Recovery of Rare Earth Elements from Calciothermic Reduction Slag by Sulfation Roasting–Water Leaching Method

Abstract

The calciothermic reduction slag (CRS) generated in heavy rare earth metal production, is rich in rare earth elements (REE) and highly amenable to recovery. In the present study, the CRS was treated with a H₂SO₄ roasting–water leaching method for the recovery of REEs. The feasibility of this process was confirmed by thermodynamic analysis. Key roasting and leaching factors governing the leaching efficiency of REE were identified and optimized. The maximum REE extraction efficiency reached 94.65% under the optimal conditions: roasting at 150 °C for 240 min with 15 mL of H₂SO₄, followed by water leaching at 20 °C for 60 min at a liquid–solid ratio of 15:1. Results of XRD, SEM, and EDS revealed that the REEs in the CRS were transformed into water-soluble rare earth sulfates after roasting. In the leaching process, the rare earth sulfate is efficiently extracted, whereas CaSO₄ has low solubility in water. A CaSO₄ product with a 98.10% purity was obtained with a calcium recovery of 90.79%, and the removal rate of fluorine in the CRS was 99.99%. The leaching kinetics of the REEs follow a diffusion plus interfacial transfer model with an apparent activation energy of –46.45 kJ·mol⁻¹. This study demonstrates that sulfation roasting–water leaching is a viable route for the comprehensive utilization of CRS.

1. Introduction

Thanks to their outstanding physical and chemical properties, rare earth metals are widely used across multiple industries, including metallurgy, military, high-performance materials, petrochemicals, and renewable energy [1,2,3,4]. By the end of 2024, rare earth production across the globe totaled about 390,000 tons. China, as the world’s largest producer, contributed about 270,000 tons to this total. Rare earth metals are primarily produced using two key techniques: molten salt electrolysis and calciothermic reduction. The former method is generally used for extracting light rare earth elements from rare earth chloride, whereas the latter is designed for the extraction of medium and heavy rare earth metals from rare earth fluoride. During the thermal reduction process, rare earth fluorides are reduced to metals via calcium metal under high-temperature vacuum conditions, simultaneously generating CRS with fluorite (CaF₂) as its primary constituent [5,6,7]. Approximately 2 tons of CRS are created per ton of refined rare earth metal. The rare earth element (REE) content in CRS ranges from 2% to 13% and exists in fluoride, metallic, and oxide forms [8]. The REEs in CRS, while less abundant than those in molten salt slag (10% to 80%) [9], are predominantly valuable elements like terbium and dysprosium, which hold considerable economic significance [10]. At present, the lack of practical recycling technologies leads companies to commonly store CRS as a stopgap measure. As rare earth resources dwindle rapidly and environmental protection requirements grow stricter, developing innovative approaches for the efficient utilization of CRS has become increasingly critical.

The academic focus on extracting rare earth elements from CRS has intensified in recent times. Chen et al. treated the CRS with a direct acid leaching technique; however, due to the limited solubility of rare earth fluorides in acidic media, the rare earth leaching rate is capped at 65% [11]. Therefore, converting rare earth fluorides in CRS into acid-soluble compounds is crucial for achieving efficient rare earth element extraction. Experimental results from Xia et al. revealed that utilizing NaOH roasting followed by an acid leaching treatment with CRS could obtain an exceptional REE recovery efficiency of 92.30% [12]. The use of additives (NaOH, Na₂SiO₃, and Na₂CO₃) was explored by Liang et al. for CRS leaching, resulting in REE leaching rates of 92.30%, 99.05%, and 94.09%, respectively. [13,14]. Huang et al. showed that the extraction rate of REEs reached 95.48% by using a CaO roast–acid leaching process on CRS [15]. These processes facilitate the transformation of rare earth fluorides into their oxide forms. Nevertheless, sodium-based roasting requires water washing for fluoride removal, generating large volumes of low-concentration fluoride-containing wastewater, while calcium-based roasting necessitates excessive amounts of additives, with residual additives leading to excessive acid consumption during subsequent processing. Additionally, the fluoride and calcium components in CRS remain underutilized [16]. Therefore, it has become imperative to establish an economically viable recycling approach for the complete utilization of CRS.

