Surface Tension and Related Ions Behavior of Silicate Melts of the Na₂O–K₂O–SiO₂–CaF₂ System under Ultrasonic Irradiation

Abstract

Under ultrasonic irradiation, surface tension and structure of molten slags of Na₂O–K₂O–SiO₂–CaF₂ system were measured. Results showed that, the slags were degraded gradually with the increase of ultrasonic intensity and silica tetrahedron radicals with oxygen vacancies generated in the melts. In fluorine free slag, the dissociated radicals coordinate with free oxygen anions, which lead to the decrease of electrostatic force between ions and the related decrease of surface tension. While in fluorine bearing slag, the surface tension increases for the reason that the free fluorine anions dominate the behavior of coordination, which lead to the increase of electrostatic force between ions. The increase of surface tension will promote the down flow of mold flux between the gap of solidified steel shell and the mold copper wall, which will facilitate the lubrication between them.

1. Introduction

Surface tension of molten slag is one of the most important physical properties, which measures the tendency of metallurgy phenomena occurring in metallurgy processes such as extraction, refining, casting. Typically, during the continuous casting of liquid steel, the entrapment of liquid mold flux into the liquid steel and the gap filling of liquid mold flux between the solidified shell of slab/billet and the copper cooling wall are closely related to the surface tension, which should be regulated properly avoiding defects in slab/billet.^1,2)

Liquid mold flux is a typical silicate melt, in which, Si⁴⁺ act as network formation cations commonly coordinate with four oxygen anions forming tetrahedron of SiO₄⁴⁻, which can further connect together forming chain-type or ring-type species via bridging oxygen (BO). The other cations of Ca²⁺, Na⁺, Mg²⁺ act as network modifiers maintaining charge balance with anions.

Ultrasonic is known as the cavitation effect, which causes complex sonochemistry effects induced by high-speed liquid jets and strong shear forces, and has been applied in many industrial processes for dispersion, crushing and degradation.^{3,4,5,6,7,8,9,10,11,12)} Especially, organic polymers with macromolecules can be degraded to radicals under ultrasonic irradiation accompanying with variation in physical properties of density, viscosity and surface tension.^{13,14,15,16,17,18,19,20,21)} Considering the similarities between silicate melts and organic polymers in both structure and bond energy (Si–O vs. C–C or C–N),²²⁾ we predicted and validated the degradation of the silicate melts under ultrasonic.²³⁾

Surface tension is a direct measure of the interionic forces acting at the surface of liquid, which are governed by the structure in the melt essentially.²⁴⁾ The degradation of structure under ultrasonic irradiation will surely lead to the variation of surface tension. In this study, the structure and surface tension of the melts of Na₂O–K₂O–SiO₂–CaF₂ system under ultrasonic irradiation were measured, and the ion behavior and its effect on the surface tension were discussed.

2. Materials and Methods

Two types of fluorine free and bearing slag of the Na₂O–K₂O–SiO₂–CaF₂ system were designed, and four intensities were selected for ultrasonic irradiation. The slag compositions, the experimental system, and the methods and operation for slag sampling, surface tension detection, and Raman spectra detection are as follows.

2.1. Slag Composition

Mold flux is one kind of typical silicate melt bearing fluorine, which commonly composes of oxides and fluorides of alkali and alkaline earth metals. In this study, the experimental slag (marked B) of the Na₂O–K₂O–SiO₂–CaF₂ system was designed according to the content of industrial mold flux. To clarify the effect mechanism of fluorine on the variation of surface tension, fluorine free slag (marked A) was also designed considering the similar melting temperature with the industrial mold flux. Table 1 shows the details of the two experimental slags. The melting temperature of slag A and slag B is 1173 K and 1179.5 K respectively, and the theoretical polymerization of slag A and slag B representing as NBO/T (the number of no-bridging oxygen per tetrahedron) is 2 and 1 respectively.

Table 1. Chemical compositions of experimental slags (mass, %). #, the deigned composition; +, the average composition of three experiments under different ultrasonic intensity before surface tension measurement; &, the average composition after surface tension measurement. After surface tension detection, the contents of Li₂O, Na₂O and CaF₂ decrease slightly from that of before surface tension detection, this may be the reason of evaporation of alkali metal oxides and fluoride under high temperature. In theory, the content decrease of alkali metal oxides and fluoride will lead to an increase in polymerization, which is different from the depolymerization results under ultrasonic field, so, the effect of variation of content on the slag structure and surface tension is not discussed in this paper.

