Kinetics of Generation of Magnesium Vapor of Novel Magnesia-based Desulfurizer for External Desulfurization of Hot Metal

Abstract

To overcome the drawbacks of the high price and low efficiency of the desulfurizer which is used in spraying magnesium desulfurization, an in-situ desulfurization method that using the high temperature in hot metal to generate magnesium vapor was proposed. In this paper, the generation of magnesium vapor using new desulfurizer was investigated, and the conclusions are: with the temperature increasing, diffusion velocity of magnesium vapor and reaction rate were accelerated; with the increasing of flow rate of carrier gas, the reaction rate were accelerated; the apparent activation energy of the generation of magnesium vapor using new Magnesia-based desulfurizer was 69.57 kJ/mol, and the process was controlled by diffusion.

1. Introduction

Due to the increasing strict requirements on steel quality, controlling the content of sulfur which is the main impurity element has become more important.¹⁾ External desulfurization of hot metal has proven to be suitable technology for controlling sulfur content due to its low cost.

Magnesium, as a refining agent, is widely used for the treatment of various high quality steels owing to its strong affinity for non-metallic elements such as oxygen, sulfur, phosphorus and nitrogen.^2,3,4,5) While being desulfurized from molten iron with passivation magnesium particles, the equilibrium constant of desulfurization reaction is 2.06×10⁴, slag can not only reduce largly, temperature drop is not high, but also high quality steel can be obtained.^6,7,8,9) At present, there are insufficiencies on the hot iron pretreatment with magnesium.

Firstly, the equilibrium relation between the magnesium vapor pressure P_Mg and the temperature T is given by, lgP_Mg=−6802/T+9.993, where P_Mg is magnesium vapor pressure, Pa; and T is absolute temperature, K. The results show that when the temperature is between 1573 K and 1723 K, magnesium vapor pressure, P_Mg, varies between 0.405 Mpa and 0.993 MPa.¹⁰⁾ Therefore, it will produce an explosive reaction when magnesium is added into hot metal too quickly, which will lead to serious splash and significant loss of hot metal, and the utilization efficiency of magnesium will be decreased, so magnesium particles for desulfurization must be passivated.

Secondly, saturated vapor pressure of Mg is very large at high temperature during the desulfurization process, so the very large bubbles are inevitably formed and cannot be disintegrated and widely dispersed in the bath. In theory, should the sulfur content of 1 ton hot metal be decreased by 0.001%, 0.00758 kg passivation magnesium needs to be consumed, decreases 0.030%, 0.227 kg magnesium needs to be consumed. In fact, the sulfur content of 1 ton hot metal decreases 0.030%, about 0.5 kg passivation magnesium is consumed, so utilization efficiency of magnesium vapor is only 40%–50% in hot metal pretreatment.^11,12,13)

Thirdly, Mg, as an active metal, needs to be passivated when desulfurization and its price is very high. So the desulfurization price is very high when using magnesium as a desulfurizer in steelmaking process. For example, in North America, the price of magnesium for desulfurization is average 1.5 dollars per pound, magnesium mass for desulfurization is 300–500 g per 1 ton, and the price is 1.0–1.5 dollars, equivalent to 8.0–12.5 RMB.¹⁴⁾

For these reasons, in foreign countries, some institutes have begun to investigate new desulfurizers to replace the magnesium based desulfurizer or by other processes to make up deficiencies of magnesium;^{15,16,17,18,19)} but in China, studies about this field is rare. As the focus of this work, to make up the deficiencies of magnesium, our group elaborates on a new magnesia based desulfurizer to replace the magnesium based desulfurizer.

The key of MgO-based desulfurization is the generating of Magnesium vapor. The generation rate of magnesium vapor affects the desulfurization efficiency directly. In this paper, the impacts of temperature, ratio of MgO in reducing agent, carrier gas flow rate and diameter of pellets on reaction rate were investigated by the experimental research of macro kinetics, and the theoretical bases of MgO-based desulfurizer of external hot metal desulfurization were provided.

