Computation-Driven Mechanistic Investigation and Data-Driven Discovery of Transition Metal-Based Catalysts

Date of Award

Fall 1-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Chemistry

First Advisor

Batista, Victor

Abstract

Catalysis is central to chemical synthesis, energy conversion, and environmental remediation, yet many challenges remains. This thesis presents three complementary theoretical studies that address this gap through mechanistic modeling and predictive tools synergizing DFT and data-driven tools. The first chapter examines CuPd nanoalloys for selective nitrite reduction, revealing how alloy composition and surface chemistry dictate divergent reaction pathways. These insights clarify the origins of high selectivity toward ammonia formation and guide rational catalyst design for environmental applications. The second chapter explores hydrogenation of internal alkynes on Pt and Pd catalysts using amine as hydrogen donors. This mechanistic framework explains the observed cis-selectivity and informs the design of isomer-selective hydrogen donor systems. It also uncovers a unique isomerization step exclusive to Pt surfaces and demonstrates how different amines modulate stereoselectivities and rates. The final chapter introduces a machine learning pipeline for predicting reaction outcomes in transition metal catalysis. By combining DFT-derived descriptors with experimental data, the model achieves accurate predictions across phosphine-ligated Ir, Cu, and Pd systems, illustrating how physics-based features can enable data-efficient catalyst optimization. Together, these advancements highlight the capability of integrating computational and experimental insights, which can foster efficient, physics-informed catalyst development for a multitude of applications.

This document is currently not available here.

Share

COinS