Reaction Behaviour between Cerium Ferroalloy and Molten Steel during Rare Earth Treatment in the Ultra-low Carbon Al-killed Steel

Abstract

In this work, the melting process and reaction behaviour of cerium ferroalloy in the liquid ultra-low carbon interstitial free steel was investigated through a special designed hot crucible experiment using standard metallographic techniques including SEM, EDS, XRF, and XRD in combination with the thermodynamic software FactSage 7.2. The evolution mechanism of cerium-containing phases in the cerium ferroalloy during the melting process was proposed. The cerium-containing phases stretched and migrated along the direction of heat flow, in further the arms of the network structures became thinner and finally dissolved into the molten steel. It was concluded that the thinner of network structures’ arms, the higher of the Ce content in cerium-containing phases. During the melting process, the cerium content in the cerium-enriched phase was increased from 43.9 wt% in the matrix of the cerium ferroalloy to 90.44 wt% in the reaction zone, and the average width of the network arms was decreased from 18 µm to 2 µm accordingly. The main reaction products in the interface among cerium ferroalloy, molten steel, and slag were Ce₂O₃ and CeAlO₃.

1. Introduction

Rare earth elements including Sc, Y, and 15 lanthanides have been applied successfully in many fields such as metallurgy, chemistry, and manufacturing engineering.^1,2,3,4) Due to their similar atomic structure and unique electronic layer structure, rare earth elements possessed great advantages in modifying inclusions, increasing purification and alloying of steel, which could provide better mechanical properties and corrosion resistance to the final products.^5,6,7,8,9) As the increase in demand for high quality steel and environment-friendly metallurgical technology in China, more and more attention has been paid to the application of rare earth in steel.^{10,11,12,13,14,15,16)} It has been proved that that with addition of La in non-oriented electrical steel, the favorable texture of (100) and (110) increases, and the thermal conductivity and the wear-resisting property of medium manganese steel can be raised by the interaction between rare earth and alloying elements.^17,18) S. K. Kwon et al.¹⁹⁾ found that the initial Mn(Cr)-silicate inclusions in stainless steel transformed to Al₂O₃ rich inclusions by Al addition, and then Ce₂O₃ was enriched by Ce addition due to local enrichment of Ce around Al₂O₃ particles, resulting in the formation of AlCeO₃-type inclusions. Huang et al.²⁰⁾ studied the modification mechanism of cerium on inclusions in H13 steel and revealed that the cerium can effectively cause the transformation of inclusions from MgO·Al₂O₃ to CeAlO₃, Ce₂O₃, and Ce₂O₂S as the cerium content is increased from 0 to 0.03 wt%. Li et al.²¹⁾ have concluded that Ce₂O₂S and CeS were the main reaction products in saw wire steel with the cerium content below 0.0277 wt%, and when the content of cerium was more than 0.0389 wt%, the transformation mechanisms of both Ce–S–O–P and Ce–P–O inclusions were based on double-layer structures. Yang et al.²²⁾ have concluded that the irregular Al₂O₃ and MnS can be modified into regular RE-type inclusions, and the impact toughness can be improved accordingly with a proper content of RE elements, for example around 0.14 wt% Ce, in high-carbon chromium bearing steel. Kim et al.²³⁾ studied the resistance to pitting corrosion of a super duplex stainless steel with rare earth metals addition and found that the formation of (Mn, Cr, Si, Al, Ce) oxides and (Mn, Cr, Si, Ce) oxides could improve the resistance to pitting corrosion.

In order to study the application of rare earth elements in a more comprehensive and systematical way, some researchers^{24,25,26,27,28,29)} have developed the thermodynamic database for the binary alloy system RE-X (RE: Ce, La, Gd, Nd, Sm, Lu, Ho, Er; X: Ag, Cr, Mn, V, Zn, Bi). However, few detailed results were reported on an actual melting process and the reaction behavior between cerium ferroalloy and molten steel. In this paper, the melting process and reaction behaviour of cerium ferroalloy in the liquid ultra-low carbon Al-killed steel was investigated through a hot crucible experiment. The formation and transformation of Ce-type inclusions and the reaction mechanism between cerium ferroalloy, molten steel and slag were studied based on the hot crucible experiments and thermodynamic calculations.

