(IUCr) Identification and characterization of a novel Mn-N nitride formed in Fe-Mn-N alloy

Volume 36
Part 1
Pages 103-108
February 2003

Received 30 April 2002
Accepted 4 November 2002

Identification and characterization of a novel Mn-N nitride formed in Fe-Mn-N alloy

M. Gouné,^a A. Redjaïmia,^a ^* T. Belmonte^a and H. Michel^a

^aLaboratoire de Science et Génie des Surfaces, UMR-CNRS 7570, Ecole des Mines de Nancy, Parc de Saurupt, 54042 Nancy Cedex, France
Correspondence e-mail: redjamia@mines.u-nancy.fr

An unexpected phase, formed throughout the ferritic matrix of an Fe-Mn (1.62 wt% Mn) alloy during a nitriding treatment at 843 K for 8 h, is analysed. This new phase, labelled [theta] ', is a metastable Mn nitride. It adopts the form of plates, with length 200 nm and width 20 nm. Its crystal structure is established by electron microdiffraction in conjunction with group-theory analysis. This nitride crystallizes in the tetragonal system and belongs to the space group P 4₂/m 2/m 2/c with the following lattice parameters: $[a_{\theta'}]$ = 2.876 Å and $[c_{\theta'}]$ = 5.752 Å. The [theta] '-phase crystal lattice is oriented with respect to the surrounding ferritic matrix according to the cube-on-cube orientation relationship, namely: $[(100)_{\theta'} || (200)_\alpha]$ , $[(010)_{\theta'} || (020)_\alpha]$ and $[(002)_{\theta'} || (002)_\alpha]$ .

Keywords: alloys; nitrides; microalloyed steel; electron microdiffraction; group theory.

1. Introduction

The attractive properties of microalloyed steels combined with nitrides make them excellent candidate materials for nitriding applications (Baird, 1966; Lightfoot & Jack, 1975; Jack, 1975). Diffusing nitrogen interacts with the alloying elements to form their respective nitrides as a fine dispersion (Mittemeijer & Jack, 1985). These elements, notably chromium, vanadium and manganese, under most circumstances combine preferentially and the induced precipitates are an important source of strengthening (Podgurski & Davis, 1981). Such behaviour is related to the interaction of substitutional alloying elements, their concentration, the strength of the thermodynamic interaction between nitrogen and substitutional solutes, and finally to the nitriding temperature.

The precipitation of manganese nitride in iron has received considerable attention (Baird, 1966; Enrietto, 1962; Jonsson-Holmqvist et al., 1973; Sato et al., 1988). Indeed, homogeneous precipitation has been studied in Fe-Mn (0.2-4 wt%) alloys aged under constant nitrogen conditions in NH₃-H₂ gas mixtures in the temperature range 753-853 K. It is well established that this nitriding treatment leads to a plateau in the hardness profile, the origin of which is related to the precipitation of a [theta] -Mn₆N₅ phase. The latter takes place in the ferritic matrix, forming disc-shaped particles. However, the nitriding treatment of the ferritic matrix of Fe-1.62 wt% Mn alloy at 843 K during 8 h exhibits a hardness profile with a sharp peak. As far as we are aware, this peak phenomenon is the first of its kind to be observed in this field. During the microstructural analyses, we found that, in addition to the [theta] -Mn₆N₅ phase, the hardening peak is induced by an unexpected phase, which is formed throughout the ferritic matrix. This new phase, labelled [theta] ' in this investigation, is a metastable Mn-N nitride, which is detectable only by transmission electron microscopy (TEM) and is not observed either in an optical microscope or by X-ray diffraction because of the small size and amount of the precipitate.

In order to understand the formation of this [theta] '-phase and its effect on the mechanical properties of the alloy, we present, in this paper, its crystal structure including the point group, lattice parameters and space group. The crystal structure is established by electron microdiffraction in conjunction with group-theory analysis.

