Current progress and future challenges in rare-earth-free permanent magnets

  Permanent magnets (PM) are critical components for electric motors and power generators. Key properties of permanent magnets, especially coercivity and remanent magnetization, are strongly dependent on microstructure.

  Understanding metallurgical processing, phase stability and microstructural changes are essential for designing and improving permanent magnets.

  The widely used PM for the traction motor in electric vehicles and for the power generator in wind turbines contain rare earth elements Nd and Dy due to their high maximum energy product. Dy is used to sustain NdFeB's coercivity at higher temperature.

  Due to the high supply risk of rare earth elements (REE) such as Dy and Nd, these elements are listed as critical materials by the U.S. Department of Energy and other international institutes.

  Other than Dy, finer grain size is also found to have effect on sustaining coercivity at higher temperature.

  A proper control of phase stability and microstructures has direct impact on mitigating REE supply risk.

  Compared to rare earth PMs, non-rare earth (non-RE) PMs typically have lower maximum energy products, however, given their small supply risks and low cost, they are being intensively investigated for less-demanding applications.

  The general goal for the development of non-RE PMs is to fill in the gap between the most cost-effective but low performing hard ferrite magnet and the most expensive but high performing RE PMs.

  In the past five years great progress has been made toward improving the microstructure and physical properties of non-RE PMs.

  Several new candidate materials systems were investigated, and some have showed realistic potential for replacing RE PMs for some applications.

  In this article, we review the science and technology of various types of non-RE materials for PM applications. These materials systems include Mn based, high magnetocrystalline anisotropy alloys (MnBi and MnAl compounds), spinodally decomposing alloys (Alnico), high-coercivity tetrataenite L10 phase (FeNi and FeCo), and nitride/carbide systems (such as a" based, high saturation magnetization Fe16N2 type phase and Co2C/Co3C acicular particle phase).

  The current status, challenges, potentials as well as the future directions for these candidates non-RE magnet materials are discussed.

  1. Introduction

  2. MnAl

  3. MnBi

  4. Alnico type magnet systems

  5. Tetrataenite L10 FeNi

  6. L10-FeCo

  7. HfCo and ZrCo based systems

  8. Carbides

  9. Iron nitride (a00-Fei6N2)

  10. Summary and future challenges

  11. Acknowledgements

  12. References

  Nomenclature

  RE········· ···Rare Earth

  REE············Rare Earth Element

  PM········· ···Permanent Magnet

  f.u.···········Formula Unit

  Tc ············Curie Temperature

  β ········ ····Density

  Ms ············Saturation Magnetization

  Mr ············Remanent Magnetization

  Br ············Remanent Magnetic Flux Density

  K1 ············Magnetocrystalline Anisotropy Field

  Ha ············Anisotropy Field

  Hci ············Intrinsic Coercivity

  (BH)max········· Maximum Energy Product

  1. Introduction

  NdFeB magnets are widely used for conversion between electricity and mechanical energy.

  Unfortunately, magnetic properties of NdFeB is strongly temperature dependent [1]. The (BH)max of the NdFeB magnet falls sharply once temperature exceeds 373 K (as shown in Fig. 1).

  Permanent magnet motors and generators using NdFeB magnets must consider some form of a cooling mechanisms in order to dissipate the heat generated by eddy current and friction.

  Better thermal stability (up to 50% increase of (BH)max at 473 K) can be achieved by adding Pr and Dy [2]. Unfortunately, Dy is scarce in the U.S. The U.S. Department of Energy (DOE) categorized Dy as the single most critically threatened strategic metal to U.S [3]. The demand of Nd-Fe-B magnets rapidly increases as the world pursues energy efficiency and renewable energy.

  According to the 2014 and 2017 USGS Minerals Commodities Summaries (Fig. 2), the average price of Dy2O3 jumped from $245/kg in 2010 to $1410/kg in 2011, then gradually retreated to $185/kg in next 5 years; the price of Nd2O3 also jumped, from $88/kg in 2010 to $195/kg in 2011, then gradually retreated to $39/kg in next 5 years [4].

  In 2016, the price of Nd2O3 started to increase again. By the summer of 2018, it has reached $69/kg, while that of the Dy2O3 reached $236/kg [5]. The increase in Nd2O3 average price is the combined result of increased demand in REE, crack-downs by the Chinese government on smuggling REE, and consolidation of small REE mines.

  Interestingly, there appears to be a divergence between the average prices of the two oxides, which were coupled in past decade. This divergence is evidence of successful efforts to reduce the usage of the most critical element Dy. In 2017, China and some European countries announced that internal combustion engine will be phased out in one to two decades.

  With foreseeable large increase in the number of electrical vehicles and their dependence on permanent magnet traction motors, the trend of increasing Nd price is expected to continue.

  Strategies to address the REE criticality issue include increasing and diversifying supply and reducing demand.

  While several mines outside of China, in Australia, Vietnam and the US, have opened and have begun production of REE elements, the most desired heavy REE elements, in particular, Dy and Tb, remain low in supply because none of these newly opened mines are rich in heavy REE reserve.

  To reduce the demand of heavy REE elements, alternative magnet technologies that use much less Dy have been developed.

  Significant progress has been made in the direction of reducing the grain size of NdFeB magnet and relying on the size of the fine grains to provide the needed coercivity [6,7].

  According to a market report produced by MCP [8] in 2015, the sales of Ferrite, NdFeB, SmCo, and Alnico, are $4344 M, $2927 M, $722 M, and $355 M respectively.

  About half of the market is occupied by hard ferrite because of its low cost, and because its properties are good enough for most motors that do not have a demanding requirement on power density. However, for those motors that do have a tight weight and size limitation, such as the traction motor in electric vehicles, REE magnets are the only choice; as for those motors that have moderate requirement on power density, a magnet with moderate energy product at a lower price is desired.

  Unfortunately, there is a performance gap between hard ferrite magnets (<5 MGOe) and REE magnets (>30 MGOe). Alnico would be a good choice if its cost were in between ferrite and NdFeB magnets. The current practice is to use smaller or diluted REE magnets.

  This miniaturization approach often reduces overall system cost, but it does impose avoidable demand on the already strained REE resources.

  With increasing REE prices, this approach will become too expensive.

  A long-term approach to reducing use of REE is to develop non-RE magnets that can fill in this gap between hard ferrite and REE magnets.

  Table 1 lists the properties and prices of various rare earth and non-rare earth magnets in 2016 and projected values in 2022.

  The table uses $/kg/kG/kOe to represent the cost properties ratio, which is the magnet cost ($/kg) divided by remanent magnetization (kG) and by coercivity (kOe).

  For any non-RE magnet to be a viable alternative, it needs to offer a better $/kg/kG/kOe value than that of ferrite and NdFeB.

Fig. 1. Temperature dependence of (BH)max for most commercial permanent magnets. The value in parentheses in (BH)max at 298 K. (Data extracted from Ref. [1], [184], [219]).

  Fig. 1. Temperature dependence of (BH)max for most commercial permanent magnets. The value in parentheses in (BH)max at 298 K. (Data extracted from Ref. [1], [184], [219]).

Fig. 2. REE Oxides prices over the past eight years. The 2010 to 2017 data is extracted from the USGS Minearl Commodity Summaries; the 2018 data was obtained from mineral.com on May 9, 2018.

  Fig. 2. REE Oxides prices over the past eight years. The 2010 to 2017 data is extracted from the USGS Minearl Commodity Summaries; the 2018 data was obtained from mineral.com on May 9, 2018.

  2. MnAl

  Manganese metal by itself is antiferromagnetic but when alloyed with other elements, it can be a strong ferromagnet.

  While quite a few Mn-based systems exhibit ferromagnetism, including Mn-B, Mn-Ga, Mn-Ge, Mn-Sb and Mn-As, only Mn-Al and Mn-Bi alloys have potential as a gap magnet. MnAl has strong magnetocrystalline anisotropy, good resistance to corrosion, and low density; perhaps more importantly, its key constituent elements Mn and Al are abundant. Yet, the metallurgy and magnetism of the t-MnAl are remarkably complex. It took nearly two decades from the 1960s to develop commercial MnAl magnets possessing only half of the theoretically possible energy density and exhibiting a fairly modest coercivity [9].

  The metastable t-MnAl compound with the tetragonal L10 structure is formed at 51e58 at.% Mn [10,11].

  In the equilibrium Mn-Al alloys [12] (Fig. 3), this range of compositions correspond to mixture of the two phases, g2-Mn5Al8 and solid solution of Al in b-Mn. In the ideal equiatomic and perfectly ordered t-MnAl compound, the Mn and Al atoms occupy the (0,0,0) and (%,%,%) sites, respectively.

Mole Fraction MnFig. 3. Phase diagram of Mn-Al alloy system [12] with approximate range of metastable t phase indicated. The ferromagnetic t structure forms through different mechanisms; one of them consists in ordering of the Mn and Al atoms in paramagnetic £ structure followed by shear transformation (see text).

  Fig. 3. Phase diagram of Mn-Al alloy system [12] with approximate range of metastable t phase indicated. The ferromagnetic t structure forms through different mechanisms; one of them consists in ordering of the Mn and Al atoms in paramagnetic £ structure followed by shear transformation (see text).

  Ab initio calculations have produced somewhat controversial results for the magnetic ground state: ferromagnetism [13e15] with a magnetic moment of 2.37e2.40 mB/f.u. (i.e., 162e164 emu/g) and a magnetocrystalline anisotropy constant of 1.50e1.77 MJ/m3, but also antiferromagnetism [16]. Both the theory [13,14,17] and experiment [18 ,19 ] agree that in the real compounds, which are neither equiatomic nor perfectly ordered, the Mn atoms occupying the “wrong” (%,%,%) sites have their magnetic moments ordered antiferromagnetically with the rest Mn atoms.

  The experimentally observed room-temperature Ms of the bulk t-MnxAhoo_x is at its highest 112 emu/g with x = 51; and it decreases with increasing x(Fig.4)[20].The Tcand Haincreasewithx, reaching at x = 56 values of 650 K and 5.3 T, respectively. The addition of carbon increases the Ms to the maximum value of 128 emu/g, while simultaneously decreasing the Tc and Ha. Based on these Ms values and densities [18,21] of5.16g/cm3 and 5.10g/cm3, the upperlimits of(BH)maxforthe MneAl and MneAleC alloys can be estimated as 13.2 MGOe and 16.8 MGOe, respectively.

  Magnetic hysteresis of the MneAl alloys is usually associated [22,23] with domain wall pinning at stacking faults and antiphase boundaries (although the latter ones were also shown to serve as nucleation sites [24,25]), as well as at distortions in the crystal structure caused by the local stress fields of dislocations [26].

  Pinning at twin boundaries e which are also commonly present in this material e was found to be weak [27], and micromagnetic simulations [25] singled out the twins as defects particularly harmful for both the coercivity and squareness of the hysteresis loop.

  It should be noted that although the nucleation-controlled coercivity is rarely brought into consideration in the context of the MneAl alloys [28], the demagnetization behavior of small recrystallized t grains present in the hot-deformed materials [27]is consistent with the nucleation mechanism.

Table 1 Comparison of motor magnet price and properties in 2016 and 2022 (estimated).

Table 1 Comparison of motor magnet price and properties in 2016 and 2022 (estimated).

  Fig. 4. Curie temperature, room-temperature saturation magnetization and roomtemperature anisotropy field of t-MnxAl100-x and t-Mn54Al46Cy compounds according to Pareti et al. [20].

  Fig. 4. Curie temperature, room-temperature saturation magnetization and roomtemperature anisotropy field of t-MnxAl100-x and t-Mn54Al46Cy compounds according to Pareti et al. [20].

  The challenges for MnAl to achieve the entitled maximum energy product come from the difficulty in obtaining high phase fraction of the t-MnAl phase and from the difficulty in achieving the microstructure with the magnetic easy axis of each grain properly aligned. The t phase usually forms from the hexagonal e phase (a chemically disordered A3 structure) which is stable above 1136 K. The maximum undercooling of the e-MnxAl100-x phase is expected at x = 55 at.% [29], which may explain the off-stoichiometric compositions of the t phase.

  Unfortunately, the typical approach for developing coercivity using the ball milling method is accompanied by a pronounced decrease of the Ms.

  This decrease may be caused by the accumulation of stacking faults inthet crystallites [30]; the shorter Mn-Mn distances through the stacking faults supposedly result in antiferromagnetic coupling in the regions adjacent to the faults.

  A deformation-induced site disorder is another possible cause [22], although Ms was found to decline even in the absence of any detectible disorder [31]. At least in some cases, the loss of the magnetization must simply reflect a partial loss of the t phase, as deformation is known to further destabilize this metastable structure [31e34]. Part ofthe lost magnetization can be recovered through an additional low-temperature annealing [35].

  The t-MnAl phase is typically obtained by controlled cooling of the e phase or by quenching followed by aging at 573e823 K.

  It can also be produced mechanically [36,37], supposedly being induced by microstrain generated in the e phase through milling. Studies suggest several modes of the e / t transformation.

  The transformation may begin with the formation of ordered e0 domains of the orthorhombic B19 structure in the e phase [38].

  The misfit between the e0 and e lattices leads to a coherency strain and the eventual formation and accumulation of stacking faults. The latter is equivalent to emergence of the fcc stacking sequence (ABC) instead of the hcp stacking sequence (ABAB) of the £ phase. Finally, atomic ordering converts the transitional fcc structure into the t phase. This e / fcc / t transformation is sometimes referred to as “shear followed by ordering."

  In the other mode, schematically shown in Fig. 3, the disordered e phase first transforms into the ordered e0 phase [39,40].

  The subsequent shear in the close-packed (100)e0 planes transforms the e0 structure into the t phase. This e / e' / t transformation maybe described as “ordering followed by shear."

  Formation of the t phase may also occur through a compositionally invariant, diffusional nucleation-and-growth “massive” transformation [41] in which nucleation of the t phase at the grain boundaries of the parent e phase is followed by growth via the motion of mostly incoherent hetero-phase interface segments. The massive and displacive mechanisms were shown to operate within the same sample [42], and may even be considered within a framework of a hybrid, displacive-diffusional mechanism.

  The e / t transformation mode is also sensitive to the annealing temperature [43]: in the Mn54Al46 alloy, for example, it is displacive below 648 K, but both displacive and massive at higher temperatures.

  Formation of the t-MnAl does not necessarily involve the e phase.

  An off-stoichiometric t-MnxAl100-x (x z 60) crystallizes in MneAl thin films [44e46], particularly when they are grown epitaxially [47e49].

  The direct formation of the t phase was reported by high-pressure synthesis [50], electrodeposition [51] and by annealing of amorphous MneAleC alloys at the relatively high temperature of 1173 K [52].

  It might have also occurred in the specific alloys solidified through splat quenching [53], meltspinning [36], strip-casting [54] and after induction melting [55], although observation of the t phase in as-solidified alloys does not completely rule out a transient formation of the e phase.

  Decomposition of the metastable t-MnAl is accelerated by deformation [33,34] and is slowed down by the addition of up to z1.5 at.% C [56].

  The latter effect has been explained [57] as inhibition of Al and Mn diffusion by the interstitial C atoms. Several other elements were also reported to stabilize the t-MnAl phase against decomposition: Cu, Fe, Ni, Co, Cr, Ti, Mo, B and Zn [58,59]; whereas a few percent of Ga reportedly make the phase thermodynamically stable [60].

  The highest room-temperature coercivities reported for the MneAl alloys had been obtained in thin films, with their Hc reaching 10.7 kOe [61].

  Melt-spinning of the MneAl(eC) alloys often produces nearly pure e phase [36,62e65], but a Hc obtained by simply annealing the melt-spun ribbons to induce the e / t transformation does not exceed 2 kOe [63e68]. Somewhat higher Hc values, up to 3.3 kOe, were obtained via mechanical alloying [69e71].

  Grinding or milling of MneAl alloys already consisting of the t phase can develop Hc of 4.5e6.5 kOe [31,32,55,65,72e75].

  If the MneAl(eC) alloys consisting of the e phase were intensively milled before being converted into the t, the resulting nanostructure exhibited Hc of 2.5e5 kOe [31,35,37,68,76e79].

  Studies on comparing milling of the e phase with milling of the t phase [31,35,79] produced mixed results, which show the importance of the specific milling conditions. Relatively high Hc values of 7.5 kOe and 5.6 kOe were obtained via electrodeposition [51] and arc-discharge [80], respectively. Surfactant-assisted milling [31,74,79] produced flake-shaped particles. Micron-size MneAl(eC) particles exhibiting Hc of 3.9 kOe and submicron particles (2 kOe) have been prepared via spark erosion [81] and mechanochemical synthesis [82], respectively.

  Consolidation of the magnetically hard MneAl(eC) powders into magnets was attempted via spark plasma sintering [83,84], warm compaction [85] and equal-channel angular extrusion [86]. The latter technique allowed for a relatively high Hc, up to 4.4 kOe, but the magnets consolidated from powders lacked crystallographic texture which is critical for the maximum use of their magnetization. When deformed, the MneAl magnets do develop a [001] texture perpendicular to the deformation [85,87] through mechanical twinning on {111} planes.

  Unfortunately, with respect to functional MnAl magnets, the achievements of the 1960s and 1970s still remain unsurpassed. The best combinations of remanence and coercivity are those obtained by swaging the MneAl alloys at room temperature [72,88,89] or extruding MneAleC at approximately 973 K [9], [39], [90] (the addition of C made it possible to deform the metastable t-phase at elevated temperatures; the swaged/extruded magnets are often also alloyed with Ti, B, or Ni). The warm extrusion followed by annealing yielded anisotropic magnets exhibiting, despite the fairly modest Hc of 3 kOe, the highest (BH)max of 7e8 MGOe [9,87,90].

