Dissolution Rate and Interfacial Behaviors of Alumina Particle in Molten Slag Studied by Single Hot Thermocouple Technique

Abstract

In view of the deficiencies of traditional methods to study the dissolution kinetics of solid oxides in molten slag, a novel method based on single hot thermocouple technique (SHTT) is proposed in this paper. The feasibility of this method is verified and the controlling means of experiment reproducibility is explored. And the effect of slag basicity and Li₂O content on the dissolution behavior of Al₂O₃ in mold flux is investigated via SHTT. The results show that: 1) Under the condition that the density of the Al₂O₃ particle is slightly higher than that of slag and the mass ratio of Al₂O₃ particle to liquid slag is less than 2%, the relative standard error of Al₂O₃ dissolution rate is within 10%. 2) The effect rule of slag basicity on the Al₂O₃ dissolution rate studied via SHTT is same as that of rotating cylinder method. The dissolution rate of Al₂O₃ increases with the increase of slag basicity. When the basicity increases from 1.0 to 1.2, the dissolution rate increases significantly, which is due to the formation of xCaO∙yAl₂O₃ or xCaO∙yAl₂O₃∙zSiO₂ intermediate compounds with low melting point at the Al₂O₃ boundary layer. 3) With the increase of the Li₂O content in mold flux, the dissolution rate of Al₂O₃ first increases and then decreases. The decrease in dissolution rate is caused by the formation of high-melting MgAl₂O₄ at the boundary of Al₂O₃ particle.

1. Introduction

In the steel production process, it is a quite common phenomenon that solid oxides contact with molten slag and be dissolved, resulting in a series of physical and chemical reactions. For instance, the liquid slag erodes the refractory materials such as blast furnace, converter, ladle, tundish and nozzle. The commonly used refractory materials are Al₂O₃ and MgO. Likewise, the solid slag-making materials interact with the liquid slag during steelmaking and secondary refining. For another example, the high-melting nonmetallic inclusions (The most common inclusion is Al₂O₃) are absorbed by molten slag such as refining slag and mold flux during secondary refining and continuous casting. The temperature control curve can be set arbitrarily according to requirement and the dissolution behaviors of solid oxides in molten slag at any temperature and time can be tested and analyzed, that is the research of the dissolution kinetics of solid oxides in molten slag. That will be of great significance to the control of refractory quality, the development of new refractory and the control of the cleanliness of liquid steel.

At present, there are two main methods to study the dissolution kinetics of solid oxides in molten slag, rotating cylinder method^1,2,3,4,5) (RCM) and confocal scanning laser microscope^{6,7,8,9,10,11)} (CSLM). Several researchers have used these two methods to investigate the dissolution kinetics of solid oxides including Al₂O₃ in molten slag. RCM is to immerse the Al₂O₃ rod into molten slag, rotate it according to set rate, measure the size and weight of the Al₂O₃ rod regularly, and calculate the average dissolution rate of Al₂O₃. CSLM is to place a spherical Al₂O₃ particle in molten slag and observe the dissolution process of Al₂O₃ directly from above through a confocal scanning laser microscope. Although these two methods have been widely recognized and applied, there are still some deficiencies about them. Among them, RCM cannot be used for in-situ observation. Although CSLM can realize in-situ observation, however, its facility is expensive and complicated to operate. In addition, CSLM requires that the molten slag does not contain transition metal oxides such as MnO, otherwise the slag is opaque to the laser and the solid oxide particle cannot be observed. Taking consideration of the deficiencies of the aforementioned methods, it is significant to propose a new method which can overcome those deficiencies.

The single hot thermocouple technique (SHTT) is frequently applied to observe the melting and crystallization of slag. Its main features are as follow: (i) The technology of video display and image recognition are combined in this technique. And the whole process of experiment is displayed in real time by video, consequently the in-situ observation can be realized. (ii) The facility is easy to operate and low in cost. (iii) A platinum-rhodium thermocouple is used as both heating component and temperature measuring component in this technique and thus the temperature control precision is reliable. The maximum temperature can reach 1800°C.¹²⁾ The heating and cooling rate are fast, so that the experimental time can be shorten (The heating rate is 0–30°C/s,^13,14) the cooling rate can reach 300°C/s). (iv) Because of the thin slag film in this technique, the slag has a good transparency, which may be able to make up for the problem that the CSLM cannot observe the oxide particle in the slag containing transition metal oxides such as MnO. Furthermore, the in-situ sample that Al₂O₃ particle be dissolved in molten slag can be obtained via cooling at a rate of 300°C/s. And the microcosmic morphology and interfacial behaviors of Al₂O₃ particle can be detected by means of the scanning electron microscope (SEM), electron probe microanalysis (EPMA) or some other microanalysis approaches. Based on the characteristics above, SHTT is capable of making up for some deficiencies of RCM and CSLM. Therefore, in the previous work,¹⁵⁾ the SHTT has been attempted to evaluate the dissolution rate of solid oxide in molten slag.

