Optimization of the Interfacial Properties between Mold Flux and TiN Substrate Through the Regulation of B₂O₃

Abstract

Titanium nitride (TiN) inclusions are easy to precipitate in the high temperature processing of titanium alloying steels, which tends to introduce numerous surface defects on the final continuous casting slabs. This study utilizes B₂O₃ to regulate the interfacial properties between the designed mold fluxes and TiN, with the aim to resolve above problems. The results show that the spreading behavior of the mold flux on the TiN substrate is enhanced, and the interfacial contact angle starts to drop at a lower temperature (from 1473 K for Sample 1 to 1343 K for Sample 4) with the addition of 0–9 wt.% B₂O₃, as the melting behavior of the designed mold fluxes has been improved. The interfacial reactions between the TiN substrate and molten fluxes are also promoted with the addition of B₂O₃, where more bubbles are observed in the tested mold fluxes samples. For Sample 1 without B₂O₃, quite a few TiN particles couldn’t be dissolved and remains in the matrix phase, where the major formed phase is perovskite (CaTiO₃) that would deteriorate the high temperature properties of mold flux severely. However, most TiN particles have been dissolved in the optimized mold fluxes, as major of them have reacted with mold fluxes, resulted in the more generation of titanium oxides phase in the samples. In addition, the calculated phase diagram of CaO–SiO₂–TiO₂ slag system under different B₂O₃ contents indicates that the formation and precipitation of CaTiO₃ can be effectively inhibited by the addition of B₂O₃.

1. Introduction

Titanium as an effective alloying element has been well used in the design of advanced steels, as it could easily react with carbon to form titanium carbides, resulted in the improvement of the corrosion resistance and high temperature mechanical properties for steels.^1,2,3,4) However, titanium nitride (TiN) inclusions would be formed at the same time during the process of steelmaking and continuous casting, since the added titanium tends to react with the residual nitrogen in the molten steel.^5,6,7) Although most TiN inclusions can be removed in the refining furnace and tundish, some of them may still enter into the casting mold through the sub-merged nozzle. Therefore, if those in-mold inclusions couldn’t be eliminated thoroughly, they would be captured and entrapped in the front of initial solidified dendrites during the continuous casting process, which in turn to introduces lots of TiN inclusions related surface defects, such as the so-called “black slivers”, micro-cracks and pin holes, which are usually appearing on the surface of the cold-rolled strips,^8,9,10) and degrade the general quality of product greatly.

Mold powders as an effective functional material have been widely used in the process of continuous casting, and play essential roles in the process of continuous casting, such as preventing the molten steel from oxidation,¹¹⁾ modifying the horizontal heat transfer in the mold,^12,13) lubricating the solidified shell,¹⁴⁾ and absorbing the inclusions from the molten bath,^15,16) etc. Therefore, it is of great importance if TiN inclusions could be effective absorbed by the top liquid mold flux and get dissolved, when they float up from the molten steel bath. Thus, these inclusions inside the mold could be clearly eliminated. The dissolution of TiN inclusions inside the mold flux is very crucial, as the precipitation of TiN high temperature inclusions inside the liquid mold flux would introduce the operation problems. The research from Mukongo et al.¹⁷⁾ suggested that the viscosity of mold flux increases significantly with the pickup of TiN, and they declared that the precipitation of perovskite (CaTiO₃) should be responsible for the fast increase of viscosity. Tokovol et al.¹⁸⁾ also pointed out that the contact between the TiN inclusions and the mold flux could lead to a higher melting temperature of the mold flux and form the metal-slag conglomerates (crust). Consequently, the deterioration of mold flux properties would further affect the shell lubrication and in-mold heat transfer control, sequentially causes resulted in severe surface defects of the solidified slab.

