Reaction Behavior of Kaolinite in Sulfur-Bearing Sodium Aluminate Solution under the Simulated Bayer Process

Abstract

Over a billion tons of high-sulfur bauxite has not been utilized effectively currently in China, because the pyrite existing in the bauxite poses a range of hazards during the Bayer process. A novel idea was proposed to remove sulfur by the silicon-containing minerals in bauxite reacting with sulfur species in sodium aluminate solution to form sulfur-bearing desilication products (SDSP) for discharge with the red mud in the Bayer process. This study investigated the reaction behavior between kaolinite and different sulfur-containing ions under the simulated Bayer process conditions, elucidating the desulfurization rate variation and formation mechanism of SDSPs. The thermodynamic calculations suggest that the reaction between kaolinite and sulfur-bearing sodium aluminate solution to form SDSPs can occur spontaneously. The experimental results demonstrated that various SDSPs can be produced through the reaction of kaolinite and sulfur-containing ions in sodium aluminate solution during the simulated Bayer process, resulting in various desulfurization efficiencies, while the desulfurization process will not result in additional alkali consumption. Increasing the kaolinite dosage, extending the reaction time, and elevating the reaction temperature all contribute positively to enhancing desulfurization efficiency. Kaolinite reacted with ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ in sodium aluminate solution to generate Na₈Al₆Si₆O₂₄S₂O₃·2H₂O, achieving a desulfurization rate exceeding 90% under optimized conditions. Under the simulated Bayer digestion process conditions at elevated temperature, the desulfurization rates of kaolinite ranked in ascending order as ${\mathrm{S}}^{2-}$ < ${\mathrm{SO}}_3^{2-}$ < ${\mathrm{SO}}_4^{2-}$ < ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$. Kaolinite reacted with ${\mathrm{SO}}_4^{2-}$ and ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ to form cancrinite type SDSPs, and a superior desulfurization rate can be achieved. This work can provide a theoretical foundation and technological support for the efficient utilization of high-sulfur bauxite by the Bayer process.

1. Introduction

The alumina output of China exceeded 80 million tons in 2022, but domestic high-quality bauxite resources are increasingly short, resulting in a large amount of bauxite needing to be imported from abroad, and the foreign bauxite dependence ratio of China was over 50% in recent years [1]. Meanwhile, China has nearly 2 billion tons of high-sulfur bauxite that has not been fully exploited and utilized [2], which has the characteristics of high aluminum–silicon ratio and high sulfur content, and so, how to use it economically and efficiently has become a hot research topic. In general, bauxite with a sulfur content exceeding 0.7% can be classified as high-sulfur bauxite [3]. However, in reality, the sulfur content in high-sulfur bauxite is usually more than 1%, of which the most harmful sulfur mineral exists in the form of pyrite (FeS₂). When treating high-sulfur bauxite by the Bayer process, the pyrite will react with sodium aluminate solution during the high-temperature digestion process to form ${\mathrm{S}}^{2-}$ in solution, leading to a series of hazards [4,5,6].