In the present work, the dissolution of solid Al₂O₃ inclusion in the mold flux based on CaO–SiO₂ system is researched, which aims to establish a method based on SHTT to investigate the dissolution rate of solid oxide in slag. On this basis, the effect of slag basicity and Li₂O content on the dissolution rate and interfacial behaviors of alumina in mold flux are investigated.

2. Experiment

2.1. Experimental Principle

The schematic diagram of SHTT is shown in Fig. 1. The principles of this technique have been introduced in detail by Kashiwaya.¹³⁾

Fig. 1.

The schematic of SHTT. (Online version in color.)

During experiments, the dissolution process of Al₂O₃ particle is present on the screen in the form of video through image acquisition system, and it is recorded at a rate of 1 sheet per second like Fig. 2 (Fig. 2(a) shows the slag without transition metal oxide, and Fig. 2(b) shows the slag containing MnO). The size information of the Al₂O₃ particle can be obtained via common graphics processing software (The scale of each picture is determined by a standard Pt–Rh thermocouple adopted in experiments, and the diameter of Pt–Rh wire is 500 μm, as shown in Fig. 3). And the number in the lower left corner of each picture indicates the time information of dissolution process. According to the size and time information of the dissolution of Al₂O₃ particle, the dissolution rate v (g.cm⁻².min⁻¹) can be calculated.

Fig. 2.

Dissolution of alumina particle in molten slag. (Online version in color.)

Fig. 3.

Standard Pt–Rh thermocouple. (Online version in color.)

In this study, the dissolution rate v is represented by the dissolved mass of Al₂O₃ per unit time per unit area. At a certain moment, the instantaneous dissolution rate of Al₂O₃ can be expressed by Eq. (1):   

$$v=\frac{dW}{Adt}$$(1)

where W (g) is the mass of the Al₂O₃ particle, A (cm²) is the contact area between the Al₂O₃ particle and molten slag. Since the Al₂O₃ particle is completely immersed in molten slag (The thickness of the slag film is about 1.8 mm, which is much larger than the diameter of Al₂O₃ particle), A is the surface area of Al₂O₃ particle.

When Bui⁴⁾ and Lee¹⁶⁾ investigated the dissolution kinetics of Al₂O₃ in molten slag via rotating cyclinder method, the Eq. (2) was used to caculate the average dissolution rate of Al₂O₃.   

$$\overline{v}=\frac{\mathrm{\Delta }W}{A_0{t}_0}$$(2)

where ΔW (g) presents the mass change of the Al₂O₃ rod before and after dissolution, A₀ (cm²) is the initial contact area between the Al₂O₃ rod and molten slag.

In the rotating cyclinder method, since the surface area of the Al₂O₃ rod changes little before and after dissolution, the contact area in Eq. (2) is regarded as a constant in the calculation process. However, in this study, due to the small size of the Al₂O₃ particle, the surface area changes significantly before and after dissolution. As the surface area decreases, the dissolution rate will gradually increase. If the contact area A is regarded as a constant, the calculation result of the average dissolution rate of Al₂O₃ will be smaller than the actual value. In order to obtain a result close to the actual situation, the average surface area of Al₂O₃ particle before and after dissolution is used in this study to calculate the average dissolution rate of Al₂O₃, as shown in Eq. (3).   

$$\overline{v}={\displaystyle \frac{\mathrm{\Delta }W}{{\displaystyle \frac{\left(A_0+0\right)}{2}}{t}_0}}={\displaystyle \frac{2\mathrm{\Delta }W}{A_0{t}_0}}$$(3)

where ΔW (g) presents the mass change of Al₂O₃ particle before and after dissolution, that is the total mass of the Al₂O₃ particle. A₀ (cm²) is the initial surface area of the Al₂O₃ particle. Because the density and size of the Al₂O₃ particle are known, ΔW and A₀ can be expressed as Eqs. (4) and (5).   

$$\mathrm{\Delta }W=\frac{4}{3}\pi R_0^3\rho$$(4)

  

$$A_0=4\pi R_0^2$$(5)

Hence, the average dissolution rate of Al₂O₃ can be expressed as Eq. (6).   

$$\overline{v}=\frac{2\rho R_0}{3t_0}$$(6)

where R₀ is the initial radius of Al₂O₃ particle, t₀ is the total time of dissolution process, ρ is the density of Al₂O₃ particle.

