Activation Energy of Alumina Dissolution in FexO Bearing Slags

1. Introduction

In line with South Korea's 2030 carbon neutrality goal, the steel industry is also increasingly interested in electric arc furnaces, which emit less carbon dioxide than blast furnaces. Accordingly, research on electric arc furnace processes is necessary [1]. The electric arc furnace process injects oxygen into the steel to reduce operating time, oxidation refining, etc [2]. The presence of oxygen in the steel can cause problems such as corrosion and hot shortness. Therefore, a deoxidation process is essential [3]. The deoxidation process is mainly carried out using aluminum, which is a strong deoxidizer [4]. When tapping molten steel, aluminum is added to the ladle furnace to deoxidize it [5]. This process produces Al₂O₃ inclusions, which cause several problems including fatigue failure of the steel and nozzle clogging [6]. It is therefore important to remove Al₂O₃ inclusions as slag.

There are two ways to remove Al₂O₃ inclusions. The first is Ca treatment, where the Al₂O₃ inclusions are removed by adding Ca to transform the Al₂O₃ inclusion solid phase to a liquid CaO-Al₂O₃ phase [7]. However, this method has the disadvantage of interaction with the bottom lining refractory, which causes corrosion of the refractory. In addition, CaS inclusions, which are as harmful as Al₂O₃ inclusions, are easily formed [6]. The second is to float and separate Al₂O₃ inclusions to the slag/metal interface in a ladle refining process and then dissolve and remove them from the top layer of slag [8]. This method has been used not only in electric furnace processes, but also in blast furnace processes. For this reason, the dissolution behavior of non-metallic inclusions in slag has been extensively studied during the past 30 years.

Sridhar et al. [9] studied the dissolution behavior of Al₂O₃ particles in CaO-SiO₂-Al₂O₃-MgO slag with temperature as a variable. They reported that the dissolution of Al₂O₃ is dominated by mass transfer through the boundary layer. Q. Shu et al. [10] investigated the effect of Na₂O addition on the dissolution behavior of cylindrical Al₂O₃ in CaO-Al₂O₃-MgO-SiO ₂ slag. They reported that the dissolution mechanism of Al₂O₃ is boundary layer mass transfer, and the dissolution rate of cylindrical Al₂O₃ increases with a decrease of slag viscosity and increase of thermodynamic driving force with Na₂O addition. Yi. K. Wi et al. [11] investigated the dissolution behavior of Al₂O₃ and MgO particles within Al₂O₃-CaO-MgO slag, with temperature as the variable. Their research revealed that the dissolution kinetics of Al₂O₃ particles are primarily influenced by diffusion processes, whereas the dissolution of MgO particles adheres to chemical reaction kinetics. C. Ren et al. [12] examined how Al₂O₃ particles dissolve within CaO-Al₂O₃-SiO₂ slag, considering variations in slag composition and temperature. They reported that the rate-limiting step of Al₂O₃ particle dissolution in CaO-Al₂O₃-SiO₂ slag is diffusion in liquid slag, and an increase of C/A and C/S increased the dissolution rate of Al₂O₃ particles. L. Holappa et al. [13] studied the dissolution behavior of Al₂O₃ particles and MgO⋅Al₂O₃ particles in CaO-SiO₂-Al₂O ₃-MgO slag with basicity as a variable. They reported that Al₂O₃ particles and MgO⋅Al₂O₃ particles dissolve slowly in slag with low basicity and the dissolution rate of the particles increases in slag with low viscosity. H. Um et al. [5] investigated the dissolution behavior of Al₂O₃ particles in CaO-SiO₂-Al₂O₃-Fe_xO slag with Fe_xO content as a variable. They reported that as the Fe_xO content increased from 0 to 20wt%, the dissolution rate of Al₂O₃ particles increased due to the decrease in viscosity. However, when the Fe_xO content increased to 30wt%, the dissolution rate did not increase due to the formation of the CA6 phase at the Al₂O₃ particle interface. Y. Park et al. [14] studied the dissolution behavior of wall-type Al₂O₃ in CaO-Al₂O₃-Fe_xO-MgO-SiO ₂ slag with C/A and Fe_xO content as variables. They reported that increasing temperature, increasing C/A, and increasing Fe_xO content increased the dissolution rate of Al₂O₃. They also conducted temperature-variable experiments on slag having one composition to derive the mass transport activation energy of Al₂O₃ dissolution, which they reported to be 193.6 kJ/mole. S. Yeo et al. [15] conducted a study on the dissolution behavior of Al₂O ₃ particles in CaO-Al ₂O₃-SiO₂ slag with Al₂O₃ composition and temperature of the slag as variables. The investigators found that the dissolution rate increased with increasing activity of Al₂O₃. In addition, they reported that the diffusion activation energy of Al₂O₃ dissolution was in a range of about 320 to 490 kJ/mole depending on the composition of Al₂O₃ in the slag.

