Abstract
Precipitationhardening highentropy alloys (PHHEAs) with good strength−ductility balances are a promising candidate for advanced structural applications. However, current HEAs emphasize nearequiatomic initial compositions, which limit the increase of intermetallic precipitates that are closely related to the alloy strength. Here we present a strategy to design ultrastrong HEAs with highcontent nanoprecipitates by phase separation, which can generate a nearequiatomic matrix in situ while forming strengthening phases, producing a PHHEA regardless of the initial atomic ratio. Accordingly, we develop a nonequiatomic alloy that utilizes spinodal decomposition to create a lowmisfit coherent nanostructure combining a nearequiatomic disordered facecenteredcubic (FCC) matrix with highcontent ductile Ni_{3}Altype ordered nanoprecipitates. We find that this spinodal order–disorder nanostructure contributes to a strength increase of ~1.5 GPa (>560%) relative to the HEA without precipitation, achieving one of the highest tensile strength (1.9 GPa) among all bulk HEAs reported previously while retaining good ductility (>9%).
Introduction
Highentropy alloys (HEAs), which contain at least four principal elements in (near)equiatomic ratios, have attracted extensive attention because of their interesting properties^{1,2,3,4,5,6}. In the early stage of the development of HEAs, researchers wanted to seek singlephase solidsolution alloys because they considered that intermetallics are brittle and may degenerate the properties of HEAs^{1,2,3,4}. However, the fact that in most engineering alloys, secondary phases contribute significantly to the alloy properties is also true in HEAs; many reported HEAs that can overcome the strength–ductility tradeoff contain two or more phases^{7,8,9,10,11,12,13,14,15,16,17}. Proper control of the type, shape, size, volume fraction, and distribution of precipitation phases is critical for the development of the HEAs with high strength while retaining sufficient ductility^{6,7,8,13,14,15,16,17}. Ni_{3}Altype ordered L1_{2} phases, which are ductile and coherent with FCC matrices^{18}, are one important class of strengthening phases in FCCbased alloys such as superalloys, and the strengthening effects depend prominently on the volume fractions of these precipitates^{19,20,21}. Similar to conventional FCC alloys, the HEAs strengthened by L1_{2} precipitates also possess good balances between strength and ductility^{5,13,14,15,16,17}. However, the current alloying concept of HEAs, which emphasizes nearequiatomic initial compositions, intrinsically limits the increase in the contents of these ductile intermetallic strengthening phases^{7,8,13,14,15,16,17}. As schematically shown in Fig. 1a, nearequiatomic HEAs tend to form brittle intermetallics (generally nearequiatomic, e.g. NiAltype B2 and complex σ phases)^{7,8,13}; formation of ductile L1_{2} phases requires a high Ni/Al ratio. As a result, inclusion of L1_{2} precipitates into FCC HEAs has to decrease the concentration of the L1_{2} forming elements to a value <7 at% in order to obtain the nearequiatomic matrices, e.g. Al_{0.3}CoCrFeNi^{13} and (CoCrFeNi)_{94}Ti_{2}Al_{4}^{17}. In other words, based on the alloying strategy/concept of HEAs, a high Ni/Al ratio that promotes the formation of L1_{2} phases can only be achieved by reducing the L1_{2} former because the Ni additions in nearequiatomic HEAs are commonly <25 at%. Such low concentrations of the L1_{2} former lead to difficulty in developing stronger HEAs with large L1_{2} volume fractions^{14,15,16,17}.
To obtain a HEA with highcontent L1_{2} strengthening phases, we use a different way to increase the Ni/Al ratio by adding more Ni element rather than reducing Al concentration. This composition design aims to ensure that the phase separation can generate a nearequiatomic FCC matrix in situ while forming highcontent L1_{2} phases, creating PHHEAs with superior strength (Fig. 1b). Based on this strategy, we develop a nonequiatomic system Al_{0.5}Cr_{0.9}FeNi_{2.5}V_{0.2}, which has a high Ni concentration of ~50 at% and a high Ni/Al ratio of ~5. Moreover, our alloy excludes Co to reduce the cost and contains ~4 at% V to stabilize and strengthen L1_{2} phases^{20}. This welldesigned nonequiatomic system can form a PHHEA in situ by utilizing spinodal decomposition to create a lowmisfit coherent nanostructure composed of a nearequiatomic disordered FCC matrix and highcontent ordered L1_{2} nanoprecipitates (>50 vol%). This distinctive spinodal order–disorder nanostructure can resist dislocation motion by providing strong diffuse obstacles and by creating antiphase boundaries (APBs) on the slip planes of these ordered L1_{2} nanophases^{19}, contributing to an ultrahigh strength while retaining good ductility and workhardening capacity. We anticipate that this alloying strategy can offer more possibilities for the compositional design of HEAs, and will open new avenues for development of ultrastrong and ductile HEAs superior to existing commercial alloys.
