Swapping One Element Decides Whether Mixed High-Entropy Alloys Turn Glassy

When two chemically distinct high-entropy alloys are forced to mix under extreme shear, whether the interface turns into a glass is decided not by the violence of the deformation but by the identity of a single element. In a study published open access in the Journal of Materials Science, Shabnam Taheriniya, Harald Rösner, Sergiy V. Divinski and Gerhard Wilde co-deformed the face-centered cubic CoCrFeMnNi Cantor alloy together with two body-centered cubic refractory Senkov alloys using high-pressure torsion. When the refractory partner contained tantalum, HfNbTaTiZr, the forced intermixing produced an amorphous interfacial phase containing all ten constituent elements. When tantalum was replaced by vanadium, HfNbTiVZr, the same processing route instead yielded a heavily defected but fully nanocrystalline intermixed region. The comparison isolates chemistry, rather than deformation alone, as the decisive factor behind interfacial amorphization in compositionally complex metal composites.

High-entropy alloys typically contain five or more principal elements in a single solid solution, and their behavior is shaped by high configurational entropy, severe lattice distortion, sluggish diffusion in some cases, and what the field calls the cocktail effect: nonlinear or synergistic responses arising from the collective interactions of many elements, which cannot be predicted from the properties of the individual constituents. The authors argue that such emergent responses should be favored in environments where competing interactions, local chemical heterogeneity, dense interfaces, defect structures and non-equilibrium states coexist. Severe plastic deformation provides a way to engineer exactly those conditions deliberately. Among severe deformation techniques, high-pressure torsion is particularly powerful because it combines very high hydrostatic pressure with extreme shear strain, granting access to microstructures and structural states that conventional processing cannot reach. The method has previously been used to create ultrafine and nanocrystalline structures in high-entropy alloys, to drive non-equilibrium intermixing in aluminum-nickel multilayers, to mechanically alloy the otherwise immiscible aluminum-zinc system, and even to synthesize bulk Cantor alloy directly from its elemental constituents.

The experimental design rested on a controlled one-element substitution. Disks of 7.5 millimeter diameter and 0.5 millimeter thickness were cut from homogenized rods, polished to a mirror finish, etched in a 1:3 mixture of concentrated nitric and hydrochloric acid, and stacked in a quasi-constrained high-pressure torsion device with sandblasted anvils to prevent slippage. Mechanical bonding between the Cantor and Senkov disks was achieved by 20 revolutions at a hydrostatic pressure of 9 gigapascals at ambient temperature, with the upper anvil rotating at 1 revolution per minute. One nanocomposite paired Cantor with the tantalum-containing Senkov alloy; the other paired the same Cantor alloy with the vanadium-containing variant. For reference, single-phase disks of Cantor, Senkov-Ta and Senkov-V were processed under identical conditions. All material extraction came from the disk edge, where the imposed strain is highest and the microstructure most homogeneous, and transmission electron microscopy lamellae and atom probe needles were taken from the same intermixed edge region, only a few micrometers apart, to allow direct correlation between techniques.

The mechanical characterization of the single-phase alloys already revealed a clear split. The as-received vanadium-containing Senkov alloy showed the highest hardness among the starting materials, a consequence of its finer grain structure of roughly 250 micrometers and stronger solid-solution strengthening; the atomic-size mismatch parameter rises from about 3.8 percent in HfNbTaTiZr to 5.9 percent in HfNbTiVZr. After high-pressure torsion, every alloy hardened substantially through grain refinement and strain hardening, and the vanadium alloy remained the hardest single-phase material. In tensile testing, however, the HPT-processed vanadium alloy stood out for a different reason: it reached a maximum tensile stress of approximately 1.94 gigapascals and an energy to fracture of 110.7 megajoules per cubic meter while retaining a fracture strain of about 11.1 percent. The processed Cantor and tantalum-containing Senkov alloys reached maximum tensile stresses of roughly 1.27 and 1.03 gigapascals respectively, with comparable fracture strains near 11.5 and 12.0 percent. The vanadium alloy’s advantage was therefore not greater ductility but its ability to sustain far higher stresses at similar deformability.

Co-deformation then produced two mechanically divergent composites. Optical and electron microscopy showed that mixing between Cantor and the tantalum-containing Senkov alloy spread extensively, reaching toward the disk center, whereas in the vanadium-containing composite intermixing stayed confined to the disk edge and did not expand even when the number of revolutions was increased from 10 to 20. The authors attribute this to co-deformation compatibility: the softer, more finely refined tantalum alloy deforms readily with the ductile Cantor constituent, promoting lamellar refinement and repeated mixing, while the harder vanadium alloy resists it. Both nanocomposites were harder than their processed constituents, but the rankings reversed between indentation and tension. The tantalum-containing composite showed the higher Vickers and local nanoindentation hardness, yet the vanadium-containing composite sustained a maximum tensile stress of about 1.60 gigapascals and absorbed 74.6 megajoules per cubic meter before fracture, against roughly 1.03 gigapascals and 39.6 megajoules per cubic meter for the tantalum composite, with fracture strains of about 9.4 and 7.4 percent respectively. Nanoindentation mapping of the vanadium composite resolved three statistically distinct hardness populations, consistent with its nanocrystalline lamellar intermixed structure.

