Effect of Sulfur on Carburization and Melting Behavior of Iron by CO Gas

Abstract

In-situ observation of carburization and melting behavior of sulfur saturated iron by CO gas and carbon concentration measurement were carried out to discuss the effect of sulfur. The time for starting the melt formation on the sulfur saturated iron is twice longer than that on iron without sulfur because the average carburization rate of former iron is very slow compared with that of latter one. On the other hand, the time from reaching the saturated carbon concentration to starting the melt formation is short. The reason may be that the carburization rate is large when the carbon concentration reaches to the saturated one and the critical energy for the melt formation is smaller. The carburization to sulfur saturated iron is mixed control of desorption reaction of oxygen and sulfur on the iron surface and diffusion of carbon and sulfur into iron.

1. Introduction

A certain amount of sulfur contains in iron ore and coal. It is harmful element for ironmaking, and sulfur concentration in the steel product should be controlled at lower level. It is well known that gasification reaction of coal is inhibited because sulfur adsorbs on the active point of carbon.¹⁾ It delays the reduction of iron ore and carburization to metallic iron.^2,3) The reason is that sulfur is surface active element like oxygen. Fruehan³⁾ reported that the rate of carburization reaction decreased with increasing sulfur concentration by the experiment of carburization under CO–H₂–H₂S atmosphere. In this system, the following two reactions proceeds.   

$${\text{H}}_{\text{2}}+\text{CO}\leftrightarrow {\text{H}}_{\text{2}}\text{O}+\underset{\_}{\text{C}}$$(1)

  

$$\text{2CO}\leftrightarrow {\text{CO}}_{\text{2}}+\underset{\_}{\text{C}}$$(2)

The change in surface carbon concentration on the basis of the effect of sulfur is expressed as follows.   

$$\frac{dC}{dt}=\left(k_1{p}_{H_2}{p}_{CO}+k_2{p}_{CO}^2\right)\frac{{\phi }_S}{a_S}$$(3)

Here, k_i is positive rate constant for reaction i. p_i, φ_S, and a_S are partial pressure of i, activity coefficient of sulfur on the metallic iron surface, and sulfur activity in iron, respectively. φ_S/a_S can be expressed by (1-θ). θ is coverage factor of adsorbed element. However, they did not consider the dissociative adsorption of C and O in case of CO gas carburization. Therefore, coverage factor was discussed only based on the adsorption of sulfur.

Sulfur affects the gasification of carbonaceous materials. Kasaoka et al.¹⁾ reported that the gasification rate of carbon pronouncedly decreased by existence of H₂S gas, and COS gas formed by the decomposition reaction of H₂S. Hayashi et al.⁴⁾ reported the effect of sulfur on the reduction of iron oxide by H₂ gas. The adsorbed sulfur on the surface of reduced iron prevents oxygen removal at the coexisting point of gas-Fe-FeO and the nucleation of metallic iron. In these reports on the effect of sulfur using H₂S gas, carburization reaction is complicated because the gas contains hydrogen and two types of carburization reaction of Eqs. (1) and (2) should be considered as reported by Fruehan.³⁾

The carburization to γ-iron by CO gas is controlled by the mixed rate-determining step, which consists of the carbon diffusion in γ-iron and chemical reaction at the gas-solid interface.⁵⁾ The iron melting by CO gas carburization has an incubation time which depends on the chemical reaction on the surface and surface tension.⁶⁾ Surface tension of iron decreases with increasing sulfur concentration in iron because of surface active element. Therefore, an incubation time may decrease by the addition of sulfur. In this study, the mechanism of the carburization and melting behavior of sulfur saturated iron by CO gas was evaluated.

