Composition and microstructural control of Sm₂Fe₁₇N₃ powders: a promising candidate for next-generation permanent magnets

Abstract

Sm₂Fe₁₇N₃ powders have drawn much attention in both fundamental research and industrial applications due to their excellent magnetic properties. The magnetic properties of Sm₂Fe₁₇N₃ powders mainly depend on the composition and microstructure. Herein, the intrinsic and extrinsic magnetic properties of Sm₂Fe₁₇N₃ powders are introduced, where the correlation between the composition, microstructure, and magnetic properties is discussed. Based on the correlation, we provide a detailed description of the synthesis methods and regulation strategies for Sm₂Fe₁₇N₃ powders, aiming to promote the design and preparation of Sm₂Fe₁₇N₃ powders with better magnetic properties.

---

1. Introduction

Sm₂Fe₁₇N₃ is expected to be a promising candidate for the next-generation rare-earth permanent magnets due to the huge anisotropy field (14 T), large saturation magnetization (1.54 T), and high Curie temperature (475 °C).^1–3 In addition to the intrinsic magnetic properties comparable to or better than Nd₂Fe₁₄B, the resistivity of Sm₂Fe₁₇N₃ is higher than that of Nd₂Fe₁₄B, which means lower current loss at high frequency. Thus, Sm₂Fe₁₇N₃ is more suitable for high-frequency applications, such as magnets in electric vehicles.^4–6

Up to now, Sm₂Fe₁₇N₃ has been widely used in the form of resin-bonded magnets because it decomposes at around 873 K, making sintering difficult.^7,8 Nevertheless, it has been demonstrated that there is a high possibility of achieving high-performance sintered Sm₂Fe₁₇N₃ magnets by a low-oxygen powder metallurgy process.^9,10 Whether in the case of the resin-bonded or sintered Sm₂Fe₁₇N₃ magnets, one of the most important issues is to prepare Sm₂Fe₁₇N₃ powders with excellent coercivity, remanence, and maximum energy product.³ Generally, Sm₂Fe₁₇N₃ powder is obtained by the nitridation of Sm₂Fe₁₇ powders. In the Sm–Fe binary system, Sm₂Fe₁₇ is formed by the peritectic reaction between solid Fe and liquid Sm–Fe, which would easily result in residual soft magnetic α-Fe, deteriorating the coercivity and remanence of Sm₂Fe₁₇N₃.^11,12 Thus, several techniques have been developed to prepare single-phase Sm₂Fe₁₇ powders, such as rapid-quenching,^11,13 reduction–diffusion,^14,15 hydrogenation–disproportionation–desorption–recombination (HDDR),^16,17 and so on.^18,19

The nitridation process directly determines the nitrogen content and distribution in Sm₂Fe₁₇N_x powders. The magnetic anisotropy would vary with the nitrogen content in Sm₂Fe₁₇N_x and it is considered that Sm₂Fe₁₇N₃ would exhibit the best comprehensive magnetic properties.²⁰ In addition to the nitrogen content, the nitrogen distribution is also important. Nonuniform nitrogen distribution in Sm₂Fe₁₇N₃ would deteriorate the magnetic properties because the region with weaker magnetic anisotropy would act as nucleation centers of reverse domains. Therefore, Sm₂Fe₁₇N₃ powder with homogenous nitrogen content has been pursued to achieve excellent magnetic properties.

The coercivity of Sm₂Fe₁₇N₃ powders is mainly dominated by the nucleation of reverse domains. Thus, the coercivity of Sm₂Fe₁₇N₃ powders is sensitive to the microstructure and the key to enhancing coercivity is to prohibit nucleation of reverse domains.^21–23 In theory, Sm₂Fe₁₇N₃ powders with a critical single-domain size and a smooth as well as undamaged surface would exhibit the largest coercivity.^24–26 It should be noted that the size of Sm₂Fe₁₇N₃ powders for the highest coercivity would deviate from the critical single-domain size due to surface oxidation and defects. Thus, many efforts have been made to regulate the size of Sm₂Fe₁₇N₃ powders and inhibit surface oxidation for coercivity optimization.^14,27

Herein, we will briefly introduce bonded Sm₂Fe₁₇N₃ magnets as well as the intrinsic and extrinsic magnetic properties of Sm₂Fe₁₇N₃ powders and then the efforts in the design and preparation of high-performance Sm₂Fe₁₇N₃ powders will be discussed in detail. Last but not least, the prospects for Sm₂Fe₁₇N₃ will be addressed.

2. Sm₂Fe₁₇N₃ magnets

2.1. Bonded Sm₂Fe₁₇N₃ magnets

MagValley, Sumitomo Metal Mining Co. Ltd, Nichia Corporation, and Daido Steel Co. Ltd have commercialized bonded Sm₂Fe₁₇N₃ magnets, which are cheaper than bonded Nd–Fe–B magnets due to the lower cost of Sm. Specifically, bonded Sm₂Fe₁₇N₃ magnets exhibit higher resistance to corrosion and heat than bonded Nd–Fe–B magnets. As a result, they have found widespread applications in products such as disk suction devices, automation robots, magnetic sensors, and more.

Bonded Sm₂Fe₁₇N₃ magnets are produced using compression and injection molding techniques, which are commonly employed for the production of other bonded rare-earth permanent magnets. Isotropically bonded Sm₂Fe₁₇N₃ magnets have a maximum energy product ranging from 9 to 14 MGOe, which exceeds that of bonded Nd–Fe–B magnets (5–11 MGOe). Additionally, the irreversible flux loss for isotropically bonded Sm₂Fe₁₇N₃ magnets without surface coating, when exposed to air at 120 °C and hot water at 70 °C for 1000 hours, is both less than 10%. In contrast, the irreversible flux loss for bonded Nd–Fe–B magnets under the corresponding environment is approximately 25% and 40%, respectively. Consequently, isotropically bonded Sm₂Fe₁₇N₃ magnets are considered to have the most superior magnetic properties among the bonded permanent magnets. Despite the significant advancements in isotropically bonded Sm₂Fe₁₇N₃ magnets, anisotropically bonded Sm₂Fe₁₇N₃ magnets exhibit lower coercivity (15–17 kOe) and maximum energy product (12–16 MGOe) compared to anisotropically bonded Nd–Fe–B magnets (15–18 kOe, 21–23 MGOe).²⁸ Therefore, it is essential to design and prepare high-performance Sm₂Fe₁₇N₃ powders because the magnetic properties of these powders limit the performance of the resulting bonded magnets.

2.2. Intrinsic magnetic properties

It is well-known that the intrinsic magnetic properties rely on the chemical composition and crystal structure.^29–31 Both Sm₂Fe₁₇ (Fig. 1a) and Sm₂Fe₁₇N₃ (Fig. 1b) crystallize in rhombohedral structure with the space group of R3̄m.^28,32 The Fe sublattice in Sm₂Fe₁₇ exhibits basal-plane anisotropy, resulting in the basal-plane anisotropy of Sm₂Fe₁₇. The small distance of dumbbell Fe–Fe sites leads to negative exchange interactions, responsible for the low Curie temperature of Sm₂Fe₁₇ (135 °C). Therefore, Sm₂Fe₁₇ could not be applied as permanent magnets.

Fig. 1 Side and top views of crystal structures of (a) Sm₂Fe₁₇, (b) Sm₂Fe₁₇N₃, and (c) Sm₂Fe₁₇N_x (x > 3).

The introduction of nitrogen atoms into the Sm₂Fe₁₇ unit cell would result in Sm₂Fe₁₇N_x with an unchanged unit cell structure and profoundly changed intrinsic magnetic properties. Generally, the nitrogen atom could only occupy the 9(e) octahedral interstices in Sm₂Fe₁₇ considering the radius of the nitrogen atom. Thus, the concentration of interstitial nitrogen in Sm₂Fe₁₇N_x would range from 0 to 3. With the increase of x, the magnetic properties become better and Sm₂Fe₁₇N₃ would exhibit the best comprehensive magnetic properties. It should be noted that Sm₂Fe₁₇N_x with x > 3 has been obtained and the extra nitrogen atoms would occupy the 3b or 18 g interstitial sites along the c-axis (Fig. 1c). However, nitrogen atoms at 3b or 18 g interstitial sites would reduce the magnetic moment, Curie temperature, and easy-axis character.³³

As shown in Fig. 1b, the Sm atoms (6c) are coordinated by the triangular interstitial nitrogen atoms (9e) in Sm₂Fe₁₇N₃. A strong negative electric field gradient is created on Sm atoms, which changes the second-order crystal-field coefficient A₂₀.³⁴ The Sm–N bonding contributes to the axial anisotropy of the Sm sublattice and the room-temperature first-order anisotropy constant significantly increases from −0.8 MJ m⁻³ for Sm₂Fe₁₇ to 8.6 MJ m⁻³ for Sm₂Fe₁₇N₃.³⁵

In addition to the change of the anisotropy field, the introduction of nitrogen atoms to Sm₂Fe₁₇ would also enhance the Curie temperature and magnetic moment. The interstitial nitrogen atoms induce an increase in the average Fe–Fe distances, which leads to a high saturation magnetization of Sm₂Fe₁₇N₃ (1.54 T). Besides, the increase of average Fe–Fe distance induces the increase of exchange interactions (from negative to positive), responsible for the high Curie temperature of Sm₂Fe₁₇N₃ (475 °C).³⁶ With a high anisotropy field, saturation magnetization, and Curie temperature, Sm₂Fe₁₇N₃ could be applied as permanent magnets.

In spite of the enhancement in magnetic properties induced by interstitial N atoms in Sm₂Fe₁₇, interstitial C or B atoms could not change the magnetic properties profoundly, which is limited by the small absorption content. For Sm₂Fe₁₇C_x, the maximum amount of C atoms is at x = 1.5, which shows a volume expansion of 2.3% compared to Sm₂Fe₁₇, although the value of x can reach 2.3 or 2.5 due to the higher reactivity of C atoms in the gas phase. The absorption limit of B atoms is much lower in Sm₂Fe₁₇, typically around x = 0.2, leading to a volume expansion of 0.3%. The change in magnetic properties largely depends on the degree of lattice expansion. However, the lattice expansion induced by N or C atoms is far lower than that of N atoms. Thus, the introduction of B atoms would almost not change the magnetic properties. The increase in saturation magnetization and Curie temperature induced by C atoms is about half that of N atoms.³⁷

2.3. Extrinsic magnetic properties

The magnetic properties of Sm₂Fe₁₇N₃ powders are critical to the performance of their resin-bonded magnets. The coercivity, remanence, and maximum energy products of Sm₂Fe₁₇N₃ powders are commonly concerned. The coercivity of anisotropic Sm₂Fe₁₇N₃ powders is mainly controlled by the nucleation of reverse domains, which could be enhanced by the optimization of composition, particle size, or surface quality.¹

Concerning the composition, it is necessary to avoid the soft magnetic phases. The soft magnetic phase does not specifically refer to the α-Fe or Sm-rich phase but also contains a nitrogen-poor region in Sm₂Fe₁₇N_x powders. It is quite easy to induce the α-Fe or Sm-rich phase impurities during the preparation of Sm₂Fe₁₇ powders. Nitrogen atoms gradually diffuse into the interior of Sm₂Fe₁₇ powders during the nitridation process, which easily results in the inner core being poor in nitrogen. The nitrogen-poor region displays lower magnetic anisotropy, acting as a nucleation center during the magnetization reversal process.

