Hume-Rothery Symposium on First-Principles Materials Design: Interface First-principle Method with Thermodynamics and Kinetics
Sponsored by: TMS Functional Materials Division, TMS Structural Materials Division, TMS: Alloy Phases Committee
Program Organizers: Bin Ouyang, Florida State University; Mark Asta, University of California, Berkeley; Geoffroy Hautier, Dartmouth College; Wei Xiong, University of Pittsburgh; Anton Van der Ven, University of California, Santa Barbara
Monday 8:30 AM
March 20, 2023
Room: Cobalt 501C
Location: Hilton
Session Chair: Mark Asta, University of California Berkey
8:30 AM Keynote
William Hume-Rothery Award Lecture: Ab initio Thermodynamics and Kinetics from Alloys to Complex Oxides: Gerbrand Ceder1; 1University of California, Berkeley
Any method for ab initio thermodynamics requires the combination of highly accurate energetics with the ability to integrate over the highly different timescales by which a system explores its degrees of freedom. The coarse-grained cluster expansion offers, in principle, such an approach to obtain thermodynamic information with the accuracy of ab-initio electronic structure methods. I will demonstrate the evolution of the cluster expansion and its current use in the design of highly complex, technologically relevant materials. With the design and property prediction of materials reaching unprecedented levels of success, I will argue how our focus needs to shift to applying ab-initio methods to understand and predict the synthesis of materials.
9:10 AM Invited
Double Descent, Linear Regression, and Fundamental Questions in Alloy Model Building: Gus Hart1; 1Brigham Young University
Though many data science concepts are just glosses on ideas that predate the data science revolution by years or even decades, some suggest altogether new approaches or raise fundamental questions. The phenomenon of double descent behavior in neural networks defies intuition and may seem to violate the "no free lunch" theorem. Is double descent behavior peculiar to neural networks? Or is it more general? We illustrate double descent in a simple linear regression model and then revisit basic questions in alloy model building, using the cluster expansion and machine learned interatomic potentials as illustrations. How is convergence impacted by the range of interaction? Or the order of an n-body interaction? How completely must we span configuration space with our expansions? We address these questions from the perspective of both mathematics and physics and discuss the implications for practical alloy models.
9:40 AM Invited
Linking Phenomenological Theories of Materials to Electronic Structure: Anton Van der Ven1; Brian Puchala2; Derick Ober1; 1University of California, Santa Barbara; 2University of Michigan
Phenomenological theories of equilibrium and non-equilibrium properties play a crucial role in how we understand and design materials. One challenge is to establish rigorous links between the electronic structure of a material and the thermodynamic and kinetic functions that inform phenomenological theories. In most applications, a statistical mechanics approach is crucial due to the importance of temperature and entropy. In this talk, I will describe coarse-graining schemes with which to generate free energy descriptions as a function of symmetry adapted order parameters. Many materials phenomena couple chemistry with mechanics, requiring free energy descriptions that integrate concentration and strain degrees of freedom. Free energy models that incorporate the crystallographic changes associated with plastic deformation are also increasingly of interest. Uncertainty quantification techniques will be described to propagate errors at the electronic structure level to the phenomenological scale. Examples from structural and battery applications will be used to illustrate the approach.
10:10 AM Break
10:30 AM Invited
Holistic Integration of Experimental and Computational Data and Simple Empirical Models for Diffusion Coefficients of Metallic Solid Solutions: Wei Zhong1; Ji-Cheng Zhao1; 1University of Maryland
Large amounts of both experimental and computational diffusion coefficients are available in the literature for careful assessments of the degree of agreements between the datasets. Such systematic assessments allow us to employ the best of both datasets and provide complementary data to establish the most reliable diffusion databases. The power of such holistic integration will be illustrated with unary and binary solid solutions. In addition, simple and yet generally applicable semi-empirical models of diffusion coefficients of binary and multicomponent metallic solid solutions are developed. Only one-fitting parameter is required to fully describe all the diffusion coefficients as a function of composition and temperature of a binary solution. Future calculations of this parameter and impurity (dilute) diffusion coefficients using first principles and machine-learning methods would be extremely valuable for the establishment of reliable diffusion (atomic mobility) databases for kinetic simulations of alloys.