Extraction of alumina and silica from high-silica bauxite by sintering with sodium carbonate followed by two-step leaching with water and sulfuric acid

Abstract

Efficient utilization of high-silica bauxite and minimization of bauxite residue are of great significance for the sustainable development of the alumina industry. In this paper, a novel process is proposed to extract Al₂O₃ and SiO₂ from high-silica bauxite without residue discharge, that is, sintering bauxite with Na₂CO₃ followed by two-step leaching with water and sulfuric acid. The effects of the sintering parameters on the process were investigated, and the phase transformations during sintering and leaching were revealed by using phase diagram, thermogravimetric analysis (TGA) and differential scanning calorimetric (DSC), X-ray diffraction (XRD), scanning electron microscopy (SEM) and energy dispersive spectrometer (EDS) methods. When the mixture of the high-silica bauxite and Na₂CO₃ with mole ratio of Na₂O/(Al₂O₃ + SiO₂) of 1 was sintered at 950 °C for 30 min, diaspore and kaolinite were primarily converted into Na_1.95Al_1.95Si_0.05O₄ and an amorphous phase, respectively. In the water leaching process, Na_1.95Al_1.95Si_0.05O₄ was dissolved while the amorphous phase underwent some transformations to form the water leaching residue, resulting in ∼84% of Al₂O₃ being extracted for alumina production. In the sulfuric acid leaching process, the amorphous phase in the water leaching residue dissolved, resulting in ∼13% of Al₂O₃ and ∼86% of SiO₂ being extracted for the production of polyaluminium ferric sulfate (PAFS) and silica gel, respectively. The silica gel had a high purity, containing more than 88% of SiO₂ after drying.

---

1 Introduction

Silica is a major impurity in bauxite ores, which is commonly present as quartz (SiO₂) and phyllosilicate minerals such as kaolinite (Al₄[Si₄O₁₀](OH)₈), illite (K_xAl₄Si_8−xAl_xO₂₀(OH)₂) and pyrophyllite (Al₂[Si₄O₁₀](OH)₂).^1,2 When producing alumina from bauxite using the Bayer process, the reactive silica is transformed into insoluble sodium aluminosilicate hydrate known as desilication product (DSP), resulting in a severe loss of alumina and caustic soda.^3–5 As a result, high-silica bauxite, in which reactive silica is above ∼8%, cannot be economically processed by the conventional Bayer process.² With the increase of alumina production, high-grade bauxite is rapidly consumed, leading to an increase in the proportion of high-silica bauxite. Efficient development and utilization of high-silica bauxite has become an urgent and significant challenge for the alumina industry.

Many methods have been developed to extract alumina from high-silica bauxite, which can be broadly classified into four main categories, as shown in Table 1. Among these methods, most are at the experimental stage and only a few have been applied industrially. The flotation-Bayer process has been widely applied in the alumina refineries, but it introduces large amounts of hazardous organic reagents and produces tailings totaling of about 25% of the initial bauxite ores in weight.^25,26 The lime Bayer process has been also employed to process a portion of high-silica bauxite, but it cannot economically process the bauxite with a mass ratio of alumina to silica (A/S) lower than 5.⁷ For processing high-silica bauxites with A/S < 5, the lime-soda sintering process and its various combinations with the Bayer process are the preferred choice.²⁷ However, they are barely economically viable due to the high operating costs. In addition, the extraction of silica is rarely considered by the methods listed in Table 1. Only in the chemical beneficiation of high-silica bauxite, silica is extracted into alkali solutions and is used to produce silicate products such as 4A zeolite.^28,29 In other methods, silica is left in the bauxite residue.

Table 1 Classification of methods for processing high-silica bauxite

Category	Description	Limitation	Method	Development stage
Pre-desilication of high-silica bauxite followed by the Bayer process	Improving the A/S of high-silica bauxite by physical, chemical and biological beneficiation methods to meet the requirement of the Bayer process	Physical beneficiation discharges a lot of tailings, wasting serious resources. Chemical beneficiation is difficult in recycling desilication solutions and high in production costs. Biological beneficiation has a long processing period and an unstable desilication efficiency	Flotation^6	Industrial application
			Alkali leaching^7	Experimental stage
			Roasting-alkali leaching^8	Experimental stage
			Reduction roasting-alkali leaching^9	Experimental stage
			Biological desilication^10	Experimental stage
The Bayer process variations	Converting the reactive silica into hydrogarnet or calcium silicate by adding lime and adjusting digestion conditions, resulting in a lower loss of caustic soda	Adding large amount of lime increases the burden of solid–liquid separation and the sodium carbonate content in sodium aluminate solution. Increasing alkali concentration leads to a difficult solid–liquid separation and a greater evaporation energy consumption	The lime Bayer process^11	Industrial application
			The calcification–carbonization process^12	Experimental stage
			The hydro-chemical process^13	Experimental stage
			The sub-molten salt process^14	Experimental stage
Sintering processes	Converting silica into insoluble dicalcium silicate during sintering and separating it with soluble sodium (or calcium) aluminate during leaching	Sintering processes are high in energy consumption and poor in economic efficiency	The lime-soda sintering process^15	Industrial application
			The lime-soda sintering process variations^16,17	Experimental stage
			The lime sintering process^18	Stop production after application
Acid methods	Processing bauxite with acid or ammonium salt to get aluminum salt solution and leaving silica in the residue	Acid has a serious corrosion on equipment, while ammonium salt has a potential ammonia leak. In addition, impurities in aluminum salt solutions are difficult and costly to remove	Ammonium sulfate roasting-water leaching^19,20	Experimental stage
			Acid leaching^21–23	Experimental stage
			Thermal or mechanical activation-acid leaching^24	Experimental stage

