Investigating the Effect of Reaction Temperature on the Extraction of Calcium from Ironmaking Slag: A Kinetics Study ^†

Abstract

Beneficiation of iron and steel slags have a great potential in mitigating high levels of anthropogenic carbon dioxide (CO₂) gas emissions owing to their high contents of calcium and magnesium. Carbonation of calcium and magnesium leaching solutions produce precipitates that promise to act as fluxing agents in various industries. There is also a need to study the physico-chemical factors affecting the extraction of these elements and their kinetic behaviour. A calcium-rich blast furnace ironmaking slag milled to 80%—75 µm was used in an experiment to study the effect of reaction temperature on the dissolution of calcium. An orbital shaking incubator was used by applying the one-factor-at-a-time method for a 2 h leaching test. Reaction temperature was controlled at 30 °C, 50 °C, 70 °C, and 100 °C at constant agitation speed of 200 rpm, 0.5 M of lixiviant concentration and solid-to-liquid ratio kept at 200 g/L. CH3COOH, (NH₄)₂SO₄, and NH₄Cl were tested individually as lixiviants. AAS, XRD and EDX were used to analyse the leach filtrate and residue. Consequently, the findings showed that temperature is the critical operating parameter that yielded exceptionally high calcium extraction. CH₃COOH yielded 80–90% of calcium extracted under high reaction temperatures. The kinetic data for the effect of reaction temperature leaching data fitted one shrinking core model equation, which described diffusion-controlled leaching reactions. The diffusion controlled by a product layer model was fitted and the activation energy were calculated as 19.4930 kJ/mol for NH4Cl solvent. (NH₄)₂SO₄ and CH₃COOH did not fit into any model because of undefined kinetic data and formation of gelatinous silica layer during leaching process.

1. Introduction

The steel industry is globally the largest energy consuming manufacturing sector and accounts for 7–12% of anthropogenic greenhouse gas emissions [1]. Measures are hence being developed to improve energy efficiency and reduce CO₂ emissions such as scrap steel recycling, but more drastic measures are needed if reduction goals are to be met in a reasonable timeframe [2]. During manufacturing of iron and steel, a significant amount of about 10–15% by mass of slag is produced [3]. These materials are made up mainly of CaO, SiO₂, MgO, Al₂O₃, and MnO [4]. The main mineral phases in steel slags are dicalcium and tricalcium silicates, calcium aluminium silicates, calcium oxides, and periclase.

Central to the understanding of leaching kinetics are the shrinking core models, namely, shrinking core-shrinking particle, shrinking core-constant particle size, and shrinking particle, which are regarded as the first types of models developed to describe better heterogeneous and non-catalytic reaction kinetics [5,6]. According to [7], for chemically controlled reactions the model can be depicted as follows:

$$1-{\left(1-X\right)}^{\frac{1}{3}}=\frac{K_C{M}_B{C}_A}{{\rho }_B\alpha r_0}t=k_{r}t$$

(1)

Consequently, if the kinetics determined by the model seems to be diffusion-controlled, then the equation changes as follows:

$$1-\frac{2}{3}X-{\left(1-X\right)}^{\frac{2}{3}}=\frac{2M_{B}D{C}_A}{{\rho }_B\alpha r_0}t=k_{r}t$$

(2)

where $X=\frac{C}{C_0}=$ Fractional conversion at a given temperature; $K=\frac{K_s{C}_A}{R_{0 }\rho }=$ The apparent rate constant (s⁻¹); $K_S=$ Rate constant of the reaction; ${\rho }_B=$ Density of the metal extracted (g/mL); $r_0=$ Radius of the unreacted particle (m); $C_{A }=$ Initial concentration of leachant (mol/mL); $M_B=$ Molecular mass of metal extracted (g/mol); D = Diffusion coefficient of the mineral species

The authors in [7] also proposed a shrinking core model for solid–liquid controlled leaching reactions. The mathematical expression of the mixed leaching mechanisms is given as [7]:

$$\frac{1}{3}\ln \left(1-X\right)+{(1-X)}^{-\frac{1}{3}}-1=kt$$

(3)