The detrimental effects of ${\mathrm{S}}^{2-}$ primarily encompass the following aspects [7,8,9,10,11,12], such as increased alkali consumption, serious steel equipment corrosion, Fe-contamination of alumina product, reduced alumina digestion ratio of bauxite and seed precipitation ratio et al. Many researchers have developed various technologies aimed at mitigating the detrimental impacts of sulfur in bauxite during the production process. The proposed desulfurization methods can be primarily classified into sulfur removal in bauxite by pretreatment and solution desulfurization by additives. The desulfurization technologies for bauxite pretreatment reduce the sulfur entering the Bayer process system from the source, with main methods including roasting [13,14], flotation [15], electrolysis [16,17], microbiology [18], and microwave treatment [19]. It is difficult to achieve complete sulfur removal from bauxite in the pretreatment processes, resulting in the continuous accumulation of sulfur in the sodium aluminate solution during the Bayer process, and still requiring desulfurization of the solution. In the solution desulfurization methods, additives are usually added to mitigate the detrimental effects of ${\mathrm{S}}^{2-}$ through precipitation or oxidation. ${\mathrm{S}}^{2-}$ can be effectively removed by incorporating additives such as ZnO [9], Fe(OH)₂ [20], NaFeO₂ [21], special seeds [22] to facilitate their conversion into corresponding precipitates. Alternatively, the addition of oxidants [23] can convert ${\mathrm{S}}^{2-}$ into ${\mathrm{SO}}_4^{2-}$, followed by the addition of barium compounds, lime, etc., to remove sulfur by precipitation or evaporative crystallization. The aforementioned measures can effectively decrease the sulfur content in bauxite or sodium aluminate solution and alleviate negative impact caused by sulfur-containing ions on production. Nevertheless, due to problems such as high expenses or intricate procedures, the large-scale utilization of high-sulfur bauxite by the Bayer process for alumina production remains limited.

In the Bayer process, silica-containing minerals presented in bauxite readily react with sodium aluminate solution to form hydrated sodium aluminum silicate (DSP, also known as desilication products). In practical production, sodium aluminate solution usually contains a certain concentration of impurity ions, which invariably impact the composition of DSP. For example, different anions such as ${\mathrm{Cl}}^{-}$, ${\mathrm{OH}}^{-}$, ${\mathrm{S}}^{2-}$, ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_4^{2-}$, ${\mathrm{CO}}_3^{2-}$ can be embedded in the formed DSP with various embedding capacities in sodium aluminate solution [24,25]. In our previous work [26], it was preliminarily verified that the silicate minerals in bauxite have exhibited desulfurization ability to some extent under Bayer process conditions. Silicate minerals can react with various sulfur ions species in sodium aluminate solution to generate sulfur-bearing desilication products (SDSP), which can increase the sulfur content in red mud and reduce the sulfur content in the solution. Namely, using kaolinite in bauxite as a desulfurizer during the Bayer process, the obtained SDSPs will be introduced into the red mud for discharge, which can effectively mitigate the accumulation and detrimental effects of sulfur in the Bayer process. However, the added lime promotes the formed SDSPs to convert to sulfur-free hydrated sodium calcium aluminosilicates [27], causing sulfur to re-enter the sodium aluminate solution system and subsequently decreasing the desulfurization efficiency of silicon minerals. The use of non-lime additives can prevent the formation of hydrated sodium calcium aluminosilicates and provide new approaches for enhancing the desulfurization capability of silicon minerals. Additionally, a reported work [28] investigated the formation of SDSPs using synthetic silicon-containing material in sodium aluminate solution, and the results showed that the sulfur content in SDSPs is positively correlated with the content of generated cancrinite, while both ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$ facilitate the formation of cancrinite, thereby enhancing the efficiency of desulfurization.

Under the Bayer process conditions, the direct desulfurization process using silicate minerals becomes highly intricate due to the diverse range of sulfur-containing ions present. Therefore, selecting kaolinite as a representative silicate mineral, this study investigated the reaction between various sulfur ions (${\mathrm{S}}^{2-}$, ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$) and kaolinite under the simulated Bayer digestion conditions. By examining the desulfurization efficiency of kaolinite and analyzing the phase composition and microstructure of obtained SDSPs, the desulfurization reaction behaviors and mechanism of kaolinite were clarified. The results can provide theoretical support for directly desulfurizing using silicate minerals in bauxite during the utilization of high-sulfur bauxite by the Bayer process.