2. Experimental Procedure

The experiments were designed to investigate the melting and reaction behavior between cerium ferroalloy and molten steel, as shown in Fig. 1. A bomb-shape sampler with inner diameter of 35 mm and height of 95 mm was chosen to obtain and fill the liquid steel whose chemical composition is listed in Table 1. There are two fixtures at the heights of 30 mm and 65 mm respectively from the bottom inside the sampler to ensure that cerium ferroalloy blocks can be fixed in the sampler. During the experiment, the molten steel of 1600°C covered by a thin layer of slag with chemical composition in Table 2 was poured into the sampler. As filling of molten steel, the liquid level grew and reached the positions of cerium ferroalloy blocks. Two reaction interfaces between molten steel and cerium ferroalloy block, and between slag and cerium ferroalloy block were then formed accordingly in the aforementioned two fixtures. These operations ensured that the cerium ferroalloy block reacted with the molten steel and slag respectively, where the difference in the reaction process of the cerium ferroalloy with them could be investigated afterwards and the same held for the melting process. The sampler was covered by a cap and cooled down to room temperature under natural condition. The metallographic specimens were prepared for further analysis with SEM/EDS. The thermodynamic software FactSage 7.2 was employed to predict the phase transition and distribution under given conditions in order to better study the melting and reaction process of the cerium ferroalloy.

Fig. 1.

Preparation of the reaction interfaces between the cerium ferroalloy and molten steel or slag. (Online version in color.)

Table 1. Chemical composition of Al-killed steel/wt%.

C	Si	Mn	P	S	Al	Ti	T.O	N
0.0013	0.0050	0.1300	0.0110	0.0050	0.0450	0.0600	0.0022	0.0025

Table 2. Chemical composition of slag/wt%.

SiO2	Al2O3	CaO	MgO	FeOx
8	32	45	7	8

The initial compositions and phases of the cerium ferroalloy were characterized in Fig. 2 using XRD, XRF and SEM. The average Ce and Fe contents were 9.51 wt% and 78.60 wt%, respectively. The main crystal phase of cerium in the cerium ferroalloy is Ce₂Fe₁₇. According to the SEM results, three characteristic regions marked with 1, 2 and 3 in the cerium ferroalloy were found as shown in Fig. 2(c), and it was noticed that the lighter of the phase’s color is, the higher of the cerium content. The highest cerium content referring to the point 3 reached 43.9 wt%. Figure 3 shows that the elements of Fe and Ce contents present opposite distribution; and the cerium-containing phase presents a network structure. It is concluded that the cerium content in the cerium ferroalloy is not uniform.

Fig. 2.

Characterization of cerium ferroalloy: (a) Chemical composition of cerium ferroalloy based on XRF; (b) Characterization in the cerium ferroalloy based on XRD; (c) Micro-structures in cerium ferroalloy by SEM. (Online version in color.)

Fig. 3.

Mapping of elements Ce and Fe in the cerium ferroalloy.

3. Results and Discussion

3.1. Melting Process of Cerium Ferroalloy in the Molten Steel

The melting process of cerium ferroalloy in the molten steel is described in Fig. 4. The molten steel poured into the sampler reacts with the cerium ferroalloy to form a reaction zone. The cerium-enriched phases in the cerium ferroalloy and reaction zone changed gradually during the melting process. In the initial cerium ferroalloy, the cerium-enriched phase displayed a bright white network microstructure where the average width of arms of network structure is 18 μm, as indicated by red double arrows. During the melting process, the network microstructures stretched and migrated along the direction of heat flow. The cerium content in the cerium-enriched phase was increased from 43.9 wt% in the matrix of the cerium ferroalloy to 90.44 wt% in the reaction zone. The average width of the network arms was decreased from 18 μm to 2 μm accordingly. In the reaction interface, the concentration of Ce and O is very high, which indicates that the reaction between Ce and Al occurs in this interface (Fig. 5).

Fig. 4.

