Orientador: Helenice de Oliveira Florentino Silva / Resumo: A escassez de combustíveis fósseis e a demanda por fontes alternativas de energia renovável e limpa são impulsionadores para o desenvolvimento de biocombustíveis, tais como o biohidrogênio. Este gás é conhecido por seu alto valor calorífico, extrema leveza e baixa densidade, além de, ao ser queimado, produzir apenas vapor d'água e calor. Dentre os modos de produção, destaca-se a fermentação escura, a qual gera biohidrogênio e subprodutos através do tratamento microbiológico de resíduos agroindustriais. O objetivo deste trabalho foi modelar matematicamente este bioprocesso, estudar suas propriedades à luz da teoria de estabilidade, além de propor um modelo de otimização que determine uma combinação das concentrações de substrato e bactérias, tais que maximizem o rendimento da produção de biohidrogênio. Foram propostas duas heurísticas para a resolução do modelo, Algoritmo de Busca em Vizinhança Variável e Algoritmo Memético. Os resultados das simulações numéricas mostraram que o modelo obtido corrobora com a dinâmica bioquímica e microbiológica do bioprocesso. / Abstract: The lack of fossil fuels and the demand of alternative, renewable and clean energy sources promote development in biofuels, as biohydrogen. It is known by its high heat, extreme lightness and low density and also when it burns, the products are only steam and energy. Among the productions ways, we highlight dark fermentation, which generates biohydrogen and subproducts through organic waste microbiological treatment. The aim of this work was to mathematical model this bioprocess, to study its properties via stability analysis, besides to propose a optmization model to determine a combination of substract and bacteria concentrations, in order to maximize biohydrogen production yield. To solve the mathematical model have been proposed two heuristics: Variable Neighboor Search Algorithm and Memetic Algorithm. Numeric simulations showed that the mathematical model corroborates with the bioprocess microbial and biochemical dynamics. / Mestre

La fermentation sombre est un procédé utilisant des déchets organiques dont le passage à l'échelle pilote est limité par les rendements de production d'hydrogène trop faibles ainsi que par l'utilisation des sous-produits. Cette étude a pour premier objectif d'étudier l'effet du pH, de la combinaison du pH et de la concentration en substrat, du prétraitement du substrat et de l'adaptation de l'inoculum sur la fermentation sombre de trois types de déchet différents. Il a notamment été montré que la biodégradabilité des substrats joue un rôle majeur dans le choix des paramètres opérationnels utilisés pour optimiser la production d'hydrogène. De plus, la faisabilité et la stabilité à long terme de la production d'hydrogène par le procédé de fermentation sombre ont été mises en évidence en utilisant des déchets agroalimentaires et du petit lait dans deux réacteurs thermophiliques fonctionnant en mode semi-continu. En particulier, il a été discuté l'influence de la charge organique (OLR), du temps de rétention hydraulique (HRT) et de l'addition de co-substrats (fumier de buffle) comme source d'alcalinité. Cette étude a montré que la combinaison de ces trois paramètres peut jouer un rôle important sur le pH et la stabilité de la production d'hydrogène. De plus, les sous-produits de la fermentation sombre ont été utilisés pour produire de l'hydrogène via la photo-fermentation, alors que les déchets générés par le couplage de la fermentation sombre et de la photo-fermentation ont été valorisés pour la production de méthane par digestion anaérobie. Ce concept de bioraffinerie basé sur la conversion en trois étapes des déchets agroalimentaires augmente le rendement énergétique global du procédé. Par ailleurs, il a été montré le potentiel important du procédé de photo-fermentation pour la production de polyhydroxybutyrate (polymère), parallèlement à celle d'hydrogène. De même, l'utilisation de la fermentation par voie sèche dans une bioraffinerie concept apparaît prometteuse vis à vis de la production de bioénergie et de molécules telles que les acides organiques et les alcools / Low biohydrogen (H2) yields and use of process by-products from dark fermentation (DF) of waste biomass is limiting its scaled-up application. This study aims to investigate the effects of culture pH, combination of substrate concentration and culture pH, pre-treatment of substrate and inoculum adaptation in H2 yields during the DF of three different wastes biomass. The study showed that the biodegradability of the substrates is important for the selection and application of optimum operational parameters aimed at enhancing H2 production. Moreover, long-term operational feasibility and stability of dark fermentative H2 production was demostrated using food waste and cheese whey in two semi-continuous thermophilic DF reactors. The effect of organic loading rates (OLRs), hydraulic retention times (HRTs) and co-substrates (buffalo manure) addition as a source of alkalinity on culture pH and H2 production stability was discussed. The study showed that combination of OLR, HRT and co-substrate addition could play an important role in the culture pH and stability of H2 production. Furthermore, the by-products of DF process was utilized for H2 production via photo fermentation (PF), while the waste stream generated from coupling of DF and PF processes was converted to methane in anaerobic digestion (AD). The three-step conversion of food waste in a biorefinery concept increased the total energy yields. Moreover, PF also showed a good potential for concomitant production of H2 and polyhydroxybutyrate (biopolymer). Likewise, dry fermentation could be promising to a biorefinery concept based on waste biomass for the production of bioenergy and biochemicals (organic acids and alcohols)

