Innovative Method for the Mass Preparation of α″-Fe₁₆N₂ Powders via Gas Atomization

Abstract

The iron nitride materials, especially α″-Fe₁₆N₂, are considered one of the most promising candidates for future rare-earth-free magnets. However, the mass production of α″-Fe₁₆N₂ powders as a raw material for permanent magnets is still challenging. In this work, starting from iron lumps as a raw material, we have managed to prepare the α″-Fe₁₆N₂ powders via the gas atomization method, followed by subsequent nitriding in an ammonia–hydrogen gas mixture stream. The particle size was controlled by changing the gas atomization preparation conditions. X-ray diffractograms (XRD) analyses show that the prepared powders are composed of α″-Fe₁₆N₂ and α-Fe phases. The α″-Fe₁₆N₂ volume ratio increases with decreasing powder size and increasing nitriding time, reaching a maximum of 57% α″-Fe₁₆N₂ phase in powders with size below 32 ± 3 μm after 96 h nitridation. The saturation magnetization reaches the value of 237 emu/g and a reasonable coercivity value of 884 Oe. Compared to the saturation magnetization values of α-Fe powders, the α″-Fe₁₆N₂ powders prepared through our proposed approach show an increase of up to 10% in saturation and demonstrate the possibility of mass production of α″-Fe₁₆N₂ powders as precursors of permanent magnets without rare earths.

1. Introduction

Permanent magnets are indispensable components of modern life and technology, and their influence is constantly growing [1,2]. Today, Nd-Fe-B-based magnets are the most widely used type of permanent magnets due to their outstanding magnetic performance. However, the worries about environmental degradation from rare earth mining, cost, and availability issues have led to a global concern to search for alternatives for rare-earth-free permanent magnets. Thus, it is expected that the next generation of permanent magnets will not contain rare earths, will come from abundant resources, and will have low costs. There are several types of alternative material systems to rare-earth magnets, such as L1₀-FeNi [3,4], MnBi [5,6,7], and α″-Fe₁₆N₂. Among them, the α″-Fe₁₆N₂ compound with giant saturation magnetization contains widely available, inexpensive, and completely non-polluting elements, making it a promising candidate for rare-earth-free magnets. After the first report by Kim and Takahashi on the giant saturation magnetization of about 290 emu/g of the α″-Fe₁₆N₂ phase in the evaporated Fe-N thin film, many researchers have made sustained efforts on both the fundamental [8,9] and practical aspects of preparing of the α″-Fe₁₆N₂ material. Thus, several attempts have been successfully made to develop magnetic materials containing α″-Fe₁₆N₂ phase in the form of thin films [10,11,12,13], foils [14,15], rods [16], ribbons [17], nanocones [18], and nanoparticles and powders [19,20], by using different preparation methods. The reported values of saturation magnetizations were widely scattered from 230 to 315 emu/g and appeared to be inconsistent and sometimes contradictory. However, the value of saturation magnetizations is mainly determined by the amount of α″-Fe₁₆N₂ phase relative to the impurity phases such as α-Fe, γ-Fe-N (austenite), or γ-Fe₄N, in the magnetic material [21]. Therefore, obtaining the magnetic material with pure α″-Fe₁₆N₂ phase in sufficiently large quantities is a challenge. On the other hand, the preparation of permanent magnets is naturally related to the precursor material, namely powders. Therefore, various attempts have been made to prepare the α″-Fe₁₆N₂ powders by mechanical ball milling of pure iron with a solid nitrogen source, ammonium nitrate [20,22,23], and nanopowders via chemical methods [19,24,25], using a two-step route where several kinds of Fe-oxides such as α-Fe₂O₃, γ-Fe₂O₃, and Fe₃O₄ are reduced to α-Fe in hydrogen atmosphere followed by a low-temperature nitriding. However, the mass production of iron nitrides with nanometer-sized Fe sources has critical drawbacks in manufacturing processes as well as economics. For example, iron oxide particles undergo significant coarsening during the reduction process, resulting in a detrimental effect on the final magnetic properties [19], in addition, due to their small size, the α″-Fe₁₆N₂ nanoparticles are highly susceptible to oxidation. Oxidation resistance is slightly modified by coating the α″-Fe₁₆N₂ nanoparticles with either Al₂O₃ or SiO₂ in a core–shell arrangement [26,27], but the coating can also contribute to the degradation of the magnetic properties of the resulting particles. A step towards mass production of the α″-Fe₁₆N₂ powders was made by introducing the mechanical ball milling of Fe powders together with solid NH₄NO₃ powders for time periods of up to 60 h [22]. The value of the saturation magnetization was about 210 emu/g, the coercivity of 854 Oe, and the volume fraction was about 70%. In a most recent work [23] amorphous powders of Fe₇₈Si₉B₁₃ and solid powders of NH₄NO₃ were milled for extended periods of up to 80 h. Although the formation of the α″-Fe₁₆N₂ phase was observed, the volume fraction was low, resulting in magnetization of very low saturation at 72.5 emu/g and a moderate coercivity of 541 Oe. A typical and affordable method for the mass production of micrometric metal and alloy powders is the gas atomization technique [28,29,30,31]. The idea is to transfer the kinetic energy from a high-velocity gas jet expanded through a nozzle, to a jet of liquid metal, resulting in fragmentation and breaking down into metal droplets. The liquid metal droplets subsequently cool and solidify into metal powder. To date, there is still no report on an approach to preparing α″-Fe₁₆N₂ powders via the gas atomization method.

