EP4626627A1 - Direct synthesis of l10 feni and other metal alloys from metal-cyano/nitrosyl complexes - Google Patents

Abstract

An effective and easily scaled-up chemical synthesis route to obtain L1₀ Iron-Nickel alloy from cyano/nitrosyl-metal complex precursor salts, for sustainable permanent magnets free of crittical rare-earth elements, with a (BH)_max ranging in between that 5 of hexaferrites (up tO ~45 kJ/m³) and rare-earth-based permanent magnets (up to ~500 kJ/m³).

Description

DIRECT SYNTHESIS OF L1 FENI AND OTHER METAL ALLOYS FROM METAL- CYANO/NITROSYL COMPLEXES DESCRIPTION SUMMARY OF THE INVENTION The present invention relates to an easily scaled-up chemical synthesis route exploiting the use of crystalline cyano/nitrosyl- metal complex salt precursors to obtain L1 Iron-Nickel alloy nano-powders for permanent magnets free of critical rare-earth elements. STATE OF THE ART Rare-earth elements (REEs) are essential components for the industry, the economy, and the daily life. They are used in various high-technology applications and are crucial for the production of high-performance permanent magnets used in different sectors, such as electronics, renewable energy technologies, robotics, electric vehicles, aerospace applications, and many others, etc. Rare-earth elements, e.g., neodymium, praseodymium, dyspro- sium, and terbium, are especially important as they allow achiev- ing the highest energy efficiency currently available. Despite accounting for only 25 percent of the total rare-earth production volume, these elements represent a staggering 80 to 90 percent of the rare-earth elements' market value. This emphasizes their crit- ical importance in various industries and underscores their eco- nomic significance. Currently, the mining and processing of REEs are primarily concentrated in a small number of countries, with China being the dominant player in this industry. This concentration of supply poses a significant risk, as any sudden disruption in the supply chain could have far-reaching consequences for various sectors, the economy, and the everyday lives of individuals residing in countries where REE mining and processing are limited or non- existent, which includes the majority of countries worldwideMore- over, the mining and refining processes of REEs require very large amounts of energy and water, while generating vast volumes of GHG emissions, such as to make the replacement of REEs highly desir- able for environmental purposes along with economic and geopolit- ical reasons. To address the REE criticality issues, an intense computa- tional and experimental activity has been carried out to develop effective solutions that allow reducing the demand and use of REEs, including the optimization of existing materials, the de- velopment of new hard magnetic phases, and the recycling/reuse of End-of-Life magnets. About the optimization/development of mate- rials, a significant work has been made in the direction of opti- mizing the microstructure of known materials by reducing the grain size (down to the nano-scale regime) and enhancing the grain iso- lation to increase the magnetic coercivity and thus the perfor- mance. In addition, meaningful improvements can be achieved by developing exchange-coupled composites consisting of a homogene- ous distribution of nanostructured hard and soft ferromagnetic phases in intimate contact. Although there is room for improving the performance of conventional materials, a major achievement would be to develop new hard magnetic materials based on non- critical, easily accessible and reusable/recyclable elements (e.g. Fe, Ni, Mn, Al) with a high saturation magnetization (^M > 1.2 T) and magnetic anisotropy (K > 1 MJ/m), capable of match- ing/surpassing the performance of the state-of-art RE-based per- manent magnets, or bridging the gap between hexaferrites and RE- based permanent magnets in terms of cost-performance. Different binary and ternary alloys have been investigated for this purpose, and, among the different options, the chemically ordered L1 FeNi alloy, existing in the composition range Fe Ni – Fe Ni , and consisting of planes of pure Fe and Ni atoms alternating along the c-axis of the tetragonal fct unit cell, has attracted a great deal of attention, owing to its large saturation magnetization (1.6 T) and high uniaxial magnetic anisotropy (~ 10 J/m), which would lead to a theoretical maximum energy product (BH) of ~0.45MJ/m approaching that of Nd-Fe-B. These values, along with a rather high Curie temperature (up to ~ 550 ° C) and an excellent chemical stability, make the L1 FeNi alloy particularly attrac- tive for applications in various fields (e.g., automotive sector, green technologies, aerospace, energy, catalysis, biomedicine, and robotics) as a replacement for the currently used materials containing critical materials such as REEs or Pt group metals. However, despite its excellent properties, the fabrication of the L1 phase is extremely complex due to the low order/disorder tran- sition temperature (T ~ 320 ° C), above which the fcc A1 phase, with low magnetic anisotropy, forms. Although the L1 phase is stable at low temperature, to favor the chemical ordering, the process temperature must be lower than T but, at the same time, high enough to guarantee an adequate atomic interdiffusion. Owed to the low T value, the low atomic mobility below the transition temperature kinetically limits the formation of the L1 phase that naturally occurs only in meteorites, which have cooled through billions of years at extremely low cooling rates (1 K per million years). Various strategies have