Effect of B₂O₃ Addition to Slag on the Dynamic Change Behavior of Interfacial Tension between Liquid Iron and Molten Slag

Abstract

The interfacial tension between liquid steel and molten slag is an important consideration for the continuous casting process because it significantly affects the inclusion of mold flux in liquid steel when the metal/slag interface is disturbed. Specifically, dynamic changes in interfacial tension are observed when liquid iron and molten silicate slag are contacted. This behavior is explained well by a mechanism in which oxygen generated by the decomposition of SiO₂ adsorbs at the interface, temporally decreasing interfacial tension, followed by the desorption of this oxygen from the interface to the bulk metal, recovering the interfacial tension. B₂O₃ has emerged as an alternative to fluorides as a component of mold flux that controls its physicochemical properties, such as liquid viscosity and crystallization behavior. However, its effect on interfacial-tension dynamics is not currently understood. Accordingly, in this study, the dynamic interfacial tension between liquid iron and molten B₂O₃-bearing slag was investigated by the floating lens method. When B₂O₃-bearing slag was used, the interfacial tension significantly decreased to a minimum in the initial stage then slowly increased and finally recovered to the value observed when the B₂O₃-free slag was contacted with liquid iron. The decomposition of both B₂O₃ and SiO₂, providing oxygen to the metal/slag interface, was proposed as a reason for the significant decrease in interfacial tension. Furthermore, it was proposed that the B₂O₃ decomposition occurs continuously, which causes the slow increase of interfacial tension after the initial stage.

1. Introduction

The interfacial tension between liquid metal and molten slag is one of the most important physical properties for many metallurgical processes. Generally, this property determines the stability of a metal/slag interface. Specifically, when the interfacial tension is high, the metal/slag interface tends to remain flat. In contrast, with low interfacial tension, the interface tends to become unstable and inclusion occurs easily upon small disturbances.

In continuous casting processes in steelmaking, for example, mold flux provided at the top surface of the liquid steel forms molten slag consisting of oxide and fluoride components, generating a metal/slag interface. The molten slag acts as a lubricant between the liquid steel and mold. Thus, slag inclusion into the liquid steel should be minimized for clean steel production. Accordingly, the interfacial tension in such systems must be carefully considered when optimizing mold flux composition.

It is commonly accepted that the interfacial tension between liquid steel and molten slag should be kept high to prevent slag inclusion.^1,2) However, the dynamic change of interfacial tension, which has been observed for various kinds of iron-based alloy and silicate slag compositions,^{3,4,5,6,7,8,9,10,11)} should also be considered.

Dynamic change of interfacial tension occurs when a redox reaction takes place at the metal/slag interface. For example, Riboud et al.⁵⁾ used X-ray transmission analysis to observe the dramatic change in interfacial tension when an Fe-5 mass% Al alloy droplet was introduced to a molten-SiO₂-based slag bath, which was attributed to the interfacial redox reaction between the SiO₂ in the slag and the Al in the alloy occurred at the interface. In contrast, Tanaka et al.¹²⁾ investigated the interfacial tension between liquid iron with a low content of Al and a molten-SiO₂-based slag by the floating lens method, where a slag droplet was contacted with the iron surface and the contact angle at the iron/slag/gas interface was measured. Even in this case, the observed interfacial tension changed dynamically with holding time.

Conversely, it is known that the iron/slag interfacial tension at the equilibrium state decreases monotonously with the oxygen content of the liquid iron, regardless of iron and slag composition.^{13,14,15,16,17,18,19)} In addition, oxygen is recognized as a surface-active element that significantly decreases the surface tension of liquid iron and iron/slag interfacial tension by its adsorption at the interface.^20,21)

Based on the above evidence, Tanaka et al.¹²⁾ proposed a mechanism to explain the dynamic change in iron/slag interfacial tension as follows: before the liquid iron and molten slag are contacted, the liquid iron has very low oxygen and silicon contents. At the beginning of the iron/slag contact, the SiO₂ in the slag decomposes to form oxygen and silicon atoms at the iron/slag interface. The oxygen atoms tend to be adsorbed at the interface, temporarily decreasing the interfacial tension. Then, as the desorption of oxygen atoms from the interface to the bulk iron proceeds, the interfacial tension initially presents a minimum and then increases because the amount of adsorbed oxygen at the interface gradually decreases. The mechanism proposed above has been successfully extended to a mathematical model that simulates dynamic changes in interfacial tension quantitatively.^22,23,24)

