Study on the Melting Temperature of CaF2-CaO-MgO-Al2O3-TiO2 Slag under the Condition of a Fixed Ratio of Titanium and Aluminum in the Steel during the Electroslag Remelting Process

During the process of electroslag remelting (ESR) of steel containing titanium and aluminum, the activity ratio between titania and alumina in CaF2-CaO-MgO-Al2O3-TiO2 slag must be fixed in order to guarantee the titanium and aluminum contents in the ESR ingots. Under the condition of fixed activity ratio between titania and alumina in the slag, the melting temperature of slag should be investigated to improve the surface quality of ESR ingots. Therefore, this paper focuses on finding a kind of slag with low melting temperature that can be used for producing steel containing titanium. In the current study, the thermodynamic equilibrium of 3[Ti] + 2(Al2O3) = 4[Al] + 3(TiO2) between SUS321 steel and the two slag systems (CaF2:MgO:CaO:Al2O3:TiO2 = 46:4:25:(25 − x):x and CaF2:MgO:CaO:Al2O3:TiO2 = 46:4:(25 − 0.5 x):(25 − 0.5 x):x) are studied in an electrical resistance furnace based on Factsage software. After obtaining the equilibrium slag with fixed activity ratio between titania and alumina, the melting temperatures of the two slag systems are studied using slag melting experimental measurements and phase diagrams. The results show that the slag systems CaF2:MgO:CaO:Al2O3:TiO2 = 46:4:25:(25 − x):x, which consists of pre-melted slag S0 (CaF2:MgO:CaO:Al2O3 = 46:4:25:25) and pre-melted slag F1 (CaF2:MgO:CaO:TiO2 = 46:4:25:25), can not only control the aluminum and titanium contents in steel, but also have the desired low melting temperature property.


Introduction
Electroslag remelting (ESR) [1,2] is one of the processes used to produce high quality special steels. During ESR process, the slag plays important roles in chemical composition and surface quality of ingot. On the one hand, the slag CaF 2 -CaO-MgO-Al 2 O 3 -TiO 2 should have the fixed activity ratio of lg(a 3 TiO 2 /a 2 Al 2 O 3 ) to guarantee the thermodynamic equilibrium of 3[Ti] + 2(Al 2 O 3 ) = 4[Al] + 3(TiO 2 ) and the ratio of lg(w 3 [Ti] /w 4 [Al] ) in steel. On the other hand, the slag should also have a low melting temperature to improve the surface quality of the ESR ingots. Especially for superalloy or stainless steel with a melting temperature lower than 1370 • C (1643 K), the melting point of the slag used for ESR of superalloy and stainless steel should be lower than 1270 • C (1543 K). Therefore, it is essential to investigate the melting temperature of CaF 2 -CaO-MgO-Al 2 O 3 -TiO 2 slag under the condition of fixing activity ratio of titania and alumina in the slag during the ESR process.
The studies on CaF 2 -CaO-MgO-Al 2 O 3 -TiO 2 slag used for steel containing titanium are mainly divided into two categories: one is about the effect of TiO 2 on the physical property of slag, and the other is about the effect of each slag component on the activities of TiO 2 and Al 2 O 3 . Shi [3][4][5] studied the effect of TiO 2 on the crystallization behavior of CaF 2 -CaO-Al 2 O 3 -MgO-TiO 2 slag and pointed out that TiO 2 has large effect on the physical property of the slag. Duan [6,7] studied the effect of each slag component on the activities of Al 2 O 2 and TiO 2 , and determined an appropriate slag to be used for ESR of superalloys based on experiments and thermodynamics. Jiang [8][9][10][11][12][13][14][15][16] investigated the thermodynamic equilibrium of 3[Ti] + 2(Al 2 O 3 ) = 4[Al] + 3(TiO 2 ) and the effect of slag components on activities of Al 2 O 2 and TiO 2 in CaF 2 -CaO-MgO-Al 2 O 3 -TiO 2 slag system, and then TiO 2 in slag was calculated to control the titanium and aluminum contents in ESR ingot. However, the above researches [17][18][19][20][21] did not comprehensively consider the physical properties and thermodynamic equilibrium of 3[Ti] + 2(Al 2 O 3 ) = 4[Al] + 3(TiO 2 ). Under the condition of controlling the titanium and aluminum contents in steel, the optimized slag with low melting temperature cannot be acquired according to the studies above.
To the best of the authors' knowledge, under the condition of fixing activity ratio between titania and alumina in slag, investigation on the melting temperature of slag has not been reported so far. In the present work, the thermodynamic equilibrium of 3[Ti] + 2(Al 2 O 3 ) = 4[Al] + 3(TiO 2 ) was studied by the slag-metal reaction in a resistence furnace and the Factsage software. After obtaining the equilibrium slag with fixed activity ratio between titania and alumina, the melting temperature of CaF 2 -CaO-Al 2 O 3 -MgO-TiO 2 slag was studied by slag melting experimental measurements and phase diagram. At last, the slag design diagram consisting of lg(a 3 TiO 2 /a 2 Al 2 O 3 ) isoactivity lines and slag phase diagram (CaF 2 :MgO:CaO:Al 2 O 3 :TiO 2 = 46:4:x:y:z, x + y + z = 50) was made for acquiring the optimized CaF 2 -CaO-Al 2 O 3 -MgO-TiO 2 slag with low melting temperature.