Reaction rates are significantly promoted by temperature [8,9]. The most pronounced observations resulting from this variable is that the temperature must be controlled at all levels if they are to produce reliable kinetic experiments. In addition, a serious consideration of temperature provides crucial information about reaction interaction and mechanisms. Studies by Van’t Hoff and Arrhenius were the starting point for all theories of the temperature dependence of reaction rate constants. These observations indicate a close relationship between reaction kinetics and equilibrium constants. The temperature dependence of the equilibrium constant K differs with temperature T, is clearly depicted by the Van’t Hoff equation, as indicated in Equations (4)–(7):

$$\frac{dlnK}{dT}=\frac{\Delta H^0}{R{T}^2}$$

(4)

where R = Gas constant (8.314 J/K/mol); ∆H° = Standard enthalpy change in the reaction; K = Ratio $\frac{K_1}{K_{-1}}$ of the rate constants K₁ and K₋₁ for the forward and reverse reactions.

Similarly,

$$\frac{dlnK}{dT}=\frac{E_a}{R{T}^2}$$

(5)

where E_a represents the activation energy which corresponds to the standard enthalpy of reaction ∆H° in the Van’t Hoff equation. The integral of both sides of the equation with respect to temperature T gives the following product.

$$lnK=lnA-\frac{E_a}{RT}$$

(6)

where $lnA$ is constant of integration with respect to T. The reaction rate constant K is then depicted as follows:

$$K=A{e}^{-\raisebox{1ex}{$E_a$}\!\left/ \!\raisebox{-1ex}{$RT$}\right.}$$

(7)

A graphical representation of $lnK$ against the inverse of temperature $\left(\frac{1}{T}\right)$ is a linear function that determines the slope$-\raisebox{1ex}{$E_a$}\!\left/ \!\raisebox{-1ex}{$R$}\right.$. Diffusion controlled reactions have an approximate activation energy of less than 21 kJ/mole while reaction and intermediate controlled mechanisms have approximate activation energies of 21–40 kJ/mole and 40–100 kJ/mole, respectively [10,11,12].

Authors who have done work on the topic of extraction of calcium from ironmaking slag have highlighted the need for a fundamental insight with regards to mechanisms that govern the leaching of calcium from iron and steel slag [13,14,15]. Currently, there is little research information available on the impact of varying physical and chemical parameters during dissolution process of calcium from steel slags. The main objective of this study is to investigate the influence of temperature on the extraction of calcium from the calcium-rich ironmaking slag and thereby establish the reaction-controlling steps during leaching of calcium by using two shrinking core models (SCMs).

2. Materials and Methods

2.1. Materials

A conventional leaching test work was conducted using well-sealed Erlenmeyer flasks rotated in a portable and adjustable orbital shaking incubator. The equipment has the following features: temperature controller, shaker speedometer, and shaker timer. Other equipment such as a small-scale rod mill, steel rods, a Jones Riffler sampler, a filter press, and an air-compressor were also used. The reagents used for this study include ammonium chloride (NH₄Cl), ammonium sulphate ((NH₄)₂SO₄), and acetic acid (CH₃COOH).

2.2. Experimental Procedure

A piece of blast furnace ironmaking slag, 20 mm in size, was blended using a Jones Riffler sampler followed by the dry milling process. The particle size reduction of the sample was achieved using a laboratory scale rod mill for slag particle size reduction down to 80% passing 75 µm. Leaching tests were conducted in 250 mL Erlenmeyer flasks in a portable orbital shaking incubator. For the extraction tests, the slurry was prepared by pouring 20 g of air-dried blast furnace slag powder into the 100 mL pre-conditioned lixiviant solution. The leach slurry agitation rate was maintained at 200 rpm for a period of two hours at temperatures of 30 °C, 50 °C, 70 °C, and 100 °C. The slurry was then filtered through a 0.2 µm membrane filter placed on the air-supplied filter press. In addition, the amount of filtrate was placed in the 200 mL Erlenmeyer flask for analysis. Approximately 30 mL of filtrate was stored in a 50 mL polyethylene bottle and diluted using a 100× dilution factor for chemical analysis. The filter cake on the membrane filter was dried for 24 h at 80 °C and stored in a vacuum desiccator for material characterisation. A 200 mL bottle of distilled water was used to dilute 2 mL of the filtrate for each lixiviant and their corresponding concentrations. Then, 10 mL of the diluted leach liquor from each lixiviant was used in determining the after-leach results using the atomic absorption spectrometry) (AAS) analyser.