Sulfation roasting, a pretreatment process that introduces sulfur-containing agents (e.g., H₂SO₄, sulfides, or SO₂ gas) to convert metal compounds into water-soluble sulfates at elevated temperatures, is widely employed in cobalt/nickel metallurgy, zinc smelting, and rare earth mineral processing. When applied to bastnasite, this method converts rare earth elements into rare earth sulfates, which are subsequently leached via cold-water immersion, while fluorine is transformed into HF gas. The gaseous HF is then absorbed by Na₂CO₃ to generate NaF as a byproduct. Inspired by this approach, this study adopts sulfuric acid roasting to treat CRS, aiming to achieve a comprehensive utilization of its rare earth, calcium, and fluorine components.

Systematic investigations were carried out to evaluate the impact of calcination and leaching conditions on REE recovery rates, and kinetic calculations for the leaching process were performed. The transformation behavior of calcium and fluorine during roasting and leaching processes was also investigated, offering insights into the integrated utilization of these elements.

2. Materials and Methods

2.1. Materials

The CRS sample was sourced from a rare earth metallurgical plant in Ganzhou, Jiangxi Province, China. Chemical analysis indicated that the CRS contained 2.33% rare earth oxides (REOs), 36.74% fluorine, 51.01% calcium, 0.09% aluminum, and 0.02% iron. The rare earth element distribution showed Tb₄O₇ (94.20%) as the dominant phase, followed by Nd₂O₃ (2.42%), Y₂O₃ (0.70%), Dy₂O₃ (0.63%), and Tm₂O₃ (0.42%). As shown in the XRD patterns (Figure 1), the CRS contains CaF₂ ( PDF#35-0816) as the primary crystalline phase, along with small amounts of Ca(OH)₂ (PDF#44-1481) and CaCO₃ (PDF#-05-0586). Before use, the CRS was ground to 100% to pass through a 0.074 mm sieve. All chemical reagents employed in this study were analytical-grade materials sourced from Xilong Scientific Co., Ltd. (Guangdong, China), with deionized water being utilized for all experimental procedures.

The results of SEM and EDS analyses on the CRS are, respectively, displayed in Figure 2 and

Table 1. Elemental composition (wt%) obtained by EDS analysis at specified points in Figure 2.

Point	O	F	Ca	Tb	Phase
1	0.17	-	0.99	98.83	Terbium metal
2	12.80	3.18	16.29	67.73	Terbium oxide
3	1.79	38.67	59.24	0.29	Fluorite
4	0.60	-	1.19	98.20	Terbium metal
5	6.72	-	1.30	91.98	Terbium oxide
6	2.14	41.93	55.61	0.32	Fluorite
7	1.92	41.85	55.96	0.27	Fluorite
8	1.20	-	1.46	97.33	Terbium metal
9	0.22	-	1.79	97.99	Terbium metal

. It is revealed by observations that REEs are predominantly present in the CRS in the forms of metals and oxides, with a substantial fraction occurring as isomorphic substitutions in the CaF₂ lattice. Although several rare earth particles measure around 20 μm, most exist at the micron scale (below 10 μm) and are tightly encapsulated within CaF₂ matrices. The presence of microfine rare earth metal particles presents a significant obstacle to effective extraction via physical separation and leaching. Additionally, the isomorphic presence of REEs within CaF₂ makes them inherently resistant to direct leaching by acids [17].

