Effect of Nitrogen on the Formation and Evolution of Non-Metallic Inclusions in Fe–Al–Ti–N Alloy

Abstract

The effects of nitrogen content on the formation and changing behaviors of non-metallic inclusions in the four kinds of Fe–Al–Ti–N alloys during heating at 1473 K were studied by observation and analysis of inclusions by FE-SEM and EDS. At the inner part of the specimens, TiN-based inclusions were mainly observed. The size of TiN single phase particles observed in as-cast specimens was much larger in the specimens with larger nitrogen content. The fraction of the number of TiN-based inclusions increased by heating in the specimens with larger nitrogen content, while it decreased in the case of smaller nitrogen content due to the formation of TiS inclusions during heating. The growth of TiN single phase particles in larger nitrogen content specimens was more significant than those in smaller nitrogen content specimens. At the outer part of the specimens, TiN formed after heating on the already existing oxide inclusions. In addition, the size of TiN on the oxide surface formed during quenching also increased by heating. The fraction of the number of oxide inclusions with TiN increased in all specimens by heating. TiN phase on the oxide surface in as-cast specimens was bigger in the case of larger nitrogen contents. TiN growth during heating was also enhanced in larger nitrogen content specimens.

1. Introduction

Titanium-alloyed steels have been widely used for various products such as automobile sheets, or heavy plates for ship construction. Since titanium has strong affinity with dissolved oxygen, aluminum is commonly added for deoxidation reaction before titanium addition for economical reason. Therefore, Al–Ti complex deoxidation is a common process for Ti-alloyed steel. The added aluminum and titanium form various types of non-metallic inclusions such as oxide, nitride, carbide, sulfide and so on.

Some of the formed Ti-based inclusions have advantage for the control of microstructure during solidification or heat treatment, such as prevention of austenite grain growth by pinning effect and assistance of finer ferrite microstructure formation.¹⁾ Therefore, formation and evolution behaviors of non-metallic inclusions in solid state steel are also quite important to understand the role of inclusions on the formation of microstructure of steel. Many investigations regarding the formation mechanisms of inclusions during solidification²⁾ have been conducted and the effects of inclusions on the solidification microstructure³⁾ and the microstructure in solid state steel during heating⁴⁾ have been clarified so far. Especially, Fujimura et al. have reported that the formed titanium nitride has an advantage to produce the equiaxed cast structure during continuous casting.⁵⁾ Nitrogen content is one of the significant parameter to bring out the positive effects of inclusions by TiN formation during solidification or heat treatment.

The goal of this research is to utilize Al or Ti-based inclusions after Al–Ti complex deoxidation for improving the microstructure, and the objective of the present study is to clarify the evolution behaviors of various inclusions in the solid state alloy by observing inclusions in as-cast and heated Fe–Al–Ti–N alloys. Especially, the effects of nitrogen content in the alloy and heating on the inclusion in terms of size, composition and number density were discussed based on the observed phenomena.

2. Experimental

2.1. Specimen

About 100 g of electrolytic iron were melted in an Al₂O₃ Tammann tube (O.D.: 30 mm, I.D.: 24 mm, height: 150 mm) by using an induction furnace at 1873 K in Ar or Ar–N₂ atmosphere. After melting, aluminum and titanium were added in sequence with the interval of 2 min. Ar–N₂ gas mixture was introduced to add nitrogen in the melt. After holding 3 min, the melt was quenched by water. To clarify the formation and evolution behaviors of inclusions with or without oxide inclusions initially formed by complex deoxidation, the preformed oxide inclusions were collected at the outside of the melt by the strong stirring. Therefore, an induction furnace was used in the present study. The compositions of obtained specimens were analyzed by ICP-OES for soluble Al and Ti, and by combustion analyzer (LECO TC600 and CSLS 600) for total oxygen, nitrogen and sulfur. Sulfur is mainly introduced from the electrolytic iron which contained 0.002 mass% of sulfur.