2.2. Experimental Scheme

2.2.1. Sample Design

In this study, the dissolution of alumina inclusion in mold flux based on CaO–SiO₂ system is investigated, and the purpose of this study is to establish a new method based on the SHTT to study the dissolution kinetics and interfacial behaviors of solid oxides in molten slag. The feasibility of this new method should be determined first. The effect of slag basicity on the dissolution rate of Al₂O₃ has been studied by several researchers,^4,17,18) and the experimental results and theories can be used for reference. Therefore, the feasibility of this method is verified via the effect of slag basicity on the dissolution rate of Al₂O₃ in molten slag. And the fluxes for experiments were designed as No. 1–5 in Table 1.

Table 1. Chemical compositions of mold fluxes (wt%).

	CaO	SiO2	CaF2	F−	Na2O	Al2O3	MgO	Li2O	R((CaO+ 0.72CaF2)/SiO2)
1	22.60	47.05	14.35	7	8	5	3	–	0.7
2	25.20	44.43	14.35	7	8	5	3	–	0.8
3	27.56	42.09	14.35	7	8	5	3	–	0.9
4	29.66	39.99	14.35	7	8	5	3	–	1.0
5	33.29	36.36	14.35	7	8	5	3	–	1.2
6	29.66	39.99	14.35	7	6	5	3	2	1.0
7	29.66	39.99	14.35	7	4	5	3	4	1.0
8	29.66	39.99	14.35	7	3	5	3	5	1.0
9	29.66	39.99	14.35	7	2	5	3	6	1.0

Li₂O is a strong fluxing agent which can significantly reduce the melting point and viscosity of mold flux. However, some researchers^19,20,21) have found that the crystallization of slag will be promoted when the Li₂O content in molten slag is overly high, which may affect the dissolution of Al₂O₃. This study attempts to explore the effect of Li₂O on dissolution kinetic of Al₂O₃ via SHTT. Therefore, the No. 4 and No. 6–9 fluxes in Table 1 were designed for experiments.

2.2.2. Density of the Al₂O₃ Particle

The density²²⁾ of molten slag in continuous casting is generally 2.5–2.7 g/cm³. The Al₂O₃ particle will settle in short order after entering the molten slag when the density of the Al₂O₃ particle is much higher than that of slag. As shown in Fig. 4, due to the curvature of the slag surface, only the magnified image of Al₂O₃ particle can be observed from above under the refraction of light. And the true size of the Al₂O₃ particle cannot be obtained because of the unknown curvature of slag surface. Therefore, the density of the Al₂O₃ particle should be slightly higher than that of slag when the SHTT is applied to study the dissolution kinetics of Al₂O₃ in molten slag. In this study, the sintered Al₂O₃ particles with a density of 3.15 g/cm³ were selected for experiments.

Fig. 4.

The refraction path of light in liquid slag.

2.2.3. Effect of the Mass Ratio of Al₂O₃ Particle to Liquid Slag on Experimental Reproducibility

According to Eq. (4), the dissolution rate of Al₂O₃ particle in molten slag is related to the concentration gradient between the interface and slag inside. During the continuous casting, the volume of liquid slag in mold is sufficient to absorb the inclusions coming from steel. Due to the low mass ratio of Al₂O₃ inclusions to liquid slag, the dissolution of Al₂O₃ inclusions will not influence the concentration of Al₂O₃ component in molten slag significantly and thus affect the concentration gradient. However, because of the limitation of the Pt–Rh thermocouple size in this study, the mass of liquid slag used in each experiment is only about 3–4 mg. When the size and mass of the Al₂O₃ particle are overly large, the dissolution of Al₂O₃ will markedly increase the concentration of Al₂O₃ component in the slag, which will lead to a decrease in the concentration gradient of Al₂O₃ between the interface and slag inside. And the diffusion of the Al₂O₃ component in the slag will be hindered, resulting a decrease of the dissolution rate of Al₂O₃ particle. It may even happen that the Al₂O₃ particle cannot be completely dissolved. Therefore, the mass ratio of the Al₂O₃ particle to slag has an influence on the experimental reproducibility. Since the density of Al₂O₃ particles is constant and the volume of slag used in each experiment is stable within a small range, the effect of the mass ratio of Al₂O₃ particle to slag on the experimental reproducibility can be regarded as the effect of Al₂O₃ particle size. Experiments were conducted based on No. 4 slag to investigate the effect of Al₂O₃ particle size on the average dissolution rate.

