Introduction
Metals and alloys that combine high strength and ductility are highly desirable for engineering applications because they can withstand substantial stress and strain before failure. However, strength and ductility are generally mutually exclusive1, compelling researchers to strike a compromise between these two properties to achieve optimal performance. To overcome this trade-off, metallurgists have made tremendous efforts to develop new alloy composition, to introduce strengthening factors while retaining ductility, such as coherent precipitates[2](https://www.nature.com/articles/s41467-025-64871-4#ref-CR2 “Jiang, S…
Introduction
Metals and alloys that combine high strength and ductility are highly desirable for engineering applications because they can withstand substantial stress and strain before failure. However, strength and ductility are generally mutually exclusive1, compelling researchers to strike a compromise between these two properties to achieve optimal performance. To overcome this trade-off, metallurgists have made tremendous efforts to develop new alloy composition, to introduce strengthening factors while retaining ductility, such as coherent precipitates2,3 and short-range ordering4,5, or to incorporate hard and soft crystallographic phases within a single alloy6,7. Another effective strategy involves microstructural design, where hierarchical or gradient microstructures are engineered through novel manufacturing processes to sequentially trigger multiple deformation mechanisms at different stages8.
Eutectic high-entropy alloys (EHEAs) exemplify a promising compositional design strategy by integrating ductile face-centred cubic (FCC) phases and strong body-centred cubic (BCC) phases9 to achieve a favourable balance between strength and ductility. Recent advances in additive manufacturing (AM), particularly laser powder bed fusion (L-PBF), offer new opportunities to further enhance the properties of such alloy systems10. L-PBF not only enables previously unattainable geometric complexities in metallic parts but also produces unique microstructures and superior properties that are beyond the reach of conventional manufacturing methods11. Particularly, the extremely rapid solidification inherent to L-PBF leads to significantly refined, non-equilibrium microstructural features that contribute to high strength of EHEA12. However, retaining high ductility in L-PBF alloys remains challenging, as such microstructural refinement may restrict strain accommodation.
Here, we leverage the L-PBF approach to fabricate a hierarchically heterostructured Al19Co20Fe20Ni41 EHEA that simultaneously achieves high strength and high ductility through synergistic deformation mechanisms and persistent strain hardening. The enhanced mechanical performance of the as-printed EHEA is attributed to its increased fraction of the ductile FCC phase, uniformly dispersed coherent nanoprecipitates, nanolamellar structures, hierarchical microstructure heterogeneity, and the activation of deformation-induced nanovoids—all of which contribute to the high strength and extensive strain hardening capacity. Our strategy highlights a pathway towards high-performance structural parts by combining composition design and advanced manufacturing techniques.
Results
Microstructure design of eutectic high-entropy alloys
The design concept of HEAs, which combines multiple principal elements, enables numerous innovative composition to be tailored for specific applications13. A key criterion in HEA design is the valence electron concentration (VEC), which serves as a predictor for phase constituents and stability14. Empirical studies suggest that a eutectic dual-phase microstructure tends to form when the VEC is between 6.87 and 8.014. Although these thresholds may vary slightly across different HEA systems15, it is generally observed that higher VEC values favour the formation of ductile FCC phases, while lower VEC values promote the development of stronger BCC phases. Among the EHEAs, AlCoCrFeNi2.1 with a VEC value of 7.7 is the most well-known and exhibits exceptional strength but moderate ductility12. Aiming for a better balance of strength and ductility in EHEAs processed by L-PBF, we selected Al19Co20Fe20Ni41, which has a higher VEC of 8.07—close to the threshold between EHEAs and FCC-HEAs (Fig. 1a and Supplementary Table 1)—to increase the volume fraction of the ductile FCC phase.
