Process–microstructure–property relationships in additive manufacturing of high-entropy alloys: recent advances, challenges and future perspectives

1 Introduction

High-entropy alloys (HEAs) are near-equimolar alloys containing five or more elements, often forming stable single-phase solid solutions instead of intermetallic compounds, although many HEAs are intentionally designed with multiphase microstructures for enhanced properties. This “high-entropy” concept opened an enormous, largely unexplored compositional space. The four core effects of HEAs, viz. Higher entropy, sluggish diffusion, severe lattice distortion, and cocktail effect (Yeh, 2013; Zhang et al., 2014) are attributed to their exceptional mechanical performance that often exceeds that of conventional alloys. Figure 1 illustrates the four core effects of HEAs and their corresponding influence on HEA behaviour and properties.

Their suitability for demanding sectors including aerospace, automotive, energy, biomedical, and electronics has further increased interest in their manufacturing/processing routes (Ren et al., 2025; Jalali and Nayebpashaee, 2025). Figure 2 shows some of the potential applications of HEAs. Conventional processing methods for HEAs such as vacuum arc melting (Dehury et al., 2024), induction casting (Dudala et al., 2024), mechanical alloying (Ji et al., 2014), and physical vapour deposition (Kumar et al., 2024) are well established but have certain limitations, such as casting, where multiple principal elements are remelted and flipped multiple times to ensure the chemical homogeneity of the alloys (Dehury and Gautam, 2024). Additionally, a post-processing step is usually carried out to further tailor the microstructure or phase formation of the produced HEAs to achieve desired properties. For certain HEAs, high cooling rates are essential to suppress the formation of undesirable intermetallic phases (Gu et al., 2012). Consequently, these limitations hinder the fabrication of complex-shaped HEA components and restrict the scalability of conventional processing techniques for large-scale production (Li, 2018).

Recently, there has been considerable interest in producing these alloys by additive manufacturing (AM) technologies. AM offers several advantages over conventional fabrication techniques, including the fabrication of complex geometries with high precision and accuracy with minimal material waste, high freedom of design, and also higher heating and cooling rates (Gu et al., 2012; Frazier, 2014). Due to high cooling rates and repeated thermal cycling, AM generates ultrafine, non-equilibrium microstructures that often differ sharply from their cast counterpart. Rapid cooling rates also enhance chemical homogeneity in HEAs, suppress unwanted phase transformations by limiting diffusion, and enable fabrication of large, complex components with high precision, efficiency, reduced defects, and superior geometric flexibility (Abdullah and Peng, 2024). The study of microstructural evolution, property enhancement, and defect mitigation in additively manufactured (AMed) HEAs has attracted significant attention. As a result, AMed HEA research has emerged as a rapidly growing field. This review summarises recent advances, microstructural Evolution, mechanical Properties, key challenges, and future directions in AMed HEAs.

2 Recent advances in AM-HEA fabrication

2.1 Additive manufacturing processes applied to HEAs

There are several AM technologies that have been adapted for use in HEA fabrication, each with distinct thermal histories and microstructures. The following provides an overview of these techniques, as well as their relevant pros and cons. Figure 3 shows some of the most important AM technologies.

2.1.1 Laser powder bed fusion (LPBF)

This is one of the most well-studied techniques. It employs a focused laser to selectively melt layers of powder. LPBF is capable of producing HEAs with refined cellular or columnar-dendritic microstructures, while reducing the formation of intermetallic compounds and elemental segregation owing to its exceptionally high cooling rates (10⁵–10⁷ K s^–1) (Jarlöv et al., 2024; Ren et al., 2022). This process combines rapid melting and solidification coupled with successive layer thermal cycles and directional heat removal through previously solidified material. Such a thermal history restricts atomic long-range diffusion and elemental partitioning, leading to the stabilisation of many HEAs in a metastable solid solution (Jarlöv et al., 2024). Rapid cooling helps maintain single-phase Face-Centered Cubic (FCC) or Body-Centered Cubic (BCC) solid solutions and prevents the separation of equilibrium phases (Diebel et al., 2026a). Additionally, LPBF encourages the development of dense dislocation networks and small cellular or sub-cellular structures, usually ranging from 0.2 to 1.0 μm within primary grains (Shu et al., 2025). Strong thermal gradients also support the elongation of columnar grains in the build direction, although grain refinement into equiaxed forms can be achieved by adding micro-alloying elements or through particle-induced grain pinning (Moura et al., 2025).

The process of melting and remelting each layer partially homogenises the previously deposited layer, resulting in highly dense structures that often negate the need for post-processing (Wang et al., 2018) and improve mechanical properties (Me et al., 2023). Other benefits include the rapid development of alloys through mechanical alloying, exceptional geometric precision allowing for unprecedented design flexibility, and the cost-effective production of specialised low-volume components (Jarlöv et al., 2024). For example, Brif et al. (2015) reported that LPBF-fabricated CoCrFeNi exhibited nearly twice the yield strength of its as-cast counterpart. LPBF has also demonstrated feasibility for fabricating refractory HEAs (RHEAs) such as WMoTaNbV, highlighting their potential for aerospace and chemical industry applications. However, the rapid melting and solidification cycles inherent to LPBF face challenges, including thermal cycling, partial remelting, residual stresses, and microcrack formation near the print/substrate interface, increasing hot-cracking susceptibility in superalloys (Lu et al., 2020). High residual stresses can further induce thermal cracking in alloys such as CoCrFeNi. Therefore, careful optimisation of laser energy density minimises residual stress as much as possible (Zhang et al., 2024). Additionally, porosity and poor process repeatability can also affect mechanical properties and reliability; therefore, stable processing strategies are necessary to control defects and parameter optimisation to ensure reproducible mechanical properties (Burgio and Moeini, 2025; Zhang et al., 2025).