In the present study, we focus on the effect of the addition of B₂O₃ to molten slag on the dynamic change behavior of iron/slag interfacial tension. B₂O₃ has recently emerged as an alternative mold flux component to replace fluorides, which are becoming increasingly undesirable from an environmental perspective.²⁵⁾ Although B₂O₃ is a network-former in oxide form, its addition effectively decreases the viscosity of molten slag.^26,27,28) Additionally, B₂O₃ also has a significant effect on the thermophysical properties of F-free slag to promote the precipitation of B-bearing crystalline phases, which is expected to contribute to mild cooling of liquid steel.^{29,30,31,32,33,34)} In terms of interfacial tension, it is expected that the B₂O₃ addition will influence the dynamic change behavior of the interfacial tension because consideration of the chemical equilibria between the alloying component, the oxygen dissolved in liquid iron, and the corresponding oxide compound suggests that B₂O₃ decomposition will provide more oxygen than Al₂O₃ decomposition but less than SiO₂ decomposition.³⁵⁾ Therefore, when a B₂O₃-bearing silicate slag is contacted with liquid iron, the B₂O₃ as well as the SiO₂ in the slag will provide oxygen atoms to the iron/slag interface upon their decompositions, which will cause a decrease in interfacial tension dynamically. However, it is unclear how significantly the interfacial tension dynamically changes when B₂O₃-bearing slag is contacted with liquid iron.

Accordingly, the present study aims to clarify the effect of B₂O₃ addition on iron/slag interfacial tension behavior using the floating lens method.

2. Experimental

2.1. Sample Preparation

In this study, one iron sample (I) (provided by Nippon Steel Corporation, Tokyo, Japan) and three different types of slag (X, Y, Z) were treated. The chemical compositions of samples are represented in Tables 1 and 2. The B₂O₃ content in the slag samples was varied, whereas the SiO₂/CaO/Al₂O₃ ratio was fixed at 40:40:20 (by mass%). To prepare the slag samples, powdered calcium carbonate (special grade, FUJIFILM Wako Chemicals Co., Ltd., Osaka, Japan) was calcined at 1223 K in air to prepare CaO powder, Then, it was mixed with silicon dioxide (special grade, Kishida Chemical Co., Ltd., Osaka, Japan) and aluminum oxide (FUJIFILM Wako Chemicals Co., Ltd., Osaka, Japan) in the powder state to the above composition ratio. The mixture was then melted in air in a Pt–Rh crucible by holding at 1823 K for 3 h, and the B-free slag sample X was obtained by quenching the melt on the copper plate. Next, slags Y and Z were synthesized by heating a mixture of slag X with B₂O₃ (special grade, Kanto Chemical Co., Inc., Tokyo, Japan) at 1823 K for 3 h and then quenching the melt.

Table 1. Chemical composition of iron sample I (mass%).

Si	Mn	Al	P	S	N	B
0.029	0.09	0.010	0.01	0.005	0.001	<0.0001

Table 2. Chemical compositions of slag samples X, Y, and Z (mass%).

	SiO2	CaO	Al2O3	B2O3
Slag X	40.0	40.0	20.0	0
Slag Y	39.0	39.0	19.5	2.5
Slag Z	38.0	38.0	19.0	5.0

2.2. Apparatus

The experimental setup for the floating lens method was comparable to that used in our previous work.¹²⁾ Briefly, 80 g of the iron sample was placed in an flat alumina crucible, which was set in the center of the furnace. Then, 0.8 g of the slag sample crushed into powder and shaped as a pellet was placed on a Mo plate above the iron sample. The furnace atmosphere was taken to vacuum and replaced with Ar gas (99.999 vol%, dehydrated by silica gel and magnesium perchlorate and deoxidized by heated magnesium tips), and then the furnace temperature was increased using the graphite heater inside the furnace. When the temperature reached 1823 K, the iron sample was completely melted. Oxygen concentration in the liquid iron with the identical composition was previously evaluated as around 15 mass ppm,¹²⁾ by dipping an oxygen sensor with respect to Cr/Cr₂O₃ reference electrode. Then, the molten slag droplet (held by a Mo rod) was contacted with the liquid iron. The slag/iron contact was held for 1–3 h while the shape of the slag droplet was monitored by a CCD camera. ImageJ software was used to measure the contact angle at the iron/slag/gas interface from the captured images. Considering the balance between the surface tension of the liquid iron, the surface tension of the molten slag, and the interfacial tension between the iron and the slag, the following equation is derived to determine the interfacial tension:   

$${\sigma }_{\mathrm{F}\mathrm{e}/\mathrm{S}\mathrm{l}\mathrm{a}\mathrm{g}}=\sqrt{{\sigma }_{\mathrm{F}\mathrm{e}}^2+{\sigma }_{\mathrm{S}\mathrm{l}\mathrm{a}\mathrm{g}}^2-2{\sigma }_{\mathrm{F}\mathrm{e}}\cdot {\sigma }_{\mathrm{S}\mathrm{l}\mathrm{a}\mathrm{g}}\cdot cos{\theta }_{\mathrm{i}}}$$(1)

where σ_Fe/Slag is the interfacial tension between iron and slag, σ_Fe is the surface tension of the liquid iron, σ_Slag is the surface tension of the molten slag, and θ_i is the apparent contact angle.