2. Experimental

2.1. Sample Preparation

Electrolytic iron (99.99%, 10×20 mm) with the thickness of 0.5 mm was prepared. The powder mixture of iron (99.9%) and iron sulfide (99.9%) was set in alumina boat. It was inserted in quartz tube, and the tube was vacuum-encapsulated. This tube was heated at 1573 K. The temperature of this tube was controlled within ±12 K. The mixture ratio of Fe and FeS was approximately 5:1, which is solid-liquid coexistent state of Fe–S system at the heat treatment temperature. In the quartz tube, vapor pressure of sulfur is controlled by the alloy of Fe–FeS. And, sulfur activity on the surface of γ-Fe was determined by this vapor pressure. The sulfur concentration of γ-Fe reaches to solidus when the sample is equilibrium with the gas. The equilibrium of this system attains by the diffusion of sulfur in γ-Fe. Crank⁷⁾ derived the general solution of Fick’s law for the diffusion of plane sheet of thickness, l, in a region 0 < x < l with variable surface concentration. The carbon concentration at position x and time t, C(x, t) is expressed by the equation,   

$$\begin{array}{l}\frac{C-C_0}{C_1-C_0}=1-\frac{4}{\pi }\sum _{n=0}^{\infty }\frac{{\left(-1\right)}^n}{2n+1}\\ exp\left\{\raisebox{1ex}{$-D_S{\left(2n+1\right)}^2{\pi }^{2}t$}\!\left/ \!\raisebox{-1ex}{$4l^2$}\right.\right\}cos\frac{\left(2n+1\right)\pi x}{2l}\end{array}$$(4)

C₀ and C₁ are initial and surface concentration of sulfur, respectively. The diffusion coefficient of sulfur in γ-Fe, D_s=9.3×10⁻¹² m²s⁻¹ was calculated from the experimental results at 1573 K reported by Seibel et al.⁸⁾ When the sulfur concentration at the center of plate became 97% of solidus, the sulfur concentration was considered as constant. As a result, the treatment time was 57.6 ks at 1573 K.

The obtained sample was confirmed without the formation of FeS and S on the surface of γ-Fe by XRD analysis. After that, the sample was machined into a cylindrical shape in 2 mm diameter, and the surface was polished. The sample with the thickness of 0.46 ± 0.02 mm was subsequently set in Al₂O₃ tube (2 mm ID, 3 mm OD, 2 mm L). The edge of the sample was covered with ZrO₂–SiO₂ cement to avoid the kind of edge effect on melting.⁶⁾ To maintain the carbon activity of CO gas unity, small pieces of graphite (99.99%) were placed around the sample. This sample with graphite in the crucible was used for in situ observation.

2.2. Carburization Experiment

The behavior of carburization and melting iron was observed in situ by a confocal scanning laser microscope incorporating an infrared image furnace, which was described in the previous study⁶⁾ in detail. Al₂O₃ crucible with the sample was introduced into the chamber, and then, the chamber was purged with CO gas for 3.6 ks. The sample was subsequently heated to the desired temperatures, 1523–1623 K in 60 s and kept for various time periods. The temperature was controlled using a B-type thermocouple placed at the sample holder. The temperature of sample surface was calibrated using melting points of Ag, Cu and Ni, which was set at the same position as the iron sample. Some uncertainly in temperature was ±5 K. CO gas was passed through the chamber at various flow rates of 5×10⁻⁶ Nm³/s during the carburization experiments, which are controlled by a mass flow controller. The sample surface was observed through the experiments. After the experiments, the sample was cooled by turning off the furnace. A vertical cross section was analyzed by electron probe microanalysis (EPMA) and carbon concentration profile was measured. The profiles were measured at three different positions of each samples and the analytical area was 8 μm². On the other hand, the carbon concentrations at surface were determined from fifteen analysis points in the surface area of the radius of 25 μm to avoid the effect of microsegregation. Furthermore, concentrations of carbon and sulfur in the iron sample before and after carburization experiment were measured by infrared absorption method after combustion in high-frequency induction furnace. And those of oxygen and nitrogen were also measured by inert gas fusion infrared and thermal conductivity detection technique.

3. Results

3.1. Concentration of Oxygen, Nitrogen, Carbon, and Sulfur in the Sample

Table 1 lists the concentration of oxygen, nitrogen, carbon, and sulfur in the sample before and after carburization experiment at 1583 K. For comparison, those in electrolytic iron were also shown. Sulfur concentration in electrolytic iron is below 1 ppm, which is very low. On the other hand, that in the sample before carburization experiment is 0.037 mass%. This value is corresponding to the solidus at 1573 K, which is the heat treatment temperature for sulfurization, based on the phase diagram for Fe–S system. Carbon concentration slightly increases by the heat treatment for sulfurization. However, it can be neglected because its change is very small compared with that by carburization experiment. After experiment, sulfur concentration decreases to approximately 0.1 mass% in all carburization time. It means that carburization and desulfurization reactions simultaneously occurred.