Fig. 1: Microstructure design for the Al19Co20Fe20Ni41 eutectic high-entropy alloy (EHEA) produced by laser powder bed fusion (L-PBF).
a Valence electron concentration (VEC) values of several representative BCC, eutectic, and FCC HEAs13. b Volume fraction of equilibrium phases at different temperatures predicted by computer coupling of phase diagrams and thermochemistry (CALPHAD). c Scanning electron micrograph (SEM) of the chemically etched EHEA showing the FCC/BCC nanolamellar structure. d Bright-field transmission electron microscopy (TEM) image revealing the pre-existing dislocations (indicated by white arrows) and nanoprecipitates (indicated by yellow dash circles) within the FCC/BCC lamellae. e Electron backscatter diffraction (EBSD) maps demonstrating the uniform, mesoscale heterogeneity of phase distribution and grain morphology.
Upon optimizing the L-PBF process parameters (Supplementary Fig. 1), the as-printed Al19Co20Fe20Ni41 EHEA displays a eutectic microstructure consisting of alternating FCC and BCC lamellae, with average thicknesses of 208 ± 71 nm and 112 ± 40 nm, respectively (Fig. 1c). Transmission electron microscopy (TEM) imaging (Supplementary Fig. 2a, b) reveals two classic crystallographic orientation relationships, Kurdjumov–Sachs (K–S) and Nishiyama–Wassermann (N–W), co-existing between the FCC and BCC lamellae, consistent with previously reported phenomena in EHEAs16. These semi-coherent phase boundaries, with a small lattice mismatch of 2.1% (Supplementary Fig. 2c, d), exhibit lower interfacial energy and enhance both the strength and ductility of the alloy12,16. Within the ultrafine FCC/BCC lamellae, pre-existing dislocations and nanoprecipitates smaller than 10 nm are captured under TEM imaging and diffraction patterns (Fig. 1d and Supplementary Fig. 2a, b), which are beneficial for enhancing the yield strength of the alloy.
Based on the phase diagram predicted by CALPHAD simulations (Fig. 1b), the Al19Co20Fe20Ni41 alloy is expected to solidify into simple FCC and ordered BCC (B2) phases17,18. The high content of Al facilitated the direct formation of B2 structure with long-range chemical ordering during solidification19. Due to the rapid cooling rate of the L-PBF process, the FCC and BCC nanolamellae in the as-printed Al19Co20Fe20Ni41 predominantly retained their solid-solution state at high temperatures. The energy dispersive spectroscopy (EDS) mapping (Supplementary Fig. 2e) and atom probe tomography (APT) analysis (Supplementary Table 2) reveal distinct compositional partitioning: Co, Fe, and Ni are enriched in the FCC lamellae, while Al concentrates in the BCC lamellae. This measured elemental partitioning aligns with the CALPHAD predictions for the high-temperature FCC and BCC (B2) phases (Supplementary Fig. 3a, b).
In addition to the ultrafine microstructure resulting from the rapid cooling rate, the nature of the AM process also imparts microstructural heterogeneity to the as-printed EHEA. As shown in Supplementary Fig. 4, the nonuniform cooling rates within each melt pool result in the inhomogeneous phase distribution of the FCC and BCC phases. Specifically, in the keyhole-mode melt pool, the central bottom regions were exposed to relatively higher cooling rates20, leading to a higher volume fraction of the BCC phase retained in these regions (Supplementary Fig. 4b). Consequently, the central bottom region contains a higher fraction of coarse BCC lamellae, whereas the side branches exhibit a dominant FCC phase distribution with ultrafine BCC lamellae (Supplementary Fig. 4c). Additively manufactured materials can be considered built by overlapping melt pools track by track and layer by layer. By utilizing a large hatch spacing of 120 µm and a scan rotation of 67° between adjacent layers, the L-PBF process generated a mesoscale structure consisting of uniformly distributed BCC-rich and BCC-lean “bricks” across the entire EHEA build volume (Fig. 1e). Coupled with the microscale grains, nanoscale lamellar structures, and smaller nanoprecipitates, this hierarchical heterogeneity introduces varying deformability at multiple length scales. The collaborative interactions between the nanoprecipitates and lamellae matrix, FCC and BCC lamellae, and BCC-rich and BCC-lean “bricks” synergistically enhance strength, accommodate large deformation, and improve strain hardening capability, as will be demonstrated below.