Application of bioengineering technologies for enhanced biological hydrogen production is a promising approach that may play a vital role in sustainable energy. Due to the ability of several naturally occurring microorganisms to generate hydrogen through varying metabolic processes, biological hydrogen has become an attractive alternative energy and fuel source. One area of particular interest is the production of biological hydrogen in organically-rich engineered systems, such as those associated with waste treatment. Despite the potential for high energy yields, hydrogen yields generated by bacteria in waste systems are often limited due to a focus on microbial utilization of organic material towards cellular growth rather than production of biogas. To address this concern and to improve upon current technological applications, metabolic engineering approaches may be applied to known hydrogen producing organisms. However, to successfully modify metabolic pathways, full understanding of metabolic networks involved in expression of microbial traits in hydrogen producing organisms is necessary. Because microbial communities associated with hydrogen production are capable of exhibiting a number of phenotypes, attempts to apply metabolic engineering concepts have been restricted due to limited information regarding complex metabolic processes and regulatory networks involved in expression of microbial traits associated with biohydrogen production. To bridge this gap, this dissertation focuses on identification of phenotype-related biochemical processes within sets of phenotype-expressing organisms. Specifically, through co-development and application of evolutionary genome-scale phenotype-centric comparative network analysis tools, metabolic and cellular components related to three phenotypes (i.e., dark fermentative, hydrogen production and acid tolerance) were identified. The computational tools employed for the systematic elucidation of key phenotype-related genes and subsystems consisted of two complementary methods. The first method, the Network Instance-Based Biased Subgraph Search (NIBBS) algorithm, identified phenotype-related metabolic genes and subsystems through comparative analysis of multiple genome-scale metabolic networks. The second method was the multiple alignments of metabolic pathways for identification of conserved metabolic sub-systems in small sets of phenotype-expressing microorganisms. For both methodologies, key metabolic genes and sub-systems that are likely to be related to hydrogen production and acid-tolerance were identified and hypotheses regarding their role in phenotype expression were generated. In addition, analysis of hydrogen producing enzymes generated by NIBBS revealed the potential interplay, or cross-talk, between metabolic pathways. To identify phenotype-related subnetworks, three complementary approaches were applied to individual, and sets of phenotype-expressing microorganisms. In the first method, the Dense ENriched Subgraph Enumeration (DENSE) algorithm, partial "prior knowledge" about the proteins involved in phenotype-related processes are utilized to identify dense, enriched sets of known phenotype-related proteins in Clostridium acetobutylicum. The second approach utilized a bi-clustering algorithm to identify phenotype-related functional association modules associated with metabolic controls of phenotype-related pathways. Last, through comparison of hundreds of genome-scale networks of functionally associated proteins, the á, â-motifs approach, was applied to identify phenotype-related subsystems. Application of methodologies for identification of subnetworks resulted in detection of regulatory proteins, transporters, and signaling proteins predicted to be related to phenotype-expression. Through analysis of protein interactions, clues to the functional roles and associations of previously uncharacterized proteins were identified (DENSE) and hypotheses regarding potentially important acid-tolerant mechanisms were generated (á, â-motifs). Similar to the NIBBS algorithm, analysis of functional modules predicted by the bi-clustering algorithm suggest cross-talk is occurring between pathways associated with hydrogen production. The ability of these phenotype-centric comparative network analysis tools to identify both known and potentially new biochemical process is important for providing further understanding and insights into metabolic networks and system controls involved in the expression of microbial traits. In particular, identification of phenotype-related metabolic components through a systems approach provides the underlying foundation for the development of improved bioengineering technologies and experimental design for enhanced biological hydrogen production.

Dans un contexte de transition énergétique, ce travail de recherche s’inscrit dans la dynamiqued’explorer la potentialité de nouvelles biomasses pour la production d’un vecteur énergétique propre,l’hydrogène (H2). Au cours de cette thèse, un procédé de production d’H2 par fermentation endogèneet sans dépense énergétique supplémentaire, a été élaboré à partir de biomasses viticoles, etl’influence de paramètres opératoires a été testée. Enfin, le type de biomasses influe sur laproduction d’H2, non seulement par sa composition biochimique, mais aussi par leur microfloreendogène qui oriente le métabolisme. La microflore endogène responsable de la production d’H2 àpartir de ‘bourbes’ (50 ± 6 L/Lbourbes avec un rendement de 2,0 ± 0,2 mol/mol) appartient à la familleClostridiaceae et est associée aux voies acétate/butyrate, alors que celle des ‘marcs’ (20 ± 4L/kgmarcs avec un rendement de 1,3 ± 0,3 mol/mol) appartient aux Enterobacteriaceae et estassociée aux voies acétate/éthanol. / In the context of energy transition, this work deals with the exploration of new potential biomasses forthe production of a clean energy vector, hydrogen (H2). During this thesis, a process of H2 productionby an endogenous fermentation from winery waste has been developed without any additionalenergetic consumption. The influence of operating parameters was studied. Finally, the type ofbiomass has an influence on the H2 production, not only because of initial biochemical composition,but also because of the endogenous microflora which orientates the metabolism. The endogenousmicroflora responsible of H2 production from ‘solid grape residues’ biomass (50 ± 6 L.L-1biomass with a2.0 ± 0.2 mol.mol-1 yield) through acetate/butyrate pathways belongs to Clostridiaceae family whilethe one responsible of H2 production from ‘grape pomace’ biomass (20 ± 4 L.kg-1biomass with a 1.3 ±0.3 mol.mol-1 yield) through acetate/ethanol pathways belongs to Enterobacteriaceae family.