The particle size of Sm₂Fe₁₇N₃ powders depends on the magnetic domain structure and magnetic anisotropy energy. Magnetic domains arise from dipole interactions. When the particle is small enough, the energy for the domain wall is larger than the external magnetostatic energy, which means the formation of a single-domain particle. The diameter for the critical single-domain particle (D_sd) could be expressed as follows:

D_sd ≈ 18(AK₁)^1/2/μ₀M_s²

where A represents the exchange stiffness.²⁵ The D_sd value for Sm₂Fe₁₇N₃ is about 300 nm.³⁸

As the size multi-domain Sm₂Fe₁₇N₃ powders decreases, the coercivity would increase. The coercivity would reach a maximum when the size reduces to D_sd where the particles have a single-domain structure. To reduce the size of single-domain Sm₂Fe₁₇N₃ powders, the magnetic anisotropy energy (K₁V, where V represents the volume) would decrease, and the coercivity would also decrease. Hence, the particle size of anisotropic Sm₂Fe₁₇N₃ powders should be regulated to D_sd for magnetic properties optimization. It should be noted that D_sd would deviate from the theoretical value because of the surface defects.²⁴

In terms of the surface, it should be smooth and undamaged to reduce nucleation of reverse domains. The reverse domain would nucleate by a spontaneous fluctuation and the nucleus volume is on the order of δ_w³ where δ_w is the domain wall width. Once the nucleus forms, it propagates; otherwise, it is pinned. Due to the small size of δ_w for Sm₂Fe₁₇N₃ (3.7 nm), even the surface asperities would act as nucleation centers.³⁹ Thus, the surface quality for Sm₂Fe₁₇N₃ powders should be good enough to achieve higher magnetic properties.

Based on the understanding of the intrinsic and extrinsic magnetic properties of Sm₂Fe₁₇N₃ powders, controlling the composition and microstructure of Sm₂Fe₁₇N₃ powders is discussed in the subsequent section.

3. Preparation of Sm₂Fe₁₇N₃ powders

Generally, Sm₂Fe₁₇N₃ powders are prepared by nitriding Sm₂Fe₁₇ master alloy powders. Herein, the preparation and nitridation of Sm₂Fe₁₇ master alloy powders is discussed in detail.

3.1. Preparation of Sm₂Fe₁₇ master alloy powders

The common strategies for the preparation of Sm₂Fe₁₇ powders are rapid quenching, reduction–diffusion, and HDDR. As shown in Fig. 2a, Sm₂Fe₁₇ was formed by the peritectic reaction between liquid Sm–Fe (L_Sm–Fe) and solid Fe at 1010 °C, which grows along the Fe/L_Sm–Fe interface (Fig. 2b (I → II)). And the equation is L_Sm–Fe + Fe → Sm₂Fe₁₇. From II to III in Fig. 2b, the peritectic transformation occurs, and L_Sm–Fe and solid Fe transform directly into Sm₂Fe₁₇. As the Sm₂Fe₁₇ phase extends towards both L_Sm–Fe and solid Fe gradually (Fig. 2b (III → IV)), the peritectic reaction is complete. Thus, the peritectic reaction would cost a lot of time, easily resulting in an incomplete reaction. The incomplete peritectic reaction resulted in residual soft magnetic α-Fe and Sm-rich phases. The soft magnetic α-Fe acted as nucleation centers of the reverse domain, adversely affecting the magnetic properties of Sm₂Fe₁₇N₃, while the Sm-rich phase partially decomposed into SmN and α-Fe during the subsequent nitridation process.¹¹ Besides, the evaporation of Sm at high temperatures also made the preparation of single-phase Sm₂Fe₁₇ difficult.^40–42 Therefore, obtaining single-phase Sm₂Fe₁₇ master alloy powders was essential for high-performance Sm₂Fe₁₇N₃. In this section, Sm₂Fe₁₇ master alloy powders prepared by different methods will be overviewed.

Fig. 2 (a) Sm–Fe binary phase diagram, (b) schematic illustration of the peritectic reaction between L_Sm–Fe and solid Fe.

3.1.1. Rapid quenching. Generally, a rapid quenching process for Sm₂Fe₁₇ powders is as follows: preparation of Sm₂Fe₁₇ ingot by induction melting or arc melting, homogenization of Sm₂Fe₁₇ ingot, melt-spinning or strip casting, annealing, and pulverization.^43–46 The obtained Sm₂Fe₁₇ powders with the size of several tens of micrometers are nitrided and further crushed to achieve Sm₂Fe₁₇N₃ powders with a size of a few micrometers, which could display hard magnetic behaviors.

Katter and co-workers prepared Sm₂Fe₁₇ by rapid quenching and then obtained Sm₂Fe₁₇N_x by nitridation for the first time.¹³ During the melting process, Sm evaporation led to a decrease in Sm concentration, leading to the formation of the SmFe₉ phase with a TbCu₇-type structure after quenching. Therefore, a moderate excess of Sm is required to compensate for Sm loss. Quenching rates would also affect the phase composition. With an increase in the wheel velocity, the amount of α-Fe would decrease. However, when the quenching rate is above 15 m s⁻¹, the SmFe₉ phase would be observed. Since the magnetic properties of SmFe₉N_x are inferior to those of Sm₂Fe₁₇N_x, it is necessary to avoid the formation of the SmFe₉ phase. In addition to SmFe₉, α-Fe is always present in rapid-quenched ribbons and could be eliminated by annealing. However, the annealing process may result in Sm loss. The Sm₂Fe₁₇N_x powders prepared by rapid-quenched ribbons with the composition, quenching rate, annealing temperature, and time of Sm₁₂Fe₈₈, 60 m s⁻¹, 730 °C, and 15 min, respectively, showed the best magnetic properties with the coercivity, remanence, and maximum energy product of 2.10 T, 0.73 T, and 8.24 MGOe, respectively. However, such powders were magnetically isotropic and remained soft α-Fe, resulting in a kink on the demagnetization curve.

To achieve better magnetic properties, Sm₂Fe₁₇N₃ powders should be magnetically anisotropic and free of secondary phases (Sm-rich phase and α-Fe). It has been demonstrated that Sm₂Fe₁₇ ingots with fine columnar grains of the same orientation could be obtained by controlling the cooling rate and crystallization process, which lays the foundation for the preparation of anisotropic Sm₂Fe₁₇N₃ powders.⁴⁷ Concerning the secondary phase, it was expected that a strip-casting technique with an enhanced cooling rate could suppress α-Fe in Sm₂Fe₁₇ strips.⁴⁸ Thus, the annealing process would also be simplified.

Kolodkin and co-workers prepared Sm_2.4Fe₁₇ ingots by induction melting, and then the strip-casting technique was followed, resulting in Sm_2.08Fe₁₇ strips.⁴⁹ It was found that the volume of the secondary phase (Sm-rich phase and α-Fe) in Sm_2.08Fe₁₇ strips was lower than that in Sm_2.4Fe₁₇ ingots. The secondary phase in Sm_2.08Fe₁₇ strips could be eliminated by annealing at 1000 °C for 1 h, whereas α-Fe remained in the bulk Sm_2.4Fe₁₇ alloy even after annealing at 1050 °C for 20 h. In addition, the strip-casting process refined the grain size of Sm₂Fe₁₇, which could promote the nitridation process. The Sm_2.08Fe₁₇ strips were crushed and nitrided, obtaining Sm_2.08Fe₁₇N₃ powders with a coercivity of 0.93 T.

Furthermore, Yang and co-workers prepared single-phase Sm₂Fe₁₇ strips using a strip-casting process without annealing the Sm₂Fe₁₇ ingot.⁵⁰ After the nitridation process, Sm₂Fe₁₇N₃ was ball-milled into powders with different diameters. As shown in Fig. 3a, the coercivity increased with a decrease in particle size and coercivity could reach 1.75 T when the particle size decreased to 1.3 μm. It should be noted that the relationship between coercivity and size was consistent with the aforementioned single-domain theory. In contrast, the saturation magnetization decreased with a decrease in particle size, which may be ascribed to surface oxidation. In addition, the temperature-dependent magnetic properties were also studied (Fig. 3b). The coercivity could increase to 3.61 T when the temperature decreased to 10 K. Increasing the temperature to 373 K, the saturation magnetization decreased to 128 emu g⁻¹ and the remanence temperature coefficient was lower than that of Nd₂Fe₁₄B.

Fig. 3 (a) Dependencies of coercivity and saturation magnetization on particle size. (b) Hysteresis loops of the powders ground for 4 h at 10, 300, and 373 K. (a) and (b) Reproduced from ref. 50 with permission from the American Institute of Physics, copyright 2023. (c) The BSE image of the cross-section of nitrided SmFeCuN particles, (d) normalized demagnetization curves of the SmFeCuN powders. (c) and (d) Reproduced from ref. 11 with permission from Elsevier BV, copyright 2019.

In addition to the optimization of the rapid quenching process, many efforts have been made to investigate nonmagnetic phases equilibrated with Sm₂Fe₁₇. Zhu and co-workers introduced Cu into the Sm–Fe system and achieved a three-phase region of SmCu, SmCu₂, and Sm₂Fe₁₇.¹¹ Thus, an ingot free of a α-Fe and Sm-rich phase could be obtained in the presence of nonmagnetic SmCu and SmCu₂, which caused less damage to the magnetic properties of Sm₂Fe₁₇N₃. The back-scattered electron (BSE) image (Fig. 3c) of the nitrided particles indicated the presence of nitrided Sm–Cu. The anisotropic Sm₂Fe₁₇N₃ powders with Cu addition showed a coercivity and maximum energy product of 1.39 T and 31 MGOe, respectively (Fig. 3d).

Though single-phase anisotropic Sm₂Fe₁₇ powders could be achieved by a rapid quenching process, it seems that the elimination of the magnetic secondary phase still requires long-term annealing, which is not beneficial for industrial applications. Maybe, there is a compromise between the cost and magnetic performance by introducing some other elements into the Sm–Fe binary system to hinder the formation of soft magnetic secondary phases.