2. Experimental Methods

2.1. Materials

In this paper, the sodium aluminate solution was prepared by dissolving industrial-grade aluminum hydroxide into a sodium hydroxide solution. All other reagents used in the experiments were of analytical grade. The concentration of caustic alkaline (N_k, calculated as Na₂O) and aluminum oxide (Al₂O₃) in the prepared sodium aluminate solution was determined by titration, then the mole ratio of Na₂O_k to Al₂O₃ (α_k) can be calculated. Finally, the obtained sodium aluminate solution was sealed and stored in a plastic container. The desired concentration of sulfur-bearing sodium aluminate solution can be obtained by diluting the pre-prepared sodium aluminate solution and adding a certain amount of analytical grade sulfide compounds (including Na₂S·9H₂O, Na₂S₂O₃·5H₂O, Na₂SO₃ or Na₂SO₄).

The silicate mineral used in this study was kaolinite, and its XRD patterns are presented in Figure 1.

In Figure 1, it can be seen that characteristic peaks were clear and sharp, with its primary mineral phases being high-purity kaolinite (Al₂Si₂O₅(OH)₄) and halloysite (Al₂Si₂O₅(OH)₄·2H₂O).

2.2. Method

2.2.1. Digestion Process

The digestion experiments were conducted in a low-pressure reaction kettle heated with glycerol (<140 °C) or a high-pressure reaction kettle heated with molten salt (>150 °C), and the temperature control accuracy of both reactors is ±1 °C.

The experimental procedure was as follows. A measured amount of kaolinite was added to a steel reaction vessel, followed by the addition of 100 mL sulfur-bearing sodium aluminate solution. Four steel balls (two with a diameter of 18 mm and two with a diameter of 8 mm) were then introduced to enhance stirring. The sealed reaction vessel was placed in an oven that had been preheated to the specified temperature using glycerol or molten salt and rotated for the designated time period. After cooling, the resulting slurry was filtrated and washed by boiling water, and then the filtered residue was dried in a vacuum drying oven at 50 ± 1 °C. Finally, the dry residue was weighed, and the sulfur content and phase composition were measured.

2.2.2. Characterization

The silica content in digestion residues of kaolinite was determined using the molybdenum blue spectrophotometric method (GB/T 6610.3-2003). The determination of sodium oxide content was performed by employing the flame photometric method (GB/T 6610.5-2003). The sulfur content of digestion residues was determined by carbon-sulfur analyzer (HDS3000, Hunan Huade, China). The mineral phases were identified by X-ray diffraction (XRD, D/max 2550VB, Rigaku, Japan) using Cu Kα radiation at a scan rate of 8°·min⁻¹, while the micromorphology and micro area composition analyses were performed using back-scattered scanning electron microscopy (SEM, MIRA3-LMH, TESCAN, Czech Republic) and energy-dispersive X-ray spectrometry (EDS, X MAX20, Oxford, England), respectively.

2.2.3. Desulfurization Rate Calculation

The sulfur removal efficiency of kaolinite was characterized by the desulfurization rate, which can be quantified using Equation (1).

$$\mathsf{\eta }=\frac{m_s\times {\omega }_s}{m_0}\times 100$$

(1)

where η is desulfurization rate (%), m_s is the mass of obtained SDSPs after digestion (g), ω_s is the sulfur content in obtained SDSPs (wt. %), m₀ is the total mass of sulfur added in the sodium aluminate solution (g).

3. Results and Discussion

3.1. Thermodynamic Calculation of the SDSP Formation Reaction

Firstly, thermodynamic calculations were conducted to investigate the reactions between kaolinite and various sulfur-bearing sodium aluminate solution. Equations (2)–(5) illustrate possible reactions between different types of sulfur-containing ions and kaolinite in the formation of SDSPs.

$$3{\mathrm{Al}}_2{\mathrm{O}}_3·2\mathrm{Si}{\mathrm{O}}_2·2{\mathrm{H}}_2\mathrm{O}+8{\mathrm{Na}}^{+}+6{\mathrm{OH}}^{-}+{\mathrm{S}}^{2-}={\mathrm{Na}}_8{\mathrm{Al}}_6{\mathrm{Si}}_6{\mathrm{O}}_{24}\mathrm{S}+9{\mathrm{H}}_2\mathrm{O}$$

