The Generalized Cluster Expansion: A Tool for Representing Structure-Property Relationships

点击翻译按钮获取中文摘要

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).TECHNICAL SUMMARYStructure-property relationships have long guided materials discovery and improved our understanding of the mechanisms underlying diverse physical phenomena. This award supports developing a set of theoretical and computational tools to determine and efficiently represent the relationship between the atomic structure of a compound and a given material property. The starting point is the cluster expansion formalism, which can express the relationship between a scalar material property and a crystalline alloy?s atomic configuration. This project generalizes this formalism to tensor-valued and nonuniform/nonisotropic material properties.The resulting generalized cluster expansion coupled with first-principles calculations will be applied in three general areas:(1) Phase field modeling: the generalized cluster expansion can encode the relationship between various phase fields representing the state of order of an alloy and tensorial properties, such as elastic constants or lattice parameters that enter the mesoscopic equation of motion of the simulated system.(2) Automated discovery of atomistic mechanisms: An illustrative application is the determination of the atomic-level origins of specific features in the phonon spectrum through the use of environment-dependent force constants tensors. Here, each term in the generalized cluster expansion would provide independent geometrical insight regarding the mechanisms at work.(3) Materials discovery and/or optimization: The generalized cluster expansion will be used to demonstrate an efficient computational combinatorial screening of a large number of candidate superlattices in search of one with the desired combination of properties relevant for device design.Software tools developed in this project will be included within the Alloy Theoretic Automated Toolkit, a widely used software package that the PI has made available on the world-wide web. This project will promote integration of research and education as well as diversity in science by contributing to the Materials Partnership Program established by Caltech?s Center for the Science and Engineering of Materials which recruits undergraduates from California State University, Los Angeles in collaborative research projects with Caltech.NON-TECHNICAL SUMMARYThis award supports theoretical and computational research that will develop new computational algorithms and tools for determining the relationship between the structure of a specific, possibly candidate, material and its properties. The PI will extend an existing approach to enable inclusion of a larger set of materials properties that enhance the power and utility of the method. The research effort contributes to the effort to discover new materials with desired properties using computation and starting only with the identity of the constituent atoms. This research may lead to the discovery of new materials with desired properties for technological applications, or optimizing the properties of existing materials or materials structures. New materials for optoelectronic devices are a specific focus of this project. Software tools developed in this project will be made available to the broader materials research community and will contribute to its cyberinfrastructure.