Determination of Fe₃O₄ Content and Total Nonhydraulic Minerals in Steel Slag

Abstract

The nonhydraulic minerals (Fe₃O₄, RO phase, Fe) in slag are important indicators for evaluating the pozzolanic activity and detecting the quality of the slag activation processing technology. Fe₃O₄ is an important characteristic mineral among the nonhydraulic minerals. In order to accurately assess the pozzolanic activity of steel slag powder and to monitor the quality of the activation process of steel slag powder for separate nonhydraulic minerals, it is imperative to precisely determine the nonhydraulic mineral content within the steel slag. Further refinement and enhancement are required for both the HNO₃ dissolution method used in determining Fe₃O₄ content in steel slag, as well as for the EDTA-DEA-TEA (ethylenediamine tetraacetate sodium-diethylamine-triethanolamine) dissolution method employed in determining total nonhydraulic minerals, due to potential deviations caused by challenging impurity separations. The results show that the content of Fe₃O₄ is determined by 10%HNO₃-20%NaOH-chemical analysis method, which solves the problem that the impurities of refractory materials (quartz, corundum, mullite) and amorphous phase affects the content determination in HNO₃ dissolution method. The total amount of nonhydraulic minerals (Fe₃O₄, RO phase, Fe) was determined by the EDTA-NaOH-TEA dissolution method, which solved the problem that the incomplete dissolution of C₂F in the EDTA-DEA-TEA dissolution method affected the content determination. The maximum error between the content determination value and the theoretical calculation value of the two methods is less than 0.50%. The improved Fe₃O₄ and total nonhydraulic mineral quantification methods are feasible and reliable.

1. Introduction

The steel slag is a by-product generated during the process of steel production in the industrial sector. As per the 2022 statistics from the National Bureau of Statistics, China’s annual crude steel output has reached 1.018 billion tons, resulting in an annual production of 146 million tons of steel slag; however, its resource utilization rate remains below 30% [1,2]. During smelting, various pollutants present in iron ore inevitably transfer to the steel slag, leading to their accumulation and causing significant environmental pollution in the surrounding areas [3,4].

The use of steel slag powder as a cement admixture or concrete additive is the main way to achieve high-value utilization of steel slag [5,6]. This is because steel slag powder contains mineral components similar to Portland cement, such as dicalcium silicate (C₂S) and tricalcium silicate (C₃S), which belong to hydraulic minerals and give steel slag certain cementitious activity [7,8]. However, there are also nonhydraulic minerals in it, such as Fe₃O₄, metallic Fe, and RO, which have high hardness, are difficult to grind, and have low reactivity. This results in slow hydration development of steel slag and lower early strength, which is the main reason for the low utilization rate of steel slag [9]. The mineral separation method is a continuously developing processing technology in the field of steel slag processing. Its objective is to separate nonhydraulic minerals (RO phase, Fe₃O₄, Fe) from crushed and ground steel slag. This process aims to enhance the relative content of hydrated active minerals (C₂S, C₃S) in steel slag powder and significantly improve its activity level. The highly active steel slag obtained after separation can achieve an activity index equivalent to S75 grade slag powder [10], while the nonhydraulic mineral product resulting from acetic acid solution leaching has a total iron (TFe) content of approximately 57.17%, meeting the minimum TFe content requirement (53.0%) specified in the iron concentrate standard, thus making it suitable for utilization as raw materials in iron production [11]. The accurate determination of the nonhydraulic mineral content is essential for evaluating the quality of both products and assessing the effectiveness of the separation process. This enables timely adjustment of separation process parameters, optimization of technical flow, and improvement in separation efficiency. Consequently, it provides a scientific basis for the application of steel slag powder.

The commonly used mineral phase quantitative methods are the XRD quantitative method, petrographic method, and chemical phase analysis. XRD quantitative methods mainly include an internal standard method, K value method, adiabatic method, and Rietveld full spectrum fitting method, etc. [12,13]. All these methods determine the phase content according to the principle that the phase diffraction peak intensity in the sample to be tested is proportional to the content. There are many kinds of mineral phases in steel slag, and the diffraction peaks of each mineral phase overlap with each other [14,15]. And the RO phase is the solid solution of bivalent metal oxides such as MgO, FeO, and MnO. These bivalent metal oxides have no definite stoichiometric ratio, so the RO phase has no definite diffraction peak and diffraction intensity, which is difficult to determine by the XRD method.

By observing the area fraction of each phase in multiple fields of view, the petrographic method can be statistically converted into mineral phase mass fraction, which has some disadvantages, such as time and effort and human factors [16,17]. Now, the petrographic analysis technology combines the automatic image analysis system (QEM) with the scanning electron microscope system (SEM), which greatly improves the efficiency of mineral phase analysis [18]. However, the quantification ratio of each bivalent metal oxide of the RO phase in steel slag cannot be determined, and there is no standard spectrum available at present that cannot be used for phase quantification in steel slag.

