFE é um processo promissor e já existem inúmeras unidades industriais operando na China, França, Alemanha, etc. A estimativa preliminar dos custos de manufatura dos extratos de erva-baleeira, alecrim, camomila, erva-doce e pimenta-do-reino obtidos por SFE e SD foram: US$ 914,00/kg e US$ 10.350,00/kg, US$ 214,00/kg e US$ 369,00/kg, US$ 277,00/kg e US$ 2.798,00/kg, US$ 216,00/kg e US$ 272,00/kg, US$ 198,00/kg e US$ 232,00/kg, respectivamente. Estes custos foram comparados ao preço de venda de similares produzidos no Brasil e/ou exterior. A produtividade anual para cada processo foi confrontada / Abstract: This study had the objective to compare the cost of manufacturing (COM) of vegetable extracts and volatile oils of aromatic, condimental and medicinal plants, natives or adaptable to cultivation in Brazil, obtained by extraction processes that are clean and harmless to the environment. A technical-economical study on supercritical fluid extraction (SFE) and steam distillation (SD) for obtaining of vegetable extracts and volatile oils was done. The methodology to estimate the cost of manufacturing used on this study produces estimation classified as COM class 4 or 5, being a preliminary estimation used in the decision of go or not go of the project. The cost of manufacturing estimations were done using information from the overall extraction curves (OEC) experimentally determined in this study (rosemary + steam, chamomile + steam, anise + steam, C. verbenacea + steam, C. verbenacea + CO2, C. zehntneri + CO2) and information obtained from literature for OECs (rosemary + steam, rosemary + CO2, anise + steam, anise + CO2, ginger + CO2, chamomile + CO2, black pepper + steam, black pepper + CO2, fennel + CO2). The overall extraction curves were fitted to a spline containing two or three lines: the first line describes the mass transfer rate for the constant extraction rate period (CER), the second line describes the falling extraction rate period (FER), and the third line describes the diffusion controlled rate period (DC). The scale-up procedure assumed that both the global yield and the time of processing of the industrial scale are linearly correlated to laboratory scale ones as long as the ratio of the solvent mass to the raw material mass is kept constant. The COMs of the extracts obtained by SD and SFE were estimated from beginning to end of the process at specified time intervals along the OECs. For the SFE process the COM was also estimated for two preselected times of process: (i) a time of cycle corresponding to CER period and denoted as tCER, and (ii) a time of cycle corresponding to the intercept of the CER and DC lines denoted in this study as tCER2. The tCER parameter is widespread in literature as being a good estimative for the time of processing for the SFE since approximately 70% of extraction bed is exhausted along this period. For the OECs adjusted to a linear spline containing three lines the time of cycle was chosen as tCER2. The simulation of the separation extract/solvent done for fennel + CO2 and C. zehntneri + CO2 system, showed that the operation conditions of this stage have great influence on the final composition of the extract. The operation pressure of 40 bar in the flash tank was considered as a good choice, since a minimum loss of CO2 (in the extract exit stream) and the extract loss (in the CO2 exit stream) were estimated. The study of the separation stage is a promising area, and still little explored when compared to the extraction stage. The impact of the fixed cost of investment, raw material cost, utilities cost, operational-labor cost and waste treatment cost on the COM were evaluated. This study was important since it broke the paradigm that the fixed cost of investment would be the factor of major impact on the COM of the extracts produced by SFE. The raw material cost was predominant on the composition of the COM of the extracts from most vegetable matrices obtained by SFE and SD. SFE is a promising process, and there already are many industrial units operating in China, France, Germany, etc. The preliminary estimative of the manufacturing costs of the extracts of C. verbenacea, rosemary, chamomile, anise and black pepper obtained by SFE and SD were: US$ 914.00/kg and US$ 10,350.00/kg, US$ 214.00/kg and US$ 369.00/kg, US$ 277.00/kg and US$ 2,798.00/kg, US$ 216.00/kg and US$ 272.00/kg, US$ 198.00/kg and US$ 232.00/kg, respectively. These costs were compared to the selling prices of similar products produced in Brazil and/or abroad. The annual productivities of each process were confronted. / Doutorado / Engenharia de Alimentos / Doutor em Engenharia de Alimentos