Table 1. Oxygen, nitrogen, carbon, and sulfur concentrations in iron before and after carburization experiment.

		O (mass%)	N (mass%)	C (mass%)	S (mass%)
Before exp.	Electrolytic iron	0.01604	0.00040	0.00186	0.00009
	Iron saturated with sulfer	0.00843	0.00000	0.03814	0.03703
	Experimental time (s)	Beginning of melting (s)		C (mass%)	S (mass%)
After exp.	1922	1756		0.96949	0.00956
	2400	2210		1.11380	0.01085
	2040	1886		0.97397	0.01194
	1500	–		0.85497	0.01338
	1500	–		0.97711	0.01125

3.2. In situ Observation of Carburization and Melting of Sulfur Saturated Iron

The change in surface appearance of sulfur saturated iron during heating is shown in Fig. 1. Figure 1(a) shows the appearance for 3 s after starting of heating. Left side of the figure, which is dark area, is ZrO₂–SiO₂ cement, and other area is iron surface. There is no change from original surface. The polishing flaws in a single direction are observed. Figure 1(b) shows the appearance at 1089 K for 35 s. A grain boundary of α-Fe appears as shown at the white arrow. At 1390 K, grain boundary of γ-Fe is observed as shown in Fig. 1(c) and grain boundary of α-Fe disappears because of α → γ phase transformation. Figure 1(d) shows the appearance at 1583 K for 30 s. There is no flow on the surface of iron while the polishing flaws disappears and the appearance becomes smooth. Namely, it can be concluded that iron does not start to melt immediately by CO gas after reaching to the eutectic temperature of Fe–C system.

Fig. 1.

Top view of sulfur saturated iron sample a) at room temperature, b) at 1089 K, c) at 1390 K during heating, and d) for 30 s attaining at 1583 K.

The melting behavior of sulfur saturated iron heated up to 1583 K is shown in Fig. 2. This sample is same as that of Fig. 1. The melt formation on the surface of iron was observed for 1.8 ks after reaching to the target temperature. The radius of the liquid iron was approximately 70 μm when the melt formation started. The liquid area gradually expands. And impurities move on the surface of liquid.

Fig. 2.

Top view of sulfur saturated iron sample heated up to 1583 K for a) 1790 s, b) 1795 s, c) 1800 s, and d) 1805 s.

3.3. Change in Carbon Concentration in Iron

Figure 3 shows the carbon concentration profiles in sulfur saturated iron for 0.3, 0.6, and 1.5 ks at 1583 K. Carbon concentration decreases from the surface to inside of iron, and shows the constant values at all conditions. The shape of carbon concentration profiles of these three conditions is almost similar while the average carbon concentration increases with carburization time. The surface carbon concentration increases with increasing carburization time. However, it does not reach to the solidus concentration, Cs for 1.5 ks.

Fig. 3.

Carbon concentration profiles of γ-Fe for different carburization time at 1583 K.

Change in carbon concentration on the surface of sulfur saturated iron and iron without sulfur⁶⁾ with carburization time at 1583 K is shown in Fig. 4. The dashed line and arrows mean the solidus concentration Cs at this temperature and the starting time of melting, respectively. The surface carbon concentration of sulfur saturated iron gradually increases with increasing the carburization time. However, its rate is much slower than that of iron without sulfur. The increasing rate of the surface carbon concentration becomes diminished for approximately 1000 s, and it rapidly increases for 1300 s. After reaching the solidus concentration, the melt formation on the surface of sulfur saturated iron starts soon. This behavior is different from that of iron without sulfur, and cannot be expressed by the equation reported for iron without sulfur.⁶⁾ It means that sulfur affects the rate-determining step of carburization.

Fig. 4.

Change in carbon concentration on the surface of sulfur saturated iron and iron without sulfur with carburization time at 1583 K.

Figure 5 shows carbon concentration profiles in sulfur saturated iron for 0.6 ks at different temperatures. The dashed lines are the solidus concentration at the temperature. The profiles at all conditions show similar behavior each other.

