Ce travail consiste à étudier différentes réactions photochimiques en dispositifs microfluidiques en utilisant la lumière UV/visible, par le biais de catalyseurs recyclables métalliques et non-métalliques, pour la synthèse organique aux applications pharmaceutiques et industrielles. En outre, l’utilisation des systèmes microfluidiques dans des chemins optiques miniaturisés de 500 µm résultant en une amélioration d’illumination. Les mesures d’actinométrie chimique confirment que ≈ 98% de la lumière émise atteint le mélange de réaction dans un réacteur fluidique Mikroglas® Dwell Device largement utilisé dans la littérature. Différentes réactions de cycloaddition [2 + 2] utilisées en synthèse totale ont été testées en utilisant un sensibilisateur sous UV. La réaction est quantitative après 2 h contre 47% après 10 h en batch. La polymérisation radicalaire contrôlée (ATRP) a été étudiée en utilisant le catalyseur photorédox éosine Y sous illumination à base de LED vertes. Six heures d'irradiations sont suffisantes pour fournir des polymères mono dispersés aux applications variables (plastiques, latex ...). Ces catalyseurs non-métalliques sont d'une importance capitale car ils sont plus respectueux de l'environnement. Former de nouvelles liaisons C-C et C-O est le cœur de la synthèse organique. L’utilisation de sources LEDs UV et d’un catalyseur photorédox nous a permis de former des produits d'addition (> 99%), à partir d’une part de sels de trifluoroborates et de TEMPO ou d’accepteurs de Michaël, d’autre part, après 2,5 min d'irradiation contre 8-24 h en batch.Ce travail montre clairement l’apport des systèmes microfluidiques pour l’accélération de réactions photochimiques. / In order to mimic nature’s highly energy efficient photosynthesis reaction, this work focuses on photochemical reactions using UV/visible light, metal based recyclable catalysts and metal free catalysts in flow to synthesize organic material that have pharmaceutical and industrial applications. The utilized microfluidic systems have small path lengths (500 μm) resulting in improved illumination. Using chemical actinometry, it was shown that ≈ 98% of the light supplied reached the reaction mixture inside the widely used Mikroglas® Dwell device. [2+2] cycloaddition, used in total synthesis, was tested in flow using a sensitizer under UV. The optimized reaction was quantitative after 2 h vs. 47% after 10 h in literature’s batch system. Metal free ATRP was assessed using the commercial Eosin Y in flow with green LEDs. Only 6 h of irradiation were enough to give narrow dispersed polymers that have wide applications (plastics, latex…). Metal free catalysts are of critical importance as they are more ecofriendly. Forming new C-C and C-O bonds is the heart of organic synthesis. Using UV LEDs and a photoredox catalyst, adducts of trifluoroborate salts with TEMPO and with Michaël acceptors were obtained (>99%) after only 2.5 min of irradiation in flow compared to 8-24 h in batch. Our results highlight the impact of miniaturization on accelerating photochemical reactions. Less time and energy usage, improved yields and strictly linear kinetic graphs are main features of flow technology. In addition, miniaturization requires less safety precautions rendering it favorable for large scale industry. This work supports considering the microfluidic technology for greener industrial systems.

L’atome de fluor est un élément essentiel de notre quotidien. Il est indispensable pour le développement des batteries, de la réfrigération (Fréon), des cristaux liquides qui constituent nos écrans de téléphone ou encore des matériaux (Téflon®). Mais c’est surtout dans les sciences du vivant que le fluor joue un rôle primordial. L’introduction d’un atome de fluor modifie les propriétés physico-chimiques d’une molécule, permettant ainsi de moduler et d’améliorer profondément son activité biologique. Son introduction dans les molécules organiques représente donc un défi majeur pour les chimistes, qui nécessite sans cesse le développement de nouveaux réactifs de fluoration et perfluoroalkylation.Parmi ceux-ci, les sulfoximines perfluorées sont des réactifs de perfluoroalkylation électrophile, nucléophile ou radicalaire. De plus, les sulfoximines perfluorées possèdent des propriétés singulières ayant des applications en sciences des matériaux et du vivant.Mes travaux de thèse s’inscrivent dans la volonté de notre laboratoire de mettre au point une nouvelle voie d’accès générale aux sulfoximines fluorées ainsi qu’à la synthèse de sulfoximines hautement fonctionnalisées. Ma thèse a également pour but l’étude des sulfilimino iminiums, dont la synthèse dérive de celle des sulfoximines et qui sont des réactifs très efficaces et polyvalents pour des réactions perfluoroalkylation par catalyse photoredox. / Fluorine atom is essential in our everyday life. It is necessary for the development of battery, refrigeration (Fréon), liquid crystals which constitute the screens of phones, or materials (Téflon®). But its main role is in life sciences. The introduction of a fluorine atom modifies the physical and chemical properties of organic molecules, allowing to modulate and to enhance their biological activities. Its introduction in organic molecules constitutes a key challenge for chemists, which necessitates continually the development of new reagents for fluoration or perfluoroalkylation reactions. Among these, perfluorinated sulfoximines are electrophilic, nucleophilic or radical perfluoroalkylating reagents. Moreover, perfluorinated sulfoximines have peculiar properties with uses in material or life sciences.My PhD work falls within the project of our laboratory to develop a new general acces to perfluorinated sulfoximines and the synthesis of highly functionalized sulfoximines. My PhD work also deals with the synthesis of sulfilimino iminiums, derived from sulfoximines, which are efficient and versatile reagents for visible light-induced perfluoroalkylation reactions.

