Chemical and Mechanical Factors on Phosphorus Dissolution Behavior from P-concentrated Slag

Abstract

Since it has been urgently demanded to secure the secondary phosphorus resources domestically in addition to the establishment of a stable supplying path of phosphate ore from overseas, the recovery of phosphorus from steelmaking slags is attracting attention. The phosphorus-enriched slag obtained by dephosphorizing the high phosphorus hot metal, which was prepared by the reduction of conventional steelmaking slag, is thought to be useful as a raw phosphorus resource. It was reported in our previous paper that phosphorus can be effectively separated from P-enriched slag by the citric acid leaching method. In this study, the effects of citric acid concentration in solution, solution temperature, slag/solution mass ratio, and P-enriched slag composition on phosphorus dissolution behavior were additively investigated. A phosphorus dissolution from the P-enriched slag was significantly promoted by increasing the citric acid concentration in solution. However, even if the solution temperature was increased during leaching, the speed and ratio of phosphorus dissolution from the P-enriched slag did not change. With increasing the slag/solution mass ratio, the concentration of each element in solution increased, while their dissolution ratios decreased. Furthermore, the phosphorus dissolution ratio was suppressed in the case of the P-enriched slag with low CaO and high P₂O₅ concentrations. Appropriate conditions for various factors affecting phosphorus dissolution behavior from P-enriched slag were discussed.

1. Introduction

Phosphorus is primarily used in agriculture, and approximately 88% of domestic demand goes towards fertilizer production.¹⁾ P also finds extensive use in industrial sectors such as pharmaceuticals, plating agents, semiconductor etching agents, and, more recently, as an electrolyte in lithium-ion batteries,¹⁾ making it a vital element in modern society. However, the production of phosphate ore, the raw material, is concentrated among the top three countries, accounting for 74% of global production,¹⁾ leading to international supply concerns. The European Commission identified phosphate ore in 2014 and phosphorus (P₄) in 2017 as “critical raw materials,”^2,3) signifying resources with high supply risk and economic importance. Consequently, securing phosphorus resources is a critical component of national resource strategies. In this context, Japan, reliant on imported phosphorus ores,¹⁾ urgently needs to ensure stable secondary phosphorus resources, and this has prompted various research initiatives. In the 1970s, phosphorus was identified in sewage,⁴⁾ leading to research on its removal from wastewater to combat eutrophication for aquatic environment preservation in lakes and marshes. Subsequent studies have explored phosphorus recovery using removal technologies, with reported benefits of using the resultant phosphate fertilizers for fertilization.^{5,6,7,8,9,10,11,12,13,14,15,16)} Additionally, sewage, livestock, and industrial wastes have been identified as candidate of secondary phosphorus resources, and research is ongoing into applying sewage treatment technology to recover phosphorus from such waste.^{17,18,19,20,21)} Among the secondary resources of phosphorus, steelmaking slag has attracted much attention in recent years. Matsubae-Yokoyama et al.²²⁾ analyzed the material flow of phosphorus in Japan and demonstrated that the total amount of phosphorus contained in steelmaking slag is comparable to that in imported phosphate ores. Various methods for recovering phosphorus from the steelmaking slag have been reported.^23,24,25) One proposed process is the dephosphorization of high-phosphorus pig iron at high temperatures, producing slag with high concentrations of phosphoric acid (P-enriched slag) , which can be effectively used as phosphate fertilizer while reducing the total amount of slag emissions.²⁶⁾ The P-enriched slag obtained is concentrated with phosphoric acid up to approximately 30 mass%,²⁷⁾ comparable to natural phosphate ores.²⁸⁾ Yellow phosphorus, a purified phosphorus product, is used in industries that require elemental phosphorus¹⁾ and in fertilizer applications, and is highly valuable. Therefore, developing a separation process of phosphorus from P-enriched slags is crucial. The authors investigated the application of the acid leaching method to separate phosphorus from P-enriched slag, examining the effects of the pH of the leachate, acid type, leaching agitation method, slag cooling conditions, and high-temperature oxidation treatment of the slag.^29,30) These studies revealed that phosphorus leaching was enhanced by lowering the leachate pH and adding an organic acid that formed complex ions within the leachate. Additionally, employing a pot mill for simultaneous grinding and leaching proved effective. In addition, during slag preparation, quenching the P-enriched slag from high temperatures, followed by oxidation treatments, was beneficial for phosphorus leaching.

To scale leaching operations of P-enriched slag to an industrial level, it is essential to increase productivity by improving the extraction ratio and leaching speed of phosphorus and handling larger amounts of slag. Leaching, a hydrometallurgical process for various nonferrous ores, has been successfully commercialized, and numerous studies have examined the effects of ore particle size, ore/solution ratio, agitation speed, temperature, pressure, and coexisting ion concentration on leaching to enhance the extraction ratio and speed of target elements from the ore.^{31,32,33,34,35,36,37)} These studies indicate that smaller ore particle sizes and ore/solution ratios, higher agitation speeds, and increased leachate temperatures elevate the leaching speed. High-pressure leaching of oxide ores is economically advantageous, as it yields a high extraction ratio of target elements and reduces chemical consumption.³⁸⁾ Given that P-enriched slag consists of oxides, it likely shares properties with oxide ores, suggesting that the aforementioned factors similarly affect P-enriched slag leaching. In this study, leaching experiments were conducted using a pot mill combined with a pH controller to explore the effects of citric acid concentration, leachate temperature, slag mass-to-solution volume ratio (S/L ratio), and slag composition on the leaching behavior of each element.

2. Experimental

2.1. Phosphorus-enriched Slag

First, 2.5 tons of hot metal, with an initial phosphorus concentration of 1.02 mass%, was melted at 1400°C in an induction furnace (inner diameter 1.0 m) lined with magnesia. Oxygen was blown onto the hot metal surface at a rate of 1.40 Nm³/min/t from the top nozzle, and argon gas was blown from the bottom of furnace at 0.10 Nm³/min/t to stir the hot metal. Simultaneously, 2.61 kg/min/t of lime was supplied from a sub-nozzle with 0.28 Nm³/min/t of argon gas to the hot metal surface, and dephosphorization was conducted for 25 min. After the treatment, the slag was removed, placed in a steel container, and cooled under atmospheric conditions. The slag was then coarsely crushed to a grain size of approximately 3 mm. Following the removal of metallic iron particles by magnetic separation, the slag was further crushed and sieved into 25–53 μm fractions for use in leaching experiments (hereafter referred to as “slag A”). To assess the impact of slag composition on each leaching condition, “slow-cooled slag” from the previous report³⁰⁾ (hereinafter referred to as “slag B”) was also processed similarly for this experiment.

