Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation




Mining Engineering

First Advisor

Dr. Rick Honaker

Second Advisor

Dr. Joshua Werner


Rare earth elements (REEs) are a group of 15 elements in the lanthanide series along with scandium and yttrium. They are often grouped together because of their similar chemical properties. As a result of their increased application in advanced technologies and electronics including electric vehicles, the demand of REEs and other critical elements has increased in recent decades and is expected to significantly grow over the next decade. As the majority of REEs are produced and utilized within the manufacturing industry in China, concerns over future supplies to support national defense technologies and associated manufacturing industries has generated interest in the recovery of REEs from alternate sources such as coal and recycling.

A solvent extraction (SX) process and circuit was developed to concentrate REEs from dilute pregnant leach solutions containing low concentrations of REEs and high concentrations of contaminant ions. The separation processes used for concentrating REEs from leachates generated by conventional sources are not directly applicable to the PLS generated from coal-based sources due to their substantially different composition. Parametric effects associated with the SX process were evaluated and optimized using a model test solution produced based on the composition of typical pregnant leach solution (PLS) generated from the leaching of pre-combustion, bituminous coal-based sources. Di-2(ethylhexyl) phosphoric acid (DEHPA) was used as the extractant to selectively transfer the REEs in the PLS from the aqueous phase to the organic phase. The tests performed on the model PLS found that reduction of Fe3+ to Fe2+ prior to introduction to the SX process provided a four-fold improvement in the rejection of iron during the first loading stage in the SX circuit. The performances on the model system confirmed that the SX process was capable of recovering and concentrating the REEs from a dilute PLS source. Subsequently, the process and optimized parametric values were tested on a continuous basis in a pilot-scale facility using PLS generated from coal coarse refuse. The continuous SX system was comprised of a train of 10 conventional mixer settlers having a volume of 10 liters each. A rare earth oxide (REO) concentrate containing 94.5% by weight REO was generated using a two- stage (rougher and cleaner) solvent extraction process followed by oxalic acid precipitation.

The laboratory evaluations using the model PLS revealed issues associated with a third phase formation. Tributyl Phosphate (TBP) is commonly used as a phase modifier in the organic phase to improve the phase separation characteristics and prevent the formation of a third phase. The current study found that the addition of TBP affected the equilibrium extraction behavior of REE as well as the contaminant elements., The effect on each metal was found to be different which resulted in a significant impact on the separation efficiency achieved between individual REEs as well as for REEs and the contaminant elements. The effect of TBP was studied using concentrations of 1% and 2% by volume in the organic phase. A Fourier Transform Infrared (FTIR) analysis on the mixture of TBP and DEHPA and experimental data quantifying the change in the extraction equilibrium for each element provided insight into their interaction and an explanation for the change in the extraction behavior of each metal. The characteristic peak of P-O-C from 1033 cm-1 in pure DEHPA to 1049 cm-1 in the 5%DEHPA-1%TBP mixture which indicated that the bond P-O got shorter suggesting that the addition of TBP resulted in the breaking of the dimeric structure of the DEHPA and formation of a TBP-DEHPA associated molecule with hydrogen bonding.

The experimental work leading to a novel SX circuit to treat dilute PLS sources was primarily focused on the separation of REEs from contaminant elements to produce a high purity rare earth oxide mix product. The next step in the process was the production of individual REE concentrates. To identify the conditions needed to achieve this objective, a thermodynamic model was developed for the prediction of distribution coefficients associated with each lanthanide using a cation exchange extractant. The model utilized the initial conditions of the system to estimate the lanthanide complexation and the non-idealities in both aqueous and organic phases to calculate the distribution coefficients. The non-ideality associated with the ions in the aqueous phase was estimated using the Bromley activity coefficient model, whereas the non-ideality in the organic phase was computed as the ratio of the activity coefficient of the extractant molecule and the metal extractant molecule in the organic phase which was calculated as a function of the dimeric concentration of the free extractant in the organic phase. To validate the model, distribution coefficients were predicted and experimentally determined for a lanthanum chloride solution using DEHPA as the extractant. The correlation coefficient defining the agreement of the model predictions with the experimental data was 0.996, which is validated the accuracy of the model. As such, the developed model can be used to design solvent extraction processes for the separation of individual metals without having to generate a large amount of experimental data for distribution coefficients under different conditions.

Digital Object Identifier (DOI)