Results
Material properties
Tensile engineering stress–strain curves (Fig. 2a) show that although the solutiontreated (ST) sample is soft (yield strength of ~274 MPa), remarkable strength increases occur in the samples after 72% cold rolling followed by ageing for 1 h at 600 °C and 700 °C (hereinafter named 600 A and 700 A, respectively). The maximum strengthening observed in 600 A is >1500 MPa (strength increment >560%), achieving ultrahigh yield strength (YS) of 1810 MPa and ultimate tensile strength (UTS) of 1905 MPa. Moreover, the YS and UTS of 700 A are 1570 MPa and 1763 MPa, respectively, which are ~1200 MPa (~430%) higher than those of the ST sample. In addition, both 600 A and 700 A exhibit good ductility (~9–10%) and excellent workhardening capacity (uniform elongation to fracture). The exceptional combination of properties makes our alloy Al_{0.5}Cr_{0.9}FeNi_{2.5}V_{0.2} superior to most traditional alloys for potential advanced structural applications.
Phase constituent
After cold rolling and ageing, the grains in 600 A and 700 A become very fine (Supplementary Fig. 1). However, the strength increases resulting from grain boundaries are only 212 MPa in 600 A and 284 MPa in 700 A (see Methods and Supplementary Fig. 2 for calculated details), indicating that the ultrahigh strength of the samples arises from other mechanisms. Synchrotron highenergy Xray diffraction (HEXRD) spectra of both 600 A and 700 A show superlattice peaks of the ordered L1_{2} structure and characteristic peaks of the bodycenteredcubic (BCC) structure, in addition to characteristic peaks of an FCCstructured matrix (Fig. 2b). Quantitative analysis of the volume fraction of each phase based on the HEXRD results reveals that in the two samples, the L1_{2} phases are the major precipitates (~50% in 600 A and ~44% in 700 A), whereas the contents of the BCC phases are barely ~6%, confirming the validity of our strategy for designing PHHEAs with highcontent ductile L1_{2} strengthening phases. Interestingly, only two regions (bright matrix and dark precipitates with low volume fractions) are found in highangle annular darkfield (HAADF) scanning transmission electron microscopy (STEM) images (Fig. 2c, d). However, selected area electron diffraction (SAED) patterns (insets in Fig. 2c, d) indicate that the samples indeed contain three phases―two symmetric spots exist near {200}_{FCC} in addition to the superlattice patterns of the FCC and L1_{2} phases. The interplanar spacings of the two spots are approximately 0.205 nm, which is consistent with the spacings of {110}_{BCC} calculated from the HEXRD spectra (~0.204 nm), suggesting that these lowcontent dark precipitates are of BCC structure and that the bright region consists of the FCC matrix and highcontent L1_{2} strengthening phases.
Spinodal order–disorder nanostructure
To reveal the microstructure of the FCC matrix and L1_{2} phases, we conducted atomicresolution HAADFSTEM using atomic mass contrast and atom probe tomography (APT). We find an interesting nanostructure consisting of two fully coherent nanoscale regions in atomicresolution images (Fig. 3a), in which brighter columns represent heavier elements (e.g. Ni) whereas darker columns refer to lighter atoms (e.g. Al). The intensity profile along the cyan arrow across the two regions in Fig. 3a reveals that heavier and lighter elements are arranged alternately in one region, whereas in the other region the intensity is uniform (Fig. 3b). Fast Fourier transform (FFT) of the region with uniform intensity presents an FCCstructured diffraction pattern, whereas that of the alternately arranged region demonstrates an L1_{2}type superlattice pattern. Interestingly, the two nanophases seem to be tangled and have no abrupt changes in composition (composition continuously changing without abrupt interfaces), and the lattice parameters of the two highly coherent phases are almost the same (0.3582 nm for FCC and 0.3583 nm for L1_{2}). Consequently, these L1_{2} phases are invisible in Fig. 2c, d. Hereinafter we mainly present the APT results of the detailed microstructure of 600 A because the microstructures of 600 A and 700 A are similar (Supplementary Fig. 3).
The threedimensional morphologies of the ordered L1_{2} nanoprecipitates and disordered FCC matrix are revealed by reconstructing 63 at% Ni and 15 at% Cr isoconcentration surfaces, respectively, based on the APT data (Fig. 3c, d). The tangled nanostructure can be seen more clearly, where the interconnected disordered FCC matrix serves as the frame and is filled with the ordered L1_{2} phases (also interconnected). Interestingly, this tangled nanostructure with diffuse coherent precipitate−matrix interfaces seems to be an isotropic spinodal structure^{22}. To further investigate the composition and morphology at the phase interfaces, we selected an analysis box of 30 × 20 × 10 nm^{3} from Fig. 3d and reconstructed an isoconcentration surface with a higher Cr content (22 at%) within it. The onedimensional concentration profiles of phase interfaces (Fig. 3e, f) demonstrate that Fe slightly enriches at the L1_{2}−L1_{2} interface and that all elements continuously change from FCC to L1_{2} without an abrupt FCC−L1_{2} interface (diffuse interface width ~4 nm). The onedimensional concentration profiles across multiple FCC and L1_{2} phases also exhibit longrange periodic composition fluctuations without abrupt changes in composition (Supplementary Fig. 4). These results suggest the spinodally decomposed nature of the phase separation^{22}. The mean compositions of the two phases confirm that the FCC matrix is a phase conforming to the nearequiatomic alloying concept of HEAs and that our alloying strategy for producing PHHEAs in situ by phase separation is valid (Supplementary Fig. 4c). Moreover, the APT results indicate that the BCC phases are Crrich lamellar phases with the interfaces slightly enriched by Fe (Supplementary Fig. 5).
