Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2026 Sep 15;11(38):58045–58053. doi: 10.1021/acsomega.6c07645

Kinetic Control of Rare Earth Element Separation via Interfacial Solvent Extraction and Low-Gradient Magnetic Fields

Bailey Lake †,‡,§, Theo Siegrist ‡,§,∥, Hadi Mohammadigoushki ‡,§,∥, Munir Humayun ‡,§,⊥, Jamel Ali ‡,§,∥,*
PMCID: PMC13629374  PMID: 42824741

Abstract

The production of energy-efficient technologies and advanced electronics requires high-purity critical minerals, such as individually separated rare earth elements (REEs). Chemical REE separation processes rely on small differences in physicochemical properties (i.e., ionic charge and radii) to transport selective metal ions across liquid–liquid interfaces, making selective separation time-consuming and expensive. Incorporating another intrinsic property, ionic magnetic moment, alongside ionic radius may improve interfacial transport. We report here a novel batch solvent extraction system, termed magnetically assisted solvent extraction (MA-SX), controlled by the orientation of low-gradient magnetic fields generated by permanent magnets. We observe that the concentration of paramagnetic REE ions and the extractant ligand, di­(2-ethylhexyl)­phosphoric acid (D2EHPA), significantly influence ion transfer kinetics across an aqueous–organic interface under varying magnetic field gradient orientations. Comparisons among REEs with different magnetic moments show that MA-SX behavior correlates with ionic magnetic moments. Extraction of binary mixtures containing Dy­(III) and Y­(III) suggests paramagnetic and diamagnetic ions co-aggregate at low concentrations, reducing separation in the presence of low-gradient magnetic fields. At total metal concentrations above 50 mM, Dy­(III) and Y­(III) partition, with Y­(III) preferentially extracting into the organic phase due to D2EHPA’s slight affinity toward Y­(III).

graphic file with name ao6c07645_0010.webp

graphic file with name ao6c07645_0009.webp

1. Introduction

The demand for a U.S. domestic critical mineral supply chain is recent motivation for creating efficient and sustainable rare earth element (REE) separation techniques. Solvent extraction, also referred to as liquid–liquid extraction, is currently the most common method for separating individual REEs. In traditional liquid–liquid extraction (T-LLE), an aqueous phase containing the metal ion(s) is mixed with an organic phase containing the extractant(s). After mixing, the phases settle, and selective metal–ligand complexes reside in the organic phase. T-LLE of selective REE ions is based on small differences in ionic radii among REEs, making efficient industrial-scale separation a challenge requiring several extraction stages. However, differences in the ionic magnetic moments of REE far outweigh the differences in ionic radii and charge (Table S1) and could be used to enhance the extraction kinetics and selectivity in REE separation systems.

Magnetic separation and field-induced cluster formation of species larger than ions (e.g., colloidal systems, magnetic nanoparticles) have been investigated over the past two decades. − Recent reports on magnetic separation and magnetomigration of paramagnetic ions show a similar clustering behavior occurring with ions in aqueous solutions and silica gels exposed to magnetic field gradients. − However, magnetomigration of dissolved paramagnetic ions in aqueous solutions has been shown to occur over relatively long time scales because the magnetic force acting on individual hydrated ions is extremely small compared to thermal motion and diffusion. Previous work has reported only a 20% decrease in paramagnetic ion concentration in regions away from the magnet after several days and a 0.5–2% concentration increase near the magnet after 90 min, demonstrating that magnetic field gradients alone are generally insufficient for rapid or practical separation of dissolved rare-earth ions. , Force balance analysis indicates that the migration/capture of individual ions by a magnetic force is insufficient and that the observed effect must arise from interactions of ion clusters where the combined magnetic forces are large enough to overcome competing forces (e.g., diffusive and viscous forces). − Magnetic field gradients (B·∇B) have proven to be the dominant influence in particle separation, but the particle interactions and aggregation mechanism are not fully understood.

The potential of combining magnetic and chemical separation mechanisms has been demonstrated by Higgins et al., who selectively crystallized heavy rare-earth elements through temperature-induced concentration gradients while applying external magnetic field gradients. Their study reported separation factors approximately twofold higher for heavy and light rare-earth element pairs exhibiting large differences in magnetic susceptibility. These findings demonstrate that integrating magnetic field gradients with established chemical separation processes may offer an effective strategy for improving rare-earth separations.

In addition to induced cluster formation, magnetic field gradients can induce secondary flows and convection within paramagnetic fluids. , An enhancement in convection near an interface of a solvent extraction system will enhance mixing, thereby increasing extraction kinetics. This effect is particularly important in solvent extraction because even modest increases in interfacial mixing can accelerate metal–ligand encounters and increase overall mass transfer beyond what would occur by diffusive transport alone.

