Extraction of Rare-Earth Elements from Silicate-Based Ore through Hydrometallurgical Route

Abstract

The European Union and several countries/regions classified the rare-earth elements (REEs), such as lanthanum, cerium, neodymium, and scandium, as critical due to the risk of supply interruption. For this reason, the growing demand for REEs has resulted in forgotten reserves receiving economic interest. So, the search for new sources and the development of chemical process is important, such as silicate-based ore. Since there is almost no literature on the extraction of REEs from this source, a new approach was developed in the present study. Direct leaching and acid baking were studied using sulfuric acid. The effect of the acid concentration, temperature, solid-liquid ratio, oxidizing/reducing medium, and acid dosage were studied. Results showed that the extraction of REEs achieved up to 80% at 90 °C in oxidizing medium, and scandium and iron achieved 13.5% and 65.0%, respectively. For the acid baking experiments, the results were better than direct leaching for REEs at over 85%. The scandium leaching rate was lower than direct leaching. On the other hand, the extraction of iron was lower in acid baking than direct leaching. The iron and scandium extraction rates were higher in lower temperatures (<200 °C) and acid dosages, achieving 50% and 6.3%, respectively. Future studies should explore thermal treatment before acid leaching.

1. Introduction

The rare-earth elements denote a group of 17 elements that includes scandium, yttrium, and lanthanides (elements with atomic numbers 57–71). The term is not used because they are rare, since their concentration in the Earth’s crust ranges from 150–220 ppm (copper is 55 ppm and zinc is 70 ppm, for instance), but due to their chemical similarity. Despite this, these elements are rarely concentrated in minable ore deposits, making the extraction of rare-earth elements an economical and technical challenge [1,2].

The market is controlled by China. As a result, the European Union classified the rare-earth elements as critical materials and the risk of supply interruption in the middle term [3]. For this reason, the search for new sources of rare-earth elements is necessary to serve the growing market. For this reason, several studies have focused on sources of new rare-earth elements to supply the market [4,5,6,7].

On the other hand, extraction from primary resources (e.g., ores) is still necessary [8,9,10,11,12]. The main minerals of rare-earth elements in the world are monazite (phosphate-(Ce,La,Nd,Th)PO₄), xenotime (phosphate-YPO₄), and bastnasite (carbonate-(Ce,La,Y)CO₃F) [13]. China is the largest producer (140,000 tons), followed by the USA (38,000 tons), Burma (30,000 tons), and Australia (17,000 tons). Moreover, the reserves are concentrated in China (44 Mt), Vietnam (22 Mt), Brazil (21 Mt), and Russia (12 Mt). In 2015, global production was 130,000 tons; in 2021, it is expected to reach up to 240,000 tons [14].

Among the extractive routes currently adopted, the hydrometallurgical route has the advantage of producing high-purity products even at low concentrations, as in the case of rare-earth elements. The processing route can vary according to the rare-earth mineral. Monazite extraction, for instance, a phosphate mineral, can be achieved by sulfuric acid digestion (200–220 °C) and further purification steps, such as precipitation and solvent extraction; alkali treatment at 140–150 °C with sodium hydroxide 70% is also used. Bastnasite ore (REECO₃F, where REE = rare-earth element) can be processed by hydrochloric acid 10%, sulfuric acid concentrated at 480 °C, or calcination before leaching. Further steps also include precipitation and solvent extraction [1].

In all the cases above mentioned, different rare-earth elements can be obtained. In the case of scandium, the most valuable among them is rarely found, and its extraction is commonly limited to uranium, thorium, niobium, and rare-earth element production as a by-product. Different extractive approaches can be used, such as alkali fusion, acid leaching under pressure, and chlorination (hydrochloric acid in high temperatures) [2].

A hydrometallurgical route has been explored for the extraction of rare-earth elements from bauxite residue and nickel laterite waste since these elements’ concentration is up to 100 mg/kg or lower [15,16,17,18,19]. Borra et al. (2015) studied the direct leaching of bauxite residue to extract rare-earth elements, achieving up to an 80% extraction efficiency [15]. As reported in the literature, these materials are rich in rare-earth elements [20].

There are other hydrometallurgical techniques for the extraction of rare-earth elements. For instance, Chaikin et al. (2020) used water leaching and alkali leaching from an intermediate product of bauxite sintered in a Russian alumina refinery (electrostatic precipitator dust). According to the authors, water leaching was necessary to remove sodium and aluminum as silicate and then increase the content of rare-earth elements. Alkali leaching occurred at 240 °C and sodium aluminate solution was obtained (also known as Bayer liquor), and the rare-earth elements were concentrated in the solid phase [21].