For the TiN dissolution study, Ozturk¹⁹⁾ measured the solubility of TiN in the liquid mold flux at 1873 K (1600°C), and the results showed that its solubility is promoted with the addition of silica content in the slag. Michelic et al.¹⁶⁾ investigated the dissolution of TiN in both SiO₂-rich slags and Al₂O₃-rich slags through the thermodynamic calculation and experimental tests, and implied that the SiO₂-based slag shows a higher solubility than the latter one. So, the composition of mold flux has a big impact on the TiN dissolution behavior.

Therefore, the dissolution of TiN inclusions and the inhibition of perovskite (CaTiO₃) precipitation inside the mold flux, when they were absorbed during the continuous casting process, are they two key issues to be tackled down for the improvement of continuous casting of Ti-alloys steels. Consequently, this paper proposed a valued study to optimize the interfacial behavior between the designed mold fluxes and TiN through B₂O₃ regulation, such that the TiN particles could be effectively dissolved and the original formed CaTiO₃ would be changed to other phase. In addition, the reason to chose B₂O₃ is because it is a high-efficiency flux agent, which can decrease the viscosity and lower the melting temperature of mold flux and other slags.^{13,20,21,22,23,24,25)} The interfacial studies in this paper focused on the spreading and wetting behaviors of the molten mold fluxes on the TiN substrates at the temperature of 1723 K (1500°C). Then the detailed interfacial reactions and phase transformations during the tests have also been investigated.

2. Experimental Methods

2.1. Sample Preparation

The chemical compositions of the mold fluxes used in this study are listed in Table 1. Among them, Sample 1 is an existing commercial mold flux for the casting of a typical titanium-stabilized stainless steel, which is considered as the benchmark. Samples 2, 3 and 4 are newly designed mold fluxes with the variation of B₂O₃ content from 3 wt.% to 9 wt.%. All the flux samples were prepared by mixing reagent grade chemicals, such as CaCO₃, SiO₂, Al₂O₃, MgO, Na₂CO₃, CaF₂ and B₂O₃ in a blender first, and then melted in an induction furnace. The melting temperature and holding time were set at 1773 K (1500°C) and 10 mins., respectively, to ensure the elimination of bubbles and the homogenization of chemical composition. After that, the molten slags were quenched onto a water-cooled copper plate to obtain a fully glassy disk. The final sample for the sessile drop test was obtained via grinding the glassy slag disk into a cube with an edge of 5 mm.

Table 1. Chemical compositions of the mold fluxes used in this study (wt.%).

	Sam.	CaO	SiO2	Al2O3	Na2O	F	MgO	B2O3
Design	1	39.81	36.19	6.00	9.00	7.00	2.00	0
	2	38.24	34.76	6.00	9.00	7.00	2.00	3.00
	3	36.67	33.33	6.00	9.00	7.00	2.00	6.00
	4	35.10	31.90	6.00	9.00	7.00	2.00	9.00
Pre-melted	1	39.85 ± 0.01	36.23 ± 0.01	6.03 ± 0.02	8.96 ± 0.02	6.92 ± 0.001	2.01 ± 0.01	0.00 ± 0.00
	2	38.36 ± 0.01	34.87 ± 0.01	5.95 ± 0.02	8.91 ± 0.02	6.96 ± 0.001	1.97 ± 0.01	2.98 ± 0.01
	3	36.81 ± 0.01	33.46 ± 0.01	5.94 ± 0.02	8.88 ± 0.02	6.95 ± 0.001	2.03 ± 0.01	5.93 ± 0.01
	4	35.17 ± 0.01	31.98 ± 0.01	6.01 ± 0.02	8.93 ± 0.02	6.96 ± 0.001	2.00 ± 0.01	8.95 ± 0.01

The chemical compositions of the pre-melted fluxes were also analyzed by ion-selective electrode method (ISE; PF-2-01, Shanghai LEICI Corporation), inductively coupled plasma optical emission spectroscopy (ICP-OES; SPECTROBLUE, SPECTRO Corporation) and X-ray fluorescence (XRF; S4 Pioneer, Bruker AXS, Karlsruhe, Germany). The analysis results were also shown in Table 1, and the results suggested the volatilization losses of components in these mold fluxes could be ignored.

