您当前的位置: 首页 > 网页快照
Solvent-driven fractional crystallization for atom-efficient separation of metal salts from permanent magnet leachates - Nature Communications
Results and discussion .
Experimental approach .
Gaseous DME is employed as a saturation agent to induce crystallization of transition metal or lanthanide sulfates from mixed metal salt magnet leachates. DME is not thought to interact directly with other solutes; instead, DME reduces the quantity of the free water (i.e., water that is not bound within a solvation environment) to fulfill its own hydration requirements 62 , 63 . A reduction in free water, reduced water activity, and/or liquid phase microstructuring 79 induce salt precipitation. This mechanism suggests that the solid salt product compositions would be similar (if not identical) to those produced in energy-intensive evaporative precipitation processes conducted at equivalent temperatures. Increasing reaction chamber head pressure increases the amount of DME dissolved into solution 80 , 81 , resulting in crystallization of a metal salt. The experimental apparatus is depicted conceptually in Fig.? 1a ; a jacketed glass vessel sealed at both ends and connected with fittings, valves, and tubing permits DME sparging of the leachate within the inner chamber while reaction temperature is controlled in the outer chamber via a water bath. In the photograph shown in Fig.? 1b , the solution becomes saturated with DME, inducing crystallization that occurs primarily on the nucleation scaffold in contact with the liquid phase, depleting metal ion salt(s) from the aqueous solution. Within the reactor, we employ stainless-steel mesh as a nucleation scaffold to increase the nucleation density and facilitate the recovery of the crystallization products 82 . Once the liquid phase is evacuated from the reaction column, precipitates are captured on, and subsequently recovered from, the nucleation scaffold. By reducing the pressure exerted on the treated solution, gaseous DME can be recovered efficiently for reuse (see process flowsheet, Fig.? 1c ). The treated solution (Fig.? 1d ) remains unchanged in its chemical character (i.e., pH, salts contained in solution), apart from the removal of precipitated salts, and is thus suitable for reuse in leaching or other hydrometallurgical processes. Experiments were conducted at ~62?psig, pursuant to the internal vapor pressure of the DME tank at ambient temperatures.
Fig. 1: DME-FC apparatus and process schematic. a Schematic depicting the experimental apparatus whereby DME gas is sparged into an aqueous solution at elevated pressure, permitting dissolution of DME into the liquid. Reaction temperature is controlled via a water bath, recirculation is carried out through a gear pump, and crystallization of metal salts occurs on the nucleation scaffold. b Photograph of the experimental apparatus during treatment of the Sm-Co leachate. c Process schematic depicting the DME-FC solid-liquid separation followed by a gas-liquid separation to recover and reuse DME with high efficiency. d Photograph of the experimental apparatus after FC of CoSO 4 from the leachate, showing the visible change in CoSO 4 concentration concurrent with crystal growth on the nucleation scaffold.
Full size image
DME-FC was studied experimentally in conjunction with two separate REE-rich leachates: one leachate was produced from Sm-Co magnet swarf, whereas a more complex leachate was generated from a real-world mixed magnet recycling feed, containing both Nd-Fe-B and Sm-Co magnet swarf. The sensitivity of the solutes to temperature shifts 73 , 74 , 75 (Fig.? 2a ) facilitates the separation of transition metal and lanthanide sulfates from the mixed metal salt solutions. Divergent aqueous solubility vs. temperature trends exist for the metal sulfates contained in the two leachates: from 10 to 50?°C, lanthanide sulfate solubilities decrease with temperature while transition metal sulfate solubilities increase (Fig.? 2a ) 73 , 74 , 75 . Initial metal concentrations in the Sm-Co leachate (Fig.? 2b ) and Nd-Fe-B mixed magnet leachate (Fig.? 2c ) were measured by inductively coupled plasma optical emission spectroscopy (ICP-OES). Experimental observations regarding DME-FC rate and related crystal size suggest that in this system, crystallization rate and crystal size vary with temperature; at lower temperatures (e.g., 20?°C), the crystallization rate is enhanced by the increased concentration of DME in water 80 , 81 as determined by the headspace pressure of the DME tank, ~62?psig 81 . Higher solution viscosity at lower temperatures 83 may also lead to greater turbulence in gas bubble flow and more rapid dissolution of DME into the aqueous phase 84 .
Fig. 2: Solubility vs. temperature trends for dissolved species and initial leachate metal concentrations. a Molal solubility limits vs. temperature for DME 81 (at sufficient pressure to condense a liquid DME phase) and metal sulfates contained in the two studied permanent magnet leachates in the temperature range from 10–50?°C 73 , 74 , 75 . ICP-OES measurement of metal concentrations in the? b Sm-Co magnet leachate and c Nd-Fe-B mixed magnet leachate. Source data are provided as a Source Data file.
