Leveraging computational chemistry to model photochemical reactions and calculate experimental observables

Abstract

Predictive atmospheric models are essential to inform political and social decisions regarding
the behavior of chemicals and pollutants in the atmosphere. However, experimental studies
investigating the photochemistry of some key volatile organic compounds, which play a crucial
role in atmospheric chemical reactions, can be extremely challenging or even impossible to
conduct due to their short lifetimes. This is even more true when considering the photochemistry
of transient species whose reactivity is often inferred based on proxy molecules or simply
neglected in current atmospheric models. To address this issue, this thesis proposes a theoretical
protocol for calculating the photolysis rate constant, a critical experimental observable for
atmospheric models, in silico. Using the developed protocol, we demonstrated its effectiveness in
studying the photochemistry of one exponent of the α-hydroperoxycarbonyls and its implications
in atmospheric chemistry. Our research not only provided new insights into the significance
of studying the photochemistry of volatile organic compounds but also brought attention to
certain shortcomings of commonly used theoretical methods. Specifically, we uncovered a novel
flaw in the algebraic diagrammatic construction at the second order – a promising candidate
for studying volatile organic compounds. Finally, our developed protocol was utilized to explore
the photodissociation of Lewis adducts, with ammonia borane being the initial focus of our
investigation. Despite being an excellent material for hydrogen storage, its photochemistry had
not been studied before. Our investigation revealed new insights into the unique electronic
properties of the dative bond driving the photolysis of Lewis adducts. In particular, we offered
a comprehensive study of the mechanisms underlying the photodissociation of Lewis adducts,
paving the way for exciting and innovative applications of these molecules.

Date of Award	3 Oct 2023
Original language	English
Awarding Institution	University of Bristol
Supervisor	Natalie Fey (Supervisor) & Basile F E Curchod (Supervisor)