In this paper, we report the successful preparation of α″-Fe₁₆N₂ powders using the gas atomization method, followed by subsequent nitriding in the ammonia–hydrogen gas mixture stream.

2. Materials and Methods

The preparation of α″-Fe₁₆N₂ powders was carried out in three steps. In the first step, Fe ingots of about 20–25 g were prepared via the arc melting technique from commercial iron lumps (Alfa Aesar) with a purity higher than 99.99%. In the second step, the resulting Fe ingots were used to prepare Fe-N powders via the gas atomization method. This step involves the transformation of a jet of liquid Fe into fine droplets using a high-velocity nitrogen gas. In the last step, the Fe-N powders were nitrided in an NH₃ (92 vol%) + H₂ (8 vol%) gas mixture stream at a flow rate of 100 mL/min, at 150 °C for periods between 12 and 96 h. After cooling to room temperature under the same gas mixture stream, the nitrided powders were immediately transferred to a nitrogen-filled glove box without exposure to air to prepare samples for structural, morphological, and magnetic measurements. The structural analysis was carried out via X-ray diffraction (XRD) in a Bruker AXS D8-Advance diffractometer (Mannheim, Germany) over the angular range 20–90°, with a step of 0.02° and a counting time of 2 s per step, using conventional Cu-Kα incident radiation. The Rietveld refinements of the XRD patterns were performed using the MAUD software package. The microstructure of the sample was investigated using a Scanning Electron Microscope, FIB/FE-SEM Cross-Beam Carl Zeiss NEON 40 EsB (Oberkochen, Germany), equipped with an energy dispersive X-ray spectroscopy (EDS) module. Magnetic measurements were performed using a vibrating sample magnetometer (VSM) (Lake Shore VSM 7410, Westerville, OH, USA) in a maximum applied field of 20 kOe, at room temperature. The design and operation of the gas atomization facility is described in detail in our previous work [32]. However, for an easy understanding of the atomization process with the homemade equipment used for the preparation of powders, we will reproduce here the gas atomization sketch (Figure 1) as well as some particular characteristics of the configuration we used in the present work. Thus, we mention that the iron ingot was heated and melted via induction in a quartz crucible, and the temperature of the melt before ejection was 1588 °C; both for melt ejection and atomization, respectively, nitrogen gas was used at the ejection pressure of 1 bar and the atomization pressure of 20 bar; the width of the atomizing gas outlet nozzle, which is in the form of an annular slot concentric with the melt jet, was 300 µm; the angle formed between the direction of the melt jet and that of the atomization gas jet was 45°; and the distance between the melt ejection nozzle and the particle collector was 120 cm. In order to prepare powders with different mean particle diameters, the ejection nozzle diameter varied between 100 and 200 µm. Before the melt ejection, the atomization and melting chambers were pressurized for 5 min with nitrogen gas at exposure pressures between 0.5 and 3 bar, to expose the liquid iron to the nitrogen gas.

3. Results and Discussions

The scanning electron microscope (SEM) images shown in Figure 2 give a short qualitative overview of the powders obtained for different diameters of the ejection nozzles and an exposure pressure of 0.5 bar.