In order to reduce the difficulty of operation and facilitate the observation of the dissolution process of Al₂O₃, the Al₂O₃ particle should not be excessively tiny. Therefore, the size of Al₂O₃ particles selected in this study is 212–500 μm.

2.3. Experimental Procedure

Analytical reagents with a purity of 98% were adopted to prepare fluxes in the light of Table 1. Each flux was thoroughly mixed into a graphite crucible and homogenized in a silicon-molybdenum resistance furnace at 1300°C. Then the homogenized slag was poured into a water-cooled copper to quench it. The quenched samples were cooled in air after being dried at a 100°C oven. Finally, each sample was ground in an agate mortar to 200 mesh for SHTT.

The thermocouple was bent into a convex shape to facilitate adding powder sample (To keep the consistency of the temperature field, a shaped mold is used to ensure that the shape of the thermocouple in each experiment remained consistent). The powder sample was moistened with alcohol hence it can be adhered to the surface of thermocouple. A spoon with constant volume was used to add sample so that the mass of slag in each experiment was approximate. The temperature control curve of experiments is shown in Fig. 5. There are two main purposes of heat preservation at 1500°C for 100 s. One is to reduce bubbles in the slag to prevent affecting the observation. The other is to homogenize the temperature and ingredient of molten slag. The temperature of molten slag was remained at 1450°C during the dissolution process of Al₂O₃ particle. The Al₂O₃ particle was added into the molten slag via a long Pt–Rh wire. After the Pt–Rh wire was wetted with alcohol, the Al₂O₃ particle can adhere to the Pt–Rh wire due to the interfacial tension between alcohol and Al₂O₃ particle. Then the Al₂O₃ particle was moved above the slag, and the alcohol evaporated rapidly due to high temperature. Thus, the Al₂O₃ particle fell to the surface of molten slag and entered the slag.

Fig. 5.

Temperature control curve.

Since the Al₂O₃ particle cannot be guaranteed to keep spherical during dissolution process, the equivalent diameter was used to present the size of Al₂O₃ particle during data processing. The commonly used equivalent diameters are equivalent volume diameter, equivalent surface area diameter and equivalent projection area diameter. In this study, because the Al₂O₃ particle can only be observed from above, the equivalent projection area diameter is used to represent the equivalent diameter of the Al₂O₃ particle.

The specific method to obtain the equivalent projection area diameter of Al₂O₃ particle is as follows. The experimental pictures were processed via a common graphics processing software to obtain the projection area S of Al₂O₃ particle. And the radius R can be calculated via Eq. (7).   

$$\text{R}=\sqrt{\frac{S}{\pi }}$$(7)

In order to observe the interfacial behaviors of the Al₂O₃ particle during dissolution process, it is necessary to obtain the in-situ samples of the Al₂O₃ particle dissolved in slag. The specific method of sample preparation is as follows. The temperature curve was set according to Fig. 5. The slag was heat to the target temperature, and then an Al₂O₃ particle was add into the slag. After the Al₂O₃ particle was dissolved for 40 s, the temperature control program was immediately stopped, and the sample was quickly cooled to room temperature at a rate of up to 300°C/s. The sample was embedded into the base of epoxy resin, and the surface was smoothed until the Al₂O₃ particle was exposed. After the sample was sprayed with gold, a FESEM was used to detect it.

3. Results and Discussion

3.1. Effect of Al₂O₃ Particle Size on the Experimental Reproducibility

The effect of Al₂O₃ particle diameter on the dissolution rate of Al₂O₃ is shown in Fig. 6. The average dissolution rate of Al₂O₃ increases correspondingly with the decrease of the diameter of Al₂O₃ particle. And the average dissolution rate tends to be stable when the diameter of Al₂O₃ particle is less than 350 μm. At this point, the mass ratio of the Al₂O₃ particle to molten slag is about 2%. Therefore, the diameter of Al₂O₃ particle should not exceed 350 μm when the dissolution kinetics of Al₂O₃ in molten slag is studied via SHTT, and the mass ratio of Al₂O₃ particle to molten slag is less than 2%.

Fig. 6.

Effect of particle size on the average dissolution rate of alumina.

3.2. Effect of Basicity on the Dissolution Rate of Al₂O₃ Particle

The Al₂O₃ particles with a density of 3.15 g/cm³ and a diameter of less than 350 μm were used to test the dissolution rate of Al₂O₃ in No. 1 to 5 slags in Table 1 respectively, and each group of slags was tested four times. The average dissolution rate of Al₂O₃ in each slag and the relative standard error were calculated, and the results are shown in Table 2. It can be seen from the calculation results that the relative standard error of the dissolution rate of Al₂O₃ in each slag is within 10% when the diameter of the Al₂O₃ particle is controlled less than 350 μm.