2. Preparation of MgO-based Desulfurizer and Experiments on Magnesium Vapor Generation

2.1. New MgO-based Desulfurizer

The dolomite phase composition in experiments was CaMg(CO₃)₂, the chemical composition (wt%) was: MgO, 21.73; CaO, 31.05; Al₂O₃, 0.16; SiO₂, 0.44; Na, 0.02; Fe, 0.07; K, 0.005. 75FeSi was chosen as the reducing agent, and its chemical composition (m/m%) is Si, 75.6; Al, 1.24; S, 0.091; C, 0.015. The binder is composite binder, which was mixed with organic and inorganic binder in ratios. XRD analysis results of Dolomite and ferrosilicon are shown in Figs. 1 and 2.

Fig. 1.

Result of XRD analysis about dolomite.

Fig. 2.

Result of XRD analysis about ferros-silicon (Si–Fe).

The material ratio of pellets, Dolomite:FeSi:Fluorite was 100:12:2. Raw materials were mixed and then shaped as green pellets by disc pelletizer, and 3% of total pellet mass composite binder was added continuously to assist the shaping of pellets. Then the pellets were gradient fired in Argon gas atmosphere, and finally the MgO-based desulfurizer was obtained.

Figure 3 is a schematic diagram of the experimental apparatus. First magnesium vapor was produced by reduction in the pellets of desulfurizer which happened in an immersion tube and was heated by heat conduction from the molten iron, at the meantime, magnesium vapor was injected into the molten iron by Ar carrier gas through the underpunch of the tube, finally, desulfurization reaction occurred and the concentration of sulfur fell. The desulfurizer was pressed to thin section of 16 mm diameter at 80 Mpa and put into the tube, the mass of molten iron was 5 kg, the temperature was 1300–1500°C, and the Ar flow rate was 0.06–0.12 m³/h. The desulfurization process was examined by taking a sample from the melt at appropriate time intervals for analyzing sulfur and carbon concentrations.

Fig. 3.

Device of experiment.

The reduction process was examined by taking all the desulfurizer from the tube at appropriate time for analyzing its Mg residue. The desulfurizer was digested by aqua regia, and its Mg residue was examined by ICP.

2.2. Thermodynamic Calculations

There are two reactions in the desulfurization process, one is the reduction in the desulfurizer in the pellets, also the reaction produced Mg vapor, the other is the reaction in-situ of Mg vapor and [S] in the hot metal. Because the second reaction is researched by many researchers, In this paper, the first reaction is investigated. During the desulfurization process, the main reactions of producing magnesium vapor are shown as following.

The overall reduction of light burned dolomite and ferrosilicon is given by   

$$2\mathrm{M}\mathrm{g}\mathrm{O}\cdot \mathrm{C}\mathrm{a}\mathrm{O}\mathrm{\hspace{0.17em} }(\mathrm{s})+\mathrm{S}\mathrm{i}\mathrm{\hspace{0.17em} }(\mathrm{s})=2\mathrm{M}\mathrm{g}\mathrm{\hspace{0.17em} }(\mathrm{g})+2\mathrm{C}\mathrm{a}\mathrm{O}\cdot \mathrm{S}\mathrm{i}{\mathrm{O}}_2\mathrm{\hspace{0.17em} }(\mathrm{s})$$(1)

  

$$\mathrm{\Delta }{G}^0=414\mathrm{\hspace{0.17em} }650-225.64T/\mathrm{K}\mathrm{\hspace{0.17em} }\text{(}\mathrm{J}/\mathrm{m}\mathrm{o}\mathrm{l}\text{)}$$(2)

It is seen from Fig. 4 that the beginning reaction temperature of light-burned dolomite and ferrosilicon is 1565°C. This temperature is under standard state. In the actual experiments, the desulfurizer put in an immersion tube were heated by heat conduction in the molten iron to produce magnesium vapor, at the same time, magnesium vapor was injected into the molten iron by Ar carrier gas through the underpunch in the tube. In the reaction atmosphere, there are two gases, one is Mg vapor, and the other one is Ar. The actual vapor pressure of magnesium is rather much less than 100 kPa, so the thermodynamic study is done on non-standard state as following.   

$$\mathrm{\Delta }G=\mathrm{\Delta }{G}^0+\mathrm{R}T\mathrm{\hspace{0.17em} }\mathrm{l}\mathrm{n}{K}_{\mathrm{p}}$$(3)

Fig. 4.

Variation with temperature of ΔG^Θ of reaction.