2.2. Leaching

The slag-leaching experimental apparatus, comprising a nylon mill pot and nylon-coated steel balls, has been detailed previously.^29,30) For leaching, aqueous citric acid solutions of 0.002, 0.005, 0.010, or 0.100 mol/L were utilized as initial solutions, adjusted to pH 4 by adding reagent grade granular NaOH while stirring with a magnetic stirrer. 800 mL of the initial solution and 1.00 g of P-enriched slag powder were introduced into the slag-leaching apparatus, and the nylon mill pot was immediately set to rotate at approximately 90 r.p.m., achieving 80% of the critical mill speed. During leaching, the pH of the leachate was continuously monitored using a pH glass electrode immersed in the leachate, whereas a 1 mol/L HCl solution with the same citric acid concentration as the initial solution was dropped from an automatic pH controller to maintain a steady pH of 4.

The solution temperature during leaching was established at three levels: room temperature (293–298 K), 323, and 343 K. A cartridge heater coated with polytetrafluoroethylene (PTFE) resin was inserted into the leachate to heat it to 323 and 343 K. At predetermined intervals, 2 to 5 mL of leachate was sampled using a plastic syringe attached to a membrane filter cartridge (0.2 μm pore size). To counteract evaporation and maintain the leachate volume, particularly during experiments at 323 and 343 K, distilled water was added continuously in drops. The leachate volume at the time of collection was calculated by comparing the sodium concentration in the collected leachate with that in the initial solution, and the concentration of each element in the leachate was accordingly corrected.

In experiments where the S/L ratio varied, 4 or 8 g of slag A was introduced into 800 mL of 0.010 mol/L citric acid solution at room temperature, maintaining a constant pH of 4, whereas the 0.010 mol/L citric acid–1 mol/L HCl solution was added dropwise.

Following the leaching experiments, the leachate containing the slag and hydration reaction products was suction filtered through a membrane filter (0.45 μm pore size), and the resultant solid phase was dried at 20°C in an incubator for a minimum of 48 h.

2.3. Analysis

The elemental concentrations in the sampled leachate were determined using inductively coupled plasma atomic emission spectrometry (ICP-AES). The slag composition was determined via ICP-AES after alkaline fusion in a platinum crucible with 0.1 g of P-enriched slag powder combined with 2.0 g of reagent-grade sodium carbonate (Na₂CO₃) and 1.0 g of reagent-grade boron trioxide (B₂O₃), followed by a warm soak in dilute hydrochloric acid. Elemental mapping and quantification of each mineral phase were performed on the polished surfaces of the P-enriched slag powder and the residue powder after leaching, using scanning electron microscopy with energy-dispersive X-ray spectrometry (SEM-EDS).

3. Results and Discussion

3.1. Mineral Phases of Phosphorus-enriched Slag

Table 1 presents the analysis results for slags A and B. The Fe_tO concentration represents the FeO concentration, recalculated from the total iron content. Slag A exhibits a lower CaO concentration but a higher P₂O₅ concentration compared to slag B. The Fe_tO concentrations, 23.3 mass% for slag A and 20.7 mass% for slag B, are not significantly different. The basicity, calculated by the ratio of (mass% CaO)/(mass% SiO₂), is 7.41 for slag A and 5.12 for slag B, significantly higher than that of conventional steelmaking slag. Moreover, the (mass% CaO)/(mass% P₂O₅) ratio, referred to as the C/P ratio, is 0.93 for slag A and 1.82 for slag B. The latter exceeds the C/P ratio of 1.18 found in the pure 3CaO∙P₂O₅ (C₃P) compound, whereas the former remains below it. Elemental mapping of the cross-sections of slags A and B by SEM-EDS after polishing are shown in Figs. 1 and 2, respectively. In Fig. 2, the C₃P phase is labeled “a,” the wüstite phase is labeled “b,” and the glassy phase is labeled “c,” whereas metallic iron is observed in other fields of view, as previously reported.³⁰⁾ The mineral phases in slag A (Fig. 1) are similar to those in slag B: “a” represents the C₃P phase, “b” the wüstite phase, and “c” the glassy phase. The metallic iron phase is also observed in other fields of view, as in slag A. It is confirmed that the constitution of the mineral phases is consistent between slags A and B, although the morphology of each mineral phase differs.

Table 1. Analyzed chemical compositions of P-concentrated slags (mass%).

	CaO	SiO2	P2O5	FetO	Al2O3	MgO	MnO
Slag A	30.4	4.1	32.7	23.3	1.2	3.0	5.3
Slag B30)	42.5	8.3	23.4	20.7	1.6	2.2	1.3

Fig. 1. Elemental mapping on the cross section of slag A obtained by EDS a: C₃P, b: wüstite and c: glass.

Fig. 2. Elemental mapping on the cross section of slag B obtained by EDS a: C₃P, b: wüstite and c: glass.

To clarify the impact of slag composition on phosphorus enrichment in the P-enriched phase, the following analysis was conducted. Quantitative analysis by SEM-EDS was performed at 25–70 points for the P-enriched and glassy phases of each slag to determine the phosphorus distribution ratio (L_P), as expressed in Eq. (1):

  

$$L_{\text{P}}=\frac{({\text{mass\%P}}_2{\text{O}}_5)\text{in}\mathrm{\hspace{0.17em} }{\text{C}}_3\text{P}}{{\text{(mass\%P}}_2{\text{O}}_5\text{)in}\mathrm{\hspace{0.17em} }\text{liquid}}$$(1)