2.1.2 Selective electron beam melting (SEBM)

This method combines high vacuum, high beam power, high scan rate, and high-temperature powder-bed preheating, which supports processing of HEAs (Hui et al., 2025; Chen, 2024). All of these factors control the thermal history of the material and subsequently the microstructure, defects, and the mechanical properties of the material. The vacuum environment eliminates oxidation and contamination and allows the use of more reactive and high-melting-point elements typically found in HEAs (Körner, 2016).

Preheating of the powder bed between 600 °C–1100 °C is a unique feature of SEBM and is performed using a defocused electron beam (Ron et al., 2023). When compared to LPBF, this process reduces thermal gradient and residual stress and thus is very helpful in the fabrication of RHEAs, which are prone to cracking (Wang and Tang, 2016). The relatively lower cooling rates associated with SEBM also reduce lattice distortion and thermal shock, thereby promoting the formation of equiaxed grains over fine dendritic structures observed in LPBF (Diebel et al., 2026b).

SEBM additionally provides rapid solidification rates of 10⁴–10⁵ K/s, precise local process management, and the capability to produce complex shapes directly from 3D-CAD models (Chen, 2024). These features often result in more refined microstructures and better mechanical properties than conventional casting methods for HEAs (Taghian et al., 2025; Zhang et al., 2020; Kuwabara et al., 2018). The defect formation shows that high-temperature processing conditions reduce cracking caused by residual stress and ensure that the refractory components are completely melted (Popov et al., 2019) (Wang et al., 2019). Nonetheless, the combination of high vacuum and beam power can cause selective evaporation of elements with high vapour pressures, such as Al, Mn, and Cu, resulting in unintended changes in alloy composition (Popov et al., 2019). Therefore, it is imperative to carefully optimise process parameters to control the porosity and compositional variation. Moreover, the dimensional accuracy and surface quality of the SEBM process may be affected by the application of larger beam diameter and thicker layers than in the LPBF process (Chen, 2024). Overall, SEBM represents a promising route for HEAs, as it enables complex, low-stress parts with often strong mechanical performance; however, its broader use still depends on tighter control of evaporation, segregation, porosity, and process stability.

2.1.3 Directed energy deposition (DED)

It is an AM process that uses focused thermal energy (laser, electron beam, or arc) to melt and deposit powder or wire feedstock layer by layer (Chen, 2024). DED offers several advantages over Powder Bed Fusion (PBF), such as a larger component size, higher deposition rates and better material utilisation, as powder-bed size limitation and material waste are mitigated. The typical cooling rate range for DED is intermediate (10²–10⁴ K/s), and the temperature gradients are high, which lead to thermal cycling, reheating and deposition of successive layers at the site of the targeted regions (Li et al., 2023). The different thermal profile of DED can also lead to different microstructures than LPBF or conventional casting, which can vary significantly in terms of their heterogeneity, hierarchy, and anisotropy (Li et al., 2023). With rapid solidification, grain refinement and/or a decrease in elemental segregation is encouraged in DED. Furthermore, due to its low dilution and small heat-affected zone, it can effectively eliminate cracking, deformation, metallurgical changes, and improve the production quality and efficiency (Jeong et al., 2024). However, slower solidification compared to PBF encourages the growth of epitaxial columnar grains in the build direction. Dynamic recrystallisation (DRX) may occur due to localised micro-strain and repeated thermal cycling, partially breaking up columnar structures and resulting in finer grains. Moreover, the slow diffusion typical of HEAs can limit elemental partitioning and support the stabilisation of single- or dual-phase FCC/BCC solid solutions (Cagirici et al., 2024).

Despite these advantages, defects such as interlayer cracking, lack of fusion, incomplete powder melting, and porosity may be introduced during the DED process that could negatively impact the integrity and performance of the fabricated component (Wang et al., 2023; Xiao et al., 2022). High scanning speed is associated with solidification cracking, and high-angle grain boundaries have a higher tendency to form cracks (Chen and Qi, 2022). In addition to this, spherical inclusions of Mn-oxide were found in the upper part of the deposited material in DED-fabricated CoCrFeMnNi, while the LPBF-fabricated counterparts usually show balling and lack-of-fusion defects (Savinov et al., 2021). Nevertheless, DED-processed HEAs can demonstrate high hardness and elevated-temperature strength owing to severe lattice distortion. However, columnar grains and crystallographic texture may introduce mechanical anisotropy (Lee et al., 2026). More importantly, the strength and elongation of CoCrFeMnNi produced using DED are greater than those of the LPBF counterparts, which can be attributed to the in-situ annealing during the thermal cycling process that removes the tensile stresses of the parts (Ron et al., 2023).

