Linking the pressure dependence of the structure and thermal stability to α- and {eta}-relaxations in metallic glasses

Glasses derive their functional properties from complex relaxation dynamics that remain enigmatic under extreme conditions. While the temperature dependence of these relaxation processes is well-established, their behavior under high-pressure conditions remains poorly understood due to significant experimental difficulties. In this study, we employ cutting-edge experimental techniques to probe the pressure evolution of the relaxation spectrum in a Zr46.8Ti8.2Cu7.5Ni10Be27.5 metallic glass across gigapascal pressure ranges. Our findings reveal two distinct relaxation mechanisms under high pressure: In the \b{eta}-relaxation regime, compression drives the system with reduced atomic mobility and enhanced structural disorder, without significant density changes. Conversely, α-relaxation under pressure promotes density-driven structural ordering that improves thermal stability. Notably, the transition between these regimes occurs at a constant T/Tg,P ratio, independent of applied pressure. These results provide crucial insights for decoupling the competing structural and relaxation contributions to glass stability, establishing a systematic framework for tailoring glass properties through controlled thermo-mechanical processing.

💡 Research Summary

In this work the authors investigate how hydrostatic pressure combined with temperature controls the structural state and thermal stability of a Zr‑based metallic glass (Vit4, composition Zr46.8Ti8.2Cu7.5Ni10Be27.5). Samples were pre‑compressed at 7 GPa over a wide temperature range (298 K to 693 K), covering cold compression in the glassy state, hot compression in the supercooled liquid, and intermediate conditions. After decompression to ambient pressure, each glass was characterized by single‑shot flash differential scanning calorimetry (FDSC) at a heating rate of 200 K s⁻¹ and by synchrotron X‑ray diffraction.

The calorimetric data reveal two distinct regimes. When the compression temperature exceeds the glass transition temperature at the applied pressure (Tcomp ≥ 643 K), the glass transition onset (Tg,onset) shifts upward by about 6 K and the fictive temperature (Tf) drops, indicating a more stable, higher‑density glass. This regime corresponds to the activation of the primary (α) relaxation under pressure, which promotes cooperative atomic rearrangements, densification, and structural ordering. Conversely, for Tcomp below Tg,P, a sub‑Tg endothermic peak appears, the so‑called “shadow glass transition,” which is linked to the secondary (β) relaxation. In this low‑temperature regime the activation energy derived from Kissinger analysis is close to 28 RTg, characteristic of β‑relaxation, and the glass shows a modest pressure‑induced rejuvenation (Tf increases).

Kissinger plots obtained from FDSC scans at several heating rates provide activation energies (Ea) for the dominant relaxation process. At ambient pressure the β‑relaxation Ea ≈ 28 RTg and the α‑relaxation Ea ≈ 42 RTg, in agreement with literature. Under 7 GPa the crossover between these two values shifts by roughly 30 K toward higher temperatures, meaning that at a given temperature the α‑process is suppressed while the β‑process dominates. For samples quenched from the supercooled liquid at 7 GPa, Ea for the α‑relaxation is about 4 % lower than for the 1 atm counterpart, reflecting a pressure‑induced reduction of the liquid’s fragility and a transition toward a stronger (more Arrhenius‑like) supercooled liquid. This is corroborated by a broader overshoot in the DSC curves, a hallmark of a wider glass‑transition interval.

X‑ray diffraction shows that, despite the high pressures, the overall density change is modest (≤ 2 %). However, the structural signature differs between the two regimes: β‑relaxation under pressure leads to increased structural disorder without a significant volume change, whereas α‑relaxation produces a more densely packed random network, akin to the high‑density amorphous states observed in covalent glasses.

A key finding is that the transition between the β‑dominated and α‑dominated regimes occurs at a constant reduced temperature T/Tg(P), independent of the absolute pressure. This scaling implies that the pressure effect can be captured by a simple shift of the glass‑transition temperature (dTg/dP ≈ 38 K for 7 GPa), while the underlying relaxation dynamics retain their temperature dependence.

Overall, the study provides a comprehensive picture of how hydrostatic pressure can be used to decouple and independently tune the secondary and primary relaxation channels in metallic glasses. By selecting appropriate pressure–temperature paths, one can engineer glasses with tailored thermal stability, density, and mechanical performance, opening new avenues for the design of high‑performance amorphous alloys.