Computationals studies of order, disorder and stability in metal halide perovskite photovoltaic absorbers

Abstract

Hybrid organic-inorganic perovskites have attracted much interest for their potential use as photovoltaic absorbers in highly efficient, cheaply manufactured photovoltaic (PV) devices. This thesis studies several topics of interest to hybrid organic-inorganic perovskite (HOP) photovoltaics, including the dynamic behaviour of the organic component in methylammonioum lead iodide (MAPbI3), the moisture induced degradation of typical hybrid perovskite materials and the possibility of engineering ferroelectric photovoltaic devices using hybrid perovskite-like materials. The observation of hysteresis in current-voltage curves for most hybrid perovskite devices suggests either dynamic movement of charged ions or a ferroelectric response of the absorber material to applied electric fields. The possibility that alignment or anti-alignment of organic cations, which are molecular dipoles, could be causing hysteresis in hybrid perovskite devices was examined using semi-classical Monte-Carlo simulations, informed by density functional theory calculations and experimental quasi-elastic neutron scattering experiments. I show that the Monte Carlo model does indeed suggest the possibility of either ordered low temperature states, and that the nature of these states (ferroelectric or anti-ferroelectric) is determined by the magnitude of a parameter related to the distortion of the inorganic PbI6 framework when methylammonium cations are aligned or anti-aligned. I also present a basic model based on the Monte Carlo and quasi-elastic neutron scattering results in that chapter which gives a characteristic timescale for domain wall motion within MAPbI3 devices. The timescale for domain wall motion across a typical device is calculated to be on the order of milliseconds, which is deemed to fast to be consistent with observed current-voltage hysteretic behaviour in hybrid perovskite solar cell devices. These results constitute the first theoretical results of this thesis. Despite rapid advances in photovoltaic efficiency of hybrid perovskite PV devices, their viability as a mainstream commercial technology is impeded by their susceptibility to a variety of environmental conditions, including the presence of airborne water vapour. The second part of this thesis studies the hydration of HOPs using electronic structure methods and ab initio thermodynamics based upon density functional theory calculations. The calculations presented show that ab initio thermodynamics can capture the key trends observed in experiments, but that it is difficult to reproduce experimental phase boundaries accurately using this technique (as is explained in that chapter this is not unexpected). A model of the entropy associated with the presence of a rotationally mobile organic cation within the perovskite is also introduced and used to augment the basic ab initio thermodynamic calculations. It is shown that the magnitude of this entropic contribution could play a key role in the location of the thermodynamic phase boundary between MAPbI3 and its hydrated structures. Finally, a mechanism for the hydration of MAPbI3 is proposed and outlined. The final part of this thesis discusses the application of the theory of hybrid improper ferroelectricity, whereby changes in the typical perovskite structure (for example by creating n = 2 layers of ABX3 perovskite intercalated with single layers with rock-salt AX stoichiometry) to metal halide perovskite based materials. It is shown through DFT calculations of polarisation and in depth analysis of the symmetries present in crystals studied that Ruddlesden-Popper and superlattice structures based on CsPbI3 metal halide perovskites should indeed exhibit hybrid improper ferroelectricity. The electronic structures of the Ruddlesden-Popper Cs3Pb2I7 is compared with that of CsPbI3 and it is shown that the Ruddlesden-Popper has a larger band gap. It is explained that this is due to changes in orbital overlap between lead and iodide states within the structure. A discussion is given of why hybrid improper ferrroelectricity is a useful property to have for solar cell absorber materials, what this means for devices based on these materials and key challenges for moving forward with this line of research is given.