2.1.4 Additive friction stir deposition (AFSD)

This process combines the principles of friction stir welding and AM in a fully solid-state process, eliminating melting and significantly reducing defects such as solidification cracking. In the AFSD process, the feedstock is melted by frictional heating and severe plastic deformation above the solidus temperature but below the melting temperature of the material (Mishra et al., 2022). Typically, the maximum processing temperature is approximately 0.7–0.8 T_m. Since no liquid-to-solid phase transformation occurs, this eliminates the high thermal gradients and solidification conditions that lead to dendritic growth, segregation and cracking that occur in fusion-based AM (Yu and Mishra, 2021; Gotawala and Yu Hang, 2023; Ge et al., 2023). The severe shear deformations and sub-solidus processing conditions promote dynamic recrystallisation (DRX), producing highly refined, equiaxed grains rather than the columnar or coarse grains commonly associated with fusion-based processes (McClelland et al., 2024; Yu and Mishra, 2021). Nevertheless, spatial variations in temperature and material flow can generate microstructural heterogeneity within the deposited material. For instance, in the case of Al₀.₃₅CoCrFeNi HEA, phase variations between the advancing and retreating sides were observed, with increased formation of Al-rich BCC particles on the retreating side. These particles promoted Zener pinning and restricted grain growth (McClelland et al., 2024).

One of the major advantages of AFSD is its inherent defect tolerance. The absence of a liquid phase eliminates solidification defects such as hot cracking, keyhole porosity or lack-of-fusion pores, which allows the production of highly dense deposits with improved structural reliability (Korgancı and Bozkurt, 2024; Nouranga et al., 2025). Furthermore, AFSD can also improve the mechanical performance through grain refinement and homogeneous microstructures. AFSD prevented brittle intermetallic formation at the particle–matrix interfaces, which increased the strength and ductility of AA5083 reinforced with CoCrFeNi HEA particles (Chen, 2024).

Despite these advantages, AFSD is more appropriate for relatively low melting point materials at present. The use of high-temperature alloys and HEAs is still in its early stages, and there is a need to develop special wear-resistant tooling that can withstand extreme thermomechanical conditions (Mishra et al., 2022). Additionally, precise control of build-direction geometry is essential for smooth edges (Garcia et al., 2020), while low technology readiness, process-control limitations, and property-uniformity issues continue to restrict widespread adoption (Yasa et al., 2024). The next section discusses feedstock considerations and powder quality effects on AM component microstructure and properties.

2.2 Feedstock considerations and powder quality

Up to now, the commercial use of HEAs has still not succeeded, primarily due to the lack of availability of the feedstock material and comprehensive qualification of new alloys, especially for applications in aerospace and energy sectors (Gerdt et al., 2023). Metal powders and wires are important feedstocks for AM production of complex and high-performance parts (DebRoy et al. et al., 2018). Powders are widely used in PBF techniques such as selective laser melting (SLM) and selective electron beam melting (SEBM), where fine, uniformly sized particles are preferred for improving mechanical properties and reducing porosity. The conventional production of HEA powders is based on mechanical alloying (MA), which, however, suffers from low powder quality in terms of contamination and irregular particle shape, making them not suitable for powder-bed AM. Water atomization is a cost-effective method, but the irregular powders obtained have low flowability and packing density, limiting their use for PBF-based HEA production (Wang et al., 2019).

Gas-atomised powder is the most established HEA AM feedstock route (Chen, 2024). For powder bed HEA AM, gas atomization is the dominant route because it yields pre-alloyed powders that are typically spherical in nature, with less contamination, compositionally homogeneous, and produced at AM-relevant scale (Chen, 2024, Torralba et al., 2019). Figure 4 illustrates equiatomic CoCrFeMnNi HEA powder produced by vacuum induction melting and argon gas atomization. The as-received powder exhibited predominantly spherical particles, with some irregular morphologies and numerous satellites. These satellites formed due to collisions between partially solidified droplets caused by turbulent flow within the atomization chamber (Wang et al., 2019). The gas-atomised HEA powders exhibited suitable density, flowability, and particle size distribution for AM applications, making gas atomization a promising route for large-scale HEA powder production (Wang et al., 2019).

However, powder-based additive manufacturing is limited by low deposition efficiency and poor material utilisation, making fabrication of large components challenging (Shen et al., 2021). Metal wires, on the other hand, are used as feedstock in wire-based DED processes such as Wire Arc AM, Wire Laser AM, and Wire Electron Beam AM. These methods have gained attention due to their high deposition rates, cost-effectiveness, and ability to manufacture large, customised, near-net-shape components (Ron et al., 2023). Wire feedstock is generally produced through extrusion or drawing processes. For the first time, Shen et al. (2021) developed a novel combined cable wire (CCW) with multi-element composition as filler material for wire arc additive manufacturing of non-equiatomic AlCoCrFeNi HEA. The CCW consisted of seven 0.6 mm diameter metal wires arranged with one central and six peripheral wires, including two Ni, two Al, one Fe, one Co., and one 304 stainless steel wire, as shown in Figure 5.

Both metal powders and wires require strict quality control in AM. Powder characteristics, such as particle size distribution, influence flowability and layer density, whereas wire diameter and composition consistency determine deposition reliability and final part quality.