Submitted by Marilene Donadel (marilene.donadel@unioeste.br) on 2017-10-25T23:42:57Z No. of bitstreams: 1 Daniela_C_O_Bueno_2016.pdf: 2802809 bytes, checksum: ff40fea77fdb8dbe45d71a56b942127c (MD5) / Made available in DSpace on 2017-10-25T23:42:57Z (GMT). No. of bitstreams: 1 Daniela_C_O_Bueno_2016.pdf: 2802809 bytes, checksum: ff40fea77fdb8dbe45d71a56b942127c (MD5) Previous issue date: 2016-02-29 / The trees are part of our daily life: in urban areas, as shadow production elements, landscaping, water infiltration, reduced thermal sensation, attenuation of noise, breaking wind, oxygen release / absorption of pollutants and as micro habitats for wildlife; and in rural areas, some cultures such as pine are distinguished primarily as energy sources (wood) in the industrial sector. In all cases the common characteristic is the generation of waste is the pruning process and / or processing / production of wood processing. In most cases these residues are intended to landfills, discarding the potential use as raw material for the production of essential oils. On that basis, the aim of this study was to develop a technology for essential oil extraction using an alternative method based on the incremental innovation. Therefore, the methodology included six stages: 1) characterization of the raw material; 2) Technology prospecting; 3) alternative technology proposition; 4) Prototyping through design, construction and assembly components; 5) Operational tests in real operating condition; and 6) the requirements of novelty evaluation, inventive step, sufficiency descriptive and industrial application for intellectual property protection. The result of the work was a technological product, pilot-scale production of essential oil residues from tree pruning and firewood production endowed simultaneously generating steam and carbon dioxide. The resulting equipment introduced innovative features and was the object of protection with the National Institute of Industrial Property (INPI) under No. 2015 032186-0 BR 20 with the following attributes: possibility of operation of the modular systems either independently or together, possibility of continuous operation or in batch, using as extracting fluid vapor and / or carbon dioxide and versatility in size and enforcement capacity in various materials that have, in their composition, essential oils. The work also had as a development technology for use in saunas which was also object patent (Patent No. BR 20 2015 032185-1). The project in question is inserted in the area of technological innovation and in its early stages prioritized the development of technology and in the second phase should be complemented with the optimization of the operational part. Thus, it contributes to environmental technology for the recovery of waste from pruning and / or industrial wood processing. / As árvores fazem parte do nosso cotidiano: na área urbana, como elementos de produção de sombra, paisagismo, infiltração da água, redução da sensação térmica, atenuação da poluição sonora, quebra de ventos, liberação de oxigênio/absorção de poluentes e como micro hábitats para a fauna; e na área rural, algumas culturas como o pinus primam essencialmente como fontes energéticas (lenha) do setor industrial. Em todos os casos a característica comum é a geração de resíduos, seja do processo de poda e/ou do processamento/beneficiamento na produção de lenha. Na maioria dos casos esses resíduos são destinados aos aterros sanitários, descartando-se o potencial do aproveitamento, como matéria prima para a produção de óleos essenciais. Partindo deste pressuposto, o objetivo deste trabalho foi desenvolver uma tecnologia para extração de óleo essencial utilizando um método alternativo com base da inovação incremental. Para tanto, a metodologia contemplou seis etapas: 1) Caracterização da matéria prima; 2) Prospecção tecnológica; 3) Proposição de tecnologia alternativa; 4) Prototipagem por meio de projeto, construção e montagem de componentes; 5) Testes operacionais na condição real de operação; e 6) Avaliação dos requisitos de novidade, atividade inventiva, suficiência descritiva e aplicação industrial para proteção intelectual. O resultado do trabalho foi um produto tecnológico, em escala piloto, para produção de óleo essencial de resíduos da poda de árvores e da produção de lenha dotado de gerador simultâneo de vapor e gás carbônico. O equipamento resultante apresentou características inovadoras e foi objeto de proteção junto ao Instituto Nacional de Propriedade Industrial (INPI) sob o Nº BR 20 2015 032186-0 com os seguintes atributos: possibilidade de operação dos seus sistemas modulares de maneira independente ou em conjunto, possibilidade de operação de forma contínua ou em batelada, utilizando como fluído extrator o vapor e/ou gás carbônico e versatilidade no seu tamanho e capacidade de aplicação em diversos materiais que possuem, na sua composição, óleos essenciais. O trabalho teve também como desdobramento uma tecnologia para utilização em saunas o qual também foi objeto de patenteamento (Patente Nº BR 20 2015 032185-1). O projeto em questão se encontra inserido na área de inovação tecnológica e na sua fase inicial priorizou o desenvolvimento da tecnologia e na segunda fase deve ser complementado com a otimização da parte operacional. Desta forma, contribui-se com a tecnologia ambiental para o aproveitamento dos resíduos de podas e/ou do beneficiamento de lenha industrial.

