Engineered Artificial Minerals (EnAMs): Concept, Design Strategies, and Case Studies

Abstract

With the continuous development of easily accessible resources, the exploitation of complex mineral resources, metallurgical waste slag containing high-value metals, and secondary resources is gradually becoming a mainstream trend. Due to the complex distribution characteristics of elements in these resources, efficient recycling is difficult to achieve. A phase reconstruction strategy has been proposed to address the distribution forms of elements. The phase reconstruction strategy employs pyrometallurgical methods to subject complex resources to high-temperature smelting and cooling crystallization. In the cooling crystallization process, the target elements in melt are selectively enriched into engineered artificial minerals (EnAMs). Then, the target elements can be recovered by subsequently separating these EnAMs. However, the concept of and design strategies for EnAMs are still unclear. In this review, the concept of EnAMs is proposed based on previous studies. This review explores how to design EnAMs by phase equilibrium studies and utilizing geochemical behaviors. Additionally, the application cases of EnAMs in treating challenging tantalum–niobium and rare earth element (REE) resources, secondary resource recycling, and metallurgical slag were collected. Furthermore, the challenges and future perspectives of EnAMs for complex resources are discussed.

1. Introduction

Mineral resources are the necessary material foundation for humanity’s advancement towards modernization, automation, and intelligence. Most cutting-edge technologies, such as the batteries of electric vehicles, the high-temperature alloys used in the aerospace field, the semiconductor materials used in 5G communication, and the rare earth permanent magnets used for artificial intelligence robots, rely on the stable supply of key metal resources. The occurrence of mineral resources not only determines the path and speed of technological development but also profoundly affects the layout of global industrial chains and resource competition. According to statistics, over 50% of lithium carbonate resources in China are used for lithium battery production [1], and 56% of platinum-group metals (PGMs) in the world are used in automotive catalysts [2]. However, with the increasing depletion of high-grade primary mineral resources and the acceleration of global industrialization, resource shortages and environmental pollution problems are becoming increasingly prominent. The traditional linear economic model of “resource consumption–product manufacturing–pollution emission” is increasingly struggling to sustain the needs of sustainable development. Resource development is gradually shifting to handle complex mineral resources, metallurgical waste slag rich in high-value metals, and secondary resources in urban mines. The comprehensive utilization of these three types of resources has become a breakthrough important to solving resource constraints and achieving a circular economy.

Complex mineral resources, especially rare and precious metal resources, are characterized by low grades, fine particle sizes, and symbiosis with other elements. Many resources have huge reserves, but they cannot be developed and utilized due to the limitations of resource development technologies. Taking niobium resources as an example, the current niobium resource output from Brazil and Canada accounts for more than 90% of the world’s total [3]. Actually, there are many large niobium deposits outside of Brazil and Canada [4], such as the Bonga deposit in Angola (824 Mt, 0.48% Nb₂O₅), the Mabounie deposit in Gabon (401.2 Mt at 1.11% Nb₂O₅), the Ghurayyah deposit in Saudi Arabia, and Bayan Obo in China (2.2 Mt at 0.13% Nb₂O₅) [5].

In the field of secondary resource recycling, the circular economy concept centered on secondary resource regeneration has become a global consensus. Valuable elements in the secondary resource can be recovered, such as in the recovery of precious metals (gold, silver, rare earth elements, and copper), as well as harmful substances like lead and mercury, from electronic waste (e.g., discarded appliances or printed circuit boards) [6,7,8]; lithium from spent lithium-ion batteries [9,10]; and platinum from spent catalysts [11]. The concentration of platinum-group metals (PGMs), lithium, and cobalt in urban mines (such as electronic waste, spent catalysts, and spent lithium-ion batteries) is significantly higher than that in primary ores [12,13,14]. For instance, mass-produced computer motherboards contain approximately 200–250 g·t⁻¹ of gold and about 80 g·t⁻¹ of palladium, while mobile phones can contain up to 350 g·t⁻¹ of gold and 130 g·t⁻¹ of palladium [15]. The content in used ceramic catalytic converters and spent metal catalysts has reached 2596 g·t⁻¹ and 7874 g·t⁻¹, respectively [16]. This is clearly much higher than the gold or PGM content in primary ores (averaging < 10 g·t⁻¹) [12]. Aside from PGMs, the content of rare metals in rechargeable batteries and hybrid/electric vehicle (HEV) batteries overs 180 g·kg⁻¹ [17,18,19]. Permanent NdFeB magnets in hard disks, microphones, and speakers contain neodymium (Nd, 190–350 g·kg⁻¹) and dysprosium (Dy, 14–45 g·kg⁻¹) [19,20]. Therefore, an economic advantage arises from the higher concentrations of valuable metals in electronic and electrical waste compared to ores. E-waste recycling offers a way to recover rare earth metals and other critical raw materials, reducing primary production [21,22,23]. With the advancement of new energy technologies, it is expected that by 2030 the market for rechargeable batteries in electric vehicles will exceed 7000 GWh, while the overall battery market will surpass 10,000 GWh [24], resulting in even more secondary resources. In 2019, approximately 53.6 million tons of e-waste were generated worldwide, and it is estimated that this amount will increase to around 74 million tons by 2030. If current trends continue, the amount could even rise to 120 million tonnes by 2050. In Germany, the amount of electrical and electronic equipment waste (WEEE) was 12.5 kg per capita, slightly above the EU average (10.5 kg per capita). This large volume of e-waste to be processed and the complex recycling processes need the further development of e-waste recycling and integration across the technical, economic, social, and environmental spheres [25,26].