---

Bauxite residue is a solid waste left over after extracting alumina from bauxite. When high-silica bauxite is used as a feedstock for alumina production, more bauxite residue is discharged compared to processing high-grade bauxite.³⁰ At present, most bauxite residues are sent to dams for storage, taking up large amounts of land and incurring significant costs for the construction and maintenance of the dams.^31,32 In addition, long-time storage of bauxite residue poses a potential threat to the surrounding environment and may cause safety accidents such as leakage and dam failure.³³ Therefore, minimization and recycling of bauxite residue should be considered in the production of alumina. However, the alkali methods listed in Table 1 rarely take this aspect into account. Only the calcification–carbonization process proposed to recycle bauxite residue for preparing cement.³⁴ Compared with the alkali methods, the acid methods produce less residue which is highly enriched in silica and can be used as a raw material for inorganic silica products.^19,23

Alkali methods are well established in the production of alumina, while acid methods have advantages in the minimization and recycling of bauxite residue. Therefore, an attempt can be made to combine the alkali and acid methods. In our previous studies, a novel combined alkali-acid process was proposed to extract aluminum, silicon and iron from silver tailings.^35,36 In the present study, the process was adapted to realize the extraction of Al₂O₃ and SiO₂ from high-silica bauxite without residue discharge, as shown in Fig. 1. First, high-silica bauxite is activated by sintering with Na₂CO₃. Second, water leaching is conducted at atmospheric pressure to extract most of Al₂O₃ from the sintered sample into the sodium aluminate solution, which can be used to produce alumina by desilication, carbonation decomposition and calcination. Third, sulfuric acid leaching is conducted at atmospheric pressure to extract SiO₂ and the remaining Al₂O₃ from the water leaching residue into the acid leaching solution, which can be used to produce silica gel and polyaluminium ferric sulfate (PAFS) by aging, separation, concentration and polymerization. This study aims to investigate the effects of sintering temperature (850–1050 °C), sintering time (10–50 min) and the mole ratio of Na₂O/(Al₂O₃ + SiO₂) (0.6–1.4) on the extraction of Al₂O₃ and SiO₂, and to reveal the process mechanism using phase diagram, thermogravimetric analysis (TGA) and differential scanning calorimetric (DSC), X-ray diffraction (XRD), scanning electron microscope (SEM) and energy dispersive spectrometer (EDS) techniques.

Fig. 1 Flowsheet for extraction of alumina and silica from high-silica bauxite.

2 Experimental

2.1 Raw materials

The high-silica bauxite sample used in this study had been described in our previous study, which has a high Al₂O₃ content of 62.92%, a high SiO₂ content of 16.72%, a low A/S of 3.76, and a mineral composition (Fig. 2(a)) dominated by diaspore (63.4%) and kaolinite (31.2%).³⁷ The particle size distribution (Fig. 2(b)) of the high-silica bauxite indicates that 80% of the particles are distributed in the range of 3.15–21.74 μm with a D50 value of 9.48 μm. The SEM images (Fig. 2(c) and (d)) and EDS results (Table 2) of the high-silica bauxite indicate that diaspore particles are primarily in massive and slaty structure, while kaolinite particles are primarily in layer structure. The particle distribution and morphological characteristics of the high-silica bauxite favor its sintering reaction with Na₂CO₃.

Fig. 2 Characterizations of the high-silica bauxite: (a) XRD pattern; (b) particle size distribution; (c and d) SEM images.

Table 2 Elemental compositions of bauxite particles from EDS analysis (wt%)

Element	O	Al	Si	Ca	Ti	Fe
Spot 1	54.27	25.20	17.86	—	1.56	1.11
Spot 2	51.06	27.13	19.55	—	0.71	1.55
Spot 3	53.36	41.46	3.38	—	0.66	1.24
Spot 4	46.03	43.19	4.86	3.37	1.22	1.33
Average	51.18	34.25	11.41	0.84	1.04	1.31

---

All the chemical reagents used in the study are analytically pure. The concentration of concentrated sulfuric acid is 18 mol L⁻¹, which was diluted with distilled water to the specified concentration.

2.2 Methods and processes

The high-silica bauxite was mixed evenly with various amounts of Na₂CO₃ in a ceramic mortar. The mixtures were transferred to a muffle furnace and sintered for a certain time after the temperature is raised to a specified level. The sintered samples were cooled to room temperature and then ground to below 74 μm in particle size for leaching. The experimental ranges of Na₂CO₃ dosage, sintering temperature and sintering time were determined by phase diagram calculation, simultaneous thermal analysis and previous experience, respectively.

The sintered samples were leached in distilled water at 75 °C for 15 min with a solid/liquid ratio of 1 g/15 mL. Water leaching residues were separated from the slurries, washed five times with hot distilled water, and then dried at 105 °C for 2 h. The dry residues were leached in a 2 mol L⁻¹ sulfuric acid solution at room temperature for 10 min with a solid/liquid ratio of 1 g/10 mL. After acid leaching, no solid–liquid separation was performed. The acid leaching solutions were aged at 85 °C for 1.5 h to precipitate silica gel, which was subsequently separated and washed with hot distilled water until neutral. The silica gels were dried at 110 °C for 12 h to obtained silica powders.

The contents of Al₂O₃ and SiO₂ in solid samples were analyzed by complexometric titration and spectrophotometry (or hydrofluoric acid gravimetry), respectively. The recovery efficiencies of Al₂O₃ and SiO₂ were calculated according to the following equations.

