Scientists look to diatoms and proteins to pull rare earths from waste

Coal ash, red mud, and mine tailings are usually treated as pollution to be contained rather than as a resource. A research team led by Worcester Polytechnic Institute (WPI) has received a $3.3 million award from the National Science Foundation's Growing Convergence Research program to see whether biological processes can recover the silica, rare earth elements, and other critical minerals trapped inside these waste streams, using less energy and fewer harsh chemicals than conventional methods.

The five-year, two-phase project is led by Mingjiang Tao of WPI's Department of Civil, Environmental, and Architectural Engineering, with Carrick Eggleston and Yan Wang as co-principal investigators. Researchers from George Mason University, the University of California San Diego, the University of Massachusetts Amherst, and the University at Buffalo are also involved. Tao said the goal is to use as much of each waste stream as possible, separating out valuable elements while turning the rest into useful products.

The scale of what is at stake is large. The team estimates that U.S. coal ash landfills alone hold about 11 million tons of rare earth elements, worth an estimated $8.4 billion, which is nearly eight times the country's current raw domestic reserves, according to the WPI announcement. These elements are needed for electronics, clean-energy technologies, transportation, and national security applications.

The WPI-led project plans to draw on organisms that already do something related to what the researchers want: diatoms, sea sponges, and certain plants use biological molecules and organic scaffolds to capture dissolved silicon and assemble it into intricate silica structures under mild conditions, rather than the high heat and heavy chemical processing that industry currently relies on. The team wants to adapt these mechanisms into lower-energy methods that break down the silica-rich parts of industrial waste, convert that silica into useful materials, and free the rare earth elements bound up inside. The work combines biology, geochemistry, materials science, metallurgy, and computational chemistry, and the team plans to use computational modeling and artificial intelligence to design biomolecules and predict how they interact with waste materials.

A different but related approach already has experimental results behind it. A team reported in Frontiers in Bioengineering and Biotechnology describes a protein-based biosorbent, a material designed to grab specific ions out of a solution, built from what they call thermo-responsive genetically encoded REE-selective elastin-like polypeptides, or RELPs. These proteins bind rare earth elements and can then be separated out simply by cycling the temperature up and down, rather than through chemical washes. In their tests, the RELPs retained about 95% of their original binding capacity even after four reuse cycles, and could raise the purity of a rare earth solution by up to 100,000-fold when isolating one representative element from a simulated low-concentration mixture.

The RELP work builds on the discovery of a natural protein called lanmodulin, which the same research group had previously identified as an effective and highly selective binder of light rare earth elements. Earlier attempts to make lanmodulin reusable by locking it onto solid beads reduced how well it worked: one method left the protein retaining only about 67% of its original binding activity, and another preserved about 80%. The RELP method avoids that immobilization step, relying instead on the protein's own temperature-driven behavior to separate it from solution.

Beyond engineered proteins, a range of microorganisms — bacteria, algae, fungi, yeast, and archaea — are already known to interact with rare earth elements in nature, and researchers are exploring ways to harness that for recovery, according to a mini-review in a PubMed Central-hosted journal focused on microalgae. The review describes three general mechanisms: bioleaching, in which organisms produce acids that dissolve metals out of a solid material; biosorption, a passive process where the elements stick to binding sites on cell walls; and bioaccumulation, an active process in which living cells take metals inside themselves and store them, often using specialized molecules to keep the metals from causing harm. The review notes that a low recovery rate and limited selectivity have so far restricted how far these microbial methods can be scaled up, though economic analyses reportedly suggest bio-recovery from waste could still be cost-effective.

Coal and its byproducts are getting particular attention as a target for this kind of recovery. A review published in the Journal of Environmental Management estimates that coal ash could potentially yield about 312,000 tons of rare earth elements annually, an amount the authors say far exceeds global demand, and surveys extraction approaches including green solvents, nature-based methods, and bioleaching and biosorption. Separately, the WPI announcement points to red mud, mine tailings, concrete debris, waste glass, and metallurgical slag as further silicon-rich waste streams containing untapped critical minerals.

Much of this work remains at an early or exploratory stage. The WPI-led project is only just beginning its five-year timeline, and the team says it still needs to evaluate whether the technologies it develops can be scaled up economically for industrial use. The RELP protein study describes itself as the first direct experimental evidence of this particular protein recovering rare earth elements from a real low-grade waste feedstock in batch testing, a more limited setting than continuous industrial operation.

The microalgae review is explicit that low recovery rates and poor selectivity are still holding back large-scale use of microbial methods, even as researchers pursue strain engineering and other strategies to improve efficiency. Reviews of the broader field also list a wide range of competing techniques — ionic liquids, deep eutectic solvents, chemical roasting, and others — without settling on which, if any, will prove most practical at scale. Taken together, the sources describe a set of promising early results and active research directions rather than a proven industrial process.