Here, “(mass%P₂O₅)in C₃P” is the P₂O₅ mass concentration in the C₃P phase and “(mass%P₂O₅)in liquid” is the P₂O₅ mass concentration in the glassy phase. The relationships between L_P and slag basicity, Fe_tO concentration in the slag, and C/P ratio of the slag are shown in Figs. 3(a)–3(c), respectively. Figures 3(d)–3(f) illustrate the relationship between the FeO concentration in the P-enriched phase and the corresponding values. The reason for the wide range of L_P values in slag B is discussed below. In the case of conventional steelmaking slag, when the basicity is higher than the nose top of 2CaO∙SiO₂ (C₂S) in the CaO–SiO₂–Fe_tO system phase diagram (approximately 3.2), the phosphorus distribution ratio between C₂S phase and liquid slag decreases from 63 to 19 as the basicity increases and as the Fe_tO concentration decreases.³⁹⁾ In the case of the P-enriched slag used in this study, as shown in Fig. 3(a), L_P also decreases with the increase in basicity; however, the L_P is less than 19, which is lower than that of conventional steelmaking slag. Meanwhile, the relationship between L_P and Fe_tO concentration in the slag shown in Fig. 3(b) is opposite to that of conventional steelmaking slag. This is because FeO does not dissolve in the C₂S phase (in which a low concentration of C₃P is dissolved), a P-enriched phase in conventional steelmaking slags. Conversely, FeO dissolves at higher concentrations in the C₃P phase of the P-enriched slag as the Fe_tO concentration in the slag increases, as shown in Fig. 3(e). Consequently, the phosphorus concentration in the C₃P phase for P-enriched slag with high Fe_tO concentration is relatively lower than for that with low Fe_tO concentration, resulting in a lower L_P. Figure 4 illustrates the relationship between the FeO and SiO₂ concentrations in the C₃P phase. The low SiO₂ concentration in the C₃P phase of both P-enriched slags suggests the formation of a solid solution phase primarily consisting of C₃P, rather than a C₂S–C₃P phase dominated by C₂S. Therefore, we attempt to correlate the phosphorus distribution ratio L_P between the solid/liquid phases and the FeO and SiO₂ concentrations in the P-enriched slag with the C/P ratio, rather than with slag basicity.

Fig. 3. Relationships between phosphorus distribution ratio and FeO content in C₃P and basicity, Fe_tO content and C/P ratio in slag.

Fig. 4. Relationships between FeO and SiO₂ contents in C₃P phases.

The CaO–Fe_tO–P₂O₅ system phase diagram⁴⁰⁾ (Fig. 5) is presented in molar concentration notation. The dashed line corresponding to the C/P molar ratio of 3 for C₃P (1.18 by mass) is also shown. When the P₂O₅ concentration in the slag falls below the dashed line, L_P is higher because the P₂O₅ concentration in the liquid phase, in equilibrium with C₃P between α and β, is lower. Conversely, when the P₂O₅ concentration in the slag exceeds the dashed line, the L_P is lower due to the higher P₂O₅ concentration in the liquid phase between γ and δ. The slag composition used in this study is indicated in Fig. 5. The P₂O₅ concentration in slag A is above the dashed line, whereas that in slag B is below it, indicating the relationship between the C/P ratio of the slag and L_P as shown in Fig. 3(c). A slag C/P ratio above the dashed line in Fig. 5 (indicating a high P₂O₅ concentration in the slag) suggests that the CaO available to form C₃P is insufficient, leading to a reduction in C₃P production and, consequently, reduction of phosphorus enrichment in the solid phase. This results in a reduced L_P, as illustrated in Fig. 3(c). In Fig. 5, there is no solid solution of FeO in C₃P, but the authors were able to synthesize a C₃P–Fe₃(PO₄)₂ solid solution.³⁰⁾ Considering that Fe²⁺, abundant in the slag melt, compensates for the lack of CaO by substituting into the divalent cation site in the C₃P phase (originally occupied by Ca²⁺) to form a C₃P–Fe₃(PO₄)₂ solid solution, the increase in FeO concentration in the C₃P phase can be explained by Figs. 3(f) and 4. In other words, it is reasonable to consider the relationship between variations in L_P, along with the changes in FeO concentration in the C₃P phase, and the C/P ratio. In contrast, in slag B, where the C/P molar ratio exceeds 3 (1.18 by mass), the high SiO₂ concentration in the C₃P phase, as shown in Fig. 4, can be attributed to excess CaO reacting with SiO₂ in the slag to form C₂S, thereby increasing the amount of dissolved C₂S in the C₃P phase. The concentration of dissolved C₂S in the C₃P phase is influenced by the progress of the dissolution reaction during high-temperature synthesis and the slow cooling process. However, in an experimental system like this synthesis procedure, where the reaction vessel is large and overall equilibrium is challenging to achieve, the progress of the C₂S solid solution reaction may vary across different parts of the slag. Therefore, the SiO₂ concentration in the C₃P phase in slag B varied significantly, as shown in Fig. 4, and this variation in the C₂S solid solution concentration is presumed to have caused the wide range in the L_P value of slag B (Fig. 3). In contrast, when CaO is deficient, as in slag A, the SiO₂ concentration in the C₃P phase is reduced, as shown in Fig. 4. This is because C₂S is less soluble, resulting in a narrower range of L_P values, as shown in Fig. 3.

Fig. 5. Composition of slag A and B plotted in CaO–Fe_tO–P₂O₅ phase diagram.⁴⁰⁾

3.2. Effect of Citric Acid Concentration on Leaching

Figure 6 shows the changes in the concentration of each element in the leachate with respect to time at varying citric acid concentrations during a leaching experiment with an S/L ratio of 1.25 g/L of slag A to solution. Increasing the citric acid concentration from 0.002 to 0.010 mol/L increases the phosphorus and calcium concentrations from 47 to 67 and 67 to 98 mg/L, respectively, after 24 h. A further increase in concentration from 0.010 to 0.100 mol/L results in the phosphorus and calcium concentrations doubling to 127 and 193 mg/L, respectively. In contrast, the concentrations of iron, silicon, and manganese after 24 h increase from 95 to 112, 20 to 24, and 29 to 39 mg/L, respectively, as the citric acid concentration increased from 0.002 to 0.100 mol/L. These increases are not as pronounced as those observed for phosphorus and calcium. Figures 7(a)–7(c) show the relationship between the dissolution ratio (R_M) of each element after 24 h of leaching and the citric acid concentration, temperature, and S/L ratio, respectively. Here, R_M is calculated using Eq. (2).²⁵⁾

  

$$R_M=\frac{C_M\times V}{W\times x_M/100}$$(2)

Fig. 6. Comparison of dissolution behavior of elements eluated from slag A in the solution with various concentration of citric solution at room temperature in the experiment of S/L=1.25 g/L.

Fig. 7. Dissolution ratios of elements at the 24 h leaching with various experimental conditions.