Slag-Metal Reaction Experiments in Resistance Furnace
SUS321 stainless steel produced by Dongbei special steel group Co. Ltd, Dalian, China was used in current study. Its chemical composition is listed in Table 1. The chemical  compositions of Slag S0F1-82, S0F1-64, S0F2-82 and S0F2-64 are listed in Table 2, and the chemical compositions of pre-melted slag S0, F1 and F2 are listed in Table 3. Each slag-metal reaction experiment is carried out with 80 g slag and 50 g steel by using a resistance furnace, as shown in Figure 1. The heating unit is made of molybdenum disilicide. The temperature of the liquid metal is continuously measured by means of a B-type reference thermocouple produced by Kejing material technology Co. Ltd, Hefei, China. Argon is used to protect the slag-metal reaction system from top and bottom of the furnace at the rate of 2 Nl/min.  To the best of the authors' knowledge, under the condition of fixing activity ratio between titania and alumina in slag, investigation on the melting temperature of slag has not been reported so far. In the present work, the thermodynamic equilibrium of 3[Ti] + 2(Al2O3) = 4[Al] + 3(TiO2) was studied by the slag-metal reaction in a resistence furnace and the Factsage software. After obtaining the equilibrium slag with fixed activity ratio between titania and alumina, the melting temperature of CaF2-CaO-Al2O3-MgO-TiO2 slag was studied by slag melting experimental measurements and phase diagram. At last, the slag design diagram consisting of lg / isoactivity lines and slag phase diagram (CaF2:MgO:CaO:Al2O3:TiO2 = 46:4:x:y:z, x + y + z = 50) was made for acquiring the optimized CaF2-CaO-Al2O3-MgO-TiO2 slag with low melting temperature.

Slag-Metal Reaction Experiments in Resistance Furnace
SUS321 stainless steel produced by Dongbei special steel group Co. Ltd, Dalian, China was used in current study. Its chemical composition is listed in Table 1. The chemical compositions of Slag S0F1-82, S0F1-64, S0F2-82 and S0F2-64 are listed in Table 2, and the chemical compositions of pre-melted slag S0, F1 and F2 are listed in Table 3. Each slag-metal reaction experiment is carried out with 80 g slag and 50 g steel by using a resistance furnace, as shown in Figure 1. The heating unit is made of molybdenum disilicide. The temperature of the liquid metal is continuously measured by means of a B-type reference thermocouple produced by Kejing material technology Co. Ltd, Hefei, China. Argon is used to protect the slag-metal reaction system from top and bottom of the furnace at the rate of 2 Nl/min.   The experimental procedures can be described as follows. Firstly, 50 g of steel and 80 g of slag are placed into a MgO crucible with 30 mm in inner diameter and 70 mm in depth. Then the crucible is placed in a graphite crucible with molybdenum wire for suspension. After the whole crucible is placed in the chamber, the power is switched on and the furnace is heated to the experimental temperature (1823 K (1550 • C)) at a rate of 8 K/min.
After the furnace temperature was held for 60 min at 1823 K (1550 • C) [8,9], the crucible was dropped into liquid water quickly. The contents of Si, Al and Ti in each steel sample are analyzed by the inductively coupled plasma-mass spectroscopy (ICP-MS) technique and the concentrations of Al 2 O 3 , TiO 2 and MgO in slag samples are analyzed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES). The results are listed in Table 4.

