Migration Behavior of the MgO and its Influence on the Reduction of Fe₃O₄–MgO Sinter

Abstract

To investigate the migration behavior of MgO in oxidization-reduction of the vanadium titano-magnetite (VTM) ore and the influence of its content on the reduction, the solid solution was prepared by annealing MgO and Fe₃O₄ and its oxidized products were isothermally reduced under H₂–N₂ atmosphere. It was found that the sinter with high MgO content showed faster reduction rate due to the formation of pores and the increase of active sites caused by the diffusion of Mg²⁺ into Fe₃O₄. In contrast, the as-oxidized sinter with high MgO content showed the slowest reduction rate due to the formation of high content of (MgO)_x·(FeO)_1−x (x=0.239–0.77) in the reduction process. MgO migrated from the solid solution and combined with partial Fe₂O₃ to form MgFe₂O₄ in the oxidization stage. In the reduction stage, MgO migrated outward as the oxidized sinter was reduced to Fe₃O₄ and FeO, while migrated inwards in the stage of Fe formation. MgO content in per mole formed FeO decreased as the sinter was reduced to FeO, while increased in solid solution during the Fe generation stage.

1. Introduction

Vanadium titano-magnetite (VTM) ore is a kind of multielements-coexistent mineral which mainly contains V, Ti and Fe. The reserves vanadium and titanium in china account for about 11.6% and 35.17% of that all over the world, respectively.¹⁾ It is an important source of vanadium and titanium. However, Ti and Fe elements are symbiosis closely to each other in the ore, and V is as isomorphism hosting in the lattice of titano-magnetite, which restricts its comprehensive utilization.²⁾ Therefore, recent years, many scholars have carried out extensive research on how to make comprehensive use of Fe, V and Ti elements, such as pre-reduction-electric furnace process,^3,4) molten salt roasting process,⁵⁾ reduction roasting- magnetic separation,⁶⁾ etc.

Because of the low processing costs and high recovery rates of valuable elements, the pre-reduction-electric furnace smelting process is the most promising processes. Obviously, reduction is an essential procedure for this technology. However, it is difficult to be reduced due to its compact structure and complex phases.⁷⁾ To solve this problem, many researchers adopted pre-oxidization to strengthen the reduction processes by controlling the ore phases and structure.^7,8,9,10,11) Studies have shown that dense structures can become porous, which was beneficial for the mass transfer and reaction.^9,12) From the view of phase transformation, FeTiO₃ and Fe₂TiO₄ could convert to Fe₂O₃ and TiO₂ by pre-oxidization, which lowered the requirement for the reduction potential.¹⁰⁾ Moreover, Tang et al.¹³⁾ studied the coupled effect of TiO₂, V₂O₅, and Cr₂O₃ during the oxidization process. Actually, the content of gangue components is very high in vanadium-titanium magnetite ore. MgO naturally exists in the form of solid solution in most iron ores, in which the content of MgO is more than 2.5 wt.% in some iron ores,¹⁴⁾ even as high as 7.8% in the Panzhihua vanadious titanomagnetite Ore.¹⁵⁾ Obviously, the existence of MgO also influences the reduction performance of the natural iron ores.¹⁶⁾ The reduction of iron oxide is a continuous and gradual process, Fe and O elements migrate constantly during the whole reduction process. Li¹⁷⁾ found that MgO was able to be dissolved in iron oxides (FeO_n) and formed solid solution [(1−x)MgO.Fe_xO]O.Fe₂O₃, the migration rate of Fe²⁺ was decreased. Consequently, the reduction rate declined.

Our previous research⁹⁾ found that MgO could be enriched in the process of oxidation. But, the formed phases and the migration behavior during the oxidization-reduction process are not clear. Moreover, the documents mentioned previous on VTM oxidization have not discussed the migration behavior of MgO. Therefore, in order to solve this problem, this paper adopted Fe₃O₄ and MgO as raw material to prepare sinter, analyzed the migration of MgO in the course of oxidation-reduction and its influence on the reduction properties of the sinter.

2. Experimental

2.1. Materials

High purity gases (H₂, CO₂, CO and N₂) were supplied by Beijing Hua Yuan Gas Chemical Industry Co., Ltd. Analytical grade of Fe₃O₄ and MgO was supplied by Beijing Chemical Reagent Company.