In addition, the dissolution behavior of Al₂O₃ in slags under various conditions has been studied [16,17,18,19,20]. However, since electric arc furnace slags use Fe scrap as raw material, the content of Fe_xO in the slag increases [2]. For this reason, the composition of the slag used in electric furnace research must include FexO. However, there has been little research on the dissolution behavior of non-metallic inclusions in slag containing Fe_xO. In addition, the temperature of electric arc furnace slag varies from 1550 °C to 1700 °C, but few studies have been conducted with temperature as a variable in slag containing Fe_xO [21]. For this reason, the dissolution behavior of Al₂O₃ particles in CaO-SiO₂-Al₂O₃-Fe_xO slag was studied in this work with temperature and Fe_xO content as variables.

2. Materials and Methods

2.1. Sample Preparation

Table 1 shows the chemical composition and diameter of the Al₂O₃ particles used in the experiment. The shape of Al₂O₃ particles is spherical, with a diameter of 500 ± 2.5 μm (Goodfellow Cambridge limited, Huntingdon, UK). The average weight of the particles was 0.25 mg and the purity of Al₂O₃ is 99.9%.

Table 2 shows the chemical composition and basicity of the slag used in the experiments. The basicity was fixed at 1, and the Fe_xO content was varied from 10 to 20 wt%. The amount of slag used in each experiment was set to 4mg for stability of the experiment. The slag was prepared by mixing CaO powder prepared by calcining CaCO₃ at 1200 °C with Al₂O₃ and Fe_xO powder individually and melting them in a high frequency induction furnace.

2.2. Single Hot Thermocouple Apparatus (SHT Apparatus)

In this study, we observed the dissolution behavior of Al₂O₃ particles using a SHT apparatus. Figure 1 is a schematic diagram of the SHT apparatus. It consists of a B-type thermocouple to melt slag and dissolve Al₂O₃ particles, a DHT controller to control and check the temperature of the B-type thermocouple, a video camera to observe and record the dissolution behavior of Al₂O₃ particles in real time, and an optical microscope.

The SHT apparatus has several advantages, including the ability to inject Al₂O₃ particles at the desired temperature range, quenching at 300 °C/s using the DHT controller, ease of preparing quenching specimens, and real-time observation of the melting behavior of Al₂O₃ particles with a video camera and optical microscope.

2.3. Experimental Conditions

Figure 2 shows the process of the Al₂O₃ particle dissolution experiment using the SHT apparatus. First, the + and – poles of the B-type thermocouple are welded to form an oval shape, and then mounted on the copper tip. The slag is then placed on the B-type thermocouple and the temperature is raised to 8 °C/s using a DHT controller to melt the slag. When the experimental temperature (1550, 1575, 1600 °C) is reached, Al ₂O₃ particles are added to the melted slag to dissolve it. Subsequently, after the setting time (120, 240, 360 sec), quenching was performed using the DHT controller.

As mentioned earlier, the SHT apparatus has the advantage of allowing observation of the dissolution behavior of Al₂O₃ particles in real time. Several studies using this apparatus have taken advantage of this by observing the dissolution behavior of inclusions in real time [5,15,22,23]. However, in this study, there was a problem that the dissolution behavior of Al₂O₃ particles could not be observed in real time due to the opacity of the slag at high temperature because it contains Fe_xO, a transition metal. To solve this problem, in this experiment, quenching specimens were prepared by exploiting one of the advantages of the SHT apparatus, i.e., that quenching at 300 °C/s is possible. The quenching specimens were then polished and the diameter was measured in four directions, as shown in Figure 3a, to check the dissolution rate of Al₂O₃ particles. However, using this method, only a small portion of the diameter of the Al₂O₃ particles may be exposed. For this reason, the largest diameter value found by repeating the polishing several times was set as the representative value of the Al₂O₃ particle dissolution rate, as shown in Figure 3b.

The quenching specimens were also analyzed by SEM to identify the reaction layer at the interface between the slag and Al₂O₃ particles. No compounds were observed at the interface in the specimens.

3. Results and Discussion

3.1. Dissolution Behavior of Al₂O₃ Particles according to Temperature and Fe_xO content in Slag

The SHT apparatus was used to assess the dissolution behavior of Al₂O₃ particles in slag with changing Fe_xO content. The experiment temperatures were 1550, 1575, and 1600 °C, and the experiment times were 120, 240, and 360 seconds for each condition. Experiments were performed at least three times for each condition for reproducibility.