2. Material and experimental procedure

High-purity iron-manganese targets were prepared by the Goodfellow Laboratory. The chemical composition of the as-received specimens is given in Table 1. After mechanical polishing, different samples were salt-bath nitrided at 843 K for 20 h and subsequently water-quenched to avoid phase transformations during cooling.

Table 1
Chemical composition (wt%) of the as-received Fe-Mn alloy

Fe	Mn	Si	Mo	Ti	N	C	S	P
Balance	1.62	0.015	0.010	0.005	<0.001	<0.001	0.004	0.005

Metallographic observations of these samples show that during the nitriding treatment, two kinds of layers, [epsilon] -Fe₂N_1-x and [gamma] '-Fe₄N, were developed above the diffusion zone, the [alpha] -Fe layer. In detailed studies, efforts were made to characterize these two nitrides. As part of the continuing investigation, here we are only concerned with the precipitation phenomenon occurring throughout the diffusion zone.

Thin foils were prepared for transmission electron microscopy (TEM) by grinding the different slices to 20 µm thickness and by double-jet electropolishing in a solution of 5% perchloric acid in 95% II-butoxyethanol at 40 V potential. They were investigated in a Philips CM12 transmission electron microscope operated at 120 kV employing a cold stage and double-tilt goniometer. The diffraction patterns were obtained with a nearly parallel beam focused on a very small area of the thin foil. An X-ray diffractometer, equipped with a source of Co K [alpha] radiation, was employed to determine with accuracy the lattice parameter of the body centred cubic (b.c.c.) ferritic matrix, i.e. a = 2.876 Å.

The hardness profiles of these nitrided specimens exhibit an unexpected and sharp peak at 100 µm from the interface between the [gamma] '-Fe₄N compound layer and the diffusion zone. In order to explain the origin of this maximum hardness, transmission electron microscopy and electron diffraction investigations have been conducted. The latter investigations have been performed on thin slices cut between 10 and 150 µm from the [gamma] '/ [alpha] interface and over 170 µm. The hardness profile after nitriding is depicted in Fig. 1. The nitrided Fe-1.62 wt% Mn substrate shows a smooth transition between the nitrided and non-nitrided zones. We can divide the profile into two regions: region I, localized near to and around the peak of hardness, and region II, localized after it, as indicated in Fig. 1.

Figure 1
Vickers hardness values of Fe-1.62 wt% Mn nitrided for 8 h at 843 K as a function of depth (µm).

3. Experimental results and discussion

3.1. Microstructural features

The nitriding of the Fe-1.62 wt% Mn samples at 843 K during 20 h originates in very fine precipitation. The latter is composed of two types of intragranular particles in the b.c.c. ferritic matrix, in the area close to the diffusion zone and beneath the [gamma] '-Fe₄N layer. This precipitation is related to the maximum depicted on the hardness profile (Fig. 1). It has been pointed out that over 170 µm, one type of precipitate has a disc shape in the observation plane, with a mean size of about 300 nm (Fig. 2). The disc-shaped particles have been the subject of studies in different papers (Baird, 1966; Enrietto, 1962; Jonsson-Holmqvist et al., 1973; Sato et al., 1988). At 843 K, their morphology as well as their size is a characteristic of the equilibrium state of manganese nitride. However, the nature and the composition of this precipitate have not been absolutely elucidated (Baird, 1966; Jonsson-Holmqvist et al., 1973; Sato et al., 1988). Since these studies, the Fe-Mn-N system has been assessed both experimentally and theoretically (Huang, 1989; Otsuka et al., 1977; Kreiner & Jacobs, 1992). The iron-rich corner of the Fe-Mn-N ternary phase diagram has been calculated from data reported in a paper published elsewhere (Gouné et al., 2002) and confirmed by the use of the Thermocalc database. It is pointed out in Fig. 3 that the most stable nitride under our conditions (temperature, pressure and composition) is clearly [theta] -Mn₆N₅. This precipitate crystallizes in the hexagonal lattice and belongs to space group P 6/m m m with the following lattice parameters: a = 4.211 Å and c = 4.145 Å. Thus, it can reasonably be concluded that disc-shaped precipitates correspond to the [theta] -Mn₆N₅ nitride.