η_A1 = [1 − (c_A2/c_S2)/(c_A1/c_S1)] × 100% (1)

η_A2 = (m₂ × c_A2 − m₃ × c_A3)/(m₁ × c_A1) × 100% (2)

η_AT = η_A1 + η_A2 (3)

η_S = m₃ × c_S3/(m₁ × c_S1) × 100% (4)

where η_A1, η_A2 and η_AT represent the recovery efficiencies of Al₂O₃ in water leaching, acid leaching and total process, respectively, η_S represent the recovery efficiency of SiO₂, m₁, m₂ and m₃ are the masses of sintered sample, water leaching residue and silica powder, respectively (g), c_A1, c_A2 and c_A3 are the contents of Al₂O₃ in sintered sample, water leaching residue and silica powder, respectively (%), and c_S1, c_S2 and c_S3 are the contents of SiO₂ in sintered sample, water leaching residue and silica powder, respectively (%).

2.3 Characterizations of solid samples

Particle size distribution was analyzed by a Bettersize 2000 laser particle size analyzer. Chemical compositions were determined by a Shimadzu 1800 X-ray fluorescence (XRF) spectrometer. X-ray diffraction (XRD) patterns were recorded on a Rigaku D/MAX 2000 X-ray diffractometer and the mineral compositions were identified using the Jade software. Thermo-gravimetric analysis (TGA) and differential scanning calorimetric (DSC) were performed using a Mettler Toledo 1/1600 simultaneous thermal analyzer at a heating rate of 10 °C min⁻¹ in a nitrogen atmosphere. The micromorphology and element distribution were conducted by a scanning electron microscopy equipped with energy disperse X-ray spectrometry (MLA650F, F.E.I.). Specific surface area was analyzed by a Brunauer–Emmett–Teller (BET) surface area analyzer (Nova Station A, Quantachrome).

3 Results and discussion

3.1 Setting of experimental parameters

In the sintering process, Na₂CO₃ dosage, sintering temperature and sintering time are the key parameters affecting the extraction of Al₂O₃ and SiO₂. A theoretical analysis must be performed to rationalize the experimental range of these parameters.

The dosages of sintering additives are usually determined by the contents of the major components in bauxite. In the traditional lime-soda sintering process, the dosages of Na₂CO₃ and CaO are controlled by the molar ratios of Na₂O/Al₂O₃ ≈ 1 and CaO/SiO₂ ≈ 2 to ensure the formations of NaAlO₂ and Ca₂SiO₄, respectively.^16,38 In this study, both Al₂O₃ and SiO₂ would react with Na₂CO₃ due to the absence of CaO. Therefore, the molar ratio of Na₂O/(Al₂O₃ + SiO₂) (N/(A + S)) should be used as the indicator for controlling Na₂CO₃ dosage.

Phase diagram calculations are generally used to model the equilibrium behavior of multi-component systems at various compositions and temperatures.³⁹ The experimental range of N/(A + S) could be determined by analyzing the equilibrium phase changes in the Na₂O–Al₂O₃–SiO₂ system. The phase diagram of the Na₂O–Al₂O₃–SiO₂ system was calculated using the phase diagram module of FactSage software based on the FACT oxide compounds and solutions (FToxid) databases, as shown in Fig. 3. When the mole fraction of Na₂O is less than 0.41, Al₂O₃, SiO₂ or NaAl₉O₁₄ appear as stable phases, indicating that Na₂O is insufficient to fully react with Al₂O₃ and SiO₂. When the mole fraction of Na₂O is above 0.57, Na₂O is present as one of the stable phases, indicating its excess in the system. As the mole fraction of Na₂O is in the range of 0.41–0.57, the stable phases are NaAlO₂ solid solution coupled with Na₂SiO₃, Na₄SiO₄ or a liquid slag. Among them, NaAlO₂ solid solution is the effective phase for alumina extraction. Therefore, it is more appropriate for the mole fraction of Na₂O between 0.41–0.57, which is converted to N/(A + S) of 0.69–1.33. Considering that minor amounts of Fe₂O₃, TiO₂ and other components in bauxite would have a certain influence on Na₂CO₃ dosage, the experimental range of N/(A + S) was extended to 0.6–1.4 in this study.

Fig. 3 Calculated phase diagram of the Na₂O–Al₂O₃–SiO₂ system (A – Al₂O₃, N – Na₂O, S – SiO₂, NA – NaAlO₂, NA₉ – NaAl₉O₁₄, NS – Na₂SiO₃, N₂S – Na₄SiO₄, N₅S – Na₁₀SiO₇, NAS₆ – NaAlSi₃O₈, Mul – mullite, Nph – nepheline, Sl – slag, s – solid, l – liquid, s.s. – solid solution).