(2)

$$3{\mathrm{Al}}_2{\mathrm{O}}_3·2\mathrm{Si}{\mathrm{O}}_2·2{\mathrm{H}}_2\mathrm{O}+8{\mathrm{Na}}^{+}+6{\mathrm{OH}}^{-}+{\mathrm{S}}_2{\mathrm{O}}_3^{2-}={\mathrm{Na}}_8{\mathrm{Al}}_6{\mathrm{Si}}_6{\mathrm{O}}_{24}{\mathrm{S}}_2{\mathrm{O}}_3+9{\mathrm{H}}_2\mathrm{O}$$

(3)

$$3{\mathrm{Al}}_2{\mathrm{O}}_3·2\mathrm{Si}{\mathrm{O}}_2·2{\mathrm{H}}_2\mathrm{O}+8{\mathrm{Na}}^{+}+6{\mathrm{OH}}^{-}+{\mathrm{SO}}_3^{2-}={\mathrm{Na}}_8{\mathrm{Al}}_6{\mathrm{Si}}_6{\mathrm{O}}_{24}\mathrm{S}{\mathrm{O}}_3+9{\mathrm{H}}_2\mathrm{O}$$

(4)

$$3{\mathrm{Al}}_2{\mathrm{O}}_3·2\mathrm{Si}{\mathrm{O}}_2·2{\mathrm{H}}_2\mathrm{O}+8{\mathrm{Na}}^{+}+6{\mathrm{OH}}^{-}+{\mathrm{SO}}_4^{2-}={\mathrm{Na}}_8{\mathrm{Al}}_6{\mathrm{Si}}_6{\mathrm{O}}_{24}\mathrm{S}{\mathrm{O}}_4+9{\mathrm{H}}_2\mathrm{O}$$

(5)

In the above reactions, the ${∆}_{\mathrm{f}}{\mathrm{G}}^{⊝}$ of SDSPs at different temperatures were estimated using an approximate algorithm for the Gibbs free energy of complex silicate minerals introduced in reference [29], combined with thermodynamic data of simplified compounds and ionic components obtained from references [30,31,32]. The calculations were conducted using classical thermodynamic algorithms. Consequently, the relationship between the ${∆}_{\mathrm{r}}{\mathrm{G}}^{⊝}$ of reactions (2) to (5) and different temperatures was determined, as shown in Figure 2.

From Figure 2, it can be seen that when the reaction temperature ranged from 323 to 573 K, the ${∆}_{\mathrm{r}}{\mathrm{G}}^{⊝}$ for reactions (2) to (5) were all negative, indicating that the reactions of ${\mathrm{S}}^{2-}$, ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$ with kaolinite to form SDSP can occur spontaneously in sodium aluminate solution. Notably, reaction (5) displayed the smallest ${∆}_{\mathrm{r}}{\mathrm{G}}^{⊝}$ value among these reactions, suggesting the higher propensity for the formation of SDSP through the reaction between ${\mathrm{SO}}_4^{2-}$ and kaolinite, while the tendencies decreased in order for ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_3^{2-}$ and ${\mathrm{S}}^{2-}$.

The formula for DSP can be expressed as Na₂O·Al₂O₃·2SiO₂·(Na, X)·nH₂O, where X can be OH⁻, ${\mathrm{Al}\left(\mathrm{OH}\right)}_4^{-}$, S²⁻, ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$. According to the molecular formulas of different generated SDSP, it was evident that compared to ${\mathrm{S}}^{2-}$, ${\mathrm{SO}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$, the removal of ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ resulted in a higher desulfurization efficiency by forming Na₈(Al₆Si₆O₂₄)S₂O₃ when consuming equal amounts of Na and Al. Therefore, various sulfur ions species reacting with kaolinite to generate SDSPs are thermodynamically feasible, and ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ is likely more suitable to remove by generating SDSPs.