The dynamic changes of the cerium ferroalloy during melting process: (a) Initial phases of cerium ferroalloy; (b) Phases in the reaction zone between cerium ferroalloy and molten steel; (c) Compositions of different areas marked in the reaction zone.

Fig. 5.

Elements distribution in reaction zone between molten steel and cerium ferroalloy, (a) reaction interface, (b) element mapping of Ce, (c) element mapping of Fe, (d) element mapping of Al, (e) element mapping of O.

As shown in Fig. 5(a), the upper left part of the area is IF steel matrix, and the lower right part of the area is cerium ferroalloy matrix. It can tell from Figs. 5(b) and 5(c) that the overlap band of elements with high concentration Ce and low concentration Fe is the boundary of reaction interface. The uniformity distribution of Al in the reaction zone as shown in Fig. 5(d) indicates the diffusion of Al is limited during this process. The concentration of Ce (see Fig. 5(b)) and O (see Fig. 5(e)) is enriched on the reaction interface with a good coupling correspondence, which proves that the Ce₂O₃ is formed through the reaction between Ce and O on the interface.

The Ce–Fe binary phase diagram calculated by FactSage 7.2 using the thermodynamic database of RE-X binary alloy^{24,25,26,27,28,29,30)} is employed to reveal the evolution process of phases during the melting process, as shown in Fig. 6. When the initial content of Ce in the enrichment zone of cerium ferroalloy is 43.9 wt%, the Ce₂Fe₁₇ and CeFe₂ should coexist at temperatures below 1203 K. It is noticed that CeFe₂ is not observed by XRD in original cerium ferroalloy. This implies that a phase transition from Ce₂Fe₁₇ to CeFe₂ could occur at the early stage of cerium ferroalloy melting, which causes the increase of the cerium content in the cerium-enriched phase. When the temperature reaches 1203 K, the CeFe₂ begins to gradually transform into Ce₂Fe₁₇ and specific liquid phase with high Ce content which results in the decrease of the cerium content in the remaining cerium-enriched solid phases. The variation in cerium content associated with the transformations of the cerium-enriched phases is in agreement with the results in Fig. 4(c).

Fig. 6.

Calculation results of phase diagram in Ce–Fe binary systems. (Online version in color.)

Combination of Fig. 4, the Fig. 7 describes the evolution mechanism of cerium-containing phases in the cerium ferroalloy during the melting process. When the cerium ferroalloy contacts with the molten steel, a reaction interface between cerium ferroalloy and molten steel is formed firstly (Stage I). Then, a heat-affected zone is formed and grows as the heat flow diffuses from the molten steel to the cerium ferroalloy along the interface (Stage II). The cerium-enriched phases of cerium ferroalloy inside the heat affected zone stretches and migrates along the direction of heat flow; during this process, the arms of the network structures of the cerium-enriched phases become thinner. A part of arms of the network structures in the cerium-containing phases are dissolved into the molten steel (Stage III to Stage V). The heat transfer from the molten steel to cerium ferroalloy makes a thin solidified layer next to the heat affected zone, and the Ce dissolved in the molten steel is transformed into new cerium-enriched phase during the solidification (Stage VI). During the whole process, the Ce content in the cerium-enriched phase increases from initial content of 43.9 wt% to the peak of 90.44 wt% then has a slightly decrease to 88.14 wt%. According to the phase diagram in Fig. 6, it is inferred based on the initial composition that the cerium-containing phase is composed of Ce₂Fe₁₇ and CeFe₂, although the dominant phase is characterized as Ce₂Fe₁₇ by XRD in Fig. 2(b). When the temperature reaches above 1203°C, the CeFe₂ disappears and a liquid with higher cerium concentration first appears; then the Ce₂Fe₁₇ phase completely disappears by further increasing the temperature to more than 1500°C. Actually, the cerium-containing phase with Ce concentration above 88 wt% is formed during the liquid cerium-enriched solution undergoing a slow natural cooling process, that promotes the formation of γ-phase of cerium at the temperature of 873°C and the micro-segregation of Ce in solid phased.

Fig. 7.

Evolution mechanism of cerium-containing phases in the cerium ferroalloy during the melting process. (Online version in color.)

