Room-temperature super-elongation in high-entropy alloy nanopillars

Nanoscale metallic materials often display very high strength but poor tensile ductility due to premature failure arising from limited dislocation sources and interactions, which leads to deformation localization and plastic instability. Prior approaches enabling large elongations in nanowires frequently involve surface diffusion or migration of a single/few twin boundaries, but these mechanisms typically operate at lower stresses, are sensitive to surface quality, or show sharp post-yield stress drops. This raises the open question: can large, uniform tensile ductility be achieved at room temperature under the high stress levels typical of metallic nanopillars/nanowires? The authors target this by studying single-crystalline CoCrFeNi high-entropy alloy (HEA) nanopillars, hypothesizing that intrinsic chemical heterogeneity and the resulting distribution of local stacking fault energies (SFEs) can promote spatially distributed activation of deformation twinning and dislocation slip, thereby suppressing strain localization and enabling super-elongation at high flow stress.

The paper situates its study within prior findings that nanoscale metals, despite high strength, often fail prematurely in tension with localized shear or brittle fracture. Large elongations in some noble metal nanowires have been achieved via twin boundary migration; however, these often show sharp stress drops after yield and low sustained flow stress. Surface diffusion–assisted deformation can also yield high stretchability in very small or irregularly shaped nanocrystals but is sensitive to surface conditions (oxidation, coatings) and geometry. The authors emphasize that such mechanisms differ from displacive plasticity and question their generality for well-defined nanopillars. HEAs, with multiple principal elements, introduce compositional fluctuations and packing variations that create a spectrum of local SFEs, fundamentally influencing deformation mechanisms. Recent studies also indicate that chemical short-range order (SRO) in HEAs can tune mechanical properties and stacking fault energies, motivating investigation of how atomic-level chemical heterogeneity affects twinning/slip and ductility in nanoscale HEAs.

• Materials and characterization: Bulk equiatomic CoCrFeNi HEA was prepared by arc melting (remelted 5 times), cast into a Cu mold, homogenized at 1473 K for 24 h, cold rolled (65% reduction), and annealed at 1373 K for 1 h to obtain recrystallized equiaxed grains (~120 µm). Single-phase FCC structure confirmed by high-energy synchrotron XRD; EBSD verified microstructure and orientations; EDS confirmed near-equiatomic uniform distribution.
• Specimen fabrication: Single-crystalline dog-bone nanopillars were fabricated via PFIB/Ga-FIB from grains oriented along <100>, <110>, or <111> for loading, with zone axis near <110> to observe defects. Typical gauge dimensions ~350 nm (length) × ~70 nm (width) × ~70–73 nm (thickness). Thickness homogeneity verified by EELS (~73 nm average).
• In situ TEM tensile testing: Conducted at room temperature in a JEOL 2100 TEM (200 kV) using Hysitron PI 95 picoindenter with a nano-gripper, displacement control at 1 nm/s equivalent to strain rate ~2.8×10^-3 s^-1 (gauge length 350 nm). Engineering stress/strain computed from SEM-measured cross-section and initial gauge length. Deformation recorded in video; SAED used to monitor orientation changes. Post-mortem atomic-resolution HAADF-STEM (Themis-Z, 300 kV) characterized twins, twin-thickening, and interactions.
• Atomistic simulations (LAMMPS): Large-scale MD simulations of single-crystalline CoCrFeNi <110> nanopillars with rectangular cross-sections and width-thickness ratio λ = 1–2; length ~90 nm, thickness ~15 nm, width varied (~15–30 nm), T = 300 K. Initial dislocations introduced in some samples by pre-strain (5% compression at 300 K) to create different initial dislocation densities ρ0. NPT relaxation followed by tensile loading at strain rate 5×10^8 s^-1 (NVT). Tracked stress-strain, twin fraction, dislocation density, von Mises atomic shear strain; CNA and DXA used for structure/defect analysis. Also simulated pure Al, Ni, Ag <110> nanopillars, and Al0.5CoCrFeNi and CoCrNi to compare mechanisms and ductility.
• Generalized planar fault energy (GPFE) calculations: Constructed 8064-atom HEA cells; generated 5000 independent random solid-solution configurations to compute GPFE curves by rigidly shifting {111} planes (intrinsic/extrinsic SF, twin nucleation/growth). Statistical distributions of intrinsic SFE and twin fault energy extracted; compared to pure Ni.
• Dislocation mobility simulations: Modeled a 1/2<110>{111} edge dislocation in HEA vs Ni in 20 nm cubic cells under constant shear stress or constant strain rate at 300 K to assess glide behavior, dissociation width variability, and resistance.
• Energetic analysis: Derived critical shear stress for surface nucleation of twinning partials using parameters including shear modulus, Burgers vector, loop radius (5–10 nm), Poisson’s ratio, correction factor for corner nucleation, and SFE (~30–34.8 mJ/m^2), then converted to applied normal stress via Schmid factor (0.471) for <110> tension.
• Orientation dependence: Also tested <100> and <111> oriented pillars experimentally for comparison.