3 Microstructural evolution, anisotropy, and post-processing effects in AMed HEAs

3.1 Microstructural evolution in AMed HEAs

A major recent trend is the deliberate engineering of heterogeneous microstructures that combine both fine and coarse grains, dual-phase structures, and gradient layers within the AM-built HEAs to improve strength and ductility simultaneously. Rapid cooling in layer-wise deposition minimises composition and phase segregation and brittle intermetallic formation, while promoting grain refinement and enhanced strength, making AM a promising route for HEA fabrication (Mishra and Thapliyal, 2021, Leijon et al., 2021). It is important to note that the local solidification and heat transfer within the melt pool govern cellular morphology and growth. The non-equilibrium conditions of AM produce distinctive fine-grained or columnar microstructures that differ fundamentally from those of conventionally processed alloys. AMed HEAs typically exhibit columnar grains due to directional heat flow and epitaxial solidification. This promotes strong crystallographic texture along the build direction, producing anisotropic microstructures with direction-dependent properties (Piglione et al., 2018, Joseph et al., 2017). However, equiaxed grains can also emerge when constitutional undercooling promotes heterogeneous nucleation and fragmented dendrites, reducing thermal gradients through scan strategy, which increases the nucleation ahead of the interface (Zhang et al., 2022, Guan et al., 2019). However, the large thermal gradients during AM generate heterogeneous tensile and compressive residual stresses between successive layers, which may lead to distortion and crack initiation that significantly degrade the structural integrity and performance of fabricated components (Zhang et al., 2022). Experimental observation has also revealed a columnar-to-equiaxed transition during microstructural evolution within the molten pool. For instance, Li et al. (2021) fabricated single-track CoCrFeMnNi HEA components using laser melting deposition and investigated microstructural evolution across the molten pool. The study revealed a columnar-to-equiaxed grain transition along the deposition direction. Figure 6 shows the Optical Microscopy and inverse pole figure mapping, which showed distinct microstructural regions consisting of planar, columnar, and equiaxed grains. Planar grains dominated the initial molten pool region (Figure 6c), while epitaxially grown columnar grains prevailed in the central region (Figure 6d). Near the molten pool boundary (Figure 6e), the microstructure transformed into uniformly distributed equiaxed grains with sizes ranging from 10 to 50 μm. Similarly, Sui et al. (2022) reported that DED-fabricated equimolar AlCoCrFeNi HEA evolved from columnar grains to equiaxed dendritic structures along the building direction.

3.2 Anisotropy mechanism

The layer-by-layer fabrication nature of additive manufacturing inherently introduces crystallographic and morphological anisotropy. Directional solidification during processing promotes the development of strong textures, which are typically aligned with the build direction, whereas the degree of texture formation is influenced by the volumetric energy density and scanning strategy (Zhang et al., 2022; He et al., 2022).

Zhang et al. (2023) investigated the influence of stripe and chessboard laser scanning strategies on the microstructure, crystallographic texture, and mechanical behaviour of LPBF-fabricated Fe_49.5Mn₃₀Co₁₀Cr₁₀C_0.5 interstitial-strengthened HEA. Their study revealed that mechanical anisotropy originates from heterogeneous grain morphology and texture evolution induced by different scan patterns. The stripe scanning strategy produced pronounced anisotropy because of the formation of elongated columnar grains preferentially aligned along the build direction and a strong <110> texture. Although both stripe and chessboard samples exhibited a dominant <110> orientation parallel to the build direction, the stripe-scanned alloy additionally developed a strong <001> texture along the transverse direction, resulting in greater microstructural heterogeneity and orientation-dependent mechanical properties (Zhang et al., 2023).

Wang et al. (2020) investigated the microstructural, mechanical, and corrosion anisotropy of SLM-fabricated CoCrFeMnNi HEA. The anisotropy originates from a columnar grain with FCC structure formed under the steep thermal gradients inherent to the SLM process. These elongated FCC grains grow preferentially along the build direction, resulting in orientation-dependent mechanical behaviour. Consequently, the specimens tested in the XOY plane exhibited higher yield strength (YS) and ultimate tensile strength (UTS) but lower elongation compared with those in the XOZ plane.

Similarly, Liu L. et al. (2023) examined the mechanical and structural anisotropy of SLM-fabricated (FeCoNi)₈₆Al₇Ti₇ HEA. The alloy developed epitaxial columnar grains aligned with the build direction, together with a cellular substructure formed by directional solidification and cyclic thermal exposure. Variations in grain size and crystallographic texture across different orientations were identified as the primary sources of yield-strength anisotropy. In contrast, tensile ductility anisotropy was mainly governed by the orientation-dependent crack propagation behaviour associated with the melting pool boundaries and grain boundaries.

More broadly, studies on LPBF/SLM-processed HEAs have shown that crystallographic texture can alter the dominant deformation mechanism from dislocation slip to deformation twinning, significantly influencing strain hardening and ductility. Consequently, the mechanical performance becomes strongly dependent on the loading direction, even when the phase constitution remains unchanged (Dixit et al., 2023; Charmi et al., 2021).

3.3 Post-processing effects

Post-processing treatments are essential to unlock AM-HEAs’ full potential by homogenising microstructures, relieving residual stresses, and reducing defects. Significant improvements in mechanical properties, including strength, hardness, fatigue resistance, and ductility, have been reported following post-processing treatments.

For example, annealing of SLM-fabricated FeCoCrNi HEA demonstrated the critical role of temperature in microstructural evolution, with an optimal grain refinement temperature of approximately 700 °C. Annealing reduced yield and ultimate tensile strengths while markedly improving elongation to failure, indicating a favourable strength–ductility trade-off (Lin et al. et al., 2020; Lin et al., 2020).

Hot isostatic pressing (HIP) is widely employed to enhance AM-HEAs by eliminating microcracks, microporosity, and residual stresses. HIP also promotes chemical homogenization; for instance, in DLF-fabricated Al_0.3CoCrFeNi alloy, it effectively dissolved grain-boundary segregations and secondary phases (Gu et al., 2012). Similarly, annealing treatments can tailor mechanical performance through stress relief, grain growth, and controlled phase evolution, with the final properties depending strongly on treatment temperature (Chen et al., 2018).

