Can Mining Wastewater Become a Source of Rare Earths? A Pilot Study Says Yes – If We Control the Chemistry
In brief
Mining sites often discharge vast amounts of water that contain trace amounts of rare‑earth elements (REEs); the metals behind wind turbines, EV motors, and electronics. A new Swedish‑led pilot study shows that this “wastewater” can be turned into a feeder stream for REE recovery by chaining three steps: EC (electrocoagulation, a low‑chemical pre‑cleaning), sorbents (engineered materials that grab REEs from water), and HFSLM (hollow‑fibre supported liquid membranes, a compact solvent‑extraction membrane). In month‑long pilot runs on real mine water, the approach demonstrated >90% REE recovery under controlled acidity, while using 10–100× less organic solvent than traditional mixer–settler plants, an encouraging signal for greener, local supply of critical materials.
Acronyms explained
- REE(s) – Rare‑earth elements (e.g., La, Nd, Dy, Y).
- EC – Electrocoagulation (uses electricity and aluminium electrodes to form flocs that remove interfering metals).
- HFSLM – Hollow‑fibre supported liquid membrane (a membrane pre‑loaded with an organic extractant—here D2EHPA in kerosene—that moves REEs from dilute feed into a small acid “strip” solution).
- TRL – Technology Readiness Level (scale for maturity; TRL 4–6 indicates pilot validation).
The strategic context: climate tech needs cleaner REE supply
Rare earths underpin high‑efficiency motors, generators and optics; demand is rising, while primary production is concentrated and often solvent‑intensive. Converting dilute, already‑pumped waters from mines and industrial sites into secondary REE sources could diversify supply, cut transport and shrink solvent inventories and acidic wastes, if separation works at ng/L–µg/L concentrations and in the presence of “disturbing” ions like Zn²⁺ and Ca²⁺.
What this new work set out to prove
Within the Mistra TerraClean programme, IVL and partners (Stockholm University, RISE, KTH, Axolot Solutions) tested three modular steps, individually but with integration in mind:
- EC pre‑treatment to remove divalent transition metals while leaving trivalent REEs in solution;
- Functionalised sorbents (silica, lignin‑chitosan, polybenzoxazine‑carbon and amidoxime‑rich hydrochars) to pre‑concentrate REEs;
- HFSLM to selectively separate and up‑concentrate individual REEs at pilot scale.
Pre‑treat once, separate better later
Pilot EC (aluminium electrodes) on Lovisagruvan mine water, spiked at µg/L REE levels, showed a useful selectivity window: at inlet pH ≈ 4.8–5.0, EC removed ~79% of Zn, Pb and Cd while keeping 84–99% of REEs in solution, boosting REE‑to‑transition‑metal ratios by up to two orders of magnitude, exactly what downstream steps need. Above pH ~5.5, that selectivity collapses as both REEs and interferents precipitate; at pH ~8, nearly everything drops out, which is good for discharge but not for enrichment. The process uses electricity and consumable aluminium (common at mine sites) and produces an Al(OH)₃ sludge manageable in existing waste frameworks, though energy optimisation and tight pH control are essential.
EC can pre‑condition large water volumes cheaply, trimming the metal “noise” and letting the REEs flow to the next stage.
Sorbents: picking and packing REEs
Six materials were bench‑marked for La³⁺: amine‑grafted silica (SiAP) reached ~55 mg/g capacity with fast kinetics; D2EHPA‑impregnated silica (D2‑SiAP) traded capacity for ~3× higher affinity, favouring trace‑level capture; diglycolamide‑silica (DG‑SiAP) showed slower uptake; lignin–chitosan (ZLC) was pH‑tolerant but low‑capacity (~5 mg/g); PBz‑modified carbons offered moderate capacity with electro‑swing regeneration potential; amidoxime‑augmented hydrochars (AAC‑1/2) exhibited high theoretical capacities (≈70–80 mg/g) at bench conditions. Regeneration and breakthrough curves remain to be proven in pilot columns—the next logical step.
Different waters, different tools. Silica sorbents serve compact columns with short contact times; PBz carbons promise electro‑regeneration; bio‑based sorbents add renewability. Scaling the best candidates to columns is a near‑term priority.
HFSLM: compact separation with solvent use slashed
The pilot HFSLM (8 m² polypropylene fibres; 10 vol% D2EHPA in kerosene; HNO₃/HCl strip) ran continuously for ~1 month, extracting REEs from tap and real mine water and achieving >90% efficiency when the feed pH was kept below ~2. Under these conditions it demonstrated selective separation among La, Nd, Dy, Y via pH control and cut organic solvent consumption by 10–100× compared with mixer–settlers reported in literature.
Critically, pH control is the master knob: at pH ≈ 3, D2EHPA forms gels/third phases with REE complexes, slashing capacity and even blocking pumps; below pH ~2, stability and transport improve markedly. Extractant losses were small in short trials but matter at scale; periodic replenishment and regeneration (ethanol wash, dry, re‑impregnate) restored ~95% flux.