In Eq. (2), C_M is the concentration of element M in the leachate (mg/L), V is the volume of the leachate (L), W is the weight of the slag (mg), and x_M is the mass concentration of element M (mass%) in the initial slag. R_M represents the ratio of the amount of dissolved element M in the solution to the total amount of element M in the slag. Figure 7(a) indicates that the dissolution ratios of phosphorus and calcium increase from 0.2 to 0.7 as the citric acid concentration rises. The dissolution ratios of iron and manganese also increase, albeit to a lesser extent than phosphorus and calcium. Conversely, silicon is nearly completely dissolved at citric acid concentrations of 0.005 mol/L or higher. Previously, it has been reported that the dissolution of the glassy phase is facilitated in acidic solutions.^{41,42,43,44,45,46)} In this leaching experiment, the dissolution ratio of silicon was nearly 100%, indicating that almost all the glassy phase, which is rich in silicon, is dissolved after 24 h of leaching. Meanwhile, the dissolution ratios of phosphorus, calcium, and manganese are less than 70%, and that of iron is less than 50%, suggesting that the dissolution of the P-enriched and wüstite phases is incomplete within 24 h, even at a citric acid concentration of 0.100 mol/L. Although the dissolution ratio of each element may be underestimated due to precipitation in the leachate, such precipitates are not detected by X-ray diffraction (XRD) analysis of the residue after leaching because of their amorphous phase nor are they confirmed by EDS analysis, as described below. Similar to Fig. 1, a wüstite phase is observed, and as shown in Fig. 7(a), the dissolution ratio of iron does not significantly alter with increasing citric acid concentration. Although iron ions form complex ions with citric acid, which become more readily dissolved in the leachate,³⁰⁾ the influence of citric acid content on the leaching of the wüstite solid phase, where iron is the predominant component, is likely minimal. Thus, we propose that an increase in citric acid concentration in the leachate may enhance the dissolution of the P-enriched phase (C₃P) while inhibiting the dissolution of the iron oxide phase.

3.3. Effect of Leachate Temperature

The impact of leachate temperature on the dissolution behavior of each element is shown in Fig. 8 for the leaching experiment using slag A. For phosphorus and calcium, the effect of temperature on concentration change is minor. Conversely, for iron, silicon, and manganese, the initial increase in concentration is observed at elevated leachate temperatures, but this effect plateaus after 2 h. The concentration of these elements increases slowly during leaching at room temperature, but reaches the concentrations observed at 323 K after 12 h. Figure 7(b) shows the relationship between the dissolution ratio after 24 h of leaching and the leaching temperature. The variation in the dissolution ratio of each element with temperature is negligible. Du et al.⁴⁷⁾ reported that the leaching speed of phosphorus from conventional steelmaking slag increased in the early stages of leaching with increased leachate temperature; however, the change in the dissolution ratio of phosphorus after 120 min was insignificant. In contrast, in this experiment utilizing P-enriched slag, neither the initial phosphorus leaching speed (Fig. 8) nor the phosphorus dissolution ratio after an extended period (Fig. 7(b)) is affected by the leachate temperature. Therefore, it is determined that the effects of temperature on the phosphorus leachability of conventional steelmaking slag and P-enriched slag differs.

Fig. 8. Comparison of dissolution behavior of elements from slag A in 0.010 mol/L citric solution at various leaching temperature in the experiment of S/L=1.25 g/L.

To evaluate the temperature dependence of the stability region of the various phosphates that may form in the M²⁺ (or M³⁺)−PO₄³⁻−citrate system solutions in this experiment, the relationship between the concentrations of various metal ions and phosphorus at 298, 323, and 343 K for pH 4 and a citric acid concentration of 0.010 mol/L was calculated using the solubility products of the phosphate formation reactions. Citrate ions are present in aqueous solution as (C₆H₅O₇)³⁻, (HC₆H₅O₇)²⁻, (H₂C₆H₅O₇)⁻ and (H₃C₆H₅O₇)⁰,⁴⁸⁾ and phosphate ions also vary in form depending on the pH.⁴⁹⁾ Thus, the general equations for the various phosphate formation reactions involving metal citrate complexes are shown in Eqs. (3), (4), (5), (6), (7):

  

$$\begin{array}{l}3\mathrm{F}\mathrm{e}{({\mathrm{H}}_{\mathrm{c}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7)}_{\mathrm{a}}{}^{2-\mathrm{a}(3-\mathrm{c})}+2{\mathrm{H}}_{\mathrm{d}}\mathrm{P}{\mathrm{O}}_4{}^{(3-\mathrm{d})-}+8{\mathrm{H}}_2\mathrm{O}\\ =\mathrm{F}{\mathrm{e}}_3{(\mathrm{P}{\mathrm{O}}_4)}_2\cdot 8{\mathrm{H}}_2\mathrm{O}(\mathrm{s})+3a{\mathrm{H}}_{\mathrm{b}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7{}^{(3-\mathrm{b})-}\\ +(3a\text{(}c-b\text{)}+2d){\mathrm{H}}^{+}\end{array}$$(3)

  

$$\begin{array}{l}\mathrm{F}\mathrm{e}{({\mathrm{H}}_{\mathrm{c}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7)}_{\mathrm{a}}{}^{3-\mathrm{a}(3-\mathrm{c})}+{\mathrm{H}}_{\mathrm{d}}\mathrm{P}{\mathrm{O}}_4{}^{(3-\mathrm{d})-}+2{\mathrm{H}}_2\mathrm{O}\\ =\mathrm{F}\mathrm{e}\mathrm{P}{\mathrm{O}}_4\cdot 2{\mathrm{H}}_2\mathrm{O}(\mathrm{s})+a{\mathrm{H}}_{\mathrm{b}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7{}^{(3-\mathrm{b})-}\\ +(a\text{(}c-b\text{)}+d){\mathrm{H}}^{+}\end{array}$$(4)

  

$$\begin{array}{l}10\mathrm{C}\mathrm{a}{({\mathrm{H}}_{\mathrm{c}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7)}_{\mathrm{a}}{}^{2-\mathrm{a}(3-\mathrm{c})}+6{\mathrm{H}}_{\mathrm{d}}\mathrm{P}{\mathrm{O}}_4{}^{(3-\mathrm{d})-}+2{\mathrm{H}}_2\mathrm{O}\\ =\mathrm{C}{\mathrm{a}}_{10}{(\mathrm{P}{\mathrm{O}}_4)}_6{(\mathrm{O}\mathrm{H})}_2(\mathrm{s})+10a{\mathrm{H}}_{\mathrm{b}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7{}^{(3-\mathrm{b})-}\\ +(10a\text{(}c-b\text{)}+6d+2){\mathrm{H}}^{+}\end{array}$$(5)