Développement de méthodes d'α,β-fonctionnalisation d'amines et formation d'hétérocycles optiquement actifs via l'utilisation de l'organocatalyse et de la catalyse photoredox. / Development of methods α,β-functionalization of amines andformation of optically active heterocycles via the use of the organocatalysis and thephotoredox catalysis.

In the last decade photoredox catalysis has emerged as an important new tool for organic chemists. The especially mild conditions and the broad range of reactions accessible using this methodology had a beneficial effect on the exploitation of radical reactions on otherwise labile substrates. Herein we report our work in this fast developing area and our efforts into the double functionalisation of styrenoid double bonds. We disclosed a new methodology for the room temperature photoredox catalysed alkoxy- and amino-arylation of styrenes using diaryl iodonium tetrafluoroborates and diazonium salts as aryl radical precursors. This methodology allows the successful regioselective coupling of three disparate components together and can be expanded to a wide range of alcohol nucleophiles, nitriles and water in moderate to good yields. The mild conditions employed permit the effective reaction of electron-rich styrenes and the tolerance of halogen functionalities, thus opening the possibility to further molecular elaboration. We then moved to explore the possibility of oxymethylnitrilation of styrenes and of the sysnthesis of heterocyclic cores via internal trapping with a nucleophile. Pleasingly, we were able to develop a mild and general methodology for the methylnitrilation of styrenes using simple and cheap bromoacetonitrile and photoredox catalysis. Furthermore, the synthesis of tetrahydrofuran and dihydrofuran cores was achieved in a single step, allowing us to synthesise tricyclic cores, maintaining functionisable handles such as halogens and ester groups. Finally, we decided to explore the possibility to add an azide functionality. After extensive optimisation, we were pleased to discover reaction conditions allowing for a switchable reactivity: under light irradiation we could perform an azidation reaction followed by addition of a nucleophile of choice; excluding the light from the reaction conditions, we could perform a double azidation reaction. The mild reaction conditions ensured the previously observed tolerance of functional groups; furthermore, we used a more sustainable copper-based photoredox catalyst.

Supramolecular chemistry allows the rapid formation of complex systems through self-assembly. These systems often possess unique properties not observed for conventional covalent constructs and have potential applications in areas such as sensing, drug delivery and catalysis. Metallosupramolecular container compounds have been shown to catalyse reactions with both regio- and stereo-selectivity in methods analogous to enzyme type catalysis. Separately, visible-light photoredox catalysis has recently gained considerable interest as an efficient, green and mild method for the rapid synthesis of many chemical compounds. In order to combine the favourable properties of both supramolecular catalysis and visible-light photoredox catalysis, a number of photoredox active metallosupramolecular assemblies were designed, synthesised and analysed. Initial steps were taken to stabilise a known iridium-based M6L4 luminescent cage compound to allow guest encapsulation to take place. The incorporation of isocyanide donors as strong ligands improved the stability of model compounds but synthesis of an analogous three-dimensional assembly was unsuccessful. Instead, a “complex-as-ligand” approach was taken that allowed the straight-forward formation of Pd2L4 systems from a range of photoactive iridium complexes. Importantly, unlike many other photoactive systems, the complexation to palladium did not drastically affect the photoredox properties of the constituent iridium complexes. Multiple approaches were then taken to improve both the stability and guest binding properties of the photoactive assembly in an effort to achieve supramolecular photoredox catalysis. Utilising a model system, a general method was thus developed for enhancing the association constants of neutral guests in organic solvents by switching to large, non-coordinating counter ions that provided reduced competition for the internal binding site. In combination with this increased binding affinity, a range of guest properties were adjusted by association with the hydrogen bond donor environment of the internal cavity. The encapsulation of quinone based oxidants led to unexpected and novel reaction pathways not observed in the bulk phase. As such, this work represents a significant advancement in development of metallosupramolecular systems capable of regio- and stereo-selective photoredox catalysis.