2.2. Experimental Procedures

Analytical grade Fe₃O₄ and different amount of MgO (A.R.) were uniformly mixed by the slurry method. Then, the slurry was dried and pressed into discs with compressive stress of 30 MPa. The pellets were roasted a tubular furnace (as shown in Fig. 1) at 1373 K under CO/CO₂ atmosphere (CO/CO₂ = 10/90; gas velocity 1.0 L/min) for 5 hours. Afterward, the pellets were crushed into powder and sieved to 74–450 μm.

Fig. 1.

Experimental setup used for the sinter preparation and reduction. 1-gas cylinder; 2-shutoff valve; 3-mass flow controller; 4-reactor; 5-electric resistance furnace; 6-thermocouple; and 7-temperature controller.

Pre-oxidation was performed in a furnace at 1223 K for 4 h. Reduction reaction was also carried out with 90%N₂-10%H₂ gas mixture (1.0 L/min) at 1073 K for different times in a tubular furnace. After a desired annealing time, the reactor was removed from the hot zone of the furnace under the protection of N₂ and quenched directly by spraying water.

2.3. Analysis and Characterization

Phase analysis was detected on PANalytical X’pert diffractometer with Cu Ka radiation (k=1.5408 Ǻ). Morphology and elements distribution of the products were analyzed on a scanning electron microscopy (SEM, JSM-6700F)/energy dispersive spectrometry (EDS, Noran System six).

It is assumed that the weight decrease was solely caused by the loss of oxygen, the reduction degree (X) of the iron oxides was calculated as follows:   

$$X=\frac{m_{actrual}}{m_{theoretical}}$$(1)

Where, m_actrual was the actual mass loss of the reactant after the reduction; m_theoretical was the theoretical mass loss of the sample.

3. Results and Discussion

3.1. Analysis of the Sinters and its Oxidized Samples

Figure 2 illustrated the XRD patterns of the annealed samples with different MgO content. Only Fe₃O₄ phase could be detected when the pure Fe₃O₄ disc was treated under the 10%CO-90%CO₂ atmosphere, indicating Fe₃O₄ has not been reduced or oxidized. However, the final phases of the sinter contained Fe₃O₄, MgFe₂O₄ and (MgO)_0.219·(FeO)_0.781 when MgO was added into Fe₃O₄. Based on the EDS analysis results (Table 1), MgO evenly distributed in Fe₃O₄ as its content was low, and the detected n_Mg/n_Fe value was close to the theoretical n_Mg/n_Fe value (point 1 and point 2). However, two phases existed in the sinter when MgO content was high, which belong to MgFe₂O₄ (point 3) and (MgO)_x·(FeO)_1−x (x=0.219) (point 4), respectively. The results were in good agreement with the XRD results. Therefore, the reaction could be described as follows:   

$$\begin{array}{c}\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3·(1-\mathrm{x})\mathrm{F}\mathrm{e}\mathrm{O}+(1+\mathrm{x})\mathrm{M}\mathrm{g}\mathrm{O}=\mathrm{M}\mathrm{g}\mathrm{O}·\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3\\ +{(\mathrm{M}\mathrm{g}\mathrm{O})}_{\mathrm{x}}·{(\mathrm{F}\mathrm{e}\mathrm{O})}_{1-\mathrm{x}}\mathrm{\hspace{0.17em} }(\mathrm{x}=0.219)\end{array}$$(2)

Fig. 2.

XRD patterns of the sinter with different MgO content.

Table 1. EDS results of the sinters with different MgO content.

Points	Elements (atm.%)			(nMg/nFe)detected	(nMg/nFe)theoretical
	Fe	Mg	O
1	42.33	3.01	54.66	0.07	0.08
2	44.06	9.45	46.50	0.21	0.21
3	36.73	14.69	48.59	0.40	0.34
4	40.48	11.49	48.03	0.28	0.34
5	29.60	15.40	55.01	0.52	–
6	48.85	0.92	50.23	0.01	–

The results demonstrated that the addition of MgO promoted the formation of magnoferrite (MgFe₂O₄) and solid solution ((MgO)_x·(FeO)_1−x, x=0.219).