Solution treatment of SEBM-fabricated CoCrFeNiTi HEA improved ductility without sacrificing strength through the dissolution of the Ni3Ti phase (Fujieda e t al. et al., 2017). Likewise, in SLM-produced CoCrFeMnNi HEA, heat treatment above 900 °C reduced dislocation density and homogenised chemical segregation, although prolonged annealing promoted the formation of Cr-rich σ phase (Zhang et al., 2022).

4 Mechanical properties of additively manufactured HEA

The mechanical properties of the AM parts are found to depend on the density and the microstructure. These properties are highly sensitive to process parameters such as laser power, energy input, scan speed, and thermal history. This is because these parameters control the texture, segregation, cell size, and residual stress. Moreover, better-tuned parameters reduce defects and anisotropy, while poorly chosen energy input introduces pores or cracks and degrades ductility (Ron et al., 2023). Compared to the parts produced using traditional methods, the mechanical properties of AMed HEAs are higher owing to the rapid solidification observed in AM techniques that leads to a finer microstructure and increased dislocation density (Wu et al., 2024a). Moreover, the same HEA system was found to exhibit varied tensile properties depending on the AM techniques utilised and the processing parameters chosen (Herzog et al., 2016). For instance, Xiang et al. (2019) showed that the mechanical properties of CoCrFeMnNi can be tailored by adjusting the laser power and scanning strategy. Increasing the laser power from 1000 to 1400 W transformed the dominant ⟨001⟩ texture into a random texture, significantly altering the alloy’s mechanical performance.

A comparative study of laser-directed energy deposition (LDED) and LPBF HEAs showed that both processes resulted in the formation of a heterogeneous microstructure comprising columnar grains, solidification cells and dislocation networks (Liu Y. et al., 2023). LPBF-manufactured samples, on the other hand, have a major <001> texture along the build direction, smaller solidification cells, and less pronounced segregation of the elements compared to LDED-fabricated samples, which have a mixed <011>/<111> texture, coarser solidification cells, and greater elemental segregation. The mixed texture is found to increase flow stress and deformation twinning, and larger cells are found to increase dislocation storage and hinder dislocation motion. Consequently, LDED samples develop pronounced planar slip and microband formation, which in turn results in sustained strain hardening and superior ductility as compared to LPBF counterparts (Liu Y. et al., 2023).

LPBF-fabricated HEAs have shown exceptional mechanical performance, with yield strengths approaching 1.3 GPa and uniform elongations of approximately 14%, surpassing many additively manufactured metallic alloys (Ren et al., 2022). The high strength of these alloys is primarily due to the presence of alternating FCC/BCC nanolamellar structures, where the stronger BCC lamellae provide significant strengthening and strain hardening. Simultaneously, the hierarchical microstructure comprising dual-phase nanolamellae embedded within microscale eutectic colonies enhances work-hardening capability and tensile ductility while promoting nearly isotropic mechanical behaviour. Kim et al. (2020) further investigated the microstructure, temperature-dependent mechanical properties, and deformation behaviour of SLM-fabricated CoCrFeMnNi HEA. Figure 7 presents (a) compressive stress-strain curves and (b) temperature-dependent yield strengths. The alloy exhibited superior compressive strength compared to annealed (Jang et al., 2016) and other AMed counterparts (Chew et al., 2019; Wang et al., 2019), attributed to the synergistic effects of dislocation networks, refined grains, and nanosized Mn₂O₃ precipitates. At 25–600 ^oC, the superior strength was governed by fine grains, dislocation networks, and precipitates, whereas at ≥700 ^oC it was primarily associated with dislocation density and precipitate strengthening. Slip and deformation twinning dominated deformation at 25–600 ^oC, while slip accompanied by partial recrystallisation became prevalent above 700 ^oC (Kim et al., 2020).

Recent studies have shown the exceptional mechanical performance of SLM-fabricated eutectic HEAs. The as-built Al₁₈Co₃₀Cr₁₀Fe₁₀Mo₁Ni₃₀W₁ eutectic HEA exhibited outstanding strength over a wide temperature range (−196 ^oC–760 ^oC), representing one of the highest strengths reported for AM-produced HEAs (Yang et al., 2021). This performance was attributed to ultrafine eutectic lamellae (150–200 nm) arranged within small lamellar colonies (2–6 μm). Similarly, LPBF-processed Fe₄₀Mn₂₀Co₂₀Cr₁₅Si₅ HEA exhibited an ε-HCP-dominated microstructure and demonstrated an excellent strength–ductility balance through the combined effects of transformation-induced plasticity (TRIP) and twinning-induced plasticity (TWIP) (Agrawal et al., 2020). Nonetheless, AM-induced heterogeneous phase architectures can also produce complex deformation behaviour, including tension–compression asymmetry (TCA) (Bai et al., 2022). In an additively manufactured dual-phase AlCoCrFeNi_2.1 (FCC + BCC) HEA, TCA originates from deformation mechanisms active in the FCC phase during compression, including stacking fault–dislocation, stacking fault–stacking fault, and twin–dislocation interactions. In contrast, tensile deformation is dominated by dislocation slip. The alternating FCC/BCC lamellar heterostructure promotes stress concentration at phase boundaries, leading to rapid crack propagation and premature failure under tension, whereas crack growth is slower under compression, delaying failure. Figure 8 shows an ultimate tensile strength vs. fracture elongation plot for various AMed HEAs (Zhang et al., 2022).

