Experimental Study on Strengthening Carbothermic Reduction of Vanadium-Titanium-Magnetite by Adding CaF2

The effects and reduction mechanisms of carbothermic reduction of vanadium–titanium–magnetite were studied by adding various mass fractions of CaF2 ranging from 0%, 1%, 3%, 5% to 7%. The results showed that the proper CaF2 addition could strengthen the carbothermic reduction of vanadium–titanium–magnetite while the excessive amounts will weaken the promotive effect, hence the appropriate dosage was determined to be 3 mass%. The CaF2 was favorable for the carbon gasification reaction, where it increased the partial pressure of CO inside briquette and caused the lattice distortion of vanadium–titanium–magnetite. The reaction improved the reduction process and accelerated the reduction rate. The appearance of 3CaO·2SiO2·CaF2 and other complex compounds with low melting point facilitated the aggregation and growth of the slag and the iron, which increased the concentration of iron grains and the aggregation level of the slag.


Introduction
Vanadium-titanium-magnetite is a valuable resource [1][2][3][4][5][6]. Vanadium-titanium-magnetite has a variety of species, good conditions for storage, and high comprehensive utilization value. Vanadium-titanium-magnetite refining could recover a variety of valuable metal elements and obtain a variety of industrial products, which plays an important role in the metallurgical industry. The main refining technique for vanadium titanium magnetite is the "blast furnace-converter" process [7,8], which is limited by the fact that titanium is completely unrecoverable, it causes a huge waste of titanium resources and simultaneously environmental pollution. To realize the efficient utilization of vanadium-titanium-magnetite has been an industrial goal of related fields at home and abroad [9][10][11][12]. Solid-state reduction technology could get a better reduction effect at a lower temperature. However, due to the combination of different titanium compounds and iron compounds in vanadium-titanium-magnetite, the complex and dense mineral structure are formed, which makes reduction difficult [13,14]. Compared with the reduction of common iron-bearing minerals, vanadium-titanium-magnetite needs higher energy and lower production efficiency.
At present, there are a lot of previous researches regarding the reinforced carbothermic reduction of ilmenite [15][16][17]. However, few researches can be found regarding-vanadium titanium-magnetite. Song et al. [15] studied the carbothermic reduction process of ilmenite concentrates with added sodium borate, demonstrating it can effectively improve the metallization degree of the carbothermic reduction and reduce the reduction temperature. Liu et al [16] investigated the effects of alkali metal additives on the pre-oxidation and the solid-state reduction reaction of ilmenite, indicating that many additives could promote the reaction. Borax (Na 2 B 4 O 7 ) was found to have an especially strong effect on the degree of reduction and metallization. Wu et al [17] discovered that alkali oxides could accelerate reduction reaction. Mono-metal oxide is better than binary, ternary, or quaternary oxides. Different alkali compounds as an accelerator could increase the metallization rate in different degrees [18]. Chen et al. [19] studied the various influencing factors of carbothermal reduction of vanadium-titanium-magnetite, and analyzed the reasons for the difficulty of reduction of vanadium-titanium-magnetite. The research on ilmenite reduction provided a lot of important information for the reinforced carbothermic reduction of vanadium-titanium-magnetite. The measure of adding catalysts is simple in operation, and the strengthening effect is obvious. CaF 2 is a flux commonly used in the metallurgical industry [20,21]. It is used in titanium-containing blast furnace slag, which reduces the viscosity and improves the stability. In addition, CaF 2 could be used not only as a reducing agent, but also as a carbon activation agent to make the carbon porous and subsequently more reactive. However, the effect of additives CaF 2 on vanadium titanium magnetite is not clear.
In the present study, the effect of adding CaF 2 on the isothermal carbothermal reduction process of vanadium-titanium-magnetite was studied. Meanwhile, the reduction mechanism was also studied using X-ray diffraction (XRD), scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) to provide empirical guidance for the high efficient industrial utilization of the vanadium-titanium-magnetite.