The original TiN substrates (99.99 wt.%) were purchased from HeFei Crystal Technical Material Corporation. All TiN substrates were polished with SiC sandpapers with the grit size down to 2000 mesh to eliminate the influence of surface roughness on the interfacial properties.

2.2. Sessile Drop Test

The investigations on the interfacial spreading, wettability and interactions between the mold fluxes and TiN substrate were conducted by using sessile drop method. The apparatus was mainly composed of four parts, including a horizontal heating furnace with a high-purity alumina tube, atmosphere controlling system, cooling water circulating system and image acquisition system, as shown in Fig. 1. During the test process, the cube-shaped mold flux sample was placed on the top of the TiN substrate and pushed into the center of the alumina tube. Then, the high-purity (99.999 vol.%) argon gas was introduced into the furnace at a flow rate of 300 ml/min after deoxidized and denitrided by slowly passing through a gas purifier contained Cu, Al, and Mg getters. The partial pressures of oxygen (PO₂) and nitrogen (PN₂) at the exit of the furnace were detected to be about 1 × 10⁻¹⁸ and 1 × 10⁻⁶ atm, respectively, by the oxygen and nitrogen monitor systems (MOT500-M, Korno Company). The flux sample and substrate were heated up at a rate of 5 K/min and held at 1773 K (1500°C) for 30 mins. The preset holding temperature in present study was 1773 K (1500°C) since it is close to the temperature at the interface of molten steel/liquid mold flux in the continuous casting mold.^5,17) The spreading behavior of mold flux on the TiN substrate during the whole test was observed and recorded by the image acquisition system. The contact angle was obtained through the image processing by using Image J software package (Version 1.48, National Institutes of Health).

Fig. 1.

Schematic figure of the sessile drop apparatus. (Online version in color.)

The mold flux and TiN substrate after the sessile drop test were air quenched and sampled for further analyzing. The energy dispersive spectroscopy (EDS; EDX-GENESIS 60S, America EDAX Corporation) was used to identify the elements distribution, and the scanning electron microscopy (SEM; JSM-6360LV, Japanese Electronics Company) was utilized to analyze the interface morphology at the interface between mold flux and TiN substrate. In addition, the melting temperature regions of the mold fluxes were tested by single hot thermocouple technique (SHTT) with a heating rate of 15 K (°C)/s. The details of the SHTT apparatus and the melting test procedure have been described in previous articles.^{26,27,28,29,30)}

3. Results and Discussion

3.1. Effect of B₂O₃ on the Interfacial Wettability

Figure 2 shows the preset heating curve and snapshots of the spreading behavior of Sample 1 on the TiN substrate during the sessile drop test. The shape of the flux sample is observed to change with the temperature rising. The whole test process could be roughly divided into three stages based on the variation of the sample profile. The mold flux first keeps its cubic shape on the substrate when the temperature is relatively low in Stage I. However, the image field becomes brighter gradually with the increase of heating temperature from 293 K (20°C) to 1473 K (1200°C). Then, the shape of flux sample starts to deform at 1473 K (1200°C), and forms a spherical cap during the melting region till 1503 K (1230°C) in Stage II. Finally, the flux droplet gets more fluidity when it is fully melted and spread continuously on the substrate surface, as the heating temperature further rises in Stage III, where bubbling can be observed on the top of flux sample. Similar bubbling phenomenon was also observed by Misra et al.⁵⁾ via using high temperature confocal scanning laser microscopy (CSLM).

Fig. 2.

The preset heating curve and snapshots of the spreading behavior of Sample 1 on the TiN substrate. (Online version in color.)