This thesis deals with the batch controlling of a distillation column model. The thesis is divided into several parts. The first part, which is more theoretical, summarizes the description of the batch controlling, the distillation process and the equipment used, including the model of the distillation column itself. The second part is focused on the practical implementation. The beginning of this part displays different types of the models produced by the Standard 88. This section is followed by a description of a device database in the environment of FactoryTalk Batch Equipment Editor, creation of a logical phase codes using functions of the PhaseManager in the environment of RSLogix5000 and a creation of a recipe in the environment of FactoryTalk Batch Recipe Editor. The last part of this thesis deals with the visualization created in the FactoryTalk View Studio SE and final testing of the entire process.

[EN] Anaerobic digestion is a sustainable and technically sound way to valorise citrus waste if the inhibitory effect of the citrus essential oil (CEO) is controlled. Several strategies have been proposed to overcome these difficulties: keeping the organic loading rate (OLR) in low values to avoid excess dosage of inhibitor, supplementing the citrus waste with nutrient and buffering solutions or pre-treating the citrus waste in order to reduce the CEO concentration, either by recovery or by degradation of the CEO. Nevertheless, although some of them have been proven successful in recovering/degrading the CEO, none of them has been applied at full scale operation. The main objective of this thesis is to study the effect of the limonene (the main component of CEO) on the anaerobic digestion of citrus waste and to evaluate different strategies to improve this process. In a first approach to the problem, the effect and dynamics of the limonene in the anaerobic digestion process is studied in batch mode. The biochemical methane potential of several citrus waste types was assessed. The inhibitory concentration of limonene for the anaerobic digestion process was estimated also, observing a certain adaptation degree. Different strategies to avoid inhibition of the anaerobic digestion by limonene were studied in batch mode, namely biological treatment by fungi of the Penicillium genus, steam distillation and ethanol extraction. All treatments decreased the limonene concentration in the orange peel, with different efficiencies. Methane potential and production rate in the batch anaerobic digestion of the pretreated orange peel were not affected by the biological treatment, but an increase was observed after steam distillation and also after ethanol extraction. This effect was attributed to the removal of minority compounds of the CEO. Energy balance was negative for steam distillation and positive for the other two tested strategies. Continuous anaerobic co-digestion experiments of orange and mandarin peel with chicken and pig manure allowed observing the importance of the limonene dosage on the inhibitory effect. Systemic inhibition was observed in the mixture with higher limonene concentration, with symptoms of inhibition on methanogenesis, protein hydrolysis pathway, sulphate reduction and acetogenesis. The degradation of the limonene produced inhibitory compounds as well, causing persistent inhibition effects even after almost complete limonene degradation. Continuous anaerobic co-digestion of pretreated citrus waste with cow manure allowed for stable processes when the pretreatment was able to remove the limonene with high efficiency and without producing other inhibitory compounds (such as ¿-terpineol in the biological treatment). Thus, the pretreatments allowing for better results in terms of process stability were mechanical removal of the flavedo and ethanol extraction of the limonene. The anaerobic digestion is able to degrade the limonene, but its by-products can be even more inhibitory than the limonene itself. Therefore, it is concluded that to apply CEO recovery strategies before anaerobic digestion is recommended, since these could possibly the valorization of CEO as added value product and to increase biogas production. / [ES] La digestión anaerobia es una vía sostenible y técnicamente viable para valorizar los residuos cítricos si se controla el efecto inhibidor del aceite esencial cítrico (AEC). Se han propuesto varias estrategias para superar esta dificultad: mantener la velocidad de carga orgánica (VCO) en valores bajos para evitar dosis excesivas del inhibidor, suplementar el residuo cítrico con soluciones