High-nitrogen steels (HNS) are a family of stainless steel grades combining excellent corrosion resistance with desirable mechanical properties. Owing to the progress in processing technologies, some commercial grades of HNS are available for applications in power generation and oil field industries as well as biomaterials.[1,2] However, the following factors have raised obstacles to wider application of HNS: the need for high-pressure melting because of the low solubility of nitrogen in iron melts and the unusual brittle fracture at high-nitrogen contents above 0.7 wt pct.[3] A new alloy system, using carbon (C) as a major alloying element together with nitrogen (N), has been proposed to overcome the aforementioned difficulties and to obtain the excellent properties in a cost-effective way.[3–5] With regards to the alloying of C in austenitic stainless steels, many researchers[1] have pointed out the drawbacks associated with the degradation in the corrosion resistance resulting from the formation of harmful carbides. Recent investigations, however, have shown that simultaneous alloying of N and C (N + C) not only increases the strength level effectively[3,4] but also improves the resistance to localized corrosion.[5]

Precipitation behavior has been regarded as an important feature of austenitic stainless steels, because their mechanical and corrosion properties are profoundly dependent on the formation of second phases.[6,7] During isothermal aging in the range of 973 K to 1273 K (700 °C to 1000 °C), several types of carbides, intermetallic compounds, and nitrides tend to precipitate on grain boundaries (GB) as well as on the grain interior of an austenite (γ) matrix.[6,7] In commercial austenitic stainless steel such as AISI 304, the major precipitates are known as M23C6-type carbide (M contains metal atoms such as Cr, Mo, and Fe) and intermetallic sigma (σ) phase, and extensive literature is available that is associated with their formation,[6–8] crystallography,[9] and related property change, especially sensitization and deterioration in mechanical properties.[7] Precipitation characteristics, however, have been considered to be a complicated phenomenon in that small changes in the chemical composition, thermomechanical processing, and aging conditions can significantly influence the formation of second phases, and in some cases, it is not easy to distinguish one type of precipitate from another by only their morphologies.[7,9]

In austenitic HNS, the precipitation behavior has been studied with a focus on the effect of nitrogen addition on the precipitation reaction of the second phases,[9,10] and the precipitation of M2X-type (M = Cr, Mo, Fe, and Mn; X = N, C) nitride, simply designated as Cr2N,[11,12] together with its detrimental effects on properties of HNS.[13–15] The present authors showed that the Cr2N precipitation occurred sequentially at grain boundaries (intergranular), by cellular, and by intragranular form within the matrix.[11] The time-temperature-precipitation behavior could be summarized as follows: (1) the high-temperature region [above 1223K (950 °C)] had mainly coarse intergranular Cr2N; (2) the nose temperature region was intergranular Cr2N → cellular Cr2N → intragranular Cr2N + sigma (σ); and (3) the low-temperature region [below 1023 K (750 °C)] was intergranular Cr2N → cellular Cr2N → intragranular Cr2N + σ + chi (χ) + M7C3 carbide.[14]

To our knowledge, however, the investigation on the precipitation of N + C austenitic stainless steel has not been conducted yet, although the precipitation reaction can be varied by the presence of alloyed C. In the present investigation, the precipitation behavior of new N + C austenitic stainless steel is investigated in terms of the precipitation sequence of second phases and their crystallographic features. Some similarities and differences in the precipitation behavior between N and N + C austenitic stainless steels are briefly discussed.

