Microstructure and Mechanical Properties of Low-Density, B2-Ordered AlNbZrTi_x Multi-Principal Element Alloys

Abstract

Low-density multi-principal element alloys (MPEAs) combining a high specific strength and considerable ductility have remained a research hotspot, due to their promising prospects for energy-saving industrial applications. Light Ti-containing AlNbZrTi_x (x = 1−3) MPEAs were designed and prepared by induction melting and annealing. As the Ti content increases, the microstructure of these MPEAs evolves from dual phase (B2-ordered and Zr₅Al₃-type structure) into a single-phase B2-ordered structure, while the density reduces by ~8.7%, from ~5.85 g·cm⁻³ (x = 1) to ~5.34 g·cm⁻³ (x = 3). Unexpectedly, the AlNbZrTi_x (x = 1, 2, 3) alloys possess high specific yield strengths of ~270 kPa·m³·kg⁻¹, ~221 kPa·m³·kg⁻¹, >208 kPa·m³·kg⁻¹, along with excellent fracture strains of ~17.8%, 21.8%, and >50%, respectively. These combined compressive properties are superior to the reported data of most BCC/B2-dominant MPEAs. The deformation mechanism of the B2-ordered structure is explained as a dislocation-based mechanism, accompanied by antiphase domains. Here, the effect of Ti on the microstructure and compressive properties of AlNbZrTi_x MPEAs was investigated, providing scientific support for the development of advanced low-density materials.

1. Introduction

Multi-principal element alloys (MPEAs), also known as high-entropy alloys because of their high mixing entropy of four or more principal elements, have received extensive attention from the scientific community, due to their unique microstructure and unexpected mechanical properties [1,2,3]. To date, various preparation approaches have produced single-phase solid-solution MPEAs, e.g., face-centered cubic (FCC) MPEAs with excellent room-temperature ductility and cryogenic damage tolerance [4]; body-centered cubic (BCC) MPEAs with superior strength [5,6]; and hexagonal-close-packed (HCP) MPEAs combined with rare-earth elements [7]. Unfortunately, most high-strength BCC MPEAs so far have been affected by a low ductility at room temperature, suggesting a BCC strength–ductility trade-off dilemma [8,9]. For example, Senkov et al. reported that systematic BCC Al-Mo-Nb-Ta-Ti-Zr MPEAs exhibited a room-temperature fracture strain of ≤4.1%; although, among which, Al_0.5Mo_0.5NbTa_0.5TiZr alloy possessed a highest compressive strength of 2350 MPa [5]. Research efforts into improving the room-temperature ductility of BCC-based MPEAs should be urgent, for further industrial applications.

Low-density (e.g., lightweight) designs of MPEAs with high performance and energy efficiency have been extensively reported recently, promoting the development of second-generation structural materials [10]. In addition to the previously published AlLiMg-based FCC MPEAs with low density (<3 g·cm³) [11], another type of low-density MPEAs with BCC structure has been designed through selecting elements with a light weight (e.g., Al and Ti) and high melting points (Cr, Nb, V, and Zr) [12,13]. For example, Yurchenko et al. revealed that ordered BCC (B2) AlNbTiVZr_x (x = 0−1.5) alloys possessed a low density of 5.53–5.87 g·cm³ and high room-temperature yield strength of 1000–1535 MPa [6]. Senkov et al. proposed a Cr-Nb-Ti-V-Zr alloy system that had an excellent yield stress, even at 1000 °C [12,14]. Among these Cr-Nb-Ti-V-Zr alloys, BCC NbTiVZr and NbTiV₂Zr could be plastically deformed up to 50% of strain without fracture, but two Cr-added CrNbTiZr and CrNbTiVZr alloys had limited fracture strains of 6% and 3%, respectively [14]. This is because the Cr element added to the two BCC alloys promotes the formation of Laves secondary phase, thereby deteriorating the ductility. Low-density MPEAs with a single-phase BCC structure usually possess a better ductility than those with multiple phases [8,15]. Fortunately, light Al elements, differently from the heavy Cr element, can stabilize the BCC structures of MPEAs and simultaneously lighten their weight [16]. Thus, our previous publication, replacing Cr by Al in CrNbTiZr alloy, found that the density of a B2-based AlNbTiZr with mirror Zr₅Al₃-type phase was reduced to ~5.85 g·cm³ and the Laves secondary phase disappeared [17]. Meanwhile, Qiu et al. reported that the B2 phase is more stable than the disordered BCC phase [13]. Moreover, a light Ti element was demonstrated to be a BCC/B2 stabilizer [18,19]. To overcome the poor ductility in most lightweight Al-containing MPEAs, an increment of Ti content, to stabilize single-phase B2 structure, may provide an effective pathway for a better strength–ductility combination at room temperature. Here, the effect of light Ti content on the microstructure and mechanical properties of AlNbZrTi_x MPEAs was investigated, aiming to simultaneously obtain a low density, high specific yield strength (strength-to-density ratio, SYS), and superior ductility at room temperature.

