poster-icon
Description
Maged Abdelsamie1 Junwei Xu1 Michael Toney1

1, SLAC National Accelerator Laboratory, Stanford University, Menlo Park, California, United States

Organo-metal-halide perovskites have gained tremendous attention as potential materials for photovoltaics, demonstrating efficiencies approaching the best silicon solar cells. Many approaches have been adopted to manipulate perovskite formation including anti-solvent processing, compressed-gas treatment, and post-deposition thermal annealing, where films can be deposited using different coating techniques such as spin-coating or blade-coating. Understanding the role of processing strategies on crystallization pathways is of crucial importance as crystallization strongly affects the perovskite film microstructure, its stability, and devices performance. Moreover, crystallization pathways become more complicated for perovskites with a mixed stoichiometric mixture such as CsxFA1−xPbI3 due to the thermodynamic and kinetic competition to form secondary phases. Herein, using time-resolved x-ray scattering, we investigate the film formation of Cs-FA-containing perovskites with stoichiometry [Cs0.15FA0.85PbI3] in situ during spin coating, blade coating, and the subsequent post-deposition thermal annealing, while different processing approaches such as anti-solvent [chlorobenzene (CB)] and compressed-gas (N2) treatments were applied during film casting.
We show how different processing routes affect the competition between the formation of the non-perovskite δ-phase and perovskite α-phase during film formation. When either anti-solvent or compressed-gas treatment is used, both δ-phase and α-phase are induced during casting, with the δ-phase more dominant in the as-cast film. However, each approach works with different mechanisms; while anti-solvent induces immediate crystallization from the bulk wet film, applying compressed N2 works by depleting volatiles from the top-surface leading to surface-induced crystallization that occurs after reaching supersaturation. When neither treatment is applied, the as-cast film is mostly amorphous with little non-perovskite δ-phase formation. In addition, we show the evolution of phases during thermal annealing. Our results reveal that, for the non-treated films, crystallization of perovskite α-phase occurs predominantly from amorphous phase rather than δ-phase to α-phase transformation during thermal annealing. However, direct phase transformation of δ-phase to α-phase is more dominant for CB- and N2- treated films, while perovskite crystallization occurs from an initially ordered film compared to non-treated films. Furthermore, our findings reveal that using blade-coating promotes completely different crystallization pathways than spin-coating, while solvothermal direct crystallization of perovskite α-phase occurs predominantly, leading to overcoming the non-perovskite δ-phase formation without the need for post-deposition thermal annealing. Our work highlights the importance of real-time investigation of film formation which can provide an in-depth understanding of the mechanisms of perovskite formation and help to establish processing-microstructure-functionality relationships.

Tags