The high-interstitial-alloyed (HIA) austenitic stainless steel, whose chemical composition in weight percent was 18.12Cr; 9.66 Mn; 0.22Si; 0.01Mo; 0.38N; and 0.38C, was fabricated using a commercial vacuum induction melting furnace (VIM 4 III-P; ALD, Berlin, Germany). After homogenization at 1523 K (1250 °C) for 2 hours under argon atmosphere, the ingots were hot rolled into sheets 4-mm thick, followed by air cooling. The specimens were solution-annealed at 1423 K (1150 °C) for 1 hour in the γ single-phase region[4] and then quenched in water. They were isothermally aged at 1123 K (850 °C) for various aging times from 10 to 106 seconds in argon atmosphere and quenched in water. After aging, the specimens were polished and chemically etched using a Glyceregia reagent (20 mL hydrochloric acid, 10mL nitric acid, and 30mL glycerin). The overall precipitation behavior of the aged specimens was observed using a scanning electron microscope (SEM; JSM- 5800, JEOL, Japan) operating at 20 kV. To identify the precipitated phase and change in its morphology, microscopic observations using scanning transmission electron microscopy (STEM, JEM 2100F; JEOL, Japan) with an acceleration voltage of 200 kV were conducted on each aged specimen. The thin foils for STEM observation were prepared in a twin-jet electrolytic polishing apparatus using a solution containing 15 pct perchloric acid and 85 pct methanol. The detailed analyses of selected area diffraction (SAD) patterns were carried out using Desktop Microscopist V2.2 software (Lacuna Lab).[11]

Figure 1 shows a series of SEM micrographs of Fe-18Cr-10Mn-0.38N-0.38C alloy aged at 1123 K (850  °C) for various aging times. At an early stage of aging (103 seconds), the second phases with bright gray contrast preferentially form along grain boundaries (intergranular precipitation, Figure 1), and the cellular (or discontinuous) precipitation starts to form in some boundaries, as indicated by arrows in Figure 1. Almost all the grain boundaries are covered with second phases at this aging time of 103 seconds. As aging time continues, the cellular precipitation comprising alternating precipitate and γ matrix prevails (cellular precipitation), and the volume fraction of cellular colonies overall increases, as shown in Figure 1. Although the cellular precipitation of second phases is a dominant form in this stage of aging, precipitation also occurs along annealing twin boundaries and gradually grows along the boundaries. Note that the cellular precipitation does not occur in all grains, although the areas covered by cellular precipitation (cellular colonies) have expanded.

Fig. 1
figure 1

SEM micrographs showing the change in morphology of precipitates formed during isothermal aging at 1123 K (850 °C): intergranular precipitation (103 s, cellular or discontinuous precipitation starts to form in some boundaries, as indicated by arrows), cellular precipitation (104 and 105 s), and intragranular precipitation indicated by dotted circles (106 s), respectively

Full size image

Long-term aging (106 seconds) at 1123 K (850 °C) led to the changes in precipitation behavior of second phases, and a typical example showing intragranular precipitation (indicated by dotted circle) is shown in Figure 1. Figure 2 represents the magnified SEM images of intragranular precipitates and other type of precipitate formed after 106 seconds of aging. In the case of grains that are free from the cellular precipitation, the small rod-like precipitates form within grain interior and they are aligned parallel to the specific directions relative to the γ matrix (intragranular precipitation). In addition to the intragranular type of precipitates, another type of second phase is also observed in the form of coarse and irregular shapes with darker contrast (indicated by arrows in Figure 2(b)).

Fig. 2
figure 2

Magnified SEM micrographs of (a) intragranular precipitates and (b) other types of precipitates indicated by arrows taken from the specimen aged at 1123 K (850 °C) for 106 s

Full size image

Based on the SEM observations shown earlier, the precipitation occurs in a sequence of intergranular (along grain boundary) → cellular (or discontinuous) → intragranular (within grain interior) form with aging time. Although a detailed description of the precipitation reaction including the formation of cellular precipitates requires more study, the precipitation sequence and morphology of second phases observed in HIA was similar to that of HNS, as previously reported in the literature.[12–15]

Figure 3(a) shows a STEM bright-field (BF) image and corresponding SAD pattern of intergranular precipitates taken from the specimen aged at 1123 K (850 °C) for 103 seconds. The intergranular precipitates formed at an early stage of aging are identified as M23C6 carbide. Its crystal structure is a face-centered cubic (fcc) structure (space group: \( {\text{Fm}}\bar{3}{\text{m}} \)) with a lattice parameter of 1.060 nm.[6–9] The M23C6 is crystallographically related to the lattice of one of the adjacent γ grains in accordance with the cube–cube orientation relationship (OR).