Recent advances in additive manufacturing have enabled HEAs with engineered heterogeneous architectures and laminated heterostructures that enhance strength–ductility synergy through pronounced dislocation–interface interactions (Bai et al., 2023; Gao et al., 2025). Bai et al. (2023) fabricated a CoCrNi-Fe₅₀Mn₃₀Co₁₀Cr₁₀ laminated HEA featuring a unique Mortise–Tenon architecture that introduced wavy interfaces in addition to conventional flat interfaces. The larger chemically heterogeneous interfacial region of the Mortise- Tenon architecture promoted stronger interface–dislocation interactions, resulting in a yield strength approximately 36.5% higher than rule-of-mixture predictions. Interface strengthening contributed about 23% of the total yield strength, while the Mortise-Tenon interlocking geometry preserved work-hardening capability and ductility (Bai et al., 2023).

A mortise-and-tenon-inspired mechanically interlocked network (MIN) was developed in SLM-fabricated Fe_28.0 Co_29.5Ni_27.5Al_8.5Ti_6.5 HEA. The MIN architecture effectively redistributes grain-boundary stress concentrations and suppresses crack propagation during deformation, achieving a tensile strength of ∼1152 MPa with 28% elongation (Wu et al., 2024b).

In another study, multi-material LPBF with staggered printing was used to fabricate heterogeneous bimetallic structures from AlCuCoCrFeNi and MnCoCrFeNi HEAs. Planar, rotating-grating, and checkerboard architectures exhibited dual-phase microstructures comprising fine BCC and coarse columnar FCC grains. Strong interfacial bonding through “dual-phase intercalation” enhanced interfacial strengthening and crack resistance, with the checkerboard structure showing particularly high impact resistance and energy absorption (Huang et al., 2025).

Hierarchical heterostructures in AM-HEAs activate multiscale deformation mechanisms, promoting sustained hetero-deformation-induced (HDI) strain hardening and an improved strength–ductility balance (Ke et al., 2025; Gao et al., 2025; Bai et al., 2025a). In AlCoCrFeNi₂.₁ eutectic HEAs, hierarchical structures comprising eutectic colonies, alternating FCC/BCC phases, and nanoprecipitates enable coordinated deformation through slip transmission and dislocation–precipitate interactions, enhancing HDI hardening and strength–ductility synergy (Ke et al., 2025). Similarly, LPBF-processed nanolamellar Al₁₉Co₂₀Fe₂₀Ni₄₁ EHEA achieved a yield strength exceeding 1.3 GPa with ∼20% uniform elongation, attributed to coherent nanoprecipitates, nanolamellar structures, hierarchical heterogeneity, and deformation-induced nanovoids within BCC lamellae (Gao et al., 2025).

In AM (FeCoNi)₈₆Al₇Ti₇ HEA, FCC/L1₂ coherent matrices, incoherent L2₁ precipitates, and chemical cells provide complementary strengthening mechanisms. The FCC matrix accommodates plastic deformation, while L1₂ precipitates strengthen through ordering and L2₁ precipitates through Orowan bypassing. Notably, lens-shaped regions within the L2₁ phase undergo stress-induced martensitic transformation, further contributing to deformation accommodation and mechanical performance (Bai et al., 2025a).

Nonetheless, the mechanical properties of additively manufactured materials at cryogenic temperatures have attracted increasing research interest because of their unique deformation mechanisms and potential for achieving an enhanced strength–ductility balance. For instance, SLM-fabricated nano oxide-reinforced CoCrFeMnNi HEA exhibited exceptional cryogenic strength due to abundant deformation twins formed at cryogenic temperatures (Kim et al., 2022). In metastable HEAs, decreasing temperature generally reduces FCC stability and stacking-fault energy, thereby promoting transformation-induced plasticity (TRIP). However, the specific transformation pathway depends strongly on alloy chemistry and AM-induced microstructural heterogeneity (Kim et al., 2021; Zuo et al., 2025).

The temperature-dependent deformation behaviour of Co-free FeMnSiNiAl HEA further demonstrates this effect (Zuo et al., 2025). At room temperature, plastic deformation is dominated by dislocation slip, accompanied by a modest contribution from deformation-induced FCC→HCP martensitic transformation. At 223 K, enhanced FCC→HCP transformation significantly strengthens the TRIP effect, resulting in a simultaneous improvement in strength and ductility. The subsequent FCC→HCP→BCC sequential transformation provides additional plastic-strain accommodation and delays strain localisation, thereby enhancing ductility. In contrast, at 173 and 77 K, rapid FCC→HCP transformation during the early stages of deformation suppresses the subsequent sequential transformation, leading to substantially higher strength but reduced ductility (Zuo et al., 2025).

A similar strengthening mechanism has been reported in SLM-fabricated (Fe₅₀Mn₃₀Co₁₀Cr₁₀)_100−*x*Si_*x* alloys, where Si addition increases lattice distortion and lowers the stacking-fault energy, particularly at 77 K. These changes promote TRIP and consequently improve both strength and ductility (Li and CHANG, 2025).

More recently, AMed Co₃₄Cr₂₀Fe₃₄Ni₆Mn₆ metastable HEA demonstrated an anomalous TRIP response at 77 K through a reversible γ→ε→γ transformation (Bai et al., 2025b). This reversible transformation substantially enhances strain hardening, approximately doubling the tensile strength while maintaining excellent ductility. Moreover, the dominant deformation mechanisms evolve with temperature: at 298 K, TRIP acts synergistically with full dislocation slip, whereas at 77 K, TRIP progressively competes with and ultimately dominates over dislocation-mediated plasticity (Bai et al., 2025b).

