Rare earths for a greener Earth

how mineral science can unlock the potential of rare earth elements

Apatite, Natural History Museum of Los Angele, photo by Stan Celestian.

Rare earth elements (REEs) are a group of similar chemical elements that have become an important research area in the mineral and materials sciences communities due to their potential use in energy generation and storage, for a greener Earth. The element neodymium (Nd) makes very powerful magnets, lanthanum (La) and cerium (Ce) are used for petroleum catalytic conversion and air pollution controls, gadolinium (Gd) and europium (Eu) are used as phosphors in lasers/digital screens/medical imaging, and all the above, plus lutecium (Lu) and samarium (Sm), are used in smart missiles, guidance systems, communication systems, and many other applications. To list all applications would be time-consuming, so suffice it to say they are essential in our modern society. REEs have to be mined (or recycled) and chemically extracted to be used, so finding the most efficient method of extraction and separation is essential for reducing environmental impact.

REEs are not rare elements on Earth, but they are rare to find in a geological deposit of sufficient quantity for economical mining. Once a deposit is found, there are often many different REEs within the rock and minerals. This ore must be extracted, the REE-bearing minerals dissolved, and the REEs themselves separated into pure substances for later use. One REE-bearing ore mineral is bastnäsite, and the museum has a few samples in the collection. Many minerals in the museum collection have REEs, but only in trace concentrations.

Orange flat crystal that has a hexagonal in shape.
Bastäsite, Ce(CO3)(OH), an ore mineral for REEs from Trimouns, France. Image from RRUFF R060283.

It turns out that these REEs are very challenging to use in materials development as they have practically indistinguishable chemical behavior. Their electron configuration makes their molecular bonding geometry and environment nearly identical. To be technical, the s-, p-, and d-orbitals, which are the dominant places electrons can be found for bonding, are identical for the REEs (with a few exceptions). The only difference between the electron configuration is in the f-orbitals, whose energy is located somewhere between the s- and d-orbitals.

This shared chemistry explains why the REEs are at the bottom of the periodic table, separated from the rest as an insert into a single slot on the main table, because they can behave chemically as if they were all the same element.

Rows of different shaped electron orbit elipses.
A representation of electron positional probability orbitals. Blue are s-orbitals (sharp), yellow are p-orbitals (principle), red are d-orbitals (diffuse), and green are the f-orbitals (fundamental). These are historical names, today, they are represented by various quantum numbers (l = 0 for s, 1 for p, 2 for d, and 3 for f). Image source.

The f-orbitals are not influential in determining what kinds of bonds are made to other atoms, and therein lies the trouble. If the only major difference between the REEs are the f-orbital configurations, and these orbitals don’t have a role in chemical bonding, then how can they be separated from one another?

In general, to separate atoms from one another, one must rely on each atom’s different bonding environments. For example, sodium is easy to separate from potassium, because the two elements have very different atomic sizes, and different physical and chemical properties, despite having the same electronic charge. However, the REEs have very similar atomic size (although lanthanide contraction helps change their shape), the same atomic charge, and the same bonding geometry in crystals and molecules, but more on that later. However, the slight differences in the f-orbital configuration do allow for their separation (see this review for current methods), but it takes very precise chemical processes to do so.

Studying REEs is extremely challenging work, even more frustrating when you want to find ways of being very selective for a specific REE you want to separate. Since there are not many minerals that have REE as a major chemical component, this has limited the ability to study f-orbital mineral crystal chemistry in great detail.

My strategy over the past 10+ years has been to take minerals and expose them to high concentrations of REEs, and monitor the processes in real-time. Only then am I able see how minerals react and absorb the REEs to better quantify their crystal chemical properties. For example, I can take a microporous mineral (a mineral with structural holes that are not much bigger than the size of atoms), and put it in a solution with different types of REEs. I prefer titanium silicates (such as zorite and sitinakite) and zirconium silicates (such as gaidonnayite and umbite) as starting materials because they have somewhat flexible crystal structures and they are quite stable, which makes them useful for industrial/environmental applications.

A black cubic shaped crystal on a white rock.
Loparite (black), (Na,Ce,Sr)(Ce,Th)(Ti,Nb)2O6, an ore mineral for REEs. From the Kola Peninsula, Russia. Image from RRUFF R070251.