2.3. Experimental Procedures

To investigate the effect of B₂O₃ addition on the dynamic change of the iron/slag interfacial tension, the following experiments were conducted:

(i) Evaluating the dynamic interfacial tension between the liquid iron and the B₂O₃-bearing molten slag: Iron I in the liquid state was directly contacted with the molten slag (X, Y, or Z), to determine the effects of B₂O₃ addition to slag on the changing behavior of the iron/slag interfacial tension. In this case, the dynamic interfacial tension was determined over time using Eq. (1), where σ_Fe was estimated using the empirical equation proposed by Ogino et al.³⁶⁾ as functions of oxygen and sulfur concentrations, and σ_Slag was estimated using the semi-empirical model proposed by Nakamoto et al.³⁷⁾ under the assumption that slag composition does not change during iron/slag contact.

(ii) Evaluating the influence of B₂O₃ in slag on the dynamic change of interfacial tension: iron sample II was prepared by the following ways: first, in the same way as explained above for interfacial tension measurements, the molten slag of composition X (B₂O₃-free slag) was dropped on the surface of liquid iron I and kept for 2 h at 1823 K, to transfer sufficient silicon from the slag to the iron for it to reach steady state at the interfacial area. Because our previous work revealed that the interfacial tension between iron sample I and SiO₂-based slag reached the final value and Si transfer from slag to iron tends to become steady state within 30 minutes after the slag/iron contact,¹²⁾ the above holding time was adequate to transfer sufficient silicon. Next, the slag/iron sample was cooled to room temperature in the furnace. Finally, the iron sample II was prepared by removing the contacted slag from the solidified iron. Iron III, for which both the Si and B transfers from the slag to the iron are saturated, was prepared by contacting liquid iron I with molten B-bearing slag Y at 1823 K for 3 h by the same way as mentioned above. Then, iron and slag samples with the following combinations were contacted in the molten state: (i) iron II/slag X, (ii) iron II/slag Y, and (iii) iron III/slag Y. For these cases, the apparent contact angle (θ_i in Eq. (1)) was used as a measure to evaluate the dynamic change of the interfacial tension, whereas the interfacial tension was not determined because the oxygen contents of liquid iron II and III were unknown. However, unless σ_Fe and σ_Slag change significantly during the iron/slag contact, it may be assumed that the dynamic change of the contact angle reflects that of the iron/slag interfacial tension.

For each kind of experiment, the chemical compositions of the slag and iron samples around the contacted area were determined by quickly solidifying the iron and then slag in the furnace after different holding times, cutting out the contacted area, and analyzing it by induction coupled plasma (ICP) analysis.

3. Results and Discussion

3.1. Dynamic Change of Interfacial Tension between Liquid Iron and Molten B₂O₃-bearing Slag

Figure 1 shows images of the slag droplet contacted with the liquid iron surface at different holding times, where one kind of iron (I) and three kinds of slag (X, Y, Z) were treated in combination. For each case, the slag droplet was held on the flat surface of the liquid iron, and the contact angle at the iron/slag/gas interface was monitored. For B₂O₃-free slag X, no specific change in the slag droplet shape is observed during the contact. However, for B₂O₃-bearing slags Y and Z, the slag droplet spreads on the iron surface immediately upon contact and becomes narrower with time.

Fig. 1.

Images of a slag droplet contacted with liquid iron at different holding times: (a) iron I and slag X (B₂O₃ = 0 mass%), (b) iron I and slag Y (B₂O₃ = 2.5 mass%), (c) iron I and slag Z (B₂O₃ = 5.0 mass%).

Figure 2 shows the changes in apparent contact angles at the iron/slag/gas interface for each sample against holding time, where the moment of iron/slag contact is defined as 0 min. For each case, the contact angle initially decreases to a minimum at ~3 min after contact, then it gradually increases to reach its final value in the later stage. For B₂O₃-bearing slags Y and Z, the following two features are observed: first, the decrease of the contact angle in the initial stage is significant, and the extent of this change increases as B₂O₃ content increases. Second, after the minimum, the contact angle monotonously increases over a long period until it reaches its final value.