2.3. Characterisation Techniques

2.3.1. Atomic Absorption Spectroscopy (AAS)

Leaching liquor samples were analysed by Varian AAS (AA140) to identify the chemical composition of the leached metals. Standard samples of calcium were prepared at concentrations of 5 ppm, 10 ppm, 20 ppm, and 40 ppm in order to identify whether analysis made are in correspondence. Acetylene was used as the fuel for the burner and air as the supporting oxidant. Operational conditions for analysis of calcium were set accordingly. Calcium cathode lambs were also used to detect the concentration of ion metals in the samples. Highly concentrated samples were diluted to be able to be read within the maximum range of the AAS.

2.3.2. X-ray Diffractometer (XRD)

A Bruker D8 advance powder X-ray diffractometer (XRD) was used to analyse the crystallographic structure of the blast furnace slag using Ca filtered Cu Kα radiation at 40 kV and 40 mA. The specimens were scanned at a scan speed of 0.1 s/step in the 2$\theta$ range of 5 to 100° at ambient temperature. The phases were identified using X Pert High Score Plus software.

2.3.3. Energy-Dispersive X-ray Spectrometry (EDX)

Leach residues were analysed by using an EDX spectrometry. EDX is a chemical micro-analysis technique normally used in together with Scanning Electron Microscopy (SEM). The components of this instrument include a sensitive X-ray detector, a liquid nitrogen is for cooling and a suitable software to collect and provide the readings for the energy spectra.

3. Results

The percentage of calcium extracted increased from about 40% at 30°C to about 90% at 100 °C when ammonium chloride was used as the lixiviant (Figure 1). Similarly, with acetic acid the percentage calcium extracted increased from about 60% at 30 °C to about 100% at 100 °C (Figure 1). On the other hand, with ammonium sulphate, calcium extraction remained low (<20%) at all temperatures.

Figure 2 represents residue XRD patterns for 0.5 M NH₄Cl, CH₃COOH, and (NH₄)₂SO₄ leached at 100 °C. The final characterisation results show that most of the calcium present is in the form of an aluminoakermanite ((Ca,Na)₂(Al,Mg,Fe²⁺)) phase and not composed of the majority of calcium silicates found in previous studies [16]. There are some traces of metallic nickel as well as niobium and titanium enclosed in the form of NbTi(OH)₃. All three lixiviants were found to have similar minerals encapsulated in their leach residue but at different quantities as indicated by the peak intensities varying from one lixiviant to another. When CH₃COOH was tested at 0.5 M ionic strength it was found that the aluminoakermanite had lower intensity than those of NH₄Cl and (NH₄)₂SO₄.

EDX spectrum analysis was performed on (NH₄)₂SO₄, NH₄Cl and CH₃COOH slag residue samples of 0.5 M leached at 100 °C. Figure 3 shows that CH₃COOH solvent dissolved calcium significantly as there was only traces of calcium (Ca) in the slag residue. Other peaks shown in Figure 3 are those of potassium (K), manganese (Mn), oxygen (O), magnesium (Mg), aluminium (Al), silicon (Si), lead (Pb), and phosphorus (P). Figure 4 represents the EDX spectrum of slag residue from the NH₄Cl 0.5 M solution leached at 100 °C. The slag residue consists of calcium occurring in three peaks with varying spectrum counts. In addition, there are also elements such as K, Mn, O, Mg, Al, Si, and Pb. Ammonium sulphate ((NH₄)₂SO₄) was found to be less efficient in dissolving calcium from its bearing slag in all variables experimented in this study. Figure 5 displays slag residue of (NH₄)₂SO₄ at 0.5 M leached at 100 °C. The EDX spectrum peaks for (NH₄)₂SO₄ are similar with NH₄Cl with slightly different peaks of various elements.

Pearson correlation coefficients (R²) for NH₄Cl data are displayed in Figure 6 and Figure 7. It was observed that the increase in R² was influenced by the increase in reaction temperature for both chemical and diffusion-controlled mechanisms. There was a significant increase of R² from 30 to 50 °C for both mechanisms and an insignificant drop in the correlation coefficient as the temperature further increased to 100 °C. The results on R² when NH₄Cl was tested as a solvent do not fully concur with those of [17] because there is a sudden drop on the Pearson correlation coefficients at temperatures above 50 °C. Although, there is no concurrence on temperatures above 50 °C particularly on Blast Furnace (BF) ironmaking slag. This may be because the slag behaves differently from natural ores that were previously studied.