The diffraction peaks shifted to the higher angle as MgO content increased from 0% to 15%, and the interplanar crystal space (d) decreased gradually, indicating the lattice parameters of Fe₃O₄ have been changed due to the diffusion of MgO. It is well known that Fe₃O₄ can be expressed as Fe₂O₃·FeO or FeO_x, the ionic radii of Mg²⁺ (0.66 Å) and Fe²⁺ (0.78 Å) are similar with each other. Therefore, Mg²⁺ could easily get into the lattice of magnetite by solid-phase diffusion, replace some of the Fe²⁺ in the magnetite lattice, and form continuous solid solution and magnesio spinels (MgFe₂O₄). As a result, the dense internal structure became porous one under the effect of Mg²⁺, as illustrated in Fig. 3. Moreover, the size and amount of the pores obviously increased with the increase of MgO content. According to the XRD results, no MgO diffraction peaks could be detected, which indicated that all MgO has combined with Fe₃O₄ to form MgFe₂O₄ and (MgO)_0.432·(FeO)_0.568.

Fig. 3.

Cross section images of the sinter with different MgO content (a) 0% MgO; (b) 5% MgO; (c) 10% MgO; (d) 15% MgO.

The magnetite (FeO·Fe₂O₃) has a spinel structure as no MgO was introduced, however, it could be replaced by a mixed spinel of type (Fe_(1−x)Mg_x)O·Fe₂O₃ with the addition of MgO. According to the report,¹⁸⁾ the course of their migration was FeO·Fe₂O₃→(Fe,Mg)O·Fe₂O₃→(Mg,Fe)O·Fe₂O₃→MgO·Fe₂O₃. Therefore, the diffraction peaks of MgFe₂O₄ gradually became stronger as MgO content increased.

Figure 4 showed the XRD pattern and the cross section image of the oxidized sinter. According to the XRD result (Fig. 4 (a)), Fe₃O₄ and (MgO)_0.432·(FeO)_0.568 phases disappeared when the sinter was treated under air atmosphere, and were replaced by Fe₂O₃ and MgFe₂O₄. The results indicated that Fe₂O₃ formed by the oxidization of FeO and reacted with MgO to form MgFe₂O₄, which was well coincided with the reported result.¹⁷⁾ Compared with the sinter with 5% MgO, two phases apparently separated and enriched each other in the oxidized sample. Based on the EDS results (Table 1), the dark phase (point 5) and light phase (point 6) belong to MgFe₂O₄ and Fe₂O₃, respectively. The internal structure of the sinter with 5% MgO was compact, while the oxidized sample showed relatively loose (Fig. 4(b)), which might be attributed to the migration of MgO and the transformation of cubic lattice to hexagonal lattice during the oxidization. In addition, the EDS results demonstrated that almost all MgO has been migrated from the solid solution and combined with Fe₂O₃ to form MgFe₂O₄ in the oxidization process. Therefore, MgFe₂O₄ content in the oxidized sinters was 24.21% and 72.86% as the amount of MgO was 5% and 15%, respectively. Obviously, the reduction process of oxidation products mainly involves the reduction of Fe₂O₃ and MgFe₂O₄.

Fig. 4.

XRD pattern and cross section image of the oxidized sample (5%MgO).

3.2. Reduction Performance of the Samples

To investigate the influence of MgO content on the reduction performance of the original sinter and the oxidized one, both samples were reduced under the 10%H₂-90%N₂ atmosphere, as illustrated in Fig. 5. The results indicated that MgO content significantly influenced the reduction rate of the samples. For the untreated samples, the reduction rate significantly increased with the increase of MgO content. However, Bahgat¹⁹⁾ observed the opposite result that the reduction rate of wüstite doped with MgO decreased due to the declining porosity volume of the sample. As stated previously, the dense internal structure became porous one under the diffusion effect of the Mg²⁺. Obviously, the formation of the pores enhanced the diffusion of the reducing gas and increased the contact areas. Therefore, the formation of the pores promoted the reduction and offset the adverse effect caused by MgO.

Fig. 5.

Reduction curves of the (a) sintered sample and the (b) as-oxidized sample.

