INTERVIEW: Aclara bets on ionic-clay feed, AI to accelerate rare earth supply chain

Aclara Resources is betting that its Latin American ionic-clay deposits and use of artificial intelligence can help it overcome two of the biggest obstacles facing emerging Western rare-earth producers: securing heavy-rare-earth feedstock and mastering complex separation processes.

The company is developing projects in Chile and Brazil to supply Project Dynamo, its planned separation facility in Louisiana, as part of an integrated rare earth supply chain outside China. Aclara says its ionic-clay deposits can produce material with fewer impurities than conventional hard-rock sources, potentially simplifying downstream separation. The company is also developing an AI-enabled digital twin using data from its Virginia Tech pilot facility to model plant performance, accelerate optimization and reduce ramp-up risk.

The replica has already separated neodymium-praseodymium oxide at 99.5% purity, while Aclara expects to separate terbium in December 2026, dysprosium in the first quarter of 2027, and yttrium, samarium, and gadolinium around June 2027.

Platts, part of S&P Global Energy, spoke with Aclara CEO Ramón Barúa and COO Hugh Broadhurst about the opportunities and technical challenges involved in establishing a mine-to-magnet rare earth supply chain outside China.

The interview has been edited for clarity and length.

Platts: What is Aclara developing, how advanced are the projects and when do you expect to begin commercial production?

Hugh Broadhurst: We are developing two mines in Chile and Brazil. They are ionic-clay deposits focused on heavy rare earths, with our patented process technology for sustainable extraction of heavy rare earths from ionic clays. Those feed into our separation project, Project Dynamo in Louisiana, where we will separate the rare earths and also have an on-site metallization facility to convert the oxides from the separation into metals and alloys. We have our permit for Chile and aim to have it up and running by the end of 2028, along with our separation plant. Brazil will be shortly thereafter. We also announced the completion of basic engineering for Project Dynamo.

Ramón Barúa: We can have this ambition timeline because we have been investing in these projects during the past eight years. We have invested more than $300 million to develop these assets.

In Brazil, we have already converted the resources into reserves. In Chile, we have measured and indicated resources that will be converted into reserves in the first part of the first quarter of next year, supported by a feasibility study. We have done an enormous amount of metallurgical testing, both in independent labs, and we have built pilot plants in Chile and Brazil to demonstrate the technology. In Chile, we have the permits. In Brazil, we are waiting for the licencia previa, the EIA, to be approved. We expect that to happen in the first quarter of next year.

China took decades to master rare earth separation, and the process is often described as much an art as a science. Why does Aclara believe it can develop that capability within a much shorter timeframe?

Barúa: I think the main reason people have not been doing separation in the US is not due to the complications and difficulties of the process, but mostly because of the lack of feedstock, the lack of material to process through those plants.

The chemistry used for separation is called solvent extraction, and it is performed every day in the Western world. It is a 50-year-old chemistry, so it is not new. How you optimize it is definitely tricky, not to make it work, but to optimize it. If you put rare earths with a lot of impurities into a separation plant, you are going to have a very tough time. What we have been able to do in Chile and Brazil is produce a carbonate that is 99% pure. If you are feeding the separation plant something that has 99% purity, it is going to reduce the complexity of the separation.

What makes ionic-clay feedstock different from the hard-rock deposits being developed by most Western producers?

Barúa: When you run an extraction on an ionic-clay deposit, you get a much cleaner concentrate.

With a monazite deposit, you have a rock, and you have to break very strong chemical bonds using acid, water, energy, pressure, and temperature. When you break it, you also liberate radioactivity and other things that you then need to deal with. In ionic clays, the ions or atoms of the rare earths are adsorbed onto the clays. They are attached to the surface. That is why our process does not use crushing or milling, because those chemical bonds are very soft.

The grade is lower than in monazite deposits, but the extraction process is incredibly simple. We are going to do it in a controlled environment, so recoveries will be much better. We can be very selective, extracting the rare earths rather than the impurities, which allows us to start with something much cleaner.

How will artificial intelligence help reduce the time needed to ramp up the separation plant?

Barúa: This is a type of problem that is beautifully suited for artificial intelligence. The complexity of the separation plant is that it is a very long process; once you put material in at one end, you need to wait around two months to figure out what happens at the other end. That's why time is so important when you want to change a parameter.

But if you have a digital twin model and you understand how the dynamics of the model work, you can really reduce the amount of time. That is the whole name of the game: who can optimize production from a separation plant as quickly as possible?

The facility requires approximately 500 mixer settlers to be in chemical equilibrium at the same time. That is why the pilot plant at Virginia Tech is so important. It is a full replica of what we are building in Louisiana and allows us to demonstrate the technology, train the workforce and generate the data needed by the computing models that will use AI.

The people operating the plant will have more than two years of experience by the time they get to Louisiana. We believe that will strongly minimize the ramp-up risk, which is probably the hardest part of the supply chain.

Where do you see new sources of demand for rare earths emerging through the 2030s?

Barúa: When we started, we saw demand coming primarily from electric vehicles. We have seen electric vehicle demand slow down a bit in recent years. Electromobility remains important, and wind turbines are also a big source of demand.

What has emerged that was not there before is robotics, not only humanoids but also industrial robots that AI is now making feasible. Robots require a lot of rare earths. That market does not fully exist yet, but it is growing quite rapidly.

For us, yttrium was a byproduct for which we had no market. But with the arrival of AI and the need for data centers, data centers need a lot of energy, and gas turbines use coatings made from zirconia and yttrium.

There are also applications emerging for gadolinium, samarium, ytterbium and other metals sourced primarily from ionic clays in China. Originally, our investment depended on our ability to sell the magnetics. Now we have more products. That starts to approach the Chinese model, because China has a market for all 17 rare earths.

How soon could the West reduce its dependence on China for rare earths?

Barúa: On NdPr, it is sooner. We already have alternatives with Lynas, MP and others.

With heavy rare earths, I think the best alternative right now is Aclara. If we are able to put our foot on the throttle, by the end of 2028, we could be there. It is not going to be easy. It is going to be a challenge, but that is the timeline.