Microstructure and Mechanical Properties of TaNbVTiAlx Refractory High-Entropy Alloys

A series of TaNbVTiAlx (x = 0, 0.2, 0.4, 0.6, 0.8, and 1.0) refractory high-entropy alloys (RHEAs) with high specific strength and reasonable plasticity were prepared using powder metallurgy (P/M) technology. This paper studied their microstructure and compression properties. The results show that all the TaNbVTiAlx RHEAs exhibited a single BCC solid solution microstructure with no elemental segregation. The P/M TaNbVTiAlx RHEAs showed excellent room-temperature specific strength (207.11 MPa*cm3/g) and high-temperature specific strength (88.37 MPa*cm3/g at 900 °C and 16.03 MPa*cm3/g at 1200 °C), with reasonable plasticity, suggesting that these RHEAs have potential to be applied at temperatures >1200 °C. The reasons for the excellent mechanical properties of P/M TaNbVTiAl0.2 RHEA were the uniform microstructure and solid solution strengthening effect.


Introduction
With the development of the aerospace and nuclear industries, the service temperature of traditional nickel-base superalloys is close to its melting point [1]. Therefore, the development of new refractory alloys should be accelerated. In the last decade, refractory high-entropy alloys (RHEAs), which contain five or more refractory elements with a concentration between 5 at.% and 35 at.% [2], have attracted much attention due to their high melting points and outstanding high-temperature mechanical properties. Studies have shown that at 1600 • C, NbMoTaW RHEA has a higher yield strength than the Inconel 718 and Haynes 230 alloys, which can reach 405 MPa [3]. Deriving from this alloying design strategy, a series of RHEAs have been developed, such as MoNbTaV [4], VNbMoTaW [3], and NbMoTaWTi [5]. These RHEAs are considered backup alloys for high-temperature structural materials. However, such alloys generally exhibit lower plasticity and poor machinability, making them unusable for industrial applications.
Many researchers have done a lot of work to improve the processability of RHEAs [6][7][8]. Chen et al. [9] reported that the plasticity of RHEAs can be improved by reducing the number of valence electrons in a single-phase BCC solid solution. Sheikh et al. [10] reported a new RHEA (Hf 0.5 Nb 0.5 Ta 0.5 Ti 1.5 Zr) which achieved fracture strength of 1000 MPa and fracture plasticity of 20%. Guo et al. [11] prepared a novel NbTaTiV RHEA through powder metallurgy (P/M) method, which has a yield strength of 1.37 GPa and a fracture strain of 23%. Unfortunately, although these RHEAs have high strength and acceptable plastic strain, their density is still too high, which hinders the practical application.  Figure 1 illustrates the macroscopic morphology of elemental Ta, Nb, V, Ti, and Al powders. The average particle size and impurity content are presented in Table 2. The Ta, Nb, V, and Ti powders have irregular shapes and the powder sizes are not larger than 30 µm. The Al powder has a nearly spherical shape with particle size less than 20 µm. powders have irregular shapes and the powder sizes are not larger than 30 μm. The Al powder has a nearly spherical shape with particle size less than 20 μm.  The raw powders (purity > 99.5 wt.%) with different compositions were mixed and ball-milled in a stainless steel ball mill tank. The mass ratio of stainless steel ball: powder was 10:1 and the ball mill was operated at 150 rpm for 8 h. The milled powders were then sintered by using SPS (D25/3, FCT, Rauenstein, Germany). The powders were heated to 700 °C without pressure, and then the temperature was further increased to 1700 °C by applying pressure of 30 MPa; the holding time was 10 min. The heating rate during the entire sintering process was 100 °C/min. Throughout the sintering process, the high-purity argon gas charged into the furnace as a protective atmosphere to prevent oxidation.
Test specimens (d 6 × 9 mm 3 ) for room-temperature compression and high-temperature compression were cut from the sintered RHEAs by using an electrical discharge machining (EDM) method. An INSTRON-5569 test system was used for room-temperature compression experiments with a compression rate of 10 −3 s −1 . High-temperature compression experiments were carried out with a Gleeble-3180 device with a strain rate of 10 −3 s −1 at 900 °C, 1000 °C, 1100 °C, and 1200 °C, respectively.
Particle size distributions of the powders were measured by a laser particle size analyzer (Mastersizer 3000, Malvern, UK). The impurity elements in the powders were detected by an elemental analyzer (TCH600, LECO, San Jose, America). Phase analyses were performed by Cu-Ka target X-ray diffractometer (XRD, D/max 2500, CORP, Tokyo, Japan). Microstructures were observed via electron microscopy (SEM, G3-US, FEI, Hillsboro, America) with an electron backscatter diffraction (EBSD) device. A field emission probe microscope (EPMA, JXA-8530F, JEOL, Tokyo, Japan) was used to characterize the element distribution of the samples.  The raw powders (purity > 99.5 wt.%) with different compositions were mixed and ball-milled in a stainless steel ball mill tank. The mass ratio of stainless steel ball: powder was 10:1 and the ball mill was operated at 150 rpm for 8 h. The milled powders were then sintered by using SPS (D25/3, FCT, Rauenstein, Germany). The powders were heated to 700 • C without pressure, and then the temperature was further increased to 1700 • C by applying pressure of 30 MPa; the holding time was 10 min. The heating rate during the entire sintering process was 100 • C/min. Throughout the sintering process, the high-purity argon gas charged into the furnace as a protective atmosphere to prevent oxidation.