Selective dissolution by chemical reagents is one of the effective ways to obtain mineral phase content. Gutteridge W [19] and Hou [20] employed a salicylic acid-methanol (SAM) and potassium hydroxy-sucrose (KOSH) solution to dissolve the silicate phase and mesophase in steel slag, as well as extract the primary RO phase present in steel slag. Xu [21] further investigated the chemical phase analysis method for determining the presence of metal Fe, Fe₃O₄, and RO phases in steel slag. The metal Fe content in steel slag was selectively dissolved using iodoethanol, while no other minerals apart from metal Fe were solubilized by this solvent. The amount of dissolved substance corresponded to the metal Fe content. The Fe₃O₄ and refractory impurities are obtained by dissolving steel slag powder in a 10%HNO₃ solution. Subsequently, the beaker is placed on a strong magnet to allow the adsorption of Fe₃O₄ onto the magnetic pole surface. Horizontal shaking of the beaker facilitates the suspension of non-magnetic impurities above the mixed liquid, which can then be poured out. However, it was observed during practical operation that the complete separation of Fe₃O₄ and refractory impurities proved challenging. The total amount of nonhydraulic minerals was determined by dissolving steel slag powder with a composite solution of EDTA-DEA-TEA to obtain the RO phase, Fe₃O₄, Fe, and refractory impurities. It was observed that effectively separating nonhydraulic minerals and refractory impurities using the same physical separation method posed challenges. Therefore, further exploration is necessary to develop a determination method for Fe₃O₄ and total nonhydraulic minerals in steel slag.

Based on Hou’s proposed method for determining Fe₃O₄ and nonhydraulic mineral content, this paper aims to elucidate the impurity composition in the original method and further enhance and refine the technique to accurately determine the content of Fe₃O₄ and the total amount of nonhydraulic minerals in steel slag. The operational parameters of the optimized determination methods are determined through experiments, ensuring their accuracy is verified. An optimized method for determining the content of Fe₃O₄, supplemented with 20%NaOH dissolution and chemical analysis, eliminates the influence of insoluble minerals (refractory material impurities, amorphous phase) on the determination of Fe₃O₄ content in the 10%HNO₃ dissolution method. The optimized method for determining the total amount of nonhydraulic minerals, using NaOH instead of the original composite solution’s DEA, eliminates the influence of C₂F in the EDTA-DEA-TEA dissolution method on the determination of the total amount of nonhydraulic minerals. This study provides a theoretical foundation for evaluating the effectiveness of steel slag separation processes.

2. Materials and Methods

2.1. Raw Materials

The steel slag samples were provided by Shaanxi Longgang Co., Ltd., Xi’an, China, and the steel slag powder was obtained by vibration grinding until all the samples were passed through the 45 μm standard sieve and marked as S. The chemical composition of sample S is shown in

Table 1. Chemical composition of S/%.

Sample	Fe2O3	MgO	CaO	SiO2	Al2O3	MnO	P2O5	TiO2	Cr2O3	SO3	V2O5	Total
S	25.16	6.57	39.38	14.95	3.06	4.95	2.74	1.53	0.29	0.38	0.49	99.50

, and the mineral composition of S determined by XRD is shown in Figure 1a. The main minerals include calcium silicate (C₂S + C₃S), dicalcium ferrite (C₂F), Fe₃O₄, RO phase, and a small amount of metal Fe.

The Fe₃O₄ analysis with a pure optical microscope measured the maximum particle size of the powder sample as 20 μm, with purity ≥ 99.7% powder.

Metal iron powder (Fe) is a pure reagent for analysis; the maximum particle size of the sample measured under the microscope is 35 μm, with purity ≥ 99.8%.

The RO phase is a pure Fe-Mg RO phase with a FeO/MgO molar ratio of 1.28 synthesized at high temperature. The maximum particle size of the sample measured under an optical microscope is 25 μm.

Quartz, corundum, mullite pure powder, and the sample’s size were measured under an optical microscope range of 25 μm~45 μm, purity ≥ 99%, collectively referred to as refractory impurities. The mineral compositions of the three samples are shown in Figure 1b.

2.2. Sample Characterization

The chemical composition of steel slag powder was determined using an X-ray fluorescence spectrometer ( BRUKEROPTICS, Saarbrücken, Germany), while the mineral phase components of the sample were analyzed using a D/Max2200 X-ray diffractometer (Rigaku, Tokyo, Japan) operating at 40 kV and 26 mA. The scanning step size was set to 0.02°, with a scanning range of 5~90°and a scanning rate of 2°/min. A Field-Emission Scanning Electron Microscope (FEI Company, Portland, OR, USA) was utilized for the observation of the microscopic distribution and aggregation patterns of typical mineral phases in the samples, while an Energy Dispersion Spectrometer (FEI Company, Portland, OR, USA) was employed to determine the chemical element composition within microregions of these mineral phases. Multiple point scanning was conducted for each mineral phase, followed by averaging to obtain their respective chemical compositions. The TFe content in the sample was determined according to YB/T 148-2015 “Method for Determination of Total Iron Content in Steel Slag” [22].

