Résumé : The development of molecular receptors for anion recognition has become an important aspect of supramolecular chemistry. In this thesis, we focused our attention to the study of systems for fluoride recognition in water. Fluoride is indeed an anion of interest due to its implication in environmental and health related issues. Furthermore, its small size and high hydration energy make its recognition in water particularly challenging.

Most of the synthetic systems reported for fluoride recognition have been extensively studied in organic solvents (DMSO, acetonitrile) using tetrabutylammonium fluoride (TBAF) as the source of fluoride. In many cases, titra- tion behaviours are observed that cannot be ascribed to a classical 1:1 binding isotherm, deprotonation problems of Brønsted-Lowry acid type of receptors aside. In the first part of our thesis we investigated, using a uranyl-salophen re- ceptor which recognizes fluoride via Lewis Acide/Base interactions, the origin of the unusual titration behaviour. Via UV/vis, 19F and 1H NMR spectroscopies, we have been able to highlight that the equilibrium between the fluoride and the corresponding bihalide ion, HF−2 , which is inevitably generated along with the hydroxide anion in situ due to trace amounts of water, can be at the ori- gin of this singular behaviour. Our results put to light that when undertaking titrations with fluoride in DMSO, the fluoride–bihalide equilibrium can affect the data and that the latter species can even be the dominating species at low TBAF concentrations. When varying the solvent from DMSO to acetonitrile, the s-shape titration curves observed by UV/vis are no longer observed for the uranyl-salophene receptor that we studied. The fluoride-bifluoride equilibrium is still present but both of the anions generated in this process are recognised by the uranyl-salophene receptor with similar affinity constants above 10^6 M−1.

The second part of our work was devoted to finding ways to solubilize anion receptors that are efficient in organic solvents, into an aqueous environment. Two approaches were investigated: (i) grafting of the receptors onto silica nanoparticles and (ii) the micellar incorporation of the receptors. For the first strategy, we developed two silylated urea-based receptors. These receptors were first studied in organic solvents (DMSO and/or acetonitrile) where they showed selectivity, among halides, towards fluoride. Once grafted on the silica nanoparticles, due to the fact that hydroxyl groups and solvent molecules are present in the silica matrix, fluoride recognition was not possible.

We explored it with different simple H-bond based urea receptors in the second srategy. With the cationic surfactants, cetyl trimethyl ammonium chloride and bromide, the counter-ions of the micelles interfere with the fluoride recognition. With the neutral surfactant triton X-100, the incorporation of the anion receptors proved to be difficult. Moreover, the variations observed in the UV/vis spectra upon titrations were too small to be able to make any conclusions about fluoride recognition.