Date of Award

August 2021

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Chemical Engineering

Committee Member

Rachel RG Getman

Committee Member

Leah LC Cassabianca

Committee Member

Mark MR Roberts

Abstract

Discovering optimal materials for a given application has become extremely difficult due to the vast scope of structure possibilities and requires substantial computational expense to evaluate even one of all possible structures. An increase in the structural complexity can offset combinatorial explosion and could take months, if not years, even with the advanced computational architecture, to screen every candidate.Here in this study, we propose a computational approach based on statistical learning combined with DFT based computations that can effectively screen over the entire range of possible candidates and predict the material’s electronic properties with high fidelity. Specifically, we use advanced machine learning algorithms such as gradient boosted decision trees and electronic structure bandgap calculation data to screen perovskite structures as alternatives for lead-free solar cells. Perovskites are compounds with chemical formula ABX3 where A and B are cations and X is an anion that bonds to both cations. The perovskite class offers compositional flexibility which allows us to tune the structure to obtain better solar absorption efficiency. Using machine learning, we could establish a structure-property correlation, by mapping the attributes of the structures to their bandgaps, which enabled us to screen over all compounds within the perovskite class thus drastically accelerating our search of the optimal lead-free alternatives for solar cells. With purpose of making the dataset more robust, this study explored the complexity of the composition effect by evaluating the substitution of one or more elements in different proportions and arrangements over the possible sites available in the base perovskite structure on the material's bandgap.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.