Date of Completion

5-12-2016

Embargo Period

6-1-2017

Keywords

Quantum mechanics, machine learning, surface chemistry, atomistic modeling, thermodynamics, kinetics

Major Advisor

Rampi Ramprasad

Associate Advisor

Ranjan Srivastava

Associate Advisor

Avinash M. Dongare

Field of Study

Chemical Engineering

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

Surface chemistry is a phenomenon manifesting itself in several key areas; catalysis, materials fabrication, and emissions mitigation, to name a few. At the present time, atomistic computational driven efforts to study such processes are dominated by models based on quantum mechanics. Their flexibility in studying diverse chemistries, along with the ability to predict accurate thermodynamic and kinetic insights of surface processes, makes them increasingly popular. From ultra-low temperature and pressure to normal operating conditions these methods are now commonly utilized. Nevertheless, the computational burden inherent in the method renders it insufficient to keep up with the current need for quick discovery, i.e. predicting properties of millions of permutations of materials or the meticulous analysis of a chemical reaction on a material. Consequently, a push to go beyond traditional design and characterization practices to explain materials chemistry is becoming necessary.

In this thesis, a new framework that combines quantum mechanics with data-driven machine learning methods is put forth. The premise of such an approach is to mine and find patterns within data and in doing so come up with human fathomable relationships, to help accelerate discovery. Here, I focus on model development, which begins by generating data, identifying descriptors for a process, learning from the data and culminating with model validation. This then enables accelerated estimation of thermodynamic and kinetic properties of surface processes. Two detailed examples of this hybrid approach are discussed; (i) a guided and targeted catalyst design framework to identify optimal dopants to enhance thermochemical dissociation of H2O, and (ii) a force predictive framework (commonly known as force field) to rapidly compute forces on atoms, so as to extend dynamic simulations to length and time scales beyond current quantum mechanical methods.

Share

COinS