Year: 2020
Type: dissertation
Abstract: Optoelectronic materials used for blue light-emission today have low efficiencies, poor chromaticity and short operational lifetimes. Perovskites are attractive alternatives due to their exceptional optoelectronic properties; however, their impact is hindered due to the presence of lead. I report the use of Density Functional Theory (DFT) and machine learning to accelerate the discovery of new metal-free perovskites, where lead ions are replaced with organic molecules. I found acceptable errors in predicting bandgaps when using the optimized structures for the perovskites obtained using DFT. Since only 40% of the total DFT time in calculating bandgaps is consumed in determining these structures, this machine learning approach is 2.3 times faster in calculating bandgaps compared to traditional methods. Specifically, the proposed graph convolution neural network predicts the bandgaps with an error of 0.275 eV. This strategy enables the screening of wide chemical spaces of organic molecules for new metal-free perovskites.
Source: TSpace (University of Toronto)
Author: Aravind Bathini Ramaraju
Language: English
Subfield: Materials Chemistry
Field: Materials Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3153270240