  3. MnBi

  MnBi exhibits a first-order structural transition between temperatures of613K and 628K [91]. The low temperature phase (LTP) possesses the desired magnetic properties. Researchers also refer this phase as a-MnBi. The Mn-Bi phase diagram is shown in Fig. 5 [92].

  Key phase transformations are:

  1) at 628 K upon heating, a-MnBi / MniaBi + Bi-rich liquid;

  2) at 719 K upon heating MniosBi / Mn + Bi-rich liquid;

  3) at 613 K during cooling MniogBi / a-MnBi + Mn;

  4) at 535 K during cooling, MnBi + Bi / Bi-rich liquid.

Fig. 5. Mn-Bi phase diagram. Reprinted with permission of ASM International.
Fig. 5. Mn-Bi phase diagram. Reprinted with permission of ASM International.

  Currently the generally accepted composition, lattice structure and parameters for the high temperature phase (b-MnBi) at 630 K are Mn2.23Bi1.88, Pmma, and a = 5.959 A, b = 4.334 A, c = 7.505 A, a = b = g = 90°, respectively [93]. For the LTP phase (a-MnBi) at 300 K, they are Mn50Bi50, P63/mmc, and a = b = 4.290 A, c = 6.126 A, a = b = 90°, g = 120°, respectively. For the a-MnBi at 20 K, they are Mn5°Bi50, Cmcm, and a = 4.269 A, b = 7.404 A, and c = 6.062 A, a = b = g = 90°, respectively.

  The magnetic structure of MnBi goes through a spinreorientation process with temperature increasing from 0 K. Neutron and X-ray diffractions carried out by Cui [94], Yang [95] andMcGuire[96] showed that the magnetic easy axis of MnBi is inplane at lower temperature and starts to reorient out-of-plane to c-axis at about 90 K, and completely aligned with c-axis when temperature is>250K; and the magnetic moments at 10, 80, 300, and 400 K are 4.18, 4.24, 3.84, and 3.56 mB/f.u. Zarkevich estimated the magnetic moment at 0 K as 3.96 mB with site-projected moments of 4.231 and —0.273 mB on Mn and Bi, respectively [97]. These recent results mostly agree with the results obtained by Robert [98] and Andresen [99] in the 60s'. Andresen also showed that above the transition between 613 and 633 K, 15% of the Mn atoms have left their regular lattice sites and occupied the large trigonal bipyra-midal holes of the NiAs structure. This also explains the observed contraction of the c-axis.

  Saturation magnetization of a-MnBi is relatively low, and it decreases rapidly with increasing temperature. At 10, 80, 300, and 400 K, the magnetization of a-MnBi at 9 T are 88.5, 89.2, 81.3, and 75.3 emu/g, respectively. Assuming a density of 8.9 g/cm3, the magnetization of a-MnBi at 300 K is 9.1 kG. The a-MnBi has large magnetocrystalline anisotropy, about 1.6MJ/m3 at 300K, and it increases to 2.2 MJ/m3 at 400 K. Among known hard magnetic materials, a-MnBi and (Ba, Sr) Hexaferrites are the two materials with intrinsic coercivity (Hci) increasing with temperature [100,101]. Yang showed that the Hci of resin-bonded a-MnBi magnet rapidly increases from 13 kOeat 300K to 20kOeat 400K, resulting in a positive temperature coefficient of 0.54%/K [102]. In comparison, the Hci of BaFe12O19 increases from a relatively low value at much slower rate, from 3.5 kOe at 300 K to 4.5 kOe at 420 K (0.2%/K) [101]; and the Hci of NdFeB decreases sharply from 14 kOe at 300 K to 4 kOe at 423 K (—0.6%/K) [1].

  A sample exhibiting large coercivity typically has large magnetocrystalline anisotropy energy and fine grains. For a-MnBi powder with 45 mm size, the coercivity at room temperature is typically about 2 kOe. It can quickly increase to 13 kOe once the particle size is reduced to 5 mm. The theoretical maximum energy product (BH)max of a-MnBi is estimated to be 20 MGOe at 300 K and 17 MGOe at 400 K using the formula (BH)max = MS/4, where Ms is the saturation magnetization (9.1 kG and 8.4 kG at 300 K and 400 K, respectively).In2015,17MGOehasbeendemonstratedwitha1 mm film by Zhang and Sellmyer [103]; and 8.7 MGOe was prepared in bulkform(~1 cm3 cube) byCui's group [104] and 7.3 MGOe(3cm3) byKim's group [105]. Interestingly, Guillaud in his 1943 PhD thesis had predicted 18 MGOe could be reached if fully dense and pure MnBi phase can be obtained [106].

  The reason for the bulk sample exhibiting less than half of the entitled maximum energy product can be attributed to the difficulty in obtaining high purity a-MnBi powder and to the difficulty in compacting the powder into a fully dense and fully aligned bulk. The first difficulty originates from the peritectic reaction between Mn and Bi, which causes Mn to solidify and segregate from the Mn-Bi liquid during the solidification process. The kinetics of the segregated Mn reacting with the excessive Bi to form MnBi phase is slow at 628 K, making it difficult for the precipitated Mn and B to react and form MnBi. The second difficulty is caused by the eutectic reaction between Bi and MnBi at 535 K. Any bulk magnet fabrication process involving temperature higher than 535 K will cause phase decomposition leading to loss of magnetization. Clearly, for the development of MnBi based permanent magnet, the challenge is improving remanent magnetization while maintaining its coercivity above 8 kOe. This focus is different from most of the hard-magnetic materials, for which coercivity is the major concern.

  The effect of initial composition on the formation of a-MnBi content was investigated by Cui's group [107]. Fig. 6 summarizes their results. It shows that in order to maximize the a-MnBi content, one has to first maximize the amount of b-MnBi, which in turn requires excessive 4.3 at.% Mn. And consequently, the maximum amount of a-MnBi in the final product is limited to about 81 mol% (91 wt%). The obtained a-MnBi powders irreversibly gain weight when exposed to air and cycled between room temperature and 623 K. The weight gain is attributed to thermal decomposition ofa-MnBi to Mn and Bi, and the subsequent reaction of Mn with oxygen forming MnO.

  The sintering temperature for the fabrication of a-MnBi bulk magnet needs to be lower than the eutectic reaction temperature 535 K. At such low temperature, reaching over 99% density will require large stress. However, too large of a compaction stress will damage powder alignment causing remanent magnetization to decline. Currently, after alignment in a 1.8 T field followed by a gentle compaction with 10 MPa stress, the green density may reach 65% while the Mr/Ms ratio reaches 0.9. Subsequent cold-iso-press at 345 MPa and annealing at 500 K for 1 h will yield about 88% density with 6.6 kOe coercivity and 8.3 MGOe at room temperature. It is possible warm-iso-press at 500 K with large stress may further improve density to over 96% and maximum energy product over 10 MGOe.

  One approach to improve the energy product of a-MnBi is to take advantage of its large magnetocrystalline anisotropy and exchange couple the a-MnBi with a soft phase such as Fe, FeB, Co and FeCo. In an ideal situation, as simulated by Li's group using micro-magnetic finite element method and assuming a perfect ferromagnetic exchange coupling, the a-MnBi-Fe composite can exhibit a(BH)max as high as 40 MGOe [108]. Such impressive potential inspired intensive theoretical and experimental investigations.

  Zarkevich and Johnson used density functional theory to explore the interfacial energy differences between antiparallel (AFM) and parallel (FM) magnetic-moment alignments across the interface for MnBi/CoxFe1-xexchange-coupled layers [109e111]. They found that a-MnBi will ferromagnetically exchange couple with Co, but not with Fe. It is also possible for a-MnBi to ferromagneticallyexchange couple with Fe-Co alloy, but the amount of Fe cannot be more than 50 at.%. For the Co case, the maximum thickness of Co layer is only about 2 nm.

  Gao and Takeuchi [109] validated the prediction ofthe exchange coupling between MnBi and FeCo using thin film method. They deposited Bi and Mn layers sequentially on Si (100) substrates at room temperature bya high-vacuum magnetron sputtering system. Fig. 7 shows their results.

  MnBi thin films grow with strong columnar texture leading to robust out-of-plane magnetic anisotropy. By comparison, most soft magnetic thin films possess in-plane anisotropy due to shape anisotropy. The perpendicularlyexchange-coupled configuration of a-MnBi/CoxFe1-x bilayers as shown in Fig. 7a with Co at the interface leads to a significant energy-product enhancement. Fig. 7b shows typical room-temperature out-of-plane hysteresis loops of MnBi (20nm), MnBi (20nm)/Co (3 nm) bilayer, and MnBi (20nm)/ Co (5 nm) bilayer. The pure a-MnBi film shows strong perpendicular anisotropy with out-of-plane coercivity of 1.6 T and saturation magnetization of0.8T. Bycomparison, MnBi (20 nm)/Co (3 nm)and MnBi (20 nm)/Co (5 nm) bilayer films both exhibit reduced coercivity and enhanced remanence while maintaining a relatively square loop, indicating that an exchange coupling is established between the MnBi and Co films. A (BH)max of 25 MGOe was achieved with 3 nm of Co layer.

  In parallel to the thin film approach, researchers have been working on solution based nano-synthesis of MnBi-FeCo core-shell. Xuand Hong[112]used a magnetic self-assembly process to obtain exchange-coupled hard/soft a-MnBi/FeeCo core/shell structured composites. Their MnBi particles were obtained through milling of coarse powder, and the FeCo shell was coated later through chemicalsynthesis.Henkelplotsoftheobtaineda-MnBi/FeCocore-shell particles indicated exchange coupling dominates in a-MnBi/ FeCo (95/5 wt%) composites, and magnetostatic interaction dominates in a-MnBi/FeeCo (90/10wt%) composites, indicating the limitation in the amount of the soft phase can that can be coupled. Dai and Ren [113] also demonstrated the exchange coupling between a-MnBi and FeCo using nano-synthesis method. The difference is that the a-MnBi nanoparticle were obtained using metalredox method and FeCo nanowires were electro-spun. In both cases, the overall energy product of the composite was less than 6 MGOe, mainly due to the large amount of impurities that are difficult to remove after the solution based chemical synthesis.

  The a-MnBi based bulk magnet may be commercialized if its energy product exceeds 10 MGOe at room temperature. The current state of the art result is close to this target. The raw materials cost of MnBi is less than $15/kg as of January 2018. The approach to push the current 8.7 MGOe to 10 MGOe is through a novel synthesis approach that can yield 99% pure a-MnBi powder with <5 mm particle size and through a large scale warm-iso-press system that can densify the aligned compact to over 96% density. Exchange coupling ofMnBi with Co has an attractive outcome but cleanly and uniformly coating MnBi powder with just 3 nm of Co is a grand challenge.

Fig. 6. Phase fractions analysis by XRD (a) before and (b) after vacuum heat treatment [107].
Fig. 6. Phase fractions analysis by XRD (a) before and (b) after vacuum heat treatment [107].
Fig. 7. Large energy product enhancement in perpendicularly coupled MnBi/CoxFe1-x bilayers: (a) micromagnetic schematic of the bilayer spin structure in a moderately high reverse field. The arrows indicate the spin directions in the MnBi and CoxFe1-x layers. t0 is the critical thickness above which the magnetization of soft layer has a negative contribution to the total magnetization; (b) experimental hysteresis loops of a pure MnBi, MnBi/Co (3 nm) and of MnBi/Co (5 nm) bilayers, and simulated hysteresis of MnBi/Co (3 nm) bilayer (green line); (c) (BH)max of MnBi/CoxFe1-x bilayers for different soft-layer compositions and thicknesses; (d) experimental (BH)max as a function of Co concentration x of CoxFe1-x for bilayers with 3 nm of soft layer [109]. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 7. Large energy product enhancement in perpendicularly coupled MnBi/CoxFe1-x bilayers: (a) micromagnetic schematic of the bilayer spin structure in a moderately high reverse field. The arrows indicate the spin directions in the MnBi and CoxFe1-x layers. t0 is the critical thickness above which the magnetization of soft layer has a negative contribution to the total magnetization; (b) experimental hysteresis loops of a pure MnBi, MnBi/Co (3 nm) and of MnBi/Co (5 nm) bilayers, and simulated hysteresis of MnBi/Co (3 nm) bilayer (green line); (c) (BH)max of MnBi/CoxFe1-x bilayers for different soft-layer compositions and thicknesses; (d) experimental (BH)max as a function of Co concentration x of CoxFe1-x for bilayers with 3 nm of soft layer [109]. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

  4. Alnico type magnet systems

  This class of alloys contains Al, Ni, Co, thus its name. Although not as powerful as REE magnets, Alnico's remarkable temperature stability and good mechanical properties earned its small (4e5%) but steady market share along with ferrite, SmCo and NdFeB. The development of Alnico can be traced back to 1931, when Tokuhichi Mishima discovered the magnet steel containing nickel and aluminum [114]. It was the first modern permanent magnet, exhibiting 0.43 kOe of coercivity and 9.4 kG remnant magnetization. Improvements in energy density were made in the 1950s and 1960s by optimizing alloy design and processing [115,116]. Adding Co lowers the magnetization but improves coercivity (Alnico-2); with Co content nearly doubled (Alnico-5), theenergyproductwas improved by nearly 5 times. Alnico-8 was developed for high coercivity (1.7kOe). Addition ofTi, further increase ofCo, and the novel thermal magnetic treatment all contributed to this rather remarkable improvement in that period of time. After the development of the much more powerful SmCo5 magnet by Karl Strnat and his colleagues in 1966, interests in further development of Alnico magnets faded away. However, in 2014, based on a micro-magnetic model and TEM/APT results, Zhou et al. estimated the theoretical (BH)max of Alnico could reach 20 MGOe [117]. This prediction reignited interest in Alnico.

  Alnico coercivity is primarily due to shape anisotropy of nanometer size FeCo-rich a1 rods embedded in nonmagnetic AlNi-richa2 phase, rather than the magnetocrystalline anisotropy that is typical of rare earth based permanent magnets. While eclipsed by rare earth-based alloys due to their superior coercivity, Alnico has a number of attractive characteristics: high magnetization (AlniCo-5, 12.8 kG), low thermal coefficients (Br- 0.02%/K and H& —0.015%/K), in particular high Curie temperature (1073 K). The normal operating temperature for Alnico-8 can be as high as 800 K, at which about 90% ofroom temperature magnetization can still be retained. Depending on the compositions, the properties can vary greatly. Table 2 shows the compositions and magnetic properties ofvarious grades of Alnico alloys. Zhou et al. carried out atom probe tomography studies ofthese grades ofAlniCo [117,118]. The compositions of the FeCo-rich a1 phase and AlNi-rich a2 phase for Alnico 5e7, 8 and 9 are listed in Table 3 [117].

  Table 2

 

  Composition and magnetic properties of commercial Alnico alloys. Isotropic (unoriented) grades: Alnico 2,3,4; Anisotropic (oriented) grades: Alnico 5,6,8,9.

Table 2    Composition and magnetic properties of commercial Alnico alloys. Isotropic (unoriented) grades: Alnico 2,3,4; Anisotropic (oriented) grades: Alnico 5,6,8,9.

  Table 3

  Compositions of the FeCo-rich a1 phase and AlNi-rich a2 phase for AlNiCo 5e7, 8 and 9.
Table 3  Compositions of the FeCo-rich a1 phase and AlNi-rich a2 phase for AlNiCo 5e7, 8 and 9.

  Mechanical properties of Alnico are superior than most of the permanent magnets, except the bonded REE magnets. The tensile strength for Alnico, SmCo5, and NdFeB are 350, 41, and 82 MPa, respectively. And the fracture toughness (KIC) of these three magnetic materials are 13.3, 5.5, and 1.9MPam0.5, respectively[119].

  The first challenge for Alnico is low coercivity. The low coercivity originates from its low magnetocrystalline anisotropy (0.26-0.32 MJ/m3) and its reliance on the shape anisotropy mechanism. While this mechanism maintains stability over a wide range of temperature, the magnitude of coercivity is limited. Even with perfectly ordered FeCo precipitate rods perfected aligned and embedded in a non-magnetic matrix, the maximum coercivity is predicted to be less than7kOe[117, 120]. The second challenge forAlnico is its high raw materials cost. In 2017, its materials cost is comparable to that of the NdFeB based magnets. Cobalt, which is the most expensive component in Alnico, is needed for magnetization and coercivity. To minimize the cobalt content without adverse impact on magnetic properties would require an optimization of the current fabrication process and compositions to induce magnetocrystalline anisotropy. In addition to the raw materials cost, the high energy cost incurred by prolonged annealing time (3e6 days) at high temperatures, as well as the need for machining of final magnet shapes, all contribute to the high cost of the final part ($71/kg in 2016), which is even higher than that of the NdFeB-Dy magnet ($60/kg in 2016).

  Alnico-9 has the highest energy product of ~10.5 MGOe with Br at11.2kGand Hciat1.5kOe. TheAlnico-5-7alloyhasthehighestFe content thus giving it the highest Br (13.5 kG), but it also has the lowest Hci (0.74 kOe). The higher (BH)max of Alnico -5-7 and Alnico —9 is due to their grain alignment. The grain alignment is achieved by solidifying in molds with a large temperature gradient which biases the growth of the high temperature cubic phase along their <100> axes. Other grades are isotropic (random oriented grains) and are produced either by casting or sintering powders. In all Alnico alloys, the shape anisotropy arises from the spinodal decomposed phases (the FeCo-rich a1 and the AlNi-rich a2 phase). In Alnico-8 and 9 alloys, the growth of a1 is biased by the application of magnetic field while annealing near the spinodal onset temperature [121,122].