3. Results and Discussion

3.1. Determination of Fe₃O₄ Content

3.1.1. Principle and Operation Procedure

• Principle

In a 10%HNO₃ solution, the insoluble minerals present in steel slag include Fe₃O₄, refractory impurities such as quartz, corundum, and mullite, as well as an amorphous phase formed during the dissolution process. Initially, the amorphous phase is eliminated using a 20%NaOH solution. Subsequently, the TFe content of the sample is determined through chemical analysis after removing the crystalline phase to calculate the Fe₃O₄ content.

• Operation procedure

The determination process of Fe₃O₄ is divided into the following three steps:

Accurately weigh 1 g of sample S and place it in a beaker containing 80 mL of 10%HNO₃ solution. Seal the beaker mouth with a plastic film and stir mechanically at 300 rpm for 3 h at a constant temperature of 30 °C. Filter the residue to obtain the sample residue, rinse it with distilled water several times, and dry it at 105 °C until constant weight.

The residue sample was mixed with 20%NaOH solution in a 100 mL beaker (Teflon) at a solid-to-liquid ratio of 1 g/4 mL. The beaker mouth was sealed with plastic film, and the mixture was mechanically stirred at 300 rpm at a constant temperature of 90 °C for 55~65 min. The filtered mixture was dried at 105 °C after being washed with deionized water at 75 °C several times. The weight of the residue was then measured.

The sample is fused with a mixture of sodium carbonate and boric acid after being dissolved in 10%HNO₃ and 20%NaOH. The fused material is dissolved in hydrochloric acid to form a solution. First, ammonium chloride is added, followed by ammonia water, to precipitate the iron and aluminum elements in the solution. The precipitate is dissolved in hydrochloric acid to form a 250 mL solution. A volume of 100 mL of the solution is then taken and titrated with EDTA standard solution. The TFe content is calculated based on the volume of EDTA standard solution consumed. The content of the sample Fe₃O₄ is equal to the content of the Fe element in the sample multiplied by 1.38.

3.1.2. Feasibility Analysis

• HNO₃ dissolution method

Sample S was dissolved in 10%HNO₃ solution at 30 °C for 3 h, and the residue sample obtained by extraction and filtration was labeled S₁, and the XRD pattern of S₁ was shown in Figure 2. The sample S₁ contains two main mineral phases: a crystalline phase and an amorphous phase. The crystalline phase consists of mullite (3Al₂O₃∙2SiO₂), quartz (SiO₂), corundum (α-Al₂O₃), as well as refractory impurities mixed with Fe₃O₄ from the iron and steel smelting process, while other minerals in the steel slag are completely dissolved. This finding is consistent with the literature [23]. The main chemical reaction equations during the dissolution of steel slag powder with a 10%HNO₃ solution are shown in Equations (1)–(5).

C₂S and C₃S react with HNO₃:

$${\mathrm{C}}_2\mathrm{S}+4\mathrm{H}\mathrm{N}{\mathrm{O}}_3\to 2\mathrm{C}\mathrm{a}{(\mathrm{N}{\mathrm{O}}_3)}_2+\mathrm{S}\mathrm{i}{\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O}$$

(1)

$${\mathrm{C}}_3\mathrm{S}+6\mathrm{H}\mathrm{N}{\mathrm{O}}_3\to 3\mathrm{C}\mathrm{a}{(\mathrm{N}{\mathrm{O}}_3)}_2+\mathrm{S}\mathrm{i}{\mathrm{O}}_2+3{\mathrm{H}}_2\mathrm{O}$$

(2)

C₂F reacts with HNO₃:

$${\mathrm{C}}_2\mathrm{F}+4\mathrm{H}\mathrm{N}{\mathrm{O}}_3\to 2\mathrm{C}\mathrm{a}{(\mathrm{N}{\mathrm{O}}_3)}_2+\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3+2{\mathrm{H}}_2\mathrm{O}$$

(3)

The main chemical composition of the RO phase is FeO and MgO, so the RO phase reacts with HNO₃:

$$\mathrm{M}\mathrm{g}\mathrm{O}+2\mathrm{H}\mathrm{N}{\mathrm{O}}_3\to \mathrm{M}\mathrm{g}{(\mathrm{N}{\mathrm{O}}_3)}_2+{\mathrm{H}}_2\mathrm{O}$$

(4)

$$\mathrm{F}\mathrm{e}\mathrm{O}+2\mathrm{H}\mathrm{N}{\mathrm{O}}_3\to \mathrm{F}\mathrm{e}{(\mathrm{N}{\mathrm{O}}_3)}_2+{\mathrm{H}}_2\mathrm{O}$$

(5)

In order to verify the preservation of Fe₃O₄ and refractory impurities in a 10%HNO₃ solution, four pure materials, including Fe₃O₄, quartz, corundum, and mullite, were dissolved in the aforementioned solution following the specified procedure. The residual rates of these materials were then determined and are presented in

Table 2. Residual ratio of 4 kinds of pure materials dissolved in HNO₃ solution.