Slag	Na2O			K2O			SiO2			CaF2
	#	+	&	#	+	&	#	+	&	#	+	&
A	30.00	29.22	28.56	25.00	24.36	24.07	45.00	46.21	47.34	/	/	/
B	12.95	12.55	11.74	14.00	13.38	13.15	43.05	44.67	45.33	30.00	29.32	28.89

2.2. Experimental System

As shown in Fig. 1, three subsystem for slag heating, ultrasonic treating and surface tension detecting are assembled together. The heating system consists of computer for operating software to control the temperature and the heating rate of the furnace, MoSi₂ heating bars and a proportional integral differential (PID) controller connecting with a Pt–Rh thermocouple for a constant temperature fluctuation within 2 K.

Fig. 1.

Experimental system for ultrasonic treatment and surface tension detection. A, ultrasonic probe, the semi-diameter of the cylindrical probe is 25 mm. The semi-diameter and the depth of the cylindrical groove is 18 mm and 70 mm, respectively. During experiment, the molten slag level is about 40 mm; B, molybdenum cylindrical tube; C, molten slag; D, molybdenum ring linked wire; E, furnace tube; F, MoSi₂ heater; G, Pt-Rh thermocouple; H, insulating brick; I, furnace shell. (Online version in color.)

The ultrasonic treating system consists of ultrasonic generator, ultrasonic transducer, ultrasonic amplitude transformer and ultrasonic probe (only the ultrasonic probe is in Fig. 1). The ultrasonic generator can output power up to 2000-Watt with frequency of 20 kHz via controlling the input voltage and current. The ultrasonic transducer, amplitude transformer and probe were made of one kind of high-temperature alloy to maintain high temperature strength under high operating temperature. Especially, the ultrasonic probe was designed a cylindrical groove as container of molten slag, which facilitates the transmission of ultrasonic into the molten slag to minimize the power loss. In this study, considering the occurrence of cavitation effect in the melt, three ultrasonic intensities (ultrasonic power divided by area of groove bottom) of 0.303 W/mm² (600 Watts), 0.509 W/mm² (1000 Watts) and 0.708 W/mm² (1400 Watts) were investigated.

Maximum-pull method was employed for the surface tension measurement. The surface tension detection system consists of an electronic balance (precision 0.0001 g), a molybdenum ring linked wire, a thin-walled molybdenum cylindrical tube (wall thickness 1.0 mm, height 20 mm, external diameter 13 mm) and a computer for movement controlling of furnace and data recording. To meet the measurement of surface tension, the ultrasonic probe was fixed on the furnace to move up and down with the furnace simultaneously.

2.3. Surface Tension Measurement

Before experiment, the detection system was calibrated by using distilled water at room temperature. In the experiments, 50 g mixture of the raw materials of slag was filled into the groove of the probe then heated to temperature of 1293 K and hold on 30 minutes for composition homogenization. During surface tension detection (at least 5 times), the rim of the molybdenum cylinder was dipped just below the surface of the molten slag for 2 minutes to wet absolutely, after 10 minutes of persistent ultrasonic applying, the level of the molten slag slowly lowered with the furnace at a speed of 20 mm·min⁻¹ until the maximum excess weight was recorded by the computer automatically when the meniscus attached to the rim ultimately fails. To avoid the fluctuation induced by ultrasonic, the furnace lowered just at the same time of ultrasonic stopping.

2.4. Slag Sampling

In experiments, a stainless tube with an inner diameter of 5 mm was used as sampler, after 10 minutes of ultrasonic applying, the tube was withdrawn and quenched into ice water as quick as possible (less than 3 s) to maintain the melt structure at high temperature. Before withdrawing, the stainless tube was immerged into the slag for no fewer than 5 minutes to reach the temperature of the molten slag. To verify the chemical composition, slag sampling also conducted before and after ultrasonic applying and followed by composition analysis via Inductively Coupled Plasma method. The X-ray diffraction results showed that all the quenched samples were amorphous, which indicated that the operation was appropriate in sampling.