Discussion
To form a nearequiatomic matrix in situ, we add a high content of Cr (~18 at%) into our alloy. Such a high addition promotes the change of the phase separation mechanism to spinodal decomposition^{23} and reduces the difference in the lattice parameters between the FCC matrix and L1_{2} nanophases^{24}, thereby stabilizing the highdensity coherent interfaces^{25} and lowering the driving force for competitive coarsening^{26}. During phase separation, both Cr and Fe are rejected from the L1_{2} phases, exhibiting similar cluster trends. We observe that the Zener ratios (a dimensionless number used to quantify the anisotropy of cubic crystals; the larger the deviation from one, the stronger the anisotropy) of Cr and V are <1, whereas those of the other elements are greater than one, and Cr has the highest elastic parameters^{21,27}, suggesting that such a cluster can reduce the anisotropy of the matrix and stabilize this distinctive isotropic spinodal nanostructure. Since the spinodal nanostructure is isotropic, no preferred orientation exists during phase separation, leading to no additional satellite spots appearing in the SAED patterns (Fig. 2c, d). It should be noted that phase separation phenomena including spinodal decomposition are common in HEAs^{1,28,29,30,31,32,33,34,35,36,37,38}. It is therefore that this work can utilize them to create a PHHEA in situ with highcontent L1_{2} precipitates from a nonequiatomic system. More importantly, such a spinodal nanostructure should be closely related to the ultrahigh strength of our alloy because it contains ordered L1_{2} nanoprecipitates with high density and volume fraction, which can introduce abundant lowmisfit, coherent, nanoscale interfaces^{39} that can effectively strengthen metals, e.g. nanotwins^{39,40,41} and nanoprecipitates^{26}.
In our HEAs, the lattice constant and modulus of the ordered L1_{2} precipitates approximate to those of the disordered FCC matrix, and no new FCC−L1_{2} interfaces form after dislocations shear these L1_{2} precipitates with diffuse coherent interfaces. Therefore, strengthening by coherent strain, modulus difference, and interfacial energy increase is negligible^{15,16,17}. Instead, the L1_{2} nanophases were formed by spinodal decomposition. This spinodal nanostructure can resist dislocation motion by providing strong diffuse attractive obstacles to trap moving dislocations and by creating APBs on the slip planes of these ordered L1_{2} nanophases when they are sheared by dislocations^{19}. The contribution of the spinodal nanostructure to the YS of 600 A is calculated to be 1158 MPa (see Methods in details). Superposing additional strengthening from grain boundaries and the BCC phases, the total increase in YS is estimated to be 1524 MPa, which agrees with the experimental result of 1536 MPa, confirming that this spinodal order–disorder nanostructure can potently strengthen our alloys. Furthermore, no pileup dislocations are found in the FCC + L1_{2} region (Supplementary Fig. 6), suggesting that the diffuse lowmisfit coherent FCC−L1_{2} interfaces can minimize the elastic strain accumulation resulting from dislocation shear and hence prevent crack initiation at these interfaces, contributing to good ductility.
We compare the tensile properties of our PHHEA with those of reported HEAs in Fig. 4. It is clear that the tensile properties depend on the phases present in alloys. FCC alloys are ductile but relatively soft, whereas BCC alloys are hard but relatively brittle, and strength–ductility combinations of multiphase alloys are generally better than those of singlephase ones. In particular, our alloys (600 A and 700 A) have significant advantages over existing bulk HEAs, including the highest YS and UTS and a good strength−ductility balance, showing the enormous potential for advanced structural applications.
Our alloying strategy not only involves the change to the initial compositions (nonequiatomic), but also requires the formation of a nearequiatomic/highentropy solidsolution matrix in situ after separating out enough strengthening phases. The latter requirement is the primary difference relative to the previous HEA design concepts, which only consider the initial compositions regardless of whether the final primary phases are highentropy solid solutions or not, leading to the possible loss of the HEA advantages. As a result, our alloying strategy can introduce desired strengthening phases with controllable contents while still possessing the advantages of HEAs, and the corresponding HEA has several unique advantages over previous reported HEAs, as well as other materials strengthened by abundant lowmisfit coherent nanotwins^{39,40,41}/nanoprecipitates^{26}. First, our selected strengthening phases are ductile, highly coherent and low misfit with matrix (i.e. they can compatibly deform with the matrix), implying that increasing their content does not significantly degenerate the ductility. Controllable microstructure (by either alloying or processing design) provides numerous possibilities for the optimization of properties; for instance, the strength of our alloy can be controlled within a very wide range, at least, of ~270–1900 MPa, and the largest elongation is >40%. Furthermore, benefiting from the advantages of the multiprincipalelement/highentropy matrix, high Cr addition can improve the resistance to corrosion and oxidation, and high Al content reduces the alloy density. These advantages present the potential for the application of our HEA to many rigorous service conditions, such as those subjected to very large loading and involving highly corrosive environment. More importantly, the spinodal nanostructure achieves an amazing strengthening effect (>1.5 GPa), good ductility (>9%), and excellent workhardening capacity (uniform elongation to fracture), which are higher than those induced by independent spherical hard nanoparticles reported recently (~1.1 GPa, 8.2%, and 3.8% uniform elongation, respectively)^{26} because of the higher volume fractions and the different strengthening mechanisms. It should be noted that the exciting strengthening effect of the isotropic spinodal structure with ductile Ni_{3}Altype L1_{2} nanophases in our HEA is different from that of the anisotropic spinodal structures with NiAltype B2 phases reported previously^{42,43}. The tensile properties of these reported HEAs with the anisotropic spinodal structures were unavailable and the compressive properties were just reported. The strength data obtained from the two loading modes are inappropriate to be analogized. Many HEAs that have very high compressive strength are too brittle to bear tensile loading^{43,44,45}, leading to inferior competitiveness compared with commercial alloys. These results imply that although the spinodal structures are common in HEAs, the strengthening mechanism of anisotropic spinodal structures is different from that of the isotropic spinodal structures in our HEA. The new strengthening mechanism of the isotropic spinodal nanostructure demonstrates another classical example of metallic strengthening by lowmisfit coherent nanoscale interfaces/precipitates, revealing a new strategy to potently strengthen alloys without severely losing ductility. Overall, the distinctive microstructure and excellent mechanical properties of our HEAs present the possibility of development of new HEAs with ultrahigh strength, sufficient ductility, and resistance to corrosion and oxidation to compete with traditional structural materials.