Moreover, acid baking has been explored for the extraction of rare-earth elements due to the low acid consumption and faster kinetics. Anawati and Azimi (2019) studied the acid baking of bauxite residue to extract rare-earth elements. Concentrated sulfuric acid was mixed with the material and heated at 200–400 °C. Then, the resulting material was leached with water for the extraction of rare-earth elements, where equilibrium was achieved within 2 h at 90 °C [22]. Kim and Azimi (2020) studied the extraction of slag rich in rare-earth elements using acid baking, and the authors also stated that water leaching achieved a plateau after 120 min in all cases, and the kinetic rate was faster at 90 °C than lower temperatures [23]. Zou et al. (2021) demonstrated that lanthanum and cerium extraction also achieved a plateau after 120 min from rare-earth polishing powder wastes after the baking process [24]. Similar results were obtained by Pan et al. (2021) regarding the extraction of rare-earth elements from coal fly ash. First, alkali roasting was carried out to convert the phases to be water soluble. Then, water leaching was responsible for the dissolution of sodium silicate compounds before acid leaching. The extraction efficiency of rare-earth elements was 89% [25].

Due to the increasing interest in such elements and the control of fewer countries of their extraction, the search for new resources is necessary. Among the potential resources of rare-earth elements are silicates and oxides. In the case of silicate-based ores, there is a lack of literature due to scarcity. However, due to the growing demand for rare-earth elements, these reserves have been considered even with lower rare-earth elements than the recognized minerals [13].

There are only a few works in the literature about direct leaching or acid baking of rare-earth elements from silicate-based ores, probably because such minerals are less common than phosphates (monazite and xenotime) and carbonates (bastnasite) [13]. However, since these elements’ consumption has grown over the years, it is necessary to search for new potential resources of rare-earth elements.

For this, the present study’s goal was the extraction of rare-earth elements from a silicate-based ore via the hydrometallurgical route. Sulfuric acid was used as a leaching agent. First, direct leaching experiments were carried out in an acid medium for 8 h. The effects of the solid-liquid ratio, concentration, and temperature were studied. The use of sodium dithionite (Na₂S₂O₄) and hydrogen peroxide (H₂O₂) as reducing and oxidizing agents, respectively, was evaluated. Secondly, acid baking was studied by varying the sulfuric acid/ore ratio and temperature. For a comparison with acid baking, dry digestion was carried out at room temperature (25 °C). After the sulfation process, water leaching was carried out at 90 °C for 2 h. The samples were characterized using XRD, SEM/EDS, and ICP-OES.

2. Materials and Methods

2.1. Materials

The silicate-based ore rich in rare-earth elements was supplied from a Canadian site (primary ore). The particle size distribution was measured by sieves 45, 75, 125, 212, 300, 425, 600, 850, 1180, and 1700 µm in size. Quantification of the main elements was carried out in X-ray fluorescence (XRF, PANalytical Epsilon 3 XL, Malvern, UK). Minor elements were determined using alkali fusion by mixing 0.5 g of the ore with 1.5 g of sodium carbonate and 1.5 g of sodium tetraborate decahydrate at 1100 °C for 30 min followed by dissolution in HCl media.

The mineralogical assessment was carried out using the X-ray diffraction technique (XRD, Rigaku MiniFlex 300, Tokyo, Japan). The analyzed sample was pooled and scanned from 3° to 100° (2 θ) with a 4°/min rate and 0.02 step. The powder morphology was studied using scanning electron microscopy with a backscattered electron detector and coupled with energy dispersion (Phenom model ProX, Norcross, GA, USA).

Sulfuric acid (H₂SO₄) (95–98%, Química Moderna), nitric acid (HNO₃) (65%, Neon), H₂O₂ (29%, Synth), and Na₂S₂O₄ (99%, MetaQuímica) were analytical grade. Acid solutions and calibration curves were elaborated with ultra-pure water.

2.2. Experimental Design

The samples were ground to a −0.5 mm particle size using a motorized pestle/mortar mill (Marconi) to ensure sample homogeneity and dried at 60 °C for 24 h. The experimental error was calculated, as previously depicted [26]. The leaching experiment was carried out at 25 °C, the solid-liquid ratio was 1/10, 4 mol/L of H₂SO₄, and a 200 rpm stirring speed for 8 h. The calculated standard deviation was 3.2%.

2.2.1. Direct Leaching

The experiments used to evaluate the effect of the acid concentration, solid-liquid ratio, reducing and oxidizing agent, and temperature were carried out in sealed Erlenmeyer flasks (250 mL) under stirring (200 rpm) and temperature control (25 °C). To study the effect of temperature, the experiments were performed in glass reactors (200 mL) fitted with a reflux condenser and placed on a hot plate with a magnetic stirring system.

The parameters explored in the present study are depicted in

Table 1. Parameters studied in the direct leaching experiments for the extraction of rare-earth elements from silicate-based ore.