2.2. Methods

The recovery process is illustrated in Figure 3. In a typical run, 20 g of CRS was blended uniformly with varying volumes of H₂SO₄ (10–20 mL in 2.5 mL increments) in a corundum crucible. The mixture was then transferred to a preheated muffle furnace (KSL-1400X, Kejing, Heifei, China) and roasted at controlled temperatures (100–300 °C in 50 °C increments) for different durations (60–300 min in 60 min intervals). Post-roasting, samples were air-cooled to ambient temperature.

For leaching, a temperature-regulated water bath maintained precise thermal conditions. The roasted product was added to deionized water at a predetermined liquid–solid ratio and agitated at 400 rpm using a mechanical stirrer. After the prescribed leaching period, vacuum filtration isolated the leachate from residual solids. REE concentrations in the filtrate were quantified by EDTA titration [18], enabling leaching efficiency calculations.

The chemical compositions of the CRS and the leach residue were analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS). Phase composition identification was conducted using an X-ray diffractometer (XRD, Ultima IV, Rigaku, Tokyo, Japan) equipped with Cu Kα radiation and a secondary monochromator. Scans were performed over a 2θ range from 10° to 80°. Microstructural and elemental analyses were carried out using scanning electron microscopy coupled with energy-dispersive spectroscopy (SEM-EDS, MLA650F, FEI, Hillsboro, OR, USA). Prior to SEM observation, the CRS was mounted in epoxy resin and polished. For the leached residue, one subset of samples was examined directly, while another was similarly mounted and polished. All samples were sputter-coated with a thin layer of gold to ensure adequate electrical conductivity during imaging.

3. Results and Discussion

3.1. Thermodynamic Analysis

The chemical reactions potentially occurring during the sulfation roasting process of the CRS are listed in

Table 2. Possible chemical reactions and their ΔG in the roasting system.

Equation	Reactions	ΔG(kJ·mol−1)
(1)	2Tb + 3H2SO4 = Tb2(SO4)3 + 3H2(g)	0.0005T2 + 0.6122T − 1510.80
(2)	Tb2O3 + 3H2SO4 = Tb2(SO4)3 + 3H2O(g)	0.0005T2 + 0.4496T − 415.04
(3)	2TbF3 + 3H2SO4 = Tb2(SO4)3 + 6HF(g)	0.0003T2 − 0.3870T + 9.57
(4)	CaF2 + H2SO4 = CaSO4 + 2HF(g)	0.00006T2 − 0.2307T − 4.65
(5)	CaCO3 + H2SO4 = CaSO4 + H2O(g) + CO2(g)	0.00001T2 − 0.1421T − 135.01
(6)	Ca(OH)2 + H2SO4 = CaSO4 + 2H2O(g)	−0.00003T2 − 0.0048T − 211.04

[19,20], and their thermodynamic feasibility was analyzed using FactSage 8.3. The results are, respectively, presented in Figure 4 and Table 2 [21,22].

Figure 4 shows the ΔG for Equations (1)–(6). Across the temperature interval of 0 °C to 400 °C, the ΔG of Equation (3) exhibits a gradual transition from positive to negative values with an increasing temperature. This thermodynamic behavior indicates that the reaction between H₂SO₄ and TbF₃ becomes increasingly spontaneous under elevated temperatures, and Equation (4) exhibits a similar trend. For Equations (5) and (6), the ΔG is negative and becomes even more negative as the temperature rises. In contrast, although the ΔG of Equations (1) and (2) is also negative, it becomes less negative with increasing temperatures. The thermodynamic results demonstrate that converting rare earth components in CRS into rare earth sulfates through sulfation roasting is feasible. Additionally, Ca and F will be transformed into CaSO₄ and HF, respectively.

3.2. Influence of Roasting Parameters on REE Leaching

The effects of the roasting temperature, time, and H₂SO₄ dosage on the REE leaching efficiency were evaluated under standardized leaching conditions (20 °C, 60 min, liquid–solid ratio of 12:1 mL/g).