In addition to complex mineral resources, metal resources in considerable amounts are contained in metallurgical waste slag. The term metallurgical waste slag refers to solid waste generated during pyrometallurgical processes, such as steel slag, blast furnace slag, and copper slag; e.g., fayalitic slags, which contain valuable elements such as scandium (Sc), titanium (Ti), nickel (Ni), copper (Cu), and iron (Fe), as well as harmful elements such as chromium (Cr) and lead (Pb) [27]. Originated from the smelting of electronic scrap incorporated into copper production, fayalitic slag is mainly formed by SiO₂ and FeO, but often contains CaO and other oxides, which consequently implies a high oxygen content. Since most valuable metals are present in diluted concentrations and show high oxygen affinity, the consequent loss of metallurgical waste is inevitable. According to statistics, smelteries globally produce approximately 190–290 million tons of steel slag and 70 million tons of copper slag annually [28]. However, the recovery rate of metals in global metallurgical waste slag is low, and the problems of resource waste and environmental pollution coexist.

To address the challenges of reusing complex mineral resources and metallurgical waste rich in high-value metals, as well as secondary resources in urban mines, the re-crystallization strategy is employed to change the distribution forms of elements, as shown in Figure 1. After a rough concentration process, target elements in these resources are induced to a directional and selective enrichment in melts, resulting in what are known as engineered artificial minerals (EnAMs). Then, the target elements can be recovered by subsequently separating these EnAMs. This strategy provides a solution for the beneficiation and recycling of complex resources. In this review, the concepts, design methods, and application cases of EnAMs are summarized. The challenges of and future perspectives related to EnAMs are also proposed at the end of the manuscript.

2. Concept of EnAMs

Engineered artificial minerals (EnAMs) are synthetic mineral phases with a high element concentration in slag created through high-temperature melting–cooling crystallization or phase transformation processes, which disrupt the original elemental distribution states [29,30,31]. By artificially designing and controlling the chemical composition of metallurgical slags, EnAMs enable the targeted centralized enrichment of specific elements. During the high-temperature melting stage in the heating process, chemical bonds within the raw materials are broken, resulting in a disordered atomic arrangement. The subsequent crystallization stage during the cooling process provides opportunities for atomic reorganization. Through the precise regulation of slag composition, tailored physicochemical environments are established to guide the selective combination of target elements, ultimately forming highly concentrated mineral phases. EnAMs mimic natural mineral formation processes, sharing similarities with the cooling and solidification of magmatic melts, including elemental partitioning between phases, migration dynamics, dissolution–precipitation behaviors, and component volatilization. However, critical differences exist between artificial and natural mineralization, particularly in terms of pressure conditions (e.g., atmospheric vs. crustal hyperpressure environments) and formation timescales (days vs. geological epochs). Hence, the design and synthesis of EnAMs require interdisciplinary collaboration across metallurgical engineering, chemistry, ore-forming geochemistry, and mineralogy to optimize phase engineering under controlled industrial conditions.

3. Design of EnAMs

Because the formation mechanisms of engineered artificial minerals (EnAMs) share similarities and differences with those of natural minerals, the design of EnAMs can be achieved by phase equilibria studies and simulating ore-forming geochemical environments, as shown in Figure 2. When the simulated physicochemical ore-forming environment satisfies the conditions for mineral formation, the designed artificial minerals may exhibit phase consistency or crystal structure consistency with natural minerals. When the composition and properties of the raw materials differ from those during elemental mineralization, new phases can be formed as carriers for element enrichment. Therefore, the design of EnAMs primarily involves considering phase equilibrium relationships and simulating ore-forming geochemical environments.

3.1. Phase Equilibrium Study

Phase equilibrium is an important theory for studying the equilibrium state of matter in multiphase systems [32,33]. In the design of EnAMs, phase equilibrium studies play a decisive role in understanding the phases formed by key elements during the cooling crystallization process. Phase diagrams based on the phase equilibrium provide a visual understanding of the types and quantities of phases that may exist in a system, as well as their interconversion relationships, under different temperature, pressure, and composition conditions [34]. Phase diagrams can be obtained from thermodynamic calculation and experimental analysis. Phase stability and phase diagram predictions are primarily conducted using thermodynamic software such as FactSage, Pandat, and the OpenCalphad (OCA) package. The experimental analysis mainly includes characterizing phases and their composition in different conditions through characterization tools, including X-ray powder diffraction and electron probe microanalysis.