3.2. Desulfurization Efficiency of Kaolinite

The reaction between kaolinite and various sulfur-containing ions in sodium aluminate solution was investigated, and the effects of reaction time, temperature, and dosage of kaolinite on the desulfurization efficiency were studied in detail. Additionally, the phase composition, micromorphology, and micro-area components of the obtained digestion residues were also analyzed.

3.2.1. Effects of Reaction Time

In the simulated Bayer process condition, the effects of reaction time on the desulfurization efficiency of kaolinite were investigated under 100 °C (pre-desilication temperature) and 260 °C (bauxite digestion temperature). The results of kaolinite desulfurization results at 100 °C are depicted in Figure 3.

As shown in Figure 3, the desulfurization rate of kaolinite increased with prolonging reaction time at 100 °C in different sulfur-bearing sodium aluminate solutions. In the ${\mathrm{S}}^{2-}$-bearing sodium aluminate solution, desulfurization rate of kaolinite was about 3.5% in 10 min, and reached approximately 5.3% with a gradual increase in 120 min. By contrast, the desulfurization rate of kaolinite increased significantly for ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$ as the reaction time prolonged. Additionally, the highest desulfurization rate of kaolinite reached approximately 34% after a reaction time of 120 min in ${\mathrm{SO}}_3^{2-}$-bearing sodium aluminate solution. As for ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$, the desulfurization rates of kaolinite were both around 20% in 120 min. The results indicate that kaolinite exhibited a certain degree of desulfurization under the pre-desilication reaction conditions; however, the achieved highest desulfurization rate was only about 30% during the ${\mathrm{SO}}_3^{2-}$ removal reaction.

To improve the desulfurization efficiency of kaolinite, the desulfurization reactions were conducted at 260 °C, namely the conventional diaspore digestion temperature (as depicted in Figure 4). In Figure 4, with the increase in reaction time, a significant increase in desulfurization rates of kaolinite can be clearly observed after reacting with sodium aluminate solution containing ${\mathrm{S}}^{2-}$, ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$ at 260 °C. During the initial 10 min of the reactions, the desulfurization rates of kaolinite increased obviously, while the reaction time extended to 20 min, and the desulfurization rates grew slowly. The obtained desulfurization rates of kaolinite were approximately 13.3%, 87.6%, 57.3% and 70.2% for removing ${\mathrm{S}}^{2-}$, ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$, respectively. Further prolonging the reaction time to 60 min, the desulfurization rate remained stable, and no significant changes were observed.

From the above information, it can be inferred that increasing the reaction time and temperature are both beneficial for enhancing the desulfurization efficiency of kaolinite. Kaolinite exhibits a certain degree of desulfurization capability in sodium aluminate solution under Bayer digestion process at elevated temperature, especially in ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ removal process. According to desulfurization rates of different sulfur-bearing ions through SDSP formation at 260 °C, the consecutive sequences are ${\mathrm{S}}^{2-}$ < ${\mathrm{SO}}_3^{2-}$ < ${\mathrm{SO}}_4^{2-}$ < ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$.

3.2.2. Effects of Kaolinite Dosage

The aforementioned analysis revealed that kaolinite removed approximately 90% of ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ during the digestion process at elevated temperature in sodium aluminate solution. Consequently, subsequent investigations focused on exploring the reaction between kaolinite and ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$-bearing sodium aluminate solution. The effects of kaolinite dosage on desulfurization efficiency and phase composition of SDSPs are illustrated in Figure 5.