The phases change during the reduction process was illustrated in Fig. 6, the diffraction peaks of the Fe₃O₄ gradually disappeared, accompanied with the formation of (MgO)_x·(FeO)_1−x (x=0.231–0.593) phase. Obviously, in the course of reduction, the amount of the Fe₃O₄ phase decreased, while the amount of (MgO)_x·(FeO)_1−x (x=0.231–0.593) phase increased. The x value increased with prolonging of the reduction time, which attributed to the gradually decrease of FeO content. Therefore, the reaction process could be described as follows:   

$$\mathrm{F}{\mathrm{e}}_3{\mathrm{O}}_4\to \mathrm{F}{\mathrm{e}}_{1-\mathrm{x}}\mathrm{O}\to \mathrm{F}\mathrm{e}$$(3)

  

$${(\mathrm{M}\mathrm{g}\mathrm{O})}_{\mathrm{x}}\cdot {(\mathrm{F}\mathrm{e}\mathrm{O})}_{1-\mathrm{x}}\to \mathrm{F}\mathrm{e}$$(4)

  

$$\mathrm{M}\mathrm{g}\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_4\to {(\mathrm{M}\mathrm{g}\mathrm{O})}_{\mathrm{x}}\cdot {(\mathrm{F}\mathrm{e}\mathrm{O})}_{1-\mathrm{x}}\to \mathrm{F}\mathrm{e}$$(5)

Fig. 6.

Process of phases change during the reduction of the sample with 15% MgO (a) sintered sample; (b) as-oxidized sample.

Fig. 7.

XRD patterns of the oxidized sample reduced to (a) Fe₃O₄; (b) FeO.

It was worth mentioning that the sintered sample with 15% MgO could not be completely reduced and the maximum reduction degree was only about 80% after reacted for 2 h. This phenomenon might be attributed to the formation of (MgO)_x·(FeO)_1−x. The results indicated that (MgO)_x·(FeO)_1−x was difficult to be reduced. Some documents pointed out that the existence of MgO retarded the reduction due to its barrier effect.^19,20,21) Actually, the barrier effect of MgO addressed in the literatures on the reduction is the formation of the solid solution ((MgO)_x·(FeO)_1−x) which is difficult to be reduced.

Compared with the reduction property of the original sinters, the reduction rate of the oxidized samples reflected significant difference. The oxidized sample with high MgO content showed slower reduction rate than that with low MgO content. During the reduction process, the diffraction peaks (Fig. 6(b)) of MgFe₂O₄ disappeared quickly accompanied with the appearance of the (MgO)_x·(FeO)_1−x phase. The results demonstrated that MgFe₂O₄ had the same reduction performance as Fe₂O₃. As stated previously, more and larger pores were generated and almost all MgO combined with Fe₂O₃ to form MgFe₂O₄ during the oxidization process. Therefore, the gap of the reduction rate of the oxidized samples with different MgO content obviously decreased. However, it was worth noting that the reduction rate of the sample with 5% MgO was much faster than the sample without MgO addition, which was contrary to the results reported by El-Geassy²²⁾ who found that 1% MgO only enhanced the initial reduction rate then hindered the reduction. The promotion effect resulting from doping with MgO was attributed to both an increase in porosity and an increase in available active sites owing to the intrusion of foreign cations in the Fe₂O₃ lattice.²³⁾ Therefore, the porosity and available active sites significantly increased as more MgO was introduced, which was beneficial for the reduction.

(MgO)_x·(FeO)_1−x phase still could be detected as prolonged the reduction time, which proved that (MgO)_x·(FeO)_1−x phase formed by the reduction of MgFe₂O₄ was unbeneficial for the further reduction. Compared with the untreated one, the diffraction peaks of (MgO)_x·(FeO)_1−x phase generated by the reduction of the oxidized sample disappeared faster, indicating that the oxidized sample was easily reduced. The higher MgO content was, the greater difference between the actual and theoretical reduction appeared. By analyzing the phase evolution processes, the diffraction peaks of MgFe₂O₄ rapidly disappeared, indicating MgFe₂O₄ could be easily reduced to Fe₃O₄ and further to FeO. Therefore, the formation of (MgO)_x·(FeO)_1−x was the key material to limit the further reduction.

