The alloy belongs to the AlCoCrFeNi family, a class of materials loosely defined as containing five or more principal elements in near-equal proportions. Such compositions pair the strength of an ordered body-centered cubic B2 phase with the ductility of a face-centered cubic (FCC) phase, and they tolerate higher concentrations of aluminum and chromium than conventional structural alloys, offering greater latitude to form protective oxide scales. Above 800 °C, however, achieving oxidation resistance becomes substantially harder: alloy survival depends on maintaining a dense, slow-growing, adherent scale, and only a few oxides qualify. Chromia (Cr2O3) and alumina (Al2O3) are the principal protective scales for structural alloys, but chromia-forming alloys become limited near 1000 °C, where Cr2O3 converts to volatile CrO3 and the scale thins faster than it can reform.

The as-cast material used in the study was a tilt-cast ingot consisting of near-equal fractions of a Cr–Co–Fe-rich FCC phase and a Ni–Al–Ti-rich B2 phase arranged in a dual-phase dendritic microstructure, with a Ni–Zr intermetallic nucleating at phase boundaries. Annealing at 1100 °C for 50 hours followed by water quenching coarsened and broke down the lamellar interior structure, yielding a more spatially uniform distribution of the two phases without changing phase identity or fraction. The B2 + BCC designation refers to a single continuous body-centered cubic constituent: an ordered Ni–Al–Ti-rich B2 matrix containing a high number density of coherent, disordered Cr-rich BCC domains formed by spinodal decomposition. The two are treated as one constituent because they share a continuous lattice and are not separable at the length scale of the morphology examined.

Cyclic oxidation tests ran in static air with 5-hour heating cycles followed by 20 minutes of cooling. At 850 °C, a diffusion-limited regime, both conditions showed rapid initial mass gain over the first three cycles before transitioning to slower parabolic growth. After 50 hours the as-cast condition had gained 1.24 mg/cm2 versus 1.52 mg/cm2 for the annealed condition. Fitting the parabolic region gave rate constants of 27.0 × 10−13 g2 cm−4 s−1 for the as-cast material and 102.0 × 10−13 g2 cm−4 s−1 for the annealed material, a factor of 3.8 difference, with the two conditions indistinguishable during the initial transient. The annealed condition departed from parabolic behavior after 40 hours, recording no further net mass gain over the final two cycles.

At 1050 °C the picture changed to a spallation-dominated regime. The as-cast condition lost mass from the first cycle onward, while the annealed condition gained slightly before turning to continuous loss; cumulative changes after 50 hours were −51.26 and −55.46 mg/cm2 respectively, a marginal difference of roughly 8 percent. Here the rate is governed by the mechanical stability of the scale rather than transport through it. Cross-sectional imaging revealed subsurface voids extending 100–200 μm deep in the as-cast sample but 500–600 μm in the annealed sample, preferentially associated with FCC regions and consistent with a Kirkendall-type mechanism in which the outward flux of aluminum and titanium generates a compensating inward vacancy flux that coalesces where boundary sink capacity is insufficient.

X-ray diffraction and electron microscopy showed that all conditions developed the same three-layer architecture: an outer oxide, an inner oxide, and a metallic FCC transition layer at the oxide–metal interface. The transition layer is consumed B2+BCC, formed as aluminum and titanium are preferentially extracted to the oxidation front, shifting the local thermodynamic balance toward FCC stability. Spinel-structured and Cr2O3 oxides dominated all surfaces, with minor Al2O3 and monoclinic ZrO2. Notably, neither aluminum nor titanium reached the surface in sufficient quantity to form a continuous external Al2O3 scale. Aluminum partitioning into the B2 phase likely holds its effective flux below the Wagner threshold for the transition from internal to external oxidation, leaving protection to a Cr2O3 framework that codominant spinel renders discontinuous.

Titanium behaved as a temperature-gated participant. At 850 °C, TiO2 rutile peaks were prominent and titanium concentrated in the outer oxide, where it likely promotes densification and adhesion. At 1050 °C, rutile peaks were strongly suppressed, surface titanium dropped from 12–18 at% to under 1 at% in most phases, and titanium was consumed within the inner oxide, oxidized in place before it could migrate outward. The authors interpret this as a kinetic gate: the faster-advancing oxidation front at 1050 °C outpaces titanium redistribution. Chromium, by contrast, acts as a morphology-independent scaffold, present as bulk solute in both phases and released near the front by dissolution of spinodal Cr-rich domains, so its supply is effectively equivalent regardless of phase arrangement.

The mechanistic interpretation is that the pathway by which morphology acts switches with temperature. In the diffusion-limited regime at 850 °C, the fine lamellar B2+BCC network of the as-cast condition places aluminum and titanium reservoirs within short diffusion distances of any surface point, enabling quasi-continuous supply, while the coarsened, isolated B2+BCC pools of the annealed condition force episodic supply across longer diffusion distances, amplified by phase boundaries acting as short-circuit diffusion pathways orders of magnitude faster than bulk transport. In the spallation-dominated regime at 1050 °C, redistribution is outpaced by front advancement, so the local phase identity at each position when the front arrives determines the resulting oxide chemistry, a framework consistent with Gesmundo and Viani’s theory of selective oxidation in two-phase alloys.

As a third, processing-based route to varying morphology at fixed composition, the team applied spot friction processing (SFP), a solid-state technique using a pinless tungsten–rhenium tool at 1000 rpm for 5 seconds. SFP refined both phases to approximately 1.9 μm without phase transformation, preserving a measured phase balance of 0.495 FCC and 0.505 B2+BCC, and redistributed the phases uniformly through a process zone about 250–300 μm deep. In a 20-hour test at 1050 °C, limited because the processed region has finite depth, the SFP condition maintained net mass gain for the first 10 hours and accumulated a mass change of −9.96 mg/cm2 at 20 hours, versus −23.11 mg/cm2 for as-cast and −22.63 mg/cm2 for annealed, a reduction in mass loss of more than 50 percent. The refined sample showed no substrate voids, more laterally continuous oxide layers, and ZrO2 redistributed from continuous columns into dispersed grains. The authors note that only a single sample was tested per condition and that the contribution of preexisting α-Cr phases introduced by processing cannot be fully separated, but the results establish phase morphology itself as a lever for designing oxidation-resistant multiphase high-entropy alloys.

Subject of Research: How phase morphology controls high-temperature cyclic oxidation in a dual-phase AlCoCrFeNi(CuTiZr) high-entropy alloy

Article Title: Morphology-controlled oxidation in a dual-phase AlCoCrFeNi(CuTiZr) high-entropy alloy

Article References: Lastovich, M., Bamisaye, O. S., Tsai, F.-Y., Uddin, M. J., Maledi, N., Klenam, D., Perkins, C., Bodunrin, M., & Gwalani, B. (2026). Morphology-controlled oxidation in a dual-phase AlCoCrFeNi(CuTiZr) high-entropy alloy. Journal of Materials Science. https://doi.org/10.1007/s10853-026-13782-w

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13782-w

Keywords: high-entropy alloys, oxidation, phase morphology, AlCoCrFeNi, chromia scale, spallation, spot friction processing, high-temperature materials, Kirkendall voids, parabolic kinetics, B2 phase, FCC phase