Solution	Residual Ratio/%
	Fe3O4	3Al2O3·2SiO2	SiO2	α-Al2O3
10%HNO3	99.99	99.98	99.98	99.98

. It is noteworthy that all four pure materials exhibit residual rates equal to or exceeding 99.98%, indicating that no reaction occurred between them and the 10%HNO₃ solution at a temperature of 30 °C.

In the diffraction angle range of 16~30°, a diffuse peak envelope is observed, which corresponds to the X-ray scattering caused by the presence of an amorphous phase. This phenomenon can be attributed to the decomposition of silicate minerals within the reaction structure with HNO₃, resulting in relatively compact aggregation of SiO₂ particles and subsequent local concentration leading to dense precipitates. Additionally, newly formed silica gel is generated during dissolution. The bonding structure of aluminate minerals is mainly composed of silico–oxygen bonds and alumino–oxygen bonds, which exist in the form of [AlO₄] tetrahedra or [AlO₆] coordination polyhedra [23]. Under acidic conditions, the tetrahedral [AlO₄] or [AlO₆] species undergoes decomposition by H⁺ ions, resulting in the formation of small molecular groups such as Al(OH)₄⁺. Eventually, these dissolved Al(OH)₄⁺ species precipitate to form amorphous Al(OH)₃. The amorphous Al(OH)₃, along with silica gel and refractory impurities, is filtered together with Fe₃O₄ into the residue. This phenomenon has been confirmed through studies on the dissolution mechanism of steel slag minerals under acidic conditions [24,25].

In the original method [21], the beaker was placed on the magnetic pole surface of a strong magnet for physical separation. Due to the magnetic properties of Fe₃O₄, it would adhere to the magnet’s surface while non-magnetic impurities were suspended above the mixed liquid and then poured out through horizontal shaking by artificial means. However, during actual operation, amorphous phases adhered to Fe₃O₄ surfaces, making it difficult to separate refractory impurities from Fe₃O₄ particles, resulting in a high determination of Fe₃O₄ content. Therefore, further exploration and improvement of the method for determining Fe₃O₄ content is necessary.

• Removal of amorphous phase

The principle of alkaline dissolution for dissolving the amorphous phase is that OH^- can effectively destroy the amorphous phase structure formed under acidic conditions, causing it to decompose rapidly and completely. Sodium hydroxide solution shows good selective dissolution in mineral phase separation processes [26,27]. Many researchers have reported the dissolution mechanism of amorphous phases in NaOH solution [28,29], and the literature has verified through experiments that most of the amorphous phases in fly ash are dissolved in a 20%NaOH solution, with the SiO₂ leaching rate reaching a maximum of 42.13% [30].

The sample S₁ obtained after 10%HNO₃ treatment was dissolved in 20%NaOH solution to eliminate the influence of amorphous relative Fe₃O₄ content determination. Among them, the reaction equation of NaOH solution with silica gel and amorphous Al(OH)₃ is as follows:

$$2\mathrm{N}\mathrm{a}\mathrm{O}\mathrm{H}+\mathrm{S}\mathrm{i}{\mathrm{O}}_2\to \mathrm{N}{\mathrm{a}}_2\mathrm{S}\mathrm{i}{\mathrm{O}}_3+{\mathrm{H}}_2\mathrm{O}$$

(6)

$$\mathrm{N}\mathrm{a}\mathrm{O}\mathrm{H}+\mathrm{A}\mathrm{l}{\left(\mathrm{O}\mathrm{H}\right)}_3\to \mathrm{N}\mathrm{a}\mathrm{A}\mathrm{l}{\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O}$$

(7)

The change of mineral phase types of the sample after dissolution by 20%NaOH solution at 90 °C for 1 h is shown in Figure 3. By comparing the phase composition of the sample after dissolution of 10%HNO₃ solution and 10%HNO₃-20%NaOH solution, it is found that the peak envelope in the range of 16~30° disappears, and only the diffraction peak of crystal minerals is found in the figure.

In order to ensure that the NaOH solution does not affect mineral phases other than the crystal phase, four pure materials, including Fe₃O₄, quartz, corundum, and mullite, were dissolved in a 20%NaOH solution following the aforementioned procedure. The residual rates of these materials were determined and are presented in

Table 3. Residual ratio of pure material dissolved in 20%NaOH solution.

Solution	Residual Ratio/%
	Fe3O4	3Al2O3·2SiO2	SiO2	α-Al2O3
20%NaOH	99.98	99.97	99.98	99.97

. The dissolution rates of all four pure materials are found to be less than 0.03%. At a constant temperature of 90 °C, the 20%NaOH solution selectively dissolves only the amorphous phase in the residue; Fe₃O₄ does not react with the NaOH solution, while mullite can react and dissolve in dilute alkali solutions above 160 °C [30]. Corundum possesses a stable crystal structure and typically remains unreactive towards acids and bases at both room temperature and high temperatures. Quartz is also stable under normal pressure at 90 °C and exhibits low reactivity towards NaOH solution.