de nutrientes y tampón o pretratar el residuo cítrico para reducir la concentración de AEC, ya sea por recuperación o degradación del mismo. Sin embargo, aunque algunas de ellas han sido exitosas para recuperar o degradar el AEC, ninguna ha sido llevada a escala industrial. El objetivo principal de esta tesis ha sido estudiar el efecto del limoneno (el componente principal del AEC) en la digestión anaerobia de los residuos cítricos y evaluar diferentes estrategias para mejorar este proceso. En una primera aproximación al problema, se estudió el efecto y la dinámica del limoneno en la digestión anaerobia en discontinuo. Se evaluó el potencial bioquímico de metano de varios tipos de residuo cítrico y se estimó la concentración inhibitoria del limoneno para el proceso de digestión anaerobia, observando un cierto grado de adpatación. Se estudiaron diferentes estrategias para evitar la inhibición de la digestión anaerobia por limoneno en modo discontinuo: tratamiento biológico mediante hongos del género Penicillium, arrastre de vapor y extracción con etanol. Todos los tratamientos disminuyeron la concentración de limoneno en la piel de naranja, con diferentes eficiencias. El potencial de metano y la velocidad de producción en la digestión anaerobia discontinua de la piel de naranja pretratada no se vieron afectados por el tratamiento biológico, pero se observó un incremento tras el arrastre de vapor y la extracción con etanol. Este efecto se atribuyó a la extracción de compuestos minoritarios del AEC. El balance energético fue negativo para el arrastre de vapor y positivo para las otras dos estrategias. Los experimentos de co-digestión anaerobia en continuo de piel de naranja y mandarina con gallinaza y purín porcino permitieron observar la importancia de la dosis de limoneno sobre el efecto inhibitorio. Se observó inhibición sistémica en la mezcla con mayor concentración de limoneno. La degradación del limoneno produjo a su vez compuestos inhibitorios, con lo que el efecto inhibitorio persistió incluso tras la completa degradación del limoneno. La co-digestión anaerobia en continuo de residuo cítrico pretratado y estiércol de vacuno permitió tener procesos estables cuando el pretratamiento era capaz de eliminar el limoneno con alta eficiencia y sin producir otros compuestos inhibitorios (como ¿-terpineol en el caso del tratamiento biológico). Así, la eliminación mecánica del flavedo y la extracción del limoneno con etanol fueron los pretratamientos que dieron mejor resultado en términos de estabilidad del proceso. La digestión anaerobia es capaz de degradar el limoneno, pero sus subproductos pueden ser aún más inhibitorios que el propio limoneno. Por ello, se concluye que son recomendables los pretratamientos de recuperación de AEC, ya que permiten su valorización como producto de valor añadido y, a la vez, incrementar la producción de biogás mediante digestión anaerobia. / [CAT] La digestió anaeròbia és una via sostenible i tècnicament viable per valoritzar els residus cítrics si es controla l'efecte inhibitori de l'oli essencial cítric (OEC). S'han proposat diverses estratègies per tal de superar aquesta dificultat: mantenir la velocitat de càrrega orgànica (VCO) en valors baixos per a evitar dosis excessives de l'inhibidor, suplementar el residu cítric amb solucions de nutrients i tamponants o pretratar el residu cítric per a reduir la concentració d'OEC, ja siga per recuperació o per degradació del mateix. Tanmateix, encara que algunes d'elles han estat exitoses per recuperar o degradar l'OEC, cap ha estat escalada a nivel industrial. L'objectiu principal d'aquesta tesi ha estat estudiar l'efecte del limonè (el component principal de l'OEC) en la digestió anaeròbia dels residus cítrics i avaluar diferents estratègies per a millorar aquest procés. En una primera aproximació al problema, es va estudiar l'efecte i la dinàmica del limonè a la digestió anaeròbia en discontinu. Es va avaluar el potencial bioquímic de metà de diferents tipus de residu cítric i es va estimar la concentració inhibitòria del limonè per al procés de digestió anaeròbia, observant-se un cert grau d'adaptació. S'estudiaren diferents estratègies per a evitar la inhibició de la digestió anaeròbia per limonè en modus discontinu: tractament biològic amb fongs del gènere Penicillium, destil·lació amb vapor i extracció amb etanol. Tots els tractaments reduiren la concentració de limonè a la pell de taronja, però amb eficiències diferents. El pre-tractament biològic no va influir en el potencial de metà ni en la velocitat de producció de la digestió