2.5. Structure Detection

The melt structure was detected via Raman spectrometer (HR800, Horiba Jobin Yvon, France) equipped with He–Ne exciting laser at room temperature, and the detection parameters are as follows: 488 nm in laser wavelength, 20 milliwatts in laser power, 300 μm in confocal hole diameter, 50 s in acquisition time. To quantify spectroscopic information of structures, Gaussian curve-fitting was conducted after data smoothing and baseline correction of the raw Raman data.

3. Results and Discussion

Figure 2 shows the variation of the surface tension of slag A and slag B under ultrasonic irradiation. They show an obviously different variation tendency that the surface tension of slag A decreases gradually with the increase of ultrasonic intensity, on the contrary, the surface tension of slag B increased gradually with the increase of ultrasonic intensity. Additionally, there is a difference in the variation degree of surface tension. For the fluorine free slag A, the surface tension is 299.1 mN/m (the average of experimental results) in its natural state, under the ultrasonic irradiation, the surface tension decreases to 281 (ultrasonic intensity 0.303 W/mm²), 274.3 (0.509 W/mm²), and 273.8 mN/m (0.708 W/mm²) respectively, and the corresponding reduction degree is 6.05%, 8.30% and 8.47%. For the fluorine bearing slag B, the surface tension increases from 296.2 to 300.5, 304.8 and 307.2 mN/m, and the corresponding increase degree is 1.46%, 2.91% and 3.71% respectively, which is significantly lower than that of slag A. The differences in variation tendency and degree of surface tension may closely relate to the ion behavior in the melt, which will be discussed combining with the ion behaviors under ultrasonic irradiation.

Fig. 2.

Variation of surface tension of fluorine free (A) and bearing (B) slag under ultrasonic irradiation. (Online version in color.)

In this study, the surface tension of molten slag was detected just after the cessation of ultrasonic irradiation, which may deviate from the real value under ultrasonic irradiation. To evaluate the deviation degree of surface tension, the results of continuous transformation of viscosity measured in the previous study was employed.²⁵⁾ As shown in Fig. 3, after the cessation of ultrasonic irradiation, the viscosity both slag A and slag B reverse to its nature values under no ultrasonic irradiation, for the slag A, the reversing time under ultrasonic intensity of 0.303 W/mm², 0.509 W/mm², and 0.708 W/mm² is 3.5, 4.5 and 3.0 minutes respectively, for the slag B, the reversing time is 8.0, 8.5 and 5.5 minutes respectively. The surface tension detecting time (from the same time of ultrasonic stopping to the recording time of surface tension value) is no more than 0.25 minutes, which is just in the initial stage of viscosity reversion. Naturally, viscosity is the reflection of the bulk structure of melt, after the cessation of ultrasonic irradiation, the bulk structure reverses to its natural state gradually, which will give rise to the reconstruction of the structure in the surface layer of the melt to build a new balance between them. Theoretically, a period of time is need to achieve a new balance, that is to say, the reversion of surface tension will be later than the reversion of viscosity, so the detecting result of surface tension can be look as the real value under the effect of ultrasonic irradiation.

Fig. 3.

Viscosity transformation after cessation of ultrasonic irradiation. The reversing time is the duration from the ultrasonic stopping time (25 minute) to the first appearance time of viscosity, after which there are at least two points with lower viscosity. (Online version in color.)

Temperature is an important factor affecting surface tension, under ultrasonic irradiation, some portion of the mechanical energy of ultrasonic would transfer to heat for the reason of cavitation effect, which would lead to an increase in slag temperature and the related variation of surface tension. In this study, to estimate the effect of temperature variation on the surface tension, the real-time temperature was measured via a K-type thermocouple during the ultrasonic applying. Figure 4 shows the furnace temperature and the slag temperature on the condition of ultrasonic intensity 0.509 W/mm², which is similar to that of ultrasonic intensity 0.303 W/mm² and 0.708 W/mm². Take slag B for example, before ultrasonic applying, the average temperature was 1295.9 K, after ultrasonic applying (60 s), the temperature increased gradually to a maximum of 1298.6 K, when the ultrasonic stopped (280 s), the temperature decreased and stabilized at the same temperature that of before ultrasonic applying. The temperature of furnace varied with the slag temperature as the similar tendency, while the fluctuation degree was only 1 K. The results show that the PID can satisfactorily control the temperature of furnace and slag within a narrow range. There are no obvious different in surface tension within the narrow temperature fluctuation range, so, the variation of surface tension can be attributed to the change of structure under ultrasonic irradiation.

