Algorithms for Improved Engineering of Chemistry

William Marsh Rice University, Houston TX

Investigators

Abstract

Abstract Proposal Title: Algorithms for Improved Engineering of Chemistry Proposal Number: CTS-0114263 Principal Investigator: Michael Deem Institution: University of California Los Angeles The objective of this project is the development of strategies for searching for the best materials property using combinatorial techniques coupled with Monte Carlo methods. The techniques to be developed are aimed at improving on the current grid search methods, which are often cumbersome to employ. Monte Carlo strategies are expected to provide significant improvement, and the ability to incorporate a priori knowledge should further improve the search strategy. Several variables can be manipulated in order to seek the material with the optimal figure of merit. Variables such as composition as well as film thickness and deposition method can be investigated. The work will focus on the theoretical aspects of the problem. Virtual libraries and materials will be constructed, and the Random Phase Volume model will be used to assess the validity of the methods. Techniques developed will be applied to data generated by an industrial collaborator. A wide range of improved material properties such as superconductivity, magnetoresistance or catalytic activity have been sought using combinatorial methods. The overall goal of high-throughput methods is to hasten discovery for improved processes and materials properties.