Fig. 3
figure 3

STEM BF images and corresponding SAD patterns of M23C6: (a) intergranular M23C6, (\({\text{z}} = \left[ {\bar{1}12} \right]_{{{\upgamma}}} \parallel \left[ {\bar{1}12} \right]_{{{\text{M}}_{23} {\text{C}}_{6} }} \), only diffraction spots from M23C6 are indexed) taken from a specimen aged at 1123 K (850 °C) for 103 s and (b) cellular M23C6 (\({\text{z}} = \left[ {110} \right]_{{{\upgamma}}} \parallel \left[ {100} \right]_{{{\text{M}}_{23} {\text{C}}_{6} }} \), only diffraction spots from γ are indexed) taken from a specimen aged at 1123 K (850 °C) for 104 s, respectively

Full size image

The STEM micrograph of cellular precipitation comprising alternating γ matrix and precipitate and a corresponding SAD pattern are depicted in Figure 3(b). Based on the analyses of the SAD pattern, the cellular type precipitates are also determined to be M23C6 carbide, and they are also related to the γ by the cube–cube OR—\( \left[ {001} \right]_{{{\upgamma}}} \parallel \left[ {001} \right]_{{{\text{M}}_{23} {\text{C}}_{6} }} \), and \( \left( {100} \right)_{{{\upgamma}}} \parallel \left( {100} \right)_{{{\text{M}}_{23} {\text{C}}_{6} }} \), which is consistent with previous reports for the M23C6 reported in commercial austenitic stainless steels such as AISI 304.[6–8] As aging time increases, the self-coarsening of M23C6, perhaps accompanied by partial dissolution, is also similar to the spheroidization of cementite in pearlitic steels.[16] From Figures 3(a) and (b), it is interesting to note that the stacking faults (SFs) are frequently observed in the vicinity of the M23C6 carbide, and the SFs formed near cellular M23C6 are aligned parallel to other {111} planes of γ that were not occupied by the M23C6 carbide. Although the morphology and sequence of precipitation of second phases in austenitic HNS are both similar to that of the investigated alloy, the formation of SFs near Cr2N precipitates has not been reported in the literature.[12–15]

The intragranular precipitation of M23C6 carbides preferentially occurred in the grains where the cellular precipitation of M23C6 did not take place. Figure 4 shows a STEM BF image and corresponding SAD patterns depicting the intragranular M23C6 taken from the specimen aged at 1123 K (850 °C) for 105 seconds. The intragranular M23C6 are related to each other by two distinct variants with a peculiar OR. In addition, the formation of SFs in γ matrix is much more noticeable in the vicinity of intragranular M23C6, and the longitudinal direction of SF is parallel to the growth direction of the intragranular M23C6. The observed variants of intragranular M23C6 have two different zone axes (i.e.,\( \left[ {001} \right]_{{{\text{M}}_{23} {\text{C}}_{6} }} \) Figure 4(b) and \( \left[ {22\bar{1}} \right]_{{{\text{M}}_{23} {\text{C}}_{6} }} \) Figure 4(c), and both of them are parallel to [001]γ of γ matrix). The latter zone axis can be obtained using a transformation matrix for twinning[2,17]; namely, two variants are related to each other by the following twin relation:

Fig. 4
figure 4

(a) STEM BF image of intragranular M23C6 and SAD patterns of M23C6 showing (b) cube–cube orientation taken from a specimen aged at 1123 K (850 °C) for 105 s, (c) twin-related M23C6, (d) two variants, and (e) calculated SAD pattern (filled circles represent the spots from the twin-related variant), respectively