2. Materials and Methods

The AlNbZrTi_x (x = 1, 1.5, 2, 3) samples were prepared by vacuum induction melting. Each ingot was melted at least five times with pure Al, Nb, Zr, and Ti (≥99.9 wt.%). The produced samples were annealed at 1473 K for 5 h in a vacuumed quartz tube, and subsequently furnace cooled to room temperature for further investigation [6,12,17]. The crystal structure of the annealed samples was characterized using X-ray diffraction (XRD, Bruker D8 Advance, Karlsruhe, Germany) with Cu-Kα radiation. XRD measurement was conducted from 20° to 100° at a scanning rate of 4°/min. The microstructure of the annealed samples was studied using transmission electron microscopy (TEM, FEI Titan Themis, Hillsboro, OR, USA), scanning transmission electron microscopy (STEM), and scanning electron microscopy (SEM, Zeiss Sigma 350, Jena, Germany). Thin samples for TEM observations were prepared by ion milling using 3 keV Ar ions and cooled by liquid N₂ during ion milling. The SEM was accompanied by backscatter electron (BSE) technology, energy-dispersive spectrometer (EDS) technology, and electron backscatter diffraction (EBSD, Oxford Aztec, Abingdon, UK) technology. The samples for SEM-BSE observation were treated by mechanical polishing, and those for EBSD measurement were electro-polished in a solution of 10% perchloric acid and 90% alcohol at sub-zero temperatures [17]. The EBSD test was performed at scanning steps of 5 μm.

The density of the annealed samples was measured using the hydrostatic weighting method, and three samples of each alloy were tested to obtain the average density value. The samples for compression tests were cut from annealed ingots with dimensions of ~4.7 × 4.7 × 7.7 mm³. Compression tests of at least three specimens were carried out at room temperature, using a strain rate of 1 × 10⁻⁴ s⁻¹. The microstructure measurement of the 20%-strained AlNbZrTi₃ sample was conducted on the longitudinal section.

3. Results and Discussion

3.1. Phase Formation

The phase formation of alloys is dependent on the Gibbs free energy (∆G_mix) formula, as presented in Equation (1) [20]. The negative ∆G_mix is associated with reducing the mixing enthalpy (∆H_mix, Equation (2)) and increasing the mixing entropy (∆S_mix, Equation (3)), which favors the formation of a single-phase solid solution. Related abbreviations/nomenclature is shown in

Table A1. Abbreviations/Nomenclature.

Abbreviations	Nomenclature	Definition
MPEA		Multi-principal element alloy
FCC		Face-centered cubic
BCC		Body-centered cubic
HCP		Hexagonal-close-packed
B2		Ordered body-centered cubic
SYS		Specific yield strength (strength-to-density ratio)
XRD		X-ray diffraction
TEM		Transmission electron microscopy
STEM		Scanning transmission electron microscopy
SEM		Scanning electron microscopy
BSE		Backscatter electron
EDS		Energy-dispersive spectrometer
EBSD		Electron backscatter diffraction
HAADF		High-angle annular dark-field
FFT		Fourier transformation
IPF		Inverse pole figure
KAM		Kernel average misorientation
SAED		Selected area electron diffraction
	∆Gmix	Gibbs free energy
	∆Hmix	Mixing enthalpy
	∆Smix	Mixing entropy
	δ	Atomic misfit
	Ω	Multi-component solid solution rule
	VEC	Valence electron concentration
	Tm	Average calculated melting point
	Ti	Melting point of ith pure metal
	ci	Atomic fraction of the ith constituent
	cj	Atomic fraction of the jth constituent
	${\Delta H}_{mix}^{ij}$	ith-jth-constituent mixing enthalpy
	R	Gas constant
	ri	ith-constituent radius
	(VEC)i	Valence electron concentration of the ith constituent