Fig. 4.

Effect of ultrasonic irradiation on the temperatures of slag and furnace. (Online version in color.)

To verify the variation of melt structure under different intensity of ultrasonic irradiation, the bands denoting different melt structures were extracted from the Raman spectra via Gauss curve-fitting method. Figure 5 shows the Raman spectra and the curve-fitting results of slag A under different ultrasonic intensities between frequency of 800 and 1200 cm⁻¹. In silicate melt, Raman peaks reflect the Si–O stretching in melt structures of Qⁱ (i = 0, 1, 2, 3, the number of bridging oxygen per tetrahedron) with different polymerization.^{26,27,28,29,30,31,32)} Figure 5 shows that there are three obvious peaks in the Raman spectra, one locates between frequency of 846 and 857 cm⁻¹, the other locates between frequency of 969 and 977 cm⁻¹ and the third locates between frequency of 1054 and 1065 cm⁻¹, which are assigned to Q⁰, Q² and Q³ accordingly. Figure 6 show the Raman spectra and the curve-fitting results of slag B under different ultrasonic intensities between frequency of 800 and 1200 cm⁻¹. There are three peaks in the Raman spectra, but they show obvious different in patter from slag A, which in consistent with the fact that the higher polymerization of slag B (NBO/T=1). The most significant peak locates near frequency of 1110 cm⁻¹ is assigned to Q³ species, the other two peaks locate near 830 and 940 cm⁻¹ were assigned to Q⁰ and Q² respectively.

Fig. 5.

Gaussian curve-fitted Raman spectra of fluorine free slag A under ultrasound irradiation. In silicate melt, Raman peaks reflect the Si–O stretching in melt structures of Qⁱ (i = 0, 1, 2, 3, the number of bridging oxygen per tetrahedron) with different polymerization. Figure 4 shows that there are three obvious peaks in the Raman spectra, one locates between frequency of 846 and 857 cm⁻¹, the other locates between frequency of 969 and 977 cm⁻¹ and the third locates between frequency of 1054 and 1065 cm⁻¹, which are assigned to Q⁰, Q² and Q³ accordingly.^{25,26,27,28,29,30)} Should be noted that the Raman spectra both fluorine free and bearing slag show no obvious peak near frequency of 900 cm⁻¹, which indicates that the content of Q¹ species is too low to be detected. (Online version in color.)

Fig. 6.

Gaussian curve-fitted Raman spectra of fluorine bearing slag B under ultrasound irradiation. There are three peaks in the Raman spectra, but they show obvious different in patter from slag A, which in consistent with the fact that the higher polymerization of slag B (NBO/T=1). The most significant peak locates near frequency of 1110 cm⁻¹ is assigned to Q³ species, the other two peaks locate near 830 and 940 cm⁻¹, and were assigned to Q⁰ and Q² respectively. Different from slag A, there is a shoulder near frequency of 1030 cm⁻¹ in the Raman spectra of slag B, which is attributed to the symmetrical stretching vibration of Si–O bond associated with bridging oxygen in species of Q¹, Q² and Q³.³¹⁾ Similar with slag A, there is no signal of Q¹ species, so the band near frequency of 1030 cm⁻¹ was assigned to Q² and Q³ jointly. To avoid double counting, the band area fitted to the frequency near 1030 cm⁻¹ was ignored in the quantification of structure species. (Online version in color.)