  The improvements in properties in the 1950s and 60s were based primarily on empirical studies. A number of minor alloying elements, primarily Cu, Ti and Nb, are added to the base alloy (8e13wt% Al, 13e28wt% Ni, 0e42wt% Co with the balance Fe), to promote columnar growth and enhance coercivity. Recent research has used modern characterization tools and modeling to understand the spinodal decomposition [123e126], with the goal of controlling the uniformity of shape of the spinodal and chemical segregation within the spinodal, all ofwhich affect coercivity. Fig. 8 shows a typical microstructure of an Alnico-8 alloy prepared at Ames Laboratory.

  These recent investigations demonstrated how closely coupled the chemistry and the processing are to optimizing the size, shape and distribution of the nanometer size magnetic a1 phase [123]. The evolution of the primary spinodal phases (a1 and a2) alone is not sufficient to explain the development of the coercivity in the Alnico —8 and 9 grades. The initial magnetic anneal only accounts for about half the coercivity. The lower temperature annealing, which is performed for a much longer time after the magnetic anneal (referred to as the draw cycle), accounts for the remaining coercivity [127]. During this draw cycle, the initial mosaic pattern of the faceted a1 phase undergoes a subtle change with little variation in the spacing. However, small Cu-rich clusters coarsen and coalesce to form extended rods along the length of the a1 phase. These Cu-rich rods then undergo a shearing as they transform from a bcc to an fcc structure and the rods now have a more ellipsoid shape in the transverse sections. In addition, a small Ni3Al phase forms in between the a1, a2 and the Cu-rich rods, further separating the magnetic a1 nanostructures. In the grains with their <100> within ~10° of the applied magnetic field direction, the a phases have aspect ratios in excess of 10:1 and their ends appear to be tapered. This geometry is nearly ideal based on micromagnetic calculations [124]. Furthermore, these calculations suggest that coercivity could be nearly doubled if the diameter of the a1 can be reduced to <20 nm.

  To address the concern on the high cost associated with the high Co content, Palasyuk developed the Rigid Band Approximation approach that assumes that structure ofelectron bands ofthe alloy remains not distorted (i.e., rigid) even after certain chemical modifications. With this approach, he developed the so-called “Colean” Alnico alloys, of which the magnetic properties are comparable to those Alnico grades with high Co contents [128]. One of alloy he developed was derived from the commercial cast Alnico grade 8H with composition near Fe30.5Co34Ni11.7Al14.3Ti7Cu2.5. The new alloy reduced the cobalt content by 42%, resulting a new composition comprising Fe37.7Co19.6Ni18.9Al14.3Ti7Cu2.5. The intrinsic coercive force of this new alloy is Hc = 1.48 kOe and remanent magnetic flux Br = 0.75 T.

  The current progress implies there is a potential pathway to produce a higher energy product, high temperature Alnico alloy. The magnet must have well aligned or single grains whose <100> axes are nearly parallel to the applied magnetic field. The time at the temperature for magnetic annealing must be only long enough to bias the spinodal growth but short enough to keep their diameters under 20 nm. The subsequent draw then must be finely tuned to allow for segregation of the Al, Cu, Co, Ni and Ti to evolve into the complex mosaic pattern with a coherent a1, a2 interface separated by the intervening Ni3Al phase forms in between the Cu-rich rods. With these requirements satisfied, the coercivity of the Alnico alloy could exceed 3 kOe and the (BH)max could exceed 20 MGOe at room temperature.

  5. Tetrataenite L10 FeNi

  L10-FeNi is one of the few non-REE materials that has the potential to reach the level of the (BH)max of RE permanent magnets. The tetragonal distortion associated with the chemical ordering gives rise to large crystalline anisotropy (1.1-1.3 MJ/m3). When coupled with large magnetic moment ofalternating layers ofFe and Ni atoms, L10-FeNi is expected to exhibit (BH)max of 56 MGOe at room temperature [129]. The interest in L10-FeNi phase can be traced backto 2010 when Wasson, from the Institute ofGeophysics and Planetary Physics at UCLA, reported his investigation of the composition and size relation of the iron meteorite collected from Northwest Africa [130]. Among the 53 samples, he found the high-Ni ataxite sample NWA 6259 (1.8 kg in mass) is magnetic enough to pick-up a large steel object. After a series of microstructural and magnetic characterization on a specimen cut from NWA 6259, Mubarok et al. concluded the magnetic properties of NWA 6259 is the result ofL10 phase in the meteorite. They estimated the uniaxial anisotropy of the L10-FeNi phase as 1.1e1.3MJ/m3 and the theoretical energy product (BH)max at room temperature is 56 MGOe. The abundance ofthe key constituent elements along with its huge theoretical (BH)max make FeNi one ofthe most interesting and most intensively studied magnetic materials.

Fig. 8. High-angle-annular-dark-field (HHADF) scanning transmission electron microscopy (STEM) image and corresponding energy dispersive X-ray spectroscopy elemental mappings show a mosaic microstructure in an alnico 8 alloy prepared in Ames Laboratory.

  Fig. 8. High-angle-annular-dark-field (HHADF) scanning transmission electron microscopy (STEM) image and corresponding energy dispersive X-ray spectroscopy elemental mappings show a mosaic microstructure in an alnico 8 alloy prepared in Ames Laboratory.

  The lattice structure for the L10-FeNi is a chemically ordered fct superstructure with alternating monatomic layers of Fe and Ni along the c-axis. The composition ratio of naturally formed L10-FeNi is Fe: 50.47±1.98at.% and Ni: 49.60 ± 1.49 at.%. The lattice parameters of L1°-FeNi were estimated as a = b = 3.582 A and c = 3.607 A (c/a = 1.007), and those of disordered FeNi were a = b = c = 3.603 A [131]. Tetrataenite has a Curie temperature of 823 K [132], and a density of 8.275 g/cm3. The saturation magnetization Ms is estimated to be 14.7 kG, anisotropy field Ha = 14.4 kOe, and anisotropy constant K = 0.84 MJ/m3 [133]. This anisotropy constant is smaller than the original 1.3 MJ/m3 value derived by Pauleve in 1968 based on neutron diffraction and refinement [134].

  The formation of L10-FeNi phase is difficult due to the sluggish diffusion at the order-disorder transition temperature 593 K [135]. Bulk L10-FeNi is only found naturally in meteorites that have cooled over billions of years at extremely low cooling rates. Astronomers believe these iron-meteorites were once part of metal mass situated 50e200 km below the surface of asteroid-like body, which allow them to experience an extraordinarily slow cooling rates of 1e5Kpermillionyearsforthetemperatureintervalof973 to623 K to fosters long-range L10 chemical ordering [136]. Scientists were able to artificially grow L10-FeNi film [137] [138], and used the biasing stress from substrate to study the magnetism of L10-FeNi [139]; however, bulk synthesis has not yet been achieved in any laboratory to date.

  Since the atomic diffusion of Fe and Ni is extremely slow at the order-disorder transition temperature near 593 K, the formation of L10-FeNi phase is difficult using conventional annealing technique. Researchers have developed various techniques to promote the diffusion and the formation of L1°-FeNi. A study by Neel et al. in 1964 showed that L10-FeNi could be obtained by irradiating it with neutrons below 593 K [140]. Electron irradiation was also applied to accelerate the diffusion and induce the formation ofL10-FeNi [141]. Geng et al. attempted to enhance the diffusion rate by creating excessive vacancies via high-energy mechanical alloying, but the method was not very successful [142]. Shima et al. reported the preparation of L10-FeNi films by an alternate monatomic layer deposition technique [137]. The films were prepared by repeating 50 alternating Fe (001 ) and Ni (001 ) layers on MgO (001 ) substrates at various temperatures in the range between 353 and 673 K. They reported the degree of long-range order reached a maximum value of0.6±0.2 based on the ratio ofthe intensity ofthe fundamental and superlattice peaks on x-ray diffraction patterns, and the uniaxial magnetocrystalline anisotropy of 0.6 MJ/m3 at 513 K.

  Lima and Drago explored a novel low temperature cyclic oxiereduction process to produce L10-FeNi at low temperature [143]. In their experiments, the nickel capped iron grains were subjected to a sequence of oxidation and reduction at 623 K by changing the fluxing atmosphere in tube furnace. The oxidizing atmosphere consisted of a mixture of 90%/10% Ar/O2 during the oxidation step and was followed by a reducing atmosphere of pure hydrogen. They reported 19% of L10-FeNi ordered phase was obtained. Lee et al. reported the formation of L10-FeNi by severe plastic deformation using high-pressure torsion [144]. Severe plastic deformation could introduce a high density of lattice defects to the material, which may significantly enhance the atomic diffusion. Makino et al. demonstrated a new idea that L10-FeNi-based magnets could be realized by crystalizing an amorphous alloy based on ~ Fe50Ni50 with a crystallization temperature close to the order-disorder transition temperature [145]. They annealed amorphous Fe42Ni41.3Si8B4P4Cu0.7 ribbon at 673 K for 288 h, and obtained the ordered L10-FeNi phase along with a-Fe and Fe3B phases. The volume fraction ofL10-FeNi phase was low, resultingin low saturation magnetization Ms ~100 emu/g and coercivity of 0.7kOe.

  Goto et al. took another novel approach to synthesize L10-FeNi phase [146]. In this nitrogen insertion and topotactic extraction method, FeNiN, which has the same ordered arrangement as L10-FeNi, was formed by nitriding A1-FeNi powder with ammonia gas. Subsequently, FeNiN is denitrided by topotactic reaction to derive single-phase L10-FeNi. The method results in an artificially made L10-FeNi phase with an order parameter of 0.71, the highest reported value thus far. The transformation of disordered-phase FeNi into the L10 phase increased the coercive force from 0.2kOe to 1.8kOe.

  Based on the assumption that interfacial strain may induce the tetragonal distortion of FeNi, Gong et al. took the template assisted epitaxial core/shell synthesis approach [147]. They used the phase transformation of AuCu cores to induce a surface stress that effectively triggers a tetragonal reconstruction of FeNi shell within the critical thickness range. This strategy could be further extended to induce tetragonal distortion of various potential hard magnetic materials (FeCo, FeMn and MnNi) by using different templates [148e151].

  Among all the non-REE magnetic materials, L10-FeNi is the only one that has the potential to rival the energy product of RE permanent magnets without concerning on supply and cost of raw materials. Several bulk methods and solution-based synthesis approaches have produced definite and positive results, although noneofthemcanproduceabulkmaterialwithmorethan19wt%of the L10-FeNi phase. With intensive efforts, it is possible to significantly improve the purity of the feedstock. However, feedstock purity is only the first part ofthe problem, the ultimate challenge is to produce a highly textured fully dense bulk magnet at a temperature lower than the order-disorder transition temperature at 593K. This challenge is similar to that ofthe MnBi alloy.

  6. L10-FeCo

  L10-FeCo appears to possess more desirable magnetic properties than L10-FeNi because of its theoretical giant uniaxial magnetocrystalline anisotropy (10 MJ/m3) and saturation magnetization (2.3e2.4 mB/atom) [152]. The tetragonal distortion in FeCo alloy is responsible for the potentially giant magnetic anisotropy energy and large saturation magnetic moment (Fig. 9.) However, the high cobalt content (~51 wt%) raises a concern on the viability of such material to replace REE based magnet. Despite the limited knowledge on the thermodynamics of the phase formation, there are efforts on the synthesis of L10-FeCo. Most of the efforts took the template approach, using the epitaxial substrate to induce tetra-gonality of the grown FeCo layers. Metallic substrates such as Pd (001),Ir(001) and Rh(111) caninducetetragonaldistortioninFeCo due to the lattice mismatch, resulting in FeCo films with a strong perpendicular anisotropy, which is dependent on the film composition and thickness [153,154]. Anderson et al. reported the experimental realization of tetragonal Fe-Co alloys as a constituent of Fe0.36Co0.64/Pt superlattices with huge perpendicular magnetocrystalline anisotropy energy, reaching 210 meV/atom, and a saturation magnetization of 2.5 mB/atom at 40 K, in qualitative agreement with theoretical predictions [155]. Moulas et al. re-portedamaximummagneticanisotropyenergyof0.5 meV/atom in 1-ML thick film of FeCo on Pt(111), which is close to that observed for the L10-FePt [156]. The addition ofthird elements (e.g. C, B) to FeCo was found to help stabilize the tetragonal distortion and maintain the large anisotropyofthe FeCo layer [157e159]. Gao et al. reported a large magnetization (900 emu/cm3) and enhanced perpendicular coercive fields (2e3 kOe) oflowWconcentrationFe-Co-W films, but they concluded that the coercivity was mostly due totheshapeanisotropyofplatelet-likegrains[160].Hasegawaetal. obtained nanopatterned FeCo(Al) alloy thin films with a large coercivity of 6kOe [161]. In addition to thin films, chemically derived L10-AuCu was also applied as the template, which induced the tetragonality of the FeCo shell, giving a coercivity of 0.85 kOe [151]. Shen et al. utilized the phase transformation of Au nanowire template to trigger the tetragonal distortion ofFeCo shell [148]. To date, no bulk L10-FeCo phase has been successfully synthesized in any laboratory.

  7. HfCo and ZrCo based systems

  Co-rich intermetallic compounds such as HfCo7 and Zr2Co11 were studied as the hard phase for exchange-coupled nanocomposites [162e170]. Their attractive magnetic properties include large magnetocrystalline anisotropy constant K1, large saturation magnetization Ms, and high Curie temperature Tc. For HfCo7, these properties are K = 1.4 MJ/m3, Ms = 11.8 kG, and Tc = 751 K; and for Z「2Co11, they are K -1.35 MJ/m3, Ms - 9.7 kG, and Tc - 783 K [163]. While the HfCo7 and Zr2Co11 compounds have been known for morethantwodecades[171e173],thestabilityandcontrolofphase purityofthese compounds have been always a challenge. HfCo7isa metastable phase and can be formed only at high temperature of about 1323e1403 K for a single composition of 12.5 at.% of Hf as shown in the Co-Hf phase diagram (Fig. 10a) [174].

Fig. 9. a) An ordered FeCo alloy with the CsCl structure. The tetragonal distortion in the lattice leads to the change of c/a ratio. b) A superlattice consisting of monolayers of Fe-Co and Pt [155]. c) Calculated uniaxial magnetic anisotropy energy Ku (upper panel) and saturation magnetic moment Ms (lower panel) of tetragonal Fe1-xCox as a function of the c/a ratio and the Co concentration x [152].

  Fig. 9. a) An ordered FeCo alloy with the CsCl structure. The tetragonal distortion in the lattice leads to the change of c/a ratio. b) A superlattice consisting of monolayers of Fe-Co and Pt [155]. c) Calculated uniaxial magnetic anisotropy energy Ku (upper panel) and saturation magnetic moment Ms (lower panel) of tetragonal Fe1-xCox as a function of the c/a ratio and the Co concentration x [152].

  According to the Zr-Co phase diagram, Zr2Co11 seems to be a stable phase for the temperature range of573e1527 K at 15.4 at.% of Zr (Fig. 10b) [175]. However, the rhombohedral Zr2Co11 with a space group R32 (a = 4.69 A and c = 24.0 A as schematically shown in (Fig. 10c) is the hard-magnetic structure that is found to be a high-temperature metastable phase [170,176,177]. A calculation of its free energy also suggests that the rhombohedral Zr2Co11 structure is stable only at above 1198 K [176]. At lower temperatures, the rhombohedral structure transforms into a comparatively soft-magnetic orthorhombic Zr2Co11 phase along with hexagonal Co and cubic Zr6Co23 [170,177]. Similarly, the melt-spun HfCo7 is shown to form an orthorhombic structure (a = 4.72 A, b = 4.28 A, and c = 8.07 A as shown in (Fig. 10d) [178]. For the compositions close to HfCo7, a formation of a new phase Hf2Co11 is also reported along with Hf6Co23 and a-Co for the bulk alloys prepared by arcmelting and subsequent annealing at above 973 K [179]. Interestingly, the Hf2Co11 phase also crystallizes into an orthorhombic structure with a = 4.7 A, b = 38 A, and c = 8.3 A and exhibits magnetic properties similar to those of HfCo7. In another study, the melt-spun Hf-Co alloys with a range of composition HfCo6+d (0.6 < d < 1) is shown to form orthorhombic structure with a = 4.7 A, b = 8.2 A, and c = 38.6 A along with Co and Hf2Co? upon annealing at 1313 K [180].

Fig. 10. Bulkphasediagrams:(a)Co-H[174]and(b)Co-Zr[175].Crystalstructures:(c)rhombohedralZr2Co11[176]andorthorhombicHfCo7[178].TheXRDpatternsofnanoparticles (NPs) and melt-spun bulk alloys: (e) HfCo7 (Reproduced with permission from Ref. [162]. and Copyright (2012), American Institute of Physics) and (f) Zr2Co11 (Reproduced with permission from Ref. [168]. Copyright (2013 WILEY-VCH Verlag GmbH & Co.).