. Owing to the component complexity of MPEAs, this explanation has its limitations. In order to predict the phase formation for MPEA design, three parameters, δ, Ω and VEC, were proposed [21,22]. The above parameters were defined as follows:

$$\mathsf{\Delta }{G}_{\mathrm{mix}}={\mathsf{\Delta }H}_{\mathrm{mix}}-T_m{\mathsf{\Delta }S}_{\mathrm{mix}},$$

(1)

$$\mathsf{\Delta }{H}_{\mathrm{mix}}={\sum }_{i=1,i\ne j}^n{4}_{\mathrm{CiCj}}\mathsf{\Delta }{H}_{\mathrm{mix}}^{\mathrm{ij}},$$

(2)

$$\Delta S_{\mathrm{mix}}=-R{\sum }_{i=1}^n{c}_i{\mathrm{lnc}}_i,$$

(3)

$$T_m={\sum }_{i=1}^n{c}_i{T}_i,$$

(4)

$$\mathsf{\delta }={({\sum }_{i=1}^n{c}_i{(1-r_i/{\sum }_{i=1}^n{c}_i{r}_i)}^2)}^{1/2},$$

(5)

$$\Omega =T_m\Delta S_{\mathrm{mix}}/\left|\Delta H_{\mathrm{mix}}\right|,$$

(6)

$$\mathrm{VEC}={\sum }_{i=1}^n{c}_i{(\mathrm{VEC})}_{i,}$$

(7)

where T_m is the average calculated melting point, c_i and c_j represent atomic fraction of the i_th and j_th constituents, ${\mathsf{\Delta }\mathrm{H}}_{\mathrm{m}\mathrm{i}\mathrm{x}}^{\mathrm{i}\mathrm{j}}$ denotes i_th-j_th-constituent mixing enthalpy, R is gas constant, T_i is melting point of i_th pure metal, δ denotes atomic misfit, r_i is i_th-constituent radius, Ω represents a multi-component solid solution rule, and VEC is the valence electron concentration. When Ω ≥ 1.1 and δ ≤ 6.6%, a single-phase solid-solution structure becomes easier to form [22,23,24]. Additionally, VEC is helpful for predicting a solid–solution structure. Guo et al. [25] revealed that the range of VEC < 6.87, 6.87 ≤ VEC ≤ 8, and VEC ≥ 8 favors the stabilization of BCC and FCC, and the coexistence of BCC and FCC phases, respectively.

The i_th-j_th-constituent mixing enthalpy in the AlNbZrTi_x alloys is listed in

Table 1. The i_th-j_th-constituent mixing enthalpy in the AlNbZrTi_x alloys [26].

${\mathrm{∆}\mathrm{H}}_{\mathrm{m}\mathrm{i}{x}}^{\mathrm{i}\mathrm{j}}$	Al	Nb	Zr	Ti
Al	-	−18.2	−43.7	−29.5
Nb	-	-	3.9	2
Zr	-	-	-	−0.2
Ti	-	-	-	-

. The parameters calculated according to Equations (2) to (7) are listed in

Table 2. The calculated parameters of the AlNbZrTi_x alloys.

