Stability of Cementite under CO–CO₂–H₂ Gas Mixture at 1200 K

Abstract

Suppression of CO₂ discharged from iron and steelmaking companies is an example of the biggest issues for the protection of global environment and sustainable growth of steelmaking industry. One of the efforts made to decrease the emission of CO₂ in ironmaking process is blowing of hydrogen gas into blast furnace. Hydrogen gas can reduce iron oxide and form harmless H₂O. Cementite (Fe₃C) may be formed by introduction of hydrogen into blast furnace and play an important role on carburization and smelting behavior of reduced iron.

In the present work, Fe₃C sample was held at 1200 K under various CO–CO₂–H₂ gas mixtures to clarify the stability of Fe₃C phase. It was confirmed that metastable Fe₃C phase will decompose into Fe and C, and FeO will form under high CO₂ partial pressure. Also, it was found that existence of H₂ will suppress decomposition of Fe₃C. It was confirmed that CO₂ gas will be continuously converted into CO by C formed by decomposition of Fe₃C.

1. Introduction

Suppression of CO₂ emission is one of the largest issues for the protection of global environment and also sustainable growth of steelmaking industry. Efforts are made to solve this problem such as Japanese national project COURSE50.¹⁾ In this project, hydrogen-rich reformed coke oven gas will be introduced into blast furnace. In conventional ironmaking processes, CO₂ gas is generated when iron ore is reduced with CO gas. On the other hand, H₂O gas is generated instead of CO₂ in hydrogen reduction, and therefore this method can be regarded as an environmentally friendly ironmaking process. By increase of H₂ partial pressure in blast furnace, Fe₃C may form in this process. Many studies have been carried out for the Fe₃C formation.^{2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17)} Also, Fe₃C is a metastable phase and its decomposition behavior was investigated by one of the authors.¹⁸⁾ It is reported by Satoh et al.^2,3) and Gudenau et al.¹⁹⁾ that Fe₃C may enhance carburization and smelting of reduced iron. Formation of Fe₃C in blast furnace may have good influence on CO₂ suppression from ironmaking process. Hence, stability of Fe₃C in blast furnace will be important. From different viewpoints, metastable Fe₃C can be used as a material to convert CO₂ into CO and chemically decrease CO₂ emission. In the present work, Fe₃C sample was held at 1200 K under various CO–CO₂–H₂ gas mixtures to clarify the stability of Fe₃C phase. This experimental temperature was chosen to observe the stability of Fe₃C above the cohesive zone in blast furnace.

2. Experimental

Powder of Fe₃C was produced by the following procedure.^2,3) Reagent grade Fe₂O₃ was sieved using a filter with 75 μm, and 0.7 g of the sieved sample was placed on Al₂O₃ boat. Sample was placed in a horizontal silica tube (O.D. 25 mm, I.D. 21 mm, L. 900 mm)) inside an electric resistance furnace. The sample temperature was heated to 800 K under purified Ar flow of 0.05 L/min. After the sample temperature reached 800 K, gas was switched to 10vol%CO₂-H₂ mixed gas with flow rate of 0.2 L/min and held for 80 min. Then gas was switched to purified Ar and the sample was quenched by pulling out the horizontal silica tube from the furnace. The sample was analyzed by XRD to confirm the presence of Fe₃C phase.

Weighed 0.2 g of produced Fe₃C was placed on Al₂O₃ boat. Sample was placed in a same horizontal silica tube inside an electric resistance furnace as explained above. The sample temperature was heated to 1200 K at heating rate of 40 K/min under CO–CO₂–H₂ gas mixture with flow rate of 0.2 L/min and was held at 1200 K for 30 min. Also, some experiments changing holding time were conducted. Then the sample was cooled to room temperature at cooling rate of 35 K/min. The sample was analyzed by XRD for phase detection and phase abundance ratio was determined. Calibration was conducted using Fe₃C–Fe and FeO–Fe powder mixture as standard samples by obtaining the relation between mixed ratio and detected peak height ratio of Fe₃C/Fe and FeO/Fe. Peaks of 82.3°/2θ, 41.92°/2θ and 49.18°/2θ were utilized for Fe, FeO and Fe₃C, respectively. Carbon concentration of the sample was determined by infrared adsorption technique. Also, analysis of out gas was conducted by gas chromatography for selected samples.