For all the powders, the general particle shape is spherical, and their median diameter increases with the diameter of the ejection nozzle. Thus, the median diameter of the obtained powders is 32 ± 3 µm when an ejection nozzle with a diameter of 100 µm was used, 58 ± 2 µm for 150 µm, and 96 ± 2 µm for 200 µm. It can be observed that all the powders show a certain amount of satellites, and the finest ones show even particles that have not finished formation. This might be due to the higher cooling rate of the fine particles, and therefore, some of them fail to take a spherical shape before solidification. It should be noted, however, that the use of ejection nozzles with diameters of the order of hundreds of µm has allowed the preparation of powder with diameters of the order of tens of µm at a low atomization pressure of only 20 bar. Additionally, in the experiments performed in this work, no measures were taken to reduce the appearance of satellites or to avoid particle deformation.

Figure 3 shows the Rietveld refined XRD diffractograms of Fe-N powders with different median diameters, obtained by atomization using the melting nozzle with diameters of 100 µm, 150 µm, 200 µm, and an exposure pressure of 0.5 bar.

The values of goodness factor, ${\chi }^2={\left(\raisebox{1ex}{$R_{wp}$}\!\left/ \!\raisebox{-1ex}{$R_{exp}$}\right.\right)}^2$, obtained from the Rietveld refinement for all samples are between 1 and 2, which indicates the quality of the fit. Indeed, in Figure 3, a good agreement can be observed between the experimentally observed and the calculated patterns. The results of XRD patterns analysis indicate that the structure of the powders with diameters of 32 ± 3 µm and 58 ± 2 µm consists of a mixture of α-Fe and α′-Fe₈N crystalline phases, while for the larger powders with a diameter of 96 ± 2 µm, the γ-F₄N phase also appears. For the phase analyses, the cards of indices PDF 98-065-4562 (α′-Fe₈N), PDF 98-015-9354 (iron), and PDF 98-007-9980 (γ-F₄N) were used. The α′-Fe₈N phase percentage decreases from 31.74% corresponding to the powder with diameters of 32 ± 3 µm to 27.29% for those with diameters of 58 ± 2 µm, and 18.55% for 96 ± 2 µm, respectively, suggesting that the percentage of phases depends on the size of the powders. In fact, it most likely depends on the different speeds of cooling of the powders, an argument in this sense being the appearance of the γ-F₄N phase in large-sized powders due to insufficiently rapid quenching.

The formation mechanism of the α″-Fe₁₆N₂ phase consists of the transformation of γ-FeN (austenite) by its quenching to α′-Fe₈N (martensite) and subsequent tempering to α″-Fe₁₆N₂ [33,34]. Since the α′-Fe₈N phase is the precursor of the desired α″-Fe₁₆N₂ phase, it is therefore desirable that the powders contain a large amount of the α′-Fe₈N phase. But the percentage of the α′-Fe₈N phase is directly proportional to the percentage of the γ-FeN phase, which in turn is dependent on the amount of nitrogen dissolved in the liquid iron. The solubility of nitrogen in iron has been successfully described by Sievert’s law, according to which the amount of nitrogen dissolved in the melt is directly proportional to the square root of the gas pressure above the melt, in fact, the exposure pressure [35]. Therefore, to increase the content of the α′-Fe₈N phase in the powders obtained via atomization, we either increase the nitrogen exposure pressure during the induction melting process or decrease the ejection nozzle diameter to prepare powders with even smaller sizes. Due to technological limitations related to the fact that the melting temperature of iron is close to the softening temperature of the quartz tube in which it is heated, reducing the ejection nozzle diameter further was not possible because the nozzle would clog and the quartz tube would break under pressure; therefore, we chose to increase the exposure pressure from 0.5 to 3 bar, keeping the ejection nozzle diameter at 100 µm.

The results from the analysis of X-ray diffraction patterns (not shown here) corresponding to Fe-N powders atomized at nitrogen pressure of 3 bar during the induction melting process showed that the α′-Fe₈N phase fraction increased to 43.32% from 31.74%, which was for the powders prepared at a pressure of 0.5 bar. Therefore, an additional increase in nitrogen pressure during induction melting could lead to an additional increase in the α′-Fe₈N phase content in the atomized powders. However, the atomization plant used by us is a homemade one and does not allow an increase higher than 3 bar, but the challenge remains open for other researchers. Next, only the powders with median diameters of 32 ± 3 µm obtained at 3 bar exposure pressure were nitrided, for which the α′-Fe₈N phase percentage is 43.32%.

Figure 4 shows the phase evolution by XRD patterns of the nitrided powders for 24, 48, and 96 h at a temperature of 150 °C.