Fig. 2.

Change in apparent contact angle at the iron/slag/gas interface against holding time, showing the effect of B₂O₃ content in slag (the moment of iron/slag contact is defined as 0 min).

It should be noted that the surface tension of the slag (σ_Slag) decreases upon adding B₂O₃,^37,38) which decreases the apparent contact angle to maintain the force valence in the horizontal direction at the iron/slag/gas interface. Therefore, the interfacial tension should be derived to compensate for the effect of the decrease in slag surface tension on the contact angle.

Figure 3 shows the dynamic changes in interfacial tension between liquid iron and molten slag, where Eq. (1) and the measured contact angle were used to derive the interfacial tension value. Because we assumed that both the surface tensions of iron and slag do not change during holding time, the changes of the interfacial tensions represent a similar tendency to that of the contact angles (Fig. 2). Here, it should be noted that although the contact angle at 0 min could not be accurately measured because the slag droplet wetted both the liquid iron surface and the molybdenum plate, one can assume that the interfacial tensions at 0 min should have been highest in the profiles and comparable among the samples because they are determined by the initial oxygen content in the liquid iron. Thus, while the B₂O₃ content is only several weight percent, the interfacial tension between liquid iron I and the B₂O₃-bearing slags (Y and Z) exhibit a very sudden decrease in the initial stage. Then, the interfacial tensions in the later stage gradually increase to reach their final values, which are comparable to the final interfacial tension value between iron I and the B₂O₃-free slag (X).

Fig. 3.

Change in interfacial tension between liquid iron and molten slag against holding time, where Eq. (1) and apparent contact angle (Fig. 2) are used, while the surface tensions of liquid iron and molten slag at the initial compositions are estimated.

Figure 4(a) shows the Si and B contents of the liquid iron against holding time, while Fig. 4(b) shows the SiO₂ and B₂O₃ contents in the molten slag against holding time. In this case, the liquid iron I and slag Y (desired B₂O₃ content was 2.5 mass%) are contacted. The results in Fig. 4(a) demonstrate that the Si content in the liquid iron monotonously increases during the iron and slag contact, while the B content also increases but plateaus in the later stage. In contrast, as shown in Fig. 4(b), both the SiO₂ and B₂O₃ contents slightly decrease during iron and slag contact. The B₂O₃ loss may have partially included the influence of its vaporization. However, the effect of the B₂O₃ loss as shown in Fig. 4(b) on the surface tension of molten slag was estimated as too small to be recognized by Nakamoto model,³⁷⁾ therefore its effect on the interfacial tension evaluated by Eq. (1) was also regarded as negligible.

Fig. 4.

(a) Si and B contents of liquid iron I, and (b) SiO₂ and B₂O₃ contents of molten slag Y against holding time after iron/slag contact.

The above results indicate that Si and B elements are transferred from the slag to iron upon the decomposition of SiO₂ and B₂O₃ at the iron/slag interface, and that the Si and B supplies are maintained for a long time while the iron and slag are in contact.

The dynamic change behavior of the interfacial tension between iron I and B₂O₃-bearing slags (Y and Z) shown in Fig. 3 may be explained as follows: First, the sudden decrease in interfacial tension in the initial stage is due to oxygen supply to the iron/slag interface by the decomposition of both SiO₂ and B₂O₃ in the slag, as follows:   

$${\left(\mathrm{S}\mathrm{i}{\mathrm{O}}_2\right)}_{\mathrm{S}\mathrm{l}\mathrm{a}\mathrm{g}}\to \underset{\_}{\mathrm{S}\mathrm{i}}+2\underset{\_}{\mathrm{O}}$$(2)

  

$${\left({\mathrm{B}}_2{\mathrm{O}}_3\right)}_{\mathrm{S}\mathrm{l}\mathrm{a}\mathrm{g}}\to 2\underset{\_}{\mathrm{B}}+3\underset{\_}{\mathrm{O}}$$(3)

where A denotes elemental A distributed in the liquid iron. While Si and B elements are diffused from the interface to the bulk iron, oxygen tends to be adsorbed at the interface and decreases the interfacial tension. Unlike the case in which the liquid iron I and B₂O₃-free slag (X) are contacted, the decomposition of B₂O₃ (Eq. (3)) as well as SiO₂ occurs at the interface and thus abundant oxygen is supplied to the interface. Particularly, because B₂O₃ has a lower surface tension in the pure liquid state compared with other oxide components in the slag,^37,38) B₂O₃ would be segregated at the surface of the molten slag before the slag is contacted with the liquid iron. Table 3 shows the calculated surface tensions and B₂O₃ surface-area contents for slags X, Y, and Z, indicating that the B₂O₃ surface-area contents are ~10-times those in the bulk. Therefore, B₂O₃ decomposition at the interface is further promoted at the beginning of iron–slag contact.