Alloys	Tm (K)	∆Hmix (kJ·mol−1)	∆Smix (J·K−1·mol−1)	δ (%)	Ω	VEC
AlNbZrTi	1935.74	−21.4250	11.53	3.85	1.04	4
AlNbZrTi1.5	1936.56	−19.6642	11.38	3.70	1.12	4
AlNbZrTi2	1937.22	−18.1440	11.08	3.57	1.18	4
AlNbZrTi3	1938.21	−15.6778	10.33	3.34	1.28	4

. With the increment of the Ti in these MPEAs, the δ critical value below of 6.6% decreased from 3.85% (x = 1) to 3.34% (x = 3), by 13.2%, and the Ω increased from 1.04 (x = 1), to exceed the critical value of 1.1 (x = 1.5, 2, 3). This suggests the easier formation of a single-phase solid–solution structure. Additionally, the calculated VEC value (4) of these alloys belongs to the VEC range (<6.87), which favors the stabilization of the BCC structure. Thus, this means that the Ti element stabilized the single-phase BCC structure in the AlNbZrTi_x alloys.

Interestingly, a specialty of a Ω value (1.04) below the critical value (1.1) appears in AlNbZrTi alloy, which promotes the ability to form a multi-phase structure. Our previous report confirmed that the AlNbZrTi alloy exhibited dual phases (B2 and Zr₅Al₃-type structure) [17]. It is speculated that the lower mixing enthalpy of Al-Zr components (Table 1) is responsible for the Zr₅Al₃-type phase. Meanwhile, the ordering of the BCC structure observed in Al-Ti-containing MPEAs, i.e., AlNbTiVZr_x [6] and AlTiVCr [13], was ascribed to the negative mixing enthalpy of Al with other elements (Table 1). This becomes a dominant factor in facilitating the formation of B2. Based on the site occupations in ordered Ti alloys [6,27], B2 AlNbZrTi_x MPEAs should have one sublattice preferably occupied by Ti and the other sublattice preferably occupied by Al, while Nb and Zr are present on both sublattices.

3.2. Microstructural Characterization

The XRD patterns of the AlNbZrTi_x alloys are shown in Figure 1. It can be seen that AlNbZrTi alloy contains a combination of a dominant BCC and minor Zr₅Al₃-type structure, as also identified in our previous report [17]. The superlattice peak at ~27° demonstrates the ordering of the BCC structure (CsCl-type B2, Pm3m). As the Ti content increases in AlNbZrTi_x (x = 1.5−3), the XRD intensity of the Zr₅Al₃-type phase decreases (x = 1.5) and finally disappears (x = 2, 3). It can be speculated that the Ti increment in these MPEAs results in the formation of a single-phase B2 structure. Additionally, the {110}_B2 peak at ~38° reveals a right shift and suggests a reduction in the lattice parameters of the B2 phase, which were calculated to be ~0.3336 nm, ~0.3325 nm, ~0.3317 nm, and ~0.3309 nm in the AlNbZrTi_x (x = 1, 1.5, 2, 3) alloys, respectively. This reduction of B2 lattice parameters is attributed to a weakened lattice distortion induced by a reduced δ value (Table 2).

To understand the phase variations, SEM-BSE images of AlNbZrTi_x (x = 1–3) alloys are shown in Figure 2. As seen in Figure 2a, the AlNbZrTi alloy possesses an equiaxed-grain B2 matrix (Region I), accompanied b polygonal (Region II) and rod-like (Region III) phases. The polygonal phase is at the matrix interior, while the rod-like phase is distributed discontinuously along the B2 grain boundaries. In AlNbZrTi_1.5 alloy, the polygonal and rod-like phase becomes irregular and forms a continuous net along the B2 grain boundaries, respectively (Figure 2b). Additionally, the amount of secondary phase decreases. In AlNbZrTi₂ and AlNbZrTi₃ alloys, the B2 matrix remains equiaxed and the secondary phase disappears. This trend in the amount of secondary phase is consistent with the XRD peak intensity of Zr₅Al₃-type structure in Figure 1. Meanwhile, the EDS results, as listed in

Table 3. EDS results of the AlNbZrTi_x alloys.

