Year: 2019
Type: dissertation
Abstract: Reaching the full potential of optoelectronic materials is often hindered by the years of necessary trial-and-error. Perovskites are an example of materials having exceptional optoelectronic properties, but require improvement with respect to stability and toxicity as they approach commercialization. Exploring new types of perovskites is key to achieving these goals. In this thesis I develop an accelerated materials discovery pipeline aimed at discovering new perovskite materials. This pipeline incorporates image recognition that detects crystals via convolutional neural networks with 95% accuracy and uses parameter exploration to predict an optimal material with experimental data. With this framework, I discovered a new type of perovskite single crystal, (3-PLA)2PbCl4, that employs a new ligand, 3-PLA, offering avenues to higher efficiency and more stable devices. This work develops a framework for discovering and optimizing materials in a wide chemical space and provides the groundwork for identifying new materials that lie beyond known chemical spaces.
Source: TSpace (University of Toronto)
Author: Jeffrey Kirman
Language: English
Subfield: Materials Chemistry
Field: Materials Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3121956847