Should be noted that, the Raman spectra both of fluorine free slag A and fluorine bearing slag B show no obvious peak or shoulder near frequency of 900 cm⁻¹, which indicates that the content of Q¹ species is too low to be detected. Repeat detections of Raman spectrum showed no obvious different, this may be considered as the nature characteristic of the experimental slag system of Na₂O–K₂O–SiO₂–CaF₂. Additionally, the Raman spectra under ultrasonic irradiation also show the same results, which can be explained that Q⁰ dissociates from Q³ with higher polymerization degree easier than Q² under ultrasonic irradiation to generate more Q² than Q¹. Figures 5 and 6 show that, the spectrum near 900 cm⁻¹ shows an increase with intensity of ultrasonic, this mostly may be the contribution of increase of Q¹ species, nevertheless, for its weak signal, the frequency near 900 cm⁻¹ was not assigned to Q¹ and was ignored in structure discussion.

In Fig. 6, it also shows a shoulder near frequency of 1030 cm⁻¹, which is attributed to the symmetrical stretching vibration of Si–O bond associated with bridging oxygen(BO) in species of Q¹, Q² and Q³, which is another signal of Q² and Q³ species.³²⁾ For the ignoring of Q¹ species, the band near frequency of 1030 cm⁻¹ can be only assigned to Q² and Q³ jointly. In this study, the stretching vibration of Si–O bond associated with no-bridging oxygen (NBO) is used to quantify the content of Qⁱ, to avoid double counting, the band area fitted to the frequency near 1030 cm⁻¹ was eliminated during quantification of structure species.

According to the above fitting rules, the abundances of structure species were quantified and the results of area percentage are listed in Table 2. Figure 7 shows clearly that the area percentage of the highest polymerized species of Q³ decreases with the increasing of ultrasonic intensity, correspondingly, the area percentage of the lower polymerized species of Q² and Q⁰ increase, which denotes that the silicate melts were degraded under ultrasonic irradiation. Figure 7 also shows that, with the increasing of intensity of ultrasonic, the reduction degree of Q³ and the increase degree of Q² and Q⁰ of slag A decreases obviously, especially, the percentages of structure species show less different at higher ultrasonic intensity. But in slag B, the variation degrees of structure species increase continuously with the increasing of ultrasonic intensity, this difference will be discussed near the end of this paper.

Table 2. Band area percentages of structure units Q⁰, Q², and Q³ in slag A and B.

Ultrasonic intensity W/mm2	Slag A			Slag B
	Q0	Q2	Q3	Q0	Q2	Q3
0	4.397	43.438	52.165	2.754	20.800	76.446
0.303	7.267	51.074	41.659	3.426	21.957	74.617
0.509	7.500	52.794	39.706	5.539	25.223	69.238
0.708	9.566	52.787	37.647	8.434	29.270	62.296

Fig. 7.

Variation of the contents of structure species under ultrasonic irradiation. (Online version in color.)

In silicate melt, Raman peaks commonly reflect the stretching of bond of Si–O in melt structures, if the stretching bond changes, the Raman peaks will shift correspondingly. Figure 5 shows that, the peak of Q⁰ locates near wavenumber of 850 cm⁻¹, and the peak of Q² locates near wavenumber of 970 cm⁻¹, they are all higher than that (in Fig. 6) in fluorine bearing slag B. The shift of Raman peaks to a lower frequency are in consistent with the behavior of fluorine ions in silicate melts that the joining of fluorine into the tetrahedron will lead to the decrease of frequency of relevant structure species.^33,34,35,36) As shown in Fig. 6, the peak of Q² under ultrasonic intensity of 0.303 W/mm² locates near wavenumber of 950 cm⁻¹, when the ultrasonic intensity increases to 0.509 W/mm², the peak of Q² shift to near wavenumber of 930 cm⁻¹, this indicates the more joining of fluorine anion into the tetrahedron under stronger ultrasonic intensity.

