Advances in Integrated Extraction of Valuable Components from Ti-Bearing Slag

Abstract

Ti-bearing blast furnace slag (TBS), a byproduct of vanadium–titanium magnetite smelting, serves as an important secondary resource for titanium recovery. However, the complex mineralogical composition and finely dispersed nature of titanium in TBS present significant challenges for efficient extraction. This review systematically examines four major titanium extraction routes: hydrometallurgical leaching, pyrometallurgical smelting, molten salt electrolysis, and selective precipitation, focusing on their limitations and recent improvements. For instance, conventional acid leaching suffers from acid mist release, a colloidal formation that hinders titanium recovery, and waste acid pollution. The adoption of concentrated sulfuric acid roasting activation effectively suppresses acid mist emission and prevents colloidal generation. Pyrometallurgical approaches are hampered by high energy consumption and substantial carbon emissions, which can be alleviated through the use of gaseous reductants to enhance reaction efficiency and reduce environmental impact. Molten electrolysis faces issues such as polarization and undesirable dendritic deposition; these are mitigated by employing liquid metal cathodes integrated with vacuum distillation to achieve high-purity titanium products. Selective precipitation struggles with strict crystallization conditions and low separation efficiency, though advanced techniques like supergravity separation show improved extraction performance. We propose an integrated technical strategy termed “Online conditioning driven by waste heat-mineral phase reconstruction-directional crystallization-optimized liberation.” This approach utilizes the inherent waste heat of slag combined with electromagnetic stirring to enhance homogeneity and promote efficient titanium recovery, offering a sustainable and scalable solution for industrial TBS treatment.

1. Introduction

Titanium is a strategic metal critical for advanced aerospace, marine, and biomedical applications due to its exceptional specific strength, corrosion resistance, and biocompatibility [1,2,3,4]. As shown in Figure 1 [5,6], it is extensively applied in cutting-edge fields. Figure 2 [5,6] further depicts the escalating annual demand for titanium resources in China.

China possesses dominant global titanium reserves, accounting for approximately 40% of worldwide deposits [6,7]. The majority of these resources are concentrated in the Panzhihua vanadium–titanium magnetite deposit, which represents over 90% of China’s total titanium resources and constitutes a typical multi-element symbiotic mineral resource [8,9,10,11]. The metallurgical processing of these ores for iron production generates substantial quantities of Ti-bearing blast furnace slag (TBS) with TiO₂ content ranging from 20% to 29% [12]. In TBS, titanium is finely dispersed within refractory mineral phases, including perovskite, anosovite, and diopside [13,14], making conventional recovery methods economically challenging. This results in annual accumulation of TBS exceeding one million tons in the Panxi region [15], representing both a significant loss of strategic resources and considerable environmental risks due to potential heavy metal leaching [16].

Consequently, developing efficient utilization strategies for TBS has become imperative, which are principally categorized into two approaches: non-titanium extraction and titanium extraction. Current non-extractive utilization approaches primarily involve applications in construction materials such as cement and concrete aggregates [17,18], with additional uses in photocatalyst synthesis, ceramics, and glass manufacturing [19,20,21,22]. These applications offer limited value as they fail to recover the valuable titanium component [23], highlighting the urgent need for developing efficient titanium extraction technologies.

The approach of titanium extraction utilization primarily encompasses four methodologies: hydrometallurgical processing, pyrometallurgical smelting, molten electrolysis, and selective phase separation and extraction [24,25]. Compared to the preparation of non-titanium extraction functional materials, titanium extraction technology enables effective recovery of titanium components to some degree, demonstrating superior economic value. However, these techniques universally share persistent challenges in industrial applications, including secondary pollution from residual tailings, excessive energy consumption, and suboptimal Ti recovery rates [26]. Additionally, these four methods each have their own limitations. For instance, the hydrometallurgical process [23,24,25,26,27] imposes stringent requirements on raw material quality (requiring a high-grade concentrate of TiO₂ ≥ 45%) [28]. Furthermore, the acid-leaching method suffers from severe equipment corrosion issues, coupled with challenges in waste acid recovery and treatment, as well as significant environmental pollution concerns. Pyrometallurgical smelting is plagued by prohibitive energy intensity during the high-temperature carbonization process, which fails to align with the national strategic goals of “carbon neutrality and carbon peak” currently issued [29,30]. Molten electrolysis confronts challenges of electrode corrosion and stringent requirements for specialized electrode materials [31,32]. The selective precipitation method suffers from excessively high heat-treatment temperatures, prolonged processing durations, difficulties in the homogenization of modifiers in molten slag, semi-coagulated core, and poor dissociation degree of monomers [33,34].

Currently, most of the review studies on titanium extraction processes focus on the introduction and evaluation of traditional processes. Nevertheless, there remains a notable gap in systematic reviews of emerging technologies addressing critical industrial-upgrading aspects such as low-carbon metallurgical system reconstruction, directional dissociation regulation of mineral phases, and multi-element synergistic co-enrichment.

In light of the prevalent challenges of high energy consumption, elevated costs, process complexity, and industrialization barriers in titanium extraction, this paper conducts an in-depth investigation of recent advancements by related researchers regarding industrial upgrading aspects like low-carbon energy consumption reduction, optimized dissociation, and enrichment of valuable elements. By probing the limitations of the aforementioned titanium extraction technologies and performing a comprehensive analysis, the present work explores titanium extraction technology solutions characterized by high efficiency, economic viability, low energy consumption, and environmental sustainability with substantial industrialization potential. Eventually, this study proposes an integrated strategy, designated as “ Online conditioning driven by waste heat–mineral phase reconstruction–directional crystallization–optimized liberation”, which aims to offer innovative solutions for breaking through the bottlenecks, establishing theoretical and technical foundations for industrial-scale implementation of titanium extraction technology.

2. Classification, Properties, and Components of Ti-Bearing Slags (TBS)

2.1. Classification and Properties of TBS

The formation of TBS occurs as a byproduct during ironmaking processes utilizing titanium-containing iron ores (including vanadium–titanium magnetite and ilmenite), representing a crucial secondary resource [35]. During ironmaking, Fe and V are reduced by carbon and migrate into the iron phase to form vanadium-rich hot metal, while Ti remains in the slag phase, ultimately forming Ti-bearing slag (TBS). In the Panxi region, approximately 45–47 wt.% of titanium from raw vanadium–titanium magnetite ore reports to titanium concentrate through mineral processing, with the remaining portion entering iron tailings [36,37]. Subsequent treatment of iron tailings yields electric furnace titanium slag, while iron concentrate processed through blast-furnace smelting or rotary hearth furnace direct reduction and electric furnace melting separation generates Ti-bearing blast furnace slag and molten separation titanium slag, respectively [38,39,40]. Figure 3 illustrates the formation pathways of different titanium slag categories.

As shown in Figure 3, Ti-bearing slags in the Panxi region can be classified into three categories based on processing routes: blast furnace slag (BFS), electric molten slag (EMS), and electric furnace slag (EFS). The titanium content in these slags varies significantly depending on their formation processes, with their respective TiO₂ contents and characteristics summarized in

Table 1. Classification and Characteristics of Titanium-Bearing Slag.

Classification	TiO2 Content wt.%	Properties
BFS	20–29%	generated during blast furnace ironmaking; large production; low grade of TiO2 content; complex mineral composition; intricate phase structures; great difficulty in comprehensive utilization.
EFS	≥75%	generated during electric furnace slag; high grade of TiO2 content; critical feedstock for both sulfate-process titanium dioxide production and chloride-process sponge titanium manufacturing.
EMS	45–68%	obtained via direct reduction-iron extraction followed by electric furnace melting separation of vanadium–titanium magnetite; high titanium grade; low impurity content; superior potential for value-added utilization.

Note: All data in this table are adapted from reference [40].

.

2.2. The Chemical Components and Mineral Phases of TBS

The composition of TBS varies significantly depending on the vanadium–titanium magnetite smelting process employed. Based on the above analysis, the chemical compositions of the three types of TBS differ considerably. BFS is represented by the Ti-bearing blast furnace slag produced by Pangang Group. Its main chemical components include TiO₂, Al₂O₃, MgO, CaO, SiO₂, V₂O₅, and MnO₂, along with trace elements such as Fe, V, Mn, P, S, etc. [40]. EMS is typified by the molten-separated titanium slag obtained from the smelting reduction ironmaking process of Panzhihua vanadium–titanium magnetite concentrate. It has a relatively high TiO₂ content, reaching approximately 45%, accompanied by other oxides and carbides [41]. EFS is exemplified by the electric furnace titanium slag produced via electric furnace smelting of Panzhihua titanium concentrate, with its TiO₂ content as high as 76% [42].