According to the compositions of resources, a phase diagram can be obtained through experiments and calculations. The corresponding formation conditions of enriched phases, such as the composition and temperature range, are also determined in the phase diagram. Taking a common ternary system A-B-C phase diagram as an example, by analyzing the various phase regions within the diagram, it is possible to determine that the target phase can be formed when the three components A, B, and C are within a specific compositional range and the temperature is controlled within a certain interval. In studies on the recovery and utilization of smelting slag containing titanium resources, high-temperature melting–cooling crystallization is employed to transform titanium-containing phases. Phase equilibrium studies based on the CaO-SiO₂-TiO₂ system can determine that titanium is mainly enriched in perovskite, as shown in Figure 3 [35]. Perovskite should be chosen as a targeted EnAM for recycling titanium. By further adding the components of the phase diagram, the changing trends in the perovskite crystallization region and crystallization temperature can be obtained in complex multi-component systems containing MgO [36,37,38], Al₂O₃ [39], and Cr₂O₃ [35]. In multi-component systems, the interaction between different elements can affect the crystal structure and chemical composition of the phases. For example, adding 5 wt% Cr₂O₃ to the CaO-SiO₂-TiO₂ ternary phase diagram proves that the addition of Cr₂O₃ promotes the crystallization of perovskite by extending the primary phase region of perovskite to a higher SiO₂ region, thereby facilitating the recovery of Ti [35]. Additionally, the influence of oxygen partial pressure on crystallization can also be determined by a CaO-SiO₂-Ti_xO_y phase diagram [40,41,42]. In the study of crystallization behaviors of lithium-bearing slag, the existence of manganese (Mn) affects the phases of lithium-bearing EnAMs [43]. The presence of manganese and its valence changes will form Li-Mn-Al spinel solid solutions or lithium manganates [44]. By adding MgO and Al₂O₃, hausmannite is effectively suppressed and a solid solution between the spinels of Mn²⁺Mn³⁺₂O₄, MgAl₂O₄ and Mn²⁺Al₂O₄ is formed [45]. The calculated equilibrium solidification process of the Li₂O-Al₂O₃-SiO₂-CaO-MnO system proves that γ-lithium aluminate as primarily crystalline was precipitated as the temperature decreased [46]. When the temperature decreases to around 1350 °C, only γ-lithium aluminate forms a liquid. Hence, controlling the crystallization temperature can effectively control the crystalline phases.

In addition to being based on resource composition, phase diagrams can also be designed based on the composition of anticipated EnAMs. The designed phase diagrams are used to study phase equilibrium relationships and determine crystallization conditions. Then, in the actual batching process, the corresponding raw materials can be accurately weighed according to this compositional range to ensure that the subsequent crystallization process proceeds in the direction of forming the target phase. For example, if titanium-containing resources are to be converted into ilmenite (TiFeO₃) minerals or pseudobrookite, the phase equilibrium relationship can be studied by designing a multi-component system phase diagram containing iron (Fe²⁺ or Fe³⁺) [47,48]. As shown in Figure 3, the designed phase diagram determines the crystallization interval for the formation of TiFeO₃ and the corresponding relative content of each component. Additionally, when constructing apatite (Ca₅(PO₄)₃(F,Cl,OH)) as target EnAMs in a targeting slag containing rare earth elements (REEs), the precipitation region of apatite can be determined through phase equilibrium studies of multi-component systems containing phosphorus, calcium, and rare earths [49].

Phase diagrams also can guide the cooling process [50]. The cooling process is a critical step in the formation of EnAMs, and phase diagrams provide important guidance for controlling the cooling process. Different cooling rates and cooling paths lead to different degrees of atomic diffusion and rearrangement, which in turn affects the final formed phases [51,52]. For example, the long-term storage of stainless-steel slag containing a large amount of chromium can cause hexavalent chromium to pollute the environment. Therefore, designing EnAMs enriched in chromium can promote the green manufacturing practice of the metallurgical industry. When chromium slag reaches phase equilibrium in the range of 800–1600 °C, chromium is mainly enriched in the amorphous phase, diopside and akermanite, and spinel [53]. Based on a comprehensive judgment of the enrichment rate of chromium in EnAMs and the economy of crystallization temperature, it is considered that 800 °C is conducive to the maximum stabilization of Cr in the spinel phase [53]. Therefore, the slag/melt can be quickly cooled down to 800 °C to allow the spinel phase to preferentially precipitate, avoiding chromium in diopside and akermanite. A reasonable cooling regime can be formulated based on the crystallization temperature range and transformation temperature of each phase in the phase diagram. The accurate control of the cooling process based on the phase diagram can realize the effective adjustment of EnAMs.

3.2. Utilizing Geochemical Behaviors

In the formation of natural minerals, the distribution of most rare metals is relatively concentrated. For example, pyrochlore is the main carrier mineral for niobium. The carrier minerals of niobium in the Brazilian Catalão and Araxá mines, and the Canadian Niobec niobium mine are pyrochlore [54]. Gallium resources are mainly concentrated in bauxite deposits [55,56]. Therefore, the design of EnAMs for rare elements can refer to the ore-forming characteristics of elements. Based on the ore-forming environment of minerals, such as temperature, pressure, composition, and redox environment [57], a simulated geochemical formation environment should be constructed. This method is limited in its application due to economic constraints. In addition, simulating the geochemical process environment of ore formation is difficult to construct. However, in the design process of artificial minerals, the ore-forming principle of geochemical and the physical-chemical environment can still provide support in terms of design direction. Simultaneously, when artificial minerals are consistent with that of natural minerals, the existing beneficiation, sorting, and separation technologies are mature, leading to high applicability.

1.