Variables	Conditions
Solid-liquid ratio (S/L)	1/5–1/50
Concentration	0.5–4.0 mol/L
Na2S2O4	1–10%
H2O2	1–10%

. The effect of the solid-liquid ratio was studied for 1/5, 1/10, 1/25, and 1/50, and further, the H₂SO₄ concentration was evaluated for the values of 0.5, 1.0, 2.0, and 4.0 mol/L. The oxidizing and reducing leaching were studied using H₂O₂ and Na₂S₂O₄, considering the concentrations of 1%, 2.5%, 5%, and 10% (${\mathrm{v}}_{{\mathrm{H}}_2{\mathrm{O}}_2}$/${\mathrm{v}}_{\mathrm{acid} \mathrm{solution}}$ and ${\mathrm{v}}_{{\mathrm{Na}}_2{\mathrm{S}}_2{\mathrm{O}}_4}$/${\mathrm{v}}_{\mathrm{acid} \mathrm{solution}}$, respectively). The effect of temperature was studied at 25, 45, 60, and 90 °C.

After the leaching procedure, the mixture was first centrifuged at 3000 rpm for 10 min and then filtered with a 2 µm quantitative filter paper. The leaching residue was washed using ultra-pure water and dried at 60 °C for 24 h for the XRD analyses. The leach solution and the washing water were diluted in HNO₃ 4% for chemical analyses in an inductively coupled plasma optical emission spectrometer (ICP-OES–Agilent Technologies 70 series). The iron concentration was analyzed in AAS (Shimadzu AA-7000).

2.2.2. Dry Digestion and Acid Baking

Crushed samples were mixed with 0.6–1.5 mL of H₂SO₄ 98% in a porcelain crucible with a glass rod to homogenize the mixture. The temperatures studied were 25 (dry digestion), 200, 300, and 400 °C. Further, water leaching was studied at 90 °C for 2 h and the solid/liquid ratio was 1/10.

3. Results and Discussion

3.1. Characterization and Economic Analysis of Silicate-Based Ore

The chemical composition of the ore was analyzed to provide a basis for the extraction efficiency calculations. The major elements in the form of oxides are shown in

Table 2. Major chemical components in the ore.

Compounds	wt%
SiO2	39.4
Fe2O3	36.3
CaO	8.79
Al2O3	4.61
TiO2	2.32
Na2O	1.64
MgO	1.44
K2O	1.42
MnO	1.09
P2O5	0.43

. Silicon is the main element in the ore, followed by iron, calcium, aluminum, and titanium. Sodium, magnesium, potassium, manganese, and phosphorous oxides wereconcentrated above 0.4%.

The concentrations of zirconium and rare-earth elements and the economic importance of each element in the silicate-based ore are presented in

Table 3. Zirconium and rare-earth elements composition of the ore sample and their economic value in the silicate-based ore.

Elements	Concentration (mg/kg)	Economic Value (in US$)
Zr	8090	6.0%
Ce	5420	0.8%
La	2330	0.3%
Nd	2140	7.3%
Y	1100	0.2%
Pr	626	3.2%
Th	375	1.7%
Sm	358	0.1%
Gd	268	0.4%
Dy	238	2.8%
Sc	191	76.8%

. From the characterization, it is evident that the ore is rich in zirconium and both heavy and light rare-earth elements. Additionally, it is rich in scandium (191 mg/kg). Among the minor concentration elements, lanthanum, cerium, neodymium, and yttrium are the main rare-earth elements where the total is 1.1 wt%. On the other hand, the most valuable element is scandium (76.8%), followed by neodymium and zirconium.

The cumulative particle size distribution is shown in Figure 1. In the silicate-based ore, 90% of the particles are smaller than 1412.9 µm (1.41 mm) and 50% are smaller than 348.5 µm (0.35 mm). Since the particle size distribution of the sample is considered high, and since its effect in acid leaching was not explored in the present study, the ore sample was ground to −0.5 mm.

Scanning electron microscopy (SEM) was used to study the sample’s surface morphology, as presented in Figure 2. The EDS spectrum shows the distribution of the chemical elements. As shown, the sample matrix is mainly composed of silicon and oxygen, indicating the presence of silicates as the main compounds. Calcium and iron were also identified. Among the minor elements presented in Table 3, zirconium, lanthanum, cerium, neodymium, and yttrium were identified. These results are in line with the chemical characterization.

Figure 3 shows the XRD of the sample. The main phases identified were dickite (Al₂Si₂O₅(OH)₄), ferrohornblende ((Na,K)Ca₂(Fe,Mg)₅(Al,Si)₈O₂₂(OH)₂), fayalite (Fe₂(SiO₄)), hedenbergite (CaFeSi₂O₆)), and albite ((Na,Ca)Al(Si,Al)₃O₈). As previously observed in the SEM/EDS analyses, the main mineral phases identified were silicates. It was impossible to confirm the rare-earth elements’ mineral phases due to the low concentration in the ore.