3.2. Reaction Behavior between Cerium Ferroalloy and the Molten Steel

Figure 8 shows the typical reaction products in the reaction zone between cerium ferroalloy and molten steel. Three main products including Al₂O₃, CeAlO₃ and Ce₂O₃ were observed in the reaction zone. The cerium-containing inclusions were brighter in color than Al₂O₃ based inclusions. According to the SEM results, the cerium-containing phases including Ce₂Fe₁₇, CeFe₂ and γ-phase of cerium participate in reaction with the oxygen and aluminum in the steel at different stages, which leads to the formation of CeAlO₃ and Ce₂O₃. In order to evaluate the formation condition of cerium-containing inclusions, the equilibrium module in FactSage 7.2 is applied. In the reaction zone, the dissolution process of cerium-containing phase occurs gradually. Therefore, the amount of Ce involved in the reaction is limited and a local equilibrium is defined accordingly. For example, 1 g molten steel is in equilibrium with a gradual increase of Ce and the results are shown in Fig. 9. It is seen that the phases calculated can cover the aforementioned main phases, i.e. Al₂O₃, CeAlO₃ and Ce₂O₃, but another main phase CeAl₁₁O₁₈ predicted is not shown in Fig. 8. In addition, the activity of CeS set as a dormant phase is given in Fig. 9. When the phase activity is greater than 1, the phase starts to be a stable phase. As such, the phase CeS should form at higher Ce contents, but it is not observed. This may result from the kinetics barrier.

Fig. 8.

Typical reaction products in the reaction zone between cerium ferroalloy and molten steel. (Online version in color.)

Fig. 9.

Distribution of precipitates in the reaction zone between cerium ferroalloy and molten steel. (Online version in color.)

It is interesting to note that the distance of the main products from the reaction interface seems to follow the sequence: Al₂O₃ < CeAlO₃ < Ce₂O₃. This is consistent with the predictions by FactSage 7.2. It is seen from Fig. 9 that the first phase to form is Al₂O₃, where the presence of Ce is not necessary. In other words, Al₂O₃ inclusions may exist before the addition of the cerium ferroalloy. With dissolution of cerium-containing phases in molten steel within the reaction zone, the phase CeAlO₃ forms, followed by Ce₂O₃. Meanwhile, the oxygen activity is decreased in the molten steel. When the Ce content reaches a threshold value, the stable inclusions are only Ce₂O₃. This indicates that Ce has a higher priority to consume the oxygen in the molten steel than Al. However, the inclusions Al₂O₃ and CeAlO₃ formed previously do not be melted again in practice due to the relatively high melting point.

Furthermore, the different precipitates Ce₂O₃ and CeAlO₃ are proposed to result from the reactions given in Table 3.^31,32) It should be mentioned that 1% (mass percent) dilute solution was chosen as the standard state for activity calculation of solutes, while for the activity calculation of inclusions the standard state is the pure substance. The corresponding interaction coefficients at 1873 K are listed in Table 4.^{31,32,33,34,35)} Different activity coefficients f_i (i = Ce, O, Al) were obtained by using the Wagner’s model, and then the Gibbs free energy at 1873 K was calculated by Eq. (1). The calculation results in Fig. 10 show that the Ce₂O₃ and CeAlO₃ can be formed under present conditions.   

$$\begin{array}{l}\mathrm{\Delta }\text{G}=\mathrm{\Delta }{\text{G}}^{\theta }\\ \hspace{1em}\hspace{1em}+\text{RT}\mathrm{\hspace{0.17em} }\text{ln}{\displaystyle \frac{\text{1}}{{\text{(}{f}_{\text{A}}\text{wt[\%A])}}^{\text{x}}\cdot {\text{(}{f}_{\text{B}}\text{wt[\%B])}}^{\text{y}}\cdot {\text{(}{f}_{\text{C}}\text{wt[\%C])}}^{\text{z}}}}\end{array}$$(1)

Table 3. Standard Gibbs free energy of possible reactions.

Reaction Equations in Molten Steel	ΔGθ = A + B × T, KJ mol−1
	A	B
2[Ce] + 3[O] = Ce2O3(s)	−1431.090	0.360
[Ce] + [Al] + 3[O] = CeAlO3(s)	−1366.460	0.364

Note: The 1 mass pct infinite dilute solution was chosen as the standard state for activity calculation of solutes and the pure substance was chosen as that for activity calculation of inclusions.

