The maximum theoretical energy efficiency of fuel cells is much larger than those of the steam-power-turbine cycles that are currently used for generating electrical power. Similarly, direct hydrocarbon fuel cells, DHFCs, can theoretically be much more efficient than hydrogen fuel cells. Unfortunately the current densities (overall reaction rates) of DHFCs are substantially smaller than those of hydrogen fuel cells. The problem is that the exchange current density (catalytic reaction rate) is orders of magnitude smaller for DHFCs. Other work at the University of Ottawa has been directed toward the development of polymer electrolytes for DHFCs that operate above the boiling point of water, making corrosion rates much slower so that precious metal catalysts are not required. Propane (liquefied petroleum gas, LPG) was the hydrocarbon chosen for this research partly because infrastructure for its transportation and storage in rural areas already exists. In this work nickel based catalysts, an inexpensive replacement for the platinum based catalysts used in conventional fuel cells, were examined using density functional theory, DFT. The heats of propane adsorption for 3d metals, when plotted as a function of the number of 3d electrons in the metal atom, had the shape of a volcano plot, with the value for nickel being the peak value of the volcano plot. Also the C-H bond of the central carbon atom was longer for propane adsorbed on nickel than when adsorbed on any of the other metals, suggesting that the species adsorbed on nickel was less likely to desorb than those on other metals. The selectivity of the propyl radical reaction was examined. It was found that propyl radicals

Hydrogen bonds are of great interest in the solid state due to their importance in structural, functional and dynamical properties of chemical systems. Moderate hydrogen bonds have been linked with proton transfer, whereas the short, strong hydrogen bonds enable proton migration. Previous work in our group on relatively simple hydrogen bonded adducts relied on the combination of ab initio computational modelling (molecular dynamics) with variable temperature diffraction results (X-ray and neutron). These demonstrated that the interplay of these techniques was successful in studying the phenomena of proton transfer and migration. The present work follows on from that, and focuses on the effects of temperature and pressure on proton transfer and migration using both experimental and computational methods. The systems studied continue to encapsulate adducts with N…O and O…O hydrogen bonds. The study of the adduct formed between squaric acid and 4,4’-bipyridine was found to exhibit proton transfer associated with a single-crystal to single-crystal phase transition at 450 K that is coupled to a colour change (yellow to red). X-ray and neutron diffraction initially revealed the heavy atom structure and secondly the location of the hydrogen atoms along the moderate N…O hydrogen bond (ca. 2.6 Å). Computational modelling supported this and deduced the reason for the striking colour change. Pressure studies also determined that the adduct underwent two phase-transitions with a similar colour change, indicating that proton transfer is also a factor here, but with powder patterns different from the high temperature form, indicating that further polymorphs for this interesting system must exist. In an attempt to lower the temperature at which proton transfer would occur the base was changed to one of a more basic nature, i.e. co-crystallisation of squaric acid and 2,2’-dimethyl–4,4’-bipyridine was pursued. This lead to the formation of two red crystals that were found to posses the base doubly protonated at all temperatures studied (from 300 K to 100 K). The adduct of N,N-dimethylurea with phosphoric acid was obtained from a systematic study designed to follow the success of a previously reported system that showed proton migration (the adduct of urea and phosphoric acid). The new material was found to crystallise as the 2:1 adduct and maintained the short, strong hydrogen bonds characteristic of the parent structure. As part of the systematic approach undertaken throughout the research presented here, co-crystallisation of a combination of acids and bases were attempted in order to synthesise new materials containing short, strong hydrogen bonds. These yielded the adducts between oxalic acid and 2,2’-dimethyl-4,4’-bipyridine, and oxamic acid and 4,4’-bipyridine. In addition to these adducts some compounds ended up reacting to create new ones, e.g. the fusing of dimethyl urea and squaric acid to give N-(2-hydroxycyclobutene-3,4-dione)-N’,N’-dimethylurea and N-(2- hydroxycyclobutene-3,4-dione)-N,N'-dimethylurea, while a new polymorphs of one of the precursors on its own was also obtained (N,N’-dimethylurea). The resulting co-crystallisations did not all follow the design quite as intended. They did, however, yield interesting new structures, some of which have the potential to be proton migration and transfer systems.