Experimental Materials
The material of vanadium-titanium-magnetite used in the current study was obtained from Liaoning, China. The ore was ground into fines with a size distribution of 98% less than 74 µm. Its component analysis and phase analysis are shown in Table 1 and Figure 1, respectively. The additive was analytic pure CaF 2 with a size distribution of 100% less than 74 µm. The reductant was high purity graphite powder with 99.8% fixed carbon. degree of the carbothermic reduction and reduce the reduction temperature. Liu et al [16] investigated the effects of alkali metal additives on the pre-oxidation and the solid-state reduction reaction of ilmenite, indicating that many additives could promote the reaction. Borax (Na2B4O7) was found to have an especially strong effect on the degree of reduction and metallization. Wu et al [17] discovered that alkali oxides could accelerate reduction reaction. Mono-metal oxide is better than binary, ternary, or quaternary oxides. Different alkali compounds as an accelerator could increase the metallization rate in different degrees [18]. Chen et al. [19] studied the various influencing factors of carbothermal reduction of vanadium-titanium-magnetite, and analyzed the reasons for the difficulty of reduction of vanadium-titanium-magnetite. The research on ilmenite reduction provided a lot of important information for the reinforced carbothermic reduction of vanadiumtitanium-magnetite. The measure of adding catalysts is simple in operation, and the strengthening effect is obvious. CaF2 is a flux commonly used in the metallurgical industry [20,21]. It is used in titanium-containing blast furnace slag, which reduces the viscosity and improves the stability. In addition, CaF2 could be used not only as a reducing agent, but also as a carbon activation agent to make the carbon porous and subsequently more reactive. However, the effect of additives CaF2 on vanadium titanium magnetite is not clear.
In the present study, the effect of adding CaF2 on the isothermal carbothermal reduction process of vanadium-titanium-magnetite was studied. Meanwhile, the reduction mechanism was also studied using X-ray diffraction (XRD), scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) to provide empirical guidance for the high efficient industrial utilization of the vanadium-titanium-magnetite.

Experimental Materials
The material of vanadium-titanium-magnetite used in the current study was obtained from Liaoning, China. The ore was ground into fines with a size distribution of 98% less than 74 µ m. Its component analysis and phase analysis are shown in Table 1 and Figure 1, respectively. The additive was analytic pure CaF2 with a size distribution of 100% less than 74 µ m. The reductant was high purity graphite powder with 99.8% fixed carbon.  As shown in Table 1, the iron grade of vanadium-titanium-magnetite is relatively low and the As shown in Table 1, the iron grade of vanadium-titanium-magnetite is relatively low and the MgO content is only 0.29%. The impurity content is high. The contents of SiO 2, Al 2 O 3 , and CaO are 6.28%, 1.74%, and 3.72% respectively. The TiO 2 content reaches 12.96%, which is difficult to utilize by traditional ironmaking technology. Figure 1 depicts the diffraction peaks of vanadium-titanium-magnetite concentrates, where the ores with superior crystallization are predominantly composed of iron, ilmenite (FeTiO 3 ), ulvospinel (Fe 2 TiO 4 ) and titanomagnetite (Fe 2.75 Ti 0.25 O 4 ).

Experimental Process and Characterization Method
The main apparatus used for the experiment included a vacuum drying oven (DZF-6050), a molding press, a resistance furnace with a rated temperature of 1600 • C, an electronic balance, a data acquisition system, a thermocouple, and a temperature controller. The schematic diagram of the experimental apparatus is shown in Figure 2.