Table 4. Interaction coefficient of elements.

Elements	C	Si	Mn	P	S	Al	Ti	O	N	Ce
Ce	−0.077	—	0.13	1.77	−10.32	−2.58	−3.62	−106	−6.612	0.0069
O	−0.421	−0.066	−0.021	0.07	−0.133	−1.17	−1.8	−0.20	−0.12	−64
Al	0.091	0.0056	—	0.033	0.041	0.045	0.004	−6.6	−0.01	−0.52

Note: 1% (mass percent) dilute solution was chosen as the standard state for activity calculation of solutes.

Fig. 10.

Variation in the Gibbs free energy of Ce₂O₃ and CeAlO₃ with respect to different content of Ce. (Online version in color.)

3.3. Reaction Behavior between Cerium Ferroalloy and Slag

As aforementioned, the reaction interface between slag and cerium ferroalloy block was obtained by filling the molten steel covered by a thin layer of slag into the sampler where the liquid-slag contacted with cerium ferroalloy block that fixed in the sampler at the height of 65 mm, then the sample was cooled down to room temperature under natural condition. Actually, various interfaces between cerium ferroalloy and slag are formed during this process as shown in Fig. 11. A common feature of these interfaces is that a layer of Ce₂O₃ with thickness of 5 to 10 μm is formed on the side where the slag layer is in contact with the cerium ferroalloy; meanwhile a thin FeO layer exists between the Ce₂O₃ layer and matrix, and the FeO is a relatively low content as compared to Ce₂O₃ observed in the contacting area. Because the liquid slag is composed of SiO₂, Al₂O₃, CaO, MgO and FeO, as a separate phase, the layer of FeO does not seems to have strong correlation with the liquid slag. It is inferred that the layer of FeO is formed by the air oxidation of the reaction zone during the natural cooling process due to the volume shrinkage which causes gaps between cerium ferroalloy and Ce₂O₃ layer. The liquid Fe in the reaction zone could be from either slag reduction products or ferroalloy reaction products. This is indicated, for example, by a drop in FeO content of final slag compared to that of initial liquid-slag. It can be seen from Figs. 11(b) and 11(c) that separate phase FeO is not observed if no interface gap between cerium ferroalloy and Ce₂O₃ layer; on the contrary, the separate phase FeO is formed between the interface gap, as shown in Figs. 11(a) and 11(d).

Fig. 11.

Different reaction interfaces between cerium ferroalloy and slag. (Online version in color.)