3. Results and Discussion

Figure 1 shows the XRD pattern of sample heated at 1200 K under 70vol%CO-30vol%CO₂ gas mixture. Fe₃C changed to Fe during experiment. Only peaks of Fe were found for experiments longer than 10 min holding at 1200 K. Peaks of C is not found by XRD, however, carbon existed in each samples and its content in sample after holding 0 min, 10 min, 30 min and 50 min holding were 1.03mass%C, 0.385mass%C, 0.325mass%C and 0.267mass%C, respectively. It was confirmed that apparent change of sample after 30 min holding was small. Hence, most of the experiments were conducted under the condition of 30 min holding at 1200 K.

Fig. 1.

XRD pattern sample heated at 1200 K under 70vol%CO-30vol%CO₂ gas mixture.

The XRD pattern of Fe₃C sample heated under Ar at 1200 K and held for 0 min is shown in Fig. 2 with sample heated under 70vol%CO-30vol%CO₂ gas mixture. It can be seen that Fe and Fe₃C peaks exist after the experiment when heated under Ar. As reported in the previous paper,¹⁸⁾ when Fe₃C is heated under Ar it will decompose into Fe and C, and nanosize C will be included in metallic Fe phase as composites and some C particle will be flown out from the furnace. By changing the gas from Ar to70vol%CO-30vol%CO₂, C content at 0 min holding time changed from 2.30mass%C to 1.03 mass% and this concentration difference is due to the oxidation reaction of C particle by CO₂ in gas mixture.

Fig. 2.

XRD pattern sample heated at 1200 K under Ar or 70vol%CO-30vol%CO₂ gas and held for 0 min.

Experimental input gas composition, observed phases with abundance ratio and C content after experiment are shown in Table 1. Observed phases after experiment and stability region of Fe containing phases are shown in Fig. 3. Phases present after experiments were Fe₃C, FeO, Fe and C. Since the size of C phase is small, it exists in metallic Fe phase. Stability region were determined by using NIST-JANAF data²⁰⁾ and HSC Chemistry 7.0.²¹⁾ The abundance ratio of Fe₃C was lower than 15 mass% and most of the initial Fe₃C decomposed into Fe and C by heating to 1200 K and further oxidized by gas mixture. The experiments were conducted at the condition that Fe or FeO is thermodynamically stable. However, it was clear that Fe₃C existed after 30 min when partial pressure of CO₂ was lower than approximately 0.4. This may be due to rapid reaction of H₂ with oxygen or Fe oxide reported by Sasaki and Belton²²⁾ causing more reduced condition than equilibrium state at the gas-solid interface by H₂ addition as suggested and reported by Hino et al.³⁾ Gas composition region where Fe₃C can exist after holding 30 min at 1200 K was much wider than thermodynamic estimation. It can be suggested that some fraction of Fe₃C formed in blast furnace will still exist when H₂ gas is added and may enhance carburization and smelting of reduced iron.^2,3,19)

Table 1. Experimental input gas composition, observed phase with abundance ratio and C content of sample after 30 min holding at 1200 K.

input gas composition			Ccontent in sample after experiment	phase abundance ratio after experiment
vol%H2	vol%CO	vol%CO2	mass%C	mass% Fe3C	mass% Fe	mass% FeO
50	35	15	3.37	13	87	0
50	40	10	3.42	11	89	0
60	28	12	2.68	11	89	0
40	42	18	3.22	12	88	0
30	49	21	3.75	11	89	0
30	56	14	3.30	11	89	0
20	64	16	4.04	15	85	0
10	63	27	2.76	10	90	0
30	35	35	2.20	8	92	0
50	25	25	1.98	8	92	0
50	15	35	1.69	8	92	0
30	21	49	0.92	0	50	50
10	27	63	0.10	0	0	100
10	36	54	0.12	0	0	100
30	28	42	1.82	8	92	0
30	14	56	0.28	0	9	91
20	40	40	2.09	9	91	0
20	24	56	0.19	0	0	100
20	32	48	1.02	0	37	63
0	70	30	0.33	0	100	0

Fig. 3.