L'hydrogène a révélé un grand potentiel en tant que vecteur d'énergie du futur polyvalent et non polluant, offrant une densité d'énergie élevée et une conversion efficace en puissance utilisable. La fermentation noire est l'un des procédés de production biologique les plus prometteurs, mais doit encore surmonter des défis majeurs, notamment des taux de production d'hydrogène faibles (HPR) et des rendements en hydrogène (HY), avant que son application industrielle ne devienne économe en énergie et en coûts. Dans ce travail, nous avons cherché à optimiser la production d’hydrogène par fermentation noire de Thermotoga neapolitana. Les principaux objectifs étaient d'améliorer le HPR et de maintenir une HY élevée en utilisant différentes approches pour contrecarrer les limitations des processus et prévenir les inhibitions les plus pertinentes. En outre, un développement du procédé à flux continu privilégié par l'industrie a été prévu. Une augmentation de la concentration de biomasse initiale de 0,46 à 1,74 g CDW / L dans les essais biologiques en lots a entraîné une augmentation de plus de 2 fois de la HPR jusqu'à 654 (± 30) mL / L / h sans affecter négativement l'HY. Cependant, alors que la productivité volumétrique augmentait, le HPR spécifique (par unité de biomasse) était négativement corrélé avec le HPR et la concentration de biomasse. Par la suite, nous avons étudié la sursaturation en hydrogène de la phase liquide (H2aq) dans des essais biologiques par lots. À 100 tr / min d'agitation, H2aq est sursaturé jusqu'à 3 fois la concentration à l'équilibre. L'augmentation de la vitesse d'agitation diminuait l'accumulation de H2aq jusqu'à atteindre un équilibre entre l'hydrogène en phase gazeuse et liquide en agitation à 500 tr / min à de faibles concentrations cellulaires. Une augmentation de 200 à 600 tr / min a réduit progressivement l'H2aq de 21,9 (± 2,2) à 8,5 (± 0,1) mL / L et a presque doublé le HPR, révélant une corrélation directe entre les deux paramètres. De même, l’ajout de supports K1 et de recirculation de biogaz riche en H2 (GaR) a permis de contrecarrer l’accumulation de H2aq. En accélérant le processus en augmentant la concentration de biomasse dans les réacteurs jusqu'à 0,79 g CDW / L, le GaR s'est révélé plus efficace pour éliminer l'H2aq que l'agitation à 500 tr / min. L'application de GaR à 300 et 500 tr / min a augmenté le HPR d'environ 260% à 850 (± 71) mL H2 / h / L, par rapport à une agitation à 300 tr / min, atteignant une HY de 3,5 mol / mol de glucose. Nous avons démontré qu'un transfert de masse gaz-liquide insuffisant conduit à une accumulation de H2aq qui inhibe le rendement, mais plus encore le taux de fermentation à l'obscurité. Dans la phase finale de ce projet, nous avons réussi à maintenir la production d'hydrogène à débit continu. L'augmentation de la concentration en glucose de l'alimentation, de 11,1 à 41,6 mM, a entraîné une diminution de l'HY de 3,6 (± 0,1) à 1,4 (± 0,1) mol H2 / mol de glucose. L’HPR a augmenté simultanément jusqu’à environ 55 ml / L / h à 27,8 mM de glucose, tandis qu’une augmentation supplémentaire du glucose dans l’alimentation animale à 41,6 mM n’a pas augmenté les concentrations en HPR et en AA. Pour rechercher si des taux élevés d'AA limitaient le processus, la concentration en AA de l'alimentation a été progressivement augmentée. Cependant, cela n'a révélé aucun effet négatif sur la fermentation à l'obscurité en continu jusqu'à 240 mM de nourriture AA et, tout au long des 110 jours de fermentation en continu, l'HY a augmenté de 47%. La réduction du THS de 24 à 7 h a également entraîné une augmentation du taux de HPR de 82 (± 1) à 192 (± 4) mL / L / h, tout en diminuant le HY. De manière concomitante, le H2aq accumulé est directement corrélé au HPR atteignant 15,6 mL / L pour un THS de 7 h et à une agitation de 500 tr / min. L'application de GaR a efficacement neutralisé la sursaturation en H2aq et a permis d'obtenir le HPR le plus élevé de 277 mL / L à un THS de 5 h / Hydrogen has revealed a great potential as a versatile and non-polluting energy carrier of the future providing a high energy density and an efficiently conversion to usable power. Dark fermentation is one of the most promising biological production processes, but still has to overcome major challenges, most importantly low hydrogen production rates (HPRs) and hydrogen yields (HYs), before its industrial application becomes cost- and energy-efficient.In this work, we aimed to optimize the hydrogen production via dark fermentation by Thermotoga neapolitana. The main objectives were to enhance the HPR and maintaining a high HY using different approaches to counteract process limitations and prevent the most relevant inhibitions. Furthermore, a development of the industrially preferred continuous-flow process was projected. An increase of the initial biomass concentration from 0.46 to 1.74 g CDW/L in batch bioassays