Table 3. Surface tensions and B₂O₃ surface-area contents for slag samples calculated by a semi-empirical model.³⁷⁾

	B2O3 content in bulk (mol%)	B2O3 content in surface (mol%)	Surface tension, σSlag [mN·m−1]
Slag X	0	0	434
Slag Y	1.7	16.4	421
Slag Z	3.4	28.1	409

Additionally, B₂O₃ addition also influences slag viscosity and SiO₂ activity, as shown in Table 4 (the data were calculated using the FactSage software and the latest FACT oxide database.³⁹⁾ The detail of the viscosity model can be found elsewhere.⁴⁰⁾) As B₂O₃ content increases, viscosity slightly decreases while the SiO₂ activity increases. These changes could also increase oxygen supply to the interface and thus decrease the interfacial tension, because mass transfer becomes easier when viscosity decreases and SiO₂ decomposition (Eq. (2)) is promoted when SiO₂ activity increases. However, these indirect influences would not be significant because these changes are small.

Table 4. Viscosity, SiO₂ activity, and B₂O₃ activity of liquid slag at 1823 K calculated by FactSage.³⁹⁾

	Viscosity of slag, η [Pa·s]	SiO2 activity, $a_{\mathrm{S}\mathrm{i}{\mathrm{O}}_2}$	B2O3 activity, $a_{{\mathrm{B}}_2{\mathrm{O}}_3}$
Slag X (B2O3 = 0 mass%)	0.51	0.09	0
Slag Y (B2O3 = 2.5 mass%)	0.50	0.12	2.3 × 10−5
Slag Z (B2O3 = 5.0 mass%)	0.48	0.14	1.5 × 10−4

Next, the slow increase of the interfacial tension in the later stage may be explained by the continuous supply of SiO₂ and B₂O₃ from the bulk slag to the interface, as indicated in Fig. 4. This phenomenon occurs for the Si and B transfer to the liquid iron because these elements readily diffuse from the interface to the bulk iron. As the elemental oxygen adsorbed on the interface gradually diffuses to the bulk iron, the SiO₂ and B₂O₃ provide more oxygen to the interface upon their decomposition, which delays the recovery of interfacial tension. The more critical the decrease of interfacial tension in the initial stage, the longer it takes to recover to the final value. Here, because the Al content of iron I is low, the effect of Al oxidation on the removal of adsorbed oxygen at the interface is negligible.

Finally, irrespective of B₂O₃ addition, the final values of interfacial tension are comparable. This may be explained as follows: first, the chemical equilibrium between the SiO₂ in the slag and the Si in the iron provide more oxygen to the liquid iron than that between the B₂O₃ in the slag and the B in the iron.³⁵⁾ Second, the activity of the SiO₂ in the slag is much higher than that of the B₂O₃, as shown in Table 4. Consequently, the final value of the interfacial tension is mainly determined by the oxygen content in the liquid iron equilibrated with the Si in the iron and the SiO₂ in the slag.

3.2. Effect of B₂O₃ in Slag on the Dynamic Change of Iron/slag Interfacial Tension

To verify that B₂O₃ addition to slag directly affects the dynamic change of the iron/slag interfacial tension by the mechanism explained above, the changes in the apparent contact angles were measured for the following combinations: (i) iron II/slag X, (ii) iron II/slag Y, and (iii) iron III/slag Y. The method to prepare the iron II and III samples is represented in the section 2.3. The iron II was prepared for Si transfer from B₂O₃-free slag (X) to iron to be saturated, while the iron III was prepared for both Si and B transfer from B₂O₃-bearing slag (Y) to iron to be saturated. The iron I was contacted with slag Y for 3 h to make iron III, where this holding time was adequate because the interfacial tension between iron I and slag Y reached final value (Fig. 3) and the B transfer tended to become steady state (Fig. 4) at 3 h after iron/slag contact.

Figure 5 shows the apparent contact angles at the iron/slag/gas interface for the above combinations (i)–(iii) against holding time. For cases (i) and (iii), the changes in the contact angle are within ± 0.5° and thus negligible. For case (ii), however, dynamic change in the contact angle is observed. It significantly decreases in the initial stage, presents a minimum at ~3 min, and then increases throughout the later stage. Unlike the case in which iron I and slag Y are contacted (Fig. 2), the rate of increase in the interfacial tension is initially high. However, it gradually decreases. After a comparatively long time (Run2 in Fig. 5), the contact angle finally plateaus.