anaeròbia discontínua de la pell de taronja. En canvi, s'observà un increment d'ambdós paràmetres amb la destil·lació amb vapor i l'extracció amb etanol. Aquest efecte es va atribuir a l'extracció de components minoritaris de l'OEC. El balanç energètic va ser negatiu per a la destil·lació amb vapor i positiu per a les altres dues estratègies. Els experiments de co-digestió anaeròbia en continu de pell de taronja i mandarina, amb gallinassa i purins de porc varen permetre observar la importància de la dosi de limonè sobre l'efecte inhibitori. Es va observar inhibició sistèmica a la mescla amb major concentració de limonè. La degradació del limonè va produir inhibidors, de manera que l'efecte inhibitori va persistir fins i tot després de la completa degradació del limonè. La co-digestió anaeròbia en continu de residu cítric pretractat i fem de boví va permetre tenir processos estables quan el pretractament era capaç d'eliminar el limonè amb alta eficiència i sense produir altres molècules inhibitòries (com α-terpineol en el cas del tractament biològic). Així, l'eliminació mecànica de l'epicarpi i l'extracció del limonè amb etanol foren els pretractaments que donaren millor resultat en termes d'estabilitat del procés. La digestió anaeròbia és capaç de degradar el limonè, però els seus subproductes poden ser encara més inhibitoris que el propi limonè. Per això, es conclou que son recomanables els pretractaments de recuperació de OEC, ja que permenten la seva valorització com productes de valor afegit i, a la vegada, incrementar la producció de biogàs mitjançant digestió anaeròbia. / Ruiz Fuertes, B. (2015). Effect of limonene on anaerobic digestion of citrus waste and pretreatments for its improvement [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/58769 / TESIS

Les régions de Midi-Pyrénées (France) et de Chongqing (Chine) sont riches en plantes aromatiques et médicinales dites oubliées (ou médiévales). Afin de valoriser pleinement les différentes bio-molécules extractibles de ces plantes, le concept de MAP-raffinerie a été créé et appliqué à une sélection de plantes issues de ces deux régions. Plusieurs technologies d’extraction utilisant l’eau comme solvant vert (hydrodistillation, distillation à la vapeur et extraction par eau sub-critique) ont ainsi été employées et leur impact tant sur la composition des huiles essentielles que sur la récupération des molécules anti-oxydantes a été évalué. Dans un premier temps, une liste de plantes aromatiques et médicinales oubliées, voire sous-utilisées dans les deux régions a été établie selon des règles de sélection prédéfinies. Six plantes modèles de la région de Midi-Pyrénées (Tussilago farfara L., Calendula arvensis L., Robinia pseudoacacia L., Geranium robertianum L., Cytisus scoparius L. et Spartium junceum L.) et trois plantes de la région de Chongqing (Tussilago farfara L., Citrus aurantium L. et Saussurea costus) ont finalement été retenues. Puis, le concept de MAP-raffinerie a été appliqué à ces plantes afin d’étudier leur possible valorisation globale. L’étude des compositions chimiques des extraits volatils des racines de Tussilago farfara L. et de Calendula arvensis L., ainsi que des boutons de fleurs de Spartium junceum L. a été réalisée par GC et GC-MS pour la première fois. Les principaux composés chimiques dans l’extrait volatil de racines de Tussilago farfara L. étaient des hydrocarbures sesquiterpéniques et des composés aliphatiques tandis que les principaux composés chimiques dans l’extrait volatil de racines de Calendula arvensis L. étaient des sesquiterpènes oxygénés, des monoterpènes oxygénés et des diterpènes oxygénés. L’extrait volatil de boutons de fleurs de Spartium junceum L. était principalement composé de composés aliphatiques. Par ailleurs, les résultats de l’évaluation des capacités anti-oxydantes des extraits (par les tests DPPH, ABTS, FRAP, ORAC et Folin-Ciocalteu) ont montrés que plusieurs plantes comme Cytisus scoparius L., Tussilago farfara L., Citrus aurantium L. ou Robinia pseudoacacia L. pourraient être des sources potentielles d’anti-oxydants naturels. D’un point de vue technologique, les comparaisons de l’utilisation de l’hydrodistillation (HD), de la distillation à la vapeur (SD) et de l’extraction par eau sub-critique (SWE) ont montrées que si la HD et la SD ont des effets limités sur la composition des huiles essentielles, la HD semble être une méthode plus efficace pour la récupération des composés anti-oxydants à partir des résidus de distillation que la SD tandis que la SWE s’avère être une technologie prometteuse pour l’extraction