Chapter one is an introduction to the less common coordination and oxidation chemistry of palladium; complexes containing Pd-OR, Pd-NR2 and those in the oxidation states of +IV. An outline of PdII/IV catalysed ligand-directed oxidative functionalisation is also included. Chapter two covers the design and synthesis of a range of tethered N-heterocyclic carbene (NHC) complexes of Pd. In addition, the syntheses of a number of new tethered NHC ligands are described. The use of Density Functional Theory (DFT) to model the complexes in this thesis was explored. Chapter three describes the synthesis and characterisation of PdIV halide complexes. The relevance of these compounds to PdII/IV catalysed ligand-directed oxidative functionalisation is explored. DFT was used to probe the reaction pathway for N-bromosuccinimide and iodobenzene dichloride. Chapter four examines reactions with oxidants used to form C-O and C-C bonds. The reaction pathway for iodobenzene diacetate was investigated using DFT. Chapter five contains experimental details and characterising data for the compounds reported.

Negative Poisson’s ratio, or auxetic, materials present the possibility of designing structures and components with tailored or enhanced mechanical properties. This thesis explores the phenomenon of auxetic behaviour in cubic crystals using classical and quantum modelling techniques and assesses the validity of these techniques when predicting auxetic behaviour in cubic elemental metals. These techniques are then used to explore the mechanism of this behaviour. The findings of the atomistic modelling are then used as a template to create networks of bending beams with tailored Poisson’s ratio behaviour.

The structural and electronic properties of covalently and non-covalently functionalized graphene are investigated by means of first-principles density-functional-theory. The electronic characteristics of non-covalently functionalized graphene by a planar covalent organic framework (COF) are investigated. The aromatic central molecule of the COF acts as an electron donor while the linker of the COF acts as an electron acceptor. The concerted interaction of donor acceptor promotes the formation of planar COF networks on graphene. The distinctive electronic properties of covalently functionalized fluorinated epitaxial graphene are attributed to the polar covalent C–F bond. The partial ionic character of the C–F bond results in the hyperconjugation of C–F σ-bonds with an sp2 network of graphene. The implications of resonant-orbital-induced doping for the electronic and magnetic properties of fluorinated epitaxial graphene are discussed. Isolation of single-walled carbon nanotubes (SWNTs) with specific chirality and diameters is critical. Water-soluble poly [(m- phenyleneethynylene)- alt- (p- phenyleneethynylene)], 3, is found to exhibit high selectivity in dispersing SWNT (6,5). The polymer’s ability to sort out SWNT (6,5) appears to be related to the carbon–carbon triple bond, whose free rotation allows a unique assembly. We have also demonstrated the important role of dispersion forces on the structural and electronic stability of parallel displaced and Y-shaped benzene dimer conformations. Long-range dispersive forces play a significant role in determining the relative stability of benzene dimer. The effective dispersion of SWNT depends on the helical pitch length associated with the conformations of linkages as well as π-π stacking configurations. We have revisited the constructing schemes for a large family of stable hollow boron fullerenes with 80 + 8n (n = 0,2,3,...) atoms. In contrast to the hollow pentagon boron fullerenes the stable structures constitute 12 filled pentagons and 12 additional hollow hexagons. Based on results from density-functional calculations, an empirical rule for filled pentagons is proposed along with a revised electron counting scheme. We have also studied the relative stability of various boron fullerene structures and structural and electronic properties of B80 bucky ball and boron nanotubes. Our results reveal that the energy order of fullerenes strongly depends on the exchange-correlation functional employed in the calculation. A systematic study elucidates the importance of incorporating dispersion forces to account for the intricate interplay of two and three centered bonding in boron nanostructures.