Full size image
$$ \frac{1}{3}\left[ {\begin{array}{*{20}c} {\bar{1}} & 2 & 2 \\ 2 & {\bar{1}} & 2 \\ 2 & 2 & 1 \\ \end{array} } \right]\left[ {\begin{array}{*{20}c} 0 \\ 0 \\ 1 \\ \end{array} } \right]\, = \,\left[ {\begin{array}{*{20}c} 2 & 2 & {\bar{1}} \\ \end{array} } \right] $$

Moreover, the characteristic Moiré fringe is observed around the interface between intragranular M23C6 and γ matrix, suggesting that the misfit strain might be generated in the vicinity of γ matrix during the formation of intragranular M23C6.

On prolonged aging of 106 seconds at 1123 K (850 °C), other types of precipitates formed mainly in the γ grain interior. Figure 5(a) shows the BF image of intragranular precipitation of M2X-type (M = Cr, Mo, Fe, and Mn; X = N and C) nitride and its SAD pattern taken from the specimen aged at 1123 K (850 °C) for 106 seconds. Based on the analysis of the SAD pattern, the M2X-type nitride (simply designated as Cr2N) has a trigonal structure (space group:\( {\text{P}}\bar{3}1{\text{m}} \)) with lattice parameters of a = 0.4800 and c = 0.4472 nm and is related to γ matrix by the following crystallographic OR: \( \left[ {\bar{1}10} \right]_{{{\upgamma}}} \parallel \left[ {\bar{1}100} \right]_{{C{\text{r}}_{2} {\text{N}}}} \) and \( \left( {111} \right)_{{{\upgamma}}} \parallel \left( {0001} \right)_{{{\text{Cr}}_{2} {\text{N}}}} \). A detailed description of the crystal structure and crystallography of Cr2N is given in our previous work.[11]

Fig. 5
figure 5

STEM BF images and corresponding SAD patterns of Cr2N and sigma phase taken from a specimen aged at 1123 K (850 °C) for 106 s; (a) BF image and SAD pattern \(({\text{z}} = \left[ {\bar{1}10} \right]_{{{\upgamma}}} \parallel \left[ {\bar{1}100} \right]_{{{\text{Cr}}_{ 2} {\text{N}}}}) \) of Cr2N, and (b) BF image and SAD pattern (z = [112] σ ) of sigma phase

Full size image

In addition to the intragranular Cr2N, another precipitate with coarse irregular morphology was also observed in the cell boundaries as well as γ grain interior, as presented in Figure 5(b). Based on the analysis of the SAD pattern, the precipitate is identified as an intermetallic σ phase that has a tetragonal structure (space group: \( {\text{P}}4_{2} /{\text{mnm}} \)) with a lattice parameter of a = 0.880 and c = 0.454 nm.[6] The formation of σ phase is associated with the C- or N-depleted zone adjacent to the M23C6 in commercial austenitic[6] or Cr2N in austenitic HNS. The formation of σ phase in HIA can also be explained by the previous mechanism in which the interstitial-depleted zone close to main precipitates induced the formation of σ phase.[18]

For the austenitic HNS, the main precipitate formed during isothermal aging has been reported to be Cr2N, and its precipitation occurs in the forms of intergranular, cellular, and intragranular precipitation with aging time.[12–15] In contrast, the addition of C to the austenitic stainless steel with a similar composition changed the type of major precipitate from M2N to M23C6, although the ratio of carbon to nitrogen in the alloy is 1:1. But the precipitation sequence and morphology of second phases in HIA were similar—intergranular → cellular → intragranular form as shown in Figure 1.

