Comparison of Oxidation Behavior of Various Reactive Elements in Alloys during Electroslag Remelting (ESR) Process: An Overview

Abstract

The oxidation behavior of various reactive elements, such as Al/Ti, B/Si, and rare earth metals (REM), by electroslag remelting (ESR) type slag has been investigated utilizing the systematic thermodynamic analysis based on the calculated activity in the fluoride-containing slags by the ion and molecule coexistence theory (IMCT). The results indicate that the IMCT model can be reliably applied to calculate the activity of each component in the ESR type slag. The oxidation behavior of the various reactive elements is completely different by changing the same component in the slag, such as CaO and Al₂O₃, during the ESR process. Therefore, it is indispensable to find a key parameter to control the homogeneity of the reactive elements in remelted ingots by comparing the effects of the activity of each component and temperature on the equilibrium content of the various reactive elements in alloy melts. The results demonstrate that the oxidation loss of Al/Ti, B/Si, and REM (Ce and La) can be effectively prevented by employing TiO₂, B₂O₃/SiO₂, and CaO during the ESR process. Oxidation behavior of B/Si and Ce(La)/Al is weakly susceptible to temperature fluctuation compared with that of Al/Ti in alloy melts, which can be controlled by adding TiO₂.

1. Introduction

Electroslag remelting (ESR) is an advanced secondary refining technology to produce clean, fully dense, and homogeneous castings of steels and alloys by removal of undesirable elements and non-metallic inclusions.^1,2,3) However, the alloying elements, such as Al/Ti, B/Si, rare earth metals (REM), in consumable electrode react with oxides in the slag during the ESR process,^4,5,6) i.e., FeO, SiO₂, TiO₂, and Al₂O₃. Hence, the reactive elements cannot be maintained uniformly from the bottom to the top of the resultant ingots and subsequently inadequate mechanical properties of the metallic materials.¹⁾

Since the 1970s, the control of oxidation of alloying elements in ingots during the ESR process has been of interest to researchers and thus many efforts have been made on the above-mentioned issues. Pateisky et al.⁷⁾ tried to develop the relationship between the Al, Ti, and Si content in steel and the SiO₂ and TiO₂ content in slag by statistical evaluation of the results of laboratory-scale experiments and 320 ESR heats. Chen et al.⁸⁾ investigated the effect of slag composition on the Mg, Al, and Ti content in nickel-based alloy ingots and built the functions between the reactive element contents and melting rate as well as the area of interface between slag bath and metal pool. The equilibrium content of the reactive alloying elements in remelted ingots is related to alloy composition, slag composition, and temperature, and hence the establishment of the relationship between equilibrium content of the reactive alloying elements in remelted ingots and slag composition at different temperatures is an essential method to prevent the oxidation loss of the reactive elements during the ESR process at various remelting conditions.⁹⁾

Therefore, a linear relationship between the composition ratio of [%Ti]³/[%Al]⁴ in GH8825 nickel-based alloy and $X_{\mathrm{T}\mathrm{i}{\mathrm{O}}_2}^3$/$X_{\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3}^2$ in slag at different CaO contents has been adopted by Jiang et al.¹⁰⁾ to control the oxidation loss of Al and Ti in the nickel-based alloy by the CaF₂–CaO–Al₂O₃–MgO–TiO₂–SiO₂ slag at 1823 K. Recently, Duan et al.¹¹⁾ systemically analyzed the effect of slag component in the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag on the equilibrium Al content in 0.02C-0.2Si-3.1Mo-18.3Cr-4.9Nb-0.43Al-1.1Ti Ni-based alloy (i.e., Inconel^® 718, mass%) at the temperature range from 1773 K to 1973 K. From the viewpoint of kinetics, Hou et al.^12,13,14) published a series of papers on the prediction of change of Al and Ti contents along with the height of 1Cr21Ni5Ti ESR ingots utilizing mass transfer model.

Compared with the control of oxidation loss of Al and Ti in steels and alloys during the ESR process, the fundamental studies on the control of the homogeneous distribution of other reactive alloying elements, such as B/Si, and REM is scarce. Kim et al.¹⁵⁾ and Peng et al.¹⁶⁾ designed an appropriate composition of the CaF₂–CaO–Al₂O₃–SiO₂(–MgO)–B₂O₃ slag for electroslag remelting of 9CrMoCoB steel to keep the Si and B content within the optimum range and further testified by the results of the pilot experiments in industrial ESR furnace. For REM, Weng et al.,¹⁷⁾ Wen et al.,¹⁸⁾ Niu et al.,¹⁹⁾ and Ren et al.²⁰⁾ determined the yield of Ce, La, and Y in the remelted ingots after the ESR process. However, the oxidation behavior of the abovementioned reactive elements by the CaF₂–CaO–Al₂O₃ based slag is completely different during the ESR process, which should be elaborated to design appropriate slag composition to maintain the homogenous distribution of the alloying elements in ingots during the ESR process.

As previously mentioned, the activities of the component in slag and alloy melts are important to the thermodynamic and kinetic analysis of oxidation loss of the reactive elements in alloys during the ESR process. However, there are few studies on the experimental determination of the activity of components in liquid slag containing fluoride,^{21,22,23,24,25,26)} which is likely because the CaF₂-containing slag melts are volatile at high temperatures.^27,28,29) Meanwhile, the thermodynamic model for calculating activity in fluoride-containing slag is rarely reported up to now. Because the basic meaning of mass action concentration is almost consistent with the traditionally applied activity defined in the classical metallurgical thermodynamics, in which pure solid matter is chosen as a standard state and mole faction is selected as a concentration unit. The ion and molecule coexistence theory (IMCT) has gradually been approved by numerous investigators to calculate the mass action concentration (i.e., activity) of structural units in liquid oxide systems.^30,31)

To date, the application of IMCT has also been further expanded to CaF₂-containing slags used for the ESR process.^32,33) FactSage^TM software (Hereinafter referred to as thermodynamic calculation software) has been popularly applied in the iron- and steelmaking process to calculate some thermodynamic parameters and physicochemical properties of metallurgical melts.³⁴⁾ However, the accuracy of the calculated activities by the aforesaid methods has been rarely compared. In the present study, therefore, both the activities in CaF₂ containing slags calculated by the IMCT model and the thermodynamic calculation software are used to compare each other and further investigate the effect of slag composition and temperature on the oxidation behavior of various reactive elements, such as Al/Ti, B/Si, and Ce(La)/Al, in alloy melts, which is to find a key parameter to restraint the oxidation loss during the ESR process.

2. Evaluation of Activity in CaF₂–CaO–Al₂O₃ Based Slags

2.1. Activities of Al₂O₃ and TiO₂ in CaF₂–CaO–Al₂O₃–MgO–TiO₂ Slag

The addition of TiO₂ into the CaF₂–CaO–Al₂O₃ slag is an effective method to prevent the oxidation loss of Al and Ti in steels and alloys during the ESR process, especially for the initial temperature-rising (called ramp-up) period.^9,13,35) From the early discussion, although the activities of Al₂O₃ and TiO₂ in the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag is the indispensable parameters, to the best of the present authors’ knowledge, the experimental determination of activities of Al₂O₃ and TiO₂ in the slag has not been reported so far. Therefore, the activities of CaF₂, CaO, Al₂O₃, and SiO₂ in the CaF₂–CaO–Al₂O₃ and CaF₂–CaO–SiO₂ slag measured by Zaitsev et al.^25,26) are firstly used to testify the feasibility of the IMCT model. Zaitsev et al.²⁵⁾ determined the activities of CaF₂, CaO, and Al₂O₃ in the liquid CaF₂–CaO–Al₂O₃ slag using the Knudsen effusion method with a mass-spectrometric analysis of evaporation products at the temperature range from 1600 K to 1830 K. According to the assumption of IMCT,³⁰⁾ a total of 7 kinds of the complex molecules can be found in the CaF₂–CaO–Al₂O₃ slag at metallurgical temperatures, such as 3CaO·Al₂O₃, 12CaO·7Al₂O₃, CaO·Al₂O₃, CaO·2Al₂O₃, CaO·6Al₂O₃, 3CaO·2Al₂O₃·CaF₂, and 11CaO·7Al₂O₃·CaF₂, and the modeling process of the IMCT has been described elsewhere.³³⁾