Simulating Element Affinity

According to Goldschmidt’s classification of elements based on their geochemical behavior during ore formation and preferred host phases, elements are broadly divided based on their affinity into lithophile, siderophile, chalcophile and atmophile elements [58,59,60]. Based on the geochemical affinity of the element, lithophile, siderophile, and chalcophile elements preferentially enter silicate melts, metallic (iron) phases, or sulfides, respectively, during phase partitioning. Lithophile elements (Li, Na, K, Mg, Ca, Al, Si, etc.) tend to combine with oxygen and are enriched in the silicates of the earth’s crust and mantle. Siderophile elements (Fe, Ni, Co, Mn, Mo, W, Au, etc.) readily combine with metallic iron and are enriched in the earth’s core. Chalcophile elements (Cu, Zn, As, Se, Pb, S, Ag, Cd, In, Sn, Sb, Te, Hg, Bi, Tl, etc.) tend to combine with sulfur, forming sulfide minerals. The affinity of Ru, Rh, Pd, Re, Os, Ir, and Pt can be either chalcophile or siderophile, depending on the ore-forming environment. The element affinity in the actual pyrometallurgy is similar with geochemistry, as shown in Figure 4 [61]. lithophile elements correspond to metallurgical slag, such as blast furnace slag, converter slag, refining slag, etc. These slags often include more than two kinds of Al₂O₃, SiO₂, FeO, MgO, CaO. The chalcophile elements correspond to matte in the metallurgical process, usually copper matte or nickel/cobalt matte (such as Cu₂S, Ni₃S₂, CoS). siderophile correspond to ferrophiles or alloys. The atmophile elements correspond to the gas phase in the metallurgical process.

Based on the Goldschmidt classification, the path for designing target minerals has a general direction. When designing chalcophile elements, EnAMs can be designed as metal sulfides formed in low-oxygen content environments. The transformation and recovery of metals can be achieved. For example, lead, arsenic, silver, gold, platinum, and copper transform into metal sulfides as matte, which can be separated from slag [62,63,64]. When treating copper smelting slag under pressurized oxygen atmosphere and low acidity conditions, iron and silicon in the slag transforms and reconstructs into hematite and amorphous silica, respectively [65]. Regarding iron-containing slag, the iron in the slag can form EnAMs in the form of oxides under oxygen-rich conditions, such as hematite (Fe₂O₃) [66] or (Mn,Mg)_xFe_3−xO₄ [67]. Under reducing conditions, using iron as an element enrichment carrier can form iron-based alloy type to achieve element enrichment, such as Mo [68].

2.

Simulating Geochemical Environments of Ore Formation

First, it is necessary to identify the main minerals of the element in nature. Based on mineralogical information of the element carrier minerals, the geochemical environments of ore formation can be determined. In spite of the geochemical environment of element ore formation is difficult to simulate during application, the partial environmental characteristic of temperature, melt acidity/basicity, slag composition, and redox conditions during the element ore formation process can be simulated through pyrometallurgical methods. In terms of compositional design, in addition to determining the multi-component basicity, the addition of appropriate non-metallic elements to the slag may increase or decrease the migration rate of elements in the melt, indirectly achieving element enrichment. For example, rare earth elements (REEs) migrate more efficiently in fluids when the slag is in the condition of sulfate-rich [69].

In geochemistry, according to the different main components in magma, the types of magma deposit mainly include ultramafic rocks [70], carbonatites [71], alkaline rocks [72,73], etc. For example, alkaline rocks contain a high total alkali content (K₂O + Na₂O), and alkaline rocks are enriched in large ion lithophile elements (Ba and Rb), high field strength elements (Y, Zr, and Nb), and rare earth elements (REEs). Although the geochemical behavior of ore formation in magma cannot provide accurate guidance for EnAMs, it can provide research directions during the design process. For example, when designing engineered artificial niobium-bearing minerals (EnANMs), niobium mainly exists in the form of pyrochlore minerals, and pyrochlore is a common mineral in carbonatite and alkaline rock magma mineralization [74]. Generally, pyrochlore can locally accumulate in large quantities in carbonatite-alkaline rocks. Therefore, it can be determined that the ore-forming environment of pyrochlore is related to the alkalinity of the melt. In the study of pyrochlore solubility, pyrochlore group minerals are more common in peralkaline granite systems, while columbite is in peraluminous systems [75]. This is mainly because the solubility of pyrochlore in magma changes with the ratio of Al₂O₃/CaO+Na₂O+K₂O (A/CNK). The change in A/CNK value is basically consistent with the change in basicity in the melt. In the actual niobium concentrate conversion process, the transformation of niobium into EnANMs with perovskite or pyrochlore crystal structures can be achieved by controlling the content of alkaline oxides and acidic oxides in the slag [76]. In addition, according to the element composition of natural pyrochlore group minerals, the high field strength elements Nb, Ta, and Ti have similarities in the distribution of mineral types [77,78,79]. Similar crystallization behavior and migration paths may appear during the crystallization of the Nb, Ta, and Ti containing slag multi-element system. The geochemistry behaviors of pyrochlore group in different compositions in Figure 5 show that betafite exists in low Nb/Ti ratio and pyrochlore exists in high Nb ratio [77,80]. In the actual transformation behaviors in slag as shown in Figure 5b, as improving the Nb/Ti ratio of slag, the niobium-bearing phase transforms from calciobetafite to pyrochlore [81]. In addition, Thomas Schirmer et al. studied Ta and Ti containing slag and found that the behavior of Ta in synthetic slag was similar to that of Ti [30]. Ta and Ti are enriched in the corresponding perovskite. This result also illustrates that the crystallization behavior of some elements is similar with geochemical behavior. There are few cases of designing artificial minerals by simulating ore-forming geochemical environments, and further research is still needed to discover the connections and differences between EnAMs crystallization and nature ore-forming processes.