  Fig. 10. Bulkphasediagrams:(a)Co-H[174]and(b)Co-Zr[175].Crystalstructures:(c)rhombohedralZr2Co11[176]andorthorhombicHfCo7[178].TheXRDpatternsofnanoparticles (NPs) and melt-spun bulk alloys: (e) HfCo7 (Reproduced with permission from Ref. [162]. and Copyright (2012), American Institute of Physics) and (f) Zr2Co11 (Reproduced with permission from Ref. [168]. Copyright (2013 WILEY-VCH Verlag GmbH & Co.).

  These results indicate that rapid quenching is essential to stabilize the rhombohedral Zr2Co11 and orthorhombic HfCo7, following the compound formation and crystallization at high temperatures. These metastable structures are stabilized in nanoparticles produced by gas-aggregation cluster deposition [162e164,168] and melt-spun nanocrystalline bulk alloys [163,167,170,177]. For example, x-ray diffraction patterns of HfCo7 and Zr2Co11 produced by the non-equilibrium cluster-deposition and melt-spinning methods are indexed to the orthorhombic and rhombohedral structures shown in Fig. 10c and d, respectively. In addition, the cluster-deposition method also provides a good control over phase purity, easy-axis alignment, and nanostructuring.

  First-principle calculations show a uniaxial anisotropy for the rhombohedral Zr2Con structure with K1 = 1.4 MJ/m3 and Ms = 10.3 kG [176]. The calculations indicate an easy-plane anisotropy for the orthorhombic HfCo7 structure with an effective anisotropy constant Kef = 3.5 MJ/m3 and Ms = 11.4 kG [178,181]. The theoretical results are in a similar range and reasonable agreement with the experimental data obtained for melt-spun Zr2Con (K1 = 1.4MJ/m3 and Ms = 9.7 kG) and HfCo7 (K1 = 1.4 MJ/ m3 and Ms - 11.8kG) [163].

  Table 4

  The measured Hc, Js, and (BH)max at 300 K and Tc for bulk Hf-Co and Zr-Co alloys.

Table 4    The measured Hc, Js, and (BH)max at 300 K and Tc for bulk Hf-Co and Zr-Co alloys.

  Furthermore, as shown in Table 4, the underlying magnetic anisotropies of these structures are clearly reflected by appreciable coercivities observed for various Zr-Co and Hf-Co alloys, in which the hard-magnetic phase is claimed to be rhombohedral Zr2Co11 and orthorhombic HfCo7 or Hf2Co11, respectively.

  The obtained bulk Hf-Co and Zr-Co alloys exhibit lower energy products as compared to the theoretical maximum energy products (BH)max for HfCo7 (32.5 MGOe) and Zr2Co11 (26 MGOe). This is mainly due to the presence of secondary phases and additive elements that significantly reduce the magnetization and randomly oriented grains that limit the remanent magnetization Mr to only ~50% of saturation magnetization Ms. Since the crystallization temperatures for bulk HfCo7 and Zr2Co11 are well above their corresponding Curie temperatures, the easy-axis alignment process for achieving high Mr/Ms ratio using magnetic field during the crystal growth is hindered.

Fig. 11. Exchange-coupled nanocomposites: (a) Deposition process and (b) EDS (energy-dispersive x-ray spectroscopy) color maps showing Hf, Co, and Fe distribution on a HAADF image of an Hf-Co:Fe-Co film [Adapted from Ref. [164]]. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

  Fig. 11. Exchange-coupled nanocomposites: (a) Deposition process and (b) EDS (energy-dispersive x-ray spectroscopy) color maps showing Hf, Co, and Fe distribution on a HAADF image of an Hf-Co:Fe-Co film [Adapted from Ref. [164]]. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

  Using the cluster-deposition method, Zr2Co11 and HfCo7 nanoparticles form the desired high-anisotropy structures during the gas-aggregation process in the cluster chamber, without the requirement of a subsequent high-temperature thermal annealing as shown in Fig. 11a [162,168]. This direct ordering substantially improves the phase purity and alsoprovidesunique opportunity to align the easy axes of nanoparticles using a magnetic field of about 5 kOe, prior to deposition on substrates [162]. As compared to the melt-spun bulk alloys, the aligned nanoparticles show improved room-temperature magnetic properties for Z^Con (He = 4.4 kOe, Ms - 10.2 kG, and M』Ms - 0.88) and HfCo? (h - 8.7 kOe, Ms - 10.9 kG, and Mr/Ms - 0.82) [164,168].

  More importantly, these aligned high-anisotropy nanoparticles are also co-deposited with soft magnetic Fe65Co35 phase using a single-step process in the cluster-deposition system as shown in Fig. 11a[164]. The result is an aligned exchange-coupled nanocomposite film in which highly anisotropic nanoparticles are uniformly coated and surrounded with the Fe-Co phase. The high-angle annular dark-field (HAADF) image, as shown in Fig. 11b, depicts the size and spatial arrangement of the hard Hf-Co phase and the soft Fe-Co phase.

  The magnetization of the soft phase is perfectly exchange-coupled with the aligned hard magnetic nanoparticles, as clearly shown in the room-temperature hysteresis loops measured along the easy- and hard-axis directions for a Hf-Co:Fe-Co nanocomposite film having 7 vol.% of Fe-Co (Fig. 12a). This Hf-Co:Fe-Co nanostructure exhibits high Mr/Ms = 0.90, Hc = 10.1 kOe, and Ms = 11.3 kG along the easy axis, which subsequently translates into a room-temperature energy product (BH)max ofabout 20.3 MGOeas shown in Fig. 12b[164]. Similarly, the cluster-assembled Zr-Co:Fe-Co exchange-couple nanocomposite films show a room temperature(BH)maxofabout19.5MGOefor15vol.%ofFe-Co[168].

  A permanent-magnet material should maintain high coercivities Hc and high remanent magnetization Mr at elevated temperatures, typically up to 453 K for high-performance motor applications. As shown in Fig. 12c, appreciable values ofHc and Mr are observed for the nanoparticles ofZ^Con (Hc = 4.2—1.7 kOe and Mr - 9.6—7.9 kG) and HfCoy (Hc - 8.7T.5 kOe and Mr - 9.2—9.0 kG) in the temperature range of 300—453 [183]. This performance is helpful to retain a significant amount of energy product at elevated temperatures for the cluster-assembled exchange-coupled nano-compositefilms.Fig.12dshowsthetemperature-dependentenergy products for various permanent-magnet materials [184]. Interestingly, the Hf-Co:Fe-Co nanocomposite film having 7 vol.% of Fe-Co exhibits a (BH)max of 17.1 MGOe at 453 K, which is comparable to that of SmCo5 and superior than other rare-earth-free materials such as CoPt, Alnico, and MnAlC.

  It is worthwhile to develop alternative chemical methods for fabricating Hf-Co and Zr-Co nanoparticles for bulk magnetic application. Wet-chemical methods have been successful in producing several permanent-magnet nanoparticles in large quantities [185]. Such method has potential to produce nanoparticles in large quantities.

Fig. 12. Aligned Hf-Co:Fe-Co exchange-coupled nanocomposite film having 7 vol.% of Fe-Co: (a) Easy- and hard-axis hysteresis loops and (b) B and BH curves measured at 300 K (Adapted from Ref. [164]). (c) Hc and Jr (i.e.Mr) for HfCo7 (solid circles) and Zr2Co11 (hollow circles) nanoparticles at elevated temperatures (Adapted from Refs. [164] and [183]). (d) The temperature-dependent energy products of the Hf-Co: Fe-Co nanocomposite film are compared with other permanent-magnet materials (Adapted from Ref. [184]).

  Fig. 12. Aligned Hf-Co:Fe-Co exchange-coupled nanocomposite film having 7 vol.% of Fe-Co: (a) Easy- and hard-axis hysteresis loops and (b) B and BH curves measured at 300 K (Adapted from Ref. [164]). (c) Hc and Jr (i.e.Mr) for HfCo7 (solid circles) and Zr2Co11 (hollow circles) nanoparticles at elevated temperatures (Adapted from Refs. [164] and [183]). (d) The temperature-dependent energy products of the Hf-Co: Fe-Co nanocomposite film are compared with other permanent-magnet materials (Adapted from Ref. [184]).

Fig. 13. (a) Saturation magnetization (Ms) and coercivity (Hc) for the cobalt carbide nanoparticles as a function the ratio between the volume fraction of Co2C and Co3C phases. The solid and dotted lines are guide to the eye (Adapted from Ref. [186]). (b) The experimental x-ray diffraction pattern (black curve) is fitted using Rietveld analysis for the orthorhombic structure with a Pnma space group (red curve) (Adapted from Ref. [190]). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

  Fig. 13. (a) Saturation magnetization (Ms) and coercivity (Hc) for the cobalt carbide nanoparticles as a function the ratio between the volume fraction of Co2C and Co3C phases. The solid and dotted lines are guide to the eye (Adapted from Ref. [186]). (b) The experimental x-ray diffraction pattern (black curve) is fitted using Rietveld analysis for the orthorhombic structure with a Pnma space group (red curve) (Adapted from Ref. [190]). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

  The films approach successfully demonstrated the high (BH)max as well as temperature stability of the exchange-coupled nanocomposite (20.3 MGOe at room temperature and 17.1 MGOe at 453 K). Future efforts should focus on large scale production of nanoparticles, accurate controlling of powder easy-axis alignment, and maintaining the nanostructure that is essential for exchangecoupling during the densification process.

  8. Carbides

  The synthesis ofCo3C and Co2C nanoparticles with decent roomtemperature coercivity of 3.4 kOe using wet-chemical approach attracted significant interest [186e189]. This method uses a polyol reduction chemistry that offers high yields of nanoparticles, and also provides an appreciable control over size, shape, composition, and structure [186]. While the Co3C phase crystallizes into an orthorhombic structure (Pnma, a = 5.03 A, b = 6.73 A, and c = 4.48 A), the Co2C also forms an orthorhombic structure (Pnnm, a = 4.45 A, b = 4.37 A, and c = 2.90A) [186]. The cobalt carbide nanoparticles often have a mixture of Co3C and Co2C phases [186,187] and their magnetic properties such as coercivity and saturation magnetization stronglydepend on the ratio between the volume fraction of Co2C and Co3C ratio as shown in Fig. 13a[186].

  An increase of coercivity upon increasing the volume fraction of Co3C in Fig. 13a indicates that the Co3C phase possesses higher magnetocrystalline anisotropy than the Co2C phase. For example, single-phase Co3C nanoparticles with an average particle size of about 8.1 nm have shown an improved anisotropy constant of 0.75 ± 0.1 MJ/m3 and a high blocking temperature of 571 K, above which superparamagnetic material loses its preferred direction of magnetization in zero magnetic fields [189]. This result was also supported by the first-principle calculations. The study claims that the cobalt carbide nanoparticles consist of cobalt layers separated by carbon atoms and the separation between the occupied and unoccupied d-states is reduced due to the partial mixing of the p-states of carbon with Co d-states which leads to large magnetocrystalline anisotropy, high blocking temperature, and high coercivity in the Co3C nanoparticles [189]. It is also worth noting that cobalt carbide phases decompose above 700 K, and this temperature is higher than the required operation temperature (玉453 K) for the most of the permanent-magnet applications.

  Fe3C also forms the orthorhombic structure (Pnma, a = 5.027 A, b = 6.713 A, and 4.477 A) [190]. Although the structure of Fe?C is similar to that of the Co3C phase, first-principle calculations shows substantially low magnetic anisotropy constant for the Fe3C structure (0.05 MJ/m3). The low anisotropy is also reflected by the experimental results of Fe3C nanoparticles that exhibit a coercivity of only about 0.36 kOe [191]. Similarly, a nanostructured Fe3C material with a grain size of 5.2 nm also exhibit a substantially low blocking temperature Tb = 20 K [192], and the equation KiVnp = 25 kBT yields an anisotropy constant K1 z 0.09 MJ/m3, where Vnp and kB are the particle volume and Boltzmann constant, respectively.
 

  Table 5

  Structural and magnetic properties of ternary Fe3-xCoxC alloys (Adapted from Ref.190).

Table 5 Structural and magnetic properties of ternary Fe3-xCoxC alloys (Adapted from Ref.190).

  Surprisingly, the substitution of a Co atom for Fe in Fe3C has shown an improved anisotropy constant in nanoparticles (K1 z 4.6 MJ/m3)[193] and melt-spun bulk alloys (K1 z0.96MJ/ m3) [190]. The Fe2CoC phase forms the structure similar to that of the Fe3C as shown in the x-ray diffraction pattern of the melt-spun Fe2CoC (Fig. 13b). The experimental x-ray diffraction pattern of the melt-spun Fe2CoC shows a good agreement with the simulated x-ray diffraction pattern using the orthorhombic structure with a space group Pnma and a = 5.055 A, b = 6.756 A, and 4.505 A [190]. In addition, computational simulations using an adaptive genetic algorithm search and densityfunctional theoryalso have identified several Fe3_xCoxC phases (0.5 < x < 2) with appreciable magnetic properties as shown in Table 5 [190].

  9. Iron nitride

  Iron nitride attracts considerable interest because of its exceptionally high magnetization and because its elements are the most earth abundant among all magnetic materials. Iron nitrides can also offer versatile magnetic properties originated from various crystal structures with different nitrogen concentrations in the lattice [194]. Among them, metastable a00-Fe16N2 with an ordered bodycentered tetragonal structure has been regarded as a promising candidate of non-REE magnets. Due to the tetragonal distortion induced by interstitial N atoms in the lattice, a00-Fe16N2 exhibits considerable magnetocrystalline anisotropy (1.0 MJ/m3) and large saturation magnetization (>20 kG).

  The a00-Fe16N2 was initially discovered as an intermediate phase during the tempering of nitrogen-martensite, when the FeN (a0) phase decomposes to Fe4N(g0)[195]. Kim and Takahashi synthesized the compound in thin film form and reported a high saturation magnetization ofabout 2200 emu/cm [3], which is larger than the value of Fe-Co alloy [196]. This giant saturation magnetization drew great attention from the magnetic materials community. However, it had been extremely difficult to reproduce these results for manyyears, until researchers at Hitachi inJapan reported a close value obtained using single-crystal a00-Fe16N2 films grown by molecular beam epitaxy method [197,198]. Nonetheless, both excitement and controversy ensued.

  The crystal structure of metastable a00-Fe16N2 was first described by Jack in 1951 as a body-centered tetragonal symmetry (space group I4/mmm) based on the a-Fe (bcc) structure, with N atoms occupying octahedral interstices [195]. The lattice parameters were found as a = 5.72 A and c = 6.29 A (c/a = 1.1). The occupation of interstitial sites with nitrogen induces the tetragonal distortion and result in three Fe positions: (0,1/2,1/4), (0, 0, 0.31), and (1/4, 1/4, 0) with site symmetries 4d (4 m2), 4e (4 mm), and 8 h (mm), respectively [194,195]. According to the established stoichiometry, a nitrogen concentration of 11.1 at.% is necessary to obtain a00-Fe16N2.

  The reason a00-Fe16N2 attracted great interest from magnetic community is mainly due to its potentially giant saturation magnetization, first reported in 1972 [196]. However, due to the difficulty in synthesizing a00-Fe16N2, these impressive results on magnetic properties of a00-Fe16N2 are hard to reproduce. The reported values appeared to be inconsistent and sometimes contradictory [199]. The saturation magnetizations reported on films, foils and powder are highly scattered, from 230 to 315 emu/g [196,199e202]. The magnetocrystalline anisotropy constant was estimated to be 1.0MJ/m3 [202,203]. Curie temperature was estimated to be around 813 K by using the Langevin function [197]. The theoretical energy product is estimated to be comparable to those state-of-the-art RE permanent magnets (45 MGOe) [202,204]. A recentreporton a00-Fe16N2 foils showing a magnetic energy product ofupto20MGOeatroomtemperatureprovedthepossibilityofa00-Fe16N2 as gap magnets [204]. Despite of the controversy these results raised, the magnetic materials community general agreed that a00-Fe16N2 is a semi-hard material that has potentials to be used as gap magnet [205,206].

  One disadvantage of the nitrides is their moderately low decomposition temperature. This greatly limits their usage in motor applications where the working temperature may be as high as 453 K. The decomposition of a00-Fe16N2 creates undesirable magnetic phases such as a-Fe and g0-Fe4N. Studies on the decomposition a00-Fe16N2 reported very different values (473e673 K) [197,207], which were indirectly measured by assuming the large decrease of magnetization is caused by the decomposition of a00-Fe16N2. Impurities could be responsible for these inconsistent results. A quantitative study on the thermal stability of high-purity a00-Fe16N2 nanoparticles revealed that more than 90% of a00-Fe16N2 could decompose into a-Fe and g0-Fe4N at 473 K under inert gaseous conditions [207]. At an elevated temperature of513 K, the decomposition time became even less than 1 h. A relatively low working temperature of 355 K was suggested for a00-Fe16N2 devices in order to maintain their performance better than 99% for a long service life. Perhaps more importantly, the thermal stability issue makes it extremely difficult to use the traditional sintering or warm compaction methods to fabricate bulk magnets.

  Thin films ofa00-Fe16N2 have been fabricated by several methods including sputtering [202,208], nitrogen ion implantation [209,210] and molecular beam epitaxy [197]. The thin film research focused more on the saturation magnetization of a00-Fe16N2 than its anisotropy. Wang's group presented a systematic study on magnetism of a00-Fe16N2 thin films which were fabricated using a unique low-energy and plasma-free sputtering process [202]. The sputtered thin films exhibited a saturation magnetization up to 26.8 kG, which was explained by a first-principle calculation based on LDA + U method. The magnetocrystalline anisotropy (1 MJ/m3) and energy product (135 MGOe) were also estimated in their study [202,204].