Experimental Process and Characterization Method
The main apparatus used for the experiment included a vacuum drying oven (DZF-6050), a molding press, a resistance furnace with a rated temperature of 1600 °C , an electronic balance, a data acquisition system, a thermocouple, and a temperature controller. The schematic diagram of the experimental apparatus is shown in Figure 2. The isothermal experiment was used to clarify the mechanism that CaF2 had on the carbothermic reduction process of vanadium-titanium-magnetite. The C/O was 1.2, and a certain amount of vanadium-titanium-magnetite and high purity graphite powder were weighed. The CaF2 was added as 0%, 1%, 3%, 5%, and 7%.The powders were mixed homogeneously in acetone. The cylindrical samples were 10 mm × 10 mm created via a molding press under 10 Mpa. The samples were dried for 2 h in a vacuum drying oven, and then loaded in the high-temperature experiment apparatus to conduct the isothermal reduction experiment. The experiments were finished at 1100 °C , 1200 °C , 1300 °C and 1400 °C , respectively. The exhaust gas was introduced into a container containing NaOH solution for the treatment of SiF4. After the samples were kept at the experimental temperature, they were rapidly cooled in argon and analyzed via XRD and SEM-EDS. To characterize the effect of CaF2 on the reduction process, the reaction fraction was selected as the evaluation index, seen in Equation (1). (1) f--reaction fraction, %; △Wt--the weight loss of the carbonaceous briquettes after the reduction time (t), g; △Wmax--the maximum weight loss of the carbonaceous briquettes, g.

The Thermodynamic Analysis of Carbothermic Reduction
The carbothermic reduction process of vanadium-titanium-magnetite could be divided into two parts [22,23]. One part is that the iron oxides are directly reduced by C, the other is the indirect The isothermal experiment was used to clarify the mechanism that CaF 2 had on the carbothermic reduction process of vanadium-titanium-magnetite. The C/O was 1.2, and a certain amount of vanadium-titanium-magnetite and high purity graphite powder were weighed. The CaF 2 was added as 0%, 1%, 3%, 5%, and 7%.The powders were mixed homogeneously in acetone. The cylindrical samples were ϕ10 mm × 10 mm created via a molding press under 10 Mpa. The samples were dried for 2 h in a vacuum drying oven, and then loaded in the high-temperature experiment apparatus to conduct the isothermal reduction experiment. The experiments were finished at 1100 • C, 1200 • C, 1300 • C and 1400 • C, respectively. The exhaust gas was introduced into a container containing NaOH solution for the treatment of SiF 4 . After the samples were kept at the experimental temperature, they were rapidly cooled in argon and analyzed via XRD and SEM-EDS. To characterize the effect of CaF 2 on the reduction process, the reaction fraction was selected as the evaluation index, seen in Equation (1). (1) f --reaction fraction, %; W t --the weight loss of the carbonaceous briquettes after the reduction time (t), g; W max --the maximum weight loss of the carbonaceous briquettes, g.

The Thermodynamic Analysis of Carbothermic Reduction
The carbothermic reduction process of vanadium-titanium-magnetite could be divided into two parts [22,23]. One part is that the iron oxides are directly reduced by C, the other is the indirect reduction which used CO as the reductant, containing carbon dissolution reaction as well as the reduction of iron compounds. The possible reactions are listed in Table 2. Figure 3 shows the relationship between the G and the temperature. Table 2. The main reactions during the reduction process.

Temperature
Range / • C Equation 400-800 800-1400 (4) 400-700 700-1400 (8) reduction which used CO as the reductant, containing carbon dissolution reaction as well as the reduction of iron compounds. The possible reactions are listed in Table 2. Figure 3 shows the relationship between the △G and the temperature.  Table 2 shows that these reactions are mainly endothermic. So it is favorable for improving reactions at high temperature. At the early stage, the low CO partial pressure inside briquette causes the reduction process to be primarily the solid-solid reaction between carbon and iron compounds. As the temperature increases, ferrotitanium compound reacts with CO easily.

