This thesis describes a new technique for the preparation of ultrathin Pd based membranes supported on a hollow-fiber α-alumina substrate for H₂ separation. The effectiveness of the membranes is demonstrated in the ethanol steam reforming (EtOH SR) reaction in a membrane reactor (MR) for H₂ production. The membrane preparation technique uses an electric-field to uniformly deposit Pd nanoparticle seeds on a substrate followed by deposition of Pd or Pd-Cu layers on the activated surface by electroless plating (ELP). The well distributed Pd nanoparticles allow for enhanced bonding between the selective layer and the substrate and the formation of gas tight and thermally stable Pd or Pd-Cu layers as thin as 1 µm, which is a record in the field. The best Pd membrane showed H₂ permeance as high as 5.0 × 10⁶ mol m²s⁻¹Pa⁻¹ and stable H²/N₂ selectivity of 9000 - 7000 at 733 K for 5 days. The Pd-Cu alloy membrane showed H₂ permeance of 2.5 × 10⁶ mol m⁻²s⁻¹Pa⁻¹ and H₂/N₂ selectivity of 970 at the same conditions. The reaction studies were carried out with a Co-Na/ZnO catalyst both in a packed bed reactor (PBR) and in a MR equipped with the Pd or Pd-Cu membrane to evaluate the benefits of employing membranes. For all studies, ethanol conversion and hydrogen product yields were significantly higher in the MRs compared to the PBR. Average ethanol conversion enhancement and hydrogen molar flow enhancement were measured to be 12 % and 11 % in the Pd MR and 22 % and 19 % in the Pd-Cu MR, respectively. These enhancements of the conversion and product yield can be attributed to the shift in reaction equilibria by continuous hydrogen removal by the Pd based membranes. The comparative low enhancement in the Pd MR was found to be the result of significant contamination of Pd layer by CO or carbon compounds deposition during the reaction. A one-dimensional modeling of the MR and the PBR was conducted using identical conditions and their performances were compared with the values obtained from the experimental study. The model was developed using a simplified power law and the predicted values matched experimental data with only minor deviations indicating that the model was capturing the essential physicochemical behavior of the system. Enhancements of ethanol conversion and hydrogen yield were observed to increase with rise in space velocity (SV), which could be explained by the increase in H₂ flux through the membranes with SV in the MRs. / Ph. D.

Oxidation reactions are of prime importance at an industrial level and correspond to a huge market. Oxidation reactions are widely practiced in industry and are thoroughly studied in academic and industrial laboratories. Achievements in oxidation process resulted in the development of many new selective oxidation processes. Environmental protection also relies mainly on oxidation reactions. Remarkable results obtained in this field contributed to promote the social image of chemistry which gradually changes from being the enemy of nature to becoming its friend and savior. This study dealt with two aspects regarding oxidation process. The first aspect represented an experimental study for the partial oxidation of benzene to phenol using Pd membrane in the gaseous phase. The second part was a theoretical study for some of the advanced oxidation process (AOPs) which are applied for contaminant destructions in polluted waters. Niwa and coworkers reported a one step catalytic process to convert benzene to phenol using Pd membrane. According to their work, this technique will produce a higher yield than current cumene and nitrous oxide based industrial routes to phenol. A similar system to produce phenol from benzene in one step was studied in this work. Results at low conversion of benzene to phenol were obtained with a different selectivity from the reported work. High conversion to phenol was not obtained using the same arrangement as the reported one. High conversion to phenol was obtained using a scheme different from the one reported by Niwa et al1. It was found that producing phenol from benzene is not related to Pd-membrane since phenol was produced by passing all reactants over a Pd catalyst. Within the studied experimental conditions, formation of phenol was related to Pd catalyst since Pt catalyst was not capable of activating benzene to produce phenol. Other evidence was the