The common silicate minerals of the rare-earth elements are: Allanite ((Ce,Ca,Y)₂(Al,Fe²⁺,Fe³⁺)₃(SiO₄)₃(OH)), gadolinite (Y₂Fe²⁺Be₂Si₂O₁₀), zircon ((Zr,REE)SiO₄), thortveitite ((Sc,Y)₂Si₂O₇), cascandite (Ca(Sc,Fe³⁺)HSi₃O₉), gadolinite (Y₂Fe²⁺Be₂Si₂O₁₀), and jervisite ((Na,Ca,Fe²⁺)(Sc,Mg,Fe²⁺)Si₂O₆) [2]. For this reason, it is worth concluding that the rare-earth elements could be in the ore as silicates.

Preliminary magnetic separation was carried out and the mineral phases and their respective percentages are presented in

Table 4. Mineral phases of the magnetic and non-magnetic fraction of the silicate-based ore.

Magnetic Fraction		Non-Magnetic Fraction
Hedenbergite	14.28%	Phlogopite	9.14%
Hilairite	1.49%	Hedenbergite	15.20%
Ilmenite	14.91%	Zirconia	4.73%
Zirconia	1.45%	Fayalite	11.81%
Fayalite	10.87%	Diopside	21.20%
Diopside	31.46%	Hastingsite	1.65%
Ferrohornblende	10.94%	Albite	22.02%
Magnetite	14.60%	Ferrohornblende	14.25%

. Magnetic separation was performed at the lab scale, where the magnetic fraction represented 11% of the raw material (silicate-based ore) and the non-magnetic fraction was 89%. The concentrations of the main elements/compounds are depicted in

Table 5. Characterization of the magnetic and non-magnetic fraction of silicate-based ore.

Compounds	Magnetic Fraction	Non-Magnetic Fraction	Elements	Magnetic Fraction	Non-Magnetic Fraction
SiO2	25.6	42.3	Zr	N.D.	8210
Fe total	36.54	23.14	Ce	N.D.	4960
CaO	7.21	9.05	La	N.D.	2210
Al2O3	1.79	4.97	Nd	N.D.	2060
TiO2	7.36	1.48	Y	N.D.	995
			Pr	N.D.	565
			Th	N.D.	369
			Sm	N.D.	339
			Gd	N.D.	247
			Dy	N.D.	226
			Sc	164	214

N.D.—not detected.

. The results demonstrated 80–90% of the rare-earth elements and 90% of scandium, 90% of silicon, and 88% of zirconium, which shows that the main valuable elements are hosted in the silicates and zircon mineral phases. Only 20% of iron is magnetic (mainly as ilmenite) and presented in the magnetic fraction. As a conclusion, the magnetic separation did not concentrate the rare-earth element or remove the impurities, and the leaching experiments were carried out using the raw material.

The extractive process of rare-earth elements from silicate ores can be accomplished using different approaches. In the case of gadolinite, for instance, it can be performed using (a) alkali fusion + water leaching; (b) chlorination; and (c) acid leaching (H₂SO₄, HCl, or HNO₃) and further precipitation as oxalate [1]. The disadvantages of alkali fusion are the high NaOH consumption, difficulty in washing and filtering due to the high viscosity, and the large amount of water used for the recovery of excessive NaOH [2]. Moreover, as in the case of scandium extraction, where the alkali fusion step is used and further acid leaching, the consumption of acid for both the neutralization of the alkali and extraction of the elements can make the process economically unfeasible. This has been observed for scandium extraction from bauxite residue [15,22,27].

For this reason, in the present study, direct leaching and acid baking were explored. The elements analyzed were iron, zirconium, titanium, and the rare-earth elements. All rare-earth elements can be extracted in an acid medium, as proposed by acid leaching. In the case of dry digestion or acid baking using H₂SO₄, the purpose is to convert the rare-earth elements into sulfate salts that are water soluble. The same can occur in direct leaching by H₂SO₄. In the case of HNO₃ and HCl, they have been replaced due to the leaching efficiency, costs, and industrial problems related to corrosion.

Reid et al. (2017) demonstrated that the extraction rate of rare-earth elements using mineral acids from bauxite residue might achieve similar results. The scandium leaching rate was 32%, 45%, and 40% using HNO₃, HCl, and H₂SO₄ 1.5 mol/L at 90 °C for 30 min. Similar results were obtained for neodymium [28]. Zhang et al. (2020) demonstrated that H₂SO₄ could be used for scandium extraction silicate ore [29]. Nevertheless, there is a lack of literature about H₂SO₄ leaching of rare-earth elements from silicate-based ore.

3.2. Direct Leaching

3.2.1. Effect of the Solid-Liquid Ratio and Acid Concentration

The experiments were performed in Erlenmeyer flasks at 25 °C for 8 h using a stirring speed of 200 rpm. The experiments were carried out with 100 mL of H₂SO₄ 4 mol/L, varying the mass of the ore. The leaching rates of iron, zirconium, lanthanum, cerium, neodymium, and scandium are shown in Figure 4 The titanium content was lower than the detection limit of the equipment. The redox potential of the leaching liquor was measured at 630 mV.