Slag Melting Temperature Tests
During industrial ESR of steel containing different titanium and aluminum contents process, the TiO 2 powder combined with pre-melted slag CaF 2 -CaO-Al 2 O 3 -MgO are added into the water cooling molds of the ESR furnace. In order to prevent the TiO 2 powder from volatilizing with the air flow during the slag addition process, a new pre-melted slag S0 without TiO 2 and a pre-melted slag F1(F2) with high TiO 2 are designed. Their compositions are listed in Table 3. By mixing S0 and F1 in the ratio of 88:12, the slag S0F1-1 in Table 3 was acquired. Slag S0F1-2 and S0F1-3 can be acquired when the ratios of S0:F1 are 76:24 and 64:36, respectively. Slag S0F2-1, S0F2-2 and S0F2-3 can be acquired when the ratios of S0:F2 are 88:12, 76:24 and 64:36, respectively.
Slag melting experiments were carried out by using a high temperature specimen deformation method. A diagram of the test system is shown in Figure 2. In order to evaluate the melting behavior, the pre-melted slag powders were compressed into cylindrical samples of 3 mm diameter and 3 mm high. For each test, the slag sample was placed at the centre of a corundum substrate which was then located within the hot zone of a molybdenum wire furnace. The furnace was heated at 10 • C/min up to the slag melting temperature, which is defined as the temperature at which the cylindrical specimen attained a hemispherical shape. The melting temperature of slag was measured using a high temperature microscope, and the results are listed in Table 3.

Slag Melting Temperature Tests
During industrial ESR of steel containing different titanium and aluminum contents process, the TiO2 powder combined with pre-melted slag CaF2-CaO-Al2O3-MgO are added into the water cooling molds of the ESR furnace. In order to prevent the TiO2 powder from volatilizing with the air flow during the slag addition process, a new pre-melted slag S0 without TiO2 and a pre-melted slag F1(F2) with high TiO2 are designed. Their compositions are listed in Table 3. By mixing S0 and F1 in the ratio of 88:12, the slag S0F1-1 in Table 3 was acquired. Slag S0F1-2 and S0F1-3 can be acquired when the ratios of S0:F1 are 76:24 and 64:36, respectively. Slag S0F2-1, S0F2-2 and S0F2-3 can be acquired when the ratios of S0:F2 are 88:12, 76:24 and 64:36, respectively.
Slag melting experiments were carried out by using a high temperature specimen deformation method. A diagram of the test system is shown in Figure 2. In order to evaluate the melting behavior, the pre-melted slag powders were compressed into cylindrical samples of 3 mm diameter and 3 mm high. For each test, the slag sample was placed at the centre of a corundum substrate which was then located within the hot zone of a molybdenum wire furnace. The furnace was heated at 10 °C/min up to the slag melting temperature, which is defined as the temperature at which the cylindrical specimen attained a hemispherical shape. The melting temperature of slag was measured using a high temperature microscope, and the results are listed in Table 3.