  

$$\begin{array}{l}\mathrm{A}\mathrm{l}{({\mathrm{H}}_{\mathrm{c}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7)}_{\mathrm{a}}{}^{3-\mathrm{a}(3-\mathrm{c})}+{\mathrm{H}}_{\mathrm{d}}\mathrm{P}{\mathrm{O}}_4{}^{(3-\mathrm{d})-}\\ =\mathrm{A}\mathrm{l}\mathrm{P}{\mathrm{O}}_4(\mathrm{s})+a{\mathrm{H}}_{\mathrm{b}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7{}^{(3-\mathrm{b})-}+(a\text{(}c-b\text{)}+d){\mathrm{H}}^{+}\end{array}$$(6)

  

$$\begin{array}{l}\mathrm{M}\mathrm{n}{({\mathrm{H}}_{\mathrm{c}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7)}_{\mathrm{a}}{}^{2-\mathrm{a}(3-\mathrm{c})}+{\mathrm{H}}_{\mathrm{d}}\mathrm{P}{\mathrm{O}}_4{}^{(3-\mathrm{d})-}\\ =\mathrm{M}\mathrm{n}\mathrm{H}\mathrm{P}{\mathrm{O}}_4(\mathrm{s})+a{\mathrm{H}}_{\mathrm{b}}{\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{O}}_7{}^{(3-\mathrm{b})-}+(a\text{(}c-b\text{)}+d-1){\mathrm{H}}^{+}\end{array}$$(7)

The solubility products for various phosphates listed in the Minteq.v4 database⁵⁰⁾ were used for the calculations, with the lowest solubility at pH 4 being applied. However, because no values for AlPO₄ were available in this database, the equilibrium constants (K) and the enthalpy change (Δ_rH) were considered:⁵¹⁾

  

$$\begin{array}{l}\mathrm{A}{\mathrm{l}}^{3+}+\mathrm{H}\mathrm{P}{\mathrm{O}}_4{}^{2-}=\mathrm{A}\mathrm{l}\mathrm{P}{\mathrm{O}}_4(\mathrm{s})+{\mathrm{H}}^{+}\\ \mathrm{l}\mathrm{o}\mathrm{g}{K}_8=7.2087,\mathrm{\hspace{0.17em} }{\mathrm{\Delta }}_r{H}_{8,\mathrm{\hspace{0.17em} }298 \mathrm{K}}=96.6313\mathrm{\hspace{0.17em} }\mathrm{k}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}\end{array}$$(8)

The enthalpy change, Δ_rH, for each reaction is assumed to be constant across temperatures. The temperature dependence of the equilibrium constants for the formation of phosphate and complex ions, as well as the equilibrium in solution of phosphate and citrate ions, are derived from the van’t Hoff equation (Eq. (9)) and the equilibrium constant values at 298 K:

  

$${\left(\frac{\partial lnK}{\partial T}\right)}_P=\frac{{\mathrm{\Delta }}_{r}H}{R{T}^2}$$(9)

The reported Δ_rH values were zero for the decomposition reactions of Fe₃(PO₄)₂∙8H₂O, Ca₁₀(PO₄)₆(OH)₂, and MnHPO₄, as well as for some iron-citrate, aluminum-citrate, and manganese-citrate complex ion formation reactions. The results, which combine the metal ion and phosphorus concentrations for all phosphate formation reactions involving citrate ions, are depicted in Fig. 9. These results indicate that temperature has a negligible effect on the relationship between metal ion and phosphorus concentrations for the phosphates. The experimental data are also plotted in Fig. 9. Compared to the calculated values, the experimental data are consistent with the calculated equilibrium concentrations of each phosphate in the Mn²⁺−PO₄³⁻−citrate−water system, suggesting that the formation of MnHPO₄ may limit phosphorus leaching under these experimental conditions. Additionally, the minimal temperature dependence of the experimental results aligns with the calculated solubility product of MnHPO₄. The XRD patterns of the slag before and after leaching are presented in Fig. 10 for 2θ = 10°–30°. The C₃P peak is evident in both samples; however, the first peak of wüstite, located at 42.03° according to Crystallographica Search-Match which is software of XRD pattern analysis, is not shown in the figure. Compared to the XRD pattern of the slag before leaching (Fig. 10(a)), halo pattern characteristic of amorphous materials is clearly observed at 2θ = 17°–25° in the residue after leaching (Figs. 10(b)–10(d)), indicating the formation of secondary amorphous products in the leachate. SEM observations and electron probe microanalysis of the residue after leaching did not reveal any secondary products with the composition corresponding to MnHPO₄, possibly because of the formation of a very thin MnHPO₄ amorphous film on the slag powder surface.

Fig. 9. Solubility of phosphate at various temperatures at pH=4 and 0.010 mol/L citric solution.

Fig. 10. X-ray diffraction peaks obtained by XRD analysis of slag A before and after leaching at various temperatures.

According to the line shown in Fig. 9, the solubility of calcium phosphates, such as hydroxyapatite, is high at pH 4; thus, their formation does not inhibit phosphorus leaching. Conversely, iron and aluminum phosphates exhibit low solubility in weakly acidic to weakly basic solutions, but their high complexation stability constants with citric acid lead to the formation of stable complex ions, which prevent phosphate precipitation. Manganese, although having a relatively high complexation stability constant with citric acid, does not reach the levels of iron or aluminum, and manganese phosphates display low solubility in weakly acidic to weakly basic solutions. As a result, manganese dissolved in a citric acid-containing solution may form manganese phosphates, reducing the phosphorus equilibrium concentration. Therefore, it is advisable for the P-enriched slag to be free of MnO. Furthermore, Fig. 7(b) suggests that temperature control is not necessary during the actual leaching process for slags with a composition analogous to the experimental system presented.

3.4. Effect of Slag/Solution Ratio

Figure 11 shows the dissolution behavior and dissolution ratio of phosphorus when the S/L ratio of slag A in the solution is varied. A higher S/L ratio results in a greater concentration of phosphorus in the leachate from the initial stage of leaching (Fig. 11(a)), but a lower phosphorus dissolution ratio (Fig. 11(b)).

Fig. 11. Dissolution behavior (a) and dissolution ratio (b) of phosphorus eluated from slag A in 0.010 mol/L citric solution at room temperature as a function of S/L ratio.

