Integrating Advanced Morphology Measurements, Kinetic Monte Carlo Simulations, and Electrical Device Characterization
Abstract: Organic photovoltaics (OPVs) have attracted major attention over the last two decades for their potential as a cheap, renewable energy technology with unique application potential unavailable to traditional inorganic technologies. Key to the development of this technology has been an understanding and control of the thin-film morphology from the molecular to mesoscale. To produce efficient devices, charge carriers must be quickly extracted from the active layer after generation, but in a competing process, oppositely charged carriers can meet and recombine. In high-performance devices, the bimolecular charge recombination process represents the dominant loss mechanism and further mitigating this loss is one of the major avenues toward continued performance improvements. However, rational chemical and processing design to both maximize the extraction rate and minimize the recombination rate has been difficult because of gaps in fundamental understanding of which factors impact each competing process.
Among a variety of modeling methods, kinetic Monte Carlo (KMC) simulations have been a uniquely powerful tool due to their ability to capture the nanoscale morphological details while retaining the ability to simulate full devices, thereby acting as a coarse-grained model that bridges the gap between atomistic and continuum methods. In this talk, I will highlight my use of KMC simulations with model morphologies to develop novel physical models that capture how the complex morphology impacts the bimolecular charge recombination kinetics and charge carrier transport. Building on this work, I will outline my current efforts to integrate transmission electron microscope tomography data into my new open-source KMC simulation software tool to refine and improve these models. I will then describe my recent efforts to develop an impedance-based experimental method for quantifying the recombination and transport behavior with high accuracy and demonstrate how improved data collection and sharing has allowed an assessment of competing device models. With increased adoption of this technique, machine learning techniques could produce improved models that capture the complex processing-structure-property relationships and accelerate the development of new materials for OPVs.
Note: Interested individuals are invited to contact Noah Paulson (firstname.lastname@example.org) to set up a meeting with Dr. Heiber.