Observed phases after 30 min holding at 1200 K and calculated stability region of Fe containing phases.

Carbon content in the sample is shown in Fig. 4. In the sample, C mainly exist as remained Fe₃C or C particle in metallic Fe phase. It can be seen that C content decrease with increase of CO₂ partial pressure of gas mixture due to oxidation of C. Addition of H₂ up to approximately 30 vol% will increase the C content due to achieving reduced condition at interface by the reason explained above. Effect of H₂ on C content in sample can be seen from Fig. 5. Addition of H₂ to 70vol%CO-30vol%CO₂ gas will increase C content in sample after experiment. It was reported in the previous paper¹⁸⁾ that composite of Fe and nanosize C formed by decomposition of Fe₃C will rapidly melt. High C content in sample will have positive effect on lowering the smelting temperature of reduced iron in blast furnace.

Fig. 4.

Carbon content in sample after 30 min holding at 1200 K.

Fig. 5.

Effect of H₂ on carbon content after experiment.

Relationship between Fe₃C abundance ratio and C content in sample for experiments that Fe phase formed is shown in Fig. 6. With increase of Fe₃C abundance ratio, C content of sample after experiment increased. Considering that Fe₃C includes 6.69mass%C, how C exists in sample can be estimated. Carbon in the sample may exist as nanosize C or Fe₃C phase. It can be seen from Fig. 6 that C exists more as nanosize C than as Fe₃C phase. Also, from mass balance calculation, C content in metallic Fe phase can be obtained as shown in Fig. 7. Carbon content increased with increase of Fe₃C ratio. Carbon content exceeded C solubility in Fe at 1200 K²³⁾ and this result supports that C particle is suspended in metallic Fe phase. It can be considered that the following reactions occurred during the experiment.   

$${\text{Fe}}_3\text{C}\to 3\text{Fe}+\text{C}\left(\text{in Fe}\right)$$(1)

  

$$\text{C}\left(\text{in Fe}\right)+{\text{CO}}_2\to 2\text{CO}$$(2)

If reaction (1) is faster than reaction (2), C content in metallic Fe phase will not change with Fe₃C abundance ratio. However, C content in metallic Fe phase clearly decreased with decrease of Fe₃C ratio. Hence, it is considered decomposition reaction of Fe₃C is slower than reaction (2) and C particles in metallic Fe phase can efficiently react with CO₂.

Fig. 6.

Relation between Fe₃C abundance ratio and C content in sample after experiment.

Fig. 7.

Relation between Fe₃C abundance ratio and calculated C content in Fe after experiment.

The change of out gas composition is shown in Fig. 8 for experiment using 20vol%H₂-32vol%CO-48vol%CO₂ gas mixture. When the sample is heated to 1200 K CO gas concentration increased to approximately 48 vol% and gas composition was constant during holding at 1200 K. Phases detected after experiment were Fe and FeO, and C content of the sample was 1.02mass%C. It can be considered that Fe₃C decomposed into Fe–C composite and C in metallic Fe phase reacted with CO₂ and was continuously converted into CO. Therefore, Fe₃C itself may be used as an agent to continuously convert CO₂ into CO and chemically decrease CO₂ emission.

Fig. 8.

Change of gas composition with time during experiment.