been proposed to favor the for- mation of the ordered phase under accessible conditions and time scales; however, none of them can guarantee the production of the L1-FeNi alloy with levels of performance, yield, costs, and en- vironmental sustainability that can be used for the production of the material on industrial scale. The first attempts date back to the 1960s and involve the use of neutron beams to accelerate the diffusion processes and induce the formation of the ordered phase. Despite the good results (K ~ 3·10 J/m), the proposed technique is not scalable for the mass production of permanent magnets, as it requires the use of nuclear reactors or particle accelerators. In 2001 E. Lima and co-authors proposed an innovative method that involves a series of oxidation and reduction cycles at low temperature (623 K) of iron granules coated with nickel. The rather complex and long process leads to the formation of FeNi alloys with a low percentage of ordered phase (~19% in the best case). In 2013, Y. Hayashi and co-authors discussed the possibility of obtaining the L1-FeNi phase by re- duction with CaH of Fe and Ni chlorides mixtures at different concentrations of the two metals. Under the best experimental conditions, coercivity values of approximately 67·10 T were ob- tained, which the authors associated to the formation of the or- dered phase. The same authors argue that an excess of Ni is present in the alloy, which is responsible for a strong reduction of the saturation magnetization (~100 KAm/Kg) that makes the material unsuitable for the development of permanent magnets. In the same year, a US patent was filed (US 2013/0186238 A1), detailing a method for producing ferromagnetic FeNi alloy powders through the thermal treatment of a solution containing FeCl and NiCl salts, expressed as FeCl∙2HO∙NiCl∙2HO. It is evident to a skilled technician that FeCl∙2HO∙NiCl∙2HO is not a structured and lay- ered compound displaying bonds between atoms, but rather repre- sents a simple mixture of two distinct salts. Despite the authors' claim of achieving a chemically ordered L1 phase, there is no reported structural analysis (such as X-ray diffraction, electron diffraction, or Mössbauer spectroscopy) to support this assertion. Furthermore, the powders exhibit a relatively high coercivity of approximately 50·10 T but a rather low saturation magnetization of around 0.6 MA/m. This suggests the presence of other magnetic phases, specifically antiferromagnetic Fe-oxides, which could po- tentially explain the increased coercivity through coupling with the ferromagnetic A1 phase of FeNi (known as the exchange bias effect). In 2014 S. Lee and co-authors described the possibility of obtaining the L1-FeNi alloy through the application of an intense plastic deformation by high pressure torsion (6GMp) for a time ranging between 10 and 40 days. Although plastic deformation seems to favor atomic diffusion through the introduction of a high-density lattice defects, the authors do not describe the magnetic properties of the final products, thus making impossible to evaluate the real effectiveness of the method, which however remains unsuitable for industrial production, because of the time- consuming nature of the process. Next, in 2015, Y. Geng and co- authors employed grinding techniques to mix metallic Fe and Ni powders (10 - 90 hours), introducing defects that can favor the diffusion of the elements during subsequent heat treatments; how- ever, the magnetic characteristics of the obtained materials are compatible with a low magnetic anisotropy FeNi phase. Also in 2015, A. Makino and co-authors demonstrated that permanent magnets based on the L1-FeNi alloy can be obtained by crystallizing an amorphous Fe Ni SixB PCu (x = 2 to 8 at. %) alloy at a crystallization temperature close to that of FeNi order-disorder transition. Although the hysteresis loops seem to indicate the formation of a phase with high magnetic anisotropy, the volume fraction of the ordered phase is rather low (<10%), resulting in low saturation magnetization values (~ 100 KAm/Kg) that prevents its use for the production of permanent magnets. Another inter- esting approach involves the use of AuCu nanoparticles with L1 structure to be employed as templating agents to obtain, through the surface stress, a tetragonal reconstruction of a FeNi shell grown on top. The method allows obtaining systems with high mag- netic anisotropy (up to 7·10 J/m) and order parameter S = 0.45; however, the proposed strategy is working only for few nanometers thick FeNi shells, above which the AuCu tetragonalizing effect is lost. Furthermore, although it is not clearly discussed by the authors, the presence of a significant amount of non-magnetic templating material should result in an important reduction in the saturation magnetization of the whole material and, therefore, in the performance of any possible magnet. In addition to the methods mentioned above, several works show that the temperature necessary to obtain the ordered phase can be significantly reduced in pre-ordered thin films consisting of alternating layers of pure Ni and Fe that mimic the final structure of the L1 phase . Although thin film systems can be useful for applications in the field of ICT and sensors, the very small quantities of material produced do not allow them to be used for the manufacture of permanent bulk magnets. The idea of using preordered systems to obtain the L1-FeNi alloy has recently been exploited by S. Goto and co-authors; their strategy includes the formation of an in- termediate FeNiN phase, which has the same ordered chemical