resulted in a more than 2-fold enhancement of the HPR up to 654 (±30) mL/L/h without negatively affecting the HY. However, while the volumetric productivity increased the specific HPR (per unit of biomass) was negatively correlated with the HPR and the biomass concentration. Subsequently, we investigated the supersaturation of hydrogen in the liquid phase (H2aq) in batch bioassays. At 100 rpm agitation H2aq supersaturated up to 3 times the equilibrium concentration. Increasing the agitation speed diminished the accumulation of H2aq until an equilibrium between the gas and liquid phase hydrogen was reached with 500 rpm agitation at low cell concentrations. A raise from 200 to 600 rpm gradually reduced H2aq from 21.9 (± 2.2) to 8.5 (± 0.1) mL/L and approximately doubled the HPR, revealing a direct correlation between the two parameters. Similarly, the addition of K1 carrier and H2-rich biogas recirculation (GaR) successfully counteracted the accumulation of H2aq. Accelerating the process by increasing the reactors biomass concentration up to 0.79 g CDW/L, GaR revealed to be more efficient in removing H2aq than 500 rpm agitation. The application of GaR at 300 and 500 rpm enhanced the HPR by approximately 260% up to 850 (± 71) mL H2/h/L, compared to a sole 300 rpm agitation, reaching a HY of 3.5 mol/mol glucose. We demonstrated that an insufficient gas-liquid mass transfer leads to the accumulation of H2aq which inhibits the yield but even more so the rate of dark fermentation. In the final phase of this project we successfully maintained continuous-flow hydrogen production. Increasing the feed glucose concentration from 11.1 to 41.6 mM diminished the HY from 3.6 (± 0.1) to 1.4 (± 0.1) mol H2/mol glucose. The HPR increased concomitantly up to approximately 55 mL/L/h at 27.8 mM of glucose, whereas a further increase of feed glucose to 41.6 mM did not enhance the HPR and the AA concentration. To investigate whether high levels of AA limited the process, the feed AA concentration was gradually increased. However, this revealed no negative effect on continuous dark fermentation up to 240 mM of feed AA and, throughout the 110 days of continuous fermentation, the HY increased by 47%. Decreasing the HRT from 24 to 7 h also led to a HPR enhancement from 82 (± 1) to 192 (± 4) mL/L/h, while decreasing the HY. Concomitantly, the H2aq accumulated, directly correlated to the HPR reaching 15.6 mL/L at an HRT of 7 h and 500 rpm agitation. The application of GaR efficiently counteracted the supersaturation of H2aq and allowed the highest HPR of 277 mL/L at a HRT of 5 h

Hydrogen is re-emerging as a serious alternative to fossil fuels. It is a clean gas with high energy density and its combustion only generates water vapour. Nevertheless, the hydrogen industry has a significant carbon footprint since this gas is mostly derived from fossil fuels reforming processes. This project focusses on the development of sustainable alternatives to conventional hydrogen production, in which approaches based on dark fermentation (DF) using an inexpensive residue from the brewery industry as primary feedstock are presented. Firstly, a fungal pre-treatment (FT) was proposed to degrade a high-strength brewery waste slurry (BWS) to obtain an effluent with a lower concentration of chemical oxygen demand (COD) but rich in readily fermentable sugars for the ensuing DF, thus improving hydrogen yields (HY). Secondly, microbial electrolysis and fuel cells (MECs and MFCs) were proposed to assist DF, generating electricity in MFCs while improving HY by MECs. Coupling both microbial electrochemical technologies sequentially after DF did not show any advantage. However, promising results were obtained for electricity and hydrogen production when taking a single-staged approach. Treating BWS directly by MFCs produced 2.0 watts/g COD consumed, while the DF process assisted simultaneously by MECs (DF/MEC) produced 1.6 times more hydrogen than DF alone. An average HY of 2.32 ± 0.06 mol H₂/mol glucose was attained between both DF/MEC and DF after FT, hence approaching the theoretical value of 2.4 mol H₂/mol glucose, representing roughly a 50% improvement compared to DF alone. With an overall COD reduction above 76%, the DF after FT exhibited the highest energy conversion rate per substrate consumed (6.3 kJ/g COD). As valuable by-products obtained, up to 31 g/L of fungal biomass, which is appreciated in many state-of-the-art biomaterials applications, was produced by using BWS. While in the DF/MEC process, 18 g/L of butyric acid were generated, which is three times more than with DF alone. Butyric acid being the precursor to butanol and building block of biodegradable thermoplastics, this result is not without significance. The proposed approaches not only valorize BWS but also validate their economic and environmental attractiveness as promising alternative hydrogen production methods.