To understand the decrease in phosphorus dissolution ratio as the S/L ratio increased, the average phosphorus leaching speed $({\overline{v}}_{P,t_n})$ between time t_n−1 and t_n was calculated from the phosphorus dissolution ratio for each S/L ratio using Eq. (10):

  

$${\overline{v}}_{P,t_n}=\frac{R_{P,t_n}-R_{P,t_{n-1}}}{t_n-t_{n-1}}$$(10)

where $R_{P,t_n}$ and $R_{P,t_{n-1}}$ are the phosphorus dissolution ratios at t_n and t_n−1, respectively. Figure 12 shows the change in the average phosphorus leaching speed over time. Initially, there is no significant difference in the average phosphorus leaching speed at each S/L ratio; however, with the increase in time, a higher S/L ratio correlates with a reduced leaching speed, indicating a delay in phosphorus leaching. Figure 13 shows the SEM image of the residue after leaching, showing that a smaller S/L ratio results in a finer particle size of the residue. Given that the mill pot shape, number of mill balls, and rotation speed are the same across all leaching experiments, the milling power input per unit time, E, remains constant regardless of the S/L ratio. Denoting the milling time as t and the slag volume as W, the total milling power input per unit slag weight (e) is given by e = t·E/W. As milling progresses and t increases, a larger W and a smaller E/W ratio result in a smaller increment of e, which is thought to influence the progress of milling. Generally, in mineral leaching, a larger surface area equates to a greater leaching speed.⁵²⁾ As shown in Fig. 13, a smaller S/L ratio leads to more advanced milling and a finer grain size; thus, phosphorus leaching is considered to progress more readily, as shown in Fig. 12.

Fig. 12. Average dissolution rate of phosphorus from slag A in 0.010 mol/L citric solution at room temperature.

Fig. 13. SEM images of residues after leaching in 0.010 mol/L citric solution at room temperature.

Based on the discussion of the dissolved phosphorus concentration and dissolution ratio (Fig. 11), phosphorus leaching speed (Fig. 12), observation of residue after leaching (Fig. 13), and input power, the dissolution behavior when the S/L ratio varies in pot-mill leaching can be summarized as follows: In the early stage of leaching, mineral phases with high water solubility, such as the C₃P–C₂S phase (a P-enriched phase with dissolved SiO₂) and the glassy phase, are leached. Assuming that the glassy phase, rather than the C₃P–C₂S phase, is dissolved in the early stage for the slag used in this experiment, the leaching speed of this easily soluble phase is so rapid that the dependence on the S/L ratio is negligible, resulting in no significant difference in the initial phosphorus leaching speed, regardless of the S/L ratio. From the middle stage of leaching, if phosphorus primarily dissolves from the C₃P phase, which contains less dissolved SiO₂ and more dissolved FeO than the mineral phase with high water solubility, resulting in a slower leaching speed, phosphorus leaching is enhanced by milling slag with a low S/L ratio and high milling power per unit slag weight. This is because the mineral phase is more readily leached when the interface area with the solution is large. Consequently, if leaching operations are conducted at high S/L ratios to increase batch throughput, it is advisable to increase the milling power to enhance the dissolution ratio.

3.5. Effect of Slag Composition

Figure 14 compares the dissolution ratio of each element from slag A at pH 4 with the results for slag B.³⁰⁾ The phosphorus dissolution ratios for slags A and B are 0.36 and 0.43, respectively, with a lower calcium dissolution ratio observed in slag A. This suggests that the dissolution of C₃P, a P-enriched phase in slag A, may be impeded. This could be due to the lower SiO₂ concentration and higher FeO concentration within the C₃P phase, which are less conducive to its dissolution,³⁰⁾ as evidenced by Fig. 4. Conversely, the dissolution ratio of silicon was lower in slag B, which has a higher SiO₂ concentration in the P-enriched phase, as shown in Fig. 4. As reported previously,³⁰⁾ this is attributed to the precipitation of silicon as a secondary product, along with a significant amount of dissolved elements, whereas the P-enriched phase of slag B has a composition that is more readily dissolved. Although the iron dissolution ratio from slag B is expected to decrease because of the formation of secondary products, it is higher than that of slag A because some of the dissolved iron ions remain stable in solution by forming complex ions with citrate ions.

Fig. 14. Comparison of dissolution ratio of elements eluated from slag A and slag B in 0.010 mol/L citric solution at pH=4.

The leachability of phosphorus varies with slag composition, particularly as the composition of the P-enriched phase changes. Therefore, for the dephosphorization of high-phosphorus pig iron, it is preferable to concentrate the phosphorus in a phase that is easily dissolved. In the dephosphorization of conventional hot metal, the phosphorus distribution between the C₂S–C₃P phase and liquid slag, along with that between the liquid slag and hot metal, can be leveraged to enhance dephosphorization efficiency.^53,54) Therefore, it is expected that dephosphorization of high-phosphorus pig iron will be promoted by using solid-liquid coexisting slag in which the C₃P phase and the liquid phase exist. To establish a phosphorus separation process that includes phosphorus enrichment in the slag and subsequent leaching from the P-enriched slag, it is necessary to generate a highly soluble P-enriched phase (with high SiO₂ and low FeO concentrations) in a solid-liquid coexisting slag at high temperatures. To achieve this, the C/P ratio of the slag may be better to exceed 1.18, as discussed in Section 3.1.

4. Conclusion

In this study, the effects of citric acid concentration, leaching temperature, slag/solution ratio, and slag composition on the leaching behavior of each element from CaO–SiO₂–FeO–P₂O₅ P-enriched slag with a high P₂O₅ concentration were investigated, and the followings were revealed:

(1) The constitution of the mineral phase did not change despite variations in slag composition. However, as the CaO concentration in the slag decreased and the P₂O₅ concentration increased, the phosphorus distribution ratio between the P-enriched phase and slag melt decreased, the FeO concentration in the P-enriched phase increased whereas the SiO₂ concentration decreased.

(2) Increasing the citric acid concentration in the leachate facilitated phosphorus dissolution. In particular, when the concentration was increased from 0.010 to 0.100 mol/L, the phosphorus dissolution ratio increased more than twice, from about 0.3 to 0.7.

(3) Varying the solution temperature during leaching increased the rate of dissolution of iron and silicon in the early stages but did not influence the dissolution behavior of phosphorus and calcium.

(4) A higher slag/solution ratio led to increased concentrations of each element in the leachate; however, it also resulted in reduced phosphorus leaching speed and dissolution ratio because the slag grain size did not decrease during leaching.

