Les transferts de nitrène représentent un outil synthétique très intéressant pour former simplement une liaison C-N à partir d’une liaison C-H. Notre laboratoire a développé des précurseurs de nitrène chiraux : les sulfonimidamides. Leur utilisation a abouti, en présence de catalyseurs de rhodium, à des réactions d’amination C-H hautement diastéréosélectives. Ce projet de thèse s’inscrit dans la continuité de ces travaux. Dans un premier temps, la synthèse totale de la Dibromophakellstatine a été envisagée, impliquant comme étape clé, une étape d’amination C-H en position pseudo benzylique. Le projet n’ayant pas abouti, une séquence réactionnelle de quelques étapes a été développée à partir de l’amination C-H d’éthers d’énols et de benzocyclobutènes. Ainsi, plusieurs motifs perhydroindoles ont pu être préparés avec de bons rendements et d’excellentes diastéréosélectivités dans la plupart des cas. Dans le cadre d’une chimie plus éco-compatible, il a ensuite été envisagé de limiter l’introduction d’iode hypervalent dans les conditions de l’amination C-H. Pour cela, une première approche a consisté à utiliser les haloamines comme précurseurs de nitrène, cependant sans résultat satisfaisant. Une autre alternative a été d’introduire un oxydant, respectueux de l’environnement, permettant la réoxydation de l’iodobenzène formé en cours de réaction en une espèce de nouveau réactive (I(III)). De même, ces derniers résultats plutôt décevants ne permettent pas de s’affranchir de l’introduction de dérivé iodé en quantité stoechiométrique. / Catalytic nitrene transfers are useful tools in organic synthesis for the efficient conversion of a C-H bond into a C-N bond. In this context, our group has recently reported the use of sulfonimidamides as efficient chiral nitrene precursors in the rhodium-catalyzed stereoselective C-H amination of hydrocarbons. These PhD studies follow on from this work; it aims, on one hand, at applying the catalytic C-H amination in total synthesis, and, on the other hand, at searching for more sustainable reactions conditions. The first part of the manuscript reports our initial investigations devoted to the synthesis of Dibromophakellstatine. The strategy was based on a key step of C-H amination of a pseudo benzylic position but did not prove successful. A second application deals with the synthesis of polycyclic nitrogen compounds that relies on the catalytic C-H amination of enol ethers and benzocyclobutenes. A 3- to 4-step scheme, thus, allows the efficient access to perhydroindole scaffolds that are isolated in good yields and excellent diastereoselectivity. The second part deals with the search for sustainable reaction conditions that will avoid the use of stoichiometric amounts of hypervalent iodine reagents. These are indeed responsible for the production of excess iodobenzene. A first approach involves the use of haloamines as nitrene precursors but it did not lead to satisfying results. Attention has thus been paid to the use of benign oxidants allowing the in situ generation of an iodine(III) species from PhI. An extensive screening of reagents and reaction parameters has led to uncover a first significant result in the case of indan that, however, does not prove reproducible.

The Chlorine-Ammonia Process was developed recently as a preoxidation process to minimize the formation of bromate during ozonation of the waters containing a significant bromide concentration. Chlorine is added first, followed by ammonia 5-10 minutes later, with the goal of sequestering bromide in monobromamine before the subsequent ozonation step. The goal of this research was to improve the Chlorine-Ammonia Process by introducing a very short prechlorination step (i.e., 30 seconds before addition of ammonia) to minimize overall disinfection by-product formation. Also, in this strategy, formation of a powerful halogenating agent, HOBr, is minimized and bromochloramine (NHBrCl) is used predominantly instead of monobromamine to sequester bromide during ozonation. To support this improved approach to bromide sequestration, this study examined the formation and decay of bromochloramine as a function of operating conditions, such as pH and Cl2/N ratio, and refined a chemical kinetic model to predict haloamine concentrations over time. Two natural organic matter (NOM) sources were used in this study (Lake Austin, Texas and Claremore Lake, Oklahoma) to study the effect of NOM on monochloramine and total chlorine decay after 30 seconds of prechlorination. The rate of the reaction between haloamines and fast and slow sites on the NOM was estimated. A kinetics model was developed to model total chlorine decay after a short prechlorination time. The model is based on the Unified Haloamine Kinetic Model developed by Pope (2006). Pope`s model failed to model the initial monochloramine concentration after 30 seconds prechlorination time as well as the monochloramine and total chlorine decay over time. The modified model shows an excellent prediction of monochloramine and total chlorine decay after 30 seconds prechlorination time at pH range of 6.5-8.0 and over a carbonate buffer concentration range of 2-10 mM. The model includes a new bromochloramine decay scheme via the reaction with monochloramine and with itself. In addition, new rate constants for the reaction of HOCl with bromide ion and reaction of HOBr with monochloramine were added. The hypobromous acid formation rate was found to be an acid-catalyzed reaction, which confirms the finding of Kumar et al. (1987). A new value of the acid catalysis effect of hydrogen ion was estimated. New terms were introduced to the hyprobromous acid formation rate including the acid catalysis effect of bicarbonate, carbonic acid, and ammonium ion. In addition, the reaction of HOBr with monochloramine to form bromochloramine was found to be an acid-catalyzed reaction, and a new value of the rate constant was estimated. / text