2.2. Procedure of Heating Experiments

A cylindrical piece of the alloy (O.D.: 24 mm, height: from 13 to 15 mm) was machined from the produced alloy and the piece was encapsulated into a low carbon steel ampoule (O.D.: 32 mm, I.D.: 24 mm, height: 20 mm) by Ar arc-welding at high purity Ar atmosphere of approximately 0.5 atm for the inhibition of oxidization of sample. The ampoule was then heated at 1473 K which is normal heat treatment temperature for 3 h in an electric furnace. As-cast and heated specimens were embedded in the phenolic resin, and polished by SiC papers and diamond suspensions up to 0.25 μm. Approximately 100 inclusions in an as-cast or a heated specimen were observed and analyzed by FE-SEM (JEOL JSM-7001FA) and EDS (JEOL JED-2300). The compositions of Al, Ti, Fe, S, N and O were measured. In the present study, the size of an inclusion was defined as the maximal diameter of that particle.

3. Results and Discussion

Four kinds of specimens were prepared for heating experiments as summarized in Table 1. The compositions of specimens were also shown on the previously reported stable oxide phase diagram at 1873 K⁶⁾ in Fig. 1. At deoxidation temperature, the stable oxide phase of the specimens A and B is expected to be Al₂O₃, while that of the specimens C and D is to be Ti₃O₅.

Table 1. Compositions of specimens (mass%).

Specimen	Sol. Al	Sol. Ti	Total O	Total N	Total S
A	0.0274	0.089	0.0005	0.0005	0.0017
B	0.0392	0.093	0.0007	0.0058	0.0012
C	0.0003	0.109	0.0012	0.0013	0.0020
D	0.0009	0.123	0.0010	0.0055	0.0017

Fig. 1.

Compositions of specimens on the stable oxide phase diagram at 1873 K.

3.1. Observed Area of Inclusions in As-cast Specimens

The inclusions observation areas on the cross section of the cylindrical shape as-cast specimens were classified into two regions by morphology of inclusions. Figure 2 shows the SEM images of observed typical inclusions in the specimen D and the schematic representation of the cross section of the cylindrical shape as-cast specimen. The area named as the “Outer layer” is the cylindrical pipe shape with the thickness of approximately 700 μm and the “Inner layer” is the inside of the outer layer. The difference in morphology between two regions could be observed, since the preformed oxide inclusions were removed from the “Inner layer” by the inductive force as planned. Description of inclusions as “X+Y” means that there are two different X and Y phases in one inclusion particle.

Fig. 2.

Schematic drawing of the classification of the observation region by inclusion type and SEM images of typical inclusions observed in the specimen D.

3.2. Inclusions in As-cast Specimens

3.2.1. Type of Inclusions

Figures 3 and 4 show the observed inclusion types at the outer and inner layers in all as-cast specimens, respectively.

Fig. 3.

Observed inclusion types in the outer layer of as-cast specimens.

Fig. 4.

Observed inclusion types in the inner layer of as-cast specimens.

In the outer layer, mostly Al₂O₃(+TiN) inclusions were observed in the four specimens. In the specimen A, some of the Al₂O₃+TiO_x(+TiN) inclusions were observed. And some of Al₂O₃+TiO_x+TiN and TiS-based inclusions were detected in the specimen C. In the specimen D, a fewer number of TiN particles were observed. Oxide inclusions were produced by Al–Ti deoxidation reaction and accumulated by the inductive flow of the melt and the adhesion onto the crucible surface. The soluble nitrogen content of each specimen at TiN saturating condition at 1873 K was calculated by Eq. (1) with soluble Al and Ti contents shown in Table 1 and following interaction parameters; $e_{\text{N}}^{\text{Ti}}$ =−0.503,⁷⁾ $e_{\text{N}}^{\text{N}}$ =0,⁷⁾ $e_{\text{N}}^{\text{Al}}$ =0.01,⁷⁾ $e_{\text{Ti}}^{\text{Ti}}$ =0.042,⁷⁾ $e_{\text{Ti}}^{\text{N}}$ =−2.04,⁷⁾ and $e_{\text{Ti}}^{\text{Al}}$ =0.026.⁸⁾ The calculated nitrogen content was in the range from 0.0162 to 0.0222 mass%, which was much larger than total nitrogen content in all specimens shown in Table 1. Therefore, TiN inclusions observed as independent particles or attached on the surface of oxide inclusions were formed due to the segregation of Ti and N during quenching.   