Alloys	Region	Chemical Compositions (at.%)
		Al	Nb	Zr	Ti
AlNbZrTi	Nominal	25	25	25	25
	I	23.1	24.7	25.2	27
	II	34.2	12.9	41.1	11.8
	III	35.1	12.8	41.2	10.9
AlNbZrTi1.5	Nominal	22.2	22.2	22.2	33.4
	I	21.9	21.5	22.7	33.9
	II	33.5	12.9	38.3	15.3
	III	28.7	12.9	40.9	17.5
AlNbZrTi2	Nominal	20	20	20	40
	I	19.9	19.3	20.5	40.3
AlNbZrTi3	Nominal	16.7	16.7	16.7	49.9
	I	16.3	17.0	16.9	49.8

, reveal that the matrix composition was close to the nominal composition; where the deviation between the matrix and nominal composition is less than ~7.6%. Both the polygonal and rod-like phases are enriched in Zr and Al, and the content of Zr is greater than that of Al. This Zr-Al-rich phenomenon confirms the Zr₅Al₃-type structure of the secondary phase shown in Figure 2. This Zr₅Al₃-type phase is also observed in AlNbTiVZr_x (x = 0.1–1.5) alloys, because the enthalpy of the binary Zr₅Al₃-type phase formation was very negative (−51.5 kJ·mol⁻¹) [6,28].

Al-Ti-containing MPEAs, such as Al_0.5NbTa_0.8Ti_1.5V_0.2Zr [29] and Al_0.25NbTaTiZr [5], have been reported to usually possess a nanoscale BCC/B2 mixture with similar lattice parameters, which is difficult to identify by XRD analysis. To distinguish disordered BCC and/or B2, an aberration-corrected STEM high-angle annular dark-field (HAADF) image of AlNbZrTi₃ is present in Figure 3. The 1/2{100} lattice fringes with lightly dark contrast and the corresponding Fourier transformation (FFT) inset show a weak {100} superlattice spot, confirming a single B2 structure rather than a nanoscale BCC/B2 mixture. The spacing of (100) and $(01\overline{1})$ planes in Figure 3 indicates a lattice parameter of ~0.34 nm, in approximate agreement with the XRD results. The alternately dark and bright distribution of the 1/2{100} lattice fringes confirms the preferable entrance of light Al atoms into one sublattice of the CsCl-type B2 structure. In summary, when the Ti content is increased in AlNbZrTi_x (x = 1~3), the microstructure transforms from a mixture of B2 matrix and mirror Zr₅Al₃-type phase, to a single-phase B2 structure, confirming that the Ti element stabilizes the B2 structure rather than the B2/BCC mixture.

3.3. Mechanical Property: Compressive Stress–Strain Analysis

Compressive engineering stress–strain curves of the AlNbZrTi_x (x = 1−3) alloys at room temperature are presented in Figure 4a. As the Ti content increases, the yield strength continuously declines, from ~1579 MPa (x = 1) to ~1111 MPa (x = 3). The corresponding mechanical properties are also listed in

Table 4. The actual density and mechanical properties of the AlNbZrTi_x alloys.

Alloys	Actual Density (g·cm−3)	Yield Strength (MPa)	Fracture Strain (%)	Strength-to-Density Ratio (SYS, kPa·m3·kg−1)
AlNbTiZr	5.85	1579	17.8	270
AlNbTiZr1.5	5.72	1364	7.8	238
AlNbTiZr2	5.56	1227	21.8	221
AlNbTiZr3	5.34	1111	>50	>208

. The downward trend in the yield strength of AlNbZrTi_x alloys is explained by two aspects. First, the strengthening contribution of the brittle Zr₅Al₃-type phase to the yield strength becomes less efficient, because the amount of Zr₅Al₃-type hard phase decreases as the Ti content increases in the MPEAs (Figure 2). Second, a drop in ∆H_mix (Table 2) suggests a lower chemical affinity of the multiple elements in the B2 matrix, which may be responsible for the decreased yield strength [26].