Strength from nanolamellae and nanoprecipitates
Through leveraging the HEA composition design and the unique AM benefits, the resulting Al19Co20Fe20Ni41 EHEA processed by L-PBF exhibits high yield strength of 1311 MPa, surpassing those as-cast and thermo-mechanically processed counterparts18. More importantly, this high yield strength is coupled with a large uniform elongation of 20%, highlighting its high strength–ductility synergy. This synergy is further boosted by a sustained strain hardening process, culminating in an ultimate tensile strength of 1630 MPa. Compared to previously reported additively manufactured alloys with notable as-printed tensile properties (Fig. 2b and Supplementary Table 3), the Al19Co20Fe20Ni41 alloy fabricated in this work demonstrates a significant breakthrough in strength–ductility synergy, underscoring the benefit of our strategy. Because of the 67° rotation between successive layers employed to disrupt epitaxial grain growth and randomize the crystallographic texture and lamellar orientation, the mechanical anisotropy is relatively weak in the as-printed EHEA (Fig. 1e, Supplementary Fig. 5, and Supplementary Table 4). The alloy exhibits slightly lower strength and higher ductility under tensile loading along the build direction than that loaded perpendicular to the build direction.
Fig. 2: Tensile properties of the Al19Co20Fe20Ni41 EHEA produced by L-PBF.
a Engineering tensile stress–strain tensile curve of the as-printed Al19Co20Fe20Ni41 EHEA compared to those of as-cast and thermo-mechanically processed counterparts18. b Comparison of tensile yield strength versus uniform elongation between the as-printed Al19Co20Fe20Ni41 EHEA in this work and other notable additively manufactured high-strength alloys from literature, including steels26,40,41, Ti-based alloys42,43,44, Ni-based alloys45,46,47, HEAs12,48,49,50, and a bulk metallic glass (BMG)51. The specific data and relevant references are summarized in Supplementary Table 3.
The high yield strength of the EHEA alloys produced by L-PBF is widely attributed to their nanolamellar structures and pre-exiting dislocations (Fig. 1d and Supplementary Fig. 2c)10,12, which originate from the rapid cooling rate and the large thermal strain during the L-PBF process. In this work, an interesting feature is the presence of nanoprecipitates within both FCC and BCC lamellae (Fig. 3a), which is expected to inhibit dislocation motion. Atomic-resolution TEM imaging and the superlattice reflections observed in diffraction patterns (Fig. 3b, c) confirm that the ordered structures of FCC (L12) nanoprecipitates and BCC (B2) matrix, respectively. A similar precipitation phenomenon was previously reported in as-cast EHEA AlCoCrFeNi2.1, where ordered L12 and disordered BCC nanoprecipitates were embedded within disordered FCC and ordered B2 matrix17. These nanoprecipitates share the same crystal structure as the matrix phases but differ in the degree of chemical ordering. APT analysis further reveals chemical clustering associated with the ordered FCC (L12) nanoprecipitates, enriched in Al and Ni, and the disordered BCC nanoprecipitates, enriched in Fe and Co, compared to their respective matrix phases (Fig. 3d and Supplementary Table 2). The average diameters of the L12 and BCC nanoprecipitates are estimated around 5 nm. Therefore, in the present EHEA with a relatively high fraction of the soft FCC phase, the coherent ordered nanoprecipitates help retain the high yield strength through precipitate strengthening21,22.
Fig. 3: Coherent nanoprecipitates within the lamellar structures of the as-printed EHEA.
a Bright-field TEM image showing the appearance of nanoprecipitates in FCC and BCC lamellae. The FCC (L12) and BCC nanoprecipitates are indicated by green and yellow arrows, respectively. b High-resolution high-angle annular dark-field imaging (HAADF) under STEM mode and the corresponding fast Fourier transform (FFT) pattern showing the atomic structure of ordered L12 precipitates along the [001] zone axis. c High-resolution TEM (HRTEM) image and the corresponding FFT pattern showing the atomic structure of BCC (B2) lamellae along the [011] zone axis. d Atom probe tomography (APT) analysis of L12 and BCC nanoprecipitates. The precipitates are located using iso-concentration surfaces of Ni + Al > 64 at% and Fe + Co > 43 at%, respectively. Line scans across the precipitates reveal Ni and Al enrichment in L12 precipitates, while BCC nanoprecipitates are rich in Co and Fe.