  Bulk a00-Fe16N2 can be synthesized through ammonia nitridation. According to Jack's original method [195], bulk a00-Fe16N2 can be obtained by rapidly quenching nitrogen austenite to form nitrogen martensite, followed by tempering at low temperature. However, fabricationofbulka00-Fe16N2 magnet with a large enough (BH)max remains a grand challenge. Direct nitridation of pure bulk iron was reported to fabricate wire-shaped a00-Fe16N2 magnet with a(BH)maxofupto9 MGOe[211].Afree-standinga00-Fe16N2 foilwith a(BH)max of up to 20 MGOe was fabricated by a nitrogen ionimplantation technique, albeit the volume fraction of a00-Fe16N2 in the foil is only 35% [204]. In this approach, the implantation flux was increased from 2 x 1016 to 1 x 1017/cm2 to optimize the coercivity and saturation magnetization to 1.9 kOe and 245 emu/g, respectively, which results in a (BH)max of 20 MGOe (Fig. 14).

  A chemical synthesis approach based on the reduction and low-temperature nitridation ofoxide fine particles showed the potential in obtaining large quantities of high-purity a00-Fe16N2 powder. In this approach, the nitridation condition and particle size play important roles in the purity of a00-Fe16N2 [203,212,213]. Several non-magnetic oxides such as SiO2 [214] and Al2O3 [215]were coated on the particles to prevent sintering during the process. The moisture and oxygen contents must be carefully controlled atalow level throughout the fabrication process and storage in order to prevent oxidation. A good example presented by Ogawa et al. demonstrated that single-phase a00-Fe16N2 powder was synthesized in gram quantity by reducing several Fe-oxide particles (e.g. a-Fe2O3, g-Fe2O3, Fe3O4) followed with subsequent ammonia nitridation [203]. Interestingly, the powder exhibited saturation magnetization of 234 emu/g at 5 K and a magnetocrystalline anisotropy constant of 0.96 MJ/cm3, which are comparable to the values of the compound in thin film form [199]. High-purity a00-Fe16N2 particles with 2.3 MGOe were obtained recently via spray drying method [216]. Nitridation of pure iron was also possible whenballmillingmethodwasutilized[217]. The a”-Fe16N2 powder prepared by ball milling was compacted into a dense disk shape by shock compaction using a gas gun, and a00-Fe16N2 remained stable under shock compaction [217]. However, the saturation magnetization and coercivity of the feedstock powder were reduced after the shock compaction from 201 emu/g and 0.85 kOe to 175 emu/g and 0.55 kOe. The authors attributed this degradation of the magnetic properties to oxidization. A high-pressure low-temperature sintering technique was developed to suppress the phase decomposition. The method maintained most of the saturation magnetization ofthe feedstock powder (177 emu/g) while consolidating the powder into 77% green density [218]. However, the coercivity sharply decreased from 2.3 kOe to 1.3 kOe. Such decrease was attributed to the loss of magnetic shape anisotropy and the magnetic interaction among powders.

Fig. 14. a) In-plane hysteresis loops for the samples with different fluences at room temperature. b) X-ray diffraction pattern of the a”-Fei6N2 foil.

  Fig. 14. a) In-plane hysteresis loops for the samples with different fluences at room temperature. b) X-ray diffraction pattern of the a”-Fei6N2 foil.

  Despite the continuous debate on the saturation magnetization of a00-Fe16N2 and the challenges in achieving the desired magnetic properties in bulk form, the a”-Fe16N2 compound remains as an attractive material for permanent magnet applications due to its cheap price and near unlimited supply of raw materials. The coercivity of a00-Fe16N2 is low, especially in the bulk form. The intrinsically poor thermal instability of this compound prevents the utilization of traditional techniques to improve its microstructure and coercivity. The success in preparing high-purity a00-Fe16N2 powder in large quantities offers a promising approach to fabricate bonded magnets. However, more efforts are needed to improve coercivity and thermal stability.

  10. Summary and future challenges

  The challenge for non-rare earth based permanent magnet are that most of them do not simultaneously exhibit high magnetization and high coercivity. High coercivity can be achieved by either the materials intrinsic high magnetocrystalline anisotropy, fine particle/grain size, or shape anisotropy. High remanent magnetization requires high saturation magnetization and high degree of alignment of the grains along magnetic easy axis. Challenges differ, depending on the materials system.

  For MnAl(C), the highest bulk (BH)max ~ 7.5 MGOe has been realized only at relatively low Hci ~ 3 kOe and imperfect texture (Br ~ 6 kG). A much higher Hci of 10.7 kOe was reported for MnAl thin films, whereas the saturation magnetization of the MnAl(-C) alloys allows for a Br up to 8.2 kG and, therefore, for a (BH)max up to 16.8 MGOe. The metastable nature of the MnAl compound has been the major obstacle to obtaining simultaneously high degree of texture and high Hci through the standard manufacturing methods like the powder metallurgy. The forty-year-long stalemate may arguably be broken either by providing thermodynamic stability without losing too much of the intrinsic magnetic properties or by designing an unorthodox processing specifically tuned to the material.

  For MnBi, the current state of the art room temperature bulk properties are (BH)max ~8.7 MGOe, Hci ~ 6.5 kOe, and Br ~ 5.8 kG, which are far less than the thick film results ((BH)max ~16.3 MGOe, Hci ~ 19.5 kOe, and Br ~ 8.3 kG) demonstrated by Zhang et al. [103]; or the powder results ((BH)max ~11.9 MGOe, Hci ~ 13.1 kOe, and Br ~ 7.1 kG) by Cui et al. [219]. The challenges for MnBi based permanent magnet are originated from its low decomposition temperature at 535 K. Such low temperature makes it difficult to obtain fine particles via ball milling process. Study shows even using low energy ball milling, 8 h of milling will cause over 10% of the powders to decompose [220]. A possible solution is to anneal the decomposed powder at 560 K after ball-milling, which may encourage the decomposed powder to form a-MnBi again. Sintering is typically the last step of bulk magnet fabrication designed to achieve nearly-full density. But for MnBi, the upper limit of the sintering temperature is only 530 K. Such low temperature can only marginally improve the density of the green compact, by less than 1 g/cm [3]. One possible solution is to increase the pressure of warm compaction or the pressure of cold-iso-press prior to the last sintering step. In addition to the low remanent magnetization caused by the powder decomposition, the bulk magnet fabrication also faces the challenge of significant reduction in coercivity (from >12 kOe to <6 kOe) once the green density exceeding 75%. Such reduction may attribute to the proximity effect, for which closely packed large and small particles are magnetically coupled. Once the magnetic moment of the large particle is reversed by a small external field, that of the small particle immediately follow, causing early demagnetization for the whole magnet body. Possible approach is either making sure all particles are single crystal and sufficiently small (e.g. < 1 mm), or coating powders with a nonmagnetic layer to ensure all particles are magnetically de-coupled.

  For Alnico, the commercial grade with the highest performance is Alnico-9 ((BH)max ~10.5 MGOe, Hci ~ 1.5 kOe, and Br~ 11.2 kG), while the theoretical performance is ((BH)max ~20.7 MGOe, Hci ~ 3.1 kOe, and Br ~ 13.5 kG). Hci will be increased by reducing the rod diameters to less than 20 nm while maintaining a more ellipsoid shape and reducing 'shorting' between rods [124]. Squareness of the loop is controlled by alignment of the gains so increasing texture will be key to increasing (BH)max of the Alnico 8 class alloys. With the a1 rod diameter already at ~50 nm, the challenge becomes how to control the size and the shape of the a1 phase during the spinodal decomposition occurred during the magnetic annealing and the subsequent low temperature annealing process. In addition to the challenge on magnetic properties, the recent surge of cobalt price due to the large demand by lithium-ion battery created a deep concern on the price of the Alnico magnet. One possible approach is through composition optimization, where cobalt content can be reduced to as low as 15% [128].

  For L10-FeNi and L10-FeCo, the thermodynamic pathway for the phase formation is either extremely sluggish or non-exist, making it much more difficult to overcome with any conventional metallurgical approach. L10-FeNi has the potential to rival the energy product of RE magnet without concern on supply and cost of the raw materials. The tetragonal distortion associated with the chemical ordering gives rise to the large magnetocrystalline anisotropy. When coupled with large magnetic moment of alternating layers of Fe and Ni atoms, L10-FeNi is expected to exhibit a (BH)max of 56 MGOe at room temperature [129]. However, the formation of L10-FeNi phase is difficult due to the sluggish diffusion of Fe and Ni atoms at the order-disorder transition temperature 593 K. Scientists were able to artificially grow L10-FeNi film [137,138] and used the biasing stress from substrate to study the magnetism of L10-FeNi [139]. Bulk synthesis has not yet been achieved in any laboratory to date. L10-FeCo appears to possess more desirable magnetic properties than L10-FeNi because of its theoretically calculated giant uniaxial magnetocrystalline anisotropy (10 MJ/m3) and saturation magnetization (2.3e2.4 mB/atom) [152]. Unfornately, the tetragonal distortion of FeCo cubic phase is thermodynamically unfavored. It appears that novel synthesis methods involving either templating or far-from-equilibrium solid state processing are possible approaches to tackle these challenges. Adding third or fourth elements to distort the FeCo lattice is also a possible approach.

  Aligned Hf-Co and Zr-Co nanoparticle films exhibit appreciable room-temperature permanent-magnet properties: HfCo7 (Hci = 5.4kOe; Br = 9.1 kG, and (BH)max = 13.2 MGOe) and Z^Con (Hd - 4.1 kG, Br - 9.4 kG, and (BH)max = 16.6 MGOe). When the nanoparticles are combined with a high-magnetization Fe-Co phase to form exchange-coupled nanostructured films, the roomtemperature (BH)max improve to about 20 MGOe [164,168]. As compared to nanoparticles, the melt-spun Hf-Co and Zr-Co alloys have shown only moderate permanent-magnet properties: Hf2ConB (Hci = 3.6 kOe; Br - 6.2 kG, (BH)max = 6.7 MGOe) [167] and Z^Con (Hci - 2.2 kOe; Br - 6.3 kG, (BH)max = 5.2 MGOe) [163]. The phase purity and random orientation of hard-magnetic grains are main issues, which significantly reduce Hci, Br and (BH)max for Hf-Co and Zr-Co based bulk alloys. Similarly, it is essential to obtain single phase high-anisotropy Co3C nanoparticles and subsequently align their easy axes to achieve higher energy products than the currently reported room-temperature value (2.6 MGOe) for cobalt carbides [186]. In addition, forbulkmagneticapplications, scale-up methods must be developed to produce nanoparticles in bulk quantities and create nanostructures similar to those obtained in the case of Hf-Co and Zr-Co based nanocomposite films [164,168]. As compared to Mn- and Fe-based permanent-magnet materials, the higher material cost is an important issue for using Hf-Co, Zr-Co, and Co-C as bulk magnets. However, these materials can have potential uses in the form of thin films for microelectromechanical systems (MEMS), data storage, and spintronics applications.

  For a00-Fe16N2, the challenge is on how to achieve coercivity higher than 2 kOe in bulk form and on how to avoid phase decomposition. Bulk magnet fabricating with reasonable magnetic properties is still an obstacle that has not been overcome. The highest performance was reported on free-standing foils with a coercivity up to 1.9 kOe and a (BH)max up to 20 MGOe at room temperature [204]. Currently, the most promising approach to fabricate bulk a00-Fe16N2 permanent magnets is nitridation followed by post annealing. Similar to other non-RE magnets, there is a conflict between the lower temperature for a00-Fe16N2 phase stability and the higher temperature desired by densification. Possible approach is to couple low-temperature annealing with other solid state processes to achieve the desired microstructure. In addition to the bulk magnet fabrication, synthesis of a00-Fe16N2 with high purity is also a challenge. Nitridation of iron powder is a relatively easy way to obtain a00-Fe16N2 with high purity. The coercivity ofa00-Fe16N2 powder was reported as high as 2.2 kOe [215], which could be further improved by texturing.Jiang et al. suggested nitrogen ion implantation as an alternative nitriding approach to fabricate a00-Fe16N2 in bulk form. The maximum magnetic energy product was pushed up to 20 MGOe by controlling the ion implantation condition and the two-step post-annealing process [204]. Applying external stress and magnetic field are also a promising approach since they may induce tetragonal distortion and assist the martensitic (a00) phase formation [211].

  Among these candidates non-RE magnets, MnAl, MnBi, and a00-Fe16N2 are more developed than the others. Interestingly, these three materials all share similar challenges on purity of the feedstock powder and thermal stabilityofthe desired phase during bulk magnet fabrication process.

  Acknowledgements

  The research on MnBi was primarily supported by the U.S. Department of Energy, Advanced Research Projects Agency-Energy under Award Number 11/CJ000/09/03 and Energy Efficiency and Renewable Energy under Award Number DE-EE0007794. The Alnico work at Ames Laboratory was supported by the Department of Energy-Energy Efficiency and Renewable Energy, Vehicles Technology Office, PEEM program, under Contract No. DE-AC02-07CH11358 for the operation of Ames Laboratory (USDOE). The research at Nebraska was primarily supported by the U.S. Department of Energy, Office of Basic Energy Sciences under Award DE-FG02-04ER46152 and performed in part at the Nebraska Nanoscale Facility, Nebraska Center for Materials and Nanoscience, which is supported by the NSF under Award NNCI: 1542182, and the Nebraska Research Initiative (NRI).

  References

  [1] S. Liu, G.E. Kuhl, Development of New High Temperature and High Performance Permanent Magnet Materials. No. UDR-tr-2000-00092, DAYTON UNIV OH RESEARCH INST, 2000.

  [2] Z.W. Liu, R.V. Ramanujan, H.A. Davies, Improved thermal stability of hard magnetic properties in rapidly solidified REeTMeB alloys, J. Mater. Res. 23 (10) (2008) 2733e2742.

  [3] Diana Bauer, David Diamond, Jennifer Li, David Sandalow, Telleen Paul, Brent Wanner, US Department of Energy Critical Materials Strategy, 2010.

  [4] USGS Mineral Commodities Summaries, 2017, 2014, 2010.

  [5] The 2018 REE oxides prices were quoted from mineralprices.com on 5/9/ 2018.

  [6] P. Nothnagel, K.-H. Muller, D. Eckert, A. Handstein, The influence of particle size on the coercivity of sintered NdFeB magnets, J. Magn. Magn Mater. 101 (1e3) (1991) 379e381.

  [7] Sagawa, Masato. Sintered NdFeB magnet and method for manufacturing the same. U.S. Patent 9,589,714, issued March 7, 2017.

  [8] Analysts, Global Industry. Magnetic Materials, a Global Strategic Business Report, 2010.

  [9] T. Ohtani, N. Kato, S. Kojima, K. Kojima, Y. Sakamoto, I. Konno, M. Tsukahara, T. Kubo, Magnetic properties of Mn-Al-C permanent magnet alloys, IEEE Trans. Magn. 13 (5) (1977) 1328e1330.

  [10] Hiroshi Kono, On the ferromagnetic phase in manganese-aluminum system,

  J. Phys. Soc. Jpn. 13 (12) (1958) 1444e1451.

  [11] A.J.J. Koch, P. Hokkeling, M.G. vd Steeg, K.J. De Vos, New material for permanent magnets on a base of Mn and Al, J. Appl. Phys. 31 (5) (1960) S75eS77.

  [12] Adarsh Shukla, Arthur D. Pelton, Thermodynamic assessment of the Al-Mn and Mg-Al-Mn systems, J. Phase Equilibria Diffusion 30 (1) (2009) 28e39.

  [13] Akimasa Sakuma, Electronic structure and magnetocrystalline anisotropy energy of MnAl, J. Phys. Soc. Jpn. 63 (4) (1994) 1422e1428.

  [14] J.H. Park, Y.K. Hong, S. Bae, J.J. Lee, J. Jalli, G.S. Abo, N. Neveu, S.G. Kim,

  C.J. Choi, J.G. Lee, Saturation magnetization and crystalline anisotropy calculations for MnAl permanent magnet, J. Appl. Phys. 107 (9) (2010), 09A731.

  [15] Priyanka Manchanda, Arti Kashyap, Jeffrey E. Shield, L.H. Lewis, Ralph Skomski, Magnetic properties of Fe-doped MnAl, J. Magn. Magn Mater. 365 (2014) 88e92.

  [16] Alexander Edstr€om, Jonathan Chico, Jakobsson Adam, Anders Bergman, Rusz Jan, Electronic structure and magnetic properties of L 1 0 binary alloys, Phys. Rev. B 90 (1) (2014), 014402.

  [17] Kanika Anand, J.J. Pulikkotil, S. Auluck, Study of ferromagnetic instability in t-MnAl, using first-principles, J. Alloy. Comp. 601 (2014) 234e237.

  [18] Poul Bernard Braun, Jacob Adriaan Goedkoop, An X-ray and neutron diffraction investigation of the magnetic phase Al0. 89Mn1. 11, Acta Crys-tallogr. 16 (8) (1963) 737e740.

  [19] Ying-chang Yang, Wen-wang Ho, Chin Lin, Ji-lian Yang, Hui-ming Zhou, Jia-xuan Zhu, Xiang-xin Zeng, Bai-shen Zhang, Lan Jin, Neutron diffraction study of hard magnetic alloy MnAlC, J. Appl. Phys. 55 (6) (1984) 2053e2054.