Influence of Additive Amounts of CaF2
The experiments were performed at various conditions, with 0%, 1%, 3%, 5%, and 7% CaF2 in order to explore the influence of CaF2 on the carbothermic reduction process. The results are shown in Figure 4.  Table 2 shows that these reactions are mainly endothermic. So it is favorable for improving reactions at high temperature. At the early stage, the low CO partial pressure inside briquette causes the reduction process to be primarily the solid-solid reaction between carbon and iron compounds. As the temperature increases, ferrotitanium compound reacts with CO easily.

Influence of Additive Amounts of CaF 2
The experiments were performed at various conditions, with 0%, 1%, 3%, 5%, and 7% CaF 2 in order to explore the influence of CaF 2 on the carbothermic reduction process. The results are shown in Figure 4.   Figure 4 shows that the addition of CaF2 can accelerate the reduction process. When the reduction temperature was 1100 °C , the reaction fraction rose by prolonging the reduction time, although the reaction could not reach equilibrium within the scope of the experiment time. The reaction process increased as the additive content increased and the temperature reduced with 1% CaF2. The reaction process reduced when more than 3% CaF2 was added. The higher the reduction temperature, the less time required for reduction equilibrium. When CaF2 of 3% was added, the equilibrium time for the carbothermic reduction at 1200 °C , 1300 °C , and 1400 °C was 22 min, 10 min, and 6 min, respectively. The reaction fraction was close to 90% when 3% of the CaF2 was added at 1200 °C . When the temperature was over 1300 °C , the change of reduction fraction was relatively small, reaching about 95% at 1400 °C . When CaF2 of 7% was added, the reaction fraction increased, but the increasing trend was not obvious. The optimum conditions were obtained at the reduction temperature of 1300 °C and additive amounts of 3%.

Morphological Analysis
The non-CaF2 and the 3% CaF2 were chosen for the experiments at 1300 °C . Figure 5 shows the SEM-EDS results for the products obtained with a reduction time of 5 min, 10 min, 20 min, and 30 min, respectively.  Figure 4 shows that the addition of CaF 2 can accelerate the reduction process. When the reduction temperature was 1100 • C, the reaction fraction rose by prolonging the reduction time, although the reaction could not reach equilibrium within the scope of the experiment time. The reaction process increased as the additive content increased and the temperature reduced with 1% CaF 2 . The reaction process reduced when more than 3% CaF 2 was added. The higher the reduction temperature, the less time required for reduction equilibrium. When CaF 2 of 3% was added, the equilibrium time for the carbothermic reduction at 1200 • C, 1300 • C, and 1400 • C was 22 min, 10 min, and 6 min, respectively. The reaction fraction was close to 90% when 3% of the CaF 2 was added at 1200 • C. When the temperature was over 1300 • C, the change of reduction fraction was relatively small, reaching about 95% at 1400 • C. When CaF 2 of 7% was added, the reaction fraction increased, but the increasing trend was not obvious. The optimum conditions were obtained at the reduction temperature of 1300 • C and additive amounts of 3%.

Morphological Analysis
The non-CaF 2 and the 3% CaF 2 were chosen for the experiments at 1300 • C. Figure 5 shows the SEM-EDS results for the products obtained with a reduction time of 5 min, 10 min, 20 min, and 30 min, respectively.