Strain hardening by coherency and heterogeneity
A remarkable phenomenon in this work is the strong and steady strain hardening throughout the entire plastic deformation regime, which is maintained at a high hardening rate of 2–4 GPa (Fig. 4a). The TEM investigation on the uniformly deformed region (20% strain) of a fractured EHEA sample elucidates the strain hardening behaviour stemming from nanoprecipitates and nanolamellae. Specifically, the pinning effect of nanoprecipitates impedes dislocation motion and promotes the formation of stacking faults, as evidenced by Fig. 4c and Supplementary Fig. 6a. In addition, the formation of Lomer–Cottrell locks of the two intersected slip systems around the nanoprecipitates (Fig. 4c) stabilizes the stacking fault network and further blocks dislocation motion. Because the ordered nanoprecipitates are naturally coherent with the disordered matrix lamellae, these nanoprecipitates minimize the strain concentration and could allow dislocations to cut through more easily2,22, thereby helping to retain the ductility of the alloy.
Fig. 4: Strain hardening behaviour of the as-printed Al19Co20Fe20Ni41 EHEA.
a Strain hardening curve of the as-printed EHEA corresponding to Fig. 2a. The orange arrow indicates the inflexion point in the strain hardening curve. b MD simulations showing the dislocation behaviour in the respective FCC and BCC lamellae at different stages of deformation. The stacking faults are marked by orange atoms and the dislocations are coloured in dark and light blue based on their types. c HRTEM image showing the formation of stacking faults and Lomer–Cottrell locks. d BF-STEM image demonstrating the increase of dislocations (indicated by yellow arrows) in both FCC and BCC lamellae, and the pile-up of dislocations along lamellar interfaces (indicated by white arrows). e HAADF-STEM image showing slip transmission across the FCC/BCC lamellar interface, indicated by yellow arrows. The inserted diffraction pattern indicates the Kurdjumov–Sachs orientation relationship of the lamellar interface.
In the as-printed Al19Co20Fe20Ni41 EHEA, the nanolamellar structures are the fundamental and predominant microstructural units. The strain hardening contribution of the nanolamellar structures is investigated through a combination of molecular dynamics (MD) simulation and TEM observation (Fig. 4 and Supplementary Fig. 6). During the initial stage of deformation (Stage I, Fig. 4b), the hard BCC lamellae remain elastic while the plastic deformation initiates preferentially in the softer FCC lamellae via the emission of partial dislocations from the FCC/BCC phase boundaries16. Strain partitioning between FCC and BCC lamellae produces hetero-deformation induced (HDI) strengthening23. As the deformation proceeds with pronounced strain hardening of FCC lamellae, the BCC lamellae gradually become plastically deformed (Stages II and III, Fig. 4b). Figure 4d and Supplementary Fig. 6b clearly indicate the increased density of dislocations in both the FCC and BCC lamellae compared to the as-printed state (Fig. 1d).
The lamellar interfaces impede the motion of dislocations, effectively blocking the slip transmission from the soft FCC lamellae into hard BCC lamellae (Fig. 4d and Supplementary Fig. 6b). The massive pile-up of dislocations along the lamellar interfaces shown in Fig. 4d is consistent with the MD result (Stage III, Fig. 4b), contributing to the substantial strain hardening. These local pile-ups at the interface may serve as emission sources to activate the slip systems within the BCC lamellae (Fig. 4e). On the other hand, with the accumulation and nucleation of dislocations from interfaces, the semi-coherent FCC/BCC interfaces also allow efficient slip transmission across lamellar interfaces, sustaining the continuity of plastic deformation16,24. This can be evidenced by the straight and continuous slip traces across lamellar interfaces exhibiting the K–S relationship in Fig. 4e, and by the dislocation transmission across the lamellar interface with the N–W relationship in Supplementary Fig. 6b. Between these two orientation relationships, the K–S relationship is generally more favorable for slip transmission across the semi-coherent interfaces due to better crystallographic alignment16. Additionally, because of the increased fraction of the FCC phase in the Al19Co20Fe20Ni41 alloy, minor interlinked FCC lamellae are not fully constrained by the hard BCC lamellae. This enables the activation of minor deformation twins in the FCC lamella from the phase boundaries (Supplementary Fig. 6c, d), which alleviates the localized stress concentration and enhances the strain hardening capacity25.