In order to investigate the effect of B₂O₃ on the wettability of mold flux on the TiN substrate, the snapshots of the spreading behavior of different mold fluxes with varying B₂O₃ contents on the TiN substrates are given in Fig. 3(a), and their contact angles are also shown in Fig. 3(b). The general spreading pattern of Samples 2, 3 and 4 on TiN substrate is similar to that of Sample 1, as shown in Fig. 3(a). However, the temperatures, at which the shape of the flux sample starting to deform and forming a spherical cap decreases. The bubbling temperature of flux droplet also decreases sharply with the addition of B₂O₃ content. These phenomena suggest that B₂O₃ has a great impact on the spreading behavior of the mold flux on the TiN substrate.

Fig. 3.

Wettability of mold fluxes on the TiN substrate at different temperatures. (a) the spreading process, (b) the variation of contact angle. (Online version in color.)

The contact angles between the mold fluxes and TiN substrates are plotted versus heating temperatures and shown in Fig. 3(b). The contact angles for all fluxes samples remain at 90 deg. in the low temperature region corresponding to Stage I in Fig. 2. However, during Stage II when the samples start to melt, the contact angle drops sharply in a very different temperature region corresponding to Stage II, where Sample 1 starts to drop from1473 K (1200°C) up to 1503 K (1230°C), and its contact angle decreases from 90 deg. to 26 deg. However, the Stage II ranges from 1403 K (1130°C) to 1433 K (1160°C) for Sample 2, from 1373 K (1100°C) to 1403 K (1130°C) for Sample 3, and from 1343 K (1070°C) to 1373 K (1100°C) for Sample 4, respectively. Combing the observations in Figs. 3(a) and 3(b), it can be inferred that the temperature at which the contact angle starts to drop decreases with the addition of B₂O₃. In Stage III, the contact angle is observed to reduce slightly with the continuous heating of the sample.

In order to further confirm above results, the melting behavior of above mold fluxes was investigated by SHTT, and their results are shown in Fig. 4. It can be seen that Sample 1 starts to melt at 1436 K (1163°C) as suggested by the deviation of the responding temperature away from the preset line. Then the Sample 1 is fully melted at 1542 K (1269°C), at which the responding temperature curve gets back to the preset line. The melting range is 1326 K (1053°C) to 1452 K (1179°C) for Sample 2, and it reduces to 1306 K (1033°C)–1435 K (1162°C) for Sample 3, and further reduces to 1339 K (1066°C)–1418 K (1145°C) for Sample 4. It is obvious that the melting temperature region is getting lowered with the addition of B₂O₃, which is consistent with the tendency of Stage II corresponding to the sharp dropping of contact angles. Therefore, it can be concluded that the addition of B₂O₃ would enhance the melting capability of mold flux system, such that their melting temperature region could be lowered as suggested in Fig. 4, and the sample would deform earlier resulted in a significant contact angle drop in Stage II. Thus, the addition of B₂O₃ promotes the spreading and wettability of mold flux on TiN substrate; sequentially enhances the assimilation of the TiN inclusions in the mold flux. Besides, it should be mentioned that the melting temperature range of the mold fluxes obtained by SHTT is higher than the melting region of Stage II in the sessile drop tests, due to the higher heating rate in SHTT tests.

Fig. 4.

Melting processes of the mold fluxes. (a) Sample 1, (b) Sample 2, (c) Sample 3, (4) Sample 4. (Online version in color.)