Cette étude vise à étudier l'effet de la composition de substrats organiques solides sur les performances de production d'hydrogène, les voies métaboliques associées et les changements des communautés microbiennes dans un réacteur discontinu (sCSTR). L'hydrogène est un vecteur énergétique idéal qui a gagné en intérêt scientifique au cours de la dernière décennie. L'H2 produit par voie biologique, ou biohydrogène, peut être produit par des procédés de fermentation sombre où les déchets organiques sont traités et avec la production de molécules à haute valeur ajoutée. Cependant, l'effet de la composition des déchets organiques solides sur la production de biohydrogène dans la fermentation sombre n'a pas encore été clairement élucidé. Au cours de cette étude, une revue bibliographique a été réalisée sur la production d'hydrogène à partir de déchets agricoles. Cette revue montre qu'une large gamme de performances en hydrogène peut être observée principalement en raison de la variabilité dans les compositions en même type de substrats et des conditions expérimentales appliquées. Après avoir optimisé un protocole de test de potentiel biohydrogène (BHP), une grande variété de substrats organiques solides visant à couvrir un grand panel de déchets a été testée pour fournir des données comparables à analyser. Les résultats d'une régression PLS ont montré que seuls les sucres solubles ou facilement disponibles éteint corrélaient avec la production d'hydrogène. En outre, les rendements d'hydrogène corrélaient aussi bien avec l'accumulation de butyrate, principale voie productrice de bioH2. Un modèle prédictif du rendement en hydrogène en fonction de la teneur en sucres a été proposé. Ensuite, des expériences ont été menées en réacteur semi-continu (sCSTR) avec le topinambour comme substrat solide. Il a été montré qu'une faible charge organique favorisait une production continue d'hydrogène tandis que l'accroissement de la charge organique introduisait la présence de voies concurrentes à la production d'hydrogène. De plus, les profils des empreintes moléculaires basées sur l'ADNr 16s ont montré que l'augmentation de la charge organique avait un impact significatif sur la diversité microbienne en favorisant l'implantation de microorganismes ne produisant pas d'hydrogène tels que des bactéries lactiques. / This study aims to investigate the effect of solid substrates composition on hydrogen production performances, metabolic pathways and microbial community changes in batch reactor and their dynamics in semi continuous reactors (sCSTR). Hydrogen is an ideal energy carrier which has gained scientific interest over the past decade. Biological H2, so-called biohydrogen, can especially be produced by dark fermentation processes concomitantly with value-added molecules (i.e. metabolic end-products), while organic waste is treated. However, the effect of solid organic waste composition on biohydrogen production in dark fermentation has not yet been clearly elucidated. In this study, a bibliographic review was made on hydrogen production from agricultural waste. This survey on literature showed that diverse performances were reported on hydrogen production due to the variability in substrate compositions and experimental conditions. After having optimized a protocol of biohydrogen potential test (BHP), a wide variety of organic solid substrates aiming to covering a large range of solid waste was tested to provide a comparable data analysis. The results of a PLS regression showed that only soluble carbohydrates or easily available carbohydrates correlated with hydrogen production. Furthermore, hydrogen yields correlated as well with butyrate H2-producing pathway which is consistent with the literature knowledge. A predictive model of hydrogen yield according to carbohydrate content was proposed. Then, experiments were carried out in sCSTR with Jerusalem artichoke tubers as a case study. It was shown that low organic loading rate favored continuous hydrogen production while higher organic loading introduced hydrogen competition pathways and decreased the overall hydrogen yields. Moereover, 16S rRNA gene based CE-SSCP profiles showed that increasing OLR had a significant effect on the microbial diversity by favoring the implementation of microorganisms not producing hydrogen, i.e. lactic acid bacteria.