Figure 8 represents the Arrhenius plot for chemical reaction-controlled mechanism in NH₄Cl solvent. The Pearson correlation coefficient (R²) for the chemically controlled mechanism is 0.126. The value of R² that is far less than 0.9 are considered to be indicative of an activation energy that does not correlate with any of the shrinking core models [18]. The calculated activation energies (E_a) from the Arrhenius graphs, where chemical reaction controlled and diffusion-controlled models were plotted and were calculated as 4.8504 kJ/mol and 19.4930 kJ/mol, respectively. The R² of 0.126 and an activation energy of 4.8504 kJ/mol confirmed that chemically controlled model does not influence the reaction on the dissolution of calcium from ironmaking slag.

Figure 9 is an Arrhenius plot for diffusion-controlled mechanism. The R² value of 0.9701 and the activation energy of 19.4930 kJ/mol confirmed that the rate-controlling step of NH₄Cl solvent is diffusion by product layer.

Pearson correlation coefficients (R²) for (NH₄)₂SO₄ solvent data are shown on Figure 10 and Figure 11. Both chemical reaction and diffusion-controlled mechanisms showed a consistence with the increase in R² as the reaction temperature increased. At 30 °C no reaction mechanism was taking place as there was no R² value for all the reaction data fitted in chemical reaction, diffusion and mixed mechanism. The increase in reaction temperature from 50 to 100 °C resulted in a constant increase in R² for both chemical reaction and diffusion-controlled models.

In Figure 12, the activation energy that was determined by chemically controlled mechanism was 2.1234 kJ mol⁻¹ and it was found to be below the range of any model fitted. In addition, both R²-value from the Arrhenius plot as well as the activation energy confirm that the dissolution of calcium by (NH₄)₂SO₄ is not controlled by the chemical reaction. Furthermore, the Arrhenius plot for diffusion-controlled mechanism was graphically undefined.

Pearson correlation coefficients (R²) for the dissolution rates of calcium from ironmaking slag with CH₃COOH used as a solvent were obtained from the data depicted in Figure 13 and Figure 14. The reaction temperature is the most important variable in chemical reactions. There is a trend in the improvement of R² with increase in reaction temperature from 30 to 70 °C for both chemical and diffusion-controlled mechanisms, but the values of R² dropped significantly at 100 °C. The drop in R² at 100 °C is attributed to the formation of silica gel that was reported from experiments where CH₃COOH solvents were used at high temperatures [19].

4. Discussion

4.1. The Effect of Reaction Temperature on the Extraction Calcium from Ironmaking Slag

Figure 1 displays the influence of increasing the temperature of a leaching incubator to 30 °C to observe the change in calcium recovery. All three lixiviants possessed linear curves at different standpoints of calcium recovered in a 30 min interval leaching reaction of a maximum of 120 min. (NH₄)₂SO₄ could extract only 18% of calcium in a period of 120 min while NH₄Cl reached a peak at 42%. Consequently, CH₃COOH and NH₄Cl had a significant effect on the dissolution of calcium at a reaction temperature of 30 °C and achieving a peak point at 62%.

Regardless of the lixiviant ionic strength, calcium recovery is altered significantly by acidity of the lixiviant [20]. A weak base performed exceptionally well compared to basic ammonium salts. This is supported by the similarities in calcium recovery in this experimental work and the literature. Acidic leaching environments seemed to give a promise of dissolving calcium, irrespective of residence time, ionic strength, or solid–liquid ratio [21]. In addition, dissolution of calcium by (NH₄)₂SO₄ and NH₄Cl as lixiviants is possible at low temperatures (e.g., 30 °C).

NH₄Cl could attain reasonable calcium extraction percentage with increasing dissolution temperatures. (NH₄)₂SO₄ has shown that it remains the weakest among the selected lixiviants at 30 °C. Several studies were conducted at temperature ranges above the standard state, and it has been shown that temperature influences the extraction efficiencies, particularly on BOS, SS, and HMD slags. The temperature of range 30 to 70 °C with aqueous solutions of ammonium acetate and ammonium chloride increases the calcium dissolution efficiency with increasing temperatures. In addition, calcium extraction from steel converter and HMD slags from ladle slag were tested with using acetic acid and ammonium salts at 30 °C [22]. The present study, using blast furnace slag has shown similarities in calcium dissolution with other types of feedstocks reported in the literature [22]. In all cases calcium leaching was greatly influenced by the chemical composition of slag and enhanced by temperature increase.