In order to determine the experimental range of sintering temperature, the thermal behavior of the mixture of bauxite and Na₂CO₃ with N/(A + S) = 1 during heating was analyzed by TGA/DSC under a nitrogen atmosphere, as shown in Fig. 4(a). There were three endothermic peaks at 89 °C, 508 °C and 849 °C, and one exothermic peak at 1068 °C on the DSC curve. The endothermic peaks at 89 °C and 508 °C were primarily attributed to the removal of adsorbed water and structural water, respectively. Accordingly, significant weight losses were observed around 89 °C and 508 °C on the TG curves. After removing the structural water, diaspore (α-Al₂O₃·H₂O) and kaolinite (Al₂O₃·2SiO₂·2H₂O) were transformed into alpha-alumina (α-Al₂O₃) and metakaolinite (Al₂O₃·2SiO₂), respectively, which can further react with Na₂CO₃ to form NaAlO₂ solid solutions (Na_2−xAl_2−xSi_xO₄, x ≤ 0.55) eventually.³⁷ Therefore, the weight loss in the TG curve above 600 °C was primarily attributed to the release of CO₂ from the reaction between bauxite and Na₂CO₃. The third endothermic peak was located at about 849 °C, where the melting of Na₂CO₃ significantly increased the reaction rate between bauxite and Na₂CO₃,⁴⁰ resulting in a significant increase in the weight loss rate of the TG curve. Thus, the sintering temperature should be set above 849 °C to improve the sintering efficiency and shorten the sintering time. The TG curve was no longer significantly losing weight upon reaching a temperature of about 1050 °C, indicating that Na₂CO₃ was fully involved in the reaction with bauxite. Therefore, the experimental upper limit of the sintering temperature could be set at 1050 °C. In summary, the range of sintering temperature investigated in this study was set as 850–1050 °C. According to Fig. 3, the stable phases at 850–1050 °C were primarily NaAlO₂ solid solution and a liquid slag. The exothermic peak at 1068 °C indicated that some exothermic reactions occurred in the temperature range of 1000–1100 °C. In addition, the TGA/DSC results under nitrogen atmosphere are highly similar to those under air atmosphere (Fig. 4(b)) that we studied previously,³⁷ indicating that different atmospheres and variable oxygen exposure have no obvious impact on the sintering process.

Fig. 4 TG and DSC curves of the mixture of bauxite and Na₂CO₃ with N/(A + S) = 1 under nitrogen (a) and air (b)³⁷ atmospheres.

The sintering time depends on the reaction rate, which in turn depends on the sintering temperature. In our previous studies, the sintering time was usually set to 30 min.^36,37 In this study, the experimental range of sintering time was set to 10–50 min.

3.2 Effects of sintering parameters on the extraction of alumina and silica

The effects of sintering temperature, sintering time, and N/(A + S) on the extraction of Al₂O₃ and SiO₂ were investigated sequentially according to the set experimental ranges, and the results were shown in Fig. 5 and 6.

Fig. 5 Effects of sintering temperature (a), sintering time (b), and N/(A + S) (c) on the recovery efficiencies of Al₂O₃.

Fig. 6 Effects of sintering temperature (a), sintering time (b), and N/(A + S) (c) on the recovery efficiency of SiO₂ and the purity of silica powder.

As shown in Fig. 5, the sintering parameters have a significant impact on the recovery efficiencies of Al₂O₃. Under the conditions of N/(A + S) = 1 and a sintering time of 30 min, the recovery efficiency of Al₂O₃ in water leaching process (η_A1) was first increased and then decreased when the sintering temperature was increased from 850 °C to 1050 °C and reached a maximum value of 84.2% at 950 °C. Under the conditions of N/(A + S) = 1 and a sintering temperature of 950 °C, η_A1 increased significantly when the sintering time was prolonged from 10 min to 30 min. Upon further prolonging the sintering time to 50 min, η_A1 was stabilized at about 84%. Under the conditions of a sintering temperature of 950 °C and a sintering time of 30 min, η_A1 continued to increase from 37.3% to 87.1% when N/(A + S) was increased from 0.6 to 1.4, and the increase became gradually lower. The recovery efficiency of Al₂O₃ in acid leaching process (η_A2) showed an opposite trend to that of η_A1, while the total recovery efficiency of Al₂O₃ (η_AT) showed a similar trend to that of η_A1 with lower rangeability. η_AT was maintained above 90%, up to 98.3%.

As shown in Fig. 6, the sintering parameters had little effect on the recovery efficiency of SiO₂ (η_S), but had a significant effect on the purity of silica powder (P_S). η_S was mainly distributed in the range of 85–89% with slight variations. Silica powder had a high purity of 87–90% under the conditions of sintering temperature of 950–1050 °C, sintering time of 30–50 min or N/(A + S) = 1–1.4. But under other conditions, the purity of silica powder was lower.

According to the above experimental results, the optimal sintering parameters were selected as sintering temperature of 950 °C, sintering time of 30 min and N/(A + S) = 1. Under these conditions, η_A1, η_A2, η_AT, and η_S were 84.2%, 13.5%, 97.7% and 86.4%, respectively. The silica powder obtained at optimal sintering parameters had a purity of 88.2%. The chemical compositions of the corresponding solid samples are shown in Table 3.

Table 3 Chemical composition of the solid samples obtained at optimal sintering parameters (wt%)

Component	Al2O3	SiO2	Fe2O3	CaO	Na2O	TiO2	LOI
Sintered sample	42.3 ± 1.0	11.4 ± 0.4	2.7 ± 0.2	0.6 ± 0.1	39.2 ± 0.8	1.6 ± 0.1	1.7 ± 0.3
Water leaching residue	20.0 ± 0.6	34.4 ± 0.9	9.5 ± 0.3	2.3 ± 0.1	9.2 ± 0.5	5.5 ± 0.3	16.2 ± 1.9
Silica powder	1.1 ± 0.2	88.2 ± 1.7	0.6 ± 0.2	0.5 ± 0.1	0.6 ± 0.1	2.6 ± 0.3	6.5 ± 1.5

---

3.3 Phase transformations

The phase transformations in the sintered samples and water leaching residues obtained at various sintering parameters were analyzed by XRD, and the results are shown in Fig. 7.

Fig. 7 XRD patterns of sintered samples (a, c and e) and water leaching residues (b, d and f).

