Back to search

Cornell University

Improving the Accuracy, Transferability, and Efficiency of Hybrid Density Functional Theory

Abstract

dc:description.abstract

Computing the ground state energy for a given molecular or condensed phase system is important for the understanding and theoretical prediction of chemical reaction thermodynamics and kinetics as well as molecular and materials properties. These tasks have far-reaching applications within chemistry, physics, and materials science alike. Consequently, the last several decades have seen the development of numerous so-called electronic structure theories to address this problem. Among these, density functional theory (DFT) has skyrocketed in popularity due to its favorable trade-off between accuracy and computational cost, resulting in hundreds of density functional approximations (DFAs), most of which can be classified into empirical/non-empirical approximations used primarily for molecular/condensed phase systems, respectively. In this thesis, I will discuss the development of both DFAs that bridge this historical gap and algorithms to enable their application to large-scale condensed phase systems. In Chapter 1, I use DFT to develop a rational design procedure for tuning the mechanical properties of metallopolymers by altering metal-ligand interactions. In Chapter 2, I design and construct a large benchmark database of short-range non-covalent interactions for the testing and development of density functional approximations. In Chapter 3, I design and construct a large benchmark database of conformational energies for the testing and development of density functional approximations. In Chapter 4, I describe a general framework that unites constraint-satisfying and data-driven strategies for the development of new DFAs with greater accuracy and transferability. I then extend this framework to the direct optimization of DFAs on data that cannot be easily described as simple energy differences (i.e. molecular geometries and electron densities). Finally, in Chapter 5, I develop a linear-scaling, nearly black-box algorithm for the evaluation of the exact exchange interaction of hybrid DFAs in large-scale (possibly heterogeneous) finite-gap condensed-phase systems. Taken in concert, these chapters argue for the development of benchmark quantum chemical databases for the construction of data-driven constraint-satisfying hybrid DFAs with extended applicability for molecular and condensed phase systems alike.

Degree

thesis:*
Name thesis:degree_name
Ph. D., Chemistry and Chemical Biology
Level thesis:degree_level
Doctor of Philosophy
Discipline thesis:degree_discipline
Chemistry and Chemical Biology
Grantor
Cornell University
Year dc:date.issued
2023

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Sparrow, Zachary
Chair dc:contributor.chair
  • DiStasio, Robert
Committee members dc:contributor.committeemember
  • Hines, Melissa
  • Loring, Roger

Subjects

dc:subject × 6

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Bibid
16176682
Dc Identifier Other
ProQuest Submission ID: 13596
ProQuest Publication ID: 30426910

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Sparrow, Zachary. Improving the Accuracy, Transferability, and Efficiency of Hybrid Density Functional Theory. Doctor of Philosophy thesis, Cornell University, 2023. https://hdl.handle.net/1813/114151