Chemistry is on the verge of a new era where the attention of chemists has shifted from covalent bonds to noncovalent interactions and their use as a predictable way to guide reactions pathways and product formation. Nature synthesizes elegant molecules under mild conditions and the designed syntheses have been demonstrated to be largely dependent on recognition, self-assembly and templating effects between molecular building blocks. Although covalent synthesis in fluidic medium via supramolecular control has been achieved with limited success, organic solid state has been of particular interest since it avoids solvent effects, and is able to provide unique materials with remarkable stereoselectivity under environment-friendly conditions. Although reactions in solids have resulted in a number of remarkable discoveries in chemistry and materials science, solid-state synthesis is generally not considered as a mainstream synthetic medium and solid-state reactions are seldom appreciated as an efficient way to access molecular targets. Owing to the limited number of solid-state reactions and the uncontrollable nature of crystal packing, solid state has not been utilized readily as a primary synthetic medium. In this context, reactions conducted in multicomponent molecular assemblies or co-crystals have been attracting much attraction in recent years as a general way of controlling the reactivity of molecules in solid state. A molecular component in the multicomponent molecular solid, acting as a linear template, has been shown to preorganize molecules in a modular way via intermolecular interactions and engineer their physical and/or chemical properties. The [2+2]photodimerization of olefins is a successful demonstration how templated solid state synthesis can efficiently synthesize complex targets that are synthetically challenging via conventional routes. In this dissertation, the generality and synthetic applicability of the templated synthetic approach in solid state will be described. How supramolecular interactions in molecular co-crystals precisely guide covalent bond formation in order to construct complex molecular targets will be demonstrated. Finally, co-crystallization will be shown as a general way to control optical properties in crystals.

The RNA World hypothesis suggests that RNA, or a proto-RNA, existed in an early form of life that had not yet developed the ability to synthesize protein enzymes. This hypothesis, by some interpretations, implies that nucleic acid polymers were the first polymers of life, and must have therefore spontaneously formed from simple molecular building blocks in the "prebiotic soup." Although prebiotic chemists have searched for decades for a process by which RNA can be made from plausible prebiotic reactions, numerous problems persist that stand in the way of a chemically-sound model for the spontaneous generation of an RNA World (e.g., strand-cyclization, heterogeneous backbones, non-selective ligation of activated nucleotides). The Molecular Midwife hypothesis, proposed by Hud and Anet in 2000, provides a possible solution to several problems associated with the assembly of the first nucleic acids. In this hypothesis, nucleic acid base pairs are assembled by small, planar molecules that resemble molecules which are known today to intercalate the base pairs of nucleic acid duplexes. Thus, the validity and merits of the Molecular Midwife hypothesis can be, to some extent, explored by studying the effects of intercalation on the non-covalent assembly of nucleic acids. In this thesis, I explore the role of the sugar-phosphate backbone in dictating the structure and thermodynamics of nucleic acid intercalation by using 2′,5′-linked RNA intercalation as a model system of non-natural nucleic acid intercalation. The solution structure of an intercalator-bound 2′,5′ RNA duplex reveals structural and thermodynamic aspects of intercalation that provide insight into the origin of the nearest-neighbor exclusion principle, a principle that is uniformly obeyed upon the intercalation of natural (i.e. 3′,5′-linked) RNA and DNA. I also demonstrate the ability of intercalator-mediated assembly to circumvent the strand-cyclization problem, a problem that otherwise greatly limits the polymerization of short oligonucleotides into long polymers. Together, the data presented in this thesis illustrate the important role that the nucleic acid backbone plays in governing the thermodynamics of intercalation, and provide support for the proposed role of intercalator-mediated assembly in the prebiotic formation of nucleic acids.