ar- rangement as the L1-FeNi alloy . The method allows to get the best results in terms of phase purity and magnetic hardness (^H ~ 0.17 T, M ~ 140 KAm/Kg, S = 0.71) achieved so far. However, the approach is rather complex and expensive as it involves mul- tiple heat treatments (up to 400 °C and 50 hours) in significantly high flows of hydrogen (1 L/min) and ammonia (5 L/min) for the treatment of only one gram of product, effectively limiting its use for a low-cost massive production of FeNi-based permanent magnets. PROBLEM AND SOLUTION The problem of the invention is to provide a scalable and effective route for the synthesis of highly ordered L1 FeNi alloy nano-powders to be used as building blocks for manufacturing rare- earth free permanent magnets with a (BH) ranging in between that of hexaferrites (up to ~45 kJ/m) and rare-earth-based permanent magnets (up to ~500 kJ/m). This purpose is achieved through a process for the prepara- tion of a bimetallic/ trimetallic alloy having chemically ordered crystal structure, characterized in that it involves the following steps: reduction of a cyano/nitrosyl-metal complex salt having alternating mononuclear layers in a reactor through a reducing agent, supplying heat to the reactor, with a heating rate equal or greater than 0.1 °C/min, at a temperature between 200°C-450°C, for a reaction time between 12 and 72 hours.Subclaims describe preferred features of the invention. BRIEF DESCRIPTION OF THE DRAWINGS Further characteristics and advantages of this invention are an- yway more apparent when reading the following detailed description of a preferred embodiment, which is given by way of example only, and with reference to the annexed drawings, wherein: fig. 1 is a schematic representation of preferred embodi- ments of a reduction process of the present invention; fig.2 illustrates XRD (X-Ray Diffraction) patterns of four different samples of binary Fe-Ni alloys of the present invention prepared under different conditions: (a) S290N, (b) S290T, (c) S400N, (d) S290N-HCl; fig.3 are TEM (Transmission Electron Microscopy) images of an L1 FeNi bimetallic alloy sample (S290N) of the present inven- tion at different magnifications; fig. 4 reports on EDX (Electron Dispersive X-ray) analysis: (a) EDX spectrum, (b) normalized line profile, (c) elemental map- ping of a FeNi L1 sample (S290N); fig. 5(a) illustrates the field dependent magnetization loops measured at room temperature of FeNi alloys of the present invention prepared under different conditions; fig. 5(b) illustrates the trend of coercivity (µH) and saturation magnetization (M) as a function of experimental con- ditions of FeNi alloy samples of the present invention; fig. 6(a, c, e) are Mössbauer spectra measured at 300 K of FeNi alloys of the present invention prepared under different conditions: (a) S290N, (c) S290T, (e) S400N; and fig. 6(b, d, f) illustrate a distribution of the hyperfine fields of the two sub-spectra, corresponding to the ferromagnetic components of the samples of the present invention: (b) S290N, (d) S290T, (f) S400N. DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT Some binary alloys, such as, for example and according to a preferred embodiment, the Iron-Nickel alloy, can exist near the equiatomic composition in two different forms: the A1 phase, hav- ing a face-centered cubic structure (fcc), disordered from a chem- ical point of view, with a random distribution of Fe and Ni atoms in the cubic cell (3.586 <a < 3.627 Å), and an L1 phase, wherein the L1 phase possesses a chemically ordered face centered tetrag- onal structure (fct), where the two atoms are found arranged in a super-lattice, consisting of monoatomic layers composed of, in this case, Fe and Ni atoms, alternating along the c-axis of the unit cell (3.5826 < a = b < 3.59235 Å, 3.5850 < c < 3.6002 Å). In particular, the chemical ordering of the L1 phase, gives rise to a high uniaxial magnetic anisotropy (of the order of 10 MJ/m) along the stacking direction, leading to a high coercivity (H). In contrast, the lack of chemical ordering of an A1 alloy makes said phase magnetically soft, namely it has low magnetic anisotropy and H values. This invention provides a process that can directly transform a two-metal Fe/Ni-based chemically ordered crystalline salt into a chemically ordered alloy through a single reduction step. Although the L1 phase is thermodynamically stable at room temperature, the stable phase at high temperature is the A1 phase. Therefore, to favor the chemical ordering, the process temperature must be below the order/disorder temperature (T ) associated with the A1-L1 phase transition. However, it is crucial to ensure that the temperature remains sufficiently high to facilitate atomic inter-diffusion. Due to the relatively low T value of the FeNi alloy (approximately 320 °C in bulk form), the limited atomic mobility below the transition temperature imposes kinetic constraints on the formation of the L1 phase. Consequently, the direct synthesis of an FeNi L1 alloy is exceptionally challeng- ing. in any case, the temperature must be high enough to ensure adequate atomic inter-diffusion. Owed to the low T value of the FeNi alloy (~320 °C in bulk form), the low atomic mobility below the transition temperature limits from a kinetic point of view the formation of the L1 phase, so that the direct synthesis of an Fe-Ni L1 alloy is extremely complicated, not to say impossible. The present invention proposes a solution to the problem above mentioned, allowing a preparation of an L1 alloy under time- effective and environmentally friendly experimental conditions, employable both on a small scale, as in a laboratory, and on a large scale, as in an industrial plant. The problem underlying the invention is solved by using precursors such as metal com- plexes that already are made up of alternating planes of Fe and Ni atoms that mimic the atomic arrangement of the L1 structure. The intrinsic atomic order present in the starting salt allows reducing the energy required to order the metallic atoms, thus driving the formation of the L1 phase, which is obtained by thermal reduction of the precursor salts at lower temperature sand reaction times with respect to other proposed strategies. The general concept of the approach and its feasibility were already proved by the authors of the present invention in a pre- vious study . The concept was successfully applied to the synthe- sis of MPt (M = Fe, Co, Ni) L1 alloys using M(HO)PtCl as a precursor salt. Building upon this previous achievement, the cur- rent invention marks the first-ever application of the same con- cept to the preparation of FeNi alloy. Additionally, finding the right ligand to facilitate and improve the synthesis of chemically ordered structures was not an immediate or easy task. The choice of CN groups as metal ligands was not random. In fact, the dative bond is a weak bond, much weaker than an ionic or covalent bond. This allows H to be used as a reductant at low temperatures. since these temperatures must be lower than the transition tem- perature to order-disorder (To-d) of many alloys considered espe- cially the FeNi L10 alloy which has a To-d of 320°. Experiments were carried out with compounds of the NiFeF type, but in this case the reduction temperature with H2 was higher than 450°, well above the To-d of FeNi L10. According to a preferred embodiment, the bimetallic L1-FeNi alloy according to the invention is obtained either through a thermal decomposition reduction reaction of Nickel Nitroprusside (NiFe(CN)NO)∙xHO) or through a thermal decomposition reduction reaction of Iron tetracyanonickelate (Fe(HO)Ni(CN)∙xHO). The hydration level (xHO) is not important for those reactions. When a precursor is exposed to a reducing agent, the oxidation state of the metal salts changes to zero, reaching the metal form. The counterions and ligands, being volatile, spontaneously remove themselves from the alloy once the reaction chamber is flushed with an inert gas. Examples of contemplated mentioned reducing agents are a reducing gas, preferably H, or a reducing salt, such as a hydride source, e.g., sodium boron hydride, lithium aluminum hydride, calcium hydride. Alternatively, a mixture of noble metals coupled to a reducing gas can be employed. The process of the present invention is easily scalable, environmentally friendly and versatile and can be easily extended for the synthesis of many other nearly equiatomic binary alloys (e.g., FeCo, CoNi, CuFe, CuNi) with a chemically ordered structure and even for the syn- thesis of ternary alloys of interest for multiple applications. This can be achieved by appropriately varying the metal cations in the starting salts. Fig.1 represents a reduction process of the invention. For clarity of representation, in the crystallographic structures of the precursor complex salts (a = Ni Nitroprussides, b = Fe Te- tracianonikelate) only the metal atoms are highlighted (Ni in dark gray and Fe light gray), while all the other elements (C, N, O, H) are shown in white. The salts have an ordered arrangement of the Fe and Ni atoms on alternating planes highlighted reminiscent of that of the same atoms in the L1 structure of FeNi. Both Nickel nitroprusside and Iron tetracyanonickelate are examples of salts having alternating mononuclear layers. The synthesis of the L1 phase of the FeNi alloy is, in general, extremely complex because of low atomic mobility present below the order/disorder transition temperature (T ), which lim- its from a kinetic point of view the formation of the ordered phase. Despite numerous efforts, experimental results are still far from theoretical predictions, and proposed approaches gener- ally involve complex and expensive protocols that cannot be easily scaled up for industrial production or, in other cases, are char- acterized by low L1 phase yields and poor overall performance, as discussed above. To overcome the limitations of the few approaches developed to date, a synthesis process is proposed, one which involves ac- ceptable reaction times, temperatures, and reagent flows. The process, described in detail below, is based on a low-temperature thermal decomposition-reduction reaction of precursors, having stoichiometric ratio of the two metals between 0.975 and 1.025:0.975 and 1.025. Favorite examples of these precursors are the metal-cyano/nitrosyl complexes of Fe and Ni, such as Nickel Nitroprusside (NiFe(CN)NO∙xHO) and Iron tetracyanonickelate (Fe(HO)Ni(CN)∙xHO x4HO) or other complexes from the same family metal-cyano/nitrosyl. Note that the hydration level of the mole- cules is not critical for the synthesis. Both Nickel Nitroprusside and Iron tetracyanonickelate consist of alternating planes of Fe and Ni atoms that mimic the layered atomic arrangement of the L1 structure. Preferably, the metallic salt is Iron tetracy- anonickelate (Fe(HO)Ni(CN)∙xHO) and the heating rate is 1 °C/min and the temperature is 290°C and the reducing agent is H, the acid is HCl (0.1M) or the metallic salt is Nickel Nitroprusside (NiFe(CN)NO∙xHO) and the heating rate is 1 °C/min and the tem- perature is 290°C and the reducing agent is H, the acid is HCl (0.1M). The atomic ordering of the precursors facilitates the for- mation of the ordered phase, which is achieved in a single step, through an incremental temperature reduction of the precursors, such as nickel nitroprusside or iron tetracyanonickelate. The starting materials, in this case, consist exclusively of non- critical elements (C, H, O, N, Fe, Ni) and are obtained via known and reliable procedures from commercially available starting ma- terials (see synthesis procedure of precursors below). The analyses, detailed below, show the effectiveness of the method to obtain nanometer-sized (20 to 120 nm) powders in a single step. In some embodiments, nanometer-sized powders are covered with a protective layer of carbon and contain more than 50 percent of L1 phase. In addition, nanometer-sized powders have high values of coercivity up to 66 ·10 T, with significant potential for improving, and saturation magnetization values (130 - 140 KAm/Kg) close to the bulk one (154 KAm/Kg), the slightly difference being due to the nanostructured nature of the sample and the presence of a Carbon shell, To demonstrate the feasibility and the effectiveness of the proposed synthesis method, four representative samples prepared under the conditions described in Table 1 are discussed. For all samples, the reaction time (t) was set at 48 hours, while the processing temperature (T) was fixed at values either slightly lower or slightly higher than the order/disorder transition tem- perature (T ). Furthermore, the effect of a post-synthesis treat- ment with a 0.1 M HCl solution used to increase the L1 phase percentage by selectively removing the A1 phase, which is less resistant to oxidation, will also be discussed. Table 1. Experimental conditions: precursor complex, processing temperature (T), reaction time (t), post synthesis treatment. Post syn- Sample Precursor Complex T (°C) t (h) thesis ID treatment 290 (< S290N NiFe(CN)NO)x5HO 24 T ) 290 (< S290T Fe(HO)Ni(CN)x4HO 24 T ) 400 (> S400N NiFe(CN)NO)x5HO 24 T ) S290N- 290 (< NiFe(CN)NO)x5HO 24 HCl HCl T ) Fig.2 shows experimental XRD spectra of the four samples discussed in the present invention. All samples show an XRD pattern (Fig. 2) that is characteristic of the FeNi metal alloy. Other phases, such as Fe/Ni oxides, are not detected. A univocal attribution of the reflection peaks to the tetragonal or to the cubic phase is extremely challenging. The similar size of Fe and Ni lead to a small tetragonality (the ratio c/a is close to 1) and a small peak splitting; the slight difference in X-ray scattering factors for Fe and Ni leads to extremely weak superlattice peaks (which originate from the alternate stacking of Fe and Ni planes along the c-axis), being the intensity of the strongest one, at 2θ ≈ 25°, only the 0.3% of that of the fundamental peak (111). In addition, since only a fraction of the tetragonal phase L1 is present, as shown later by Mössbauer measurements, there is a further reduction in the intensity of the superlattice peaks, which is difficult to detect, even using very intense X-ray sources, such as synchrotron facilities. The TEM images in (Figs 3 a-d) of FeNi L1 emphasize that the samples consist of micrometer-size aggregates of nanoparticles with sizes ranging from 20 to 120 nm. A deeper inspection shows that most of the particles are rounded (see dashed white circles, fraction 1, F1, Figs. 3 a and b) with a mean size of 40 – 50 nm, while a minority fraction consists of spherical particles with sizes in the 20 to 40 nm range (see dashed white circle, fraction 2, F2, Figs. 3 a and b) and few larger particles that rarely reach 120 nm (see dashed white circle, fraction 3, F3, Figs. 3 a and b). Most of the nanoparticles’ aggregates are homogeneously cov- ered with a very thin carbon layer of about 3-7 nm (Figs 3 c and d) that shows less contrast than the inner part. This carbon layer proved useful in preventing the oxidation of the alloy. TEM-EDX analysis (Figs. 4 a-c) on different areas of the sample confirms that the theoretical stoichiometry of the alloy is maintained, considering both point and selected-area measure- ments. This is also demonstrated by following the composition along a line: the Ni profile completely overpowers that of Fe, showing that Ni and Fe form a 1:1 alloy. The elemental mapping shown in Figure 4 c is a further confirmation that Ni and Fe are homogeneously distributed through- out the particles. Figure 5a shows the room-temperature magnetic response of the samples as the external magnetic field changes. The trends of coercivity (^H) and saturation magnetization (M), as determined by hysteresis cycles, and are shown in Figure 5b as a function of experimental conditions. All samples show high saturation magnetization values (130 to 140 KAm/Kg) close to the bulk value (154 KAm/Kg), confirming the absence of other phases such as Fe and Ni oxides. The small differences are in fact due to both the nanometer structure of the material and the presence of a thin layer of carbon around the particles which reduces, albeit minimally, the percentage of the magnetic phase to the total. In contrast, a significant change in the coercive field (^H) is observed as the experimental conditions change. For pro- cess temperatures below T , the coercive field assumes high val- ues (S290N: ^H ≈ 56·10 T; S290T: (^H ≈ 44·10 T) which are not compatible with the A1 phase of the FeNi alloy, thus suggesting the presence of particles with ordered structure. This hypothesis is reinforced by the fact that the coer- civity decreases to ~19 mT when the reaction is conducted at a temperature above T (sample S400N), i.e., under conditions that thermodynamically favor the formation of phase A1. A further in- crease in coercivity up to 66·10 Tis observed following treatment with hydrochloric acid, which can preferentially solubilize the disordered phase, which is nevertheless present, thereby increas- ing the final performance via augmenting the percentage of the L1 phase. The formation of the L1 phase is confirmed by the Mössbauer measurements shown in Figure 6 (a, c, e) for samples S290N, S290T and S400N. The Mössbauer technique is sensitive to the local Fe surround, which varies with the chemical order of the alloy, and thus allows the identification of the possible presence of the L1 phase. For proper interpolation of the spectra, it is necessary to consider three magnetic sub-spectra corresponding to two fer- romagnetic phases and one paramagnetic phase, which is present in minimal amounts (< 2%), the white dots represent experimental data of the samples, while the black curve represents an interpolation of the experimental data of the present invention. The individual sub-spectra forming the interpolation curve are shown by thick lines in black, dark grey and light grey. Figs. 6 (b, d, f) show that two principal domains are present and associated to the fer- romagnetic sub-spectra: the high-field component (mean value: 32.5 – 33 T) can be attributed to the A1 phase, while the low field component (mean value: 29.5 – 30 T) can be ascribed to the L1 phase, as disclosed by K. Mibu et al. . Analysis of the Mössbauer spectra clearly demonstrates the presence of both phases in all three samples. Analyses of hysteresis loops and Mössbauer spectra confirm that the ordered arrangement of the atoms in the precursor com- plexes is maintained in the reduction reaction, favoring the for- mation of the ordered phase. However, when the process temperature is higher than T , the thermodynamics of the process tends to favor the formation of the disordered phase by counteracting the initial salt order with a consequent reduction in the percentage of phase L1 and thus in coercivity values. To obtain coercivity values useful to produce permanent magnets, it is advantageous to conduct the synthesis at temperatures below T . However, acceptable results can also be obtained by operating at temperatures slightly above T . In the present case, acceptable results are obtained by operating at temperatures up to 450 °C. As shown in table 1, sample S400N was heated to 400°C. The L1-FeNi powders obtained by the method of the present invention can be potentially used for multiple applications in different fields (e.g., automotive, green technologies, aerospace, energy, catalysis, biomedicine, and robotics), and in particular as a starting material for making medium/high performance permanent magnets free of critical elements, reaching performances close to those of magnets containing rare-earth elements (e.g., Nd-Fe-B) and providing materials which are requested by the market and up to now unavailable. In addition, the alloy according to this invention can be synthesized starting from abundant, low cost and easily recyclable and reusable elements. The synthesis method is also extremely versatile and can be extended to the synthesis of other chemically ordered binary or ternary metal alloys (containing for example Fe, Co, Ni, Mn, etc. such as CoFe and CoNi), through the use of other metal- cyano/nitrosyl complexes such as, for example, CoFe(CN)NO∙xHO and Co(HO)Ni(CN)∙xHO. Non-magnetic alloys, such as CuNi, of interest for other purposes (e.g., catalysis) can also be synthesized using Cu(HO)Ni(CN)∙xHO as a precursor. In particular, the present synthesis method can be exploited to obtain alloys having a chemically ordered structure such as L1 or a body centered tetragonal (BCT). As an example, a BCT CoFe alloy is obtainable using the above cited precursors. Unlike previous published and patented processes, the process disclosed by this invention is a cost-effective process, scalable in a relatively easy way, with high yield and low environmental impact. Reduction of the starting salts can be performed by employing a reducing gas, such as, for example, H, or other reducing agents, such as, for example, alkali or alkaline earth metal hydrides and mixtures of noble metal powders and H. The process is conducted at a temperature (T) between 200 °C and 450 °C (i.e., in the range of T ~ 320 °C) and for a reaction time (t) varying in the range of 12 to 72 hours. As an alternative to the above-described process, the precursor salt can be placed in a shuttle, for example of alumina, and said shuttle can be inserted in the center of a quartz tube of a tubular furnace. The tube and its contents can first be flushed for an interval of 5 -15 min with an inert gas, such as Ar (flow rate: 0.01-1 L/min) and then with a reducing gas, such as H with an initial flow rate of 0.01-1 L/min, which preferably can be reduced in the range of 0.001-0.05 L/min at steady state. During this step, the furnace is brought to the desired temperature with a heating rate greater than 0.1/min and maintained at the desired temperature for a reaction time (t) between 12 and 72 hours. At the end of the process, the furnace is cooled down to room temperature; then, the quartz tube and its contents are washed with an inert gas, such as Ar, and the sample is finally extracted. Eventually, the reduction can occur coupled to a magnetic field to increase the percent of ordered phase, in fact, by imposing an external magnetic field, the alloy has a higher chance to maintain or even improve its ordered disposition. The samples are then treated with an inorganic acid solution, preferably a hydrohalic acid, having a concentration between 0.01- 1.5 M, for reaction time between 5 and 30 min. Afterwards, the alloy can be subjected to an aqueous solvent wash, preferably with distilled water. The aim for this last step is to increase the percentage of L1 phase present by selective removal of the less oxidation resistant A1 phase. Synthesis of precursors (not according to the invention) Nickel Nitroprusside. All necessary chemicals were supplied by Sigma-Aldrich and used without further purification. NiFe(CN)NO∙xHO crystals with Fe and Ni in the stoichiometric ratio 1: 1 are obtained starting from aqueous solutions of sodium hydrate nitroprusside – NaFe(CN)NO ∙xHO – and nickel nitrate hydrate – Ni(NO)∙xHO –. In a typical synthesis, 3 mM of Ni(NO)∙xHO and 3 mM of NaFe(CN)5NO∙xHO are weighed and the two separated solutions are prepared with 50 ml of distilled water. Ni(NO)∙xHO solution is slowly added to 50 ml of 3mM sodium nitroprusside solution while stirring. The reaction immediately begins, leading to the formation of a grey NiFe(CN)NO ∙xHO precipitate, according to the reaction: 2NaNO + 3HO The precipitate is washed abundantly with distilled water by centrifugation to eliminate the sodium nitrate formed; then, it is dried under vacuum at room temperature. The final dried product is grey. The yield is 85%. Iron Tetracyano-nickelate. All necessary chemicals were supplied by Sigma-Aldrich and used without further purification. Crystalline FeNi(CN)∙xHO powders with Fe and Ni in the stoichiometric ratio 1:1 are obtained through an exchange reaction between Ammonium Iron (II) sulfate hexa-hydrate – (NH)Fe(SO)∙6HO – (Mohr's salt) and Potassium tetra- cyanonickelate (II) Hydrate – KNi(CN)∙xHO –. In a typical synthesis, 3 mM of Mohr's salt and 3 mM of KNi(CN)∙xHO are weighed and two separate solutions are prepared with 50 ml of distilled water. Mohr's salt solution is dropwise added to the KNi(CN)xHO solution or vice versa. The reaction starts immediately leading to the formation of a yellow Fe(H2O)Ni(CN)∙xHO precipitate, according to the reaction: The precipitate is washed abundantly with distilled water by centrifugation, to eliminate residual sulphates. The yellow- orange precipitate is separated and dried under vacuum at room temperature. The final dried product is light orange. The yield is 90%. With the procedures described above it is also possible to obtain ternary or quaternary alloys simply by varying the stoichiometry and the nature of the countercations in the precursor salts. For example, Co Ni Fe can be obtained as a co- precipitation product from the reaction: with different metals (e.g., Co Ni Fe, CoNi, CoFe, CuNi, CdNi) to be used in many fields such as catalysis, medicine (hyperthermia), etc. EXAMPLE I SYNTHESIS OF A L1 Fe-Ni ALLOY FROM Nickel Nitroprusside (NiFe(CN)NO∙xHO) 200 mgs of Nickel Nitroprusside (NiFe(CN)NO∙xHO) are placed in a tubular furnace. The reduction process takes place under H atmosphere and under incremental temperature, starting from room temperature to 290°C, and increasing by 1°C/min. The reduction reaction proceeds for 24h. H flow rate was initially 0.1 L/min, and it became 0.01 L/min under steady conditions. When the reduction reaction is finished, the flow of H is stopped and the tubular furnace is turned off. When the furnace has reached room temperature, the product is taken out and washed with 0.1M HCl for 15 minutes and then washed with distilled water. The process disclosed by this invention is a cost-effective process, high yield and low environmental impact. EXAMPLE II SYNTHESIS OF A L1 Fe-Ni ALLOY FROM Iron tetracyanonickelate (Fe(HO)Ni(CN)∙xHO) 200 mgs of Iron tetracyanonickelate (Fe(HO)Ni(CN)x4HO) are placed in a tubular furnace. The reduction process takes place under H atmosphere and under incremental temperature, starting from room temperature to 270°C, and increasing by 0.5°C/min. The reduction reaction proceeds for 24h. H flow rate was initially 0.15 L/min and it became 0.01 L/min under steady conditions. When the reduction reaction is finished, the flow of H is stopped and the tubular furnace is turned off. When the furnace has reached room temperature, the product is taken out and washed with 0.15 M HCl for 20 minutes and then washed with distilled water. The process according to this invention is a cost-effective process, high yield, and low environmental impact. Anyway, it is understood that the invention should not be considered limited to the particular embodiments illustrated above, which make up only exemplary embodiments thereof, but that more variants are possible, all under reach of the skilled person, without departing from the scope of the invention itself, as de- fined by the appended claims. EXPERIMENTAL Powder X-ray diffraction (XRD) measurements were performed on a Seifert 3003 TT diffractometer equipped with a secondary graphite monochromator and using Cu kα radiation, (l = 1.5418 Å) to investigate the crystallographic structure. Rietveld refine- ments of the experimental data were performed using MAUD soft- ware51. Field-dependent magnetic properties were investigated by us- ing a commercial vibrating sample magnetometer (MicroSense Model 10) equipped with an electromagnet that can supply a maximum mag- netic field of 2T. Transmission electron microscopy (TEM) images were obtained using a JEOL JEM 1400 Plus (Jeol Ltd., Tokyo, Japan) operating at 120 kV and equipped with a EDX detector. The specimens were pre- pared by dropping an octane dispersion of the samples on a 200- mesh carbon-coated copper grid. Mössbauer spectrometry measurements were performed at 25°C, with the aid of a Co/Rh gammaray source in a transmission scheme with a triangular velocity form. The hyperfine structure was mod- eled by means of a least-squares fitting procedure involving Zee- man sextets composed of Lorentzian lines. To describe the broad- ening of lines, several magnetic subcomponents (hyperfine magnetic field distribution) have been considered. The isomer shift (IS) values were referred to that of α-Fe at 25°C. REFERENCES 1. [1] Mandal, S.; Debata, M; Sengupta, P; Basu, S. Critical Reviews in Solid State and Materials Sciences. 2022, https://doi.org/10.1080/10408436.2022.2107484 2. Néel, L.; Pauleve, J.; Pauthenet, R.; Laugier, J.; Dautreppe, D. Magnetic Properties of an Iron-Nickel Single Crystal Ordered by Neutron Bombardment. J. Appl. Phys. 1964, 35, 873–876. https://doi.org/10.1063/1.1713516. 3. Lima, E.; Drago, V. A New Process to Produce Ordered Fe50Ni50 Tetrataenite. Phys. Status Solidi Appl. Res. 2001, 187, 119–124. https://doi.org/10.1002/1521-396X(200109)187:1<119::AID- PSSA119>3.0.CO;2-L. 4. Hayashi, Y.; Gotou, S.; Mizuguchi, M.; Kotsugi, M.; Kitou, Y.; Okuno, E.; Takanashi, K. Synthesis and Characterization of L10-FeNi Powders. J. Magn. Soc. Japan 2013, 37, 198–201. https://doi.org/10.3379/msjmag.1303r019. 5. Lee, S.; Edalati, K.; Iwaoka, H.; Horita, Z.; Ohtsuki, T.; Ohkochi, T.; Kotsugi, M.; Kojima, T.; Mizuguchi, M.; Takanashi, K. Formation of FeNi with L10-Ordered Structure Using High-Pres- sure Torsion. Philos. Mag. Lett. 2014, 94 (10), 639–646. https://doi.org/10.1080/09500839.2014.955546. 6. Geng, Y.; Ablekim, T.; Koten, M. A.; Weber, M.; Lynn, K.; Shield, J. E. Defect Generation and Analysis in Mechanically Al- loyed Stoichiometric Fe-Ni Alloys. J. Alloys Compd. 2015, 633, 250–255. https://doi.org/10.1016/j.jallcom.2015.02.038. 7. Makino, A.; Sharma, P.; Sato, K.; Takeuchi, A.; Zhang, Y.; Takenaka, K. Artificially Produced Rare-Earth Free Cosmic Magnet. Sci. Rep. 2015, 5, 16627 (1-8). https://doi.org/10.1038/srep16627. 8. Gong, M.; Ren, S. Phase Transformation-Driven Surface Recon- struction of FeNi Nanostructures. Chem. Mater. 2015, 27, 7795−7800. https://doi.org/10.1021/acs.chemmater.5b03736. 9. Kojima, T.; Ogiwara, M.; Mizuguchi, M.; Kotsugi, M.; Kogan- ezawa, T.; Ohtsuki, T.; Tashiro, T. Y.; Takanashi, K. Fe-Ni Com- position Dependence of Magnetic Anisotropy in Artificially Fabri- cated L10-Ordered FeNi Films. J. Phys. Condens. Matter 2014, 26, 064207 (1-10). https://doi.org/10.1088/0953-8984/26/6/064207. 10. Takanashi, K.; Mizuguchi, M.; Kojima, T.; Tashiro, T. Fab- rication and Characterization of L10-Ordered FeNi Thin Films. J. Phys. D. Appl. Phys. 2017, 50, 483002 (1-9). https://doi.org/10.1088/1361-6463/aa8ff6. 11. Goto, S.; Kura, H.; Watanabe, E.; Hayashi, Y.; Yanagihara, H.; Shimada, Y.; Mizuguchi, M.; Takanashi, K.; Kita, E. Synthesis of Single-Phase L10-FeNi Magnet Powder by Nitrogen Insertion and Topotactic Extraction. Sci. Rep. 2017, 7, 13216 (1-7). https://doi.org/10.1038/s41598-017-13562-2. 12. G. Varvaro, P. Imperatori, S. Laureti, C. Cannas, A. Ardu, P. Plescia, A. Capobianchi, Synthesis of L10 alloy nanopar- ticles. Potential and versatility of the pre-ordered Precursor Reduction strategy, J. Alloy. Compd. 846 (2020) 13. K. Mibu, T. Kojima, M. Mizuguchi K. Takanashi. Local structure and magnetism of L10-type FeNi alloy films with perpen- dicular magnetic anisotropy studied through Fe nuclear probes. J. Phys. D: Appl. Phys. 48 (2015) 205002)

Claims

CLAIMS 1) A process for the preparation of a bimetallic/ trimetallic alloy having chemically ordered crystal structure, characterized in that it involves the following steps: reduction of a cyano/ni- trosyl-metal complex salt having alternating mononuclear layers in a reactor through a reducing agent, supplying heat to the reactor, with a heating rate equal or greater than 0.1 °C/min, at a temperature between 200°C-450°C, for a reaction time between 12 and 72 hours. 2) A process as in claim 1, wherein said chemically ordered structure is a L1 structure or a body centered tetragonal (BCT) structure. 3) A process as in claim 1 or 2, wherein the cyano/nitrosyl- metal complex salt is a Transition Bivalent Metal (TBM) tetracy- anonickelate (TBM(HO)Ni(CN)x4HO) or a Transition Bivalent Metal Nitroprusside (TBMFe(CN)NO)x5HO)); and said Transition Bivalent Metal (TBM) is chosen from the following list: Co, Cu, Cd, Mn. 4) A process as in 1), characterized in that the cyano/ni- trosyl-metal complex salt is Nickel Nitroprusside (NiFe(CN)NO)x5HO). 5) A process as in 1), characterized in that the cyano/ni- trosyl-metal complex salt is Iron tetracyanonickelate (Fe(HO)Ni(CN)x4HO). 6) A process as in 1) or 4), characterized in that the cyano/nitrosyl-metal complex salt is Iron tetracyanonickelate (Fe(HO)Ni(CN)x4HO) and in that the heating rate is 0.1 or more °C/min and in that the temperature is between 200 °C-450 °C or more and in that the reducing agent is H or other reductive agents chosen from NaBH, CaH. 7) A process as in 1) or 3), characterized in that cy- ano/nitrosyl-metal complex salt is Nickel Nitroprusside (NiFe(CN)NO)x5HO) and in that the heating rate is 0.1 or more °C/min and in that the temperature is between 200 °C-450 °C and in that the reducing agent is H or other reductive agents chosen from NaBH, CaH 8) Use of a bimetallic alloy having an L1 ordered crystal structure, being composed by the metals Fe and Ni, wherein the L1 alloy has a coercivity up to 66 mT and a saturation magnetization higher than 130-140 KAm/kg. FeNi L1 alloy could be used as a starting material in a preparation of a permanent magnet after a process improvement.