Fig. 5.

Carbon concentration profiles of sulfur saturated iron carburized at 1523 K, 1583 K, and 1623 K.

4. Discussion

4.1. Desorption Reaction of Oxygen and Sulfur on the Iron Surface

For the reaction kinetics between CO gas and sulfur saturated solid iron, it is necessary to consider the following six steps.

(1) CO gas supply to the surface of iron

(2) Dissociative adsorption reaction of CO gas on the iron surface

(3) Desorption reaction of adsorbed oxygen on the iron surface

(4) Diffusion of adsorbed carbon into the iron

(5) Desorption reaction of adsorbed sulfur on the iron surface

(6) Diffusion of sulfur in the iron to the surface

In this study, the step (1) can be neglected from the rate-determining steps because CO gas flow did not affect described above. Shatynski et al.⁹⁾ reported that the rate of the step (2) has five times higher than that of the (3) at 1273 K. Therefore, the step (2) is not rate-determining step. In this study, consequently, the steps (3)–(6) are discussed as the rate-determining step. The rate equation of carburization reaction considered these four steps was developed.

The carburization reaction on the surface of pure iron without sulfur by CO gas can be expressed by   

$${\text{CO}}_{\left(\text{g}\right)}\stackrel{{\text{k}}_{\text{5}}^{\pm }}{⟷}{\text{C}}_{\left(\text{inFe}\right)}+{\text{O}}_{\left(\text{ad}\right)},k_5^{+}/k_5^{-}=K_5$$(5)

  

$${\text{CO}}_{\left(\text{g}\right)}+{\text{O}}_{\left(\text{ad}\right)}\stackrel{{\text{k}}_{\text{6}}^{\pm }}{⟷}{\text{CO}}_{\text{2}}{}_{\left(\text{g}\right)}$$(6)

where subscript ad expresses adsorbed species on iron. If sulfur is added to iron, not only oxygen and carbon but also sulfur adsorbs on the iron surface. Adsorbed sulfur reacts with CO gas, and COS gas generates. This reaction can be expressed as follows,   

$${\text{CO}}_{\left(\text{g}\right)}+{\text{S}}_{\left(\text{ad}\right)}\stackrel{{\text{k}}_{\text{7}}^{\pm }}{⟷}{\text{COS}}_{\left(\text{g}\right)}$$(7)

On the iron surface, these three reactions should be considered. Here, equilibrium coefficient, K₅, and the rate equation of reaction (6) and (7) are expressed as follows,   

$$K_5=\frac{a_C{\theta }_O}{\left(1-{\theta }_O-{\theta }_S\right)P_{CO}}$$(8)

  

$$\frac{d{O}_{\left(ad\right)}}{dt}=-k_6^{+}{P}_{CO}{\theta }_O+k_6^{-}{P}_{C{O}_2}\left(1-{\theta }_S-{\theta }_O\right)$$(9)

  

$$\frac{d{S}_{\left(ad\right)}}{dt}=-k_7^{+}{P}_{CO}{\theta }_S+k_7^{-}{P}_{COS}\left(1-{\theta }_S-{\theta }_O\right)$$(10)

where θ_O and θ_S are the coverage factor of adsorbed oxygen and sulfur, respectively. Second term on the right side of Eq. (9) can be neglected because partial pressure of CO is 1 atm and the generated amount of CO₂ is very small.

To discuss the effect of second term on the right side of Eq. (10), sulfur concentration in iron before carburization experiment and partial pressure of COS gas in equilibrium with CO gas of 1 atm were calculated as follows.

Change in standard Gibbs energy change of the following reaction, ΔG° can be expressed as follows.^10,11)   

$${\text{CO}}_{\left(\text{g}\right)}+\frac{\text{1}}{\text{2}}{\text{S}}_{\text{2}\left(\text{g}\right)}={\text{COS}}_{\left(\text{g}\right)}$$(11)

  

$$\mathrm{\Delta }G{°}_{\left(11\right)}/{\text{Jmol}}^{-1}=76.501T-89\mathrm{\hspace{0.17em} }491$$(12)

  

$$\frac{\text{1}}{\text{2}}{\text{S}}_{\text{2}\left(\text{g}\right)}={\text{S}}_{\left(\text{inFe}\right)}$$(13)