Overall, these findings underscore the critical role of AM-induced microstructural characteristics in controlling cryogenic deformation mechanisms and, consequently, the mechanical performance of HEAs.

5 High entropy alloy composites via additive manufacturing

High entropy alloy composites (HEACs) are emerging metal matrix composites formed by incorporating secondary phases such as borides, carbides, or nitrides into HEA matrices. These materials exhibit enhanced mechanical and functional properties. For example, NbC-reinforced AlCoCrFeNi_x HEA fabricated by laser cladding showed improved wear resistance and hardness due to grain refinement and solution strengthening (Li X. et al., 2019). Li J. et al. (2019) fabricated CrMnFeCoNi-based HEACs using laser melting deposition with 0, 5, and 10 wt% WC additions and evaluated their mechanical properties. Tensile stress–strain results at room temperature and 873 K (Figure 9) showed that WC addition significantly increased both yield and tensile strengths, although ductility decreased. The enhanced strength was attributed to the formation of Cr₂₃C₆ reinforcing precipitates, which promoted heterogeneous grain nucleation and impeded slip-band propagation, thereby improving the overall mechanical performance of the composites.

Similarly, FeCoCrAlCu/ZrO₂ HEACs produced by LMD demonstrated superior wear resistance to FeCoCrAlCu HEA owing to fine, compact microstructures and nanocrystalline phases (Li et al., 2017). In another study, Li et al. (2020) fabricated CNT-reinforced HEA composites on TA2 titanium alloy using LMD of FeCoCrAlCu-based powders. The CNT-modified composites exhibited improved corrosion and high-temperature oxidation resistance. This was attributed to dispersion strengthening and lattice distortion induced by nanoscale ceramic reinforcements. The following section briefly discusses the potential applications and the challenges associated with additively manufactured HEAs.

6 Applications and challenges of AMed- HEA

6.1 Applications of AMed-HEA

Recent advances indicate that HEAs can replace conventional materials with superior performance in aerospace, biomedical, energy, marine, and nuclear applications. AMed HEAs offer optimised microstructures and enhanced properties, driving increasing research interest in their development and applications. Arif et al. (2022) reported that DLD- and SLM-fabricated HEAs exhibited superior mechanical properties compared to conventional manufacturing routes, enabling their application in large-scale and complex 3D industrial components. Busachi et al. (2015) demonstrated the suitability of wire arc AM-based manufacturing for defence applications. Liu et al. (2017) discussed different aerospace applications that can benefit from different AM processes. Kuma et al. (2023) described the development of equiatomic TiTaNbMoZr HEA by the plasma μ - arc additive manufacturing (μ-PAAM) process and its suitability for knee implant applications. It was reported that TiTaNbMoZr HEA offers potential to overcome limitations of existing materials, emerging as a promising metallic biomaterial for knee implants. Similarly, eutectic HEAs processed via AM have been reported to develop alternating nano-lamellar microconstituents spacing within colonies, consequently exhibiting impressive strength–ductility balance and thus have the potential to be used for structural applications (Gupta et al., 2024). Nonetheless, RHEAs made with AM offer several superior properties such as high-temperature stability, little thermal expansion, and excellent corrosion and wear resistance. These properties make them a potential candidate in applications involving thermal protection systems, turbine parts, and rocket nozzles, high-temperature fuel cells, as well as other energy-generating and storage systems (Abdullah and Peng, 2024).

6.2 Challenges of AMed-HEA

Despite significant progress in the field of AM HEAs, several challenges still exist in this field. The key ones include the achievement of precise microstructural control, chemical homogeneity, dimensional accuracy and surface finish, control of undesirable phase formation, densification, mitigating cracking and porosity, among others (Cagirici et al., 2024). The presence of such defects degrades the mechanical properties. Hence, it is important to find ways to optimise the microstructure and hence the mechanical properties of AMed HEA. Additionally, achieving homogeneous net-shape components remains challenging when combining refractory and transition metals due to significant differences in their thermophysical properties. In other words, fabrication and investigation of materials with large melting point differences is another challenge. Furthermore, scaling up RHEA to large-scale production with AM is challenging and expensive. Hence, to fully realise the application potential of AMed HEAs as both structural and functional materials, a strategic approach involving process optimisation and advanced material design is essential.

7 Discussion

Recent studies demonstrate that AM has evolved from an alternative fabrication method to a powerful enabling technology for HEAs. By treating process parameters as design tools, AM enables tailored microstructures and heterostructures, overcoming traditional strength–ductility limitations and shifting materials development toward a microstructure-by-design approach. Table 1 presents a summary of important findings in the additive manufacturing of HEAs, considering composition, process, key parameters, morphology, defects, mechanical properties, etc.

TABLE 1. Summary table of important parameters in the fabrication of HEAs via. AM process.

The current section highlights emerging directions, including machine learning (ML)-assisted design and development of AMed HEAs, while discussing the associated challenges in integrating ML approaches for AM-HEA research. Finally, it concludes with remarks on the future prospects and research trajectory of AM-HEAs.