The chemical compositions of these three typical Ti-bearing slags (TBS) are listed in

Table 2. Main chemical components of the three types of EMS, BFS, and EFS (wt. %).

Type	TiO2	Al2O3	MgO	CaO	SiO2	TFe	MFe	FeO	V2O5	Ti2O3	TiC	MnO2	MnO	P2O5	S	Others
EMS	45.32	12.26	7.58	2.91	13.10	11.09	0.27	14	2.00	0.167	0.3	-	1.44	-	-	-
BFS	23.35	11.09	7.06	28.64	25.44	2.82	-	-	0.20	-	-	0.75	-	0.022	0.12	0.50
EFS	76.03	2.40	7.39	1.72	3.20	8.46	-	6.15	0.215	5.07	-	-	-	-	0.25	0.335

Note: Data in this table for BFS, EMS, and EFS, adapted from references [40,41,42], respectively.

, which clearly illustrates the differences in chemical compositions of titanium slags generated under three distinct smelting processes.

These chemical components constitute intricate, fine-grained, and dispersed Ti-bearing mineral phases with distinct structures, such as perovskite ((Ca, Mg)O·TiO₂), anosovite (m(AO·TiO₂)·n(B₂O₃·TiO₂)), rutile(TiO₂), diopside(CaMgSi₂O₆), pseudobrookite (Fe₂TiO₅), spinel ((Ca, Mg, Fe, Mn) O·(Fe, Al, V)₂O₃), and silicate glassy phases [43,44,45]. The crystallographic structural parameters of these key Ti-bearing phases, which are decisive for the kinetics and efficiency of titanium extraction, are summarized in

Table 3. Crystallographic structural parameters of major Ti-bearing phases.

Phase	Molecular Formula	Density (p/cm)−3	Crystal Structure	Geometric Morphology	Key Unit Cell Parameters
Ilmenite	FeTiO3	4.20–5.20	Trigonal Crystal System	Tabular, granular	a = 5.09 Å, c = 14.07 Å
Perovskite	CaTiO3	4.10	Cubic Crystal System	Spindle shaped	a = 3.85 Å
Anosovite	Ti3O5	4.68–4.79	Orthorhombic Crystal System	Bundled	a = 9.12 Å b = 3.78 Å c = 5.03 Å
Rutile	TiO2	4.20–4.30	Tetragonal Crystal System	Tetragonal prismatic, acicular	a = 4.59 Å c = 2.96 Å
Pseudobrookite	Fe2TiO5	4.49	Orthorhombic Crystal System	Acicular	a = 9.74 Å b = 3.74 Å c = 5.09 Å
Sphene	CaTiSiO5	3.50	Monoclinic Crystal System	Wedge shaped, tabular	a = 6.55 Å b = 8.70 Å c = 7.43 Å β = 119.3°

Note: All data in this table are generated via the online simulation platform Materials Explorer.

.

Figure 4 reveals the typical crystal ball-and-stick models of Ti-bearing mineral phases. Based on the crystallographic parameters of typical Ti-bearing minerals listed in Table 3 and molecular ball-and-stick models as shown in Figure 4, the molecular structural characteristics significantly influence the design and efficiency of titanium extraction processes. The trigonal layered structure of ilmenite (FeTiO₃, a = 5.09 Å, c = 14.07 Å), as shown in Figure 4a, characterized by ionic bonding differences between Fe²⁺ and Ti⁴⁺, enables high-efficiency magnetic separation, yet requires reductive roasting to convert Fe³⁺ to Fe²⁺ for enhanced magnetism [46,47]. In contrast, rutile (TiO₂), as shown in Figure 4d, with its tetragonal chain-like structure (Ti–O bond length = 1.94 Å), dominated by covalent Ti–O octahedral networks, demands high-temperature (>1000 °C) or strong alkaline fusion to overcome chemical inertness [48]. The cubic symmetry of perovskite (CaTiO₃, Pm-3m), as shown in Figure 4b, exposes Ti⁴⁺ in a shared-vertex TiO₆ framework, facilitating rapid acid leaching, though requiring suppression of Ca²⁺ dissolution byproducts [49,50]. The orthorhombic mixed-valence (Ti³⁺/Ti⁴⁺) structure of anosovite (Ti₃O₅), as shown in Figure 4c, and the acicular morphology of pseudobrookite (Fe₂TiO₅), as shown in Figure 4e, optimize physical separation and acid-leaching kinetics via high density and surface area, respectively [51,52]. Additionally, the monoclinic silicate network of sphene (CaTiSiO₅, β = 119.3°), as shown in Figure 4f, necessitates alkaline fusion to release encapsulated Ti⁴⁺. Overall, the crystal symmetry, unit cell parameters, bonding types, and geometric morphology of titanium minerals collectively govern ion diffusion pathways, chemical activity, and impurity interference. These insights advocate process-specific strategies: high-symmetry minerals suit ambient-pressure acid leaching, covalent-dominated structures require high-temperature chlorination, and complex minerals demand integrated physical preconcentration and multi-stage chemical decomposition for efficient and selective titanium recovery [53,54].

The predominant mineral phases of BFS in Panxi include perovskite, Panxi titaniferous diopside, titanium-enriched diopside, Mg–Al spinel, and titanium carbonitride [55]. EMS primarily contains anosovite solid solution, Ti₃O₅ solid solution, TiC solid solution, magnesium aluminate spinel, cryptocrystalline silicate glass, and metallic iron droplets. EFS is mainly composed of anosovite solid solution, Ti₃O₅ solid solution, silicate glassy matrix, and free TiO₂ [56,57].

The physical and chemical properties of Ti-bearing phases are governed by their crystal/molecular structures. For instance, rutile-type TiO₂ exhibits a dense crystalline structure with high chemical stability, while anatase-type TiO₂ possesses a relatively loose structure and higher reactivity. The titanium extraction kinetics from these phases are influenced by their crystal structures through critical factors including reaction interfaces, chemical bond nature/rupture mechanisms, diffusion rates, and crystal defects. By analyzing these structures, the inherent characteristics can be precisely identified, providing fundamental insights for optimizing subsequent titanium extraction processes [58,59].

3. Analysis of the Current Technologies for Ti Extraction from TBS

The titanium extraction technology from TBS involves the migration, enrichment, and separation of titanium constituents via physical, chemical, or metallurgical methods for its efficient recovery and value-added utilization of titanium resources [60]. This approach offers superior economic viability compared to non-extractive methods. Four principal industrial strategies are prevalent: hydrometallurgical leaching, pyrometallurgical smelting, molten electrolysis, and selective precipitation [61]. This section will delve into the technical limitations of these methods, such as high energy consumption, equipment corrosion, and environmental pollution [62], and explore potential improvements based on recent research.

3.1. Hydrometallurgical Leaching

Hydrometallurgical leaching represents a pivotal technology based on chemical dissolution–crystallization principles to achieve high-value conversion of titanium components in the field of titanium recovery from TBS. By means of acid leaching or alkali dissolution, this process dissolves titanium from Ti-bearing raw materials (e.g., ilmenite, rutile, etc.) into the aqueous phase system. The titanium-related products are then prepared by subsequent processes such as purification, enrichment, and hydrolysis stages [63,64]. Acid-based titanium extraction technologies primarily include the sulfuric acid process, ammonium sulfate process, and hydrochloric acid process. Among them, the sulfuric acid process features a simple process, wide adaptability to raw materials, and relatively established industrial application maturity [65].

Although the conventional sulfuric acid process for extracting valuable components from high-titanium slag achieves relatively good results, it typically employs sulfuric acid with a concentration below 90% [66,67]. To achieve optimal extraction efficiency, the acid dosage generally far exceeds the theoretical stoichiometric requirement of the slag components [68]. Furthermore, the reaction is conducted at temperatures near the boiling point of the acid at the employed concentration. Dilute sulfuric acid has a high moisture content and is prone to undergo redox reactions with reducing components in slag at high temperatures. This leads to the generation of significant amounts of acid mist and the discharge of large volumes of waste acid, resulting in environmental pollution and equipment corrosion [69]. Additionally, during the leaching process, CaO in slag directly reacts with dilute sulfur acid to form CaSO₄ precipitate, which tends to precipitate in the form of ultrafine particles and agglomerate into gypsum colloids (CaSO₄·2H₂O), resulting in difficult filtration and loss of target elements such as Ti due to adsorption [70].