  

$$\mathrm{\Delta }G{°}_{(13)}/{\text{Jmol}}^{-1}=18.41T-125\mathrm{\hspace{0.17em} }281$$(14)

Using Eqs. (12) and (14), the formation reaction of COS from CO and S in liquid iron can be expressed as follows.   

$${\text{CO}}_{(\text{g})}+{\text{S}}_{(\text{inFe})}={\text{COS}}_{(\text{g})}$$(15)

  

$$\mathrm{\Delta }G{°}_{(15)}/{\text{Jmol}}^{-1}=58.091T+35\mathrm{\hspace{0.17em} }790$$(16)

In this chemical equation, sulfur solves in liquid iron. In this study, however, sulfur solves in solid iron as solid solution. In Eq. (13), existing state of sulfur is different from that in this study. Near the carburization temperature, sulfur concentration is saturated with solid iron. Therefore, activity of sulfur in iron sample is same as that in sulfur saturated iron.

Activity coefficient of sulfur in liquid iron at the temperature from 1773 K to 1873 K was reported by Ishii et al.¹¹⁾   

$$\text{log}({\gamma }_{\text{S}})=\left(-120/\text{T}+0.018\right)\cdot \left[\%\text{S}\right]$$(17)

where γ_S is activity coefficient of sulfur in liquid iron based on the standard for infinite dilute solution. [%S] is expressed by mass. Using this equation, sulfur activity in solid iron at 1583 K can be calculated. Saturated sulfur concentration in solid iron can be obtained by the phase diagram¹²⁾ as 0.0367 mass% at 1583 K. Using these values, partial pressure of COS was calculated, and it was 2.33×10⁻⁶ atm. COS gas generates only from the desorption reaction on the iron surface because supplied gas does not include COS gas. However, the generated COS gas flows out soon by strong gas flow in the chamber due to large temperature distribution in the chamber. Therefore, it seems that partial pressure of COS gas is almost 0. Consequently, the second term of right side in Eq. (10) can be neglected.

Considering these, the rate Eqs. (9) and (10) can be expressed as follows,   

$$\frac{d{O}_{\left(ad\right)}}{dt}=-k_6^{+}{P}_{CO}{\theta }_O$$(18)

  

$$\frac{d{S}_{\left(ad\right)}}{dt}=-k_7^{+}{P}_{CO}{\theta }_S$$(19)

4.2. Change in Sulfur Concentration on Iron Surface

Carbon amount reached to the iron surface by carburization reaction is equal to the carbon flux on the iron surface. Therefore, the carburization rate can be expressed as follows.   

$$V_{C-S}=-\frac{\rho }{100}{D}_C\frac{dC}{dx},\hspace{1em}x=0$$(20)

where V_C–S, ρ, and D_C are the carburization rate to saturated iron, density of iron, and diffusion coefficient of carbon in iron. Positive reaction of Eq. (5) proceeds because the vacancy of adsorbed site forms due to desorption reaction of oxygen and sulfur by CO gas. This supplies carbon on the iron surface. The supply rate of carbon to the surface is controlled by the Eqs. (5), (6), and (7), and diffusion of sulfur into solid iron. The reason why diffusion of sulfur affects this rate is that the vacancy of adsorbed site is covered by sulfur diffused from bulk iron, which is described below. Therefore, the carburization rate by CO gas can be expressed as follows.   

$$V_{C-S}=k_6^{+}{P}_{CO}{\theta }_O-\frac{d{\theta }_S}{dt}$$(21)

where first and second terms of right side are about oxygen and sulfur adsorption, respectively. From Eq. (8), the coverage factor of oxygen is as follows.   

$${\theta }_O=\frac{K_5{P}_{CO}}{a_C+K_5{P}_{CO}}\left(1-{\theta }_S\right)$$(22)

The Eqs. (21) and (22) were substituted into the Eq. (20), and the coverage factor of sulfur is expressed as follows.   

$$\frac{d{\theta }_S}{dt}=\frac{K_5{k}_6{P}_{CO}^2}{a_C+K_5{P}_{CO}}\left(1-{\theta }_S\right)+\frac{\rho }{100}{D}_C\frac{dC}{dx}$$(23)

Here, K₅ and k₆ at 1583 K are 1.98×10⁻² and 6.57×10⁻⁵, respectively, which is reported by Zhang et al.¹³⁾ Density of iron and diffusion coefficient of carbon at 1583 K are 7.3×10³ kgm⁻³ and 4.0×10⁻¹⁰ m²s⁻¹. Partial pressure of CO gas is 1 atm. θ_S, a_C, and dC/dx are the function of time variables, and a_C, and dC/dx can be calculated from the experimental results. Therefore, the coverage factor, θ_S can be calculated.