Results
Test specimens (d 6 × 9 mm 3 ) for room-temperature compression and high-temperature compression were cut from the sintered RHEAs by using an electrical discharge machining (EDM) method. An INSTRON-5569 test system was used for room-temperature compression experiments with a compression rate of 10 −3 s −1 . High-temperature compression experiments were carried out with a Gleeble-3180 device with a strain rate of 10 −3 s −1 at 900 • C, 1000 • C, 1100 • C, and 1200 • C, respectively.
Particle size distributions of the powders were measured by a laser particle size analyzer (Mastersizer 3000, Malvern, UK). The impurity elements in the powders were detected by an elemental analyzer (TCH600, LECO, San Jose, America). Phase analyses were performed by Cu-Ka target X-ray diffractometer (XRD, D/max 2500, CORP, Tokyo, Japan). Microstructures were observed via electron microscopy (SEM, G3-US, FEI, Hillsboro, America) with an electron backscatter diffraction (EBSD) Entropy 2020, 22, 282 4 of 13 device. A field emission probe microscope (EPMA, JXA-8530F, JEOL, Tokyo, Japan) was used to characterize the element distribution of the samples. Figure 2 shows the XRD patterns of the TaNbVTiAl x RHEAs. From Figure 2a, it can be seen that the TaNbVTiAl x RHEAs had three diffraction peaks, which corresponded to the (110), (200), and (211) crystal planes, respectively. The positions of these crystal planes were consistent with those of the diffraction peaks of a BCC single phase. Therefore, all the TaNbVTiAl x RHEAs exhibited a typical BCC solid solution microstructure. As shown in Figure 2b, the diffraction peak of the (110) shifted significantly from 39.26 • to 39.68 • as the Al content increased. It means that the (110) crystal plane lattice parameter gradually reduced with the increase of the Al content. Table 3 lists the crystal structures, calculated lattice parameters, theoretical lattice parameters, calculated melting temperatures, and theoretical densities of these alloys. The calculated lattice parameters were obtained based on the XRD results by a formula as follows [18]:

Microstructures of the TaNbVTiAl x RHEAs
where d is the crystal surface spacing; θ is the diffraction angle; λ is the X-ray diffraction wavelength (λ = 1.5406 Å); a is the lattice constant; h, k, l are the crystal surface index.   In addition, the theoretical lattice parameter of these RHEAs can be calculated by the element's theoretical lattice parameter of the RHEA. The calculation formula is as follows [19]: where c i is the content of the i component (at.%); a i is the lattice constant of the i component. As seen in Table 3, with an increase of Al content, the calculated lattice parameter decreased from 3.243 Å to 3.210 Å, while the theoretical lattice parameter increased from 3.230 Å to 3.394 Å. The reason for the difference is that the p-layer saturated electron orbit of Al atom was easy to hybridize with the d-layer unsaturated electron orbit of transition metal to form a covalent bond, while the length of the covalent bond was small [20]. Therefore, with the increase of Al content, the hybrid effect was more obvious, and the lattice constant decreased. The atomic radius of Al is smaller than that of Ta, Nb, and Ti, and obviously larger than that of V. Moreover, the single-phase BCC solid solution structure was always present in these alloys, which indicated that Al could easily form a solid solution with the other elements, and enhance the high-entropy effect of the alloy. Therefore, the increase of the Al element led to the increase of lattice distortion, and eventually, led to the difference between calculated lattice parameters and theoretical lattice parameters. Similar results have been reported in NbTiMoVAl x RHEAs [18]. The theoretical densities of TaNbVTiAl x RHEAs were calculated by the following formula [21]: where the c i , M i , and ρ i represent the concentration, molar mass, and theoretical density of the i component, respectively.
As the Al content increased, the density of the TaNbVTiAl x RHEAs decreased gradually from 9.16 g/cm 3 of Al0 alloy to 7.89 g/cm 3 of Al1.0 alloy, which was very close to that of nickel-based superalloys [22]. Figure 3 illustrates the inverse pole figure (IPF) and grain distribution maps of the TaNbVTiAl x RHEAs in the transverse direction. The average grain sizes were 69, 101, 106, 135, 147, and 187 µm for the Al0, Al0.2, Al0.4, Al0.6, Al08, and Al1.0, respectively. The results indicated that the higher the Al atom ratio, the larger the grain size of TaNbVTiAl x RHEAs, which may be because of the higher the atomic ratio of aluminum, the lower the melting point of the alloy. At the same sintering temperature, the lower the melting point of the alloy, the closer it was to the sintering temperature and the faster the grain growth rate, resulting in an increase of grain size. In addition, there were no obvious textures in the as-sintered RHEAs, as shown in Figure 3. Figure 4 shows the EPMA results of the TaNbVTiAl 1.0 RHEA. It can be found that the distribution map of all elements shows a single color without obvious bright spots. Therefore, the distribution of all elements was uniform and no significant segregation was observed.      1932 MPa, and 15%, respectively. With an increase in the Al content to Al0.2, the yield strength and compressive strength increased to 1835 MPa and 2217 MPa, indicating that the addition of Al alloy elements significantly improved the strength, while the compressive strain decreased to about 10%. With the further increase of Al content, a gradual decrease in compressive strength occurred, while the plasticity continued to decrease. When x = 1.0, the yield strength, compressive strength, and plastic strain decreased to only 1450 MPa, 1619 MPa, and 2.5%, respectively. Table 4 lists the mechanical properties of the typical RHEAs reported in the literature and this work. It is shown that the Al0.2 RHEA had the highest specific yield strength, which is higher than most of the RHEAs, such as NbMoTaW [3], TaNbHfZr [23], TiZrNbTa [19], TaNbHfZrTi [24], Al 0.4 Hf 0.6 NbTaTiZr [12], and Al 0.21 HfNbTiZr [25] RHEAs, with reasonable compressive plasticity of about 10%. Compared with NbTiVTaAl x [15] prepared by arc melting, the specific strength was greatly improved. Figure 6 shows the morphologies of the fracture surface of the TaNbVTiAl x RHEAs. The fracture surfaces of all the RHEAs show classic rivers and step shapes, suggesting that the fracture mode was a typical brittle cleavage fracture.  Figure 5 illustrates the room-temperature compressive stress-strain curves of the TaNbVTiAlx RHEAs. The yield strength, compressive strength, and plastic strain of the Al0 alloy were 1391 MPa, 1932 MPa, and 15%, respectively. With an increase in the Al content to Al0.2, the yield strength and compressive strength increased to 1835 MPa and 2217 MPa, indicating that the addition of Al alloy elements significantly improved the strength, while the compressive strain decreased to about 10%. With the further increase of Al content, a gradual decrease in compressive strength occurred, while the plasticity continued to decrease. When x = 1.0, the yield strength, compressive strength, and plastic strain decreased to only 1450 MPa, 1619 MPa, and 2.5%, respectively. Table 4 lists the mechanical properties of the typical RHEAs reported in the literature and this work. It is shown that the Al0.2 RHEA had the highest specific yield strength, which is higher than most of the RHEAs, such as NbMoTaW [3], TaNbHfZr [23], TiZrNbTa [19], TaNbHfZrTi [24], Al0.4Hf0.6NbTaTiZr [12], and Al0.21HfNbTiZr [25] RHEAs, with reasonable compressive plasticity of about 10%. Compared with NbTiVTaAlx [15] prepared by arc melting, the specific strength was greatly improved. Figure 6 shows the morphologies of the fracture surface of the TaNbVTiAlx RHEAs. The fracture surfaces of all the RHEAs show classic rivers and step shapes, suggesting that the fracture mode was a typical brittle cleavage fracture.       Figure 7 shows the high-temperature compressive properties of the Al0.2 RHEA. It is shown that the Al0.2 alloy had a yield strength of 783 MPa at 900 • C (specific yield strength was about 88.37 MPa*cm 3 /g). As can be seen from Figure 7b, the high-temperature (<1000 • C) specific strength was better than that of the typical NbMoTaW RHEA [3], TaNbHfZrTi RHEA [26], NbTaTiV RHEA [11], Ni-based IN718 alloy, and Haynes 230 alloy [3]. This is because the single-phase BCC structure with multiple components and melting elements had slower element diffusion at higher temperatures [27]. Therefore, the high-temperature softening resistance of the alloy could be improved. At 1200 • C, the Al0.2 RHEA still kept a yield strength of 142 MPa, suggesting that the material has the possibility for use at high temperatures (>1200 • C).