The comparisons between the calculated and determined activities in the CaF₂–CaO–Al₂O₃ slag are shown in Fig. 1. It can be observed that the activities of CaF₂, CaO, and Al₂O₃ calculated by IMCT have a small deviation from the measured results at the temperature range from 1600 K to 1830 K. The compositions of the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag listed in the literature were employed to calculate the activity of each component using different methods.^11,35,36) The comparisons of the activity of each component in the slag calculated by the thermodynamic calculation software (FactSage Ver. 7.3, used module: Equilib; used databases: FactPS and FT oxide) and the IMCT model are shown in Fig. 2. It can be seen that the activities calculated by thermodynamic software are consistent with those determined by the IMCT model within acceptable scatters, while the activity of Al₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag calculated by thermodynamic software is slightly higher compared with that by IMCT model, which has been proved by Ju et al.³⁷⁾ in the work of the slag-metal equilibrium reaction between CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag and 0.01C-0.13Si-3.2Mo-21Cr-1.7Cu-0.12Al-1.0Ti Ni-based alloy (i.e., Incoloy^® 825, mass%) at the temperature range from 1773 K to 1973 K. These results indicate that the IMCT model can be applied to calculate the activities of Al₂O₃ and TiO₂ in the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag. Therefore, the relationship between the activity coefficients of Al₂O₃ (γ_Al2O3) and TiO₂ (γ_TiO2) in 37CaF₂-25CaO-25Al₂O₃-3MgO (mass%) slag by adding up to 10 mass% TiO₂ can be evaluated by employing the IMCT model at the temperature range from 1773 K to 1973 K as shown following equations, respectively.   

$$log {\gamma }_{\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3}=-\frac{215.3}{T}-0.249\mathrm{\hspace{0.17em} } \left(1\mathrm{\hspace{0.17em} }773\mathrm{\hspace{0.17em} } \mathrm{K}<T<1\mathrm{\hspace{0.17em} }973\mathrm{\hspace{0.17em} } \mathrm{K}\right)$$(1)

  

$$log {\gamma }_{\mathrm{T}\mathrm{i}{\mathrm{O}}_2}=-\frac{4\mathrm{\hspace{0.17em} }532.7}{T}+1.687\mathrm{\hspace{0.17em} } \left(1\mathrm{\hspace{0.17em} }773\mathrm{\hspace{0.17em} } \mathrm{K}<T<1\mathrm{\hspace{0.17em} }973\mathrm{\hspace{0.17em} } \mathrm{K}\right)$$(2)

Fig. 1.

Comparisons between the calculated and measured activities in the CaF₂–CaO–Al₂O₃ slag in the temperature range of 1600 K to 1830 K by Zaitsev et al.²⁵⁾ (Online version in color.)

Fig. 2.

Comparisons of activity of each component in the CaF₂–CaO–Al₂O₃–(MgO)–TiO₂ slag calculated by the thermodynamic calculation software and IMCT model using the slag composition reported by Duan et al.,¹¹⁾ Yang et al.,³⁵⁾ and Ju et al.³⁶⁾ at the temperature range from 1773 K to 1873 K. (Online version in color.)

2.2. Activities of SiO₂ and B₂O₃ in CaF₂–CaO–Al₂O₃–B₂O₃–SiO₂ Slag

Zaitsev et al.²⁶⁾ also determined the activity of SiO₂ in the CaF₂–CaO–SiO₂ slag using the Knudsen effusion method with a mass-spectrometric analysis of evaporation products at 1823 K, and the results are listed in Table 1. According to the hypothesis of IMCT,³⁰⁾ total 5 kinds of complex molecules can be found in the CaF₂–CaO–SiO₂ slag,³³⁾ such as 3CaO·SiO₂, 3CaO·2SiO₂, 2CaO·SiO₂, CaO·SiO₂, and 3CaO·2SiO₂·CaF₂. To certify the accuracy of the IMCT model, the activity of SiO₂ in the CaF₂–CaO–SiO₂ slag measured by Zaitsev et al.,²⁶⁾ Isaksson et al.,³⁸⁾ Sommerville et al.,²⁴⁾ and Hsu et al.,³⁹⁾ as well as calculated by thermodynamic calculation software are also listed in Table 1 for comparison. It can be noted from Table 1 that both calculated activities of SiO₂ in the CaF₂–CaO–SiO₂ slag by the IMCT model and the computational software are in good agreement with the measured values by Zaitsev et al.²⁶⁾ at 1823 K. These results demonstrate that the IMCT model can be reliably applied to predict the activity of SiO₂ in CaF₂-containing slags.

Table 1. Comparisons between the calculated and measured activity of SiO₂ in the CaF₂–CaO–SiO₂ slag at 1823 K.

Mole fraction			Measured values				Calculated values
xCaF2	xCaO	xSiO2	Zaitsev20) (1823 K)	Isaksson32) (1823 K)	Sommerville18) (1723 K)	Hsu33) (1723 K)	by IMCT	by FactSage
0.37	0.52	0.11	2.3×10−5	4.0×10−4	–	7.0×10−8	2.8×10−5	1.3×10−5
0.35	0.50	0.15	4.1×10−5	8.0×10−4	5.2×10−3	4.0×10−6	7.6×10−5	1.8×10−5
0.34	0.47	0.19	1.8×10−4	2.4×10−3	1.1×10−3	8.0×10−5	2.4×10−4	4.6×10−5
0.55	0.34	0.11	9.2×10−5	1.0×10−5	5.8×10−3	1.0×10−6	1.8×10−4	2.5×10−5
0.14	0.32	0.54	2.5×10−4	4.2×10−3	9.6×10−3	6.0×10−5	3.0×10−1	5.9×10−1

The previous researchers have experimentally determined the activity of B₂O₃ in the oxide slag system.^{24,40,41,42,43,44)} However, the measurement of the activity of B₂O₃ in CaF₂-containing slags has not been reported so far. Xu et al.⁴⁵⁾ calculated the surface tension of liquid CaO–Al₂O₃–MgO–B₂O₃–SiO₂ slag based on the calculated activity of each component in the slag using the IMCT model. Recently, Li et al.⁴⁶⁾ used the same method to calculate the activity of components in the CaO–SiO₂–Al₂O₃–Na₂O–MgO–CaF₂–BaO slag. Therefore, the activity of B₂O₃ in the CaF₂–CaO–Al₂O₃–B₂O₃–SiO₂ slag is also predicted by using the IMCT model at 1823 K in the present study, and the results are listed in Table 2. The calculated activity coefficient of B₂O₃ in the slag at 1823 K is γ_B2O3 = 4.79 × 10⁻⁴ by employing the IMCT model, which is much higher than that calculated by thermodynamic software. Peng et al.¹⁶⁾ reported that the γ_B2O3 = 2.45 × 10⁻⁴ in the CaF₂–CaO–Al₂O₃–MgO–SiO₂–B₂O₃ slag at 1823 K using the agglomerated electron phase model. By comparing the activity coefficient of B₂O₃ calculated by different authors, the calculated γ_B2O3 by the IMCT model and the agglomerated electron phase model have the same order of magnitude. The value γ_B2O3 determined by Peng et al.¹⁶⁾ is slightly lower than the result calculated by the IMCT model, which is likely due to the presence of MgO content reducing the activity coefficient of B₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–SiO₂–B₂O₃ slag. The feasibility of the γ_B2O3 calculated by the IMCT model has been verified by both the laboratory scale experiments and industrial trials. The relative discussion will be proved in the following section.

Table 2. Comparisons between the calculated activity of SiO₂ and B₂O₃ in the CaF₂–CaO–Al₂O₃–B₂O₃–SiO₂ slag by utilizing the IMCT model and the thermodynamic calculation software at 1823 K.

Slag composition (mass%)				Activity of SiO2 and B2O3 in the slag
CaF2	CaO	Al2O3	SiO2+B2O3	$a_{\mathrm{S}\mathrm{i}{\mathrm{O}}_2}^{\mathrm{I}\mathrm{M}\mathrm{C}\mathrm{T}}$	$a_{{\mathrm{B}}_2{\mathrm{O}}_3}^{\mathrm{I}\mathrm{M}\mathrm{C}\mathrm{T}}$	$a_{\mathrm{S}\mathrm{i}{\mathrm{O}}_2}^{\mathrm{F}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{S}\mathrm{a}\mathrm{g}\mathrm{e}}$	$a_{{\mathrm{B}}_2{\mathrm{O}}_3}^{\mathrm{F}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{S}\mathrm{a}\mathrm{g}\mathrm{e}}$
48.0	22.0	25.0	5.0	4.23×10−4	–	3.17×10−4	–
47.9	22.0	25.0	5.1	4.34×10−4	1.24×10−7	3.27×10−4	7.54×10−14
47.9	22.0	24.9	5.2	4.46×10−4	2.58×10−7	3.37×10−4	3.56×10−13
47.9	21.9	24.9	5.3	4.58×10−4	4.02×10−7	3.48×10−4	9.36×10−13
47.8	21.9	24.9	5.4	4.71×10−4	5.59×10−7	3.59×10−4	1.92×10−12
47.8	21.9	24.9	5.5	4.84×10−4	7.28×10−7	3.70×10−4	3.46×10−12
47.7	21.9	24.8	5.6	4.97×10−4	9.10×10−7	3.82×10−4	5.70×10−12
47.7	21.8	24.8	5.7	5.11×10−4	1.11×10−6	3.95×10−4	8.83×10−12
47.6	21.8	24.8	5.8	5.25×10−4	1.32×10−6	4.08×10−4	1.31×10−11
47.6	21.8	24.8	5.9	5.39×10−4	1.54×10−6	4.22×10−4	1.86×10−11
47.5	21.8	24.7	6.0	5.54×10−4	1.79×10−6	4.36×10−4	2.59×10−11