3.3. Migration Behavior of MgO during the Reduction Process

When the oxidized sinter was reduced to Fe₃O₄, the diffraction peaks belong to MgFe₂O₄ disappeared quickly accompanied with the appearance of MgO as prolonged the reduction time. By analyzing the composition of the different phases of the reduced sample, it could be found that the detected n_Mg/n_Fe value (point 1 in Table 2) in the dark area was lower than that in the oxidized one, the detected n_Mg/n_Fe value (point 2 in Table 2) in the light area was higher than that in the oxidized one. The results indicated that partial MgO spread out from MgFe₂O₄ and combined with Fe₃O₄ to form solid solution. Therefore, the results demonstrated that MgFe₂O₄ could be reduced to Fe₃O₄ according to the reaction (6).   

$$3\mathrm{M}\mathrm{g}\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_4+\mathrm{C}\mathrm{O}=3\mathrm{M}\mathrm{g}\mathrm{O}+2\mathrm{F}{\mathrm{e}}_3{\mathrm{O}}_4+\mathrm{C}{\mathrm{O}}_2$$(6)

Table 2. EDS results of the sample reduced to Fe₃O₄.

Points	Elements (atm.%)			nMg/nFe	Reduction degree (%)
	Fe	Mg	O
7	33.81	12.86	53.34	0.38	11.28
8	43.25	2.5	54.24	0.06	13.41

As the reduction proceeded, the clear phase interface gradually disappeared, which meant that both Fe₂O₃ and MgFe₂O₄ were gradually reduced. As could be seen in Fig. 8, the porosity obviously increased as the oxygen element lost and the iron and magnesium elements migrated. Due to the limited binding between MgO and Fe₃O₄, there still existed two separated phases. Compared with the oxidized sample, the two phases became irregular and the light color phase gradually decreased in the reduced sample. By analyzing the distribution of Mg element (as could be seen in Table 2), the content of Mg element contained in the light area increased, while decreased in the dark phase. Moreover, the as-oxidized sample reduced for 10 min contained (Fe_(1−x)Mg_x)O·Fe₂O₃ (x=0.64) and MgO. The above detected results illustrated that the Mg element migrated from the reduced MgFe₂O₄ and partially combined with Fe₃O₄ again. The reduction degree also could be calculated based on following formula:   

$$(\mathrm{n}{\mathrm{O}}_1-\mathrm{n}{\mathrm{O}}_{\mathrm{x}})\mathrm{\hspace{0.17em} }*100/\mathrm{n}{\mathrm{O}}_1$$

Fig. 8.

Texture of the oxidized sample with 10% MgO reduced to Fe₃O₄ stage for different time (a) 3 min; (b) 10 min.

Where, nO₁ was the O element content in the oxidized sample; nO_x was the O element content in the sample reduced for different times.

When the oxidized sample was reduced to Fe_1−xO under the 50%CO-50%CO₂ atmosphere, the diffraction peaks of MgFe₂O₄ rapidly disappeared less than 1 min due to the high reduction potential. Only (MgO)_0.239·(FeO)_0.761 could be detected after reacted for 10 min, the results indicated that MgFe₂O₄ could be easily reduced to Fe₃O₄ and further reduced to (MgO)_0.239·(FeO)_0.761. Correspondingly, the two clear phases also gradually disappeared and formed a homogeneous phase, as shown in Fig. 9. According to the EDS results (Table 3), it was obvious that MgO content gradually increased in the loose area, while decreased in the dense area as the reduction time extended. When the sample was reduced for 30 min, the n_Mg/n_Fe (0.19) value in the gray area (point 5 in Table 3) was close to the average value (0.22). Moreover, the homogeneous phase was formed with some big holes in the size of 3–5 μm. The result demonstrated that part of MgO migrated out from the reduced MgFe₂O₄ phase and combined with other FeO to form (MgO)_0.239·(FeO)_0.781.

Fig. 9.

Texture of the oxidized sample with 10% MgO reduced to Fe_1−xO stage for different times (a) 1 min; (b) 3 min; (c) 30 min.

Table 3. EDS results of the samples reduced to FeO and Fe.

