First-Principles Molecular Dynamics Simulations of Silicate Liquids: Structure, Diffusion and Viscosity at Mantle Conditions

硅酸盐液体是地球化学和热演化的主要媒介，因为它们形成了早期的岩浆海洋，并在地球表面以岩浆的形式出现，在地壳和地幔中以部分熔融的形式出现。岩浆作用负责大洋和大陆地壳的起源和持续形成，并以捕虏体的形式将我们关于内陆成分的主要线索之一带到地表。由于硅酸盐液体与其源区在密度、化学扩散率、粘度和体积组成上的差异，岩浆的生成和输送是导致质量和热量输送的最有效的地质过程之一。硅酸盐液体可能在地球早期历史中扮演了更重要的角色，当时熔融可能更广泛，可能已经延伸到更深的地方。因此，为了更好地理解行星的演化，我们需要在地幔压力-温度条件下的相关熔体性质的知识方面取得重大进展。在这个项目中，建议应用第一性原理计算和可视化技术相结合的方法来研究广泛的压力、温度和组成范围内硅酸盐熔体的结构、扩散和粘度。这种方法本质上是无参数的，有望为实验方法提供理想的补充，并可以为结构和成键的基本物理性质和行为提供重要的见解。建议的具体活动包括：1)将成分范围扩大到不含/不含挥发物(H2O和CO2)的天然熔体(MgO-CaO-Na2O-K2O-Al_2O_3-TiO_2-SiO_2系统)。计算的密度、热焓和结构随压力和温度的变化有望加深我们对多组分熔体体系中浮力、成键、挥发性组分的形态、多态和混合热力学的理解。2)通过对自扩散系数和粘度系数的第一性原理预测，研究了硅酸盐熔体的输运性质。位置-时间序列的原子可视化将使我们深入了解压缩和传输现象的微观机制，以及扩散/粘度对温度、压力和组成的复杂依赖。3)继续研究硅酸盐玻璃的结构和压缩机制，以进一步了解液体结构背后的能量学，并丰富与玻璃状态下地质相关成分的广泛实验文献的联系。该提案的统一主题是大系统的第一性原理模拟，需要探索真实的组成，准确地计算动力学性质，并成功地捕捉玻璃结构的本质。PI可以在本地访问足够的资源来执行这种密集的模拟。这项拟议的研究实质上是利用计算科学的思想和技术来挑战地球材料调查中的问题。它将对许多领域产生影响，包括地球化学、岩石学、地球物理学、计算矿物/材料物理和科学可视化，它将培养学生拥有多学科的经验和专业知识。

Silicate liquids are primary agents of chemical and thermal evolution in the Earth as they form early magma ocean, and appear as magmas in the surface of the planet and as partial melts in the crust and mantle. Magmatic processes are responsible for the origin and ongoing formation of the oceanic and continental crust, and for bringing to the surface one of our primary clues to the composition of the interior in the form of xenoliths. Because of the contrast in density, chemical diffusivity, viscosity, and bulk composition between silicate liquids and their source regions, the generation and transport of magma is among the most efficient geological processes leading to mass and heat transport. Silicate liquids may have played an even more important role in the Earth's early history when melting may have been more widespread and may have extended to greater depths. A better understanding of planetary evolution thus requires major advances in our knowledge of relevant melt properties at mantle pressure-temperature conditions.In this project, it is proposed to apply a combination of first-principles computational and visualization techniques to investigate structure, diffusion and viscosity of silicate melts over broad range of pressure, temperature and composition. This approach, being parameter free in the nature, is expected to provide the ideal complement to experimental approaches, and can provide important insights into the fundamental physical properties and behavior in structure and bonding. The specific activities proposed include: 1) Expanding the range of composition towards sampling natural melts (MgO-CaO-Na2O-K2O-Al2O3-TiO2-SiO2 system) with/out volatiles (H2O and CO2). The calculated densities, enthalpies, and structures as a function of pressure and temperature are expected to enhance our understanding of buoyancy, bonding, speciation of volatile components, polymorphism, and thermodynamics of mixing in a multi-component melt system. 2) Investigating the transport properties of silicate melts through first-principles predictions of the self-diffusion and viscosity coefficients. Atomistic visualization of the position-time series will allow us gain insight into the microscopic mechanisms of compression and transport phenomena, and into the complex dependence of diffusion/viscosity on temperature, pressure and composition. 3) Continuing the study of structure and compression mechanisms of silicate glasses as a way of gaining additional insight into the energetics underlying liquid structure, and in order to enrich contact with the extensive experimental literature on geologically relevant compositions in the vitreous state. The unifying theme of the proposal is the first-principles simulations of large systems needed to explore realistic compositions, to accurately compute dynamical properties and to successfully capture the essence of glass structures. The PI has local access to sufficient resources to carry out such intensive simulations. The proposed research is essentially an exploitation of ideas and techniques of computational science to challenging problems in the investigation of Earth materials. It will have impact on a number of fields including geochemistry, petrology, geophysics, computational mineral/materials physics, and scientific visualization, and it will train students to have a multidisciplinary experience and expertise.