directe de ces molécules à partir des plantes. Si la composition minérale de l’eau lors de l’hydrodistillation n’a que des effets très limités sur les rendements d’extraction, les teneurs en ions calcium et bicarbonate des eaux ont par contre des effets décroissants significatifs sur la capacité anti-oxydante et sur la teneur phénolique totale des extraits aqueux et méthanoliques. Au vue de ces résultats, un concept amélioré de MAP-raffinerie a été développé en intégrant une extraction à l’eau sub-critique pour l’extraction des composés anti-oxydants des résidus d’extraction primaire. Selon ce nouveau concept, cinq extraits peuvent être obtenus à partir des matières végétales: un extrait volatil, un extrait aqueux, un extrait méthanolique, un extrait à l’eau sub-critique et in fine un résidu solide. Les premiers résultats ont montrés que la "MAP-raffinerie améliorée" augmente de manière significative la récupération des antioxydants par rapport à la MAP-raffinerie originale et permet d’envisager une valorisation plus facile du résidu solide en agro-matériaux du fait de sa faible teneur en eau résiduelle. / In both Midi-Pyrénées region (France) and Chongqing region (China), there are rich and underutilized medicinal and aromatic plants (MAP). Aiming at fully exploiting different molecules in these plants, the concept of MAP-refinery was developed and applied to several underutilized medicinal and aromatic plants in these two regions. Several water-based green extraction technologies of natural products (e.g. hydrodistillation, steam distillation and subcritical water extraction) were also investigated to look at their effects on essential oil composition and antioxidants recovery from selected plants. Firstly, lists of forgotten or underutilized medicinal and aromatic plants in both regions were established according to the rules of selection. From the lists, six plants in the Midi-Pyrénées region (Tussilago farfara L., Calendula arvensis L., Robinia pseudoacacia L., Geranium robertianum L., Cytisus scoparius L. and Spartium junceum L.) and three plants in the Chongqing region (Tussilago farfara L., Citrum aurantium L. and Saussurea costus) were finally selected for investigations. Then the MAP-refinery was applied to the selected plants in two regions in order to realise their global valorisation. Volatile extracts composition in the roots of Tussilago farfara L. and Calendula arvensis L., as well as flower buds of Spartium junceum L. were firstly investigated. The main chemical compounds in volatile extract from Tussilago farfara L. roots were sesquiterpene hydrocarbons and aliphatic compounds while main chemical compounds in volatile extract from Calendula arvensis L. roots were oxygenated sesquiterpenes, oxygenated monoterpenes and oxygenated diterpenes. The volatile extract from flower buds of Spartium junceum L. was mainly composed of aliphatic compounds. Antioxidant capacity evaluation results (by DPPH, ABTS, FRAC, ORAC and Folin-Ciocalteu tests) showed that several plant samples like Cytisus scoparius L., Tussilago farfara L., Citrum aurantium L. and Robinia pseudoacacia L. could be potential sources of natural antioxidants. Comparisons of hydrodistillation (HD), steam distillation (SD) and subcritical water extraction (SWE) showed that HD and SD had limited effects on essential oil composition but HD, SD and SWE had significant impacts on the recovery of antioxidants. Hydrodistillation seemed to be a better method for recovery of antioxidant compounds from residues of distillation than steam distillation. However, SWE appeared to be a more efficient method for direct extraction of antioxidant molecules (or phenolic compounds) from plants. In the hydrodistillation process, mineral contents in water were found to have very limited effects on yields of extracts but calcium and bicarbonate ions, had significant decreasing effects on antioxidant capacity and total phenolic content of both aqueous and methanolic extracts. Finally, an improved MAP-refinery was developed. Subcritical water was used for further extraction of antioxidant compounds from residues in original MAP-refinery. In this way, five parts could be obtained from plant materials: volatile extract, aqueous extract, methanolic extract, subcritical water extract and the final residue. The results showed that the improved MAP-refinery significantly increased the recovery of antioxidants compared with original MAP-refinery. This promising process will also allow a better valorisation of the final solid residue due to the lower content of residual water.