Full size image
DME-driven fractional crystallizations under controlled temperatures .
DME-FC was conducted on the leachates at two separate temperatures. In FC treatments applied to both leachates (initial concentrations given in Fig.? 2b, c ) at 31?°C, the increased solubility of transition metal sulfates (FeSO 4 and CoSO 4 ) maintains Fe and Co in solution and motivates crystallization of Ln-rich products (Fig.? 3a, b ). Conversely, treatment of the Sm-Co leachate at a lower temperature of 20?°C leverages the increased solubility of Sm in solution to preferentially crystallize the higher value Co fraction on the stainless-steel scaffold (Fig.? 3a ). This ability to combine solvent-driven and temperature-driven FC is advantageous over evaporatively driven precipitations, where temperature control is more complex. However, it may also be more challenging to control concentration gradients in evaporatively driven processes, as during the evaporation of water, the solute must be redistributed at a diffusion rate that matches or exceeds rates of nucleation processes to ensure uniform behavior 85 . In contrast, DME is distributed through a salt solution more rapidly and uniformly than a precipitation process in the studied systems, as demonstrated by the increase in the aqueous solution volume long before turbidity or macroscopic crystals are observed.
Fig. 3: Compositional data for solid products of DME-FC. a Solid products obtained from DME-FC of the Sm-Co magnet leachate at 20 and 31?°C. b Solid product obtained from DME-FC of the Nd-Fe-B mixed magnet leachate at 31?°C. ICP-OES acquired mass percent compositions are plotted at left (arrows indicate the shift from the original leachate composition to solid product composition). Associated separation factors are listed in Table? 1 . Photographs of the solid products as precipitated on nucleation scaffolds are shown at right. Source data are provided as a Source Data file.
Full size image
Separation efficacy for DME-FC was quantified with a separation factor, \(\alpha\) , a common method to evaluate SX and IX processes 86 , 87 . The separation factor is the ratio of distribution coefficients, \({K}_{d}^{,}\) , as determined in this work by the ratio of metal in the solid product relative to that in the original aqueous phase, Eq. ( 1 ):
$${\alpha }_{{{{{{{\rm{Co}}}}}}}/{{{{{{\rm{Sm}}}}}}}}=\frac{{K}_{{d{{{{{\rm{Co}}}}}}}}^{{\prime} }}{{K}_{{d{{{{{\rm{Sm}}}}}}}}^{{\prime} }}=\frac{\frac{{{{{{{\rm{Co}}}}}}}\,{{{{{{\rm{Mass}}}}}}}\, \% \,{{{{{{\rm{in}}}}}}}\,{{{{{{\rm{Solid}}}}}}}\,{{{{{{\rm{Product}}}}}}}}{{{{{{{\rm{Co}}}}}}}\,{{{{{{\rm{Mass}}}}}}}\, \% \,{{{{{{\rm{in}}}}}}}\,{{{{{{\rm{Initial}}}}}}}\,{{{{{{\rm{Aqueous}}}}}}}}}{\frac{{{{{{{\rm{Sm}}}}}}}\,{{{{{{\rm{Mass}}}}}}}\, \% \,{{{{{{\rm{in}}}}}}}\,{{{{{{\rm{Solid}}}}}}}\,{{{{{{\rm{Product}}}}}}}}{{{{{{{\rm{Sm}}}}}}}\,{{{{{{\rm{Mass}}}}}}}\, \% \,{{{{{{\rm{in}}}}}}}\,{{{{{{\rm{Initial}}}}}}}\,{{{{{{\rm{Aqueous}}}}}}}}}$$
(1)
Separation factors for transition metal—lanthanide separations range from 48–730 (see Table? 1 ), with the highest selectivity achieved in the crystallization of the Sm-rich product from the Sm-Co magnet leachate.
Table 1 Separation factors for DME-FC treatments of Sm-Co magnet leachate and Nd-Fe-B mixed magnet leachate for products depicted in Fig.? 3 . Full size table
X-ray diffraction (XRD) patterns of known references 88 , 89 were compared to experimental XRD datasets obtained for Co-rich solids and Sm-rich solids, showing agreement with CoSO 4 ·6H 2 O 88 and Sm 2 (SO 4 ) 3 ·8H 2 O 89 (Supplementary Fig.? 1 ). These results indicate that the solid products are recovered as sulfates, corresponding to their solubilized form in the aqueous leachate.