These challenges stem from the inherently low reactivity of refractory oxides in TBS towards dilute acid. To address these drawbacks, the mechanism of activation, roasting with concentrated sulfuric acid coupled with dilute sulfuric acid leaching, has emerged as the primary promising improvement direction for current sulfuric acid-based processes [71]. This improvement is scientifically grounded in the capability of concentrated sulfuric acid to convert insoluble oxides (e.g., Ti₂O₃, TiO₂) into soluble sulfates (e.g., TiOSO₄) through sulfation reactions, as thermodynamically demonstrated in Figure 5. Concentrated sulfuric acid exerts an enhancing effect on the reactivity of Ti, Mg, and Al in the slag. Via reactions with concentrated sulfuric acid (as shown in Equations (1)–(6)), water-insoluble oxides in the slag, such as Ti₂O₃, TiO₂, MgO, Al₂O₃, Fe₂O₃, and FeO, are preferentially converted into water-soluble sulfate (i.e., TiOSO₄, MgSO₄, Al₂(SO₄)₃, and Fe₂(SO₄)₃), which can be efficiently dissolved in the subsequent dilute sulfuric acid-leaching step. Meanwhile, CaO also reacts with concentrated sulfuric acid to produce sparingly water-soluble sulfate CaSO₄, as shown in Equation (7), which exists in its original coarse particle solid form in the roasted sand residue [72,73,74,75,76]:

$${\mathrm{TiO}}_2 + {\mathrm{H}}_2\mathrm{S}{\mathrm{O}}_4 = {\mathrm{TiOSO}}_4 + {\mathrm{H}}_2\mathrm{O}$$

(1)

$${\mathrm{Ti}}_2{\mathrm{O}}_3+3{\mathrm{H}}_2{\mathrm{SO}}_4=2{\mathrm{TiOSO}}_4+3{\mathrm{H}}_2\mathrm{O}+\mathrm{S}{\mathrm{O}}_2$$

(2)

$$\mathrm{MgO}+{\mathrm{H}}_2{\mathrm{SO}}_4={\mathrm{MgSO}}_4+{\mathrm{H}}_2\mathrm{O}$$

(3)

$${\mathrm{Al}}_2{\mathrm{O}}_3+3{\mathrm{H}}_2\mathrm{S}{\mathrm{O}}_4={\mathrm{Al}}_2{(\mathrm{S}{\mathrm{O}}_4)}_3+3{\mathrm{H}}_2\mathrm{O}$$

(4)

$${\mathrm{Fe}}_2{\mathrm{O}}_3+3{\mathrm{H}}_2\mathrm{S}{\mathrm{O}}_4={\mathrm{Fe}}_2{(\mathrm{S}{\mathrm{O}}_4)}_3+3{\mathrm{H}}_2\mathrm{O}$$

(5)

$$2\mathrm{FeO}+4{\mathrm{H}}_2\mathrm{S}{\mathrm{O}}_4={\mathrm{Fe}}_2{(\mathrm{S}{\mathrm{O}}_4)}_3+4{\mathrm{H}}_2\mathrm{O}+\mathrm{S}{\mathrm{O}}_2$$

(6)

$$\mathrm{CaO}+{\mathrm{H}}_2\mathrm{S}{\mathrm{O}}_4=\mathrm{C}\mathrm{a}\mathrm{S}{\mathrm{O}}_4+{\mathrm{H}}_2\mathrm{O}$$

(7)

$${\mathrm{SiO}}_2+{\mathrm{H}}_2\mathrm{S}{\mathrm{O}}_4\stackrel{Gibbs simulation assumption}{\Rightarrow }\mathrm{S}\mathrm{i}\left(\mathrm{S}{\mathrm{O}}_4\right)+{\mathrm{H}}_2\mathrm{O}$$

(8)

The Ellingham diagram for the reactions between components in high-titanium blast furnace slag and concentrated sulfuric acid is shown in Figure 5. The Gibbs free energy change (∆G) for each reaction, as a function of temperature, was computationally determined using the HSC Chemistry 9 software. The software calculates ∆G values by minimizing the total Gibbs energy of the system, utilizing its integrated and extensively validated thermodynamic database. The resulting data points for all reactions were then exported to Origin 2024 for graphical representation and to generate the comprehensive ∆G-T diagram shown in Figure 5. Based on the calculated Gibbs free energy values for each chemical reaction, the results are shown in Figure 5. It is evident that all values are negative at temperatures below 160 °C except SiO₂. This further theoretically demonstrates the feasibility of the reactions of Equations (1)–(7) within this temperature range. The ΔG-T plot of Figure 5 demonstrates that the Gibbs free energy for the reaction between SiO₂ and concentrated sulfuric acid remains persistently positive across 0–400 °C, confirming that no spontaneous reaction occurs as shown in Equation (8). Therefore, SiO₂ and CaSO₄ can be easily separated from the leachate by filtration in the subsequent dilute sulfuric acid-leaching step.

The calcine generated from concentrated sulfuric acid roasting is mixed with dilute sulfuric acid at a specific liquid-to-solid ratio, and leaching is performed under stirring conditions at 50–90 °C. During this process, soluble sulfates in the calcine (i.e., TiOSO₄, MgSO₄, Al₂(SO₄)₃, and Fe₂(SO₄)₃) are fully dissolved into the liquid phase, enabling the transfer of target elements from the solid phase to the liquid phase. In contrast, insoluble impurities, such as SiO₂ and CaSO₄, remain as solid residues. The separation of the leachate from the residues is achieved through filtration, providing a high-quality liquid feedstock for the subsequent purification and recovery of target elements [77].

The hydrolysis process ultimately decomposes titanium oxysulfate into precipitated metatitanic acid and sulfuric acid, as shown in Equation (9). The precipitated metatitanic acid obtained from hydrolysis undergoes dehydration through high-temperature calcination to form titanium dioxide, as illustrated in Equation (10) [78]:

$${\mathrm{TiOSO}}_4 + 2{\mathrm{H}}_2\mathrm{O}\rightleftharpoons {\mathrm{H}}_2{\mathrm{TiO}}_3\downarrow + {\mathrm{H}}_2\mathrm{S}{\mathrm{O}}_4$$

(9)

$${\mathrm{H}}_2\mathrm{T}\mathrm{i}{\mathrm{O}}_3=\mathrm{T}\mathrm{i}{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}$$

(10)

Figure 6 illustrates the process where high-titanium slag undergoes roasting with concentrated sulfuric acid and leaching with dilute sulfuric acid. After filtering out impurities (CaSO₄ and SiO₂), a filtrate containing valuable elements (TiOSO₄) is obtained, which is subsequently subjected to hydrolysis and calcination to produce titanium oxide.

Recent research has achieved some progress. For instance, He S.Q.’s team developed a four-stage integrated process route that involves activation by concentrated sulfuric acid roasting, extraction of Ti, Mg, and Al by dilute sulfuric acid leaching, preparation of TiO₂ by boiling hydrolysis, and preparation of hydrotalcite by coprecipitation hydrothermal treatment for Panzhihua water-quenched high-titanium blast furnace slag (TiO₂ = 22.3 wt.%) [79].

Through experimental research on the extraction of Ti, Al, and Mg components from high-titanium blast furnace slag using concentrated sulfuric acid, the research group determined the optimal process parameters: an acid-to-slag ratio of 1.4, a roasting temperature of 130 °C, and a roasting duration of 40 min. Under these conditions, the extraction rates reached 82.85% for Ti and over 90% for both Mg and Al. In the experimental study on enhancing the filterability of calcined slag slurry through dilute sulfuric acid leaching, it was found that the optimal process parameters are as follows: sulfuric acid concentration of 0.6 mol/L, leaching temperature of 60 °C, and leaching time of 40 min. Under these conditions, the formation of gypsum colloids (CaSO₄·2H₂O) is effectively prevented, with anhydrite instead precipitating as coarser and uniformly distributed particles. This morphology resists compact packing, thereby achieving complete liquid–solid separation within a short timeframe. Through optimization of the forced hydrolysis process under boiling conditions, metatitanic acid with uniform particle size distribution was obtained. Subsequent calcination yielded TiO₂, while magnesium–aluminum hydrotalcite was synthesized via a co-precipitation hydrothermal method. This integrated process route effectively achieves cascade recovery and resource utilization of valued elements (Ti, Mg, and Al) from high-titanium blast furnace slag [79].