Figure 12 presents the typical characteristics of reaction interface between cerium ferroalloy and slag. It can be seen from Fig. 12(a) that the bright gray Ce₂O₃ layer distributes around the black slag. Combining the aforementioned analysis, the reaction mechanism of cerium ferroalloy with slag layer is proposed. Firstly, the temperature of cerium ferroalloy increases after contacting the liquid slag, then the Ce in cerium ferroalloy diffuses along the direction of heat flow, and starts to react with the liquid slag. The oxygen potential is higher in liquid slag than in cerium ferroalloy, that drives the free oxygen transferring from the liquid slag to reaction interface, finally the bright gray Ce₂O₃ layer is formed next to the slag. It is very critical for present reaction of Ce and O to ensure the sufficient transfer of cerium and oxygen from the bulk to reaction interface. As shown in Fig. 12(b), the elements of Ca, Mg, Al, Si mainly distribute in the slag bulk; while element Fe has a relative uniform distribution in both phases of slag and Ce₂O₃ but maintains a low concentration. Is also can be found that the element Ce does not appear in the final slag which implies that the Ce has been transformed to Ce₂O₃ before diffusing into the slag phase. It is expected that the formation of Ce₂O₃ cause a barrier to slag transfer to the active reaction interface, and therefore the oxygen source required for the reaction with Ce is limited. Based on the aforementioned reaction process, the molten slag next to the reaction interface is considered as active and can be involved in the reaction. As such, the reaction taking place in the interface is simulated with a local equilibrium by FactSage using the FToxid and FTmisc databases. That is, the cerium ferroalloy of 1 g, taken as an example, is in equilibrium with a gradual increase of molten slag. As shown in Fig. 13, the results calculated show Ce₂O₃ as a major phase which is in agreement with the experimental observation. However, the observed separate FeO phase should not directly result from the reaction on the interface, since FeO is only a component of the solid solution phase MeO in Fig. 13. It should be mentioned that the MeO#1 or MeO#2 indicate the presence of a miscibility gap. The presence of another phase, i.e. AlCeO₃ is noticed, although it is not observed in the reaction zone. This is dependent on the kinetics of oxygen transfer. A. Katsumata et al.³⁴⁾ reported that the composition of deoxidation products took only compounds of Al₂O₃, CeAl₁₁O₁₈, CeAlO₃ and Ce₂O₃ using both Ce and Al for deoxidation, and their relevant thermodynamic calculation also indirectly supported present results. However, we did not observe the phases of AlCeO₃ and CeAl₁₁O₁₈ in present condition. S. K. Kwon et al.³⁵⁾ considered that the cerium reacts with alumina to form Ce₂O₃, and the Ce₂O₃ can react with alumina in a sequence resulting in a formation of CeAlO₃, and then CeAlO₃ continues to react with alumina to form CeAl₁₁O₁₈. The absence of AlCeO₃ and CeAl₁₁O₁₈ implies that there is not enough alumina to support the further reactions. In experiment by S. K. Kwon et al., the Ce reacted with alumina refractory, and the alumina supposed as unity is not the limitation step which promotes the formation of AlCeO₃ and CeAl₁₁O₁₈; In present experiment, high Ce content and low Al content in cerium ferroalloy determine the liquid slag tending to react with Ce preferentially, meanwhile the further reaction cannot occur carried out before oxygen is transferred from slag to reaction interface.

Fig. 12.

Elements mapping of reaction interface between cerium ferroalloy and slag.

Fig. 13.

Distribution of precipitates in the reaction zone between cerium ferroalloy and molten slag.

It is proved that the cerium-containing phases are reactive with oxidizing slag and oxides inclusions. As thus, the prerequisite conditions to get a high recovery rate of cerium during the melting process are to maintain low slag oxidation and less oxides in the molten steel. The conditions of adding cerium ferroalloy have a strong influence on the recovery rate of cerium.

4. Conclusions

The melting process and reactive behavior between cerium ferroalloy and molten steel were investigated through a special designed hot crucible experiment, where the mechanism for the formation and transformation of Ce-type inclusions is illustrated. The main findings can be summarized as follows:

(1) The cerium content was relevant with the cerium-containing phases. It is noticed that the lighter of the phase’s color is, the higher of the cerium content. The average cerium content in ferroalloy was 9.51 wt%, and the main cerium-containing phase in cerium ferroalloy is Ce₂Fe₁₇ with highest cerium content of 43.9 wt%.

(2) The evolution mechanism of cerium-containing phases in the cerium ferroalloy during the melting process was proposed. The cerium-containing phases stretched and migrated along the direction of heat flow, in further the arms of the network structures became thinner and finally dissolved into the molten steel. It was concluded that the thinner of network structures’ arms, the higher of the Ce content in cerium-containing phases.

(3) A layer of Ce₂O₃ was formed around the slag at the reaction zone of the cerium ferroalloy with slag layer, whereas the Ce₂O₃ and CeAlO₃ could be formed easily in the liquid steel under present conditions. The mechanism for the formation of aforementioned different precipitates in the reaction zone is proposed through a comprehensive analysis on the results from both thermodynamic calculations and melting experiments.

Acknowledgements

This work was supported by the Fundamental Research Funds for the Central Universities (FRF-AT-18-002) and Open Project of State Key Laboratory of Advanced Special Steel, Shanghai Key Laboratory of Advanced Ferro-metallurgy, Shanghai University (SKLASS 2020-02). The authors wish to express their gratitude to the foundation for providing financial support.