result of a blank experiment, where no catalyst was used. From this experiment no phenol was produced. A kinetic model for the advanced oxidation process using ultraviolet light and hydrogen peroxide (UV/H2O2) in a completely mixed batch reactor has been tested for the destruction of humic acid in aqueous solutions. Known elementary chemical reactions with the corresponding rate constants were taken from the literature and used in this model. Photochemical reaction parameters of hydrogen peroxide and humic acid were also taken from the literature. Humic acid was assumed to be mainly destroyed by direct photolysis and radicals. The rate constant for the HA- reaction was optimized from range of values in the literature. Other fitted parameters were the rate constant of direct photolysis of hydrogen peroxide and humic acid. A series of reactions were proposed for formation of organic byproducts of humic acid destruction by direct photolysis and radicals. The corresponding rate constants were optimized based on the best fit within the range of available published data. This model doesn't assume the net formation of free radicals species is zero. The model was verified by predicting the degradation of HA and H2O2 for experimental data taken from the literature. The kinetic model predicted the effect of initial HA and H2O2 concentration on the process performance regarding the residual fraction of hydrogen peroxide and nonpurgeable dissolved organic carbon (NPDOC). The kinetic model was used to study the effect of the presence of carbonate/bicarbonate on the rate of degradation of NPDOC using hydrogen peroxide and UV (H2O2/UV) oxidation. Experimental data taken from literature were used to test the kinetic model in the presence of carbonate/bicarbonate at different concentrations. The kinetic model was able to describe the trend of the experimental data. The kinetic model simulations, along with the experimental data for the conditions in this work, showed a retardation effect on the rate of degradation of NPDOC due to the presence of bicarbonate and carbonate. This effect was attributed to the scavenging of the hydroxyl radicals by carbonate and bicarbonate. A kinetic model for the degradation of methyl tert-butyl ether (MTBE) in a batch reactor applying Fenton's reagent (FeII/ H2O2) and Fenton-like reagent (Feo/ H2O2) in aqueous solutions was proposed. All of the rate and equilibrium constants for hydrogen peroxide chemistry in aqueous solutions were taken from the literature. Rate and equilibrium constants for ferric and ferrous ions reactions in this model were taken from the reported values in the literature, except for the rate constant for the reaction of ferric ions with hydrogen peroxide where it was fitted within the range that was reported in the literature. Rate constant for iron dissolution was also a fitted parameter. The mechanism of MTBE degradation by the hydroxyl radicals was proposed based on literature studies. The kinetic model was tested on available experimental data from the literature which involved the use of Fenton's reagent and Fenton-like reagent for MTBE degradation. The degradation of MTBE in Fenton's reagent work was characterized to proceed by two stages, a fast one which involved the reaction of ferrous ions with hydrogen peroxide (FeII/H2O2 stage) and another, relatively, slower stage which involved the reaction of ferric ions with hydrogen peroxide (FeIII/H2O2 stage). The experimental data of MTBE degradation in the FeII/H2O2 stage were not sufficient to validate the model, however the model predictions of MTBE degradation in the FeIII/H2O2 stage was good. Also, the model was able to predict the byproducts formation from MTBE degradation and their degradation especially methyl acetate, and tert-butyl alcohol. The effect of each proposed reaction on MTBE degradation and the byproducts formation and degradation was elucidated based on a sensitivity analysis. The kinetic model predicted the degradation of MTBE for Fenton-like reagent for the tested experimental data. Matlab (R13) was used to solve the set of ordinary nonlinear stiff differential equations that described rate of species concentrations in each advanced oxidation kinetic model. Niwa, S. et al., Science 295 (2002) 105

