Column model. For a fixed bed with void fraction ε, bulk density ρb, flow u and axial dispersion D: ε ∂c/∂t + ρb ∂q/∂t + u ∂c/∂z = ε D ∂²c/∂z², with a linear-driving-force uptake ∂q/∂t = k (q* − q). Equilibrium is a competitive Langmuir isotherm q* = qmax KLi c /(1 + KLi c + KNa cNa): lots of sodium crowds lithium out of sites. The rate constant follows Arrhenius, k = k25 exp[Ea/R (1/298 − 1/T)], which is why geothermal brine is attractive: it is already hot. The scheme is finite-volume upwind with an implicit cell-wise uptake step, so it conserves lithium exactly and is stable at any rate constant.
Cycle. Adsorb until the outlet reaches the chosen fraction of the feed (breakthrough), swap columns, then strip the loaded column with fresh eluent (modelled as a much weaker isotherm, K reduced 50×). The eluate is more concentrated than the brine by the concentration factor, which then feeds a precipitation step; 1 g of Li gives 5.32 g of Li2CO3 (73.89 / 13.88).
Honest simplifications. One lumped sorbent (single pseudo-component for Li and a lumped competitor for everything else), constant bed properties, no pH, no fouling/silica scaling, no sorbent ageing or dissolution, perfect elution switch, no heat balance (the column temperature is simply set), no power or water costs. The numbers (8 mg/g, K = 0.02 L/mg) are of the right order for lithium sorbents but are illustrative, not data for a specific product. Models and textures were built in Blender.
Independent checks (run on load in your browser):
| running… |