When the temperature is higher than 1107°C in the reaction, there is only Mg gas phase in the reaction. The reaction equilibrium constant follows as:   

$$K_{\mathrm{p}}=P^{\mathrm{c}}{}_{\mathrm{M}\mathrm{g}}/P^{\mathrm{\Theta }\mathrm{c}}$$(4)

  

$$\mathrm{l}\mathrm{n}K\mathrm{p}=\mathrm{c}\mathrm{l}\mathrm{n}\mathrm{\hspace{0.17em} }{P}_{\mathrm{M}\mathrm{g}}/P^{\mathrm{\Theta }}=-\mathrm{\Delta }{G}^0/\mathrm{R}T\mathrm{\hspace{0.17em} }\mathrm{\hspace{0.17em} }\mathrm{P}\mathrm{a}$$(5)

The variation with temperature of magnesium vapor P_Mg/P^Θ is shown as Fig. 5.

Fig. 5.

Variation with temperature of magnesium vapor P_Mg/P^Θ.

Figure 5 shows that in the molten iron temperature do not require a high degree of vacuum to make reducing reaction. When P_Mg=8 kPa, the reaction happens at 1300°C.

3. Results and Discussion

3.1. Impacts of Temperature on Magnesium Vapor Generation

Figure 6 shows the relationship between the magnesium content and time under varying temperature. The experimental condition was: reducing agent ratio: 1.0, carrier gas: Ar, gas flow rate: 0.1 m³/h.

Fig. 6.

The relationship between time and magnesium content at different temperature.

Figure 6 indicates that the reaction rate is fast in the initial stage of the reaction (0–16 min). Because in this stage, the reaction was in the intact pellet but reacting only on the surface of the whole pellet, while the main FeSi reacted with calcined dolomite was the free silicon. Thus the reaction rate was fast. In the middle stage of reaction, with the influence of the diffusion rate of magnesium vapor, reaction in the pellets was limited, which led to decreased reaction rate. In the later reaction stage, with the reaction interface promoting into interior, the escape distance and resistance of generated magnesium vapor were also increased. And calcined dolomite was mainly reacting with silicated iron, which led to reduced reaction rate. According to data above, the rising of temperature was not only beneficial to reaction improvement, but also improved the magnesium vapor diffusion and fastened reaction.

3.2. Impacts of Reducing Agent Ratio on Magnesium Vapor Generation

Results of different ratio of reducing agent were shown in Fig. 7. The experimental condition was:

Fig. 7.

The relationship between time and magnesium content at different ratio.

Temperature: 1400°C, carrier gas: Ar, gas flow rate: 0.1 m³/h, mass of FeSi: 1.0, 1.1 and 1.2 times the ratio of chemical reaction.

The graph shows that with the amount of FeSi was increasing, the reaction was expedited. Because the Si in 75FeSi mainly composed of 71.04% Si_free and 28.96% FeSi₂, and the reactivity of Si_free was better than that of FeSi₂. Si_free was abundant in the initial stage of the reaction in different ratio, which is the reason of the generation rates of magnesium were similar in the initial stage of the reaction. With the consumption of Si_free and the growth of the ratio of FeSi, the decline rate of proportion of Si_free in FeSi to generate unit mass Magnesium vapor decreased, and that was the reason of the reaction rate became more diverse in the later reaction stage with ratio growth.

3.3. Impacts of Gas Flow Rate on Magnesium Generation

Because the splash could cause by high carrier gas flow rate, several gas flow rates were chosen to make data have practical guidance significance, which were 0.06 m³/h, 0.08 m³/h, 0.1 m³/h and 0.12 m³/h. The results were shown in Fig. 8, and experimental condition was:

Fig. 8.

The relationship between time and magnesium content at different gas flow rate.

Temperature: 1400°C, carrier gas: Ar, reducing agent ratio: 1.0.

The figure indicates that the reaction rate was quite slow when gas flow was 0.06 m³/h and increases with gas flow rate growth. When gas glow rate reached 0.12 m³/h, reaction rate changed insignificantly. Because the growth of gas flow rate enhanced magnesium escaping, which led to partial pressure of magnesium vapor and improved the reaction. On the other hand, when further increased gas flow rate to a certain figure, the room temperature carrier gas decreased the reaction temperature and influenced reaction, which indicated in Fig. 8 when gas flow rate was 0.12 m³/h. Therefore, the gas flow rate growth can fasten reaction in a certain range, while will have little impact on reaction rate after exceeding a upper limit, which depend on the heating capacity of heating environment on carrier gas.

