[P2-75] Why the structure-property relationship in metallic glasses should be established beyond short-range order: Insight from potential energy landscape
For crystals, structures provide all information needed for predicting material properties. However, what determines non-crystalline solids properties remains elusive for many years. Extensive work has been performed to identify structures playing an important role in glass, but a key question arises that what is the hidden rule of structural feature that can predict properties. Here we calculate an atom’s activation energy (the system’s long-time property) for thermally activated relaxation with the Activation-Relaxation Technique (ART) and correlate the searched local potential energy landscape with several of the successful structural predictors. We find a common nature in the successful structural predictors that spatial correlation of structural information matters a lot once they tend to determine an atom’s properties. There exists a critical correlation length of about sub-nanometer which is corresponding to the second shell of the pair correlation function of glassy structures. We further demonstrate this concept by manipulating the cutoff distance of local structural entropy - one of the successful structural feature - that only if this local structure is defined beyond the short-range order it can predict activation of local atom rearrangement in the model metallic glass. In this way, we question the prevailing approach of materials science aimed at identifying simple structural motifs responsible for metallic glass properties.