To investigate how the roasting temperature influences REE leaching, tests were conducted at 100–300 °C while holding the roasting time (60 min) and H₂SO₄ volume (20 mL) constant; the outcomes are presented in Figure 5a. As shown in Figure 5, elevating the roasting temperature from 100 °C to 150 °C sharply increased the REE extraction yield from 33.00% to 82.49%; however, a subsequent decline to 67.73% occurred at 300 °C. Increasing the roasting temperature is beneficial for the conversion of rare earth components into rare earth sulfates, thereby improving the subsequent rare earth leaching rate. However, when the roasting temperature exceeds 150 °C, the rapid volatilization of H₂SO₄ leads to a reduction in its effective dosage within the system [23]. Thus, 150 °C was established as the optimal temperature for the subsequent experiments.

With the H₂SO₄ dosage fixed at 20 mL and the roasting temperature at 150 °C, the influence of the roasting duration (60, 120, 180, 240, and 300 min) on REE leaching was examined. Figure 5b demonstrates that extending the roasting duration from 60 min to 240 min resulted in a significant enhancement of the REE leaching efficiency, rising from 82.49% to 92.04%. Extending the roasting time facilitates the conversion of rare earth components into water-soluble rare earth sulfates, thereby improving the REE leaching rate. However, prolonging the roasting process beyond 240 min up to 300 min showed a negligible improvement for the REE extraction. Consequently, 240 min was identified as the most suitable roasting time for optimal results.

To evaluate the impact of the H₂SO₄ dosage (10–20 mL) on the REE leaching efficiency, tests were conducted under fixed conditions (150 °C, 240 min). Figure 5c illustrates that increasing the H₂SO₄ dosage from 10 mL to 15 mL corresponds to a rise in the REE leaching efficiency, which climbs from 78.58% to 92.04%. In chemical reactions, sulfuric acid preferentially reacts with calcium rather than rare earth elements due to calcium’s higher reactivity as an alkaline earth metal. The leaching efficiency of REEs exhibited a clear dependence on the H₂SO₄ dosage. Below 15 mL, insufficient acid availability limited the rare earth extraction. At 15 mL, the stoichiometric acid demand was fully satisfied, achieving maximum leaching rates. Further increases beyond 15 mL showed no significant change, establishing 15 mL as the optimal dosage.

In conclusion, the optimal roasting conditions for the REE leaching from CRS were identified as a temperature of 150 °C, time of 240 min, and H₂SO₄ dosage of 15 mL.

3.3. Influence of Leaching Parameters on REE Leaching

The leaching process is primarily governed by three key parameters: the leaching temperature, time, and liquid-to-solid ratio [24]. A comprehensive study was performed to assess the influence of these variables on the recovery of REEs from roasted CRS under optimum calcination parameters.

The REE extraction was systematically examined at leaching durations of 30–90 min (15 min intervals) under controlled conditions of a 20 °C leaching temperature and a liquid-to-solid ratio of 12:1. As depicted in Figure 6a, the REE extraction efficiency improved significantly from 71.20% to 92.04% when the leaching duration was extended from 30 to 60 min. Further prolonging the leaching time produces little improvement in the extraction yield of REEs. Thus, the optimal leaching time was selected as 60 min.

To evaluate the impact on the REE extraction, leaching temperatures (20–80 °C) were tested under fixed conditions of a 60 min leaching duration and a 12:1 liquid–solid ratio. As presented in Figure 6b, the REE leaching efficiency exhibited an inverse temperature relationship, declining from 92.04% to 72.07% as the leaching temperature increased from 20 °C to 80 °C. This negative correlation stems from the temperature-dependent solubility of rare earth sulfates in aqueous media, with lower temperatures favoring higher REE leaching rates [25,26]. Therefore, 20 °C was chosen for all subsequent leaching experiments.