Figure 2
Bright-field TEM image showing [theta]

-Mn₆N₅ particles precipitated throughout the ferritic matrix in connection with the region II of Fig. 1

Figure 3
Isothermal section of the Fe-rich corner of the Fe-Mn-N phase diagram at 843 K calculated by Gouné et al. (2002

[Gouné, M., Belmonte, T., Redjaïmia, A., Weisbecker, P., Fiorani, J. M. & Michel, H. (2002). Mater. Sci. Eng. A. In the press.]

) and confirmed by the use of the Thermocalc database.

The second type of precipitate is formed by an unexpected compound, labelled [theta] ' in this work. It adopts the form of plates of length 200 nm and width 20 nm (Figs. 4, 5a and 6). Beyond a distance of about 170 µm, the [theta] ' precipitates disappear in favour of the [theta] phase. This behaviour is reflected in the decrease in the hardness profile. The habit plane of the [theta] ' precipitates has been determined by carrying out tilting experiments in the transmission electron microscope to determine the electron beam orientation at which the plate/matrix interface has a minimum thickness. This orientation is accurately parallel to $[\langle 001\rangle_{\alpha}]$ , indicating that the habit plane favoured for precipitation is parallel to $[\{ 001\}_{\alpha}]$ . The growth directions of the [theta] ' plates were determined by trace analysis (Figs. 5a and 5b). Such analyses, involving a number of orientations of the ferritic matrix, established that the growth directions of the [theta] ' particles are $[\langle 001\rangle _{\alpha}]$ . Thus the number of variants of the plate-shaped particles in a given matrix grain is equal to three (n = 3). This result is also corroborated by the extent of the intensity streaking in the $[ [ 200 ]^*_{\alpha} ]$ and $[ [ 020 ]^*_{\alpha}}]$ directions in the diffraction patterns recorded along the $[ [ 001 ]_{\alpha}]$ zone axis (Fig. 5b). The streaks arising from the [theta] '-phase particles and running completely through each reflection, forming a cross-grid, are much larger than the kinematical theory would predict (Edington, 1974). This effect is related to the specific shape and to the high density of the [theta] '-phase particles. The TEM micrograph (Fig. 6) recorded along $[[001]_\alpha]$ shows three mutually orthogonal variants, two of which are clearly visible while the third one is edge-on as indicated by circles in Fig. 6.

Figure 4
Bright-field TEM image showing the [theta]

' phase and [theta]

phase in region I of Fig. 1

Figure 5
(a) Centred dark-field TEM image showing [theta]

'-phase particles. (b) Related composite diffraction pattern recorded along the [001] zone axis of the ferritic matrix. The reflections with a low intensity belong to the [theta]

' phase. The reflection at half the distance between 000 and 020 is used for the dark-field image.

Figure 6
Bright-field TEM image showing three mutually orthogonal variants, two of which are clearly visible while the third one is edge-on as indicated by circles.

The nitriding treatment (843 K, 8 h) followed by an ageing at 843 K during 20 h leads to a dissolution of [theta] ' plates and a coarsening or impingement of the disc-shaped [theta] phase (Fig. 7). This observation indicates that the [theta] ' phase is a metastable product of the ferritic matrix.

Figure 7
Bright-field TEM image showing a coarsening or impingement of the disc-shaped [theta]

-phase particles and dissolution of [theta]

' plates after an ageing heat treatment at 843 K during 20 h.