References

• 1)   P. J.  Zhou,  J. J.  Yu,  X. F.  Sun,  H. R.  Guan,  X. M.  He and  Z. Q.  Hu: Mater. Sci. Eng. A, 551 (2012), 236. https://doi.org/10.1016/j.msea.2012.04.117
• 2)   X. L.  Li,  S. M.  He,  X. M.  Zhou,  Y.  Zou,  Z. G.  Li,  A. G.  Li and  X. H.  Yu: Mater. Charact., 95 (2014), 171. https://doi.org/10.1016/j.matchar.2014.06.016
• 3)   Y.  Meng,  C. L.  Yan,  X. P.  Yang and  X. H.  Ju: ISIJ Int., 60 (2020), 534. https://doi.org/10.2355/isijinternational.ISIJINT-2019-428
• 4)   E. T.  Stromme,  H. B.  Henderson,  Z. C.  Sims,  M. S.  Kesler,  D.  Weiss,  R. T.  Ott,  F. Q.  Meng,  S.  Kassoumeh,  J.  Evangelista,  G.  Begley and  O.  Rios: JOM, 70 (2018), 866. https://doi.org/10.1007/s11837-018-2861-9
• 5)   S.  Imashuku and  K.  Wagatsuma: Metall. Mater. Trans. B, 51 (2020), 79. https://doi.org/10.1007/s11663-019-01732-8
• 6)   X.  Li,  Z. H.  Jiang,  X.  Geng,  M. J.  Chen and  L. Z.  Peng: ISIJ Int., 59 (2019), 1552. https://doi.org/10.2355/isijinternational.ISIJINT-2019-094
• 7)   H.  Torkamani,  S.  Raygan,  C. G.  Mateo,  J.  Rassizadehghani,  Y.  Palizdar and  D.  San-Martin: Met. Mater. Int., 24 (2018), 773. https://doi.org/10.1007/s12540-018-0084-9
• 8)   S.  Jeon,  S.  Kim,  M.  Choi,  J.  Kim,  K.  Kim and  Y.  Park: Corros. Sci., 75 (2013), 367. https://doi.org/10.1016/j.corsci.2013.06.020
• 9)   H.  Wang,  Y. P.  Bao,  J. G.  Zhi,  C. Y.  Duan,  S.  Gao and  M.  Wang: ISIJ Int., 61 (2021), 657. https://doi.org/10.2355/isijinternational.ISIJINT-2020-053
• 10)   Z. Y.  Cai,  B.  Song,  L. F.  Li,  Z.  Liu and  X. K.  Cui: Int. J. Miner. Metall. Mater., 26 (2019), 565. https://doi.org/10.1007/s12613-019-1765-1
• 11)   S.  Song and  S.  Sun: Steel Res. Int., 87 (2016), 1435. https://doi.org/10.1002/srin.201500425
• 12)   Q. F.  Guan,  Q.  Jiang,  J. R.  Fang and  H.  Jiang: ISIJ Int., 43 (2003), 784. https://doi.org/10.2355/isijinternational.43.784
• 13)   A. E.  Somers,  B. R. W.  Hinton,  C.  de Bruin-Dickason,  G. B.  Deacon,  P. C.  Junk and  M.  Forsyth: Corros. Sci., 139 (2018), 430. https://doi.org/10.1016/j.corsci.2018.05.017
• 14)   H.  Torkamani,  S.  Raygan,  C.  Garcia Mateo,  J.  Rassizadehghani,  J.  Vivas,  Y.  Palizdar and  D.  San-Martin: Metals, 7 (2017), 377. https://doi.org/10.3390/met7090377
• 15)   S. H.  Song,  Y.  Xu,  X.  Chen and  X.  Jiang: J. Rare Earth., 34 (2016), 1062. https://doi.org/10.1016/S1002-0721(16)60135-7
• 16)   L. J.  Wang,  Y. Q.  Liu,  Q.  Wang and  K. C.  Chou: ISIJ Int., 55 (2015), 970. https://doi.org/10.2355/isijinternational.55.970
• 17)   L. F.  Fan,  R.  Zhu,  J. Z.  He and  B.  Lu: ISIJ Int., 58 (2018), 2348. https://doi.org/10.2355/isijinternational.ISIJINT-2018-392