The change in carbon concentration on the iron surface as shown in Fig. 4 was approximated by the following function.   

$$C\left(t\right)=5.84\times {10}^{-10}{t}^3-1.86\times {10}^{-6}{t}^2+2.11\times {10}^{-3}t$$(24)

The relation between surface carbon concentration and activity of carbon was reported by Ellis et al.¹⁴⁾ as follows,   

$$a_c=\frac{1.07C(t)}{100-19.5C(t)}exp\left(\frac{4\mathrm{\hspace{0.17em} }798.6}{T}\right)$$(25)

The slope of carbon concentration at the iron surface, which is corresponding to carbon flux at the surface of iron, can be calculated from the cross-section profile as shown in Fig. 3. The calculated results of carbon flux for 0.3, 0.6, and 1.5 ks are plotted on Fig. 6. The slope of carbon concentration increases with passing the carburization time, and further increase in the time makes it decrease. Using quadratic function passed through the origin, which is most simple function, the following equation was developed by least square method using these three data.   

$$J_C=-\frac{\rho D_C}{100}\frac{dC}{dx}\left(kg/m^{2}s\right)=-3.78\times {10}^{-10}{t}^2+7.21\times {10}^{-7}t$$(26)

Fig. 6.

Change in calculated carbon flux at the iron surface with carburization time at 1583 K.

The value of initial θ_S is required when the function of θ_S is calculated. However, it is very difficult to calculate it because there is few data of surface tension of solid iron considering the impurity at high temperature. Assuming that the ratio of surface excesses of oxygen and sulfur is equal to the ratio of these coverage factors and this value does not change even if temperature and state of iron change, θ_S was estimated roughly. The value surface excess for F–S–O system was reported as 0.744×10⁻⁵ mol and 0.735×10⁻⁵ mol at 0.05 mass%S and 0.015 mass%O, respectively.¹⁵⁾ These concentrations of oxygen and sulfur is similar to those of this study. Therefore, it can be estimated that the ratio of coverage factor of oxygen and sulfur is approximately 1. At initial stage of the carburization experiment in this study, only two types of element as oxygen and sulfur determine the coverage factor because adsorbed elements on the iron are only these elements. Accordingly, θ_S was determined as 0.5.

The Eqs. (24) and (26) were substituted into the Eq. (23), and θ_S was calculated as shown in Fig. 7. θ_S gradually decreases with increasing the carburization time.

Fig. 7.

Change in calculated coverage factor of sulfur on the solid iron with carburization time at 1583 K.

4.3. Rate-determining Step of Carburization to Sulfur Saturated Iron by CO Gas

Not only desorption reaction of oxygen on the iron surface but also coverage factor of sulfur which fixes oxygen coverage is important for carburization reaction to sulfur saturated iron. At the late stage of carburization, carburization reaction accelerates by amount of adsorbed site of oxygen increases due to decreasing that of sulfur. Therefore, four steps described in the section 4.1 cannot be neglected for the rate-determining step. Namely, the carburization to sulfur saturated iron by CO gas is mixed control of four steps as desorption reaction of oxygen and sulfur on the iron surface and diffusion of sulfur and carbon in the iron.

