Rapid computational screening of materials for energy storage applications
File(s)
Author(s)
Onwuli, Anthony
Type
Thesis
Abstract
Energy storage systems are vital for enabling the distribution and storage of renewable energy and the development of electric vehicles. Solid-state electrolytes will be key to scaling up the development of safe batteries for this application. Computational materials design enables us to discover new materials at a lower cost and in a shorter timeframe than an Edisonian experimental synthesis approach.
In the first results chapter of this thesis (Chapter 4), the concept of quantifying the search space of stoichiometric garnets is presented using structure substitutions and a geometric descriptor for phase stability. Following this was a high-throughput density functional theory screening to determine the phase stability of these compounds. In Chapter 5, the high-throughput materials design workflow previously introduced is leveraged for the discovery of novel sodium halide electrolyte materials. A large compositional space of 3,710 quaternary materials was screened and the phase, electrochemical, and interfacial stability, as well as the Na-ion diffusivity, were analysed through a mixture of high-throughput DFT calculations and machine-learning molecular dynamic simulations. The work described in Chapter 6 delved more into materials informatics in which high-dimensional representations of the chemical elements were analysed to see if they reproduce chemical trends and show potential applications for structure substitution based on chemical similarity. To end this thesis, in the final results chapter (Chapter 7), the concept of machine learning representations of compounds was expanded upon by designing of ionic representations which were proven to enhance structure-agnostic learning for properties derived from the electronic structure such as the band gap.
Many of the results and ideas presented in this thesis may pave the way for the potential design of new materials, but more broadly aim to synthesise many of the methods for materials informatics and materials modelling and extend their application into both functional and novel applications.
In the first results chapter of this thesis (Chapter 4), the concept of quantifying the search space of stoichiometric garnets is presented using structure substitutions and a geometric descriptor for phase stability. Following this was a high-throughput density functional theory screening to determine the phase stability of these compounds. In Chapter 5, the high-throughput materials design workflow previously introduced is leveraged for the discovery of novel sodium halide electrolyte materials. A large compositional space of 3,710 quaternary materials was screened and the phase, electrochemical, and interfacial stability, as well as the Na-ion diffusivity, were analysed through a mixture of high-throughput DFT calculations and machine-learning molecular dynamic simulations. The work described in Chapter 6 delved more into materials informatics in which high-dimensional representations of the chemical elements were analysed to see if they reproduce chemical trends and show potential applications for structure substitution based on chemical similarity. To end this thesis, in the final results chapter (Chapter 7), the concept of machine learning representations of compounds was expanded upon by designing of ionic representations which were proven to enhance structure-agnostic learning for properties derived from the electronic structure such as the band gap.
Many of the results and ideas presented in this thesis may pave the way for the potential design of new materials, but more broadly aim to synthesise many of the methods for materials informatics and materials modelling and extend their application into both functional and novel applications.
Version
Open Access
Date Issued
2024-03
Date Awarded
2024-09
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Walsh, Aron
Butler, Keith
Sponsor
Engineering and Physical Sciences Research Council
Grant Number
EP/T51780X/1
Publisher Department
Materials
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)