Here, we combine T-LLE with low-gradient magnetic fields (B·∇B < 100 T2/m) to explore the transfer kinetics of ions in individual and binary metal solutions across an aqueous–organic interface where the organic phase contains the ligand di­(2-ethylhexyl)­phosphoric acid (D2EHPA), industrially known as extractant P204, and the aqueous phase contains the metal analyte(s). There are two pathways by which a metal ion can be extracted by D2EHPA: (1) metal complexation at the interface; (2) D2EHPA transfer into the aqueous phase boundary layer followed by diffusion back into the organic phase as a metal–ligand complex. , High extraction efficiencies of a paramagnetic metal ion by D2EHPA into an organic phase result in a bulk paramagnetic fluid responsive to external magnetic field gradients. This magnetic response could enable bulk magnetically driven motion for enhanced phase contact.

In this work, we investigate how the orientation of low-gradient magnetic fields and the ligand and metal concentrations alter extraction kinetics of paramagnetic ions. Here, extraction kinetics refer to the rate at which extraction occurs and the factors that govern how rapidly the system approaches equilibrium. To confirm the correlation of the magnetic effects with the magnetic moment, REE ions with varying magnetic moments were employed in a magnetically assisted solvent extraction (MA-SX) system. To explore the magnetic separation of binary mixtures, aqueous solutions containing paramagnetic and diamagnetic REE ions were employed.

2. Materials and Methods

The magnetic field gradients were produced by 2 × 2 × 1-inch N52 magnets purchased from K&J Magnetics, Inc. A 3-axis magnetic field meter was used to collect the magnetic field data at different distances away from the side face of the magnet. The magnetic field gradients (B·∇B) were calculated using eq , and the experimental data collected with the field meter.

(B·∇B)=BxdBxdx 1

The organic phase was formed by dissolving bis­(2-ethylhexyl) phosphate (97%) in anhydrous dichloromethane purchased from Millipore Sigma and used without further purification. The hydrated metal chloride salts, DyCl3·6H2O, GdCl3·6H2O, NdCl3·6H2O, and YCl3·6H2O, were supplied by Millipore Sigma or Chem-Impex. The salts were dissolved in 0.05 M HNO3 (trace metal grade) to create individual and binary metal aqueous solutions. The two phases were added to a 10 mm pathlength quartz cuvette. The quartz cuvette was rinsed with HNO3 to render a hydrophilic surface, producing a vertical aqueous–organic interface (Figure S1) for small sample volumes.

For all experiments, except where the effects of ligand concentration were investigated, 450 μL of the organic phase (1000 mM D2EHPA) was added to a cuvette followed by 450 μL of the aqueous phase. After placing a disposable cuvette cap on the cuvette, it was immediately transferred to the 3D-printed holder designed with corner slots for the cuvettes around the magnet, as shown in Figure S2 of the Supporting Information. At time points of 5, 20, 40, 60, and 80 min, 5 μL of the aqueous phase was removed and diluted to 5 mL with 2 % HNO3 to be analyzed by inductively coupled plasma optical emission spectrometry (Agilent 5800 ICP-OES). All extraction experiments were performed in triplicate, and the reported standard deviations represent the variability among the three independent experiments. To minimize potential effects of concentration gradients within the unstirred aqueous phase, aliquots for ICP-OES analysis were consistently collected from approximately the center of the aqueous phase using the same sampling procedure for all experiments. The reproducibility of triplicate measurements, reflected by the low standard deviations, indicates that the sampling approach provided representative concentration measurements for each experimental condition. The concentrations at each time point as well as the initial concentration were used to calculate the percent extractions (% E) according to eq :

%E=[Mn+]0−[Mn+]t[Mn+]0×100 2

where [M n+]0 and [M n+] t are the REE ion concentrations in the aqueous phase at the start and at the time of sample collection, respectively.

The same procedure was used to conduct experiments on the effect of the ligand concentration except the volume of each phase was decreased to 400 μL of each phase to maintain a vertical interface with the lower concentrations of D2EHPA.

3. Results and Discussion

3.1. Effect of Magnet Field Gradients

To evaluate how magnetic field gradients and their orientation influence the extraction of paramagnetic REE ions, the interface was placed in two orientations relative to the permanent magnet’s field and compared to a control extraction performed without a magnetic field. Figure illustrates the two positions of the magnetic extraction systems. The organic phase is closest to the magnet in Position 1, and the aqueous phase is closest to the magnet in Position 2. Importantly, the two phases are not mixed to solely address the transfer of ions across a mechanically static interface. Therefore, the extraction is initiated by concentration gradients and progressed by the complexation of the metal with D2EHPA at the interface or in the aqueous phase boundary layer.

3.