Table 2. Average dissolution rate of Al₂O₃ and relative standard error.

Basicity	0.7	0.8	0.9	1.0	1.2
Average dissolution rate 1 (g.cm−2.min−1)	0.0074	0.0094	0.0140	0.0308	0.1076
Average dissolution rate 2 (g.cm−2.min−1)	0.0072	0.0082	0.0134	0.0326	0.1152
Average dissolution rate 3 (g.cm−2.min−1)	0.0078	0.0096	0.0152	0.0354	0.1208
Average dissolution rate 4 (g.cm−2.min−1)	0.0086	0.0102	0.0146	0.0314	0.1060
Relative standard error (%)	7.99	8.97	5.42	6.27	6.13

The relationship between the average dissolution rate of Al₂O₃ and the slag basicity is plotted in Fig. 7. The average dissolution rate of Al₂O₃ gradually increases as the slag basicity increases, which is the same as the previous studies of several researchers via other methods.^4,16,18) According to the structural theory of silicate slag,^23,24,25) the viscosity characteristics of molten slag mainly depend on the connection degree of the Si–O tetrahedron network. When the oxygen-silicon ratio (OSI) in slag gradually increases, the connection mode of the Si–O network transitions from skeleton, layer, and chain to island, and the viscosity decreases accordingly. As the basicity of the molten slag increases, the oxygen-silicon ratio in the slag increases, and the viscosity of the molten slag decreases. That facilitates the mass transfer in slag, resulting in an increase in the dissolution rate of Al₂O₃. Furthermore, Fig. 7 indicates that the dissolution rate of Al₂O₃ increases slowly when the slag basicity increases from 0.7 to 1.0. However, when the basicity of slag increases from 1.0 to 1.2, the dissolution rate of Al₂O₃ increases significantly. In order to study the cause of this phenomenon, the interfacial behaviors of Al₂O₃ particles during dissolution process were investigated.

Fig. 7.

Effect of slag basicity on the dissolution rate of alumina. (Online version in color.)

The Al₂O₃ particle was respectively dissolved in the slag with basicity of 0.8 and 1.2 for 40 s, and the in-situ samples were obtained via a cooling rate of 300°C/s. The FESEM was used to detect the samples, and the results are shown in Figs. 8 and 9. When the Al₂O₃ particle is dissolved in the slag with a basicity of 0.8, the boundary of the Al₂O₃ particle is tight and the boundary with slag is distinct. When the slag basicity is 1.2, the boundary of Al₂O₃ particle is loose and a transition layer can be observed around the Al₂O₃ particle. The composition of the transition layer was measured by EDS, and the results are shown in Fig. 10. The main elements of the transition layer were Al, O and Ca. As shown in Fig. 9, the atomic number percentage was measured at three points in the transition layer, and the results are shown in Table 3. The main elements of 1# are Al, O, and Ca; the main elements of 2# are Al and O; and 3# is mainly composed of Al, O, Ca, and Si. The results indicate that xCaO∙yAl₂O₃ or xCaO∙yAl₂O₃∙zSiO₂ intermediate compounds exist in the transition layer. The phenomenon of the precipitation of xCaO∙yAl₂O₃ and xCaO∙yAl₂O₃∙zSiO₂ was also found in previous studies.^4,16,26) Therefore, when the Al₂O₃ is dissolved in the slag with a basicity of 1.0–1.2, it may react with the CaO and SiO₂ in the slag to form low-melting intermediate compounds, resulting in a significant increase in the dissolution rate of Al₂O₃.

Fig. 8.

Morphology of alumina particle in the molten slag with a basicity of 0.8.

Fig. 9.

Morphology of alumina particle in the molten slag with a basicity of 1.2.

Fig. 10.

EDS result of alumina particle boundary in the slag with a basicity of 1.2. (Online version in color.)

Table 3. Compositions of transition layer (%).

Point	O	Al	Ca	Si	Mg	F	Na
1#	41.46	49.51	8.39	0.35	0.09	0	0.20
2#	48.90	47.19	0.90	0.28	1.37	0	1.37
3#	42.70	40.88	7.43	4.58	2.02	0.19	2.20

3.3. Effect of Li₂O Content on the Dissolution Rate of Al₂O₃

The influence of Li₂O content in No. 4 and No. 6–9 slags on the dissolution rate of Al₂O₃ is shown in Fig. 11, and the dissolution rate of Al₂O₃ was tested three times in each slag respectively. It can be known from Fig. 11 that the dissolution rate of Al₂O₃ increases first and then decreases as the Li₂O content increases. In order to research the cause of above phenomenon, the in-situ samples were prepared respectively in the slags with Li₂O content of 2%, 4%, and 6%. And the microcosmic interfaces of Al₂O₃ particles were observed via the FESEM, the results are shown in Fig. 12.