Fig. 7(a) shows the XRD patterns of the sintered samples obtained at different sintering temperatures. At 850 °C, the crystallographic phases in the sintered sample were primarily Na_1.95Al_1.95Si_0.05O₄ and Na_1.75Al_1.75Si_0.25O₄. In addition, trace amounts of α-Al₂O₃ and Na₂CO₃ were also present in the sintered sample, indicating that the reaction between bauxite and Na₂CO₃ had not yet fully occurred. As the sintering temperature increased to 950 °C, the diffraction peaks of Na_1.75Al_1.75Si_0.25O₄, α-Al₂O₃ and Na₂CO₃ weakened, while those of Na_1.95Al_1.95Si_0.05O₄ increased in intensity, indicating that Na_1.75Al_1.75Si_0.25O₄, α-Al₂O₃ and Na₂CO₃ reacted to form Na_1.95Al_1.95Si_0.05O₄. At 950 °C and above, only Na_1.95Al_1.95Si_0.05O₄ and Na_1.75Al_1.75Si_0.25O₄ were detected in the sintered samples, and the former was gradually converted into the latter with increasing temperature. According to the TGA/DSC analysis (Fig. 4), the transformation of Na_1.95Al_1.95Si_0.05O₄ into Na_1.75Al_1.75Si_0.25O₄ was exothermic. Fig. 7(b) shows the XRD patterns of the water leaching residues corresponding to the sintered samples in Fig. 7(a). At 850 °C, only α-Al₂O₃ was detected in the water leaching residue. As the sintering temperature increased to 1050 °C, the diffraction peaks of α-Al₂O₃ weakened and disappeared, while those of Na_1.75Al_1.75Si_0.25O₄ arose and increased in intensity. At 950 °C and above, only Na_1.75Al_1.75Si_0.25O₄ were detected in the water leaching residues. In addition, an amorphous peak was present in where 2θ is in the range of 20–40°, indicating the presence of amorphous phase in the water leaching residues.

The XRD patterns of the sintered samples and water leaching residues obtained at different sintering times were shown in Fig. 7(c) and (d), respectively. The changes in the XRD patterns at 10–30 min presented similar trends to those at 850–950 °C in Fig. 7(a) and (b). When the sintering time prolonged from 30 min to 50 min, the XRD patterns had no obvious changes anymore.

The XRD patterns of the sintered samples obtained at different N/(A + S) were shown in Fig. 7(e). When N/(A + S) was 0.6, Na_1.75Al_1.75Si_0.25O₄, α-Al₂O₃ and Na_1.95Al_1.95Si_0.05O₄ were detected in the sintered sample, in which Na_1.75Al_1.75Si_0.25O₄ played a dominant role. As N/(A + S) increased from 0.6 to 1.4, Na_1.75Al_1.75Si_0.25O₄ and α-Al₂O₃ transformed into Na_1.95Al_1.95Si_0.05O₄ entirely. In addition, the diffraction peaks of Na₂CO₃ and NaAlO₂ arose at N/(A + S) of 1.2 and 1.4, respectively. The presence of Na₂CO₃ in the sintered samples suggested an over-dosing of it. The XRD patterns of the water leaching residues corresponding to the sintered samples in Fig. 7(e) were shown in Fig. 7(f). When N/(A + S) was 0.6, Na_1.75Al_1.75Si_0.25O₄ and α-Al₂O₃ were detected in the water leaching residue. As N/(A + S) increased from 0.6 to 1.4, the diffraction peaks of Na_1.75Al_1.75Si_0.25O₄ and α-Al₂O₃ weakened and disappeared, leaving only those of amorphous phases.

Comparing the major phases in the sintered samples (Fig. 7(a), (c) and (e)) with those in the corresponding water leaching residues (Fig. 7(b), (d) and (f)), it can be known that Na₂CO₃, Na_1.95Al_1.95Si_0.05O₄ and NaAlO₂ were entirely dissolved during water leaching, while Na_1.75Al_1.75Si_0.25O₄ was partially dissolved, and α-Al₂O₃ was not dissolved. Therefore, the extraction of Al₂O₃ in water leaching process was primarily related to the formation and transformation of Na_2−xAl_2−xSi_xO₄ (x = 0, 0.05 and 0.25) in the sintered samples. According to Fig. 5, η_A1 was relatively low when Na_1.75Al_1.75Si_0.25O₄ was dominant in the sintered samples, and significantly increased with the transformation of Na_1.75Al_1.75Si_0.25O₄ into Na_1.95Al_1.95Si_0.05O₄. When Na_1.95Al_1.95Si_0.05O₄ was further converted into NaAlO₂, η_A1 increased only slightly. In the water leaching process, few SiO₂ in the sintered sample was leached out with the dissolution of Na_1.75Al_1.75Si_0.25O₄ and Na_1.95Al_1.95Si_0.05O₄, while most of SiO₂ was left in the water leaching residue. After water leaching, α-Al₂O₃, Na_1.75Al_1.75Si_0.25O₄, and an amorphous phase were retained in the water leaching residues, where one of the latter two phases dominated. According to the experimental phenomena, the water leaching residues without α-Al₂O₃ was almost completely dissolved during acid leaching, indicating that Na_1.75Al_1.75Si_0.25O₄ and the amorphous phase have excellent acid solubility. As a result, almost all the remaining Al₂O₃ and SiO₂ in the water leaching residue were extracted, leading to high levels of η_AT, η_S and the purity of silica powder. In contrast, η_AT was lower when α-Al₂O₃ was present in the water leaching residue, indicating that α-Al₂O₃ was difficult to dissolve during acid leaching. As a result, the purity of silica powder was also lower.