3.1.2. Reduction–diffusion. The reduction–diffusion process has been widely applied to prepare rare-earth permanent magnetic powders, where rare-earth oxides were reduced and then rare earth metals and transition metals are diffused into each other, resulting in the formation of an alloy powder.^51–55 The industrial production of Sm₂Fe₁₇N₃ powders prepared by the reduction–diffusion and nitridation process has been achieved by Sumitomo Co. Ltd and Nichia Corporation.¹ Economically speaking, using relatively inexpensive Sm₂O₃ instead of the expensive and hard-to-preserved Sm metal could bring about a significant reduction in cost.⁵⁶

In the early stage, commercial micro-sized Sm₂O₃ and Fe were applied as precursors to verify the feasibility of the reduction–diffusion process.⁵⁷ As shown in Fig. 4a, the precursors were mechanically milled with reduction agent (Ca or CaH₂ powders) and then heated to a temperature higher than the melting point of Ca. Molten Ca would reduce Sm₂O₃ and then inter-diffusion of Sm and Fe would be responsible for the formation of Sm₂Fe₁₇, similar to the recombination of Sm and Fe in the HDDR process to some content. A moderate excess of Sm₂O₃ was required to compensate for Sm evaporation, critical to the generation of single-phase Sm₂Fe₁₇ powders without impurities (α-Fe or Sm-rich phases).⁵⁸ It has been demonstrated that excessive Sm₂O₃ may lead to the formation of a Sm-rich phase, which could be washed with dilute acid.⁵⁹ Besides, the amount of reduction agent should be larger than the theoretical value to guarantee that all Sm₂O₃ could be reduced.

Fig. 4 (a) Fabrication process for Sm₂Fe₁₇N₃ powders by reduction–diffusion from commercial micro-sized Sm₂O₃ and Fe applied as precursors. BSE images of the cross-section for the Sm₂Fe₁₇N₃ coarse powder converted from (b) washed Sm₂Fe₁₇ powders and (c) unwashed Sm₂Fe₁₇ powders. (b) and (c) Reproduced from ref. 60 with permission from the Institute of Physics, copyright 2011.

The formation mechanism of Sm₂Fe₁₇ powders using the reduction–diffusion process has been investigated by Asaki and co-workers in detail.⁶¹ The reaction kinetics was strongly related to the reaction temperature, which affected the reduction rate of Sm₂O₃ and the state of Sm. At 1300 K, the reduction rate of Sm₂O₃ was not high enough and Sm existed in a solid state. Thus, the inter-diffusion of Sm and Fe resulted in the formation of Sm₂Fe₁₇ and the growth rate of Sm₂Fe₁₇ followed the parabolic law. When the temperature was above the melting point of Sm (1347 K), the reduction rate of Sm₂O₃ increased and Fe would dissolve in the Sm or Sm–Ca melt. Sm₂Fe₁₇ and SmFe₂ would precipitate during the cooling process. Therefore, the high temperature may be not good for the formation of single-phase Sm₂Fe₁₇ because the reaction is similar to the peritectic reaction. Besides, the high temperature would lead to the growth and agglomeration of Sm₂Fe₁₇ powders, and the magnetic properties of Sm₂Fe₁₇N₃ prepared by the nitridation of such Sm₂Fe₁₇ powders would be degraded. Therefore, the temperature for the reduction–diffusion process should be in the range of the melting points of Ca and Sm.

Intuitively, the by-product CaO and the residual reduction agent should be directly eliminated by water washing after the reduction–diffusion process. Due to the low solubility of generated Ca(OH)₂, a long-term washing was necessary, which inevitably led to oxidation and corrosion of the obtained Sm₂Fe₁₇ powders.^62,63 The surface oxidation and corrosion were not beneficial to nitridation. Although Sm₂Fe₁₇ powders were exposed to a careful nitridation process, it is still possible that some Sm₂Fe₁₇ powders may not be fully nitrided, which would deteriorate the coercivity, saturation magnetization, and squareness ratio. Fortunately, the nitridation process was modified by Ishikawa and co-workers, where Sm₂Fe₁₇ powders were nitrided without washing because the surface of such Sm₂Fe₁₇ powders was quite clean.⁶⁰ The nitrogen distribution was studied by BSE imaging of the cross-section of coarse Sm₂Fe₁₇N_x powders. As shown in Fig. 4b, there were some white cores in the BSE image of Sm₂Fe₁₇N_x converted from washed Sm₂Fe₁₇, which corresponded to un-nitrided Sm₂Fe₁₇. In the BSE image of Sm₂Fe₁₇N_x prepared from unwashed Sm₂Fe₁₇ powders (Fig. 4c), white cores could almost not be observed. The ratio of the white core area to the whole powder area in Fig. 4b and c was 0.73% and 0.04%, respectively, which proved that unwashed Sm₂Fe₁₇ powders with a clean surface could promote the nitridation process dramatically. Hence, nitridation of unwashed Sm₂Fe₁₇ powders has been widely used for the preparation of Sm₂Fe₁₇N₃ powders nowadays.

After the nitridation process, the by-product CaO and residual reduction agent still need to be removed. Using a direct water washing process, a metastable SmFe₅ phase with poor hard magnetic properties was detected using synchrotron X-ray diffraction, which is attributed to the crystallization of the Sm-rich phase driven by the reaction heat between Ca and H₂O.⁵⁹ In addition, oxygen could stabilize SmFe₅, and the higher the oxygen content, the worse the magnetic properties and heat resistance. To achieve higher magnetic performance, the washing process was carried out in a glove box and ethylene glycol was applied to eliminate CaO and Ca. However, the efficiency is far lower than that of the water-washing process. Fortunately, a much more practical strategy has been developed by Okada and co-workers.⁵⁹ Before the washing process, the powder was subjected to dilute air at ambient temperature to slowly oxidize residual Ca. The slowly-oxidized powders could be directly washed with water or weak acid in air. The magnetic properties of Sm₂Fe₁₇N₃ powders washed after pre-oxidation were better than those washed without pre-oxidation.

Concerning the mechanical mixing of Sm₂O₃, Fe, and Ca powders to prepare Sm₂Fe₁₇ powders, there remained two major problems. Insufficient mixing would result in impurities in the products due to the inadequate reaction. In addition, the size of the obtained Sm₂Fe₁₇ powders is quite large, unfavorable for the magnetic properties. Fortunately, the Sm–Fe–O precursors have been designed using the wet-chemistry method to control the phase and size of Sm₂Fe₁₇ and hinder agglomeration. The design of Sm–Fe–O precursors for Sm₂Fe₁₇N₃ powders with high performance is discussed in the next section.

3.1.3. HDDR. HDDR process refined the grain size of coarse Sm₂Fe₁₇ powders without significantly changing the original particle size before the nitridation process.⁶⁴ Thus, nitridation of such Sm₂Fe₁₇ powders becomes easier and would generally result in magnetically isotropic Sm₂Fe₁₇N_x powders.

Under the H₂ atmosphere, when the temperature increased to 250–300 °C, Sm₂Fe₁₇ would absorb H atoms, resulting in the formation of Sm₂Fe₁₇H_x. The higher the pressure of H₂, the easier it is for H atoms to enter the lattice of Sm₂Fe₁₇. Increasing the temperature to 500 °C, the disproportionation reaction would proceed near the grain boundaries and SmH₂ as well as α-Fe with a microcrystalline or amorphous structure would be generated. With a further increase in temperature, SmH₂ and α-Fe fully crystallized. The grain size and morphology of SmH₂ and α-Fe were essential to the control of the size and microstructure of recombined Sm₂Fe₁₇. Therefore, the disproportionation of Sm₂Fe₁₇H_x was conducted at around 750 °C to achieve a small grain size of SmH₂ and α-Fe.¹⁷ Once the temperature was high enough, SmH₂ and α-Fe could recombine to form Sm₂Fe₁₇ even in an H₂ atmosphere, whereas high temperature induced grain growth.

To obtain recombined Sm₂Fe₁₇ powders with fine grain size, SmH₂ and α-Fe with small grain sizes were subjected to vacuum conditions. When the temperature was higher than 600 °C, desorption of SmH₂ and recombination of highly reactive Sm and α-Fe occurred almost simultaneously. Recombination of Sm and Fe was ascribed to the inter-diffusion of Sm and Fe, and a low temperature is necessary to hinder the grain growth.⁶⁴

Even after the HDDR process, some un-recombined α-Fe and large recombined grains may be detected, whereas single-phase Sm₂Fe₁₇ powder with a uniform and fine grain is essential for the coercivity of Sm₂Fe₁₇N_x powders. Homma and co-workers have demonstrated that sufficient time for the recombination of Sm and α-Fe is a prerequisite to avoid residual α-Fe.⁶⁵ However, the prolonged recombination process may lead to Sm loss and grain growth. Later, Kwon and co-workers applied the HDDR treatment repeatedly on the Sm₂Fe₁₇ powders to eliminate un-recombined α-Fe and large recombined grains as well as obtain homogeneous and fine Sm₂Fe₁₇ grains.⁶⁶ Even though the strategy worked, the process cost too much time.

In order to fully achieve the recombination process in a short period and avoid residual α-Fe, some other elements have been introduced into the Sm–Fe binary system to regulate the disproportionation and recombination process, which might also affect the phase and microstructure of recombined Sm₂Fe₁₇.^66,67 For example, Ga addition decreases the recombination temperature, making it much easier for full recombination, and the lower recombination temperature is beneficial for inhibiting grain growth.⁶⁸

It has been anticipated that magnetically anisotropic Sm₂Fe₁₇N₃ powders could be obtained by the HDDR process. However, there were almost no reports of anisotropic Sm₂Fe₁₇N₃ powders prepared using the HDDR process. Fortunately, anisotropic Nd₂Fe₁₄B powders have been fabricated using the HDDR process. Strictly speaking, the grains in anisotropic Nd₂Fe₁₄B powders by the HDDR process were highly textured. The texture could be ascribed to the crystallographic orientation of Fe₂B grains which “memorize” the orientation of initial Nd₂Fe₁₄B powders.⁶⁹ It has been reported that the disproportionation of Sm₂Fe₁₇ could result in SmH₂ with the same orientation whereas the orientation could not be maintained after the recombination process.⁷⁰ Hence, more efforts are deserved to prepare anisotropic Sm₂Fe₁₇N₃ powders with uniform and fine grain size using the HDDR process and the microstructure change during the HDDR process should be of special concern.

3.1.4. Other processes. Previously, homogenized Sm₂Fe₁₇ ingots were directly crushed to coarse powders and then nitrided to prepare Sm₂Fe₁₇N₃ powders.^71,72 Due to the large size of Sm₂Fe₁₇ grain, anisotropic Sm₂Fe₁₇N₃ powders could be obtained. However, the large grain size does not lead to nitridation. In addition, the homogenization of Sm₂Fe₁₇ ingots costs too much time. So, the investigation of such methods for Sm₂Fe₁₇ and Sm₂Fe₁₇N₃ powders is declining.

Mechanical alloying for the preparation of Sm₂Fe₁₇ powders has also been studied. The high-energy ball milling of Sm and Fe powders and the following nitridation would result in nanocrystalline Sm₂Fe₁₇N₃ powders. Therefore, such Sm₂Fe₁₇N₃ powders are magnetically isotropic.^18,19,73 The coercivity of such Sm₂Fe₁₇N₃ powders could reach up to 4.4 T due to the strong pinning.⁷⁴ However, the remanence of such powders is not high, limiting the maximum energy product.

