Control of Silica Gel Formation in the Acidic Leaching of Calcium Aluminate Slags with Aqueous HCl for Al Extraction

Abstract

The present research article investigates the extraction of aluminum (Al) from an engineered CaO-Al₂O₃-SiO₂ slag by leaching with aqueous HCl under atmospheric pressure conditions. The slag is the by-product of an aluminothermic reduction process to produce metallurgical-grade silicon (Si) in a more sustainable way compared to conventional carbothermic reduction. One major challenge in the acidic leaching of aluminosilicate materials, like the slag treated in this study, is the possibility of SiO₂ gel formation during the leaching process. Extensive SiO₂ gel formation can make the separation of pregnant leach solution (PLS) from the leached residue impossible. Consequently, any acid leaching of aluminosilicate materials must be optimized for avoidance of these gelation phenomena. The present study first explores the leaching behavior of the calcium aluminosilicate slag in relation to the S/L ratio, with all other hydrometallurgical factors remaining stable (20.2% wt. HCl solution, 80 °C, optimized stirring rate), to determine at which value of this parameter SiO₂ gelation occurs. After determining the optimum S/L ratio for this system, an analysis of the behavior of Si in the PLS is presented, and the mechanism of SiO₂ gelation is explained based on critical assessment of these results against data provided from the scientific literature. It was found that the value of the pH of the PLS and the concentration of dissolved CaCl₂ and AlCl₃ are critical factors for the formation of filterable amorphous SiO₂. Under the optimum leaching conditions, PLS containing 37 g/L Al was obtained with concurrent avoidance of gelation phenomena.

1. Introduction

Metallothermic reduction is a chemical process whereby a reactive metal is used as a reducing agent for a less reactive metal oxide. Metallothermic reduction processes for production of metals are not uncommon, while other applications include the preparation of alloys and the production of nanostructured carbon, silicon, and others. A comprehensive review of the subject was recently performed by Z. Xing et al. [1].

Metallurgical applications based on metallothermic reduction processes are gaining renewed attention due to their potential to substitute the use of carbon and, consequently, reduce CO₂ emissions. Focus has been given to aluminothermic reduction processes, utilizing Al metal from a variety of secondary sources and wastes such as Al-dross and Al-scrap, thereby promoting more circular and sustainable value chains. For instance, H. Güney et al. studied the aluminothermic reduction of chromite concentrates for the production of FeCr [2], C. Stinn et al. the aluminothermic reduction of manganese sulfide to produce Mn master alloy, and J. Safarian and A. Kudyba et al. the aluminothermic reduction of MnO slags in process that could be tailored to produce Mn, AlMn, and FeMn [3,4].

The SisAl process applies the principles of aluminothermic reduction to demonstrate an alternative, more sustainable path to the energy- and carbon-intensive submerged arc furnace (SAF) process for producing Si alloys [5]. As shown in Figure 1, in the SisAl process Al metal (originating from dross, end-of-life scrap, etc.) reacts with a liquid CaO-SiO₂ slag to produce metallurgical-grade Si (MG-Si). The process can also be modified to produce 4N Si (high-purity silicon 99.99%) or an Al-Si alloy. In either case, a calcium aluminate slag (CA slag) is produced as the by-product of the reduction process.

Crucial for the sustainability of the SisAl process is the separation of CaO and Al₂O₃ contained in the CA slag. The goal is to recycle CaO to the CaO-SiO₂ slag-making step before the aluminothermic reduction and produce either metallurgical-grade Al₂O₃ (MG-Al₂O₃ or MGA) or high-purity Al₂O₃ (HP-Al₂O₃ or HPA). To achieve the separation of CaO and Al₂O₃, the CA slag undergoes hydrometallurgical treatment. Both an alkaline and an acid hydrometallurgical process route can be examined. This study is concerned only with the acidic route.

