UC Santa Barbara

Modern engineering applications use alloys with up to 10 different elements to meet performance requirements. Designing new alloys in such a vast multicomponent space is difficult. Even with first-principles methods to expedite the quest for new alloys, improvements to materials search algorithms are required to fully sift through the multidimensional chemical space. One possible search strategy is to exploit the natural hierarchy between crystal structures. Many crystal structures are related to a high symmetry parent crystal structure via a group/subgroup symmetry relationship. While many ordered structures are traditionally viewed as distinct crystals, they can be thought of as derivative structures of a parent crystal. The first half of this dissertation describes a framework for alloy design that is based on a parent-derivative crystal structure hierarchy. Chemical trends in this hierarchy are illuminated and new structural transformation pathways between important parent crystals are described.

The second half of this dissertation uses a similar philosophy to explore a different set of materials; it examines the effects of chemical trends on the electrochemical properties of materials used for energy storage. To meet rising energy demands with intermittent renewable energy sources such as solar and wind power, society needs to create high performance secondary batteries for energy storage. Li-ion batteries have quickly risen to be a prime candidate to meet energy storage needs, and new advances are being made to promote sustainable earth-abundant materials, higher charge rates, and larger capacities. Some of these new advances include the investigation of Na-ion and Mg-ion batteries. Spinel intercalation compounds are well-known to facilitate high rate and high voltage Li-ion batteries, but less is known about their Na-ion and Mg-ion counterparts. When exploring new materials, it helps to have a strong fundamental understanding of how changes in chemistry or structure affect electrochemical performance of battery materials. This dissertation uses a combination of electronic structure calculations and statistical mechanics methods to study the thermodynamic and kinetic properties of a wide range of spinel intercalation compounds.

The identification of a crystal structure hierarchy was achieved by analyzing existing crystal structure databases using a crystallographic mapping algorithm to find representative parent crystal structures from which many derivative structures are formed. We found that 73% of binary intermetallic compounds found on the ICSD to be derived from only 20 unique parent crystal structures. There are important crystallographical relationships, such as the Bain path and the Burgers path, that connect parent crystals within our hierarchy. These crystallographic relationships facilitate structural phase transformations which are important for shape-memory alloys, magnetocalorics, or self-assembling block co-polymers. A crystal structure mapping algorithm was used to show the existence of 7 new structural transformation pathways between orderings on simple parent crystal structures to others belonging to the top 20 most common parent crystal structures. We use high-throughput density functional theory (DFT) to probe which elemental combinations are most likely to cause spontaneous structural transformations without an energy barrier. We find multiple chemical combinations that lead to barrierless transformation pathways. These barrierless transformations suggest a nucleation mechanism that does not require large structural fluctuations that are often energetically costly.

In our investigation of electrochemical properties of energy storage materials, we examine the spinel structure which has a general formula AMX2 where A is a guest cation, M is a transition metal and X is a chalcogenide. The electrochemical properties of spinel can be affected by many factors including (i) ionicity of the MX2 framework, (ii) guest cation radius, and (iii) guest cation oxidation state. We conducted a systematic study and determined that guest cation radius and MX2 ionicity play a significant role in guest cation site preference which, in turn, affects the electrochemical properties of spinel. The insights of this study suggest that large cations in an oxide spinel creates a desirable energy landscape for high-rate capable batteries. Using kinetic Monte Carlo simulations and a model cluster expansion Hamiltonian, we identify a topological pitfall that makes spinel prone to highly correlated cation diffusion at intermediate to high guest cation concentrations. This undesirable property is dependent on the strength of the nearest-neighbor repulsion of diffusing cations within spinel. A strong repulsion results in a dependence on large vacancy clusters to mediate diffusion. This dependence indicates that a percolating network of large vacancy clusters is necessary to facilitate long range diffusion. Our results show that the detrimental effects can be mitigated by reducing the strength of the repulsion or increasing the connectivity of the percolating network. Although our focus is on the fully-ordered spinel structure, the conclusions drawn from this study apply to other close-packed anion hosts such as disordered rocksalt electrodes and partially ordered spinels.