2.3. Activities of Ce₂O₃, La₂O₃ and Al₂O₃ in CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃(–La₂O₃) Slag

Ueda et al.⁴⁷⁾ measured the activity of Al₂O₃ in the CaO–Al₂O₃–Ce₂O₃ slag by using the chemical equilibrium reaction between 1.5 g of Cu and 0.5 g of CaO–Al₂O₃–Ce₂O₃ slag in a graphite crucible under a CO atmosphere for 20 h at 1773 K. Meanwhile, Kitano et al.⁴⁸⁾ used the molten silver as a reference metal to investigate the activities of CaO, Al₂O₃, and Ce₂O₃ in the slag under the similar experimental conditions. An evaluation of activities for Ce₂O₃-containing slags was performed by Wu et al.⁴⁹⁾ by formulating the thermodynamic model based on the IMCT.³¹⁾ They found that the calculated results agreed with the data reported by Ueda et al.,⁴⁷⁾ indicating that the IMCT model can be used to predict the activity of components in Ce₂O₃-containing slags. Similar conclusions were also obtained by Zheng and Liu.⁵⁰⁾ Wang et al.⁵¹⁾ studied the effect of Ce₂O₃ on the vaporization mechanism of fluoride in the CaF₂–CaO–Al₂O₃–Ce₂O₃ slag by the thermogravimetric analysis and the IMCT model, and the results showed that the fluoride vaporization of the slag under vacuum condition at 1823 K was inhibited by Ce₂O₃ addition.

For La₂O₃-containing slags, Wang et al.⁵²⁾ presented data of the slag-metal equilibrium experiments between 8 g of La–Sn alloy and 10 g of CaF₂–Al₂O₃–La₂O₃ slag in a graphite crucible under pure CO gas at 1873 K and further determined the activity of La₂O₃ in the slag. Based on the assumption of the IMCT,³⁰⁾ Wu et al.⁵³⁾ found that 10 kinds of complex molecules can be found in the CaF₂–CaO–Al₂O₃–MgO–La₂O₃ slags, such as CaO·Al₂O₃, CaO·2Al₂O₃, CaO·6Al₂O₃, 3CaO·Al₂O₃, 12CaO·7Al₂O₃, MgO·Al₂O₃, 3CaO·3Al₂O₃·CaF₂, 11CaO·7Al₂O₃·CaF₂, Al₂O₃·La₂O₃, and 11Al₂O₃·La₂O₃. The comparisons between the calculated activities of La₂O₃ by the IMCT model and measured values by Wang et al.⁵²⁾ in the CaF₂–Al₂O₃–La₂O₃ slag at 1873 K are shown in Fig. 3, from which it can be seen that the calculated activity of La₂O₃ in the slag exhibits an almost 1:1 correlation with the measured one within some scatters. Similar behavior has been reported by Zhang et al.⁵⁴⁾ in the CaF₂–CaO–Al₂O₃–Nb₂O₅–La₂O₃–TiO₂–SiO₂–FeO slag. Therefore, the calculated activities of Ce₂O₃, La₂O₃, and Al₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃(–La₂O₃) slags using the IMCT model are also applied in the present study to investigate the oxidation behavior of Ce(La)/Al in alloy melts during the ESR process.

Fig. 3.

Comparisons of calculated activity of La₂O₃ by the IMCT model and determined values by Wang et al.⁵²⁾ in the CaF₂–Al₂O₃–La₂O₃ slag at 1873 K.⁵³⁾ (Online version in color.)

3. Effect of Slag Components in ESR-type Slag on Oxidation Behavior of Reactive Elements

For electroslag remelting of the alloys containing Al and Ti, the previous researchers were inclined to adjust TiO₂ content in the CaF₂–CaO–Al₂O₃–(MgO)–TiO₂ slags to avoid the oxidation loss of the reactive elements. Two main reasons are responsible for this phenomenon: (1) the change of Al content in the molten alloys is sensitive to the content of TiO₂ compared with the other components in the CaF₂–CaO–Al₂O₃–(MgO)–TiO₂ slags; (2) the amount of TiO₂ should be added into the slag increases with increasing temperature, which is the benefit to avoid the oxidation loss of Ti content in the molten alloys during the first temperature-rising (called ramp-up) period of the ESR process. One of the present authors systemically investigated the effect of each component in the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag on the equilibrium Al content for a given Ti content in 0.02C-0.2Si-3.1Mo-18.3Cr-4.9Nb-0.43Al-1.1Ti Ni-based alloy at the temperature range from 1773 K to 1973 K.¹¹⁾ It is interesting from Fig. 4 that the equilibrium Al content in the Ni-based alloy decreases logarithmically with the increase of TiO₂ content in the slag, indicating that the content of TiO₂ in the slag has a significant effect on the Al content in the alloy. This is the reason why the previous researchers frequently use TiO₂ to control the homogenous distribution of Al and Ti content in remelted ingots during the ESR process based on Eq. (3).⁷⁾   

$$\begin{array}{l}4\left[\mathrm{A}\mathrm{l}\right]+3\left(\mathrm{T}\mathrm{i}{\mathrm{O}}_2\right)=2\left(\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3\right)+3\left[\mathrm{T}\mathrm{i}\right],\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\mathrm{\Delta }G°=675\mathrm{\hspace{0.17em} }900-190.3\mathrm{\hspace{0.17em} } T\left(\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}\right)\end{array}$$(3)

Fig. 4.

Effect of TiO₂ on the equilibrium Al content for a given Ti content (1.0 mass%) in nickel-based alloy at the temperature range from 1773 K to 1973 K.⁹⁾ (Online version in color.)

On the other hand, it can be acquired different conclusions about the relationship between the additive amount of TiO₂ and temperature from different views of Fig. 4: (1) in the Y-direction, the equilibrium Al content increases with increasing temperature for a fixed TiO₂ content. After the equilibrium Al content is equal to the initial Al content in the electrode (0.38 mass% for example), i.e., the intersections between the equilibrium Al content line and iso-concentration ([Al] = 0.38 mass%) line, the increment of Al content is caused by the oxidation loss of Ti in the alloy with increasing temperature, which demonstrates that Ti is easily oxidized compared with Al in the alloy when the temperature increases. Similar results have been proved by Hou et al.^13,14) and Yang and Park.³⁵⁾ (2) in the X-direction: the addition of TiO₂ into the slag can be quantificationally determined at the temperature between 1673 K to 1973 K. The relationship between the additive amount of TiO₂ and temperature has been described elsewhere in detail.⁹⁾

The 9CrMoCoB heat resistant steel, i.e., precipitate hardened 9–12 mass% Cr ferritic/martensitic steel has been widely applied in ultra-supercritical power plants in consideration of their low thermal expansion and good thermal conductivity.⁵⁵⁾ However, the strengthening phase M₂₃C₆ (M=Cr or Fe) in the steel would inevitably coarsen during long-term service at high temperatures.^56,57) A valid strategy is an addition of approx. 0.01 mass% B into the steel to enhance the abovementioned mechanical properties by retarding the coarsening of M₂₃C₆ particles and formation of fine grains in the steel.⁵⁸⁾ However, the strong chemical reaction between unstable oxide SiO₂ in slag and B in the steel occurs at the slag-metal interface during the ESR process as shown in Eq. (4),¹⁵⁾ contributing to the uneven distribution of the B and Si in remelted ingots.¹⁶⁾   

$$\begin{array}{l}4\left[\mathrm{B}\right]+3\left(\mathrm{S}\mathrm{i}{\mathrm{O}}_2\right)=2\left({\mathrm{B}}_2{\mathrm{O}}_3\right)+3\left[\mathrm{S}\mathrm{i}\right],\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{0.5em}\mathrm{\Delta }G°=-97\mathrm{\hspace{0.17em} }920-91.2\mathrm{\hspace{0.17em} } T\left(\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}\right)\end{array}$$(4)

Therefore, the effect of each component in the CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag on equilibrium B content in molten 9CrMoCoB steel has been investigated based on the calculated activities of SiO₂ and B₂O₃ as listed in Table 2, and the relationship between (%B₂O₃)/(%SiO₂) ratio in the slag and equilibrium B content in the steel at the temperature range from 1823 K to 1973 K is displayed in Fig. 5.

Fig. 5.

