Solvent extraction involved strongly coupled phenomena, sometimes acting at the smallest scales of matter (e.g. the molecular, or even particulate scale), and which affect their performance at large scale. The development of new processes and equipment is therefore mainly based on empirical approaches and macroscopic concepts (such as theoretical stage heights). These models are the result of many years of studies carried out from the laboratory to large-scale pilots, in order to properly integrate the effects of the complex chemistry and transport mechanisms involved. Although validated and widely used for the dimensioning of current industrial units (such as the Orano La Hague plant), they have the disadvantage of not being robust to changes in chemical nature and/or equipment, which may limit their relevance to face the new challenges of the chemistry for low energies carbon.
Among the potential applications of chemical process engineering to support the energy transition, liquid-liquid extraction is a key unit operation of the circular economy, both for nuclear (multi-recycling of plutonium) and for new technologies for energy (strategic materials). Given the urgent need to reduce process development costs and times in the hypercompetitive context of recycling, effort is today focused on the search for more efficient and more robust models, relevant for the simulation of innovating chemical processes. In this context, the new methodology we propose is based on the complementarity of i) miniaturized chemical reactors, which makes it possible to drastically reduce the volumes and flow rates of products handled in the exploratory phases, while widening the range of chemical conditions tested [1], and ii) fluid mechanics studies, thanks to instrumented hydrodynamic devices and the development of image processing tools [2]. The presentation will illustrate, in the case of liquid-liquid contactors, how the currently developed models can valuably improve the prediction of the interfacial area [3-6], the residence time and the apparent transfer kinetics of the droplets [7-9].