The optimal sintering parameters (950 °C, 30 min and N/(A + S) = 1) were selected to compare the identified phases in Fig. 7 with the thermodynamic predictions of Fig. 3. The identified phases in the sintered sample were primarily Na_2−xAl_2−xSi_xO₄ (x = 0.05 and 0.25), which are NaAlO₂ solid solutions predicted by thermodynamics. In addition to the NaAlO₂ solid solutions, a liquid slag was also predicted by thermodynamics, which contained most of the SiO₂ in the Na₂O–Al₂O₃–SiO₂ system. The crystallinity of the sintered sample was calculated to be 80.3%, indicating the presence of an amorphous phase, which is the main Si-bearing phase in the sintered sample and corresponds to the liquid slag predicted by thermodynamics.

3.4 Micromorphology

In order to further reveal the process mechanism, the micromorphology of the sintered sample and water leaching residue obtained at optimal sintering parameters were observed and analyzed.

Fig. 8 shows the XRD pattern and SEM images results for the sintered sample. According to Fig. 8(b), the particles in the sintered sample could be divided into two groups. As shown in Fig. 8(c), the first kind of particles were loose and porous aggregates of fine lamellar crystals, which is favorable for water leaching. According to Fig. 8(a), the crystalline phase in the sintered sample was primarily Na_2−xAl_2−xSi_xO₄ (x = 0.05 and 0.25). Combined with the EDS results of spot 5–7 (Table 4), it could be concluded that the first kind of particles mostly consisted of Na_2−xAl_2−xSi_xO₄ (x = 0.05 and 0.25). As shown in Fig. 8(d), the second kind of particles were in the form of dense blocks with smooth surfaces and numerous microscopic pores inside. The micromorphology indicated that the second kind of particles were amorphous, which verifies the presence of an amorphous phase in the sintered sample. According to Table 4, the second kind of particles (spot 8–10) contained less Na, Al and more Ca, Ti, Fe compared to the first kind of particles (spot 5–7). The average A/S of these two kinds of particles were 4.23 and 2.61, respectively, while that of the sintered sample was 3.71. Based on the difference in A/S, the proportion of the first and second kinds of particles was calculated to be about 68% and 32%, respectively.

Fig. 8 XRD pattern (a) and SEM images (b–d) of the sintered sample obtained at optimal sintering parameters.

Table 4 Elemental compositions of the sintered sample particles from EDS analysis (wt%)

Element	O	Na	Al	Si	Ca	Ti	Fe
Spot 5	33.82	31.52	27.02	5.68	0	0.64	1.32
Spot 6	22.81	32.3	31.2	6.42	0	3.53	3.74
Spot 7	32.39	36.3	18.09	3.82	0.38	0.48	1.21
Average (5–7)	29.67	33.37	25.44	5.31	0.13	1.55	2.09
Standard deviation	5.99	2.56	6.70	1.34	0.22	1.72	1.43
Spot 8	33.83	26.88	16.52	5.32	0.36	1.31	3.71
Spot 9	34.11	29.02	14.93	4.94	0.78	1.63	3.44
Spot 10	33.36	32.71	15.04	5.45	1	3.99	8.45
Average (8–10)	33.77	29.54	15.50	5.24	0.71	2.31	5.20
Standard deviation	0.38	2.95	0.89	0.27	0.33	1.46	2.82

---

Fig. 9 shows the XRD pattern and SEM images of the water leaching residue obtained at optimal sintering parameters. According to Fig. 9(a), the water leaching residue was primarily composed of amorphous phases. As shown in Fig. 9(b)–(d), the amorphous particles were primarily massive, with rough surfaces and numerous dissolution pores, which was inferred to be the result of partial dissolution and secondary reactions of the amorphous phase particles in the sintered sample during water leaching. The EDS results of the amorphous particles in the sintered sample (Table 4) and water leaching residue (Table 5) show that the contents of Na and Al decreased after water leaching, while those of Si, Ca, Ti and Fe increased relatively, further confirming the above inference. As a result, the A/S of the amorphous particles decreased to 0.56, which was similar to that of the water leaching residue (0.58). In addition, the particles in the water leaching residue have a size of less than 100 μm and a porous structure, which favors acid leaching.

Fig. 9 XRD pattern (a) and SEM images (b–d) of the water leaching residue obtained at optimal sintering parameters.

Table 5 Elemental compositions of the water leaching residue particles from EDS analysis (wt%)

Element	O	Na	Al	Si	Ca	Ti	Fe
Spot 11	44.16	9.37	11.06	18.09	2.16	3	5.87
Spot 12	40.75	10.52	12.34	19.06	2.2	3.99	11.14
Spot 13	41.64	8.75	11.06	16.9	2.55	2.89	6.95
Average (11–13)	42.18	9.55	11.49	18.02	2.30	3.29	7.99
Standard deviation	1.77	0.90	0.74	1.08	0.21	0.61	2.78

---

3.5 Discussion of the mechanism and advantages of the novel process

Based on the above analysis results, the phase transformations of the Al- and Si-bearing minerals during sintering and leaching were revealed. In addition, the formation mechanism and application of silica gel were discussed.

In this study, diaspore and kaolinite are the main Al- and Si-bearing minerals in the high-silica bauxite, respectively. In the sintering process, diaspore and kaolinite were primarily converted into Na_2−xAl_2−xSi_xO₄ (x = 0, 0.05 and 0.25) and an amorphous phase by fully reacting with sufficient Na₂CO₃. In the water leaching process, Na_2−xAl_2−xSi_xO₄ (x = 0, 0.05 and 0.25) dissolves completely or partially depending on the x value, while the amorphous phase partially dissolves and undergoes some secondary reactions. The residual amorphous phase and Na_2−xAl_2−xSi_xO₄ (x = 0.25) formed the main body of the water leaching residue and were almost completely dissolved in the subsequent acid leaching process. If the dosage of Na₂CO₃ is insufficient or the sintering reaction does not fully occur, a part of diaspore would be dehydroxylated to form α-Al₂O₃, which is insoluble in both the water leaching and acid leaching processes. Fig. 10 shows the proposed mechanism at optimal sintering parameters.