3

Open in a new tab

Percent extractions of Dy­(III) by 1000 mM D2EHPA from an aqueous solution of 10 mM DyCl3 without a magnetic field (control) and in two position relative to the magnetic field.

The magnetic fields and field gradients produced by the permanent magnets were measured. In Figure (a) the magnetic field and gradient intensity 1.3 mm away from the side face of the magnet, where the first phase starts due to the thickness of the quartz cuvette walls (1.25 mm), are shown. The magnetic field gradient reaches a maximum of 55 T2/m at the edge of the magnet. In Figure (b) the magnetic field and gradients 6.1 mm away from the side face, approximately where the vertical interface sits in a 10 mm pathlength cuvette, are shown. At the interface, the magnetic field gradients reached only 4.5 T2/m. As shown in Figure , the cuvettes were placed at the corners of the magnets where the magnetic field gradients are highest (Figure ). Because paramagnetic ions rapidly align with the local magnetic field, the observed magnetic effects are governed by the field magnitude and gradient rather than the polarity of the magnet. Therefore, the orientation of the north or south pole toward the extraction system is not expected to influence the reported results. The magnetic effects these extraction systems experience are classified as a low-gradient magnetophoresis process as the gradients are less than 100 T2/m.

2.

2

Open in a new tab

Experimental data of the magnetic field strength (left y-axis) and gradient (right y-axis) at (a) 1.3 mm away from the side face and approximately the beginning of the phase closest to the magnet (b) 6.1 mm away from the side face approximately at the organic–aqueous interface. Shaded region indicates location of the cuvette (Z = 0 mm).

1.

1

Open in a new tab

Schematic for the magnetically controlled extraction of paramagnetic rare earth ions from the aqueous phase to the organic phase. (left) The organic phase closest to the magnet. (right) The aqueous phase closest to the magnet.

Although the magnetic field gradients are considered to be low, they had an unexpectedly large effect on the extraction kinetics when 10 mM DyCl3·6H2O was contacted with 1000 mM D2EHPA. At 40 minutes, Position 1 enhanced the extraction kinetics by 12 %, and Position 2 depleted the extraction by 30 % compared to the control (no magnetic field).

It is evident by the extraction results of the control that additional flows are generated within the system enhancing the contact of Dy­(III) with the interface beyond what would occur by thermally driven displacement alone. The thermally driven displacement of an individual Dy­(III) ion due to Brownian motion is

⟨x2⟩=2Dt 3

where t is time (s) and ⟨x 2⟩ is the particle’s mean squared displacement. In 1 hour an individual Dy­(III) ion would only have an average displacement less than half the distance the aqueous phase takes up in the cuvette (approximately 5 mm of the 10 mm pathlength cuvette), suggesting that less than half the ions in the aqueous phase could travel to the interface to be extracted by D2EHPA. Because the controlled extraction (no magnetic field) of Dy­(III) in 1 h is much over 50 % extraction (82% extraction), as seen in Figure , there must be flows generated within the systems enhancing the contact of Dy­(III) with the interface. The dimensionless Peclet number (Pe) describing the relationship between convection and diffusion can be expressed as

Pe=NconvNdiff=L·uDDy(III) 4

where L is the length scale, u is the aqueous phase velocity magnitude, and D Dy(III) is the diffusivity of trivalent dysprosium in water (5.57 × 10–6 cm2/s) from Table S1. Initially, because the bulk fluid velocity of the aqueous phase is approximately 0 cm/s, Pe is close to zero. Although diffusion is the dominating force for the movement of the bulk aqueous phase, there are additional factors increasing the extraction of Dy­(III) at the interface from what would be expected by diffusion alone: (1) interfacial turbulence due to Marangoni effects or undesired mechanical agitation; (2) interfacial convection due to local concentration and density changes. −

The enhancement of the MA-SX in Position 1 is likely due to increased secondary flows in the organic phase upon the extraction of Dy­(III). It is hypothesized that once enough metal is extracted into the organic boundary layer at the interfacial region, the organic phase is then mixed at a higher rate than the control due to magnetically induced convection. The significantly lower extraction rate observed for Position 2 is likely due to an increased accumulation of metal–ligand complexes at the interface as they are directed toward the region of higher magnetic field gradients. Because these complexes are not soluble in the aqueous phase, they remain localized at the interface, where their increased interfacial population may reduce the rate at which fresh extractant molecules participate in interfacial exchange. One possible contributor to the reduced extraction kinetics observed in Position 2 is the magnetomigration of paramagnetic ions toward regions of higher magnetic field gradients within the aqueous phase. However, previous work by another group reports only a 0.5–2% increase in local ion concentration near the magnet after 80 min, depending on the initial concentration, suggesting that magnetomigration alone is unlikely to account for the magnitude of the observed decrease in extraction kinetics. We instead propose that the magnetic force opposes transport into the organic phase, increasing the residence time or local concentration of paramagnetic metal–extractant complexes in the organic phase interfacial region. While the interface may already be substantially populated with complexes, an increased local accumulation of these complexes could reduce the rate at which fresh extractant molecules participate in interfacial exchange, thereby slowing the overall extraction kinetics without completely inhibiting mass transfer. Figure is a schematic of the proposed MA-SX pathways of Dy­(III) in the relative positions to the magnet.