CaF 2 +0%
(d)  Figure 5 shows that the two types of samples could participate internally in the reduction reaction to generate bits of metallic iron quickly. The vanadium-titanium-magnetite particles reacted rapidly with the surrounding carbon and the CO created by the reaction. The particles were eroded gradually from outside to inside and the interface tended to be smooth. The iron grain size in the inner section of the reduction product with 3% CaF2 was larger. Due to the erosion of CaF2 and the occurrence of the reduction reaction, the surface of internal particles was relatively rough. As the reduction time increased, the iron phase inside samples increased constantly, and the grain size increased. However, the content or grain size of the iron phase inside briquettes adding CaF2 was better than others. This phenomenon demonstrated that CaF2 could enhance the chemical reaction rate. The larger the partial pressure of CO, the faster the diffusion rate was into the particles to seize  Figure 5 shows that the two types of samples could participate internally in the reduction reaction to generate bits of metallic iron quickly. The vanadium-titanium-magnetite particles reacted rapidly with the surrounding carbon and the CO created by the reaction. The particles were eroded gradually from outside to inside and the interface tended to be smooth. The iron grain size in the inner section of the reduction product with 3% CaF 2 was larger. Due to the erosion of CaF 2 and the occurrence of the reduction reaction, the surface of internal particles was relatively rough. As the reduction time increased, the iron phase inside samples increased constantly, and the grain size increased. However, the content or grain size of the iron phase inside briquettes adding CaF 2 was better than others. This phenomenon demonstrated that CaF 2 could enhance the chemical reaction rate. The larger the partial pressure of CO, the faster the diffusion rate was into the particles to seize the oxygen bound to the iron. This suggested that CaF 2 expedited the Boudouard reaction to accelerate the CO generation rate, which raised the partial pressure of reducing gas. When the reduction time was 30 min, there were a large number of metal irons in the products with either none or 3% CaF 2 . The mineral segregation results in a homogeneous distribution of metal iron around or inside the non-ferrous particles. The CaF 2 could not only facilitate the segregation of metal iron grains, but also erode the surface of particles to make them rough.

CaF 2 +3%
The grain sizes of the metal iron varied with the reduction time and they were analyzed with "Image-Pro-Plus", all the complete grains in the picture are selected out, and the average grain size is calculated. The results were shown in Figure 6. The CaF 2 had a significant effect on the formation, aggregation, and growth of the metal iron grains, and the effect was more evident with longer reduction time.
the oxygen bound to the iron. This suggested that CaF2 expedited the Boudouard reaction to accelerate the CO generation rate, which raised the partial pressure of reducing gas. When the reduction time was 30 min, there were a large number of metal irons in the products with either none or 3% CaF2. The mineral segregation results in a homogeneous distribution of metal iron around or inside the non-ferrous particles. The CaF2 could not only facilitate the segregation of metal iron grains, but also erode the surface of particles to make them rough.
The grain sizes of the metal iron varied with the reduction time and they were analyzed with "Image-Pro-Plus", all the complete grains in the picture are selected out, and the average grain size is calculated. The results were shown in Figure 6. The CaF2 had a significant effect on the formation, aggregation, and growth of the metal iron grains, and the effect was more evident with longer reduction time.  Figure 7 shows the XRD results of the reduction products that were reacted with none or 3% CaF2 for 5 min, 10 min, 20 min, and 30 min. The principal phases (shown in Figure 7) were Fe, FeTiO3, TTM, some Fe2TiO4 and FeTi2O5 after  Figure 7 shows the XRD results of the reduction products that were reacted with none or 3% CaF 2 for 5 min, 10 min, 20 min, and 30 min.