The α″-Fe₁₆N₂ has a structure very similar to α´-Fe₈N except for the arrangement of nitrogen atoms. In the α′-Fe₈N phase, the N atoms are disordered, while in the α″-Fe₁₆N₂ phase, they are ordered. For this reason, the most main reflections from the α″-Fe₁₆N₂ phase overlap with those from the α′-Fe₈N phase. As can be seen in Figure 4, the nitrided powders showed a diffraction peak at θ = 46.57° corresponding to the (103) planes, which comes from the ordering of nitrogen atoms and represents the so-called “fingerprint” of the α″ phase [36]. Therefore, this is direct evidence of the occurrence of the α″-Fe₁₆N₂ tetragonal phase (space group: I4/mmm) (PDF card no. 01-075-2143) in the powders prepared via the gas atomization method followed by their subsequent nitriding in gaseous ammonia. The calculated lattice parameters of the α″-Fe₁₆N₂ phase were a = 5.73Å and c = 6.29Å, which are in good agreement with the reported data provided by Kim and Takahashi [34]. Also, it should be mentioned that the (220), (400), and (413) peaks of α″-Fe₁₆N₂ shared the intensity with α-Fe(110), α-Fe(200), and α-Fe(112), respectively. It can be observed that the intensity of the (103) peak increased with nitriding time, together with a reduction in α-Fe peaks. In addition, for samples nitrided for 48 or 96 h, the appearance of other new diffraction peaks associated with the α″-Fe₁₆N₂ phase is observed. Thus, the volume ratio of the α″-Fe₁₆N₂ phase increases with the nitriding time from 38% for powders nitrided for 24 h, reaching 57% after 96 h.

Figure 5a–c show the hysteresis loops of the powders after 24, 48, and 96 h of nitridation time at a temperature of 150 °C. To compare the magnetic properties of nitrided powders with those of powders similar in size from pure Fe, Fe powders atomized in argon gas were prepared, and the corresponding hysteresis loop is shown in Figure 5d.

The Fe powders atomized in argon showed a saturation magnetization of 215 emu/g and a coercivity about 28 Oe. The saturation magnetization obtained in these powders is lower than the theoretical value of 218 emu/g, most likely due to surface defects or possibly superficial oxidation. For nitrided powders, both the coercivity and the saturation magnetization of the powders increase once the nitriding time increases. Thus, the coercivity increases from approximately 358 Oe for the 24 h nitrided powders to 884 Oe for the 96 h nitrided powders, and the saturation magnetization increases from 224 emu/g to 237 emu/g. Compared to the saturation magnetization of Fe powders atomized in argon, the 96 h nitrided powders show an increase in the saturation magnetization by approximately 10%. The higher values of the coercivity and the saturation magnetization for the sample nitride 96 h are caused by the increase in the percentage of α″-Fe₁₆N₂ phase, according to XRD results. However, under the preparation conditions applied in this study, the volume fraction of α″-Fe₁₆N₂ is still low. Therefore, although the method described here opens a new way of preparation of α″-Fe₁₆N₂ powders, further future research is needed to optimize the process parameters and increase the volume fraction of α″-Fe₁₆N₂.

4. Conclusions

The α″-Fe₁₆N₂ powders were successfully obtained via the gas atomization method of pure iron in nitrogen gas, followed by subsequent nitriding in an ammonia–hydrogen gas mixture stream. The atomized powders presented a structure consisting of a mixture of α′-Fe₈N, γ-F₄N, and α-Fe phases. By tuning the parameters of the atomization process, the volume fraction of the α′-Fe₈N phase, which represents the precursor of the desired phase α″-Fe₁₆N₂, was increased up to 43.32%. Subsequent nitriding of the powders allowed the transformation of the α′-Fe₈N phase into the α″-Fe₁₆N₂ phase. The volume fraction of the α″-Fe₁₆N₂ phase in the nitride powders reached a maximum of 57% for which the powders showed a saturation magnetization (Ms) of 237 emu/g, remanence (Mr) of 101 emu/g, and develop reasonable magnetic hardness with coercivity (Hc) of 884 Oe. Taking into account that the ratio of the α″-Fe₁₆N₂ phase could be controlled by modifying the nitridation parameters, their further optimization could increase the volume of the α″-Fe₁₆N₂ phase and improve the magnetic properties of powders. However, the present results open a new route for the mass-producing of α″-Fe₁₆N₂ powder as precursors of permanent magnets without rare earths.