4.

4

Open in a new tab

Schematic of the proposed extraction pathways for the MA-SX system in (left) Position 1 and (right) Position 2.

To confirm that the observed behavior in Figure arises from the combined chemical and magnetic effects of the MA-SX process, the transfer of Dy­(III) into dichloromethane in the absence of D2EHPA was evaluated (Figure S3). Under the conditions, only 7–13% of Dy­(III) was transferred from the aqueous phase to the organic phase, and no clear influence of the magnetic field was observed over the investigated time points. Additionally, MA-SX experiments for the extraction of 10 mM Dy­(III) using 1000 mM D2EHPA were conducted with a horizontal aqueous–organic interface (Figure S4). Because of the sample holder geometry, only magnetic Position 1 could be investigated by placing the cuvette on the top corner of the magnet. In this configuration, Position 1 produced an extraction percentage approximately 10 % higher than the control at minute 40. The overall extraction rates observed for both the control and Position 1 with the horizontal interface are likely enhanced by increased mixing during sample preparation, particularly when the aqueous phase is introduced into the organic phase. The limitation of only observing Position 1 with this sample holder motivated the implementation of the vertical interface MA-SX system for straightforward investigation of magnetic field gradient orientation effects by simply rotating the cuvette prior to placement adjacent to the magnet.

3.2. Effect of Dy Concentration on the Percent Extraction

Across all studied Dy­(III) concentrations (100 mM, 50 mM, 10 mM, and 1 mM), the influence of magnetic field gradients was observed for either Position 1 or Position 2. At the lowest concentration, 1 mM DyCl3, Position 1 produced a clear extraction kinetic enhancement across multiple time points. At 20, 40, and 60 min, the percent extraction of 1 mM Dy­(III) was 78.3, 93.8, and 97.4%, respectively in magnetic Position 1, while the control displays 68.4, 87.5, and 94.0% at the same time points (Figure ). The increased transport of Dy­(III) in Position 1 into the organic phase at such low concentrations may suggest enhancement due to magnetically induced convection rather than direct magnetomigration. In contrast, Position 2 shows no extraction depletion at this concentration, indicating that too few paramagnetic metal–ligand complexes build up at the interface to impede extraction kinetics.

5.

5

Open in a new tab

Percent extraction of (a) 100 mM, (b) 50 mM, (c) 10 mM, and (d) 1 mM DyCl3 by 1000 mM D2EHPA without a magnetic field (control) and in two positions relative to the magnetic field.

As discussed in Section , there was both an enhancement of 12% extraction and depletion of 30% extraction in Positions 1 and 2, respectively, compared to the control, at 10 mM DyCl3. At 50 mM DyCl3, both magnetic positions influenced the kinetics. For the orientation of Position 1, the extraction enhancement appears primarily in the early stages (within 5 min) and diminishes near 20 min, where the extractions of MA-SX in Position 1 and control converge due to rapid extraction kinetics. At minute 5, the extraction percentage reaches 67.1% without the presence of external field gradients and 77.4% when the organic phase is in the region of higher magnetic field gradients (Position 1). The extraction remained suppressed for 80 min in Position 2. At intermediate Dy­(III) concentrations (10–50 mM), the magnetic enhancement is likely the result of a balance between increased formation of paramagnetic metal–ligand complexes and sufficient time for the magnetic field gradient to influence their transport. Under these conditions, the interfacial concentration of complexes is high enough for magnetic effects to become significant, while the extraction kinetics have not yet reached the rapid approach to steady state observed at the highest Dy­(III) concentration.

At the highest concentration observed, 100 mM DyCl3, Position 1 shows no deviation from the control, both reaching an extraction steady state around 84 % extraction just after 5 min. The increased kinetics with higher metal loading enables the system to approach steady state rapidly, limiting the duration over which the magnetic field gradient-induced effects can contribute to mass transfer. This decrease in the overall extraction efficiency reflects saturation of the organic phase boundary layer at lower ligand-to-metal ratios around 10:1 (1000 mM D2EHPA:100 mM Dy­(III)). A supersaturated interface prevents complete Dy­(III) extraction regardless of the magnetic field gradient orientation. A similar behavior has been reported for the extraction of Ho­(III) in a magnetic solvent extraction system. At 100 mM DyCl3, Position 2 behaves similarly to the extraction of 50 mM Dy­(III). There was a significant depletion in the extraction kinetics lasting for the duration of this kinetic investigation.