$$\begin{array}{l}\text{TiN}(\text{s})=\underset{\_}{\mathrm{T}\mathrm{i}}(\text{mass}\%)+\underset{\_}{\text{N}}(\text{mass}\%)\\ logK=–19\mathrm{\hspace{0.17em} }800/T+7.78\end{array}$$(1) ⁷⁾

TiS formation in the outer layer of the specimen C could be explained as follows. Assuming the interaction parameters during quenching are equal to those at 1873 K,^7,8) such as $e_{\text{S}}^{\text{O}}$ =−0.27, $e_{\text{S}}^{\text{S}}$ =−0.046, $e_{\text{S}}^{\text{N}}$ =0.01, $e_{\text{S}}^{\text{Ti}}$ =−0.18, $e_{\text{S}}^{\text{Al}}$ =0.041, $e_{\text{Ti}}^{\text{O}}$ =−1.62, $e_{\text{Ti}}^{\text{S}}$ =−0.27, $e_{\text{Ti}}^{\text{N}}$ =−2.04, $e_{\text{Ti}}^{\text{Ti}}$ =0.042 and $e_{\text{Ti}}^{\text{Al}}$ =0.026, activity of sulfur in the specimen C is 0.00191. It is the largest value among four specimens. On the other hand, activity of titanium in the specimen D is the largest of 0.121. As a result, the product of the sulfur and titanium activities of the specimen C is the largest value as 0.00021 and thus it is expected that the formation of TiS inclusion in the specimen C is most advantageous.

In the case of the specimens A and B, Al₂O₃ is thermodynamically the stable oxide at 1873 K as shown in Fig. 1. On the other hand, TiO_x is the stable oxide at the metal composition of the specimens C and D. Nevertheless, Al₂O₃ inclusions were also mainly observed in those specimens. It would be explained by that the holding time after Ti addition is too short to convert the firstly formed Al₂O₃ inclusions into the stable TiO_x inclusions completely. In the experimental conditions for the specimens C and D, the order of the additions of Al and Ti is significantly important for the formation of oxide inclusions.

In the inner layer, mostly TiN and Al₂O₃+TiN, and some Al₂O₃ and Al₂O₃+TiO_x inclusions were observed in the specimens A and C. In the case of the specimens B and D, TiN inclusions were mainly observed and some Al₂O₃ and Al₂O₃+TiN inclusions were also detected. The fraction of the number of TiN single-phase particles in the specimens A and C containing from 0.0005 to 0.0013 mass%N was 42 and 76% of observed inclusions, respectively. On the contrary, in the case of the specimens B and D, the fraction of the number of TiN inclusions was more than 87% of total inclusions. The increase in N content of the alloy resulted in the increase of independently precipitated TiN single phase particles.

3.2.2. Size of TiN Single Phase Particles in the Inner Layer

Figures 5 and 6 show the size distribution of TiN single phase particles in the inner layer at similar metal composition with different nitrogen content. The size of TiN single phase particles in the specimen B (0.0058 mass%N) is much larger than that in the specimen A (0.0005 mass%N). In the case of the specimens C and D, the same phenomenon was observed. This tendency indicates that the formation of larger TiN inclusions easily happened in the specimen with larger nitrogen content than smaller nitrogen content during quenching.

Fig. 5.

Size distribution of the observed TiN inclusions in the inner layer of the specimens A and B.

Fig. 6.

Size distribution of the observed TiN inclusions in the inner layer of the specimens C and D.

3.2.3. Formation of Al₂O₃–TiO_x Inclusions

In the specimens A and C with relatively small nitrogen contents, some (Al₂O₃+TiO_x)-based inclusions with the diameter of more than 3.5 μm were observed. Figure 7 shows the SEM images of the observed (Al₂O₃+TiO_x)-based inclusions in the specimens A and C. It consists of two phases, namely Al-rich oxide and Ti-rich oxide phase. Figure 8 shows mole fractions of Al, Ti and Fe in (Al₂O₃+TiO_x)-based inclusions in the specimen A on the Al–Ti–Fe ternary composition diagram calculated from measured compositions of Al₂O₃–TiO_x–FeO oxide inclusions, since the measured peaks are mainly Al, Ti and Fe. The oxide phase rich in Ti could be TiO_x, since the Al peak detected by EDS in Ti-rich phase would be from the Al-rich phase because of the fine structure of Ti-rich phase. In the case of the specimens B and D, (Al₂O₃+TiO_x)-based inclusions were not detected.