A 50 °C leach temperature is compared with the calcium recovery from calcium-rich slag. The rate at which the reaction temperature was increased resulted in a significant decrease in calcium dissolution by (NH₄)₂SO₄ as it resulted in 18% of calcium recovery after 120 min of leaching time. On the other hand, CH₃COOH and NH₄Cl achieved 75% recovery after 90 min of leaching time. [22] experimented with temperature as the factor affecting the reaction rate under the same condition and an average of 65% and 60% calcium recovery was achieved respectively when using CH₃COOH as the solvent. Similarly, the present study deduced a potential of about 76% calcium recovered by CH₃COOH after 120 min of leaching reaction suggesting that the current method was more efficient.

Correlations of calcium extraction capability at 70 °C were reported by [23]. Both studies achieved an average calcium recovery of 70% under similar conditions to the present study. These studies were aimed at determining the influence of weak acid CH₃COOH on the dissolution rate of calcium from its feedstock. With temperature increase, the rate of calcium dissolution capability for (NH₄)₂SO₄ decreased significantly over the entire leaching period (Figure 1). This trend was continuous from lower temperatures of 30 °C and 50 °C, which further supports the contention that sulphate ions weaken with increasing reaction temperature. As for NH₄Cl, the calcium recovery increased significantly as the reaction temperature was elevated. As a result, the increase in temperature has a positive influence in NH₄Cl to dissolve calcium over the entire 30 min interval-leaching period of the maximum residence time of 120 min. Meanwhile, elevated incubator temperatures have shown similar dissolution trends in the range of 50 to 70 °C for both NH₄Cl and CH₃COOH. In addition, the NH₄Cl became active in dissolving calcium from blast furnace slag earlier than CH₃COOH especially after 90 min of calcium dissolution process. However, calcium recovery dropped significantly after 90 min when considering NH₄Cl as a solvent. It could be deduced from this that the maximum leaching recovery of calcium by NH₄Cl was achieved after 90 min and decreased drastically during the last dissolution time interval. Calcium dissolution by the CH₃COOH solution kept increasing insignificantly in the last interval whilst NH₄Cl was decreasing, until they both had an equal recovery after the maximum dissolution time.

At a temperature of 100 °C, there was an observation that 100% of calcium can be extracted from ironmaking slag after a maximum leaching period of 120 min. This was only accomplished by CH₃COOH and NH₄Cl whereas, (NH₄)₂SO₄ performed poorly at all temperature operating conditions. In addition, silica gelling was also observed at a temperature of 100 °C. Silica gelling only formed with CH₃COOH solutions. Furthermore, solid–liquid separation was somewhat hindered but calcium recovery seemed not to be compromised given the fact that this lixiviant was found to be the best in dissolving calcium from the slag.

4.2. XRD Analyses of Blast Furnace Ironmaking Slag Leaching Residues

The residue XRD patterns indicated that aluminoakermanite ((Ca,Na)₂(Al,Mg,Fe²⁺)) and NbTi(OH)₃ were the dominant phases after the leaching process. Many iron and steel slags are made up mainly of CaO, SiO₂, MgO, Al₂O₃, and MnO. The main mineral phases in steel slags are dicalcium and tricalcium silicates, calcium aluminium silicates, calcium oxides, and periclase. The findings from this study did not show any common mineral phases with that of the literature and this could be the fact that slags are very reactive and they are capable of changing properties during leaching[23].

4.3. EDX Analyses of Blast Furnace Ironmaking Slag Leaching Residues

EDX spectrum analysis that was performed gives an indication that calcium is present in different forms. There are still peaks of calcium present in the residue and this can better indicate that calcium was still available as free lime or as solid solutions. In addition, there is no relevant EDX spectrum analyses investigated on the same ironmaking slag to concur these findings in the literature reviewed.