The initial formation of second phases along the grain boundary is regarded as a common feature of precipitation in austenitic stainless steels because the grain boundaries are preferential sites for the diffusion of alloying elements.[7] After almost complete coverage of second phases along grain boundaries, the precipitation in the next stage proceeds into the discontinuous or intragranular form depending on the alloy system. Manna et al. performed a systematic study on the discontinuous precipitation.[19] Owing to its practical importance related to the degradation of steel properties, extensive studies on the cellular (or discontinuous) precipitation of Cr2N in austenitic HNS has been performed,[1,12–15,20–23] and the following distinct models have been suggested: long-range diffusion of nitrogen in Cr-Ni steels[12] and intergranular and bulk diffusion of chromium in Cr-Mn steels.[20,21] Kikuchi et al.[12] showed that the cellular precipitation of Cr2N was a non-steady-state growth process in which the migration of cell boundaries decelerated with aging time and terminated even though N supersaturation still remained in the untransformed matrix. In contrast, Presser and Silcock[20] envisaged that the nucleation rather than growth of cellular precipitation was the controlling step, and that the bulk diffusion of Cr governed the volume fraction of cellular precipitates. Later, Vanderschaeve et al.[21] reported that two processes, first the intergranular diffusion of Cr and then its bulk diffusion, governed the cellular precipitation of Cr2N. With regard to the cellular (or discontinuous) precipitation of M23C6 in austenitic stainless steels, limited results are available, but they are mainly related to the phenomenological analysis dealing with the identification of cellular precipitates.[6–8] Voice and Faulkner[22] reported the discontinuous precipitation of M23C6 in Nimonic 80A and considered it a pseudo-discontinuous precipitation because of partial carbon–diffusion controlled growth of precipitates. The discontinuous precipitation of M23C6 showed a strong dependence on grain-boundary nucleation as well as on the growth direction in the receding grain. Although the detailed description of the nature of cellular precipitation on M23C6 carbide requires more work and is beyond the scope of this study, the cellular precipitation of M23C6 in HIA may be understood based on the non-steady-state growth process controlled by an interstitial element (N or C), as suggested by Kikuchi et al.[12] as well as by Voice and Faulkner.[22] Another possible reason can also be found in the orientation of the grain boundary. Matsuoka et al.[23] reported that the cellular precipitation of Cr2N in 19Cr-5Ni steel was caused by grain boundaries with misorientation angles higher than 22 deg, that the cellular colonies grew faster, and that a better organized two-phase lamellar structure developed as the boundary misorientation increased. A similar dependence on cellular precipitation of the misorientation angles of grain boundaries was also confirmed based on a statistical approach for the EBSD measurements over 40 grains with cellular precipitation, but the preferential occurrence of cellular precipitation was observed in misorientation angles around 40 deg, which is known as a grain boundary with higher mobility in fcc materials.[24] However, for the comprehensive understanding of the nature of cellular precipitation observed in HIA, systematic studies should be carried out.

In contrast to the precipitation behavior of austenitic HNS, it is interesting to note that the characteristic Moiré fringe usually appears along the interface between M23C6 and γ matrix. Because the two phases have the same crystal structure (fcc) and because the perfect correspondence between two lattices exists following the cube-on-cube OR,[6–9] even though the lattice parameter of M23C6 carbide is three times larger than that of austenite, the difference in lattice parameters can be considered a probable source of SF formation. In the case of aged HNS, the morphology and sequence of precipitation of second phases are both similar to those of the investigated alloy, but the formation of SFs did not prevail for austenitic HNS even though the sigma phase was also observed in prolonged aging condition. Therefore, it is conceivable that the formation of SFs is closely related to M23C6 carbide rather than sigma phase, although the volume change induced by the formation of hard sigma phase can also be a possible source of SF formation.

Precipitation behavior of austenitic stainless steel containing both N and C was investigated and could be summarized as follows:

  1. 1.

    The main precipitate was M23C6 carbide, and its morphology changed in a sequence of intergranular → cellular → intragranular form with increasing aging time.

  2. 2.

    The intragranular M23C6 carbides have twin-related variants, and their formation is accompanied by the formation of Moiré fringe at the M23C6/γ interface and SF surrounding γ matrix.

  3. 3.

    Additional aging produced two types of precipitates—Cr2N and intermetallic σ phase.