Fig. 11.

Effect of Li₂O content in molten slag on the dissolution rate of alumina. (Online version in color.)

Fig. 12.

Morphology of alumina particles in the molten slag with different Li₂O content. (a)-2% Li₂O, (b)-4% Li₂O, (c)-6% Li₂O.

As shown in Fig. 12, when the Li₂O content in molten slag is 2%, no evident intermediate compound is observed at the boundary of Al₂O₃ particle. When the Li₂O content in slag is 4%, there are fine flocculent substances at the boundary of Al₂O₃ particle. And there are some agglomerates precipitate at the boundary of Al₂O₃ particle when the Li₂O content in slag is 6%. The boundary layer of Al₂O₃ particles were detected via EDS to affirm the compositions of the precipitation, and the results are shown in Figs. 13, 14 and 15. According to the results, the main elements of the boundary of Al₂O₃ particle are Al and O when the Li₂O content in slag is 2%. Hence the main composition of boundary is Al₂O₃. When the Li₂O content in slag is 4%, the main elements of the flocculent substance at the boundary are Al, O, and Mg, so the composition of flocculent substance is presumed to be MgAl₂O₄. When the Li₂O content is 6%, the main elements of the agglomeration precipitating at the boundary of Al₂O₃ particle are Al, O, and Mg. And the agglomeration shows obvious shape characteristic of high-melting MgAl₂O₄. Therefore, it is considered that the agglomeration is also MgAl₂O₄. The similar phenomenon that Al₂O₃ reacts with MgO in the molten slag to precipitate MgAl₂O₄ was also found by Kim²⁷⁾ and Valdez¹⁸⁾ in their studies.

Fig. 13.

EDS result of alumina particle boundary in the slag containing 2% Li₂O. (Online version in color.)

Fig. 14.

EDS result of alumina particle boundary in the slag containing 4% Li₂O. (Online version in color.)

Fig. 15.

EDS result of alumina particle boundary in the slag containing 6% Li₂O. (Online version in color.)

Combining the change law of Al₂O₃ dissolution rate with the Li₂O content of slag in Fig. 11 and the interfacial behaviors of Al₂O₃ particles in Figs. 12, 13, 14, 15, the effect of Li₂O content on the dissolution rate of Al₂O₃ can be analyzed. Li₂O can significantly reduce the melting point and viscosity of the slag.^{21,28,29,30,31)} When the Li₂O content in molten slag is low, it is conducive to the mass transfer in the slag as the Li₂O content increases, thus the dissolution of Al₂O₃ is promoted. When the Li₂O content in molten slag exceeds 4%, high-melting MgAl₂O₄ precipitates at the boundary of the Al₂O₃ particle, which forms a barrier layer around the Al₂O₃ particle and hinders the diffusion of solutes in the slag. Hence the dissolution rate of Al₂O₃ shows a decreasing trend.

It is worth noting that the MgO content in the slag in this study is only 3%, and the CaO content is about 30%. And the high-melting MgAl₂O₄ precipitates under the condition of such low MgO content, the cause of this phenomena should be discussed. According to the slag ion structure theory, the basic ions that make up slag at high temperature are simple ions Ca²⁺, Mg²⁺, O²⁻, F⁻, etc., and complex anions SiO₄⁴⁻, PO₄³⁻, AlO₂⁻, etc. And the ions in molten slag can move freely. According to their respective electrostatic potentials, the distribution of cations and anions in the slag shows microscopic inhomogeneity and becomes the ordered ionic groups that form compounds. These ordered ionic groups follow the principle that the cations and anions with high electrostatic potential distribute together to form strong ion pairs, while the cations and anions with low electrostatic potential distribute together to form weak ion pairs. Among them, the electrostatic potential of the basic oxide Mg²⁺ is 3.08, which is obviously higher than that of other basic oxide cations, i.e., higher than Ca²⁺ (1.89), Na⁺ (1.05), K⁺ (0.72) and Li⁺ (1.67). And some scholars³²⁾ have found that AlO₂⁻ has a higher electrostatic potential than SiO₄⁴⁻. Therefore, according to the principle of strong ion pair, Mg²⁺ tends to form ordered ionic groups with AlO₂⁻ at the interface of Al₂O₃ particle in this study, resulting in the precipitation of MgAl₂O₄. As shown in Fig. 16,³³⁾ the quaternary phase diagram of CaO-SiO₂-Al₂O₃-5MgO shows that there is a MgAl₂O₄ spinel region when the Al₂O₃ content becomes high in the slag containing 5% MgO. Comprehensive analysis shows that there is a strong thermodynamic trend of MgAl₂O₄ precipitating at the interface of the Al₂O₃ particle in this study. Therefore, with the increase of Li₂O content, the melting point and viscosity of slag decreases significantly, which leads to the increase of the ion migration rate in slag. When the Li₂O content in the slag exceeds 4%, the migration rate of Mg²⁺ can meet the kinetic conditions of MgAl₂O₄ precipitation. Thus, the MgAl₂O₄ spinel with high melting point precipitates at the interface of Al₂O₃ particle, which becomes a barrier layer to hinder the Al₂O₃ particle in dissolution. Consequently, the dissolution rate of Al₂O₃ shows a decreasing trend.