The degradation of silicate melts must relate to the broken of Si–O (BO) bond in the higher polymerized chain or ring type species. In this study, according to the Raman spectra (Figs. 5 and 6), the most likely way is that a portion of Q³ were degraded to Q⁰ and Q² under ultrasonic irradiation, the ions behaviors in the melts can be illustrated as Fig. 8. Near ultrasonic cavitation site, the broken of Si–O (BO) bonds, shown as step 1 in Fig. 8, generate the un-fully coordinated silica tetrahedron (${\mathrm{Q}}_{\mathrm{u}\mathrm{n}}^0$) radicals with one oxygen vacancy. In silicate melt, except coordination with Si⁴⁺ forming tetrahedron, some of oxygen and fluorine anions exist in free type of O²⁻ and F⁻ simultaneously. After dissociated from Q³, some portion of ${\mathrm{Q}}_{\mathrm{u}\mathrm{n}}^0$ species can reconnect reverting to the nature state at the site of no ultrasonic cavitation occurs, while the others can coordinate with the free O²⁻ or F⁻ which fill into the vacancies in ${\mathrm{Q}}_{\mathrm{u}\mathrm{n}}^0$, shown as step 2 in Fig. 8. In the fluorine free silicate melt of slag A, ${\mathrm{Q}}_{\mathrm{u}\mathrm{n}}^0$ can only connect with free O²⁻ generating fully coordinated ${\mathrm{Q}}_{\mathrm{O}}^0$, while in the fluorine bearing silicate melt of slag B, there are sufficient of free F⁻ substituting for O²⁻ to coordinate with ${\mathrm{Q}}_{\mathrm{u}\mathrm{n}}^0$ generating fully coordinated ${\mathrm{Q}}_{\mathrm{F}}^0$ via forming Si–F bond.

Fig. 8.

Behaviors of ions in the silicate melt under ultrasonic irradiation (K⁺ and Ca²⁺ were eliminated for clarity). (Online version in color.)

Surface tension is the measure of intermolecular forces acting at the surface,³⁷⁾ in ionic melt, surface tension is governed by the electrostatic force between cations and anions, which can be estimated via the equation shown as follow,   

$$I=\mathrm{k}{\mathrm{Z}}_{\mathrm{a}}{\mathrm{Z}}_{\mathrm{c}}{\mathrm{e}}^2/{\left(r_{\mathrm{a}}+r_{\mathrm{c}}\right)}^2$$

where k is Coulomb constant, 9.0×10⁹ Nm²/C², e is unit charge, 1.6×10⁻¹⁹ C, Z_a and Z_c, r_a and r_c is ionic valencies and radii for anion and cation, respectively. Figure 8 shows that, generating one fully coordinated ${\mathrm{Q}}_{\mathrm{O}}^0$ or ${\mathrm{Q}}_{\mathrm{F}}^0$ accompanies consumption of one free anion of O²⁻ or F⁻, which will lead to the variation of electrostatic force.

In the experimental slag system of this study, ionic valencies and radii of cations of Na⁺, K⁺, Ca²⁺ remain constant under ultrasonic irradiation, accordingly, the electrostatic force between ions varies only with the variation of anions of O²⁻, F⁻ and tetrahedral species of Q⁰ (assuming the radius of ${\mathrm{Q}}_{\mathrm{O}}^0$ equal to that of ${\mathrm{Q}}_{\mathrm{F}}^0$). Taking Na⁺ as an example, the electrostatic forces between Na⁺ and O²⁻, F⁻ and Q⁰ can be calculated according to the equation, and the results together with ionic radii are listed in Table 3.³⁸⁾ It shows that, the electrostatic force between Na⁺ and O²⁻ is the strongest, that between Na⁺ and F⁻ is the weakest, and that between Na⁺ and Q⁰ is in the middle.

Table 3. Ionic radii and electrostatic forces in the experimental slags.

Ionic radius, ×10−9 m				Electrostatic force, ×10−9 N
Na+	O2−	F−	SiO44−	Na+−F−	Na+−O2−	Na+−Q0 ($\mathrm{S}\mathrm{i}{\mathrm{O}}_4^{4-}$)
0.118	0.121	0.133	0.269	4.14	8.27	6.15

In the slag A, anions of O²⁻ is consumed generating Q⁰ species, that is to say, the electrostatic forces between some portion ions of Na⁺ and O²⁻ in the nature state are substituted to weaker electrostatic forces between Na⁺ and Q⁰, which lead to the decrease of surface tension under ultrasonic irradiation. In slag B, as same as slag A, anions of O²⁻ also are consumed leading to the decreasing of surface tension. Except consumption of O²⁻, abundant anions of F⁻ are consumed generating Q⁰ species, that is, the electrostatic forces between Na⁺ and F⁻ in the nature state are substituted to stronger electrostatic forces between Na⁺ and Q⁰, which lead to the increasing of surface tension. Additional, the degree of polymerization of slag B (NBO/T≈1) is higher than slag A (NBO/T≈2), that is, the content of free O²⁻ in slag B is lower than that in slag A, so the decreasing effect on surface tension should be weaker than that in slag A. Combining the fact of increasing of surface tension, it can be concluded that anions of F⁻ dominate the coordination process. The obvious lower variation degree of surface tension of slag B also can be attribute to the withdrawn of the consumption of O²⁻ on surface tension.