Over the past decade, the field of photoredox catalysis has garnered increasing amounts of attention in the organic chemistry community due to its wide applicability in sustainable free radical-mediated processes. Several examples have demonstrated that under carefully optimized conditions, efficient and highly selective processes can be developed through excitation of a photosensitizer using inexpensive, readily available light sources. Furthermore, these reactions can generally be performed under milder conditions than thermal reactions, as all the energy required to overcome the reaction barrier is supplied by light. Despite all these recent advancements in the field, many of these discoveries often lack in depth investigations into the excited state kinetics and underlying mechanisms. Furthermore, the vast majority of these transformations are photocatalyzed by ruthenium and iridium polypyridyl complexes. Not only are these precious metal catalysts extremely costly, but these metals are also known to be toxic, limiting their potential use in the development of pharmaceutical protocols. Herein, we present our solutions to these shortcomings, which involve a three-prong approach in the development of novel protocols, understanding the underlying mechanisms through detailed kinetic analysis, and by the development of new tools to facilitate mechanistic investigation for practitioners who may not possess specialized photochemical equipment. In this work, we were the first to demonstrate that radicals derived from amines, commonly employed as “sacrificial” electron-donors, can also act as reducing agents in photoredox transformations. We also present examples in which Methylene Blue, an inexpensive, non-toxic organic dye, can be employed as a viable alternative to ruthenium complexes for photoredox transformations. By employing a photosensitizer with more favourable excited state kinetics for electron-transfer, we successfully demonstrated that Methylene Blue could be used to increase the efficiency of a previously developed photoredox transformation. While employing organic dyes is an excellent strategy to lowering the cost of photoredox transformations, another viable strategy is to employ heterogeneous semiconductors. Titanium dioxide is an example of a semiconductor which is often employed in photocatalytic applications due to its low cost, desirable redox properties, and high chemical stability which allows for continued use. However, titanium dioxide has seen limited use in organic synthesis due to the requirement of UV irradiation for excitation. Herein, we present a process which led to the discovery of visible light photochemistry with titanium dioxide, generated through the adsorption of indole substrates creating a new, visible light absorbing complex. Employing this strategy, we were able to promote the photocatalytic Diels–Alder reaction of indoles with electron-rich dienes, giving access to valuable tetrahydrocarbazole scaffolds. Finally, in order to facilitate the characterization of chain processes in photoredox catalysis, we have successfully developed a visible light actinometer based on the ubiquitous photocatalyst, Ru(bpy)3Cl2. This actinometer offers many advantages compared to other visible light actinometers, such as completely eliminating the need for spectral matching, as the actinometer is also the photocatalyst. This technique should provide researchers with a mechanistic tool to properly characterize chain propagation in the transformation of interest.