HFSLM compresses an extraction plant into a vertical cylinder, with a low solvent inventory and high surface‑area‑to‑volume ratio, an enviable fit for modular, near‑source REE harvesting. The operating envelope is tight but navigable.
Why this matters (environment, economics, geopolitics, Nordic relevance)
- Environmental: By design, HFSLM keeps only ~hundreds of millilitres of organic phase inside fibres rather than tons circulating in mixer–settlers, and EC avoids bulk chemical dosing, together promising lower solvent footprints and less acidic waste per kg REE recovered. Early life‑cycle risk mapping (LCBROM) highlights solvent/acid use, membrane replacement, and ethanol‑wash streams as impacts to monitor, but the direction of travel is greener than status‑quo circuits.
- Economic: Mines already pump and treat water; turning that cost centre into a revenue stream improves site economics. The study flags unknowns, energy for EC, membrane lifetime beyond one month, true acid/solvent consumption, but also shows continuous operability and matrix tolerance on Swedish mine water, which reduces integration risk. A techno‑economic analysis is a stated next step.
- Geopolitics: Local REE recovery from Nordic waters would diversify supply for European clean‑tech value chains, reducing exposure to single‑region refining bottlenecks, especially for magnet materials (La/Nd/Dy) that anchor wind and EV growth.
- Nordic innovation: The project links IVL, KTH, SU, RISE, Axolot Solutions, a regional R&D stack capable of pushing from TRL 4–5 to integrated pilots at mines in Sweden/Finland/Norway. Priorities include 3–6‑month runs, sorbent column pilots, alternative diluents to kerosene, and membrane recycling protocols.
What’s proven vs. What’s pending
Proven in this study
- Selective EC window (pH ~4.8–5.2) that enriches REEs by up to 100× relative to Zn/Pb/Cd.
- Bench‑scale sorbents with either high capacity (SiAP; AAC‑1/2) or high affinity (D2‑SiAP), and practical kinetics for column design.
- Pilot HFSLM operating ~1 month with >90% efficiency at pH < 2, showing individual REE selectivity and far lower solvent inventory than conventional extraction.
Still to crack
- Membrane durability beyond a month; quantifying acid/solvent consumption under steady industrial duty.
- Sorbent regeneration and breakthrough over ≥20 cycles; scale‑up to pilot columns.
- Gelation control above pH ~3 (notably with Y–D2EHPA) via chemistry tweaks, alternate diluents/fibres, or process control.
Why this approach deserves scale‑up
Three signals cut through the noise. First, selectivity exists where it counts, EC at mildly acidic pH purges the worst interferents while sparing REEs. Second, pilot‑scale HFSLM works on real water, not just pristine lab solutions, and it does so with a drastically smaller solvent inventory, exactly the lever industry needs to de‑risk environmental performance and footprint. Third, the toolkit is modular: sites can adopt EC for compliance and add sorbents and HFSLM as value‑recovery economics mature. With longer pilot campaigns, quantified consumables, and robust regeneration playbooks, this becomes a realistic pathway for mines to produce REE‑rich concentrates from water they already move, transforming a liability into a strategic resource for the energy transition.
Actionable next steps (for operators, policymakers, investors, researchers)
- Mine operators: Trial EC at pH 4.8–5.2 on side‑streams; sample before/after for Zn/Pb/Cd vs. REE retention; evaluate energy per m³ and sludge handling.
- Pilot consortia: Build sorbent columns (SiAP and AAC‑1/2) ahead of HFSLM; target 3–6‑month runs with continuous TP/TOC and pressure‑drop monitoring as early‑warning KPIs.
- Policymakers: Support field pilots at Nordic mines, paired with rapid permitting for modular HFSLM skids and funding for membrane/diluent alternatives and recycling.
- Investors: Focus on platforms that reduce acid/solvent OPEX and offer retrofit‑friendly modules; diligence on membrane lifetime and regeneration yields is key.
- Researchers: Mechanistic work on third‑phase/gel suppression; electro‑swing regeneration of PBz carbons; techno‑economic and LCA benchmarks against mixer–settlers.
Reference
- IVL Swedish Environmental Research Institute (2025). Can mining wastewater be turned into a source of REE? Separation of Rare Earth Elements in water with potentially disturbing elements (Report C11074). Authors: Strandberg, J., Fischer, S., Tjus, K., Hedman, F., Kärnman, T., Ragnar, M., Mukherjee, S., Edlund, U., Prabhakar, R., Ullah, L., & Hedin, N. Mistra TerraClean. ISBN 978‑91‑7883‑779‑3.
(The report also reviews and builds on prior studies of supported liquid membranes, sorbents and solvent extraction; see its internal bibliography for full details.)