3.4. Magnesium Vapor Generation Mechanism

The reduction reaction experiments were conducted in 1300–1450°C, and the reduction rates of Magnesium in different time were obtained. Due to the reaction materials decreasing and control of reactants on reaction progress increasing in the later reaction period, only the results in early 16 min were used to calculate the activation energy.

The reduction rate can be calculated by the formula: b=(M1−M2)/(M1*a). In the formula, b stands for the magnesium reduction rate of the desulfurizer. M1 is the desulfurizer mass when the reaction begins. M2 is the desulfurizer mass when the reaction is finished. a is the initial magnesium content of the desulfurizer.

4 lines were obtained by fitting data in different temperature, whose slopes indicated the macroscopic characterization rate K in different temperature. The results were shown in Figs. 9 and 10.

Fig. 9.

Fitting results between reduction rate and reaction time at different temperature.

Fig. 10.

Relations of lgk and 1000/T.

According to Arrhenius equation: $lg\text{k}=-\frac{E_{\text{a}}}{2.\text{3}03\text{R}T}+\text{A}$ , k was apparent rate constant, yielded by slopes of fitting lines. R was gas constant, T was thermodynamic temperature (K), A was a constant. After obtaining slopes by plotting lgk and 1000/T, the apparent activation energy (E_a) of dolomite reduction can be calculated. In the range of 1300–1450°C, E_a was 69.57 kJ/mol.

3.5. Discussion on Reaction Mechanism

Raw materials in reaction temperature mainly composed of solid dolomite and liquid silicon, and Magnesium vapor was generated, thus this reaction was solid-liquid-gas system. Normally, multi-phase reaction includes the followingsteps: reactant diffusing to reaction interface, reactant reacting on reaction interface and resultant leaving reaction interface.

In this research, the reduction reaction in pellets had the following features: (1) Though being totally mixed and heated in high temperature, reducing agent and dolomite powder interacted well, and liquid state FeSi in high temperature also improved reaction. (2) This reaction was determined as endothermic reaction by kinetic analysis; however, due to the improvement of blowing gas, the heat transfer rate was enhanced remarkably, thus the control of exterior heat transfer on reaction was insignificant. (3) The Magnesium vapor escaping impacted on the Magnesium partial pressure in reaction interface, according to Arrhenius equation, the escaping of Magnesium vapor had noticeable effects on macroscopic reaction.

In the reduction experiments, blocks at different response time are taken out and cut. Due to the different color of raw material (ferrosilicon is black, lightly burned dolomite is faint yellow, the general color of material is gray as shown), obvious interface stratification can be found in the section layer (as shown in Fig. 11). The outside of interface is brown, inside of interface is gray, and deepest color of interface is black. Along with the extending of time, interface layer is pushed to the inside of material in order, until the interface layer disappears at 30 min of reaction time. This indicates the reaction is an outside-in process.

Fig. 11.

Photos of reactant cross-section in different reaction time.

In the figure 5 min, 10 min, 15 min, 20 min (especially the Figure 15 min and 20 min), we can see that when the reaction time reach the preset time, the desulfurizer in the blowing tube is removed, reaction in the material block is basically terminated. Mg vapor did not quickly leave reaction interface but slowly leave, and when finally leave, there is large amounts of oxygen in the environment that caused the magnesium vapor react with oxygen on the surface of material block to generate white magnesium oxide as shown in figure. White matter was not found significantly in the interior of the desulfurizer. It also suggests that when experiment is end, under the condition of no carrier gas, while desulfurization still maintains high temperature in a short period of time, but the reaction is basically terminated, and there is only magnesium vapor. Mg steam spreads quickly at high temperature, and finally magnesium oxide appears on the surface of material block.