Slag-Metal Reaction Experiments Results
The results of slag-metal reaction experiments in resistance furnace are shown in Table 4. Due to the existence of unstable oxides SiO2 and FeO, both aluminum and titanium are lower than them in the steel before experiments. If assuming that the slag-metal reaction of 3[Ti] + 2(Al2O3) = 4[Al] + 3(TiO2) in Table 4 reaches thermodynamic equilibrium [8,9], the activity coefficients of alloy element in steel and oxide component in slag can be experimental measured based on thermodynamics. At the slag-metal interface under 1550 °C, the following Reaction (1) will take place [22,23]. After substituting Ti, Al, Al2O3 and TiO2 of Table 4 into Equation (2), the activity coefficient value of Equation (2) in Exp.S0F1-82, Exp.S0F2-82, Exp.S0F1-64 and Exp.S0F1-64 are experimental measured as −3.57, −3.21, −3.42 and −2.89, as shown in Table 4.

Slag-Metal Reaction Experiments Results
The results of slag-metal reaction experiments in resistance furnace are shown in Table 4. Due to the existence of unstable oxides SiO 2 and FeO, both aluminum and titanium are lower than them in the steel before experiments. If assuming that the slag-metal reaction of 3[Ti] + 2(Al 2 O 3 ) = 4[Al] + 3(TiO 2 ) in Table 4 reaches thermodynamic equilibrium [8,9], the activity coefficients of alloy element in steel and oxide component in slag can be experimental measured based on thermodynamics. At the slag-metal interface under 1550 • C, the following Reaction (1) will take place [22,23]. After substituting Ti, Al, Al 2 O 3 and TiO 2 of Table 4 into Equation (2), the activity coefficient value of Equation (2)  is the activity coefficient of Equation (2). During the slag-metal reaction experiments, the MgO in slag after experiments was increased to 10% because of MgO crucible being eroded by slag. In order to investigate the thermodynamic equilibrium of SUS321 steel and slag S0-F1(F2) further, the activity coefficients of Ti and Al in steel are calculated by Equation (3) and the value of lg( f 3 [Ti] / f 4 [Al] ) is considered as −0.12. The interaction parameters [24][25][26] used in present study are listed in Table 5 Table 4.  During the slag-metal reaction experiments, the MgO in slag after experiments was increased to 10% because of MgO crucible being eroded by slag. In order to investigate the thermodynamic equilibrium of SUS321 steel and slag S0-F1(F2) further, the activity coefficients of Ti and Al in steel are calculated by Equation (3) Table 4.    Table 6, which will be used as points in Figure 4.   Table 6, which will be used as points in Figure 4.   It is also can be seen that the activity coefficients of Equation (2) in each slag listed in Table 6 are different, it is different CaO content in slag that changes the activity coefficients of TiO2 and Al2O3, which has been studied based on ion and molecular coexistence theory (IMCT) in the previous study [8,9]. The changes of lg / with CaO are calculated based on Factsage software, as shown in Figure 3b. The lg / decreases with the increase of CaO in slag, which means the ratio of lg / in slag should increase with the increase of CaO under the condition of fixed lg / in slag. The changes of S0: F1 and S0: F2 ratios in slag mixtures with lg / are calculate according to Equations (1)- (3) and Factsage software at the temperature of 1550 °C, as shwon in Figure 4. It is can be seen that the points listed in Table 6 are in good agreement with the calculated results based on thermodynamics and Factage software. The slag S0-F1 containing high CaO needs large ratio of lg / to guarantee the thermodynamic equilibrium of 3[Ti] + 2(Al2O3) = 4[Al] + 3(TiO2) and the ratio of lg / in steel. If the titanium and aluminum contents in steel are given, the mixture ratios between pre-melted slag S0 and F1(F2) can be acquired according to Figure 4. In order to compare the melting temperature of two slag systems S0-F1 and S0-F2 under the condition of fixing the titanium and aluminum contents in steel, the steel with lg / 4.10 and 5.03 combined with corresponding slag systems S0-F1 and S0-F2 are calculated, as shown in Table 7. It is also can be seen that the activity coefficients of Equation (2) in each slag listed in Table 6 are different, it is different CaO content in slag that changes the activity coefficients of TiO 2 and Al 2 O 3 , which has been studied based on ion and molecular coexistence theory (IMCT) in the previous study [8,9]. The changes of lg(γ 3 TiO 2 /γ 2 Al 2 O 3 ) with CaO are calculated based on Factsage software, as shown in Figure 3b. The lg(γ 3 TiO 2 /γ 2 Al 2 O 3 ) decreases with the increase of CaO in slag, which means the ratio of lg(w 3 TiO 2 /w 2 Al 2 O 3 ) in slag should increase with the increase of CaO under the condition of fixed lg(γ 3 TiO 2 /γ 2 The changes of S0:F1 and S0:F2 ratios in slag mixtures with lg(w 3 [Ti] /w 4 [Al] ) are calculate according to Equations (1)- (3) and Factsage software at the temperature of 1550 • C, as shwon in Figure 4. It is can be seen that the points listed in Table 6 are in good agreement with the calculated results based on thermodynamics and Factage software. The slag S0-F1 containing high CaO needs large ratio of lg(w 3 TiO 2 /w 2 Al 2 O 3 ) to guarantee the thermodynamic equilibrium of 3[Ti] + 2(Al 2 O 3 ) = 4[Al] + 3(TiO 2 ) and the ratio of lg(w 3 [Ti] /w 4 [Al] ) in steel. If the titanium and aluminum contents in steel are given, the mixture ratios between pre-melted slag S0 and F1(F2) can be acquired according to Figure 4. In order to compare the melting temperature of two slag systems S0-F1 and S0-F2 under the condition of fixing the titanium and aluminum contents in steel, the steel with lg(w 3 [Ti] /w 4 [Al] ) 4.10 and 5.03 combined with corresponding slag systems S0-F1 and S0-F2 are calculated, as shown in Table 7. Then the melting temperatures of thermodynamic equlibrium slag systems in Table 7 are measured in slag melting experiments, and the results are listed in Table 3.

