UC Santa Barbara

Halide perovskite materials have emerged as a potentially disruptive technology in the field of photovoltaics with devices exceeding 20% power conversion efficiencies. Crystallizing in the ABX3 perovskite structure, these materials incorporate main-group cations (Pb2+, Sn2+, Ge2+) on the B-site, halide anions (Br–, Cl–, I–) on the X-site and large monovalent cations (Cs+, CH3NH3+) on the A-site. Record-breaking materials are achieved by tuning the band gap through halide substitution on the X-site and by increasing structural stability through the use of mixtures of organic cations and inorganic alkali metals on the A-site. In addition to configurational degrees of freedom associated with different alloying strategies, vibrational contributions to the free energy play a large role in the phase evolution of these materials and result in structural phase transitions due to octahedral tilting of the metal-halide sublattice. The phase evolution can be further complicated by the presence of order-disorder transitions due to orientational degrees of freedom of the molecular A-site cations. Anharmonic dynamic fluctuations on all three sublattices give rise to a highly polarizable and deformable lattice which plays a role in the remarkable optoelectronic properties observed in halide perovskites.

In this thesis we examine the role of orientational, vibrational, and configurational degrees of freedom in the phase evolution of halide perovskite materials using first-principles electronic structure calculations. In particular, density functional theory calculations reveal anisotropic molecular motion in hybrid perovskite CH3NH3PbI3 as well as a highly anharmonic energy landscape due to octahedral tilting displacement modes across all inorganic halide perovskites.

To link first principles calculations to finite-temperature thermodynamics we make use of cluster expansion effective Hamiltonians applied to both configurational and vibrational degrees of freedom in conjunction with Monte Carlo simulations. In particular, we predict temperature-composition phase diagrams for halide substitution in six Cs-based perovskite binary alloy systems where solid solutions are suppressed by the size-mismatch of end-members. Additionally, we employ machine learning methods to parameterize an anharmonic vibrational cluster expansion enabling both ab initio prediction of finite temperature phase transitions as well as a unique opportunity to investigate local structure at high temperatures from first principles.