The unique combination of metallic bonds and disordered atomic structure makes the metallic glasses promising functional materials as well as important candidate for scientific studies. However, the discovery of new glass-forming compounds is limited by the low-efficiency trial and error protocol due to the lack of knowledge on phase-selectivity between glass and crystal state. A primary obstacle in understanding the phase-selectivity is the description of disordered structure in metallic glasses (MG). In this thesis, we first propose a new scheme for systematically estimating the short-range order (SRO), medium-range order (MRO), and chemical orders in MGs. Using a model binary Lennard-Jones (LJ) system, we show that the five-fold symmetry can emerge in the medium-range level. Then, the role of MRO in glass stability is examined by comparing two glass-forming LJ systems with similar SRO, but very different MRO and glass formability (GFA), providing new insights into the phase-selectivity. We find that the MRO with rich pentagons not only induce the geometrical frustration, but also controls the crystallization paths by suppressing the local composition variations. Finally, the candidate ‘defect’ in MGs, termed soft spots are studied because of their central role in strain-triggered phase selection. Through the vibrational density of states (VDOS) obtained directly from molecular dynamics (MD) simulations, we propose a new softness parameter that can be explicitly related to the real space motions of atoms. The physical nature of atomic softness is clarified as the ability to rearrange. The soft particles keep rearranging rapidly, and the high rearranging rate give rise to the rich low-frequency VDOS. At the same time, the hard particles do not rearrange and do not contribute to the low-frequency intensity. Structural analysis shows that the soft particles are essentially in disordered local structures, where a representative packing motif cannot be identified. On the other hand, the hard particles have well-defined local order, even medium-range order.
| Date of Award | 2021 |
|---|
| Original language | English |
|---|
| Awarding Institution | - The Hong Kong University of Science and Technology
|
|---|
| Supervisor | Michael Scott ALTMAN (Supervisor) & Li Huang (Supervisor) |
|---|
Computational study of phase selectivity in supercooled metallic liquids
FAN, X. (Author). 2021
Student thesis: Doctoral thesis