The chemical process I’m exploiting is called ion exchange (see my other posts like this one, or this one, for a deeper discussion). Briefly, the pore spaces inside these microporous minerals already have atoms in them, usually metals like sodium or potassium. If you place a microporous mineral in some water that contains other dissolved metals, then the mineral could go through a chemical process that exchanges the metals from the water for the ones that were already in the crystal.

Ion exchange is a spontaneous process — an element from the environment (the guest element) is preferentially absorbed into the mineral’s micropores and the host element (e.g., sodium or potassium) is kicked out of the crystal. The specific chemical driving force behind this process is not well understood and is a very active area of research. During ion exchange, the crystal as a whole must maintain a neutral charge throughout this process. If a calcium atom (with a charge of 2+) is absorbed in the mineral, then two sodium atoms (each with a charge of 1+) must be kicked out. REEs usually have a charge of 3+, therefore three sodium atoms would have to be kicked out. So the space that was normally occupied by 3 atoms, would then be occupied by only one atom. This can be problematic.

Imaging a wall in a building, where that wall has 2" x 4" x 8' pieces of wood (studs) holding it up for strength, spaced out every 18 inches. If you were to take out three studs and replace them with one, you might think that the whole structure could collapse, and it could if you did that throughout the entire building. But you might be able to get away with exchanging the studs on one or two walls, without affecting the structural integrity of the building much. Such is the case with minerals.

a sponge-like solid with one of the channels through the spongy structure exposed.
Microporous minerals have a sponge-like structure. Host cations (blue) can be exchanged for guest cations (green). In many cases, the ion exchange results in a more stable crystal structure. Image my own.

Keeping the crystal from collapsing is only half the battle. The more difficult problem starts when figuring out each REE’s unique chemistry, so that they can be separated. Some microporous minerals are well known as highly selective atomic sieves, and this sieving process relies on very subtle chemical changes in the ingoing and outgoing atoms. As luck would have it (finally some luck), subtle differences in crystal chemistry are controllable within minerals. By changing the shape of the pore size slightly, or the temperature at which the exchange process is happening, or making small modifications in the acidity of the water, and other controlling factors, one can begin to find the differences that allow the mineral to be incredibly selective to absorb the REE of interest, and in turn, how the f-orbital can influence bonding in a crystal.

Very dark brown crystal with a pointed top on a grey backgroud.
Zircon, ZrSiO4. NHMLA 24046, Australia. Not a principle REE ore mineral, but REEs can be found in some specimens in high concentration. Photo by Stan Celestian.

To check all possible configurations of mineral types and pore spaces for REE separation potential might be a daunting task, but fortunately, these combinations are not infinite. There are a little more 90 known microporous minerals that are of interest (excluding the 140+ zeolite species as of 2020), and more are discovered every year. Of those ~90, maybe a dozen show promise in being able to be both REE selective and structurally stable ion exchangers. This is a starting point. Once a mineral demonstrates the ability to exchange an REE, then the work begins on figuring out how selective it is: how to speed up the exchange process, how to repel the REE (which tells you what controls its attraction to the mineral), and possibly to add molecular levers inside the crystal like a Venus fly trap (see this blog post) to capture and hold onto the REE.

By tuning the crystal structure properties through careful experiments, the ion exchange process will eventually get better and better (as has been demonstrated in other studies, e.g. see these papers 1, 2, 3, 4). The outcomes of this type of fundamental research can then facilitate various applications, including those that mitigate environmental damages and create sustainable alternatives.

Here are two quick case studies of what REE exchange into various microporous minerals looks.

Example 1: Gadolinium into Sodium Sitinakite

Gadolinium is used as a contrast agent in MRI biomedical imaging of the brain blood vessels. Passing chemicals through the blood-brain barrier is tough because humans have evolved to keep the brain protected from foreign chemicals. Organic molecules containing Gd have been developed by Bayer for this specific purpose, but not without problems.

It was unknown if Gd could be exchanged into microporous materials, or what the exchange mechanisms might be if it did. Cesium (Cs+) and strontium (Sr2+) exchange into sitinakite had been previously described (3, 4), and served as an initial guide for studying REE ion exchange. It was first assumed that the REE would follow suit, that the ion exchange process would be similar to the 1+ and 2+ cations already studied. Interestingly, the ion exchange processes were different than those determined for Cs and Sr. It was found that the REE exchange processes were significantly different than those previously reported for alkali and alkaline earth metals. The Gd started to go in the structure very fast, paused while short while, and speed back up again. In total, there were at least four different times where something happened to the crystal structure that affected the exchange of Gd.