Slag Melting Temperature Results
The halfsphere melting temperature and flowing melting temperature results for the slag listed in Table 3 are shown in Figure 5. It is clear that: (i) with a CaO/(Al 2 O 3 + TiO 2 ) ratio = 1, the melting temperature of the slag S0-F1 is lower than slag S0-F2 with a CaO/Al 2 O 3 ratio = 1 under the condition of increasing TiO 2 content in slag; (ii) when the TiO 2 content reaches more than 9%, the melting temperatures of slag S0F1-3 and S0F1-4 are much lower than that of slag S0F2-4 and S0F2-5; (iii) the melting temperatures of slag S0F1-4 is much lower than that of slag S0F2-4 in Table 7 under the condition of fixing steel with lg(w 3 [Ti] /w 4 [Al] )5.03; (iv) the melting temperature of the slag S0-F2 system (CaO/Al 2 O 3 ratio = 1) decreases first and then increases with the increase of TiO 2 content, which has been described in detail based on SHTT, SEM and XRD in [3]. As the description of conclusion in [3], TiO 2 addition from 0 to 6.43 mass% inhibited crystallisation behaviour of CaF 2 -CaO-MgO-Al 2 O 3 ESR type slag, whereas the further TiO 2 addition up to 9.73 mass% greatly enhanced the crystallisation tendency.