3.2. Nitridation

N atoms enter 9(e) octahedral interstices in the Sm₂Fe₁₇ lattice through a gas–solid reaction, resulting in the formation of Sm₂Fe₁₇N_x powders.^75,76 First, N₂ or NH₃ is physically adsorbed on the surface of Sm₂Fe₁₇ powders and decomposed into N atoms with high activity. Then, N atoms diffused into Sm₂Fe₁₇ powders to form the nitrided phase. Finally, N atoms diffused from the fully nitrogenated phase to the poorly nitrogenated phase due to the concentration gradient and the diffusion of N atoms proceeded by voidal diffusion in Sm₂Fe₁₇.^77,78 The N atoms in the 9(e) site jumped to the thermodynamically unstable 18(g) site and then to a new 9(e) site. Due to the anisotropic lattice structure of Sm₂Fe₁₇, the diffusion of N atoms was anisotropic.⁷⁹ The diffusion in the c plane is isotropic and the diffusivity in the c plane was about 0.3 times that along the c axis. It should be noted that the diffusion of N atoms required energy to break the strong bonding between N atoms and neighboring atoms and the energy was almost independent of the nitrogen source.⁷⁷

Based on Fick's law, the size of Sm₂Fe₁₇ powders markedly affected the diffusion of N. Numerous reports have proved that Sm₂Fe₁₇ powders with a size of tens of micrometers could be nitride, and it has been demonstrated that the N diffusion coefficient increased with the decrease in particle size (5–500 μm).⁷⁸ So many efforts have been made to optimize the nitridation process for the improvement of nitridation efficiency. Generally, the nitridation process was regulated by the nitrogen source, temperature, pressure, time, and magnetic field, which will be discussed in this section.

3.2.1. Nitrogen source, temperature, pressure, and time. It was quite easy to regulate the nitrogen source, temperature, pressure, and time during the nitridation process, which would affect the nitridation kinetics and nitrogen content in the final products. Generally, a nitrogen source could be N₂, a mixture of N₂ and H₂, NH₃, or a mixture of NH₃ and H₂. With the increase in nitridation temperature, the nitridation efficiency would increase. However the nitridation temperature should be lower than 550 °C to avoid decomposition. In terms of pressure, the increase in gas pressure could promote the solid–gas reaction. Thus, higher pressure of the nitrogen source is beneficial to the nitridation of Sm₂Fe₁₇. Even if the nitrogen source, temperature, and pressure were optimal, sufficient time was necessary for the homogeneous distribution of the N atom.⁸⁰

Using N₂ as a nitrogen source, Sm₂Fe₁₇N_2.7 powder that was not fully nitrided was usually achieved under atmospheric pressure due to the low activity of N₂. Therefore, the increase in N₂ pressure was investigated to increase the nitrogen content. Koyama and co-workers studied the nitridation of Sm₂Fe₁₇ under N₂ with a pressure ranging from 0.01 to 6 MPa.^81,82 As shown in Fig. 5a, with N₂ pressure below 0.05 MPa, the nitridation process is mainly controlled by the diffusion of N. Thus, Sm₂Fe₁₇ powders were hard to be fully nitrided and N atoms in the obtained Sm₂Fe₁₇N_x powders were distributed in a gradient. Once the N₂ pressure was above 0.1 MPa, the grain growth of the fully nitrided phase would be dominant and Sm₂Fe₁₇N₃ could be achieved. In addition, higher pressure of N₂ could inhibit the formation of α-Fe during the nitridation process. From an empirical point of view, the nitridation rate at N₂ pressure of 0.1 MPa should be higher than that of 0.05 MPa. However, it was revealed that the nitrogen rate with the N₂ pressure of 0.1 MPa was lower than that at 0.05 MPa in the early nitridation stage. In other words, low N₂ pressure promoted the formation of incompletely nitrided Sm₂Fe₁₇N_x and high N₂ pressure was beneficial to the formation of Sm₂Fe₁₇N₃. Therefore, the nitridation efficiency could be raised by the combined pressure process with N₂ pressure of 0.05 and 0.1 MPa in early and later stages rather than by just increasing N₂ pressure.

Fig. 5 (a) Schematic of the nitridation process of Sm₂Fe₁₇ powders at different N₂ pressure. (a) Reproduced from ref. 82 with permission from Elsevier, copyright 2022. (b) Effect of H₂ content on the composition. (b) Reproduced from ref. 76 with permission from Elsevier, copyright 2004. (c) Effect of magnetic field on the nitrogen content and the saturation moment. (c) Reproduced from ref. 83 with permission from the Japan Institute of Metals, copyright 2020.

In addition to increasing the N₂ pressure, a mixture of N₂ and H₂ was also applied to adjust the nitrogen content and nitridation efficiency. As mentioned earlier, Sm₂Fe₁₇ powders would absorb H, which would induce strain, resulting in an incoherent interface between Sm₂Fe₁₇N_x and Sm₂Fe₁₇. The incoherent interface was beneficial to the diffusion of N atoms. Besides, the absorption of H would also lead to cracks in the powders, increasing the diffusion paths. It was thought that H atoms entered the lattice of Sm₂Fe₁₇, which resulted in lattice expansion, beneficial to N atoms diffusion. However, the introduction of H atoms did almost not change the activation energy for diffusion.⁷⁸ Using the plasma of a mixture of N₂ and H₂ as a nitrogen source, Gama and co-workers obtained Sm₂Fe₁₇N₃ powders with a nitridation temperature of 400 °C.⁷⁶ As shown in Fig. 5b, when the concentration of H₂ was low, poor-crystalline or amorphous Sm₂Fe₁₇N₈ and Sm₂Fe₁₇N₁₁ phases rich in nitrogen would be generated. Though excessive N atoms could be desorbed by annealing, they resulted in distortion or destruction of the crystal lattice. Increasing H₂ concentration, the nitrogen-rich phase would reduce and Sm₂Fe₁₇N₃ would become the dominant phase. However, the high H₂ concentration may lead to the formation of α-Fe.

For higher nitridation efficiency, NH₃ was applied as a nitrogen source, which decomposed into N₂ with high activity and H₂. Sm₂Fe₁₇N₁₀ could be easily obtained under the NH₃ atmosphere. To avoid irreversible lattice distortion induced by excessive N atoms, NH₃ was mixed with H₂ to suppress the activity of N atoms.^84,85 Due to the high nitridation efficiency, the mixture of NH₃ and H₂ was widely used despite the potential dangers.

As mentioned earlier, the nitridation efficiency was improved by regulation of nitrogen source, plasma, pressure increase of nitrogen source, and so on. However, the nitridation process still needed several to dozens of hours and a little Sm₂Fe₁₇N_x may decompose during the nitridation process. Therefore, efforts are deserved for the optimization of the nitridation process to achieve short-time nitridation.

3.2.2. Magnetic field. It has been demonstrated that a magnetic field could promote the nitridation of Sm₂Fe₁₇ because of the reaction involved with a gain in magnetic free energy (G_M) under an external magnetic field. Just considering Zeeman energy, G_M could be expressed as
where M, T, and H represented the magnetic moment, temperature, and magnetic field, respectively.^86,87 The Curie temperature of Sm₂Fe₁₇ is far lower than the nitridation temperature, so nitridation of Sm₂Fe₁₇ would theoretically become easier in favor of G_M. Furthermore, the diffusion of N atoms from the fully nitrogenated phase to the poorly nitrogenated phase would also be enhanced by the external magnetic field for the magnetic moment of poor nitrogenated Sm₂Fe₁₇ was inferior to that of Sm₂Fe₁₇N₃.

Koyama and co-workers performed the nitridation of Sm₂Fe₁₇ under a zero-field and in-field of 5 T.⁸³ As shown in Fig. 5c, compared to Sm₂Fe₁₇N_x powders prepared under a zero-field, the magnetic field induced 0.4–0.6 N atoms increase per unit cell. Inspiringly, with the assistance of a magnetic field, Sm₂Fe₁₇N_2.9 has been prepared at an N₂ pressure of 0.1 MPa at 743 K, attributed to that G_M approximately corresponded to a thermal energy of 35 K. For a more comprehensive study of the magnetic field effect, magnetostriction should also be considered, although it is not quite easy.

4. Magnetic properties optimization

As mentioned earlier, surface quality and particle size optimization of Sm₂Fe₁₇N₃ powders could effectively inhibit the nucleation of the reverse domain, resulting in the enhancement of magnetic properties. In addition, some other elements are doped into Sm₂Fe₁₇N₃ for higher remanence, coercivity, Curie temperature, oxidation resistance, and so on.

4.1. Surface quality

Sm₂Fe₁₇ strips or ribbons should be roughly crushed for nitridation and then the nitrides were further pulverized for coercivity optimization. The crushing process would cause damage to the surface which is not beneficial to the magnetic properties.

The surface quality of crushed Sm₂Fe₁₇ powders is not commonly concerned. Recently, it was pointed out that the pulverization process resulted in poor crystallinity and nanocrystalline grains on the surface of Sm₂Fe₁₇ powders even though the jet milling was carried out in a glove box.⁸⁸ During the jet milling process, the impact, friction, and shear led to transient high temperatures at the powder surface. Due to that, residual oxygen in the jet-milling chamber would absorb on the surface and the region rich in Sm and O as well as α-Fe would generate. It is anticipated that the wet milling process is applied to crush Sm₂Fe₁₇ whereas the effect of solvent on the nitridation process is not studied.

Sm₂Fe₁₇N₃ powders with an average diameter (D₅₀) ranging from 1.1 μm to 3.0 μm have been obtained by jet-milling in a very low oxygen atmosphere.⁸⁹ With the decrease in powder size, the crystallinity became weaker, which was ascribed to the nanocrystalline region emerging at the surface induced by rapid heat during the milling process. Once the size was reduced to 1.3 μm, the best comprehensive magnetic properties could be achieved with a coercivity, remanence, and maximum energy product of 1.27 T, 147 emu g⁻¹, and 43.6 MGOe, respectively. Although the magnetic properties are excellent, it could be further optimized because the poor crystallinity deteriorated the degree of alignment (DOA) of the particles, further optimization of the magnetic properties is possible. It should be noted that Sm₂Fe₁₇N₃ powders with a size of 1.3 μm had a flake-like morphology which was not beneficial to magnetic alignment. Thus, DOA of such Sm₂Fe₁₇N₃ powders by wet-milling to hinder transient surface heating and morphology optimization is required to increase mobility.

In order to deliver the heat generated by friction, Zhang and co-workers applied a cryo-milling system to crush the Sm₂Fe₁₇N₃ powders.⁹⁰ Compared to room-temperature milling, a low-temperature would result in a smaller particle size (Fig. 6a) and less nitrogen loss (Fig. 6b). More importantly, the hysteresis loop revealed that the low-temperature milling process would inhibit the formation of soft magnetic matter (Fig. 6c). Anyway, the low-temperature ball milling is an effective process to hinder the oxidation and nitrogen loss.