Fig. 10 Schematic diagram of the process mechanism at optimal sintering parameters.

According to Table 3, the chemical composition of the water leaching residue was dominated by SiO₂, Al₂O₃ and Fe₂O₃, which underwent the following reactions during acid leaching.^41,42

M₂SiO_4(s) + 2H₂SO_4(aq) → H₄SiO_4(aq) + 2MSO_4(aq) (5)

Al₂O_3(s) + 3H₂SO_4(aq) → Al₂(SO₄)_3(aq) + 3H₂O (6)

Fe₂O_3(s) + H₂SO_4(aq) → Fe₂(SO₄)_3(aq) + 3H₂O (7)

Accordingly, the compositions of the acid leaching solution were dominated by H₄SiO₄, Al₂(SO₄)₃, and Fe₂(SO₄)₃. H₄SiO₄ will gradually polymerize in the acid leachate, forming dimers, trimers, cyclic oligomers and up to silica gels,^43,44 as shown in Fig. 11. Pure silica gel was obtained using the method described in Section 2.2, as shown in Fig. 12(a). The silica gel has excellent water absorbing properties, containing 90–95% water, which can be processed as a water retention agent, desiccant, or absorbent. The silica powder (Fig. 12(b)) obtained by drying silica gel has a high purity of 87–90%, which is primarily amorphous according to Fig. 12(c). As shown in Fig. 12(d) and (e), the amorphous silica particles are primarily layered and thus have a specific surface area of up to ∼406 m² g⁻¹. Therefore, the silica gel can also be used as a raw material for the production of silica-based products. After separating out silica gel, the remaining solution had Al₂(SO₄)₃ and Fe₂(SO₄)₃ as the main components, which can be used to produce PAFS.³⁵

Fig. 11 Diagram of the formation mechanism of silica gel.⁴³

Fig. 12 Photographs of silica gel (a) and silica powder (b); XRD pattern (c) and SEM images (d and e) of silica powder.

The above process realizes the comprehensive utilization of high silica bauxite and has obvious advantages compared with the traditional lime-soda sintering process, as shown in Table 6. First, lime is not introduced into the sintering process, reducing the material flow and improving the production efficiency. Second, the sintering temperature is reduced from 1250–1350 °C to 950 °C, resulting in lower sintering energy consumption. Third, SiO₂ and Fe₂O₃ are also extracted and utilized with high levels of recovery in addition to Al₂O₃. Fourth, residue is fully utilized without discharge.

Table 6 Comparison between the novel process and the lime-soda sintering process

Process	Sintering parameters	Recovery efficiency			Emission of residue
		Al2O3	SiO2	Fe2O3
The lime-soda sintering process	1250–1350 °C	88–92%	—	—	>1.5 t per t-Al2O3
	C/S = 2, N/A = 1
The novel process	950 °C	97.7% (84.2% for alumina, 13.5% for PAFS)	86.4%	95.9%	0
	N/(A + S) = 1

---

4 Conclusions

The extraction of Al₂O₃ and SiO₂ from high-silica bauxite was realized by sintering with N₂CO₃ followed by two-step leaching with water and acid. During the sintering of bauxite with N₂CO₃, diaspore and kaolinite were transformed into α-Al₂O₃, Na_2−xAl_2−xSi_xO₄ (x = 0, 0.05 and 0.25) and an amorphous phase. In the subsequent two-step leaching processes, α-Al₂O₃ was insoluble, while the other two phases were almost completely dissolved. The formation of α-Al₂O₃ could be avoided by adjusting sintering parameters. The optimal sintering parameters were determined experimentally to be sintering temperature of 950 °C, sintering time of 30 min, and N/(A + S) of 1. Under these conditions, ∼97% of Al₂O₃ and ∼86% of SiO₂ were extracted for the production of alumina, PAFS and silica gel. The purity of silica gel was positively correlated with the total recovery efficiency of Al₂O₃. The dried silica gel reached a purity of ∼88% at optimal sintering parameters. These findings suggest an idea for the comprehensive utilization of high-silica bauxite and the minimization of residue.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors are grateful for the financial support from the Collaborative Innovation Center Project of the Shaanxi Provincial Department of Education, China (No. 21JY027).