Relationship between equilibrium B content for a given Si content (0.07 mass%) in the 9CrMoCoB steel and (%B₂O₃)/(%SiO₂) ratio in the CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag in the temperature range from 1823 K to 1973 K.⁵⁹⁾ (Online version in color.)

It is noted that the equilibrium B content increases with the increase of (%B₂O₃)/(%SiO₂) irrespective of temperature. For a given (%B₂O₃)/(%SiO₂) ratio, the equilibrium B content slightly increases with the temperature increases, which indicates that the change of the equilibrium B content in the molten steel is less sensitive to the increase of temperature. Unlike the control of Al and Ti content in alloys during the ESR process, the B₂O₃ content in the CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag can be maintained the same content at the different temperatures, and it cannot result in the severe oxidation loss of B and Si in the steel. Noted that the B₂O₃ in liquid slag is likely to evaporate at high temperature. Dai et al.⁴²⁾ measured the activity of B₂O₃ in the CaO–B₂O₃ and MgO–B₂O₃ slags in a semi-sealed graphite container at 1723 K, and they found that the volatilization loss of B₂O₃ in the slag is higher when the equilibrium experiment was conducted in an open graphite crucible than that in a semi-sealed graphite crucible.

Therefore, the electroslag remelting of 9CrMoCoB steel is recommended to carry out in a sealed ESR furnace and B₂O₃ content in the CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag is moderately higher than the designed slag composition, which is to reduce fluctuation of B content in remelted ingot caused by the evaporation of B₂O₃ in the CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag. The thermodynamic calculations shown in Fig. 5 have been proved by slag-metal equilibrium experiments between CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag and 9CrMoCoB steel at 1823 K,⁵⁹⁾ and it has been successfully applied in industrial tests. From the calculated results, the equilibrium B content in the liquid steel is about 24 to 33 ppm when Si content is in the range of 0.06 to 0.07 mass%, which is in good agreement with the determined values as shown in Fig. 6. These results imply that the activity of B₂O₃ in the 48CaF₂-22CaO-25Al₂O₃-xSiO₂-yB₂O₃ slag (mass%, x+y = 0–6 mass%) by the IMCT model is correct and the IMCT model can be reliably applied to calculate the activity of B₂O₃ in the CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag, which can be used to design slag composition to control Si and B content in 9CrMoCoB steel during the ESR process.

Fig. 6.

(a) Picture of remelted ESR ingot, and (b) Si and B content along with the radial direction of the remelted ingot.⁵⁹⁾

Rare earth elements, such as Ce and La, are added into steels and alloys to improve fatigue and creep properties as well as depress temper embrittlement.⁶⁰⁾ Al₂O₃ in the CaF₂–CaO–Al₂O₃ slag is easily reduced by Ce in steel during the ESR process according to the chemical reaction in Eq. (5).⁶¹⁾   

$$\begin{array}{l}2\left[\mathrm{C}\mathrm{e}\right]+\left(\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3\right)=\left(\mathrm{C}{\mathrm{e}}_2{\mathrm{O}}_3\right)+2\left[\mathrm{A}\mathrm{l}\right],\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\mathrm{\Delta }G°=-109\mathrm{\hspace{0.17em} }330-11.1\mathrm{\hspace{0.17em} } T\left(\text{J/mol}\right)\end{array}$$(5)

Therefore, it is very important to understand the reaction mechanism between Al₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag and Ce in the alloy at various slag compositions and temperatures from viewpoint of thermodynamics. Baligidad and Radhakrishna investigated the effect of process parameters on the recovery of Ce in Fe-10.5Al-0.8C-(0.1, 0.3)Ce (mass%) alloys during the ESR process and found that the high recovery of Ce was achieved by using Ca as a deoxidizer during the ESR.⁶²⁾ Weng et al.¹⁷⁾ studied the control technique of Ce content in 1Cr17 ferritic stainless steel in a 1 tonne ESR furnace.

In the present work, the chemical compositions of the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag and 1Cr17 steel from Ref. 17) were used to testify the feasibility of thermodynamic analysis of the effect of each component in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag on equilibrium Ce content in the liquid steel at different temperatures. From Eq. (5), the relationship between the equilibrium Ce content and slag composition, alloy composition, as well as temperature can be acquired by simple mathematical derivation as shown in Eq. (6).   

$$\begin{array}{l}log\left[\%\mathrm{C}\mathrm{e}\right]=\\ \frac{1}{2}\left(log\frac{a_{\mathrm{C}{\mathrm{e}}_2{\mathrm{O}}_3}}{a_{\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3}}+2log f_{\mathrm{A}\mathrm{l}}+2log\left[\%\mathrm{A}\mathrm{l}\right]-2log f_{\mathrm{C}\mathrm{e}}-\left(\frac{5\mathrm{\hspace{0.17em} }710}{T}+0.58\right)\right)\end{array}$$(6)

where f_i is the activity coefficient of component i in the metal phase in reference to the 1 mass% standard state, a_i is the activity of component i in slag phase with reference to the Raoultian standard state. T is the absolute temperature (K). The activities of Al₂O₃ and Ce₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag are computed by the IMCT model; The activity coefficients of Al and Ce in the liquid steel are calculated by Wagner equation⁶³⁾ and the interaction parameters listed in Table 3.^64,65) The slag-metal reaction temperature at the tip of the electrode does not superheat more than 20 K to 30 K,⁶⁶⁾ compared with the liquidus temperature of 1Cr17 steel (1713 K), which was calculated by the thermodynamic calculation software. Some researchers have experimentally and numerically determined the temperature profile of slag bath and metal pool.^67,68,69) Thus, the reaction temperature at the interface between slag and metal pool can be taken as 1973 K based on the previous studies. Comparisons between the calculated and measured Ce content in remelted 1Cr17 steel ingots by Weng et al.¹⁷⁾ are listed in Table 4, from which it can be observed that the calculated equilibrium Ce contents for a fixed Al content (0.04 mass%) are consistent with the measured results in remelted 1Cr17 steel ingots using various slag compositions, indicating that the IMCT model can be used to calculate the activities of Al₂O₃ and Ce₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag.

Table 3. Interaction coefficient used in the present study.⁶⁴⁾

$e_i^j$	C	Si	Cr	Al	S	O	Ce
Al	0.091	0.0056	0.009665)	0.045	0.03	−6.6	−0.043
Ce	−0.077	–	–	−2.25	−39.8	−5.03	−0.003

Table 4. Comparisons between the measured and calculated Ce content for a given Al content (0.04 mass%) in remelted 1Cr17 steel ingots.

Slag composition (mass%)						Ce content in steel (mass%)
Exp.	CaO	CaF2	Al2O3	MgO	Ce2O3	Measured17)	Calculated
E1	8	66	16	10	–	–	–
E2	8	60	12	5	15	0.05	0.049
E3	8	48	12	5	27	0.13	0.124

The relationships between equilibrium content of Ce for a given Al content (0.04 mass%) in the steel and each component in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag at the temperature range from 1773 K to 1973 K are shown in Fig. 7. It is clear that the equilibrium Ce content in the molten steel increases with increasing CaO and Ce₂O₃ content, whereas decreases with the increase of Al₂O₃ content in the slag at different temperatures. Ren et al.²⁰⁾ investigated the change of CaF₂ and MgO content in the slag has little influence on the equilibrium Ce content in the liquid steel, which results from the activity of Al₂O₃ and Ce₂O₃ is not sensitive to the change of CaF₂ and MgO content in the CaF₂–CaO–Al₂O₃–MgO– Ce₂O₃ slag. Similar results have also been reported by other researhers.^11,25,70)

Fig. 7.

Relationship between equilibrium content of Ce for a given Al content (0.04 mass%) in the steel and (a) CaF₂, (b) CaO, (c) Al₂O₃, (d) MgO, and (e) Ce₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag in at the temperature range from 1773 K to 1973 K. (Online version in color.)

Meanwhile, it can be noted that the effect of temperature on the equilibrium Ce content can be nearly negligible, which is similar to the control of Si and B content in 9CrMoCoB steel as mentioned above. It can be obtained from Fig. 7 that the relative order of importance of the components in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag for reducing the oxidation loss of Ce in molten 1Cr17 steel is ‘CaO ≈ Al₂O₃ > Ce₂O₃ > MgO > CaF₂.’ That is to say, the adjustment of CaO (or Al₂O₃) and Ce₂O₃ content in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag is an effective method to control the Ce content in 1Cr17 steel during the ESR process. The equilibrium Ce content shows a positive relationship with Al content in the molten alloy as shown in Eq. (6), therefore, an appropriate increase of Al content in the electrode is in favor of reduction of oxidation loss of Ce in 1Cr17 steel.