These results show that the orientation of magnetic position 1 enhances extraction when the kinetics are not already rapid or inhibited by interfacial saturation. Position 2 consistently slows extraction, where interfacial accumulation may slow mass transfer.

3.3. Effect of Ligand Concentration

The concentration of D2EHPA also had a large influence not only on the extraction efficiency but on the magnitude of the influence of the magnetic-assisted separation. The percent extractions of Dy­(III) over 6 h from a 10 mM DyCl3 solution by three different ligand concentrations (e.g. 1000 mM, 100 mM, and 10 mM D2EHPA) are reported in Figure . The 10 mM D2EHPA extraction system reached an equilibrium state at 12–14% extraction before minute 30. There were not enough paramagnetic ions complexed at the interface at that point to be affected by the magnet field gradients and, hence, no effects of magnetic Positions 1 and 2 with 10 mM D2EHPA. The influence of the magnet was clear in the first 3 h of extraction by 100 mM D2EHPA, but all extraction systems, the control and the 2 magnetic positions, reach equivalent steady states around 70 % Dy­(III) extraction. Extraction by 1000 mM D2EHPA reached 99% Dy­(III) extraction in the control and magnetic Position 1 in 3 h, while only 78% Dy­(III) was extracted in magnetic Position 2. All extraction systems with 1000 mM D2EHPA reach equivalent steady states within 6 h. As the D2EHPA concentration increases, a greater fraction of Dy­(III) ions can rapidly form paramagnetic metal–ligand complexes at the interface, increasing the population of species susceptible to magnetic field gradient effects. This increase in interfacial complex formation likely contributes to the enhanced magnetic response observed at extractant concentrations greater than 10 mM D2EHPA. These results support our hypothesis that the magnetic field gradient orientation imposed at the beginning of the experiment influences the extraction kinetics, while the equilibrium extraction remains unchanged.

6.

6

Open in a new tab

Influence of ligand concentration (1000 mM, 100 mM, and 10 mM D2EHPA) on the percent extraction of Dy­(III) without a magnetic field (control) and in two positions relative to the magnetic field. HL represents the extractant D2EHPA.

3.4. Effect of Analyte Magnetic Susceptibility

The percent extractions of metal ion with varying magnetic moments from individual metal solutions in the absence of an applied magnetic field (control) and in magnetic Positions 1 and 2 are shown in Figure (a–d). The difference in the extraction of the control and the two magnetic positions increases as the ionic magnetic moments increase from Y­(III) to Dy­(III).

7.

7

Open in a new tab

Percent extraction of analytes with different ionic magnetic moments by 1000 mM D2EHPA with no magnetic field (control) and in 2 positions relative to the magnetic field. (a) Dy­(III) (10.6), (b) Gd­(III) (7.9), (c) Nd­(III) (3.7), (d) Y­(III) (0). (e) The difference in the percent extraction of Position 1 (P1) and Position 2 (P2) at 40 min for the different REE ions.

Figure (e) displays the differences in the percent extractions of MA-SX in Position 1 (P1) and Position 2 (P2) versus the magnetic moments of metal ions in the individual metal solutions. For Position 1, the extraction differences relative to the control for Dy­(III), Gd­(III), Nd­(III), and Y­(III) (at 40 min) were 12.36%, 12.31%, 5.69%, and 1.47%, respectively, while for Position 2 the differences were 29.65%, 24.33%, 6.93%, and 4.30%, respectively. In both Position 1 and Position 2, the magnitude of the extraction difference generally increased with ionic magnetic moment, with the smallest effect observed for diamagnetic Y­(III) and the largest effects observed for the highly paramagnetic Gd­(III) and Dy­(III). This trend suggests that the magnetic field effect is correlated with the magnetic properties of the extracted ion. The effective magnetic moments of REE in this system dictate the intensity of their magnetic response. Dy­(III), among the highest magnetic moments of REE, experiences the strongest magnetic force, thereby exhibiting the greatest magnetic effect on extraction efficiency. Gd­(III), with a moment of 7.9 μB, is also considered strongly paramagnetic and was significantly influenced by magnetic-assisted separation. While the Nd­(III) extraction was still influenced by the presence of the magnetic field gradients, we observe a noticeably weaker response, as expected. Lastly, employing the effectively diamagnetic ion, Y­(III), resulted in no meaningful interaction with the applied magnetic field. The transport of Y­(III) across the aqueous–organic interface was governed solely by the chemical extractant D2EHPA. Notably, Dy­(III) and Y­(III) have almost identical ionic radii. Figure (a) and (d) show that these two ions behave similarly in the control position (no magnetic field) as expected for a chemical extraction based exclusively on ionic radii and charge. The direct correlation of magnetic effects to the ionic magnetic moment of the species in the MA-SX system suggests its potential for separation of highly paramagnetic REE ions from those with a lower magnetic response.