4.4. The Effect of Reaction Rate-Controlling Mechanisms on the Leaching of Calcium from Ironmaking Slag

To compare with previous work, [23] found that an activation energy of a chemically-controlled process is usually greater than 41.8 kJ/mol while for diffusion-controlled reactions E_a is generally within the range of 4.18–12.55 kJ/mol. In addition, [23] suggested that the reaction depends on the diffusion by the product layer when the activation energy is between 8.37 and 20.92 kJ/mol, whereas reported figures of 12.55–25.08 kJ/mol. The results obtained from this study concur with the findings from [24]. There is a wide range of findings from previous work on reaction-controlling mechanisms; this is because there is also a wide range of experimental conditions and leaching variables. In some cases, leaching reactions were assisted by catalysts, while others are run with unassisted and/or blended solvents. Nonetheless, the activation energies obtained in the present work confirmed that leaching of calcium from CaO bearing slag was controlled by diffusion by the product layer since they fall in the range of 12.55–25.08 kJ/mol.

During leaching with (NH₄)₂SO₄ solvent, low recoveries of calcium were observed from analyses. A similar problem has been previously observed with the diffusion-limiting silica (SiO₂) layering in silicate minerals such as serpentine. The main reason for the low dissolution of calcium and magnesium from silicate minerals by reaction mechanisms is the increase of particle radius due to increasing silica-rich layer that remains passive during leaching reaction. Therefore, the ordinary shrinking core model does not apply due to formation of silica gelatinous layer. In order to correct the model prediction, two-stage leaching is sometimes used to decrease silica gelling that core the core of the mineral [24]. High-silica content on the blast furnace slag consequently created resistance to calcium extraction when (NH₄)₂SO₄ solvents were used.

Although CH₃COOH solvents attained high recoveries of calcium with increasing temperature, thick SiO₂ gelatinous layer in the residue was observed after the leaching reaction. Filtration process efficiency was also affected by the silica layer. Arrhenius plots for chemically and diffusion-controlled mechanisms during fitting in CH₃COOH solvent kinetic data were also graphically undefined, similar to the (NH₄)₂SO₄ diffusion model. The recovery of calcium indicates that the rate-limiting step of dissolution of the dominant mineral in the ironmaking slag (aluminoakermanite) in CH₃COOH is most likely due to formation of a passive silica layer. The silica-rich layer remains passive during dissolution of magnesium and calcium from silicate minerals [24].

Mixed reaction, chemical reaction, and diffusion-controlled mechanisms have no importance in interpreting the leaching rates of calcium when CH₃COOH and (NH₄)₂SO₄ were used as lixiviants due to the formation of silica layer. The leaching behaviour of high-silica minerals such as wollastonite and serpentine sometimes have dissolution rates where kinetic models are inappropriate. Nevertheless, the rate-controlling steps of the leaching reaction are influenced by the increase in size of the particle radius at high temperatures. The dissolution rate of calcium and magnesium can still increase over time. Consequently, leaching still took place and CH₃COOH lixiviant showed best leaching results in calcium dissolution, despite the failure to generate reasonable kinetic data. Selective flotation of quartz and multi-stage leaching of high-silica minerals that are rich with calcium and magnesium were previously investigated. However, selective separation of free quartz did not eliminate the problem of the formation of gelatinous silica layer during leaching of calcium and magnesium [24].

5. Conclusions

The experimental results from this study show that extraction of calcium from ironmaking slag by leaching is feasible. The best calcium recovery was achieved by acetic acid; ammonium salts were less efficient. One hundred percent extraction can be achieved with a solid–liquid ratio of approximately 200 g/L, reaction temperatures of 100 °C, and 0.5 M acetic acid. As the reaction temperature was increased from 50 °C up to 100 °C, silica gelling was observed, which hindered filtration of a leach liquor from a residue. Silica gelling did not compromise calcium recovery, but it appears to have affected the reaction rates. The XRD analyses of the leach residue showed high concentrations of calcium in the form of silicates, also known as aluminoakermanite ((Ca,Na)₂(Al,Mg,Fe²⁺)). EDX analysis of the residue also deduced that there is free calcium present in the ironmaking slag that was investigated. The kinetic data for the effect of reaction temperature leaching data fitted the shrinking core model equation, 1 − 2/3x − (1 − x)^2/3, which described diffusion-controlled leaching reactions. The diffusion controlled by a product layer model was fitted and the activation energy was calculated as 19.4930 kJ/mol for NH₄Cl solvent. The kinetic data for (NH₄)₂SO₄ and CH₃COOH solvents did not fit into any model because of formation of gelatinous silica layer during leaching process.