Methods
Samples preparation
Master Al_{0.5}Cr_{0.9}FeNi_{2.5}V_{0.2} alloy ingots were prepared by arcmelting pure metals (Al, Cr, Fe, Ni, and V with purity >9.9%) under highpurity argon atmosphere. During preparation, the ingots were inverted and remelted at least four times to improve chemical homogeneity. These master alloy ingots were subsequently casted into a copper mold with a dimension of 50 × 13 × 30 mm^{3}. To reduce elemental segregation, the ascast alloys were solutiontreated at 1200 °C for 24 h followed by water quenching. The solutiontreated alloys were then coldrolled in steps at room temperature with a total rolling reduction ratio of 72%. Finally, the asrolled samples were aged at 600 and 700°C for 1 h followed by air cooling.
Tensile testing
Flat, dogboneshaped tensile samples with a gauge length of 10 mm, a width of 3 mm and a thickness of 1 mm were cut by electrical discharge machining, and polished with 2000grit SiC papers. Tensile tests were performed by a CMT4305 universal electronic tensile testing machine at room temperature using a nominal strain rate of 1 × 10^{−3} s^{−1}.
Microstructural characterization
Grain morphology was characterized by a Zeiss Axio Observer A1m inverted optical microscope (OM) and a JEOL JSM7001F scanning electron microscope with a TSL OIM 6.2 system for electron backscatter diffraction (EBSD) analysis. The specimens were ground with 2000grit SiC papers, then mechanically polished using 2.5 and 0.5grit diamond pastes, and finally polished using oxide polish suspension. The surfaces of the OM specimens were corroded by Marble’s corrosive liquid to show grain boundaries. The EBSD scanning step is 0.05 μm, and the data are processed and analyzed by TSL OIM 6.2 software.
Phases existing in the current alloys were investigated by a synchrotronbased highenergy Xray diffraction technique, at the 11IDC beam line of the Advanced Photon Source, Argonne National Laboratory, USA. The wavelength λ of the high energy Xray used here is 0.011725 nm. To obtain precise lattice parameters a, we calculated the lattice parameters of all peaks in the HEXRD spectra, and plotted them versus f(θ) = (cos^{2}θ/sinθ + cos^{2}θ/θ)/2, where θ is the Bragg angle of the diffraction peaks^{46}. We then fitted the plot using the linear least squares method and extrapolated to f(θ) = 0, i.e. θ = 90°, where the lattice parameter a is the precise value. The results showed that although the lattice parameter of the ST sample is 3.5962 ± 0.0006 Å, the lattice parameter of the FCC matrix of 600 A (3.5815 ± 0.0008 Å) with a high Cr content was decreased to a value very close to that of the L1_{2} precipitates (3.5832 ± 0.0009 Å). The ± values mean the standard deviation given by the linear least squares fitting. The mean size (particle diameter d) of the ordered L1_{2} precipitates was determined according to the Scherrer equation d = Kλ/βcosθ^{47}, where β is the full width at half maximum (FWHM) of the diffraction peak (in radian) and K is a constant and approximates to 0.9 for cubic symmetry^{48}. Many factors such as fine particle size, chemical heterogeneities, stress gradients, and microstresses can cause broadening of the (100) superlattice peak of the L1_{2} phases. Fortunately, in addition to fine particle size, other factors broaden both the superlattice and fundamental peaks. The difference between the FWHM of the (100) superlattice peak (~0.043°) and that of the fundamental peaks (~0.026°) in 600 A is the broadening resulting from the fine particle size^{49}, affording a mean diameter of the L1_{2} phases of ~35 nm in 600 A. The highest peaks of FCC + L1_{2} (111) and BCC (110) are clearly overlapped. To reduce error, we performed the quantitative phase analysis based on the secondary high peak of each phase, i.e. the (200) fundamental peak of the FCC + L1_{2}, the (100) superlattice peak of the L1_{2}, and the (211) peak of the BCC. Another difficulty is lied in distinguishing the contributions of FCC and L1_{2} to the (200) fundamental peak. Here we assume that (1) the L1_{2} phases is of perfect order, and (2) the contributions of the FCC and L1_{2} to the peak intensity depend on only their contents. Marking FCC, L1_{2}, and BCC as subscripts 1, 2, and 3, respectively, there is:
where I_{i j} is the intensity of the i peak for the j phase, f_{j} is the fraction volume of the j phase, and C_{i j} is a parameter related to phase composition, Bragg angle, lattice structure, and phase nature. For synchrotron source, the absorptivity and Bragg angle are rather small, and hence C_{i j} can be evaluated according to the APT result. The volume fractions of the FCC, L1_{2}, and BCC phases in 600 A were calculated to be 43.7%, 50.1%, and 6.2%, respectively. It should be noted that these values may not be absolutely precise, but they can serve as reasonable references for analysis of the strengthening mechanism in our ultrastrong PHHEAs.