XRD Analysis
the oxygen bound to the iron. This suggested that CaF2 expedited the Boudouard reaction to accelerate the CO generation rate, which raised the partial pressure of reducing gas. When the reduction time was 30 min, there were a large number of metal irons in the products with either none or 3% CaF2. The mineral segregation results in a homogeneous distribution of metal iron around or inside the non-ferrous particles. The CaF2 could not only facilitate the segregation of metal iron grains, but also erode the surface of particles to make them rough.
The grain sizes of the metal iron varied with the reduction time and they were analyzed with "Image-Pro-Plus", all the complete grains in the picture are selected out, and the average grain size is calculated. The results were shown in Figure 6. The CaF2 had a significant effect on the formation, aggregation, and growth of the metal iron grains, and the effect was more evident with longer reduction time.  Figure 7 shows the XRD results of the reduction products that were reacted with none or 3% CaF2 for 5 min, 10 min, 20 min, and 30 min. The principal phases (shown in Figure 7) were Fe, FeTiO3, TTM, some Fe2TiO4 and FeTi2O5 after 5 min. The additive could be detected from the samples of 3% CaF2, indicating that the CaF2 did not react, basically remaining at the earlier stage. When the reduction time was 10 min, the CaF2 did not change the product phases. The peak values of Fe and FeTiO3 were high. The reaction time extended and the CaF2 with mixed mineral compositions including CaO and SiO2 were in the reaction. 3CaO·2SiO2·CaF2 is a low-melting complex compound, appeared, which was favorable for diffusion and iron phases aggregation. The production phases consisted of Fe, FeTi2O5 and a slight amount of titanomagnetite (TTM) with a 30 min reaction time. The behavior of the main phases in the reduction The principal phases (shown in Figure 7) were Fe, FeTiO 3 , TTM, some Fe 2 TiO 4 and FeTi 2 O 5 after 5 min. The additive could be detected from the samples of 3% CaF 2 , indicating that the CaF 2 did not react, basically remaining at the earlier stage. When the reduction time was 10 min, the CaF 2 did not change the product phases. The peak values of Fe and FeTiO 3 were high. The reaction time Minerals 2020, 10, 219 8 of 11 extended and the CaF 2 with mixed mineral compositions including CaO and SiO 2 were in the reaction. 3CaO·2SiO 2 ·CaF 2 is a low-melting complex compound, appeared, which was favorable for diffusion and iron phases aggregation. The production phases consisted of Fe, FeTi 2 O 5 and a slight amount of titanomagnetite (TTM) with a 30 min reaction time. The behavior of the main phases in the reduction process is shown in Figure 8. process is shown in Figure 8. Phase of raw Process phase Final Products Figure 8. The schematic diagram of the main phase transformation.

FeO·
The "Jade" software (Version 6.5) was used to interpret the XRD data of the 20 min reaction time in Figure 8. The XRD data of the crystal lattice parameters of the main phases amended by the "Rietveld" methods [21] are listed in Table 3.  Table 3 shows that the CaF2 had particular influences on the crystal structure of the ferrotitanium phase contained in vanadium-titanium-magnetite during the reduction process. In the TTM phase, the crystal cell stretched along the axis c, decreased slightly along the axis a, and its volume expanded. The crystal lattice parameter shortened from 5.07824Å to 5.07623Å , value c increased from 13.43312Å to 13.45201Å , and volume expanded from 346.42Å 3 to 346.63Å 3 . It can be observed that the change rules of the crystal lattice parameter of FeTiO3, FeTi2O5 and Fe2TiO4 are similar. The CaF2 distorted the crystal structure, which benefited the transformation of the TTM phase from the rhombohedral system to other reduction product crystal system and lowered the reduction difficulty between iron compounds with C and CO.

Mechanism Analysis of CaF2
According to the analysis results depicted in Figure 5, Figure 7 and Table 3, the schematic diagram of the samples reduction process with adding CaF2 is given in Figure 9. The CaF2 had little effect on the reduction process at the initial stage. As the reaction layer thickened, the reduction degree of vanadium-titanium-magnetite rose, the iron grains proliferated, and the slag phase formed. The CaF2 reacted with the high-melting-point phase to generate a certain amount of 3CaO·2SiO2·CaF2, which had a low-melting-point and could be a "lubricant" to the reaction layer The "Jade" software (Version 6.5) was used to interpret the XRD data of the 20 min reaction time in Figure 8. The XRD data of the crystal lattice parameters of the main phases amended by the "Rietveld" methods [21] are listed in Table 3.  Table 3 shows that the CaF 2 had particular influences on the crystal structure of the ferrotitanium phase contained in vanadium-titanium-magnetite during the reduction process. In the TTM phase, the crystal cell stretched along the axis c, decreased slightly along the axis a, and its volume expanded. The crystal lattice parameter shortened from 5.07824Å to 5.07623Å, value c increased from 13.43312Å to 13.45201Å, and volume expanded from 346.42Å 3 to 346.63Å 3 . It can be observed that the change rules of the crystal lattice parameter of FeTiO 3 , FeTi 2 O 5 and Fe 2 TiO 4 are similar. The CaF 2 distorted the crystal structure, which benefited the transformation of the TTM phase from the rhombohedral system to other reduction product crystal system and lowered the reduction difficulty between iron compounds with C and CO.