Figure 2 also shows that, with the increasing of intensity of ultrasonic, the reduction degree of surface tension of slag A decreases obviously. Especially, the reduction degree shows less different between ultrasonic intensity of 0.509 W/mm² (8.30%) and 0.708 W/mm² (8.47%). In slag B, the reduction degree of surface tension decreases indistinctively with the ultrasonic intensity. Comparing Figs. 2 with 7, the variation tendencies are all consistent with the degradation tendencies of melt structures with ultrasonic intensity, which also closely related to the reconfiguration behaviors of ions. In slag A, there are not sufficient free anions of O²⁻ to fully coordinate with abundant ${\mathrm{Q}}_{\mathrm{u}\mathrm{n}}^0$ species generated under high ultrasonic intensity, most portion of ${\mathrm{Q}}_{\mathrm{u}\mathrm{n}}^0$ reconnect with Q² reverting to the nature state. While in slag B, there are sufficient ions of free F⁻, which can meet the increasing demand of coordination of ${\mathrm{Q}}_{\mathrm{u}\mathrm{n}}^0$ with increasing of ultrasonic intensity.

If the ultrasonic can be transmitted into the gap zone between the solidified shell and the cooling wall, as one kind of fluorine bearing slag, the surface tension of mold flux will increase, which will lead to the increase of work adhesion between the two phase of mold flux and the solidified steel will increase, furtherly, will promote the flow of mold flux into the gap with the downward movement of the slag/billet.

4. Conclusions

In molten slags of Na₂O–K₂O–SiO₂–CaF₂ system, radicals of silica tetrahedron with one oxygen vacancy dissociate from the higher polymerized structures under ultrasonic irradiation, which can coordinate with free anions of oxygen and fluorine, this lead to the changes of electrostatic force between ions. In fluorine free slag, free oxygen dominates the reconfiguration process, which lead to the decrease of electrostatic force and surface tension, while in fluorine bearing slag, free fluorine dominates the reconfiguration process, which lead to the increase of electrostatic force and surface tension.

Acknowledgment

We acknowledge the supports of the National Natural Science Foundation of China (grant number 51974075, 51674069), the National key R & D Program of China (grant number 2017YFC0805100), the Open Funds of State Key Laboratory of Metal Material for Marine Equipment and Application (SKLMEA-K201911), and the Fundamental Research Funds for the Central Universities of China (grant number N182506001, N180725008, N182503034).