Zhou L.S. [80] also adopts concentrated sulfuric acid (95.0–98.0 wt.%)-roasting activation to significantly enhance the leaching efficiency of Ti-bearing blast at 130 °C, achieving notable leaching rates of valuable metallic elements, including Ti, Mg, and Al. The reaction mechanism of concentrated sulfuric acid roasting was systematically elucidated through chemical reaction thermodynamic analysis. Experimental parameters were optimized to determine the appropriate liquid-to-solid ratio and identify the optimal reaction temperature. The leaching rates of Ti, Mg, Al, Fe, and Ca from Ti-bearing blast furnace slag after concentrated sulfuric acid-roasting activation reached higher than 85%, 95%, 95%, 45%, and 24%, respectively, all exceeding those of the same slag without roasting activation under identical conditions. Figure 7 obviously shows that the leaching rates of valuable elements achieved via the concentrated sulfuric acid-roasting process are significantly higher than those without this process.

The synergistic “calcination conversion-leaching dissolution” process not only overcomes the inherent low solubility of refractory oxides through calcination, but also achieves targeted transfer of valuable elements from solid to liquid phase via leaching. This dual mechanism ensures high-efficiency extraction of titanium and other valuable elements.

Another pressing challenge in acid-based titanium extraction processes is the treatment of spent acid, the disposal cost and environmental hazard of which directly restricts industrial-scale application. The spent acid generated from the sulfuric acid process typically contains less than 20% sulfuric acid and is contaminated with metal ions (such as iron, aluminum, and calcium) and solid suspensions. Direct discharge of such low-concentration spent acid would not only cause severe environmental pollution but also result in the loss of valuable sulfur resources. Therefore, effective treatment of spent acid is of great significance. The core tasks of the treatment are to remove impurities through pretreatment and conduct deep purification by adding dehydrating agents and adopting multi-effect evaporation, so as to achieve concentration, recovery, and reuse of spent acid. The main technological process of the treatment of spent acid is shown in Figure 8.

To address this issue, Pang H.Y. et al. proposed a coupled Chemical Dehydration-Multi-Effect Evaporation Waste Acid Regeneration Process for effective spent-acid management. This process comprises two key aspects: first, addition of FeSO₄·H₂O triggers metastable phase transformation (as shown in Equations (11) and (12)), achieving chemical dehydration via heptahydrate crystallization (ΔH = −58 kJ·mol⁻¹). This elevates the spent acid concentration from 20 wt.% to 45 wt.% with 62% water reduction. Second, three-stage countercurrent evaporation (vacuum: 0.08–0.09 MPa) further concentrates acid to 70 wt.%, while FeSO₄·7H₂O is regenerated to FeSO₄·H₂O via 120 °C drying (>95% recyclability), reducing operational costs by 41% versus conventional evaporation [81]:

$$\mathrm{FeS}{\mathrm{O}}_4·{\mathrm{H}}_2\mathrm{O}\leftrightharpoons {\mathrm{Fe}}^{2+}+ \mathrm{S}{\mathrm{O}}_4^{2-}+ {\mathrm{H}}_2\mathrm{O}$$

(11)

$${\mathrm{Fe}}^{2+}+\mathrm{S}{\mathrm{O}}_4^{2-}+7{\mathrm{H}}_2\mathrm{O}\leftrightharpoons {\mathrm{FeSO}}_4·7{\mathrm{H}}_2\mathrm{O}\downarrow$$

(12)

While this method alleviates scaling issues and minimizes environmental impacts, it overlooks interference from Al³⁺ (>1.2 g/L) and SiO₂ colloids (<0.5 μm) in waste acid during crystallization, potentially accelerating evaporator scaling to 1.2 mm/month during long-term operation [82].

Future research directions for hydrometallurgical titanium extraction will be addressed from the following perspectives: developing TiO₂–SO₂ synergistic capture technology for in situ sulfur resource immobilization; constructing impurity (Al³⁺, SiO₂) directional separation–passivation systems to extend equipment lifespan; and optimizing multi-process parameter coupling via machine learning as well as establishing a digital twin model for hydrometallurgical titanium extraction [83].

3.2. Pyrometallurgical Titanium Extraction Methods

Pyrometallurgical titanium extraction methods refer to a set of metallurgical processes that utilize high-temperature conditions to extract titanium from its ores (primarily ilmenite and rutile). These methods are characterized by high-temperature reactions, molten-phase operations, and the separation of titanium from impurities through thermal-driven chemical or physical transformations. Based on the type of reducing agent, typical pyrometallurgical methods include high-temperature carbonization, low-temperature chlorination, and metallothermic reduction (e.g., using aluminum or magnesium) [84].

3.2.1. Carbothermal Reduction

The conventional carbothermal reduction route, exemplified by the high-temperature carbonization–low-temperature chlorination process, faces significant sustainability challenges despite its industrial application. These include prohibitively high energy consumption and substantial CO₂ emissions, primarily due to the reliance on solid carbon reductants and extreme operating temperatures.

Subsequent research has, therefore, focused on developing innovative solutions to mitigate these drawbacks. The primary improvement directions involve replacing solid carbon with gaseous reductants (H₂, CO, and CH₄) to enhance reaction efficiency and reduce emissions, and adopting advanced heating technologies like plasma to lower overall energy intensity. The scientific basis for these improvements lies in the superior kinetics and mass transfer properties of gaseous reactants, as well as the high energy density and precise control offered by plasma arcs.

High-temperature carbonization–low-temperature chlorination process: TiC is synthesized via carbothermal reduction in high-titanium slag and coke in an electric furnace at 1600–1800 °C, followed by low-temperature chlorination (400–600 °C) to generate gaseous TiCl₄. The TiCl₄ vapor is subsequently condensed, purified, and separated to obtain titanium tetrachloride with a purity ≥ 99.5% [85]. Key reactions involve sequential carbonization and chlorination steps as shown in Equations (13) and (14):

$${\mathrm{TiO}}_2 + 3\mathrm{C}\to \mathrm{TiC} + 2\mathrm{CO} (\Delta H = +520 \mathrm{kJ}/\mathrm{mol})$$

(13)

$$\mathrm{TiC}+2{\mathrm{Cl}}_2\to {\mathrm{TiCl}}_4\uparrow + \mathrm{C} (\Delta H=-198 \mathrm{kJ}/\mathrm{mol})$$

(14)

Despite advantages such as abundant reductant availability (e.g., coke and coal powder) and high processing capacity, this method faces critical challenges including excessive energy intensity (>1200 kWh/t of electricity (>4.32 GJ/ton slag) and >200 kg of coke/coal per ton of slag) and significant CO₂ emissions from coke/coal combustion (≥1.5 t CO₂/t product) which conflict with “dual-carbon” strategic goals [86,87].

To align with “dual carbon” goals and address the existing challenges in pyrometallurgical titanium extraction technologies, two parallel approaches, including adaptation of novel clean reductants to replace traditional carbon-based reduction methods and implementation of optimized heating technologies, should be taken into consideration in improvement efforts [88,89,90].

Gaseous reductants (e.g., H₂, CO, and CH₄) exhibit significantly superior performance in establishing efficient gas–solid interfaces and improving mass/heat transfer efficiency and reaction kinetics, compared with traditional solid carbonaceous reductants. This innovation provides opportunities for optimizing carbothermal titanium extraction processes [90,91]. Gaseous reducing agents offer three key advantages: firstly, their excellent fluidity and diffusivity enable rapid dispersion around Ti-bearing materials; secondly, their full contact with the materials expands the reaction interfaces and accelerates the reduction kinetics; and, thirdly, they enhance the temperature distribution within the reaction system, minimizing local overheating and undercooling, thus establishing more stable reaction conditions [91,92,93].

