(208f) Ligand-Assisted Displacement Chromatography for Rare Earth Elements Separations

Rare earth elements (REE), which include scandium, yttrium, and 15 elements of the lanthanides, are critical components of many high-technology products. Since the ions of REE have the same valence and little differences in size and other physical and chemical properties, they cannot be separated using conventional ion exchange processes. Spedding and Powell demonstrated the feasibility of ligand-assisted displacement chromatography (LAD) for REE purification using polymeric ion exchangers and ligand elution on the preparative chromatography scale in the 1950s and 1960s. Three lanthanides were recovered with relatively high purity (>99%) and with yields from 83 to 93%. However, each separation run took more than three weeks, resulting in very low adsorbent productivity. More importantly, the complex mechanisms in this system were not well understood, and no systematic design/optimization method or scaling rules have been reported for the systems with mass transfer effects.

In this study, general LAD design and optimization methods were developed to recover high-purity REE with high yield and high productivity. The methods were tested with the purification of three REE. The results showed that the average yield of high-purity (>99%) products was more than 96%, and sorbent productivity was two orders of magnitude higher than that of Spedding and Powell. Rate model simulations were developed and verified using the experimental data and literature data. The verified model and model parameters were used to elucidate the dynamic separation phenomena in LAD. The simulation results showed that the mechanism of displacement is different from that of conventional displacement chromatography. The reaction of ligand with presaturant, which has the highest affinity for the ligand, drives the isotachic train in LAD. After the presaturant, solute with the highest affinity to the ligand elutes next. The constant pattern of shock waves forms due to the reactions of ligand with solutes. The shock layer thickness depends on the ligand selectivity relative to sorbent selectivity and mass transfer resistance. The maximum yield and productivity were greatly affected by the mass transfer resistances, selectivity, feed compositions, feed loading volume, and effective ligand concentration. These important parameters were incorporated into key dimensionless groups, which can be used in the efficient design and optimization method to optimize product purity, yield, or productivity for LAD systems at different scales.