• 18)   S. C.  Chen,  H. X.  Ye and  X. Q.  Lin: Int. J. Miner. Metall. Mater., 24 (2017), 670. https://doi.org/10.1007/s12613-017-1449-7
• 19)   S. K.  Kwon,  Y. M.  Kong and  J. H.  Park: Met. Mater. Int., 20 (2014), 959. https://doi.org/10.1007/s12540-014-5022-x
• 20)   Y.  Huang,  G. G.  Cheng,  S. J.  Li and  W. X.  Dai: Steel Res. Int., 89 (2018), 1800371. https://doi.org/10.1002/srin.201800371
• 21)   Y.  Li,  M.  Sun,  Z. H.  Jiang,  C. Y.  Chen,  K.  Chen,  X. F.  Huang,  S.  Sun and  H. B.  Li: Metals, 9 (2019), 54. https://doi.org/10.3390/met9010054
• 22)   C. Y.  Yang,  Y. K.  Luan,  D. Z.  Li and  Y. Y.  Li: J. Mater. Sci. Technol., 35 (2019), 1298. https://doi.org/10.1016/j.jmst.2019.01.015
• 23)   S.  Kim,  S.  Jeon,  I.  Lee and  Y.  Park: Corros. Sci., 52 (2010), 1897. https://doi.org/10.1016/j.corsci.2010.02.043
• 24)   Z.  Li,  X. P.  Su,  F. C.  Yin and  C. T.  Chen: J. Alloy. Compd., 299 (2000), 195. https://doi.org/10.1016/S0925-8388(99)00686-6
• 25)   S.  Gao,  M.  Wang,  J. L.  Guo,  H.  Wang,  J. G.  Zhi and  Y. P.  Bao: Steel Res. Int., 90 (2019), 1900194. https://doi.org/10.1002/srin.201900194
• 26)   F. C.  Yin,  M. H.  Huang,  X. P.  Su,  P.  Zhang,  Z.  Li and  Y.  Shi: J. Alloy. Compd., 334 (2002), 154. https://doi.org/10.1016/S0925-8388(01)01757-1
• 27)   F. C.  Yin,  X. P.  Su,  Z.  Li,  P.  Zhang,  M. H.  Huang and  Y.  Shi: J. Alloy. Compd., 307 (2000), 202. https://doi.org/10.1016/S0925-8388(00)00812-4
• 28)   H. Y.  Qi,  G. X.  Huang,  R.  Liu,  K.  Zhang,  L.  Liu and  Z. P.  Jin: J. Alloy. Compd., 497 (2010), 336. https://doi.org/10.1016/S0925-8388(00)00812-4
• 29)   C. P.  Wang,  X.  Chen,  X. J.  Liu,  F. S.  Pan and  K.  Ishida: J. Alloy. Compd., 458 (2008), 166. https://doi.org/10.1016/j.jallcom.2007.03.120
• 30)   Z. M.  Cao,  X. Y.  Song and  Z. Y.  Qiao: Chin. J. Rare Met., 32 (2008), 216 (in Chinese). https://doi.org/10.13373/j.cnki.cjrm.2008.02.026
• 31)   Y. D.  Li,  C. J.  Liu,  T. S.  Zhang,  M. F.  Jiang and  C.  Peng: Metall. Mater. Trans. B, 48 (2017), 956. https://doi.org/10.1007/s11663-016-0873-9
• 32)   A.  Vahed and  D. A. R.  Kay: Metall. Trans. B, 7 (1976), 375. https://doi.org/10.1007/BF02652708
• 33)   H.  Li,  Y. C.  Yu,  X.  Ren,  S. H.  Zhang and  S. B.  Wang: J. Iron Steel Res. Int., 24 (2017), 925. https://doi.org/10.1016/S1006-706X(17)30135-8
• 34)   A.  Katsumata and  H.  Todoroki: Iron Steelmaker, 29 (2002), 51.
• 35)   S. K.  Kwon,  J. S.  Park and  J. H.  Park: ISIJ Int., 55 (2015), 2589. https://doi.org/10.2355/isijinternational.ISIJINT-2015-125