Fig. 16.

CaO-SiO₂-Al₂O₃-5MgO quaternary diagram.

The above results indicate that a high-melting MgAl₂O₄ barrier layer can precipitate around the Al₂O₃–C nozzle with high Al₂O₃ content when a certain amount of MgO is added to the mold flux during the continuous casting. That can reduce the erosion speed of the nozzle, and increase the service life of the nozzle. This is especially important for the mold fluxes for high-speed continuous casting that pursue low melting point and low viscosity. On the contrary, for the continuous casting process of the high-aluminum steel with [Al]> 0.5%, [Al] in the liquid steel will inevitably react with the SiO₂ in the mold flux. A large amount of Al₂O₃ will enter the slag to change the viscosity and crystallization performance of slag, which affect the lubrication performance of the slag. In order to prevent the precipitation of high-melting MgAl₂O₄ in the slag with high Al₂O₃ content and further deteriorate the lubrication performance of the slag, the MgO content in mold flux must be strictly controlled.

4. Conclusions

The dissolution of solid Al₂O₃ inclusion in the mold flux based on CaO–SiO₂ system is investigated in this study, which aims to establish a method based on SHTT to investigate the dissolution rate of solid oxide in slag. On this basis, the effect of slag basicity and Li₂O content on the dissolution rate and interfacial behaviors of alumina in mold flux are investigated. The main conclusions are as follows.

(1) The density of Al₂O₃ particles should be slightly higher than that of molten slag when the dissolution kinetics of Al₂O₃ in molten slag was studied via SHTT. The equivalent diameter of the Al₂O₃ particle should be less than 350 μm, and the mass ratio of the Al₂O₃ particle to molten slag should not exceed 2%. The relative standard error of experiment under these conditions is within 10%. The influence of slag basicity on the dissolution rate of Al₂O₃ studied via SHTT is same as that of previous studies.

(2) The dissolution rate of Al₂O₃ in molten slag increases gradually as the slag basicity increases. The dissolution rate of Al₂O₃ increases slowly when the slag basicity is 0.7–1.0, and it increases significantly when the slag basicity is 1.0–1.2. In the slag with a basicity of 1.0–1.2, the dissolution of Al₂O₃ is promoted through the formation of the low-melting xCaO∙yAl₂O₃ and xCaO∙yAl₂O₃∙zSiO₂ which generated via the reaction of CaO, SiO₂ and Al₂O₃.

(3) The dissolution rate of Al₂O₃ increases first and then decreases with the increase of the Li₂O content in slag. The decrease of the dissolution rate is caused by the high-melting MgAl₂O₄ precipitating around the boundary of Al₂O₃ particle when the Li₂O content in slag exceeds 4%.

Acknowledgments

The authors wish to thank Chongqing University (in China) and the National Natural Science Foundation of China [grant number 51574050] for the financial support.