The laboratory-scale experiment between CaF₂–CaO–Al₂O₃–Ce₂O₃ slag and Ni-20%Cr-Ce alloy in a 100 kg ESR furnace was carried out by Ren et al.,²⁰⁾ who found the same tendency that the yield of Ce in the remelted Ni–Cr–Ce ingot increased with the increase of CaO and Ce₂O₃ content in the CaF₂–CaO–Ce₂O₃ slag as shown in Fig. 8(a), whereas the presence of Al₂O₃ in the slag would tend to oxidize the rare earth element. Comparing with CaF₂–CaO–Al₂O₃–Ce₂O₃ and CaF₂–CaO–Al₂O₃–La₂O₃ slags,^51,53) there are 7 kinds of identical complex molecules and the form of the complex molecules containing rare earth oxide is similar, viz., Ce₂O₃·Al₂O₃, Ce₂O₃·11Al₂O₃, La₂O₃·Al₂O₃, and La₂O₃·11Al₂O₃, according to the assumption of the IMCT.³⁰⁾ Therefore, it is reasonable to consider that the oxidation behavior of La in the alloy by the CaF₂–CaO–Al₂O₃–La₂O₃ slag is similar to that of Ce by the CaF₂–CaO–Al₂O₃–Ce₂O₃ slag.

Fig. 8.

Relationship between (a) Ce₂O₃ and (b) La₂O₃ content in the CaF₂–CaO–Al₂O₃–Ce₂O₃–(La₂O₃) slag and yield of Ce and La in remelted ingots, respectively.²⁰⁾ (Online version in color.)

On the other hand, Gordy and Thomas determined the optical basicity of Ce₂O₃ and CaO as 1.1 and 1.0, respectively, which means that Ce₂O₃ and CaO have a similar influence on the control of Ce/Al content and improvement of cleanness of alloys.⁷¹⁾ Yang et al.⁷²⁾ investigated the effect of CaO–Al₂O₃–MgO–SiO₂–Ce₂O₃ slag containing different Ce₂O₃ contents on the cleanness of the Al-killed steel at 1873 K, and they found that the total oxygen content of the steel decreased from 100 ppm to 25 ppm after the addition of 10 mass% Ce₂O₃. Gao et al.⁷³⁾ reported a similar trend. Long et al.⁷⁴⁾ also pointed out that the 46CaO-39Al₂O₃-10SiO₂-5Ce₂O₃ (mass%) slag has a low viscosity that is beneficial to application in the practical refining process.

The value of optical basicity of La₂O₃ (1.07) is almost the same as the Ce₂O₃,⁷⁵⁾ and thus a similar mothed can be used to control La/Al content in the ingot during the ESR process, viz., increasing CaO or La₂O₃ content in the CaF₂–CaO–Al₂O₃–La₂O₃ slag. Ren et al.²⁰⁾ also studied the yield of La during electroslag remelting of Fe-5Al-25Cr-0.036La-0.0008O (mass%) alloy using the CaF₂–CaO–Al₂O₃–La₂O₃ slag, and the results are illustrated in Fig. 8(b). As expected, the yield of La in the Fe–Al–Cr–La–O alloy increases with increasing CaO and La₂O₃ content in the slag, while they also pointed out that it is hardly improved when the addition of CaO content is up to 15 mass%. The GH188 cobalt-based superalloy was utilized as the consumable electrode and remelted using three different slag compositions. The alternate current and secondary voltage applied during the normal refining were 2400 to 3400 A and 47 V, respectively. Based on the experimental results, Niu et al.¹⁹⁾ found that the oxidation loss of La in GH188 ingot was mainly caused by oxygen potential in slag bath, and it can be prevented when Al was used as a deoxidizer. Additionally, La in the alloy can react with Al₂O₃ in the CaF₂–CaO–Al₂O₃ based slag according to reaction in Eq. (7),¹⁸⁾ resulting in the loss of the reactive element during the ESR process.   

$$\begin{array}{l}2\left[\mathrm{L}\mathrm{a}\right]+\left(\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3\right)=\left(\mathrm{L}{\mathrm{a}}_2{\mathrm{O}}_3\right)+2\left[\mathrm{A}\mathrm{l}\right],\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\mathrm{\Delta }G°=-243\mathrm{\hspace{0.17em} }320-49.6\mathrm{\hspace{0.17em} } T\left(\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}\right)\end{array}$$(7)

4. Effect of CaO on Oxidation Behaviour of Reactive Elements

The CaO in the CaF₂–CaO–Al₂O₃ based slag is an important component to affect the cleanness of remelted ingots, such as deoxidation, desulfurization, and removal of non-metallic inclusions, after the ESR process.^{1,3,4,76,77,78,79)} From the early discussion, the oxidation loss of Al/Ti, Si/B, and Ce/Al in steels and alloys can be effectively prevented by adjusting TiO₂, B₂O₃, and CaO (or Al₂O₃) in the CaF₂–CaO–Al₂O₃ based slags during the ESR process. CaO has a significant influence on the oxidation behavior of the aforementioned reactive elements by altering the activities TiO₂, B₂O₃, and Al₂O₃ in the slag. Meanwhile, the oxidation behavior of the preceding various elements is different even merely the CaO content is changed in the slag during the ESR process, which should be elucidated in the viewpoint of thermodynamics.

The slag-metal experiments between CaF₂–CaO–Al₂O₃–MgO–TiO₂–SiO₂ slag (100 g) and 1Cr21Ni5Ti steel (650 g) were conducted in a magnesia crucible at 1823 K to 1873 K, and the metal samples were taken by Hou et al.⁸⁰⁾ at different time intervals (0, 10, 20, 30, and 40 min) after addition of the slag to study the effect of CaO content on the loss of Al and Ti in the steel. According to the results, they found that the slag with low CaO content at low slag temperature was suitable for electroslag remelting of 1Cr21Ni5Ti steel. Hou et al.^10,81) also suggested that continuous addition of TiO₂ into the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag with high CaO content not only control the homogeneity of Al and Ti in remelted 1Cr21Ni5Ti (GH8825) ingot during the temperature-rising (ramp-up) period of ESR process but also improve the cleanness of 1Cr21Ni5Ti steel. From the abovementioned theoretical and experimental results proposed by Hou et al.,^10,80,81) they investigated the effect of two kinds of slag composition, such as low and high CaO content in the CaF₂–CaO–Al₂O₃–MgO–TiO₂(–SiO₂) slags, on the oxidation behavior of Al and Ti in 1Cr21Ni5Ti steel. It can be obtained that the corresponding method for the controlling homogenous distribution of Al and Ti in the 1Cr21Ni5Ti ingot is different for the slag containing different CaO contents. In the case of low CaO content in the slag, the oxidation loss of Al and Ti can be controlled only by changing slag temperature, while TiO₂ should be continuously added in case of the slag containing high CaO.

The relationship between the equilibrium amount of Al in the molten 0.02C-0.2Si-3.1Mo-18.3Cr-4.9Nb-0.43Al-1.1Ti Ni-based alloy at a given Ti content (1.0 mass%) and CaO concentration in the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag at a temperature range of 1773 K to 1973 K has been investigated by the present authors. It can be observed from Fig. 9(a) that the intersection point between the equilibrium Al content line and iso-concentration line ([Al]=0.38 mass%) will be shifted to the left-hand side when temperature increases that are exactly opposite tendency compared with the results shown in Fig. 4(b), indicating that CaO content in the slag should be dropped when slag temperature goes up, in other words, change of TiO₂ content in the slag is an effective approach to control oxidation loss of Al and Ti content in the alloy, especially for the temperature-rising (ramp-up) period during the ESR process. It can also be seen from Fig. 9(a) that the equilibrium Al content increases with an increase of CaO content regardless of temperature, giving rise to the fact that the homogeneity of the reactive element Al and Ti in the alloy can be conveniently controlled by adjusting slag temperature when the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag containing low CaO content (e.g., CaO = 0 mass%) according to the above discussion. However, when the slag contains high CaO content (e.g., CaO = 35 mass%), the equilibrium Al content is higher than the initial Al content 0.38 mass%, resulting in the oxidation of Ti by Eq. (3). After the addition of 15% TiO₂ into the 35% CaO-containing slags as presented in Fig. 9(c), the equilibrium Al content is equal to the initial Al content at 1973 K, suggesting that both Al and Ti are not susceptible to the oxidation loss during the ESR process.

Fig. 9.

Relationship between the equilibrium amount of Al in the molten nickel-based alloy at a given Ti concentration ([Ti]=1.0 mass%) and CaO concentration in the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag over a temperature range of 1773 K to 1973 K: (a) 5 mass% TiO₂, (b) 10 mass% TiO₂, (c) 15 mass% TiO₂. (d) Relationship between the equilibrium content of CaO in the slag and temperature.⁸²⁾ (Online version in color.)