According to the mass ratio of m(Fe₃O₄):m(SiO₂):m(α-Al₂O₃):m(3Al₂O₃·2SiO₂) = 7:1:1:2, the mixed samples of four pure materials were prepared. The mixed samples were dissolved by 20%NaOH, and the quality of the mixed samples basically did not change before and after dissolution. It is proven that there is no mutual interference between mineral components in the process of determination. Therefore, it can be determined that the amorphous phase is the only soluble component of NaOH.

• Chemical analysis

Chemical analysis can calculate the content of mineral phases by determining the element content. For example, the f-CaO content in steel slag can be determined by the ethylene glycol-EDTA chemical analysis method. According to the principle of f-CaO reacting with ethylene glycol, calcium glycolate is generated, and the calcium glycolate content is determined by the consumption of EDTA, thereby allowing the f-CaO content in the slag to be measured [31]. Similarly, the f-MgO content is mainly determined by the ethylene glycol-iodine ethanol method and the ammonium nitrate-ethanol method [32,33]. However, the titration methods used for the two mineral phases above are for the determination of the total Ca and Mg content in slag, which leads to an overestimation of the calculated mineral phase content. In this study, the impurity mineral phases in the steel slag treated with 10%HNO₃-20%NaOH were all oxides of silicon and aluminum and did not contain other Fe-containing minerals. Therefore, the Fe₃O₄ content was calculated by determining the TFe content of the treated sample through chemical analysis, which was more accurate than other chemical analysis methods for determining the mineral phase content.

In order to validate the feasibility of the chemical analysis method, the sample with the amorphous phase removed was selected as the research object, labeled S₁′, and the impurity chemical composition of sample S₁′ was studied by XRD and SEM-EDS analysis. In order to enrich refractory impurities, wet magnetic separation was performed on sample S₁′, and a comparison was made between the intensity of diffraction peaks in XRD patterns obtained from raw materials before magnetic separation and tailings after magnetic separation (Figure 4). The intensity of the diffraction peak for Fe₃O₄ and refractory impurities shows a significant difference. Specifically, after magnetic separation, there is a decrease in the diffraction peak (35.44°PDF#97-015-9925) corresponding to the Fe₃O₄ mineral phase in the tailings compared with that of raw materials. On the contrary, there is an evident increase in diffraction peaks associated with quartz, corundum, and mullite mineral phases. This indicates that wet magnetic separation greatly affects Fe₃O₄ separation. Additionally, from Figure 4, it can be observed that refractory material impurities become more concentrated in tailings with higher relative content; notably absent are any discernible diffraction peaks related to other iron-containing minerals besides Fe₃O₄.

The SEM images depict the mineral phase of raw materials before magnetic separation (Figure 5a) and tailings after magnetic separation (Figure 5b). The chemical composition analysis conducted using EDS is presented in

Table 4. Chemical composition of 4 mineral phases.

Mineral	Chemical Formula
Quartz	SiO2·0.008Al2O3·0.002Fe2O3
Corundum	Al2O3·0.046SiO2·0.004Fe2O3
Mullite	3.000Al2O3·2.583SiO2·0.123Fe2O3
Fe3O4	Fe3O4·0.053SiO2·0.027Al2O3

. From Figure 5a, it can be observed that Fe₃O₄ is densely distributed throughout as bright white plates or small particles within a grayish-white base color. However, Figure 5b shows a black base color with a significant decrease in bright white mineral particles. This indicates that some Fe₃O₄ present in the tailings samples has been separated as a concentrate during the magnetic separation process, leading to reduced content within the tailings.

The refractory impurities of quartz and mullite are distributed more uniformly in the visual field. Under SEM, quartz appears as dark gray slab-shaped structures, while mullite exhibits a light gray distribution in small block or elliptical granular forms. Corundum is observed as light gray strips with relatively lower content distribution. According to the SEM image analysis, the refractory impurities in the tailings are enriched, and the types of impurity mineral phases do not change significantly. No other mineral phases are detected except quartz, corundum, and mullite, which is consistent with the XRD analysis results. Based on the chemical composition of these three refractory impurities’ mineral phases, it can be concluded that only a small amount of Fe₂O₃ exists in solid solution within these minerals (approximately 0.043 mol). Therefore, it is determined that these mineral phases are mixed into the steelmaking process or during stacking and transportation as refractory impurities rather than inherent impurities within the steel slag itself. Most of the Fe elements exist in the form of Fe₃O₄ mineral phases. Estimating Fe₃O₄ content by measuring TFe content is considered feasible.

The improved method was used to determine the Fe₃O₄ content in sample S, which was found to be 5.16%. To verify the applicability of the method to steel slag samples and the stability of the data, the following linear extrapolation mathematical model was used for calculation and verification. Steel slag powder was mixed with Fe₃O₄ sample (purity 99.7%) in a ratio of 9:1. The theoretical calculated value is 5.16% × 0.9 + 99.70% × 0.1 = 14.61%. When repeating the experiment three times, the actual average value determined was 14.54%. As shown in the Figure 6, the maximum error between the measured and theoretically calculated values is less than 0.50% for all three determinations. The measured values are stable and consistent with the calculated values, which proves that this method is applicable to the determination of Fe₃O₄ content in steel slag.