Mechanism Analysis of CaF 2
According to the analysis results depicted in Figure 5, Figure 7 and Table 3, the schematic diagram of the samples reduction process with adding CaF 2 is given in Figure 9. The CaF 2 had little effect on the reduction process at the initial stage. As the reaction layer thickened, the reduction degree of vanadium-titanium-magnetite rose, the iron grains proliferated, and the slag phase formed. The CaF 2 reacted with the high-melting-point phase to generate a certain amount of 3CaO·2SiO 2 ·CaF 2 , which had a low-melting-point and could be a "lubricant" to the reaction layer and the interstitial flow, promoting the migration of the iron and the aggregation of slag. Vanadium-titanium-magnetite with the CaF2 enabled the equilibrium of the original Boudouard reaction to be broken and then to catalytically form a new one, seen in Equations (13) and (15). 2CaF2 + 2C + SiO2 = 2Ca(g) + SiF4(g) + 2CO(g) (13) Ca(g) + 4CO(g) = CaC2 + 2CO2(g) (14) 2CaC2 + SiF4(g) + 6CO2(g) = 2CaF2 + SiO2 + 10CO(g).
The favorable reaction activity of the Ca (g) was not the restrictive step of the reaction, but rather led to its higher reaction rate. Equations (14) and (15) were combined to form the catalytic reaction cycle, which sped up the Boudouard reaction and produced CO, which improved the reduction process of the ferrotitanium compounds. The carbon surface participated in the Boudouard reaction that intensified the surface erosion degree of the high purity graphite powder. The larger contact area between the carbon particles and the vanadium-titanium-magnetite would update carbon layers to pile up the active points into the carbon surface and to build up the solid-solid reduction.

Conclusions
The addition of CaF2 facilitated the carbothermic reduction of the vanadium-titaniummagnetite, enhanced the reaction fraction, and the continuously increasing additive amounts and reduction temperature led to the reaction fraction tending to increase slowly. The appropriate condition was 3% CaF2 at 1300 °C , where reaction fraction could reach up to 95%.
The favorable reaction activity of the Ca (g) was not the restrictive step of the reaction, but rather led to its higher reaction rate. Equations (21) and (22) were combined to form the catalytic reaction cycle, which sped up the Boudouard reaction and produced CO, which improved the reduction process of the ferrotitanium compounds. The carbon surface participated in the Boudouard reaction that intensified the surface erosion degree of the high purity graphite powder. The larger contact area between the carbon particles and the vanadium-titanium-magnetite would update carbon layers to pile up the active points into the carbon surface and to build up the solid-solid reduction.

Conclusions
The addition of CaF 2 facilitated the carbothermic reduction of the vanadium-titanium-magnetite, enhanced the reaction fraction, and the continuously increasing additive amounts and reduction temperature led to the reaction fraction tending to increase slowly. The appropriate condition was 3% CaF 2 at 1300 • C, where reaction fraction could reach up to 95%.
The CaF 2 accelerated the diffusion and migration of the metal iron points, extended the grain size, eroded the surface of the vanadium-titanium-magnetite particles, destroyed the crystal structure of the non-metallic reduction products in the specimens, promoted the aggregation of the particles like titanomagnetite, ilmenite, and ferrous brookite, increased the lattice distortion of the reduced phase, and improved the reduction capacity.
Vanadium-titanium-magnetite with CaF 2 disrupted the equilibrium of the original Boudouard reaction which formed a new catalytic reaction, and the larger contact area between the carbon particles and the vanadium-titanium-magnetite updated the carbon layers to pile up the active points in the carbon surface.