À la lumière du changement climatique et de l'épuisement des réserves de combustibles fossiles, l'innovation dans les technologies énergétiques vertes et durables devient un défi crucial. La fabrication de produits chimiques consomme de grandes quantités d'énergie et est responsable d'une part importante des émissions mondiales de carbone. Dans ce contexte, l'électrosynthèse, alimentée par de l'électricité renouvelable, peut remplacer de nombreux procédés thermochimiques industriels pour générer des carburants, des produits chimiques et des engrais. Plutôt que de nous concentrer sur des domaines qui ont reçu beaucoup d'attention ces dernières années (par exemple, l'électrolyse de l'eau et la réduction du CO2), nous avons exploré les domaines émergents de l'électrosynthèse hétérogène pour lesquels il existe un besoin substantiel. Dans le chapitre 3, nous soulignons l'importance de concevoir des électrocatalyseurs avec des sites actifs bien définis. Nous rapportons l'utilisation de la chimie réticulaire pour concevoir un système de modèle électrocatalytique à base d'organo-métallique conducteur avec des sites actifs moléculaires M-O4 pour l'oxydation électrochimique du 5-hydroxyméthylfurfural (HMFOR). L'activité des MOF portant des sites actifs Ni-O4 (Ni-CAT) et Co-O4 (Co-CAT) a été analysée avec des techniques spectroscopiques électrochimiques et operando pour élucider le mécanisme de réaction se produisant à la surface. Les expériences électrochimiques révèlent que le Co-CAT a un potentiel d'apparition plus précoce pour activer le HMFOR, par rapport à la plupart des catalyseurs établis, tandis que le Ni-CAT présente une cinétique plus rapide pour la conversion du 5-hydroxyméthylfurfural (HMF) en acide 2,5-furandicarboxylique (FDCA) . Nous avons déterminé que Ni-CAT atteignait des rendements de FDCA (notre molécule cible) de 98,7 %. L'efficacité faradique peut atteindre 86,8% d'efficacité faradique. La spectroscopie infrarouge indique le HMF avec un groupe aldéhyde lié à la surface comme intermédiaire clé dans le cycle catalytique, qui se forme une fois que l'oxydation M (II \ III) se produit. Ce travail illustre l'avantage d'utiliser des sites actifs moléculairement définis couplés à la spectroscopie operando pour fournir des informations fondamentales sur une variété de réactions électrosynthétiques et ouvrir la voie à la conception future de catalyseurs. Suite à ce projet, nous nous sommes tournés vers l'utilisation d'un réacteur à membrane sélective pour l'hydrogène afin d'explorer de nouveaux concepts de réaction et de catalyseurs. La clé ici était d'utiliser une feuille de Pd comme matériau qui réduisait les protons en *H dans un compartiment aqueux et transférait l'hydrogène dans un compartiment organique où il hydrogénait le réactif de choix. À l'aide d'un réacteur à membrane, nous avons pu séparer physiquement la réduction électrochimique de l'hydrogène et la chimie de l'hydrogénation d'une manière qui contournait l'utilisation du gaz H2 qui serait autrement nécessaire. Nous choisissons comme point de départ un produit chimique produit industriellement en excès, l'acétonitrile. Le réacteur à membrane Pd est appliqué pour hydrogéner complètement la liaison C≡N de l'acétonitrile. Avec succès, nous avons obtenu de l'ammoniac et de l'acétaldéhyde comme produits de réaction à un potentiel de début record de 0,4 V vs Ag/AgCl. Enfin, en concevant soigneusement une cellule spectroélectrochimique unique, nous avons pu effectuer des mesures spectroscopiques infrarouges pour visualiser le processus de réaction dans la membrane Pd et par conséquent proposé un mécanisme unique de réaction d'hydrolyse de l'imine (Chapitre 4). Dans le chapitre 5, nous choisissons d'innover dans un domaine émergent : la formation de liaisons électrochimiques C-N à partir de réactifs de petites molécules (par exemple CO2, NH3). Le mécanisme conventionnel de formation de liaisons électrochimiques C-N est basé sur le CO2RR électrochimique. Dans ce chapitre, nous proposons une stratégie orthogonale pour activer simultanément le CO2 et les N-réactifs en appliquant respectivement des impulsions de potentiel négatives et positives. Les nanoparticules de Cu sont utilisées comme catalyseur modèle, le CO2 agit comme réactif C et le NH3 agit comme réactif N pour le couplage C-N. Dans des conditions optimisées dans lesquelles la couverture *NH2 est maintenue à l'état stable tandis que Cu reste métallique, l'électrolyse pulsée augmente