Fig. 5.

Change in apparent contact angle at the iron/slag/gas interface for the following combinations: (i) iron II and slag X, (ii) iron II and slag Y, (iii) iron III and slag Y.

Figure 6 shows the (a) Si and B content in the iron, and (b) the SiO₂ and B₂O₃ contents in the slag for case (ii) (i.e., iron II/slag Y contact) against holding time. Unlike the case in which iron I and slag Y are contacted (Fig. 4), the changes in iron Si content in and slag SiO₂ content are not significant. However, the B content in the iron increases and the B₂O₃ content in the slag decreases as much as the case where iron I and slag Y were contacted. The above tendency indicates that the B transfer from the slag to the iron mainly occurred through the iron/slag interface while the Si transfer was not significant during the iron II and the slag Y contact in the molten state.

Fig. 6.

(a) Si and B contents in liquid iron II, and (b) SiO₂ and B₂O₃ contents of molten slag Y against holding time after iron/slag contact.

As shown in Fig. 5, the fact that the contact angle is fixed in case (i) indicates that the change in the interfacial tension does not occur because the Si transfer from the slag to the iron is already saturated and hence the oxygen content at the interface does not change. Similarly, the result for case (iii) indicates that the interfacial tension does not change because both the Si and B transfer are saturated. However, in case (ii), Si transfer is almost saturated, but B transfer is not saturated. Hence, when the iron II is contacted with B₂O₃-bearing slag Y, B transfer from the slag to the iron is dominant. Thus, the decreasing tendency of the contact angle in the initial stage is because this B transfer accompanies oxygen supply to the interface upon the decomposition of B₂O₃ (Eq. (3)), which decreases the interfacial tension. The amount of the oxygen supplied is so small compared with the volume of liquid iron that the increase of oxygen content in the iron would be negligible. While the oxygen at the interface is gradually desorbed by diffusion to the bulk iron, the above oxygen supply by B transfer through B₂O₃ decomposition continues until the B transfer becomes saturated. This delays the recovery of the interfacial tension and thus alters the change in contact angle in the later stage.

Consequently, it is verified that B₂O₃ in slag directly affects the dynamic change of the interfacial tension between liquid iron and molten slag. If the Si transfer from the slag to the iron is saturated, the decomposition of B₂O₃ occurs to supply elemental B and oxygen to the interface. When elemental B transfers from the interface to the bulk iron, oxygen tends to be adsorbed at the interface and decreases the interfacial tension. The recovery of the interfacial tension proceeds in the later stage by the desorption of oxygen at the interface, while oxygen is continuously supplied until B transfer from the slag to the bulk is saturated.

4. Conclusions

In this study, the effect of B₂O₃ addition to slag on the dynamic change of the interfacial tension between liquid iron and molten slag was studied experimentally by the floating lens method. The conclusions obtained are summarized as follows:

(1) Compared with the case where B₂O₃-free silicate slag is contacted with liquid iron, the dynamic change of the interfacial tension when B₂O₃-bearing slag is contacted with liquid iron exhibits two features: first, the interfacial tension significantly decreases in the initial stage of iron/slag contact. Second, after presenting a minimum, the interfacial tension gradually increases for over time, reaching a final value that is comparable to that for B₂O₃-free silicate slag contacting with liquid iron. As B₂O₃ content increases, the above effects on interfacial tension become more significant. Chemical analyses of iron and slag samples revealed that increases in iron Si and B contents as well as decreases in slag SiO₂ and B₂O₃ contents occur continuously during iron/slag contact. Thus, the above behaviors in interfacial tension were attributed to the supply of a large amount of oxygen to the interface by the decomposition of the B₂O₃ as well as the SiO₂ in the slag. The adsorption of the supplied oxygen significantly decreases the interfacial tension in the initial stage. Additionally, the oxygen supply is maintained over time while the adsorbed oxygen diffuses to the bulk iron until the Si and B transfers from the slag to the iron are saturated, which delays the increase of the interfacial tension in the later stage.

(2) If the Si transfer from the slag to the iron is saturated after the liquid iron is pre-contacted with the B₂O₃-free silicate slag, the apparent contact angle between the liquid iron and the B₂O₃-bearing slag changes dynamically: it initially decreases to exhibit a minimum, then gradually increases over time to its final value. The observed changes of the contact angle reflect the dynamic change of the interfacial tension. Chemical analyses indicated that the Si transfer from the slag to the iron is not predominant, but instead that B transfer through B₂O₃ decomposition at the is dominant during iron/slag contact. Therefore, in this case, the interfacial tension dynamically changes owing to the additional supply of oxygen by B₂O₃ decomposition at the interface, which is driven by B transfer from the slag to the iron.