Fig. 7.

Observed (Al₂O₃+TiO_x)-based inclusions in as-cast specimens A and C.

Fig. 8.

Mole fraction of (Al₂O₃+TiO_x)-based inclusions in the specimen A on the Fe–Al–Ti ternary composition diagram.

3.3. Evolution Behavior of Inclusions during Heating at 1473 K

3.3.1. Type Change of Inclusions

Figure 9 shows the observed inclusion in the outer layer of the heated specimens. In all alloys, Al₂O₃ and Al₂O₃+TiN inclusions were mainly observed. In the specimen A, some Al₂O₃+TiO_x(+TiN) inclusions were observed. Some Al₂O₃+TiO_x+TiN, Al₂O₃+TiS, TiS and TiN inclusions were also detected in the specimen C. In the specimens B and D, some TiN inclusions were observed.

Fig. 9.

Observed inclusions in the outer layer of heated specimens.

Figure 10 shows the observed inclusion in the inner layer of the heated specimens. In the specimen A, mostly TiN, TiS and Al₂O₃+TiN inclusions were observed, and some TiN+TiS, Al₂O₃+TiS, Al₂O₃+TiN+TiS and Al₂O₃+TiO_x inclusions were also observed. In the case of the specimen B, almost all inclusions were TiN and some Al₂O₃+TiN inclusions were also detected. In the case of the specimen C, TiN, TiS and TiN+TiS were main inclusion. Some Al₂O₃, Al₂O₃+TiN(+TiS) and Al₂O₃+TiS inclusions were observed. TiN(+TiS), TiS and Al₂O₃+TiN(+TiS) inclusions were observed in the specimen D.

Fig. 10.

Observed inclusions in the inner layer of heated specimens.

In the outer layer, the fraction of the number of TiN-based inclusions increased in all specimens after heating. TiN formed during heating in all alloys. On the other hand, the fraction of the number of TiS-based inclusions increased in the inner layer by heating. The fraction of the number of TiS-based inclusions in the specimens A and C (0.0005 and 0.0013 mass%N, respectively) increased to 38 and 68%, respectively. On the other hand, TiS inclusions were not detected in the specimen B and observed TiS-based inclusions were 32% of total inclusions in the specimen D.

Comparing between the specimens B and D, TiS formation in the inner layer of the specimen D could be explained as follows. The significant difference between the inner layer of the specimen B and that of the specimen D is the existence of Al₂O₃ inclusions and metal compositions. Influence of the existence of Al₂O₃ inclusions was insignificant due to the number density of Al₂O₃ inclusion in the inner layer is very low in both specimens. Therefore, the reason of TiS formation in the specimen D would be the metal composition. Assuming the interaction parameters are inversely proportional to temperature and using those reported at 1873 K,^7,8) the activity of sulfur in the specimens B and D was calculated to be 0.00115 and 0.00159, respectively. In addition, the activity of titanium in the specimens B and D was 0.0908 and 0.120, respectively. Therefore, it is expected that TiS formation in the specimen D is much favorable than that in the specimen B.

However, only limited evolution of TiS-based inclusions was observed in specimens with larger nitrogen content and similar Al and Ti compositions. In the case of the specimens B and D, most of observed inclusions were still TiN-based inclusions. This difference is caused by the preferential evolution and growth of TiN than TiS at larger N content.