After the complete melting of the flux samples, which corresponds to Stage III in Fig. 2, the contact angle still decreases slightly with the increase of heating temperature. The continuous attenuation of contact angle at the high temperatures is mainly caused by the interfacial reaction occurs at the flux droplet/substrate interface, besides the influence of temperature. The thermodynamic calculations of the possible interfacial reactions have been conducted in previous paper,³⁰⁾ in which TiN can react with SiO₂ and oxygen inside the mold flux through Reactions (1) and (2). The migrations of elements and charges during the chemical reactions may lower the interfacial tension and enhance the wettability of molten flux versus TiN substrate. In fact, the microstructure analyses in the section 3.2 also reveal the existing of interfacial chemical reactions. Therefore, the interfacial reaction is another important factor impacting the interfacial wettability and spreading behavior.   

$$\begin{array}{l}2\mathrm{T}\mathrm{i}\mathrm{N}(\mathrm{s})+2(\mathrm{S}\mathrm{i}{\mathrm{O}}_2)=2\mathrm{S}\mathrm{i}\mathrm{T}\mathrm{i}(\mathrm{s})+2{\mathrm{O}}_2(\mathrm{g})+{\mathrm{N}}_2(\mathrm{g})\\ \mathrm{\Delta }\mathrm{G}{}_{(1)}{}^0=2\mathrm{\hspace{0.17em} }153\mathrm{\hspace{0.17em} }470-1\mathrm{\hspace{0.17em} }309.74\mathrm{T}\mathrm{\hspace{0.17em} }(\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l})\end{array}$$(1)

  

$$\begin{array}{l}2\mathrm{T}\mathrm{i}\mathrm{N}(\mathrm{s})+2{\mathrm{O}}_2(\mathrm{g})=2\mathrm{T}\mathrm{i}{\mathrm{O}}_2(\mathrm{s})+{\mathrm{N}}_2(\mathrm{g})\\ \mathrm{\Delta }\mathrm{G}{}_{(2)}{}^0=-1\mathrm{\hspace{0.17em} }205\mathrm{\hspace{0.17em} }050+739.85\mathrm{T}\mathrm{\hspace{0.17em} }(\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l})\end{array}$$(2)

Furthermore, the contact angles curves are observed to fluctuate and overlap in Stage III shown in Fig. 3(b), especially when the temperature is above 1523 K (1250°C). So, it is very difficult to evaluate the influence of B₂O₃ content on the interfacial contact angle between mold flux and TiN substrate in stage III. The main factor to introduce this fluctuation is the bubbling, where N₂ is formed and released due to the substrate (inclusion) and flux reactions. The bubbling would disturb the surface of the flux droplet and lead to obvious deviation for the measurements of contact angle.

3.2. Effect of B₂O₃ on the Interfacial Microstructure

Figure 5 shows the top view of above samples after sessile drop tests. The mold fluxes completely wet the TiN substrates since the surface of the substrates is fully covered by the mold fluxes. Meanwhile, the surfaces are not smooth, where concave, convex and bubbles are occurring. These bubbles are formed due the interfacial reactions as described in above section. The broken and non-broken nitrogen bubbles on the top of the samples lead to these concave or convex circles after the molten flux being cooled down.

Fig. 5.

Top views of the samples after the sessile drop tests. (a) Sample 1, (b) Sample 2, (c) Sample 3, (d) Sample 4. (Online version in color.)

In order to study the microstructure in the vicinity of the interface, the cross-sectional profiles of above 4 samples shown in Fig. 5 were subjected to SEM analysis and theirs photos are shown in Fig. 6. The bottom part of these figures is TiN substrate, whereas the upper part is mold flux. Although it is easy to distinguish the substrate from flux, there is no clear line at the interface of mold flux/steel.^31,32) It is mainly due to the intensive interfacial reactions occurred between mold flux and TiN substrate as suggested by Reactions (1) and (2).

Fig. 6.

Cross-sectional profile of the samples after the sessile drop tests. (a) Sample 1, (b) Sample 2, (c) Sample 3, (d) Sample 4.

A lot of holes can be also found in the mold fluxes besides the appearance of the circles on the top surface of the samples as shown in Fig. 5. These phenomena correlate well with the bubbling images in Figs. 2 and 3(a). In addition, by comparing the four cross-sectional profiles of the samples, the porosity inside the mold fluxes samples is observed to increase from Sample 1 to Sample 4, which means the addition of B₂O₃ tends to increase the interfacial reactions. The reason for that is because both the melting temperature and viscosity of the mold flux are reduced by the additional B₂O₃, which provides better kinetic conditions for the interfacial chemical reactions; thereby, more N₂ gas and the resulted holes are formed.