To quantify the recovery efficacy of DME-FC in the treatment of the two leachates, recovery fractions were calculated based on the concentration of metals in the original and treated solutions. ICP-OES results for treated Sm-Co leach solutions show 95.9% Co recovery in the 20?°C treatment and 62.5% Sm recovery in the 31?°C treatment (see Supplementary Fig.? 2 ). Results for the Nd-Fe-B mixed magnet leach solution treated at 31?°C show 76.1% Ln recovery. These high metal recoveries are consistent with a process determined by molar solubilities of the salt rather than mass-based solubilities 61 . Experiments were ended once extensive solid product had formed; as such, it is unlikely that these recovery fractions represent a thermodynamic endpoint. Prolonged experiments were avoided; once a fraction of the crystallizing salt has been depleted from leachate, the solution composition has changed such that a different metal salt composition is preferred in crystallization. Under such circumstances, the subsequent FC product is no longer representative of treatment of the initial leach solution, and separate metal salts may crystallize on distinct surfaces (an example optical microscope image is shown in Supplementary Fig.? 3 ). The simultaneous growth of multiple crystal types in spatially distinct locations indicates that the product crystal structure is also a factor in the process selectivity. Further investigation is required to determine the tradeoffs between recovery fraction and product purity and to model the separation mechanism. It is important to note that changes in solution composition do not fully balance with the compositions of the sampled crystals; this may occur due to changes in the solution temperature and pressure during the evacuation of the reaction chamber, precipitation losses within sampling hardware, or simultaneous growth of differing crystals. Complete compositional data for original and treated solutions is tabulated within Supplementary Table? 1 .
In the Co-rich solid product, Sm is largely excluded; however, the Co:Fe ratio is similar to that of the original leachate (11.13 vs. 13.20). This is likely a reflection of the similar crystallographic structures of CoSO 4 and FeSO 4 90 , 91 , which permits their co- crystallization within the same crystal lattice 92 . This is supported by minimal Co and Fe entrainment in the trivalent Sm 2 (SO 4 ) 3 crystalline product, which is not amenable to Co/FeSO 4 crystallization within its lattice. Compositional data indicate that initial solution compositions affect separation factors of crystallizations. Moreover, solute solubilities are more deterministic of crystallized product than the solute concentration or its nearness to saturation. Process temperature also plays an important role; for example, at 20?°C, transition metal sulfates are produced, while at 31?°C, lanthanide sulfates are produced. The results also highlight the importance of product crystal structure, as it is possible to crystallize more than one metal salt within the same solid product (e.g., Fe 0.x Co 0.y SO 4 ).
Sequential fractional crystallizations in stages and passes .
Metal-rich solutions can be treated via two distinct sequential methods, in passes to precipitate distinct fractions from a single solution, and also in stages by dissolving a solid product and treating the resulting solution (see Fig.? 4a ). In DME-FC passes, a solution is exposed to a set of conditions defined by temperature, DME pressure/concentration, and treatment duration/salt recovery fraction to produce an initial precipitate product. This solution can then be treated under differing conditions to produce additional solid product(s). In this way, a lanthanide-rich product can be initially precipitated at higher temperatures and removed from the solution, followed by a subsequent crystallization at lower temperatures to isolate a transition metal-rich product. This protocol is demonstrated in the treatment of the Sm-Co magnet leachate via ICP-OES measurement of metal concentrations in treated solutions in Supplementary Fig.? 4 .
Fig. 4: Sequential treatments in stages and passes in DME-FC. a Scheme depicting stages and passes in DME-FC. DME-FC stages involve treatment of the initial leachate ( \({{{{{{\mathscr{l}}}}}}}_{0}\) ), dissolution of the the solid product (e.g., \({{{{{{\rm{s}}}}}}}_{1}\) ) in water, followed by treatment of the resulting solution to yield a higher-purity solid (e.g., \({{{{{{\rm{s}}}}}}}_{11}\) ). DME-FC passes are successive treatments of the liquid stream to recover chemically distinct solid products ( \({{{{{{\rm{s}}}}}}}_{1}\) , \({{{{{{\rm{s}}}}}}}_{2}\) ) from the same aqueous solution. Temperature control can enhance separations in passes (e.g., initial solid fraction \({{{{{{\rm{s}}}}}}}_{1}\) recovered at temperature \({{{{{{\rm{T}}}}}}}_{1}\) , followed by treatment at a different temperature ( \({{{{{{\rm{T}}}}}}}_{2}\) ) to recover a separate solid fraction ( \({{{{{{\rm{s}}}}}}}_{2}\) )), and can also be used to enhance purification via DME-FC stages. As the chemical character of the solution remains unchanged, albeit with reduced salt concentration, liquid products are suitable for further hydrometallurgical processes upstream or downstream of the separation. ICP-OES metal compositions depicting the purification from original solution ( \({{{{{{\mathscr{l}}}}}}}_{0}\) ), first solid product ( \({{{{{{\rm{s}}}}}}}_{1}\) ), and solid product produced through staging ( \({{{{{{\rm{s}}}}}}}_{11}\) ) with treatments at 25?°C for b Sm-Co magnet leachate and c Nd-Fe-B mixed magnet leachate. Sum of the lanthanide elements (Nd, Pr, Sm, and Dy) is given as Ln. Separation factors for both b and c are listed in Table? 2 . Source data are provided as a Source Data file.