As shown in Figure 5a, it was evident that increasing the kaolinite dosage led to a remarkable improvement of the desulfurization rate. When the kaolinite dosage reached 120$\mathrm{g}·{\mathrm{L}}^{-1}$, the desulfurization rate can exceed 95%, indicating that the ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ removal efficiency was directly correlated with the mass of reacted kaolinite in the digestion process at elevated temperature in sodium aluminate solution. When the amount of added kaolinite was insufficient, the desulfurization rate was suboptimal, and the obtained digestion residues primarily consisted of Na₈(Al₆Si₆O₂₄)S₂O₃·2H₂O, while Na₈(Al₆Si₆O₂₄)S₂O₃·2H₂O and Na₈(Al₆Si₆O₂₄)(OH)₂·2H₂O were both detected in the residue (Figure 5b) with the surplus kaolinite (120$\mathrm{g}·{\mathrm{L}}^{-1}$) added. Nevertheless, it should be noted that a high content of silicate minerals in the Bayer process can diminish alumina digestion efficiency from bauxite and lead to more alumina loss, thereby negatively impacting production. Hence, under the simulated digestion process conditions in the Bayer process, a favorable efficiency in ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ removal needs to consider both desulfurization rate and an appropriate dosage of kaolinite.

3.2.3. Effects of Reaction Temperature

The reaction temperature has a significant influence on the desulfurization process of kaolinite in sodium aluminate solution. The desulfurization efficiency of kaolinite after reacting with ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$–bearing sodium aluminate solution was investigated within the temperature range of 180~280 °C, as shown in Figure 6.

From Figure 6a, it was evident that the desulfurization rate of kaolinite gradually increased with rising temperature from 180 °C to 240 °C, and the desulfurization rate of kaolinite reached approximately 80%. However, as the reaction temperature further elevated from 240 °C to 280 °C, the desulfurization rate remained relatively constant at approximately 80%, suggesting that the maximum desulfurization efficiency was achieved with the added kaolinite. According to the XRD patterns of SDSPs obtained at different temperatures (Figure 6b), Na₈(Al₆Si₆O₂₄)S₂O₃·2H₂O and Na₈(Al₆Si₆O₂₄)(OH)₂·2H₂O were formed by kaolinite at 180 °C, and the phases transformed into a single phase of Na₈(Al₆Si₆O₂₄)S₂O₃·2H₂O with the increase in the reaction temperature from 220 °C to 280 °C. Elevating the digestion temperature facilitates the generation of Na₈(Al₆Si₆O₂₄)S₂O₃·2H₂O, consequently enhancing the desulfurization efficiency of kaolinite. Therefore, considering that the alumina digestion efficiency of diaspore in high-sulfur bauxite will be improved obviously in practice, the selected appropriate digestion temperature is 260 °C.

3.2.4. Digestion Residues Characterization

In order to further clarify the differences in desulfurization efficiency of kaolinite reacting with different sulfur-bearing sodium aluminate solutions, XRD analysis for the phases of obtained residues under diverse conditions was conducted, as depicted in Figure 7.

The results depicted in Figure 7 demonstrate that the SDSPs were formed by kaolinite reacting with sulfur-bearing sodium aluminate solution, while the types of SDSPs were various with the species of sulfur ions. As reported in the research [26,33], sodalite-type DSPs are produced by the reaction of kaolinite in sodium aluminate solution with ${\mathrm{S}}^{2-}$ or ${\mathrm{SO}}_3^{2-}$, while molecular formulas are proposed to be Na_8.14(Al₆Si₆O₂₄)(SO₄)_1.14S_0.86, respectively. However, after the reaction with ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ or ${\mathrm{SO}}_4^{2-}$ in sodium aluminate solution, kaolinite transformed to cancrinite-type DSPs with the molecular formulas of Na₈(Al₆Si₆O₂₄)(S₂O₃)·2H₂O and Na₈(Al₆Si₆O₂₄)SO₄·3H₂O, respectively. Li et al. [28] have confirmed that the sulfur content in DSP is related to the content of formed cancrinite, indicating that the generation of cancrinite is beneficial to the desulfurization efficiency. From the molecular formula of SDSP, it can be inferred that each individual molecule of the SDSPs produced by the reaction between ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ and kaolinite in sodium aluminate solution contained two sulfur atoms, while other SDSPs obtained in the sodium aluminate solution by adding ${\mathrm{S}}^{2-}$, ${\mathrm{SO}}_3^{2-}$ or ${\mathrm{SO}}_4^{2-}$ contained only one sulfur atom. When sulfur existed in the form of ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ in a sodium aluminate solution and reacted with kaolinite, it led to a significant reduction in the sulfur content in the solution. However, in ${\mathrm{S}}^{2-}$-bearing sodium aluminate solution, the conversion of kaolinite to obtained Na_8.14(Al₆Si₆O₂₄)(SO₄)_1.14S_0.86 was not readily achieved, resulting in minimal incorporation of sulfur into the SDSP. Therefore, the sulfur contents in obtained SDSPs were different with the sulfur-containing ions species, resulting in the variations in the desulfurization efficiency of kaolinite. In comparison, kaolinite exhibited the highest desulfurization rate for ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ removal in sodium aluminate solution.