The variation of the diameter of Al₂O₃ particles in slags with different Fe_xO content at each temperature is shown in Figure 4. Slag0 and Slag3 are based on previous papers [5,15]. In the case of slag3, it was not possible to conduct experiments at temperatures above 1550 °C due to the short circuit of B-type thermocouples at high temperatures. Also, in the case of slag0, the temperature deviation was increased by 50 °C, and hence there are no data at 1575 °C.

Figure 4a shows the dissolution behavior of Al₂O₃ particles at 1550 °C. The diameter of the Al₂O₃ particles decreases linearly with the dissolution time, and the dissolution rate increases as the Fe_xO content in the slag increases. However, for slag3, the dissolution rate did not increase with increasing Fe_xO content. These experimental results are ascribed to the generation of the CA6 phase at the interface of Al₂O₃ particles and slag under the experimental conditions of slag3, which changed the dissolution process of the particles into an inter-compound chemical reaction [5]. Figure 4b shows the dissolution behavior of Al₂O₃ particles at 1575 °C. The results of the experiment at 1575 °C show that the diameter of the Al₂O₃ particles decreases linearly with the dissolution time, and the dissolution rate increases with an increase of Fe_xO content in the slag. Figure 4c shows the dissolution behavior of Al₂O₃ particles at 1600 °C. The results of the experiment at 1600 °C show that the diameter of the Al₂O₃ particles decreases linearly with the dissolution time, and the dissolution rate increases with an increase of Fe_xO content in the slag. However, at 1600 °C, in the slag2 experiment, the Al₂O₃ particles completely dissolved before 360 seconds. For this reason, additional experiments were conducted at 300 seconds for an accurate interpretation.

Table 3 lists the particle diameters of the Al₂O₃ particles for each condition.

3.2. Analysis of Slag/Al₂O₃ Particle Interface by SEM

As previously described, H. Um et al. [5] reported that for slag3, a CA6 phase was created at the interface of slag and Al₂O₃ particles, which changed the dissolution process of Al₂O₃ particles. In addition, Park et al. [8] reported that a ring-shaped compound was formed along the particle/slag interface depending on the slag composition. For this reason, to better understand the dissolution behavior of Al₂O₃ particles in the experimental slag, the interface of slag and Al₂O₃ particles was analyzed by SEM. Figure 5 shows images of the interface of slag1, slag2, and Al₂O₃ particles under the condition of 1550 °C.

It can be seen that no compounds are formed at the interface of slag1, 2, and Al₂O₃ particles. It was also confirmed that no compounds were formed at the interface at 1550℃ or at 1575 °C and 1600 °C. S. Yeo et al. [15] and Taira et al. [24] reported that the dissolution rate of Al₂O₃ in CaO-SiO₂-Al₂O₃ slag is controlled by diffusion in the boundary layer. Also, H. Um et al. [5] reported that the dissolution of Al₂O₃ particles in CaO-SiO₂-Al₂O₃-Fe_xO slag is controlled by diffusion in the boundary layer of slag and Al₂O₃ particles if no compounds are generated at the interface of slag and Al₂O₃ particles. Therefore, it is believed that for the dissolution mechanism of Al₂O₃ particles in slag1 and slag2 it is only necessary to consider the behavior by diffusion in the boundary layer.

3.3. Dissolution Mechanism of Al₂O₃

Solid Al₂O₃ particles can be dissolved by liquid slag through the following process [5].

$${Al}_2{O}_{3\left(s\right)}={Al}_2{O}_{3\left(l\right)} \dots \mathrm{Chemical} \mathrm{kinetics} \mathrm{of} \mathrm{the} \mathrm{reaction} \mathrm{at} \mathrm{the} \mathrm{interface}$$

(1)

$${Al}_2{O}_{3\left(l\right)}=\left({Al}_2{O}_3\right) \dots \mathrm{Liquid} \mathrm{phase} \mathrm{mass} \mathrm{transfer}$$

(2)

In other words, the dissolution of solid Al₂O₃ particles in liquid slag can be controlled by a chemical reaction or liquid phase mass transfer [8,24,25,26]. It is interpreted that the dissolution of Al₂O₃ particles in liquid slag is controlled by liquid phase mass transfer in the boundary layer unless a compound is generated at the interface of Al₂O₃ particles and slag [9,27,28]. As noted above, no compounds were generated at the interface of Al₂O₃ particles and slag under the present experimental conditions. Therefore, the rate-controlling step of Al₂O₃ particle dissolution in this experiment can be interpreted as liquid phase mass transfer at the boundary layer.