Detailed microstructure was characterized by a JEOL ARM200F sphericalaberration corrected transmission electron microscope (TEM), and atomicresolution images were taken by a highangle annular darkfield (HAADF) detector. The TEM specimens with dimensions of Ф3 mm × 0.5 mm were cut with a diamond wire saw and thinned to 40–60 μm by 3000 and 5000grit SiC papers. After mechanically polishing, twinjet electropolishing was conducted using a mixed solution of HNO_{3}:CH_{3}OH = 1:4 at a temperature of ~235 K with a direct voltage of 28 V and a current of 55 mA. Final polishing was conducted on a Gatan 691 precision ion polishing system using a voltage of 2 kV for 7.5 min and a voltage of 1 kV for 15 min.
Atom probe tomography (APT) analysis for threedimensional distribution of elements in materials was performed by a CAMECA local electrode atom probe LEAP^{TM} 4000X SI at a target specimen temperature of 20 K, under a pulsing UV laser with a pulse energy of 40 pJ, a pulse rate of 200 kHz, and an ion collection rate of 0.5% per pulse. A twostage electropolishing procedure was used to fabricate a sharp tip specimen (apex radius <50 nm) for APT analyses. The first stage electropolishing was performed using an electrolyte containing 25 volume per cent (vol.%) perchloric acid in acetic acid at 15 V and room temperature, and the second stage was by a 2 vol.% perchloric acid in 2butoxyethanol solution at 20 V. Reconstruction and quantitative analysis of the APT data were conducted on a CAMECA IVAS version 3.6.8 software. The compositions of the FCC, L1_{2} and BCC phases were estimated by averaging the core concentrations of the proximity histograms produced by reconstructing 22% Cr, 63% Ni, and 50% Cr (in at%) isoconcentration surfaces, respectively.
Estimation of strengthening by various mechanisms
For nearequiatomic HEAs, strengthening by solute atoms is difficult to evaluate because distinguishing between the solvent and solute is difficult. Fortunately, solution strengthening should be similar for a given system, and hence the difference in the strength of ST alloys is attributed to various grain sizes. The intercept term σ_{0} in the Hall–Petch equation (Δσ_{gs} = σ_{0} + k_{g}∙d^{−0.5}, where k_{g} is the Hall–Petch coefficient and d is the mean grain size)^{21} can be considered an alloy constant composed of solution strengthening Δσ_{ss} and lattice friction stress σ_{fs}. This consideration can reasonably simplify the estimation of the strengthening effect in multiprincipalelement HEAs. To obtain different grain sizes, we annealed the asrolled samples at 1200°C for various hold times (10, 30, and 120 min), and measured the grain sizes and tensile properties of these asannealed samples. Using Hall–Petch equation to fit the experimental results (Supplementary Fig. 2), σ_{0} and k_{g} were determined to be 270.0 MPa and 378.6 MPa∙μm^{0.5}, respectively, where σ_{0} = σ_{fs} + Δσ_{ss}. The mean grain sizes of 600 A and 700 A are 2.82 ± 0.77 and 1.62 ± 0.53 μm, respectively. The ± values mean the standard deviation given by EBSD. The contributions of grain boundaries to the yield strengths of the two samples are 212 and 284 MPa, respectively. Since recovery and recrystallization have been completed in all the samples after the final heat treatment, dislocation strengthening is negligible.
For spinodal structure, dislocations will be trapped by strong diffuse obstacles. The equation for evaluation of strengthening by spinodal structure is^{19}:
where M = 3.06 is the Taylor factor, k_{s} = 0.122∙[π(1 − υ)/(1 − 2υ)]^{2/3} is a material constant, υ = 0.31 is the Poisson ratio, A = (C − C_{0})/3 is the amplitude of that modulation (in at%), λ is the wavelength of the composition modulation, and η = a^{−1}∙da/dC is the lattice strains produced by spinodal decomposition, in which a is the lattice constant for the undecomposed sample, da/dC is the variation in the lattice constant with the composition C. According to the HEXRD data, the lattice constants of the undecomposed ST sample and FCC matrix in 600 A are 0.3596 and 0.3582 nm, respectively. Since the current system (600 A sample) is isotropic, E = 162 GPa is the elastic modulus, Y = E/(1 − υ) = 234.8 GPa is an elasticdependent parameter, G = 61.8 GPa is the shear modulus, and b = \(\sqrt 2 a/2\) is the Burgers vector. According to Supplementary Fig. 4b, λ is evaluated to be ~40 nm, whereas this value approximates to 70 nm (2d) based on the HEXRD result. However, the value evaluated based on the APT data is quite local while that calculated by the Scherrer equation is based on an assumption of spherical particles. Therefore, we use a λ value of 60 nm for calculation. In the current multiprincipalelement system, it is difficult to evaluate the value of A. We observe that Fe and Cr exhibit similar cluster trends, and the content of Ni + Al + V is greater than 60 at% according to the APT data. We therefore assume that Ni + Al + V is the solvent for simplification. After ageing 1 h at 600°C, the solute concentrations change from 39.34 at% in the ST sample to 54.57 at% in the FCC matrix (rich region) based on the APT data. Accordingly, A = 5.1 at% and Δσ_{sds} is calculated to be 406 MPa.