The RNA world hypothesis posits that at some stage in the development of life, RNA functioned as both an informational polymer and a catalyst for important reactions. However, many questions remain as to how RNA molecules might have evolved into living organisms. This thesis uses computer simulations to model processes thought to be important to the development of an RNA world. First, a model is discussed which describes non-enzymatic polymerization of single-stranded RNA from different kinds of activated nucleotides, a necessary first step towards an RNA world. It was found that a system undergoing polymerization of RNA from 5′-activated triphosphates or imidazolides behaves differently from an equilibrium system undergoing reversible polymerization reactions from 2′,3′-cyclic monophosphates, for example. In the 5′-triphosphate case, the system is not in equilibrium but rather in a state of circular reaction flux that must be maintained by an external source of phosphates. This model is then adapted to investigate non-enzymatic template-directed replication of RNA strands. It is found that this process fulfills all the necessary requirements to function as a metabolism which maintains a difference between the outside non-living environment and the internal environment of the cell. Finally, byproducts arising from the template copying mechanism in this model are discussed, including the development of highly regular sequence patterns in the strand population due to selection for the ability to form duplexes with neighbouring strands. Altogether, this thesis illustrates new implications, potential pitfalls, and possibilities of the RNA world hypothesis for the origin of life. In particular, it emphasizes the fundamental link between the processes of replication and metabolism, both of which must have been crucial to the functioning of the earliest protocells. This link has been largely overlooked in scientific literature on the topic to date. / Thesis / Master of Science (MSc) / For millennia, humanity has told stories about the origin of life. Since the 1960s, scientists have hypothesized that RNA is a key player in this origin story. RNA can both hold information and catalyze chemical reactions, meaning only one molecule is needed for both these crucial functions. However, many questions remain about how this would work in practice. This project used computer simulations to model steps along the path from RNA to living organisms. First, a model was developed for the formation of single-stranded RNA from building block molecules. The model was then expanded to include copying of existing RNA strands, and it was found that this process constitutes a metabolism. Finally, it was discovered that over time the copying process produces simple patterns in the sequence of building blocks that make up the RNA strands. Altogether, these findings emphasize the link between replication and metabolism in early cells.

In the past 30 years, the discovery of capabilities of nucleic acids far beyond their well-known information-bearing capacity has profoundly influenced our understanding of these polymers. The discovery by the Cech and Altman labs that nucleic acids could perform catalytic functions, coupled with the Gold and Szostak groups’ demonstration of the de novo evolution of nucleic acids that bind arbitrary ligands, has resulted in a proliferation of newfound roles for these molecules. Nucleic acids have found utility in both engineered systems, such as aptamer therapeutics, as well as in newly appreciated roles in extant organisms, such as riboswitches. As a result of these discoveries, many have pondered the potential importance of the dual (catalytic and informational) roles of nucleic acids in early evolution. A high-yielding synthetic route for the nonenzymatic polymerization of nucleic acids, based on the aqueous self-assembly of their components, would provide a powerful tool in nucleic acid chemistry, with potential utility in prebiotic and contemporary nucleic acid systems alike – however, such a route remains elusive. In this thesis, I describe several steps towards such a synthetic route. In these systems, a nucleic-acid binding ligand drives the assembly of short DNA and RNA duplexes, promoting the production of long nucleic acid polymers, while suppressing the production of short, cyclic species. Additionally, the use of a reversible covalent linkage allows for the production of long polymers, as well as the incorporation of previously cyclized products into these polymers. I also report several explorations of novel base pairings, nucleic acid-ligand interactions, and nucleic acid-ion interactions that have informed our studies of self-assembling nucleic acid systems.

This thesis describes supramolecular approaches to porphyrin nanorings. Cyclic porphyrin arrays resemble natural light harvesting systems, and it is of interest to probe the photophysical effects of bending the porphyrin aromatic π-system. A general overview of the synthesis and photophysical properties of porphyrins and their arrays is carried out in Chapter 1. The electronic structure of porphyrins is examined, and how conformational effects in oligomers, such as inter-porphyrin torsional angle and backbone bending influence the π-conjugation pathway. The structures of light harvesting complexes are discussed. Chapter 2 describes the design and synthesis of a complementary 12-armed template designed to coordinate linear porphyrin oligomers in the correct conformation for cyclisation to give a cyclic porphyrin dodecamer. Chapter 3 demonstrates two approaches to a cyclic porphyrin dodecamer ring. Firstly, a classical templating approach using the 12-armed template is described. The limitations of this approach in the quest for larger nanorings are discussed. Vernier templating, which utilises a mismatch in the number of binding sites between a ligand and its receptor is introduced as a general strategy to the synthesis of large nanorings. This is demonstrated by the synthesis of cyclic dodecamer from a linear porphyrin tetramer and a hexadentate template via a figure-of-eight intermediate. The general utility of the Vernier method to large nanorings is explored in Chapter 4 with steps towards the synthesis of a cyclic tetracosamer, consisting of 24 porphyrin subunits. In preliminary experiments, an improved route to the cyclic porphyrin octamer is described. Finally, the photophysical properties of the nanoring series are explored in Chapter 5 as a function of size and conformation. Femtosecond photoluminescence spectroscopy shows that even in cyclic dodecamer, exciton delocalisation over the entire porphyrin backbone occurs on a sub-picosecond timescale, and parallels are drawn with the dynamics of natural light harvesting complexes.