4. Applications and Case Studies

4.1. Challenging Niobium Resources

The Bayan Obo niobium resource is characterized by a wide variety of niobium-bearing minerals, complex intergrowth relationships, and fine mineral particle size [82,83,84,85,86]. However, the complex mineralogical properties have hindered the efficient development and utilization of niobium through conventional physical beneficiation processes and equipment [87]. The Nb₂O₅ grade of the Bayan Obo low-grade niobium rougher concentrate can only be increased from 0.1%–0.2% to 1%–5% via physical beneficiation [87]. Addressing the challenge of inefficient development and utilization of the Bayan Obo niobium resources, a novel three-step combined beneficiation and metallurgy approach of “selective and precise pre-concentration, mineral phase transformation, industrial-scale EnAMs enrichment” has been proposed to overcome the mineralogical complexity. Firstly, to enhance the niobium grade, the raw materials are pre-enriched to obtain low-grade niobium rougher concentrate. Then, the low-grade niobium rougher concentrate undergoes mineral phase transformation, converting various niobium-bearing minerals into engineered artificial niobium-bearing minerals (EnANMs). Finally, the transformed slag containing EnANMs is crushed and enriched through beneficiation to achieve the utilization of niobium resources.

To clarify the crystallization behavior of niobium-bearing slag during the cooling process, the CaO-SiO₂-Nb₂O₅-La₂O₃ multi-component system was designed for optimizing slag design [88,89,90,91,92,93,94]. CaNb₂O₆, LaNbO₄, Ca₂Nb₂O₇, and the liquid phase are the main carriers of niobium after mineral phase transformation. CaNb₂O₆ preferentially precipitates in the low La₂O₃ region and under high CaO/SiO₂ ratio conditions. Ca₂Nb₂O₇ is stable at 1200 °C with medium CaO content. To further clarify the crystallization behavior of niobium in the slag under multi-component conditions, Lifeng Sun et al. conducted research on the CaO-SiO₂-Nb₂O₅-Fe₂O₃-TiO₂ five-component phase diagram at 1100 °C and 1200 °C, respectively [95,96]. At 1200 °C, the coexistence of CaTiO₃ and Ca₁₀Nb₂Si₆O₂₇ is dominant. At 1100 °C, a large amount of CaSiO₃ phase precipitates, inhibiting the formation of CaTiO₃. The influences of TiO₂ on the precipitates of niobium-bearing phase can be obtained. TiO₂ competes with Nb₂O₅ to enter the perovskite (CaTiO₃) phase, forming a Ca(Ti,Nb)O₃ solid solution. In addition, niobium partially transfers to Ca₂Nb₂O₇. Multi-component phase equilibrium studies have revealed that the crystallized niobium minerals after phase equilibrium mainly include perovskite structure, pyrochlore structure, and liquid phase.

Based on the theoretical foundation of phase equilibrium, Zhang Bo et al. conducted a mineral phase reconstruction study on Bayan Obo tailings by controlling the cooling crystallization method after high-temperature smelting reduction [97,98]. After carbothermal smelting reduction in Bayan Obo tailings at 1600 °C, the iron forms spherical iron beads, which are completely separated from the slag. After smelting reduction crystallization regulation of Bayan Obo tailings at 1600 °C, perovskite, geikielite, forsterite precipitated in the slag as shown in Figure 6a [98]. Through elemental distribution analysis, Nb, Ti and REEs are enriched in perovskite. Nb-bearing minerals including niobite, pyrochlore, and bastnaesite in the tailings were mainly transformed into perovskite phase. REEs bearing minerals, including pyrochlore, aeschynite, and bastnaesite, were transformed into perovskite and geikielite. The distribution rates of Nb, REEs, and Ti in the perovskite phase are 94.52%, 63.10%, and 48.82%, respectively. In addition, during the solidification and crystallization process of silicon-iron bath smelting reduction, a large amount of perovskite precipitates from niobium-containing slag, and rare earth elements are also dissolved in the perovskite [99]. By adjusting w(CaO)/w(SiO₂) ratio in the range of 1.2–1.9 in the slag and maintaining the slag at 1100 °C for 4 h, the area fraction of perovskite in the final slag exceeds 37%. Mineral phase reconstruction transforms complex and diverse niobium mineral phases into REEs-bearing EnANMs.