One key origin of the high ductility (over 20%) of the as-printed nanolamellar Al19Co20Fe20Ni41 alloy is the enhanced HDI strain hardening originating from the hierarchical heterogeneity. During the plastic deformation, the interaction between the nanoprecipitates and the lamellar matrix produces localized strain gradients21. Moreover, the mutual deformation constraints between the FCC and BCC lamellae generate nanoscale strain gradients, inducing the formation of geometrically necessary dislocations (GNDs)27. Beyond the nanolamellar hetero-deformation behaviour, the heterogeneous BCC-rich and BCC-lean zones (Fig. 1e) introduce additional mesoscale strain gradients into the as-printed EHEA, which are accommodated by extra GNDs. These hierarchical strain gradients further saturate dislocation density and lead to enhanced HDI strain hardening, maintaining the strain hardening rate sufficiently high to postpone localized deformation that results in necking23. The estimated HDI stress (Supplementary Fig. 7) in this work, reflecting the offset between back stress and forward stress, remains in the range of 750–900 MPa during plastic deformation. Therefore, the hierarchical heterogeneity manipulated by L-PBF facilitates the increased GNDs, sustains a high strain hardening rate, and improves the ductility.
Nanovoids formation during plastic deformation
Another critical factor leading to the high ductility is the formation of nanovoids in the BCC lamellae during the latter stage of plastic deformation, which turns out to prevent catastrophic failure. In SEM observation (Fig. 5a), nanovoids (~100 nm in diameter) form within the BCC lamellae at a strain of 10%, initiating from the phase boundaries. With increasing strain (15–20%), more nanovoids appear and propagate within the BCC lamellae. Instead of causing rapid rupture of the lamellae, these nanovoids grow under the confinement of the adjacent FCC lamellae (already hardened at this stage), which effectively delays the failure of the alloy28. At the final stage of deformation, the coalescence of these nanovoids breaks the constraints of the FCC lamellae, resulting in the fracture of the EHEA (Supplementary Fig. 8a).
Fig. 5: Nanovoid formation in the EHEA during plastic deformation.
a SEM images showing the evolution of nanovoids (indicated by red arrows) with increasing strain. b, c Molecular dynamic (MD) simulations showing the atomistic evolution during plastic deformation and the corresponding relative Hardy stress distribution (along the tensile direction). The FCC, BCC, hexagonal close-packed (HCP), and amorphous atoms are colored in pink, blue, yellow, and gray, respectively. The localized Hardy stress of an atom was obtained by averaging tensile stress within its cutoff distance of 1 nm.
The MD simulation results in Fig. 5b, c demonstrate the nucleation and growth of nanovoids near the lamellar interfaces under a triaxial tensile stress state. The tendency for nanovoid formation might be attributed to the solid-state amorphization of the BCC phase in the Al19Co20Fe20Ni41 alloy, which is driven by the localized stress concentration29 (Fig. 5b and Supplementary Fig. 8b). In a triaxial stress environment, volumetric expansion stretches atomic bonds uniformly in all spatial directions. Under high hydrostatic tensile stress, the uniform bond stretching destabilizes local atomic coordination, effectively disrupting the long-range crystalline order and promoting a crystalline-to-amorphous structural transition. As deformation proceeds, the local amorphous regions further evolve into nanovoids, facilitating stress relaxation (Fig. 5c)30. The formation of nanovoids reduces localized stress concentrations by redistributing stress into adjacent crystalline regions, which may further propagate amorphization and drive nanovoid enlargement under continued deformation.