Efficient Al₂O₃ production from alternative resources is a research topic of high technological and strategic significance. MG-Al₂O₃ is almost universally produced by refining bauxite with the Bayer process [6]. Bauxite of acceptable quality for the Bayer process is found in specific geographic locations around the globe, mostly around the tropics. Consequently, the world’s primary aluminum production is dependent on the uninterrupted supply of this high-quality bauxite. Therefore, the production of Al₂O₃ from alternative primary or secondary raw materials by novel metallurgical technologies can potentially diversify the value chain of the primary aluminum industry and, at the same time, increase the strategic autonomy of non-bauxite-producing countries in Al metal. For this reason, the EU [7], USA [8], and Canada [9] have included bauxite and/or aluminum in their corresponding critical raw materials lists.

Both acid and alkaline methods for the extraction of aluminum from alternative raw materials are being investigated, most notably from coal fly ash [10,11,12] and aluminosilicate minerals, e.g., from kaolin or other clays [13,14,15,16,17]. Notably, HCl leaching of kaolinitic clays is close to achieving industrial application in Russia [18]. CA slags produced either from primary aluminum resources (e.g., low-grade bauxite or kaolin) [19,20,21,22] or industrial by-products (e.g., bauxite residue) [23] are processed exclusively by an alkaline route, which includes leaching with aqueous Na₂CO₃ and precipitation of Al(OH)₃ from the resulting PLS by CO₂ purging. The present research contributes to this gap in the literature by studying the processing of CA slags via an acid route.

Acidic processes are evaluated for Al extraction from aluminous raw materials when their SiO₂ concentration is unacceptably high for the Bayer process, favoring these materials due to the low solubility of SiO₂ in acid solutions. On the other hand, if Si in the raw material is found in silicates (e.g., Al-silicates, Ca-silicates etc.) that dissolve in the acid solution, the concentration of dissolved Si can exceed its equilibrium solubility, producing supersaturated solution. Under these circumstances, it polymerizes, and it can produce amorphous silica in colloidal form, which upon ageing can transform into a precipitate or a gel [24]. An essential review of the dissolution behavior of several common silicate compounds was performed by B. Terry [25,26]. The most common silicates encountered in calcium aluminate slags like those of the SisAl process (e.g., Ca₂SiO₄) are soluble in acid solutions and are termed as “gelatinizing” according to the aforementioned review; i.e., Si congruently dissolves with the other metals (Ca and Al) and tends to form supersaturated solutions.

Thus, the purpose of the present research work is dual: (1) to evaluate the effect of S/L ratio in the extraction of Al and Si from the slag and its contribution to the occurrence SiO₂ gelation and (2) to further investigate the conditions under which SiO₂ gelation occurs and, if possible, to define the parameters that contribute to this phenomenon. For this study, a 20.2% wt. HCl solution was used, while the other hydrometallurgical parameters have been defined in previous studies with similar slags and are as follows: leaching temperature 80 °C and stirring rate 300 rpm [27].

2. Materials and Methods

2.1. Laboratory/Analytical Equipment, Reagents, and Software Used in This Study

Leaching experiments were conducted in 0.25 L, round-shaped, borosilicate glass vessels. The reactor setup consisted of a mechanical stirring unit, a PTFE-coated stirrer, and a heating mantle, both controlled by an external PLC unit. A PTFE-coated thermocouple was immersed in the reactor and connected to the PLC unit for temperature control. A water-cooled borosilicate glass condenser was adjusted on the vessel lid for condensing vapors. The pH of the resulting solutions was measured by a Metrohm™ glass electrode (Type LL-Ecotrode Plus) and a Metrohm™ electronic pH-meter (Type 826 pH-Mobile). Ceramic Buchner funnels and Macherey Nagel (MN) filters (MN640 d and MN 615 ¼) were used for solid/liquid separation.

Leaching solutions were prepared with concentrated hydrochloric acid (37% HCl, 1.18 g·mL⁻¹) and distilled H₂O. Reagent-grade Li₂B₄O₇, LiBO₂, and KNO₃ were used for fusion of the solid samples. Concentrated nitric acid (70% HNO₃, 1.413 g·mL⁻¹) and distilled H₂O were used for the preparation of the aqueous solutions used in dissolving the fused samples for chemical analysis.