FAN G.Q. et al. from Chongqing University employed a H₂–CO–CH₄ mixed gas system (H₂ = 2:1, flow rate: 5–10 L/min) as a novel reductant, achieving efficient gas–solid interfacial reactions that reduced the reduction temperature to 1200 °C with 37% lower energy consumption compared to conventional methods [94].

Furthermore, FAN G.Q. et al. developed a low-temperature CH₄–H₂–N₂ mixed gas reduction-carbonitriding process for Ti-bearing blast furnace slag, replacing high-temperature carbonization. The process route involves two key stages: First, TiO₂ reduction to generate Ti (C, N, O) by-products, and, in addition, low-temperature chlorination in Cl₂ atmosphere to produce TiCl₄. Figure 9 shows the process of CH₄–H₂–N₂ mixed gas reduction and low-temperature chlorination. The optimized protocol demonstrates enhanced phase transformation efficiency while minimizing energy-intensive thermal treatment steps [95].

Key reactions of the process are as shown in Equations (15)–(19) [94,95]:

methane thermal desorption reaction:

$$\begin{array}{c}\mathrm{F}\mathrm{e}\\ {\hspace{1em}\hspace{1em}\mathrm{CH}}_4\rightleftarrows \mathrm{C} + 2{\mathrm{H}}_2\end{array}$$

(15)

water–gas conversion reaction:

$$\begin{array}{c}\mathrm{F}\mathrm{e}\\ {\mathrm{C} + \mathrm{H}}_2\mathrm{O}\rightleftarrows \mathrm{C}\mathrm{O} +{ \mathrm{H}}_2\end{array}$$

(16)

Boudouard reaction:

$$\begin{array}{c}\hspace{1em}\hspace{1em}\mathrm{F}\mathrm{e}\\ {\mathrm{C} + \mathrm{CO}}_2\rightleftarrows 2\mathrm{C}\mathrm{O}\end{array}$$

(17)

carbonitrided reaction:

$$\begin{array}{c}\mathrm{F}\mathrm{e}\\ \hspace{1em} {\mathrm{CaTiO}}_3 + 3\mathrm{C}{\mathrm{H}}_4 \rightleftarrows \mathrm{T}\mathrm{i}\mathrm{C} + \mathrm{C}\mathrm{a}\mathrm{O} + 2\mathrm{C}\mathrm{O} + 6{\mathrm{H}}_2\end{array}$$

(18)

$$\begin{array}{c}\mathrm{F}\mathrm{e}\\ {\mathrm{CaTiO}}_3+2\mathrm{C}{\mathrm{H}}_4+0.5{\mathrm{N}}_{2 }\rightleftarrows \mathrm{T}\mathrm{i}\mathrm{N}+\mathrm{C}\mathrm{a}\mathrm{O}+2\mathrm{C}\mathrm{O}+4{\mathrm{H}}_2\end{array}$$

(19)

Through the implementation of gaseous reduction process, the remarkable results have been achieved: conversion efficiency has been enhanced to 97.71% by means of optimized gas–solid reaction kinetics; energy intensity has been reduced by lowering the operating temperature from 1600 °C (conventional methods) to 1200 °C, with energy consumption decreased to <75% of that of traditional approaches; reactant porosity has been enhanced via urea/sawdust additives, leading to a 20–30% reduction in gas transport resistance; and a 40% acceleration in Ti(C,N) nucleation kinetics has been enabled by Fe₂O₃-catalyzed cracking mechanisms [95,96].

Plasma-assisted carbothermic reduction, a novel heating technology, utilizes direct current (DC)-extended arc technology, where argon ionization between graphite electrodes generates high-temperature plasma (>10,000 °C) through Joule heating, enabling rapid reaction kinetics. The rotating plasma arc enhances heat/mass transfer by homogenizing molten pool temperatures [97].

Samal et al. employed a 30 kW DC-extended arc plasma reactor, achieving complete reduction within 25 min through two core mechanisms: first, argon plasma generation-sustaining ultrahigh temperatures (>10,000 °C) for efficient TiO₂→TiC conversion; second, intensified plasma rotation eliminating thermal gradients (<±50 °C variance) to enhance mass/heat transfer [98]. Figure 10 illustrates the working principle of the plasma heating reactor.

This approach demonstrates notable merits: (1) 30–40% lower energy consumption versus conventional electric furnaces via staged reduction and minimized thermal dissipation; (2) precise reaction controllability can be realized via adjustable current (300 A) and voltage (50 V) parameters, preventing localized overheating through optimized temperature field regulation; and (3) sustainability is further enhanced by argon shielding that limits oxidation losses (<5% Ti wastage), while recyclable slag phases (e.g., Ca–Al–Ti oxides) contribute to closed-loop material flows [98,99].

Future research directions for this novel technology will primarily focus on two critical aspects: First, the integration of plasma reactors with sustainable energy systems (e.g., solar photovoltaic or nuclear power) will be systematically investigated to achieve carbon-neutral titanium extraction. Second, an advanced intelligent control framework incorporating machine learning algorithms will be developed to optimize key process parameters (such as electrical current density, etc.), thereby enhancing operational efficiency and enabling precise prediction of phase equilibria in complex metallurgical systems [99,100].

3.2.2. Metallothermic Reduction

Traditional single-stage metallothermic reduction processes are often hampered by issues such as localized heat accumulation, which triggers side reactions and impurity formation, as well as inefficient separation between the reductant and products, compromising purity and yield.

To address these limitations, the multi-stage reduction methodology has been developed as a key technological advancement. This approach provides a stronger scientific basis for efficient recovery by strategically employing different reductants in a stepwise manner. Its effectiveness is grounded in (1) the controlled distribution of exothermic heat to prevent overheating and suppress impurities, and (2) the principle of leveraging the differing affinities of reductants (e.g., Mg, Ca) for oxygen across various titanium oxide valence states to achieve deep deoxygenation.

This method employs strongly reducing metals (e.g., Al, Mg, and Na) to reduce titanium compounds (e.g., Ti_xO_y, Ti_xCl_y) into metallic titanium or titanium alloys. Representative industrial processes include the Kroll process (Mg reduction of TiCl₄), Hunter process (Na reduction of TiCl₄), and aluminothermic reduction [101,102]. These techniques fundamentally rely on exothermic displacement reactions between molten metals and titanium precursors under controlled atmospheres [103]. You J. et al. adopted an aluminothermic reduction process to directly synthesize FeTi50-A grade ferrotitanium alloy based on the properties of perovskite resources [104]. Using perovskite as the titanium source, this process achieves efficient alloy formation by controlling Al/Ti molar ratios and reaction temperatures. Compositional analysis reveals 50.2 wt.% Ti content with impurity levels (C, Si, and P) below GB/T 3282-2012 [105] industry standards, offering a green alternative for ferro-titanium production [106,107].

Conventional metallothermic reduction for titanium extraction frequently suffers from localized heat accumulation due to concentrated exothermic reactions, triggering side reactions and consequent impurity formation [108]. Concurrently, inefficient separation between reductants and products compromises titanium purity and yield. These limitations, coupled with challenges in achieving continuous operation, substantially constrain industrial application efficiency. However, the multi-stage reduction methodology effectively addresses these issues. Through stepwise regulation of reaction parameters (e.g., temperature, reductant ratios), it distributes exothermic heat to prevent localized overheating and suppress side reactions [109,110,111]. Simultaneously, with intermediate products being separated stepwise, this approach enhances both reductant utilization efficiency and titanium product purity, thereby establishing foundations for continuous production [112,113].

The multi-stage metallothermic reduction process comprises three critical phases: the first stage, exemplified by metal magnesium as the reductant, involves primary self-propagating Mg-thermal reduction, in which rapid exothermic reactions (Equation (20)) generate sub-stoichiometric titanium oxides (Ti_xO, x < 1) with kinetics controlled by Mg/TiO₂ molar ratio and ignition temperature; the second stage is acid leaching separation, where HCl effectively removes MgO and unreacted Mg (Equations (21) and (22)); however, low solubility of intermediate oxides (e.g., MgTiO₃, Mg₂TiO₄) results in 5% titanium loss; The third stage, using metal calcium as the reductant, entails Ca thermal deep reduction, in which the process is conducted at 980–1050 °C with a Ca/Ti_xO molar ratio >1.5:1, reducing oxygen content to 0.25 wt.% to meet TF-0-grade specifications (Equation (23)) [114,115]. Notably, temperatures exceeding 1100 °C promote refractory calcium titanate phases (e.g., CaTiO₃, Ca₃Ti₂O₇), drastically lowering deoxygenation efficiency (<85%). The key chemical reactions of the process are shown as Equations (21)–(23) [116,117].

