SI2-SSE: Automated Statistical Mechanics for the First-Principles Prediction of Finite Temperature Properties in Hybrid Organic-Inorganic Crystals

University Of California-Santa Barbara, Santa Barbara CA

Investigators

Abstract

This project seeks to advance computational capabilities in materials science by developing new theoretical and computational tools to predict temperature dependent properties of complex crystalline materials containing organic molecules. The recent discovery that hybrid organic-inorganic compounds can achieve remarkable photovoltaic conversion efficiencies has led to the recognition that a fundamental understanding of these complex compounds is urgently needed and that first-principles computational tools are necessary to enable a prediction of their intrinsic materials properties. The room temperature properties of hybrid organic-inorganic compounds are strongly affected by thermal excitations. Important electronic, thermodynamic and kinetic properties of these compounds therefore cannot be predicted directly with quantum mechanical approaches alone, but require statistical mechanics tools that account for the effects of temperature. A major objective of this project is the development of highly automated statistical mechanics software tools to predict materials properties where disorder due to alloying, atomic vibrations and molecular rotations are rigorously accounted for. These tools will greatly enhance the ability to predict the properties of complex materials from first principles, thereby enabling the directed design of a broad class of new materials with applications in a wide variety of technologies, including energy conversion and storage, carbon capture and organic electronics. The fundamental scientific insights to be generated by this study on hybrid organic/inorganic compounds will lead to invaluable design principles to enable the further improvement of these compounds for photovoltaic applications. The proposed activity will also educate and train graduate students in computational materials science, a field that is increasingly recognized as invaluable in the design and rapid implementation of new materials. Modern first-principles electronic structure methods have reached a remarkable level of accuracy and ease of use, making them invaluable tools in the design of new materials. Electronic structure methods by themselves, however, do not explicitly account for the role of temperature on thermodynamic and kinetic properties. The properties of many promising materials for energy storage and conversion applications and for transportation applications depend sensitively on temperature due to large entropic contributions arising from atomic-scale excitations and disorder. Most materials of technological relevance are characterized by configurational disorder due to alloying and many high temperature phases are dynamically stabilized by large anharmonic vibrational excitations. Entropic contributions to equilibrium and non-equilibrium properties are especially important in a new class of hybrid organic-inorganic perovskites that show great promise as photovoltaic materials. These compounds belong to a class of crystalline materials that can host molecular species in large interstitial cages and exhibit a wide range of atomic and molecular excitations already at room temperature. Optimal photovoltaic properties are achieved by alloying on all three sublattices of the ABX3 perovskite crystal, leading to configurational disorder in addition to molecular and vibrational excitations. A statistical mechanics approach is therefore essential to accurately predict the electronic, thermodynamic and kinetic properties of these materials. The aim of this project is to develop a statistical mechanics framework and an accompanying highly automated software infrastructure that rigorously accounts for all relevant configurational, vibrational and molecular degrees of freedom in crystalline solids containing interstitial molecular species. The prediction of finite temperature thermodynamic and kinetic properties will rely on effective Hamiltonians that serve to extrapolate highly accurate first-principles electronic structure calculations within Monte Carlo simulations. A major activity of the project is the creation of a highly automated statistical mechanics software package called a Clusters Approach to Statistical Mechanics (CASM) to predict the finite temperature properties of multicomponent crystalline materials from first principles. The application of these tools in a first-principles study of alloyed hybrid organic-inorganic perovskites will generate a fundamental scientific understanding of the relative importance of the various atomic and molecular excitations on electronic structure, phase stability and ionic transport properties.