Quantitative elemental chemical analysis of wet samples was performed by atomic absorption spectroscopy (AAS) and inductively coupled plasma optical emission spectroscopy (ICP-OES). In more detail, AAS analysis was performed with a PerkinElmer™ PinAAcle 900T atomic absorption spectrometer. ICP-OES analysis was performed with a PerkinElmer™ Optima 800 optical emission spectrometer.

The mineralogical characterization of the materials was carried out by X-ray diffraction (XRD) on powder samples. The experiments were performed on an X’Pert Pro diffractometer (PANalytical, Almelo, The Netherlands) with CuKα radiation at 1200 W (30 mA and 40 kV), and the diffraction patterns were collected between 10 and 75° 2θ in 0.02° steps, with 2 s per step.

The XRD analysis included (a) the qualitative analysis using DIFFRAC. EVA V5.1 software (Bruker AXS, Karlsruhe, Germany) and the ICDD databases PDF-4+ 2022 and PDF-4 Minerals 2022 [28] and (b) the quantification of the raw material using DIFFRAC. TOPAS V6 software (Bruker AXS). In the content of this quantitative analysis, the indirect method of the external standard (NIST-SRM 640f Si powder) [29], as described by O’Connor and Raven [30], was also implemented to investigate the presence of any amorphous material.

2.2. Materials Used

The CA slag was supplied by the Department of Materials Science and Engineering of the Norwegian University of Science and Technology (NTNU). For the production of the slag, 810 g of aluminum dross (containing 72% wt. Al) was mixed with 1993 g CaO-SiO₂ slag in a graphite crucible. The mixture was heated to 1650 °C and remained at this temperature for 1h. Then, the crucible with its contents was allowed to cool freely to room temperature.

Prior to the leaching study, grinding, sampling, and characterization of the slag were performed. In more detail, a representative sample of the CA slag was ground to a particle size −100 μm (cumulative 100% mass of material passed through the −100 μm sieve) and then dried at 105 °C for at least 24 h. Dried samples were fused with a mixture of Li₂B₄O₇/LiBO₂/KNO₃ and then dissolved into 10% HNO₃ solution. The liquid samples were analyzed by AAS and/or ICP-OES to determine the slag’s composition in the major metal elements.

The chemical composition of raw material is shown in

Table 1. Chemical analysis of calcium aluminate slag used in this study.

Al2O3	CaO	SiO2	MgO	Fe2O3	Total
49.4%	37.5%	15.2%	0.3%	0.1%	102.5%

in equivalent oxide % wt., dry basis.

The results of the XRD analysis are presented in Figure 2 and

Table 2. Quantitative XRD analysis of the calcium aluminate slag used in this study.

CaAl2O4 (Krotite)	Ca2Al2SiO7 (Gehlenite)	Ca12Al14O33 (Mayenite)	Si (Silicon)	C (Graphite)	SiC (Moissanite 3C)	Amorphous Content	Rwp 1
32.6%	13.6%	28.6%	1.3%	0.4%	3.7	21.3%	6.82%

¹ R_wp or the R-weighted pattern corresponds to the statistical performance of the quantitative analysis.

. According to our findings, the main constituents are krotite (CaAl₂O₄), mayenite (Ca₁₂Al₁₄O₃₃), gehlenite (Ca₂Al₂SiO₇), and amorphous material. Furthermore, silicon metal (Si) and moissanite-3C (SiC) are present in minor quantities, while a trace amount of graphite (C) is detected.

The oxidation of these Si impurities is probably responsible for the increased equivalent SiO₂ concentration reported in the chemical analysis of the slag (Table 1), resulting in a sum of equivalent oxides slightly exceeding 100%.