• Exceptional ductility: <110>-oriented CoCrFeNi HEA nanopillars exhibited uniform room-temperature tensile super-elongation up to ~110% at high flow stress of 0.6–1.0 GPa. <100> and <111> orientations showed limited plasticity due to shear localization from dislocation glide and deformation twinning, respectively.
• Mechanism: Ultralarge uniform ductility stems from spatially distributed and synergistic activation of deformation twinning and dislocation slip. Deformation proceeds via nucleation, propagation, interaction, thickening, and coalescence of numerous nanotwins, with extensive twin–dislocation interactions that delocalize strain and delay necking.
• Microstructural evolution: In situ TEM showed initial twin nucleation on {111} from free surfaces, followed by activation of a second twinning system, twin–twin interactions leading to detwinning of early twins (Twin 1) and dominance and thickening of later twins (Twin 2) via partial glide along twin boundaries. Eventually, the entire pillar became fully twinned with a reorientation from <110> to <100> before failure. Typical nanotwin thickness ~1 nm; fully twinned areas confirmed by SAED/HAADF.
• MD corroboration: Simulations reproduced experimental behavior across λ and initial ρ0, showing twin fractions increasing to ~90% with strain, with dislocations maintaining plasticity as twins saturate. Strain maps showed no strong shear localization during uniform flow. Twin coalescence accelerates twinned area growth; notch-like flaws from twin–twin interactions did not trigger localization due to continued twin extension.
• Statistical energetics: GPFE calculations for 5000 random HEA configurations revealed a wide spread of energy pathways; intrinsic SFE and twin energies show substantial scatter (HEA SFE ~30–34.8 mJ/m^2 used in estimates). This variability provides a broad distribution of local stress thresholds for ongoing partial nucleation and twin propagation, enabling sustained, distributed plasticity.
• Critical stress consistency: Estimated critical twinning partial nucleation stress τcrit ≈ 628–903 MPa (loop radius 5–10 nm, SFE ~30–34.8 mJ/m^2). With a Schmid factor of 0.471 for <110> tension, the corresponding applied normal stress is ~1.33–1.92 GPa, consistent with experimental yield strengths; initial dislocations and TB–surface junctions can lower barriers locally.
• Dislocation behavior: HEA dislocations display rugged lines and variable dissociation widths with higher barriers for equivalent velocity than Ni, reflecting severe lattice distortion and local heterogeneities, contributing to higher flow stresses.
• Comparisons: Pure Al and Ni <110> nanopillars show ~30% elongation; Ag ~60%, all with localization near twin–twin interactions, underscoring the unique distributed plasticity in HEAs. Similar super-elongation mechanisms were observed in Al0.5CoCrFeNi and CoCrNi nanopillars, indicating generality to multi-principal element alloys with low/variable SFEs.

The study addresses the long-standing challenge of achieving large, uniform tensile ductility in nanoscale metals at high stress. In contrast to surface diffusion–mediated stretchability or single-TB migration in noble metals, the HEA nanopillars achieve super-elongation via robust, spatially distributed displacive mechanisms. Atomic-level chemical heterogeneity in HEAs generates a broad distribution of local SFEs and thus a spectrum of stress thresholds for partial dislocation nucleation and twin growth. This promotes persistent, staggered activation of multiple twinning/slip events throughout the volume that continuously redistribute strain, suppress instability, and maintain high flow stress. The sequence of twinning (nucleation, interaction, detwinning of early twins, thickening/coalescence of later twins) and the eventual reorientation to <100> underpin sustained uniform plastic flow until failure. Simulations corroborate these observations and show the generality across geometry and initial dislocation content, as well as across different multi-principal alloys. Orientation dependence is pronounced: <110> enables favorable Schmid factors for twinning and coordinated slip, while <100> and <111> exhibit early localization. The findings thus provide a mechanistic basis for designing reliable nanoscale devices by leveraging compositional heterogeneity to tailor local SFEs and activate synergy between twinning and slip.

The work demonstrates room-temperature uniform super-elongation exceeding 100% at high flow stress (0.6–1.0 GPa) in single-crystalline <110> CoCrFeNi HEA nanopillars. In situ TEM, HAADF-STEM, MD simulations, and energetic analyses reveal that ultrahigh ductility arises from spatially and synergistically coordinated deformation twinning and dislocation slip driven by atomic-level chemical heterogeneity and a broad distribution of local SFEs. Extensive nanotwin nucleation, interaction, thickening, and coalescence, coupled with twin–dislocation interactions, delocalize strain and delay necking; the lattice reorients from <110> to <100> via pervasive twinning before failure. The mechanisms are general to multi-principal element alloys with low/variable SFEs. Future work should systematically quantify the effects of initial dislocation density, width–thickness ratio, sample shape/morphology, and atomic ordering (including SRO) on twin/dislocation evolution, and explore compositional tuning to optimize the balance of strength and ductility for nanoscale device reliability.

• Orientation sensitivity: Super-elongation was achieved in <110>-oriented pillars; <100> and <111> showed limited ductility due to early localization, limiting generality across orientations.
• Scale and geometry: Findings pertain to nanoscale pillars with specific gauge dimensions; size effects and geometry (e.g., aspect ratio, cross-section) may influence mechanisms and were only partially explored (λ = 1–2 in MD).
• FIB-induced defects: Despite cleaning, FIB preparation introduces defects that can affect initial dislocation content and nucleation sites; while MD examined different initial dislocation densities, experimental quantification is limited.
• Simulation conditions: MD strain rates (5×10^8 s^-1) are much higher than experimental rates; although mechanisms matched, rate effects could influence quantitative thresholds.
• Material scope: While similar behavior was observed in Al0.5CoCrFeNi and CoCrNi, broader compositional spaces and environmental effects (e.g., oxidation, surface conditions) were not comprehensively assessed.