(5) Changes in slag composition affected the SiO₂ and FeO concentrations in the C₃P phase, subsequently affecting the phosphorus dissolution ratio. To obtain a C₃P phase with increased SiO₂ and decreased FeO concentrations, the CaO/P₂O₅ mass ratio of the slag would be better to be at least 1.18.

In summary, to enhance phosphorus dissolution from P-enriched slag, it is advantageous to increase the citric acid concentration, increase the milling power input, and control the slag composition. However, managing the leaching temperature is not necessary.

Acknowledgments

This paper is based on results obtained from a project, P12004, subsidized by the New Energy and Industrial Technology Development Organization (NEDO).

References

• 1)   Japan  Oil, Gas and Metals National Corporation: Material Flow of Mineral Resources, 2018, 33, Phosphorus, https://mric.jogmec.go.jp/wp-content/uploads/2019/03/material_flow2018_P.pdf (in Japanese), (accessed 2022-10-06).
• 2)  European Commission, 2014: On the Review of the List of Critical Raw Materials for the EU and the Implementation of the Raw Materials Initiative, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52014DC0297&from=CS, (accessed 2022-10-06).
• 3)  European Commission, 2017: On the 2017 List of Critical Raw Materials for the EU, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52017DC0490, (accessed 2022-10-06).
• 4)   S.  Nambu,  S.  Kunikane,  G. F.  Lee,  W.  Rast and  A.  Jones: Suishitsu-Odaku-Kenkyu, 1 (1978), 157 (in Japanese). https://doi.org/10.2965/jswe1978.1.157
• 5)   J.  Zoltek: J. Water Pollut. Control Fed., 46 (1974), 2498. https://www.jstor.org/stable/25038300, (accessed 2022-10-07).
• 6)   I.  Hirasawa,  T.  Tanaka and  N.  Iwai: Suishitsu-Odaku-Kenkyu, 6 (1983), 229 (in Japanese). https://doi.org/10.2965/jswe1978.6.229
• 7)   H.  Katsuura and  Y.  Ueno: Kankyo-Gijutsu, 27 (1998), 397 (in Japanese). https://doi.org/10.5956/jriet.27.397
• 8)   Y.  Yoshino: Kankyo-Gijutsu, 7 (1978), 371 (in Japanese). https://doi.org/10.5956/jriet.7.371
• 9)   I.  Yall,  W. H.  Boughton,  R. C.  Knudsen and  N. A.  Slinclair: Appl. Microbiol., 20 (1970), 145. https://doi.org/10.1128/am.20.1.145-150.1970
• 10)   T.  Fukase,  M.  Shibata and  Y.  Miyaji: Suishitsu-Odaku-Kenkyu, 5 (1982), 309 (in Japanese). https://doi.org/10.2965/jswe1978.5.309
• 11)   M.  Kitagawa: Kankyo-Gijutsu, 2 (1973), 267 (in Japanese). https://doi.org/10.5956/jriet.2.267
• 12)   Y.  Shigetomi,  T.  Watanabe and  Y.  Hori: Kankyo-Gijutsu, 7 (1978), 378 (in Japanese). https://doi.org/10.5956/jriet.7.378
• 13)   T.  Kojima: Doctoral Thesis, Saga University, Graduate School of Sciences and Engineering, (2004).
• 14)   A.  Furuhata: Nogyo-oyobi-Engei, 89 (2014), 429 (in Japanese). https://agriknowledge.affrc.go.jp/RN/2010871630.pdf, (accessed 2022-10-19).
• 15)   K.  Oshita,  M.  Iwashita,  M.  Takaoka and  N.  Takeda: Kankyo-Kogaku-Kenkyu-Ronbunshu, 40 (2003), 395 (in Japanese). https://doi.org/10.11532/proes1992.40.395
• 16)   Y.  Iwai,  T.  Jozuka,  T.  Kobayashi,  T.  Kameya,  Y.  Miyake,  T.  Komatsu,  S.  Takagi and  F.  Mishina: Haikibutsu-Shigen-Junkan-Gakkai-Ronbunshi, 20 (2009), 203 (in Japanese). https://doi.org/10.3985/jjsmcwm.20.203
• 17)   T.  Sekito,  Y.  Dote,  K.  Kaikake,  S.  Masuda and  Y.  Suzuki: Doboku-Gakkai-Ronbunshu G, 64 (2008), 88 (in Japanese). https://doi.org/10.2208/jscejg.64.88
• 18)   K.  Ueda,  R.  Sakamoto,  J.  Kabaya and  M.  Tokuo: Hitz Giho, 74 (2013), 71 (in Japanese). https://www.hitachizosen.co.jp/hitz-tech/pdf/2013h25_3_04.pdf, (accessed 2022-10-19).
• 19)   T.  Sasaki,  H.  Kudo,  S.  Kawai,  Y.  Sato,  T.  Abe and  R.  Sugawara: Jpn. J. Soil Sci. Plant Nutr., 84 (2013), 367 (in Japanese). https://doi.org/10.20710/dojo.84.5_367
• 20)   M.  Miwa: EICA, 14 (2009), 27 (in Japanese). http://eica.jp/search/pdf_browse.php?file=c_14_1_27.pdf&id=766, (accessed 2024-11-06).
• 21)   S.  Kataki,  H.  West,  M.  Clatke and  D. C.  Baruah: Waste Manage., 49 (2016), 437. https://doi.org/10.1016/j.wasman.2016.01.003
• 22)   K.  Matsubae-Yokoyama,  H.  Kubo,  K.  Nakajima and  T.  Nagasaka: J. Ind. Ecol., 13 (2009), 687. https://doi.org/10.1111/j.1530-9290.2009.00162.x
• 23)   H.  Kubo,  K.  Matsubae-Yokoyama and  T.  Nagasaka: ISIJ Int., 50 (2010), 59. https://doi.org/10.2355/isijinternational.50.59
• 24)   H.  Ono,  A.  Inagaki,  T.  Masui,  H.  Narita,  T.  Mitsuo,  S.  Nosaka and  S.  Gohda: Tetsu-to-Hagané, 66 (1980), 1317 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.66.9_1317