The stress relaxation associated with the nanovoid formation prevents intergranular cracks and reboots the transfer of plastic deformation across the FCC/BCC interfaces. The occurrence of nanovoids (Fig. 5a) matches well with the inflexion point in the strain hardening curve at ~9% strain (Fig. 4a), where the stress concentration at the lamellar interfaces is alleviated. During the subsequent plastic deformation, the mobile dislocations could also be hindered by these nanovoids and amorphous regions (Supplementary Fig. 8c), thus suppressing substantial drop in the strain hardening rate31. Hence, the dual role of nanovoids within BCC lamellae contributes to the enhanced ductility of the EHEA.
Discussion
Our work highlights a path to designing advanced structural alloys with high strength–ductility synergy by integrating multiple deformation mechanisms. By integrating composition design with an advanced manufacturing process, the Al19Co20Fe20Ni41 EHEA achieves a combination of strength- and ductility-improving factors, including an increased fraction of the ductile phase, nanolamellar structures dispersed with coherent nanoprecipitates, hierarchical strain gradients, and the potential to trigger deformation-induced nanovoids that are confined by adjacent hardened lamellae. These microstructural features synergistically contribute to the high strength and extensive strain hardening capacity of the as-printed alloy.
Previously reported manufacturing processes have also demonstrated the capability to create heterogeneous microstructures with superior mechanical properties18,25,32. However, these approaches generally rely on complex processing routes, such as multiple-pass thermomechanical treatments, and face limitations in shaping flexibility18,25 and scalability32. The resulting materials are typically manufactured in simple geometries, such as thin sheets or rods, which restricts their applications or incurs high costs during subsequent machining into complex final products. In contrast, L-PBF enables the direct fabrication of high-performance components with complex, application-specific geometries. With ongoing advancements of AM systems, the capability to fabricate larger components continues to expand (from millimeters to meters). Furthermore, the unique layer-by-layer fabrication mode of AM offers greater freedom in locally tailoring microstructural heterogeneity. The strategy in this work provides a cost-effective approach for fabricating high-value alloy components and can be readily extended to the design of other dual-phase HEAs or heterogeneous metallic structural materials requiring certain geometrical complexity and scalable manufacturing.
Methods
Sample preparation
The Al19Co20Fe20Ni41 EHEA samples in this work were fabricated by an EOS M290 L-PBF machine using gas-atomized powder (10–53 μm in diameter) with a nominal composition shown in Supplementary Table 5. The maximum power, wavelength, and beam size of the ytterbium fibre laser were 370 W, 1060 nm, and 100 μm, respectively. All print batches were pre-heated at 80 °C and conducted in an argon atmosphere. To optimize the process parameters for printing the EHEA, we printed cubic EHEA samples (10 mm × 10 mm × 10 mm) using laser power of 210–350 W, scanning speeds of 600–1200 mm/s, hatch spacings of 60–120 μm, and scan rotation angles of 0°, 67°, and 90°. Upon the screen of the process parameters (Supplementary Fig. 1a), we refined process window to ensure the relative densities of the printed EHEA cubes higher than 99.8% (Supplementary Fig. 1b). The microstructure and mechanical properties shown in this work were produced using an optimized parameter set including a laser power of 350 W, a scanning speed of 800 mm/s, a hatch spacing of 120 μm, a layer thickness of 40 μm, and a scan rotation of 67°. The cuboid samples (40 mm length × 10 mm width × 10 mm height) were printed for mechanical tests and were cut into dog-bone-shaped tensile plates (gauge area, 8 mm length × 2 mm width × 1 mm thickness) by electric discharge machining.
CALPHAD simulation
Thermodynamic calculation of phase diagram (CALPHAD) and non-equilibrium Scheil–Gulliver solidification simulations were performed using the Thermo-Calc software33 (version 2024b) and the TCHEA7 database. The equilibrium phase diagram of the Al19Co20Fe20Ni41 EHEA was calculated within a temperature range of 0–1400 °C. The equilibrium phase compositions of the high-temperature phases (i.e., simple FCC and ordered BCC (B2)) were estimated above 700 °C where the phases were stable. The Scheil–Gulliver simulation was performed assuming no and infinite diffusion in the solid and liquid, respectively.