3.1.3. Parameter Selection

In order to design an appropriate process flow to achieve complete dissolution of the amorphous phase, it is crucial to clarify the dissolution characteristics of the amorphous phase under different conditions. The reaction temperature and reaction time are important operating parameters in the alkaline solution process. Research in reference [34,35] shows that when the reaction temperature is high (60 °C and 75 °C), the reaction degree of the amorphous phase of fly ash can reach the theoretical maximum value after about 336 h of reaction, while at lower temperatures (20 °C and 40 °C), the reaction degree can only reach 14.1% and 39.1%. Prolongation of the reaction time may promote the production of silicoaluminate and reduce the dissolution of the amorphous phase. In this paper, based on the existing literature, the experiment was conducted in the range of high temperatures, 60~90 °C, so that the dissolution rate was relatively improved so as to achieve the purpose of completely dissolving the amorphous phase in a short time.

To eliminate the influence of the amorphous phase on the determination of Fe₃O₄ content, the amorphous phase was dissolved in 20%NaOH solution. Sample S₁ is the residue sample obtained after the dissolution of 10%HNO₃ solution. Under the condition of an alkaline dissolution time of 1 h, the residual rate of sample S₁ is determined in the range of 60–100 °C, and the optimal alkaline dissolution temperature is determined. As can be seen from

Table 5. Alkaline dissolution temperature and residual ratio of samples S₁.

Sample	Residual Ratio/%
	60 °C	70 °C	80 °C	90 °C	100 °C
S1	68.76	67.23	65.32	64.15	64.15

, when it is set at 90 °C, the dissolution amount is basically stable, and the amorphous phase of the sample is completely dissolved, as detected by XRD. The maximum solubility has been reached.

Figure 7 shows the change of sample residual rate after sample S₁′ reacts in 20%NaOH solution at 90 °C for 20~90 min. It can be seen from the figure that the residual rate decreases with the increase of time within 65 min and reaches 69.73% with the increase of reaction time at 20 min. The residue rate decreased by 5.58% and reached the minimum residue rate of 64.15% at 55 min~65 min. However, after 65 min, the residue rate increased slightly. In order to analyze the reasons for the trend change, XRD analysis was conducted on the solid samples obtained at the above time points.

Figure 8 shows the XRD patterns of the products obtained with 20%NaOH solution at 90 °C at different reaction times. When the reaction time reached 20 min, the peak envelope uplift area collapsed obviously, and most of the amorphous phase reacted with NaOH into the solution. When at 55–65 min, the residual rate was stable and reached the maximum, and the amorphous phase was completely dissolved in 20%NaOH solution. When the reaction time reached 70 min, the diffraction peak of the aluminosilicate phase appeared. This is because, with the increase of alkali dissolution time after the amorphous phase is completely dissolved, the content of SiO₃²⁻ and AlO₂⁻ in the solution increases and further combines to form aluminosilicate [30], which increases the quality of the solid sample. In order to ensure that the amorphous phase is fully dissolved and prevent the secondary reaction from generating new impurity minerals, 55 min to 65 min is selected as the alkali dissolution time.

3.2. Determination of Total Nonhydraulic Minerals

3.2.1. Principle and Operation Procedure

• Principle

The original determination method used an EDTA-DEA-TEA composite solution to dissolve steel slag powder, and the incomplete dissolution of C₂F resulted in the total amount of nonhydraulic minerals and the determination error. On this basis, the optimization was carried out. The weakly alkaline DEA in the original composite reagent was replaced by a strongly alkaline NaOH solution, and the operation parameters were adjusted to achieve the purpose of eliminating C₂F.

• Operation procedure

Take a total of 55.80 g EDTA and 10.0 g NaOH, adding water to prepare 1000 mL of EDTA-NaOH solution. Take 167 mL of EDTA-NaOH solution, 33 mL TEA (1 + 2), and 200 mL H₂O, and adjust the pH value of the solution with NaOH solution (50 g∙L⁻¹) to the range of 12.50~13.00 under the instruction of the pH meter; a total of 2 g of sample S was immersed in an EDTA-NaOH-TEA solution, and the mouth of the beaker was sealed with a plastic film. The solution was mechanically stirred at 30 °C at a speed of 300 rpm for 1 h. Subsequently, the solution was pumped and filtered, and the resulting solid sample was then dried at a temperature of 105 °C. The weight of the dried sample was measured, and by subtracting the residual impurity content from the refractory impurity content, the total amount of nonhydraulic minerals present in the sample was calculated.