The nitrogen and sulfur contents equilibrated with solute Ti in each specimen were calculated by Eqs. (2) and (3) for solubility products of TiN and TiS.^7,9) By using these thermodynamic data and analyzed compositions, the supersaturation ratios of TiN and TiS, SR_TiN and SR_TiS, defined as Eqs. (4) and (5), respectively, were calculated.   

$$\begin{array}{c}\text{TiN}\left(\text{s}\right)=\text{Ti}{\left(\text{mass}\%\right)}_{\text{in}\mathrm{\hspace{0.17em} }\gamma }+\text{N}{\left(\text{mass}\%\right)}_{\text{in}\mathrm{\hspace{0.17em} }\gamma }\\ \text{log}{\left[\text{mass}\%\text{Ti}\right]}_{\text{eq}. \text{in}\mathrm{\hspace{0.17em} }\gamma }^{}{\left[\text{mass}\%\text{N}\right]}_{\text{eq}. \text{in}\mathrm{\hspace{0.17em} }\gamma }^{}=–\frac{13\mathrm{\hspace{0.17em} }860}{T}+3.75\end{array}$$(2) ⁷⁾

  

$$\begin{array}{c}\text{TiS}\left(\text{s}\right)=\text{Ti}{\left(\text{mass}\%\right)}_{\text{in}\mathrm{\hspace{0.17em} }\gamma }+\text{S}{\left(\text{mass}\%\right)}_{\text{in}\mathrm{\hspace{0.17em} }\gamma }\\ \text{log}{\left[\text{mass}\%\text{Ti}\right]}_{\text{eq}. \text{in} \gamma }^{}{\left[\text{mass}\%\text{S}\right]}_{\text{eq}. \text{in} \gamma }^{}=–\frac{13\mathrm{\hspace{0.17em} }975}{T}+5.43 \end{array}$$(3) ⁹⁾

  

$$S{R}_{\text{TiN}}=\frac{\left[\text{mass}\%\text{Ti}\right]\left[\text{mass}\%\text{N}\right]}{{\left[\text{mass}\%\text{Ti}\right]}_{\text{eq}. \text{in}\mathrm{\hspace{0.17em} }\gamma }{\left[\text{mass}\%\text{N}\right]}_{\text{eq}. \text{in}\mathrm{\hspace{0.17em} }\gamma }}$$(4)

  

$$S{R}_{\text{TiS}}=\frac{\left[\text{mass}\%\text{Ti}\right]\left[\text{mass}\%\text{S}\right]}{{\left[\text{mass}\%\text{Ti}\right]}_{\text{eq}. \text{in}\mathrm{\hspace{0.17em} }\gamma }{\left[\text{mass}\%\text{S}\right]}_{\text{eq}. \text{in}\mathrm{\hspace{0.17em} }\gamma }}$$(5)

where [mass%Ti]_{eq. in γ}, [mass%N]_{eq. in γ} and [mass%S]_{eq. in γ} are, respectively, Ti, N and S contents in γ iron equilibrated with solid TiN or TiS.

The supersaturation ratio of TiN in the specimens A, B, C and D were 20, 246, 65 and 309, respectively. On the other hand, the supersaturation ratio of TiS in each specimen was in the range from 1.2 to 2.5. Although the supersaturation ratios required for TiN and TiS precipitation in γ iron at 1473 K have not been reported so far, TiN precipitation in the specimens B and D would be preferable than TiS precipitation during heating. Therefore, TiN-based inclusions were still mainly observed after heating in the specimens B and D.

3.3.2. Size Change of TiN in the Inner Layer

Figure 11 shows the average size of TiN inclusions in the inner layer of all as-cast and heated specimens. In the specimens A, B and D, the average size of TiN inclusion in the inner layer increased by heating. Therefore, it can be concluded that TiN particles grew during heating.

Fig. 11.

Average size of TiN inclusions in the inner layer of all as-cast and heated specimens.

Maximum size of TiN inclusions in the specimens B and D with larger nitrogen content was much larger than that in specimens with smaller nitrogen content. From this result, it is considered that the growth of TiN inclusions occurred more easily in metal with larger nitrogen content than smaller nitrogen content during heating.

In the case of the specimen C, the average size of TiN inclusions decreased by heating due to the decrease of the observed number of TiN single phase inclusions. The fraction of the number of TiN+TiS inclusions substantially increased from 0 to 33% during heating by TiS formation on the surface of TiN inclusions.