• 25)   T.  Teratoko,  N.  Maruoka,  H.  Shibata and  S.-y.  Kitamura: High Temp. Mater. Processes, 31 (2012), 329. https://doi.org/10.1515/htmp-2012-0032
• 26)   H.-j.  Li,  H.  Suito and  M.  Tokuda: ISIJ Int., 35 (1995), 1079. https://doi.org/10.2355/isijinternational.35.1079
• 27)   K.  Nakase,  A.  Matsui,  N.  Kikuchi,  Y.  Miki,  Y.  Kishimoto,  I.  Goto and  T.  Nagasaka: J. Manuf. Sci. Prod., 13 (2013), 39. https://doi.org/10.1515/jmsp-2012-0038
• 28)   L.  Bilali,  M.  Benchanaa, K. El harfi, A. Mokhlisse and A. Outzourhit: J. Anal. Appl. Pyrolysis, 73 (2005), 1. https://doi.org/10.1016/j.jaap.2004.08.004
• 29)   T.  Iwama,  R.  Inoue,  K.  Nakase and  S.  Ueda: Tetsu-to-Hagané, 109 (2023), 1 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2022-061
• 30)   T.  Iwama,  R.  Inoue,  K.  Nakase and  S.  Ueda: ISIJ Int., 63 (2023), 799. https://doi.org/10.2355/isijinternational.ISIJINT-2022-450
• 31)   Y.  Liu,  T.  Qi,  J.  Chu,  Q.  Tong and  Y.  Zhang: Int. J. Miner. Process., 81 (2006), 79. https://doi.org/10.1016/j.minpro.2006.07.003
• 32)   S.  Aydoğan,  M.  Erdemoğlu,  A.  Aras,  G.  Uçar and  A.  Özkan: Hydrometallurgy, 84 (2006), 239. https://doi.org/10.1016/j.hydromet.2006.06.001
• 33)   S.  Aydogan,  G.  Ucar and  M.  Canbazoglu: Hydrometallurgy, 81 (2006), 45. https://doi.org/10.1016/j.hydromet.2005.10.003
• 34)   S.  Espiari,  F.  Rashchi and  S. K.  Sadrnezhaad: Hydrometallurgy, 82 (2006), 54. https://doi.org/10.1016/j.hydromet.2006.01.005
• 35)   M. G.  Bodas: Hydrometallurgy, 40 (1996), 37. https://doi.org/10.1016/0304-386X(94)00076-F
• 36)   L.  Wang,  C.  Liao,  Y.  Yang,  H.  Xu,  Y.  Xiao and  C.  Yan: J. Rare Earths, 35 (2017), 1233. https://doi.org/10.1016/j.jre.2017.07.001
• 37)   Ş.  Kaya and  Y. A.  Topkaya: Miner. Eng., 24 (2011), 1188. https://doi.org/10.1016/j.mineng.2011.05.004
• 38)   Y.  Ozaki,  T.  Okabe and  Y.  Kagawa: J. MMIJ, 130 (2014), 93. https://doi.org/10.2473/journalofmmij.130.93
• 39)   R.  Inoue and  H.  Suito: ISIJ Int., 46 (2006), 174. https://doi.org/10.2355/isijinternational.46.174
• 40)   M.  Hasegawa,  M.  Iwase,  K.  Wakimoto and  A.  McLean: Scand. J. Metall., 32 (2003), 47. https://doi.org/10.1034/j.1600-0692.2003.10610.x
• 41)   X.  Gao,  N.  Maruoka,  S.  Kim,  S.  Ueda and  S.-y.  Kitamura: J. Sustain. Metall., 1 (2015), 304. https://doi.org/10.1007/s40831-015-0030-8
• 42)   T.  Iwama,  C.-m.  Du,  X.  Gao,  S.-j.  Kim,  S.  Ueda and  S.-y.  Kitamura: ISIJ Int., 58 (2018), 1351. https://doi.org/10.2355/isijinternational.ISIJINT-2017-658
• 43)   T.  Iwama,  C.-m.  Du,  S.  Kozumi,  X.  Gao,  S.  Ueda and  S.-y.  Kitamura: ISIJ Int., 60 (2020), 400. https://doi.org/10.2355/isijinternational.ISIJINT-2019-298
• 44)   S.  Koizumi,  X.  Gao,  S.  Ueda and  S.-y.  Kitamura: Tetsu-to-Hagané, 105 (2019), 664 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2018-120
• 45)   S.  Koizumi,  X.  Gao,  S.  Ueda and  S.-y.  Kitamura: J. Sustain. Metall., 7 (2021), 87. https://doi.org/10.1007/s40831-020-00320-4
• 46)   S.  Koizumi,  X.  Gao,  S.  Sukenaga,  K.  Kanehashi,  T.  Takahashi,  S.  Ueda and  S.-y.  Kitamura: J. Sustain. Metall., 7 (2021), 583. https://doi.org/10.1007/s40831-021-00357-z
• 47)   C.-m.  Du,  X.  Gao,  S.  Ueda and  S.-y.  Kitamura: J. Sustain. Metall., 6 (2020), 724. https://doi.org/10.1007/s40831-020-00309-z
• 48)   P. C.  Hidber,  T. J.  Graule and  L. J.  Gauckler: J. Am. Ceram. Soc., 79 (1996), 1857. https://doi.org/10.1111/j.1151-2916.1996.tb08006.x
• 49)   H. M.  Azam,  S. T.  Alam,  M.  Hasan,  D. D. S.  Yameogo,  A. D.  Kannan,  A.  Rahman and  M. J.  Kwon: Environ. Sci. Pollut. Res., 26 (2019), 20183. https://doi.org/10.1007/s11356-019-04732-y
• 50)  U.S. Environmental Protection Agency: MINTEQA2/PRODEFA2, A Geochemical Assessment Model for Environmental Systems-User Manual Supplement for Version 4.0, https://www.epa.gov/ceam/minteqa2prodefa2-geochemical-assessment-modelenvironmental-systems-user-manual-supplement, (accessed 2022-12-10).
• 51)  The Geochemist’s Workbench: Thermodynamic datasets, https://www.gwb.com/thermo.php, (accessed 2022-12-10).
• 52)  Editorial Board of Introduction to Metallurgical Chemistry Series (Kinzoku-Kagaku-Nyumon-Shiri-zu-Henshu-Iinkai): Extractive Metallurgy (Kinzoku-Kagaku-Nyumon-Shiri-zu-Dai-3-Kan Kinzoku-Seiren-Kogaku), The Japan Institute of Metals (Nippon-Kinzoku-Gakkai), Sendai, (1999), 118 (in Japanese).
• 53)   K.  Ito,  M.  Yanagisawa and  N.  Sano: Tetsu-to-Hagané, 68 (1982), 342 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.68.2_342
• 54)   H.  Suito and  R.  Inoue: ISIJ Int., 46 (2006), 180. https://doi.org/10.2355/isijinternational.46.180