The extraction efficiency of REEs was investigated across a liquid-to-solid ratio range of 6:1 to 18:1, maintaining constant operational parameters of a 60 min leaching duration and a 20 °C process temperature. Figure 6c reveals that when the liquid-to-solid ratio increased from 6:1 to 15:1, the extraction yield of REEs climbed from 67.73% to 94.65%. This trend occurs mainly because of the decrease in the quantity of solids per amount of lixiviant, lower viscosity of the mixture, increase in the diffusion coefficient, ease of convection, and facilitation of diffusion [27]. Raising the liquid-to-solid ratio beyond 15:1 to 18:1 did not meaningfully alter the REE extraction; therefore, 15:1 was selected as the optimum ratio.

Comprehensive optimization revealed that the highest REE extraction efficiency (94.65%) was obtained under the following conditions: a 60 min leaching time, a 20 °C leaching temperature, and a liquid-solid ratio of 15:1.

In terms of the rare earth leaching efficiency, the leaching rate achieved in the present study is comparable to that reported by Huang et al. (95.48%) [15] but lower than the result obtained by Liang et al. (99.05%) [13,14]. This difference may be attributed to the fact that the rare earth content in the slag used by Liang et al. was nearly twice as high as that in the present study. Regarding roasting conditions, the optimal temperature and time in this work were 150 °C and 240 min, respectively, compared with 1000 °C and 90 min in the study by Huang et al. [15] and 850 °C and 120 min in that by Liang et al. [13,14]. Therefore, the sulfuric acid roasting method employed in the present study for treating CRS may offer greater feasibility compared to the calcification roasting and sodium roasting processes.

3.4. Characterization of Leached Slag

Table 3. Elemental makeup of the post-leaching residue (wt%).

Elements	REO	Ca	F	${\mathbf{S}\mathbf{O}}_4^{\mathbf{-}2}$
Assay	0.068	28.94	0.084	69.79

shows that the purity of the CaSO₄ obtained was as high as 98.10%, and the recovery rate of the calcium was 90.79%, so it can be widely used in construction, agriculture, and other fields [28,29,30,31]. The leached slag contained only 0.068% REOs, with a corresponding fluorine content of 0.084%. It can be seen from the calculation that the defluorination rate of the CRS in the process of the sulfation roasting reaches 99.99%, and the HF in the flue gas can be separated and purified into HF products following the production process of HF, or NaF products can be prepared by absorption with the Na₂CO₃ solution.

The XRD pattern of the leached slag (Figure 7) exclusively exhibited the diffraction peaks of CaSO₄ (PDF#-37-1496), indicating the complete conversion of the original CRS components (CaF₂, CaCO₃, and Ca(OH)₂) to CaSO₄ during the sulfation roasting process.

Figure 8 and

Table 4. Elemental composition (wt%) obtained by EDS analysis at specified points in Figure 8b.

Point	O	F	S	Ca	Tb	Phase
1	45.14	0.48	22.45	31.85	0.07	Calcium sulfate
2	44.78	0.23	22.74	32.14	0.10	Calcium sulfate
3	43.82	0.40	22.50	33.19	0.09	Calcium sulfate

, respectively, display the SEM micrographs and EDS data for the leached residue obtained under optimum conditions. Figure 8 show that the size of the CaSO₄ particles is less than 3 μm, and no rare earth metals or oxides were found. The EDS results revealed that CaSO₄ particles contain only 0.07%–0.19% terbium. The results verify that the REEs originally hosted in CRS in multiple phases were completely transformed into sulfates during roasting and then readily dissolved in the subsequent water leaching.