7.1 Emerging directions: machine learning-assisted alloy/process design

Designing complex HEAs with desirable properties by a “trial-and-error” method is a significant challenge to the scientific community due to the chemical complexity and phases. Additionally, optimising AM processing parameters for HEAs remains challenging. Machine learning (ML) offers an efficient approach to rapidly screen alloys, predict optimal processing conditions, and capture complex nonlinear relationships from large datasets. Compared with traditional empirical methods, ML enables faster, cost-effective, and data-driven prediction, reducing experimental efforts while improving process optimisation (Li et al., 2022). ML avoids the high computational costs and intensive calibration requirements as compared with physics-based simulations. The metallurgical complexity and rapid solidification phenomena inherent to AM create highly nonlinear process–structure–property relationships that traditional modelling approaches struggle to capture accurately. This motivates the adoption of ML techniques, and increasingly hybrid ML–physics frameworks, for predictive alloy design and process optimisation in AM (Koul, 2026).

The integration of ML with AM is expected to accelerate HEA development and promote industrial adoption. ML algorithms such as logistic regression, random forest, decision tree, K-nearest neighbour, support vector machine, and ANN have been widely explored for predicting phases in HEAs (Huang et al., 2019; Zhou et al., 2019). Wen et al. developed a materials design strategy integrating an ML-based surrogate model with experimental design algorithms to identify high-hardness HEAs within the AlCoCrCuFeNi alloy system (Wen et al., 2019). Bhandari et al. (2021) employed a random forest (RF) regression-based ML model to predict the high-temperature yield strength of MoNbTaTiW and HfMoNbTaTiZr HEAs at 800 °C and 1200 °C. The predicted values showed strong agreement with experimental results, demonstrating the model’s accuracy and reliability. Similarly, in another study, Islam et al. (2018) employed a neural network-based machine learning framework to identify data patterns using an experimental dataset and classify phase selection in MPEAs using experimental datasets. Apart from research studies, many well-known companies or organisations have started adopting ML techniques for alloy design in AM to develop new metallic alloys faster and improve material properties (Koul, 2026).

Even though there are many advantages of ML in AMed HEAs, there are also some challenges, such as selecting the best algorithms and descriptors, as there are too many combinations of models and material descriptors to choose from. Secondly, there is a complex process–structure–property relationship in metallic AM, with multiple scale interactions between alloy composition, processing parameters and microstructure. This complexity makes prediction and optimisation tricky using the traditional trial-and-error method. Inaccurate ML predictions can be due to the large number of data required, so it is important to choose data-efficient ML models that can make accurate predictions with a smaller number of training data. The ability of ML to tackle these challenges is crucial for the scalability, viability and application of ML models in the design and discovery of metallic alloys for AM applications.

7.2 Future trajectory of AMed-HEA research

• Additive manufacturing of HEAs requires further optimisation through microstructure tailoring, alloy composition adjustment, and post-processing to minimise defects, stabilise phases, and achieve superior mechanical performance.
• Current AMed-HEA research focuses on hardness, tensile, and compressive properties. However, fatigue, fracture toughness, creep, magnetic behaviour, oxidation, irradiation, hydrogen storage, wear, and corrosion resistance are underexplored.
• Metal additive manufacturing for HEAs has enormous industrial potential. Achieving widespread application necessitates sophisticated alloy design, a better knowledge of AM processes, the development of innovative and hybrid production techniques, and cost-effective scaling solutions. These developments will improve AM capabilities and enable large-scale production across a variety of industrial industries.
• Furthermore, most studies focus on CoCrFeMnNi HEAs; future research should explore lightweight and refractory HEAs, which remain challenging for 3D printing.

8 Conclusion

This section summarises the current state of knowledge on additive manufacturing of HEAs, identifies important unresolved scientific and technological difficulties, and outlines promising future research directions.

• Additive manufacturing has now emerged as a transformative processing route. It allows for complex shapes, customised microstructures and improved strength and ductility.
• Tailored microstructures, heterogeneous architectures and improved strength-ductility have been achieved using LPBF, SEBM, DED and AFSD methods for the fabrication of HEAs. Rapid solidification and repeated thermal cycling inherent to AM processes enable refined microstructures, suppression of elemental segregation, and the development of heterogeneous architectures that are difficult to achieve using conventional production routes.
• There are a number of scientific and technological challenges that remain unanswered. The reliability and reproducibility of AMed HEAs are still hindered by the appearance of defects, residual stresses, anisotropic microstructures, compositional variations due to elemental evaporation, and challenges in achieving phase stability. Moreover, the present research has focused predominantly on a few alloy systems, such as the CoCrFeMnNi group, and lightweight and refractory alloys are still relatively under-studied. Long-term performance characteristics such as fatigue resistance, fracture toughness, creep behaviour, oxidation resistance, corrosion resistance, irradiation tolerance, and hydrogen-related degradation also require systematic investigation.
• Future research needs include the development of advanced alloy design, microstructure tailoring, optimisation of the post-processing methods, development of lightweight alloys and refractory materials, fatigue, creep, fracture, corrosion, oxidation, irradiation, and wear, and the application of machine learning to expedite alloy and process designs. Furthermore, advanced post-processing methods, real-time process monitoring, and multi-scale characterisation methods will play key roles in defect reduction and reliability of the structure.
• There are additional opportunities to explore multi-material architectures, HEA composites, functionally graded materials and hybrid manufacturing processes that could enable new applications in the aerospace, energy, biomedical, nuclear and defence industries.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Kamran A Khan, Khalifa University, United Arab Emirates

Reviewed by: Zhiyong Ji, Beihang University, China

Yunjian Bai, Shandong University, China