A First-Principles Study of Highly Anharmonic and Dynamically Disordered Solids

Sammanfattning: This thesis is a first-principles theoretical investigation of solid materials with high degrees of anharmonicity. These are systems where the dynamics of the constituent atoms is too complex to be well-described by a set of uncoupled harmonic oscillators. While theoretical studies of such systems pose a significant challenge, they are under increasing demand due to the prevalence of these sytems in next-generation technological applications. Indeed, very anharmonic systems are ubiquitous in envisioned materials for future solid-state batteries and fuel-cells, thermoelectrics and optoelectronics. In some of these cases, the anharmonicity is a “side-effect” that simply has to be dealt with in order to accurately model certain properties, while in other cases the anharmonicity is the origin of the high-performance of the material.There are two main parts to the thesis: The first is on materials with perovskite-related structures. This is a very large class of materials, members of which are typically highly anharmonic, not least in relation to a series of complex phase transformations between different structural modifications. In the thesis, I have studied a specific class of such phase-transformations that relate to tilting of the framework of octahedra that make up the structure. The oxide CaMnO3 and a set of inorganic halide perovskites were taken as model systems. The results shed some light on the experimentally observed differences between the local and average atomic structure in these systems. I have further studied Cs2AgBiBr6, a member of the so-called lead-free halide double perovskites. I rationalized its temperature induced phase transformation and found high degrees of anharmonicity and ultra-low thermal conductivity. Finally, I studied the influence of nuclear quantum effects, which are often ignored in computational modelling, on the structure and bonding in the hybrid organic-inorganic lead-halide perovskite, CH3NH3PbI3.The second part of the thesis deals with theoretical studies of the phase stability of dynamically disordered solids. These are solids which have some sort of time-averaged long-range order, characteristic of a crystalline solid, but where the anharmonicity is so strong that the basic concept of an equilibrium atomic position cannot be statically assigned to all atoms. Examples include certain solids with very fast ionic conduction, so called superionics, and solids with rotating molecular units. This absence of equilibrium atomic positions makes many standard computational techniques to evaluate phase-stability inapplicable. I outline a method to deal with this issue, which is based on a stress-strain thermodynamic integration on a deformation path from an ordered variant to the dynamically disordered phase itself. I apply the method to study the phase stability of the high-temperature δ-phase of Bi2O3, which is the fastest know solid oxide ion conductor, and to Li2C2, whose high temperature cubic phase contains rotating C2 dimers.The thesis exemplifies the need to go beyond many of the standard approximations used in first-principles computational materials science if accurate theoretical predictions are to be made. This is true, in particular, for many emerging material classes in the field of energy materials.