3.5. Leaching Kinetic Analysis

Rare earth leaching constitutes a classic heterogeneous liquid–solid reaction system. When rare earth particles are assumed to have a spherical morphology, their leaching behavior can be characterized using the unreacted shrinking-core model (SCM) [32]. Potential limiting factors for the leaching process include the following:

Reaction-controlled mechanism:

$$\begin{array}{c}1-{\left(1-\mathrm{x}\right)}^{1/3}={\mathrm{k}}_1\mathrm{t}\end{array}$$

(7)

Diffusion-controlled kinetics in three dimensions at the interfacial boundary:

$$\begin{array}{c}1-{\displaystyle \frac{2}{3}}\mathrm{x}-{\left(1-\mathrm{x}\right)}^{2/3}={\mathrm{k}}_2\mathrm{t}\end{array}$$

(8)

A novel kinetic model proposed by Dickinson et al. [33] characterizes the leaching process as being controlled by both interfacial mass transfer and transmembrane diffusion. The associated rate expression is presented as follows:

$$\begin{array}{c}{\displaystyle \frac{1}{3}}\ln \left(1-\mathrm{x}\right)-1+{\left(1-\mathrm{x}\right)}^{-1/3}={\mathrm{k}}_3\mathrm{t}\end{array}$$

(9)

where k_i denotes the apparent rate constant for each controlling step, t represents the leaching duration, and x signifies the REE extraction fraction.

Based on optimal roasting parameters, the leaching characteristics of roasted samples were investigated at various temperatures and durations. A kinetic model was developed to represent the reaction-controlled leaching mechanism. The rare earth recovery efficiency served as an indicator of the reaction progression, with Figure 9 illustrating the correlation between the leaching time and leaching yield under varied temperatures.

Table 5. Correlation coefficient (R²) values for the three kinetic models across varying leaching conditions.

Parameters	Values	Chemical Reaction Control		Three-Dimensional Diffusion Control		New Kinetic Model
		1 − (1 − x)1/3 = k1t		1 − 2x/3 − (1 − x)2/3 = k2t		1/3ln(1 − x) − 1+(1 − x)−1/3 = k3t
		K1	R2	K2	R2	K3	R2
T/℃	20	0.00828	0.83915	0.00334	0.96147	0.01029	0.90232
	30	0.00689	0.80181	0.00263	0.94651	0.00533	0.93820
	40	0.00573	0.73771	0.00203	0.91232	0.00327	0.97908
	50	0.00433	0.60014	0.00135	0.78336	0.00170	0.89707

and Figure 10 show that the new kinetic model consistently fits the leaching data best, with R² > 0.89 across all tests. Using Equations (7)–(9), we calculated the temperature-dependent rate constants for the REE leaching. These rate constants (k) were subsequently used to construct an Arrhenius plot (Figure 11) by graphing ln(k) versus 1/T, which yielded a linear relationship through a regression analysis. Based on the slope of the Arrhenius plot in Figure 11, the apparent activation energy for the REE leaching was calculated to be −46.45 kJ/mol. The phenomenon of the negative activation energy can be attributed to the fact that the solubility of rare earth sulfates in aqueous solutions decreases with increasing temperatures. Model calculations and the derived activation energy indicate that REE leaching is governed by a mixed mechanism of diffusion and interfacial mass transfer.

4. Conclusions

The CRS was treated with the H₂SO₄ roasting–water leaching method for the recovery of REEs. From the present investigation, the following conclusions can be drawn:

(1)

The leaching efficiency of REEs is highly dependent on both roasting and leaching parameters. The optimal conditions were determined to be 15 mL of H2SO4, roasting at 150 °C for 240 min, followed by water leaching at 20 °C for 60 min with a liquid–solid ratio of 15:1, achieving a maximum REE extraction efficiency of 94.65%.

(2)

REE in the CRS were completely transformed into water-soluble sulfates during roasting, which were subsequently dissolved during water leaching.

(3)

A CaSO4 product with 98.10% purity was obtained with a calcium recovery of 90.79%. The removal rate of fluorine in the CRS was 99.99%, which makes the comprehensive utilization of the CRS possible.

(4)

The leaching kinetics of REEs were best described by a combined diffusion and interfacial mass transfer model, with an apparent activation energy of –46.45 kJ·mol−1.