Primary self-propagating reaction:

$${\mathrm{TiO}}_2 + \mathrm{nMg}\to {\mathrm{Ti}}_{\mathrm{x}}\mathrm{O} + \left(2 - \mathrm{x}\right)\mathrm{MgO} (\mathrm{x} < 1)$$

(20)

leaching separation reaction:

$$\mathrm{Mg}+2\mathrm{HCl}={\mathrm{MgCl}}_2+{\mathrm{H}}_2\mathrm{O}$$

(21)

$$\mathrm{TiO}+2\mathrm{HCl}={\mathrm{TiCl}}_2+{\mathrm{H}}_2\mathrm{O}$$

(22)

deep reduction reaction:

$${\mathrm{Ti}}_{\mathrm{x}}\mathrm{O}+\mathrm{Ca}\to \mathrm{Ti}+\mathrm{CaO}$$

(23)

Fan S.G. from Northeastern University proposed a “two-stage reduction of TiO₂ via sequential Mg-thermal and Ca-thermal process” leveraging electronegativity differences between different reducing agents and the theory of titanium–oxygen affinity across valence states to produce titanium powder synthesis [118]. The process flow diagram is shown in Figure 11.

The experimental research shows that multi-stage reduction can lower the activation energy, which facilitates the reaction and improves its efficiency. In the process, stepwise reduction lowers the energy barrier for the gradual deoxygenation of TiO₂ (e.g., from TiO₂ to low-valent titanium oxides and then to metallic Ti) by regulating reaction conditions (such as temperature, type, and ratio of reducing agents). This enables the otherwise difficult deep deoxygenation reaction to proceed under milder conditions, reducing reliance on extreme conditions like high temperatures and thus lowering energy consumption. Additionally, the reduced activation energy accelerates the reaction rate, shortens the production cycle, and minimizes the formation of by-products caused by violent reactions, ultimately improving the purity and yield of titanium products and laying a foundation for the industrial application of the process [119,120,121].

3.3. Molten Electrolysis Technology for Titanium Extraction

Conventional molten electrolysis processes for titanium extraction, while promising, are fundamentally constrained by two critical issues: energy-intensive polarization phenomena and uncontrolled dendritic titanium deposition at solid cathodes. These problems lead to high energy consumption, low current efficiency, and contaminated or morphologically poor titanium products, especially when dealing with complex feeds like Ti-bearing blast furnace slag (TBS).

In response to these challenges, subsequent research has pioneered innovative solutions. The most significant technological improvements involve the adoption of liquid metal cathodes (LMCs) to fundamentally alter the reduction and deposition interface, and the integration of electrolysis with downstream vacuum distillation to achieve high-purity separation. The scientific basis for the effectiveness of LMCs lies in their ability to enhance mass transfer, reduce activation energy for reduction, and ensure a homogeneous electrode surface, thereby simultaneously mitigating both concentration and electrochemical polarization while preventing dendritic growth.

Molten electrolysis employs alkali or alkaline metal oxides/halides as electrolytes for titanium extraction at temperatures exceeding their melting points, powered by an external electrical source [122]. This methodology is categorized into two primary approaches: molten salt electrolysis and molten oxide electrolysis (MOE).

In molten salt electrolysis, titanium fluoride-based compounds (e.g., TiF₄, K₄TiF₆, Na₂TiF₆) are commonly incorporated into the electrolyte system, which is typically mixed with alkali metal fluorides (LiF, NaF, and KF) or alkaline earth metal fluorides (CaF₂) to adjust the melting point, ionic conductivity, and stability of the molten medium [123]. These titanium fluoride species dissolve to form Ti-containing ions (e.g., TiF₆²⁻, TiF₄), which migrate directionally under an electric field. Titanium ions are reduced to metallic titanium at the cathode through electron transfer. In contrast, MOE, exemplified by the FF-Cambridge process, enables direct electrolytic reduction in solid TiO₂. During this process, oxygen in TiO₂ is ionized, dissolved into the molten salt, and discharged at the anode, while metallic titanium forms at the cathode [124,125].

For molten salt electrolysis systems where oxygen is generated at the anode (e.g., when TiO₂ or oxide-containing feeds are used, leading to O²⁻ oxidation: 2O²⁻ → O₂(g) + 4e⁻), and two key strategies are employed to ensure the production of high-purity metallic titanium: First, the aforementioned titanium fluoride-based electrolyte inhibits the dissolution of anode-generated oxygen into the molten medium—fluoride ions (F⁻) exhibit stronger affinity for alkali/alkaline earth cations than O²⁻, reducing the risk of oxygen re-incorporation into the cathode region. Second, coupling electrolysis with post-treatment processes (e.g., vacuum distillation, as discussed later) further removes residual oxygen: the low vapor pressure of oxygen facilitates its evacuation under vacuum, while the separation of titanium from liquid metal cathodes (e.g., Sn, Sb) avoids oxygen entrapment in the final titanium product.

Electrochemical smelting offers advantages such as simplified processes, cost efficiency, and reduced wastewater generation, positioning it as a promising alternative to the Kroll method [126]. However, inherent challenges, including electrolytic polarization and electrode deposition, critically affect energy consumption, product quality, and process scalability. Polarization arises from the disparity between electron migration rates and electrode reaction kinetics, comprising electrochemical polarization and concentration polarization. Electrochemical polarization: Caused by sluggish electrode reaction rates, leading to deviations from the equilibrium potential. Concentration polarization: Results from reactant depletion at the electrode surface, further disrupting potential equilibrium [127]. Uncontrolled titanium deposition on cathodes induces dendritic growth, surface roughness, and resistive “nodules,” impairing electron transfer and increasing energy demands. Co-deposition of impurities (e.g., Si, Al) further compromises titanium purity, particularly when utilizing Ti-bearing secondary resources like blast furnace slag (TBS) [128,129]. Molten salt electrolysis demands high-purity feedstocks to avoid impurity interference, while MOE faces challenges such as poor electrolyte conductivity, electrode corrosion, and preferential reduction in silica in TBS, which generates silicon contamination [130].

To mitigate or eliminate issues such as increased energy consumption and reduced electrolysis efficiency caused by polarization phenomena, specific depolarization measures need to be taken, including the use of depolarizers, optimization of electrode materials, improvement of electrolysis conditions (e.g., controlling appropriate temperature, adjusting current density, and enhancing stirring), and improvement of electrolysis equipment design [128,129,130]. In recent years, many scholars have conducted extensive research on solving the problems of polarization and electrode deposition in the electrolytic preparation of metals.

Effective strategies, such as a liquid metal cathode and a hybrid MOE-vacuum distillation process, have been proven effective in achieving depolarization and improving electrode deposition in molten oxide electrolysis.

Liquid metal cathodes (LMCs), typically comprising Sn-/Sb/Cu-based low melting point metallic systems, have been developed as advanced electrodes to address polarization and deposition challenges in molten oxide electrolysis (MOE) processes. The mechanism is illustrated in Figure 12. Under the influence of an applied external electric field, Ti⁴⁺ cations within the molten slag migrate electromagnetically towards the liquid metal cathode. Upon reaching the interface between the molten slag and the liquid metal cathode, Ti⁴⁺ undergoes electrochemical reduction at the cathode surface via the following reaction: Ti⁴⁺ + 4e⁻ → Ti. The resultant metallic titanium atoms subsequently dissolve and diffuse into the bulk of the liquid metal cathode, forming a homogeneous Ti-bearing alloy. Concurrently, anions such as O²⁻ within the molten electrolyte migrate towards the anode. At the anode surface, these anions are oxidized according to the following reaction: 2O²⁻ → O₂(g) + 4e⁻, liberating oxygen gas.