A field in bloom: photoredox catalysis has allowed chemists access to highly reactive intermediates via the photo-mediated excitation of transition metal complexes and organic dyes for the mild generation of free radicals. These complexes and dyes are designed based on Nature’s blueprints of light-harvesting biomolecules that transform solar energy (photons) into chemical energy during photosynthesis. Light-mediated chemical activation is regarded as one of the most sustainable forms of chemical activation being that the energy provided by the sun is considered renewable and largely underutilized and presents an attractive avenue for research and development of new transformations that are mild, efficient, and waste-limiting in organic synthesis. Radical chemistry and photochemistry are united in their inherent ability to undergo single (or photoinduced) electron transfers by one-electron reaction modes. Combining these unique fields, photoredox catalysis has emerged as a mild and efficient alternative to classic alkyl radical generation using hazardous initiators and organostannanes. Photoredox catalysis has been dominated by ruthenium- and iridium-based polypyridyl complexes. These complexes are limited by their inherent redox potentials, restricting their reactivity towards relatively activated bonds. Nonactivated bromoalkanes and arenes are considered challenging substrates to engage using redox chemistry and typically only accessible in the realm of organostannane chemistry. Described herein are the efforts towards the discovery of free radical based organic transformations derived from nonactivated bromoalkanes and arenes mediated by photochemical excitation of polynuclear gold(I) complexes as photoredox catalysts. This work represents some of the first uses of a photoredox catalyst in the reduction of substrates having such high reduction potentials and offers a practical and useful alternative to classic radical reactions mediated by initiators (peroxides, persulfates, and azo compounds) and toxic organostannanes (Bu3SnH). Using gold based photoredox catalysts, the research conducted has provided many methodological advancements for the mild and efficient formation of carbon-carbon bonds using nonactivated bromoalkanes and a large collection of radical acceptors. Establishing the use of these photoexcited polynuclear gold(I) complexes in the context of classic radical reactions in organic synthesis was important for their validation as useful photocatalysts. First, the Ueno-Stork cyclization of nonactivated bromoalkanes was used to demonstrate the powerful reducing capabilities of the excited-state gold(I) complexes. Next, a photo-mediated variant of the Appel reaction was described, where the transformation of an alcohol to a bromoalkane was achieved using carbontetrabromide and N,N-dimethylformamide through the intermediacy of a Vilsmeier-Haack reagent. In combination with the hydrodebromination chemistry developed with photoexcited polynuclear gold(I) complexes, a photo-mediated one-pot formal deoxygenation reaction of alcohols was described; a useful alternative to the organostannane mediated Barton-McCombie deoxygenation reaction. Finally, in the field of medicinal chemistry, the functionalization of heteroarenes is of high interest for the discovery of drug candidates and bioactive molecules. In this respect, one of the most useful reactions for the functionalization of heteroarenes by alkyl radicals is the Minisci reaction using silver salts, carboxylic acids, and persulfates. Detailed are the efforts for the development of a photo-mediated redox-neutral improvement of the Minisci reaction, needing only gold(I) photocatalyst and nonactivated bromoalkane in the presence of heteroarenes. Overall, the work described in this thesis represents the push for mild and efficient alternatives to the relatively harsh conditions and/or toxic reagents and byproducts associated with classic radical chemistry. These studies demonstrate the ability to control highly reactive alkyl radical intermediates with the goal of their broader application in synthetic organic chemistry. The use of photoexcited polynuclear gold(I) complexes as potent reductants compared to ruthenium- and iridium-based polypyridyl complexes is illustrated through the genesis of highly reactive alkyl radicals from nonactivated bromoalkanes.

Thesis advisor: Matthias M. Waegele / Studying the visible light-driven photoredox catalysis coupled with transition-metal complexes is of overriding importance in the development of synthetic strategy. Comparing to conventional thermal catalysis, reactions catalyzed and/ or initiated by photon energy are not only attractive for establishing a more sustainable system, but also for their unique reactivity that has previously been inaccessible. However, one issue draws our attention is that such photoredox catalytic schemes often suffer from a limited substrate scope. To develop more efficient and effective synthetic strategies applicable to broader range of substrates, it is of our interest to construct an functional and reliable instrument to identify the critical mechanistic steps that lead to low product yield. To this end, we designed a time-resolved visible-pump/ infrared-probe spectroscopic measurement technique to monitor reaction dynamics in-situ. Using our transmission infrared setup, we effectively demonstrated in-situ photoexcitation and decay process of Tris(2,2′-bipyridyl)dichlororuthenium(II) hexahydrate in deuterated acetonitrile. In addition, to optimize signal resolution, an electronic filter was installed in one of the data-collecting channels to allow for concurrent AC-coupled and DC-coupled signal recording. A series of chopper wheel experiments was conducted to assure the functionality of the system and reliability of obtained data. / Thesis (MS) — Boston College, 2020. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Chemistry.