3.2.2. Feasibility Analysis

• EDTA-DEA-TEA dissolution method

The specific operation of the EDTA-DEA-TEA dissolution method is as follows: dilute 400 mL of the composite solvent (93 g EDTA, 173 mL DEA, 250 mL TEA) to 4800 mL at a dilution ratio of 1:12 (pH = 11.20), accurately weigh 5 g of steel slag powder sample, pour it into the prepared composite solution, seal the beaker mouth with plastic film to ensure its sealing, and stir mechanically at 300 rpm for 1.5 h at a constant temperature of 30 °C. Remove the beaker and place it on the magnetic pole of the strong magnet, wait for the residue to settle fully, pour out the residue and grind it to below 6 μm, then pour the ground residue back into the composite solution and continue to stir mechanically at 30 °C for 1.5 h. Filter to obtain the residue sample, rinse with distilled water several times, and dry at 105 °C to constant weight. The XRD pattern of sample S dissolved with EDTA-DEA-TEA solution is shown in Figure 9. After sample S is dissolved by EDTA-DEA-TEA composite solution for 3 h, the active mineral silicate phase (C₂S and C₃S) is reactively dissolved, but the characteristic diffraction peak of C₂F still exists, indicating that in addition to refractory impurities, the dissolved residue still contains C₂F, which affects the accuracy of total nonhydraulic mineral determination.

Therefore, after being dissolved by the EDTA-DEA-TEA composite solution, the steel slag powder contains insoluble refractory impurities (quartz, corundum, mullite) and incompletely dissolved inherent mineral components of C₂F co-existing with three nonhydraulic minerals. Because nonhydraulic minerals have magnetic properties, the original method placed a container containing the mixed solution on the surface of a strong magnet to physically separate the minerals. The nonhydraulic minerals adhered to the magnetic surface, while other mineral components were poured out artificially by shaking the mixture horizontally. However, during actual operation, it was found that the mineral phase particles were wrapped around each other, making it difficult to separate them, resulting in an overestimated total value of the nonhydraulic minerals, and the method needed to be further optimized.

• EDTA-NaOH-TEA dissolution method

According to Wang et al.’s [36] research, C₂F in slag is an active mineral, but its reaction activity is lower than that of C₂S and C₃S [37]. The composition of the complex solution, the pH value of the solution, and the reaction temperature are important parameters for the selective separation of minerals [38]; therefore, the complete dissolution of C₂F can be achieved by adjusting the solvent composition of the EDTA-DEA-TEA complex solution, the pH value of the solution, and the reaction temperature.

In aqueous EDTA, EDTA exists in seven forms: H₆Y²⁺, H₅Y⁺, H₄Y, H₃Y⁻, H₂Y²⁻, HY³⁻, and Y⁴⁻. Of these forms, only Y⁴⁻ can coordinate directly with metal ions. As the alkalinity of the solution increases, the concentration of Y⁴⁻ also increases, thus enhancing the complexing ability of metal ions. EDTA-NaOH-TEA solution can selectively dissolve clinker, gypsum and carbonate components in cement samples [39]. In the composite solution, the addition of NaOH reduces acidity and increases Ca²⁺ complexing capacity. Therefore, the addition of NaOH ensures a strongly alkaline environment, allowing EDTA to form a stable water-soluble complex with Ca²⁺. TEA is complexed with Al³⁺ and Fe³⁺ to promote the complete dissolution of the silicate phase and mesophase in steel slag.

XRD analysis was conducted on the solid samples obtained after dissolving the composite solution of EDTA-DEA-TEA and EDTA-NaOH-TEA, as depicted in Figure 10. The comparison reveals a complete disappearance of the diffraction peak corresponding to C₂F when using the EDTA-NaOH-TEA composite solution, indicating that all active minerals have been dissolved.

Pure materials of Fe₃O₄, RO phase, Fe, and three kinds of refractory impurities are dissolved by the above operation steps. The results of the residual rate are shown in

Table 6. Residual ratio of pure material dissolved in EDTA-NaOH.

Solution	Residual Ratio/%
	Fe3O4	RO Phase	Fe	3Al2O3·2SiO2	SiO2	α-Al2O3
EDTA-NaOH-TEA	99.98	99.95	99.96	99.97	99.95	99.96

. The dissolution rate of pure materials of three kinds of nonhydraulic minerals does not exceed 0.05%. The insoluble mineral dissolved by EDTA-NaOH-TEA solution is the combination of the total nonhydraulic mineral and the impurity of the refractory material. According to the dissolution method of 10%HNO₃-20%NaOH, the impurity content of refractory materials can be obtained, and the total amount of nonhydraulic minerals is subtracted.

The total amount of nonhydraulic minerals in sample S is 27.65% by the improved method. The following linear extrapolation mathematical model was used to calculate and test, and the steel slag powder was mixed with three nonhydraulic mineral pure materials at a ratio of 9:1 to check the stability and accuracy of the determination results.

The steel slag powder and RO phase pure sample (purity 100%) are mixed in proportion. The mixed sample is recorded as A1, and the theoretical calculation value is 27.65% × 0.9 + 100% × 0.1 = 34.89%. The average value of the mixed sample is 34.60%, and the error is 0.29%.

