Growing energy needs and the climate emergency require a rapid transition to completely carbon-free energy, by mixing renewable energies and sustainable nuclear power. In this context, the precipitation of plutonium and uranium in the form of oxalate constitutes a key step in the industrial process of recycling spent fuel. A detailed understanding of the crystallization mechanisms of these oxalates thus constitutes a major challenge for better management of these operations.

However, it is now widely accepted that ions in solution assemble into crystals via a series of non-crystalline transient states, which fundamentally contradicts all classical nucleation theories used in precipitation models. In particular, we have demonstrated in recent years that rare earth oxalate crystals (Eu, Nd, Ce, Tb), some used to experimentally simulate the recycling of uranium and plutonium, form via liquid, reagent-rich nanodroplets which separate from the aqueous solvent. This behavior modifies the view hitherto retained for the precipitation of these oxalates and leads us to question the behavior of actinide oxalates.

The aim of this thesis is to confirm or refute that transient mineral droplets also form during the formation of uranium and plutonium oxalates, and to determine whether crystallization transients impact the precipitation models used to calibrate the recycling process of nuclear fuel. This study will not only impact precipitation processes used in recycling, but will also advance a fundamental question about long-debated “non-classical” crystallization.