4. Conclusion

Stability of Fe₃C sample was investigated held at 1200 K under various H₂–CO–CO₂ gas mixtures. It was confirmed that most of metastable Fe₃C phase will decompose into Fe and C, and FeO was formed at high CO₂ partial pressures and Fe₃C existed after 30 min holding at 1200 K when partial pressure of CO₂ is lower than approximately 0.4. Also, it was found that existence of H₂ will suppress decomposition of Fe₃C Addition of H₂ up to approximately 30 vol% to gas mixture will increase the C content after experiment. It was confirmed that CO₂ gas will be continuously converted into CO by C formed by decomposition of Fe₃C.

Acknowledgements

Authors wish to acknowledge valuable discussion with Prof. T. Nagasaka of Tohoku University, and the members of Research Group of Carbon Recycling Iron Making System, Division of Environmental, Energy and Social Engineering, The Iron and Steel Institute of Japan.

References

• 1)  The Japan Iron and Steel Federation: http://www.jisf.or.jp/course50/index_en.html, (accessed 2014-05-30).
• 2)   K.  Satoh,  T.  Noguchi,  T.  Miki,  Y.  Sasaki and  M.  Hino: ISIJ Int., 51 (2011), 1269.
• 3)   K.  Satoh,  T.  Noguchi and  M.  Hino: Steel Res. Int., 81 (2010), 834.
• 4)   H. J.  Grabke,  E. M.  Muller-Lorentz and  A.  Schneid: ISIJ Int., 41 (2001), S1.
• 5)   S.  Hayashi and  Y.  Iguchi: ISIJ Int., 37 (1997), 16.
• 6)   S.  Hayashi and  Y.  Iguchi: ISIJ Int., 37 (1997), 345.
• 7)   S.  Hayashi and  Y.  Iguchi: ISIJ Int., 38 (1998), 1053.
• 8)   A. N.  Conejo and  G. P.  Martins: ISIJ Int., 37 (1997), 967.
• 9)   A.  Schneider,  G.  Inden,  H. J.  Grabke,  Q.  Wei,  E.  Pippel and  J.  Woltersdorf: Steel Res. Int., 71 (2000), 179.
• 10)   A.  Schneider,  H.  Viefhaus,  G.  Inden and  H. J.  Grabke: Mater. Corros., 51 (2000), 338.
• 11)   A.  Schneider and  G.  Inden: Calphad, 31(2007), 141.
• 12)   F.  Bonnet,  F.  Ropital,  P.  Lecour,  D.  Espinat,  L.  Gengembre,  Y.  Verthier and  P.  Marcus: Surf. Interface Anal., 43 (2002), 418.
• 13)   J.  Zhang and  O.  Ostrovski: ISIJ Int., 41 (2001), 333.
• 14)   J.  Zhang and  O.  Ostrovski: ISIJ Int., 41 (2001), 340.
• 15)   E.  Park,  J.  Zhang,  S.  Thomson,  O.  Ostrovski and  R.  Howe: Metall. Mater. Trans. B, 32B (2001), 839.
• 16)   Y.  Iguchi,  K.  Hori and  T.  Shibata: ISIJ Int., 44 (2004), 992.
• 17)   G.  Li,  J.  Ma,  H.  Ni,  Q.  Shen and  F.  Tsukihashi: ISIJ Int., 46 (2006), 981.
• 18)   T.  Miki and  K.  Ishii: ISIJ Int., 54 (2014), 29.
• 19)   H. W.  Gudenau,  D.  Senk,  S.  Wang,  D. M.  Martins and  C.  Stephany: ISIJ Int., 45 (2005), 603.
• 20)   NIST-JANAF  Thermochemical Tables, 4th ed., ed. by M. W. Chase, Jr., J. Phys. Chem. Ref. Data, Washington, D.C., (1998).
• 21)   HSC  Chemistry, Version 5.11., Outokumpu Research Oy, Pori, Finland, (2002).
• 22)   Y.  Sasaki and  G. R.  Belton: Metall. Trans. B., 11B (1980), 221.
• 23)   H.  Ohtani,  M.  Hasebe and  T.  Nishizawa: Trans. Iron Steel Inst. Jpn., 24 (1984), 857.