3.2. Structure of the metastable ' nitride

The absence of high-order Laue zones and the extent of intensity streaking observed in different zone-axis patterns recorded from this metastable [theta] ' precipitate makes it difficult to obtain the two-dimensional and specially the three-dimensional information required to determine its crystal structure by convergent-beam electron diffraction (Buxton et al., 1976) or microdiffraction (Morniroli & Steeds, 1992; Redjaïmia & Morniroli, 1994). Nevertheless, we were able to overcome these difficulties. Most information was obtained by recording suitable diffraction along two common parallel axes of the [theta] ' phase and the ferritic matrix. The determination of the metastable [theta] '-phase structure was then carried out by a thorough analysis of these patterns in conjunction with crystallographic symmetry (Portier & Gratias, 1982; Cahn & Kalonji, 1982).

3.3. Symmetry analysis

It is well established that the morphology of transformation products forming in a solid matrix is of great importance in material science. The equilibrium shape and the habit plane, respectively adopted and developed between the transformation products and the parent phase, have been extensively investigated. Both the morphology and the variant number of precipitates, which adopt orientation relationships with the matrix, can be understood in terms of group theory as developed by different authors (Portier & Gratias, 1982; Cahn & Kalonji, 1982). This approach, based on the shared symmetry elements of the two point groups, could be applied to explain the metastable [theta] '-phase features.

First of all, let us consider $[G^{\alpha}]$ and $[G^{\theta '}]$ to be the point symmetry groups of the ferritic matrix and of the nitride precipitates, respectively. The intersection point group of $[G^{\alpha}]$ and $[G^{\theta '}]$ is represented by their common symmetry elements when the precipitate adopts an orientation relationship with the matrix. This intersection point group, labelled H (H = $[G^{\alpha} \cap G^{\theta '}]$ ), is one of the 32 crystallographic point groups and a subgroup of the matrix and the precipitate point groups. One can assert with Cahn & Kalonji (1982) that it is an error to relate the morphology of the precipitate to the symmetry of the matrix or of the precipitate. It is suggested that the precipitate crystal adopts a form consistent with the symmetry of the subgroup H. Another important concept of the crystallographic symmetry is the index of the subgrouping which provides the number of precipitate variants of a given orientation relationship (Portier & Gratias, 1982; Cahn & Kalonji, 1982). This number, n, for a precipitate developed in a matrix is defined as the index of H in $[G^{\alpha}]$ , and is determined as the ratio n = m/h, where m is the order of the matrix and h is the order of the intersection point group, H. In fact, the order of each group represents its symmetry-element number (Buerger, 1978). This symmetry analysis concept has been successfully applied to determine the number of variants and to characterize the morphology of precipitates of different systems (Dahmen & Westmacott, 1986; Muddle & Polemar, 1989; Redjaïmia et al., 1993; Donnadieu et al., 1998; Redjaïmia & Metauer, 2001).

3.3.1. Determination of H = G G^'

The point group $[G^{\alpha}]$ of the ferritic matrix is m $[\bar{3}]$ m (4/m $[\bar{3}]$ 2/m) and its order is m = 48. The fact that the precipitation of the [theta] ' plates leads to three variants (n = 3) with equivalent symmetries in each grain of the matrix is indicative that the order, h, of the intersection point group H is equal to 16 (h = m/n = 48/3 = 16). This order of sixteen suggests that the only possibility for the intersection point group is H (16) = 4/m m m (4/m 2/m 2/m), which belongs to the tetragonal system (Hahn, 1988). This point group dictates a plate morphology (Phillips, 1971), which is consistent with that experimentally observed (Figs. 4, 5a and 6). It is clear that the resulting shared symmetry elements fulfil the following relation:

$[H = G ^{\alpha}\,\, \cap \,\,G ^{\theta'} = {{4}\over{m}}\,{\bar{3}}\,{{2}\over{m}} \,\,\cap\,\, G ^{\alpha '} = {{4}\over{m}} \,{{2}\over{m}} \,{{2}\over{m}}.]$

3.3.2. Determination of G^' point group

The point group which fulfils the previous requirement has to contain three mutually orthogonal axes: one fourfold axis and two twofold axes. Each of which is orthogonal to a mirror. Of the 32 crystallographic point groups, only the following two can be selected for $[G^{\theta '}]$ , namely:

$[ m\, {\bar{3}}\, m \,\left({{4}\over{m}}\, {\bar{3}}\, {{2}\over{m}} \right)\quad \rightarrow\quad {\rm Cubic\,\, system}]$

and

$[{{4}\over{m}}\, m\, m \,\left({{4}\over{m}}\, {{2}\over{m}} \,{{2}\over{m}} \right) \quad\rightarrow\quad {\rm Tetragonal\,\, system}.]$

To discriminate between the two possible crystal systems, cubic or tetragonal, in which the [theta] ' phase crystallizes, experimental electron diffraction patterns along parallel zone axes common to the ferritic matrix and the [theta] ' phase have been recorded. Close examination of the composite electron diffraction patterns recorded along $[[001 ]_{\alpha}]$ and $[[{\bar{1}}10 ]_{\alpha '}]$ , reported in Figs 5(b) and 8, respectively, leads to two possible orientation relationships. These correspond to the cube-on-cube and the Bain orientation relationships. The former consists of making parallelism between planes and directions with the same crystallographic indexes and the latter is derived from the cube-on-cube one by a relative rotational angle of 45° along $[[001]_{\alpha '}]$ , i.e. the $[[100]_{\alpha}]$ direction is set parallel to $[[110]_{\theta '}]$ .

Figure 8
Composite electron diffraction pattern recorded along the [ $[\bar{1}]$ 10] zone axis of the ferritic matrix. The reflections with low intensity belong to the [theta]

' phase.

Cubic system assumption. In the case in which the [theta] ' phase belongs to the cubic system, it is clear that neither the cube-on-cube nor the Bain orientation relationships lead to the intersection point group, $[H^{\theta '}_{\alpha}]$ = 4/m 2/m 2/m. These orientations have to be discarded. The first one is compatible with H = m $[\bar{3}]$ m (4/m $[\bar{3}]$ 2/m) and the second one is consistent with the tetragonal system. These two situations are not in agreement with the suggested assumption. Therefore, we are only left with the tetragonal system.

Tetragonal system assumption. In the case of the tetragonal system, both orientation relationships evoked previously are possible because, on the one hand it leads to the intersection point group H = 4/m 2/m 2/m and, on the other hand, the analysis of the possible lattice parameters is consistent with the tetragonal system. Therefore, the metastable [theta] ' phase crystallizes in the tetragonal system and belongs to the 4/m 2/m 2/m point group. The crystallographic characterization of the [theta] ' phase has to be completed by the determination of the space group and the lattice parameters.

3.3.3. Space-group determination and lattice parameters

The diffraction patterns along the main zone axes of the ferritic matrix were found to exhibit zone axes patterns from the [theta] ' phase. In order to determine the [theta] '-phase space group, it was necessary to build its reciprocal space. We therefore tilted the sample along two orthogonal $[ \langle {110} \rangle _{\alpha}]$ directions and recorded the superimposed diffraction patterns of the [theta] ' phase and the ferritic matrix. In this way, incident beams in the $[\langle {001} \rangle _{\alpha}]$ , $[\langle {011} \rangle _{\alpha}]$ and $[\langle {111} \rangle _{\alpha}]$ directions were sufficient to cover the two-phase reciprocal lattices, as illustrated in Fig. 9. The projections of first layer on the zeroth layer along $[ [{001}]_{\alpha}]$ , $[\langle {100} \rangle _{\alpha}]$ { $[[{100} ]_{\alpha}]$ and $[[{010} ]_{\alpha}]$ } and $[\langle {110}\rangle _{\alpha}]$ { $[ [{110} ]_{\alpha}]$ and $[[{1\bar{1}0} ]_\alpha]$ } are analysed. Their comparisons with the simulated microdiffraction patterns of Fig. 9c of Morniroli & Steeds (1992) give the Bravais lattice (P) in connection with the following individual extinction symbols:

Figure 9
Schematic representation of [theta]

' and

superimposed reciprocal lattices. The underlined hkl values are for the [theta]

' phase and those not underlined are for the ferritic matrix.