For better observation of the elements distribution and morphologies of the precipitated phases, the box zones at positions I, II, III and IV in Figs. 6(a) through 6(d) inside mold fluxes bodies are further magnified and shown in Fig. 7, where Figs. 7(a1), 7(b1), 7(c1) and 7(d1) show the SEM images of these magnified zones, and the other figures show the EDS results of typical points A, B and C in Figs. 7(a1) through 7(d1).

Fig. 7.

SEM images and EDS spectra results of the typical positions at the cross-sectional profile of the samples. (a1)–(d1) the SEM images, (a2)–(d2) EDS of points A, (a3)–(d3) EDS of points B, (a4)–(d4) EDS of points C. (Online version in color.)

The morphology of the phases in Fig. 7(a1) is quite different from other three SEM images, in which lots of the white tiny particles are gathering around Point A. The EDS results of Points A are shown in Figs. 7(a2) through 7(d2), and suggest that these particles are composed of elements Ti and N, which confirms that they are from the top surface of substrate. Therefore, it indicates that top part of the substrate is melted and those dissolved particles penetrate into the mold flux body. As the interracial reactions are not significant, they would remain in the matrix mold flux body. However, only a small amount of the particles are found in Figs. 7(b1) through 7(d1). Above phenomena could be explained as lot of bubbles produced at the interface due to the chemical reactions; it could break down the top part of the substrate and bring those particles into the molten flux with the help of buoyancy force. The reason why there are much more TiN residues in Sample 1, it may be due to the fact that the melting temperature and viscosity of Sample 1 are higher than other samples. The entered TiN particles cannot be dissolved and reacted completely in the molten flux due to the poor kinetic conditions.

Once TiN particles enter molten mold flux, they could react with (SiO₂) and (O) and form TiO₂, SiTi and N₂ as suggested by Reactions (1) and (2). Therefore, SiTi is found at points B nearby TiN particles in Figs. 7(a1) through 7(d1), even though its amount is not significant. The EDS spectra for SiTi are shown in Figs. 7(a3) through 7(d3). Points C in Figs. 7(a1) through 7(d1) represent the major phase of mold flux samples, and their EDS results suggest that it is composed of O, Ca and Ti (perovskite) for Sample 1, whereas it changes to O and Ti (Titanium oxides) for Samples 2, 3 and 4, as shown in Figs. 7(a4) to 7(d4). The major components in Point C Sample A (Fig. 7(d1)) suggest that the phase might be CaTiO₃ (perovskite), as the produced titanium oxides, such as TiO₂ from Reaction (2) would easily combine with CaO to form this high melting point phase according to Reaction (3).   

$$\begin{array}{l}\mathrm{T}\mathrm{i}{\mathrm{O}}_2(\mathrm{s})+(\mathrm{C}\mathrm{a}\mathrm{O})=\mathrm{C}\mathrm{a}\mathrm{T}\mathrm{i}{\mathrm{O}}_3(\mathrm{s})\\ \mathrm{\Delta }\mathrm{G}{}_{(3)}{}^0=-75\mathrm{\hspace{0.17em} }636.33-10.80\mathrm{T}\mathrm{\hspace{0.17em} }(\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l})\end{array}$$(3)

In order to figure out why there is no CaTiO₃ phase in Samples 2, 3 and 4, the phase diagram of CaO–SiO₂–TiO₂ slag system under different B₂O₃ contents at 1773 K (1500°C) has been calculated by Factsage (Version 7.2, Thermfact/CRCT and GTT-Technologies), as shown in Fig. 8. It can be seen that the liquid region expands and the precipitation region of CaTiO₃ shrinks greatly in the phase diagram with the addition of B₂O₃ from 0 wt.% to 9 wt.%. The variation tendency indicates that B₂O₃ can inhibit the formation and precipitation of CaTiO₃. This may be the main reason why the major phase in Sample 1 is CaTiO₃, whereas it is titanium oxides in the other three samples. The inhibition of the CaTiO₃ precipitation is beneficial for the performances of the mold flux, since it precipitates at high temperature, and increases the mold flux viscosity , resulted in the degradation of mold flux lubrication performance.