Mechanical Properties of the TaNbVTiAlx RHEAs
better than that of the typical NbMoTaW RHEA [3], TaNbHfZrTi RHEA [26], NbTaTiV RHEA [11], Ni-based IN718 alloy, and Haynes 230 alloy [3]. This is because the single-phase BCC structure with multiple components and melting elements had slower element diffusion at higher temperatures [27]. Therefore, the high-temperature softening resistance of the alloy could be improved. At 1200 °C, the Al0.2 RHEA still kept a yield strength of 142 MPa, suggesting that the material has the possibility for use at high temperatures (>1200 °C).

Phase Prediction
Predicting the phase structure of high-entropy alloys (HEAs) is a challenging task. At present, researchers have found a semiempirical method to determine the generation of a solid solution in HEAs [28]. According to the literature, the main factors affecting structures are as follows: the enthalpy of mixing (−15 kJ/mol ≤ ΔHmix ≤ 5 kJ/mol), radius asymmetry (δ < 6.6%) and entropy/enthalpy ratio (Ω > 1.1) [29,30]. Studies have [31,32] calculated the valence electron concentration (VEC) values through the relevant parameters of the HEAs, thereby obtaining the structure type of the solid solution. When the VEC ≤ 6.87, the HEAs generally exhibit BCC solid solution structure. If the VEC is between 6.78 and 8, they usually show a BCC + FCC two-phase solid solution structure. While when the VEC ≥ 8, the HEAs are mainly FCC solid solution phase. The calculation formula for the main parameters is as follows [33]:

Phase Prediction
Predicting the phase structure of high-entropy alloys (HEAs) is a challenging task. At present, researchers have found a semiempirical method to determine the generation of a solid solution in HEAs [28]. According to the literature, the main factors affecting structures are as follows: the enthalpy of mixing (−15 kJ/mol ≤ ∆Hmix ≤ 5 kJ/mol), radius asymmetry (δ < 6.6%) and entropy/enthalpy ratio (Ω > 1.1) [29,30]. Studies have [31,32] calculated the valence electron concentration (VEC) values through the relevant parameters of the HEAs, thereby obtaining the structure type of the solid solution. When the VEC ≤ 6.87, the HEAs generally exhibit BCC solid solution structure. If the VEC is between 6.78 and 8, they usually show a BCC + FCC two-phase solid solution structure. While when the VEC ≥ 8, the HEAs are mainly FCC solid solution phase. The calculation formula for the main parameters is as follows [33]:  [34]; (Tm)i is the theoretical melting temperature of the i component element; VECi is the valence electron concentration of the i component element [31]. Referring to the above semiempirical criteria, the calculation results of TaNbVTiAl x RHEAs are shown in Table 5. It is shown that all the TaNbVTiAl x RHEAs were located in the region of BCC solid solution. The prediction results were basically consistent with the conclusions of this study.

Strengthening Mechanism
The TaNbVTiAl x RHEAs prepared by a P/M method exhibited high specific strength and reasonable plasticity both at room temperature and high temperature. The high yield strength of the RHEAs may have come from the uniform microstructure and the solid solution strengthening effect, while the reasonable plasticity may have resulted from the ductile TaNbVTi matrix, which has similar BCC microstructure.
The yield strength and compressive strength of the TaNbVTiAl x RHEAs first increased and then decreased with the increase of Al content. Compression plasticity was decreasing. The Al0.2 RHEA had the highest yield strength and compressive strength. According to the traditional theory of solution strengthening [15], the yield strength should increase with the increase of Al content, but the actual situation was different. This shows that the traditional solid solution strengthening theory cannot fully explain the relationship between strength and solute Al. Moreover, Al atoms formed covalent bonds with other alloy atoms, which made the strengthening method more complex.
In addition to solution strengthening, the effect of grain size should also be considered. Fine grains can produce fine grain strengthening benefits. With the increase of Al content, the grain size increased from 69 µm to 187 µm. This is one reason for the decrease in strength and plasticity. This change in compressive strength may have been caused by the change in lattice constant, which resulted from the addition of element Al. The higher the Al content, the smaller the lattice constant of TaNbVTiAl x RHEAs. With the increase of Al content, the effect of covalent bond formed by the hybridization of p-layer saturated electron orbit and d-layer orbit of transition metal was more obvious. The covalent bond had a small length, which led to the decrease of the lattice constant. The change trend of the lattice constant was consistent with that of crystal surface spacing. As the Al content increased, the lattice constant decreased, so the interplanar spacing became smaller and the dislocation became more difficult to slip, making the yield strength increase [35,36]. When the Al content exceeded Al0.2, continuing to increase the Al content and lowering the interplanar spacing may have made the dislocations difficult to move, and thus generated dislocation accumulation and local stress concentration. When the stress concentration cannot be released, the alloy will break. This may be the main reason why the compressive strength and compression plasticity of TaNbVTiAl x RHEAs decreased when the Al content exceeded Al0.2. These results are highly consistent with the as-cast TaNbVTiAl x [15] and NbTiMoVAl x [18] RHEAs.