CAREER: A Novel Computational Thermodynamics Framework with Intrinsic Chemical Short-Range Order

非技术总结这个职业奖项支持计算和理论研究、实验和教育，以提高计算材料相图的能力。相图是材料的路线图，表明原子在不同的温度、组成、压力或其他变量下是如何组织的。金属和合金的计算热力学模型可以揭示它们在不同温度和成分下的稳定性，为控制合金的性质提供指导。然而，目前计算热力学方法常用的数学模型中缺少对原子尺度短程有序的描述。当原子相互结合形成固体时，它们倾向于将特定种类的原子作为邻居，这导致了原子尺度(短程)有序。这一职业奖支持PI和他的团队通过开发一种新的基于集群的模型来追求计算热力学的根本改进，该模型将原子尺度有序纳入主流热力学建模框架。这个新的框架是在一个软件平台上实现的，并可供广泛的材料科学界使用。原子尺度有序的计算预测在几种合金中得到了实验验证，这些合金使用强高能X射线来获得不同类型的原子位置图像。开发的计算框架代表了一种新的方法，使原子尺度的秩序能够被用于材料设计，这可能导致新的复杂的浓缩合金或可能出现在汽车、飞机或其他应用中的改进的商业合金。这一职业奖项为教育和推广活动提供支持，包括：通过弗吉尼亚大学的NanoDays活动指导当地的K-12学生，通过弗吉尼亚-北卡罗来纳少数民族参与联盟为代表不足的少数族裔学生举办年度暑期计划，开发与热力学相关的课程和学习模块，以及创建用于热力学概念3D可视化的巨蟒代码库，使这个主题有趣和直观。技术总结这个职业奖支持计算和理论研究、实验和教育，以提高计算相图和材料，特别是金属和合金的热力学性质的能力。PI旨在探索一种基于团簇的热力学方法，该方法能够使用相图计算(CALPHAD)方法预测多组分材料的化学短程有序(SRO)。CALPHAD是材料计算热力学建模的主流方法。然而，目前在CALPHAD中使用的子晶格模型是一个具有理想混合熵的平均场模型。这使得CALPHAD不足以正确描述合金中的有序-无序转变或化学SRO，如Guinier-Preston区或纳米级团簇，这些对合金的机械性能至关重要。第一性原理合金理论可以用团簇变分方法(CVM)或团簇展开方法来描述SRO，但由于组态变量较多，一般仅限于二元或三元体系。PI计划将CVM和CALPHAD的独特优势结合起来，通过将SRO融入CALPHAD和基于集群的解决方案模型来开发混合框架。关键是利用Fowler-Yang-Li变换将云服务器中繁琐的簇概率分解为更少的基簇的点/点概率，从而大大减少了多组分系统的最小化变量数量。构型自由能和非构型自由能(例如振动和弹性自由能)被分开建模，以了解它们各自对相稳定性的影响。采用了现代高效的算法来最小化基于簇的非线性自由能函数。生成的代码是在开源平台OpenCALPHAD中实现的。所预测的化学SRO在选定的合金中得到了同步辐射X射线实验的验证。这种混合的CVM-CALPHAD框架代表了一种新的热力学建模方法，使原子尺度的有序性能够被用作材料设计的维度，这可能导致新型复杂的浓缩合金。该职业奖为教育和外展活动提供支持，包括：通过弗吉尼亚大学的NanoDays活动指导当地的K-12学生，通过弗吉尼亚-北卡罗来纳州少数族裔参与联盟为代表不足的少数族裔学生举办年度暑期计划，开发与热力学相关的课程和学习模块，以及创建用于热力学概念3D可视化的Python代码库，使这门学科有趣和直观。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

NONTECHNICAL SUMMARYThis CAREER award supports computational and theoretical research, experiment, and education to advance the ability to calculate phase diagrams of materials. Phase diagrams are roadmaps of materials that indicate how atoms organize themselves at different temperatures, compositions, pressures, or other variables. Computational thermodynamic modeling of metals and alloys could reveal their stability under various temperatures and compositions, providing guidance for manipulating the properties of alloys. However, currently missing in the mathematical models commonly used by computational thermodynamic approaches is the description of atomic-scale short-range order. When atoms bond with each other to form a solid, they prefer to have specific kinds of atoms as their neighbors which leads to atomic-scale (short-range) order. This CAREER award supports the PI and his team to pursue a fundamental improvement in computational thermodynamics by developing a new cluster-based model which incorporates atomic-scale order into the mainstream thermodynamic modeling framework. This new framework is implemented in a software platform and accessible to the broad materials science community. The computational prediction of atomic-scale order is experimentally validated in several alloys with experiments that use intense high-energy X-rays to obtain different kinds of images of the locations of atoms. The developed computational framework represents a new methodology that enables atomic-scale order to be exploited for materials design, which potentially leads to novel complex concentrated alloys or improved commercial alloys that may appear in automobiles, airplanes, or other applications.This CAREER award provides support for education and outreach activities including: mentoring local K-12 students through the NanoDays events at the University of Virginia, holding annual summer programs for underrepresented minority students through the Virginia-North Carolina Alliance for Minority Participation, developing a thermodynamics-related course and learning modules, and creating a Python code library for 3D visualization of concepts in thermodynamics, making this subject fun and intuitive.TECHINICAL SUMMARYThis CAREER award supports computational and theoretical research, experiment, and education to advance the ability to calculate phase diagrams and thermodynamic properties of materials, particularly metals and alloys. The PI aims to explore a cluster-based thermodynamic approach that enables the prediction of chemical short-range order (SRO) in multicomponent materials using the CALculation of PHAse Diagram (CALPHAD) method. CALPHAD is a leading method for computational thermodynamic modeling of materials. However, the current approach used in CALPHAD, the sublattice model, is a mean-field model with the ideal entropy of mixing. This makes CALPHAD inadequate for properly describing order-disorder transitions or chemical SRO in alloys, such as the Guinier-Preston zones or nanoscale clusters, which are critical for alloy mechanical properties. First-principles alloy theories, using the cluster variation method (CVM) or the cluster expansion method, can describe SRO but are generally limited to binary or ternary systems due to the large number of configuration variables. The PI plans to develop a hybrid framework by marrying unique advantages from CVM and CALPHAD through incorporating SRO into CALPHAD with a novel cluster-based solution model. The key is to use the Fowler-Yang-Li transform to decompose the cumbersome cluster probabilities in CVM into fewer site/point probabilities of the basis cluster, thereby considerably reducing the number of minimizing variables for multicomponent systems. The configurational and non-configurational, for example vibrational and elastic, free energies are modeled separately to gain insight into their respective effects on phase stability. Modern, efficient algorithms are employed to minimize the non-linear cluster-based free energy functions. The resulting codes are implemented in an open-sourced platform, OpenCALPHAD. The predicted chemical SRO is validated in selected alloys with synchrotron X-ray experiments. This hybrid CVM-CALPHAD framework represents a new methodology for thermodynamic modeling that enables atomic-scale order to be exploited as a dimension for materials design, which potentially leads to novel complex concentrated alloys. This CAREER award provides support for education and outreach activities including: mentoring local K-12 students through the NanoDays events at the University of Virginia, holding annual summer programs for underrepresented minority students through the Virginia-North Carolina Alliance for Minority Participation, developing a thermodynamics-related course and learning modules, and creating a Python code library for 3D visualization of concepts in thermodynamics, making this subject fun and intuitive.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.