Strengthening by ordered coherent precipitates Δσ_{os} occurs when dislocations shear these ordered phases and create APBs on the slip planes of them. We find that the distances between two coupled dislocations (~20–40 nm, as shown in Supplementary Fig. 6d) approximate to the mean diameter of the ordered L1_{2} phases (~30 nm), suggesting that the coupling of the dislocation in the pairs should be moderate (between strong and weak), and that the alloy should be aged to near the peak strength condition. Therefore, the YS increase Δσ_{os} is given by^{19}:
in which f is the volume fraction of the ordered precipitates, r is the average radius of these ordered phases, and γ_{APB} = 0.2 J m^{−2} is the APB energy that is adopted from the data of Ni_{3}Al precipitates^{50}. Assuming that the L1_{2} precipitates are near spherical, f ≈ 50% and r is assumed to be 1/4λ (~15 nm) in 600 A; hence Δσ_{os} is calculated to be 752 MPa. It should be noted that in our PHHEAs the ordered L1_{2} nanoprecipitates are irregular and interconnected rather than spherical. The increases in YS calculated based on independent spheres is therefore not absolutely accurate and may underestimate the area of APBs and the strengthening effect. Nevertheless, the results can offer reasonable references to analyze the strengthening caused by the current spinodal order–disorder nanostructure.
Considering the BCC phases as the precipitates within FCC + L1_{2} matrix, we find that dislocations can shear and pass through these precipitates (Supplementary Fig. 6a–c). It is noteworthy that two orientation relationships, i.e. \([\bar 111]_{{\mathrm{BCC}}}//[{\mathrm{0}}11]_{{\mathrm{FCC}}}\) and \((110)_{{\mathrm{BCC}}}//(11\bar 1)_{{\mathrm{FCC}}}\), exist between the BCC phases and matrix (Supplementary Fig. 7). These observations mean that these BCC phases share the slide planes with the matrix, and that the specific energy of new interfaces created by shearing is low, leading to weak chemical strengthening. Moreover, elastic lattice strain induced by semicoherent precipitate–matrix interfaces is low enough to be negligible. Instead, the difference in shear moduli between the BCC precipitates and matrix is high (ΔG = G_{BCC}–G_{matrix} = 100–85 = 15 GPa, where G_{BCC} is adopted from the data of a Cr−22 at% Fe alloy^{51,52} and G_{matrix} is a value selected between the shear modulus of CoCrFeNi^{17} and that of CoCrNi^{16}). Therefore, modulus strengthening Δσ_{ms}, which should be the primary contribution to the strength increase, is given by^{19}:
where m is a constant and is equal to 0.85 for FCCbased alloys. In the 600 A sample, the BCC phases are plateshaped with dimensions of ~500 × 50 × 10 nm^{3}. For simplification, we assume the BCC phases to be spherical with diameters of 50 nm because the projected lengths of the BCC phases at the slide planes approximate to this value. Accordingly, Δσ_{ms} is calculated to be 154 MPa. It is worth pointing out that this assumption significantly increases the number density of the BCC phases for a fixed f, leading to overestimation of the contribution of the BCC phases to the yield strength. Nevertheless, strengthening by the BCC precipitates is confirmed to be low, and the calculated result agrees well with the actual increases in strength.
Data availability
The data that support the findings of this study are available from the corresponding author on reasonable request.
References
 1.
Yeh, J.W. et al. Nanostructured highentropy alloys with multiple principal element: novel alloy design concepts and outcomes. Adv. Eng. Mater. 6, 299–303 (2004).
 2.
Zhang, Y. et al. Microstructures and properties of highentropy alloys. Prog. Mater. Sci. 61, 1–93 (2014).
 3.
Pickering, E. J. & Jones, N. G. Highentropy alloys: a critical assessment of their founding principles and future prospects. Int. Mater. Rev. 61, 183–202 (2016).
 4.
Gludovatz, B. et al. A fractureresistant highentropy alloy for cryogenic applications. Science 345, 1153–1158 (2014).
 5.
Liu, W. H., Yang, T. & Liu, C. T. Precipitation hardening in CoCrFeNibased high entropy alloys. Mater. Chem. Phys. 210, 2–11 (2018).
 6.
Miracle, D. B. & Senkov, O. N. A critical review of high entropy alloys and related concepts. Acta Mater. 122, 448–511 (2016).
 7.
Liu, W. H. et al. Ductile CoCrFeNiMo_{X} high entropy alloys strengthened by hard intermetallic phases. Acta Mater. 116, 332–342 (2016).
 8.
Ming, K., Bi, X. & Wang, J. Precipitation strengthening of ductile Cr_{15}Fe_{20}Co_{35}Ni_{20}Mo_{10} alloys. Scr. Mater. 137, 88–93 (2017).
 9.