As shown in Figure 4a and Table 4, the fracture strain of the AlNbZrTi_x (x = 1, 2, 3) alloys were ~17.8%, 21.8%, and >50%, respectively, except for AlNbZrTi_1.5, which had a poor fracture strain of ~7.8%. Note that no macroscopic fracture was observed in the AlNbZrTi₃ alloy before 50% strain. In these MPEAs consisting of a B2 matrix and Zr₅Al₃-type phase (Figure 2), the former is relatively softer and provides a larger contribution to the ductility than the latter. It is well known that a high δ in the MPEAs is a dominant factor in causing severe lattice distortion, which would deteriorate the ductility [10]. Thus, as the Ti content increases in the AlNbZrTi_x (x = 1, 2, 3) alloys, the improvement of the fracture strain can be attributed to the reduced δ (Table 2) and gradual disappearance of the Zr₅Al₃-type phase. Formation of a single-phase B2 structure is beneficial for improving the ductility in these MPEAs. Exceptionally, compared to AlNbZrTi alloy, AlNbZrTi_1.5 alloy possesses a lower δ, but a poor fracture strain. The morphology and distribution transition of a hard Zr₅Al₃-type phase (Figure 2) provide a reasonable explanation. When Ti was added to AlNbZrTi_x alloys from x = 1 to x = 1.5, rod-like Zr₅Al₃-type phase becomes a continuous net distributed along the B2 grain boundaries and is a cause of stress-induced intergranular fracture; while irregular Zr₅Al₃-type phase at the matrix interior has some sharp tips, where the deformation-induced stress is concentrated. Consequently, the fracture strain becomes poor only in the AlNbZrTi_1.5 alloy.

The actual densities of the AlNbZrTi_x (x = 1, 1.5, 2, 3) MPEAs were measured as ~5.85 g·cm⁻³, ~5.72 g·cm⁻³, ~5.56 g·cm⁻³, and ~5.34 g·cm⁻³, respectively. This suggests a lowered density of those MPEAs through the addition of the light-weight Ti element. These AlNbZrTi_x MPEAs are lighter than most reported MPEAs (≥6 g·cm⁻³) [6,8]. Regarding yield strength and actual density, the SYS versus compressive fracture strain of the AlNbZrTi_x and previously reported BCC/B2-dominant MPEAs are displayed in Figure 4b. As seen, the fracture strain of most single-phase BCCs (~5–20%) is better than most multi-phase MPEAs (≤ ~10%). The SYS of the AlNbZrTi_x (x = 1, 1.5, 2, 3) MPEAs were calculated as ~270 kPa·m³·kg⁻¹, ~238 kPa·m³·kg⁻¹, ~221 kPa·m³·kg⁻¹, >208 kPa·m³·kg⁻¹, respectively. Except for AlNbZrTi_1.5 with a limited ductility of ~7.8%, the AlNbZrTi_x (x = 1, 2, 3) have, so far, a better combination of the SYS and fracture strain than any of the other BCC/B2-dominant alloys; with a fracture strain of <50% [30,31,32,33]. This suggests a combination of high strength, low density, and excellent ductility for the AlNbZrTi_x (x = 1, 2, 3) MPEAs.