Points	Elements (atm.%)			nMg/nFe	Reduction degree (%)
	Fe	Mg	O
9	38.01	14.26	47.72	0.37	29.78
10	52.02	0.51	47.47	0.01	32.38
11	40.95	11.81	47.25	0.29	33.04
12	46.36	5.92	47.72	0.13	36.16
13	48.23	9.09	42.68	0.19	46.35
14	42.85	15.68	41.48	0.36	45.22
15	36.48	32.52	30.99	0.89	60.90
16	86.87	5.70	7.43	–	98.26

Though MgO had the similar migration behavior in the two reduction stages, the amount of migrated MgO in the FeO reduction stage was much more than that in the Fe₃O₄ reduction stage. It is well known that the radius of Mg²⁺ (dMg²⁺=0.66 Å) is smaller than that of Fe²⁺ (dFe²⁺=0.78 Å).²⁴⁾ Therefore, according to the crystallography theory, Mg²⁺ can easily get into the lattice of the FeO, replace some Fe²⁺ ions and completely form continuous isomorphism.²⁵⁾ Moreover, MgO was completely mutually soluble in FeO at 1023 K.¹⁷⁾ Consequently, MgO has totally combined with FeO and formed (MgO)_x·(FeO)_1−x phase. Therefore, the amount of migrated MgO in the Fe_1−xO stage was much higher than that in the Fe₃O₄ stage and the value of x decreased as Fe₃O₄ was totally reduced to FeO. It was worth to noting that the internal structure became porous one due to the remove of oxygen and the migration of MgO. From the view point of reduction, the porous structure was beneficial for the further reduction.

Therefore, under the 50%CO-50%CO₂ atmosphere, the oxidized sample was reduced by the reactions as follows:   

$$\mathrm{M}\mathrm{g}\mathrm{O}·\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3+\mathrm{C}\mathrm{O}=(\mathrm{M}\mathrm{g}\mathrm{O})·(\mathrm{F}\mathrm{e}\mathrm{O})+\mathrm{C}{\mathrm{O}}_2$$(7)

  

$$\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3+\mathrm{C}\mathrm{O}=\mathrm{F}\mathrm{e}\mathrm{O}+\mathrm{C}{\mathrm{O}}_2$$(8)

  

$$\mathrm{x}\mathrm{M}\mathrm{g}\mathrm{O}+(1-\mathrm{x})\mathrm{F}\mathrm{e}\mathrm{O}={(\mathrm{M}\mathrm{g}\mathrm{O})}_{\mathrm{x}}\cdot {(\mathrm{F}\mathrm{e}\mathrm{O})}_{1-\mathrm{x}}$$(9)

When the oxidized sinter was reduced to Fe, (MgO)_x·(FeO)_1−x (x=0.77) could still be detected though the reduction degree was as high as 92%, as shown in Fig. 6(b). The results proved that the formation of solid solution helped to stabilize the crystal lattice, which had a negative effect on the reduction. The morphology evolution (Fig. 10) showed that the iron oxide was reduced from the edge of the pores, which indicated that the iron oxides reduction followed the un-reacted shrinking core model.²⁶⁾ Because the existence of Mg²⁺ ions reduced the migration rate of Fe²⁺ during the reduction process, Fe²⁺ at the interface was difficult to migrate into crystal lattice, Fe²⁺ accumulated at the interfaces and formed lots of metallic iron cores under the reduction gas, and then the iron cores grew and formed a dense metallic iron layer. As a result, the core-shell structure coated by Fe appeared when the metallization degree was high. During the reduction process, MgO exsoluted from the solid solution as the oxygen element was removed from the iron oxide. However, no peaks attributed to MgO could be detected by XRD, indicating MgO did not exist as a separate phase. Therefore, MgO separated from the reduced Fe and moved toward the unreduced FeO, and combined with them to form solid solution. As a result, the solid solution was covered by the Fe layer, which further hindered its reduction.

Fig. 10.

Back scatter images of the sample reduced to metallic Fe (a) 10 min; (b) 60 min.