Wani, I. S. et al. Tailoring nanostructures and mechanical properties of AlCoCrFeNi_{2.1} eutectic high entropy alloy using thermomechanical processing. Mater. Sci. Eng. A 675, 99–109 (2016).
 10.
Li, Z., Pradeep, K. G., Deng, Y., Raabe, D. & Tasan, C. C. Metastable highentropy dualphase alloys overcome the strengthductility tradeoff. Nature 534, 227–230 (2016).
 11.
Lu, Y. et al. Directly cast bulk eutectic and neareutectic high entropy alloys with balanced strength and ductility in a wide temperature range. Acta Mater. 124, 143–150 (2017).
 12.
Huang, H. et al. Phasetransformation ductilization of brittle highentropy alloys via metastability engineering. Adv. Mater. 29, 1701678 (2017).
 13.
Li, D. et al. Highentropy Al_{0.3}CoCrFeNi alloy fibers with high tensile strength and ductility at ambient and cryogenic temperatures. Acta Mater. 123, 285–294 (2017).
 14.
Gwalani, B. et al. Optimizing the coupled effects of HallPetch and precipitation strengthening in a Al_{0.3}CoCrFeNi high entropy alloy. Mater. Des. 121, 254–260 (2017).
 15.
Wang, Z. G. et al. Effect of coherent L1_{2} nanoprecipitates on the tensile behavior of a fccbased highentropy alloy. Mater. Sci. Eng. A 696, 503–510 (2017).
 16.
Zhao, Y. L. et al. Heterogeneous precipitation behavior and stackingfaultmediated deformation in a CoCrNibased mediumentropy alloy. Acta Mater. 138, 72–82 (2017).
 17.
He, J. Y. et al. A precipitationhardened highentropy alloy with outstanding tensile properties. Acta Mater. 102, 187–196 (2016).
 18.
Liu, C. T. & Stiegler, J. O. Ductile ordered intermetallic alloys. Science 226, 636–642 (1984).
 19.
Ardell, A. J. Precipitation hardening. Metall. Trans. A 16A, 2131–2165 (1985).
 20.
Reed, R. C. The Superalloys: Fundamentals and Applications. (Cambridge university press, New York, 2006).
 21.
Courtney, T. H. Mechanical Behavior of Materials. Second edn. (McGrawHill, Boston, 2000).
 22.
Cahn, J. W. Phase separation by spinodal decomposition in isotropic systems. J. Chem. Phys. 42, 93–99 (1965).
 23.
Viswanathan, G. B. et al. Precipitation of ordered phases in metallic solid solutions: a synergistic clustering and ordering process. Scr. Mater. 65, 485–488 (2011).
 24.
BoothMorrison, C. et al. Effects of solute concentrations on kinetic pathways in Ni–Al–Cr alloys. Acta Mater. 56, 3422–3438 (2008).
 25.
Lu, K. Stabilizing nanostructures in metals using grain and twin boundary architectures. Nat. Rev. Mater. 1, 16019 (2016).
 26.
Jiang, S. et al. Ultrastrong steel via minimal lattice misfit and highdensity nanoprecipitation. Nature 544, 460–464 (2017).
 27.
Lenkkeri, J. T. & Lahteenkorva, E. E. An investigation of elastic moduli of vanadiumchromium alloys. J. Phys. F Metal Phys. 8, 1643–1651 (1978).
 28.
Singh, S., Wanderka, N., Murty, B. S., Glatzel, U. & Banhart, J. Decomposition in multicomponent AlCoCrCuFeNi highentropy alloy. Acta Mater. 59, 182–190 (2011).
 29.
Santodonato, L. J. et al. Deviation from highentropy configurations in the atomic distributions of a multiprincipalelement alloy. Nat. Commun. 6, 5964 (2015).
 30.
Otto, F. et al. Decomposition of the singlephase highentropy alloy CrMnFeCoNi after prolonged anneals at intermediate temperatures. Acta Mater. 112, 40–52 (2016).
 31.
Wen, L. H. et al. Effect of aging temperature on microstructure and properties of AlCoCrCuFeNi highentropy alloy. Intermetallics 17, 266–269 (2009).
 32.
Singh, S., Wanderka, N., Kiefer, K., Siemensmeyer, K. & Banhart, J. Effect of decomposition of the CrFeCo rich phase of AlCoCrCuFeNi high entropy alloy on magnetic properties. Ultramicroscopy 111, 619–622 (2011).
 33.
Wang, W.R. et al. Effects of Al addition on the microstructure and mechanical property of Al_{X}CoCrFeNi highentropy alloys. Intermetallics 26, 44–51 (2012).
 34.
Pickering, E. J., Stone, H. J. & Jones, N. G. Finescale precipitation in the highentropy alloy Al_{0.5}CrFeCoNiCu. Mater. Sci. Eng. A 645, 65–71 (2015).
 35.
Tong, C.J. et al. Mechanical performance of the Al_{X}CoCrCuFeNi highentropy alloy system with multiprincipal elements. Metall. Mater. Trans. A 36A, 1263–1271 (2005).
 36.
Tong, C.J. et al. Microstructure characterization of Al_{X}CoCrCuFeNi highentropy alloy system with multiprincipal elements. Metall. Mater. Trans. A 36A, 881–893 (2005).
 37.