Compared to binary B2 materials with inherent brittleness, e.g., CoTi, CoAl, and NiAl [34], the single-phase B2 AlNbZrTi₃ alloy reported here has exceptional room-temperature ductility. To uncover the underlying compressive deformation mechanisms of the single-phase B2 structure, the microstructure of 20%-strained AlNbZrTi₃ was examined using EBSD, TEM, and STEM-HAADF, as present in Figure 5. The EBSD inverse pole figure (IPF) image in Figure 5a shows elongated grains, in which the color gradients suggest compression-induced deformation bands. Notably, two typical deformation band zones, denoted by A and B in Figure 5a, have misorientation relations of 20°~25°<110> and 30°~40°<123>, which are inconsistent with potential twin relations in B2 materials [35]. This suggests the deformation bands are not twins. Similarly, Yamaguch et al. [34] also reported that superdislocation should exhibit a twinning sense. In Figure 5b, the EBSD kernel average misorientation (KAM) image shows local misorientations, mainly ranging from 0.2° to 1.8°, and delimits some lenticular-like structures spreading from the grain boundaries into the grain interior. The low-magnification TEM image in Figure 5c displays the two dislocation slip bands responsible for the local misorientations and for the lenticular-like structures, while the high-magnification TEM micrograph in the lower-left inset shows the dislocation morphology. This dislocation-mediated lenticular-like structure was also observed in the deformation of a previous BCC TiHfZrTaNb alloy [36]. The inset of the [011]_B2 selected area electron diffraction (SAED) pattern in the upper-right corner in Figure 5c indicates that the phase structure is B2 or a B2/BCC mixture. Although the atomic resolution STEM-HAADF technique can only image the edge component of dislocations, but is not able to see screw dislocations, it can easily trace the antiphase domains caused by the motion of dislocations. As massive atomic resolution STEM-HAADF observations revealed few edge- or mixed-type dislocations, but antiphase domains were frequently observed, it is reasonable to suggest the existence of extensive screw dislocations. An example of an antiphase domain is shown in the [011]_B2 STEM-HAADF micrograph in Figure 5d, in which the antiphase boundary is marked by a red line. Meanwhile, unlike the results of the phase structure indicated in the upper-right corner in Figure 5c, Figure 5d confirms a single-phase B2 structure with dark 1/2{100} lattice fringes in the compressed sample; indicating that the deformation process did not induce the B2/BCC transition, differently from the B2 AlNbTiVZr_x alloys [6]. In summary, these results show that the remarkable compressive properties of the low-density B2 AlNbZrTi_x are closely related to a dislocation-based ductility mechanism accompanied by antiphase domains, as also demonstrated in other Al-Ti containing B2 MPEAs by Feuerbacher et al. [37].

4. Conclusions

The effect of Ti content on the microstructure and room-temperature mechanical properties of AlNbZrTi_x (x = 1−3) MPEAs was investigated here. The following conclusions were drawn:

(1)

As Ti content increases, the microstructure of the MPEAs transforms, from a mixture of B2 matrix and Zr₅Al₃-type phase (x = 1, 1.5) to a single-phase B2 structure (x = 2, 3). When x = 1, the Zr₅Al₃-type phase exhibits polygonal and rod-like shapes. When x = 1.5, the polygonal particles become irregular and the rod-like phase evolves into a continuous net along the B2 grain boundaries.

(2)

The Ti element in these alloys can stabilize a single-phase B2 structure, rather than a nanoscale BCC/B2 mixture, which can also be predicted by three previously-proposed parameters (Ω ≥ 1.1, δ ≤ 6.6%, and VEC < 6.87). These MPEAs exhibited a steady decrease in B2 lattice parameters, from ~0.3336 nm (x = 1) to ~0.3309 nm (x = 3), which is attributed to weaken lattice distortion caused by a reduced δ value.

(3)

The actual densities of the AlNbZrTi_x (x = 1, 1.5, 2, 3) MPEAs reduced by ~8.7%, from ~5.85 g·cm⁻³ (x = 1) to ~5.34 g·cm⁻³ (x = 3), and were lower than that of most reported MPEAs (≥6 g·cm⁻³). Thus, the specific yield strengths (SYS) of the AlNbZrTi_x (x = 1, 1.5, 2, 3) MPEAs were calculated to be ~270 kPa·m³·kg⁻¹, ~238 kPa·m³·kg⁻¹, ~221 kPa·m³·kg⁻¹, >208 kPa·m³·kg⁻¹, respectively.

(4)

The excellent fracture strain of AlNbZrTi_x (x = 1, 2, 3) alloys were ~17.8%, 21.8%, and >50%, respectively. This Ti-addition-induced ductility improvement was attributed to the reduced δ and gradual disappearance of the Zr₅Al₃-type phase. An exception was AlNbZrTi_1.5 with a limited ductility of ~7.8%, which was explained through the distinct morphology and distribution of hard Zr₅Al₃-type phase.

(5)

The combined compressive properties (SYS and fracture strain) of AlNbZrTi_x (x = 1, 2, 3) were superior to the reported data of most BCC/B2-dominant MPEAs. The deformation mechanism of the B2 structure is closely related to a dislocation-based ductility mechanism, accompanied by antiphase domains. Our results provide an explanation for the high strength and exceptional room-temperature ductility of low-density B2 MPEAs.