Lixa v0.1.2: an open-source, image-informed framework for closed-loop design of leaching processes for domestic critical-mineral recovery, from the pore to the plant scale

Shahram Asgari 

Leaching takes a metal out of a rock and into a solution. In the United States, rare earth elements, lithium and scandium occur mostly in rocks where the metal is thinly spread through a porous matrix: weathered clays, coal ash, lithium claystones and bauxite residue. For such rocks the leaching recipe (reagent, concentration, temperature, particle size, liquid-to-solid ratio and contact time) decides whether a deposit can be worked, and the pore structure of the rock sets how fast the solution reaches the metal. The recipe is still found by months of bench tests, without the pore structure. Lixa is a Python framework that joins the two. From an image of the rock it reads porosity, connected porosity, specific surface, tortuosity and permeability. These numbers tune a multi-host kinetic model built from shrinking-core dissolution, Langmuir ion exchange and a reagent balance that includes carbonate and silicate gangue. A differential-evolution search then finds the recipe with the highest net value per tonne within reagent and effluent limits, and the same kinetics project it to a heap or to a well field, where the image permeability sets the injection pressure. Each numerical part was checked against an exact solution: lattice Boltzmann flow to 0.12 percent, column transport to 0.004 in relative concentration, kinetics to better than one part in a thousand. The shipped presets for clay-hosted rare earths and coal fly ash are consistent with published recovery ranges. For a weathered-clay ore the search returns ammonium sulfate near 0.3 mol L⁻¹ at room temperature, the concentration used in practice, in 6 s. Lixa is a screening tool that gives the first design point for bench work; it does not replace a feasibility study.