[en] This research thesis deals with the back-end cycle of the nuclear fuel by improving, modernizing and optimizing the processes used for all types of fuels which are to be re-processed. After a presentation of the industrial context and of the state of the art concerning dissolution kinetic data for uranium dioxide and mixed oxide, the author proposes a model which couples dissolution kinetics and hydrodynamics of a solid in presence of auto-catalytic species, in order to better understand phenomena occurring at the solid-liquid-gas interface. The next part reports dissolution experiments on a non-radioactive material (copper) and out of a nuclear environment. Then, the author identifies steps which are required to transpose this experiment within a nuclear environment. The first results obtained on uranium dioxide are discussed. Recommendations for further studies conclude the report