Photoredox catalysis has been a flourishing field in synthetic organic chemistry. Organic chemists have been inspired by Nature and the conversion of photons into potential energy by light-harvesting biomolecules. Recent developments in photoredox catalysis have led to a rapid increase in development of new methodologies in synthetic organic chemistry. The use of transition metal photocatalysts and organic dyes in photo-mediated processes has been proven to be an effective alternative to the harsh and toxic reaction conditions that area needed in classical radical formation. Photoredox catalysis eliminates the need of initiators, stoichiometric additives, and strong oxidants, and allows for the highly efficient formation of new C-C bonds under mild conditions. Photochemistry and radical chemistry work in unison in their ability to undergo photoinduced single electron transfers (SET) or photoinduced electron transfers (PET) and enable one electron reaction pathways. A plethora of photocatalysts have been developed, mainly using Ir- and Ru-based polypyridyl complexes. Polynuclear gold complexes have come to light in the last decade as another class of photocatalysts. This bench stable complex is marked by its unique photophysical and electrochemical properties, most notable the relatively long-lived excited state. This triplet excited state can be used as a powerful reductant or oxidant when irradiated with UVA light. A class of organic substrates that can be used when working with this gold photocatalyst, is nonactivated bromoalkanes, which could not be used if working with other photocatalysts. First, the alkylative semi-pinacol using gold photoredox chemistry and nonactivated bromoalkanes was described. A new mode of reactivity of the gold binuclear photocatalyst was found where it was shown to work as a photocatalyst and a Lewis acid. Next, a follow-up to that report was the halogen atom transfer radical addition (ATRA) using gold photoredox catalysis. A mild ATRA reaction was presented where the dual reactivity of the gold photocatalyst was exploited. Mild bromine and iodine transfer reactions, without the use of strong oxidants or toxic additives, are largely unknown, and a metal-based mechanistic pathway was proposed to explain this transformation. Minisci-type alkylation is of high-interest in the field of medicinal chemistry and drug discovery. With this is mind, a photoredox catalysed Minisci reaction was presented, where the alkylation of an activated heteroarenes was achieved by HAT via chlorine atom generation. Knowing this, the alkylation using primary alcohols was presented, were a the 𝛼-alkoxy radical is formed after a HAT by chlorine atom. When secondary alcohols were used, a reduction of the heteroarene occurred and was described. Finally, a photo-mediated [3 + 2] cycloaddition using N-aryl cyclopropylamines and α, β- unsaturated carbonyl systems was described. This simple method that was presented does not require the use of photocatalysts or added additives, as it is self-catalyzed. The reaction is proceeding through a single electron transfer (SET) and offers a wide scope for the synthesis of N-arylaminocycloalkyl compounds. Overall, the collection of work described in this thesis represents the growth of photoredox catalysis in organic synthetic chemistry and the ability to form highly reactive radical without the need of harsh conditions, toxic reagents, or strong oxidants. The use of binuclear gold(I) complexes as a photocatalyst with unique photophysical and electrochemical properties was shown. Compared to Ir- and Ru-based polypyridyl complexes, which cannot react with nonactivated bromoalkanes, the binuclear gold(I) complexes offer broader redox potentials and a newfound dual mode of reactivity. Furthermore, photo-mediated synthetical useful reactions were shown. The application of photoredox catalysis in synthetic chemistry will continue to flourish, and this work is sample of all the possibilities that a simple photon can bring.

L’objectif principal des travaux de recherche décrits dans ce manuscrit de thèse, est de concevoir de nouvelles méthodes de synthèse en chimie radicalaire et organométallique, de les associer dans une catalyse duale afin qu’elles puissent être appliquées à l’élaboration de briques moléculaires élaborées. Nous avons développé dans une première étude une alternative photocatalytique à la réaction de désoxygénation classique de Barton-McCombie d’alcools secondaires et tertiaires par l’hydrure de tributylétain. Celle-ci repose sur l’utilisation d’un précurseur O-thiocarbamate dérivé d’alcools pouvant être réduit par catalyse photorédox. Une étude mécanistique, basée sur des expériences de fluorescence et de voltammétrie cyclique, est aussi présentée. Dans une seconde étude, une élégante méthode de génération de radicaux alkyles non stabilisés, par photooxydation de « ate-complexes » de bore ou d’espèces hypervalentes à base de silicium est présentée. Ces radicaux sont piégés en présence d’accepteurs radicalaires ou engagés dans une catalyse duale avec des complexes de Nickel. / Visible-light photoredox catalysis has emerged as a very powerful strategy to generate radical species replacing more and more tin-mediated or stoichiometric redox methodologies. The main objective of the research described in this Ph. D. thesis is to develop new synthetic methodologies in radical and organometallic chemistry, and merge them in a dual catalysis process for the preparation of elaborated molecular building blocks. In a first study, we report a photocatalytic alternative of Barton-McCombie deoxygenation based on a visible-light photoreduction of O-thiocarbamates derived from secondary and tertiary alcohols. A mechanistic investigation is presented based on fluorescence quenching and cyclic voltammetry experiments. In a second study, a challenging method of generation of unstabilized alkyl radicals by photooxidation of borate salts or hypervalent silicon species is reported. These radicals are trapped by free radical scavengers or engaged in a photoredox/nickel dual catalysis.