First-principles calculations have been carried out to investigate structural and electronic properties of graphene on SiC and diamond substrates and for a study of doping of fluorographene with various surface adsorbates. New insight is given into the problem of the decoupling of the graphene layers from SiC substrates after epitaxial growth. Mechanisms of hydrogen penetration between graphene and SiC(0001) surface, and properties of hydrogen and fluorine intercalated structures have been studied. Energy barriers for diffusion of atomic and molecular hydrogen through the interface graphene layer with no defects and graphene layers containing Stone-Wales defect or two- and four-vacancy clusters have been calculated. It is argued that diffusion of hydrogen towards the SiC surface occurs through the hollow defects in the interface graphene layer. It is further shown that hydrogen easily migrates between the graphene layer and the SiC substrate and passivates the surface Si bonds, thus causing the graphene layer decoupling. According to the band structure calculations the graphene layer decoupled from the SiC(0001) surface by hydrogen intercalation is undoped, while that obtained by the fluorine intercalation is p-type doped. Further, structure and the electronic properties of single and double layer graphene on H-, OH-, and F- passivated (111) diamond surface have been studied. It is shown that graphene only weakly interacts with the underlying substrates and the linear dispersion of graphene pi-bands is preserved. For graphene on the hydrogenated diamond surfaces the charge transfer results in n-type doping of graphene layers and the splitting of conduction and valence bands in bilayer graphene. For the F- and OH-terminated surfaces, charge transfer and doping of graphene do not occur. Finally, the possibility of doping fluorographene by surface adsorbates have been investigated. The structure and electronic properties of fluorographene with adsorbed K, Li, Au atoms, and F4-TCNQ molecule are described. It is shown that adsorption of K or Li atoms results in electron doping of fluorographene, while Au atoms and F4-TCNQ introduce deep levels inside the band gap. The calculated value of the fluorographene work function is extremely high, 7.3 eV, suggesting that p-type doping is difficult to achieve.

Theoretical investigation of the structure and stability of neutral and cationic TixOy cluster series (where y =2x-1, 2x, 2x+1) have been performed. The lowest lying structures for the neutral clusters are usually found in the singlet state. Generally, in bulk and in the case of the neutral TixOy clusters, the 2x cluster series is relatively more abundant than the 2x-1 and the 2x+1 cluster series. But in the case of cationic TixOy clusters, the 2x-1 series is more abundant. To understand the origin of the stability of the TixO2x-1+ clusters, we use density functional theory within the NRLMOL set of codes. Different analyzing factors such as ionization potential, TiO2 removal energy, oxygen removal energy, binding energy per atom and HOMO-LUMO gap have been used to examine the relative stability of TixO2x-1+ clusters. After analyzing the above criteria, we find that the ionization potential and HOMO-LUMO gap are more reliable, as the low ionization potential of the 2x-1 series generally implies low HOMO-LUMO gap and suggests that the 2x-1 cluster series more likely prefer to remain as cations. To further confirm this, we examine the density of states of Ti3O5 and Ti3O6 which show a larger HOMO- HOMO-1 gap in case of Ti3O5, indicating that the cluster would like to lose an electron for enhancing electronic stability.

Despite over 100 years of scientific research, a full understanding of superconductivity remains elusive. While it is known that the electron-phonon interaction is responsible for the formation of Cooper pairs in conventional superconductors, many superconductors exhibit behaviour suggestive of more exotic pairing mechanisms. In this thesis, two novel superconducting materials are considered, monolayer transition metal dichalcogenide, MoS₂, and iron-based superconductor, LaFeAsO_1−xF_x. The former is ideal for the study of the electron-phonon interaction, as it not only has potential applications as an atomically thin transistor, but also displays a dome-shaped superconductive state as a function of electron doping. In the latter, the superconductive state emerges from a magnetic parent compound upon flourine doping. Its high critical temperature is thought to be enhanced by magnetic fluctuation rather than being purely phonon-mediated. By using novel first-principles techniques, the electron-phonon interaction in electron doped single-layer MoS₂ is investigated. The superconducting gap is calculated using the Migdal-Eliashberg theory, and by considering the electronic structure and lattice dynamics in this material, an explanation is provided for the experimentally observed doping-dependent critical temperature in this material. The origin of the doping-induced transition from a magnetic phase to a nonmagnetic phase in LaFeAsO_1−xF_x is determined. A new model to capture the effects of the fluorine dopants is developed, which has implications for the electron-phonon interaction in this material.