The leaching rate of zirconium was lower than 0.5%; it was 0.4% for a solid-liquid ratio of 1/5 and slightly increased to 0.5% for 1/50. For iron, the leaching rate increased from 33.2% to 53.3% as the solid-liquid ratio increased from 1/5 to 1/50. This occurred as the increase in the solid-liquid ratio increases the amount of acid for each part of the ore’s leaching reaction.

In the rare-earth elements, there was virtually no difference in the increase in the solid-liquid ratio. In the case of lanthanum, cerium, and yttrium, for instance, the leaching rate increased from 61.0%, 57.7%, and 47.7% (1/5) to 64.9%, 61.0%, and 52.0% (1/50), respectively. Scandium extraction was up to 3%.

Figure 5 shows the effect of the acid concentration on the leaching of silicate-based ore. The iron extraction increased from 19.7% (0.5 mol/L) to 24.8% (1.0 mol/L) and then remained constant as the concentration of H+ increased. This might have occurred due to the reaction of iron oxides present in the ore, which are more easily leached than other iron compounds [30] than ilmenite [31], for instance. In this case, hydrochloric acid with sodium bifluoride and metallic iron is used as a reducing agent for titanium extraction [32].

The leaching of scandium slightly increased as the acid concentration increased to 2.4%. Zirconium leaching remained the same (up to 0.45%). The extraction of lanthanum, cerium, neodymium, and yttrium decreased as the H₂SO₄ concentration increased. The concept of salvation can explain the decrease in the leaching efficiency. As the concentration of electrolyte increases, the proportion of water/electrolyte decreases, resulting in fewer water molecules for the leaching reaction because cations and anions hold them tightly in the solution [28,33].

As shown in Equations (1) and (2), the leaching of rare-earth elements by H₂SO₄ generates their ions in sulfate media. As the concentration of sulfate anions increases, the rare-earth elements precipitate as sulfate salts [34]. This may explain why the leaching of rare-earth elements declined:

$$RE{E}_2{O}_3+H_{2}S{O_4}_{\left(aq\right)} \to RE{E}^{+3}+S{O}_4{}^{-2}+H_{2}O$$

(1)

$$RE{E}^{+3}+S{O}_4{}^{-2} \to RE{E}_2{\left(S{O}_4\right)}_3$$

(2)

3.2.2. Oxidizing and Reducing Leaching Experiments

H₂O₂ and Na₂S₂O₄ were used as oxidants and reducing agents, respectively, in the leaching process. In the case of H₂O₂, it was studied due to the synthesis of silica gel during silicates’ leaching. Alkan et al. (2018) studied the oxidizing leaching of bauxite residue to avoid silica gel formation. As stated by the authors, the compound is formed during the acid leaching of silicates, and its synthesis decreases the leaching rate since a part of the liquor generated in the leaching is trapped. By the same token, H⁺ is consumed during the synthesis of silica gel. Additionally, the silica gel causes practical process problems due to the difficulty in the solid-liquid separation step [35].

In the case of Na₂S₂O₄ being used as a reducing agent, it was used to reduce the redox potential during the reaction. Figure 6 shows the Pourbaix diagram of the Fe-S-H₂O system, where the pH of the leach solution and the initial redox potential and the decrease as the Na₂S₂O₄ is used in the reaction are highlighted. According to Luo et al. (2015), and as demonstrated in the Pourbaix diagram, in reducing medium, the leaching of iron could be beneficial and the target metals present in these compounds will be released [30].

Since the same rare-earth elements can be found in iron compounds, such as scandium [22], the conversion into ferrous iron and thereby the release of the elements by a reducing agent to lower the potential of the leaching reaction can be highly beneficial. Luo et al. (2015) stated that the use of a reducing agent in the leaching of nickel laterite increased the leaching of iron and nickel from 40% to up to 60% [30].

Figure 7 shows the leaching rate of iron, titanium, zirconium, and rare-earth elements after the addition of hydrogen peroxide and sodium dithionite into the reaction. The extraction efficiency of oxidizing leaching increased as H₂O₂ was added into the system to reach 2.5% in all cases and then slightly decreased. The exception was iron, which declined from 29.3% (2.5 v/v% of H₂O₂) to 9.3% (10.0 v/v% of H₂O₂). As depicted in Figure 6, for H₂SO₄ 2.0 mol/L without oxidant agent, iron can be present in the solution as ferrous iron and as iron sulfate in the solid phase, where it occurs after the leaching reaction. However, in the oxidizing leaching reaction, where the redox potential increased to 1000 mV, iron oxide is not fully leached, and only a part generates iron sulfate ions.