References

1. N. Zhang, A. V. Nguyen and C. Zhou, Adv. Colloid Interface Sci., 2018, 254, 56–75.
2. P. Smith, Hydrometallurgy, 2009, 98, 162–176.
3. Z. Wu, H. Lv, M. Xie, L. Li, H. Zhao and F. Liu, Ceram. Int., 2022, 48, 18676–18686.
4. X. Pan, H. Wu, H. Yu and S. Bi, Hydrometallurgy, 2020, 197, 105469.
5. N. Kawashima, L. Shi, N. Xu, J. Li and A. R. Gerson, Hydrometallurgy, 2016, 159, 75–86.
6. B. Gibson, D. G. Wonyen and S. C. Chelgani, Miner. Eng., 2017, 114, 64–73.
7. Y. Wu, X. L. Pan, Y. J. Han and H. Y. Yu, Trans. Nonferrous Met. Soc. China, 2019, 29, 2627–2637.
8. Y. Xu, J. Li, C. Chen, Y. Lan and L. Wang, JOM, 2020, 72, 2705–2712.
9. X. B. Li, H. Y. Wang, Q. S. Zhou, T. G. Qi, G. H. Liu, Z. H. Peng and Y. L. Wang, Trans. Nonferrous Met. Soc. China, 2019, 29, 416–423.
10. Q. Wang, X. F. Sheng, L. Y. He and Y. Shan, Miner. Eng., 2018, 128, 179–186.
11. G. Lü, T. Zhang, X. Zhu, C. Zheng, Y. Wang, W. Zhang and Z. Zhang, Chin. J. Chem. Eng., 2019, 27, 1965–1972.
12. Y. Wang, T. Zhang, G. Lyu, L. Ma and W. Zhang, J. Cleaner Prod., 2021, 290, 125828.
13. C. Liu, S. Ma, S. Zheng, Y. Luo, J. Ding, X. Wang and Y. Zhang, Hydrometallurgy, 2018, 175, 224–231.
14. S. N. Wang, S. L. Zheng and Y. Zhang, Chin. J. Process Eng., 2007, 7, 967–972.
15. A. Alp and A. O. Aydin, Can. Metall. Q., 2002, 41, 41–46.
16. J. Pei, X. Pan, Y. Zhang, H. Yu and G. Tu, J. Environ. Chem. Eng., 2021, 9, 106754.
17. T. Le, S. Ju, L. Lu, J. Peng, L. Zhou and S. Wang, Hydrometallurgy, 2017, 169, 124–134.
18. A. B. ElDeeb, V. N. Brichkin, M. Bertau, Y. A. Savinova and R. V. Kurtenkov, Appl. Clay Sci., 2020, 196, 105771.
19. Y. J. Xu, H. X. Xin, H. M. Duan, Y. D. Li, Y. Wu, J. Lin and Y. C. Zhai, J. Cent. South Univ., 2022, 29, 22–31.
20. D. Tian, X. Y. Shen, Y. C. Zhai, P. Xiao and P. Webley, J. Iron Steel Res. Int., 2019, 26, 578–584.
21. D. Valeev, A. Shoppert, D. Dogadkin, T. Romashova, T. Kuz'mina and C. Salazar-Concha, Hydrometallurgy, 2023, 215, 105994.
22. Y. Zhang, J. Zhang, L. Wu, L. Tan, F. Xie and J. Cheng, J. Hazard. Mater., 2021, 404, 124044.
23. Y. Zhao, Y. Zheng, H. He, Z. Sun and A. Li, J. Environ. Chem. Eng., 2021, 9, 104770.
24. T. S. Barry, T. Uysal, M. Birinci and M. Erdemoğlu, Min., Metall., Explor., 2019, 36, 557–569.
25. S. Ma, Z. Wen, J. Chen and S. Zheng, Miner. Eng., 2009, 22, 793–798.
26. J. Y. Ma, Z. B. Li and Q. G. Xiao, AIChE J., 2012, 58, 2180–2191.
27. W. Liu, J. Yang and B. Xiao, Int. J. Miner. Process., 2009, 93, 220–231.
28. Y. Li, X. Pan, Z. Lv, H. Wu and H. Yu, Miner. Eng., 2022, 187, 107805.
29. Y. Xu, C. Chen, Y. Lan, L. Wang and J. Li, J. Mater. Res. Technol., 2020, 9, 7418–7426.
30. Y. Wang, T. Zhang, Y. Zhang, G. Lyu and W. Zhang, Miner. Eng., 2019, 138, 139–147.
31. G. Power, M. Grafe and C. Klauber, Hydrometallurgy, 2011, 108, 33–45.
32. X. Liu, Y. Han, F. He, P. Gao and S. Yuan, J. Hazard. Mater., 2021, 420, 126542.
33. P. S. Reddy, N. G. Reddy, V. Z. Serjun, B. Mohanty, S. K. Das, K. R. Reddy and B. H. Rao, Waste Biomass Valorization, 2021, 12, 1185–1217.
34. Y. Wang, T.-a. Zhang, G. Lyu, F. Guo, W. Zhang and Y. Zhang, J. Cleaner Prod., 2018, 188, 456–465.
35. J. Chang, A. Pan, Y. Ma, Y. Sun and S. Hu, Minerals, 2023, 13, 105.
36. J. Chang, A. Pan, Y. Ma, Y. Sun, S. Hu and K. Li, Metall. Res. Technol., 2022, 119, 114.
37. Y. Sun, A. Pan, Y. Ma, J. Chang, K. Li and S. Hu, Miner. Eng., 2022, 187, 107782.
38. T. Le, S. H. Ju, A. V. Ravindra, X. T. Li and Q. Wang, JOM, 2019, 71, 831–837.
39. C. Dessemond, G. Soucy, J.-P. Harvey and P. Ouzilleau, Minerals, 2020, 10, 519.
40. A. Kosminski, D. P. Ross and J. B. Agnew, Fuel Process. Technol., 2006, 87, 1051–1062.
41. G. Hu, H. Tang, D. He, W. Sun and L. Wang, Miner. Eng., 2021, 173, 107180.
42. H. Tanvar and B. Mishra, Sustainable Mater. Technol., 2022, 33, e00466.
43. G. Alkan, B. Yagmurlu, S. Cakmakoglu, T. Hertel, Ş. Kaya, L. Gronen, S. Stopic and B. Friedrich, Sci. Rep., 2018, 8, 5676.
44. D. Voßenkaul, A. Birich, N. Müller, N. Stoltz and B. Friedrich, J. Sustain. Metall., 2017, 3, 79–89.

---