Except applying phase diagrams in the reconstruction of Nb-bearing resources, simulating geochemical environments of ore formation was also applied. In the geochemistry behaviors of niobium-bearing minerals as shown in Figure 6b, the pyrochlore formed in the magma with high Al₂O₃/CaO+Na₂O+K₂O (A/CNK) ratio [75]. Perovskite exists in the undersaturated SiO₂ rocks or magmas [100]. Hencn, pyrochlore and perovskite form in different content of acidic oxides and alkaline oxides. In the crystallization behaviors study of Nb-bearing phases in slag, Wensheng Han et al. conducted mineral phase transformation research on pre-selected low-grade niobium rougher concentrate through high-temperature smelting-cooling crystallization process [76]. Under an oxygen atmosphere, the transformed iron is present in the form of hematite and pseudobrookite; the transformed niobium is present in the form of loparite (perovskite structure) or calciobetafite (pyrochlore structure). As shown in Figure 6b, the niobium-bearing phase is transformed from calciobetafite to loparite by adding alkaline oxides (such as CaO, K₂O) to increase the slag basicity. For example, when the CaO/SiO₂ ratio increases from 0.5 to 1.4, the niobium-bearing phase in the slag changes from calciobetafite to loparite. By adding acidic oxides (such as Al₂O₃, P₂O₅) to decrease the slag basicity, the transformation of the niobium-bearing phase from loparite to calciobetafite is observed. For example, after adding 10% Al₂O₃ or P₂O₅, the diffraction peak intensity of loparite decreases, while the diffraction peak of calciobetafite appears despite the CaO/SiO₂ ratio is 1.4. The effect of slag basicity on niobium-bearing phase is consistent with the A/CNK value change in natural pyrochlore in the mineralization environment. Basicity regulation in slag can realize the directional transformation of various niobium minerals to calciobetafite (pyrochlore structure) and loparite (perovskite structure) [76]. The niobium content in different niobium-bearing phases varies with the change in basicity. The niobium content in the calciobetafite phase formed at lower basicity is relatively high. The mass fraction of Nb₂O₅ in calciobetafite reach 29.38%, and the total mass fraction of TiO₂, Nb₂O₅, and rare earth oxides (REO) in calciobetafite phase is 71.94%. The niobium content in the loparite phase formed at higher basicity is relatively low, and the mass fraction of Nb₂O₅ in loparite is 21.02%. Obviously, designing calciobetafite as EnAMs is more conducive to Nb enrichment. By studying the growth behavior of calciobetafite at different temperatures, a cooling process for EnAMs was designed [101]. When the melt is kept at 1245 °C for 3 h, calciobetafite does not crystallize, and hematite and pseudobrookite crystallize. When the temperature was maintained at 1200~1230 °C is beneficial to the growth of calciobetafite, and the temperature at 1140~1170 °C is beneficial to the nucleation of calciobetafite. During the cooling process of the melt, the maximum equivalent diameter of calciobetafite reaches 98.01 μm, which fulfills the granularity requirement for subsequent beneficiation.

4.2. Secondary Lithium Resource Recycling

When pyrometallurgical techniques are used to recover metals from secondary sources, such as lithium, nickel, and cobalt from spent batteries, these metals are reduced thermodynamically and hence separated from the slag by density [102,103]. However, lithium can not be separated and recovered because lithium has high oxygen affinity, consequently remaining in the slag [104]. Thomas Schirmer et al. conducted research on the design and control of EnAMs in lithium-containing slag systems, employing phase equilibrium design and calculations to study lithium-bearing EnAMs as enrichment carriers [45,105]. The primary lithium-bearing EnAMs includes lithium aluminate (LiAlO₂), spinel, lithium aluminum silicates, and calcium aluminum silicates. Among them, LiAlO₂ preferentially crystallizes at high temperatures, enabling efficient lithium enrichment; the crystal morphology (isometric crystals) of LiAlO₂ facilitates subsequent flotation separation [106,107]. And LiAlO₂ has the highest lithium content, approximately 10.5 wt%, which makes it a suitable lithium artificial mineral. In the study of the influence of components on LiAlO₂ crystallization, the increase in Mg/Al ratio promotes the preferential crystallization of high-melting-point spinel, leading to a reduction in LiAlO₂ crystallization, then affecting the distribution of Li in the minerals [45]. If the melt contains phosphorus, it may lead to the formation of Li₃PO₄ and inhibit the crystallization of lithium aluminate (LiAlO₂) [105]. Therefore, low-phosphorus and low-MgO content systems are the preferable designs to achieve efficient lithium enrich in LiAlO₂.

During the recovery lithium from Li(Co_1−x−yNi_xMn_y)O₂ electrode materials, Alena Wittkowski et al. designed a Li₂O-CaO-SiO₂-Al₂O₃-MgO-MnO_x system to simulate slag and investigate the influence of manganese on the formation of lithium aluminate (LiAlO₂) [43]. Manganese in the slag is enriched in the form of spinel solid solutions (Li_2xMn²⁺_1−x)_1+x(Al_2−z, Mn³⁺_z)O₄. The enrichment of partial lithium and aluminum in spinel solid solution results in the decrease of LiAlO₂ crystallization. D.A. Abrue et al. designed a multi-element phase equilibrium study to reduce the formation of LiMn₂O₄ [108]. By employing the CALPHAD method and combining phase diagram data from binary/ternary systems (such as CaO-SiO₂, Li₂O-SiO₂, Li₂O-MnO_x, MnO_x-SiO₂), a thermodynamic database for multi-component systems can be established to predict phase equilibrium and crystallization paths. By controlling the oxygen partial pressure and cooling path of the slag, the precipitation of Li₂SiO₃ (high lithium content) and LiMnO₂ can be directionally controlled. Effective suppression of Lithium Manganate (LiMn₂O₄) promotes lithium enrichment in EnAMs.