Liquid Metal Cathodes offer several notable benefits [131,132,133,134]. Firstly, the high fluidity of liquid metal cathodes (e.g., Sn, Bi, and Cu) promotes the rapid diffusion of metal ions (such as Ti⁴⁺) to the cathode surface, effectively alleviating concentration polarization caused by slow ion migration. Simultaneously, convective flow within the liquid cathode facilitates the removal of reaction products from the electrode interface, further enhancing mass transfer efficiency. Secondly, the unique physicochemical properties of liquid metals reduce the activation energy required for titanium ion reduction, thereby minimizing electrochemical polarization and improving both reaction kinetics and current efficiency. Thirdly, the superior electrical conductivity and fluidity of liquid cathodes ensure a homogenized electric field distribution across the electrode surface, preventing localized current hotspots and dendritic growth. This is exemplified in molten Ti-bearing slag (TBS) electrolysis, where liquid cathodes (e.g., Sn, Sb) selectively suppress silicon co-deposition while enabling efficient titanium recovery, addressing a critical limitation of solid cathodes.

As demonstrated by Jiao H.D., replacing the solid cathode in composite anode electrolysis systems with a liquid metal cathode (LMC) enables the formation of non-dendritic liquid deposits, significantly simplifying product collection [135]. Furthermore, in molten oxide electrolysis (MOE), the use of LMCs promotes the diffusion of reduced metallic species (e.g., Ti) into the liquid cathode matrix. This diffusion mechanism effectively suppresses the secondary oxidation of titanium, thereby reducing oxygen content in the final product. The resulting Ti-containing liquid metal can be directly utilized as a titanium intermediate alloy without additional purification steps.

Pu Z.H. et al. employed a combined process of molten oxide electrolysis (MOE) and vacuum distillation refining to treat Ti-bearing blast furnace slag, producing titanium alloys and titanium metal [136]. Figure 13 presents the technological process of titanium extraction via liquid metal molten oxidation electrolysis and vacuum distillation. Building upon the liquid metal cathode electrolysis approach, this study leveraged the significant difference in saturated vapor pressure between titanium and tin (or antimony) to refine the electrolytically produced tin-cathode and antimony-cathode alloys via vacuum distillation. Under the optimized distillation parameters of 1500 °C, 1 Pa vacuum, and a duration of 4 h, the vacuum distillation refining of the tin-cathode alloy yielded an Sn-Ti alloy with a titanium content exceeding 80%. Similarly, subjecting the antimony-cathode alloy to vacuum distillation refining at 1100 °C, 1 Pa, and for 3 h resulted in titanium metal with a purity surpassing 98% [136,137].

Future research priorities include optimizing electrode architectures and electrolyte formulations through in-depth studies of polarization and deposition mechanisms, which are pivotal for enhancing process efficiency and stability [138].

3.4. Selective Precipitation Titanium Extraction Technology

The industrial application of selective precipitation technology for titanium enrichment from complex slags like TBS is fundamentally constrained by two inherent challenges: thermodynamic–kinetic conflicts in multiphase systems that make simultaneous control over perovskite, anosovite, and rutile formation exceedingly difficult, and crystal growth–separation trade-offs where measures to promote growth often increase slag viscosity and hinder subsequent physical separation due to low density contrasts.

Subsequent research has, therefore, focused on innovative pathways to decouple these challenges. The primary technological improvements have evolved along two key fronts: the precise control over the crystal morphology of the enriched Ti-bearing phase to drastically improve its subsequent separation efficiency, and the development of intensified separation processes to overcome the inherent limitations of conventional methods. The scientific basis for the effectiveness of these strategies is thoroughly elaborated in the studies that follow, providing a stronger rationale for titanium recovery.

The fundamental principle of selective precipitation for titanium extraction involves creating tailored physicochemical conditions to drive the selective migration and enrichment of valuable elements (e.g., Ti) from dispersed mineral phases into designed target phases, governed by a “chemical-potential-gradient-driven” mechanism [139,140].

Grounded in physicochemical thermodynamics, solid-state physics, and solidification principles, this theory of “selective precipitation technology” systematically addresses the physicochemical behavior of metallurgical slags. And it was applied to separate titanium components from Ti-bearing blast furnace slag characterized by complex mineral phases and multi-element coexistence. The technical framework of this theory comprises three stages: first, selective enrichment: Creating favorable conditions (e.g., oxygen potential, additive ratios) facilitates the selective mitigation and enrichment of the valuable elements dispersed in various mineral phases into the target phases (e.g., perovskite, anosovite, and rutile) under the chemical potential gradient; second, selective precipitation and growth: Rational control of modifiers and cooling rates enlarges crystal sizes of target phases in slag; and third, selective separation: Utilizing mineral processing methods (e.g., gravity separation, flotation) to isolate enriched phases based on density, morphology, and chemical stability [141,142,143,144].

Despite its theoretical promise, the industrial implementation of selective precipitation technology for titanium enrichment faces two fundamental constraints. On one hand, thermodynamic–kinetic conflicts arise from multiphase coexistence. For instance, perovskite (CaTiO₃) nucleation requires precise oxygen potential control (Δμ₀ = −120 to −80 kJ/mol) and cooling rates (0.5–5 °C/min), while anosovite ((AO·2TiO₂) x·(B₂O₃·TiO₂) y) crystallization exhibits extreme sensitivity to SiO₂ content (42.0–44.9 wt.%), where feedstock variability destabilizes phase control. On the other hand, crystal growth–separation trade-offs wherein viscosity modulation (η < 5 Pa·s) for target phase growth (e.g., rutile) risks additive-induced glass formation (η > 10 Pa·s) and encapsulation, compounded by low density contrast (Δρ < 0.5 g/cm³) limiting gravity separation to <50% efficiency [145,146,147,148,149,150].

To address the limitations of selective precipitation technology, efforts must encompass the use of advanced additives, optimization of separation processes, and application of multiscale modeling to bridge lab-to-industry gaps. Building on these foundational measures, recent advancements have further focused on three key fronts: crystallographic optimization, process innovation, and equipment development [151,152,153,154].

Han et al. demonstrated that rutile morphology control is critical for efficient settling and separation of titanium components in Ti-bearing mixed molten slag [155,156]. The conceptual diagram of the 100 kg pilot-scale test setup is shown in Figure 14. And the flow chart of the technology for rutile transformation, settling, and separation of titanium components in Ti-bearing mixed molten slag is shown in Figure 15.

Based on the related derivation of Stocke’s settling formula (Equation (24)), the study found that the spherical rutile crystal exhibits the highest settling velocity, followed by the cubic rutile crystal, while the cuboid one shows the slowest [157,158]:

$$v=2r^2\Delta \rho g/9\eta$$

(24)

The laboratory research has shown that the microscopic morphology of the rutile crystal in the modified slag will change from cuboid to cube, then to sphere when the TiO₂ content of raw materials increased from 27 wt.% to 47 wt.% [155,156,157,158]. The expanded experiment results indicate that the volume fractions of rutile crystals in the lower part modified slag significantly more than that in the upper shown in Figure 16.

Based on equilibrium thermodynamic analysis of the CaO–SiO₂–MgO–Al₂O₃–TiO₂ quinary slag system and combined with investigations into the standard Gibbs free energy of relevant reactions, the formation mechanism of rutile in modified slag was comprehensively elucidated. Through systematic parameter control, the synergistic effect of moderate SiO₂ addition (5–10 wt.%) and precisely regulated oxidation time (126 s) was conducive to drive the phase evolution from metastable titanium phase (perovskite, anosovite, and titanite) toward thermodynamically stable rutile, reducing glass phase formation (<5 vol.%) and slag viscosity (η < 3 Pa·s), thereby improving titanium mobility (Diffusion coefficient: D > 1 × 10⁻⁸ m²/s). However, excessive SiO₂ content (>15 wt.%) induced a substantial increase in slag viscosity (η > 10 Pa·s) while simultaneously facilitating glass phase formation [155,156,157,158,159,160,161].

Guo et al. designed a supergravity separation system which leverages liquid–solid coexistence to achieve phase partitioning via high-speed centrifugal devices, enabling selective extraction of the enriched fraction [162,163,164]. The principle of hypergravity separation is shown in Figure 17.

The gravity coefficient is the ratio of the supergravity acceleration to the standard acceleration, and the calculation formula is shown as Equation (25) [162]:

$$G=\sqrt{g^2+(w^2{R}^2)}/g$$

(25)

The relationships in the formula are as follows:

G—Gravity coefficient;

ω—Rotational angular velocity (rad/min);

g—Standard gravitational acceleration (g = 9.80 m/s²);

R—The distance between the axis of rotation and the center of the sample (m).