If the rate-controlling step in the dissolution of Al₂O₃ in slag is liquid phase mass transfer in the boundary layer, then the relationship between the mass transfer flux and mass transfer coefficient of Al ₂O₃ particle dissolution can be expressed by the mass transfer equation as follows [29].

$$J=-k(C_i-C_b)$$

(3)

where J is the mass transfer flux; k is the mass transfer coefficient in the slag; C_i and C_b are the Al₂O₃ content at the interface and in bulk slag; and ${(C}_i-C_b)$ is the driving force of the dissolution.

If the Al₂O₃ particles being dissolved are spherical, then Equation (3) can be transformed into the following dissolution rate equation [29].

$$\frac{dr}{dt}=-k(C_i-C_b)M/\rho$$

(4)

where r is the radius of the Al₂O₃ particles; dr/dt is the dissolution rate; M is the molecular weight of Al₂O₃; and $\rho$ is the slag density.

The dissolution rate can be calculated from the experimental data, as shown in Figure 6, and the slag density and driving force of the dissolution can be obtained using Factsage^7.3TM, and hence the mass transfer coefficient according to temperature and Fe_xO content can be derived using Equation (4). The physical properties and mass transfer coefficients for each temperature and Fe_xO content are summarized in Table 4.

The mass transfer coefficient increases with the Fe_xO content in the slag and with increasing melting temperature.

3.4. Activation Energy

The activation energy of mass transfer, the dissolution mechanism of Al₂O₃, can be quantitatively measured. In this study, the dissolution rate of Al₂O₃ particles was measured with temperature as a variable, and the mass transfer coefficient was calculated accordingly. By graphing the reciprocal of the mass transfer coefficient and temperature using the Arrhenius equation, which is expressed as follows, the activation energy for mass transfer can be derived [30].

$$k=k_0\cdot \mathrm{e}\mathrm{x}\mathrm{p}(-\frac{E_k}{RT})$$

(5)

where k is the mass transfer coefficient; k₀ is the pre-exponential constant; R is the universal gas constant; T is the absolute temperature; and E_k is the activation energy of mass transfer.

To explain how E_k is derived, substituting logarithms into the above expression, we can express it as a function of ln(k) and temperature, as follows.

$$\ln k=-\frac{E_k}{R}\frac{1}{T}+\ln k_0$$

(6)

This expression tells us that the slope of the ln(k)-1/T graph is $-\frac{E_k}{R}$. Therefore, multiplying this value by -R gives the activation energy for mass transfer. Cho et al. calculated the mass transfer coefficient by measuring the dissolution behavior of Al₂O₃ in CaO-SiO₂-Al₂O₃ slag with temperature as a variable and derived the E_k of Al₂O₃ dissolution using the above method [29]. In this study, E_k was also calculated through the above process and is shown in Figure 7.

The E_k of Slag1 was calculated to be 159 kJ/mole and the E_k of Slag2 was calculated to be 182 kJ/mole. In addition, the E_k of slag with Fe_xO calculated in this study and the E_k values of slag without Fe_xO calculated in other studies are summarized in Table 5.

From Table 5, the E_k of the experimental slag with Fe_xO is lower than that of the slag without Fe_xO, indicating that the Al₂O₃ dissolution in the slag with Fe_xO is faster compared to that in the slag without Fe_xO.

4. Conclusions

In this study, the dissolution behavior of Al₂O₃ was studied by varying the Fe_xO content in the slag from 10 to 20wt% and the temperature as 1550 °C, 1575 °C, and 1600 °C. The results were as follows.

(1)

The dissolution rate increased linearly as the Fe_xO content of the slag increased from 0 to 20 wt% and the dissolution temperature increased from 1550 to 1600 °C.

(2)

Through a SEM analysis it was observed that no compound was formed at the interface of Al₂O₃ particles and slag. Therefore, the rate step of Al₂O₃ particle dissolution is interpreted as liquid phase mass transfer.

(3)

The mass transfer coefficient was obtained using the dissolution rate equation. The mass transfer coefficient increased with increasing Fe_xO content in the slag and increasing dissolution temperature.

(4)

The mass transfer coefficient is plotted in a graph as a function of temperature and the E_k values of slag1 and slag2 (159 and 182 kJ/mole, respectively) are found using the Arrhenius equation.

(5)

The E_k of Al₂O₃ mass transfer in slag containing Fe_xO in this study was lower than the E_k of slag without Fe_xO.