A new technology for rare-earth elements chemical separation has been licensed to Marshallton Research Laboratories, a North Carolina-based manufacturer of organic chemicals for a range of industries.

Developed by scientists from Oak Ridge National Laboratory and Idaho National Laboratory in the Department of Energy’s Critical Materials Institute (CMI), the technology provides insight into how to separate in-demand rare-earth elements cost-effectively, which could shift the industry to benefit producers in the United States.

The unique electronic properties of rare-earth elements (REEs)—a group of 17 metallic elements that includes 15 lanthanides plus yttrium and scandium—make them critical for producing electronics, optical technologies, alloys and high-performance magnets. These powerful, permanent magnets are vital to clean energy technology and defense applications.

Individual REEs do not occur in minable concentrations in the Earth’s crust, but are naturally mineralized together and must be chemically separated to use for technological applications. Their physical and chemical similarities make them extremely difficult and costly to separate while generating a lot of waste. Extraction and separation of REEs for technological applications occurs largely overseas—predominantly in China.

To meet the growing need for these materials and to limit the nation’s reliance on foreign sources, ORNL and INL scientists working under the banner of CMI, a DOE Energy Innovation Hub led by Ames Laboratory, have applied their expertise in chemical synthesis, separations, and engineering to design and produce new extraction agents based on diglycolamide (DGA) ligands and a corresponding process for separating lanthanides that outperforms current technology.

At Marshallton, our purpose is to become a domestic, strategically reliable supplier of DGA extractants for rare-earth elements. We expect to service pilot-plant and commercial operations in ore processing, recovery from mining tailings and recycling. We’re excited to further explore what these new extractants can achieve.

—Mac Foster, co-owner of Marshallton and a collaborator on the technology

REEs are commercially separated using liquid-liquid extraction, which uses ligands—organic molecules composed of carbon, hydrogen, oxygen, and nitrogen atoms—as extractants to selectively bind the REE ions. An oily solvent containing the extractant is vigorously mixed with an REE-rich aqueous solution, then allowed to separate in the same manner as oil and vinegar for salad dressing.

During this process, the REEs get transferred into the organic solvent forming complexes with the extractant molecules. DGAs show higher affinity for lanthanides with smaller ionic radius, which allows individual REEs to be separated from one another in multiple stages.

Our goal was to identify an extractant that surpasses the performance of the state-of-the-art ligands that are currently used in industry. The compound widely used is a phosphorous based extractant, called PC88A, and since its selectivity is relatively low, a lot of separation stages are required along with generation of additional waste throughout the process.

—ORNL’s Santa Jansone-Popova

Selectivity refers to the degree to which a solvent prefers one metal over another and is described by a unit called separation factor. For example, when seeking to separate adjacent lanthanides neodymium and praseodymium—both used in high-powered magnets—the phosphorus-based extractant’s separation factor is around 1.2, which is very low.

You have to run the extraction many, many times to separate adjacent lanthanides completely. We need to improve the economics of the process, reduce the waste, reduce the complexity—limit the steps it takes to achieve separation.

—Santa Jansone-Popova

ORNL’s Chemical Science Division had been experimenting with an alternative DGA called TOGDA, which has a separation factor of 2.5—already a big improvement over the phosphorus-based extractant. However, a key variable in the economics of the process is loading capacity—how many grams per liter of extractants can be held in the organic solvent without adverse reactions. TODGA could only handle about one-fifth of what the phosphorous-based extract could.

The extractant concentrations we were limited to were not adequate compared to the industry standard. At higher concentrations, we run into things like gelling or precipitation, which are detrimental to the process. If you think of the process as a conveyor belt, we want to be able to load that conveyor belt up as high as we can, or at least competitive with what industry does, to make it cost effective.

—Kevin Lyon, INL chemical engineer who tested and developed the process design for the licensed technology

Jansone-Popova recognized that by chemically modifying the structure of DGAs, she might improve their properties and their efficiency in extracting REEs.

Her team at ORNL began a systematic approach to making structural changes to the DGA ligands by adding a range of substituents known as alkyls—fatty organic groups that exclusively contain hydrogen and carbon atoms. These groups can be arranged into different structural configurations. For example, their length and shape can be altered, branches created or linear chains transformed into cyclic arrangements.

The ORNL team passed the trial ligands off to Lyon to test under industrial operating conditions using a counter-current solvent extraction system—a series of vessels that mix and settle the materials to separate out REE compounds through a sequence of liquid-liquid extraction stages.

During the mixing, the ligands attract the metal ions using electron-rich donor groups, binding the metal ions in a coordinated manner. Extracting certain lanthanides over others depends on ligands having the right number and arrangement of functional groups—atoms within a molecule that can maintain functionality independently of other atoms in the molecule—as well as the size of the ligands and their ability to mix with the oily organic solvent.

The ORNL team designed, synthesized and tested a library of chemically modified ligands, in collaboration with Lyon, narrowing the field of novel agents for industrial application that could potentially outperform state-of-the-art technology in REE selectivity. Each agent performs differently based on its physical arrangement and the electronic activity it prompts.

The TOGDA extractant, when saturated with REE ions, would rapidly transform from the liquid phase into a gel or precipitate. The new DGA ligands allow the system to remain homogenous even at higher extractant concentrations and maintain good selectivity.

—Santa Jansone-Popova

In separating REEs, the new ligands achieved a selectivity range of 2.5–3.1, a significant improvement for these critical materials.

The team then took on the challenge of scaling up the process to be viable for industry use.

Most REE extractants have a separation factor of about 1.5 for adjacent lanthanides across the series—if we get to 2, that’s good. If we get to 2.5, that’s really starting to save some money. If we can get to 3, we’re really happy. We’ve gotten to 6.7 with one of Santa’s ligands.

—ORNL’s Bruce Moyer, who leads the CMI focus area for diversifying supply and is a collaborator on the licensed technology

Resources

• Diana Stamberga, Mary R. Healy, Vyacheslav S. Bryantsev, Camille Albisser, Yana Karslyan, Benjamin Reinhart, Alena Paulenova, Mac Foster, Ilja Popovs, Kevin Lyon, Bruce A. Moyer, and Santa Jansone-Popova (2020) “Structure Activity Relationship Approach toward the Improved Separation of Rare-Earth Elements Using Diglycolamides” Inorganic Chemistry 59 (23), 17620-17630 doi: 10.1021/acs.inorgchem.0c02861