Acknowledgements

This work was financially supported by Japan Society Promotion Society KAKENHI (Grants-in-Aid for Scientific Research) Number 17H03437. We thank warmly to Mr. Takahiro Nakano and Mr. Takumi Kageyama (graduated students, Osaka University, Japan) for their efforts on the preliminary experimental works. We thank Dr Jay Freeman at Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.

References

• 1)   M.  Hanao and  M.  Kawamoto: Tetsu-to-Hagané, 93 (2007), 362 (in Japanese). https://doi.org/10.2355/tetsutohagane.93.362
• 2)   K.  Watanabe,  K.  Tsutsumi,  M.  Suzuki,  M.  Nakada and  T.  Shiomi: ISIJ Int., 49 (2009), 1161. https://doi.org/10.2355/isijinternational.49.1161
• 3)   P.  Kozakevitch,  G.  Urbain and  M.  Sage: Rev. Metall., 52 (1955), 161.
• 4)   H.  Ooi,  T.  Nozaki and  H.  Yoshii: Trans. Iron Steel Inst. Jpn., 14 (1974), 9. https://doi.org/10.2355/isijinternational1966.14.9
• 5)   P. V.  Riboud and  L. D.  Lucas: Can. Metall. Q., 20 (1981), 199. https://doi.org/10.1179/cmq.1981.20.2.199
• 6)   H.  Gaye,  L. D.  Lucas,  M.  Olette and  P. V.  Riboud: Can. Metall. Q., 23 (1984), 179. https://doi.org/10.1179/cmq.1984.23.2.179
• 7)   A.  Sharan and  A. W.  Cramb: Metall. Mater. Trans. B, 26 (1995), 87. https://doi.org/10.1007/BF02648981
• 8)   I.  Jimbo,  Y.  Chung and  A. W.  Cramb: ISIJ Int., 36 (1996), S42. https://doi.org/10.2355/isijinternational.36.Suppl_S42
• 9)   Y.  Chung and  A. W.  Cramb: Philos. Trans. R. Soc. A, 356 (1998), 981. https://doi.org/10.1098/rsta.1998.0202
• 10)   Y.  Chung and  A. W.  Cramb: Metall. Mater. Trans. B, 31 (2000), 957. https://doi.org/10.1007/s11663-000-0072-5
• 11)   M. A.  Rhamdhani,  K. S.  Coley and  G. A.  Brooks: Metall. Mater. Trans. B, 36 (2005), 591. https://doi.org/10.1007/s11663-005-0050-z
• 12)   T.  Tanaka,  H.  Goto,  M.  Nakamoto,  M.  Suzuki,  M.  Hanao,  M.  Zeze,  H.  Yamamura and  T.  Yoshikawa: ISIJ Int., 56 (2016), 944. https://doi.org/10.2355/isijinternational.ISIJINT-2015-654
• 13)   K.  Mukai,  T.  Kato and  H.  Sakao: Tetsu-to-Hagané, 59 (1973), 55 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.59.1_55
• 14)   K.  Ogino,  S.  Hara,  T.  Miwa and  S.  Kimoto: Tetsu-to-Hagané, 65 (1979), 2012 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.65.14_2012
• 15)   K.  Ogino,  S.  Hara,  T.  Miwa and  S.  Kimoto: Trans. Iron Steel Inst. Jpn., 24 (1984), 522. https://doi.org/10.2355/isijinternational1966.24.522
• 16)   K.  Nakashima and  K.  Mori: ISIJ Int., 32 (1992), 11. https://doi.org/10.2355/isijinternational.32.11
• 17)   H.  Sun,  K.  Nakashima and  K.  Mori: ISIJ Int., 37 (1997), 323. https://doi.org/10.2355/isijinternational.37.323
• 18)   T.  Tanaka and  S.  Hara: Z. Metallkd., 90 (1999), 348.
• 19)   H.  Sun,  K.  Nakashima and  K.  Mori: ISIJ Int., 46 (2006), 407. https://doi.org/10.2355/isijinternational.46.407
• 20)   G. R.  Belton: Metall. Trans. B, 7 (1976), 35. https://doi.org/10.1007/BF02652817
• 21)   T.  Tanaka and  S.  Hara: Steel Res., 72 (2001), 439. https://doi.org/10.1002/srin.200100149