The micromorphology and micro area composition of DSPs obtained in different reaction conditions were further detected through SEM and EDS analyses. The SEM results are presented in Figure 8, and the composition analysis results of selected sites (1#–10#) using SEM-EDS in Figure 8 are listed in

Table 1. The SEM-EDS composition analysis of micro-areas at different sites in Figure 8.

Sites	Element Mass Fraction (wt%)				Na2O/SiO2	Al2O3/SiO2	Added Sulfur Compounds
	Al2O3	SiO2	Na2O	S
1#	20.62	39.27	38.14	-	0.53	1.03	None
2#	18.51	43.80	46.18	-	0.42	0.95
3#	20.34	38.25	43.84	0.12	0.53	0.87	Na2S·9H2O
4#	18.57	45.28	49.65	-	0.41	0.91
5#	22.90	32.81	35.42	2.98	0.70	0.93	Na2S2O3·5H2O
6#	20.46	31.26	36.64	3.51	0.65	0.85
7#	23.43	31.66	36.43	0.81	0.74	0.87	Na2SO3
8#	20.58	34.57	45.86	1.79	0.60	0.75
9#	21.05	29.37	34.16	1.84	0.72	0.86	Na2SO4
10#	22.51	29.58	31.65	0.45	0.76	0.93

.

Based on the results depicted in Figure 8a,b, it was evident that the morphology of the DSP was obtained from the reaction of kaolinite in sodium aluminate solution or ${\mathrm{S}}^{2-}$-bearing aluminum solution, both of which exhibited a lamellar stacking state. The EDS analysis results in Table 1 further confirmed the similarity in composition between site 1# and site 3#, as well as between site 2# and site 4#. The sulfur content at site 3# was only 0.12%, indicating a minimal involvement of sulfur in the reaction to form SDSPs in ${\mathrm{S}}^{2-}$-bearing sodium aluminate solution. Consequently, the effect of ${\mathrm{S}}^{2-}$ on the morphology of DSP formed from kaolinite reaction was inappreciable. The distinct mottled fragment morphology was observed on the surface of SDSPs obtained in the sodium aluminate solution contains ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$, ${\mathrm{SO}}_3^{2-}$ or ${\mathrm{SO}}_4^{2-}$, respectively. Additionally, the fragments on the surface of ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$-containing DSP (Figure 8c) and ${\mathrm{SO}}_4^{2-}$-containing DSP (Figure 8e) were more abundant, with elongated substances formed as well. Although the fragments were also found on the surface of the SDSPs obtained in ${\mathrm{SO}}_3^{2-}$-bearing sodium aluminate solution, the morphology was more similar to that of Figure 8a,b. Furthermore, sulfur contents were clearly detected at sites 5# to 10# (Table 1), especially at sites 5# and 6# obtained in ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$-bearing sodium aluminate solution, the sulfur content reached 2.98% and 3.51%, respectively, indicating that the desulfurization efficiency was significantly enhanced by generating ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$-containing SDSP. The results were consistent with the variation trend of desulfurization rate shown in Figure 4.