Fig. 6 (a) The mean particle sizes of Sm₂Fe₁₇N_x nanoflakes milled at different times. (b) The nitrogen content of Sm₂Fe₁₇N_x nanoflakes milled at room temperature and low-temperature as a function of milling time. (c) Hysteresis loops of the magnetically aligned Sm₂Fe₁₇N_x nanoflakes milled for 4h at room temperature and low-temperature. (a)–(c) Reproduced from ref. 90 with permission from Elsevier, copyright 2020. (d) The morphology image of the milled Sm₂Fe₁₇N₃ powders ground in heptane with 100 wt% OA, (e) The hysteresis loop of loosely packed Sm₂Fe₁₇N₃ powders at room temperature, and (f) the curves of coercivity loss with time in air for coated and uncoated powders. (d)–(f) Reproduced from ref. 91 with permission from Elsevier, copyright 2014.

Yang and co-workers studied the effect of the milling medium on the morphology and magnetic properties of Sm₂Fe₁₇N₃ powders.⁹¹ Sm₂Fe₁₇N₃ powders with a size of 100 μm were milled with the assistance of a solution of heptane and oleic acid, resulting in the formation of closely packed Sm₂Fe₁₇N₃ powders with a dimensional size and thickness of about 0.8 μm and 0.1 μm, respectively (Fig. 6d). Oleic acid could promote the milling efficiency and was beneficial to the homogeneous pulverization and size reduction. However, the combination between oleic acid and Sm₂Fe₁₇N₃ powders was found to be ineffectual. To improve the dispersibility of milled Sm₂Fe₁₇N₃ powders, coarse Sm₂Fe₁₇N₃ powder was milled in a mixture of heptane, oleic acid, and silane. The silane coupling agent was coated on the surface, resulting in the formation of loosely packed Sm₂Fe₁₇N₃ powders. The loosely packed Sm₂Fe₁₇N₃ powders showed comprehensive magnetic properties with coercivity, remanence, and maximum energy product of 1.30 T, 155 emu g⁻¹, and 40 MGOe (Fig. 6e), respectively. The surface coating of Sm₂Fe₁₇N₃ powders with silane improved the resistance to the external environment. Compared to Sm₂Fe₁₇N₃ powders milled in heptane and oleic acid, the coercivity of silane-coated Sm₂Fe₁₇N₃ powders decreased more slowly with time (Fig. 6f), further indicating the effective binding between silane and powder surface.

4.2. Particle size

Although pulverization of coarse Sm₂Fe₁₇N₃ powders by ball milling was assisted by surfactants, surface defects, such as edges and strains, were hard to avoid, resulting in lower coercivity compared to Sm₂Fe₁₇N₃ powders with similar size converted from Sm₂Fe₁₇ powders prepared by using a reduction–diffusion process.³ In addition, Sm₂Fe₁₇N₃ powders might decompose during the ball milling process, deteriorating the magnetic properties.

The nitridation process would almost not lead to powder growth and agglomeration due to the low nitridation temperature. Thus, the preparation of monodispersed Sm₂Fe₁₇ powders is a prerequisite for the fabrication of Sm₂Fe₁₇N₃ powders with excellent magnetic properties. The high-temperature reduction–diffusion process easily led to the growth and sintering of Sm₂Fe₁₇ powders, and nitridation of such powders would result in Sm₂Fe₁₇N₃ powders with poor magnetic properties. Generally, reduction–diffusion process optimization and precursor design have been applied to hinder the sintering and agglomeration of Sm₂Fe₁₇ powders.

The reduction–diffusion temperature regulation is quite easy, directly affecting the growth and sintering of Sm₂Fe₁₇ powders. Hono and co-workers studied the influence of the preparation temperature on the size of Sm₂Fe₁₇N₃ powders.¹⁴ The Sm–Fe complex was obtained using a sol–gel method and then calcined to remove carbon in the atmosphere. Hydrogen reduction was followed to obtain a mixture of SmO_x and Fe. With the mixture as a precursor, Sm₂Fe₁₇ powders were synthesized using the reduction–diffusion process and converted into Sm₂Fe₁₇N₃ powders by nitridation. As displayed in Fig. 7a–c, Sm₂Fe₁₇N₃ powders with sizes ranging from 0.69 to 3.49 μm could be obtained by regulating the calcination, hydrogen reduction, and reduction–diffusion temperatures. The low reaction temperature favored small particle size. At the calcination, hydrogen reduction, and reduction–diffusion temperatures of 500, 700, and 900 °C, respectively, Sm₂Fe₁₇N₃ powders with a size of 0.69 ± 0.17 μm could be obtained (Fig. 7c) and exhibited a coercivity of 2.32 T (Fig. 7d). The large coercivity was attributed to the small particle size and the smooth surface. It should be noted that the Sauter mean diameter of the Sm₂Fe₁₇N₃ powders, obtained by laser diffraction, was larger than that counted by the SEM image, indicating that agglomeration of Sm₂Fe₁₇ powders was hard to avoid by decreasing the temperature. It was inspired that the reduction–diffusion temperature decreased to below 600 °C.⁹² LiCl–KCl with a eutectic temperature of 352 °C was applied as a solvent for Ca, which could guarantee the reduction of Sm. Although Sm₂Fe₁₇ was not achieved, the low reduction–diffusion temperature was important to further hinder the growth and agglomeration.

Fig. 7 SEM images of the Sm₂Fe₁₇N₃ particles with average sizes of (a) 3.49 μm, (b) 1.86 μm, and (c) 0.69 μm. (d) Demagnetization curves of Sm₂Fe₁₇N₃ powders shown in Fig. 7a–c. (a)–(d) Reproduced from ref. 14 with permission from Elsevier, copyright 2016.

To further reduce the size of Sm₂Fe₁₇N₃ powders, precursor design is also an effective strategy. By ultrasonic spray pyrolysis and hydrogen reduction (USP–HR), Che and co-workers prepared the precursor in which Sm₂O₃ and α-Fe were embedded uniformly in particles with an average size of 0.48 μm, which could promote the reduction–diffusion process.⁹³ After the reduction–diffusion and nitridation processes, Sm₂Fe₁₇N₃ powders with an average size of 0.61 μm have been fabricated and exhibited a coercivity of 1.47 T. Recently, reduction–diffusion and nitridation of Sm–Fe–O precursors prepared using the UPS–HR process was carried out in a rotary furnace, which reduced the size distribution of Sm₂Fe₁₇N₃. And the Sm₂Fe₁₇N₃ powders with an average diameter of 0.4 μm showed a coercivity of 3.17 T.¹⁵

In addition to the UPS-HR process, coprecipitation was also applied to prepare Sm–Fe–O precursors with small particle sizes.^59,94 The coprecipitated Sm(OH)₃ and Fe(OH)₃ were crushed until the mean size of secondary particles was smaller than 1 μm. Following the hydrogen reduction, reduction–diffusion, and nitridation process, submicron-sized Sm₂Fe₁₇N₃ powders could be obtained. Sm₂Fe₁₇N₃ powders with mean diameters of 0.6 μm (Fig. 8a) and 0.9 μm (Fig. 8b) have been obtained at reduction–diffusion temperatures of 900 °C and 950 °C. As shown in Fig. 8c, the Sm₂Fe₁₇N₃ powders with a mean diameter of 0.6 μm exhibited a coercivity of 2.28 T, which was directly washed with water. The coercivity could be optimized to 2.81 T by slow-oxidization before washing to avoid the formation of metastable SmFe₅. However, the saturation and remanence magnetizations were 134 and 100 emu g⁻¹. The low remanence ratio was ascribed to the agglomeration.

Fig. 8 SEM images of Sm₂Fe₁₇N₃ powders obtained at a reduction–diffusion temperature of (a) 900 °C and (b) 950 °C, and corresponding demagnetization curves with a mean diameter of (c) 0.6 μm and (d) 0.9 μm oriented in resin. (e) SEM images after disintegration and (f) demagnetization curves of the disintegrated powders. (a)–(f) Reproduced from ref. 94 with permission from Elsevier, copyright 2017.

Compared to the Sm₂Fe₁₇N₃ powders with a size of 0.6 μm, Sm₂Fe₁₇N₃ powders with a mean diameter of 0.9 μm had a smaller coercivity of 1.81 T and a larger saturation and remanence magnetization of 142 and 110 emu g⁻¹ (Fig. 8d), indicating the agglomeration of larger particles was partially inhibited. Even so, the maximum energy product of the Sm₂Fe₁₇N₃ powders with a size of 0.9 μm was only 22.8 MGOe. To further enhance the remanence and maximum energy product, Sm₂Fe₁₇N₃ powders with a size of 0.9 μm were subjected to ball milling for disintegration. As shown in Fig. 8e, the ball milling process effectively reduces the agglomeration of powders. After disintegration for 3 h, the Sm₂Fe₁₇N₃ powders showed excellent comprehensive magnetic properties with coercivity, saturation magnetization, remanence, and maximum energy product of 1.43 T, 156 emu g⁻¹, 141 emu g⁻¹ and 42.8 MGOe, respectively (Fig. 8f), indicating that disintegration of the agglomerated powders promotes magnetic alignment. The decreased coercivity was attributed to the defects resulting from the ball milling. Inspiringly, the submicron-sized Sm₂Fe₁₇N₃ powders displayed good thermal stability with a temperature coefficient of coercivity being −0.36%/K.

Upon the design of the precursor, efforts have been made to inhibit the agglomeration of submicron-sized Sm₂Fe₁₇N₃ powders. Okada and co-workers employed a hydrothermal method to prepare Fe₂O₃/CaCO₃, which was subsequently impregnated with Sm(NO₃)₃·6H₂O. After the hydrogen reduction, reduction–diffusion, and nitridation process, the resulting Sm₂Fe₁₇N₃ powders had a size of 0.6 μm, yet the remanence ratio was only 75%.⁹⁵ It seems as though the precursor design did not work as expected. However, monodispersed submicron-sized SmCo₅ powders have been obtained based on a similar precursor design.^51,52 It is worthwhile to design an Sm–Fe–O precursor for the preparation of monodispersed Sm₂Fe₁₇N₃ powders, which would possess both high coercivity and remanence.