3.5. Extraction of a Paramagnetic and Diamagnetic Mixture

In industrial processes, metal ions with different magnetic moments often exist as mixtures. Utilizing these differences to separate them from one another would be beneficial in developing an efficient and selective separation of such mixtures. Here, we examine the extraction and separation of Dy­(III) from Y­(III) from a binary mixture of DyCl3 and YCl3. Figure shows the extraction percentages of mixtures with 4 different initial concentrations, where Dy­(III) and Y­(III) equally make up the total concentration. For example, in the system with a 100 mM binary mixture, there was 50 mM DyCl3 and 50 mM YCl3.

8.

8

Open in a new tab

Percent extraction of Dy­(III) and Y­(III) from different initial total metal concentrations of a binary mixture containing DyCl3 and YCl3 by 1000 mM D2EHPA without a magnetic field (control) and in two positions relative to the magnetic field. (a) 100 mM, (b) 50 mM, (c) 10 mM, and (d) 1 mM.

At 1 mM total metal concentrations, no differences between the control or magnetic orientation are observed due to the low concentration of the paramagnetic species (0.05 mM DyCl3). For the intermediate total mixture concentrations of 50 mM and 10 mM, both Dy­(III) and Y­(III) exhibit identical extraction rates across the control and both magnetic orientations. This consistency suggests that the two ions exist together in clusters at equivalent ratios regardless of D2EHPA’s slight affinity toward Y­(III), the differences in their magnetic properties, and the orientation of the magnetic field gradients. Such co-aggregation of paramagnetic and diamagnetic ions was also observed by Benhal et al. when capturing transition metal ions in a high gradient magnetic separator. It has been unclear how magnetic field gradients effect the ratio of ions in clusters. One might expect that paramagnetic ions would cluster together in field gradients due to dipole–dipole interactions and because the field pulls paramagnetic particles toward the same region. However, these data suggest not only that Dy­(III) and Y­(III) exist together in clusters but also that the clusters have equal amounts of the respective ions, as shown by their identical extraction percentages. This indicates that cluster formation in the Dy–Y–Cl system is governed primarily by ionic radii and ionic charge, rather than by magnetic properties.

At an initial total metal concentration of 100 mM, the chemical selectivity of D2EHPA becomes apparent. The control and Position 1 show identical extraction rates for Dy­(III) and likewise for Y­(III), but the values for Y­(III) are consistently higher than those for Dy­(III). However, Position 2 follows the same behavior as the lower concentrations, demonstrating that Dy­(III) and Y­(III) complexes are clustering together, but the slightly higher extraction percentage of Y­(III) relative to Dy­(III) further suggests that these clusters may not contain equal amounts of the two metal ions. Overall, it was not until higher concentrations that we observe partitioning of Dy­(III) and Y­(III), which arises from D2EHPA’s chemical preference.

Conclusions

In summary, we report the influence of low-gradient magnetic fields on the extraction of REE ions, specifically Dy­(III), Gd­(III), Nd­(III), and Y­(III), across aqueous–organic interfaces. Previously, magnetically induced changes in the extraction behavior have been reported. , However, the present work examines the influence of the magnetic field gradient orientation on the kinetics of metal-ion transfer. Overall, we have shown the ability to control the kinetics of the transfer of metal ions across an aqueous–organic interface in a batch magnetically assisted solvent extraction (MA-SX) system with the orientation of magnetic field gradients. The effect of the magnetic field gradients increases with concentration of the paramagnetic ion Dy­(III) until the aqueous–organic interface becomes fully saturated, decreasing the equilibrium extraction percentage near a ligand-to-metal ratio of 10:1. The greatest difference between the magnet Position 1, where we expected to see enhanced extraction, and magnet Position 2, where we expected to have a depleted extraction percent compared to the control, is with a ligand to metal ratio of 20:1, 50 mM DyCl3, and 1000 mM D2EHPA. Similarly, the ligand concentration plays a dominant role in the magnetic effect on the extraction efficiency of the system. Increasing the ligand concentration results in a prolonged and enhanced extraction efficiency by the magnetic field. Here, it was also confirmed that the kinetics of the extraction can be controlled but the extraction equilibrium is unchanged by the application of a magnetic gradient. Additionally, the greater the magnetic moment of the metal analyte in single-element extraction, the larger the effect of the magnetic field gradients on the extraction kinetics. However, when paramagnetic and diamagnetic ions (i.e., Dy­(III) and Y­(III)) are extracted from low concentrations (10–50 mM) of binary mixtures, they are extracted at the same rate and are equivalently affected by the magnetic field gradients, suggesting that they exist in equal amounts within clusters. Overall, the kinetic hindrance of Position 2 may offer a new separation mechanism, whereby REE extraction is selectively delayed relative to other elements, potentially enhancing separation factors under nonequilibrium operating conditions, a concept that will be examined in future work. Future efforts should also aim to use both chemical and magnetic methods to improve the separation of ions with significantly different ionic magnetic moments.