2.3. Experimental Methodology

As mentioned earlier, the leachability of calcium aluminate slags in HCl solutions was demonstrated by our previous work [27]. The simplified chemical equation showing the dissolution of CaAl₂O₄ in aqueous HCl is given in Equation (1).

$${\mathrm{C}\mathrm{a}\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_{4\left(\mathrm{s}\right)}+8{\mathrm{H}\mathrm{C}\mathrm{l}}_{\left(\mathrm{a}\mathrm{q}\right)}\to {\mathrm{C}\mathrm{a}\mathrm{C}\mathrm{l}}_{2\left(\mathrm{a}\mathrm{q}\right)}+{2\mathrm{A}\mathrm{l}\mathrm{C}\mathrm{l}}_{3\left(\mathrm{a}\mathrm{q}\right)}+{4\mathrm{H}}_2\mathrm{O}$$

(1)

A corresponding chemical equation can be provided for mayenite (Ca₁₂Al₁₄O₃₃), leading to the same products. Consequently, the process of leaching aluminum from a calcium aluminate slag is limited either by the solubility of aluminum chloride in the PLS or, as mentioned earlier, by the formation of a silica gel.

This study aims to explore the maximum amount of slag that can be leached in a single-stage leaching process under atmospheric pressure conditions while simultaneously avoiding silica gel formation in the PLS. In

Table 3. Experimental conditions applied in this leaching study.

S/L Ratio (%)	Temperature (°C)	Agitation (rpm)	(HCl) (%w/w)	Volume of Solution (mL)	Duration (h)
10	80	300	20.2	150	2
12.5
15
20

, the experimental conditions applied in this study are summarized. All tests were performed in three replicates to ensure reproducibility of the results.

The following experimental procedure was applied to all tests. The leaching solution was prepared by mixing deionized H₂O and concentrated HCl (37% w/w). First, 0.15 L of the leaching solution was obtained and transferred to the reactor. The reactor was sealed, and mechanical stirring was applied at a rate of 300 rpm. The solution was preheated to 80 °C, and the slag was inserted when the set temperature was reached, which did not fluctuate more than ±2 °C. The insertion of the slag in the reactor marked the start of the leaching test. If a sample needed to be drawn during the experiment, stirring was momentarily stopped and reinitiated after the sample was extracted. The leaching test was concluded when its predetermined duration was reached. At this stage, stirring and heating were turned off, the reactor lid was removed, and the pulp was filtered to separate the PLS from the residue. Afterwards, the PLS was allowed to cool freely to room temperature and then transferred to a 0.25 L borosilicate volumetric flask. The volume of the solution was adjusted to the standard volume of the flask. Subsequently, the solution was transferred into a vessel for storage, and samples were drawn for chemical analysis. The leaching residue was washed with deionized H₂O. The solution acquired from the washing process was also transferred to a 0.25 L volumetric flask, its volume adjusted to the standard volume of the flask, and a sample drawn for chemical analysis. The washed residue paste (the filter cake) was then placed in a drying oven and allowed to dry for 24 h at 105 °C. Afterward, a sample of the dried residue was stored for the XRD analysis.

The % wt. extraction of a metal (Me = Al, Ca, Si, Mg, Fe) in the PLS and the washing solutions is defined as the mass of each metal dissolved in solution divided by its corresponding mass contained in the slag before leaching, as described by Equation (2).

$$\% \mathrm{w}\mathrm{t}. \mathrm{M}\mathrm{e} \mathrm{e}\mathrm{x}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}=\frac{m_{\mathrm{M}\mathrm{e} \mathrm{d}\mathrm{i}\mathrm{s}\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{v}\mathrm{e}\mathrm{d}}}{m_{\mathrm{M}\mathrm{e} \mathrm{s}\mathrm{l}\mathrm{a}\mathrm{g}}}\times 100$$

(2)

where m_{Me dissolved} is the mass of a metal dissolved in solution as defined by the chemical analysis of the solution, while m_{Me slag} its mass in the amount of slag that underwent leaching. Finally, the overall % wt. extraction for each metal is determined as the sum of its % wt. extraction in the PLS and its % wt. extraction in the washing solution.

3. Results

3.1. Metal Extraction Results

The extraction curves for Al, Ca, Mg, and Fe as a function of the S/L ratio are shown in Figure 3a. The corresponding curve for Si is shown in Figure 3b. The concentration of each metal in the PLS is shown in

Table 4. Filterability, average metal concentrations (at 25 °C), and average % wt. metal extraction in the PLS at the different S/L ratios tested.