Full size image
In DME-FC stages, initial solid products are dissolved in water and resulting solutions are treated in a subsequent crystallization, resulting in two-stage solid products of increased purity. To investigate this process, an intermediate temperature of 25?°C was selected to promote uniform crystal growth. DME-FC experiments conducted at 25?°C yielded solid products that were dissolved in water to saturation, then treated again under the same conditions (Fig.? 4a ). The sequential product obtained from the Sm-Co leachate was Co-rich, with higher purity than the treatment carried out at 20?°C (Fig.? 4b ). Similarly, sequential treatment of the Nd-Fe-B mixed magnet leachate at 25?°C produced a Fe- and Co-rich solid product with minimal Ln entrainment (Fig.? 4c ). In hydrometallurgical extractions of REEs from primary feedstocks, the leachate contains large amounts of Fe that must be precipitated through pH neutralization 93 or carried over with the lanthanides during SX 18 . DME-FC may provide a means to selectively precipitate Fe from REE-bearing solutions without chemical consumption before SX, reducing the extractant required to load REE onto organic.
Separation factors calculated for two-stage DME-FC are given in Table? 2 . The total separation factors for separating the transition metals from lanthanides were measured to be 528 in the Nd-Fe-B mixed magnet leachate and 102 for the Sm-Co magnet leachate. Separation factors for the two-stage process were measured as 9.69 and 54.5 in the Nd-Fe-B mixed magnet leachate, and 1.96 and 52.3 in the Sm-Co magnet leachate. This progressive increase in separation factor highlights the impact of the initial solution composition on separation efficacy: in DME-FC, selectivity increases with the concentration of the major component. In many separations, the selectivity of the process declines with increasing concentration 94 . In membrane processes, the rejection of the membrane declines as concentration increases 95 , increasing the concentration of minor components in the liquor. When separation factors are relatively independent of concentration due to chemical interactions, as is the case with SX and IX, increasing purity becomes more difficult as the minor component declines. In the case of DME-FC, the separation factor is enhanced as the minor component declines. This ability of crystallization processes to access high purity is commonly used in industrial processes such as float zone refining of single crystal silicon 96 .
Table 2 Separation factors for staged treatments of the Sm-Co magnet leachate and the Nd-Fe-B mixed magnet leachate at 25?°C as depicted in Fig.? 4 . Full size table
Separation factors for transition metal (Co/Fe) separations are relatively low for both systems. Many factors influence these results, including the presence of the lanthanide ions and relative concentrations of Fe and Co in the initial solutions. While Fe and Co concentrations are within an order of magnitude in the studied solutions, molal solubility of CoSO 4 is twice that of FeSO 4 at 25?°C. Given that Fe tends to be incorporated in transition metal-rich solid products (even where it appears as a minor component, as in the Sm-Co magnet leachate), small solubility differences can have a significant impact on DME-FC separations.
In summary, DME-driven FC was demonstrated in the separation of REE and transition metal salts from industrially generated magnet wastes. DME-FC was applied to two separate leachates, one comprising only Sm, Co, and Fe, and a more complex leachate containing Nd, Pr, Dy, Sm, Fe, and Co. Depending on the temperature (20–31?°C), the process can be tailored to selectively yield either transition metal-rich or lanthanide-rich solid products. High recoveries are observed for separations obtained in DME-FC (62.5–95.9% recovery), indicating that high-yield sequential processing can be achieved with a limited number of steps. Staged treatment of the leachate, involving dissolution of the solid products and treatment of the resulting solutions at 25?°C, produced high purity transition metal salt products (>99.5% transition metal). The selectivity of the process increases with the concentration of the major component, suggesting that DME-FC may be an effective processing route to generate high purity products.
DME-driven FC offers non-toxic separation of valuable elements from a mixed salt solution, avoiding requirements of stoichiometric chemical consumption. The selectivity in solid products avoids the energy costs of distillation and the introduction of additional ions to the working fluid (salt metathesis). This process presents opportunities for significant reductions in downstream environmental effects associated with state-of-the-art hydrometallurgical purifications. DME-driven FC is a versatile separation that can be integrated with existing separations such as SX to reduce reagent usage, waste generation, and energy consumption. .
From:
系统抽取对象
机构     
(1)
系统抽取主题     
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)