à la fois le taux de formation et la sélectivité des produits C-N urée, formamide et acétamide de 3 à 20 fois. En étendant le champ d'application à des réactifs C et N supplémentaires, ainsi qu'au couplage C-S, cette nouvelle approche démontre davantage sa valeur générale en électrosynthèse. / In light of climate change and depleting fossil fuel reserves, innovating green and sustainable energy technologies becomes a critical challenge. Chemical manufacturing consumes large amounts of energy and is responsible for a substantial portion of global carbon emissions. Against this backdrop, electrosynthesis, powered by renewable electricity, can replace many industrial thermochemical processes to generate fuels, chemicals, and fertilizers. Rather than focusing on areas that have received much attention in recent years (e.g. water electrolysis and CO2 reduction), we explored emerging areas within heterogeneous electrosynthesis for which there is a substantial need. In chapter 3, we highlight the importance of designing electrocatalysts with well defined active sites. We report the use of reticular chemistry to design a conductive metal organic framework-based electrocatalytic model system with molecular M-O4 active sites for electrochemical oxidation of 5-hydroxymethylfurfural (HMFOR). The activity of MOFs bearing Ni-O4 (Ni-CAT) and Co-O4 (Co-CAT) active sites were analyzed with electrochemical and operando spectroscopic techniques to elucidate the reaction mechanism occurring on the surface. Electrochemical experiments reveal that Co-CAT has an earlier onset potential for enabling HMFOR, relative to most established catalysts, while the Ni-CAT shows faster kinetics for the conversion of 5-hydroxymethylfurfural (HMF) to 2,5-furandicarboxylic acid (FDCA). We determined that Ni-CAT achieved FDCA (our target molecule) yields of 98.7% yield. The faradic efficiency can reach out to 86.8% faradic efficiency. Infrared spectroscopy points to HMF with a surface-bound aldehyde group as the key intermediate in the catalytic cycle, which forms once the M(II\III) oxidation occurs. This work illustrates the advantage of utilizing molecularly defined active sites coupled with operando spectroscopy to provide fundamental insights into a variety of electrosynthetic reactions and pave the way for future catalyst design. Following this project, we turned to the use of a hydrogen-selective membrane reactor to explore more new reaction and catalysts concepts. The key here was using a Pd foil as a material that reduced protons to *H at an aqueous compartment and transferred the hydrogen through to an organic compartment where it hydrogenated the reactant of choice. Using a membrane reactor, we could physically separate electrochemical hydrogen reduction and hydrogenation chemistry in a manner that circumvented the use of H2 gas as would otherwise be necessary. We choose a chemical that is industrially produced in excess, acetonitrile, as a starting point. The Pd membrane reactor is applied to fully hydrogenate the C≡N bond of acetonitrile. Successfully, we obtained ammonia and acetaldehyde as reaction products at a record onset potential of 0.4 V vs Ag/AgCl. Finally, by carefully designing a unique spectroelectrochemical cell, we were able to carry out infrared spectroscopic measurements to visualize the reaction process in Pd-membrane and consequently proposed a unique imine-hydrolysis reaction mechanism (Chapter 4). In Chapter 5, we choose to innovate in an emerging area: electrochemical C-N bond formation from small molecule reactants (e.g. CO2, NH3). The conventional electrochemical C-N bond formation mechanism is based on electrochemical CO2RR. In this chapter, we propose an orthogonal strategy to simultaneously activate CO2 and N-reactants by applying negative and positive potential pulses, respectively. Cu nanoparticles are used as a model catalyst, CO2 acts as the C-reactant, and NH3 acts as the N-reactant for C-N coupling. Under optimized conditions in which *NH2 coverage is maintained at steady state while Cu remains metallic, pulsed electrolysis increases both the rate of formation and the selectivity of the C-N products urea, formamide and acetamide by 3-20 times. By extending the scope to additional C- and N-reactants, as well as C-S coupling, this new approach further demonstrates its general value in electrosynthesis.