Supplementary Material

ao6c07645_si_001.pdf (260.4KB, pdf)

Acknowledgments

The methods and results presented in this manuscript are covered by a provisional patent application filed with the U.S. Patent and Trademark Office on October 6, 2025. The authors acknowledge Muhammad Garba for collecting the magnetic field strength data used to determine the magnetic field gradients of the 2 × 2 × 1-inch N52 magnets. Research was sponsored by the Army Research Office and was accomplished under Grant Number W911NF-22-1-0265. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the Army Research Office or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein. A portion of this work was performed at the National High Magnetic Field Laboratory, which is supported by the National Science Foundation Cooperative Agreement No. DMR-2128556 and the State of Florida. This work was supported by the Center for Rare Earths, Critical Minerals, and Industrial Byproducts, through funding provided by the State of Florida.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c07645.

  • Additional experimental details, including control experiment data and images of the experimental setup (PDF)

The authors declare no competing financial interest.

References

  1. Xie F., Zhang T. A., Dreisinger D., Doyle F.. A Critical Review on Solvent Extraction of Rare Earths from Aqueous Solutions. Min. Eng. 2014;56:10. doi: 10.1016/j.mineng.2013.10.021. [DOI] [Google Scholar]
  2. Lake B., Siegrist T., Albrecht T. E., Mohammadigoushki H., Humayun M., Ali J.. Recent Advances in Rare Earth Element Recovery: Liquid–Liquid Extraction and Magnetophoretic Separation. Ind. Eng. Chem. Res. 2025;64:19781. doi: 10.1021/acs.iecr.5c02137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Tsouris C., Scott T. C.. Flocculation of Paramagnetic Particles in a Magnetic Field. J. Colloid Interface Sci. 1995;171(2):319–330. doi: 10.1006/jcis.1995.1186. [DOI] [Google Scholar]
  4. Faraudo J., Andreu J. S., Calero C., Camacho J.. Predicting the Self-Assembly of Superparamagnetic Colloids under Magnetic Fields. Adv. Funct. Mater. 2016;26(22):3837–3858. doi: 10.1002/adfm.201504839. [DOI] [Google Scholar]
  5. De Las Cuevas G., Faraudo J., Camacho J.. Low-Gradient Magnetophoresis through Field-Induced Reversible Aggregation. J. Phys. Chem. C. 2008;112(4):945–950. doi: 10.1021/jp0755286. [DOI] [Google Scholar]
  6. Andreu J. S., Camacho J., Faraudo J.. Aggregation of Superparamagnetic Colloids in Magnetic Fields: The Quest for the Equilibrium State. Soft Matter. 2011;7(6):2336–2339. doi: 10.1039/c0sm01424a. [DOI] [Google Scholar]
  7. Ditsch A., Lindenmann S., Laibinis P. E., Wang D. I. C., Hatton T. A.. High-Gradient Magnetic Separation of Magnetic Nanoclusters. Ind. Eng. Chem. Res. 2005;44(17):6824–6836. doi: 10.1021/ie048841s. [DOI] [Google Scholar]
  8. Moeser G. D., Roach K. A., Green W. H., Alan Hatton T., Laibinis P. E.. High-Gradient Magnetic Separation of Coated Magnetic Nanoparticles. AIChE J. 2004;50(11):2835–2848. doi: 10.1002/aic.10270. [DOI] [Google Scholar]
  9. Fujiwara M., Kodoi D., Duan W., Tanimoto Y.. Separation of Transition Metal Ions in an Inhomogeneous Magnetic Field. J. Phys. Chem. B. 2001;105(17):3343–3345. doi: 10.1021/jp003562d. [DOI] [Google Scholar]
  10. Fujiwara M., Chie K., Sawai J., Shimizu D., Tanimoto Y.. On the Movement of Paramagnetic Ions in an Inhomogeneous Magnetic Field. J. Phys. Chem. B. 2004;108(11):3531–3534. doi: 10.1021/jp0303523. [DOI] [Google Scholar]
  11. Fujiwara M., Mitsuda K., Tanimoto Y.. Movement and Diffusion of Paramagnetic Ions in a Magnetic Field. J. Phys. Chem. B. 2006;110(28):13965–13969. doi: 10.1021/jp061482p. [DOI] [PubMed] [Google Scholar]
