Application of a Dy3Co0.6Cu0.4Hx Addition for Controlling the Microstructure and Magnetic Properties of Sintered Nd-Fe-B Magnets

The focus of new technologies on the formation of inhomogeneous distributions of heavy rare-earth metals (REMs) in hard magnetic Nd–Fe–B materials is of scientific importance to increase their functional properties, along with preserving existing sources of heavy REMs. This paper focused on the coercivity enhancement of Nd2Fe14B-based magnets by optimizing the microstructure, which includes the processes of grain boundary structuring via the application of a Dy3Co0.6Cu0.4Hx alloy added to the initial Nd–Fe–B-based powder mixtures in the course of their mechanical activation. We have studied the role of alloying elements in the formation of phase composition, microstructure, the fine structure of grains, and the hysteretic properties of hard magnetic Nd(R)2Fe14B-based materials. It was shown that the Dy introduction via the two-component blending process (the hydrogenated Dy3Co0.6Cu0.4 compound is added to a powder mixture) resulted in the formation of the core-shell structure of 2–14–1 phase grains. The efficient improvement of the coercivity of Nd(RE)–Fe–B magnets, with a slight sacrifice of remanence, was demonstrated.


Introduction
Researchers have made many attempts to reduce the heavy rare-earth (RE) consumption Nd-Fe-B sintered magnets with high-coercivity. Some progress has been achieved using Dy and/or Tb in various forms to realize approaches named grain boundary diffusion (GBD) [1][2][3] and grain boundary structuring (GBS) [4][5][6][7][8]. The application of binary mixtures allows one to improve the structure of the boundary phases and grain boundaries of the main magnetic phase and to realize the diffusion of a required component of the alloy directly through the boundaries. It has been demonstrated that by controlling the process time and temperature of GBD processes, the coercivity of the magnet can be greatly enhanced, without sacrificing the remanence. It was shown in our previous studies that hydrogenated Tb and Dy additions allowed us to enhance the coercivity with a slight decrease in the remanence [9] and increase the stability of the magnet properties during annealing at the low-temperature [10], respectively.
The grain boundary restructuring, with rare-earth-rich low-melting compounds added to low-alloyed Nd-Fe-B-based compositions in the course of technological processing, was realized when using (Pr,Nd) 6 Fe 13 Cu [4], Dy 32.5 Fe 62 Cu 5.5 [5], Dy 69 Ni 31 [6], Dy 88 Mn 12 (wt.%) [11], Pr 34.4 Co 65.6 (wt.%) [12], and Dy 82.3 Co 17.7 (wt.%) [13], which is a low-melting eutectic composition. It was shown that the intrinsic coercivity evidently increased when using Dy 82.3 Co 17.7 and the maximum intrinsic coercivity was achieved when its content was 2 wt.%. At the same time, the remanence and maximum-energy product decreased slightly as the Dy 82.3 Co 17.7 content increased. By adding a small amount of Dy 82.3 Co 17.7 , the coercivity improved greatly, and the irreversible loss decreased sharply. The increase in the Curie temperature of the magnets suggests that Co atoms have been incorporated into the 2:14:1 main phase. A well-developed a core-shell structure is formed in these magnets.
The experiments with REM-M-H compounds (rare earth metal-transition metal(s)-hydrogen), which are added at the stage of mechanical milling and alloying, were performed to realize the optimum microstructure, nano-heterogeneous distribution of heavy REMs (Dy or Tb) within a grain, and economically alloyed composition of magnets, which assumes, in particular, the distribution of heavy REMs within the near-grain boundary areas. Such a heavy-REM distribution allows us to (1) locally increase the coercive force and decrease the probability of the formation of reverse domains at grain boundaries; (2) limit the substitution of heavy REM for neodymium in the matrix phase and, thus, decrease the probability of decreasing magnetization and remanence; and (3) decrease the amount of heavy REMs, which is required to reach the given increase in the coercive force. The latter circumstance determines the possibility of the development of physico-chemical and technological foundations of resource-saving technology, the possibility of decreasing the material costs and prices of products manufactured from the new alloys, and the possibility of substantially widening the functionality of the materials.
Thus, by applying compositions with a heavy rare-earth metal, the outer region of the Nd 2 Fe 14 B matrix grains was enriched during the sintering process and substitutes for Nd were used in the matrix grains to form the (Nd,Dy) 2 Fe 14 B core-shell phase.
This paper focused on optimizing the microstructure of the near-stoichiometric Nd 2 Fe 14 B-based magnet, which included the grain boundary diffusion and grain boundary structuring processes via the application of a hydrogenated Dy 3 Co 0.6 Cu 0.4 H x composition added to a powder mixture.

Experimental
The strip casting technique was used for the preparation of the base Nd-24.0, Pr-6.5, Dy-0.5, B-1.0, Al-0.2, Fe-balance alloy (wt.%). The strip-cast alloy was subsequently subjected to hydrogen decrepitation process, which was realized during heating to 270 • C in a hydrogen flow at a pressure of 0.1 MPa and holding at this temperature for 1 h.
The Dy 3 (Co 1−x Cu x ) alloy with x = 0.4 was produced by the arc melting of the starting components (distilled Dy of 99.9% purity, Co of ≥ 99.25% purity, and oxygen-free Cu of 99.95% purity) in an argon atmosphere using a water-cooled copper bottom and a non-consumable tungsten electrode. The ingot was homogenized at 600 • C for 90 h and subjected to hydrogenation under conditions used for the strip-casting alloy, namely, upon heating to 270 • C in a hydrogen flow at a pressure of 0.1 MPa and subsequent 1 h heating at this temperature (Regime 1 was used to manufacture the magnet), and upon heating to 700 • C in a high-purity hydrogen atmosphere and holding at this temperature for 1 h in a glass Sieverts-type apparatus (Regime 2 was used for investigations). In the case of heating at 700 • C, the hydrogenation up to the Dy 3 Co 0.6 Cu 0.4 H x composition with x = 8.26 was realized. It is expected that such a hydrogen content accords with the complete hydrogenation of dysprosium to a dysprosium hydride.
The mixture of hydrogen-decrepitated strip-cast Nd(RE)-Fe-B alloy and the Dy 3 Co 0.6 Cu 0.4 H x alloy (Regime 1) was milled for 40 min to an average particle size of 3 µm using a vibratory mill and isopropyl alcohol medium. After wet pressing of the pulp in a transverse magnetic field of 1500 kA/m, compacts were sintered at 1080 • C for 2 h and optimally heat treated (HT) at 500 • C for 2 h. Then, samples of the magnet were subjected to low-temperature heat treatment in the temperature range 400-900 • C, with subsequent quenching in N 2 .
The phase composition of the Dy 3 Co 0.6 Cu 0.4 and Dy 3 Co 0.6 Cu 0.4 H x (x = 8.26) alloys was investigated by X-ray diffraction (XRD) analysis using an Ultima IV (Rigaku») diffractometer (equipped with a "D/teX" detector, CuKα radiation) and a Philips X'Pert 1 diffractometer, respectively; the scanning step was 0.001 • . X-ray diffraction patterns were processed, and the phase composition of the alloy was determined using PowderCell software. Data on the crystal structure type, lattice parameters, and the crystallographic positions of atoms in the Dy-Co, Dy-Cu, and H-Dy systems [14][15][16] were used to simulate theoretical XRD patterns.
An Quanta 450 FEG high-resolution field emission gun scanning electron microscope (FEI Company, Fremont, USA) equipped with an energy-dispersive spectroscopy (EDS, EDAX Inc., Mahwah, USA) microprobe was used to investigate the structure, chemical composition, and distribution of magnet components (X-ray mapping) of the addition and magnet sample. The mean particle size was evaluated by means of a MasterSizer 3000 laser diffraction particle size analyzer (Malvern Panalytical Ltd, Malvern, United Kingdom). The hysteretic properties of the magnet sample were measured at room temperature (RT) using an automatic hysteresis graph system MH-50 (Walker Scientific Inc., Worcester, USA). The differential thermal analysis (DTA) and thermogravimetric analysis were performed under an argon atmosphere with a heating/cooling rate of 30 • C/min using a STA 449 F3 Jupiter installation (Netzsch Holding, Selb, Germany). Figure 1 shows the X-ray diffraction pattern of the Dy 3 Co 0.6 Cu 0.4 alloy subjected to prolonged annealing in an argon atmosphere. The reflections belong to the main Dy 3 (Co,Cu) phase and the Dy(Cu,Co) phase based on DyCu [14,15]. The analysis of the crystal structures of the found compounds and theoretical XRD patterns constructed for the simulated structures allowed us to determine variations in the lattice parameters of the Dy(Cu 1−y Co y ) and Dy 3 (Co 1−x Cu x ) phases alloyed with Co and Cu, respectively (see Table 1). As seen, the alloying of the binary compounds with Co and Cu did not change the crystal structure type of the compounds. In accordance with the binary phase diagrams [14,15], the phases present in the alloy are alloyed compositions of the binary compounds. compacts were sintered at 1080 °C for 2 h and optimally heat treated (HT) at 500 °C for 2 h. Then, samples of the magnet were subjected to low-temperature heat treatment in the temperature range 400-900 °C, with subsequent quenching in N2.

X-Ray Diffraction Analysis
The phase composition of the Dy3Co0.6Cu0.4 and Dy3Co0.6Cu0.4Hx (x = 8.26) alloys was investigated by x-ray diffraction (XRD) analysis using an Ultima IV (Rigaku») diffractometer (equipped with a "D/teX" detector, CuKα radiation) and a Philips X'Pert 1 diffractometer, respectively; the scanning step was 0.001°. X-ray diffraction patterns were processed, and the phase composition of the alloy was determined using PowderCell software. Data on the crystal structure type, lattice parameters, and the crystallographic positions of atoms in the Dy-Co, Dy-Cu, and H-Dy systems [14][15][16] were used to simulate theoretical XRD patterns.
An Quanta 450 FEG high-resolution field emission gun scanning electron microscope (FEI Company, Fremont, USA) equipped with an energy-dispersive spectroscopy (EDS, EDAX Inc., Mahwah, USA) microprobe was used to investigate the structure, chemical composition, and distribution of magnet components (x-ray mapping) of the addition and magnet sample. The mean particle size was evaluated by means of a MasterSizer 3000 laser diffraction particle size analyzer (Malvern Panalytical Ltd, Malvern, United Kingdom). The hysteretic properties of the magnet sample were measured at room temperature (RT) using an automatic hysteresis graph system MH-50 (Walker Scientific Inc., Worcester, USA). The differential thermal analysis (DTA) and thermogravimetric analysis were performed under an argon atmosphere with a heating/cooling rate of 30 °C/min using a STA 449 F3 Jupiter installation (Netzsch Holding, Selb, Germany). Figure 1 shows the x-ray diffraction pattern of the Dy3Co0.6Cu0.4 alloy subjected to prolonged annealing in an argon atmosphere. The reflections belong to the main Dy3(Co,Cu) phase and the Dy(Cu,Co) phase based on DyCu [14,15]. The analysis of the crystal structures of the found compounds and theoretical XRD patterns constructed for the simulated structures allowed us to determine variations in the lattice parameters of the Dy(Cu1−yCoy) and Dy3(Co1−xCux) phases alloyed with Co and Cu, respectively (see Table 1). As seen, the alloying of the binary compounds with Co and Cu did not change the crystal structure type of the compounds. In accordance with the binary phase diagrams [14,15], the phases present in the alloy are alloyed compositions of the binary compounds.    The phase composition of the alloy was also confirmed by the EDS microanalysis, see Figure 2 and Table 2. The microstructure consisted of Dy 3 (Co 1−x Cu x ) (x~0.4) dendrites (point 1 in Figure 2) and Dy(Cu 1−y Co y ) + Dy 3 (Co 0.6 Cu 0.4 ) mixture (point 2 in Figure 2) found in the interdendritic regions. The composition of the Dy(Cu 1−y Co y ) phase cannot be accurately determined by the EDS analysis because of its small size, since the surrounding matrix is analyzed along with this very small inclusion. However, the increased content of copper is evident in this mix area. The phase composition of the alloy was also confirmed by the EDS microanalysis, see Figure 2 and Table 2. The microstructure consisted of Dy3(Co1−xCux) (x ∼ 0.4) dendrites (point 1 in Figure 2) and Dy(Cu1−yCoy) + Dy3(Co0.6Cu0.4) mixture (point 2 in Figure 2) found in the interdendritic regions. The composition of the Dy(Cu1−yCoy) phase cannot be accurately determined by the EDS analysis because of its small size, since the surrounding matrix is analyzed along with this very small inclusion. However, the increased content of copper is evident in this mix area.  As is shown in Table 1, the substitution of Cu for Co in Dy3(Co1−xCux) (with regard to the solubility of Cu and Co in Dy3Co and DyCu, respectively) changed the lattice parameters: the lattice parameters b and c increased as the radius of Cu atoms (0.128 nm) was higher than that of the Co atoms (0.125 nm), whereas the lattice parameter a decreased. This is likely to be due to the fact that copper atoms substitute for cobalt atoms only at certain sites.

X-Ray Diffraction Analysis
We assumed that the solidification of the alloy occurs via the primary formation of the Dy3Cobased phase by peritectic reaction; the DyCu-based compound is the secondary phase. According to the Co-Dy phase diagram, the solidification path may include the formation of the Dy12Co7-based phase by peritectic reaction.

Interaction of Dy3(Co,Cu) Alloy with Hydrogen
The saturation of the Dy3Co0.6Cu0.4 alloy with hydrogen led to the embrittlement of the alloy (i.e., the powder material suitable for further introduction of the composition into the Nd-Fe-B magnetic alloy powder during cooperative milling was obtained). Figure 3a shows the x-ray diffraction analysis data for the Dy3Co0.6Cu0.4 alloy subjected to hydrogenation (Regime 2). The hydrogenated composition contained DyH2 [17] and DyH3 [18] hydrides. Other reflections corresponded to the Dy3(Co,Cu) phase; it is likely that small quantities of the Dy3(Co,Cu) and Dy(Cu,Co) phases did not react with hydrogen. After hydrogenation, copper and cobalt may be present in the form of a fine mixture.  As is shown in Table 1, the substitution of Cu for Co in Dy 3 (Co 1−x Cu x ) (with regard to the solubility of Cu and Co in Dy 3 Co and DyCu, respectively) changed the lattice parameters: the lattice parameters b and c increased as the radius of Cu atoms (0.128 nm) was higher than that of the Co atoms (0.125 nm), whereas the lattice parameter a decreased. This is likely to be due to the fact that copper atoms substitute for cobalt atoms only at certain sites.
We assumed that the solidification of the alloy occurs via the primary formation of the Dy 3 Co-based phase by peritectic reaction; the DyCu-based compound is the secondary phase. According to the Co-Dy phase diagram, the solidification path may include the formation of the Dy 12 Co 7 -based phase by peritectic reaction.

Interaction of Dy 3 (Co , Cu) Alloy with Hydrogen
The saturation of the Dy 3 Co 0.6 Cu 0.4 alloy with hydrogen led to the embrittlement of the alloy (i.e., the powder material suitable for further introduction of the composition into the Nd-Fe-B magnetic alloy powder during cooperative milling was obtained). Figure 3a shows the X-ray diffraction analysis data for the Dy 3 Co 0.6 Cu 0.4 alloy subjected to hydrogenation (Regime 2). The hydrogenated composition contained DyH 2 [17] and DyH 3 [18] hydrides. Other reflections corresponded to the Dy 3 (Co,Cu) phase; it is likely that small quantities of the Dy 3 (Co,Cu) and Dy(Cu,Co) phases did not react with hydrogen. After hydrogenation, copper and cobalt may be present in the form of a fine mixture.  Figure 3b shows the x-ray diffraction analysis data of the alloy Dy3Co0.6Cu0.4Hx subjected to thermal dehydrogenation (upon heating during DTA). The sample was heated up to 700 °C ( Figure  3). After heating, the presence of DyH2 and small quantities of the Dy3(Co,Cu) and Dy(Cu,Co) phases were detected; DyH3 was absent. The presence of a thin mechanical mixture of Cu and Co is also possible.
According to the DTA data (Figure 4), the decomposition of DyH3 started at a temperature of ~314 °C, which agreed with the literature data [16]. Between ~314 °C and ~690 °C, no thermal effects were identified. Above ~690 °C, in accordance with the Dy-H [16] diagram, the solid solution of hydrogen in dysprosium decomposed to form dysprosium. However, the thermal effects at temperatures above 600 °C can correspond to the melting of one of the metallic phases of the alloy; nevertheless, the thermal effect corresponding to ~690 °C is accompanied by a significant weight loss. The observed formation of Dy hydrides indicates the possibility of the hydrogenated Dy3Co0.6Cu0.4 alloy to be used as additions in manufacturing sintered Nd-Fe-B magnets.

Microstructure and Electron Microprobe Analysis of Sintered NdFeB-Based Magnet
In accordance with the microprobe analysis data shown in Table 3, the microstructure of a magnet prepared from the powder mixture with 2 wt.% Dy3Co0.6Cu0.4Hx (Regime 1) was characterized by the presence of four structural components differing in the chemical composition, see Figure 5 (the phases are indicated by red numbers).  Figure 3b shows the X-ray diffraction analysis data of the alloy Dy 3 Co 0.6 Cu 0.4 H x subjected to thermal dehydrogenation (upon heating during DTA). The sample was heated up to 700 • C (Figure 3). After heating, the presence of DyH 2 and small quantities of the Dy 3 (Co,Cu) and Dy(Cu,Co) phases were detected; DyH 3 was absent. The presence of a thin mechanical mixture of Cu and Co is also possible.
According to the DTA data (Figure 4), the decomposition of DyH 3 started at a temperature of~314 • C, which agreed with the literature data [16]. Between~314 • C and~690 • C, no thermal effects were identified. Above~690 • C, in accordance with the Dy-H [16] diagram, the solid solution of hydrogen in dysprosium decomposed to form dysprosium. However, the thermal effects at temperatures above 600 • C can correspond to the melting of one of the metallic phases of the alloy; nevertheless, the thermal effect corresponding to~690 • C is accompanied by a significant weight loss. The observed formation of Dy hydrides indicates the possibility of the hydrogenated Dy 3 Co 0.6 Cu 0.4 alloy to be used as additions in manufacturing sintered Nd-Fe-B magnets.  Figure 3b shows the x-ray diffraction analysis data of the alloy Dy3Co0.6Cu0.4Hx subjected to thermal dehydrogenation (upon heating during DTA). The sample was heated up to 700 °C ( Figure  3). After heating, the presence of DyH2 and small quantities of the Dy3(Co,Cu) and Dy(Cu,Co) phases were detected; DyH3 was absent. The presence of a thin mechanical mixture of Cu and Co is also possible.
According to the DTA data (Figure 4), the decomposition of DyH3 started at a temperature of ~314 °C, which agreed with the literature data [16]. Between ~314 °C and ~690 °C, no thermal effects were identified. Above ~690 °C, in accordance with the Dy-H [16] diagram, the solid solution of hydrogen in dysprosium decomposed to form dysprosium. However, the thermal effects at temperatures above 600 °C can correspond to the melting of one of the metallic phases of the alloy; nevertheless, the thermal effect corresponding to ~690 °C is accompanied by a significant weight loss. The observed formation of Dy hydrides indicates the possibility of the hydrogenated Dy3Co0.6Cu0.4 alloy to be used as additions in manufacturing sintered Nd-Fe-B magnets.

Microstructure and Electron Microprobe Analysis of Sintered NdFeB-Based Magnet
In accordance with the microprobe analysis data shown in Table 3, the microstructure of a magnet prepared from the powder mixture with 2 wt.% Dy3Co0.6Cu0.4Hx (Regime 1) was characterized by the presence of four structural components differing in the chemical composition, see Figure 5 (the phases are indicated by red numbers).

Microstructure and Electron Microprobe Analysis of Sintered NdFeB-Based Magnet
In accordance with the microprobe analysis data shown in Table 3, the microstructure of a magnet prepared from the powder mixture with 2 wt.% Dy 3 Co 0.6 Cu 0.4 H x (Regime 1) was characterized by the presence of four structural components differing in the chemical composition, see Figure 5 (the phases are indicated by red numbers).    Table 2).
The chemical composition of matrix grains (Phase 1 in Figure 5a) was close to the stoichiometric (Nd,R)2Fe14B composition. The presence of Dy in the matrix alloy did not allow us to unambiguously conclude the formation of the core-shell structure, but the presence of cobalt in 2:14:1 phase grains demonstrates the possibility of micro-alloying through the use of hydrogenated low-melting Cocontaining compounds (the melting temperature was lower than the sintering temperature of Nd-Fe-B magnets). The Nd-rich phase (Phase 2 in Figure 5a) was characterized by a variable composition. Phase 3 (Figure 5a) corresponded to the oxide phases. In accordance with the literature data [19,20], they may be based on NdO, Nd2O3, or NdO2. The presence of a phase based on Fe-Nb in triple junctions (TJ) was observed (Phase 4, Figure 5b). This fact may be related to impurities in the industrially prepared alloy matrix.
The distribution of rare earth elements, Co and Cu in the matrix grains, and in the intergranular Nd-rich phases (phase 2 in Figure 5a) in the sintered magnets prepared from the powder mixture with 2 wt.% of Dy3Co0.6Cu0.4Hx addition was also investigated by x-ray mapping (see Figure 6). The nonuniform Dy distribution within the 2:14:1 phase grains could be observed. The depletion of triple junctions of Co and their enrichment in Cu should be noted in the case of the addition of Dy3Co0.6Cu0.4Hx. The presence of reactive Dy powder (originating from DyH2 that was decomposed during sintering) ensures the diffusion of Dy atoms to the 2:14:1 phase lattice, since the atomic radius of Dy atoms is lower than that of Nd atoms. This led to ousting Nd atoms to peripheral areas. The diffusion coefficient of Nd atoms is lower than that of Dy atoms [21]; thus, the diffusion of Dy is more significant. Such an inequality of diffusion flows of atoms caused lattice stresses and resulted in the inhomogeneous Dy and Nd(Pr) distribution over the 2:14:1 phase grains. The core-shell structure (Dy-enriched shell and Dy-depleted core) is evident in Figure 6.  Table 2).
The chemical composition of matrix grains (Phase 1 in Figure 5a) was close to the stoichiometric (Nd,R) 2 Fe 14 B composition. The presence of Dy in the matrix alloy did not allow us to unambiguously conclude the formation of the core-shell structure, but the presence of cobalt in 2:14:1 phase grains demonstrates the possibility of micro-alloying through the use of hydrogenated low-melting Co-containing compounds (the melting temperature was lower than the sintering temperature of Nd-Fe-B magnets). The Nd-rich phase (Phase 2 in Figure 5a) was characterized by a variable composition. Phase 3 (Figure 5a) corresponded to the oxide phases. In accordance with the literature data [19,20], they may be based on NdO, Nd 2 O 3 , or NdO 2 . The presence of a phase based on Fe-Nb in triple junctions (TJ) was observed (Phase 4, Figure 5b). This fact may be related to impurities in the industrially prepared alloy matrix.
The distribution of rare earth elements, Co and Cu in the matrix grains, and in the intergranular Nd-rich phases (phase 2 in Figure 5a) in the sintered magnets prepared from the powder mixture with 2 wt.% of Dy 3 Co 0.6 Cu 0.4 H x addition was also investigated by X-ray mapping (see Figure 6). The nonuniform Dy distribution within the 2:14:1 phase grains could be observed. The depletion of triple junctions of Co and their enrichment in Cu should be noted in the case of the addition of Dy 3 Co 0.6 Cu 0.4 H x . The presence of reactive Dy powder (originating from DyH 2 that was decomposed during sintering) ensures the diffusion of Dy atoms to the 2:14:1 phase lattice, since the atomic radius of Dy atoms is lower than that of Nd atoms. This led to ousting Nd atoms to peripheral areas.
The diffusion coefficient of Nd atoms is lower than that of Dy atoms [21]; thus, the diffusion of Dy is more significant. Such an inequality of diffusion flows of atoms caused lattice stresses and resulted in the inhomogeneous Dy and Nd(Pr) distribution over the 2:14:1 phase grains. The core-shell structure (Dy-enriched shell and Dy-depleted core) is evident in Figure 6.  The other components of the Dy3Co0.6Cu0.4Hx composition (i.e., Cu and Co) are also useful additions for Nd-Fe-B-based magnets. It is evident from Figures 6 and 7 that Co evinced the tendency to incorporate the 2:14:1 phase grains, while the Cu enriched triple junction phases. The role of Cu in Figure 6. Co, Cu, and Dy mapping in 2:14:1 phase grains and triple junction phases of the Nd-Fe-B sintered magnet prepared from the powder mixture with 2 wt.% Dy 3 (Co,Cu). The red circle indicates the depletion of 2:14:1 phase grain in Dy (i.e. the formation of core-shell structure).  The other components of the Dy3Co0.6Cu0.4Hx composition (i.e., Cu and Co) are also useful additions for Nd-Fe-B-based magnets. It is evident from Figures 6 and 7 that Co evinced the tendency to incorporate the 2:14:1 phase grains, while the Cu enriched triple junction phases. The role of Cu in The other components of the Dy 3 Co 0.6 Cu 0.4 H x composition (i.e., Cu and Co) are also useful additions for Nd-Fe-B-based magnets. It is evident from Figures 6 and 7 that Co evinced the tendency to incorporate the 2:14:1 phase grains, while the Cu enriched triple junction phases. The role of Cu in the grain-boundary restructuring and positive effects of Co on the coercivities of Nd-Fe-B magnets were reported in our previous work [22] and were also considered in [23][24][25][26][27][28][29][30][31].

Dependence of the Coercive Force ( j H c ) on the Heat Treatment Temperature
The magnetic properties ( j H c ) of the magnets (see Table 4 and Figure 8) prepared with the hydrogenated Dy 3 Co 0.6 Cu 0.4 alloy were lower than those in the case of the application of the addition of the DyH 2 [31]. One of the causes is the incomplete hydrogenation of the alloy (see Figure 3, XRD data) and, therefore, the incomplete occurrence of the grain boundary diffusion of the available Dy. The small quantity of the Dy 3 (Co,Cu) phase present in the Dy 3 Co 0.6 Cu 0.4 alloy was subjected to hydrogenation. However, the value of B r in the case of Dy 3 Co 0.6 Cu 0.4 H x was higher than that in the case of DyH 2 , which may be due to a difference in the Dy content in the chemical composition of the 2:14:1 phase. The difference in the rare-earth metal and Cu contents in the Nd-rich phases provided a lower value of H k in the case of magnets with 2 wt.% Dy 3 Co 0.6 Cu 0.4 H x . The hysteretic properties of the Nd-Fe-B magnet, without the addition of hydride after optimal HT, are also shown in Table 4 for comparison. the grain-boundary restructuring and positive effects of Co on the coercivities of Nd-Fe-B magnets were reported in our previous work [22] and were also considered in [23][24][25][26][27][28][29][30][31].

Dependence of the Coercive Force (jHc) on the Heat Treatment Temperature
The magnetic properties (jHc) of the magnets (see Table 4 and Figure 8) prepared with the hydrogenated Dy3Co0.6Cu0.4 alloy were lower than those in the case of the application of the addition of the DyH2 [31]. One of the causes is the incomplete hydrogenation of the alloy (see Figure 3, XRD data) and, therefore, the incomplete occurrence of the grain boundary diffusion of the available Dy. The small quantity of the Dy3(Co,Cu) phase present in the Dy3Co0.6Cu0.4 alloy was subjected to hydrogenation. However, the value of Br in the case of Dy3Co0.6Cu0.4Hx was higher than that in the case of DyH2, which may be due to a difference in the Dy content in the chemical composition of the 2:14:1 phase. The difference in the rare-earth metal and Cu contents in the Nd-rich phases provided a lower value of Hk in the case of magnets with 2 wt.% Dy3Co0.6Cu0.4Hx. The hysteretic properties of the Nd-Fe-B magnet, without the addition of hydride after optimal HT, are also shown in Table 4 for comparison. We assumed that the optimal HT for magnets of this type was in the range of 475 to 500 °C, as in the case of the magnets considered in [32][33][34]. Subsequent HT in this temperature range, which is performed after the optimal heat treatment (500 °C), will lead to an increase in the coercive force of magnets with 2 wt.% Dy3Co0.6Cu0.4Hx. Table 4. Hysteretic properties of sintered magnets prepared from the powder mixtures with 2 wt.% Dy3Co0.6Cu0.4Hx and DyH2 and optimally heat treated at 500 °C for 2 h; Br = remanence of magnetic flux density; jHc = coercivity of magnetic polarization; Hk = parameter adopted as a criterion of coercivity (i.e., the magnetic field determined at 0.9 x Br); (BH)max = maximum energy product; HT = heat treatment.  Figure 9 shows the variations of the coercive force (jHc) with changing heat treatment (HT) temperature. As can be seen from the data, after low-temperature HT in a range of 475-500 °C, jHc demonstrated an abrupt increase. We assumed that the optimal HT for magnets of this type was in the range of 475 to 500 • C, as in the case of the magnets considered in [32][33][34]. Subsequent HT in this temperature range, which is performed after the optimal heat treatment (500 • C), will lead to an increase in the coercive force of magnets with 2 wt.% Dy 3 Co 0.6 Cu 0.4 H x . Table 4. Hysteretic properties of sintered magnets prepared from the powder mixtures with 2 wt.% Dy 3 Co 0.6 Cu 0.4 H x and DyH 2 and optimally heat treated at 500 • C for 2 h; B r = remanence of magnetic flux density; j H c = coercivity of magnetic polarization; H k = parameter adopted as a criterion of coercivity (i.e., the magnetic field determined at 0.9 × B r ); (BH) max = maximum energy product; HT = heat treatment.  Figure 9 shows the variations of the coercive force ( j H c ) with changing heat treatment (HT) temperature. As can be seen from the data, after low-temperature HT in a range of 475-500 • C, j H c demonstrated an abrupt increase.

Conclusions
The phase composition of the Dy3Co0.6Cu0.4 alloy in the initial homogenized and hydrogenated states was studied. The alloy in the homogenized state was multiphase and contained the Dy3(Co,Cu) and Dy(Cu,Co) phases. During the hydrogenation of the alloy, the disproportionation or hydrogenolysis process took place, which, regardless of the multiphase composition of the initial alloy, resulted in the formation of DyH2-3 hydride and a fine (Co + Cu) mixture with small trace quantities of Dy3(Co,Cu) and Dy(Cu,Co).
The study of the sintered Nd(RE)-Fe-B magnet prepared from the strip-cast alloy showed that Dy introduction via the two-component blending method (the hydogenated Dy3Co0.6Cu0.4 compound was added to the powder mixture) resulted in the formation of the core-shell structure of 2-14-1 phase grains. The efficient enhancement of the coercivity of Nd(RE)-Fe-B magnets, with a slight sacrifice of remanence, was demonstrated.
The positive effect of REM-alloy hydrogenated additions to the Nd-Fe-B powder mixture allows the possibility of introducing various components to the permanent magnets (heavy REMs, elements structuring grain boundaries, and restricting the magnet grain growth) at the preparation stage, rather than at the alloy-melting stage. This gives the possibility of using a unified initial alloy for the manufacture of magnets with improved (high-coercive or high-performance) magnetic characteristics. Funding: This study was carried out within the project "Development of physico-chemical and engineering foundations for the initiation of innovative resources-economy technology of high-power and high-coercivity (Nd,r)-Fe-B (R = Pr, Tb, Dy, Ho) low-REM permanent magnets", projects No. LTARF18031 funded by the Ministry of Education, Youth and Sports of the Czech Republic and No. 14.616.21.0093 (unique identification number: RFMEFI61618x0093) funded by the Ministry of Science and Higher Education of the Russian Federation. The SEM/EDS investigation and particle size analysis were performed using the research infrastructure of the Regional Materials Science and Technology Centre, VSB -Technical university of Ostrava (Czech Republic), and XRPD analysis, DTA/TG analysis, and the study of magnetic characteristics were carried out at the Center of Collaborative Access for Functional Nanomaterials and High-Purity Substances, Baikov

Conclusions
The phase composition of the Dy 3 Co 0.6 Cu 0.4 alloy in the initial homogenized and hydrogenated states was studied. The alloy in the homogenized state was multiphase and contained the Dy 3 (Co,Cu) and Dy(Cu,Co) phases. During the hydrogenation of the alloy, the disproportionation or hydrogenolysis process took place, which, regardless of the multiphase composition of the initial alloy, resulted in the formation of DyH 2-3 hydride and a fine (Co + Cu) mixture with small trace quantities of Dy 3 (Co,Cu) and Dy(Cu,Co).
The study of the sintered Nd(RE)-Fe-B magnet prepared from the strip-cast alloy showed that Dy introduction via the two-component blending method (the hydogenated Dy 3 Co 0.6 Cu 0.4 compound was added to the powder mixture) resulted in the formation of the core-shell structure of 2-14-1 phase grains. The efficient enhancement of the coercivity of Nd(RE)-Fe-B magnets, with a slight sacrifice of remanence, was demonstrated.
The positive effect of REM-alloy hydrogenated additions to the Nd-Fe-B powder mixture allows the possibility of introducing various components to the permanent magnets (heavy REMs, elements structuring grain boundaries, and restricting the magnet grain growth) at the preparation stage, rather than at the alloy-melting stage. This gives the possibility of using a unified initial alloy for the manufacture of magnets with improved (high-coercive or high-performance) magnetic characteristics. Funding: This study was carried out within the project "Development of physico-chemical and engineering foundations for the initiation of innovative resources-economy technology of high-power and high-coercivity (Nd,r)-Fe-B (R = Pr, Tb, Dy, Ho) low-REM permanent magnets", projects No. LTARF18031 funded by the Ministry of Education, Youth and Sports of the Czech Republic and No. 14.616.21.0093 (unique identification number: RFMEFI61618x0093) funded by the Ministry of Science and Higher Education of the Russian Federation. The SEM/EDS investigation and particle size analysis were performed using the research infrastructure of the Regional Materials Science and Technology Centre, VSB-Technical university of Ostrava (Czech Republic), and XRPD analysis, DTA/TG analysis, and the study of magnetic characteristics were carried out at the Center of Collaborative Access for Functional Nanomaterials and High-Purity Substances, Baikov Institute of Metallurgy and Materials Science, Russian Academy of Sciences. The MasterSizer 3000 particle size analyzer (Malvern) was acquired within the "Development of research and development basis of RMSTC" project (No. CZ.1.05/2.1.00/19.0387) within the frame of the operational program "Research and Development for Innovations" financed by structural funds and the state budget of the Czech Republic.

Conflicts of Interest:
The authors declare no conflict of interest.