4.4. Starting Time of Melt Formation and its Driving Force

From the observation of melting behavior as shown in Fig. 2, the minimum radius of liquid iron for the melt formation was estimated at approximately 70 μm at 1583 K. Using the previous method,⁶⁾ the calculation of the critical size of liquid is tried. However, the contact angle between solid and liquid interface of sulfur saturated iron has not been reported. The value of this contact angle is smaller than that of iron without sulfur because the contact angle decreases with decreasing the surface tension and sulfur addition leads to decreasing the surface tension. Therefore, the value of the contact angle of sulfur saturated iron is less than 20° which is the value of iron without sulfur. It is assumed that this value is 10°. Accordingly, the critical size was estimated as 403 μm. And the supersaturation degree was as 0.014 mass%, which is lower than that of iron without sulfur (0.024 mass%). It means that the required driving force of liquid formation of sulfur saturated iron is smaller than that of iron without sulfur.

However, much longer time for the liquid formation of sulfur saturated iron is required. It is more than twice for that of iron without sulfur. The reason is the effect of sulfur on the carburization reaction on the iron surface as shown in Fig. 4. The incubation time⁶⁾ which is defined as the time required for the carbon concentration to reach the carbon concentration when the radius is critical size at the beginning of melting from saturated concentration based on the phase diagram becomes short. The incubation time depends on the rate of CO gas carburization because it is the time for accruing the critical energy to form liquid after reaching the saturated concentration. In this study, the carburization rate decreases after 0.6 ks carburization, and then it increases again. And the rate is larger than that of iron without sulfur. The reason is the decrease in the amount of adsorbed sulfur. By these two factors, the incubation time of sulfur saturated iron is shorter than that of iron without sulfur.

5. Conclusions

The carburization and melt formation mechanism of sulfur saturated iron was studied by in situ observation of melting behavior and the analysis of the sample. The obtained results are summarized as follows.

(1) The melt of iron-carbon alloy formed on the surface of iron with the thickness of 0.46 mm by CO gas carburization at 1583 K for approximately 1.9 ks. This is twice longer than that of iron without sulfur.

(2) The carburization reaction of sulfur saturated iron by CO gas is mixed control of four steps described as follows.

a) Desorption reaction of oxygen on the iron surface

b) Desorption reaction of sulfur on the iron surface

c) Diffusion of carbon in solid iron

d) Diffusion of sulfur in solid iron

(3) The incubation time of the melt formation of sulfur saturated iron is shorter than that of iron without sulfur. The reason is that the critical energy for melt formation decreases and the carburization rate is large when carbon concentration reaches near solidus.

References

• 1)   S.  Kasaoka,  Y.  Sakata and  M.  Shimada: J. Fuel Soc. Jpn., 63 (1984), 861.
• 2)   S.  Hayashi,  Y.  Iguchi and  J.  Hirao: Tetsu-to-Hagané, 71 (1985), 281.
• 3)   R. J.  Fruehan: Metall. Trans., 4 (1973), 2129.
• 4)   S.  Hayashi and  Y.  Iguchi: Tetsu-to-Hagané, 73 (1987), 1668.
• 5)   X.  Ye,  H.  Numata and  S.  Haruyama: J. Jpn. Inst. Met., 49 (1985), 529.
• 6)   T.  Murakami,  H.  Fukuyama and  K.  Nagata: ISIJ Int., 41 (2001), 416.
• 7)   J.  Crank: The Mathematics of Diffusion, Oxford University Press, London, (1975), 47.
• 8)   M. G.  Seibel: Rev. Met. Mem. Sci., 6 (1964), 431.
• 9)   S. R.  Shatynski and  H. J.  Grabke: Arch. Eisenhüttenwes, 49 (1978), 129.
• 10)   I.  Barin: Thermochemical Data of Pure Substances, VCH Pubulishers, New York, (1995), 277.
• 11)   F.  Ishii and  T.  Fuwa: Tetsu-to-Hagané, 67 (1981), 736.
• 12)   O.  Kubaschewski: IRON-Binary Phase Diagram, Springer-Verlag, Berlin, Heidelberg, New York, (1982), 80.
• 13)   X.  Zhang,  R.  Takahashi,  T.  Akiyama and  J.  Yagi: Tetsu-to-Hagané, 83 (1997), 299.
• 14)   T.  Ellis,  I. M.  Davidson and  C.  Bodsworth: J. Iron Steel Inst., 201 (1963), July, 582.
• 15)   M.  Divakar,  J. P.  Hajra,  A.  Jakobsson and  S.  Seetharaman: Metall. Mater. Trans. B, 31B (2000), 267.