In the overall process of EnAMs design, dissociation, and separation, Hao Qiu et al. used thermodynamic calculations to design and control lithium-bearing EnAMs in the Li₂O-CaO-Al₂O₃-SiO₂-MnO system, promoting lithium enrichment in the γ-LiAlO₂ phase and inhibiting its dispersion in other mineral phases as shown in Figure 7 [46]. The slag is cooled and crystallized, followed by beneficiation and leaching to recover lithium. Through the design and control of the slag, the mass fraction of γ-LiAlO₂ phase was 27.4%. And other lithium-bearing phase, the mass fractions of β-eucryptite (β-LiAlSiO₄) and lithium manganese silicate (Li₂MnSiO₄) in slag are only 2.2% and 3.8%, respectively. The lithium mainly enriches in γ-LiAlO₂, and the remainder enriches in β-eucryptite (β-LiAlSiO₄) and lithium manganese silicate (Li₂MnSiO₄). Some lithium may be distributed in the amorphous glass phase. After crushing and dissociating the cooled slag, flotation was performed [108]. The flotation process uses sodium oleate (NaOl) as a collector and sodium hexametaphosphate (SHMP) as a depressant for separation. A lithium concentrate with a grade of 10.94% Li₂O and a recovery rate of 30.45% were obtained. In addition, Sophie Acker et al. designed a novel imidazole-2-thione reagent for lithium aluminate (LiAlO₂) [109]. The recovery rates of LiAlO₂ and spodumene are 88.9% and 71.8%, respectively. While the recovery rate of gehlenite gangue was less than 15%, which demonstrates excellent selectivity for EnAMs. Future research and application of new reagents might better enable the recovery of Li-bearing EnAMs.

Except using the flotation method for mineral separation, Javadi and co-workers also employed electrostatic separation to separate lithium aluminate from mixed powders [31]. The enrichment of lithium aluminate can be significantly improved when the proportion of particles with high free surface area values increases.

4.3. Recovery and Crystallization of Rare Earth from Slag

The recovery of Ce and La from rare earth slags is primarily achieved through mainstream recovery processes including hydrometallurgy, electrochemistry, bio-hydrometallurgy, and pyrometallurgy. Among them, promoting innovation in high-temperature recovery technologies can foster the sustainable recovery and utilization of rare earth resources [110]. The high-temperature melting-cooling crystallization method can be used to construct engineered artificial rare earth elements-bearing phases in the slag.

To study the crystallization behaviors of REEs in slag, the phase equilibrium relationships of multi-component systems such as CaO-SiO₂-La₂O₃ [111], CaO-SiO₂-Al₂O₃-Ce₂O₃ [112,113], CaO-SiO₂-CaF₂-Ce₂O₃ [114], CaO-SiO₂-CeO₂-10.0 wt% Al₂O₃-5.0 wt% MgO [115], and CaO–SiO₂–CaF₂(30 wt%)-P₂O₅(10 wt%)-Ce₂O₃ [49] at different temperatures have been widely studied. Researchers have revealed that the main rare earth element-bearing phases include Ca_xLa_4.67−x(SiO₄)₃O_1−0.5x (containing 70 wt% La₂O₃), La₂Ca₃(SiO₃)₆ (containing 40 wt% La₂O₃), Ce_4.667−xCa_x(SiO₄)₃O_1−0.5x, Ce_9.33−xCa_x(SiO₄)₄O_5−0.5xF₂, Ca₅(PO₄)₃F, and silicophosphate apatite (Ca_5−xCe_x[(Si,P)O₄]₃F). Because the Ca_xLa_4.67−x(SiO₄)₃O_1−0.5x, Ce_4.667−xCa_x(SiO₄)₃O_1−0.5x, britholite-(Ce), silicophosphate apatit, and Ca₅(PO₄)₃F phases have a high content of REEs, these phases are potential candidate EnAMs. According to the phase equilibrium results, it is promising to design an environment where only rare earth element-bearing phases are preferentially precipitated in the liquid phase [114,116]. As shown in Figure 8a, only the Ce_9.33−xCa_x(SiO₄)₄O_5−0.5xF₂ phase forms in zoon 9. Then rare earth element-bearing phases can be extracted by solid–liquid separation technology.

Further research on the crystallization behavior of rare earth element-bearing phases has revealed that the crystallization model is mainly diffusion-controlled [116,117,118]. The crystallization model suggests the cooling rate of slag should be in the range of 20–30 °C/min [116]. Regarding the growth behavior of rare earth element-bearing EnAMs, an increase in P₂O₅ content shortens the incubation time of EnAMs because P₂O₅ raises the initial crystallization temperature [118]. The P₂O₅ also can increase the nucleation rate because it reduces the activation energy for crystallization [118]. Hence, P₂O₅ can promote the growth of rare earth element-bearing EnAMs in certain conditions [119]. In addition, P₂O₅ content also affects the transformation of apatite (Ca_2.02Ce_7.98Si₆O₂₆) and britholite (Ca_5−xCe_x[(SiO₄)_x(PO₄)_3−x]F) [117]. Conversely, increasing the CaF₂ content in slag has the opposite effect on incubation time and nucleation rate compared to phosphorus [118].

In a separation method for REEs, Xi Lan et al. used centrifugal force generated by rotation to separate and recover REEs from the slag system because of the density difference between rare earth element-bearing EnAMs and slag, as shown in Figure 8b [120]. In the centrifugal equipment, a counterweight and a heating furnace are symmetrically fixed on the centrifugal rotor. The heating furnace continuously heats during the rotation process to stabilize the temperature. After the separation process, the rare earth element-bearing EnAMs move to the bottom of the slag, as shown in Figure 8c. In Figure 8c, Cerium fluorosilicate (Ce₂O₃: 67.23 wt%) was selectively separated from the CaO-SiO₂-CaF₂-Ce₂O₃ system, achieving a rare earth recovery rate of 98.23% [120]. Similarly, britholite with varying Ce₂O₃ contents (17.87–46.93 wt%) was separated from the CaO-SiO₂-CaF₂-P₂O₅-Ce₂O₃ system, with rare earth recovery rates reaching 95.58%–98.65% [121]. In the practical application of Bayan Obo rare earth concentrate, the major rare earth elements (Ce, La, Pr, and Nd) were enriched in EnAMs as the temperature decreased. Cerium was almost entirely transformed into the Ce-oxide phase, lanthanum was present in the La-ferrate phase, and praseodymium and neodymium were mainly enriched in apatite. After supergravity separation, the recovery rates of Ce in the Ce-oxide phase, La in the La-ferrate phase, and Pr and Nd in the apatite phase reached 98.80%, 96.81%, 92.12%, respectively [122].