4.3. Elemental substitution

As mentioned earlier, the introduction of other elements into the Sm–Fe binary system has been explored to prevent the formation of α-Fe and Sm-rich phases. Though the magnetic properties of Sm₂(Fe_1−xM_x)₁₇N₃ are dominated by interstitial N atoms, where M represents the substitution elements, such as Co or Mn, substitution has a significant effect on the magnetic properties.^96,97

Co-substitution changes the lattice parameters, which is responsible for the optimization of the overall intrinsic magnetic properties with an appropriate substitution content. The Curie temperature and magnetocrystalline anisotropy could be enhanced by small Co contents, whereas the saturation magnetization is slightly optimized by Co substitution. The best comprehensive magnetic properties could be achieved when the content of Co is 20%.⁹⁸

Different from Co doping, Mn substitution enhances the coercivity due to changes in the microstructure. Anisotropic Sm₂Fe₁₇N₃ powders show a single-crystal feature, whereas anisotropic Sm₂(Fe, Mn)₁₇N_x powders have a cell-like structure.^99–101 Sm₂(Fe, Mn)₁₇N_x powders consist of crystalline and amorphous phases. The crystalline phase is Sm₂(Fe, Mn)₁₇N_x, and the amorphous phase is rich in Mn and N. With the increase in nitrogen content, the fraction of the amorphous phase increases and the crystalline phase would be surrounded by the amorphous phase. Additionally, the magnetic domain would become smaller with the increase in nitrogen content. Magnetic domain walls would be pinned by the amorphous phase during the magnetic reversal process, thus enhancing the coercivity. Thus, Sm₂(Fe, Mn)₁₇N_x with x larger than 3 may display higher coercivity. It should be noted that crystalline Sm₂(Fe, Mn)₁₇N_x grains in Sm₂(Fe, Mn)₁₇N_x powders retain their orientation, ensuring that the powders are magnetically anisotropic. Despite the coercivity optimization, Mn substitution leads to a decrease in saturation magnetization.

Additionally, Sm can be partially substituted by Nd or Pr to alter the intrinsic magnetic properties of Sm₂Fe₁₇N_x.^102,103 However, such substitutions of rare earth elements result in a decrease in the anisotropic field and Curie temperature to varying degrees, due to the easy plane anisotropy of Nd₂Fe₁₇N_x and Pr₂Fe₁₇N_x.

It is evident that the progress in the elemental substitution of Sm₂Fe₁₇N₃ was not good enough to improve the magnetic properties. Referring to the great progress and wide applications of Nd₂Fe₁₄B-based magnets with elemental substitution,^104–106 multi-element substitution may be a promising strategy for magnetic properties and thermal stability optimization.

From a global perspective, great progress has been made in high-performance anisotropic Sm₂Fe₁₇N₃ powders by rapid-quenching and reduction–diffusion processes, whereas large-scale production of high-performance anisotropic Sm₂Fe₁₇N₃ powders remains a great challenge, which is essential to the fabrication of bonded Sm₂Fe₁₇N₃ magnets. The two methods suffer from complex and time-consuming processes, which suggest high costs. More importantly, controlling the composition and size of Sm₂Fe₁₇N₃ powders is hard to achieve. Especially, the nitridation process, which is sensitive to the atmosphere, pressure, and surface of Sm₂Fe₁₇, could easily lead to uneven nitridation or the presence of soft magnetic Fe. So, the key to the mass production of high-performance anisotropic Sm₂Fe₁₇N₃ powders may be the development of advanced equipment that could achieve homogeneous reaction and controlled loss of Sm.

5. Conclusion and perspective

In this review, we have summarized the fundamentals, synthesis methods, and regulation strategies of Sm₂Fe₁₇N₃ powders, guiding the design and preparation of Sm₂Fe₁₇N₃ powders with enhanced magnetic properties. The perspectives on Sm₂Fe₁₇N₃ powders are discussed as follows.

On the one hand, achieving precise control over the chemical composition, such as preparation of single-phase Sm₂Fe₁₇ without a significant loss of magnetic properties and enhancement of the nitridation efficiency, remains a significant challenge. The introduction of additional elements into the Sm–Fe binary system may offer opportunities to optimize the formation and nitridation of Sm₂Fe₁₇. Moreover, elemental substitution has the potential to enhance thermal stability and oxidation resistance. On the other hand, further investigations into microstructure control of Sm₂Fe₁₇N₃ powders are essential for magnetic properties improvement. The ideal state for Sm₂Fe₁₇N₃ powders would involve achieving a critical single-domain size with a smooth and non-oxidized surface. Therefore, concerted efforts should be directed towards attaining this ideal state, either through the pulverization of large powders or via the reduction–diffusion process.

Data availability

This is a review, the data should be found from original articles in the references.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the National Key R&D Program of China (2022YFB3505900), the Natural Science Basic Research Program of Shaanxi (2024JC-YBQN-0406), and the Qinchuangyuan Project of Shaanxi Province (QCYRCXM-2022-155, QCYRCXM-2022-267, QCYRCXM-2022-295).