$[\eqalign{&P\,\, - \,\, \bullet\,\, \bullet \qquad{\rm for}\quad [ 001 ]_{\theta ^{\,\prime}},\cr & P\,\, \bullet \,\, - \,\,\bullet \qquad{\rm for}\quad [100 ]_{\theta ^{\prime}} \cr &P \,\,\bullet \,\,\, \bullet\,\,\,\,\, c \qquad{\rm for}\quad [ 1\bar{1} 0 ]_{\theta ^{\,\prime}}.}]$

The determination of the individual extinction symbols is based on the nature of the shift and of the periodicity difference between the first and zeroth layers.

The addition of the three individual extinction symbols gives the partial extinction symbol P - - c. This extinction symbol reveals the presence of an axial c-glide plane, which is in our case parallel to $[ ({110} )_{\theta '}]$ . The c-glide plane is responsible for the $[(0 \,0\, 2n + 1 )_{\theta '}]$ reflection extinctions, i.e. the reflections along $[ [{001} ]_{\theta '}]$ . According to Table 3.2 of the International Tables for Crystallography (Hahn, 1988), this extinction symbol P - - c is in agreement with the space group P 4₂/m m c, or P 4₂/m 2/m 2/c in its full notation.

3.3.4. Orientation relationships and lattice parameters

The orientation relationship between the [theta] ' phase and the ferritic matrix is deduced from diffraction patterns recorded along common and parallel zone axes. The derived orientation relationship expressed by exact parallelism between the corresponding planes can be written as

$[\eqalign{& (100)_{\theta^{\prime}} || (200)_\alpha \, ,\cr & ( 010 )_{\theta^{\prime}}|| ( 020 )_\alpha \, , \cr & (002 )_{\theta^{\prime}} || ( 002 )_\alpha \, . }]$

The ferritic matrix parameter, $[a_{\alpha}]$ = 2.876 Å, characterized by X-ray diffraction, is used to calibrate the electron diffraction patterns and to deduce the lattice parameters of the tetragonal [theta] ' phase, namely,

$[a_{\theta '} = a_{\alpha} = 2.876\,\hbox{\AA}]$

and

$[c_{\theta '} = 2 \times a_{\alpha} = 5.752\,\hbox{\AA}.]$

4. Conclusion

Nitriding treatment of the ferritic matrix of Fe-1.62 wt% Mn alloy at 843 K during 8 h produces a hardness profile with a sharp peak. As far as we are aware, this peak phenomenon is the first of its kind to be observed during the nitriding treatment of an Fe-Mn ferritic matrix. During the microstructural analyses, it is shown that, in addition to the [theta] -Mn₆N₅ nitride, the hardening peak is induced by an unexpected phase, which is formed throughout the ferritic matrix. This new phase, labelled [theta] ' phase in this investigation, is a metastable Mn-N nitride. It is detected only by transmission electron microscopy (TEM) and is not observed in an optical microscope or by X-ray diffraction because of the small size and amount of the precipitate. This nitride crystallizes in the tetragonal system and belongs to the space group P 4₂/m 2/m 2/c with the following lattice parameters: $[a_{\theta'}]$ = 2.876 Å and $[c_{\theta'}]$ = 5.752 Å. The [theta] '-phase crystal lattice is oriented with respect to the surrounding ferritic matrix according to the cube-on-cube orientation relationship, namely: $[(100)_{\theta'}\, || \,(200)_\alpha]$ , $[(010)_{\theta'} \,|| \,(020)_\alpha]$ and $[(002)_{\theta'}\, ||\, (002)_\alpha]$ .

The volume of the tetragonal [theta] ' phase is given by $[V_{\theta'}]$ = $[2a_{\alpha}^{3}]$ = 47.577 Å³. In comparison with the volume per atom for Fe (11.777 Å³) and for Mn (12.245 Å³), it is clear that for this [theta] ' phase there is room for no more than four metallic atoms in the cell. The remaining space will be occupied by nitrogen atoms. The number of metallic atoms has to be connected with the four metallic atoms in the austenite [gamma] phase, the cell volume of which is $[V_{\gamma}]$ = 44.362 Å³.

