(260c) Constant-Pattern Design Method for Separating Ternary Mixtures of Rare Earth Elements Using Ligand-Assisted Displacement Chromatography | AIChE

(260c) Constant-Pattern Design Method for Separating Ternary Mixtures of Rare Earth Elements Using Ligand-Assisted Displacement Chromatography

Authors 

Choi, H. - Presenter, Purdue University
Harvey, D. M., Purdue University
Wang, N. H. L., Purdue University
Ding, Y., Purdue University
The rare earth elements (REEs) consist of the 15 elements of the lanthanide series, yttrium, and scandium. REEs are important components of many high-tech products, and as a result, the demand for REEs is growing. Traditionally, REEs are mined from ores that contain a mixture of several different elements. For most applications, high purity of a single REE is required, so the ores must be separated and purified. However, since REEs have similar physicochemical properties, the separation and purification is a challenging and expensive process. Ligand-assisted displacement chromatography (LAD) was developed to separate REEs in the 1950s and 1960s using commercially available ion exchange sorbents and ligands. While previous studies have demonstrated the feasibility of LAD for separating REEs with high purity (>99%) and high ligand efficiency, the productivity for these experiments were very low. Without a systematic process to design and optimize LAD systems with mass transfer effects, LAD systems were designed using a trial and error approach.

In this study, a constant-pattern design method is developed to separate REEs with high purity and high productivity for a given target yield. When the mixed band regions in an LAD system reach a constant pattern state, the yields of high purity products will be maximized. A general map was developed to show the relationship between the minimum column length to reach a constant pattern state and key dimensionless groups, which include selectivity, loading fraction, cut, and overall mass transfer coefficients. Given a target yield, intrinsic parameters, and material properties, operating conditions were calculated using the general map without trial and error. The design method was verified experimentally for different target components, yields, ligand concentrations, and feed compositions. The yields of the target components were within 3% of the expected values, demonstrating the robustness of the design method. High purity (>99%) of REEs were obtained from experiments with two orders of magnitude (x800) higher productivity than literature results. In cases with a minimum required yield, the component with the lowest feed concentration can be the limiting component, even in cases where it does not have the lowest selectivity.