Doctor of Philosophy / Department of Chemistry / Christine M. Aikens / Small silver and gold clusters (less than 2 nm) display a discrete absorption spectrum characteristic of molecular systems whereas larger particles display a strong, broad absorption band in the visible. The latter feature is due to the surface plasmon resonance, which is commonly explained by the collective dipolar motion of free electrons across the particle, creating charged surface states. The evolution between molecular properties and plasmon is investigated. Time-dependent density functional theory (TDDFT) calculations are performed to study the absorption spectrum of cluster-size silver and gold nanorods. The absorption spectrum of these silver nanorods exhibits high-intensity longitudinal and transverse modes (along the long and short axis of the nanorod respectively), similar to the plasmons observed experimentally for larger nanoparticles. These plasmon modes result from a constructive addition of the dipole moments of nearly degenerate single-particle excitations. The number of single-particle transitions involved increases with increasing system size, due to the growing density of states available. Gold nanorods exhibit a broader absorption spectrum than their silver counterpart due to enhanced relativistic effects, affecting the onset of the longitudinal plasmon mode. The high-energy, high-intensity beta-peak of acenes also results from a constructive addition of single-particle transitions and I show that it can be assigned to a plasmon. I also show that the plasmon modes of both acenes and metallic nanoparticles can be described with a simple configuration interaction (CI) interpretation. The evolution between molecular absorption spectrum and plasmon is also investigated by computing the density of states of spherical thiolate-protected gold clusters using a charge-perturbed particle-in-a-sphere model. The electronic structure obtained with this model gives good qualitative agreement with DFT calculations at a fraction of the cost. The progressive increase of the density of states with particle size observed is in accordance with the appearance of a plasmon peak. The optical properties of nanoparticles can be tuned by varying their composition. Therefore, the optical behavior of the bimetallic Au[subscript](25-n)Ag[subscript]n(SH)[subscript]18[superscript]- cluster for different values of n using TDDFT is analyzed. A large blue shift of the HOMO-LUMO absorption peak is observed with increasing silver content, in accordance with experimental results.

Doctor of Philosophy / Department of Chemistry / Christine M. Aikens / Water oxidation can be considered as the ‘holy grail’ of renewable energy research, where water is split into constituent molecular hydrogen and oxygen. Hydrogen is a very efficient energy source that is both clean and sustainable. The byproduct of hydrogen combustion is water, which in turn can be reused as the source for hydrogen generation. Natural water splitting is observed during photosynthesis in the oxygen-evolving complex of photosystem II, which consists of a CaMn₄O₄ cubane core. Herein, we report in silico approaches to understand bottom up catalytic design of model transition metal oxide complexes for water splitting. We have employed density functional theory to investigate model ligand-free architectures of cobalt and manganese oxide dimer (Mn₂(μ-OH)(μ-O)(H₂O)₃(OH)₅, Mn₂(μ-OH)₂(H₂O)₄(OH)₄, Mn₂(μ-OH)₂(H₂O)₂(OH)₂(O(CH)₃O)₂, Co₂(μ-OH)₂(H₂O)₄(OH)₄ and cubane (Co₄O₄(H₂O)₈(OH)₄, Mn₄O₄(H₂O)[subscript]x(OH)[subscript]y x = 4-8, y = 8-4) complexes. The thermodynamically lowest energy pathway on the cobalt dimer catalyst proceeds through a nucleophilic attack of a solvent water molecule to a Co(V)-O radical moiety whereas the pathway on the cubane catalyst involves a geminal coupling of a Co(V)-O radical oxo group with bridging oxo sites. The lowest energy pathway for the fully saturated Mn₂O₄•6H₂O (Mn₂(μ-OH)(μ-O)(H₂O)₃(OH)₅) and Mn₂O₃•7H₂O (Mn₂(μ-OH)₂(H₂O)₄(OH)₄) complexes occur through a nucleophilic attack of a solvent water molecule to Mn(IV½)O and Mn(V)O oxo moieties respectively. Out of all the oxidation state configurations studied for the manganese cubane, we observed that Mn₄(IV IV IV IV), Mn₄(III IV IV IV), and Mn₄(III III IV V) configurations are thermodynamically viable for water oxidation. All three of these reaction pathways proceed via nucleophilic attack of solvent water molecule to the manganese oxo species. The highest thermodynamic energy step in manganese dimer and cubane complexes corresponds to the formation of the manganese oxo species, which is a significant feature that reoccurred in all these reaction pathways. We have also employed multireference and multiconfigurational calculations to investigate the Mn₂(μ-OH)₂(H₂O)₂(OH)₂(O(CH)₃O)₂ system. The presence of Mn(IV)O[superscript]• radical moieties has been observed in this catalytic pathway. These simplest models of cobalt and manganese with water-derived ligands are essential to understand microscopic properties that can be used as descriptors in designing future catalysts.