To further investigate the impact of generated SDSPs on alkali consumption under the Bayer process, the sodium content (expressed as mass percentage of Na₂O), silicon content (expressed as mass percentage of SiO₂), and sodium-to-silicon ratio (N/S, mass ratio of Na₂O to SiO₂) in the obtained DSPs under different conditions were analyzed. The corresponding results are presented in

Table 2. The contents of Na₂O and SiO₂ in DSPs.

	Added Sulfur Compounds	Na2O (wt%)	SiO2 (wt%)	Na2O/SiO2 (N/S)
1	None	19.46%	32.56%	0.60
2	Na2S·9H2O	18.32%	30.95%	0.59
3	Na2S2O3·5H2O	19.25%	31.98%	0.60
4	Na2SO3	19.36%	30.88%	0.63
5	Na2SO4	19.40%	33.32%	0.58

(ρ(Na₂O): 230.29 $\mathrm{g}·{\mathrm{L}}^{-1}$; α_k: 2.98; 260 °C, 1 h; dosage: kaolinite 100$\mathrm{g}·{\mathrm{L}}^{-1}$, 2#-Na₂S·9H₂O 37.50 $\mathrm{g}·{\mathrm{L}}^{-1}$, 3#-Na₂S₂O₃·5H₂O 19.40 $\mathrm{g}·{\mathrm{L}}^{-1}$, 4#-Na₂SO₃ 19.68 $\mathrm{g}·{\mathrm{L}}^{-1}$, 5#-Na₂SO₄ 22.18 $\mathrm{g}·{\mathrm{L}}^{-1}$).

.

From Table 2, both sulfur-free DSP and sulfur-containing DSPs obtained from kaolinite reacting with different sodium aluminate solutions exhibited a N/S of approximately 0.60. By comparing the molecular formulas of various SDSP, it can be inferred that the formed Na₈Al₆Si₆O₂₄S₂O₃ type SDSP exhibited superior sulfur removal efficiency under equivalent Na and Al consumption. Sulfur-containing compounds such as sodium sulfide or sodium thiosulfate can substitute the NaOH or NaAl(OH)₄ in common DSP during the desulfurization process. To sum up, desulfurization through the formation of SDSPs using kaolinite in sodium aluminate solution will not increase the overall alkali consumption of the system, which is one of the advantages of using kaolinite for sulfur removal.

4. Conclusions

(1) The thermodynamic calculations suggest that kaolinite can transform to SDSP in various sulfur-containing sodium aluminate solutions spontaneously. The experimental results demonstrated that kaolinite exhibited various desulfurization efficiency on different sulfur-containing ions under the simulated Bayer digestion process conditions at elevated temperature, with their desulfurization rates ranked in ascending order as ${\mathrm{S}}^{2-}$ < ${\mathrm{SO}}_3^{2-}$ < ${\mathrm{SO}}_4^{2-}$ < ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$.

(2) The reaction of kaolinite with ${\mathrm{S}}^{2-}$ and ${\mathrm{SO}}_3^{2-}$ yielded sodalite-type SDSP, while ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ and ${\mathrm{SO}}_4^{2-}$ promoted the formation of cancrinite-type SDSPs (Na₈Al₆Si₆O₂₄S₂O₃·2H₂O and Na₈Al₆Si₆O₂₄SO₄·3H₂O, respectively). Notably, the cancrinite-type SDSPs exhibited enhanced desulfurization efficiency. Increasing the kaolinite dosage, extending the reaction time, and elevating the reaction temperature all contribute to improving the desulfurization efficiency of kaolinite in ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$-bearing sodium aluminate solution.

(3) The desulfurization by transforming kaolinite to SDSP will not result in additional alkali consumption. The removal of ${\mathrm{S}}_2{\mathrm{O}}_3^{2-}$ was found to be more suitable to remove by forming SDSP, and the desulfurization rate exceeded 90% under optimized conditions.