  [20] L. Pareti, F. Bolzoni, F. Leccabue, A.E. Ermakov, Magnetic anisotropy of MnAl and MnAlC permanent magnet materials, J. Appl. Phys. 59 (11) (1986) 3824e3828.

  [21] M.A. Bohlmann, J.C. Koo, J.H. Wise, Mn-Al-C for permanent magnets, J. Appl. Phys. 52 (3) (1981) 2542e2543.

  [22] R. McCurrie, J. Rickman, P. Dunk, D. Hawkridge, Dependence of the permanent magnet properties of Mn 55 Al 45 on particle size, IEEE Trans. Magn. 14 (5) (1978) 682e684.

  [23] H. Zijlstra, Coercivity and wall motion, IEEE Trans. Magn. 6 (2) (1970) 179e181.

  [24] J.P. Jakubovics, T.W. Jolly, The effect of crystal defects on the domain structure of Mn-Al alloys, Phys. B+C 86 (1977) 1357—1359.

  [25] Simon Bance, Florian Bittner, Thomas G. Woodcock, Ludwig Schultz, Thomas Schrefl, Role of twin and anti-phase defects in MnAl permanent magnets, Acta Mater. 131 (2017) 48—56.

  [26] F. Bittner, J. Freudenberger, L. Schultz, T.G. Woodcock, The impact of dislocations on coercivity in L10-MnAl, J. Alloy. Comp. 704 (2017) 528—536.

  [27] J. Thielsch, F. Bittner, T.G. Woodcock, Magnetization reversal processes in hot-extruded t-MnAl-C, J. Magn. Magn Mater. 426 (2017) 25e31.

  [28] J.D. Livingston, A review of coercivity mechanisms, J. Appl. Phys. 52 (3) (1981) 2544e2548.

  [29] Ch Muller, H.H. Stadelmaier, B. Reinsch, G. Petzow, Metallurgy of the magnetic t-phase in Mn-Al and Mn-Al-C, Z. Metallkd. 87 (7) (1996) 594e597.

  [30] H. Zijlstra, H.B. Haanstra, Evidence by Lorentz microscopy for magnetically active stacking faults in MnAl alloy, J. Appl. Phys. 37 (7) (1966) 2853e2856.

  [31] Michael J. Lucis, Timothy E. Prost, Xiujuan Jiang, Meiyu Wang, Jeffrey E. Shield, Phase transitions in mechanically milled Mn-Al-C permanent magnets, Metals 4(2)(2014) 130e140.

  [32] Hong Jian, Konstantin P. Skokov, Oliver Gutfleisch, Microstructure and magnetic properties of MneAleC alloy powders prepared by ball milling, J. Alloy. Comp. 622 (2015) 524e528.

  [33] Phase transformation of Mn-Al alloy powders, Scripta Metall. Mater. 28 (1) (1993) 133e138.

  [34] F. Bittner, L. Schultz, T.G. Woodcock, The role of the interface distribution in the decomposition of metastable L10-Mn54Al46, J. Alloy. Comp. 727 (2017) 1095e1099.

  [35] Jung-Goo Lee, Xiao-Lei Wang, Zhi-Dong Zhang, ChuI-Jin Choi, Effect of mechanical milling and heat treatment on the structure and magnetic properties of gas atomized MneAl alloy powders, Thin Solid Films 519 (23) (2011) 8312e8316.

  [36] Luke G. Marshall, Ian J. McDonald, L.H. Lewis, Quantification of the strain-induced promotion of t-MnAl via cryogenic milling, J. Magn. Magn Mater. 404 (2016) 215e220.

  [37] J. Rial, M. Villanueva, E. Cespedes, N. Lopez, J. Camarero, L.G. Marshall,

  L. H. Lewis, A. Bollero, Application of a novel flash-milling procedure for coercivity development in nanocrystalline MnAl permanent magnet powders, J. Phys. Appl. Phys. 50 (10) (2017) 105004.

  [38] Vintaiki Ez, va udovenko, Is belyatsk, Nn luarsabi, sy makushev, Formation of ferromagnetic tau-phase in manganese-aluminum alloys, Fizika Metallov I Metallovedenie 38 (2) (1974) 398e409.

  [39] Sea Kojima, T. Ohtani, N. Kato, K. Kojima, Y. Sakamoto, I. Konno,

  M. Tsukahara, T. Kubo, Crystal transformation and orientation of Mn-Al-C hard magnetic alloy, AIP Conf. Proc. 24 (1) (1975) 768e769.

  [40] J.J. Van Den Broek, H. Donkersloot, G. Van Tendeloo, J. Van Landuyt, Phase transformations in pure and carbon-doped Al45Mn55 alloys, Acta Metall. l (27) (1979) 1497e1504.

  [41] D.P. Hoydick, E.J. Palmiere, W.A. Soffa, On the formation of the metastable L1 {sub o} phase in manganese-aluminum-base permanent magnet materials, Scripta Materialia 36 (2) (1997).

  [42] Alla S. Sologubenko, Peter Mullner, Helge Heinrich, Gernot Kostorz, On the plate-like t-phase formation in MnAleC alloys: dedicated to Professor Dr. Peter Neumann on the occasion of his 65th birthday, Z. Metallkd 95 (6) (2004) 486e491.

  [43] Ping-Zhan Si, Hui-Dong Qian, Chul-Jin Choi, Jihoon Park, Sangho Han, Hong-Liang Ge, Kiran P. Shinde, In situ observation of phase transformation in MnAl (C) magnetic materials, Materials 10 (9) (2017) 1016.

  [44] A. Morisako, M. Matsumoto, M. Naoe, Synthesis of ferromagnetic t phase of Mn-Al films by sputtering, J. Appl. Phys. 61 (8) (1987) 4281e4283.

  [45] G.A. Fischer, M.L. Rudee, Effect ofmagnetic annealing on the t-phase ofMnAl thin films, J. Magn. Magn Mater. 213 (3) (2000) 335e339.

  [46] Z.C. Yan, Y. Huang, Y. Zhang, G.C. Hadjipanayis, W. Soffa, D. Weller, Magnetic and structural properties of MnAl/Ag granular thin films with L10 structure, Scripta Materialia 53 (4) (2005) 463e468.

  [47] T. Sands, J.P. Harbison, M.L. Leadbeater, S.J. Allen Jr., G.W. Hull, R. Ramesh, V.G. Keramidas, Epitaxial ferromagnetic t-MnAl films on GaAs, Appl. Phys. Lett. 57 (24) (1990) 2609e2611.

  [48] W. Van Roy, J. De Boeck, H. Bender, C. Bruynseraede, A. Van Esch, Gustaaf Borghs, Structural and magnetic investigations of epitaxial ferromagnetic t MnAl films grown on GaAs/AlAs by molecular-beam epitaxy, J. Appl. Phys. 78(1) (1995) 398e404.

  [49] T.M. Rosier, Y.W. He, N.A. El-Masry, Ferromagnetic t-MnAl epitaxially grown on (100) GaAs substrates by pulsed laser deposition, Mater. Lett. 26 (4e5) (1996) 227e231.

  [50] Y. Kinemuchi, A. Fujita, K. Ozaki, High-pressure synthesis of L1 0 MnAl with

  near-stoichiometric composition, Dalton Trans. 45 (27) (2016)

  10936e10941.

  [51] G.R. Stafford, B. Grushko, R.D. McMichael, The electrodeposition of Al-Mn ferromagnetic phase from molten salt electrolyte, J. Alloy. Comp. 200 (1e2) (1993) 107e113.

  [52] Tetsuji Saito, Magnetic properties of MneAleC alloy powders produced by mechanical grinding, J. Appl. Phys. 97 (10) (2005), 10F304.

  [53] Y.J. Kim, J.H. Perepezko, Formation of a metastable ferromagnetic t phase during containerless melt processing and rapid quenching in Mn-Al-C alloys, J. Appl. Phys. 71 (2) (1992) 676e680.

  [54] Zhuyin Shao, Hui Zhao, Jiling Zeng, Yinfeng Zhang, Wenyun Yang, Youfang Lai, Shuai Guo, et al., One step preparation of pure t -MnAl phase with high magnetization using strip casting method, AIP Adv. 7 (5) (2017) 056213.

  [55] Hailiang Fang, Sofia Kontos, Angstro€m Jonas, Johan Cedervall, Svedlindh Peter, Klas Gunnarsson, Sahlberg Martin, Directly obtained t-phase MnAl, a high performance magnetic material for permanent magnets, J. Solid State Chem. 237 (2016) 300e306.

  [56] W. Dreizler, A. Menth, Transformation kinetics of the ferromagnetic alloy Mn-Al-C, IEEE Trans. Magn. 16 (3) (1980) 534e536.

  [57] J.H. Huang, P.C. Kuo, Influence ofcarbon on the phase transformation kinetics and magnetic properties of Mn-Al alloys, Mater. Sci. Eng., B 22 (2e3) (1994) 256e260.

  [58] N.I. Vlasova, G.S. Kandaurova, YAS Shur, N.N. Bvkhanova, Magnetic properties and crystal structure of Mn-Ai alloys, Phys. Met. Metallogr. 51 (4e6) (1981) 207.

  [59] H.X.Wang,P.Z.Si,W.Jiang,J.G.Lee,C.J.Choi,J.J.Liu,Q.Wu,M.Zhong,H.L.Ge, Structural stabilizing effect of Zn substitution on MnAl and its magnetic properties, Open J. Microphys. 1 (02) (2011) 19.

  [60] T. Mix, F. Bittner, K.-H. Muller, L. Schultz, T.G. Woodcock, Alloying with a few atomic percent of Ga makes MnAl thermodynamically stable, Acta Mater. 128 (2017) 160e165.

  [61] S.H. Nie, L.J. Zhu, J. Lu, D. Pan, H.L. Wang, X.Z. Yu, J.X. Xiao, J.H. Zhao, Perpendicularly magnetized t-MnAl (001) thin films epitaxied on GaAs, Appl. Phys. Lett. 102 (15) (2013)152405.

  [62] C.T. Lee, K.H. Han, I.H. Kook, Woong Kil Choo, Phase and lattice parameter relationships in rapidly solidified and heat-treated (Mn 0.53 Al 0.47) 100— x C x pseudo-binary alloys, J. Mater. Res. 7 (7) (1992) 1690e1695.

  [63] Yunlong Geng, Michael J. Lucis, Pamela Rasmussen, Jeffrey E. Shield, Phase transformation and magnetic properties of rapidly solidified Mn-Al-C alloys modified with Zr, J. Appl. Phys. 118(3) (2015) 033905.

  [64] I. Janotova, P. Svec Sr., P. Svec, I. Mat'ko, D. Janickovic, J. Zigo, M. Mihalkovic, J. Marcin, I. Skorvanek, Phase analysis and structure of rapidly quenched Al-Mn systems, J. Alloy. Comp. 707 (2017) 137e141.

  [65] Wei Lu, Junchao Niu, Taolei Wang, Kada Xia, Zhen Xiang, Yiming Song, Hong Zhang, Satoru Yoshimura, Hitoshi Saito, Low-energy mechanically milled t-phase MnAl alloys with high coercivity and magnetization, J. Alloy. Comp. 675 (2016) 163e167.

  [66] Z.W. Liu, C. Chen, Z.G. Zheng, B.H. Tan, Raju V. Ramanujan, Phase transitions and hard magnetic properties for rapidly solidified MnAl alloys doped with C, B, and rare earth elements, J. Mater. Sci. 47 (5) (2012) 2333e2338.

  [67] E. Fazakas, L.K. Varga, F. Mazaleyrat, Preparation of nanocrystalline MneAleC magnets by melt spinning and subsequent heat treatments, J. Alloy. Comp. 434 (2007) 611e613.

  [68] Anurag Chaturvedi, Rumana Yaqub, Ian Baker, A comparison of t-MnAl particulates produced via different routes, J. Phys. Condens. Matter 26 (6) (2014) 064201.

  [69] D.C. Crew, P.G. McCormick, R. Street, MnAl and MnAlC permanent magnets produced by mechanical alloying, Scripta Metall. Mater. 32 (3) (1995).

  [70] Tetsuji Saito, Magnetic properties of MneAl system alloys produced by mechanical alloying, J. Appl. Phys. 93 (10) (2003) 8686e8688.

  [71] O. Obi, L. Burns, Y. Chen, T. Fitchorov, S. Kim, K. Hsu, D. Heiman, L.H. Lewis, V.G. Harris, Magnetic and structural properties of heat-treated high-moment mechanically alloyed MnAlC powders, J. Alloy. Comp. 582 (2014) 598e602.

  [72] N. Makino, Y. Kimura, M. Suzuki, On the effect of pulverizing and swaging on the magnetic properties of MnAl compound, J. Jpn. Inst. Metals 28 (1964) 483e489.

  [73] Tetsuji Saito, Magnetic properties of MneAleC alloy powders produced by mechanical grinding, J. Appl. Phys. 97 (10) (2005), 10F304.

  [74] P. Saravanan, Jen-Hwa Hsu, V.T.P. Vinod, Miroslav Cernfk, S.V. Kamat, Coercivity enhancement in Mn-Al-Cu flakes produced by surfactant-assisted milling, Appl. Phys. Lett. 107 (19) (2015) 192407.

  [75] P. Saravanan, Jen-Hwa Hsu, V.T.P. Vinod, Miroslav Cernfk, S.V. Kamat, MWCNT reinforced t-Mn-Al nanocomposite magnets through spark plasma sintering, J. Alloy. Comp. 695 (2017) 364e371.

  [76] Q. Zeng, I. Baker, J.B. Cui, Z.C. Yan, Structural and magnetic properties of nanostructured MneAleC magnetic materials, J. Magn. Magn Mater. 308 (2) (2007) 214e226.

  [77] Qi Zeng, Ian Baker, Zhi-cheng Yan, Nanostructured MneAl permanent magnets produced by mechanical milling, J. Appl. Phys. 99 (8) (2006), 08E902.

  [78] J.Y. Law, J. Rial, M. Villanueva, N. Lopez, J. Camarero, L.G. Marshall, J.S. Blazquez, et al., Study of phases evolution in high-coercive MnAl powders obtained through short milling time of gas-atomized particles, J. Alloy. Comp. 712 (2017) 373e378.

  [79] K.P. Su, J. Wang, H.O. Wang, D.X. Huo, L.W. Li, Y.Q. Cao, Z.W. Liu, Strain-induced coercivity enhancement in Mn51Al46C3 flakes prepared by surfactant-assisted ball milling, J. Alloy. Comp. 640 (2015) 114e117.

  [80] Jung-Goo Lee, Li Pu, Chul-Jin Choi, Xing-Long Dong, Synthesis of MneAl alloy nanoparticles by plasma arc discharge, Thin Solid Films 519 (1) (2010) 81e85.

  [81] A.E. Berkowitz, J.D. Livingston, J.L. Walter, Properties of Mn-Al-C magnets prepared by spark erosion and other rapid solidification techniques, J. Appl. Phys. 55 (6) (1984) 2106e2108.

  [82] A.M. Gabay, G.C. Hadjipanayis, Application of mechanochemical synthesis to manufacturing of permanent magnets, JOM (J. Occup. Med.) 67 (6) (2015) 1329e1335.

  [83] P. Saravanan, V.T.P. Vinod, Miroslav Cernfk, A. Selvapriya, Dibyendu Chakravarty, S.V. Kamat, Processing of MneAl nanostructured magnets by spark plasma sintering and subsequent rapid thermal annealing, J. Magn. Magn Mater. 374 (2015) 427e432.

  [84] Alexandre Pasko, Martino Lobue, E. Fazakas, L.K. Varga, Frederic Mazaleyrat, Spark plasma sintering of MneAleC hard magnets, J. Phys. Condens. Matter 26 (6) (2014) 064203.

  [85] Rajasekhar Madugundo, Ozlem Koylu-Alkan, C. George, Hadjipanayis. Bulk Mn-Al-C permanent magnets prepared by various techniques, AIP Adv. 6 (5) (2016) 056009.

  [86] Anurag Chaturvedi, Rumana Yaqub, Ian Baker, Microstructure and magnetic properties of bulk nanocrystalline MnAl, Metals 4 (1) (2014) 20—27.

  [87] Y. Sakamoto, S. Kojima, K. Kojima, T. Ohtani, T. Kubo, Crystal orientation and its formation process of an anisotropic Mn-Al-C permanent magnet, J. Appl. Phys. 50 (B3) (1979) 2355—2357.

  [88] M.A. Bohlmann, Saturation magnetization of swaged Mn-Al. Proceedings of the Seventh Conference on Magnetism and Magnetic Materials, Springer, Boston, MA, 1962, pp. 1315—1316.

  [89] K. Kamino, T. Kawaguchi, M. Nagakura, Magnetic properties of MnAl system alloys, IEEE Trans. Magn. 2 (3) (1966) 506—510.

  [90] A. Yanagitani, A. Okawa, Y. Tanaka, Characteristics of MnAlC rapidly solidified powders, Mater. Sci. Eng., A 181 (1994) 1355—1358.

  [91] Tu Chen, W. Stutius, The phase transformation and physical properties of the MnBi and Mn 1.08 Bi compounds, IEEE Trans. Magn. 10 (3) (1974) 581—586.

  [92] H. Okamoto, Bi-mn Phase Diagram, ASM International, Materials Park, OH, 1990.

  [93] K. Cenzual, L.M. Gelato, M. Penzo, E. Parthe, Inorganic structure types with revised space groups, Acta Crystallogr. B 47 (4) (1991) 433—439.