Based on the CaO–SiO₂–MgO–Al₂O₃–TiO₂ quinary slag system, this study identified a solid–liquid coexistence zone (1200–1337.4 °C) through optimized process parameters. Employing a custom-designed supergravity separation system under the conditions of G = 600, 1280 °C, and 5 min separation, rutile with a purity of 95.56% was obtained, demonstrating a 60% efficiency improvement over conventional gravity separation methods. However, temperature fluctuations (±20 °C) risk rutile remelting, and industrial energy consumption exceeds traditional processes by 25% [162,163,164,165].

The primary focus of future development will lie in the following three aspects: first, the development of low-viscosity and green additives (e.g., CeO₂, La₂O₃) for effective glass phase suppression; second, the application of multiscale modeling approaches such as the coupling of density functional theory with computational fluid dynamics (DFT + CFD) to achieve thermodynamic–kinetic synergy; and third, the advancement of high-temperature resistant materials (e.g., SiC ceramics), which are essential for ensuring industrial feasibility [166,167,168,169,170].

3.5. Comparative Analysis of Titanium Extraction Processes from TBS

A comparative summary of the key titanium extraction processes discussed in this review, including their applicable TBS composition, optimum operational conditions, and performance metrics, is provided in

Table 4. Comparative summary of titanium extraction processes from TBS.

Extraction Method	TBS Composition (TiO2 wt.%)	Optimal Conditions	Ti Recovery/Outcome	Key Advantages	Key Challenges	Ref.
Hydrometallurgical Leaching
Conventional H2SO4	20–29% (BFS)	Acid conc. < 90%, T ~ boiling point	~80% (estimated)	Simple process, wide adaptability	High acid consumption, severe corrosion, waste acid, and gypsum colloids	[66,67,68,69,70]
Activation Roasting-Leaching	20–29% (BFS)	Acid conc. > 92%, Act. Roasting Temp.130 °C, Act. Roasting Temp.40 min, Acid-to-slag = 1.4	>82.85%	Prevents gypsum colloids, improves filtration, and lowers acid use	Spent acid treatment required, process complexity	[71,79]
Pyrometallurgical Reduction
Carbothermal Reduction	≥75% (EFS)	1600–1800 °C, Coke reductant	80–90%	Abundant reductant, high capacity	Excessive energy (>4.32 GJ/t), high CO2 emissions (≥1.5 t/t)	[85,86,87]
Gaseous Reduction (H2/CO/CH4)	22–25% (BFS)	1200 °C, H2/CO/CH4	97.71% conversion	Lower temp, lower energy consumption (↓37%), superior kinetics	Requires a gas handling system, process control	[94,95,96]
Molten Electrolysis
Liquid Metal Cathode (MOE)	~22–25% (BFS)	Current density: ~0.5–0.8 A/cm2, T: 900–1000 °C, Cathode: Sn or Cu	92–98%	Prevents dendritic growth, suppresses Si co-deposition	Requires downstream separation	[131,132,133,134,135]
MOE + Vacuum Distillation	~22–25% (BFS)	Electrolysis: (as above); Distillation: 1500 °C/1 Pa/4 h (Sn) or 1100 °C/1 Pa/3 h (Sb)	>98% purity	Produces high-purity Ti metal or alloy	High energy consumption for distillation	[136,137]
Selective Precipitation and Separation
Rutile Transformation-Gravity Sedimentation	~22–25% (BFS, Modified)	SiO2 addition: 5–10%, Oxidation time: 126 s, Controlled cooling	98.73%	Forms separable rutile crystals, lower energy than supergravity	Requires precise composition and thermal control, slower separation speed	[155,156,157,158,159,160,161]
Supergravity Separation	~22–25% (BFS, Modified)	temp.1300 °C, super-gravity G = 800, duration 6 min	95.37% purity	High separation efficiency, high-purity product	Sensitive to temperature fluctuations, high energy use	[162,163,164,165]

. This synthesis aims to facilitate a clear comparison of the technological maturity, efficiency, and applicability of each method.

4. Innovative Insights on Industrial Titanium Extraction

Despite acknowledging the partial efficacy of current mainstream titanium recovery technologies in recovering titanium components, they face critical technical bottlenecks in industrial-scale applications. These limitations, including secondary pollution risks, excessive energy consumption, and suboptimal recovery rates, were previously outlined. Crucially, when scaling to capacities exceeding 500,000 tons per annum (t/a), existing processes generally lack research on process stability verification under continuous operation. Furthermore, a systematic research model addressing the long-term operational impacts on equipment corrosion, slag phase reconstruction, and product quality has yet to be established [171,172,173,174].

Ti-bearing slags, byproducts of ironmaking with annual production exceeding 10 million tons, exhibit high tapping temperatures (1400–1550 °C) and contain substantial waste heat (1.5–2.0 GJ/ton) [172,173,174,175,176]. Conventional approaches involving cooling followed by crushing and sorting suffer from waste heat dissipation, titanium dispersion in glassy matrices, and prohibitive separation costs [177,178,179,180,181,182]. Industrial-scale processing (>500,000 t/a) poses significantly greater challenges than laboratory-scale tests (100–1000 kg batches), including semi-solidification defects and excessive thermal gradients during large-scale slag cooling [183,184,185,186,187,188,189,190]. These induce heterogeneous crystal growth, hindering efficient titanium enrichment into the targeted mineral phase. To address these issues and fully harness the waste heat of the slag, our research group proposes an integrated technological framework: “Online conditioning driven by waste heat–mineral phase reconstruction–directional crystallization–optimized liberation”.

This technology innovatively adopts the integration of electromagnetic stirring (EMS) and high-speed inert-gas injection to enhance melt convection, supplemented by an auxiliary heating system. The conceptual schematic of this technology is illustrated in Figure 18. This synergistic configuration achieves precise control of radial temperature gradients below 10 °C. Under electromagnetic stirring, the internal structure of the melt progressively achieves homogenization through forced convection. Furthermore, the injection of inert gas generates abundant microbubbles that, under the synergistic coupling effects of buoyancy force, initial injection velocity, and electromagnetic stirring, facilitate the uniform dispersion of conditioning agents within the molten matrix. This multiphase interaction mechanism effectively enhances the spatial uniformity of composition distribution, thereby accomplishing the desired metallurgical homogenization effect through coordinated bubble transport and electromagnetic-driven mass transfer processes [191,192,193,194,195,196,197].

This technology can effectively prevent semi-solidification defects and eliminate excessive thermal gradients during the large-scale slag cooling process through the synergistic action of electromagnetic stirring (EMS) and high-speed inert-gas injection, thereby eliminating heterogeneous crystal growth. The dissolution and diffusion of conditioning agents in the molten slag will be well driven by efficiently harnessing high-temperature waste heat from slag, achieving homogeneous distribution within the melt via the synergistic EMS-gas-heating system. This homogeneous melt environment and optimized crystal growth state can further promote the directional enrichment of titanium components into titanium-rich phases, ultimately improving the recovery rate of titanium components in titanium slag, which is also one of the core objectives of this study. This provides an innovative pathway for industrial titanium extraction technology.

5. Conclusions

This review has critically analyzed technologies for titanium extraction from Ti-bearing slag (TBS). The complex mineralogy and fine dispersion of titanium within stable phases present significant challenges for efficient recovery. Based on the assessment, the following conclusions are drawn:

(1)

TBS is a multi-component byproduct wherein titanium is primarily locked within refractory minerals. This intricate structure necessitates aggressive extraction conditions, often leading to high energy consumption and environmental concerns.

(2)

Comparative analysis of mainstream technologies reveals a performance–energy trade-off. Hydrometallurgical methods are improved by activation roasting, while pyrometallurgical routes benefit from cleaner gaseous reductants to lower energy intensity. For molten electrolysis, liquid metal cathodes not only prevent dendritic growth but also resolve electrolytic deposition and polarization. In selective separation, techniques with supergravity integration and optimized crystal growth kinetics enable high-purity products, albeit with sensitivity to process parameters.

(3)

Future development must focus on low-carbon processes (use of hydrogen for pyrometallurgical treatment) and the synergistic recovery of all valuable elements via hydrometallurgical treatment. The proposed strategy of “Online conditioning–mineral phase reconstruction–directional crystallization–optimized liberation” is recommended to harness the slag’s waste heat, prevent inhomogeneous crystallization, and enable efficient titanium recovery on an industrial scale.