Comprehensively analyzing the experimental results at oxidization and each reduction stage, the migration behavior of MgO throughout the oxidization-reduction process could be described as follows:

Figure 11 clearly illustrated the migration behavior of MgO during the oxidization-reduction process. The migration behavior of the whole oxidization-reduction process could be divided into two types: MgO reacted with Fe₂O₃ to form MgFe₂O₄, MgO migrated into FeO by diffusion and substitution of the lattice to form solid solution. The first type mainly happened in the sintering and oxidization process, in which the migration process of MgO was inclined to chemical process. This conclusion could be explained that the Mg²⁺ could not diffuse into the lattice of the Fe³⁺ because the Mg²⁺ is much bigger than the Fe³⁺. Another type could be found in the sintering and reduction process. Mg²⁺ can easily diffuse into the Fe²⁺ lattice for that the Mg²⁺ radius is smaller than that of the Fe²⁺. However, the amount of MgO diffused into the FeO was significant different in the Fe₃O₄ and reduced products. For instance, both the composition of the solid solution formed in the sintering process and the sample reduced to FeO were (MgO)_0.231·(FeO)_0.769. Obviously, the content of MgO in per mole FeO generated during the reduction was much lower than that in the Fe₃O₄. It is well known that the formed Fe_1−xO during the reduction is non-stoichiometric, which can cause the different lattice parameter and the defect sites. Moreover, the amount of FeO gradually decreased as the sinter was reduced to FeO. Therefore, the decrease of MgO content in per unit FeO could be interpreted that the lattice parameter and the defect sites of the FeO has changed, and the amount of FeO gradually reduced. In the third stage, the content of MgO in per mole FeO gradually increased as the solid solution was reduced to Fe.

Fig. 11.

Scheme of MgO migration behavior during the oxidization-reduction of the sinter.

In summary, whether the first or the second reduction stage, its essence of the solid solution formation was generated by the Mg²⁺ totally or partially substituted the Fe²⁺. Obviously, the migration behavior of MgO was closely related to the content and distribution of the Fe_1−xO. Though MgO combined with Fe₂O₃ by reaction with each other, the content was quite limited due to the larger radius of Mg²⁺. Therefore, the direction and amount of the migrated MgO changed with the change of the Fe_1−xO. As described by Du,²⁷⁾ to effectively prevent the adhesion during the fluidization reduction, MgO should be added into system as the iron ore powder was reduced to FeO. However, MgO easily reacted with FeO to form solid solution that hindered the further reduction. Moreover, a compact shell (Fe)-core ((MgO)_x·(FeO)_1−x) structure formed due to the opposite migration direction of the Fe²⁺ and Mg²⁺, which further hindered the reduction of the gas into the internal reductive solid solution. Therefore, no matter which stage to add MgO, it will eventually react with FeO to form solid solution (MgO)_x·(FeO)_1−x.

4. Conclusions

The migration behavior of MgO in the oxidization-reduction processes of Fe₃O₄–MgO sinter, and its content influence on the reduction have been investigated. The results indicated that the reduction rate of the sinter increased with the increase of MgO content because the porosity and available active sites increased some pores were formed by the diffusion of Mg²⁺ into the Fe₃O₄. The oxidized sinter with high MgO content showed slower reduction rate than that with low MgO content due to the formation of high content of (MgO)_0.219·(FeO)_0.781 in the reduction process. During the oxidization process, MgO migrated from the solid solution and combined with some Fe₂O₃ to form MgFe₂O₄. MgO exhibited different migration behavior in the different reduction stages, for instance, MgO migrated outward as the oxidized sinter was reduced to Fe₃O₄ and FeO, while migrated inward in the FeO reduced to Fe stage. MgO content in per mole FeO generated during the reduction decreased as the sinter was reduced to FeO, while increased in the stage of the Fe generation.

Acknowledgments

The authors wish to gratefully acknowledge the financial support from the National Natural Science Foundation of China (Grant No. 21736010 and No. 51404228).