L'hydrogène, qui constitue une solution alternative et durable à l’usage d’énergies fossiles, est produit essentiellement par reformage de combustibles fossiles (95%). Des filières de production plus soucieuses de l'environnement sont envisagées. Deux familles de technologies sont explorées: 1) par décomposition thermochimique ou électrochimique de l'eau et 2) à partir de différentes sources de biomasse. Parmi celles-ci, les cellules d'électrolyse microbienne ou «Microbial electrolysis cell (MEC)» permettent de produire de l'hydrogène par électrolyse de la matière organique. Une MEC consiste en une cathode classique qui assure la production d'hydrogène par la réduction électrochimique de l'eau, associée à une bioanode qui oxyde des substrats organiques en dioxyde de carbone. Ce processus d'oxydation n'est possible que grâce au développement sur l'anode d'un biofilm microbien électroactif qui joue le rôle d'électro-catalyseur. Par rapport aux procédés courants d'électrolyse de l'eau, une MEC requière un apport énergétique 5 à 10 fois plus faibles. En outre, les procédés « classiques » de production de bio-hydrogène par voie fermentaire en cultures mixtes convertissent des sucres avec des rendements limités à 2-3 moles d'hydrogène par mole d'hexose tout en coproduisant des acides organiques. Alimenté par de l'acétate, une MEC produit au maximum 3 moles d'hydrogène/mole d'acétate. Le couplage de la fermentation à un procédé d'électrolyse microbienne pourrait donc produire de 8 à 9 moles d'hydrogène/mole d'hexose, soit un grand pas vers la limite théorique de 12 moles d'hydrogène/mole d'hexose. L'objectif de cette thèse est d'analyser les liens entre la structure des communautés microbiennes dans les biofilms électroactifs et en fermentation, les individus qui les composent et les fonctions macroscopiques (électroactivité du biofilm, production d'hydrogène) qui leur sont associées dans des conditions permettant de réaliser le couplage des deux procédés. L'originalité de cette étude a été de travailler en milieu salin (30-35 gNaCl/L), favorable au transport de charges dans l'électrolyte de la MEC. Dans un premier temps, la faisabilité de la fermentation en conditions salines (3-75 gNaCl/L) a été démontrée en lien avec l'inhibition de la consommation de l'hydrogène produit et une forte prédominance d'une nouvelle souche de Vibrionaceae à des concentrations en sel supérieures à 58 gNaCl/L. D'autre part, la mise en œuvre de biofilms électroactifs dans des conditions compatibles avec la fermentation sombre a permis la sélection d'espèces dominantes dans les biofilms anodiques et présentant des propriétés électroactives très prometteuses (Geoalkalibacter subterraneus et Desulfuromonas acetoxidans) jusqu'à 8,5 A/m². En parallèle, la sélection microbienne opérée lors d'une méthode d'enrichissement utilisée pour sélectionner ces espèces à partir d'une source d'inoculum naturelle sur leur capacité à transférer leurs électrons à des oxydes de Fer(III) a été étudiée. Une baisse des performances électroactives du biofilm liée à une divergence de sélection microbienne dans ces deux techniques de sélection mène à limiter le nombre de cycle d'enrichissement sur Fer(III). Cependant, l'enrichissement sur Fer(III) reste une alternative efficace de pré-selection d'espèces électroactives qui permet une augmentation de rendement faradique de 30±4% à 99±8% par rapport au biofilm obtenu avec un inoculum non pré-acclimaté. Enfin, l'ajout d'espèces exogènes issues de la fermentation sombre sur le biofilm électroactif a révélé une baisse de l'électroactivité du biofilm se traduisant par une diminution de la densité de courant maximale produite. Cette baisse pourrait s'expliquer par à une diminution de la vitesse de transfert du substrat due à un épaississement apparent du biofilm. Cependant, un maintien de sa composition microbienne et de la quantité de biomasse laisse supposer une production d'exopolymères (EPS) dans le biofilm en situation de couplage. / Nowadays, alternative and sustainable solutions are proposed to avoid the use of fossil fuel. Hydrogen, which constitutes a promising energy vector, is essentially produced by fossil fuel reforming (95%). Environmentally friendly production systems have to be studied. Two main families of technologies are explored to produce hydrogen: 1) by thermochemical and electrochemical decomposition of water and 2) from different biomass sources. Among those last ones, microbial electrolysis cells (MEC) allow to produce hydrogen by electrolysis of organic matter. A MEC consists in a classical cathode, which provides hydrogen production by electrochemical reduction of water, associated to a bio-anode that oxidizes organic substrates into carbon dioxide. This process is only possible because of the anodic development of an electroactive microbial biofilm which constitutes an electrocatalyst. In comparison to classical water electrolysis process, a MEC requires 5 to 10 times less electrical energy and therefore reduces the energetic cost of produced hydrogen. Furthermore, classical process of dark fermentation in mixed cultures converts sugars (saccharose, glucose) to hydrogen with a limited yield of 2-3 moles of hydrogen per mole of hexose because of the coproduction of organic acids (mainly acetic and butyric acids). Fed with acetate, a MEC can produce up-to 3 moles of hydrogen per mole of acetate. Therefore, the association of these two processes could permit to produce 8 to 9 moles of hydrogen per mole of hexose, which represents a major step toward the theoretical limit of 12 moles of hydrogen per mole of hexose.Therefore, this work aims at analyzing the relationship between microbial community structures and compositions and the associated macroscopic functions (biofilm electroactive properties, hydrogen production potential) in electroactive biofilms and in dark fermentation in conditions allowing the coupling of the two processes. The originality of this study is to work in saline conditions (30-35 gNaCl/L), which favors the charges transfer in the MEC electrolyte.First of all, feasibility of dark fermentation in saline conditions (3-75 gNaCl/L) has been shown. This was linked to an inhibition of produced hydrogen consumption and the predominance of a new Vibrionaceae species at salt concentrations higher than 58 gNaCl/L. Secondly, electroactive biofilm growth in conditions compatibles to dark fermentation (pH 5.5-7 and fed with different organic acids) allowed to select dominant microbial species in anodic biofilms that present promising electroactive properties (Geoalkalibacter subterraneus and Desulfuromonas acetoxidans) with maximum current densities up to 8.5 A/m². In parallel, the microbial selection occurring during iron-reducing enrichment method used to select species from a natural inoculum source and based on their capacity to transfer electrons to iron oxydes (Fe(III)) has been studied. A decrease of electroactive performances of the biofilm linked to the divergence of microbial selection led to a limitation of the number of iron-enrichment steps. However, enrichment on Fe(III) presents an efficient alternative to pre-select electroactive species with an increase of coulombic efficiency from 30±4% to 99±8% in comparison with a biofilm obtained with a non-acclimated inoculum. Finally, the addition of exogenous bacteria from a dark fermenter on the electroactive biofilm revealed a decrease of electroactivity with a decrease of maximum current density produced. This diminution could be explained by a lower substrate transfer due to an apparent thickening of the biofilm. Nevertheless, the stability of microbial composition and of bacterial quantity on the anode suggests that a production of exopolymers (EPS) occurred.