Fig. 8.

Phase diagram of CaO–SiO₂–TiO₂ system slag with different B₂O₃ content at 1773 K (1500°C). (Online version in color.)

4. Conclusions

This study focuses on the effect of B₂O₃ on the interfacial behaviors between the mold flux droplets and the TiN substrate, where the spreading behavior, contact angle variation, microstructure and element distribution in the vicinity of the interface were investigated. Some important conclusions are summarized as follows:

(1) The spherical cap temperature and bubbling temperature of the mold fluxes droplets decrease with the addition of B₂O₃ content, which suggests that B₂O₃ can enhance the spreading of the mold flux on the TiN substrate.

(2) The temperature region, in which the contact angle drops sharply, are 1473 K (1200°C)–1503 K (1230°C) for Sample 1, 1403 K (1130°C)–1433 K (1160°C) for Sample 2, 1373 K (1100°C)–1403 K (1130°C) for Sample 3, and 1323 K (1050°C)–1373 K (1100°C) for Sample 4. In addition, the variation tendency of this temperature region is consistent with the reduction of melting temperature range of the mold fluxes observed by SHTT.

(3) The interfacial reactions between TiN substrate and molten fluxes are enhanced by the addition of B₂O₃, and the interfacial reactions can improve the wettability, and lead to the bubbling in the molten flux samples, where concave or convex circles on the surface of the samples, and holes inside of mold flux are observed.

(4) The main phase inside Sample 1 is CaTiO₃, and quite a few of TiN particles are observed to precipitated inside the sample with some amount of SiTi crystals nearby. However, very few TiN particles are found in the B₂O₃ containing samples, as most of them are reacted with mold flux to form titanium oxides and tiny TiSi. The calculated phase diagram indicates that B₂O₃ can inhibit the formation and precipitation of CaTiO₃. Therefore, the regulation of B₂O₃ could effectively dissolve TiN inclusions by the enhancement of the spreading and interfacial reactions, resulted in major formation of titanium of oxides and the inhibition of CaTiO₃.

Acknowledgment

This work was supported by the national natural science foundation of China (51874363, U1760202), the Natural Science Foundation of Hunan Province (2019JJ40345), and Hunan Scientific Technology projects (2018RS3022, 2018WK2051).