References

• 1)   P.  Tan,  H. P.  Hu and  L.  Zhang: Trans. Nonferr. Met. Soc. China, 21 (2011), 1414.
• 2)   Y. L.  Sui,  Y. F.  Guo,  A. Y.  Travyanov,  T.  Jiang,  C.  Feng and  G. Z.  Qiu: Rare Met., 35 (2016), 954.
• 3)   W.  Li,  G. Q.  Fu,  M. S.  Chu and  M. Y.  Zhu: Steel Res. Int., 87 (2016), 1.
• 4)   S. P.  Yang,  J.  Wang,  D. U.  Xin and  J.  Liu: Min. Metall. Eng., 34 (2014), 87.
• 5)   D.  Chen,  L.  Zhao,  Y.  Liu,  T.  Qi,  J.  Wang and  L.  Wang: J. Hazard. Mater., 244–245 (2013), 588.
• 6)   M. S.  Jena,  H. K.  Tripathy,  J. K.  Mohanty,  J. N.  Mohanty,  S. K.  Das and  P. S. R.  Reddy: Sep. Sci. Technol., 50 (2015), 1221.
• 7)   S. S.  Liu,  Y. F.  Guo,  G. Z.  Qiu and  T.  Jiang: Rare Met., 3 (2014), 1.
• 8)   D. S.  Chen,  B.  Song,  L. N.  Wang,  T.  Qi,  Y.  Wang and  W. J.  Wang: Miner. Eng., 24 (2011), 864.
• 9)   F.  Pan,  Z.  Du,  M.  Zhang and  H.  Sun: ISIJ Int., 57 (2017), 413.
• 10)   H.  Sun,  A. A.  Adetoro,  F.  Pan,  Z.  Wang and  Q.  Zhu: Metall. Mater. Trans. B, 48B (2017), 1898.
• 11)   W.  Li,  G. Q.  Fu,  M. S.  Chu and  M. Y.  Zhu: Steel Res. Int., 87 (2016), 1.
• 12)   Y. L.  Sui,  Y. F.  Guo,  T.  Jiang and  G. Z.  Qiu: J. Alloy. Compd., 706 (2017), 546.
• 13)   J.  Tang,  M.  Chu,  C.  Feng,  F.  Li,  Y.  Tang and  Z.  Liu: ISIJ Int., 56 (2016), 1342.
• 14)   T.  Hu,  X.  Lv,  C.  Bai, Z. Lun and G. Qiu: Metall. Mater. Trans. B, 44 (2013), 252.
• 15)   F.  Pan,  Q. S.  Zhu,  D. U.  Zhan and  H. Y.  Sun: J. Iron Steel Res. Int., 23 (2016), 1160.
• 16)   H.  Sun,  A. A.  Adetor,  Z.  Wang,  F.  Pan and  L.  Li: ISIJ Int., 56 (2016), 936.
• 17)   G. H.  Li,  Z. K.  Tang,  Y. B.  Zhang,  Z. X.  Cui and  T.  Jiang: Ironmaking Steelmaking, 37 (2010), 393.
• 18)   S. Y.  Jin,  A. A.  Bhattacharyya and  C. A.  Radlowski: Ind. Eng. Chem. Res., 30 (1991), 1444.
• 19)   M.  Bahgat,  K. S. A.  Halim,  M. I.  Nasr and  A. H. A.  El-Geassy: Steel Res. Int., 78 (2007), 443.
• 20)   S. K.  Gupta,  V.  Rajakumar and  P.  Grieveson: Metall. Trans. B, 18 (1987), 713.
• 21)   S. K.  Gupta,  V.  Rajakumar and  P.  Grieveson: Metall. Trans. B, 20 (1989), 735.
• 22)   A. A.  El-Geassy: Ironmaking Steelmaking, 26 (1999), 41.
• 23)   A.  Ünal and  A. V.  Bradshaw: Metall. Trans. B, 14 (1983), 743.
• 24)   U. S.  Yadav,  B. D.  Pandey,  B. K.  Das and  D. N.  Jena: Ironmaking Steelmaking, 29 (2002), 91.
• 25)   M.  Zhou,  S. T.  Yang,  T.  Jiang and  X. X.  Xue: Ironmaking Steelmaking, 42 (2015), 217.
• 26)   S. S.  Liu,  Y. F.  Guo,  G. Z.  Qiu,  T.  Jiang and  F.  Chen: Trans. Nonferr. Met. Soc. China, 24 (2014), 3372.
• 27)   Z.  Du,  Q.  Zhu,  Y.  Yang,  C.  Fan,  F.  Pan,  H.  Sun and  Z.  Xie: Steel Res. Int., 87 (2016), 1.