References

• 1)   M.  Iguchi,  J.  Yoshida,  T.  Shimizu and  Y.  Mizuno: ISIJ Int., 40 (2000), 685.
• 2)   K.  Watanabe,  K.  Tsutsumi,  M.  Suzuki,  M.  Nakada and  T.  Shiomi: ISIJ Int., 49 (2009), 1161.
• 3)   K. S.  Suslick: Sci. Am., 260 (1989), 80.
• 4)   K. S.  Suslick: Science, 247 (1990), 1439.
• 5)   S. J.  Doktycz and  K. S.  Suslick: Science, 247 (1990), 1067.
• 6)   S.  Mitragotri,  D.  Blankschtein and  R.  Langer: Science, 269 (1995), 850.
• 7)   Y. T.  Didenko and  K. S.  Suslick: Nature, 418 (2002), 394.
• 8)   M. W. A.  Kuijpers,  D.  van Eck,  M. F.  Kemmere and  J. T. F.  Keurentjes: Science, 298 (2002), 1969.
• 9)   W. H.  Suh and  K. S.  Suslick: J. Am. Chem. Soc., 127 (2005), 12007.
• 10)   J. E.  Park,  M.  Atobe and  T.  Fuchigami: Ultrason. Sonochem., 13 (2006), 237.
• 11)   L.  Gaete-Garretón,  Y.  Vargas-Hernández,  M. G.  Cares-Pacheco,  J.  Sainz and  J.  Alarcón: Ultrasonics, 51 (2011), 581.
• 12)   F.  Chemat,  N.  Rombaut,  A. G.  Sicaire,  A.  Meullemiestre,  A. S.  Fabiano-Tixier and  M.  Abert-Vian: Ultrason. Sonochem., 34 (2017), 540.
• 13)   A. M.  Basedow and  K. H.  Ebert: Adv. Polym. Sci., 22 (1977), 83.
• 14)   G. J.  Price and  P. F.  Smith: Polym. Int., 24 (1991), 159.
• 15)   G. J.  Price and  P. F.  Smith: Eur. Polym. J., 29 (1993), 419.
• 16)   G. J.  Price and  P. F.  Smith: Polymer, 34 (1993), 4111.
• 17)   G. J.  Price,  A. J.  White and  A. A.  Clifton: Polymer, 36 (1995), 4919.
• 18)   G.  Madras and  S.  Chattopadhyay: Polym. Degrad. Stab., 71 (2001), 273.
• 19)   M. T.  Taghizadeh and  A.  Mehrdad: J. Polym. Sci., Part B: Polym. Phys., 42 (2004), 445.
• 20)   S. P.  Vijayalakshmi and  G.  Madras: Polym. Degrad. Stab., 90 (2005), 116.
• 21)   A. V.  Mohod and  P. R.  Gogate: Ultrason. Sonochem., 18 (2011), 727.
• 22)   D. R.  Lide: Handbook of Chemistry and Physics, 90th Ed., CRC Press, Boca Raton, (2009), 9-64.
• 23)   Y.  Min,  Z. X.  Shi and  C. J.  Liu: Ultrason. Sonochem., 39 (2017), 727.
• 24)   M.  Allibert,  H.  Gaye,  J.  Geiseler,  D.  Janke,  B. J.  Keene,  D.  Kirner,  M.  Kowalski,  J.  Lehmann,  K. C.  Mills,  D.  Neuschutz,  R.  Parra,  C.  Saint-Jours,  P. J.  Spencer,  M.  Susa,  M.  Tmar and  E.  Woermann: Slag Atlas, Verlag Stahleisen GmbH, Dusseldorf, (1995), 403.
• 25)   Y.  Min,  J.  Luo and  C. J.  Liu: Ultrason. Sonochem., 55 (2019), 289.
• 26)   S. A.  Brawer and  W. B.  White: J. Chem. Phys., 63 (1975), 2421.
• 27)   B. O.  Mysen,  D.  Virgo and  C. M.  Scarfe: Am. Mineral., 65 (1980), 690.
• 28)   T.  Furukawa,  K. E.  Fox and  W. B.  White: J. Chem. Phys., 75 (1981), 3226.
• 29)   P.  McMillan and  B.  Piriou: Bull. Mineral., 106 (1983), 57.
• 30)   P.  McMillan: Am. Mineral., 69 (1984), 645.
• 31)   M.  Tatsumisago,  T.  Minami,  N.  Umesaki and  N.  Iwamoto: Chem. Lett., 15 (1986), 1371.
• 32)   B. O.  Mysen and  J. D.  Frantz: Chem. Geol., 96 (1992), 321.
• 33)   P.  Dumas,  J.  Corset,  W.  Carvalho,  Y.  Levy and  Y.  Neuman: J. Non-Cryst. Solids, 47 (1982), 239.
• 34)   K.  Yamamoto,  T.  Nakanishi,  H.  Kasahara and  K.  Abe: J. Non-Cryst. Solids, 59–60 (1983), 213.
• 35)   B. O.  Mysen and  D.  Virgo: Phys. Chem. Miner., 12 (1985), 77.
• 36)   B. O.  Mysen and  D.  Virgo: Contrib. Mineral. Petrol., 91 (1985), 205.
• 37)   S.  Hara and  K.  Ogino: ISIJ Int., 29 (1989), 477.
• 38)   Q.  Gao,  Y.  Min and  M. F.  Jiang: Metall. Mater. Trans. B, 49 (2018), 1302.