References

• 1)   X.  Yu,  R. J.  Pomfret and  K. S.  Coley: Metall. Mater. Trans. B, 28 (1997), 275.
• 2)   S.  Xiang,  X.  Lv,  B.  Yu,  J.  Xu and  J.  Yin: Metall. Mater. Trans. B, 45 (2014), 2106.
• 3)   Q.  Shu,  Y.  Wang,  J.  Li,  Y.  Liu,  P.  Li and  K.  Chou: ISIJ Int., 55 (2015), 2297.
• 4)   A. H.  Bui,  H. M.  Ha,  I. S.  Chung and  H. G.  Lee: ISIJ Int., 45 (2005), 1856.
• 5)   S.  Jansson,  V.  Brabie and  P.  Jönsson: Scand. J. Metall., 34 (2005), 283.
• 6)   S.  Sridhar and  A. W.  Cramb: Metall. Mater. Trans. B, 31 (2000), 406.
• 7)   J.  Liu,  F.  Verhaeghe,  M.  Guo,  B.  Blanpain and  P.  Wollants: J. Am. Ceram. Soc., 90 (2007), 3818.
• 8)   F.  Verhaeghe,  J.  Liu,  M.  Guo,  S.  Arnout,  B.  Blanpain and  P.  Wollants: J. Appl. Phys., 103 (2008), 023506.
• 9)   A. B.  Fox,  M. E.  Valdez,  J.  Gisby,  R. C.  Atwood,  P. D.  Lee and  S.  Sridhar: ISIJ Int., 44 (2004), 836.
• 10)   J. Y.  Choi,  H. G.  Lee and  J. S.  Kim: ISIJ Int., 42 (2002), 852.
• 11)   M.  Valdez,  K.  Prapakorn,  A. W.  Cramb and  S.  Seetharaman: Steel Res., 72 (2001), 291.
• 12)   J. H.  Welch: J. Sci. Instrum., 31 (1954), 458.
• 13)   Y.  Kashiwaya,  C. E.  Cicutti,  A. W.  Cramb and  K.  Ishii: ISIJ Int., 38 (1998), 348.
• 14)   L.  Zhou,  W.  Wang,  F.  Ma,  J.  Li,  J.  Wei,  H.  Matsuura and  F.  Tsukihashi: Metall. Mater. Trans. B, 43 (2012), 354.
• 15)   F.  Chen,  P.  Tang,  G.  Wen and  S.  Gu: 149th Annual Meeting and Exhibition of the Minerals, Metals and Materials Society, Springer, Cham, (2019), 57.
• 16)   M.  Lee,  S.  Sun,  S.  Wright and  S.  Jahanshahi: Metall. Mater. Trans. B, 32 (2001), 25.
• 17)   S.  Taira,  K.  Nakashima and  K.  Mori: ISIJ Int., 33 (1993), 116.
• 18)   M.  Valdez,  K.  Prapakorn,  A. W.  Cramb and  S.  Sridhar: Ironmaking Steelmaking, 29 (2002), 47.
• 19)   B.  Lu,  K.  Chen,  W.  Wang and  B.  Jiang: Metall. Mater. Trans. B, 45 (2014), 1496.
• 20)   H.  Liu,  G.  Wen and  P.  Tang: ISIJ Int., 49 (2009), 843.
• 21)   L.  Zhou,  H.  Li,  W.  Wang,  D.  Xiao,  L.  Zhang and  J.  Yu: Metall. Mater. Trans. B, 49 (2018), 2232.
• 22)   K. C.  Mills and  C.-Å.  Däcker: The Casting Podwers Book, Springer International Publishing, Cham, (2017), 352.
• 23)   D. S.  Sanditov and  A. A.  Mashanov: Glass Phys. Chem., 36 (2010), 41.
• 24)   J. S.  Cheng,  Z. F.  Xiao,  K.  Yang and  H.  Wu: Ceram. Int., 39 (2013), 4055.
• 25)   L.  Coudurier,  D. W.  Hopkins and  I.  Wilkomirsky: Fundamentals of Metallurgical Processes, Pergamon Press, London, (1978), 191.
• 26)   J.  Li,  Q.  Shu,  Y.  Liu and  K.  Chou: Ironmaking Steelmaking, 41 (2014), 732.
• 27)   M. S.  Kim,  S. W.  Lee,  J. W.  Cho,  M. S.  Park,  H. G.  Lee and  Y. B.  Kang: Metall. Mater. Trans. B, 44 (2013), 299.
• 28)   G. R.  Li: J. Iron Steel Res. Int., 10 (2003), 6.
• 29)   H.  Kim and  I.  Sohn: ISIJ Int., 51 (2011), 1.
• 30)   G.  Wen,  S.  Sridhar,  P.  Tang,  X.  Qi and  Y.  Liu: ISIJ Int., 47 (2007), 1117.
• 31)   H. Y.  Chang,  T. F.  Lee and  T.  Ejima: ISIJ Int., 27 (1987), 797.
• 32)   M. E.  Grice,  J. S.  Murray and  P.  Politzer: J. Mol. Graph., 12 (1994), 169.
• 33)   D.  Springorum: Slag Atlas, Verlag Stahleisen, Düsseldorf, (1995), 160.