Notes and references

1. J. M. D. Coey, Perspective and prospects for rare earth permanent magnets, Engineering, 2020, 6, 119–131.
2. R. W. McCallum, L. Lewis, R. Skomski, M. J. Kramer and I. E. Anderson, Practical aspects of modern and future permanent magnets, Annu. Rev. Mater. Res., 2014, 44, 451–477.
3. K. Takagi, Y. Hirayama, S. Okada, W. Yamaguchi and K. Ozaki, Novel powder processing technologies for production of rare-earth permanent magnets, Sci. Technol. Adv. Mater., 2021, 22, 150–159.
4. N. Imaoka, E. Kakimoto, K. Takagi, K. Ozaki, M. Tada, T. Nakagawa and M. Abe, Progress with insulating nanocomposites based on ferrite plating of Sm₂Fe₁₇N₃ micropowders, J. Magn. Magn. Mater., 2019, 476, 613–621.
5. J. Cui, M. Kramer, L. Zhou, F. Liu, A. Gabay, G. Hadjipanayis, B. Balasubramanian and D. Sellmyer, Current progress and future challenges in rare-earth-free permanent magnets, Acta Mater., 2018, 158, 118–137.
6. T. Horikawa, M. Yamazaki, M. Matsuura and S. Sugimoto, Recent progress in the development of high-performance bonded magnets using rare earth-Fe compounds, Sci. Technol. Adv. Mater., 2021, 22, 729–747.
7. A. Hosokawa and Y. Hirayama, Cold consolidation and coercivity enhancement in Sm₂Fe₁₇N₃ magnets by high-pressure torsion, Scripta Mater., 2023, 226, 115251.
8. C. Lu, J. Zhu, J. Gong and X. Gao, A method to improving the coercivity of sintered anisotropic Sm-Fe-N magnets, J. Magn. Magn. Mater., 2018, 461, 48–52.
9. K. Takagi, R. Soda, M. Jinno and W. Yamaguchi, Possibility of high-performance Sm₂Fe₁₇N₃ sintered magnets by low-oxygen powder metallurgy process, J. Magn. Magn. Mater., 2020, 506, 166811.
10. W. Yamaguchi, R. Soda and K. Takagi, Role of surface iron oxides in coercivity deterioration of Sm₂Fe₁₇N₃ magnet associated with low temperature sintering, Mater. Trans., 2019, 60, 479–483.
11. C. Lu, X. Hong, X. Bao, X. Gao and J. Zhu, Changing phase equilibria: a method for microstructure optimization and properties improvement in preparing anisotropic Sm₂Fe₁₇N₃ powders, J. Alloys Compd., 2019, 784, 980–989.
12. V. Kraposhin, A. Kamynin, E. Khotulev and A. Everstov, Study of phase composition, structure and magnetic properties of high-coercivity alloys based on Sm₂Fe₁₇ system depending on technology of their manufacture, J. Phys. Conf. Ser., 2018, 1134, 012029.
13. M. Katter, J. Wecker and L. Schultz, Structural and hard magnetic properties of rapidly solidified Sm-Fe-N, J. Appl. Phys., 1991, 70, 3188–3196.
14. Y. Hirayama, A. Panda, T. Ohkubo and K. Hono, High coercivity Sm₂Fe₁₇N₃ submicron size powder prepared by polymerized-complex and reduction–diffusion process, Scripta Mater., 2016, 120, 27–30.
15. S. Okada, E. Node, K. Takagi and R. Hashimoto, Synthesis of ultra-high coercivity Sm₂Fe₁₇N₃ powder by homogeneous reduction-diffusion with rotary furnace, J. Alloys Compd., 2023, 960, 170726.
16. M. Röhrig, R. G. T. Fim, L. T. Quispe, F. J. G. Landgraf, C. H. Ahrens, C. C. C. Plá and P. A. P. Wendhausen, On the feasibility of using Sm₂Fe₁₇N_x powders obtained via HDDR process for laser powder bed fusion of bonded permanent magnets, J. Magn. Magn. Mater., 2023, 565, 170273.
17. J. Ye, Y. Liu, M. Li, S. Gao and M. Tu, Structural change in Sm₂Fe₁₇ alloys during homogenization and HD processing, Mater. Charact., 2007, 58, 108–113.
18. J. Pan, J. Shentu, C. He, D. Zeng, F. Gao, G. Dodbiba and T. Fujita, Preparation of temperature-sensitive rare-earth-iron alloy fine particles using mechanical alloying and sintering, IEEE Magn. Lett., 2023, 14, 1–5.
19. I. Nehdi, L. Bessais, C. Djega-Mariadassou, M. Abdellaoui and H. Zarrouk, X-ray and Mössbauer studies of Sm₂Fe_17-xCr_x materials synthesized by mechanical alloying followed by an appropriate short annealing, J. Alloys Compd., 2003, 351, 24–30.
20. K. Kobayashi, T. Akiya and H. Kato, Magnetic anisotropy in Sm₂Fe₁₇N_x (0 < x < 3) of intermediate nitrogen concentrations, J. Alloys Compd., 2006, 408–412, 1404–1408.
21. L. Ye, F. Wang, Y. Liu, H. Zhou, L. Liu, Y. Ding, Y. Sun and A. Yan, New structural features and pinning-evolved coercivity mechanism: a potential route for developing high coercivities in anisotropic Sm-Fe-N material, J. Mater. Res. Technol., 2024, 30, 451–460.
22. J. M. D. Coey, P. Stamenov, S. B. Porter, M. Venkatesan, R. Zhang and T. Iriyama, Sm-Fe-N revisited; remanence enhancement in melt-spun Nitroquench material, J. Magn. Magn. Mater., 2019, 480, 186–192.
23. K. H. Kim, J. Kim, D. Shindo, I. Takashi and K. Ohmori, Magnetization process in Sm₂Fe₁₇N₃ fine powders studied using Lorentz microscopy and electron holography, Mater. Trans., 2010, 51, 1302–1307.
24. J. Xu, K. Zhu, S. Gao and Y. Hou, Rare earth permanent magnetic nanostructures: chemical design and microstructure control to optimize magnetic properties, Inorg. Chem. Front., 2021, 8, 383–395.
25. Z. Ma, J. Mohapatra, K. Wei, J. P. Liu and S. Sun, Magnetic nanoparticles: synthesis, anisotropy, and applications, Chem. Rev., 2023, 123, 3904–3943.
26. S. Wang, J. Xu, W. Li, S. Sun, S. Gao and Y. Hou, Magnetic nanostructures: rational design and fabrication strategies toward diverse applications, Chem. Rev., 2022, 122, 5411–5475.
27. R. Skomski, K. H. Müller, P. A. P. Wendhausen and J. M. D. Coey, Magnetic reversal in Sm₂Fe₁₇N_y permanent magnets, J. Appl. Phys., 1993, 73, 6047–6049.
28. J. M. D. Coey and T. Iriyama, Bonded Sm-Fe-N permanent magnets, Woodhead Publishing, 2022, pp. 305–342.
29. E. A. Gorbachev, E. S. Kozlyakova, L. A. Trusov, A. E. Sleptsova, M. A. Zykin and P. E. Kazin, Design of modern magnetic materials with giant coercivity, Russ. Chem. Rev., 2021, 90, 1287.
30. X. Yu, M. Yue, W. Liu, Z. Li, M. Zhu and S. Dong, Structure and intrinsic magnetic properties of MM₂Fe₁₄B (MM=La, Ce, Pr, Nd) alloys, J. Rare Earths, 2016, 34, 614–617.
31. Z. Li, M. Yue, X. Liu, H. Zhang, W. Liu, Y. Li and Z. Zhang, Tuning the structure and intrinsic magnetic properties of Ce₂Fe₁₄B alloys by elimination of CeFe₂ with La substitution, J. Magn. Magn. Mater., 2020, 505, 166747.
32. S. Aich, D. Satapathy and J. Shield, Rapidly solidified rare-earth permanent magnets: processing, properties, and performance, CRC Press, 2017, 453–508.
33. S. Brennan, X. L. Rao, R. Skomski, N. Dempsey and J. M. D. Coey, Magnetic properties of Sm₂Fe₁₇N_x, x = 3.9, J. Magn. Magn. Mater., 1996, 157–158, 510–511.
34. H. S. Li and J. M. Cadogan, Variation of A₂₀ with nitrogen content in Sm₂Fe₁₇N_3−δ compounds, Solid State Commun., 1991, 80, 905–908.
35. R. Skomski and D. J. Sellmyer, Anisotropy of rare-earth magnets, J. Rare Earths, 2009, 27, 675–679.
36. C. N. Christodoulou and T. Takeshita, Preparation, structural and magnetic properties and stability of interstitial Sm₂Fe₁₇-carbonitrohydrides, J. Alloys Compd., 1993, 198, 1–24.
37. H. Horiuchi, U. Koike, H. Kaneko, T. Kurino and H. Uchida, Effects of N, C and B additions on the Sm₂Fe₁₇ crystal structure and magnetic properties, J. Alloys Compd., 1995, 222, 131–135.
38. J. Wang, W. Cai, J. Zheng, J. Zhu, X. Zhang, H. Chen, L. Qiao, Y. Ying, W. Li, J. Yu, J. Li, Y. Su and S. Che, Preparation of ultrafine Sm₂Fe₁₇N₃ magnetic powders without ball milling, J. Mater. Res. Technol., 2023, 27, 6688–6695.
39. J. M. D. Coey, Magnetism and Magnetic Materials, Cambridge University Press, 2010, 1–617.
40. Z. Ma, H. Liu and Z. Luo, A general method for the large-scale preparation of multifunctional magnetic Sm-Fe/Co nanomaterials, Adv. Funct. Mater., 2023, 33, 2301350.
41. G. Deng, Q. Jing, X. Wang, G. He and X. Ye, Synthesis mechanism of Sm₂Fe₁₇ alloy produced in reduction-diffusion process, J. Rare Earths, 2010, 28, 420–424.
42. R. Kuchi, D. Schlagel, T. A. Seymour-Cozzini, J. V. Zaikina and I. Z. Hlova, Exploiting mechanochemical activation and molten-salt-assisted reduction-diffusion approach in bottom-up synthesis of Sm₂Fe₁₇N₃, J. Alloys Compd., 2024, 980, 173532.
43. Y. Luo, D. Yu, H. Li, W. Zhuang, Y. Mao, X. Chen, K. Li and Y. Liu, Structure and permanent magnetic properties of SmFe_x (x = 3-8) melt spun ribbons during heat treatment, J. Rare Earths, 2014, 32, 960–964.
44. Z. Sun, H. Zhang, S. Zhang, J. Wang and B. Shen, Structure and magnetic properties of Sm₂Fe_17-xMn_x compounds, J. Phys. D: Appl. Phys., 2000, 33, 485.
45. M. Xing, J. Han, Y. Zhang, S. Liu, Z. Chen, C. Wang, J. Yang, H. Du, Y. Yang and M. Yue, Nitrogenation effect of Sm₂Fe₁₇ alloys prepared by strip casting technique, J. Appl. Phys., 2015, 117, 17A732.
46. J. Yang, H. Chen, Y. Zhang, Y. Yang, X. Chen, S. Liu, C. Wang, J. Hang, H. Du, G. Lian and Y. Yang, Magnetostrictions of Sm₂Fe₁₇ and Sm₂Fe₁₇N₃, IEEE Trans. Magn., 2011, 47, 3621–3624.
47. G. Lin, W. Bi, X. Bao and W. Cao, Preparation of Sm₂Fe₁₇ columnar grains ribbons by rapid quenching, Adv. Mater. Res., 2014, 1004–1005, 367–370.
48. M. Xing, J. Han, F. Wan, S. Liu, C. Wang, J. Yang and Y. Yang, Preparation of anisotropic Sm₂Fe₁₇N_X magnetic materials by strip casting technique, IEEE Trans. Magn., 2013, 49, 3248–3250.
49. D. A. Kolodkin, A. G. Popov, A. V. Protasov, V. S. Gaviko, D. Y. Vasilenko, S. Kavita, D. Prabhu and R. Gopalan, Magnetic properties of Sm_2+αFe₁₇N_x powders prepared from bulk and strip-cast alloys, J. Magn. Magn. Mater., 2021, 518, 167416.
50. D. Liang, W. Yang, X. Wang, Q. Xu, J. Han, S. Liu, C. Wang, H. Du, X. Zhu, T. Yuan, Z. Luo and J. Yang, Study of the anisotropic Sm₂Fe₁₇N₃ powders with high performance, AIP Adv., 2023, 13, 025104.
51. C. Li, Q. Wu, Z. Ma, H. Xu, L. Cong and M. Yue, A novel strategy to synthesize anisotropic SmCo₅ particles from Co/Sm(OH)₃ composites with special morphology, J. Mater. Chem. C, 2018, 6, 8522–8527.
52. H. Tang, M. A. H. Mamakhel and M. Christensen, Enhancing the coercivity of SmCo₅ magnet through particle size control, J. Mater. Chem. C, 2020, 8, 2109–2116.
53. J. Xu, K. Zhu, W. Li, X. Wang, Z. Yang, Y. Hou and S. Gao, First-order-reversal-curve analysis of rare earth permanent magnet nanostructures: insight into the coercivity enhancement mechanism through regulating the Nd-rich phase, Inorg. Chem. Front., 2021, 8, 1975–1982.
54. Z. Ma, M. Yue, H. Liu, Z. Yin, K. Wei, H. Guan, H. Lin, M. Shen, S. An, Q. Wu and S. Sun, Stabilizing hard magnetic SmCo₅ nanoparticles by N-Doped graphitic carbon layer, J. Am. Chem. Soc., 2020, 142, 8440–8446.