A ce jour, le contexte énergétique mondial est dominé par une utilisation massive des énergies fossiles non-renouvelables et épuisables par nature. La production de biohydrogène de 2ème génération issu de déchets organiques par le procédé de fermentation sombre constitue donc une solution attractive pour diversifier le mix énergétique actuel. Dans ce cadre, l’objectif de ce travail est d’étudier l’influence de la qualité du mélange sur l’efficacité de la voie fermentaire sombre. En effet, les conditions d’agitation mécanique (type d’agitateur, vitesse d’agitation) et la viscosité du digestat (fonction des intrants en cours de culture), comptent parmi les paramètres abiotiques les moins étudiés à ce jour dans ce procédé. Or, l’agitation joue un rôle clé puisqu’elle doit permettre non seulement d’homogénéiser la phase liquide riche en bactéries, en substrats organiques, en métabolites et en biogaz soluble, mais aussi de favoriser les échanges de matière liquide-bactéries et liquide-gaz. Cependant, pour atteindre la qualité de mélange requise, il faut faire face à deux contraintes : d’une part il faut maintenir un niveau acceptable de stress mécanique pour les bactéries du consortium ; d’autre part, la puissance mécanique consommée par l’agitation doit rester limitée pour assurer la viabilité économique du procédé. Dans ce travail, les effets combinés de la viscosité du digestat et de la vitesse d’agitation des mobiles sur la production de biohydrogène dans un bioréacteur ont été étudiés dans un premier temps. Les résultats ont montré une influence significative de ces deux facteurs sur la productivité en biohydrogène qui a pu être reliée au nombre adimensionnel de Reynolds et au régime d’écoulement du digestat. Un maximum de productivité a été observé lors de la transition laminaire-turbulent. Dans un deuxième temps, des méthodes de détermination du temps de mélange (conductimétrie, décoloration chimique, Fluorescence Induite par Nappe Laser) et du transfert de matière liquide-gaz (désoxygénation/oxygénation) ont été mises en oeuvre dans les mêmes conditions de viscosité et d’agitation afin de rechercher les étapes limitantes pouvant expliquer les évolutions observées lors des essais de fermentation. Les résultats ont montré que transfert interfacial et mélange ne sont limitants qu’en régime laminaire, alors que les faibles productivités en régime turbulent résultent vraisemblablement d’une interaction entre la turbulence et les agrégats bactériens. Ensuite, l’écoulement dans le bioréacteur a été modélisé par une approche de type Mécanique des Fluides Numérique (CFD) et analysé par une méthode de Vélocimétrie par Images de Particules (PIV) afin de déterminer les échelles spatiales locales de la turbulence et de pouvoir les comparer à la dimension caractéristique des agrégats bactériens. Les mesures locales confirment les hypothèses émises à partir des valeurs moyennes observées. Finalement, un modèle de type ADM1 (Anaerobic Digestion Model N°1) standard a été modifié en prenant en compte les ions lactate et un modèle hydrodynamique de type « cascade de cellules » dans le but de simuler la production de biohydrogène en systèmes batch et continu. Les simulations sont en bon accord avec les résultats expérimentaux dans les deux modes de culture en supposant un réacteur parfaitement mélangé. En conclusion, l’ensemble de ce travail confirme que la viscosité du digestat et les conditions de mélange sont effectivement des paramètres essentiels à prendre en compte pour l’optimisation et l’extrapolation du procédé de fermentation sombre. / The global energy trends are currently dominated by a massive use of fossil non-renewable energy sources which are progressively depleting. In this way, the production of second-generation biohydrogen production from organic wastes by the dark fermentation process offers, therefore, an attractive solution to diversify the present energy mix. Within this framework, the aim of this work is to investigate the effect of the efficiency of the mixing process on dark fermentation. The conditions of mechanical agitation (mixer type, mixing speed) and the viscosity of the digestate (which depends on the variability of influent substrate concentration) are, indeed, among the abiotic factors that have been the most disregards up to now in this bioprocess. For example, mixing plays a key role because agitation conditions must ensure on the one hand the homogenization of the liquid phase enriched in bacteria, in organic substrate, in soluble metabolites, and in soluble biogas, and in the other hand promote liquid-to-bacteria and liquid-to-gas mass transfer. However, to reach the desired degree of mixing, two constraints must be faced: firstly, an acceptable level of mechanical stress must be maintained on the microbial consortium, and secondly, mechanical power input due to mixing must comply with the economic sustainability of the process. In this work, the combined effects of digestate viscosity and agitation