Therefore, it can be obtained that the thermodynamic analysis results shown in Fig. 9 can be used to explain why the two kinds of methods adopted by Hou et al.^10,80,81) for controlling Al and Ti content during electroslag remelting of 1Cr21Ni5Ti steel when the CaF₂–CaO–Al₂O₃–MgO–TiO₂(–SiO₂) slag includes low and high CaO content. From the results reported by Duan et al.,^9,11) Jiang et al.,¹⁰⁾ Yang and Park,³⁵⁾ and Hou et al.,^13,80,81) it can be noted that the oxidation behavior of Al/Ti in molten Ni- and Fe-based alloys by the CaF₂–CaO–Al₂O₃–MgO–TiO₂–SiO₂ slag is the same at different temperature even though the alloy grade is different. Consequently, it is believed that the oxidation behavior of Al/Ti is strongly dependent on the different components in slag during the electroslag remelting of the alloys containing Al and Ti.

Al₂O₃ is an indispensable component in ESR slag that usually contains 20 to 30 mass%, which can increase the generated heat by dropping the electrical conductivity of the slag because the liquid slag acts as a resistive medium for the conversion of electric energy into thermal energy.^83,84,85,86) The effect of Al₂O₃ on the equilibrium Al content in the molten 0.02C-0.2Si-3.1Mo-18.3Cr-4.9Nb-0.43Al-1.1Ti Ni-based alloy at the temperature range from 1773 K to 1973 K has been investigated in the present authors’ previous study.¹¹⁾ The influence of CaO and Al₂O₃ on the oxidation behavior of Al/Ti in the nickel-based alloy melts is nearly the same. However, the previous researchers prefer changing CaO content rather than Al₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–TiO₂ slag since high CaO content combined with TiO₂ addition in the slag can minimize the use of CaF₂ content, which is not only to promote the homogeneity of Al and Ti and cleanness of ingots,⁸⁷⁾ but also to reduce the specific energy consumption of the ESR process.^88,89,90)

In the previous section, Fig. 7(b) shows the equilibrium Ce content in molten 1Cr17 steel and CaO content in the CaF₂–CaO–Al₂O₃–Ce₂O₃ slag at the temperature range from 1773 K to 1973 K. The effect of CaO on the activities of Al₂O₃, Ce₂O₃, and Al₂O₃·Ce₂O₃ at 1873 K calculated by IMCT model is shown in Fig. 10. Given the hypothesis of the IMCT,³¹⁾ the Ce₂O₃ can react with Al₂O₃ to produce the complex molecules Al₂O₃·Ce₂O₃ and 11Al₂O₃·Ce₂O₃ in the slag melts. The activities of Al₂O₃·Ce₂O₃ and Al₂O₃ decrease, while the activity of Ce₂O₃ increases with the addition of CaO in the slag. Hence, the equilibrium Ce content in the molten steel increases with increasing CaO content in the slag as shown in Fig. 7(b). Similar results have been reported by Wang et al.⁹¹⁾ Meanwhile, Wang and Zou found that activity of La₂O₃ increased with increasing CaO/SiO₂ ratio in the CaF₂–CaO–SiO₂–La₂O₃ slag at 1773 K.⁹²⁾

Fig. 10.

Relationship between calculated activities of Al₂O₃, Ce₂O₃ as well as Al₂O₃·Ce₂O₃ and CaO content in the CaF₂–CaO–Al₂O₃–Ce₂O₃ at 1873 K. (Online version in color.)

However, the equilibrium B content decreases with the CaO content from 0 to 10 mass% in the CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag at a constant temperature, while the equilibrium B content stays approximately the same level as the CaO content changes from 15 to 30 mass% at any given temperature as illustrated in Fig. 11. In the case of the slag with low CaO content, the equilibrium B content is found to increase with an increase in temperature for a constant slag composition, indicating that Si is easily prone to oxidation than B in the steel when temperature increases, which can be restrained by adding SiO₂ into the slag. Another efficient method is that the CaO content is controlled in the range of 15 to 30 mass% in the CaF₂–CaO–Al₂O₃–SiO₂–B₂O₃ slag because the equilibrium content of B equals the initial B content (49 ppm) at the remelting temperature of 1973 K.

Fig. 11.

Relationship between the equilibrium amount of B for a given Si content (0.07 mass%) in the molten 9CrMoCoB steel at a temperature range from 1823 K to 1973 K.⁵⁹⁾ (Online version in color.)

5. Conclusions

The activities of components in fluoride-containing slags are calculated by the ion and molecule coexistence theory (IMCT) and the thermodynamic calculation software and further compared. Additionally, the oxidation behavior of various reactive elements, such as Al/Ti, B/Si, and rare earth metals (REM) such as Ce and La, by ESR type slag has been investigated from the perspective of thermodynamics based on the calculated activity in the fluoride-containing slags by the IMCT model. The main findings of the present work can be summarized as follows:

(1) The accuracy of the calculated activity of TiO₂ in the CaF₂–CaO–Al₂O₃–TiO₂(–MgO) slags and that of Ce₂O₃ in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃ slag by the IMCT model has been proved by the slag-metal experimental results. Meanwhile, although the calculated activity of B₂O₃ in the CaF₂–CaO–Al₂O₃–B₂O₃–SiO₂ slag by the IMCT model is higher than that by the thermodynamic calculation software at 1823 K due to the experimental measurement of the activity of B₂O₃ in the slag is rarely reported, the calculated γ_B2O3 = 4.79 × 10⁻⁴ by the IMCT model at 1823 K is also testified by the slag-metal experimental results. These results imply that the IMCT model can be reliably applied to calculate the activity of components in ESR type slag.

(2) The oxidation of the reactive elements Al/Ti, B/Si, and La(Ce)/Al in alloy melts can be effectively prevented by adding TiO₂, B₂O₃, and CaO in the CaF₂–CaO–Al₂O₃ based slags during the ESR process, while the reactive elements in different types of alloy melts lead to the activity differences, which only affects the equilibrium content of reactive elements in the alloy melts for a given slag composition and temperature.

(3) Both CaO and Al₂O₃ are key components in ESR type slags, however, the oxidation behavior of Al/Ti, B/Si, and La(Ce)/Al in alloy melts by changing CaO and Al₂O₃ content in the slag is different. The different methods should be adopted to control the homogeneity of Al and Ti in ingots when the slag contains high and low CaO content, respectively. In the case of the slag with high CaO content, the extra addition of TiO₂ should be needed, whereas adjustment of slag temperature is enough in the case of low CaO content slag during the ESR process. The CaF₂–CaO–Al₂O₃–B₂O₃–SiO₂ with high CaO content is suitable for electroslag remelting of the alloys containing B due to the equilibrium contents of Si and B being slightly affected by the remelting temperature. The increase of Al₂O₃ content in the CaF₂–CaO–Al₂O₃–MgO–Ce₂O₃/La₂O₃ slags can deteriorate the uniform distribution of Ce(La) and Al in ingots during remelting the alloys containing Ce(La), which can be restrained by the addition of Ce₂O₃ (or La₂O₃) into the slag. However, the addition amount of Ce₂O₃ (or La₂O₃) should be higher than that of CaO.

Acknowledgments

This work was supported by the research fund of Hanyang University, Korea (HY-2021). Also, the authors express their appreciation to Dr. Dong Soo KIM, Doosan Heavy Industries & Construction, Korea, for a fruitful discussion regarding the commercial ESR process.