	Filterability	Average Metal Concentration in PLS					Average % wt. Metal Extraction in PLS
S/L (%)	Silica Gel Observed	Ca (g/l)	Al (g/l)	Si (mg/L)	Fe (mg/L)	Mg (mg/L)	Ca	Al	Si	Fe	Mg
10	No	28.9	26.4	53	13	272	88.9%	82.7%	0.6%	72.5%	77.2%
12.5	No	32.6	31.0	129	18	248	95.3%	87.9%	1.1%	67.6%	77.9%
15	No	40.1	34.5	299	14	275	86.4%	76.5%	3.2%	47.8%	58.5%
20	Yes	58.5	45.3	3400	19	402	97.4%	80.2%	21.4%	45.0%	63.5%

. Moreover, in Table 4, the observation or lack thereof of silica gelation phenomena is reported. Finally, when SiO₂ gelation was observed, complete filtration of the residue was not feasible.

3.2. Behavior of Dissolved Si in Relation to the pH

In Figure 4, the logarithm of Si concentration (in mmol/L) measured at each S/L ratio value is depicted (left-hand side axis) as well as the corresponding variation of pH (right-hand side axis). Finally, in

Table 5. The concentration of dissolved Si in PLS in mg/L, in mmol/L, and the logarithm of the latter and the pH of the PLS in each S/L ratio tested.

S/L (%)	Si (mg/L)	Si (mmol/L)	Log Si (mmol/L)	pH of PLS
10	53	1.9	0.27	0.01
12.5	129	4.6	0.66	0.15
15	299	106.5	2.03	1.35
20	3400	121.1	2.08	1.98

, the data used for the curves shown in Figure 4 are presented.

3.3. Mineralogical and Chemical Analysis of Residues

The mineralogical analysis of the residues produced at 10%, 15%, and 20% S/L are presented in Figure 5a–c, respectively. The analyses of the residues produced in the 12.5% S/L tests are omitted, as they have similar diffraction patterns to the ones of the 10% S/L tests. The qualitative XRD analysis of the residues reveals that the pure calcium aluminates, i.e., krotite and mayenite, were completely dissolved after 2 h of leaching in all residues. On the other hand, gehlenite was still observed in the 10% and 15% S/L ratio tests, indicating that its leaching kinetics are slower than the corresponding kinetics of the pure calcium aluminates. Similarly, the Si metal and the polymorphs of SiC were not dissolved in any of the leaching tests. It is noted that two polymorphs of SiC, namely moissanite-3C and moissanite-2H, were detected in the leaching products. The latter was not evident in the raw material (Figure 2) due to the detection limit of the diffractometer used in this study.

Another important finding of the XRD analysis is the increased amount of amorphous phases observed in the residues with increasing S/L ratio. This can be attributed to the formation of amorphous SiO₂, which is further discussed in the following Section 4.2.

Finally, since the value of S/L ratio found to produce non-gelatinizing solutions was determined at 15%, the residues from these tests were chemically analyzed. The average chemical composition of the residue in equivalent oxide basis, including the determination of LOI, is shown in

Table 6. Chemical analysis of the residue of the 15% S/L tests.

Al2O3	CaO	SiO2	MgO	Fe2O3	LOI	Total
12.9%	9.3%	36.4%	0.1%	0.3%	41.8%	100.8%

. The residues, as expected, are predominantly composed of SiO₂. The Al₂O₃ leaching ratio is estimated at 87.1%. The presence of CaO and Al₂O₃ indicates that gehlenite was indeed not completely leached and verifies the incomplete % wt. extraction results.

4. Discussion

4.1. Leaching of Aluminium and Behavior of the Impurities

The extraction curves shown in Figure 3 and the corresponding % wt. extraction results presented in Table 4 provide evidence that the extraction of Al from the CA slag used in this study is efficient across the range of S/L ratio values tested. Under all conditions examined, more than 80% of the Al was extracted from the slag. The mineralogical analysis of the residues (Figure 5a–c) reveals that the aforementioned leaching efficiency is owed to the complete leaching of the pure calcium aluminate phases (krotite and mayenite).

