Cryogenic Cycling Rejuvenates Zr₅₀Cu₄₀Al₁₀ Bulk Metallic Glass: The Role of Casting Temperature on Microstructure and Properties

Abstract

This study investigates how cryogenic cycling treatment (CCT) rejuvenates a Zr₅₀Cu₄₀Al₁₀ bulk metallic glass (BMG). When cast at a high temperature, the alloy forms a homogeneous amorphous matrix that does not generate internal stress during CCT, preventing rejuvenation. Conversely, casting at a lower temperature introduces nano‑scale heterogeneities that induce internal stress during thermal cycling, leading to increased free volume and enhanced plasticity. These findings demonstrate that synthesis parameters can be tuned to control microstructural heterogeneity and, consequently, the rejuvenation response of metallic glasses.

Background

Bulk metallic glasses (BMGs) are prized for their exceptional mechanical properties—high fracture strength and a large elastic limit—thanks to their long‑range disordered microstructures. Rapid quenching is essential to suppress crystallization during solidification, yielding a non‑equilibrium state with higher configurational energy than crystalline counterparts. Upon annealing, BMGs tend to relax toward a lower energy state, often resulting in embrittlement. However, BMGs can also be driven to a higher energy state through rejuvenation techniques such as severe plastic deformation or recovery annealing.

Recently, deep cryogenic cycling treatment (DCT) has emerged as a novel method to rejuvenate BMGs. By cyclically cooling and heating between room temperature and cryogenic temperatures (~77 K), internal stresses arise in heterogeneous amorphous structures, driving atomic rearrangements that increase free volume. In this work, we examine the DCT response of Zr₅₀Cu₄₀Al₁₀ (at.%) BMGs fabricated at two distinct casting temperatures, evaluating the resulting microstructures, thermal behavior, and mechanical performance.

Methods

Sample Preparation

High‑purity Cu, Zr, and Al (≥99.9 %) were arc‑melted in a Ti‑gettered Ar atmosphere to form a master alloy. The alloy was cast into a copper mold to produce 2‑mm‑diameter rod samples. Two casting conditions were used: a high temperature (HT) condition at 9 A and a low temperature (LT) condition at 7 A. DCT was performed using a custom instrument that cycles samples between 293 K and 113 K for 30 cycles (DCT30).

Characterization

Phase identification was carried out by X‑ray diffraction (XRD, Cu Kα). Microstructure was examined with transmission electron microscopy (TEM, JEOL JEM‑2100F, 200 kV). Differential scanning calorimetry (DSC, heating rate 20 K min⁻¹) provided glass transition (T_g) and crystallization (T_x) temperatures, as well as relaxation enthalpy (ΔH_relax). Density was measured by Ar gas pycnometry. Compression tests (strain rate 5 × 10⁻⁴ s⁻¹) assessed mechanical properties. All measurements were repeated on at least four specimens to ensure reproducibility.

Results and Discussion

HT Samples

Figure 1a shows XRD patterns for HT As‑cast and DCT30 samples, both displaying only the broad amorphous halo—no crystalline peaks were observed. DSC thermograms (Figure 1b) revealed T_g of 690 K and T_x of 780 K for As‑cast, and 688 K / 781 K for DCT30, indicating negligible change after CCT. Isothermal annealing at 740 K (1.07 T_g) yielded identical crystallization incubation times (t_x ≈ 12.5 min) for both states, further confirming the absence of rejuvenation.

Relaxation enthalpy was calculated from the specific‑heat difference (ΔC_p) using

ΔHrelax = ∫RTTΔCp dT

yielding ΔH_relax ≈ 12.6 J g⁻¹ for As‑cast and 12.9 J g⁻¹ for DCT30—essentially identical, corroborating the lack of rejuvenation. TEM images (Figure 2a,b) further confirmed a homogeneous amorphous structure. Compression tests (Figure 2c) showed fracture strengths of ~2000 MPa and plastic strains of 0.3 % for both samples, with no improvement after CCT.

LT Samples

LT samples displayed a similar broad amorphous XRD halo (Figure 3a). However, DSC curves (Figure 3b) revealed a longer crystallization incubation time for DCT30 than for As‑cast, and a higher relaxation enthalpy (ΔH_relax), indicating successful rejuvenation. Density measurements (Table 1) showed a significant reduction from 6.957 g cm⁻³ (As‑cast) to 6.931 g cm⁻³ (DCT30). Using the free‑volume relation

x = [2(ρc − ρ)]/ρ

the free‑volume fraction increased from 1.30 % (As‑cast) to 2.06 % (DCT30). TEM images (Figure 4b,c) revealed nano‑scale clusters embedded in the amorphous matrix, likely B2‑CuZr phases, which are responsible for generating internal stresses during CCT. Compression tests (Figure 4a) demonstrated a substantial increase in plastic strain—from 2.8 % to 4.3 %—and a modest rise in fracture strength (~2050 MPa) after CCT, confirming the mechanical benefits of rejuvenation.

Mechanism of Rejuvenation

The presence of nano‑scale heterogeneities in LT samples introduces internal stresses during thermal cycling, described by

σα = Δα ΔT [2EcEa]/[(1+νa)Ec + 2(1 − 2νc)Ea]

Using literature values (Δα ≈ 1.3 × 10⁻⁵ K⁻¹, ΔT ≈ 180 K, E_c ≈ 77 GPa, E_a ≈ 123 GPa, ν_c ≈ 0.385, ν_a ≈ 0.383) yields σ_α ≈ 34 MPa, sufficient to drive local atomic rearrangements and increase free volume. In contrast, the homogeneous HT structure cannot generate such stresses, preventing rejuvenation.

Conclusions

High‑temperature casting of Zr₅₀Cu₄₀Al₁₀ produces a fully homogeneous amorphous matrix that does not respond to cryogenic cycling. Low‑temperature casting introduces nano‑scale heterogeneities that, upon DCT, generate internal stresses leading to increased free volume, improved plasticity, and overall rejuvenation. These findings offer a straightforward route to tailor BMG microstructure—and consequently, its thermal and mechanical behavior—by adjusting casting temperature.

Abbreviations

BMG

Bulk Metallic Glass

DCT

Deep Cryogenic Cycling Treatment

DCT30

Thermal treatment with 30 cycles

DSC

Differential Scanning Calorimetry

GFA

Glass Forming Ability

HT

High Casting Temperature

LT

Low Casting Temperature

TEM

Transmission Electron Microscopy

XRD

X‑Ray Diffraction