References

• 1)   Z. B.  Li: Electroslag Metallurgy Theory and Practice, Metallurgical Industry Press, Beijing, (2010), 79.
• 2)   Z. H.  Jiang: Electroslag Metallurgy, Science Press, Beijing, (2015), 3.
• 3)   A.  Mitchell: J. Vac. Sci. Technol., 7 (1970), S63. https://doi.org/10.1116/1.1315924
• 4)   F.  Reyes-Carmona and  A.  Mitchell: ISIJ Int., 32 (1992), 529. https://doi.org/10.2355/isijinternational.32.529
• 5)   S. C.  Duan,  H. J.  Guo,  X.  Shi,  J.  Guo,  B.  Li,  S. W.  Han and  W. S.  Yang: Chin. J. Eng., 40 (2018), 53. https://doi.org/10.13374/j.issn2095-9389.2018.s1.009
• 6)   S. F.  Medina and  A.  Cores: ISIJ Int., 33 (1993), 1244. https://doi.org/10.2355/isijinternational.33.1244
• 7)   G.  Pateisky,  H.  Biele and  H. J.  Fleischer: J. Vac. Sci. Technol., 9 (1972), 1318. https://doi.org/10.1116/1.1317029
• 8)   C. X.  Chen,  R. F.  Gao and  W. X.  Zhao: Acta Metall. Sin., 20 (1984), 243. https://www.ams.org.cn/CN/Y1984/V20/I3/243
• 9)   S. C.  Duan,  X.  Shi,  F.  Wang,  M. C.  Zhang,  Y.  Sun,  H. J.  Guo and  J.  Guo: Metall. Mater. Trans. B, 50 (2019), 3055. https://doi.org/10.1007/s11663-019-01665-2
• 10)   Z. H.  Jiang,  D.  Hou,  Y. W.  Dong,  Y. L.  Cao,  H. B.  Cao and  W.  Gong: Metall. Mater. Trans. B, 47 (2016), 1465. https://doi.org/10.1007/s11663-015-0530-8
• 11)   S.  Duan,  X.  Shi,  M.  Mao,  W.  Yang,  S.  Han,  H.  Guo and  J.  Guo: Sci. Rep., 8 (2018), 5232. https://doi.org/10.1038/s41598-018-23556-3
• 12)   D.  Hou,  Z. H.  Jiang,  Y. W.  Dong,  W.  Gong,  Y. L.  Cao and  H. B.  Cao: ISIJ Int., 57 (2017), 1400. https://doi.org/10.2355/isijinternational.ISIJINT-2017-147
• 13)   D.  Hou,  Z. H.  Jiang,  Y. W.  Dong,  W.  Gong,  Y. L.  Cao and  H. B.  Cao: ISIJ Int., 57 (2017), 1410. https://doi.org/10.2355/isijinternational.ISIJINT-2017-148
• 14)   D.  Hou,  F. B.  Liu,  T. P.  Qu,  Z. H.  Jiang,  D. Y.  Wang and  Y. W.  Dong: ISIJ Int., 58 (2018), 876. https://doi.org/10.2355/isijinternational.ISIJINT-2017-687
• 15)   D. S.  Kim,  G. J.  Lee,  M. B.  Lee,  J. I.  Hur and  J. W.  Lee: Int. Symp. on Liquid Metal Processing & Casting 2015 (LMPC2015), IOP Publishing, Leoben, (2016), 12002. https://doi.org/10.1088/1757-899x/143/1/012002
• 16)   L.  Peng,  Z.  Jiang and  X.  Geng: Calphad, 70 (2020), 101782. https://doi.org/10.1016/j.calphad.2020.101782
• 17)   J.  Weng,  W.  Shi and  H.  Xie: Metalurgija, 57 (2018), 67. https://hrcak.srce.hr/189366
• 18)   T.  Wen,  H.  Zhang,  X.  Li,  L.  Yu,  Y.  Ren,  H.  Liu and  L.  Zhang: JOM, 70 (2018), 2157. https://doi.org/10.1007/s11837-018-3049-z
• 19)   Y. J.  Niu,  H. W.  Liu,  Z. W.  Zhang,  N.  An,  Y.  Gao and  J. J.  Tian: Chinese Material Conference 2017, Springer Singapore, Singapore, (2018), 731. https://doi.org/10.1007/978-981-13-0104-9_78
• 20)   Y.  Ren,  L.  Zhang,  L.  Yu and  W.  Fang: Metall. Res. Technol., 113 (2016), 405. https://doi.org/10.1051/metal/2016017
• 21)   M.  Allibert and  C.  Chatillon: Can. Metall. Q., 18 (1979), 349. https://doi.org/10.1179/cmq.1979.18.3.349
• 22)   K. C.  Mills,  J. S.  Powell,  J. W.  Bryant and  B. J.  Keene: Can. Metall. Q., 20 (1981), 93. https://doi.org/10.1179/cmq.1981.20.1.93
• 23)   M.  Iwase,  H.  Akizuki,  E.  Ichise and  Y.  Tanaka: Steel Res., 57 (1986), 436. https://doi.org/10.1002/srin.198600800
• 24)   I. D.  Sommerville and  D. A. R.  Kay: Metall. Trans., 2 (1971), 1727. https://doi.org/10.1007/BF02913898
• 25)   A. I.  Zaitsev,  N. V.  Korolyov and  B. M.  Mogutnov: J. Chem. Thermodyn., 22 (1990), 531. https://doi.org/10.1016/0021-9614(90)90145-G
• 26)   A. I.  Zaitsev,  A. D.  Litvina and  B. M.  Mogutnov: J. Chem. Thermodyn., 24 (1992), 1039. https://doi.org/10.1016/S0021-9614(05)80015-1
• 27)   J. H.  Park and  D. J.  Min: Steel Res. Int., 75 (2004), 807. https://doi.org/10.1002/srin.200405846
• 28)   Y.  Liu,  Z.  Zhang,  G.  Li,  Y.  Wu,  D.  Xu and  B.  Li: Vacuum, 158 (2018), 6. https://doi.org/10.1016/j.vacuum.2018.09.027
• 29)   M.  Shinmei and  T.  Machida: Metall. Trans., 4 (1973), 1996. https://doi.org/10.1007/BF02665432
• 30)   J.  Zhang and  P.  Wang: Calphad, 25 (2001), 343. https://doi.org/10.1016/S0364-5916(01)00054-2
• 31)   J.  Zhang: Computational Thermodynamics of Metallurgical Melts and Solutions, Metallurgical Industry Press, Beijing, (2007), 241.
• 32)   S. C.  Duan,  C.  Li,  X. L.  Guo,  H. J.  Guo,  J.  Guo and  W. S.  Yang: Ironmaking Steelmaking, 45 (2018), 655. https://doi.org/10.1080/03019233.2017.1318547
• 33)   X.  Yang,  C.  Shi,  M.  Zhang and  J.  Zhang: Steel Res. Int., 83 (2012), 244. https://doi.org/10.1002/srin.201100233
• 34)   I. H.  Jung and  M. A.  Van Ende: Metall. Mater. Trans. B, 51 (2020), 1851. https://doi.org/10.1007/s11663-020-01908-7
• 35)   J. G.  Yang and  J. H.  Park: Metall. Mater. Trans. B, 48 (2017), 2147. https://doi.org/10.1007/s11663-017-0994-9
• 36)   J. T.  Ju,  K. S.  Yang,  G. H.  Ji, J. L. An and S. W. Liu: Chin. J. Eng., 42 (2020), 119. https://doi.org/10.13374/j.issn2095-9389.2020.03.07.s01
• 37)   J.  Ju,  Z.  Zhu,  K.  Yang,  G.  Ji, J. An and C. Shi: Rare Met. Mater. Eng., 50 (2021), 3550. http://www.rmme.ac.cn/rmme/article/html/20210168
• 38)   I.  Isaksson and  K.  Oberg: Scand. J. Metall., 2 (1973), 1.
• 39)   Y. S.  Hsu,  Y. H.  Chou and  F. L.  Lui: Acta Metall. Sin., 7 (1964), 24. https://www.ams.org.cn/CN/Y1964/V7/I1/24
• 40)   W. J.  Knapp and  H.  Flood: J. Am. Ceram. Soc., 40 (1957), 246. https://doi.org/10.1111/j.1151-2916.1957.tb12612.x
• 41)   M.  Boike,  K.  Hilpert and  F.  Müller: J. Am. Ceram. Soc., 76 (1993), 2809. https://doi.org/10.1111/j.1151-2916.1993.tb04020.x
• 42)   C.  Dai,  X.  Zhang and  L.  Shui: Metall. Mater. Trans. B, 26 (1995), 651. https://doi.org/10.1007/BF02653887
• 43)   Z.  Huang,  Z.  Yang,  Y.  Su,  S.  Tong and  Z.  Wang: J. Chem. Thermodyn., 27 (1995), 1429. https://doi.org/10.1006/jcht.1995.0150
• 44)   X.  Huang,  K.  Asano,  T.  Fujisawa,  Z.  Sui and  C.  Yamauchi: ISIJ Int., 36 (1996), 1360. https://doi.org/10.2355/isijinternational.36.1360
• 45)   J. F.  Xu,  J. C.  Li,  W. P.  Weng,  M. Q.  Sheng,  D.  Chen,  J. Y.  Zhang and  Y.  Chen: Chin. J. Nonferrous Met., 27 (2017), 206. https://doi.org/10.19476/j.ysxb.1004.0609.2017.01.026
• 46)   B.  Li,  X.  Geng,  Z.  Jiang,  Y.  Hou and  W.  Gong: Metals, 11 (2021), 165. https://doi.org/10.3390/met11010165