Previous works concerned with diffusion of nitrogen in ferrite have not provided any information about this unstable [theta] ' phase. As far as we are aware, the crystallographic data are the first data published on this topic. This lack of information is probably due to the fact that [theta] '-phase precipitation occurs in a very thin diffusion domain.

A continuing investigation is to be started to characterize the chemical composition of this [theta] ' phase by electron energy loss spectroscopy (PEELS), to explain its formation mechanism and optimize this source of alloy strengthening.

References

Baird, J. D. (1966). J. Iron Steel Inst. 204, 1122-1130.
Buerger, M. J. (1978). Elementary Crystallography. Cambridge, Massachussetts: MIT Press.
Buxton, B. F., Eades, J. A., Steeds, J. W. & Rackham, G. M. (1976). Philos. Trans. R. Soc. 281, 171-194.
Cahn, J. W. & Kalonji, G. (1982). International Conference on Solid-Solid Phase Transitions, Pittsburg, Vol. 1, pp. 3-14. (Trans. Metall. Soc. AIME.)
Dahmen, U. & Westmacott, K. H. (1986). Mater. Res. Soc. Symp. Proc. 62, 217-228.
Donnadieu, P., Carshughi, F., Redjaïmia, A., Diot, C. & Lapasset, G. (1998). J. Appl. Cryst. 31, 212-222.
Edington, J. W. (1974). Practical Electron Microscopy in Materials Science, Vol. 3, p. 67. Eindhoven: Philips.
Enrietto, J. F. (1962). Trans. Metall. Soc. AIME, 224, 43-48.
Gouné, M., Belmonte, T., Redjaïmia, A., Weisbecker, P., Fiorani, J. M. & Michel, H. (2002). Mater. Sci. Eng. A. In the press.
Hahn, T. (1988). Editor. International Table for Crystallography. Dordrecht: Reidel.
Huang, W. (1989). Calphad, 13, 243-252.
Jack, K. H. (1975). Proceedings of Heat Treatment, Vol. 73, pp. 39-50. London: The Metals Society.
Jonsson-Holmqvist, B., Grieveson, P. & Jack, K. H. (1973). Scand. J. Metall. 2, 35-38.
Kreiner, G. & Jacobs, H. (1992). J. Alloys Compd. pp. 345-362.
Lightfoot, B. J. & Jack, D. H. (1975). Proceedings of Heat Treatment Vol. 73, pp. 59-65. London: The Metals Society.
Mittemeijer, E. J. & Jack, D. H. (1985). Nitrogen Met. Alloys, A, 16-20.
Morniroli, J. P. & Steeds, J. W. (1992). Ultramicroscopy, 45, 219-239.
Muddle, B. C. & Polemar, I. J. (1989). Acta Metall. 37, 777-789.
Otsuka, N., Hanawa, Y. & Nagakura, S. (1977). Phys. Status Solidi A, 43(Fasc. 2 pK), 127-129.
Phillips, F. C. (1971). An Introduction to Crystallography, 4th ed. Singapore: Longman.
Podgurski, H. H. & Davis, F. N. (1981). Acta Metall. 29, 1-9.
Portier, R. & Gratias, D. (1982). J. Phys. Suppl. 43(C4), 17-34.
Redjaïmia, A. & Metauer, G. (2001). J. Mater. Sci. 36, 1717-1725.
Redjaïmia, A. & Morniroli, J. P. (1994). Ultramicroscopy, 53, 305-317.
Redjaïmia, A., Ruterana, P., Metauer, G. & Gantois, M. (1993). Philos. Mag. A, 67, 1277-1286.
Sato, K., Takahashi, K., Inoue, Y., Yamazaki, T. & Nitoh, Y. (1988). J. Mater. Sci. 23, 4445-4449.

research papers