From above results, it is suggested that around 0.005 mass% of nitrogen content is necessary to form TiN phase on the Al- or Ti-based oxide inclusions preferentially than TiS during heating at 1473 K in the case of less than 0.002 mass% of sulfur content in the Al-killed Ti alloyed steel.

3.3.3. Size Change of TiN Phase in Al₂O₃+TiN Inclusion in the Outer Layer

Figure 12 shows the difference between the average size of Al₂O₃+TiN inclusions and that of Al₂O₃ core in Al₂O₃+TiN inclusions, which is l_TiN in Eq. (6).   

$$l_{\mathrm{T}\mathrm{i}\mathrm{N}}=\left({\text{size of Al}}_{\text{2}}{\text{O}}_{\text{3}}+\text{TiN inclusion}\right)–\left({\text{size of Al}}_{\text{2}}{\text{O}}_{\text{3}}{\text{ core of Al}}_{\text{2}}{\text{O}}_{\text{3}}+\text{TiN inclusion}\right)$$(6)

From that, the change of TiN phase size in the Al₂O₃+TiN inclusion by heating can be determined. The tendency of size distribution of Al₂O₃ phase (Al₂O₃ single phase and Al₂O₃ phase in Al₂O₃+TiN) did not change in all specimens except the specimen C. At the smaller nitrogen content such as the specimens A and C, the difference between two sizes slightly increased or decreased by heating. On the contrary, in metals with larger nitrogen content such as the specimens B and D, the difference certainly increased by heating. Since the fraction of the number of Al₂O₃+TiN inclusions increased in all specimens as previously mentioned in 3.3.1, TiN phase newly formed on the Al₂O₃ inclusion during heating. In addition, TiN phase of Al₂O₃+TiN inclusions formed during quenching also grew by heating.

Fig. 12.

Average of l_TiN in all as-cast and heated specimens.

4. Conclusions

The observation of inclusions formed by Al–Ti deoxidation has been conducted for specimens with different metal compositions such as Al and N contents. The oxide inclusions were mostly found at the edge and TiN inclusions were observed at the inner part of as-cast specimens. The effect of Al content on the formation of inclusion during heating was not observed. After heating, the size of TiN single phase particle and that of TiN phase in Al₂O₃+TiN inclusion increased. In addition, increase in N content resulted in the enhancement of single phase TiN inclusion formation or the precipitation of TiN on the surface of oxide inclusions and inhibition of TiS formation. Therefore, in the case of less than 0.002 mass% of sulfur content in the Al-killed Ti alloyed steel, around 0.005 mass% of nitrogen content is necessary to form TiN phase on the Al- or Ti-based oxide inclusions preferentially than TiS during heating at 1473 K.

References

• 1)   N.  Kikuchi,  S.  Nabeshima,  Y.  Kishimoto,  T.  Matsushita and  S.  Sridhar: ISIJ Int., 47 (2007), 1255.
• 2)   S. C.  Park,  I. H.  Jung,  K. S.  Oh and  H. G.  Lee: ISIJ Int., 44 (2004), 1016.
• 3)   J. H.  Park: Calphad, 35 (2011), 455.
• 4)   D.  Yu,  D. P.  Dunne,  T.  Chandra and  F. J.  Barbaro: Mater. Trans. JIM, 37 (1996), 1554.
• 5)   H.  Fujimura,  S.  Tsuge,  Y.  Komizo and  T.  Nishizawa: Tetsu-to-Hagané, 87 (2001), 707.
• 6)   H.  Matsuura,  K.  Nakase and  F.  Tsukihashi: CAMP-ISIJ, 23 (2010), 956.
• 7)   M.  Hino and  K.  Ito (Eds.): Thermodynamic Data for Steelmaking, Tohoku University Press, Sendai, (2010), 42, 236, 255, 259.
• 8)   W. Y.  Kim,  J. O.  Jo,  C. O.  Lee,  D. S.  Kim and  J. J.  Pak: ISIJ Int., 48 (2008), 17.
• 9)   X.  Yang,  D.  Vanderschueren,  J.  Dilewijns,  C.  Standaert and  Y.  Houbaert: ISIJ Int., 36 (1996), 1286.