Figure 8. (a) Isothermal phase equilibria diagram of the CaO–SiO₂–CaF₂–Ce₂O₃ system at 1373 K with a fixed CaF₂ component (reproduced with permission from [114], published by Elsevier Ltd. and Techna Group S.r.l., 2020) All rights reserved. (b) Operation schematic illustrations of the centrifugal device [120], and (c) vertical section of the sample via supergravity and the corresponding results [120].

Figure 8. (a) Isothermal phase equilibria diagram of the CaO–SiO₂–CaF₂–Ce₂O₃ system at 1373 K with a fixed CaF₂ component (reproduced with permission from [114], published by Elsevier Ltd. and Techna Group S.r.l., 2020) All rights reserved. (b) Operation schematic illustrations of the centrifugal device [120], and (c) vertical section of the sample via supergravity and the corresponding results [120].

Current studies in phase diagram construction and crystallization kinetics analysis have clarified the formation mechanism and separation conditions of rare earth element-bearing EnAMs. Supergravity technology provides a new pathway for efficient recovery. However, the thermodynamic database for REE systems still needs improvement, and further research is needed on complex component interactions and non-equilibrium crystallization behavior. In the future, the combination of artificial intelligence and high-throughput experimental techniques can accelerate phase diagram prediction and promote the sustainable utilization of rare earth resources.

EnAM design is applied not only in REE slags but also in blast furnace titanium slag to design perovskite as an EnAM for titanium recovery [123,124,125,126]. In high-iron-content nickel slag/blast furnace slag, EnAMs such as (Ni,Fe)Fe₂O₄ [127], Magnetite [128,129,130], and (Mn,Mg)yFe_3−yO₄ [67] are designed. Iron and associated elements, such as the Co, Ni, and Cu in EnAMs, can be recovered by magnetic separation. From stainless steel chromium slag, spinel EnAMs were designed for chromium immobilization [53,131,132]. In vanadium converter slag, V-bearing spinel EnAMs were designed for vanadium enrichment [133]. Therefore, most of the high-value elements and harmful elements in slag can be recovered and transformed by constructing EnAMs.

5. Challenges and Future Perspectives

5.1. Optimization of EnAMs Design

On the one hand, based on thermodynamic phase simulations and development of databases for the mineralogical phases, the conversion rate of EnAMs should be enhanced. After slag cooling and crystallization, target elements are not only enriched in EnAMs but also in other gangue phases and amorphous glass phases, leading to difficulties in sorting various minerals and low recovery rates.

On the other hand, during EnAMs design, the magnetic, density, and dielectric properties of gangue should be considered. Gangue minerals that exhibit physical properties and differences relative to EnAMs facilitate subsequent mineral processing.

5.2. Improving Economic Viability

To achieve the separability of EnAMs after liberation, a relatively long cooling and crystallization time is required to ensure the particle size of EnAMs, which results in the low economic viability of the process. In the future, research should focus on the crystallization kinetics of the slag system, and additive engineering to promote the rapid growth of EnAMs. Designing new high-temperature conversion equipment powered by renewable electricity achieves the application of green electricity in EnAMs and improves economic viability.

5.3. Physical Separation of Fine-Grained (<20 μm) EnAMs

To reduce the cost of metal recovery processes, physical separation equipment for fine-grained minerals should be vigorously developed. If efficient separation of fine-grained minerals can be achieved based on magnetic, density, or conductivity differences, it will greatly reduce the crystallization time during slag cooling. Thereby, effective physical separation improves economic viability while reducing the environmental pollution caused by flotation or leaching processes.

6. Conclusions

Constructing engineered artificial minerals (EnAMs) is an effective strategy for sustainable resource recovery from complex mineral resources, metallurgical waste slags, and secondary urban mines. By leveraging high-temperature phase reconstruction processes, EnAMs enable the selective enrichment of target elements (e.g., Nb, REEs, Li, Cr, and V) into designed crystalline phases during crystallization melting. In comparison, phase equilibrium studies are more effective and scientific than utilizing geochemical behaviors when designing EnAMs. However, in certain specific cases, the metallogenic geochemical behavior of elements can also provide rough guidance for EnAM design. The limitations of geochemical behaviors in phase reconstruction highlight the need for further exploration of the relationship study between geochemistry and pyrometallurgy. The application of EnAMs in metal recovery from complex resources has achieved some success. During the transformation process, target elements can be enriched selectively in EnAMs. However, further research is still needed on the design, dissociation, and beneficiation of EnAMs. In the future, electricity generated from new energy sources can be applied to the energy required for preheating and cooling in the process of mineral phase reconstruction. This will provide a clean approach for the mineral phase reconstruction of complex resources. Overall, designing EnAMs to solve the development of complex resources has broad prospects.