55. B. Shen, C. Yu, A. A. Baker, S. K. McCall, Y. Yu, D. Su, Z. Yin, H. Liu, J. Li and S. Sun, Chemical synthesis of magnetically hard and strong rare earth metal based nanomagnets, Angew. Chem., Int. Ed., 2019, 58, 602–606.
56. J. C. Boareto, J. Soyama, M. D. V. Felisberto, R. Hesse, A. V. A. Pinto, T. R. Taylor and P. A. P. Wendhausen, Sm₂Fe₁₇ prepared by calciothermic reduction-diffusion using different iron powders, Mater. Sci. Forum, 2007, 534–536, 1365–1368.
57. S. Sugimoto, Current status and recent topics of rare-earth permanent magnets, J. Phys. D: Appl. Phys., 2011, 44, 064001.
58. H. Chen, J. Zheng, L. Qiao, Y. Ying, L. Jiang and S. Che, Preparation of Sm₂Fe₁₇ alloy by reduction-diffusion process, Rare Met., 2018, 37, 989–994.
59. S. Okada, E. Node, K. Takagi, Y. Fujikawa, Y. Enokido, C. Moriyoshi and Y. Kuroiwa, Synthesis of Sm₂Fe₁₇N₃ powder having a new level of high coercivity by preventing decrease of coercivity in washing step of reduction-diffusion process, J. Alloys Compd., 2019, 804, 237–242.
60. T. Ishikawa, K. Yokosawa, K. Watanabe and K. Ohmori, Modified process for high-performance anisotropic Sm₂Fe₁₇N₃ magnet powder, J. Phys. Conf. Ser., 2011, 266, 012033.
61. T. Tanabe, Y. Nagai, T. Kubota and Z. Asaki, Formation of Sm-Fe intermetallic compounds by the reduction-diffusion process with CaH₂, Mater. Trans., 1992, 33, 1163–1170.
62. J. Xu, J. Zheng, H. Chen, L. Qiao, Y. Ying, W. Cai, W. Li, J. Yu, M. Lin and S. Che, Corrosion mechanism of H₂O on Sm₂Fe₁₇ and nitrogenation process of corroded Sm₂Fe₁₇, J. Rare Earths, 2019, 37, 638–644.
63. R. Matsuda, M. Matsuura, N. Tezuka, S. Sugimoto, T. Ishikawa and Y. Yoneyama, Preparation of highly heat-resistant Sm-Fe-N magnetic powder by reduction-diffusion process, Mater. Trans., 2020, 61, 2201–2207.
64. X. Zhao, Z. Zhang, W. Liu, Q. Xiao and X. Sun, Structural and magnetic properties of Sm-Fe-N magnets prepared by hydrogenation and nitrogenation processes, J. Magn. Magn. Mater., 1995, 148, 419–425.
65. M. Okada, K. Saito, H. Nakamura, S. Sugimoto and M. Homma, Microstructural evolutions during HDDR phenomena in Sm₂Fe₁₇N_x compounds, J. Alloys Compd., 1995, 231, 60–65.
66. K. Hae-Woong, Study of the effect of a repeated HDDR treatment on the coercivity of two rare-earth-transition metal hard magnetic materials, J. Phys. D: Appl. Phys., 2001, 34, 36.
67. J. Sun, C. Cui, Y. Zhang, L. Li, J. Gao and Y. Liu, Structural and magnetic properties of Sm₂Fe_17-xNb_x (x = 0-4) alloys prepared by HDDR processes and their nitrides, Rare Met., 2006, 25, 129–137.
68. A. Teresiak, O. Gutfleisch, N. Mattern, B. Gebel and K. H. Müller, Phase formation and crystal structure of Sm₂Fe_17-yGa_y compounds during hydrogen disproportionation and desorption recombination (HDDR-process), J. Alloys Compd., 2002, 346, 235–243.
69. H. Sepehri-Amin, T. Ohkubo, K. Hono, K. Güth and O. Gutfleisch, Mechanism of the texture development in hydrogen-disproportionation–desorption-recombination (HDDR) processed Nd-Fe-B powders, Acta Mater., 2015, 85, 42–52.
70. M. Xing, J. Han, Z. Lin, F. Wan, S. Liu, C. Wang, Q. Xu, J. Yang and Y. Yang, Self-organized nanorods in disproportionated Sm₂Fe₁₇ and NdFe_10.5Mo_1.5 alloys, J. Magn. Magn. Mater., 2013, 326, 201–204.
71. N. Mommer, M. Kubis, M. Hirscher, M. Gerlach, J. V. Lier, K. H. Müller and H. Kronmüller, Measurement of N and C diffusion in Sm₂Fe₁₇ by magnetic relaxation, J. Alloys Compd., 1998, 279, 113–116.
72. J. Wang, W. Li, X. Zhong, Y. Gao, W. Qin, N. Tang, W. Lin, J. Zhang, R. Zhao, Q. Yan and F. Yang, Study on high performance Sm₂Fe₁₇N_x magnets, J. Alloys Compd., 1995, 222, 23–26.
73. C. H. Lee and Y. S. Kwon, Effect of mechanical alloying on the formation of Sm₂Fe₁₇N_x compound, Met. Mater. Int., 2002, 8, 151–154.
74. W. Liu, Q. Wang, X. Sun, X. Zhao, T. Zhao, Z. Zhang and Y. Chuang, Metastable Sm-Fe-N magnets prepared by mechanical alloying, J. Magn. Magn. Mater., 1994, 131, 413–416.
75. H. Uchida, T. Kawanabe, S. Tachibana, K. Kinoshita, Y. Matsumura, M. Tada, H. H. Uchida, H. Kaneko, T. Kurino and M. Sato, Nitrogen absorption by Sm₂Fe₁₇, J. Ceram. Soc. Jpn., 2006, 114, 896–901.
76. J. D. Ardisson, R. C. Araujo, W. A. A. Macedo, A. I. C. Persiano and S. Gama, The effect of atmosphere composition in plasma nitrogenation of Sm₂Fe₁₇, J. Magn. Magn. Mater., 2004, 272–276, 1859–1861.
77. C. N. Christodoulou and N. Komada, On the atomic diffusion mechanism and diffusivity of nitrogen atoms in Sm₂Fe₁₇, J. Alloys Compd., 1994, 206, 1–13.
78. H. Uchida, S. Tachibana, T. Kawanabe, Y. Matsumura, V. Koeninger, H. H. Uchida, H. Kaneko and T. Kurino, Diffusion behavior of N atoms in Sm₂Fe₁₇, J. Alloys Compd., 1995, 222, 107–112.
79. K. Kobayashi, M. Ohmura, Y. Yoshida and M. Sagawa, The origin of the enhancement of magnetic properties in Sm₂Fe₁₇N_x (0<x≲3), J. Magn. Magn. Mater., 2002, 247, 42–54.
80. C. Cui, J. Sun, R. Wang and Z. Liang, The effect of nitridation on the microstructure and lattice parameters of fine grained Sm₂Fe₁₇N_x, Superlattice. Microst., 2006, 39, 406–413.
81. H. Fujii, K. Tatami and K. Koyama, Nitrogenation process in Sm₂Fe₁₇ under various N₂-gas pressures up to 6 MPa, J. Alloys Compd., 1996, 236, 156–164.
82. J. Takahashi, Y. Mitsui, M. Onoue, R. Kobayashi and K. Koyama, Nitridation kinetics of Sm₂Fe₁₇ probed using Mössbauer spectroscopy, J. Magn. Magn. Mater., 2022, 554, 169295.
83. K. Koyama, M. Onoue, R. Kobayashi, Y. Mitsui, R. Y. Umetsu and Y. Uwatoko, Magnetic field effect on nitrogenation of Sm₂Fe₁₇, Mater. Trans., 2020, 61, 1487–1491.
84. T. Iriyama, K. Kobayashi, N. Imaoka, T. Fukuda, H. Kato and Y. Nakagawa, Effect of nitrogen content on magnetic properties of Sm₂Fe₁₇N_x (0 < x < 6), IEEE Trans. Magn., 1992, 28, 2326–2331.
85. Y. Wei, K. Sun, Y. Fen, J. Zhang, B. Hu, Y. Wang, X. Rao and G. Liu, Structural and intrinsic magnetic properties of Sm₂Fe₁₇N_y (y = 2-8), J. Alloys Compd., 1993, 194, 9–12.
86. Y. Mitsui, R. Y. Umetsu, K. Koyama and K. Watanabe, Magnetic-field-induced enhancement for synthesizing ferromagnetic MnBi phase by solid-state reaction sintering, J. Alloys Compd., 2014, 615, 131–134.
87. M. Onoue, R. Kobayashi, Y. Mitsui, R. Y. Umetsu, Y. Uwatoko and K. Koyama, In-field heat treatment effect on nitridation of Sm₂Fe₁₇, Mater. Trans., 2019, 60, 2179–2182.
88. A. Hosokawa, K. Suzuki, W. Yamaguchi and K. Takagi, Mechanism of anomalous α-Fe formation from stoichiometric Sm₂Fe₁₇ jet-milled powder during post-pulverization annealing, Acta Mater., 2021, 213, 116981.
89. A. Hosokawa, W. Yamaguchi, K. Suzuki and K. Takagi, Influences of microstructure on macroscopic crystallinity and magnetic properties of Sm-Fe-N fine powder produced by jet-milling, J. Alloys Compd., 2021, 869, 159288.
90. Y. Li, F. Wang, J. P. Liu, F. Wang, S. Wang and J. Zhang, Fabrication of remarkably magnetic-property-enhanced anisotropic Sm₂Fe₁₇N_x nanoflakes by surfactant assisted ball milling at low temperature, J. Magn. Magn. Mater., 2020, 498, 166191.
91. X. Ma, L. Li, S. Liu, B. Hu, J. Han, C. Wang, H. Du, Y. Yang and J. Yang, Anisotropic Sm-Fe-N particles prepared by surfactant-assisted grinding method, J. Alloys Compd., 2014, 612, 110–113.
92. J. Kim, S. Okada and K. Takagi, Low-temperature synthesis of Sm-Fe binary compounds by reduction-diffusion process using an eutectic salt solvent, Rare Met., 2023, 42, 1724–1731.
93. J. Zheng, S. Tian, K. Liu, W. Cai, Y. Tang, L. Qiao, Y. Ying, W. Li, J. Yu, Y. Liu and S. Che, Preparation of submicron-sized Sm₂Fe₁₇N₃ fine powder by ultrasonic spray pyrolysis-hydrogen reduction (USP-HR) and subsequent reduction-diffusion process, AIP Adv., 2020, 10, 055119.
94. S. Okada, K. Suzuki, E. Node, K. Takagi, K. Ozaki and Y. Enokido, Preparation of submicron-sized Sm₂Fe₁₇N₃ fine powder with high coercivity by reduction-diffusion process, J. Alloys Compd., 2017, 695, 1617–1623.
95. S. Okada, K. Suzuki, E. Node, K. Takagi, K. Ozaki and Y. Enokido, Improvement of magnetization of submicron-sized high coercivity Sm₂Fe₁₇N₃ powder by using hydrothermally synthesized sintering-tolerant cubic hematite, AIP Adv., 2017, 7, 056219.
96. B. Hu, X. Rao, J. Xu, G. Liu, F. Cao, X. Dong, H. Li, L. Yin and Z. Zhao, Magnetic properties of Sm₂(Fe_1-xM_x)₁₇N_y nitrides (M = Co, Ni, Al, Ti, V), J. Magn. Magn. Mater., 1992, 114, 138–144.
97. M. Matsuura, K. Yarimizu, Y. Osawa, N. Tezuka, S. Sugimoto, T. Ishikawa and Y. Yoneyama, Preparation of Mn-diffused Sm-Fe-N core-shell powder by reduction-diffusion process, J. Magn. Magn. Mater., 2019, 471, 310–314.
98. M. Katter, J. Wecker, C. Kuhrt, L. Schultz and R. Grössinger, Structural and intrinsic magnetic properties of Sm₂(Fe_1−xCo_x)₁₇N_y, J. Magn. Magn. Mater., 1992, 114, 35–44.
99. K. Majima, M. Ito, S. Katsuyama and H. Nagai, Structure and magnetic properties of Sm₂(Fe,Mn)₁₇N_x coarse powders with high coercivity, J. Appl. Phys., 1997, 81, 4530–4532.
100. A. Hosokawa and K. Takagi, Multi-scale electron microscopy of overnitrided Sm-Fe-Mn-N powder, AIP Adv., 2018, 8, 055031.
101. A. Yasuhara, H. S. Park, D. Shindo, T. Iseki, N. Oshimura, T. Ishikawa and K. Ohmori, Microstructures and magnetic domain structures in Sm₂(Fe,Mn)₁₇N_δ powders studied by analytical electron microscopy and Lorentz microscopy, J. Magn. Magn. Mater., 2005, 295, 1–6.
102. M. Katter, J. Wecker, C. Kuhrt, L. Schultz, X. C. Kou and R. Grössinger, Structural and intrinsic magnetic properties of (Sm_1−xNd_x)₂Fe₁₇N≈2.7 and (Sm_1−xNd_x)₂(Fe_1−zCo_z)₁₇N≈2.7, J. Magn. Magn. Mater., 1992, 111, 293–300.
103. Y. Zeng, Z. Lu, N. Tang, X. L. R. Zhao and F. Yang, Structural, magnetic and microscopic physical properties of (Sm, Pr)₂Fe₁₇ and their nitrides, J. Magn. Magn. Mater., 1995, 139, 11–18.
104. W. Chen, J. Jin, J. Yu, L. Zhou, B. Peng, S. Fu, X. Liu, G. Bai and M. Yan, Outstanding anti-corrosion performance in Nd-Fe-B permanent magnets by constructing a hydrophobic triplex surface coating, J. Mater. Chem. C, 2023, 11, 6884–6893.
105. Z. Yu, X. Liu, J. He, J. Cao, W. Fan, Y. Wu, H. Yu, Z. Liu, Z. Wang and E. Niu, Tb-Cu grain boundary diffusion effects on single- and multi-main-phase Nd-Fe-B based magnets, J. Mater. Chem. C, 2023, 11, 645–655.
106. J. Zhang, X. Liao, K. Xu, J. He, W. Fan, H. Yu, X. Zhong and Z. Liu, Enhancement in hard magnetic properties of nanocrystalline (Ce,Y)–Fe–Si–B alloys due to microstructure evolution caused by chemical heterogeneity, J. Mater. Chem. C, 2020, 8, 14855–14863.

---

Footnote

† These authors contributed equally.

---