| Summary: | Thesis: S.M., Massachusetts Institute of Technology, Department of Materials Science and Engineering, 2016. === Thesis: S.M., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2016. === This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. === Cataloged from student-submitted PDF version of thesis. === Includes bibliographical references (pages 45-52). === Accelerating the discovery of advanced materials is essential for human beings. However, the traditional trial-and-error way of developing materials is often very empirical and time- consuming. In 2011, the launch of Materials Genome Initiative marked a large-scale collaboration between computer scientists and materials scientists to deploy proven computational methods to predict, screen, and optimize materials at an unparalleled scale and rate. This thesis is based on this idea. Finding a suitable cathode material for Mg batteries has been one of the key challenges to the next-generation multi-valent battery technology. In this thesis, a high-throughput computation system is proposed to solve such problem. I tested the high-throughput structures applying traditional NEB calculations schemes and find out it is very different to scale traditional NEB method to a high-throughput application. Then I proposed a new scheme for estimating migration minimum- energy path (MEP) geometry and energetics (PathFinder and ApproxNEB). By testing our methodology against standard NEB calculations and literature values, we find that the PathFinder algorithm can reliably predict the geometry of cation migration MEP within 0.2 Å at negligible computational cost. Furthermore, we find that the ApproxNEB calculation scheme yields activation barriers for migration within an error bound of 20 meV while using significantly fewer computational resources than NEB. We envision that our methods can be used to accelerate NEB calculations, as well as to provide a robust estimation criterion for migration barriers in ionic materials for highthroughput computational screening of materials. Based upon these two newly developed methods, coupled with EndPointFinder, I developed two functional high-throughput applications (ApproxNEB for estimating migration barriers and PathFinder for calculating migration geometric paths), and have already put PathFinder high-throughput system into production and calculate around 2000 structures. === by Ziqin Rong. === S.M.
|
|---|