As expected, high % wt. extraction values were also observed for Ca, exceeding 85% across the range of S/L values tested in this study. The non-complete extraction of Al and Ca in this study can be attributed to the presence of gehlenite. Again, the analysis of the leached residues (Figure 5a–c) reveals the presence of gehlenite in all residues up to 15% S/L. This is an indication that gehlenite is probably also present in the residues of the 20% S/L ratio tests, but its major peaks are masked under the increased background caused by the amorphous phases of the precipitate. Overall, the PLS produced in this study contain almost equal concentrations of Ca and Al. Consequently, the subsequent separation of Ca from the leaching solution and the recycling of the corresponding HCl consumed for its dissolution are critical factors that need to be addressed in future studies.

Probably the most interesting observation is the value of % wt. Al extraction of the 20% S/L ratio tests. This is the value of the S/L ratio for which extensive SiO₂ gelation was observed in the PLS. This indicates that the leaching of the CA slag in aqueous HCl could be performed at even higher S/L ratio if SiO₂ gelation phenomena were not observed. Consequently, the limiting factor in this hydrometallurgical extraction process is the gelation of SiO₂ in the PLS, and leaching conditions will inevitably be dictated by this phenomenon.

Concerning the dissolution of impurities, according to the extraction curves of Figure 3a,b, these can be divided into three groups: (a) Ca, (b) Mg and Fe, and (c) Si. The case of Ca was already addressed earlier in this section. The second group of impurities, Mg and Fe, behave similarly with increasing S/L ratio, as shown in Figure 3a. Their % wt. extraction values are lower than those of Ca and Al, and they decrease with increasing S/L ratio. Investigation of the behavior of these impurities is beyond the scope of the present research that focuses on the behavior of Al and Si and will be treated in future optimization studies of the process. Finally, the following section focuses on the behavior of Si, which is the most important impurity.

4.2. Silica Gelation Phenomena and Mechanism

Figure 3b and Figure 4, along with the data presented in Table 5, reveal some key features about the occurrence of SiO₂ gelation in this system. Firstly, increasing the S/L ratio from 15% to 20% results in an eruptive increase of the dissolved Si from ≈300 mg/L to more than ≈3.4 g/L (Figure 3b and Table 5). This increase is accompanied by the formation of the SiO₂ gel. At the same time, the pH of the PLS increases to ≈2 in the 20% S/L tests. Overall, in all tests performed up to 15% S/L, where no silica gelation was observed, the pH value of the solution was well below 2 (Table 5).

At pH = 2, the isoelectric point (IEP) of SiO₂ is found. Moreover, the precipitation behavior of SiO₂ in aqueous systems greatly depends on this property. Studies in the literature suggest that the agglomeration and precipitation paths of SiO₂ in aqueous solutions with pH values below the IEP are different compared to those observed in solutions with pH values above the IEP. In more detail, E.A. Gorrepati et al. showed that in acidic solutions with pH values <2, SiO₂ particle formation and growth occurs in two stages [31]. First, mono-silicic acid disappears rapidly through a dimerization reaction, forming nanoparticles. These primary particles flocculate, and the mean floc diameter increases exponentially with time, leading to the precipitation of amorphous SiO₂ particles. Both the rate of monomer disappearance and the flocculation rate increase rapidly with increasing HCl concentration. Moreover, the same study showed that the addition of chloride salts accelerates both stages of this precipitation. The acceleration of the flocculation rate of SiO₂ nanoparticles depends on the type of chloride salt added, in the following order (starting with the salt that showed the highest acceleration and the rest in decreasing order):

AlCl₃ > CaCl₂ > MgCl₂ > NaCl > CsCl > no salt

A. Lazaro et al. [32] confirmed the observation of Gorrepati et al. and, in addition, developed a model for silica particle growth below the isoelectric point.

