Title:	Computational search for new clean energy materials
DNr:	SNIC 2015/1-151
Project Type:	SNIC Medium Compute
Principal Investigator:	Surendra Saxena <saxenas@fiu.edu>
Affiliation:	Florida International University
Duration:	2015-04-29 – 2016-05-01
Classification:	10304 10403 10105
Homepage:	http://Www.cesmec.fiu.edu/
Keywords:

Abstract

The primary objective of this project is to design new materials using recently developed evolutionary algorithm USPEX (Universal Structure Predictor: Evolutionary Xtallography) . It allows one to predict the stable structure of a given compound at given conditions (pressure, temperature) just from the knowledge of the chemical composition and using no experimental information. We will also focus on the key issue of producing a hydride that will have appropriate volume and density to be useful for hydrogen storage. Theoretical modeling and simulations are very well suited to predict properties and optimize the composition of hydrogen storage systems. Such guidance from theory can greatly help to improve the efficiency of the development process. Defect chemistry is at the center stage of the fundamental investigation into metal oxides, as it allows us to determine thermodynamic equilibria, electrical conductivity, superconductivity, catalytic activity, electromechanical properties, and many other properties. Investigation on defect chemistry has been carried out for almost a century mostly with analytical defect chemistry analysis combined with transport experiments. Various new approaches have been adopted on the investigation of defect chemistry in the last two decades, including first-principles calculations, and analytical and numerical models. On the other hand, extensive efforts have been made to develop the thermodynamic databases for metal oxides using the computational thermodynamics approach. However, because of the lack of communications between the computational thermodynamics and defect chemistry research communities, no effort has yet been made to systematically investigate quantitative defect chemistry based on computational thermodynamics and its application for various properties. Based on our exciting preliminary results linking computational thermodynamics, defect chemistry, and electrical and thermomechanical properties, we propose to develop the quantitative defect chemistry analysis to accurately capture the effect of temperature, p(O2), and stoichiometry for multicomponent metal oxides. We will use a deep integration of first-principles calculations, computational thermodynamics, analytical defect chemistry analysis, and transport experiments and utilize the most common perovskite families as a case study. By the end of this project, we will demonstrate that this integrated and general approach can be applied to other metal oxides used in many different applications. First-principles calculations have been successfully adopted to provide the critical phase equilibria and thermochemistry data at 0K for the thermodynamic modeling, including both the stoichiometric compounds and solution phases. Perovskites are solution phases with many end members. Although the Gibbs energies of these end members are difficult to measure because some are unstable, they are also required to model the solution phases. In the literature, some end members are traditionally assigned arbitrary values. Our work will provide the accurate enthalpy of end member formation using first-principles calculations, which will improve the accuracy of the Gibbs energy descriptions for perovskite. Besides the standard application of the first-principles calculations, they can also be adopted to calculate the defect energetics of various perovskite-related defect reactions. In corporation with the computational thermodynamics, we may expand the defect energetics prediction into multicomponent systems.