The leaching of lanthanum, cerium, and neodymium increased from up to 65% (without hydrogen peroxide) to 85% (5.0 v/v% of hydrogen peroxide). In the case of scandium, its efficient rate increased from 2.6% (without hydrogen peroxide) to 4% (5.0 v/v% of hydrogen peroxide). The leaching of titanium increased in the oxidizing leaching. The H₂SO₄–H₂O₂ system contributes to the formation of titanium peroxo sulfate, which is soluble during leaching, as shown in Equations (3) and (4) [35]:

$$Ti{O}_2+H_{2}S{O}_4 \to \left[TiO\right]S{O}_4$$

(3)

$$\left[TiO\right]S{O}_4+H_2{O}_2 \to \left[TiO-O\right]S{O}_4+H_{2}O$$

(4)

The leaching rate in reducing medium is presented in Figure 7b. As observed, the H₂SO₄–Na₂S₂O₄ system increased the leaching of rare-earth elements from around 60% (without Na₂S₂O₄) to 80% (1.0 wt% of Na₂S₂O₄) and 70% (2.5 wt% of Na₂S₂O₄). The same behavior was observed for the other elements. A decrease in the leaching efficiency above 2.5 wt% of Na₂S₂O₄ was observed. This is not because of the reducing redox potential but is due to the acid consumption by Na₂S₂O₄. As previously reported, in an acid medium, Na₂S₂O₄ reacts with H⁺ and releases H₂S [36,37,38,39]. For this reason, as more reduction was added into the leaching process, the extraction rate decreased, and consequently, the pH of the solution increased (until 3.0). As a result, the leaching efficiency achieved lower values for 10 wt% of Na₂S₂O₄ than 2.5 wt%.

Comparing the oxidizing and reducing leaching reaction, the H₂SO₄–H₂O₂ system is more beneficial to the leaching of rare-earth elements than the H₂SO₄–Na₂S₂O₄ system, where the extraction of lanthanum, cerium, and neodymium was up to 80%. The concentration of yttrium and scandium in the solution increased by more than 17% when 1 wt% hydrogen peroxide was added into the reaction. For 1 wt% of Na₂S₂O₄, the yttrium content increased by 12%, and no difference was observed for scandium.

3.2.3. Thermodynamic Experiments

The effect of temperature on the oxidizing leaching of silicate-based ore was studied from 25 to 90 °C. The results are presented in Figure 8, which shows that the increase in the temperature benefited the rare-earth elements’ leaching. Lanthanum, cerium, neodymium, and yttrium extraction slightly increased as a function of the temperature. In iron, titanium, zirconium, and scandium, the extraction was even more accentuated: 65.0%, 31.1.%, 3.1%, and 13.5% at 90 °C, 4.0 mol/L of H₂SO₄, and 1 v/v% of hydrogen peroxide, respectively.

As observed in the experiments carried out with 2.0 mol/L of H₂SO₄ and 1 v/v% of hydrogen peroxide (Figure 8b), the extraction of lanthanum, cerium, and neodymium slightly decreased as the temperature increased from 25 to 90 °C. This might have occurred due to the decomposition of hydrogen peroxide as the temperature increased. At 25 °C, the oxidant agent acts with the acid in the leaching reaction; however, hydrogen peroxide decomposes faster in increasing temperatures, which is even more favorable in an acid medium [35].

On the other hand, the leaching of iron, zirconium, titanium, and scandium increased as the temperature increased. This may indicate that scandium, the most valuable rare-earth element in silicate-based ore, is strongly related to its minerals. Due to its low concentration, scandium was not detected in the SEM/EDS and XRD analyses. Moreover, it is already known that it is spread out in different minerals and is not concentrated in a few compounds. Scandium occurs in trace concentrations in up to 800 minerals because it also substitutes the major elements, such as iron and aluminum, in the ores [2,17]. Its extraction from primary resources is difficult.

Figure 9 shows the correlation between the scandium leaching and iron, zirconium, and rare-earth elements. A low linear correlation r² < 0.9 was observed for zirconium and cerium. This shows that scandium is spread out in different minerals in the silicate-based ore. The linear correlation (r²) between lanthanum and cerium was 0.98 and up to 0.9 between iron and rare-earth elements. This indicates that the rare-earth elements might be in the same minerals containing iron. As the iron leaching increased from 30% (25 °C) to 65% (90 °C) and the rare-earth elements increased from up to 60% (25 °C) to 80% (90 °C), not all iron compounds contain rare-earth elements.

The effect of temperature can be expressed by the Arrhenius equation as depicted in Equations (5) and (6), where K is the reaction rate, A is the frequency factor, Ea is the activation energy, R is the gas constant (8.3145 J/(mol.K)), and T is the temperature [29,40]. The results are shown in

Table 6. Arrhenius data for the leaching of iron, zirconium, lanthanum, cerium, neodymium, yttrium, and scandium, varying the temperature. Experimental conditions: 100 mL of H₂SO₄ 4.0 mol/L; 1 v/v% H₂O₂; solid-liquid ratio = 1/10; t = 8 h; stirring speed = 200 rpm.