During the leaching of the CA slag performed in this study, SiO₂ gelation did not occur as long as the pH of the PLS was <2 (S/L ratio values from 10% to 15%). Extensive gelation was observed only when the pH of the PLS approached the value of the pH at the IEP of SiO₂ (pH = 2). That condition was met in 20% S/L tests. Finally, as shown by Equation (2), AlCl_3(aq) and CaCl_2(aq) are produced during the leaching process investigated. However, according to the findings of E.A. Gorrepati et al. that were mentioned earlier, these salts accelerate primary SiO₂ particle flocculation the most.

Thus, to confirm the SiO₂ particle agglomeration mechanism, two residue samples from the 15% S/L test were analyzed by laser diffraction particle size analysis. The first residue was obtained at 90 min of leaching and the second at 120 min, i.e., at the end of the leaching test. The results of the particle size analysis are shown in Figure 6. It can be observed that after 90 min of leaching (solid, orange-colored curve), there are still nanometer-sized SiO₂ particles present. These correspond to the primary amorphous SiO₂ particles suggested by the model proposed by E.A. Gorrepati et al. The molar concentration of the solution in equivalent AlCl₃ and CaCl₂ is 1.3M and 1.0M, respectively. As suggested by E.A. Gorrepati et al.’s model, the presence of these chloride salts accelerates the flocculation rate of the primary particles. Indeed, already in 90 min of leaching, the majority of the particles were between 1 μm and 10 μm in diameter, and larger particles over 50 μm in diameter had already formed. Furthermore, according to the model described earlier, the presence of these salts should accelerate the growth of the SiO₂ particles, which was clearly observable in the residues after 120 min of leaching (solid, blue-colored curve). Indeed, the primary submicron particles were almost eliminated, and the majority of particles were between 10–100 μm in diameter. The measured mean particle size, D(v,0.5), increased from 5.21 μm to 27.51 μm when the duration of leaching was increased from 90 to 120 min.

In the light of these findings, it was ascertained that SiO₂ gelation during the leaching of CA slags in aqueous HCl is directly associated with the pH of the PLS. When the pH is below the IEP of SiO₂, i.e., pH < 2, no SiO₂ gelation phenomena are observed, and the mean diameter of particles increases fast due to the presence of dissolved AlCl₃ and CaCl₂. Under these circumstances, a residue with good filtration properties is produced. When the IEP is reached at ≈20% S/L ratio, SiO₂ gelation is the dominant precipitation mechanism, and the resulting residue is unfilterable.

5. Conclusions

This study has shown that the leaching of calcium aluminate slags with aqueous HCl for Al extraction is a promising metallurgical process that can pave the way for introducing a sustainable approach of the SisAl process for the co-production of metallurgical-grade Si and Al₂O₃. In more detail, the following conclusions can be drawn from the research results:

• The crucial design parameter for successful leaching is the avoidance of SiO₂ gelation. In the course of the experimental work, it was shown that efficient leaching of Al was achieved from the slag, even at conditions that favored SiO₂ gelation. Leaching with a 15% S/L ratio led to the production of stable, non-gelatinizing PLS containing 34.5 g/L Al;
• Concerning the composition of slags suitable for this process, the phase of gehlenite must be avoided, as its leaching kinetics in aqueous HCl solutions appear to be slower compared to the leaching of pure calcium aluminates;
• SiO₂ gelation can be controlled by monitoring the pH of the PLS. Non-gelatinizing residues are produced for pH < 2. This phenomenon was attributed to the behavior of SiO₂ in aqueous solutions with pH values below its isoelectric point (IEP);
• When the pH of the PLS at the end of the leaching process is <2, the dissolved CaCl₂ and AlCl₃ salts accelerate the particle growth of amorphous SiO₂ particles, leading to the formation of a filterable leached residue. In this study, during leaching with a 20.2% w/w HCl solution and a 15% S/L ratio at 80 °C, an increase in the D50of the amorphous SiO₂ particles was observed from 5.21 μm to 27.51 μm when the duration of leaching increased from 90 min to 120 min.