Quantized aging mode in metallic glass-forming liquids

Physical aging exhibits complex off-equilibrium dynamics and its studies are of particular importance for understanding the nature of glassy state. We investigate the aging process of supercooled Cu₅₀Zr₅₀ glass-forming liquid numerically and theoretically. The relaxation time spectra based on molecular dynamics simulations show that the aging dynamics features the superposition of discrete homogeneous dynamic modes. The visiting probabilities of these modes decay toward slower relaxation regime, which is tuned by aging time from the compressed to the stretched exponential manner. Such multi-mode dynamics is found to be accompanied by the quantized hierarchy of mobile clusters with identical size increment. With this finding, we propose an aging model, which describes the aging dynamics as the aging-dependent weighted average over quantized dynamic modes. The two-time correlation functions associated with aging thus are separated into aging-time dependent factor and aging independent relaxation-time term. The model well predicts the heterogeneous dynamics in aging and equilibration of supercooled metallic liquids, and reveals the intrinsic aspects influencing aging process.