The steel slag powder and Fe₃O₄ pure sample (purity 99.7%) are mixed proportionally, the mixed sample is recorded as A2, and the theoretical calculation value is 27.65% × 0.9 + 99.70% × 0.1 = 34.86%. The average value of the mixed sample is 34.53%, and the error is 0.33%.

The steel slag powder and Fe sample (purity 99.8%) are mixed proportionally, the mixed sample is recorded as A3, and the theoretical calculation value is 27.65% × 0.9 + 99.80% × 0.1 = 34.87%. The average value of the mixed sample is 34.55%, and the error is 0.32%.

As shown in Figure 11, the error between the measured mean value and the calculated value of the above three mixed samples is less than 0.50%, which proves that the method is suitable for the determination of the total amount of nonhydraulic minerals in steel slag.

3.2.3. Parameter Selection

In the first experiment, the temperature was set at 75 °C, the pH value of the solution was adjusted to 11.6, three samples of 2 g S were weighed and poured into the solution for 1, 2, and 3 h, and then pumped and filtered, dried at 105 °C, weighed, and the residual rate was measured, as shown in

Table 7. Time and residual rate of S dissolution.

Sample	Residual Ratio/%
	1 h	2 h	3 h
S	33.40	32.81	32.80

. After 1 h, the dissolved amount of S still showed a slight increase. A small C₂F diffraction peak was also found in the XRD pattern of the sample after the reaction for 1 h. With the extension of the reaction time, the dissolution amount basically reached a stable level after 2 h, and C₂F was completely dissolved.

At 75 °C, water evaporation not only has a significant impact on the stability of the experiment but also consumes a lot, and the experimental conditions are harsh. In order to meet the needs of low consumption and simple experimentation, it is necessary to explore the operating conditions for the complete dissolution of C₂F at 30 °C, which is close to room temperature, by adjusting the alkalinity.

In experimental plan 2, the reaction temperature was set at 30 °C, while the pH of the solution was utilized as an indicator of alkalinity, with all other operational parameters remaining unchanged. The dissolution of sample S was measured under eight different pH conditions, as shown in Figure 12. When pH is increased from 11.60 to 12.50, the residual rate decreases from 37.67% to 32.80%. When pH is at three nodes below 12.50, 12.80, and 13.0, the residual rate stabilizes at about 32.80%, and the dissolved amount also reaches the maximum. Phase analysis was performed on the solid samples obtained by the reaction when the solution pH was 11.6, 12.8, and 14.0. As shown in Figure 13, when pH is 11.6, the diffraction peak of C₂F still exists, and the dissolution degree of C₂F is relatively low and cannot be completely dissolved.

When the pH reaches 12.8, C₂F disappears. On the one hand, as the pH increases, the alkalinity of the complex solution also increases, enhancing the concentration of Y⁴⁻ in EDTA aqueous solution and thereby improving the complexing ability of metal ions. On the other hand, TEA demonstrates its capability to form complexes with Ca²⁺ under alkaline conditions. Additionally, through their vacant electron pairs on oxygen and nitrogen atoms, TEA molecules can establish stable five-membered ring chelate structures with Ca²⁺, Fe³⁺, and Al³⁺. Under highly alkaline conditions, an increase in OH^- ions within the solution facilitates their combination with these three ions to form relatively stable complexes that promote silicate phase and mesophase dissolution in steel slag [40]. When pH increased to 14.0, a C₂F diffraction peak appeared. The reason was that when pH increased to 14, alkali concentration increased, NaOH particles were enriched on the surface of steel slag powder, and the reaction was difficult to continue, so the pH range was selected to be 12.50~13.00.

4. Conclusions

In this paper, based on the determination methods of Fe₃O₄ content and total nonhydraulic minerals proposed by Hou, the impurity composition in the original method is clarified, and the two determination methods are further improved and perfected. Specific conclusions are as follows.

The 10%HNO₃ dissolution method is used to determine the content of Fe₃O₄. In addition to Fe₃O₄, impurities in refractory materials (quartz, corundum, and mullite) and the amorphous phase produced during the dissolution process with HNO₃ also exist in the insoluble minerals.

The optimized Fe₃O₄ content determination method includes an additional step of 20%NaOH dissolution and chemical analysis. The amorphous phase is selectively dissolved by the 20%NaOH method. The solid sample is treated at 90 °C for 55~65 min, and the amorphous phase is completely dissolved. The TFe content of the sample after removing the amorphous phase is determined by chemical analysis, and the Fe₃O₄ content is calculated.

The total amount of nonhydraulic minerals was determined using the EDTA-DEA-TEA solution method. In addition to impurities in refractory materials, the incompletely dissolved C₂F also affects the accurate determination.

The optimized method for determining the total amount of nonhydraulic minerals uses NaOH instead of the original DEA in the composite solution. At 30 °C, the pH is adjusted to 12.50~13.00, and the sample is dissolved for 1 h. After C₂F is completely dissolved, the insoluble residue minus the refractory material impurity is the total amount of nonhydraulic minerals.