This report presents the results from testing three separate process modules for the recovery of rare earth elements (REE) from dilute mine and industrial waters. The background to the work is the growing need to develop new, more efficient and environmentally benign methods than those currently available, as existing technologies often suffer from limitations in selectivity, cost-effectiveness and environmental performance. The evaluated modules – electrocoagulation, functionalized sorbents and HFSLM (hollow-fibre supported liquid membrane) – were assessed individually as potential components of a future integrated process chain for REE extraction. The work was carried out within the Mistra TerraClean research programme by IVL in collaboration with researchers from Stockholm University, RISE and KTH. Pretreatment by electrocoagulation Pilot-scale electrocoagulation using aluminium electrodes was tested on mine water from the Lovisa mine spiked with REE.
The process showed selective removal of divalent transition metals (Zn²⁺, Pb²⁺, Cd²⁺), while trivalent lanthanides remained in solution. This selectivity increased the REE/transition metal ratio by up to two orders of magnitude, potentially reducing the burden on subsequent process steps by lowering the concentration of interfering elements. Energy consumption was identified as a primary concern requiring optimisation, and the narrow pH window (4.8–5.2) demands careful control for successful implementation. Development of functionalized sorbents Six sorbent materials were synthesised and evaluated at bench scale for uptake of lanthanum (La³⁺) from aqueous solutions. Aminopropyl-functionalised silica (SiAP) exhibited the highest adsorption capacity at 55 mg/g La³⁺, while D2EHPA-impregnated silica (D2-SiAP) showed a three-fold increase in affinity despite lower overall capacity.
Lignin-chitosan composites displayed broad pH tolerance but limited capacity (5 mg/g), and benzoxazine-based carbons showed moderate performance with good chemical stability. All sorbents were tested only in single-element solutions under controlled conditions. Regeneration protocols and breakthrough behaviour remain uncharacterised, and no scale-up to pilot columns has yet been conducted. HFSLM Separation at Pilot Scale HFSLM was tested using real mine water spiked with REE concentrations ranging from ng/L to µg/L. One month of continuous operation was achieved with over 90% efficiency, representing a significant improvement compared with previously published results. Through careful pH control, selective separation of individual rare earth elements was demonstrated in the tested matrices.
The technique reduced solvent consumption by one to two orders of magnitude compared with conventional mixer–settler systems reported in the literature. pH control was found to be critical, including for maintaining membrane function, as pH levels above 3 led to gel formation, substantially reducing capacity. Losses of the extractant (D2EHPA in kerosene) were relatively small during operation, although some replenishment will be required upon scale-up. Reducing extractant losses will lower operating costs—given the high cost of the chemical—and decrease overall solvent use. Integrated assessment Overall, this research provides important insights into membrane-based recovery of rare earth elements (REE) from dilute streams and demonstrates that selective separation of individual REE is achievable through pH control. However, the technologies are still in a developmental phase.
Realising their full potential will require addressing the operational and economic challenges identified here, particularly those related to long-term stability and cost-effectiveness. At the same time, the compact system design, reduced consumption of organic solvents and the demonstrated selectivity constitute strong incentives for continued development. Conclusions The research demonstrates the technical feasibility of individual process modules for REE recovery from dilute streams. Each module shows potential within specific operational windows—electrocoagulation for selective metal removal at controlled pH, sorbents for pre-concentration of REE and HFSLM for final separation and up-concentration.
The modules, however, are positioned at different technology readiness levels (TRL): HFSLM is demonstrated at pilot scale (TRL 4–5), electrocoagulation partially at pilot scale (TRL 5–6) and the sorbents remain at bench scale (TRL 3–4). Progress towards industrial implementation will require evaluating the technologies within an integrated process context, assessing their economic viability and demonstrating long-term operational performance. Priorities include scaling up the most promising sorbents to pilot columns, extending HFSLM testing to 3–6 months and subsequently testing all three modules in series based on the actual water chemistry. The results constitute valuable proof-of-concept data but should be interpreted as highly promising early-stage research requiring substantial further development before industrial application is feasible.
Stockholm: IVL Svenska Miljöinstitutet, 2025.