Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department


Major Professor

Ivan I. Oleynik, Ph.D.

Committee Member

Sagar Pandit, Ph.D.

Committee Member

Humberto Rodriguez Gutierrez, Ph.D.

Committee Member

Arjan Van Der Vaart, Ph.D.

Committee Member

Aidan Thompson, Ph.D.


Carbon, Density Functional Theory, High Pressure, Pentazolate, Polynitrogen


The prediction of the structure of a crystal given only the constituent elements is one of the greatest challenges in both materials science and computational science alike. If one were to try to predict a novel crystal by brute force, meaning by arranging the atoms in every possible position of the unit cell and optimizing the geometry to find the energy minima of the potential energy surface, the amount of computer resources required to complete the calculation on the timescale of a few years would vastly exceed the currently installed computational capacity of the entire world. Fortunately, several methods have been developed to circumvent this problem, allowing for the prediction of the structure of many crystals on an attainable timescale.

This dissertation focuses on the first-principles prediction of novel materials at extreme conditions. An area of active research is the search for novel high-nitrogen-content energetic materials (EMs), which promise to achieve higher energy release while reducing harmful environmental impacts of traditional CHNO explosives. Over the last several years, there have been several attempts to synthesize an isolated cyclo-N_{5}^{-} anion at ambient pressure. The cyclo-N_{5}^{-} anion contains N-N conjugated (between single and double) bonds and would prove to be a promising candidate for next generation EMs due to the substantial energy release upon conversion of conjugated N-N bonds to triple bonds of N_{2} gas. This dissertation research investigated a series of metal-nitrogen compounds at high pressures with the goal of predicting crystal structures and compositions as well as resulting properties of the novel class of EMs. By applying pressure to high-nitrogen content compounds, one can hope to break the strong bonds found in the precursor materials to initiate new nitrogen bonding geometries that could prove to be metastable at ambient conditions. Stimuli can then be applied to the metastable materials, pushing them towards the stable state, inducing the large release in energy.

In another closely related project, the crystal structure prediction method has been extended to the study of carbon under uniaxial shock-like compression. The high pressure (~TPa) phase diagram of carbon is of great interest for several reasons- namely its use in inertial confinement fusion and the discovery of carbon rich exoplanets. Shock compression is the only viable route to achieve such high pressures in experiment. In order the capture this anisotropic compression environment, the crystal structure prediction method is adapted to search for carbon crystals under uniaxial compression. A series of crystals called diamond polytypes are identified as accessible via shock compression experiments owing to the fact that they are energetically competitive to cubic diamond, and in fact preferred to hexagonal diamond. The project pursued two important goals: to explore the metastability of carbon phases under shock compression as well as to generate a multitude of novel carbon structures to be used for training a quantum accurate machine learning interatomic potential for carbon. Machine learning potentials are urgently sought for large-scale quantum-accurate molecular dynamics simulations of shock compressed materials.