	Fe	Ti	Zr	La	Ce	Nd	Y	Sc
Activation energy (J/mol)	11,217.87	39,404.73	31,617.41	4661.09	5387.19	6830.66	4541.61	25,448.91
Frequency factor (×10−2)	3.38	5.90 × 10−4	9.73 × 10−2	26.19	19.83	12.73	29.48	0.14
Linear correlation (r2)	0.8911	0.9639	0.8492	0.788	0.7051	0.8401	0.8663	0.9098

. The activation energy for iron, titanium, zirconium, and scandium was 11.3, 39.4, 31.6, and 25.4 kJ/mol, respectively. For lanthanum, cerium, neodymium, and yttrium, the activation energy was 4.7, 5.4, 6.8, and 4.5 kJ/mol, respectively:

$$K=A{e}^{\frac{-Ea}{RT}}$$

(5)

$$\mathrm{Ln}K=\ln A-\frac{Ea}{R.T}$$

(6)

3.3. Dry Digestion and Acid Baking Experiments

Although direct leaching can achieve the extraction rate for rare-earth elements, different leaching methods have been studied. Dry digestion and acid baking are a few of the methods studied in the literature for the extraction of rare-earth elements. One of the reasons is the recovery of H₂SO₄ in the process. Rivera et al. (2018) reported that dry digestion has more economic benefits than direct leaching for the extraction of rare-earth elements [41]. In acid baking, where the source is mixed with H₂SO₄ and heated at 200–400 °C, SO₂ released from the pyrometallurgical step can be recovered for H₂SO₄ production and then recycled. Nevertheless, direct leaching consumes less energy than acid baking [22].

Moreover, the main problem in direct leaching of silicate ores by H₂SO₄ is silica gel formation. According to Terry (1983), the action of acids in the silicate results in the following [29,42]:

(a)

Complete destruction of the silicate structure and cation release, generating silica gel;

(b)

Partial dissolution of the silicate structure and cation release, which leads the silica compound in the solid phase; and

(c)

The acid does not react with the silicate.

Anawati and Azimi (2019) studied the acid baking of bauxite residue to extract rare-earth elements. The authors stated that the water leaching achieved equilibrium within 2 h of the process. Moreover, water leaching at 90 °C achieved a faster kinetic reaction than lower temperatures [22]. Kim and Azimi (2020) studied the extraction of slag rich in rare-earth elements using acid baking, and the authors also stated that the water leaching achieved a plateau after 120 min in all cases, and the kinetic rate was faster at 90 °C than lower temperatures [23].

Zou et al. (2021) demonstrated that lanthanum and cerium extraction also achieved a plateau after 120 min in leaching after the baking process from rare-earth polishing powder wastes [24]. As depicted by Demol et al. (2019), the reaction time of water leaching after acid baking of rare-earth concentrates varied between 0.08 and 3 h [13]. Because of this, in the present study, a leaching time of 2 h was adopted.

3.3.1. Effect of Acid Dosage

The experiments were performed at 400 °C, and the acid dosages studied were: 0.6, 1.0, 1.3, and 1.5 mL for each gram of ore. The mixture was homogenized at room temperature and then heated to the desired temperature for 2 h. The water leaching was carried out using ultra-pure water in a solid-liquid ratio of 1/10 at 90 °C for 2 h.

The effect of the acid dosage is shown in Figure 10. The extraction of cerium was 92% for 0.6 mL of acid dosage for each gram of sample. This is higher than the direct leaching experiments, where up to an 80% extraction was achieved. Similar results were observed for lanthanum and neodymium.

Yttrium extraction was higher for indirect leaching than acid baking at 400 °C. The same was observed for iron and scandium. During heating, iron compounds are converted to iron sulfate, which is soluble in water. Equations (7) and (8) show the reaction of iron with H₂SO₄ concentrated at 150–250 °C, generating iron sulfates. In temperatures up to 400 °C (Equations (9) and (10)), iron oxide (hematite) is formed, releasing sulfur oxide. Basic iron sulfate can also form. In both cases, the iron compound is insoluble [13].

According to the equations, the increase in the H₂SO₄ dosage increased the generation of insoluble iron compounds. As these compounds bear scandium and its release during water leaching is directly related to iron extraction, the leaching of scandium decreased. Moreover, it is reported in the literature that at high temperatures, the rare-earth element compounds formed during acid baking are insoluble in the water leaching [13].

Moreover, after the water leaching, the leach solution’s pH was up to 1.5, which indicates that most of the H₂SO₄ was released as gas in the baking step. As shown in Equations (7)–(10), the same might occur in the leaching of rare-earth elements: first, sulfate compounds formed; and due to the temperature, decomposition occurred, forming oxides.

