Degree of short-to-medium-range order in metallic glasses

非技术摘要使用的大多数金属和合金都是晶体，其内部结构具有长程有序的特征。在过去的一个世纪里，人们对它们的结构进行了深入的研究，为它们的性质提供了具体的基础。这种结构-性质范式在材料科学领域得到了很好的确立。相比之下，通过快速冷却母体液体获得的金属玻璃(MGs)具有类似液体的非晶态结构，内部没有可识别的微观结构。虽然缺乏长程周期性，但MGs确实有结构，在最近邻原子配位的水平上，以及在几个纳米尺度的长度尺度上的位置关联。这种结构性组织很难描述，也仍然知之甚少。因此，对镁的原子结构和性质之间的联系的理解远远落后于对传统金属的理解。本项目旨在利用计算机模拟的方法来解决MGS的中短程排序问题。详细的结构信息有望揭示这些玻璃的宏观行为，从而在设计它们的应用时能够更好地做出判断。在教育方面，本研究对丰富材料科学重点学科，特别是材料结构这门核心学科具有重要意义。许多计算机分子动力学模拟适合本科生参加高级设计/研究体验课程，提供了有趣的案例研究，以与以前在课堂上学到的熟悉的晶体对称性进行比较。本项目的目标是1)量化金属玻璃(MGs)中短程有序(SRO)的实现程度和程度，以及2)可靠地识别不同类型MGs中的中程有序(MRO)特征模式。这种对非晶态固体有序性的基本理解是一个基本的材料科学问题，但到目前为止还很难实现，因为到目前为止，无论是通过对散射数据的拟合，还是通过分子动力学(MD)模拟中的液体猝灭，所构建的MG构型本质上远低于实验玻璃的有序性。该项目旨在开发高效的分子动力学模拟路线，能够产生具有与真实世界分子格子相似的势能、原子密度、弹性常数以及有序度的分子格子结构。一个焦点是算法开发，以加速MD在搜索更深能量盆地的过程中。该计划将使用1)延长老化，2)循环应力辅助老化，以及3)新的删除-添加协议，建立在模拟晶体崩溃或气相沉积获得的非晶态结构的基础上。初步结果表明，这些方法(以及它们的组合)可以在最终结晶的过程中获得低能量的玻璃态。捕捉到的构型提供了一个前所未有的机会，可以在类似于实验室玻璃的组成和饱和度水平上揭示SRO/MRO，从而允许与剪切转变、松弛和其他玻璃特性背后的原子机制联系在一起。预计这一结果将是朝着在MG中建立具体结构-财产关系迈出的重要一步。

Nontechnical abstractMost metals and alloys in use are crystals, with an internal structure characterized by long-range order. Their structures have been well studied over the past century, providing concrete basis of their properties. Such a structure-property paradigm is well established in the field of materials science. In contrast, metallic glasses (MGs) obtained by fast cooling of their parent liquids have a liquid-like amorphous structure with no discernable microstructure inside. Though lacking long-range periodicity, MGs do have structure, at the level of nearest-neighbor atomic coordination and positional correlations on the length scale of a couple of nanometers. This structural organization is difficult to characterize and remains poorly understood. As a result, the understanding of the links between the atomic structure and properties in MGs lags far behind that achieved for conventional metals. This project aims to resolve the short-to-medium range order in MGs, using computer simulation approaches. The detailed structural information is expected to shed light on the macroscopic behavior of these glasses, enabling better judgement in designing their applications. In terms of education, this research has significance in enriching key materials science subjects, in particular the core subject of Structure of Materials. Many of the computer molecular dynamics simulations are amenable to the participation of undergraduate students in Senior Design/Research Experience courses, providing interesting case studies to compare with the familiar crystal symmetry learnt previously in the classroom. Technical abstractThe goal of this project is to 1) quantify the degree and extent of short-range order (SRO) realizable in metallic glasses (MGs), and 2) reliably identify the characteristic patterns of medium-range order (MRO) in different types of MGs. This basic understanding of the ordering in amorphous solids is a fundamental materials science issue, but has been difficult to achieve so far because the MG configurations constructed to date, either by fitting to scattering data or by liquid quenching in molecular dynamics (MD) simulations, are inherently far less ordered than experimental glasses. The project aims to develop efficient MD simulation routes that are capable of producing MG structures having potential energy, atomic density, elastic constants, and therefore degree of order all similar to real-world MGs. A focus is algorithm development to accelerate MD in the search of deeper energy basins. The plan is to use 1) prolonged aging, 2) cyclic-stress-assisted aging, and 3) a new deletion-addition protocol, building upon amorphous structures obtained from simulated collapse of crystals or vapor deposition. Preliminary results suggest that these approaches (and their combination) can access low-energy glassy states along the way towards eventual crystallization. The configurations captured provide an unprecedented opportunity to uncover the SRO/MRO at a level of make-up and saturation that resembles those in laboratory glasses, to allow connections with atomistic mechanisms underlying shear transformations, relaxation and other glass properties. The results are expected to be a major step-forward towards establishing concrete structure-property relations in MGs.