Reaction rate of pellets was closely related with Magnesium vapor escaping, which was caused by the difference of Magnesium partial pressure on radial distribution. During the reaction, a relative low Magnesium partial pressure atmosphere was created by carrier gas. Magnesium vapor was generated in such a condition, and then carried out by carrier gas. In the early reaction period, pellets reacted quickly due to the low Magnesium partial pressure. In the inner layer of pellet, Magnesium partial pressure was relative high due to the small amount of diffused carrier gas and bad generated Magnesium vapor ventilation, which leaded to a low reaction rate. After that, changes of Magnesium partial pressure caused the reaction rate difference on radial distribution, and reaction interface was formed. Reactants inside the interface reacted slowly due to the impacts of Magnesium vapor pressure, while reactants out of interface were almost totally reacted, and the highest reaction rate appeared on the interface.

4. Conclusion

In this study, we propose a novel desulfurizer. The thermodynamics calculations and experiments on Magnesium vapor generation of the new desulfurizer were investigated. The following can be concluded as:

(1)　Through thermodynamic calculation, the reaction temperature of magnesium vapor production is 1565°C in the standard state and decreased with the decreasing partial pressure of magnesium vapor P_Mg. When P_Mg=8 kPa, the reaction happens at 1300°C.

(2)　Kinetics experiments on Magnesium vapor generation shows that with the increase of temperature, not only the reaction but also the Magnesium vapor escaping were improved. With the increasing of carrier gas flow rate, reaction rate also grew. When it gas flow rate reached 0.12 m3/h, reaction rate remained stable.

(3)　According to Arrhenius equation, the apparent activation energy of Magnesium vapor generation was calculated, at 69.57 kJ/mol, which indicated the reaction was a diffusion control process.

Acknowledgments

This research was financially supported by the National Basic Research Program of China (2013CB632606-2).

References

• 1)   B.  Sun,  X.  Chen and  C.  Zhang: J. Iron Steel Res., 15 (2003), 1.
• 2)   J.  Ruan and  F.  Zhang: Wuhan Univ. (Natural Science), 11 (2004), 48.
• 3)   S.  Yamaguchi,  T.  Uemura,  H.  Nashiwa and  H.  Sugita: Ironmaking Steelmaking, 5 (1977), 276.
• 4)   D. A. R.  Kay,  S. V.  Subramanian and  R. V.  Kumar: Can. Inst. Min.Metall., 17 (1986), 43.
• 5)   G. A.  Irons and  C.  Celik: Ironmaking Steelmaking, 19 (1992), 136.
• 6)   H.  Sandberg: Ironmaking Steelmaking, 5 (1977), 280.
• 7)   G. A.  Irons and  R. L.  Guthrie: Ironmaking Steelmaking, 3 (1981), 114.
• 8)   J.  Yang,  K.  Okumura and  M.  Kuwabara: ISIJ Int., 41 (2001), 965.
• 9)   Y.  Liu,  M.  Sano,  T. A.  Zhang,  Q.  Wang and  J. C.  He: ISIJ Int., 49 (2009), 17.
• 10)   Y.  Liu,  T. A.  Zhang,  M.  Sano,  Q.  Wang,  X. D.  Ren and  J. C.  He: Trans. Nonferrous Met. Soc. China, 21 (2011), 1896.
• 11)   Y.  Liu,  M.  Sano,  Q.  Wang,  T. A.  Zhang and  J. C.  He: CAMP-ISIJ, 20 (2007), 48.
• 12)   Y.  Liu,  T. A.  Zhang and  Q.  Wang: Indust. Furnace, 29 (2007), 1.
• 13)   X. H.  Sun,  X. M.  Xia and  C. J. V.  Long: Iron Steel, 42 (2007), 27.
• 14)   V. M.  Safonov,  A. N. V.  Smirnov and  K. E.  Pismarev: Russ. Metall. (Metally), 12 (2010), 1087.
• 15)   D. G.  Zhao,  J. S.  Li and  S. H.  Wang: Adv. Mater. Res., 568 (2012), 17.
• 16)   R.  Babak: Int. J. Adv. Manuf. Technol., 46 (2010), 191.
• 17)   N. A.  Felipe,  F. G.  Felipe and  A. S. T.  Jorge: Metalurgiae Materiais, 65 (2012), 233.
• 18)   G. P.  Liu,  Y.  Deng,  L. P.  Wu and  L.  Xiong: Steelmaking, 30 (2014), 1.
• 19)   C. C.  Xu,  A. J.  Wang,  L. H.  Liao and  Y. F.  Xiong: Wisco Technol., 30 (2014), 1.