• 47)   S.  Ueda,  K.  Morita and  N.  Sano: ISIJ Int., 38 (1998), 1292. https://doi.org/10.2355/isijinternational.38.1292
• 48)   R.  Kitano,  M.  Ishii,  M.  Uo and  K.  Morita: ISIJ Int., 56 (2016), 1893. https://doi.org/10.2355/isijinternational.ISIJINT-2016-201
• 49)   C.  Wu,  G.  Cheng,  H.  Long and  X.  Yang: High Temp. Mater. Process., 32 (2013), 207. https://doi.org/10.1515/htmp-2012-0119
• 50)   X.  Zheng and  C.  Liu: Metall. Mater. Trans. B, 52 (2021), 3183. https://doi.org/10.1007/s11663-021-02245-z
• 51)   X.  Wang,  Y.  Liu,  G.  Li,  Q.  Wang,  T.  Li and  Y.  Cao: Vacuum, 185 (2021), 109997. https://doi.org/10.1016/j.vacuum.2020.109997
• 52)   C. Z.  Wang,  S. Q.  Ye,  D. Y.  Yu and  W. Q.  Guo: Acta Metall. Sin., 20 (1984), 357. https://www.ams.org.cn/CN/Y1984/V20/I5/357
• 53)   C. C.  Wu and  G. G.  Cheng: J. Iron Steel Res., 25 (2013), 19. https://doi.org/10.13228/j.boyuan.issn1001-0963.2013.09.011
• 54)   B.  Zhang,  M. F.  Jiang,  J.  Qi and  C. J.  Liu: J. Northeast. Univ. (Nat. Sci.), 32 (2011), 524. http://xuebao.neu.edu.cn/natural/EN/volumn/volumn_1171.shtml
• 55)   C.  Jia,  Y.  Liu,  C.  Liu,  C.  Li and  H.  Li: Mater. Charact., 152 (2019), 12. https://doi.org/10.1016/j.matchar.2019.04.006
• 56)   K.  Hamada,  K.  Tokuno,  Y.  Tomita,  H.  Mabuchi and  K.  Okamoto: ISIJ Int., 35 (1995), 86. https://doi.org/10.2355/isijinternational.35.86
• 57)   K.  Guguloth and  N.  Roy: Mater. Charact., 146 (2018), 279. https://doi.org/10.1016/j.matchar.2018.10.011
• 58)   J.  Kasl and  D.  Jandová: Mater. Sci. Forum, 782 (2014), 179. https://doi.org/10.4028/www.scientific.net/MSF.782.179
• 59)   S. C.  Duan,  M. J.  Lee,  D. S.  Kim and  J. H.  Park: J. Mater. Res. Technol., 17 (2022), 574. https://doi.org/10.1016/j.jmrt.2022.01.033
• 60)   J.  Yu,  Z.  Yu and  C.  Wu: JOM, 40 (1988), 26. https://doi.org/10.1007/BF03258907
• 61)   A.  Karasev and  H.  Suito: Metall. Mater. Trans. B, 30 (1999), 249. https://doi.org/10.1007/s11663-999-0054-1
• 62)   R. G.  Baligidad and  A.  Radhakrishna: J. Mater. Sci., 37 (2002), 5021. https://doi.org/10.1023/A:1021035615317
• 63)   C.  Wagner: Thermodynamics of Alloys, Addison-Wesley Press, Cambridge, MA, (1952), 47.
• 64)   Y. Q.  Liu,  L. J.  Wang,  J. B.  Guo,  X. J.  Hu and  K. C.  Chou: Chin. J. Nonferrous Met., 23 (2013), 720. https://doi.org/10.19476/j.ysxb.1004.0609.2013.03.018
• 65)   M.  Hino and  K.  Ito: Thermodynamic Data for Steelmaking, Tohoku University Press, Sendai, (2010), 259.
• 66)   M. E.  Fraser and  A.  Mitchell: Ironmaking Steelmaking, 3 (1976), 279.
• 67)   A.  Mitchell and  S.  Joshi: Metall. Trans., 4 (1973), 631. https://doi.org/10.1007/BF02643068
• 68)   M.  Kawakami,  K.  Nagata,  M.  Yamamura,  N.  Sakata,  Y.  Miyashita and  K.  Goto: Tetsu-to-Hagané, 63 (1977), 2162 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.63.13_2162
• 69)   K.  Narita: Tetsu-to-Hagané, 63 (1977), 1996 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.63.13_1996
• 70)   A. I.  Zaitsev,  N. V.  Korolyov and  B. M.  Mogutnov: J. Chem. Thermodyn., 22 (1990), 513. https://doi.org/10.1016/0021-9614(90)90144-F
• 71)   W.  Gordy and  W. J. O.  Thomas: J. Chem. Phys., 24 (1956), 439. https://doi.org/10.1063/1.1742493
• 72)   X.  Yang,  H.  Long,  G.  Cheng,  C.  Wu and  B.  Wu: J. Rare Earths, 29 (2011), 1079. https://doi.org/10.1016/S1002-0721(10)60602-3
• 73)   X.  Gao,  L.  Zhang,  X.  Qu,  Y.  Luan and  X.  Chen: Metall. Res. Technol., 117 (2020), 501. https://doi.org/10.1051/metal/2020041
• 74)   H.  Long,  G. G.  Cheng,  B.  Wu and  Y.  Wu: J. Chin. Soc. Rare Earths, 28 (2010), 721. https://oversea.cnki.net/KNavi/JournalDetail?pcode=CJFD&pykm=XTXB
• 75)   T.  Honma,  Y.  Benino,  T.  Fujiwara,  T.  Komatsu,  R.  Sato and  V.  Dimitrov: J. Appl. Phys., 91 (2002), 2942. https://doi.org/10.1063/1.1436292
• 76)   C.  Shi: ISIJ Int., 60 (2020), 1083. https://doi.org/10.2355/isijinternational.ISIJINT-2019-661
• 77)   C. B.  Shi,  Y.  Huang,  J. X.  Zhang,  J.  Li and  X.  Zheng: Int. J. Miner. Metall. Mater., 28 (2021), 18. https://doi.org/10.1007/s12613-020-2075-3
• 78)   C. B.  Shi,  X. C.  Chen,  H. J.  Guo,  Z. J.  Zhu and  H.  Ren: Steel Res. Int., 83 (2012), 472. https://doi.org/10.1002/srin.201100200
• 79)   Q.  Wang,  R.  Wang,  Z.  He,  G.  Li,  B.  Li and  H.  Li: Int. J. Heat Mass Transf., 125 (2018), 1333. https://doi.org/10.1016/j.ijheatmasstransfer.2018.04.168
• 80)   D.  Hou,  Z.  Jiang,  Y.  Dong,  Y.  Cao,  H.  Cao and  W.  Gong: Ironmaking Steelmaking, 43 (2016), 517. https://doi.org/10.1080/03019233.2015.1110920
• 81)   D.  Hou,  Z. H.  Jiang,  T. P.  Qu,  D. Y.  Wang,  F. B.  Liu and  H. B.  Li: J. Iron Steel Res. Int., 26 (2019), 20. https://doi.org/10.1007/s42243-018-0107-2
• 82)   S. C.  Duan,  M. J.  Lee,  J. H.  Park and  H. J.  Guo: JOM, 74 (2022), 1228. https://doi.org/10.1007/s11837-021-05139-2
• 83)   S.  Hara,  H.  Hashimoto and  K.  Ogino: Trans. Iron Steel Inst. Jpn., 23 (1983), 1053. https://doi.org/10.2355/isijinternational1966.23.1053
• 84)   A.  Mitchell and  J.  Cameron: Metall. Mater. Trans. B, 2 (1971), 3361. https://doi.org/10.1007/BF02811617
• 85)   K.  Morinaga,  Y.  Ohta and  T.  Yanagase: J. Jpn. Inst. Met., 50 (1986), 907. https://doi.org/10.2320/jinstmet1952.50.10_907
• 86)   K.  Ogino,  H.  Hashimoto and  S.  Hara: Tetsu-to-Hagané, 64 (1978), 225 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.64.2_225
• 87)   S. C.  Duan,  X.  Shi,  M. C.  Zhang,  B.  Li,  W. S.  Yang,  F.  Wang,  H. J.  Guo and  J.  Guo: Metall. Mater. Trans. B, 51 (2020), 353. https://doi.org/10.1007/s11663-019-01729-3
• 88)   S. C.  Duan and  H. J.  Guo: Steel Res. Int., 91 (2020), 1900634. https://doi.org/10.1002/srin.201900634
• 89)   K.  Narita,  T.  Onoye,  T.  Ishii and  T.  Kusamichi: Tetsu-to-Hagané, 64 (1978), 1568 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.64.10_1568
• 90)   P.  Presoly,  J.  Korp and  R.  Schneider: Arch. Metall. Mater., 53 (2008), 567. http://www.imim.pl/files/archiwum/Vol2_2008/artykuly/Presoly.pdf
• 91)   C.  Wang,  S.  Ye,  Y.  Hu and  Q.  Du: Acta Metall. Sin., 22 (1986), 77. https://www.ams.org.cn/CN/Y1986/V22/I4/77
• 92)   C.  Wang and  Y.  Zou: Acta Metall. Sin., 16 (1980), 190. https://www.ams.org.cn/CN/Y1980/V16/I2/190