But what makes this exchange speed up and slow down, and what forces are acting upon the REE within the micropores of the mineral? What chemical processes are happening during the pause, and how can those be measured? All questions still need to figured out.

Example 2: Terbium, Europium, & Yttrium into Sodium Zorite

Structure of zorite with open tunnels through crystal structure. Many of the atoms are reduced to various polyhedra shapes.
Atomic crystal structure of Y exchanged zorite after Celestian et al. (2016). Yttrium is in the oval channels as green/white spheres. Sodium is yellow/white, oxygens are red, silicon is blue, and titaium is aqua. This is an example of site selective exchange into the mineral. Only one of the two different sodium sites were exchanged, which leaves the structure much more stable. In addition, the Y can likely be removed by hydrogen or another competitive cation because the structure did not collapse and the symmetry around the cation site did not change, which is an indication of reversibility. The above work was funded by the National Science Foundation, also see this poster.

Zorite is another titanium silicate used for many industrial processes (known to industry as ETS-4) ranging from catalysis to gas separation. Zorite is interesting in that it has both TiO6 and TiO5 polyhedra. The TiO5 pyramid has an unbounded oxygen atom at the apex, making this a potentially reactive site for REE interactions. Given the short distance of the Ti-O bond, if an REE is in proximity to this site, then electron transfer should occur between the Ti and REE — resulting in new type of catalyst. However, it was unknown which (if any) REEs could be exchanged into the structure.

In contrast to the sitinakite study above where the exchange occurred similarly for europium (Eu) and gadolinium (Gd), yttrium (Y) behaves very differently than either terbium (Tb) or Eu. In fact, Y was the only one of the three to leave the crystal intact for single crystal diffraction studies for atomic structure analysis. The Tb and Eu exchanges broke the crystal into pieces, indicating they placed large stresses on the crystals. The Y exchange is rapid, and occurs in a single step (as compared to the Eu/Gd exchange into sitinakite which occurred in multiple complex steps).

So why is there such a large difference between Y and Tb/Eu exchanges? Can zorite be modified so that Tb/Eu can be selected for by ion exchange over the other REEs including Y?

The ultimate goal of in this REEs research is to use as little energy as possible during separation while maintaining high element selectivity.


Microporous materials might be the answer, but we have to see how this works when scaled up. If it does work, then less energy means less electricity, less carbon footprint, less waste into landfills and oceans, and more efficient mining and separation. Microporous materials are recyclable and reusable. Further, the chemical properties of certain microporous materials change in useful ways when they contain a REE. For example, when Eu is inside a titanium silicate, it turns into a photocatalyst capable of breaking down really tough toxic chemicals (like asphalt shingles and roads, this work is on-going in my lab). REE-bearing microporous materials can also be used to separate hydrogen gas from other gasses, making them potentially useful in the supply-chain for fuel cell technology. And they can be used as medical imaging contrast agents when Gd is in their crystal structure (see Gd toxicity in organic molecules). The possibilities are numerous, making time spent to figure all the chemistry, atomic structures, and ion exchange mechanisms worthwhile.

There are still many unanswered questions about REE chemistry and ion separation. New materials need to be synthesized, experiments using a wide range of analytical techniques must be performed, and collaborative research between academics and industry leaders will undoubtably improve our understanding of these processes.

Aaron Celestian is Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, Adjunct Associate Professor in the Department of Earth Sciences at the University of Southern California, Affiliate Research Scientist at NASA Jet Propulsion Lab. He researches how minerals interact with their environments and with living things, and how those minerals can be used to solve problems like climate change, pollution, and disease.

Title photo: Apatite, Natural History Museum of Los Angeles, photo by Stan Celestian with color modifications by me. Some apatite deposits could contain economic quantities of REEs.



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Aaron Celestian, Ph.D.

Aaron Celestian, Ph.D.


Keeping science accessible. Researching how minerals can be used to solve problems like climate change, pollution, and disease. @ NHMLA, USC, NASA-JPL