Figure 11 shows the X-ray diffractogram of the samples after acid baking. Iron sulfate was identified in the samples after acid baking. However, hematite was also identified. As depicted, the mineral-phase fayalite was decomposed; on the other hand, other silicate phases did not fully react with the concentrated H₂SO₄:

$$F{e}_2{O_3}_{\left(s\right)}+3H_{2}S{O_4}_{\left(l\right)} \to F{e}_2{{\left(S{O}_4\right)}_3}_{\left(s\right)}+3H_2{O}_{\left(g\right)}$$

(7)

$$F{e}_3{O_4}_{\left(s\right)}\left(s\right)+4H_{2}S{O_4}_{\left(l\right)} \to F{e}_2{{\left(S{O}_4\right)}_3}_{\left(s\right)}+FeS{O_4}_{\left(s\right)}\left(s\right)+4H_2{O}_{\left(g\right)}$$

(8)

$$F{e}_2{{\left(S{O}_4\right)}_3}_{\left(s\right)} \rightleftarrows F{e}_2{O_3}_{\left(s\right)}+3S{O_3}_{\left(g\right)}$$

(9)

$$F{e}_2{{\left(S{O}_4\right)}_3}_{\left(s\right)} \to F{e}_{2}O{\left(S{O}_4\right)}_2+S{O_3}_{\left(g\right)}$$

(10)

3.3.2. Effect of Temperature

The baking process temperature was evaluated between 200 and 400 °C using a 1 g of ore / 0.6 mL of H₂SO₄. Dry digestion (25 °C) was also tested for the same proportion. Ultra-pure water was used for leaching at 90 °C for 2 h. The results are shown in Figure 12. The pH of the liquor of the water leaching increased from 0.2 (25 °C) to 1.5 (300 and 400 °C). Equation (11) depicts the decomposition reaction of H₂SO₄ in the baking process. As the temperature increases, there will be less acid to react with the ore, and sulfur oxide and water will be released as gases [13]:

$$H_{2}S{O_4}_{\left(l\right)} \to S{O_3}_{\left(g\right)}+H_2{O}_{\left(g\right)}$$

(11)

For cerium extraction, the highest extraction occurred at 400 °C. Indeed, acid baking at 400 °C released more sulfur oxides than at lower temperatures, which explains the low leaching rates.

There was almost no effect on neodymium and yttrium extraction as the temperature of the acid baking increased. At 200 °C, lanthanum achieved the highest extraction rate (87.0%), and zirconium (4.9%)

Compared to direct leaching, acid baking increased the extraction of rare-earth elements. For iron extraction, the extraction rate was higher at lower temperatures (25 and 200 °C). In the case of scandium, the highest extraction rate was obtained in dry digestion and acid baking at 200 °C (6.3% and 5.7%, respectively), which is lower than the rate obtained in direct leaching at 90 °C (13.5%).

Varying the acid dosage (0.6 and 1.5 mL) for the baking process at 200 °C, as presented in Figure 12b, the extraction of rare-earth elements slightly increased. According to the results presented here, it is concluded that dry digestion or acid baking could partially destroy the silicate structures.

As observed, the extraction rate of lanthanum, cerium, and neodymium was higher than direct leaching. On the other hand, the scandium, titanium, zirconium, and yttrium rates were lower. For iron, no difference was observed between direct leaching and dry digestion or acid baking.

4. Conclusions

The goal of the present study was to extract rare-earth elements from silicate-based ore. Due to the increasing interest in such elements and the control of fewer countries of their extraction, the search for new resources is necessary. There is a lack of studies in the literature on the extraction of rare-earth elements from silicates ores. In this study, direct leaching, dry digestion, and acid baking were studied. H₂SO₄ was used as a leaching agent. The ore was mainly composed of silicon (SiO₂ = 39.4%), iron (Fe₂O₃ = 36.3%), calcium (CaO = 8.8%), aluminum (Al₂O₃ = 4.6%), and titanium (TiO₂ = 2.3%). The zirconium content was 8090 mg/kg. Among the minor elements, the REO content was 1.3%, mainly comprising lanthanum, cerium, neodymium, yttrium, and scandium. The d50 and d90 were 1412.9 and 348.5 µm, respectively. The main mineral phases were dickite, ferrohornblende, fayalite, hedenbergite, and albite. The results of the direct leaching showed that the extraction of rare-earth elements was up to 80% at 90 °C and 1 v/v% of H₂O₂. Scandium, iron, titanium, and zirconium extraction was 13.5%, 65.0%, 31.2%, and 3.1%, respectively. The solid-liquid ratio had almost no effect on the extraction of rare-earth elements. As the amount of acid solution increased, the extraction of iron also increased. The same was observed when the H₂SO₄ concentration was varied. The effect of oxidization and reduction on the leaching process was tested, and the use of H₂O₂ was beneficial to the extraction of rare-earth elements. The leaching reaction was endothermic, and the activation energy for lanthanum, cerium, neodymium, yttrium, and scandium was 4.7, 5.4, 6.8, 4.5, and 25 kJ/mol, respectively. The extraction of rare-earth elements using acid baking was up to 80%. The acid dosage and temperature had almost no effect. The scandium leaching rate was lower than direct leaching. The iron and scandium extraction rates were higher at low temperatures (25 and 200 °C). Further studies may explore thermal treatment to release the rare-earth elements from silicate mineral phases before acid extraction.