An Atom-Probe Tomography and Lattice Kinetic Monte Carlo Study of Phase Separation in Ni-Al-Based Alloys from the Atomic Scale Up to Link with Continuum Theories

技术概要：在浓缩的多组分过饱和固溶体的相分离的早期阶段的动力学途径是至关重要的最终微观结构的发展，这是在宏观尺度上控制材料的物理和机械性能的重要技术。在理论方面，动力学途径描述的成核，生长和粗化，其理论意味着在基体中的扩散，转移的原子穿过基体/沉淀物界面与局部平衡的假设。最新的成就，高产量的原子探针层析成像（APT）和快速晶格动力学蒙特卡罗（LKMC）模拟，使人们有可能在原子尺度上描述相分离。热力学和动力学的必要参数推导出第一性原理计算和实验拟合的LKMC参数。在相干沉淀的情况下，相分离仅仅是由原子种类的不均匀分布中的无数空位跳跃留下的足迹。由于相同的空位跳跃在平衡固溶体中起作用，因此在固溶体中的动力学途径和扩散特征之间建立了联系。我们的APT实验结果直接表明，在固态扩散的相关效应（通量耦合）控制的早期阶段的动力学途径相干相分离。这些效应被所有现存的成核理论的变体所忽略，它们都依赖于扩散矩阵的简化形式。我们开发了一种变革性的方法，采用最先进的APT与LKMC模拟和扩散理论相分离，实验研究动力学途径，并通过模拟和扩散理论对其进行建模。我们已经证明了这种方法的独特性集中的Ni-Al-Cr合金。APT和LKMC在模型镍基高温合金中的结合使用为发展二元（Ni-Al）和浓缩多元合金（Ni-Al-Cr和Ni-Al-Mo）的新形核理论提供了独特的数据基础（实验和LKMC结果）。经典成核理论假设原子通量是这样的，分解路径遵循构型自由能表面的最陡斜率，而通量的动力学耦合是为什么这通常是不正确的一个原因：另外，所涉及的不同的扩散系数是另一个原因。我们的方法将考虑耦合通量的影响，导致第二相的成核。在这种新方法中，界面宽度应受扩散机制的影响（与Cahn-Hilliard形式主义不同）。此外，我们建议实现一个并行LKMC模拟代码的多组件系统的基础上完美的时间同步，这是基于由马丁内斯等人开发的代码。在LANL，将加速LKMC模拟。拟议中的研究将导致一个完全不同的方式查看动力学途径相分离的固溶体，因此，成核，生长和coarsening.NON-TECHNICAL摘要：我们独特的方法相分离使得广泛使用的局部电极原子探针（LEAP）断层扫描仪在西北大学中心原子探针断层扫描（NUCAPT）。NUCAPT是美国唯一的大学设施，它已经并将对本科生勤工俭学，高级论文，本科生研究经验（REU），MS，博士和博士后学生：在过去四年中，许多代表性不足的学生群体使用了NUCAPT。在西北大学材料科学与工程系，六位教授和他们的学生广泛使用NUCAPT。此外，NUCAPT还拥有来自美国其他大学以及国家和工业实验室的用户。基于原子探针断层扫描的研究结果发表在档案文献中，我们的网站http://nucapt.northwestern.edu有一个图书馆，可以从那里下载PDF文件。通过NUCAPT，我们已经教育了大量的人在原子探针断层扫描的应用范围广泛的材料科学和工程问题，这是重要的表征材料在亚纳米尺度。此外，我们正在与法国的材料科学家就这项研究的主题进行互动，并为美国和国外的工业公司提供服务。美国没有其他人这样做，以至于我们正在实现这一目标。

TECHNICAL SUMMARY: The kinetic pathways in the early stages of phase separation of a concentrated multicomponent supersaturated solid-solution are critical to the development of the final microstructure, which is of significant technological importance for controlling the physical and mechanical properties of a material at the macroscopic scale. On the theoretical side, kinetic pathways are described in terms of nucleation, growth and coarsening, whose theory implies diffusion in the matrix, transfer of atoms across a matrix/precipitate interface together with the assumption of local equilibrium. The most recent achievements of high-yield atom-probe tomography (APT) and fast lattice kinetic Monte Carlo (LKMC) simulations make it possible to describe phase separation at the atomic scale. The necessary parameters for the thermodynamics and kinetics are deduced from first-principles calculations and experimental fits for the LKMC parameters. In the case of coherent precipitation, phase separation is simply the footprint left by the myriad of vacancy jumps in the inhomogeneous distribution of atomic species. Because the same vacancy jumps are at work in the equilibrium solid-solution, a link is established between features of kinetic pathways and diffusion in a solid-solution. Our APT experimental results demonstrate directly that correlation effects (flux-couplings) in solid-state diffusion control the early stage kinetic pathway for coherent-phase separation. Such effects are neglected by all extant variants of nucleation theory, which all rely on a simplified form of the diffusion matrix. We have developed a transformative approach that employs state-of-the-art APT coupled with LKMC simulations and diffusion theory for phase separation, which studies the kinetic pathways experimentally and models it via simulations and diffusion theory. And we have demonstrated the uniqueness of this approach for concentrated Ni-Al-Cr alloys. The combined use of APT and LKMC in model Ni-base superalloys provides a unique data-base (experimental and LKMC results) for developing a new theory of nucleation in binary (Ni-Al) and concentrated multi-component alloys (Ni-Al-Cr and Ni-Al-Mo). Classical nucleation theory assumes that the atomic fluxes are such that the decomposition path follows the steepest slope of the configurational free energy surface, whereas kinetic coupling of the fluxes is one reason why this isn't generally correct: also the disparate diffusivities involved is another reason. Our approach will take into account the coupled-flux effects that lead to nucleation of a second phase. In this new approach the interfacial width should be affected by the diffusion mechanism (unlike in the Cahn-Hilliard formalism). Additionally, we propose to implement a parallel LKMC simulation code for multi-component systems based on perfect time synchronicity, which is based on a code developed by Martinez et al. at LANL that will accelerate the LKMC simulations. The proposed research will lead to a completely different way of viewing kinetic pathways for phase separation of a solid-solution, and hence, nucleation, growth and coarsening.NON-TECHNICAL SUMMARY: Our unique approach to phase separation makes extensive use of a local-electrode atom-probe (LEAP) tomograph in the Northwestern University Center for Atom-Probe Tomography (NUCAPT). NUCAPT is unique university facility in the US, which has had and will have a significant impact on the research efforts of undergraduate work-study, senior thesis, research experiences for undergraduates (REU), M.S., Ph.D. and postdoctoral students: during the last four years many underrepresented groups of students have used NUCAPT. In the Department of Materials Science and Engineering at Northwestern six professors and their students make extensive use of NUCAPT. Additionally, NUCAPT has users from other US universities, and both national and industrial laboratories. Research results based on atom-probe tomography are published in the archival literature and our web site, http://nucapt.northwestern.edu , has a library from which reprints can downloaded as PDF files. Through NUCAPT we have educated a large number of people in the application of atom-probe tomography to a wide range of materials science and engineering problems, which is important for characterizing materials at the subnanoscale scale. Additionally, we are interacting with a materials scientist in France on the subject matter of this research and are supplying service to industrial companies in the US and abroad. There is no one else in the US doing this to the extent that we are accomplishing this objective.