Tung, C.C. et al. On the elemental effect of AlCoCrCuFeNi highentropy alloy system. Mater. Lett. 61, 1–5 (2007).
 38.
Manzoni, A., Daoud, H., Volkl, R., Glatzel, U. & Wanderka, N. Phase separation in equiatomic AlCoCrFeNi highentropy alloy. Ultramicroscopy 132, 212–215 (2013).
 39.
Lu, K., Lu, L. & Suresh, S. Strengthening materials by engineering coherent internal boundaries at the nanoscale. Science 324, 349–352 (2009).
 40.
Lu, L., Chen, X., Huang, X. & Lu, K. Revealing the maximum strength in nanotwinned copper. Science 323, 607–610 (2009).
 41.
Li, X., Wei, Y., Lu, L., Lu, K. & Gao, H. Dislocation nucleation governed softening and maximum strength in nanotwinned metals. Nature 464, 877–880 (2010).
 42.
Hanna, J. A., Baker, I., Wittmann, M. W. & Munroe, P. R. A new highstrength spinodal alloy. J. Mater. Res. 20, 791–795 (2011).
 43.
Xia, S., Yang, X., Chen, M., Yang, T. & Zhang, Y. The Al effects of Cofree and Vcontaining highentropy alloys. Metals 7, 18 (2017).
 44.
Liu, S., Gao, M. C., Liaw, P. K. & Zhang, Y. Microstructures and mechanical properties of Al_{X}CrFeNiTi_{0.25} alloys. J. Alloy Compd. 619, 610–615 (2015).
 45.
Laktionova, M. A., Tabchnikova, E. D., Tang, Z. & Liaw, P. K. Mechanical properties of the highentropy alloy Ag_{0.5}CoCrCuFeNi at temperatures of 4.2–300 K. Low Temp. Phys. 39, 630–632 (2013).
 46.
Nelson, J. B. & Riley, D. P. An experimental investigation of extrapolation methods in the derivation of accurate unitcell dimensions of crystals. Proc. Phys. Soc. 57, 160–177 (1945).
 47.
Patterson, A. L. The scherrer formula for Xray particle size determination. Phys. Rev. 56, 978–982 (1939).
 48.
Langford, J. I. & Wilson, A. J. C. Seherrer after sixty years: a survey and some new results in the determination of crystallite size. J. Appl. Crystallogr. 11, 102–113 (1978).
 49.
Ungár, T. Microstructural parameters from Xray diffraction peak broadening. Scr. Mater. 51, 777–781 (2004).
 50.
Kozar, R. W. et al. Strengthening mechanisms in polycrystalline multimodal nickelbase superalloys. Metall. Mater. Trans. A 40, 1588–1603 (2009).
 51.
Lenkkeri, J. T. The elastic moduli of some bodycentred cubic titaniumvanadium, vanadiumchromium and chromiumiron alloys. J. Phys. F Metal. Phys. 10, 611–618 (1980).
 52.
Razumovskiy, V. I., Ruban, A. V. & Korzhavyi, P. A. Firstprinciples study of elastic properties of Cr and Ferich FeCr alloys. Phys. Rev. B 84, 024106 (2011).
Acknowledgements
Y.X. acknowledges support by the National Natural Science Foundation of China (NSFC) (grant numbers 51471035 and 51701018). Y.J.L. acknowledges the Project Funded by China Postdoctoral Science Foundation (grant number 2017M620019). X.L. acknowledges support by the NSFC (grant numbers 11522218 and 51420105001). Use of the Advanced Photon Source is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DEAC0206CH11357. We thank Dr. F.K. Chiang at National Institute of CleanandLowCarbon Energy, China for help in the STEM experiments, thank Dr. S.B. Jin at the Materials Characterization Facility of Nanjing University of Science and Technology for help in the APT experiments, and thank Prof. Z.H. Nie at Beijing Institute of Technology for help in the HEXRD analysis.
Author information
Author notes
Affiliations
Contributions
Y.X., F.W., X.L., and Y.J.L. designed the study. Y.J.L., L.j.W., Y.r.W., Y.X., and X.L. analyzed the data and wrote the paper. L.j.W., B.C., T.C., and Q.X. carried out the main experiments. Y.r.W. conducted the STEM characterization and interpreted the results. Q.W. and G.S. conducted the APT experiments and interpreted the results. Y.W. and Y.R. conducted the synchrotron experiments and analyzed the results. L.W., F.W., and H.C. reconfirmed the mechanical properties. All authors contributed to the discussion of the results, and commented on the manuscript.
Corresponding authors
Correspondence to Yunfei Xue or Xiaoyan Li or Fuchi Wang.
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Received
Accepted
Published
DOI
Further reading

High entropy alloy: A promising matrix for highperformance tungsten heavy alloys
Journal of Alloys and Compounds (2019)

Localordering mediated configuration stability and elastic properties of aluminumcontaining high entropy alloys
Intermetallics (2019)

Effect of cooling rate on microstructure and mechanical properties of Al0.3CoCrFeNi highentropy alloy
Materials Research Express (2019)

Enhanced strength–ductility synergy in ultrafinegrained eutectic highentropy alloys by inheriting microstructural lamellae
Nature Communications (2019)

Thermodynamic modeling of elastic mismatch strain energy on epitaxial growth of GaInN thin films
Journal of Alloys and Compounds (2019)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.