  12. Benhal P., Garba M., Ali J., Siegrist T., Humayun M., Mohammadigoushki H.. Dynamics of Transition Metal Ion Transport in High-Gradient Magnetic Fields. J. Phys. Chem. A. 2025;129(15):3401–3410. doi: 10.1021/acs.jpca.4c08312. [DOI] [PubMed] [Google Scholar]
  13. Rassolov P., Ali J., Siegrist T., Humayun M., Mohammadigoushki H.. Magnetophoresis of Paramagnetic Metal Ions in Porous Media. SOFT MATTER. 2024;20(11):2496–2508. doi: 10.1039/D3SM01607B. [DOI] [PubMed] [Google Scholar]
  14. Kolczyk K., Kutyla D., Wojnicki M., Cristofolini A., Kowalik R., Zabinski P.. Separation of Rare Earth Metals Ions in a Static Magnetic Field. Magnetohydrodynamics. 2016;52(4):541–547. doi: 10.22364/mhd.52.4.11. [DOI] [Google Scholar]
  15. Kołczyk K., Wojnicki M., Kutyła D., Kowalik R., Żabiński P., Cristofolini A.. Separation of Ho3+ in Static Magnetic Field. Arch. Metall. Mater. 2016;61(4):1919–1924. doi: 10.1515/amm-2016-0308. [DOI] [Google Scholar]
  16. Higgins R. F., Cheisson T., Cole B. E., Manor B. C., Carroll P. J., Schelter E. J.. Magnetic Field Directed Rare-Earth Separations. Angew. Chem.-Int. Ed. 2020;59(5):1851–1856. doi: 10.1002/anie.201911606. [DOI] [PubMed] [Google Scholar]
  17. Butcher T. A., Coey J. M. D.. Magnetic Forces in Paramagnetic Fluids. J. Phys. Condens. Matter. 2023;35(5):053002. doi: 10.1088/1361-648X/aca37f. [DOI] [PubMed] [Google Scholar]
  18. Ortmann K., Bidmon A., Eckert K., Lei Z.. Increasing Rare Earth Extraction Kinetics by Tailored Magnetic Fields. Commun. Phys. 2025;8(1):427. doi: 10.1038/s42005-025-02393-7. [DOI] [Google Scholar]
  19. Sun P., Binter E. A., Liang Z., Brown M. A., Gelis A. V., Benjamin I., Bera M. K., Lin B., Bu W., Schlossman M. L.. Antagonistic Role of Aqueous Complexation in the Solvent Extraction and Separation of Rare Earth Ions. ACS Cent. Sci. 2021;7(11):1908–1918. doi: 10.1021/acscentsci.1c00960. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Stevens, G. W. ; Perera, J. M. ; Grieser, F. . Interfacial Aspects of Metal Ion Extraction in Liquid-Liquid Systems. Rev. Chem. Eng. 2001, 17 10.1515/REVCE.2001.17.2.87. [DOI] [Google Scholar]
  21. De Las Cuevas, G. ; Faraudo, J. ; Camacho, J. . Low-Gradient Magnetophoresis through Field-Induced Reversible Aggregation. J. Phys. Chem. C 2008, 112, 945–950 10.1021/jp0755286. [DOI] [Google Scholar]
  22. Berg, H. C. Random Walks in Biology; Princeton University Press, 2025. [Google Scholar]
  23. Wojtaszek, K. ; Cristofolini, A. ; Popoli, A. ; Siedlecka, K. K. ; Socha, R. P. ; Owińska, M. ; Wojnicki, M. . Influence of Magnetic Field on the Kinetics of Ho­(III) Solvent Extraction Using D2EHPA: A Comprehensive Study. Small Methods 2025, 9, 2402002 10.1002/smtd.202402002. [DOI] [PubMed] [Google Scholar]
  24. Berg J.. Interfacial Hydrodynamics: An Over View. Can. Metall. Q. 1982;21(2):121–136. doi: 10.1179/cmq.1982.21.2.121. [DOI] [Google Scholar]
  25. Berg J. C., Morig C. R.. Density Effects in Interfacial Convection. Chem. Eng. Sci. 1969;24(6):937–946. doi: 10.1016/0009-2509(69)87002-8. [DOI] [Google Scholar]
  26. Berg J. C., Haselberger G. S.. Mass Transfer during Interfacial Convection. Chem. Eng. Sci. 1971;26(3):481–485. doi: 10.1016/0009-2509(71)83020-8. [DOI] [Google Scholar]
  27. Wojtaszek K., Kołczyk-Siedlecka K., Cebula F., Michalek T., Wojnicki M.. Analysis of the Kinetics Ions Ho­(III), Pr­(III), Gd­(III) Transport at the Liquid-Liquid Interface in the Presence of a Magnetic Field Gradient. Magnetohydrodyn. 2022;58(4):515–521. doi: 10.22364/mhd.58.4.16. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao6c07645_si_001.pdf (260.4KB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES