Structure and Properties of Heat-Resistant Alloys NiAl–Cr–Co–X (X = La, Mo, Zr, Ta, Re) and Fabrication of Powders for Additive Manufacturing

The NiAl–Cr–Co–X alloys were produced by centrifugal self-propagating high-temperature synthesis (SHS) casting. The effects of dopants X = La, Mo, Zr, Ta, and Re on combustion, as well as the phase composition, structure, and properties of the resulting cast alloys, have been studied. The greatest improvement in overall properties was achieved when the alloys were co-doped with 15% Mo and 1.5% Re. By forming a ductile matrix, molybdenum enhanced strength characteristics up to the values σucs = 1604 ± 80 MPa, σys = 1520 ± 80 MPa, and εpd = 0.79%, while annealing at T = 1250 ℃ and t = 180 min improved strength characteristics to the following level: σucs = 1800 ± 80 MPa, σys = 1670 ± 80 MPa, and εpd = 1.58%. Rhenium modified the structure of the alloy and further improved its properties. The mechanical properties of the NiAl, ZrNi5, Ni0.92Ta0.08, (Al,Ta)Ni3, and Al(Re,Ni)3 phases were determined by nanoindentation. The three-level hierarchical structure of the NiAl–Cr–Co+15%Mo alloy was identified. The optimal plasma treatment regime was identified, and narrow-fraction powders (fraction 8–27 µm) characterized by 95% degree of spheroidization and the content of nanosized fraction <5% were obtained.


Introduction
Nickel aluminide NiAl-based alloys are promising for designing the next-generation aircraft and aerospace products characterized by low specific weight and improved heat resistance, elevated-temperature strength, and creep resistance in the temperature range of 700-1100 • C [1][2][3][4]. The challenges related to application of the conventional casting technologies for product manufacturing because of low fracture toughness and difficulties with subsequent mechanical machining are the factors constraining the practical use of NiAl-based materials [1,[4][5][6].

Materials and Methods
The cast alloy was produced by centrifugal SHS casting [21]. The synthesis was performed on a radial centrifugal setup upon exposure to high gravity up to 300 g. The overall schematic diagram of the centrifugal setup used in this study was earlier presented in references [21,23]. Due to the design of the setup, it was possible to set the rotational speed of the centrifuge rotor in a controlled manner to ensure the target acceleration level. A distinctive feature of this technology is that the relatively cheap oxide raw material is used and high flame temperature (2100-3500 • C) is attained. The chemical flowchart for the basic composition of the alloy can be written as follows:  Table 1 lists the grades and characteristics of the initial components for preparing exothermic powder mixtures. The dopants were added into the reaction mixture so that the basic alloy was obtained. Experiments on co-doping the alloy with molybdenum and rhenium were additionally performed. The preparation scheme of exothermic mixtures involved drying the components in SNOL-type drying cabinets at 90 • C for 1 h, dosing the reagents, mixing, and placing the mixture into graphite molds. Mixing was performed in an MP4/0.5 planetary ball mill for 15-20 min; the drum volume was 1 L; the ball-to-powder weight ratio was 1:5. The combustion temperature of the mixtures was higher than the melting point of the final synthesis products, thus making phase segregation possible due to gravity separation of the molten metal and the cinder. The highest degree of conversion of the target product into a metal ingot was achieved at the optimal acceleration value. The calculated composition of the NiAl-Cr-Co-(X) alloys and concentrations of the doping agents (X) are listed in Table 2. Components Zr, Ta, Re, and La were added to the reaction mixture as pure elements, while molybdenum was added in the oxide form. Tantalum and rhenium were added into the mixture in the form of powders, while lanthanum and zirconium, as metal chips 1-2 mm long and ≤100 µm thick. Boron (0.01 wt.%) was added to all the alloys under study in order to improve their casting characteristics and increase ductility. The effect of heat treatment on the microstructure and mechanical properties of the cast samples was studied by annealing the samples in an SShVL-0.6.2/16I2 vacuum pittype furnace with a heat shield without a ceramic liner at temperatures T 1 = 850 • C, T 2 = 1150 • C, and T 3 = 1250 • C and residual pressure of 0.066-0.106 Pa for 3 h. Heat capacity was calculated according to the increase in temperature of water in the calorimeter filled with water where the samples heated to 100 • C had been immersed. The temperature at which alloy melting started and ended was determined by the method of torsional vibrations on a system for melt viscosity measurements (a viscometer) using cylindrical samples 16 mm in diameter and 30 mm high [48].
Impurity analysis was performed on a Foundry-Master LABFoundry-Master OE750 optical emission spectrometer (Hitachi, Japan). Gas impurity contents were determined on a TC-436 analyzer, and carbon impurities were detected on a CS-230 IH (LECO) analyzer. The compression tests were conducted on a LF-100KN universal testing machine (Walter + Bai AG, Switzerland) according to the GOST standard 25.503-97. The disintegration of cast ingots involved stepwise milling in a VEB LKS5 jaw crusher and comminution in an Activator-4M planetary ball mill (Russia) to a particle size of ≤45 µm. The finegrained fraction with particle size of 20-40 µm was separated by air classification on a Golf-2 laboratory-scale centrifugal classifier (GeFest, Moscow, Russia). The precursor powder was treated with a flow of thermal plasma generated by electric arc discharge. The surface of the spherical powder was cleaned to remove condensed nanoparticles by ultrasonic treatment of the powder in a liquid. The granulometric composition of the particles was determined by laser beam diffraction on an ANALYSETTE 22 MicroTec plus laser diffractometer (Fritch GmbH, Idar-Oberstein, Germany). Bulk weight and flowability were determined according to the GOST standards 16440-94 and 20899-94, respectively.
The phase composition was determined by X-ray diffraction analysis on a D2 PHASER diffractometer (Bruker AXS GmbH, Karlsruhe, Germany) using Cu-Kα radiation within the range of 2θ = 10-140 • . The microstructural studies were conducted on an S-3400N scanning electron microscope (Hitachi, Tokyo, Japan) coupled with a NORAN System 7 X-ray Microanalysis System energy-dispersive spectrometer (Thermo Scientific, Waltham, Massachusetts, USA). The effect of annealing on the crystal structure and phase composition of the NiAl-Cr-Co+15%Mo alloy was studied by high-resolution transmission electron microscopy (HRTEM) on a JEM-2100 microscope (Jeol, Akishima, Japan) using a Gatan 650 Single Tilt Rotation Analytical Specimen Holder (Gatan Inc., Pleasanton, CA, USA). The elemental composition of the phases was measured by energy-dispersive X-ray spectroscopy (EDXS) in the scanning TEM (STEM) mode. Ultra-thin foils for HRTEM studies were prepared by ion etching on a PIPS II System setup (Gatan Inc., Pleasanton, CA, USA). The Young's modulus, hardness, and elastic recovery for individual structural components were determined on polished samples by nanoindentation on a Nanohardness Tester (CSM Instruments, Peuseux, Switzerland) at a load of 20 mN, loading rate of 40 mN/min, and exposure to the load for 5 s. The indentation curves were processed according to standard test method ASTM E2546−15.

Results and Discussion
Thermodynamic analysis of the adiabatic combustion temperature (T ad ) using the THERMO software [34] showed that T ad was 2300-2400 • C for all the compositions under study, being noticeably higher than the melting point of the products. A video of the combustion process and visual inspection of the samples demonstrated that the target product is the cast alloy (an ingot).
Earlier, V. Sanin et al. [21] showed that when synthesizing cast metallic materials by centrifugal SHS casting, exposure to external forces generated in centrifugal setups is among the key tools for affecting structure formation and composition of the resulting products. Visual inspection of the samples synthesized at different acceleration values demonstrated that in the absence of acceleration (a = 1 g), the synthesis products are characterized by high porosity and contain a significant amount of oxide inclusions Al 2 O 3 and gas pores 0.2-1.5 mm in size ( Figure 1). Increasing acceleration generated by a rotating rotor of the centrifugal SHS setup [21] increases the degree of phase separation and reduces the bulk concentration of inclusions. However, at accelerations below a = 50 g, inclusions are retained within the bulk of the sample, while the upper portion of the ingot contains a large shrinkage cavity. The ingots fabricated in the acceleration range a = 150-300 g had no noticeable inclusions or residual porosity. The samples removed from the molds could be easily separated into two layers: the lower layer was the target alloy; the upper layer was corundum Al 2 O 3 . molds could be easily separated into two layers: the lower layer was the target alloy; the upper layer was corundum Al2O3.
A comparative data analysis revealed no significant differences in the degree of phase separation of the synthesis products and dispersion of the mixture during synthesis of the samples with basic composition [24] and the compositions analyzed in this study. However, slight variation in combustion velocity (Uf) of the mixtures was detected.   At least three ingots 80 mm in diameter and 25-30 mm high were synthesized for each composition, and samples for the subsequent tests were cut from these ingots. Curve analysis ( Figure 2) and optical spectroscopy studies of the ingots fabricated at different accelerations inferred that the optimal acceleration is a = 150 ± 5 g. An analysis of optical images on the transverse and longitudinal cross-sections of the samples revealed no residual non-metallic Al2O3 inclusions. Small-sized Al2O3 inclusions resulting from incomplete phase separation were observed at lower a values. No noticeable structural changes occurred when the acceleration was increased above 150 ± 5 g. Table 3 shows the chemical composition of the target products of synthesis of multicomponent NiAl-Cr-Co-X alloys; Table 4 lists the impurity contents. An analysis of the data demonstrated that concentrations of the main components and dopants are close to the calculated ones ( Table 1). The only exception is lanthanum whose content in the target product was significantly lower than the calculated value. The noticeable deviation from A comparative data analysis revealed no significant differences in the degree of phase separation of the synthesis products and dispersion of the mixture during synthesis of the samples with basic composition [24] and the compositions analyzed in this study. However, slight variation in combustion velocity (U f ) of the mixtures was detected. Figure 2 shows the combustion velocity U f of the analyzed mixture compositions as a function of acceleration. The reported data are the average of three experimental values.
However, slight variation in combustion velocity (Uf) of the mixtu 1 g 20 g 50 g 150 g  At least three ingots 80 mm in diameter and 25-30 mm hig each composition, and samples for the subsequent tests were cut fr analysis ( Figure 2) and optical spectroscopy studies of the ingots accelerations inferred that the optimal acceleration is a = 150 ± 5 g images on the transverse and longitudinal cross-sections of the s sidual non-metallic Al2O3 inclusions. Small-sized Al2O3 inclusions plete phase separation were observed at lower a values. No notice occurred when the acceleration was increased above 150 ± 5 g.  At least three ingots 80 mm in diameter and 25-30 mm high were synthesized for each composition, and samples for the subsequent tests were cut from these ingots. Curve analysis ( Figure 2) and optical spectroscopy studies of the ingots fabricated at different accelerations inferred that the optimal acceleration is a = 150 ± 5 g. An analysis of optical images on the transverse and longitudinal cross-sections of the samples revealed no residual non-metallic Al 2 O 3 inclusions. Small-sized Al 2 O 3 inclusions resulting from incomplete phase separation were observed at lower a values. No noticeable structural changes occurred when the acceleration was increased above 150 ± 5 g. Table 3 shows the chemical composition of the target products of synthesis of multicomponent NiAl-Cr-Co-X alloys; Table 4 lists the impurity contents. An analysis of the data demonstrated that concentrations of the main components and dopants are close to the calculated ones ( Table 1). The only exception is lanthanum whose content in the target product was significantly lower than the calculated value. The noticeable deviation from the calculated value can be attributed to the fact that due to the high affinity for oxygen, most of La participated in the reaction of oxide reduction and competed with the main reducing agent (Al). The Al 2 O 3 -based cinder phase contained 0.3 wt.% of La. Therefore, doping with lanthanum exhibited little effect for fabricating the alloy by centrifugal SHS casting. During any metallurgical process, ingots always contain impurities. SHS metallurgy is not an exception: the target product contains impurities even when the synthesis regimes have been properly chosen, so their mechanical properties can be deteriorated [1,21,39,40]. Thus, the ingots contain up to 0.1% of iron impurity due to the presence of 0.5 wt.% Fe in the initial powder of nickel oxide NiO. Table 5 show the X-ray diffraction (XRD) data of the synthesis products and the mechanical properties of individual phases determined by selective indentation. β-NiAl is the main phase for all the synthesized compositions. The Ta-doped cast alloy contains the intermetallic compound Ni 3 (AlTa), while doping it with Zr yields the ZrNi 5 phase.

Figures 3 and 4 and
An analysis of the indentation curve and the XRD data obtained by studying the NiAl-Cr-Co alloys doped with Mo, Zr, Ta, and Re revealed the typical mechanical properties of the NiAl, ZrNi 5, Ni 0.92 Ta 0.08 , and (Al,Ta)Ni 3 phases, as well as the hypothetical Al(Re,Ni) 3 phase, which was not detected by XRD. The values of these properties are summarized in Table 5.
When performing an analysis of the indentation data, we selected similar curves and classified them into groups in accordance with the experimentally determined phase composition of the alloy. The averaged hardness and Young's modulus were then calculated for each phase. When doing so, it was taken into account that the intermetallic compound NiAl was the predominant phase (content >85%). Figure 4 shows an example of this approach. One can see that in addition to the curves typical of the NiAl phase (with the indent depth of~330 nm), the doped NiAl-Cr-Co+0.5%Zr sample has another two indentation curves characterized by indent depth of >430 nm. This more ductile phase can presumably correspond to Ni-rich ZrNi 5 intermetallic compound.   An analysis of the indentation curve and the XRD data obtained by studying the NiAl-Cr-Co alloys doped with Mo, Zr, Ta, and Re revealed the typical mechanical properties of the NiAl, ZrNi5, Ni0.92Ta0.08, and (Al,Ta)Ni3 phases, as well as the hypothetical Al(Re,Ni)3 phase, which was not detected by XRD. The values of these properties are summarized in Table 5.
When performing an analysis of the indentation data, we selected similar curves and classified them into groups in accordance with the experimentally determined phase composition of the alloy. The averaged hardness and Young's modulus were then calculated for each phase. When doing so, it was taken into account that the intermetallic compound NiAl was the predominant phase (content >85%). Figure 4 shows an example of this approach. One can see that in addition to the curves typical of the NiAl phase (with the     Since microcracks propagate in intermetallic materials predominantly at intergrain boundaries via a longer path (to go round inclusions) [15,16], it is fair to assume that the alloy containing molybdenum precipitates will be characterized by increased ductility at room temperature.
The intergrain space of the Zr-doped alloy ( Figure 6) was found to contain zone seg- An analysis of the microstructures of the Ta-doped alloy ( Figure 7) showed that the intergrain space contains precipitates of the Laves phase (Cr2Ta). The Ni3AlTa phase is supposedly formed along the grain boundaries (the light-gray areas). Figure 8b shows the results of microprobe analysis along the line within the intergrain space. One can see that interlayers of Cr-based solid solution Cr(Co,Ni,Al) reside between NiAl grains.  The analysis of the microstructures of the Re-doped alloy ( Figure 9) showed that s of all the structural components decreased twofold compared to that for other alloys. Do ing heat-resistant alloys with rhenium especially affects their structural stability at hi temperatures, behavior of the material upon primary creep, and oxidative characterist [3]. Doping with rhenium to produce alloys using the centrifugal SHS casting technolo was investigated for the first time in this study. Elemental analysis demonstrated that added amount of the component completely passes into the metal ingot, without a losses for formation of the cinder phase. It is important to mention that rhenium is hom geneously distributed over the intergrain space of the Cr-based solid solution. The cont of rhenium in the solid solution ranges from 2 at. % (dark gray regions) to 12 at. % (lig colored regions) ( Figure 9). Since microcracks propagate in intermetallic materials predominantly at intergrain boundaries via a longer path (to go round inclusions) [15,16], it is fair to assume that the alloy containing molybdenum precipitates will be characterized by increased ductility at room temperature.
The intergrain space of the Zr-doped alloy ( Figure 6) was found to contain zone segregations of Ni5Zr. The characteristic size of the structural components and the principle of microstructure formation are similar to those for the developed Hf-doped CompoNiAl-M5-3 alloy [23][24][25][26]. It can be assumed that this microheterogeneous structure will not cause deterioration of mechanical properties. Materials 2021, 14, x FOR PEER REVIEW 14 of 32 Taking into account the features of structure formation of the molybdenum-containing phases, additional experiments were conducted, aiming to synthesize the NiAl-Cr-Co alloy with increased Mo concentration (up to 15%) and co-doping it with molybdenum and rhenium (15 wt.%Mo + 1.5 wt.%Re). In the former case, the objective was to increase the ductility of the alloy at room temperature by increasing the bulk content of the ductile (Cr, Mo) phase. Additional doping with rhenium was performed to obtain a fine-grained structure. Table 6 summarizes the elemental composition of the NiAl-Cr-Co+15%Mo and NiAl-Cr-Co+15%Mo+1.5%Re alloys, while Table 7 lists the impurity content. Table 6. Elemental composition of the NiAl-Cr-Co+15%Mo and NiAl-Cr-Co+15%Mo+1.5%Re alloys.
* NiAl-Cr-Co-(X).  An analysis of the microstructures of the Ta-doped alloy ( Figure 7) showed that the intergrain space contains precipitates of the Laves phase (Cr 2 Ta). The Ni 3 AlTa phase is supposedly formed along the grain boundaries (the light-gray areas). Figure 8b shows the results of microprobe analysis along the line within the intergrain space. One can see that interlayers of Cr-based solid solution Cr(Co,Ni,Al) reside between NiAl grains.
The analysis of the microstructures of the Re-doped alloy ( Figure 9) showed that size of all the structural components decreased twofold compared to that for other alloys. Doping heat-resistant alloys with rhenium especially affects their structural stability at high temperatures, behavior of the material upon primary creep, and oxidative characteristics [3]. Doping with rhenium to produce alloys using the centrifugal SHS casting technology was investigated for the first time in this study. Elemental analysis demonstrated that the added amount of the component completely passes into the metal ingot, without any losses for formation of the cinder phase. It is important to mention that rhenium is homogeneously distributed over the intergrain space of the Cr-based solid solution. The content of rhenium in the solid solution ranges from 2 at. % (dark gray regions) to 12 at. % (light-colored regions) (Figure 9).
Taking into account the features of structure formation of the molybdenum-containing phases, additional experiments were conducted, aiming to synthesize the NiAl-Cr-Co alloy with increased Mo concentration (up to 15%) and co-doping it with molybdenum and rhenium (15 wt.%Mo + 1.5 wt.%Re). In the former case, the objective was to increase the ductility of the alloy at room temperature by increasing the bulk content of the ductile (Cr, Mo) phase. Additional doping with rhenium was performed to obtain a fine-grained structure. Table 6 summarizes the elemental composition of the NiAl-Cr-Co+15%Mo and NiAl-Cr-Co+15%Mo+1.5%Re alloys, while Table 7 lists the impurity content.  The XRD pattern of the synthesis products are shown in Figure 10. β-NiAl, (Cr, Mo) solid solution, and the accompanying phases shown in Table 8  The XRD pattern of the synthesis products are shown in Figure 10. β-NiAl, (Cr, Mo) solid solution, and the accompanying phases shown in Table 8 are the main phases for both compositions. The sample with 15 wt.%Mo + 1.5 wt.%Re contains the Laves phase MoRe2. The results of microstructural studies of the alloys with increased Mo and Mo + Re contents are shown in Figures 11 and 12, respectively. *-NiAlCrCo Figure 10. The XRD pattern of the NiAl-Cr-Co-(X) alloys, X= 15%Mo and 15%Mo + 1.5%Re. The results of microstructural studies of the alloys with increased Mo and Mo + Re contents are shown in Figures 11 and 12, respectively. The alloy doped with 15% Mo has a cellular structure ( Figure 11) and consists of the following phases: NiAl, Cr-based solid solution, Mo-based solid solution, and the (Ni,Cr,Co)3Mo3C phase that was identified by a more thorough study.
Precipitates of the MoRe2 phase are additionally observed within the structure of the alloy doped with 15%Mo + 1.5%Re ( Figure 12). It is expected that the dispersed Cr(Mo) and MoRe2 particles will have a favorable effect on strength properties of the alloy (and primarily on resistance to viscoplastic flow) due to deceleration of mobile matrix disloca-   Table 9 summarizes the properties of SHS ingots: the melting point Tmelt, density ρ, heat capacity Cv, hardness, ultimate compression strength σucs, the offset yield stress σys, and the degree of plastic deformation εpd. Samples of the alloys co-doped with molybdenum and rhenium exhibited the highest strength characteristics (Figure 13a). Thus, the ultimate compressive strengths of the alloys doped with 2.5% Mo and 15% Mo were σucs = 1586 and 1728 MPa, respectively, while σucs of the alloy co-doped with Mo and Re was 1800 MPa. These values are comparable with strength of the CompoNiAl-M5-3 alloy subjected to vacuum induction melting [25]. The alloys doped with Re and Ta were also characterized by a relatively high ultimate compressive strength. The alloy doped with 15% Mo has a cellular structure ( Figure 11) and consists of the following phases: NiAl, Cr-based solid solution, Mo-based solid solution, and the (Ni,Cr,Co) 3 Mo 3 C phase that was identified by a more thorough study.
Precipitates of the MoRe 2 phase are additionally observed within the structure of the alloy doped with 15%Mo + 1.5%Re (Figure 12). It is expected that the dispersed Cr(Mo) and MoRe 2 particles will have a favorable effect on strength properties of the alloy (and primarily on resistance to viscoplastic flow) due to deceleration of mobile matrix dislocations [38]. Table 9 summarizes the properties of SHS ingots: the melting point T melt , density ρ, heat capacity C v , hardness, ultimate compression strength σ ucs , the offset yield stress σ ys , and the degree of plastic deformation ε pd . Samples of the alloys co-doped with molybdenum and rhenium exhibited the highest strength characteristics (Figure 13a). Thus, the ultimate compressive strengths of the alloys doped with 2.5% Mo and 15% Mo were σ ucs = 1586 and 1728 MPa, respectively, while σ ucs of the alloy co-doped with Mo and Re was 1800 MPa. These values are comparable with strength of the CompoNiAl-M5-3 alloy subjected to vacuum induction melting [25]. The alloys doped with Re and Ta were also characterized by a relatively high ultimate compressive strength.  The resulting experimental data demonstrate that formation of the ductile phase in the intergrain space had a positive effect on mechanical properties, while the characteristics of the alloys can be further improved during the subsequent technological stages, including HIP [36] and SLM, in a manner similar to that for the CompoNiAl-M5-3 alloy. The resulting experimental data demonstrate that formation of the ductile phase in the intergrain space had a positive effect on mechanical properties, while the characteristics of the alloys can be further improved during the subsequent technological stages, including HIP [36] and SLM, in a manner similar to that for the CompoNiAl-M5-3 alloy.
Melt crystallization rates achieved during alloy production by centrifugal SHS casting can be as high as 20-25 • C/s [21]. Supersaturated solid solutions are formed at these crystallization rates of multicomponent melts; the phases exist in a non-equilibrium state; and the samples contain residual stresses. Therefore, vacuum annealing usually favorably affects strength and ductility [49]. Heat treatment of the NiAl-Cr-Co-Hf alloy at T = 850 • C and p = 10 −2 Pa for 3 h simultaneously increased strength and ductility due to concentration stratification of the supersaturated chromium-based solid solution and precipitation of strengthening α-Cr nanoparticles (sized less than 45 nm) and the Heusler phase Ni 2 AlHf (3-5 nm) [24].
In this study, annealing was performed for the alloys doped with 15% Mo and 15%Mo + 1.5%Re and was shown to have a favorable effect on mechanical properties of the alloys. Figure 13b,c shows that strength and ductility increase noticeably as the annealing temperature rises from 850 to 1250 • C. There is a plastic deformation region at 1250 • C, corresponding to residual deformation ε pd = 2.01 and 6.15% for the alloys doped with 15%Mo and 15%Mo + 1.5%Re, respectively. It is important to mention that Re also increases the degree of plastic deformation because the grain structure of the alloy is refined as it is uniformly distributed along the intergrain boundaries predominantly within chromium-and molybdenum-based solid solutions. Figure 13 compares our findings with the data reported in ref. [24] for the Ni 41 Al 41 Cr 12 Co 6 alloy. The diagram was re-plotted with allowance for rigidity of the testing machine and the Young's modulus determined graphically, which is equal to~200 GPa for the NiAlbased alloys [1,4]. This alloy is characterized by high ultimate compressive strength σ ucs = 2250 MPa. Annealing at 1250 • C made it possible for the alloy doped with 15% Mo to approach this value, while the alloy doped with 15%Mo + 1.5%Re has reached it (Table 10). Plastic deformation of the NiAl-Cr-Co+15%Mo+1.5%Re alloy annealed at 1250 • C is higher than that of the Ni 41 Al 41 Cr 12 Co 6 alloy by 1.92% due to precipitation of the viscous (Cr,Mo) phase as interdendritic interlayers. Unlike the integral mechanical properties evaluated during the compression tests, nanoindentation measurements showed that local mechanical properties (hardness H and the Young's modulus E) as a function of annealing temperature were reduced by 10-12% ( Figure 14). This can be possibly related to the coherence loss at the interface between the nanosized disc-shaped Cr-based precipitates and the supersaturated solid solution via the mechanism of Guinier-Preston structural transformation, which takes place in NiAl-Cr-Co-Hf alloys at temperatures above 850 • C as it was determined earlier in ref. [24]. Figure 14 shows that this thermal behavior of local properties is typical of the samples cut from the ingots both along and across their axis. This was proved by the texture formed during centrifugal SHS casting, which has led to anisotropy of the properties.     By comparing Figures 13 and 14, one can infer that local disordering during annealing increases the content of the plastic component of strain ε pd in the compression tests. Figure 15 shows the microstructures of the NiAl-Cr-Co+15%Mo alloy before (a) and after annealing at 1150 • C (b) and 1250 • C (c).     One can see that the alloy structure is more homogeneous in the range of maximal working temperatures for this class of materials (1150 • C); NiAl grains become smaller. Growth of dendritic grains, the (Ni,Cr,Co) 3 Mo 3 C phase, and the (Cr,Mo) solid solution took place at annealing temperature of 1250 • C. The structural components of the NiAl-Cr-Co+15%Mo alloy after annealing at 1150 • C (a) and 1250 • C (b) are shown in Figure 16.
One can see that the alloy structure is more homogeneous in the range of maximal working temperatures for this class of materials (1150 °C); NiAl grains become smaller. Growth of dendritic grains, the (Ni,Cr,Co)3Mo3C phase, and the (Cr,Mo) solid solution took place at annealing temperature of 1250 °C. The structural components of the NiAl-Cr-Co+15%Mo alloy after annealing at 1150 °C (a) and 1250 °C (b) are shown in Figure 16. In order to investigate the crystal structure of the components of the NiAl-Cr-Co+15%Mo alloy before and after annealing at 1250 °C, we studied ultrathin foils made of this alloy by HRTEM and electron diffraction. Figure 17a,b shows the bright-field (BF) TEM images of the characteristic structure of the NiAl-Cr-Co+15%Mo alloy near the interface boundary of a dendritic cell. According to the EDXS data, dendrites were composed of the solid solution of chromium and cobalt in β-NiAl (Table 11, spectrum 1). A feature of this material was that elongated grains of the molybdenum-containing phases (1-2 µm wide) ( Table 11, spectra 2-5) were formed in the interdendritic space owing to the effect of non-equilibrium conditions of melt crystallization at excessive molybdenum and chromium contents during SHS casting [21,24]. The relatively slow cooling down of the ingots in air contributed to additional precipitation of nanosized (<100 nm) particles of the excessive phase (composition: Cr, 65.92 at.%; Mo, 25.39 at.%; Ni, 5.62 at.%; and Al, 2.07 at.%) in the dendrite bodies. These precipitates increased the resistance to plastic deformation due to dispersion strengthening of the NiAl matrix as previously demonstrated in references [31,36,50,51]. In order to investigate the crystal structure of the components of the NiAl-Cr-Co+15%Mo alloy before and after annealing at 1250 • C, we studied ultrathin foils made of this alloy by HRTEM and electron diffraction. Figure 17a,b shows the bright-field (BF) TEM images of the characteristic structure of the NiAl-Cr-Co+15%Mo alloy near the interface boundary of a dendritic cell. According to the EDXS data, dendrites were composed of the solid solution of chromium and cobalt in β-NiAl (Table 11, spectrum 1). A feature of this material was that elongated grains of the molybdenum-containing phases (1-2 µm wide) ( Table 11, spectra 2-5) were formed in the interdendritic space owing to the effect of non-equilibrium conditions of melt crystallization at excessive molybdenum and chromium contents during SHS casting [21,24]. The relatively slow cooling down of the ingots in air contributed to additional precipitation of nanosized (<100 nm) particles of the excessive phase (composition: Cr, 65.92 at.%; Mo, 25.39 at.%; Ni, 5.62 at.%; and Al, 2.07 at.%) in the dendrite bodies. These precipitates increased the resistance to plastic deformation due to dispersion strengthening of the NiAl matrix as previously demonstrated in references [31,36,50,51].
A more detailed analysis of the fine structure of dendritic cells revealed coherent (Cr,Mo) precipitates sized 10-20 nm, which is demonstrated by the HRTEM image of the β-phase taken along the [111] zone axis (Figure 17c). The unit cell parameter of NiAl calculated using the SAED pattern (the inset on the left-hand side in Figure 17c) was a = 2.952 Å, being higher than the tabulated value (a = 2.887 Å) by 2.3%. The increase in the unit cell parameter of the matrix phase was presumably caused by dissolution of molybdenum, chromium, and cobalt atoms in this phase as proved by the results of EDXS analysis of the chemical composition of the dendrite body (Table 11, spectrum 1). The crystal structure of dendritic cells was characterized by high number density of defects in the form of alternating dark-and light-contrast lines oriented along <200> (Ni, Fe)Al. These defects resulted from plastic deformation of the matrix by passing numerous dislocations of the crystal lattice on the <110> planes upon exposure of the alloy to compressive stress, which is proved by the presence of additional reflections in the SAED pattern (the inset in Figure 17c) exhibiting characteristic mirror-type dislocation with respect to the reflections with stronger intensity. The IFFT-filtered [001] HRTEM image of the crystal structure of the NiAl/(Cr, Mo) interface from the area indicated with A in Figure 17c demonstrates that there is an a/2{110} misfit dislocation (marked with T), thus indicating that even the excessive precipitates (≤14 nm in size) in the deformed alloy have lost complete coherence.  Figure 17e.
A more detailed analysis of the fine structure of dendritic cells revealed coherent (Cr,Mo) precipitates sized 10-20 nm, which is demonstrated by the HRTEM image of the β-phase taken along the [1 11] zone axis (Figure 17c). The unit cell parameter of NiAl calculated using the SAED pattern (the inset on the left-hand side in Figure 17c) was a = 2.952 Å, being higher than the tabulated value (a = 2.887 Å) by 2.3%. The increase in the unit cell parameter of the matrix phase was presumably caused by dissolution of molybdenum, chromium, and cobalt atoms in this phase as proved by the results of EDXS analysis of the   The BF TEM image of the structure of the annealed NiAl-15Mo-1 taken along the [1 11] (Ni,Co,Cr)3Mo3C zone axis is shown in Figure 18a. An diffusion-controlled growth of strengthening Cr(Mo) nanoprecipitates a grain boundaries up to the submicron size (150-400 nm) and increased cobalt concentrations in the (Ni,Co,Cr)3Mo3C phase (Table 11, spectrum 6 references [24,26,31,44], precipitation of dispersed particles of the excessiv took place at the stage when the ingots were cooled down in the vacuum electric furnace as a result of concentration stratification of supersaturated Cr(Mo) in the intermetallic matrix. Changes in the composition of interden ers are caused by its diffusion saturation with doping elements as dispe (Cr,Mo) are dissolved in β-NiAl during the heat treatment of the alloy. T pattern (marked with F in Figure 17a) proved that there is the (Ni,Co,Cr)3M a = 10.74 Å) having a suprastructure. The decrease in the lattice paramete denum-containing phase by 5.5% for the annealed alloy can be caused b molybdenum concentration in this phase is reduced.
No coherent coupling of atomic planes of crystal lattices betw (Ni,Co,Cr)3Mo3C and β-NiAl phases has been detected as indicated by the image of (Ni,Co,Cr)3Mo3C crystal structure at the interface with the dend the matrix phase (Figure 18b). Despite the deformed state of the alloy, n were revealed in the phase under study. The inset in Figure 18b also de supra-structure of the crystal lattice of the phase being analyzed.  The BF TEM image of the structure of the annealed NiAl-15Mo-12Cr-6Co alloy taken along the [1 11] (Ni,Co,Cr)3Mo3C zone axis is shown in Figure 18a. Annealing caused diffusion-controlled growth of strengthening Cr(Mo) nanoprecipitates along the inter grain boundaries up to the submicron size (150-400 nm) and increased chromium and cobalt concentrations in the (Ni,Co,Cr)3Mo3C phase (Table 11, spectrum 6). According to references [24,26,31,44], precipitation of dispersed particles of the excessive Cr(Mo) phas took place at the stage when the ingots were cooled down in the vacuum chamber of th electric furnace as a result of concentration stratification of supersaturated solid solution Cr(Mo) in the intermetallic matrix. Changes in the composition of interdendritic interlay ers are caused by its diffusion saturation with doping elements as dispersed inclusion (Cr,Mo) are dissolved in β-NiAl during the heat treatment of the alloy. The [1 11] SAED pattern (marked with F in Figure 17a) proved that there is the (Ni,Co,Cr)3Mo3C phase (fcc a = 10.74 Å) having a suprastructure. The decrease in the lattice parameter of the molyb denum-containing phase by 5.5% for the annealed alloy can be caused by the fact tha molybdenum concentration in this phase is reduced.
No coherent coupling of atomic planes of crystal lattices between the main (Ni,Co,Cr)3Mo3C and β-NiAl phases has been detected as indicated by the [1 11] HRTEM image of (Ni,Co,Cr)3Mo3C crystal structure at the interface with the dendritic unit cell o the matrix phase (Figure 18b). Despite the deformed state of the alloy, no linear defect were revealed in the phase under study. The inset in Figure 18b also demonstrates th supra-structure of the crystal lattice of the phase being analyzed.  Figure 17d shows the HRTEM image of the crystal structure of the multicomponent phase in the Mo-Cr-Ni-Co system (Table 11, spectra 2 and 3). According to the quantitative ratio between the main components in the phase, we have put forward a hypothesis that a continuous series of solid solutions with a bcc structure has been formed. However, the detected Mo-based phase had an fcc crystal lattice oriented along the [114] zone axis. This was confirmed by the results of measuring angles between the main direction vectors in the SAED pattern (the inset in Figure 17d), which fully corresponded to the tabulated values and were equal to 63.9 • and 50.1 • [52,53]. The lattice parameter of the phase under study calculated using the recorded SAED pattern with allowance for the Miller indices was a = 11.37 Å. Taking into account the determined lattice parameters, one can state that a complex carbide (Ni,Co,Cr) 3 Mo 3 C crystallizes in the interdendritic space. Along with boron, the impurity carbon interacts with (Cr, Mo) at grain boundaries to give rise to carbides, thus being involved in dispersion hardening of the alloy by slowing down grain boundary diffusion [54,55]. The IFFT filtered image of the crystal structure (marked with B) shows the ordered arrangement of atoms of the second crystal lattice between the atoms of the first one, being indicative of superstructure formation via the defect-ordering mechanism [51,54]. Figure 17e shows the results of analyzing the crystal structure of a grain with composition Mo, 78 ± 2 at. % and Cr, 22 ± 2 at. % (Table 11, spectra 4 and 5) near the interface boundary with the dendritic cell of β-NiAl. According to the recorded SAED pattern (the inset in Figure 17e), we have identified the lattice type and parameters for the phase under study. Taking into account the angles between the main crystal directions (61.2 • {2201} and 28.2 • {2020}) of the hcp lattice and the calculated lattice constants (a = 5.92 Å and c = 6.41 Å), a complex boride (Mo 0.8 Cr 0.2 ) x B y was formed instead of (Mo,Cr) solid solution. Boron could not be identified at the stage of phase composition measurements, since an X-Max80 T silicon drift detector (Oxford Instruments, High Wycombe, UK) was used in the study.
The FFT image from β-NiAl dendritic cell (marked with C) compared to the [1216] SAED inset shows that there is no coherent bonding of crystal lattices of the phases at the interface with (Mo 0.8 Cr 0.2 ) x B y (Figure 17e), since the main crystal direction vectors do not coincide. Nevertheless, the analyzed phases have close orientation, so it might be possible to provide conditions when boride precipitates will be coherent. This will cause additional dispersion hardening of the material when elastic strain fields generated by the crystal lattices of two phases impede the gliding of matrix dislocations.
The IFFT filtered [1216] HRTEM image recorded for the area indicated with D in Figure 17e shows the high-density stacking faults oriented along the atomic planes {2201} (Figure 17f). This stacking fault (SF) with an additional similar layer was likely to be produced by aggregation of interstitial atoms [44]. Hence, plastic deformation of the (Mo 0.8 Cr 0.2 ) x B y phase exposed to compressive strain is accompanied by dissociation of total dislocations into partial ones.
The BF TEM image of the structure of the annealed NiAl-15Mo-12Cr-6Co alloy taken along the [111] (Ni,Co,Cr) 3 Mo 3 C zone axis is shown in Figure 18a. Annealing caused diffusion-controlled growth of strengthening Cr(Mo) nanoprecipitates along the intergrain boundaries up to the submicron size (150-400 nm) and increased chromium and cobalt concentrations in the (Ni,Co,Cr) 3 Mo 3 C phase (Table 11, Figure 17a) proved that there is the (Ni,Co,Cr) 3 Mo 3 C phase (fcc, a = 10.74 Å) having a suprastructure. The decrease in the lattice parameter of the molybdenum-containing phase by 5.5% for the annealed alloy can be caused by the fact that molybdenum concentration in this phase is reduced.
No coherent coupling of atomic planes of crystal lattices between the main (Ni,Co,Cr) 3 Mo 3 C and β-NiAl phases has been detected as indicated by the [111] HRTEM image of (Ni,Co,Cr) 3 Mo 3 C crystal structure at the interface with the dendritic unit cell of the matrix phase (Figure 18b). Despite the deformed state of the alloy, no linear defects were revealed in the phase under study. The inset in Figure 18b also demonstrates the supra-structure of the crystal lattice of the phase being analyzed.
The bodies of the dendritic cells were found to contain highly dispersed (Cr,Mo) inclusions sized~21 nm (Figure 18c). The observed coupling of atomic planes in the HRTEM image of (Cr,Mo)/NiAl interface taken along a common [111] zone axis, as well as the superposition of the diffraction spots in the FFT patterns for the matrix and the precipitates confirm that they are coherent. Therefore  Figure 18c. This fact also explains the nature of the dark-colored nanosized zones in the crystal structure of the material, which mainly reside around a coherent precipitate of the excess phase; their presence is caused by elastic deformations of the lattice around the dislocation centers.
between their crystal lattices. The development of plastic deformation in the alloy due to compressive stress resulted in accumulation of a/2 [1 11] (011) edge dislocations formed by extra half-planes of the same sign as demonstrated by the IFFT filtered area marked with G in Figure 18c. This fact also explains the nature of the dark-colored nanosized zones in the crystal structure of the material, which mainly reside around a coherent precipitate of the excess phase; their presence is caused by elastic deformations of the lattice around the dislocation centers. Further spheroidization studies were conducted for the most promising alloy with composition NiAl-Cr-Co + 15%Mo. The precursor powder having a target composition was obtained by mechanical grinding of the cast SHS ingots. The spherical powder for additive manufacturing SLM machines was fabricated using the plasma spheroidization method.
The precursor powder of the alloy under study had a characteristic fragmented morphology of particles (Figure 19a). Most particles were ≤30 µm in size; however, the powder also contained relatively coarse particles sized up to 50 µm. It is important to mention that the classified powder contained no submicron-sized particles that abruptly deteriorate the technological properties of the powder and have an unfavorable effect on the stability of Further spheroidization studies were conducted for the most promising alloy with composition NiAl-Cr-Co + 15%Mo. The precursor powder having a target composition was obtained by mechanical grinding of the cast SHS ingots. The spherical powder for additive manufacturing SLM machines was fabricated using the plasma spheroidization method.
The precursor powder of the alloy under study had a characteristic fragmented morphology of particles (Figure 19a). Most particles were ≤30 µm in size; however, the powder also contained relatively coarse particles sized up to 50 µm. It is important to mention that the classified powder contained no submicron-sized particles that abruptly deteriorate the technological properties of the powder and have an unfavorable effect on the stability of plasma spheroidization, thus causing excessive evaporation and condensation of the material on the surface of coarse particles [30]. Figure 19b shows the integral and differential particle size distribution of the NiAl-Cr-Co+15%Mo powder after mechanical grinding in a planetary ball mill followed by air classification. Particle size of the precursor powder ranged between 7 and 79 µm. The average particle diameter D av was 33.9 µm. The resulting powder was characterized by bimodal particle size distribution. The mode corresponding to the first peak was 17 µm, while the mode for the second (higher) peak was 49 µm. The distribution quantiles D 10 , D 50 , and D 90 were 12.3, 31.6, and 60.7 µm, respectively. The content of the fraction <20 µm and >40 µm was 35% and 38%, respectively. also contained relatively coarse particles sized up to 50 µm. It is important to mention that the classified powder contained no submicron-sized particles that abruptly deteriorate the technological properties of the powder and have an unfavorable effect on the stability of plasma spheroidization, thus causing excessive evaporation and condensation of the material on the surface of coarse particles [30].
- Figure 19. The morphology of the comminuted particles (a) and the granulometric composition (b) of the NiAl-Cr-Co+15%Mo alloy. Figure 19b shows the integral and differential particle size distribution of the NiAl-Cr-Co+15%Mo powder after mechanical grinding in a planetary ball mill followed by air classification. Particle size of the precursor powder ranged between 7 and 79 µm. The average particle diameter Dav was 33.9 µm. The resulting powder was characterized by bimodal particle size distribution. The mode corresponding to the first peak was 17 µm, while the mode for the second (higher) peak was 49 µm. The distribution quantiles D10, D50, and D90 were 12.3, 31.6, and 60.7 µm, respectively. The content of the fraction <20 µm and >40 µm was 35% and 38%, respectively.
To choose the optimal powder spheroidization regime, we conducted a series of experiments aiming to evaluate the effect of enthalpy of plasma flow, the composition of the plasma supporting gas, and flowability of the precursor powder on the spheroidization degree and the intensity of powder evaporation yielding condensed nanoparticles. The study was conducted at the following design and technological parameters: plasma torch power Npl = 6.6-13.8 kW; Ar and Ar + H2 were used as plasma supporting gases; the flowability of the plasma supporting gas Gpl.gas = 2.7 m 3 /h; the enthalpy of plasma flow Ipl = 1.04-1.91 kW h/m 3 ; and powder feed rate Gpowder = 1.5-6.0 kg/h.
Having analyzed the results of the studies, we found that the degree of spheroidization of the product increased from 80% to 92% with rising enthalpy of plasma flow. A powder with 65-99% degree of spheroidization can be produced depending on the feed rate of the precursor powder. The use of hydrogen-containing thermal plasma increases heat capacity in the "gas-processed material" system. This results in a higher rate of heating of the particles being processed and an increase in the degree of powder spheroidization to 99%. The intensity of powder evaporation depending on various treatment parameters was determined experimentally. The content of the nano-sized fraction in the powder ranged from 7.9 to 13.5 wt.%.
When using argon plasma with the powder feed rate of 3 kg/h and the enthalpy of plasma flow Ipl = 1.71 kW h/m 3 , the resulting powder was characterized by the degree of spheroidization of 92% and nanoparticle content of 10% (Figure 20a). Before conducting the structural analysis and measuring particle size distribution, the powder was subjected to ultrasonic treatment in a liquid to remove condensed particles from its surface. To choose the optimal powder spheroidization regime, we conducted a series of experiments aiming to evaluate the effect of enthalpy of plasma flow, the composition of the plasma supporting gas, and flowability of the precursor powder on the spheroidization degree and the intensity of powder evaporation yielding condensed nanoparticles. The study was conducted at the following design and technological parameters: plasma torch power N pl = 6.6-13.8 kW; Ar and Ar + H 2 were used as plasma supporting gases; the flowability of the plasma supporting gas G pl.gas = 2.7 m 3 /h; the enthalpy of plasma flow I pl = 1.04-1.91 kW h/m 3 ; and powder feed rate G powder = 1.5-6.0 kg/h.
Having analyzed the results of the studies, we found that the degree of spheroidization of the product increased from 80% to 92% with rising enthalpy of plasma flow. A powder with 65-99% degree of spheroidization can be produced depending on the feed rate of the precursor powder. The use of hydrogen-containing thermal plasma increases heat capacity in the "gas-processed material" system. This results in a higher rate of heating of the particles being processed and an increase in the degree of powder spheroidization to 99%. The intensity of powder evaporation depending on various treatment parameters was determined experimentally. The content of the nano-sized fraction in the powder ranged from 7.9 to 13.5 wt.%.
When using argon plasma with the powder feed rate of 3 kg/h and the enthalpy of plasma flow I pl = 1.71 kW h/m 3 , the resulting powder was characterized by the degree of spheroidization of 92% and nanoparticle content of 10% (Figure 20a). Before conducting the structural analysis and measuring particle size distribution, the powder was subjected to ultrasonic treatment in a liquid to remove condensed particles from its surface. Complete powder spheroidization was not achieved at these parameters of plasma treatment (irregular-shaped particles were present in the range of 10-20 µm and 30-50 µm). Most spherical particles contained Al 2 O 3 inclusions caused by evaporation of the material, oxidation of aluminum, and condensation of the material on the surface. According to the laser diffraction data (Figure 20b), the spheroidized powder had a unimodal particle size distribution in the range of 5-44 µm. The characteristic dimensions were as follows: D av = 18.0 µm; D 10 = 8.4 µm; D 50 = 16.4 µm; and D 90 = 30.0 µm.
Due to the use of the Ar-H 2 mixture as a plasma supporting gas, the enthalpy of plasma flow was increased to 1.9 kW h/m 3 and the degree of powder spheroidization rose to 98% (Figure 21a) as energy density of the plasma flow per unit of surface area of the processed material was increasing. However, the content of condensed particles also rose to 11.6%. An analysis of the morphology of powder particles revealed no irregular-shaped sharp-cornered particles and a significantly smaller content of Al 2 O 3 inclusions. Some particles contain satellites sized 1-10 µm. Figure 21b shows the granulometric composition of spherical particles at these process parameters. The increasing enthalpy of plasma flow made it possible to narrow the range of particle size distribution to 6-26 µm. The powder is characterized by unimodal particle size distribution with D av = 13.53 µm. The distribution quantiles were as follows: D 10 = 9.0 µm; D 50 = 13.2 µm; and D 90 = 18.7 µm. Figure 20. The morphology (a) and granulometric composition (b) of the NiAl-Cr-Co+15%Mo alloy powder at Ipl = 1.71 kW h/m 3 , Gpowder = 3 kg/h, with Ar used as a plasma supporting gas.
Complete powder spheroidization was not achieved at these parameters of plasma treatment (irregular-shaped particles were present in the range of 10-20 µm and 30-50 µm). Most spherical particles contained Al2O3 inclusions caused by evaporation of the material, oxidation of aluminum, and condensation of the material on the surface. According to the laser diffraction data (Figure 20b), the spheroidized powder had a unimoda particle size distribution in the range of 5-44 µm. The characteristic dimensions were as follows: Dav = 18.0 µm; D10 = 8.4 µm; D50 = 16.4 µm; and D90 = 30.0 µm.
Due to the use of the Ar-H2 mixture as a plasma supporting gas, the enthalpy of plasma flow was increased to 1.9 kW h/m 3 and the degree of powder spheroidization rose to 98% (Figure 21a) as energy density of the plasma flow per unit of surface area of the processed material was increasing. However, the content of condensed particles also rose to 11.6%. An analysis of the morphology of powder particles revealed no irregular-shaped sharp-cornered particles and a significantly smaller content of Al2O3 inclusions. Some particles contain satellites sized 1-10 µm. Figure 21b shows the granulometric composition of spherical particles at these process parameters. The increasing enthalpy of plasma flow made it possible to narrow the range of particle size distribution to 6-26 µm. The powder is characterized by unimodal particle size distribution with Dav = 13.53 µm. The distribution quantiles were as follows: D10 = 9.0 µm; D50 = 13.2 µm; and D90 = 18.7 µm. The studies revealed the most efficient powder spheroidization regime. The 95% degree of spheroidization was achieved by using the optimal design and technological parameters; the content of the nano-sized fraction was 5%. The following process characteristics of the spherical powder were identified: flowability, 20.5 s; bulk density, 4.04 g/cm 3 Figure 21. The morphology (a) and granulometric composition (b) of the NiAl-Cr-Co+15%Mo alloy powder at I pl = 1.9 kW h/m 3 , G powder = 3 kg/h, with Ar + H 2 used as a plasma supporting gas.
The studies revealed the most efficient powder spheroidization regime. The 95% degree of spheroidization was achieved by using the optimal design and technological parameters; the content of the nano-sized fraction was 5%. The following process characteristics of the spherical powder were identified: flowability, 20.5 s; bulk density, 4.04 g/cm 3 . Figure 22c shows the granulometric composition of the resulting powder. The characteristics of the powder are as follows: D av = 14.8 µm; D 10 = 10.5 µm; D 50 = 14.5 µm; and D 90 = 19.7 µm. The powder is characterized by unimodal particle size distribution within the range of 8-27 µm. Figure 22c shows the granulometric composition of the resulting powder. The characteristics of the powder are as follows: Dav = 14.8 µm; D10 = 10.5 µm; D50 = 14.5 µm; and D90 = 19.7 µm. The powder is characterized by unimodal particle size distribution within the range of 8-27 µm. Figure 22. The morphology (a,b), differential, and integral distribution (c) of the spherical powder of the NiAl-Cr-Co+15%Mo alloy produced in the efficient treatment regime. Figure 22a,b shows the morphology of the resulting NiAl-Cr-Co+15%Mo alloy powder. Powder particles have a regular spherical shape, and almost no Al2O3 inclusions are detected.
The microstructure of the surface and the transverse cross-section of the NiAl-Cr-Co+15%Mo alloy powder after plasma treatment is shown in Figure 23. The powder has a characteristic dendritic grain structure with unit cell dimension of 0.2-3 µm; the grains are formed by supersaturated solid solution of dopants (Cr, Co, Mo) in the NiAl matrix. The interdendritic space contains continuous interlayers of Cr(Mo, Co) solid solution. The resulting powders will be used in further studies of the laser powder bed fusion (LPBF) process.  Figure 22. The morphology (a,b), differential, and integral distribution (c) of the spherical powder of the NiAl-Cr-Co+15%Mo alloy produced in the efficient treatment regime. Figure 22a,b shows the morphology of the resulting NiAl-Cr-Co+15%Mo alloy powder. Powder particles have a regular spherical shape, and almost no Al 2 O 3 inclusions are detected.

Conclusions
The microstructure of the surface and the transverse cross-section of the NiAl-Cr-Co+15%Mo alloy powder after plasma treatment is shown in Figure 23. The powder has a characteristic dendritic grain structure with unit cell dimension of 0.2-3 µm; the grains are formed by supersaturated solid solution of dopants (Cr, Co, Mo) in the NiAl matrix. The interdendritic space contains continuous interlayers of Cr(Mo, Co) solid solution. The resulting powders will be used in further studies of the laser powder bed fusion (LPBF) process. Figure 22c shows the granulometric composition of the resulting powder. The characteristics of the powder are as follows: Dav = 14.8 µm; D10 = 10.5 µm; D50 = 14.5 µm; and D90 = 19.7 µm. The powder is characterized by unimodal particle size distribution within the range of 8-27 µm. Figure 22. The morphology (a,b), differential, and integral distribution (c) of the spherical powder of the NiAl-Cr-Co+15%Mo alloy produced in the efficient treatment regime. Figure 22a,b shows the morphology of the resulting NiAl-Cr-Co+15%Mo alloy powder. Powder particles have a regular spherical shape, and almost no Al2O3 inclusions are detected.
The microstructure of the surface and the transverse cross-section of the NiAl-Cr-Co+15%Mo alloy powder after plasma treatment is shown in Figure 23. The powder has a characteristic dendritic grain structure with unit cell dimension of 0.2-3 µm; the grains are formed by supersaturated solid solution of dopants (Cr, Co, Mo) in the NiAl matrix. The interdendritic space contains continuous interlayers of Cr(Mo, Co) solid solution. The resulting powders will be used in further studies of the laser powder bed fusion (LPBF) process.