  [94] Jun Cui, Jung-Pyung Choi, Guosheng Li, Evgueni Polikarpov, J. Darsell, Matthew J. Kramer, Nikolai A. Zarkevich, et al., Development of MnBi permanent magnet: neutron diffraction of MnBi powder, J. Appl. Phys. 115 (17) (2014) 17A743.

  [95] Y.B. Yang, X.G. Chen, S. Guo, A.R. Yan, Q.Z. Huang, M.M. Wu, D.F. Chen, Y.C. Yang, J.B. Yang, Temperature dependences of structure and coercivity for melt-spun MnBi compound, J. Magn. Magn Mater. 330 (2013) 106—110.

  [96] Michael A. McGuire, Huibo Cao, Bryan C. Chakoumakos, Brian C. Sales, Symmetry-lowering lattice distortion at the spin reorientation in MnBi single crystals, Phys. Rev. B 90 (17) (2014) 174425.

  [97] Nikolai A. Zarkevich, L.-L. Wang, Duane D. Johnson, Anomalous magneto-structural behavior of MnBi explained: a path towards an improved permanent magnet, Apl. Mater. 2 (3) (2014) 032103.

  [98] B.W. Roberts, Neutron diffraction study of the structures and magnetic properties of manganese bismuthide, Phys. Rev. 104 (3) (1956) 607.

  [99] A.F. Andresen, W. Ha€lg, P. Fischer, E. Stoll, The magnetic and crystallographic properties of MnBi studied by neutron diffraction, Acta Chem. Scand. 21 (6) (1967).

  [100] X. Guo, X. Chen, Z. Altounian, J.O. Str€om-Olsen, Magnetic properties of MnBi prepared by rapid solidification, Phys. Rev. B 46 (22) (1992) 14578.

  [101] H. St€ablein, Hard ferrites and plastoferrites, Handb. Ferromagn. Mater. 3 (1982) 441—602.

  [102] J.B. Yang, William B. Yelon, William Joseph James, Qingsheng Cai, M. Kornecki, Samit Roy, Naushad Ali, Ph l'Heritier, Crystal structure, magnetic properties and electronic structure of the MnBi intermetallic compound, J. Phys. Condens. Matter 14 (25) (2002) 6509.

  [103] Wenyong Zhang, Parashu Kharel, Valloppilly Shah, Lanping Yue, J. David, Sellmyer. High-energy-product MnBi films with controllable anisotropy, Physica Status Solidi (B) 252 (9) (2015) 1934—1939.

  [104] Narayan Poudyal, Xubo Liu, Wei Wang, V. Vuong Nguyen, Yilong Ma, Kinjal Gandha, Kevin Elkins, et al., Processing of MnBi bulk magnets with enhanced energy product, AIP Adv. 6 (5) (2016) 056004.

  [105] Sumin Kim, Hongjae Moon, Hwaebong Jung, Su-Min Kim, Hyun-Sook Lee, Haein Choi-Yim, Wooyoung Lee, Magnetic properties of large-scaled MnBi bulk magnets, J. Alloy. Comp. 708 (2017) 1245—1249.

  [106] Charles Guillaud, Ferromagnetisme des alliages binaires de manganese, PhD diss, Gauthier-Villars, 1943.

  [107] Jun Cui, Jung-Pyung Choi, Evgueni Polikarpov, Mark E. Bowden, Wei Xie, Guosheng Li, Zimin Nie, Nikolai Zarkevich, Matthew J. Kramer, Duane Johnson, Effect of composition and heat treatment on MnBi magnetic materials, Acta Mater. 79 (2014) 374—381.

  [108] Y.Q. Li, M. Yue, J.H. Zuo, D.T. Zhang, W.Q. Liu, J.X. Zhang, Z.H. Guo, W. Li, Investigation of magnetic properties of MnBi-Fe nanocomposite permanent magnets by micro-magnetic simulation, IEEE Trans. Magn. 49 (7) (2013) 3391—3393.

  [109] T.R. Gao, L. Fang, S. Fackler, S. Maruyama, X.H. Zhang, L.L. Wang, T. Rana, et al., Large energy product enhancement in perpendicularly coupled MnBi/ CoFe magnetic bilayers, Phys. Rev. B 94 (6) (2016) 060411.

  [110] John P. Perdew, John A. Chevary, Sy H. Vosko, Koblar A. Jackson, Mark R. Pederson, Dig J. Singh, Carlos Fiolhais, Atoms, molecules, solids, and surfaces: applications of the generalized gradient approximation for exchange and correlation, Phys. Rev. B 46 (11) (1992) 6671.

  [111] Georg Kresse, Jurgen Hafner, Ab initio molecular-dynamics simulation of the liquid-metal—amorphous-semiconductor transition in germanium, Phys. Rev. B 49 (20) (1994) 14251.

  [112] Xia Xu, Yang-Ki Hong, Jihoon Park, Woncheol Lee, Alan M. Lane, Jun Cui, Magnetic self-assembly for the synthesis of magnetically exchange coupled MnBi/Fe—Co composites, J. Solid State Chem. 231 (2015) 108—113.

  [113] Qilin Dai, John Q. Xiao, Shenqiang Ren, Solution processed MnBi-FeCo magnetic nanocomposites, Nano Res. 9 (11) (2016) 3222—3228.

  [114] Tokushichi, Mishima. Magnet steel containing nickel and aluminium. U.S. Patent 2,027,994, issued January 14, 1936.

  [115] Robert B. Campbell, Carl A. Julien, Structure of alnico V, J. Appl. Phys. 32 (3) (1961) S192—S194.

  [116] C.A. Julien, F.G. Jones, Alpha-Sub-gamma phase in alnico 8 alloys, J. Appl. Phys. 36 (3) (1965) 1173—1174.

  [117] Lin Zhou, Michael K. Miller, Ping Lu, Liqin Ke, R. Skomski, H. Dillon, Q. Xing, et al., Architecture and magnetism of alnico, Acta Mater. 74 (2014) 224—233.

  [118] Lin Zhou, Wei Tang, Liqin Ke, Wei Guo, Jonathan D. Poplawsky, Iver

  E. Anderson, Matthew J. Kramer, Microstructural and magnetic property evolution with different heat-treatment conditions in an alnico alloy, Acta Mater. 133 (2017) 73—80.

  [119] J.A. Horton, J.L. Wright, J.W. Herchenroeder, Fracture toughness of commercial magnets, IEEE Trans. Magn. 32 (5) (1996) 4374—4376.

  [120] B.D. Cullity, Introduction to Magnetic Materials, Addison-Wesley, Reading, MA, 1972. Google Scholar (1994).

  [121] John W. Cahn, On spinodal decomposition in cubic crystals, Acta Metal-lurgica 10 (3) (1962) 179—183.

  [122] Yoshiro Iwama, Masaharu Takeuchi, Spinodal decomposition in Alnico 8 magnet alloy, Trans. Jpn. Inst. Met. 15 (5) (1974) 371—377.

  [123] Lin Zhou, Michael K. Miller, H. Dillon, A. Palasyuk, S. Constantinides, R.W. McCallum, I.E. Anderson, M.J. Kramer, Role ofthe applied magnetic field on the microstructural evolution in Alnico 8 alloys, Metall. Mater. Trans. 1 (1) (2014) 27—35.

  [124] Liqin Ke, Ralph Skomski, Todd D. Hoffmann, Lin Zhou, Wei Tang, Duane

  D. Johnson, Matthew J. Kramer, Iver E. Anderson, C.-Z. Wang, Simulation of alnico coercivity, Appl. Phys. Lett. 111 (2) (2017) 022403.

  [125] I.E. Anderson, A.G. Kassen, E.M.H. White, L. Zhou, W. Tang, A. Palasyuk,

  K.W. Dennis, R.W. McCallum, M.J. Kramer, Novel pre-alloyed powder processing of modified alnico 8: correlation of microstructure and magnetic properties, J. Appl. Phys. 117 (17) (2015) 17D138.

  [126] Aaron G. Kassen, Emma MH. White, Wei Tang, Liangfa Hu, Andriy Palasyuk, Lin Zhou, Iver E. Anderson, Compression molding and novel sintering treatments for alnico Type-8 permanent magnets in near-final shape with preferred orientation, JOM (J. Occup. Med.) 69 (9) (2017) 1706—1711.

  [127] Lin Zhou, Emma White, Liqin Ke, David A. Cullen, M. K. Miller, Ping Lu, S. Constantinides, R.W. McCallum, I. E. Anderson, and M. J. Kramer, Microstructure and coercivity in alnico 9, J. Magn. Magn Mater., (to appear).

  [128] Palasyuk, Andriy, Ralph W. McCallum, Iver E. Anderson, Matthew Kramer, Lin Zhou, and Wei Tang. Cobalt-Lean alnico alloy. U.S. Patent Application 15/ 330,754, filed May 4, 2017.

  [129] A. Mubarok, N. Bordeaux, E. Poirier, F.E. Pinkerton, J. Gattacceca, P. Rochette, R. Reisener, L.H. Lewis, J.I. Goldstein, Microstructural and magnetic characterization of the NWA 6259 iron meteorite, Meteoritics Planet Sci. 76 (2013).

  [130] John T. Wasson, Relationship between iron-meteorite composition and size: compositional distribution ofirons from North Africa, Geochem. Cosmochim. Acta 75 (7) (2011) 1757—1772.

  [131] M. Kotsugi, H. Maruyama, N. Ishimatsu, N. Kawamura, M. Suzuki, M. Mizumaki, K. Osaka, et al., Structural, magnetic and electronic state characterization of L1 0-type ordered FeNi alloy extracted from a natural meteorite, J. Phys. Condens. Matter 26 (6) (2014) 064206.

  [132] Peter PWasilewski, Magnetic characterization of the new magnetic mineral tetrataenite and its contrast with isochemical taenite, Phys. Earth Planet. In. 52 (1—2) (1988) 150—158.

  [133] Eric Poirier, Frederick E. Pinkerton, Robert Kubic, Raja K. Mishra, Nina Bordeaux, Arif Mubarok, Laura H. Lewis, Joseph I. Goldstein, Ralph Skomski, Katayun barmak, Intrinsic magnetic properties of L10 FeNi obtained from meteorite NWA 6259, J. Appl. Phys. 117 (17) (2015) 17E318.

  [134] J. Pauleve, A. Chamberod, K. Krebs, A. Bourret, Magnetization curves of Fe—Ni (50—50) single crystals ordered by neutron irradiation with an applied magnetic field, J. Appl. Phys. 39 (2) (1968) 989—990.

  [135] J.F. Albertsen, J.M. Knudsen, N.O. Roy-Poulsen, L. Vistisen, Meteorites and thermodynamic equilibrium in fcc iron-nickel alloys (25-50% Ni), Phys. Scripta 22 (2) (1980) 171.

  [136] J.F. Albertsen, Tetragonal lattice of tetrataenite (ordered Fe-Ni, 50-50) from 4 meteorites, Phys. Scripta 23 (3) (1981) 301.

  [137] T. Shima, M. Okamura, S. Mitani, K. Takanashi, Structure and magnetic properties for L10-ordered FeNi films prepared by alternate monatomic layer deposition, J. Magn. Magn Mater. 310 (2) (2007) 2213—2214.

  [138] Takayuki Kojima, Misako Ogiwara, Masaki Mizuguchi, Masato Kotsugi, Tomoyuki Koganezawa, Takumi Ohtsuki, Taka-Yuki Tashiro, Koki Takanashi, Fe—Ni composition dependence of magnetic anisotropy in artificially fabricated L1 0-ordered FeNi films, J. Phys. Condens. Matter 26 (6) (2014) 064207.

  [139] Andreas Frisk, Thomas PA. Hase, Svedlindh Peter, Erik Johansson, Gabriella Andersson, Strain engineering for controlled growth of thin-film FeNi L10, J. Phys. Appl. Phys. 50 (8) (2017) 085009.

  [140] Louis Neel, J. Pauleve, R. Pauthenet, J. Laugier, D. Dautreppe, Magnetic properties of an irondnickel single crystal ordered by neutron bombardment, J. Appl. Phys. 35 (3) (1964) 873—876.

  [141] A. Chamberod, J. Laugier, J.M. Penisson, Electron irradiation effects on ironnickel invar alloys, J. Magn. Magn Mater. 10 (2—3) (1979) 139—144.

  [142] Yunlong Geng, Tursunjan Ablekim, Mark A. Koten, Marc Weber, Kelvin Lynn, Jeffrey E. Shield, Defect generation and analysis in mechanically alloyed stoichiometric Fe—Ni alloys, J. Alloy. Comp. 633 (2015) 250—255.

  [143] E. Lima Jr., V. Drago, A new process to produce ordered Fe50Ni50 tetrataenite, Physica Status Solidi (A) 187 (1) (2001) 119—124.

  [144] Seungwon Lee, Kaveh Edalati, Hideaki Iwaoka, Zenji Horita, Takumi Ohtsuki, Takuo Ohkochi, Masato Kotsugi, Takayuki Kojima, Masaki Mizuguchi, Koki Takanashi, Formation of FeNi with L 10-ordered structure using high- pressure torsion, Phil. Mag. Lett. 94 (10) (2014) 639e646.

  [145] Akihiro Makino, Parmanand Sharma, Kazuhisa Sato, Akira Takeuchi, Yan Zhang, Kana Takenaka, Artificially produced rare-earth free cosmic magnet, Sci. Rep. 5 (2015) 16627.

  [146] Sho Goto, Hiroaki Kura, Eiji Watanabe, Yasushi Hayashi, Hideto Yanagihara, Yusuke Shimada, Masaki Mizuguchi, Koki Takanashi, Eiji Kita, Synthesis of single-phase L 1 0-FeNi magnet powder by nitrogen insertion and topotactic extraction, Sci. Rep. 7 (1) (2017) 13216.

  [147] Maogang Gong, Shenqiang Ren, Phase transformation-driven surface reconstruction of FeNi nanostructures, Chem. Mater. 27 (22) (2015) 7795e7800.

  [148] Jian Shen, Maogang Gong, Qilin Dai, Shenqiang Ren, Surface-stress-induced phase transformation of ultrathin FeCo nanowires, ACS Appl. Mater. Interfaces 8 (1) (2015) 31e36.

  [149] Jian Shen, Qilin Dai, Shenqiang Ren, Phase transformation controlled tetra-gonality of MnNi-based nanocrystals, Nanotechnology 27 (10) (2016), 10LT01.

  [150] Maogang Gong, Xin Jin, Ridwan Sakidja, Shenqiang Ren, Synergistic strain engineering effect of hybrid plasmonic, catalytic, and magnetic coreeshell nanocrystals, Nano Letters 15 (12) (2015) 8347e8353.

  [151] Maogang Gong, Alec Kirkeminde, Manfred Wuttig, Shenqiang Ren, Phase transformation-induced tetragonal FeCo nanostructures, Nano Lett. 14 (11) (2014) 6493e6498.

  [152] Till Burkert, Lars Nordstro€m, Olle Eriksson, Olle Heinonen, Giant magnetic anisotropy in tetragonal FeCo alloys, Phys. Rev. Lett. 93 (2) (2004) 027203.

  [153] Aimo Winkelmann, Marek Przybylski, Feng Luo, Yisheng Shi, Jochen Barthel, Perpendicular magnetic anisotropy induced by tetragonal distortion of FeCo alloy films grown on Pd (001), Phys. Rev. Lett. 96 (25) (2006) 257205.

  [154] F. Yildiz, M. Przybylski, X.-D. Ma, J. Kirschner, Strong perpendicular anisotropy in Fe 1- x Co x alloy films epitaxially grown on mismatching Pd (001), Ir (001), and Rh (001) substrates, Phys. Rev. B 80 (6) (2009) 064415.

  [155] Gabriella Andersson, Till Burkert, Warnicke Peter, Matts Bjo€rck, Biplab Sanyal, Cyril Chacon, Claudia Zlotea, Lars Nordstro€m, Per Nordblad, Olle Eriksson, Perpendicular magnetocrystalline anisotropy in tetragonally distorted Fe-Co alloys, Phys. Rev. Lett. 96 (3) (2006) 037205.

  [156] Geraud Moulas, Anne Lehnert, Stefano Rusponi, Zabloudil Jan, Corina Etz, Safia Ouazi, Manuel Etzkorn, et al., High magnetic moments and anisotropies for Fex Co1- x monolayers on Pt (111), Phys. Rev. B 78 (21) (2008) 214424.

  [157] Delczeg-Cziijak, K. Erna, Edstrom Alexander, M. Werwinski, Rusz Jan, Natalia V. Skorodumova, Levente Vitos, Olle Eriksson, Stabilization of the tetragonal distortion of Fe x Co 1- x alloys by C impurities: a potential new permanent magnet, Phys. Rev. B 89 (14) (2014) 144403.

  [158] Ludwig Reichel, Giannopoulos George, Sandra Kauffmann-Weiss, Martin Hoffmann, Darius Pohl, Edstro€m Alexander, Steffen Oswald, et al., Increased magnetocrystalline anisotropy in epitaxial Fe-Co-C thin films with spontaneous strain, J. Appl. Phys. 116 (21) (2014) 213901.

  [159] Ludwig Reichel, Ludwig Schultz, Darius Pohl, Steffen Oswald, Sebastian F€hler, M. Werwinski, Edstrom Alexander, K. Erna, Delczeg-Czir-jak, Rusz Jan, From soft to hard magnetic FeeCoeB by spontaneous strain: a combined first principles and thin film study, J. Phys. Condens. Matter 27 (47) (2015) 476002.