• 22)   P.  Ni,  T.  Tanaka,  M.  Suzuki,  M.  Nakamoto and  P. G.  Jönsson: ISIJ Int., 58 (2018), 1979. https://doi.org/10.2355/isijinternational.ISIJINT-2018-303
• 23)   P.  Ni,  T.  Tanaka,  M.  Suzuki,  M.  Nakamoto and  P. G.  Jönsson: ISIJ Int., 59 (2019), 737. https://doi.org/10.2355/isijinternational.ISIJINT-2018-496
• 24)   P.  Ni,  T.  Tanaka,  M.  Suzuki,  M.  Nakamoto,  M.  Ersson and  P. G.  Jönsson: ISIJ Int., 59 (2019), 2024. https://doi.org/10.2355/isijinternational.ISIJINT-2019-007
• 25)   J.  Gao,  G.  Wen,  Q.  Liu,  W.  Tan and  P.  Tang: J. Non-Cryst. Solids, 409 (2015), 8. https://doi.org/10.1016/j.jnoncrysol.2014.11.014
• 26)   Z.  Wang,  Q.  Shu and  K.  Chou: Steel Res. Int., 84 (2013), 766. https://doi.org/10.1002/srin.201200256
• 27)   L.  Zhang,  W.  Wang,  S.  Xie,  K.  Zhang and  I.  Sohn: J. Non-Cryst. Solids, 460 (2017), 113. https://doi.org/10.1016/j.jnoncrysol.2017.01.031
• 28)   T. M.  Yeo,  J. W.  Cho,  M.  Alloni,  S.  Casagrande and  R.  Carli: J. Non-Cryst. Solids, 529 (2020), 119756. https://doi.org/10.1016/j.jnoncrysol.2019.119756
• 29)   B.  Lu,  W.  Wang,  J.  Li,  H.  Zhao and  D.  Huang: Metall. Mater. Trans. B, 44 (2013), 365. https://doi.org/10.1007/s11663-012-9784-6
• 30)   L.  Zhou,  W.  Wang and  K.  Zhou: ISIJ Int., 55 (2015), 1916. https://doi.org/10.2355/isijinternational.ISIJINT-2015-193
• 31)   J. Y.  Park,  G. H.  Kim,  J. B.  Kim,  S.  Park and  I.  Sohn: Metall. Mater. Trans. B, 47 (2016), 2582. https://doi.org/10.1007/s11663-016-0720-z
• 32)   J.  Yang,  J.  Zhang,  Y.  Sasaki,  O.  Ostrovski,  C.  Zhang,  D.  Cai and  Y.  Kashiwaya: Metall. Mater. Trans. B, 48 (2017), 2077. https://doi.org/10.1007/s11663-017-0997-6
• 33)   W.  Wang,  D.  Cai and  L.  Zhang: ISIJ Int., 58 (2018), 1957. https://doi.org/10.2355/isijinternational.ISIJINT-2018-232
• 34)   L.  Zhang,  W.  Wang and  I.  Sohn: J. Non-Cryst. Solids, 511 (2019), 41. https://doi.org/10.1016/j.jnoncrysol.2019.01.035
• 35)  The Japan Society for the Promotion of Science, The 19th Committee on Steelmaking, ed.: Steelmaking Data Sourcebook, Gordon and Breach Science Publishers, New York, (1988), 10.
• 36)   K.  Ogino,  K.  Nogi and  C.  Hosoi: Tetsu-to-Hagané, 69 (1983), 1989 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.69.16_1989
• 37)   M.  Nakamoto,  T.  Tanaka,  L.  Holappa and  M.  Hämäläinen: ISIJ Int., 47 (2007), 211. https://doi.org/10.2355/isijinternational.47.211
• 38)   X.  Shi,  Q.  Wang,  X.  Niu,  C.  Li and  K.  Lu: J. Am. Ceram. Soc., 89 (2006), 3222. https://doi.org/10.1111/j.1551-2916.2006.01215.x
• 39)   C. W.  Bale,  A. D.  Pelton,  W. T.  Thompson,  P.  Chartland,  S.  Decterov,  I.-H.  Jung,  J.  Melancon,  G.  Eriksson,  K.  Hack and  S.  Peterson: FactSage, CRCT (Montreal, Canada), GTT-Technologies (Aachen, Germany), http://www.factsage.com/, (accessed 2021-08-31).
• 40)   E.  Brosh,  A. D.  Pelton and  S. A.  Decterov: Int. J. Mater. Res., 103 (2012), 537. https://doi.org/10.3139/146.110639