Microstructure characterization
The EHEA samples were prepared following standard metallographic procedures for density measurement and microstructure characterization. The relative densities as function of process parameters were quantified by optical microscopic imaging of polished cross-sections. We employed scanning electron microscope (SEM) and transmission electron microscope (TEM) to characterize the hierarchical as-printed features. We conducted backscattered electron (BSE) imaging, secondary electron (SE) imaging, electron backscatter diffraction (EBSD, Oxford Instrument, Symmetry), and Transmission Kikuchi Diffraction (TKD)–energy dispersive spectroscopy (EDS, Oxford Instrument, Ultim Max 170) detection on a JEOL JSM-7800F Prime field emission SEM. The EBSD data were acquired using Oxford Instruments AZtec software (version 6.1) with step sizes ranging from 0.1 µm to 0.25 µm. The step size and corresponding indexing rate for each scan are summarized in Supplementary Table 6. A MATLAB (version R2018A) toolbox, MTEX (version 5.6.1)34, was used to analyze the captured EBSD data. Grain reconstruction was carried out using a 15° misorientation threshold to define grain boundaries, which were subsequently smoothed with a smoothing parameter of 3. A polished sample was etched by Marble’s reagent for 30 s for the quantification of nanolamellar thickness, which was averaged from at least ten SEM images.
To deeply understand the strengthening and strain hardening behavior of the as-printed EHEA, the TEM and STEM imaging were performed on an aberration-corrected JEOL ARM-300F at 300 kV. The STEM–EDS measurement was carried out to estimate respective compositions of FCC and BCC lamellae, using an aberration-corrected JEOL ARM-200F equipped with an EDS detector (Oxford Instruments) at 200 kV. TEM samples were twin-jet electropolished by a Tenupol-5 automatic system, using a mixed solution of 6% perchloric acid and 94% ethanol at −40 °C.
Atom probe specimens were prepared using a Tescan Amber X2 Plasma focused ion beam (PFIB) using standard in-situ lift-out approaches. TKD mapping was used for FCC and BCC phase identification with a 30 kV acceleration voltage, 1 nA beam current and an Oxford Instruments Symmetry S3 EBSD camera. Atom probe data was collected using a Cameca Invizo 6000 ultra-wide field-of-view atom probe equipped with a deep-UV laser. Data acquisition was conducted at a temperature of 50 K, laser energy of 200 pJ, pulse rate of 200 kHz, and target detection rate of 6%. Data reconstruction was performed using the Cameca IVAS 6.3 software. For the precipitate analysis cylindrical regions of interest of 30 × 30 x 120 nm3 were extracted from the FCC and BCC phases. Proximity histograms (proxigrams) were extracted with a step size of 0.05 nm, whereas complex ion decomposition and background correction were applied to calculate average matrix and precipitate compositions. Average precipitate diameters were calculated from the iso-concentration surface extent distributions across the x-, y-, and z-dimensions.
Mechanical testing
The tensile tests were conducted using a strain rate set at 0.001 s−1 on a SHIMADZU AG-X plus machine equipped with a TRViewX video extensometer. The tests were repeated five times. To evaluate the HDI stress contribution in the as-printed EHEA, we carried out loading–unloading cycles during three tensile tests at various plastic strains (i.e., 2.5%, 5%, 7.5%, 10%, 12.5%, 15%, 17.5%, and 20%) on the same mechanical tester (Supplementary Fig. 7). Upon loading to each specific strain level, the samples were unloaded to 20 N in a force-control mode at a rate of 20 N/s before being re-loaded to the next designated strain level.
Molecular dynamics simulations
Molecular dynamics (MD) simulations were performed to investigate the deformation behaviour of the Al19Co20Fe20Ni41 EHEA, using empirical embedded atom method (EAM) potentials35,[36](#ref-CR36 “Zhou, X. W., Johnson, R. A. & Wadley, H. N. G. Misfit-energy-increasing dislocations in va