References

• 1)   L. H.  Boulton and  A. J.  Betts: Br. Corros. J., 26 (1991), 287.
• 2)   S. M.  Gateman,  L. I.  Stephens,  S. C.  Perry,  R.  Lacasse,  R.  Schulz and  J.  Mauzeroll: NPJ Mater. Degrad., 2 (2018), Article No. 5.
• 3)   Z.  Chen,  Y.  Zhang,  S.  He,  Z.  Li and  Q.  Wang: Can. Metall. Q., 58 (2019), 464.
• 4)   E.  El-Kashif,  K.  Asakura,  T.  Koseki and  K.  Shibata: ISIJ Int., 44 (2004), 1568.
• 5)   P.  Misra,  S.  Sridhar and  A. W.  Cramb: Metall. Mater. Trans. B, 32 (2001), 963.
• 6)   R. B.  Tuttle and  S. A.  Kottala: Arch. Foundry Eng., 18 (2018), 129.
• 7)   A.  Ito,  H.  Suito and  R.  Inoue: ISIJ Int., 52 (2012), 1196.
• 8)   Q.  Ren,  J.  Li,  B.  Xu,  L.  Jiang and  H.  Li: Mater. Res. Innov., 19 (2015), S9-377.
• 9)   W.  Gao: Adv. Mater. Res., 402 (2012), 221.
• 10)   K.  Mills and  A.  Fox: ISIJ Int., 43 (2003), 1479.
• 11)   L.  Zhou,  W.  Wang,  J.  Wei and  B.  Lu: ISIJ Int., 53 (2013), 665.
• 12)   J.  Cho,  H.  Shibata,  T.  Emi and  M.  Suzuki: ISIJ Int., 38 (1998), 440.
• 13)   T.  Yeo,  J.  Cho,  M.  Alloni,  S.  Casagrande and  R.  Carli: J. Non-Cryst. Solids, 529 (2020), 119756.
• 14)   L.  Zhou,  Z.  Pan,  W.  Wang and  J.  Chen: Steel Res. Int., 91 (2020), No. 3, 1900503.
• 15)   W.  Wang,  E.  Gao,  L.  Zhou,  L.  Zhang and  H.  Li: J. Iron Steel Res. Int., 26 (2019), No. 4, 355.
• 16)   S. K.  Michelic and  C.  Bernhard: Scanning, 2017 (2017), 2326750.
• 17)   T.  Mukongo,  P. C.  Pistorius and  A. M.  Garberscraig: Ironmaking Steelmaking, 31 (2004), 135.
• 18)   R.  Ozturk: Metall. Trans. B, 23 (1992), 523.
• 19)   O. K.  Tokovoi and  D. V.  Shaburov: Steel Transl., 43 (2013), 678.
• 20)   L.  Zhou,  W.  Wang,  B.  Lu,  G.  Wen and  J.  Yang: Met. Mater. Int., 21 (2015), 126.
• 21)   H.  Wang,  G.  Li,  B.  Li,  X.  Zhang and  Y.  Yan: J. Iron Steel Res. Int., 17 (2010), 18.
• 22)   W.  Wang,  S.  Dai,  L.  Zhou,  J.  Zhang,  W.  Tian and  J.  Xu: Ceram. Int., 46 (2020), 3631.
• 23)   S.  Ren,  J.  Zhang,  L.  Wu,  W.  Liu,  Y.  Bai,  X.  Xing,  B.  Su and  D.  Kong: ISIJ Int., 52 (2012), 984.
• 24)   W.  Wang,  S.  Dai,  L.  Zhou,  T.  Zhang,  W.  Tian and  J.  Xu: Ceram. Int., 46 (2020), 13460.
• 25)   J.  Qi,  C.  Liu and  M.  Jiang: ISIJ Int., 58 (2018), 186.
• 26)   L.  Zhou,  W.  Wang,  F.  Ma,  J.  Li,  J.  Wei,  H.  Matsuura and  F.  Tsukihashi: Metall. Mater. Trans. B, 43 (2012), 354.
• 27)   L.  Zhou,  W.  Wang,  J.  Wei and  K.  Zhou: ISIJ Int., 55 (2015), 821.
• 28)   Y.  Kashiwaya,  C.  Cicutti,  A.  Cramb and  K.  Ishii: ISIJ Int., 38 (1998), 348.
• 29)   Y.  Kashiwaya,  C.  Cicutti and  A.  Cramb: ISIJ Int., 38 (1998), 357.
• 30)   L.  Zhou,  Z.  Pan,  W.  Wang,  J.  Chen,  L.  Xue,  T.  Zhang and  L.  Zhang: Metall. Mater. Trans. B, 51 (2020), 85.
• 31)   L.  Zhou,  J.  Li,  W.  Wang and  I.  Sohn: Metall. Mater. Trans. B, 48 (2017), 1943.
• 32)   W.  Wang,  J.  Li,  L.  Zhou and  J.  Yang: Met. Mater. Int., 22 (2016), 700.