conditions on the fermentative biohydrogen production in the bioreactor were studied first. Experimental results highlighted a significant effect of these factors on biohydrogen productivity which could be expressed as function of the purely hydrodynamic dimensionless Reynolds number and of the prevailing flow regime. Hydrogen production was maximized in the transition region between laminar and turbulent flow conditions. Secondly, experimental measuring methods of mixing time (conductimetric, chemical decolorization and Planar Laser Induced Fluorescence techniques) and mass transfer (dynamic deaeration/aeration) were implemented in the same conditions of viscosity and agitation conditions so as to investigate the possible limiting steps that could explain the trends observed in the mixed cultures. The results proved that mixing and liquid-gas transfer was slower than hydrogen production rate only in the laminar flow regime, while low production rate under turbulent flow conditions might stem from an interaction between turbulent eddies and bacterial aggregates. Then, the flow field in the bioreactor was simulated using a CFD (Computational Fluid Dynamics) methodology and analyzed experimentally using PIV (Particle Image Velocimetry) to determine the characteristic turbulent length scales and to compare them to the characteristic size of the bacterial aggregates. Local measurements confirmed the assumptions made from average values derived from power input data. Finally, a modified ADM1 model (Anaerobic Digestion Model N°1) was developed to simulate the biohydrogen production, accounting for lactate ions and non-ideal mixing, under batch and continuous culture conditions. Simulations fairly agree with experimental data in both modes of cultures assuming perfect mixing condition. As a conclusion, the present work as a whole confirms that digestate viscosity and mixing conditions constitute key parameters that must be considered for process optimization and for the scale-up of dark fermentation.

Limitation of fuels reserves and contribution of fossil fuels to the greenhouse effect leads to develop a new, clean and sustainable energy. Among the various options, biohydrogen appears as a promising alternative energy source. The fermentative hydrogen production process holds a great promise for commercial processes. Hydrogen production by fermentative bacteria is a very complex and greatly influenced by pH. This paper presents biohydrogen production by bacterial strain Clostridium sp. Tr2. Operational pH strongly affected its hyrogen production. Its gas production rate as well as obtained gas product were roughly increase twice under controlled pH at 6 than non-controlled condition. Dark fermentation for hydrogen production of strain Tr2 was performed under bottle as well as automatic fermenter scale under optimal nutritional and environmental conditions at 30oC, initial pH at 6.5, then pH was controlled at 6 for bioreactor scale (BioFlo 110). Bioreactor scale was much better for hydrogen production of strain Tr2. Clostridium sp. Tr2 produced 0.74 L hydro (L medium)-1 occupying 72.6 % of total gas under bottle scale while it produced 2.94 L hydro (L medium)-1 occupying 95.82 % of total gas under fermenter scale. Its maximum obtained hydrogen yield of Clostridium sp. Tr2 under bioreactor scale Bioflo 110 in optimal medium with controlled pH 6 was 2.31 mol hydro (mol glucose)-1. / Dự trữ nhiên liệu có giới hạn và việc sử dụng nhiên liêu hoá thạch góp phần không nhỏ gây hiệu ứng nhà kính dẫn đến cần phải phát triển năng lượng mới, sạch và bền vững. Trong số các giải pháp, hydro sinh học xuất hiện như một nguồn năng lượng thay thế đầy hứa hẹn. Quá trình lên men sản xuất hydro có tiềm năng lớn để áp dụng trong sản xuất thương mại. Tuy nhiên qúa trình này rất phức tạp và chịu ảnh hưởng lớn bởi pH. Nghiên cứu này trình bày sản xuất hydro sinh học do chủng vi khuẩn Clostridium sp. Tr2. Quá trình sản xuất hydro của chủng này bị ảnh hưởng mạnh mẽ bởi pH thay đổi trong quá trình lên men. Tốc độ tạo khí cũng như lượng khí thu được của chủng này tăng gần gấp đôi trong môi trường có duy trì pH ở pH 6 so với môi trường không kiểm soát pH. Quá trình lên men tối sản xuất hydro của chủng Tr2 được thực hiện ở quy mô bình thí nghiệm cũng như bình lên men tự động trong điều kiện môi trường tối ưu ở 30oC, pH ban đầu 6.5, ở qui mô bình lên men tự động (BioFlo 110), pH môi trường sau đó được duy trì ổn định ở pH 6. Lên men sản xuất hdyro của chủng Tr2 trong bình lên men tự động tốt hơn rất nhiều so với lên men trong bình thí nghiệm. Clostridium sp. Tr2 chỉ tạo ra được 0,74 L hydro (L medium)-1 chiếm 72,6 % tổng thể tích khí thu được ở điều kiện lên men bình thí nghiệm trong khi chủng này sản xuất được 2,94 L hydro (L medium)-1 chiếm 95,82 % tổng thể tích khí ở điều kiện lên men tự động. Sản lượng hydro thu được lớn nhất của chủng này trong bình lên men tự động BioFlo 110 trong trong môi trường tối ưu có kiểm soát pH tại pH 6 là 2,31 mol hydro (mol glucose)-1.