The structural origin of this divergence emerged from correlative scanning transmission electron microscopy, energy-dispersive X-ray spectroscopy, four-dimensional scanning transmission electron microscopy nanobeam diffraction and atom probe tomography. Atomic-scale intermixing occurred in both systems, with intermixed compositions involving more than five elements drawn from both parent alloys, so the different outcomes cannot be ascribed to the presence or absence of mechanical mixing itself. Quantitative comparison of the ten-element interfaces showed nearly identical atomic-size mismatch, 9.75 percent for the tantalum system versus 9.71 percent for the vanadium system, and only modestly different global pair-interaction variance, 6.08 versus 4.23 kilojoules per mole. What differed sharply were the local interactions: the tantalum-Cantor mixing enthalpy of minus 16.4 kilojoules per mole with a heterogeneity parameter of 3.72 kilojoules per mole far exceeded the vanadium-Cantor values of minus 8.2 and 1.72 kilojoules per mole, yielding a local segregation parameter of 61.01 for tantalum against 14.11 for vanadium. In particular, the strongly negative tantalum-nickel interaction of about minus 22 kilojoules per mole drives nickel partitioning toward tantalum-rich regions, and this chemically driven redistribution competes with lattice reconstruction, increasing local frustration and favoring loss of long-range crystalline order.

The amorphous interfacial phase that forms only in the tantalum-bearing composite proved remarkably durable. Correlative differential scanning calorimetry and in situ heating in the transmission electron microscope, both run at 20 kelvin per minute up to 680 degrees Celsius, revealed three overlapping thermal events: recovery and dislocation rearrangement below about 200 degrees Celsius, recovery and grain growth near 350 degrees Celsius, and deformation twinning with the onset of phase separation at higher temperatures. Because the composite thermogram superposes signals from three microstructurally distinct regions, the authors deconvoluted it mathematically and experimentally; the total enthalpy of the bulk composite signal was approximately 2.03 kilojoules per mole, and the overlapping character of the events ruled out a conventional Kissinger analysis. After isothermal heating at 680 degrees Celsius for 15 minutes in the microscope, and after bulk annealing at the same temperature for 20 minutes, the fcc, bcc and amorphous phases persisted as nanoscale laminates even as a chromium-rich sigma phase and a nickel-rich fcc phase developed. The authors note that thin-film effects, such as manganese evaporation, altered concentrations but not the final phase constitution, which matched the bulk-annealed material. Notably, intermixed layers thinner than about 100 nanometers remained free of decomposition products, while decomposition appeared only in Senkov-rich regions exceeding that thickness.

Fluctuation electron microscopy based on variable-resolution nanobeam diffraction quantified how the amorphous state matured. In Stratton-Voyles plots of the normalized variance peak height against the inverse square of the probe size, the as-deformed amorphous phase followed the trend of the phase-change material Ge2Sb2Te5, whereas the annealed state aligned with conventional metallic glasses such as a zirconium-based bulk glass, exhibiting a medium-range order correlation length of approximately 1.3 nanometers. The authors interpret this as partial relaxation of the heavily damaged amorphous-crystalline mixture produced during deformation, with the ten-element chemistry itself stabilizing the glass: multiple competing crystallization pathways, configurational disorder and sluggish diffusion hinder nucleation of any single crystalline phase, and a ten-component crystalline single-phase solid solution is thermodynamically unlikely. At higher temperatures, differential thermal analysis to 1100 degrees Celsius showed eventual complete crystallization into several crystalline phases each containing more than five principal elements, moving phase evolution away from the original alloy compositions entirely.

Functional behavior emerged alongside structure. Severe plastic deformation alone transforms the nearly paramagnetic Cantor alloy ferromagnetic at room temperature, associated with deformation-induced chemical fluctuations and nickel-rich nanoscale regions, and differential phase contrast imaging shows randomly oriented, spatially isolated magnetic regions. Co-deformation with the tantalum Senkov alloy produced substantially stronger and more continuous ferromagnetic contrast, together with antiferromagnetic regions, reflecting the broader distribution of local chemical environments. Annealing then drove cocktail-effect-driven phase separation into iron-manganese-chromium-rich and cobalt-nickel-rich domains, increasing magnetic contrast, enhancing coercivity and strengthening magnetic ordering at cryogenic temperatures down to 10 kelvin, although the lamellar alignment of ferromagnetic nickel-cobalt-rich and antiferromagnetic chromium-rich domains lowered the peak magnetization relative to the processed Cantor alloy. The authors present this as a hierarchical magnetic-hardening mechanism in which deformation, phase separation and cooling each contribute. Overall, the work demonstrates that elemental selection at heterophase interfaces governs whether forced mixing yields glass, nanocrystal, or new magnetic functionality, providing a physical basis for engineering emergent behavior in chemically complex alloys.

Subject of Research: Element-specific amorphization and emergent magnetic behavior in high-pressure torsion mixed high-entropy alloy nanocomposites

Article Title: Impact of crystal lattice and alloy composition on formation and stability of amorphous high entropy metal composites under forced mechanical mixing

Article References: Taheriniya, S., Rösner, H., Divinski, S. V., & Wilde, G. (2026). Impact of crystal lattice and alloy composition on formation and stability of amorphous high entropy metal composites under forced mechanical mixing. Journal of Materials Science. https://doi.org/10.1007/s10853-026-13935-x

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13935-x

Keywords: high-entropy alloys, high-pressure torsion, severe plastic deformation, amorphization, Cantor alloy, Senkov alloy, cocktail effect, nanocomposites, thermal stability, ferromagnetism, atom probe tomography, metallic glass