Slag Melting Temperature Results
The halfsphere melting temperature and flowing melting temperature results for the slag listed in Table 3 are shown in Figure 5. It is clear that: (i) with a CaO/(Al2O3 + TiO2) ratio = 1, the melting temperature of the slag S0-F1 is lower than slag S0-F2 with a CaO/Al2O3 ratio = 1 under the condition of increasing TiO2 content in slag; (ii) when the TiO2 content reaches more than 9%, the melting temperatures of slag S0F1-3 and S0F1-4 are much lower than that of slag S0F2-4 and S0F2-5; (iii) the melting temperatures of slag S0F1-4 is much lower than that of slag S0F2-4 in Table 7 under the condition of fixing steel with lg / 5.03; (iv) the melting temperature of the slag S0-F2 system (CaO/Al2O3 ratio = 1) decreases first and then increases with the increase of TiO2 content, which has been described in detail based on SHTT, SEM and XRD in [3]. As the description of conclusion in [3], TiO2 addition from 0 to 6.43 mass% inhibited crystallisation behaviour of CaF2-CaO-MgO-Al2O3 ESR type slag, whereas the further TiO2 addition up to 9.73 mass% greatly enhanced the crystallisation tendency.
The phenomena whereby 'the melting temperature of the slag S0F2 system decreases first and then increases with the increase of TiO2 content' and 'the melting temperature of the slag S0F1 system decreases with the increase of TiO2 content' is further explained according to the phase diagram of CaF2-CaO-MgO-Al2O3-TiO2 (CaF2 = 46% and MgO = 4%) calculated by Factsage, as shown in Figure 6. It can be seen that slag S0-F1 is closer to the low melting point region with the increase of F1:S0 ratio. TiO2 addition from 0 to 6 mass% promotes S0-F2 to approach the low melting point region, whereas the further TiO2 addition up to 9 mass% makes S0-F2 away from the low melting point region.  The phenomena whereby 'the melting temperature of the slag S0F2 system decreases first and then increases with the increase of TiO 2 content' and 'the melting temperature of the slag S0F1 system decreases with the increase of TiO 2 content' is further explained according to the phase diagram of CaF 2 -CaO-MgO-Al 2 O 3 -TiO 2 (CaF 2 = 46% and MgO = 4%) calculated by Factsage, as shown in Figure 6. It can be seen that slag S0-F1 is closer to the low melting point region with the increase of F1:S0 ratio. TiO 2 addition from 0 to 6 mass% promotes S0-F2 to approach the low melting point region, whereas the further TiO 2 addition up to 9 mass% makes S0-F2 away from the low melting point region.

The Optimized Low Melting Temperature Slag Used for Steel Containing Ti and Al
It is final goals to acquire the optimized slag with low melting temperature under the condition of fixing lg(w 3 [  [27], the melting temperature of slag S0-F1 would be decreased further according to Figure 7.

The Optimized Low Melting Temperature Slag Used for Steel Containing Ti and Al
It is final goals to acquire the optimized slag with low melting temperature under the condition of fixing lg / ratio between Ti and Al contents in steel. During ESR of steel containing lg / = 5.03 and 4.10, the lg / in slag can be calculated as −4.51 and −5.44 according to Equations (2)  The melting temperature of CaF2-CaO-MgO-Al2O3-TiO2 slag systems would increase with the decrease of CaO content. In addition, with the increase of CaO in slag due to the reaction of 3CaF2 + Al2O3 = 2AlF3 (g) + 3CaO during long term ESR process [27], the melting temperature of slag S0-F1 would be decreased further according to Figure 7.

The Optimized Low Melting Temperature Slag Used for Steel Containing Ti and Al
It is final goals to acquire the optimized slag with low melting temperature under the condition of fixing lg / ratio between Ti and Al contents in steel. During ESR of steel containing lg / = 5.03 and 4.10, the lg / in slag can be calculated as −4.51 and −5.44 according to Equations (2)  The melting temperature of CaF2-CaO-MgO-Al2O3-TiO2 slag systems would increase with the decrease of CaO content. In addition, with the increase of CaO in slag due to the reaction of 3CaF2 + Al2O3 = 2AlF3 (g) + 3CaO during long term ESR process [27], the melting temperature of slag S0-F1 would be decreased further according to Figure 7.

Conclusions
The melting temperature of two slag systems and thermodynamic equilibrium of (2) The melting temperature of slag S0-F1 with a CaO/(Al 2 O 3 + TiO 2 ) ratio = 1 is lower than that of slag S0-F2 with a CaO/Al 2 O 3 ratio = 1. Especially for thermodynamic equilibrium slag containing high TiO 2 , the melting temperature of S0-F1 slag CaF