  [160] T.R. Gao, Y.Q. Wu, S. Fackler, I. Kierzewski, Y. Zhang, A. Mehta, M.J. Kramer, I. Takeuchi, Combinatorial exploration of rare-earth-free permanent magnets: magnetic and microstructural properties of Fe-Co-W thin films, Appl. Phys. Lett. 102 (2) (2013) 022419.

  [161] Takashi Hasegawa, Shunsuke Kanatani, Miyu Kazaana, Kairi Takahashi, Kohei Kumagai, Maiko Hirao, Shunji Ishio, Conversion of FeCo from soft to hard magnetic material by lattice engineering and nanopatterning, Sci. Rep. 7 (1) (2017) 13215.

  [162] Balamurugan Balamurugan, Bhaskar Das, V.R. Shah, Ralph Skomski, X.Z. Li, David J. Sellmyer, Assembly of uniaxially aligned rare-earth-free nanomagnets, Appl. Phys. Lett. 101 (12) (2012) 122407.

  [163] B. Balamurugan, B. Das, W.Y. Zhang, R. Skomski, D.J. Sellmyer, HfeCo and ZreCo alloys for rare-earth-free permanent magnets, J. Phys. Condens. Matter 26 (6) (2014) 064204.

  [164] Balamurugan Balasubramanian, Pinaki Mukherjee, Ralph Skomski, Priyanka Manchanda, Bhaskar Das, J. David, Sellmyer. Magnetic nanostructuring and overcoming Brown's paradox to realize extraordinary high-temperature energy products, Sci. Rep. 4 (2014) 6265.

  [165] Michael A. McGuire, Rios Orlando, Evolution of magnetic properties and microstructure of Hf2Co11B alloys, J. Appl. Phys. 117 (5) (2015) 053912.

  [166] Michael A. McGuire, Orlando Rios, Nirmal J. Ghimire, Michael Koehler, Hard ferromagnetism in melt-spun Hf2Co11B alloys, Appl. Phys. Lett. 101 (20) (2012) 202401.

  [167] H.W. Chang, M.C. Liao, C.W. Shih, W.C. Chang, C.C. Yang, C.H. Hsiao, H. Ouyang, Hard magnetic property enhancement ofCo7Hf-based ribbons by boron doping, Appl. Phys. Lett. 105 (19) (2014) 192404.

  [168] Balamurugan Balasubramanian, Bhaskar Das, Ralph Skomski, Wenyong Y. Zhang, David J. Sellmyer, Novel nanostructured rare-earth-free magnetic materials with high energy products, Adv. Mater. 25 (42) (2013) 6090e6093.

  [169] Zhipeng Hou, Feng Su, Shifeng Xu, Jinbao Zhang, Chunji Wu, Dan Liu, Beipei Wei, Wenquan Wang, Magnetic properties, phase evolution, and microstructure of the CoeZreV ribbons, J. Magn. Magn Mater. 346 (2013) 124e129.

  [170] W.Y. Zhang, X.Z. Li, S. Valloppilly, Ralph Skomski, Jeffrey E. Shield, David J. Sellmyer, Magnetism of rapidly quenched rhombohedral Zr2Co11-based nanocomposites, J. Phys. D Appl. Phys. 46 (13) (2013) 135004.

  [171] K.H.J. Buschow, J.H. Wernick, G.Y. Chin, Note on the Hf-Co phase diagram, J. Less Common Met. 59 (1) (1978) 61 e67.

  [172] C. Gao, H. Wan, G.C. Hadjipanayis, High coercivity in non-rare-earth containing alloys, J. Appl. Phys. 67 (9) (1990) 4960e4962.

  [173] T. Ishikawa, K. Ohmori, Hard magnetic phase in rapidly quenched Zr-Co-B alloys, IEEE Trans. Magn. 26 (5) (1990) 1370e1372.

  [174] Hiroaki Okamoto, Phase Diagram for Binary Alloys, Desk Hand Book, 2000, p. 248.

  [175] Hiroaki Okamoto, Phase Diagram for Binary Alloys, Desk Hand Book, 2000, p. 263.

  [176] X. Zhao, M.C. Nguyen, W.Y. Zhang, C.Z. Wang, Matthew J. Kramer, David J. Sellmyer, X.Z. Li, et al., Exploring the structural complexity of intermetallic compounds by an adaptive genetic algorithm, Phys. Rev. Lett. 112 (4) (2014) 045502.

  [177] G.V. Ivanova, N.N. Shchegoleva, A.M. Gabay, Crystal structure of Zr2Co11 hard magnetic compound, J. Alloy. Comp. 432 (1e2) (2007) 135e141.

  [178] Bhaskar Das, Balamurugan Balamurugan, Pankaj Kumar, Ralph Skomski, V.R. Shah, Jeffrey E. Shield, Arti Kashyap, David J. Sellmyer, ${\rm HfCo} _ {7} $-Based rare-earth-free permanent-magnet alloys, IEEE Trans. Magn. 49 (7) (2013) 3330e3333.

  [179] Xingxu Lu, Kaiming Cheng, Shuhong Liu, Kai Li, Feng Zheng, Yong Du, Experimental investigation of phase equilibria in the CoeHf system, J. Alloy. Comp. 627 (2015) 251e260.

  [180] X.-Z. Li, Y.-L. Jin, M.-Y. Wang, J.E. Shield, R. Skomski, D.J. Sellmyer, Electron diffraction study of cobalt-rich Hf-Co, Intermetallics 75 (2016) 54e61.

  [181] P. Kumar, Arti Kashyap, Balamurugan Balamurugan, Jeffrey E. Shield, David J. Sellmyer, Ralph Skomski, Permanent magnetism of intermetallic compounds between light and heavy transition-metal elements, J. Phys. Con-dens. Matter 26 (6) (2014) 064209.

  [182] L.Y. Chen, H.W. Chang, C.H. Chiu, C.W. Chang, W.C. Chang, Magnetic properties, phase evolution, and coercivity mechanism of Co x Zr 98- x B 2 (x= 74- 86) nanocomposites, J. Appl. Phys. 97 (10) (2005) 10F307.

  [183] David J. Sellmyer, Balamurugan Balamurugan, Bhaskar Das, P. Mukherjee,

  R. Skomski, George C. Hadjipanayis, Novel structures and physics of nanomagnets, J. Appl. Phys. 117 (17) (2015) 172609.

  [184] M.J. Kramer, R.W. McCallum, I.A. Anderson, Steven Constantinides, Prospects for non-rare earth permanent magnets for traction motors and generators, JOM (J. Occup. Med.) 64 (7) (2012) 752e763.

  [185] Narayan Poudyal, J. Ping Liu, Advances in nanostructured permanent magnets research, J. Phys. Appl. Phys. 46 (4) (2012) 043001.

  [186] V.G. Harris, Y. Chen, A. Yang, S. Yoon, Z. Chen, A.L. Geiler, J. Gao, et al., High coercivity cobalt carbide nanoparticles processed via polyol reaction: a new permanent magnet material, J. Phys. Appl. Phys. 43 (16) (2010) 165003.

  [187] Yajing Zhang, Girija S. Chaubey, Chuanbing Rong, Yong Ding, Poudyal Narayan, Po-ching Tsai, Qiming Zhang, J. Ping Liu, Controlled synthesis and magnetic properties of hard magnetic CoxC (x= 2, 3) nanocrystals, J. Magn. Magn Mater. 323 (11) (2011) 1495e1500.

  [188] Mehdi Zamanpour, Steven Bennett, Parisa Taheri, Yajie Chen, G. Vincent, Harris, Magnetic properties and scale-up of nanostructured cobalt carbide permanent magnetic powders, J. Appl. Phys. 115 (17) (2014) 17A747.

  [189] El-Gendy, A. Ahmed, Meichun Qian, Zachary J. Huba, Shiv N. Khanna, Everett

  E. Carpenter, Enhanced magnetic anisotropy in cobalt-carbide nanoparticles, Appl. Phys. Lett. 104 (2) (2014) 023111.

  [190] S.Q. Wu, B. Balamurugan, X. Zhao, S. Yu, Manh Cuong Nguyen, Y. Sun,

  S. R. Valloppilly, D.J. Sellmyer, K.M. Ho, C.Z. Wang, Exploring new phases of Fe3- xCoxC for rare-earth-free magnets, J. Phys. Appl. Phys. 50 (21)(2017) 215005.

  [191] Xiaobai Wang, Peng Zhang, Jiajia Gao, Xiaodong Chen, Hua Yang, Facile synthesis and magnetic properties of Fe3C/C nanoparticles via a solegel process, Dyes Pigments 112 (2015) 305e310.

  [192] Yu-Hua Lee, T.C. Han, Jung-Chun Andrew Huang, Magnetic properties of Fe 3 C nanograins embedded in carbon matrix, J. Appl. Phys. 93 (10) (2003) 8462e8464.

  [193] El-Gendy, A. Ahmed, Massimo Bertino, Dustin Clifford, Meichun Qian, Shiv

  N. Khanna, Everett E. Carpenter, Experimental evidence for the formation of CoFe2C phase with colossal magnetocrystalline-anisotropy, Appl. Phys. Lett. 106 (21) (2015) 213109.

  [194] Sayan Bhattacharyya, Iron nitride family at reduced dimensions: a review of their synthesis protocols and structural and magnetic properties, J. Phys. Chem. C 119 (4) (2015) 1601e1622.

  [195] Kenneth Henderson Jack, The occurrence and the crystal structure of a-iron nitride; a new type of interstitial alloy formed during the tempering of nitrogen-martensite, Proc. Roy. Soc. Lond. a 208 (1093) (1951) 216e224.

  [196] T.K. Kim, Minoru Takahashi, New magnetic material having ultrahigh magnetic moment, Appl. Phys. Lett. 20 (12) (1972) 492e494.

  [197] Y. Sugita, K. Mitsuoka, M. Komuro, H. Hoshiya, Y. Kozono, M. Hanazono, Giant magnetic moment and other magnetic properties of epitaxially grown Fe16N2 single-crystal films, J. Appl. Phys. 70 (10) (1991) 5977e5982.

  [198] Yutaka Sugita, Hiromasa Takahashi, Matahiro Komuro, Masukazu Igarashi, Ryo Imura, Takashi Kambe, Magnetic and electrical properties of singlephase, single-crystal Fe16N2 films epitaxially grown by molecular beam epitaxy, J. Appl. Phys. 79 (8) (1996) 5576e5581.

  [199] Migaku Takahashi, H. Shoji, a 00-Fe16N2 problemdgiant magnetic moment  or not, J. Magn. Magn Mater. 208 (3) (2000) 145e157.

  [200] J.M.D. Coey, The magnetization of bulk a ‘Fe16N2, J. Appl. Phys. 76 (10) (1994) 6632e6636.

  [201] Yutaka Sugita, Hiromasa Takahashi, Matahiro Komuro, Katsuya Mitsuoka, Akimasa Sakuma, Magnetic and Mo€ssbauer studies of single-crystal Fe16N2 and Fe-N martensite films epitaxially grown by molecular beam epitaxy, J. Appl. Phys. 76 (10) (1994) 6637e6641.

  [202] Jian-Ping Wang, Nian Ji, Xiaoqi Liu, Yunhao Xu, C. Sanchez-Hanke,

  Yiming Wu, F.M.F. De Groot, Lawrence F. Allard, Edgar Lara-Curzio, Fabrication of $\hbox {Fe} _ {16}\hbox {N} _ {2} $ films by sputtering process and experimental investigation of origin of giant saturation magnetization in $\hbox {Fe} _ {16}\hbox {N} _ {2} $, IEEE Trans. Magn. 48 (5) (2012) 1710e1717.

  [203] Tomoyuki Ogawa, Yasunobu Ogata, Ruwan Gallage, Naoya Kobayashi, Naoaki Hayashi, Yoshihiro Kusano, Shinpei Yamamoto, et al., Challenge to the synthesis of a''-Fe16N2 compound nanoparticle with high saturation magnetization for rare earth free new permanent magnetic material, APEX 6 (7) (2013) 073007.

  [204] Yanfeng Jiang, Md Al Mehedi, Engang Fu, Yongqiang Wang, Lawrence

  F. Allard, Jian-Ping Wang, Synthesis of Fe 16 N 2 compound free-standing foils with 20 MGOe magnetic energy product by nitrogen ion-implantation, Sci. Rep. 6 (2016) 25436.

  [205] R.W. McCallum, Laura H. Lewis, R. Skomski, M.J. Kramer, I.E. Anderson, Practical aspects of modern and future permanent magnets, Annu. Rev. Mater. Res. 44 (2014) 451e477.

  [206] R. Skomski, J.M.D. Coey, Magnetic anisotropydhow much is enough for a permanent magnet? Scripta Mater. 112 (2016) 3e8.

  [207] Shinpei Yamamoto, Ruwan Gallage, Yasunobu Ogata, Yoshihiro Kusano, Naoya Kobayashi, Tomoyuki Ogawa, Naoaki Hayashi, Kaori Kohara, Migaku Takahashi, Mikio Takano, Quantitative understanding of thermal stability of a00-Fe 16 N 2, Chem. Commun. 49 (70) (2013) 7708e7710.

  [208] D.C. Sun, E.Y. Jiang, M.B. Tian, C. Lin, X.X. Zhang, Epitaxial single crystal Fe16N2 films grown by facing targets sputtering, J. Appl. Phys. 79 (8) (1996) 5440e5442.

  [209] Kensuke Nakajima, Shoichi Okamoto, Nitrogen-implantation-induced transformation of iron to crystalline Fe16N2 in epitaxial iron films, Appl. Phys. Lett. 54 (25) (1989) 2536e2538.

  [210] H. Shinno, K. Saito, Effects of film thickness on formation processes of Fe16N2 in nitrogen ion-implanted Fe films, Surf. Coating. Technol. 103 (1998) 129e134.

  [211] Yanfeng Jiang, Vivekanand Dabade, Lawrence F. Allard, Edgar Lara-Curzio, Richard James, Jian-Ping Wang, Synthesis of a00— Fe 16 N 2 compound anisotropic magnet by the strained-wire method, Phys. Rev. Appl. 6 (2) (2016) 024013.

  [212] I. Dirba, C.A. Schwo€bel, L.V.B. Diop, M. Duerrschnabel, L. Molina-Luna, K. Hofmann, P. Komissinskiy, H.-J. Kleebe, O. Gutfleisch, Synthesis, morphology, thermal stability and magnetic properties of a 00-Fe16N2 nanoparticles obtained by hydrogen reduction of g-Fe2O3 and subsequent nitrogenation, Acta Mater. 123 (2017) 214e222.

  [213] Jing Li, Wei Yuan, Xiaoling Peng, Yanting Yang, Jingcai Xu, Xinqing Wang, Bo Hong, Hongxiao Jin, Dingfeng Jin, Hongliang Ge, Synthesis of fine a 00-Fe16N2 powders by low-temperature nitridation of a-Fe from magnetite nanoparticles, AIP Adv. 6 (12) (2016) 125104.

  [214] Rizka Zulhijah, Asep Bayu Dani Nandiyanto, Takashi Ogi, Toru Iwaki, Keitaro Nakamura, Kikuo Okuyama, Gas phase preparation of spherical coreeshell a00-Fe 16 N 2/SiO 2 magnetic nanoparticles, Nanoscale 6 (12) (2014) 6487e6491.

  [215] Takashi Ogi, Qing Li, Shinji Horie, Akito Tameka, Toru Iwaki, Kikuo Okuyama, High-purity core-shell a 00-Fe16N2/Al2O3 nanoparticles synthesized from a-hematite for rare-earth-free magnet applications, Adv. Powder Technol. 27 (6) (2016) 2520e2525.

  [216] Kim, Gyeong Su, Kwang-Won Jeon, Jung-Goo Lee, Hyeongjun Kong, Hyoungjeen Jeen, Se-Hun Kwon, Youn-Kyoung Baek, High purity a-Fe 16 N2 particles with tunable sphere structures via spray drying method, J. Magn. 22 (4) (2017) 590e595.

  [217] Yanfeng Jiang, Jinming Liu, Pranav K. Suri, Greg Kennedy, Naresh N. Thadhani, David J. Flannigan, Jian-Ping Wang, Preparation of an a 00-Fe16N2 magnet via a ball milling and shock compaction approach, Adv. Eng. Mater. 18 (6) (2016) 1009e1016.

  [218] Kenta Takagi, Misaho Akada, Kimihiro Ozaki, Naoya Kobayashi, Tomoyuki Ogawa, Yasunobu Ogata, Migaku Takahashi, High-pressure sintering behavior of a 00-Fe16N2 nanopowder, J. Appl. Phys. 115 (10) (2014) 103905.

  [219] Jun Cui, Jung-Pyung Choi, Guosheng Li, Evgueni Polikarpov, J. Darsell, N. Overman, M. Olszta, et al., Thermal stability of MnBi magnetic materials, J. Phys. Condens. Matter 26 (6) (2014) 064212.

  [220] Wei Xie, Evgueni Polikarpov, Jung-Pyung Choi, Mark E. Bowden, Kewei Sun, Jun Cui, Effect of ball milling and heat treatment process on MnBi powders magnetic properties, J. Alloy. Comp. 680 (2016) 1e5.

How to choose and buy a strong neodymium magnet? ALBMagnets is a professional company for strong magnet design and manufacturing,
providing you with reliable N35, N38, N42, N52, N42SH and other grade super neodymium magnets and SmCo rare earth magnets.

Can't Find The Magnets You Need?