RUI: Simulation and Improved Analysis of Data from PAC and other Hyperfine Methods for Studying Local Atomic Jumps and Long Range Diffusion in Intermetallic Compounds

技术支持：超精细方法，如核磁共振，穆斯堡尔和扰动角相关（PAC）光谱是敏感的电子和磁性结构的变化，在约1纳米的探针原子，所使用的技术。因此，超精细方法能够检测局部原子跳跃。因此，超精细方法有可能使研究人员能够确定有序合金中的操作扩散机制，并研究纳米晶材料中的原子运动。目前，超精细方法获得的数据分析的复杂性阻碍了这些技术在扩散研究中的广泛应用。复杂性出现在随机模型的发展中，该模型将原子的跳跃与它们引起的时变相互作用联系起来。PI将扩展Winkler和Gerdau提出的方法，以整合超精细方法的理论光谱计算。在过去，对计算能力的限制要求研究人员使用Winkler和Gerdau的方法来执行耗时的活动，例如（1）使用对称参数解析地减少大型矩阵或（2）开发矩阵特征值问题的数值解的解析近似。现代计算机足够强大，人们现在可以将随机模型的数值解直接纳入最小二乘拟合软件，用于直接从随机模型分析光谱。将开发足够通用的计算机代码，以支持所有类型的超精细相互作用和具有任何自旋状态的原子核的随机模型，并帮助网络基础设施（CI）的发展。该代码将允许计算理论光谱的超精细方法，可用于最小二乘拟合的实验数据。除了有助于数据分析外，计算机代码还将支持不同实验条件下的光谱模拟，例如不同的晶体结构，缺陷模型，成分，温度和扩散机制。所提出的工作的一个重要部分将利用模拟来确定在什么样的实验条件下可以使用超精细方法来区分有序合金中的一种扩散机制。非技术性：固体中原子的运动对于材料的选择和加工至关重要。虽然这项工作的重点将是金属间化合物的扩散，但快速开发用于扩散分析和模拟的随机模型的能力也将使对陶瓷和半导体材料感兴趣的研究人员受益。此外，这将是有益的纳米粒子中的原子运动的研究和自旋涨落的研究。

TECHNICAL: Hyperfine methods such as Nuclear Magnetic Resonance, Mossbauer, and Perturbed Angular Correlation (PAC) spectroscopies are sensitive to changes in electronic and magnetic structure within about 1 nm of the probe atoms that are used by the techniques. Hyperfine methods are therefore capable of detecting local atomic jumps. As a consequence, hyperfine methods have the potential to allow researchers to determine operative diffusion mechanisms in ordered alloys and to study atomic motion in nanocrystalline materials. At present, complications in analysis of data obtained by the hyperfine methods hold back widespread application of these techniques in studies of diffusion. The complications arise in the development of stochastic models that connect jumping of atoms to the time varying interactions they induce. PIs will expand on the approach set forth by Winkler and Gerdau to integrate the calculation of theoretical spectra for the hyperfine methods. In the past, constraints on computational power have required researchers who use Winkler and Gerdau's approach to perform time consuming activities such as (1) reducing large matrices analytically using symmetry arguments or (2) developing analytic approximations to numerical solutions of matrix eigenproblems. Modern computers are powerful enough that one can now incorporate numerical solution of the stochastic models directly into least-squares-fitting software for analyzing spectra directly from the stochastic model. Computer code that is sufficiently general to support stochastic models for all types of hyperfine interactions and for nuclei with any spin state will be developed, and aid cyberinfrastructure (CI) development. The code will allow calculation of theoretical spectra for the hyperfine methods that can be used in least-squares-fits of experimental data. In addition to helping in data analysis, the computer code will support simulation of spectra under varying experimental conditions such as for different crystal structures, defect models, compositions, temperatures, and diffusion mechanisms. A significant portion of the proposed work will utilize simulations to determine under what experimental conditions one can use hyperfine methods to distinguish one type of diffusion mechanism from another in ordered alloys. NON-TECHNICAL: The movement of atoms in solids is of fundamental importance in selection and processing of materials. While the emphasis of this work will be on diffusion in intermetallic compounds, the ability to quickly develop stochastic models for analysis and simulation of diffusion would benefit researchers interested in ceramic and semiconducting materials as well. Moreover, it would be beneficial for studies of atomic motion in nanoparticles and for studies of spin fluctuations.